article
stringlengths 0
745k
| abstract
stringlengths 0
92.7k
|
|---|---|
recently new narrow scalar resonances @xmath0 and @xmath1 were discovered at babar @xcite , cleo @xcite and belle @xcite .
they are not expected to be standard hadrons because their masses are close to hundred mev smaller than the ones predicted in quark models for positive parity @xmath6 quark - antiquark mesons @xcite .
for instance the experimental candidate to the first quark - antiquark @xmath6 meson is the @xmath7 @xcite .
this is confirmed by lattice simulations @xcite for quark - antiquark mesons .
the existence of non - mesonic and non - hadronic multiquarks has been suggested form the onset of the quark model @xcite .
these positive parity @xmath6 resonances are interpreted by beveren and rupp @xcite as poles in the s matrix of the coupled channels of mesons and meson pairs .
they were also predicted by nowak , rho and zahed @xcite as chiral partners of the well known negative parity pseudoscalar @xmath8 and @xmath9 .
they are as well interpreted as k - d molecules by barnes , close and lipkin @xcite , similar to the deusons anticipated by tornqvist @xcite . in the same way i find here that the new @xmath6 resonances can be understood as tetraquarks , or equivalently as s - wave @xmath10-@xmath11 molecules bound by the short range attraction . the experimental discovery of these hadron also suggests the existence of a new class of multiquarks . in this paper
i study the family of all possible narrow tetraquark and pentaquark resonances where the quark @xmath12 , or the @xmath2 kaon play a crucial role .
the strangeness @xmath13 pentaquark @xmath14 is more difficult to bind and was recently addressed with the same techniques of this paper in reference @xcite .
a chiral invariant framework @xcite is used to compute microscopically , at the quark level , the masses of this new class of hadrons . the resulting mechanism which provides the binding of the @xmath2 tetraquarks and pentaquarks is equivalent to the short range attraction of hadrons . here
a standard quark model ( qm ) hamiltonian is assumed , @xmath15 where each quark or antiquark has a kinetic energy @xmath16 with a constituent quark mass , and the colour dependent two - body interaction @xmath17 includes the standard qm confining and hyperfine terms , @xmath18 \ . \label{potential}\ ] ] moreover the hamiltonian includes a quark - antiquark annihilation term @xmath19 which is the result of spontaneous chiral symmetry breaking .
this paper is organised in sections . in section
ii the qm is reviewed , together with the resonating group method ( rgm ) @xcite which is adequate to study multiquark states , where several quarks overlap .
the rgm , together with chiral symmetry , produces short range hadron - hadron potentials , which can be either repulsive ( hardr core repulsion ) or attractive . in section
iii a criterion is derived to discriminate which systems bind and which are unbound .
this criterion is applied to find , among the s - wave hadrons , the candidates to trap a kaon . in section
iv the binding energy is computed for the selected positive parity mesons @xmath20 and negative parity baryons @xmath21 . finally the results are presented and discussed in section v.
for the purpose of this paper the details of potential ( [ hamiltonian ] ) are unimportant , only its matrix elements matter .
the hadron spectrum constrains the hyperfine potential , @xmath22 the quark - antiquark annihilation potential @xmath19 is also constrained when the quark model produces spontaneous chiral symmetry breaking @xcite .
the annihilation potential @xmath23 is present in the @xmath24 salpeter equation , @xmath25 \left ( \begin{array}{c } \phi^+ \\
\phi^- \end{array } \right ) = m_\pi \left ( \begin{array}{c } \phi^+ \\
-\phi^- \end{array } \right ) \label{pion bs}\ ] ] where the @xmath24 is the only hadron with a large negative energy wave - function , @xmath26 . in eq .
( [ pion bs ] ) the annihilation potential @xmath23 cancels most of the kinetic energy and confining potential @xmath27 .
this is the reason why the pion has a very small mass . from the hadron spectrum and using eq .
( [ pion bs ] ) the matrix elements of the annihilation potential are determined , @xmath28 where this result is correct for the annihilation of @xmath29 or @xmath30 quarks . when a strange quark is present , the corresponding matrix element is smaller by a factor @xmath31 which is a power of the constituent quark mass ratio @xmath32 ( 200,80)(0,0 ) ( 20,5 ) ( 120,30)(0,0 ) ( 20,0 ) ( 100,100)(0,0 ) ( 0,20)(1,0)100 ( 50,5)@xmath33 ( 60,60)(-1,-4)10 ( 65,41)@xmath34 ( 10,60)(1,0)100 ( 60,70)@xmath35 ( 20,0 ) ( 100,100)(0,0 ) ( -15,18)@xmath36 ( 0,20 ) ( 110,18)@xmath37 ( 102,20 ) ( -5,58)@xmath38 ( 10,60 ) ( 120,58)@xmath39 ( 112,60 ) the qm of eq .
( [ hamiltonian ] ) reproduces the meson and baryon spectrum with quark and antiquark bound - states ( from the heavy quarkonium to the light pion mass ) .
moreover the rgm was first used in hadronic physics by ribeiro @xcite to show that in exotic hadron - hadron scattering , the quark - quark potential together with the pauli repulsion of quarks produces a repulsive short range interaction .
for instance this explains the @xmath40 hard core repulsion , preventing the collapse of nuclei .
this @xmath40 hard core repulsion is supposed to also occur in several hadron - hadron interactions , and this explains why many multiquarks systems are not stable .
however deus and ribeiro @xcite used the same rgm to show that , in non - exotic channels , the quark - antiquark annihilation could produce a short core attraction .
recently it was shown that in the particular case of the low energy @xmath41 system in the chiral limit , the short range attraction and repulsion exactly cancel @xcite , resulting in a adler zero and the weinberg theorem . addressing a tetraquark system with the @xmath41 quantum numbers
, it was shown that the qm also fully complies with the chiral symmetry , including the pcac theorems @xcite .
therefore the qm is adequate to address the new @xmath6 hadrons , which were predicted by nowak , rho and zahed in an effective chiral model . in this paper
the qm and the rgm are applied to @xmath2 multiquarks .
the rgm @xcite computes the effective multiquark energy using the matrix elements of the microscopic quark - quark interactions .
any multiquark state can be decomposed in combinations of simpler colour singlets , the baryons and mesons .
this can be illustrated with the colour structure of a multiquark with two quark - antiquark pairs , @xmath42 .
assuming that this tetraquark is a colour singlet , each quark - antiquark pair can either be a colour singlet or a colour octet .
nevertheless the octet - octet state can be described with an exchange operator and with the singlet - singlet state , @xmath43 and there is only one anti - symmetrised state , @xmath44 where @xmath45 is the quark and antiquark anti - symmetriser .
therefore the multiquarks can be described with an anti - symmetrised basis of baryons and mesons .
the multiquark may only be a bound state , or a narrow resonance , if these baryons and mesons are sufficiently attracted .
thus the problem of multiquark stability can be technically reduced to the problem of the binding of baryons or mesons .
the rgm produces both the energy of a multiquark state and the effective hadron - hadron interaction .
the energy of the multiquark is computed with the matrix elements of the hamiltonian ( [ hamiltonian ] ) .
the wave functions of quarks are arranged in anti - symmetrised overlaps of simple colour singlet hadrons . as an example , in fig .
[ rgm coordinates ] , a tetraquark system is arranged in a pair of mesons @xmath23 and @xmath46 .
once the internal energies @xmath47 and @xmath48 of the two hadronic clusters are accounted , @xmath49 where @xmath50 is the anti - symmetrizer , the remaining energy of the meson - baryon or meson - meson system is computed with the overlap of the inter - cluster microscopic potentials , @xmath51| \phi_a \phi_b \rangle \nonumber \\ & & / \langle \phi_b \ , \phi_a | ( 1+p_{ab})(1- p_{13 } ) | \phi_a \phi_b \rangle \ , \label{overlap kernel } \end{aligned}\ ] ] where @xmath52 stands for the exchange of particle @xmath53 with particle @xmath54 , see fig .
[ rgm overlaps ] .
in what concerns the annihilation potential , it only occurs in non - exotic channels .
then it is clear from eqs .
( [ sum rules ] ) that the annihilation potential provides an attractive ( negative ) overlap .
this confirms that the hard core can be attractive for non - exotic channels where annihilation occurs . in what concerns the quark - quark(antiquark ) potential , it may also contribute to exotic channels .
because the potential @xmath17 is assumed to be proportional to the colour dependent @xmath55 it is clear that it can only contribute together with an exchange interaction , which provides a color octet , see eq .
( [ octet ] ) .
moreover the spin independent part of the interaction vanishes .
for instance in the meson - meson overlap of eq .
( [ overlap kernel ] ) the overlap of @xmath56 essentially cancels with the overlap of @xmath57 . the only potential which may contribute is the hyperfine potential , proportional to @xmath58 . in the present case where the kaon is a spin singlet ,
the minus phase from colour is inverted by a minus phase from spin .
i find that the total colour and spin matrix element is a hyperfine splitting , @xmath59 the flavour trace is quite simple and the spatial integral provides a geometrical overlap .
all the corresponding factors are positive , an therefore the quark - quark interaction results in a repulsive interaction .
these results are independent of the particular quark model that one chooses to consider , providing it is chiral invariant
. therefore two opposite classes of diagrams exist .
the exchange diagrams produce a repulsive interaction , which turns out to be proportional to the hyperfine quark - quark(antiquark ) interaction .
the annihilation diagrams produce an attractive interaction , which turns out to be proportional to the spin - independent quark - quark(antiquark ) interaction .
i arrive at the attraction / repulsion criterion for the short range hadron - hadron interaction , + - _ whenever the two interacting hadrons have quarks ( or antiquarks ) with a common flavour , the repulsion is increased , + - when the two interacting hadrons have a quark and an antiquark with the same flavour , the attraction is enhanced _ .
this paper is dedicated to the class of resonances which can be understood as a s=-1 kaon @xmath60 or @xmath61 strongly trapped by a s - wave hadron .
the criterion shows that all hadrons with an antiquark @xmath62 or @xmath63 or with a a quark @xmath12 allow the exchange overlap @xmath64 , and this would certainly contribute to repulsion @xcite .
although systems with both a short range repulsion and a short range attraction may still bind @xcite , i specialize here in systems where clearly there is no hard core repulsion . in what concerns attraction
, a quark @xmath29 or @xmath30 is needed in the s - wave partner of the kaon , in order to produce annihilation .
therefore the isospin of the partner of the kaon needs to be close to the opposite isospin of the kaon .
this excludes the mesons @xmath65 .
moreover i also specialise in systems which may achieve a remarkable stability , where the kaon partner is a hadronic resonance with a very narrow width .
this not only essentially restricts the kaon partner to s - wave mesons and baryons , it also excludes the meson @xmath66 and the baryon @xmath67 because they are very wide , due to the decay with a pion production .
the pion is also excluded because it is too light to bind to the kaon , all that one may get is a very broad resonance , the kappa resonance @xcite , which has been recently confirmed by the scientific community . therefore the hadrons which are best candidates to strongly bind the kaons @xmath68 are the s - wave hadrons with flavor @xmath69 .
this is expected to result in the @xmath70 @xcite , the @xmath71 @xcite the @xmath72 @xcite , the @xmath73 and several other resonances .
a convenient approximation for the meson and baryon s - wave wave - functions is the harmonic oscillator goundstate wave - function , @xmath74 where , in the case of vanishing external momenta @xmath75 and @xmath76 , the momentum integral in eq .
( [ overlap kernel ] ) is simply @xmath77 .
the coordinates of the incoming @xmath78 functions are illustrated in fig .
[ rgm coordinates ] , while the coordinates of the outgoing @xmath79 have the quark 1 and 3 exchanged .
i summarise @xcite the effective potentials computed for the different channels , @xmath80 where the colour and spin factors contribute respectively with @xmath81 and @xmath82 , @xmath83 is of the order of 430 mev and the geometrical factor is @xmath84 .
the remaining factor is the flavour factor .
the parameter @xmath85 is the only one that is model dependent , and it will be determined with a fit of experimental binding energies .
the estimation of @xmath85 is an important by - product of this method because the hadronic size can not be estimated directly by the hadronic charge radius which is masked by the vector meson dominance .
( 100,150)(0,0 ) ( -70,0 ) ( 0,140)@xmath86 ( 30,40)@xmath87 to study binding one has to proceed to the finite momentum case . then the effective potentials in eq .
( [ zero p ] ) turn out to be multiplied by the gaussian separable factor , @xmath88 \int \frac{d^3 p'_\lambda}{(2 \pi)^3 } \exp\left[- \frac { { p'_\lambda}^2}{2 \beta^2}\right]$ ] . in the exotic channels with exchange diagrams only
, this result can be prooved @xcite , and moreover the new parameter @xmath89 .
this occurs because the overlaps decrease when the relative momentum of the hadrons @xmath23 and @xmath46 increases . in the non - exotic channels with annihilation diagrams
the present state of the art of the rgm does not allow a precise determination of the finite momentum overlap .
nevertheless , it is expected that eventually the overlap decreases due to the geometrical wave - function overlap in momentum space . here
the precise determination of @xmath90 does not affect the results .
i order to reduce the number of parameters , and for simplicity , in this paper i assume that @xmath91 . for other approaches that also lead to a separable potential , see for instance ref .
this parametrisation of the schrdinger equation in a separable potential , @xmath92 |\psi_\lambda>=0 \ , \end{aligned}\ ] ] enables the use of standard techniques @xcite to exactly compute the scattering @xmath93 matrix , @xmath94 the binding occurs when the @xmath93 matrix has a pole for a negative relative energy , @xmath95 in fig .
[ g0 of energy ] the function @xmath96 is plotted .
@xmath96 is real for a negative relative energy @xmath97 and is complex for positive @xmath97 .
the binding only occurs if , @xmath98 using fig . [ g0 of energy ]
it is possible to determine the parameter @xmath85 which reproduces the experimentally measured binding energy of the @xmath99 . with a binding energy of @xmath100 mev ,
the corresponding parameter @xmath85 is 285 mev .
this corresponds to a radius of @xmath101 fm for the nucleon .
i now compute the binding energies of the kaonic strongly bound molecules , or equivalently strange multiquarks .
these multiquarks are divided into two different families which are respectively coupled and decoupled to pionic channels .
the computations are straightforward and the results are displayed in tables [ missing couple channels ] and [ binding energies ] . .
this table summarises the parameters @xmath102 and binding energies @xmath46 ( in mev ) @xcite for the channels closed to pion decay .
the italic binding energy @xmath103 of the @xmath104 is fitted from experiment . [
cols="^,^,^,^,^,^,^",options="header " , ] i find that , in a microscopic and chiral invariant calculation , the new hadrons appear as tetraquarks , or as pentaquarks . in a macroscopic hadron - hadron interaction perspective , the resulting mechanism which provides the binding is the short range strong attraction of hadrons . technically is is not possible distinguish a microscopic multiquark from a macroscopic strongly bound s - wave molecule , because the hadrons totaly overlap .
the purely molecular pattern only appears in different cases , when there is some repulsion which produces the clustering of quarks , and this may happen in the pentaquark family of exotics @xcite .
nevertheless the molecular perspective is convenient to estimate the mass of the multiquark because the binding energies are not very large , and the mass is essentially the sum of the kaon plus hadron mass , with a relatively small binding energy . the chiral excitation perspective , present in the chiral soliton model @xcite or in the chiral lagrangian , @xcite may also be eventually equivalent to the multiquark , or to the two - hadron narrow bound molecule .
suppose that a flavour singlet quark - antiquark pair @xmath105 or @xmath106 is created in a given hadron @xmath107 .
when the resulting multiquark @xmath108 remains bound , we have a state with an opposite parity to the original @xmath107 .
the reversed parity occurs due to the intrinsic parity of fermions and anti - fermions .
moreover , because s - waves are expected to have the lowest mass , this is equivalent to adding a pseudoscalar meson to the original @xmath107 .
the finite masses of the kaon and of the pion prevent the existence of a large number of pseudoscalar mesons in the multiquark family of the @xmath0 . in this family
the mass shift is of the order of the kaon mass minus the binding energy minus the strange - light quark mass difference @xmath109 , and this results in a mass shift close to 350 mev .
the same mass shift is expected in the whole family of the @xmath0 .
thus the multiquarks @xmath108 studied in this paper can be regarded as the chiral partners of the original @xmath107 , in agreement with nowak , rho and zahed .
nevertheless the standard @xmath110 quark - antiquark @xmath6 meson remains a chiral partner of the @xmath111 one , where the mass shift is a result of the angular momentum and spin excitations . in the present framework both the quark - antiquark creation and the angular momentum excitation
chiral excitations can be understood .
the presented results only depend on the hadronic size @xmath85 and not on the details of the quark - quark interactions , because the assumptions are quite simple .
i also assumed that the inverse radius parameters @xmath85 and @xmath90 are identical for all channels , although small channel dependences are expected .
moreover i neglected the meson exchange interactions because they are expected to be smaller than the hard core interaction . in the same way i did not consider the s - channel coupling to a single meson or baryon .
more importantly , in the second family of multiquarks , the coupling to pionic channels was neglected although this coupling conserves isospin .
it would be particularly interesting to study these neglected effects in the second family studied in this paper , in order to check if they correct the computed binding which seems too strong .
because the research in this direction will be model dependent , this will be done elsewhere .
the short range attraction studied in this paper is quite general and can also be applied to the different candidates to multiquarks which have been discovered quite recently @xcite
. this will also be applied elsewhere .
i thank emilio ribeiro for discussions on the rgm , and george rupp for discussions on hadronic resonances .
i am very grateful to gonalo marques for discussions on the family of hadronic molecules , and on the algebraic computations of this paper .
00 b. aubert _ et al . _ [ babar collaboration ] , phys .
lett . * 90 * , 242001 ( 2003 ) [ arxiv : hep - ex/0304021 ] .
d. besson _ et al .
_ [ cleo collaboration ] , phys . rev .
d * 68 * , 032002 ( 2003 ) [ arxiv : hep - ex/0305100 ] .
p. krokovny _ et al .
_ [ belle collaboration ] , phys .
lett . *
91 * , 262002 ( 2003 ) [ arxiv : hep - ex/0308019 ] ; hep - ex/0307052 ; hep - ex/0307041 .
s. godfrey and n. isgur , phys .
d * 32 * , 189 ( 1985 ) ; s. godfrey and r. kokosi , phys .
d * 64 * , 1679 ( 1991 ) .
g. s. bali , phys .
d * 68 * , 071501 ( 2003 ) [ arxiv : hep - ph/0305209 ] .
review of particle physics , k. hagiwara et al .
, phys . rev . *
d 66 * , 010001 ( 2002 ) .
r. l. jaffe , phys . rev .
d * 15 * 281 ( 1977 ) .
d. strottman , phys .
d * 20 * 748 ( 1979 ) .
e. van beveren and g. rupp , phys.rev.lett.*91 * , 012003(2003 ) [ arxiv : hep - ph/0305035 ] .
nowak , m. rho and i. zahed , phys . rev.d * 48 * , 4370 ( 1993 ) hep - ph/9209272 .
w.a . bardeen and c. t. hill , phys . rev.d
* 49 * , 409 ( 1994 ) hep - ph/9304265 .
t. barnes , f. e. close and h. j. lipkin , phys .
d * 68 * , 054006 ( 2003 ) [ arxiv : hep - ph/0305025 ] .
n. a. tornqvist , phys .
lett . * 67 * , 556 ( 1991 ) ; n. a. tornqvist , arxiv : hep - ph/0308277 .
p. bicudo and g. marques , accepted for publication in phys .
d. , rapid communication [ arxiv : hep - ph/0308073 ] ; p. bicudo , arxiv : hep - ph/0403146 .
p. j. bicudo and j. e. ribeiro , phys .
d * 42 * , 1611 ( 1990 ) ; 1625 ( 1990 ) ; 1635 ( 1990 ) .
j. bicudo , phys .
c * 60 * , 035209 ( 1999 ) .
p. bicudo , s. cotanch , f. llanes - estrada , p. maris , j. e. ribeiro and a. szczepaniak , phys .
d * 65 * , 076008 ( 2002 ) [ arxiv : hep - ph/0112015 ] ; p. bicudo , phys .
c * 67 * , 035201 ( 2003 ) .
p. bicudo , m. faria , g. m. marques and j. e. ribeiro , arxiv : nucl - th/0106071 . j. wheeler , phys . rev . *
52 * , 1083 ( 1937 ) ; * 52 * , 1107 ( 1937 ) .
j. e. ribeiro , z. phys .
c * 5 * , 27 ( 1980 ) .
j. dias de deus and j. e. ribeiro phys .
d * 21 * , 1251 ( 1980 ) .
p. bicudo and j. e. ribeiro , z. phys .
c * 38 * , 453 ( 1988 ) ; p. bicudo , j. e. ribeiro and j. rodrigues , phys .
c * 52 * , 2144 ( 1995 )
. i. bender , h. g. dosch , h. j. pirner and h. g. kruse , nucl .
phys . a * 414 * , 359 ( 1984 ) .
t. barnes and e. s. swanson , phys . rev .
c * 49 * 1166 ( 1994 ) ; j. s. hyslop , r. a. arndt , l. d. roper and r. l. workman , phys .
d * 46 * 961 ( 1992 ) .
e. van beveren , c. dullemond , c. metzger , j. e. ribeiro , t.a .
rijken , and g. rupp , z. phys .
c 30 , 615 ( 1986 ) . j. e. ribeiro , phys .
d * 25 * , 2406 ( 1982 ) ; e. van beveren , zeit .
c * 17 * , 135 ( 1982 ) ; j. e. ribeiro , arxiv : hep - ph/0312035 .
f. cannata , j. dedonder and l. lesniak , z. phys .
a * 334 * , 457 ( 1989 ) .
d. jido , j. a. oller , e. oset , a. ramos and u. g. meissner , nucl .
a * 725 * , 181 ( 2003 ) [ arxiv : nucl - th/0303062 ] ; e. oset , a. ramos and c. bennhold , phys .
b * 527 * , 99 ( 2002 ) [ erratum - ibid .
b * 530 * , 260 ( 2002 ) ] [ arxiv : nucl - th/0109006 ] .
e. e. kolomeitsev and m. f. m. lutz , phys .
b * 582 * , 39 ( 2004 ) [ arxiv : hep - ph/0307133 ] ; m. f. m. lutz and e. e. kolomeitsev , nucl .
a * 730 * , 392 ( 2004 ) [ arxiv : nucl - th/0307039 ] ; c. garcia - recio , m. f. m. lutz and j. nieves , phys . lett .
b * 582 * , 49 ( 2004 ) [ arxiv : nucl - th/0305100 ] .
t. nakano _ et al .
_ [ leps collaboration ] , phys .
* 91 * , 012002 ( 2003 ) [ arxiv : hep - ex/0301020 ] .
v. v. barmin _ et al . _ [ diana collaboration ] , phys . atom .
nucl . * 66 * , 1715 ( 2003 ) [ yad .
* 66 * , 1763 ( 2003 ) ] [ arxiv : hep - ex/0304040 ] .
s. stepanyan _ et al . _ [ clas collaboration ] , phys .
lett . * 91 * , 252001 ( 2003 ) [ arxiv : hep - ex/0307018 ] .
j. barth _ et al .
_ [ saphir collaboration ] , arxiv : hep - ex/0307083 . c. alt _ et al . _ [ na49 collaboration ] , arxiv : hep - ex/0310014 . k. abe _ et al . _ [ belle collaboration ] , arxiv : hep - ex/0308029
. j. z. bai _ et al . _
[ bes collaboration ] , phys .
lett . * 91 * , 022001 ( 2003 ) [ arxiv : hep - ex/0303006 ] . | i study the @xmath0 and @xmath1 discovered at babar , cleo and belle , and find that they belong to a class of strange @xmath2 tetraquarks and pentaquarks , which is equivalent to the class of kaonic molecules bound by short range attraction . in this class of hadrons a kaon is strongly trapped by a s - wave meson or baryon . to describe this class of multiquarks the resonating group method
is applied to a standard quark model with chiral symmetry breaking , and the short range kaon - meson(baryon ) interactions are extracted .
a criterion is derived to classify the attractive channels .
i conclude that the mesons @xmath3 , and the baryons @xmath4 clearly belong to the new hadronic class of the @xmath0 and @xmath1 .
the hadrons @xmath5 possibly belong to a related family . |
one of the open questions in cosmology today is the state of the intergalactic medium ( igm ) . in the present popular cosmological models , a putative cosmological constant accounts for perhaps 2/3 of the energy density of the universe , and a similarly mysterious cold dark matter component accounts for 2/3 of the remainder . of order 10% of the energy density
is supposed to be accounted for by baryons @xcite as obtained from primordial nucleosynthesis calculation .
surprisingly , the vast majority of the baryons in the local universe is as of yet undetected .
the directly observed known components consisting of stars , as well as hot and cold gas in galaxies , groups and clusters , only account for a few percent of this baryon budget @xcite .
cluster of galaxies are known to possess a large gas fraction , of order @xmath1 @xcite .
this gas fraction , if representative of the universe at large , is sufficient to account for the majority of baryons in the universe @xcite .
since gravity acts on all particles equally , one expects this ratio within the virial radius of clusters to be representative of the universal fraction . in the standard bottom - up picture of structure formation , clusters are the most massive objects in the universe , and therefore the latest to form .
this suggests that the majority of baryons is in a diffuse intergalactic form , which forms the intracluster medium when clusters collapse .
if the baryons were in other forms , for example compact objects which formed early in the universe , it would be extremely difficult to understand how those baryons would be released back into diffuse form when clusters form .
@xcite showed that the igm must have been heated by sources other than the gravitational energy of collapse and its resulting shock waves .
its direct detection would be of significant interest to quantitatively understand the distribution and state of the igm , and thereby infer its thermal history .
unfortunately , hot diffuse gas is difficult to detect in emission .
luckily , the cosmic microwave background ( cmb ) scatters off all free electrons through inverse compton scattering , allowing us to `` see '' the igm through scattering , in analogy to absorption .
this is known as the sunyaev - zeldovich effect @xcite , which is becoming routinely observable .
it becomes the dominant source of cmb fluctuations on arc - minute scales .
several experiments to conduct deep blank field or all sky search are under development .
the upcoming cmb experiments such as amiba ( @xcite ) , @xcite , @xcite , etc . and the current cbi ( @xcite ) are capable of detecting this angular scale . using multi - frequency information
, the sz fluctuation can be disentangled from the primary anisotropies at all angular scales @xcite .
on the theory part , various approaches have been explored to compute the sz angular power spectrum . amongst these ,
the press - schechter formalism is perhaps the most widely used due to its versatility and ease of implementation @xcite .
the authors above adopted the cluster gas model , while @xcite used simulations and @xcite used a simplified gas pressure bias model motivated from simulations to probe the statistics of the sz effect .
one of the shortcomings of the sz effect is its lack of redshift information .
since scattering is independent of redshift , the sz effect on one hand allows a direct probe of the igm to high redshift , on the other hand makes it challenging to disentangle the contributions arising from different redshifts . in this paper , we apply the hierarchical clustering model and non - linear perturbation theory to directly compute the sz effect .
this analytical method enables us to to extract distance information of intergalactic gas from the cross correlation of the sz effect with galaxy surveys and at the same time , check the results of both simulations and the press - schechter formalism . in section [ sec : ps ] , we develop our gas pressure model and in section [ sec : stat ] the sz statistics including power spectrum and bispectrum .
section [ sec : cross ] contains our results on the sz - galaxy cross correlation and the method to extract the redshift distribution of the sz effect .
the effect of non - gravitational heating and the method to extract it from overall gas pressure power spectrum are discussed in section [ sec : ngheating ] .
we discuss the potential inaccuracies arising from the approximations in section [ sec : discussion ] .
the paper concludes with section [ sec : conclusion ] .
the temperature distortion caused by the sz effect @xcite is : @xmath2 \equiv -2ys(x)\ ] ] where @xmath3 is the direction on the sky , @xmath4 , and the scattering function @xmath5 when @xmath6 ( rayleigh - jeans tail ) . in this limit
the sz effect results in an apparent cooling of the cmb background .
the `` @xmath7 '' parameter is defined as @xmath8 here , @xmath9 and @xmath10 are the temperature and number density of free electrons , respectively .
@xmath11 is the gas pressure . @xmath12 and
@xmath13 is the gas density weighted mean temperature .
@xmath14 is the proper distance , @xmath15 is the scale factor , @xmath16 is the comoving distance , @xmath17^{1/2}}\ ] ] describes the geometric effect of the curved universe , @xmath18 for open , flat and closed universes respectively , and @xmath19 is the curvature radius .
@xmath20 is the present cosmological matter density .
@xcite showed that the intergalactic medium has most likely been preheated by non - gravitational sources to @xmath21 kev per nucleon in order to be consistent with the observed upper bounds from the x - ray background .
we adopt the model of @xcite to express the gas distribution as a convolution of the matter distribution : @xmath22 this equation expresses gas as being less clumped than dark matter , partly due to the required preheating .
the effective radius in the top - hat window function , which is the gas heating radius , has the typical value @xmath23 mpc from the x - ray background constraint @xcite . for simplicity
, we adopt the gaussian window @xmath24 ( @xmath25mpc corresponds to @xmath26 mpc top hat window ) .
the window function for specific heating models can be obtained from hydrodynamic simulation by comparing the gas power spectrum to the dark matter power spectrum ( ma and pen , 2000 ) .
we also need to relate the gas temperature with the density .
first , we consider the gravitational heating . we adopt the cosmic energy theorem @xcite for the gas temperature model , which specifies the ratio between gravitational binding energy and total kinetic energy @xmath27 .
the pressure depends on the thermalized fraction of @xmath27 .
the translational kinetic energy is thermalized from the energy released when particles shell cross . a model of the thermalized energy
is thus given by the difference in energy between two particles separated by a non - linear scale in lagrangian space , which is the distance at which they can be expected to have shell crossed . the exact procedure amounts to solving the non - linear evolution equations directly .
but we can treat the effect statistically in a linear fashion . in the initial linear evolution ,
the gravitational potential remains constant .
after virialization , the gravitational energy at a fixed location remains almost constant . in an eulerian description , we can describe the energy of particles at a final virialized location as the energy released as a particle travels from its initial position to the final virialized location . while the initial position is not exactly known , we take a spherical average over the non - linear scale to average over all possible initial locations .
we thus have @xmath28 @xmath29 is the gravitational potential , @xmath30 .
@xmath31 ( hereafter call the `` electron window function '' ) is the potential averaged over the non - linear scale @xmath32 .
we choose a gaussian window function @xmath33 with the the non - linear scale @xmath32 , which for @xmath34 is @xmath35mpc .
hereafter we will adopt this value of @xmath32 .
@xmath36 is the mass fraction of the hydrogen in baryonic matter .
then , @xmath37/k^2 \equiv \delta(k ) f_e(k)$ ] . here , @xmath38 is the fourier transform of the electron window function . equation ( [ eqn : temp ] ) has some unphysical statistical properties .
the spatial average of the temperature , for example , is exactly zero , and for our model using gaussian random fields , it will be negative in half the volume .
but for purposes of modeling the sz effect , temperature is only observable when multiplied by density , and regions of positive temperature will have high density , while the negative temperature regions only contribute negligibly to the sz effect .
this model is not meant to be an exact description , but hopefully captures the statistical properties , while being simple and thus exactly solvable . in order to estimate the accuracy of these assumptions ,
we first compute the gas density weighted temperature and the mean @xmath7 parameter .
we define @xmath39 , the power spectrum @xmath40 and the variance @xmath41 .
we adopt the initial power spectrum @xcite @xmath42 ( here , @xmath43 is in unit of @xmath44/mpc and we choose the harrison - zeldovich - peebles scale invariant spectrum @xmath45 corresponding to @xmath46 ) , cluster - normalized density fluctuation at @xmath47mpc by @xmath48 [ @xmath49 for @xmath50cdm and @xmath51 for ocdm ] @xcite and the @xcite fitting formula to convert the linear @xmath52 to the nonlinear @xmath53 . then @xmath54 @xmath55 @xmath56 is the fourier transform of the gas window function in equation ( [ eqn : gasw ] ) , @xmath57 is the present baryonic matter density and @xmath58 is the mass fraction of baryonic matter in gaseous form . since cluster gas fractions @xcite are comparable to the baryon fraction obtained from big bang nucleosynthesis , we expect that @xmath59 .
we use @xmath60 , @xmath61 and the dimensionless hubble constant @xmath62 .
the results for scdm , ocdm and @xmath50cdm are shown in table [ tble : tbar ] and figure [ fig : temp ] . in the absence of non - gravitational heating effects , we find the present @xmath63 is around @xmath64 kev and the mean sz temperature distortion @xmath65k .
they are all consistent with the simulations and press - schechter formalism results @xcite .
@xmath66 & & & + @xmath67 & & & + @xmath68mpc ) & @xmath69&@xmath70&@xmath69 & @xmath70&@xmath69&@xmath70 + @xmath71/kev & 0.3 & 0.35 & 0.30 & 0.35 & 0.33 & 0.38 + @xmath72 & 1.4 & 1.7 & 2.9 & 3.5 & 2.3 & 2.6 + we define the pressure correlation function @xmath73 t(x)[1+\delta_g(x+r ) ] t(x+r)\rangle/\bar{t_g}^2 $ ] .. here , @xmath74 . these two definitions only differ by a constant @xmath75 and the resulting power spectrums only differ when @xmath76 by a dirac function .
we choose this definition because this is the most convenient way to deal with sz effect .
] then the pressure power spectrum @xmath77^{-2 } \times \nonumber \\ & & \left ( \frac{1}{(2 \pi)^6 } \int b_4({\bf k_1},{\bf k_2},{\bf k_3},{\bf k_4};z ) w_g(k_1 ) f_e(k_2,z ) w_g(k_3 ) f_e(k_4,z ) d^3 k_2 d^3 k_4 \right . \\ & & + \frac{1}{(2 \pi)^3 } \int w_g(k_1 ) p(k_1,z ) f_e(k_2,z ) p(k_2,z)\left[w_g(k_1)f_e(k_2,z)+w_g(k_2)f_e(k_1,z)\right]d^3 k_2 \nonumber\\ & & + 2 \frac{1}{(2 \pi)^3 } \int b_3(k_1,k_2,k;z ) w_g(k_1 ) f_e(k_2,z ) f_e(k , z ) d^3
k_2 \nonumber \\ & & + p(k , z ) f_e^2(k , z)\ ) \nonumber\end{aligned}\ ] ] statistical isotropy implies , @xmath78 .
the first term in the parentheses is from the non - gaussian ( connected ) part of the four - point correlation @xmath79 .
@xmath80 is the density polyspectrum .
the second term is from the gaussian part of @xmath79 .
this part also produces a term @xmath81 in @xmath82 , which only affects the spatially averaged quantities such as @xmath83 and will be included explicitly only when it is observable .
the third term is from the three - point correlation @xmath84 .
@xmath85 is the density bispectrum .
the last term is from the raw temperature correlation @xmath86 .
we adopt the hierarchical model @xcite which allows all higher order correlations to be expressed as the sum of products of two point correlations over all configurations , so @xmath87 @xmath88 is the two point power spectrum .
@xmath89 denotes different configurations .
coefficients @xmath90 generally depend on configurations . in the highly non - linear regime , they degenerate @xcite . because the sz effect is mostly contributed by the highly nonlinear regime ,
we adopt the saturated @xmath91 in the highly non - linear regime : @xmath92 and @xmath93 from hept ( the hyper - extended perturbation theory ) @xcite . here , @xmath94 is the linear power spectrum index and we choose the value of @xmath94 at @xmath95 . in this framework , @xmath96\end{aligned}\ ] ]
where @xmath97 and @xmath98 .
results are shown in figures [ fig : bias ] and [ fig : pressure ] .
calculations show that the non - gaussian term is dominant in the spectrum even when @xmath43 is small and this non - gaussian term itself is mostly contributed by the nonlinear regime .
besides the reason that the electron window function sweeps off most contribution from the linear regime , these behaviors reflect the strong correlation between @xmath99 and @xmath100 and the domination of the highly nonlinear ( non - gaussian ) regime .
this argues that the isotropic simplification for @xmath101 is at least self - consistent .
we define the gas pressure bias @xmath102 as @xmath103 .
the amplitude of @xmath102 varies only weakly with time .
this is an expected consequence of the hierarchical model , argued as followed .
the time dependence through the density power spectrum is basically canceled because eq .
( [ pressure ] ) shows that the dependence of @xmath104 over time is roughly @xmath105 .
the hierarchical model implies that roughly @xmath106
. then @xmath107 .
the only remaining time dependence arises from the evolution of the electron window function .
its effect is to eliminate the contribution from scales larger than several @xmath32 . since @xmath32 decreases with redshift , its evolution moves the peak of the bias to smaller scales , which should have higher bias .
the bias approaches a constant ( @xmath108 for all three cosmological models ) at large scale , but it drops quickly at very small scale ( due to the gas window function ) .
this result is consistent with @xcite .
one point worthy of mention is that , the behavior that @xmath102 approach a constant at large scale does not validate the constant bias model , which always produces a unit pressure - density cross correlation coefficient . but
as we will see in sec .
[ sec : cross ] , this is not the case . the pressure fluctuation variance peaks roughly at the scale of cluster core radii ( @xmath109 mpc ) , which is the direct result of the gas window function ( @xmath110 mpc ) .
the sz effect on cmb anisotropies is fully described by the @xmath94-point correlation functions between temperature distortion @xmath111 in different sky directions . the cmb anisotropy is decomposed into the harmonic components by : @xmath112
the cmb power spectrum is defined as @xmath113 .
the bispectrum is defined as @xmath114 .
they have been discussed by many authors as mentioned in the introduction section .
our model successfully reproduces these results .
in contrast to the press - schechter models , our model relies only on direct statistical correlators that have been proposed semi - analytically with only a few free dimensionless parameter , the @xmath115 saturation values , that have been verified in simulations . in the press - schechter model ,
the shape of the correlation function depends directly on the assumed structure of halos , which is a free function of the model and empirically measured in simulations . the radially averaged structure of halos , however , is not necessarily a direct predictor of correlations , unless one assumes halos to have no substructure .
it is reassuring to see that our approach , which originates from the diametrically opposite theoretical bases from press - schechter formalism , produces consistent results and agrees with simulations .
since we are interested in small angular scale where the line - of - sight cmb survey depth is much larger than any correlation length , we adopt limber s equation @xcite . in the rayleigh - jeans region
we get the sz power spectrum : @xmath116 where @xmath117 the results are shown in figure [ fig : cl ] and [ fig : zcon ] .
we find that ( 1 ) the sz effect exceeds the primary cmb anisotropy at @xmath118 .
( 2 ) the @xmath119 of scdm is much smaller than that of ocdm and @xmath50cdm .
this is due to the smaller @xmath67 and faster drop of the gas temperature in scdm with increasing redshift .
( 3 ) because @xmath120 ( @xmath121 in linear regime and @xmath122 in stable clustering regime ) and @xmath123 , we at once find that @xmath124 .
( 4 ) the peak contribution to @xmath119 depends on the angular scale , varying from @xmath125 at @xmath126 to @xmath127 for @xmath128 .
smaller angles probe the more distant universe .
these results are consistent with existing simulations and press - schechter formalism results .
the sz bispectrum is the projection of the pressure bispectrum @xmath129 defined by @xmath130 since @xmath131 , @xmath132 is related to 6-point density correlation function and in principle can be computed using our model .
but due to the complexity of the 6-point hierarchy and the absence of extensive numerical tests , we adopt the approach of @xcite and assume a simple pressure bias model where @xmath133 with @xmath134 .
the pressure bias model states that the pressure is a linear convolution over the density .
we note that this model builds in certain assumptions , which are invalid for treating cross - correlation issues address in section [ sec : cross ] .
it presumably provides a quick order of magnitude estimator , but fails on qualitative issues like galaxy - sz cross - correlation coefficients . in the bias model
the pressure bispectrum is given as @xmath135 the sz skewness parameter , @xmath136 is the easiest observable projection of the sz bispectrum , so we show its computation as an example . instead of working in multipole space , we derive following equations adopting the same approximation as limber s equation : @xmath137\ ] ] @xmath138\end{aligned}\ ] ] where @xmath139 and @xmath140 .
@xmath141 is the mean temperature sz distortion . @xmath142 and @xmath143 are from the corresponding gaussian term in the correlation function , as explained in eq .
( [ pressure ] ) . as a reminder
, @xmath144 is the geometric function defined in eq .
( [ eqn : cx ] ) . though the unsmoothed @xmath145 is not directly observable due to the absence of the window function , from the theory viewpoint , it represents the raw non - gaussianity properties .
our calculations find that @xmath146 ( table [ tble : tbar ] ) .
we defined the skewness parameter ( [ eqn : t3 ] ) as the ratio of differential measurements , which might result from interferometers in the raleigh - jeans regime .
for comparison , @xcite and @xcite considered the absolute temperature distortion and defined the skewness parameter as @xmath147 .
with multi - frequency information the absolute @xmath7 could be measured at each point in space , and that definition of skewness could in principle be observed .
the results are not too different : in our model , @xmath148 , which is similar to the result of @xcite with the maximum virialized mass @xmath149 .
further work considering window function filtering and the general calculation of the sz bispectrum can be done following the method of @xcite .
in the sz effect , the redshift distribution of intergalactic gas is lost . taking advantage of the cross correlations with other surveys having redshift information
, we may be able to statistically extract igm 3-d correlations and evolution . because the sz effect is mostly contributed by nonlinear structures at @xmath150
, it should have a strong cross correlation with galaxies at that redshift range .
the general idea is to use a galaxy survey with coarse redshift information , where the distance information only needs enough accuracy to resolve the time evolution of the correlation function , for example @xmath151 , which can be reached by photometric redshift surveys .
one then correlates each redshift bin with the sz map , and obtains the relative contribution of that sz slice to the total projected map .
several questions that pose themselves in this procedure are : how big a survey area and depth does one need ?
what fraction of the total sz fluctuations can be accounted using cross - correlations ?
are there optimal procedures of cross - correlating ?
since sz fluctuations are a non - linear function of the galaxy field , what fraction of the signal will ever be accounted for by this approach ?
do we expect the sz - galaxy cross - correlation coefficient to depend on redshift ?
how big an uncertainty might arise from a time and scale dependent galaxy bias ? in this section
we will address each of these questions using our model , and obtain quantitative estimates .
we assume a linear bias model for the galaxy number overdensity @xmath152 . the bias @xmath153 is taken to be a constant for simplicity .
then , the projected galaxies number overdensity can be expressed as @xmath154 @xmath155 is the selection function ( which we will want to vary _ a posteriori _ using the photometric redshift information ) , @xmath156 is the redshift survey depth and @xmath144 is the geometric function defined before .
the corresponding multipole moments are : @xmath157 where @xmath158 .
the multipole moments of sz - galaxy cross correlation are : @xmath159 with @xmath160 .
@xmath161 is the fourier transform of @xmath162 .
@xmath163 we have used @xmath164 .
it is useful to define the cross correlation coefficient @xmath165 in @xmath161 , the @xmath85 term is dominant . in @xmath80
there are 16 hierarchical terms and in @xmath85 there are three hierarchical terms .
different term dominates in different regions .
calculation shows that , roughly there are three regions : ( a ) @xmath166
. 4 terms ( @xmath167 ) dominate @xmath80 and 2 terms ( @xmath168 ) dominate @xmath85 .
@xmath169 ( @xmath170 ) is small , contribution to @xmath104 and @xmath171 are mostly from nonlinear regimes with @xmath172 . ] .
( b ) @xmath173 .
each hierarchical term in @xmath85 and @xmath80 has about the same contribution to @xmath161 and @xmath174 , respectively .
@xmath175 ( @xmath176 ) .
this region contributes most of the sz effect , as seen from figure [ fig : pressure ] . unless explicitly notified
, hereafter we will adopt the value of @xmath177 in this region .
we only show the result of this region in figure [ fig : r ] .
( c ) @xmath178 .
this is the opposite case to the case ( a ) .
@xmath179 ( @xmath170 ) .
no significant time dependence is found .
the corresponding cross correlation coefficients in multipole space are : @xmath180^{1/2 } } \label{eqn : corr}\ ] ] we now use the photometric redshift information from the galaxy survey to vary the selection function @xmath181 .
we pick the weighting which maximizes ( [ eqn : corr ] ) .
this allows the measurement of one number in the sz to yield a full function of redshift @xmath182 .
since ( [ eqn : corr ] ) depends on one variable @xmath183 , we can in principle measure an optimal redshift weighting @xmath155 at each @xmath183 .
the cross correlation variation has allowed us to measure a two dimensional cross - correlation function from a one dimensional observable in the sz and a two dimensional observable in the galaxies . the selection function to maximize @xmath184 is obtained from the variation @xmath185 .
we denote this selection function as @xmath186 and find that @xmath187 here , @xmath188 is a constant to be determined later by the observational data .
the optimized @xmath189 is : @xmath190^{1/2}= \left[\frac{\int_0^{z_g } \frac{\delta^4_{sz , g}}{\delta_g^2}c(x ) x(dx / dz)dz}{\int_0^{z_{cmb } } \delta_{sz}^2 c(x ) x(dx
/ dz)dz}\right]^{1/2}\ ] ] results are shown in figure [ fig : corr ] . following the same estimation as in @xmath191 , @xmath192 . the observed @xmath193 may be smaller because of the limited galaxy survey depth .
the observationalrocedure to extract the redshift information is as follows .
1 . start with a random guess for @xmath155 , e.g. @xmath194 .
2 . given a photometric galaxy survey and sz survey , measure @xmath195 , @xmath196 and @xmath197 from angular correlation functions @xmath198 , @xmath199 and @xmath200 , respectively .
3 . vary @xmath181 to maximize @xmath184 for a specific @xmath183 , therefore obtain @xmath186 .
4 . apply eq.([eqn : phim ] ) to infer @xmath201 , up to a constant @xmath202 , from the directly measured @xmath203 from the galaxy survey .
furthermore , eq . ( [ eqn : szg ] ) enables us to infer @xmath188 from the observed @xmath204 .
@xmath205 c(z ) x(z)dx / dz dz \right]/c_{l , obs}^{sz , g}\ ] ] 5 .
combine eq.([eqn : rkz ] ) and eq.([eqn : phim ] ) to obtain the time resolved sz power spectrum , up to a factor @xmath206 .
@xmath207\ ] ] in our model @xmath208 is almost a constant @xmath209 independent of cosmological models or redshifts ( see figure [ fig : r ] ) .
this gives a theoretical estimate of @xmath210 .
@xmath211 nomalizes the observed sz anisotropy multipole @xmath212 and the sz temperature variance .
( [ eqn : cl ] ) to integrate @xmath213 obtained above with @xmath214 and compare with @xmath212 , we will obtain the averaged @xmath211 .
@xmath215/\alpha^2(l)c(x)x(z)dx / dz dz \right]\ ] ] 7 .
follow the same steps to obtain @xmath201 and @xmath216 at different @xmath183 . since for @xmath195 at different angular scale @xmath183 , @xmath217
is determined roughly by @xmath218 ( @xmath219 is the redshift with peak contribution to @xmath119 .
[ fig : zcon ] ) , we can even get some idea about the scale dependence of @xmath191 .
thus , the total projected sz autocorrelation give a consistency check on the reconstructed time resolved power spectrum from the galaxy - sz cross correlation .
furthermore , the galaxy bias and its time dependence have completely dropped out of the calculation , and are thus not expected to affect the results at all .
then , in principle , sz - galaxy correlation plus sz cmb anisotropy provide a consistent and powerful method to extract all time evolution information of the igm pressure power spectrum .
noise and cosmic variance put constraints on the feasibility of our procedure .
( 1 ) limitation of cmb resolution degrades our method .
the measured range of @xmath43 is @xmath220\sim[1_1/3000z , l_2/3000z ] h$]/mpc . here , @xmath221 $ ] is the range of the cmb experiment . in order to detect the peak of @xmath222 ( around @xmath223/mpc as shown in fig .
[ fig : pressure ] ) , @xmath224 . for cbi ( @xmath225 )
, we are only able to detect @xmath226 .
amiba will measure @xmath227 and @xcite will measure @xmath228 .
they will allow us to measure the gas power spectrum up to @xmath229 and @xmath230 , respectively .
( 2 ) observational errors impose further constraints .
suppose that the galaxy survey covers a fraction @xmath231 of the sky and the @xmath232-th survey region ( for example , if the redshift accuracy is @xmath233 , then we can divide the galaxies into redshift bins with @xmath234 , @xmath235 , etc . )
have @xmath236 observed galaxies .
the cmb observation covers a fraction @xmath237 of sky and the @xmath119 is averaged over the band @xmath238 $ ] .
then the galaxy number count causes the poisson error : @xmath239^{-1/2}\ ] ] ( 3 ) the cosmic variance of the @xmath119 also cause errors . recalling that @xmath240 and @xmath241
, we get : @xmath242 here , @xmath243 is the density dispersion over smoothing scale @xmath244mpc .
we already use the typical value of @xmath245 and @xmath246 . the corresponding error caused in @xmath247
is : @xmath248^{1/4}\ ] ] recalling that @xmath249 , requiring a @xmath250 accuracy on @xmath213 would impose that ( a ) @xmath251 .
each survey regions must be large enough in order to contain sufficient number of galaxies and must be small enough to ignore the evolution .
we may choose redshift bands of each survey region @xmath252 .
then the number of galaxies observed @xmath253 has to satisfy @xmath254 . for sdss ( @xcite ) , which covers one quarter of the sky and probes more than one million galaxies with photometric redshift up to @xmath255 ,
the requirement @xmath256 is easily satisfied .
the measurement of the intergalactic gas at redshift z requires the galaxy survey at least up to that redshift ( equation [ eqn : phim ] ) , so we need deeper galaxy survey in order to probe the gas beyond @xmath257 .
( b ) @xmath258 . for cmb experiments with relatively lower resolution
, larger sky coverage is required .
for example , though planck only measures @xmath259 , it covers the whole sky and therefore satisfies this condition . for those with much higher resolution such as amiba and submillimeter telescope
, the required sky coverage can be relaxed to the order of @xmath260 .
this variation method only depends on the assumption that the cross correlation coefficient @xmath191 is approximately a constant , which is the direct result of the hierarchical model and has only weak dependence on the gas model and cosmologies .
furthermore , the hierarchical model is strongly supported by the consistency of the cmb sz power spectrum dependence on @xmath67 and the behavior of the gas bias between our model and simulations .
since the averaged @xmath191 is measurable , our method does not rely much on the theoretical value of @xmath191 and thus observationally consistent .
pen ( 1999 ) has shown that the igm has most likely been preheated by non - gravitational energy sources with energy injection @xmath261 kev per nucleon .
this section is devoted to consider this effect . because the relation between the gravitational heating and the non - gravitational heating is very uncertain , we only consider two extreme cases .
the first one is that the non - gravitational heating is perfectly correlated with the gravitational heating , then we can change eq . (
[ eqn : temp ] ) to : @xmath262 we will append all former results from gravitational heating with a superscript a ( adiabatic ) . here , @xmath263 represents the ratio of the non - gravitational heating and the gravitational heating . @xmath264 .
@xmath261 kev corresponds to @xmath265 .
all former results are not affected by this change except @xmath63 , @xmath266 , @xmath267 and @xmath268 due to the dependence @xmath269 .
figure [ fig : temp ] and [ fig : cl ] need to be changed correspondingly .
figure [ fig : zcon ] and [ fig : corr ] will change only when @xmath270 is time - dependent .
all other figures remain the same .
the second case is that non - gravitational heating is uncorrelated with local density , then , eq .
( [ eqn : temp ] ) changes to : @xmath271 @xmath272 \bar{t_g}^a$ ] .
then , the corresponding new results following the same definitions are : @xmath273/(1+\beta)^2 \nonumber \\ & = & \frac{(b_p^a)^2 + 2 \beta r^{a}b_p^{a}+\beta^2}{(1+\beta)^2 } \times p(k , z ) \ ( { \rm when } \ k\neq 0)\end{aligned}\ ] ] @xmath274=b\times\frac{r^{a}b^{a}_p+\beta}{1+\beta } p(k , z)\ ] ] @xmath275 @xmath276 here , @xmath277 is the fourier transform of @xmath278 .
the expression of the optimal selection function @xmath247 ( eq . [ eqn : phim ] ) is not affected at all .
equation ( [ eqn : newr ] ) tells us that @xmath279 . as expected , when @xmath280 , we reduce to our former results and when @xmath281 , we obtain the bias model .
there are three regions : ( a ) @xmath282/mpc .
we always have @xmath281 , @xmath283 and @xmath284 .
our redshift extraction method works well with @xmath285 in this region .
because this region is dominated by non - gravitational heating , high resolution cmb experiments are able to measure @xmath213 in this region and enable us to directly obtain the information of the non - gravitational heating .
since @xmath102 does not approach zero as before , @xmath286 increases with increasing @xmath43 .
this behavior will move the peak of sz power spectrum to larger @xmath183 while increasing the amplitude of sz power spectrum .
( b ) @xmath287/mpc .
then , @xmath288 , so @xmath289 will decrease roughly by a factor @xmath290 . the resulting @xmath177 is bigger than our former results , but it still remains roughly a constant with respective to space and time , so our redshift extraction method works well .
( c ) the intermediate region ( the range @xmath291 , roughly @xmath292 . ) , @xmath177 varies from @xmath293 to 1 .
our redshift deprojection method still works , but with a larger error .
the discussion of the observational requirement is not affected at all .
the new @xmath294 should differ only a little bit from the former result due to its dimentionless definition .
roughly , @xmath295 , bispectrum @xmath296 and skewness @xmath297 .
our knowledge on the non - gravitational heating is very limited , and presumably depends on the poorly understood physics of star formation and supernovae dynamics .
it is not even known if the heating was pre or post structure formation .
we have surveyed two simple but different models of non - gravitational heating to demonstrate the range of effects it may have on the correlations .
the above equations show the basic procedure to include the non - gravitational heating and the possible effects .
so we do not plan to constrain to a highly hypothetical model and do the calculation .
for the purpose of estimation , we may either assume a step function for @xmath298 ( @xmath299 for @xmath300 .
otherwise , @xmath301 . ) or @xmath302 . the upper limit of @xmath7 parameter @xmath303 from cobe / firas @xcite put constraints on the value of @xmath304 and @xmath305 . for example ,
when @xmath306 kev , @xmath307 or @xmath308 .
further simplification can be made by taking the fitting formula of @xmath309 and @xmath310 .
we show some examples ( figure [ fig : newbias],[fig : newpp ] and [ fig : newr ] ) .
when combining cmb experiments with galaxy surveys , we have four observables : @xmath119 , @xmath311 , @xmath312 and @xmath313 . if @xmath191 is not a strong function of time and scale , we can solve for @xmath314 . if the behavior of @xmath315 as discussed above
is correct , we obtain the fifth observable @xmath316 from the small scale behavior of @xmath213 .
let us now address some potential systematic shortcomings in our simplified model .
( 1 ) our model relies on the adopted gas - dark matter density relation ( eq .
[ eqn : gasw ] ) , the ( non - local ) temperature - density relation ( eq .
[ eqn : temp ] ) and the non - gravitational heating . in case
that only the gravitational heating is included , apart from intrinsic problems addressed in section [ sec : ps ] , our model contains two free parameters : the gas - dark matter smoothing length , and the temperature - potential smoothing length .
these parameters only weakly affect the pressure power spectrum and pressure bias , which are defined in terms of dimensionless functions .
the normalization is expressed in terms of the mass weighted temperature , @xmath63 .
we see in table [ tble : tbar ] that the gas smoothing only has a small effect .
a preheating of 1 kev as proposed by @xcite has a factor of order unity impact on the temperature , and about factor of 4 effect on the cmb anisotropy spectrum and 8 on the bispectrum , while the skewness will decrease by about 50% .
we provide a convenient method to include such effects .
hydro simulations are able to measure the gas - dm relation and the @xmath317 relation and will further improve our work . fortunately , the main goal of our model the extraction of the 3-d gas information is least affected , since it depends only on the cross - correlation coefficient ( [ eqn : rkz ] ) and figure [ fig : r ] , which does not vary a lot even including non - gravitational heating . ( 2 ) the galaxy bias model we adopt may be too simple . in our redshift extraction method of the gas pressure power spectrum , the only bias dependency comes from @xmath191 . we speculate that , even for a realistic galaxy bias model , @xmath191 should be close to a constant , which still enable us to extract @xmath213 .
( 3 ) @xmath318 .
the value of @xmath318 in hept is a function of the power index of a power law power spectrum .
we have extended it to the cdm power spectrum and choose the power index at @xmath319 .
because sz effect is mainly contributed by the non - linear regime with @xmath320 , hept is applicable and the resulting @xmath101 does not vary a lot .
more importantly , @xmath321 only has a weak dependence on @xmath94 .
this behavior ensures that @xmath191 has least dependence on our assumptions about @xmath318 .
we have presented a new tool to compute power spectra and other statistics of sz fluctuations based on hierarchical clustering and scaling .
this approach describes the two point correlation of non - linear observables , such as the sz @xmath7 parameter in terms of two point correlation functions , which directly maps to the observed angular power spectrum .
we have demonstrated that this approach is feasible , and produces results consistent with simulations and press - schechter approaches .
we then addressed the problem of measuring the redshift evolution of the sz effect through cross correlation with galaxy surveys containing coarse photometric redshift information .
we have shown that a variational method allows the redshift deconvolution which does not depend on galaxy biasing . the only model dependent quantity is the cross correlation coefficient @xmath177 between galaxies and gas pressure , whose value has least dependence on the gas density model and the gas temperature model .
our quantitative estimates suggest that , when combining cmb experiments either with high resolution such as amiba and south pole submillimeter telescope or with high sky coverage such as planck and deep broad photometric galaxy surveys such as sloan , our method is able to extract the redshift evolution information of intergalactic gas , such as the full time resolved gas pressure power spectrum , even without requiring the knowledge of galaxy bias
. it is even capable of disentangling the contribution from the gravitational heating and those from the non - gravitational heating .
this method serves as a powerful probe to this primary component of cosmic baryons to high redshift .
the model also provides an alternative to simulations and the press - schechter formalism to calculate the cmb sz power spectrum and bispectrum .
we successfully reproduce , when no non - gravitational heating presents , the mass weighted gas temperature @xmath322 kev ) , the pressure power spectrum , the pressure bias ( @xmath323 , but scale dependent ) , the mean sz temperature distortion ( @xmath324 k ) , the sz power spectrum and the skewness parameter ( @xmath325 ) . with our transform formulas ,
non - gravitational heating is easily included and we have estimated its effects on the correlations .
amiba , 2003 , http://www.asiaa.sinica.edu.tw/research/amiba/amiba.html atrio - barandela , f. and mucket , j. , 1999 , , 515 , 465 bernardeau , f. , 1994 , , 291 , 697 cole , s. and kaiser , n. , 1988 , , 233 , 637 cooray , a. , hu , w. and tegmark , m. , 2000a , submitted to apj , astro - ph/0002238 cooray , a. , 2000b , submitted to prd , astro - ph/0005287 cbi , http://astro.caltech.edu/$^{\sim}$tjp/cbi/[http://astro.caltech.edu/$^\{\sim}$tjp/cbi/ ] danos , r. and pen , u.l .
, 1998 , astro - ph/9803058 davis , m. , et al .
1985 , , 292 , 371 da silva , a.c . ,
et al . , 1999 , submitted to mnras , astro - ph/9907224 fixsen , d.j .
, et al . , 1996 , , 473 , 576 fry , j.n . , 1984 , , 279 , 499 fukugita , m. , hogan , c.j .
and peebles , p.j.e , astro - ph/9712020 lange , a.e .
2000 , astro - ph/0005004 komatsu , e. and kitayama , t. , 1999 , , 526 , l1 ma , c.p . and pen , u.l .
, in preparation , 2000 makino , n. , and suto , y. , 1993 , , 405 , 1 matsubara , t. , 1999 , , 525 , 543 peacock , j.a . and dodds , s.j .
1996 , , 280 , l19 peacock , j. , _ cosmological physics , cambridge university press , 1999 peebles , p.j.e . , _ the large - scale structure of the universe , 1980 , princeton university press peebles , p.j.e . , 1983 , , 263 , l1 pen , u.l .
, 1997 , new astronomy , 2 , 309 pen , u.l .
, 1998 , , 498 , 60 pen , u.l . , 1999 , , 510 , l1 planck , 2007 , http://astro.estec.esa.nl/sa-general/projects/planck/ refregier , a. , komastu , e. , spergel , d.n . and pen , u.l .
, 1999 , submitted to prd , astro - ph/9912180 scoccimarro , r. and frieman , j. , 1999 , , 520 , 35 seljak , u. , burwell , j. and pen , u.l . , 2000 , submitted to prd , astro - ph/001120 submillimeter telescope , http://cfa-www.harvard.edu/$^{\sim}$aas/tenmeter[http://cfa-www.harvard.edu/$^\{\sim}$aas/tenmeter ]
tytler , d. , omeara , j.m . , suzuki , n. and lubin , d. 2000 , astro - ph/0001318 sloan digital sky survey , http://www.sdss.org white , s.d.m . ,
navarro , j.f . ,
evrard , a.e . ,
frenk , c.s .
1993 , nature , 366 , 429 zeldovich , y.b . and sunyaev , r. , 1969 , ap&ss , 4 , 301 _ _ | we apply the hierarchical clustering model and non - linear perturbation theory to the cosmological density and temperature fields .
this allows us to calculate the intergalactic gas pressure power spectrum , sz anisotropy power spectrum , skewness and related statistics .
then we show the effect of the non - gravitational heating .
our model confirms recent simulations yielding mass weighted gas temperature @xmath0 kev and reproduces the power spectra found in these simulations .
while the sz effect contains only angular information , we show that it is possible to extract the full time resolved gas pressure power spectrum when combined with galaxy photometric redshift surveys by using a variation on the cross - correlation .
this method further allows the disentanglement of the gravitational and the non - gravitational heating . |
a key goal of the large hadron collider ( lhc ) physics programme is to reveal the properties of the electroweak symmetry breaking ( ewsb ) sector in nature .
the atlas and cms data from the 7 and 8 tev runs of lhc show that the properties of the newly discovered higgs particle are compatible with those expected for the standard model ( sm ) higgs boson @xmath6 .
yet , it still needs to be determined whether the scalar sector consists of one @xmath7 doublet or has a richer structure , with additional states .
a particularly appealing feature of extensions of the sm scalar sector , like two - higgs - doublet - models ( 2hdms ) , is that they could successfully explain the generation of the cosmic matter - antimatter asymmetry via electroweak baryogenesis @xcite .
a key requirement in this context is that the electroweak phase transition ( ewpt ) in the early universe be strongly first order , which for the sm with a @xmath8 gev higgs is not the case @xcite .
here we review the results of @xcite , showing that the characteristic signature of a strongly first order ewpt in 2hdms is the decay @xmath9 , and analyzing its promising search prospects at the 14 tev run of lhc in the @xmath10 and @xmath11 final states .
this signature could then provide a connection between the generation of the cosmic matter - antimatter asymmetry in the early universe and searches for new physics at lhc .
the scalar sector of a 2hdm contains two scalar doublets @xmath12 . in the following
we assume for simplicity no charge - parity ( cp ) violation in the scalar sector .
then , apart from the recently observed ( cp - even ) higgs boson @xmath6 , the 2hdm physical spectrum contains a charged scalar @xmath2 , a cp - even scalar @xmath0 , and a cp - odd scalar @xmath1 . after fixing the electroweak ( ew ) vacuum expectation value ( vev ) @xmath13 gev and the higgs mass @xmath8 gev , the remaining parameters in the scalar potential are : the masses @xmath14 , @xmath15 , @xmath16 , two angles @xmath17 and @xmath18 and a dimensionful parameter @xmath19 . here
the angle @xmath18 is defined such that when @xmath20 , the state @xmath0 decouples from gauge bosons , and @xmath6 has sm - like properties , known as the _ alignment limit _ ( see @xcite for a review of 2hdms , and @xcite for details on the definition of parameters and the conventions chosen ) . in order to study the strength of the ewpt in 2hdms , a scan over the new physics parameters @xmath14 , @xmath15 , @xmath16 , @xmath21 , @xmath22 and @xmath19 is performed .
the yukawa type of 2hdm considered is irrelevant for the ewpt ( all types couple in the same way to the top quark ) , while experimental constraints do differ between types .
we choose a type i 2hdm , which is less constrained than a type ii , and thus provides a better gauging of the impact of a first order ewpt on the 2hdm parameter space .
the scan is interfaced to 2hdmc @xcite and higgsbounds @xcite to select points in parameter space that satisfy stability , unitarity , perturbativity , ew precision constraints and collider bounds .
flavour constraints ( coming mainly from @xmath23 for type i 2hdm @xcite ) and constraints from measured higgs signal strengths on @xmath21 and @xmath24 ( see e.g. @xcite ) are also included .
point in our scan satisfying the above constraints are called _ physical points_. for them , the strength of the ewpt is computed via the thermal 1-loop effective potential ( see @xcite for details ) . in figure
[ fig : ewpt ] we show heat - maps of physical points ( _ left _ ) and points with a strongly first order ewpt ( _ right _ ) in the planes ( @xmath25 ) and ( @xmath26 ) .
altogether , a strong ewpt as needed for electroweak baryogenesis strongly favours a rather heavy cp - odd scalar @xmath1 ( @xmath27 gev ) , together with a large mass splitting @xmath28 , as shown in figure [ fig : ewpt ] ( bottom )
. it also favours the light higgs @xmath6 to have sm - like properties , _
i.e. _ small @xmath24 and moderate @xmath21 @xcite , the range of @xmath24 leading to a strong ewpt shrinking as the cp - even state @xmath0 becomes heavier . )
_ bottom _ : ( @xmath26)-plane .
the dotted - black line corresponds to @xmath29.,scaledwidth=63.0% ] [ fig : ewpt ]
the results of the previous section point towards the @xmath9 decay channel ( see also @xcite ) as a characteristic signature of 2hdms with a strong ewpt , to be searched for at the upcoming 14 tev run of the lhc .
this decay is strongly favoured both by the large phase space available and by the @xmath31 dependence of the coupling @xmath32 , unsuppressed in the alignment limit .
this is in contrast with the decay @xmath33 , which vanishes in that limit ( since @xmath34 ) , and is therefore suppressed in this scenario ( see figure [ fig : br ] ) .
the competing decay channels would then be @xmath35 and possibly @xmath36 .
the former is subdominant for @xmath37 ( figure [ fig : br ] ) .
the presence of the latter depends on the splitting @xmath38 .
ew precision observables require @xmath2 to be close in mass to either @xmath0 or @xmath1 @xcite , which makes @xmath36 either kinematically forbidden or similar to @xmath9 , and here we assume for simplicity @xmath39 . in the following we analyze two prototypical scenarios which feature @xmath40 gev , @xmath41 and @xmath42 gev , and with @xmath43 ( benchmark a ) and @xmath44 ( benchmark b ) , as shown in figure [ fig : br ] .
these two benchmarks characterize the two dominant alternatives for the subsequent decay of @xmath0 : @xmath45 dominates very close to the alignment limit , while away from it @xmath46 ( and to a lesser extent @xmath47 ) is the dominant decay mode . as a function of @xmath14 for @xmath42 gev , @xmath48 , @xmath40 gev , @xmath49 ( benchmark a , solid lines ) and @xmath50 ( benchmark b , dotted lines ) .
_ right _ : main branching ratios of @xmath0 as a function of @xmath14 ( same benchmark parameters as in _ left_).,title="fig:",scaledwidth=45.0% ] as a function of @xmath14 for @xmath42 gev , @xmath48 , @xmath40 gev , @xmath49 ( benchmark a , solid lines ) and @xmath50 ( benchmark b , dotted lines ) .
_ right _ : main branching ratios of @xmath0 as a function of @xmath14 ( same benchmark parameters as in _ left_).,title="fig:",scaledwidth=45.0% ] [ fig : br ] this discussion highlights the fact that for 2hdms with a strongly first order ewpt , the corresponding smoking gun " signature at lhc will either be @xmath51 or @xmath52 , depending on how close the 2hdm is to the alignment limit .
we now analyze the search prospects in the @xmath55 and @xmath56 channels at the 14 tev run of the lhc using the defined benchmarks a and b from section [ section3 ] and considering in both cases a mass for the cp even scalar @xmath57 gev .
concentrating first on benchmark a , which corresponds to the @xmath55 final state , the two main sm backgrounds are : ( i ) @xmath58 production ( with @xmath59 ) , ( ii ) qcd @xmath60 production ( with @xmath61 ) , while the rest of potential backgrounds ( e.g. @xmath62 production and associated production of a higgs @xmath6 and a @xmath63 boson ) are found to be practically negligible @xcite .
regarding our analysis , we implement the type - i 2hdm in feynrules
@xcite and use madgraph5@xmath64amc@xmath65nlo @xcite to generate both signal and background analysis samples , which are then passed on to pythia @xcite and delphes @xcite for parton showering , hadronization and a detector simulation . for event selection
we require the presence of two isolated same flavour ( sf ) leptons in the final state with @xmath66 , @xmath67 and @xmath68 2.5 ( 2.7 ) for electrons ( muons ) , together with two b - tagged jets in the event with @xmath69 , @xmath70 and @xmath71 2.5 ( see @xcite for details ) , and the subsequent cut - flow analysis is presented in table [ table1 ] .
we define the signal region as @xmath72 gev and @xmath73 gev ( small b - jet energy loss due to showering is expected ) , and show the @xmath74 and @xmath75 distributions after cuts in figure [ fig : dist1 ] ( left ) ( various contributions stacked and for an integrated luminosity of @xmath76 ) .
the results from table [ table1 ] show that a discovery value @xmath77 may be obtained already with @xmath78 fb@xmath79 in the limit that only statistical uncertainties are important .
.event selection ( see section iii ) and background reduction in the @xmath55 final state .
cross section ( in fb ) is shown after successive cuts for the signal @xmath80 and the dominant backgrounds @xmath81 and @xmath82 . [ cols="^,^,^,^",options="header " , ] ( _ left _ ) and @xmath75 ( _ right _ ) distributions after analysis cuts , with the various contributions stacked ( for an integrated luminosity of @xmath83 ) .
right : @xmath84 ( _ left _ ) and @xmath85 ( _ right _ ) distributions after event selection , with the various contributions stacked ( for an integrated luminosity of @xmath86).,title="fig:",scaledwidth=48.0% ] ( _ left _ ) and @xmath75 ( _ right _ ) distributions after analysis cuts , with the various contributions stacked ( for an integrated luminosity of @xmath83 ) . right : @xmath84 ( _ left _ ) and @xmath85 ( _ right _ ) distributions after event selection , with the various contributions stacked ( for an integrated luminosity of @xmath86).,title="fig:",scaledwidth=48.0% ] allowing a departure from the alignment limit ( our benchmark b ) the decay channel @xmath87 , @xmath88 provides the best discovery prospects ( @xmath89 , leading to a final state @xmath90 has been considered in @xcite ) .
the main background is diboson ( @xmath62 ) production with @xmath91 ( other backgrounds like @xmath92 , @xmath93 and @xmath94 are very small even before cuts ) . for event selection ,
we require the presence of four isolated leptons in the final state with @xmath66 gev , @xmath95 gev , and require that one lepton pair ( opposite sign , same flavour ( sf ) ) reconstructs @xmath96 within @xmath97 gev .
the leading order cross sections at lhc 14 tev after event selection for the signal and @xmath62 background are respectively @xmath98 fb and @xmath99 fb . we note that a @xmath63-veto on the remaining lepton pair would greatly suppress the @xmath62 background , as would a veto on sf leptons ( this at the expense of reducing the signal by a factor 2 ) . defining the transverse mass variables @xmath84 and @xmath85 @xmath100 ( @xmath101 are the two sf leptons most closely reconstructing @xmath96 )
, a signal region of @xmath102 gev ( see figure [ fig : dist1 ] , right ) allows to extract a clean signal @xcite .
our final signal cross section is 1.41 fb , which compared to a background of 1.7 fb reaches a significance of 5 with @xmath103 fb@xmath79 . altogether , this analysis highlights that the decay @xmath9 , being a ` smoking gun ' signature of 2hdm scenarios with a strongly first order ewpt , can be probed at the 14 tev run of lhc in either of the two dominant final states , thus providing a powerful probe of the ewpt and ew cosmology at the lhc .
i am deeply indebted to glauber dorsch , stephan huber and ken mimasu , from which i learned a lot by working together , and who made our collaboration really fun .
i also want to thank veronica sanz and michael ramsey - musolf for useful discussions and comments , and the organizers of hpnp2015 for their great hospitality during the conference .
the work of j.m.n . is supported by the people programme ( marie curie actions ) of the european union seventh framework programme ( fp7/2007 - 2013 ) under rea grant agreement pief - ga-2013 - 625809 .
n. turok and j. zadrozny , nucl .
b * 358 * , 471 ( 1991 ) ; j. m. cline , k. kainulainen and a. p. vischer , phys . rev .
d * 54 * , 2451 ( 1996 ) [ hep - ph/9506284 ] ; l. fromme , s. j. huber and m. seniuch , jhep * 0611 * ( 2006 ) 038 [ hep - ph/0605242 ] ; j. m. cline , k. kainulainen and m. trott , jhep * 1111 * , 089 ( 2011 ) [ arxiv:1107.3559 [ hep - ph ] ] . k. kajantie , m. laine , k. rummukainen and m. e. shaposhnikov , nucl .
b * 466 * ( 1996 ) 189 [ hep - lat/9510020 ] ; k. kajantie , m. laine , k. rummukainen and m. e. shaposhnikov , phys .
* 77 * ( 1996 ) 2887 [ hep - ph/9605288 ] .
a. celis , v. ilisie and a. pich , jhep * 1307 * , 053 ( 2013 ) [ arxiv:1302.4022 [ hep - ph ] ] ; m. krawczyk , d. sokolowska and b. swiezewska , j. phys .
ser . * 447 * , 012050 ( 2013 ) [ arxiv:1303.7102 [ hep - ph ] ] ; b. grinstein and p. uttayarat , jhep * 1306 * , 094 ( 2013 ) [ erratum - ibid . * 1309 * , 110 ( 2013 ) ] [ arxiv:1304.0028 [ hep - ph ] ] ; c. -y .
chen , s. dawson and m. sher , phys .
d * 88 * , 015018 ( 2013 ) [ arxiv:1305.1624 [ hep - ph ] ] ; o. eberhardt , u. nierste and m. wiebusch , jhep * 1307 * , 118 ( 2013 ) [ arxiv:1305.1649 [ hep - ph ] ] ; g. belanger , b. dumont , u. ellwanger , j. f. gunion and s. kraml , phys . rev .
d * 88 * ( 2013 ) 075008 [ arxiv:1306.2941 [ hep - ph ] ] ; b. dumont , j. f. gunion , y. jiang and s. kraml , phys . rev .
d * 90 * ( 2014 ) 035021 [ arxiv:1405.3584 [ hep - ph ] ] ; m. gorbahn , j. m. no and v. sanz , arxiv:1502.07352 [ hep - ph ] .
n. d. christensen and c. duhr , comput .
commun . * 180 * , 1614 ( 2009 ) [ arxiv:0806.4194 [ hep - ph ] ] ; c. degrande , c. duhr , b. fuks , d. grellscheid , o. mattelaer and t. reiter , comput .
commun . * 183 * , 1201 ( 2012 ) [ arxiv:1108.2040 [ hep - ph ] ] .
j. alwall , m. herquet , f. maltoni , o. mattelaer and t. stelzer , jhep * 1106 * , 128 ( 2011 ) [ arxiv:1106.0522 [ hep - ph ] ] ; j. alwall , r. frederix , s. frixione , v. hirschi , f. maltoni , o. mattelaer , h .- s .
shao and t. stelzer _ et al .
_ , jhep * 1407 * ( 2014 ) 079 [ arxiv:1405.0301 [ hep - ph ] ] . | a cosmological first order electroweak phase transition could explain the origin of the matter - antimatter asymmetry in the universe .
such a phase transition does not occur in the standard model , while it becomes possible with the existence of a second higgs doublet in nature .
we obtain the properties of the new scalars @xmath0 , @xmath1 and @xmath2 that lead to such a phase transition , showing that its characteristic signature at lhc would be the observation of the decay @xmath3 .
we analyze the lhc search prospects for this decay in the @xmath4 and @xmath5 final states , showing that either one is promising at the early stages of the 14 tev run . |
the arctic ocean soundscape is defined by its sea ice cover .
internal frictional shearing , thermal stress fracturing , and interaction within leads generate distinct sounds that can exceed 100 db re 1@xmath2 pa@xmath3 hz@xmath4 . at the same time , the upward refracting sound speed profile and nearly year round ice cover create a propagation channel that preserves low frequency signals while attenuating higher frequency components .
this unique environment depends strongly on the properties of the arctic sea ice , including percentage of areal cover , thickness ( age ) , and lateral extent . in the past decade , the arctic sea ice has dramatically reduced in thickness as well as annual extent,@xcite resulting in unknown changes to the soundscape .
sea ice noise and the arctic soundscape properties have historically been an area of interest in underwater acoustics.@xcite@xmath5@xcite measurements of ice noises have shown that they are highly non
gaussian,@xcite varying in characteristic by the generating mechanism,@xcite but often more prevalent near ice ridges.@xcite@xmath5@xcite the cumulative ambient noise levels generated by ice noise have been shown to correlate with environmental variables like wind , air pressure , and temperature.@xcite@xmath5@xcite@xmath5@xcite@xmath5@xcite near the marginal ice zone ( miz ) , where the ice is subject to increased wave forcing , noise levels have been shown to increase as much as 10 db from those further under the ice cap.@xcite@xmath5@xcite it is well known that the sea ice is a strong scatterer that attenuates high frequencies at a much higher rate than the open ocean,@xcite although the exact attenuation coefficients depend on the local sea ice structure and have yet to be determined.@xcite due in large part to biological activity and experimental accessibility , the western arctic soundscape near the beaufort sea@xcite@xmath5@xcite@xmath5@xcite@xmath5@xcite has been studied more extensively than the eastern arctic soundscape near and north of the fram strait .
studies north of 85@xmath6n are extremely rare . in april 2013 ,
a bottom moored vertical hydrophone array was deployed at ice camp barneo near 89n , 62w .
the experiment was designed to study the propagation properties and transmission loss under the sea ice , as well as the northern polar soundscape . around
april 15 , 2013 the mooring cable failed .
the subsurface float rose to the surface where it remained constrained by the sea ice , with the array hanging below .
it drifted southward with the transpolar current toward the fram strait , recording ambient noise as scheduled .
microcat pressure measurements ( see sec . [
subsec : microcats ] ) showed that the array was vertical under its own weight during much of the transit .
the resulting data record the spatiotemporal variation of the far northern arctic soundscape ( @xmath7 85 n ) . in this study ,
the dataset is analyzed and the observations are interpreted in terms of previous studies of the arctic soundscape . the paper is organized as follows . in sec .
ii , the acoustic experiment is described , data processing methods are explained , and collection of supplementary environmental data is discussed . in sec .
iii , select noise events are discussed . in sec .
iv , the results of statistical soundscape analyses in both time and depth are presented . in the final sec .
v , arctic soundscape power estimates from previous studies are compared with summer 2013 .
the goal of this paper is to establish an understanding of soundscape contributors and sound levels in the northeastern arctic during summer 2013 .
p1.8 cm < p2 cm < p2 cm < p1.7 cm < hydrophone # & hydrophone depths ( m ) & microcat depths ( m ) & microcat # + 1 & 12 & 5 & 1 + 2 & 26.5 & 25 & 2 + 3 & 41 & - & - + 4 & 55.5 & 50 & 3 + 5 & 70 & - & - + 6 & 84.5 & - & - + 7 & 99 & 100 & 4 + 8 & 113.5 & - & - + 9 & 128 & - & - + 10 & 142.5 & - & - + 11 & 159.1 & 150 & 5 + 12 & 178.1 & - & - + 13 & 199.9 & 201 & 6 + 14 & 224.8 & - & - + 15 & 253.4 & 250 & 7 + 16 & 286.1 & - & - + 17 & 323.6 & - & - + 18 & 366.5 & 350 & 8 + 19 & 415.6 & - & - + 20 & 471.8 & 450 & 9 + 21 & 536.2 & - & - + 22 & 610 & 600 & 10 + a 600 m long bottom - moored acoustic receiving array was deployed at ice camp barneo , 8923.379n , 6235.159w , on april 14 , 2013 .
twenty - two omnidirectional hydrophones were spaced along the array , with phones 110 separated by 14.5 m and phones 1122 separated by logarithmically increasing spacing starting at 16.5 m ( table [ table : array ] , fig .
[ fig : array ] ) .
the topmost hydrophone was located 12 m below the subsurface float .
the hydrophones recorded underwater sound for 108 minutes six days per week ( sunday through friday ) , starting at 1200 utc , with a sampling frequency of 1,953.125 hz .
acoustic recordings are available for 119 days between april 29 and september 20 , 2013 .
the raw acoustic recordings were scaled to be in units of instantaneous sound pressure using the analog - to - digital conversion parameters , the gain , and the hydrophone receiving sensitivity given by the manufacturer .
the system noise floor was taken from a model combining the known self noise of its individual components .
the system was also tested in a faraday cage and by calculating the coherence between multiple sensors recording noise in a quiet room , which both fit well with the modeled system noise floor .
median spectral estimates were created by segmenting data across a given time period into 4096-point windows ( @xmath8 2 s ) , taking a 16,384-point fast fourier transform of each window , and calculating the median of the individual spectral estimates .
the frequency bins are 0.12 hz for these estimates .
spectrograms were estimated separately using 512-point windowed segments ( @xmath8 0.25 s ) zero - padded to 2048 points ( df @xmath9 1 hz ) .
all data were recorded at 84.5 m ( hydrophone # 6 ) unless otherwise noted .
ten sea - bird sbe 37sm / smp microcat instruments were co - located with the hydrophones , spaced 25 , 50 , 50 , 50 , 50 , 100 , 100 , and 150 m apart . the topmost microcat was located 5 m below the subsurface float ( table [ table : array ] , fig .
[ fig : array ] ) .
the microcats began recording on april 28 and sampled continuously until september 19 .
the sampling period of the top four instruments was 480 s , the next five 380 s , and the deepest 300 s. a xeos technologies kilo iridium - gps mooring location beacon located on top of the subsurface float began transmitting alarm messages on may 3 , 2013 , indicating that the mooring had prematurely surfaced .
the reported position at the time of surfacing was 8850.30n , 5117.91w , 63 km from the deployment location .
analysis of an acoustic survey on april 14 , following deployment of the mooring , revealed that the acoustic release was significantly shallower than expected .
the implication is that the mooring failed shortly after deployment , but the subsurface float was trapped beneath sea ice , preventing the location beacon from obtaining gps positions or transmitting alarm messages until it was exposed on may 3 .
the float drifted southward in the transpolar drift .
there were frequent gaps in transmissions from the location beacon , which are presumed to coincide with periods when the subsurface float was covered by sea ice ( fig .
[ fig : map ] ) .
the buoy was recovered on september 21 , 2013 , at 8403.50n , 00305.83w ( fig .
[ fig : float ] ) . the mooring line was found to have parted immediately above the anchor ( fig .
[ fig : array ] ) .
a 56day discontinuous timeseries of the bathymetry along the array drift path between may 3 and september 21 was created using the international bathymetric chart of the arctic ocean from the national centers for environmental information ( fig . [
fig : bathy ] ) .
the georeferenced polar stereo projection bathymetry grid was indexed at the desired coordinate locations to obtain the timeseries data .
acoustic propagation along the drift path is ray dominated and confined primarily to the top 200 m , so the variation in bathymetry has a minimal effect on the measured ambient noise ( fig .
[ fig : ssp ] ) . low frequency ( @xmath10 hz ) cable strum was observed but kept in the spectral estimates for comparison purposes .
strong spectral bands were also observed , exceeding 100 db re 1 @xmath2pa@xmath3 hz@xmath4 and extending to the nyquist frequency ( 976.56 hz ) ( fig .
[ fig : filters ] ) . periods of unexpectedly low pressures ( depth ) on the microcats corresponded to the periods of high acoustic power ( fig . [ fig : pressures ] ) .
it therefore seems unlikely that these high spectral levels are due to propagating acoustic noise . with the buoyant subsurface float constrained to the surface ,
flow past the mooring lifts and thus tilts the array and reduces the microcat pressures ( depths ) .
potential non - acoustic noise sources associated with flow past the mooring include strumming , flow noise , and/or mechanical vibrations .
to remove data affected by flow noise , the median microcat pressure for each day was computed .
the pressure on the deepest microcat ( 600 m ) had the largest variation between days and was used as an indicator of flow - related noise ( fig .
[ fig : mucats ] ) . by comparing the good and bad spectrograms ( fig .
[ fig : filters ] ) with their median microcat pressures , it was found that most corrupted data had a median pressure level below 604.9 dbars .
therefore days with @xmath11 dbars were not used .
this method selected 19 days for further analysis : april 30 , may 1 , 2 , 7 , 8 , 9 , 12 , 14 , june 16 , 18 , july 3 , 14 , 19 , 24 , august 2 , and september 10 , 18 , 19 , 20 .
the near polar arctic under - ice soundscape is generated by sea ice , wind , anthropogenic and biologic noise sources that travel long distances within the subsurface propagation duct .
the following section discusses specific examples of underwater noise for the eastern arctic soundscape of summer 2013 .
the soundscape observed during the array transit consists primarily of background noise from distant events .
the spectrum is characterized by a broad spectral peak at 1020 hz and a power fall off above @xmath8100 hz ( fig .
[ fig : background_noise ] , log frequency ) .
the hydrophone received pressure timeseries shows that the background noise and broadband noise vary by a factor of 10 , despite appearing similar in spectrogram estimates ( fig .
[ fig : background_noise ] , [ fig : broadband_noise ] ) .
the hydrophone timeseries also emphasizes the wide variability within a recording period due to ice noise ( fig .
[ fig : broadband_noise ] ) . the median spectral estimate for each spectrogram
can be compared to the daily and monthly median spectral estimates ( fig .
[ fig : dailymonthly ] ) , demonstrating the smoothing effect of using a longer recording for the median spectral estimate .
ice noises were observed to be either broadband or tonal in nature , lasting from 1 to 100 s. broadband ice noise extended across the frequency band ( fig .
[ fig : broadband_noise ] , log frequency ) . tonal ice noises are single frequency or harmonic signatures modulated in time ( figs .
[ fig : squeak1]-[fig : squeak3 ] , linear frequency ) . xie and
farmer@xcite demonstrated that constant
frequency ice tonals could be modeled as resonance in an infinitely long sea ice block of uniform height , density , and velocity generated by frictional shear stress on its edge .
the nonlinear tonals observed here may indicate anomalies in the local height or composition of the sea ice or a frictional stress that is velocity dependent ( fig .
[ fig : squeak1 ] ) .
the degree of nonlinearity varies between hydrophone recordings ( fig .
[ fig : squeak2 ] ) , indicating that significant changes in ice properties and dynamics may occur within the spatiotemporal span of 23 array drift days .
another interesting case is ocean swell modulated ice tonals , with period @xmath89 s ( fig .
[ fig : squeak3 ] ) .
ocean waves impinging on the sea ice edge can generate seismic or flexural waves that propagate within the sea ice , if the frequency ice product is less than about 300 hz
m.@xcite swell modulated ice tonals observed on the receiving array suggest that these effects can be seen as far as 230 km from the ice edge .
broadband pressure pulses generated by airguns are used to image the ocean bottom subsurface during seismic surveys . at long distances ,
higher frequencies ( f @xmath7 100 hz ) are attenuated by scattering at the water - ice boundary .
the resulting pulses can be observed on hydrophone receivers at f @xmath12 50 hz .
distant noise from seismic surveys can be observed almost daily in the fram strait during summer months .
for example , airgun surveys were observed on 9095% of days between july and september 2009.@xcite airgun survey pulses were observed on 11 out of 19 selected recording days between may 7 and sep .
[ fig : airgun ] , linear frequency ) . due to irregular sampling , the 13 seismic surveys recorded per month represented seismic surveys on 43%100% of the recording days in may through september .
these pulses arrived at about @xmath13 4from horizontal ( fig .
[ fig : beam ] ) , as determined by incoherently averaged 2.1 s bartlett beamformer estimates across june 16 . for other recording days
, the pulse arrivals ranged between 515 .
the shallow arrival angles indicate that these low
frequency signals travel great distances within the subsurface sound propagation channel .
location , type and date of surveys in norwegian territory are available through the norwegian petroleum directorate . according to these data ,
the array was between 1800 and 3500 km distant from seismic surveys at the beginning of its transit in april and 1000 to 3000 km at the end in september . using the advanced microwave scanning radiometer-2 ( amsr-2 ) 89-ghz channel satellite dataset ( ) ,
@xcite the array was found to be about 1000 km distant from the ice edge in april and 200 km distant in september ( fig .
[ fig : source_distances ] ) .
it is likely that the signals observed arrived from the closest surveys .
thus airgun signals received at 7080 db re 1 @xmath2pa@xmath3 hz@xmath4 had propagated approximately 800 km in open water and anywhere from 200 to 1000 km under the ice . frequent seismic activity and low
magnitude earthquakes occur where the north american / eurasian plate boundary meets the gakkel ridge . because the nearest seismic stations are land based and up to 1000 km away , many of these earthquakes
are not registered with the global seismic network . a receiving array deployed in the lincoln sea
has been successful in detecting and locating earthquakes originating from this juncture.@xcite earthquakes can be observed acoustically through the @xmath0 phase arrival .
the @xmath0 phase is an acoustic pressure wave coupled into the water column from the ocean bottom at an anomaly ( for example , a seamount ) .
the @xmath0 phase arrives after the @xmath14 and @xmath15 seismic arrivals but is the most visible arrival .
the differences in arrival time between the three waves can be used to localize the earthquake if appropriate propagation velocities are known .
@xmath0 phase arrivals were observed during the array transit in summer 2013 , indicating that arctic basin earthquakes also contribute to the low frequency soundscape ( fig .
[ fig : earthquake ] ) .
the earthquake center was estimated to be about 100 km from the receiving array .
in this section , spectral estimates from different months at several depths are compared to establish the spatiotemporal dependence of ambient noise in the eastern arctic during summer 2013 .
three days were selected to examine the daily variation in median ambient noise levels : may 7 , july 14 , september 10 ( fig . [ fig : dailymonthly ] ) .
the 10@xmath16 , 50@xmath16 ( median ) , and 90@xmath16 percentiles for may 2013 show the variation in spectral shape between the most and least frequent generation mechanisms .
for example , cable strum is visible at f @xmath12 10 hz in the 90@xmath16 percentile .
other peaks near 10100 hz may be caused by flow
related noise or seismic survey activity .
the 10@xmath16 percentile is defined by single , broad peak at 15 hz due to distant sources within the sound channel .
the daily median estimates progress from least to most peaked , with may 7 showing a peak to peak ( 15 to 900 hz ) difference of 26 db and september 10 showing a peak to peak difference of 44 db .
this trend indicates that the earlier spatiotemporal recordings contain more higher frequency noises whereas the later recordings were influenced by lower frequency sources .
the trend of increasing peakedness in the median spectral is more apparent in the monthly median spectra ( fig .
[ fig : dailymonthly ] ) . at low frequencies ( f @xmath12 100 hz ) ,
an increase in median ambient noise levels corresponds to decreasing distance to the ice edge and to seismic survey activity . at higher frequencies ( f @xmath7 100 hz ) median
ambient noise levels decrease with time , corresponding to fewer ice creak signatures observed in spectrograms . when limited data was available , such as in august , the median monthly spectra was likely to show evidence of flow related noise i.e. at about 75 and 90 hz , despite efforts to remove this noise .
it is interesting to note these spatiotemporal differences in median ambient noise levels with proximity to the marginal ice zone ( miz ) ; the effect of the miz on ambient noise was previously studied only for distances less than 150 km @xcite .
the decrease in high frequency median ambient noise levels may represent a physical change in the spatial or temporal ice stress field .
the median ambient noise depth profile during summer 2013 was found to depend on frequency . at low frequencies ( 1050 hz ) , a local peak in median ambient noise
is centered in the subsurface propagation channel at the depth where ice reflections and upward refracting rays converge ( fig .
[ fig : depth ] ) .
below the sound channel and at higher frequencies ( 200500 hz ) , the median noise level is constant with depth as expected for an isotropic distribution of surface sources ( fig .
[ fig : depth ] ) .
there is a strong signal during may at about 400 600 m depth that is likely due to flow related noise , which could not be completely removed from the lower hydrophones using the noise filtering method ( sec .
[ sssec : filtering ] ) .
[ fig : thaaw_framiv ] cccccccccc + & location ( lat , lon ) & experiment & dates & 15 hz & 50 hz & 100 hz & 500 hz & 1 khz + & 86n 56.9w & may june 2013 & 05/2013 & 76.5 & 66 & 60.2 & 43.7 & - + & 89n 1e & & 06/2013 & & & & & + & 86n 1.3e & july sep .
2013 & 07/2013 & 78.7 & 64.9 & 55.6 & 37.6 & - + & 83.8n 4.5e & & 09/2013 & & & & & + & 83n 20e & fram iv@xcite & 04/1982 & 90 & 79.5 & 73 & 60 & 53 + + & 82n 168e & mellen , marsh 1985@xcite & 0910/1961 & 72 & 70 & 61 & 51 & 40 + & 75n 168w & & 0509/1962 & 63 & 64 & 49 & 37 & 32 + & & & & - & 75 & 72 & 61 & 52 + & 78.5n 105.25w & ice pack i @xcite & 27/04/1961 & 50 & 42 & 38 & 37 & 20 + & & & 28/04/1961 & 58 & 52 & 51 & 52 & 51 + & 74.5n 115.1w & ice pack ii @xcite & 9/23/1961 & - & 57 & 56 & 52 & 43 + & beaufort sea & prl@xcite & april 1975 & 73 & 68 & 62 & 48 & 43 + & & & & ( 10 hz ) & ( 32 hz ) & & + & @xmath872n 142w & aidjex@xcite & 08/1975 & 6585 & 6575 & - & - & 3855 + & & & 11/1975 & 7090 & 6588 & - & - & 4070 + & & & 02/1976 & 6590 & 6090 & - & - & 3570 + & & & 05/1976 & 6588 & 6090 & - & - & 3768 + & & & & & & & & + & 71n 126.07w & kinda et al .
2013@xcite & 11/2004 & 68 & 69 & 66 & 58 & 54 + & & & 06/2005 & & & & & + & 72.46n 157.4w & roth et al .
2011@xcite & 09/2008 & 84 & 80 & 74 & 60 & 56 + & & & 03/2009 & 84 & 70 & 62 & 48 & 48 + & & & 05/2009 & 76 & 61 & 56 & 44 & 44 + [ tab : spectrallevels ] the summer 2013 median ambient noise results are here compared with historical estimates from both western and eastern arctic stations .
the median spectral estimates for may 2013 was below , but similarly structured to , a composite spectral estimate from april 1982.@xcite the peak at 15 hz appears less prominent at lower frequencies in 2013 than in 1982 . in comparison ,
a spectral estimate recorded in the beaufort sea in april 1975 shows comparable ambient noise levels and structure to 2013 but does not extend to lower frequencies ( fig .
[ fig : comparisons]).@xcite the differences in these spectra may be caused by environmental factors or by experimental factors , including recording length and post processing methods , which were not published alongside the results .
the broad peak at 15 hz in all estimates can be attributed to distant ice and seismic survey noises propagating in the sound channel,@xcite as higher frequencies are more attenuated and lower frequencies have long wavelengths compared to the channel .
the higher frequencies fall off at a consistent rate of f@xmath1 .
it is likely that an increased cumulative ice noise source level increases high frequency noise , thus reducing the fall off rate ( fig .
[ fig : dailymonthly ] ) .
[ fig : results ] demonstrates the wide variability in arctic ambient noise estimates across frequency , year , and study .
this variability arises from a complex relationship between the arctic soundscape and both environmental and anthropogenic factors , such as sea ice percent cover , sea ice age / thickness , barometric conditions and wind patterns , local subsurface currents , seismic survey activity , and marine biologic activity .
the studies shown indicate that , without correction for environmental factors , there is not a significant trend in the arctic soundscape power levels between 1960 and 2013 , but that frequency dependent ambient noise levels are within a 3040 db range for both regions of the ice covered arctic .
between april and september , 2013 , a 22element vertical hydrophone array recorded the eastern arctic for 108 minutes / day between 89n , 62w and svalbard ( north of the fram strait ) .
the data were analyzed to produce spectral estimates of the median soundscape and demonstrate that ice noise and seismic airgun surveys were the dominant noise sources .
the median ambient noise level for may 2013 was below a similar eastern arctic estimate from april 1982,@xcite but comparable to a western arctic estimate from april 1975.@xcite a multi decadal summary of arctic soundscape studies demonstrates that the estimated ambient noise levels depend strongly on experimental and environmental parameters and that there is not a significant trend in ambient noise levels among the studies examined .
we would like to thank john colosi for providing the microcat data , john kemp and the whoi mooring operations and engineering group for their assistance , and hanne sagen and the norwegian coast guard for assistance in recovering the array mooring .
this work is supported by the office of naval research under award numbers n00014 - 13 - 1 - 0632 and n00014 - 12 - 1 - 0226 .
any opinions , findings , and conclusions or recommendations expressed in this publication are those of the authors and do not necessarily reflect the views of the office of naval research . daily sea ice concentration , defined as the areal percentage of satellite imagery above a certain brightness level , was obtained from the advanced microwave scanning radiometer-2 ( amsr-2 ) 89-ghz channel satellite dataset,@xcite provided in a 4 km x 4 km gridded format from the institute of environmental physics , university of bremen , germany ( fig .
[ fig : environmental ] ) .
the sea ice concentration ranges from 0 ( no ice ) to 100 ( solid ice ) .
the georeferenced latitude and longitude grids were transformed into regular latitude and longitude grids with 0.1@xmath6 resolution and the ice concentration was interpolated to the array location .
n - s and e - w wind velocity components were obtained from the european center for medium - range weather forecasts ( ecmwf ) ( fig .
[ fig : environmental ] ) .
the era - interim reanalysis from which the data was drawn uses 30 minute time steps and a spectral t255 geographic resolution corresponding to 79 km spacing on a reduced gaussian grid.@xcite local wind velocity for the 10 m isobar at the array location was extracted .
r. lindsay and a. schweiger , @xmath17arctic sea ice thickness loss determined using subsurface , aircraft , and satellite observations,@xmath18 _ t.c . _ * 9 * , 269283 ( 2015 ) .
i. dyer , @xmath17the song of sea ice and other arctic ocean melodies,@xmath18 in _ arctic technology and policy _ , edited by i. dyer and c. chryssostomidis ( hemisphere , new york , 1984 ) , pp .
w. c. cummings , o.i .
diachok , and j. d. shaffer , @xmath17acoustic transients of the marginal sea ice zone : a provisional catalog,@xmath18 _ naval research laboratory memorandum report 6408 _ , dtic number ada214142 . naval research lab , washington , d.c .
veitch and a.r.wilks , @xmath17a characterization of arctic undersea noise,@xmath18 _ j. acoust .
* 77 * , 989999 ( 1985 ) .
kinda , y. simard , c. gervaise , j.i .
mars , and l. fortier , @xmath17arctic underwater noise transients from sea ice deformation : characteristics , annual time series , and forcing in beaufort sea,@xmath18 _ j. acoust .
am . _ * 138 * , 20342045 ( 2015 ) .
diachok , @xmath17effects of sea ice ridges on sound propagation in the arctic ocean,@xmath18 _ j. acoust .
* 59 * , 11101120 ( 1976 ) .
b.m . buck and j.h .
wilson , @xmath17nearfield noise measurements from an arctic pressure ridge,@xmath18 _ j. acoust .
* 80 * , 256264 ( 1986 ) .
greene and b.m .
buck , @xmath17arctic ocean ambient noise,@xmath18 _ j. acoust .
* 36 * , 1218 ( 1964 ) .
makris and i. dyer , @xmath17environmental correlates of pack ice noise,@xmath18 _ j. acoust . soc .
* 79 * , 1434 1440 ( 1986 ) .
makris and i. dyer , @xmath17environmental correlates of arctic ice - edge noise,@xmath18 _ j. acoust .
am . _ * 90 * , 32883298 ( 1991 ) .
kinda , y. simard , c. gervaise , j.i .
mars , and l. fortier , @xmath17under - ice ambient noise in the eastern beaufort sea , canadian arctic , and its relation to environmental forcing,@xmath18 _ j. acoust .
* 134 * , 7787 ( 2013 ) .
o.i . diachok and r.s .
winokur , @xmath17spatial variability of underwater ambient noise at the arctic ice - water boundary,@xmath18 _ j. acoust .
* 55 * , 750753 ( 1974 ) . f. geyer , h. sagen , g. hope , m. babiker , and p. f. worcester , @xmath17identification and quantification of soundscape components in the marginal ice zone,@xmath18 _ j. acoust .
* 139 * , 18731885 ( 2016 ) .
o. diachok , @xmath17arctic hydroacoustics,@xmath18 , _ cold reg .
_ * 2 * , 1861201 .
d. tollefsen and h. sagen , @xmath17seismic exploration noise reduction in the marginal ice zone,@xmath18 _ j. acoust .
* 136 * , el4752 ( 2014 ) .
e. h. roth , j. a. hildebrand , s. m. wiggins , and d. ross , @xmath17underwater ambient noise on the chuckchi sea continental slope from 2006 - 2009,@xmath18 _ j. acoust .
* 131 * , 104110 ( 2012 ) . c. l. berchok , p. j. clapham , j. crance , s. e. moore , j. napp , j. overland , m. wang , p. stabeno , m. guerra , and c. clark , @xmath17passive acoustic detection and monitoring of endangered whales in the arctic ( beaufort , chukchi ) and ecosystem observations in the chukchi sea : biophysical moorings and climate modeling,@xmath18 annual report 2012 , contract m09pc00016 ( akc 083 ) , bureau of ocean energy management , regulation , and enforcement , anchorage , alaska . j. k. lewis and w. w. denner , @xmath17arctic ambient noise in the beaufort sea : seasonal space and time scales,@xmath18 _ j. acoust .
* 82 * , 988997 ( 1987 ) .
jensen , w.a .
kuperman , m.b . porter , and h. schmidt , _ computational ocean acoustics _
( springer , new york , new york , 2011 ) , pp . 2728 . y. xie and d. m. farmer , @xmath17the sound of ice break - up and floe interaction,@xmath18 _ j. acoust
_ * 91 * , 22152231 ( 1992 ) .
stein , @xmath17interpretation of a few ice event transients,@xmath18 _ j. acoust .
soc . am . _
* 83 * , 617622 ( 1988 ) .
moore , k.m .
stafford , h. melling , c. berchok , .
wiig , k.m .
kovacs , c. lydersen , and j. richter menge , @xmath17comparing marine mammal acoustic habitats in atlantic and pacific sectors of the high arctic : year long records from fram strait and the chukchi plateau,@xmath18 _ polar biol .
* 35 * , 475480 ( 2012 ) .
g. spreen , l. kaleschke , and g.heygster @xmath17sea ice remote sensing using amsr - e 89 ghz channels,@xmath18 _ j. geophys .
res . _ * 113 * , c02s03 ( 2008 ) .
sohn and j.a .
hildebrand , @xmath17hydroacoustic earthquake detection in the arctic basin with the spinnaker array,@xmath18 _ b. seismol .
* 91 * , 572579 ( 2001 ) . r.h .
mellen and h.w .
marsh , @xmath17underwater sound in the arctic ocean,@xmath18 , accession number ad718140 .
navy underwater sound laboratory , new london , connecticut , 1965 .
a.r milne and j.h .
ganton , @xmath17ambient noise under arctic sea ice,@xmath18 _ j. acoust .
* 36 * , 855863 ( 1964 ) .
b. buck , @xmath17 preliminary under ice propagation models based on synoptic ice roughness,@xmath18 _ prl tr30 _ ( may 1981 ) .
seattle , washington .
dee , et al . , @xmath17the era - interim reanalysis : configuration and performance of the data assimilation system,@xmath18 _ q. j. r. meteorolog
. soc . _ * 137 * , 1477870x ( 2011 ) . | the soundscape in the eastern arctic was studied from april to september 2013 using a 22 element vertical hydrophone array as it drifted from near the north pole ( 8923n , 6235w ) to north of fram strait ( 8345n 428 w ) .
the hydrophones recorded for 108 minutes on six days per week with a sampling rate of 1953.125 hz .
after removal of data corrupted by nonacoustic flow related noise , 19 days throughout the transit period were analyzed .
major contributors include broadband and tonal ice noises , seismic airgun surveys , and earthquake @xmath0 phases .
statistical spectral analyses show a broad peak at about 15 hz similar to that previously observed and a mid frequency decrease at f@xmath1 .
the median noise levels reflect a change in dominant sources , with ice noises ( 200500 hz ) decreasing and seismic airgun surveys ( 1050 hz ) increasing as the array transited southward .
the median noise levels were among the lowest of the sparse reported observations in the eastern arctic , but comparable to the more numerous observations of western arctic noise levels . |
the fate of the two dimensional delocalized states ( ds s ) in a magnetic field became recently the subject of intensive experimental @xcite and theoretical @xcite studies .
the goal of these studies is to trace the position of ds s as magnetic field decreases .
more than a decade ago it was predicted that with decreasing magnetic field each ds departs from the center of the landau level ( ll ) and floats up in energy ( levitation scenario @xcite ) .
this prediction was derived from the non interacting scaling theory of the quantum hall effect @xcite .
since the position of a ds determines the boundary between the insulating and the quantum hall phases , such a floating implies that within some range of concentrations the electron gas undergoes an insulator metal insulator transition with decreasing magnetic field .
later it was argued that this behavior should persist in the presence of interactions @xcite .
experimental observation of the insulator metal insulator transition was reported by several groups @xcite .
it is obvious just from the electron
hole symmetry that the deviation of ds from the center of landau level is possible only due to the disorder induced ll mixing .
numerical simulations @xcite indeed support the levitation scenario .
analytical theory of the levitation was developed only for the region of magnetic fields where the departure of the ds from the center of ll is relatively small @xcite .
it was assumed @xcite that the random potential is smooth , so that the structure of electronic states is described by the network model of chalker and coddington @xcite . in this model ,
delocalization results from the tunneling of an electron through the saddle points of a random potential which are connected by equipotential lines .
it was demonstrated that the ll mixing changes _ on average _ the transmittancy of a saddle point in such a way that it becomes smaller than 1/2 for the energy at the center of the ll .
this means that to achieve the 1/2 average transmittancy , the energy should be shifted upwards , which is equivalent to the levitation . however , in a smooth potential the ll mixing is generally weak since it is associated with the large momentum transfer . short
range potential is much more effective in this respect .
that is why in this paper we consider the situation when a small portion of short range impurities is present in a sample in addition to a smooth potential . within the chalker
coddington model the shift in position of the ds results from the change in the transmittancy of a saddle point due to ll mixing .
then it is obvious that impurities , responsible for this shift , are those that fall in the vicinity of saddle points .
we assume the sign of the impurity potential to be random and show that the net effect of such impurities is the reduction of the transmittancy at a given energy , and , hence , the levitation of ds s , even if the mixing of ll s by a smooth potential is neglected .
the plan of the paper is the following . in sec .
[ sec : h ] we consider the motion of an electron in a strong magnetic field and a saddle point potential with a short range impurity located nearby .
we derive the system of equations for the motion of the guiding center using the procedure developed by fertig and halperin @xcite . in sec .
[ sec : r ] we demonstrate that coupling of ll s by an impurity leads to the renormalization of its effective strength . in sec .
[ sec : t ] a general expression for the transmission coefficient is derived and different limits are analyzed . in sec .
v the impurity induced change in the average transmittancy is studied . in sec .
vi the magnitude of levitation of ds s , which results from the reduction of the average transmittancy is , estimated .
the hamiltonian we consider has the form @xmath2 where @xmath3 is a potential of the saddle point ; @xmath4 is the short range potential of an impurity centered at the point @xmath5 .
@xmath6 is chosen in such a way that @xmath7 .
we choose a symmetric gauge , * a*=@xmath8 , where @xmath0 is the magnetic field . to calculate the transmission coefficient of the saddle point
we adopt the approach developed by fertig and halperin @xcite .
it was demonstrated @xcite that the hamiltonian @xmath9 can be separated into two parts describing respectively the motion of a guiding center and a cyclotron motion . the unitary transformation which provides this separation is as follows : @xmath10 where @xmath11 is the magnetic length ( @xmath12 stands for the cyclotron frequency ) .
the new operators satisfy the commutation relations @xmath13= \left[\frac{\partial}{\partial x},\frac{\partial}{\partial s } \right]=0 $ ] , @xmath14=\left[\frac{\partial}{\partial s } , s \right]=1 $ ] .
the parameter @xmath15 is determined by the equation @xmath16 where @xmath17 is the characteristic length defined as @xmath18 .
the condition that the potential is smooth can be quantitatively expressed as @xmath19 , which means that @xmath20
. then the solution of eq.([8 ] ) , can be written as @xmath21 as a result of the transformation ( 4)(7 ) , the hamiltonian @xmath9 takes the form @xmath22 where the energies @xmath23 in the limit @xmath20 are given by @xmath24 it is seen that the variable @xmath25 describes the motion of the guiding center whereas @xmath26 is responsible for the cyclotron motion .
we search for the solution of the schrdinger equation , @xmath27 , in the form @xmath28 where @xmath29 are the eigenfunctions of the harmonic oscillator @xmath30 the coefficients @xmath31 satisfy the system of equations @xmath32 with @xmath33 defined as @xmath34 first , it can be readily seen that , since the function @xmath35 is nonzero only when both arguments are smaller than @xmath36 , where @xmath37 is the radius of the impurity potential , one can take @xmath38 out of the integral at point @xmath39 .
we also see that after switching to the @xmath40 variables the impurity potential became an operator .
thus , to find its action on the function @xmath41 one should perform the fourier transformation @xmath42 where we again took into account that @xmath43 . as a result
the differential operator @xmath44 in the argument of @xmath35 can be replaced by @xmath45 , so that the right hand side takes the form @xmath46 now the integration over @xmath26 and @xmath47 can be easily performed using the short range character of @xmath35 .
this picks the values @xmath48 and @xmath39 and generates the factor @xmath49 . after that it is convenient to express the component @xmath50 in terms of @xmath41 using the inverse fourier transformation .
finally we obtain @xmath51 where we have introduced the notations @xmath52 the paremeter @xmath53 is determined as @xmath54 in the second equality we have substituted @xmath55 using the representation ( [ 15 ] ) .
( [ 19 ] ) , ( [ 21 ] ) form a closed system of equations describing the motion of the guiding center . in the absence of a short range impurity , @xmath56 , so that ( [ 19 ] ) reduces to the system of identical schrdinger equations for a particle in an inverted parabolic potential .
the transmission coefficient for this potential is well known : @xmath57 and is independent of the number of ll .
since we are interested in scattering of an electron with energy close to the center of some ll , say @xmath58 , by the saddle point potential , we can assume @xmath59 for any @xmath60 .
the condition @xmath19 guarantees that @xmath61 .
this means that each of eqs .
( [ 19 ] ) with @xmath60 contains a big parameter @xmath62 .
then for @xmath60 one can neglect the first two terms in ( [ 19 ] ) and get the solution @xmath63 substituting this solution into eq .
( [ 21 ] ) , enables us to express the constant @xmath53 in terms of only one unknown function @xmath64 @xmath65 once @xmath53 is determined , one can write a closed equation for @xmath66 @xmath67 where @xmath68 is defined as @xmath69 with @xmath70 given by @xmath71 we see that the parameter @xmath70 describes the effect of all other ll s on the motion of an electron on the level @xmath58 .
it is apparent from ( [ 30 ] ) that this effect reduces to the renormalization of the scattering strength @xmath72 .
important is that , since the impurity is short range , it causes a strong local mixing of ll s so that many levels with @xmath60 contribute to the renormalization constant @xmath70 .
indeed , the terms in the sum ( [ k ] ) fall off with the number of ll as @xmath73 . in the case
@xmath74 all the terms in the sum ( [ k ] ) have the same sign , so , strictly speaking , the sum diverges logarithmicaly .
the cutoff value of @xmath75 is determined by the following condition . for large enough numbers of ll
@xmath76 is a rapidly oscillating function , the typical period of oscillations being @xmath77 .
taking @xmath76 out of integral in ( [ 17 ] ) is justified only if this period is much larger than the dimensionless size of the impurity potential @xmath78 .
this leads to the cutoff value @xmath79 .
thus we obtain @xmath80 for @xmath81 the sum in ( [ k ] ) contains both positive and negative contributions .
positive terms are those with @xmath82 , while the terms with @xmath83 are negative . for large enough @xmath58
( but much smaller than @xmath84 ) we have @xmath85 thus the magnitude of the renormalization constant @xmath70 decreases with the number of ll .
in this section we will drop for simplicity the ll index of the transmission coefficient . without an impurity , the solutions of eq .
( [ 28 ] ) are the parabolic cylinder functions @xcite @xmath86 and @xmath87 , where @xmath88 . using the asymptotic form of @xmath89 @xmath90
one can see that it is the function @xmath86 that has the
right behavior at @xmath91 ( no incoming wave at @xmath92 ) .
the amplitude transmission coefficient , @xmath93 , emergies from the comparison of the amplitudes of the transmitted and the incoming waves in ( [ as1 ] ) , ( [ as2 ] ) @xmath94 it is easy to see that @xmath95 , where @xmath96 is given by ( [ 22 ] ) . with the right hand side in ( [ 28 ] ) present , it is convenient to search for a solution of eq .
( [ 28 ] ) in the form of a linear combination of the functions @xmath89 @xmath97 \ .\ ] ] the expressions for the coefficients @xmath98 are obtained by multiplying eq .
( [ 28 ] ) by @xmath99 and integrating over @xmath100 .
one has @xmath101 where @xmath102 and @xmath103 are defined as @xmath104 and the function @xmath105 is determined by the normalization condition @xmath106 using the asymptotics ( [ as1 ] ) , ( [ as2 ] ) it can be shown that @xmath107 upon substituting ( [ 44 ] ) into ( [ 43 ] ) , we obtain @xmath108.\end{aligned}\ ] ] for @xmath109 only the poles contribute to the integrals in eq .
( [ 49 ] ) . in the limit @xmath110
, @xmath111 is just a transmitted wave , whereas @xmath112 represents the combinations of waves traveling in both directions .
collecting the components corresponding to the transmission , we obtain the following general expression for the transmission coefficient @xmath113 \right|^2 .\ ] ] to find the constant @xmath103 one should multiply ( [ 49 ] ) by @xmath114 and integrate .
this will generate @xmath103 in the left - hand side ; the solution of the resulting equation for @xmath103 has the form @xmath115 } .\ ] ] with @xmath103 given by ( [ 47 ] ) , the final result for the transmission coefficient reads @xmath116 } { 1-\frac{\tilde{v}_0^r}{\sqrt{2}}\int \limits_{- \infty}^{\infty } \frac{d\varepsilon^{\prime } } { ( \varepsilon-\varepsilon^{\prime } ) } \
\alpha^{-1 } ( \varepsilon^{\prime } ) \left[\left|i^+(\varepsilon^{\prime } ) \right|^2 + \left|i^- ( \varepsilon^{\prime } ) \right|^2 \right ] } \right|^2 .\ ] ] the denominator of ( [ tg ] ) can be presented as @xmath117 , where the functions @xmath118 and @xmath119 are defined as @xmath120 \ , \\
\label{tetaf } \theta ( \varepsilon)&= & \frac{\pi}{\sqrt{2 } } \
\alpha^{-1 } ( \varepsilon ) \left[\left|i^+(\varepsilon ) \right|^2 + \left|i^- ( \varepsilon ) \right|^2 \right]\ , \end{aligned}\ ] ] where the symbol @xmath121 stands for principal part .
the meaning of the functions @xmath118 and @xmath119 can be understood in the following way .
it is known that without a smooth potential , a short range impurity pulls out a state from the center of a degenerate ll and forms a bound state @xcite . in the presence of a saddle point potential the energy position of the bound state is determined by the condition : @xmath122 . on the other hand , since the potential ( [ 2 ] ) `` bends '' the ll , the bound state becomes degenerate with the continuum of states at the ll and , thus , acquires a finite width which is described by the function @xmath123 .
it is seen that @xmath123 represents the sum of two terms , which correspond to the widths associated with the outcome to the left and to the right , respectively . when the impurity is located
exactly at the saddle point ( @xmath124 ) we have @xmath125 , so that ( [ tg ] ) reduces to @xmath126 ^ 2 } { ( 1-\tilde{v}_0^r \ \sigma(\varepsilon))^2 + ( \tilde{v}_0^r \ \theta(\varepsilon))^2}.\ ] ] for this particular location of the impurity , the functions @xmath118 and @xmath119 are , respectively , odd and even functions of energy .
we have evaluated these functions for the lowest ll .
note first that for @xmath127 and arbitrary @xmath128 the integrals @xmath129 and @xmath130 can be expressed in terms of the parabolic cylinder functions if one substitutes into ( [ iint ] ) the integral representation @xcite of @xmath131 .
then one gets @xmath132 since @xmath133 , where @xmath134 is the @xmath135function , we get the following expression for @xmath136 at @xmath124 @xmath137 then the functions @xmath119 and @xmath118 take the form @xmath138 @xmath139 @xmath123
is shown in fig .
1 together with the function @xmath118 calculated numerically .
the asymptotic behavior of @xmath140 is as follows : @xmath141 , for @xmath142 ; @xmath143 , for @xmath144 .
the energy dependence of the transmission coefficient for different amplitudes of the impurity potential , @xmath145 , is shown in fig .
2 . at zero energy @xmath140 vanishes
, so that ( [ t0 ] ) simplifies to @xmath146 fig .
3 shows the transmission coefficient at @xmath147 as a function of @xmath145 .
without an impurity , @xmath96 obeys the relation : @xmath148 which is the manifestation of the symmetry of the saddle point ( transmission of an electron with energy @xmath149 can be viewed as a reflection for an electron with energy @xmath150 ) . obviously ,
when the impurity is randomly positioned around the saddle point and the sign of @xmath68 is random , this symmetry relation should hold _ on average_. indeed , it can be demonstrated that for an arbitrary position of an impurity , @xmath151 , the transmission coefficient satisfies the relation @xmath152 to prove ( [ gs ] ) , one should rewrite ( [ tg ] ) in the form @xmath153 where the functions @xmath154 and @xmath155 are defined as @xmath156 , \\
\label{b } \lambda(\varepsilon)&=&\pi \sqrt{2 } \
\alpha^{-1 } ( \varepsilon ) \ i^+(\varepsilon)(i^- ( \varepsilon ) ) .^*\end{aligned}\ ] ] the function @xmath154 is real and turns to zero for an impurity located at @xmath124 ; @xmath157 is , generally speaking , a complex function .
it is obvious that @xmath158 .
one can verify that under the transformation @xmath159 the integrals @xmath129 and @xmath130 with an accuracy of phase factors are transformed into @xmath160 and @xmath161 respectively ; with the same accuracy the conjugated values @xmath162 and @xmath163 are transformed into @xmath164 and @xmath165 .
then the rules of transformation for the functions @xmath166 , and @xmath155 are as follows : @xmath167 . using these properties , the relation ( [ gtc ] )
can be easily proved .
the meaning of this relation is that the transmittancy of the saddle point , averaged over the amplitude @xmath145 , position @xmath151 , and energy @xmath149 with any _ symmetric _ distribution function , is equal to @xmath168 .
the averaging over @xmath149 has the meaning of averaging over the background value of the saddle point potential , @xmath169 , ( this value was set zero in ( [ 2 ] ) ) .
our main observation is that for a symmetric distribution of the `` bare '' amplitudes @xmath170 , the distribution of renormalized amplitudes , @xmath171 , which are defined by eq .
( [ 30 ] ) , is _
asymmetric_. the origin of the asymmetry is the ll mixing .
indeed , denote with @xmath172 the distribution function of @xmath170 ( we assume that @xmath173 ) . then using ( [ 30 ] ) one can easily find the distribution function of @xmath145 @xmath174 the degree of asymmetry of @xmath175 , due to the finite renormalization constant @xmath70 , is determined by the product @xmath176 , where @xmath177 is the width of the distribution function @xmath178 .
consider , for example , the lorentzian distribution : @xmath179 .
then @xmath175 takes the form @xmath180}.\ ] ] we see that @xmath181 is also a lorentzian , but it is centered at @xmath182 . for @xmath183 the shift of the center , @xmath184 , is relatively small as compared to the width @xmath177 .
however for @xmath185 we get a narrow peak at @xmath186 with the width @xmath187 .
this means that all the impurities , which were with equal probability repulsive and attractive before the renormalization , become effectively repulsive after the renormalization .
the asymmetry in @xmath145 makes the average transmittancy of the saddle point smaller than @xmath168 . to demonstrate it
, it is convenient to study the combination @xmath188 for the same @xmath145 .
moreover , since @xmath189 , the renormalized impurity strength @xmath190 is much larger than @xmath191 , so that in calculating this combination one can expand ( [ gtc ] ) with respect to the small parameter @xmath192 . using the transformation rules for the functions @xmath193 and @xmath155 , established above , we obtain @xmath194 where @xmath195 is defined as @xmath196 we see that it is the sign of the parameter @xmath197 that determines the sign of deviation of the average transmittancy from @xmath168 .
consider first the simplest case @xmath124 .
then we have @xmath198 , _ im_@xmath199 and _ re_@xmath200 .
hence , the parameter @xmath197 reduces to @xmath201.\ ] ] it follows from ( [ delt ] ) that @xmath197 is positive if the condition @xmath202 is met .
1 illustrates that this indeed is the case for any energy @xmath149 .
as the displacement @xmath151 increases , the parameter @xmath197 falls off due to the general decay of the integrals @xmath203 .
however , @xmath197 remains positive on average . the easiest way to see
it is to consider the limit of large displacements : @xmath204 . in this limit @xmath129 and @xmath130
acquire big phases .
this results in a big phase of the function @xmath205 .
consequently , _ re_@xmath155 and _ im_@xmath155 oscillate rapidly with the change of the impurity position ( roughly as @xmath206 and @xmath207 ) .
thus the contributions of the first two terms in ( [ delta ] ) average out .
when considering the third term , we note that @xmath208 depends only on the magnitude of the displacement ( it can be shown that @xmath209 ) , i.e. @xmath208 is constant if @xmath210 is constant .
the form of the function @xmath118 depends on the angular position of the impurity when @xmath210 is fixed , but , averaged over the angular position , @xmath118 is an odd function of energy .
one can check formally that @xmath211 changes sign when @xmath149 changes sign .
therefore , the product @xmath212 is positive on average .
thus we conclude that average @xmath197 is positive both for small and large displacements , and , consequently , the average transmittancy of the saddle point is diminished due to the impurity induced ll mixing .
in order to compensate for the reduction of the average transmittancy , the energy position of the delocalized state in a smooth potential shifts up with respect to the center of ll .
let us estimate this shift ( levitation ) . denote with @xmath213 the concentration of the impurities . clearly , to produce a significant effect on the transmittancy
, the impurity should be located within the interval of the order of magnetic length from the center of the saddle point ( @xmath214 ) .
the probability to find such an impurity is @xmath215 .
for @xmath214 we have @xmath216 in eq .
( [ exp ] ) .
thus , the magnitude of the reduction of the average transmittancy is of the order of @xmath217 . on the other hand
, we have established that if an impurity is strong enough , the ll mixing renormalizes its amplitude to @xmath218 . as a result ,
in the presence of impurities , the delocalized state corresponds to such an energy for which , in the absence of impurities , the average transmittancy of the saddle point exceeds @xmath168 by @xmath219 . since without an impurity the transmission coefficient @xmath220 has the energy scale @xmath221 , determined by eq .
( [ 12 ] ) , we get the following estimate for the magnitude of the levitation @xmath222 @xmath223 it is instructive to compare the levitation with the spacing , @xmath224 , between the ll s .
one gets @xmath225 where we have used the expression ( [ kaa ] ) for @xmath70 . in our consideration
we have assumed that there is only a single impurity near the the saddle point , which implies that @xmath226 .
we have also assumed that @xmath227 .
then the above estimate shows that under the assumptions adopted the relative levitation is small .
it also shows that the magnetic field dependence of the levitation is @xmath228 . in numerical simulations
@xcite the authors restricted the study to the two lowest ll s .
they found that , due to the ll mixing , the lower ds shifts up whereas the upper ds shifts down in energy ( `` attraction '' of ds s ) .
our result is consistent with this observation .
indeed , if only two ll s are considered , the renormalization constant , @xmath70 , would be positive for the lower level and negative for the upper level .
note in conclusion , that our main result
the reduction of the transmittancy due to the ll mixing , was derived in the limit of the strong renormalization of the impurity potential .
the criterion for that , formulated in the previous section , is @xmath229 , @xmath177 being the typical value of the dimensionless potential @xmath170 . using eqs .
( [ kaa ] ) and ( [ 20 ] ) this criterion can be rewritten in terms of the `` bare '' amplitude and size of the impurity potential as @xmath230 . on the other hand , in our approach
we have treated the impurity as point
like @xcite and , thus , neglected the change of the electron wave function within the radius @xmath36 .
this is justified if the condition @xmath231 is met .
the second condition seems to restict the validity of the theory to the region @xmath232 .
it appears , however , that the condition @xmath233 is not relevant .
we address this question in the appendix and show that as soon as @xmath234 , it renormalizes to @xmath235 ( which is equivalent to @xmath236 ) , as we have assumed .
since the presence of the saddle point potential is not important for the renormalization procedure , we will trace the renormalization for the case when @xmath169 is absent . in this case
the effect of an impurity is just the formation of the bound states which split off the ll s . in the symmetric gauge with an impurity at the origin
, the schrdinger equation allows the separation of variables .
the radial wave function , @xmath237 , for a zero orbital moment , satisfies the equation @xmath238r=0.\ ] ] we will analyze the solutions of ( [ a1 ] ) , corresponding to high ll s , for which the calculations can be performed semiclassically without invoking the hypergeometric function .
then the bohr - sommerfeld condition for the energy levels , @xmath239 , reads @xmath240 where @xmath241 is an additional phase shift acquired at the origin ( @xmath242 for @xmath56 ) .
this shift should be found by matching the solutions at @xmath243 and at @xmath244 .
important is that for large @xmath245 the magnetic potential , @xmath246 , in ( [ a1 ] ) comes into play only at large @xmath247 . for smaller @xmath248
one can neglect the magnetic potential .
then the solution of ( [ a1 ] ) ( which is finite at @xmath249 ) can be written as @xmath250 where @xmath251 and @xmath252 are , respectively , the bessel and the neumann functions of order zero ; @xmath253 , and @xmath254 .
the constants @xmath255 determine the phase shift @xmath241 .
indeed , at @xmath256 the functions @xmath251 and @xmath252 oscillate as @xmath257 and @xmath258 , so that @xmath259 .
the continuity conditions for @xmath237 and @xmath260 at @xmath261 can be written as @xmath262 solving this system yields the following expression for @xmath241 @xmath263 the case we are interested in is @xmath264 .
this allows to use the small argument asymptotics for the functions depending on @xmath265 .
one gets @xmath266,\ ] ] where @xmath267 is the euler constant . in the limit of a point
like impurity , considered by gredeskul and azbel@xcite , we have @xmath231 ( which is equivalent to @xmath268 ) , and eq.([a6 ] ) simplifies to @xmath269.\ ] ] we see that @xmath6 enters into the phase shift in the same `` renormalized '' form as in eq .
( [ 30 ] ) for @xmath145 . if the renormalization is weak , we get @xmath270 , which leads to the standard result , @xmath271 , for the binding energy .
as @xmath6 exceeds @xmath272 ( but remains smaller than @xmath273 ) we get from ( [ a7 ] ) @xmath274 so that @xmath6 disappears from the phase shift and , correspondingly , from the binding energy .
this is equivalent to the conclusion that @xmath145 renormalizes to @xmath275 .
most importantly , eq.([a8 ] ) remains valid when @xmath6 gets _ much larger _ than @xmath273 , and the approach , adopted in sec .
[ sec : h ] , is not applicable any more .
indeed , as @xmath276 increases , the combination @xmath277 becomes either oscillating ( for @xmath278 ) or monotonously increasing ( for @xmath279 ) function with a typical magnitude much larger than unity .
then the result ( [ a8 ] ) immediately follows from ( [ a6 ] ) .
i. glozman , c. e. johnson , and h. w. jiang , phys .
* 74 * , 594 ( 1995 ) .
i. glozman , c. e. johnson and h. w. jiang , phys .
b * 52 * , 14348 ( 1995 )
. i. glozman , c. e. johnson and h. w. jiang , in _ proceedings of the 6th international conference on hopping and related phenomena _ , ed . by o. millo and z. ovadyahu , jerusalem ( 1995 ) .
s. v. kravchenko , w. mason , j. e. furneaux , and v.m .
pudalov , phys .
lett . * 75 * , 910 ( 1995 ) .
j. e. furneaux , s. v. kravchenko , w. e. mason , g. e. bowker , and v. m. pudalov , phys .
b * 51 * , 17227 ( 1995 ) . t. v. shahbazyan and m. e. raikh , phys . rev
. lett . * 75 * , 304 ( 1995 ) .
d. z. liu , x. c. xie , and q. niu , preprint ( cond - mat/9504010 ) .
v. kagalovsky , b. horovitz , and y. avishai , in _ proceedings of the 6th international conference on hopping and related phenomena _ , ed . by o. millo and z. ovadyahu , jerusalem ( 1995 ) .
d. e. khmelnitskii , phys .
105a * , 182 ( 1984 ) .
r. b. laughlin , phys .
lett . * 52 * , 2304 ( 1984 ) .
h. levine , s. b. libby , and a. m. m. pruisken , phys .
lett . * 51 * , 1915 ( 1983 ) .
s. kivelson , d .- h .
lee , and s .- c .
zhang , phys .
b * 46 * , 2223 ( 1992 ) .
h. w. jiang , c. e. johnson , k. l. wang , and s. t. hannahs , phys .
* 71 * , 1439 ( 1993 ) . c. e. johnson and h. w. jiang , phys .
b * 48 * , 2823 ( 1993 ) .
t. wang , k. p. clark , g. f. spencer , a. m. mack , and w. p. kirk , phys .
* 72 * , 709 ( 1994 ) .
r. j. f. hughes , j. t. nicholls , j. e. f. frost , e. h. linfield , m. pepper , c. j. b. ford , d. a. ritchie , g. a. c. jones , e. kogan , and m. kaveh , j. phys
. condens .
matter * 6 * , 4763 ( 1994 ) .
k. yang and r. n. bhatt , preprint ( cond - mat/9510066 ) .
t. ando , j. phys .
* 53 * , 3126 ( 1984 ) . j. t. chalker and p. d. coddington , j. phys .
c * 21 * , 2665 ( 1988 ) .
h. fertig and b. i. halperin , phys .
b * 36 * , 7969 ( 1987 ) . , edited by a. erdlyi , ( mcgraw - hill , new york , 1953 ) , vol .
t. ando , j. phys .
. jpn . * 36 * , 1521 ( 1974 ) .
e. m. baskin , l. n. magarill , and m. v. entin , zh .
75 * , 723 ( 1978 ) [ sov .
jetp * 48 * , 365 ( 1978 ) ] .
v. a. geiler and v. a. margulis , teor . mat .
58 * , 461 ( 1984 ) [ theor .
phys . * 58 * , 302 ( 1984 ) ] .
s. a. gredeskul , y. avishai , and m. ya .
azbel , europhys . lett .
* 21 * , 489 ( 1993 ) .
m. ya . azbel and b. i. halperin , phys .
b * 52 * , 14098 ( 1995 ) . s. a. gredeskul and m. ya .
azbel phys .
b * 49 * , 2323 ( 1994 ) . | the effect of a short range impurity on the transmission through a saddle point potential for an electron , moving in a strong magnetic field , is studied .
it is demonstrated that for a random position of an impurity and random sign of its potential the impurity induced mixing of the landau levels diminishes _ on average _ the transmission coefficient .
this results in an upward shift ( levitation ) of the energy position of the delocalized state in a smooth potential . the magnitude of the shift is estimated
. it increases with decreasing magnetic field @xmath0 as @xmath1 . |
the skyrme model is an effective theory of qcd constructed by pion fields alone .
baryons are incorporated as topological solitons in this model , called skyrmions , and the topological charge corresponds to the baryon number @xmath1 @xcite .
skyrme introduced a hedgehog ansatz for the pion fields to obtain @xmath2 spherically symmetric soliton solutions .
the detailed analysis for the property of the @xmath2 skyrmion as a nucleon was performed in ref . upon quantisation of the collective coordinate .
later braaten and carson obtained @xmath0 skyrmions numerically and showed that @xmath0 skyrmions are axially symmetric .
interestingly , multi - skyrmions with @xmath3 exhibit various discrete symmetries analogously to multi - bps monopoles .
the einstein - skyrme system in which skyrme fields couple with gravity was first considered by luckock and moss where the schwarzschild black hole with skyrme hair was obtained numerically .
this is a counter example of the no - hair conjecture for black holes .
they observed that the presence of the horizon in the core of skyrmion unwinds the skyrmion , leaving fractional baryon charge outside the horizon .
the full einstein - skyrme system was solved later to obtain spherically symmetric black holes with skyrme hair and regular gravitating skyrmions . in this paper
we shall study regular and black hole solutions of @xmath0 skyrmions with axisymmetry .
the recent study by hartmann , kleihaus and kunz showed the einstein - yang - mills - higgs theory possess axially symmetric monopole and black hole solutions .
therefore , we shall follow their numerical technique to solve the einstein - skyrme model with axial symmetry .
computing the energy density and charge density , we shall show how gravity affects to the skyrmion .
after a long history , the skyrme model was proved to be an effective theory of qcd by e. witten in ref . @xcite . at low energy the @xmath4 chiral symmetry of the strong interaction is broken spontaneously and the skyrme lagrangian respects this symmetry . the skyrme model
coupled to gravity is defined by the lagrangian , @xmath5 where @xmath6\left[u^{-1}\partial_{\rho}u , u^{-1 } \partial_{\sigma}u\right]\right ) \label{ls } \\ { \cal l}_g & = & \frac{1}{16\pi g}r \label{lg}\end{aligned}\ ] ] and @xmath7 is an @xmath8 valued scalar field of pions .
@xmath9 mev is the pion decay constant . to solve the field equations , ansatz for the metric and skyrme fields are required .
let us introduce an ansatz for the metric given in ref .
@xmath10 where @xmath11 .
the axially symmetric skyrme fields can be parameterised by @xmath12 with @xmath13 and @xmath14 .
the integer @xmath15 corresponds to the winding number of the skyrme fields and for @xmath0 we have @xmath16 . in terms of @xmath17 and @xmath18
, the lagrangian takes the form @xmath19 where @xmath20 \\ { \cal l}_s^{(2)}&=&-\frac{1}{2a^{2}r^{2}}\left(\frac{f}{m}\right)^{2 } \biggl[\left(\partial_{r}f\partial_{\theta}\theta-\partial_{\theta}f \partial_{r}\theta\right)^{2 } \\ & & + \frac{n^{2}}{\sin^{2}\theta}\frac{m}{l}\left\{(\partial_{r}f)^{2 } + \frac{1}{r^{2}}(\partial_{\theta}f)^{2}\right\ } \sin^{2}\theta \\ & & + \frac{n^{2}}{\sin^{2}\theta}\frac{m}{l } \left\{(\partial_{r}\theta)^{2}+\frac{1}{r^{2}}(\partial_{\theta}\theta)^{2 } \right\}\sin^{2}f\sin^{2}\theta \biggr]\sin^{2}f \,\ , .
\label{}\end{aligned}\ ] ] the baryon current in curved spacetime is obtained by taking the spacetime covariant derivative @xmath21 , @xmath22 the baryon number then is given by integrating @xmath23 over the hypersurface @xmath24 , @xmath25 where @xmath26 and @xmath27 are the values at the inner and outer boundary respectively .
in flat spacetime we have @xmath28 which gives @xmath0 . in the presence of a black hole
, the integration should be performed from the horizon to infinity , which change the values of @xmath29 and allow the @xmath1 to take a fractional value of less than two .
this situation can be interpreted as the black hole absorbing a skyrmion .
the energy density is given by the time - time component of the stress - energy tensor @xmath30 + \frac{1}{2a^{2}r^{2}}\frac{f^{2}}{m^{2 } } \left[\frac{}{}(\partial_{[r}f\partial_{\theta]}\theta)^{2 } \right .
\nonumber \\ & & \left .
+ \frac{n^{2}}{\sin^{2}\theta}\frac{m}{l } \left\{(\partial_{r}f)^{2}+\frac{1}{r^{2}}(\partial_{\theta}f)^{2 } \right\}\sin^{2}\theta \right .
\nonumber \\ & & \left .
+ \frac{n^{2}}{\sin^{2}\theta}\frac{m}{l } \left\{(\partial_{r}\theta)^{2}+\frac{1}{r^{2}}(\partial_{\theta}\theta)^{2 } \right\}\sin^{2}f\sin^{2}\theta\right]\sin^{2}f .
\end{aligned}\ ] ]
let us consider the boundary conditions for the chiral fields and metric functions . at the origin @xmath31 , the metric functions are regular , which requires the boundary conditions @xmath32 at infinity @xmath33 , the metric is asymptotically flat .
thus we have @xmath34 the boundary conditions for the skyrme fields at @xmath35 are given by @xmath36 the axially symmetric condition leads to the boundary conditions on the axes @xmath37 as @xmath38 at the horizon @xmath39 , the time - time component of the metric satisfies @xmath40 regularity of the metric at the horizon requires @xmath41 the boundary conditions for @xmath42 and @xmath43 at the horizon are obtained by expanding them at the horizon and inserting into the field equations which are derived from @xmath44 and @xmath45 respectively .
thus we have @xmath46 the condition that the spacetime is asymptotically flat requires @xmath47 the boundary conditions for @xmath17 and @xmath48 at infinity remain the same as in flat spacetime @xmath49 for the solution to be axially symmetric , we have @xmath50 regularity on the axis and axisymmetry impose the boundary conditions on @xmath17 and @xmath18 as @xmath51 under these boundary conditions , we shall solve the einstein equations and the matter field equations numerically .
let us introduce dimensionless coordinate and coupling constant @xmath52 then the free parameters are the horizon @xmath53 and the coupling constant @xmath54 for black hole solutions while for regular solutions the @xmath54 is the only free parameter
. we shall take @xmath55 as decoupling of gravity from the matter , effectively @xmath56 . in figs .
[ fig : reg_ed0 ] , [ fig : reg_ed3 ] are the energy densities per unit @xmath57 and @xmath58 @xmath59 of the regular solutions with @xmath60 , @xmath61 respectively . as @xmath62 becomes larger , the energy density gets concentrated in a smaller region .
thus the gravity influences for the matter to be more compact and denser .
the shape remains torus irrespective of the value of @xmath63 .
we show the baryon density per unit @xmath57 and @xmath64 @xmath65 in fig .
[ fig : reg_b0 ] .
the shape is similar to the energy density , but its dependence on the @xmath62 is rather small .
note that the energy density and the baryon density both vanish at @xmath66 .
this can be readily checked by inserting the skyrme functions expanded around @xmath67 into @xmath68 and @xmath69 respectively .
we found that there exists no regular solution for @xmath70 . in figs .
[ fig : ed0 ] , [ fig : ed15 ] are the energy densities of the black hole solutions with @xmath60 , @xmath71 respectively . as @xmath62 becomes larger , the energy density becomes smaller and sparse .
this can be interpreted that the black hole absorbs more skyrmions for a larger coupling constant .
the shape is slightly distorted in the background of the black hole so that one can see the spherically symmetric horizon in the center of the skyrmion .
[ fig : b0 ] is the baryon density around the black hole .
as in the case of the regular solution , the dependence of the baryon density on the value of the coupling constant is small . it can be checked that the energy and baryon density vanish at @xmath72 in the same manner as the regular solution . inserting the metric functions as well as the skyrme functions expanded around the horizon instead
, one can also see that the energy and baryon density vanish at the horizon .
the domain of existence of the black hole solution is shown in fig . [
fig : domain ] . there exist minimum and maximum value of @xmath53 and @xmath54 beyond which no black hole solutions exist .
therefore the regular skyrmion solutions can not be recovered from the black hole solutions by taking the limit of @xmath73 unlike the case of @xmath2 . in fig .
[ fig : baryon_number ] is the dependence of the baryon number on @xmath53 .
one can see that the baryon number decreases as the black hole grows in size .
this figure confirms that the baryon number is no longer conserved due to the black hole absorbing the skyrmion .
we have obtained static axially symmetric regular and black hole skyrmions numerically . as is shown in figs . [
fig : reg_ed0 ] , [ fig : reg_ed3 ] , [ fig : ed0 ] , [ fig : ed15 ] , the energy density is toroidal in shape for the regular and black hole solutions .
however , in the black hole case , the shape is distorted around the spherically symmetric horizon .
the energy density depends on the coupling constant . for the regular solution ,
as the coupling constant becomes larger , the energy density distribute in a smaller region and become denser . on the other hand , for the black hole solution , the energy density becomes sparse due to the black hole absorbing more skyrmions .
the baryon density shows a similar shape with the energy density and its dependence on the value of the coupling constant is rather small .
however , one can see from fig .
[ fig : baryon_number ] that it depends on the horizon size such that the baryon density decreases in increase of @xmath53 .
since there are minimum values of @xmath53 , the black hole solution can not recover a regular solution by taking the limit of @xmath74 unlike the @xmath2 spherically symmetric case .
as our further works , it will be interesting to consider skyrmion black holes with @xmath75 which have discrete symmetries .
the inclusion of gauge fields may also be possible to study the interaction between a monopole black hole and a deuteron .
extending the model to higher dimensions will be also an exciting problem , leading to the study of deuteron black holes which may be observed in lhc in future .
maeda k. , tachizawa t. , torii t. and maki t. , stability of nonabelian black holes and catastrophe theory , _ phys .
lett . _ , 1994 , 72 , 450453 ; torii t. , maeda k. and tachizawa t. , nonabelian black holes and catastrophe theory 1 .
neutral type , _ phys .
_ , 1995 , d51 , 15101524 . | it has been known that a @xmath0 skyrmion is axially symmetric .
we consider the skyrme model coupled to gravity and obtain static axially symmetric regular and black hole solutions numerically .
computing the energy density of the skyrmion , we discuss the effect of gravity to the energy density and baryon density of the skyrmion . |
gaussian states , such as coherent states , squeezed states , and two - mode squeezed states , are of fundamental to describe basic nature of quantum mechanical phenomena .
they play a central role in experiments of quantum optics and quantum information science @xcite .
gaussian states are characterized by their covariance matrices with the first and second order moments of canonical quadrature variables .
similarly , gaussian operations are characterized by the transformation of the covariance matrices , and there have been various experimental demonstrations of quantum information processing within the gaussian framework @xcite . in turn
, it has been known that non - gaussian resource is necessary to implement a universal quantum computation @xcite .
this fact strongly motivates us to study the property of non - gaussian quantum states .
there have been many works to investigate the properties of non - gaussian wave functions @xcite and non - gaussian entangled states @xcite .
an outstanding example of non - gaussian quantum states is a single photon state .
generation method of the single photon state and its non - gaussian properties have been widely investigated not only for its primitive quantum nature but also for potential application in quantum information science @xcite .
quantum measurements to manipulate a few number of photons have been experimentally demonstrated @xcite and such methods are useful to generate non - gaussian entanglement @xcite .
entanglement of gaussian states has been successfully described by their covariance matrices @xcite .
on the other hand , there have been many proposed entanglement criteria based on the measurements of higher order moments of canonical quadrature variables @xcite .
generally , the criteria with higher order moments are thought to be powerful to detect non - gaussian entanglement , however , the criteria based on the covariance matrices are sensitive to some of non - gaussian entangled states .
thus , it is not so clear in what condition we need higher order moments for entanglement verification .
hence , it would be valuable to investigate limitations of entanglement criterion with the lower order moments of the covariance matrix and show the advantage of higher order criteria @xcite .
thereby , the following two steps would be important : ( i ) to identify a set of non - gaussian entangled states whose inseparability can not be verified by the measurement of the covariance matrix .
( ii ) to find higher order entanglement criterion which can verify the inseparability of these states .
the photonic number states are eigenstates of a harmonic oscillator and their properties are simply described by using the annihilation and creation operators .
such a formalism enables us to describe an orthonormal basis consist of entangled states @xcite .
it is likely that such an orthonormal basis gives an insightful example to proceed the step ( i ) . to proceed the step ( ii ) ,
there have been a sequence of works to present the separable conditions based on su(2 ) and su(1,1 ) commutation relations @xcite .
these conditions can be derived from the uncertainty relations for the operators described by the products of canonical variables , and are thought to be the lowest order criteria to go beyond the gaussian framework . in this paper , we consider two classes of non - gaussian entangled states generated from the product of the number states by using two - mode gaussian unitary interactions .
we show that , for many of these states , the covariance matrix is compatible with the covariance matrix of separable gaussian states and their inseparability can not be verified by the measurements of the first and second moments of canonical variables .
we also identify a couple of continuous - variable entanglement criteria with higher order moments to verify the inseparability of these states .
this paper is organized as follows .
we investigate the property of the covariance matrix for the non - gaussian entangled states in sec .
we consider the entanglement criteria with higher order moments for the non - gaussian entangled states in sec .
the results are summarized in sec iv .
in this section we apply the separable condition based on the covariance matrix of the density operator to two families of non - gaussian entangled states , and show limitations on the detection of entanglement from the measurement of the covariance matrix .
we consider a two - mode system @xmath0 described by quadrature variables with the canonical commutation relations @xmath1= [ \hat x_b , \hat p_b]=i$ ] .
let us define a vector form for the set of quadrature variables as @xmath2 the covariance matrix of a two - mode density operator @xmath3 is defined by @xmath4^t \rangle = : \left ( \begin{array}{cc } a & c \\ c^t & b \\ \end{array } \right ) , \label{covama}\end{aligned}\ ] ] where @xmath5 denotes the expectation value of the observable @xmath6 and @xmath7 .
the submatrices @xmath8 , @xmath9 , and @xmath10 are real symmetric 2-by-2 matrices .
their determinants and @xmath11 are invariant under the local gaussian unitary transformations . in terms of these determinants ,
the separable criterion @xcite is given by @xmath12 this is the necessary and sufficient condition for separability of two - mode gaussian states . for non - gaussian states ,
this is a necessary condition for separability .
hence , we may define a class of purely non - gaussian entangled states by the set of entangled states which fulfill the condition of eq .
( [ sep1 ] ) .
an important property of this class of states is that their covariance matrices can not be transform into the covariant matrices of the entangled gaussian states by local gaussian operations and classical communications . in the next two subsections we consider two families of entangled states given by acting gaussian unitary interactions on the product of number states as in fig .
[ fig : sep - fig1.eps ] .
we will specify the regime of parameters which indicates the purely non - gaussian entangled states in each family . and
@xmath13 specify the action of the two - mode squeezer and the action of the beamsplitter , respectively ( see main text ) .
] let us write the annihilation operators of mode @xmath8 and mode @xmath9 by @xmath14 and @xmath15 .
this implies the bosonic commutation relations @xmath16=1 $ ] , @xmath17=1 $ ] , and @xmath18= [ a , b^\dagger ] = 0 $ ] .
the number states of the local modes are defined by @xmath19 and @xmath20 together with the conditions for the vacuums , @xmath21 and @xmath22 . by using this standard notation ,
the two - mode - squeezed ( tms ) vacuum state in the schmidt decomposed form is defined by @xmath23 .
the tms vacuum state is a gaussian state and it is known that any two - mode pure gaussian state can be transformed into the form of eq .
( [ deftmsv ] ) by local gaussian unitary operations .
let us define the annihilation operators crossing over the separation of @xmath8 and @xmath9 by @xmath24 where the unitary operator @xmath25 is the two - mode squeezing operator .
the new field operators fulfill the bosonic commutation relations @xmath26 = [ \hat b_\xi , \hat b_\xi ^\dagger ] = 1 $ ] and @xmath27=[\hat a_\xi,\hat b_\xi^\dagger]= 0 $ ] .
the tms vacuum state corresponds to the vacuum of the new field operators as we have @xmath28 and @xmath29 .
the tms number state @xcite is defined by the number state of the non - local modes as @xmath30 .
it is entangled whenever @xmath31 and its wave function is non - gaussian except for @xmath32 @xcite . from eqs .
( [ defa ] ) and ( [ basisi ] ) we can confirm the relation @xmath33 .
this implies that the tms number states can be generated by acting the two - mode squeezing operation on the product of number states as in fig .
[ fig : sep - fig1.eps](a ) .
the quadrature moments for the tms number states are routinely calculated by using the following relations on the annihilation and creation operators : @xmath34 the elements of the covariance matrix are determined to be @xmath35 where we write @xmath36 .
this implies an explicit formula of the quantity @xmath37 in eq .
( [ sep1 ] ) as @xmath38 from the condition @xmath39 we can see that the final term is negative , i.e. , @xmath40 .
hence , the separable inequality @xmath41 in eq .
( [ sep1 ] ) is violated when @xmath42 the parameter regime of this inequality can be depicted in fig .
[ fig : ens - sep - fig1.eps ] . from eq .
( [ joken1 ] ) , we can specify the subset of the tms number states whose inseparability can not be confirmed by the measurement of the covariance matrix .
we can see that the majority of the tms number states belong to the class of the purely non - gaussian entangled states .
, @xmath43 ) assigns a tms number state @xmath44 . in the gray regime , eq .
( [ joken1 ] ) is fulfilled , and entanglement of the corresponding tms number states can be verified by the measurement of the covariance matrix . on the outside ( and the boundary ) of this regime , the covariance matrix of @xmath45 has to be compatible with a separable gaussian state , and its inseparability can not be confirmed by the measurement of the covariance matrix .
here , we set the parameter @xmath46 in eq .
( [ joken1 ] ) . ]
let us consider the beam - splitter transformation described by @xmath47 where @xmath48 is the unitary operator of the beamsplitter and we assume @xmath49 .
the field operators fulfill the bosonic commutation relations @xmath50 = [ d_r , d_r ^\dagger ] = 1 $ ] and @xmath51= [ c_r , d_r^\dagger]= 0 $ ] . in the case of
@xmath52 we have the half - beamsplitter transformation @xmath53 in the limit of @xmath54 , there is no interaction between the modes and we have a trivial transformation @xmath55 whereas we have another trivial transformation @xmath56 in the limit of @xmath57 up to the phase factor .
the beam - splitter transformation produces another family of orthonormal states [ see , fig .
[ fig : sep - fig1.eps ] ( b ) ] , @xmath58 , the state @xmath59 is a separable gaussian state .
except for this case the bs number states are non - gaussian and entangled as we prove in the following .
let us assume the non - zero - photon case of @xmath60 . recall that any gaussian states are non - orthogonal with each other . since , a non - zero - photon state @xmath61 is orthogonal to the vacuum state from the orthonormal relation @xmath62 it can not be a gaussian state . from eqs .
( [ defc ] ) and ( [ p - basisi ] ) , we can verify that the bs number state @xmath63 has schmidt decomposition with the schmidt basis @xmath64 and the schmidt rank is at most @xmath65 .
it is direct to see that the coefficients of the edges of the schmidt basis @xmath66 and @xmath67 is non - zero .
this implies that one can find entanglement from the subspace spanned by @xmath68 and @xmath69 .
hence , the bs number states are non - gaussian entangled states except for the case of @xmath70 .
the inseparability of the bs number states can be also proven from the theorem for beam - splitter entangler @xcite ( see also @xcite ) .
note that a bs number state is described by a finite number of schmidt bases , and it is thought to be an entity of discrete - variable systems . on the contrary ,
schmidt rank of the tms number states is infinite , and they are thought to be entities of purely infinite - dimensional systems .
the quadrature moments for the bs number states are readily calculated by using the following relations on the annihilation and creation operators , @xmath71 the elements of the covariance matrix @xmath72 of eq .
( [ covama ] ) are determined to be @xmath73 where @xmath74 . from these relations and eq .
( [ sep1 ] ) , we have @xmath75 this implies that the separable condition @xmath41 of eq .
( [ sep1 ] ) is always satisfied for the bs number states .
( this result is also given by the fact @xmath76 with lemma of ref .
@xcite , that is , gaussian states with @xmath77 are separable . ) .
hence , they are classified to the purely non - gaussian entangled states
. it might be interesting that a relation similar to eq .
( [ joken1 ] ) can be derived for the bs number states when we consider a higher order entanglement condition given by hillery and zubairy @xcite : @xmath78 for the bs number states of eq .
( [ p - basisi ] ) we have @xmath79 \right\ } , \nonumber \\ { \langle a b^\dagger \rangle}&= & \frac{r}{1 + |r|^2 } ( m - n ) .
\label{eq18}\end{aligned}\ ] ] then , the entanglement condition of eq .
( [ jokenhz ] ) turns out to be @xmath80 this is an inverse proportional condition on the photon numbers of entangled modes for the inseparability similar to eq .
( [ joken1 ] ) [ see also , fig . [
fig : ens - sep - fig1.eps ] ] .
the condition of eq .
( [ nm - qua ] ) is satisfied only for the cases of @xmath81 with @xmath82 .
this is an advantage of the measurements of higher order moments for detection of the purely non - gaussian entanglement although it can detect the entanglement of a rather small portion of the bs number states .
now , it has turned out that the measurement of the covariance matrix is not useful to detect the entanglement for a significant portion of the photonic non - gaussian entangled states .
we thus consider higher order criteria in the subsequent section .
in this section we consider the separable conditions based on the operators that form su(2 ) algebra and su(1,1 ) algebra @xcite .
we show that the measurements of quadrature moments to the fourth order are sufficient to verify the present non - gaussian entangled states except for a small portion of the bs number states .
we also show that the entanglement of this small portion is detectable by the measurements of further higher order moments based on the idea of ref . @xcite .
let us define the set of operators @xmath83 where @xmath84 and @xmath85 are number operators of the local modes .
the operators @xmath86 with @xmath87 form the su(2 ) algebra while the operators @xmath88 with @xmath87 form the su(1,1 ) algebra @xcite .
the commutation relation @xmath89= ik_x$ ] implies the uncertainty relation @xmath90 . from its partial transposition
we have the separable condition in eq .
( 10 ) of ref .
@xcite @xmath91 { \langle \delta^2 k_z \rangle } \ge \frac{1}{4 } | { \langle j_x \rangle}| ^2 , \label{10sun11}\end{aligned}\ ] ] where the partial transposition of the second system with respect to the number basis is calculated by the relation @xmath92 . similarly , starting from the uncertainty relation @xmath93 we have another separable condition @xmath94 { \langle \delta^2 j_z \rangle } \ge \frac{1}{4 } | { \langle k_x \rangle}| ^2 . \label{m10sun11}\end{aligned}\ ] ] in the next two subsections we investigate the role of these separable conditions on the detection of the non - gaussian entanglement .
the entanglement of the present non - gaussian states is induced by the unitary operators @xmath95 in eq .
( [ defa ] ) and @xmath48 in eq .
( [ defc ] ) , and the hamiltonians can be connected with a partial transposition as it maps @xmath96 .
hence , it might be insightful to find general relationships between the non - gaussian entanglement and the separable conditions with the terms in the hamiltonian .
the tms number states of eq .
( [ basisi ] ) are simultaneous eigenstates of the number operators @xmath97 and @xmath98 . from the definition of @xmath99 in eq .
( [ def211 ] ) and the relations in eq .
( [ defa ] ) we have @xmath100 this implies that the tms number states are eigenstates of @xmath99 and @xmath101 . in turn , we can write @xmath102.\end{aligned}\ ] ] then , its expectation value for the tms number state @xmath44 of eq .
( [ basisi ] ) is calculated to be @xmath103 this implies @xmath104 .
consequently , for the tms number states of eq .
( [ basisi ] ) , the left - hand side of eq .
( [ m10sun11 ] ) is zero and right - hand side of eq .
( [ m10sun11 ] ) is positive .
hence , we can confirm the inseparability of the tms number states from the violation of eq .
( [ m10sun11 ] ) . since the separable condition of eq .
( [ m10sun11 ] ) is violated by the tms number state @xmath44 for any integers @xmath105 , @xmath43 , and @xmath106 , it might be valuable to emphasize the mechanism of the entanglement detection .
on one hand , the tms number states exhibit strong classical correlation on the photon number as we have @xmath107 . on the other hand , they show coherence between the ( schmidt ) bases @xmath108 as the pair creation and annihilation of the photons occur coherently .
this keeps the term @xmath109 non - zero .
therefore , we can say that the strong classical correlation and the coherence on the schmidt bases signify the existence of entanglement .
note that there is a different entanglement criterion that can verify the entanglement of the tms number states based on the measurement of quadrature moments to the fourth order @xcite .
this criterion is formulated to verify the entanglement of the tms number state when the parameter @xmath110 is specified .
note also that test of a sequence of high order separable conditions for tms number states have been reported in ref .
@xcite . the bs number states of eq .
( [ p - basisi ] ) are simultaneous eigenstates of the number operators @xmath111 and @xmath112 . from the relations in eq .
( [ defc ] ) and the definition of @xmath113 in eq .
( [ def211 ] ) we have @xmath114 hence , the bs number states are eigenstates of @xmath113 , and we have @xmath115 . in turn , from eq .
( [ eq18 ] ) and definition of @xmath116 we have @xmath117 for the bs number state @xmath118 of eq .
( [ p - basisi ] ) .
this implies that the left - hand side of eq .
( [ 10sun11 ] ) is zero and the right - hand side of eq .
( [ 10sun11 ] ) is positive whenever @xmath119 .
hence , we can confirm the inseparability of the bs number state from the violation of eq .
( [ 10sun11 ] ) except for the case of @xmath120 .
to cover the case of @xmath121 , we consider another separable condition proposed in ref .
following eq . ( 11 ) of @xcite , let us define @xmath122 , \nonumber \\ h_y & = & \frac{1}{2 i } [ ( a^\dagger b^\dagger)^2 - ( a b ) ^2 ] , \nonumber \\ n_+ & = & \frac{1}{4 } ( n_a + n_b ) , \nonumber \\
\tilde l_x & = & \frac{1}{2 } [ ( a^\dagger b)^2 + ( a b^\dagger)^2 ] , \nonumber \\
\tilde l_y & = & \frac{1}{2i } [ ( a^\dagger b)^2 - ( a b^\dagger)^2 ] . \end{aligned}\ ] ] from the commutation relation @xmath123 = ih_x$ ] we have the uncertainty relation @xmath124 .
partial transposition on this inequality leads to the separable condition in eq .
( 14 ) of @xcite , @xmath125 { \langle \delta^2 n_{+ } \rangle } \ge \frac{1}{4 } | { \langle \tilde l _ x \rangle}| ^2 , \label{14sun11}\end{aligned}\ ] ] where @xmath126\otimes [ b^2 , ( b^\dagger ) ^2 ] $ ] . from a direct calculation of the commutation relations @xmath127=4 a^\dagger a + 2 $ ] and @xmath128=4 b^\dagger b + 2 $ ] we can see that @xmath129 is a positive operator .
for the bs number states of eq .
( [ p - basisi ] ) we have @xmath130 , and the left - hand side of eq .
( [ 14sun11 ] ) is zero . in turn , by picking the terms which include the same number of annihilation and creation operators from the expression @xmath131 \nonumber \\ & & \times [ d ^2 + ( r^ * c ) ^2 - 2{r^ * } c d ] \nonumber \\ & & + [ ( d^\dagger ) ^2 + ( r c ^\dagger ) ^2 - 2 r c^\dagger d ^\dagger ] [ ( c ^2 ) + r ^2 d^2 + 2 r c d ] \}\nonumber \\
\end{aligned}\ ] ] we have @xmath132\nonumber . \\
\ ] ] this implies @xmath133 for the case of @xmath121 .
hence , the left - hand side of eq .
( [ 14sun11 ] ) is zero and right - hand side of eq .
( [ 14sun11 ] ) is non - zero for the bs number states with @xmath134 .
therefore , the violation of the condition of eq .
( [ 14sun11 ] ) can verify the entanglement of the bs number states with @xmath121 .
similar to the situation that the tms number states violates the condition of eq .
( [ m10sun11 ] ) , the mechanism of entanglement detection of the bs number states can be explained by the coexistence of strong classical correlation and coherence on the photon - number basis .
the classical correlation is confirmed from @xmath135 or @xmath136 while the non - zero coherence of @xmath137 or @xmath138 is observed on the ( schmidt ) bases with single - photon exchange @xmath139 or the ( schmidt ) bases with two - photon exchange @xmath140 for @xmath141 and @xmath142 , respectively .
therefore , we can observe basically the same structure in non - gaussian entanglement for the tms number states and the bs number states .
that satisfies @xmath143 .
there are eleven solutions for @xmath144 : ( 0,1 ) , ( 1,5 ) , ( 5 , 20 ) , ( 20 , 76 ) , ( 76 , 285 ) , ( 285 , 1065 ) , ( 1065 , 3976 ) , ( 3976 , 14840 ) , ( 4840 , 55385 ) , ( 55385 , 206701 ) , and ( 206701 , 771420 ) . the solutions @xmath145 are not displayed in this figure . ]
it might be worth to mention the existence of the bs number states that do not violate the higher order separable condition of eq .
( [ 14sun11 ] ) .
the set of such bs number states is specified by the condition @xmath146 , namely , it can be specified by the pairs of non - negative integers @xmath147 that fulfill @xmath143 .
the distribution of the pairs @xmath147 for @xmath148 is shown in fig .
[ fig : sep - fig3.eps ] .
hence , the separable conditions of eq .
( [ 14sun11 ] ) is also widely useful to verify the entanglement of the bs number states but there is a small portion of the states which fulfill the condition .
consequently , at the present , we require two separable conditions in order to verify whole class of the bs number states based on the measurement of quadrature moments .
it is in a sharp contrast to the case of the tms number states , in which the entanglement can be verified by the violation of the single separable condition of eq .
( [ m10sun11 ] ) .
we have considered two photonic families of entangled states generated from the products of number states by using two - mode gaussian entangling operations and investigated their inseparability associated with the entanglement criterion based on the measurements of quadrature moments .
we have specified the parameter regime of the states where the entanglement can not be verified by the measurement of the covariance matrix , which covers the case of the measurement of quadrature moments to the second order .
we have also shown a couple of continuous - variable entanglement criteria to verify whole the non - gaussian entanglement of the two families .
an interesting finding is that the measurement of the covariance matrices is useless to detect substantial part of the non - gaussian entanglement . on the other hand
, it has been shown that the measurements to the fourth order moments are sufficient for the entanglement verification except for a very small portion of the non - gaussian states .
generation of non - gaussian entangled states from the number states and gaussian operation would be an important goal of experiments and systematic analysis of such states might be an insightful step to comprehend quantum correlation beyond the gaussian framework .
this work was supported by the grant - in - aid for the global coe program `` the next generation of physics , spun from universality and emergence '' from the ministry of education , culture , sports , science and technology ( mext ) of japan .
bennett and d.p .
divincenzo , nature ( london ) 404 , 247 - 255 ( 2000 ) .
nielsen and i.l .
chuang , _ quantum computation and quantum information _ , ( cambridge university press , cambridge , 2000 ) .
cerf , g. leuchs , and e.s .
polzik ( eds ) , _ quantum information with continuous variables of atoms and light _ , ( imperial college press , 2007 ) .
braunstein and p. van loock , * 77 , * 513 ( 2005 ) .
g. adesso and f. illuminati , j. phys . a math .
40 7821 ( 2007 ) .
s. lloyd and s.l .
braunstein , 82 , 1784 ( 1999 ) .
menicucci , p. van loock , m. gu , c. weedbrook , t.c .
ralph , and , m.a .
nielsen , 97 , 110501 ( 2006 ) .
m. gu , c. weedbrook , n.c .
menicucci , t.c .
ralph , and p. van loock , 79 , 062318 ( 2009 ) .
l. mista , r. filip , and j. fiurasek , 65 , 062315 ( 2002 ) ; a.p .
lund , t.c .
ralph , and p. van loock , arxiv : quant - ph/0605247v1 ; f. dellanno , s.de siena , l. albano , and f. illuminati , phys .
a 76 , 022301 ( 2007 ) .
f. dellanno , s.de siena , and f. illuminati , open sys . & info .
13 , 383 ( 2006 ) .
r. m. gomes , a. salles , f. toscano , p. h. souto ribeiro , and s. p. walborn , pnas 106 , 21517 ( 2009 ) . s .-
y . lee and h. nha , phys .
rev . a 82 , 053812 ( 2010 ) .
lee , s .- w .
ji , h .- j .
kim , and h. nha , phys .
a 84 , 012302 ( 2011 ) .
r. namiki , j. phys .
jpn . , 79 , 013001 ( 2010 ) .
v. parigi , a. zavatta , m.s .
kim and m. bellini , science 317 , 1890 , ( 2007 ) ; a. zavatta , v. parigi , m.s .
kim , and m. bellini , new j. phys . 10 123006 ( 2008 ) ; a. zavatta , s. viciani and m. bellini , science , 306 , 660 ( 2004 ) . j. wenger , r. tualle - brouri and p. grangier , phys .
92 , 153601 ( 2004 ) ; a. ourjoumtsev , h. jeong , r. tualle - brouri and p. grangier .
nature ( london ) 448 , 784 ( 2007 ) .
h. takahashi , k. wakui , s. suzuki , m. takeoka , k. hayasaka , a. furusawa , m. sasaki , phys .
lett . 101 , 233605 ( 2008 ) .
n. namekata , y. takahashi , g. fujii , d. fukuda , s. kurimura , and s. inoue , nature photonics 4 , 655 ( 2010 ) .
a. ourjoumtsev , a. dantan , r. tualle - brouri and p. grangier , phys .
98 , 030502 ( 2007 ) .
h. takahashi , j. s. neergaard - nielsen , m. takeuchi , m. takeoka , k. hayasaka , a. furusawa , m. sasaki , nat .
photonics 4 , 178 ( 2010 ) .
e. shchukin and w. vogel , 95 , 230502 ( 2005 ) .
agarwal and a. biswas , new j. phys . 7 , 211 ( 2005 ) .
m. hillery and m.s .
zubairy , 96 , 050503 ( 2006 ) .
d. mchugh , v. buzek , and m. ziman , 74 , 050306(r ) ( 2006 ) .
h. nha and j. kim , 74 , 012317 ( 2006 ) . h. nha , 76 , 014305 , ( 2007 ) . | we consider two classes of non - gaussian entangled states generated from the product of number states with the action of the beamsplitter or the two - mode squeezer .
it is shown that , for many of these states , the covariance matrix is compatible with the covariance matrix of separable gaussian states and their separability can not be verified by the measurements of the first and second moments of canonical variables .
we identify a couple of continuous - variable entanglement criteria with higher order moments to verify these non - gaussian entanglement . |
the study of cosmic microwave background ( cmb ) photons using the _ planck _ @xcite and bicep2 @xcite data sets has opened up a new constraint on inflationary models @xcite . specifically , the detailed analysis of _ planck _ data @xcite constrains single scalar - field models of slow - roll inflation , to have very low tensor - to - scalar fluctuation ratio @xmath0 , with a scalar spectral index @xmath1 possessing no appreciable running . the upper bound set by the _ planck _ collaboration @xcite on tensor - to - scalar fluctuation ratio , related to the absence of the b - modes of polarization in the cmb , is @xmath2 , but their favored regions ( higher than 95% c.l . )
point towards a more stringent bound of @xmath3 . from the theoretical view point one can see that this is in agreement with the so - called starobinsky - type ( or @xmath4 , with @xmath5 denoting the scalar space - time curvature ) inflationary models @xcite .
last year , however , the bicep2 team @xcite made an important claim : to have made the first measurement of b - mode polarization in the cmb radiation .
this measurement was initially interpreted as indication for gravitational waves at the time of the last scattering , with a tensor - to - scalar ratio @xmath6 .
if such claims were confirmed , they would constitute the first experimental observation of ( transverse ) primordial ( possibly quantum ) metric fluctuations .
for the scalar spectal index , it has been found that @xmath1 and @xmath7 , in agreement with the _ planck _ data @xcite . since
then a great effort has been spent in order to compare the bicep2 tensor - to - scalar ratio against the choice of inflationary paradigm ( @xmath4 @xcite , chaotic @xcite , inverse power law @xcite , hilltop @xcite , natural @xcite , supersymmetry @xcite , d - flation @xcite and the like ) .
furthermore , during this period there was an intense debate as to whether the bicep2 signal is indubitably due to primordial gravitational waves , or is polluted by gravitationally - lensed e - modes and galactic foregrounds .
recently , in ref . @xcite it was stressed that magnetized dust associated with radio loops due to supernova remnants might affect the signal received by bicep2 . in the case of polarization effects by galactic dust , refs .
@xcite and @xcite have shown that the cosmological value of tensor - to - scalar ratio could be very small @xmath8 . within this framework ,
the recent analysis on the foreground dust in the bicep2 region released by the _ planck _ collaboration @xcite points to a significant foreground pollution which means that the bicep2 b - mode polarization data can not be used as evidence for primordial cmb polarization .
recently , the joint analysis of bicep2/keck array and _ planck _ data appeared in the literature @xcite . this analysis has placed an upper bound in the tensor - to - scalar ratio , namely @xmath9 at @xmath10 significance level . in this context , the planck team has repeated the inflationary analysis by using the _ planck 2015 _ data and essentially confirm the _
planck 2013 _ results : @xmath11 , @xmath12 and @xmath2 @xcite .
the crux of these studies is that the potential energy of the scalar field , is not really known and one must introduce it using phenomenological arguments , starting with the simplest possibilities .
there has been intense debate and speculation about the functional form of the potential energy @xmath13 .
various candidates have been proposed in the literature , such as a power law , inverse power law , exponential and so on ( for review see @xcite and references therein ) but exact solutions are not abundant and always possess special mathematical features .
some time ago , a special solution for a spatially flat friedmann - lematre - robertson - walker ( flrw ) spacetime with a perfect fluid with a constant equation of state parameter * * @xmath14 * * ( where , for radiation @xmath15 and for dust @xmath16 ) and a scalar field with a constant equation of state @xmath17 was found in @xcite .
in particular , rubano and barrow @xcite ( see also @xcite and @xcite ) showed that under of specific conditions we can solve the einstein equations when the potential @xmath13 has the interesting hyperbolic form : @xmath18 ^{b } , \label{pot}\ ] ] where the constant @xmath19 is @xmath20 and @xmath21 note that @xmath22 and @xmath23 are the usual cosmological parameters ( although @xmath23 denotes any possible matter content according to the appropriate choice of @xmath24 ) . if we trace the late universe ( dustlike matter @xmath25 ) we have @xmath16 .
we remind the reader that @xmath26 denotes the equation of state parameter of the dark energy usually parametrized by @xmath27 , with @xmath28 and @xmath29 being the pressure and density of the dark energy fluid .
obviously the exponent @xmath30 plays an important role in the cosmic dynamics , since it is related with the equation of state parameter , @xmath26 . in this context , inverting eq.([bb ] ) one can prove that the equation of state parameter reduces to @xmath31 .
note that the accelerated expansion of the universe poses the restriction @xmath32 which implies @xmath33 , where @xmath34 .
the potential ( [ pot ] ) has some interesting geometric characteristics . under specific conditions
, it behaves either as an exponential or as a power - law . if @xmath35 ( or @xmath36 ) , we find @xmath37 ( or @xmath38 ) , where @xmath39 ( see also @xcite ) .
the initial motivation in @xcite was to use potential ( [ pot ] ) to describe the late - time acceleration of the universe , but we can also apply eq.([pot ] ) to the very early states of the cosmic evolution when the cosmic fluid is dominated by the radiation and the inflaton components . *
* this might then provide a unified picture of inflation and dark energy in which both eras are described by the potential of eq.([pot ] ) .
it is the purpose of this work to demonstrate compatibility of the rubano and barrow @xcite scenario with the _ planck 2015 _ data , taking into account the foreground ambiguities clouding the _ planck 2015 _ data , as mentioned previously .
the structure of the paper is as follows . the slow - roll inflation and its connection to the hyperbolic potential of ref .
@xcite are reviewed in section ii . in section
iii we study the performance of eq.([pot ] ) against the _
planck 2015 _ data .
finally , our conclusions are summarized in section iv .
diagram for the hyperbolic potential of @xcite using @xmath40 .
the contours borrowed from _
planck 2015 _ @xcite .
the area which is plotted over the contours corresponds to the hyperbolic rubano & barrow @xcite potential ( see eq.([pot1 ] ) .
the points correspond to chaotic ( solid point ) and starobinsky ( star ) inflation respectively.,scaledwidth=48.0% ] diagram .
the area corresponds to the hyperbolic rubano & barrow @xcite potential ( for more details see the caption of figure 1).,scaledwidth=48.0% ]
let us present here the basic ingredients in the context of single - field inflation .
assuming an inflaton field @xmath41 with potential energy @xmath13 , the slow - roll parameters are given by @xmath42@xmath43@xmath44 where the prime denotes derivatives with respect to @xmath41 and @xmath45 .
the corresponding spectral indices are defined in terms of the slow - roll parameters , as usual @xcite , by @xmath46@xmath47@xmath48 in this framework , the number of e - folds is written as by @xmath49 which denotes the value at the horizon crossing for which the scalar amplitude is @xmath50 , where the energy scale of inflation is @xmath51gev . ]
@xmath52 where @xmath53 is the value of the inflaton field at the end of inflation , namely @xmath54 .
now , let us focus on the potential of eq.([pot ] ) which is written as @xmath55 where @xmath56 is the scale in units of @xmath57 .
recall that in the dark - energy era the corresponding constants are related with the cosmological parameters [ see eqs.([aa ] ) and ( [ bb ] ) ] .
therefore , in the early universe one may expect that the corresponding constants in eq.([pot1 ] ) are not necessarily equal to those derived by @xcite using arguments from the late universe where pressureless matter dominates radiation ( for comparison with the observational data see @xcite ) .
when radiation dominates we have @xmath15 and so from the second equality of eq.([bb ] ) we arrive at @xmath58 note that the restriction @xmath59 implies that the potential contains a critical point . now inserting eq.([pot1 ] ) into the slow - roll parameters we obtain after some simple algebra @xmath60@xmath61 , \label{nn1}\]]@xmath62 .
\label{xx1}\ ] ] combining eq.([ee1 ] ) with eq.([nn1 ] ) and eq.([xx1 ] ) we can write @xmath63 in terms of @xmath64 : @xmath65 \label{nn2}\ ] ] @xmath66 .\ ] ] in this case , the number of e - folds becomes @xmath67 .
\label{efold1}\ ] ] inverting the above , we can express the inflaton as @xmath68 by @xmath69 .
\label{efold2}\ ] ] in order to proceed with the analysis we need to know the values of @xmath70 and @xmath53 .
firstly , it is natural to consider that the number of e - folds lies in the interval @xmath71 $ ] . here
, we set it to 55 , for concreteness .
secondly , using the constraint @xmath54 and eq.([ee1 ] ) we can estimate the value of the scalar field at the end of inflation to be @xmath72 where @xmath73 .
since @xmath74 the scale @xmath56 obeys the restriction @xmath75 .
the point of this section is to test the viability of the potential ( [ pot1 ] ) at the inflationary level , involving the latest cosmological data . in particular , the combined analysis between _
planck 2015 _ and various data such as wmap , high-@xmath76 data , and baryonic acoustic oscillations ( bao ) , shows that the scalar spectral index is @xmath11 .
for the scalar spectral index in this work we use @xmath77 .
furthermore , as mentioned in the introduction , the analysis of the _ planck _ collaboration places an upper limit on the tensor - to - scalar ratio is concerned , @xmath78 which is in agreement with the joint analysis of bicep2/keck array and _ planck _ @xcite .
let us now briefly present our results .
in figures 1 and 2 we show the confidence contours in the @xmath79 and @xmath80 planes which are provided by the planck team @xcite . on top of
that we present the area for the individual sets of @xmath79 which are based on the potential ( [ pot1 ] ) , whereas in figure 2 we display the corresponding area in the case of @xmath80 . obviously , our results are consistent with those of _
planck 2015_. specifically , as it can be seen from figure 1 , the tensor - to - scalar fluctuation ratio could reach the value of @xmath81 which is in a good agreement within @xmath82 with that of bicep2/keck array/_planck _ results ( @xmath83 see figure 9 in @xcite ) .
notice that in order to derive @xmath84 the constants in eq.([pot1 ] ) need to obey the following inequalities : @xmath85 [ or inverting eq.([bbw ] ) we obtain @xmath86 and @xmath87 . concerning the running spectral index we obtain @xmath88 which is consistent with that of _ planck 2015 _ , namely @xmath77 .
moreover , we find that the running of the scalar spectral index does not change significantly as a function of @xmath89 ( see figure 2 ) .
regarding the inflaton field at the beginning of inflation , we obtain @xmath90 ; while at the end of inflation we require @xmath91 .
furthermore , we find that the allowed region in which our results satisfy the @xmath92 observational restrictions of _ planck 2015 _ is @xmath93 and @xmath94 .
thus the data applied to slow - roll inflation place constraints on @xmath30 , although the value of scale @xmath56 has only a lower limit at @xmath92 level .
finally , we would like to compare our results with those found using other potentials . specifically , in the case of the chaotic inflation @xmath95 @xcite , the corresponding slow - roll parameters are written as @xmath96 , @xmath97 and @xmath98 .
the latter implies @xmath99 , @xmath100 and @xmath101 . using @xmath102 and @xmath40 we obtain @xmath103 , @xmath104 and @xmath105 .
this also corresponds to the slow - roll regime of intermediate inflation @xcite with hubble rate during inflation given by @xmath106with @xmath107 and @xmath108 gives @xmath109 exactly to first order and @xmath110 . on the other hand the starobinsky inflation @xcite
, namely @xmath111 $ ] , leads to the following slow - roll predictions @xcite : @xmath112 and @xmath113 , where @xmath114 .
furthermore , for starobinsky inflation following the notations of @xcite we find that the running spectral index is given by @xmath115 . to this end using @xmath40 we obtain @xmath116 .
the above slow - roll parameters are indicated by the solid points ( chaotic inflation ) in figures 1 and 2 while the stars represent values for the starobinsky inflation . _ _
in the light of the _ planck 2015 _ results , a debate is taking place in the literature about the best implementation of the inflationary paradigm . in the current article
we would like to contribute to this discussion . employing the hyperbolic dark - energy scalar - field potential of rubano and barrow @xcite
( see also @xcite ) we study the performance of this model as a description of inflation .
we find that the hyperbolic inflation turns out to be quite promising in the context of the new data from _
planck 2015_. specifically , using the scalar - field potential ( [ pot1 ] ) , we calculate the slow - roll parameters analytically and then we compare the corresponding predictions against the observational data .
we find that currently hyperbolic inflation is consistent with the results provided by _
planck 2015 _ within @xmath82 uncertainties .
the combination of ref @xcite with these calculations provide an overall cosmological investigation of the potential given by eq.([pot1 ] ) .
we find that the hyperbolic structure of this potential leads to a viable model which can be used separately to understand the main properties of both inflation and dark energy in the presence of a single perfect fluid .
finally , following the notations of @xcite we can provide a unified picture of dark energy and inflation using the sum of the potentials ( [ pot ] ) and ( [ pot1 ] ) @xmath117 + a\;\mathrm{sinh}^{b}(\phi / f)\ ] ] where @xmath118 and the constants @xmath119 , @xmath120 , @xmath30 , are given by eqs.([aa ] ) , ( [ bb ] ) and ( [ bbw ] ) .
note that similar considerations hold for the early dark - energy model @xcite .
* acknowledgments . *
sb acknowledges support by the research center for astronomy of the academy of athens in the context of the program _ tracing the cosmic acceleration_. jdb acknowledges stfc support .
also , we would like to thank n. e. mavromatos for useful comments and suggestions . for planck constraints on inflationary models ,
see p. a. r. ade _ et al .
_ [ planck collaboration ] , astron .
astrophys . , * 571 * , a22 ( 2014 ) ;
for a general survey of planck results , including measurement of cosmological parameters , see p. a. r. ade _ et al . _
[ planck collaboration ] .
p. a. r. ade _ et al .
_ [ planck collaboration ] , astron .
astrophys . * 571 * , a16 ( 2014 ) a. a. starobinsky , phys .
b * 91 * , 99 ( 1980 ) ; a. s. goncharov and a.d .
linde , sov .
jetp 59 , 930 ( 1984 ) ; j.d .
barrow , nucl .
b 296 , 697 ( 1988 ) ; j.d . barrow and s. cotsakis , phys . lett .
b 214 , 515 ( 1988 ) ; k - i .
maeda , phys.rev .
d37 , 858 ( 1988 ) l. a. urena - lopez , t. matos , phys .
d * 62 , * 081302 ( 2000 ) ; v. sahni and a. starobinsky , int .
d * 9 * 373 ( 2000 ) ; j.a.e .
carrillo , j.m .
silva and j.a.s .
lima , astr .
astroph . : new phenomena and new states of matter in the universe , proceedings of the third workshop ( iwara07 ) ( arxiv:0806.3299 ) j.d .
barrow , phys . lett .
b * 235 * , 40 ( 1990 ) ; j.d .
barrow and a.r .
liddle , phys .
d * 47 * , r5219 ( 1993 ) ; j.d .
barrow , a.r .
liddle and c. pahud , phys .
d * 74 * , 127305 ( 2006 ) ; j.d .
barrow , m. lagos and j. magueijo , phys .
d * 89 * , 083525 ( 2014 ) | rubano and barrow have discussed the emergence of a dark energy , with late - time cosmic acceleration arising from a self - interacting homogeneous scalar field with a potential of hyperbolic power type . here , we study the evolution of this scalar field potential back in the inflationary era . using the hyperbolic power potential in the framework of inflation , we find that the main slow - roll parameters , like the scalar spectral index , the running of the spectral index and the tensor - to - scalar fluctuation ratio can be computed analytically . finally , in order to test the viability of this hyperbolic scalar field model at the early stages of the universe , we compare the predictions of that model against the latest observational data , namely _
planck 2015_. |
in compact qed ( cqed ) there are two distinct phases .
the weak coupling regime corresponds to the coulomb phase with massless photons and has been investigated in very detail in the extremely successful program of perturbation theory .
however , there is also a second regime realized in the strong coupling region , which has become an important research topic since the 1980s . here , magnetic monopoles cause the creation of electric flux tubes leading to a linear potential and therefore to confinement . although the confinement mechanism in qed seems to be rather different from the one in qcd , the investigation of the former gives some inference to the confinement in qcd , which is far more complex to investigate due to its non - abelian structure . in the weak coupling phase
the monopoles are supposed to disappear but they can survive in meta - stable states .
there has been intensive discussion whether the phase transition between these two phases is of first or second order . for a long time it seemed that this transition is of first order@xcite .
some later investigations suggested a second order phase transition@xcite .
recent studies@xcite strongly suggest a first order transition again .
the investigation of the order of the phase transition is of vital importance since the continuum limit can only be obtained in the case of a second order transition .
here we give a short summary of compact qed on the lattice .
the wilson action@xcite of qed is given by @xmath4 a common choice for the definition of the field - strength is @xmath5 where the physical angle @xmath6 $ ] is obtained by splitting off @xmath7 the number of dirac strings penetrating a plaquette @xmath8 @xmath9 and @xmath10 are gauge - dependent contrary to @xmath11 . in our simulations we use the definition due to lang
et al.@xcite @xmath12 where the field strength is a continuous function of @xmath13 and takes into account @xmath14-periodicity .
this definition is achieved by a variation of the wilson action@xcite and is therefore in agreement with the gau law on the lattice . as will be described further below , definition ( [ ft ] )
allows to investigate continuous shifts of monopoles by arbitrary non - integer distances , a property which we need for our investigations .
as mentioned in the introduction an interesting feature of @xmath15 gauge theory is the appearance of magnetic monopoles .
so we will shortly recall the fundamental relations for dirac monopoles .
dirac considered a 3-dimensional ball surrounding a magnetic point charge yielding the following relation @xmath16 the boundary of a boundary like @xmath17 is zero , which implies that the vector potential diverges at a single point on the sphere for non - vanishing @xmath18 .
one is lead to the picture of monopoles connected by arbitrary lines with a diverging vector potential along them .
these lines are the famous dirac strings .
their position is gauge dependent .
so , there are two kinds of divergences . first the `` true '' physical divergences of monopoles and the unphysical `` gauge '' divergences of dirac strings .
furthermore , dirac showed that the electric charge @xmath19 is quantized by the existence of a single magnetic charge @xmath18 since the wave function of an electric charge transported along @xmath20 is modified by @xmath21 transporting @xmath19 around the boundary @xmath17 of eq .
( [ dirac ] ) , the non - ambiguity of @xmath22 demands @xmath23 setting @xmath24 yields for the smallest ( @xmath25 ) charge quantum @xmath26
dirac monopoles in compact u(1 ) theory can be identified according to degrand and toussaint@xcite as follows .
the plaquette angle @xmath27\ ] ] has no direct physical meaning since the action has a periodicity of @xmath14 only .
furthermore , the sum of six plaquette angles enclosing a cube vanishes since each link of the cube is counted twice but in opposite direction . in the language of forms
the cube is the exterior differential of the six plaquettes ( respectively angles ) , which are the exterior differentials of their links . due to
the bianchi identity ( `` the boundary of a boundary is zero '' ) the sum of the six plaquette angles of a cube vanishes .
dirac strings can be closed or start and end at magnetic monopoles .
thus , the magnetic charge in a cube , respectively the number of monopoles in it is given by the sum of in - going minus out - going strings .
it is an integer multiple of @xmath28 .
the corresponding four - current is given by @xmath29 where we used the ( forward ) lattice derivative @xmath30 .
obviously , this current is source - free ( @xmath31 ) .
identifying monopoles this way allows to count the number of monopoles in each cube only . for the identification of monopole positions inside cubes we have to choose another method .
this second method , which is due to ref .
@xcite detects monopoles by measuring the magnetic fluxes leaving cubes .
the distribution of the flux over six plaquettes surrounding a cube reflects the position of the monopole inside .
then , the monopole current is given by the dual maxwell equations , which read in covariant notation @xmath32 with the dual field strength - tensor @xmath33 .
using the definition of the dual field strength - tensor and eqs .
( [ ft ] ) and ( [ me ] ) yields for the magnetic four - current on the lattice @xmath34 evidently , this current is also conserved ( @xmath35 ) .
the magnetic charge in the cube at position @xmath36 is given by @xmath37 in this definition the monopole charge in each cube is not an integer multiple of @xmath18 . only for infinitely separated monopoles investigated in the classical limit
, the charge inside a large surface enclosing a monopole is an integer multiple of @xmath18 since @xmath38 .
this leads to the picture of a single cube yielding the main contribution to the magnetic charge of the monopole , while the neighboring cubes contribute with a few percent .
since this charge distribution is independent of the lattice constant monopoles are still `` point - like '' .
this definition of monopole currents was investigated in detail in ref .
it was shown that the definition ( [ q ] ) is necessary to get agreement with the dual superconductor picture of confinement and the dual london equation@xcite .
using eq .
( [ q ] ) we introduce a monopole s center of mass @xmath39 depending on the flux distribution around the monopole where @xmath40 corresponds to space - like cubes with a positive contribution to the magnetic charge and correspondingly for @xmath41 . with definition ( [ com ] )
it is possible to obtain positions of monopoles which are not restricted to integer values .
for the realization of a classical monopole - configuration on the lattice we discuss two possibilities : * since a plaquette is part of the surface of four cubes , a non - vanishing plaquette value corresponds to non - zero magnetic currents in these cubes . on the dual lattice the corresponding links
assemble a plaquette . following this line of thought , fixing the value of an @xmath42-plaquette leads to a rotating current in the @xmath43-plane .
further fixing of all @xmath42-plaquettes with given coordinates @xmath44 and arbitrary @xmath45 and closing them periodically at the boundary of the lattice yields two currents in @xmath45-direction but with opposite sign . for
sufficiently large plaquette angles the two currents in @xmath45-direction represent a static monopole - antimonopole pair .
thus , arbitrary monopole configurations can be realized by fixing corresponding plaquettes appropriately . + after specifying the constraints for a static monopole pair we minimize the action by local cooling . in fig .
[ jthe ] the sum of the absolute values of the magnetic charges in one time slice @xmath46 in dependence on the angle of the plaquette separating the monopole pair is shown .
+ . besides the measured data and the result due to the definition of degrand - toussaint an analytical approximation formula
is plotted which is given by @xmath47 . ] + using definition ( [ dgt ] ) yields zero charge for plaquette angles in the interval @xmath48 $ ] followed by a discrete jump at @xmath49 to @xmath50 when a dirac string penetrates the plaquette . using definition ( [ zach ] ) this jump
is spread out ( see fig . [ jthe ] ) .
this is caused by the fact that the charge of the monopole in definition ( [ zach ] ) is distributed over several cubes , allowing for charge annihilation for a monopole pair at close distance .
furthermore , values for the approximation formula @xmath51 are plotted in fig .
this expression is obtained by considering two large closed surfaces each surrounding one monopole and assuming that all plaquettes @xmath52 , except the tuned one @xmath53 , have small values @xmath54 .
+ * constraints with local chemical potentials @xmath55 introduced at single cubes can be used to create monopoles at given positions .
the potentials are introduced in the action as follows @xmath56 with the plaquette angle @xmath54 , the chemical potential @xmath57 , the magnetic current @xmath58 and the sum @xmath59 extending over all cubes . fixing a chemical potential @xmath60 for a static monopole pair at a distance of one lattice spacing followed by local cooling preserves such a monopole - antimonopole pair for values @xmath61 only , see fig .
+ introduced at neighboring cubes . ]
+ one can observe that the charge for the monopole pair created at @xmath62 is approximately @xmath63 and increases continuously to a value near @xmath50 .
for smaller values of @xmath64 the chemical potential is not strong enough to keep apart a monopole pair at a distance of one against their magnetic attraction .
the monopoles approach each other with decreasing @xmath64 , therefore the magnetic charge densities of monopoles and antimonopoles overlap leading to gradual annihilation . for increasing @xmath64
the monopoles are closer at the centers of the cubes and their overlap decreases .
since the charge of the monopole is distributed over more than a single cube a value of @xmath50 can not be attained for a monopole pair at distance one . in a strict sense ,
only infinitely separated monopoles obtain the value @xmath50 .
furthermore , as soon as a monopole pair is created one would expect local minima of the action @xmath65 .
indeed this is true for @xmath61 as can be seen in fig .
[ sethe ] . +
[ cr][cr]@xmath66 according to eq .
( [ sgmu ] ) for a monopole pair in neighboring cubes in dependence on the phase @xmath67 of the plaquette separating the monopoles for various values of @xmath64.,title="fig : " ] in fig .
[ sethe ] the action @xmath65 in dependence on the value @xmath67 of the plaquette , which separates the monopole pair is shown for various values of @xmath64 .
one can clearly recognize local minima of @xmath65 for @xmath61 .
the potential @xmath68 of a static monopole pair in the classical limit corresponds to the action @xmath69 , since the kinetic energy @xmath70 vanishes . using the monopole definition ( [ dgt ] ) due to degrand - toussaint@xcite
the monopole charge is contained entirely in single cubes and the potential can be determined for integer distances only . due to the electro - magnetic duality
a monopole pair feels a @xmath0-potential at large separations .
we investigate the monopole - antimonopole potential not only for integer distances but for arbitrary real ones . for this purpose
monopoles must be movable inside cubes .
this can be achieved by tuning of appropriate plaquettes .
[ moving ] shows the result of such tuning . and the positions of monopoles .
the vertical straight line at @xmath71 indicates the position of the tuned plaquette @xmath67 . introducing a chemical potential @xmath72 at @xmath73 and tuning of @xmath67 followed by cooling leads to an antimonopole in the cube at @xmath73 and a monopole in the vicinity of the tuned plaquette . ] the chemical potential is set to @xmath72 at the cube centered at @xmath74 , while the angle @xmath67 of the plaquette with its center at @xmath75 is tuned . after choosing appropriate values of @xmath67 cooling
generates an antimonopole near @xmath74 the center of the cube with non - vanishing @xmath55 .
the position of the monopole depends strongly on the value of @xmath67 . for @xmath76
the plaquette is pierced by a dirac - string and the position of the monopole is larger than two . with decreasing @xmath67 the monopole is continuously shifted towards the antimonopole .
one can observe that for @xmath77 the monopole is shifted slightly to an @xmath36-value below @xmath78 , which is due to magnetic attraction .
calculations for large separations with and without chemical potentials show that the wilson part of the action @xmath65 does not differ . using the above procedure it is possible to generate monopole pair configurations at arbitrary real distances . the potential of a static monopole pair in dependence on their distance yields fig . [ action ] . of the tuned plaquette which allows to vary the monopole position .
] the coordinate @xmath79 of the tuned plaquette @xmath67 varies from @xmath80 to @xmath81 in integer steps .
the values of @xmath79 used to determine the potential are indicated in the figure .
it is interesting to observe that the well - known @xmath0-potential is superimposed by oscillations with a wave length of lattice spacing @xmath82 .
these oscillations are obviously due to the granularity of the lattice .
local minima are located at integer distances , while the maxima can be found at half - integer separations .
this behavior of the potential at non - integer distances may lead to the well - known observation that some monopole loops survive the transition from the confined to the coulomb phase .
@xcite and @xcite report strong indications that this phase transition is of first order .
a vanishing of the two - state signal was detected in refs .
@xcite,@xcite on a lattice with a topology homo - topic to an s@xmath83-sphere .
this was explained by the fact that for such a topology monopole loops may contract and disappear without restriction .
a discretization of an s@xmath83 where the curvature is distributed almost homogeneously over the lattice favors a contraction of arbitrary monopole loops and decreases the influence of meta - stabilities like those appearing in fig .
[ action ] .
magnetic monopoles defined by their charge distributions can take arbitrary positions on a discrete space - time lattice . with this definition
the potential can be calculated for arbitrary non - integer distances between monopoles .
local chemical potentials provide an efficient tool to create monopoles on the lattice and to realize arbitrary monopole configurations .
magnetic monopoles feel a coulomb potential in analogy to electric point - like charges , if they are located in the centers of the lattice cubes .
the above results demonstrate deviations from the @xmath0-behavior when the monopoles are shifted away from the centers . we find it interesting to investigate whether the meta - stabilities appearing in the potential due to the resulting hysteresis are strong enough to influence the order of the phase transition .
methods to remove the influence of the granularity of the lattice will be a subject of our further investigations .
this work was supported in part by `` fonds zur frderung der wissenschaften '' ( fwf ) under contract p16910-n12 ( p.k .
, m.p . ) and by `` fonds zur frderung der wissenschaften '' ( fwf ) under contract p15015-n08 ( h.b .
, m.p . ) .
99 j. jersk , t. neuhaus and p. m. zerwas , phys .
b * 133 * , 103 ( 1983 ) .
g. bhanot , nucl .
b * 205 * , 168 ( 1982 ) .
b. lautrup and m. nauenberg , phys .
b * 95 * , 63 ( 1980 ) .
r. gupta , m. a. novotny , and r. cordery , phys .
b * 172 * , 86 ( 1986 ) . c. b. lang , phys .
lett . * 57 * , 1828 ( 1986 ) . c. b. lang , nucl .
b * 280 * , 255 ( 1987 ) .
m. vettorazzo and p. de forcrand , nucl .
b * 686 * , 85 ( 2004 ) , [ arxiv : hep - lat/0311006 ] .
g. arnold , b. bunk , t. lippert and k. schilling , nucl .
. suppl . * 119 * , 864 ( 2003 ) , [ arxiv : hep - lat/0210010 ] .
k. g. wilson , phys .
d * 10 * , 2445 ( 1974 ) . c. b. lang and c. rebbi , phys .
d * 35 * , 2510 ( 1987 ) .
m. zach , m. faber , w. kainz , p. skala , phys .
b * 358 * , 325 ( 1995 ) , [ arxiv : hep - lat/9508017 ] .
v. singh , r. h. haymaker , d. a. browne , phys .
d * 47 * , 1715 ( 1993 ) .
[ arxiv : hep - lat/9206019 ] t. degrand and d. toussaint , phys .
d * 22 * , 2478 ( 1980 ) . c. b. lang , t. neuhaus , nucl .
suppl . * 34 * , 543 ( 1994 ) , [ arxiv : hep - lat/9311030 ] .
j. jersak , c. b. lang , t. neuhaus , phys .
d * 54 * , 6909 ( 1996 ) , [ arxiv : hep - lat/9606013 ] . | we investigate the influence of the granularity of the lattice on the potential between monopoles . using the flux definition of monopoles
we introduce their centers of mass and are able to realize continuous shifts of the monopole positions .
we find periodic deviations from the @xmath0-behavior of the monopole - antimonopole potential leading to local extrema .
we suppose that these meta - stabilities may influence the order of the phase transition in compact qed .
h. bozkaya@xmath1 , m. faber@xmath2 , p. koppensteiner , m. pitschmann@xmath3 |
gravitational lensing is and will be playing very important roles in astrophysics @xcite @xcite .
it covers the large region in scale from the brown dwarf to the supercluster of galaxies .
microlensing uncovers the dark matter in the halo and the planetary system in the neighbourhood of the sun @xcite . in this letter
we deal with a special macrolensing , that is , the multiple quasar supposed to be caused by a single galaxy in the foreground .
it enables us to reveal the structure of the lensing galaxy without the ambiguity of surrounding galaxies .
it also gives us the information of the large scale structure of the universe through the angular diameter distance .
one of the most popular models of lensing galaxy is the spheroid with softened power law behaviours @xcite .
the physical implication of this model is very clear and it is easy to compare with observations , such as luminosity distribution and ellipticity , etc .
however these merits are also the demerits of this model since it is too restrictive to model shape .
we are not sure whether the lensing galaxy is elliptical or spiral .
it may be even complex system composed of several galaxies .
this model is also analytically non - tractable . on the other hand multipole expansion model ( hereafter mpe model )
is more general than the spheroid model .
mpe model is applicable to any complex system as far as the lensing objects are compact with respect to impact parameters .
also it is analytically tractable . using it , we are able to estimate the structure of the lensing object systematically .
that is , the more we incorporate multipole components , the more we can know the detailed structure of the lensing object . from the observational side
, gravitational lensing effect seems to be very promissing .
deep space survey by hst is now working and sloan digitized sky survey will work soon @xcite .
however the present status of observation is not still sufficient to uncover the affirmative results .
for instance , parameter fitting is very sensitive to the image positions @xcite whose observation can not be precise in many cases .
we need more detailed and more precise data to be adjusted .
in such a situation it is also very useful to study the same lensing phenomena by several models , which reveal the common and therefore true features of lensing objects . in the previous paper @xcite we discussed about the relationship between multipole moments and caustic singularities , and
applied it to pg1115 + 080 . in this treatise
it is also shown that complex coordinates are very useful .
the merit of mpe model is that we can probe into the detailed structure systematically by steping up higher order of multipoles .
an arbitrarily complex but compact lensing object is analytically tractable as far as each @xmath0-pole moment is discussed separately .
so in this article we study the structure of @xmath0-pole term systematically .
if we sum up all contributions , its treatment goes beyond analytical survey and is devoted to the numerical calculations .
two point - mass lens is situated at the antipodal point of multipole expansion model .
two point - mass model consists of a series of multipole moments .
however it allows analytical survey as a whole though a straightforward application of catastrophe theory still needs numerical calculations . in order to understand the essential physical implication of any event ,
analytical survey is indispensable .
we apply the above mentioned models to pg1115 + 080 and discuss their common and discrepant characters among them .
this paper is organized as follows . in section two
we review the metamorphoses of caustic for later use whose detailed arguments are given in our previous work @xcite .
this metamorphoses are applied to mpe model in section three and to the two point - mass model in section five .
section four is devoted to the analysis of the detailed structure of @xmath0-pole contribution in mpe model .
mpe model is applied to the multiple lens system of pg1115 + 080 in section six .
the obtained numerical results are compared with other models , especially with elliptical lens model in section seven .
section eight is devoted to discussions .
for later studies , we summarize the metamorphoses of the caustics .
this is mainly based on our previous work @xcite .
using complex variables , singularities are classified in compact form .
critical line is @xmath1 where @xmath2 is fermat potential and @xmath3 is complex representation of image position , @xmath4 .
@xmath5 is complex conjugate of @xmath3 .
subscripts @xmath6 , @xmath7 on @xmath2 denotes the derivative @xmath8 and so on . here
we use the poisson equation @xmath9 @xmath10 is the normalized surface mass density and is equal to zero on the light path for non - transparent model which we will consider in this article . then the critical line is reduced to @xmath11 by the diagonalization in real coordinates .
@xmath12 ( @xmath13 ) corresponds to @xmath14 ( @xmath15 ) .
cuspoid sequences are characterized by @xmath16 @xmath17 is the derivative operator along the critical line , which is given by @xmath18 their explicit forms are @xmath19 for the cusp ( @xmath20 ) singularity , @xmath21 for the swallowtail ( @xmath22 ) and @xmath23 for the butterfly ( @xmath24 ) .
it should be noticed that the higher cuspoids must also satisfy the conditions of the lower ones .
for instance , butterfly must satisfy eqs .
( [ cocrit ] ) , ( [ cocusp ] ) and ( [ coswallow ] ) as well as eq .
( [ cobutterfly ] ) .
umbilic is @xmath25 as is easily checked this singularity is not realized for the non - transparent model .
beak - to - beak and lips are characterized by @xmath26 @xmath27 @xmath28 the explicit form of eq.([beaklip ] ) is @xmath29 @xmath30 takes the form @xmath31 for non - transparent model . therefore eq.([lipbeaks ] ) leads to beak - to - beak .
fermat s potential is @xmath32 using the expansion form of logarithmic function , @xmath33 we obtain the multipole expansion form of fermat s potential , @xmath34 thus mpe model becomes a non - transparent model from the convergence requirement of taylor expansion .
this property seems to have interrupted popularity in gravitational lensing .
however this model is very useful and applicable to many lensing events and deserves further investigations as will be discussed .
we can transform the potential into , @xmath35 for this potential , critical condition and cusp condition become @xmath36 @xmath37 = 0~ , \label{cuspi}\ ] ] @xmath38 > 0~ , \label{cuspr}\ ] ] where @xmath39 and @xmath40 .
i m ( re ) denotes the imaginary ( real ) part .
unfortunately , if we consider the terms up to quadrupole moment , mpe model is already beyond analytical calculation . however , analytical survey is desirable for the essential physical interpretation .
so we consider the case of @xmath0-pole only , which allows analytical survey . introducing the variables , @xmath41 eqs.([crit ] ) , ( [ cuspi ] ) and ( [ cuspr ] )
are rewritten into , @xmath42=0~ , & \\ & \re ( \hbox{cusp } ) \equiv x^3+(2\an+3)\bn x^2y+\bn^2\{2(\an^2 + 3)xy^2-(\an^2 - 6\an+3)x^2\ } & \nonumber \\
\label{cuspre } & + 3\bn^3(\an^2 + 2\an-1)xy+\an\bn^4(3\an-2)x+\bn^3y(2y+\an\bn)^2>0~ , & \\
\label{condxy } & x \ge y^2~. & \end{aligned}\ ] ] the last condition follows from the just definition of @xmath43 and @xmath44 . as for the solution of @xmath45 , there are @xmath46 solutions for @xmath47 on lens plane and @xmath48 solutions for @xmath49 . at first
, we examine the cusp conditions , eqs.([cuspim ] ) and ( [ cuspre ] ) , and eq.([condxy ] ) . when @xmath50 , @xmath51 , @xmath49 , eq.([cuspre ] ) becomes @xmath52
therefore a part of region that satisfies the cusp conditions is @xmath49 for @xmath53 or @xmath54 ( @xmath55 ) .
further at the points , @xmath56 , the beak - to - beak condition is held .
when @xmath57 , eq.([cuspim ] ) becomes @xmath58 after some calculations , we can obtain the region that satisfies eq.([cuspre ] ) and eq.([condxy ] ) , @xmath59 we call the characteristic points , @xmath60 , @xmath61 and @xmath62 defined by @xmath63 @xmath64 the region in which the cusp conditions is satisfied is depicted in fig.1 for @xmath65 .
+ fig.1 + for the other values of @xmath48 , topological structure does not change .
next we examine the qualitative behaviour of the critical condition , eq.([critii ] ) .
firstly , irrespective of the value of @xmath66 , the curve eq.([critii ] ) passes through the definite point , @xmath67 and passes above @xmath60 , @xmath68 for @xmath69 , the curve has a locally minimum point and @xmath44 tends to @xmath70 at @xmath71 .
furthermore , for fixed @xmath43 , the value of @xmath44 increases as @xmath66 does . from the above arguments ,
the number of the solutions of the critical condition and the cusp conditions changes when the curve eq.([critii ] ) * is tangential to @xmath49 for @xmath54 or , * passes through @xmath62 , or * is tangential to @xmath72 for eq.([region ] ) .
each case is depicted in fig.2 for @xmath73 .
+ fig.2(a ) + + fig.2(b ) + for the case(0 ) , the value of @xmath66 is @xmath74 and the tangential point is @xmath61
. therefore beak - to - beak singularity appears . for the case(1 ) , the value of @xmath75 is @xmath76 as known from graphical consideration , then cusps split into more cusps or merge to less cusps
. therefore butterfly singularity appears . for the case(2 ) , only when @xmath77 , the lines of eqs.([critii ] ) and ( [ cuspim ] ) are tangential to each other in the region eq.([region ] ) and the value of @xmath75 is @xmath78 and then cusps merge to fold , @xmath51 , swallowtail singularity appears .
we surmmarize the result : * for @xmath79 , @xmath80 * * @xmath81 + there are three solutions on @xmath49 line and two solutions for @xmath82 , so @xmath83 cusps on lens plane .
* * @xmath84 + after beak - to - beak singularity appears , there is a solution on @xmath49 line and two solutions for @xmath82 , so @xmath85 cusps on lens plane . *
* @xmath86 + after swallowtail singularity appears , there is a solution on @xmath49 line and no solution for @xmath82 , so one cusp on lens plane . * for @xmath87
* * @xmath81 + there are three solutions on @xmath49 line and one solution for @xmath82 , so @xmath88 cusps on lens plane .
* * @xmath89 + after beak - to - beak singularity appears , there is a solution on @xmath49 line and one solution for @xmath82 , so @xmath90 cusps on lens plane .
* * @xmath91 + after butterfly singularity appears , there is a solution on @xmath49 line and two solutions for @xmath82 , so @xmath92 cusps on lens plane .
* * @xmath86 + after swallowtail singularity appears , there is a solution on @xmath49 line and no solution for @xmath82 , so @xmath48 cusps on lens plane . * for @xmath93
* * @xmath81 + there are three solutions on @xmath49 line and one solution for @xmath82 , so @xmath88 cusps on lens plane .
* * @xmath89 + after beak - to - beak singularity appears , there is a solution on @xmath49 line and one solution for @xmath82 , so @xmath90 cusps on lens plane . * * @xmath94 + after butterfly singularity appears , there is a solution on @xmath49 line and no solution for @xmath82 , so @xmath48 cusps on lens plane .
critical curve produced by the lens galaxy of pg1115 + 080 makes us imagine that the lens galaxy may be approximated by the two point - mass lens [ fig.3 ] .
+ fig.3 +
this model is interesting as its own right from the following reason . in the previous section we studied the complex lens object by decomposing it into @xmath0-pole contribution .
two point - mass model is very simple one but is composed of composite multipoles .
so in this section we analyse this model in the analogous manner as in the previous section .
it is found that this model is qualitatively different from the mpe model .
this model has been discussed by erdl and schneider @xcite .
a straightforward application of catastrophe theory to this model still enforces numerical calculations .
however we can derive singularities in simple analytical ways by pursuing the critical conditions of @xmath95 defined in eq.([fnoexp ] ) .
the surface mass density of two point - mass lens is given by @xmath96~~,\ ] ] here @xmath97 is total mass and @xmath98 . @xmath99 and @xmath100 are locations of two point masses in the lens surface .
fermat potential @xmath101 is given by @xmath102 in complex representation with @xmath103 and @xmath104 .
we can rescale the potential into @xmath105\ ] ] up to irrelevant constant . here
@xmath106 , @xmath107 and @xmath108 .
lens equation @xmath109 gives @xmath110 critical curve which is given by @xmath111 in general system of coordinates takes the following form , @xmath112 introducing the new complex variable @xmath113 , @xmath114 in eq.([2pocrit ] ) is factorized out as @xmath115 using the polar coordinates @xmath116 , critical curve eq.([factorized ] ) is rewritten as @xmath117 where @xmath118 with @xmath119 .
beak - to - beak condition @xmath120 becomes @xmath121 from eq.([fnodef ] ) .
eq.([bbcondi ] ) is equivalent to @xmath122 and @xmath123 and , therefore , to @xmath124 for @xmath125 or to @xmath126 beak - to - beak appears at the points where four real parameters @xmath127 , @xmath128 , @xmath129 and @xmath130 satisfy three equations ( [ fnodef ] ) and ( [ case1 ] ) or ( [ case2 ] ) .
thus metamorphoses of the caustics are appeared in the critical behaviours of @xmath131 .
firstly we consider case i ) .
@xmath132 from the definitions , @xmath133 , @xmath134 ( @xmath135 for case i ) ) and @xmath136 takes the minimum value @xmath137 at @xmath138 for fixed @xmath127 .
@xmath139 we proceed to study the critical behaviour of @xmath137 .
@xmath140 where @xmath141 @xmath142 is a monotonically increasing function of @xmath143 and the root of @xmath144 is given by @xmath145 , @xmath146 the behaviour of @xmath137 is depicted in fig.4 . -
+ fig.4 + -
nextly we consider case ii ) where @xmath147 .
we define @xmath148 .
then @xmath149 is a monotonically decreasing function of @xmath127 and has no critical point . on the other hand , @xmath150 and @xmath151 where @xmath152 @xmath153 is a monotonically increasing function of @xmath127 and has only one real solution @xmath154 to @xmath155 , @xmath156 the behaviour of @xmath157 is depicted in fig.5 . - + fig.5
+ - the above discussions in case i ) and case ii ) together with the facts that @xmath158 are summarized as the contour map of @xmath95 [ fig.6 ] . - + fig.6
it is found that the critical curve changes its topology when we cross the contour @xmath159 passing through @xmath160 .
we denote the @xmath114 values corresponding to the contour @xmath159 as @xmath161
. then @xmath162 and @xmath163 which agree with the results of erdl and schneider @xcite .
in this section , we apply our mpe model to the lensing galaxy of multiple quasar pg1115 + 080 .
we use deflection potential up to the @xmath164-pole expansion here . about this lensed quasar
, we have got the observed data of positions and relative amplifications of the images @xcite .
( we have adopted the new data for only @xmath165 from angonim - willaime et al .
@xcite ) the positions of the images are as follows ; @xmath166 and the relative amplifications of the images are as follows ; @xmath167 lens equation up to the @xmath164-pole expansion is given by @xmath168 here @xmath169 is source position , @xmath170 is image position , and @xmath171 , @xmath172 , @xmath173 are dipole , quadrupole and @xmath164-pole , respectively .
these are all dimensionless complex numbered quantities normalized by suitable power of mass ( see eq.([renormalize ] ) ) .
image amplification is given by @xmath174 we obtain values of the parameters included in eq.([fiteq ] ) which reproduce the observed data well .
these parameter values are as follows ; if we assume einstein - de sitter universe , this mass would be @xmath176 . here
redshifts of the lensing galaxy and the source quasar are @xmath177 and @xmath178 , respectively .
hubble constant is @xmath179km / sec / mpc .
calculated image positions are @xmath180 and calculated relative amplifications are @xmath181 then , we should check the validity of approximation in multipole expansion up to @xmath164-pole .
it must hold the following conditions .
@xmath182 and the best fit parameters ( [ para ] ) give following values .
@xmath183 this result roughly holds the condition @xmath184 .
so this mpe model up to @xmath164-pole has been proved to be valid up to @xmath164-pole moment .
we performed parameter fitting of this lensed quasar using mpe model up to quadrupole in the previous article @xcite .
the values of source position , dipole and quadrupole in eq.([para ] ) are similar to those values in previous fitting .
so they are consistent .
we have shown in the previous sections that deflection potential up to the quadrupole expansion reproduces multiple images well .
its critical curve takes the similar form produced by two point - mass lens @xcite .
so it may be meaningful to check that two point - mass lens model on the single lens plane well resembles with the lens of pg1115 + 080 .
also it is very helpful for the study of lensing object that we analyze the same lensing system by qualitatively different models . as a typical example , we consider softened power - law spheroid models .
fermat potential given by eq .
( [ pofermat ] ) is expanded by multipoles as @xmath185 here @xmath186 and so on , which are the real representations of eqs.([fermul ] ) - ( [ polemul ] ) . in this case
we have six independent parameters @xmath187 besides source positions @xmath188 . here
we assume that mpe model up to quadrupole moments is good approximation as we have shown and that two point - mass lens can be considered to be their prototype
. then we must be able to fit the parameters @xmath187 from eq.([mulpara ] ) so as to give the same @xmath189 as the mpe model .
this is always possible . for eq.([mulpara ] ) consists of five equations and we have six parameters to be adjusted .
whether two point - mass model is good approximation or not , therefore , depends on the fact that the higher multipole terms subsequent to quadrupole in eq.([mulpsi ] ) is small or not . here
higher multipole terms is fixed by @xmath190 determined by eq.([muldef ] ) . as is easily checked it is impossible .
therefore two point - mass lens model can not be a good approximation of pg1115 + 080 .
this model is also called elliptical lens model whose mass density is given by @xmath191 with @xmath192 the principal axes of spheroid ( @xmath193 ) are related with the coordinate of optical axis and lens surface ( @xmath194 ) by @xmath195 that is , optical axis @xmath3 is tilted to @xmath196 by @xmath197 .
here @xmath198 is the core radius and @xmath199 is the constant central density .
@xmath200 is the eccentricity . for @xmath201 total
mass is divergent , mass density has a cut - off @xmath202 this model has been applied to pg1115 + 080 by narasimha et al .
however they used the observational data different from ours , therefore we can not simply compare their results with ours .
elliptical lens model is qualitatively different model from mpe model .
the former is transparent and the latter non - transparent .
if we are only satisfied by reproducing image positions , both model may pass our requirements .
so we must try parameter fitting to more various and precise data .
the detail of this model has been discussed in a separate form @xcite , so we just quote the result [ table 1 , table 2 ] .
+ table 1 +
+ table 2 +
spls has a large fitting zone for its transparency compared with mpe . as a whole this model gives rather non - compact lensing object unlike the compact aspect of narasimha et al .
this seems to be incompatible with the observed lensing object @xcite .
observed image amplifications are better fitted by mpe model than by elliptical lens model .
mpe model is not so popular as the elliptical lens model with or without external shear since most of the gravitational lensings are thought to be caused by the transparent global systems composed of many galaxies .
however , the recent observations support that the lens of pg1115 + 080 is a compact object whose optical extent is within the observed innermost image b @xmath203 kpc @xcite @xcite . mpe model is neither transparent nor nonsingular . therefore it is free from the general theorems about gravitational lensing ( see chapter 5.4 of schneider et al .
@xcite ) . indeed ,
in fig.3 we have 4 images of odd parity and 2 images of even parity .
there gives rise , of course , no inconsistency .
we must wait further refinement of model check in order that our model parameters are physically affirmative .
for that purpose , we need more precise observations such as image shape , time delays and spectroscopy profile @xcite , etc . in this connection
it is in order to comment on time delay .
time delay is given by @xmath204 for @xmath205 case , @xmath206 for @xmath207 case , @xmath208 time delays for each case are given in table 3 .
+ table 3 + for comparison
we have also listed the respective time delay based on the elliptical lens model [ table 2 ] .
estimations by mpe model are roughly coincident with the observation by vanderriest et al.@xcite . on the contrary ,
estimations by spls in table 2 are too short , which denies the possibility that its amplification discrepancy with observation might be remedied by the time variation of amplification of source qso .
naively we may conclude that the lensing object of pg1115 + 080 is compact also from the theoretical side .
consequently mpe model deserves serious surveys . in this connection
it should be remarked that the dipole term plays an important role in our theory : the presence of dipole moment means that the origin of the coordinates ( which is the optical center ) is away from the center of mass of the lensing galaxy .
there are the arguments which insist that this deviation may be the cause of spiral arms .
also mpe model gives better fit to the observational data than elliptical lens model in many points @xcite .
this may supports the view point that the lensing galaxy is a spiral galaxy , whose equipotential contour is different from spheroid .
there is the observational support of this viewpoint from the analysis of spectroscopy @xcite@xcite . from these facts ,
non zero dipole component may play any role in the formation of spiral arms .
though we applied it only to pg1115 + 080 in this article , mpe model is very useful for the wide class of lensing phenomena .
it gives systematically the multipole moments of the lensing object , irrespective to isolated lens or not .
this serves crucially to the construction of more concrete and more precise structure of lensing object .
we are grateful to dr .
hiroshi yoshida for useful comments .
fig.1 : : the region that satisfies the cusp condition .
solid line is @xmath49 , dotted line is re(cusp)=0 and dashed line is im(cusp)=0 .
three lines pass through the points @xmath60 and @xmath61 .
the cusp condition is satisfied in the region where the dashed line is in the common region of the interiors of the solid and the dotted line .
between @xmath60 and @xmath61 , the dashed line passes very close to but in the exterior of the dotted line .
fig.2 : : the behaviours of the critical condition .
solid line is @xmath49 , dashed line is im(cusp)=0 and dotted line is the critical condition .
fig.2(a ) is the @xmath65 case
. the dotted lines of ( 0 ) , ( 1 ) and ( 2 ) correspond to the cases ( 0 ) , ( 1 ) and ( 2 ) discussed in the text , respectively .
fig.2(b ) is the @xmath209 case .
fig.3 : : shapes of images , critical line and caustics for the multiple quasar pg1115 + 080 ( kakigi et al .
. crosses ( dotted regions ) are the observed ( calculated ) image positions .
it has been assumed that source qso is spherical .
fig.4 : : the critical behaviour of @xmath210 .
critical points @xmath211 and @xmath212 in @xmath213 plane are @xmath214 and @xmath215 , respectively .
we can set @xmath216 without spoiling generality and therefore @xmath217 .
fig.5 : : the critical behaviour of @xmath157 .
critical point @xmath218 is given by @xmath219 .
fig.6 : : the contour map of @xmath131 .
origin is the point of heavier point mass .
@xmath131 diminishes in the direction of @xmath220 .
@xmath212 and @xmath218 are the points where beak - to - beaks appear .
table 1 : : fitting parameters of spls ; ( a ) k=3 case and ( b ) k=2 case . table 2 : : calculated image positions ( and time delays lagging behind c in @xmath221 days ) of spls ; ( a ) k=3 case and ( b ) k=2 case . table 3 : : time delays lagging behind c in @xmath221 days of mpe model . | non - transparent models of multipole expansion model and two point - mass model are analyzed from the catastrophe theory .
singularity behaviours of @xmath0-pole moments are discussed .
we apply these models to triple quasar pg1115 + 080 and compare with the typical transparent model , softened power law spheroids .
multipole expansion model gives the best fit among them .
gravitational lensing , multipole expansion model . |
neutron stars are the remants of the gravitational collapse of massive stars during a type - ii , ib or ic supernova event .
their masses and radii are typically of the order of @xmath3 ( @xmath4 g being the mass of the sun ) and @xmath5 km , respectively . with central densities in the range of @xmath6 times the normal nuclear matter saturation density ,
@xmath7 g/@xmath8 ( @xmath9 @xmath10 ) , neutron stars are most likely among the densest objects in the universe @xcite .
these objects are an excellent observatory to test our present understanding of the theory of strong interacting matter at extreme conditions , and they offer an interesting interplay between nuclear processes and astrophysical observables . conditions of matter inside neutron stars are very different from those one can find in earth , therefore , a good knowledge of the equation of state ( eos ) of dense matter is required to understand the properties of neutron stars . nowadays , it is still an open question which is the true nature of neutron stars . traditionally the core of neutron stars has been modeled as a uniform fluid of neutron - rich nuclear matter in equilibrium with respect to the weak interaction ( @xmath11-stable matter ) .
nevertheless , due to the large value of the density , new hadronic degrees of freedom are expected to appear in addition to nucleons .
hyperons , baryons with a strangeness content , are an example of these new degrees of freedom .
contrary to terrestial conditions , where hyperons are unstable and decay into nucleons through the weak interaction , the equilibrium conditions in neutron stars can make the inverse process happen .
hyperons may appear in the inner core of neutron stars at densities of about @xmath12 .
their presence on the neutron star interior leads to a softening of the eos and cosequently to a reduction of the maximum mass .
other neutron star properties , such as their thermal and structural evolution , can be also very sensitive to the composition , and therefore to the hyperonic content of neutron star interiors .
in particular , the cooling of neutron stars may be affected by the presence of hyperons , since they can modify neutrino emissivities and can allow for fast cooling mechanisms .
furthermore , the emission of gravitational waves in hot and rapidly rotating neutron stars due to the so - called r - mode instability can also be afffected by the presence of hyperons in neutron stars , because the bulk viscosity of neutron star matter is dominated by the contribution of hyperons as soon as they appear in the neutron star interior . in the following i review briefly the so - called hyperon puzzle and other implications of hyperons on the properties of neutron stars .
the presence of hyperon in neutron stars was considered for the first time in the pioneering work of ambartsumyan and saakyan in 1960 @xcite . since then , their effects on the properties of these objects have been studied by many authors using either phenomenological @xcite or microscopic @xcite approaches for the neutron star matter eos with hyperons .
phenomenological approaches , either relativistic or non - relativistic , are based on effective density - dependent interactions which typically contain a certain number of parameters adjusted to reproduce nuclear and hypernuclear observables , and neutron star properties .
relativistic mean field ( rmf ) models @xcite and skyrme - type interactions @xcite are among the most commonly used ones within this type of approach .
microscopic approaches , on the other hand , are based on realistic two - body baryon - baryon interactions that describe the scattering data in free space .
these realistic interactions have been mainly constructed within the framework of a meson - exchange theory @xcite , although recently a new approach based on chiral perturbation theory has emerged as a powerful tool @xcite . in order to obtain the eos one has to solve then the very complicated many - body problem @xcite .
a great difficulty of this problem lies in the treatment of the repulsive core , which dominates the short - range behavior of the interaction .
although different microscopic many - body methods have been extensively used to the study of nuclear matter , up to our knowledge , only the brueckner hartree
fock ( bhf ) approximation @xcite of the brueckner bethe
goldstone theory , the @xmath13 approach @xcite , the dirac
hartree fock theory @xcite , and very recently the auxiliary field diffusion monte carlo method @xcite , have been extended to the hyperonic sector .
all these approaches agree that hyperons may appear in the inner core of neutron stars at densities of @xmath14 as it has been said . at such densities ,
the nucleon chemical potential is large enough to make the conversion of nucleons into hyperons energetically favorable .
this conversion relieves the fermi pressure exerted by the baryons and makes the eos softer , as it is illustrated in panel ( a ) of fig .
[ f : fig1 ] for a generic model with ( black solid line ) and without ( red dashed line ) hyperons . as a consequence ( see panel ( b ) ) the mass of the star , and in particular the maximum one , is substantially reduced . in microscopic calculations ( see _ e.g. , _ refs .
@xcite ) , the reduction of the maximum mass can be even below the canonical " one of @xmath15 @xcite .
this is not the case , however , of phenomenological calculations for which the maximum mass obtained is still compatible with the canonical value .
in fact , most relativistic models including hyperons obtain maximum masses in the range @xmath16 @xcite .
although the presence of hyperons in neutron stars seems to be energetically unavoidable , however , their strong softening of the eos leads ( mainly in microscopic models ) to maximum masses not compatible with observation .
the solution of this so - called `` hyperon puzzle '' is not easy , and it is presently a subject of very active research , specially in view of the recent measurements of unusually high masses of the millisecond pulsars psr j1903 + 0327 ( @xmath17 ) @xcite , psr j1614 - 2230 ( @xmath1 ) @xcite , and psr j0348 + 0432 ( @xmath2 ) @xcite which rule out almost all currently proposed eos with hyperons ( both microscopic and phenomenological ) . to solve this problem
it is necessary a mechanism that could eventually provide the additional repulsion needed to make the eos stiffer and , therefore the maximum mass compatible with the current observational limits .
three different mechanisms that could provide such additional repulsion that have been proposed are : ( i ) the inclusion of a repulsive hyperon - hyperon interaction through the exchange of vector mesons @xcite , ( ii ) the inclusion of repulsive hyperonic three - body forces @xcite , or ( iii ) the possibility of a phase transition to deconfined quark matter at densities below the hyperon threshold @xcite . in the following i
briefly revise these three possible solutions .
this solution has been mainly explored in the context of rmf models ( see _ e.g. , _ refs .
@xcite ) and it is based on the well - known fact that , in a meson - exchange model of nuclear forces , vector mesons generate repulsion at short distances . if the interaction of hyperons with vector mesons is repulsive enough then it could provide the required stiffness to explain the current pulsar mass observations .
however , hypernuclear data indicates that , at least , the @xmath18n interaction is attractive @xcite .
therefore , in order to be consistent with experimental data of hypernuclei , the repulsion in the hyperonic sector is included only in the hyperon - hyperon interaction through the exchange of the hidden strangeness @xmath19 vector meson coupled only to the hyperons . in this way
, the onset of hyperons is shifted to higher densities and neutron stars with maximum masses larger than @xmath20 and a significant hyperon fraction can be successfully obtained .
for further information the interested reader is referred to any of the works that have explored this solution in the last years .
it is well known that the inclusion of three - nucleon forces in the nuclear hamiltonian is fundamental to reproduce properly the properties of few - nucleon systems as well as the empirical saturation point of symmetric nuclear matter in calculations based on non - relativistic many - body approaches .
therefore , it seems natural to think that three - body forces involving one or more hyperons ( _ i.e. , _ nny , nyy and yyy ) could also play an important role in the determination of the neutron star matter eos , and contribute to the solution of the hyperon puzzle .
these forces could eventually provide , as in the case of the three - nucleon ones , the additional repulsion needed to make the eos stiffer at high densities and , therefore , make the maximum mass of the star compatible with the recent observations .
this idea was suggested even before the observation of neutron stars with @xmath21 ( see _ e.g. , _ ref .
@xcite ) , and it has been explored by some authors in the last years @xcite .
however , the results of these works show that there is not yet a general consensus regarding the role of hyperonic three - body forces on the hyperon puzzle .
whereas in refs .
@xcite these forces allow to obtain hyperon stars with @xmath20 , in ref .
@xcite the larger maximum mass that they can support is @xmath22 , and the results of ref .
@xcite are not conclusive enough due to their strong dependence on the @xmath18nn force employed .
therefore , it seems that hyperonic three - body forces are not the full solution to the hyperon puzzle , although , most probably they can contribute to it in a very important way .
the interested reader is referred to these works for the specific details of the calculations .
several authors have suggested that an early phase transition from hadronic mater to deconfined quark matter at densities below the hyperon threshold could provide a solution to the hyperon puzzle .
therefore , massive neutron stars could actually be hybrid stars with a stiff quark matter core .
the question that arises in this case is then whether quarks can provide the sufficient repulsion required to produce a @xmath20 neutron star . to yield maximum masses larger than @xmath20 ,
quark matter should have two important and necessary features : ( i ) a significant overal quark repulsion resulting in a stiff eos , and ( ii ) a strong attraction in a particular channel resulting in a strong color superconductivity , needed to make the deconfined quark matter phase energetically favorable over the hadronic one @xcite .
several models of hybrid stars with the necessary properties to generate @xmath20 neutron stars have been proposed in the recent years @xcite .
conversely , the observation of 2@xmath23 neutron stars may also helped to impose important constraints on the models of hybrid and strange stars with a quark matter core , and improve our present understanding of the hadron - quark phase transition . here
the interested reader is also referred to the original works for detail information on this possible solution .
neutron stars are formed after a successful supernova explosion .
properties of newly born neutron stars are affected by thermal effects and neutrino trapping .
these two effects have a strong influence on the overall stiffness of the eos and the composition of the star .
in particular ( see _ e.g. , _
@xcite ) matter becomes more proton rich , the number of muons is significantly reduced , and the onset of hyperons is shifted to higher densities . in addition , the number of strange particles is on average smaller , and the eos is stiffer in comparison with the cold and neutrino - free case .
a very important implication of neutrino trapping in dense matter is the possibility of having metastable neutron stars and a delayed formation of a `` low - mass '' ( @xmath24 ) black hole .
this is illustrated in fig .
[ f : fig2 ] for the case of a bhf calculation of ref .
the figure shows the gravitational mass @xmath25 of the star as a function of its baryonic mass @xmath26 .
if hyperons are present ( panel ( a ) ) , then deleptonization lowers the range of gravitational masses that can be supported by the eos from about @xmath27 to about @xmath28 ( see dotted horizontal lines in the figure ) .
since most of the matter accretion on the forming neutron star happens in a very early stages after birth ( @xmath29 s ) , with a good approximation , the neutron star baryonic mass stays constant during the evolution from the initial proto - neutron star configuration to the final neutrino - free one .
then , for this particular model , proto - neutron stars which at birth have a gravitational mass between @xmath30 ( a baryonic mass between @xmath31 ) will be stabilized by neutrino trapping effects long enough to carry out nucleosynthesis accompanying a type - ii supernova explosion .
after neutrinos leave the star , the eos is softened and it can not support anymore the star against its own gravity .
the newborn star collapses then to a black hole @xcite . on the other hand , if only nucleons are considered to be the relevant baryonic degrees of freedom ( panel ( b ) ) , no metastability occurs and a black hole is unlikely to be formed during the deleptonization since the gravitational mass increases during this stage which happens at ( almost ) constant baryonic mass . if a black hole were to form from a star with only nucleons , it is much more likely to form during the post - bounce accretion stage .
the cooling of the newly born hot neutron stars is driven first by the neutrino emission from the interior , and then by the emission of photons at the surface .
neutrino emission processes can be divided into slow and fast processes depending on whether one or two baryons participate .
the simplest possible neutrino emission process is the so - called direct urca process ( @xmath32 , @xmath33 ) .
this is a fast mechanism which however , due to momentum conservation , it is only possible when the proton fraction exceeds a critical value @xmath34 to @xmath35 @xcite .
other neutrino processes which lead to medium or slow cooling scenarios , but that are operative at any density and proton fraction , are the so - called modified urca processes ( @xmath36 , @xmath37 ) , the bremsstrahlung ( @xmath38 ) , or the cooper pair formation ( @xmath39+\nu+\bar \nu$ ] , @xmath40+\nu+\bar \nu$ ] ) , this last operating only when the temperature of the star drops below the critical temperature for neutron superfluidity or proton superconductivity . if hyperons are present in the neutron star interior new neutrino emission processes , like _
@xmath41 , may occur providing additional fast cooling mechanisms .
such additional rapid cooling mechanisms , however , can lead to surface temperatures much lower than that observed , unless they are suppressed by hyperon pairing gaps .
therefore , the study of hyperon superfluidity becomes of particular interest since it could play a key role in the thermal history of neutron stars .
nevertheless , whereas the presence of superfluid neutrons in the inner crust of neutron stars , and superfluid neutrons together with superconducting protons in their quantum fluid interior is well established and has been the subject of many studies , a quantitative estimation of the hyperon pairing has not received so much attention , and just few calculations exists in the literature @xcite .
it is well known that the upper limit on the rotational frequency of a neutron star is set by its kepler frequency @xmath42 , above which matter is ejected from the star s
equator @xcite .
however , a neutron star may be unstable against some perturbations which prevent it from reaching rotational frequencies as high as @xmath42 , setting , therefore , a more stringent limit on its rotation @xcite .
many different instabilities can operate in a neutron star . among them , the so called r - mode instability @xcite , a toroidal mode of oscillation whose restoring force is the coriolis force , is particularly interesting .
this oscillation mode leads to the emission of gravitational waves in hot and rapidly rotating neutron stars though the chandrasekhar
schutz mechanism @xcite .
gravitational radiation makes an r - mode grow , whereas viscosity stabilizes it .
therefore , an r - mode is unstable if the gravitational radiation driving time is shorter than the damping time due to viscous processes . in this case
, a rapidly rotating neutron star could transfer a significant fraction of its rotational energy and angular momentum to the emitted gravitational waves .
these waves , potentially detectable , could provide invaluable information on the internal structure of the star and constraints on the eos .
bulk ( @xmath43 ) and shear ( @xmath44 ) viscosities are usually considered the main dissipation mechanism of r- and other pulsation modes in neutron stars .
bulk viscosity is the dominant one at high temperatures ( @xmath45 k ) and , therefore , it is important for hoy young neutron stars
. it is produced when the pulsation modes induce variations in pressure and density that drive the star away from @xmath11-equilibrium . as a result
, energy is dissipated as the weak interaction tries to reestablish the equilibrium . in the absence of hyperons or other exotic components ,
the bulk viscority of neutron star matter is mainly determined by the reactions of direct and modified urca processes . however , has soon as hyperons appear new mechanisms such as weak non - leptonic hyperon reactions ( @xmath46 , @xmath47 ) , direct ( @xmath48 , @xmath49 ) and modified hyperonic urca ( @xmath50 , @xmath51 ) , or strong interactions ( @xmath52 , @xmath53 , @xmath54 ) contribute to the bulk viscosity and dominate it for @xmath55 .
several works have been devoted to the study of the hyperon bulk viscosity @xcite .
the interested reader is referred to these works for detailed studies on this topic .
the time depencende of an r - mode oscillation is given by @xmath56 , where @xmath57 is the frequency of the mode , and @xmath58 is an overall time scale of the mode which describes both its exponential growth due to gravitational wave emission as well as its decay due to viscous damping .
it can be written as @xmath59 .
if @xmath60 is shorter than both @xmath61 and @xmath62 the mode will exponentially grow , whereas in the opposite case it will be quickly damped away .
for each star at a given temperature t one can define a critical angular velocity @xmath63 as the smallest root of the equation @xmath64 .
this equation defines the boundary of the so - called r - mode instability region .
a star will be stable against the r - mode instability if its angular velocity is smaller than its corresponding @xmath63 .
on the contrary , a star with @xmath65 will develope an instability that will cause a rapid loss of angular momentum through gravitational radiation until its angular velocity falls below the critical value . on panel ( a ) of fig .
[ f : fig3 ] it is presented , as example , the r - mode instability region for a pure nucleonic ( black solid line ) and a hyperonic ( red dashed line ) star with @xmath66 @xcite .
the contributions to the bulk viscosity from direct and modified nucleonic urca processes as well as from the weak non - leptonic process @xmath67 included in the calculation are shown in the panel ( b ) of the figure .
clearly the r - mode instability is smaller for the hyperonic star .
the reason being simply the increase of the bulk viscosity due to the presence of hyperons which makes the damping of the mode more efficient .
i have briefly reviewed the main effects of hyperons on the properties of neutron . in particular , i have revised the problem of the strong softening of the eos , and the consequent reduction of the maximum mass , due to the presence of hyperons .
i have discussed three different solutions proposed to tackle this problem : ( i ) the inclusion of a repulsive hyperon - hyperon interaction through the exchange of vector mesons ( ii ) the inclusion of repulsive hyperonic three - body forces , and ( iii ) the possibility of a phase transition to deconfined quark matter at densities below the hyperon threshold .
the role of hyperons on the properties of proto - neutron stars , as well as on the cooling properties of newly born neutron stars , and on the so - called r - mode instability has been also re - examined .
the author is very grateful to the organizers of sqm 2015 for their invitation and to d. unkel for interesting discussions .
this work is partly supported by the project pest - oe / fis / ui0405/2014 developed under the inititative qren financed by the ue / feder through the program compete-``programa operacional factores de competitividade '' , and by `` newcompstar '' , cost action mp1304 .
9 shapiro s l and teukolsky s a , _ black holes , white dwarf and neutron stars _ ( wiley , new york ) 1983 ; glendenning n k , _ compact stars : nuclear physics , particle physics and general relativity _ , 2nd edition ( springer , berlin ) 2000 ; haensel p , potekhin a y and yakovlev d g , _ neutron stars 1 : equation of state and structure , astrophysics and space science library series _ , vol .
* 326 * ( springer , new york ) 2007 .
glendenning n k 1982 phys .
b * 114 * , 392 ; glendenning n k 1985 astrophys .
j. * 293 * , 470 ; glendenning n k 1987 z. phys .
a * 326 * , 57 ; glendenning n k and moszkowski s a 1991 phys .
lett . * 67 * , 2414 ; weber f and weigel m k 1989 nucl .
phys . a * 505 * , 779 .
knorren r , prakash m and ellis p j 1995 phys .
c * 52 * , 3470 ; schaffner j and mishustin i 1996 phys .
c * 53 * , 1416 ; huber h , weber f , weigel m k and schaab ch 1998 int
e * 7 * , 310 .
balberg s and gal a 1997 , nucl .
a * 625 * , 435 ; balberg s , lichtenstadt i and cook g b 1999 astrophys .
* 121 * , 515 ; lanskoy d e and yamamoto y 1997 phys .
c * 55 * , 2330 ; tretyakova t y and lanskoy d e 1999 eur .
j. a * 5 * , 391 ; cugnon j , lejeune a and schulze h - j 2000 phys .
c * 62 * , 064308 ; vidaa
i , polls a , ramos a and schulze h - j 2001 phys .
c * 64 * , 044301 ; zhou x - r , schulze h - j , sagawa h , wu c - xand zhao e - g 2007 phys .
c * 76 * , 034312 ; zhou
x - r , polls a , schulze h - j and vidaa i 2008 , phys .
c * 78 * , 054306 .
schulze , m. baldo , u. lombardo , j. cugnon , and a. lejeune , phys .
b * 355 * , ( 1995 ) 21 ; h .- j .
schulze , m. baldo , u. lombardo , j. cugnon , and a. lejeune , phys .
c * 57 * , ( 1998 ) 704 ; m. baldo , g. f. burgio , h .- j .
schulze , phys .
c * 58 * , ( 1998 ) 3688 ; m. baldo , g. f. burgio , h .- j .
schulze , phys .
c * 61 * , ( 2000 ) 055801 ; i. vidaa , a. polls , a. ramos , m. hjorth - jensen , v. g. j. stoks , phys .
c * 61 * , ( 2000 ) 025802 ; i. vidaa , a. polls , a. ramos , l. engvik , and m. hjorth - jensen , phys
c * 62 * , ( 2000 ) 035801 ; h .- j .
schulze , a. polls , a. ramos , and i. vidaa , phys .
c * 73 * , ( 2006 ) 058801 ; h .- j .
schulze and t. rijken , phys .
c * 84 * , ( 2011 ) 035801 .
maessen p m m , rijken t a and de swart j j 1989 phys .
c * 40 * , 2226 ; rijken th a , stoks v g j and yamamoto y 1999 phys
c * 59 * , 21 ; stoks v g j and rijken th a 1999 phys .
rev c * 59 * , 3009 ; rijken th a 2006 phys . rev .
c * 73 * , 044007 ; rijken th a and yamamoto y 2006 phys .
c * 73 * , 044008 .
keil w and janka h - th 1996 astron . and astrophys .
* 296 * 145 ; bombaci i 1996 astron . and astrophys .
* 305 * 871 ; prakash m , bombaci i , prakash m , ellis p j , knorren r and lattimer j m 1997 phys . rep .
. balberg s and barnea n 1998 phys .
c * 57 * 409 ; takatsuka t and tamagaki r 1999 prog .
* 102 * 1043 ; takatsuka t , nishizaki s , yamamoto y and tamagaki r 2000 prog .
. phys . * 105 * 179 ; takatsuka t , nishizaki s , yamamoto y and tamagaki r 2002 prog .
. suppl . *
146 * 279 ; vidaa i and tol ' os 2004 phys .
c * 70 * 028802 ; zhou x - r , schulze h - j , pan f and drayer j p 2005 phys .
lett . * 95 * 051101 ; wag y n and shen h 2010 phys .
c * 81 * 025801 .
chandrasekhar s 1970 phys .
* 24 * 611 ; friedman j l and schutz b f 1978 astrophys .
j. * 221 * 937 ; friedman j l and schutz b f 1978 astrophys . j. * 222 * 281 ; friedman j l 1978 commun . math . phys . * 62 * 247 .
langer w d and cameron a g w 1969 astrophys .
space sci . * 5 * 213 ; jones p b proc .
r. soc . a 1971 * 323 * 111 ; levin y 1999 astrophys . j. * 517 * 328 ; jones p b 2001 phys .
* 86 * 1384 ; jones p b 2001 phys .
d * 64 * 084003 ; lindblom l and owen b j 2002 phys .
d * 65 * 0653006 ; haensel p , levenfish k p and yakovlev d g 2002 astron . and astrophys . * 381 * 1080 ; van dalen e n e and dieperink a e 2004 phys .
c * 69 * 025802 ; chatterjee d and bandyopadhyay d 2006 phys .
d * 74 * 023003 ; bondarescu r , teukolsky s a and wasserman i 2007 phys .
d * 76 * 064019 ; chatterjee d and bandyopadhyay d 2008 astrophys .
j. * 680 * 686 ; gusakov m e and kantor e m 2008 phys .
d * 78 * 083006 ; sinha m and bandyopadhyay d 2009 phys .
d * 79 * 123001 ; patruno a 2010 astrophys
. j. * 722 * 909 ; jha t k , mishra h and sreekanth v 2010 phys .
c * 82 * 025803 . | in this work i briefly review some of the effects of hyperons on the properties of neutron and proto - neutron stars .
in particular , i revise the problem of the strong softening of the eos , and the consequent reduction of the maximum mass , induced by the presence of hyperons , a puzzle which has become more intringuing and difficult to solve due the recent measurements of the unusually high masses of the millisecond pulsars psr j1903 + 0327 ( @xmath0 ) , psr j1614 - 2230 ( @xmath1 ) , and psr j0348 + 0432 ( @xmath2 ) . some of the solutions proposed to tackle this problem are discussed . finally
, i re - examine also the role of hyperons on the cooling properties of newly born neutron stars and on the so - called r - mode instability . |
the celebrated phenomenological ginzburg - landau ( gl ) theory of superconductivity was introduced in the seminal paper by ginzburg and landau @xcite and has been proving ever since to be one of most fruitful and universal tools in the description basic properties of superconductors .
the examples of its success range from the prediction of the abrikosov vortex state @xcite to the recent advance in understanding the complex vortex phase diagram of the high - temperature superconductors @xcite and the properties of mesoscopic superconducting systems @xcite .
the gorkov s derivation of the gl equations from the bcs theory @xcite that have put the gl theory on a firm microscopic basis and related phenomenological constants to material parameters , completed the construction of the gl theory of superconductors .
the gl theory describes well the properties of almost all superconductors near transition temperature and is successful even when dealing with the superconductors with quite complicated band structures .
the notorious exception from the rule is recently discovered multiband superconductor magnesium diboride ( mgb@xmath0 ) .
as it was shown @xcite , the gl theory applies to mgb@xmath1 only in the very immediate vicinity of the transition temperature , @xmath2 , i.e. within the interval which turns out to be much more narrow than the usual one given by the condition @xmath3 .
the origin of this narrowing of the interval of validity lies in the sophisticated band structure of the material and reflects the specific interplay between its single - electron and superconducting characteristics .
namely , the brillouin zone of mgb@xmath1 consists of the two families of bands : the quasi - two - dimensional @xmath4-bands with the strong superconductivity and the weakly superconducting three - dimensional @xmath5-bands . due to the large difference in anisotropy ,
the c - axis coherence length of the @xmath5-bands , @xmath6 , is much larger than the c - axis coherence length of the @xmath4-bands , @xmath7 .
formation of the global coherent superconductivity with the unique order parameter implies the appearance of the associated unique effective coherence length @xmath8 , which turns out to be much smaller than @xmath6 at almost any temperature . as a consequence ,
the gl applicability interval shrinks to the parametrically narrow range of temperatures where @xmath9 . beyond this temperature range
the system is strongly nonlocal along the @xmath10-axis ; to describe such a nonlocality one has to employ a generalized nonlocal gl model @xcite .
one of the spectacular manifestations of the non - gl behavior in mgb@xmath1 is the strong temperature dependence of the @xmath11 anisotropy close to @xmath2@xcite .
the description of superconducting fluctuations is one of the major fields for the application of the gl theory .
since the standard gl approach turns out to be insufficient for mgb@xmath1 , a generalization of the gl theory is required in order to describe its fluctuation properties .
the variation of the order parameter on the scales smaller than the largest intrinsic coherence length means that the usually assumed local approximation does not hold anymore and the corresponding short - wavelength fluctuations have to be taken into account . in the present paper
we develop a nonlocal theory of superconducting fluctuations that applies to a strongly anisotropic two - band superconductor @xcite and show that the short - wavelength fluctuations are essential in it near the critical temperature .
we start with the derivation of the microscopic fluctuation propagator for a such two - band model
. then we use it to nonlocal two - band gl functional and the corresponding time - dependent gl ( tdgl ) equations .
this , in particular , allows us to calculate the kinetic and thermodynamic observable quantities including the fluctuational specific heat and conductivity . as we have already stated , the main source of the non - gl behavior is the nonlocality in the @xmath12 direction , i.e. , the strong wave - vector dependence of c - axis coherence length .
the conventional local form of the gl equations turns out to be valid only within the narrow interval of temperatures , @xmath13 , where @xmath14 is the relative interband interaction constant which will be specified below . beyond this interval ,
the superconducting correlations in the @xmath5-band become nonlocal and their contribution to the effective coherence length rapidly decreases . far away from @xmath15 the effective c - axis coherence length
is determined only by the @xmath4-band . in other words ,
the effective c - axis coherence diverges for @xmath16 faster than it could be expected from the naive gl extrapolation , @xmath17 , started from high temperatures .
this also leads to the decrease of the effective anisotropy factor @xmath18 . as a consequence , the temperatures dependencies of all fluctuation corrections exhibit the characteristic crossovers between the dominating @xmath4-band regime ( far away from @xmath15 ) and the `` true '' gl regime ( very close to @xmath15 ) .
for example , the @xmath12-axis component of the paraconductivity diverges faster in the immediate vicinity of @xmath15 than one could expect from high - temperature extrapolation using the aslamazov - larkin formula @xcite while the fluctuation specific heat and @xmath19 component of the paraconductivity diverge slower than the corresponding extrapolations .
we will obtain the temperature dependencies of these fluctuation corrections .
the bcs theory was generalized to the case of the two - band electron spectrum long time ago @xcite and has been extended recently to include the specific features of magnesium diboride in refs . .
we briefly overview this theory rewriting it in terms of green s function formalism .
the two - band bcs hamiltonian is given by @xmath20 where @xmath21 and @xmath22 are the creation and annihilation field operators in the heisenberg representation for quasiparticle in band @xmath23 with momentum @xmath24 and spin @xmath4 , @xmath25 is the quasiparticle spectrum , @xmath26 , @xmath27 are the fermi velocity and momentum of the @xmath23-band .
the matrix nature of the electron - electron interaction @xmath28 in ( [ bcsham ] ) reflects the possibility of the interband interactions .
the free electron green s functions for each band have the usual form @xmath29 , where @xmath30 is the time - ordering operator . in the matsubara representation , @xmath31 with @xmath32 being the fermionic matsubara frequencies .
now we turn to calculation of the fluctuation propagator @xmath33 which characterizes the properties of fluctuation cooper pairs and their effect on observable quantities of superconductor above @xmath2.@xcite note first of all , that the interband electron interactions do not result in the cooper pairing of the electrons from different bands but rather lead to the transfer of the pairing correlations between the bands .
indeed , the cooper pairing means the appearance of superconducting - type correlations between two similar states obeying the condition of the time reversal symmetry .
this means that pairing is possible only for electrons belonging to the same band , otherwise the electron states are too diverse ( in terms of the plane waves description , their momenta are not the opposite ) and the integral of the product of their wave functions is zero .
thus the formation of unique condensate should be understood as the result of the _ intraband _ electron correlations and subsequent travelof the cooper pair from one band to another due to off - diagonal interaction components @xmath34 and @xmath35 .
hence , in terms of diagrams , the entrance and exit lines of the fluctuation propagator must belong to the same bands ( see fig .
[ fig - dyson ] ) and it can be presented as the @xmath36 matrix @xmath37 where @xmath38 are the bosonic matsubara frequencies . the corresponding dyson equation for fluctuation propagator can be written in the ladder approximation as ( see fig . [ fig - dyson ] ) @xmath39 where @xmath40 is the matrix polarization operator which is determined by two green s functions loop @xcite . without interband electron scattering
the only nonzero components of the operator @xmath33 are the diagonal ones . ) for the fluctuation propagator @xmath33 ( wavy line ) .
the greek letters indicate the band indices .
the black dot represents the coupling - constants matrix @xmath28 , the loop represents the polarization operator @xmath41 , and the shaded triangle represents the scattering vertex @xmath42.,width=326 ] turning to the role of electron scattering , let us note that at the first sight the short - range impurity potential might lead to the scattering of electrons all over the whole fermi surface which would result in the equalization of all gap values and reduction of @xmath2 .
however , it is well established now that this does not happen in mgb@xmath0 : even in the samples with rather strong intraband scattering , interband scattering remains rather weak .
this property gave possibility to fabricate samples with very high upper critical fields ( up to 50 tesla ) which have transition temperatures only slightly smaller than clean material@xcite .
weakness of the interband scattering has been explained by mazin _
_ in ref . who argued that different parity symmetry of the @xmath4- and @xmath5- orbitals leads to strong reduction of the impurity scattering matrix element between these bands . for simplicity
, we will completely neglect the interband scattering . in this case
the off - diagonal components of the polarization operator vanish .
the averaging over impurities position will result just in the usual renormalization of the green s functions : @xmath43 ( @xmath44 is the corresponding intraband scattering time ) and to the appearance of the scattering vertex part @xmath45 in the expression for the polarization operator @xmath46 the scattering vertex part can be calculated in the ladder approximation@xcite . in the case of a dirty metal @xmath47 this gives@xcite @xmath48 leading to the following result for the polarization operator @xmath49 , \label{polar_expl}%\ ] ] where @xmath50 is the density of states in the band @xmath23 , @xmath51 is the debye frequency and @xmath52 is the digamma - function .
we introduced above the notation for the diffusion - coefficient tensor acting in the momentum and band spaces : @xmath53 the diffusivities @xmath54 determine the band coherence lengths as @xmath55 . for magnesium
diboride the ratio @xmath56 is large , @xmath57 , due to the large difference between the band fermi velocities in the @xmath12 direction @xcite .
this parameter will play an important role in the following consideration .
the inversion of the dyson equation ( [ dyson1 ] ) gives @xmath58 we will use now this general expression to reconsider the main properties of two - band superconductivity and to describe the corresponding fluctuation properties . the critical temperature of a two - band superconductor is determined by the condition @xmath59 taken at @xmath60 and @xmath61 : @xmath62 with @xmath63 being the euler constant .
introducing the coupling - constants matrix as @xmath64{cc}% \nu_{1}g_{11 } & \nu_{2}g_{12}\\ \nu_{1}g_{21 } & \nu_{2}g_{22}% \end{array } \right ) , \ ] ] we can find that the transition temperature is determined by its largest eigenvalue , @xmath65 with @xmath66 , and it is given by the bcs - type equation : @xmath67 following ref . , we introduce the inverse coupling - constants matrix @xmath68{cc}% w_{11 } & -w_{12}\\ -w_{21 } & w_{22}% \end{array } \right ) \equiv\widehat{\lambda}^{-1}-\tilde{\lambda}^{-1}\hat{i } = \frac{1}{\det\widehat{\lambda}}\left ( \begin{array } [ c]{cc}% \!\sqrt { \lambda_{-}^{2}/4\!+\!\lambda _ { 12}\lambda_{21}}\!-\!\lambda_{-}/2 & -\lambda_{21}\\ -\lambda_{12 } & \!\sqrt{\lambda_{-}^{2}/4 \!+\!\lambda_{12}\lambda_{21}}\!+\!\lambda_{-}/2% \end{array } \right ) , \ ] ] where @xmath69 is the unit matrix and @xmath70 . from the definition of the matrix @xmath71
is evident that it is degenerate , @xmath72 having in mind applications of our theory to mgb@xmath0 , we will use the numerically computed effective coupling constants for this compound @xcite : @xmath73{cc}% w_{11 } & w_{12}\\ w_{21 } & w_{22}% \end{array } \right ) = \left ( \begin{array } [ c]{cc}% 0.088 & 0.535\\ 0.424 & 2.56 \end{array } \right ) .\ ] ] we see now that apart from the already mentioned large ratio @xmath74 ( [ ratio ] ) , another characteristic small parameter , the relative interband coupling , @xmath75 appears@xcite . in what follows
we will demonstrate that it is the interplay between these two parameters that defines the rich picture of fluctuations that occurs beyond the traditional gl theory . in terms of the notations introduced above
, the inverse matrix for the propagator ( [ propa ] ) assumes the form @xmath76{cc } w_{11}+\epsilon+\beta_{1}\left ( |\omega_{k}|,q\right)&-w_{12}\\ -w_{21 } & w_{22}+\epsilon+\beta_{2}\left ( |\omega_{k}|,q\right ) \end{array } \right ) \left ( \begin{array}[c]{cc } \nu_{1 } & 0\\ 0 & \nu_{2}% \end{array } \right ) , \label{linvers}%\ ] ] where @xmath77 , @xmath78 , \ ] ] and @xmath79{l}% $ x,\ \
x\ll1$\\ $ \ln x+c_{e}+\ln(8/\pi^{2}),\ x\gg1$% \end{tabular } \right . ,
\label{beta1}\ ] ] note that due to the identity @xmath80 , the matrix @xmath81 is symmetric . calculating @xmath82 and inverting the matrix , one can present @xmath83 as @xmath84{cc}% \widetilde{w}_{22}+\widetilde{\theta}\left [ \epsilon+ \beta_{2}\left(|\omega_{k}|,q\right ) \right ] & \widetilde{w}_{21}\\ \widetilde{w}_{12 } & \widetilde{w}_{11}+\widetilde{\theta}\left [ \epsilon+\beta_{1}\left ( |\omega_{k}|,q\right)\right ] \end{array } \right ) \left ( \begin{array } [ c]{cc}% \nu_{1}^{-1 } & 0\\ 0 & \nu_{2}^{-1}% \end{array } \right ) } { \epsilon+\left [ \widetilde{w}_{11}+\widetilde{\theta}% \epsilon\right ] \beta_{2}\left(|\omega_{k}|,q\right)+\left [ \widetilde{w}_{22}+\widetilde{\theta}% \beta_{2}\left(|\omega_{k}|,q\right)\right ] \beta_{1}\left(|\omega_{k}|,q\right ) } , \label{propagator1}\ ] ] where we introduced the following notations @xmath85 since we are interested in the most singular contribution with respect to @xmath86 , we omitted the terms @xmath87 and @xmath86 as compared to @xmath88 . in order to derive microscopically the tdgl equation and study the effect of fluctuations on transport phenomena one has to perform the analytical continuation of the propagator ( [ propagator1 ] ) from the set of the imaginary matsubara frequencies @xmath89 , @xmath90 to the upper half - plane of the complex frequencies .
since the function @xmath91 has poles at points @xmath92 the analytical continuation to the real frequencies @xmath93 can be performed by the simple substitution @xmath94 in the argument of the function ( [ beta1]).@xcite when this argument is small ( @xmath95 ) the function @xmath96 is reduced to the fourier transform of the diffusion operator : @xmath97 \rightarrow - i\frac{\pi\omega}{8t}+\frac{\pi\hat{d}_{\alpha}q^{2}}{8t}.\ ] ] performing the analytical continuation , using @xmath98 , and accounting for the small parameters of the model ( @xmath99 ) , we obtain the analytically continued propagator @xmath100 in the form @xmath101{cc}% 1/\nu_{1 } & \frac{\widetilde{w}_{21}}{\nu_{2}\left ( 1+\widetilde{\theta}% \beta_{2}\left(-i\omega , q\right)\right ) } \\
\frac{\widetilde{w}_{12}}{\nu_{1}\left ( 1+\widetilde{\theta}\beta_{2}\left(-i\omega , q\right)\right ) } & \frac{s_{12}+\widetilde{\theta}\left [ \epsilon+\beta_{1}\left(-i\omega , q\right)\right ] } % { \nu_{2}\left ( 1+\widetilde{\theta}\beta_{2}\left(-i\omega , q\right)\right ) } % \end{array } \right ) . \label{progen1}%\ ] ] below we will analyze the behavior of this matrix in different temperature intervals .
let us take the reduced temperature so small that both @xmath102 and @xmath103 functions can be expanded .
at the end of this subsection we will arrive at the analytical criterion for this condition to be satisfied .
in this case the propagator ( [ progen1 ] ) significantly simplifies ( we use here the definition @xmath104 ) : @xmath105{cc}% 1/\nu_{1 } & \widetilde{w}_{21}/\nu_{2}\\ \widetilde{w}_{12}/\nu_{1 } & s_{12}/\nu_{2}% \end{array } \right ) , \label{progl}%\ ] ] where @xmath106 , and the effective coherence length components @xmath107 is given by @xmath108 typically @xmath109 and , due to the presence of a small coefficient @xmath110 , the contribution from the @xmath5-band to the transversal component of the effective coherence length ( @xmath111 ) turns out to be small and can be ignored . on the other hand , the c - axis motion in the @xmath5-band , due to the large value of the ratio @xmath112 can play an important role for the longitudinal component of the effective coherence length @xmath113 and when @xmath114 it may significantly exceed @xmath115 .
the characteristic momentum @xmath116 is determined by the diverging gl coherence length : @xmath117 therefore the standard gl theory is valid only at temperatures so close to @xmath2 that @xmath118 and where both @xmath96 functions can be expanded .
the corresponding condition acquires the form @xcite @xmath119 let us return to the propagator ( [ propagator1 ] ) and analyze its behavior in all the vicinity of the transition , @xmath120 , performing allowed expansions and simplifications .
the diffusion in the first band is slow , therefore we can expand the corresponding @xmath96-functions . on the other hand , the function @xmath121 $ ] can be expanded only with respect to @xmath93 and @xmath122 and one has to keep the full nonlinear dependence on @xmath116 . as a result
, we obtain the expression similar to ( [ progl ] ) except for the appearance of the explicit @xmath123-dependence of the gl coefficients @xmath124 and @xmath125 @xmath126{cc}% 1/\nu_{1 } & \frac{\widetilde{w}_{21}}{\nu_{2}\left ( 1+\widetilde{\theta}% \beta_{2}\right ) } \\ \frac{\widetilde{w}_{12}}{\nu_{1}\left ( 1+\widetilde{\theta}\beta_{2}\right ) } & \frac{s_{12}+\epsilon\widetilde{\theta}+\widetilde{\theta}\xi_{1,a}^{2}% q_{a}^{2}}{\nu_{2}\left ( 1+\widetilde{\theta}\beta_{2}\right ) } % \end{array } \right ) .
\label{lgeneral}%\ ] ] with @xmath127 , \label{gammanongl}\\ \widetilde{\xi}_{x}^{2}\left ( q_{z}\right ) & = \xi_{1x}^{2}+\frac { s_{12}+\epsilon\widetilde{\theta}}{1+\widetilde{\theta}\beta_{2 } ( q_{z})}\beta^{\prime}_2 ( q_{z})% \xi_{2x}^{2},\label{xinongl1}\\ \widetilde{\xi}_{z}^{2}\left ( q_{z}\right ) & = \xi_{1,z}^{2}+\frac{s_{12}% } { 1+\widetilde{\theta}\beta_{2 } ( q_{z})}\frac{\beta_{2 } ( q_{z})}{\xi_{2,z}^{2}q_{z}^{2}}% \xi_{2,z}^{2}. \label{xinongl2}%\end{aligned}\ ] ] here @xmath128 one can see that when the condition ( [ glcondition ] ) is satisfied , the characteristic @xmath116 ( see ( [ char ] ) ) is small and the expressions ( [ gammanongl ] ) , ( [ xinongl1 ] ) , and ( [ xinongl2 ] ) reproduce the tdgl coefficients . in the region
@xmath129 the argument @xmath130 of the @xmath96 function ( [ betadef2 ] ) for the essential momenta @xmath116 becomes large and it can not be expanded anymore .
this means that the gradient expansion needed for validity of the gl regime fails . in this region of temperatures
the contribution of the @xmath5-band rapidly decreases and the main role passes to the @xmath4-band .
typically , the @xmath5-band strongly contributes to @xmath131 and gives only small corrections to @xmath132 and @xmath133 .
now we have in place all the elements required for the microscopic calculations of fluctuation effects in a two - band superconductor . however , because of the complex band structure and necessity to take into account the short wavelength fluctuations the diagrammatic calculations present itself a bulky calculus . in
what follows we establish another route to address fluctuation phenomena .
namely , we will re - derive the gl functional and extend the standard gl scheme to treatment of fluctuations for the two - band superconductor , and then apply this modified gl approach to calculations of specific heat and paraconductivity .
knowing the explicit form of the fluctuation propagator ( [ linvers ] ) , one can write down corresponding gl free - energy functional @xmath134 the complete procedure of its microscopic derivation is given in ref . and here we will present only the specific expressions for gl coefficients corresponding to the model under consideration .
the quadratic in order parameter @xmath135 part of the gl functional @xmath136 is expressed in terms of the linearized gl hamiltonian density @xmath137 @xmath138 with @xmath139{cc}% \nu_{1}\xi_{1,a}^{2}q_{a}^{2 } & 0\\ 0 & \nu_{2}\left [ \xi_{2,x } ^{2}q_{\parallel}^{2}\beta^{\prime}_2(q_z)+\beta_2(q_z ) \right ] \end{array } \right ) , \label{hgl}%\ ] ] and @xmath140 . here
the matrix @xmath141{cc}% \nu_{1}\left ( w_{11}+\epsilon\right ) & -\nu_{2}w_{12}\\ -\nu_{1}w_{21 } & \nu_{2}\left ( w_{22}+\epsilon\right ) \end{array } \right)\ ] ] plays the role of the gl coefficient @xmath142 . in general
, we have to keep nonlinear dependence on @xmath116 in @xmath143 meaning that @xmath144 in eq .
( [ f_gl ] ) is not a simple second - order differential operator .
note that the similar generalization of the gl functional has been performed by maki @xcite in order to describe a dirty superconductor in the vicinity of the @xmath145 line .
the coefficients in the fourth order term of the gl functional @xmath146 do not change their form with respect to the noninteracting bands case:@xmath147 with @xmath148 and @xmath149 .
now one can write down the linearized tdgl equation in the form @xcite @xmath150 \delta_{\beta } \!=\!0\ ] ] with the matrix of tdgl dynamic coefficients @xmath151{cc}% \nu_{1 } & 0\\ 0 & \nu_{2}\beta^{\prime}_2(q_z ) \end{array } \right ) \label{gammamat}%\ ] ] which directly follows from the dynamic part of the fluctuation propagator .
we see that there are two essential differences from the standard gl approach : ( i ) strong nonlocality along @xmath10-direction ( ii ) two - component character of the order parameter . in the case where the parameter @xmath88 is large
, one can reduce the functional with two order parameters to the functional for the single band - averaged order parameter.@xcite the tdgl operator @xmath153 is the @xmath36 matrix defined on the band index space .
it is convenient to diagonalize it .
the eigenvalues @xmath154 and the normalized eigenstates @xmath155 obey the equation @xmath156 \psi_{m,\beta}=l_{m}^{-1}(\mathbf{q},\omega)\psi_{m,\alpha } , \label{spectral}%\ ] ] where @xmath157 is the mode index .
superconducting instability corresponds to the vanishing of the @xmath158 eigenvalue at @xmath159 .
calculating the determinant of ( [ spectral ] ) and equating it to zero one finds @xmath160 with@xmath161.\end{aligned}\ ] ] [ [ gl - regime ] ] gl regime + + + + + + + + + + + in the case of small @xmath93 , @xmath86 and @xmath162 , corresponding to the gl region of temperatures , one can find the simplified expressions for the eigenvalues of energy : @xmath163 \label{eigenvaluegl1}\\ l_{2}^{-1 } & \approx\nu_{1}w_{11}+\nu_{2}w_{22}. \label{eigenvaluegl2}%\end{aligned}\ ] ] the expression ( [ eigenvaluegl1 ] ) for @xmath164 reproduces the gl relation ( [ progl ] ) while @xmath165 is not singular .
[ [ beyond - gl - regime ] ] beyond gl regime
+ + + + + + + + + + + + + + + + + + in generally , the value of function @xmath166 may not be small . in the absence of that the smallness the more general expansions for eigenvalues have to be used : @xmath167 we will need further the eigenvector for the singular mode , @xmath158 , in the zero order with respect to the small parameters @xmath168@xmath169 where @xmath170 now we are prepared to revise the results of the standard fluctuation theory and to generalize them for the case of a two - band superconductor .
before going to detail calculations of different fluctuation properties , it is instructive to estimate the relative strength of fluctuations in the available two - band material , mgb@xmath0
. it is characterized by the magnitude of the ginzburg - levanyuk parameter : @xmath171 where @xmath172 g @xmath173 is the flux quantum , @xmath174 and @xmath175 are the in - plane london penetration depth and coherence length , and @xmath176 is the anisotropy parameter in the limit @xmath177 . making use of the values of parameters typical for the clean mgb@xmath0 crystals : penetration depth @xmath178 cm ( ref . ) , coherence length @xmath179 cm ( ref . ) and the anisotropy coefficient @xmath180 ( ref . ) , we obtain @xmath181 , what means that fluctuations in this compound are weak .
this conclusion is not so surprising , since mgb@xmath0 is known to be a good metal with large concentration of charge carriers .
therefore , identifying experimentally the contribution of fluctuations in the clean mgb@xmath0 crystals is a challenging task . on the other hand ,
the amplitude of fluctuation is expected to be much higher in disordered films or in crystals with large number of substitution impurities .
we stress , however , that the effects discussed in this paper hold only as long as scattering does not mix bands and this limits an applicability of our theory for strongly disordered materials .
another complication is that increasing disorder in magnesium diboride is usually accompanied by doping of the @xmath4-band leading to modification of material parameters ( e.g. , decreasing the @xmath4-band anisotropy ) @xcite .
we start with the calculation of the fluctuation contribution to the free energy of the two - band superconductor above the critical temperature .
it is determined by the partition function @xmath182 : @xmath183 , which , in its turn , can be expressed in terms of the determinant of the gl matrix hamiltonian ( [ hgl ] ) @xmath184 where @xmath185 is an insignificant dimensional constant . separating the most singular fluctuation contribution ( see eq .
( [ lgeneral ] ) ) , one can find @xmath186 with @xmath187 .
the corresponding contribution to the specific heat is given by @xmath188 note that the denominator of the the logarithm argument coincides with the denominator of the fluctuation propagator ( [ lgeneral ] ) .
introducing the reduced variable @xmath189 , we rewrite this result in the form convenient for numerical evaluation : @xmath190 where the dimensionless function @xmath191 is defined as @xmath192 it weakly depends on temperature and has the following asymptotics @xmath193{ll}% \left(1+s_{12}r\right)^{-1/2}=\xi_{1,z}/\widetilde{\xi}_{z},&\text{for } \epsilon\ll 1/r+s_{12}\\ \sqrt{1-(s_{12}/\epsilon)\ln\left [ cr\left ( \epsilon - s_{12 } \right)\right ] } , & \text{for } 1/r+s_{12}\ll \epsilon\ll s_{12}\ln ( rs_{12})\\ 1-\frac{s_{12}}{2\epsilon}\frac{\ln\left ( c_{1}r\epsilon\right ) } % { 1+\tilde{\theta}\ln\left ( c_{1}r\epsilon\right ) } , & \text{for } s_{12}\ln ( rs_{12})\ll\epsilon\ll1 \label{interpol } \end{array } \right.\ ] ] with @xmath194 .
the formulas ( [ specheat1])-([kappa ] ) work in the entire region @xmath120 .
the intermediate asymptotic in @xmath195 appears only in the case @xmath196 which is valid for mgb@xmath1 .
@xcite one can see that the main difference between the gl and non - gl regions in the temperature dependence of the fluctuation heat capacity correction is the change in the coefficient of the @xmath197 dependence in factor @xmath198 .
the paraconductivity in the phenomenological gl approach can be expressed via the eigenvalues of the gl hamiltonian and the matrix elements of the `` velocity '' operator @xmath199 .
the approach of ref . can be extended it to the case of two - band model .
the general formula for the paraconductivity tensor then is : @xmath200 .
\label{sigma - general}%\end{aligned}\ ] ] with @xmath201 ( we restored dimensional units in this formula ) .
let us stress that the summation here is performed over the mode indices rather than the band indices in other parts of the paper .
the main contribution to the paraconductivity comes from the projection to the singular mode ( the first term in the square brackets in eq.([sigma - general ] ) ) .
keeping only these terms and using results ( [ eigenvalue1 ] ) for the eigenvalue @xmath202 and ( [ eigenvector ] ) for eigenvector , we derive @xmath203 ^{2}% } { \left [ \epsilon+\xi_{1,i}^{2}q_{i}^{2}+s_{12}\tilde{\beta}_{2}\right ] ^{3}\ , } \label{lastcond}%\ ] ] with @xmath204 , @xmath205 and @xmath206 ( @xmath207 ) .
below we evaluate the in - plane and z - axis components of paraconductivity separately .
in the case of an in - plane paraconductivity we can neglect the small renormalization of the in - plane velocity in formula ( [ lastcond ] ) , perform the @xmath122 integration and get@xmath208 one can see that the in - plane conductivity has exactly the same temperature dependence as the fluctuation specific heat and therefore can be represented in the form analogous eq.([specheat ] ) : @xmath209 where the function @xmath210 is defined by eq .
( [ kappa ] ) .
the above results show that @xmath211 has a 3d character but diverges slower than the specific @xmath212-law .
numerically calculated dependence @xmath213 with parameters typical for mgb@xmath0 , @xmath214 , @xmath215 and @xmath216 , is shown in the left panel of fig .
[ fig - sigmas ] together with the gl asymptotic for @xmath217 and the single @xmath4-band curve . taking the estimate for the coherence length of mgb@xmath0 crystals as @xcite @xmath218 nm
, we find the typical scale for the fluctuation correction , @xmath219 , as @xmath220 40 [ @xmath93 cm]@xmath221 .
this scale must be much higher in dirty mgb@xmath0 films .
the in - plane paraconductivity in mgb@xmath0 films has been studied recently in ref .
but in the analysis of the experimental data the two - band nature of mgb@xmath0 was not taken into account .
nevertheless , it was found that the paraconductivity indeed diverges slower than @xmath212 , in agreement with our results .
( function @xmath222 , see eqs .
( [ sigma - x ] ) and ( [ kappa ] ) ) .
this curve also gives the fluctuation specific heat in units @xmath223 , see eq .
( [ specheat1 ] ) . _
right panel _ shows the z - axis component in units of @xmath224 calculated from eq.([sigma - z ] ) .
both components were calculated using parameters @xmath214 , @xmath215 and @xmath216 . for comparison
we also show the ginzburg - landau asymptotics ( dotted lines ) and @xmath4-band contribution to paraconductivity ( dashed lines ) .
one can see that in - plane component diverges slower and z - axis component diverges faster than the paraconductivity in a single - band anisotropic superconductor . in the @xmath225-plot
we indicate three characteristic regions with different functional dependencies of fluctuation parameters.,width=326 ] the longitudinal component of paraconductivity @xmath226 is given by the eq .
( [ lastcond ] ) . here
the velocity renormalization turns out to be essential and it can not be omitted . performing integration with respect to @xmath227 and using the introduced above reduced variable @xmath228 ,
the result can be rewritten as @xmath229 ^{2}}{\left [ 1+u^{2}+(s_{12}/\epsilon ) \tilde{\beta}_{2}(u)\right ] ^{2}\,}. \label{sigma - z}\ ] ] in the ginzburg - landau regime , @xmath230 , we obtain@xmath231 in the regime of nonlocal fluctuations ( @xmath232 ) , out of the gl region , the behavior @xmath226 becomes quite peculiar .
essential contributions to the integral in eq.([lastcond ] ) occur from two regions of @xmath116 : from @xmath233 in arguments of functions @xmath234 and @xmath235 ( mainly @xmath5-band contribution ) and from @xmath236 ( mainly @xmath4-band contribution ) .
this corresponds to the ranges @xmath237 and @xmath238 in the reduced integral eq .
( [ sigma - z ] ) .
we evaluate separately contributions from these ranges , or , what is the same , from @xmath5- and @xmath4- bands in the most interesting case @xmath239 relevant for mgb@xmath0 . in the range
@xmath240 one can drop @xmath4-band terms in eq .
( [ sigma - z ] ) . as we consider the regime @xmath241 we can neglect the term with @xmath242 in denominator . dimensionalizing the integral
, one can obtain the following result for this term @xmath243 ^{2}.\nonumber\end{aligned}\ ] ] numerical calculation gives @xmath244 and @xmath245 . in the range @xmath238
the main contribution occurs from the @xmath4-band and the terms proportional to @xmath110 ( @xmath5-band terms ) can be treated as small perturbations .
expansion with respect to these terms leads to the following result for the paraconductivity contribution@xmath246 with@xmath247 \\ & \approx-\frac{\ln\left ( c_3r\epsilon\right ) } { 1+\tilde{\theta}\ln\left ( c_4r\epsilon u^{2}\right ) } .\end{aligned}\ ] ] with @xmath248 and @xmath249 .
let us note that in this range of temperatures the contribution from the @xmath5-band comes with the negative sign , as in the cases of the heat capacity and in - plane paraconductivity .
this means that the @xmath5-band contribution changes sign with increasing temperature . comparing contributions ( [ sigma - zpi ] ) and ( [ sigma - zsigma ] )
, we find that the @xmath5-band term dominates in the interval of temperatures @xmath250 .
therefore in the case @xmath196 the @xmath10 axis paraconductivity @xmath226 has three asymptotic regimes : @xmath251{ll}\left(1+s_{12}r\right)^{1/2},&\text{for } \epsilon \ll 1/r+s_{12}\\ \sqrt{r}s_{12}^{2}c_{2}(\tilde{\theta})/\epsilon^{3/2},&\text{for } 1/r+s_{12}\ll \epsilon\ll \frac{\left ( rs_{12}\right ) ^{4/3}}{r}\\ 1-\frac{s_{12}}{2\epsilon}\frac{\ln\left ( c_{1}r\epsilon\right ) } { 1+\tilde{\theta}\ln\left ( c_{1}r\epsilon\right ) } , & \text{for } \max[1/r+s_{12 } , s_{12}\left ( rs_{12}\right ) ^{1/3}]\ll\epsilon\ll1 \end{array } \right . .\
] ] in the case @xmath252 the intermediate asymptotic disappears . the right panel of fig .
[ fig - sigmas ] shows the numerically calculated dependence of @xmath253 with parameters specified in the captions . for comparison
, we also show the gl asymptotics and the @xmath4-band contribution .
in this paper we have presented a microscopic derivation generalizing the conventional gl description onto two - band superconductors .
we have further applied the developed approach to the investigation of the fluctuation phenomena in mgb@xmath0 .
the important feature of our approach is that the derived _ nonlocal _ gl functional takes into account not only the long - wavelength fluctuations ( as is the case of the conventional gl theory ) , but also the short - wavelength fluctuations , which enables to significantly extend the range of validity of the gl technique .
this approach not only permits the study of the thermodynamic characteristics of the superconductors , but provides a convenient technique for calculation of transport coefficients .
it is necessary to underline that we succeeded to solve analytically the problem of accounting for short - wavelength fluctuations due to the specific for our model large difference in the intra - band diffusion coefficients @xmath254 .
fortunately , this assumption corresponds to the practically interesting case of magnesium diboride .
one can see that the main qualitative effect that determines the unusual behavior of fluctuations in the anisotropic two - band model consists in globalization of superconductivity in the immediate vicinity of transition temperature ( @xmath255 ) .
it occurs due to the possibility for electrons of both bands to participate in fluctuation pairing and to exchange of fluctuation pairs between them .
formally , this process manifests itself in the appearance of long range superconducting correlations on the scale @xmath256 . turning the role of fluctuations in mgb@xmath0
, we refer to the formulas for paraconductivity in a one - band 3d anisotropic superconductor @xcite : @xmath257 we see that the growth of the effective @xmath258 results in the noticeable increase of the @xmath10-axis paraconductivity in the immediate vicinity of critical temperature ( @xmath259 ) with respect to its high temperature ( @xmath260 ) extrapolation based only on the @xmath4-band fluctuation pairing .
the crossover between these two 3d regimes occurs in the narrow interval of temperatures @xmath261 where the boson degrees of freedom corresponding to the pairings in @xmath5-band rapidly freeze out , leading to the fast decrease ( @xmath262 ) of @xmath10-axis paraconductivity .
we want to stress that this temperature dependence appears due to the noticeable contribution of short - wavelength fluctuations in this range of temperatures . in the case of the in - plane component of paraconductivity and the fluctuation part of heat capacity , the situation is the opposite .
the coherence length @xmath263 appears in the denominator of the corresponding expressions , so @xmath211 and @xmath264 turn out to be suppressed in the vicinity of transition with respect to their extrapolation formulas from the high - temperature behavior .
similar situation occurs in the temperature dependencies of the in - plane upper critical field and of anisotropy of the upper critical field @xcite .
the obtained results coincide with that ones of the diagrammatic approach but the proposed gl description has an advantage of being more physically transparent , economic , and universal . the fact that the results derived via the gl equations coincide with those of our microscopic consideration is a convincing cross - check of the phenomenological description .
the same approach can be used to derive other fluctuation properties such as the magnetic susceptibility and the field dependencies of conductivity and magnetization in the vicinity of transition .
this work was supported by the u.s .
doe , office of science , under contract # w-31 - 109-eng-38 .
a.a.v . acknowledges the support of the firb project of the italian ministry of science and education .
l. p. gorkov , zh . eksp .
fiz . * 36 * , 1918 , ( 1959 ) [ sov .
jetp , * 9 * , 1364 ( 1959 ) ] ; l. p. gorkov and t.k .
melik - barkhudarov , zh . eksp .
fiz . * 45 * , 1493 , ( 1963 ) [ sov .
jetp , * 18 * , 1031 ( 1964 ) ] .
eltsev , s. lee , k. nakao , n. chikumoto , s. tajima , n. koshizuka , and m. murakami , phys .
b * 65 * , 140501(r ) ( 2002 ) ; physica c * 378 - 381 * , 61 ( 2002 )
. l. lyard , p. samuely , p. szabo , t. klein , c. marcenat , l.paulius , k.h.p .
kim , c.u .
jung , h .- s .
lee , b. kang , s. choi , s .-
lee , j. marcus , s. blanchard , a.g.m .
jansen , u. welp , g. karapetrov , w.k .
b * 66 * , 180502(r ) ( 2002 ) ; m. angst , r. puzniak , a. wisniewski , j. jun , s. m. kazakov , j. karpinski , j. roos , and h. keller , phys .
lett . * 88 * , 167004 ( 2002 ) . strictly speaking , the band structure of mgb@xmath0
is composed of four bands : two @xmath265bands and two @xmath5-bands ( see , e.g. , ref . ) .
however , in real samples the scattering rate between the bands of the same family is usually high while the scattering of electrons between the different families is rather weak.@xcite in this situation scattering homogenizes superconductivity inside both band families and superconductivity in magnesium diboride is effectively described by the two - band model . to simplify terminology , in the rest of the paper we use term `` @xmath4-band '' ( `` @xmath5-band '' ) for the family of @xmath4-bands ( @xmath5-bands ) .
r. h. t. wilke , s. l. budko , p. c. canfield , d. k. finnemore , r.j .
suplinskas , and s. t. hannahs , phys . rev .
lett . * 92 * , 217003 ( 2004 ) ; v. braccini , _ et al .
_ cond - mat/0402001 ; m. angst , s. l. budko , r. h. t. wilke , p. c.canfield cond - mat/0410722 t. yildirim , o. glseren , j. w. lynn , c. m. brown , t. j. udovic , q. huang , n. rogado , k. a. regan , m. a.hayward , j. s. slusky , t. he , m. k. haas , p. khalifah , k. inumaru , and r. j. cava , phys .
* 87 * , 037001 ( 2001 ) . | a theory of fluctuations in two - band superconductor mgb@xmath0 is developed .
since the standard ginzburg - landau ( gl ) approach fails in description of its properties , we generalize it basing on the microscopic theory of a two - band superconductor . calculating the microscopic fluctuation propagator
, we build up the nonlocal two - band gl functional and the corresponding time - dependent gl equations .
this allows us to calculate the main fluctuation observables such as fluctuation specific heat and conductivity . |
bok globules are the most suitable sites for studying the evolution and formation of isolated low mass stars @xcite .
aligned dust grains in the interstellar medium ( ism ) and the bok globule polarize the radiation through dichroic extinction . the grains are aligned with their short axis preferentially aligned with the local magnetic field , resulting in a polarization parallel to the field , as projected onto the plane of the sky ( lazarian et al .
1997 ) .
optical polarimetric observations of dark cloud can give information about the magnetic field orientation in the low density edge regions of clouds , whereas the polarimetric observation in the infrared and submillimetre range can map the field orientation in the higher density central regions of the clouds .
thus the resulting polarization map outlines the geometry of the magnetic field lines projected onto the plane of the sky , which in turn helps to study the relationship between structure kinematics , and embedded magnetic field during various stages of star formation ( @xcite,@xcite,@xcite , vrba , coyne & tapia ( 1981 ) , @xcite , @xcite , @xcite , @xcite , @xcite , @xcite , @xcite , @xcite ) .
@xcite and @xcite modeled the observed variation in polarization with extinction along different lines of sight , in terms of a magnetic field with distinct random and uniform components .
@xcite argued that most of the observed fall - off in @xmath7 is due to the turbulence of the magnetic field in the medium .
@xcite predicted that @xmath8 @xmath9 a@xmath10 , which is closed to the observed dependence .
it has been investigated later that jones model may not be applicable to individual clouds .
@xcite studied the dust grain alignment in the taurus dark cloud ( tdc ) and showed a strong systematic trend in polarization efficiency with extinction which can be expressed by a power law @xmath8 @xmath9 a@xmath11 , in the magnitude region ( 0 @xmath12 25 ) .
they explained this phenomenon of decrease in polarization with increase in extinction by considering a number of factors ( poor grain alignment , grain growth , and or changes in grain shape or composition ) .
@xcite also did a detailed study of interstellar polarization efficiency toward molecular clouds to understand the grain alignment mechanism in different regions of interstellar medium .
they studied the dependence of polarization efficiency on visual extinction by probing background field stars , and found that this dependency can be well expressed by a power law @xmath13 @xmath9 @xmath14 ( @xmath15 is the optical depth ) .
@xcite also showed that , among all the available grain alignment models , the radiative torque mechanism could more efficiently explain the observed polarization in the quiescent regions of the cloud .
in this work , we have probed cb56 , cb60 and cb69 at the v - band , with a scientific goal to measure the optical polarization of the stars background to the globules and to map the magnetic fields within these globules .
the extinction of field stars in each cloud are estimated using a standard technique suggested by @xcite and @xcite .
the extinction of each field stars in the cloud and background star polarization will help to understand the dynamical behavior of the cloud .
some works on cloud cb56 and cb60 were reported by a few investigators .
both near infrared imaging and millimetre dust emission of cb60 were studied by @xcite and @xcite respectively and found that it is about 1500 parsec(pc ) away from the sun . @xcite
also studied the millimetre dust emission of cb69 and reported a distance of 500pc .
this is only marginally smaller than the distance of 600 pc derived by @xcite .
further , @xcite mapped the magnetic fields within cb56 at r - band .
this article is arranged as follows . in section 2
we discuss the observational procedures and details of the data reduction techniques are discussed in section 3 .
this is followed by results and discussion in section 4 ; we end with appropriate conclusions in section 5 .
polarimetric observation of three bok globules cb56 , cb60 and cb69 , have been made at the v - band with exposure time of 900 seconds , using 2-m telescope of girawali observatory at inter university center for astronomy and astrophysics ( iucaa ) , pune ( igo , latitude : 19@xmath165@xmath17 n , longitude : + 73@xmath1640@xmath17 e , altitude = 1000 m ) , india on 4th and 5th march , 2011 .
these globules were selected from the catalogue of @xcite , depending on their visibility at the time of observation .
the observational details of the field containing three bok globules , are listed in table 1 ( columns 1 , 2 and 3 give serial number , globule identification and position angle ( in degree ) , respectively ; columns 4 and 5 give 2000.0 epoch right ascension ( ra ) and declination ( dec . ) , respectively ; columns 6 and 7 give the galactic coordinates , respectively ; column 8 gives the dimension of the globules and columns 9 and 10 gives the date of observation and ut , respectively ) . [ cols="^,^,^,^,^,^,^,^,^,^,^,^,^,^ " , ] [ tab9 ] table 7 , 8 and 9 show the calculated values of @xmath18 , @xmath19 , polarization efficiency ( @xmath20/a@xmath21 ) and probable location of the main sequence field stars of cb56 , cb60 and cb69 respectively . to identify the location of different field stars in a cloud we use statistical approach and define : @xmath22_{avg } - \sigma$ ] , where @xmath23_{avg}$ ] is the average extinction of the cloud and @xmath24 is the standard deviation of e(b - v ) .
if e(b - v ) of any star is less than @xmath25 , then it may be considered as foreground to the cloud .
the star whose @xmath18 is more than @xmath26 is considered background whereas the value within @xmath25 and @xmath26 is considered either within the cloud or maybe background to the cloud . for cb56 ,
the value of @xmath27 ( coefficient of determination computed between @xmath18 and @xmath0 along the observed line of sight ) is 0.14 , which gives the proportion of the variance ( fluctuation ) of polarization that can be predicted from the @xmath18 .
it means the polarization of the starlight along the observed line of sight is weakly correlated with the extinction .
the average extinction and standard deviation are @xmath23_{avg}$ ] = 0.62 ( mag ) and @xmath24 = 0.16 . in table 7
, we have categorized the observed field stars of cb56 , in three distinct groups on the basis of their location .
stars with serial no .
14 and 16 , have @xmath28 which may be located foreground to the cloud . thus excluding these two stars the average value of @xmath18 of rest 11 stars is 0.66 ( mag ) and the value of @xmath29 . on the other hand , stars with serial no .
1 , 3 , 7 , 12 , 13 and 17 have @xmath30 , so they may be considered to be located background of the cloud .
however , the @xmath18 of stars with serial no . 2 , 4 , 8 , 10,and 11 , lies within @xmath25 and @xmath26 , so they may be either within the cloud or located background to the cloud .
thus , these stars fall into third group whose location can not be claimed with certainty .
star with serial no .
18 could nt be located as the bvr magnitude of this star is not available in the vizier database .
stars with serial no . 5 , 6 , 9 and 15 ( please see table [ tab4 ] )
are not in main sequence class so we have excluded them from present analysis . in the observed field of cb60 ,
star having serial no .
9 ( in table 5 ) is identified to be variable star ( v517 pup ) which is confirmed from simbad database .
so we have excluded this from our present analysis .
the average extinction and standard deviation are given by @xmath23_{avg}$ ] = 0.46 ( mag ) and @xmath24 = 0.08 .
the @xmath31 is estimated to be 0.09 .
table 8 indicates the location of observed field stars of cb60 which are categorized following the argument as discussed earlier for identifying the location of the field stars of cb56 . excluding the foreground star ( serial no .
7 ) , the mean @xmath18 is 0.48 ( mag ) , and the value of @xmath32 is 0.11 . stars with serial no . 1 , 2 , 10 and 13 ( please see table [ tab5 ] ) are not in main sequence class
so we have excluded them from present analysis .
table 9 displays the location for the observed field stars of cb69 .
the average extinction and standard deviation of cb69 field stars are @xmath23_{avg}$ ] = 0.96 ( mag ) and @xmath24 = 0.43 where @xmath33 .
we identified the location of some field stars of this cloud , one foreground star ( 13 ) , five background stars ( 1 , 3 , 4 , 9 and 11 ) and five stars within the cloud ( 2 , 6 , 8 , 10 , and 12 ) .
the average @xmath18 excluding the foreground star is 1.01 ( mag ) , and the value of @xmath32 is 0.24 .
however , the location of star number 7 could nt be identified as the bvr magnitude is not available in the vizier database of astronomical catalogues .
stars with serial no . 5 , 14 , 15 and 16 ( please see table [ tab5 ] ) are not in main sequence class
so we have excluded them from present analysis .
@xcite have probed 248 small molecular clouds ( mostly bok globules ) , using deep iras photometry and @xmath34co spectroscopy . comparing their co peak line temperature and co line width
they have divided those clouds into three types a , b , and c. the a type cloud contains the maximum number of clouds ( 74@xmath35 of 248 clouds ) where gas temperature are cool ( @xmath368.5k ) and with little turbulent gas motion .
the b type cloud has unusually warm temperature , but with narrow line width and the c type cloud has quite a broad line width , implying unusual dynamic activity , whereas the gas temperature are cool .
it is observed from @xcite that cloud cb56 and cb69 are identified as a type cloud , and cb60 as b type cloud .
figures 5a5f represent the distribution of _ p _ and @xmath37 in the v - band , for the field stars observed in the bok globules : cb56 , cb60 and cb69 .
cloud cb56 has two iras sources ( 07125 - 2503 ; 07125 - 2507 ) where its major axis has a position angle of 170@xmath38 ( @xcite ) .
the mean value for polarization and their position angle for the cb56 field stars are @xmath39 = 1.08@xmath40 , and @xmath41= 150.94@xmath42 with a standard deviation of @xmath43 = 0.56@xmath40 and @xmath44 = 10.09@xmath45 respectively .
the polarization vector of most of the stars is aligned in a common direction which indicates the local magnetic field of the cloud . the alignment of polarization vectors in local magnetic field is quite expected as a type clouds show less dynamical activities and turbulence .
thus , they have better alignment of the grains .
it is further noticed from fig .
1 that the direction of the local magnetic field of the cloud is almost parallel to the galactic plane .
thus we can infer that the field direction in this cloud is similar to that of the galactic field .
further , significant variation in the polarization value of star 18 ( please see table 4 ) from mean polarization value , suggests that it may be located either within the cloud or maybe located background to the cloud .
it is worth mentioning that bvr magnitudes are not available for this star , so it is not possible to comment its location from photometry . from figure [ fig2 ]
, it can be seen that the polarization vectors in cb60 ( b - type cloud ) seem to be directed in some common direction of the local magnetic field of the cloud .
the direction of local field ( @xmath46 ) is not aligned in the direction of gp ( 40@xmath38 ) which represents that the magnetic field within the cloud is different from the galactic field .
also it is to be noted that b type cloud has some unusual warm temperature .
it has three iras sources ( 08026 - 3122 ; 08029 - 3118 ; 08022 - 3155 ) ( @xcite ) . the average polarization and their position angle of the cb60 field stars are @xmath39 = 1.30@xmath40 , and @xmath41= 155.15@xmath42 with a standard deviation of @xmath43 = 0.94@xmath40 and @xmath44 = 10.67@xmath38 respectively .
the polarization value of star 9 is unusually high from mean polarization value ( please see table 5 ) , which is also a known variable star v517 pup .
it can be observed from figure [ fig3 ] that for the cloud cb69 ( a type cloud ) the polarization vectors are well aligned but the general direction is not along the direction of the gp ( @xmath47 ) .
this cloud has six iras sources ( 16594 - 3315 , 16595 - 3311 , 16586 - 3304 , 16589 - 3315 , 16590 - 3313 , 16590 - 3305 ) ( @xcite ) . the average polarization and their position angle of the cb60 field stars are @xmath39 = 2.00@xmath40 , and @xmath41= 155.8@xmath42 with a standard deviation of @xmath43 = 1.12@xmath40 and @xmath44 = 13.06@xmath38 respectively .
the degree of linear polarization varies ( lower / higher ) depending on the location of the stars in the cloud .
similar variation ( alignment / non - alignment ) of the polarization position angle with gp are generally observed depending on the membership of the stars ( i.e. foreground / background / within the cloud ) .
thus , the efficiency of polarization depends on both , the properties of dust grain and on the environment in which they exist .
so the study of variation of the background star polarization with extinction ( a@xmath21 ) and polarization efficiency ( defined as @xmath48/a@xmath21 ) , can help us in understanding the nature of the dust and the magnetic field associated with the dark cloud ( @xcite , @xcite ) .
the alignment of non spherical dust grains are generally measured by the ratio of degree of polarization produced for a given amount of extinction ( @xmath20/a@xmath21 ) , also known as the _ polarization efficiency _ @xcite .
@xcite observed the linear polarization of 180 nearby stars in u , b , v and r bands where they studied the relation of the maximum polarization @xmath49 with the color excess @xmath18 at a wavelength @xmath50 , and found that the polarization efficiency along different line of sight varies within an upper limit : @xmath51/e(b - v)@xmath52 9.0 @xmath35 mag@xmath53 .
@xcite calculated the theoretical upper limit of the polarization efficiency of the grains due to selective extinction which is @xmath52 14@xmath35 mag@xmath54 .
however , the observational upper limit of polarization efficiency , is found to be @xmath55 3@xmath40mag@xmath56 a factor of four less than the predicted value for the ideal scenario @xcite .
figure [ fig6 ] ( a ) , [ fig7](a ) and [ fig8](a ) are the plot of @xmath57 vs @xmath20 for the three clouds cb56 , cb60 and cb69 .
it can be seen that the value of @xmath58 for all the stars lies below the line , drawn by using the relation : @xmath59 ( mag ) .
vs @xmath20 ( where @xmath60 = 3.1 ) .
the solid line in the figure represents upper limit of polarization efficiency ( @xmath20=3.0 a@xmath21 ) , and the dotted line is the unweighted best fit line of the field stars of cb56 .
( b ) lower panel is the plot of a@xmath21 vs @xmath20/a@xmath21 ( where @xmath60 = 3.1 ) .
the solid line is the unweighted best fit power law ( @xmath61 ) for the background field stars of cb56 .
, title="fig : " ] + vs @xmath20 ( where @xmath60 = 3.1 ) .
the solid line in the figure represents upper limit of polarization efficiency ( @xmath20=3.0 a@xmath21 ) , and the dotted line is the unweighted best fit line of the field stars of cb60 .
( b ) lower panel is the plot of a@xmath21 vs @xmath20/a@xmath21 ( where @xmath60 = 3.1 ) .
the solid line is the unweighted best fit power law ( @xmath62 ) for the background field stars of cb60 . ]
vs @xmath63 ( where @xmath60 = 3.1 ) .
the solid line in the figure represents upper limit of polarization efficiency ( @xmath64=3.0 a@xmath21 ) , and the dotted line is the unweighted best fit line of the field stars of cb69 .
( b ) lower panel is the plot of a@xmath21 vs @xmath65 @xmath66 ( where @xmath60 = 3.1 ) .
the dotted line is the unweighted best fit power law ( @xmath67 ) for the background field stars of cb69 . ]
several investigators ( @xcite , @xcite , @xcite , @xcite and @xcite ) in past showed that the efficiency of polarization decrease systematically with the increase in extinction and it can be represented by a power law @xmath68/a@xmath69 for some index @xmath70 . figure [ fig6 ] ( b ) , [ fig7](b ) and [ fig8](b )
are the plots of a@xmath71 vs @xmath72 a@xmath73 for the three clouds cb56 , cb60 and cb69
. the unweighted power law fits to all the background field stars in the cloud cb56 , cb60 and cb69 are given by @xmath74 @xmath75 @xmath76 these power law explain the dependence of polarization efficiency on visual extinction .
our calculated power laws are consistent with the findings of @xcite and @xcite for tdc .
it is to be noted that our analysis is based on v - band data set whereas their findings are based on k - band data set .
it may be concluded that the linear polarization of the starlight due to aligned dust grains in these three clouds is more efficient in regions of low extinction as compared to high obscured lines of sight .
thus the characteristics of material composing these bok globules are consistent with the material found in diffuse interstellar medium ( ism ) .
however present analysis is restricted to limited number of data points .
1 . three bok globules ( cb56 , cb60 and cb69 ) have been probed at the v - band , in order to map the magnetic field within the observed region of the cloud .
the distribution of @xmath77 and @xmath37 in the v - band are studied for these clouds .
it has been found that the polarization vectors of most of the stars in these globules are aligned in some common direction which indicate the direction of local magnetic field in the globules .
the local magnetic field of the cloud cb56 ( @xmath78 = 237.92 @xmath79 , @xmath80 ) is almost aligned with the galactic field whereas the alignment is not observed in cb60 ( @xmath78 = 248.89 @xmath79 , @xmath81 ) and cb69 ( @xmath78 = 351.23 @xmath79 , @xmath82 = 5.14@xmath38 ) .
@xmath18 for the observed field stars are calculated from the observed bvr magnitudes .
these values are estimated for @xmath83 = 3.1 . from the estimated @xmath18 ,
the probable location of each field stars is identified .
the polarization efficiency of these clouds is also studied , which shows decrease in polarization efficiency with an increase in extinction along the observed line of sight .
this indicates that the non spherical dust grains in these globules , is consistent with dust grains found in the diffuse interstellar medium .
it is to be noted that the decrease in polarization could , in some cases , result from a change in the direction of the magnetic field along the line of sight , not necessarily a decrease in alignment efficiency . also power laws calculated for cb56 , cb60 and cb69 are in good agreement with the findings of both @xcite and @xcite for tdc .
we gratefully acknowledge iucaa , pune for making telescope time available .
the reviewers of this paper are highly acknowledged for their constructive comments which definitely helped us to improve the quality of the paper .
we also acknowledge the use of the vizier database of astronomical catalogues namely ucac4 @xcite , nomad @xcite , denis database @xcite .
this work is supported by the department of science & technology ( dst ) , government of india , under serc - fast track scheme ( dy . no . serb / f/1750/2012 - 13 ) . | we present the polarization maps of three selected bok globules cb56 , cb60 and cb69 constructed using a v - band data from a ccd imaging polarimeter . the aim of this work is to measure the optical polarization ( @xmath0 ) of background field stars in order to determine the polarization efficiency , @xmath1 .
we find that the local magnetic field of the cloud cb56 is almost aligned with the galactic field , but not in cb60 and cb69 .
a trend of decreasing polarization efficiency with increasing extinction ( @xmath2 ) is observed : it can be well represented by a power law , @xmath3 , where @xmath4 , @xmath5 and @xmath6 for cb56 , cb60 and cb69 respectively .
this indicates that the linear polarization of the starlight due to aligned dust grains in these clouds is produced more efficiently in low extinction regions , compared with high obscured lines of sight .
[ firstpage ] polarization
ism : clouds dust , extinction ism : individual objects : cb56 ism : individual objects : cb60 ism : individual objects : cb69 |
mesons and baryons have been intensively studied in the framework of potential models for a long time @xcite . moreover , since qcd is a nonabelian gauge theory , the gluons are able to form a bound state , at least in principle .
these bound states of two , three or more gluons are called glueballs .
the theoretical computation of the glueball spectrum is a problem which currently deserves much interest , either in lattice qcd @xcite , or within the framework of effective models @xcite . other so - called
exotic " hadrons are hybrid mesons : mesons in which the color field is in an excited state .
they can be seen as a three - body @xmath0 state , with a constituent gluon .
again , hybrid mesons are the subject of many works , in lattice qcd @xcite , and effective models @xcite .
the experimental detection of exotic hadrons is also an active field of research , although no definitive conclusion has been drawn @xcite .
the confrontation between theoretical approaches and experimental data appears thus particularly interesting in order to confirm or invalidate some exotic candidates .
let us now turn our attention to effective qcd models .
it is a well - known fact that the short - range interaction between two heavy quarks is given by the fermi - breit potential @xcite .
this corresponds to the one - gluon exchange process between the quarks , and contains fine structure terms like spin - orbit interaction and tensor force for example .
effective models with a linear confining potential and the fermi - breit potential have been showed to correctly describe the meson and baryon spectrum for a long time @xcite .
apart from phenomenological arguments , the static potential between a quark and an antiquark has also been intensively studied in numerous theoretical works . at large distances , typically for distances larger than @xmath1 gev@xmath2 , the static potential indeed appears to be linearly rising ( see for example the lattice calculations of ref .
@xcite ) . at short distances
however , the static potential has been shown to be no longer linear and nonperturbative but rather perturbative and coulombic , just as is the static limit of the fermi - breit interaction @xcite .
logically , the same kind of models linear confinement and coulomb term at long- and short - range respectively has since been applied to describe exotic hadrons , like glueballs @xcite and hybrids @xcite . however ,
if the derivation of the fermi - breit potential between two quarks is a standard procedure , it is not the case for the interactions between two gluons ( see for example the discrepancies between refs .
moreover , to our knowledge , the equivalent of the fermi - breit potential between a quark and a gluon has not yet been computed . since recent references , based on lattice qcd computations ,
validate the picture of a glueball as a bound state of constituent gluons @xcite and of a hybrid meson as a @xmath0 system @xcite , the accurate knowledge of the gluon - gluon and quark - gluon potentials appears to be crucial point to further understand the properties of these exotic hadrons .
it is thus of interest to compute all these potentials in a systematic way , within a single formalism , in order to obtain a set of potentials which should be applicable to most hadrons , either exotic or not .
these computations are performed in this work .
our paper is organized as follows . in sec .
[ gene ] , we introduce the general procedure for extracting the effective potential corresponding to a particular matrix element . before applying it to all the possible second order qcd - diagrams related to bound states
, we have to know the non relativistic wave functions of the _ in _ and _ out _ states .
this is presented in sec .
[ inout ] .
then , we compute the effective potentials for the diagrams involving pairs of quarks and antiquarks in sec .
[ qqint ] . in sec .
[ ggint ] , we consider the interactions between two gluons , and the interactions between ( anti)quarks and gluons are treated in sec .
[ qgint ] .
the exchanged gluons are massless in our formalism . since it is not the case in some approaches
( see refs .
@xcite ) , we then discuss in sec .
[ applic ] the influence of a possible nonzero mass for the exchanged gluons . in sec .
[ valid ] , we discuss the validity of the potentials which are computed .
since we want our potentials to be relevant in various systems
exotic or non exotic hadrons , we compute in sec .
[ colfact ] the color factors appearing in the various potentials we obtain .
finally , we draw some conclusions in sec .
[ conclu ] .
we also added two appendices , where we recall the feynman rules for qcd ( appendix [ feyn ] ) and some useful fourier transforms ( appendix [ fourier ] ) .
let us consider a general feynman diagram of order @xmath3 , with @xmath4 the strong coupling constant + @xmath5 denote the color indices of the particles @xmath6 , @xmath7 , @xmath8 , and @xmath9 . as the goal of this work is to compute effective potentials for bound states of quarks , antiquarks , or gluons , we will restrict ourselves to diagrams where the initial and final particle content are the same : @xmath10 and @xmath11 .
the application of feynman rules in momentum space , summed up in appendix [ feyn ] , leads us formally to the following matrix element ( in units where @xmath12 ) @xmath13 where @xmath14 is the exchanged @xmath15-momentum , and @xmath16 are the properly normalized wave functions of the _ in _ and _ out _ states . in our notations ,
@xmath17 label the particles , @xmath18 , and @xmath19 .
these wave functions can be schematically decomposed as @xmath20 with @xmath21 the color wave function , and @xmath22 the spatial wave function in momentum space , @xmath23 being the @xmath15-momentum of the particle .
the explicit form of these wave functions in the nonrelativistic limit will be given in sec .
[ inout ] .
then , the crucial and most complex step of the procedure is to rewrite @xmath24 as @xmath25\psi^a_{{\rm i},1}\psi^b_{{\rm i},2}.\ ] ] @xmath26 and @xmath27 are the momentum and spin of the particle @xmath28 . in the latter ,
bold quantities will always denote a vector .
the spin operators appear through the rearrangement of the wave functions , and @xmath29 is a color operator emerging from the vertices .
we will show in sec .
[ colfact ] that this operator turns out to be a real number when correctly contracted with the color functions of the particles , @xmath21 .
we denote @xmath30 this color factor .
once the decomposition ( [ mif ] ) is known , the effective potential we want to find is readily obtained in position space by a fourier transform , denoted as @xmath31 , which changes @xmath32 in @xmath33 , the relative position of the two interacting particles : @xmath34.\ ] ] we used the definition @xmath35 , which has the advantage of simplifying the fourier transforms .
all the fourier transforms we need are listed in appendix [ fourier ] .
we have at last to make a remark which will be valid in all our work : the effective potentials , and all the other quantities appearing in the latter , will be computed at the order @xmath36 , with @xmath37 the speed of light .
this is a standard approximation scheme , justified because the most relevant contributions as spin - orbit , tensor interaction , , arise from these first relativistic corrections .
quarks and antiquarks are spin-@xmath38 particles , so they are described by dirac spinors , denoted as @xmath39 for the quarks and @xmath40 for the antiquarks , with @xmath41 the momentum of the particle and @xmath42 their color indices . the wave function of a dirac spinor , whose mass is @xmath43 , can be computed at the order @xmath36 thanks to a foldy - wouthuysen transformation @xcite . in our notations , it can be recast in the form ( [ fodef ] ) , and reads @xmath44 with @xmath45 the color functions , and @xmath46 where @xmath47 is a two component spinor and @xmath48 are the pauli matrices ( see appendix [ feyn ] ) . in order to rewrite the matrix element on the form ( [ mif ] ) , we will have to isolate a factor @xmath49 for each wave function .
the following useful relations hold for both @xmath50 and @xmath51 @xcite [ rel_q ] @xmath52 with @xmath53 , @xmath54 .
an important relation between the pauli matrices is @xmath55 in which @xmath56 and @xmath57 are two vectorial operators .
the spin operator of a quark or an antiquark is defined by @xmath58 . a massive , colored
, spin-@xmath6 particle is described by the wave function @xmath59 where @xmath60 is a polarisation vector satisfying the proca equation with the transverse condition @xmath61 relation ( [ gluc ] ) together with the normalisation condition @xmath62 leads to the nonrelativistic expression @xmath63 with @xmath64 in agreement with refs .
the polarization vector @xmath65 does not depend on the momentum @xmath41 anymore .
the factor to isolate in the matrix element will now be given by @xmath66 , by analogy with the dirac spinors .
the spin operators for a spin-@xmath6 particle are defined by the matrices @xmath67 , where @xmath68 @xmath69 is the levi - civita symbol .
the following useful relation involves the spin-@xmath6 matrices @xmath70_{ij}.\ ] ] for qcd applications as those we consider here , the spin-@xmath6 particles which have to be taken into account are the gluons .
these are massless particles , so , one can wonder if formula ( [ epsdef1 ] ) makes sense . in our framework
however , exchanged gluons , which are true massless gauge particles , have to be distinguished from the constituent gluons of the @xmath71 and @xmath72 states .
these ones are confined inside a hadron , and one can consider that even if the bare gluon mass is zero , they acquire a nonzero constituent mass due to the confining interaction .
@xmath43 has thus here to be understood as the constituent gluon mass in a particular bound state , which typically has a value around @xmath73 gev @xcite .
this argument has already been used in refs .
@xcite , and will be further developed in sec .
[ bfapp ] .
let us note that the same argument holds in the case of the massless @xmath74 and @xmath75 quarks , whose constituent mass is around @xmath76 gev @xcite .
as an illustration of the general method presented in the previous section , we begin by the best - known diagram , that is the one - gluon exchange between two quarks . applying the feynman rules of appendix [ feyn ] , we can write the matrix element corresponding to the diagram in the form ( [ mif0 ] ) , which is @xmath77 the color operator is given by @xmath78 and the conserved current of the quarks by @xmath79 we omitted to note the color functions in order to simplify the notations . the explicit computation of the color factors will be done in sec .
[ colfact ] .
let us call @xmath80 the color factor corresponding to @xmath81 .
as the current @xmath82 is obviously known thanks to the relations ( [ rel_q ] ) , we only need to specify the form of the propagator @xmath83 in order to develop eq .
( [ qqstep1 ] ) .
let us note that @xmath84 an interesting gauge choice is then to set @xmath85 in the propagator ( [ ddef ] ) .
this leads to the expressions @xmath86 putting all these elements together , one can rewrite eq .
( [ qqstep1 ] ) in the form ( [ mif ] ) , i. e. @xmath87w_1w_2,\ ] ] with @xmath88 in the latter , the @xmath37 factors will not be written anymore in order to simplify the notations .
the last step is to compute the fourier transform of @xmath89 , as formulated in eq .
( [ pot_four ] ) . by using the relations of appendix [ fourier ] , we obtain the following effective potential @xmath90-\frac{1}{m_1m_2r^3}\left[\bm l_1\cdot\bm s_2-\bm l_2\cdot\bm s_1\right]\nonumber\\ & & + \frac{1}{m_1m_2r^3}\left[\bm s_1\cdot\bm s_2 - 3\frac{(\bm s_1\cdot\bm r)(\bm s_2\cdot\bm r)}{r^2}\right]-\left.\frac{8\pi}{3m_1m_2}\bm s_1\cdot\bm s_2\delta^3(\bm r)\right\},\end{aligned}\ ] ] where @xmath91 .
this is the well - known fermi - breit interaction @xcite .
it is interesting to rewrite this contribution by using [ change ] @xmath92 which are simply the relative and center of mass variables of the interacting pair , and also by introducing new spin variables as follows
@xmath93 potential ( [ pot_1 ] ) then becomes @xmath94\right.\nonumber\\ & & -\left(\frac{m^2_2+m^2_1 + 4m_1m_2}{4m_1 ^ 2m_2 ^ 2}\right)\frac{\bm l\cdot\bm s } { r^3}+\frac{1}{4m_1m_2}\frac{\bm \delta\cdot\bm \omega}{r^3}+ \left(\frac{m_1 ^
2-m_2 ^ 2}{4m_2 ^ 2m_1 ^ 2}\right)\frac{\bm l\cdot\bm \delta}{r^3}\nonumber\\ & & + \frac{(m_2-m_1)}{4(m_1+m_2)m_1m_2}\frac{\bm s\cdot\bm \omega}{r^3}+\left.\frac{1}{2m_1m_2}\left[\frac{\bm s^2}{r^3}-3\frac{(\bm s\cdot\bm r)^2}{r^5}\right]\right\}.\end{aligned}\ ] ] it can be checked that the skewsymmetric spin - orbit interaction vanishes for two particles of equal masses . indeed , in this case
, the wave function must either have a good symmetry under the exchange @xmath95 if both particles are identical , or a good @xmath96-parity for a particle - antiparticle pair .
in a two - body system , @xmath97 can be imposed by working in the center of mass frame .
if we considered an antiquark interacting with a quark or an antiquark , we would have found the same expression ( [ pot_1 ] ) , but with a different color factor , since the color operator would be different .
these possible color operators are listed in eq .
( [ color : qq ] ) , and denoted as @xmath98 respectively . in the particular case of the interaction between a quark and an antiquark of the same flavor ,
the contribution of a new diagram must be added , that is the annihilation diagram : the flavor of the outcoming pair is assumed to be identical to the incoming one . with the same indices as for the scattering diagram , we find @xmath99 d_{\mu\nu}\left[\bar u'_1\gamma^\nu\ , v'_2\right ] , \end{split}\ ] ] with @xmath100 the independence of the matrix element ( [ qqbstep1 ] ) in @xmath101 is readily checked .
consequently , the relevant part of the gluon propagator in the @xmath102-channel is given by @xmath103 because we can suppose that the momentum of the created gluon is very smaller that its rest energy .
the wave functions ( [ uvdef ] ) can be taken at the lowest order , that is @xmath104 after a rearrangement of formula ( [ qqbstep1 ] ) thanks to the relation @xmath105 one finds @xmath106 \left(w_2\right)^{\rm c } w_1.\ ] ] @xmath107 has been replaced by the corresponding factor @xmath108 , and we defined the charge conjugation as @xmath109 after a fourier transform , we get the annihilation potential @xmath110\delta^3(\bm r),\\ & = & { \cal c}_4 \frac{\pi\alpha_s}{m^2 } \bm s^2\delta^3(\bm r),\end{aligned}\ ] ] which has logically the form of a contact interaction .
moreover , it is a projector on a spin-@xmath6 state : the quark - antiquark pair must have the quantum numbers of a gluon in order to make the annihilation possible .
the color factor @xmath108 is indeed also a projector , on the octet state ( see table [ tab : color : qq ] ) .
let us note that our potential ( [ pot_2 ] ) is in agreement with the one of ref .
@xcite .
the spin - dependent corrections coming from qcd are computed with the background perturbation theory in ref .
@xcite for the @xmath111 pair in a meson . in our notations
, the result reads @xmath112\left[\frac{1}{r}\partial_r \epsilon(r)+\frac{2}{r}\partial_r v_1(r)\right]+\frac{\bm s_2\cdot\bm l_1-\bm s_1\cdot\bm l_2}{\mu_1\mu_2}\ \frac{1}{r}\partial_r v_2(r ) \\&+\frac{\bm s_1\cdot\bm s_2}{3\mu_1\mu_2}v_4(\bm r)+\frac{1}{3\mu_1\mu_2}\left[3\frac{(\bm s_1\cdot\bm r)(\bm s_2\cdot\bm r)}{r^2}-\bm s_1\cdot\bm s_2\right]v_3(r ) , \end{split}\ ] ] where @xmath113 and @xmath114 are rather complicated functions of the quark correlators .
fortunately , at the order @xmath3 they take simpler expressions , which are @xmath115 one can check that , by injecting the relations ( [ corr_def ] ) into the potential ( [ v_simo ] ) , one formally obtains the spin - dependent part of the fermi - breit potential ( [ pot_1 ] ) for a meson , that is for a color factor equal to @xmath116 .
the only difference is that , in the background field theory , the bare masses @xmath117 are systematically replaced by @xmath118 , which are interpreted as dynamical , or constituent quark masses . if one sees the potentials we are computing as perturbations of an effective hamiltonian ( typically a spinless salpeter one with linear confining potential ) , the @xmath118 can be defined as @xcite @xmath119 where the averages are computed with the unperturbed wave functions .
the advantage of this approach is that the @xmath118 are always nonzero ; the development of the matrix elements in powers of @xmath120 is thus relevant , even for light particles .
consequently , the equivalence between potentials ( [ pot_1 ] ) and ( [ v_simo ] ) , suggest that in all our effective potentials , the quark masses should be seen as the dynamical ones , the @xmath118 defined by eq .
( [ mudef ] ) .
the same arguments also hold for gluons : even if the exchanged gluons are seen as massless gauge particles , it is usually considered that the constituent gluons , in a glueball for example , acquire a mass because of the confining interaction .
this point of view can also be checked with the background perturbation theory ( see ref .
@xcite ) , and has already given good results in the description of glueballs @xcite .
thus , in our case , the gluon mass " appearing in eq .
( [ epsdef1 ] ) has to be understood as the constituent one .
the computation of the effective potential coming from the scattering diagram between two gluons , is rather similar to the one of the scattering diagram between two quarks .
the analog of formula ( [ qqstep1 ] ) is now @xmath121 with the color operator @xmath122 and the conserved current @xmath123\epsilon^{\mu}_\alpha\nonumber\\ & & + \left[(q - p'_\alpha)\cdot\epsilon_\alpha\right]\epsilon^{*\mu}_{\alpha } .\end{aligned}\ ] ] with the aid of the definitions ( [ epsdef1 ] ) and ( [ spin1def ] ) , and the property @xmath124 we can write the different components of @xmath125 as @xmath126\right\}\bm \epsilon_\alpha\\ \bm g_\alpha(\bm q)&=&2p_\alpha^0\left[-\frac{1}{2mc}(2\bm p_\alpha-\bm q)\cdot\bm\epsilon_{\alpha}^{*}\ , \bm\epsilon_\alpha+\frac{\bm q\cdot\bm\epsilon^*_{\alpha}}{mc}\ \bm\epsilon_\alpha - \frac{\bm q\cdot\bm\epsilon_\alpha}{mc}\ , \bm\epsilon^*_{\alpha}\right].\end{aligned}\ ] ] as the gluon propagator is given by relations ( [ propadef ] ) , the matrix element ( [ mscatdiag ] ) becomes @xmath127 , \end{split}\ ] ] where we used the obvious notation @xmath128 and @xmath129 . by simplification of eq .
( [ mscatdia2 ] ) , and with formula ( [ spin1prop ] ) , one can arrive at the analog of eq .
( [ qqstep2 ] ) , that is @xmath130\bm\epsilon_1\bm\epsilon_2,\ ] ] with @xmath131 the constituent gluons are assumed to have the same effective mass : this is clear because of the symmetry of the problem .
the effective potential in position space is thus @xmath132\right.\nonumber\\ & & -\frac{1}{m^2r^3}\left [ \bm l_1\cdot\bm s_2 - \bm l_2\cdot\bm s_1\right]+ \frac{1}{m^2}\left[\frac{\bm s_1\cdot\bm s_2}{r^3}-3\frac{(\bm s_1\cdot\bm r)(\bm s_2\cdot\bm r ) } { r^5}\right]-\frac{8\pi}{3m^2}\bm s_1\cdot\bm s_2\delta^3(\bm r)\nonumber\\ & & \left.+\frac{2}{m^2r^3}-\frac{3}{2m^2r^5}\left[(\bm s_1\cdot\bm r)^2+(\bm s_2\cdot\bm r)^2\right]\right\}.\end{aligned}\ ] ] this potential is very similar to ( [ pot_1 ] ) , except in two terms .
the first one is the contact interaction , which is stronger than in the quark - quark case , and the second one is , at first sight , a new interaction term , written in the last line of eq .
( [ pot_3 ] ) .
the existence of this term has already been pointed out in ref .
@xcite , but here we go further than this reference because we completely remove the @xmath133 from our expression . actually , these terms are a part of the tensor force , because if we rewrite the potential ( [ pot_3 ] ) with the new variables defined by ( [ change ] ) , we obtain @xmath134\right.\nonumber\\ & & -\frac{3}{2m^2}\frac{\bm l\cdot\bm s } { r^3}+\frac{1}{4m^2}\frac{\bm \delta\cdot\bm \omega}{r^3}+\left.\frac{1}{2m^2}\left[\frac{\bm s^2}{r^3}-3\frac{(\bm s\cdot\bm r)^2}{r^5}\right]\right\}.\end{aligned}\ ] ] the annihilation of two gluons , represented by the diagram is a special case because its contribution is always zero .
first of all , the color operator associated to this diagram is @xmath135 it is shown in sec .
[ colfactc ] that the corresponding color factor is non zero only if the gluon pair is in an antisymmetric color octet .
but even in this case , a calculation similar to the one performed in sec .
[ anniqq ] shows that the matrix element is zero at the order @xmath36 .
indeed , as the gluon propagator in the @xmath102-channel can be taken of the form ( [ ddef2 ] ) , the wave functions and momenta will simply be given by their lowest order approximation @xmath136 respectively .
this approximation causes the matrix element to vanish in every case , as it is mentioned in ref .
@xcite .
the seagull diagram possesses a unique vertex : the four - gluon one .
applying the corresponding feynman rules together with the jacobi identity @xmath137 we obtain the following matrix element @xmath138 an additional @xmath38 factor is present in eq .
( [ sgstep1 ] ) .
this is due to the degeneracy of the different diagrams . indeed ,
if we attribute a weight @xmath6 to the gluon - gluon scattering diagram , we have to attribute a weight @xmath38 to the annihilation and to the seagull diagrams @xcite .
the effective potential corresponding to the seagull diagram should clearly be a contact - like interaction .
it is thus sufficient to inject the lowest - order expressions ( [ epslow ] ) to simplify eq .
( [ sgstep1 ] ) .
the following relations have also to be used [ propsg ] @xmath139 & = & \bm\epsilon^*_\alpha\bm s^2_\alpha\bm\epsilon_\alpha.\end{aligned}\ ] ] moreover , @xmath140 since such terms are higher order relativistic corrections .
the effective potential in momentum space reads @xmath141
+ { \cal c}_5 \left[\bm s_1\cdot\bm s_2 - 2\right]\right\}.\ ] ] introducing the total spin of the system , we get in position space @xmath142 + { \cal c}_5 \left[\frac{1}{2}\bm s^2 - 4\right]\right\}\delta^3(\bm r ) . % \end{split}\ ] ] the complete potential to use to describe the interactions between two gluons is thus the sum of potentials ( [ pot_3bis ] ) and ( [ pot_sg ] ) .
the understanding of hybrid mesons currently deserves much interest in experimental as well as in theoretical physics . from a quark model point of view
, hybrid mesons can be seen as a quark - antiquark - constituent gluon bound state . in the framework of a potential model ,
the dominant part of the interaction is a confining term which is typically given by @xmath143 @xcite .
then , short - range interactions should be added in order to build a spin - dependent model .
the quark - antiquark potential has been computed in sec .
[ qqint ] ; in this case , the @xmath111 pair is in a color octet and couples to the constituent gluon in order for the total system to be in a color singlet .
but , the short - range potentials between the constituent gluon and the quark or the antiquark have , up to our knowledge , never been computed .
consequently , the knowledge of ( anti)quark - gluon interactions is a missing element in the building of accurate potential models of hybrid mesons .
that is why we compute these interactions in the present section , applying the same procedure as the one used up to now .
there are three diagrams to consider when one deals with quark - gluon interactions @xcite .
we begin with the following scattering diagram , in the following , the indices @xmath144 will be replaced by @xmath145 in order to clearly distinguish between the symbols related to the gluon and to the quark respectively . applying the feynman rules , we can write the corresponding matrix element in form , @xmath146 where the color factor is @xmath147 and where @xmath148 and @xmath149 were already computed in eqs . and
. using the propagator ( [ propadef ] ) , the matrix element can rewritten as @xmath150.\end{aligned}\ ] ] after the same rearrangements as those which were done in secs .
[ qqint ] and [ ggint ] , eq .
( [ qgmfi ] ) becomes @xmath151\bm\epsilon w,\ ] ] with @xmath152 one can check that the quark part of potential ( [ pot_40 ] ) is equal to the one of eq .
( [ eq22 ] ) , and that the gluon part is the same as in eq .
( [ inter1 ] ) , which is rather coherent .
the last term of this expression is actually not explicitly written in the quark - quark case .
it is indeed eliminated thanks to the properties of the pauli matrices through the relation @xmath153 consequently , it is only a constant term , which will be part of the contact interaction after a fourier transform . in the gluon - gluon scattering , the gluonic term in @xmath154
was naturally included in the tensor force . in the quark - gluon case ,
the masses @xmath155 are not equal , and this term has thus to be treated separately . in position space ,
the effective potential is given by @xmath156-\frac{1}{m_gm_qr^3}\left[\bm l_g\cdot\bm s_q-\bm l_q\cdot\bm s_g\right]\nonumber\\ & & + \frac{1}{m_gm_q}\left[\frac{\bm s_g\cdot\bm s_q}{r^3}-3\frac{(\bm s_g\cdot\bm r)(\bm s_q\cdot\bm r)}{r^5}\right]\left .
+ \frac{1}{m_g^2r^3}-\frac{3}{2m_g^2}\frac{(\bm s_g\bm r)^2}{r^5}\right\}.\end{aligned}\ ] ] rewriting it with the appropriate variables we have @xmath157\right.\nonumber\\ & & + \pi\left(\frac{11}{3m_gm_q}-\frac{2}{3m^2_g}-\frac{1}{2m^2_q}-\frac{4}{3m_gm_q}\bm s^2\right)\delta^3(\bm r)-\left(\frac{m^2_q+m^2_g+4m_gm_q}{4m_g^2m_q^2}\right)\frac{\bm l\cdot\bm s } { r^3}\nonumber\\ & & + \frac{1}{4m_gm_q}\frac{\bm \delta\cdot\bm \omega}{r^3}+\frac{(m_q - m_g)}{4m_qm_g(m_q+m_g)}\frac{\bm s\cdot\bm \omega}{r^3 } + \left(\frac{m_g^2-m_q^2}{4m_q^2m_g^2}\right)\frac{\bm l\cdot\bm \delta}{r^3}\nonumber\\ & & \left.+\frac{1}{2m_gm_q}\left[\frac{\bm s^2}{r^3}-3\frac{(\bm s\cdot\bm r)^2}{r^5}\right]+\frac{m_q - m_g}{m^2_gm_q}{\cal q}(\bm r)\right\},\end{aligned}\ ] ] with @xmath158.\ ] ] the physical meaning of this term can be clarified by using the following relation involving the spin-@xmath6 matrices @xcite @xmath159 @xmath160 are the polarisation matrices , defining the quadrupole tensor .
consequently , eq . ( [ q1 ] ) can be rewritten as @xmath161 a quadrupolar interaction is thus present in the potential .
such a term is logically absent in symmetric configurations , i.e. quark - quark and gluon - gluon systems , but appears in this asymmetric quark - gluon system .
the two remaining diagrams are analog to those of the compton scattering in qed .
we begin by the following annihilation diagram , whose matrix element reads @xmath162 where the color factor is @xmath163 the quark propagator reads @xmath164 in the approximation @xmath165 .
@xmath166 is the mass of the propagating quark , which we choose to be equal to @xmath167 . using the lowest order approximations for the wave functions and the momenta , and thanks to formula @xmath168 we find @xmath169 w\bm\epsilon.\ ] ] the effective potential , @xmath170\delta^3(\bm r),\ ] ]
is a projector on a spin @xmath38 state because the quark - gluon system must have the quantum numbers of a quark in order to form a single quark .
if one now considers an antiquark instead of a quark , the only modification in the potential ( [ qgpot2 ] ) will be the color operator , given by @xmath171 the second compton - like diagram is this diagram is no longer an annihilation one .
it can indeed be seen as an exchange of a(n ) ( anti)quark between a ( anti)quark - gluon pair .
we will thus work in breit s frame , as in the case of the usual scattering diagrams : @xmath172 .
the exchanged quark has a constituent mass @xmath167 , which is always nonzero in our formalism .
consequently , we expect to find an expression in which a yukawa potential appears , whose contribution will in every case be less important than the one of the quark - gluon scattering diagram since the exchanged gluons are massless .
this leads us to assume that only the lowest order expression of the effective potential will be relevant .
one obtains easily for this potential @xmath173\frac{{\rm e}^{-m_q r}}{r},\ ] ] @xmath174 corresponding to the color operator @xmath175 let us remark that if one deals with a quark instead of an antiquark in this diagram , the only change in potential ( [ qgpot3 ] ) will be again the color operator , given by @xmath176
in the computation of all the potentials that we presented up to now , the constituent gluons were considered as massive particles because of confinement , but the exchanged gluons remained massless gauge particles . the advantage of this assumption , already used in ref .
@xcite , is to treat the diagrams involving constituent ( anti)quarks and/or gluons on an equal footing .
moreover , it allows to compute the potentials with the standard feynman rules corresponding to the qcd lagrangian .
however , in some other approaches related to potential models of glueballs , it is argued that the exchanged gluons have to be massive as well in order for all the gluons to have an equal status @xcite . in these references , consequently , a gluonic mass term has to be added to the qcd lagrangian . for completeness
, we discuss here the consequences of such a term .
actually , the feynman rules will not be modified at the level of the gluon propagator : the coulomb form in position space , denoted hereafter as @xmath177 , will turn into the propagator of a massive vector particle , typically @xmath178 .
thus , at the first order in the relativistic corrections , only the scattering diagrams will be affected by this modification .
the corresponding scattering potentials , denoted here as @xmath179 , and given by eqs .
( [ pot_1bis ] ) , ( [ pot_3bis ] ) , and ( [ pot_4 ] ) , can be rewritten as @xmath180\right.\nonumber\\ & & + \left(\frac{m^2_2+m^2_1 + 4m_1m_2}{4m_1 ^ 2m_2 ^ 2}\right)\bm l\cdot\bm s \
\frac{u'}{r}-\frac{1}{4m_1m_2}\bm \delta\cdot\bm \omega\ \frac{u'}{r}- \left(\frac{m_1 ^ 2-m_2 ^ 2}{4m_2 ^ 2m_1 ^ 2}\right)\bm l\cdot\bm \delta\ \frac{u'}{r}\nonumber\\ & & -\frac{(m_2-m_1)}{4(m_1+m_2)m_1m_2}\bm s\cdot\bm \omega\ \frac{u'}{r}+\left.\frac{1}{6m_1m_2}\left(u''-\frac{u'}{r}\right)\left[\bm s^2 - 3\frac{(\bm s\cdot\bm r)^2}{r^2}\right]\right\},\end{aligned}\ ] ] @xmath181\right.\nonumber\\ & & + \frac{3}{2m^2}\bm l\cdot\bm s\ \frac{u'}{r}-\frac{1}{4m^2}\bm \delta\cdot\bm \omega\ \frac{u'}{r}\left.+\frac{1}{6m^2}\left(u''-\frac{u'}{r}\right)\left[\bm s^2 - 3\frac{(\bm s\cdot\bm r)^2}{r^2}\right]\right\},\end{aligned}\ ] ] @xmath182\right.\nonumber\\ & & -\left(\frac{11}{12m_gm_q}-\frac{1}{6m^2_g}-\frac{1}{8m^2_q}-\frac{1}{3m_gm_q}\bm s^2\right)\bm\nabla^2u\left(\frac{m^2_q+m^2_g+4m_gm_q}{4m_g^2m_q^2}\right)\bm l\cdot\bm s \ \frac{u'}{r}\nonumber\\ & & -\frac{1}{4m_gm_q}\bm \delta\cdot\bm \omega\ \frac{u'}{r}-\left(\frac{m_g^2-m_q^2}{4m_q^2m_g^2}\right)\bm l\cdot\bm \delta\ \frac{u'}{r}-\frac{(m_q - m_g)}{4(m_g+m_q)m_gm_q}\bm s\cdot\bm \omega\ \frac{u'}{r}\nonumber\\ & & + \frac{1}{6m_gm_q}\left(u''-\frac{u'}{r}\right)\left[\bm s^2 - 3\frac{(\bm s\cdot\bm r)^2}{r^2}\right]\left .
-\frac{m_q - m_g}{2m^2_gm_q}\left(u''-\frac{u'}{r}\right)\frac{\bm r q\bm r}{r^2}\right\}.\end{aligned}\ ] ] the symmetrizations on the noncommuting operators have not been explicitly written in order to clarify to formula .
it appears that all the scattering potentials we compute almost share the same structure , the only difference coming in the coefficients of @xmath183 , typically responsible for the contact interaction .
the quadrupolar term is trivially zero except for the ( anti)quark - gluon interactions . in this last case
, we denoted the gluon and the ( anti)quark masses by @xmath184 and @xmath167 respectively , in order to avoid confusion .
it is worth mentioning that , if @xmath185 is the coefficient of @xmath183 in the scattering potential @xmath186 , it can be checked that @xmath187 it appears obviously that the differences only depend on the nature of the interacting particles , each quark bringing a contribution @xmath188 , while this contribution is equal to @xmath189 for a gluon .
clearly , the coulomb form of @xmath89 comes from the propagator of a massless particle .
following ref .
@xcite , the expressions ( [ sqqpot ] ) , ( [ sggpot ] ) , and ( [ sqgpot ] ) are valid for any @xmath190 , this function being the fourier transform of the propagator of an exchanged vector particle .
thus , for a massive gluon , one should simply replace @xmath190 by a yukawa potential @xmath191 .
moreover , following eq .
( [ green ] ) , @xmath192 .
various gluon - gluon potentials involving massive gluons can be found in the literature ( see for example refs .
@xcite ) , but they are neither equal between each other , nor equivalent to our potential . in recent works
@xcite , the coefficients of the different terms ( @xmath193 , @xmath194 , ... ) of the potential between two heavy quarks was found numerically by lattice qcd computations .
these fitted coefficients agree with the fermi - breit potential .
it could be interesting to find , by the same procedure , the same coefficients for two static sources in a color octet .
this could lead to identify the most relevant gluon - gluon potential .
moreover , an accurate computation of various hadrons spectra will be able to check the relevance of one approach or another , through a comparison with experimental data or with lattice qcd .
the effective potentials we computed up to now involve feynman diagrams of order @xmath3 , without loop .
it is well known however that the inclusion of the one - loop diagrams leads to the conclusion that the strong coupling constant depends on the exchanged momentum through @xmath195 where @xmath196 is the number of quark flavors whose masses are lower than @xmath197 , and where @xmath198 is the famous lambda - qcd parameter , whose value is around @xmath199 mev . clearly , @xmath200 is smaller and smaller for increasing @xmath197 , qcd becoming perturbative at high momentum . as an illustration ,
it is nowadays accepted that @xmath201 @xcite , with @xmath202 gev . for @xmath203 however , qcd is nonperturbative at low momentum since the strong coupling constant ( [ asr ] ) blows up .
this low - momentum domain precisely concerns the bound states physics we are interested in .
consequently , it appears necessary to wonder whether the potentials we obtain are valid or not to describe hadronic bound states .
lattice qcd clearly shows that the static potential between a quark and an antiquark is nicely fitted by a cornell shape , that is @xmath204 , for @xmath205 gev@xmath206 and @xmath207 @xcite .
such a form suggests that the total energy between the quarks can be roughly separated into a confining ( nonperturbative ) part at long - range ( @xmath208 ) , and a residual " , short - range , coulombic part ( @xmath209 ) . in this picture , @xmath210 should be interpreted as a small , effective , strong coupling constant .
we point out that , although the cornell potential correctly matches the data , it should only be seen as an interesting approximation of the real static potential , for example because it neglects the interplay between perturbative and nonperturbative effects at intermediate distances . moreover , the calculations of ref .
@xcite forbid the presence of a linear term at short distances .
but , the cornell potential is clearly the simplest way to interpolate between a coulomb potential at short distances and a linear confining one at long distances , both asymptotic behaviors being confirmed by theoretical studies @xcite .
anyway , the short - range part of the static potential appears to be mostly given by a perturbative coulomb term , i.e. the static approximation of all the potential we computed up to now .
let us note that strong arguments indicate that this picture is also valid not only for bound states of heavy quarks and antiquarks , but also for bound states of massless gluons @xcite , thus in the light particle sector .
several attempts have been made in order to compute @xmath200 at all orders , especially by solving the dyson - schwinger equations ( see a rather complete list of references in ref .
all these approaches qualitatively lead to @xmath211 where @xmath212 is a monotonic function such that @xmath213 and @xmath214 . however , the explicit formula giving @xmath212 is different following the different works .
equation ( [ asr2 ] ) states that the strong coupling constant remains finite for @xmath215 , and tends to a maximal value @xmath216
one speaks of freezing " of the strong coupling constant .
the value of @xmath216 is still a matter of research , but no theoretical argument forbids @xmath217 , and some studies even allow for @xmath216 to be as low as @xmath218 @xcite , just around the typical values which are used in potential models @xcite .
such a freezing scenario is coherent with the perturbative point of view that we adopted , and justifies in particular the assumption that the lowest order feynman diagrams already give a relevant picture of the short - range interactions between two confined particles . for completeness ,
it should be pointed out that , although providing a coherent framework to understand to success of potential models and how to build them , the freezing of the strong coupling constant is not universally accepted , see for example the lattice computations of ref .
@xcite in which no freezing is observed . another important ingredient of our computations is a nonrelativistic expansion of the wave functions " ( dirac spinors and polarization vectors ) in powers of @xmath219 .
such an expansion is clearly valid for heavy quarks since @xmath220 @xmath221 . in this last case
, our quark - quark potential reduces to the usual fermi - breit one , which is widely used in potential models . for light quarks and massless constituent gluons
, it is readily checked that @xmath222 , and the nonrelativistic expansion is formally no longer valid . however , although no theoretical argument is able to justify it , the extension of the fermi - breit potential to light mesons and baryons has been shown to reproduce the experimental data for a long time with a quite good agreement @xcite .
even the glueball spectrum , computed within a potential model where we used the gluon - gluon potential that we obtained in this work , is in agreement with lattice qcd when massless constituent gluons are used @xcite . by doing explicit computations , one is thus led to the empirical conclusion that the first relativistic corrections in @xmath219 contain most of the physically relevant information .
but , it is clear that light quarks are at the limit of the validity range of potential approaches . for what concerns the constituent gluons ,
it is worth mentioning that , since the work of ref .
@xcite , several studies have given theoretical arguments favoring massive constituent gluons , with a mass which could be as high as @xmath223 gev .
in this case , our gluon - gluon potentials remain valid , but with @xmath224 , as we argued in sec .
[ applic ] . at such high masses , @xmath225 becomes small enough for the nonrelativistic expansion to be justified .
the introduction of massive constituent gluons , in gluon - gluon as well as in ( anti)quark - gluon potentials , could then bypass the conceptual problems caused by the nonrelativistic expansion .
we have now to compute the value of the various color operators @xmath226 that we introduced in the previous sections . up to now
, we did not impose the two interacting particles to be in a particular representation of @xmath227 .
let us illustrate this with the @xmath228 interaction in a baryon .
the color functions of the quarks were of course chosen to be in the fundamental representation of @xmath227 , but we did not explicitly asked that the @xmath228 pair had to be in the conjugate representation in order to make a singlet when coupled to the third quark . to include this point in our discussion , we have to contract the color operators with the appropriate clebsh - gordan coefficients ( cgc ) of the _ in _ and _ out _ states .
let @xmath229 be the generic color indices of the two incoming particles , and @xmath230 those of the outcoming particles . in all our potentials
, @xmath226 has thus to be replaced by @xmath231 where @xmath232 are the cgc .
@xmath96 stands for the color index of the coupling , which is the same for the in and out states ( the color is conserved during the interaction ) . in every cases
, one can check that the result is independent of @xmath96 .
let us now compute the @xmath232 , in a similar way as it is done it ref .
let @xmath233 be the color wave function of a particle in the representation @xmath234 of @xmath227 .
in other words , @xmath235 with @xmath236 the generators of the representation @xmath237 and @xmath238 defining an infinitesimal gauge transformation ( in this section , the vectors will be denoted with an arrow for convenience ) . because the @xmath232 are general cgc that couple the representations @xmath234 and @xmath239 to form the representation @xmath240 , we can find them by requiring @xmath241= i\vec\epsilon\cdot(\vec{t}^{\bm z})^{\ d}_c \left[\phi^{\bm x}_af^{ab}_d\phi^{\bm y}_b\right].\ ] ] moreover , the cgc computed thanks to eq .
( [ cg ] ) have to be normalised with the relation @xmath242 our method can be in principle applied to any representation of @xmath227 .
but in this work , we will restrict ourselves to systems involving only the representations @xmath243 , @xmath244 , @xmath245 and @xmath246 , which are particularly relevant for hadron spectroscopy .
in particular , the quarks , antiquarks and gluons are in the @xmath244 , @xmath245 , and @xmath246 representations .
some cgc are computed in table [ tab : cgc ] .
as it is shown in the second column , they can be used to describe mesons , baryons , glueballs , hybrids , etc .
we recall that @xmath247 is the levi - civita symbol , and that the constants @xmath248 and @xmath249 are defined by the well - known relations @xmath250=2i \sum_c f^{ab}_c \lambda^c,\ ] ] @xmath251 .normalised cgc for two particles in the respective representations @xmath234 and @xmath239 of @xmath227 , coupled so that they are in the representation @xmath240 . the indices
@xmath252 take values from @xmath6 to @xmath253 , and @xmath254 from @xmath6 to @xmath255
. the color indices of the two particles are the upper indices .
the lower one is the color of their coupling . [
cols="<,<,<",options="header " , ] [ tab : color : qg ]
we have computed in this paper the effective potentials arising from second order feynman diagrams involving quarks , antiquarks , or gluons .
our expressions remain defined for the lightest particles , because the masses appearing in our potentials are seen as the constituent ones , not the bare ones .
for instance , in the nucleon , the constituent mass of the @xmath256 quarks are about @xmath199 mev . in glueballs ,
the typical value for the gluon constituent mass is @xmath257 mev @xcite .
it is worth mentioning that a recent comparison between quark models and large @xmath258 expansion of qcd also supports the development of the relativistic corrections in powers of @xmath259 @xcite .
we firstly re - obtained the well - known fermi - breit potential for two interacting quarks .
then , we computed the effective potential between two constituent gluons , and we derived an effective potential for the quark - gluon interaction and showed that in this case a quadrupolar interaction is present , due to the asymmetry of the system .
our results are valid at the order @xmath36 , which ensures us to have correct hyperfine terms , like spin - orbit , tensor force ,
as some studies argue that a nonzero mass must be confered to the exchanged gluons , we showed how to include in a simple way such a mass starting from our expressions .
whether the exchanged gluons are massive or not clearly leads to very different potentials .
the predictions of these two approaches could be compared , for example to the glueball spectra obtained in lattice qcd , in order to check their relevance . however , we stress that the goal of this work is not to decide whether the gluons are massive or not : we computed general effective potentials in a systematic way from qcd , and were led to expressions which remain valid for light particles .
furthermore , for the first time , we obtained a quark - gluon interaction potential .
as an important check of the framework that we presented here , we can mention that we recently used it to compute light and heavy meson as well as two - gluon glueball spectra @xcite .
it can be seen in this last reference that the agreement between experimental data and our theoretical spectra is very satisfactory , and that the glueball spectrum we get is mostly located in the error bars of the lattice qcd one .
consequently , we think that the various potentials which are presented in this study can be successfully used as short - range terms within potential models describing not only usual hadrons , but also exotics like two or three - gluon glueballs , and hybrid @xmath0 mesons , which currently deserve much interest . motivated by the results of ref .
@xcite , we think that our results can be seen as a basis for further computations of exotic hadron spectra , that we leave for future works .
the authors thank dr .
claude semay for many fruitful and stimulating discussions .
we recall in this section the feynman rules we use .
they are identical to those of ref .
@xcite up to some constant normalisation factors , which were changed to obtained more suitable expressions for our calculations . for the _ in _ and _ out _ states wave functions
, we have : [ feyn0 ] where @xmath23 is the @xmath260momentum of the particle . as color indices
, we used @xmath261 for a particle in the adjoint representation , and @xmath262 for a particle in the fundamental representation .
the quark propagator and the gluon propagator in lorentz gauge read [ feyn1 ] @xmath263 is defined as @xmath264,\ ] ] where @xmath101 is the gauge parameter , and @xmath265 is the minkowski metric .
@xmath266 are used as space - time indices . as usual , @xmath267 , with @xmath268
the dirac matrices , satisfying @xmath269 these matrices can be explicitly represented by @xmath270 with @xmath48 the pauli matrices .
we also need the contributions of the different vertices .
they are given by [ feyn2 ] where @xmath271 are the gell - mann matrices , and @xmath272 are the structure constants of @xmath227 .
finally , for each matrix element , we have to add a factor @xmath273 for the energy - momentum conservation . @xmath274 and @xmath275 are the initial and final @xmath15-momentum respectively .
we present in this section useful fourier transforms that are taken from ref . @xcite .
we define the fourier transform of a function @xmath276 as @xmath277= \int\frac{d^3\bm q}{(2\pi)^3}\;e^{-i\bm{q\cdot r}}g(\bm q),\ ] ] where the bold quantities always denote a vector . then , one can check that the following equalities hold @xmath278&= & \frac{{\rm e}^{-mr}}{r}\quad m\geq0 , \\
% { \cal f}\left[\frac{4\pi(\bm{a\cdot q})(\bm{b\cdot q } ) } { \bm q^4}\right]&= & \frac{1}{2r}\left[\bm{a\cdot b}-\frac{(\bm{a\cdot r})(\bm{b\cdot r})}{r^2}\right],\\ % { \cal f}\left[\frac{4\pi(\bm{a\cdot q})(\bm{b\cdot q } ) } { \bm q^2+m^2}\right]&= & -\left[(\bm{a\cdot\nabla})(\bm{b\cdot\nabla})+\frac{1}{3}(\bm{a\cdot b})\bm\nabla^2 \right ] \frac{{\rm e}^{-mr}}{r},\label{eqbiz}\\ % { \cal f}\left[\frac{4\pi i\bm{a\cdot}(\bm q\times\bm p)}{\bm q^2+m^2}\right]&=&-[\bm{a\cdot}(\bm r\times\bm p)]\frac{1}{r}\frac{\partial}{\partial r}\frac{{\rm e}^{-mr}}{r},\end{aligned}\ ] ] with @xmath279 and @xmath280 vectors or matrix vectors .
all these identities hold for massless propagators , but the limit @xmath281 has to be taken before applying the derivatives . let us recall the relation @xmath282 which is useful in the computation of eq .
( [ eqbiz ] ) . 99 a.
de rjula , h. georgi , and s. l. glashow , phys .
d * 12 * , 147 ( 1975 ) .
s. godfrey and n. isgur , phys .
d * 32 * , 189 ( 1985 ) . c. j. morningstar and m. peardon , phys .
d * 60 * , 034509 ( 1999 ) [ hep - lat/9901004 ] ; y. chen _ et al .
d * 73 * , 014516 ( 2006 ) [ hep - lat/0510074 ] . j. m. cornwall and a. soni , phys .
lett b * 120 * , 431 ( 1982 ) .
t. barnes , z. phys .
c - particles and fields * 10 * , 275 ( 1981 ) .
a. b. kaidalov and yu .
a. simonov , phys . atom
. nucl . * 63 * , 1428 ( 2000 ) ( yad .
fiz . * 63 * , 1428 ( 2000 ) ) [ hep - ph/9911291 ] .
hou , c .- s . luo and g .-
wong , phys .
d * 64 * , 014028 ( 2001 ) [ hep - ph/0101146 ] .
f. brau and c. semay , phys .
d * 70 * , 014017 ( 2004 ) [ hep - ph/0412173 ] . v. mathieu , c. semay , and b. silvestre - brac , phys .
d * 74 * , 054002 ( 2006 ) [ hep - ph/0605205 ] . c. michael , hep - ph/0308293 , and references therein .
p. bicudo , m. cardoso , and o. oliveira , arxiv:0704.2156 .
d. horn and j. mandula , phys .
d * 17 * , 898 ( 1978 ) .
f. buisseret and c. semay , phys .
d * 74 * , 114018 ( 2006 ) [ hep - ph/0610132 ] . v. mathieu , c. semay , and f. brau , eur .
j. a * 27 * ( 2006 ) 225 [ hep - ph/0511210 ] .
a. p. szczepaniak and p. krupinski , phys .
d * 73 * , 034022 ( 2006 ) [ hep - ph/0511083 ] . f. j. llanes - estrada and s. r. cotanch , phys .
b * 504 * , 15 ( 2001 ) [ hep - ph/0008337 ] .
e. klempt , hep - ph/0404270 , and references therein . l. landau and e. lifchitz , _ thorie quantique relativiste _ ( ed .
mir , moscou , 1972 ) .
m. lscher and p. weisz , jhep * 0207 * , 049 ( 2002 ) [ hep - lat/0207003 ] .
a. pineda , j. phys .
g * 29 * , 371 ( 2003 ) .
f. buisseret and c. semay , eur .
j. a * 33 * , 87 ( 2007 ) [ hep - ph/0611216 ] .
l. l. foldy and s. a. wouthuysen , phys .
* 78 * , 29 ( 1950 ) .
p. j. silva and o. oliveira , nucl .
b * 690 * , 177 ( 2004 ) [ hep - lat/0403026 ] . c. semay , b. silvestre - brac , phys . rev .
d * 51 * , 1258 ( 1995 ) .
a. simonov , hep - ph/9911237 .
a. m. badalian and b. l. g. bakker , phys .
d * 66 * , 034025 ( 2002 ) [ hep - ph/0202246 ] .
d. horn and j. mandula , phys .
d * 17 * , 898 ( 1978 ) ; f. buisseret and c. semay , phys .
d * 74 * , 114018 ( 2006 ) [ hep - ph/0610132 ] . j .-
derendinger , _ thorie quantique des champs _ ( presses polytechniques et universitaires romandes , paris , 2001 ) .
d. a. varshalovich , a. n. moskalev , and v. k. khersonskii , _ quantum theory of angular momentum _ ( world scientific publishing , singapore , 1988 ) .
m. koma , y. koma and h. wittig , pos lat2005 , 216 ( 2006 ) [ hep - lat/0510059 ] , y. koma and m. koma , nucl .
b * 769 * , 79 ( 2007 ) [ hep - lat/0609078 ] .
b. gromes , nucl .
b * 131 * , 80 ( 1977 ) . w .-
et al . _ , j. phys .
g * 33 * , 1 ( 2006 ) .
g. s. bali , phys .
* 343 * , 1 ( 2001 ) [ hep - ph/0001312 ] .
a. c. aguilar , a. mihara , and a. a. natale , phys .
d * 65 * , 054011 ( 2002 ) [ hep - ph/0109223 ] . f. buisseret , phys . rev .
c * 76 * , 025206 ( 2007 ) [ arxiv:0705.0916 ] .
* 95 * , 052002 ( 2005 ) [ hep - lat/0503005 ] .
p. carruthers , _ introduction to unitary symmetry _
( interscience publishers , usa , 1966 ) . f. j. yndurin , _ the theory of quark and gluon interactions _ ( springer - verlag , berlin , 1999 ) . c. semay , f. buisseret , n. matagne , and fl .
stancu , phys .
d * 75 * , 096001 ( 2007 ) [ hep - ph/0702075 ] . | we systematically compute the effective short - range potentials arising from second order qcd - diagrams related to bound states of quarks , antiquarks , and gluons .
our formalism relies on the assumption that the exchanged gluons are massless , while the constituent gluons as well as the lightest quarks acquire a nonvanishing constituent mass because of confinement .
we recover the well - known fermi - breit interaction in the heavy quark limit .
such an effective potential has been proved in the past to be relevant for the building of accurate potential models describing usual hadrons ( mesons and baryons ) .
the general potentials we obtain in this work are also expected to be useful in the understanding of exotic hadrons like glueballs and @xmath0 hybrids . in particular , we compute for the first time an effective short - range quark - gluon potential , and show the existence of a quadrupolar interaction term in this case .
we also discuss the influence of a possible nonzero mass for the exchanged gluons . |
cataclysmic variables ( cvs ) are compact binaries containing an accreting white dwarf ( wd ) and a low - mass main - sequence ( ms ) donor star ( see * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* for reviews ) .
since the maximum size that the progenitor of the wd can reach during its evolution is much larger than the current binary separation , it is believed that the binary orbit must have significantly shrunk through so - called common envelope ( ce ) evolution , during which the primary star has lost most of its envelope and left a degenerate core , i.e. , the proto - wd @xcite . if surviving ce evolution , the post - common envelope binary ( pceb ) can evolve to be a cv by two ways .
( 1 ) if the secondary is originally of low - mass ( @xmath0 ) , angular momentum loss ( aml ) via magnetic braking ( mb ) and/or gravitational radiation ( gr ) may reduce the orbit so that the secondary overflows its roche - lobe ( rl ) and initiates mass transfer within a hubble time .
( 2 ) a more massive secondary star can fill its rl due to nuclear expansion . in that case the mass transfer first proceeds on a thermal timescale of the donor star , and the binary may appear as a super - soft x - ray source ( sss , * ? ? ? * for a review ) . after the mass ratio reverses , the binary evolution may follow the similar path as in the former case .
the wd mass problem concerns the discrepancy in the mass distribution for isolated wds and wds in cvs , which is in contradiction with theoretical expectations @xcite . while the average mass of isolated wds is @xmath1 ( e.g. , * ? ? ?
* ; * ? ? ?
* ) , the wds in cvs are remarkably more massive , with an average mass @xmath2 ( * ? ? ?
* ; * ? ? ?
* and references therein ) , and this discrepancy is unlikely to be due to the observational bias toward cvs with more massive wds , which tend to be brighter and hence easier to be discovered @xcite .
considering the fact that wds in pcebs have an average mass around @xmath3 , similar to that of isolated wds @xcite , there are three possible resolutions to the wd mass problem .
( 1 ) the wds have increased their masses during the cv evolution , but this idea is disfavored since numerical calculations ( e.g. , * ? ? ?
* ; * ? ? ?
* ) have shown that all ( or even more ) of the accreted matter is lost during nova outbursts .
( 2 ) a significant fraction of cvs originate from pcebs with initially more massive secondaries @xcite , and during the thermal - timescale mass transfer phase there is stable burning of hydrogen and helium on the surface of the wds , leading to mass growth . using a binary population synthesis ( bps ) method
, @xcite found that this model can produce a large number of massive wds , but still with too many helium wds and evolved donor stars , both inconsistent with observations .
( 3 ) mass transfer in pcebs with low - mass wds are dynamically unstable , provided that there is extra aml besides mb and gr .
such binaries would merge rather than evolve into cvs @xcite .
the aim of this paper is to examine the feasibility of the above models with detailed binary evolution calculations . for resolution
( 2 ) we note that @xcite employed the semi - analytic method of @xcite ( with modifications by * ? ? ?
* ) to deal with thermal - timescale mass transfer .
the actual evolution of the mass transfer rate should be more complicated than this simplified estimate ( e.g. , * ? ? ?
. moreover , a semi - analytic criterion for stable hydrogen burning @xcite was adopted , which possibly overestimates the net accumulation efficiency .
recently , @xcite presented the results of the parameter space for net mass accumulation through simulations of a long series of hydrogen and helium flashes on wds with mass @xmath4 .
we employ them to calculate how much mass a @xmath5 wd in binaries can grow . for resolution ( 3 ) , @xcite showed that assuming an empirical model for consequential angular momentum loss ( caml ) that increases with decreasing wd mass can greatly reduce the number of cvs with low - mass wds .
if this idea is correct , the effect of the caml should also be present in cvs .
@xcite recently also reconstructed the cv evolution by use of the mass - radius relation for the secondary stars , which can be used to constrain the caml model parameters .
the result is also helpful in understanding the physics behind the caml .
this paper is organized as follows .
we describe the input physics included in the binary evolution calculations in section 2 .
the calculated results on the wd mass growth are presented in section 3 . in section 4
, we examine the parameter space for dynamically unstable mass transfer under caml and its influence on the cv evolution .
our conclusions and discussion are in section 5 .
we use an updated stellar evolutionary code developed by @xcite to follow the evolution of pcebs , which consist of a co wd of mass @xmath6 and a ms donor star of mass @xmath7 .
we assume solar chemical composition ( @xmath8 , @xmath9 , and @xmath10 ) for the donor star .
the ratio of the mixing length to the pressure scale height and the convective overshooting parameter are set to be 2.0 and 0.12 , respectively .
the stellar opal opacities are taken from @xcite and @xcite .
the growth of an accreting wd depends on how much mass can be accumulated during the hydrogen and helium burning phases , and there have been extensive investigations on this subject @xcite .
the mass growth rate @xmath11 of a wd can be described as follows , @xmath12 where @xmath13 is the accumulation efficiency during hydrogen burning , @xmath14 the accumulation efficiency during helium burning , and @xmath15 the mass transfer rate .
@xcite simulated the long - term evolution of wds with a range of masses ( @xmath16 ) and accretion rates ( @xmath17yr@xmath18 ) of hydrogen - rich material to determine the efficiency of mass retention for each parameter combination .
we fit their numerically calculated results for @xmath13 .
if the mass accretion rate exceeds @xmath19yr@xmath18 , we fix the hydrogen accumulation efficiency to be that for @xmath19yr@xmath18 , as this rate is actually higher than the traditional critical accretion rate for stable hydrogen burning on a @xmath20 wd considered here ( e.g. , * ? ? ?
we adopt the prescriptions suggested by @xcite for @xmath14 , because the calculations of @xcite for helium flashes were only limited to @xmath21 and @xmath22 wds .
a comparison of the magnitudes of @xmath14 with the same @xmath6 and @xmath23 in @xcite and @xcite shows that the former is generally higher , so our following results on the wd growth might be regarded as upper limits . in our calculation
, the excess matter @xmath24 is assumed to be ejected from the wd in the form of isotropic wind , carrying its specific am @xcite .
aml due to gr @xcite and mb @xcite is also included .
in our calculations , we set the initial wd masses to be @xmath25 to examine its evolution with mass accretion and thermonuclear burning .
this mass range is roughly in accord with the average mass of wds in pcebs .
we do not consider lower - mass wds since reasonable estimate of the accumulation efficiencies are not available for them .
the initial masses of the secondary / donor stars are chosen to be in the range of @xmath26 .
figures 1 summarizes the outcome of the binary evolution in the initial orbital period - donor mass plane for a @xmath27 wd . the parameter space is divided into five regions : ( 1 ) the forbidden region where the secondary s radius exceeds its rl radius at the beginning , ( 2 ) the region where the mass transfer is subject to ( delayed ) dynamical instability , ( 3 ) the region with no rl - overflow within a hubble time , ( 4 ) the region with the wd mass growth @xmath28 , and ( 5 ) the region where the wd can effectively increase mass ( by more than @xmath29 ) .
the red and blue circles represent the final wd masses in the range of @xmath30 and @xmath31 , respectively .
the red and blue solid triangle denote the binaries with same mass growth range respectively , but unable to evolve into cvs .
figure 2 is similar to fig . 1 but for @xmath32 wds .
the red , blue , and pink circles represent the final wd masses in the range of @xmath33 , @xmath34 , and @xmath35 , respectively .
the triangles have the same meanings as in fig . 1 .
details of the evolutionary sequences for different sets of initial parameters are demonstrated in figs .
3 - 7 . in each figure
, the left and right panels exhibit the evolution of the mass transfer rate ( solid line ) and the white dwarf mass ( dashed line ) , and the evolution of the donor mass ( solid line ) and the orbital period ( dashed line ) , respectively . in fig . 3 ,
the binary initially consists of a @xmath36 wd and a @xmath37 secondary in a 1.26 day orbit .
mass transfer is driven by mb at a rate @xmath38 yr@xmath18 , and the wd has no chance to grow mass , so only part of the evolutionary tracks are plotted here . in figs . 4 - 6 the donor mass
is increased to be @xmath39 , @xmath40 , and @xmath41 , respectively .
they represent the cases of the wd growth with increasing amount .
their distinct feature is the early sss stage when the donor star is more massive than the wd .
larger mass ratio @xmath42 leads to higher mass retention efficiency , and the final wd masses in these binaries are @xmath43 , @xmath44 , and @xmath45 , respectively . in the subsequent phases the mass transfer rates decreases continuously and
the wd masses do not change . in fig .
7 the donor mass is taken to be @xmath46 .
the mass transfer firstly proceeds on a thermal timescale , then rises rapidly due to delayed dynamical instability . to roughly estimate the fraction of pcebs with efficient wd mass growth by thermal - timescale mass transfer , we incorporate our calculated results with the bps study of pcebs and cvs by @xcite . according to figs . 2 and 3 in that paper
, we find that with the ce efficiency parameter @xmath47=0.2 and 0.6 , about @xmath48 of the binaries with an initial @xmath27 wd can evolve into cvs with the final wd mass @xmath49 . for an initial @xmath32 wd , around @xmath50 of the binaries can evolve into cvs with the final wd mass @xmath51 .
thus thermal - timescale mass transfer seems unable to significantly change the distribution of the wd mass .
we then move to resolution ( 3 ) . in the standard model of the cv evolution ,
aml is caused dominantly by mb above the period gap and solely by gr below the period gap @xcite .
extra aml might amplify the mass transfer rate or even cause the mass transfer to be dynamically unstable , since the rl radius of the donor star shrinks more rapidly than the stellar radius @xcite .
@xcite proposed that , for a specific form of caml that increases with decreasing wd mass , the parameter space for ( delayed ) unstable mass transfer can be significantly enlarged , especially for pcebs with low - mass wds .
thus these binaries are likely to merge after experiencing a rapid mass transfer phase .
this increases the average mass of the wds in binaries that successfully evolve to be long - lived cvs . by following the evolutionary paths of cvs
, @xcite found the the best fit of the observational data gives the aml rates to be @xmath52 times the gr rate below the period gap and @xmath53 times the mb rate above the period gap , indicating that extra aml mechanism(s ) works at least below the period gap if we believe that the mb effect has switched off .
this kind of aml is most likely to be related to mass loss during the cv evolution . assuming that all the transferred mass is lost from the binary during nova outbursts taking the specific am of the wd
, we can write the mass transfer rate as follows @xcite @xmath54 where @xmath55 represents the change in the donor star s radius due to thermal or nuclear evolution , @xmath56 is the binary orbital am , @xmath57 the systemic aml rate caused by mb and/or gr , and @xmath58 the adiabatic mass - radius exponent of the donor star ( i.e. , @xmath59 ) . for low - mass ms stars
, we can neglect the change in the stellar radius due to evolution , i.e. , @xmath60 . for cvs below the period gap ,
let @xmath61 ( where @xmath62 is the rate of aml by gr ) we can rewrite eq .
( 2 ) as @xmath63 however , since the normalized gr - induced aml rate should not be larger than unity , this phenomenological form simply reflects some more complicated processes that drive the mass transfer .
@xcite examined several potential caml mechanisms for mass loss including isotropic wind from the wd , outflow from ( inner / outer ) lagrangian points , and formation of a circumbinary ( cb ) disk .
they showed that formation of a cb disk or outflow from the outer lagrangian point seems to be able to account for the extra 1.47gr aml rate .
note that @xcite considered each caml mechanism separately , while in real situation there may be different ways of mass loss simultaneously ( e.g. , * ? ? ?
* ) . in the following we assume that mass loss occurs in two ways :
( 1 ) a fraction @xmath64 of the lost mass leaves the binary in the form of isotropic wind from the wd , and ( 2 ) the other @xmath65 part forms a cb disk surrounding the system .
the total rate of the caml is then given by @xmath66\frac{\dot{m}_2}{m_2},\ ] ] where @xmath67 = 1.5 is the ratio of the radius of the cb disk to the binary separation @xcite .
the first and second terms on the right hand side of eq .
( 4 ) represent the contribution from isotropic wind and the cb disk , respectively .
similar to eq .
( 2 ) , we can write the mass transfer rate for cvs below the period gap as @xmath68}.\ ] ] since eq . ( 3 ) and
( 5 ) should match each other , we get @xmath69\right\ } = \frac{5}{6}+\frac{\zeta}{2}-\frac{q}{3(1+q)}-\frac{q^2}{1+q}.\ ] ] after transformation , we obtain @xmath70/2.47}{\gamma(1+q)^2-q^2},\ ] ] where @xmath71 .
figure 8 shows @xmath72 , the fraction of the lost matter that forms a cb disk , as a function the mass ratio @xmath73 for @xmath6 = 0.6 @xmath74 and @xmath75 .
it can be seen that @xmath72 decreases with @xmath73 , ranging from about 0.1 to 0.4 when @xmath76 .
we then calculate the mass transfer sequences in a number of cvs adopting typical values of @xmath64 .
we regard the mass transfer to be dynamically unstable if the mass transfer rate rapidly rises to @xmath77yr@xmath18 . in fig . 9
we show the boundary between stable and unstable mass transfer in the @xmath78 plane .
the thin solid and dotted lines correspond to @xmath79 and 0.3 , respectively . also shown with the thick solid line
is the stability limit in @xcite .
compared with @xcite , it can be seen that for low - mass donors , the critical mass ratio for dynamically unstable mass transfer becomes significantly smaller when mass loss through a cb disk is considered .
this can be understood as follows .
it is evident that if the denominator of the right hand of eq .
( 5 ) is @xmath80 then the mass transfer is dynamically unstable @xcite .
this means that the mass transfer is stable only when @xmath81 or @xmath82 when @xmath83 , 0.5 , and 1 , we obtain @xmath84 , 0.41 , and 0.19 , respectively .
thus pcebs possessing low - mass wds , which tend to have higher initial @xmath73 @xcite , are more likely to undergo unstable mass transfer for a given @xmath72 .
figure 10 shows an example of the evolutionary sequence with runaway mass transfer for a binary with @xmath85 , @xmath86 , @xmath87 day , and @xmath79 .
in this work we examine the proposed explanations for the wd problem in cvs by calculating the pre - cv evolution with updated hydrogen accumulation efficiency and aml prescription .
simulations by @xcite on hydrogen flashes disfavor the possibility of mass growth during the cv evolution , since the mass transfer rates are too low to allow stable burning on wds . however , we caution that there exist a group of ssss ( e.g. , rx j0439.8@xmath886809 , 1e0035.4@xmath887230 , rx j0537.7@xmath887034 ) whose properties are similar to those of ordinary cvs , suggesting that the mass transfer rates in these systems are ( at least temporarily ) high enough for stable hydrogen burning ( * ? ? ?
* and references therein ) .
a possible mechanism is there is excited wind from the donor star driven by x - ray irradiation from the accreting wd @xcite .
however , this self - excited wind requires some special conditions and should not be popular in cvs @xcite .
our calculations demonstrate that in a minor fraction of pcebs the wds can effectively grow in mass through thermal - timescale mass transfer .
this means that the formation channel from binaries with initially more massive secondaries may be less efficient than in @xcite .
the main reason for the discrepancy is that the hydrogen accumulation efficiency of @xcite is considerably lower than that used by @xcite .
the best - fit aml prescriptions for present cvs by @xcite suggest that there should be extra aml mechanism , which is most likely related to mass loss during the binary evolution .
@xcite proposed a parameterized caml with the specific aml inversely proportional to the wd mass , and showed that the criterion for stable mass transfer can be significantly altered compared with in the traditional model .
the consequence is that a large fraction of the binaries with low - mass wds would undergo unstable mass transfer and probably merge with their companion stars .
the physical mechanism to account for this extra aml is still open , and there are several possibilities proposed in the literature . @xcite considered the effect of frictional aml in recurrent nova outbursts , i.e. interaction of the expanding nova envelope with the donor , on the secular evolution of cvs .
they showed that the strength of frictional aml is sensitive to the expansion velocity @xmath89 of the envelope at the location of the donor , being stronger for smaller @xmath89 .
indeed lower expansion velocities are expected for low - mass wd in cvs during nova outbursts than high - mass wds @xcite . alternatively , part of the slowly expanding ejected matter may form a cb disk or ce @xcite , and cause additional aml .
matter can also be transported into the circumbinary space via a wind / outflow from the accretion disk or the secondary star as a natural consequence of the mass transfer .
numerical simulations of the mass transfer process in cvs have demonstrated that as much as @xmath90 of the transferred matter can escape from the wd s roche lobe and end up in the circumbinary space around the cv @xcite .
infrared observations of cvs by the _ spitzer space telescope _ have revealed the presence of dust in many systems , indicating the possible origin of a cb disk @xcite . in this work
we take into account the effect of mass loss both from the wd and through a cb disk .
we show that if around @xmath91 of the matter ejected during nova eruptions forms a cb disk , mass transfer in low - mass wds are likely to be dynamically unstable .
our results are largely consistent with @xcite , in which a ce phase is involved . while the frictional aml and ce models invoke discontinuous mass loss during nova eruptions , mass loss in the cb disk model can be both continuous and discontinuous . in the former case mass loss
can be from the donor star rather the wd @xcite , and one would not expect substantial change in the orbital period after a nova outburst as in the frictional aml and ce models .
finally we note that cb disks may also help explain the large spread for mass transfer rates in cvs for a given orbital period @xcite .
we thank the referee for her / his valuable comments that helped to improve this paper .
we are also grateful to yong shao for useful discussion .
this work was supported by the natural science foundation of china under grant numbers 11133001 , 11333004 and u1331117 , the national key research and development program of china ( 2016yfa0400803 ) , and the strategic priority research program of cas under grant no .
xdb09000000 .
ablimiti , i. , xu , x .-
, & li , x .- d .
2014 , , 780 , 80 alexander , d. r. , ferguson , j. w. 1994 , , 437 , 879 bergeron , p. , saffer , r. a. , & liebert , j. 1992 , apj , 394 , 228 bisikalo , d. v. , boyarchuk , a. a. , kaigorodov , p. v. , kuznetsov , o. a. 2003 , ast .
, 47 , 809 bisikalo , d. v. , & kononov , d. a. 2010 , mmsai , 81 , 187 claeys , j. s. w. , pols , o. r. , izzard , r. g. , vink , j. , & verbunt , f. w. m. 2014 , , 563 , a83 de kool , m. 1992 , a&a , 261 , 188 dubus , g. , taam , r. , hull , c. , qatson , d. m. , & mauerhan , j. c. 2007 , , 663 , 516 eggleton , p. p. 1971
, , 151 , 351 eggleton , p.
, , 156 , 361 eggleton , p.
, , 163 , 279 hachisu , i. , kato , m. , & nomoto , k. 1996 , , 470 , l97 hachisu , i. , kato , m. , nomoto , k. , & umeda , h. 1999 , , 519 , 314 han , z. , & podsiadlowski , ph .
2004 , , 350 , 1301 hillman , y. , prialnik , d. , kovetz , a. , & shara , m. m. 2015 , mnras , 446 , 1924 hillman , y. , prinlnik , d. , kovetz , a. , & shara , m. m. 2016 , , 819 , 168 hoard , d. w. , howell , s. b. , brinkworth , c. s. , ciardi , d. r. , & wachter , s. 2007 , , 671 , 734 hoard , d. w. , kafka , s. , wachter , s. , howell , s. b. , & brinkworth , c. s. et al .
2009 , , 693 , 236 howell , s. b. , brinkworth , c. , hoard , d. w. , wachter , s. , & harrison , t. et al . 2006 , , 646 , l65 hurley , j. r. , tout , c. a. , & pols , o. r. 2002 , , 329 , 897 idan , i. , shaviv , n. j. , & shaviv , g. 2013 , mnras , 433 , 2884 kahabka , p. , & van den heuvel .
e. p. j. 1997 , , 35 , 69 kato , m. , & hachisu , i. 2004 , , 613 , l129 kepler , s. o. , kleinman , s. j. , nitta , a. , et al .
2007 , mnras , 375 , 1315 king , a. r. , & van teeseling , a. 1998 , , 338 , 965 knigge , c. , baraffe , i. , & patterson , j. 2011 , apjs , 194 , 28 kovetz , a. , & prialnik , d. 1994 , apj , 424 , 319 landau , l. d. , & lifshitz , e. m. 1975 , in course of theoretical physics , pergamon international library of science , technology , engineering and social studies ( 4th rev .
engl . ed .
; oxford : pergamon ) livio , m. , govarie , a. , & ritter , h. 1991 , a&a , 246 , 84 meng , x. , chen , x. , & han , z. 2009 , , 395 , 2103 nelemans , g. , siess , l. , repetto , s. , toonen , s. , & phinney , e. s. 2016 , apj , 817 , 69 nomoto , k. , saio , h. , kato , m. , & hachisu , i. 2007 , , 663 , 1269 paczyski , b. 1976 , in structure and evolution of close binary systems , ed .
p. eggleton , s. mitton , & j. whelan ( dordrecht : kluwer ) , iau symp .
, 73 , 75 piersanti , l. , tornamb , a. , & yungelson , l. r. 2014 , mnras , 445 , 3239 politano , m. 1996 , apj , 465 , 338 politano , m. , & weiler , k. p. 2007 ,
apj , 665 , 663 prialnik , d. , & kovetz , a. 1995 , , 445 , 789 rappaport , s. , verbunt , f. , & joss , p. c. 1983 , apj , 275 , 713 ritter , h. 2010 , mmsai , 81 , 849 rogers , f. j. , iglesias , c. a. 1992 , apjs , 79 , 507 schenker , k. , king , a. r. , kolb , u. , wynn , g. a. , & zhang , z. 2002 , , 337 , 1105 schenker k. , kolb u. , & ritter h. 1998 , mnras , 297 , 633 schreiber , m. r. , zorotovic , m. , & wijnen , t. p. g. 2016 , , 455 , l16 shao , y .
, & li , x. 2012 , apj , 745 , 165 soberman , g. e. , phinney , e. s. , & van den heuvel , e. p. j. 1997 , , 327 , 620 spruit , h. c. , & ritter , h. 1983 , a&a , 124 , 267 spruit , h. c. , & taam , r. 2001 , , 548 , 900 starrfield , s. , timmes , f. x. , iliadis , c. , et al . 2012 , balta , 21 , 76 van den heuvel , e. p. j. 1994 , in interacting binaries ( saas - fee 22 ) , ed . s. n. shore et al .
( berlin : springer ) , 263 van teeseling , a. , & king , a. r. 1998 , , 338 , 957 verbunt , f. , & zwaan , c. 1981 , , 100 , l7 warner , b. 1995 , cataclysmic variable stars ( cambridge : cambridge university press ) whelan , j. , & iben , i. 1973 , apj , 186 , 1007 wijnen , t. p. g. , zorotovic , m. , & schreiber , m. r. 2015 , , 577 , a143 williams , r. 2013 , , 146 , 55 xu , x .- j . , &
li , x .- d .
2007 , , 476 , 1283 yaron , o. , prialnik , d. , shara , m. m. , & kovetz , a. 2005 , apj , 623 , 398 zorotovic , m. , schreiber , m. r. , & gnsicke , b. t. 2011 , , 536 , a42 = 0.6 @xmath74 with different outcome .
the red and blue circles represent the final wd mass of @xmath92 , and @xmath93 , respectively .
the red and blue triangles denote the same final wd mass range respectively , but the binaries can not evolve into cvs . ]
= 0.6 @xmath74 , @xmath97 = 1.26 days and @xmath98 = 0.7 @xmath74 .
the solid and dashed curves represent the evolution of mass transfer rate and the wd mass ( left panel ) , the donor star mass and the orbital period ( right panel ) , respectively , title="fig : " ] = 0.6 @xmath74 , @xmath97 = 1.26 days and @xmath98 = 0.7 @xmath74 .
the solid and dashed curves represent the evolution of mass transfer rate and the wd mass ( left panel ) , the donor star mass and the orbital period ( right panel ) , respectively , title="fig : " ] | recent observations show that the white dwarfs ( wds ) in cataclysmic variables ( cvs ) have an average mass significantly higher than isolated wds and wds in post - common envelope binaries ( pcebs ) , which are thought to the progenitors of cvs .
this suggests that either the wds have grown in mass during the pceb / cv evolution or the binaries with low - mass wds are unable to evolve to be cvs . in this paper
, we calculate the evolution of accreting wd binaries with updated hydrogen accumulation efficiency and angular momentum loss prescriptions .
we show that thermal timescale mass transfer is not effective in changing the average wd mass distribution .
the wd mass discrepancy is most likely related to unstable mass transfer in wd binaries in which an efficient mechanism of angular momentum loss is required . |
the asymmetry of the stokes parameters with wavelength is a powerful diagnostic tool to probe the relation between flows and the magnetic field in the atmosphere of sunspots .
line asymmetries convey information about lateral ( within one resolution element ) and the line - of - sight ( los ) gradients and discontinuities of physical parameters .
since the first systematic characterization and interpretation of these asymmetries and their spatial distribution throughout a sunspot by ( * ? ? ?
* and references therein ) the complementary improvements achieved in both observational techniques and theoretical modeling , have advanced our understanding of particularly the penumbral inhomogeneities .
it is general consent that a combination of non - trivial magnetic field and flow topologies along the los are the cause of the observed penumbral stokes asymmetries .
a simple and instructive way to characterize the asymmetry of stokes v profiles is through the net - circular polarization ( ncp ) , defined as @xmath0 , where the integral is performed over one spectral line only .
the sum of the ncp of individual lines than leads to the well - known broad - band circular polarization ( bbcp ) first noticed by @xcite .
the appearance of the ncp is very distinct in the penumbra and shows a peculiar difference when between observations in the visible or in the near infrared wavelength region : in the visible at 630.2 nm the ncp maps tend to be symmetric about the disk center and sun center connection line @xcite while infrared ncp maps at 1564.8 nm reveal an antisymmetric distribution @xcite . @xcite and @xcite
demonstrate that the uncombed penumbral atmosphere @xcite mimicked by the _ moving tube model _
@xcite reproduces the azimuthal behavior of the ncp and , furthermore , convincingly show that the incisive discrepancy between the visible and the near infrared can be understood when the effects of anomalous dispersion @xcite are included .
so far penumbral ncp maps of the 630.2 nm line have been only presented before by @xcite ( in form of a contour line map ) , @xcite , @xcite and @xcite .
these maps visualize very well the symmetric behaviour of the ncp but lack the spatial resolution to reveal the presence of any potential fine structure .
in this letter we focus on high - spatial resolution spectropolarimetric measurements of a sunspot to investigate inhomogeneities in the penumbral magnetic field and flow field . instead of utilizing inversion techniques we calculate the ncp and line - of - sight ( los ) velocities and
elaborate on the properties of the same . in order to interpret the observational results we use the 3d geometric flux tube model vtube @xcite and perform subsequent radiative transfer calculations from which ncp and doppler maps can be synthesized .
the spectropolarimetric observations were obtained on 2005 june 30 , with the diffraction limited spectro - polarimeter ( dlsp , * ? ? ? * ) located at the nso dunn solar telescope ( dst ) , sacramento peak , new mexico .
the dlsp was operated in its high - resolution mode ( 00893pixel@xmath1 along the slit ) and in conjunction with the high - order adaptive optics system @xcite .
the slit width of 12microns is matched to the detector image scale along the spectrograph slit .
in spectral direction the instrument samples the solar spectrum from 630.0 nm to 630.4 nm with 2.1pmpixel@xmath1 .
a two - dimensional map was generated by moving the slit across the leading sunspot ( positive polarity ) of ar noaa 10781 located at @xmath2 ( corresponding to a viewing angle of 42@xmath3 ) with a step size of 0089 covering a total field - of - view of 589 ( 660 steps ) .
the slit was oriented along solar north - south and the scan direction of the dlsp was solar west - east . at each slit position
16 images of 30ms exposure time each were accumulated for each of the four individual modulation states leading to a total integration time of 2sec .
a full map is accomplished within 56min .
the measurements are compensated for polarization effects caused by the telescope and the polarimeter following the calibration procedure developed by @xcite .
the very good seeing in conjunction with the high - order ao system led to exceptional stable observing conditions while scanning the solar image .
we estimate our effective spatial resolution in the dlsp map to be about 04 .
we calculate maps of the net circular polarization by integrating the stokes @xmath4 signal over the 630.25 nm line .
because of its relatively minor contamination by one of the two telluric lines the 630.15 nm line is used to derive line - of - sight ( los ) doppler velocities .
we apply a fourier phase method @xcite to the @xmath5 and @xmath6 profile of the 630.15 nm line to calculate line shifts .
an absolute velocity calibration is performed by taking advantage of the telluric o@xmath7 lines ( see e.g. * ? ? ?
* ) present in the recorded spectrum .
in contrast to the astrophysical convention we assign a negative sign to redshifts and a positive sign to blueshifts : dark areas in the dopplergram move away from the observer , while bright areas move towards the observer .
well - reported properties of the penumbral ncp ( see * ? ? ?
* ; * ? ? ?
* ) are the sign change along azimuthal cuts centered around the spot center and an increased magnitude on the limb side . based on spectropolarimetric data with much higher spatial resolution than utilized before ( figure [ fig : physmaps ] )
we additionally find ( a ) evidence for a fine - structure in the ncp and ( b ) a sign - reversal of the ncp as a function of radial distance from the spot center on the center - side penumbra which has not been recognized before .
confirming previous results we find that the ncp is symmetric in the inner and middle penumbra w.r.t .
the line that connects the disk center with the spot center .
the same property applies to the los velocity map which indicates an intricate coupling between penumbral flows and the sign of the ncp .
this is corroborated by the fine structure we detect in the ncp .
the spatial distribution of the ncp reveals filamentary organized inhomogeneities throughout the penumbra with very localized enhancements of elongated shape on a spatial scale of 03 .
although we detect these enhancements with no preferred location inside the penumbra those with the largest value of the ncp appear on the limb - side penumbra in proximity to where the magnetogram ( not shown ) indicates an apparent neutral line .
they are related to anomalous ( multi - lobe ) stokes v profiles ( see also * ? ? ?
* ) and thus do not necessarily show up in the flow signal ( figure [ fig : physmaps ] , _ right _ ) .
anomalously shaped stokes v profiles indicate either spatially unresolved contributions to the resolution element or los effects of at least two differently inclined magnetic components with opposite polarity that are shifted w.r.t .
each other and viewed upon under an oblique angle .
the ncp on the center - side penumbra is correlated with normal 2-lobe antisymmetric stokes v profiles . on smaller spatial scales related to the flow channels ( figure [ fig : physmaps ] ,
_ right _ ) that carry most of the evershed flow we are unable to state conclusively that there is a good overall correlation between the ncp and the flow signatures .
figure [ fig : azimuthal ] exemplifies this statement along three azimuthal cuts through the penumbra representative for three different radial distances from spot center ( as indicated by the ellipses in figure [ fig : physmaps ] , _ left _ ) .
furthermore , we detect for the first time a difference in the symmetry properties of the ncp observed between the outer limb- and center - side penumbra , namely a sign reversal of the ncp in the outer center - side penumbra . we verify the visual impression of the ncp zero - crossing on the center - side penumbra in figure [ fig : radial ] by means of the radial dependence of the ncp when averaged over the limb - side ( 90@xmath3 - 270@xmath3 ) and center - side penumbra ( 270@xmath3 - 90@xmath3 ) separately . the limb - side segment without penumbra
does not contribute to the averaging .
a comparison of the two curves clearly demonstrates the differences between the ncp on the limb- and center - side penumbra in radial direction . on the limb side
the ncp has one sign only throughout the penumbra while on the center side the ncp reverses its sign at a radial distance of @xmath80.75@xmath9 , ( where @xmath9 denotes the sunspots radius ) peaks at 0.95@xmath9 and then falls off outside the visible penumbra .
it should be noted that the course of the radial variation of the ncp and the exact location of the zero - crossing point within the penumbra depends also on the calibration for instrumental polarization , i.e. predominantly the crosstalk from stokes @xmath10 to @xmath4 .
we estimate that our calibration for this particular crosstalk is better than 0.1% .
the ncp averaged over areas where we see no significant stokes @xmath4 signal is in the range of 1.5@xmath1110@xmath12pm .
significant changes of the radial variation of the ncp , however , take place only if the residual crosstalk from @xmath10 to @xmath4 is in the range of 1% and larger .
in order to understand the observed sign reversal of the ncp , we have carried out radiative transfer calculations from which synthetic maps of observable quantities are obtained . at first sight , one may conjecture that the observed radial sign reversal of the ncp is caused by downflows in the outer penumbra . while this is indeed a possible solution that is compatible with the observed ncp maps , the doppler maps constrain the details of the model atmosphere even further .
it turns out that a model atmosphere with strongly downwards inclined flux tubes in the outer penumbra is not able to explain simultaneously both the observed sign reversal of the ncp and the constant sign of the doppler maps on the spot s center side .
however , we find that both features can be reproduced by magnetic flux tubes that have a magnetic field strength that is stronger than the background magnetic field in the outer penumbra while it is weaker in the inner part as visualized in figure [ fig : synthetic ] ( _ top left _ ) .
the assumption is not unrealistic as two - component inversions of sunspots indicate a higher field strength in the tube component in the extreme outer penumbra ( e.g. * ? ? ?
figure [ fig : synthetic ] shows the synthetic ncp ( _ middle _ ) and doppler ( _ right _ ) maps for a penumbral atmosphere , calculated with the vtube model of @xcite .
we adopt a very generic atmospheric configuration with horizontal magnetic flux tubes of 50 km radius with a flow velocity of 10km / s , embedded in a radially symmetric background field as in @xcite .
we assume a magnetic field strength inside the flux tubes that increases linearly from the inner penumbra ( where it is 500 g weaker than the background field ) to the outer penumbra ( where it is 500 g stronger than the background field ) .
it turns out that this configuration is quite robust , and the details of the magnetic field strength affect primarily the radial location of the zero - crossing of the ncp .
the axes of the flux tubes are placed at 30 km above @xmath13 . raising the flux tubes
mostly increases the absolute value of the ncp and would require a lower filling factor to explain the observed ncp values .
the effect of a jump in the magnetic field strength along the los on the ncp has been studied for a simpler atmospheric configuration by @xcite who derived a sign rule for the ncp due to this so - called @xmath14 effect .
this expression is proportional to the difference in the magnetic field strengths between the static background and the part of the atmosphere containing the flow . as a consequence the ncp changes sign when @xmath14 changes sign .
when comparing this rule with both observational data and more realistic models , one should note that the @xmath14 effect is but one of several ncp generating effects at work .
therefore , the locations of the sign reversals of @xmath14 and the ncp do not necessarily coincide .
in particular the sum of the @xmath14 effect and changes in the inclination ( @xmath15 effect ) results in the different behaviour of the ncp on the limb and center side : on the limbside the ncp decreases with radial distance because of the increasing @xmath14 effect but does not change the sign because the @xmath15 effect contributes positive to the ncp .
one should also bear in mind the assumptions on which the findings of @xcite are based .
a discussion of the results of @xcite in the context of flux tubes in the penumbra is given by @xcite and @xcite .
our model solution is in qualitative agreement with both the observed ncp and the observed los velocity .
it should be pointed out though that while the simultaneous agreement between the ncp and doppler maps places a strong constraint on the possible atmospheric models our solution may not necessarily be unique .
in this paper , we have presented novel evidence for a spatial fine structure in the visible ncp of the 630.25 nm line .
we find for the first time that the ncp is structured in radial filaments very much alike maps of other line parameters , e.g. the intensity , equivalent width and line width ( see e.g. * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
the spatial scales of the ncp filaments are similar to those found in the observed los velocity map although we can not find a good overall correlation between the signals of the ncp and the los velocity .
we conjecture that higher spatial resolution is needed to settle this issue .
@xcite find a correlation between the azimuthal variation of the ncp and the magnetic field inclination on angular scales of @xmath16 .
although we can not provide information about the inclination at this point we find that our observations show azimuthal variations of the ncp that are only of the order of @xmath17 .
furthermore , we have observed a zero - crossing of the ncp on the outer center - side penumbra and have demonstrated that this signature can be explained by an increased magnetic field strength inside magnetic flux tubes in the outer penumbra , while the magnetic field strength is decreased in the background field ( @xmath14-effect ) . to this end , we have compared high - resolution observations with maps of the ncp and the los velocity that have been synthesized from a 3d geometric sunspot model that has been coupled with a 1d radiative transfer code ( see * ? ? ? * for details ) .
we point out that the space of possible atmospheric models that can reproduce the observations is significantly constrained by taking into account the information contained in doppler maps in addition to the ncp maps .
this paper extends the work by @xcite , @xcite , @xcite , and @xcite and clearly demonstrates the need to model the `` uncombed '' penumbra of sunspots in greater detail .
this would include also taking into account the effect of field line curvature around flux tubes as considered by @xcite . | we present novel evidence for a fine structure observed in the net - circular polarization ( ncp ) of a sunspot penumbra based on spectropolarimetric measurements utilizing the zeeman sensitive 630.2 nm line .
for the first time we detect a filamentary organized fine structure of the ncp on spatial scales that are similar to the inhomogeneities found in the penumbral flow field .
we also observe an additional property of the visible ncp , a zero - crossing of the ncp in the outer parts of the center - side penumbra , which has not been recognized before . in order to interprete the observations we solve the radiative transfer equations for polarized light in a model penumbra with embedded magnetic flux tubes .
we demonstrate that the observed zero - crossing of the ncp can be explained by an increased magnetic field strength inside magnetic flux tubes in the outer penumbra combined with a decreased magnetic field strength in the background field .
our results strongly support the concept of the uncombed penumbra . |
higgs bosons in representations of su(2)@xmath0 other than doublets occur in many extensions of the standard model . among these ,
situations of special interest are created by su(2)@xmath0 triplet scalars which occur in various scenarios , ranging from left - right symmetric models to little higgs theories @xcite .
complex scalar triplets with a hypercharge @xmath8 are particularly rich in this context in terms of their phenomenological implications .
once we allow lepton number violation by two units , the complex triplet can couple to left - handed leptons in a gauge - invariant and renormalizable manner and give rise to majorana masses for neutrinos .
no right - handed neutrinos need to be postulated in such a scenario . at the same time
, the complex triplet contains a doubly charged component ( @xmath9 ) , and the same @xmath10 interactions with charged leptons open up a very spectacular set of decay channels for this state , namely , resonant decays into a pair of like - sign leptons .
these channels not only lead to remarkably background - free signatures of the doubly charged scalars , but also demonstrate a crucial link between observations at high energy colliders and the widely discussed mechanism of neutrino mass generation .
we explore such a link in this paper , in the context of the large hadron collider ( lhc ) .
although the general idea is simple here and has been discussed earlier in different connections , including direct experimental searches at the fermilab tevatron @xcite , a number of subtle and challenging issues invariably come up in such a study , and we have done our best to address them . first of all , the presence of a @xmath11 triplet @xmath12 allows the following @xmath10 interaction with left - handed lepton doublets @xcite : @xmath13 where @xmath14 is the charge conjugation operator . in general
, the triplet can develop a vacuum expectation value ( vev , @xmath15 ) .
it thus contributes to @xmath16 majorana masses for the neutrinos , proportional to @xmath17 .
the current observations from the neutrino oscillation experiments and cosmological bounds yield @xcite @xmath18 this implies a stringent bound on @xmath15 .
for instance , if we take @xmath19 to be as small as the electron s yukawa coupling , we have @xmath20 mev . as for any @xmath16 interactions , neutrinoless double beta decay experiments ( @xmath21 )
@xcite provide a direct test .
@xmath9 also contributes to @xmath21 via @xmath22 the @xmath23 coupling is proportional to @xmath24 in the model under discussion , and with the @xmath21 bound @xcite , we have @xmath25 given the neutrino mass requirement in eq .
( [ yvev ] ) , we obtain a bound @xmath26 , which is too weak to be relevant here .
in fact , there is another direct bound on @xmath15 due to the electroweak @xmath27-parameter . in order to prevent large tree - level contributions to the @xmath27-parameter
, one needs @xcite @xmath28 although models exist in the literature @xcite , where one has complex as well as real triplets , whose combined contributions to the @xmath27-parameter cancel , thus allowing large triplet vevs at the tree level , they presuppose the existence of additional symmetries whose validity is not clear once higher order effects involving gauge couplings are included .
it is also to be noted that , if @xmath15 is allowed to be close to the upper limit from the @xmath27-parameter , then the couplings @xmath29 are forced to be @xmath30 .
since there are six of these couplings ( assuming the matrix to be real symmetric ) , they all need to be adjusted within such a small range in order to reproduce the neutrino mass matrix that fits the observed mixing pattern .
this makes the scenario rather more fine - tuned than one in which one has a much smaller @xmath15 a single quantity ) , along with the six parameters @xmath29 which are allowed to be closer to unity . in this phenomenological study
, we will assume a small @xmath15 and be guided for the parameters by the neutrino mass generation , to saturate the relation of eq .
( [ yvev ] ) .
secondly , we wish to identify the leading production channels for the doubly charged higgs at the lhc .
rather encouraging predictions about single production of the doubly charged higgs boson in @xmath31-boson fusion can be found in some earlier studies via the process @xmath32 @xcite .
however , specific features of models are often utilized in such studies , and it is difficult to maintain such optimism in a general case if the triplet vev is restricted to a smaller value . numerically , the signal rate is proportional to @xmath33 and is already too small to be observable at the lhc if @xmath34 gev .
since we are considering even smaller values of @xmath15 , we therefore concentrate instead on pair production which is largely governed by electromagnetic interactions . in principle , the production of one doubly - charged higgs in conjunction with a singly charged one , followed by the decay @xmath35 , driven by the @xmath36 gauge coupling , can generate additional signals @xcite . however , this presupposes considerable mass separation between the @xmath9 and the @xmath37 , which may not be quite expected in many scenarios such as those based on little higgs theories .
therefore , we concentrate on the pair production process as the constantly available fall - back , and remember that the doubly charged scalars produced most copiously in the drell - yan ( dy ) channel . in our analysis , we also include the production via the two - photon fusion channel .
this is motivated by the stronger electromagnetic coupling of a doubly charged particle .
comparing with a singly charged scalar , the two - photon channel will have an enhancement factor of 16 .
numerically , it provides about 10 per cent correction to the dy process , and it is comparable to the qcd corrections @xcite .
thirdly , as for the identification of the @xmath38 signal , one must consider both @xmath39 and @xmath40 , keeping in mind the interplay of the two independent parameters @xmath19 and @xmath15 that govern the two decay channels . in spite of the simple and distinctive nature of the first signal
, one can not rule out their faking by a number of sources .
we have studied in detail the event selection criteria which establish the _ bona fide _ of such signals in a convincing way .
side by side , we are also suggesting ways of isolating the signals coming from @xmath40 , which are in general more background - prone , and can dominate the decays of the doubly charged scalar in certain regions of the parameter space .
most studies on the signals of doubly charged scalars have not adequately addressed the challenges posed by backgrounds in this search channel .
furthermore , given the small values of @xmath41 , it is useful to investigate whether the doubly charged higgs can indeed be long - lived on the scale of collider detectors in any region of the parameter space of this scenario , a situation experimentalists have already looked for at the tevatron @xcite . and
finally , we wish to point out , albeit in a qualitative way , that relative strengths of the different flavor diagonal as well as off - diagonal decays @xmath42 of the doubly charged state should give us information about the matrix @xmath19 and , in turn , generate insight into the structure of the neutrino mass matrix @xcite , which is responsible for the tri - bimaximal mixing pattern suggested by observations .
we present the calculation related to both the drell - yan and two - photon production channels at the lhc in sec .
section iii contains a discussion on the decay of the doubly charged states into the @xmath43 and @xmath44 final states across the parameter space allowed by the neutrino mass constraint .
the signal observability at the lhc for both channels are discussed in detail in sec .
we summarize and conclude in sec .
as has been already discussed in the literature , the dominant chanel through which doubly charged scalar pairs can be produced at hadron colliders is the drell - yan process @xmath45 in terms of @xmath46 in the parton c.m .
frame , the parton level differential cross section for this process is @xmath47 \big\ } , \end{aligned}\ ] ] where @xmath48 is the speed of @xmath9 in the c.m . frame .
the qcd corrections to this process have been also computed , yielding a next - to leading order ( nlo ) k - factor of the order of 1.25 at the lhc energy for the entire mass range between 200 gev and 1 tev @xcite . due to the stronger electromagnetic coupling of the doubly charged scalar
, one may anticipate a sizable two - photon contribution to the pair production . comparing with a singly charged scalar
, the two - photon channel will have an enhancement factor of 16 .
we consider the dominant contribution from the collinear photons and adopt the effective photon approximation .
the event rates from effective photon contributions can be just added to the drell - yan contributions to the @xmath49 final state . for the process
@xmath50 the matrix element squared is @xmath51 the corresponding differential cross section can be obtained as @xmath52 it is interesting to see the different @xmath53-dependence of the cross sections for the dy and two - photon processes . the dy process undergoes a pure @xmath54-wave channel , while the two - photon process contains all @xmath55 waves leading to a general dependence of @xmath56 near the threshold .
furthermore , the two - photon contributions may arise from both elastic and inelastic processes , including the semi - elastic case where it is elastic on one side and inelastic on the other .
the total cross - section can thus be written as @xcite @xmath57 the treatment in this section has ignored the channels from other gauge boson fusions such as @xmath60 ( virtual in general ) .
our estimates reveal that the production through @xmath31-fusion channel is rather suppressed and more so for @xmath61-fusion .
the cost of using the effective photon approximation rather than calculating the full @xmath62 subprocess is the loss of the potential tagging jets in the forward - backward regions .
fortunately , due to the rather clean charged leptonic final states , we do not require such jet tagging to identify the signal . in fig .
[ rate ] the total cross - sections are plotted against the mass of the doubly charged scalar , showing the drell - yan as well as different types of the two - photon processes .
the highest lower bound on a doubly charged higgs mass from the tevatron search with an integrated luminosity of 240 pb@xmath63 is about 136 gev ( though in the specific context of a left - right symmetric model ) .
it can be seen from fig
. [ rate ] that with 10 fb@xmath63 at the tevatron , the mass reach may be extended approximately to 250 gev . at the lhc , the production rate is increased by a factor of 20 over that at the tevatron , reaching the order of femtobarns for a higgs mass upto as much as 600 gev , and of 0.1 fb for a mass of 1 tev .
we have also shown the three classes of contributions from the two - photon channel separately .
it is interesting to note that the rate of the elastic process is larger than that of the inelastic at the tevatron energies , while the pattern is reversed at the lhc energies .
this is because the probability of a proton remaining unbroken after the emission of a photon in an elastic process is smaller at the lhc .
the semi - elastic contribution is numerically the highest in both cases , because of a factor of 2 for the initial state interchanges .
the ratio of the two - photon contribution relative to the drell - yan channel is shown in fig .
[ ratio ] .
in general , the two - photon contributions at lhc remain about 10% of that of the drell - yan process , while the fraction is much smaller for the tevatron , the reason being that is the photon needs to come with a larger momentum fraction of the parent parton in the latter case , for which the distribution function is relatively suppressed . it should be noted that , while the drell - yan cross - section has a next - to - leading order ( nlo ) qcd k - factor of about 1.25 @xcite , only the leading order cross - section has been presented in these figures .
this is for comparison of the correction to this cross - section due to the two - photon process and the strong correction mentioned above .
higher order corrections to the two photon electromagnetic process are rather small .
cteq6l parton distribution functions have been used in the calculation , and the factorization scale is set at half the subprocess center - of - mass energy @xmath64 . and leading order @xmath65 at the lhc and the tevatron.,width=321 ] we now examine some kinematic features of the two processes considered above .
as shown in fig .
[ pt ] , while the transverse momentum ( @xmath66 ) of the scalars tends to peak at values that are significant fractions of its mass , drell - yan production leads to scalars with harder @xmath66 distributions than those from the two - photon fusion .
this is due to the forward - backward nature of the @xmath67-channels in the two - photon process at high energies .
the rapidity @xmath68 distributions in fig .
[ eta ] reflect central peaking for both channels , and the distributions are broader for the two - photon process . on the whole , in spite of some qualitative differences between these two channels ,
it is not possible effectively to separate the two production mechanisms through the kinematical considerations . in our subsequent analysis , we will add them up incoherently .
the doubly charged higgs may in principle decay into all of the following two - body final states : @xmath69 the third and fourth channels depend on the mass splitting among the members of the triplet , and also the scalar potential . the consequence of the third channel being kinematically allowed ( which makes it the dominant decay , since it is driven by the su(2)@xmath0 gauge coupling ) has been discussed , for example , in reference @xcite , in the context of a linear collider . here we take a conservative approach and assume that the mass splitting within the triplet is very small as is the case , for example , in little higgs models .
thus only the first two decays listed above occur in our scenario , mode will affect neither our analyses nor the results for the individual channels @xmath70 and @xmath71 , since the branching ratios in these channels have been used by us as free parameters . however , the inclusion of this mode can in principle alter the overall discovery limit , depending upon the mass parameters . ]
the corresponding widths being given by @xcite @xmath72 in terms of the kinematic function @xmath73 and the scaled mass variable @xmath74 .
the approximate forms are valid for @xmath75 .
the subscripts @xmath76 and @xmath77 denote the transverse and longitudinal polarizations of the @xmath31 boson .
the longitudinal @xmath31 final state becomes dominant at higher @xmath78 .
the relative strengths of the two types of decays ( @xmath79 ) depend on the couplings @xmath29 as well the triplet vev @xmath15 , which has to be less than 1 gev in order to prevent large tree - level contributions to the @xmath27-parameter unless additional model assumptions are made .
while treating @xmath29 and @xmath15 as completely free parameters can lead to practically any relative strength between the two sets of final states , we have been guided by the additional constraint of neutrino mass generation , to saturate the bound of eq .
( [ yvev ] ) . of course
, @xmath80 can be even smaller if there are right - handed neutrinos in addition , for instance , a type ii seesaw mechanism is operative .
we shall further comment on this possibility at the end of this section . in fig .
[ br ] we show the branching fractions for the @xmath81 and @xmath3 decay modes as functions of the triplet vev ( left ) and the higgs mass ( right ) , keeping the overall constraint from neutrino mass mentioned above . with a higher mass of @xmath9 ,
the @xmath3 mode overtakes @xmath2 sooner due to the fast growing @xmath82 mode , even for a relatively smaller value of @xmath83 .
while the explanation is obvious from the expressions of the decay widths , this underlines the importance of exploring the latter mode at lhc , in addition to the easily identifiable like - sign lepton pair signal . and @xmath84 as functions of ( left ) the triplet vev , for two values of the doubly charged higgs mass , and ( right ) the higgs mass for two choices of @xmath15 .
, title="fig:",width=308 ] and @xmath84 as functions of ( left ) the triplet vev , for two values of the doubly charged higgs mass , and ( right ) the higgs mass for two choices of @xmath15 . , title="fig:",width=308 ] in units of @xmath85 m for the doubly charged higgs as a function of the triplet vev , within the neutrino mass constraint .
, width=321 ] since @xmath19 and @xmath15 are both quite small , it is natural to ask whether the decay width of @xmath9 can be so small that it becomes quasi - stable in the collider experiments .
as has been mentioned earlier , a long - lived doubly charged higgs has been looked for at the tevatron , in terms of highly ionizing tracks and muon - like penetration beyond the electromagnetic and hadron calorimeters @xcite . in fig .
[ dl ] , we plot the the proper decay length @xmath86 with @xmath87 being the proper lifetime of @xmath9 , keeping within the constraint on @xmath88 from neutrino mass to saturate eq .
( [ yvev ] ) .
it can be seen that , for a light scalar @xmath89 gev with certain values of @xmath15 , the proper decay length may exceed 10 microns .
such a decay length can be enhanced to a visible displaced secondary vertex by the appropriate boost @xmath90 .
the length goes further down for a more massive scalar .
thus it can be concluded that a doubly charged higgs can not be long - lived , or quasi - stable , on the scale of a collider detector if its @xmath10 coupling has to be the mechanism operative for the generation of neutrino masses .
the observation of a long - lived @xmath9 should therefore tell us that something over and above the @xmath10 interaction is active for generating neutrino masses .
it should also be remembered that @xmath91 can have six independent elements ( assuming a real symmetric majorana neutrino mass matrix ) in general .
values of these elements are highly model - dependent and they can lead to a wide number of possibilities in flavor diagonal as well as off - diagonal decays of the @xmath9 . while there is very little guideline as to the choice of these values , it should be emphasized that actual observation of the decays into different flavor combinations can actually give us information about the structure of the neutrino mass matrix . in order to convey the general features of the collider analysis
, we assume a diagonal structure of @xmath92 , with the same value for all three flavors , as shown in fig .
we include only the electronic and muonic branching ratios added together while predicting the signal events at the hadron colliders .
this analysis can always be translated into one with a variety of structures for the @xmath91 , without any serious difference in the results presented here .
distributions for same - sign ( solid ) opposite - sign ( dotted ) dileptons in the four - lepton final state . for opposite - sign dileptons , the isolation between the two harder leptons of each sign has been selected.,width=321 ] we first consider the most spectacular final state , namely , @xmath93 driven by the decay of the @xmath9 into lepton pairs .
as mentioned earlier , we have assumed the same branching ratio for flavor diagonal decays into the three lepton channels , and included the electron and muon pairs as our characteristic signals .
the branching ratio into lepton pairs has been taken as a free parameter , since , given the neutrino mass constraint , the decay rate into @xmath94 is automatically determined by it .
we consider both electrons and muons . to simulate the detector effects on the energy - momentum measurements , we smear the electromagnetic energy and the muon momentum tracking by a gaussian distribution
whose width is @xcite @xmath95 the signal consists of two like sign dilepton pairs with the same invariant mass . as can be seen from fig .
[ dr ] , the @xmath96 distribution ( where @xmath97 ) of the like - sign pairs peaks at a smaller value than that of opposite sign dileptons , thereby indicating a spatial separation of the pairs , at least when the doubly charged scalars are sufficiently boosted .
we have checked that this separation remains noticeable well beyond @xmath98 gev .
in addition , we have employed the following acceptance criteria for the events : * a veto on any opposite sign dilepton pair invariant mass being close to the @xmath61-boson mass : @xmath99 .
* @xmath100 15 gev , and the hardest leptons has @xmath101 30 gev .
* @xmath102 2.8 .
the above signal has _
prima facie _ no standard model background .
however , fake backgrounds may arise from @xmath31 bosons in conjunction with misidentified jets and/or leptons from heavy flavor decays . in order to suppress such backgrounds
, we recommend the following steps : * reconstruct the invariant mass of each like - sign dilepton pair , and insist on their being equal within about 5 per cent .
while the smearing procedure is seen to keep the invariant masses of the signal pairs equal well within this limit , backgrounds are largely eliminated by it . * demand a minimum @xmath96-isolation of 0.5 between each lepton and any nearby jet . *
demand little missing transverse energy : @xmath103 gev . plane , including the cuts discussed in the text , for an integrated luminosity of 300 fb@xmath63.,width=321 ] in fig .
[ evt ] we present event contours in the space spanned by the @xmath9 mass and the branching fraction @xmath104 for an integrated luminosity of 300 fb@xmath63 .
the drell - yan contribution has been multiplied here by the nlo k - factor 1.25 , and the two - photon contribution has also been included . in the absence of backgrounds
after all of the cuts discussed above , the contour corresponding to three events can be taken as the @xmath105 c.l .
discovery limit for an integrated luminosity of 300 fb@xmath63 .
the event rates suggest a high degree of detectability of the signal , even for rather small branching ratios . given the acceptance criteria listed above , it can be concluded from the contours that a doubly charged higgs up to the mass range of 1 tev can be detected at the lhc if the @xmath2 decay mode has a branching ratio on the order of @xmath106 per cent . on the lower side ,
even a branching ratio of 10 ( 5 ) per cent allows a @xmath105 c.l .
search limit of 500 ( 350 ) gev .
a recent similar study has appeared @xcite for the dy process and the leptonic decay mode of @xmath9 only .
their numerical results , wherever overlapping , are in agreement with ours
. events
. the solid ( dotted ) line corresponds to event rates coming from each doubly charged scalar decaying into two like - sign @xmath31 s without ( with ) basic cuts as described in the text .
event rates from various types of background are also shown , without the basic cuts .
( b ) right : equal event contours in the @xmath107 plane , including the cuts discussed in the text , for an integrated luminosity of 300 fb@xmath63 .
the dashed curve is a reproduction of the 3-event contour in fig .
, title="fig:",width=308 ] events
. the solid ( dotted ) line corresponds to event rates coming from each doubly charged scalar decaying into two like - sign @xmath31 s without ( with ) basic cuts as described in the text .
event rates from various types of background are also shown , without the basic cuts .
( b ) right : equal event contours in the @xmath107 plane , including the cuts discussed in the text , for an integrated luminosity of 300 fb@xmath63 .
the dashed curve is a reproduction of the 3-event contour in fig .
, title="fig:",width=308 ] next , let us consider the @xmath3 channel . as can be seen from fig .
[ br ] , this channel may become dominant even for rather small values of the triplet vev , especially if the @xmath9 mass is on the higher side .
it is therefore desirable to devise search strategies for this channel , leading to a final state consisting of @xmath108 . in order to confirm the nature of a doubly charged state , and to reveal its resonant production , we propose to reconstruct the events by looking for two like - sign @xmath31 s through a pair of like - sign dileptons and
the remaining two in their hadronic decays which allow complete reconstruction of @xmath78 .
the branching fraction for this decay chain is taken to be @xmath109 the predicted cross sections including the above branching fraction are shown in fig .
[ ww](a ) , assuming br(@xmath40 ) to be 100% .
the rather small branching fraction in eq .
( [ br4w ] ) is a price we pay for cleaner signals , and consequently the event rate tends to be low for a high @xmath9 mass .
for instance , it may only have a handful events for @xmath110 tev with an integrated luminosity of 300 fb@xmath63 , even before any acceptance cuts .
we again start with some basic cuts " , where the leptons are subject to the same acceptance criteria as before . in addition
, we demand * @xmath111 * @xmath112 , @xmath113 the jet energies are also smeared using the same gaussian formula @xcite as in eq .
( [ ecal ] ) , but with @xmath114 as seen from the solid and dotted curves in fig .
[ ww](a ) , these criteria do not affect the signal in a significant way , especially for heavier masses . in fig .
[ ww](b ) we translate the signal cross section to equal event contours in the @xmath107 plane , including the cuts discussed above and both @xmath2 and @xmath115 , for an integrated luminosity of 300 fb@xmath63 . for comparison , we reproduce the 3-event contour in fig .
[ evt ] by the darshed curve , assuming @xmath116 .
this illustrates the complementarity between these two channels for different values of the br s . as for the sm backgrounds ,
the purely electroweak production of @xmath117 final states is rather modest @xcite , as indicated in fig .
[ ww](a ) with the branching fraction of eq .
( [ br4w ] ) included .
a next background leading to @xmath117 final states from @xmath118 production @xcite is larger but the kinematics will look very different from the signal processes due to the four extra hard @xmath119 jets .
it is easy to demonstrate that these backgrounds can be effectively suppressed well below the signal by judicial kinematical cuts .
the largest background arises from @xmath120 production .
in addition , if the @xmath121 mode from the sm higgs boson ( @xmath122 ) has a substantial branching ratio , then the @xmath123 production @xcite may constitute a severe background . as shown in @xcite , this production cross - section for @xmath124
can be of the order of 150 fb for a higgs mass of 200 gev when it has unsuppressed decays into two @xmath125s , and this can contribute enough 4-jet plus like - sign dilepton events to swamp our signals .
we have found that the event selection strategies which can suppress the leading @xmath120 background are also effective in reducing the fake events coming from @xmath124 .
therefore , we outline below a step - by - step procedure for reducing the leading background , as summarised in table [ tab : i ] .
the relatively smaller backgrounds of @xmath126 and @xmath117 are also taken care of by the same procedure . for illustration , we first choose a doubly - charged higgs mass of 300 gev .
in the results presented , the branching ratio for @xmath127 has been taken as unity .
first of all , we wish to select events with four jets and thus we veto events with additional central jets in the region , @xmath128 then we experiment with progressively stronger cuts on the ( leading ) lepton and jet @xmath66 .
furthermore , the four jets will pair up to reconstruct two @xmath129 peaks , and we thus demand @xmath130 gev for the pair closest to the @xmath129 peak and the other pair at the same time . as seen from the table , this reduces both the @xmath120 and @xmath131 backgrounds rather drastically , but not to the extent necessary for discerning the signal .
therefore , one should note that the pair production of the heavy higgs bosons has a high energy threshold .
we accordingly define a cluster transverse mass for the whole system @xmath132 which should start from a threshold around @xmath133 .
we impose a cut of @xmath134 gev on this variable . for the @xmath38 kinematics on the leptonic side ,
we can only define a transverse mass due to the missing neutrinos @xmath135 and a cut of 300 gev on this quantity proves to be effective in further reducing the background . on the hadronic side
, the 4-jet invariant mass should peak at the location of @xmath78 .
taking into account the detector resolution , we look for the events in the window @xmath136 gev , which reduces the background by another order of magnitude .
two sets of numerical values for the cuts have been presented in table [ tab : i ] , to establish the fact that the successive criteria applied here can reduce the backgrounds by more than two orders of magnitude , while the survival efficiency for the signal remains as high as @xmath137 .
this example illustrates how the rather distinctive kinematical features of the signal allow us to effectively suppress the sm backgrounds , after which the signal observation is mainly a statistical issue .
.the signal and the leading background rates for @xmath138 .
the rates after imposing each selection criterion , as described in the text , are shown .
@xmath140 for @xmath120 calculation . [ cols="^,^,^,^,^,^,^,^",options="header " , ]
we have investigated the visibility of the pair - production of doubly charged scalars at the lhc .
such a scalar is assumed to belong to an su(2)@xmath0 triplet which can generate majorana masses for left - handed neutrinos through the vev of its neutral component .
pair - production , in spite of its relative kinematical suppression , has the advantage of being relatively model - independent , and is not driven by the vev of the triplet .
we find that , while the contribution comes largely from the drell - yan process , two - photon fusion also contributes at the level of 10% at the lhc due to the substantially enhanced electromagnetic coupling .
this , we emphasize , is comparable to the qcd correction to the drell - yan channel , and must be included in a complete and accurate estimate .
signatures of the pair - produced scalars have been investigated over the entire parameter space where viable neutrino masses can arise from the scalar triplet .
we have covered both the regions where they decay dominantly into like - sign dileptons and like - sign @xmath31 s .
the former are largely free from backgrounds , and a mass range upto about @xmath141 to @xmath142 can be probed with an integrated luminosity of 300 @xmath143 , if the doubly charged higgs has a branching ratio of at least @xmath144 in this channel . in the latter channel , we have suggested optimal ways of eliminating the standard model backgrounds with high efficiency for the signal retention .
the signal observation is mainly a statistical issue , and a doubly charged scalar of mass up to about 700 gev should be identifiable , if it dominantly decays into a pair of like - sign @xmath31 s . with the spirit of neutrino mass generation of eq .
( 2 ) and the signal complementarity between the @xmath145 and @xmath146 channels as shown in fig .
[ ww](b ) by the 3-event contours , we claim a complete coverage for @xmath147 gev with any arbitrary decay of these two modes .
there has been great interest of searches for long - lived doubly charged scalars .
we found that in order to have the requisite contribution to neutrino masses , it is not possible for the scalar to be long - lived , although it is possible to leave a perceptible decay gap at the detector for @xmath148 gev .
any observation of long - lived scalars of this kind will therefore be a pointer towards some alternative theory of neutrino masses such as the type ii seesaw mechanism .
finally , we wish to emphasize that a study of the decays of the doubly charged scalar into different lepton flavors , diagonal as well as off - diagonal , is a very useful probe of the mechanism whereby neutrino masses arise from interaction with an su(2)@xmath0 scalar triplet .
this probe is eminently within the scope of the lhc .
we thank patha konar for helpful discussions . the work of th and kw
is supported in part by the u.s .
department of energy under grant no .
de - fg02 - 95er40896 , by the wisconsin alumni research foundation .
bm would like to thank the department of atomic energy , india , for support through the xth and xith 5-year plan projects .
si is supported by nnsfc , ncet and huoyingdong foundation .
bm and si would like to acknowledge the hospitality of the phenomenology institute , university of wisconsin - madison while the work was in progress and completed .
99 g. senjanovic and r. n. mohapatra , phys .
d * 12 * , 1502 ( 1975 ) .
r. n. mohapatra and j. c. pati , phys .
d * 11 * , 2558 ( 1975 ) .
h. georgi and m. machacek , nucl .
b * 262 * , 463 ( 1985 ) .
j. f. gunion , j. grifols , a. mendez , b. kayser and f. i. olness , phys .
d * 40 * , 1546 ( 1989 ) .
n. arkani - hamed , a. g. cohen and h. georgi , phys .
b * 513 * , 232 ( 2001 ) [ arxiv : hep - ph/0105239 ] .
n. arkani - hamed , a. g. cohen , e. katz , a. e. nelson , t. gregoire and j. g. wacker , jhep * 0208 * , 021 ( 2002 ) [ arxiv : hep - ph/0206020 ] .
m. schmaltz , jhep * 0408 * , 056 ( 2004 ) [ arxiv : hep - ph/0407143 ] .
t. han , h. e. logan , b. mcelrath and l. t. wang , phys .
d * 67 * , 095004 ( 2003 ) [ arxiv : hep - ph/0301040 ] .
d. acosta _ et al .
_ [ cdf collaboration ] , phys . rev .
lett . * 93 * , 221802 ( 2004 ) [ arxiv : hep - ex/0406073 ] .
d. acosta _ et al .
_ [ cdf collaboration ] , phys . rev . lett . * 95 * , 071801 ( 2005 ) [ arxiv : hep - ex/0503004 ] . v. m. abazov _ et al . _
[ d0 collaboration ] , phys .
* 93 * , 141801 ( 2004 ) [ arxiv : hep - ex/0404015 ] .
j. a. coarasa , a. mendez and j. sola , phys .
b * 374 * , 131 ( 1996 ) [ arxiv : hep - ph/9511297 ] .
e. ma , m. raidal and u. sarkar , nucl .
b * 615 * , 313 ( 2001 ) [ arxiv : hep - ph/0012101 ] .
p. h. frampton , m. c. oh and t. yoshikawa , phys .
d * 66 * , 033007 ( 2002 ) [ arxiv : hep - ph/0204273 ] .
e. j. chun , k. y. lee and s. c. park , phys .
b * 566 * , 142 ( 2003 ) [ arxiv : hep - ph/0304069 ] .
t. han , h. e. logan , b. mukhopadhyaya and r. srikanth , phys .
d * 72 * , 053007 ( 2005 ) [ arxiv : hep - ph/0505260 ] .
j. y. lee , jhep * 0506 * , 060 ( 2005 ) [ arxiv : hep - ph/0501118 ]
. e. accomando _ et al .
_ , arxiv : hep - ph/0608079 . w. m. yao _ et al . _
[ particle data group ] , j. phys .
g * 33 * , 1 ( 2006 ) . for recent review
elliott and j. engel , j. physics . * g30 * r183 ( 2004 ) [ arxiv : hep - ph/0405078 ] .
g. belanger , f. boudjema , d. london and h. nadeau , phys .
d * 53 * , 6292 ( 1996 ) [ arxiv : hep - ph/9508317 ] .
d. london , arxiv : hep - ph/9907419 .
p. benes , a. faessler , f. simkovic and s. kovalenko , phys .
d * 71 * , 077901 ( 2005 ) [ arxiv : hep - ph/0501295 ] .
m. hirsch , h. v. klapdor - kleingrothaus and o. panella , phys .
b * 374 * , 7 ( 1996 ) [ arxiv : hep - ph/9602306 ] . c. aalseth _
et al . _ ,
arxiv : hep - ph/0412300 . j. f. gunion , r. vega and j. wudka , phys .
d * 43 * , 2322 ( 1991 ) . j. f. gunion , r. vega and j. wudka , phys .
d * 42 * , 1673 ( 1990 ) .
r. s. chivukula and h. georgi , phys .
b * 182 * , 181 ( 1986 ) .
r. vega and d. dicus , nucl .
b*329 * , 533 ( 1990 ) m. muhlleitner and m. spira , phys .
d * 68 * , 117701 ( 2003 ) [ arxiv : hep - ph/0305288 ]
. m. drees , r. m. godbole , m. nowakowski and s. d. rindani , phys .
d * 50 * , 2335 ( 1994 ) [ arxiv : hep - ph/9403368 ] .
s. chakrabarti , d. choudhury , r. m. godbole and b. mukhopadhyaya , phys .
b * 434 * , 347 ( 1998 ) [ arxiv : hep - ph/9804297 ] .
a. hektor , m. kadastik , m. muntel , m. raidal and l. rebane , arxiv:0705.1495 [ hep - ph ] . v. d. barger , t. han and h. pi , phys .
d * 41 * , 824 ( 1990 ) .
v. d. barger , a. l. stange and r. j. n. phillips , phys .
d * 45 * , 1484 ( 1992 ) .
w. beenakker , s. dittmaier , m. kramer , b. plumper , m. spira and p. m. zerwas , phys .
lett . * 87 * , 201805 ( 2001 ) [ arxiv : hep - ph/0107081 ] .
s. dawson , l. h. orr , l. reina and d. wackeroth , phys .
d * 67 * , 071503(r ) ( 2003 ) [ arxiv : hep - ph/0211438 ] . | we study the pair production of doubly charged higgs bosons at the large hadron collider ( lhc ) , assuming the doubly charged higgs to be part of an su(2)@xmath0 triplet which generates majorana masses for left - handed neutrinos .
such pair - production has the advantage that it is not constrained by the triplet vacuum expectation value , which tends to make the single production rate rather small .
we point out that , in addition to the drell - yan ( dy ) production mechanism , two - photon processes also contribute to @xmath1 production at a level comparable to the qcd corrections to the dy channel .
decays of the doubly charged higgs into both the @xmath2 and @xmath3 modes are studied in detail to optimize the signal observation over the backgrounds .
doubly charged scalars should be observable at the lhc with 300 @xmath4 integrated luminosity in the @xmath5 channel upto the mass range of 1 tev even with a branching fraction of about @xmath6 , and in the @xmath7 channel upto a mass of 700 gev .
such a doubly charged higgs , if it is a member of a triplet generating neutrino masses , can not be long - lived on the scale of collider detectors although it might lead to a displaced secondary vertex during its decay if it is lighter than about 250 gev . madph-07 - 1488 + hri - p-07 - 05 - 001 |
polar codes @xcite provably achieves the capacity of channels by using the ( low complexity ) decoding algorithm , in the limit of infinite block length . at short block lengths polar codes under decoding tend to exhibit a poor performance . in @xcite
it was suggested that such a behavior might be due , on one hand , to an intrinsic weakness of polar codes and , on the other hand , to the sub - optimality of decoding w.r.t . decoding .
an improved decoding algorithms were proposed in @xcite , while the structural properties of polar codes ( e.g. , their distance properties ) were studied , among others , in @xcite .
the minimum distance properties of polar codes can be improved by resorting to concatenated schemes like the one of @xcite , where the concatenation of polar codes with an outer code is considered .
this solution , together with the use of the list decoding algorithm of @xcite , allows short polar codes to become competitive against other families of codes @xcite .
a theoretical characterization of the performance of concatenated -polar codes is still an open problem .
furthermore , for a fixed code length , a concatenated scheme can be realized with several combinations of the component codes parameters ( e.g. , one may choose various polynomials and polar codes designed for different target signal - to - noise ratios ) . in this paper
, we provide an analysis of the concatenation of polar codes with binary cyclic outer codes .
the analysis is carried out by introducing concatenated ensembles and by deriving by following the well - known uniform interleaver approach @xcite . in the analysis , the knowledge of the of the inner polar code
is required . however , the polar code calculation through analytical methods is still an unsolved problem .
hence , we restrict our attention to short , high - rate polar codes for which the problem can be solved through a pragmatic approach . more precisely , we consider the dual code of the selected polar code and then we find its by listing the codewords .
subsequently , by using the generalized macwilliams identity @xcite , we obtain the of the original polar code .
the of the outer cyclic code , instead , is computed by following the method presented in @xcite . by adopting the uniform interleaver approach , we subsume the existence of an interleaver between the inner and the outer code , and
obtain the average performance of an ensemble composed by the codes obtained by selecting all possible interleavers .
our analysis shows that the performance of the concatenated scheme with and without interleaver ( as proposed in @xcite ) may differ substantially .
similarly , by considering both and outer codes , we show that the choice of the outer code plays an important role in the short block length regime .
the paper is organized as follows .
notation and definitions are introduced in section [ sec : definitions ] .
the distance spectrum analysis of the concatenated scheme is discussed in section [ sec : upperunionbound ] .
numerical results are reported in section [ sec : numericalresults ] .
finally , section [ sec : conclusion ] concludes the paper .
let @xmath0 be the hamming weight of a vector .
we denote as @xmath1 and @xmath2 the outer cyclic code length and dimension , respectively , while @xmath3 and @xmath4 identify the same parameters for the inner polar code .
therefore , the code rates of the outer and inner code are @xmath5 and @xmath6 , respectively .
the two codes can be serially concatenated on condition that @xmath7 , thus the overall code rate of the concatenated code is @xmath8 .
given a binary linear code @xmath9 , its is defined as @xcite @xmath10 where @xmath11 is the number of codewords @xmath12 with @xmath13 . in this work
we focus on systematic polar codes ( the reason of this choice will be discussed later ) so the first @xmath4 bits of a codeword @xmath12 coincide with the information vector @xmath14 , yielding @xmath15 with @xmath16 being the parity vector .
the of a code @xmath9 is @xmath17 where @xmath18 is the multiplicity of codewords @xmath12 with @xmath19 and @xmath20 .
the enumeration of the codeword weights entails a large complexity even for small code dimensions . in order to overcome this problem and obtain the of the considered polar codes
, we focus on short , high - rate polar codes and we exploit the generalized macwilliams identity @xcite .
this approach was followed also for cyclic codes , for example , in @xcite to compute the of several .
denote by @xmath21 the dual code of @xmath22 .
given the dual code @xmath23 , we can express the original code @xmath24 as @xcite @xmath25 where @xmath26 is the cardinality of the dual code .
when the is of interest , a significant reduction of the computational cost can be achieved by considering systematic codes .
for such reason , in this work we have used only systematic inner polar codes . in the case of a systematic code @xmath9 , it is convenient to derive the from the @xmath27 defined as @xmath28 where @xmath29 is the multiplicity of codewords @xmath12 with @xmath19 and @xmath30 , with @xmath31 and @xmath32 .
hence , starting from the of the dual code @xmath33 , we have @xcite @xmath34 then , the is obtained as @xmath35
given an ensemble of binary linear codes @xmath36 , the expected @xmath37 of a random code @xmath38 under decoding over a with erasure probability @xmath39 can be upper bounded as @xcite @xmath40 \leq p_b^{(s)}(n , k,\epsilon ) \nonumber \\ & + \sum_{e=1}^k \binom{n}{e } \epsilon^e(1-\epsilon)^{n - e } \min\left\{1,\sum_{\omega=1}^e\binom{e}{n}\frac{\bar{a}_\omega}{\binom{n}{\omega}}\right\ } \label{eq : upperunionbound}\end{aligned}\ ] ] where @xmath41 $ ] is the average multiplicity of codewords with @xmath20 and @xmath42 represents the of an ideal code , with parameters @xmath3 and @xmath4 .
the in is applicable to every code ensemble whose expected is known . in order to use , the of the concatenated code ensemble must be known .
when dealing with concatenated codes , it is commonplace to consider a general setting including an interleaver between the inner and the outer codes .
the concatenated code ensemble is hence given by the codes obtained by selecting all possible interleavers . the special case without any interleaver can then be modeled as an identity interleaver . from @xcite
, the of a concatenation formed by an inner polar code and an outer cyclic code can be obtained from the cyclic code and the polar code as @xmath43 where @xmath44 is the weight enumerator of the outer code and @xmath45 is the input - output weight enumerator of the inner code ( we remind that the average multiplicities resulting from are , in general , real numbers ) .
the ensemble @xmath36 contains the codes generated by all possible interleavers .
thus , also bad codes ( i.e. , characterized by bad error rate performance ) belong to the ensemble .
it is clear that the bad codes adversely affect the obtained through causing a too pessimistic estimate of the error probability obtained through , with respect to that achieved by properly designed codes .
a simple way to overcome this issue is to divide @xmath36 into the bad and good code subsets , and then derive the only of good codes through the expurgated @xmath36 @xcite .
in fact , @xmath46 , where @xmath47 and @xmath48 denote the good and the bad codes ensemble , respectively . in this work we have assumed @xmath49 , hence at least one code belongs to the good codes ensemble .
therefore , when the expurgation method is adoptable ( i.e. , the first term of the has a codewords multiplicity less than @xmath50 ) , through and the average performance of the good codes subset is derived .
studying the dual codes and exploiting macwilliams identities allow considerable reductions in complexity of exhaustive analysis as long as the original code rate is sufficiently large .
this will be the case for the component codes considered next , which are characterized by @xmath51 .
therefore , in our case and can effectively be exploited to calculate @xmath44 and @xmath45 in .
in this section we consider several examples of polar - cyclic concatenated codes and assess their performance through the approach described in the previous sections .
our focus is on short , high rate component codes , which allow to perform exhaustive analysis of their duals in a reasonable time .
our results are obtained considering a with erasure probability @xmath39 . as known , in the polar code construction a fixed value of the error transition probability is considered .
therefore , as usual in literature , in each of the following examples the polar code is designed by using @xmath52 .
all performance curves provided next are obtained through .
we consider codes with @xmath53 bits and a or a outer code , with @xmath1 according to the polar code dimension . in this work ,
we have used a @xmath54 and a @xmath55 with the following generator polynomials @xmath56 @xcite : * @xmath54 : @xmath57 ; * @xmath58ccitt : @xmath59 . instead ,
regarding the code , we study two codes that have the same redundancy as the -8 and -16 codes . thus , for the same @xmath60 value , a performance comparison of the results achieved by using a or a outer code is feasible and fair . in order to keep the selected polar codes unchanged ,
we have considered shortened codes . for the case without interleaver considered in @xcite ,
the generator matrix @xmath61 of the concatenated code can be obtained from the cyclic and polar code generator matrices @xmath62 and @xmath63 , respectively , as @xmath64 when we instead consider the more general case with an interleaver between the inner and outer codes , we have @xmath65 where @xmath66 is an @xmath67 permutation matrix representing the interleaver . considering all codes in @xmath36 ,
leads to their and hence to the average performance in terms of , according to the uniform interleaver approach . in order to have an idea of the gap between this average performance and those of single codes in the ensemble , in each of the following figures we include the performance of codes randomly picked in @xmath36 . without changing the outer and inner code ,
this can be easily done by introducing a random interleaver between the cyclic code and the polar code in the concatenated scheme .
hence , in this case , the matrix @xmath66 in is a random permutation matrix . for readability reasons , in addition to the solution without interleaver , in each of the following examples , we have considered only @xmath68 random interleavers ; however , the obtained results allow to address some general conclusions .
moreover , when the results of the expurgated @xmath36 differs from that achieved by the , its performance in terms of is also considered .
the of the considered polar code alone is also included as a reference .
clearly , the comparison between the performance of the polar code and those of the concatenated schemes is not completely fair at least because of the different code rates ; however , in this way , the performance gain achieved through concatenation can be pointed out .
figures [ fig : polar64 - 48-crc8 ] and [ fig : polar64 - 48-bch48 - 40 ] show the of the concatenated scheme , with and without random interleaver , formed by a polar code with @xmath69 and an outer cyclic code with @xmath70 ( i.e. , @xmath71 ) in terms of . in fig .
[ fig : polar64 - 48-crc8 ] and fig .
[ fig : polar64 - 48-bch48 - 40 ] a -8 and a ( 48 , 40 ) outer code is considered , respectively , the latter obtained by shortening the ( 255 , 247 ) code . in both the examples , the result obtained through the corresponds to an average behavior , while some interleaver configurations achieve a smaller .
this trend is due to the different minimum distance @xmath72 of the codes .
the results in figs .
[ fig : polar64 - 48-crc8 ] and [ fig : polar64 - 48-bch48 - 40 ] are very similar , showing that , for this particular case , there is no substantial difference in performance arising from the different type of outer code . this conclusion is also supported by the results in tab . [
tab : minimumdistance ] , where the number of concatenated schemes with random interleaver corresponding to a minimum distance value is reported . from these figures
we can observe the performance gain introduced by the concatenated scheme with respect to the ( 64,48 ) polar code used alone , that has @xmath73 . in both the examples
the and the concatenated code without interleaver have @xmath73 and @xmath74 , respectively .
code over the under decoding .
performance of the ( 64,48 ) polar code alone is also reported.,width=340 ] .number of concatenated codes with random interleaver for a given minimum distance value [ cols="^,^,^,^",options="header " , ] in fig .
[ fig : polar64 - 48-crc16 ] and fig .
[ fig : polar64 - 48-bch48 - 32 ] the of the concatenated codes , with and without interleaver , composed by a polar code with @xmath69 and an outer code with @xmath75 ( i.e. , @xmath76 ) in terms of is plotted . in figs . [
fig : polar64 - 48-crc16 ] and [ fig : polar64 - 48-bch48 - 32 ] a -16 and a ( 48 , 32 ) outer code is considered , respectively , the latter obtained by shortening the ( 255 , 239 ) code . differently from the previous figures ,
the obtained with the expurgated is now available . as in figs .
[ fig : polar64 - 48-crc8 ] and [ fig : polar64 - 48-bch48 - 40 ] , the curve obtained through the well describes the ensemble average performance , while we see that the curve corresponding to the expurgated belongs to the group of best codes .
also in these cases the performance gap between the ( 64,48 ) polar code alone and the concatenated codes is remarkable . in both the examples
the has @xmath74 , while the expurgated and the concatenated scheme without interleaver have @xmath77 .
however , from fig . [
fig : polar64 - 48-bch48 - 32 ] , we can observe that , on the contrary to fig .
[ fig : polar64 - 48-crc16 ] , all curves of concatenated codes are very close .
in fact , for the case in fig . [
fig : polar64 - 48-bch48 - 32 ] only the results in @xmath74 but with a codewords multiplicity equal to 0.0336 , instead the realizations of concatenated codes have @xmath77 , as shown in tab .
[ tab : minimumdistance ] .
therefore , differently from the , in this case the use of a code is able to increase the minimum distance of the concatenated code ( also for the solution without interleaver ) ; thus , for this specific case , the code should be preferred to the code .
code over the under decoding .
performance of the ( 64,48 ) polar code alone is also reported.,width=340 ] in all the previous figures the curve of the concatenated scheme without interleaver proposed in @xcite falls within the group of best performing codes .
this may lead to the conclusion that this configuration always produces a good result ; in reality this trend is not preserved for any choice of the code parameters .
an example of the latter kind is shown in fig .
[ fig : polar64 - 56-crc16 ] , where a polar code with @xmath78 and a -16 code ( i.e. , @xmath79 and @xmath71 ) are considered . from the figure we observe that , in this case , the introduction of a random interleaver can improve the scheme without interleaving . also for this example , tab .
[ tab : minimumdistance ] summarizes the number of concatenated schemes with interleaver for each value of the code minimum distance . in this case
the polar code has @xmath80 , while both the and the concatenated scheme without interleaver have @xmath73 .
instead , the expurgated belongs to the group of good codes with @xmath74 .
we have found a similar result also by using the -8 code in place of the -16 but it is omitted for the sake of brevity .
so , these counter examples ( others can be found ) clearly demonstrate that the use of a selected interleaver may be beneficial from the error rate viewpoint .
code over the under decoding .
performance of the ( 64,56 ) polar code alone is also reported.,width=340 ]
in this paper , the analysis of the performance of the concatenation of a short polar code with an outer binary linear block code is addressed from a distance spectrum viewpoint .
the analysis is carried out for the case where an outer cyclic code is employed together with an inner systematic polar code . by introducing an interleaver at the input of the polar encoder ,
we show that remarkable differences on the block error probability at low erasure probabilities can be observed for various permutations .
the variations are due to the change in the overall concatenated code minimum distance ( and minimum distance multiplicity ) induced by the choice of the interleaver .
bounds on the achievable error rates under maximum likelihood decoding are obtained by applying the union bound to the ( expurgated ) average weight enumerators .
the results point to the need of careful optimization of the outer code , at least in the short block length regime , to attain low error floors .
e. arikan , `` channel polarization : a method for constructing capacity - achieving codes for symmetric binary - input memoryless channels , '' _ ieee trans .
inf . theory _
55 , no . 7 , pp . 30513073 , jul .
2009 .
m. bardet , v. dragoi , a. otmani , and j. p. tillich , `` algebraic properties of polar codes from a new polynomial formalism , '' in _ proc .
inf . theory ( isit ) _ , barcelona , spain , jul .
2016 , pp . 230234 .
a. bhatia , v. taranalli , p. h. siegel , s. dahandeh , a. r. krishnan , p. lee , d. qin , m. sharma , and t. yeo , `` polar codes for magnetic recording channels , '' in _ proc .
theory workshop ( itw ) _ , jerusalem , israel , apr .
2015 , pp . 15 .
j. k. wolf and r. d. blakeney , `` an exact evaluation of the probability of undetected error for certain shortened binary crc codes , '' in _ proc .
21st ieee military commun .
_ , vol . 1 , san diego , ca , usa , oct .
1988 , pp .
287292 .
f. chiaraluce and r. garello , `` extended hamming product codes analytical performance evaluation for low error rate applications , '' _ ieee trans . on wireless commun .
_ , vol . 3 , no . 6 ,
. 23532361 , nov . | in this paper , the analysis of the performance of the concatenation of a short polar code with an outer binary linear block code is addressed from a distance spectrum viewpoint .
the analysis targets the case where an outer cyclic code is employed together with an inner systematic polar code .
a concatenated code ensemble is introduced placing an interleaver at the input of the polar encoder .
the introduced ensemble allows deriving bounds on the achievable error rates under maximum likelihood decoding , by applying the union bound to the ( expurgated ) average weight enumerators .
the analysis suggests the need of careful optimization of the outer code , to attain low error floors . |
spectroscopic analyses of stars with planets ( swps ) have revealed that they , as a group , exhibit higher metallicity than randomly selected samples of nearby stars @xcite .
@xcite reviewed the three explanations proposed to account for this difference .
one of these , the `` self enrichment '' hypothesis , posits that swps accreted sufficient mass of metals to enhance the metallicity of their photospheres by observable amounts .
@xcite suggested that accretion of metals by a star might be seen via its effects on the correlation between its photospheric abundances and the elemental condensation temperatures ( t@xmath0 ) . this would be expected if , for example , a star accretes rocky material that formed significantly inside the water ice radius in the protoplanetary disk
. such material would be relatively more depleted in volatiles ( e.g. , c , n and o ) than in refractories ( e.g. , al , ca and ti ) .
material in the inner disk could find its way into the central star , for example , from gravitational scattering by a migrating giant planet .
@xcite were the first to search for anomalous trends between abundances and t@xmath0 among swps .
they compared the slopes from some 15 elements in 30 swps and 102 comparison stars .
they found 5 or 6 swps with significantly higher abundance - t@xmath0 slopes than the comparison stars .
@xcite also noted that these stars tended to be hotter .
however , @xcite and @xcite analyzed 14 and 12 swps , respectively , finding no significant correlations between abundances and t@xmath0 .
@xcite studied six common proper motion main sequence stars . for one pair , hd 219542 a and b
, they found a significant trend between the differences in their abundances and t@xmath0 .
@xcite also studied the pair , but they failed to find such a trend .
taken together , these studies leave the question of abundance - t@xmath0 trends unresolved . over the last two years
, many more swps have been discovered and analyzed using detailed spectroscopic methods . in this study
we revisit the question of abundance - t@xmath0 trends among swps using the most recent data sets .
we discuss the sample preparation , analysis and interpretation of results .
we also search for a trend between the solar photospheric and meteoritic abundance differences and t@xmath0 .
a weakness of the smith et al .
study was the heterogeneous nature of their abundance data . in order to avoid possible systematic offsets among different sets of abundance determinations and between the swp and comparison stars samples ,
the best strategy is to employ abundance data from a single source . with this in mind , we have selected for our analysis published data for swps and comparison stars from the following studies : c , s , zn , cu @xcite ; na , mg , al @xcite ; si , ca , sc , ti , v , cr , mn , ni @xcite .
these studies make use of the same analysis methods .
the atmospheric parameters for the swps and comparison stars are taken from @xcite .
the data culled from these studies formed the first version of our database .
it contains abundance data for 76 swps and 37 comparison stars .
more elements could have been included , but the number of stars lacking a complete set of abundance determinations increases as the number of elements is increased . still , some of the stars in the database lack abundance determinations for the full suite of elements listed above . to mitigate this deficit ,
we also included abundance determinations from @xcite , @xcite and @xcite for the following elements : c , s , na , al and mn .
the offsets between the abundances reported in the two sets of studies for each element were calculated using stars common to both studies and then applied prior to importing these data into the database .
typically , these other abundance determinations allowed us to fill in missing data for two or three swps per element . even with these additions ,
the database still contains missing data ; 64 swps and 33 comparison stars in the database are missing at most one element abundance value . since it is important that the same elements be analyzed for the swps and comparison stars , only these 97 stars were included in the final version of the database .
one of the analyses described in the next section make use of the uncertainties in the abundance determinations .
we adopted the uncertainties reported in each respective study . in a few instances uncertainties
were reported as `` 0 '' ( mostly for mn ) in the original studies ; in such cases , an average value of the uncertainty determined from other stars in the study was adopted instead .
values for t@xmath0 for each element are from table 8 of @xcite .
the t@xmath0 value for a given element corresponds to the temperature at which 50% of it condenses out of the gas phase as the temperature is reduced .
the values of t@xmath0 for the elements included in the database range from 40 k for c to 1659 k for sc ; the c abundance , in particular , provides considerable leverage in calculating the slope between abundance and t@xmath0 for each star .
for each star in the database , we determined a linear least squares solution between the abundances and t@xmath0 values .
two versions of least squares analysis were employed .
the first makes use of weighted least squares using the uncertainties in the abundances .
the results of this analysis are shown in fig . 1 with the t@xmath0 slope values plotted against [ fe / h ] .
four stars , all swps , deviate significantly from the general trend ; hd 114783 is much higher than all the other points , while hd 19994 , hd 9826 and hd 179949 are slightly low .
slope is plotted against [ fe / h ] for swps ( dots ) and comparison stars ( open circles ) the t@xmath0 slope for each star was determined from a weighted least squares analysis.,width=316 ] the second method is a standard unweighted linear least squares analysis .
we show the results of this analysis in fig . 2 .
in this case only one star , hd 5133 a comparison star , deviates significantly from the others .
the overall scatter among the points in fig .
2 is also smaller than in fig . 1 .
while in principle weighted least squares should be preferable to unweighted least squares , in this case it is not . in order to obtain reliable results from weighted least squares ,
it is necessary to know the uncertainties accurately .
those elements for which the uncertainties in the abundances are underestimated will be given too much weight in the solution .
to illustrate this , we plot abundances against t@xmath0 for hd 9826 and hd 114783 in fig .
3 . from the weighted least squares analysis ,
the slope for hd 9826 was found to be negative , while it was positive for hd 114783 . in each case , the derived slope is strongly biased by 2 or 3 data points that have much smaller uncertainties than the other data ( see fig .
however , for both stars the slopes derived form the standard least squares analysis yielded values indistinguishable from zero . for the remainder of the analysis we will rely on the unweighted analysis results . for hd 9826 ( panel a ) and hd 114783 ( panel b).,width=316 ] as noted by @xcite ,
a trend between the t@xmath0 slope and t@xmath1 can be a signature of accretion .
hotter stars have thinner outer convection zones and are thus more likely to display the effects of accretion . to test for this
, we performed a multiple linear regression analysis on the unweighted t@xmath0 slopes with effective temperature ( t@xmath1 ) and [ fe / h ] as independent parameters .
the resulting equation for the t@xmath0 slopes of the swps is : @xmath2 t@xmath3}$ ] and the equation for the comparison stars is ( excluding the outlier ) : @xmath4 t@xmath5}$ ] the zero - points and t@xmath1 coefficients differ marginally at the one-@xmath6 level .
we show the two fits in fig . 4 .
slope corrected for [ fe / h ] trends plotted against effective temperature .
the solid line is a least squares fit to the swps , and the dashed line is a fit to the comparison stars.,width=316 ] @xcite noted that the signature of accretion of high t@xmath0 material in the sun might be detectable by comparing the solar photospheric abundances to those of the ci carbonaceous chondrite meteorites ( hereafter , chondrites )
. the elemental abundances in chondrites closely track those in the solar photosphere .
indeed , this is often cited as strong evidence for : 1 ) the common origin of the sun and meteorites and 2 ) the primitive nature of the chondrites . however , the chondrites are deficient relative to the solar photosphere in the most volatile elements , h , he , c , n and o. if the sun has accreted rocky material , then its atmosphere should have been enriched in elements with high t@xmath0 values .
the differences between the photospheric and chondritic abundances should display an increasing trend with t@xmath0 . to test for a trend with t@xmath0 in the sun , we calculated the differences between the solar photospheric and chondritic abundances tabulated in two recent studies : table 1 of @xcite and table 1 of @xcite .
we selected elements from these studies with photospheric and chondritic abundance uncertainties less than or equal to 0.15 dex ; the only exceptions to this rule are li , which we excluded since it is known to have been depleted in the sun s envelope via proton capture over its lifetime , and be since it was possibly slightly depleted by the same process .
the uncertainties in the abundance differences were calculated by summing the photospheric and chondritic abundances in quadrature .
the resulting abundance differences are shown in figs . 5 and 6 . for 43 elements from table 1 of @xcite .
the weighted least squares fit is shown as a solid line , and the unweighted fit is shown as a dashed line .
the dotted line corresponds to zero difference.,width=316 ] for 50 elements from table 1 of @xcite .
all else as in fig .
5.,width=316 ] the weighted and unweighted least squares fits shown in fig . 5
give very similar solutions .
weighted least squares gives a slope of @xmath7 dex k@xmath8 , and unweighted least squares gives @xmath9 dex k@xmath8 ; the @xmath10-intercept is about @xmath11 dex .
thus , the slope is significant at about the @xmath12 level .
if we simply compare the number of positive and negative abundance differences for the low and high values of t@xmath0 , we also find support for a positive slope .
for example , the differences are negative for 6 elements , positive for 2 elements and zero for one element for t@xmath13 k ; the differences are negative for 12 elements , positive for 17 elements and zero for 5 elements for t@xmath14 k. applying the same analysis to the data in fig . 6
, we obtain a weighted least squares slope of @xmath15 dex k@xmath8 ; unweighted least squares gives @xmath16 dex k@xmath8 ; the @xmath10-intercept is about @xmath17 dex for the weighted fit and @xmath18 dex for the unweighted fit .
it is obvious from fig .
6 why the weighted solution is different from the unweighted one ; three elements ( rb , w and au ) deviate markedly from zero difference , and they have large uncertainties .
none of these elements were included in the @xcite study , because their photospheric uncertainties were too large .
thus , the weighted least squares solution is preferred in this case . for the @xcite data , the differences are negative for 8 elements and positive for 4 elements for t@xmath13 k
; the differences are negative for 6 elements , positive for 27 elements and zero for 5 elements for t@xmath14 k.
elemental abundances have been found to correlate with t@xmath0 in the atmospheres of a few types of stars .
@xcite showed that elemental abundances correlate with t@xmath0 in the atmospheres of many rv tauri variables ; @xcite showed that the type ii cepheid st pup shows a similar trend . @xcite and @xcite argued that the abundance pattern in the a star j37 in ngc 6633 is best explained by accretion of material with high t@xmath0 elements .
our analysis of the swps t@xmath0 slopes above did not reveal significant differences relative to the comparison stars , either as a general trend or as individual outliers .
two of the stars originally flagged by smith et al . as having high abundance - t@xmath0 slopes , hd 52265 and hd 217107 ,
were found to be normal .
our examination of the solar system abundances revealed a trend with t@xmath0 in two recent datasets .
the simplest interpretation of this trend , assuming it is real , is that the sun s photosphere has been enriched with high t@xmath0 elements . the quality and quantity of the data plotted in figs .
5 and 6 are not sufficient to determine if the linear trend can be extrapolated below t@xmath19 k. if it can be extrapolated to the most volatile elements , then we estimate that elements with the highest t@xmath0 were enriched in the sun s photosphere by about 0.10 dex ; most elements would be enriched by about 0.07 dex .
if , instead , the abundance differences between chondrites and the solar photosphere are constant for t@xmath13 k , then this implies that high t@xmath0 elements have been enriched in the sun s photosphere by about 0.07 dex .
thus , the sun s interior metallicity would be smaller than the surface metallicity by about 0.07 dex .
the c , n and o abundances in the sun s photosphere should be representative of the primordial solar system abundances . while the simplest interpretation of the trends in figs .
5 and 6 is that the sun s photosphere has been enriched in high t@xmath0 elements , it is also possible that elements with t@xmath0 near 600 k have been depleted in the chondrites . certainly , for elements with t@xmath20 k ( e.g. c , n and o ) the depletion in chondrites is very large .
is it safe to assume that elements with @xmath21 t@xmath22 k are not depleted in chondrites to within the measurement errors ? to answer this , we need to determine if elements with t@xmath0 just below 600 k are depleted in chondrites .
@xcite ( footnote 15 ) noted that the photospheric abundance of in is 0.74 dex larger than the chondritic abundance and cited this as possible evidence for large depletion of in in the chondrites . at the time , however , the published t@xmath0 estimate for in was 470 k and that of cd was 430 k , but the cd chondritic and photospheric abundances are similar .
if in is depleted in chondrites , then cd should be as well , given its lower value of t@xmath0 .
this discrepancy seemed to disprove the depletion hypothesis for in and leave its high photospheric abundance unexplained .
k. in is represented by the point at 536 k.,width=316 ] the new t@xmath0 values published by @xcite appear to resolve this discrepancy .
they are 536 k for in and 652 k for cd .
@xcite lists only a few other elements with t@xmath0 values between 200 and 600 k , and only in has a reliable photospheric abundance determination in this range .
br , hg and tl have t@xmath0 values of 546 , 252 and 532 k , respectively . of these , only tl has a photospheric abundance determination .
@xcite quote a photospheric tl abundance range of 0.72 to 1.1 from a sunspot spectrum ; this compares to a chondritic abundance of 0.81 .
we looked for the two tl i features at 3775 and 5350 in the kurucz solar flux atlas @xcite , but we could only set an upper limit on the tl abundance near 1.2
. it would be worthwhile to revisit the solar tl abundance with a new umbral spectrum .
we plot in fig .
7 all the elements from fig . 5 as well as those elements with t@xmath0 values @xmath23 k with reliable photospheric abundances
. it would be helpful if the abundance differences below t@xmath19 k displayed a simple trend with t@xmath0 .
this would allow us to estimate the amount of depletion in chondrites for elements with higher t@xmath0 .
unfortunately , the few elements in fig . 7 with t@xmath0 values below 600 k do not display a simple trend , so we can not easily extrapolate to values of t@xmath0 just above 600 k. nevertheless , given that in is significantly depleted in chondrites , it seems likely that elements with t@xmath0 just above 600 k are depleted in by at least a few hundredths of a dex .
previous searches for anomalous correlations between photospheric elemental abundances and t@xmath0 values among swps have suffered from either small or heterogeneous samples .
we have revisited this topic using a homogeneous set of published abundance data .
we fail to find any significant anomalies with respect to t@xmath0 among a sample of 64 swps and 33 comparison stars .
nevertheless , we encourage others to revisit this topic as more swps are discovered and subjected to detailed spectroscopic analyses .
the signature of accretion may reveal itself either as a general trend of abundance - t@xmath0 slopes with t@xmath1 or as anomalously high abundance - t@xmath0 slopes in a few hotter stars .
it would also be helpful to extend the comparison stars sample above [ fe / h ] @xmath24 .
relative to the chondritic abundances , the solar photospheric abundances do display a weak trend with t@xmath0 .
it is seen in two recent compilations of solar systems abundances .
the implications of this are far ranging , and we encourage additional work on improving photospheric abundance determinations .
i thank the anonymous referee for very useful comments that improved the paper .
ashwell j. f , jeffries r.d .
, smalley b. , deliyannis c.p . ,
steinhauer a. , king j.r . , 2005 , mnras , 363 , l81 asplund m. , grevesse n. , sauval a. j. , 2005 , in bash f. n. , barnes , t. g. , eds , asp conference series , cosmic abundances as records of stellar evolution and nucleosynthesis , in press ( astro - ph/0410214v2 ) beirao p. , santos n.c . , israelian g. , mayor m.,2005 , a&a , 438 , 251 bodaghee a. , santos n.c . , israelian g. , mayor m. , 2003 , a&a , 404 , 715 ecuvillon a. , israelian g. , santos n. c. , et al . , 2004 , a&a , 426 , 619 fischer d. a. , valenti j. , 2005 , apj , 622 , 1102 giridhar s. , lambert d. l. , reddy b.e . , gonzalez g. , yong d. , 2005 , apj , 627 , 432 gonzalez g. , 1997 , mnras , 285 , 403 gonzalez g. , 2003 , rev mod phys , 75 , 101 gonzalez g. , 2005 , in preparation gonzalez g. , laws c. , 2000 , aj , 119 , 390 gonzalez g. , laws c. , tyagi s. , reddy b. e. , 2001 , aj , 121 , 432 gonzalez g. , wallerstein g. , 1996 , mnras , 280 , 515 gratton r.g .
, bonanno g. , claudi r.u . , cosentino r. , desidera s. , lucatello s. , scuderi s. , 2001 , a&a , 377 , 123 kurucz r. l , furenlid i. , brault j. , testerman l. , 1984 , solar flux atlas from 296 to 1300 nm , national solar observatory lambert d. l. , mallia e.a . , smith g. , 1972 , solar physics , 26 , 250 laws c. , gonzalez g. , 2003 , apj , 595 , 1148 laws c. , gonzalez g. , walker k.m . , et al .
, 2003 , aj , 125 , 2664 lodders k. , 2003 , apj , 591 , 1220 sadakane k. , ohkubo y. , takeda y. , sato b. , kambe e. , aoki w. , 2002 , pasj , 54 , 911 sadakane k. , ohkubo y. , honda , s. , 2003 , pasj , 55 , 1005 santos n. c. , israelian g. , mayor m. , 2004 , a&a , 415 , 1153 smith v.v . , cunha k. , lazzaro d. , 2001 , aj , 121 , 3207 takeda y. , sato b. , kambe e. , aoki w. , honda s. , kawanomoto s. , masuda s. , izumiura h. , et al . , 2001 , pasj , 53 , 1211 | results from detailed spectroscopic analyses of stars hosting massive planets are employed to search for trends between abundances and condensation temperatures .
the elements c , s , na , mg , al , ca , sc , ti , v , cr , mn , fe , ni and zn are included in the analysis of 64 stars with planets and 33 comparison stars .
no significant trends are evident in the data .
this null result suggests that accretion of rocky material onto the photospheres of stars with planets is not the primary explanation for their high metallicities .
however , the differences between the solar photospheric and meteoritic abundances do display a weak but significant trend with condensation temperature .
this suggests that the metallicity of the sun s envelope has been enriched relative to its interior by about 0.07 dex .
[ firstpage ] sun : abundances stars : abundances . |
as a result of the discovery of pnictide superconductors by kamihara
_ et al . _ , @xcite in 2008 there has been intense interest in this new family of materials . in the past two years their quality
has greatly improved and large single crystals have become available .
however , the nature of the superconducting state and the corresponding gap structure of the pnictide superconductors is not settled . while the pairing mechanism , related to magnetic fluctuations , and basic structure of the extended @xmath7-wave state are generally agreed upon , @xcite details of possible gap anisotropy are currently actively investigated , often with opposite conclusions .
experiments from angle resolved photoemission spectroscopy ( arpes ) , @xcite andreev - reflection spectroscopy , @xcite and specific heat @xcite measurements support the fully - gapped model . but penetration depth measurements , @xcite nuclear quadrupole resonance ( nqr ) , and specific heat measurements @xcite have been interpreted in terms of nodal gap structure .
these measurements indicate that the superconducting order parameter might be fundamentally different in different classes of these materials , further complicating interpretation of the experiments .
moreover , even in a single compound , but with different amount of doping , various gap signatures have been seen , from fully gapped , to nodal character .
for example , calorimetric,@xcite transport@xcite and optical conductivity@xcite measurements in overdoped ba(fe@xmath11co@xmath12)@xmath2as@xmath2 ( @xmath13 ) show nodal character , and similarly in underdoped samples , @xcite whereas at optimal doping @xmath14 results are contradictorily interpreted either in terms of a more or less isotropic gap@xcite or a strongly anisotropic gap .
@xcite the fact that the gap structure might be doping - dependent is a possibility .
theory predicts that in multi - band materials , such as pnictides , one may naturally find states with `` accidental '' nodes , even in the most symmetric @xmath8 configuration . in most cases the nodes appear on the outer electronic fermi surface sheets , depending on the values of the interaction parameters , which are functions of doping .
@xcite the outstanding question now is what is the form and position of nodal lines on the electron fermi surface .
@xcite other nodal locations ( e.g. on central hole pocket ) have also been considered .
@xcite in this work we report nuclear magnetic resonance ( nmr ) measurements on single crystals of ba(fe@xmath0co@xmath1)@xmath2as@xmath2 .
nmr has advantages in that it can probe bulk characteristics , is sensitive to the electronic structure within a penetration depth of the surface of a single crystal , @xmath15 nm , and is less susceptible to surface conditions as might be the case in stm and arpes experiments . in principle , knight shift data from nmr can be an appropriate indicator of fully gapped or nodal gap structure . for the former ,
the knight shift has the temperature dependence of the yosida function @xcite or normal fluid density , and for a gap with line nodes the temperature dependence should be linear at low temperatures .
@xcite our goal was to investigate the knight shift at sufficiently high magnetic fields that vortex contributions to the temperature dependence of the local field are minimized , and at sufficiently low temperatures to identify the gap structure .
it is also important to determine independently that the vortex structures have a static distribution , that is to say there is a solid vortex state .
then the vortex configuration is stable and can not introduce temperature dependence to the lineshape , affecting the knight shift analysis .
consequently we have measured the spin - spin relaxation time , @xmath9 , as a function of temperature and magnetic field and found it to be a good indicator for the onset of vortex dynamics .
we focus on measurements of the knight shift below the temperature where vortices become frozen .
we have performed @xmath3as nmr studies on a single crystal of ba(fe@xmath11co@xmath16)@xmath2as@xmath2 ( @xmath17 ) with zero field @xmath18 k , varying temperatures from 2 k to 200 k with external magnetic field from @xmath19 t to @xmath10 t parallel to the @xmath20-axis of the sample .
the measurements were performed at northwestern university and the national high magnetic field laboratory in tallahassee , florida .
the crystal was grown at ames laboratory by the self - flux method , @xcite having dimensions of [email protected]@xmath210.7 mm@xmath22 and mass of 49.3 mg . a hahn echo sequence ( @xmath23echo )
was used for spectra , knight shift , linewidth , and some of our spin - spin relaxation time , @xmath24 , measurements , where @xmath25 is the delay time .
a typical @xmath26 pulse length was 4 @xmath27s , defined as that which gave maximum echo intensity .
we used frequency sweeps when a single pulse did not cover a sufficiently wide frequency range .
delay times for acquisition of the spectra were varied from 70 to 200 @xmath27s depending on temperature . the knight shift and
linewidth measurements were determined from gaussian fits to the line shape .
the gyromagnetic ratio of the bare nucleus @xmath3as , @xmath28 7.2919 mhz / t , was used as the reference for the knight shift .
linewidths were defined as full - width - at - half - maximum ( fwhm ) of the gaussian distribution . for @xmath24 measurements from the hahn echo sequence , we varied @xmath25 from 100 @xmath27s to 1.40 ms , and determined the rate from the initial portion of the recovery .
in high magnetic fields , @xmath29 t in a resistive magnet , we used the carr - purcell - meiboom - gill ( cpmg ) sequence which is resistant to magnetic field fluctuations , @xcite ( @xmath30echo@xmath31 and @xmath32 indicate orthogonal rf phases ) .
typically , the refocusing time , @xmath33 , was chosen to be 250 @xmath27s .
in the course of these spin - spin relaxation measurements our comparison of cpmg and hahn echo experiments revealed the existence of strong field fluctuations intrinsic to the sample .
in all cases , the pulse sequence repetition time in the superconducting state was taken to be of order the spin - lattice relaxation time , @xcite @xmath34 , and was increased with decreasing temperature .
our field swept spectra of @xmath3as nmr and the shift of their peak frequencies with temperature are shown in fig .
[ fig1 ] . from the satellite transition we find the quadrupolar coupling , @xmath35 mhz ( 0.35 t ) , similar to the value reported by ning _
et al . _
@xcite the quadrupolar satellite transitions are suppressed because of disorder in the electric field gradient ( efg ) tensor , most likely from co doping , from which the central transition is largely immune . in part
a ) of this figure there is a small component of the spectrum having a negligible knight shift ( 0.003% ) .
it is clear that these nuclei are from impurity or defect sites having a different orbital electronic shift .
the spectra in the lower panel are obtained from the central transition .
similar observations were made by ning _
et al . _
@xcite our measurements of the temperature variation of the total knight shift , @xmath36 , of the @xmath3as central transition are presented in fig .
spin - singlet pairing is evident from the sharp decrease of @xmath37 at @xmath38 .
the onset of this decrease at 18 k in 11.6 t coincides with the @xmath38 measured from the onset of the drop of the resistance reported by ni _
et al . _
@xcite in the same magnetic field .
this is in contrast to the cuprates , especially the underdoped cuprates , where the knight shift starts to decrease well above @xmath38 , indicating the presence of a pseudogap . at low temperatures in the superconducting state ,
we observe that the knight shift decreases , almost linearly with temperature .
we express the total knight shift as @xmath39 .
the spin part of the shift is @xmath40 , where @xmath41 is the spin susceptibility , and @xmath42 is the hyperfine field . both the orbital , @xmath43 , and the quadrupolar shift , @xmath44 , are expected to be temperature independent .
additional magnetic shifts which we call @xmath45 could be associated with impurities , @xcite and in principle , could be temperature dependent , although this is not seen in cuprates , @xcite nor do we find any evidence for this in the present work . in the normal state of ba(fe@xmath0co@xmath1)@xmath2as@xmath2 below 170 k , we find the knight shift can be described by an activated process
. however , there is a substantial range at low temperatures where the knight shift is essentially temperature independent , fig .
we use an arrhenius form , @xmath46 , where we find @xmath47 k , @xmath48 , and @xmath49 with the fit shown by the ( red ) curve in the figure . at high temperatures , 90 170 k ,
the total knight shift is proportional to the bulk susceptibility , @xcite giving a hyperfine field of @xmath50 koe/@xmath51 , comparable to the value of 18.8 koe/@xmath51 reported by kitagawa _
et al . _
@xcite therefore the thermally activated process modifies the local density of states and directly affects the spin susceptibility .
however , neither the origin of this high - temperature behavior , nor its relation to superconductivity is known .
the first observations of this effect , along with a similar analysis , was performed by ning _
_ @xcite giving @xmath52 k for their 8 % co crystal of ba(fe@xmath53co@xmath54)@xmath2as@xmath2 . in the superconducting state
, we found that the total knight shift gradually decreases on cooling from 12 k to 2 k in 11.6 t , figures 2 and 3 .
similar behavior is consistently observed for higher magnetic fields up to 20 t , fig .
if one takes the view that this is a linear temperature dependence , then such behavior can be associated with gap nodes at the fermi surface , similar to cuprate superconductors with @xmath55-wave symmetry .
@xcite the knight shift @xmath56 is proportional to the susceptibility @xmath57 where @xmath58 - normal state susceptibility , @xmath59 is the fermi - dirac distribution function , and @xmath60 is the density of states . for temperatures @xmath61 @xmath62/[2\mu h ] \approx - \partial f(e)/\partial e$ ] and for nodal quasiparticles
@xmath63 , resulting in a linear temperature dependence @xmath64 .
this is a much steeper rise than what we observe .
thus , we can exclude the possibility of nodes simultaneously present on both hole and electron sheets .
however , nodes might be present on one of the pockets , presumably electronic , which in combination with the fully gapped hole pocket would give a slower @xmath65-dependence .
another possibility is that the observed temperature dependence is higher power than linear , and the electronic superconducting gap in ba(fe@xmath0co@xmath1)@xmath2as@xmath2 is either anisotropic with weak `` accidental '' nodes , or has only a minimum with typical energy scale @xmath66 , or even isotropically gapped , with strong pairbreaking due to impurities .
@xcite in this regard , the existence of impurities affecting the local magnetic fields is evident in the temperature dependence of the nmr linewidth in the normal state , fig .
comparison of our knight shift measurements for @xmath67 t with the above model and different gap structures are shown in fig .
the spin susceptibility is calculated according to eq .
[ eq : chi ] , where the shift of the energy levels in magnetic field is given by parameter @xmath68 ( @xmath69 t ) , where @xmath51 is the bohr magneton , and @xmath70 k is an estimate of the zero - field transition temperature in the absense of impurities .
the density of states is computed in the two - band model used previously and described in references 30 and 46 .
the gap on the hole fs is isotropic , @xmath71 , while on the electron fs it is a function of the angle @xmath72 , with @xmath73 being the isotropic gap , and @xmath74 gives four equally spaced nodes .
the impurity scattering is given by the concentration parameter @xmath75 , _ interband _ potential fraction ( @xmath76 ) , and an _ intraband _ scattering phase shift ( @xmath77 ) .
the zero - temperature shift @xmath78 is not well - defined experimentally and is left as a free parameter . below @xmath79
the temperature dependence is stronger and it is plausible that this can be associated with diamagnetic screening currents or vortex supercurrents which do not cancel as effectively in the intervortex region at low fields . at high temperature ( @xmath80 ) , the discrepancy between our data and the model is likely due to thermal fluctuations of vortices that introduce a temperature dependence to the nmr lineshape in a region where vortices are being pinned .
it is just in this region of temperature that slow field fluctuations from vortices are manifest in the linewidth , fig . 5 , and in spin - spin relaxation , discussed in the next section , fig
we observe a systematic trend in the data of increasing knight shift , @xmath81 , with increasing field , fig . 4 , which might be accounted for by considering the vortex core contributions to the density of states . from fig
. 3 , in the temperature range @xmath82 where we can compare the knight shift data with the theoretical model
, we find that more or less all electronic gap structures that we have considered in the @xmath8 symmetry class can fit the data .
one might say that the intermediate cases of anisotropic gaps with either minima @xmath83 or close nodes @xmath84 are less suitable , especially in the strong scattering case , leaving surprisingly a possibility of either an isotropically gapped electronic fs or one having a gap @xmath85 . a more reliable identification of the angular structure could be possible in lower magnetic fields ( @xmath86 inset in fig.3(a ) ) , but the analysis would be quite complicated due to the effects of vortices and diamagnetic screening currents .
different anisotropy parameters @xmath87 in this model give a reasonable coverage of the phase space of possibilities for the spectrum of quasiparticles .
we checked that with a more realistic three dimensional fermi surface and the gap suggested in ref .
results do not change much ( dashed line in the inset of fig.3(a ) that appear exactly between the isotropically gapped @xmath73 and purely nodal @xmath74 cases ) .
the linewidth of the central transition , fig .
[ fig5 ] , full - width - at - half - maximum ( fwhm ) , was determined by fitting the spectra to a gaussian function .
the fit deviates slightly near the wings but at low temperatures the lineshape appears to be symmetric . the linewidth increases with decreasing temperature in the normal state .
below @xmath38 , there is a decrease , followed by an upturn at @xmath88 11 k in 11.6 t. these features closely parallel all of the aspects of the nmr linewidth previously reported by chen _
et al . _
@xcite for @xmath89o nmr on high quality single crystals of bi@xmath90sr@xmath90cacu@xmath90o@xmath91 ( bscco ) . for cuprate superconductors@xcite this similar behavior in the normal state was identified with impurities .
its classic signature is a curie or curie - weiss temperature dependence to the linewidth and a constant knight shift .
the basic argument is that impurities introduce strong polarization of the nearby conduction electrons which produce a spatially oscillating spin density that couples through the hyperfine interaction to the nuclei ( rkky - interaction ) .
consequently , the average local field is not significantly perturbed , so the knight shift is unaffected ; but , a large distribution of knight shifts gives a significant broadening of the nmr spectrum .
the polarization of paramagnetic impurities is typically quite temperature dependent ( curie - like ) , and is the likely origin for the temperature dependence in the linewidth below @xmath92 k. at higher temperatures there is a possible additional temperature dependence from the activated process that is evident in the knight shift shown in fig . 2 .
the success we have had with the above model in understanding the @xmath89o nmr linewidth in bscco , @xcite serves as a guide for our interpretation of the impurity contributions to the @xmath3as nmr linewidth in ba(fe@xmath0co@xmath1)@xmath2as@xmath2 .
we focus on the range of temperature where the knight shift is mainly temperature independent , @xmath93 k. we fit the linewidth to the following phenomenological relation , @xmath94 since there may also be a distribution of knight shifts with a thermally - activated origin , we allow for this contribution with the additional term , @xmath95 , which goes beyond the impurity model used for bscco.@xcite here , @xmath96 k , obtained from a fit to @xmath37 in the normal state with @xmath97 with parameters , @xmath98 , and @xmath99 .
overall , the fit is reasonable , capturing the essential characteristics of the normal state behavior of the linewidth with a reduction in the superconducting state as the impurity local field distribution becomes gapped following the temperature dependence of the knight shift . although impurity effects can account for the temperature dependent linewidth , we can not comment on what type of impurity might be responsible .
the upturn of the linewidth at @xmath100 k in 11.6 t in the superconducting state is the mark of vortex freezing .
it occurs at @xmath101 in @xmath102 t , and indicates that there is a substantial temperature range below @xmath38 where fluctuating vortices are in a liquid - like phase .
we can compare this with yba@xmath90cu@xmath103o@xmath104 ( ybco ) @xcite and overdoped bscco @xcite where vortex melting for the same magnetic field occurs at @xmath105 and @xmath106 respectively .
calculations of the melting transition in high magnetic fields for pnictide superconductors have been performed by murray and tesanovic @xcite . however , a true , thermodynamic , vortex - liquid phase in ba(fe@xmath0co@xmath1)@xmath2as@xmath2 would be surprising since both the mass anisotropy@xcite and @xmath38 are small ; for example , as compared to cuprate superconductors .
it was argued by yamamoto _
et al . _
@xcite that a confluence of weak thermal fluctuations and a very broad distribution of pinning forces gives rise to an irreversibility field significantly less than the upper critical field in crystals of ba(fe@xmath107co@xmath108)@xmath2as@xmath2 .
generally the nmr spectrum is sensitive to field fluctuations on time scales slower than the larmor precession period which is tens of nanoseconds in our case .
consequently , the sharp variation in the linewidth that we observe at @xmath109 suggests that vortex dynamics are slower than this for @xmath110 .
to investigate this further , we have performed spin - spin relaxation experiments discussed in the next section .
usually , the vortex contribution to the nmr or @xmath27sr linewidth@xcite is interpreted in terms of the penetration depth . in ba(fe@xmath0co@xmath1)@xmath2as@xmath2
the linewidth broadening from the vortex solid state can be seen in fig .
[ fig5 ] as a difference between data ( represented by a black dashed line ) and our model for the vortex free linewidth ( solid blue curve ) discussed above .
the vortex component of our linewidth at the lowest temperatures , @xmath111 khz ( @xmath112 gauss ) is much less than predicted by ginzburg - landau theory @xcite which is 33 khz for a penetration depth from @xmath27sr of @xmath113 nm , @xcite or 15 khz for the value , @xmath114 nm , from scanning squid and magnetic force
measurements.@xcite the most likely reason for this discrepancy is c - axis vortex disorder discussed by brandt .
@xcite for example , at very low magnetic fields , @xmath115 t , this plane - to - plane disorder is well - established in bscco from @xmath27sr and nmr measurements @xcite and can be expected in ba(fe@xmath0co@xmath1)@xmath2as@xmath2 since in - plane disorder and strong pinning has already been established by scanning probe measurements @xcite and small angle neutron scattering .
to explore possible effects of vortex dynamics , already indicated in our nmr linewidth measurements , we have investigated spin - spin relaxation .
there are two classes of spin relaxation experiments that we perform .
the results measured by hahn echo methods are presented in fig . 6 and 7 , and by the cpmg pulse sequence in fig . 6 and 8 . the nuclear spin dephasing times , @xmath116 and @xmath117 , from these experiments can be viewed semi - classically as a measure of field fluctuations along the direction of the applied magnetic field .
the dephasing rate is a summation of contributions from spin - lattice relaxation ( redfield contribution ) , local fluctuations in the external magnetic field , and fluctuations in the internal fields such as those from the nuclear dipole - dipole interaction , or from vortex dynamics .
the hahn echo method is generally useful for homogeneous relaxation ( all nuclei are equivalent ) in a steady external field .
however , if the local field fluctuates inhomogeneously , or there is nuclear spin diffusion in a spatially inhomogeneous steady field,@xcite then the cpmg method can be helpful in identifying the time scale of the fluctuations through variation of the spin refocusing time , @xmath25 ( see the experimental methods section ) .
@xcite from our relaxation measurement data on ba(fe@xmath0co@xmath1)@xmath2as@xmath2 shown in fig . 6 , it is immediately apparent that @xmath117 is significantly longer than @xmath116 and that in the normal state , both are much longer than the temperature independent dipolar contribution , @xmath118 ms , calculated from the rigid lattice limit , shown as a red dashed line .
for all cpmg experiments shown here @xmath33 was set to be relatively short , @xmath119s to @xmath120s .
we infer that significant dephasing takes place on time scales longer than @xmath120s and , since this is in the normal state , the process responsible has nothing to do with vortices .
rather it is an indication of the presence of slow field fluctuations from an unidentified source .
in fact , by varying the refocusing time , @xmath33 , we have determined that this dephasing is effectively eliminated in the cpmg results which we report here .
we defer our discussion of the mechanism for the hahn echo relaxation to later in this section .
the temperature dependence of @xmath117 in the normal state can be understood from the temperature dependence of spin - lattice relaxation , through the redfield contribution .
@xcite this behavior can be expressed phenomenologically as , @xmath121 .
@xcite taking @xmath34 from ning _
et al . _
@xcite we find @xmath122 = 5.7 and @xmath123 ms , where @xmath9 is the spin - spin relaxation time after correction for this contribution . since @xmath124 is long enough below @xmath38 ( @xmath125 150 ms ) we may take @xmath117 to be equal to @xmath9 in the superconducting state .
nonetheless , we find that there is a remarkable dip in both @xmath117 and in @xmath116 at @xmath109 , fig . 6 to 8 , which we identify with thermal fluctuations of vortices .
this follows from the fact that the temperature of the minimum in @xmath9 coincides with the vortex freezing temperature apparent in the nmr linewidth , presented in fig . 5 for 11.6 t and that this dip at @xmath126 appears just below the upper critical field , fig .
vortex dynamics contribute similarly in these two types of spin - spin relaxation experiments although the contribution to the hahn echo can be an order of magnitude larger than for the cpmg , especially at higher magnetic fields .
this means that the principal fluctuation time scale must be of order @xmath116 , but longer than the refocusing time , @xmath127s , in the cpmg measurement . in previous work on @xmath89o nmr in ybco
it was found that the spin - spin relaxation rate , @xmath128 , had a peak below @xmath38 .
@xcite krmer _
et al . _
@xcite proposed a dynamical charge - density - wave state in yba@xmath2cu@xmath129o@xmath130 coupled to the nucleus through the quadrupolar interaction .
however , bachman _
et al . _
@xcite found that this peak was an artifice of vortex dynamics which become slow enough in the vortex solid to be observable in @xmath9 .
they found a lorentzian component to the spectral density for field fluctuations that abruptly onsets at the irreversibility temperature .
the fact that the onset temperature was magnetic field dependent indicated that the phenomenon was not related to a charge density wave .
this argument holds equally well in the present case for ba(fe@xmath0co@xmath1)@xmath2as@xmath2 . here , as for the cuprates , we associate the temperature of the maximum spin - spin relaxation rate ( the dip in the relaxation time ) with the temperature where the vortex dynamics rapidly slow down , _
i.e. _ the irreversibility temperature .
for comparison , in fig .
9 we plot , along with our data , the irreversibility curve from prozorov _ et al . _
@xcite obtained at lower magnetic fields for similar crystals from the ames laboratory .
our data includes the temperatures of anomalies from the nmr linewidth ( spectrum ) and @xmath9 measurements we have discussed previously .
the cpmg sequence was used to obtain @xmath9 at the higher magnetic fields , 20 t to 28 t because the hahn echo experiment is adversely affected by magnetic field fluctuations endemic to high field resistive magnets .
@xcite we also compare our measurements of the superconducting transition with those of ni _ et al . _
@xcite providing a consistent vortex phase diagram for ba(fe@xmath0co@xmath1)@xmath2as@xmath2 .
now we return to a discussion of the large @xmath9 we have observed , as compared with the expected temperature independent upper bound established by the nuclear dipole - dipole interaction , the red dashed line in fig .
6 . in order for @xmath9 to be longer than the dipole limit
there must be some form of coherent averaging which in liquids is the well - known phenomenon of motional averaging . in metals
there is a similar mechanism , a type of spin - motional averaging , which is attributable to a strong indirect interaction between adjacent nuclei , @xmath3as in our case .
@xcite for this process to work the coupled nuclei must be resonant , that is to say they must have the same larmor frequency to within the small band given by the bare nuclear dipole - dipole interaction .
this requirement is imposed by the necessity to conserve energy in the spin ` flip - flop ' process for spin - spin relaxation .
the indirect interaction is a coupling between nuclei , mediated by the conduction electrons , involving an rkky type polarization and the hyperfine interaction . for metallic elements with only one spin isotope
this leads to averaging of the dipolar local fields , called exchange narrowing , and can substantially increase @xmath9 , a factor of 5.6 in the case of platinum .
@xcite for elements with more than one spin isotope the nuclei are non - resonant and the interaction leads to @xmath9 shorter than the dipolar limit , as is the case for thallium . from the theory of exchange narrowing
@xcite we have , @xmath131 where @xmath132 is the frequency of the exchange interaction and @xmath133 is the bare dipolar interaction between nuclei . from this relation
we find , @xmath134 khz , which is larger than the 4 khz , @xcite observed for metallic platinum , but not unreasonably large given that @xmath3as is 100 % naturally abundant as compared to 33.8 % for @xmath135pt .
although this explanation can account for the longer @xmath24 than expected from the dipolar interaction , it does not explain the shorter phase coherence time found from hahn echo measurements as compared with the cpmg sequence .
there are two possibilities for the source of low frequency field fluctuations that could give rise to the difference between hahn and cpmg results . in the first
, we consider fluctuating magnetic fields , either from impurities or residual effects of the suppressed antiferromagnetism inherent to the bafe@xmath2as@xmath2 system .
however , it is hard to reconcile this explanation with our observation that @xmath116 is independent of magnetic field and temperature , as is evident in fig .
7 after allowing for the vortex contribution .
one might expect that thermally driven , magnetic field fluctuations would be both temperature and magnetic field dependent over the wide range we have covered .
a second possibility is that the rather large indirect interaction leads to nuclear spin diffusion that produces dephasing in a locally inhomogeneous , static , magnetic field .
we have estimated the diffusion coefficient from the indirect interaction@xcite to be @xmath136 @xmath137/s which , in order to explain our hahn echo results , would require an average magnetic field gradient of @xmath138 1.5 @xmath21 10@xmath139 g / cm .
associating a magnetic field inhomogeneity of this size with residual magnetism in the sample could provide the temperature independent mechanism that would be required to account for the hahn echo results .
the length scale for the distribution of such magnetism is constrained by the linewidth to be less than @xmath140 nm .
although it may seem that the latter explanation is the more likely , further work will be necessary to confirm or disprove either of these suggestions .
we have studied @xmath3as nmr spectra and spin - spin relaxation on a single crystal of optimally doped ba(fe@xmath0co@xmath1)@xmath2as@xmath2 from room temperature to well below the superconducting transition temperature in magnetic fields as high as 28 t. the temperature dependence of the knight shift is compatible with a general , two - band @xmath8 symmetry scenario for the superconducting states .
the details , nodes or full gap , can not be determined from the present work .
comparison with an elementary 2d model might indicate that the order parameter in this compound has nodes but they have relatively weak weight compared to the fully gapped portions of the fermi surface .
we have identified a sharp signature of irreversibility in the temperature dependence of our measured spin - spin relaxation from which we have established a vortex phase diagram up to high magnetic fields .
we have found that there is a very strong indirect exchange interaction in this compound providing a possible explanation for our observations of spin - spin relaxation attributed to nuclear spin diffusion in static magnetic field gradients .
we thank c.a .
collett , w.j .
gannon , j. li , and j. pollanen for critical reading of the manuscript and acknowledge support from the department of energy , basic energy sciences under contracts no .
de - fg02 - 05er46248 ( northwestern university ) and no .
de - ac02 - 07ch11358 ( ames laboratory ) .
work at high magnetic field was performed at the national high magnetic field laboratory with support from the national science foundation and the state of florida . a.b.v .
acknowledges support from nsf grant dmr-0954342 . 99 natexlab#1#1bibnamefont # 1#1bibfnamefont # 1#1citenamefont # 1#1url # 1`#1`urlprefix[2]#2 [ 2][]#2 , , , , * * , ( ) . , * * , ( ) . , , , , * * , ( ) . , , , * * , ( ) .
, , , , , * * , ( ) . , , , , , , , , , , , , , , , * * , ( ) . , , , , , , , , , , , * * , ( ) . , , , , , , , , , , , , , * * , ( ) . , , , , , , * * , ( ) . , , , , , , * * , ( ) . , , , , , , , , , , , , , , ,
* * , ( ) . , , , , , , , , * * , ( ) . , , , , , , , , * * , ( ) .
, , , , , , , , * * , ( ) . , , , , , , , , , , , , , , , * * , ( ) . ,
, , , , , , , , , , ( ) , arxiv.org:1011.4808v1 .
, , , , , , , , ( ) , arxiv.org:1009.1091v1 . , , , , , , , , * * , ( ) . , , , , , , , , , ,
* * , ( ) . , , , , , , , , , * * , ( ) .
, , , , , , , , , ( ) , arxiv.org:1012.3436v1 . , , , , , , , ,
, ( ) , arxiv.org:1003,2959v2 . , , , , , , , * * , ( ) . , , , , , , , , * * , ( ) .
, , , , , , , * * , ( ) . , , , * * , ( ) . , , , ,
* * , ( ) . , , , * * , ( ) .
, , , , , * * , ( ) .
* * , ( ) . , * * , ( ) . , , , , , , * * , ( ) . , , , , * * , ( ) . , , ,
* * , ( ) . ,
* * , ( ) .
, , , , , * * , ( ) . , , , , , , , * * , ( ) . , , , , , , , , * * , ( ) . , , , , , , , , * * , ( ) . , , , , , , , , , , * * , ( ) . , , , , * * , ( ) . , , , , , * * , ( ) . , , , , , * * , ( ) . , , , , , , , * * , ( ) . , , , , ,
* * , ( ) . , , , , * * , ( ) .
, , , , , , , * * , ( ) . , , , , , ,
* * , ( ) . , , * * , ( ) . , , , , , , , , , , , , , * * , ( ) . ,
, f , * * , ( ) . , * * , ( ) .
, , , , , , , , , , , , , , , , , * * , ( ) . , , , , , , , , , , * * , ( ) . , * * , ( ) . , * * , ( ) . , , , , , , , , , , doi:10.1038/nphys1835 . , , , , , , , , , ( ) , arxiv.org:1011.1520v1 . , , , , , , , * * , ( ) . , , , , , , , , , , , ,
* * , ( ) . , , * * , ( ) .
i. yu , a.a.v .
gibson , e.r .
hunt , and w.p .
halperin , phys .
* 44 * , 348 ( 1980 ) ; i. yu and w.p .
halperin , phys .
b * 47 * , 15830 ( 1993 ) . , * * , ( ) . , * * , ( ) . , , , , , , , * * , ( ) .
, , , , , * * , ( ) . , , , * * , ( ) .
, , , , * * , ( ) . , * * , ( ) . , , , , , , , , , , , * * , ( ) . ,
* * , ( ) . , * * , ( ) . , , * * , ( ) .
as nmr spectra of ba(fe@xmath0co@xmath1)@xmath2as@xmath2 .
a ) single crystal spectra were obtained from a field sweep at 84.88 mhz . from the quadrupolar satellite transitions , shown magnified for clarity ,
@xmath141 was found to be 2.58 mhz ( 0.35 t ) .
b ) spectra of the central transition are shown at different temperatures , normalized in area.,scaledwidth=35.0% ] co@xmath1)@xmath2as@xmath2 .
the temperature dependence of the knight shift consists of three parts : at low temperature ( below the vortex freezing temperature ) , near the superconducting transition ( the arrow indicates @xmath38 ) , and in the normal state . the drop in knight shift at the superconducting transition
is evidence for a spin - singlet state . in the normal state
the knight shift follows an activated behavior , with activation energy of 394 k. the solid curve ( red ) is the fit to this activated process .
the low temperature behavior shows a decreasing knight shift from which we find @xmath142 by extrapolating to zero temperature .
, scaledwidth=40.0% ] co@xmath1)@xmath2as@xmath2 .
the theoretical curves ( solid lines ) are shown for several possible gap structures on the electronic pocket , given by the form @xmath72 , for intermediate ( panel a ) and strong scattering ( panel b ) . to compare these results with the experiment we took @xmath143 ( @xmath69 t ) .
inset in panel ( a ) shows results for the clean case and at low fields ( @xmath86 ) - note that aside from the low-@xmath65 region all anisotropic gaps , @xmath144 , and a model with more realistic 3d fermi surface , behave in approximately the same way . in high field and for weak impurity scattering ( a ) ,
all gap models are consistent with experiment at low @xmath65 . in the unitary limit ( b ) , the best fit is given by either combination of two isotropic gaps @xmath145 ( @xmath73 ) , or , surprisingly , by an isotropic hole @xmath146 and d - wave like electron gaps @xmath147 ( @xmath74 ) .
it is possible that the dip in the data , @xmath148 , is a consequence of changes in the nmr lineshape owing to vortex , section iv . , scaledwidth=49.0% ] as central transition at @xmath102 t. the upturn of the linewidth at @xmath149 11 k , indicates freezing of the vortex liquid@xcite ( the inset shows an expanded view of this transition ) .
the solid blue curve , down to @xmath109 , is a fit to the model described in the text .
below @xmath109 the solid blue curve is a calculation from this model , absent any vortex contributions to the linewidth .
the dashed black line is a guide - to - the - eye representing the data .
the difference at low temperatures between the solid and dashed curves in the inset is the vortex contribution to the linewidth which we find to be @xmath150 khz .
, scaledwidth=40.0% ] t. measurements of spin - spin relaxation times with hahn echo and cpmg methods are shown for comparison .
a superconducting magnet was used to rule out @xmath9 reduction due to field fluctuation .
a ) the slow increase with decreasing temperature down to @xmath38 corresponds to reduction in the redfield contribution , _
i.e. _ from spin - lattice relaxation .
b ) an expanded view at low temperature below @xmath38 , shows that there is a decrease in both relaxation times .
we associate this with vortex dynamics providing a maximum contribution to the rates at @xmath109 , taken from fig .
the red dashed line indicates @xmath9 in the dipolar limit , 0.7 ms.,scaledwidth=40.0% ] , for various magnetic fields from a superconducting magnet .
there is a clearly defined temperature , @xmath109 , for a minimum in @xmath9 at each magnetic field which we associate with vortex freezing .
the minimum @xmath9 becomes even smaller than the @xmath9 from the dipolar limit ( 0.7 ms ) when the magnetic field increases .
additionally , @xmath116 shows a decrease in the normal state just above @xmath38 at @xmath151 t.,scaledwidth=40.0% ] , for various magnetic fields .
the transition temperature@xcite ( black upward arrow ) and the vortex freezing temperature ( red downward arrow ) correspond to abrupt changes in relaxation .
the dashed black arrows indicate the minimum relaxation time from the hahn echo method.,scaledwidth=40.0% ] | the superconducting state of an optimally doped single crystal of ba(fe@xmath0co@xmath1)@xmath2as@xmath2 was investigated by @xmath3as nmr in high magnetic fields from 6.4 t to 28 t. it was found that the knight shift is least affected by vortex supercurrents in high magnetic fields , @xmath4 t , revealing slow , possibly higher order than linear , increase with temperature at @xmath5 , with @xmath6 .
this is consistent with the extended @xmath7-wave state with @xmath8 symmetry but the precise details of the gap structure are harder to resolve .
measurements of the nmr spin - spin relaxation time , @xmath9 , indicate a strong indirect exchange interaction at all temperatures . below the superconducting transition temperature abrupt changes in vortex dynamics lead to an anomalous dip in @xmath9 consistent with vortex freezing from which we obtain the vortex phase diagram up to @xmath10 t. |
denote by @xmath6 the orientable compact real surface of genus @xmath1 . a * projective structure * on @xmath6
is given by an atlas @xmath7 of coordinate charts ( local homeomorphisms ) @xmath8 such that the transition functions @xmath9 are restrictions of moebius transformations @xmath10 .
there is a unique maximal atlas defining the projective structure above , obtained from the previous one by adding all charts @xmath11 when @xmath12 runs over @xmath13 .
a projective structure induces a * complex structure * on @xmath6 , just by pulling - back that of @xmath3 by the projective charts .
we will denote by @xmath0 the corresponding riemann surface ( complex curve ) .
[ e : universalcovering](*the universal cover * ) let @xmath0 be a compact riemann surface having genus @xmath1 and consider its universal cover @xmath14 . by the riemann mapping theorem , we can assume that @xmath15 is either the riemann sphere @xmath3 , or the complex plane @xmath16 or the unit disk @xmath17 depending wether @xmath18 , @xmath19 or @xmath20 .
we inherit a representation of the fundamental group @xmath21 whose image @xmath22 is actually a subgroup of @xmath13 . all along the paper , by abuse of notation
, we will identify elements @xmath23 with their image @xmath24 .
the atlas defined on @xmath0 by all local determinations of @xmath25 defines a projective structure on @xmath0 compatible with the complex one .
indeed , any two determinations of @xmath26 differ by left composition with an element of @xmath22 .
we thus see that any complex structure on @xmath6 is subjacent to a projective one .
in fact , for @xmath27 , we will see that there are many projective structures compatible to a given complex one ( see theorem [ t : quadraticdifferentials ] ) .
we will refer to the projective structure above as the * canonical projective structure * of the riemann surface @xmath0 : it does not depend on the choice of the uniformization of @xmath15 .
we now give other examples .
[ e : quotient](*quotients by kleinian groups * ) let @xmath28 be a subgroup acting properly , freely and discontinuously on some connected open subset @xmath29 .
then , the quotient map @xmath30 induces a projective structure on the quotient @xmath0 , likely as in example [ e : universalcovering ] .
there are many such examples where @xmath15 is neither a disk , nor the plane .
for instance , * quasi - fuchsian groups * are obtained as image of small perturbations of the representation @xmath31 of example [ e : universalcovering ] ; following @xcite , such perturbations keep acting discontinuously on some * quasi - disk * ( a topological disk whose boundary is a jordan curve in @xmath3 ) .
[ e : schottkygroups](*schottky groups * ) pick @xmath32 disjoint discs @xmath33 and @xmath34 in @xmath3 , @xmath27 . for @xmath35 , let @xmath36 be a loxodromic map sending the disc @xmath37 onto the complement @xmath38 . the group @xmath28 generated by @xmath39 acts properly , freely , and discontinuously on the complement @xmath15 of some closed set contained inside the disks ( a cantor set whenever @xmath40 )
. the fundamental domain of this action on @xmath15 is given by the complement of the disks and the quotient @xmath41 is obtained by gluing together the boundaries of @xmath42 and @xmath37 by means of @xmath43 , @xmath44 .
therefore , @xmath0 is a compact riemann surface of genus @xmath1 .
this picture is clearly stable under small deformation of the generators @xmath43 and we thus obtain a complex @xmath45 dimensional family of projective structures on the genus @xmath1 surface @xmath6 ( we have divided here by the action of @xmath13 by conjugacy ) . given a projective atlas and starting from any initial coordinate chart @xmath46 , one can extend it analytically along any path @xmath47 starting from @xmath48 .
indeed , after covering @xmath47 by finitely many projective coordinate charts , say @xmath46 , @xmath49 , ... , @xmath50 , one can modify them step by step in the following way .
first of all , since @xmath51 on @xmath52 , one can replace the chart @xmath53 by @xmath54 which is well - defined on @xmath55 , extending @xmath56 .
next , we replace @xmath57 by @xmath58 which , on @xmath59 , coincide with @xmath60 .
step by step , we finally arrive at the chart @xmath61 which , by construction , is the analytic continuation of @xmath56 along @xmath47 .
therefore , the local chart @xmath46 extends ( after lifting on the universal covering ) as a global submersion on the universal cover @xmath62 which is called the * developping map * of the projective structure .
the developping map is moreover holomorphic with respect to the complex structure subjacent to the projective one . by construction
, the monodromy of @xmath63 along loops takes the form @xmath64 ( @xmath65 is the coordinate on @xmath15 and @xmath66 , the canonical action of @xmath67 on @xmath15 ) .
in fact , @xmath68 is the composition of all transition maps @xmath69 encoutered along @xmath47 for a given finite covering of projective charts : with notations above , setting @xmath70 , we have @xmath71 .
it turns out that @xmath68 only depends on the homotopy class of @xmath47 and we inherit a * monodromy representation * @xmath72 the image @xmath22 of @xmath31 will be called monodromy group .
the developping map @xmath63 is defined by the projective structure up to the choice of the initial chart @xmath46 above : it is unique up to left composition @xmath73 , @xmath74 .
therefore , the monodromy representation is defined by the projective structure up to conjugacy : the monodromy of @xmath73 is @xmath75 .
conversely , any global submersion @xmath76 on the universal covering @xmath77 satisfying ( [ f : developpingmap ] ) is the developping map of a unique projective structure on @xmath6 .
we note that condition ( [ f : developpingmap ] ) forces the map @xmath78 to be a morphism .
the developping map of the canonical projective structure ( see example [ e : universalcovering ] ) is the inclusion map @xmath79 of the universal cover of @xmath0 .
more generally , when the projective structure is induced by a quotient map @xmath80 like in example [ e : quotient ] , then the developping map @xmath63 is the universal cover @xmath81 of @xmath15 and the monodromy group is @xmath22 . in example
[ e : schottkygroups ] , the open set @xmath15 is not simply connected ( the complement of a cantor set ) and the developping map is a non trivial covering .
thus the corresponding projective structure is not the canonical one .
similarly , the developping map of a quasi - fuchsian group is the uniformization map of the corresponding quasi - disk and is not trivial ; the projective structure is neither the canonical one , nor of schottky type .
[ e : thesphere](*the sphere * ) given a projective structure of the riemann sphere @xmath3 , we see that the developping map @xmath82 is uniform ( no monodromy since @xmath83 is trivial ) . therefore , @xmath63 is a global holomorphic submersion ( once we have fixed the complex structure ) and thus @xmath84 .
consequently , the projective structure is the canonical one and it is the unique projective structure on @xmath3 . for similar reason ,
we remark that the monodromy group of a projective structure on a surface of genus @xmath27 is never trivial .
[ e : thetorus](*the torus * ) let @xmath85 be a lattice in @xmath16 , @xmath86 , and consider the elliptic curve @xmath87 .
the monodromy of a projective structure on @xmath0 is abelian ; therefore , after conjugacy , it is in one of the following abelian groups : * the linear group @xmath88 , * the translation group @xmath89 , * the finite abelian dihedral group generated by @xmath90 and @xmath91 .
the canonical projective structure on @xmath0 has translation monodromy group @xmath22 . on the other hand , for any @xmath92
the map @xmath93 is the developping map of a projective structure on @xmath0 whose monodromy is linear , given by @xmath94 we inherit a @xmath19-parameter family of projective structures parametrized by @xmath92 ( note that @xmath95 is not a submersion ) .
we will see latter that this list is exhaustive .
in particular , all projective structures on the torus are actually affine ( transition maps in the affine group ) . the projective structures listed in example [ e : thetorus ] are actually * affine structures * : the developping map takes values in @xmath16 with affine monodromy .
[ t : gunningaffine ] all projective structures on the elliptic curve @xmath96 , are actually affine and listed in example [ e : thetorus ] above .
there is no projective structure having affine monodromy on surfaces @xmath6 of genus @xmath40 .
in particular , the dihedral group is not the holonomy group of a projective structure on the torus . here , we only prove that the list of example [ e : thetorus ] exhausts all affine structures on compact riemann surfaces .
in example [ e : quadraticdifferentialtorus ] , we will see that there are no other projective structure on tori than the affine ones .
let @xmath76 be the developping map of a projective structure with affine monodromy on the compact riemann surface @xmath97 : @xmath63 is an holomorphic local homeomorphism satisfying @xmath98 choose a holomorphic @xmath19-form @xmath99 on @xmath0 and write @xmath100 .
here , we identify @xmath99 with its lifting on the universal covering .
since @xmath63 is a local diffeomorphism , @xmath101 has no zeroes and @xmath102 is a holomorphic on @xmath15 , vanishing exactly on zeroes of @xmath99 and poles of @xmath63 .
moreover , the monodromy of @xmath102 is that of @xmath101 , given by @xmath103
. therefore , the meromorphic @xmath19-form @xmath104 has no monodromy : it defines a meromorphic @xmath19-form on @xmath0 having only simple poles , located at the zeroes of @xmath99 and poles of @xmath63 , with residue @xmath105 .
following residue theorem , @xmath106 has actually no poles : @xmath63 is holomorphic , @xmath99 does not vanish and thus genus @xmath107 .
this proves the second assertion of the statement .
now , assume @xmath107 , @xmath108 and @xmath109 .
the @xmath19-form @xmath106 above is holomorphic and thus takes the form @xmath110 for some constant @xmath111 . in other words
, we have @xmath112 and we obtain after integration * @xmath113 when @xmath114 , * @xmath115 when @xmath116 for constants @xmath117 and @xmath118 . after left composition by an affine map , which does not affect the affine structure , we can set @xmath119 and @xmath120 and @xmath63 belongs to the list of example [ e : thetorus ] .
we see from the proof that the projective structures on @xmath96 are naturally parametrized by @xmath16 , namely the constant map @xmath121 , which is not clear from the description of example [ e : thetorus ] ( we see @xmath122 plus one point ) .
one can recover this by choosing conveniently the integration constants @xmath123 and @xmath124 in the proof above . indeed , consider the alternate family of developping maps given by @xmath125 the map @xmath126 is clearly holomorphic on @xmath127 and makes the developping maps @xmath128 into an holomorphic family parametrized by @xmath111 .
moreover , the corresponding holonomy representations are given by @xmath129 and we see the affine motions with common fixed point @xmath130 converging to translations while @xmath131 . when we set @xmath107 in example [ e : schottkygroups ] , we have @xmath132 and @xmath22 is generated by a single map @xmath133 .
the quotient @xmath41 is the elliptic curve with lattice @xmath134 .
the complex structure varies with @xmath135 and very few projective structures on the torus are obtained by this way .
in fact , we see in example [ e : thetorus ] that , for generic values of @xmath136 , the monodromy group of the corresponding projective structure is not discrete ( @xmath136 is not @xmath137-commensurable with @xmath19 and @xmath138 ) . in order to generalize the arguments involved in the proof of theorem [ t : gunningaffine ] for genus @xmath40 riemann surfaces
, we have to replace @xmath139 by the * schwartzian derivative * of @xmath63 @xmath140 recall that , for any holomorphic functions @xmath63 and @xmath1 , we have @xmath141 given a projective structure on a riemann surface @xmath0 , consider the schwartzian derivative of the corresponding developping map @xmath142 . for any @xmath143 , we deduce from property ( [ f : developpingmap ] ) of @xmath63 that @xmath144 in other words , the * quadratic differential * @xmath145 is invariant under @xmath22 and gives rise to a quadratic differential on the riemann surface @xmath0 .
we note that @xmath146 is holomorphic . indeed , outside the poles of @xmath63
, @xmath147 is not vanishing , thus @xmath148 is holomorphic and @xmath149 well . on the other hand , at a pole of @xmath63 , one can replace @xmath63 for instance by @xmath150 , which is not relevant for the schwartzian derivative , and go back to the previous argument . by this way , we canonically associate to any projective structure on @xmath0 an holomorphic quadratic differential @xmath146 on @xmath0 , i.e. a global section of @xmath151 , where @xmath152 is the canonical line bundle over @xmath0 . conversely , given any holomorphic quadratic differential @xmath145 on the riemann surface @xmath0 , one can solve locally the differential equation @xmath153 in @xmath63 and recover the coordinate charts of a projective structure on @xmath0 ( compatible with the complex one ) : the fact is that any two ( local ) solutions of @xmath153 differ by left composition by a moebius transformation .
[ e : quadraticdifferentialtorus]in genus @xmath19 case , any holomorphic quadratic differential takes the form @xmath154 for a constant @xmath111 ( @xmath155 is still the trivial bundle ) .
in fact , @xmath156 , where @xmath157 . on the other hand ,
any solution of @xmath158 gives rise to a solution of @xmath159 ; therefore , the projective structure defined by @xmath146 is actually subjacent to the affine structure defined by @xmath160 .
this concludes the proof of theorem [ t : gunningaffine ] .
we note that the space of affine structures forms a two fold covering of the space of projective structures ( the choice of the square root @xmath160 ) . of course
, this comes from the fact that the @xmath161 affine structures given by @xmath128 and @xmath162 ( with notations of example [ e : thetorus ] ) do not define distinct projective structures . for genus
@xmath40 riemann surfaces , the dimension of @xmath163 can be computed by riemann - roch formula , and we obtain [ t : quadraticdifferentials ] the set of projective structures on a complex riemann surface @xmath0 of genus @xmath40 is parametrized by the @xmath45-dimensional complex vector space @xmath163 . in this vector space ,
@xmath164 stands for the canonical structure of example [ e : universalcovering ] .
a natural question arising while studying projective structures is to understand , for a given surface @xmath6 , the nature of the monodromy ( or riemann - hilbert ) mapping @xmath165 on the left - hand side , @xmath166 denotes the set of all projective structure on @xmath6 up to isomorphism ; on the right - hand side , @xmath167 is the set of representations of the fundamental group in @xmath13 up to conjugacy : @xmath168 let us first consider the genus @xmath107 case . from gunning s theorem [ t :
gunningaffine ] , the left - hand side can be viewed as a @xmath16-bundle over the modular orbifold @xmath169 where @xmath170 denotes the upper - half plane whose fiber at a given complex structure is the affine line of holomorphic differentials . nevertheless , to avoid dealing with orbifold points , we prefer to deal with the parametrization of affine structures by @xmath171 given by the map @xmath172 here , the base @xmath170 is the space of marked complex structures on the torus , up to isomorphism , and the fiber over @xmath138 is the affine line of differentials @xmath173 , @xmath65 the variable of @xmath170 .
since all projective structures are actually affine , we can replace@xmath174 by @xmath175 where @xmath176 is the group of affine transformations .
once we have fixed generators @xmath19 and @xmath138 for the fundamental group of @xmath109 , the set @xmath177 identifies with the complex @xmath178-dimensional subvariety @xmath179 ( here , we see the condition for the commutativity ) .
since any linear representation does not occur as monodromy representation of an affine structure on the torus , we consider the quotient @xmath180 of the complement of @xmath181 in this variety .
it is easy to see that @xmath182 can be identified with the @xmath161-dimensional complex manifold @xmath183)\ ; \
( a_1 - 1)b_\tau=(a_\tau-1)b_1\}\subset \mathbb c^*\times \mathbb c^*\times \mathbb p^1\ ] ] where @xmath184 $ ] denotes homogeneous coordinates on @xmath3 .
the projection @xmath185 is just the blow - up of the point @xmath186 and the exceptional divisor consists in euclidean representation .
finally , the monodromy map is described by @xmath187 ) , & c\not=0\\ ( 1,1,[1:\tau]),&c=0\end{array}\right.\ ] ] looking at the differential of the monodromy map above , we see that it has always rank @xmath161 and _ the monodromy map is an holomorphic local diffeomorphism _ ; it is moreover _ injective and proper in restriction to each fiber @xmath188_. its image is the complement of the real torus @xmath189 , or we better should say , its lifting on @xmath182 : the complement of @xmath190 in restriction to the exceptional divisor . * but * the monodromy map _ is neither injective , nor a covering map onto its image _ : for instance , for any @xmath191 , @xmath192 , and for any @xmath193 , the two affine structures @xmath194 have the same monodromy representation . in particular , the injectivity is violated for arbitrarily close complex structures .
on the other hand , the monodromy of the canonical structure @xmath195 occurs only for this structure .
consider now the genus @xmath40 case .
it follows from gunning s theorem [ t : quadraticdifferentials ] above that the set @xmath166 of projective structures on the genus @xmath40 surface @xmath6 can be viewed as a complex @xmath196-dimensional space .
indeed , if we denote by @xmath197 the teichmller space of complex marked structures on @xmath6 viewed as an open subset of @xmath198 , then @xmath166 is parametrized by the rank @xmath45-vector bundle @xmath199 over @xmath197 whose fiber over a given complex structure @xmath0 is the space of quadratic differentials @xmath163 . by theorem [ t :
gunningaffine ] , the monodromy representation can not be affine in the case @xmath200 . the image of @xmath201 by the monodromy map is thus included in the subset of * irreducible * representations @xmath202 where @xmath203 is the set of affine representations * up to @xmath13-conjugacy*. one can check ( see @xcite ) that @xmath204 forms a non - singular complex manifold of dimension @xmath196 .
thus , the monodromy map can locally be described as a holomorphic map between open subsets of @xmath205 and the following result makes sense ( see proof in section [ s : bundlepicture ] ) .
[ t : localdiffeomorphism ] the monodromy map is a local diffeomorphism . in @xcite , it is moreover proved that the monodromy map is symplectic with respect to symplectic structures that can be respectively canonically defined on both spaces ( see @xcite ) .
the restriction of the monodromy map to each fiber @xmath206 of @xmath207 over @xmath208 is injective . in other words
, we have the following result whose proof will be given in section [ sec2.2 ] .
[ t : poincare ] given a compact riemann surface @xmath0 , any two projective structures are the same if , and only if , they have the same monodromy representation ( up to @xmath13 ) .
it is clear that the monodromy map is not surjective .
first of all , by theorem [ t : gunningaffine ] , its image is contained in @xmath209 .
on the other hand , the space of representations @xmath210 falls into @xmath161 connected components , namely the component of those that can be lifted as @xmath211 and the other ones .
since the monodromy map is continuous ( actually holomorphic ) and since the monodromy of canonical projective structures can be lifted to @xmath212 , it becomes clear that the image of the monodromy map will be in the former component .
finally , notice that the monodromy representation can not be in @xmath213 , i.e. conjugated to a group of rotations of the sphere , otherwise we could pull - back the invariant spherical metric of @xmath3 by the developping map , giving rise to a curvature @xmath105 metric on the surface , impossible except in the trivial case @xmath18 ( see example [ e : thesphere ] ) .
the main result in the field , which has been conjectural for decenies , is the following .
[ t : gallokapovichmarden ] consider the genus @xmath1 surface @xmath6 , @xmath40 .
an homomorphism @xmath214 is the monodromy representation of a projective structure on the @xmath6 if , and only if , @xmath31 can be lifted as @xmath215 and the image of @xmath31 is , up to @xmath13-conjugacy , neither in the affine group @xmath216 , nor in the rotation group @xmath217 .
let @xmath76 be the developping map of a projective structure on @xmath0 ( here we fix the underlying complex structure ) and consider its graph @xmath218 .
the fundamental group @xmath219 acts on the product @xmath220 in as follows : for any @xmath23 , set @xmath221 where @xmath222 is the canonical action of @xmath219 on the universal cover and @xmath68 is the monodromy of the projective structure along @xmath47 .
this action of @xmath219 is proper , free and discontinuous since its projection on @xmath15 is so . by consequence
, we can consider the quotient : @xmath223 the projection @xmath224 defined by @xmath225 , where @xmath14 is the universal cover , is preserved by the action and induces a global submersion @xmath226 making _
@xmath227 into a @xmath3-bundle over @xmath0 .
the graph of @xmath63 also is invariant under the action ( consequence of ( [ f : developpingmap ] ) ) thus defining a section _
@xmath228 finally , the horizontal foliation defined by @xmath229 is also preserved and defines a foliation @xmath5 transversal to all @xmath3-fibres on @xmath227 .
since the developping map @xmath63 is regular , its graph is transversal to the horizontal foliation and @xmath4 is transversal to @xmath5 . in this situation
, we say that the @xmath230-bundle @xmath227 is flat .
the triple @xmath231 is well - defined by the projective structure up to analytic isomorphism of @xmath3-bundles .
conversely , given a @xmath3-bundle @xmath232 , a foliation @xmath5 on @xmath227 transversal to @xmath233 and a section @xmath234 transversal to @xmath5 , then the ( unique ) projective structure on @xmath3-fibres can be transported , transversely to the foliation @xmath5 , inducing a projective structure on the section @xmath235 , and thus on its @xmath233-projection @xmath0 . in the recent terminologoly of @xcite , such triple @xmath231 are called @xmath236-opers .
[ rm13]given an homomorphism @xmath237 , one can at least construct the pair @xmath238 as above .
this foliated surface is called the * suspension * of the representation @xmath31 , also known as the * flat @xmath3-bundle * associated to @xmath31 .
conversely , consider a flat @xmath3-bundle , i.e. a pair @xmath238 where @xmath232 is a @xmath3-bundle and @xmath5 is a foliation transversal to @xmath233
. then one can associate to it a representation @xmath31 in the following way . over any sufficiently small open subset @xmath239
, one can construct a trivializing coordinate @xmath240 for the flat bundle , that is to say inducing an isomorphism in restriction to each fibre and such that the level curves @xmath241 are local leaves of the foliation @xmath5 .
in fact , @xmath242 is uniquely determined after choosing the local @xmath5-invariant sections @xmath243 along which @xmath242 takes values @xmath164 ,
@xmath19 and @xmath244 respectively .
such flat coordinate is well defined up to left - composition by a moebius transformation ; likely as in section [ s : monodromy ] , after fixing a flat local coordinate @xmath126 over some neighborhood of the base point @xmath245 , we inherit a monodromy representation @xmath246 where the analytic continuation of @xmath126 along any loop @xmath47 satisfies @xmath247 .
it turns out that any flat @xmath3-bundle is isomorphic to the suspension of its monodromy representation just defined .
in fact , _ any two flat @xmath3-bundles are isomorphic if , and only if , they have the same monodromy representation up to @xmath13 conjugacy_. indeed , let @xmath238 and @xmath248 be flat @xmath3-bundles having flat coordinates @xmath126 and @xmath249 over @xmath250 giving rise to the same monodromy representation ; then the local isomorphism @xmath251 sending any point @xmath252 to the unique point @xmath253 satisfying @xmath254 extends uniformly as a global isomorphism of flat @xmath3-bundles @xmath255 , i.e. conjugating @xmath5 to @xmath256 and satisfying @xmath257 . in fact , since the monodromy map is clearly holomorphic , it is enough to prove that it is locally bijective .
let @xmath231 be the triple associated to a projective structure having monodromy representation @xmath258 .
for any perturbation @xmath259 of @xmath31 , the corresponding suspension @xmath248 is close to the foliated bundle @xmath238 ; if the perturbation is small enough , one can find a real @xmath260 section @xmath261 close to @xmath234 and still transversal to @xmath256 ( all of this makes sense and can be checked on the neighborhood of a fundamental domain of the universal cover @xmath220 ) .
the foliation @xmath256 still induces a projective structure on the real surface @xmath262 that , by construction , has the required monodromy .
this proves the surjectivity .
let @xmath231 be the triple associated to a projective structure @xmath263 and consider another projective structure @xmath264 close to this @xmath263 having the same monodromy representation .
the fibre bundle construction can be done in the real @xmath260 setting so that one can associate to @xmath264 a triple @xmath265 where @xmath0 is still the complex curve attached to @xmath263 and @xmath266 is now a real @xmath260 section transversal to @xmath5 ; we note that the pair @xmath238 is the same for @xmath263 and @xmath264 since they have the same monodromy representation .
if @xmath264 is close enough to @xmath263 , say in the @xmath260 category , then @xmath267 is close to @xmath4 ; one can therefore unambiguously define a @xmath260 diffeomorphism @xmath268 by following the leaves of the foliation from one section to the other one . by construction
, the projective structures induced by @xmath5 on both sections are conjugated by @xmath102 .
the diffeomorphism @xmath269 actually integrates the quasi - conformal structure induced by @xmath264 on @xmath0 ; it is close to the identity .
since the monodromy map is not globally injective , the injectivity argument of the previous proof can not be carried out for sections @xmath4 and @xmath267 that are not closed enough : the set of @xmath260 sections transversal to @xmath5 may have infinitely many connected components as it so happens in the case of affine structures on the torus . similarly ,
the surjectivity argument of the proof can not be globalized : when the monodromy representation @xmath270 eventually becomes reducible for instance , there does not exist @xmath260 section transversal to @xmath5 anymore . following theorem [ t : gallokapovichmarden ] , the existence of a @xmath260 section transversal to @xmath5 is possible if , and only if , @xmath5 is the suspension of a non elementary representation @xmath31 ( lifting to @xmath271 ) ! from this point of view , theorem [ t : gallokapovichmarden ] looks like a very subtle transversality result .
motivated by the fibre bundle picture of section [ s : bundlepicture ] , we developp here the study of riccati foliations on @xmath3-bundles over compact riemann surfaces .
let @xmath232 be a @xmath3-bundle over a compact riemann surface @xmath0 : @xmath227 is a smooth surface and the fibers of @xmath233 are rational , isomorphic to @xmath3 .
we also say that @xmath227 is a * ruled surface*. another @xmath3-bundle @xmath272 is * analytically equivalent * to the previous one if there is an holomorphic diffeomorphism @xmath273 such that @xmath274 .
we recall some basic facts ( see @xcite ) . on open charts @xmath275 on @xmath0 ,
the bundle becomes analytically trivial ( see @xcite ) : we have holomorphic diffeomorphisms ( trivializing coordinates ) @xmath276 on overlapping charts @xmath277 , the transition maps take the form @xmath278 where @xmath279 and @xmath280 the @xmath3-bundle is equivalently defined by the collection @xmath281 by lifting conveniently the transition maps into @xmath282 , we may view a @xmath3-bundle as the projectivization @xmath283 of a * rank @xmath161 vector bundle * @xmath284 over @xmath0 .
moreover , another vector bundle @xmath285 will give rise to the same @xmath3-bundle if , and only if , @xmath286 for a line bundle @xmath287 over @xmath0 .
the classification of @xmath3-bundles is thus equivalent to the classification of rank @xmath161 vector bundles up to tensor product by a line bundle . from the * topological * point of view , due to the fact that @xmath288 , there are exactly @xmath161 distinct @xmath289-bundles over a compact real surface . from the * birational * point of view ,
any @xmath3-bundle is equivalent to the trivial bundle : there are infinitely many holomorphic sections @xmath234 ; after choosing @xmath178 distinct ones @xmath290 , @xmath291 and @xmath292 , one defines a birational transformation @xmath293 commuting with @xmath233 by sending those sections respectively to @xmath294 , @xmath295 and @xmath296 . when the @xmath178 sections are * disjoint * , the transformation @xmath102 is actually biregular and @xmath227 is the trivial bundle @xmath297 . the * analytic classification * is a much more subtle problem .
if @xmath227 admits @xmath161 disjoint sections , say @xmath298 , we then say that the bundle is * decomposable * : one can choose trivialization charts sending those two sections respectively onto @xmath294 and @xmath299 , so that @xmath227 may be viewed as the compactification @xmath300 of a line bundle @xmath287 .
recall that line bundles are analytically classified by the picard group @xmath301 .
any two elements @xmath302 have the same compactification @xmath227 if , and only if , @xmath303 or @xmath304 : we just exchange the role of @xmath290 and @xmath292 ( see proof of proposition [ p : riccatitorus ] ) . for instance , on @xmath305 , @xmath306 and the compactification of @xmath307 ( or @xmath308 ) , @xmath309 , gives rise to the hirzebruck surface @xmath310 .
it follows from birkhoff s theorem @xcite that all @xmath3-bundle is decomposable on @xmath3 and is thus one of the @xmath310 above .
an important analytic invariant of a @xmath3-bundle over a curve @xmath0 is the minimal self - intersection number of a section @xmath311 for a decomposable bundle @xmath312 , @xmath313 , we have @xmath314 . for an undecomposable bundle , nagata proved in @xcite that @xmath315 and all those possibilities occur . from the * homological * point of view
, @xmath316 is generated by the homology class of @xmath290 and @xmath63 where @xmath290 is any holomorphic section and @xmath63 any fibre .
let us choose @xmath290 with minimal self - intersection : @xmath317 the homology class of any other holomorphic section is @xmath318 with @xmath319 : it has self - intersection @xmath320 in particular , the intersection number of holomorphic sections are either all even , either all odd : @xmath321 mod @xmath161 is the * topological invariant * of the bundle . on the other hand , if @xmath322 and @xmath4 are not homologous then the intersection number @xmath323 must be non negative and we deduce that @xmath324 : when @xmath325 , this implies that _
@xmath290 is the unique holomorphic section having negative self - intersection _ ; there is a gap between @xmath326 and @xmath321 .
[ t : atiyah ] beside compactifications of line bundles , there are exactly @xmath161 undecomposable @xmath3-bundles over an elliptic curve , @xmath327 and @xmath328 , with invariant @xmath329 and @xmath330 respectively .
a @xmath3-bundle @xmath227 is * flat * ( in the sense of steenrod @xcite ) when a trivializing atlas can be choosen with constant transition maps @xmath331 ( not depending on @xmath65 ) .
this means that this atlas defines by the same time a foliation @xmath5 transversal to the fibres on @xmath227 , namely the horizontal foliation defined by @xmath229 in trivializing coordinates , see remark [ rm13 ] .
[ t : weil ] the flat @xmath3-bundles over @xmath0 are all undecomposable bundles and all those arising as compactification of elements of @xmath332 .
the pairs @xmath238 are classified by @xmath333 .
a riccati foliation on the bundle @xmath232 is a singular foliation ( see definition in @xcite ) @xmath5 on @xmath227 which is transversal to a generic fibre . in trivialization charts @xmath334 ,
it is defined by a riccati differential equation @xmath335 , @xmath336 meromorphic in @xmath65 , whence the name .
the poles of the coefficients correspond to vertical invariant fibres for the foliation . outside of those poles ,
the leaves of the foliation are graphs of solutions for the riccati equation .
the foliation @xmath5 arising in the fibre bundle picture of section [ s : bundlepicture ] is a * regular * riccati foliation .
nevertheless , we will need to deal with singular foliations later .
one can define the monodromy representation of a riccati foliation as @xmath337 a classical theorem due to poincar asserts that , in the regular case , the monodromy representation characterizes the riccati foliation as well as the @xmath3-bundle supporting it up to analytic equivalence .
one can view a riccati foliation @xmath5 on the @xmath3-bundle @xmath283 as the projectivization of a meromorphic linear connection @xmath338 on the vector bundle @xmath284 .
in fact , given a ( meromorphic linear ) connection @xmath339 on the determinant bundle @xmath340 , there is a unique connection @xmath338 on @xmath284 lifting @xmath5 and such that @xmath341 .
indeed , over a local coordinate @xmath275 , the bundle @xmath284 is trivial and a connection @xmath338 is just a meromorphic system @xmath342 and the trace of @xmath338 is the rank @xmath19 connection defined by @xmath343 the projection of @xmath338 on @xmath344 is therefore the riccati equation defined in affine coordinate @xmath345 by @xmath346 clearly , @xmath338 is uniquely defined by @xmath5 and @xmath339 .
notice finally that the line bundle @xmath347 admits a linear connection @xmath339 without poles if and only if it belongs to @xmath348 .
we start recalling some usefull homological formulae from @xcite .
first of all , let us introduce @xmath349 , the * tangent bundle * of @xmath5 , which is a line bundle on the total space @xmath227 defined as follows . in trivializing charts @xmath350 ,
the riccati foliation is also defined by the meromorphic vector field @xmath351 .
the leaves of @xmath5 are just complex trajectories of the vector field @xmath352 . after choosing a global ( meromorphic ) vector field @xmath353 on @xmath0
, one can write @xmath354 for a meromorphic function @xmath355 on the chart @xmath356 so that the new meromorphic vector fields @xmath357 glue together into a global meromorphic vector field @xmath284 on @xmath227 still defining @xmath5 at a generic point .
one can think of @xmath284 as the lifting of @xmath353 by the ( meromorphic ) projective connection defined @xmath5 on the bundle .
then , @xmath349 is the line bundle defined by the divisor of @xmath284 , i.e. @xmath358 . if @xmath359 denotes the number of invariant fibre ( counted with the multiplicity of the corresponding pole for @xmath352 ) , then the homology class of @xmath349 is given by @xmath360 . given a curve @xmath4 on @xmath227 ,
each component of which is not invariant by @xmath5 , then the number of * tangencies * @xmath361 counted with multiplicities is given by ( see @xcite , p. 23 ) @xmath362 for instance , if @xmath363 is a section , we immediately deduce that @xmath364 .
consider two projective structures on @xmath0 ( compatible with the complex structure of @xmath0 ) having the same monodromy representation : by the construction given in section [ s : bundlepicture ] , they correspond to triples @xmath231 and @xmath265 with common @xmath3-bundle and riccati foliation .
since @xmath5 is regular and the section @xmath4 defining the first projective structure is transversal to @xmath5 , we have @xmath365 , @xmath366 , and we deduce that @xmath367 . on the other hand , @xmath368 should be non negative and we obtain @xmath369 and @xmath370 : in the genus @xmath40 case , @xmath371 is the unique section having negative self - intersection in @xmath227 , and by the way @xmath372 . in genus
@xmath164 case , there is nothing to show ; in genus @xmath19 case , the result follows directly from formula ( [ f : linearmonodromytorus ] ) and theorem [ t : gunningaffine ] .
another important formula is the camacho - sad index theorem ( see @xcite ) . given a curve @xmath4 on @xmath227 invariant by @xmath5 , the self - intersection number of @xmath4 equals the sum of camacho - sad index of @xmath5 along this curve .
when @xmath5 is regular , all invariant curves are smooth and all camacho - sad index vanish : when @xmath5 is regular , any invariant curve @xmath4 has zero self - intersection . for instance ,
if @xmath5 has affine monodromy , then the fixed point gives rise to an invariant section @xmath373 .
we deduce that @xmath329 and @xmath292 realizes this minimal self - intersection number .
in particular , we recover the fact that a projective structure on a genus @xmath40 curve can not have affine monodromy since the corresponding bundle has invariant @xmath374 .
more generally , if the monodromy of @xmath5 has a finite orbit ( e.g. a finite group of the infinite dihedral group ) , then @xmath5 has an invariant curve @xmath375 and formulae @xmath376 together with @xmath377 and @xmath378 show that @xmath379 .
again , this is not the monodromy of a projective structure whenever @xmath40 . in the particular case
where the monodromy of @xmath5 is linear , we have @xmath161 invariant disjoint sections @xmath290 and @xmath292 showing that the bundle @xmath227 is actually a compactification of a line bundle @xmath380 .
we should emphasize that any two line bundles @xmath287 and @xmath381 have the same compactification if , and only if , @xmath303 or @xmath304 .
indeed , we first note that , for a bundle @xmath227 satisfying @xmath382 , any two sections @xmath4 and @xmath267 are disjoint if , and only if , they have @xmath164 self - intersection ( and are distinct ) .
the compactification @xmath300 of a line bundle @xmath287 always has the two canonical disjoint sections @xmath290 and @xmath292 .
now , a diffeomorphism @xmath383 between the compactifications of @xmath161 non trivial line bundles has to preserve or permute the two canonical sections ; in the former case , @xmath102 is actually an equivalence of line bundles ; in the latter case , @xmath102 restricts to the fibres as @xmath91 and invert the monodromy .
of course , @xmath300 is trivial if , and only if , @xmath287 is trivial as a line bundle .
it follows that when @xmath0 has genus @xmath19 the corresponding set of equivalence classes of @xmath3-bundles may be thought as @xmath384 .
a bundle @xmath227 obtained by suspension of a representation @xmath385 is topologically trivial ( @xmath321 even ) if , and only if , @xmath31 can be lifted as a representation @xmath386 .
there is an algebraic and somewhat technical notion of ( semi- ) stability of vector bundles of arbitrary rank on riemann surfaces due to mumford , see for instance @xcite .
we can define the ( semi- ) stability of a @xmath230-bundle @xmath344 by the same requirement to the rank @xmath161 vector bundle @xmath284 .
it turns out that a @xmath3-bundle is * stable * ( resp . * semi - stable
* ) when @xmath387 ( resp .
@xmath388 ) .
it is known that if such a bundle occur along an algebraic ( resp .
analytic ) family , it occurs for a zariski open subset of the family .
there is a theorem of narashimhan and seshadri characterizing stable bundles on a compact riemann surface @xmath0 by means of a precise , but some technical , construction in terms of unitary representations of the fundamental group of @xmath0 .
we present here a more comprehensible consequence , see corollaries 1 and 2 of @xcite : let @xmath0 be a compact riemann surface of genus @xmath200 .
then a holomorphic vector bundle of degree zero is stable if and only if it arises from an irreducible unitary representation of the fundamental groups @xmath389 of @xmath0 .
a holomorphic vector bundle on @xmath0 arises from a unitary representation of the fundamental group if and only if each of its undecomposable components is of degree zero and stable . applying this general result to our situation
we obtain that the map @xmath390 which to a representation @xmath391 associate the @xmath3-bundles obtained by suspension ( forgetting the flat structure ) induces a bijection from the set of irreducible representations @xmath392 up to @xmath217 conjugacy onto the set of isomorphism class of @xmath3-bundles with invariant @xmath387 and even ( not fixed ) .
the complete analytic classification of @xmath3-bundles ( including unstable ones ) over curves of genus @xmath161 has been achieved by the works of atiyah @xcite and maruyama @xcite .
the analytic classification of rank @xmath161 stable vector bundles over curves of arbitrary genus from the algebraic point of view ( in contrast with narasimhan - seshadri s approach ) has been done by tyurin in @xcite ( see also @xcite for a survey in arbirtrary rank ) .
given a point @xmath252 on ( the total space of ) a @xmath3-bundle @xmath232 , we will denote by @xmath393 the new @xmath3-bundle obtained after elementary transformation centered at @xmath252 : after blowing - up the point @xmath252 , @xmath393 is obtained by contracting the strict transform of the fiber passing through @xmath252 .
the strict transform of a section @xmath4 passing through @xmath252 ( resp .
not passing through @xmath252 ) is a section of the new bundle having self - intersection @xmath394 ( resp .
@xmath395 ) .
all birational transformations between @xmath3-bundles over curves are obtained by composing finitely many elementary transformations . on the other hand ,
any @xmath3-bundle over a curve is birational to the trivial bundle .
for instance , let @xmath396 be a divisor on @xmath0 and let @xmath397 be the point on the zero section of the ( total space of the ) line bunde @xmath398 over @xmath399
. then @xmath400)}.\ ] ] similarly , if @xmath401 lie on the infinity section of @xmath402 over @xmath403 , then @xmath404)}.\ ] ] now , recall that , as a consequence of abel theorem , the map @xmath405-[x_{1}]-\cdots-[x_{g}])\ ] ] is surjective for any @xmath406 : it follows that ( compactification of ) line bundles of degree @xmath164 can be obtained after applying at most @xmath32 elementary transformations to the trivial bundle . in @xcite ,
maruyama and nagata proved that an undecomposable bundle can be obtained from the trivial one after at most @xmath407 elementary transformations . on the other hand
, we note that the minimal number of elementary transformations needed to trivialize all decomposable bundle is unbounded : for a line bundle of large degree @xmath408 , one need at least @xmath359 elementary transformations . first , we would like to make explicit the correspondance between the point of view of quadratic differentials , and that one of bundle triples .
consider the triple @xmath231 associated to a projective structure on the curve @xmath0 .
one can reduce @xmath227 to the trivial bundle and @xmath4 to the infinity section @xmath299 either locally , by a fibre bundle isomorphism , or globally on @xmath0 , by birational transformation . here
below , we adopt the later point of view ; everything can be carried out _ mutatis mutandis _ in the local regular setting .
after a birational trivialization like above , @xmath5 becomes possibly singular , but is now defined by a global riccati equation @xmath410 where @xmath411 are meromorphic @xmath19-forms on @xmath0 .
this trivialization is unique up to birational transformation of the form @xmath412 where @xmath123 and @xmath124 are meromorphic function on @xmath0 , @xmath413 .
let us see how such change of coordinate acts on the equation . a change of coordinate of the form @xmath414 transforms the riccati equation into @xmath415 although a change of coordinate @xmath416 yields @xmath417 after a combination of those two transformations , we can choose @xmath418 and @xmath419 arbitrary ( with @xmath420 ) and then @xmath47 is uniquely determined by the projective structure .
let us show how to compute it from the developping map @xmath63 of the projective structure .
let us go back to the universal cover where the riccati foliation is given by @xmath421 and @xmath4 is the graph of @xmath63 ( see section [ s : bundlepicture ] ) . by a preliminary change of coordinate @xmath422 , we have now @xmath423 and the equation becomes @xmath424 .
a second change of coordinate @xmath425 yields @xmath426 , and @xmath4 is still the zero section @xmath427 . in the case of an affine structure on a torus ,
the later riccati equation is well - defined : the corresponding triple @xmath231 is then given by : @xmath428 for some @xmath111 . in the general projective case ,
it is more convenient to send the section @xmath4 to the infinity : in the coordinate @xmath429 , @xmath5 is defined by @xmath430 and @xmath4 , by @xmath431 .
we finally apply the change of coordinate @xmath432 and obtain @xmath433 where @xmath434 is the schwartzian derivative of @xmath63 with respect to the variable @xmath65 .
unfortunately , @xmath435 is not a global @xmath19-form .
moreover , @xmath65 is a transcendental variable that we do not want to deal with when we are considering a triple @xmath231 . in general , by birational trivialization of the bundle , one can assume @xmath4 at infinity and , after choosing a global holomorphic @xmath19-form @xmath418 on @xmath0 , reduce the riccati foliation to the special form @xmath436 with @xmath47 meromorphic on @xmath0 . here , @xmath418 plays the role of @xmath435 , that is @xmath65 is replaced by a variable @xmath353 such that @xmath437 ; this takes sense at least at a generic point of @xmath0 where everything is regular . setting @xmath438 , the change of coordinate @xmath439 transforms equation ( [ eq : riccatischwarzian ] ) into ( [ eq : riccatinormalform ] ) ; after computation we find @xmath440 . using ( [ f : schwartziancomposition ] ) , one finally obtains @xmath441 where @xmath442 is the schwarzian derivative with respect to @xmath353 and deduce let @xmath231 be a triple defining a projective structure on @xmath0 .
let @xmath443 be bundle coordinates over @xmath444 , @xmath445 , such that @xmath446 then the projective coordinates @xmath63 on @xmath15 are the solutions of @xmath447 .
following @xcite , the maximally unstable undecomposable @xmath3-bundle @xmath227 corresponding to projective triples @xmath448 can be trivialized after @xmath32 elementary transformations ( here @xmath369 is even ) .
the birational tranformation constructed above to put @xmath5 into the normal form ( [ eq : riccatinormalform ] ) however needs much more elementary transformations .
indeed , at a point where @xmath449 has a zero of order @xmath450 , i.e. @xmath451 , the expression @xmath452 has a pole of order @xmath161 and thus @xmath453 has a pole of order @xmath454 .
in fact , @xmath455 and the birational change of coordinate takes the form @xmath456 @xmath457 elementary transformations are needed at this point .
now , we look for a sharp birational trivialization of @xmath227 , that is to say with exactly @xmath32 elementary transformations . for any choice of global meromorphic @xmath19-forms @xmath418 and @xmath419 , there is a unique birational transformation of the form @xmath458 putting the initial riccati equation ( [ eq : riccatischwarzian ] ) into the form ( [ e : globalriccati ] ) and we have @xmath459 each zero or pole of @xmath418 ( or @xmath123 ) gives rise to an elementary transformation : if we choose @xmath418 holomorphic , we already get @xmath460 elementary transformations with the first change of coordinate .
we would like now @xmath461 be holomorphic ( as much as possible ) .
the sum of residues of @xmath462 is @xmath460 : we can construct a meromorphic @xmath19-form @xmath419 having the same principal part as @xmath462 , plus one extra simple pole ( at , say , @xmath252 ) with residue @xmath463 .
the final change of coordinate @xmath464 is therefore a combination of @xmath32 elementary transformations : the change of coordinate @xmath414 goes from the trivial bundle to @xmath465 with @xmath460 elementary transformations ; the ultimate transformation @xmath466 has one simple pole corresponding to a succession of @xmath161 generic elementary transformations of the same fibre ( compare @xcite ) .
setting @xmath467 , @xmath468 , the differential equation @xmath469 is transformed into @xmath470 then the following goes back to schwarz : any solution @xmath63 to the differential equation @xmath471 takes the form @xmath472 where @xmath473 and @xmath474 are independant solutions of ( [ e : scalarequation ] ) .
a straightforward computation shows that @xmath475 provided that @xmath473 and @xmath474 are solutions of ( [ e : scalarequation ] ) .
any other solution @xmath471 takes the form @xmath476 , a quotient of two other solutions .
in fact , one can take @xmath472 with @xmath477 and @xmath478 .
one can easily generalize the notion of projective structure to the branching case by considering triples @xmath231 with @xmath4 generically transversal to @xmath5 : branching points of the structure are those points @xmath399 over which @xmath4 has a contact with @xmath5 .
the local projective chart then takes the form @xmath479 where @xmath480 is the order of contact .
more generally , one can consider a singular riccati foliation @xmath5 generically transversal to @xmath4 , or equivalently linear equation @xmath469 with @xmath102 meromorphic . by the way
, projective structures on the @xmath178-punctured sphere ( @xmath178 simple poles ) correspond to the gauss hypergeometric equation , on the @xmath481-punctured sphere , to the heun equation and on the punctured torus , to the lam equation .
let @xmath482 be an effective divisor on @xmath483 .
consider the set @xmath484 consisting in all the projective structures on @xmath483 branched over the points @xmath485 with ramification order @xmath486 ( see @xcite ) .
notice that the case @xmath487 corresponds to genuine projective structures on @xmath483 .
as before we can describe the elements of @xmath484 as triples @xmath488 , where @xmath489 is a @xmath230-bundle with structural group @xmath490 , @xmath491 is a transversely projective foliation transverse to the fibres , @xmath4 is a section such that @xmath492 is a tangency point with @xmath491 of order @xmath493 for each @xmath494 and outside these points @xmath4 is transverse to @xmath491 .
projecting to @xmath483 the branched projective structure induced by @xmath491 on @xmath4 we obtain an orbifold complex structure @xmath0 over @xmath495 .
we can make a finite number of elementary transformations centered at the tangency points @xmath492 in order to obtain a birationally equivalent triple @xmath496 , where @xmath497 is a singular riccati foliation and @xmath261 is a holomorphic section everywhere transverse to @xmath497 .
applying the same transversality arguments of the proof of hejal s theorem to @xmath496 one shows that the monodromy mapping @xmath498 is also a local diffeomorphism .
* @xmath499 is linear and @xmath500 is the compactification of a line bundle ; @xmath227 is trivial if , and only if , @xmath501 .
* @xmath502 is euclidean and either @xmath503 is the semi - stable undecomposable bundle , or @xmath227 is the trivial bundle ; we are in the latter case if , and only if , @xmath504 .
* @xmath505 and @xmath506 is the stable undecomposable bundle .
it is easy to verify that all representation @xmath507 appear in the statement .
we have already noticed that a linear representation gives rise to the compactification of a line bundle ( this is almost the definition ) .
in fact , for linear representations , we have the exact sequence of sheaves @xmath508 where @xmath509 is the sheaf of invertible holomorphic functions and the morphism @xmath510 is given by @xmath511 . from the corresponding exact sequence of cohomology groups , we deduce the following one @xmath512 the first non trivial morphism associates to an holomorphic @xmath19-form @xmath146 the homomorphism @xmath513 while the second one is the suspension . in our particular case where @xmath0 is an elliptic curve , we finally deduce @xmath514 and the first alternative of the statement follows .
the suspension of an euclidean representation gives rise to a bundle with a section @xmath292 having @xmath164 self - intersection .
if there is another section @xmath290 disjoint from @xmath292 , then it should be either transversal , or invariant by @xmath5 from ( [ e : tang ] ) : in the first case , @xmath290 provides a projective structure on @xmath0 and the monodromy satisfies @xmath504 by guning theorem [ t : gunningaffine ] ; in the second case , the monodromy has two fixed points and is trivial , so is the bundle . in the remaining case
where there is no disjoint section from @xmath292 , the bundle is undecomposable with invariant @xmath329 and we conclude with atiyah theorem [ t : atiyah ] that @xmath503 . finally , if @xmath505 is the irreducible representation , we note that @xmath31 can not be lifted to @xmath271 and thus @xmath321 is odd . on the other hand , weil theorem [ t : weil ]
tells us that @xmath227 must be undecomposable ( being flat with @xmath515 ) . from atiyah theorem [ t : atiyah ] , the only possibility is @xmath506 .
it follows from @xcite that all degree @xmath164 line bundles as well as @xmath327 can be obtained after @xmath161 elementary transformations of the trivial bundle . in order to obtain @xmath516 ,
a third one is needed .
we use this approach to provide an algebraic family of flat bundles and riccati foliations .
let @xmath517 , @xmath518 and consider @xmath519 .
fix trivializing coordinates @xmath520 and , for simplicity , set @xmath521 .
this is irrelevant since all flat @xmath3-bundles over @xmath0 admit a one parameter group of automorphism lifting the action of @xmath522 ( see description of proposition [ p : riccatitorus ] ) .
after elementary transformation at the point @xmath252 , we obtain the bundle @xmath523)}$ ] having one section @xmath292 with @xmath330 self - intersection and a special point @xmath524 , on the fiber over @xmath525 but not on @xmath292 , through which all sections having @xmath105 self - intersection intersect .
indeed , @xmath105-sections come from horizontal sections of the trivial bundle . here
, we use the fact that there is no holomorphic section of homology type @xmath526 on @xmath527 , otherwise it would be the graph of a regular covering @xmath528 .
case 1 : @xmath531 with @xmath532 and @xmath533 .
after vertical automorphism , one may assume @xmath534 .
the sections @xmath535 and @xmath536 respectively give rise to disjoint sections @xmath290 and @xmath292 on @xmath227 having @xmath164 self - intersection .
we are in @xmath332 case : @xmath312 .
the generic horizontal section @xmath537 gives rise to a section @xmath4 on @xmath227 intersecting @xmath290 at @xmath525 and @xmath292 at @xmath538 ; in other words , @xmath4 is a meromorphic section of @xmath287 with divisor @xmath539-[u_0]$ ] on @xmath0 : @xmath540-[0])$ ] corresponds to @xmath541 .
case 2 : @xmath542 is on the fibre over @xmath525 but is neither @xmath524 , nor on @xmath292 .
then , @xmath503 is the indecomposable bundle . indeed
, assume that there exists a section @xmath4 on @xmath227 disjoint from @xmath292 .
it then comes from a section of @xmath543)}$ ] disjoint from @xmath292 and passing through @xmath542 , itself coming from a section of @xmath544 intersecting @xmath545 only at @xmath525 , without multiplicity .
we have already seen that this can not happen . here
, we have parametrized all topologically trivial flat bundles by the line bundle @xmath547)$ ] , see @xcite .
we now want to parametrize all regular riccati foliations on topologically trivial bundles .
the natural way to do this is to provide an explicit family of riccati equations on the trivial bundle @xmath544 having appearant singular fibres whose desingularization span all regular riccati foliations . for instance , consider a linear riccati foliation defined on the bundle @xmath548-[0])}$ ] , @xmath532 .
apart from the invariant sections @xmath290 and @xmath292 , the leaves are multisections without zero or pole ; after trivialisation of the bundle , those multisections @xmath549 have now a simple pole over @xmath525 and a simple zero over @xmath538 ( and still have linear monodromy ) : their logarithmic derivative @xmath550 is a meromorphic @xmath19-form on @xmath0 having exactly @xmath161 simple poles , one at @xmath164 with residue @xmath330 and one at @xmath546 with residue @xmath105 . in other words ,
the riccati equation defining the singular foliation after trivialization of the bundle is @xmath551 indeed , the @xmath19-form @xmath552 has a simple pole at @xmath525 with residue @xmath330 since its principal part is given by @xmath553 and @xmath554 has a double pole at @xmath525 ; the other poles may come from the two zeroes of @xmath555 , namely @xmath556 , but @xmath557 is actually regular since the numerator @xmath558 also vanishes at this point : by residue theorem , @xmath538 is a simple pole with residue @xmath105 .
of course , any other @xmath19-form having the same principal part must differ by a holomorphic @xmath19-form , namely @xmath559 , @xmath111 .
we have omited from our discussion the case @xmath560 is an order @xmath161 point which can be treated like @xmath525 .
after two elementary transformations of @xmath561 centered at the points @xmath562 and @xmath563 , we obtain by this way all ( linear ) foliations on the bundle @xmath564-[0 ] ) $ ] while @xmath136 runs over @xmath16 .
this does not provide yet a universal family for linear connections on @xmath0 since the limit of the riccati foliation while @xmath565 is the vertical fibration : for @xmath566 in a compact set not intersecting @xmath567 , @xmath535 and @xmath536 , we have @xmath568 in other words , the @xmath19-form @xmath569 tends uniformly to @xmath570 on the compact set , so does the foliation .
we would like to complete this @xmath16-bundle over @xmath532 with the family @xmath571 , @xmath572 , of linear connections on the trivial bundle @xmath573 ( @xmath574 ) .
a way to obtain it from our large family is obviously to set @xmath575 and take the limit while @xmath565 with @xmath572 fixed . in other words , in the parameter space @xmath576 we consider only the limit at @xmath577 while @xmath565 with a special direction .
the good global parameter space is obtained after separating the germs of curves @xmath578 .
this is done after @xmath161 elementary transformations on @xmath544 : first we blow - up @xmath579 by setting @xmath580 , and then we blow - up @xmath581 by setting @xmath582 , so that @xmath583 coincides with the expected parameter @xmath584 .
the resulting parameter space is the * affine bundle * @xmath585 where @xmath292 is the unique @xmath164-section of the undecomposable bundle @xmath327 .
now , we construct a fine moduli space as follows .
consider the product @xmath586 with global coordinates @xmath587 , and equipp the bundle over @xmath576 with the riccati foliation @xmath551 this can be seen as an algebraic foliation on the total space .
now , apply the elementary transformations with center along the surfaces @xmath588 and @xmath589 .
then , we modify the base @xmath590 by two elementary transformations so that we obtain @xmath327 as a base and the foliation extends as a linear connections all along @xmath574 .
the euclidean connections on @xmath327 are given by : @xmath591 indeed , one can check that the reduction of the singularity over @xmath525 yields @xmath327 ; on the other hand , it is clear that monodromy is given by translations .
this can be obtained also as a limit of our previous family of connections , or better from @xmath592 which is equivalent to the previous one by the change of coordinate @xmath593 .
now , instead of taking limit along curves @xmath575 as @xmath594 with @xmath584 constant , we take limit along @xmath595 , @xmath596 constant , i.e. @xmath597 .
we have on convenient compact sets : @xmath598 so that @xmath599 the space of linear connections on @xmath0 is an affine @xmath16-bundle over @xmath601 that we have identified with @xmath602 : it is defined by gluing the chart @xmath603 with the chart @xmath604 by the transition map @xmath605 the space of linear representations of @xmath219 is @xmath606 . in the main chart @xmath576 ,
the analytic connection is given by @xmath607 introducing weierstrass zeta function @xmath608 , one can write @xmath609 and integrate the differential equation above by means of the weierstrass sigma function : the general solution is therefore given by @xmath610 , @xmath117 , and the monodromy is given by the homomorphism @xmath611 finally we obtain the full monodromy mapping @xmath612 as a particular case of narasimhan - seshadri theorem , the unitary representations @xmath614 form a smooth real @xmath161-dimensional torus transversal to the foliation and cutting each leaf once .
it is the space of the leaves .
it inherits , from the transversal complex foliation , a complex structure , namely the structure of @xmath0 .
the euclidean foliations defined on @xmath602 by @xmath615 , @xmath596 , are sent to the linear foliations @xmath616 , @xmath617 .
the space of linear connections is equipped with a group law given by tensor product ; it is just the pull - back of the natural group law on @xmath606 .
we thus get an analytic isomorphism between two algebraic groups that are not algebraically equivalent . one can compute the group law on @xmath602 as follows .
given two connections @xmath618 the tensor product is a connection of the form @xmath619 with @xmath620 ( group law on @xmath332 ) .
then , @xmath621 is determined by the fact that @xmath622 for a meromorphic function @xmath63 on @xmath0 .
looking at the principal part of the left hand side , one see that @xmath63 must have divisor @xmath623+[u_2]-[u_3]-[0]$ ] so that , up to a scalar , we have @xmath624 after computations , one finds that @xmath625 m. brunella , _ birational geometry of foliations _ , available electronically at http://www.impa.br/publicacoes/monografias/abstracts/brunella.ps .
monografas de matemtica .
instituto de matemtica pura e aplicada ( impa ) , rio de janeiro , ( 2000 ) .
r. c. gunning , _
analytic structures on the space of flat vector bundles over a compact riemann surface _ , several complex variables , ii ( proc .
maryland , college park , md . , 1970 ) , pp .
lecture notes in math .
185 , springer , berlin , 1971 .
r. c. gunning , _ affine and projective structures on riemann surfaces _ , riemann surfaces and related topics : proceedings of the 1978 stony brook conference , annals of math . studies n. 97 , 225244 ( 1981 ) .
n. m. katz , _ an overview of deligne s work on hilbert s twenty - first problem _ , mathematical developments arising from hilbert problems .
proceedings of symposia in pure mathematics , vol .
american mathematical society , providence , r. i. , ( 1976 ) , 537557 . | a projective structure on a compact riemann surface @xmath0 of genus @xmath1 is given by an atlas with transition functions in @xmath2 .
equivalently , a projective structure is given by a @xmath3-bundle over @xmath0 equipped with a section @xmath4 and a foliation @xmath5 which is both transversal to the @xmath3-fibers and the section @xmath4 . from this latter geometric bundle picture
, we survey on classical problems and results on projective structures . by the way
, we will recall the analytic classification of @xmath3-bundles .
we will give a complete description of projective ( actually affine ) structures on the torus with an explicit versal family of foliated bundle picture . |
_ ab - initio _ hartree - fock ( hf ) and configuration interaction ( ci ) methods are standard tools in computational chemistry nowadays and various program packages are available for accurate calculations of properties of atoms and molecules
. for solids , hf calculations have become possible , on a broad scale , with the advent of the program package crystal @xcite . however , the problem of an accurate treatment of electron correlation is not fully settled ( for a survey see @xcite ) .
although the absolute value of the hf energy is usually much larger than the correlation energy , the correlation energy is very important for energy differences .
for example , the o@xmath1 ion is not stable at the hf level , and correlations are necessary in order to obtain even qualitative agreement with the experimental result for the electron affinity of oxygen . in solid state physics , nio is a well known example of a system which is insulating due to correlations .
the most widely used method to include correlations in solids is density functional theory ( dft ) @xcite .
dft has also recently become quite popular for a computationally efficient treatment of exchange and correlation in molecules .
however , a systematic improvement towards the exact results is currently not possible with dft .
wave - function - based methods are more suitable for this purpose . in the last years , quantum monte - carlo calculations have been performed for several systems @xcite .
correlations are included here by multiplying the hf wavefunction with a jastrow factor .
an approach more closely related to quantum chemistry is the local ansatz @xcite where judiciously chosen local excitation operators are applied to hf wavefunctions from crystal calculations .
some years ago , an incremental scheme has been proposed and applied in calculations for semiconductors @xcite , for graphite @xcite and for the valence band of diamond @xcite ; here information on the effect of local excitations on solid - state properties is drawn from calculations using standard quantum chemical program packages . in a recent paper @xcite we showed that this method can be successfully extended to ionic solids ; we reported results for the correlation contribution to the cohesive energy of mgo . in the present article
, we apply the scheme to the cohesive energy of cao , as a second example . in addition , we show how correlations affect the lattice constants of mgo and cao . for these systems ,
several calculations have been performed at the hf level with the crystal code @xcite as well as with inclusion of correlations using dft @xcite .
the method of increments can be used to build up correlation effects in solids from local correlation contributions which in term may be obtained by transferring results from suitably embedded finite clusters to the infinite crystal .
it has been fully described in @xcite , and a formal derivation has been given within the framework of the projection technique @xcite .
thus , we will only briefly repeat the main ideas .
+ ( a ) starting from self - consistent field ( scf ) calculations localized orbitals are generated which are assumed to be similar in the clusters and in the solid .
+ ( b ) one - body correlation - energy increments are calculated : in our specific case these are the correlation energies @xmath2 , @xmath3 , @xmath4 , ... of localized orbital groups which can be attributed to x@xmath5 ( x = mg , ca ) or o@xmath6 ions at ionic positions @xmath7 , @xmath8 , @xmath9 , ... each localized orbital group is correlated separately . + ( c ) two - body increments are defined as non - additivity corrections : @xmath10 where @xmath11 is the correlation energy of the joint orbital system of @xmath12 .
+ ( d ) three - body increments are defined as @xmath13 - \nonumber \\ & [ \delta\epsilon(ab ) + \delta\epsilon(ac ) + \delta\epsilon(bc)].\end{aligned}\ ] ] similar definitions apply to higher - body increments .
+ ( e ) the correlation energy of the solid can now be expressed as the sum of all possible increments : @xmath14 of course , this only makes sense if the incremental expansion is well convergent , i.e. if @xmath15 rapidly decreases with increasing distance of the ions at position @xmath7 and @xmath8 and if the three - body terms are significantly smaller than the two - body ones .
a pre - requisite is that the correlation method used for evaluating the increments must be size - extensive : otherwise the two - body increment @xmath15 for two ions @xmath7 and @xmath8 at infinite distance would not vanish . in our present work , we used three different size - extensive approaches , cf .
2.1 . finally , the increments must be transferable , i.e. they should only weakly depend on the cluster chosen . in this section we want to give a brief description of the correlation methods used . in the averaged coupled - pair functional ( acpf @xcite ) scheme ,
the correlation energy is expressed in the form @xmath16=\frac { < \psi_{scf}+\psi_c|h - e_{scf } |\psi_{scf}+\psi_c > } { 1+g_c<\psi_c|\psi_c>}\ ] ] with @xmath17 being the scf - wavefunction ( usually of the spin - restricted hartree - fock type ) and @xmath18 the correlation part of the wavefunction , @xmath19 @xmath20 is chosen as @xmath21 in order to make the expression ( 2 ) approximately size - consistent ( @xmath22 being the number of correlated electrons ) . for more details ( and the extension to multi - reference cases ) , see @xcite .
+ in the coupled - cluster singles and doubles ( ccsd @xcite ) scheme , the wavefunction is expressed with the help of an exponential ansatz : @xmath23 @xmath24 ( @xmath25 ) are creation ( annihilation ) operators of electrons in orbitals which are occupied ( @xmath25 , @xmath26 ) or unoccupied ( @xmath27 @xmath28 ) in the scf wavefunction .
+ finally , in the ccsd(t ) scheme , three particle excitations are included by means of perturbation theory as proposed in @xcite .
+ we used these three methods to compare their quality in applications to solids .
it turns out that acpf and ccsd give very similar results , while ccsd(t ) yields slighty improved energies @xcite .
altogether , the results are not strongly dependent on the methods and no problem arises , therefore , if only one method should be applicable ( as is the case for low - spin open - shell systems , where ccsd and ccsd(t ) are not yet readily available ) .
all calculations of this work were done by using the program package molpro @xcite .
for calculating the correlation contribution to the cohesive energy of cao , we closely follow the approach of ref . @xcite . for oxygen we choose a @xmath29 $ ] basis set @xcite .
calcium is described by a small - core pseudopotential replacing the @xmath30 , @xmath31 and @xmath32 electrons @xcite , and a corresponding @xmath33 $ ] valence basis set ( from ref .
@xcite , augmented with polarization functions @xmath34=0.863492 , @xmath35=2.142 and @xmath36=1.66 ) is used .
all orbitals are correlated , with the exception of the o @xmath30 core . in particular ,
the correlation contribution of the outer - core ca @xmath37 and @xmath38 orbitals is explicitly taken into account .
we did not use a large - core ( x@xmath5 ) pseudopotential and a core polarization potential ( cpp ) for treating core - valence correlation as was done in the case of mgo @xcite , because ca is close to the transition metals and excitations into @xmath39-orbitals are important .
the influence of the latter on the x@xmath5 core can not be well represented by a cpp since the @xmath40 orbitals are core - like themselves ( cf . the discussion in @xcite ) .
correlating the ca outer - core orbitals explicitly , using the small - core ( ca@xmath41 ) pseudopotential , we circumvent this problem . using this approach , we performed test calculations for the first and second ionization potential of the ca atom ( table 1 ) and calculated spectroscopic properties of the cao molecule ( table 2 ) . in both cases , we obtain good agreement with experiment .
we first calculated one - body correlation - energy increments .
for ca@xmath5 , the results are virtually independent of the solid - state surroundings .
this was tested by doing calculations for a free @xmath42 and a @xmath42 embedded in point charges .
( a cube of 7x7x7 ions was simulated by point charges @xmath432 , with charges at the surface planes / edges / corners reduced by factors 2/4/8 , respectively . ) in the case of @xmath44 , of course , the solid - state influence is decisive for stability , and we took it into account by using an embedding similar to that of ref.@xcite : the pauli repulsion of the 6 nearest @xmath42 neighbours was simulated by large - core pseudopotentials @xcite , while the rest of a cube of 7x7x7 lattice sites was treated in point - charge approximation again .
a nacl - like structure with a lattice constant of 4.81 was adopted .
( the experimental value for the lattice constant is 4.8032 at a temperature of t=17.9 k @xcite ) .
we performed similar calculations for various other finite - cluster approximations of the cao crystal , in order to insure that the results are not sensitive to lattice extensions beyond the cube mentioned above .
the results for the one - body correlation - energy increments are shown in table 3 .
it is interesting to note that the absolute value of the ca @xmath45 increment is larger than in the case of mg , although the electron density in the valence region of ca is lower than for mg .
the larger correlation contribution for ca can be rationalized by the fact that excitations into low - lying unoccupied @xmath39-orbitals are much more important for ca than for mg .
this is a result which would be difficult to explain by density functional theory : in a local - density framework , higher density leads to a higher absolute value of the correlation energy . in ref .
@xcite we argued that the increment in correlation energy @xmath46
@xmath47 @xmath48 is not just twice the increment @xmath49 @xmath50 @xmath51 .
however , comparing the increments @xmath52 ( embedded @xmath53 @xmath51 and @xmath52 ( embedded @xmath54 @xmath48 one finds a factor very close to two .
this can be seen from table 4 where we compare the increments in the case of mgo .
thus , for the embedded species linear scaling is appropriate as in the case of the gas - phase isoelectronic series ne , ne@xmath55 , ne@xmath5 : there , the increments in correlation energy are 0.0608 h ( ne@xmath5 @xmath56 ne@xmath55 ) and 0.0652 h ( ne@xmath57 @xmath56 ne ) @xcite .
table 4 also shows that the correlation contribution to the electron affinity of the oxygen atom is _ smaller _ for the embedded species than in the gas phase .
this is due to the fact that energy differences to excited - state configurations become larger when enclosing o@xmath58 in a solid - state cage .
once again , this is at variance with a lda description as the electron density in the case of the embedded @xmath59 is more compressed than in the case of a free @xmath59 . in figure 1
we show the charge density distribution of o@xmath6 , again in the case of mgo .
we used basis functions on both o and mg ; the mg @xmath30 , @xmath31 and @xmath32-electrons are replaced by a pseudopotential .
one recognizes the minimum near the mg@xmath5 cores , where the pauli repulsion prevents the oxygen electrons from penetrating into the mg@xmath5 core region .
this way , the solid is stabilized .
the 6@xmath60 contour line , counting from mg to o , is the line which represents a density of 0.002 a.u .
this is the density which encloses about 95 % of the charge and was proposed as an estimate of the size of atoms and molecules @xcite .
the sum of the intra - ionic correlation - energy increments discussed in this subsection turns out to yield only @xmath61 60 % of the correlation contribution of the cohesive energy of cao .
this percentage is quite similar to that obtained for mgo @xcite , at the same level .
thus , although mgo and cao are to a very good approximation purely ionic solids , the inter - atomic correlation effects to be dealt with in the next subsection play an important role . when calculating two - body correlation - energy increments , point charges or pseudopotentials surrounding a given ion have to be replaced by real ions . in the case of an additional real @xmath44 , its next - neighbour shell also has to be replaced by a cage of pseudopotentials simulating @xmath62 . this way the increments shown in table 3 are obtained .
it turns out that the ca - o increments are much more important than the o - o increments , while ca - ca increments are negligibly small .
the changes with respect to mgo @xcite can easily be rationalized : on the one hand , the lattice constant is larger than in the case of mgo ( 4.81 vs. 4.21 ) , which reduces the van der waals interaction and makes the o - o increments smaller . on the other hand ,
the polarizability of @xmath62 is by a factor of more than 6 higher than that of @xmath63 ( see for example ref .
@xcite ) , which leads to large ca - o increments .
we show the van der waals - like decay in figure 2 by plotting the two - body increments o - o for cao from ccsd calculations ( without including weight factors ) . by multiplying with the sixth power of the distance
, one can verify the van der waals - law .
plots for the other two - body increments are qualitatively similar .
three - body increments contribute with less than 2 % to the correlation piece of the bulk cohesive energy and may safely be neglected , therefore .
a survey of the convergency pattern of the incremental expansion , for both cao and mgo , is given in figs .
3 and 4 .
adding up the increments of sections 3.2 and 3.3 ( cf . table 5 ) , we obtain between 71 and 78 % of the experimental correlation contribution to the cohesive energy which we define as the difference of the experimental cohesive energy ( 11.0 ev , @xcite ) plus the zero - point energy ( which is taken into account within the debye approximation and is of the order 0.1 ev ) minus the hf binding energy ( 7.6 ev , ref .
the percentage obtained is slightly less compared to the case of mgo @xcite where 79 to 86 % were recovered .
one of the reasons for this difference is that we used a cpp in the case of mgo which covers nearly 100% of the core - valence correlation contributions in mg , while the explicit treatment of that correlation piece for ca was less exhaustive .
another reason is that on the hartree - fock level @xmath64-functions for ca ( which are not yet implemented in crystal ) would probably increase the cohesive energy and lower the experimental correlation contribution . finally , as in the case of mgo
, a significant part of the missing correlation energy should be due to basis set errors for the o atom .
the total cohesive energy recovered in our calculations is in the range of between 91 and 93 % of the experimental value .
our results are compared in table 5 to those from density functional calculations .
we choose the results from @xcite where a correlation - only functional was used and 0.078 to 0.097 h of the correlation contribution to the cohesive energy were obtained , depending on the specific correlation functional used .
at the hartree - fock level , the lattice constant is in good agreement with experiment for mgo @xcite , whereas there is a deviation of 0.05 in the case of cao @xcite .
it is interesting , therefore , to study the influence of correlation effects on lattice constants . in tables 6 and 7
, we give the necessary increments for mgo and cao , respectively .
we find two main effects of correlations . on the one hand
, the van der waals - interaction leads to a reduction of the lattice spacing since the attractive interaction is of the form @xmath65 and obviously stronger at shorter distance . on the other hand
, we find that the intra - ionic correlation of the @xmath66-ion forces a larger constant . this can be understood from the argument that excited configurations are lower in energy and mix more strongly with the ground - state determinant if the @xmath66 is less compressed as explained in section 3.2 .
adding up all these contributions ( cf .
table 8) , they are found to nearly cancel in the case of mgo and to lead to a reduction of only 0.01 . for obtaining this result , we applied a linear fit to the correlation energy and superimposed it on the hf potential curve of refs .
we checked the validity of the linear approximation by calculating selected increments at other lattice constants . in the case of cao ,
the van der waals - interaction is more important and the lattice constant is reduced to 4.81 which is in nice agreement with the experimental value .
the lattice constants seem to be in better agreement with the experimental values than those calculated from density functional theory for mgo @xcite and cao @xcite , where deviations of @xmath43 2 % are found .
this is similar to earlier findings for semiconductors @xcite .
we determined the correlation contribution to the cohesive energy of cao using an expansion into local increments recently applied to mgo . making use of quantum - chemical _ ab - initio _ configuration - interaction calculations for evaluating individual increments , we obtain @xmath61 80 % of the expected value .
the missing energy is probably mainly due to the lack of @xmath36 and higher polarization functions in our one - particle basis set .
the computed lattice constants show deviations of less than 1% from the experimental values .
we found two correlation effects on the lattice constants : the inter - atomic van der waals - force leads to a reduction , whereas intra - atomic correlations of the @xmath66 ions lead to an increase of the lattice constant .
the main difference between cao and mgo is the reduced importance of the inter - atomic o - o correlations in cao ( due to the larger lattice constant ) and the higher importance of the ca - o correlations ( due to the higher polarizability of ca@xmath5 ) .
compared to dft , the numerical effort of our scheme is significantly higher .
however , we feel that the advantage of the present approach is the high quality and stability of the results both for atoms , ions as well as for solids .
another advantage is the possibility of a systematic improvement by using larger basis sets .
we think that the method of local increments is capable now of being routinely applied to ionic systems , and a systematic study on alkali halides is underway .
an extension to open - shell systems such as nio is also a project currently under investigation .
we would like to thank prof .
p. fulde for supporting this work and prof .
w. c. nieuwpoort ( groningen ) for interesting suggestions .
we are grateful to prof .
werner ( stuttgart ) for providing the program package molpro .
99 r. dovesi , c. pisani , and c. roetti , int .
j. quantum chem . *
17 * , 517 ( 1980 ) ; c. pisani , r. dovesi , and c. roetti , lecture notes in chemistry , vol . 48 ( springer , berlin , 1988 ) p. fulde in _ electron correlations in molecules and solids _ , springer series in solid state sciences , vol . 100 ( springer , berlin , 1993 ) for a review about dft , see , e.g. r. o. jones , o. gunnarsson , rev .
* 61 * , 689 ( 1989 ) s. fahy , x. w. wang , s. g. louie , phys
. rev . lett . * 61 * , 1631 ( 1988 ) ; s. fahy , x. w. wang , s. g. louie , phys .
b * 42 * , 3503 ( 1990 ) ; x .- p .
li , d. m. ceperley , r. m. martin , phys .
b * 44 * , 10929 ( 1991 ) ; l. mits , r. m. martin , phys . rev
72 * , 2438 ( 1994 ) g. stollhoff , p. fulde , j. chem
. phys . * 73 * , 4548 ( 1980 ) h. stoll , phys . rev .
b * 46 * , 6700 ( 1992 ) ; h. stoll , chem .
lett . * 191 * , 548 ( 1992 ) b. paulus , p. fulde , h. stoll , phys .
b * 51 * , 10572 ( 1995 ) h. stoll , j. chem . phys . *
97 * , 8449 ( 1992 ) j. grfenstein , h. stoll , p. fulde , chem .
. lett . * 215 * , 611 ( 1993 ) k. doll , m. dolg , p. fulde , h. stoll , phys .
b * 52 * , 4842 ( 1995 ) m. caus , r. dovesi , c. pisani , c. roetti , phys . rev .
b * 33 * , 1308 ( 1986 ) r. dovesi , c. roetti , c. freyria - fava , e. apr , v.r .
saunders , and n.m .
harrison , philos .
london ser .
a * 341 * , 203 ( 1992 ) ; w. c. mackrodt , n. m. harrison , v. r. saunders , n. l. allan , m. d. towler , e. apr , r. dovesi , philos . mag .
a * 68 * , 653 ( 1993 ) m. caus , a. zupan , chem .
. lett . * 220 * , 145 ( 1994 ) m. i. mccarthy , n. m. harrison , phys .
b * 49 * , 8574 ( 1994 ) m. catti , g. valerio , r. dovesi , m. caus , phys . rev .
b * 49 * , 14179 ( 1994 ) n. c. pyper , phil .
london a * 352 * , 89 ( 1995 ) t. schork , thesis , university stuttgart ( 1992 ) r. j. gdanitz , r. ahlrichs , chem
. phys . lett . *
143 * , 413 ( 1988 ) f. koester , h. kmmel , nucl .
* 17 * , 477 ( 1960 ) ; j. ek , j. chem . phys . * 45 * , 4256 ( 1966 ) ; j. paldus , j. ek , phys .
a * 5 * , 50 ( 1972 ) ; r. f. bishop , theor .
acta * 80 * , 95 ( 1991 ) ; g. d. purvis iii , r. j. bartlett , j. chem .
* 76 * , 1910 ( 1982 ) k. raghavachari , g. w. trucks , j. a. pople , m. head - gordon , chem .
. lett . * 157 * , 479 ( 1989 ) it was pointed out by h .- j .
werner that ccsd(t ) as implemented in ref .
23 can not be applied to localized orbitals but only to canonical ones ( see m. j. o. deegan , p. j. knowles , chem .
. lett . * 227 * , 321 ( 1994 ) ) . in our scheme localization is necessary , but we can estimate the magnitude of the concomitant errors by calculating two - body increments using _ different _ clusters which makes localization unnecessary : a cluster with two oxygen atoms , e.g. , in order to get @xmath11 and a cluster with one oxygen atom to get @xmath67 .
we compared the increments @xmath68 calculated by using localized / canonical orbitals and found that the error for the most important increment is less than 0.01 ev including the weight factor .
the acpf and ccsd results are not affected from this problem .
molpro is a package of _ ab initio _ programs written by h .- j .
werner and p.j .
knowles , with contributions from j. almlf , r.d .
amos , m.j.o .
deegan , s.t .
elbert , c. hampel , w. meyer , k. peterson , r. pitzer , a.j . stone , and p.r .
taylor ; the cpp program was written by a. nicklass .
werner and p.j .
knowles , j. chem .
* 82 * , 5053 ( 1985 ) ; p.j .
knowles and h .- j .
werner , chem .
phys . lett . * 115 * , 259 ( 1985 ) ; h .- j .
werner and p.j .
knowles , j. chem .
* 89 * , 5803 ( 1988 ) ; p.j .
knowles and h .- j .
werner , chem .
. lett . * 145 * , 514 ( 1988 ) ; h .- j .
werner and p.j .
knowles , theor .
acta * 78 * , 175 ( 1990 ) ; c. hampel , k. peterson , and h .- j .
werner , chem .
lett . * 190 * , 1 ( 1992 ) ; p.j .
knowles , c. hampel , and h .- j .
werner , j. chem .
* 99 * , 5219 ( 1993 ) + t.h . dunning , jr . ,
j. chem . phys . * 90 * , 1007 ( 1989 ) m. kaupp , p. v. r. schleyer , h. stoll , h. preuss , j. chem . phys . * 94 * , 1360(1991 ) w. mller , j. flesch , and w. meyer , j. chem . phys .
* 80 * , 3297 ( 1984 ) p. fuentealba , l. v. szentply , h. preuss and h. stoll , j. phys .
b * 18 * , 1287 ( 1985 ) k .- h .
hellwege and a. m. hellwege ( eds . ) , landolt - brnstein tables , group iii , vol .
7b1 ( springer , berlin , 1975 ) e.r .
davidson , s.a .
hagstrom , s.j .
chakravorty , v.m .
umar , and ch .
froese - fischer , phys .
a * 44 * , 7071 ( 1991 ) r. f. bader , w. h. henneker , p. e. cade , j. chem .
46 * , 3341 ( 1967 ) crc handbook of chemistry ad physics , 75@xmath60 edition , editor : david r. lide ( crc press , boca raton , 1994/1995 ) c. e. moore , atomic energy levels , nsrds - nbs 35 / vol .
i - iii , nat .
standards ( washington , dc ) k. p. huber , g. herzberg , molecular spectra and molecular structure : iv .
constants of diatomic molecules ( van nostrand , new york , 1979 ) j. a. irvin , p. j. dagdigian , j. chem .
phys . * 73 * , 176 ( 1980 ) | a recently proposed computational scheme based on local increments has been applied to the calculation of correlation contributions to the cohesive energy of the cao crystal . using _ ab - initio _ quantum chemical methods for evaluating individual increments , we obtain @xmath0 of the difference between the experimental and hartree - fock cohesive energies .
lattice constants corrected for correlation effects deviate by less than 1% from experimental values , in the case of mgo and cao .
accepted by phys .
rev . |
diffractive interactions are characterized by the presence of at least one non - exponentially suppressed large rapidity gap ( lrg ) in the final state , defined as a region in pseudorapidity devoid of particles .
the origin of lrgs is attributed to a color - singlet exchange with vacuum quantum numbers , referred to as pomeron ( @xmath0 ) exchange . soft diffractive interactions ( with no hard scale )
can not be calculated within perturbative qcd ( pqcd ) , and traditionally have been described by models based on regge theory .
model predictions generally differ when extrapolated from pre - lhc energies ( @xmath1 tev ) to 7 tev at lhc .
thus , measurements of diffractive cross sections at the lhc provide a valuable input for understanding diffraction and improving the modeling of diffraction in current event generators .
the first cms measurements of single - diffractive ( sd ) , double - diffractive ( dd ) , and forward rapidity gap cross are presented .
results are based on 2010 data when the lhc was running in a low pile - up scenario , most suitable for event selection based on the lrg signature .
r0.5 diffractive cross sections have been measured @xcite based on a minimum - bias sample at @xmath2 tev .
the sd and dd events ( fig . [ fig1 ] ) were separated using the castor calorimeter , which covers the very forward region of the experiment , -6.6 @xmath3 -5.2 .
minimum - bias events were selected by requiring a signal above noise level in any of the bsc ( beam scintillator counter ) devices ( @xmath4 ) and the presence of at least two energy deposits in the central detector ( @xmath5 ) .
no vertex requirement was imposed .
diffractive events were selected by requiring the presence of a forward rapidity gap reconstructed at the edge of the central detector or central gap .
the forward gap on the positive ( negative ) side was reconstructed in terms of the variable @xmath6 ( @xmath7 ) defined as the highest ( lowest ) @xmath8 of a particle reconstructed in the detector .
the central gap was reconstructed as @xmath9 , with @xmath10 ( @xmath11 ) defined as the closest - to - zero @xmath8 of a particle reconstructed on the positive ( negative ) @xmath8-side of the central detector , with an additional requirement of activity on both sides of the detector .
figure [ fig2 ] shows distributions of @xmath6 , @xmath7 and @xmath12 compared to predictions of pythia8-mbr @xcite .
diffractive events appear as a flattening of the exponentially falling distributions of non - diffractive ( nd events with rapidity gaps due to random multiplicity fluctuations , and dominate the regions of low @xmath6 , high @xmath7 , or high @xmath12 . the event sample after the @xmath13 selection ( fig
. [ fig2 ] , middle ) was used to extract sd and dd cross sections .
the sample consists of approximately equal numbers of sd and dd events for which one of the dissociated masses is low and escapes detection in the central detector .
subsamples enhanced in sd and dd events were selected by requiring an absence or a presence of an energy deposit ( above threshold ) in the castor calorimeter . the sd / dd separation with castor
is presented in fig .
[ fig3 ] , showing the distribution of the variable @xmath14 calculated from all energy deposits in the detector . for sd events
, @xmath14 approximates the incoming - proton momentum loss .
these distributions were used to measure the differential sd cross section as a function of @xmath14 , and the differential dd cross section as a function of @xmath15 for @xmath16 ( castor acceptance ) , after subtracting the background contribution to the signal ( dd to sd and nd to dd ) .
results are compared to mc models in figs .
[ fig4 ] ( left ) and ( middle ) , respectively .
the predictions of pythia8-mbr are shown for two values of the @xmath17 parameter of the pomeron trajectory ( @xmath18 ) , @xmath19 and @xmath20 .
both values describe the measured sd cross section within uncertainties , while the dd data favor the smaller value of @xmath17 .
the predictions of pythia8 - 4c and pythia6 describe well the measured dd cross section , but fail to describe the falling behavior of the data ( see details in @xcite ) . the total sd cross cross section integrated over the region @xmath21 ( @xmath22 gev )
was measured to be @xmath23 mb ( dissociation of either proton ) .
the event sample after the @xmath24 selection ( fig . [ fig2 ] , right ) was used to extract the differential dd cross section as a function of the central - gap width , @xmath25 . the cross section for @xmath26 , m@xmath27 gev and m@xmath28 gev
is presented in fig .
[ fig3 ] ( right ) .
the total dd cross cross section integrated over this region was measured to be @xmath29 mb . , and dd cross section vs @xmath25 ( right ) , compared to pythia6 , pythia8 - 4c and pythia8-mbr mc .
errors are dominated by systematic uncertainties ( hf calorimeter energy scale , and hadronization and diffraction model).,title="fig:",scaledwidth=66.0% ] , and dd cross section vs @xmath25 ( right ) , compared to pythia6 , pythia8 - 4c and pythia8-mbr mc .
errors are dominated by systematic uncertainties ( hf calorimeter energy scale , and hadronization and diffraction model).,title="fig:",scaledwidth=33.0% ] the inclusive differential cross section @xmath30 for events with a forward rapidity gap was measured , with @xmath31 defined as the larger of gaps within @xmath32 .
a bayesian - unfolded and fully corrected cross section for particles with @xmath33 mev and @xmath34 was extracted and compared to pythia6-z2 * @xcite , pythia8 - 4c , and pythia8-mbr ( @xmath19 and @xmath35 ) predictions .
figure [ fig5 ] shows a comparison to the atlas measurement @xcite .
the cms measurement extends the atlas result by 0.4 units of gap width .
the two results are in agreement within the uncertainties , which at cms are dominated by hf energy scale , hadronization , and diffraction model used .
k. goulianos , _ hadronic diffraction : where do we stand ? _ , arxiv : hep - ph/0407035 ( 2004 ) ; k. goulianos , _ diffraction in qcd _ , arxiv : hep - ph/0203141 ( 2002 ) ; k. goulianos , _ a new statistical description of hadronic and @xmath37 multiplicities _ , phys .
b193 * 151 ( 1987 ) . | we present measurements of soft single- and double - diffractive cross sections , as well as of forward rapidity gap cross sections at 7 tev at the lhc , and compare the results to other measurements and to theoretical predictions implemented in various monte carlo simulations . |
continuous quantum phase transitions ( qpt ) and behaviour of quantum systems in vicinity of the corresponding quantum critical points attract a great research interest@xcite both in theory and in experiment .
there are manifestations of such effects in cuprate superconductors , in iron pnictides and in other materials .
qpt are most pronounced in low dimensional systems , but they also occur in three dimensional ( 3d ) systems . in the present paper we consider quantum systems
, therefore 3d means 3d+time . a well known material that manifests the 3d magnetic quantum critical behaviour is tlcucl@xmath1 . magnetic and other properties of the compound has been extensively studied experimentally .
@xcite under normal conditions the material is nonmagnetic , while a magnetic field and/or pressure drives a qpt to the magnetically ordered phase .
an analogous material kcucl@xmath1 have similar properties .
@xcite it has been understood that tlcucl@xmath1 consists of a 3d set of spin dimers and the qpt in the magnetic field can be considered as bose condensation of the spin dimer triplet excitations ( triplons ) .
@xcite the pressure induced qpt can be also described in terms of triplons.@xcite inelastic neutron scattering from the magnetically ordered phase in the vicinity of the pressure induced qpt @xcite clearly indicates two transverse and one longitudinal spin - wave excitation modes as one would expect for condensed triplons . according to the data @xcite the longitudinal spin - wave excitation has a finite width .
generically this is also expected because of the possibility for the longitudinal mode to decay to two transverse spin - wave excitations .
however to the best of our knowledge this width has never been calculated . in the present work
we perform such a calculation and compare the result with the experimental data .
the width is related to theparameters of the bose condensate of magnons observed in the same compound in the external magnetic field .
on the technical side in the present work we use the effective quantum field theory approach .
as we pointed out , a description of the system in terms of the spin dimer triplon technique is also possible . the spin dimer technique accounts for both the short range and the long range dynamics on equal footing .
on the one hand , this is a strength of this description . on the other hand ,
the description is quite technically involved if one needs to analyze the interaction effects that we are after .
the technical complexity is a price for the universality .
the effective quantum field theory method adopted in the present work , allows us to account for the interaction effects much more accurately .
the method does not account for the short range dynamics , but the short range effects are irrelevant in the vicinity of a qpt .
we would like also to note that in the present paper we use terms `` spin waves '' and `` magnons '' on equal footing , having in mind magnetic excitations on both sides of the qpt .
the qpt that we consider here is a continuous zero temperature transition .
therefore , the short range dynamics are not changed in the qpt , only the long range dynamics are changed . in this situation
it is quite natural to apply the effective field theory approach to describe the transition @xcite .
a similar approach was used to analyse the longitudinal mode in a quasi - one dimensional antiferromagnet.@xcite technically this approach is much simpler than the triplon description . the effective field theory must be written in terms of the vector field @xmath2 that describes staggered magnetization as well as the long - wave - length magnetic excitations .
there is no net magnetization on both sides of the qpt , only the staggered magnetization , hence the effective landau - ginzburg lagrangian corresponding to the nonlinear @xmath3-model describes the system , @xmath4\right)^2- \frac{c^2}{2}\left({\bm \nabla}{\vec \varphi}\right)^2 -\frac{1}{2}m^2{\vec \varphi}^2 -\frac{1}{4}\alpha[{\vec \varphi}^2]^2\nonumber\\\end{aligned}\ ] ] we note the standard logic for this effective lagrangian .
( i ) the elastic energy must be proportional to @xmath5 ( stiffness ) .
( ii ) the system consists of spins with the same gyromagnetic ratio @xmath6 . therefore the larmor theorem must be valid : if there is a state of the system without magnetic field , then the same state in a uniform magnetic field @xmath7 must precess with frequency @xmath8 ( hereafter we set @xmath9 , where @xmath10 is bohr magneton . ) ( iii ) the static energy can depend on the magnetic field only quadratically because the system has no net magnetization .
note that instead of the hard constraint , @xmath11 , usually used in the nonlinear @xmath3-model , we impose the `` soft '' quartic interaction @xmath12 ^ 2 $ ] .
the qpt is due to the mass term , @xmath13 in tlcucl@xmath1 the qpt is driven by external hydrostatic pressure @xmath14 and this is reflected in eq.([m ] ) , @xmath15 is just a coefficient . in a more general situation pressure
has to be replaced by a generalized `` coupling constant '' . when @xmath16 and there is no magnetic field , @xmath17 , the mass squared is positive and this corresponds to the magnetically disordered phase with the gapped triple degenerate spin - wave dispersion @xmath18 .
when @xmath19 the mass squared is negative and this results in a nonzero expectation value of @xmath20 that describes the spontaneous antiferromagnetic ordering with a gapped longitudinal mode and two gapless goldstone modes .
since we will compare our results with experimental data for tlcucl@xmath1 we need to determine the corresponding spin - wave velocity @xmath22 .
tlcucl@xmath1 has a monoclinic structure @xcite with a unit cell very close to a rectangular prism with lattice parameters @xmath23 , @xmath24 , @xmath25 and an angle @xmath26 only slightly deviating from a right angle ( @xmath27 ) .
the spin wave dispersion over the brillouin zone was determined by inelastic neutron scattering and the dispersion can be very well fit by the following formula @xcite @xmath28 here @xmath29 , @xmath30 are dimensionless momenta .
the dispersion has minimum at @xmath31 . near the minimum the dispersion squared
can be expanded in powers of momentum , @xmath32 , where @xmath33 , @xmath34 , @xmath35 , and the coefficients @xmath36 are expressed in terms of parameters from ( [ js ] ) .
diagonalization of the quadratic form gives the spin - wave velocities @xmath37 , @xmath38 the expansion ( [ c123 ] ) is generally valid only at small momenta , @xmath39 , however , accidentally the simple quadratic dependence along the `` softest '' direction @xmath40 is valid practically up to the boundary of the brillouine zone , @xmath41 .
therefore , magnetism in tlcucl@xmath1 is three - dimensional up to the energy / temperature @xmath42 above the gap . at a higher energy / temperature
the magnetism becomes two - dimensional .
taking into account eq.([c123 ] ) the stiffness term @xmath43 in the effective action ( [ lls ] ) has to be replaced by @xmath44\ .\end{aligned}\ ] ] in the magnetically disordered phase , @xmath16 the lagrangian ( [ lls ] ) yields the following triple degenerate spectrum @xmath45 comparing with the gap measured in neutron scattering , ref .
@xcite , and presented in fig .
[ figa ] by blue triangles we determine the value of @xmath46 , @xmath47 . here
we take into account that @xmath48 . ) and the magnetically ordered ( @xmath49 ) phases .
the circles , diamonds , triangles etc , show experimental data from ref .. the lines , solid and dashed show the theoretical values of the gaps .
we take the convention that the staggered magnetization in the ordered phase is directed along the z - axis , the corresponding longitudinal gap is shown by the topmost curve .
the goldstone transverse y - magnon is gapless in the ordered phase , while the x - magnon is slightly gapped due to the spin - orbit anisotropy .
, scaledwidth=40.0% ] the experimental data @xcite presented in fig .
[ figa ] demonstrate some anisotropy that is not accounted in the lagrangian ( [ lls ] ) . at the moment we disregard the anisotropy and return to this point later . on the other side of the qpt , @xmath19 , the field @xmath50 attains spontaneously the following expectation value @xmath51 stands for classical expectation value .
we chose the classical field to be directed along the z - axis in the spin space .
further standard analysis of the effective lagrangian shows that there are two transverse gapless goldstone x- and y - modes @xmath52 and one gapped longitudinal z - mode with dispersion @xmath53 the `` longitudinal '' gap agrees pretty well with the data shown in fig . [ figa ] by red triangles and diamonds .
to calculate the width of the longitudinal spin wave we need to know the coefficient @xmath54 in the quartic term in ( [ lls ] ) . to determine the coefficient we turn to the analysis of bose condensation of magnons in the external magnetic field at zero pressure .
the lagrangian ( [ lls ] ) results in the following energy density @xmath55 ^ 2\ .\end{aligned}\ ] ] here @xmath56 since @xmath16 . in the energy
we skip all terms with derivatives since we are interested in bose condensation in a homogeneous system . for magnetic field @xmath57 ,
the energy is minimized when @xmath58 , there is no spontaneous magnetization . for magnetic field @xmath59
( i.e. when the gap is closed ) , the energy is minimum when @xmath60 thus , there is a spontaneous staggered magnetization perpendicular to the direction of the magnetic field .
the magnetization can be arbitrarily rotated around the direction of the magnetic field , therefore , the system retains a u(1 ) symmetry as it should be for a bose condensate .
substitution of the expectation value ( [ 12 ] ) back to eq.([en ] ) gives the following energy density of the condensate @xmath61 in the second equality in this equation we assume that @xmath62 , i.e. we are close to the phase transition point at @xmath63 .
this concludes the analysis of bose condensation in terms of the effective action .
it is well known that the bose condensation of magnons can be also analysed in terms of usual schrodinger equation within the hartree - fock - popov approximation .
such an analysis has been performed earlier in ref . .
in the hartree - fock - popov approximation the density energy of the condensate at zero temperature reads @xmath64 where @xmath65 is the condensate density and @xmath66 is the effective repulsion between magnons .
the arrow `` @xmath67 '' shows the result of the energy minimization with respect to @xmath65 .
comparing ( [ en3 ] ) and ( [ hfp ] ) we express the field theory parameter @xmath54 is terms of the hartree - fock - popov parameter @xmath66 , @xmath68 .
fit of data @xcite for bose condensation in tlcucl@xmath0 performed in ref
. gave the value of the effective repulsion , @xmath69 .
the data in ref . is taken at zero pressure , therefore @xmath70 .
all in all we get @xmath71 according to the fit in ref . , there is a margin of error of about 20% in the values of @xmath66 and @xmath54 .
@xmath72 corresponds to the longitudinal magnon , @xmath73 corresponds to transverse goldstone magnons .
, scaledwidth=20.0% ] decay of the longitudinal magnon is due to the quartic term in ( [ lls ] ) and is described by the diagram shown in fig .
[ figld ] . to calculate the decay amplitude one needs first to rewrite the lagrangian in terms of magnons in the magnetically ordered phase .
after substitution of @xmath74 in eq.([lls ] ) ( the condensate field @xmath75 is given by eq.([vf ] ) ) we find the quadratic and the cubic part of the lagrangian @xmath76 the stiffness term @xmath77 is determined in eq.([cr ] ) .
standard quantization of the quadratic lagrangian gives ( we set @xmath78 ) @xmath79 where @xmath80 and @xmath81 are corresponding annihilation and creation operators .
the decay matrix element shown in fig .
[ figld ] is determined after the operators ( [ rw ] ) are substituted in the cubic lagrangian @xmath82 , @xmath83 where @xmath84 is the momentum of the initial longitudinal magnon and @xmath85 , @xmath86 are the momenta of the final transverse magnons .
the coefficient 2 in ( [ m ] ) is due to two different ways for pairing of creation and annihilation operators .
the decay width is given by fermi golden rule @xmath87 stands to avoid the double counting of final bosonic states .
integration in ( [ gr ] ) is straightforward , the answer is @xmath88 in this equation @xmath89 are the components of the vector @xmath90 . .
the circles and squares , show experimental data from ref .. the solid line shows the theoretical value calculated without account of the spin - orbit anisotropy .
the dashed line shows the theoretical value with account of the spin - orbit anisotropy .
, scaledwidth=40.0% ] according to ( [ gr1 ] ) the width @xmath91 in a reasonable agreement with experimental data from ref . presented in fig.[figb ] . according to eqs .
( [ gr1]),([c123]),([av ] ) the theoretical value for the ratio @xmath92 is @xmath93 . the corresponding theoretical curve for @xmath94 versus pressure is shown in fig . [ figb ] by the black solid line together with experimental data @xcite shown in the same figure .
we stress that we calculate not only the functional dependence of the gap on the pressure .
we calculate the absolute value of the gap .
overall the agreement between the theory and the experiment is quite good having in mind experimental error bars , fig.[figb ] , as well as 20% theoretical uncertainty in the value of @xmath54 .
the agreement can be further improved if one accounts for the spin - orbit anisotropy .
we discuss the anisotropy in the next section .
data presented in fig.[figa ] clearly indicate an easy plane anisotropy in the magnetically ordered phase : one of the transverse magnons is gapped with the constant gap @xmath95 where the subscript `` a '' stands for anisotropy . to incorporate the anisotropy in our description we have to add the term @xmath96 to the effective lagrangian ( [ lls ] ) . with account of the anisotropy
the triple degeneracy of the dispersion in the magnetically disordered phase , @xmath16 , is lifted , the z- and y - modes have the same dispersion ( [ sp1 ] ) as before while the x - mode has a higher energy @xmath97 the corresponding gaps ( @xmath98 ) are plotted by lines in fig.[figa ] versus pressure . the dispersion in the magnetically ordered phase , @xmath49 , is the following @xmath99 the corresponding gaps ( @xmath98 ) are also plotted by lines in fig.[figa ] versus pressure . overall agreement between the theory and the experimental data presented in the same fig.[figa ]
is excellent . the anisotropy gap influences the width of the longitudinal magnon only via limitation of the decay phase space .
a straightforward calculation shows that eq.([gr1 ] ) has to be modified in the following way @xmath100 \ .\end{aligned}\ ] ] here @xmath101 is the step function , @xmath102 if @xmath103 , otherwise @xmath104 .
the calculated decay width is shown in fig.[figb ] by the dashed line .
the agreement with experiment is excellent .
the decay width of the longitudinal magnon in the vicinity of the o(3 ) quantum critical point of a three - dimensional quantum antiferromagnet has been calculated using the effective field theory approach .
the results are quite generic .
we also compared our results with experimental data for tlcucl@xmath0 where the quantum phase transition can be driven by hydrostatic pressure and/or by external magnetic field .
the parameters of the effective action were determined from available data for inelastic neutron scattering and for bose condensation of magnons in the external magnetic field .
having determined these parameters , we calculated the decay width .
the result agrees very well with measurements .
we are grateful to b. normand for stimulating discussions and to j. oitmaa for important comments .
s. sachdev and b. keimer , phys . today * 64 * , february , 29 ( 2011 )
. k. takatsu , w. shiramura and h. tanaka , j. phys .
jpn . , * 66 * , 1611 ( 1997 ) .
a. oosawa , m. ishii , and h. tanaka , j. phys .
cond . mat .
, * 11 * , 265 ( 1999 ) .
t. nikuni , m. oshikawa , a. oosawa , and h. tanaka , phys .
lett . , * 84 * , 5868 ( 2000 ) .
n. cavadini , g. heigold , w. henggeler , a. furrer , h .- u .
gdel , k. krmer , and h. mutka , phys .
b * 63 * , 172414 ( 2001 ) .
choi , g. guntherodt , a. oosawa , h. tanaka , and p. lemmens , phys .
b * 68 * , 174412 ( 2003 ) . e. ya .
sherman , p. lemmens , b. busse , a. oosawa , and h. tanaka , phys . rev .
* 91 * , 057201 ( 2003 ) . v. n. glazkov , a. i. smirnov , h. tanaka , and a. oosawa , phys .
b * 69 * , 184410 ( 2004 ) .
k. goto , n. fujisawa , t. ono , h. tanaka , and y. uwatoko , j. phys .
73 * , 3254 ( 2004 ) .
k. goto , t. ono , h. tanaka , a. oosawa , y. uwatoko , k. kakurai , and t. osakabe , aip conf .
proc . * 850 * , 1059 ( 2006 ) .
h. tanaka , f. yamada , t. ono , t. sakakibara , y. uwatoko , a. oosawa , k. kakurai , and k. goto , j. magn .
mater . * 310 * , 1343 ( 2007 ) .
f. yamada , y. ishii , t. suzuki , t. matsuzaki , and h. tanaka , phys .
b * 78 * , 224405 ( 2008 ) .
t. sakurai , a. taketani , s. kimura , m. yoshida , s. okubo , h. ohta , h. tanaka , and y. uwatoko , aip conf .
proc . * 850 * , 1057 ( 2006 ) .
k. goto , t. osakabe , k. kakurai , y. uwatoko , a. oosawa , j. kawakami , and h. tanaka , j. phys .
. jpn . * 76 * , 053704 ( 2007 ) .
m. matsumoto , b. normand , t. m. rice , and m. sigrist , phys .
lett . , * 89 * , 077203 ( 2002 ) .
m. matsumoto , b. normand , t. m. rice , and m. sigrist , phys .
b * 69 * , 054423 ( 2004 ) .
g. misguich and m. oshikawa , j. phys .
jpn . * 73 * , 3429 ( 2004 ) .
j. sirker , a. weisse and o. p. sushkov , europhys .
lett . , * 68 * , 275 ( 2004 ) .
j. sirker , a. weisse and o. p. sushkov , physica b * 359 * , 1318 ( 2005 ) .
j. sirker , a. weisse and o. p. sushkov , j. phys .
* 74 * , 129 ( suppl ) ( 2005 ) .
b. normand , m. matsumoto , o. nohadani , s. wessel , s. haas , t. m. rice , and m. sigrist , j. phys . : condens .
matter * 16 * , s867 ( 2004 ) .
s. sachdev , arxive:0910.1139 , contributed chapter to the book `` understanding quantum phase transitions , '' edited by lincoln d. carr ( taylor & francis , boca raton , 2010 ) . | we consider a three - dimensional quantum antiferromagnet in the vicinity of a quantum critical point separating the magnetically ordered and the magnetically disordered phases .
a specific example is tlcucl@xmath0 where the quantum phase transition can be driven by hydrostatic pressure and/or by external magnetic field . as expected two transverse and one longitudinal magnetic excitation
have been observed in the pressure driven magnetically ordered phase . according to the experimental data , the longitudinal magnon has a substantial width , which has not been understood and has remained a puzzle . in the present work , we explain the mechanism for the width ,
calculate the width and relate value of the width with parameters of the bose condensate of magnons observed in the same compound .
the method of an effective quantum field theory is employed in the work . |
lyman alpha emitting ( lae ) galaxies have been shown to be building blocks of milky - way type galaxies today ( @xcite , @xcite ) .
the brightness of the lyman alpha line allows one to detect these galaxies using the narrow - band technique even when the continuum is faint . as a result
, large samples of laes have been studied using sed fitting at a variety of redshifts ( _ e.g. , _
@xcite , @xcite , @xcite , @xcite , @xcite , @xcite , @xcite , @xcite , @xcite , @xcite , @xcite , , @xcite , @xcite ) .
the picture emerging from these collective studies is far from homogeneous .
while laes were once thought to be young , dust - free galaxies experiencing their first episode of star formation , many of these investigations suggest a large spread in their physical properties , and find evidence of an older , more evolved , and dustier stellar population component .
an important caveat in the interpretation of results comes from the systematic uncertainties introduced by inconsistent modeling of the stellar populations among different groups .
this issue was studied in detail by @xcite , where we showed that the use of bc03 ( @xcite ) and cb11 ( @xcite ) stellar population synthesis ( sps ) templates , at solar or variable metallicity , and with or without including nebular emission , gave rise to a scatter in the estimate of age and mass significantly larger than the statistical uncertainty for the same data .
further scatter might be created by the use of a different initial mass function ( imf ) or of inconsistent statistical estimators of the physical properties of galaxies , such as the best - fit parameters as opposed to the mean of the probability distribution employed by bayesian statistics .
line has been subtracted from the photometry and is not included in the templates used to fit the sed . in the top panel , data at @xmath3 have been offset to account for the different luminosity distance .
the dip in the sed at rest - frame wavelengths @xmath5 1 @xmath9 m is typical of low - metallicity , old stellar populations . ]
this paper focuses on one particular puzzle , the ages of lyman alpha emitters , and one particular redshift range , the gyr between @xmath10 and @xmath2 .
laes have been mostly studied through stacking analyses , because their faint continua make it hard to fit the broad - band photometry of each galaxy , although @xcite and have also attempted to fit the seds of bright individual objects at these redshifts . at @xmath10 , @xcite and @xcite ( hereafter l08 )
examined the seds of average - stacked laes , segregated according to their detection in the irac 3.6 @xmath9 m band , and found that both samples were essentially dust - free , and that laes undetected in irac were fairly young ( age @xmath5 0.15 gyr ) and laes detected in irac were old ( age @xmath5 1.6 gyr ) .
at the same redshift , @xcite also found an old median age of 0.85 gyr , while @xcite found laes to be typically young ( age @xmath11 0.1 gyr ) , although a few of the individually fitted k - detected objects appeared to be older . at redshift @xmath2 , @xcite revealed the laes to be young ( age @xmath11 0.1 gyr ) , and did not find a noticeable difference in the age of laes when classified and stacked on the basis of their irac 3.6 @xmath9 m band luminosity . reported finding a young ( age @xmath11 0.05 gyr ) and an old ( age @xmath5 1 gyr ) stellar population , of comparable mass , in @xmath12 laes , although the authors did not report whether the two - component fit was statistically favored with respect to a fit with a single stellar population .
the purpose of our work is to compare the properties of laes at @xmath13 and @xmath14 by using exactly the same pipeline to compose the stacked samples and employing the same , state - of - the - art algorithm to analyze the seds .
this procedure will discriminate between physical evolution of laes and artificial differences introduced by inconsistencies in building the stacks and modeling the stellar populations .
the lae stacks at @xmath4 are those presented in @xcite , hereafter gu11 .
we consider the full sample , which includes 216 objects , as well as the irac - bright and irac - dim subsamples , which were built using flux cuts corresponding to the irac detection criterion used by l08 , appropriately rescaled to account for the difference in luminosity distance .
these samples were composed by median stacking , which is less sensitive to outliers than average stacking .
for this reason , we also revisit the detected and undetected stacks of l08 and use median stacking , and we compose a new median - stacked sed ( the full sample at @xmath8 = 3.1 ) that comprises 70 laes from @xcite . at both redshifts ,
laes are selected based on rest - frame ly@xmath15 equivalent width @xmath16 20 and on the flux in the ly-@xmath15 line .
the flux limits used in the two surveys lead to very similar rest - frame ly-@xmath15 luminosity limits , making the samples at @xmath3 and @xmath4 suitable for a comparative study @xcite . at @xmath3
, we have ubvrizjk data from musyc @xcite and irac photometry , while at @xmath4 we also have h band data .
the photometry is computed using iraf ; we use an aperture radius of 1 arcsec in the uv , optical and nir bands , and 1.25 arcsec in the irac bands .
the fluxes are aperture - corrected to estimate the total fluxes as described in @xcite . for both samples ,
objects found in crowded irac regions , where imperfect subtraction of bright nearby neighbors would bias the photometry , are discarded before the stacking .
we cross - checked that the photometric data obtained using sextractor are compatible with those computed through iraf .
errors on the stacked photometry are determined by a bootstrap procedure , which accounts for both the photometric error and the scatter resulting from the spread in the seds of different galaxies . at both redshifts ,
the latter is the predominant cause of uncertainties ; as a result , the size of the error bars and the number of objects in each stack are effectively uncorrelated .
we use the markov chain monte carlo code galmc @xcite to fit the spectral energy distribution of these six samples , and the getdist " software from @xcite to analyze the chains .
the data are made available on the musyc public data release website at http://physics.rutgers.edu/@xmath5gawiser/musyc/data.html .
the seds of the median - stacked full samples at redshifts @xmath3 and @xmath4 are plotted in fig . [
fig : data ] .
we begin by fitting four parameters : age since the beginning of star formation ( required to be lower than the age of the universe at that redshift ) , dust content parametrized by the excess color e(b - v ) , stellar mass ( which is the integral of the star formation rate minus the mass loss due to the life cycle of stars ) , and metallicity ( using a flat prior in log @xmath17 , as discussed in @xcite ) .
we use the latest cb11 models , use the calzetti law @xcite to compute the attenuation , assume constant star formation history , favored by the previous analyses of l08 and gu11 , and adopt a salpeter imf .
we include the contribution of nebular emission as described in @xcite , following the procedure outlined by .
results from sed fitting are shown in figs .
[ fig : prob1d ] and [ fig : prob ] and summarized in table [ tab : ressed ] .
unsurprisingly , given the different shapes of their seds , we find strong differences in the properties of @xmath3 and @xmath4 lyman alpha emitters .
the higher - redshifts laes are dust - free , have strongly sub - solar metallicities , have mean masses of @xmath18 m@xmath19 , and are significantly older , with mean ages of 1 gyr , than their lower - redshift counterparts , which present a very moderate amount of dust , appear to have solar metallicity , have mean masses of @xmath20 m@xmath19 , and have mean ages of 45 myr .
results for the full , irac - bright , and irac - dim stacked samples of laes at z = 2.1 can be compared directly to the ones for the same stacks of gu11 ; the main difference is the lower amount of dust ( e(b - v ) @xmath5 0.11 vs 0.22 for the full sample ) found by this analysis , which is a direct consequence of the inclusion of nebular emission lines in our code .
we can also compare our median - stacked detected and undetected samples to the average - stacked detected and undetected samples of l08 . in this case
the main difference is the older age estimated for the undetected sample , from the 160 myr of l08 to the 450 myr found by us .
this difference comes in part from the median stacking , which alters the shape of the sed slightly , enhancing the flux in the irac 3.6 and 4.5 @xmath9 m bands by a factor of 1.8 , and in part from allowing metallicity to vary and finding @xmath21 .
the quoted l08 results assumed solar metallicity , although the authors reported , consistent with our findings , that their age estimate for the undetected population increased if @xmath22 was used .
the evolutionary picture of laes between @xmath10 and @xmath2 suggested by our results is quite puzzling .
the same type of galaxies that are found to be old and metal - poor at @xmath10 appear to be young and metal - rich only a gyr later . before trying to reflect on the implications of this scenario , we consider several possible artificial sources of such a sharp dichotomy .
we begin by examining in detail the stacked seds of fig .
[ fig : data ] . at @xmath3 ,
the best - model has a very reasonable reduced @xmath23 of 1.4 .
this is a desired property because the stacking procedure relies on the strong assumption that the typical lae " exists , and that the artificial sed built by stacking many real seds is a reasonable representation of the spectrum of this typical lae galaxy .
a reasonable reduced @xmath23 value suggests that the bootstrap uncertainties correctly account for the spread in the seds of laes and does not illuminate a template incompleteness problem .
nonetheless , there is a relevant amount of scatter in the stacked sed , with three points particularly standing out , the low measurements in the j and irac 5.8 @xmath9 m band , and the high measurement in the k band .
we re - did the sed fitting excluding each of them , and each pair , in turn , and we found that our results are extremely robust to the exclusion of these data points , with changes in the parameter estimates by less than 0.2@xmath24 .
we also note that the results at @xmath3 for both age and metallicity are in good agreement with those of @xcite , and that the analysis of @xcite favors low metallicity values .
strongly sub - solar metallicities were also found in damped ly@xmath15 systems by @xcite . at @xmath4 ,
the sed visually appears to be smoother ; however , its reduced @xmath23 has a slightly higher value of 2.2 .
this feature suggests that the stacking at this redshift might be an overly aggressive compression of the information contained in the individual seds .
breaking down the sample in irac - bright and irac - dim laes in the attempt to increase the homogeneity of the stacks does not afford a better fit , although it shows that the properties of these two stacks follow closely the ones of the full sample .
the sed fitting parameters are correlated with each other , and in particular , there is a well - know age - metallicity relation that can be seen easily as the axis of the 2-d credible region in the age - metallicity plane of fig . [
fig : prob ] .
our estimates of metallicity depend on the sps templates we use and ( although weakly ) on the flat prior on @xmath25 that we assume .
we think that the latter is physically motivated , and we have no reason not to trust the models ; however , it is possible that the observed difference in metallicity between laes at the two redshift is overestimated .
if the true metallicity at @xmath3 were higher , and the true metallicity at @xmath4 were lower , this might help reconcile the gap in the physical properties of these galaxies . to test this hypothesis
, we run our sed fitting code on the full samples at the two redshifts , holding the metallicity fixed at @xmath26 ; this value is suggested by the spectroscopic analysis of three laes at @xmath27 by @xcite .
results are shown in fig .
[ fig : syst ] and listed in table [ tab : ressed ] . as expected , the estimates for the mean ages of the stellar population shift in the desired direction , but remain incompatible at several sigma level .
this result seems to exclude metallicity as the primary source of the gap in the age of laes , at least in this stacking analysis .
the assumption of a particular functional form for the star formation history of laes can also influence the determination of ages .
we investigate this issue by repeating the sed fitting procedure on the full samples using exponentially increasing and decreasing star formation histories ( sfhs ) , @xmath28 . by sampling in @xmath29 , both increasing and decreasing sfhs
are explored as part of a contiguous parameter space , as explained in @xcite .
this parametrization is also able to capture constant or starburst sfhs for appropriate values of @xmath30/age . the results of this sed fitting run ( where the metallicities were held fixed at the best - fit value for both samples ) are shown in fig .
[ fig : syst ] ; in both cases the estimates of ages do not change significantly .
the quality of the fit does not improve significantly in either case by allowing this larger range of star formation histories , as can be seen in table [ tab : ressed ] .
the data favor a nearly constant sfh at both redshifts , with a best - fit value of @xmath30 around 4 gyr for both samples .
gu11 had also performed the sed fitting using these sfhs for the @xmath31 sample , albeit with a different algorithm , and had found similar results . a possible explanation of the difference in the physical nature of laes at @xmath3 and @xmath4 might be the presence of stellar components of different ages .
we test this hypothesis by performing the sed fitting on the full samples using two stellar populations , fixing the age of the second stellar population at 1 gyr , and introducing the stellar mass of the second stellar population as a free sed fitting parameter .
we show the results of this test for a single - burst model that would solely contain old stars , although we have also considered constant or linearly increasing sfhs for the second stellar population , as suggested by some recent observations ( _ e.g. , _
@xcite and references therein ) . for @xmath4 laes
, we also consider a few different values of metallicity and dust content for the second population . at @xmath4 ,
this test is meant to reveal an older stellar population ; at @xmath3 , it may allow us to single out two components of different ages , although we do not expect to find a large component of very young stars , since they would dominate the rest - frame uv part of the total spectrum . in the case of @xmath3 laes , adding a second old stellar population slightly improves the quality of the fit ( from @xmath32 = 1.4 to @xmath32 = 1.0 ) .
young ages for the primary stellar population are allowed , but only when the fraction of mass in this stellar population is small , as seen in the third panel of fig .
[ fig : syst ] .
for example , ages of the order of @xmath33 myr are allowed at @xmath34 confidence when the mass in young stars is less than @xmath35 of the total stellar mass . as expected
since the two stellar populations are so similar , the mass ratio in the two populations is unconstrained . for @xmath4 laes ,
the quality of the fit does not improve by adding a second stellar population , and again the age estimate for the first stellar population does not change from the value of @xmath5 50 gyr favored by the single population fit , and the mass fraction in the
old " stellar population is found to be consistent with zero , and @xmath36 at @xmath37 confidence for all cases considered .
we conclude that the two - component scenario can not be confirmed by analyzing the stacked seds , although the improvement in the @xmath23 at @xmath8 = 3.1 might indicate a heterogeneity of the stellar populations , masked by the stacking process .
the clustering analysis of laes @xcite has shown that laes at @xmath10 and @xmath2 are hosted by dark matter halos that can be in a direct progenitor - descendant relationship .
an intriguing possibility is that they might represent two different stages of the same galaxy population .
we test this hypothesis by using the same data analysis pipeline and sed fitting software ; this strategy allows us to discriminate between evolution in the galaxy properties and the effects of different assumptions in the processing of data and the modeling of the stellar populations .
we find that the physical properties of laes at @xmath3 are very different from those of laes at @xmath4 , which appear to be several times younger and more metal - rich than their higher - redshift counterparts .
if these samples are a fair representation of laes at each redshift , this result directly rules out the hypothesis that z=3.1 laes evolve into z=2.1 laes , implying that the lae phase lasts significantly less than 1 gyr .
moreover , such stark differences between emission - line selected galaxy samples become difficult to explain as the effect of 1 gyr of cosmological evolution .
we find hints that stacking analysis , meant to provide insights on the nature of the _ typical _ lae galaxy , might be overly aggressive and perhaps unable to capture the multi - component nature of laes at each redshift . to be properly investigated , these issues require constraints from sed fitting of large samples of individual objects , as well as spectroscopic constraints on the metallicity of laes .
the former have been especially challenging to acquire so far , because of the lack of deep data in the observed nir region of the spectrum .
this gap is being filled by deep yjh hst observations by the cosmic assembly near - infrared deep extragalactic survey ( candels , @xcite ) , and we plan to address this issue by using candels data in the musyc fields in a subsequent paper ( vargas et al 2012 , in preparation ) .
it is a pleasure to thank kamson lai for advice on the photometry at @xmath8 = 3.1 , the astrophysics group at rutgers , in particular saurabh jha and curtis mccully , and the anonymous referee , for useful conversations .
this research was supported in part by the american astronomical society s small research grant program . | lyman alpha emitting ( lae ) galaxies are thought to be progenitors of present - day l@xmath0 galaxies .
clustering analyses have suggested that laes at @xmath1 might evolve into laes at @xmath2 , but it is unclear whether the physical nature of these galaxies is compatible with this hypothesis .
several groups have investigated the properties of laes using spectral energy distribution ( sed ) fitting , but direct comparison of their results is complicated by inconsistencies in the treatment of the data and in the assumptions made in modeling the stellar populations , which are degenerate with the effects of galaxy evolution . by using the same data analysis pipeline and sed fitting software on two stacked samples of laes at @xmath3 and @xmath4 , and by eliminating several systematic uncertainties that might cause a discrepancy , we determine that the physical properties of these two samples of galaxies are dramatically different .
laes at @xmath3 are found to be old ( age @xmath5 1 gyr ) and metal - poor ( @xmath6 ) , while laes at @xmath4 appear to be young ( age @xmath5 50 myr ) and metal - rich ( @xmath7 ) .
the difference in the observed stellar ages makes it very unlikely that @xmath8=3.1 laes evolve directly into @xmath8=2.1 laes .
larger samples of galaxies , studies of individual objects and spectroscopic measurements of metallicity at these redshifts are needed to confirm this picture , which is difficult to reconcile with the effects of 1 gyr of cosmological evolution . |
not much is known from first principles about the topological structure at the thermal phase transition in quantum chromodynamics . how this structure changes at the deconfinement phase transition is not even clear in pure gauge theory .
what we mean by topological structure is the characterization of euclidean field trajectories in terms of instantons @xcite and abelian monopoles @xcite and their interrelation .
changes of this structure are expected to be the common mechanism for deconfinement and restoration of chiral symmetry . understanding this mechanism is important for understanding the qcd vacuum .
there was a wave of attention recently for improving lattice actions for gauge fields @xcite ( and fermions ) .
part of the interest was motivated to minimize finite lattice spacing corrections in simulation results obtained on coarse lattices .
one particular approach towards improvement was guided by the desire to construct a perfect action being a fixed point action @xcite with respect to some real space renormalization group ( rg ) transformations . in this context
the concept of inverse blocking came up which is the essence of a new method , recently proposed by degrand et al .
in particular it improves the determination of the topological charge of a lattice configuration . in the present paper
we want to explore the capability of this method , to resolve more details and different aspects of topological structure . in the neighbourhood of the deconfinement transition of the pure @xmath1 gauge theory , we want to study the space - time distribution of topological charge , correlations of the density with itself and with abelian monopole currents .
the topological charge density , when defined for quantum fields , is obscured by ultraviolet fluctuations which must be removed in some way . cooling @xcite has been invented for this purpose long ago . according to the abelian dominance scenario at large distances @xcite abelian magnetic monopoles are closely related to confinement .
the density @xcite and anisotropy @xcite of these magnetic currents are studied in the maximally abelian gauge @xcite .
these currents are quantized by construction @xcite . this does not exclude the possibility that some part of these currents are lattice artefacts which do not correspond to long distance physics .
correlations indicating a close relation between topological charge and abelian magnetic monopole currents have been studied , not addressing this question .
they have been studied so far for particular instanton configurations @xcite and , as far as the quantized vacuum is concerned , the cooling method has been employed @xcite . however , cooling is known to change the topological structure and has to be used with care .
a closer examination shows that the method of refs .
@xcite gives not only the topological charge but also permits a reasonable definition of the topological density of generic , equilibrium lattice fields with minimal effort .
a better resolution of topological and monopole structure is possible without recourse to cooling .
while the method is applied to extract the topological structure , the monopole activity is changed , too , roughly in proportion to the topological activity ( to be defined below ) .
we shall see that part of it becomes an order parameter of the confinement / deconfinement transition , according to the alleged role monopoles play for confinement . according to the scope of our present study ,
let us sketch the background of this approach @xcite restricting the discussion to the case of pure gauge field theories .
the starting point is a block spin transformation ( here of scale factor two ) which maps a fine lattice configuration of link matrices @xmath2 to a coarse lattice configuration with link matrices @xmath3 ( each single link being an element of the gauge group @xmath4 ) .
the second ingredient is a general type of action @xmath5 , written on both lattices in terms of wilson loops ( evaluated in various representations ) .
the classical perfectness of this action is guaranteed if there exists an inverse blocking transformation @xmath6 with respect to which the action saturates the inequality @xmath7 here , @xmath8 is some non - negative functional related to the blockspin transformation and @xmath9 is a kind of lagrange multiplier .
once the action and other parameters have been determined , this inequality leads to the target configuration @xmath10 of the inverse blocking by minimization of the l. h. s. , the fixed point action with respect to @xmath10 , with a constraint derived from the coarse configuration @xmath11 .
thus , @xmath10 is found as the result of an iterative relaxation process and can not be explicitely expressed .
we call the mapping @xmath12 one step of inverse blocking . similar to blocking , inverse
blocking can be iterated .
this leads to a sequence of interpolating configurations on finer and finer lattices . without giving details ,
we mention that the saturated inequality ( [ eq : inequality1 ] ) is constructive for finding the perfect action itself . it has been used @xcite to optimize the ratios between the couplings in front of various loops in different representations within the fixed point action @xmath5 , as well as the other constants related to the block spin transformation .
after this machinery has been set up , the procedure of inverse blocking remains a useful tool of analysis .
doing a sequence of inverse blocking steps one would be able to relate a continuum configuration to any given lattice configuration .
the topological charge of this continuum configuration is _ the topological charge _ related to the given lattice configuration ( the _ classically perfect topological charge _ ) .
this prescription is an important conceptual step forward compared with previous techniques of measuring topological charge on the lattice , eventually used in combination with cooling .
inverse blocking has been used before @xcite only to measure the topological charge of lattice configurations , by evaluating the phillips stone ( ps ) charge @xcite just on the next finer level .
the success indicates that one step of inverse blocking can be sufficient to have access to the classically perfect topological charge corresponding to the _ continuum _ by means of a _ lattice _ definition .
other lattice algorithms besides ps could also have been used , for instance in order to test the uniqueness of charge .
we will use in this paper a geometric charge due to lscher @xcite and the old plaquette based ( field theoretic or naive ) definition @xcite .
both have the advantage to localize the topological density in space - time which makes them suitable for our present purpose . the first concern in ref .
@xcite has been that action and topological charge of artificially constructed lattice instantons should be independent of size , if the perfect action is used and the topological charge is measured as described .
this was demonstrated for instanton configurations with @xmath13 .
there were no dislocations , with action @xmath14 ( entropic bound , instanton action @xmath15 ) and non zero topological charge , among them .
however , the perfect action and the new definition of topological charge are useful also at finite @xmath16 . in the second paper
@xcite a ( truncated ) fixed point action has been successfully tested for scaling of the deconfinement temperature , of the torelon mass , of the string tension and of the topological susceptibility . for the latter
, it has been stressed that inverse blocking has been indispensable to achieve reasonable scaling .
this has nourished the hope that the fixed point action and the inverse blocking prescription for the topological charge can avoid dislocations which otherwise ( for instance in the case of wilson s action ) are known to spoil the continuum limit of the topological susceptibility @xcite .
even the temperature dependence of the topological susceptibility has remained controversial until recently . armed with the new method , the topological susceptibility has been measured across the deconfinement transition @xcite without the ambiguities of the cooling method @xcite .
inverse blocking is not the only technique to avoid cooling .
we only want to mention the use of an improved lattice operator of the topological density @xcite .
this approach has given a much more discontinuous drop @xcite at the deconfinement transition than the inverse blocking method @xcite .
our paper is organized as follows . in section 2
we give details on the fixed point action , on blocking and inverse blocking .
in section 3 we present results on the string tension . for symmetric lattices
this serves to calibrate the lattice step size @xmath17 in accordance to simulations using the fixed point action .
the finite temperature string tension is computed ( in the confined phase ) from polyakov line correlators .
these results are indicating that the string tension is largely conserved by inverse blocking .
section 4 contains our results on the gross topological features .
the topological susceptibility obtained on inversely blocked configurations in the confined phase is quantitatively reasonable .
the temperature dependence of the topological susceptibility across the phase transition is studied . analysing inversely blocked configurations we observe , apart from the decay of the topological susceptibility , the complete suppression of spatial abelian monopole currents in the deconfined phase . in section 5
we present results on topological charge density density correlations and on monopole instanton correlations .
we visualize clusters of topological charge together with their accompanying monopole currents for typical confining configurations . in closing
, directions for further work are pointed out in section 6 .
the simplified fixed point action @xcite is parametrized in terms of only two types of wilson loops , plaquettes @xmath18 ( type @xmath19 ) and tilted @xmath20-dimensional @xmath21-link loops ( type @xmath22 ) of the form @xmath23 it contains several powers of the corresponding linear action terms as follows @xmath24 the parameter of this action have been optimized in ref .
@xcite and are reproduced in table [ tab : weights ] .
.__weight coefficients of the simplified fixed point action _ _ [ cols="<,^,^,^,^",options="header " , ] it has been demonstrated before for certain classical configurations that there is a close connection between instantons and monopole world loops .
one can visualize this in the maximally abelian gauge @xcite .
a nontrivial correlation has been measured between topological charge density and monopole currents in the monte carlo ensemble of euclidean field configurations , too @xcite .
we show here in fig .
[ fig : qlumon_corr ] the analogous normalized correlator for the original monte carlo ensemble ( generated with the fixed point action at @xmath25 in the confinement phase ) and for the corresponding ensemble of inversely blocked configurations .
the latter correlator ( for smoothed configurations ) is slightly wider .
in contrast to the topological density density correlations , this correlator can be detected in the monte carlo configurations as well . comparing this with the same ( normalized ) correlation function at @xmath26 in the deconfined phase ( not shown here ) , we find that they are identical as a function of the distance in lattice units within the error bars .
this means that this correlation length changes proportional to the inverse temperature .
we have explained in this paper how inverse blocking can be used to analyze topological structures of genuine quantum configurations , usually hidden under ultraviolet fluctuations .
apart from cooling techniques , inverse blocking is the only way to measure autocorrelations of the topological density giving information on the size of instantons .
information on monopole currents is available also for normal monte carlo configurations .
monopoles are correlated with the topological density , both in monte carlo and inversely blocked configurations . however , the density of monopole currents is reduced , roughly proportional to the topological activity , if a configuration undergoes the procedure of blocking followed by inverse blocking .
only a small part of abelian monopoles seen usually in monte carlo configurations seems to be really important for confinement .
this part survives smoothing in the confinement phase .
we come to this conclusion because the string tension is found to be largely unaffected by the removal of ultraviolet fluctuations .
this observation supports semiclassical scenarios of confinement . in the deconfinement phase , however , we find that spacelike monopoles are strongly suppressed for inversely blocked configurations while only timelike ones survive .
armed with inverse blocking , we can do better than measuring the topological density density correlation or the correlation between monopoles and topological density .
we are now able to study clustering properties of the topological charge using inversely blocked lattice configurations .
we can relate clusters of topological charge to the diluted network of monopole currents . instead of examining idealized semiclassical configurations under different aspects of topological and monopole structure , the method of inverse blocking
makes it possible to analyze generic euclidean lattice field histories in order to characterize the physical phase one is dealing with .
the results should be interesting for vacuum model builders . in view of chiral symmetry breaking and restoration
, it would be also interesting to study the influence of the inverse blocking rg transformation on the spectrum of the dirac operator .
the study of clusters of topological charge promises to give a more direct access to the parameters and correlations that should be built in an instanton model for qcd at zero and finite temperature . in figs .
[ fig : unsmooth ] and [ fig : smooth_luescher ] lattice hypercubes of one timeslice are marked if the lscher topological charge density exceeds a threshold value @xmath27 .
different colors symbolize the sign .
the original monte carlo configuration does not allow to guess any structure .
the other picture shows the structure emerging after blocking and inverse blocking .
the effect is the same using either naive or lscher s charge density
. therefore one can abandon to use complicated algorithms and use naive or improved loop
oriented definitions of topological charge density .
there are two immediate directions for further studies .
* study of the clustering properties of topological charge in individual lattice configurations and of the relation to the monopole currents . * repeating the present explorative study at higher @xmath16 with a particular emphasis on the scale dependence of topological structure .
further progress in the construction of the perfect action should be incorporated into this program .
during the time elapsed since the first submission of our article , several interesting papers have appeared which have attempted to investigate the space time topological structure of lattice gauge fields .
three methods have been used : ( improved ) cooling , fermionic ( spectral ) methods and smoothing .
the latter method had been proposed in this paper as a general , renormalization group oriented pattern recognition method for large scale structures in terms of topological charge and monopoles which conserves the string tension .
while the subsequent studies were mostly focussed on vacuum ( zero temperature ) configurations , the results are still controversial as emphasized in ref . @xcite . as a generic rule
, one can state that improved cooling @xcite is biased towards larger instantons ( because only those are stable ) while the topological susceptibility tends to be lower ; in contrast , the smoothing method ( as developed further by the boulder group @xcite ) favours smaller instanton sizes and tends to find somewhat higher topological susceptibility .
this applies to @xmath1 pure gauge theory where the results can be compared ( but not without additional assumptions ) .
we want to stress an important difference of ref .
@xcite to the work published here .
smoothing has been applied there in an iterative way that leads finally to locally classical configurations .
the number of iterations appears there as somewhat subjective parameter similar to the number of ( unimproved or improved ) cooling steps .
this goes far beyond our intention . from our point of view
, the advantage of smoothing is that it provides a minimal smoothing of quantum fluctuations in clear correspondence to a certain scale .
this scale is well defined by the highest blocking level .
the effect of smoothing on various abelian projections has been critically examined in ref .
the maximally abelian gauge is the only one which seems to exhibit long range physics also if applied to smoothed lattice fields .
while we had concentrated on the density of monopoles , it has been found there that the property of abelian dominance ( _ i.e. _ dominance of the abelian string tension ) is only partly preserved .
a large number of investigations was studying topological structures using fermionic methods going back to smit and vink @xcite .
it is important in the context of this work that these methods seem to work immediately , without smoothing or cooling techniques which otherwise would have to be applied to the monte carlo lattice gauge field configurations . investigating the spectral flow of the lowest eigenvalues of the dirac operator @xcite these authors are able to count the number of localized lumps ( and identify the respective sign ) of topological charge .
the local topological structure is thereby indirectly accessible through the localization properties of the corresponding eigenmodes .
this has been checked in refs .
@xcite with cooling , but there is no comparison so far with the topological structure detected by smoothing .
we have recently improved our relaxation algorithm for implementing inverse blocking @xcite .
we are now in the position to optimize a fixed point action of moderate complexity for a given inverse blocking parameter @xmath9 .
we could check a new parametrization of the perfect action for @xmath1 which is truncated compared with the action used in ref .
we thank the boulder group for correspondence about that and for providing the truncated action @xcite .
we can now show that the parametrization of table [ tab : weights ] is not compatible with the @xmath28 required .
this inconsistency had forced us to accept ( in this paper ) a smaller effective @xmath9 .
we have convinced ourselves that the statistical observations are independent of this choice although the inversely blocked configurations may differ locally . c. g. callan , r. dashen and d. j. gross , phys .
* d 17 , ( 1978 ) 2717 ; d. j. gross , r. d. pisarski and l. g. yaffe , _ rev .
phys . * 53 , ( 1981 ) 43 e. shuryak , phys .
reports * 264 , ( 1996 ) 357 ; t. schfer and e. v. shuryak phys .
* d 54 , ( 1996 ) 1099 ; t. schfer and e. v. shuryak , phys .
* d 53 , ( 1996 ) 6522 ; t. schfer and e. v. shuryak , _ e - print archive hep - ph/9610451 s. mandelstam , _ phys .
reports * 23c , ( 1976 ) 245 g. t hooft , nucl .
* b 190 , ( 1981 ) 455 f. niedermayer , _ nucl .
* b _ proc .
suppl . * 53 ( 1997 ) 56 ; w. bietenholz , r. brower , s. chandrasekharan and u. j. wiese , _ nucl . phys .
* b _ proc . suppl .
* 53 ( 1997 ) 921 t. a. degrand , a. hasenfratz , p. hasenfratz , f. niedermayer , and u. j. wiese , _ nucl .
* b _ proc .
suppl . * 42 ( 1995 ) 67 ; t. degrand , a. hasenfratz , p. hasenfratz and f. niedermayer , _ nucl .
* b 454 , ( 1995 ) 587 ; _ nucl .
* b 454 , ( 1995 ) 615 ; _ phys .
* b 365 , ( 1996 ) 233 ; m. blatter and f. niedermayer , _ nucl .
* b 482 , ( 1996 ) 286 t. degrand , a. hasenfratz , p. hasenfratz , p. kunszt and f. niedermayer , _ nucl . phys .
* b _ proc .
suppl . *
53 ( 1997 ) 942 t. degrand , a. hasenfratz and de - cai zhu , _ nucl
. phys .
* b 475 , ( 1996 ) 321 t. degrand , a. hasenfratz and de - cai zhu , _ nucl .
* b 478 , ( 1996 ) 349 e.m .
ilgenfritz , m. l. laursen , m. mller
preussker , g. schierholz and h. schiller , _ nucl .
* b 168 , ( 1986 ) 693 ; m. teper , _ phys .
* b 162 , ( 1985 ) 357 v. g. bornyakov , e.m .
ilgenfritz , m. l. laursen , v. k. mitrjushkin , m. mller preussker and a. j. van der sijs , _ phys .
* b 261 ( 1991 ) 116 v. g. bornyakov , v. k. mitrjushkin and m. mller - preussker , _ phys .
lett . * b 284 , ( 1992 ) 99 a. s. kronfeld , g. schierholz and u. j. wiese , _ nucl .
* b 293 , ( 1987 ) 461 ; a. s. kronfeld , m. l. laursen , g. schierholz and u. j. wiese , _ phys .
* b 198 , ( 1987 ) 516 t. a. degrand and d. toussaint , _ phys .
* d 22 , ( 1980 ) 2478 v. g. bornyakov and g. schierholz , _ phys .
* b 384 , ( 1996 ) 190 ; m. n. chernodub and f. v. gubarev , _ jetp lett .
* 62 , ( 1995 ) 100 ; a. hart and m. teper , _ phys .
* b 371 , ( 1996 ) 261 ; r. c. brower , k. n. orginos and c.i tan , _ nucl .
* b _ proc .
* 53 ( 1997 ) 488 and _ e - print archive hep - th/9610101 s. thurner , m. feurstein , h. markum and w. sakuler , _ phys .
* d 54 , ( 1996 ) 3457 ; s. thurner , h. markum and w. sakuler , in _ proceedings of confinement 95 _ , osaka 1995 , eds . h. toki et al .
( world scientific , 1996 ) 77 ; h. markum , w. sakuler and s. thurner , _ nucl .
* b _ proc .
suppl . * 47 ( 1996 ) 254 a. phillips and d. stone , _ comm .
* 103 , ( 1986 ) 599 ; a. s. kronfeld , m. l. laursen , g. schierholz , c. schleiermacher and u. j. wiese , _ comp .
* 54 , ( 1989 ) 109 m. lscher , _ comm .
* 85 , ( 1982 ) 29 i. a. fox , j. p. gilchrist , m. l. laursen and g. schierholz , _ phys .
* 54 , ( 1985 ) 749 ; a. s. kronfeld , m. l. laursen , g. schierholz and u. j. wiese , _ nucl .
* b 292 , ( 1987 ) 330 p. di vecchia , k. fabricius , g. c. rossi and g. veneziano , _ nucl .
* b 192 , ( 1981 ) 392 ; _ phys .
* b 108 , ( 1982 ) 323 ; _ phys .
* b 249 , ( 1990 ) 490 n. v. makhaldiani and m. mller
preussker , _ jetf pisma * 37 , ( 1983 ) 440 m. gckeler , a. s. kronfeld , m. l. laursen , g. schierholz and u. j. wiese , _ phys .
* b 233 , ( 1989 ) 192 a. di giacomo , e. meggiolaro and h. panagopoulos , _ phys .
* b 277 , ( 1992 ) 491 e.m .
ilgenfritz , e. meggiolaro and m. mller
preussker , _ nucl .
* b _ proc .
* 42 ( 1995 ) 496 c. christou , a. di giacomo , h. panagopoulos and e. vicari , _ phys .
d 53 , ( 1996 ) 2619 b. alles , m. delia and a. di giacomo , _ nucl .
* b _ proc .
* 53 ( 1997 ) 541 and _ e - print archive 9605013 see the second of ref .
@xcite m. falcioni , m. paciello , g. parisi and b. taglienti , _ nucl .
* b 251 , ( 1985 ) 624 ; m. albanese et al .
, _ phys .
b 192 , ( 1987 ) 163 m. campostrini , a. di giacomo , m. maggiore h. panagopoulos and e. vicari , _ phys .
* b 225 , ( 1989 ) 403 m. campostrini , a. di giacomo and h. panagopoulos , _ phys .
* b 212 , ( 1988 ) 206 ; m. campostrini , a. di giacomo , h. panagopoulos and e. vicari , _ nucl .
* b 329 , ( 1990 ) 683 m. c. chu , j. m. grandy , s. huang and j. w. negele , _ phys .
* d 49 , ( 1994 ) 6039 ; r. c. brower , t. l. ivanenko , j. w. negele and k. n. orginos , _ nucl .
* b _ proc .
suppl . * 53 ( 1997 ) 547 a. di giacomo and e. vicari , _ phys .
* b 275 , ( 1992 ) 429 ; b. alles , m. campostrini , a. di giacomo , y. gndc and e. vicari , _ phys .
* d 48 , ( 1993 ) 2284 m. garcia perez , a. gonzalez arroyo , j. snippe and p. van baal , _ nucl .
* b 413 ( 1994 ) 535 p.
de forcrand , m. garcia perez and i.o .
stamatescu , _ nucl .
* b _ proc .
* 47 ( 1996 ) 278 and 777 p. de forcrand , m. garcia perez and i.o .
stamatescu , _ nucl .
* b 499 ( 1997 ) 409 ; p.
de forcrand , m. garcia perez , j. e. hetrick and i.o .
stamatescu , _ contribution to lattice 97 , e - print archive hep - lat/9710001 e.m .
ilgenfritz and m. mller
preussker , _ phys .
* b 99 , ( 1981 ) 128 g. mnster , _ z. f. phys .
* c 12 , ( 1982 ) 43 l. del debbio , a. di giacomo , g. paffuti and p. pieri , _ nucl .
* b _ proc .
* 42 ( 1995 ) 234 and _ phys .
* b 355 , ( 1995 ) 255 ; m. n. chernodub , m. i. polikarpov and a. i. veselov , _ nucl .
* b _ proc .
* 49 ( 1996 ) 307 and _ e - print archive hep - lat/9610007 ; a. i. veselov , m. i. polikarpov and m. n. chernodub , _ jetp lett . * 63 , ( 1996 ) 411 m. c. chu and s. schramm , _ phys .
* d 51 , ( 1995 ) 4580 t. degrand , a. hasenfratz , t. g. kovacs , _ e - print archive hep - lat/9705009 ; and _ contribution to lattice 97 , e - print archive hep - lat/9709095 t. g. kovacs , z. schram , _ e - print archive hep - lat/9706012 , _ phys .
* d ( to be published ) j. smit and j. vink , _ nucl .
* b 286 , ( 1987 ) 485 , * b 303 , ( 1988 ) 36 r. narayanan and r. l. singleton , _ contribution to lattice 97 , e - print archive hep - lat/9709014 r. narayanan and p. vranas , _ e - print archive hep - lat/9702005 , _ nucl .
* b ( to be published ) d. smith , h. simma and m. teper ( ukqcd collaboration ) , _ contribution to lattice 97 , e - print archive hep - lat/9709128 t. l. ivanenko and j. w. negele , _ contribution to lattice 97 , e - print archive hep - lat/9709130 ; j. w. negele , _ contribution to lattice 97 , e - print archive hep - lat/9709129 e.m .
ilgenfritz , m. mller preussker and s. thurner , _ in progress t. g. kovacs , _ private communication _ _ _ _ _ * _ _ _ * * _ * _ _ _ _ * _ * _ _ * _ * _ * _ * _ * _ * _ * _ _ * _ * _ * _ * _ * _ * _ * _ * _ * _ * _ * _ * _ * _ * _ _ * _ * _ * _ * _ * _ * _ * _ * _ * _ * _ * _ * _ * _ * _ * _ * _ * _ * _ * _ _ * _ * _ * _ * _ * _ * _ * _ * _ * _ * _ * _ * _ * _ * _ * _ * _ * _ * _ * _ * _ * _ * _ * _ * _ * _ * _ * * _ _ * * * * _ * | renormalization group transformations as discussed recently in deriving fixed point actions are used to analyse the vacuum structure near to the deconfinement temperature .
monte carlo configurations are generated using the fixed point action .
we compare equilibrium configurations with configurations obtained by inverse blocking from a coarser lattice .
the absence of short range vacuum fluctuations in the latter does not influence the string tension .
for the inversely blocked configurations we find the following : ( _ i _ ) the topological susceptibility @xmath0 is consistent with the phenomenological value in the confinement phase , ( _ ii _ ) @xmath0 drops across the deconfinement transition , ( _ iii _ ) density and size of instantons are estimated , ( _ iv _ ) the topological density is found to be correlated to abelian monopole currents and ( _ v _ ) the density of spacelike monopole currents becomes a confinement order parameter . |
3-dimensional ( 3d ) reconstruction is one method of stereoscopic analysis and involves the process of using multiple views of a single object from different view angles in an effort to determine the intrinsic 3d shape of the object . with one viewpoint ,
only 2-dimensional feature qualities are observed and any underlying 3d structure must be inferred from theory . with more than one viewpoint , comparison of the projection effects in the images can be used to mathematically invert the images to determine an optimal 3d distribution of matter that would best reproduce the observed images , such as with rotational tomography @xcite to provide multiple viewpoints over time .
coronal mass ejections ( cmes ) are rapidly evolving , fast moving heliospheric plasma ejecta originating in the lower solar atmosphere .
early lasco @xcite and eit @xcite results included the first observations of the start of a cme @xcite and analysis included comparisons against the flux rope model . in one sample cme event ,
a prominence accelerated to 100@xmath0 and then , within the view range of 1.1 to 30 solar radii in the corona , accelerated to 200 - 400 km / s @xcite , ultimately covering 70 degrees of visual latitude .
cmes which are directed towards earth can cause disruptive effects when they hit and interact with the earth s magnetic field . to model dynamic features such as coronal mass ejections ( cmes )
, the multiple view angles must also be simultaneous or near - simultaneous .
with the two view angles from solar terrestrial relations observatory ( stereo , ) a and b satellites , optionally combined with additional viewpoints from other satellites , 3d reconstruction of rapidly evolving features such as cme is possible @xcite but the problem is underconstrained when there are too few viewpoints . forward modeling ( fm )
presumes a theoretical shape for a cme and seeks to position that shape so as to reproduce the observed data . for data with strong asymmetries ,
additional model assumptions must be added to accurately reproduce the data .
inverse modeling makes no shape assumptions when carving out an underlying density distribution that , when rendered , reproduces the observed data . for underconstrained problems ,
there are multiple possible density distributions that can reproduce the same data , so some criteria for weighting potential solutions must be employed . since cme structure is neither perfectly formed like a model , nor completely unknown as inversion assumes , we combine the two methods to make minimal assumptions on shape and extent as a constraint for the model - free inverse solution . in our hybrid reconstruction , we define the detailed 3d density distribution for the cme and derive the 3d height - time from the sun for each data frame .
this is a true 3d position for the chosen center of mass with no projection effects .
prediction of the arrival of a cme to 1au is strongly correlated with true space speed and the subsequent deduction of initial acceleration / deceleration @xcite . from the reconstruction time series of true 3d positions
, we can derive a true space speed .
our primary source of data is from the sun - earth connection coronal and heliospheric investigation ( secchi , ) instruments on board stereo , as these return synchronized images from two viewpoints using multiple detectors .
stereo is a pair of identical spacecraft , a and b , launched into a leading and trailing earth orbit .
the twin spacecraft , equipped with identical instruments , observe the sun simultaneously from two different angles , allow for the first time 3d viewing of solar features such as cmes .
the initial fields of view are partially overlapping and , for the 31 dec 2007 event , the satellites are at an angular spacing of 44 degrees .
the cor2 white light coronagraphs onboard each of the twin stereo spacecraft each cover a field of view from 2 to 15 solar radii .
all secchi cor2 images used are total brightness images acquired with an cadence of every 30 minutes .
we are primarily concerned with the density and kinematics of the 31 dec 2007 cme as derived from the cor2 observations , however , we briefly look at this cme in context . there is a pre - existing streamer , visible when median background subtraction is used ( figure [ figurecme ] ) but not visible in difference images , as it changes only slowly relative to the cor2 cadence .
a cme erupts from this active region , either as a streamer blow - out type event , a prominence eruption , or an impulsive lower event .
this temporarily changes the magnetic field topology and either completely dissipates the streamer or minimally deflects it . after the event , the basic magnetic field structure reforms and subsequently , the streamer reforms .
cmes are faint relative to the underlying f- and k - corona emission , so we use difference images to extract the cme from the background ( figure [ figureevo ] ) .
we subtract the 0:22ut quiet data frame from each image in order to better view the cme .
this subtracts the f - corona as well as any quasi - static non - cme structures .
we confine the kinematic analysis to the period during which the cme is visible in the cor2 field of view . by the time the cme has entered the cor2 field of view
, the cme front has become distorted and shows a distinct upper and lower lobe ( again , figure [ figurecme ] ) .
the cme event was first visible within the cor2 coronagraphs at around 1:52ut , and the leading edge begins to depart at or around 4:52ut .
we use base difference images as explained above , shown in figure [ figureevo ] .
we also create a running difference image consisting of a data frame minus the immediately prior ( by 30 minutes ) data frame , to extract just the most rapidly changing portion of the cme .
there is not yet one single best method in the fledgling field of solar reconstruction @xcite .
in addition to our choice of reconstruction method , we note that other methods are used to estimate the location of a cme .
image geometric constraints can localize a cme rapidly @xcite and , if the cme is fully in the field of view , deduce the total mass and center of mass @xcite .
polarization measurements allow 3d placement using a single satellite ( , ) .
methods such as tie - pointing and visual analysis @xcite manually draw where the presumed shock front or leading edge of the cme is , then use that to define the height - time profile for the cme . while the center of mass of a cme in a single image can be computed @xcite , this center of mass
can not be directly compared with that from a different satellite image unless it is known that the full cme mass is visible in both .
further , you can not define identical components , such as the upper and lower lobes seen in this 31-dec-2007 cme , in two views ( such as cor2-a and cor2-b ) unless the cme is both equally visible in both detectors and perfectly aligned
. if the cme has any tilt or asymmetry , material from one region in cor2-a will be co - mingled with material viewed in cor2-b .
the solution is to reconstruct the entire underlying density , then split that 3d distribution into the appropriately - oriented regions to extract component masses and kinematics . with stereo , we have two simultaneous viewpoints of cmes and solar features at wide angles of separation that do not necessarily show the same morphological shape in all viewpoints , in which the cme may not be fully within the field of view of all viewpoints at all times .
we use a hybrid inverse reconstruction method to fit the underlying cme density distribution without requiring distinct identifiable features nor mass conservation .
this hybrid method is a type of full - cme modeling that solves for the most likely underlying density distribution that can give rise to the data observed . from the resulting 3d solution
, we can then produce total mass and regional section masses as well as reconstructing successive frames to obtain 3d kinematic information .
we solve for the bulk cme , generating results that differ from that of leading edge and shock front methods .
we also provide an equivalent for the leading edge , where we trace the center of mass of the most rapidly changing material through a running difference method .
a 3d reconstruction problem consists of selecting data , extracting the geometry of the view , rendering the data , then applying a reconstruction method .
inverse modeling is a non - parametric method that assumes no a priori shape for the cme and uses grid minimization to solve for an underlying 3d electron density distribution by comparing renderings with the actual data views taken .
inverse modeling is constrained by the limits on the overlapping viewpoints and can not provide model insight into regions that do not contain at least two partially overlapping fields of view .
in our work we use the two stereo viewpoints , in contrast to wide - field image work @xcite and rotational tomography @xcite .
in particular , the existing data are all along the ecliptic and lack an overhead view from the pole .
in contrast to inversion methods , which are non - parametric and produce non - unique solutions , forward modeling assumes a specific parametric shape and fits the model until it produces renderings that match the actual data .
forward modeling therefore provides a likelihood that a given model shape describes the actual cme event , but does not provide information on other models distributions that could produce the same data . in essence , inverse modeling is underconstrained and forward model is overconstrained .
given enough viewpoints , an inverse modeling solution will produce the best match for the actual underlying distribution . in regions
where the solution space is under - resolved and has insufficient viewpoints , inverse modeling often produces a non - physical solution . forward
modeling produces a physical solution based on model assumptions , but predisposes the solution to only fitting that model .
our hybrid method combines model - free inverse reconstruction supplemented with weighted model renderings for regions where the inverse reconstruction problem is unresolved . this balances model - centric approaches with the unbiased solution from inverse modeling . by creating synthetic model viewpoints at a lower weight than the actual data
, the inverse reconstruction can better converge on a physically meaningful result without being overly predisposed towards a single model assumption .
the inversion computes a presumed underlying density distribution in an attempt to reproduce the data .
the data need to be fit only to the tolerance set by the noise , since cor2-a and cor2-b views will have statistical deviations as well as instrument effects . for mass that is visible in one view but not another ( such as being behind the occulter in one view ) ,
that portion of the inversion solution is less constrained by lacking the second view , but will still place mass that does not violate either data view . as the cme evolves into and out of each field of view , there will be uncertainties in the mass due to fewer constraints , but the solutions themselves do not assume equal mass visible
. the final inversion result will be self - consistent with all available data , though it does not require that the solution be unique .
we use this hybrid of inverse and forward modeling to investigate the asymmetries of the 31 dec 2007 cme as observed by the secchi cor2 white light coronagraphs onboard the twin stereo satellites .
we observe a streamer and the cme superimposed in the image ( figure [ figurecme ] ) and see evidence for interaction of the two .
we bisect the cme into an upper and lower region using the line defined by the movement of the bulk center of mass radially out from the sun .
this vector also corresponds roughly to the observed longitude and latitude of the streamer seen in the cor2 images . by combining forward and inverse modeling into a single hybrid method ,
we reconstruct a probable underlying 3d density distribution for each pair of frames at a given time . from this
, we can sum the mass to obtain a total mass and center of mass position , or divide the mass to compare different 3d regions of the distribution .
we compare the total mass ( and total cm ) to that of the distinct upper and lower regions . by comparing the distribution over time
, we obtain the kinematics for the cme as a whole as well as for the bisecting regions .
the visible ( white light ) emission of the solar k - corona is due to thomson scattering of the incident photospheric emission by hot coronal electrons .
this scattered radiation is tangentially and radially polarized .
secchi coronagraphs mask the solar disk , whose brightness is more than @xmath1 that of the corona .
the coronagraph measures the brightness integrated over the line of sight through the optically thin corona .
applying g(x , y , z ) , a geometric function for the line of sight ( depending on limb darkening ) yields , for a given distance z to the line of sight , the following total brightness @xcite : @xmath2 rendering of a presumed underlying density distribution is generally solved using raytracing , where we sum the thomson scattering contribution of each 3d voxel element stepwise along the line of sight to obtain the total brightness at the given detector pixel .
we use difference images in order to subtract the f - corona and quasi - stable structures from the data .
the data that remains is presumed to be entirely k - corona white light emission due to the rapidly cme changing material .
to reconstruct , we first fit the cme loosely using forward modeling @xcite . a flux rope shaped density distribution
is manually positioned until its shape overlays with the observed shape in both the cor2-a and cor2-b data frames .
the model assumes a symmetric cme and serves as only a rough fit to the actual cme .
this model produces a likely enclosing volume for the cme , but does not provide us with information on mass or density .
second , we use this forward model to generate a simulated polar view of the cme ( figure [ figurefms ] ) . from this
, we create a polar mask to serve as a constraining envelope .
our subsequent modeling will thus require that the cme be fully contained within this 3d cylindrical envelope .
finally , we use the cor2-a and cor2-b data with the addition of the constraining polar mask to run an inverse reconstruction .
the inversion will place density within the enveloped region in a manner which , when rendered , will recreate the original observed data .
this allows the inversion to capture any asymmetries within the actual density distribution while preventing the inversion from placing material outside what theory ( via the forward model ) suggests is the constrained volume of space for the cme .
our forward modeling technique converges on parameters of a given geometric model a croissant flux rope shell model @xcite to find the best fit that model gives to the original data .
these fits are done by eye to place a flux rope that best matches the morphology seen by the multiple data views . for inverse modeling
, we converge on a parameter - free minimization of the underlying matter distribution . using a cube ( consisting of voxels , the 3d equivalent to 2d pixels ) , we first render a possible solution volume ,
compare that render to the original data , then adjust our guess at the solution iteratively until the renders match the data . because there are multiple solutions to the underlying distribution that can potentially reproduce the original data
, we try to create the simplest underlying density distribution which can reproduce the original data .
we currently use the pixon method , discussed in detail in .
pixon is a minimum complexity , non - parametric , locally adaptive , iterative image reconstruction method .
it seeks the smoothest solution that fits the data , using adaptive elements dubbed pixons to restrict the number of elements needed for the calculation .
the pixon elements smooth the data to reduce the number of image components . given an initial image , pixon finds a non - negative least - squares image , maps that image with pixon elements , then updates the solution image with the pixon map .
a solution with fewer pixon elements is deemed simpler and therefore more likely .
since the data are all gathered from the ecliptic plane , the reconstructions are not well constrained along the polar viewpoint . using a forward model
, we can create an artificial estimate at a polar view , and use that to restrict the spatial extent of the possible solution .
so we further constrain our inverse reconstruction by using a forward model of the cme to create an outer boundary envelope which contains fully the entire cme .
this restricts the volume of space in which the inverse reconstruction can place material , while freely allowing the reconstruction to distribute the material within that envelope in order to best match the data as seen .
we reconstruct the cme as a whole using 3 different constraining envelopes , shown in figure [ figurefms ] : \a ) fm series uses a series of stepwise flux rope forward models manually fit to time pair \b ) fm max uses a single largest extent forward model ( taken as the largest of those from the single steps above ) , \c ) alt fm refers to an alternative largest extent model fit using a powell iteration , that yielded a slightly different placement and extent for the constraining element . the inversion results did not vary greatly when comparing the choice of forward model constraining envelopes .
larger envelopes obviously allow a solution that has a larger volume , but the center of mass positions for selected regions do not differ , only the distribution of the material about that center of mass .
each of the three methods produced a similar center of mass position , and even testing slight variations in the extent or placement of the constraining envelope does not alter the positions of the center of mass , despite distributing the solution into a slightly different volume .
our detailed analysis uses the fm max case with the largest extent forward model .
analyzing very fine structure using this method would be biased in cases where the structure desired conflicts with the size of the envelope .
for example , investigation of edge densities would show a dependence on the envelope choice because the envelope can affect the compactness of the solution . since the envelope only constrains the z - axis , we can separate the cme into horizonal slices without worrying about the choice of envelope biasing the results .
we split the cme into an upper and lower segment along @xmath3 ecliptic line ; defining these regions is independent of constraining envelope as well .
we binned down the @xmath4 cor2 data to @xmath5 images and created an underlying density cube of size @xmath6 .
given the cor2 field of view ( fov ) of @xmath7 at 1au , the per - pixel fov @xmath8 corresponds to a voxel size of @xmath9 .
we limit this analysis to aggregates of voxels rather than individual voxels , since at a separation angle of @xmath10 the problem is still slightly underconstrained . for the bulk cme position , we achieve with each of the 3 methods a latitude that is independent ( within @xmath11 ) of the specific constraining forward model chosen , table [ t - loc ] .
the choice of method did result in slightly larger deviations for computed center of mass longitude than for latitude , varying by @xmath12 depending on the specific forward model chosen as a constraint . while the difference images provide the whole observed cme mass , the running difference images highlight the more rapidly changing material , including the cme front . comparing both the difference and running difference reconstructions ,
the center of mass longitudes were comparable , but there was a stronger ( @xmath13 ) difference in the latitude of the cme path between the difference and running difference results .
.cme location determined from the difference and running difference hybrid reconstructions .
error brackets for the position show the effect of different assumed forward models .
latitudes and longitudes are in the heliospheric aries ecliptic reference frame .
carrington longitudes are also provided . [ cols="^,^,^,^,^",options="header " , ] the front of the 31 dec 2007 cme is highly distorted as it travels across the cor2 field of view ( 2 - 15 @xmath14 ) .
purely geometric projection effects of a flux rope model are not sufficient to reproduce this shape and we contend it is intrinsically distorted or disrupted so as to show the chevron shape of upper lobe , sparse middle , and lower lobe . in these ranges , the cme is expanding .
figure [ figuredt ] shows the `` true '' height - time profile and true space speed for this cme .
the true space speed is taken from the radial distances from the sun as derived from the 3d reconstruction .
the running difference results , also in figure [ figuredt ] , emphasizes the faster - moving material and therefore weight the front shock more heavily , with a more rapid movement compared to the bulk of the cme .
we plot the radial velocities for the cme , similar to the height - time plots , in figure [ figurekin ] .
the bulk cme motion ( for the entire cme and for each segment ) has an initial acceleration then is relatively flat at approximately @xmath15 . the more rapidly changing material , as estimated from the running difference images , has an initial cme speed of @xmath16 .
as the cme spreads out , the running difference measure converges to the slower bulk cme speed .
the upper and lower portions of the cme have different velocities and slightly different directions , reinforcing our contention that a single cme speed measure does not suffice in describing the propagation of the cme outwards .
the deviation of the base difference and running difference indicates qualitatively the expansion rate of the cme .
however , we note that running difference images include material through the entire cme and therefore provide a slower true space speed than visual tracking of just the shock front .
to measure the expansion rate of the cme , we extracted the density of occupied voxels ( those voxels containing material at or above a 1-sigma standard deviation of the spatial mean ) . within occupied voxels , we calculate a mean density for that segment , plotted in figure [ figuredensity ] .
although the cme bulk centers of mass are advancing at a similar rate , the upper portion of the cme shows clearly an increasing large expansion , and corresponding lower average density , despite a slightly slower velocity ( figure [ figurekin ] ) .
we suggest that both the upper and lower portions have similar masses and initial motion , and the asymmetries seen are not purely projection effects but reflect the inner heliosphere medium the cme is moving into .
the upper and lower regions are split along the line located along the pre - cme streamer .
the upper portion of the cme is directed along the ecliptic plane and therefore is the portion of the cme that will first arrive at 1au . the lower portion , due to the cme latitude ( @xmath17 ) and the streamer separation , can be treated as a separate entity possessing the same initial velocity impulse as the upper portion but moving into a different medium and therefore evolving differently .
hybrid reconstruction provides mass , position and density information for a cme viewed from multiple viewpoints .
in contrast to reconstruction methods that apply a monolithic single model or track only a specific feature such as a leading edge , we derive the spatial extent and kinematics of the cme as a whole , and for the separate upper and lower halves , in order to track individual components of a cme
. additional selection of sub - regions of the cme can be analyzed directly from the density cubes created by reconstruction . however , sub - regions should be defined based not on arbitrary geometric positions but on whether they are bonafide physical congregations of material .
we restricted ourselves to analyzing the upper and lower lobes as they were the most prominent asymmetrical features .
for the 31 dec 2007 cme event we studied , the chevron shape of the cme , with the bright observed upper and lower lobes and fainter middle region , is intrinsic to the cme and not just a projection effect .
the evolution of the upper and lower regions are distinct enough that they should be modeled as separate entities .
the upper region is initially moving faster but has more deceleration than the lower region . while the estimated masses of the upper and lower regions are very similar ( at roughly 50% the total cme mass ) , the upper region has a consistently lower reconstructed density and therefore occupies a larger volume of space than the lower region .
the expansion of both regions maintain self - similar shapes but are sufficiently distinct from each other that the rates of their expansion differ
. some cmes ( e.g. 12-dec-2008 , paper in preparation ) have a strongly symmetric shape throughout their propagation , while others such as this 31-dec-2007 event show asymmetry increasing as they propagate .
the cause of such chevron shapes is not yet understood but such inhomogeneities have been seen since the soho era @xcite .
comparison of the true space speed of the two portions during the cor2 observations supports the contention that cme evolution in the inner heliosphere can be expressed as self - similar , shape preserving expansion into an interplanetary medium of differing density @xcite .
the upper and lower portions had an identical initial impulse , substantiated by the strongly symmetric shape of the cme at distances closer to the sun , which develops into a chevron shape as time evolves .
exact symmetry between the upper and lower regions depart over time as the top is rapidly becoming more rarefied as the cme evolves .
such asymmetry should be considered when determining the true space speeds and evolution of the cme at increasing distances from the sun .
the authors thank the stereo secchi consortia for supplying their data .
portion of this work were funded by the secchi mission .
the secchi data used here were produced by an international consortium of the naval research laboratory ( usa ) , lockheed martin solar and astrophysics lab ( usa ) , nasa goddard space flight center ( usa ) , rutherford appleton laboratory ( uk ) , university of birmingham ( uk ) , max - planck - institut for solar system research ( germany ) , centre spatiale de liege ( belgium ) , institut doptique theorique et appliquee ( france ) , and institut dastrophysique spatiale ( france).w brueckner , g. e. , howard , r. a. , koomen , m. j. , korendyke , c. m. , michels , d. j. , moses , j. d. , socker , d. g. , dere , k. p. , lamy , p. l. , llebaria , a. , bout , m. v. , schwenn , r. , simnett , g. m. , bedford , d. k. and eyles , c. j. : 1995 , _ solar phys . _ * 162 * , 357 delaboudinire , j .-
p . , artzner , g. e. , brunaud , j. , gabriel , a. h. , hochedez , j. f. , millier , f. , song , x. y. , au , b. , dere , k. p. , howard , r. a. , kreplin , r. , michels , d. j. , moses , j. d. , defise , j. m. , jamar , c. , rochus , p. , chauvineau , j. p. , marioge , j. p. , catura , r. c. , lemen , j. r. , shing , l. , stern , r. a. , gurman , j. b. , neupert , w. m. , maucherat , a. , clette , f. , cugnon , p. , and van dessel , e. l. : 1995 , _ solar phys . _
* 162 * , 1 - 2 , 291 | we present a hybrid combination of forward and inverse reconstruction methods using multiple observations of a coronal mass ejection ( cme ) to derive the 3d true height - time plots for individual cme components .
we apply this hybrid method to the components of the 31 dec 2007 cme .
this cme , observed clearly in both the stereo a and stereo b cor2 white light coronagraphs , evolves asymmetrically across the 15 solar radius field of view within a span of three hours .
the method has two reconstruction steps .
we fit a boundary envelope for the potential 3d cme shape using a flux rope - type model oriented to best match the observations . using this forward model as a constraining envelope ,
we then run an inverse reconstruction solving for the simplest underlying 3d electron density distribution that can , when rendered , reproduce the observed coronagraph data frames .
we produce plots for each segment to establish the 3d or `` true '' height - time plots for each center of mass as well as for the bulk cme motion , and use these plots along with our derived density profiles to estimate the cme asymmetric expansion rate . |
information about the health outcomes in many epidemiological studies is obtained from multiple data sources or over a certain time - period with multiple observations . the multiple data sources provide multiple measures of the same underlying variable , measured on a similar scale . as an example @xmath0
adults are chosen for high - blood pressure study at the age of @xmath6 and are asked about their diet , smoking and drinking habits and their blood pressures are measured .
the same subjects over the course of time are monitored again on the basis of the same variables choice as before .
this illustrates the standard data collecting exercise to measure health risk .
once such a data is available we can begin the risk modelling to ascertain the factors which contribute towards high - blood pressure and those that contribute towards low - blood pressure .
to put it mathematically , given @xmath0 subjects and @xmath1 data sources or time - points the data is collected as a @xmath7 dimensional vector of covariates , @xmath8 where @xmath9 and @xmath10 .
given such a vector the outcome is reported as a variable @xmath4 for the @xmath2 subject and from the @xmath11 data source .
then we can construct a @xmath1 dimensional vector @xmath12 as @xmath3 . in a @xmath1 dimensional space the above vector for a given value of @xmath13 is a point . since @xmath9 the total number of points in the @xmath1 dimensional space equal @xmath0 .
these @xmath0 points follow a certain distribution which is a - priori unknown . in the conventional analysis the outcome variables @xmath5s are treated as independent variables and nothing
is assumed about the joint distribution in the @xmath1 dimensional space .
the assumption about the independence is not correct but as we will see in section [ [ tradition ] ] this does not affect the statistical analysis that we intend to carry out . + we propose a novel analysis tool to carry out health risk modelling by transforming the @xmath1 dimensional a - priori unknown density to that of a gaussian density whilst keeping the shannon entropy constant . to do
so we transform the @xmath0 number of @xmath1 dimensional vectors in a basis set consisting of divergence - free vector fields .
the condition that the basis set can only consist of divergence - free vector fields enforces the entropy conserving condition .
entropy conserving condition can be enforced by having volume preserving maps and our choice of the basis set represents a flow of incompressible fluid @xcite hence is a volume preserving map . to determine the coefficients of these basis vectors such that the @xmath1 dimensional density is a gaussian , karplus theorem
is used @xcite .
we will demonstrate in section[[vecs ] ] that how this theorem allows for determination of the basis coefficients in such a way that the @xmath1 dimensional density is transformed to a gaussian .
+ the paper is divided as follows , in section[[tradition ] ] the traditional approach to model health risk is reviewed and a novel approach is proposed . in section[[basis ] ] the construction of basis set consisting of high dimensional vector fields is shown . in section[[vecs ] ] the question of determining the coefficients of the basis set is settled .
the algorithm and program structure is discussed in section[[algo ] ] . in section[[test ] ] a test case is computed to see how the transformation works in practice .
finally we conclude our work done so far and make proposals for the future work .
consider a trial to model high - blood pressure with @xmath0 subjects monitored over @xmath1 time - period .
the response obtained from the @xmath2 subject can be written as a @xmath1 dimensional vector as @xmath14 traditionally a joint distribution for @xmath0 subjects is not specified .
instead a working generalized linear model ( glm ) to describe the marginal distribution of @xmath4 as in liang , zeger ( 1986 ) @xcite is used , @xmath15\label{marginal}\ ] ] if the output @xmath4 is a binary random variable then the parameters for the above exponential family are @xmath16 , \theta_{ij}=log\left [ \frac{\mu_{ij}}{1-\mu_{ij } } \right ] , b(y_{ij},\theta_{ij})=0\ ] ] the probability of a favorable outcome if @xmath4 is a binary random variable can be modelled via a logit function as @xmath17 ) = \vec{x}_{ij}\vec{\beta},\label{logit}\ ] ] where @xmath18 represents the favourable outcome _
i.e _ low risk of high - blood pressure whereas @xmath19 corresponds to high - risk of high - blood pressure .
+ given eq[[marginal ] ] the log - likelihood function can be written as @xmath20 to determine the regression parameters @xmath21 we differentiate the log - likelihood with respect to @xmath21 , this gives the following equation to estimate @xmath22s @xmath23 this is the traditional approach to determine the regression parameters that help us evaluate the factors which contribute towards high or low health risks given the data .
equation[[assump ] ] was derived assuming that the outcome @xmath4 is a binary random variable , however similar equation can be derived when the outcome @xmath4 is of any other type .
+ in this approach it is assumed that all the @xmath5s are independent observations . this assumption although not correct , yields the estimates for @xmath21 which are valid but their variances are not .
however using techniques such as empirical variance estimator valid standard errors can be obtained@xcite .
having reviewed the traditional approach we now present our idea .
as remarked earlier the joint distribution of the @xmath0 subjects is not specified in the @xmath1 dimensional space as this is an a - priori unknown .
however if the unknown distribution is transformed to that of a @xmath1 dimensional gaussian then a valid analysis tool can be developed .
we have developed an algorithm and a program which does this transformation in an entropy conserving way _ i.e _ the shannon entropy is preserved during the transformation .
+ to preserve the shannon entropy requires transformation of the high dimensional vectors in a basis set consisting of divergence - free vector fields as that represents a volume preserving map thereby preserving shannon entropy .
such a construction of orthocomplete basis set is available in any dimensions @xcite .
we use this mathematical construct to transform then @xmath0 number of @xmath1 dimensional vectors . to determine the basis coefficients such that the @xmath1 dimensional density is transformed to that of a gaussian karplus theorem discussed in section[[vecs ] ]
is invoked .
once the coefficients are determined the @xmath1 dimensional joint distribution of the subjects is a gaussian having the same shannon entropy as the starting distribution .
this has been implemented in entra .
+ once the joint distribution for the @xmath0 subjects is known log - likelihood function can be written for such a distribution and the regression parameters determined thereby . in the next section
we show the construction of the divergence - free vector fields used in our program entra .
a mathematically rigorous construction of divergence free smooth vector fields in any dimensions was provided in @xcite .
we use it to construct a basis set consisting of @xmath1 dimensional divergence free vector fields . to do
so we define the following @xmath24 matrix valued operator @xmath25 here the first term is a @xmath24 dimensional laplacian operator , @xmath26 being the @xmath24 identity matrix and the second term consists of column and row vectors of the gradient operator in @xmath1 dimensions .
this operator acts on a smooth scalar function which we construct from a @xmath1 dimensional vector @xmath27 as @xmath28 where the symbol @xmath29 is the euclidean distance between two @xmath1 dimensional vectors .
@xmath30 is chosen to be @xmath31 where @xmath32 is the spacing between the basis vectors .
the vector @xmath33 is chosen as a constant @xmath34 where @xmath35 goes from @xmath36 .
now we define a matrix valued function by applying the operator in eq[[operator ] ] to the scalar field in eq[[scalar ] ] as @xmath37 e^{- \arrowvert\arrowvert \vec{x}-\vec{x}_l\arrowvert\arrowvert^2/(2\sigma^2 ) } \end{aligned}\ ] ] it was proven rigorously that the columns of the above matrix consist of divergence free vector fields @xcite . for a given choice of centre we therefore obtain @xmath24 dimensional vector field . from the results in @xcite @xmath38 .
+ for a given centre @xmath33 there are @xmath1 number of @xmath1 dimensional mutually orthogonal basis vectors .
hence for each centre we have a complete basis set .
due to such a construction the basis vectors enforce the divergence free condition strictly .
each vector has a unique coefficient @xmath39 attached to it,@xmath40 . in the next section
we demonstrate this for a simple @xmath41 case .
to demonstrate how the vector field looks like we take a simple 2d case and define the scalar operator in eq[[scalar ] ] as @xmath42 the matrix valued function @xmath43 becomes @xmath44 the vectors @xmath45 and @xmath46 defined from the columns of the above matrix as below are then divergence free , @xmath47 @xmath48 any linear combination of these divergence free fields is also divergence free .
we plot these fields in fig.[div ] for @xmath49 and @xmath50 . and @xmath46 respectively . ,
title="fig:",scaledwidth=52.0% ] ; and @xmath46 respectively .
, title="fig:",scaledwidth=52.0% ] from the above plot we can see that for the @xmath41 case we have two mutually orthogonal divergence free basis vectors which constitute the complete basis set in two dimensions .
the result holds in general for any dimensions @xcite .
in the @xmath1 dimensional space of the outcome variable the @xmath0 subjects are represented by @xmath0 points .
these @xmath0 points are arranged according to a certain density @xmath51 .
karplus theorem states that for a given covariance gaussian distribution maximizes entropy .
mathematically this can be written as @xmath52=s[\rho_g]-s[\rho]\geq 0.\label{basic}\ ] ] here @xmath53 is the gaussian density .
the covariance matrix is defined as @xmath54 here the symbol @xmath55 denotes the ensemble average over the @xmath0 subjects implying that @xmath56 is a @xmath24 dimensional matrix .
@xmath27 denotes a @xmath1 dimensional vector or a point in the configuration space . + the equality sign in eq[[basic ] ] holds only if the underlying density distribution is a gaussian .
now we introduce the transformation @xmath57 that preserves the entropy , i.e , @xmath58=s[\rho].\ ] ] with this transformation we want to deform the density @xmath51 towards a gaussian density , this then becomes the following minimization problem @xcite @xmath59 \right).\ ] ] here @xmath60 is the group of all the smooth entropy preserving transforms .
since the transformation leaves the entropy unchanged and as the entropy of a gaussian density is proportional to the determinant of the covariance matrix , to solve the above minimization problem we have to minimize the determinant of the covariance matrix of the gaussian density .
since by karplus theorem the covariance of @xmath53 is same as that of @xmath51 , we can use the covariance in eq[[covar ] ] and write the above minimization problem as @xmath61.\ ] ] as a consequence of the above corollary if we have a basis set consisting of divergence free vector fields under which @xmath1 dimensional vectors for @xmath0 subjects are transformed , then the basis coefficients can be determined by minimizing the covariance matrix determinant of the transformed vectors with respect to the basis coefficients . in the next section
we describe the construction of the basis set consisting of divergence free smooth vector fields .
this construction is implemented in the program package entra , that i have developed .
given @xmath0 subjects the outcome variable for each subject is a @xmath1 dimensional vector[[master ] ] .
we compute the covariance matrix eq[[covar ] ] which is a @xmath24 dimensional matrix . the matrix is then diagonalized by an orthogonal transformation @xmath62 .
the covariance matrix @xmath56 can then be written as @xmath63 with @xmath64 being the eigenvalue matrix and columns of @xmath62 matrix being the eigenvectors of @xmath56 .
we construct a @xmath65 dimensional vector by appending all the @xmath0 number of @xmath1 dimensional vectors and label it as trajectory @xmath66 . to transform the trajectory to principal coordinates where the mean of the trajectory is centered at @xmath67 we project the eigenvectors onto the trajectory to get mean - centered trajectory as @xmath68 for the trajectory @xmath69 we choose two @xmath1 dimensional vectors @xmath70 and @xmath71 and transform them in the vector field shown before as @xmath72 @xmath73 the coefficients are chosen by minimizing the determinant of matrix @xmath56 with respect to the basis coefficients @xmath74 .
the minimization is performed via conjugate gradients method @xcite .
the process is repeated for all the @xmath0 number of @xmath1 dimensional vectors .
this in the end yields a trajectory @xmath75 which has the least determinant of the covariance matrix and the underlying density has the same shannon entropy as that of our starting system .
+ the classes that build up the core of the program are entra , trajectory and grid
. objects of class trajectory represent real - valued arrays .
the arrays are stored as high dimensional vectors and the dimensionality of the vectors is defined by the grid class .
the grid class also defines the number and spacing of the basis vectors .
the methods provided by trajectory and grid classes allow to initialize the corresponding arrays , to manipulate them , and to store ( load ) them to ( from ) files .
to start using the program few parameters have to be provided these are : + long nsources = j ; + long nsubjects = n ; + double deltx = spacing between basis functions ; + long ngpsx = number of basis functions = l ; + long max iter = maximum iteration to find basis coefficients ; +
the goal of this example is to generate a trajectory having a random underlying density and then to transform it towards a trajectory having gaussian underlying density . to generate a random trajectory we use the utility @xmath76 of eigen to generate random matrix of any dimension .
the parameters chosen for this test are as follows : + long nsources=10 ; + long nsubjects=1000 ; + double deltx = 0.05 ; + long ngpsx=80 ; + long maxiter=500 ; results of the transformation ( after a single iteration cycle ) are shown in fig.[[results ] ] . in this example
we transform a @xmath77 dimensional vector having @xmath78 configurations .
the data is along these @xmath77 axises which are labelled as @xmath79 .
here @xmath80 , _
i.e _ @xmath81 component of the @xmath1 dimensional vector .
we plot two dimensional subspace of original @xmath77 dimensional space along different axises as labelled in fig.[[results ] ] . to prove that the transformation is indeed entropy conserving , we computed subspace entropy via histogramming for the 2d subspaces shown in fig.[[results ] ] for the original and the transformed subspaces .
+ the matlab code file along with the output to compute entropy of subspaces is provided here : + * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * + matlab file to estimate entropy via histogramming for the plot in fig.[[results]]a + * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * + load simulation_examplehist.dat x1=simulation_examplehist(:,8 ) ; x2=simulation_examplehist(:,9 ) ; x3=simulation_examplehist(:,11 ) ; x4=simulation_examplehist(:,12 ) ; x11=[x1;x3 ] ; x12=[x2;x4 ] ; x = [ x11,x12 ] ; plotmatrix(x ) ; defaultn=500 ; error(nargchk(1 , 2 , nargin ) ) ; if nargin @xmath82 n = defaultn ; end x = double(x ) ; xh = hist(x ( : ) , n ) ; xh = xh / sum(xh ( : ) ) ;
i = find(xh ) ; h = -sum(xh(i ) .
* log2(xh(i ) ) ) ; initialentropy = h ; display(initialentropy ) ; y1=simulation_examplehist(:,2 ) ; y2=simulation_examplehist(:,3 ) ; y3=simulation_examplehist(:,5 ) ; y4=simulation_examplehist(:,6 ) ; y11=[y1;y3 ] ; y12=[y2;y4 ] ; y=[y11,y12 ] ; plotmatrix(y ) ; error(nargchk(1 , 2 , nargin ) ) ; if nargin @xmath82 n = defaultn ; end y = double(y ) ; yh = hist(y ( : ) , n ) ; yh = yh / sum(yh ( : ) ) ; i = find(yh ) ; htwo = -sum(yh(i ) .
* log2(yh(i ) ) ) ; transformedentropy = htwo ; display(transformedentropy ) ; entropydifference = initialentropy - transformedentropy ; display(entropydifference ) ; * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * + result of the above file + * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * + @xmath83 run simpleentropy initialentropy = 8.7636 transformedentropy = 7.5625 entropydifference = @xmath84 + * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * result for the plot fig.[[results]]b + @xmath83 run simpleentropy initialentropy = 8.7770 transformedentropy = 7.6632 entropydifference = @xmath85 + * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * result for the plot fig.[[results]]c + @xmath83 run simpleentropy initialentropy = 8.7754 transformedentropy = 7.6483 entropydifference = @xmath86 two dimensional histogram plots for initial and transformed configurations are shown in fig.[[results ] ] .
now we can also compute entropy in higher dimension via histogramming . as the data is 30 dimensional
we now look at the data along first three axises @xmath87 as seen in fig.[[threedorg ] ] , in this plot we plot the data along one axis versus another as labelled along with histogram along each axis .
we also plot the same plot for the transformed data as seen in fig.[[threedtrans ] ] .
we compute the entropy of the three dimensional data using a matlab code similar to above and its results are shown here : * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * result for computation of entropy for 3d data @xmath83 run highdimentropy initialentropy = 8.8220 transformedentropy = 7.5756 entropydifference = @xmath88 .
more details in the text , scaledwidth=100.0% ] .
more details in the text , scaledwidth=100.0% ] these results imply that in just a single iteration cycle the unknown configuration space density is transformed to a gaussian density with entropy being conserved approximately . clearly @xmath78 points are not sufficient to get an accurate entropy estimate and more statistics is required .
this will form part of the work to be done .
entra package has been presented . the aim of this package is to do data transformation on high dimensional data sets as found in epidemiology . with this transformation
the underlying high dimensional density function is transformed that to a high dimensional gaussian and due to the nice properties associated with a gaussian distribution the further data analysis can be accomplished easier than before .
\1 ) the appropriate choice of the basis set vectors .
the number of basis vectors depend on the data and need to be appropriately estimated beforehand .
how exactly that can be achieved needs to be determined .
+ 2 ) building an example with enough statistics to be able to prove that the entropy conservation is maintained to a high degree of accuracy .
+ 3 ) furthermore developing full file support for epidemiologists to enable them to load their data and get the transformed data .
+ 4 ) also , developing complete regression analysis to estimate conditional probabilities of the likes in equation[[logit ] ] in the ecosystem of entra .
i. andricioaei , and m. karplus , on the calculation of entropy from covariance matrices of the atomic fluctuations , j. chem .
phys . , 115:6289 - 6292 , 2001 . m. karplus and j.n .
kushick , method for estimating the configurational entropy of macromolecules .
macromolecules , 14(2):325332 , 1981 .
schlitter , j. ( 1993 ) .
estimation of absolute entropies of macromolecules using the covariance matrix
. chemical physics letters , * 215 * , 617621 .
liang k - y , zeger sl .
longitudinal data analysis using generalized linear models .
biometrika 1986 ; * 73*:13 22 . | the traditional approach of health risk modelling with multiple data sources proceeds via regression - based methods assuming a marginal distribution for the outcome variable .
the data is collected for @xmath0 subjects over a @xmath1 time - period or from @xmath1 data sources .
the response obtained from @xmath2 subject is @xmath3 . for @xmath0 subjects
we obtain a @xmath1 dimensional joint distribution for the subjects . in this work
we propose a novel approach of transforming any @xmath1 dimensional joint distribution to that of a @xmath1 dimensional gaussian keeping the shannon entropy constant .
this is in stark contrast to the traditional approaches of assuming a marginal distribution for each @xmath4 by treating the @xmath5s as independent observations .
the said transformation is implemented in our computer package called entra . |
we present a study of parametric interactions between laser pulses in a coherent ensemble of two - level quantum objects ( `` atoms '' ) .
the main attention is paid to the processes that can be obtained in the strong coupling regime of light - matter interaction , which is achieved , when ( i ) the high frequency of field - matter excitation exchanges exceeds the decoherence rates , ( ii ) the external field is not strong enough to entirely determine the evolution of a system : it is the reemission field that plays a key role and provides the collective behaviour of atoms .
recently , this regime has attracted attention in quantum optics with both atoms and nanostructures such as quantum wells and dots ( cavity qed , dicke effects , microcavity exciton - polariton parametric scattering , squeezing and entanglement @xcite ) . in both fields , it is considered as one of the key models for quantum information processing ( qip ) . in this context , the works on single or small number of objects ( photons , atoms , excitons ) as qubits should be mentioned @xcite . during last years , there is a growing interest in the study of macroscopic ensembles of quantum objects , which can be used as collective elements for qip protocols @xcite .
quantum memories @xcite , sources of single photons @xcite , entanglement of ensembles @xcite and new sources of entangled fields based on optically dense atomic ensembles were recently proposed .
the advantages of such collective objects in contrast to single - particle qubits were highlighted @xcite .
investigations of the collective effects in the strong coupling regime may be of especial importance , because in this regime , some purely quantum properties of the phenomena do not decrease with increasing number of particles @xcite , which will be important for qip with both discrete and continuous variables . in ref .
@xcite , a scheme for generation of twin polaritons ( coupled light - matter excitations ) in semiconductor microcavities with quantum wells was developed .
it was based on the parametric scattering of microcavity exciton - polaritons in the strong coupling regime . in this scheme
, very strong quantum correlations between polaritons were obtained inside a cavity . nevertheless , since only one polariton branch of the dispersion curve was used in the parametric process proposed , quantum correlations in the outgoing light fields were essentially reduced . in this report , we present a study of parametric processes both in a cavity and free space , which arise due to collective energy exchange between field and matter systems ( i.e. , beating between two polariton branches of the dispersion curve ) @xcite .
the most significant manifestation of the strong coupling regime consists in the resonant density - dependent splitting of the dispersion curve into two polariton branches .
the splitting appears , when the collective coupling constant ( cooperative frequency ) of a medium @xmath0 exceeds all rates of incoherent relaxations @xmath1 : @xmath2 where @xmath3 and @xmath4 are the dipole moment and frequency of a transition , @xmath5 and @xmath6 are the density and number of atoms , @xmath7 is the single - atom coupling constant .
thus , in dense media , photons with equal wave vectors but different frequencies exist .
if a spatial spectrum of a problem is fixed by a single - mode cavity , beating between these frequencies corresponds to vacuum rabi oscillations .
a spectral doublet , arising in this case , represents density - dependent vacuum rabi splitting of a cavity mode , which is proportional to @xmath0 ( 1 ) .
the origin of this effect can be traced to linear interaction of light and dipoles in a passive cavity @xcite .
the condition of weak field ( which can not destroy polariton dispersion ) is reduced to the statement that the photon number is smaller then the number of atoms . in this report
, we present our results related to the strong light - matter coupling in an active system : a multimode broadband laser with an intracavity narrowband coherent resonant medium without population inversion .
the condition of the strong coupling ( 1 ) can be fulfilled with respect to that macroscopic resonant absorber .
a phenomenon of spectrum condensation was analysed .
it consists in the dramatic modification of a lasing spectrum under transition from the weak- to strong - coupling regime of light - matter interaction : instead of usual saturated absorption line , a bright narrow doublet of generation appears .
self - splitting of a dye laser spectrum at different moments of a ne discharge afterglow , which corresponds to decreasing atomic density in an intracavity cell , @xmath8 nm . ] figure 1 presents dye laser spectra obtained with a neon discharge as an intracavity absorber at different densities @xmath5 ( up to @xmath9 @xmath10 corresponding to @xmath11 up to 10 ghz ) of neon atoms in the metastable state , which was used as a lower state of a two - level system .
it was shown , that the density dependence of the doublet splitting is in agreement with that of @xmath0 ( 1 ) , which determines the vacuum rabi splitting . moreover , the the existence of a density threshold , which is also determined by eq .
( 1 ) , was demonstrated . on the basis of the cooperative parametric resonance model @xcite ,
this effect was treated as lasing at collective vacuum rabi sidebands due to parametric interactions between laser modes .
the main difference of the free - space interaction consists in the fact that the spatial spectrum of a problem is not fixed by a cavity . as a consequence , under the linear propagation of a short pulse , no coherent density - dependent spectral features ( such as vacuum rabi splitting in a cavity )
can be extracted from spectral measurements , except for the trivial appearance of an incoherent absorption line . nevertheless , in our work , we have shown that such coherent features can be obtained under nonlinear parametric interaction of laser pulses ( the details of the theory can be found in ref .
@xcite , experimental results are presented in ref .
in contrast to cavity oscillations , they were shown to appear due to free - space optical ringing , which does not originate from beating between waves of equal wave vectors , but from the successive beats between polaritonic wave packets of equal group velocities .
long coherent oscillations can be obtained due to considerable reducing of the group velocity in a dense medium and due to its essential variation over the broad spectrum of a short laser pulse .
the oscillation frequency at the initial stage of the ringing @xmath12 should exceed the decoherence rates @xmath1 : this frequency increases with propagation distance @xmath14 and medium density @xmath5 .
thus , it is proportional to the number of interacting atoms , which directly demonstrates the superradiant character of the field reemission by the atomic ensemble .
transmission spectra of the probe field .
( a ) linear transmission ( a ) and @xmath15 ( b ) , @xmath16 ; ( b ) parametric amplification at @xmath17 ( a ) and dumping at @xmath18 ( b ) for @xmath16 ; ( c ) parametric amplification at @xmath19 for @xmath20 ( a ) and for @xmath21 ( b ) ; @xmath22 , pump area @xmath23 . ] we considered the nonlinear parametric interaction between two short laser pulses propagating in a dense resonant atomic ensemble .
optical ringing was shown to significantly affect the propagation and amplification of a probe field under its interaction with a nearly copropagating pump .
depending on the probe - pump time delay @xmath24 ( the probe pulse precedes the pump for @xmath25 ) , the probe transmission spectra show either a specific doublet or dip , which corresponds to parametric sideband amplification or dumping of radiation [ cf .
the widths of these features are greater than the width of an incoherent absorbtion contour [ shown in fig .
2(a ) ] , they are determined by the density - dependent field - matter coupling constant and increase during the propagation [ cf .
2(b ) , ( c ) ] . the spectral features in fig .
2 can be explained as a result of parametric forward scattering of a probe pulse on spatiotemporal modulations of the population difference , appearing due to the optical ringing in a pump @xcite .
the characteristic frequencies , which increase with the number of interacting atoms , correspond to the frequency of optical ringing .
the condition of the weak fields , which do not destroy dispersion and collective behaviour of atoms , in this case , is equivalent to the requirement of the small input pulse area ( so that @xmath26 ringing should not be shadowed by 2@xmath27 solitons and rabi flopping ) .
contrary to strong - field parametric processes ( e.g. , due to transient rabi flopping or stationary mollow - boyd effect , which are determined by the pump amplitude ) , the density- and coordinate - dependent spectra of the probe display the importance of free - space collective oscillations and can not be obtained in the framework of a single - atom model . | we consider parametric interactions of laser pulses in a coherent macroscopic ensemble of resonant atoms , which are possible in the strong coupling regime of light - matter interaction .
the spectrum condensation ( lasing at collective vacuum rabi sidebands ) was studied in an active cavity configuration .
parametric interactions under the strong light - matter coupling were proved even in free space .
in contrast to bichromatic beats in a cavity , they were shown to appear due to interference between polaritonic wave packets of different group velocities . , , , , , quantum optics , dense atomic ensembles , dicke superradiance , collective effects , resonant parametric processes , polaritons |
the subject of arithmetic constraints on reals has attracted a great deal of attention in the literature .
for some reason arithmetic constraints on integer intervals have not been studied even though they are supported in a number of constraint programming systems .
in fact , constraint propagation for them is present in ecl@xmath0ps@xmath1 , sicstus prolog , gnu prolog , ilog solver and undoubtedly most of the systems that support constraint propagation for linear constraints on integer intervals . yet , in contrast to the case of linear constraints see notably @xcite we did not encounter in the literature any analysis of this form of constraint propagation .
in this paper we study these constraints in a systematic way .
it turns out that in contrast to linear constraints on integer intervals there are a number of natural approaches to constraint propagation for these constraints . to define them we introduce integer interval arithmetic that is modeled after the real interval arithmetic , see e.g. , @xcite
there are , however , essential differences since we deal with integers instead of reals .
for example , multiplication of two integer intervals does not need to be an integer interval . in passing by we
show that using integer interval arithmetic we can also define succinctly the well - known constraint propagation for linear constraints on integer intervals . in the second part of the paper
we compare the proposed approaches by means of a set of benchmarks .
we review here the standard concepts of a constraint and of a constraint satisfaction problem .
consider a sequence of variables @xmath2 where @xmath3 , with respective domains @xmath4 associated with them .
so each variable @xmath5 ranges over the domain @xmath6 . by a _ constraint _
@xmath7 on @xmath8 we mean a subset of @xmath9 . given an element @xmath10 of @xmath9 and a subsequence @xmath11 of @xmath8 we denote by @xmath12 $ ] the sequence @xmath13 .
in particular , for a variable @xmath5 from @xmath8 , @xmath14 $ ] denotes @xmath15 .
a _ constraint satisfaction problem _ , in short csp , consists of a finite sequence of variables @xmath8 with respective domains @xmath16 , together with a finite set @xmath17 of constraints , each on a subsequence of @xmath8 . we write it as @xmath18 , where @xmath19 and @xmath20 . by a _ solution _
to @xmath21 we mean an element @xmath22 such that for each constraint @xmath23 on a sequence of variables @xmath8 we have @xmath24 \in c$ ] .
we call a csp _ consistent _ if it has a solution and _ inconsistent _ if it does not
. two csps with the same sequence of variables are called _ equivalent _ if they have the same set of solutions . in what follows
we consider csps the constraints of which are defined in a simple language and identify the syntactic description of a constraint with its meaning being the set of tuples that satisfy it .
we view _ constraint propagation _ as a process of transforming csps that maintains their equivalence . in what follows
we define this process by means of proof rules that act of csps and preserve equivalence .
an interested reader can consult @xcite for a precise explanation of this approach to describing constraint propagation . to define the arithmetic constraints use the alphabet that comprises * variables , * two constants , 0 and 1 , * the unary minus function symbol ` @xmath25 ' , * three binary function symbols , ` + ' ,
@xmath25and ` @xmath26 ' , all written in the infix notation . by an _ arithmetic expression _ we mean a term formed in this alphabet and by an _ arithmetic constraint _ a formula of the form @xmath27 where @xmath28 and @xmath29 are arithmetic expressions and @xmath30 .
for example @xmath31 is an arithmetic constraint . here
@xmath32 is an abbreviation for @xmath33 and similarly with the other expressions .
if ` @xmath26 ' is not used in an arithmetic constraint , we call it a _ linear constraint _ . by an _ extended arithmetic expression _ we mean a term formed in the above alphabet extended by the unary function symbols ` @xmath34 ' and ` @xmath35{.}$ ] ' for each @xmath36 and the binary function symbol ` @xmath37 ' written in the infix notation .
for example @xmath38{(y^2 \cdot z^4)/(x^2 \cdot u^5 ) } \label{eq : extended}\ ] ] is an extended arithmetic expression . here
, in contrast to the above @xmath32 is a term obtained by applying the function symbol ` @xmath39 ' to the variable @xmath40 .
the extended arithmetic expressions will be used only to define constraint propagation for the arithmetic constraints .
fix now some arbitrary linear ordering @xmath41 on the variables of the language . by a _ monomial _ we mean an integer or a term of the form @xmath42 where @xmath43 , @xmath44 are different variables ordered w.r.t .
@xmath41 , and @xmath45 is a non - zero integer and @xmath46 are positive integers .
we call then @xmath47 the _ power product _ of this monomial . next , by a _ polynomial _ we mean a term of the form @xmath48 where @xmath49 , at most one monomial @xmath50 is an integer , and the power products of the monomials @xmath51 are pairwise different . finally , by a _ polynomial constraint _
we mean an arithmetic constraint of the form @xmath52 , where @xmath28 is a polynomial with no monomial being an integer , @xmath30 , and @xmath53 is an integer .
it is clear that by means of appropriate transformation rules we can transform each arithmetic constraint to a polynomial constraint .
for example , assuming the ordering @xmath54 on the variables , the arithmetic constraint ( [ eq : arithcon ] ) can be transformed to the polynomial constraint @xmath55 so , without loss of generality , from now on we shall limit our attention to the polynomial constraints . next ,
let us discuss the domains over which we interpret the arithmetic constraints . by an
_ integer interval _ , or an _ interval _ in short , we mean an expression of the form @xmath56\ ] ] where @xmath45 and @xmath53 are integers ; @xmath57 $ ] denotes the set of all integers between @xmath45 and @xmath53 , including @xmath45 and @xmath53 . if @xmath58 , we call @xmath57 $ ] the _ empty interval _ and denote it by @xmath59 . finally , by a _ range _
we mean an expression of the form @xmath60 where @xmath40 is a variable and @xmath61 is an interval .
to reason about the arithmetic constraints we employ a generalization of the arithmetic operations to the sets of integers . for @xmath62 sets of integers we define the following operations : * addition : @xmath63 * subtraction : @xmath64 * multiplication : @xmath65 * division : @xmath66 * exponentiation : @xmath67 for each natural number @xmath49 , * root extraction : @xmath68{x } : = { \mbox{$\{{x \in { \cal z } \mid x ^n \in x}\}$}},\ ] ] for each natural number @xmath49 .
all the operations except division are defined in the expected way .
we shall return to it at the end of section [ sec : third ] . at the moment it suffices to note the division operation
is defined for all sets of integers , including @xmath69 and @xmath70 .
this division operation corresponds to the following division operation on the sets of reals introduced in @xcite : @xmath71 for a ( n integer or real ) number @xmath45 and @xmath72 we identify @xmath73 with @xmath74 and @xmath75 with @xmath76 . to present the rules we are interested in we
shall also use the addition and division operations on the sets of real numbers .
addition is defined in the same way as for the sets of integers , and division is defined above . in @xcite
it is explained how to implement these operations .
further , given a set @xmath77 of integers or reals , we define @xmath78 @xmath79 when limiting our attention to intervals of integers the following simple observation is of importance .
[ note:1 ] for @xmath62 integer intervals and @xmath45 an integer the following holds : * @xmath80 , @xmath81 are integer intervals .
* @xmath82 is an integer interval .
* @xmath83 does not have to be an integer interval , even if @xmath84 or @xmath85 .
* @xmath86 does not have to be an integer interval .
* for each @xmath87 @xmath88 does not have to be an integer interval .
* for odd @xmath87 @xmath35{x}$ ] is an integer interval . * for even @xmath87 @xmath35{x}$ ] is an integer interval or a disjoint union of two integer intervals .
for example we have @xmath89 + [ 3 .. 8 ] = [ 5 .. 12],\ ] ] @xmath90 - [ 1 .. 8 ] = [ -5 .. 6],\ ] ] @xmath91 \cdot [ 1 .. 2 ] = { \mbox{$\{{3,6}\}$}},\ ] ] @xmath92/[-1 .. 2 ] = { \mbox{$\{{-5,-4,-3,2,3,4,5}\}$}},\ ] ] @xmath93/[-1 .. 2 ] = { \cal z},\ ] ] @xmath94 ^ 2 = { \mbox{$\{{1,4}\}$}},\ ] ] @xmath38{[-30 .. 100 ] } = [ -3 .. 4],\ ] ] @xmath95{[-100 .. 9 ] } = [ -3 .. 3],\ ] ] @xmath95{[1 .. 9 ] } = [ -3 .. -1 ] \cup [ 1 .. 3].\ ] ] to deal with the problem that non - interval domains can be produced by some of the operations we introduce the following operation on the subsets of the set of the integers @xmath96 : @xmath97 for example @xmath98/[-1 .. 2 ] ) = [ -5 .. 5]$ ] and @xmath99/[-1 .. 2 ] ) = { \cal z}$ ] . to define constraint propagation for the arithmetic constraints on integer intervals we shall use the integer set arithmetic , mainly limited to the integer intervals .
this brings us to the discussion of how to implement the introduced operations on the integer intervals .
since we are only interested in maintaining the property that the sets remain integer intervals or the set of integers @xmath96 we shall clarify how to implement the intersection , addition , subtraction and root extraction operations of the integer intervals and the @xmath100 closure of the multiplication , division and exponentiation operations on the integer intervals .
the case when one of the intervals is empty is easy to deal with .
so we assume that we deal with non - empty intervals @xmath101 $ ] and @xmath102 $ ] , that is @xmath103 and @xmath104 .
[ [ intersection - addition - and - subtraction ] ] intersection , addition and subtraction + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + it is easy to see that @xmath105 \cap [ c .. d ] = [ max(a , c) .. min(b , d)],\ ] ] @xmath105 + [ c .. d ] = [ a+c\ .. \ b+d],\ ] ] @xmath105 - [ c .. d ] = [ a - d\ .. \ b - c].\ ] ] so the interval intersection , addition , and subtraction are straightforward to implement . [
[ root - extraction ] ] root extraction + + + + + + + + + + + + + + + the outcome of the root extraction operator applied to an integer interval will be an integer interval or a disjoint union of two integer intervals .
we shall explain in section [ sec : first ] why it is advantageous not to apply @xmath100 to the outcome .
this operator can be implemented by means of the following case analysis .
_ suppose @xmath106 is odd
. then @xmath68{[a .. b ] } = [ { \left \lceil \sqrt[n]{a } \right \rceil } .. { \left \lfloor \sqrt[n]{b } \right \rfloor}].\ ] ] _ case 2 .
_ suppose @xmath106 is even and @xmath107 .
then @xmath68{[a .. b ] } = { \mbox{$\emptyset$}}.\ ] ] _ case 3 .
_ suppose @xmath106 is even and @xmath108 .
then @xmath68{[a .. b ] } = [ -{\left \lfloor |\sqrt[n]{b}| \right \rfloor } .. -{\left \lceil |\sqrt[n]{a^{+}}| \right \rceil } ] \cup [ { \left \lceil |\sqrt[n]{a^{+}}| \right \rceil } .. { \left \lfloor |\sqrt[n]{b}| \right \rfloor } ] \ ] ] where @xmath109 . [ [ multiplication ] ] multiplication + + + + + + + + + + + + + + for the remaining operations we only need to explain how to implement the @xmath100 closure of the outcome .
first note that @xmath110 \cdot [ c .. d ] ) = [ min(a ) ..
max(a ) ] , \ ] ] where @xmath111 .
using an appropriate case analysis we can actually compute the bounds of @xmath112 \cdot [ c .. d])$ ] directly in terms of the bounds of the constituent intervals .
[ [ division ] ] division + + + + + + + + in contrast , the @xmath100 closure of the interval division is not so straightforward to compute .
the reason is that , as we shall see in a moment , we can not express the result in terms of some simple operations on the interval bounds .
consider non - empty integer intervals @xmath113 $ ] and @xmath102 $ ] . in analyzing the outcome of @xmath114/[c ..
d])$ ] we distinguish the following cases . _ case 1 .
_ suppose @xmath115 $ ] and @xmath116 $ ] .
then by definition @xmath114/[c .. d ] ) = { \cal z}$ ] .
for example , @xmath117/[-2 .. 8 ] ) = { \cal z}.\ ] ] _ case 2 .
_ suppose @xmath118 $ ] and @xmath119 .
then by definition @xmath120/[c .. d ] ) = { \mbox{$\emptyset$}}$ ] .
for example , @xmath121/[0 .. 0 ] ) = { \mbox{$\emptyset$}}.\ ] ] _ case 3 .
_ suppose @xmath118 $ ] and @xmath122 and @xmath123 .
it is easy to see that then @xmath124/[c .. d ] ) = [ -e .. e],\ ] ] where @xmath125 . for example ,
@xmath126/[-2 .. 5 ] ) = [ -100 .. 100].\ ] ] _ case 4 .
_ suppose @xmath118 $ ] and either @xmath127 and @xmath128 or @xmath129 and @xmath130 .
then @xmath120/[c .. d ] ) = int([a .. b]/([c .. d ] - { \mbox{$\{{0}\}$}}))$ ] .
for example @xmath131/[-7 .. 0 ] ) = int([1 .. 100]/[-7 .. -1]).\ ] ] this allows us to reduce this case to case 5 below . _ case 5 .
_ suppose @xmath132 $ ] .
this is the only case when we need to compute @xmath120/[c ..
d])$ ] indirectly .
first , observe that we have @xmath133/[c .. d ] ) { \mbox{$\:\subseteq\:$}}[{\left \lceil min(a ) \right \rceil } .. { \left \lfloor max(a ) \right \rfloor}],\ ] ] where @xmath134 .
however , the equality does not need to hold here . indeed , note for example that @xmath135/[9 .. 11 ] ) = [ 16 .. 16]$ ] , whereas for @xmath136 , @xmath137 we have @xmath138 and @xmath139 .
the problem is that the value 16 is obtained by dividing 160 by 10 and none of these two values is an interval bound .
this complication can be solved by preprocessing the interval @xmath140 $ ] so that its bounds are actual divisors of an element of @xmath57 $ ] .
first , we look for the least @xmath141 $ ] such that @xmath142\ \exists u \in { \cal z}\ u\cdot c ' = x$ ] . using a case analysis ,
the latter property can be established without search .
suppose for example that @xmath143 and @xmath144 . in this case , if @xmath145 , then @xmath146 has the required property .
similarly , we look for the largest @xmath147 $ ] for which an analogous condition holds .
/ [ c .. d ] ) = [ \lceil\textit{min\/}(a)\rceil .. \lfloor\textit{max\/}(a)\rfloor]$ ] , where @xmath149 .
[ [ exponentiation ] ] exponentiation + + + + + + + + + + + + + + the @xmath100 closure of the interval exponentiation is straightforward to implement by distinguishing the following cases .
_ case 1 . _
suppose @xmath106 is odd
. then @xmath133^n ) = [ a^n .. \ : b ^n].\ ] ] _ case 2 .
_ suppose @xmath106 is even and @xmath150 .
then @xmath133^n ) = [ a^n .. \ : b ^n].\ ] ] _ case 3 .
_ suppose @xmath106 is even and @xmath151
. then @xmath133^n ) = [ b^n .. \ : a^n].\ ] ] _ case 4 .
_ suppose @xmath106 is even and @xmath152 and @xmath153
. then @xmath133^n ) = [ 0 .. max(a^n , b^n ) ] .\ ] ] given now an extended arithmetic expression @xmath28 each variable of which ranges over an integer interval , we define @xmath154 as the integer interval or the set @xmath96 obtained by systematically replacing each function symbol by the application of the @xmath100 operation to the corresponding integer set operation .
for example , for the extended arithmetic expression @xmath155{(y^2 \cdot z^4)/(x^2 \cdot u^5)}$ ] of ( [ eq : extended ] ) we have @xmath156{int(int(y^2 ) \cdot int(z^4))/int(int(x^2 ) \cdot int(u^5))}),\ ] ] where @xmath40 ranges over @xmath8 , etc .
the discussion in the previous subsection shows how to compute @xmath154 given an extended arithmetic expression @xmath28 and the integer interval domains of its variables .
the following lemma is crucial for our considerations .
it is a counterpart of the so - called ` fundamental theorem of interval arithmetic ' established in @xcite .
because we deal here with the integer domains an additional assumption is needed to establish the desired conclusion .
[ lem : correctness ] let @xmath28 be an extended arithmetic expression with the variables @xmath157 .
assume that each variable @xmath5 of @xmath28 ranges over an integer interval @xmath158 .
choose @xmath159 for @xmath160 $ ] and denote by @xmath161 the result of replacing in @xmath28 each occurrence of a variable @xmath5 by @xmath162 .
suppose that each subexpression of @xmath161 evaluates to an integer .
then the result of evaluating @xmath161 is an element of @xmath154 .
* the proof follows by a straightforward induction on the structure of @xmath28 .
even though we focus here on arithmetic constraints on integer intervals , it is helpful to realize that the integer interval arithmetic is also useful to define in a succinct way the well - known rules for constraint propagation for linear constraints . to this end consider first a constraint @xmath163 , where @xmath3 , @xmath164 are non - zero integers , @xmath165 are different variables , and @xmath53 is an integer .
to reason about it we can use the following rule parametrized by @xmath166 $ ] : _ linear equality _
where * for @xmath168 @xmath169 * @xmath170 - \{j\ } } a_i \cdot x_i)/a_j\big{)}.\ ] ] note that by virtue of note [ note:1 ] @xmath171 - \{j\ } } int(a_i \cdot d_i))/a_j.\ ] ] to see that this rule preserves equivalence suppose that for some @xmath172 we have @xmath173 .
then for @xmath166 $ ] we have @xmath174 - \{j\ } } a_i \cdot d_i)/a_j\ ] ] which by the correctness lemma [ lem : correctness ] implies that @xmath175 - \{j\ } } a_i \cdot x_i)/a_j\big{)},\ ] ] i.e. , @xmath176 . next , consider a constraint @xmath177 , where @xmath178 and @xmath53 are as above .
to reason about it we can use the following rule parametrized by @xmath166 $ ] : _ linear inequality
_ @xmath179 where * for @xmath168 @xmath169 * @xmath180 - \{j\ } } a_i \cdot x_i)/a_j)\ ] ] to see that this rule preserves equivalence suppose that for some @xmath172 we have @xmath181 .
then @xmath182 - \{j\ } } a_i \cdot d_i $ ] . by the correctness lemma [ lem : correctness ] @xmath183 - \{j\ } } a_i \cdot d_i \in int(b - \sigma_{i \in [ 1 .. n ] - \{j\ } } a_i \cdot x_i),\ ] ] so by definition @xmath184 - \{j\ } } a_i \cdot x_i)\ ] ] and consequently @xmath185 - \{j\ } } a_i \cdot x_i)/a_j\ ] ] this implies that @xmath176 .
we now move on to a discussion of constraint propagation for the arithmetic constraints on integer intervals . to illustrate the first approach consider the following example .
consider the constraint @xmath186 and the ranges @xmath187 $ ] and @xmath188 $ ] .
we can rewrite it as @xmath189{\frac{40 + x}{y } } \right \rfloor}\ ] ] since @xmath40 assumes integer values .
the maximum value the expression on the right - hand side can take is @xmath190{140 } \right \rfloor}$ ] , so we conclude @xmath191 . by reusing ( [ eq : x3 ] ) ,
now with the information that @xmath192 $ ] , we conclude that the maximum value the expression on the right - hand side of ( [ eq : x3 ] ) can take is actually @xmath190{45 } \right \rfloor}$ ] , from which it follows that @xmath193 . in the case of
@xmath194 we can isolate it by rewriting the original constraint as @xmath195 from which it follows that @xmath196 , since by assumption @xmath197 .
so we could reduce the domain of @xmath40 to @xmath198 $ ] and the domain of @xmath194 to @xmath199 $ ] .
this interval reduction is optimal , since @xmath200 and @xmath201 are both solutions to the original constraint @xmath202 .
more formally , we consider a polynomial constraint @xmath203 where @xmath204 , no monomial @xmath50 is an integer , the power products of the monomials are pairwise different and @xmath53 is an integer .
suppose that @xmath157 are its variables ordered w.r.t .
@xmath41 .
select a non - integer monomial @xmath205 and assume it is of the form @xmath206 where @xmath43 , @xmath207 are different variables ordered w.r.t .
@xmath41 , @xmath45 is a non - zero integer and @xmath46 are positive integers .
so each @xmath208 variable equals to some variable in @xmath209 .
suppose that @xmath210 equals to @xmath211 .
we introduce the following proof rule : _ polynomial equality _
@xmath212 where * for @xmath168 @xmath169 * @xmath213{int\left(({b- \sigma_{i \in [ 1 .. m ] - \{{\ell}\ } } m_i})/s\right)}\ \right)\ ] ] and @xmath214 to see that this rule preserves equivalence choose some @xmath172 .
to simplify the notation , given an extended arithmetic expression @xmath29 denote by @xmath215 the result of evaluating @xmath29 after each occurrence of a variable @xmath5 is replaced by @xmath15 .
suppose that @xmath216 .
then @xmath217 - \{{\ell}\ } } m'_i,\ ] ] so by the correctness lemma [ lem : correctness ] applied to @xmath218 - \{{\ell}\ } } m'_i$ ] and to @xmath28 @xmath219 - \{{\ell}\ } } m_i)/int(s).\ ] ] hence @xmath220{int(b- \sigma_{i \in [ 1 .. m ] - \{{\ell}\ } } m_i)/int(s)}\ ] ] and consequently @xmath221{int\left(({b- \sigma_{i \in [ 1 .. m ] - \{{\ell}\ } } m_i})/s\right)}\ \right)\ ] ] i.e. , @xmath176 .
note that we do not apply @xmath100 to the outcome of the root extraction operation .
for even @xmath222 this means that the second operand of the intersection can be a union of two intervals , instead of a single interval . to see why this is desirable , consider the constraint @xmath223 in the presence of ranges @xmath224 $ ] , @xmath225 $ ] . using the @xmath100 closure of the root extraction
we would not be able to update the lower bound of @xmath40 to 5 .
next , consider a polynomial constraint @xmath226 below we adopt the assumptions and notation used when defining the _ polynomial equality _ rule . to formulate the appropriate rule
we stipulate that for extended arithmetic expressions @xmath28 and @xmath29 @xmath227 with @xmath228 . to reason about this constraint we use the following rule : _ polynomial inequality _
@xmath229 where * for @xmath168 @xmath169 * @xmath230{int\left(^{\leq}(b- \sigma_{i \in [ 1 .. m ] - \{{\ell}\ } } m_i)/s\right)}\ \right)\ ] ] to prove that this rule preserves equivalence choose some @xmath172 .
as above given an extended arithmetic expression @xmath29 we denote by @xmath215 the result of evaluating @xmath29 when each occurrence of a variable @xmath5 in @xmath29 is replaced by @xmath15 .
suppose that @xmath231 .
then @xmath232 - \{{\ell}\ } } m'_i.\ ] ] by the correctness lemma [ lem : correctness ] @xmath233 - \{{\ell}\ } } m'_i \in int(b- \sigma_{i \in [ 1 .. m ] - \{{\ell}\ } } m_i),\ ] ] so by definition @xmath234 - \{{\ell}\ } } m_i).\ ] ] hence by the definition of the division operation on the sets of integers @xmath235 - \{{\ell}\ } } m_i)/int(s)\ ] ] consequently @xmath220{^{\leq}int(b- \sigma_{i \in [ 1 .. m ] - \{{\ell}\ } } m_i)/int(s)}\ ] ] this implies that @xmath176 .
note that the set @xmath236 - \{{\ell}\ } } m_i)$ ] is not an interval .
so to properly implement this rule we need to extend the implementation of the division operation discussed in subsection [ subsec : implementation ] to the case when the numerator is an extended interval .
our implementation takes care of this . in an optimized version of this approach
we simplify the fractions of two polynomials by splitting the division over addition and subtraction and by dividing out common powers of variables and greatest common divisors of the constant factors .
subsequently , fractions whose denominators have identical power products are added .
we used this optimization in the initial example by simplifying @xmath237 to @xmath238 .
the reader may check that without this simplification step we can only deduce that @xmath239 . to provide details of this optimization , given two monomials @xmath28 and @xmath29
, we denote by @xmath240\ ] ] the result of performing this simplification operation on @xmath28 and @xmath29 .
for example , @xmath241 $ ] equals @xmath242 , whereas @xmath243 $ ] equals @xmath244 . in this approach
we assume that the domains of the variables @xmath245 , @xmath246 of @xmath205 do not contain 0 .
( one can easily show that this restriction is necessary here ) . for a monomial @xmath28 involving variables ranging over the integer intervals that do not contain 0 , the set @xmath154 either contains only positive numbers or only negative numbers . in the first case we write @xmath247 and in the second case we write @xmath248 . the new domain of the variable @xmath211 in the _ polynomial inequality _
rule is defined using two sequences @xmath249 and @xmath250 of extended arithmetic expressions such that @xmath251\ \textrm{and}\ \frac{m_i'}{s_i ' } = - [ \frac{m_i}{s } ] \ \textrm{for $ i \in [ 1 .. m]$.}\ ] ] let @xmath252 - \{{\ell}\ } \}$ ] and for an extended arithmetic expression @xmath253 let @xmath254 - \{{\ell}\ } \mid s_i ' = t\}$ ] .
we denote then by @xmath255 the polynomial @xmath256 .
the new domains are then defined by @xmath257 { ^{\leq}int\left ( \sigma_{t \in s}\ \frac{p_{t}}{t}\right)}\ \right)\ ] ] if @xmath247 , and by @xmath257 { ^{\geq}int\left(\sigma_{t \in s}\ \frac{p_{t}}{t}\right)}\ \right)\ ] ] if @xmath248 . here
the @xmath154 notation used in the correctness lemma [ lem : correctness ] is extended to expressions involving the division operator on real intervals in the obvious way .
we define the @xmath100 operator applied to a bounded set of real numbers , as produced by the division and addition operators in the above two expressions for @xmath258 , to denote the smallest interval of real numbers containing that set .
in this approach we limit our attention to a special type of polynomial constraints , namely the ones of the form @xmath259 , where @xmath28 is a polynomial in which each variable occurs _ at most once _ and where @xmath53 is an integer .
we call such a constraint a _ simple polynomial constraint_. by introducing auxiliary variables that are equated with appropriate monomials we can rewrite each polynomial constraint into a sequence of simple polynomial constraints .
this allows us also to compute the integer interval domains of the auxiliary variable from the integer interval domains of the original variables .
we apply then to the simple polynomial constraints the rules introduced in the previous section . to see that the restriction to simple polynomial constraints can make a difference consider the constraint @xmath260 in presence of the ranges @xmath261 $ ] .
we rewrite it into the sequence @xmath262 where @xmath263 are auxiliary variables , each with the domain @xmath264 $ ] .
it is easy to check that the _ polynomial equality _ rule introduced in the previous section does not yield any domain reduction when applied to the original constraint @xmath265 . in presence of the discussed optimization the domain of @xmath40
gets reduced to @xmath198 $ ] . however , if we repeatedly apply the _ polynomial equality _
rule to the simple polynomial constraint @xmath266 , we eventually reduce the domain of @xmath267 to the empty set ( since this constraint has no integer solution in the ranges @xmath268 $ ] ) and consequently can conclude that the original constraint @xmath269 has no solution in the ranges @xmath261 $ ] , without performing any search .
in this approach we focus on a small set of ` atomic ' arithmetic constraints .
we call an arithmetic constraint _ atomic _ if it is in one of the following two forms : * a linear constraint , * @xmath270 .
it is easy to see that using appropriate transformation rules involving auxiliary variables we can transform each arithmetic constraint to a sequence of atomic arithmetic constraints . in this transformation , as in the second approach , the auxiliary variables are equated with monomials so we can easily compute their domains .
the transformation to atomic constraints can strengthen the reduction .
consider for example the constraint @xmath271 and ranges @xmath272 $ ] , @xmath273 $ ] , and @xmath274 $ ] . the first approach without optimization and the second approach can not find a solution without search . if , as a first step in transforming this constraint into a linear constraint , we introduce an auxiliary variable @xmath275 to replace @xmath276 , we are effectively solving the constraint @xmath277 with the additional range @xmath278 $ ] , resulting in only one duplicate occurrence of a variable instead of two . with variable @xmath275 introduced ( or using the optimized version of the first approach ) constraint propagation alone
finds the solution @xmath279 , @xmath280 , @xmath281 , @xmath282 .
we explained already in section [ sec : intermezzo ] how to reason about linear constraints .
( we omitted there the treatment of the disequalities which is routine . )
next , we focus on the reasoning for the multiplication constraint @xmath283 in presence of the non - empty ranges @xmath284 , @xmath285 and @xmath286 . to this end
we introduce the following three domain reduction rules : _ multiplication 1 _
@xmath287 _ multiplication 2 _
@xmath288 _ multiplication 3 _
@xmath289 the way we defined the multiplication and the division of the integer intervals ensures that the _ multiplication _ rules _ 1,2 _ and _ 3 _ are equivalence preserving .
consider for example the _ multiplication 2 _ rule .
take some @xmath290 and @xmath291 such that @xmath292
. then @xmath293 , so @xmath294 and a fortiori @xmath295 .
consequently @xmath296 .
this shows that the _ multiplication 2 _ rule is equivalence preserving .
the following example shows an interaction between all three _ multiplication _ rules .
[ exa : mult ] consider the csp @xmath297 , y \in [ 9 .. 11 ] , z \in [ 155 .. 161 ] \rangle}}.\ ] ] to facilitate the reading we underline the modified domains .
an application of the _ multiplication 2 _ rule yields @xmath298 } , y \in [ 9 .. 11 ] , z \in [ 155 .. 161 ] \rangle}}\ ] ] since , as already noted in in subsection [ subsec : implementation ] , @xmath299/[9 .. 11 ] ) = [ 16 .. 16]$ ] , and @xmath300 \cap int([16 .. 16 ] ) = [ 16 .. 16]$ ] .
applying now the _ multiplication 3 _ rule we obtain @xmath301 , y \in \underline{[10 .. 10 ] } , z \in [ 155 .. 161 ] \rangle}\ ] ] since @xmath299/[16 .. 16 ] = [ 10 .. 10]$ ] and @xmath302 \cap int([10 .. 10 ] ) = [ 10 .. 10]$ ] .
next , by the application of the _ multiplication 1 _ rule we obtain @xmath301 , y \in [ 10 .. 10 ] , z \in \underline{[160 .. 160 ] } \rangle}\ ] ] since @xmath303 \cdot [ 10 .. 10 ] = [ 160 .. 160]$ ] and @xmath299 \cap int([160 .. 160 ] ) = [ 160 .. 160]$ ] . so using all three multiplication rules we could solve the csp ( [ eq : example ] ) .
now let us clarify why we did not define the division of the sets of integers @xmath304 and @xmath305 by @xmath306 the reason is that in that case for any set of integers @xmath304 we would have @xmath307 .
consequently , if we adopted this definition of the division of the integer intervals , the resulting _ multiplication _ _ 2 _ and _ 3 _ rules would not be anymore equivalence preserving . indeed , consider the csp @xmath308 , y \in [ 0 .. 0],z \in [ -8 .. 10 ] \rangle}.\ ] ] then we would have @xmath309/[0 .. 0 ] = { \mbox{$\emptyset$}}$ ] and consequently by the _ multiplication _ _ 2 _ rule we could conclude @xmath310,z \in [ -8 .. 10 ] \rangle}.\ ] ] so we reached an inconsistent csp while the original csp is consistent . in the remainder of the paper we will also consider variants of this third approach that allow squaring and exponentiation as atomic constraints . for this purpose
we explain the reasoning for the constraint @xmath311 in presence of the non - empty ranges @xmath284 and @xmath285 , and for @xmath312 . to this end
we introduce the following two rules in which to maintain the property that the domains are intervals we use the @xmath100 operation of section [ sec : interval ] : _ exponentiation _
_ root extraction _ @xmath314{d_x } ) \rangle}}\ ] ] to prove that these rules are equivalence preserving suppose that for some and @xmath315 we have @xmath316
. then @xmath317 , so @xmath318 and consequently @xmath319 .
also @xmath320{d_x}$ ] , so @xmath321{d_x}$ ] , and consequently @xmath322{d_x})$ ] .
in this section we describe the benchmark experiments that were performed to compare the proposed approaches .
these experiments were performed using a single solver of the dice ( distributed constraint environment ) framework .
dice @xcite is a framework for solver cooperation , implemented using techniques from coordination programming .
it is developed around an experimental constraint solver , called opensolver , which is particularly suited for coordination .
the coordination and cooperation aspects are irrelevant from the point of view of this paper .
relevant aspects of the opensolver are : * it implements a branch - and - infer tree search algorithm for constraint solving .
the inference stage corresponds to constraint propagation and is performed by repeated application of domain reduction functions ( drfs ) that correspond to the domain reduction rules associated with the considered constraints .
* this algorithm is abstract in the sense that the actual functionality is determined by software plug - ins in a number of predefined categories .
these categories correspond to various aspects of the abstract branch - and - infer tree search algorithm .
relevant categories are : variable domain types , domain reduction functions , schedulers that control the application of the drfs , branching strategies that split the search tree after constraint propagation has terminated , and several categories corresponding to different aspects of a search strategy that determine how to traverse a search tree . all experiments were performed using the ` integerinterval ` variable domain type plug - in .
domains of this type consist of an indication of the type of the interval ( bounded , unbounded , left / right - bounded , or empty ) , and a pair of arbitrary precision integer bounds .
this plug - in , and the interval arithmetic operations on it are built using the gnu mp library @xcite .
the branching strategy that we used selects variables using the chronological ordering in which the auxiliary variables come last .
the domain of the selected variable is split into two subdomains using bisection , so the resulting search trees are binary trees . in all experiments we searched for all solutions , traversing the entire search tree by means of depth - first leftmost - first chronological backtracking . for the experiments in this paper
a drf plug - in has been developed that implements the domain reduction rules discussed in the previous sections .
the scheduler plug - in used in the benchmarks keeps cycling through the sequence of drfs , applying drfs that have been scheduled for execution . when a drf is applied , and some variable domain is modified , all drfs that depend on these changes are scheduled for execution , including possibly the one that has just been applied .
the cycling stops when no more drfs are scheduled for execution , or when the domain of a variable becomes empty . as an alternative to cycling
, the scheduler can be supplied with a _ schedule _ : a sequence of indices into the sequence of drfs .
the scheduler will then cycle through this schedule instead , and consider drfs for application in the specified order .
this is used in combination with the second and third approach , where we distinguish _ user _ constraints from the constraints that are introduced to define the values of auxiliary variables . before considering for execution a drf @xmath323 that is part of the implementation of a user constraint , we make sure that all auxiliary variables that @xmath323 relies on are updated . for this purpose , the indices of the drfs that update these variables precede the index of @xmath323 in the schedule .
if @xmath323 can change the value of an auxiliary variable , its index is followed by the indices of the drfs that propagate back these changes to the variables that define the value of this auxiliary variable . for the third approach , there can be hierarchical dependencies between auxiliary variables .
much like the hc4 algorithm of @xcite , the schedule specifies a bottom - up traversal of this hierarchy in a forward evaluation phase and a top - down traversal in a backward propagation phase before and after applying a drf of a user constraint , respectively . in the forward evaluation phase ,
the drfs that are executed correspond to rules _ multiplication 1 _ and _ exponentiation_. the drfs of the backward propagation phase correspond to _ multiplication 2 _ and _ 3 _ , and _ root extraction_. it is easy to construct examples showing that the use of hierarchical schedules can be beneficial compared to cycling through the rules .
the proposed approaches were implemented by first rewriting arithmetic constraints to polynomial constraints , and then to a sequence of drfs that correspond with the rules of the approach used .
we considered the following methods : 1a : : the first approach , discussed in section [ sec : first ] , 1b : : the optimization of the first approach discussed at the end of section [ sec : first ] that involves dividing out common powers of variables , 2a : : the second approach , discussed in section [ sec : second ] .
the conversion to simple polynomial constraints is implemented by introducing an auxiliary variable for every non - linear monomial .
this procedure may introduce more auxiliary variables than necessary .
2b : : an optimized version of approach 2a , where we stop introducing auxiliary variables as soon as the constraints contain no more duplicate occurrences of variables .
3a : : the third approach , discussed in section [ sec : third ] , allowing only linear constraints and multiplication as atomic constraints .
3b : : idem , but also allowing @xmath324 as an atomic constraint .
3c : : idem , allowing @xmath325 for all @xmath87 as an atomic constraint . approaches 2 and 3 involve an extra rewrite step , where the auxiliary variables are introduced .
the resulting csp is then rewritten according to approach 1a . during the first rewrite step the hierarchical relations between the auxiliary variables
are recorded and the schedules are generated as a part of the second rewrite step . for approaches 2b and 3 the question of which auxiliary variables to introduce is an optimization problem in itself .
some choices result in more auxiliary variables than others .
we have not treated this issue as an optimization problem but relied on heuristics .
we are confident that these yield a realistic implementation . in our experiments
we used the following benchmarks .
[ [ cubes ] ] cubes + + + + + the problem is to find all natural numbers @xmath326 that are a sum of four different cubes , for example @xmath327 this problem is modeled as follows : @xmath328,\ x_1,x_2,x_3,x_4 \in { \cal z } \rangle \end{array}\ ] ] [ [ opt ] ] opt + + + we are interested in finding a solution to the constraint @xmath329 in the integer interval @xmath330 $ ] for which the value of @xmath331 is maximal .
[ [ fractions ] ] fractions + + + + + + + + + this problem is taken from @xcite : find distinct nonzero digits such that the following equation holds : @xmath332 there is a variable for each letter .
the initial domains are @xmath333 $ ] . to avoid symmetric solutions an ordering
is imposed : @xmath334 also two redundant constraints are added : @xmath335 because division is not present in the arithmetic expressions , the above constraints are multiplied by the denominators of the fractions to obtain arithmetic constraints .
two representations for this problem were studied : * _ fractions1 _ in which five constraints are used : one equality and four inequalities for the ordering and the redundant constraints , * _ fractions2 _ , used in @xcite , in which three auxiliary variables , @xmath336 and @xmath337 , are introduced to simplify the arithmetic constraints : @xmath338 , @xmath339 , and @xmath340 . additionally , in both representations , 36 disequalities @xmath341 , @xmath342 , ...
, @xmath343 are used . [ [ kyoto ] ] kyoto + + + + + the problem is to find the number @xmath106 such that the alphanumeric equation [ cols= " < , < , < , < , < , < " , ] + for the _ cubes _ and _ opt _ problems , the constraints are already in simple form , so approaches 1a , 1b and 2b are identical .
also all non - linear terms involve either a multiplication or an exponentiation , so also approaches 2a and 3c are the same .
the results of these experiments clearly show the disadvantage of implementing exponentiation by means of multiplication : the search space grows because we increase the number of variable occurrences and lose the information that it is the same number that is being multiplied . for _ opt _ and approach 3a
, the run did not complete within reasonable time and was aborted .
columns e and i of table [ tab - elapsed ] compare the propagation achieved by our approaches with two other systems , respectively ecl@xmath0ps@xmath1 version 5.6ps@xmath1 constraint logic programming system .
see http://www-icparc.doc.ic.ac.uk/eclipse ] using the ic library , and ilog solver 5.1 using type iloint . for this purpose
we ran the test problems without search , and compared the results of constraint propagation .
a mark ` = ' means that the computed domains are the same , ` + ' that our approach achieved stronger propagation than the solver that we compare with , and ` - ' that propagation is weaker . for _ cubes _ , ecl@xmath0ps@xmath1 computes the same domains as those computed according to approach 3b , so here the reduction is stronger than for 3a , but weaker than for the other approaches .
for _ opt _ ecl@xmath0ps@xmath1 and ilog solver compute the same domains .
these domains are narrower than those computed according to approaches 3a and 3b , but the other approaches achieve stronger reduction . in all other cases except for _ kyoto _ and approach 1b the results of all three solvers are the same . for both representations for the fractions puzzle , the symbolic manipulation of approach 1b is able to achieve a significant reduction of the search tree , but this is not reflected in the timings . for _
fractions1 _ the elapsed time even increases .
the reason is that computing the domain updates involves adding intervals of real numbers .
the arithmetic operations on such intervals are more expensive than their counterparts on integer intervals , because the bounds have to be maintained as rational numbers .
arithmetic operations on rational numbers are more expensive because they involve the computation of greatest common divisors .
for _ kyoto _ the symbolic manipulation did not reduce the size of the search tree , so the effect is even more severe .
in general , the introduction of auxiliary variables leads to a reduction of the number of interval operations compared to approach 1a .
the reason is that auxiliary variables prevent the evaluation of subexpressions that did not change .
this effect is strongest for _ fractions1 _ , where the main constraint contains a large number of different power products . without auxiliary variables
all power products are evaluated for every _ polynomial equality _
rule defined by this constraint , even those power products the variable domains of which did not change . with auxiliary variables the intervals for such unmodified terms are available immediately , which leads to a significant reduction of the number of interval multiplications .
the effect that stronger reduction is achieved as a result of introducing auxiliary variables , mentioned in section [ sec : third ] , is seen for both representations of the _ fractions _ benchmark .
the effect described in section [ sec : second ] is not demonstrated by these experiments . if we do nt consider the symbolic manipulation of approach 1b , approach 3c leads to the smallest total number of interval operations in all cases , but the scheduling mechanism discussed in section [ sec - experiments ] is essential for a consistent good performance .
if for example the schedule is omitted for _ opt _ , the number of interval operations almost triples , and performance of approach 2a and 3c is then much worse than that of approach 1a .
the total numbers of interval operations in table [ tab - numbers ] do not fully explain all differences in elapsed times .
one of the reasons is that different interval operations have different costs .
especially the preprocessing of the numerator interval for integer interval division , discussed in subsection [ subsec : implementation ] , is potentially expensive , which may explain why for _ opt _ , approach 1a runs faster than approach 2a , even though the total number of interval operations is higher . among the many other factors that may be of influence , some overhead is involved in applying a drf , so if the number of applications differs significantly for two experiments , this probably influences the elapsed times as well ( _ cubes _ , 1a , 2a , _ opt _ , 1a , 2a , _ fractions2 _ , 2a , 2b ) .
the elapsed times are not the only measure that is subject to implementation details . for example
, we implemented division by a constant interval @xmath344 $ ] as multiplication by a constant , which is more efficient in our implementation .
such decisions are reflected in the numbers reported in table [ tab - numbers ] .
in this paper we discussed a number of approaches to constraint propagation for arithmetic constraints on integer intervals . to assess them we implemented them using the dice ( distributed constraint environment ) framework of @xcite , and compared their performance on a number of benchmark problems .
we can conclude that : * implementation of exponentiation by multiplication gives weak reduction . in our third approach @xmath325
should be an atomic constraint .
* the optimization of the first approach , where common powers of variables are divided out , can significantly reduce the size of the search tree , but the resulting reduction steps rely heavily on the division and addition of rational numbers . these operations can be expected to be more expensive than their integer counterparts , because they involve the computation of greatest common divisors . * introducing auxiliary variables can be beneficial in two ways : it may strengthen the propagation , as discussed in sections [ sec : second ] and [ sec : third ] , and it may prevent the evaluation of subexpressions the variable domains of which did not change . * as a result , given a proper scheduling of the rules , the second and third approach perform better than the first approach without the optimization , in terms of numbers of interval operations .
actual performance depends on many implementation aspects .
however for our test problems the results of variants 2a , 2b and 3c do not differ significantly .
in general , our implementation is slow compared to , for example , ilog solver .
a likely cause is that we use arbitrary precision integers .
we chose this representation to avoid having to deal with overflow , but an additional benefit is that large numbers can be represented exactly . a different approach would be to use floating - point arithmetic and then round intervals inwards to the largest enclosed integer interval .
this was suggested in @xcite and implemented in for example realpaver .
a benefit of this inward rounding approach is that all algorithms that were developed for constraints on the reals are immediately available .
a disadvantage is that for large numbers no precise representation exists , i.e. , the interval defined by two consecutive floating - point numbers contains more than one integer .
but it is debatable whether an exact representation is required for such large numbers .
we realize that the current set of test problems is rather limited .
in addition to puzzles , some more complex non - linear integer optimization problems should be studied .
we plan to further evaluate the proposed approaches on non - linear integer models for the sat problem .
also we would like to study the relationship with the local consistency notions that have been defined for constraints on the reals and give a proper characterization of the local consistencies computed by our reduction rules .
this work was performed during the first author s stay at the school of computing of the national university of singapore .
the work of the second author was supported by nwo , the netherlands organization for scientific research , under project number 612.069.003 .
f. benhamou , f. goualard , l. granvilliers , and j .- f .
puget . revising hull and box consistency . in _ proceedings of the 16th international conference on logic programming ( iclp99 )
_ , pages 230244 . the mit press , 1999 | we propose here a number of approaches to implement constraint propagation for arithmetic constraints on integer intervals . to this end
we introduce integer interval arithmetic .
each approach is explained using appropriate proof rules that reduce the variable domains .
we compare these approaches using a set of benchmarks . |
although the idea of inflation @xcite ( for a review , see @xcite ) started from solving the problems ( like horizon and flatness problem ) of the old hot big bang model , the key to distinguish different inflation models lies on primordial curvature perturbation @xmath2 which provides the seeds for structure formation .
inflation could more or less generate some primordial curvature perturbation since we are living in a quantum world .
the primordial curvature perturbation generated from simple ( single - field slow - roll ) inflation models is adiabatic , almost scale invariant , and gaussian .
however , if inflaton is responsible for generating primordial curvature perturbation , the constraint of cmb normalization ( i.e. @xmath3 ) is so strong that we usually need to fine - tune the parameter(s ) when building an inflation model .
furthermore , future experiments ( like planck satellite ) may detect large non - gaussianity and hence rule out simple scenarios of inflation .
this will reveal for us a nontrivial way of generating primordial curvature perturbation and one of the promising ideas to generate large non - gaussianity is through a curvaton @xcite .
non - gaussianity generated from curvaton scenario can be described by the nonlinear parameter @xmath1 , which takes the form @xmath4 where @xmath5 denotes the gaussian part of @xmath2 .
currently the upper bound of @xmath1 is roughly given by ( @xmath6 ) @xcite @xmath7 in the near future , the planck satellite will reduce the bound to @xmath8 if non - gaussainity is not detected .
curvaton is supposed to be light and subdominant during inflation .
because it is light , it can produce sizable quantum fluctuations which when stretched outside the horizon during inflation would become classical perturbations . because it is subdominat ,
the perturbations should be regarded as isocurvature perturbation after inflation when the curvaton field starts to oscillate .
the curvaton is supposed to decay after inflaton decay and at the same time transform its isocurvature perturbation to curvature perturbation .
if the curvature perturbation of the universe is from the curvaton , it could liberate the constraint on inflation and lower the scale of inflation @xcite
. it would be interesting if we can identify a field from particle physics to be the curvaton and one of the candidate is a right - handed ( rh ) sneutrino in the framework of supersymmetry ( susy ) .
the idea of a right - handed sneutrino curvaton has already be mentioned in the original paper of curvaton @xcite and been considered in @xcite . in this case , however , it is possible that we have more than one generation of the right - handed sneutrino and more than one curvaton .
the calculation for two curvaton decay has been considered in @xcite , however , the parameters used in the paper is not of direct use for particle physicists to find possible candidate for the curvatons . in this paper
we consider the decay width of the right - handed sneutrino curvaton in terms of the yukawa couplings and masses .
it turns out that there are six parameters in addition to the hubble parameter .
we then numerically scanning the parameter space in order to find solutions .
this paper is organized as followsing . in section [ sec1 ] , we review the result of @xcite in order to compare single- and two- curvaton decays in the succeeding sections . in section [ sec2 ]
, we present the formalism we use for calculating two curvaton decay . in section [ sec3 ] , we specify the parameter space and present our numerical results
. section [ sec4 ] is our conclusion . for completeness , there is also an appendix section in the end of the paper where we summarize the equations adopted from @xcite .
it is well known that in standard model of particle physics , there is no good candidate to play the role of a curvaton .
however , if we go beyond the standard model by imposing susy , there are lots of scalar fields . in this section
we review the idea of a right - handed sneutrino which plays the role of the curvaton .
we focus on the case that when the curvaton decays , it subdominates the energy density of the universe @xcite , because large non - gaussianity could be produced in this region of the parameters .
the superpotential of the rh neutrino is given by @xmath9 where @xmath10 is the rh neutrino superfield , @xmath11 and @xmath12 are the mssm higgs and lepton doublet superfields , and @xmath13 is the rh neutrino mass .
the canonical type - i seesaw mechanism gives the mass relation between light and heavy majorana neutrino masses as @xmath14 with @xmath15 denoting the vacuum expectation value of @xmath16 .
the mass squared differences revealed from the neutrino oscillation data , @xmath17 and @xmath18 @xcite , indicates that if we consider three generations of rh sneutrinos , the lightest left - handed neutrino mass can be very small while the mass of the other two generations are fixed due to the oscillation data .
this means in principle we can have a yukawa coupling arbitrarily small ( @xmath19 ) but the other two yukawa couplings should result in the neutrino masses around @xmath20 ( with a mass difference @xmath21 , namely , `` inverted ordering '' ) . on the other hand
, we can also have `` normal ording '' which implies @xmath22 , @xmath23 , and @xmath24 . from eq .
( [ yukawac ] ) we can see that for the case @xmath25 , if we have @xmath26 ( @xmath27 ) , we need @xmath28 ( @xmath29 ) .
the potential of the rh sneutrino @xmath30 can be expressed as @xmath31 for simplicity , here we do not consider the hubble - induced mass term .
it is possible that the mechanism which suppresses the hubble - induced mass term for the inflaton also suppresses that for the curvaton .
for example , this is the case for d - term hybrid inflation @xcite .
the decay rate for the rh sneutrino is @xmath32 the spectrum is given by @xmath33 if we assume that at the time @xmath34 of curvaton oscillation with energy density @xmath35 , the universe is dominated by radiation ( the decay products of inflaton ) with energy density @xmath36 . at the time of curvaton decay @xmath37 , the energy density of the universe
is given by @xmath38 .
therefore @xmath39 and @xmath40 is given by @xmath41 throughout this paper , we always use a subscript @xmath42 " to denote horizon exit .
when @xmath43 is small , the nonlinear parameter is given by @xmath44 note that here @xmath1 can never be negative because we assume the curvaton is subdominant when it decays .
as we will see in the following section , this is not true in the two - curvaton case . by using eqs .
( [ eq1 ] ) , ( [ eq2 ] ) , and ( [ eq3 ] ) , we can obtain @xmath45 and @xmath46 it is interesting to note that here the spectrum does not give constriant on the mass of the right - handed sneutrino . however , as we can see in the next section , the results
does depend on the masses when we consider the two curvatons case .
in our setup , we consider three generations of right - handed sneutrinos for type - i seesaw model and for simplicity we assume the heavist right - handed sneutrino ( with a mass we denoted as @xmath47 ) does not play any role in cosmology , namely @xmath48 . ) .
the effect is most significant for the heavist right - handed snetrino due to the largest yukawa coupling .
even if @xmath49 , it is possible that the heavist right - handed sneutrino obtains an effective mass larger than the hubble scale during inflation while other right - handed sneutrinos have masses lighter than the hubble scale . ] here we consider the case that the lighter two right - handed sneutrinos are curvatons ( we call them curvaton @xmath50 and curvaton @xmath51 ) and investigate the effects on the primordial curvature perturbation
. the decay rates of the two curvatons are @xmath52 where @xmath53 and @xmath54 are the yukawa couplings and @xmath55 and @xmath56 are their masses . quite generally , we consider @xmath57 we assume that when both of the curvaton decays , the energy density of the universe is dominated by radiation so that we can compare this with single curvaton case . the similar calculation of the energy density ratio @xmath58 as in eq .
( [ eq2 ] ) can be obtained for radiation @xmath59 , curvaton @xmath50 , and curvaton @xmath51 . at curvaton @xmath50 decay ( which is denoted by subscript 1 " throughout this paper ) , @xmath60 where the subscript @xmath61 denotes pre - existing radiation just before curvaton @xmath50 decay . at curvaton @xmath51 decay ( which is denoted by subscript 2 " throughout this paper ) , @xmath62 where the subscript @xmath63 denotes radiation just before curvaton @xmath51 decay . to linear order , the spectrum is given by @xcite @xmath64 where the parameters @xmath65 and @xmath66 can be found in the appendix and @xmath67 here we use subscript ( 1 ) " to denote first order " and ( 2 ) "
will be used to denote second order " .
if we define @xmath68 , we obtain @xmath69^{1/2}\frac{1}{3\pi}\frac{h_\ast}{a_\ast}.\ ] ] we can write the spectrum explicitly by inserting @xmath65 and @xmath66 to obtain @xmath70 ^ 2+\left[12m_p^2\sqrt{\pi}(\frac{b_{*}}{\lambda_b})+\pi{b_{*}}\left[\frac{a_{*}^2}{\lambda_a}(\frac{1}{\lambda_a}\sqrt{\frac{m_b}{m_a}}+\frac{3}{\lambda_b})+3\sqrt{\frac{m_b}{m_a}}\frac{b_{*}^2}{\lambda_a\lambda_b}\right ] \right]^2}{\left[4m_p^2+\sqrt{\pi}(\frac{a_{*}^2}{\lambda_a}+\sqrt{\frac{m_b}{m_a}}\frac{b_{*}^2}{\lambda_a } ) \right]^2\left[12m_p^2+\sqrt{\pi}(4\frac{a_{*}^2}{\lambda_a}+3\frac{b_{*}^2}{\lambda_b})\right]^2}. \nonumber \\\end{aligned}\ ] ] this is subjected to cmb normalization @xmath71 at horizon exit .
to second order , the curvature perturbation is given by @xcite @xmath72+\frac{1}{2}\left[c\zeta^2_{a(1)}+d\zeta^2_{b(1)}+e\zeta_{a(1)}\zeta_{b(1)}\right ] \label{revision3}\ ] ] where @xmath73 is the total curvature perturbation after the second curvaton decay .
as in eq .
( [ revision2 ] ) , we use subscript ( 1 ) " to denote first order " and ( 2 ) " to denote second order " .
the first order part @xmath74 is gaussian and the second order part @xmath75 is non - gaussian .
the parameters @xmath76 , @xmath77 , @xmath78 can be found in the appendix and the nonlinear parameter is @xmath79
there are six parameters : @xmath80 in addition to the hubble parameter @xmath81 and one constraint ( cmb normalization ) .
we tackle the problem numerically by scanning the parameter space .
first of all , we assume the inflaton contribution to the curvature perturbation is small , this implies @xmath82 where @xmath83 . for a typical value of @xmath84 at horizon exit , we have the bound for hubble parameter @xmath85 we will numerically find solutions by making plots for the two cases @xmath86 and @xmath87 in this paper .
the range of the curvaton field values at hubble exit are chosen to be @xmath88 the lower bound is from the requirement that the classical field value is larger than its fluctuation and the upper bound is from the requirement that the curvaton would not drive a second stage of inflation .
the masses of the curvatons are @xmath89 the lower bound is from the fact that the mass of the rh sneutrino should be larger than its soft mass which is assumed to be tev scale and we also use the constraint from eq .
( [ eq9 ] ) .
the decay rate is also constrained by eq .
( [ eq9 ] ) , @xmath90 therefore we choose @xmath91 and again we require @xmath92 therefore we choose @xmath93 we plot our results in figs .
[ fig1]-[fig3 ] for @xmath86 and figs .
[ fig4]-[fig6 ] for @xmath87 . for comparison
, we also plot the results from the single - curvaton case .
naively one may think it is possible that one of the curvaton plays no role at all and the results would be dominated by a single curvaton .
however , those plots show that both of the curvatons could play some role in generating curvature perturbation . in all of the plots ,
each point represents a solution of all six parameters subjected to cmb normalization .
we pick up a few points and list them on the table [ output ] . in the plots
@xmath1 is an output after we imposing cmb normalization .
we notice that even we assume the energy density of the curvatons is subdominant when they decays , we can still get small or negative ( as well as large ) @xmath1 .
in addition , we also find that one of the yukawa coupling can be as large as @xmath94 while the other coupling is relatively small .
this is interesting because it is consistent with the neutrino oscillation data which we discussed below eq .
( [ yukawac ] ) by the assumption that the curvatons are rh sneutrinos .
for example , in the second row of table [ output ] , we would have the light neutrino masses correspond to the heavy neutrino masses @xmath55 and @xmath56 as @xmath95 and @xmath96 respectively from eq .
( [ yukawac ] ) .
we can easily choose a @xmath97 with a larger yukawa coupling to make @xmath98 and make an inveted ordering of neutrino masses compatible with neutrino oscillation .
as another example , for the seventh row of table [ output ] , @xmath99 can be obtained and along similar arguing we can obtain a normal ordering of neutrino masses .
we would like to emphasis here that our goal is not showing that all of our parameter spaces are compatible with neutrino oscillation data , because our results can generically apply to other two - curvaton models with a similar decay rates .
however , it is interesting enough if some of them do compatible with our assumption that two rh sneutrinos can play the role of curvatons .
.[output ] list of a few points of the numerical solutions .
[ cols="^,^,^,^,^,^,^,>",options="header " , ]
we have explored the parameter space of generating primordial curvature perturbation via two right - handed sneutrino curvaton decays in the framework that there are three generations of rh sneutrinos .
we compared the results with the single - curvaton case and found that an additional rh sneutrino curvaton could have some effects and can not be neglected .
notably we may still get small or even negative @xmath1 in the case that the energy density of the curvatons sub - dominates when they decay . as can be seen from fig .
[ fig2 ] and fig .
[ fig5 ] that the field value for the solutions may be not very small compared to the planck mass ( although it has to be smaller than the planck mass in order to avoid driving a second stage of inflation )
. therefore we have to suppress the nonrenormalizable terms in the superpotential .
this can be achieved for example by judiciously assigning r - charge to the rh sneutrinos @xcite .
the purpose of this appendix is to show how to obtain parameters @xmath65 , @xmath66 , @xmath76 , @xmath77 , and @xmath78 from eqs .
( [ eq14 ] ) and ( [ eq15 ] ) .
the following equations and notations are from @xcite : + the energy density parameters ( @xmath100 ) : + @xmath101 the density of the first curvaton @xmath50 just before the first decay .
+ @xmath102 the density of the second curvaton @xmath51 at the first decay .
+ @xmath103 the density of the second curvaton @xmath51 just before the second decay .
+ @xmath104 the density of pre - existing radiation at the first ( curvaton ) decay .
+ @xmath105 the density of all radiation immediately after the first decay .
+ @xmath106 the density of all radiation just before the second decay .
+ @xmath107 the density of all radiation immediately after the second decay .
+ the full non - linear curvature perturbations : + @xmath108 the primordial perturbation after the second curvaton decay , but before the nucleosynthesis , + @xmath109 the total perturbation at the first decay , + @xmath73 the total perturbation at / after the second decay , + @xmath110 the perturbation of the first curvaton @xmath50 , + @xmath111 the perturbation of the second curvaton @xmath51 , + @xmath112 the pre - existing radiation perturbation , + @xmath113 the radiation perturbation after the second decay .
+ subscript in parenthesis : + @xmath114 the first order part of the primordial perturbation , + @xmath115 the second order part of the primordial perturbation , + @xmath116 the first order part of the total perturbation at the first decay , + @xmath117 the second order part of the total perturbation at the first decay , + @xmath74 the first order part of the total perturbation at the second decay , + @xmath75 the second order part of the total perturbation at the second decay , + @xmath118 the first order part of the first - curvaton perturbation , + @xmath119 the second order part of the first - curvaton perturbation , + @xmath120 the first order part of the second - curvaton perturbation , + @xmath121 the second order part of the second - curvaton perturbation , + @xmath122 the first order part of the pre - existing radiation perturbation , + @xmath123 the second order part of the pre - existing radiation perturbation , + @xmath124 the first order part of the radiation perturbation after the first decay , + @xmath125 the second order part of the radiation perturbation after the first decay , + @xmath126 the first order part of the radiation perturbation after the second decay , + @xmath127 the second order part of the radiation perturbation after the second decay , + we have @xmath128 where k. enqvist and m. s. sloth , nucl .
b * 626 * , 395 ( 2002 ) [ arxiv : hep - ph/0109214 ] .
t. moroi and t. takahashi , phys .
b * 522 * , 215 ( 2001 ) [ erratum - ibid .
b * 539 * , 303 ( 2002 ) ] [ arxiv : hep - ph/0110096 ] .
e. komatsu _ et al .
_ [ wmap collaboration ] , astrophys . j. suppl . * 192 * , 18 ( 2011 ) .
[ arxiv:1001.4538 [ astro-ph.co ] ] .
j. mcdonald , phys .
d * 70 * , 063520 ( 2004 ) [ arxiv : hep - ph/0404154 ] .
t. moroi and h. murayama , phys . lett .
b * 553 * , 126 ( 2003 ) [ arxiv : hep - ph/0211019 ] .
m. postma , phys . rev .
d * 67 * , 063518 ( 2003 ) [ arxiv : hep - ph/0212005 ] .
a. mazumdar and a. perez - lorenzana , phys .
d * 70 * , 083526 ( 2004 ) [ arxiv : hep - ph/0406154 ] . c. -m .
lin , k. cheung , jcap * 0906 * , 006 ( 2009 ) .
[ arxiv:0904.2826 [ hep - ph ] ] . | in this paper , we consider the effect of two right - handed sneutrino curvaton decays and investigate the parameter space .
we compare the difference of the result between single- and two - curvaton cases .
we find one yukawa coupling of the right - handed snetrinos can be as large as @xmath0 while the other one is much smaller .
when the curvatons decay , we assume both of them subdominate the energy density of the universe . we find that , unlike a single curvaton case , here a small or negative , as well as a large @xmath1 can be generated . |
core collapse supernovae are among the most important phenomena in astrophysics because of their energetics and nucleosynthesis . beginning with
the first numerical simulations conducted by colgate and white@xcite , three decades of supernova modeling have established a basic supernova paradigm . the supernova shock wave formed when the iron core of a massive star collapses gravitationally and rebounds as the core matter exceeds nuclear densities
stalls in the iron core as a result of enervating losses to nuclear dissociation and neutrinos .
the failure of this `` prompt '' supernova mechanism sets the stage for a `` delayed '' mechanism , whereby the shock is reenergized by the intense neutrino flux emerging from the neutrinospheres carrying off the binding energy of the proto - neutron star@xcite .
the heating is mediated primarily by the absorption of electron neutrinos and antineutrinos on the dissociation - liberated nucleons behind the shock .
this past decade has also seen the emergence of multidimensional supernova models , which have investigated the role convection , rotation , and magnetic fields may play in the explosion @xcite .
although a plausible framework is now in place , fundamental questions about the explosion mechanism remain : is the neutrino heating sufficient , or are multidimensional effects such as convection and rotation necessary ?
can the basic supernova observable , explosion , be reproduced by detailed spherically symmetric models , or are multidimensional models required ? without a doubt , core collapse supernovae are not spherically symmetric .
for example , neutron star kicks@xcite and the polarization of supernova emitted light@xcite can not arise in spherical symmetry .
nonetheless , ascertaining the explosion mechanism and understanding every explosion observable are two different goals . to achieve both , simulations in one , two , and three
dimensions must be coordinated .
the neutrino energy deposition behind the shock depends sensitively not only on the neutrino luminosities but also on the neutrino spectra and angular distributions in the postshock region , necessitating exact multigroup ( multi - neutrino energy ) boltzmann neutrino transport .
ten percent variations in any of these quantities can make the difference between explosion and failure in supernova models@xcite .
past simulations have implemented increasingly sophisticated approximations to boltzmann transport , the most sophisticated of which is multigroup flux - limited diffusion@xcite .
a generic feature of this approximation is that it underestimates the isotropy of the neutrino angular distributions in the heating region and , thus , the heating rate@xcite .
it is important to note that , without invoking proto - neutron star ( e.g. , neutron finger ) convection , simulations that implement multigroup flux - limited diffusion do not produce explosions@xcite .
moreover , the existence and vigor of proto - neutron star convection is currently a matter of debate@xcite .
wilson@xcite implemented an approximation to boltzmann neutrino transport by using order - of - magnitude parameterizations of the neutrino
matter weak interactions .
his models failed to produce explosions .
core collapse simulations that implemented exact boltzmann neutrino transport were completed by mezzacappa and bruenn@xcite . following this work
, we now present the findings of a core collapse , bounce , and postbounce simulation . recognizing the need for more accurate time - dependent neutrino transport in supernova models , other groups have now developed boltzmann solvers@xcite .
we model the explosion of a 13 m@xmath0 star , beginning with the precollapse model of nomoto and hashimoto@xcite .
the core collapse , bounce , and explosion were simulated with a new neutrino radiation hydrodynamics code for both newtonian and general relativistic spherically symmetric flows : agile boltztran .
boltztran is a three - flavor boltzmann neutrino transport solver@xcite , now extended to fully general relativistic flows@xcite . in this simulation , it is employed in the @xmath2 limit with 6-point gaussian quadrature to discretize the neutrino angular distributions and 12 energy groups spanning the range from 5 to 300 mev to discretize the neutrino spectra .
agile is a conservative general relativistic hydrodynamics code@xcite .
its adaptivity enables us to resolve and seamlessly follow the shock through the iron core into the outer stellar layers .
the equation of state of lattimer and swesty@xcite ( ls eos ) is employed to calculate the local thermodynamic state of the matter in nuclear statistical equilibrium ( nse ) . for matter initially in the silicon layer , the temperatures are insufficient to achieve nse . in this region ,
the radiation and electron components of the ls eos are used , while an ideal gas of @xmath3 is assumed for the nuclear component . for typical hydrodynamic timesteps ( @xmath4 millisecond ) ,
silicon burning occurs within a single timestep for t @xmath5@xcite ; therefore , when a fluid element exceeds a temperature of 5 @xmath6 in our simulation , the silicon is instantaneously burned , achieving nse and releasing thermal energy equal to the difference in nuclear binding energy between @xmath3 and the composition determined by the ls eos .
we investigated the convergence of the net neutrino heating rate as the number of gaussian quadrature points and the number of neutrino energy groups in our boltzmann simulations were varied , as in messer et al@xcite . in the heating region ,
the 4- and 6-point rates , the 6- and 8-point rates , and the 12- and 20-group rates differed by at most 5 percent , 3 percent , and 3 percent , respectively . moreover , during the course of the important first 300 ms of our simulation , the maximum variation in the total energy is @xmath7 erg , which is a few percent of the total energy , and the total lepton number is conserved to within a fraction of a percent .
note that the numerical uncertainty in the net heating rate ( which is at most 3 percent in our model ) is no larger than the uncertainty in the total energy .
therefore , any further numerical convergence in the computation of this rate would be meaningless .
figure 1 shows the radius - versus - time trajectories of equal mass shells ( 0.01m@xmath0 ) in the stellar iron core and silicon layer during the first 600 ms of postbounce evolution .
it also shows the shock and nuclear burning front trajectories . at 110
ms after bounce , the shock stalls at a radius of 230 km and then recedes for the duration of the simulation , and no explosion is launched . the shock and burning fronts divide the stellar core and silicon layer into three regions : a : silicon .
b : iron produced by infall compression and heating .
c : free nucleons and alpha particles . in figure 2 , we plot the neutrino luminosities and rms energies at 500 km in the stellar core as a function of time .
the electron neutrino luminosity decreases from its early `` burst '' value of @xmath8 erg / s as we enter the postbounce accretion phase .
( the early electron neutrino burst occurs as the shock passes the electron neutrinosphere in the core , liberating the trapped neutrinos behind it that are produced by electron capture during stellar core collapse . ) in the accretion phase , the electron neutrino luminosity reaches a maximum of @xmath9 erg / s and then decreases slowly as the mass accretion rate decreases .
the electron antineutrino and muon / tau neutrino and antineutrino luminosities rise after a hot , deleptonized `` mantle '' forms beneath the shock ( this is the region above the cold , degenerate , unshocked core , and includes the neutrinospheres ) . in this mantle ,
electron antineutrinos are produced by positron capture and all three flavors of neutrinos and antineutrinos are produced by electron positron annihilation .
the electron antineutrino luminosity reaches a maximum of @xmath10 erg / s and exhibits the same subsequent decrease with decreasing mass accretion rate .
the muon / tau neutrino and antineutrino luminosities on the other hand reach values of only @xmath11 erg / s , and thereafter decrease with time .
for all three flavors , the rms energies increase with time , owing to infall into an increasingly deep gravitational well .
relative to the electron neutrino and antineutrino rms energies , the muon / tau neutrino and antineutrino rms energies are larger , with values between 2025 mev : the muon / tau neutrinos and antineutrinos interact only via neutral currents and therefore decouple deeper in the core at higher densities .
the electron neutrino rms energies lie in the range between 1020 mev .
relative to the electron neutrino rms energies , the electron antineutrino rms energies are slightly larger because the electron antineutrinos decouple at slightly higher densities : the core material is neutron rich ; therefore , electron antineutrino absorption on protons is reduced relative to electron neutrino absorption on neutrons . in figure 3 , we plot the mass density , entropy per baryon ( in units of @xmath12 ) , electron fraction , and velocity as a function of radius for various time slices in our simulation . in the velocity profiles , the initial outward propagation of the shock is evident , as is the subsequent decrease in radius at later times .
note the increasing infall velocities below the shock as it recedes , reaching values of -6000
km / s at 600 ms after bounce . despite the decreasing shock radius and failed explosion
, there remains a heating region behind the shock , where the entropies continue to increase .
this is evident in the entropy profiles .
nonetheless , conditions remain such that an explosion does not occur in this model in the first second .
we have presented results from the simulation of the stellar core collapse , bounce , and the first 600 ms of postbounce evolution of a 13 m@xmath0 progenitor .
spherical symmetry was assumed , @xmath2 boltzmann neutrino transport was implemented , and gravity was newtonian .
no explosion was obtained . in light of our implementation of boltzmann transport , if we do not obtain explosions in models initiated from other progenitors ( see also rampp and janka@xcite ) , it would indicate that improvements in our initial conditions ( precollapse models ) and/or input physics are needed , and/or that the inclusion of multidimensional effects such as convection , rotation , and magnetic fields are required ingredients in the recipe for explosion . in the past
, it was not clear whether failures to produce explosions in spherically symmetric models were the result of transport approximations or the neglect of an important physical effect .
we will report on the general relativistic case@xcite and other models in subsequent papers .
potential improvements in our initial conditions and input physics include : improvements in precollapse models@xcite ; the use of ensembles of nuclei in the stellar core rather than a single representative nucleus ; determining the electron capture rates on this ensemble with detailed shell model computations@xcite ; the inclusion of nucleon correlations in the high - density neutrino opacities@xcite ; and the inclusion of new neutrino emissivities in dense matter@xcite .
these improvements all have the potential to quantitatively , if not qualitatively , change the details of our simulations .
thus , it is important to note that our conclusions are drawn considering the initial conditions and input physics used .
a.m. is supported at the oak ridge national laboratory , managed by ut - battelle , llc , for the u.s .
department of energy under contract de - ac05 - 00or22725 .
is supported by the national science foundation under contract ast-9877130 and , formerly , was supported by the swiss national science foundation under contract 2000 - 53798.98 .
o.e.b.m . is supported by funds from the joint institute for heavy ion research and a department of energy pecase award .
is supported by nasa under contract nag5 - 8405 and by funds from the joint institute for heavy ion research .
is supported in part by the swiss national science foundation under contract 2000 - 53798.98 and as a visiting distinguished scientist at the oak ridge national laboratory .
s.w.b . is supported by the nsf under contract 96 - 18423 and by nasa under contract nag5 - 3903 .
our simulations were carried out on the ornl physics division cray j90 and the national energy research supercomputer center cray j90 and sv-1 .
we thank the referees for their important questions and comments . | with exact three - flavor boltzmann neutrino transport , we simulate the stellar core collapse , bounce , and postbounce evolution of a 13 m@xmath0 star in spherical symmetry , the newtonian limit , without invoking convection . in the absence of convection ,
prior spherically symmetric models , which implemented approximations to boltzmann transport , failed to produce explosions .
we are motivated to consider exact transport to determine if these failures were due to the transport approximations made and to answer remaining fundamental questions in supernova theory .
the model presented here is the first in a sequence of models beginning with different progenitors . in this model
, a supernova explosion is not obtained .
we discuss the ramifications of our results for the supernova mechanism .
# 1#1 @xmath1 |
supernovae are the final stage of stellar evolution for white dwarfs and massive stars .
supernova remnants ( snrs ) , produced by the interaction of the ejected stellar material with the ambient medium @xcite , are a key link between the life of stars and the evolution of the interstellar medium ( ism ) .
moreover , they are believed to be the main production sites of galactic cosmic - rays , up to the so - called knee " in their spectrum ( at a few pev ) .
indeed the supersonic ejecta generate a strong , magnetized shock wave , that can efficiently accelerate charged particles through the process of _ diffusive shock acceleration _
a variety of multi - wavelength observations supports this scenario , especially at high energies ; see @xcite and @xcite for recent reviews , and @xcite for an up - to - date catalogue of x - ray and @xmath0-ray observations of all galactic snrs . however , if the presence of relativistic particles in snrs is well established , their nature and their actual energy distribution is still unclear . in the last decade ,
several remnants were finally detected in @xmath0-rays , but it proved difficult to disentangle the leptonic and hadronic contributions from inverse compton scattering and neutral pion decay , respectively ( e.g. @xcite ) . in the gev @xmath0-ray band ,
the characteristic pion - decay spectral signature could be detected with certainty only in a couple of middle - aged remnants @xcite , and does not require the underlying proton spectrum to extend above the tev energies . in the tev @xmath0-ray band , the hadronic scenario is favoured in several remnants because of the presence of molecular clouds that make good targets for protons ( e.g. @xcite ) , but again , due to their advanced age , these objects are not believed to still be pevatron accelerators , as required to power the galactic cosmic rays up to knee " energies .
younger , potentially more active objects ( with faster , stronger shocks ) , were finally detected as well , amongst the historical remnants .
one of the most interesting targets is tycho s snr ( g120.1 + 1.4 ) , a remnant of a type ia supernova , that has a nearly spherical shape and evolves in a relatively uniform ism . in this object
, most models point to a hadronic origin of the @xmath0-ray emission , and imply modest to very high efficiencies of acceleration at the shock @xcite .
when modeling the emission from a snr , it is important to have a joint description of the remnant and of particle acceleration ( energetic protons and electrons ) .
if particle acceleration is as efficient as is believed , especially regarding ions , then this must impact the dynamics of the flow and of the magnetic field and in turn the way acceleration works .
the back - reaction of the energetic particles on the shock is two - fold : ( i ) the pressure of particles can set the fluid in motion @xcite , and ( ii ) their current can amplify the magnetic field @xcite .
both effects happen immediately ahead of the shock , in the so - called precursor , where energetic particles are free to wander in the yet un - shocked medium .
to compute the emission from the snr , it is mandatory to solve jointly the ( magneto-)hydrodynamic equations describing the evolution of the fluid and the kinetic equations describing the acceleration of the particles .
various numerical approaches are possible to tackle this difficult problem .
direct computations of the coupled system are expensive , and require special treatments to be useful for the large scale modeling of snrs ; in particular , @xcite performed simulations in a comoving grid expanding with the ( forward ) shock , while @xcite introduced a change of variables that allows for the simultaneous handling of the forward and reverse shocks . an alternative to explicitly solving the differential equation describing
the transport of particles is to resort to monte - carlo approaches , where the particles are treated in a stochastic way while the fluid is treated with standard numerical methods ; for instance @xcite used stochastic differential equations coupled with an hydrodynamic code .
it is possible to simplify the problem by using a two - fluid approximation , that is by relying on a hydrodynamic description for both the thermal and relativistic components of the system , with the drawback of losing the energy dependence of the diffusion coefficient ; this allowed @xcite to make parametric studies .
we advocate here a hybrid approach , where a hydrodynamic description of the snr evolution is coupled to a kinetic model of acceleration .
@xcite coupled a simple acceleration model , in which the spectrum of particles is linear by parts , to a code computing self - similar hydrodynamic solutions . @xcite
coupled the same acceleration model to a code solving the euler equations .
@xcite switched to a more realistic acceleration model , and recently @xcite improved its treatment of the magnetic field . in this approach , the fluid simulated by the hydrodynamic code ( that can be called a pseudo - fluid " ) accounts for both the thermal and non - thermal populations , yet the kinetic model ensures that the mix of the two follows the expectations from non - linear dsa at any given time and position .
this allows for a realistic description of some important aspects of the acceleration mechanism , while having a relatively small overhead compared to a purely hydrodynamic simulation .
our overall aim in our ongoing modeling effort is to constrain the acceleration properties in a snr from its broadband thermal and non - thermal emission . in our preliminary work (
* hereafter paper i ) , we presented the morphological evolution of the remnant , taking into account both the development of the instabilities and the shrinking of the shocked region caused by the presence of energetic particles , in a time - dependent manner . in our subsequent paper (
* hereafter paper ii ) , we computed the thermal x - ray emission from the shocked material , showing how it is affected by the presence of energetic particles . in the present work ,
we compute the _ non - thermal _ broadband emission from the accelerated particles ( from radio to @xmath0-rays ) , again discussing non - linear effects in the interpretation of observations and now including different hypotheses for the evolution of the magnetic field .
an important aspect in our modeling is to take into account the actual morphology of the remnant , by running three - dimensional simulations , whereas the modeling efforts previously cited that couple the time evolution of particle acceleration with the hydrodynamics are one - dimensional , based on the simplifying assumption of spherical symmetry . even under the assumptions of uniform ejecta and ism , the assumption of symmetry does not hold for young remnants : the interface between the shocked ism and the shocked ejecta ( called the contact discontinuity ) is subject to rayleigh - taylor instabilities , which produce a distinctive pattern of fingers and holes @xcite .
it is mandatory to take this effect into account to properly quantify the size of the shrinked shocked ism and to reproduce spatially - resolved images of the thermal emission , as observed with x - ray satellites ( see e.g. @xcite and @xcite for tycho s snr ) . in this work
we consider that particles are accelerated at the forward shock , and so we concede that the non - thermal emission , that mostly follows the distribution of the shocked ism , has a simpler geometry than the thermal emission , that mostly follows the distribution of the ejecta .
the impact of the instabilities occurring inside the shell may be more subtle , although it should be investigated as well . in any case
, we want to be able to compute all kinds of emission within a single model , which offers the best possible description of the snr evolution .
@xcite showed the importance of the geometry of the downstream region in the computation of the emission , and @xcite further showed the importance of properly taking into account the dynamical evolution of the snr shock to obtain the particle population at a given age . in this work ,
we compute the continuous distribution of particles in both space and time , as consistently as we can in the framework of our pseudo - fluid approach .
@xcite demonstrated the importance of multi - zone versus single - zone modeling when interpreting the @xmath0-ray emission from tycho s snr and trying to assess the presence or absence of energetic protons in its shell . whereas they used an ad - hoc parametrization of the zones
, our model naturally produces the fine - grained radial structure of the shocked region .
moreover , we would like to emphasize that , to be able to reproduce the observations , one needs to compute the emissivity as seen in projection in the plane of the sky .
then each pixel of the map is obtained by integrating all the variations that occur along the line of sight , that are due both to geometry effects ( non - uniform plasma conditions in the interior ) and age effects ( non - uniform conditions for energetic particles and the magnetic field at the shock ) .
the outline of the paper is as follows .
our method is presented in section [ sec : model ] .
we briefly recall how we couple a hydrodynamic code with a kinetic model of non - linear acceleration , and we detail the treatment of the magnetic field and of the accelerated particles in the shocked ism .
our results are presented in section [ sec : results ] .
we first show slices of some useful hydrodynamic quantities , to demonstrate the effect of the particles back - reaction on the shock structure and on the downstream magnetic field , and how this impacts the energy losses of particles .
we then show projected maps of the various non - thermal emission mechanisms in different energy bands , relevant to currently available instruments .
finally we present broad - band spectra , integrated over the whole remnant , to compare the signatures of the different emissions of the two kinds of particles ( electrons vs. protons ) .
our findings are discussed in section [ sec : discussion ] .
we summarize the main observational diagnostics of efficient acceleration of particles ( electrons and protons ) inside young snrs , and we review the main physical assumptions underlying our model .
our conclusions are presented in section [ sec : conclusion ] .
in this section we present our numerical model .
we recall how we couple a hydrodynamic code and a kinetic model to run simulations of a snr undergoing efficient particle acceleration ( section [ sec : model - snr ] ) .
we discuss the fate of the magnetic field upstream of the shock , where it is affected by particle - driven instabilities ( section [ sec : model - mfa ] ) , as well as downstream of the shock , where it is advected with the flow , together with the energetic particles ( section [ sec : model - down ] ) .
finally we present the code used to compute the non - thermal emission of the particles at any point inside the shell ( section [ sec : model - em ] ) .
particles are assumed to be accelerated at the blast wave of snrs through the process of diffusive shock acceleration , in which charged particles gain energy by repeatedly crossing the velocity jump .
here we consider acceleration at the forward shock only , as in our previous work ( see paper ii for a discussion of the possible role of the reverse shock ) . throughout this paper
, we are working with the isotropic distribution function of particles in phase space @xmath2 , a function of time @xmath3 , position @xmath4 , and momentum @xmath5 . in the test - particle regime ,
the final spectrum is a power - law @xmath6 of index @xmath7 where @xmath8 is the compression ratio of the shock ( for a strong shock @xmath9 and thus @xmath10 ) .
this index in turns determines the spectrum of the emitted photons .
but an important feature of efficient acceleration is its non - linearity , which leads to a variety of spectra , hence the need for a consistent treatment of the thermal and non - thermal populations .
here we briefly outline our approach , which was introduced in paper i and detailed in paper ii .
the evolution of the remnant is described using the code ramses @xcite .
it is used to solve the euler equations on a 3d cartesian grid , which is comoving with the blast wave @xcite . the tree - based grid is adaptively refined around the forward shock that propagates in the ism , the reverse shock that propagates backwards in the ejecta , and the contact discontinuity that separates the ejecta from the ism .
these two media are a single fluid within the hydrodynamic solver , with their respective position known thanks to a passive scalar @xmath11 tracing the ejecta .
although our numerical setup is very efficient in terms of computational resources ( thanks to the adaptive mesh refinement and comoving grid ) , the shock fronts are affected by the so - called carbuncle instability @xcite because they are quasi - stationary . in paper
ii focused on the thermal emission , the effect was noticeable but could be neglected , given that most of the radiation came from the ejecta . in this paper focused on the non - thermal emission , it appeared mandatory to improve the numerical stability of the forward shock front , given that most of the radiation comes from its immediate vicinity . to this end
, we implemented a smoothing procedure within the code : in the medium located just upstream of the shock , the density and pressure are replaced by their azimuthally - averaged values .
this leads to a much cleaner shock front , while still preserving the overall remnant s evolution , and without disturbing the flow in the downstream region .
however , this restricts our ability to investigate the propagation of the forward shock in a medium that would vary strongly with azimuth .
diffusive shock acceleration is computed at each time step using the semi - analytical kinetic model of blasi ( @xcite ; @xcite ) , as explained in papers i and ii .
given the shock properties ( velocity and upstream conditions ) , this model jointly solves the particle spectrum @xmath12 and the fluid velocity profile @xmath13 .
note that it can only find stationary solutions , so , assuming that the acceleration process can be described as a succession of quasi steady - states , we re - compute the results after each hydrodynamic time - step , given the current flow conditions and the maximum energy particles have reached at that time .
we recall that the injection of protons is parametrized with the thermal leakage recipe of @xcite , controlled by the free parameter @xmath14 defined as @xmath15 where @xmath16 is the injection momentum of particles and @xmath17 is the mean thermal momentum immediately downstream of the shock . by definition , particles at @xmath16 are somewhat more energetic than the average , so that their larmor radius is a few times larger than the width of the collisionless shock . as a result
, @xmath14 is expected to be in the range 24 , which corresponds to a fraction of particles injected at the shock front , @xmath18 , ranging from roughly @xmath19 to @xmath20 . from a theoretical perspective
, the injection mechanism for electrons is less clear than for protons , but , apart from the vicinity of the minimum and maximum momenta , particles of same rigidity @xmath21 will have the same spectral shape ( @xmath5 is the particle momentum and @xmath22 its charge ) .
so , as commonly done , we normalize their spectrum with respect to the protons spectrum : @xmath23 with the normalization @xmath24 being a free parameter . the maximum energy of protons is limited by both the finite age and finite size of the remnant ,
according to the diffusion time- and length - scales @xcite , under the common assumption of bohm diffusion . for electrons ,
the maximum energy is limited by radiative losses ( more in section [ sec : model - down - part ] ) , and is computed according to the procedure outlined in @xcite .
when ions are efficiently accelerated , their pressure modifies the shock s structure : a precursor is formed ahead of the shock front , where the ambient medium is pre - accelerated and pre - heated by the streaming energetic particles @xcite .
then it is useful to make the distinction between the unperturbed ism far upstream of the shock front ( denoted hereafter by subscript 0 ) , the perturbed ism immediately upstream of the shock front ( subscript 1 ) , and the perturbed and compressed ism immediately downstream of the shock front ( subscript 2 ) .
we can now define two compression ratios : @xmath25 over the actual discontinuity ( from medium 1 to medium 2 ) , and @xmath26 over the whole structure ( from medium 0 to medium 2 ) .
the compression ratio @xmath25 ( resp .
@xmath26 ) is expected to be lower ( resp .
higher ) than in the standard adiabatic case @xmath9 . in our model , we implement this back - reaction of particles on the shock with a convenient pseudo - fluid approach : the fluid evolved by the hydrodynamic code in the shocked region accounts for both the thermal and non - thermal populations .
this is made possible through the use of an effective adiabatic index @xmath27 , which is a function of both time and space .
this variable index is set according to the acceleration model of section [ sec : model - accel ] , and embodies the effect of dsa on the flow outlined in the previous paragraph . in effect
, it makes the fluid more compressible , so that the shocked region gets denser and narrower , as required ( see @xcite and paper i ) .
this hydro - kinetic coupling is applied to the initial snr profiles , computed from the self - similar solutions of @xcite , and then after each hydrodynamic update in ramses . of course ,
ideally one would like to directly integrate the transport equation of particles over time ( as done e.g. in @xcite ) , however such simulations are computationally expensive , and still unpractical in 3d for realistic applications . in our work
we choose to focus on the description of the snr evolution .
we note that both @xcite and @xcite used the same model of acceleration in their modeling of tycho .
the former coupled it to prescribed , analytical , un - modified solutions of the shock evolution , while the later coupled it self - consistently to a hydro code as we do .
both @xcite and @xcite assumed spherical symmetry , which is appropriate after the remnant has entered the sedov - taylor phase of expansion .
our setup can accurately describe the dynamics of the shell and its interior from a young age , and as it evolves through different stages .
both theoretical and observational evidence imply that energetic particles can substantially amplify the magnetic field upstream of the shock , by exciting magnetic waves ( see @xcite and @xcite for recent reviews and references therein ) . because the magnetic turbulence controls the diffusion of the particles , this constitutes a _ second _ back - reaction loop .
this effect needs to be included when computing the non - thermal emission , because the fate of the magnetic field is critical for the emission of energetic electrons . as we shall see
, it also impacts the intensity of the shock modification , in a way that is similar to changing the injection level .
because the latter is poorly constrained , we had deliberately not included a discussion of the impact of the magnetic field in paper ii focused on the thermal emission . in this paper , we explicitly describe its evolution in the precursor under different scenarios . the shock precursor ( the region where streaming energetic particles can affect the incoming flow ) is not modeled in the hydrodynamic part of the simulation .
it is modeled inside the kinetic acceleration model , under the assumption that the diffusion coefficient @xmath28 is a fast and strictly rising function of momentum @xmath5 .
if the shock travels at speed @xmath29 , particles of momentum @xmath5 can travel a certain distance @xmath30 ahead of the shock front , where the fluid has a density @xmath31 and a velocity @xmath32 .
the shock modification then follows from the condition of conservation of the total momentum between the far unperturbed upstream medium and any point @xmath33 in the precursor : @xmath34 where @xmath35 is the dynamical pressure , @xmath36 is the thermal pressure , @xmath37 is the particle pressure computed from their distribution @xmath12 : @xmath38 with @xmath5 and @xmath39 being the momentum and velocity of the particles , and @xmath40 is the pressure of the magnetic waves .
it is the purpose of the subsequent sections to provide estimates of @xmath41 and @xmath42 .
various instabilities have been invoked to explain magnetic filed amplification ( mfa ) in the vicinity of snr shocks , both resonant and non - resonant , operating at different time- and length - scales .
here we use a simple model for the growth of the waves pressure , proposed by @xcite for the most commonly considered resonant instability : @xmath43 where @xmath44 is a free parameter ( that is also related to field damping and plasma heating , see section [ sec : model - mfa - damping ] ) .
this formula was derived under the assumption of large sonic and alfvnic mach numbers in the unperturbed upstream medium .
we recall that the sonic mach number is defined as @xmath45 where @xmath46 is the sound velocity in the medium considered , defined as @xmath47 and that the alfvnic mach number is defined as @xmath48 where @xmath49 is the alfvn velocity in the medium considered , defined as ( in gaussian units ) @xmath50 the total magnetic field at point @xmath33 is then given by @xmath51 where @xmath52 is the ordered component of the ambient field .
we note that formula ( [ eq : pw , p - res ] ) was also used by @xcite , and that a different version was proposed in @xcite and used by @xcite .
the later was obtained by expressing the alfvn speed using the amplified @xmath53 instead of the ambient @xmath52 .
this may seem more consistent , however when the field gets turbulent the very definition of @xmath54 becomes questionable as we shall see in the next section .
we note that @xcite finally opted to set the field everywhere in the precursor equal to the value predicted by the formula immediately upstream of the shock , which may seem a bit artificial .
we would like to recall that , anyway , all such formulas are only heuristic formulas , derived to encapsulate the overall behaviour of the magnetic turbulence . addressing mfa directly would require a different numerical treatment , and is well beyond the scope of this paper our point here is just to show how the inclusion of mfa can affect the computation of the emission from a realistic snr . to complement our work ,
we note that recently @xcite performed a comparison of different results obtained for mfa in the precursor : the two aforementioned recipes from @xcite , and two more models for @xmath55 that are parametrizations from simulations made by other groups .
they find that the last three models do not match observations of real snrs .
when the magnetic field is substantially amplified , the common assumption of largely super - alfvnic flows may no longer hold true .
it is important to recall that the charged energetic particles are tied to the flow because they are being scattered off the magnetic waves , so that the velocity that matters to them is not the fluid velocity @xmath56 but rather the waves velocity @xmath57 . in the mfa picture ,
alfvn waves are generated by particles that are counter - streaming the flow , so that @xmath58 at point @xmath33 .. it is known that taking @xmath59 strongly affects the spectra , but the physical justification for such a choice is far less clear than in the precursor . ]
this value is used in our acceleration model for both the transport of particles and the growth of waves . the difference between @xmath56 and the actual @xmath57
is commonly referred to as the alfvnic drift .
it had long been neglected on the assumption that @xmath60 , but it was recently realized that it should not be neglected when mfa is efficient and the magnetic field gets much higher than in the ism .
because of the alfvnic drift , particles experience smaller velocity jumps that lead to steeper spectra and less modified shocks .
this turns out to be an important ingredient in the modeling of non - linear acceleration at snr shocks @xcite .
however , one difficulty is that , when the field gets highly turbulent through mfa , it is not clear whether the velocity of the perturbations should still be taken as the alfvn velocity defined by equation ( [ eq : va ] ) , and if so whether it should be computed in the amplified magnetic field . following others @xcite , we take the alfvnic drift to be the alfvn velocity @xmath61 in an effective magnetic field @xmath62 varying between the ambient field @xmath52 and the total amplified field @xmath63 : @xmath64 with free parameter @xmath65 . mfa can also affect the plasma that is supporting the waves . if we were to assume adiabatic compression in the precursor , the thermal pressure at point @xmath33 would simply be given by @xmath66 where @xmath67 is the adiabatic index of the thermal fluid .
but several mechanisms can potentially alter the energy distribution in the precursor .
in particular , if the magnetic waves discussed in section [ sec : model - mfa - amplification ] are damped in the plasma , then the magnetic energy will be dissipated as heat in the plasma .
@xcite then proposed the following recipe ( obtained for large @xmath68 ) : @xmath69\ ] ] where the heating is parametrized by the parameter @xmath70 , that balances the factor ( @xmath71 ) in equation ( [ eq : pw , p - res ] ) .
this recipe , with @xmath72 , has often been used in kinetic models of dsa , long before mfa was established , in order to temper the effect of back - reaction and obtain reasonable shock modifications ( see the discussion in @xcite ) .
later on , it was realized that amplification and damping are competing processes , and that either can back - react on the flow @xcite as explained in the next section .
if the pressure of magnetic waves reaching the sub - shock is high ( case @xmath73 ) , then it must be included in the energy budget of the shock , and the standard rankine - hugoniot jump condition for the thermal pressure @xmath74 has to be modified according to @xmath75 this simplified formula was derived by @xcite , assuming the ordered magnetic field @xmath52 is initially parallel to the shock .
the key point is that , upstream of the shock discontinuity , the pressure of magnetic waves @xmath76 can become comparable to the thermal pressure @xmath77 , even though it is negligible compared to the dynamical pressure @xmath78 both downstream and upstream of the shock . when @xmath79 , according to equation ( [ eq : pth,2/pth,1-pw ] ) the thermal pressure is reduced , so that the effect of particle back - reaction is offset to some point : energetic particles tend to make the medium more compressible , but the magnetic field that they amplify tend to make it less so .
particles accelerated at the shock front are eventually advected downstream , where they can be traced by their non - thermal radiation . in our model , this step is post - processed : it is computed at the end of the hydrodynamic run in order to limit the numerical requirements . in paper ii , we showed how to obtain the thermodynamic properties of the thermal plasma in the shocked region for the computation of the thermal emission . similarly , in this section , we show how we reconstruct the quantities of interest for the computation of the non - thermal emission : particle spectra and magnetic field .
the acceleration model provides the spectrum of particles @xmath80 ( for both protons and electrons ) immediately downstream of the forward shock at each time @xmath81 .
we assume that particles are coupled to the gas inside the shocked region , which amounts to neglecting the diffusion of the particles of the highest energies .
this assumption , together with the assumption of a series of quasi - steady states at the shock front , seems reasonable for the bulk of particles ( that contributes most of the pressure ) , however it restricts our ability to describe the fate of the highest energy particles ( that will eventually escape the remnant anyway ) .
the particle spectra are thus advected towards the interior of the remnant , while they suffer energy losses .
the spectrum @xmath12 at the final time @xmath3 in a cell at position @xmath4 inside the shell is obtained from the relations
@xcite : @xmath82 and @xmath83 with @xmath84 and @xmath85 where @xmath86 is the time at which the material in the cell was shocked and @xmath87 is the position of the shock at that time .
@xmath88 is a constant which in gaussian cgs units reads : @xmath89 where @xmath22 is the charge of the particle ( @xmath90 or @xmath91 ) and @xmath92 its mass ( @xmath93 or @xmath94 ) .
formulae ( [ eq : f / f0 ] ) and ( [ eq : p / p0 ] ) are strictly valid for relativistic particles of @xmath95 .
the variable @xmath96 represents the adiabatic losses that shift the spectrum in energy without altering its shape ( independently of the particle mass or charge ) .
the variable @xmath97 represents the radiative losses ( in the magnetic field ) that alter the shape of the spectrum ; the effect is important only for electrons given the ratio @xmath98 .
the synchrotron loss time in a constant magnetic field @xmath99 , defined so that @xmath100 , is given for relativistic particles by @xmath101 that is for electrons @xmath102 in practice , in the code ramses , we advect the following two quantities inside the shocked region as passive scalars : @xmath103 is the cell age that gives knowledge of @xmath86 ; and @xmath104 can be thought as a loss age that permits the computation of @xmath97 ( for both protons and electrons ) . to compute the radiative losses of electrons inside the shell , as well as to compute their radiation , it is also necessary to have a description of the evolution of the magnetic field after it crossed the shock .
the value of @xmath99 immediately downstream of the shock is provided by the acceleration model . according to the discussion of section [ sec : model - mfa ] ,
the pre - shock field consists of both the ambient field and the amplified field ( equation [ eq : bp ] ) . assuming that the ambient field @xmath52
is isotropized in the vicinity of the shock , may not seem fully consistent with the models for the behavior of the magnetic field in the precursor presented in section [ sec : model - mfa ] : formulas ( [ eq : uw ] ) and ( [ eq : pth,2/pth,1-pw ] ) were nominally obtained by assuming that the magnetic field @xmath52 is initially parallel to the shock .
however a fully consistent treatment of the magnetic field would be difficult , especially at this level of modeling .
we note that @xcite used a similar approach : they assumed that the quadratic sum of the ordered and turbulent components of the field is isotropic immediately upstream of the shock .
the aim of the aforementioned formulas is just to catch some important effects , that are believed to extend when the field gets fully turbulent : the fact that the shock jump is reduced , as seen by particles , because of the waves speed @xmath57 , and the fact that the shock jump is reduced , as seen by the fluid , because of @xmath105 .
] then two out of its three components are compressed , by a factor taken to be @xmath26 . for the turbulent component
, the calculation shows that it is compressed by a factor of @xmath25 at the sub - shock @xcite . the total magnetic field immediately downstream of the shock is then given by @xmath106 we note that when mfa is efficient ( @xmath73 ) , the downstream @xmath107 is totally controlled by the turbulent component @xmath108 , so that the exact fate of @xmath52 does not matter much .
when mfa is quenched ( @xmath109 ) , then @xmath108 is negligible , and the chosen formula still allows for some reasonable compression of @xmath52 .
downstream of the shock , we assume that the magnetic flux is frozen in the plasma .
its evolution can then be re - constructed analytically , without the need to activate the mhd solver of ramses .
the two components of the field , its radial component @xmath110 ( parallel to the shock normal ) and its tangential component @xmath111 ( perpendicular to the shock normal ) evolve separately according to @xcite : @xmath112 where @xmath113 is the current radius and @xmath114 the shock radius when @xmath99 was compressed .
this computation applies to the shocked ism only , between the forward shock and the contact discontinuity .
the evolution of the magnetic field is not computed at smaller radii , since the field is thought to be extremely low in the ejecta due to adiabatic expansion @xcite .
accelerated particles produce non - thermal emission over a very wide energy band : energetic electrons generate radio to x - ray photons through interaction with the magnetic field ( synchrotron radiation ) and @xmath0-ray photons through interaction with ambient photons ( inverse compton effect ) , energetic protons generate @xmath0-ray photons through the collisional creation and subsequent decay of neutral pions .
energetic protons can also produce secondary leptons , but their own radiation is usually negligible inside snrs , and is not considered here .
we also neglect the bremsstrahlung emission from accelerated electrons , which is usually found to be a sub - dominant process in snrs ( except possibly in the center of the shell , see @xcite ) .
we already noted in section [ sec : model - down - part ] that synchrotron emission limits the maximum energy of electrons through losses , and thus needs to be computed consistently .
apart from this , the computation of the particles emission can be decoupled from the computation of the snr s evolution and of dsa at the shock . in our simulations
it is computed in post - processing , at any chosen age . to compute the non - thermal emission of particles
, we use the general - purpose code aura ( edmon , private communication ) , which is a rewrite of the code cosmicp presented in @xcite .
it makes use of the formalism of @xcite for leptonic processes , and of the formalism of @xcite and @xcite for hadronic processes . from the distribution function @xmath12 of particles and the relevant ambient parameters
, the code computes the volume emissivity of photons per unit frequency , for each radiative process of interest .
pion decay emission depends on the ambient density which is known from the hydrodynamic simulation .
inverse compton emission depends on the background radiation , here taken to be the uniform cosmic microwave background ( cmb ) .
synchrotron emission depends on the intensity of the magnetic field which is modeled according to the previous sections . in this work ,
we focus on the non - thermal emission coming from the shocked ambient medium , due to particle acceleration at the forward shock .
the computation of the non - thermal emission is thus done , cell by cell , in the shocked ism only ( between the outermost position of the contact discontinuity and the azimuthally averaged position of the forward shock ) .
it is important to understand that , even though our acceleration model at any time is approximate , we keep track of the particle acceleration history over the whole simulation , in order to compute their emission at any point in the shocked ism at the time considered .
the photon spectra are then projected to obtain 2d maps of the emission that can be directly compared with observations .
so even if the maps presented here are a snapshot of the snr at a given age , they do integrate the particle spectra in both space ( along the the line of sight ) and time ( along the shock history ) .
our purpose in this paper is not to make a detailed modeling of a specific object , but rather to point out the important physical aspects to consider when interpreting observations .
therefore , we do not explore the full parameter space , but rather use a single set of parameters appropriate for a type ia supernova such as tycho s snr ( as in our study of the morphology in paper i and our study of the thermal x - ray emission in paper ii ) .
we are considering the remnant formed by low - mass ejecta in a homogeneous medium .
we set the explosion energy to @xmath115 , and the ejecta mass to @xmath116 solar masses .
we assume the ejecta to be initially spherically symmetric , and adopt a power - law profile with index @xmath117 .
we assume a uniform ambient medium ( that is , a power - law profile with index @xmath118 ) of density @xmath119 .
we use these profiles to initialize the hydrodynamic code from self - similar solutions at a small age of @xmath120 years .
we then let the code evolve the profiles until @xmath121 years .
particles are accelerated continuously in time at the forward shock . for the injection of protons , we take @xmath122 , which produces reasonable injection fractions at the shock front of a few @xmath123 .
similar values were used by other groups in their modeling of tycho s snr : @xcite found a best fit of @xmath124 , @xcite used @xmath125 and @xcite preferred @xmath126 . for the injection of electrons , we set @xmath127 , as observed in the local cosmic - ray spectrum .
we note that the previously mentioned models of tycho s snr need smaller values : @xcite obtained @xmath128 , @xcite @xmath129 , and @xcite @xmath130 .
however @xcite proposed a model in which @xmath127 is a universal value that can fit all galactic snrs . in any case , as electrons play no dynamical role in our model , their spectra and thus their emissions could be re - scaled as needed .
we assume an ambient magnetic field @xmath131 for the diffusion of particles , representative of the average galactic field .
a key aspect discussed in this paper is the amplification of the magnetic field by particles , which in our model depends on the free parameter @xmath70 ( see equations ( [ eq : pw , p - res ] ) and ( [ eq : pth , p - alfven ] ) ) . in all our simulations , we considered the following two limit cases : * @xmath72 : the waves are all damped , so that the net magnetic field is limited to the ambient value , which is only enhanced at the shock front ( by standard compression ) .
the intensity of the back - reaction is limited by the pre - heating of the plasma in the precursor ( see section [ sec : model - mfa - damping ] ) .
* @xmath132 : the waves grow freely until saturation of the amplification process , and are advected to the sub - shock .
the intensity of the back - reaction is limited by the magnetic tension at the shock discontinuity ( see section [ sec : model - mfa - advection ] ) .
in addition , we consider two limit cases for the alfvnic drift , controlled by the parameter @xmath133 ( see equation [ eq : f_a ] ) : * @xmath134 : the net upward velocity of the waves is computed in the ambient field , and is therefore very close to the local velocity of the flow .
the inclusion of the drift is then only a minor correction in the computation of dsa , regardless of @xmath70 .
* @xmath135 : the net upward velocity of the waves is computed in the total magnetic field . in the case
@xmath72 this does not make any difference , since there is not net mfa .
but in the case @xmath132 , the inclusion of the drift is an important correction in the computation of dsa , leading to steeper spectra and less modified shocks . by comparing such extreme cases for mfa and alfvnic drift
, we encompass a large range of possibilities , and hope to account for the uncertainties in the modeling of the behavior of the magnetic field that were mentioned thorough section [ sec : model - mfa ] .
we note that our approach is orthogonal to the one by @xcite , who compared different mfa recipes , but arbitrarily set all free parameters to their median values ( @xmath136 and @xmath137 ) .
finally , we have to specify the numerical resolution of the simulations .
regarding acceleration , for the particles , the momentum grid is logarithmically sampled between @xmath138 and @xmath139 with 20 points per decade .
this is largely sufficient to get proper estimates of the shape of the spectrum and of the resulting pressure and back - reaction .
regarding transport , for both the fluid and the particles , the spatial amr grid has a formal resolution of @xmath140 , so that we produce emission maps of @xmath141 pixels in each energy bin ( as in papers i and ii ) .
note that we only simulate one octant of the remnant , and therefore obtain projected maps of one quarter of the remnant .
the code automatically refines the grid around the discontinuities in the profiles , and because of the rayleigh - taylor instabilities , most of the shocked region ends up being at the highest resolution of @xmath140 .
the rest of the shocked region , and more , is at the intermediate resolution of @xmath142 , and the rest of the grid is at the lowest resolution of @xmath143 .
it is important to see that , because we simulate an octant of a sphere embedded in a cartesian grid , the refined regions are always present along any line of sight .
so when doing the projection from 3d to 2d , every pixel inside the shell gets data that were actually computed on its size ( where it mattered ) . regarding emission , for the photons , the energy grid is logarithmically sampled with 10 points per decade over several decades in energy : from @xmath144 ev to @xmath145 ev for synchrotron , from @xmath19 ev to @xmath146 ev for inverse compton , and from @xmath147 ev to @xmath146 ev for pion decay .
this encompasses the energy ranges of all relevant instruments , as shown in the following section .
in this section , we present a variety of synthetic emission maps and spectra generated from our 3d simulations , in order to investigate the effects of acceleration on observational diagnostics of the snr , and specifically the impact of both hydrodynamic and magnetic back - reactions occurring in efficient dsa . as in paper
ii , in order to illustrate the effect of the hydrodynamic modifications , in all the plots we compare data obtained without including the back - reaction from the pressure of accelerated particles on the flow ( cases labeled off " ) so that the shock structure is unmodified , and data obtained including the back - reaction from the pressure of accelerated particles ( cases labeled on " ) so that the shock structure is modified .
it is important to understand that our on / off switch controls the hydrodynamic back - reaction only : in both cases we run the same non - linear acceleration model , so that the off cases are not in test - particle " regime in terms of magnetic field and particle spectra .
new in this paper , in order to investigate the magnetic back - reaction ( see section [ sec : parameters - mfa ] ) , we compare on each figure the two limit cases obtained with complete damping ( @xmath72 , the magnetic field remains at the ambient level ) and with no damping ( @xmath132 , the magnetic field is amplified well above the ambient value ) .
when magnetic amplification is at work , we in addition consider the effect of computing the alfvnic drift in the ordered ambient field ( @xmath134 , the drift is then negligible with our parameters ) or in the turbulent amplified field ( @xmath135 , the drift may be important if there is much amplification ) .
we first show the quantities related to the plasma and particles , on which the non - thermal emission relies ( section [ sec : results - diagnostics ] ) , before showing the emission itself ( section [ sec : results - emission ] ) . here
we show the quantities , diagnosed from the ramses code , needed to compute the emission .
we present all these data in the form of slices ( in a plane z = 0 perpendicular to the line of sight ) in order to show the inner structure of the remnant .
the fluid density inside the shocked region is shown in figure [ fig : map - np ] .
the shocked region is bounded by the forward shock at the outer edge and by the reverse shock at the inner edge .
the shocked ejecta are shaped by the development of the rayleigh - taylor instability , showing a distinctive pattern of fingers and holes near the contact discontinuity . when comparing the on vs. off cases ( right plots vs. left plots ) , it is obvious that the shocked region gets compressed when the particle pressure is taken into account , as expected ( see paper i ) . in the case with no net magnetic field amplification ( @xmath72 ) , its width is smaller by about 1/3 , and the post - shock density is higher by a factor of about 2 . the thin layer of dense , recently shocked ism is now almost in contact with the rayleigh - taylor fingers of the ejecta .
when comparing the two limit mfa cases @xmath132 vs. @xmath72 ( middle and bottom plots vs. top plots ) , no impact is visible in the off case , as expected for such a purely hydrodynamic diagnostic . for the on case , as long as the alfvnic drift is negligible ( @xmath134 , top and middle plots ) , the level of back - reaction is overall the same : it is just a little higher with efficient mfa ( @xmath132 , middle plots ) .
this result that was not obvious , given that the back - reaction is adjusted through completely different mechanisms : pre - heating in the precursor for @xmath72 , vs. magnetized sub - shock for @xmath132 .
if the alfvnic drift is important ( @xmath135 , bottom plots ) , the level of back - reaction is lower : the shock is not as modified .
quantitatively , the width of the shocked region is reduced by only about 1/4 , and the post - shock density is enhanced by only about 50% . the magnetic field downstream of the forward shock is shown in figure [ fig : map - b_fs ] .
it is only shown where the ism fraction ( @xmath148 , see section [ sec : model - hydro ] ) is @xmath149 , because its evolution is not computed in the ejecta ( grey area ) . with no net mfa ( @xmath72 , top plots ) ,
the downstream evolution of the field for a modified shock is similar to the case of an un - modified shock , but over a smaller region . in this case
, @xmath107 is just the compression of the ambient @xmath52 at the shock , with ratio @xmath113 dependent on back - reaction . according to equation ( [ eq : b2_shock ] )
, the ambient field @xmath150 is compressed by a factor @xmath151 for @xmath152 ( the reference value ) and @xmath153 for @xmath154 ( the maximal value predicted by the acceleration model ) , leading to @xmath155 as observed on the maps behind the shock front .
then @xmath107 decreases away from the shock , according to equations ( [ eq : br_down ] ) and ( [ eq : bt_down ] ) ( with the latter dominating close to the contact discontinuity , because of the drop in density ) . in the case with efficient mfa ( @xmath132 , middle and bottom plots )
, @xmath107 is much higher : it reaches hundreds of @xmath156 ( note the different scale ) . in this case , the values are slightly lower in the off vs. on cases ( by about 20% ) , because in our mfa model weaker shocks are less efficient at generating magnetic turbulence ( so computing mfa in the un - modified shock profile results in over - estimating the magnetic field ) .
also , in this case @xmath132 , not much radial variations are seen downstream of the shock , so that the variations seen in the @xmath72 case must be counter - balanced by the time evolution of the shock properties : at earlier ages ( that correspond to material close the the contact discontinuity ) the shock was stronger , inducing more efficient mfa .
we recall that in order to compute the evolution of particle spectra and their emission , we need to reconstruct the history of the material downstream of the shock ( see section [ sec : model - down ] ) .
the synchrotron radiative loss age , as defined by equation ( [ eq : tau_l ] ) , is shown in figure [ fig : map - tl_fs ] .
the higher @xmath157 , the more particles are likely to be affected by losses ( depending on their mass and energy ) .
@xmath157 rises downstream away from the shock , because material farther away from the shock crossed it at earlier times .
the maximal value , reached close to the contact discontinuity , is slightly lower for the modified shocks ( case on , right plots , vs. case off , left plots ) , following the magnetic field .
the value of @xmath157 is strongly dependent on @xmath107 , and thus strongly depends on the mfa recipe : when @xmath132 ( middle and bottom plots ) it reaches values about 2 orders of magnitude higher than when @xmath72 ( top plots ) .
when @xmath132 , its value is lower by a factor of about 2 when the alfvnic drift is computed in the amplified field ( @xmath135 , bottom plots ) compared to when it is computed in the ambient field ( @xmath134 , middle plots ) .
the quantity @xmath157 was purposely defined to be independent of the mass @xmath92 of the particle , and is used for both protons and electrons . the actual loss factor @xmath97 ( not shown here ) , given by equation ( [ eq : theta ] ) , is inversely proportional to @xmath92 and thus is very different for protons and electrons . from equations ( [ eq : f / f0 ] ) and ( [ eq : p / p0 ] ) , we see that the higher @xmath97 , the lower the momentum at which particles will be affected . for protons
@xmath158 is very high ( more than @xmath159 @xmath160 ) , so that they are not affected by radiative losses ( they still suffer adiabatic losses , through the parameter @xmath96 in the equations ) . for electrons
@xmath161 is comparable with @xmath160 , so that they are strongly affected by radiative losses ( on top of adiabatic losses ) .
this implies different maximum energies for protons and electrons . the maximum momentum of protons @xmath162 is shown in figure [ fig : map - plp ] .
it is always the highest at the shock front , and decreases downstream because of the adiabatic losses ( contrast with figure [ fig : map - tl_fs ] ) .
the values reached by @xmath162 are notably higher ( note the different scales ) when mfa is efficient : between roughly 100 and 300 tev / c for @xmath132 ( middle and bottom plots ) , versus between only 10 and 20 tev / c for @xmath72 ( top plots ) .
this is because with a much higher magnetic field the particles are better confined in the vicinity of the shock .
for the same reason , @xmath162 is slightly lower for modified shocks ( right plots vs. left plots ) . in the case of efficient mfa ( @xmath132 )
, @xmath162 is even higher when the alfvnic drift is computed in the amplified field ( @xmath135 , bottom plots ) compared to when it is computed in the ambient field ( @xmath134 , middle plots ) , which is opposite to the behaviour of the magnetic field .
this demonstrates the non - trivial impact of the drift in the non - linear acceleration process .
the maximum momentum of electrons @xmath163 is shown in figure [ fig : map - ple_fs ] .
as for protons , it is highest at the shock front and decreases downstream , and the effect is even more marked than for protons because of the combined ( adiabatic + radiative ) losses . in the case @xmath72
( top plots ) , @xmath163 remains of the same order as @xmath162 ( note the same scales ) , meaning that it is limited by the finite age and/or size of the remnant as much as by the magnetic field . in the case
@xmath132 , @xmath163 is much lower than @xmath162 , because it is limited by the radiative losses .
according to equation ( [ eq : t_sync-2 ] ) , after 500 yr one still has electrons of energies up to @xmath164 tev in a uniform @xmath165 , appropriate for the un - amplified case @xmath72 , and up to only @xmath166 tev in a uniform @xmath167 , appropriate for the amplified case @xmath132 .
in reality , the downstream magnetic field is not constant , because of the spherical geometry and shock history , yet these order of magnitude values are in broad agreement with the results obtained in figure [ fig : map - ple_fs ] . in the case of efficient mfa ( @xmath132 ) ,
@xmath163 is slightly higher when the alfvnic drift is computed in the amplified field ( @xmath135 , bottom plots ) compared to when it is computed in the ambient field ( @xmath134 , middle plots ) , which follows from the behaviour of the magnetic field .
but this is a secondary effect : the main result is that when @xmath132 highly energetic electrons ( @xmath168 tev ) are confined very close to the shock front , which will restrict their emission . using the data shown in the previous section , we compute the non - thermal emission in each cell , which
is then projected along the line of sight ( arbitrarily chosen to be the z - axis of the simulation box ) to generate synthetic maps .
the maps , in different energy bands , are presented in section [ sec : results - emission - maps ] .
spatially - integrated spectra are also presented in section [ sec : results - emission - spectra ] .
we have computed projected maps in 5 energy bands : + @xmath169 in radio from @xmath170 to @xmath123 ev , that is from 30 mhz to 30 ghz , which covers the range of lofar to effelsberg instruments ; + @xmath169 in soft x - rays from 0.3 to 10 kev , the range of chandra , xmm - newton , suzaku ; + @xmath169 in hard x - rays from 5 to 80 kev , the range of nustar ( and partially that of integral ) ; + @xmath169 in low - energy @xmath0-rays from 0.1 to 100 gev , the range of fermi and agile ; + @xmath169 in high - energy @xmath0-rays from 0.01 to 10 tev , the range of hess 2 , magic , veritas .
+ all the maps show the differential volume emissivity as computed by the code , in erg/@xmath171/s / ev , as indicated on the colour scales .
note that the angular resolution of our maps is limited by the numerical resolution of our grid , they have not been convolved with the psf of any particular instrument .
we recall that the maps presented here have 1024 pixels along each direction , at the age shown here they span 5.61 pc , hence a physical cell size of 0.0055 pc , and an angular resolution of about @xmath172 at a distance @xmath173 kpc .
the synchrotron emission from electrons extends from the radio to the x - rays .
it is shown in the radio band in figure [ fig : map - ntsy - radio ] and in the soft x - ray band in figure [ fig : map - ntsy - xsoft ] .
the morphology of the remnant in radio does not vary much depending on the various back - reaction cases .
the maximum of emission is always well behind the shock , close to the contact discontinuity , where low energy ( gev ) electrons have accumulated .
this allows us to see the imprint of the rayleigh - taylor instability , as a negative image of the fingers .
the emissivity is higher for @xmath132 vs. @xmath72
( note the different scales ) , especially when @xmath134 , because of the much higher value of @xmath107 : electrons radiate more strongly in the amplified field , while at the same time they are not affected much by radiative losses at these energies .
the morphology of the remnant in x - rays is noticeably different .
the maximum of emission is now close to the shock , where the high energy ( multi tev ) electrons are confined ( see figure [ fig : map - ple_fs ] ) , and decreases rapidly downstream ( note that the scales now span two orders of magnitude , compared to one for the radio ) .
we recall that the emissivity cut - off happens to be in the x - ray band ( see also figure [ fig : spc - back - zeta ] ) , and therefore critically depends on the fate of the electrons near @xmath174 .
the effect is much enhanced with efficient mfa ( @xmath132 ) because of the stronger losses in the amplified @xmath107 : very thin rims are visible immediately behind the shock , regardless of how the alfvnic drift is computed . as a result of the rapid drop of emissivity downstream of the shock , no more physical structures are visible inside the shell .
note that the overall emissivity levels are not as much affected by @xmath70 , as for the radio ( the same scale was used for all mfa cases ) .
moreover , there is a stronger contrast between the edge and the center of the shell in the on case , so that the rims look even thinner .
the morphology of the remnant in the hard x - ray band ( not shown here ) is similar to the soft x - ray map , but with much lower normalizations as we are sampling the steeply falling cut - off of the synchrotron emissivity .
the inverse compton emission from electrons extends over the @xmath0-ray bands .
it is shown around gev energies in figure [ fig : map - ntic - gsoft ] and around tev energies in figure [ fig : map - ntic - ghard ] .
first we see that the emissivity is lower for the on vs. off cases , as previously .
this emission does not explicitly depend on the magnetic field as the synchrotron emission , but it is still affected by the different mfa cases , because it depends on the number of emitting electrons left after losses . as a result , the emissivity is lower for @xmath132 vs. @xmath72 , by about one order of magnitude at gev energies and almost two orders of magnitude at tev energies ( note the different scales ) . mfa also changes the emission pattern , at both gev and tev energies , depending on where the emitting electrons are : widespread in the shocked ambient medium when @xmath72 ( so that the rayleigh - taylor fingers are visible ) , and close to the forward shock when @xmath132 ( so that inner structures disappear ) .
the effect is more pronounced at higher energies , because high - energy photons are emitted by high - energy electrons that are the most affected by losses , and thus restricted to the vicinity of the forward shock where they are accelerated .
we note that this effect does not depend much on the alfvnic drift .
the pion decay emission from protons extends over the @xmath0-ray bands .
it is shown around gev energies in figure [ fig : map - ntpi - gsoft ] and around tev energies in figure [ fig : map - ntpi - ghard ] .
less variations are seen than for the leptonic emissions ( note that the scale is the same for all maps in a given energy band ) .
the emission extends in most of the downstream region , it never peaks at the shock , it always reaches the contact discontinuity and therefore highlights the instabilities shaping it .
there is not much difference between the off and on cases ( except of course for the compression of the emitting region ) .
there are some differences in emissivity levels between the different mfa cases , that depend on the energy band . at gev energies ,
the emissivity is slightly reduced for @xmath132 , for both drift cases , and more so when @xmath135 .
this reflects a decreasing number of emitting protons , as spectra get different curvatures with different back - reaction recipes . at tev energies ,
the emissivity is slightly reduced for @xmath132 and @xmath135 because in that case spectra are steeper , but slightly enhanced for @xmath132 and @xmath134 , mostly because of the higher @xmath162 ( see figure [ fig : map - plp ] ) .
finally , we show the broad - band photon distribution in figure [ fig : spc - back - zeta ] , in order to compare the different emission processes for the different acceleration scenarios .
the energy bands used to produce the previous maps are labeled . for clarity , we only show here the spatially integrated spectrum , although we have computed it in every pixel of the map .
we recall that , because these emission spectra integrate the contributions of different parts of the remnant , that contain particles that were accelerated at different times in different conditions , there is not a single emitting population behind these spectra . for convenience , here we have multiplied the results by the emitting volume and by the emitted energy , to get the ( isotropic ) emissivity @xmath175 in erg / s , as in commonly done . to get the flux from the snr at the earth
, we need to multiply this emissivity by 8 ( as only one octant was simulated ) and divide by the illuminated surface @xmath176 at distance @xmath177 .
numerically , we get a translation factor of @xmath178/@xmath179 at 1 kpc . for comparison ,
fluxes observed for tycho s snr ( at a distance of @xmath180 kpc ) at various wavelengths are summarized in @xcite : in radio at 1.4 ghz : @xmath181 erg/@xmath179/s , in x - rays at 10 kev : @xmath182 erg/@xmath179/s , in @xmath0-rays above 1 tev : @xmath183 erg/@xmath179/s .
missing is the detection by fermi @xcite , for @xmath0-rays at 1 gev : @xmath184 erg/@xmath179/s .
so the fluxes we obtain are of the correct order of magnitude , although we have not attempted a fit of any particular object at this point .
we already observed that , for all three emission processes , the emissivity is on average higher towards the edge rather than in the center of the shell , regardless of the back - reaction effects .
this is expected since energetic particles are in the vicinity of the forward shock , plus there is a limb - brightening effect in projection .
the emission is always lower for the on vs. off cases ( solid vs. dashed lines ) , firstly because the emitting region is smaller and there are less emitting particles .
note however that the emission in the off case is unrealistic since it involves a lot of accelerated particles without hydrodynamic back - reaction .
more relevant to observations is the comparison of different magnetic back - reaction recipes in the on case .
there are important variations in emissivity level and cut - off energy depending on the amplification of the magnetic field ( @xmath132 , thick lines , vs. @xmath72 , thin lines ) . with a high @xmath107 ( obtained when @xmath132 ) , the emissivity level gets higher for synchrotron in radio ( by a factor of up to almost 10 ) , because the magnetic field is higher and there are as many low energy electrons .
it gets lower for inverse compton in @xmath0-rays ( by a factor of up to more than 100 ) because there are much less high - energy electrons , and it stays almost the same for pion decay in @xmath0-rays ( within a factor of about 2 ) because this hadronic process is not directly dependent on the magnetic field . with a high @xmath107
, the cut - off energy stays the same for synchrotron because , in all cases considered , the magnetic field is high enough so that the maximum energy of electrons is loss - limited . in a magnetic field @xmath99 emits synchrotron photons having energy @xmath185 , and in the loss - limited case and assuming bohm diffusion the maximum energy of electrons goes as @xmath186 , so that the maximum energy of photons goes as @xmath187 independent of @xmath99 @xcite.[note : loss - limited ] ] it gets lower for inverse compton ( by almost a factor of 10 ) because of the lower @xmath163 caused by strong radiative losses , and it gets higher for pion decay ( by more than a factor of 10 ) because of the higher @xmath162 caused by better confinement at the shock .
note that all these general behaviours are not affected by the exact value of the alfvnic drift ( @xmath134 for top plots , @xmath135 for bottom plots ) .
the latter does alter the cut - off energies , but in the same way for both kinds of particles ( towards slightly higher energies for a higher drift ) . also note that the pion - decay plateau of emission is slightly flatter for @xmath135 , which reflects a steeper spectrum of the emitting protons .
in this section , we recap our main results that are mostly pertaining to cosmic - ray acceleration at the blast wave of young type ia snrs like tycho , linking them with existing or future observations .
the most important effect highlighted in our simulations is the critical role of the magnetic field , that can be much affected by the accelerated protons , and that in turns much affects the accelerated electrons . in a cr - modified shock
, some important processes are occurring ahead of the shock front , in the shock precursor . as streaming particles
can themselves generate the magnetic turbulence that scatters them off , this makes a second back - reaction loop in dsa , in addition to the back - reaction exerted through their pressure on the fluid .
the fate of the leptonic component of the accelerated particles critically depends on the value of @xmath107 downstream of the shock . when mfa is efficient ( that is when the magnetic waves are growing and are not damped , the case @xmath132 )
, @xmath107 can easily be one order of magnitude higher than what would be expected from a mere compression of the ambient field at the shock front , regardless of the exact value of the shock compression ratio ( figure [ fig : map - b_fs ] ) .
the values we obtain , in the range of 200 to 300 @xmath188 g , are consistent with what has been inferred from observations of young snrs : for instance , @xcite estimated the downstream field of tycho s snr to be between @xmath189 and @xmath190 g .
our work is also in agreement with modeling of this particular snr : @xcite adopted the values of @xmath191 and @xmath192 g , @xcite calculated values of up to @xmath193 g , and @xcite found a current value of @xmath194 g .
because of radiative losses , these strong magnetic fields drastically limit the lifetime of multi - tev electrons that are thus confined very close to the shock ( figure [ fig : map - ple_fs ] ) . as as result , the synchrotron emission in x - rays produces a very thin rim immediately downstream of the shock ( figure [ fig : map - ntsy - xsoft ] ) .
this is in agreement with what is observed in several young snrs , in particular tycho @xcite
. mfa has mostly been discussed in the context of the synchrotron x - ray emission , but the inverse compton @xmath195ray emission also relies on the presence of high - energy electrons .
when @xmath107 is amplified , this emission is also concentrated towards the shock at energies close to the cut - off , which is around the tev ( figure [ fig : map - ntic - ghard ] ) .
so , in principle , the same kind of diagnostic for mfa would be possible in @xmath0-rays as in x - rays , if the post - shock region could be resolved .
also note that , further downstream , the non - thermal emission may bear the imprint of the rayleigh - taylor instabilities , depending on how far the particles can travel , which for electrons depends on the magnetic field and therefore on mfa .
for the synchrotron emission , we observe that radio images always show the turbulent interior of the remnant ( figure [ fig : map - ntsy - radio ] ) , whereas x - ray images never do ( figure [ fig : map - ntsy - xsoft ] ) , in agreement with observations . more interestingly ,
for the inverse compton emission , we observe that , in the tev range , the visibility of inner structures depends on mfa : they disappear when @xmath107 is efficiently amplified , leaving only the outer rim already mentioned ( figure [ fig : map - ntic - ghard ] ) .
unfortunately , current @xmath0-ray instruments are far from having the same angular resolution as current x - ray instruments , and can not resolve a snr shell with as a much details as we do in our simulations . however , the planned cherenkov telescope array ( cta ) observatory will aim at improving the spatial resolution in the tev band @xcite .
we recall that all these results depend on the value of two free parameters in our model : the level of wave damping @xmath70 ( equations [ eq : pw , p - res ] and [ eq : pth , p - alfven ] ) , and the factor @xmath133 determining the alfvnic drift ( equation [ eq : f_a ] ) , which both range between 0 and 1 .
out of these two parameters , @xmath70 is clearly the most important : @xmath133 has no noticeable effect when @xmath70 is close to 1 , and it brings a correction when @xmath70 is close to 0 . with these two parameters
, we have tried to offer a realistic , yet reasonably simple , modeling of the fate of the magnetic field around the shock front , suitable for the inclusion in large - scale simulations of snrs .
however , the process of mfa is still poorly understood , and ongoing studies keep unravelling new possible mechanisms with their share of complications ( see @xcite for a review of recent developments ) . being difficult to address directly , yet having a direct impact on the observational results , mfa definitively appears to be a critical topic in the studies of dsa .
we believe that we have bracketed a fair range of scenarios relevant to young snrs , although somewhat different recipes for the evolution of the magnetic field in the precursor could be proposed , and could possibly alter the value of @xmath107 by a factor of two or so .
we made some additional tests , using an alternative recipe for mfa proposed by @xcite ( see section [ sec : model - mfa - amplification ] ) .
we observed that the general features of the maps and spectra are of the same kind as reported here .
this model produces even higher magnetic fields , and so the general behaviours discussed here are enhanced .
however , this model leads to much less back - reaction ( for the same injection parameters ) , and so less modified shocks and steeper spectra ( in line with the modeling of tycho s snr by @xcite ) .
what this shows , is that the two back - reaction loops of dsa ( on the shock structure and on the magnetic field , via respectively the pressure and the streaming of energetic particles in the upstream ) , are not connected in any simple way . as a result
, both effects should be assessed individually in the modeling of a snr .
we have so far focused on the physical processes occurring in the upstream region , since this is where the amplification is believed to occur .
however , the subsequent fate of the magnetic field in the downstream region is not entirely settled .
@xcite argued that a high magnetic field could be efficiently damped just downstream of the shock , which would invalidate the standard interpretation of non - thermal filaments as being loss - limited .
we refer the reader to @xcite for a critical analysis of this possibility , and to @xcite and @xcite for a comparison of the two scenarios .
further downstream , the magnetic field is expected to peak close to the contact discontinuity ( e.g. @xcite ) .
it can be furthermore stretched and amplified by the development of rayleigh - taylor instabilities @xcite and may become dynamically important . our current model does not allow an investigation of these effects , that require mhd simulations . on a related matter , we note that , in our modeling , we do not let the accelerated particles interact with the contact discontinuity ( from which they are very close initially ) , or go past the contact discontinuity ( where they could in principle illuminate the dense ejecta ) . the first point would probably require a better treatment of the magnetic field close to the contact discontinuity as well , and the second point should probably be studied jointly with particle acceleration at the reverse shock ( on this topic , we refer the reader to the recent work by @xcite , and to the discussion in paper ii ) . in this paper , we focused our attention to the shocked ism , where the processes we consider are inevitable and indeed well observed . as stated in the introduction , the particles that are the easiest to track are the electrons , but the ones that are the most important for cr theory are the protons , since they make the bulk of galactic crs and can reach higher energies .
the emission at @xmath0-ray energies is of paramount importance for particle acceleration studies , because energetic protons only contribute in this band ( through pion decay ) , but unfortunately , electrons also radiate in the same energy band ( through inverse compton effect ) .
if mfa is not efficient ( @xmath72 ) , then it would be very hard to distinguish the emission mechanism based on the morphology alone ( compare top row of figure [ fig : map - ntic - gsoft ] to figure [ fig : map - ntpi - gsoft ] , and of figure [ fig : map - ntic - ghard ] to figure [ fig : map - ntpi - ghard ] ) .
if mfa is efficient ( @xmath132 ) , then the inverse compton emission could be distinguished as the component starting right at the shock front and peaking just behind it , with the pion decay emission peaking further behind , and extending far enough that it highlights the turbulent interior of the remnant provided the angular resolution is good enough . at the moment , given the instrumental limitations , distinguishing the two scenarios is only possible via spectral features .
these are the most visible at the lowest and highest energies . in between
, the shape of the spectrum depends on the shape of the parent distribution , which is much affected by non - linear effects , plus by losses for electrons , so that any modeling can hardly be conclusive . at the lower end
, the emission from protons can not extend much below @xmath196 mev ( a threshold corresponding to the mass of the @xmath197 particle ) , when the emission from electrons just decreases gradually . at the higher end
, the emission from protons can extend well above @xmath198 tev , when the emission from electrons can not .
we note that , by reducing the cut - off energy of the inverse compton emission , and increasing the cut - off energy of the pion - decay emission , efficient mfa ( case @xmath132 ) greatly helps to separate the leptonic and hadronic contributions ( see again figure [ fig : spc - back - zeta ] ) . as a result
, a clear observational detection of snrs in the 10100 tev range would be strong evidence for the presence of very energetic protons , as well as good evidence for efficient mfa .
it is therefore of paramount importance to extend the energy coverage of existing instruments .
the planned facility cta will be able to probe the tev cut - off in snrs , and thus identify the emission .
although the situation is still confused at @xmath0-ray energies for now , it shall be noted that observations in the radio to x - ray domains can greatly help to shed light on the mechanisms at play .
the properties of the synchrotron emission can be used to constrain the underlying electron distribution , and therefore the possible contribution through inverse compton ( for a given ambient photon field ) . in particular , observations of thin x - ray synchrotron rims suggest high magnetic fields , that disfavour inverse compton emission as a viable explanation of the @xmath0-ray emission .
this is in particular the case for tycho s snr @xcite .
moreover , the electrons are not believed to be able to generate these high magnetic fields that make them radiate : instabilities are driven by protons and other heavy nuclei . therefore , the radiation from electrons offers indirect diagnostics on the presence of energetic protons , and on the efficiency of particle acceleration at the blast wave . measuring efficient dsa in snrs
is an important step towards establishing them as cosmic - ray accelerators in the galaxy , however that does not tell the maximum energy achieved by the particles in these objects . for electrons , it can be directly inferred from studying the hard non - thermal x - ray emission @xcite , but it is limited by losses and can barely reach 100 tev . for protons
the situation is less clear . for tycho s snr
, veritas observations do not indicate a cut - off in the @xmath0-ray spectrum @xcite , so the maximum energy of protons can only be estimated based on the age and size of the system assuming a certain diffusion law .
for this remnant , @xcite obtained likely values between 100 and 600 tev , reaching 2 pev only in an `` extreme case '' with all parameters pushed to their limits .
we note that the amplification of the magnetic field has an opposite effect on the maximum momentum of protons , compared to electrons : a higher @xmath99 ( case @xmath132 ) , both upstream and downstream of the shock , helps to confine particles close to the velocity discontinuity , where they can gain energy . in our simple model
the acceleration timescale is directly proportional to @xmath199 on each side of the shock , where @xmath99 is the total field .
things can actually be more complicated : confining particles up to the highest energies requires having magnetic waves at all the relevant resonant wavelengths , and so amplifying the field does not always guarantee reaching higher energies @xcite . in any case , even when @xmath162 is boosted by mfa , it is still lower than the cr knee energy in our simulations , by a factor of almost ten , at less than 400 tev / c . in their modeling of tycho s snr , @xcite inferred a similar limit of 500 tev / c , while @xcite found even lower values in the range 40 to 100 tev / c .
although somewhat disappointing , we note that these results are in agreement with recent investigations of the timescales involved in the process , by the very first proponents of the mfa mechanism themselves @xcite . in this paper we considered a uniform medium , appropriate for a type ia .
we used typical values of the density and magnetic field , but these will of course vary from remnant to remnant , or even in the vicinity of a given remnant .
here we discuss how such variations affect our results , and how our model could be extended to handle more complicated cases .
we note that , with our chosen parameters , the emissivity levels in @xmath0-rays happen to be comparable for the leptonic and hadronic emissions ( see figure [ fig : spc - back - zeta ] ) .
however , all sorts of outcomes are possible , depending on the local conditions . for pion - decay ,
the emissivity scales directly with the ambient density , which can vary by several orders of magnitude in different regions of the galaxy ( or even different regions around a given snr ) , and with the density of accelerated protons , which depends on the ambient density as well as on the poorly constrained injection ratio at the shock . for inverse compton , as well as for synchrotron ,
the emissivity scales directly with the density of accelerated electrons , which was parametrized with the ratio @xmath200 of electrons over protons , which again is poorly constrained in the sources . in the end ,
the best way to distinguish the emissions is therefore through their spectral shape , as discussed in section [ sec : discussion - protons - electrons - emissions ] .
the impact of the ambient magnetic field is more complicated , as it impacts both acceleration ( of protons ) and losses ( of electrons ) , on top of controlling the synchrotron emissivity of electrons . in general
, a higher magnetic field presumably helps confining particles at the shock and reaching higher energies for protons , while it may limit the maximum energy of electrons through radiative losses . in the case with efficient mfa ( @xmath132 ) , the resulting field is completely dominated by the amplified field , so that the actual value of @xmath52 does not matter much .
any uncertainty regarding the value of @xmath52 is likely small compared to the huge difference between @xmath52 and @xmath107 . in the case with no net mfa ( @xmath72 ) , variations of @xmath52 can be seen on the spectra . in particular , as soon as @xmath52 is reduced by a factor of about two or if the compression at the shock is lower than was assumed with equation ( [ eq : b2_shock ] ) , in the case of a more ordered field
then the maximum energy of electrons may no longer be loss - limited , so that note [ note : loss - limited ] does not hold and the synchrotron cut - off energy becomes dependent on @xmath99 .
however , this case is probably not the most relevant one to observations , as discussed in section [ sec : discussion - mf - up ] . to study in more details the impact of ambient parameters , such as @xmath201 and @xmath52 , and of other free model parameters , such as @xmath14 and @xmath70
, @xcite conducted an extensive parametric study using our code presented in paper i. this resulted in the production of a library of models of snrs undergoing ( efficient ) dsa , that can be used either to assess the possible outcomes of acceleration when the ambient parameters are sufficiently known , or to constrain these ambient parameters when the acceleration efficiency could be measured . in the future
we will extend this library to include the most salient observational features discussed in paper ii and this paper .
in this paper we were only interested in the particles accelerated at the shock and subsequently trapped inside the shell .
however , protons reaching @xmath139 eventually escape the system upstream of the shock . the corresponding energy loss is included in the acceleration model , and their energy distribution can be computed @xcite , but their emission is not included in the simulations presented in this paper .
this would be especially interesting when dense targets are present in the immediate vicinity of the remnant , a situation already observed and modeled ( e.g. @xcite ) .
this situation is more likely for remnants of core - collapse supernovae , that are the final stage of massive stars , that were born inside molecular clouds .
particles accelerated at the blast wave may irradiate neighbouring clouds , providing another kind of diagnostic on their acceleration and diffusion mechanisms ( see @xcite and references therein ) .
we postpone this study to a dedicated paper .
we have developed a model that simulates the morphological and spectral evolution of a young type ia snr undergoing efficient particle acceleration .
starting from chevalier s profiles , the time - evolution of the shocked structure is computed using the 3d hydrodynamical code ramses , in a comoving grid of variable resolution .
the acceleration of particles at the forward shock is computed with blasi s non - linear semi - analytical model that provides the level of back - reaction on the fluid .
time - dependent shock modifications are implemented in the hydrodynamic code through the use of a variable adiabatic index , which is lowered at the shock front according to the contribution of the particle pressure to the total pressure .
we are thus effectively working with a pseudo - fluid , which represents both thermal and non - thermal populations . at the end of the hydrodynamic simulation , each cell located downstream of the forward shock
is processed to compute the evolution of the magnetic field , the spectra of remaining particles after losses , and their non - thermal emission : pion decay for protons , synchrotron and inverse compton for electrons .
finally the emission is projected along the line of sight , integrating variations due to the geometry and the history of the shocked ism .
this allows us to produce realistic synthetic maps of a typical remnant , in any energy band from radio to x - rays and @xmath0-rays .
a major improvement of our model , compared to our study of the thermal emission in paper ii , concerns the fate of the magnetic field .
we have implemented recipes for the amplification and transport of the magnetic field in the vicinity of the shock , that allows to probe the impact of cr - driven instabilities on their own spectra and on their emission spectra .
we have considered two limit cases for the evolution of the magnetic field in the shock precursor : either the amplified field is entirely damped in the plasma , so that the shock propagates through a pre - heated medium , or the amplified field is advected to the ( sub-)shock , which is then a magnetized shock .
interestingly , the net effect in terms of shock modifications is roughly the same for both models unless the speed of the waves that scatters particles is computed in the amplified field , in which case the second model leads to less modified shocks than the first . regardless of this complication induced by the alfvnic drift , when mfa is efficient , very high values of the magnetic field are obtained downstream of the shock ( @xmath202 g ) .
this markedly alters the spectra of electrons , especially around their maximum energy : because of the strong radiative losses , the most energetic ( @xmath203 tev ) electrons are confined close to the shock front . as a result , their emissions , at the cut - off of the synchrotron in x - rays , and at the cut - off of the inverse compton in @xmath0-rays , produce thin rims immediately behind the shock front , with no other structure visible inside the shell in contrast with leptonic emission in radio , and with hadronic emissions in @xmath0-rays , which show the imprint of the rayleigh - taylor instabilities .
the cut - off energy itself is significantly reduced for the inverse compton emission ( from almost 10 tev to about 1 tev ) , while it stays the same for the synchrotron emission ( around 10 kev ) as long as the electron s energy is loss - limited . on the other hand ,
an enhanced magnetic field can boost the maximum momentum of protons
although it still falls short from the energy of the knee in the local cr spectrum in all our simulations , with a maximum of about @xmath204 tev at the shock front . as a result , the cut - off energy of the pion decay emission is significantly increased ( from about 10 tev to about 100 tev ) . in conclusion
, mfa has a strong impact on the x - ray emission from electrons , and can help distinguish the @xmath0-ray emissions from electrons and protons , by altering both the radial distribution and the cut - off in this energy band .
an important goal of snr observations is to assess the presence of energetic protons at the shock , that are believed to be the constituents of galactic cosmic - rays . observing these protons
directly proved to be a difficult task , because they do not radiate as efficiently as electrons .
however , they can show up in two other ways : first they impact the dynamics of the shock wave ( see paper i ) , and therefore the thermal x - ray emission from the plasma ( see paper ii ) ; and second they impact the evolution of the magnetic field , and therefore the non - thermal emission from electrons in radio , x - rays and @xmath0-rays ( this paper ) .
this work completes our modeling of the emission from snrs , providing a framework for the study of particle acceleration from broadband observations of young shells .
next , we will apply this model to make a detailed modeling of some specific object such as tycho s snr or other type ia snrs . in parallel
, we shall extend our model to more complex environmental situations , such as a shock expanding inside a wind or interacting with a molecular cloud , of relevance for the case of a core - collapse snrs .
more generally , our work will help for the interpretation of x - ray and @xmath0-ray observations with current instruments ( such as the recently launched astro - h ) , as well as for the preparation of future missions ( such as the proposed athena+ ) .
this work has been partially funded by the accelrsn project anr-07-jcjc-0008 in france , and has been supported in canada by the natural sciences and engineering research council ( nserc ) through the canada research chairs program and by the canadian institute for theoretical astrophysics ( cita ) through the national fellows program .
the code used was largely developed on the nuit computing cluster at the university of manitoba , funded by the canada foundation for innovation ( cfi ) and the manitoba research innovation funds ( mrif ) .
the high - resolution simulations presented here were performed on the national computing facilities provided by westgrid ( western canada research grid ) .
we thank paul edmon for providing us with a version of the code aura used for computing the non - thermal emission .
we thank the anonymous referee for their useful comments that helped us to substantially improve the quality of our manuscript .
erg in a uniform ism of density @xmath205 ( see section [ sec : parameters - snr ] for details ) .
each plot covers 5.61 pc with a resolution of 1024 pixels in each direction .
the right plots ( on cases ) include the hydrodynamic back - reaction of accelerated particles .
the three rows show the effect of the magnetic back - reaction in different cases : with no net amplification ( @xmath72 and thus @xmath134 , top ) , with efficient amplification but still negligible alfvnic drift ( @xmath132 and @xmath134 , middle ) , and with efficient amplification and strong alfvnic drift ( @xmath132 and @xmath135 , bottom ) .
[ fig : map - np],title="fig : " ] erg in a uniform ism of density @xmath205 ( see section [ sec : parameters - snr ] for details ) .
each plot covers 5.61 pc with a resolution of 1024 pixels in each direction .
the right plots ( on cases ) include the hydrodynamic back - reaction of accelerated particles .
the three rows show the effect of the magnetic back - reaction in different cases : with no net amplification ( @xmath72 and thus @xmath134 , top ) , with efficient amplification but still negligible alfvnic drift ( @xmath132 and @xmath134 , middle ) , and with efficient amplification and strong alfvnic drift ( @xmath132 and @xmath135 , bottom ) .
[ fig : map - np],title="fig : " ] erg in a uniform ism of density @xmath205 ( see section [ sec : parameters - snr ] for details ) .
each plot covers 5.61 pc with a resolution of 1024 pixels in each direction .
the right plots ( on cases ) include the hydrodynamic back - reaction of accelerated particles .
the three rows show the effect of the magnetic back - reaction in different cases : with no net amplification ( @xmath72 and thus @xmath134 , top ) , with efficient amplification but still negligible alfvnic drift ( @xmath132 and @xmath134 , middle ) , and with efficient amplification and strong alfvnic drift ( @xmath132 and @xmath135 , bottom ) .
[ fig : map - np],title="fig : " ] , top panel ) .
no computation is performed in the shocked ejecta ( grey area ) .
[ fig : map - b_fs],title="fig : " ] , top panel ) .
no computation is performed in the shocked ejecta ( grey area ) .
[ fig : map - b_fs],title="fig : " ] , top panel ) .
no computation is performed in the shocked ejecta ( grey area ) .
[ fig : map - b_fs],title="fig : " ] ) .
model parameters are the same as for figure 1 .
note that a different scale is used for each row .
[ fig : map - tl_fs],title="fig : " ] ) .
model parameters are the same as for figure 1 .
note that a different scale is used for each row .
[ fig : map - tl_fs],title="fig : " ] ) .
model parameters are the same as for figure 1 .
note that a different scale is used for each row .
[ fig : map - tl_fs],title="fig : " ] , top panel ) . as only acceleration at the forward shock
is considered , no computation is performed in the shocked ejecta ( grey area ) . [
fig : map - plp],title="fig : " ] , top panel ) . as only acceleration at the forward shock is considered , no computation is performed in the shocked ejecta ( grey area ) .
[ fig : map - plp],title="fig : " ] , top panel ) . as only acceleration at the forward shock
is considered , no computation is performed in the shocked ejecta ( grey area ) . [
fig : map - plp],title="fig : " ] , top panel ) . as only acceleration at the forward shock
is considered , no computation is performed in the shocked ejecta ( grey area ) . [
fig : map - ple_fs],title="fig : " ] , top panel ) . as only acceleration at the forward shock
is considered , no computation is performed in the shocked ejecta ( grey area ) . [
fig : map - ple_fs],title="fig : " ] , top panel ) . as only acceleration at the forward shock
is considered , no computation is performed in the shocked ejecta ( grey area ) .
[ fig : map - ple_fs],title="fig : " ] @xmath123 ev radio band .
note that a different scale is used for the case with no net amplification of the magnetic field ( @xmath72 , top panel ) .
[ fig : map - ntsy - radio],title="fig : " ] @xmath123 ev radio band .
note that a different scale is used for the case with no net amplification of the magnetic field ( @xmath72 , top panel ) .
[ fig : map - ntsy - radio],title="fig : " ] @xmath123 ev radio band .
note that a different scale is used for the case with no net amplification of the magnetic field ( @xmath72 , top panel ) .
[ fig : map - ntsy - radio],title="fig : " ] -ray band .
note that a different scale is used for the case with no net amplification of the magnetic field ( @xmath72 , top panel ) .
[ fig : map - ntic - gsoft],title="fig : " ] -ray band .
note that a different scale is used for the case with no net amplification of the magnetic field ( @xmath72 , top panel ) .
[ fig : map - ntic - gsoft],title="fig : " ] -ray band .
note that a different scale is used for the case with no net amplification of the magnetic field ( @xmath72 , top panel ) .
[ fig : map - ntic - gsoft],title="fig : " ] -ray band .
note that a different scale is used for the case with no net amplification of the magnetic field ( @xmath72 , top panel ) .
[ fig : map - ntic - ghard],title="fig : " ] -ray band .
note that a different scale is used for the case with no net amplification of the magnetic field ( @xmath72 , top panel ) .
[ fig : map - ntic - ghard],title="fig : " ] -ray band .
note that a different scale is used for the case with no net amplification of the magnetic field ( @xmath72 , top panel ) .
[ fig : map - ntic - ghard],title="fig : " ] ) , while the bottom panel is with strong alfvnic drift ( @xmath135 ) .
each panel compares the cases with or without hydrodynamic back - reaction ( on vs. off ) , and with or without magnetic back - reaction ( @xmath132 vs. @xmath72 ) .
[ fig : spc - back - zeta],title="fig : " ] ) , while the bottom panel is with strong alfvnic drift ( @xmath135 ) .
each panel compares the cases with or without hydrodynamic back - reaction ( on vs. off ) , and with or without magnetic back - reaction ( @xmath132 vs. @xmath72 ) .
[ fig : spc - back - zeta],title="fig : " ] | supernova remnants are believed to be the major contributors to galactic cosmic rays . in this paper
, we explore how the non - thermal emission from young remnants can be used to probe the production of energetic particles at the shock ( both protons and electrons ) .
our model couples hydrodynamic simulations of a supernova remnant with a kinetic treatment of particle acceleration . we include two important back - reaction loops upstream of the shock : energetic particles can ( i ) modify the flow structure and ( ii ) amplify the magnetic field . as the latter process is not fully understood , we use different limit cases that encompass a wide range of possibilities . we follow the history of the shock dynamics and of the particle transport downstream of the shock , which allows us to compute the non - thermal emission from the remnant at any given age .
we do this in 3d , in order to generate projected maps that can be compared with observations .
we observe that completely different recipes for the magnetic field can lead to similar modifications of the shock structure , although to very different configurations of the field and particles .
we show how this affects the emission patterns in different energy bands , from radio to x - rays and @xmath0-rays .
high magnetic fields ( @xmath1 g ) directly impact the synchrotron emission from electrons , by restricting their emission to thin rims , and indirectly impact the inverse compton emission from electrons and also the pion decay emission from protons , mostly by shifting their cut - off energies to respectively lower and higher energies . |
the sunyaev - zeldovich ( s - z ) effect is a small change in the intensity of the cmb that is caused by compton scattering as the radiation passes through clusters of galaxies .
energy is transferred from the hot intracluster ( ic ) gas to the radiation and a fraction of the photons move from the rayleigh - jeans ( r - j ) to the wien side of the ( planck ) spectrum ; the radiation is said to be comptonized .
a quantitative description of the effect was given by zeldovich & sunyaev ( 1969 ) and sunyaev & zeldovich ( 1972 ) , who realized very early on the cosmological significance of this characteristic spectral signature .
first measurements of the effect in a small number of clusters were made long ago , but these were mostly low ( signal ) quality results .
s - z observations significantly improved when interferometric arrays equipped with high sensitivity receivers began to be used .
the current high - quality s - z database , consisting largely of @xmath0 ghz measurements with the bima and ovro interferometric arrays ( , carlstrom 1996 , grego 2000 , reese 2000 ) , now includes some 60 clusters . at present , the s - z cosmological results ( carlstrom 2001 ) are based almost entirely on these ( single frequency ) measurements of a sample of moderately distant clusters .
this brief review highlights recent theoretical and observational results .
the reviews by rephaeli ( 1995a ) and birkinshaw ( 1999 ) provide a more comprehensive discussion of most of the essential theoretical and observational aspects of the effect .
results from the extensive program of observations with interferometric arrays were reviewed recently by carlstrom ( 2001 ) .
accurate description of the change of cmb intensity due to interaction with fast moving electrons in clusters necessitates use of the exact frequency re - distribution function in compton scattering and a relativistic calculation .
the original description of the effect by sunyaev & zeldovich ( 1972 ) is based on a solution to the kompaneets equation , a diffusion approximation to the exact kinetic equation for the scattering .
the calculation yields a simple expression for the change of cmb intensity induced by scattering of the cmb by electrons with _
thermal _ velocity distribution i_t = i_o y g_0(x ) , where @xmath1 , and @xmath2 is the cmb temperature .
all the dependence on the properties of the cluster is in the comptonization parameter , y=(kt_e / mc^2 ) n _ t dl , an integral over the electron density ( @xmath3 ) and temperature ( @xmath4 ) ; @xmath5 is the thomson cross section .
the spectral distribution is given in terms of the non - dimensional frequency @xmath6 , g_0(x)=x^4e^x ( e^x -1)^2 ; @xmath7 is negative for @xmath8 and positive at larger values of this critical frequency , @xmath9 ghz .
the magnitude of the relative temperature change due to the thermal effect is @xmath10 in the r - j region , with @xmath11 along a line of sight ( los ) through the center of a rich cluster .
likely motion of the cluster in the cmb frame induces a _ kinematic _ ( doppler )
s - z component , i_k = - i_o h_0(x ) _ c , h_0(x ) = x^4e^x(e^x -1)^2 , where @xmath12 , with @xmath13 the line of sight component of the cluster peculiar velocity , and @xmath14 is the thomson optical depth of the cluster .
the associated temperature change of this component is @xmath15 ( sunyaev & zeldovich 1980 ) .
the above expressions for the two components of the s - z effect are sufficiently accurate only at low frequencies and low gas temperatures . an exact description of the effect is required for its use as a precise cosmological probe . since electron velocities in the ic gas
are high , the relative photon energy change in the scattering is large enough to require a calculation . using the exact probability distribution in compton scattering , and the relativistically correct form of the electron maxwellian velocity distribution , an improved calculation of @xmath16 was performed in the limit of small @xmath14 ( rephaeli 1995b ) .
results of this semi - analytic calculation demonstrate that the relativistic spectral distribution of the intensity change is quite different from that derived by sunyaev & zeldovich ( 1972 ) .
deviations from their expression increase with @xmath4 and can be substantial , as can be seen in figure 1 .
these are particularly large near the crossover frequency , which shifts to higher values with increasing gas temperature .
relativistic generalization of the sunyaev & zeldovich ( 1972 ) treatment has been discussed in many papers since 1995 .
for example , challinor & lasenby ( 1998 ) generalized the kompaneets equation and obtained analytic approximations to its solution by means of a power series in @xmath17 = @xmath18 .
this approach was adopted by itoh ( 1998 ) who improved the accuracy of the analytic approximation by expanding to fifth order in @xmath17 .
sazonov & sunyaev ( 1998 ) , and nozawa ( 1998a ) , extended the relativistic treatment also to the kinematic component obtaining the leading cross terms in the expression for the total intensity change ( @xmath19 ) which depends on both @xmath4 and @xmath13 .
an improved analytic fit to the numerical solution , valid for @xmath20 , and @xmath21 ( @xmath22 ghz ) , was given by nozawa ( 2000 ) .
in all these treatments only terms linear in @xmath14 were calculated . since in some rich clusters @xmath23 ,
the approximate analytic expansion to fifth order in @xmath17 necessitates also the inclusion of multiple scatterings , of order @xmath24 .
such terms were calculated by itoh ( 2000 ) , and shimon & rephaeli ( 2002 ) . the relativistically generalized expression for the total intensity change , @xmath25 ,
can be written ( to first order in @xmath14 ) in the form : i = i_o h_0(x ) d , where @xmath26 .
the first two terms are just the expressions for @xmath16 and @xmath27 , respectively , and the function @xmath28 includes the additional thermal and kinematic terms obtained in a calculation .
this function can be written in the form @xmath29 \\ & & - \beta_{c}\theta \biggr [ h_{1}(x ) + \theta h_{2}(x ) \biggl ] \mbox { } + \beta_{c}^{2}\biggr[1 + \theta h_{3}(x ) \biggl ] \nonumber \ , , \end{aligned}\ ] ] that includes terms to orders @xmath30 and @xmath31 .
more details on the calculation , and the definitions of the functions @xmath32 and @xmath33 ( @xmath34 and @xmath35 ) , can be found in shimon & rephaeli ( 2002 ) .
measurements with the interferometric bima and ovro arrays have been made at ( relatively ) low frequencies ( @xmath0 ghz , or @xmath36 ) for which differences between the and the above more exact expressions are small .
but measurements of a few clusters have also been made at much higher frequencies ( , up to @xmath37 in the case of the mito telescope , and even higher for some of the upcoming s - z projects ) .
use of the relativistically exact expressions for @xmath16 and @xmath27 is clearly necessary at high frequencies ( @xmath38 ghz ) .
this is especially so when the effect is used for the purpose of determining precise values of cosmological parameters .
also , since the ability to determine peculiar velocities of clusters depends very much on measurements very close to the crossover frequency whose dependence on @xmath4 is approximately given by @xmath39 ( shimon & rephaeli 2002 ) .
note that because high precision s - z work entails use of x - ray derived gas parameters , similarly accurate expressions for the x - ray bremsstrahlung emissivity have to be used ( rephaeli & yankovitch 1997 ) . in the latter paper
first order relativistic corrections to the velocity distribution and electron - electron bremsstrahlung were taken into account in correcting values of the hubble constant , @xmath40 , that were previously derived using the expression for the emissivity ( see also hughes & birkinshaw 1998 , nozawa 1998b ) .
we briefly note that there are various ways by which the scattering of the cmb in clusters can ( linearly ) polarize the radiation , generally at levels which are below 1 @xmath41k ( sazonov & sunyaev 1999 ) .
first , polarization is induced due to the cmb quadrupole component , at a maximal level of @xmath42 @xmath41k . when the cluster has a finite velocity component transverse to the line of sight , @xmath43
, there are two contributions to the polarization that are @xmath44 , and @xmath45 .
the radiation will also be polarized when scattered in a cluster with aspherical electron distribution .
the great property of the s - z effect which makes it a uniquely important cosmological probe is its ( essentially ) redshift independence .
measurements of the effect yield directly the integrated pressure of the hot ic gas , and thereby also the total mass of the cluster .
the cluster velocity along the los can be deduced from measurement of the effect close to the corossover frequency . from s - z and x - ray measurements ( the angular diameter distance , and therefore ) @xmath40 can be determined .
this method to measure @xmath40 has clear advantages over the traditional galactic distance ladder method .
it is also possible to determine the density parameter , @xmath46 , from the hubble diagram ( or from the redshift dependence of the gas mass fraction ) when a sufficiently precise database on a large sample of clusters is available .
the feasibility of detecting clusters at large redshifts strongly motivates performing number counts through cluster surveys in order to characterize the population and its cosmological evolution . also , mapping the cmb anisotropy induced by clusters can yield important information on the cluster mass function , cluster properties and evolutio of the population .
finally , the anisotropy and redshift evolution of the cmb temperature , @xmath47 , can be tested and verified through multi - frequency measurements of the effect in clusters at different sky directions and redshifts .
a brief summary of current results is given below .
sensitive ground based measurements of the thermal and kinematic components of the s - z effect with single dish telescopes face the major challenge of accounting for atmospheric emission .
the significance of astrophysical sources of confusion such as emission from galactic dust , cluster radio sources , and cmb anisotropy varies greatly with the frequency of observation and telescope beam size . over the last decade
observational s - z work has been carried out mostly with interferometric arrays .
these have major advantages over single dish telescopes , including sensitivity to specific angular scales and to signals which are correlated between array elements , insensitivity to changes in atmospheric emission , and high angular resolution that enables nearly optimal subtraction of signals from point sources .
the improved sensitivity of radio receivers made it feasible ( mainly through the use of low - noise hemt amplifiers ) to image the effect in moderately distant clusters , first with the ryle telescope ( jones 1993 ) , and then mostly with the bima and ovro arrays ( carlstrom 1996 , 2001 ) .
the extensive program of s - z observations with bima and ovro has resulted in high quality measurements ( at frequencies near @xmath0 ghz ) of about 60 clusters in the redshift range @xmath48 .
an example of the higher sensitivity and resolution of interferometric images over the traditional one dimensional drift scans , the bima image of a2163 ( carlstrom 2001 ) is shown in figure 2 above a profile of the effect in the same cluster from measurements with the small suzie array ( holzapfel 1997a ) .
the bima contour plot is superposed on a rosat x - ray image ( false color ) of the cluster ; as expected , the cluster x - ray size appears smaller than its s - z size .
the array configurations of bima and ovro telescope systems used in the ( above mentioned ) measurements are not suitable for s - z observations of nearby clusters .
a more optimal system for imaging the effect over larger angular scales is the cbi an interferometric array of 13 small ( 0.9 m ) dishes , with spatial resolution in the @xmath49 , operating in the 26 - 36 ghz spectral range .
work with the cbi began at the atacama desert ( chile ) , and has already resulted ( udomprasert 2001 ) in measurements of the effect in 9 clusters .
there are major advantages in measuring the effect at several high ( @xmath50 ghz ) frequencies ( close to and on the wien side ) where its characteristic spectral shape can be used as a powerful discriminant to control observational and systematic uncertainties .
multi - frequency measurements of the effect also enhance its scope as a cosmological probe ; an example for this will be shown in the next section .
high frequency observations were made with the suzie array , the pronaos and mito telescopes , and the diabolo bolometer .
three moderately distant clusters were measured with the suzie @xmath51 element array : a1689 & a2163 ( holzapfel 1997a , 1997b ; the s - z profile across a2163 is shown in figure 2 ) , and a1835 , which was observed at three spectral bands centered on 145 , 221 , 279 ghz ( mauskopf 2000 ) . work with suzie continued with measurements of at 6 more clusters ( s. church , private communication ) .
pronaos , a stratospheric 2 m telescope , measured the effect in a2163 at four broad spectral bands in the combined range of 285 - 1765 ghz ( lamarre 1998 ) .
this seems to have been the first detection of the effect by a balloon - borne experiment .
the mito 2.6 m telescope ( located in testa grigia in the italian alps ) originally had only a single photometer with a large @xmath52 beam and four high - frequency bands .
even so it was possible to measure the effect in the coma cluster ( de petris 2002 ) , in spite of the fact that the fluctuating atmospheric emission dominated the signals .
the sample of clusters in which the s - z effect was measured includes the distant ( z=0.45 ) cluster rxj 1347 with the largest deduced comptonization parameter , @xmath53 ( pointecouteau 1999 ) , one of 5 clusters observed with the diabolo bolometer used at the iram 30 m radio telescope .
the well understood nature of the s - z effect and its redshift independence have always provided strong motivation for using it as a cosmological probe .
availability of the high quality interferometric s - z data obtained with bima and ovro has significantly improved the scientific yield from results of s - z and x - ray measurements .
the following is only brief update on current results ; discussion of the basic methodologies can be found in the reviews by rephaeli ( 1995a ) , birkinshaw ( 1999 ) , and carlstrom ( 2001 ) .
density and temperature profiles of ic gas can be determined from x - ray and s - z measurements .
spatially resolved s - z measurements can , in principle , yield these distributions out to larger radii due to the linear dependence of @xmath16 on @xmath3 , as compared to the @xmath54 dependence of the thermal x - ray brightness profile .
the equation of hydrostatic equilibrium can then be used to find the cluster total mass , @xmath55 , and therefore also the gas mass fraction at a given radius , @xmath56 .
grego ( 2001 ) determined gas mass fractions in 18 clusters from interferometric bima and ovro measurements .
they assumed isothermal gas with the familiar density profile , @xmath57 .
the core radius , @xmath58 , and @xmath59 were determined from the s - z data ; only the temperature was taken from results of spectral x - ray measurements .
this analysis yielded the gas mass fraction in the region were the s - z data are sufficiently sensitive ( typically 1 ) . scaling relations from numerical simulations
were then used to extrapolate the gas mass fraction to a distance where the cluster mean density is 500 times the critical density ( at the appropriate redshift ) , since it is presumed that at this distance the cluster gas mass fraction well samples the universal baryon fraction ( evrard 1997 ) . in the currently popular open and flat , @xmath60-dominated cdm models , mean
values in the range @xmath61 ( where @xmath62 is the value of @xmath40 in units of @xmath63 km s@xmath64 mpc@xmath64 ) were obtained for the gas mass fraction ( carlstrom 2001 ) .
use of more realistic temperature profiles that can now be measured by xmm and _ chandra _ will reduce the substantial modeling uncertainties in these mass estimates . to measure cluster radial velocities from the kinematic component of the s - z effect observations have to be made in a narrow spectral band near the crossover frequency , where the thermal effect vanishes while the kinematic effect ( which is usually swamped by the much larger thermal component ) is maximal ( rephaeli & lahav 1991 )
this necessitates knowledge of the exact spectral shape of the thermal component near this frequency , and control of significant systematic uncertainties ( such as due to the primary cmb anisotropy ) .
suzie is the first experiment with a spectral band centered on the crossover frequency .
measurements of the clusters a1689 and a2163 ( holzapfel 1997b ) and a1835 ( mauskopf 2000 ) yielded substantially uncertain results for @xmath13 ( @xmath65 , @xmath66 , and @xmath67 km s@xmath64 , respectively ) .
balloon - borne measurements of the effect with pronaos have also resulted in a statistically insignificant value for the peculiar velocity of a2163 ( lamarre 1998 ) .
the hubble constant and the contributions to the cosmological density parameter that of matter , @xmath68 , and the cosmological constant , @xmath69 can be determined from a plot of the angular diameter distance , @xmath70 , vs. redshift ( hubble diagram ) .
the ability to infer @xmath70 from s - z and x - ray observations is essentially based on the different density dependences of comptonization and thermal bremsstrahlung .
specifically , @xmath70 can be deduced from @xmath16 , the x - ray surface brightness , and their spatial profiles .
it is ( at least currently ) unrealistic to obtain a usefully precise value of @xmath40 from measurements of any single cluster , due mostly to large systematic uncertainties .
modeling and other systematic errors can be significantly reduced by averaging over values of @xmath40 from measurements of a large number of clusters .
first , @xmath40 can be determined from a sufficiently large sample of nearby clusters ( whose angular diameter distances are little affected by the geometry of the universe ) . an initial sample of eight measured values yielded @xmath71 km s@xmath64 mpc@xmath64 ( rephaeli 1995b ) .
a similar mean value ( @xmath72 km s@xmath64 mpc@xmath64 ) was deduced by birkinshaw ( 1999 ) based on a somewhat updated data set .
the interferometric bima and ovro s - z survey provides the first relatively uniform dataset for the determination of @xmath40 .
since the set of 33 available values ( from single dish as well as interferometric measurements ) for @xmath70 includes measurements of clusters up to @xmath73 , the deduced mean value depends on the cosmological model : @xmath74 km s@xmath64 mpc@xmath64 for a flat model with @xmath75 and @xmath76 , and @xmath77 km s@xmath64 mpc@xmath64 for an open model with @xmath75 ( carlstrom 2001 ) . while this overall @xmath78 ( @xmath79 ) observational error is small , the estimated systematic uncertainty is @xmath80 . in principle , the dependence of the cluster gas mass fraction on the angular diameter distance provides a way to constrain the matter and cosmological constant density parameters . in a flat model , @xmath81 , the fit to the interferometric data yields @xmath82 ( carlstrom 2001 ) .
measurements with the cobe / firas experiment have shown that the cmb spectrum is a precise planckian with @xmath83k at the current epoch ( mather 1999 ) . in the standard cosmological model , @xmath84 , a fundamental relation which has not yet been observationally confirmed to the desired extent .
cosmological models with a purely blackbody spectrum but with a different @xmath47 dependence than in the standard model are formally , at least unconstrained by the firas measurements . also unconstrained
are models with spectral distortions that are now negligible , but may have been appreciable in the past .
thus far @xmath47 has been determined mainly from measurements of microwave transitions in interstellar clouds in which atoms and molecules are excited by the cmb ; for more on this see the review by losecco ( 2001 ) .
the temperature has been determined in the galaxy , as well as in clouds at redshifts up to @xmath85 .
results are , however , substantially uncertain due to the poorly known physical conditions in the absorbing clouds .
the use of the thermal s - z effect to measure @xmath47 has been suggested long ago .
fabbri , melchiorri & natale ( 1978 ) proposed spectral mapping of the effect near the crossover frequency whose value depends on @xmath47 .
this is very challenging given the dominating confusing signals such as due to the primary cmb anisotropy and the kinematic s - z component and the fact that in the treatment ( of the effect ) the crossover frequency depends on gas temperature , whose value has to be known very precisely if this method is to yield useful results for @xmath47 .
rephaeli ( 1980 ) proposed a more feasible method which exploits the steep frequency dependence of @xmath16 on the wien side , and the weak dependence of _ ratios _ of the intensity change at different frequencies on the properties of the cluster .
formally , in the limit such a ratio is completely independent of the comptonization parameter .
most of the dependence on the cluster parameters still drops out also in the exact description , but a weak dependence remains on the gas temperature .
when unknown , the cluster velocity introduces a small systematic uncertainty ( which is , however , much smaller than in the case of measurements near the crossover frequency ) .
thus , s - z measurements have the potential of yielding more precise values of @xmath47 than can be obtained from ratios of atomic and molecular lines . with the availability of spectral measurements of the s - z effect , the method of rephaeli ( 1980 ) has recently been employed to measure @xmath47 in the coma and a2163 clusters ( battistelli 2002 ) .
spectral measurements of coma ( @xmath86 ) included observations at central frequencies @xmath87 , @xmath88 , and @xmath89 ghz ( each with a @xmath0 ghz bandwidth ) , made with with the mito telescope ( de petris 2002 ) , and the low frequency @xmath90 ghz ( with a @xmath91 ghz bandwidth ) ovro measurement ( herbig 1995 ) . measurements of a2163 ( @xmath92 ) were made with the suzie array ( holzapfel 1997a ) at @xmath93 , @xmath94 , and @xmath95 ghz ( with @xmath96 ghz bandwidths ) , and a combined bima and ovro measurement at @xmath0 ghz ( laroque 2002 ) . the measurements yield three independent intensity ratios for each cluster ; all combinations of these ratios were compared to the theoretically predicted values .
the latter were calculated by performing integrations of the expression for @xmath16 over the spectral bands of the mito photometer , and of the other receivers when available .
fits of the measured ratios to the predicted values were performed , yielding best fit values for the cmb temperature at the redshifts of the two clusters .
( for more on the data analysis and estimates of the uncertainties , see battistelli 2002 . )
the results for the temperatures are @xmath97 k and @xmath98 k ( at 68% confidence ) .
these values are consistent with those expected from the standard relation @xmath99 .
battistelli ( 2002 ) have also tested two alternative scaling relations that are conjectured in non - standard cosmologies , @xmath100 , and @xmath101 $ ] ( e.g. , lima et al .
they determined the best fit values for the two parameters to be @xmath102 , and @xmath103 ( at 95% confidence ) . within the large uncertainties
these values are consistent with zero , so no significant deviation is found from the standard model .
losecco ( 2001 ) obtained @xmath104 and @xmath105 ( at 95% cl ) from measurements of microwave transitions .
the two sets of results are consistent .
thus , the s - z results of battistelli ( 2002 ) already provide the same level of precision even though the two clusters are at much lower redshifts than the galaxies in the observational sample used by losecco ( 2001 ) . with more precise multi - spectral s - z measurements expected in the future
, it is anticipated that the s - z method will provide a preferred alternative to the atomic and molecular lines method .
the anisotropy in the cmb induced by clusters has been extensively explored ( since it was first modeled [ rephaeli 1981 ] in the context of a simple model for ic gas evolution ) , reflecting increased realization of its significance on arcminute scales .
since this anisotropy arises from the scattering of the cmb in the evolving population of clusters , its power spectrum and cluster ( s - z ) number counts can potentially yield important information on the properties of ic gas , the cluster mass function , cosmological evolution of clusters and their gaseous contents , as well as some of the global cosmological and large scale parameters .
clearly , therefore , a quantitative description of this anisotropy entails the added need ( when compared with a calculation of the primary cmb anisotropy ) of modeling gas properties across the evolving population of clusters . the usual approach to the calculation of the s - z anisotropy is based on the press - schechter model for the cluster mass function , @xmath106 , the comoving density of clusters of mass m at redshift z. following collapse and virialization , ic gas is presumed to have reached hydrostatic equilibrium at the virial temperature , with a density distribution that is commonly assumed to have an isothermal @xmath59 profile .
the mass function is normalized by specifying the mass variance on a scale of @xmath107 , @xmath108 , a parameter that is determined from the observed x - ray temperature function by using a mass - temperature calibration .
the calibration is limited to clusters at small redshifts .
the cluster induced anisotropy has been studied at an increasingly greater degree of sophistication and detail and in wide range of cosmological and dark matter models beginning about a decade ago ( , makino & suto 1993 , bartlett & silk 1994 ) . in particular ,
colafrancesco ( 1994 ) calculated the temperature anisotropy in a flat cdm model including gas evolution ( based on results from the einstein medium sensitivity survey [ emss ] ) .
they later extended the work to other cosmological models and estimated also that many thousands of clusters are expected to be detected during the planned planck survey ( colafrancesco 1997 ) .
s - z maps and power spectra can also be generated directly from hydrodynamical simulations ( , da silva 2000 ) .
the range of cosmological models was extended to include currently viable @xmath60cdm models ( , komatsu & kitayama 1999 , molnar & birkinshaw 2000 , cooray 2000 ) .
reflecting the parameter intensive nature of the s - z anisotropy ( with the added complexity due to the need to describe the dynamical as well as hydrodynamical evolution of clusters ) , results for predicted power spectra and number counts span a wide range even when calculated for the same cosmological model . aside from obvious reasons for some of the differences , such as differing choices of gas parameters and degree of the evolution of the gas mass fraction , results differ also due to inconsistent choice of parameters whose values are coupled .
an example is a conflicting choice of the adopted value of @xmath108 , and a different mass - temperature relation than assumed in order to determine @xmath108 itself ( for more on this , see sadeh & rephaeli 2002 ) .
the main features of the power spectrum of the thermal component of the s - z anisotropy are shown in figure 3 ; results are from the work of sadeh & rephaeli ( 2002 ) who have calculated power spectra of the s - z and primary anisotropies in an array of cosmological and dark matter models .
the s - z power spectrum , based on a press & schechter mass function , was normalized by the observed x - ray luminosity function .
the primary power spectrum ( dashed line ) was calculated using the cmbfast code of seljak & zaldarriaga ( 1996 ) .
the plots are of the predicted power spectrum , @xmath109 vs. multipole @xmath110 , in a flat cosmological model with @xmath76 , and cdm density parameter @xmath75 .
ic gas was assumed to evolve in a simple manner consistent with the results of the emss survey , as parametrized by colafrancesco ( 1994 ) . in this model
the s - z anisotropy ( which is dominated by the thermal effect ; the contribution of the kinematic effect is of second order ) is appreciable already at @xmath111 . clearly , the s - z anisotropy has to be accounted for in the detailed modeling of the small scale structure of the cmb , if parameter extraction is to be precise
. the first observational evidence for s - z anisotropy could possibly be the recent cbi measurement of excess power ( at an estimated @xmath112 level ) over what is predicted in standard models for the primary anisotropy at multipoles @xmath113@xmath114 ( mason 2002 , bond 2002 ) .
the quality of the scientific yield from the many s - z images obtained over the last few years with interferometric arrays has made quite apparent the great potential of the s - z effect as a major cosmological probe .
this is just the beginning : numerous s - z projects will become operational in the near future , both bolometric multi - frequency arrays , and interferometric arrays ; these include the ground - based ( upgraded ) mito telescope , the sza , amiba , and the balloon - borne telescopes boost and olimpo
. improved sensitivity , higher spatial resolution , and expanded spectral range will greatly improve the quality of s - z measurements .
better understanding and control of systematics will continue to determine the overall value of cosmological results from s - z and x - ray measurements .
beacause of this , optimal results will likely be obtained from measurements of the effect in nearby ( @xmath115 ) clusters .
for example , higher quality s - z and x - ray measurements of a sufficiently large sample of nearby clusters are expected to result in an overall uncertainty of @xmath78 in the value of @xmath40 . | extensive recent work on the sunyaev - zeldovich ( s - z ) effect reflects major progress in observational capabilities of interferometric arrays , the improved quality of multi - frequency measurements with upcoming ground - based and stratospheric bolometer arrays , and the intense theoretical and experimental work on the small scale structure of the cosmic microwave background ( cmb ) radiation .
i briefly describe the effect and discuss its significance as a major cosmological probe .
recent results for the gas mass fraction in clusters and the hubble constant ( largely from measurements with the bima and ovro interferometric arrays ) are discussed .
also reviewed are results from the first determination of the cmb temperature at the redshifts of two clusters ( from measurements with the mito and suzie experiments ) , and recent work on the cmb anisotropy due to the s - z effect . |
the gutzwiller variational method has been used since long for the study of ground state properties of hubbard - type models for correlated fermions .
@xcite the one - band ( one orbital ) formulation has been extended in various ways to treat , e.g. antiferromagnetism @xcite or to include additional , uncorrelated orbitals .
@xcite the gutzwiller ansatz has also been investigated analytically .
@xcite however , attempts to extend the method to @xmath1 correlated orbitals per site have , sofar , led to unsatisfactory results .
@xcite + in this paper we generalize the gutzwiller method to cases of arbitrary numbers @xmath2 of correlated orbitals per site .
the extension makes it possible to study the ground state properties of realistic models for magnetic 3d transition metals and compounds . as a first application
, we examine a two - band model with more general on - site interactions and focus on the metal - insulator transition @xcite at quarter and half band fillings .
for one band , the gutzwiller variational wavefunction reads as @xmath3 \left| \psi _ 0\right\rangle \ ; , \label{1}\ ] ] where @xmath4 is an uncorrelated wavefunction on a lattice of @xmath5 sites and the operators @xmath6 measure the double occupancy of sites @xmath7 .
the variational parameter @xmath8 will reduce the total number @xmath9 of doubly occupied sites and may vary between @xmath10 ( @xmath11 ) and @xmath12 ( @xmath13 ) . here
@xmath14 are the gross numbers of spin up(down ) sites .
the norm of ( [ 1 ] ) is approximated by @xcite @xmath15 with the probabilities @xmath16 and the combinatorial weight factors @xmath17 here we have introduced net numbers of occupied sites @xmath18 and of empty sites @xmath19 . in the thermodynamic limit the sum of eq .
( [ 4 ] ) is replaced by its largest term , leading to the relation @xmath20 with @xmath21 , etc .
also the expectation values @xmath22 are found by combinatorics . here
the `` loss factors '' @xmath23 ^2 \label{12}\ ] ] with @xmath24 and @xmath25 represent the step sizes of the momentum distribution functions at the fermi level ( @xmath26 for @xmath27 ) .
when we extend the gutzwiller method to cases of @xmath0 correlated orbitals per site , the number of different site occupancies @xmath28 will increase enormously . for @xmath1 and using the notation @xmath29
@xmath30 @xmath31 @xmath32 the @xmath33 occupancies are : i ) empty ( @xmath34 ) ; ii ) four single ( @xmath35 ) ; iii ) six double ( @xmath36 ; iv ) four triple ( @xmath37 and v ) quadruple ( @xmath38 ) . in general , the @xmath39 occupancies are labeled by subscripts @xmath40 , which are assigned to the multiple orbital - spin subscripts by @xmath41 the net and gross occupancies are measured by appropriate operators , e.g. , @xmath42 there are now @xmath43 multiple occupancies so that the generalized gutzwiller operator reads as @xmath44 note , that in contrast to ref . , we found it essential to include multiple occupancies and to use net multiple occupancy operators .
the norm is given by @xmath45 with the @xmath46 factors of eq .
( [ 4 ] ) , and @xmath47 ^{-1 } \label{17}\ ] ] the sum of eq . ( [ 16 ] ) includes all sets of multiple occupancy configurations @xmath48 with @xmath49 . replacing the sum by its largest term leads to the @xmath50 relations @xmath51 keeping the gross numbers @xmath52 fixed , we get @xmath53 where @xmath54 is the number of subscripts in @xmath55 . for each of the generalized `` loss factors '' @xmath56 the evaluation of @xmath57 terms is required . nevertheless , the resulting equations for the @xmath58 factors look remarkably simple : [ 25 ] @xmath59 ^ 2 \label{25a } \\
q_{kl}^2 & = & q_{kk}q_{ll } \label{25b}\end{aligned}\ ] ] here the summation primes exclude @xmath60 , etc . obviously eqs .
( [ 25 ] ) are a straightforward generalization of eq .
( [ 12 ] ) and of results obtained in ref . .
simpler forms of the @xmath61 equations are obtained , when certain multiple occupancies are put zero , as was assumed in ref . .
recently , eqs .
( [ 25 ] ) have also been found by t. okabe .
@xcite using the generalized gutzwiller wavefunction , we can now investigate extensions of the hubbard model for arbitrary numbers of orbitals @xmath62 : @xmath63 which include , apart from general hopping terms , the on - site interaction @xmath64 , both orbital - diagonal @xmath65 and off - diagonal @xmath66 , and the on - site or hund s rule exchange @xmath67 .
the primes exclude non - physical orbital - spin combinations in the respective @xmath64 and @xmath68 summations . for the expectation values of the on - site exchange only the terms
@xmath69 survive , i.e. , all spin fluctuations are ignored . by assuming @xmath70 , it is possible to achieve the filling of incomplete @xmath71 shells in the atomic limit according to hund s rules .
thus , the hamiltonian of eq .
( [ 27 ] ) represents realistic models for the valence electron structure of both 3d transition metal elements and of magnetic 3d oxides and halides , provided that a sufficient number of orbitals and sites are included .
the @xmath72particle wavefunctions @xmath73 can be chosen to represent para- , ferro- , or antiferromagnetic states , and in principle , even more complicated cases , such as orbital ordering .
also , ligand orbitals or @xmath74 type orbitals on the @xmath71 sites may be incorporated . however , any additional complexity of the wavefunctions @xmath4 leads to further variational parameters in @xmath4 representing effective orbital energy shifts or effective hoppings , etc .
@xcite the problem of orbital density depletion caused by the reduction of multiple occupancies can be treated in analogy to ref . .
in the following , we investigate the @xmath1 hamiltonian , i.e. , we assume one atomic site on a simple cubic lattice with two , degenerate @xmath75 orbitals , and we study only paramagnetic solutions .
the single particle bands are constructed using realistic @xmath76 and @xmath77 hopping matrix elements @xmath78 @xmath79 , and @xmath80 .
there are three interaction parameters @xmath81 and @xmath82 , which , in the limit of vanishing configuration interaction are related by @xcite @xmath83 for the paramagnetic case we have @xmath84 @xmath85and @xmath86 .
all @xmath56 are equal @xmath87 ^ 2\ , .
\label{30}\end{aligned}\ ] ] the energy function for @xmath88 ( number of electrons per atom ) is @xmath89 here @xmath90 is the kinetic energy of the uncorrelated case@xmath91 @xmath92 has been minimized numerically with respect to the five variational parameters @xmath93 and has been studied at several @xmath94 values and for various sets of interaction parameters @xmath64 and @xmath95 , keeping @xmath96 according to eq . ( [ 29b ] ) .
there are two @xmath94 regions of interest .
the first is around @xmath97 ( or @xmath98 ) , the other around @xmath99 . for these integer fillings
@xmath100 we observe metal - insulator ( mi ) transitions in the @xmath101plane .
however , there are big differences between the two cases , mainly concerning the shapes of the phase diagrams in the @xmath101plane and the order of the mi transitions . for @xmath97
, the mi phase diagram is shown in fig .
[ f1]a . for @xmath102 , we find @xmath103 in agreement with ref .. for @xmath104 @xmath105 the system will avoid all multiple occupancies in the limit @xmath106 , so that there is always a mi transition .
this transition is of second order in @xmath107 , very similar to that of the one - band model .
@xcite near any critical values
@xmath108 , @xmath107 behaves like @xmath109 , when we approach @xmath110 from the metallic side , i.e. ; the effective mass @xmath111 diverges with @xmath112 .
@xcite for @xmath113 @xmath114 the system is always metallic .
in contrast , the mi transition for @xmath99 is very different ( see fig . [ f1]a ) .
the transition is generally of first order in @xmath107 .
this means that the @xmath107 values change discontinuously at critical values ( @xmath115 from a metal ( typically @xmath116 ) to an insulator ( @xmath117 ) ( see fig .
[ f1]b ) . only at the point @xmath118i.e . @xmath119
the mi transition is second order in @xmath107 .
note that apart from the point @xmath120 the ground state energy @xmath92 varies at the transition in such a way that its first derivatives with respect to all relevant system parameters changes discontinuously .
in particular we observe a finite discontinuity of the chemical potential @xmath121 at @xmath99 for all @xmath122 if @xmath64 exceed @xmath110 by any infinitesimal amount .
in the insulating region we have @xmath123 , and because of particle hole symmetry , @xmath124 for @xmath125 only the spin triplet occupancy @xmath126 prevails : @xmath127 .
the case @xmath120 is special , as there is no preference for one of the three double occupancies .
we believe , because of this kind of frustration the metallic phase extends so far out near @xmath128 . in the metallic regime
all multiple occupancies are nonzero ( see fig .
[ f2 ] ) . for @xmath125 ,
@xmath126 increases with @xmath64 , typically up to @xmath129 .
then , at @xmath110 , @xmath126 jumps up to @xmath130 , representing a low - to - high - spin transition . for certain band filling values close to @xmath99 ,
the first order step is still present in the @xmath131 and @xmath132 curves .
further away from @xmath99 , the curves smoothen , yet still vary rapidly in the range of critical parameters @xmath133 as shown in fig .
[ f2 ] . a typical phase diagram in the @xmath101plane of this `` good - to - bad''-metal or low - to - high - spin transition is shown in fig .
[ f3 ] . note that for our evaluation a homogenous phase was assumed .
however , in the vicinity of the low - to - high - spin transition , the homogeneous phase is unstable against phase separation . as a consequence there exists a surface @xmath134 in the space of @xmath135 which divides the homogeneous from the separated region ( see fig . [ f3 ] ) . at any point @xmath136 with @xmath137 ( or @xmath138 ,
the system breaks up into one part with density @xmath139 and another one with @xmath140 .
it is an open question , however , whether the phase separation is an artefact of hubbard - type models , because of , e.g. , the neglect of long range coulomb interaction .
the first order behavior of the mi transition for @xmath99 is caused by the competition of two different energy minima , now possible due to of the much larger space of variational parameters .
we expect that the first order nature will survive when we include antiferromagnetic states in the minimization .
first order , low - to - high spin transitions are also observed for two - band , itinerant ferromagnetic states .
@xcite in fact , we suspect that first order changes in the electronic structure of incomplete 3d shells may drive many phase transitions of magnetic 3d compounds .
it should be interesting to reexamine the mi phase transitions observed e.g. for @xmath141 or @xmath142@xcite with respect to their first order behavior .
very recently , a dynamical mean field study about the mi transition in a degenerate ( n=2 ) hubbard model has been presented .
@xcite the authors only consider the cases @xmath143 and/or @xmath144 ( in our notation ) . in aggreement with our results
, they find a qualitative feature , which is very similar to the @xmath145 case .
nevertheless , their more general remark that the nature of the mott transition is not qualitatively changed by orbital degeneracy is not supported by our results , concerning the cases @xmath146 .
in conclusion , we have extended the gutzwiller method to arbitrary numbers @xmath147 of correlated orbitals per site and have reported studies on the simplest @xmath1 model .
we find a wealth of new features , not present in the @xmath148 models , such as first order metal - insulator transitions or low - to - high spin transitions .
numerical studies appear to be feasible even for @xmath149 realistic models of 3d transition metals and compounds . | we have generalized the gutzwiller method to the cases of @xmath0 correlated bands and report studies on a degenerate two - band model with hund s rule type on - site interactions . at half band filling the metal - insulator transitions are usually of first order . |
a charged particle confined to the helical orbit is a source of high intensity electromagnetic radiation over a broad range of wavelengths with a number of remarkable properties such as high collimation and high degree of polarization ( see , for instance refs .
these properties have resulted in extensive applications of this radiation in a wide variety of experiments and in many disciplines . in particular , the helical motion of an electron beam is employed in helical undulators to produce circularly polarized radiation in narrow angular cone in the forward direction @xcite . in the proposal of ref . bala79 the helical undulator radiation was used to generate a polarized positron beam .
synchrotron radiation from relativistic electrons spiralling in magnetic fields is the main mechanism to explain the emissions of many objects in radio astronomy ( see @xcite and references therein ) .
most of the works on the radiation from the helical trajectory refer to radiation in free space .
it is well known that the presence of medium can essentially change the characteristics of the electromagnetic processes and gives rise to new types of phenomena such as the cherenkov , transition , and diffraction radiations . in particular , the operation of a number of devices assigned to production of electromagnetic radiation is based on the interaction of charged particles with materials ( see , for example , @xcite ) .
the synchrotron radiation from a charged particle circulating in a homogeneous medium was considered in ref .
@xcite . in this paper
it was shown that the interference between the synchrotron and cherenkov radiations leads to interesting effects .
new interesting features arise in inhomogeneous media .
in particular , the interfaces of media can be used to control the radiation flow emitted by various systems . in a series of papers started in refs .
@xcite , we have considered the simplest geometries of boundaries between two dielectrics with spherical and cylindrical symmetries . the synchrotron radiation from a charge rotating around a dielectric ball enclosed by a homogeneous medium is investigated in refs .
it was shown that if for the material of the ball and the particle velocity the cherenkov condition is satisfied , strong narrow peaks appear in the radiation intensity . at these peaks the radiated energy exceeds the corresponding quantity in a homogeneous medium by several orders of magnitude .
a similar problem with the cylindrical symmetry has been discussed in refs .
grigoryan1995,kot2000,kot2001,kotnimb . in ref .
@xcite we have developed a recurrent scheme for constructing the green function of the electromagnetic field for a medium consisting of an arbitrary number of coaxial cylindrical layers .
the investigation of the radiation from a charged particle circulating around a dielectric cylinder immersed in a homogeneous medium , has shown that under the cherenkov condition for the material of the cylinder and the velocity of the particle , there are narrow peaks in the angular distribution of the number of quanta emitted into the exterior space .
for some values of the parameters the density of the number of quanta in these peaks exceeds the corresponding quantity for the radiation in vacuum by several orders .
the radiation by a longitudinal charged oscillator moving with a constant drift velocity along the axis of a dielectric cylinder immersed in a homogeneous medium is investigated in refs .
@xcite . as in the case of the circular motion
it was shown that the presence of the cylinder provides a possibility for an essential enhancement of the radiation intensity .
the properties of the radiation from a charged particle moving along a helical orbit in homogeneous dispersive medium are investigated in @xcite . the corresponding problem for the charge moving in vacuum has been widely discussed in literature ( see , e.g. , refs . @xcite and references given therein ) . the electromagnetic field and the radiation in the case of particle following the helical path inside a dielectric cylinder immersed into a homogeneous medium are studied in refs .
recently the influence of a homogeneous transparent medium on the radiation of relativistic particles in planar undulators is considered in ref .
@xcite . in the present paper ,
we study the electromagnetic field and the radiation intensity for a charge moving in a helical orbit inside a circular waveguide with dielectric filling .
note that the radiation parts of the fields in the case of vacuum inside the waveguide are investigated in ref .
the plan of the paper is as follows . in section [ sec : fields ] we derive expressions for the electric and magnetic fields by making use of the corresponding formulae from @xcite for the geometry of a dielectric cylinder immersed into a homogeneous medium .
analytic properties of the corresponding fourier components are investigated . in section sec
: intensity the radiation fields are separated and they are presented as a superposition of the waveguide eigenmodes .
the formulae are derived for the radiation intensity of te and tm waves and numerical examples are presented .
section [ sec : conc ] concludes the main results of the paper .
consider a point charge @xmath0 moving along the helical trajectory of radius @xmath1 inside a circular waveguide with conducting walls .
we will denote by @xmath2 the radius of the waveguide and will assume that it is filled by homogeneous dielectric with permittivity @xmath3 .
the particle velocities along the axis of the waveguide ( drift velocity ) and in the perpendicular plane we denote by @xmath4 and @xmath5 , respectively . in a properly chosen cylindrical coordinate system ( @xmath6 )
the corresponding motion is described by the coordinates @xmath7where the @xmath8-axis coincides with the waveguide axis and @xmath9 is the angular velocity of the charge .
this type of motion can be produced by a uniform constant magnetic field directed along the axis of a cylinder , by a circularly polarized plane wave , or by a spatially periodic transverse magnetic field of constant absolute value and a direction that rotates as a function of the coordinate @xmath8 . in the helical undulators
the last geometry is used . in accordance with the symmetry of the problem , we present the electric and magnetic fields in the form of the fourier expansion@xmath10where @xmath11 , and in the discussion below @xmath12 for the electric and magnetic fields , respectively .
the expressions for the fourier transforms @xmath13 are obtained from the corresponding formulae derived in ref .
@xcite for the geometry of a dielectric cylinder with permittivity @xmath3 immersed into a medium with permittivity @xmath14 , taking the limit @xmath15 . as a result of this limiting procedure ,
the fourier transforms are written in the decomposed form @xmath16where the part @xmath17 corresponds to the fields generated by the charge in a homogeneous medium with permittivity @xmath18 and the part @xmath19 is induced by the presence of the waveguide . in the case @xmath20 for the homogeneous part of the magnetic field one has@xmath21with the coefficients@xmath22 .
\label{dm0p}\]]in these expressions @xmath23 , @xmath24 , @xmath25 is the bessel function , @xmath26 is the hankel function of the first kind , and @xmath27the corresponding expressions for @xmath20 are obtained from ( hml0 ) , ( [ hmz0 ] ) by the replacements @xmath28 .
the part @xmath29 induced by the presence of the waveguide is given by the formulae @xmath30where we have introduced the notation@xmath31 , \label{dmpw}\end{aligned}\]]and the prime stands for the derivative with respect to the argument of the function . the part in the electric field due to the presence of the waveguide is given by the formulae @xmath32 , \label{electricb } \\
e_{mz}^{(1 ) } & = & \frac{qic\lambda _ { 0}k_{z}}{2\pi \omega _ { m}(k_{z})\varepsilon _ { 0}}\sum_{p=\pm 1}d_{m}^{(p)}j_{m}(\lambda _ { 0}\rho ) , \label{electricc}\end{aligned}\]]where @xmath33 . the corresponding formulae for the part @xmath34 are obtained from ( [ electricb ] ) , ( [ electricc ] ) by the replacement @xmath35 .
note that in the limit @xmath36 one has @xmath37 and the fields vanish .
we could expect this result , as when the charge is on the surface of the waveguide the charge and its image compensate each other .
the formulae given above describe the total electromagnetic field inside the waveguide . to separate the parts corresponding to the radiation we need the analytic properties of the fourier components as functions on @xmath38 . from formula ( [ dmpw ] )
it follows that the function @xmath39 has singular points corresponding to the zeros of the function @xmath40 and its derivative .
note that in the second and third summands on the right of formula ( [ dmpw ] ) the singularities at the zeros of the functions @xmath41 cancel out and the function @xmath42 is analytic at these points .
we denote by @xmath43 , @xmath44 , the @xmath45th positive zero of the bessel function ( @xmath46 ) and its derivative ( @xmath47 ) : @xmath48where @xmath49 .
these zeros describe the eigenmodes for the cylindrical waveguide and are known as tm modes for the case @xmath50 and as te modes for the case @xmath47 jackson .
the corresponding modes for the projection of the wave vector on the cylinder axis are determined from the relation @xmath51 by taking into account expressions ( [ lambdaj]):@xmath52this equation has real solutions under the condition@xmath53and these solutions have the form@xmath54 .
\label{kmnpm}\]]the values @xmath55 correspond to the simple poles of the functions @xmath56 and @xmath57 .
note that if the cherenkov condition for the velocity of the particle along the axis of the waveguide is satisfied , @xmath58 , then inequality ( solcond ) is valid for all values @xmath44 . in the case
@xmath59 , condition ( [ solcond ] ) determines the maximal value for @xmath45 , which we will denote by @xmath60:@xmath61the value of the function @xmath62 at the points @xmath63 is equal to@xmath64 .
\label{ommnsigw}\]]for @xmath65 , equation ( [ eqforkz ] ) has real solutions only when @xmath66 and @xmath67 for real values @xmath3 and under the condition ( [ solcond ] ) the poles are situated on the real axis of the complex plane @xmath38 . in formula ( [ fourfields ] ) it is necessary to give the way by which these poles should be circled .
for this we note that in physical situations the dielectric permittivity is a complex quantity , @xmath68 , and the imaginary part @xmath69 determines the absorbtion in the medium . under the condition @xmath70 , for the imaginary part of @xmath55 from ( [ kmnpm ] ) one has@xmath71where @xmath72 for @xmath73 .
it can be also seen that @xmath74 .
if the cherenkov condition is not satisfied , @xmath75 , from the formulae given above it follows that @xmath76 in this case the poles @xmath77 ( @xmath78 ) are situated in the upper ( lower ) half of the complex plane @xmath38 . in the limit @xmath79 ,
deforming the integration contour we obtain the rule for avoiding the poles plotted in figure [ fig1 ] . in the case
@xmath58 the corresponding inequalities have the form@xmath80 in the way similar to that for the previous case , deforming the contour for the integration over @xmath38 , we obtain the avoidance rule for the poles given in figure [ fig2 ] .
in this section we consider the radiation field propagating inside a cylindrical waveguide with perfectly conducting walls for large distances from the charge .
first of all let us show that the part corresponding to the first term on the right of formula ( [ fml ] ) does not contribute to the radiation field .
this directly follows from the estimate of the integral over @xmath38 on the base of the stationary phase method . as in the integral over @xmath38
the phase @xmath81 has no stationary points , for large values @xmath82 , under the condition that the preexponential function belongs to the class @xmath83 .
it follows from here that the radiation field is determined by the singularity points of the preexponential function .
as it has been mentioned before , for the integral over @xmath38 the only singular points are the poles of the function @xmath39 at the points @xmath84 , determined by relations ( [ kmnpm ] ) . to find the corresponding contribution , we note that the integration contour over @xmath38 has the form depicted in figure [ fig1 ] for @xmath85 and the form depicted in figure [ fig2 ] for the case @xmath58 . at large distances from the charge
the integration contour can be closed by large semicircle in the upper ( lower ) half - plane for @xmath86 ( @xmath87 ) .
this choice is caused by the fact that the integrand exponentially vanishes in the upper ( lower ) half - plane . as a result when the radius of the large semicircle goes to infinity the corresponding integral vanishes .
hence , for large values @xmath88 the integral over @xmath38 , by residue theorem , is equal to the sum of residues inside the contour multiplied by @xmath89 . for @xmath90 and large @xmath45 , when @xmath91 @xmath92 , the poles @xmath63 have a finite imaginary part and the corresponding contribution exponentially vanishes in the limit @xmath93 . as a result
these poles do not contribute to the radiation field . as before we consider two cases .
when @xmath85 we close the integration contour in figure [ fig1 ] by the semicircle of large radius in the upper half - plane for @xmath86 and in the lower half - plane for @xmath94 . as a result for the radiation field
one finds @xmath95 , \label{radfieldw}\]]where @xmath96 @xmath97(@xmath98 ) corresponds to the case @xmath99 ( @xmath94 ) and we use the notation@xmath100each term in the sum on the right of formula ( [ radfieldw ] ) describes waves with the frequency @xmath101 propagating along the positive direction of the axis @xmath8 for @xmath96 and for @xmath98 , @xmath102 , and waves propagating along the negative direction of the axis @xmath8 for @xmath98 , @xmath103 . if the cerenkov condition is satisfied , @xmath66 , then closing the integration contour in figure [ fig2 ] by large semicircle in the upper or lower half - plane in dependence of the sign for @xmath104 , for the vector potential of the radiation field one finds@xmath105 , \label{radfieldw2}\]]where @xmath106 is the heaviside unit step function .
separate terms in the sum in formula ( [ radfieldw2 ] ) describe waves with the frequency latexmath:[$% direction of the axis @xmath8 for @xmath96 and for @xmath98 , @xmath108 , and along the negative direction of the axis @xmath8 for @xmath98 , @xmath103 .
note that for @xmath108 we have no waves propagating along the negative direction of the axis @xmath8 . by taking into account the formulae for the fourier components of the fields and evaluating the residues by the standard formulae of the complex analysis
, we find the following expressions for the radiation parts of the @xmath8-components of the fields:@xmath109 , \label{ez0 } \\ h_{mz}^{(1,\alpha ) } ( \mathbf{r},t ) & = & -\alpha \frac{4iqv_{\perp } j_{m , n}^{(1)3}j_{m}^{\prime } ( j_{m , n}^{(1)}\rho _ { 0}/\rho _ { 1})}{\sqrt{% \varepsilon _ { 0}}\rho _ { 1}^{3}(j_{m , n}^{(1)}-m^{2})j_{m}^{2}(j_{m , n}^{(1)})}% j_{m}(j_{m , n}^{(1)}\rho /\rho _ { 1 } ) \notag \\ & & \frac{\exp [ i(m\varphi + k_{m , n}^{(1,\alpha ) } z-\omega _ { m , n}^{(1,\alpha ) } t)]}{m\omega _ { 0}\sqrt{1+b_{m , n}^{(\sigma ) 2}\left ( \beta _ { \parallel } ^{2}-1\right ) } } , \label{hz1}\end{aligned}\]]and @xmath110 .
the transverse components are found from the formulae@xmath111where @xmath112 and @xmath113 , @xmath114 , and @xmath115 is the unit vector along the axis of the waveguide .
as we have seen , inside the waveguide the radiation is presented in the form of waves with discrete set of the values for the projection of the wave vector on the waveguide axis , @xmath116 , @xmath117 , determined by formula ( [ kmnpm ] ) .
having the radiation fields we consider the mean energy lost per unit time,@xmath118the radiation intensity @xmath119 is presented as the sum of intensities on separate modes:@xmath120the term with @xmath65 is present only then the condition @xmath121 is satisfied and the corresponding parts have the form@xmath122each term in the sums of these formulae corresponds to the radiation with the frequency given by the formula @xmath123 . for the radiation intensities on harmonics @xmath124 one has the formulae@xmath125for @xmath85 the upper limit of the summation over @xmath45 is defined by relation ( [ defnmax ] ) .
otherwise this limit is determined by the dispersion law for the dielectric permittivity @xmath3 through the condition @xmath126 .
it can be seen that for the case @xmath127 the expression for @xmath128 obtained from formulae ( [ itm ] ) and ( ite ) coincides with the corresponding formulae in ref .
taking @xmath129 , from formulae ( [ itm ] ) and ( [ ite ] ) we obtain the corresponding results for the radiation from a particle circulating in the plane perpendicular to the waveguide axis @xcite . in accordance with ( [ solcond ] ) , for given @xmath130 and @xmath45
the necessary condition for the presence of the radiation is the condition@xmath131now by taking into account the relation @xmath132 for the zeros of the bessel functions , we conclude that under the conditions @xmath133 and @xmath134 , with @xmath135 , there is no radiation inside the waveguide though the particle moves with acceleration .
if @xmath85 and the condition@xmath136 takes place then the intensity for the te waves defined by formulae ( ite ) goes to infinity .
however , under these conditions the absorption in the medium ( and also in the walls of the waveguide ) becomes important and the imaginary part of the dielectric permittivity should be taken into account .
formulae ( [ itm ] ) and ( [ ite ] ) are valid under the condition@xmath137 -1 , \label{conddiv}\]]where @xmath138 . instead of
@xmath38 we can introduce the angular variable @xmath139 , the values @xmath140 for which are related with the quantities @xmath141 by the formula@xmath142the quantities @xmath62 and @xmath143 are expressed via @xmath144 by the formulae@xmath145the possible values @xmath140 are determined by formulae ( [ kmnpm ] ) and ( [ teta]):@xmath146note that the singularity in the radiation intensity ( [ ite ] ) noted above corresponds to the values of the angular variable determined by the condition@xmath147 in the reference frame moving along the direction of the axis @xmath8 with the velocity @xmath4 the angle corresponding to @xmath148 is equal to @xmath149 . from ( [ costet ] ) the following relations can be seen@xmath150where@xmath151and @xmath152 is the cherenkov angle related to the drift velocity @xmath4 . hence , for @xmath59 the waves with @xmath96 ( @xmath98 ) are those which in the reference frame moving with velocity @xmath4 along the waveguide axis , propagate along the positive ( negative ) direction of the axis @xmath8 .
for @xmath58 the waves with @xmath96 ( @xmath98 ) propagate inside ( outside ) the cherenkov cone @xmath153 . from formulae
( [ itm ] ) , ( [ ite ] ) it follows that the number of radiated quanta does not depend on @xmath154 .
in particular , for @xmath59 , the same number of quanta is radiated inside and outside the cone @xmath155 .
for the case @xmath127 we could expect this result from the problem symmetry , as in the reference frame moving along the direction of the axis @xmath8 with the velocity @xmath156 we have a symmetric situation under the reflection with respect to the charge rotation plane .
now let us consider the radiation intensity in the limiting case of large values of the waveguide radius , @xmath157 .
in this limit the main contribution into the radiation intensity comes from large values @xmath45 and we can use the asymptotic formula ( see , for instance , abramovic ) @xmath158replacing the summation over @xmath45 by the integration and introducing as a new integration variable the angle @xmath139 , for the the radiation intensity one finds @xmath159where@xmath160 in this limit the frequency for the radiation along given direction @xmath161 is determined by the expression @xmath162 .
the expressions ( [ imhamtm ] ) and ( imhamte ) coincide with the corresponding formulae for the radiation in a homogeneous medium with dielectric permittivity @xmath3 .
note that in the discussed limit we have the transition @xmath163 and the upper ( lower ) sign corresponds to the angular region @xmath164 ( @xmath165 ) , where @xmath166 is defined by formula ( [ thet0 ] ) .
as an additional check for formulae ( [ itm ] ) , ( [ ite ] ) we can consider the special case @xmath167 for a fixed value @xmath168 .
this corresponds to a charge moving with constant velocity @xmath4 on a straight line @xmath169 parallel to the waveguide axis . in this limit @xmath170 and from condition ( solcond )
it follows that the radiation is present only under the cherenkov condition @xmath58 .
taking the limit @xmath171 , from formulae ( [ itm ] ) , ( [ ite ] ) we see that @xmath172hence , in the limit under consideration the tm waves are radiated only .
the corresponding frequency is given by the expression @xmath173 .
in the limit @xmath174 , by using asymptotic formula ( [ jmnas ] ) for the zeros of the bessel function , replacing the summation over @xmath45 by the integration , and using the formula @xmath175 , we can see that from ( [ omega0 ] ) the formula for the cherenkov radiation intensity in a homogeneous medium is obtained .
formula ( [ omega0 ] ) for the radiation of a charge moving parallel to the axis of the waveguide can be found , for example , in @xcite .
we have carried out numerical calculations for the number of the radiated quanta per one period of the particle orbiting,@xmath176 does not depend on @xmath154 . in figure [ fig3 ]
we have plotted the dependence of @xmath177 on the ratio @xmath178 for @xmath179 and @xmath180 corresponding to the energy 2 mev assuming that @xmath181 .
the graphs are given for @xmath182 ( left panel ) and @xmath183 ( right panel ) .
note that the location of the peaks in the number of radiated quanta for te waves is determined by formula ( [ peakcond ] ) .
in particular , for large values @xmath45 by using the asymptotic formula ( [ jmnas ] ) we see that the distance between the neighboring peaks is given by formula @xmath184 and decreases with increasing @xmath185 . as we see the main part of the radiated quanta
is emitted in the form of the te waves .
similar features take place for the radiation on other values of the harmonic @xmath130 .
for the same values of the parameters , for the number @xmath186 of the radiated quanta in the homogeneous medium evaluated from formulae ( [ imhamtm ] ) , ( imhamte ) , one finds@xmath187of course , in this case the result does not depend on @xmath2 .
[ cols="^,^ " , ] for ultrarelativistic particles the main part of the radiated energy is in the spectral range @xmath188 with @xmath189 being the plasma frequency , where the dielectric permittivity is well approximated by the formula @xmath190 . in this regime the radiation frequencies are of the order @xmath191 and for the undulator parameter @xmath192 cm we have @xmath193 . in this range
the perfect conductor boundary condition on the waveguide walls are no longer valid and for frequencies larger than the corresponding plasma frequency the waveguide becomes transparent .
the corresponding radiation intensity propagating in the exterior region is strongly directed in the forward direction and is described by the formulae given in @xcite . as regards the frequency range @xmath194 the corresponding radiation intensities are described by formulae ( itm ) , ( [ ite ] ) , where for materials with @xmath195 we can substitute @xmath196 . by taking into account formula ( [ helund ] ) for the radius of the helical orbit , from formulae ( [ itm ] ) , ( [ ite ] ) we can see that for @xmath197 the number of the radiated quanta @xmath198 is suppressed by the factor @xmath199 for both tm and te waves , whereas the corresponding frequencies are practically independent on the particle energy . for the modes with @xmath200 we have @xmath201 and @xmath202 . in this case the radiation on the tm modes dominates . in the discussion above we have considered the radiation emitted by a single particle moving along a prescribed trajectory . from the point of view of practical applications
the generalization of the obtained results for the case of the radiation from an electron beam is a next important step . in helical undulators
the beam is bunched and the features of the radiation critically depend on the ratio of the bunch length to the wavelength of the emitted radiation .
if the wavelength is smaller than the bunch length the particles in the bunch are not phase correlated and the total power radiated by the bunch is the sum of single particle parts . in the opposite limit , called the coherent spontaneous radiation regime , the wavelength of the radiation exceeds the length of the bunch and the particles radiate coherently . in this case the radiated power is enhanced by the factor of number of particles in the bunch relative to the incoherent radiation at the same wavelength .
here it is important to take into account that long wavelength radiation is suppressed by the waveguide cutoff condition and this leads to the constraints for the bunch length in order to have coherent radiation in the waveguide .
once the fields are evaluated , another interesting question is related to the influence of the radiation field on the motion of the radiating particles .
the interaction of the beam with its own radiation can induce an additional microbunching with the possibility of coherent radiation from particles of the same microbunch .
these questions require a separate consideration and we plan to address them in the future work .
a recent discussion of the beam dynamics in undulators can be found in @xcite ( see also @xcite ) .
another point which deserves a separate investigation is the role of the other processes of the particle interaction with the medium ( see also the discussion in @xcite ) .
in particular , they include ionization energy losses , particle bremsstrahlung in media , and the multiple scattering ( see , for instance , @xcite ) .
the relative role of these processes depends on the particle energy and characteristics of the medium . to our best knowledge , the previous investigations in this direction
were mainly concerned with the case of an unbounded homogeneous medium and the investigation of the effects induced by the presence of the waveguide requires a separate consideration . however , some general features can be obtained by using the corresponding results for a homogeneous medium .
first of all , as the above - mentioned processes are absent in the vacuum , the radiation discussed in the present paper is the main mechanism of the energy losses in sufficiently rarefied medium .
next , while the ionization losses by an electron rise logarithmically with energy and bremsstrahlung losses arise nearly linearly , the undulator radiation intensity arises quadratically and , hence , dominates at sufficiently high energies . for a particle moving in air the ionization losses dominate those due to the bremsstrahlung for the energies less than 100 mev . by using the standard formula , it can be seen that in the example corresponding to figures [ fig5 ] and [ fig6 ] the relative energy loss is @xmath203 per meter . an interesting possibility to escape ionization losses in the medium
was indicated in @xcite ( see also @xcite ) . in this paper
it was argued that a narrow empty channel along the particle trajectory in the solid dielectric does not affect the radiation intensity if the channel radius is less than the radiation wavelength . from the other side ,
the maximum impact parameter for ionization losses is of the order @xmath204 $ ] , where @xmath119 is the mean excitation energy for the atom of the medium , and for the channel radius larger than @xmath205 ionization losses are suppressed . as in other processes involving multiple scattering
, we expect that this effect will appear in the formula for the radiation intensity at given frequency in the form of the additional multiplicative factor and there exists a critical energy of the particle below which the multiple scattering does not effect the radiation .
we can try to estimate this factor by using the migdal formula . for the values of the parameters taken in the example above
this factor leads to the decrease of the radiation intensity by @xmath206 .
we have investigated the electromagnetic field generated by a charge moving along a helical orbit inside a circular waveguide with dielectric filling . this type of motion is involved in magnetic devices called helical undulators which are inserted into a straight sector of storage rings .
the helical undulators are used to generate circularly polarized intense electromagnetic radiation in a relatively narrow bandwidth .
the frequency of radiation is tunable by varying the beam energy and the magnetic field . in this paper
we have seen that the insertion of a waveguide into the helical undulator provides an additional mechanism for tuning the characteristics of the emitted radiation by choosing the parameters of the waveguide and filling medium .
the electric and magnetic fields are presented as the sum of two parts .
the first one corresponds to the fields of the charge in the homogeneous medium and the second one is induced by the presence of the waveguide .
the fourier components of the latter are given by formulae ( hml1w ) , ( [ hmz1w ] ) , ( [ electricb ] ) , ( [ electricc ] ) . in order to extract from the total fields the parts corresponding to the radiation
, we have investigated analytic properties of the fourier components as functions on @xmath38 .
these components have poles on the eigenmodes of the waveguide .
we have specified the ways by which these poles should be circled in the integral over @xmath38 .
for the cases @xmath85 and @xmath66 the corresponding contours are plotted in figures [ fig1 ] and [ fig2 ] . by using the residue theorem ,
the radiation fields are presented as a superposition of the eigenmodes of the waveguide corresponding to the tm and te waves .
the projection of the wave vector on the waveguide axis and the corresponding frequency are given by formulae ( [ kmnpm ] ) , ( [ ommnsigw ] ) .
we have derived formulae ( [ itm ] ) and ( [ ite ] ) for the radiation intensity emitted in the form of the tm and te waves .
limiting cases are considered and features of the radiation are investigated . in particular , we have seen that the main part of the radiated quanta is emitted in the form of the te waves .
applications of general formulae to helical undulators are given .
in particular , we have demonstrated that in the case of filled waveguide the radiation with features characteristic for ultrarelativistic particles in the empty waveguide is obtained for moderately relativistic particles .
the radiation emitted on the waveguide modes propagates inside the cylinder and the waveguide serves as a natural collector for the radiation .
this eliminates the necessity for focusing to achieve a high - power spectral intensity .
the geometry considered here is of interest also from the point of view of generation and transmitting of waves in waveguides , a subject which is of considerable practical importance in microwave engineering and optical fiber communications .
the authors are grateful to professor a.r .
mkrtchyan for general encouragement and to professor l.sh .
grigoryan , s.r .
arzumanyan , h.f .
khachatryan for stimulating discussions .
the work has been supported by grant no .
1361 from ministry of education and science of the republic of armenia . | we investigate the electromagnetic field generated by a point charge moving along a helical trajectory inside a circular waveguide with conducting walls filled by homogeneous dielectric .
the parts corresponding to the radiation field are separated and the formulae for the radiation intensity are derived for both te and tm waves . it is shown that the main part of the radiated quanta is emitted in the form of the te waves .
various limiting cases are considered .
the results of the numerical calculations show that the insertion of the waveguide provides an additional mechanism for tuning the characteristics of the emitted radiation by choosing the parameters of the waveguide and filling medium .
pacs number(s ) : 41.60.ap , 41.60.bq |
the grbm @xcite is a gamma - ray detection system onboard the bepposax satellite .
it is the secondary function of the anticoincidence shields of the pds experiment .
it is composed of four @xmath11100 @xmath2 slabs of csi(na ) scintillators operating in the 40 - 700 kev range .
each detector provides time series of the detected counts in the above energy range , with 1 s resolution as a continuous housekeeping and with better than 8 ms resolution upon trigger .
an oboard trigger is active whenever a statistically significant counting excess is simultaneously detected in at least two of the four detectors .
besides of being gamma - ray detection devices , the scintillating crystals composing the grbm are sensitive to highly ionizing particles that , leaving a large amount of energy ( @xmath1gev ) in just one shot , result in a phosphorescence phenomenon , with a consequent detection by the electronics of a large number of counts in few tens of ms , i.e. a _ spike _ in the counting rate .
when the same particle subsequently crosses two detectors , it causes the onboard logic to trigger on the event , that is electronically undistinguishable from a cosmic gamma - ray transient ( no particle anticoincidence is available to the grbm ) . the number of spikes that trigger the onboard electronics ( originating false triggers ) is of the order of 9 - 10/day .
therefore , the aim of this work is to provide a software filter that allows to make a ` safe ' first - order discrimination of the ` instrumental ' triggers from the ` cosmic - origin ' triggers .
this will allow the reduction of the huge number of onboard triggers recorded so far ( of the order of 10,000 in the first three years of bepposax operations ) , and to apply a more refined program to the generated sample of triggers .
a first drawback of the ` roughness ' of our filter is that it does not include criteria to separate gamma ray bursts from solar flares and from soft gamma repeaters events .
this task will be carried out by the ` second order ' filter .
our software filter is based on the automatic on - ground analysis of the high resolution time series , according to criteria established on the basis of the known detector / electronics behaviour and an extended study of the grbm time series .
usually , an eye inspection of a grbm light curve is sufficient to discriminate cosmic gamma - ray events from spurious events .
however , when an archival search for real grb is carried out this becomes not viable anymore , and an automatic filter is needed .
we therefore set - up an idl - based software code that implements a number of discrimination criteria to the grbm light curves .
this is a first approach to the problem and we did not make use of all the information that the grbm provides for each onboard trigger .
in fact , also 1-s resolution data are available in two energy ranges ( 40 - 700 and @xmath3700 kev ) and time - averaged energy spectra but they are not used in the analysis presented here .
the criteria implemented in the program are based on the knowledge of the instrument operational principles and on experience on the observed light curves .
it is likely that particular cases exist that were not taken into account in our code . at this time ,
the following parameters of the individual events are computed for each of the four grbm detection units : duration , rise time , simultaneity , shape , full width at half maximum .
the basic criterium of the code is to assign a ` score ' to each of these parameters , based on their comparison between pairs of detectors , with the goal of having the smallest score to the most - likely - cosmic events .
the individual scores for the different parameters then combine together to give a total score accounting for all the measured event characteristics . for the final score ,
the higher is the value the smaller is the probability of being a cosmic event .
in order to test and calibrate the efficiency of our software filter , we set - up a sample of selected events whose origin was known by different methods ( typically batse or ipn events ) . to these
a number of eye - screened false triggers were added . in figure 1
we present the results of the analysis carried out with our filter on such a sample of events .
the x - axis reports the score given to each event by the filter .
the texture classifies the events , based on the comparison with data from other satellites ( batse , ulysses , etc . ) .
the sample includes several short grbs in order to test the ability of the software to distinguish between them and spikes , but many long grbs are also included .
thus , we can define 3 score classes ( burst , spike , doubt ) , with the confidences given in table 1 ( the sum of every row is 100% ) , computed assuming multinomial distribution .
the software works on single peaks even during the evolution within a multiple - peaked event .
therefore , the results reported in the table should be intended as applicable to individual peaks in each light curve .
examples of events belonging to these three categories are presented in figure 2 .
.score calibration for the software filter [ cols="^,^,^,^ " , ] = 14.0 cm there are 3 additional special classes in the plot : no - end , no - class , no - trig .
no - end : the signal does not return to noise level before the end of the light curve ( grbm high time resolution data have a maximum coverage of 106 s ) .
this can be due to long bursts , or to a variable background ( see an example in @xcite ) .
no - class : the signal has not been analyzed .
this can happen for weak events if the software is not able to estimate the duration of the signal .
no - trig : the software is unable to find other than noise in the light curve , possibly due to very weak signals .
= 8.4 cm = 8.4 cm = 8.4 cm
the software filter has been applied to a fraction of the bepposax grbm data archive , covering more than 3 years of elapsed time , but consisting of about 603 days of satellite observing time .
the goal was to test our software on an homogenous sample of onboard triggers . a noticeable by - product of this operation is an estimate of the grb detection efficiency .
as stated above , the filter operates on individual peaks .
the result of the analysis gives : 440 peaks from cosmic events , 510 doubt cases , 4648 spikes , 58 no end , 86 no class .
if we apply the filter efficiency that was defined in the previous section , then the number of cosmic peaks become ( 367@xmath426 ) .
they belong to about 340 individual events , of which only @xmath1180 have been post - facto verified to be real cosmic events .
the results of the analysis on an unselected portion ( @xmath140% ) of the bepposax / grbm data archive presented in this paper allow to draw the following conclusions : - a number of @xmath1180 cosmic events detected by the bepposax / grbm were identified ( including a small number of solar flares and events from soft gamma repeaters ) , over a net exposure time of @xmath1545 days , leading to an estimation of the _ grbm efficiency on triggering cosmic events of @xmath10.33/day _ ( additional events are detected , but without an onboard trigger ) .
- the automatic filter has a @xmath150% efficiency on selecting real events out of false triggers ( i.e. , any event selected by the program has a @xmath150% probability of being of cosmic origin ) .
this reduced efficiency is likely due to the incompleteness of the sample on which the code was calibrated .
- the automatic filter has a @xmath390% efficiency on descarding false triggers ( i.e. , any event not selected by the program has a less than 10% probability of being a real event , based on the software calibration ) .
so , the filter allows to reduce the number of light curves to be analyzed to about 10% of the total , with an expected efficiency of more than 90% . | the bepposax gamma ray burst monitor ( grbm ) is triggered any time a statistically significant counting excess is simultaneously revealed by at least two of its four independent detectors .
several spurious effects , including highly ionizing particles crossing two detectors , are recorded as onboard triggers .
in fact , a large number of false triggers is detected , in the order of 10/day .
a software code , based on an heuristic algorithm , was written to discriminate between real and false triggers .
we present the results of the analysis on an homogeneous sample of grbm triggers , thus providing an estimate of the efficiency of the grb detection system consisting of the grbm and the software .
epsf _ proceedings of the 5@xmath0 huntsville gamma - ray bursts symposium ( 1999 ) _ |
a cycle cover of a graph @xmath6 is a collection of cycles @xmath7 such that every edge edge of @xmath6 belongs to at least one cycle of @xmath7 .
cycle covers are well studied objects , see @xcite for surveys , and in particular the length , i.e the sum of the lengths of the cycles in the cover , has received much attention .
we let @xmath8 denote the length of the shortest cycle cover of @xmath6 . in @xcite alon and tarsi made the following conjecture @xcite if @xmath6 is a cubic 2-edge connect graph then @xmath9 , where m is the number of edges in @xmath6 [ c : at ] if @xmath6 is 3-edge - colorable it is easy to see that the conjecture is true , and in fact @xmath10 , since we can use the cycle cover given by taking two pairs of edge colours . this particular covering will also be a @xmath5-covering , ie .
every edge lies in either 1 or 2 cycles , and it can be extended to a cycle double covering , where every edge belongs to exactly 2 cycles
. in @xcite kostochka proved that if conjecture [ c : at ] is true then it also holds with the added condition that the covering must be a @xmath5-covering , and zhang @xcite conjectured that for any 3-edge - connected cubic graph there is a shortest cycle cover which is a @xmath5-cover .
conjecture [ c : at ] remains open , it is known from @xcite that for both 2-edge and 3-edge - connected graphs the constant @xmath11 would be optimal , and the best current bound is @xmath12 , @xcite . in @xcite all snarks ( cyclically 4-edge connected , not 3 edge - colourable cubic graphs ) on @xmath13 vertices were generated and , among other things , the value of @xmath8 was computed .
it turns out that in this collection of graphs only two have @xmath14 , the examples are the petersen graph and a particular snark with 34 vertices , while all other have @xmath10 , and the authors made the following conjecture .
@xcite[c : s ] if @xmath6 is a snark then @xmath15 for the class of cyclically 4-edge - connected graph this conjecture gives a significant strengthening of conjecture [ c : at ] . in @xcite a snark with @xmath16 was constructed .. apart from the computational results of @xcite we also know that conjecture [ c : s ] holds for snark families where the girth grows with @xmath17 , since jackson @xcite proved that @xmath18 for 2-edge - connected cubic graphs of girth @xmath19 and @xmath20 .
additional conditions for having short cycle covers with different structural properties were given in @xcite .
in this paper we will show that conjecture [ c : s ] is true for several classical families of snarks by relating short cycle covers to the perfect matching index of a cubic graph , see section [ pm ] , where we also show that graphs with low perfect matching index satisfy zhang s conjecture
. we will also prove that for graphs with long cycles , of length at least @xmath21 , where @xmath22 , we have @xmath23 , and that graphs of oddness 2 have short cycle covers if they have small 2-factors . most of our results a proven by showing that the graphs at hand have cycle double covers which contain 2-regular subgraphs which are close to a 2-factor in size .
finally we discuss some open problems motivated by our current investigation , look at the connection to jaeger s petersen - colouring conjecture , and point out a perhaps unexpected rigidity in the structure of shortest covers of some snarks and graphs of low connectivity .
our proofs will make use of the following simple observations , neither of which are new and have been observed on different occasions by several other authors .
[ lem1 ] if a cubic graph @xmath6 has a 2-regular subgraph @xmath24 and a cdc @xmath25 such that @xmath26 then @xmath6 has a @xmath5-cycle cover of length @xmath27 typically we will be interested in cycle covers with length close to the shortest possible value @xmath4 and for the very shortest cases we can reverse the conclusion of the preceding lemma .
[ lem2 ] if @xmath6 has a cycle cover @xmath7 of length @xmath28 , where @xmath29 then the edges of weight 1 form a 2-regular subgraph @xmath24 of length @xmath21 such that @xmath30 is a cycle double cover of @xmath6 given a cycle covering @xmath7 we define the weight of an edge @xmath31 to be the number of cycles from @xmath7 which contain @xmath31 , and the weight @xmath32 of a vertex @xmath33 to be the sum of the weights of the edges incident to @xmath33 . the length of @xmath7 is then given by @xmath34 .
the lowest possible weight of a vertex is 4 , which happens if the incident edges have weights 1,1 and 2 .
we now note that in a covering of these lengths all edges must have weight 1 or 2 , since @xmath35 if there exists at least one edge of weight 3 .
next we note that the edges of weight 1 must form a 2-regular subgraph @xmath24 of @xmath6 with length @xmath21 , and c together with @xmath7 form a cycle double cover of @xmath6
in @xcite goddyn posed what is now known as the strong cycle double cover conjecture : given a cycle @xmath24 in a 2-connected cubic graph there is a cycle double cover of @xmath6 which contains @xmath24 . in @xcite this conjecture was verified for snarks on first 32 and then @xmath13 vertices . in @xcite
this result was used in order to prove that if the circumference @xmath36 of @xmath6 , ie the length of the longest cycle in @xmath6 , is close enough to @xmath17 then @xmath6 has a cdc . by following the method of that proof we can also prove that graph with large circumference have short cycle covers .
there is a function @xmath37 such that if the strong cycle double cover conjecture is true for cubic graphs with at most @xmath38 vertices then for @xmath6 such that @xmath39 we have @xmath40 and in general @xmath41 .
let @xmath24 be a longest cycle of @xmath6 and let @xmath42 be the cubic graph which is homeomorphic to the graph obtained by deleting all chords of @xmath24 .
if @xmath43 then @xmath42 can have at most @xmath38 vertices , which happens when the vertices not in @xmath24 form an independent set , and is 2-connected .
if the strong cycle double cover conjecture is true for cubic graphs on at most @xmath38 vertices then @xmath42 has a cdc @xmath44 which contains the cycle corresponding to @xmath24 .
the cycle @xmath24 together with all its chords is homeomorphic to a hamiltonian cubic graph @xmath45 with @xmath24 giving the hamiltonian cycle .
this graph has a 3-edge - colouring with colour 1 and 2 on the edges on the hamiltonian cycle and colour 3 on the chords .
each pair of colours define a 2-regular graph and those three graphs form a cdc @xmath46 of @xmath45 which contains the cycle corresponding to @xmath24 .
now @xmath47 forms a cycle double cover of @xmath6 which contains the 2-regular graph given by the colours 1 and 3 in @xmath45 .
this graph has size at least @xmath48 and the theorem follows by lemma [ lem1 ] @xmath37 exists for @xmath22 , and @xmath49 , @xmath50 @xmath51 since in this case the graph is hamiltonian . by theorem b of @xcite a cyclically 4-edge - connected cubic graph with an @xmath52-cycle @xmath24
has a cdc which contains @xmath24 , so @xmath53 by lemma [ lem1 ] , and the petersen graph shows that @xmath54 . in @xcite
it was verified that the strong cycle double cover conjecture holds for @xmath13 , so @xmath37 exists for @xmath22 .
recall that the _ oddness _ @xmath55 of a cubic graph @xmath6 is the minimum number of odd cycles in any 2-factor of @xmath6 .
graphs with oddness 0 are 3-edge - colourable and graphs with oddness 2 and 4 are known to have cycle double covers @xcite and also behave well with respect to other cycle cover / decomposition properties @xcite . for some graphs of oddness 2 we can show an optimal bound on the length of the shortest cycle cover .
if @xmath6 is cyclically 4-edge - connected and has a 2-factor with exactly two components then @xmath56 .
if there are three consecutive vertices on one the two cycles in the 2-factor with neighbours on the other cycle then @xmath57 .
there are graphs of this type with @xmath14 , e.g the petersen graph , so the bound in the theorem is at most 1 away from being sharp .
the second case of the theorem will for example apply if one of the two cycles is an induced cycle in @xmath6 , as in the petersen graph .
let @xmath58 be a 2-factor with exactly two components and let @xmath59 be three edges with endpoints on both @xmath60 and @xmath61 .
let @xmath42 be the cubic graph which is homeomorphic to the graph given by @xmath62 .
now @xmath42 has a 2-factor consisting of two even cycles @xmath63 and @xmath64 , corresponding to @xmath60 and @xmath61 , and so has a 3-edge - colouring with colours 1 and 2 on the edges of @xmath63 and @xmath64 , and colour 3 on the perfect matching @xmath65 given by @xmath66 .
we may assume that at least two of the three edges on which the endpoints of @xmath67 were in @xmath60 have colour 1 , if necessary by switching the colour 1 and 2 on @xmath63 , and likewise on @xmath64 .
the colouring now gives a cycle double covering of @xmath68 , by the cycles given by each pair of the three colours , and a corresponding cycle double covering @xmath44 of @xmath62 such that @xmath69 and @xmath70 .
similarly @xmath71 together with @xmath67 gives a graph with a cdc @xmath46 such that @xmath72 and @xmath73 .
now @xmath74 is a cdc of @xmath6 which includes the cycles given by the colours @xmath75 in @xmath42 , and by our choice of colouring the length of this 2-regular subgraph is at least @xmath76 , so by lemma [ lem1 ] we have @xmath77 if the second case of the theorem applies then we can choose the colouring of the cycle in the condition such that all three of those edge endpoints lie on an edge of colour 1 , thereby missing at most one vertex on the other cycle .
it is possible to generalize the preceding proof to the situation where the 2-factor has some even components as well , but the bound becomes weaker .
if @xmath6 is cyclically 4-edge - connected and has a 2-factor @xmath78 with exactly two odd components @xmath60 and @xmath61 , and the total number of edges in the three shortest disjoint paths from @xmath60 to @xmath61 in the multigraph obtained by contracting each cycle in @xmath78 to a vertex is @xmath79 , then @xmath80 .
let @xmath60 and @xmath61 be the two odd cycles in @xmath78 . by menger s theorem we can find three disjoint paths @xmath81 from @xmath60 to @xmath61 in @xmath6 .
we can now proceed as in the previous proof but using the three paths instead of the edges @xmath82 . since some cycles in @xmath78 can intersect more than one of the three paths @xmath83 we can not necessarily use the colouring modification used in the previous proof and so each edge of the three paths may contribute 2 edges which are not included in the ( 1,3)-coloured subgraph , thus adding @xmath84 to the length of the covering .
it is possible to make the bound for @xmath85 somewhat stronger but since the method does not seem likely to give a sharp result , and quickly becomes lengthy , we have not included that analysis here .
the perfect matching index @xmath86 of a bridgeless cubic graph is the smallest integer @xmath87 such that there exists perfect matchings @xmath88 such that @xmath89 note that @xmath6 is 3-edge - colourable if and only if @xmath90 . if fulkerson s conjecture @xcite is true then @xmath91 for all bridgeless cubic graphs @xmath6 .
the perfect matching index was studied in more detail in @xcite , where its value for several families of snarks were determined , and in @xcite it was shown that all but two of the snarks with @xmath13 vertices have @xmath92 . [ simpleprop ]
let @xmath93 be two perfect matchings in a cubic graph @xmath6 .
then @xmath94 induces a subgraph @xmath95 of @xmath6 which consists of disjoint even cycles . since every vertex of @xmath95 either has degree @xmath96 or it has degree 2 and one edge belongs to @xmath97 and the other to @xmath98 the result follows
let @xmath6 be a cubic graph .
then @xmath6 has a 5-cdc where one colour class is a 2-factor if and only if @xmath99 .
[ tauthm ] if @xmath90 then @xmath6 is 3-edge - colourable and the cdc given by the three pairs of edge colour classes is a 3-cdc in which each colour pair define a 2-factor .
if @xmath92 then we let @xmath100 be a set of four perfect matchings such that @xmath101 . by the pigeonhole principle ,
each vertex is incident to exactly one edge which is covered twice by the perfect matchings in @xmath102 .
therefore the set of edges covered by exactly two perfect matchings is a perfect matching which we denote by @xmath65 . now consider @xmath103 . by proposition [ simpleprop
] @xmath104 is a set of even subgraphs . if @xmath105 then @xmath31 is covered twice by @xmath104 and if @xmath106 then it is covered exactly once by @xmath104 .
hence @xmath107 is a 5-cdc of @xmath6 in which one colour class is a 2-factor .
conversely assume that @xmath6 has a 5-cdc @xmath108 where @xmath78 is a 2-factor of @xmath6 .
then @xmath109 is a perfect matching for @xmath110 .
let @xmath100 .
if @xmath111 then e is covered by exactly one perfect matching in @xmath102 and if @xmath112 it is covered by exactly two elements of @xmath102 . hence @xmath102 is a perfect matching cover of @xmath6 with four perfect matchings . in combination with our earlier results
this shows that graphs with @xmath113 has cycle covers of the shortest possible length .
[ th : tau ] if @xmath113 then @xmath114 by theorem [ tauthm ] @xmath6 has a cdc which contains a 2-factor and then the theorem follows by lemma [ lem1 ] .
theorem [ th : tau ] was implicitly present already in the phd thesis of celmins @xcite , with a different proof and terminology , but as far as we have been able to determine he did not publish it . that a low perfect matching index implies the existence of a 5-cdc , and thereby a short cycle cover ,
was recently also shown by steffen in @xcite . by this theorem
the snarks in several of the classical snark families have short cycle covers .
we refer to e.g @xcite for the full definition of these snarks families .
the flower snarks , goldberg snarks , and permutation snarks which are not the petersen graph have @xmath10 in @xcite it was proven the snarks in each of these families have @xmath92 . as a corollary
we also get a partial result on the conjecture of zhang @xcite mentioned in the introduction , as already pointed out by steffen @xcite .
if @xmath113 then the shortest cycle cover of @xmath6 is a @xmath5-cover .
finally we note that the situation for cubic graphs of cyclic edge - connectivity exactly 2 , or 3 , is not as simple as one might hope . as before we only need to consider graphs which can not be three - edge - coloured , so the only difference from the snark case is the cyclic connectivity .
there are standard reductions for cutting a cubic graph into two smaller graphs @xmath115 and @xmath115 at a 2-edge or 3-edge cut , adding in extra edges and/or a vertices to make the component graphs cubic .
ideally one could hope that a shortest covering of @xmath6 could be built from shortest coverings of @xmath116 and @xmath115 .
however this is not always the case . in order for this composition construction to work one
would need to know that for any edge @xmath31 in @xmath116 or @xmath115 there is a shortest cover where @xmath31 has a specified weight , in order to make sure that the cycle covers agree when we compose the two smaller graphs to form @xmath6 .
this is true for 3-edge - colourable cubic graphs , and the petersen graph , but it is not true for one of the two snarks on 18 vertices which has an edge with only one possible weight in all shortest coverings . in every 2-factor of the snark @xmath117 , shown in the left part of figure [ fig : sn18 ] , which can be extended to a cycle double covering
there is a cycle which includes the edge @xmath118 .
hence the edge @xmath118 has weight 1 in every shortest cycle cover of @xmath117 .
the observation can be proven by either a case analysis or a direct computer search .
[ fig : sn18 ] using this observation we can easily construct graphs where each component coming from a 2-edge cut has shortest cycle cover of length @xmath119 , while the original graph does not . [ ex1 ] delete the edge @xmath118 from three copies of the snark @xmath117 and then join them up to four additional vertices as shown on the right in figure [ fig : sn18 ] , where the dark circles are the copies of @xmath117 , to form a new graph @xmath120 .
if we cut away the bottom copy of @xmath117 , and add the edge @xmath118 to this copy and the edge @xmath121 , from this graph we are left with one large cubic graph @xmath122 which has a 2-factor which corresponds to a 2-factor in each of the copies of @xmath117 which includes the deleted edge @xmath118 , and so has a cycle cover of length @xmath123 .
however each such cycle cover gives weight 2 to the edge @xmath121 , which corresponds to the 2-edge cut used to separate the third copy of @xmath117 , and so can not be composed with a shortest cycle cover of @xmath117 .
hence @xmath124 in @xcite it was proven that the perfect matching index of a graph @xmath6 with a 2-edge cut is always at least the maximum of the perfect matching index of the two graphs obtained by cutting @xmath6 at a 2-edge cut . from theorem [ th : tau ] and example [ ex1 ]
it follows that the perfect matching index of a graph can be strictly greater than that of components coming from a 2-edge cut .
a we have seen snarks with @xmath8 larger than @xmath4 are rare among the small snarks , and the classical snark families , however we expect the minimum to increase for larger snarks . in @xcite jaeger made conjecture with far reaching consequences for the structure of cubic graphs .
a _ petersen_-colouring of a cubic graph @xmath6 is a map @xmath130 , where @xmath131 is the petersen graph , such that if @xmath67 are incident edges in @xmath6 then @xmath132 are incident edges in @xmath131 .
jaeger conjectured that all 2-connected cubic graphs have a petersen - colouring .
this is trivially true for 3-edge - colourable graphs and in @xcite the conjecture was verified for all snarks on @xmath13 vertices . given a petersen - coloring @xmath1 of @xmath6 and a cycle cover @xmath7 of @xmath131 we can construct a cycle cover of @xmath6 by taking the inverse image under @xmath1 of each cycle in the cover . each edge in @xmath131 makes a contribution to the length of @xmath133 which is equal to its weight in @xmath7 times the number of edges mapped onto it by @xmath1 .
using this observation , and saying that @xmath1 is balanced if the same number of edges in @xmath6 is mapped onto every edge in @xmath131 , we have : the reason for the latter point is that if the colouring is unbalanced we can always find a 9-cycle in @xmath131 to which more than the average number edges is mapped and use that as the cycle which is covered only once in the shortest cover of @xmath131 .
bill jackson . on circuit covers , circuit decompositions and euler tours of graphs . in _
surveys in combinatorics , 1993 ( keele ) _ , volume 187 of _ london math .
lecture note ser .
_ , pages 191210 .
cambridge univ . press ,
cambridge , 1993 . | in this paper we show that many snarks have shortest cycle covers of length @xmath0 for a constant @xmath1 , where @xmath2 is the number of edges in the graph , in agreement with the conjecture that all snarks have shortest cycle covers of length @xmath3 .
in particular we prove that graphs with perfect matching index at most 4 have cycle covers of length @xmath4 and satisfy the @xmath5-covering conjecture of zhang , and that graphs with large circumference have cycle covers of length close to @xmath4 .
we also prove some results for graphs with low oddness and discuss the connection with jaeger s petersen colouring conjecture . |
correlation functions of static quark and antiquark sources , represented by the polyakov loop and its hermitean conjugate , have proven to be very useful observables that allow to characterize thermal modifications of the interaction among partons at high temperature . at short distances
polyakov loop correlation functions are sensitive to the interaction between quarks and antiquarks which is transmitted through gluon exchange .
the thermal modification of gluons propagating in a hot medium is reflected in the screening of the interaction among static quark and antiquark sources .
this allows , for instance , the study of the running of the qcd coupling @xcite and its temperature dependence @xcite .
the large distance behavior of static quark correlation functions is sensitive to the confining properties of the qcd vacuum , the breaking of the string due to pair creation in the vacuum as well as the transition to a deconfined phase at high temperature and density . while at low temperature the correlation functions calculated in a @xmath4 gauge theory approach zero at large distances and
thus reflect strict confinement of quarks , they approach constant values in qcd with light dynamical quarks indicating the onset of string breaking at some characteristic distance . at zero temperature string
breaking arises through the creation of a quark - antiquark pair from the qcd vacuum . with increasing temperature more complex mechanisms involving quarks and gluons from the thermal medium also start playing a role . in the high temperature phase
static quark correlation functions approach constant values in qcd as well as in pure gauge theories suggesting that screening of static sources through a gluon cloud is the dominant mechanism for the neutralization of the color charge of static external sources .
information deduced from the temperature dependence of the change of free energy induced in a thermal medium due to the presence of external sources has been used to model the thermal modification of the potential between heavy quarks . through an analysis of the non - relativistic schrdinger equation , which may be a good approximation for heavy quarks , it has been concluded that some heavy quark bound states may also survive the transition to the high temperature phase of qcd @xcite .
recently it has been argued that also more exotic , colored bound states made up of quark antiquark pairs or quark - quark pairs may exist in a thermal medium at high temperature @xcite . to quantify such a scenario for quark - quark systems
requires a better understanding of the interaction among quarks at high temperature and the screening of colored diquark states .
first studies of quark - quark interactions at finite temperature have been performed in quenched qcd @xcite and some results for qcd with light staggered fermions have been reported at the lattice conference in 2005 @xcite . in this paper
we will focus on an analysis of static quark - quark correlation functions in 2-flavor qcd .
we will calculate free energies of diquark systems at finite temperature and compare with results obtained for quark - antiquark systems at the same temperature .
we will also calculate the induced quark number that arises from the presence of a diquark source that has non - vanishing triality in a thermal bath of quarks and gluons .
this provides information on the response of the thermal medium to the presence of external sources and allows to distinguish between the quark dominated screening mechanisms at low temperature ( string breaking ) and the gluon dominated screening at high temperature .
this paper is organized as follows . in the next section
we introduce the framework for our study of diquark correlation functions , discuss the basic setup for our numerical calculations and present some basic results on thermal properties of diquark and quark - antiquark free energies . in section
iii we present results on free energies of color anti - triplet and sextet diquark configurations calculated in coulomb gauge and discuss the gradual change from string breaking at low temperature to screening at high temperature which becomes transparent in the behavior of the net quark number induced in a thermal medium due to the presence of static quark sources . in section iv we discuss consequences of the observed casimir scaling in diquark and quark - antiquark systems for the possible existence of colored quark - quark bound states at high temperature .
we finally give our conclusions in section v.
the general framework for the analysis of thermal properties of qcd in the presence of @xmath5 static quarks and @xmath6 static antiquarks has been formulated by mclerran and svetitsky @xcite .
we concentrate here on 2 quark systems for which the lattice regularized qcd partition function with @xmath7 light dynamical quark degrees of freedom and static sources represented by the operator @xmath8 is given by , @xmath9 where @xmath10 are gauge field variables defined on the links of a 4-dimensional lattice of size @xmath11 ; @xmath12 denotes the gauge coupling , @xmath13 is the bare quark mass for the @xmath7 degenerate quark flavors .
furthermore , @xmath14 denotes the fermion matrix , for which we use the staggered fermion discretization scheme .
although all our calculations have been performed at vanishing quark chemical potential ( @xmath15 ) , we have indicated in eq .
[ qq_partition ] the dependence of the fermion determinant on @xmath15 .
we will make use of this in our discussion of the quark number induced in a thermal medium due to the presence of external sources . in eq .
[ qq_partition ] @xmath8 denotes the operator for two static quark sources separated by a distance @xmath16 .
the superscript @xmath17 indicates the color representation of the diquark system which we will specify later .
a static quark source is represented by the polyakov loop , @xmath18 and the source for antiquarks is given by the hermitean conjugate , @xmath19 .
in the definition of the polyakov loop we have introduced a renormalization constant , @xmath20 , that removes divergent self energy contributions to the static quark sources and insures that the diquark partition function normalized with the partition function in the absence of any sources , @xmath21 has a well defined continuum limit .
the renormalization constants for static quark sources have been determined previously for @xmath22 ( quenched qcd ) @xcite and 2-flavor qcd ( @xmath23 ) @xcite for the action and for values of the gauge coupling used also in this study . the normalized @xmath24 partition function , @xmath25 is the correlation function for two static quark sources separated by a distance @xmath16 .
its logarithm defines the free energy of the static diquark system , _
i.e. _ the change in free energy of a thermal medium at temperature @xmath26 that arises form the presence of two static sources , @xmath27 we note that the discussion of diquark partition functions and free energies given so far is completely analogous to the discussion usually given for the expectation value of the polyakov loop@xmath28@xmath29 .
] , @xmath30 and the related static quark free energy , @xmath31 the polyakov loop operator @xmath32 simply has been replaced by the diquark operator @xmath8 .
similarly one may , of course , also discuss the partition function and free energies of quark - antiquark systems . to be specific
we introduce here the so - called color averaged operators for a diquark and quark - antiquark system , @xmath33 like the polyakov loop also the operator defining the diquark partition function is not invariant under global @xmath34 transformations .
the static quark and diquark sources represent states of non - vanishing triality .
due to the @xmath34 symmetry of the @xmath4 gauge action partition functions with non - vanishing triality will be zero in quenched qcd . to insure a proper definition of diquark partition functions also for the pure @xmath4 gauge theory one should introduce a symmetry breaking term which is removed again after taking the thermodynamic limit . in the presence of dynamical quarks
the @xmath34 symmetry is , however , explicitly broken and the partition functions with non - zero triality are well defined .
the fermion sector of the qcd partition function , eq . [ partition ] , provides the appropriate number of quark and antiquark contributions such that only states with vanishing triality will finally contribute to the partition function .
we will elaborate further on this in section [ sec_string_breaking ] .
the analysis of diquark free energies presented in this paper has been performed on data samples that have been used previously to analyze free energies of static quark - antiquark systems @xcite .
these data samples have been generated on @xmath35 lattice for 2-flavor qcd using an improved staggered fermion action and quark masses , that correspond to a light pseudo - scalar ( pion ) mass of about @xmath36 mev .
for further details on this data sample and further details on the action and simulation parameters we refer to @xcite and references quoted therein .
the su(3 ) pure gauge data used in section [ subsec : sextet_screening ] have been obtained on lattices of size @xmath37 , where we used the tree - level symanzik improved gauge action @xcite . for further details see @xcite . in fig .
[ fig : plcqqav ] we compare the color averaged free energy of a static diquark system with that of a static quark - antiquark system both calculated in 2-flavor qcd .
the figure shows free energies as function of the separation of the two sources measured in units of the zero temperature string tension , @xmath38 . in both cases
renormalized polyakov loop operators , eq .
[ polyakov ] , have been used ; no further normalization of the relative magnitude of @xmath24 and @xmath39 free energies thus is needed nor has it been performed .
apparently the free energies of a diquark and a quark - antiquark system approach the same large distance limit at all temperatures .
in fact , as noted previously @xcite , this asymptotic value equals twice the free energy of a static quark represented by the renormalized polyakov loop , eq .
[ eq : fq ] .
this observation confirms our picture of screening at large distance ; the two static sources are screened independently and the thermal bath does not distinguish between quark and antiquark sources . at short distances
the @xmath39 free energies drop faster than the @xmath24 free energies .
this is expected from a perturbative analysis of the interaction in 2 quark systems ; the difference in casimir factors for one gluon - exchange in quark - quark and quark - antiquark systems suggests that the attractive interaction in an anti - triplet @xmath24 system is half as strong as in a singlet @xmath39 system .
the color averaged free energies receive also contributions from octet and sextet representations of a @xmath39 and @xmath24 system , respectively .
these contributions are , however , repulsive and thus are exponentially suppressed in polyakov loop correlation functions at short distances .
an additional difference in free energies of @xmath24 and @xmath39 arises from the fact that due to the non - vanishing triality of a @xmath24 state the free energy of a diquark system will receive an additional screening contribution which is not present in a @xmath39 system . to quantify the differences seen in fig .
[ fig : plcqqav ] and to explore in more detail the mechanism that leads to screening of color charges in a hot medium it is advantageous to study diquark and quark - antiquark correlation functions in definite color channels which can be defined in fixed gauges .
in addition to the color averaged diquark source introduced in eq .
[ qq_av ] we will consider here sources in different color channels . this is analogous to studies of color singlet and octet free energies performed for static quark - antiquark systems .
a system of two quarks can be either in a color anti - triplet ( antisymmetric ) or color sextet ( symmetric ) state , @xmath40 the corresponding operators defining the anti - triplet and sextet partition functions are given by @xcite @xmath41 these operators are , however , gauge dependent and the corresponding partition functions can only be studied in a fixed gauge .
we will analyze them in coulomb gauge .
the partition function for the color averaged diquark system can be written as a superposition of partition functions for color anti - triplet and sextet systems , @xmath42 it has already been observed in analogous studies of singlet and octet free energy in quark - antiquark systems that the large distance behavior of free energies does not depend on the relative color orientation of the two static sources . as shown in fig .
[ fig:36 ] this is also reflected in the large distance behavior of diquark free energies .
the same gauge invariant asymptotic value is approached by @xmath24 free energies calculated with sources in the anti - triplet and sextet representation .
we want to explore here in more detail the relation between the @xmath16-dependence of static diquark and quark - antiquark free energies . in leading order perturbation
theory the interaction in a @xmath39 singlet state is twice as strong as in a @xmath24 anti - triplet state .
this also carries over to finite temperature perturbation theory where in leading order the coulomb interaction is modified only through a debye - screening term that reflects the in - medium properties of the exchanged gluons , @xmath43 this suggest that at least at high temperatures the @xmath16-dependent @xmath24 and @xmath39 interaction part in the free energies obeys casimir scaling .
however , the gluonic screening cloud in a @xmath24 and @xmath39 system will differ as a function of distance . while we just have deduced from the long distance behavior of free energies shown in fig .
[ fig:36 ] that at infinite distances the screening clouds contribute the same amount to the free energy of a diquark and quark - antiquark system , respectively , this clearly has to be different at short distances .
a @xmath39 system in a color singlet state does not require a screening cloud at short distances ; it is color neutral by construction .
however , an anti - triplet @xmath24 state needs to be screened even when the separation between both sources vanishes .
in fact , we expect that at short distances the free energy of the screening cloud for an anti - triplet @xmath24 state is identical to that of a single static antiquark , @xmath44 .
the above considerations suggest a simple relation between free energies of anti - triplet @xmath24 states and color singlet @xmath39 systems , which should be valid at least at high temperature as well as at large and short distances , respectively , @xmath45 in fig . [
fig : compare ] we compare the left and right hand sides of eq .
[ qqdifference ] at various temperatures .
as can be seen the equality indeed holds very well at all distances for temperatures above the transition temperature , while some differences show up at intermediate distances below @xmath3 . nonetheless , at short distances the equality in eq .
[ qqdifference ] holds at all temperatures which shows that indeed a well localized anti - triplet diquark system is screened in the same way as a single antiquark .
moreover , we note here that eq .
[ qqdifference ] is found to hold also in the deconfined phase of su(3 ) pure gauge theory @xcite .
we want to elaborate here a bit more in detail on the differences seen in the screening of color singlet quark - antiquark and anti - triplet diquark free energies below @xmath3 .
as noted before the contribution of states with at least one additional quark is needed in the diquark partition function to project on the sector of vanishing triality . at low temperature
it is , in fact , a single quark which dominates the additional contributions in a diquark partition function for small separations of the two quarks . at large distances , on the other hand ,
two antiquarks will be needed to screen the color charges of the two well separated quarks .
the situation in a quark antiquark system is quite different .
no additional quarks are needed for screening of the sources at short distances , while a quark - antiquark pair is needed for screening at large distances , _
i.e. _ the net quark number will always vanish . at low temperature the screening mechanism at large distances
generally is referred to as string breaking . at high temperature
such a screening mechanism based on the presence of dynamical quarks in the thermal medium or the vacuum becomes less relevant ; screening with gluons becomes the dominant mechanism . to explore the features of screening and string breaking and the change of screening mechanisms at low and high temperature on a more quantitative level we have analyzed the net quark number induced in a thermal medium due to the presence of a diquark system separated by a distance @xmath16 .
the net quark number can be obtained from the static diquark free energy , eq .
[ eq : fx ] , by considering its parametric dependence on the quark chemical potential through the fermion determinant as indicated in eq .
[ partition ] .
a derivative of @xmath46 with respect to @xmath47 evaluated at @xmath48 then yields the net quark number , @xmath49 where @xmath50 is the quark number operator in 2-flavor qcd , n_q = .
similarly we obtain the net quark number induced by a single static quark source , @xmath51 for a quark - antiquark system it is straightforward to see that the induced net quark number indeed is strictly zero for all temperatures , _
i.e. _ @xmath52 .
however , as argued above , for a single quark or diquark system a net quark number has to be induced to generate a state with vanishing triality . in fig .
[ fig : plcqqav_c1 ] we show the net quark number induced by color averaged diquark sources separated by a distance @xmath16 .
we observe a clear separation between the short and long distance behavior of this observable . while for small separation of the sources a net quark number ( @xmath53 ) is induced , a net antiquark number ( @xmath54 ) is induced at large separations .
this effect becomes more pronounced with decreasing temperature . above the transition temperature
, however , the net quark number rapidly goes to zero .
the properties of @xmath55 become more apparent when we consider also the net quark numbers induced by static sources in an anti - triplet state . in fig .
[ fig : nq_asymptotic ] we show the values for the induced net quark number in color averaged and anti - triplet diquark systems in the large distance limit as well as for the shortest distance available , @xmath56 . at all temperatures
we find that the net quark number induced for small separations of the sources in the anti - triplet channel is positive and equals the net quark number needed to neutralize a static antiquark source . at large separations , on the other hand , it is minus twice that needed to neutralize static antiquark source . at short distances
a @xmath24 anti - triplet diquark thus is neutralized by a light quark from the thermal medium , while at large distances both quark sources are individually neutralized by a light antiquark .
while at finite temperature @xmath55 also receives contributions from other ( multi - quark ) states , this mechanism becomes increasingly dominant at low temperature .
in fact , the data shown in figs . [ fig : plcqqav_c1 ] and [ fig : nq_asymptotic ] support the expected limiting behavior , @xmath57 with @xmath58 being a typical hadronic scale , @xmath59fm , that characterizes the string breaking radius at zero temperature . in all cases considered here
the induced net quark number rapidly becomes small at high temperature .
in fact , already for @xmath60 the deviations from zero are smaller than 0.02 at all distances .
thermal fluctuations of the quark number and the largely available number of thermal gluons thus seem to be sufficient to screen the static color sources at high temperature .
the considerations of the previous section can also be carried over to sextet diquark systems .
the perturbative one - gluon exchange among quarks in a sextet representation is repulsive .
also this interaction is screened at finite temperature . at high temperature
we thus expect to find for the genuine interaction part in the sextet free energy , @xmath61 while the large distance behavior of the sextet and anti - triplet free energies is identical ( see fig . [ fig:36 ] ) , we again expect that at short distances the sextet diquark state is screened differently from the anti - triplet .
in fact , we expect that screening of a sextet diquark at short distances is identical to screening of a single fermion in a color sextet representation , _
i.e. _ when combining sextet and anti - triplet free energies in proportion to the corresponding casimir factors such that the repulsive and attractive coulombic contributions cancel we expect to find @xmath62 where @xmath63 is the free energy of a fermion in a color sextet representation given by the corresponding polyakov loop expectation value , @xmath64 where @xmath65 is obtained from standard group theoretical relations .
eq . [ f63 ]
is indeed fulfilled at all temperatures in 2-flavor qcd as well as in the deconfinement phase in su(3 ) pure gauge theory , as can be seen in fig .
[ fig : f63 ] .
we note that the expectation value of a polyakov loop in the sextet representation is , of course , gauge invariant .
the results on the short distance behavior of sextet as well as anti - triplet quark - quark free energies thus reflect gauge invariant screening properties of colored diquark systems .
the analysis of diquark and quark - antiquark free energies presented in the previous section suggests that aside from the different screening mechanisms for @xmath39 and @xmath24 systems the r - dependence of the free energies obeys casimir scaling above @xmath3 ; _ i.e. _ the interaction in a diquark system is only half as strong as in a @xmath39 system .
an immediate consequence of this is that the temperature dependent coupling defined in terms of @xmath24 systems , @xmath66 coincides with that extracted previously from color singlet free energies @xcite . the close relation between color singlet quark - antiquark free energies and anti - triplet diquark free energies , eq .
[ qqdifference ] , of course also carries over to the corresponding energies ( @xmath67 ) and entropies ( @xmath68 ) deduced from the free energy , @xmath69 , through appropriate partial derivatives with respect to temperature at fixed separation , @xmath16 , of the sources .
free energies as well as combinations of energy and entropy of static quark systems have been used to construct temperature dependent , effective potentials for heavy quark systems that allow to model the temperature dependence of heavy quark bound states in the high temperature phase of qcd .
although such constructions are not rigorous and will include model assumptions , it seems to be clear from the analysis presented in the previous section that any potential constructed for color singlet quark - antiquark systems also puts stringent constraints on the interaction in a diquark system .
we expect to have @xmath70 if a tightly bound @xmath39 systems can exist at high temperature it thus is conceivable that also @xmath24 states may persist in some temperature regime above @xmath3 .
as we have seen in the previous section the color charge of such a @xmath24 system will in addition be screened by a gluon cloud .
using eq .
[ qqpotential ] we have solved the schrdinger equation for diquark states formed from charm and bottom quarks respectively , @xmath71 \phi_i^a = m_i^a(t)\ ; \phi_i^a\ ; , \ ; a\ ; = \ ; { \rm c,}\ ; { \rm b}.\ ] ] as we do not want to go into a detailed modeling of heavy quark bound states but rather want to discuss some generic features expected to hold for @xmath24-systems given some knowledge on the behavior of @xmath39-systems , we have used a generic ansatz for the heavy quark potential , which has been used previously also in studies of charmonium and bottomonium systems , @xmath72 here we have used parameters for the coupling in the coulomb term and the linear rising confinement term of the zero temperature potential that describe well the lattice results for the static quark potential , @xmath73 and adjusted the screening mass , @xmath74 , such that the finite temperature potential reproduces the melting temperature of the charmonium ground state , @xmath75 , at a temperature consistent with lattice studies of spectral functions , @xmath76 @xcite . with this ansatz
we find dissociation temperatures for ground states in the bottomonium system ( @xmath77 ) and the corresponding diquark states given in table [ tab : boundstates ] .
as can be seen this analysis suggests that @xmath2 states are unlikely to exist in the high temperature phase of qcd .
they dissolve already at @xmath3 .
however , @xmath0 states dissolve only at temperatures of about @xmath78 .
.dissociation temperatures , @xmath79 , for heavy quark - antiquark and diquark bound states in units of the transition temperature @xmath3 and dissociation energies for these states at @xmath3 .
[ cols="^,^,^,^,^",options="header " , ]
we have presented results on free energies of static diquark systems .
a detailed comparison with corresponding results for free energies of static quark - antiquark systems shows that the relative magnitude of the r - dependent part of these free energies obeys casimir scaling although at short distances the screening of diquark states is quite different from that of quark - antiquark states ; colored diquark systems at short distances are screened like static quark sources in the same color representation . using these results in a model for heavy diquark bound states we showed that it is likely that in a quark gluon plasma @xmath0 states can form at temperatures well above the transition temperature while @xmath2 states are unlikely to exist above @xmath3 .
this work has been supported in part by contract de - ac02 - 98ch1 - 886 with the u.s .
department of energy , the helmholtz gesellschaft under grant vi - vh-041 and the deutsche forschungsgemeinschaft under grant grk 881 .
99 s. necco and r. sommer , nucl .
b * 523 * , 135 ( 2001 ) and nucl .
b * 622 * , 328 ( 2002 ) .
o. kaczmarek , f. karsch , p. petreczky and f. zantow , phys .
d * 70 * , 074505 ( 2004 ) , erratum - ibid d * 72 * , 059903 ( 2005 ) .
o. kaczmarek and f. zantow , phys .
d * 71 * , 114510 ( 2005 ) .
k. symanzik , nucl .
b * 226 * , 187/205 ( 1983 ) .
s. nadkarni , phys .
d * 33 * ( 1986 ) 3738 .
s. nadkarni , phys .
d * 34 * ( 1986 ) 3904 .
w. m. alberico , a. beraudo , a. de pace and a. molinari , arxiv : hep - ph/0612062 .
s. digal , p. petreczky and h. satz , phys . lett .
b * 514 * , 57 ( 2001 ) s. digal , p. petreczky and h. satz , phys .
d * 64 * , 094015 ( 2001 ) c. y. wong , phys .
c * 65 * , 034902 ( 2002 ) and j. phys .
g * 28 * , 2349 ( 2002 ) e. shuryak and i. zahed , phys .
c * 70 * , 021901 ( 2004 ) and phys
d * 70 * , 054507 ( 2004 ) .
d. blaschke , o. kaczmarek , e. laermann and v. yudichev , eur .
j. c * 43 * , 81 ( 2005 ) w. m. alberico , a. beraudo , a. de pace and a. molinari , phys .
d * 72 * , 114011 ( 2005 ) a. mocsy and p. petreczky , phys .
d * 73 * , 074007 ( 2006 ) c. y. wong , phys .
c * 72 * , 034906 ( 2005 ) a. nakamura and t. saito , prog .
* 112 * , 183 ( 2004 ) and phys .
b * 621 * , 171 ( 2005 ) z. fodor , c. hoelbling , m. mechtel and k. szabo , pos * lat2005 * , 310 ( 2006 ) . k. hubner , o. kaczmarek and o. vogt , pos * lat2005 * , 194 ( 2006 ) .
l. mclerran and b. svetitsky , phys .
d * 24 * , 450 ( 1981 ) .
o. kaczmarek , f. karsch , p. petreczky and f. zantow , phys .
b * 543 * , 41 ( 2002 ) s. gupta , k. hubner , o. kaczmarek , in preparation m. asakawa and t. hatsuda phys .
lett . * 92 * , 012001 ( 2004 ) .
s. datta , f. karsch , p. petreczky and i. wetzorke , phys .
d * 69 * , 094507 ( 2004 ) .
a. jakovac , p. petreczky , k. petrov and a. velytsky , arxiv : hep - lat/0611017 . | we analyze the screening of static diquark sources in 2-flavor qcd and compare results with the screening of static quark - antiquark pairs .
we show that a two quark system in a fixed color representation is screened at short distances like a single quark source in the same color representation whereas at large distances the two quarks are screened independently . at high temperatures we observe that the relative strength of the interaction in diquark and quark - antiquark systems , respectively , obeys casimir scaling .
we use this result to examine the possible existence of heavy quark - quark bound states in the high temperature phase of qcd .
we find support for the existence of @xmath0 states up to about @xmath1 while @xmath2 states are unlikely to be formed above @xmath3 . |
noise induced phenomena in nonlinear systems have attracted a great deal of attention in a variety of contexts in physics , chemistry and the life sciences .
an important example is the phenomenon of stochastic resonance ( sr ) @xcite , in which the response of the system ( output ) to external driving ( input ) is amplified and optimized for certain values of the noise parameters .
more specifically , the non - monotonic behavior of the output signal - to - noise ratio ( snr ) with the strength of the noise is a widely accepted signature of sr .
in addition , a dimensionless quantity known as the sr gain is usually defined as the ratio of the output snr over the input snr .
very recently , very large sr gains have been reported for systems formed by a finite number @xmath0 of globally coupled bistable oscillators @xcite . here
the term _ global coupling _ is used to indicate that each oscillator interacts with _ all _ other oscillators .
these systems were used years ago by kometani and shimizu @xcite as an empirical model to describe muscle contraction .
later on , desai and zwanzig @xcite gave a more detailed statistical mechanical description , finding an order - disorder transition for a variable defined as the expectation value of the position of one oscillator .
this variable is used to study the global behavior of the coupled bistable system .
desai and zwanzig focused on systems with infinitely large sizes , @xmath1 , investigating the system dynamics by analyzing the time evolution of the central cumulant moments of one - oscillator s distribution .
in addition , a gaussian approximation was proposed in order to close the cumulant moment hierarchy and obtain analytical expressions .
a similar approach is currently used as a mean field approximation in the investigation of various noise - induced phenomena such as noise - induced phase transitions @xcite .
recently , in order to study the effect of fluctuations due to the finite size of the system , pikovsky et al .
@xcite extended this approach by replacing the expectation values of one - oscillator properties @xmath2 by arithmetic means over all oscillators @xmath3 .
a gaussian approximation , including noisy terms , was derived and used to illustrate the phenomenon of system size resonance , in which the sr quantifiers display a non - monotonic behavior as a function of @xmath0 . in this paper , the work by pikovsky et al .
is extended to higher orders in the fluctuating cumulant dynamics .
the gaussian approximation is re - derived using a rigorous formalism based on it stochastic calculus and compared with other approximations .
one important goal of this paper is to explain the very large gain values observed in globally coupled bistable systems @xcite , especially when compared with those observed in uncoupled or isolated bistable systems . to that effect , it is desirable to derive a simplified theory in which the number of degrees of freedom is much smaller than the number of coupled oscillators , thus being more amenable to analytical treatment or qualitative interpretation . in this regard ,
the gaussian approximation is a practical alternative , though in principle not fully satisfactory , because it is not based on a small parameter expansion but on an uncontrolled assumption ( the neglect of cumulants higher than the second ) that is known to be not accurate even in the limit of an infinite system @xcite . in this paper , this approximation is presented , as well as other simplified models with a reduced number of degrees of freedom which are able to mimic the most important features of the system dynamics with a finite size .
these simplified models represent different approximation schemes and might be regarded as an expansion or generalization of the work by desai and zwanzig @xcite and pikovsky et al .
@xcite .
the paper is organized as follows . in the next section , the model system and the sr quantifiers
are defined .
the simplified models are presented in sec .
[ sec : eff ] .
these models are compared to the original model system by means of computer simulations in sec .
[ sec : sim ] . finally , sec .
[ sec : con ] provides a short summary and conclusions .
let us consider a set of @xmath0 interacting bistable oscillators , each one of them characterized by a single degree of freedom @xmath4 ( @xmath5 ) , whose dynamics is governed by the langevin equations @xcite @xmath6 where @xmath7 is a gaussian white noise with zero average and @xmath8 @xmath9 is a coupling parameter defining the strength of the interaction between oscillators , and @xmath10 is an external driving force of period @xmath11 . to characterize the system as a whole we define the _ collective _ or _ global _ variable @xmath12 as @xmath13
the stochastic resonance quantifiers for this variable are defined in the usual way .
the one - time correlation function , @xmath14 can be written as the sum of two contributions : a coherent part , @xmath15 which is periodic with period @xmath11 , and an incoherent part @xmath16 which decays to zero for large values of @xmath17 and reflects the correlation of the process @xmath12 at different times due to fluctuations . in the expressions
above , the notation @xmath18 indicates an average over the noise realizations and the subscript `` @xmath19 '' indicates the long time limit of the noise average , i. e. , its value after waiting for @xmath20 long enough that transients have died out .
the snr of a random signal measures the signal strength relative to its background noise .
more specifically , we calculate the output snr as @xmath21 where @xmath22 @xmath23 being the driving frequency , and @xmath24 note that the quantity @xmath25 is proportional to the so - called spectral amplification , which is another widely used sr quantifier . as the size of the system @xmath0 is increased while keeping the interaction parameter @xmath9 constant , the collective variable @xmath12 becomes less noisy , becoming completely deterministic in the limit @xmath1 . as a result
, @xmath26 diverges in that limit .
this is a consequence of the averaging process implicit in the definition ( [ eq : def : s ] ) .
the sr gain is defined as @xmath27 where @xmath28 is the snr of the collective input signal @xmath29 $ ] .
for example , for a periodic rectangular driving force of amplitude @xmath30 , the input snr is given by @xmath31 .
the sr gain ( [ eq : gaindef ] ) is a dimensionless quantity that measures the amplification of the system response with respect to the collective input signal .
the input snr @xmath28 diverges linearly with @xmath0 in the limit @xmath1 , so that the sr gain remains finite .
since in a system with coupled linear oscillators the snr of the collective process equals the snr of the collective input signal , i.e. @xmath32 , the sr gain also measures the response of the non - linear system with respect to that of a linear system subject to the same deterministic and stochastic forces .
additionally , in the absence of interaction between the bistable oscillators ( the case @xmath33 ) , the collective sr gain @xmath34 equals the sr gain of each independent oscillator @xcite .
thus , by comparing the sr gain values of the collective variable of a finite set of interacting oscillators with those observed in the case @xmath35 , we have a useful tool to highlight nonlinear effects that are a direct consequence of the coupling between the oscillators .
in this section , we define a set of stochastic processes , which we will refer to as _ fluctuating cumulants _ , in order to describe the dynamics of a finite system of coupled oscillators in terms of a reduced number of variables .
then , by using it stochastic calculus , we derive the hierarchy of equations that these cumulants obey .
a few approximation schemes are proposed for systems with a large but finite number of oscillators .
finally , we introduce a simple one - variable model in which the dynamics of @xmath12 is mimicked by using an effective potential .
the fact that the infinite system ( @xmath36 ) is completely deterministic , and the approximations described in this section are valid for large @xmath0 , makes these methods especially appropriate to study the effect of fluctuations due to the finite size of the system .
let us define the set of stochastic variables @xmath37^k,\ ] ] with @xmath38 being a positive integer .
we will refer to @xmath39 as the _ fluctuating moment _ of order @xmath38 .
note that @xmath40 . in order to obtain a hierarchy of stochastic differential equations for these variables , we need first to choose a convenient stochastic interpretation . the langevin equations ( [ eq : lang ] ) are well defined and do not depend on the stochastic interpretation . note , however ,
that the specific form of the stochastic differential equations for the fluctuating moments do depend on the stochastic calculus utilized . in the following , unless explicitly stated , it stochastic calculus is assumed .
it is customary within this calculus to use a notation to express the stochastic differential equations in which there is no explicit mention to the white noises ( see for example @xcite ) . in particular , eq .
( [ eq : lang ] ) would be written as @xmath41dt+(2d)^{1/2}{{\mathrm d}}b_i , \label{eq : lang : ito}\ ] ] where @xmath42 , with @xmath43 , is the differential of the wiener process @xmath44 with properties @xmath45 the gaussian white noise @xmath7 can be viewed as proportional to the derivative of @xmath44 , @xmath46 , though it is not an ordinary stochastic process but a generalized process and requires a special formalism to be defined rigorously ( see @xcite and references within ) . here
we will use both notations at convenience . using it differentiation rules
@xcite we find the following stochastic differential equations for the fluctuating moments @xmath47 where @xmath48^k\xi_i(t ) \label{eq : muk : def}\ ] ] and @xmath49 are @xmath50correlated noises with the following first moments @xmath51 notice that the result ( [ eq : mukmukp ] ) is only obtained from ( [ eq : muk : def ] ) when it calculus is assumed ( see appendix [ sec : apen1 ] ) .
using stratonovich calculus leads to a much more intricate expression in which the approach proposed in this paper is not applicable .
additionally , note that ( [ eq : mukmukp ] ) implies that the processes @xmath52 are not uncorrelated .
rigorously , only @xmath53 is a gaussian process .
nevertheless , it can be shown that in the asymptotic limit of a very large number of oscillators , @xmath1 , all @xmath52 tend towards a gaussian behavior ( see appendix [ sec : apen2 ] ) .
this property will allow us to rewrite eq .
( [ eq : hier : mk ] ) as a closed set of stochastic equations for the fluctuating variables in that limit .
we can define a set of fluctuating cumulants by using the formula : @xmath54 equation ( [ eq : mom : cum ] ) is the formula that relates the moments with the cumulant moments of a single - variable stochastic distribution . when the stochastic variable is gaussian , all cumulants @xmath55 , with @xmath56 , exactly vanish .
a description in terms of cumulants is preferable because , unlike a descriptions with moments , it is expected that higher order cumulants are negligible in comparison with lower order cumulants , especially if the deviation with respect a gaussian behavior is not very large . in terms of the fluctuating cumulants @xmath55 , the first three equations of the hierarchy ( [ eq : hier : mk ] ) are @xmath57 the noise terms @xmath58 and @xmath52 vanish in the formal limit @xmath1 , therefore leading to a deterministic hierarchy of equations for the fluctuating moments or cumulants .
this deterministic hierarchy is equivalent to the non - linear hierarchy obtained by desai and zwanzig in @xcite for the cumulant moments of the process @xmath59 .
in contrast to the theory presented in ref .
@xcite , which is based on the calculation of one - time expectation values , the fluctuating cumulant approach will allow us to study dynamical properties such as autocorrelation functions . when @xmath0 is finite ,
the hierarchy of equations that @xmath39 or @xmath60 obeys is not closed , since the noise processes @xmath52 depend on @xmath39 in a non - trivial way . in the next subsections we present a few approximative schemes that overcome this difficulty for systems with a large number of coupled oscillators .
if we retain the first two equations of ( [ eq : hie : cum:1])-([eq : hie : cum:3 ] ) , neglect all fluctuating cumulants @xmath55 with @xmath61 , and also neglect the term @xmath62 and the noise @xmath63 in ( [ eq : hie : cum:2 ] ) , we obtain a closed set of equations for the processes @xmath12 and @xmath64 : @xmath65 these set of equations was proposed by pikovsky et al . in ref .
this truncation scheme has been called `` the gaussian approximation '' because all fluctuating cumulants with order higher than the second one are neglected .
there is no reason to expect a priori that these higher order cumulants can be neglected in any limit , other than the hope that their contribution is small .
note , in addition , that in this scheme the @xmath66correlated noise @xmath63 is neglected without justification .
let us now focus in a third order truncation scheme .
we will retain the three eqs .
( [ eq : hie : cum:1])-([eq : hie : cum:3 ] ) , but consistently neglect @xmath67 and @xmath68 .
since each @xmath69 for @xmath70 is a gaussian process in the lowest order in @xmath71 ( see appendix [ sec : apen2 ] ) , its probability distribution is completely determined by its first moments ( [ eq : muk:0])([eq : mukmukp ] ) .
as mentioned before , the processes @xmath69 are not independent of each other .
thus , it is preferable to express them in terms of a set of independent gaussian noises @xmath72 with zero mean and @xmath73 where @xmath74 and @xmath75 . with the expansion @xmath76 we only need to select the coefficients @xmath77 so that the correlations ( [ eq : mukmukp ] ) are satisfied .
this can be achieved by using the gram - schmidt ortho - normalization method .
the result for the first two terms is @xmath78 note , however , that in these expressions the coefficients @xmath77 appear as functions of the average values of the fluctuating cumulants @xmath55 .
thus , if we plan to solve the eqs .
( [ eq : hie : cum:1])([eq : hie : cum:3 ] ) using ( [ eq : mu1])([eq : mu2 ] ) , we would have to consider the equation of motion for @xmath79 @xcite and solve the whole set of equations self - consistently .
alternatively , we could use a slightly different version of ( [ eq : mu1])([eq : mu2 ] ) in which the expected values @xmath79 are replaced by @xmath55 .
this way , the correlations ( [ eq : mukmukp ] ) for the first noise terms are also identically satisfied , and the fact that the fluctuating cumulants become deterministic in the limit @xmath1 guarantees that the proposed expressions for the noises @xmath69 are gaussian in the lowest order in @xmath71 .
since all gaussian processes are completely determined by its first two moments , both methods to generate the noise sources @xmath52 are mathematically equivalent in the asymptotic limit of large @xmath0 , though the later is more physically appealing because in this case the instantaneous value of the noise source @xmath69 in one trajectory does not depend on averages over trajectories but on single - trajectory values .
using this later procedure , the following stochastic differential equations with multiplicative noise are obtained @xmath80 this method has also the advantage that the system of equations ( [ eq : third ] ) can be solved numerically using standard stochastic algorithms @xcite .
it represents a third order approximation scheme .
notice that this scheme can be applied in a straightforward way to obtain the corresponding equations of an arbitrary order truncation of the fluctuating cumulant hierarchy .
as we increase the truncation order of the fluctuating cumulant hierarchy , as we have discussed above , a more accurate approximation is obtained .
however , the number of equations is also increased . on the other hand
, we may wonder how good a description based on a single differential equation is .
the aim is to derive a simplified model that may not mimic quantitatively but qualitatively the coupled system dynamics , in addition to being more amenable to analytical treatment .
here we consider the following single stochastic equation @xmath81 where @xmath53 is the gaussian white noise defined by ( [ eq : muk:0])([eq : etaeta ] ) , and @xmath82 an effective potential to be specified .
we can determine the effective potential uniquely by requiring the model to reproduce the equilibrium properties of the original system .
the stationary probability density @xmath83 of the langevin eq .
( [ eq : eff ] ) in the absence of external driving is given by @xcite @xmath84 where @xmath85 is a normalization constant .
therefore , by inverting ( [ eq : peq ] ) , we find an expression for the effective potential up to an additive constant @xmath86 , @xmath87 in ref .
@xcite , desai and zwanzig presented an analytical expression for the equilibrium density @xmath83 by retaining the leading term in the asymptotic expansion for large @xmath0 .
we will refer as @xmath88 to the corresponding effective potential .
this analytical solution shows that for a given value of @xmath9 there exists a @xmath89 such that for values @xmath90 greater than @xmath91 , the effective potential @xmath88 is monostable with a minimum located at @xmath92 . for @xmath93 ,
the effective potential is bistable with two minima at @xmath94 , being @xmath95 a function of @xmath9 and @xmath90 .
figure [ fig : ueff ] depicts this situation for a system with @xmath96 .
the calculated critical noise for this value of @xmath9 is @xmath97 @xcite . for a system with a finite size @xmath0
, we can calculate numerically @xmath83 by simulating the langevin equations ( [ eq : lang ] ) and computing the histogram of the collective variable @xmath12 after a sufficiently large time when the system has equilibrated .
figure [ fig : ueff ] shows the resulting effective potential @xmath98 for a system with @xmath99 oscillators .
it can be seen that the deviations with respect to the infinite size potential @xmath88 are very small , even for such a small system .
effective potential for the simple langevin model ( [ eq : eff ] ) for two systems with @xmath96 .
top panel corresponds to @xmath100 ( @xmath101 ) and the bottom panel to @xmath102 ( @xmath103 ) .
the solid lines depict the analytical solution given in ref .
@xcite , whereas the dotted lines correspond to the effective potential obtained using the simulation method described in the text for a system with @xmath99 . ]
effective dynamics for a coupled system with @xmath99 oscillators subject to a rectangular driving of amplitude @xmath104 and frequency @xmath105 .
several stochastic resonance quantifiers are depicted vs the noise strength @xmath90 . from top - left to right - bottom : the numerator of the snr ( @xmath106 ) , the denominator ( @xmath107 ) , the output snr ( @xmath26 ) and the gain ( @xmath34 ) .
the solid lines depict the numerical solution of the full langevin dynamics ( [ eq : lang ] ) .
the dotted and dashed lines correspond to the fluctuating cumulants approach truncated at the second ( [ eq : pikovsky ] ) and third order ( [ eq : third ] ) , respectively .
the results for the effective potential approach ( [ eq : eff ] ) are depicted by triangles pointing upwards ( @xmath108 ) and downwards ( @xmath109 ) .
, width=321 ] in this section , we compare numerically the predictions of the effective models presented in sec .
[ sec : eff ] in the framework of sr .
the simplified character of these models will allow us to explain in intuitive terms the highly nonlinear effects observed in the stochastic resonance quantifiers @xcite .
we will restrict our study to a periodic rectangular driving force , @xmath110 where @xmath111 , @xmath112 being the floor function of @xmath113 .
the input snr for forces of this type can be readily calculated as @xmath31 . in all cases reported here the coupling strength
is fixed to @xmath96 , the driving frequency @xmath23 to @xmath105 , and the driving amplitude to @xmath104 .
this amplitude is _ subthreshold _ in the sense that the driving force ( [ eq : driving ] ) can not induce sustained oscillations between the different attractors of the dynamics in the absence of noise ( i.e. , for @xmath114 ) .
the stochastic differential equations presented in the preceding sections were solved numerically by using weak predictor - corrector algorithms of order 2.0 @xcite .
figure [ fig : n10 ] shows several sr quantifiers as a function of the noise strength @xmath90 for a coupled system with @xmath99 oscillators .
a strong amplification of the collective response is observed , with sr gains reaching very large values , especially when compared with uncoupled systems subject to the same input signals ( see ref .
these findings were first reported in @xcite . since the numerator @xmath25 of the snr remains of the same order of magnitude for the range of noise strength values plotted , the large values of the sr gain are mainly due to the reduction of a few orders of magnitude of the denominator @xmath107 , as shown in the top - right panel of fig .
[ fig : n10 ] .
the sr quantifiers obtained with the effective potential model described by eq .
( [ eq : eff ] ) are depicted by triangles in fig .
[ fig : n10 ] .
triangles pointing upwards correspond to the effective potential @xmath108 in the asymptotic limit @xmath1 , whereas triangles pointing downwards correspond to the effective potential @xmath109 computed numerically for a system with @xmath99 . in fig .
[ fig : n10 ] , it can be seen that the later leads to a better agreement than the former for @xmath106 due to the small but appreciable discrepancies observed in fig .
[ fig : ueff ] .
however , no significant improvement is seen in the rest of the quantifiers : @xmath107 , the snr , and the gain . in general , the effective potential approach is able to describe qualitatively the phenomenon , displaying a non - monotonic behavior with a maximum at about the same value of the noise strength @xmath90 than the original system .
nevertheless , quantitatively the agreement is not so good , showing a consistent underestimation of the noise term @xmath107 by roughly a factor of 2 .
the fact that this approach underestimates the fluctuations of the collective variable is easy to understand if one takes into account that the effective potential is a mean - field - like idealization in which the real discrete , more noisy , interaction is replaced by a smoothed potential .
a slightly better quantitative agreement is obtained with the gaussian approximation described by eq .
( [ eq : pikovsky ] ) , i.e. the fluctuating cumulants approach truncated at the second order , which is depicted in fig .
[ fig : n10 ] by dotted lines .
it can be seen that the snr is in better agreement , though the sr gain around the maximum has not been improved significantly overall .
a considerably enhanced agreement is achieved by the third order approach in eq .
( [ eq : third ] ) , which is represented in fig .
[ fig : n10 ] by dashed lines .
it can be seen that this third order approximation slightly underestimates the noise term @xmath107 for large enough values of @xmath90 .
this is what one would expect , because this method neglects the higher order cumulants @xmath67 and @xmath68 in eq .
( [ eq : third ] ) [ see eqs .
( [ eq : hie : cum:2])([eq : hie : cum:3 ] ) ] , which , if present , would increase the fluctuations of the lower order cumulants .
same as in fig .
[ fig : n10 ] but for a system with @xmath115 .
, width=321 ] figure [ fig : n30 ] confirms the above discussed behavior of the simplified models for a larger system with @xmath115 .
again , the effective potential theory and the gaussian approximation provide mainly a qualitative picture , with quantitative predictions within the same order of magnitude .
the best quantitative agreement is also observed to be given by the third - order fluctuating cumulant scheme . the main differences with respect to the smaller system discussed before are in the quantity @xmath106 , which is proportional to the spectral amplification . in this case
the system is large enough so that very small differences are observed between the effective potential @xmath116 and @xmath108 .
they both provide data in very good agreement with the original system data .
in addition , notice that the gaussian approximation data for @xmath106 deviates appreciably from the system data for large enough values of @xmath90 . the fact that the third order approximation leads to a good agreement for @xmath106 indicates that the third cumulant plays an important role for the spectral amplification at these noise strength values . finally , we now use these simplified models to explain the very large gain values observed in globally coupled bistable systems @xcite , and particularly those observed in the bottom - right panels of figs . [ fig : n10 ] and [ fig : n30 ] .
these gain values are much larger than those observed in uncoupled or isolated bistable systems subject to the same rectangular input signals @xcite , and thus are due to the interaction between the oscillators . for the sake of simplicity
we will use the effective potential approximation in the following discussion , though similar arguments can be used within the other schemes presented in sec .
[ sec : eff ] . the only noisy term in eq .
( [ eq : eff ] ) is @xmath53 , which has a strength of @xmath117 .
thus , we can fix the individual noise strength @xmath90 and still being able to get rid of the noise by considering the limit @xmath1 . this way eq .
( [ eq : eff ] ) becomes deterministic and we can apply the concept of static threshold for a finite value of @xmath90 .
a simple analysis of the infinite size potential shows that a constant driving of @xmath104 is able to remove one of the two attractors of the dynamics for systems with noise strength values @xmath90 larger than @xmath118 ( and , of course , smaller than @xmath119 , the noise value where the effective potential turns monostable in the absence of driving ) . therefore , most of the data points in figs .
[ fig : n10 ] and [ fig : n30 ] correspond to _ suprathreshold _ dynamics when viewed from the perspective of the langevin equation ( [ eq : eff ] ) .
the response is expected to be more amplified with suprathreshold signals than with subthreshold signals , because , for the first ones , the presence of noise is not necessary in order to produce jumps between the two locations of the time - dependent attractor .
for instance , only gains larger than unity has been found for an isolated bistable system subject to monochromatic signals when the driving amplitude is suprathreshold @xcite , with gain values reaching a few tenths above unity .
in fact , the above consideration may well explain why gains larger than unity ( also a few tenths above unity ) are found for globally coupled bistable systems subject to monochromatic signals @xcite : the collective variable dynamics is effectively suprathreshold .
subthreshold dynamics .
same stochastic resonance quantifiers as in fig . [
fig : n10 ] but as a function of the system size @xmath0 for a fixed noise strength @xmath120 .
squares depict the simulation data corresponding to the full dynamics ( [ eq : lang ] ) .
lines are a guide to the eye .
, width=321 ] suprathreshold dynamics .
same as in fig .
[ fig : sub ] but for @xmath121 . ,
width=321 ] a numerical analysis of the deterministic version ( i.e. without the noise terms @xmath122 ) of eqs .
( [ eq : third ] ) governing the third order approximation confirms that under a constant driving of @xmath104 , there is a transition at about @xmath123 between a situation in which the system presents two attractors ( subthreshold dynamics ) and only one ( suprathreshold dynamics ) .
in fact , this analysis can also be carried out with the fluctuating cumulant theory presented in sec .
[ sec : eff ] with an arbitrary order of truncation , and thus , with an arbitrary order of accuracy .
let us illustrate the above discussion by considering two systems : one with a noise strength value @xmath120 , which is just below the transition value @xmath124 ; and one with @xmath121 , well above that transition value .
figures [ fig : sub ] and [ fig : supra ] shows the behavior of the stochastic resonance quantifiers as a function of the system size @xmath0 for these two systems under the same rectangular driving with @xmath104 and @xmath105 .
it can be seen that , for large enough @xmath0 , the size of the fluctuations ( @xmath107 ) is always reduced as @xmath0 is increased , as expected
. however , both systems display a very different behavior . in fig .
[ fig : sub ] the sr quantifiers @xmath106 , snr and the gain are observed to decrease monotonically as @xmath0 is increased .
this is what is expected for a system under subthreshold dynamics , because the driving force alone is unable to provoke jumps between the attractors and needs the presence of fluctuations . in the limit @xmath1 ,
the fluctuations , and thus the jumps , are completely suppressed . on the other hand , a very different behavior
is observed in fig .
[ fig : supra ] for a noise strength value above the transition value corresponding to this driving amplitude . in this case
, there is only one attractor , which is displaced to the positive or negative axis around the origin according to the instantaneous sign of the driving force . here
the presence of finite size fluctuations is only a nuisance to the driven oscillations of the collective variable . as a result ,
the spectral amplification and the sr gain grow with @xmath0 toward the finite values which correspond to the infinite system .
finally , let us notice that the phenomenon of _ system size resonance _
@xcite , i.e. the non - monotonic behavior of the sr quantifiers when plotted as a function of @xmath0 , could only appear in a parameter region in which the system dynamics is subthreshold .
nevertheless , this phenomenon is not observed in fig .
[ fig : sub ] .
this is due to the fact that the strength of the fluctuations is very small even for small values of the system size @xmath0 , already below the optimal noise value in which the response is maximized .
any further increase of the system size reduces the fluctuations and , thus , the jumps between the attractors .
starting from the langevin equations defining the model system , we have derived , using it stochastic calculus , a hierarchy of exact stochastic differential equations for a set of fluctuating cumulant variables , defined by using the arithmetic mean over all oscillators . due to the useful mathematical properties of it stochastic calculus , the hierarchy contains noise terms with simple autocorrelation properties .
furthermore , the approach proposed in this paper for the fluctuating cumulants is not directly applicable with stratonovich calculus . in the limit of an infinite number of oscillators , the whole hierarchy reduces to the one obtained by desai and zwanzig @xcite for the expected values of the cumulants .
in contrast to the theory presented in ref.@xcite , the fluctuating cumulant approach allows us to study a wide kind of collective dynamical properties like autocorrelation functions or the snr , in addition to effects due to finite size fluctuations .
nevertheless , the noise terms that appear in the exact hierarchy for the fluctuating cumulant variables depend in a complicated way on the fluctuating cumulants and approximations have to be taken in order to obtain a closed set of stochastic differential equations . here
it is shown that this difficulty can be overcome in the asymptotic limit of a very large number of oscillators . however , one still has the inconvenience of dealing with a hierarchy with an infinite number of equations , and a truncation scheme is desirable .
a gaussian approximation was proposed by pikovsky et al . in ref .
@xcite , and here it is presented as a second - order truncation scheme of the fluctuating cumulant hierarchy .
in addition , an arbritrary - order truncation method is proposed , with explicit expressions given for the third order only .
this third order approach turns out to provide the best quantitative agreement with the sr data .
finally , a rather simple approach based on a single variable and the use of an effective potential is proposed .
the spectral amplification of the collective variable as a function of the noise strength @xmath90 of systems with @xmath99 and @xmath115 bistable oscillators is found to be in good agreement with the predictions given by all the approximations , though small but appreciable systematic deviations are observed for the gaussian approximation for a system with @xmath115 oscillators . additionally , the effective potential theory and the gaussian approximation do not account well for the snr or the sr gain of the collective variable , though the data is within the same order of magnitude
. a systematic underestimation of the fluctuations of the collective variable is done by the effective potential approach due to the mean - field like character of this simplified theory .
the best quantitative agreement of the snr and the sr gain is given by the third - order fluctuating cumulant theory , although a small systematic underestimation of the fluctuations is still observed with this third order theory due to the neglect of higher order fluctuating cumulants .
furthermore , using any of the approximations presented , we are able to explain the very large gain values observed in refs .
specifically , it is shown that the driving amplitudes used are suprathreshold from the point of view of the effective dynamics in the range of noise strength values utilized in those works , i.e. , there is only one attractor of the dynamics under the presence of the driving force , and this attractor oscillates following the driving force .
simulation results , showing several sr quantifiers as a function of the system size @xmath0 , confirm this behavior .
this situation resembles the effect of a high - frequency signal in an isolated bistable system . in the later case
, the high - frequency signal can be removed from the description by means of an effective bistable potential with modified parameters , with the consequence that previously subthreshold driving amplitudes can become suprathreshold from the point of view of the effective potential @xcite .
in contrast , the effective dynamics induced by the high - frequency signal has been shown to provoke the opposite effect on an excitable system , being able to suppress the excitable character of the system @xcite .
this suggests that much work is needed in order to extend the present analysis to finite sets of coupled excitable systems @xcite .
this research was supported by the direccin general de enseanza superior of spain ( grant no . fis2005 - 02884 ) , the junta de andaluca , and the juan de la cierva program of the ministerio de ciencia y tecnologa .
the author thanks jess casado - pascual for helpful discussions .
in the notation commonly utilized within the framework of it calculus , eq .
( [ eq : mukmukp ] ) can be expressed as @xmath125 where @xmath126 . to prove these equations we start from the definition ( [ eq : muk : def ] ) to arrive at @xmath127^k{{\mathrm d}}b_i(t),\ ] ]
where @xmath128 .
thus @xmath129^k[y_j(t)]^l{{\mathrm d}}b_i(t ) { { \mathrm d}}b_j(t ) \nonumber\\ & = & \frac{2d}{n^2}\sum_{i}[y_i(t)]^{k+l}dt \nonumber \\ & = & \frac{2d}{n}m_{k+l}(t)dt,\end{aligned}\ ] ] where we have used ( [ eq : wienner ] ) .
similarly , using the fact that the wiener processes @xmath44 have independent increments , i.e. @xmath130 for @xmath131 , and that the increments @xmath132 are independent of @xmath133 at times @xmath134 , eq .
( [ eq : apen1:gammak ] ) is readily proven . in order to clarify the advantages of it calculus in the context of this paper ,
let us now compute the autocorrelation of @xmath69 by using stratonovich calculus . in this case
we are entitled to utilize the usual rules of differentiation of deterministic calculus .
to that aim , the novikov - furutsu theorem @xcite states that if @xmath135 is a gaussian white noise with zero mean and autocorrelation @xmath136 , then for any functional @xmath137 $ ] we have @xmath138\rangle=\int { { \mathrm d}}s \langle \xi(t)\xi(s)\rangle \langle \frac{\delta g[\xi]}{\delta \xi(s)}\rangle=2d\langle\frac{\delta g[\xi]}{\delta \xi(t)}\rangle,\ ] ] where @xmath139/\delta \xi(t)$ ] denotes the functional derivative of @xmath137 $ ] .
thus , assuming @xmath140 , @xmath141}{\delta\xi_j(t^\prime)}\rangle \nonumber \\ & = & \frac{2d}{n^2}\sum_{i , j}\big\ { \langle \xi_i(t ) y_i(t)^k \frac{\delta[y_j(t^\prime)^l ] } { \delta\xi_j(t^\prime)}\rangle+ \nonumber\\ & & \quad \langle \xi_i(t ) y_j(t^\prime)^l \frac{\delta[y_i(t)^k ] } { \delta\xi_j(t^\prime)}\rangle+ \nonumber\\ & & \quad \langle y_i(t)^k y_j(t^\prime)^l \frac{\delta\xi_i(t)}{\delta\xi_j(t^\prime)}\rangle \big\}. \label{eq : muapen}\end{aligned}\ ] ] using repeatedly the novikov - furutsu theorem and the fact that @xmath142 , we arrive at eq .
( [ eq : mukmukp ] ) plus the following two extra terms on the righthand side of eq .
( [ eq : mukmukp ] ) @xmath143 these extra terms make the problem much more difficult to deal with .
in this appendix we show that the process @xmath69 tends to a gaussian behavior as @xmath1 . first note that the third moment of @xmath69 , @xmath144 and all odd moments of @xmath69 vanish .
if @xmath69 were gaussian , all cumulants higher than the second should be zero .
this requires all odd moments of @xmath69 to vanish but also a specific functional form of the even moments @xcite .
for instance , were @xmath69 a gaussian process , the fourth moment @xmath145 would be given by @xmath146 instead , a simple calculation shows that for a finite @xmath0 the fourth moment is @xmath147 .
\label{eq : muk4:2}\end{aligned}\ ] ] therefore , by considering ( [ eq : mukmukp ] ) , it is clear that ( [ eq : muk4:1 ] ) does not equal ( [ eq : muk4:2 ] ) unless @xmath148 this identity is asymptotically correct in the limit @xmath149 , because then all fluctuating moments become deterministic . for a large enough @xmath0 , eq .
( [ eq : m2gauss ] ) will hold as a good approximation , showing a gaussian behavior in the lowest order of a @xmath71 expansion .
clearly , similar considerations apply to other even moments of @xmath69 . | the dynamics of a system formed by a finite number @xmath0 of globally coupled bistable oscillators and driven by external forces is studied focusing on a global variable defined as the arithmetic mean of each oscillator variable .
several models based on truncation schemes of a hierarchy of stochastic equations for a set of fluctuating cumulant variables are presented .
this hierarchy is derived using it stochastic calculus , and the noise terms in it are treated using an asymptotic approximation valid for large @xmath0 .
in addition , a simplified one - variable model based on an effective potential is also considered .
these models are tested in the framework of the phenomenon of stochastic resonance . in turn , they are used to explain in simple terms the very large gains recently observed in these finite systems . |
an accurate knowledge of @xmath0 and log @xmath1 allows us to deduce many aspects of a white dwarf s evolution .
currently , the best way of determining both of these quantities is by fitting a calculated model atmosphere to either the hydrogen lyman or balmer line absorption series @xcite .
this is not without it s drawbacks , however , as the shape of the lyman / balmer absorption features is also dependent upon the metal species present in the atmosphere .
therefore , inaccurate atomic data used to calculate model atmospheres may also result in an inaccurate determination of the @xmath0 and log @xmath1 . current calculations of lyman / balmer line profiles also utilise pre - calculated stark broadening tables to reduce calculation time . as with the atomic data ,
if the broadening tables are not accurate enough , then this will also propagate errors to the determined @xmath0 and log @xmath1 .
in an analysis by @xcite , the authors measured @xmath0 and log @xmath1 for a sample of da white dwarfs that had data in the lyman / balmer line regions from orfeus , hut , and the far ultraviolet spectroscopic explorer ( fuse ) in the uv , and ground based telescopes in the optical .
it was shown that the @xmath0 determined from the lyman / balmer line series disgreed for several white dwarfs , however , this was regarded as being due to systematics present in the observational data .
an improved study by @xcite considered a larger data set where all uv data originated from fuse . a uniform pipeline was developed and applied to the uv data in an attempt to reduce any systematics present .
the determined lyman / balmer @xmath0 up to @xmath250,000k was found to be in agreement for both series , however , a statistically significant discrepancy became apparent for stars with @xmath0 greater than this .
this discrepancy has affectionately been dubbed `` the lyman / balmer line problem '' . in this proceeding , we present preliminary results from our study of the lyman / balmer line problem in hot da white dwarfs .
we approach the problem from three perspectives .
the first is the opacity used to calculate model atmospheres .
this involves using atomic data from the @xcite ( ku92 hereafter ) and the @xcite ( ku11 hereafter ) data releases .
the ku92 release contains @xmath3 and @xmath4 fe / ni iv - vi transitions respectively . the second is the broadening tables used to synthesise the lyman / balmer line profiles .
two of these tables are investigated , one calculated by @xcite , and another from @xcite .
the third is the composition of the atmosphere .
we calculate model atmospheres with a pure hydrogen composition , and a metal polluted composition .
seven white dwarfs are chosen that have been observed both in the uv and optical wavebands .
all of these bar wd1254 + 233 ( gd153 ) are thought to be polluted with metals .
two of these stars , namely wd0501 + 527 ( g191-b2b ) , and wd1254 + 233 , are calibration standards @xcite and have well measured @xmath0 .
the remaining five , wd0229 - 481 , wd0556 - 375 , wd0621 - 376 , wd1738 + 669 , and wd2211 - 495 have been observed to have @xmath0 @xmath5k , and have exhibited the lyman / balmer problem .
the uv data for each of these stars is taken from fuse observations . where possible , fuse datasets taken using the lwrs aperture are used as flux losses due to source drift are minimised .
stars with multiple datasets are coadded to improve the signal to noise .
the full list of datasets used in this analysis will be included in a future publication .
optical observations are taken from several sources . in the case of wd0501 + 527 and wd1254 + 233 ,
coadded g430l observations from the space telescope imaging spectrometer ( stis ) aboard the hubble space telescope are used . for the remaining five stars , objects with declinations @xmath6
were observed by the steward observatory , while objects with declinations @xmath7 were observed by the south african astronomical observatory .
full details on the optical observations are given in @xcite .
model atmospheres used in this study are calculated using tlusty @xcite version 201 , and are synthesised with synspec version 49 @xcite .
each grid is calculated with @xmath0 ranging from 35,000 to 100,000k in steps of 2,500k , and log @xmath1 from 6.5 to 9.5 in steps of 0.25 dex .
it has been shown by @xcite that the inclusion of heavy metals in model atmospheres will affect the determined @xmath0 .
this result is investigated in this study by including two atmospheric compositions .
the first is a pure hydrogen composition , and the second is a metal polluted composition using the abundances from @xcite , with c / h=@xmath8 , n / h=@xmath9 , o / h=@xmath10 , al / h=@xmath11 , si / h=@xmath12 , p / h=@xmath13 , s / h=@xmath14 , fe / h=@xmath15 , and ni / h=@xmath16 as number fractions .
@xcite calculated new stark broadening tables , including additional souces of opacity not accounted for in @xcite s tables .
as the line shapes of the lyman / balmer series are affected by @xmath0 , log @xmath1 , and the broadening treatment , it is imperative to calculate the synthetic profiles as accurately as possible . to assess the effects of using either the @xcite and the @xcite calculations , all model grids are synthesised using either of these broadening tables in turn .
the completeness of the atomic data supplied to model atmosphere can be inferred to have an effect upon the determined @xmath0 .
radiative levitation calculations by @xcite were found to be dependent upon the number of transitions included in the calculation .
we investigate this possibility by calculating the metal polluted model atmospheres using either the ku92 dataset , or the ku11 dataset .
the difference between these datasets is the number of energy levels and transitions available for the ions fe / ni iv - vi .
ku92 is supplemented with photoionization ( pi ) cross sections calculated by the opacity project for fe iv - vi , while ni iv - vi is supplemented with pi data calculated using an hydrogenic approximation .
the ku11 data is not accompanied by pi cross section data .
we therefore calculated this data using the atomic collision package autostructure @xcite .
full details of this calculation will be provided in a future publication .
as expected , the largest lyman / balmer line discrepancies occur when pure h grids are used to measure @xmath0 . curiously , however , in the case of wd1254 + 233 , which is known to be a pure h atmosphere star , the discrepancy between the lyman / balmer line @xmath0 is not affected by the metal content , but is resolved upon using the @xcite broadening tables . in the case of the metal rich stars ,
the discrepancies between the lyman / balmer @xmath0 are greatly reduced when using a metal polluted model grid , and are improved further when these grids are synthesised using the @xcite tables . in figure
[ fig : teffres ] , we have plotted the measured @xmath0 using the lyman / balmer lines for the different types of model grid used .
we have also tabulated the @xmath0 determinations in table [ table : results ] .
interestingly , the use of ku11 data in model atmosphere calculations does not appear to have any noticable effect upon the determined @xmath0 , regardless of the stark broadening table employed .
it is for this reason we have omitted the ku11 results from figure [ fig : teffres ] .
the next step to try and improve the agreement between the lyman / balmer is to allow variations in the metal abundances .
this will be done by calculating a set of model grids for different multiples of the g191-b2b abundances used in this proceeding .
we also plan to include more stars in our sample .
lllllll & lemke & & & tremblay & & + wd name & pure h & ku92 & ku11 & pure h & ku92 & ku11 + wd0229 - 481 & @xmath17 & @xmath18 & @xmath19 & @xmath20 & @xmath21 & @xmath22 + & @xmath23 & @xmath24 & @xmath25 & @xmath26 & @xmath27 & @xmath28 + wd0501 + 527 & @xmath29 & @xmath30 & @xmath31 & @xmath32 & @xmath33 & @xmath34 + & @xmath35 & @xmath36 & @xmath37 & @xmath38 & @xmath39 & @xmath40 + wd0556 - 375 & @xmath41 & @xmath42 & @xmath43 & @xmath44 & @xmath45 & @xmath46 + & @xmath47 & @xmath48 & @xmath49 & @xmath50 & @xmath51 & @xmath52 + wd0621 - 376 & @xmath53 & @xmath54 & @xmath55 & @xmath56 & @xmath57 & @xmath58 + & @xmath59 & @xmath60 & @xmath61 & @xmath62 & @xmath63 & @xmath64 + wd1254 + 233 & @xmath65 & @xmath66 & @xmath67 & @xmath68 & @xmath69 & @xmath70 + & @xmath71 & @xmath72 & @xmath73 & @xmath74 & @xmath75 & @xmath76 + wd1738 + 669 & @xmath77 & @xmath78 & @xmath79 & @xmath80 & @xmath81 & @xmath82 + & @xmath83 & @xmath84 & @xmath85 & @xmath86 & @xmath87 & @xmath88 + wd2211 - 495 & @xmath89 & @xmath90 & @xmath91 & @xmath92 & @xmath93 & @xmath94 + & @xmath95 & @xmath96 & @xmath97 & @xmath98 & @xmath99 & @xmath100 + [ table : results ]
we have presented a preliminary report on our investigation into the lyman / balmer line problem from the perspective of the opacity used to calculate the model atmosphere , the stark broadening tables used to synthesise the lyman / balmer line profiles , and the composition of the atmosphere specified in the model .
the choice of atomic data , either ku92 or ku11 is inconsequential , and the largest improvement is observed when using the tremblay tables with metal polluted atmospheres .
fuse data was obtained from the mikulski archive for space telescopes ( mast ) .
spp and mab acknowledge the support of the science and technology facilities council ( stfc , uk ) .
jbh and ih acknowledge support from the nasa grant nng056gc46 g . | arguably , the best method for determining the effective temperature ( @xmath0 ) and surface gravity ( log @xmath1 ) of a da white dwarf is by fitting the hydrogen lyman and balmer absorption features .
however , as has been shown for white dwarfs with @xmath0>50,000k , the calculated value from the lyman and balmer lines are discrepant , which worsens with increasing temperature .
many different solutions have been suggested , ranging from the input physics used to calculate the models , to interstellar reddening .
we will focus on the former , and consider three variables .
the first is the atomic data used , namely the number of transitions included in line blanketing treatments and the photoionization cross sections .
the second is the stark broadening treatment used to synthesise the lyman and balmer line profiles , namely the calculations performed by @xcite and @xcite . finally , the third is the atmospheric content .
the model grids are calculated with a pure h composition , and a metal polluted composition using the abundances of @xcite .
we present the preliminary results of our analysis , whereby we have determined the @xmath0 for a small selection of white dwarfs .
we plan to extend our analysis by allowing metallicity to vary in future model grids . |
virtually all of the different models of radio pulsars ( e. g. , cheng & ruderman 1980 ; arons 1981 ; michel 1991 ; beskin , gurevich & istomin 1993 ) predict a common phenomenon : the presence of polar caps ( pcs ) around the neutron star ( ns ) magnetic poles heated up to x - ray temperatures by the backward accretion of relativistic particles and gamma - quanta from the pulsar magnetosphere . a typical size of the pc is estimated to be close to the radius within which the open magnetic field lines originate from the ns surface , @xmath0 km for the period @xmath1 ms ) .
however , expected pc temperatures , @xmath2 k , and luminosities , @xmath3 erg s@xmath4 , are much less certain and strongly depend on the specific pulsar model .
studying x - ray radiation from the pcs is particularly useful to discriminate between different models .
the best candidates for the investigation of the pc radiation are nearby , old pulsars of ages @xmath5 yr , including very old millisecond pulsars , for which the ns surface outside the pcs is expected to be so cold , @xmath6 k , that its thermal radiation is negligibly faint in the soft x - ray range . indeed
, available observational data allow one to assume the pc origin of soft x - rays detected from , e. g. , psr b1929 + 10 ( yancopoulos , hamilton & helfand 1994 ; wang & halpern 1997 ) , b0950 + 08 ( manning & willmore 1994 ; wang & halpern 1997 ) , and j04374715 ( becker & trmper 1993 ) . moreover , there are some indications that hard tails of the x - ray spectra of younger pulsars , b0656 + 14 and b105552 , may contain a thermal pc component ( greiveldinger et al .
1996 ) . on the other hand , nonthermal ( e. g. , magnetospheric ) x - ray radiation may dominate even in very old pulsars ( cf . becker & trmper 1997 ) . for example , @xmath7 observations of the millisecond pulsar b182124 ( saito et al .
1997 ) , whose luminosity in the 0.510 kev range exceeds that predicted by pc models by a few orders of magnitude ( see discussion in zavlin & pavlov 1997 ; hereafter zp97 ) , proved its x - rays to be of a magnetospheric origin .
thus , thorough investigations are needed in each specific case to ascertain the nature of the observed radiation .
the closest known millisecond pulsar j04374715 ( @xmath8 ms , @xmath9 yr , @xmath10 erg s@xmath4 , @xmath11 g , @xmath12 pc ) is of special interest .
important data from this pulsar have been collected with the @xmath13 ( becker & trmper 1993 ; becker et al .
1997 ) , @xmath14 ( edelstein , foster & bowyer 1995 ; halpern , martin & marshall 1996 ) and @xmath7 ( kawai , tamura & saito 1996 ) space observatories .
becker & trmper ( 1993 ) and halpern et al . ( 1996 ) showed that the spectral data can be fitted with power - law or blackbody plus power - law models .
the latter fit indicates that the x - ray emission may be , at least partly , of a thermal ( pc ) origin .
rajagopal & romani ( 1996 ) applied more realistic models of radiation emitted by ns atmospheres to fitting the thermal component and concluded that the iron atmospheres do not fit the observed spectrum .
the observations made with the @xmath13 and @xmath14 missions revealed smooth pulsations of soft x - rays with the pulsed fraction @xmath15 apparently growing with photon energy in the 0.12.4 kev @xmath13 range . since the pulsed fraction of the nonthermal radiation is not expected to vary significantly in the narrow energy range , it is natural to attribute the observed radiation to the pulsar pcs .
the pc radiation should be inevitably pulsed unless the rotation axis coincides with either the line of sight or magnetic axis .
if it were the blackbody ( isotropic ) radiation , the pulsed fraction would remain the same at all photon energies .
the energy dependence of @xmath16 can be caused by anisotropy ( limb - darkening ) of thermal radiation emitted from ns atmospheres ( pavlov et al .
1994 ; zavlin , pavlov & shibanov 1996 ) .
zavlin et al . (
1996 ) showed that even in the case of low magnetic fields characteristic to millisecond pulsars ( @xmath17 g ) the anisotropy strongly depends on photon energy and chemical composition of ns surface layers .
to interpret the @xmath13 and @xmath14 observations of , zp97 applied the ns atmosphere models with account for the energy - dependent limb - darkening and the effects of gravitational redshift and bending of photon trajectories ( zavlin , shibanov & pavlov 1995 ) . assuming the viewing angle ( between the rotation axis and line of sight ) , @xmath18 , and the magnetic inclination ( angle between the magnetic and rotation axes ) , @xmath19 , inferred by manchester & johnston ( 1995 ) from the phase dependence of the position angle of the radio polarization
, zp97 showed that both the spectra and the light curves ( pulse profiles ) of the _ entire _ soft x - ray radiation detected by @xmath13 and @xmath14 can be interpreted as thermal radiation from two hydrogen - covered pcs , whereas neither the blackbody nor iron atmosphere models fit the observations .
the approach of zp97 differs substantially from that of rajagopal & romani ( 1996 ) who did not take into account the energy - dependent anisotropy of the emergent radiation and gravitational bending ( hence , they could not analyze the pulse profiles ) , made the unrealistic assumption that pulsed and unpulsed flux components are of completely separate origin , and consequently obtained quite different parameters of the radiating region .
the simplest , single - temperature pc model of zp97 provides a satisfactory fit with typical pc radius @xmath20 km ( reasonably close to the theoretical estimate of @xmath21 km ) and temperature @xmath22 k. the corresponding bolometric luminosity of the two pcs , @xmath23 erg s@xmath4 , comprises @xmath24 of the pulsar total energy loss @xmath25 .
this value of @xmath26 is in excellent agreement with the predictions of the slot - gap pulsar model by arons ( 1981 ) .
even better fit to the observational data is provided by a model with non - uniform temperature distribution along the pc surface .
in addition , the inferred interstellar hydrogen column density towards the , @xmath27 @xmath28 was demonstrated to be well consistent with the ism properties obtained from observations of other stars in the vicinity of the pulsar .
all these results allow one to conclude that the x - ray radiation observed from is indeed of the thermal ( pc ) origin .
the results of zp97 were obtained for fixed orientations of the rotation and magnetic axes ( angles @xmath29 and @xmath30 ) and for standard ns mass @xmath31 and radius @xmath32 km .
the inferred pc radius , temperature and luminosity are almost insensitive to these four parameters .
their main effect is on the shape and pulsed fraction of the light curves .
for instance , @xmath16 decreases with increasing the mass - to - radius ratio , unless the observer can see the pc in the center of the back hemisphere of the ns , which is possible at @xmath33 and @xmath34 , where @xmath35 and @xmath36 km@xmath37 ( e. g. , zavlin et al .
the angles @xmath30 and @xmath29 can not be precisely evaluated from the radio polarization measurements because of the complicated variation of the polarization position angle across the eight - component mean radio pulse of ( manchester & johnston 1995 ) , and the true ns mass and radius may differ from the canonical values . on the other hand , the fact that the set of angles and @xmath38 adopted by zp97 fits the data does not mean that a better fit can not be obtained for another set , and it tells us nothing about the allowed domain of these parameters .
hence , fitting the light curves for variable @xmath38 , @xmath30 , and @xmath29 enables one to constrain the mass - to - radius ratio and/or the magnetic inclination and viewing angle , so that the present paper is complementary to zp97 .
we describe our approach in 2 and present the results on in 3 .
since the shape of the light curves depends on energy , the light curve fitting is coupled to the spectral fitting .
we fit the count rate spectrum ( total @xmath39 counts ) collected by the @xmath13 position sensitive proportional counter ( pspc ) with the phase - integrated model spectrum emitted from two identical , uniformly heated pcs @xmath40 apart , assuming the hydrogen composition of the surface layers ( zp97 ) .
the fitting is carried out on a grid of angles @xmath29 and @xmath30 between @xmath41 and @xmath42 at different mass - to - radius ratios @xmath43 in a range allowed by equations of state of the superdense ns matter . as shown in zp97 , both the @xmath13 and @xmath14 observations of are consistent with the applied model at the interstellar hydrogen column density of @xmath44 @xmath28 ; therefore we freeze @xmath45 at this value and obtain @xmath46 and @xmath47 from the spectral fits for each set of @xmath29 , @xmath30 and @xmath43 . with these @xmath46 and @xmath47
, we calculate the model spectral fluxes for various phases of the pulsar period and fold each of the spectra with the pspc response matrix ; this gives us the model light curve as a function of phase @xmath48 for given @xmath29 , @xmath30 and @xmath43 .
this light curve is then compared with the observed pspc light curve . for this comparison
, we used the pspc light curve for the total @xmath13 energy range , 0.12.4 kev ; its pulsed fraction ( for 17 phase bins ) was determined to be @xmath49 ( zp97 ) .
we bin the model light curve to the same number of phase bins ( @xmath50 ) and calculate the @xmath51 value , @xmath52 for each set of @xmath29 , @xmath30 and @xmath43 ( @xmath53 and @xmath54 are the observed and model numbers of counts in the @xmath55-th bin ) .
this allows us to find the best - fit parameters ( which correspond to the minimum of @xmath51 ) and confidence levels in the parameter space .
figure 1 shows the 68% , 90% and 99% confidence regions for the above - described light curve models in the @xmath29-@xmath30 plane at several values of the mass - to - radius ratio , @xmath56 .
the ratio determines the parameter @xmath57 responsible for the effects of gravitational redshift and bending of photon trajectories ( e. g. , zavlin et al .
for the @xmath43 values in figure 1 , the parameter @xmath58 varies between 0.82 and 0.73 , making visible from 74% to 91% of the whole ns surface , so that a distant observer can detect the radiation from both pcs simultaneously during almost the whole pulsar period .
the plots in figure 1 are clearly symmetrical with respect to the transformation @xmath59 because the model light curves depend only on the angle @xmath60 between the observer s direction and the magnetic axis : @xmath61 , where @xmath48 is the rotational phase .
the minimum value of the reduced @xmath51 ( @xmath62 for 17 degrees of freedom ) was obtained at @xmath63 , @xmath64 ( or vice versa ) and @xmath65 ( @xmath66 ) . in figure 1
we also show the lines of constant model pulsed fraction .
the lines for the pulsed fractions compatible with the detected value , @xmath67 , are close to the confidence contours unless @xmath29 and/or @xmath30 are close to @xmath42 ; at these large angles the model light curves have a complicated shape ( e. g. , two maxima per rotational period ) inconsistent with what is observed .
the increase of the mass - to - radius ratio enhances the gravitational bending and leads to a greater contribution from the secondary pc ( that on the back ns hemisphere ) , which suppresses the model pulsations . as a result ,
the allowed regions in figure 1 completely vanish at @xmath68 ( @xmath69 ) ( @xmath70 ) the model pulsations may grow because of appearance of strong narrow peaks from pc at @xmath71 . however , there are no such peaks in the observed light curve . ] . with decreasing the @xmath43 ratio , the confidence regions shift towards the bottom - left corner of the @xmath29-@xmath30 plane reaching a limiting position at @xmath72 ( @xmath73 ) when the effect of the gravitational bending becomes negligible , and the observer detects radiation only from the primary pc ( on the front hemisphere ) .
figure 1 provides obvious constraints on the pulsar mass - to - radius ratio .
for instance , if there were no observational information about the @xmath29 and @xmath30 values , the only constraint would be @xmath74 , or @xmath75 km , at a 99% confidence level . if , however , we adopt @xmath18 and @xmath19 , as given by manchester & johnston ( 1995 ) , then @xmath76 .
figure 2 shows the corresponding domain in the ns mass - radius diagram , restricted by the @xmath77 dependences for soft ( @xmath78 ) and hard ( ti ) equations of state of the superdense matter ( shapiro & teukolsky 1983 ) .
it follows from this picture , for example , that the radius of a ns with the canonical mass @xmath31 is within the range @xmath79 km .
another set of angles , @xmath80 and @xmath81 , was suggested by gil & krawczyk ( 1997 ) .
this set gets within the 99% confidence region only for very low mass - to - radius ratios , @xmath82 , when the gravitational effects become negligible .
this corresponds to very low masses , @xmath83 at any @xmath84 allowed by the equations of state ( fig . 2 ) .
note that for these angles and @xmath85 the secondary pc remains invisible during the whole pulsar period .
we have demonstrated that the analysis of the soft x - ray radiation emitted by pcs of radio pulsars in terms of ns atmosphere models provides a new tool to constrain the ns mass and radius , and consequently the equation state of the superdense matter in the ns interiors .
the constraints become more stringent if this analysis is combined with complementary data on the pulsar magnetic inclination and viewing angle , as we have shown using the millisecond pulsar j04374715 as an example . in principle , these angles can be inferred from the phase dependence of the radio polarization position angle . however , in the case of j04374715 this dependence is too complicated to be described by a simple rotating vector model of radhakrishnan & cooke ( 1969 ) , so that the inferred angles are very uncertain .
once an adequate model for the radio emission is found , the statistical analysis of the polarization data would result in a domain of allowed angles in the @xmath29-@xmath30 plane , and the constraints should be based upon overlapping of the confidence regions obtained from the x - ray light curve and from the radio polarization data .
this may constrain not only the @xmath38 ratio , but also the pulsar geometry .
the allowed @xmath77 domain can be further restricted if additional information on the ns mass is available .
for instance , since is in a binary system with a white dwarf companion , it is possible to estimate independently an upper limit on the ns mass .
sandhu et al .
( 1997 ) found upper limits on the white dwarf mass , @xmath86 , and on the orbital inclination , @xmath87 , that yields a restriction @xmath88 ( see their fig . 2 ) .
if further observations of the pulsar reduce the limit on @xmath89 , or a lower upper limit on @xmath90 is obtained from white dwarf cooling models , a more stringent constraint on @xmath91 would follow , thus narrowing the allowed domains of the other pulsar parameters ( @xmath84 , @xmath29 and @xmath30 ) .
the above - described analysis for is simplified by the pulsar s low magnetic field which does not affect the properties of the x - ray radiation .
the spectra and , particularly , angular distribution of radiation emerging from strongly magnetized ns atmospheres depends significantly on the magnetic field ( pavlov et al .
thus , a similar , albeit more complicated , analysis of x - ray radiation from pulsars with strong magnetic fields , @xmath92 g ( e. g. , psr b1929 + 10 and 0950 + 08 ) , would enable one to constrain also the magnetic field strength at their magnetic poles .
we thank werner becker for providing us with the @xmath13 pspc light curve .
we are grateful to joachim trmper for stimulating discussions .
the work was partially supported through nasa grant nag5 - 2807 , intas grant 94 - 3834 and dfg - rbrf grant 96 - 02 - 00177 g .
vez acknowledges the max - planck fellowship . | properties of x - ray radiation emitted from the polar caps of a radio pulsar depend not only on the cap temperature , size , and position , but also on the surface chemical composition , magnetic field , and neutron star s mass and radius . fitting the spectra and
the light curves with neutron star atmosphere models enables one to infer these parameters . as an example , we present here results obtained from the analysis of the pulsed x - ray radiation of a nearby millisecond pulsar j04374715 .
in particular , we show that stringent constraints on the mass - to - radius ratio can be obtained if orientations of the magnetic and rotation axes are known , e.g. , from the radio polarization data . |
laser - induced periodic surface structures ( lipss ) or ripples were first reported by birnbaum in 1965 @xcite who attributed them to diffraction of the incident beam in the optical system . nowadays
two other theories are discussed in the context of the background mechanisms of the lipss formation ( 1 ) interference of the incident light with the surface scattered wave @xcite and ( 2 ) self - organization processes on the surface @xcite .
in general , surface roughness allows for the coupling of the incident laser wave with surfaces that is described in the frames of the surface - scattered wave model @xcite . for metallic / metalized surfaces
, it is believed that surface scattered waves have the form of surface plasmon polaritons ( spp ) @xcite which turn to more localized plasmons with growing surface roughness @xcite .
the influence of plasmonics is supported by the fact that , in many cases , the period of the observed lsfl ( low - spatial frequency lipss ) can be predicted in the frames of this model @xcite , and that the orientation of the ripples follows the polarization of the incident light .
recent experiments on double - pulse generation of lipss on silicon surfaces @xcite , as well as the lipss period reduction upon laser irradiation of silicon in water @xcite , are well explained in the frames of the plasmonic theory . however , lipss generated by white light with a coherence length of @xmath01.4@xmath1 m cast some doubt on the purely interference hypothesis @xcite has been made based on the data of fig . 3 in ref .
@xcite with the central wavelength of the white light spectrum @xmath2 nm and the line width @xmath3 nm . ] .
raman measurements reveal that the lipss formation takes place happens in the molten phase @xcite , which allows for hydrodynamic effects in the pattern formation . for multi - pulse irradiation regimes ,
it was concluded that the plasmonic stage , which governs the lipss orientation , does not necessary determine the periodicity of the final pattern due to contribution of the thermocapillary effects @xcite .
the hydrodynamic processes can be even more important for the lipss produced by single laser pulses at relatively high fluences assuming relatively deep melting and strong ablation of material @xcite . additionally , the hydrodynamic theory is supported by the topology of the pattern , which can be transient between the lipss and cell - like structures @xcite , that is typical for the hydrodynamic systems .
the accumulated evidences suggest that , in many situations , the plasmonic mechanism can not completely explain the origin of lipss without assisting of other physical processes , especially for single laser pulses when the lipss emerge without progressive , pulse - by - pulse conversion of the surface roughness spectrum into periodic grating - like structure via gentle ablation @xcite .
the main experimental argument supporting the plasmonic origin of lipss is that the ratio between the period of the observed ripple pattern @xmath4 and the wavelength of the incident light @xmath5 fits the predictions of the plasmonic theory @xcite .
however , to the best of our knowledge , the experimental evidences of this linear dependency are summarized for several wavelengths based on works of several independent groups , which generate the lipss patterns on different samples at different processing conditions such as laser fluence , repetition rate , focusing and pulse number . in this paper
, we report on systematic studies on the period of lipss generated by _ single _ laser pulses of a picosecond laser system at three different wavelengths , of 1064 , 532 , and 355 nm on the same copper sample .
the experiments are described in section 2 . in section 3
, it is shown experimentally that , for the studied conditions , the observed lipss periodicity is considerably overestimated by the plasmonic theory .
possible mechanisms of this disagreement are discussed in section 4 .
the lipss on the surface of a polished copper sample were generated at three different wavelengths with single pulses of a diode - pumped mode - locked picosecond laser ( lumera , hyper rapid 25 , pulse duration @xmath6ps , repetition rate set at 200khz ) .
the laser system was combined with a galvanometric scanner ( scanlab ) equipped with an f - theta lens .
the single pulse irradiation was achieved by high scanning speed of the galvoscanner controlled by a computer .
the sample surface was positioned at the focal plane of the lens .
the laser pulse energy was adjusted by @xmath7-plate and a polarising beam splitter .
the fundamental harmonics 1064 nm , the frequency - doubled 532 nm and the frequency - tripled 355 nm wavelengths were used for the formation of periodic surface structures on cu .
all optical systems were identical , but consisted of elements designed for the corresponding spectral range .
the one and the same scanner was used for the visible and the infrared light .
a representative scheme of the laser setup used is shown in fig .
[ setup ] .
the structures were analyzed by means of optical and scanning electron microscopy .
as one can see from the electron microscope images ( figs .
[ uv]-[ir ] ) , the lipss , which appear after single - pulse laser ablation are not ideally periodic and contain cell - like defects .
these structures are less regular compared to multipulse - generated lipss . as a result , determination of the pattern period by fourier transform of the whole image is ineffective and the lipss periods were determined through the pattern profile measurements .
the number of peaks along the lines perpendicular to the ripples were measured in several defectless spots within the lipss - covered area for each laser wavelength .
the peak fluence was calculated as @xmath8 , where @xmath9 is the beam radius at the @xmath10 of the intensity and @xmath11 the laser pulse energy .
we assume here that @xmath12 , i.e. , that the pulses are nearly gaussian .
in this work , we investigate the wavelength dependence of the periodicity of lipss produced by single picosecond laser pulses , making an attempt to correlate it with the plasmonic theory and gain a better insight into the underlying physical mechanisms .
the plasmonic concept of lipss formation was never systematically tested before for the same metal sample irradiated under the same conditions by laser pulses of different duration .
figures [ uv][ir ] show the electron microscope images of lipss produced by single picosecond laser pulses at different wavelengths . in each figure
, laser fluence corresponds approximately to the middle of the range , in which the lipss can be observed at particular wavelengths .
nm , pulse duration @xmath13ps , peak fluence @xmath14 .
the average lipss period @xmath15 .
two irradiation spots with centers separated by approximately 10@xmath1 m are shown in the figure.,width=453 ] nm , @xmath13ps , @xmath16 .
the average lipss period @xmath17.,width=453 ] nm , @xmath13ps , @xmath18 .
the average lipss period @xmath19.,width=453 ] table [ summary ] summarizes the obtained theoretical and experimental data on the lipss periods .
considerable disagreement of the plasmonic theory with experiment , which increases with laser wavelength and can not be explained by both the statistical and the systematic error , calls for revision of the lipss formation concept .
.summary of the experimental and theoretical results .
@xmath5 is the incident light wavelength .
@xmath20 is the lipss period calculated using the simple spp model ( solid line in fig . [ lcalc ] ) .
@xmath21 is the lipss period calculated with accounting for the surface roughness ( squares in fig .
[ lcalc ] ) .
@xmath22 is the experimentally measured lipss period .
@xmath23 is the measurement uncertainty . [
cols="^,^,^,^",options="header " , ]
in the frames of the plasmonic ( or interference ) model , the spatial period of the lipss on metallic surfaces @xmath4 depends on the wavelength of the incident light @xmath5 as @xcite @xmath24 where the @xmath25 is the wavelength - dependent dielectric constant of metal .
the angle of the laser beam incidence is equal to zero for all experiments reported here .
the lipss period for copper calculated according to eq . with the data on optical properties from @xcite is shown in fig .
[ lcalc ] by solid line . are shown by dots with error bars .
, width=377 ] the simple spp model , expressed by eq . , assumes an ideally - flat sample surface .
interestingly , the calculations based on a more advanced model ( so - called sipe model @xcite ) , which introduces the shape and filling factors to describe randomly rough surfaces , @xmath26 and @xmath27 respectively , coincide with predictions of the simple spp theory , eq .
( 1 ) , with @xmath28 and @xmath29 ( as given in ref .
the lipss period predicted by this model is given in fig .
[ lcalc ] by empty squares .
the experimental data shown by dots with error bars demonstrate that , for the conditions of lipss formation studied here , both the simple and sophisticated models can predict only the overall tendency of lipss period evolution with laser wavelength but can not describe the period values . for stainless steel surfaces , gedvilas et al . @xcite
have also reported periodicities of lipss created by single picosecond pulses , which are considerably smaller than predicted in the frames of the plasmonic theory .
note that equation yields respectively @xmath0350 nm and @xmath01050 nm at 355 and 1064 nm wavelengths for optical properties of iron @xcite .
two possible interpretations of the disagreement are possible : \(1 ) during picosecond laser pulses , the optical constants of metals can considerably change due to swift heating of the conduction electrons followed by heat exchange with the lattice and lattice melting .
although at picosecond irradiation regimes of moderate fluences the non - equilibrium between the electron and lattice subsystems is not so strong as in the case of femtosecond laser pulses ( see , e.g. , @xcite ) , the simulations for the present experimental conditions based on the two - temperature model ( ttm ) @xcite reveal extremely high heating of the lattice already by the middle of the laser pulse .
thus , at @xmath5 = 355 nm , the laser pulse with peak fluence of 0.4j/@xmath30 ( fig . [ uv ] ) corresponding to the fluence in the center of irradiation spot yields electron and lattice temperatures in the excess of 5000k . under the irradiation conditions of figs .
[ greenps ] and [ ir ] , modelling gives lattice heating up to 30 000k .
although simulations of such regimes , when matter experiences heating to extreme states and can behave abnormally @xcite , calls for a more sophisticated modeling , the ttm simulations unambiguously show the tendency for enormous heating of matter .
such a level of heating assumes material melting with possible superheating of the solid phase @xcite , initiation of lattice disintegration already during first picoseconds @xcite , formation of stress and unloading waves in the surface layer of the sample @xcite .
all these processes should lead to the transient changes of the dielectric function , so that the room - temperature optical parameters @xcite are not applicable anymore for estimating of the lipss period .
however , the evolution of optical properties of metallic surfaces dynamically evolving at picosecond timescales is poorly known while the concept of plasmons on such surfaces requires reconsideration .
we keep this explanation as possible , but the detailed analysis of this possibility would require development of a new theory of the material optical response upon interaction between ultrashort laser pulses and metals that is beyond the scope of this paper .
\(2 ) the second possible explanation , on which we concentrate in this paper , is based on the the three - step model of the lipss formation : 1 .
periodically modulated electron temperature can appear due to , e.g. , interference of the incident light and the surface scattered waves or periodic modulation of the incident light which in its turn causes a modulated reflection of the sample surface .
the amplitude modulation in the electron temperature evolves into a high - amplitude modulation in the lattice temperature upon electron - lattice thermalization .
this process was studied in detail in @xcite .
hydrodynamic processes in the molten surface layer relocate material along the surface and form the lipss pattern .
possible hydrodynamic instability mechanisms are discussed in @xcite schematically this model is represented in fig .
[ 3steps ] .
( 0,0)(0,3)node[right]@xmath31 ; ( 0,0)(3,0)node[above]@xmath32 ; plot ( /3,0.3*sin(80*)+2 ) ; ( 4,0)(4,3)node[right]@xmath33 ; ( 4,0)(7,0)node[above]@xmath32 ; plot ( /3
+ 4,1.3*sin(80*)+1.5 ) ; at ( 9.25,2.1 ) lattice temperature modulation @xmath34 surface pattern formation .
efficiency of each step depends on the wavelength of the preceding modulation.,title="fig:",width=94 ] ; at ( 9.25,0.7)hydrodynamic ; at ( 9.25,0.1)instabilities ; ( 1,3.2) .. controls(1,4) .. (2,4 ) ; at ( 3,4 ) lattice temperature modulation @xmath34 surface pattern formation .
efficiency of each step depends on the wavelength of the preceding modulation.,title="fig:",width=75 ] ; ( 4,4) .. controls(5,4) .. (5,3.2 ) ; ( 6,3.2) .. controls(6,4) .. (7,4 ) ; at ( 8,4 ) lattice temperature modulation @xmath34 surface pattern formation . efficiency of each step depends on the wavelength of the preceding modulation.,title="fig:",width=75 ] ; ( 9,4) .. controls(10,4) .. (10,3.2 ) ; we stress that the spectrum of the initial modulation of the electron temperature @xmath31 is rather broad , because the spectrum of the surface roughness is not monochromatic .
moreover , the incident beam intensity is non - uniformly distributed over the irradiation spot on the sample surface ( e.g. , due to the gaussian beam profile ) that can also lead to broadening of the spectrum of the @xmath31 modulation .
the efficiency of each aforementioned step is higher or lower for different spatial periods of the instabilities , thus acting as high - pass or low - pass filters . in the frames of this 3-step model , according to our experimental observations , the complete process of the lipss formation should be a high - pass filter , because it works better for high spacial frequencies of the initial temperature modulation ( in the uv and visible ranges ) , whereas for low periods ( in the ir range ) the observed lipss period is considerably smaller than predicted by the plasmonic theory .
it has been demonstrated @xcite , that the transition between the steps 1 and 2 acts as a low - pass filter , because the growth rate of the lattice temperature modulation with larger spatial periods is higher , than that of the high - frequency temperature modulations .
consequently , in order to explain the experimental observations , the growth rate of the hydrodynamic instability must decrease with the period , i.e. , the instability should work as a high - pass filter ( or a band - pass filter with the central frequency in the short - wave range ) .
our estimating ttm simulations show that , for the three cases considered , the ablation depths are in the range of 100300 nm leaving behind a molten metal layer app .
50 nm thick .
material relocation within this layer induced by hydrodynamic instability , in particular at post - irradiation timescales , should be dependent on many factors , including molten layer depth , melt viscosity and surface tension .
these factors can provide the reasons for the deviation of the lipss period from the initial interference pattern imprinted on the surface by the interference of incident and scattered electromagnetic waves .
in this paper , it is shown that the periodicity of the lipss patterns on metal surfaces at single - pulse irradiation regimes can be explained by the mutual action of the two known mechanisms , laser light interference with the scattered electromagnetic wave , which is responsible for the lipss direction , and consequent relocation of molten material facilitated by the hydrodynamic instability of melt , which finally affects the pattern periodicity .
additional limitations on possible hydrodynamic instabilities , which participate in the melt relocation on the irradiated surface are discussed .
the instability should act as a high - pass filter , i.e. , the growth rate should increase at shorter periods of the initial surface temperature modulations . as an example , the growth rate of the ablative rayleigh - taylor instability is proportional to @xmath35 and , hence , satisfies this condition .
the temperature - modulated ablation efficiency @xcite can also be considered as a possible mechanism of the lipss formation . for a given amplitude of the lattice temperature modulation on the surface ,
the difference between the driving force of this instability and the viscous friction is larger for smaller periods . due to this ,
the final lipss period is to be shifted to shorter wavelengths and this mechanism can be considered as a high - pass filter .
tjyd , yl and nmb acknowledge the support from the european regional development fund and the state budget of the czech republic ( project biatri : cz.02.1.01/0.0/0.0/15@xmath36003/0000445 ) and from the ministry of education , youth and sports of the czech republic ( programs npu i project no . lo1602 , and large research infrastructure project no .
lm2015086 ) .
tjyd also acknowledges funding from the european commission for the marie sklodowska - curie individual fellowship ( project 657424 ) .
s. a. akhmanov , v. i. emelyanov , n. i. koroteev , v. n. seminogov , interaction of powerful laser radiation with the surfaces of semiconductors and metals : nonlinear optical effects and nonlinear optical diagnostics , sov .
uspekhi 28 ( 1985 ) 10841124 .
t. j .- y .
derrien , j. krger , t. e. itina , s. hhm , a. rosenfeld , j. bonse , rippled area formed by surface plasmon polaritons upon femtosecond laser double - pulse irradiation of silicon , opt .
express 21 ( 2013 ) 2964329655 .
t. j .- y .
derrien , r. koter , j. krger , s. hhm , a. rosenfeld , j. bonse , plasmonic formation mechanism of periodic 100-nm - structures upon femtosecond laser irradiation of silicon in water , j. appl .
( 2014 ) 074902 .
j. p. colombier , f. galerie , p. brunet , a. bruyere , f. pigeon , r. stoian , o. parriaux , plasmonic and hydrodynamic effects in ultrafast laser - induced periodic surface structures on metals , j. laser micro / nanoeng . 3 ( 2012 ) 362368 .
m. gedvilas , j. miksys , g. raciukaitis , flexible periodical micro- and nano - structuring of a stainless steel surface using dual - wavelength double - pulse picosecond laser irradiation , rsc adv . 5 ( 2015 ) 7507575080 .
k. k. anoop , a. rubano , r. fittipaldi , x. wang , d. paparo , a. vecchione , l. marrucci , r. bruzzese , s. amoruso , femtosecond laser surface structuring of silicon using optical vortex beams generated by a q - plates , appl .
phys . lett .
104 ( 2014 ) 241604 .
o. varlamova , c. martens , m. ratzke , j. reif , http://ao.osa.org/abstract.cfm?uri=ao-53-31-i10[genesis of femtosecond - induced nanostructures on solid surfaces ] , appl .
. 53 ( 31 ) ( 2014 ) i10i15 .
http://dx.doi.org/10.1364/ao.53.000i10 [ ] .
http://ao.osa.org/abstract.cfm?uri=ao-53-31-i10 j. bonse , a. rosenfeld , j. krger , on the role of surface plasmon polaritons in the formation of laser - induced periodic surface structures upon irradiation of silicon by femtosecond - laser pulses , j. appl .
106 ( 2009 ) 104910 .
s. e. kirkwood , y. y. tsui , r. fedosejevs , a. v. brantov , v. y. bychenkov , experimental and theoretical study of absorption of femtosecond laser pulses in interaction with solid copper targets , phys .
b 79 ( 2009 ) 144120 .
y. levy , t. j .- y .
derrien , n. m. bulgakova , e. l. gurevich , t. mocek , relaxation dynamics of femtosecond - laser - induced temperature modulation on the surfaces of metals and semiconductors , appl .
374 ( 2016 ) 157164 .
b. i. cho , t. ogitsu , k. engelhorn , a. a. correa , y. ping , j. w. lee , l. j. bae , d. prendergast , r. w. falcone , p. a. heimann , measurement of electron - ion relaxation in warm dense copper , sci .
6 ( 2016 ) 18843 .
e. l. gurevich , y. levy , s. v. gurevich , n. m. bulgakova , role of the temperature instabilities for formation of nano - patterns upon single femtosecond laser pulses on gold , phys .
b submitted ( 2016 ) arxiv identifier 1610.02294 . | the physical mechanisms of the laser - induced periodic surface structures ( lipss ) formation are studied in this paper for single - pulse irradiation regimes .
the change in the lipss period with wavelength of incident laser radiation is investigated experimentally , using a picosecond laser system , which provides 7-ps pulses in near - ir , visible , and uv spectral ranges .
the experimental results are compared with predictions made under the assumption that the surface - scattered waves are involved in the lipss formation .
considerable disagreement suggests that hydrodynamic mechanisms can be responsible for the observed pattern periodicity . #
1#2 lipss , ripples , ultrafast laser ablation , self - organization |
the study of critical phenomena has always been one of the most challenging and fascinating topics in physics .
one can give many examples of systems which undergo phase transitions , from familiar cases like the boiling of water in a kettle or the paramagnetic - ferromagnetic transition of iron at the curie temperature , to the more complicated case of the transition from hadronic matter to quark - gluon plasma which is likely to be obtained by high - energy heavy - ion experiments in the coming years .
particularly interesting are the second - order phase transitions , characterized by a continuous variation of the order parameter and a divergent correlation length at the threshold .
already in the 70 s one tried to use percolation theory in order to give a geometrical description of the critical behaviour of dynamical systems undergoing second - order phase transitions @xcite .
percolation has in fact several features in common with such systems : power law behaviour of the variables at criticality , scaling relations , universality .
the critical behaviour of the ising model can indeed be reformulated in terms of percolation theory : magnetization sets in when suitably defined clusters of parallel spins reach the dimensions of the system @xcite .
in particular , the critical exponents for percolation then become equal to the ising exponents .
the paramagnetic - ferromagnetic transition of the ising model is strictly related to the confinement - deconfinement transition in finite temperature @xmath1 gauge theory .
this was conjectured on the basis of effective theories and confirmed by lattice studies @xcite .
the order parameter for @xmath1 is the lattice average of the polyakov loop , and it behaves like the magnetization in the ising model ; in particular , the critical exponents are the same .
these analogies are the basis of this work .
we have tried to see whether it is possible to find a description of critical behaviour in terms of percolation also for the deconfinement transition in @xmath1 gauge theory .
we show that in a lattice regularization which corresponds to the strong coupling limit , both in two and in three space dimensions , the percolation of polyakov loop clusters ( taken to be suitably defined areas of polyakov loops @xmath2 of the same sign ) leads to the correct deconfinement temperature and to the correct critical exponents for the deconfinement .
we have achieved this result by means of computer simulations of finite temperature @xmath1 gauge theory and using standard finite size scaling tecniques to extrapolate the results to the infinite volume limit .
the percolation problem @xcite is easy to formulate : just place randomly pawns on a chessboard .
regions of adjacent pawns form _
clusters_. percolation theory deals with the properties of these clusters when the chessboard is infinitely large .
if one of the clusters spans the chessboard from one side to the opposite one , we say that the cluster _
percolates_. quantitatively , one counts how many pawns belong to each cluster and calculates two quantities : + @xmath3 the _ average cluster size _ s , defined as : @xmath4 here @xmath5 is the number of clusters of size @xmath6 and the sums exclude the percolating cluster ; this number indicates how big on average the clusters are which do not percolate .
+ @xmath3 the _ percolation strength _
p , defined as : @xmath7 by varying the density of our pawns , a kind of phase transition occurs .
we pass from a phase of non - percolation to a phase in which one of the clusters percolates .
the percolation strength p is the _ order parameter _ of this transition : it is zero in the non - percolation phase and is different from zero in the percolation phase .
if we want to study percolation in dynamical systems , for example the ising model , first of all we must define the rule to build up the clusters ( for instance we can join together all nearest - neighbour spins of the same sign ) .
then one has to find out at which temperature we have a spanning cluster .
let us call this temperature @xmath8 .
it turns out that for @xmath9 the percolation variables p and s behave in the following way : @xmath10 a. coniglio and w. klein @xcite demonstrated that for some special definition of cluster , the onset of percolation coincides with the one of magnetization ; besides , @xmath11 and @xmath12 coincide respectively with the magnetization exponent @xmath13 and the susceptibility exponent @xmath14 .
such cluster definition had already been used by fortuin and kasteleyn to show that the partition function of the ising model can be rewritten in purely geometrical terms as a sum over clusters configurations @xcite . according to the fortuin - kasteleyn prescription , two nearest - neighbouring spins of the same sign
belong to the same cluster with a probability @xmath15 ( @xmath16 , j is the ising coupling , t the temperature ) .
the result of coniglio and klein is valid in any space dimension @xmath17 ( @xmath18 ) and it is independent of the lattice geometry ( square , triangular , honeycomb , etc . ) as long as it is homogeneous .
finite temperature @xmath19 gauge theories describe the interactions of systems of mutually interacting gluons in thermal equilibrium ( @xmath20 is the number of colours ) . systems containing quarks together with gluons are certainly more appealing ; such a scenario will be explored experimentally up to very high temperatures by means of high - energy heavy ion collisions .
nevertheless , @xmath19 gauge theories are of theoretical interest ; already at this level one has a transition between confinement to deconfinement , going from a phase of bound state of gluons ( glueballs ) to a phase of free gluons .
the order parameter of the confinement - deconfinement phase transition of @xmath19 gauge theories is the lattice average of the polyakov loop @xmath21 .
it is related to the effective potential @xmath22 of a static ( mass @xmath23 ) quark - antiquark pair put in the gluons medium at temperature @xmath24 at a distance @xmath25 from each other , when @xmath25 is very big :
@xmath26 the polyakov loop is zero in the confined phase ( @xmath27 ) and different from zero in the deconfined phase ( @xmath22 finite ) .
some time ago it was conjectured by b. svetitsky and l. g. yaffe that the critical behaviour of @xmath19 gauge theories has a strong relationship with the one of simple @xmath28 spin models , with which they share a common global symmetry . in particular , in case of second order phase transitions , both models would belong to the same universality class , that is they would have the same set of critical exponents .
one simple test of the svetitsky - yaffe conjecture is provided by the @xmath1 gauge theory : numerical simulations showed that its critical exponents indeed coincide with the ones of the ising model @xcite . in some respect
the lattice configurations of @xmath1 are similar to the ones of the ising model . instead of having a spin variable at each lattice site we have the value of the polyakov loop at that site .
our aim is to see whether it is possible to build clusters of polyakov loops such that their percolation indices ( threshold , exponents ) coincide with the thermal ones . in order to do that
, we have to face two difficulties : \i ) the polyakov loop is not a two - valued variable like the spin in the ising model but a continuous one ; its values vary in a range that , with the normalization convention we use , is @xmath29 $ ] ; \ii ) the @xmath1 lagrangian is not directly a function of the polyakov loop , therefore it is not possible to extract from it the expression of the bond probability that we need to build the clusters like in the ising model . in a recent work
@xcite it was proved that the point i ) is not a problem .
if we take an ising model with continuous instead of two - valued spins , the definition of fortuin - kasteleyn clusters can be generalized by taking as a bond probability between two positive ( or negative ) nearest - neighbour spins @xmath30 and @xmath31 the following expression : @xmath32 the problem ii ) , however , is hard to overcome . in particular , it seems that there is no way to express the lattice lagrangian of @xmath1 in terms of polyakov loops for all lattice regularizations .
therefore we were forced to investigate @xmath1 in a special lattice regularization , which corresponds to the so called strong coupling limit . in this case
it was shown @xcite that the partition function of @xmath1 can be written in a form which , apart from a factor which depends on the group measure , is the partition function of the continuous ising model studied in @xcite .
the coupling @xmath33 of the spin model and the coupling @xmath13 of @xmath1 are related to each other by this relation : @xmath34 ( @xmath35 fixes the temporal lattice regularization ) .
if @xmath1 is approximately a continuous ising model , we can try to use the general definition of clusters of @xcite .
therefore we have defined our clusters as regions of like - sign polyakov loops connected by bonds distributed according to the bond weight @xmath36 ( @xmath37 and @xmath38 are the polyakov loop values at the sites @xmath39 and @xmath40 ) .
for @xmath41 the approximation of @xcite is good and we chose to investigate numerically this special case in two and three space dimensions .
we began our percolation studies performing some test runs for different lattice sizes to check the behaviour of our percolation variables around criticality .
figure 1 shows the average cluster size @xmath42 for @xmath43 in correspondence of different lattice sizes . to get the critical point of the percolation transition we used the method suggested in @xcite . for a given lattice size and a value of @xmath13 we counted how many times
we found a percolating cluster .
this number is successively divided by the total number of configurations at that @xmath13 .
we call this quantity percolation probability .
this variable is directly a scaling function , analogue to the binder cumulant in continuous thermal phase transitions .
figure 2 shows the percolation probability for the 3-dimensional case as a function of @xmath13 for @xmath44 , @xmath45 and @xmath46 . using @xmath47 and the 3-dimensional random percolation exponent @xmath48.,title="fig:",width=298 ] -0.9 cm using @xmath47 and the 3-dimensional random percolation exponent @xmath48.,title="fig:",width=298 ] -0.9 cm 0.4 cm using @xmath47 and the 3-dimensional random percolation exponent @xmath48.,title="fig:",width=298 ] -0.9 cm using @xmath47 and the 3-dimensional random percolation exponent @xmath48.,title="fig:",width=298 ] -0.9 cm the lines cross at the same point within the errors and that restricts further our @xmath13 range for the critical threshold . besides , since the percolation probability is a scaling function , we could already get clear indications about the class of critical exponents of our clusters .
in fact , if one knows the critical point and the exponent @xmath49 , a rescaling of the percolation probability as a function of @xmath50 should give us the same function for each lattice size .
3 and 4 show the rescaled percolation probability for 3-dimensional @xmath1 using @xmath47 and two different values of the exponent @xmath49 , respectively the ising value and the random percolation one .
the figures show clearly a remarkable scaling for the ising exponent and no scaling for the random percolation exponent .
-0.8 cm to evaluate the exponents @xmath13 and @xmath14 we performed high - statistics simulations in the range where the percolation probability curves cross each other .
the number of measurements we took for each value of the coupling varies from 50000 to 100000 .
we used the @xmath51 method @xcite to determine the values of the exponents .
the final results are reported in table 1 and 2 .
the agreement both in two and in three dimensions is good .
.comparison of the thermal and the percolation results for 2 + 1 su(2 ) .
[ cols="<,<,<,<,<",options="header " , ] +
we have shown that the confinement - deconfinement phase transition in finite temperature @xmath1 pure gauge theory can be described as percolation of some suitably defined clusters of polyakov loops of the same sign .
our result is valid only in the strong coupling limit , and in order to make it general a different approach seems to be necessary .
the use of effective theories for @xmath1 may help to solve the problem @xcite . | the deconfinement transition in su(2 ) gauge theory and the magnetization transition in the ising model belong to the same universality class .
the critical behaviour of the ising model can be characterized either as spontaneous breaking of the @xmath0 symmetry of spin states or as percolation of appropriately defined spin clusters .
we show that deconfinement in su(2 ) gauge theory can be specified as percolation of polyakov loop clusters with fortuin - kasteleyn bond weights , leading to the same critical exponents as the conventional order - disorder description based on the polykov loop expectation value . |
a fascinating issue in general relativity and string theory is to study time - dependent brane solutions .
the time evolution of them is intimately related to the d - brane dynamics in string theory and also there are some potential applications to cosmology and black hole physics .
we are concerned here with colliding brane solutions .
the original colliding brane solutions have been found by gibbons _ et al_. @xcite . for some generalizations of the solutions , see .
a remarkable point is that the solutions are time - dependent but still exact and analytic .
the colliding brane solutions are constructed basically by generalizing static brane solutions in supergravities . as a general property ,
the solutions give rise to friedmann - robertson - walker ( frw ) universes typically by regarding a homogeneous and isotropic part of the solution as our four - dimensional spacetime .
also , some colliding solutions may provide black holes in a frw universe when three of the transverse directions to the d - branes are regarded as the three spatial directions of our world .
although the colliding brane solutions have interesting aspects , there are some problems . as a feature of the colliding solutions , for example , a warp factor in the metric
is a linear function of time for a nontrivial dilaton .
hence , even in the case of the fastest expansion , the power is too small to exhibit a realistic expansion law as in the matter dominated era or in the radiation dominated era .
a possible resolution is to include some additional matter fields . as another application of time - dependent solutions
, one may consider condensed matter systems .
an example is a lifshitz point @xcite .
the static solutions dual to a lifshitz point have originally been shown in @xcite .
the embedding of them into string theory has been carried out in @xcite .
there exist d3-brane solutions that give rise to the lifshitz space as a near - horizon geometry @xcite . in the previous work @xcite ,
we have considered a time - dependent generalization of the solution found in @xcite . then dynamical lifshitz - type solutions @xcite have been constructed with a compactification @xcite .
the solutions describe the time evolution from the lifshitz space to an ads space and it may be interpreted as an aging phenomenon in condensed matter systems . in this paper , we first generalize the d3-brane solutions constructed in @xcite to _ multicenter _ d3-brane solutions with
_ multiple _ waves .
then we consider a t - dual picture of the generalized d3-brane solutions ( without the compactification to the lifshitz space ) . with a slight extension ,
we obtain a brane system composed of dynamical f - strings intersecting d2-branes . in the six - dimensional space transverse to both f - strings and d2-branes (
called the overall transverse space ) , the positions of the d2-branes are fixed at some points while f - strings are dynamical and tend to approach each other with time .
in general , a singularity appears before collision .
an exceptional case is that the f - string charges are the same and five directions of the overall transverse space are smeared out .
we also discuss some cosmological aspects of the solutions towards building a realistic cosmological model .
the possible models are classified based on compactifications of the internal space .
all of them give rise to the frw universe with a power - law expansion .
this manuscript is organized as follows . in sec .
[ wave ] , we first present time - dependent multicenter d3-brane solutions with multiple waves in type iib supergravity . in sec .
[ d2 ] , by performing a t - duality , we construct dynamical solutions composed of f - strings intersecting d2-branes in type iia supergravity .
the solutions describe collision of f - strings when the f - string charges are the same and five directions of the six - dimensional transverse space are smeared out . in sec .
[ cos ] we argue cosmological implications of the solutions .
we also give a classification of lower - dimensional theories with compactifications .
section [ conclusion ] is devoted to conclusion and discussion .
let us present time - dependent d3-brane solutions with multicenters and multiple waves in type iib supergravity . we restrict ourselves to the solutions that contain the dilaton @xmath2 , the axion @xmath3 and the self - dual five - form field strength @xmath4 .
then the relevant field equations are given by here @xmath5 is the hodge operator in ten dimensions .
now let us suppose the following ansatz : the two - dimensional metric @xmath6 and the six - dimensional metric @xmath7 depend only on @xmath8 and @xmath9 , respectively
. then @xmath10 and @xmath11 are constant .
the volume two - form @xmath12 is given by the constant @xmath11 is interpreted as the momentum of wave on the d3-branes in our later discussion .
the above ansatz reduces the field equations to the following form : where @xmath13 and @xmath14 denote the laplace operators on y and z spaces , respectively .
from now on , we shall focus upon a special case by imposing the conditions , where @xmath15 , @xmath16 are the two- , six - dimensional euclidean metric , respectively . then , by solving the field equations , @xmath17 and @xmath18 are given by where @xmath19 , @xmath20 , @xmath21 , @xmath22 , @xmath23 and @xmath24 are constant parameters ; @xmath25 is the asymptotic string coupling constant ; @xmath26 is the string length , related to the string tension @xmath27 as @xmath28 ; @xmath29 is the radius of @xmath30 ; and @xmath31 is the volume of y. here @xmath32 are masses of traveling waves located at @xmath33 and @xmath34 are masses of d3-branes located at @xmath35 and the @xmath30 direction is compact with period @xmath36 .
the distances in the six - dimensional space with @xmath37 are defined as the solutions with ( [ d3p : solution - w : eq ] ) and ( [ d3p : solution - h : eq ] ) describe @xmath38-center d3-branes with @xmath39 waves .
the directions along which the d3-branes and the waves extend are listed in table [ d3p ] .
the time - dependent d3-brane solutions found in @xcite correspond to the case with @xmath40 .
it is worth noting a relation to other solutions by considering a special case of the solution with ( [ d3p : solution - w : eq ] ) and ( [ d3p : solution - h : eq ] ) .
let us impose the conditions , then we can obtain simpler solutions given by @xcite here @xmath42 is a constant parameter .
the positions of wave and d3-branes are represented by @xmath43 and @xmath44 , respectively .
then the distance in the two - dimensional space with @xmath45 and @xmath46 is defined as thus the solutions with ( [ d3p : solution - w : eq ] ) and ( [ d3p : solution - h : eq ] ) can be regarded as a generalization of the solution found in @xcite .
we shall consider a t - dual picture of the solutions presented in sec . [ wave ] .
the t - dualized solutions describe a brane system which consists of dynamical f - strings intersecting d2-branes .
we also discuss collision of f - strings and the near - horizon geometries .
the starting point is the d3-brane solutions with ( [ d3p : solution - w : eq ] ) and ( [ d3p : solution - h : eq ] ) . to perform a t - duality along the @xmath30 direction
, we have to set @xmath47 .
hence @xmath48 does not depend on @xmath30 .
the t - duality relations from type iib to type iia are given by @xcite where @xmath30 is the coordinate to which the t - duality is performed .
the indices @xmath49 , @xmath50 and @xmath51 denote the other coordinates . with the relations in ( [ t :
duality : eq ] ) , the type iia metric is given by here @xmath52 is the ten - dimensional metric with the einstein frame
. then @xmath53 , @xmath54 and @xmath55 are the neveu - schwarz - neveu - schwarz two - form , the ramond - ramond ( rr ) one - form and the rr three - form , respectively
. the field strengths of them are defined as @xmath56 , @xmath57 and @xmath58 .
the resulting metric describes a brane system composed of dynamical f - strings intersecting d2-branes .
the directions along which the f - strings and the d2-branes extend are listed in table [ d2f1 ] .
let us check the equations of motion as a consistency check and consider a further generalization .
first of all , the bianchi identities for @xmath59 , @xmath60 and @xmath61 are automatically satisfied due to the t - duality relations in ( [ t : duality : eq ] ) .
then the equation of motion for @xmath59 is decomposed into a set of the five equations as follows : note that the second and the fourth equations in ( [ t : h3 - 1:eq ] ) are trivially satisfied because there is no @xmath30 dependence now .
the equation of motion for @xmath60 is reduced to where @xmath12 and @xmath62 denote the volume forms , defined as now that @xmath63 is assumed , the condition ( [ t : f2:eq ] ) leads to the constraint @xmath41 .
this means that @xmath64 .
this is consistent with the equation of motion for @xmath59 given in ( [ t : h3:eq ] ) . finally ,
from the field equation for @xmath61 , we obtain the following equation : let us suppose that @xmath65 .
then , from ( [ t : h3 - 2:eq ] ) , we obtain the relation , @xmath66 .
the solutions of @xmath18 and @xmath17 are given by , respectively , here the lengths are defined in ( [ d3p : z : eq ] ) , and the volume of five - sphere with the unit radius is given by @xmath67 .
the parameters @xmath22 are constant
. then @xmath32 and @xmath68 are mass parameters of f - strings and d2-branes , respectively . the f - strings ( d2-branes )
are located at @xmath69 ( @xmath35 ) .
note that we allowed the solution @xmath17 ( [ t : hw : eq ] ) to depend on @xmath30 again .
in fact , this generalization is possible .
the second and the fourth equations in ( [ t : h3 - 1:eq ] ) are also satisfied .
thus we will discuss a t - dual picture of the d3-brane solutions with a slight generalization .
let us next see the behavior of the solutions with ( [ t : hw : eq ] ) and ( [ t : h : eq ] ) .
first of all , the solutions are time - dependent .
it is obvious from the expression of the metric . by putting @xmath17 ( [ t : hw : eq ] ) into the metric ( [ t
: metric : eq ] ) , the following metric is obtained : where the function @xmath70 is defined by the time dependence appears through the function @xmath17 .
hence the next task is to study the time evolution of the solutions carefully .
here we shall consider it by focusing upon collision of f - strings . here
, in order to decrease the number of transverse dimensions to d2-f1 brane effectively , let us smear out some of the directions in the z space .
when the number of the smeared direction is given by @xmath71 , the functions @xmath70 and @xmath18 are rewritten as , respectively , hereafter , we will use the smeared solutions .
let us see here the asymptotic behavior of the solutions
. the harmonic function @xmath70 dominates in the limit of @xmath72 ( near a position of f - strings ) and the time dependence can be ignored .
thus the system becomes static . on the other hand , in the limit of @xmath73
, @xmath70 vanishes .
then @xmath17 depends only on time @xmath74 in the far region from f - strings and the resulting metric is given by the metric becomes singular at @xmath75 or @xmath76 .
therefore the spacetime is regular when it is restricted inside the domain specified by the conditions , here the function @xmath70 is defined in ( [ t : w2 - 2:eq ] ) .
the spacetime can not be extended beyond this region because the spacetime evolves into a curvature singularity .
note that the regular spacetime with two f - strings ends up with the singular hypersurfaces .
the dynamics of the spacetime depends on the signature of @xmath20 .
the system with @xmath77 has the time reversal one of @xmath78 . in the following
, we will study the case with @xmath78 .
then the function @xmath17 is positive everywhere for @xmath79 and the spacetime is nonsingular . in the limit of @xmath80 , the solution is approximately described by a time - dependent uniform spacetime ( apart from a position of branes , near which the geometry takes a cylindrical form of infinite throat ) .
let us study the time evolution for @xmath81 .
at the initial time @xmath82 , the spacetime is regular everywhere and has a cylindrical topology near each brane .
as time slightly goes , a curvature singularity appears as @xmath83 .
the singular hypersurface cuts off more and more of the space as time goes further .
when @xmath74 continues to increase , the singular hypersurface eventually splits and surrounds each of the f - string throats individually .
the spatial surface is composed of two isolated throats . for @xmath79 , the time evolution of the spacetime
is the time reversal of @xmath81 . for any values of
fixed @xmath84 in the regular domain , the metric ( [ t : sufrace2:eq ] ) implies that the overall transverse space tends to expand asymptotically like @xmath85 .
thus , the solutions describe static intersecting brane systems composed of f - strings and d2-branes near the positions of the branes . on the other hand , in the far region as @xmath86 ) , the solutions approach frw universes with the power law expansion @xmath85 . the emergence of frw universes is an important feature of the dynamical brane solutions .
we shall argue whether two f - strings can collide or not .
we put the two f - strings at @xmath87 and @xmath88 , where @xmath89 is a constant parameter .
in addition , we suppose that @xmath39 d2-branes are sitting at a point , namely @xmath90 and introduce the total mass of the @xmath39 d2-branes as @xmath91 it is helpful to introduce the following quantity : then the proper distance at @xmath92 between the two f - strings is represented by this is a monotonically decreasing function of time . here
@xmath42 is the mass of the d2-brane .
the behavior of the proper distance is different depending on the number of the smeared directions @xmath71 .
we shall consider it for each of the values of @xmath71 below .
let us consider the case with @xmath93 .
the proper distance is plotted in fig .
[ fig : f1 ] for the cases with @xmath94 ( left ) and @xmath95 ( right ) .
both plots mean that a singularity appears before the proper distance becomes zero .
hence the singularity between two f - strings develops before collision .
the two f - strings approach very slowly , and then the singular hypersurface suddenly appears at a finite distance . after that , the spacetime splits into two isolated brane throats
. therefore one can not see collision of the f - strings in these examples .
for the other cases with @xmath96 and @xmath97 , the result is the same . and
the parameters are taken as @xmath98 , @xmath99 , @xmath100 , @xmath101 , @xmath102 , @xmath103 .
the difference is just the value of @xmath71 .
the result is also the same and a singularity develops before collision .
the plots in ( a ) and ( b ) correspond to @xmath94 and @xmath95 , respectively .
, title="fig : " ] ( -195,100)@xmath104 ( -100,0)@xmath74 ( 5,70)@xmath30 1.5 cm and the parameters are taken as @xmath98 , @xmath99 , @xmath100 , @xmath101 , @xmath102 , @xmath103 .
the difference is just the value of @xmath71 .
the result is also the same and a singularity develops before collision .
the plots in ( a ) and ( b ) correspond to @xmath94 and @xmath95 , respectively .
, title="fig : " ] ( -195,100)@xmath104 ( -100,0)@xmath74 ( 5,70)@xmath30 + ( a ) @xmath94 ( b ) @xmath95 the next is to consider the case with @xmath105 .
we assume that the @xmath84 directions apart from @xmath106 are smeared .
the functions @xmath18 and @xmath17 are linear in @xmath107 .
hence the behavior of the proper distance is different from the previous case .
the metric is given by ( [ t : sufrace : eq ] ) . by choosing @xmath108 , the harmonic functions @xmath18 and @xmath109
are given by let us consider the solutions in the case that one f - string charge @xmath110 is located at @xmath111 and the other @xmath112 at @xmath113 .
the proper distance between the two f - strings is represented by for @xmath78 , the proper length decreases with time . in the case that @xmath114 , a singularity appears again at a certain finite time @xmath115 , while the distance is still finite . here
@xmath116 is defined as @xmath117/2a_1>0\,.\ ] ] this is the same result as the case with @xmath118 .
however , in the same charge case that @xmath119 , the distance indeed vanishes at a certain finite time @xmath115 , where @xmath116 is defined as @xmath120 hence two f - strings can collide completely . by using @xmath116 ,
the proper distance is expressed as for the values as @xmath98 , @xmath99 , @xmath100 , @xmath101 , @xmath102 , @xmath103 and @xmath121 , the proper distance @xmath104 is plotted in fig . [
fig : f1-d2 ] for the two cases ( a ) the same charges @xmath122 and ( b ) different charges @xmath123 , @xmath124 . in the case ( a ) the two f - strings can collide completely , but in the case ( b ) a singularity appears before collision , as we have already seen analytically . , @xmath99 , @xmath100 , @xmath101 , @xmath102 , @xmath103 and @xmath121 .
the difference is the values of the charges . for the plot in ( a ) , @xmath122 .
then the complete collision occurs at finite time @xmath115 . for the plot in ( b ) , @xmath123 and @xmath124 . then a singularity suddenly appears before collision again . , title="fig : " ] ( -190,100)@xmath104 ( -100,0)@xmath74 ( 5,70)@xmath30 1.5 cm , @xmath99 , @xmath100 , @xmath101 , @xmath102 , @xmath103 and @xmath121 . the difference is the values of the charges . for the plot in ( a ) , @xmath122
. then the complete collision occurs at finite time @xmath115 . for the plot in ( b ) , @xmath123 and @xmath124 . then a singularity suddenly appears before collision again .
, title="fig : " ] ( -190,100)@xmath104 ( -100,0)@xmath74 ( 5,70)@xmath30 + ( a ) @xmath125 ( b ) @xmath126 finally we shall see the near - horizon geometry of the solutions .
when all of the f - strings and d2-branes are located at the origin of the z spaces , the solutions are rewritten as here @xmath42 is introduced in ( [ l ] ) and @xmath127 is the total mass of f - strings @xmath128 in the near - horizon region @xmath129 , the dependence on @xmath74 and @xmath130 in ( [ nh : hw : eq ] ) is negligible and hence the metric is reduced to the following form : where @xmath131 has been performed .
the line elements of a two - dimensional ads space ( ads@xmath132 ) and a five - sphere with the unit radius ( @xmath133 ) are given by @xmath134 and @xmath135 , respectively .
thus the metric ( [ resulting ] ) describes a warped product of ads@xmath136 and @xmath133 .
let us consider a simpler case that the spacetime metric and @xmath17 satisfy ( [ d3p : flat2:eq ] ) .
it is convenient to perform a coordinate transformation for @xmath130 and @xmath84 as follows @xcite : here the unit vector @xmath137 parametrizes @xmath133 and satisfies the following conditions : in terms of the new coordinates ( [ nh : trans : eq ] ) , @xmath17 and @xmath18 are rewritten as where @xmath127 is a constant parameter and the f - strings are assumed to be at the origin of the y space and the z space . in the near - horizon region
@xmath129 , the metric is described by @xcite this is a warped product of a two - dimensional lifshitz spacetime with an eight - dimensional internal space . note that the lifshitz part can be rewritten as an ads@xmath132 with a coordinate transformation @xmath138 .
let us consider cosmological aspects of the solution describing dynamical f - strings intersecting d2-branes .
we first study the time dependence of the scale factors in the solutions .
then , by compactifying the extra directions , our dynamical universe is discussed .
first of all , let us see the scale factors in the solutions .
it is helpful to introduce a new time coordinate @xmath139 as where @xmath77 is assumed for simplicity .
the ten - dimensional metric is thus given by where @xmath140 is defined by recall that @xmath71 is the number of smeared dimensions and should satisfy @xmath141 .
here we assume that at least one direction of @xmath9 is not smeared in order to specify the location of our universe in the transverse space .
let us consider a simplified case that @xmath142 .
this geometry is realized in the limit @xmath143 .
then the scale factor of the four - dimensional space is proportional to @xmath144 . for the remaining spaces @xmath145 and @xmath146 , it is proportional to @xmath147 and @xmath144 , respectively .
thus , unfortunately , the solutions do not to give a realistic universe such as an accelerating expansion , a matter dominated era or a radiation dominated era @xcite .
however , there is a possibility that appropriate compactification and smearing of the extra directions may lead to a realistic expansion . in the next subsection
we will discuss this possibility .
the next is to consider some compactification and smearing of the extra directions of the solutions .
first of all , our universe has to be specified in the solutions with ten directions @xmath148 .
the time direction is identified with @xmath74 .
the remaining task is to identify the three spatial directions from the other coordinates . in the usual approach such as the brane - world scenario ,
three spatial directions are supposed to be on the branes . in the present case , however , it does not work because the spatial directions are specified with @xmath30 , @xmath45 and @xmath46 and then the space is not isotropic from the expression of the metric .
thus we have to look for another way to realize an isotropic and homogeneous three - dimensional space in the present solutions .
the remaining choice is to take the three - dimensional from the overall transverse space with @xmath84 .
that is , our universe is spanned by @xmath74 , @xmath149 , @xmath150 and @xmath151 , for example .
the @xmath106 direction is preserved to measure the position of our universe in the overall transverse space .
note that the metric depends on @xmath84 explicitly .
hence we have to smear out @xmath149 , @xmath150 and @xmath151 so as to define our universe .
thus the number of the smeared directions @xmath71 should satisfy the bound @xmath152 .
the remaining degrees of freedom to be smeared out are @xmath153 and @xmath154 .
now our universe is described by the coordinates @xmath155 .
hence it is convenient to decompose the ten - dimensional metric of the solutions into the following form : where each part of the metric is given by here @xmath156 is the metric of the four - dimensional spacetime with @xmath157 @xmath158 .
the space @xmath159 is one - dimensional with @xmath30 . the space @xmath146 is described by @xmath8 and and the space @xmath160 is three - dimensional with @xmath161 .
then @xmath162 , @xmath15 , @xmath163 are the three- , two- and three - dimensional euclidean metrics , respectively .
the next task is to derive the lower - dimensional effective theory by compactifying the extra directions .
it is necessary to take that @xmath164 in ( [ cos : metric3:eq ] ) to compactify the @xmath30 and @xmath130 directions . to find a realistic universe ,
let us compactify @xmath165-dimensional space to be a @xmath166-dimensional torus , where @xmath167 , @xmath168 and @xmath169 are the compactified dimensions for the direction of @xmath30 , @xmath146 and @xmath160 spaces , respectively .
the range of @xmath169 is given by @xmath170 because the @xmath106 direction is preserved to measure the position of the universe in the overall transverse space .
that is , the @xmath153 and @xmath154 directions may be compactified , where the direction to be compactified has to be smeared out before the compactification . on the other hand , for the compactification of the @xmath30 and @xmath130 directions ,
the smearing is not needed because we have set that @xmath164 .
we shall classify the compactifications of the solutions depending on the variety of the internal space .
the remaining noncompact space is referred to as the external space .
then the metric ( [ cos : metric : eq ] ) is recast into the following form : where @xmath171 is the metric of @xmath172-dimensional external spacetime and @xmath173 is the metric of compactified dimensions .
recall here that our four - dimensional spacetime should be described in the einstein frame .
the einstein frame metric is obtained by multiplying the extra factor and the resulting metric , which leads to new exponents , is given by where @xmath174 is the @xmath172-dimensional metric in the einstein frame . here
@xmath175 and @xmath176 are the coordinates of @xmath177- and @xmath178-dimensional relative transverse spaces , respectively .
finally @xmath179 denotes @xmath180-dimensional spaces .
the constant parameters @xmath181 , @xmath182 , @xmath183 and the cosmic time @xmath139 are defined as the power of @xmath184 of the solutions is given by in table [ ta : power ] , we list the power exponent of the fastest expansion of our @xmath172-dimensional universe in the einstein frame . unfortunately
, every exponent in table [ ta : power ] is so small that the solutions do not give rise to our expanding universe . to realize a realistic cosmological model such as in the inflationary scenario
, it would be necessary to add some new ingredients . .
the classification of compactifications of the transverse directions and the maximum value of the power exponent @xmath185 for @xmath184 .
the exponent is measured in the einstein frame . [ cols="^,^,^,^",options="header " , ]
we have constructed gravitational solutions composed of dynamical f - strings intersecting d2-branes in type iia supergravity and studied the time evolution focusing upon the collision of f - strings .
we also have discussed some applications to cosmology .
first , we have obtained time - dependent multicenter d3-brane solutions with multiple waves in type iib supergravity by generalizing the solutions found in @xcite .
then we have obtained the solutions composed of dynamical f - strings intersecting d2 branes by using a t - duality .
the solutions can also be obtained from supersymmetric static solutions by replacing a constant in the harmonic function with a linear function of time .
this procedure is applicable universally to construct a class of time - dependent brane solutions .
it is worth commenting on the time dependence of the solutions .
the solutions admit f - strings to be dynamical , but d2-branes have to be static .
the origin of this restriction is that the ansatz we imposed is too restrictive .
in fact , in recent works on similar systems , both harmonic functions may depend on time @xcite .
we have also discussed collision of f - strings .
two f - strings approach each other with time .
in general , a singularity appears before collision .
an exceptional case is that five directions in the overall transverse space are smeared out .
the dynamical solutions give rise to a singularity when @xmath186 .
the appearance of the singularity indicates the two possibilities : ( 1 ) the ansatz is simply wrong or ( 2 ) just the supergravity approximation is not valid and it is necessary to include some stringy corrections from @xmath187 or @xmath25 .
it is difficult to decide which of them applies to the present case only from the supergravity perspective
. it would be interesting to consider some generalizations to include stringy corrections .
we have also argued cosmological models obtained from the solutions by compactifying the internal space . the resulting frw universes exhibit a power - law expansion .
however , the power of the scale factor is less than @xmath188 and hence it is too small to realize a realistic expansion . to obtain a realistic expanding universe such as an inflationary scenario or matter or a radiation dominated era
, we have to include additional matter fields .
thus it is interesting to try to construct a cosmological model by studying a more complicated setup .
after all , the solutions constructed here do not give a realistic cosmological model .
however , we believe that more elaborated construction along this line would open a new window to realize realistic cosmological expansions in the context of string theory . c. p. burgess , c. nunez , f. quevedo , g. tasinato and i. zavala , `` general brane geometries from scalar potentials : gauged supergravities and accelerating universes , '' jhep * 0308 * ( 2003 ) 056 [ hep - th/0305211 ] . | we present time - dependent exact solutions composed of f - strings intersecting d2-branes in type iia supergravity .
the f - strings tend to approach each other with time in the six - dimensional space transverse to both f - strings and d2-branes .
in general , a singularity appears before collision .
an exceptional case is that the charges are the same and five directions in the transverse space are smeared out .
then we argue some applications of the solutions in building cosmological models .
the possible models are classified based on compactifications of the internal space .
all of them give rise to the friedmann - robertson - walker universe with a power - law expansion .
* dynamical f - strings intersecting d2-branes + in type iia supergravity * + kunihito uzawa@xmath0 and kentaroh yoshida@xmath1 @xmath0_department of physics , school of science and technology , + kwansei gakuin university , sanda , hyogo 669 - 1337 , japan
_ + @xmath1_department of physics , kyoto university + kyoto 606 - 8502 , japan _ |
the cosmic - ray spectrum extends from the sub - gev region to at least @xmath0 gev ( fig .
[ fig1 ] . ) up to @xmath1 gev and somewhat higher , measurements with magnetic spectrometers flown above most of the atmosphere provide good momentum resolution along with identification of the charge and mass of individual primaries .
measurements with calorimeters continue to identify individual primaries to beyond @xmath1 tev , but with larger systematic uncertainties in the energy assignment . in the pev region and beyond ,
the cosmic - ray intensity is too low for direct measurements ; only indirect measurements of air showers from the ground are possible . since the particles are not identified on an event - by - event basis , the energy spectrum derived from measurements of air showers is given as an all - particle " spectrum , in terms of energy per particle rather than energy per nucleon . in the air - shower regime , identification of the primary mass is made in one of several indirect ways on a statistical basis , complicating the search for features in spectra of individual elements .
the gyroradius of a proton in a typical galactic magnetic field of 3 @xmath2gauss is about half a parsec at one pev .
the parsec scale is also typical of the size of structures in the interstellar medium driven by supernova explosions . for particles with higher energy and larger gyroradii , diffusion in the interstellar medium may become less efficient .
estimates of the maximum energy of particles accelerated at supernova shocks are around the same energy .
it is therefore natural to guess that the knee of the cosmic - ray spectrum around 3 pev has something to do with the end ( or at least the beginning of the end ) of the galactic cosmic - ray spectrum .
peters @xcite described the consequences for energy dependence of the primary composition if the spectrum is characterized by a maximum rigidity , @xmath3 , which could be associated either with propagation or with acceleration ( or both ) .
the relation between rigidity and total energy is @xmath4 where @xmath5 is the total energy of a nucleus of charge @xmath6 and mass @xmath7 .
if cosmic rays are classified by energy per particle , as is the case for air shower measurements , then the spectrum should steepen first for protons , then for helium , then for the the cno group etc .
elsewhere @xcite i have called this sequence the `` peters cycle '' .
several air shower measurements show some evidence that the spectrum becomes progressively enriched in heavy nuclei through the knee region .
the clearest evidence for the peters sequence for several groups of nuclei comes from analysis of the kascade experiment @xcite .
data from direct measurements above 100 tev are sparse , leaving a gap that is bridged by emulsion chamber data with low statistics below a pev and by the threshold region of air shower experiments above 100 tev .
apart from the technical problems of low statistics and systematic threshold effects , there is also the problem that air shower experiments do not identify individual primary nuclei .
in addition , the efficiency of ground arrays depends strongly on primary mass in the threshold region .
nevertheless , it is possible to form a fairly consistent picture of the cosmic - ray spectrum up to the knee . and by @xmath8 for clarity ( see text ) .
, title="fig:",width=264 ] and by @xmath8 for clarity ( see text ) .
, title="fig:",width=264 ] figure [ fig2 ] shows several measurements of the primary spectrum through the knee region . up to 100 gev
there are good measurements of the spectra of protons and helium with magnetic spectrometers flown in spacecraft @xcite and high - altitude balloons @xcite . using the ams @xcite and bess @xcite measurements , together with measurements of heavier nuclei at @xmath9 gev / nucleon @xcite ,
i have converted the spectra of protons , helium , cno , ne - si and fe from energy per nucleon to total energy per particle and combined them to give the all particle spectrum at low energy ( shown as the solid line in fig .
[ fig2 ] ) . the highest energy direct measurements in which individual nuclei are identified directly are from emulsion chamber experiments .
measurements of protons and helium from runjob @xcite and jacee @xcite are shown in fig .
preliminary data on protons and helium from the atic thin ionization chamber @xcite ( not shown here ) are consistent with a smooth power - law extrapolaton between the bess and ams spectromter data at low energy and the runjob data above @xmath9 tev .
( the jacee helium data are higher . )
the solid line normalized to the all - particle spectrum at 100 gev and extrapolated to 1 pev with an @xmath10 power law is consistent with the all - particle measurements of grigorov @xcite but about a factor 1.5 above the sum of the runjob spectra at 100 tev .
( using the higher jacee measurement of helium would bring the emulsion chamber measurements into agreement with the @xmath10 extrapolation at 100 tev . )
an external motivation for using the hard extrapolation from the spectrometer measurements comes from the super - k measurement of the flux of atmospheric neutrinos , where the best fit to the data requires an even harder extrapolation of the primary spectrum above 100 gev @xcite .
even with the hard @xmath10 spectrum , the extrapolation of the direct measurements comes in somewhat below the air shower measurements .
moreover , the all - particle spectrum of runjob @xcite is somewhat below the lowest - energy air shower data , so there may be some systematic offset between the direct measurements and the measurements of air showers .
there is also a slight correlation between hardness of the fitted spectrum and primary mass in this energy range @xcite .
there is , however , no sign @xcite of the sharp steepening of the proton spectrum , below @xmath1 tev , which would be the case if the maximum rigidity accessible in most galactic cosmic - ray accelerators were around @xmath1 tv or below as originally estimated for diffusive shock acceleration by expanding supernova remnants @xcite .
although there are systematic differences among the various air shower measurements in the knee region , all show that the spectrum steepens from @xmath11 or slightly harder below @xmath12 pev to @xmath13 above @xmath9 pev .
figure [ fig2]b shows several results .
two of the measurements @xcite show evidence of structure in the knee region , while the others @xcite do not .
the figure also shows the spectra of two of the measurements ( casa - mia and tibet ) with the energy assignments of the latter shifted down by a factor of @xmath14 , which may be taken as an indication of systematic uncertainties in energy assignment .
the shapes of these two spectra agree well with each other .
some authors @xcite have suggested that a single source contributes significantly to the flux of cosmic - rays in the knee region .
the possible offset between the air - shower measurements and the extrapolation of the direct measurements leaves room for such a possibility .
others @xcite argue , however , that the overall smoothness of the spectrum indicates a single population , with a secondary acceleration mechanism boosting some of the cosmic rays accelerated by galactic supernova explosions to higher energy .
several authors @xcite approach the problem from the high - energy end by modelling the highest energy spectrum as arising from a cosmological distribution of sources and subtracting the extra - galactic population from a model of the supernova - accelerated population to see if an extra , high - energy galactic source is required to fill in the energy region above the knee ( population b " @xcite ) .
this is an open question at present .
( for a recent review and further references see @xcite . ) in any case it is interesting to estimate the power required to produce the high - energy end of the galactic cosmic - ray spectrum .
to do so , i assume a model of galactic propagation with diffusion characterized by an equivalent leaky box model with a escape time given by @xmath15 which applies for all energies . with this propagation model ( motivated by lack of anisotropy at high enery ) , the source spectrum at low energy must be @xmath16 to give the observed @xmath10 spectrum inside the galaxy . assuming a maximum rigidity of @xmath12 pv for this low - energy component , with a composition as measured by an emulsion chamber experiment @xcite , the assumed low - energy component can be subtracted from the observed spectrum .
the total power required to account for the observed spectrum up to @xmath12 eev is then @xmath17 erg / sec .
such an estimate is obviousy very model dependent , but not unreasonably large compared to what might be available in individual galactic sources .
what is needed to make progress is a precise knowledge of the energy - dependence of the major groups of nuclei ( p , he , cno , heavy ) through the knee region .
if there are several important groups of sources with different maximum rigidities then there should be a corresponding sequence of peters cycles , perhaps characterized by different compositions .
the transition to extra - galactic cosmic - rays would be characterized by a transition from heavy nuclei ( from the highest galactic source ) to the light component of a cosmological distribution of sources .
a recent summary of composition measurements with air showers in the knee region is given in the rapporteur paper of matthews at the last icrc @xcite ( see also swordy _ et al . _
overall , the evidence suggests a change in composition toward hearvier primares at higher energy as expected .
however , results of different measurements disagree in detail and the picture is not very clear . as noted in the introduction , the best evidence for the sequence of increasingly heavier groups of nuclei comes from the kascade experiment @xcite , which uses the ratio of muons to electrons as a probe of primary composition
. ratio of muons to electrons in the shower front is a classic probe of primary composition . at each interaction of the nucleons in an air shower , roughly 1/3 of the energy not retained by the projectile nucleon is transferred to the electromagnetic component of the cascade via @xmath18 and 2/3 to charged pions .
the charged pions either reinteract or decay depending on their energy and the depth ( and hence the density ) in the atmosphere where they are produced . charged pions that interact contribute further to the electromagnetic component while those that decay feed the muon component of the air shower .
comparing nuclei of the same total energy , charged pions reach the energy at which they can decay earlier in the cascade for heavy primaries than for protons because the initial energy per nucleon is lower by @xmath19 . as a consequence , the ratio of the muonic to the electromagnetic component of an air shower is larger for heavy primaries . to measure
this ratio with greatest sensitivity requires an observation near shower maximum so that the size of the shower is well correlated with the primary energy . in general , showers in the knee region and somewhat above are observed after shower maximum so that fluctuations in the relation between observed shower size and primary energy are important .
kascade is a surface array near sea level so that these fluctuations are large .
muons detected in kascade typically have energies of a few gev .
an alternative , realized with eastop - macro @xcite and with spase - amanda @xcite , is to sample the muon component with a deep underground detector and the electromagnetic component with an array on the surface above .
such a setup selects muons with sufficiently high energy at production to penetrate to the deep detector .
the high - energy muons generally come more from the fragmentation region of phase space of the hadronic interactions in the shower , whereas than the low - energy muons reflect more the less well - understood central region . on the other hand ,
the multiplicity of high - energy muons is small so that fluctuations in muon number are more important .
some of these points are illustrated in fig .
[ fig3 ] which compares @xmath20 vs. @xmath21 for high - energy ( @xmath22 tev ) muons and for low - energy muons in @xmath23 and @xmath24 ev showers simulated with corsika . in general , the mass resolution is somewhat better for the low energy muons provided that the sampling is good enough to get a good measure of the muon number in individual showers , while the energy resolution is better for coincident events with high energy muons , particularly if the surface array is at high altitude .
this complementarity was pointed out by ralph engel @xcite .
it is interesting that analyses of both eastop - macro @xcite and spase - amanda @xcite suggest an increase in the mean primary mass in the decade of energy above a pev , that is , through the knee region . as a muon detector
, macro has the advantage of resolving muon tracks , whereas amanda reconstructs the light pool generated by the bundle , which gives a measure of the energy deposition of the muons .
a limitation of both experiments is the relatively small sampling area of the deep muon detectors , which has to be accounted for in the comparison between data and simulations .
icecube , with its surface component icetop , now under construction at the south pole @xcite will have a much larger acceptance .
engineering data from the first year of operation of icecube with four icetop stations and one string demonstrate the ability of icecube to reconstruct events with few nanosecond accuracy over distances of more than two kilometers @xcite .
since january 2006 icecube has been operating with 16 surface stations and 9 strings of detectors in the ice .
each string contains 60 digital optical modules ( doms ) evenly spaced in the clear ice between 1450 and 2450 meters below the surface .
each icetop station consists of two ice cherenkov tanks separate from each other by 10 meters and located 25 m from the top of the corresponding icecube string .
each tank is instrumented with two of the same doms used in the ice . with a nominal grid spacing of @xmath25 m ,
the acceptance of the current partial array is @xmath26
m@xmath27sr , which allows detection of coincident events approaching @xmath28 ev . the acceptance of the full icecube as a three - dimensional air shower array will be approximately @xmath29 km@xmath27sr , allowing measurement of coincident events up to an eev .
it is scheduled for completion in 2011 and will be operating in the meantime as new detectors are added .
analysis of measurements of energy - dependence of the depth of shower maximum with hires @xcite suggest a change in composition from heavy to light as energy increases from @xmath28 to @xmath30 ev .
shower maximum is deeper in the atmosphere for protons than for heavy nuclei of the same total energy .
the measured depth of shower maximum increases more rapidly than expected from model calculations , indicating an increasing fraction of protons as energy increases .
a similar change was observed in analysis of the fly s eye data @xcite but at a higher energy ( above @xmath30 ev ) .
it was recognized at the time that such a change of composition could be a signature of the transition from a population of galactic cosmic rays to an extra - galactic population . in view of the more recent data , the signature of a transition appears to occur at a lower energy ( below @xmath30 ev ) . a discussion and references to cosmological scenarios in which the transition to extra - galactic cosmic rays occurs at relatively low energy is given by hillas @xcite .
several new experiments are planned or in operation that can explore the energy region from the knee to the eev region to overlap with the threshold region of the giant arrays auger @xcite and ( formerly ) agasa @xcite .
the most advanced of these if kascade - grande @xcite , a large sea - level array that includes the original kascade as a subarray .
icecube is under construction as described in
telescope array @xcite and its low - energy extension , tale @xcite , are under development in utah .
there is a propsed array of atmospheric cherenkov detectors , tunka @xcite , in russia , and use of the radio technique for a large acceptance air shower detector is being explored @xcite .
understanding the transition from a galactic to an extra - galactic population of cosmic rays is an interesting and important goal for the near future .
* acknowledgments * i thank peter niessen for the generating the data for fig .
[ fig3 ] and michael hillas for enlightening discussions .
this research is supported in part by the u.s .
department of energy under de - fg0291 er 40626 .
99 b. peters , _
nuovo cimento _ * xxii * , 800 ( 1961 ) . t. k. gaisser , _ proc . of the nato advanced study institute on neutrinos and explosive events in the universe _
( springer , 2005 ) pp .
t. antoni _ et al .
_ ( kascade ) , _ astropart . phys .
_ * 24 * , 1 ( 2005 ) .
j. alcarez _
et al . _ ( ams ) , _ phys .
b _ * 490 * , 27 and * 494 * , 193 ( 2000 ) .
t. sanuki _
_ ( bess ) , _ ap.j . _ * 545 * , 1135 ( 2000 ) .
m. boezio _
et al . _ ( caprice ) , _ ap.j . _
* 518 * , 457 ( 1999 ) .
j. j. engelmann _
_ , _ astron .
astrophys _ * 233 * 96 ( 1990 ) .
( runjob ) , _ astropart .
phys . _ * 16 * , 13 ( 2001 ) .
k. asakimori _
et al . _ ( jacee ) , _ ap.j . _ * 502 * , 278 ( 1998 ) .
_ , _ phys .
_ * d71 * 112005 ( 2005 ) .
h. s. ahn _
et al . _ ( atic ) , _ proc .
. cosmic ray conf .
( pune ) _ * 3 * 57 ( 2005 ) .
n. l. grigorov _
_ _ yad . fiz . _
* 11 * , 1058 ( 1970 ) and _ proc .
cosmic ray conf .
( hobart ) _ * 2 * , 206 ( 1971 ) .
l. g. sveshnikova _
et al . _ ( runjob ) , _ proc .
. cosmic ray conf .
( pune ) _ * 3 * , 49 ( 2005 ) .
m. ichimura _
et al . _ ( runjob ) , _ proc .
. cosmic ray conf .
( pune ) _ * 3 * , 21 ( 2005 ) .
m. hareyama _
( runjob ) , _ proc .
cosmic ray conf .
( pune ) _ * 3 * , 17 ( 2005 ) .
o. lagage & c. j. cesarsky , _ astron .
astrophys . _ * 125 * , 249 ( 1983 ) .
m. nagano _
et al . _ ( akeno ) , _ j. phys .
g _ * 10 * , 1295 ( 1984 ) . yu .
et al . _ ( msu ) , _ proc .
cosmic ray conf .
( dublin ) _ * 2 * , 85 ( 1991 ) .
m. a. k. glasmacher _
( casa - mia ) _ astropart .
* 10 * , 291 ( 1999 ) .
m. amenomori _
( tibet ) , _ ap.j . _ * 461 * , 408 ( 1996 ) .
f. arqueros _ et al . _
( hegra ) , _ astron .
_ * 359 * , 682 ( 2000 ) .
j. w. fowler _
( casa - blanca ) _ astropart .
phys . _ * 15 * , 49 ( 2001 ) .
a. d. erlykin & a. w. wolfendale , _ j. phys
. g _ * 27 * , 1005 ( 2001 ) .
w. i. axford _ ap.j
. suppl . _ * 90 * , 937 ( 1994 ) .
h. j. vlk & v. n. zirakashvili , _ proc .
. cosmic ray conf .
( tsukuba ) _ * 4 * , 2031 ( 2003 ) .
a. m. hillas , _ proc .
cosmology , galaxy formation and astroparticle physics on the pathway to the ska _ oxford ( 2006 ) .
d. r. bergman , _ et al . _
( hires ) _ proc .
. cosmic ray conf .
( pune ) _ * 7 * , 315 ( 2005 ) .
r. aloisio _ et al . _ , astro - ph/0608219 .
a. m. hillas , _ j. phys .
g : nucl . part .
phys . _ * 31 * r95 ( 2005 ) .
j. matthews , _ proc .
. cosmic ray conf .
( pune ) _ * 10 * , 283 ( 2005 ) .
s. swordy _
_ , _ astropart .
phys . _ * 18 * , 129 ( 2002 ) .
m. aglietta _
et al . _ ,
( eastop - macro ) _ astropart .
phys . _ * 20 * , 641 ( 2004 ) .
j. ahrens _ et al .
_ , ( spase - amanda ) _ astropart . phys . _
* 21 * , 565 ( 2004 ) .
r. engel , _ proc . workshop on physics at the end of the galactic cosmic ray spectrum _
j. ahrens _ et al . _
( icecube ) , _ astropart .
phys . _ * 20 * , 507 ( 2004 ) .
a. achterberg _
et al . _ ( icecube ) , _ astropart .
( to be published ) * * ( 2006 ) ( astro - ph/0604450 ) .
r. u. abassi _
( hires ) , _ astropart .
* 23 * , 157 ( 2005 ) .
d. j. bird _
_ ( fly s eye ) _ phys
lett . _ * 71 * , 3401 ( 1993 ) .
p. sommers _ et al .
_ ( auger ) _ proc .
. cosmic ray conf .
( pune ) _ * 7 * , 387 ( 2005 ) .
m. takeda _
et al . _ ( agasa ) , _ astropart .
* 19 * , 447 ( 2003 ) .
a. haungs _ et al . _
( kascade - grande ) , astro - ph/0508286 .
et al . _ ( telescope array ) , _ proc .
. cosmic ray conf .
( tsukuba ) _ * 2 * , 1025 ( 2003 ) . g.
b. thomson _
_ ( tale ) , _ proc . 29th int
. cosmic ray conf .
( pune ) _ * 8 * , 363 ( 2005 ) .
n.m . budnev _
( tunka ) , _ proc .
cosmic ray conf .
( pune ) _ * 8 * , 255 ( 2005 ) .
h. falcke _
et al . _ ( lopes ) , _ nature _ * 435 * , 313 ( 2005 ) . | several kinds of measurements are combined in an attempt to obtain a consistent estimate of the spectrum and composition of the primary cosmic radiation through the knee region . assuming that the knee is a signal of the high - energy end of a galactic cosmic - ray population , i discuss possible signatures of a transition to an extra - galactic population and how they might be detected . |
random unitary matrices are often used to describe the process of chaotic scattering @xcite , conductance in mesoscopic systems@xcite and statistics of quantum , periodically driven systems ( see @xcite and references therein ) .
they may be defined by circular ensembles of unitary matrices , first considered by dyson @xcite .
he defined circular orthogonal , unitary or symplectic ensembles ( coe , cue and cse ) , which display different transformation properties @xcite .
distribution of matrix elements and their correlations are known for these canonical ensembles @xcite .
our investigations are motivated by many successful applications of the random matrix theory to problems of quantum chaos , i.e. to the description of quantum properties of systems chaotic in the classical limit .
random matrices of three canonical circular ensembles appear to provide quantitatively verifiable predictions concerning statistical properties of quasi - energy spectra , transition amplitudes etc . for quantum chaotic systems @xcite . for systems without generalized time
reversal symmetry one should use cue , while coe consisting of unitary symmetric matrices corresponds to the time reversal invariant systems ( with integer spin ) .
the so - called circular poissonian ensemble ( cpe ) of diagonal unitary matrices with independent unimodular eigenvalues has also found applications for certain classically integrable systems . in this paper
we shall study statistical properties of composed ensembles defined by products of unitary matrices , each drawn with a given probability distribution .
products of matrices come into attention in a natural way when we consider the evolution of kicked systems .
unitary propagators transporting wave functions of such systems over one period of the kicking perturbation are products of `` free '' evolution propagators and unitary transformations corresponding to instantaneous kicks .
moreover , products of two unitary matrices appear also in the theory of chaotic scattering @xcite .
the paper is organized as follows . in section 2
we briefly recall the necessary definitions and introduce notation .
section 3 contains results concerning spectral properties of composed ensembles of random unitary matrices . the paper is completed by concluding remarks . in appendices we review methods of generating random matrices according to the invariant haar measures on the orthogonal and unitary group
circular ensembles of matrices where defined by dyson @xcite as the subsets of the set of unitary matrices .
uniqueness of the ensembles is imposed by introducing measures invariant under appropriate groups of transformations @xcite .
specifically the circular unitary ensemble ( cue ) consists of all unitary matrices with the ( normalized ) haar measure @xmath2 on the unitary group @xmath3 .
the circular orthogonal ensemble ( coe ) is defined on the set @xmath4 of all symmetric unitary matrices @xmath5 by the property of being invariant under all transformations by an arbitrary unitary matrix @xmath6 , @xmath7 where @xmath8 denotes the transposition . the normalized measure on coe
will be denoted by @xmath9 .
eigenvalues of an @xmath10 unitary matrix lie on the unit circle , @xmath11 ; @xmath12 , @xmath13 . the joint probability distribution ( jpd ) of eigenvalues for each ensemble
was given by dyson @xcite @xmath14 where @xmath15 is a normalization constant and @xmath16 equals to @xmath17 and @xmath18 for coe and cue , respectively .
this number is sometimes called repulsion parameter , since it determines the behaviour of levels spacings as @xmath19 for small @xmath20 @xcite .
the above formula with @xmath21 describes spectra of circular poissonian ensemble ( cpe ) of diagonal unitary matrices with @xmath22 independent phases drawn with uniform distribution in @xmath23 .
the set of diagonal matrices will be denoted by @xmath24 and the normalized measure on cpe , ( which is simply the product measure of @xmath22 measures on the unit circle , ) by @xmath25 . for further consideration
we will also need an ensemble of orthogonal matrices with the probability density @xmath26 defined by the ( normalized ) haar measure on the orthogonal group in @xmath22 dimension .
we shall call this ensemble as haar orthogonal ensemble ( hoe ) .
it is invariant with respect to all transformations @xmath27 , where @xmath28 and @xmath29 denote arbitrary orthogonal matrices .
the joint distribution of eigenvalues of this ensemble @xmath30 can be found in the book of girko @xcite and is recalled in the appendix a. in this appendix we propose a method of generating such matrices numerically and study some properties of their spectra .
we are interested in the spectral properties of products of unitary matrices , each pertaining to a given ensemble .
let us introduce a following notation : @xmath31 denotes a diagonal unitary matrix of cpe , @xmath32 denotes a symmetric matrix of coe , @xmath33 represents a unitary matrix of cue and @xmath34 an orthogonal matrix typical to hoe . as usual ,
the symbol @xmath35 represents a product of two concrete matrices @xmath32 and @xmath33 . on
the other hand @xmath36 will denote the composed ensemble of unitary matrices defined as the image of the mapping @xmath37 with the measure induced by this mapping in its image by the product measure @xmath38 on the cartesian product @xmath39 .
indices can be added to any matrix , if needed .
for example @xmath40 denotes a product of two symmetric matrices , while @xmath41 represents the composed ensemble defined as the image of @xmath42 , which is different from the ensemble of squared symmetric matrices @xmath43 obtained by the mapping @xmath44 . composed ensembles , their measures (
? represents an unknown measure ) , and their joint probability distribution of eigenvalues .
apart of ensembles defined in the text , symbol @xmath45 represents an arbitrary ensemble of unitary matrices and @xmath46 denotes an arbitrary positive real number .
[ cols="^,^,^,^,^",options="header " , ] main results of this paper concerning the spectra of products of unitary matrices are collected in table 1 . for convenience
we added also some previously known results .
jpd @xmath47 represents the formula ( [ joint ] ) , which depending on @xmath16 describes properties of all canonical ensembles .
last column of the table gives a reference to the further text .
some items have not been proved rigorously yet , but are based on numerical results .
we shall start the discussion of above results with an important note .
the fact that the joint probability distribution of eigenvalues characteristic to a given composed ensemble is same as , for example , for cue , does not mean at all that the measures of both ensemble are the same .
in other words , if probability measures of two ensembles are equal @xmath48 ) , then the corresponding jpd are the same @xmath49 .
reverse is not true , what explains why composition of ensembles is not transitive .
for example jpd of @xmath32 is the same as for @xmath41 but differs from this for @xmath50 .
detailed remarks and references to the table are collected below .
* a ) * let us consider an arbitrary subset @xmath45 of the unitary group @xmath3 with an arbitrary measure @xmath51 , and the mapping : @xmath52 the product measure @xmath53 in @xmath54 induces a measure in the image of @xmath55 i.e. in @xmath3 . since @xmath2 is left - invariant i.e. invariant with respect to the left multiplication by @xmath56 the same is true for the product measure i.e. @xmath57 is invariant under the transformation @xmath58 .
in consequence also the measure induced on @xmath3 is left - invariant .
there is only one ( normalized ) left - invariant measure on @xmath3 - the haar measure , hence the resulting ensemble @xmath59 is cue .
the cases ( a2-a5 ) from the table are particular examples .
* b ) * since the haar measure on @xmath3 is also right - invariant an analogous reasoning shows that @xmath60 gives the cue ensemble for an arbitrary ensemble of unitary matrices @xmath45 from which the matrices @xmath61 are drawn .
further , since @xmath59 and @xmath60 are cue so is @xmath62 for @xmath63 and @xmath61 from arbitrary ensembles @xmath64 and @xmath65 of unitary matrices ( the case a6 from the table above ) .
* c ) * similar results are valid for diagonal ( or orthogonal ) matrices . we must only substitute in the previous reasoning , cue by cpe ( or the ensemble of orthogonal matrices ) with measures @xmath25 ( or @xmath26 ) and @xmath45 by an arbitrary subset of diagonal ( or orthogonal ) matrices .
the cases d1 and d2 from the table correspond to this situation .
* d ) * it is easy to prove @xcite that the mapping @xmath66 induces in its image ( the full set of symmetric unitary matrices ) the coe measure @xmath9 i.e. in our notation @xmath67coe .
this corresponds to the b2 and b3 in the table . in the latter case
let s observe that @xmath68 , where @xmath69 and @xmath70 denotes an arbitrary diagonal unitary matrix such that @xmath71 .
the mapping @xmath72 induces , according to b ) , @xmath2 in @xmath3 which reduces b3 to b2 with @xmath33 substituted by @xmath73 .
* e ) * let , as previously , @xmath45 denote an arbitrary subset of @xmath3 with an arbitrary probabilistic measure @xmath51 . from a )
it is now clear that the composite mapping @xmath74 where @xmath55 and @xmath75 are given by ( [ f ] ) and ( [ g ] ) @xmath76 induces coe measure @xmath9 in the image @xmath4 .
consider now two following mappings @xmath77 according to the above , @xmath78 induces in its image the measure @xmath79 .
since @xmath80 they induce the same measure in their image @xmath4 and , as a consequence , @xmath81 induces @xmath9 in @xmath4 i.e. in our notation @xmath82coe for @xmath63 from an arbitrary ensemble @xmath45 .
this corresponds to the case b8 in the table and its special forms b6 and b7 . *
f ) * until now we considered the situations where the ensemble obtained by multiplication of matrices coincided with cue , coe or cpe .
our main interest consists , however , in examination of statistical properties of spectra of resulting matrices .
this allows us to investigate more general situations in which products either do not have specific symmetry properties or the induced measure is not equal to @xmath2 , @xmath26 , @xmath9 or @xmath25 . as an example
let us consider the mapping @xmath83 observe that the image of @xmath20 is the whole set @xmath3 .
indeed , it is enough to show that an arbitrary unitary matrix @xmath33 is a product of two symmetric unitary matrices . to this end
lets denote by @xmath6 an arbitrary unitary matrix which diagonalizes @xmath33 ( such a matrix @xmath6 exists since @xmath33 is unitary ) i.e. @xmath84 where @xmath31 is diagonal and unitary .
now take @xmath85 and @xmath86 .
both @xmath87 and @xmath88 are unitary and symmetric and @xmath89 .
nevertheless the measure induced on @xmath3 by the coe measures @xmath90 on @xmath42 is not equal to cue measure @xmath2 .
indeed , for all @xmath91 the matrix @xmath40 is unitary similar to @xmath92 , where @xmath93 is an arbitrary unitary , symmetric matrix such that @xmath94 , ( such a unitary , symmetric @xmath93 exists since @xmath87 is unitary and symmetric ) .
it means that the spectra of @xmath40 and @xmath95 coincide . but from e )
above we know that the mapping @xmath96 induces coe measure @xmath9 in the image @xmath4 for @xmath88 from coe and arbitrary @xmath45 .
it follows that the eigenvalues of @xmath97 are distributed according to ( [ joint ] ) with @xmath98 , which , on one side , proves that the the mapping does not give cue and , on the other side , covers the case b5 from the table .
* g ) * similar reasoning proves the validity of b4 .
indeed , observe that since @xmath99 for some unitary @xmath33 the matrix @xmath100 is unitary similar to to @xmath101 , but from already proven case b3 from the table we know that such multiplication produces coe . *
h ) * a superposition of two spectra has jpd different from canonical @xmath47 .
the two level correlations can be expressed as combination of correlations of both initial spectra ( with rescaled argument ) @xcite , while level spacing distribution may be obtained as a special case of berry robnik distribution @xcite ( for two equal chaotic layers ) . *
@xmath102 ) * conjectures based on numerical results .
conjectures indexed by the same index are equivalent .
random orthogonal matrices where generated as described in appendix a. a modified version of an algorithm for generation of random unitary matrices , first presented in ref .
@xcite , is given in the appendix b. we generated several realizations of discussed products , diagonalized them numerically and compared the level spacing distribution @xmath103 and number variance @xmath104 with known predictions of canonical ensembles @xcite .
our numerical results are valid thus in the limit of large @xmath22 ( practically @xmath105 and larger ) .
we have performed additional cross - checking by repeating calculations ( with similar results ) using random matrices generated out of eigenvectors . in order to verify or reject hypothesis concerning properties of the spectra the long range correlations where found to be more informative than spacing distribution . in fig . 1 we display number variance averaged over spectra of exemplary composed ensembles - ( @xmath106 ) typical of cue , and other ( @xmath107 , @xmath108 ) typical of coe . ( 9,8 ) ( 0.7,10.5 ) * @xmath109 ) * consider composed ensemble defined as a product of @xmath110 matrices , each pertaining to a given ensemble . for large @xmath110
we expect the product to be distributed uniformly with respect to the haar measure , thus displaying the cue - like spectral fluctuations .
this remark obviously hold if at least one matrix belongs to cue ( see ensemble a6 ) . on the other hand ,
it does not hold if all @xmath110 matrices belong to the poissonian ensemble , since their product displays the jpd @xmath111 .
observe that the jpd of the composed ensembles a5 , b5 and d1 can be written as @xmath112 = p[u_{\beta ' } ] \label{betaa}\ ] ] with @xmath113 equal to @xmath114 , and @xmath115 .
the number @xmath16 , characterizing the degree of the level repulsion , ( [ joint ] ) , for ensembles interpolating between cpe and cue may take any real value in @xmath116 $ ] .
( 9,8 ) ( 0.7,10.5 ) in order to investigate , to what extend the formula ( [ betaa ] ) may be generalized , we constructed numerically random unitary matrices pertaining to ensembles interpolating between cpe and cue as described in appendix b. fig .
2 presents level spacing distribution taken of 3000 matrices of size @xmath117 , while the value of the parameter @xmath118 , controlling the transition cpe - cue , is set to @xmath119 .
level spacing distribution of the composed ensemble defined _ via _ product of such two independent matrices is represented by open symbols .
it is closer to the cue prediction and can be approximated by distribution typical of another ensemble with larger value of the control parameter @xmath118 .
in other words , for this family of interpolating ensembles the relation ( [ betaa ] ) seems to hold with @xmath120 being an unknown function of @xmath16 satisfying @xmath121 .
( 9,8 ) ( 0.7,10.5 ) further tests of the long range correlations of the spectra allowed us to support this conjecture .
figure 3 shows the number variance @xmath104 @xcite for simple and composed interpolating ensembles for three values of the control parameter . in every case the spectra of products of two matrices ( full symbols ) are less rigid that the spectra of the simple interpolating ensemble ( open symbols ) .
this property can be understood realizing that such interpolating unitary matrices enjoy band structure , as demonstrated in fig . 4 for an exemplary matrix of size @xmath122 .
vaguely speaking , a product of two band matrices possess a band of a double width , and the spectral properties of composed ensembles are thus closer to these typical of cue .
( 9,9 ) ( 2,-3.0 ) describing quantized physical systems one encounters often a structure of one of the above mentioned composed ensembles . analyzing a concrete physical system we deal with deterministic matrices , so the assumptions concerning randomness of each matrix forming the composed ensemble can not be rigorously fulfilled .
it seems however , that the assumptions concerning randomness are too strong : we provide below examples of quantum systems which are characterized by jpd found for an appropriate composed ensemble , although some composing matrices do not display required properties of presupposed canonical ensembles . to show this one may study the statistical properties of a _ semiclassical ensemble _
, i.e. the properties of several quantum realizations of the same classical system , distinguished only by different values of the relative planck constant ( spin length ) .
let us start the discussion analyzing periodically time - dependent quantum systems
. generally speaking , jpd @xmath123 corresponds to fully chaotic systems with ( generalized ) time reversal invariance , while spectrum characterized by @xmath124 provides an evidence that such a symmetry has been broken .
let us consider the composed ensemble d6 .
a single orthogonal matrices @xmath34 pertaining to hoe does not often appear alone in the theory , nevertheless the compositions @xmath125 are crucial for many important models .
consider an exemplary periodically kicked system described by a hamiltonian @xmath126 .
its free evolution is represented by @xmath127 and the perturbation term can be written as @xmath128 , where @xmath73 is a symmetric operator and @xmath81 is the perturbation strength .
it is natural to represent the system in the eigenbasis of @xmath129 so the unitary matrix @xmath130 is diagonal .
orthogonal rotation @xmath34 allows one to change the basis into eigenbasis of @xmath73 and obtain eigenvalues of @xmath131 .
note that discussed ensemble @xmath132 corresponds just to the floquet operator @xmath133 of such a system .
we can therefore expect that if both operators @xmath134 and @xmath73 sufficiently do not commute ( so as to assure that the transition matrix @xmath34 is generic in sense of @xmath26 ) , than for generic values of the parameters @xmath135 and @xmath81 the operators @xmath136 and @xmath137 are relatively random @xcite and
the system described by floquet operator @xmath138 is chaotic .
in fact the structure ( [ sys1 ] ) is typical to several models for quantum chaos discussed in the literature ( kicked rotator @xcite , kicked top@xcite , kicked harper model @xcite ) . in ensemble
d6 it is assumed that the diagonal matrices @xmath139 are random . in the simplest chaotic kicked top model
@xmath140 , defined by the angular momentum operators @xmath141 acting on @xmath142 dimensional hilbert space as : @xmath143 @xcite , the diagonal matrix @xmath144 reads @xmath145 . due to the factor @xmath146 in the exponent for a generic value of the parameter @xmath81 the diagonal elements of the matrix @xmath144 are pseudorandom @xcite what assures the coe - like spectral fluctuations of the orthogonal top @xmath147 . to observe the @xmath123 jpd of eigenvalues characteristic to the composed ensemble d6 it is therefore sufficient ,
if at least one of the matrices @xmath148 and @xmath144 is pseudorandom . on the other hand , if both diagonal matrices @xmath149 have the structure @xmath150 , the resulting operator @xmath151 does not pertain to coe , what corresponds to the integrability of the kicked top with @xmath152 @xcite . in order to get a cue spectrum
it is necessary to break the time reversal symmetry ( or any generalized antiunitary symmetry)@xcite . as follows from example d8
this can be done by adding additional unitary term generated by a kick - perturbation @xmath153 not commuting with @xmath134 nor with @xmath73 .
this scheme corresponds exactly to the so called unitary kicked top given by @xcite @xmath154 with @xmath155 ( and arbitrary order of unitary factors ) , or cue version of kicked rotator @xcite . according to remark
f ) the systems which can be brought to a symmetric coe -like structure by a similarity transformation display spectra described by @xmath123 jpd
. therefore example b5 represented by @xmath156 leads to coe spectrum , in contrast to example c1 : @xmath157 , for which such a transformation is not possible . in the same spirit
it is sufficient to modify slightly the system ( [ sys2 ] ) into @xmath158 , or @xmath159 , so as it recovers the generalized antiunitary symmetry and its spectrum pertains to coe .
any `` unitary '' top @xmath160 , without time - reversal symmetry , may be artificially made symmetric by adding the same sequences @xmath160 of perturbation in the reverse order .
therefore @xmath161 displays coe like fluctuations of the spectra . mathematical theory of time reversible and irreversible tops
is given in @xcite , while some further examples where numerically studied in @xcite .
a product of two symmetric random matrices @xmath40 arises in the theory of chaotic scattering @xcite .
its spectrum obeys coe statistics , as follows from the example b5 .
the same statistics is characteristic to several versions of quantized baker map @xcite , which is also represented by a product of two symmetric matrices @xmath162 , although both matrices @xmath163 and @xmath164 , defined via fourier matrices , do not show the properties of coe . as a last example let us consider the piecewise affine transformation of the torus , which can be quantized as @xcite @xmath165 .
diagonal matrices , of the type discussed above , @xmath166 are pseudorandom for a generic value of the parameter @xmath167 .
albeit the symmetric fourier matrix @xmath168 is not typical to coe , the structure of @xmath169 resembles the ensemble d12 , and its spectrum confers to the predictions of cue .
let us conclude our paper with the following , summarizing remarks .
various statistical properties of products of random matrices can be interesting when studying quantum chaotic systems influenced by symmetry breaking perturbations .
we showed that using our results we can predict properties of spectra of a large class of periodically driven model systems ( kicked tops ) . from the mathematical point of view
our investigations leave many questions open . not in all cases we were able to calculate the resulting probability distribution of the composite ensembles .
in fact it was possible only in those cases where the distribution coincided with one of the `` classical '' ones ( coe , cue , cpe , hoe ) . in some cases for which we did not find the probability distribution of the ensemble we were nevertheless able to give the corresponding distribution of the eigenvalues , from which the most popular statistical measure of quantum chaotic systems , namely the distribution of neighboring levels , is easily calculable . for some other composed ensembles we provided numerical evidence for their distribution of eigenvalues applying efficient methods of constructing of random unitary ensembles of all canonical ensembles .
further investigation should resolve the problem of the full probability distributions for these composed examples and find analytical arguments for distributions of eigenvalues founded numerically .
we acknowledge financial support by polish committee of scientific research under the grant no .
2p03b 03810 .
the distribution of eigenvalues in the ensemble of random orthogonal matrices can be found in the books of girko @xcite .
we shall give here the relevant result for completeness .
the distribution density of matrices in the ensemble is the ( normalized ) haar measure on the orthogonal group @xmath170 .
the simpler situation occurs for @xmath22 odd . in this case among the eigenvalues there is one ( say @xmath171 ) equal @xmath17 or @xmath172 .
the rest of eigenphases can be grouped into pairs @xmath173 , @xmath174 . with the probability @xmath17 they are not degenerate and distributed independently of eigenvectors .
the joint probability distribution of eigenphases reads @xmath175 where the last argument @xmath176 and the alternative signs in the rest of the formula refer to @xmath177 or @xmath178 . for the slightly more complicated case of @xmath22
even consult the above cited books of girko . in order to generate numerically a random orthogonal matrix typical of hoe we employed a parametrisation of the orthogonal group defined by hurwitz in the classical paper @xcite published exactly one hundred years ago .
an arbitrary @xmath22 dimensional orthogonal matrix @xmath34 may be written as a product of @xmath179 elementary orthogonal rotations in two - dimensional subspaces .
the matrix of such an elementary orthogonal rotation will be denoted by @xmath180 .
the only nonzero elements of @xmath181 are @xmath182 from these transformations one constructs the following @xmath183 composite orthogonal rotations @xmath184 and finally forms the orthogonal transformation @xmath34 as @xmath185 uniform distribution with respect to the haar measure on the orthogonal group is achieved if the generalized euler angles @xmath186 are uniformly distributed in the interval @xmath187 , and the remaining angles @xmath188 ( for @xmath189 ) are taken from the interval @xmath190 $ ] according to the measure @xmath191 @xcite . an alternative way to generate random orthogonal matrices was recently proposed by heiss @xcite .
random orthogonal matrices may also be obtained as eigenvectors of real random symmetric matrices typical to gaussian orthogonal ensemble .
( 9,8 ) ( 0.7,10.5 ) figure 5 presents the level spacing distribution obtained of 20000 random orthogonal matrices obtained by hurwitz parametrisation for @xmath192 .
numerical data suggest that statistical properties of the spectra of random orthogonal matrices are close to the cue predictions .
in our earlier paper @xcite we have also used the hurwitz @xcite parametrisation to generate random unitary matrices .
we present it here in details for completeness of the present paper and since in the text of @xcite a slightly different , not yet verified algorithm appeared ( the numerical calculation , however , were based on the prescription given below ) .
an arbitrary unitary transformation @xmath33 can be composed from elementary unitary transformations in two - dimensional subspaces .
the matrix of such an elementary unitary transformation will be denoted by @xmath193 .
the only nonzero elements of @xmath194 are @xmath195 from the above elementary unitary transformations one constructs the following @xmath183 composite rotations @xmath196 and finally forms the unitary transformation @xmath33 as @xmath197 the angles @xmath198 , and @xmath199 are taken uniformly from the intervals @xmath200 whereas @xmath201 with @xmath202 uniformly distributed in @xmath203 if the parameter @xmath118 is set to unity then the obtained matrix is drawn from the circular unitary ensemble @xcite . in order to obtain a a family of ensembles interpolating between diagonal matrices of cpe and generic unitary matrix typical of cue we construct a product @xmath204 . diagonal matrix @xmath31 is typical of cpe , while the matrix @xmath205 is obtained according the above procedure with real parameter @xmath206 determining the intervals in eq .
( [ intv ] ) and ( [ xi ] ) .
varying the value of this parameter from zero to unity one obtains a continuous interpolation between cpe and cue @xcite .
random unitary matrices may be also constructed by taking @xmath22 eigenvectors of random hermitian matrix pertaining to the gaussian unitary ensemble . in this procedure
one must specify @xmath22 arbitrary phases of each eigenvector .
this method , albeit simple , does not allow to control parameters of the interpolating ensemble as it is possible for the hurwitz algorithm discussed above .
r. blmel and u. smilansky , phys .
* 60 * , 477 ( 1988 ) .
r. a. jalabert , j - l .
pichard , j. phys .
i france , * 5 * , 287 , ( 1995 ) . c. w. j. beenakker , preprint cont - mat/9612179 and _ to appear in rev .
1997_. f. haake , _ quantum signatures of chaos _ , springer , berlin , 1990 .
f. j. dyson , j. math .
phys . * 3 * , 140 ( 1962 ) .
m. l. mehta , _ random matrices _ , ii ed . academic press , new york , 1991 .
p. pereyra and p. a. mello , j. phys .
a * 16 * , 237 ( 1983 ) .
p. a. mello , p. pereyra , and t. h. seligman , ann .
phys.(n.y . ) * 161 * , 254 ( 1985 ) .
p. a. mello , j. phys . * a 23 * , 4061 ( 1990 ) .
e. doron , u. smilanski and a. frenkel , physica d * 50 * , 367 ( 1991 ) .
u. smilanski , in `` chaos and quantum chaos '' , lecture notes in physics * 411 * , p.57 , ed .
heiss , springer verlag , heidelberg , 1993 .
gopar , m. martinez , p.a .
mello , and h. u. baranger , j. phys . * a 29 * , 881 ( 1996 ) .
l. k. hua , _ harmonic analysis of functions of several variables in the classical domains _ , american mathematical society , providence , 1963 .
v. l. girko _ theory of random determinants _ , kluwer , dordrecht , 1990 .
k. yczkowski and m. ku , j. phys .
a * 27 * , 4235 ( 1994 ) .
f. j. dyson , j. math .
phys . * 3 * , 166 ( 1962 ) .
m. v. berry and m. robnik , j. phys . * a 17 * , 2413 ( 1984 ) .
m.ku and k.yczkowski , phys.rev .
* a44 * , 956 ( 1991 ) . f. izrailev , phys . rep .
129 , 299 ( 1990 ) .
f. haake , m. ku , and r. scharf , z. phys .
* b 65 * , 381 ( 1987 ) .
p. leboeuf , j. kurchan , m. feingold , and d.p .
arovas , phys .
* 65 * , 3076 ( 1990 ) .
grempel , r.e .
prange , and s. fishman , phys . rev . * a 29 * , 1639 ( 1984 ) .
k. yczkowski , acta phys . pol .
* b24 * , 967 ( 1993 ) . k. yczkowski , phys .
e , 1997 , _ in press _ n.l .
balazs and a. voros , ann .
( n.y . ) * 190 * , 1 ( 1989 )
. m. saraceno , ann .
( n.y . ) * 199 * , 37 ( 1990 ) . s.
de bivre , m. degli esposti and g. giachetti , comm .
phys . * 176 * , 73 ( 1996 ) . v. l. girko _ spectral theory of random matrices _ ,
nauka , moscow , 1988 , in russian .
a. hurwitz , nachr .
math .- phys.kl .
71 , ( 1897 )
. w. d. heiss , z. phys . * a 349 * , 9 ( 1994 ) . k. yczkowski and m. ku , phys .
e * 53 * , 319 , ( 1996 ) . | composed ensembles of random unitary matrices are defined via products of matrices , each pertaining to a given canonical circular ensemble of dyson .
we investigate statistical properties of spectra of some composed ensembles and demonstrate their physical relevance .
we discuss also the methods of generating random matrices distributed according to invariant haar measure on the orthogonal and unitary group . 2.0 cm 2.0 cm 2.0 cm 2.0 cm 3.0 cm * * 3.0 cm * * 3.0 cm e - mail : @[email protected] @[email protected] 1.0 cm |
let @xmath1 be an open bounded connected set of @xmath3 , with lipschitz boundary .
the classical relative isoperimetric inequality states that @xmath4 for any measurable subset @xmath5 of @xmath1 ( see , for example , @xcite,@xcite,@xcite ) . here
@xmath6 is the lebesgue measure of @xmath5 , and @xmath7 is the usual perimeter in @xmath1 .
being @xmath8 , the inequality can be written as @xmath9 for any @xmath10 such that @xmath11 .
natural questions related to the inequality are the following : finding the optimal constant @xmath12 proving that it is attained , and characterizing the minimizers .
first results in this direction can be found in @xcite or @xcite , where it is proved that @xmath13 when @xmath1 is the unit disk in @xmath3 , and it is attained at a semicircle .
more generally , in @xcite the author proves that for an open convex set @xmath1 of the plane , @xmath14 is actually a minimum .
moreover , there exists a convex minimizer of whose measure equals @xmath15 , and any minimizer @xmath5 has the following properties : * @xmath16 is either a circular arc or a straight segment .
moreover , neither @xmath5 nor @xmath17 is a circle .
* let @xmath18 be one of the terminal points of @xmath19 .
then @xmath18 is a regular point of @xmath20 and @xmath19 is orthogonal to @xmath20 . as a consequence ,
either @xmath5 or @xmath17 is convex . *
if @xmath21 , then @xmath5 is a circular sector having sides on @xmath20 . in such a case , there exists another minimizer @xmath22 which is a sector with sides on @xmath20 , having the same vertex as @xmath5 , such that @xmath23 .
furthermore , in @xcite @xmath14 is explicitly computed under the additional assumption that @xmath1 is symmetric about a point and also in special cases of convex domains .
if @xmath24 is the inradius of @xmath1 , then @xmath25 we refer the reader to @xcite for some extremal problems involving @xmath14 .
the purpose of the present paper is to find analogous results when the euclidean perimeter is replaced by an `` anisotropic '' perimeter .
more precisely , if @xmath2 is an arbitrary norm on @xmath3 , the perimeter with respect to @xmath2 for a set @xmath26 with sufficiently smooth boundary is given by @xmath27 where @xmath28 is the 1-dimensional hausdorff measure and @xmath29 is the unit outer normal to @xmath5 ( see section 2 for the precise definition ) .
we recall that in this setting it is well - known that the following isoperimetric inequality holds for any @xmath26 @xmath30 where @xmath31 and @xmath32 is polar to @xmath2 ( see @xcite,@xcite,@xcite,@xcite,@xcite ) . moreover , the equality in holds if and only if @xmath5 is homothetic to @xmath33 .
we refer to @xmath33 as the wulff shape .
our results can be summarized as follows . under suitable assumptions on @xmath2 ,
we first show that an anisotropic relative isoperimetric inequality holds .
that is : when @xmath1 is an open , bounded connected set of @xmath3 , with lipschitz boundary , then there exists @xmath34 such that @xmath35 then we prove that , for a convex set @xmath1 , @xmath36 is actually a minimum , there exists a convex minimizer of whose measure equals @xmath15 , and any minimizer @xmath5 has the following properties : * @xmath16 is either homothetic to a wulff arc ( that is an arc of @xmath37 ) or a straight segment .
moreover , neither @xmath5 nor @xmath17 is homothetic to a wulff shape .
* let @xmath18 be one of the terminal points of @xmath19 .
then @xmath18 is a regular point of @xmath20 and @xmath19 verifies the following contact angle condition with @xmath20 : @xmath38 where @xmath39 and @xmath29 are the usual unit outer normal vectors to @xmath20 and @xmath40 at @xmath18 respectively . * if @xmath21 , then @xmath5 is homothetic to a wulff sector ( see section 2 for the precise definition ) having sides on @xmath20 . in such a case
, there exists another minimizer @xmath22 which is a sector with sides on @xmath20 , having the same vertex as @xmath5 , such that @xmath23 .
furthermore , we explicitly compute @xmath36 under the additional assumption that @xmath1 is symmetric about a point .
indeed , @xmath41 where @xmath42 is defined in theorem [ thm : cost ] . for example , if @xmath1 is obtained by a rotation of @xmath43 of a level set of @xmath2 , that is @xmath44 , then @xmath45 where @xmath46 .
we recover immediately the classical result @xmath47 when @xmath2 is the euclidean norm .
the paper is organized as follows . in section 2
we give the precise definitions of anisotropic perimeter and some basic properties . in section 3
we prove the main result .
a fundamental argument is to study problem by considering the area @xmath6 fixed .
finally , we give some examples .
let @xmath48 be a @xmath49 function such that @xmath50 is strictly convex and @xmath51 moreover , suppose that there exist two positive constants @xmath52 such that @xmath53 we define the polar function @xmath54 of @xmath2 as @xmath55 where @xmath56 is the usual scalar product of @xmath3 .
it is easy to verify that also @xmath32 is a convex function which satisfies properties and .
furthermore , @xmath57 the set @xmath58 is the so - called wulff shape centered at the origin .
we will call wulff sector with vertex at the origin the set @xmath59 , where @xmath60 is an open cone with vertex at @xmath61 .
the following properties of @xmath2 and @xmath32 hold true ( see for example @xcite ) : @xmath62 let @xmath1 be an open bounded set of @xmath63 . in @xcite , the perimeter of @xmath64 in @xmath1 with respect to @xmath2 is defined as the quantity @xmath65 the equality @xmath66 holds , where @xmath67 is the reduced boundary of @xmath22 and @xmath68 is the unit outer normal to @xmath22 ( see @xcite ) .
the anisotropic perimeter of a set @xmath22 is finite if and only if the usual euclidean perimeter @xmath69 @xmath70 is finite .
indeed , by properties and we have that @xmath71 and then @xmath72 [ rem : corda ] we observe that when @xmath73 is the image of a smooth curve @xmath74 , @xmath75 $ ] , then @xmath76 coincides with the value @xmath77 by regularity of @xmath2 , the curve joining two points @xmath78 and @xmath79 which minimizes @xmath80 is the straight segment @xmath81 .
this can be shown by classical argument of calculus of variations .
we consider , for sake of simplicity , the curves @xmath82 . denoting by @xmath83 ,
the minimum of the problem @xmath84 is the solution to @xmath85 such solution is the linear function passing through @xmath86 and @xmath87 .
let @xmath64 be a bounded open set with smooth boundary , @xmath88 the unit outer normal at @xmath89 , in the usual euclidean sense .
let @xmath90 be a @xmath91 function such that @xmath92 , @xmath93 and @xmath94 on @xmath95 .
hence , @xmath96 on @xmath95 .
the anisotropic outer normal @xmath97 is defined as @xmath98 and , by the properties of @xmath2 , @xmath99 the anisotropic curvature @xmath100 of @xmath101 is @xmath102 , \quad ( x , y)\in { \partial}f.\ ] ] let @xmath103 . without loss of generality , we can locally describe @xmath95 with a @xmath91 function @xmath104x_0-\delta , x_0+\delta[\rightarrow { \mathbb r}$ ] , that is @xmath22 is the epigraph of @xmath105 near @xmath106 . by properties of @xmath2 , the anisotropic curvature @xmath107 of @xmath101 at @xmath108 can be written as @xmath109 we stress that if @xmath22 is homothetic to the wulff shape @xmath33 and centered at @xmath108 , the anisotropic outer normal at @xmath110 has the direction of @xmath111 .
indeed , being @xmath112 , for some positive @xmath113 , by property it follows that @xmath114 see figure [ fig : normal ] for an example . and @xmath115 .
when @xmath116 , the usual and the anisotropic outer normal are , in general , different . ] [ rem : constcurv ] let be @xmath117 , with @xmath118 .
it is not difficult to show ( see , for instance , @xcite , @xcite ) that the anisotropic curvature at @xmath89 is @xmath119
let @xmath1 be an open bounded connected set of @xmath3 , with lipschitz boundary .
then an anisotropic relative isoperimetric inequality holds .
namely , there exists a constant @xmath120 such that @xmath121 for every measurable set @xmath122 .
the hypotheses made on @xmath1 guarantee that a relative isoperimetric inequality holds when we consider the usual perimeter @xmath7 ( see @xcite,@xcite,@xcite ) . hence the inequality follows immediately from property .
our aim is to study , for @xmath1 bounded and convex , the best constant in the inequality , that is to find the infimum @xmath123 where @xmath124 to prove that @xmath125 is actually a minimum , and to characterize the minimizers .
furthermore , we will find the explicit value of @xmath125 in some special case .
if @xmath5 is a minimizer of , then @xmath5 solves also the following problem under volume constraint : @xmath126 the following result characterizes the minimizers of the above problem .
[ th : wulff - arc ] let @xmath1 be an open bounded connected set of @xmath3 , with lipschitz boundary .
then there exists a minimizer @xmath5 of the problem @xmath127 with @xmath128 fixed .
moreover , @xmath129 is either homothetic to an arc of @xmath130 , or a straight segment .
hence a minimizer of , if exists , has the same characterization .
the existence of a minimizer of follows by the lower semicontinuity of @xmath131 ( see @xcite ) using standard methods of calculus of variations . to prove the result
, we proceed by steps .
* step 1 . *
first , we show that a minimizer @xmath5 is locally homothetic to an arc of @xmath130 , or a straight segment . fixed @xmath132
, we can locally describe @xmath73 with a @xmath91 function @xmath90 ( see @xcite,@xcite,@xcite,@xcite ) .
that is , without loss of generality , there exists a rectangle @xmath133x_0-\delta , x_0+\delta[\times i $ ] where @xmath134 is the epigraph of @xmath135x_0-\delta , x_0+\delta[\to i$ ]
. moreover , there exists @xmath113 such that @xmath90 is the minimum of the functional @xmath136 with boundary conditions @xmath137 and @xmath138 .
the corresponding euler equation associated to @xmath139 is @xmath140x_0-\delta , x_0+\delta[,\\ v(x_0\pm\delta)=u(x_0\pm\delta ) . \end{array } \right.\ ] ] if @xmath141 , there exists a linear function @xmath142 which solves . if @xmath143 , by remark [ rem : constcurv ] , the function @xmath144 , which describes @xmath145 ( up to translation ) near @xmath146 , is a solution of . on the other hand , for any @xmath147 , the regularity on @xmath2 guarantees that the functional @xmath139 is strictly convex .
hence , @xmath148 is the unique solution of ( see also @xcite , @xcite )
. * step 2 .
* now we show that the minimizer has the same anisotropic curvature at any point .
let us take @xmath149 and @xmath150 in @xmath16 .
as in the step 1 , let us consider @xmath151x_1-\delta_1 , x_1 + \delta_1[\to i_1 $ ] and @xmath152x_2-\delta_2 , x_2 + \delta_2[\to i_2 $ ] two functions which locally describe @xmath153 . moreover , there exist @xmath154 and @xmath155 such that @xmath156 , for @xmath157 , minimizes the functional @xmath158 with boundary conditions @xmath159 .
we claim that @xmath160 .
this can be shown by arguing as in @xcite , theorem 2 .
we briefly describe the idea , and we refer to the quoted paper for the precise details .
we assume that @xmath161 .
a similar argument can be repeated in the other cases . for every @xmath162\lambda_1,\lambda_2[$ ]
there exists a function @xmath163 which is the unique minimizer to @xmath164 where @xmath165 and @xmath166x_i-\rho , x_i+\rho[$ ] , with boundary conditions @xmath167 . by convexity of @xmath2
, a comparison argument shows that @xmath168 in @xmath169 , and @xmath170 in @xmath171 . defining @xmath172
it is possible to prove that there exist two suitable positive numbers @xmath173 and @xmath174 such that @xmath175 this implies that , defining the set @xmath176 as @xmath177 \setminus \left [ c_2\cap(e\setminus\operatorname{epi}u_{r_2,2 } ) \right],\ ] ] where @xmath178 , we have that @xmath179 .
finally , we get that @xmath180 and @xmath181 where last line in the above equality vanishes , by . by minimality of @xmath182 and @xmath183 , @xmath184 , and
this contradicts the minimality of @xmath5 .
hence , @xmath160 .
* step 3 . *
we point out that the claim of step 2 assure that @xmath16 consists of wulff arcs , all with the same curvature , or straight segments .
to conclude the proof of the theorem , we have to prove that @xmath5 and @xmath129 are connected .
this can be shown by repeating line by line the proof of theorem 2 in @xcite .
the following property of the minimizers is a direct consequence of remark [ rem : corda ] .
[ prop : conv ] let @xmath1 be an open bounded connected set of @xmath3 , with lipschitz boundary .
suppose that @xmath5 is a minimizer of . if @xmath185 and @xmath16 is not a straight segment , @xmath16 is concave towards @xmath5 .
if @xmath185 and @xmath16 is strictly concave towards @xmath17 , we can consider a new set @xmath176 by adding to @xmath5 the region of @xmath1 between @xmath186 and a straight segment joining two suitable points of @xmath73 .
choosing the two points sufficiently near , we get that @xmath187 and , by remark [ rem : corda ] , @xmath188 .
this contradicts the minimality of @xmath5 .
[ thm : angle ] let @xmath1 be an open bounded convex set of @xmath3 .
suppose that @xmath5 is a minimizer of , and let @xmath18 be a terminal point of @xmath189 .
then @xmath20 at @xmath18 is @xmath190 , and @xmath191 where @xmath192 is the anisotropic outer normal to @xmath189 and @xmath39 is the usual unit outer normal to @xmath193 at @xmath18 .
[ rem : angle ] the angle condition is justified by the following natural geometric argument .
let @xmath194 be a straight line , @xmath195 . by an immediate calculation ,
the straight segment which minimizes @xmath196 between @xmath78 and @xmath197 is parallel to the straight line @xmath198 which has to satisfy the following orthogonality condition : @xmath199 using the notation of theorem [ thm : angle ] , if we consider as @xmath200 the tangent line to @xmath20 at a terminal point @xmath18 of @xmath201 , and as @xmath197 the tangent straight line to @xmath40 at @xmath18 , then and coincide .
and @xmath5 is homothetic to the wulff shape @xmath33 and centered at @xmath108 .
the tangent lines to @xmath20 at the contact points have the same direction of the anisotropic normal to @xmath40 at the same points . ]
we first assume that @xmath20 is @xmath190 at @xmath18 .
let us suppose , by contradiction , that is not verified .
the idea is to construct a new set @xmath176 such that @xmath202 .
this will contradict the minimality of @xmath5 .
to do that , we need to distinguish three cases .
first of all , we denote with @xmath197 the tangent line to @xmath20 at @xmath18 , and with @xmath203 and @xmath200 two half line with vertex at @xmath18 and towards @xmath1 such that @xmath203 is tangent to @xmath40 at @xmath18 and @xmath200 satisfies the angle condition with respect to @xmath197 . * case 1 . *
we first assume that @xmath185 and the angle between @xmath197 and @xmath203 towards @xmath5 is greater than the one between @xmath197 and @xmath200 towards @xmath5 .
we construct the straight segment @xmath204 parallel to @xmath200 joining a suitable point @xmath205 and @xmath206 .
being @xmath1 convex , we can consider the point @xmath207 .
denoted by @xmath208 the closed region delimited by @xmath209 , the arc of @xmath40 joining @xmath210 and @xmath18 and the arc of @xmath20 between @xmath18 and @xmath211 , let be @xmath212 ( see figure [ fig:3 ] ) .
( , 5.5 cm ) ( -1.5cm,0 cm ) .,title="fig : " ] ( -8cm,4.5cm)@xmath1 ( -2cm,2.6cm)@xmath5 ( -5.1cm,0.77cm)@xmath18 ( -.55cm,0.2cm)@xmath197 ( -3.1cm,4.9cm)@xmath200 ( -7.2cm,3cm)@xmath203 ( -5.92cm,2.2cm)@xmath210 ( -6.4cm,1.09cm)@xmath213 ( -6.4cm,1.55cm)@xmath211 ( -5.8cm,1.35cm)@xmath208 we choose @xmath210 sufficiently near to @xmath18 such that @xmath214 . hence @xmath176 has larger area than @xmath5 and , by remark [ rem : angle ] and lemma [ rem : corda ] also smaller anisotropic perimeter .
* now we still suppose that @xmath185 , and the angle between @xmath197 and @xmath203 towards @xmath5 is smaller that the one between @xmath197 and @xmath200 .
we construct the straight segment @xmath215 parallel to @xmath200 joining a suitable point @xmath216 and @xmath206 , the point @xmath217 and the set @xmath208 as the intersection between the triangle @xmath218 and @xmath5 ( see figure [ fig:4 ] ) .
we define @xmath219 . .
] we show that , for a suitable choice of @xmath210 , @xmath220 differently from the case 1 , inequality is not obvious because @xmath176 has both smaller perimeter and area .
hence , we explicitly calculate the right - hand side in .
denoted by @xmath221 , @xmath222 , @xmath223 , @xmath224 , the inequality becomes @xmath225 denoting by @xmath226 and @xmath227 , where @xmath228 and @xmath229 are the curves which represent @xmath230 and @xmath231 respectively , it is easy to prove that @xmath232 where @xmath233 and @xmath234 are respectively the usual lenght and the direction of @xmath230 , and @xmath235 where @xmath236 and @xmath237 are respectively the usual lenght and the direction of @xmath231 . observe that by construction , @xmath238 .
we first show replacing @xmath239 with @xmath240 and @xmath241 with @xmath242 , where @xmath243 and @xmath244 are the measures of the sets as in figure [ fig : approx ] .
by elementary properties of triangles , @xmath245 the function @xmath246 is strictly decreasing in the interval @xmath247 $ ] , with @xmath248 which is strictly positive , being @xmath238 .
this implies that , for @xmath249 , @xmath250 on the other hand , by remark [ rem : corda ] we get @xmath251 hence , being obviously @xmath252 , by , it follows for a suitable choice of @xmath210 .
* case 3 . * finally ,
if @xmath253 , we can both consider , as minimum sets , @xmath5 and @xmath17 . hence ,
if the angle condition is not verified , we can suppose , without loss of generality , that the lines @xmath254 and @xmath203 verify the hypotheses of case 2 .
if @xmath16 is a straight segment , or it is strictly concave towards @xmath5 , we can repeat line by line the same argument of case 2 . otherwise , if @xmath16 is strictly concave towards @xmath17 , proceeding as in case 1 we construct the straight segment @xmath231 , and another straight segment @xmath255 joining two suitable points of @xmath73 .
let @xmath256 and @xmath257 be as in figure [ fig : fine ] , and define @xmath258 .
choosing @xmath259 and @xmath210 in such a way that @xmath260 , since @xmath261 we obtain a contradiction , and the proof of the theorem is completed when @xmath18 is a regular point of @xmath20 .
finally , we show that @xmath73 can not join @xmath20 at a non regular point . by contradiction , suppose that @xmath20 is not regular at @xmath18 . by convexity
it has different right and left tangent straight lines , that we denote by @xmath262 and @xmath263 respectively .
clearly , the tangent line @xmath203 does not satisfy the contact angle condition with both @xmath262 and @xmath263 .
so we can repeat the arguments just considered by replacing the straight line @xmath197 with @xmath262 or @xmath263 , and obtaining a contradiction with the minimality of @xmath5 .
[ properties ] let @xmath1 be an open bounded convex set of @xmath3 , @xmath264 , and set @xmath265 be a minimizer of problem @xmath266 we have the following properties : 1 .
[ item1 ] neither @xmath265 nor @xmath267 is homothetic to a wulff shape ; 2 . [ item2 ] if @xmath268 , and @xmath269 and @xmath270 are the terminal points of @xmath271 on @xmath20 , then the left and right tangent straight lines at @xmath269 to @xmath20 do not make a cone towards @xmath272 with the analogous lines at @xmath270 .
[ item3 ] if @xmath268 and @xmath273 is not a straight segment , @xmath273 is concave towards @xmath5 .
we prove the three properties by contradiction with the minimality of @xmath265 , finding a set with same area and smaller perimeter .
let @xmath265 or @xmath272 be homothetic to a wulff shape .
since the perimeter @xmath274 is invariant up to translations in @xmath1 , we can suppose that @xmath275 touches at least at one ( regular ) point @xmath276 , and there exists a small ball @xmath277 centered at @xmath278 such that @xmath279 .
we stress that in @xmath278 the contact angle condition can not hold .
indeed @xmath280 , and by and the homogeneity of @xmath2 we should have that @xmath281 so @xmath282 and this is absurd . then arguing as in case 3 of the proof of theorem [ thm : angle ] , being @xmath265 ( or @xmath283 ) strictly convex we can add and subtract two small regions in order to get a new set with the same area and smaller perimeter ( see figure [ fig : fine ] ) . this proves .
property easily follows by the convexity of @xmath1 . indeed ,
if @xmath265 has measure smaller than @xmath284 and does not verify , we can do a suitable translation @xmath285 of @xmath275 towards the vertex @xmath286 of the cone in @xmath3 , in such a way that the set @xmath287 bounded by @xmath288 towards @xmath286 and @xmath20 , has measure @xmath289 and smaller perimeter than @xmath265 in @xmath1 ( see figure [ fig : conocontr ] ) .
this contradicts the minimality of @xmath265 .
finally , suppose that @xmath275 is concave towards @xmath290 . by property ,
the tangent straight lines at terminal points of @xmath271 either make a cone towards @xmath265 or are parallel . as in proposition [ prop : conv ] , in both cases we can add a small region to @xmath265 in order to decrease the perimeter and , similarly as in the proof of property , with a suitable translation of @xmath271 towards the vertex of the cone , keep fixed the area @xmath291 .
this proves property . in order to prove the existence of a minimizer of , we need the following technical lemma . [ lem : decr ]
let @xmath2920,+\infty[\rightarrow{\mathbb r}$ ] be a lower semicontinuous function .
suppose that for any @xmath293 there exists @xmath294 such that @xmath295.\ ] ] then @xmath296 is decreasing in @xmath2970,+\infty[$ ] . by contradiction ,
suppose that there exist @xmath298 such that @xmath299 define @xmath300 as @xmath301 the function @xmath302 is lower semicontinuous , and for any @xmath289 there exists @xmath294 such that @xmath303 , for any @xmath304 $ ] .
hence , we can define @xmath305 as @xmath306 if @xmath307 , then @xmath308 , and this contradicts . hence , suppose that @xmath309 .
being @xmath302 lower semicontinuous , @xmath310 is actually a maximum : @xmath311 but this contradicts the definition of @xmath312 . indeed ,
by the property of @xmath302 we can take @xmath313 such that @xmath314 .
hence , necessarily @xmath315 , and the proof is concluded .
[ thm : decr ] let @xmath1 be an open bounded convex set of @xmath3 .
let @xmath316 be the function defined in @xmath2970,|\omega|/2]$ ] as @xmath317 then , we have the following results hold : 1 .
@xmath316 is a decreasing lower semicontinuous function in @xmath2970,|\omega|/2]$ ] , [ en:1 ] 2 .
the sets which minimize verify the contact angle condition .
more precisely , they verify the thesis of theorem [ thm : angle ] .
[ en:2 ] we first prove that the function @xmath296 is lower semicontinuous in @xmath2970,|\omega|/2]$ ] .
let be @xmath3180,|\omega|/2]$ ] , and take a positive sequence @xmath319 such that @xmath320 .
consider @xmath321 such that @xmath322 and @xmath323 . by proposition [ properties ]
, @xmath324 is convex .
hence , by the blaschke selection theorem ( see @xcite , page 50 ) @xmath324 converges ( up to a subsequence ) to a set @xmath5 in the hausdorff metric .
being @xmath324 convex and bounded , then @xmath325 in @xmath326 strongly , and @xmath327 . using the lower semicontinuity of @xmath328
( see @xcite ) we get @xmath329 in order to prove that @xmath296 is decreasing , let be @xmath3300,|\omega|/2[$ ] fixed and consider @xmath265 , @xmath331 such that @xmath332 .
we claim that there exists a positive number @xmath333 and a family of sets @xmath334 , @xmath335 with continuously increasing area and @xmath336
. then @xmath3370,\delta_k].\ ] ] being @xmath296 lower semicontinuous in @xmath2970,|\omega|/2]$ ] , by lemma [ lem : decr ] this is sufficient to show that @xmath296 is decreasing . by theorem [ th : wulff - arc ] , @xmath271 is a straight segment or a wulff arc , and by property of proposition [ properties ] , it has two terminal points @xmath338 on @xmath20 .
we suppose that such points are regular for @xmath20 , so that by property proposition [ properties ] , the tangent lines to @xmath20 , @xmath339 at @xmath338 either are parallel or make a cone @xmath60 towards @xmath265 . in the first case , the claim follows immediately by the convexity of @xmath1 and making a suitable translation of @xmath275 .
hence , we consider the second case , and suppose without loss of generality that @xmath340 .
moreover , by property of proposition [ properties ] , @xmath271 is a straight segment , or concave towards @xmath265 .
we need to distinguish two cases for the shape of @xmath1 . * case 1 .
* @xmath341 is not contained in @xmath342 .
we set @xmath343 , @xmath344 , the region bounded by @xmath345 and @xmath342 , and @xmath346 . for sake of simplicity , we define @xmath347 .
let @xmath348 be the boundary point of @xmath349 on @xmath339 . moreover , let be @xmath350 , where @xmath351 is the tangent line to @xmath352 at @xmath348 .
( see figure [ fig : angmezzo ] ) .
now we compute area and relative perimeter of @xmath334 .
observe that the triangles @xmath353 of vertex @xmath348 , @xmath354 and @xmath338 have area @xmath355 .
we have : @xmath356 it follows that @xmath357 \le \\ \le \frac 1 \delta q(e_k ) \left[\frac{(1+\delta)^2 } { 1 + 2\delta \frac { |c|}{|e_k|}+o(\delta ) } - 1 \right ] = \\= q(e_k ) \left [ \frac{2 \left(1- \frac { |c|}{|e_k| } \right ) + o(1 ) } { 1+o(1 ) } \right].\end{gathered}\ ] ] since @xmath358 , then for @xmath359 sufficiently small we obtain that the left - hand side of is negative .
this proves , and hence , if @xmath338 are regular points of @xmath20 .
if , for example , @xmath269 is not a regular point , we can repeat the arguments just considered by replacing @xmath262 with the left or right tangent straight line . now we prove . in order to fix the ideas
, we consider the regular point @xmath269 and the straight line @xmath200 which verifies the contact angle condition with @xmath262 .
let @xmath360 be the angle between @xmath262 and @xmath200 towards @xmath265 , and @xmath361 the corresponging angle between @xmath273 and @xmath262 towards @xmath265 .
suppose by contradiction that @xmath362 . if @xmath363 , then the construction made in the proof of case 2 of theorem [ thm : angle ] allows to take @xmath176 such that @xmath364 and @xmath365 , and this contradicts the monotonicity of @xmath296 .
if @xmath366 , and @xmath367 is a wulff arc , as in case 3 of theorem [ thm : angle ] we can add and subtract two sets in order to decrease the perimeter and to preserve the area , contradicting the minimality of @xmath265 . in the case
that @xmath271 is a straight segment , we can add a small region to @xmath265 in order to decrease the perimeter and with a suitable translation , keep fixed the area @xmath291 .
finally , @xmath269 can not be a singular point for @xmath20 . otherwise , similarly as observed at the end of the proof of theorem [ thm : angle ] and proceeding as above
, we get a contradiction with the minimality of the minimizer . * case 2 .
* @xmath341 is contained in @xmath342 , that is @xmath368 .
define @xmath369 , @xmath370 , and @xmath371 such that @xmath372 first , we prove that at the terminal points of @xmath271 it holds the contact angle condition . in order to fix the ideas
, we consider the regular point @xmath373 and the straight line @xmath200 which verifies the contact angle condition with @xmath262 .
let @xmath360 be the angle between @xmath262 and @xmath200 towards @xmath265 , and @xmath361 the corresponging angle between @xmath273 and @xmath262 towards @xmath265 .
suppose by contradiction that @xmath362 .
* case 2-a * let be @xmath374 .
reasoning as in the proof of theorem [ thm : angle ] , we find @xmath176 such that @xmath365 , with @xmath375 sufficiently small .
then there exists @xmath376 such that @xmath377 , and @xmath378 .
this contradicts the minimality of @xmath265 . repeating the same argument for @xmath270 , we have that the terminal points of @xmath273 have to verify the angle condition , that is @xmath265 is homothetic to a wulff sector @xmath379 .
* case 2-b * let be @xmath380 . then , as @xmath381 , the set @xmath382 is such that @xmath383 . thanks to case 2-a , we have that @xmath384 is attained at a wulff sector , namely the set @xmath385 , for @xmath386 such that @xmath387 .
hence @xmath388 .
define @xmath389 we have that @xmath390 is finite and @xmath391 .
otherwise , there exists @xmath392 such that @xmath393 and @xmath394 , and this is a contradiction . as matter of fact , the homogeneity of @xmath2 and imply , for @xmath395 , that @xmath396 . moreover , for @xmath397 , @xmath398 .
hence by the divergence theorem we get that , for @xmath399 , @xmath400 define @xmath401 $ ] , and @xmath402 the cone made by the two half - straight lines @xmath403 , @xmath157 with origin at @xmath61 and passing through one of the two terminal points of @xmath404 $ ]
. by and the convexity of @xmath1 , we get , for an appropriate @xmath359 , @xmath405 and @xmath406 then @xmath275 must verify the contact angle condition at each @xmath338 , and this concludes the case 2-b , and is proved . in order to prove , and
hence , we observe that from , @xmath407 , for some @xmath408 .
let @xmath390 as in , and suppose that @xmath409 , otherwise is immediate , being @xmath410 , for any @xmath411 . defining @xmath401 $ ] and reasoning as in case 2-b , we get
. finally , the regularity of @xmath338 on @xmath20 follows exactly as in the case 1 , and the proof is completed .
we observe that if @xmath5 is a minimizer of , and @xmath185 , then @xmath5 is homothetic to a wulff sector with sides on @xmath20 .
otherwise , arguing as in case 1 of the proof of theorem [ thm : decr ] , we construct a new set @xmath176 with @xmath202 .
hence , @xmath412 with sides on @xmath20 .
being @xmath413 where @xmath414 , there exists another minimizer @xmath22 which is a wulff sector with sides on @xmath20 and @xmath415 .
now we are able to prove the main result .
[ thm : main ] let @xmath1 be an open bounded convex set of @xmath63 . then there exists a convex minimizer of problem whose measure is equal to @xmath284 .
more precisely , either a minimizer @xmath5 of has measure @xmath284 , or @xmath5 is homothetic to a wulff sector with sides on @xmath20 .
finally , it verifies the contact angle condition . let @xmath296 defined as in the above theorem and , being @xmath296 decreasing in @xmath2970,|\omega|/2]$ ] , it attains its minimum at @xmath416 .
now we are able to prove that has a minimum .
let @xmath287 be such that @xmath417 and @xmath418 .
let @xmath324 , @xmath419 be a minimizing sequence of problem , that is @xmath420 , @xmath421 .
otherwise , we replace @xmath324 with the minimizer of problem with volume constraint @xmath422 .
then @xmath423 passing to the limit , @xmath424 and @xmath287 is a minimizer of , whose boundary in @xmath1 is a straight segment or a wulff arc . from the proof of theorem [ thm : decr ]
it follows that if @xmath5 is another minimizer of with @xmath185 , then it is a wulff sector with sides on @xmath20 . recalling theorem [ thm : angle ] ,
the result is completely proved . in the following theorem
, we characterize the minimizers for centrosymmetric sets , and find the constant @xmath125 in . for sake of simplicity ,
if @xmath18 is a point in @xmath3 , we put @xmath425 , where @xmath80 is defined in , and @xmath426 is a curve which represent the straight segment @xmath427 joining @xmath18 with the origin @xmath428 . we observe that if @xmath429 , then @xmath430 .
[ thm : cost ] let @xmath431 be a convex bounded set , symmetric about the origin @xmath428 .
then a minimizer of is a set @xmath5 whose boundary in @xmath1 is a straight segment passing through the origin and such that @xmath432 , where @xmath433 .
hence , @xmath434 the first step is to prove the existence of a set @xmath5 enjoying the properties of the statement .
let us consider the set @xmath435 then @xmath436 meets @xmath20 at least at two symmetric regular points @xmath269 , @xmath270 .
we observe that in @xmath338 the contact angle condition is satisfied . indeed , the anisotropic outer normal to the straight segment @xmath437 is @xmath438 , where @xmath439 , @xmath157 .
denoted by @xmath440 the unit outer normal to @xmath20 at @xmath338 , being @xmath441 , we have @xmath442 .
we show that @xmath443 is the boundary in @xmath1 of the required set @xmath5 , and @xmath444 . by theorem [ thm : main ] , there exists a convex minimizer of whose measure is @xmath284 , which is a straight segment or a wulff arc . if we show that @xmath445 , where @xmath22 is a open convex subset of @xmath1 such that @xmath415 and @xmath446 is a straight segment or a wulff arc , we have done .
clearly , any straight segment passing through the origin bounds in @xmath1 a set with greater perimeter than @xmath5 and with same area @xmath284 .
we do not consider the straight segments which not contain the origin , because they bounds in @xmath1 sets with measure different from @xmath284 .
hence we can suppose that @xmath447 is a wulff arc .
obviously , @xmath448 , otherwise @xmath449 .
more precisely , denoted by @xmath79 and @xmath450 the terminal points of @xmath446 , we get that @xmath451 , where @xmath452 is bounded by @xmath20 and @xmath453 , otherwise @xmath454 , and this is impossible .
hence we can consider the straight segments in @xmath22 , @xmath455 and @xmath456 , and it is not difficult to show that @xmath457 and this concludes the proof .
[ rem : thm ] if @xmath458 , i.e. @xmath1 is obtained by a rotation of @xmath43 the @xmath459level set of @xmath2 , then theorem [ thm : cost ] gives @xmath460 where @xmath461 .
observe that any straight segment passing through the origin and joining the boundary of @xmath1 bounds a minimizer .
in particular , if @xmath462 , we recover the classical result @xmath463 ( see for instance @xcite,@xcite ) .
here we apply the results just obtained to some particular function @xmath2 . let @xmath464 defined as @xmath465 an immediate calculation gives that @xmath466 if @xmath1 is the ellipse @xmath467 , then @xmath468 , and @xmath469
. by theorem [ thm : cost ] and remark [ rem : thm ] we have @xmath470 moreover , the equality in holds if and only if @xmath201 is any straight segment passing through the origin ( see figure [ fig : ellisse ] ) .
we observe that if we compute @xmath125 for the ellipse @xmath471 , with for example , @xmath472 , then the smaller axis of the ellipse ( in the usual sense ) is the boundary of the only minimizer of ( see figure [ fig : ellisse ] ) , and the constant @xmath125 is @xmath473 is a level set of @xmath2 , and the straight segment is the boundary of the only minimizer of . in the second figure , @xmath1 is a level set of @xmath32 , and any straight segment passing through the origin is the boundary of a minimizer.,title="fig : " ] ( -7.8cm,1.37cm)@xmath474 ( -.35cm,1cm)@xmath1 we point out that the above result for @xmath1 can be obtained directly by the classical relative isoperimetric inequality for the euclidean perimeter . indeed , the anisotropic relative perimeter of a smooth set @xmath5 , whose boundary is described by @xmath475 , with @xmath476 $ ] , is @xmath477 defining @xmath478 and @xmath479 , the curve @xmath480 describe the boundary of the unit euclidean disk @xmath481 with radius @xmath200 and centered at the origin . by changing the variables in ,
we get latexmath:[\[\int_\alpha^\beta \left(\frac{(z')^2}{a^2b^2}+\frac{(w')^2}{a^2b^2}\right)^{\frac{1}{2 } } dt = \frac{1}{ab}p(\tilde e , b_1 ) \ge \frac{1}{ab}\sqrt{\frac{8}{\pi } } finally , the characterization of the minimizers is a direct consequence of the fact that in the classical relative isoperimetric inequality , the minimizers are the diameters .
hence in this case we get the relative anisotropic isoperimetric inequality by a linear trasformation , as a consequence of the classical relative isoperimetric inequality .
let us consider @xmath488 .
being @xmath1 invariant by @xmath489rotations , by theorem [ thm : cost ] and remark [ rem : thm ] we have @xmath490 where @xmath46 , and any straight segment passing through the origin bounds a minimizer .
[ ex:3 ] let @xmath2 be defined as follows : @xmath491 with @xmath492 , @xmath493 and @xmath494 .
let us consider @xmath495 .
then @xmath496 we stress that if @xmath497 , then easy computations give that @xmath498 observe that @xmath499 ( compare figure [ fig : centrosimmdoppio ] ) .
[ nonreg ] let us consider @xmath500 .
the singular behavior of @xmath2 does not allow to apply the previous results .
then , in order to prove the anisotropic isoperimetric inequality relative to @xmath1 with respect to @xmath2 , we argue by approximation . let be @xmath501 , and @xmath502
. for any set @xmath503 such that @xmath504 , we have @xmath505 and the best constant is reached by a rectangle whose boundary in @xmath1 is the straight segment joining @xmath506 and @xmath507 ( or @xmath508 and @xmath509 ) .
we can pass to the limit as @xmath510 in , obtaining @xmath511 any straight segment passing through the origin and joining the boundary of @xmath1 bounds a minimizer .
unlike the case of @xmath2 smooth ( compare remark [ rem : thm ] ) , such sets are not the only minimizers .
for example , in figure [ fig : minim ] some minimizer is represented . indeed ,
if @xmath40 is described by a lipschitz function @xmath512 , @xmath75 $ ] , the perimeter is @xmath513 then in the picture on the left - hand side of figure [ fig : minim ] , the perimeter of @xmath5 is @xmath514 and @xmath515 .
moreover , in the other picture any triangle @xmath5 such that @xmath16 is a straight segment parallel to a diagonal is a minimizer . | in this paper we prove a relative isoperimetric inequality in the plane , when the perimeter is defined with respect to a convex , positively homogeneous function of degree one @xmath0 . under suitable assumptions on @xmath1 and @xmath2 , we also characterize the minimizers . |
language is the most characteristic trait of human communication but takes on many heterogeneous forms .
dialects , in particular , are linguistic varieties which differ phonologically , gramatically or lexically in geographically separated regions @xcite .
however , despite its fundamental importance and many recent developments , the way language varies spatially is still poorly understood .
traditional methodological approaches in the study of regional dialects are based on interviews and questionnaires administered by a researcher to a small number ( typically , a few hundred ) of selected speakers known as informants @xcite .
based on the answers provided , linguistic atlases are generated that are naturally limited in scope and subject to the particular choice of locations and informants and perhaps not completely free of unwanted influences from the dialectologist .
another approach is the use of mass media corpora which provide a wealth of information on language usage but suffer from the tendency of media and newspapers to use standard norms ( the `` bbc english '' for example ) @xcite that limits their usefulness for the study of informal local variations . on the other hand ,
the recent rise of online social tools has resulted in an unprecedented avalanche of content that is naturally and organically generated by millions or tens of millions of geographically distributed individuals that are likely to speak in vernacular and do not feel constrained to use standard linguistic norms .
this , combined with the widespread usage of gps enabled smartphones to access social media tools provides a unique opportunity to observe how languages are used in everyday life and across vast regions of space . in this work ,
we use a large dataset of geolocated tweets to study local language variations across the world .
similar datasets have recently been used to map public opinion and social behavior @xcite and to analyze planetary language diversity @xcite .
preliminary results demonstrating the feasibility of this approach have thus far been limited to considering only few words or just a few geographical areas @xcite . here
, we move beyond the mere proof of concept and provide a detailed global picture of spatial variants for a specific language . for definiteness , we choose spanish as it is not only one of the most spoken in the world but it has the added advantage of being spatially distributed across several continents@xcite . several other languages such as mandarin or english have more native speakers or higher supra - regional status but their use is hindered by the limited local availability of twitter ( mandarin ) or a high abundance of homographs that percludes a detailed lexicographic analysis ( english ) .
we used the twitter gardenhose to gather an unbiased sample of all tweets written in spanish that contained gps information over the course of over two years .
language detection was performed using the state of the art chromium compact language detector@xcite software library .
the resulting dataset contained over @xmath3 geolocated tweets written in spanish distributed across the world ( see fig .
[ fig_map ] ) .
as expected , most tweets are localized in spain , spanish america and extensive areas of the united states .
these results are consistent with recent sociolinguistic data @xcite , providing an initial level of validation to our approach .
interestingly , we also find significant contributions from major non - spanish - speaking cities in latin america and western europe , likely due to considerable population of temporary settlers and tourists .
@xcite for further details and results on this dataset .
traditional approaches in dialectology have preferred rural , male informants while modern analyses include interactions with urban speakers regardless of age and gender . on average ,
twitter users are young , urban@xcite and more likely to be technologically savvy thus providing more modern perspective on the use of language .
to be able to determine exactly what the major local varieties of spanish are , we use a list of concepts and utterances selected from an exhaustive study of lexical variants in major spanish - speaking cities .
reference @xcite provides a comprehensive list of possible words representing several concepts , such as `` popcorn '' , `` car '' , `` bus '' , etc .
we selected a subset of concepts that minimized possible semantic ambiguities by ensuring that they contained no common words . in our initial set of tweets
we observed @xmath4 geolocated instances where words from our catalogue were used .
individual instances were then agregated geographically into cells of @xmath5 , which corresponds to an approximate area of @xmath6 km@xmath7 in the equator . * geographical distribution of the dominant word for the concepts computer ( left ) and car ( right ) .
* map locations are colored according to the most common expression found in the corresponding cell .
the area of the circle is proportional to the number of tweets , title="fig : " ] * geographical distribution of the dominant word for the concepts computer ( left ) and car ( right ) .
* map locations are colored according to the most common expression found in the corresponding cell .
the area of the circle is proportional to the number of tweets , title="fig : " ] finally , we define the dominant word for each concept in each geographical cell by a simple majority rule and generate a @xmath8 matrix where element @xmath9 is @xmath10 when word @xmath11 is the dominant for a given concept in cell @xmath12 and @xmath13 otherwise .
the resulting matrix has @xmath14 rows and @xmath15 columns and constitutes the dataset used for the analysis presented in the remainder of this paper .
figure [ fig_popcorn_car ] illustrates two illustrative concepts ( computer and car ) that are both associated to multiple utterances .
each utterance is represented with a different color .
we draw a circle centered on each cell with an area proportional to the number of tweets that use the corresponding expression it is clear from the map that some expressions ( _ computadora _ , _ ordernador _ , _ computador _ ) are strongly clustered in space , allowing us to easily define regional dialects characterized by the set of dominant words used to express the concepts in our list . due to the unique resolution of our data we could limit the isoglosses ( boundaries ) of the regions corresponding to each concept - word with a high degree of precision .
however , the isoglosses corresponding to different concepts can overlap and bundle rendering any simple arrangement of dialect areas almost impossible . the natural way to overcome this difficulty and
characterize the various regional dialects present in modern day spanish is to apply machine learning ( ml ) approaches to automatically cluster the @xmath16 matrix and identify which cells are closely related to one another .
we start by applying principal component analysis to reduce the dimensionality of the matrix @xmath16 .
pca determines the linear combinations of the columns ( features in ml literature ) of the matrix that explain most of the variance observed in the rows ( observations ) .
we find that by projecting the data onto the @xmath17 principal components ( see fig .
[ pca ] ) we are able to maintain over @xmath18 of the variance in the data while reducing by @xmath19 the dimension of the matrix with clear numerical advantages .
the task of identifying meaningful clusters in this matrix is now simplified .
we proceed by applying the well known @xmath20-means@xcite algorithm that iteratively refines the position of the centers of @xmath20 clusters until it finds a stable set of locations .
the main dificulty of utilizing this algorithm lies in identifying the correct number @xmath20 of clusters to utilize . here
, we apply the @xmath21 metric introduced by pham _
_ to establish the best value for @xmath20 .
we run @xmath20-means with values of @xmath20 up to @xmath22 using @xmath23 different random initializations and depict the results in fig .
[ population ] a ) . for verification purposes
, we also plot the value of the silhouette @xcite of the clusters found with each value of @xmath20 .
both metrics agree that @xmath24 is the correct number of clusters ( both curves show a extremum at that point ) , leading to two clusters of size @xmath25 ( cluster @xmath26 ) and @xmath27 ( cluster @xmath28 ) , respectively .
a geographic plot of the location of the cells belonging to each clusters ( @xmath26 and @xmath28 ) provides a fundamental clue to their meaning ( see fig . [ population ]
strikingly , we find a profound correlation between location of cells belonging to cluster @xmath26 ( red dots ) and areas of high population density .
we validate this idea using estimates of the population living within each cell provided by the landscan dataset .
hence , we plot the population distribution boxplot for each cluster in fig . [ population ]
the results clearly confirm our intuition .
cluster @xmath26 corresponds to cells with a typical population that is significantly larger than cluster @xmath28 .
this suggests a natural lexical bipartition of spanish into two superdialects .
superdialect @xmath26 is utilized by speakers in main american and spanish cities and corresponds to an international variety with a strongly urban component while superdialect @xmath28 is comprised mostly of rural areas and small towns .
our result provides some evidence that the increasing globalization of major languages leads to an homogenization that is especially apparent for the active lexicon @xcite .
cities ( our superdialect @xmath26 ) naturally exert an intrinsic linguistic centripetal force that favors dialect unification , smoothing possible lexical differences .
this leveling process present in all countries ( thereby its international denomination ) is reinforced by the rapid increase of worldwide social ties and the powerful influence of mass media precisely located in important metropolitan areas ( madrid , mexico city , miami ) @xcite.several other sociolinguistic aspects ( prestige , higher educational status ) also have a role that is more visible in urban environments .
in contrast , rural areas ( superdialect @xmath28 ) are generally more conservative and keep a larger number of characteristic lexical items and native words . as a result ,
the dialectal area corresponding to superdialect @xmath28 is much more geographically diverse and can be further split , as discussed below .
* characterization of major cluster @xmath28 * geographical representation of regional dialects . for visualization purposes we increased the size of each cell .
three well separated regions are indicated with dashed lines . ]
the size imbalance between the two clusters when combined with our intuition suggest that we can also employ the statistical procedure discussed above to further divide the largest cluster ( @xmath28 ) .
we apply @xmath20-means recursively until the remaining cluster has a similar size to the previous ones . in the end
, we obtain six well defined clusters that we display in fig .
[ cluster ] . clearly , three regions can be distinguished .
yellow dots span a wide area covering mexico , central america , the caribbean and north - western areas of south america .
green dots correspond to the southern cone while blue dots are almost exclusively accumulated within spain .
the first region is quite diverse .
in fact , smaller cells can be aggregated into two additional clusters ( depicted with magenta and orange dots in fig . [ cluster ] ) .
interestingly , the magenta and orange dots seem to be localized in the mexican plateau , the interior of central america and andean colombia , in contrast with the speech of venezuela , the antilles and coastal areas represented with yellow dots .
this division between highland and lowland varieties agrees with classifications discussed previously in the linguistics literature @xcite .
the two regions marked in fig . [ cluster ] partly reflect the settlement patterns and the formal colonial spanish administration within the empire .
conquerors and settlers occupied first the territories of mexico , peru and the caribbean , and only much later colonists established permanent residence in the southern cone , which stayed away from prestigious linguistic norms .
this strong cultural heritage that can still be observed , centuries later , in our datasets deserves to be further analysed in future works .
using a large dataset of user generated content in vernacular spanish , we analyse the diatopic structure of modern day spanish language at the lexical level . by applying standard machine learning techniques
, we find , for the first time , two large spanish varieties which are related to , respectively , international and local speeches .
we can also identify regional dialects and their approximate isoglosses .
our results are relevant to empirically understand how languages are used in real life across vastly different geographical regions .
we believe that our work has considerable latitude for further applications in the computational study of linguistics , a field full of rewarding opportunities
. one can envisage much deeper analyses pointing the way towards new developments in sociolinguistic studies ( bilingualism , creole varieties ) .
our work is based on a synchronous approach to language
. however , the possibilities presented by the combination of large scale online social networks with easily affordable gps enabled devices are so remarkable that might permit us to observe , for the first time , how diatopic differences arise and develop in time .
we thank i. fernndez - ordez for useful discussions . this product was made utilizing the landscan 2007 high resolution global population data set copyrighted by ut - battelle , llc , operator of oak ridge national laboratory under contract no .
de - ac05 - 00or22725 with the united states department of energy .
the united states government has certain rights in this data set .
neither ut - battelle , llc nor the united states department of energy , nor any of their employees , makes any warranty , express or implied , or assumes any legal liability or responsibility for the accuracy , completeness , or usefulness of the data set .
bauer , l ( 2004 ) inferring variation and change from public corpora . in _ the handbook of language variation and change _ , ed . by chambers
j k , trudgill p , and schilling - estes n , backwell publishing , 97:114 tumasjan a , sprenger t , sandner p , welpe i ( 2010 ) predicting elections with twitter : what 140 characters reveal about political sentiment . in : proceedings of the fourth international aaai conference on weblogs and social media : .
178185 . eisenstein j , oconnor b , smith n and xing , e. ( 2010 ) a latent variable model for geographic lexical variation . in : proceedings of the 2010 conference on empirical methods in natural language processing : 1277 - 1287 .
lpez morales , h ( 2005 ) ltimas investigaciones sobre lxico hispanoamericano : unidad y variedad . in : _
homenaje a jos joaqun montes giraldo :
estudios de dialectologa , lexicografa , lingstica general , etnolingstica e historia cultural _ : 333358 . | we perform a large - scale analysis of language diatopic variation using geotagged microblogging datasets . by collecting all twitter messages written in spanish over more than two years
, we build a corpus from which a carefully selected list of concepts allows us to characterize spanish varieties on a global scale .
a cluster analysis proves the existence of well defined macroregions sharing common lexical properties .
remarkably enough , we find that spanish language is split into two superdialects , namely , an urban speech used across major american and spanish citites and a diverse form that encompasses rural areas and small towns .
the latter can be further clustered into smaller varieties with a stronger regional character . * crowdsourcing dialect characterization through twitter * + bruno gonalves @xmath0 , david snchez@xmath1 + * 1 aix - marseille universit , cnrs , cpt , umr 7332 , 13288 marseille , france + * 2 universit de toulon , cnrs , cpt , umr 7332 , 83957 la garde , france + * 3 instituto de fsica interdisciplinar y sistemas complejos ifisc ( uib csic ) , e-07122 palma de mallorca , spain + @xmath2 e - mail : corresponding [email protected] + * * * |
addtoresetequationsection # 1to 3em(#1 ) # 1to 3em # 1([#1 ] ) @#1#2#3 # 1 # 1 = cmbx5 = cmbx6 = cmbx7 = cmbx8 [ 2 ] # 1(#1 ) # 1#2#1 # 1#2#1#2 _ department of physics , hiyoshi , keio university , hiyoshi , yokohama 223 , japan _
since aharonov and bohm ( ab ) have discussed a scattering problem of a charged particle by a solenoid in order to clarify the significance of the vector potential in the quantum theory @xcite , many people have considered the same problem from various viewpoints @xcite@xcite . as is well known
, there are two approaches to deal with a scattering problem in the quantum theory .
the first approach is to find a stationary state describing the scattering process by solving a time - independent schrdinger equation .
the second one is to study the time development of a wave packet with respect to a time - dependent schrdinger equation .
most people as well as aharonov and bohm have analyzed the scattering by means of the first approach , and some people have discussed the same problem with the second approach @xcite , @xcite .
as we see in the following , however , in spite of these efforts it seems not to be clear what is an incident wave in this scattering process . in this paper
we will try to answer this question with the first method because there seems to be lacking a common interpretation of the stationary wave function in the literature .
a system of charged particles interacting with the solenoid is described by a hamiltonian @xmath1 where the electromagnetic vector potential @xmath2 is given by @xmath3 in order to study the scattering of the charged particles we solve the time - independent schrdinger equation @xmath4 to find an eigenfunction which describes the scattering process of charged particles . since the hamiltonian commutes with the angular momentum
, it can be easily shown that a most general solution of is given by @xmath5 where @xmath6 denotes the bessel function of @xmath7-th order and we have put @xmath8 . in , @xmath9 s are arbitrary constants to be determined by phisycal requirements . our main interest here and in the following is to ask what should be the correct choice for the coefficients @xmath9 s to describe the scattering process .
it has been asserted by aharonov and bohm and many other authors that the incident wave for this scattering problem , when the incident beam comes into from the positive @xmath10 axis , should be a modulated plane wave @xmath11 to make the probability current of the incident wave constant . to fulfill this requirement the coefficients have been taken to be @xmath12 . in the case of nonintegral @xmath13 , however , the incident wave becomes a multi - valued function .
we may say that from the physical point of view it seems quite unsatisfactory to take such a multi - valued wave function as an incident one .
we should also note that , for sufficiently large @xmath14 , the term due to the vector potential do not contribute to the dominant part of the current , even if we take a plane wave as an incident wave function .
thus , it is not conclusive to argue that the incident modulated wave gives the condition to determine those constants as @xmath12
. it will be , therefore , instructive to reconsider the scattering problem from other viewpoints .
in this paper we try to find the wave function to describe the scattering process using two different ways .
the first one is to solve the lippmann - schwinger ( ls ) equation , which will be a standard method to consider the scattering problem of a quantum system , taking a plane wave as an incident state instead of the modulated one .
if we adopt the born expansion to solve the ls equation , we will soon meet a difficulty because the perturbative method does not work for the present problem as is shown in @xcite , @xcite .
then we find an exact solution of the ls equation with the aid of the feynman kernel .
the second method is an application of gordon s idea which is proposed to discuss the scattering problem by the coulomb potential @xcite .
the idea may have been introduced to avoid the difficulty caused by the long - ranged nature of the coulomb potential in formulating a scattering theory .
this method will also be useful to examine scattering problems by other potentials with long - range effect .
it will be shown that these two methods give the same wave function to describe the scattering state given by ab @xcite . if @xmath15 , a further discussion is needed since the solution of the schrdinger equation for infinitely thin solenoid does neither vanish nor be defined at the origin in that case . to ensure the impenetrability of the solenoid even for integral @xmath13
, we assume a solenoid to have a finite radius and consider this issue by generalizing the second method .
the plan of this paper is as follows . in section [ kernelkeisan ]
we give the feynman kernel with effects of the solenoid potential using the path integral method .
section [ lsanalysis ] is devoted to solving the ls equation exactly with the aid of the feynman kernel , and the explicit form of the wave function will be obtained . in section [ gordon1 ] gordon s method
will be argued and its generalization to a system of a solenoid with finite radius will be done in section [ gordon2 ] .
conclusions and discussions in comparison with other s results are made in section [ discuss ] .
since the complete set of the eigenfunctions @xmath16 for the hamiltonian with effects of the solenoid are known , we can immediately find an expression of the feynman kernel @xmath17 as its spectral representation @xcite , @xcite @xmath18 to verify this expression , it suffices to notice the following facts , ( i ) it obeys the time - dependent schrdinger equation , ( ii ) it is apparently single - valued with respect to both @xmath19 and @xmath20 , ( iii ) it has the correct limit @xmath21 which follows from @xmath22\ , \label{fourier - bessel}\ ] ] and @xmath23\ .
\label{deltatheta}\ ] ] if we carry out the integration with respect to @xmath24 in , we obtain @xmath25 by use of or we can proceed to solve the ls equation . in this section , however , we would like to give another derivation of because there seems to be some confusion in the path integral construction of the feynman kernel in the literature @xcite@xcite . among them
the most typical one would be the interpretation of its expression as _ the sum over winding number _ such as @xmath26 where @xmath27 comes from paths going around the solenoid @xmath28 times in anticlockwise way , and the factor @xmath29 is a one dimensional representation of the fundamental group of the configuration space . in the following , however ,
we show that _ the sum over winding number is not essential _ , and that the sum in is to be interpreted as a result of the reformulation with aid of the poisson sum formula for the expression obtained using the usual path integral method . to achieve this
we will make use of the completeness relations of both eigenvectors @xmath30 of @xmath31 with eigenvalue @xmath32 , and eigenvectors @xmath33 of @xmath34 with eigenvalue @xmath35 , in formulating the path integral . as is easily recognized
, the single - valuedness of is the consequence of these relations @xcite .
now we give the path integral derivation for the feynman kernel of this system .
defining the exponential operator by @xmath36 and using the completeness of the states @xmath30 , we obtain @xmath37 then , with the aid of the completeness of the states @xmath33 , we can express the infinitesimal version of the feynman kernel as @xmath38\ , \label{hammat1}\end{aligned}\ ] ] where @xmath39 .
after shifting the integration variable @xmath35 by @xmath40 we can rewrite as @xmath41 carrying out the gaussian integration with respect to @xmath35 and noting that @xmath42 we obtain @xmath43 , \label{shortkernel1}\end{aligned}\ ] ] where we have made a change of variables from cartesian coordinate to the polar one and @xmath44 in has been renamed by @xmath45 . to find the kernel for a finite time interval @xmath46 we need to perform @xmath47 integrations with respect to @xmath32 s in . for this purpose ,
the form of the exponent in is extremely inconvenient . to overcome the difficulty it is useful to rewrite @xmath48 where @xmath49 upon integration with respect to @xmath32 or @xmath50 , the gaussian part in the integrand will dominate for sufficiently small @xmath51 .
we may , therefore , regard components of @xmath52 as @xmath53 .
then it follows that @xmath54 ( note , however , that the same argument does not hold true for @xmath55 . ) recalling that we may discard terms of @xmath56 for @xmath57 , in the exponent of a path integral , we can replace the definition of @xmath58 by @xmath59 thus we obtain @xmath60 \nonumber\\ & = & \exp\left [ -i{\mu rr^{\prime}\over\hbar\veps } e^{i\delta}\sqrt{1+\theta_{\veps}^{2 } } \cos(\varphi-\varphi^{\prime}-\theta_{\veps})\right ] \left\{1+o(\veps^{3/2})\right\ } \nonumber\\ & = & \sum_{n=-\infty}^{+\infty } i_{|n|}\left({\mu rr^{\prime}\over i\hbar\veps}e^{i\delta}\sqrt{1+\theta_{\veps}^{2}}\right ) e^{in(\varphi-\varphi^{\prime}-\theta_{\veps } ) } \left\{1+o(\veps^{3/2})\right\ } \ . \label{pathact}\end{aligned}\ ] ] when @xmath51 becomes small , the argument of the modified bessel function grows to allow us to apply its asymptotic form . then keeping terms up to @xmath61 in the exponent , we obtain @xmath62\left\{1+o(\veps^{3/2})\right\ } \nonumber\\ & = & i_{|n+\alpha|}\left({\mu rr^{\prime}\over
i\hbar\veps}e^{i\delta}\right ) \left\{1+o(\veps^{3/2})\right\}. \label{besselch}\end{aligned}\ ] ] substituting it into , we arrive at @xmath60 \nonumber\\ & = & \sum_{n=-\infty}^{+\infty } i_{|n+\alpha|}\left({\mu rr^{\prime}\over i\hbar\veps}e^{i\delta}\right ) e^{in(\varphi-\varphi^{\prime } ) } \left\{1+o(\veps^{3/2})\right\ } \ .
\label{pathact1}\end{aligned}\ ] ] therefore the infinitesimal kernel is now rewritten as @xmath63 here a comment is in need ; in obtaining the result of we have discarded the possibility to use the modified bessel function of negative order since it breaks the regularity of the kernel at the origin .
it is now straightforward to see that the multiplication rule holds : @xmath64 since the integration with respect to the angle variable is trivial and we may make use of a formula @xmath65 which holds for @xmath66 .
repeated use of the rule ( and putting all @xmath44 s to @xmath67 after integration ) will lead us to @xmath68 thus is again obtained by the usual formulation of path integral . here
it should be noticed that the sum over winding numbers in formulating the path integral is not essential .
in this section we obtain the wave function for the scattering state of charged particles scattered by the solenoid .
it is known that , for the present problem , the born approximation fails to give a reliable answer because we can not avoid a divergent integral @xmath69 even in its first order @xcite , @xcite . the iterative method to solve the ls equation
will also be unsatisfactory by the same reason .
therefore we need to solve it in an exact way with the aid of the feynman kernel given in .
the ls equation for the system reads @xmath70 where we have taken an incident plane wave @xmath71 as an eigenstate of the free hamiltonian and @xmath72 indicates the direction of the incident beam . from , we can easily obtain the green s function in the above by laplace transform @xmath73 by putting @xmath74 , it turns out to be @xmath75 where use has been made of a formula @xmath76 , \label{besselkousiki}\ ] ] and @xmath77 is the step function . substituting and partial wave expansion of the plane wave @xmath78 into the integrand of @xmath79 , we obtain @xmath80 where @xmath81 making use of a formula of indefinite integral for cylindrical functions(represented by @xmath82 and @xmath83 for the sake of convenience ) @xmath84 we obtain @xmath85 and @xmath86 by a simple calculation with the aid of lommel s formula @xmath87 we have @xmath88 then can be rewritten as @xmath89 we thus find that the total wave function for the scattering state is given by @xmath90 it is very interesting to recognize that by putting @xmath91 the solution of ls equation coincides with the wave function obtained by ab @xcite .
but we have to remember that in solving the ls equation we take the plane wave as an incident wave and that the resulting scattered wave is given by . in spite of the fact that the total wave function is same as that of ab , both the incident wave and the scattered wave in this section are different from those of ab as a consequence
. next we proceed to find the differential cross section by use of the scattered wave . in view of and
, we notice that @xmath92 while @xmath93 for large @xmath94 .
then we easily obtain the asymptotic form of scattered wave @xmath95 from as @xmath96 where @xmath97 is found from to be @xmath98 using the asymptotic form of the hunkel functions and for @xmath97 in , we are lead to @xmath99 where the phase - shift in @xmath100-th partial wave is given by @xmath101\ge 0 ) \\ + \pi\alpha/2 & ( n+[\alpha]<0 ) \end{array}}\right.\ . \label{phaseshift}\ ] ] here and in the following we denote the integral part of @xmath13 by @xmath102 $ ] and its nonintegral part by @xmath103 to write @xmath104+\{\alpha\}$ ] .
we here introduce a regularization parameter @xmath105 for the sum in so that it is defined as an abel sum because the phase - shift does not decrease at all when @xmath106 becomes large and define @xmath107 as @xmath108\theta}\over\sqrt{2\pi k } } \left\ { { \sum_{n=0}^{+\infty}(e^{-i\pi\alpha}-1)e^{in\theta - n\eps } + \sum_{n=1}^{+\infty}(e^{i\pi\alpha}-1)e^{-in\theta - n\eps } } \right\}\ .
\label{scamp1}\ ] ] then performing the sum of geometric series and making use of a symbolic relation @xmath109 with denoting the principal value by @xmath110 , we finally obtain in terms of the scattering amplitude @xmath107 @xmath111\theta } p{1
\over e^{i\theta}-1}\right\}. \label{psasymp1}\end{aligned}\ ] ] although the total wave function has happened to have exactly same form as the result of ab as mentioned in the above , the scattering amplitude disagrees to that of ab @xcite or @xcite by the @xmath44 function term .
nevertheless , this disagreement can be discarded when we have an interest in the differential cross section only for non - forward direction(@xmath112 ) since we can not well separate the scattered and un - scattered particles in the forward direction experimentally . as far as in the non - forward direction ,
the differential cross section is thus given by @xmath113 in this sense the scattering amplitude of ab describes the physics appropriately .
however , if we take into account the unitarity of the @xmath0-matrix , the @xmath44 function for the forward direction can not be neglected as is pointed out by ruijsenaars @xcite . a more detailed description of the property of the @xmath0-matrix for this system is given in appendix [ smatrix ] .
we here consider the other approach to the problem by using a modified version of gordon s idea which has been proposed in the analysis for the scattering of a charged particle by the coulomb potential @xcite .
the essence of the method is to prepare the asymptotic region described by the free hamiltonian far distant from the solenoid in order to overcome some difficulties caused by the long range effects of the solenoid field . to this aim
we introduce a modified vector potential @xmath114 where @xmath115 is the radius of the shielded solenoid . it should be noticed that in the region @xmath116 the vector potential does not affect charged particles .
to go back to the original ab problem we just put @xmath117 after solving the schrdinger equation for this system . in this section , we first deal with the scattering by an infinitely thin solenoid(@xmath118 ) , and then we generalize the analysis to the case of a finite size(@xmath119 ) solenoid in the next section . in the asymptotic region(@xmath120 ) where the vector potential is absent , the solution of the schrdinger equation is given by eigenstates of the free hamiltonian and the wave function to describe the scattering state @xmath121 will be given by @xmath122 where @xmath123 s are constant coefficients to be determined in the following .
in the scattering region @xmath124 the wave function @xmath125 is subjected to @xmath126.\end{aligned}\ ] ] assuming the partial wave expansion for @xmath125 @xmath127 @xmath128 \psi_{\text{i},n}(r)=0\ ] ] and making a change of variable @xmath129 with @xmath130 , we obtain @xmath131 in the above we have put @xmath132 , @xmath133 , and a prime denotes the differentiation with respect to @xmath134
. the general solution of is given by a linear combination of the whittaker functions @xmath135 where @xmath136 and @xmath9 are arbitrary constants and @xmath137 is defined by @xmath138 the regularity of the wave function at the origin implies that the coefficient @xmath9 of the singular solution @xmath139 in must vanish .
therefore the solution for the scattering region is given by @xmath140 to determine the coefficients @xmath123 in and @xmath136 in , we require the continuity of the wave function itself as well as its derivative in the normal direction on the surface @xmath141 . these conditions may be imposed on each partial wave independently to give @xmath142 here @xmath143 in is defined by @xmath144 and we have put @xmath145 because it should be identified with the phase - shift of @xmath100-th partial wave .
thus , we obtain a solution in the scattering region @xmath146 and also in the asymptotic region @xmath147 we put @xmath117 in and by making use of the well - known asymptotic forms of cylindrical functions and that of the whittaker function @xmath148 for @xmath149 in to obtain the phase - shift @xmath150 then the coefficients of scattered wave in the asymptotic region is found to be @xmath151 when @xmath152 becomes large , the coefficient @xmath153 in the solution of scattering region behaves as @xmath154 .
recalling the definition of the whittaker function and a relation between the hypergeometric functions @xmath155 we observe @xmath156 by recognizing @xmath157 we finally obtain the solution for the scattering region @xmath158 as well as that for the asymptotic region @xmath159 thus we have found that in the large @xmath152 limit the solution @xmath125 in the scattering region has the same form as the one that was obtained in the previous section through ls equation .
it should be noticed , however , that in this approach the wave function @xmath121 describes the scattering state in the asymptotic region far from the solenoid . from
we find that the incident wave is given by the plane wave and that the scattered wave is given by the second term of denoted by @xmath160 @xmath161 if we take the limit @xmath162 of @xmath160 , we again obtain the same scattering amplitude as in the previous section . thus we conclude that the two approach to discuss the scattering problem of this system give the same result . here
it is better to give a comment on the relation between the methods explained here and in the previous section . in finding the scattering amplitude in section [ lsanalysis ] ,
information on the scattering has been given by the asymptotic behavior of the wave function that corresponds to @xmath125 in this section .
on the other hand , in this section , @xmath121 describes asymptotic behavior of the scattering state as has been shown above .
the fact that these two methods give the same result is the consequence of the existence of the limit @xmath117 in both @xmath125 and @xmath121 simultaneously . in other words
, we may say that the aharonov - bohm scattering problem accepts the plane wave as a piece of its asymptotic wave function . in this regard ,
we are reminded of the need of a more careful treatment in the same analysis of the scattering by the coulomb potential .
as is easily seen from or , the whole wave function of the scattering state neither vanishes nor is defined at the origin . therefore the analyses in the preceding sections are unsatisfactory on this point .
fortunately the idea developed in the previous section is easily generalized to the system with a solenoid of finite radius . repeating the same procedure with finite @xmath115
, we obtain @xmath163 where @xmath164 and @xmath165 and the wave function is assumed to vanish in the region @xmath166 . again from the solution in the asymptotic region we can easily find the scattered wave @xmath167 and its asymptotic form for large @xmath10 @xmath168 the first term in the scattering amplitude exactly cancels the corresponding term from the incident plane wave .
therefore the @xmath0-matrix for the system is just a multiplication of a complex number of unit modulus : @xmath169 on each eigenspace of the angular momentum .
hence the unitarity of the @xmath0-matrix is evident .
let us denote @xmath104+\{\alpha\}$ ] again to write @xmath170\theta}\left\ { e^{-i\pi\alpha/2}\sum_{n=0}^{\infty } e^{in\theta}a^{+}_{n}(\alpha ) + e^{i\pi\alpha/2}\sum_{n=1}^{\infty } e^{-in\theta}a^{-}_{n}(\alpha)\right\ } , \label{psisasympf2}\ ] ] where @xmath171 is given in terms of the bessel and the neumann functions by @xmath172 the total cross section is then found to be @xmath173 this result explains an interesting feature of this system : @xmath174 is apparently _ periodic in @xmath13 with period @xmath175_(not @xmath176 ) .
unlike the case @xmath118(extremely thin solenoid ) , the wave function is strictly subjected to the boundary condition @xmath177 even when @xmath15 .
thus the solenoid is completely impenetrable to the charged particles . here
let us consider the special case of @xmath15 .
the explicit form of @xmath178 is found to be @xmath179 for @xmath180 and @xmath181 for @xmath182 .
when @xmath183 tends to @xmath67 , these two formula behave in quite different ways .
the formula is nothing but a total cross section of two dimensional hard core scattering , thus tends to @xmath67 with @xmath184 as @xmath185 .
this result simply means that in the case of @xmath180 charged particles are not affected by the solenoid at all . on the other hand
, the formula grows up to @xmath186 in the same limit since it has the neumann function instead of the bessel function in the numerator of each term .
therefore the total cross section of ab scattering for @xmath182 diverges when the radius of the solenoid tends to @xmath67 .
this singular behavior is the common feature of the total cross section except the case @xmath180 .
if we notice that the partial cross section for large @xmath100 approaches immediately to @xmath187 even for finite @xmath183 , we conclude that the singularity is not the consequence of putting the radius of solenoid infinitely small .
this result implies the important fact that the vector potential can affect the charged particles even in the case of @xmath182 , which is the different conclusion from that of ab .
apart from the divergence of the total cross section considered above , the unitarity of the @xmath0-matrix on each eigenspace of angular momentum is expected from .
this fact is also recognized from the different point of view . according to the discussion given in the appendix [ appa ] ,
the generalized optical theorem , which is rewritten in terms of partial wave decomposition of a scattering amplitude @xmath188 as @xmath189 is equivalent to the unitarity of the @xmath0-matrix . in our problem @xmath190
is given by @xmath191}^{\pm}(\alpha ) e^{\mp i\pi\alpha/2}\ , \label{fnab}\ ] ] where the upper and the lower signs correspond to @xmath192\ge 0 $ ] or @xmath192 < 0 $ ] , respectively .
then the condition reads @xmath193 and is easily verified . as for the system with
infinitely thin solenoid , an explicit form of the @xmath0-matrix is found and the operator identity @xmath194 can be verified directly .
this is given in appendix [ smatrix ] .
in this article the scattering problem , first discussed by aharonov and bohm @xcite , has been reconsidered .
we have examined how the charged particle is scattered by the solenoid taking a plane wave as an incident wave in two ways ; by solving lippmann - schwinger equation ; by applying the gordon s idea to the present situation .
furthermore the scattering problem by a solenoid of finite radius has been considered from the second viewpoint to ensure the impenetrability of the solenoid .
we have shown that the two methods considered in this paper give the same wave function which represents the scattering state and that the results for the infinitely thin solenoid are exactly the same as was obtained by ab as far as the total wave function is concerned .
however , it should be stressed again that the incident wave taken in this paper is different from that of ab to result in the disagreement in the scattering amplitude of this paper to that of @xcite or @xcite by the term proportional to the @xmath44-function .
it is the appearance of the @xmath44-function of forward direction that guarantees the unitarity of the @xmath0-matrix . on this point
, our result agrees with that by ruijsenaars @xcite . in appendix [ appb ]
we discuss the result of ab and give an explanation for the reason why we need the @xmath44-function in addition to the scattering amplitude given by ab .
we also have shown that the @xmath44-function term in the scattering amplitude makes the scattering cross section nonzero even in the case of @xmath182 as well as in the case of nonintegral @xmath13 .
it is better to make a comment for the case of integral @xmath13 .
if we regard a change of variables @xmath195 as a gauge transformation for integral @xmath13 , it seems that there is no ab effect for that case .
but it is not true because this causes a change in the strength of the magnetic field at the origin .
furthermore the new wave function is not defined at the origin , which will brake the single - valuedness of the wave function .
we may here refer the same argument for the vector potential given in @xcite , @xcite .
the authors are grateful to thank professor y. ohnuki and professor s. kamefuchi for many fruitful discussions .
we provide in this appendix a note on two dimensional scattering theory for completeness .
suppose we have two solutions , @xmath196 and @xmath197 , for a scattering problem corresponding to different incident beams of a same energy .
they are assumed to have the asymptotic behavior @xmath198 where the phase factor @xmath199 has been introduced for later convenience .
if we assume the hamiltonian to be hermitian , it is straightforward to obtain @xmath200 as a consequence of the schrdinger equation . taking @xmath14 sufficiently large and using the asymptotic form of the wave functions
, we immediately find @xmath201 this is the generalized optical theorem and is nothing but the @xmath202-number version of the unitarity of the @xmath0-matrix .
to see this , let us define @xmath0-matrix for the wave function given in . from the asymptotic form of the wave function @xmath203 , \label{defsmat}\ ] ] we can find a definition of operators
@xmath204 and @xmath205 @xmath206 then unitarity of the operator @xmath204 reads @xmath207 which is equivalent to . as a special case of or ,
we can easily obtain the optical theorem just by putting @xmath208 @xmath209
taking a plane wave @xmath210 as an eigenstate of the hamiltonian(@xmath211 ) of a free particle , we obtain @xmath212 as solutions of ls equations @xmath213 here we again abbreviate @xmath214 by @xmath7 .
a matrix element of @xmath0 operator is given in terms of @xmath215 and @xmath216 by @xmath217 making use of the explicit form of @xmath218 , we can easily obtain @xmath219 where @xmath220 .
it is then straightforward to see @xmath221 in the same way @xmath222 can be verified .
therefore the @xmath0-matrix of the ab scattering is unitary . by performing the sum in
, we can further rewrite @xmath223\theta}p{1\over{e^{i\theta}-1 } } \right\}\ .
\label{smat3}\ ] ] recalling the expression for the scattering amplitude @xmath107 given in section [ lsanalysis ] , we find a fundamental operator relation @xmath224 furthermore , if we introduce common eigenstates of @xmath211 and of the angular momentum by @xmath225 the operator @xmath204 is diagonalized as @xmath226 to convince us that the solution of ls equation assures the unitarity of @xmath0-matrix as well as its commutability with @xmath211 .
the wave function of the scattering state is given by @xmath227 by use of the integral representation of the bessel functions @xmath228 , \label{besselint}\ ] ] we can immediately convert into its integral representation @xcite , @xcite , @xcite @xmath229 for @xmath230 .
( see fig .
[ fig1 ] for the contour @xmath231 . )
when @xmath13 has an integral part ( @xmath104+\{\alpha\}$ ] ) , the wave function is obtained by @xmath232(\theta+\pi)}\psi_{\{\alpha\}}$ ] .
we may , therefore , consider only the case of @xmath230 .
making a change of variable , we can further rewrite as @xmath233 where the contours @xmath234 and @xmath235 are depicted in fig .
[ fig2 ] . on change of variable @xmath236 ,
there arises a multi - valued function @xmath237 in the integrand .
therefore we need to deal it with due care .
if we recall that our solution for the scattering state , @xmath215 in appendix [ smatrix ] , has been obtained from the ls equation , we immediately notice that we have only one way to deform the contour to adopt the residue theorem to the integral on @xmath238-plane .
( see fig .
[ fig3 ] . ) another option for the deformation , fig .
[ fig4 ] , obviously corresponds to another solution @xmath216 . as is seen from fig .
[ fig3 ] , we can make use of the residue theorem to the contour integration around the unit circle only when @xmath112 .
then we obtain @xmath239 where @xmath240 if we interpret the modulated plane wave @xmath241 as an _ incident wave _ , the second term of will be regarded as a scattered wave .
then we will obtain the aharonov - bohm scattering amplitude @xmath242 with the aid of the stationary phase approximation from @xmath243 so far the amplitude @xmath242 has not been treated in any connection with the @xmath0-matrix of the theory .
here it is important to note that we can not define a @xmath0-matrix from because @xmath244 in is not defined for @xmath245 .
therefore it is inappropriate to decompose the total wave function in the form given above for considering the relation between the @xmath0-matrix and @xmath242 . to find a definition of the @xmath0-matrix for this scattering problem
, we need the asymptotic form of the total wave function @xmath246 \ .
\label{psitotasymp}\ ] ] it should be then compared with and with discussion below in appendix [ appa ] . for the present case
, the @xmath0-matrix should be defined by @xmath247 which is nothing but the result given in . by equating both expressions in and
that in , we find @xmath248 as the relation of the two scattering amplitudes .
therefore @xmath249 can not be interpreted as the total cross section . as a consequence
, @xmath242 does not obey the unitarity condition .
rather , it satisfies an operator relation @xmath250 because @xmath0-matrix itself has been shown to be unitary . in terms of the amplitude itself
, it is expressed as @xmath251 because @xmath242 satisfies @xmath252 3 y. aharonov and d. bohm , phys .
rev . * 115 * , 485 ( 1959 ) .
t. takabayashi , _ hadronic journal supplement _ * 1 * , 219 ( 1985 ) .
p. bocchieri and a. loinger , nuovo cimento , * 47a * , 475(1978 ) .
p. bocchieri , a. loinger and a. siragusa , nuovo cimento , * 51a * , 1(1979 ) . m. v. berry , eur . j. phys .
* 1 * , 240 ( 1980 ) .
s. olariu and i. lovitzu popescu , rev .
* 57 * , 339 ( 1985 ) . c. c. bernido and a. inomata , j. math
* 22 * , 715 ( 1981 ) .
g. mornandi and e. menossi , j. phys .
* 5 * , 49 ( 1984 ) c. c. gerry and v. a. singh , phys .
d * 20 * , 2550 ( 1979 ) .
a. y. shiekh , ann . phys . * 166 * , 299 ( 1986 ) .
d. stelitano , phys .
d * 51 * , 5876 ( 1995 ) .
y. ohnuki , proc .
foundations of quantum mechanics , tokyo , 1986 , pp .
117 - 126 .
w. gordon , zeits .
* 48 * , 180 ( 1928 ) .
s. n. m. ruijsenaars , ann . phys . * 146 * , 1 ( 1983 ) .
y. aharonov , c. k. au , e. c. lerner , and j. q. liang , phys .
d * 29 * , 2396 ( 1984 ) .
bengt nagel , phys .
d * 32 * , 3328 ( 1985 ) . c. r. hagen , phys .
d * 41 * , 2015 ( 1990 ) .
r. jackiw , ann . phys . * 201 * , 83 ( 1990 ) .
m. alvarez , hep - th/9510085 . | the scattering problem under the influence of the aharonov - bohm ( ab ) potential is reconsidered . by solving the lippmann - schwinger ( ls ) equation
we obtain the wave function of the scattering state in this system . in spite of working with a plane wave as an incident wave
we obtain the same wave function as was given by aharonov and bohm .
another method to solve the scattering problem is given by making use of a modified version of gordon s idea which was invented to consider the scattering by the coulomb potential .
these two methods give the same result , which guarantees the validity of taking an incident plane wave as usual to make an analysis of this scattering problem .
the scattering problem by a solenoid of finite radius is also discussed , and we find that the vector potential of the solenoid affects the charged particles even when the magnitude of the flux is an odd integer as well as noninteger .
it is shown that the unitarity of the @xmath0 matrix holds provided that a plane wave is taken to be an incident one .
# 1 |
iras observations of vega detected an infrared excess above the photospheric level , suggesting the presence of circumstellar material ( aumann 1984 ) .
follow - up observations found infrared excesses around other main - sequence stars , including @xmath8 pictoris .
coronographic imaging identified a resolved dusty disk around @xmath8 pictoris as the cause of the infrared excess ( smith & terrile 1984 ) .
@xmath8 pictoris s disk is a prototype for a young , but nearly complete planetary system .
identification of this disk was the first observational evidence of planetary systems outside our own .
many nearby stars have been investigated using iras and iso to search for far - ir photospheric excesses , and a large number have been found ( lagrange 2000 ; backman 1993 ; chen 2005 and references therein ) . due to the large beam size of iras , far - ir observations of nearby stars such as @xmath8 pictoris , @xmath9 eradani , fomalhaut , and
others have resulted in follow - up higher resolution ground based mid - ir observations that resolved their disks and probed their compositions ( lagrange 2000 ) .
stars with far - ir excesses such as @xmath9 eradani may even harbor planets ( hatzes 2000 ) . only a few main - sequence stars with mid - ir but no near - ir excesses have been detected .
aumann & probst ( 1991 ) investigated nearby main - sequence stars using 12 @xmath0 m measurements from iras , but out of the @xmath10 stars investigated , only @xmath11 stars had colors unusual enough to warrant further investigation .
follow up ground - based observations were conducted on @xmath12 stars at @xmath13 @xmath0 m to determine if the excess originated from the star .
only @xmath8 pictoris and @xmath14 leporis exhibited mid - ir excesses ( aumann & probst 1991 ) .
chen & jura ( 2001 ) measured a temperature of 370 k for the mid - ir excess in @xmath14 leporis and suggest that it is due to a massive asteroid belt .
the excess in @xmath14 leporis originates at small radii , less than 6 au , and not at large radii , @xmath1540 au , unlike other stars such as @xmath8 pictoris which have large disks extending hundreds of au ( chen & jura 2001 ) .
stars with higher disk temperatures have dust which is closer to the star , and the excess may be more analogous to the warm temperatures of asteroid belts than to the @xmath16k dust of kuiper belts .
evidence for planets around nearby stars is now overwhelming with over @xmath17 planets in @xmath18 planetary systems , most found using high resolution stellar spectroscopy capable of detecting doppler shifts in the parent star .
however , the process or processes by which these planets form is still unknown .
planetesimals may originate in circumstellar disks , formed as a by - product of star formation .
the evolutionary sequence of both inner and outer circumstellar disks from pre - main sequence to main - sequence is not well understood ( meyer & beckwith 2000 ) .
lack of observational evidence of an evolutionary disk sequence is a major hindrance in the understanding of planetary formation .
the reason for this lack of evidence is that only the nearest stars have been probed for circumstellar disks , and a relatively small number of stars with disks have been identified .
it is estimated that at least 15% of nearby a - k main - sequence stars have dusty debris disks that were detectable to iras and iso sensitivities in the far - infrared ( lagrange 2000 ; backman & paresce 1993 ) .
the _ spitzer _ legacy program includes two surveys investigating the formation of stars and the processes leading to the formation of solar systems : the `` cores to disks '' ( c2d ; evans 2003 ) and `` formation and evolution of planetary systems '' ( feps ; meyers 2001 ) projects . data from these programs will cover a large portion of the spectral energy distribution ( sed ) of stars as well as provide insight into the structure and evolution of circumstellar disks .
the focus of the c2d project is on stellar cores and very young stars with ages up to @xmath19@xmath13 myr .
the feps project focuses on older solar analogues , 3 myr to 3 gyr , to determine disk composition and characterize their evolutionary stages .
the feps and c2d projects are , however , narrowly focused in spatial coverage , with c2d covering @xmath19@xmath20 sq .
degree and feps focusing on @xmath19@xmath21 individual stars out to @xmath19@xmath22 pc . in this present study
, we have utilized a subset of the galactic legacy infrared midplane survey extraordinaire ( glimpse ; benjamin et al .
2003 ) database to try to identify main - sequence stars with mid - infrared excesses . given the sensitivity of the _ spitzer space telescope _
spitzer , we expect to be able to detect unreddened a stars to @xmath19@xmath23 kpc , f stars to @xmath19@xmath24 kpc , g stars to @xmath19@xmath25 kpc , and k stars to @xmath19@xmath26 kpc .
given these distances we will survey a much larger volume of space for stars exhibiting mid - ir excesses than any previous study .
optical spectra were obtained to determine if near and mid - infrared colors could indicate a spectral sequence . in this paper , we present results of a survey of the glimpse observation strategy validation ( osv ) region in the vicinity of the southern galactic star - forming region rcw @xmath3 , where the westerlund 2 open cluster is centered ( churchwell 2004 ) .
we obtained optical spectroscopy for a sample of stars with red or unusual near and mid - infrared colors .
we combine optical photometry and spectroscopy with near and mid - infrared photometry to model the seds of the stars .
once a star is shown to exhibit a mid - ir excess , we model the excess as a single blackbody to gain some insight into its basic properties . in section 2 , we discuss the photometry obtained with glimpse and the method of classification of spectral types . in section 3 ,
we discuss how stars with mid - ir excesses were identified , and in section 4 we compare the stars with mid - ir excesses to the prototype debris disk system , @xmath8 pictoris . in section 5
we discuss the lone o4v((f ) ) star which exhibits a mid - ir excess and its implications for the age and distance of westerlund 2 .
this is a preliminary investigation to determine if warm debris disks can detected with the glimpse survey in the mid - ir .
the glimpse project is one of six spitzer legacy programs . glimpse mapped the galactic plane in four infrared array camera ( irac ;
fazio 2004 ) band passes , @xmath27 $ ] , @xmath28 $ ] , @xmath29 $ ] , and @xmath30 $ ] @xmath0 m from @xmath31 and @xmath32 degrees ( benjamin 2003 ) .
this survey has generated a point source catalog of 3@xmath33@xmath34 objects .
the glimpse program is the largest , most sensitive , mid - infrared survey to date .
the glimpse program mapped rcw@xmath3 as part of the osv ( churchwell 2004 ) .
this region was mapped @xmath13 times with @xmath2 second exposures by irac and is therefore the deepest region surveyed by the glimpse program .
a total of 38,734 mid - infrared catalog sources were found in the osv . to be included in the point source catalog
, a source must have a signal - to - noise ratio of greater than 5:1 with at least two detections in one band and at least one detection in an adjacent band with fluxes greater than 0.6 mjy , 0.6 mjy , 2 mjy , and 10 mjy ( in bands 1 through 4 , respectively ) .
the other two bands need only have a signal - to - noise ratio of greater than 3:1 .
see the glimpse data document or mercer ( 2004 ) , churchwell ( 2004 ) , whitney ( 2004 ) , or indebetouw ( 2005 ) for further descriptions .
we obtained optical spectroscopy of @xmath35 stars in the glimpse osv field surrounding rcw@xmath3 using the hydra instrument on the ctio @xmath23 m on the night of @xmath36 march @xmath26 using three fiber configurations .
the instrument configuration used the large @xmath24 arcsecond fibers and the kgpl1 grating , which provides a dispersion of @xmath37 / pixel , a spectral resolution of @xmath38 , and wavelength coverage of @xmath19@xmath39 from @xmath19@xmath40 .
this wavelength regime covers the majority of the hydrogen balmer series , ca h and k , as well as many metal lines which are necessary for luminosity and temperature classification .
total exposure times were @xmath20 minutes for the first fiber configuration and @xmath41 minutes for the second and third fiber configurations .
sky conditions were photometric .
we selected target stars based on preliminary glimpse photometry in all @xmath23 irac bandpasses along with @xmath24mass , and guide star catalog @xmath42 ( gsc ) photometry .
the stars were selected using optical brightness and infrared colors .
the stars in the first fiber configuration consisted of high - precision parallax - collecting satellite ( hipparcos ) and tycho catalog stars and bright gsc stars with v @xmath43 13 .
bright stars with @xmath24mass / irac colors indicative of possible ir excesses were also placed in the first two fiber fields .
we selected stars with colors indicating an ir - excess or extreme reddening as those which satisfied one or more of the following criteria : flux ratios log @xmath44 @xmath15 -0.5 ( criterion 1 ) , log @xmath45 @xmath15 0.15 ( criterion 2 ) , or log @xmath46 @xmath15 -0.25 ( criterion 3 ) .
we chose these criteria to sample a representative distribution of objects with unusual infrared colors .
criterion 1 sources comprise the reddest @xmath19@xmath47 @xmath48 of sources with visual counterparts , criterion 2 sources comprise @xmath19@xmath24 @xmath48 , and criterion 3 comprise @xmath19@xmath11 @xmath48 .
color - color plots of the data set used in selecting stars are shown in figure [ color ] .
stars that met criterion @xmath26 , @xmath24 , and @xmath49 are plotted as green , red , and blue pluses respectively .
figure [ color ] gives a representative view of the color - color space occupied by the stars meeting one or more of the three criteria .
the second fiber field consisted of stars with gsc photometry and ir colors satisfying one or more of the three color criteria .
the third fiber field consisted of stars meeting one or more of the color criteria , but some stars did not have gsc v - band counterparts , allowing for the inclusion of extremely red stars .
a total of @xmath50 new stellar classifications resulted from these observations .
the stars were classified by eye using comparison spectra ( yamashita 1976 ) to within two comparison temperature classes of the atlas .
there were @xmath51 stars too faint or featureless for classification .
protostars may comprise a significant fraction of these unclassified objects because they have high extinctions and are optically faint .
there are @xmath19@xmath52 stars in the osv that had previous spectral classifications in the literature , but most of them were saturated in glimpse so their mid - ir colors are , therefore , unreliable and are not included in this investigation .
we were able to use 20 stars that had suitable glimpse colors and were previously classified in the literature .
the positions of 38,734 glimpse point source catalog objects in the osv were matched against the positions of @xmath24mass point source catalog objects using the glimpse data reduction pipeline .
we then matched the guide stars with the glimpse objects to produce v , j , h , k , [ 3.6 ] , [ 4.5 ] , [ 5.8 ] , [ 8.0 ] photometry for the 184 spectrally classified stars , although some objects lack photometry at one or more of the eight bandpasses .
the gsc v - band photometry was extracted from photometric plates , and most stars have v - band uncertainties of @xmath19@xmath53 magnitudes , ( see the gsc documentation for more information ) .
the eight photometric data points were fit with kurucz atlas9 models using temperatures and effective gravities corresponding to the nearest spectral type ( kurucz 1993 ) .
since all of the stars lie within the solar circle , we adopt the solar metallicity models .
the kurucz model surface fluxes were scaled by a factor , @xmath54 , so that the model k - band fluxes matched the observed k - band 2mass photometry .
this scale factor represents @xmath55 , the ratio of the stellar radius squared over the distance to the star squared . adopting stellar radii appropriate for the observed spectral types ( de jager , c. & nieuwenhuijzen , h. ( 1987 ) , schmidt - kaler , th .
( 1982 ) , johnson , h.l .
( 1966 ) ) , we calculated a spectrophotometric distance to each star
. these distances and their uncertainties estimated by taking into account the k - band photometric errors added in quadrature with an additional 15% uncertainty in the stellar radii are listed in the tables .
the k - band measurements were used as the basis for the distance computation because the effects of extinction are nearly negligible at 2.2 @xmath0 m for most of our stars .
our derived @xmath56 are all @xmath57 mag implying @xmath58 .
the model atmospheres were reddened with variable amounts of extinction using the li & draine ( 2001 ) extinction curve , except in the mid - infrared ( @xmath59 @xmath0 m ) where the glimpse extinction results were used ( indebetouw 2005 ) .
distance and extinction were fit iteratively as free parameters covering a grid of all reasonable values .
the best fit parameters for each star are determined by the minimum of the @xmath60 statistic .
modeling the star with the next earlier or later spectral type model available in the kurucz library ( i.e. , an a3 with an a0 or an a5 ) produced very similar reduced @xmath60 values .
this allows small errors in classification to yield similar results .
this also provides a check on classification .
if a star is grossly misclassified the fit is poor and the classification is double checked .
we performed a trial fitting procedure in which the effective temperatures and gravities of the stellar models were allowed to vary as free parameters ( i.e. , the program searched through the entire library of kurucz models to find a best fit ) . in nearly all cases , this method produced fits with lower @xmath60 , but this approach frequently converged upon temperature and gravity combinations which were grossly inconsistent with the observed spectral types .
we found that there was often a degeneracy between the amount of extinction and the effective temperature , such that models with high temperatures and low extinctions yielded similar fits as models with low temperatures and high extinctions .
we , therefore , chose to fix the effective temperatures and gravities of the models at value appropriate to the classified spectral type . our data set consisted of @xmath61 main - sequence , giant , and supergiant stars with spectral types in the osv region .
this sample included 20 stars with known spectral types from the literature which had reliable glimpse photometry as well as the 164 sources for which we obtained new classifications .
we calculated ( @xmath62 ) color excesses , e(@xmath62 ) , by taking the differences between the kurucz photospheric model and the photometric ( @xmath62 ) .
we chose the ( @xmath62 ) color because main - sequence stars should exhibit minimal color differences for these wavelengths .
this method is similar to the method used by aumann & probst ( 1991 ) , in which @xmath63 from iras was used to identify warm disk candidates .
figure [ ek8 ] shows the e(@xmath62 ) distribution of stars with ( @xmath62 ) colors .
the distribution shows a peak with a mean near e(@xmath62)=0.08 and a long tail toward larger excesses . the majority of stars , 110 , have e(@xmath62 ) between -0.2 and 0.35 with a mean of 0.08 and a dispersion of 0.09 .
there were 33 stars with excesses larger than three times the dispersion of the main group , i.e. , with an e(@xmath62 ) @xmath15 0.35 .
of these 33 stars with possible mid - ir excesses , three did not have measurements at [ 5.8 ] , and three of the stars had photometric uncertainties placing them within one standard deviation from an e(@xmath62 ) of 0.35 . in order to place a higher reliability on our detections we required that the stars have all four irac measurements , and an e(@xmath62 ) @xmath15 0.35 + 1 @xmath64 .
table @xmath26 gives the colors and uncertainties for the 18 main sequence stars satisfying the above criteria , along with their spectral types and color uncertainties .
hereafter , we refer to this group of stars as the g18 sample .
table 2 contains the stars that had an excess greater than 0.35 but did not meet all of our criteria for inclusion in table 1 .
table 2 include stars which did not have all 4 irac measurements , or stars of early spectral type where the excess could be explained by thermal bremsstrahlung emission ( see 3.1 )
. figure [ colore ] shows color - color plots of the stars with mid - ir excesses .
panel a shows that the stars with a ( @xmath62 ) excess exhibit typical colors for the j , h , k near - infrared , but in panel d they show a significant deviation from the main - sequence which can not be explained by reddening .
stars which could not be classified because of insufficient signal to noise in the optical spectra are shown only in panel b because almost all of these stars were in field 3 and did not have a v band measurement .
triangles denote stars with a mid - ir excess that did not have v - band measurements .
nearly all of the unidentified stars have @xmath45 much redder than the stars exhibiting mid - infrared excesses .
these stars are likely to be highly extinguished and could be protostars .
we analyzed the sed s of the 23 stars meeting the above excess criteria to determine whether phenomena other than disks could explain their mid - ir excesses .
an unresolved close stellar companion can not produce the observed ir - excess .
for most of the stars in table @xmath26 , their mid - ir excesses are only significant at [ 8.0 ] . a companion dwarf star of g , k , or m spectral class would raise the photometric flux at wavelengths @xmath43 @xmath26 @xmath0 m to a much greater degree than in the mid - ir , and such additional stellar components are inconsistent with the observed seds .
even with stellar additional companions , the [ 8.0 ] measurements are still above expected photospheric levels . if any of our targets are multiple star systems , the primary star dominates the sed and a diskless companion would not contribute significantly to the sed .
we investigated the glimpse residual images after psf fitting photometry and subtraction to rule out the possibility of systematic photometric errors .
there were no significant residuals in the pixels at or surrounding each star s location .
if larger photometric uncertainties in high background or highly structured background regions produced spurious signatures of ( @xmath62 ) excesses , we would expect to see as many stars with negative e(@xmath62 ) as positive e(@xmath62 ) .
the distribution in figure [ ek8 ] , however , shows only significant positive e(@xmath62 ) values . if the excess is due to only pah features , we would expect to see an enhancement at [ 3.6 ] , [ 5.8 ] , and [ 8.0 ] with respect to [ 4.5 ] since [ 4.5 ] is devoid of pah features .
none of our excess sources exhibit this enhancement .
thermal bremsstrahlung emission could possibly explain the e(@xmath62 ) values seen in the three early b stars and one o star listed in table 2 .
these stars generally have all @xmath23 irac measurements in excess of the photosphere model values .
an optically thin , thermal bremsstrahlung component was included in the modeling of their seds .
the luminosity of the bremsstrahlung component was allowed to vary as a free parameter along with extinction and distance .
these models with bremsstrahlung components fit the seds much better than the models with a single blackbody component .
a blackbody component model with a single temperature is insufficient to fit all four measurements in excess of the photosphere .
since these are ionizing stars and they fit a thermal bremsstrahlung model , we feel that this is the most likely explanation for the mid - ir excess of these four stars . in conclusion
, we note that the characteristic temperature of the blackbody component which creates a significant excess at 8 @xmath0 m , but not at shorter wavelengths , is @xmath19@xmath65 k because a 400 k blackbody peaks at 7.4 @xmath0 m . even the coolest known brown dwarfs ( burgasser et al .
2000 ; burrows et al .
2001 ) have temperatures of @xmath19@xmath66 k , and jupiter - like planets would be insufficiently luminous to create the observed infrared excesses . we conclude that the source of the mid - ir excess for the stars in table 1 is probably warm @xmath19@xmath65 k dust .
an sed of @xmath8 pictoris , the prototype debris disk system , is shown in figure [ beta ] for comparison to the excess candidates .
figure [ beta ] also shows the appropriate kurucz model .
the sed is comprised of l ( 3.45 @xmath0 m ) , m ( 4.50 @xmath0 m ) , n ( 10.1 @xmath0 m ) , and q ( 20 @xmath0 m ) measurements from backman ( 1992 ) .
the q band measurement was included so that the number of data points and the wavelength range was similar for both @xmath8 pictoris s model and the g18 stars ( 8 measurements and 4 free parameters in the model ) .
the gsc and @xmath24mass measurements were used to model @xmath8 pictoris in a manner equivalent to that for the g18 stars . comparing the @xmath24mass photometry to near - ir measurements in the literature for @xmath8 pictoris shows that the near - ir photometry is consistent , but @xmath24mass data have much larger uncertainties , approximately 20@xmath48 .
our spectrophotometric distance of 22.5 @xmath67 3.0 pc for @xmath8 pictoris is consistent with the hipparcos distance of 19.28 @xmath67 0.19 pc . in order to estimate the sizes and temperatures of any dust disks for the g18 stars
, we modeled the excess in each star as a single black body with temperature and fractional bolometric disk - to - star luminosity ratio , @xmath5 , as free parameters .
this assumption , although simplistic , yields a rough approximation to the temperatures and @xmath5 of the dust disks , given the limited number of photometric points available .
we searched a grid of blackbody temperatures and fractional disk luminosities to find best fits to the observed data .
the temperature grid ranged from 100 k to 1100 k in steps of 1 k. this model was used to estimate the most probable parameters for the dust disks . for our model of @xmath8 pictoris
, we found a single temperature for the excess to be 223@xmath68 k with the disk emitting 0.0019@xmath69 the luminosity of the star . in comparison , gillett ( 1986 )
fit a single temperature blackbody to both mid - ir @xmath70 far - ir data , from which they infer a temperature of 103 k and a fractional luminosity of @xmath71 for @xmath8 pictoris . by fitting the mid - ir data alone we may be overestimating the disk temperature or just measuring the hotter inner portion of a larger disk .
chen & jura ( 2001 ) found the color temperature of @xmath14 leporis to be 370 k , but they note that if they only fit the data at wavelengths greater than 10 @xmath0 m the disk temperature is better fit by a 230 k blackbody . backman ( 1992 ) use a two disk component for @xmath8 pictoris and find the inner disk temperatures to range from 200 - 400 + k depending on the model used .
therefore the temperature we estimate for the debris disk is consistent with the range of disk temperatures given in backman ( 1992 ) from a variety of more sophisticated models .
we investigated the likelihood of other disk temperature / fractional luminosity combinations using a monte carlo simulation to determine the uncertainties on our derived disk and star parameters .
the monte carlo code adds gaussian noise to each of the photometric data points ( based on their 1 @xmath64 photometric errors ) and then re - computes the most probable distance , visual extinction , blackbody temperature , and fractional blackbody luminosity .
figure [ cont1 ] shows a plot of the monte carlo simulation results for @xmath8 pic .
the photometry and model constrain the temperature and fractional luminosity to a small region of the parameter space .
even a few near and mid - ir photometric measurements effectively constrain the temperatures and fractional disk luminosities .
figure [ bhis ] shows the distribution functions of distance ( before adding the 15% stellar radii uncertainty ) , visual extinction , blackbody temperature , and fractional luminosity for @xmath8 pictoris .
both the blackbody temperature and fractional luminosity are well constrained in this simulation .
spectrophotometric uncertainties are primarily due to the assumption for the stellar radii uncertainties but are still well constrained . with only the visual measurement in the optical , there is some variation in extinction but for most stars the extinction is low . distributions for all the derived parameters are nearly gaussian .
we then conducted the same analysis on our candidate stars .
in a small fraction of the monte carlo simulations , @xmath43 5 @xmath48 for most stars , the best fit parameters require a fractional disk - to - star luminosity ratio greater than unity .
we discard these monte carlo iterations as unphysical . in order for the luminosity of the disk to exceed the luminosity of the star
, the star would have to be embedded in a region which required re - emission of absorbed stellar light , or generation of energy by the disk .
this would occur in early protostars , class 0 or i , but not in more evolved pre main - sequence stars .
three stars , g284.1744 - 00.5141 , g284.0110 - 00.1208 , and g283.9773 - 00.3948 , had greater than 10@xmath48 of their monte carlo simulations yield unphysical fractional disk luminosities .
these 3 stars have an excess only at 8 @xmath0 m which provides weaker constraints on the fractional disk luminosity and temperature .
table 3 lists derived parameters and their uncertainties based on the median values of the monte carlo simulation .
figure [ exce ] shows the sed for one of our g18 stars , the a3v star g284.3535 - 00.2021 . the thick solid line is the kurucz model
, the dot - dash line is the sed of the disk component , and the thin line is the combination of the kurucz model with extinction plus the sed of the disk component .
figure [ cont2 ] shows the results for 2000 monte carlo iterations , demonstrating that higher temperatures paired with lower fractional luminosities occur more readily that lower temperatures with higher fractional luminosities .
this plot of fractional disk luminosity versus temperature demonstrates that there is an anti - correlation between fractional disk luminosity and temperature .
when blackbody temperature is high , the fractional disk luminosity is low , and when temperature is low fractional disk luminosity is high .
the median temperature from the simulations is 385 k ( i.e. , @xmath19@xmath72 k warmer than @xmath8 pictoris ) , consistent with the observation that the excess appears at shorter wavelengths in g284.3535 - 00.2021 .
the models constrain the temperature to 385@xmath73 k and the fractional disk luminosity to @xmath74@xmath75 .
figure [ ghis ] shows the distribution of distance ( before adding the 15% stellar radii uncertainty ) , visual extinction , temperature , and fractional luminosity for g284.3535 - 00.2021 determined from monte carlo simulations .
this distribution is much broader than that for @xmath8 pictoris due to our limited ability to constrain maximum possible temperatures with data shortward of 8 @xmath0 m .
longer wavelength data , in both the mid and far - ir , would allow us to better constrain the temperature and fractional luminosity .
table 3 lists the derived parameters and their uncertainties for all 18 main - sequence , non o and b stars with an excess ( in the g18 sample ) . the temperatures for the excesses range from 220 - 820 k with fractional luminosities ranging from 10@xmath6 to 10@xmath7 .
figure [ ahis ] shows both temperature and fractional luminosity ratio histograms for the remaining 17 stars in the g18 sample . for star 11 a large portion of
the fractional luminosity ratio was below 10@xmath6 and is therefore off the scale .
this is because there is a significant excess at only [ 8.0 ] .
g284.0110 - 00.1208 and g284.0719 - 00.1637 have broad distributions , again due to the weaker constraints on the excess .
figure [ ahis ] demonstrates the large variation in temperature and fractional luminosity distribution between stars .
figure [ dist ] shows the distribution of both temperature and fractional luminosity for all g18 stars .
the median temperature for the sample is @xmath76 k. there does not appear to be a correlation between either temperature or fractional disk luminosity with spectral type but the sample is small .
future work may reveal a correlation of either disk temperature or fractional luminosity with spectral type .
lagrange ( 2000 ) suggest a definition of debris disks using four criteria .
these criteria include the bolometric fractional disk - to - star luminosity @xmath43 10@xmath6 , the mass of the gas and dust to be below 10@xmath7 , the dust mass significantly greater than the gas mass , and the grain destruction time much less than the stellar age .
our limited data only allow comparisons based on the fractional bolometric luminosity .
since @xmath5 @xmath43 0.1 for all of the g18 stars , these stars are debris disks or may be in a transitionary phase from protostellar to debris disk .
although these stars have blue spectra which lack emission lines , except g284.3417 - 00.2049 which exhibits h@xmath8 emission , and appear to be main - sequence , we can not rule out h@xmath77 emission .
weak h@xmath77 emission could be present , without h@xmath8 emission , which could imply that the star is pre main - sequence .
we can not rule out the possibility that the stars may have primordial disks , but the disks are most likely to be debris disks .
protostars typically have @xmath5 @xmath15 0.1 and debris disks typically have @xmath5 @xmath78 ( lagrange 2000 ; backman & paresce 1993 ) . since g284.1744 - 00.5141 , g284.0110 - 00.1208 , and g283.9773 - 00.3948 had a large number of their monte carlo iterations discarded because they were unphysical , their fractional disk luminosities may be greater than tabulated in table 3
. these stars may be protostellar in nature since the models allow higher fractional luminosities , but their derived luminosities , omitting the unphysical simulations , are still consistent with debris disks .
mid - ir spectral analysis of the stars in table 3 would reveal whether the 10 @xmath0 m silicate feature is present and would be a useful tool in the characterization of mid - ir excesses around main - sequence stars .
the absorption ( embedded protostar ) , emission ( pre main - sequence ) , or lack of the 10 @xmath0 m silicate feature ( debris disk ) is loosely related to the evolutionary progression of a circumstellar disk ( kessler - silacci 2005 ) .
however , the debris disks studied by kessler - silacci ( 2005 ) contained debris disks that only exhibited far - ir excesses .
@xmath8 pictoris , a mid - ir debris disk system , does have a 10 @xmath0 m silicate emission feature .
a useful comparison would be to determine whether this sample of mid - ir debris disk systems also exhibits a 10 @xmath0 m silicate emission feature .
one especially interesting star that exhibited a mid - ir excess is g284.2642 - 00.3156 .
the optical spectrum of this early type star is shown in figure [ spec ] . using the ob star spectral atlas of walborn & fitzpatrick ( 1990 ) we classify this star as o4v((f ) ) .
the classification is based on the relative strength of , , and the hydrogen balmer series .
figure [ spec ] shows that for this star weak 4634 - 40 - 42 appears in emission and a strong 4686 appears in absorption , which implies that the star is earlier than o5 .
is present in absorption but is very weak , which implies that the star is later than o3 since an o3 lacks absorption .
we did not see any 4058 emission or any significant 4604 or 4620 absorption , and 4686 is in absorption .
these imply that the star is not a supergiant or giant .
figure [ ostar1 ] and figure [ ostar4 ] show glimpse [ 3.6 ] and [ 8.0 ] images in the region surrounding g284.2642 - 00.3156 , including westerlund 2 .
g284.2642 - 00.3156 is marked by the white box .
the [ 3.6 ] image shows that g284.2642 - 00.3156 is surrounded by a grouping of faint stars which are partially blended with the o4 star at the 1.22 arcsecond angular resolution of the irac 3.6 @xmath0 m array .
we determine a spectrophotometric distance to this star of 3233@xmath79 pc with a visual extinction of 5.63@xmath80 .
the [ 8.0 ] image shows a nearly circular ring with radius of @xmath19@xmath81@xmath82 surrounding the o star . at a distance of 3.5 kpc ,
the ring around the star would be 0.65 pc in diameter .
this could be a wind blown shell illuminated by the star .
there is also an irregular linear feature extending 29@xmath82 from the star in the northwest direction .
this feature may be a nearby cloud illuminated by the star , or may trace the relative motion of the star through the inter - stellar medium surrounding westerlund 2 . using the images of moffat ( 1991 ) we identify g284.2642 - 00.3156 as their star @xmath83@xmath84 , which they find to be the most luminous star in the cluster .
they type this star as an o7v , but they note that its absolute luminosity is brighter than that of an o7v .
moffat conclude that since the star appears isolated , it must , by color and luminosity arguments , be a supergiant if it lies at the distance of the cluster . our new classification results in an earlier spectral type with an extinction which is still consistent with cluster membership , but a distance that is not consistent with the 7.9 kpc by moffat ( 1991 ) .
we note that the distance we find for the o4v((f ) ) star is inconsistent with the carraro & munari ( 2004 ) cluster distance of 6.3 kpc but still consistent with the wide variation of distances found for this cluster ( churchwell 2004 ) . if this star is a cluster member it would be the earliest known member .
moffat ( 1991 ) assumed a cluster age of 2 - 3 myr based on the presence of the o7 stars .
the proximity of wr20b to the cluster may imply an older age of 3 - 5 myr if it is also a cluster member ( shara 1991 ) .
the lifetime for an o4 is on the order of 3 myr ( for a 60 star with z=0.02;meynet 1994 ) .
g284.2642 - 00.3156 lies 1 arcminute from the cluster center , which corresponds to a minimum projected separation of 1 pc if the star is at a distance of 3.5 kpc . in order to travel 1 pc in 3 myr or less , a modest space velocity of @xmath15 0.3 km s@xmath85
is required .
thus the reddening , and projected separation of the o4 star are consistent with cluster membership and origin .
this implies that westerlund 2 may be younger and closer than previously thought .
however , g284.2642 - 00.3156 may have been formed subsequent to the formation of westerlund 2 as a result of triggered star formation in surrounding molecular clouds ( deharveng 2005 , oey 2005 ) .
further studies will determine if this star is a cluster member or a result of triggered star formation as well as the distance to this cluster .
high resolution spectra of both the o4 star and westerlund 2 would allow a measurement if their relative radial velocities and would help establish whether this star is a cluster member .
we found 33 stars which exhibited a mid - ir excess above photospheric levels in the field surrounding rcw@xmath3 .
we combined our new spectral classifications with known literature classifications and optical , near , and mid - infrared photometry to model the spectral energy distributions of 184 stars .
stars with mid - ir excesses span all spectral classes from b to early k. we modeled the excess for each star as a single component blackbody and found that for the g18 stars , the excess is consistent with a debris disk or some transitionary phase between primordial circumstellar disk and debris disk . for
these @xmath84 stars the additional black body component was found to have fractional bolometric disk - to - star luminosity ratios , @xmath5 , ranging from ( 10@xmath6 to 10@xmath7 ) with temperatures ranging from 220 to 820 k. these temperatures and fractional disk - to - star luminosities are consistent with warm inner dust of debris disks which could be analogous to asteroid belt type objects .
we would like to thank michael meyer for his comments and suggestions .
we would like to thank our anonymous referee for their useful comments and suggestions .
support for this work , part of the spitzer space telescope legacy science program , was provided by nasa through contract numbers ( institutions ) 1224653 ( uw ) , 1225025 ( bu ) , 1224681 ( umd ) , 1224988 ( ssi ) , 1242593 ( ucb ) , 1253153 ( umn ) , 1253604 ( uwy ) , 1256801 ( uww ) by the jet propulsion laboratory , california institute of technology under nasa contract 1407 .
brian uzpen acknowledges support by the wyoming nasa space grant consortium , nasa grant ngt-40102 40102 , wyoming nasa epscor grant ncc5 - 578 and 1253604 .
cccccccccccccccccccc 1 & g284.353500.2021 & 10250358 - 5741409 & 11.89 & 0.03 & -0.05 & 0.17 & 0.07 & 0.04 & 0.10 & 0.04 & 0.14 & 0.04 & 0.15 & 0.05 & 0.24 & 0.07 & 0.97 & 0.14 & a3v + 2 & g284.174400.5141 & 10224039 - 5751461 & 12.47 & 0.03 & -0.00 & 0.43 & 0.06 & 0.04 & 0.05 & 0.04 & 0.02 & 0.04 & 0.01 & 0.05 & 0.13 & 0.15 & 0.89 & 0.12 & a2v + 3 & g283.884200.3361 & 10213378 - 5733229 & 11.97 & 0.02 & -0.13 & 0.17 & 0.02 & 0.03 & 0.00 & 0.03 & 0.18 & 0.03 & 0.12 & 0.04 & 0.22 & 0.07 & 0.73 & 0.09 & b9v + 4 & g284.124100.2429 & 10232673 - 5736249 & 10.92 & 0.02 & 0.86 & 0.16 & 0.11 & 0.03 & 0.03 & 0.03 & 0.14 & 0.03 & 0.14 & 0.04 & 0.22 & 0.05 & 0.53 & 0.06 & a7v + 5 & g284.018500.1803 & 10230188 - 5729511 & 12.66 & 0.03 & 1.98 & 0.44 & 0.55 & 0.04 & 0.11 & 0.04 & 0.14 & 0.04 & 0.08 & 0.05 & 0.45 & 0.11 & 0.45 & 0.08 & k0v + 6 & g284.054700.5695 & 10214145 - 5750415 & 12.73 & 0.04 & 0.99 & 0.43 & 0.21 & 0.05 & 0.07 & 0.05 & 0.10 & 0.05 & 0.11 & 0.06 & 0.40 & 0.09 & 1.27 & 0.23 & b8v + 7 & g284.071900.1637 & 10232601 - 5730435 & 12.60 & 0.03 & 1.66 & 0.43 & 0.43 & 0.04 & 0.13 & 0.04 & 0.22 & 0.04 & 0.26 & 0.05 & 0.28 & 0.09 & 0.65 & 0.08 & f3iv + 8 & g284.011000.1208 & 10231333 - 5726356 & 12.46 & 0.04 & 1.54 & 0.43 & 0.33 & 0.05 & 0.06 & 0.05 & 0.20 & 0.05 & 0.14 & 0.05 & 0.18 & 0.09 & 0.49 & 0.07 & f5iv + 9 & g284.232000.1670 & 10242577 - 5736016 & 11.24 & 0.02 & 0.61 & 0.43 & 0.15 & 0.03 & 0.06 & 0.03 & 0.09 & 0.03 & 0.11 & 0.03 & 0.25 & 0.05 & 0.61 & 0.04 & a5v + 10 & g284.065800.3254 & 10224485 - 5738431 & 11.51 & 0.02 & 1.35 & 0.43 & 0.34 & 0.03 & 0.07 & 0.03 & 0.11 & 0.03 & 0.08 & 0.04 & 0.22 & 0.06 & 0.55 & 0.06 & g0v + 11 & g283.977300.3948 & 10215463 - 5739217 & 12.38 & 0.02 & 0.02 & 0.23 & 0.04 & 0.03 & 0.00 & 0.04 & 0.03 & 0.04 & 0.05 & 0.05 & -0.01 & 0.08 & 0.37 & 0.08 & b8v + 12 & g283.993500.1944 & 10224911 - 5729455 & 12.21 & 0.02 & 1.68 & 0.43 & 0.45 & 0.03 & 0.11 & 0.03 & 0.19 & 0.04 & 0.13 & 0.05 & 0.30 & 0.07 & 1.05 & 0.06 & g5v + 13 & g283.923900.5103 & 10210640 - 5743271 & 12.09 & 0.02 & 0.97 & 0.43 & 0.27 & 0.03 & 0.03 & 0.04 & 0.12 & 0.04 & 0.14 & 0.04 & 0.22 & 0.07 & 0.70 & 0.08 & f5iv / v + 14 & g283.915300.4337 & 10212181 - 5739183 & 12.80 & 0.03 & 1.34 & 0.43 & 0.30 & 0.04 & 0.07 & 0.04 & 0.05 & 0.05 & 0.07 & 0.06 & 0.41 & 0.12 & 0.94 & 0.13 & f3v + 15 & g284.047800.1686 & 10231576 - 5730119 & 13.33 & 0.05 & & & 0.81 & 0.06 & 0.19 & 0.06 & 0.24 & 0.06 & 0.18 & 0.08 & 0.79 & 0.19 & 0.99 & 0.15 & k5v + 16 & g283.907600.1997 & 10221550 - 5727151 & 10.90 & 0.02 & & & 0.95 & 0.04 & 0.22 & 0.04 & 0.18 & 0.04 & 0.14 & 0.04 & 0.22 & 0.05 & 0.51 & 0.05 & g8iv + 17 & g283.904000.3687 & 10213333 - 5735398 & 12.06 & 0.02 & & & 0.83 & 0.03 & 0.17 & 0.03 & 0.18 & 0.04 & 0.14 & 0.05 & 0.41 & 0.07 & 0.58 & 0.08 & k5v + 18 & g284.341700.2049 & 10245840 - 5741272 & 11.57 & 0.04 & & & 1.39 & 0.06 & 0.36 & 0.06 & 0.34 & 0.05 & 0.33 & 0.05 & 0.43 & 0.07 & 1.05 & 0.12 & f8ve + ccccccccccccccccccccc 19 & g283.940300.2636 & 10221243 - 5731324 & 9.81 & 0.02 & 0.52 & 0.05 & 0.07 & 0.03 & 0.06 & 0.03 & 0.51 & 0.03 & 0.71 & 0.03 & 0.92 & 0.03 & 1.32 & 0.02 & b2v(weak be ) & 1.43 + 20 & g284.172800.2039 & 10235451 - 5736004 & 10.63 & 0.02 & 0.49 & 0.06 & 0.08 & 0.03 & 0.00 & 0.03 & 0.54 & 0.03 & 0.61 & 0.03 & 0.73 & 0.05 & 1.03 & 0.09 & b1v ( weak be ) & 1.08 + 21 & g284.127700.5835 & 10220574 - 5753460 & 10.60 & 0.02 & 3.91 & 0.44 & 0.77 & 0.03 & 0.28 & 0.03 & 0.27 & 0.03 & 0.41 & 0.03 & 0.49 & 0.04 & 1.48 & 0.06 & b5vneb & 1.23 + 22 & g284.033500.2091 & 10230066 - 5731474 & 12.74 & 0.03 & 1.68 & 0.44 & 0.39 & 0.04 & 0.11 & 0.04 & 0.15 & 0.04 & 0.22 & 0.05 & 0.29 & 0.12 & 0.60 & 0.12 & f5ii & 0.57 + 23 & g284.264200.3156 & 10240243 - 5744359 & 8.65 & 0.03 & & & 0.84 & 0.04 & 0.16 & 0.05 & 0.40 & 0.05 & 0.75 & 0.05 & 1.50 & 0.03 & 3.84 & 0.03 & o4v((f ) ) & 4.08 + 24 & g284.050100.2464 & 10225795 - 5734130 & 11.50 & 0.02 & & & 0.98 & 0.03 & 0.23 & 0.03 & 0.23 & 0.03 & 0.19 & 0.03 & 0.29 & 0.06 & 0.75 & 0.06 & g8ii & 0.70 + 25 & g283.9567 + 00.1258 & 10235162 - 5712213 & 13.04 & 0.03 & & & 0.38 & 0.04 & 0.08 & 0.04 & 0.18 & 0.05 & 0.11 & 0.06 & 0.43 & 0.17 & 0.59 & 0.09 & g8iii & 0.54 + 26 & g283.9776 + 00.1738 & 10241078 - 5710353 & 13.03 & 0.02 & & & 0.39 & 0.03 & 0.07 & 0.03 & 0.09 & 0.04 & 0.10 & 0.06 & 0.45 & 0.24 & 0.63 & 0.13 & f5ii & 0.61 + 27 & g284.298000.5951 & 10230784 - 5759512 & 7.49 & 0.02 & 1.44 & 0.03 & 0.58 & 0.04 & 0.38 & 0.06 & 0.44 & 0.03 & 0.70 & 0.03 & 0.90 & 0.03 & 1.24 & 0.02 & b & 1.15 + 28 & g283.983100.5360 & 10212249 - 5746404 & 13.39 & 0.04 & 1.80 & 0.44 & 0.54 & 0.05 & 0.21 & 0.06 & 0.15 & 0.05 & 0.12 & 0.08 & & 1.00 & 1.62 & 0.20 & g2v & 1.60 + 29 & g284.010700.1372 & 10230932 - 5727250 & 14.05 & 0.05 & 2.08 & 0.45 & 0.48 & 0.05 & 0.11 & 0.05 & 0.11 & 0.07 & 0.06 & 0.09 & & 1.00 & 1.58 & 0.20 & f8v & 1.53 + 30 & g283.976400.1365 & 10225656 - 5726168 & 13.36 & 0.04 & 0.63 & 0.02 & 0.63 & 0.05 & 0.23 & 0.05 & 0.30 & 0.06 & 0.36 & 0.07 & & 1.00 & 1.09 & 0.14 & a7v & 1.11 + 31 & g284.473000.2456 & 10253880 - 5747417 & 12.33 & 0.03 & 0.65 & 0.43 & 0.18 & 0.04 & 0.03 & 0.04 & 0.06 & 0.04 & 0.09 & 0.05 & 0.23 & 0.11 & 0.34 & 0.07 & a7v & 0.36 + 32 & g283.930900.0712 & 10225509 - 5721302 & 12.58 & 0.03 & 1.91 & 0.44 & 0.53 & 0.04 & 0.11 & 0.17 & 0.04 & 0.04 & 0.01 & 0.05 & 0.24 & 0.10 & 0.44 & 0.11 & g0v & 0.40 + 33 & g283.980900.1931 & 10224464 - 5730119 & 12.78 & 0.02 & & & 0.43 & 0.04 & 0.11 & 0.04 & 0.12 & 0.04 & 0.05 & 0.05 & 0.12 & 0.17 & 0.52 & 0.21 & f8v & 0.52 + ccccccccccc 1 & g284.353500.2021 & 10250358 - 5741409 & a3v & 1.00 & 1068@xmath86 & 0.00@xmath87 & 385@xmath73 & 0.0043@xmath75 & 8200 & + 4.29 + 2 & g284.174400.5141 & 10224039 - 5751461 & a2v & 0.92 & 1398@xmath88 & 0.00@xmath89 & 300@xmath90 & 0.0059@xmath91 & 8200 & + 4.29 + 3 & g283.884200.3361 & 10213378 - 5733229 & b9v & 0.77 & 1938@xmath92 & 0.00@xmath89 & 482@xmath93 & 0.0018@xmath94 & 9520 & + 4.14 + 4 & g284.124100.2429 & 10232673 - 5736249 & a7v & 0.54 & 563@xmath95 & 0.23@xmath96 & 620@xmath97 & 0.0033@xmath98 & 7200 & + 4.34 + 5 & g284.018500.1803 & 10230188 - 5729511 & k0v & 0.43 & 593@xmath99 & 0.17@xmath100 & 801@xmath101 & 0.0078@xmath102 & 5250 & + 4.49 + 6 & g284.054700.5695 & 10214145 - 5750415 & b8v & 1.24 & 3073@xmath103 & 1.47@xmath104 & 350@xmath105 & 0.0137@xmath106 & 11900 & + 4.04 + 7 & g284.071900.1637 & 10232641 - 5730435 & f3iv & 0.62 & 963@xmath107 & 0.78@xmath108 & 803@xmath109 & 0.0081@xmath110 & 6440 & + 4.34 + 8 & g284.011000.1208 & 10231333 - 5726356 & f5iv & 0.46 & 918@xmath111 & 0.62@xmath112 & 838@xmath113 & 0.0057@xmath114 & 6440 & + 4.34 + 9 & g284.232000.1670 & 10242577 - 5736016 & a5v & 0.61 & 773@xmath115 & 0.41@xmath116 & 508@xmath117 & 0.0030@xmath118 & 8200 & + 4.29 + 10 & g284.065800.3254 & 10224485 - 5738431 & g0v & 0.55 & 493@xmath119 & 0.04@xmath120 & 567@xmath121 & 0.0051@xmath122 & 6030 & + 4.49 + 11 & g283.977300.3948 & 10215463 - 5739217 & b8v & 0.42 & 2798@xmath123 & 0.41@xmath124 & 460@xmath125 & 0.0009@xmath69 & 11900 & + 4.04 + 12 & g283.993500.1944 & 10224911 - 5729455 & g5v & 1.03 & 553@xmath95 & 0.27@xmath126 & 424@xmath127 & 0.0154@xmath128 & 5770 & + 4.49 + 13 & g283.923900.5103 & 10210640 - 5743271 & f5iv / v & 0.71 & 793@xmath129 & 0.00@xmath130 & 542@xmath131 & 0.0050@xmath132 & 6440 & + 4.34 + 14 & g283.915300.4337 & 10212181 - 5739183 & f3v & 0.93 & 1068@xmath133 & 0.39@xmath134 & 395@xmath135 & 0.0115@xmath136 & 6440 & + 4.34 + 15 & g284.047800.1686 & 10231576 - 5730119 & k5v & 0.91 & 608@xmath137 & & 654@xmath138 & 0.0309@xmath139 & 4350 & + 4.54 + 16 & g283.907600.1997 & 10221550 - 5727151 & g8iv & 0.49 & 268@xmath140 & & 723@xmath141 & 0.0066@xmath142 & 5250 & + 4.49 + 17 & g283.904000.3687 & 10213333 - 5735398 & k5v & 0.50 & 338@xmath143 & & 776@xmath144 & 0.0148@xmath145 & 4350 & + 4.54 + 18 & g284.341700.2049 & 10245840 - 5741272 & f8ve & 1.05 & 513@xmath119 & & 560@xmath146 & 0.0108@xmath69 & 6030 & + 4.49 + | _ spitzer _ irac 3.6 - 8 @xmath0 m photometry obtained as part of the glimpse survey has revealed mid - infrared excesses for @xmath1 field stars with known spectral types in a @xmath2 sq .
degree field centered on the southern galactic region rcw@xmath3 .
these stars comprise a subset of 184 stars with known spectral classification , most of which were pre - selected to have unusually red ir colors .
we propose that the mid - ir excesses are caused by circumstellar dust disks that are either very late remnants of stellar formation or debris disks generated by planet formation . of these 33 stars ,
29 appear to be main - sequence stars based on optical spectral classifications .
five of the 29 main - sequence stars are o or b stars with excesses that can be plausibly explained by thermal bremsstrahlung emission , and four are post main - sequence stars .
the lone o star is an o4v((f ) ) at a spectrophotometric distance of 3233@xmath4 pc and may be the earliest member of the westerlund 2 cluster . of the remaining 24 main - sequence stars ,
18 have seds that are consistent with hot dusty debris disks , a possible signature of planet formation .
modeling the excesses as blackbodies demonstrates that the blackbody components have fractional bolometric disk - to - star luminosity ratios , @xmath5 , ranging from 10@xmath6 to 10@xmath7 with temperatures ranging from 220 to 820 k. the inferred temperatures are more consistent with asteroid belts rather than the cooler temperatures expected for kuiper belts .
mid - ir excesses are found in all spectral types from late b to early k. 1.cm |
the study of the mechanism of the fusion - fission process in the reactions with massive nuclei is of interest of both experimentalists and theorists to obtain a favorable way of the synthesis of superheavy elements or exotic nuclei far from the stability line .
the last experiments on the synthesis of superheavy elements @xmath5=114 , 115 , 116 and 118 were successful at beam energies corresponding to 35 - 40 mev excitation energies of the compound nucleus which is enough higher than the bass barrier .
the deformed actinide nuclei were used as targets at synthesis of new superheavy nuclei in the @xmath6ca + @xmath1u , @xmath7pu , @xmath8 cm reactions @xcite .
it means that the orientation angle of the symmetry axis of target - nucleus relative to the beam direction affects on the fusion - fission mechanism .
the cross sections of events corresponding to the synthesis of superheavy elements are not higher than few picobarns @xcite and the width of evaporation residues excitation function is very narrow . at the same time
the measured cross sections of the fission fragments are several tens of millibarn @xcite and the excitation function of fission fragments yields is very wide .
it means that only a very small part of collisions in the narrow range of the beam energy leads to the formation of the evaporation residues considered as superheavy elements .
the problem is to establish this small range of beam energy as the optimal condition for the synthesis of superheavies .
the main reason leading to the small values of the evaporation residue cross - sections seems to be connected with the small survival probability @xmath9 of the heated compound nucleus against fission by evaporating neutrons .
it is well known that the @xmath9 decreases by an increase of the excitation energy @xmath10 and angular momentum @xmath11 of the compound nucleus @xcite .
but the formation of the compound nucleus in reactions with massive nuclei has a hindrance : not all of the dinuclear systems formed at capture of the projectile by the target - nucleus can be transformed into compound nuclei .
we should stress that the estimation of the formation probability is difficult by both experimental and theoretical methods .
the determination of the fusion probability from the experimental data is ambiguous due to the difficulties to identify pure fission fragments of the splitting compound nucleus from the fragments which are formed in other processes of heavy - ion collisions like fast - fission , quasifission and deep inelastic collisions . by the way , restoration of its value from the cross - sections of evaporation residues
is model - dependent . as a result
there is a field for speculations which can be clarified indirectly by the analysis of the physical results connected with the formation of the compound nucleus .
the angular distribution of reaction fragments is one of the informative quantity allowing us to study the fusion - fission mechanism of heavy - ion collisions .
the goal of the present paper is to show the ability of the method based on the dinuclear system ( dns ) concept to calculate the angular momentum distribution for the fusion - fission and quasifission fragments by analyzing the anisotropy of the angular distribution of both fusion - fission and quasifission fragments in reactions with deformed and nearly spherical target nuclei .
the partial capture , fusion and quasifission excitation functions , as well as the corresponding mean square values of angular momentum calculated in this work were used to determine the anisotropy @xmath12 of the fragment angular distribution by formula ( [ anisa ] ) as a function of the spin distribution of the fissioning systems : compound nucleus and dinuclear system .
the results were compared with the experimental data for the observed anisotropy @xmath12 of the angular distributions of fragments of the @xmath0o+@xmath1u , @xmath2f+@xmath3pb and @xmath4s+@xmath3pb reactions .
the paper is organized in the following way . in sect .
[ about ] , we discuss the possibility of the use of the anisotropy of the angular distribution of fragments to establish their origination . in sect .
[ msv ] we present the method to calculate the orbital angular momentum distribution , mean square values @xmath13 , and anisotropy @xmath12 of the angular distribution of the fission and quasifission fragments .
a short presentation about how we calculate capture , fusion , and quasifission excitation functions is given in sect .
[ method ] .
the results of the anisotropy of the fragment angular distribution contributed by the fission and quasifission are calculated and discussed in sect .
[ discuss ] .
conclusions are given in sect .
[ concl ] .
generally , authors analyzing the experimental data determine the fission cross sections by fitting of the measured angular distributions of the fission fragments . to generate the angular momentum distributions required to predict the shape of the fission angular distributions , the approximate fusion cross sections is used ( see , for example , ref .
@xcite ) . in this paper ,
the restored " fission angular distributions were used again and fitted to the measured data for the @xmath2f+@xmath3pb reactions . from the final fits the anisotropies @xmath12 and the fusion - fission cross sections as well as the value of @xmath14 at each bombarding energy were determined . in such an analysis , it was implicitly assumed that @xmath14 is independent * on * the total spin @xmath15 .
the main conclusion of the authors was that the anisotropy at the highest bombarding energies can only be reproduced by assuming that the fission barrier does no longer control the fission process when its height is less than the nuclear temperature .
the role of the quasifission process was not discussed in @xcite . in ref.@xcite ,
the fission fragment anisotropies and mass distributions were measured over a wide range of angles for the @xmath0o+@xmath1u reaction .
the authors concluded that a systematic deviation of the measured fission fragment anisotropies from the transition state model predictions confirms the validity of the correlating anomalously large anisotropies with the presence of quasifission . in refs.@xcite ,
the sensitivity of the features of fission - fragment angular distributions to nonequilibrium processes such as quasifission was shown by a quantitative analysis of the angular distributions of near - symmetric masses produced in the @xmath4s+@xmath3pb reaction .
the study of correlations between mass and angular distribution of fragments of full momentum transfer reactions allows us to separate the pure fission fragments of the compound nucleus with a compact shape @xcite .
the mass and angular momentum distributions of the reaction fragments are determined by the dynamics of collision .
the mass distribution strongly depends on the potential energy surface .
quasifission produces fragments alike the fission fragments confusing the estimation of the fusion cross section . in the quasifission process
, the compound nucleus stage is not reached .
the complete kinetic energy relaxation ( capture stage ) is a main characteristics of the quasifission reactions .
it means that quasifission takes place only after the capture of the projectile by the target - nucleus .
the mass equilibrium can be reached or not in dependence on the masses and mass asymmetry of the reactants @xcite , as well as on the dynamics of collision .
the symmetric mass distributions at all angles and symmetric angular distributions relative to @xmath16 are characteristic features for the fission decay of the completely fused system .
the angular distributions of fission fragments are often characterized by the ratio of the yield at @xmath17 ( or @xmath18 ) to the one at @xmath19 , _
i.e. _ , @xmath20 .
the angular distribution of the fission products is described in the framework of the standard statistical model ( ssm ) usually making use of the fact that the fission saddle - point configuration can be treated as a transition state between the compound system in its quasi - equilibrium state and the two separated fission fragments @xcite .
this model is used under the assumption that the final direction of fragments is given by the orientation of the nuclear symmetry axis as the nucleus passes over the fission saddle - point .
this assumption may not be justified for the heaviest systems for which the saddle- and scission - point configurations have very different shapes .
consequently , in such case , the ssm may not describe properly the angular anisotropy of pure fission fragments in reactions with massive nuclei .
certainly , the discrepancy between theoretical and experimental estimations of the angular anisotropy is caused by an imperfection of the ssm and the experimental difficulties of separating the pure fission fragments .
the models which reproduce the fusion - fission excitation functions fail to account for the fission - fragment angular - distribution data .
the reasons of the failure could be connected with the effects of the entrance channel ( presence of quasifission ) or the fission exit channel ( @xmath21-equili- brium is not reached , where @xmath21 is the projection of the total spin of the nucleus on its axial symmetry axis ) @xcite .
certainly effects of both of above - mentioned phenomena should be analyzed with the increase of the anisotropy @xmath12 in the angular distribution of reaction fragments .
vopkapic and ivanisevic in ref .
@xcite suggested that at sub - barrier energies , fusion of projectile occurs only when the prolate deformed target is oriented in the beam direction , producing a narrow initial @xmath21 distribution peaked around @xmath22 .
the @xmath21 equilibration time was also assumed to be not too short compared to the fission time . using of a time dependent and narrow @xmath21 distribution compared to predictions of the statistical saddle - point model could be envisaged and the fragment angular anisotropy could be explained .
the deviation of the experimental angular anisotropy from the statistical model picture at energies above the interaction barrier was discussed by authors of ref .
the reason of the failure of the statistical model to reproduce the measured data was considered doubtful for the application of the ssm at high energies when it is no longer valid if the fission barrier could be passed at the first attempt .
this deviation occurs at large values of angular momenta where the fission barrier height is less than the saddle - point temperature ( @xmath23 mev ) .
it means that all the properties of the fission process should be determined completely by the dynamics of the motion over the potential energy surface ( pes ) .
the measured fission data corresponding to large values of the fragment angular anisotropy @xmath12 can include a contribution of quasifission fragments leading to higher anisotropy than the ones predicted by standard statistical models @xcite since at quasifission the dinuclear system never becomes as compact to be the compound nucleus , and also the @xmath21 equilibration is probably not attained .
the experimental data ( see , for example , ref .
@xcite ) confirm events with characteristic features , particularly in association with projectiles heavier than @xmath24 mg .
so , there are two main points of view to interpret the experimentally observed angular anisotropy @xmath12 : 1 ) authors of refs .
@xcite and we , in the present paper , explain it with the contribution of the quasifission process competing with the formation of the compound nucleus ; 2 ) authors of ref .
@xcite observe a strict evidence of the anisotropy @xmath12 , explaining such an anomaly by a new version of the preequilibrium fission model @xcite .
calculations within the ssm assume the availability of a realistic spin distribution of the fissioning system . in turn
the calculation of the spin distribution of the compound nucleus formed in the heavy - ion induced reactions is a complicated task .
the extraction of a realistic spin distribution from the measured angular distribution of reaction fragments may be ambiguous due to a sufficient contribution of quasifission .
we consider the role of the entrance channel in the observed angular anisotropy .
the mean square values @xmath25 versus @xmath26 is determined .
we compare our results of @xmath12 with the available estimations extracted from the experimental data @xcite .
the angular distribution of splitting fragments of the rotating system is determined by its angular momentum distribution , namely , by the projection , @xmath21 , of the total spin vector , @xmath15 , onto the center axis of the separated fission fragments and by the moment of inertia of the fissioning system .
the total spin has no component along the beam axis ( @xmath28 ) , if the fissioning system is formed at capture ( full momentum transfer ) of a spinless projectile by a spinless target.we calculate the anisotropy @xmath12 using our results of the angular momentum distributions @xmath13 for the complete fusion and quasifission , and @xmath29 for compound nucleus is found by the rotating finite range model ( rfrm ) by sierk @xcite and for the dns is determined by our model taking into account different mutual orientations of symmetry axes of interacting nuclei .
then we can use the expression for the approximated anisotropy of the fission fragment angular distribution suggested by halpern and strutinski in ref.@xcite and griffin in ref.@xcite : @xmath30 where @xmath31 is the effective moment of inertia on the saddle point for the compound nucleus ; @xmath32 and @xmath33 are moments of inertia around the symmetry axis and a perpendicular axis , respectively .
their values are determined in the framework of the rfrm by sierk @xcite . @xmath29 and @xmath34 are functions of @xmath35 and their values for the given beam energy and orbital angular momentum are found by averaging @xmath36 by the partial fusion cross sections , similar as in formula ( [ uno ] ) .
the mean square values of the orbital angular momentum for the fusion - fission @xmath37 and quasifission @xmath38 processes are calculated by using the partial cross sections of fusion , @xmath39 , and quasifission , @xmath40 , respectively .
the above - mentioned mean square values are found by averaging over all orientation angles of the symmetry axis of deformed nuclei @xcite : @xmath41 with @xmath42 where @xmath43 or @xmath44 , and @xmath45 and @xmath46 are the orientation angles of the axial symmetry axes of the projectile and target nuclei , respectively .
the effective temperature @xmath34 at the saddle point is related to the excitation energy by the expression : @xmath47^{1/2},\ ] ] where @xmath48 and @xmath49 are the reaction @xmath48-value for the ground states of nuclei and the fission barrier height , respectively .
the @xmath49 is calculated in terms of the rfrm by sierk @xcite . for the given @xmath26
we calculate @xmath35 and its value is used to find @xmath49 .
@xmath50 is the mass number of the composite system and @xmath51 the energy carried away by the pre - saddle fission neutrons .
the last was not analyzed in this work .
an important physical quantity in formula ( [ anisa ] ) is the variance @xmath14 of the gaussian distribution of the @xmath21 projection : @xmath52 often @xmath53 is used to fit the angular distribution of fission fragments ( see ref.@xcite ) . for the estimation of the anisotropy of quasifission fragments we calculate @xmath29 for the dinuclear system taking into account the possibility of different orientation angles of its constituent nuclei ( see appendix a ) .
the excitation energy of the dinuclear system is found as a sum of the difference between the beam energy and the minimum of the potential well of the interaction potential and the @xmath48-value corresponding to a change of the excitation energy of the dinuclear system from the projectile - target configuration to the quasifission fragments . assuming that after capture the mutual orientations of the dns nuclei do not change much
, we calculate @xmath29 for collisions of the projectile and target with different time - independent orientations of their symmetry axes . under this assumption , and using our calculated mean square values @xmath13 for quasifission we determine the angular anisotropy @xmath12 of the quasifission fragments .
the effective value @xmath29 of the moment of inertia of the dinuclear system ( @xmath54 ) is found by averaging on all the @xmath55 values and orientations @xmath56 with the partial capture cross sections , for a given collision energy : @xmath57 where @xmath58 .
the values of @xmath59 for the fragments of quasifission are higher than the ones of the compound nuclei @xcite .
this kind of fission - like decay produces a high anisotropy in the angular distributions due to the large angular momentum of dns @xcite , because the partial cross section of quasifission increases by increasing of @xmath55 @xcite .
the reason is that the hindrance for the transformation of the dinuclear system into compound nucleus increases due to an increase of the intrinsic fusion barrier @xmath60 with the orbital angular momentum @xmath55 . at the same time quasifission barrier
@xmath61 decreases by increasing of @xmath55 @xcite .
we determine these barriers of the dns model in sect .
[ fusion ] .
the dissipation of the initial orbital angular momentum @xmath62 of collision during the capture process and the maximum value @xmath63 of the partial waves leading to capture are calculated by the solution of the corresponding equation of motion .
the results show that such a dissipation is considerable and the value of angular momentum after dissipation @xmath64 is about 2530% lower than the initial value @xmath62 .
this value is found by the solution of the equations of motion for the orbital angular momentum and radial motion of nuclei .
details of these calculations can be found in refs .
the possibility to calculate the spin distribution of the compound nucleus ( its angular momentum distribution ) is the advantage of the used method based on the dinuclear system concept @xcite . in the next sect .
[ capture ] we present shortly the basic points of the model .
it should be stressed that , in the case of collisions of deformed nuclei , the orientation angles @xmath65 of the symmetry axes to the beam direction play an important role at the capture and complete fusion stages .
the importance of the orientation angles of the symmetry axes of the reacting nuclei was analyzed in ref.@xcite .
the final results of the capture and complete fusion are obtained by averaging the contributions calculated for different orientation angles of the symmetry axes of the reacting nuclei with formula ( [ sigmal ] ) .
at the early stage of the reaction with massive nuclei , the complete fusion of colliding nuclei has a very strong competition with the quasifission process which decreases the probability of the compound nucleus formation . the fusion and quasifission
are considered as a two stage process @xcite : ( i ) the formation of a dinuclear system as the result of the capture of the projectile - nucleus by the target - nucleus ; ( ii ) the transition of the dinuclear system into the compound nucleus ( complete fusion ) as a special channel of its evolution .
the other alternative way of the dinuclear system evolution is quasifission .
the quasifission is the decay of the dinuclear system without formation of the compound nucleus .
both processes can produce fragments with similar characteristics as total kinetic energy and mass distributions .
the ratio of yields from both channels depends on the structure of the pes ( see , for example , fig .
[ figdns]a ) which is different for the different total mass and charge numbers . in fig .
[ figdns ] , we presented pes calculated for reactions leading to @xmath66pa .
the probability of quasifission is determined by the relief of pes of the dinuclear system calculated as a function of the relative distance and mass asymmetry . in fig .
[ figdns]b , the curve connecting minimums of the valley on the pes is the driving potential as a function of the charge asymmetry of the dns fragments .
the cut of pes for the given charge number is the nucleus - nucleus interaction potential @xmath67 .
the curve in fig .
[ figdns]c was calculated for the @xmath2f+@xmath3pb reaction .
the size of the potential well is determined by the orbital angular momentum leading to capture and the depth of the potential well is the quasifission barrier @xmath61 for the given charge asymmetry . for the interacting deformed nuclei pes
depends on the orientation angles of the symmetry axes ( see formula [ pes ] ) .
the partial capture cross section is determined by the capture probability @xmath68 which means that the colliding nuclei are trapped into the well of the nucleus - nucleus potential after dissipation of a part of the initial kinetic energy and orbital angular momentum : @xmath69 here @xmath70 is the de broglie wavelength of the entrance channel .
the capture probability @xmath71 is equal to 1 or 0 for the given beam energy and orbital angular momentum .
our calculations showed that in dependence on the beam energy , @xmath72 , there is a window for capture as a function of orbital angular momentum : @xmath73 where @xmath74 can be observed when the beam energy is large than the coulomb barrier ( @xmath75 ) .
it means that the friction coefficient is not so strong to trap the projectile into the potential well ( see fig.2 in ref.@xcite ) .
the number of the partial waves giving a contribution to the capture is calculated by the solution of the equations for the radial and orbital motions simultaneously @xcite .
they are defined by the size of the potential well of nucleus - nucleus potential @xmath76 and the values of the radial @xmath77 and tangential @xmath78 friction coefficients , as well as by the moment of inertia for the relative motion . here
@xmath79 and @xmath80 are the charge numbers , deformation parameters and orientation angles of the symmetry axes of nuclei , @xmath81 and @xmath82 correspond to quadrupole and octupole deformations , respectively .
the nucleus - nucleus interaction potential , radial and tangential friction coefficients and inertia coefficients are calculated in the framework of our model @xcite . for the capture events we calculate the competition between quasifission and complete fusion with a statistical approach @xcite .
the competition between the complete fusion and quasifission is obtained as the branching ratio between the transition of the dinuclear system from its position in the valley of the pes to the fusion lake " overcoming the intrinsic fusion barrier @xmath83 on the charge number axis ( complete fusion ) , and the decay to the quasifission sea " after overcoming the quasifission barrier @xmath61 on the radial distance axis .
the size of the potential well decreases by increasing the orbital angular momentum @xmath55 , _
i.e. _ the valley of pes becomes shallow and as result the lifetime of the dns decreases . at the same time
the intrinsic fusion barrier @xmath83 increases while the quasifission barrier @xmath61 decreases .
so , we conclude that the contribution of the quasifission increases by increasing the angular momentum for a given beam energy @xcite . the intrinsic fusion barrier @xmath83 for a given projectile - target pair is the height of the saddle - point in the valley of the pes along the axis of the dns charge asymmetry .
the pes of the dinuclear system is calculated as a sum of binding energies of interacting nuclei , nucleus - nucleus interaction potential @xmath67 and rotational energy , @xmath84 where @xmath85 , @xmath86 , @xmath87 , and @xmath88 are binding energies of the constituent nuclei of dns and compound nucleus , respectively ; @xmath89 is charge number of compound nucleus , @xmath90 is the nucleus - nucleus interaction potential of the dns nuclei ; @xmath91 and @xmath92 are rotational energies of dns and the compound nucleus . @xmath93 and @xmath94 are deformation parameters and orientation angles of axial symmetry axis of interacting nuclei .
the binding energy values are obtained from the tables @xcite ) .
the dependence of pes on the shell structure of the nuclei forming the dinuclear system and on the orbital angular momentum leads to the strong influence of the entrance channel ( @xcite ) .
the effects connected with the entrance channel appear in the partial fusion cross section @xmath95 , which is defined by the product of the partial capture cross section and the related fusion factor @xmath96 presenting the competition between complete fusion and quasifission processes : @xmath97 where @xmath98 is the partial capture cross section which is defined by formula ( [ parcap ] ) .
the details of the calculation method are described in ref.@xcite .
@xmath99 is the hindrance factor for formation of compound nucleus connected with the competition between complete fusion and quasifission as possible channels of evolution of the dns . due to nucleon transfer between the dns constituents the wide range of charge asymmetry
can be populated in dependence on landscape of the potential energy surface .
this phenomenon is taken into account by using of the branching ratio @xmath99 as the sum of ratios of the widths related to the overflowing over the quasifission barrier @xmath100 at a given mass asymmetry , over the inner fusion barrier @xmath101 on mass asymmetry axis to complete fusion and over @xmath102 in opposite direction to the symmetric configuration of dns : @xmath103 where @xmath104 is the excitation energy of dns for a given value of its charge - asymmetry configuration @xmath105 and @xmath106 ; @xmath107 is the minimum value of the nucleus - nucleus potential well ; the position of the minimum is marked as @xmath108 ( see fig .
[ figdns]c ) ; @xmath109 is the change of the @xmath48-value by changing the dns charge asymmetry ; @xmath110 is the probability of population of the configuration @xmath105 at @xmath111 , @xmath55 and given orientation angles @xmath112 .
@xmath110 was obtained by solving the master equation for the evolution of the dinuclear system ( charge ) mass asymmetry ( for details see refs .
@xmath113 and @xmath114 correspond to the point where the driving potential reaches its maximum ( see fig . [
figdns]b ) @xcite .
the branching ratio @xmath115 is calculated as a ratio of widths related to the overflowing over the quasifission barrier @xmath100 at a given mass asymmetry , over the intrinsic barrier @xmath101 on mass asymmetry axis to complete fusion and over @xmath102 in opposite direction to the symmetric configuration of dns : @xmath116 here , the complete fusion process is considered as the evolution of the dns along the mass asymmetry axis overcoming @xmath101 ( a saddle point between @xmath117 and @xmath118 ) and ending in the region around @xmath117 or @xmath119 ( fig .
[ figdns]b ) .
the evolution of the dns in the direction of the symmetric configuration increases the number of events leading to quasifission of more symmetric masses .
this kind of channels are taken into account by the term @xmath120 .
one of the similar ways was used in ref.@xcite in calculations of the evaporation residues cross sections in the reactions with actinides . the widths of these `` decays '' leading to quasifission and complete fusion can be presented by the formula of the width of usual fission @xcite : @xmath121 where @xmath122 ; @xmath123 , @xmath61 , and @xmath124 .
@xmath125 is the temperature of the dinuclear system consisting of fragments with charge numbers @xmath5 and @xmath126 : @xmath127 .
usually the value of the factor @xmath128\right)\ ] ] in ( [ gammai ] ) is approximately equal to the unit . inserting eq .
( [ gammai ] ) in ( [ gammas ] ) , we obtain the expression ( [ pcn ] ) used in our calculations : @xmath129
in this section we compare the capture , fusion and quasifission excitation functions with the available experimental data which were obtained by the analysis of the angular distribution of fission fragments .
the calculated partial capture , fusion and quasifission excitation functions in this work are used to determine the mean square values of angular momentum @xmath130 for the dinuclear system and compound nucleus , and to find the anisotropy @xmath12 of the angular distribution by formula ( [ anisa ] ) .
the results for @xmath12 are compared with the experimental data .
[ 3figo16u]a shows that at low energies the calculated cross sections are nearly equal and are in good agreement with the experimental data for the @xmath0o+@xmath1u reaction .
the measured data is related by a mixture of the complete fusion and quasifission fragments in equal proportions . in the energy interval
@xmath131 mev the contribution of fragments of the fusion - fission process dominates over the one of quasifission ( see fig .
[ 3figo16u]a ) .
an increase of the quasifission contribution by increasing the beam energy is explained by an increase of events with the formation of the dinuclear system with the large orbital angular momenta because the intrinsic fusion barrier @xmath60 increases and the quasifission barrier decreases by increasing of @xmath55 ( see refs .
@xcite ) .
the authors of ref .
@xcite in detail analyzed the angular anisotropy of fragments at low energies to show the dominant role of the quasifission in collisions of the projectile with the target - nucleus when the axial symmetry axis of the last is oriented along or near the beam direction .
it was obtained large values of the anisotropy at low energies and these data were assumed to be connected with the quasifission because a mononucleus or dinuclear system formed in the near tip collisions has the elongated shape .
this shape can be far from the one corresponding to the saddle point @xcite .
therefore , for such system , there is a hindrance at its transformation into compound nucleus .
the comparison of our calculated anisotropy values for the quasifission and complete fusion fragments with the ones which were presented in refs .
@xcite shows that the anisotropy @xmath12 connected with quasifission is very close to the experimental data ( fig .
[ 3figo16u]b ) at low energies @xmath132 mev . in spite of the quasifission and fusion - fission cross section
are close , the fusion - fission fragments show less anisotropy due to small @xmath25 values and large @xmath29 . the large anisotropy for quasifission fragments
is explained by small temperature @xmath133 and small effective moment of inertia @xmath134 . because the fragments under discussion were formed in collisions with orientation angles @xmath135 for the target symmetry axis .
figure [ 3figo16u]c shows the comparison of the calculated mean square of the angular momentum @xmath13 of the fissioning systems ( dinuclear system and compound nucleus ) with the data extracted from the measured angular distribution of fragments in ref.@xcite .
the agreement of the fusion and quasifission angular momentum distributions with the experimental data is well for all values of beam energy excluding of point @xmath136=90 mev .
the authors of ref .
@xcite concluded that in the sub - barrier region , in this reaction , the contribution of the quasifission is negligible .
this conclusion has been made by a comparison of calculated excitation functions for the evaporation residues with the experimental data @xcite .
they did not need to include a hindrance to fusion to reproduce the experimental data .
authors used the coupled - channel code ccdegen @xcite , which is based on a version of the ccfull code described in @xcite to calculate the fusion excitation function and the results were used as input for the statistical model calculation of evaporation residue cross sections by using the code hivap @xcite .
one can see that even in the case of using a deformed target - nucleus , at sub - barrier energies the yield of fusion - fission fragments are comparable with the yield of quasifission fragments . in fig.[3figf19pb]a , we show the comparison of the excitation functions for the capture , complete fusion and quasifission calculated for the @xmath2f+@xmath3pb reaction in the framework of our model with the experimental data of the fission cross sections presented in refs.@xcite .
our results for the complete fusion are in agreement with the experimental data .
it means that the contribution of quasifission in the measured data is small .
we should note that the maximum of the calculated mass ( charge ) distribution of quasifission fragments is near masses of the projectile - like and target - like fragments of the @xmath2f+@xmath3pb reaction .
the absent of the large anisotropy at the low energies is explained by the fact that the massive target nucleus has the spherical shape and @xmath137 is not so small as in the @xmath0o+@xmath1u reaction . in fig .
[ massf19 ] we show the time dependence of the charge distribution @xmath138 for yield of products of the quasifission processes .
it was obtained by solving of the master equation for the evolution of the dinuclear system ( charge ) mass asymmetry ( for details see ref .
it is seen that the light fragments of quasifission have a charge less than @xmath139 and the asymptotic value is @xmath140 .
heaviest fragments have a charge larger than @xmath141 and the largest value is @xmath142 .
this result is caused by the influence of the shell structure of interacting nuclei .
moreover , the fragments around the initial charge number @xmath139 at time @xmath14310)@xmath144s can be considered as fragments of deep - inelastic collisions when the dinuclear system is formed for the short times ( no capture ) . from the good agreement of our results on the excitation function of the complete fusion and the angular anisotropy @xmath12 for the @xmath2f+@xmath3pb reaction ( see fig . [ 3figf19pb]a and fig .
[ 3figf19pb]b ) with the experimental data presented as the cross section of the fusion - fission process we can conclude that in this reaction the fusion - fission mechanism dominates over quasifission mechanism . in fig .
[ 3figf19pb]c our theoretical results are compared with the values of @xmath59 extracted from the description of the experimental results on the angular anisotropy @xmath12 of the @xmath2f+@xmath3pb reaction @xcite .
the good agreement between the calculated and experimental results confirms the correctness of the angular momentum distribution for the complete fusion and quasifission calculated with our model .
the dominant role of quasifission can be seen in the @xmath4s+@xmath3pb reaction which is a more symmetric than the above discussed two reactions .
a sufficient role of the quasifission in this reaction was suggested by the authors of the experiment in ref.@xcite .
but they did not present quantitative results of the ratio between complete fusion and quasifission contributions .
it is well known that this is very complicated task due to the strong overlap in mass and angular distributions of the fragments from both processes .
we have theoretically analyzed the contributions of the above mentioned processes . in fig .
[ 3figs32pb]a , we compare the excitation functions for capture , complete fusion and quasifission calculated for this reaction with the experimental data for the fission cross section presented in ref .
our results for complete fusion are lower than the experimental fission cross sections .
our statement is that the data contain a large amount of contributions of the quasifission fragments together with fusion - fission fragments .
the ratio of the yields of quasifission fragments to fusion - fission fragments is larger at the lower and higher beam energies . .
at the low energies the competition between complete fusion and quasifission is very sensitive to the peculiarities of the potential energy surface .
the height of the intrinsic fusion barrier @xmath60 is comparable with the excitation energy of the dinuclear system and , therefore , there is a hindrance to complete fusion . at the highest values of beam energy
the hindrance to complete fusion appears due to the increase of @xmath60 as a function of the orbital angular momentum @xmath145 of the dinuclear system . at the same time the quasifission barrier @xmath61 decreases by increasing of @xmath145 .
the combined effect from the behaviour of the @xmath60 and @xmath61 barriers as functions of @xmath145 makes quasifission as the dominant process @xcite .
this phenomenon is common for all reactions with projectiles heavier than @xmath0o .
the fast - fission mechanism also contributes to the anisotropy of the angular distributions of fragments in all of the above mentioned reactions at large values of @xmath55 . according its definition the fast - fission mechanism takes place when complete fusion occurs at large orbital angular momentum but there is not a fission barrier for the being formed compound nucleus .
the range of angular momentum leading to the fast - fission is @xmath146 where @xmath147 is a value at which the fission barrier for compound nucleus disappear ; @xmath148 is a maximum value of @xmath55 at which complete fusion takes place .
the value of @xmath148 is different for different orientation angles of the colliding nuclei .
note that quasifission can take place at small values of angular momentum including @xmath149 in contrast to the fast - fission which occurs at @xmath150 where @xmath151 is the minimum value of angular momentum of the compound nucleus where the fission barrier disappears and @xmath152 is the maximum value of orbital angular momentum leading to capture . in fig.[spinspb ] , we present as an example our results of the angular momentum distribution of the partial fusion and quasifission cross sections for the @xmath4s+@xmath3pb reaction .
we estimated the contribution of the fast - fission in the @xmath4s+@xmath3pb reaction .
[ 3figs32pb]a shows , the fast - fission cross section is small in comparison with the one of the other processes . at large beam energies ,
the fast - fission and complete fusion cross sections are comparable .
the energy - dependences of anisotropy of the angular distribution of reaction fragments are shown in fig .
[ 3figs32pb]b .
the contribution of fusion - fission process is small in comparison with the data from ref .
this fact shows the dominance of quasifission fragments in the measured anisotropy of the angular distribution .
the dotted line in fig.[3figs32pb]b is obtained by averaging the anisotropies @xmath153 , @xmath154 and @xmath155 corresponding to the quasifission , fusion - fission and fast - fission fragments , respectively , according to the formula : @xmath156 where @xmath157 , @xmath158 , and @xmath159 are the quasifission , fusion , and fast - fission cross sections , respectively ; @xmath153 , @xmath154 , and @xmath155 are the corresponding anisotropies . in fig.[3figs32pb]c
, we compare our theoretical results with the values of @xmath59 extracted from the description of the experimental data on the angular anisotropy @xmath12 of the @xmath4s+@xmath3pb reaction in ref .
the extracted values of @xmath59 from the measured data are in good agreement with our theoretical results for the fusion - fission process .
note that at lower and higher values of the beam energy , the curve for the quasifission is closer to the experimental data .
this confirms our conclusion of the role of quasifission made above in the discussion of fig.[3figs32pb]a . in order to comment the overestimation of the measured data by our results for the quasifission excitation function
, we calculated of evolution of the charge distribution in the dinuclear system as in the case of the @xmath2f+@xmath3pb reaction .
the maximum of the charge distribution of the @xmath4s+@xmath3pb reaction splits into two peaks with increasing interaction time .
it is seen from fig.[chardisspb ] that the first maximum of the lightest components is concentrated around the charge number @xmath160 and the other maximum around the charge value @xmath161 .
this effect is connected with the influence of shell structure of the interacting nuclei .
it seems to us that the quasifission fragments around @xmath160 were not registered as fission fragments .
this is seen from fig .
but such fragments are included in our results of the quasifission contribution . as a result we overestimated the experimental data of refs.@xcite ) for the yield of binary fragments of the full momentum transfer reactions .
the fragments around the initial charge number @xmath162 at time @xmath14310)@xmath163s in fig.[chardisspb ] can be considered as fragments of deep - inelastic collisions when the dinuclear system is formed for short time ( no capture ) .
the model based on the dinuclear system @xcite was improved to study the influence of the quasifission on the angular anisotropy of the measured fission - like fragments .
we compared our calculated results with the experimental data on the excitation function of the fusion - fission fragments , anisotropy @xmath12 of the angular distributions and mean square values @xmath59 extracted from the description of the measured anisotropy @xmath12 for the @xmath0o + @xmath1u @xcite , @xmath2f + @xmath3pb @xcite and @xmath4s + @xmath3pb @xcite reactions .
the experimental studies of the angular distributions of fragments in heavy - ion reactions show distinct deviation from the ssm theory .
we explain this deviation by the presence of a contribution of the quasifission reactions .
the importance of the quasifission mechanism at the low beam energies was used to explain the large anisotropy of in the angular distribution of fragments of the full momentum transfer reaction was discussed in ref .
our results obtained taking into account contributions of different orientation angles of the symmetry axis of the deformed @xmath1u target to the measured anisotropy @xmath12 confirm this interpretation . at the low beam energies we observe
capture ( formation of the relatively long living dns ) only for @xmath135 and the angular distribution of the quasifission fragments shows large anisotropy @xmath12=1.72.0 .
the small values of @xmath164 and dns temperature @xmath133 are responsible for this phenomenon . in the reactions with the spherical @xmath3pb target ,
@xmath164 is not so small to cause in @xmath2f + @xmath3pb the similar effect observed in the @xmath0o + @xmath1u reaction .
the calculated fusion and quasifission cross sections are nearly equal and are in good agreement with the experimental data for the @xmath0o+@xmath1u reaction .
therefore , we conclude that the measured data is related by a mixture of the complete fusion and quasifission fragments .
considering the quasifission as a fission " of the dinuclear system from a not compact shape we estimate the mean square values of the angular momentum @xmath55 and anisotropy @xmath12 of the angular distribution of the reaction fragments .
the experimental data of the anisotropy @xmath12 are described if we also take into account the contribution of the quasifission fragments . for the @xmath2f + @xmath3pb reaction @xcite
the contribution of the quasifission fragments is comparable at low energies with the one of fusion - fission mechanism and the last mechanism become dominant for the beam energy @xmath165 mev .
so in the @xmath2f + @xmath3pb reaction the fusion - fission fragments give the main contribution to the measured data .
the effect of quasifission appears only at more higher beam energies .
the analysis of the measured data for the @xmath4s + @xmath3pb reaction showed the dominant role of quasifission in this reaction .
it was determined by the comparison of the calculated fusion and quasifission cross sections , anisotropies @xmath154 and @xmath153 connected with these processes , as well as @xmath166 with the corresponding experimental data .
we conclude that the appearance of the competition between quasifission and complete fusion depends on such parameters of entrance channel of reactions as mass asymmetry , orbital angular momentum and beam energy .
this conclusion supports the statement of b. back _ et al .
_ @xcite and m. tsang _
_ @xcite that the assumption of the fusion ( and formation of a truly equilibrated compound nucleus ) during the first step of the reaction is not valid in the analysis of the experimental data of fission fragments . therefore , these authors hypothesized the quasifission contribution on the experimental data in order to describe the angular anisotropy of the detected fragments .
the good agreement of our results with the experimental data shows that our model can be applied to analyze the anisotropy of angular distribution of the reaction fragments and the contribution of quasifission fragments in the measured data which depend on the charge asymmetry of reaction in the entrance channel , peculiarities of the shape and shell structure of colliding nuclei .
the authors are grateful to profs .
jolos , and a. sobiczewski for helpful discussions . this work was performed partially under the financial support of the dfg , rfbr ( grant no .
04 - 02 - 17376 ) and intas ( grant 03 - 01 - 6417 ) .
the authors ( a.k.n . and r.k.u . ) thank daad and polish - jinr cooperation program for support while staying at the giessen university and soltan institute for nuclear studies , respectively . a.k.n . and
are also grateful to the fondazione bonino - pulejo ( fbp ) of messina for the support received in the collaboration with the messina group .
authors ( a.i.m , a.k.n . and r.k.u ) are grateful to the center on science and technologies at the ministry cabinet ( grant no .
f-2.1.8 ) and the support fond of fundamental research of the academy of science of uzbekistan ( no .
64 - 04 ) for partial support .
nasirov is grateful to the universit libre de bruxelles for the support received in the collaboration .
the effective moment of inertia @xmath167 of dns which is formed in collisions with the different orientation angles of their symmetry axes relative to the beam direction is calculated by the formula ( [ jeff ] ) .
but the moments of inertia @xmath168 and @xmath169 should be the smallest and largest components , respectively .
the axis for which the value of @xmath32 is minimal is found from the condition @xmath170 ( @xmath171 is the angle between the axis @xmath32 and beam direction ) ( fig.[jdns ] ) . the @xmath33 axis is directed as a normal to the reaction plane .
the moments of inertia of nuclei are calculated as for a rigid - body system . for a quadrupole deformed nucleus the moment of inertia
is calculated by the expression : the moments of inertia @xmath168 and @xmath169 are found by using the steiner s theorem for the rigid - body moments of inertia of the dns constituents : @xmath176 @xmath177 where @xmath178 ( @xmath179 ) is the distance between the center of mass of the fragment @xmath180 ( @xmath175 ) and the axis corresponding to the largest ( smallest ) moment of inertia of the dinuclear system .
nasirov , g. giardina , a.i .
muminov , w. scheid , u.t .
yakhshiev , _ proc . of the symposium on nuclear clusters : from light exotic to superheavy nuclei , rauischholzhausen , germany , 5 - 9 august , 2002 _ , ed .
r. jolos and w. scheid , ep systema , debrecen , hungary , 2003 , p.415 - 426 ; acta physica hungarica a*19 * ( 2004 ) pp.109 - 120 . v.v .
volkov , _ contributed papers of nucleus - nucleus collison ii , visby , 1985 _ , edited by b. jakobson and k. aleclett ( north - holland , amsterdam , 1985 ) , vol . 1
, p.54 ; izv .
nauk sssr ser .
* 50 * , 1879 ( 1986 ) ; _ proceedings of the international school - seminar on heavy ion physics , dubna , 1986 _ , d7 - 87 - 68 ( dubna , 1987 ) , p.528 ; _ proceedings of the 6th international conference on nuclear reaction mechanisms , varenna , 1991 _ , edited by e. gadioli , ( ricerca scientifica ed educazione permanente , supplemento n.84 , 1991 ) , p.39 .
p. moller , j. r. nix , w. d. myers and w. j. swiatecki , at .
. data tables * 59 * , 185 ( 1995 ) .
adamian , n.v .
antonenko , and w. scheid , phys .
c * 69 * , 044601 ( 2004 ) .
p. j. siemens and a. s. jensen , _ elements of nuclei , lecture notes and supplements in physics _ , ( addison - wesley , redwood city , california , 1987 ) . | the anisotropy in the angular distribution of the fusion - fission and quasifission fragments for the @xmath0o+@xmath1u , @xmath2f+@xmath3pb and @xmath4s+@xmath3pb reactions is studied by analyzing the angular momentum distributions of the dinuclear system and compound nucleus which are formed after capture and complete fusion , respectively .
the orientation angles of axial symmetry axes of colliding nuclei to the beam direction are taken into account for the calculation of the variance of the projection of the total spin onto the fission axis .
it is shown that the deviation of the experimental angular anisotropy from the statistical model picture is connected with the contribution of the quasifission fragments which is dominant in the @xmath4s+@xmath3pb reaction .
enhancement of anisotropy at low energies in the @xmath0o+@xmath1u reaction is connected with quasifission of the dinuclear system having low temperature and effective moment of inertia .
pacs : 25.70.jj-fusion and fusion - fission reactions ; 25.70.lm-strongly damped collisions ; 25.85.ge-charged-particle-induced fission , capture , fusion , quasifission , angular anisotropy of fragments . |
recent developments in the physics of intrinsic localized modes ( ilm ) @xcite have shown that the ilms are not only conceptually important in theoretical and mathematical physics as ubiquitous fundamental modes in nonlinear , discrete physical systems @xcite but also possess innovative potential applications in fundamental science and technology in addition to their experimental observation @xcite .
areas of such applications cover coupled josephson junctions , photonic crystals , optical lattices in bose - einstein condensates , all - optical logic and switching devices , targeted breaking of chemical bonds , and so on @xcite .
very recently , the interplay between nonlinear dynamics and geometry has attracted particular attention in ilm problems @xcite .
historically , such an observation was first made when formulating model hamiltonians for the dynamics of bases in dna to take care of helical structure , to the best of the authors knowledge @xcite .
the latter problem gave rise to the intuitive introduction of the so - called sine - lattice ( sl ) model @xcite in which inter - site interactions along a given strand are taken to be sinusoidal rather than the conventional quadratic ones .
the sl model later turned out to yield a novel type of ilm , referred to as an intrinsic rotating mode ( irm ) , also termed a roto - breather @xcite .
an example of transition from oscillation to such rotation modes in simulations of molecular crystals can be found in @xcite .
another example of this type is given by curved or bent chains @xcite and long - range interactions on a fixed curved substrate @xcite .
geometric constraints naturally appear in crystalline bodies consisting of relatively rigid atomic clusters .
an important example of this class of materials are the polymorphs of silica ( sio@xmath0 ) , where the structural units , i.e. , sio@xmath1 tetrahedra are corner - linked by oxygen atoms , and the energy cost of deformation of the tetrahedra is much greater than the cost of their mutual rotations .
atoms in the almost rigid clusters move as if they were subjected to a geometrical constraint .
to describe the position of a finite - size molecular cluster , one has to introduce not only translational but also rotational degrees of freedom .
it has been demonstrated that the rotational degrees of freedom can be responsible for the incommensurate phase in quartz @xcite , negative poisson s ratio of cristobalite and quartz @xcite , and negative thermal expansion of beta - quartz @xcite .
similar effects can be observed in other materials with microscopic rotations , such as perovskites ( e.g. , srtio@xmath2 ) containing corner - linked tio@xmath3 octahedra , the kh@xmath0po@xmath1 ( kdp ) family of crystals with comparatively rigid po@xmath1 tetrahedra , among others . motivated by these investigations , we proposed in a previous paper a mechanical model in which a set of masses on a linear chain are re - arranged to slide on fixed rings @xcite .
such a model was shown to contain rich nonlinear dynamics , exhibiting various types of nonlinear modes ranging from kinks to irms .
the method employed there amounts to applying a specific geometrical constraint to a purely harmonic lattice , leading eventually to equations having the form of a extended version of the sl equations .
it is our purpose here to propose a method for studying the above - mentioned interplay between nonlinearity and geometry in a more general and transparent way .
the basic point of our method is to take three - dimensional ( 3d ) harmonic lattices as a starting point on which geometrical constraints are imposed .
since linear systems constitute a basis for studying physics and mathematics in general , we expect the present method will give much more insight into the problem . the work is composed of two parts : the first presents a general scheme of geometrical constraints , while the second complements it with a detailed study of the ilms in sine- plus linear- lattices resulting from the helical constraints that we impose on the original 3d harmonic lattice .
this paper is organized as follows : in the next section , we present a simple 3d harmonic lattice model and consider a general scheme of the geometrical constraint applied to it . in section iii , we introduce the helical constraints as an application of the general method to arrive at helical lattices described by sine - linear - lattice ( sll ) equations interpolating between sl and linear lattice .
section iv is devoted to the study of some properties of the sll equations .
generalization of the sll model is made in section v to include the effect of an on - site potential , giving rise to the ilms .
the last section is devoted to concluding remarks .
we consider a three - dimensional lattice governed by the lagrangian @xmath4 @xmath5 } , \label{lagrangian}\end{aligned}\ ] ] where @xmath6 and @xmath7 are the kinetic energy and the potential energy of the system , respectively .
the quantities @xmath8 are the dynamical variables associated with the @xmath9th atom of atomic mass @xmath10 . then , the euler - lagrange equations assume the form @xmath11 physically , eqs .
( [ eqmotion ] ) are equations of motion for harmonic lattice vibrations of a simplified version of a simple cubic lattice , in which @xmath12 represent the @xmath13 components of the displacement vector @xmath14 of the @xmath9th atom from its equilibrium position .
suppose now that there exists a single variable @xmath15 such that @xmath16 where @xmath17 and @xmath18 are functions of @xmath15 .
the lagrangian of the system is then written as @xmath19}\dot{s}^2 \nonumber \\ -\frac{k}{2}\sum_n{}\big([f(s_{n+1})-f(s_n)]^2 \nonumber \\ + [ g(s_{n+1})-g(s_n)]^2+[h(s_{n+1})-h(s_n)]^2\big ) .
\label{newlagrangian}\end{aligned}\ ] ] then , the euler - lagrange equations takes the form @xmath20}\ddot{s}_n \nonumber \\
+ \frac{m_n}{2}{\left(\frac{d}{ds}[f^{\prime}(s_n)^2+g^{\prime}(s_n)^2+h^{\prime}(s_n)^2]\right)}\dot{s}_n^2 \nonumber \\ = k{\left[f(s_{n+1})+f(s_{n-1})-2f(s_n)\right]}f^{\prime}(s_n ) \nonumber \\ + k{\left[g(s_{n+1})+g(s_{n-1})-2g(s_n)\right]}g^{\prime}(s_n ) \nonumber
\\ + k{\left[h(s_{n+1})+h(s_{n-1})-2h(s_n)\right]}h^{\prime}(s_n ) , \label{neweqmotion}\end{aligned}\ ] ] where @xmath21 with @xmath22 .
it can hence be seen from eqs .
( [ newlagrangian ] ) and ( [ neweqmotion ] ) that the geometric constraint of eqs .
( [ constraint ] ) , transforms the original linear lattice equations into nonlinear ones .
thus , depending on the choice of @xmath23 and @xmath24 , we can obtain various nonlinear equations from the linear lattice model through the one - parameter geometric constraint .
equations ( [ constraint ] ) can be generalized to the case in which , e.g. , the quantities @xmath25 are parametrized by two variables , @xmath26 .
we illustrate the above method by considering situations in which the constraint functions @xmath27 are of the form @xmath28 where @xmath29 are constants depending on the site index @xmath9 and @xmath30 . inserting eqs .
( [ helicalconstraint ] ) into eq .
( [ newlagrangian ] ) , we obtain the lagrangian of the system in the form @xmath31 } \nonumber \\ -\sum_n\frac{k}{2}\large[a_{n+1}^2+a_n^2 - 2a_{n+1}a_n\cos(\theta_{n+1}-\theta_n ) \nonumber \\ + ( b_{n+1}\theta_{n+1}-b_n\theta_n)^2\large ] .
\label{helicallagrangian}\end{aligned}\ ] ] then , the euler - lagrange equations are given by @xmath32 \nonumber \\
+ c_n{\left[b_{n+1}b_n\theta_{n+1}+b_nb_{n-1}\theta_{n-1}-2b_n^2\theta_n\right ] } , \label{helicaleqmotion}\end{aligned}\ ] ] where @xmath33 equations ( [ helicaleqmotion ] ) interpolate between a generalized version of the linear lattice @xmath34 , \label{linearlattice}\end{aligned}\ ] ] and that of the sine - lattice equation @xcite @xmath35 , \label{sinelattice}\end{aligned}\ ] ] where the coefficients of the right - hand side are all site - dependent .
( [ helicaleqmotion ] ) are referred to as sine - linear - lattice ( sll ) equations .
here we restrict ourselves to the simplest situations in which all the coefficients are site - independent .
then , dropping the subscript @xmath9 attached to the coefficients in eq .
( [ helicaleqmotion ] ) leads to @xmath36 , \label{sinelinearlattice}\end{aligned}\ ] ] where @xmath37 the non - integrable equation ( [ sinelinearlattice ] ) can be considered as relevant to helical - lattices in which the helicity is determined by the factor @xmath38 .
alternatively , eqs .
( [ sinelinearlattice ] ) may be considered as a modified version of the discrete sine - gordon equation @xcite in which the conventional on - site term @xmath39 is replaced by the first term in the right hand side of eq .
( [ sinelinearlattice ] ) .
equation ( [ sinelinearlattice ] ) can be rewritten in the form @xmath40 equilibrium solutions to eq .
( [ sll ] ) can be found from @xmath41 , which is particularly satisfied when @xmath42 for all @xmath9 .
stable equilibria are subjected to an additional condition , @xmath43 . to study the physical properties of eq .
( [ sinelinearlattice ] ) which represents the competition between the linearity and the helicity , let us consider the potential function @xmath44 associated with @xmath45 , @xmath46 we pay particular attention to its minimum points . for sufficiently large @xmath38 ( weak helicity ) , the potential @xmath47 has only one minimum at @xmath48 and the only stable state of the sll is the ground state , @xmath49 .
however , with decrease of @xmath38 , new minima appear in the potential @xmath44 , the @xmath50th minimum appearing at @xmath51 for example , @xmath52 , @xmath53 , @xmath54 .
the potential @xmath44 is shown in fig .
[ dfig1 ] for these values of @xmath38 and also for a larger magnitude , @xmath55 , when there is only one minimum . in the limiting case @xmath56 , the sll eq .
( [ sinelinearlattice ] ) degenerates to the sl equation with energetically indistinguishable minima situated at @xmath57 . as the number of minima on the potential @xmath44 increases , the variety of the stable structures of the sll eq .
( [ sinelinearlattice ] ) also increases . for @xmath58
, @xmath59 can obtain @xmath60 values ( positive or negative ) in addition to @xmath61 , which is allowed for any @xmath38 . for the particular class of equilibrium solutions expressed by eq .
( [ equilibriumcondition ] ) , configuration of the @xmath9th bond , @xmath62 , is not affected by the adjacent bonds , and one can easily create various compacton - like localized defects . for example , for @xmath63 , @xmath44 has minima at @xmath48 and @xmath64 . in fig .
[ dfig2 ] we show several stable configurations of sll at @xmath63 and @xmath65 . shown are ( a ) a kink - type structure , ( b ) a point defect in a ground state structure , ( c ) a zigzag periodic structure , and ( d ) a point defect in the zigzag structure . the defects in periodic structures presented in fig .
[ dfig2 ] can be regarded as discrete compactons @xcite because they are sharply localized and do not have exponential tails .
a vibration mode can be excited on the defect in zigzag structure presented in fig .
[ dfig2](d ) .
the mode is ( practically ) localized on three particles and it is presented in fig . [ dfig3 ] .
the existence of this mode can be easily understood
. the 0th particle in fig . [ dfig2](d ) oscillates with a frequency close to the upper edge of the ground state spectrum which , however , is outside of the frequency band of the zigzag structure . due to translational invariance , the sll model of eq .
( [ sinelinearlattice ] ) admits solutions with @xmath66 for arbitrary @xmath67 .
domain walls separating two domains with @xmath68 can also be formed . here
we consider the case when the difference @xmath69 is not equal to the distance between the minima of the potential @xmath44 eq .
( [ potentialfunction ] ) and which can give rise to moving kinks .
there are many possibilities to initiate such moving kink - like structures ; one is presented in fig .
[ dfig4 ] for @xmath65 , @xmath63 . here
the kinks were initiated by applying initial conditions @xmath70 , @xmath71 , with zero initial conditions for all other particles .
the height of the kinks is @xmath72 and it increases by increasing the initial velocity of the @xmath73th particle .
here we use periodic boundary conditions to demonstrate that the kinks survive collisions and hence are relatively robust objects .
let us consider situation in which an atom located at the site @xmath73 makes a large excursion around one of the minimum points @xmath74 of the potential , while the others only perform small amplitude oscillatory motion .
this means that @xmath75 as we expect a highly localized solution , we preserve the full nonlinearity of the equations for the central site only , while the other equations are linearized around the equilibrium position .
such a procedure leads to @xmath76 } + { \left[\cos(\theta_0)+\epsilon \right]}(\theta_1+\theta_{-1})\right ) } , \label{e17}\end{aligned}\ ] ]
@xmath77}\theta_{\pm 1 } \nonumber \\ + ( 1+\epsilon)\theta_{\pm 2}+\sin(\theta_0)+\epsilon \theta_0\big ] , \label{e18}\end{aligned}\ ] ] @xmath78 we now seek solutions to this set of equations in the form @xmath79 when , in eq .
( [ e20 ] ) , @xmath80 , all particles oscillate near one well of the potential @xmath44 , while for @xmath81 the @xmath73th particle oscillates in a different well . inserting eq .
( [ e20 ] ) into eqs .
( [ e17 ] ) , ( [ e18 ] ) and ( [ e19 ] ) and retaining solely terms linear with respect to @xmath82 and @xmath83 , we arrive at a set of equations which are of the same form as those appearing in one - impurity problems in harmonic lattice vibrations .
thus , setting @xmath84 where @xmath85 is time - independent and @xmath86 is a constant , we reduce the above set of equations to the form @xmath87 with @xmath88}(2v_0-v_1-v_{-1 } ) , \nonumber \\ l^{\prime}v_{\pm 1}=j{\left[1-\cos(\alpha)\right]}(v_{\pm 1}-v_0 ) , \nonumber \\ l^{\prime}v_n=0 \,\,\,\ , { \rm for } \,\,\,\
, n \neq 0 , \pm 1 .
\label{e23}\end{aligned}\ ] ] from the above equations , the ilms we are seeking are presumed to exist within the frequency band @xmath89}$ ] of the perfect harmonic lattice characterized by wavenumber @xmath90 .
such a nonlinear localized in - band modes are referred to as ( intrinsic ) nonlinear resonant modes ( nrm )
. then , eqs .
( [ e22 ] ) and ( [ e23 ] ) can be handled by introducing lattice green s functions @xmath91}-\omega^2-i\gamma } , \nonumber \\
\gamma \rightarrow 0_+ .
\label{e24}\end{aligned}\ ] ] analytical expressions for the lattice green s functions @xmath92 can be obtained by replacing the sum with respect to @xmath90 by an integral over the first brillouin zone as follows @xmath93 , \label{e25}\end{aligned}\ ] ] where @xmath94 in which @xmath95 is the bessel function of the @xmath9th order .
we note that the quantity @xmath96 represents the density of states of the band . in terms of
the @xmath97s so obtained , eqs .
( [ e22 ] ) and ( [ e23 ] ) are rewritten as @xmath98\big[g(n)(2v_0-v_1-v_{-1 } ) \nonumber \\
+ g(n-1)(v_1-v_0)+g(n+1)(v_{-1}-v_0)\big ] .
\label{e28}\end{aligned}\ ] ] we pay particular attention to an s - like mode having the symmetry property @xmath99 and use the identity relation , @xmath100 } \nonumber \\ -\omega^2g(n)=\delta_n , \label{e30}\end{aligned}\ ] ] to obtain @xmath101}(v_0-v_1 ) , \label{e31}\end{aligned}\ ] ] where @xmath102 and @xmath103 are kronecker s delta . in terms of the dimensionless frequency @xmath104 , an equation giving the eigenfrequency of the nrm
is obtained from eqs .
( [ e25 ] ) , ( [ e27 ] ) , and ( [ e31 ] ) as follows : @xmath105 the dimensionless squared eigenfrequency @xmath106 is therefore obtained as @xmath107 the localization of the nrm is given by the equation @xmath108 as mentioned before , here the localized mode around the local minimum point of the potential function is of resonant type , the eigenfrequency which appears close to bottom of the frequency band @xmath109}}$ ] . apart from the factor @xmath74 , which is the position of the local minimum , the localization properties of the nrm are essentially different from those of the ilm in that it exhibits oscillatory slow decay , contrary to what is the case for the ilm .
this point will be examined in detail below .
we have attempted to excite nrms in sll of eqs .
( [ sinelinearlattice ] ) .
such a numerical calculation is presented in fig .
[ dfig5 ] . here
we set @xmath65 , @xmath63 and for the initial conditions , @xmath110 , @xmath111 , with zero initial conditions for all other particles . for this choice of @xmath38
, the potential @xmath44 has minima at @xmath112 and @xmath113 .
one can see that the @xmath73th particle oscillates near the minimum of @xmath44 situated at @xmath114 , while the remaining particles oscillate near the minimum at @xmath48 . due to the rather strong interaction with the linear excitations of the lattice ,
the energy of the nrm is rapidly imparted to the lattice ( after a few oscillations ) .
the lifetime of the excitation does not increase much , either by changing the initial deviation of the @xmath73th particle from the potential minimum or by changing @xmath115 .
it is of interest to see what happens if we include on - site potentials in our original lattice .
this is done by generalizing eq .
( [ lagrangian ] ) in the form @xmath116 , \label{lwithonsite}\end{aligned}\ ] ] where @xmath117 and @xmath118 are constants . applying the helical constraint , eq .
( [ helicalconstraint ] ) leads to @xmath119 } \nonumber \\ -\sum_n\frac{k}{2}\big[a_{n+1}^2+a_n^2 - 2a_{n+1}a_n\cos(\theta_{n+1}-\theta_n ) \nonumber \\ + ( b_{n+1}\theta_{n+1}-b_n\theta_n)^2 \nonumber \\ + \frac{a_n^2(\kappa_2-\kappa_1)}{4}\cos{\left[1-\cos(2\theta_n)\right]}\big ] .
\label{lwithonsitewithconstraint}\end{aligned}\ ] ] it is seen that such a procedure adds a sine - gordon type on - site potential @xmath120 to the lagrangian .
the equations of motion for @xmath121 then assume the form @xmath122
+ l_n\big[b_{n+1}b_n\theta_{n+1}+b_nb_{n-1}\theta_{n-1 } \nonumber \\ -(b_{n+1}b_n+b_nb_{n-1})\theta_n\big ] , \label{emwithonsite}\end{aligned}\ ] ] where @xmath123 .
when all the coefficients are site - independent , the above equations reduce to @xmath124 . \label{sllwithonsite}\end{aligned}\ ] ] the presence of the on - site potential here can be speculated to induce the existence of localized modes with longer lifetime than that of the case of eq .
( [ sinelinearlattice ] ) . in the following ,
we demonstrate the existence of ilms in the model of eq .
( [ sllwithonsite ] ) .
the linear spectrum of the sll with the on - site potential , eq .
( [ sllwithonsite ] ) , is given by @xmath125 , \label{sllwithonsitespectrum}\end{aligned}\ ] ] with @xmath126}. \label{minmax}\end{aligned}\ ] ] when @xmath127 , i.e. , @xmath128 , one can look for an ilm oscillating with frequency @xmath129 so that all higher harmonics @xmath130 , with an integer @xmath131 , lie above @xmath132 .
in this situation the ilm would not interact with the lattice preserving its identity . as an example
, we take @xmath65 , @xmath63 , and @xmath133 , with @xmath134 , and @xmath135 so that the necessary condition of the existence of ilm , @xmath136 , is fulfilled .
a large - amplitude localized mode can be excited by choosing the initial deviation and/or initial velocity of a particle . as an example
, we initialize the @xmath73th particle with @xmath137 , @xmath138 with zero initial conditions for other particles and , after some stabilization period , a steady oscillatory motion is observed ( see fig .
[ dfig6 ] ) .
one can see that the ilm is highly localized .
the mode has a frequency @xmath139 , which is considerably lower than the bottom edge of the linear spectrum , @xmath140 , with all higher harmonics lying above the upper edge , @xmath132 .
the ilm can be expressed analytically assuming that @xmath141 with @xmath142 and @xmath143 .
parameters of the approximate solution eq .
( [ ilm1 ] ) can be expressed in terms of the ilm amplitude , @xmath144 : @xmath145 , \nonumber \\ a_3=\frac{(1 + 4\lambda)a_1 ^ 3}{12\left[9\frac{\omega^2}{j}-2(1+\epsilon+\lambda)\right ] } , \nonumber \\ b_1=\frac{(1+\epsilon)a_1-\frac{1}{8}a_1 ^ 3}{2(1+\epsilon+\lambda)-\frac{\omega^2}{j}}. \label{ilm1parameters}\end{aligned}\ ] ] let us analyze the obtained solution .
first of all we note that the ilm frequency , @xmath86 , can not be greater than @xmath140 , i.e. , we can only look for a mode with frequency below @xmath140 [ see eq .
( [ minmax ] ) ] , which is possible only for ilms with sufficiently large amplitude , @xmath146 for example , for @xmath63 , and @xmath133 , one has @xmath147 .
we have confirmed numerically that , for this choice of parameters , our solution can be used for @xmath148 . for larger amplitudes
the ilm is highly localized ( see fig .
[ dfig7 ] ) but the solution eq .
( [ ilm1 ] ) , ( [ ilm1parameters ] ) becomes inaccurate because it takes into account only cubic anharmonicity .
the ilm in fig .
[ dfig7 ] has amplitude @xmath149 and frequency @xmath150 , which is significantly lower the bottom edge of the linear spectrum , @xmath151 .
the mode radiates energy extremely slowly due to the interaction of higher harmonics with the linear spectrum .
note the difference in the ordinate scale for the middle panel . for small ilm amplitudes , approaching the allowed limit , eq .
( [ amplitudecondition ] ) , the solution profile becomes wider and our assumption that it is localized on three particles becomes invalid .
estimation for the limiting amplitude , eq .
( [ amplitudecondition ] ) , is also valid only for a highly localized ilm . in fact
, it is possible to excite an ilm with a very small amplitude but the width of such ilm increases considerably and its frequency approaches @xmath140 from below .
the smooth transformation of the ilm from a highly localized ( as the amplitude decreases ) into a very broad one is not surprising because for smooth solutions , @xmath152 , one has @xmath153 and the sll with the on - site potential , eq .
( [ sllwithonsite ] ) , can be approximated by the well - known continuum sine - gordon equation .
the ilm is , then , nothing but the analog of breather solution to the sine - gordon equation .
using the lorentz invariance of the sine - gordon equation one can create a moving small - amplitude breather .
we also note that , when @xmath154 and @xmath155 , the first two terms in the right hand side of eq .
( [ sllwithonsite ] ) can be neglected .
in this situation , when @xmath38 and @xmath156 are of the same order of magnitude , our solution , eqs .
( [ ilm1 ] ) and ( [ ilm1parameters ] ) , describes a highly localized ilm in the frenkel - kontorova model @xcite . the sll lattice with on - site potential supports localized modes of another type , when one particle is trapped in a potential well different from the well occupied by the other particles . to demonstrate this mode
, we take @xmath65 , @xmath157 , and @xmath158 , with @xmath159 , and @xmath160 so that the necessary condition , @xmath136 , is fulfilled . setting @xmath161 , @xmath138 with zero initial conditions for other particles , after some time a steady oscillatory motion presented in fig .
[ dfig8 ] was observed .
the @xmath73th particle was found to oscillate with @xmath162 which is below the linear vibration band but all higher harmonics are above the band .
one can see that the ilm is not highly localized .
the @xmath73th particle oscillates with the amplitude @xmath163 near the coordinate shifted by roughly @xmath164 with respect to the neutral position of all other particles [ shown in fig .
[ dfig8](b ) ] .
its nearest neighbors oscillate with nearly the same amplitude , @xmath165 , and the amplitudes decrease rather slowly with deviation from the @xmath73th particle [ see fig .
[ dfig8](a ) ] .
in this paper , we have outlined a general scheme to apply geometrical constraints to a simplified version of the 3d harmonic lattice with or without an on - site potential to obtain nonlinear dynamical lattice equations .
depending on the type of the constraints , we arrive at various nonlinear equations , being interested in the coherent structures that arise in them .
we then studied , more specifically , a helical constraint by which the original 3d harmonic lattice was shown to reduce to 1d sine - plus - linear lattice equations or helical lattice equations .
this formulation revealed several interesting features : 1 .
it provides us with a systematic derivation of the sine - lattice equations which were derived heuristically in earlier works .
2 . physically , applying the helical constraint amounts to transforming the original harmonic lattices ( possessing the hard potential ) to a nonlinear dynamical lattice that supports resonant and multi - kink modes . seeking other kinds of nonlinear modes will be an interesting topic for future study .
3 . the method developed here may be relevant to the study of nonlinear excitations in biomolecules with helical structure such as dna and proteins .
there are several areas of condensed matter physics where the existence of a complex energy landscape with a number of local minima in the potential function is presumed to play a crucial role in determining their properties .
such examples can , for instance , be found in the study of glasses and proteins .
application of the techniques employed here to these problems would be another topic warranting further investigation .
4 . finally , another relevant generalization would be to consider geometric constraints different than the helical one , and derive and study the ensuing ( reduced ) dynamical models .
s. takeno , prog .
. phys . * 75 * , 1 ( 1986 ) ; s. takeno , k. kisoda , and a. sievers , prog .
suppl . * 94 * , 242 ( 1988 ) .
a. j. sievers and s. takeno , phys .
lett . * 61 * , 970 ( 1988 ) . for recent reviews ,
see for example , s. flach and c. r. willis , phys .
, * 295 * , 181 ( 1998 ) ; d. hennig and g. tsironis , physics reports * 307 * , 333 ( 1999 ) ; p.g .
kevrekidis , k.o .
rasmussen , and a.r .
bishop , int .
b * 15 * ( 2001 ) 2833 - 2900 ; j.ch .
eilbeck and m. johansson , in _ localization and energy transfer in nonlinear systems _
, l. vazquez , r.s .
mackay , and m.p .
zorzano ( eds . ) , ( world scientific , singapore , 2003 ) , p.44 .
u. t. schwarz , l.q .
english and a.j .
sievers , phys .
lett . * 83 * , 223 ( 1999 ) .
n. k. voulgarakis , g. kalosakas , a.r .
bishop and g.p .
tsironis , phys .
b * 64 * , 020301(r ) ( 2001 ) and references cited therein .
e. tras , j.j .
mazo , and t.p .
orlando , phys .
lett . * 84 * , 741 ( 2000 ) ; p. binder , d. abraimov , a.v .
ustinov , s. flach , and y. zolotaryuk , phys .
* 84 * , 745 ( 2000 ) . for a very recent review on the latest developments ,
see , e.g. , d. k. campbell , s. flach and y. kivshar , phys . today
* 57 * , 43 ( 2004 ) .
see , e.g. , p.g .
kevrekidis , b.a .
malomed and a.r .
bishop , phys .
e * 66 * , 046621 ( 2002 ) and references cited therein .
s. yomosa , phys .
a * 27 * , 2120 ( 1983 ) ; a * 30 * , 474 ( 1984 ) .
s. homma and s. takeno , prog .
phys . * 72 * , 679 ( 1984 ) .
s. takeno and s. homma j. phjys .
jpn . * 65 * , 2547 ( 1986 ) .
s. takeno and m. peyrard , physica d * 92 * , 142 ( 1996 ) ; phys .
e * 55 * , 1922 ( 1997 ) .
k. w. wojciechowski , k. v. tretiakov , m. kowalik , phys .
e * 67 * , 036121 ( 2003 ) .
j. cuevas j. f. r. archilla , yu .
b. gaididei and f. r. romero , physica d * 163 * , 106 ( 2002 ) . y. b. gaididei , s. f. mingaleev and p. l. christiansen , phys rev .
e * 62 * , r53 ( 2000 ) .
p. g. kevrekidis , s. v. dmitriev , s. takeno , a. r. bishop and e. c. aifantis , phys .
e ( in press ) .
p. g. kevrekidis and v. v. konotop , phys .
e * 65 * , 066614 ( 2002 ) ; p.
g. kevrekidis , v. v. konotop and s. takeno , phys .
a * 299 * , 166 ( 2002 ) .
s. a. wells , m. t. dove , m. g. tucker , k. trachenko , j. phys . : condens .
matter * 14 * , 4645 ( 2002 ) .
m. vallade , b. berge , g. dolino , j. physique i * 2 * , 1481 ( 1992 ) .
s. v. dmitriev , d. a. semagin , t. shigenari , k. abe , m. nagamine and t. a. aslanyan , phys .
b * 68 * , 052101 ( 2003 ) ; d. a. semagin , s. v. dmitriev , k. abe and t. shigenari , russ . j. phys
. ch . * 77 * , s30 ( 2003 ) .
s. v. dmitriev , a. a. vasiliev and n. yoshikawa , in _ recent research developments in physics
_ , s. g. pandalai ( ed . ) , ( transworld research network , kerala , india , 2003 ) , vol .
4 , part i , pp .
267 - 286 .
h. kimizuka , h. kaburaki , y. kogure phys .
lett . * 84 * , 5548 ( 2000 ) .
m. b. smirnov , a. p. mirgorodsky , phys .
78 * , 2413 ( 1997 ) .
n. r. keskar , j. r. chelikowsky , phys .
b * 48 * , 16227 ( 1993 ) .
a. alderson , k. e. evans , phys .
. lett . * 89 * , 225503 ( 2002 ) .
m. b. smirnov , phys .
b * 59 * , 4036 ( 1999 )
. o. m. braun , y. s. kivshar , _ the frenkel - kontorova model : concepts , methods , and applications _ , ( springer - verlag , berlin , 2004 ) . of eq .
( [ potentialfunction ] ) is shown as a function of @xmath166 for different magnitudes of @xmath38 .
curves 1 to 4 correspond to @xmath55 , @xmath52 , @xmath167 , and @xmath54 .
the magnitudes @xmath168 , at which @xmath50th minimum of the potential function appears , are given by eq .
( [ newminima ] ) . ] , @xmath71 , and zero initial conditions for all other particles , presented by the set of plots @xmath171 vs. @xmath9 for different times .
the height of the kinks is @xmath72 and it increases with increase in initial velocity of @xmath73th particle .
periodic boundary conditions are used here to demonstrate the kinks robustness against collisions .
@xmath65 , @xmath63 . ] ) by means of initial conditions @xmath110 , @xmath111 , with zero initial conditions for all other particles .
@xmath65 , @xmath63 .
the @xmath73th particle oscillates near the minimum of @xmath44 situated at @xmath114 , while other particles oscillate near the minimum at @xmath48 . due to the rather strong interaction with the lattice ,
the energy of excitation is `` dissipated '' to the lattice after only a few oscillations . ] and frequency @xmath139 , which is considerably lower than the bottom edge of the linear spectrum , @xmath172 with all higher harmonics lying above the upper edge , @xmath173 .
@xmath65 , @xmath63 , and @xmath133 . ]
but for larger amplitude , @xmath149 .
the ilm has frequency @xmath150 which is significantly lower than the bottom edge of the linear spectrum , @xmath151 .
the mode radiates energy extremely slowly due to the interaction of higher harmonics with the linear spectrum .
note the difference in the ordinate scale for the middle panel .
@xmath65 , @xmath63 , and @xmath133 .
] vs. @xmath9 for different times showing the dynamics of particles ( the @xmath73th particle is not shown here because its coordinate differs from coordinates of other particles significantly , by roughly @xmath164 ) .
( b ) dynamics of @xmath73th particle .
the ilm is rather broad .
the @xmath73th particle oscillates with the amplitude @xmath163 while its nearest neighbors have nearly the same amplitude , @xmath165 , and the amplitudes decrease rather slowly with deviation from @xmath73th particle .
@xmath65 , @xmath157 , and @xmath158 . | geometrical constraints imposed on higher dimensional harmonic lattices generally lead to nonlinear dynamical lattice models .
helical lattices obtained by such a procedure are shown to be described by sine- plus linear - lattice equations .
the interplay between sinusoidal and quadratic potential terms in such models is shown to yield localized nonlinear modes identified as intrinsic resonant modes . |
in recent experiments of relativistic heavy ion collisions realized at the brookhaven rhic and at the cern lhc , the production of a deconfined and strongly coupled quark - gluon plasma ( qgp ) has been detected .
the qgp expands anisotropically and cools , reaching the hydrodinamic regime after a time of order 1 fm / c .
the relaxation towards equilibrium can be studied in the holographic approach , a framework inspired by the gauge / gravity correspondence .
this conjecture relates a strongly coupled super - yang mills gauge theory defined in a @xmath0-dimensional minkowski space ( boundary ) with a dual classical gravity theory living in a @xmath1-dimensional anti - de sitter space ( bulk ) times a compact manifold @xcite .
the evolution of the qgp initial state , anisotropic and far from equilibrium , can be examined by introducing an impulsive perturbation ( quench ) to the metric on the boundary , and then solving the einstein equations in the bulk @xcite . in our analysis ,
the external sources that distort the boundary metric are chosen to mimic processes in which a small number of collisions occur .
as the quench becomes static , the system starts to relax towards the hydrodinamic regime .
the boundary coordinates are denoted as @xmath2 , with @xmath3 the collision direction .
we assume that the geometry is invariant under boost transformations along this axis and under translations and rotations in the transverse plane @xmath4 .
the minkowski line element @xmath5 , with @xmath6 the proper time and @xmath7 the spacetime rapidity @xmath8 and @xmath9 , is perturbed through the quench @xmath10 as follows : @xmath11 the stress - energy tensor on the boundary yields a local probe of the relaxation process and a measure of the residual anisotropy in the plasma .
its components , the energy density and the pressures longitudinal and transverse with respect to the collisional axis , have been computed for different quench profiles and compared to the viscous hydrodinamics behaviour .
regardless of the considered distortion , hydrodinamization of the stress energy - tensor is a two - step process : the energy density acquires the viscous form as soon as the quench is switched off , while pressure isotropy is restored with a time delay of order of @xmath12 fm / c , for an effective temperature at the end of the quench of about @xmath13 mev @xcite .
such description can be compared to the one deduced from the time evolution of non - local observables @xcite , as presented in the next section for a particular distortion profile @xcite .
the relaxation process can be monitored through non - local probes , such as the two - point correlation function of boundary theory operators , and expectation values of wilson loops defined on the boundary @xcite .
the system under investigation is driven out - of - equilibrium by boundary sourcing using a sequence of two nearly - overlapping pulses , as shown in fig .
[ fig ] ( top left ) . .
upper right : geodesics length @xmath14 for different vales of the size @xmath15 .
lower left : difference @xmath16 between the geodesics lengths of size @xmath15 in the quenched and in the hydrodynamic geometries .
lower right : non - local probes @xmath16 , @xmath17 and @xmath18 at the isotropization time @xmath19.,scaledwidth=90.0% ] we consider the correlation function of a boundary scalar operator @xmath20 with large conformal dimension @xmath21 , between two equal - time points having a spatial separation @xmath15 in the transverse plane : @xmath22 and @xmath23 . in the holographic picture , this observable can be approximated as @xmath24 , where @xmath25 is the length of the geodesic that connects the boundary points and extends in the bulk at fixed @xmath26 .
figure [ fig ] ( top right ) shows the evolution of @xmath14 for several values of the distance @xmath15 between the points in the correlation function : the curves follow the distortion profile , with a time delay that increases with the size of the probe . in order to evaluate the degree of thermalization of the system , it is interesting to examine the difference @xmath16 between the length @xmath14 in the geometry under investigation and the same observable computed in a bulk metric reproducing the viscous hydrodynamic time dependence of the stress - energy tensor @xcite . the evolution of @xmath16 after the end of the quench is shown in fig . [ fig ] ( bottom left ) for different values of @xmath15 . each curve is plotted starting from the value of the physical time @xmath27 , at which it is no longer affected by the quench .
an analogous approximation can be used to compute the expectation value of the wilson loop along a closed contour @xmath28 of linear size @xmath15 and fixed time @xmath29 on the boundary : @xmath30 , with @xmath31 the area of the extremal surface bounded by @xmath28 and plunging in the bulk at fixed @xmath7 .
two shapes for @xmath28 have been considered : a rectangular strip of finite width @xmath15 along the @xmath32 axis and infinite length along @xmath33 , and a circular path of diameter @xmath15 in the transverse plane @xmath4 .
the areas of the extremal surfaces bounded by these contours , @xmath34 and @xmath35 respectively , and the differences @xmath17 and @xmath36 between the observables in the quenched and in the hydrodynamic geometries , have been worked out in @xcite .
the behaviour of such quantities resembles the one in fig .
[ fig ] ( top right and bottom left ) for the geodesics case .
the time at which each non - local probe reaches the viscous hydrodynamic regime depends on its size : for larger boundary separation between the geodesics extremes or among the contour points , the quantities @xmath14 , @xmath34 and @xmath35 take longer to equilibrate .
the thermalization time can be identified with the half - life @xmath37 , defined as the value of @xmath29 at which @xmath16 , @xmath17 and @xmath36 are reduced by a half with respect to their values at time @xmath27 .
the analysis reveals that @xmath37 exceeds the thermalization time obtained from the stress - energy tensor for boundary separations @xmath38 , and increases linearly with @xmath15 for larger probes .
a hierarchy among the relaxation times of the energy density , pressures and large probes emerges , indicating that thermalization is faster at uv scales .
although the three non - local observables have the same qualitative behaviour with @xmath29 and @xmath15 , their half - lives and thermalization length scales are quantitatively different . for a comparison ,
we show in fig .
[ fig ] ( bottom right ) the @xmath15-dependence of the quantities @xmath16 , @xmath17 and @xmath18 at the physical time @xmath19 at which pressure isotropy is restored in the plasma @xcite . it is possible to identify for each function a transition to a linear regime .
the inflection point of the first derivative can be used to characterize the onset of the linear behaviour , and defines the length scale above which the probe is not thermalized at the chosen time @xcite .
the rectangular wilson loop thermalizes slower than the other non - local probes . | the thermalization of a strongly coupled plasma is examined in the holographic framework through non - local observables : the equal - time two - point correlation function of a large dimension boundary operator , and wilson loops of different shapes .
the evolution of the probes from an initial far - from - equilibrium state to a hydrodynamic regime is found to depend on their size .
a hierarchy among the thermalization times of the energy density , the pressures and the large size probes , is identified : the relaxation process is faster at short distances . |
the large hadron collider opens an opportunity to explore kinematic regions where particles produced in high - energy collisions posses large transverse momenta and rapidities .
the production of electroweak bosons and jets is a vital part of tests of standard model as well as searches of physics beyond the standard model .
furthermore it has been recognized in @xcite that by studies of associated production of electroweak bosons and jets may provide a new insight into the transverse partonic structure of hadrons at small @xmath2 , where @xmath2 is the momentum fraction of the hadron taken by a parton participating in the hard collision .
furthermore , such a final state , being a combination of colorful and colorless particles , gives the opportunity for particularly interesting investigations complementary to results obtained in studies of pure jet final states @xcite and drell - yan pairs @xcite .
in particular the final state rescatterings due to soft color exchanges should have less impact on the properties of produced final state as compared to pure jet final states .
this work is motivated by a recent lhcb measurement @xcite at @xmath4 of the process @xmath5 in the forward direction within the pseudorapidity range @xmath6 .
the final - state muon and anti - muon were required to have transverse momenta @xmath7 while the leading jet was considered with two different cuts : @xmath8 and @xmath9 .
the jets were reconstructed using the anti-@xmath10 algorithm with radius @xmath11 and they were required to be separated from muon tracks on the @xmath12 plane ( azimuthal angle - pesudorapidity ) by a distance @xmath13 . the muon - pair was required to have an invariant mass within the range @xmath14 .
the rapidity constraint assures that in the partonic picture of the process , one of the initial state partons carries a rather small fraction @xmath15 of the corresponding hadron momentum @xmath16 , while the other must have a fraction @xmath17 ( c.f .
[ fig : diagram ] ) . in order to describe the process perturbatively ,
one definitely needs to go beyond the pure collinear factorization and support the calculation by a resummations . in the modern advanced approaches
this is achieved by parton showers and hadronization as implemented for example in pythia @xcite . in the present work we consider another approach ,
namely a resummation of logarithms of @xmath18 and @xmath19 , where @xmath20 is a hard scale and @xmath21 is a certain additional scale given by the imbalance of the final states on the transverse plane .
this approach captures certain aspects of the process more accurately already at lowest order of the strong coupling constant in the hard process . in the present paper , we will therefore attempt to study the process within so called high energy factorization , or more precisely , using so - called hybrid high energy factorization motivated by the works @xcite . within the asymmetric kinematic situation @xmath17 described above
, the cross section for the process under consideration can be expressed by the following formula @xmath22 where @xmath23 is the unintegrated gluon distribution for hadron @xmath24 , @xmath25 is a collinear pdf and @xmath26 is the hard cross section obtained from a gauge invariant tree - level off - shell amplitude for the process @xmath27 , and where @xmath28 refers to quarks as well as anti - quarks ( fig .
[ fig : diagram ] ) .
let us note , that the original high energy factorization prescription was designed to study inclusive small @xmath2 processes and the corresponding unintegrated gluon distribution was assumed to undergo the balitski - fadin - kuraev - lipatov ( bfkl ) evolution equation ( see e.g. @xcite ) .
for more exclusive processes it is however necessary to include the subleading bfkl effects , such as kinematic constraint ensuring energy conservation , large-@xmath2 correction , running strong coupling constant , and notably the hard scale dependence .
the last , denoted in eq .
( [ eq : hen_fact_2 ] ) as @xmath20 , turns out to be essential to describe the data under consideration . in the present work we shall not discuss the validity of the model ( [ eq : hen_fact_2 ] ) on the theoretical level .
instead , we shall test it phenomenologically against the existing data . for a more detailed review of various approaches to the small @xmath2 factorization and related issues
see e.g. @xcite . for
a derivation of hybrid high energy factorization from the dilute limit of the color glass condensate approach see @xcite .
using the formalism described in the preceding section we have computed the cross sections for @xmath29 production .
we have used two programs to calculate off - shell @xmath30 amplitude ( with @xmath3 and @xmath31 exchange ) and to cross - check the results .
the first program is a very handy library ( avhlib ) @xcite in fortran , in which the approach of @xcite is implemented .
it computes amplitudes entirely numerically and includes a full monte carlo program .
the second is the electroweak extension of the program ogime @xcite .
it calculates amplitudes analytically in a form that can be interfaced with a monte carlo program .
more specifically , the analytic expressions were implemented in the c++ code lxjet @xcite .
note , however , that since there are no final state gluons , the ordinary feynman diagram depicted in fig .
[ fig : diagram ] ( with appropriate high energy projector for the off - shell gluon as described in @xcite ) is enough to obtain the gauge invariant amplitude for this process , and all the complications discussed in @xcite do not have any impact here .
+ jet production .
the upper blob corresponds to a collinear pdf , whereas the lower one corresponds to the unintegrated gluon distribution .
the gluon entering the hard scattering is off - shell with virtuality @xmath32 .
[ fig : diagram],scaledwidth=60.0% ] for the numerical computation we use the kinematic cuts as described in the previous section , i.e. the ones used in @xcite .
we use standard values for the electroweak parameters : electroweak coupling @xmath33 , @xmath0 boson mass @xmath34 and width @xmath35 .
we work with the 5 flavour scheme and use the cteq10nlo set @xcite for the collinear pdfs . for the unintegrated gluon distributions we consider the following models ( in brackets with bold font we give our abbreviations for the models ) : * bfkl evolution with kinematic constraint , running strong coupling constant , dglap effects including the contribution from the quark sea @xcite and supplemented with the nonlinear term @xcite ; the initial condition has been fitted to hera data @xcite ( ` ks ` ) * hard scale dependent ` ks ` ; the hard scale dependence is achieved by implementing the sudakov form factor on the generated events in such a way that the total cross section remains unchanged @xcite ( ` ks+sudakov ` ) * similar to above , but the sudakov form factor is implemented to the ` ks`density in such a way that the integrated gluon density remains unchanged @xcite ( ` kutak - nonlinear - prd15 ` ) * the simplified kimber - martin - ryskin model @xcite applied to mrstw08 pdfs ; this model gives unintegrated gluon distribution that is hard scale dependent ( ` kmr ` ) * bfkl evolution with kinematic constraint , running strong coupling constant and the dglap contribution coming only from gluons and fitted to the lhc jet data @xcite ( ` lhc - fit ` ) * hard scale dependent lhc fit ( ` lhc - fit+sudakov ` ) the hard scale was chosen to be the average of the three large scales appearing in the problem : the mass of the @xmath3 boson , its @xmath36 and the @xmath36 of the jet .
we present the results for numerical simulations in figs .
[ fig : dphi10]-[fig : ptj ] and compare them with lhcb data .
[ fig : dphi10],[fig : dphi20 ] present normalized azimuthal decorrelations for @xmath37 and @xmath38 respectively ( the azimuthal decorrelation is defined as the differential cross section in the difference of azimuthal angles of the reconstructed @xmath3 boson and the leading jet ) .
we present also results for normalized differential cross sections in the transverse momentum @xmath39 of @xmath3 boson ( figs .
[ fig : ptz10],[fig : ptz20 ] ) , in rapidity @xmath40 of @xmath3 ( figs . [
fig : yz10],[fig : yz20 ] ) , in the transverse momentum @xmath41 of the jet ( fig .
[ fig : ptj ] ) , and finally in the rapidity separation between the @xmath3-boson and the jet @xmath42 ( figs . [ fig : dy10],[fig : dy20 ] ) . the shaded boxes in figs .
[ fig : dphi10]-[fig : ptj ] represent the theoretical uncertainty obtained by varying the hard scale by a factor of two .
the differential cross sections are normalized to the total cross section , as in @xcite .
we list the total cross sections for different models in table [ table1 ] .
.[table1]total cross sections obtained from different models for the unintegrated gluon density .
the model uncertainties are defined through variations of the hard scale . for the data the total uncertainty
is estimated as an average square error from statistical , systematic and luminosity errors as given in @xcite .
we do not include models ` ks+sudakov ` and ` lhc - fit+sudakov ` as the sudakov - type resummation applied there does not change the total cross section . [ cols="^,^,^",options="header " , ] let us first discuss the normalized differential cross sections .
we can conclude that the azimuthal decorrelations are described reasonably well for both jet @xmath36 cuts for most of the models .
it was however essential for this observable to include the hard scale dependence in unintegrated gluon distributions .
as this observable is the most sensitive one for the small @xmath2 effects this underlines the importance of the resummation of the logarithms @xmath43 , where @xmath20 is the hard scale provided by the hard process and @xmath21 is the transverse momentum of the gluon in unintegrated gluon distributions , on the top of the small @xmath2 logarithms .
the effect of this sudakov - type resummation was also important for the transverse momentum spectrum of @xmath3 boson .
for the other observables it has had much less impact .
since our calculations are not interfaced with any sort of final state parton shower we expected rather rough description of transverse momenta spectra .
our actual study shows however that the situation is relatively good for the spectrum of @xmath3 boson ( except very low transverse momentum ) , figs .
[ fig : ptz10]-[fig : ptz20 ] , and indeed fails for the jet spectrum , fig .
[ fig : ptj ] . this can be attributed to the fact that all unintegrated gluon distributions we have used contain contribution from pieces of splitting functions subleading at low @xmath2 ( see @xcite for an analysis of the impact of this correction on jet @xmath36 spectra ) .
this correction seems to be enough for the colorless final state while the color rescattering for the final state jet is evidently missing . also , the next - to - leading correction in the hard process is necessary to improve the description of the transverse momentum spectra .
the lhc - jet - motivated unintegrated gluon densities , which were actually fitted to the lhc jet transverse momentum spectra behave similar to the other gluon densities .
this gives one more clue that the improvement in the hard process in terms of higher order corrections or / and a resummation is necessary .
for @xmath38 the ` kutak - nonlinear - prd15 ` and the ` kmr ` models overestimate the normalized transverse momentum spectrum of the @xmath3 boson .
these models implement a hard scale dependence but the spectra they give are almost the same as from the ` ks`model , which is hard - scale - independent ( the difference is however in the uncertainty which is much bigger for the hard - scale - dependent models ) .
this means that the sudakov - type resummation in the former models has no effect for this observable with such large @xmath36 cut . on the contrary , the sudakov - type resummation with unitarity constraint ( in the sense of preserving the total cross section ) applied to the ` ks ` or ` lhc - fit ` improves the results .
interestingly , the unintegrated gluon distribution which comes from the fit to the lhc jet data does not perform better that the distributions obtained from the inclusive dis data .
this suggests that the effects of factorization breaking due to the lack of universality is rather weak at the phenomenological level . as already mentioned above , this suggests also that one needs higher order corrections in the hard process .
indeed , as observed in @xcite the @xmath36 spectra can not be described by improving the evolution of unintegrated gluon distribution itself .
one should mention , that the evolution equations that were fitted to the lhc data did not include a contribution from the sea quarks on the dense hadron side , whereas this contribution is present in the other approaches .
it is however unlikely that this can improve the situation .
present calculations did not take into account the situation where the off - shell initial - state parton is a quark or an anti - quark .
we expect this contribution to be small , as at small @xmath2 gluons dominate significantly . on the other hand
, the process @xmath44 gives an important contribution in the collinear approach .
since in reality the probed values of @xmath2 are not extremely small , such process might be important in the high energy factorization .
practical applications require however a set of unintegrated quark distributions , consistent with the unintegrated gluon distributions .
inclusion of those is left for future studies .
as seen from the table [ table1 ] our calculations strongly underestimate the total cross section as compared to the data .
this should be probably attributed to the hard multi - parton interaction ( mpi ) effects which are not taken into account in the hybrid high energy factorization .
this statement is supported by the observation that the description of the azimuthal decorrelations is very good up to the normalization , which , when corrected , will only shift the data eventually forming a `` pedestal '' .
it is also known that mpis indeed contribute only a pedestal to the azimuthal decorrelations ( see e.g. @xcite ) , as the two partons coming from independent scatterings are completely decorrelated ( to leading order ) and thus the azimuthal decorrelation distribution is flat .
the subject of mpis in the high energy factorization is however rather complicated and needs a separate study . in principle , on the dense hadron side ( i.e. for the one probed at small @xmath2 ) , the soft mpis can be partially taken into account by means of the nonlinear term in the evolution equation , as for example in the ` ks ` unintegrated gluon density or its extensions .
therefore , one has to be careful not to make a double counting when using some phenomenological models by an inclusion of double hard scattering mechanism .
actually the problem extends beyond the saturation regime as the high energy factorization contributes to higher twists as well as mpis , at least for certain observables @xcite .
these subjects are however beyond the scope of the present work .
we acknowledge useful correspondence and discussions with albert bursche , and we thank rafamaciua for providing us the necessary grid for the kmr pdf .
the work of k.k . has been supported by narodowe centrum nauki with sonata bis grant dec-2013/10/e
/ st2/00656 .
p.k . acknowledges the support of the grants de - sc-0002145 and de - fg02 - 93er40771 .
s. dooling , f. hautmann , and h. jung , _ hadroproduction of electroweak gauge boson plus jets and tmd parton density functions _ , _ phys.lett . _ * b736 * ( 2014 ) 293298 , [ http://xxx.lanl.gov/abs/1406.2994 [ 1406.2994 ] ]
. a. van hameren , p. kotko , and k. kutak , _ three jet production and gluon saturation effects in p - p and p - pb collisions within high - energy factorization _ , _ phys.rev . _ * d88 * ( 2013 ) , no . 3 094001 , [ http://xxx.lanl.gov/abs/1308.0452[1308.0452 ] ] .
m. nefedov , v. saleev , and a. v. shipilova , _ dijet azimuthal decorrelations at the lhc in the parton reggeization approach _ , _ phys.rev . _ * d87 * ( 2013 ) 094030 , [ http://xxx.lanl.gov/abs/1304.3549[1304.3549 ] ] . a. van hameren , p. kotko , k. kutak , and s. sapeta , _
small-@xmath2 dynamics in forward - central dijet decorrelations at the lhc _ , _
phys.lett . _ * b737 * ( 2014 ) 335340 , [ http://xxx.lanl.gov/abs/1404.6204 [ 1404.6204 ] ] .
a. van hameren , p. kotko , k. kutak , c. marquet , and s. sapeta , _ saturation effects in forward - forward dijet production in p@xmath45pb collisions _ , _ phys.rev . _ * d89 * ( 2014 ) , no . 9 094014 , [ http://xxx.lanl.gov/abs/1402.5065[1402.5065 ] ] .
a. lipatov , m. malyshev , and n. zotov , _ drell - yan lepton pair production at high energies in the @xmath46-factorization approach _ , _ jhep _ * 1112 * ( 2011 ) 117 , [ http://xxx.lanl.gov/abs/1110.6582 [ 1110.6582 ] ] .
s. catani and f. hautmann , _ high - energy factorization and small x deep inelastic scattering beyond leading order _ , _ nucl.phys . _ * b427 * ( 1994 ) 475524 , [ http://xxx.lanl.gov/abs/hep-ph/9405388 [ hep - ph/9405388 ] ] .
p. kotko , k. kutak , c. marquet , e. petreska , s. sapeta , _
_ , _ improved tmd factorization for forward dijet production in dilute - dense hadronic collisions _ , http://xxx.lanl.gov/abs/1503.03421 [ 1503.03421 ] .
a. van hameren , p. kotko , and k. kutak , _ multi - gluon helicity amplitudes with one off - shell leg within high energy factorization _ , _ jhep _ * 1212 * ( 2012 ) 029 , [ http://xxx.lanl.gov/abs/1207.3332 [ 1207.3332 ] ] .
m. strikman and w. vogelsang , _ multiple parton interactions and forward double pion production in pp and da scattering _ , _ phys.rev . _ * d83 * ( 2011 ) 034029 , [ http://xxx.lanl.gov/abs/1009.6123[1009.6123 ] ] . | we calculate several differential cross sections for @xmath0 and high-@xmath1 jet production in the forward rapidity region at the lhc using the hybrid high energy factorization .
we test various unintegrated gluon distributions involving subleading bfkl effects ( such as kinematic constraint , running strong coupling and dglap correction ) and compare the results with experimental data obtained by the lhcb experiment .
we find that the hard scale dependence of unintegrated gluon distributions , which effectively resums the sudakov - type logarithms on the top of the resummation of the small @xmath2 logarithms , is essential to describe the normalized azimuthal decorrelations between the @xmath3-boson and the jet . |
optical fiber grating sensors are becoming increasingly widespread in a wide range of applications @xcite .
their characteristics include immunity to electromagnetic interference , high sensitivity , fast response , and keeping their calibration with time .
furthermore they possess a high tolerance to harsh environments such as sea water , concrete , and extreme temperatures .
the measured parameter , the resonant wavelength of the grating , is usually linear in the magnitude of the measurand , e.g. temperature or strain , allowing for an accurate quantitative estimate of the measurand .
photonic crystal fibers ( pcfs ) @xcite are optical fibers , in which the cladding consists of a microstructured array of air holes running along the fiber axis . here
we consider index guiding pcfs , which have a solid core in the form of a missing air hole in the center . in index
guiding pcfs light is confined to the core by modified total internal reflection , analogous to guiding in standard optical fibers .
the holes in the cladding are most commonly placed in a periodic triangular structure , which is characterized by a lattice constant ( pitch ) and the diameter of the holes .
pcfs are typically made of silica @xcite or , more recently , polymer materials , where they are termed microstructured polymer optical fibers ( mpofs ) @xcite .
the hole structure in the cladding determines the optical properties of the pcf , which allows for a large degree of freedom in tayloring key properties , such as dispersion and group velocity .
in standard optical fibers the silica in the core contains dopants to increase the refractive index .
the refractive index of the doped silica is sensitive to uv radiation and this may be used for inscription of gratings .
this technique can not be used for pure silica pcfs , since the core is not uv sensitive . for long - period gratings ( lpgs ) in pcfs other techniques exist , such as using an electric arc @xcite , a co@xmath1 laser @xcite , or mechanical microdeformation @xcite .
otherwise the pcf can be specially made with a uv sensitive core for fiber gratings allowing for the inscription of both lpgs and bragg gratings ( bgs ) @xcite .
polymer has a smaller young s modulus and greater elastic limit than silica , which makes it a good material for measuring strain and bending .
in addition , polymer is biocompatible and can bind biomolecules directly to its surface , making it ideal for biosensing .
standard polymer optical fibers are made of poly - methyl methacrylate ( pmma ) , which is inherently photosensitive , allowing for both bgs @xcite and lpgs @xcite to be inscribed using a uv lamp .
the inherent uv sensitivity was also recently used to write bgs in mpof @xcite , whereas heat treatment under mechanical stress was used to write the first permanent lpg in an mpof @xcite .
an important property of grating sensors in silica pcfs are that they are largely temperature insensitive , showing @xmath2 1 pm / k for both bgs @xcite and lpgs @xcite .
the strain sensitivity is of the same order for lpgs and bgs in pcfs , and thus strain sensors may be realized with reduced cross sensitivity to temperature sensing .
in addition , the holes of a pcf can be infiltrated with a substance having a high temperature coefficient to reduce or enhance the temperature sensitivity @xcite .
whereas fiber - optic strain and temperature sensors use the material properties of the fiber , fiber - optic biosensors and refractive index sensors use the principle of evanescent wave sensing .
in particular , there is a growing interest in optical biochemical sensors @xcite . here
pcfs have the advantage that the analyte may be infiltrated into the air holes and thereby create a strong interaction with the probing electromagnetic field @xcite .
when using mpofs one can take advantage of different polymers with special biosensing properties , such as topas @xcite .
liquid infiltrated pcf - lpgs @xcite and pcf - bgs @xcite have been demonstrated , which is an important step towards the ultimate goal of developing a so - called _ label - free _ biosensor in a pcf .
fiber gratings potentially allow for label - free biosensing with pcfs @xcite by tracking how the grating resonance changes during operation .
this may open for new possibilities in biomedical applications @xcite .
the dispersion plays a crucial role for all types of grating sensors in standard optical fibers @xcite .
the dispersion of pcfs depends strongly on the hole structure in the cladding and thus the performance of the pcf grating sensor can be designed by an appropriate choice of pitch and hole size . in this paper
we explore the possibility of optimizing the hole structure of pcfs for sensing purposes .
we do this by analyzing the sensitivity of triangular pcf gratings for different pitch and hole diameters .
the applications under study are biosensing , refractive index sensing , strain , and temperature sensing .
we show that the sensitivity is strongly influenced by the choice of the structural parameters of the pcf .
we employ realistic values for the fibers parameters and give realistic values of pitch and hole diameter that minimize or maximize the sensitivity to different measurands .
we find that bgs are best characterized in the usual way by their sensitivity , @xmath3 , defined as the resonant wavelength shift divided by the resonant wavelength .
in contrast , we find that the lpgs are best characterized by their quality factor , @xmath0 , defined as the resonant wavelength shift divided by the full - width at half maximum ( fwhm ) of the resonance dip .
the eigenmodes of the pcfs for a temporally harmonic electrical field , @xmath4 , may be found from helmholtz eigenvalue equation , @xmath5 where @xmath6 is the angular frequency , c is the speed of light in vacuum , @xmath7 is the dielectric function , and @xmath8 is the transverse coordinate vector .
the @xmath9-axis is chosen to coincide with the fiber axis .
we look for plane - wave solutions of the form @xmath10 with @xmath11 being the propagation constant .
the effective index is defined as @xmath12 , where @xmath13 is the free - space wavevector and is related to the frequency and wavelength by @xmath14 .
the effective indices of the core mode and a resonant cladding mode are denoted by @xmath15 and @xmath16 .
the orthonormalization for eq .
( [ eq : eigenvalue ] ) is @xmath17 the pcf is typically made of two materials with a high refractive index contrast .
a general dielectric function , @xmath18 , for a two material triangular pcf structure is described by high and low index regions as seen in fig .
[ fig : analyte ] .
the periodic triangular structure is spanned by the lattice vectors @xmath19 and @xmath20 .
the general dielectric function equals @xmath21 for the base region and @xmath22 for the hole regions , with @xmath23 and @xmath24 being the refractive indices of the base and hole material , respectively .
the effective index of the core depends strongly on the wavelength , the hole diameter , and the pitch .
the hole diameter , @xmath25 , and pitch , @xmath26 , are indicated in fig .
[ fig : analyte ] . large hole diameters and
small pitch also give a small core and therefore a strong effective nonlinearity .
triangular silica pcfs are single mode for @xmath27 for all wavelengths [ ] . for small hole sizes ( @xmath28 ) the pcf is only weakly guiding , and bend losses increase drastically .
bend losses also become pronounced in the short wavelength limit , @xmath29 , regardless of hole diameter .
pcfs can be designed to have a large mode area .
the transmission spectrum of an lpg in a pcf only shows a few cladding mode resonances , whereas several resonances are seen in a standard optical fiber lpg spectrum .
the cladding modes of a pcf are different from the cladding modes of a standard optical fiber , since the fiber cross sections obey very different symmetries .
in particular a pcf obeys a six - fold rotational symmetry ( @xmath30 ) while the standard optical fiber has circular symmetry .
the effective indices of the cladding modes of a pcf are closely spaced , and this spacing is much smaller than the difference between the core effective index and the effective index of any cladding mode . in other words ,
the beat length between any two cladding modes is much longer than the beat length between a cladding mode and the core mode . in practice
we can assume @xmath31 for all cladding modes , where the fundamental space filling mode , @xmath32 , is the effective index of an infinite cladding structure spanned by the lattice vectors in fig .
[ fig : analyte ] .
we have calculated the effective index for the core mode , @xmath15 , and the fundamental space filling mode , @xmath32 , for different hole to pitch and wavelength to pitch ratios , with the base material having a refractive index @xmath33 , which is the refractive index of silica at the wavelength @xmath34 m .
the effective indices are shown in fig [ fig : neffs ] .
the effective indices decrease with increasing hole size ( @xmath35 ) or increasing normalized wavelength ( @xmath36 ) .
the effective index of the core mode is larger than that of the fsm , regardless of the fiber structure parameters ( @xmath25,@xmath26 ) .
coupled mode theory @xcite is a simple and compact theory that has been successful in accounting for the characteristics of fiber gratings . it allows us to derive the qualitative behavior of the gratings with some general , simplifying assumptions on the physical mechanisms .
the theory assumes that near resonance the electromagnetic field propagating through the fiber may be expressed as a superposition of eigenmodes of the unperturbed fiber @xmath37 , where ` @xmath38 ' denotes a resonant mode , and the @xmath39 s are amplitude coefficients that depend on @xmath9 due to the perturbation . for bragg gratings
the resonant mode is the counter propagating core mode , and for long - period gratings the resonant mode is a copropagating higher order mode , usually a cladding mode .
the coupled - mode equations constitute a set of linear coupled differential equations .
they may be expressed in a compact matrix form as @xmath40 = -i\left [ \begin{array}{cc } 0 & \kappa_i e^{i\delta_i z } \\ \kappa_i^
* e^{-i\delta_i z } & 0 \end{array } \right ] \left [ \begin{array}{c } a_\textrm{co}(z,\delta_i)\\ a_i(z,\delta_i ) \end{array } \right ] \label{eq : cmt}\end{aligned}\ ] ] where the detuning in the resonance between the core mode and the resonant mode , @xmath38 , is @xmath41 and the coupling constant is @xmath42 where the integration is over the fiber cross section , @xmath43 .
@xmath44 is the first phase matched fourier component ( @xmath45 ) in the expansion @xmath46 . in eq .
( [ eq : cmt ] ) we have neglected the diagonal terms , the so called self - coupling terms . the self - coupling arises from the perturbation of the individual modes due to the index change of the grating .
the self - coupling shifts the resonant wavelength some nm s , while otherwise leaving the spectrum of the grating unaltered . since this effect is typically small we neglect it for our purposes .
using eq .
( [ eq : detuning ] ) we obtain a resonance condition in terms of effective indices and the wavelength @xmath47 where @xmath48 is the effective index of the resonant mode .
the equation is valid both for lpgs , @xmath49 , and bgs , @xmath50 . in the following we use the bar for indicating a difference between the core and the resonant modes as done in eq .
( [ eq : res_lpg ] ) .
thus @xmath51 for the propagation constant , and for the group index we have @xmath52 , where @xmath53 is the beat length , with @xmath54 .
the resonance conditions must be solved considering the dispersion of the effective indices , since these are strongly dependent on the pitch and hole size of the pcf structure as seen in fig .
[ fig : neffs ] . couple - mode theory in the form of eq .
( [ eq : cmt ] ) will give an expression for the transmission in terms of the detuning and coupling constant .
however , experimentally the transmission is measured as function of wavelength . in fig .
[ fig : fwhm ] the fwhm for the detuning , @xmath55 , and the wavelength , @xmath56 , are defined . by definition
we have @xmath57 , and equivalently for the wavelength we have @xmath58 . to find a relation between @xmath59 and @xmath60 that incorporates the dispersion of the effective indices we expand the detuning around the resonance wavelength to first order in @xmath61 @xmath62 with this equation it is possible to obtain two equations relating @xmath63 and @xmath64 to @xmath65 and @xmath66 .
subtracting the two equations and taking the absolute value of both sides , we obtain an expression for @xmath67 and @xmath68 @xmath69 which incorporates the dispersion of the effective indices in a consistent manner .
we note that the expression is not valid at group index matching between the two resonant modes , @xmath70 . in this limit
, however , the fwhm goes to infinity , and any wavelength shifts of the sensor will become difficult to detect .
long - period gratings couple the core mode to a co - propagating high - order mode .
the difference in effective index of two such modes is typically small , hence by eq .
( [ eq : res_lpg ] ) the grating period of such a grating must be long , @xmath71 . for lpgs we put the subscript @xmath38 = fsm , and @xmath72 typically the fwhm of an lpg is of the order of nanometers .
the minimal detectable shift in resonance condition is thus set by the fwhm and not the spectral resolution of the spectrum analyzer .
the coupled - mode equations for the long - period grating may be solved analytically from eq .
( [ eq : cmt ] ) yielding the transmission coefficient @xmath73 at resonance ( @xmath74 ) the transmission is sinusoidal , @xmath75 , where @xmath76 is the length of the grating .
the first zero for @xmath77 is at @xmath78 . ignoring the wavelength dependence of the coupling constant , @xmath79 , we substitute @xmath80 and solve @xmath81 numerically .
this yields @xmath82 .
we can find the fwhm of the detuning as @xmath83 .
the wavelength fwhm of the transmission can then be found using eq .
( [ eq : fwhm ] ) @xmath84 the wavelength fwhm obviously is large when the group index matching is close to zero , @xmath85 .
in conclusion we have found an expression , characteristic for lpgs , that relates @xmath60 directly with the length of the grating , the group index and the resonant wavelength .
the expression is derived under the assumption , that @xmath86 bragg gratings couple the incident core mode to a counter - propagating core mode , hence @xmath87 and the effective index is @xmath88 and the coupling constant @xmath89 . from eq .
( [ eq : res_lpg ] ) we find that the period of such a grating must be comparable with the wavelength , @xmath90 . the fwhm of a bg is usually very narrow compared with an lpg .
although a narrow fwhm gives a high resolution for a sensor , it also puts high demands on production tolerances and the optical spectrum analyzer .
thus the minimal detectable shift in the resonant wavelength is set by the spectrum analyzer .
fluctuations in the effective index caused by imperfections in the pcf structure or the grating period along the grating enters by eq .
( [ eq : res_bg ] ) in the transmission spectrum and can inflict a significant broadening of the resonance peak @xcite . on the beneficial side one
need only to couple light into one end of the pcf , since it is possible to measure the reflection spectrum of the grating .
the coupled - mode equations for the bragg grating may be solved analytically from eq .
( [ eq : cmt ] ) yielding a transmission coefficient @xmath91 at resonance the transmission is equal to @xmath92 .
assuming that the coupling constant is constant , @xmath93 , we find numerically that the @xmath59 can be approximated reasonably well by @xmath94 according to this estimate the fwhm decreases with the grating length until a lower bound set by the coupling strength itself .
typically @xmath95 and we can neglect the second term in eq .
( [ eq : bg_fwhm_l ] ) and obtain @xmath96 .
the wavelength fwhm is then @xmath97 we note that the fwhm generally increases with the resonance wavelength squared .
in conclusion we have found an expression , characteristic for bgs , that relates @xmath60 directly with the coupling constant , the group index of the core mode and resonant wavelength .
we have made the assumption the coupling constant is independent of wavelength .
the fiber sensor is perturbed by a measurand to produce a change in the measured quantity , in our case the resonant wavelength . for quantitative and qualitative estimation of the measurand
a linear relation to the resonant wavelength is assumed .
since the resonant wavelength shifts are generally small compared to the wavelength , it is natural to employ perturbation theory and taylor expand the resonance condition to first order around the resonance wavelength .
consider a general measurand , @xmath98 .
we assume that the shift in resonant wavelength as well as in the measurand are sufficiently small , and we expand the resonance condition eq .
( [ eq : res_lpg ] ) in the resonant wavelength and the measurand to first order @xmath99\lambda_\textrm{g}.\end{aligned}\ ] ] identifying @xmath100 by eq .
( [ eq : res_bg ] ) and @xmath101 , we obtain an expression for the sensitivity @xmath102 the shift in the resonant wavelength is thus dependent on the change in the difference of the effective indices relative to the difference in group indices .
we have chosen the symbol ` @xmath3 ' for the sensitivity to match the notation of shu et al .
obviously , when @xmath103 is large the sensor will give large wavelength shifts , which will give a high sensitivity and accuracy of the measurand .
however , changing the fiber parameters not only changes the wavelength shifts but also change the full - width half maximum ( fwhm ) of the resonance dip .
thus , even though @xmath103 may be large for a certain fiber structure , the wavelength shifts will be difficult to detect if they are much smaller than the fwhm of the resonance dip .
in particular , we therefore note that the sensitivity is high close to group index matching , @xmath104 , which by eqs .
( [ eq : lpg_fwhm],[eq : bg_fwhm ] ) is also where the fwhm is wide .
we introduce the quality factor , @xmath0 , of the grating sensor , which is the wavelength shift divided by the fwhm , @xmath105 we note that @xmath0 , in contrast to @xmath3 , does not increase when @xmath106 .
perfect group index matching @xmath107 only occurs for lpgs , as we will see in the following .
we have derived two expressions for the performance of a fiber grating sensor .
the sensitivity , @xmath3 , is inversely proportional to the group index mismatch , @xmath108 , and expresses the shift in the resonant wavelength relative to the resonant wavelength .
the quality factor , @xmath0 , in contrast , expresses the wavelength shift relative to the fwhm .
ideally both the sensitivity , @xmath3 , _ and _ the quality factor , @xmath0 , should be large .
the fwhm of lpgs is usually larger than the resolution of the spectrum analyzer .
the quality factor is thus the most important measure of the performance of lpgs .
the fwhm of bgs is usually narrow , and comparable to the resolution of the spectrum analyzer .
the sensitivity is then the most important for bgs .
we employ a commercial finite element code ( comsol multiphysics @xcite ) for the calculation of the effective indices .
the code uses an adaptive mesh algorithm for generating the mesh .
the meshes consist of 60,000 - 250,000 quadratical lagrange elements .
the core mode is simulated by considering a cross section with 12 - 25 rings of air holes : 12 rings for @xmath109 , 15 rings for @xmath110 , and 25 rings for @xmath111 . outside the rings of air holes
there is a layer of air @xmath112 thick to reduce the effects of a finite cladding .
the large number of rings of air holes is chosen such that effects of a finite cladding is small .
otherwise reflections may arise from both the silica - air interface and the metallic boundary outside the air layer .
this induces an error on the effective index . in our simulation
the number of rings was sufficiently large that the relative error on the effective indices is smaller than @xmath113 .
large holes give a strong confinement of the core mode , and obviously fewer rings are needed compared to small holes .
we generally use only data where @xmath114 such that the error is negligible .
this also makes physical sense , since the pcf is only weakly guiding below this limit . for the fundamental space filling mode we use the bloch theorem and we need only to simulate a single unit cell containing only one hole .
the unit cell is spanned by the lattice vectors seen in fig .
( [ fig : analyte ] ) of an infinite triangular cladding structure .
we impose periodic boundary conditions on the unit cell .
in our simulations we solve eq .
( [ eq : eigenvalue ] ) for the propagation constants , @xmath11 , for the modes of the fiber at a fixed frequency , @xmath6 , which enables us to include material dispersion in a consistent manner . for the refractive indices of the materials we use a sellmeier expression for the silica and
use the refractive index of water given in ref .
all plots in the following are shown as function of the wavelength relative to the pitch , but at a fixed wavelength .
if the material dispersion is neglected then space and time in maxwell s equation may be scaled with the pitch of the pcf . however , in our study the material dispersion can not be neglected .
we thus choose a wavelength @xmath115 m .
the grating period and therefore the beat length for lpgs can be determined from the effective indices using eq .
( [ eq : res_lpg ] ) .
the beat length varies significantly with the hole size and pitch , as shown in fig .
[ fig : lg ] .
the resonance condition , eq .
( [ eq : res_lpg ] ) , can be satisfied over a range of wavelengths if @xmath116 . for these values
the fwhm of an lpg is very large .
the between the curves of in fig .
[ fig : lg ] the gray area indicates @xmath117 and the white area @xmath118 .
the boundary between these two areas indicates @xmath119 .
when the slopes of the curves in fig .
[ fig : lg ] are parallel with the normalized wavelength axis the resonance condition eq .
[ eq : res_bg ] is nearly matched for multiple normalized wavelengths , @xmath36 , we may thus expect that the wavelength fwhm is wide . by eq .
( [ eq : fwhm ] ) the fwhm is widest when @xmath119 , indicated by the boundary of the gray area .
it may seem surprising that the @xmath119 does not coincide with zero slope of the curves .
the curves in fig .
[ fig : lg ] , however , are evaluated for a fixed wavelength with a variable pitch , and not for a fixed pitch with variable wavelength .
the difference in normalized wavelength for group index matching and for zero slope of the curves is caused by material dispersion of silica .
for bgs @xmath120 in the resonance condition eq .
( [ eq : res_bg ] ) only varies a few percent for all wavelengths , pitch values , and hole sizes .
thus the dispersion effects on the resonance condition are small .
correspondingly , it can shown that dispersion effects on the wavelength fwhm is also small , because the group index mismatch , @xmath121 , in eq .
( [ eq : fwhm ] ) only varies slightly ( @xmath122 12% ) for index guiding pcfs , since @xmath15 always dominates over its derivative in the group index , @xmath123 .
there is a fundamental difference between pcfs and standard optical fibers is that a pcf consists of two , often very different , materials .
standard optical fibers also consist of two materials : pure silica and a core of doped silica . in practice , however , the material parameters for the doped silica may often , to a good approximation , be taken to be equal to those of pure silica . for a pcf , e.g. the strain optic response of silica is completely different from that of air , which is obviously zero for our applications . to correctly handle a two material fiber
one must determine the field energy intensity fraction , @xmath124 , inside the holes of the pcf .
the field energy fraction for the base material is then , @xmath125 .
it is not possible to simply add a contribution , e.g. stress optic , directly to the resonance condition , eq .
( [ eq : res_bg ] ) .
instead every contribution must be weighted with the fraction of field intensity in the corresponding material , as shown in the appendix .
it is possible to make some generalizations about field energy fraction in the holes , @xmath124 , for index guiding fibers . in the short wavelength limit where the wavelength is much smaller than the pitch of the structure , @xmath126 , the field
is confined to the high index base material . in the long wavelength limit , @xmath127 ,
the changes in the field are small over a length of the pitch .
this imply @xmath128 where @xmath129 is the air filling fraction , i.e. the fraction of area covered by the holes in the cross section of the fiber .
the approximation @xmath130 is natural as a first approximation since the field intensity in the air holes @xmath124 is generally much smaller than the field intensity in the base material .
but pcfs can possibly have air filling fractions of almost unity , implying that also @xmath124 can be almost unity , and therefore the effect can not be neglected .
we have shown @xmath124 for the fsm in fig .
[ fig : fufsm_air ] for air filled holes ( @xmath131 ) .
@xmath124 for the core mode of a pcf with water filled holes ( @xmath132 ) is shown in fig .
[ fig : fuco_h20 ] ) and the corresponding fsm is shown in fig . [
fig : fufsm_air ] . clearly @xmath133 is larger for the fsm than for the core mode , and it increases with increasing wavelength and increasing hole diameters @xmath35 .
the @xmath133 also increases when the index contrast between the hole and base material is decreased .
hence @xmath133 is larger for water filled pcfs than for air filled . in
the following we refer to two pcfs to as the large core and the small core pcf .
the structure parameters for the small core pcf is @xmath134 2 ) and for the large core pcf @xmath135 .
we choose the length of the gratings to be 30 mm for lpgs with additionally @xmath136 for bgs .
the sensitivity , @xmath3 , is independent of these parameters .
the quality factor , @xmath0 , depends inversely proportional on the length of the grating , when assuming constant @xmath137 . for bgs
the quality factor also depends inversely proportional to the coupling constant , @xmath138 .
refractive index sensors can be used in the application of continuous monitoring of gasses in the oil industry .
the gas flows through the pcf is detected by measuring the overall refractive index of the hole contents .
optical fibers have the advantage that there are no electric currents thus avoiding completely the possibility of generating electric sparks .
the sensitivity for refractive index sensing is also used for benchmarking the sensitivity of evanescent wave biosensors , which will be studied in the next section .
alternatively , a refractive index sensor may be used as tunable grating if the holes are filled with a substance whose refractive index can be controlled @xcite .
tunable gratings may have applications within modulation of optical signals .
infiltrating substances into pcf - lpgs that are absorbing at the resonant wavelength requires special considerations , as discussed by daxhelet et al .
@xcite . in brief
the attenuation induced by the absorption of the substance becomes significant if the modes are significantly attenuated over the length of a period of the grating .
the interference of modes in the grating becomes weakened and the resonance dip is broadened . since the grating period of an lpg is much longer than that of a bg , the lpg can be considerably affected by absorbtion , while the bg is largely unaffected even by a high absorbtion .
the change in the effective index of a mode induced by a change of refractive index of the hole contents is proportional to the fraction of the time averaged field energy intensity ( @xmath139 ) inside the holes . for bgs
the change in @xmath140 is ( see appendix for derivation ) @xmath141 where @xmath142 is the change in the refractive index , @xmath24 , of the content in the hole .
@xmath143 is the fraction of the field energy intensity of the core mode inside the holes .
the sensitivity is also proportional to the group index @xmath144 . for lpgs
the change in @xmath145 is @xmath146 it is clear from eqs .
( [ eq : gamma],[eq : q ] ) that the sensitivity and the detectability are both largely determined by @xmath124 . for bgs @xmath124 should be as large as possible . for bgs
the group index is close to the effective index , @xmath147 since the derivative is negligible as seen in fig .
[ fig : neffs ] .
thus for bgs the sensitivity is determined by @xmath143 . in fig .
[ fig : fuco_h2o ] we realize that a large @xmath148 implies a small pitch and a large hole diameter . for bgs the size of @xmath0 is largely determined by the pitch .
@xmath0 also increases with decreasing hole size , but to a lesser extent . in order to have a high @xmath0 for lpgs the difference
@xmath149 should be large .
comparing figs .
[ fig : fuco_h2o ] and [ fig : fufsm_h2o ] we realize that this also implies a small pitch and a large hole diameter . but in this case the difference falls off beyond a certain point in the wavelength pitch ratio .
this is because @xmath150 and @xmath151 approaches the same limit for large wavelengths , and thus the @xmath0 decreases , as seen in fig .
[ fig : lpg_nr_q ] . in fig .
[ fig : lpg_nr_q ] there is also indicated the group index matched line for lpg where @xmath152 .
along this line the peak of the lpg is very wide by eq .
( [ eq : fwhm ] ) . on the left side of this line
the shift in wavelength are positive for a positive @xmath142 . on the right hand side
the shift in wavelength is negative for positive @xmath153 .
the reason for this , is that @xmath154 enters in eq .
( [ eq : gamma ] ) in the denominator , and thus its sign determines the sign of the wavelength shift . for bgs group index
matching @xmath155 does not occur since for the considered index guiding pcfs the effective index is always larger than its derivative .
for the sensitivity of bgs eq .
( [ eq : bgrefrac ] ) is inserted in eq .
( [ eq : gamma ] ) , and plotted for the wavelength 1.0 @xmath156 m , in fig .
[ fig : bg_nr_sens ] . for lpgs eq .
( [ eq : lpgrefrac ] ) is inserted in the expression for the @xmath0 , eq .
( [ eq : q ] ) and plotted for the wavelength 1.0 @xmath156 m , fig .
[ fig : lpg_nr_q ] . to give some specific details we choose a typical large mode area , endlessly single mode pcf with structure values : @xmath157 m and @xmath158 .
this gives a core diameter of 10 @xmath156 m .
we choose the length of the grating to be 30 mm , and the coupling strength is @xmath159 .
we choose different wavelengths for both lpgs and bgs : 600 , 900 , and 1550 nm .
the data is presented in table [ table : refrac_lin ] .
the sensitivity increases with increasing wavelength since the evanescent wave is increasing inside the holes , corresponding to an increasing @xmath124 .
the sensitivity , @xmath3 , for lpgs is four orders of magnitude larger than the sensitivity for bgs . but
taking the fwhm of the resonance peak into account , the @xmath0 is only slightly larger for lpgs than for bgs .
the data for a small core pcf are presented in table [ table : refrac_nl ] .
the small core pcf has significantly higher sensitivity than the large core pcf .
notice that the sensitivity for the small core has opposite sign to the large core pcf .
also @xmath0 _ decreases _ with increasing wavelength .
the shift is negative because the pcf s structure parameters places the pcf on the left of the group index matching line @xmath119 indicated in fig .
[ fig : lpg_nr_q ] . the minimum detectable change in refractive index for lpgs can be calculated . for l = 30 mm
, we get a maximal @xmath160 . taking that the minimal detectable wavelength is @xmath161 we get a maximal sensitivity of @xmath162 .
the same calculation for a bg with @xmath163 gives @xmath164 .
both lpgs @xcite and bgs @xcite have been used for measuring the refractive index of a liquid analyte inside the holes of the pcf . comparing with experiment for lpgs @xcite
the pcf has parameters @xmath165 m , and the lpg has @xmath166 nm and @xmath76 = 18.2 mm .
the minimal detectable wavelength shift is @xmath167 nm . theoretically the parameters give @xmath168 = 52 ( no units ) and @xmath169 nm which agrees with the experimental value @xmath170 nm . the minimal sensitivity given the mentioned minimal detectable wavelength is theoretically @xmath171 which is in agreement with the experimental value @xmath172 .
the shift in resonances wavelength as function of sensitivity is experimentally @xmath173 = 7500 nm but gives theoretically a lower value @xmath174 = 2300 nm .
.comparison of sensitivity and the quality factor for pcf biosensors and refractive index sensors for both lpgs and bgs at different wavelengths for two different pcfs : a large core of 10 @xmath156 m and a small core pcf of 1.5 @xmath156 m . [ cols="^,^,^,^,^,^,^,^ " , ] [ table : temp_nl][table : strain_lin][table : strain_nl ] fiber optic strain sensors have found application within civil engineering , where they provide a stable and sensitive measurement of the strain in constructions .
strain is defined as the relative elongation : @xmath175 of a piece pcf of length @xmath76 .
often millistrain , @xmath176 , is used instead of strain . when the pcf is elongated , the pcf contracts transversely to the applied strain .
this is expressed by poisson s ratio which formally is the ratio between lateral and tensile strain with a negative sign . for silica glass poisson
s ratio is given by @xmath177 .
strain also changes the refractive index of materials .
the strain - optic coefficient for silica is @xmath178 @xcite . since strain changes the length , pitch , and the material refractive indices we apply partial derivatives ( the chain rule ) to the linearization of the effective index @xmath179 since @xmath180 we have @xmath181 and @xmath182 gives @xmath183 .
an increase in length of the pcf with a fixed effective index is mathematically equivalent to a fixed length with an increase in effective index , @xmath184 the derivative of the effective index with respect to the refractive index of the base material may be evaluated by inserting the strain - optic coefficient in the perturbative formula , eq .
( [ eq : deltaneff_refrac_app_base ] ) , given in the appendix .
the formula for bgs is @xmath185 while for lpgs we obtain @xmath186 eqs .
( [ eq : bgstrain],[eq : lpgstrain ] ) are inserted in the expression for the sensitivity , eq .
( [ eq : gamma ] ) and calculated for the wavelength 1.0 @xmath156 m .
the results are shown in figs .
[ fig : bg_strain_sens ] and [ fig : lpg_strain_sens ] , respectively .
the sensitivity for bgs is very close to being constant regardless of pitch , hole diameters , and wavelength , as seen in fig .
[ fig : bg_strain_sens ] and table [ table : temp_lin ] .
this can understood by that fact that @xmath15 and @xmath144 are of similar magnitudes , such that @xmath187 .
we can neglect the second term in eq .
[ eq : bgstrain ] .
we then put @xmath188 , where @xmath189 is the refractive index of silica . inserting this in the expression for the sensitivity , eq .
( [ eq : gamma ] ) , we realize that @xmath190 and @xmath191 .
this result is in very good agreement with the values for bgs displayed in table [ table : strain_lin ] and in fig .
[ fig : bg_strain_sens ] .
the quality factor , @xmath192 , for lpgs with @xmath193 tends to increase for increasing hole size , while being almost independent of wavelength , as seen in fig .
[ fig : lpg_strain_q ] . for these wavelength the first term in eq .
( [ eq : lpgstrain ] ) , the elongation of the pcf , is dominating over the strain - optic effect and the lateral strain .
for @xmath194 the sensitivity is generally lower since @xmath120 decrease for decreasing wavelengths , but also because the strain - optic effect and the lateral strain become significant .
the dips indicate a net cancelation of the terms in eq .
( [ eq : lpgstrain ] ) . to validate our theory we compare with experiments .
dobb et al .
@xcite characterized lpgs in pcfs inscribed with an electric arc . for a pcf with structure parameters @xmath195 m ,
@xmath196 m they found a shift of resonance wavelength with -2.5 pm/@xmath197 at the wavelength 1688 nm .
using the structure parameters in the theory yields -3.7 pm/@xmath197 . for a pcf with structure parameters @xmath198 m , @xmath199 m
, they found -2.08/@xmath197 at 1403 nm wavelength and the theory gives -3.2/@xmath197 .
thus we have an agreement of the right magnitude and right sign of the wavelength shift .
their pcf structure values , @xmath200 , gives a @xmath0 that is close to the dips in fig .
[ fig : lpg_strain_q ] .
the dips corresponds to a cancelation of the terms in eq .
( [ eq : lpgstrain ] ) and close to the dips the quality factor is more susceptible to errors between experimental and numerical parameters .
martelli et al .
used a pcf with a erbium - doped uv sensitive core to inscribe bgs @xcite .
their structure parameters are roughly @xmath201 m , @xmath202 m . at a wavelength of 1535
nm the wavelength shift is 1.2 pm/@xmath197 .
the theory gives 1.25 pm/@xmath197 for a pure silica pcf .
a doped core does not change the strain - optic coefficient significantly due to a doped core @xcite .
for bgs in a small core fiber with the structure parameters @xmath203 and @xmath204 m @xcite it is found experimentally a sensitivity of 1.25 pm/@xmath205 at the wavelength 1508 nm .
the core was doped with germanium to make the core uv - sensitive .
our theory gives 1.16 pm/@xmath205 in good agreement with the experimental value .
in conclusion we have found that the sensitivity for bgs is approximately constant both for temperature and strain sensing , with the coefficients @xmath206k and @xmath207 for pure silica pcfs .
the wavelength shift both for bgs and lpgs is generally of the order of 5 pm / k , and the sensitivity and quality factor are an order of magnitude large ( in k ) compared with strain ( in microstrain ) . the quality factor for lpgs for temperature and strain sensing has roughly the same characteristics .
in conclusion we have shown that the performance of lpg sensors are best characterized by the detectability , @xmath0 , rather than the sensitivity , @xmath3 .
we have calculated expressions for the sensitivity , that incorporates the material - optic responses to the external perturbation , as well as treated effects of having air holes in the structure .
using the results of a fully - vectorial finite element calculations , incorporating material dispersion , the sensitivity and detectability have been calculated for a range of pcf structure parameters , i.e. hole diameters and pitch lengths .
we have proposed expressions for the sensitivity for internal refractive index sensing , label - free biosensing , strain , and temperature sensing . for strain and temperature
we have included both thermo - optic and strain - optic effects , as wells as the effects of both longitudinal and lateral elongation .
the formulas are of general form and apply to all pcfs as well as standard optical optical fibers . for pcfs infiltrated with samples
, the @xmath0 for refractive index sensing and biosensing may be improved by several orders of magnitude by appropriate choice of the hole size and pitch . for temperature and strain index sensing with bg sensors we have shown that the sensitivity is almost constant for all pcfs . for lpgs an accurate theoretical prediction of the sensitivity for temperature sensing or
strain sensing can be difficult since the sensitivity strongly affected by the waveguide dispersion .
however , our analysis shows that the low temperature sensitivity of pcf - gratings , @xmath2085 pm / k , is generally valid for all pcfs .
l. rindorf s e - mail address is [email protected] .
the linear change in propagation constant by a small perturbation in the dielectric function , @xmath209 , is @xcite @xmath210 @xmath43 denotes the fiber cross section . by setting the norm of the orthogonal solutions to the helmholtz equation ( [ eq : eigenvalue ] ) to unity , @xmath211 where @xmath212 denotes hermitian conjugate ( the complex conjugate transposed ) , we get @xmath213 with the group index , @xmath214 .
we define the field energy intensity in the air holes of the mode @xmath215 by @xmath216 @xmath217 is the part of the fiber cross section covering holes , and @xmath218 . the linear change in effective index of mode
@xmath215 caused by a change of the refractive index of the holes is then @xmath219 where @xmath220 is the refractive index of the holes .
the change is proportional to @xmath221 the corresponding linear change in effective index caused by a change of the refractive index of the base material is @xmath222 where @xmath223 is the refractive index of the base material .
it is seen that the change is proportional to the field intensity in the base material , @xmath125 .
m. a. van eijkelenborg , m. c. j. large , a. argyros , j. zagari , s. manos , n. issa , i. bassett , s. fleming , r. c. mcphedran , c. m. de sterke , and n. a. p. nicorovici , `` microstructured polymer optical fibre , '' opt .
express * 9 * , 319 327 ( 2001 ) .
z. li , y. tam , l. xu , and q. zhang , `` fabrication of long - period gratings in poly(methyl methacrylate - co - methyl vinyl ketone - cobenzyl methacrylate)-core polymer optical fiber by use of a mercury lamp , '' opt . lett . *
30 * , 11171119 ( 2005 ) .
h. dobb , d. webb , k. kalli , a. argyros , m. large , and m. van eijkelenborg , `` continuous wave ultraviolet light - induced fiber bragg gratings in few- and single - mode microstructured polymer optical fibers , '' opt . lett . * 30 * , 32963298 ( 2005 ) .
o. frazao , j. p. carvalho , l. a. ferreira , f. m. araujo , and j. l. santos , `` discrimination of strain and temperature using bragg gratings in microstructured and standard optical fibres , '' meas .
* 16 * , 21092113 ( 2005 ) .
m. p. delisa , z. zhang , m. shiloach , s. pilevar , c. c. davis , j. s. sirkis , and w. e. bentley , `` evanescent wave long - period fiber bragg grating as an immobilized antibody biosensor , '' analytical chemistry * 72 * , 28952900 ( 2000 ) .
j. b. jensen , l. h. pedersen , p. e. hoiby , l. b. nielsen , t. p. hansen , j. r. folkenberg , j. riishede , d. noordegraaf , k. nielsen , a. carlsen , and a. bjarklev , `` photonic crystal fiber based evanescent - wave sensor for detection of biomolecules in aqueous solutions , '' opt .
* 29 * , 19741976 ( 2004 ) .
j. b. jensen , p. e. hoiby , g. emiliyanov , o. bang , l. h. pedersen , and a. bjarklev , `` selective detection of antibodies in microstructured polymer optical fibers , '' opt .
express * 13 * , 58835889 ( 2005 ) .
g. emiliyanov , j. b. jensen , o. bang , p. e. hoiby , l. h. pedersen , e. m. kjaer , and l. lindvold , `` localized biosensing with topas microstructured polymer optical fiber , '' opt . lett . * 32 * , 460462 ( 2007 ) . opt
. lett . * 32 * , 1059 ( 2007 ) . m. c. p. huy , g. laffont , y. frignac , v. dewynter - marty , p. ferdinand , p. roy , j .-
blondy , d. pagnoux , w. blanc , and b. dussardier , `` fibre bragg grating photowriting in microstructured optical fibres for refractive index measurement , '' meas .
* 17 * , 992997 ( 2006 ) .
b. r. acharya , t. krupenkin , s. ramachandran , z. wang , c. c. huang , and j. a. rogers , `` tunable optical fiber devices based on broadband long - period gratings and pumped microfluidics , '' * 83 * , 49124914 ( 2003 ) . y. park , t .- j .
ahn , y. h. kim , w .- t .
han , u .- c .
paek , and d. y. kim , `` measurement method for profiling the residual stress and the strain - optic coefficient of an optical fiber , '' appl . opt . *
41 * , 2126 ( 2002 ) . | we study the sensitivity of fiber grating sensors in the applications of strain , temperature , internal label - free biosensing , and internal refractive index sensing .
it is shown that optical dispersion plays a central role in determining the sensitivity , and the dispersion may enhance or suppress sensitivity as well as change the sign of the resonant wavelength shifts .
we propose a quality factor , @xmath0 , for characterizing lpgs . |
the muskat problem is a widely used model for the intrusion of water into oil sand .
a linear analysis was performed in @xcite where a relation , the so - called rayleigh - taylor condition , was found to determine two regimes for the problem : a stable regime , when a flat interface is stable under small deviations , and an unstable one , when fingering occurs .
nonetheless , existence and uniqueness of classical solutions has been firstly proven in @xcite by using newton s iteration method . in the last decade
the problem has received more interest and was studied by means of complex analysis @xcite , energy estimates @xcite , power series expansions @xcite , or abstract parabolic theory @xcite .
these different approaches cover a wide spectrum of questions related to the muskat problem : local well - posedness , global existence of solution , singular solutions , stability properties of equilibria .
it is worth noticing that all these papers mentioned above consider the situation when there is only one moving boundary , namely the one separating the fluids .
either one prescribes boundary conditions at two boundaries which are kept fixed during the flow or so - called far - field boundary condition are imposed .
this setting corresponds to an abstract equation with only one unknown - the interface between the fluids . in the present paper
we consider the more involved situation when there are two moving boundaries , one separating the two fluids and one separating the wetting phase from air ( assumed to be at uniform pressure equal to zero ) .
the fluids are located in a porous medium ( or a vertical hele - shaw cell ) and are assume to fill together with the dry phase ( air ) the entire void medium . moreover , we incorporate gravity and viscosity effects into the modeling as well as surface tension forces at both interfaces .
the invertibility of a bounded operator permits us to re - write the problem as an abstract non - autonomous evolution equation @xmath0 where the variable @xmath1 parametrises both unknown interfaces .
the temporal variable @xmath2 is induced into the problem by the boundary condition @xmath3 for the pressure on the bottom of the cell .
for this problem we find a generalised rayleigh - taylor condition in terms only of the boundary data @xmath3 , the viscosities @xmath4 , and densities @xmath5 of the fluids of the following form @xmath6 which determines the parabolic character of the problem in the absence of surface tension effects . when including surface tension forces at both interfaces we may drop condition .
we steadily use in this paper the subscript @xmath7 for the fluid on the bottom of the cell and @xmath8 for that above . after showing that the frchet derivative @xmath9 generates a strongly continuous and analytic semigroup , parabolic theory provides local well - posedness of the problem and the principle of linearised stability
may be applied to study the stability properties of the unique flat equilibrium which is determined for a fixed amount of fluid @xmath8 ( this quantity is preserved by the flow ) and a certain constant boundary data . when considering surface tension effects at the interface between the fluids and the more dense fluid lies above we re - discover the global bifurcation branches obtained in @xcite which consist only of finger - shaped equilibria of the muskat problem .
the exchange of stability theorem due to crandall and rabinowitz @xcite applies to this particular problem and we show that all small equilibria are unstable .
the outline of the paper is as follows : we describe in section 2 the mathematical model and present the main results .
section 3 is dedicated to the proof of the well - posedness result theorem [ t : p1 ] , and in the subsequent section we analyse the stability properties of the unique flat equilibrium as stated in theorem [ t : p2 ] . in section 5
we prove our third main result , theorem [ t : p3 ] .
the calculations leading to the representation of @xmath9 as a fourier multiplication operator are done in the appendix .
let us start this section by presenting the mathematical model of the setting described in the introduction .
given @xmath10 and @xmath11 the small hlder space @xmath12 stands for the closure of the smooth functions @xmath13 in @xmath14 we let @xmath15 denote the unit circle and functions on @xmath15 are identified with @xmath16-periodic functions on @xmath17 for later purposes we define @xmath18 as the subspace of @xmath12 consisting only of even functions , @xmath19 is the subspace of @xmath12 consisting only of functions with integral mean zero , and @xmath20 furthermore , we define the set of admissible functions to be @xmath21 each pair @xmath22 determines two open and simply connected subsets of the porous medium , seen as @xmath23 , as follows : @xmath24 & \omega(f , h):=\{(x , y)\,:\ , f(x)<y<1+h(x)\}.\end{aligned}\ ] ] let @xmath25 and @xmath26\to { \mathcal{u}}^2 $ ] be given such that , at each time @xmath27,$ ] the fluid @xmath7 is located at @xmath28 and the fluid @xmath8 at @xmath29 ( see figure [ f:1 ] ) .
the two fluids are assumed to be of newtonian type and incompressible , and both interfaces are supposed to move along with the fluids .
the problem is governed by the following system of partial differential equations : @xmath30 \delta u_-&=&0&\text{in } & \omega(f ) , \\[1ex ] { { \partial}_th+\frac{k\sqrt{1+h'^2}}{\mu_+}{\partial}_\nu u_+}&=&0 & \text{on } & \gamma(h),\\[1ex ] u_+&=&g\rho_+(1 + h)-\gamma_{d}\kappa_{{\gamma}(h)}&\text{on}&\gamma(h),\\[1ex ] u_-&=&b&\text{on}&\gamma_{-1},\\[1ex ] u_+-u_-&=&g(\rho_+-\rho_-)f+\gamma_{w}\kappa_{{\gamma}(f)}&\text{on } & \gamma(f),\\[1ex ] { { \partial}_tf+\frac{k\sqrt{1+f'^2}}{\mu_\pm}{\partial}_\nu u_\pm}&=&0 & \text{on } & \gamma(f),\\[1ex ] f(0)&=&f_0 , \\[1ex ] h(0)&=&h_0 \end{array } \right.\ ] ] for @xmath27 $ ] , where @xmath31 determines the initial domains occupied by the fluids .
we used the variable @xmath32 for parametrising the interface @xmath33 $ ] between the two fluids and @xmath34 $ ] separates the fluid @xmath8 from air .
the unit normal @xmath35 at @xmath36 [ resp .
@xmath37 is chosen such that , if @xmath38 is the tangent , the orthonormal basis @xmath39 has positive orientation .
we also write @xmath40 and @xmath41 for the curvature of @xmath42 and @xmath43 , respectively . moreover @xmath44 [ resp .
@xmath45 is the surface tension coefficient of the interface separating the fluids from air [ resp .
the fluids ] .
the potentials @xmath46 incorporate both pressure and gravity force @xmath47 , with @xmath48 the gravity constant .
the velocity fields @xmath49 , which satisfy darcy s law @xmath50 are presupposed to be equal on the boundary separating the fluid phases .
hereby , @xmath51 stands for the permeability of the porous medium . on the fixed boundary @xmath52
we prescribed the value of the velocity potential @xmath53 . for a precise deduction of we
refer to @xcite .
let @xmath54 be fixed for the following .
a pair @xmath55 is called _ classical hlder solution _ of if @xmath56,{\mathcal{v}})\cap c^{1}([0,t],(h^{1+\alpha}({\mathbb s}))^2),\\[1ex ] & \text{$u_+(t)\in\mbox{\it buc}^{2+\alpha}({\omega}(f(t),h(t)))$ and $ u_-(t)\in\mbox{\it buc}^{2+\alpha}({\omega}(f(t)))$ for $ t\in[0,t],$}\end{aligned}\ ] ] and if @xmath55 satisfies the equations of pointwise .
we defined @xmath57 to be the subset of @xmath58 given by @xmath59 where @xmath60 and @xmath61 for @xmath62 the space @xmath63 is defined as closure of the smooth functions with bounded and uniformly continuous derivatives @xmath64 in @xmath65 .
the space @xmath66 is defined similarly . moreover , since the potentials @xmath46 are determined , when knowing @xmath67 , as solutions of elliptic problems ( see section 3 ) we also refer to @xmath67 to be the solution of .
the first main result of this paper states : [ t : p1 ] let @xmath68 be given .
there exist an open neighbourhood @xmath69 of the zero function in @xmath70 such that for all @xmath71 and @xmath72 problem possesses a unique classical hlder solution @xmath73 defined on a maximal time interval @xmath74 and which satisfies @xmath75 the mapping @xmath76 @xmath77 has the same regularity as @xmath3 has . [ r:1 ] the conclusion of theorem [ t : p1 ] remains valid if @xmath78 or @xmath79 with the following modifications : if @xmath80 we have to replace @xmath70 by @xmath81 and require that @xmath82 where @xmath83 is a small neighbourhood of the zero function in @xmath84 and @xmath85 satisfies @xmath86 \label{eq : cond2 } & \frac{\mu_+-\mu_-}{\mu_++\mu_-}(c - g\rho_+)+g(\rho_+-\rho_-)<0.\end{aligned}\ ] ] when @xmath87 [ resp .
@xmath88 we replace @xmath70 by @xmath89 [ resp .
@xmath90 and request that the constant @xmath91 satisfies only equation [ resp .
eq . ] .
relation is a generalisation of the positive pressure condition imposed in @xcite to ensure well - posedness and stability of the one - phase hele - shaw problem without surface tension . indeed ,
if the fluids have the same densities and viscosity , re - writes @xmath92 which is , up to a scaling , the same condition as in @xcite .
moreover , it turns out that the muskat problem without surface tension effects studied in @xcite is similar to our problem if @xmath93 indeed , we have : [ l:1 ] the volume of fluid @xmath8 is preserved by the solutions of .
the proof is similar to that of ( * ? ? ?
* lemma 3.1 ) . in order to establish similarity between our problem when @xmath94 and that in @xcite , we determine a special solution of in the case when the volume of fluid @xmath8 is equal to @xmath16 , i.e. @xmath95 if initially @xmath96 and @xmath3 depends only on time , then @xmath97 and , by lemma [ l : kl ] , @xmath98 as long as the solution exists . if @xmath99 and @xmath100 we obtain from that if @xmath101 , then @xmath102 , thus @xmath103 is positive if @xmath104 is close to zero , meaning that the more viscous fluid drives upwards the less viscous one in the medium .
this condition has been found also in @xcite to guarantee well - posedness of the muskat problem studied therein . moreover , if the atwood number @xmath105 is zero , then tells us that the more dense fluid must lay beneath in order to guarantee well - posedness of when @xmath106 result similar to that in @xcite .
corresponding to the result in @xcite , where an optimal value for the normal velocity at which water may replace oil in the absence of surface tension effects was found , we obtain herein an optimal value for the pressure on the bottom of the medium : if the fluid below is water and that above oil , and we neglect the surface force at the interface between them , we find from an optimal value @xmath107 for the pressure on the bottom of the porous medium below which water may drive upwards oil in a stable regime ( no fingering occurs ) .
relation is obtained form in view of @xmath108 on @xmath109 the optimal value for the potential @xmath3 is @xmath110 and if the boundary value @xmath3 is close to this value we find that the solutions of fulfill @xmath111 thus water drives oil upwards .
this last assertion follows from @xmath112 since it is well - known @xcite that @xmath113 and @xmath114 ( oil is less dense and more viscous than water ) .
we infer from that if @xmath115 and @xmath116 , then @xmath117 for all @xmath118 concerning the stability properties of the stationary solution @xmath119 , which is the unique flat stationary solution of for @xmath120 and which satisfies , we state : [ t : p2 ] let @xmath121 .
then : * if @xmath122 then the flat equilibrium @xmath119 of is exponentially stable .
more precisely , there exists positive constants @xmath123 and @xmath124 such that if @xmath125 and @xmath126 satisfies , then the solution @xmath67 of exists globally and @xmath127 & \phantom{doublespace}\leq me^{-\omega t}\|(f_0,h_0)\|_{h^{2 + 2{\mathop{\rm sign}\nolimits}(\gamma_w)+\alpha}({\mathbb s})\times h^{2 + 2{\mathop{\rm sign}\nolimits}(\gamma_d)+\alpha}({\mathbb s})},\quad\forall t\geq0.\end{aligned}\ ] ] * if @xmath128 then @xmath119 is unstable .
[ r:2 ] when we study the stability of equilibria in theorem [ t : p2 ] and theorem [ t : p3 ] below we fixed a volume of fluid @xmath8 equal to @xmath129 meaning that the initial data of are presupposed to satisfy .
this setting is imposed by lemma [ l:1 ] , since the volume of fluid @xmath8 is preserved by the solutions of .
theorem [ t : p2 ] is related to the exponential stability result established in ( * ? ? ?
* theorem 5.3 ) for the muskat problem with only one free boundary and is stronger than that in @xcite , where only stability is shown .
notice that if @xmath106 then the flat solution is always stable , since @xmath130 is exactly the condition which guarantees well - posedness of . concerning the unstable case , numerical experiments @xcite
show that the interface between the fluids becomes very ramified , and dendrite like structures occur as time evolves if @xmath131 if @xmath120 and the volume of fluid @xmath8 is equal to @xmath16 , there exist also other stationary solutions of .
they appear only in the unstable regime or sufficiently close to it , that is when @xmath132 and the more dense fluid lies above in the cell .
we show that for certain small @xmath132 there exist finger - shaped stationary solutions of , and therefore we shall refer also to @xmath133 to be solution of .
given @xmath134 we define @xmath135 [ t : p3 ] let @xmath120 and @xmath136 .
if @xmath133 is a stationary solution of satisfying , then @xmath137 and @xmath138 is a solution of the laplace - young equation @xmath139 the solution of are , up to a translation , even and all even solutions of can be represented as a disjoint union @xmath140 with continuous functions @xmath141 which , near @xmath142 , are real analytic and satisfy : @xmath143 while @xmath144 is even and @xmath145 either @xmath146 or @xmath147 additionally , the equilibrium @xmath148 of problem is unstable if @xmath149 is small .
when @xmath150 we have to assume @xmath151 too . here
@xmath152 stands for euler s beta function .
notice that the stationary solutions of , which satisfy ( see in figure [ f:2 ] ) , are the same with the stationary solutions of the muskat problem studied in @xcite , where just one moving boundary is considered ( @xmath153 is chosen a priori to be zero ) . for a precise description of the global bifurcation branches
@xmath154 we refer to @xcite .
it is shown there that the situation @xmath146 may occur only for small integers @xmath155 .
in order to solve problem we re - write it as an abstract evolution equation on the unit circle . to do that we first transform system into a system of equations on fixed domains by using the unknown functions
@xmath156 let @xmath157 and @xmath158 given @xmath159 we define the mappings @xmath160 by @xmath161 respectively @xmath162 @xmath163 one can easily check that @xmath164 and @xmath165 are diffeomorphisms for all @xmath166 these diffeomorphisms induce pull - back and push - forward operators ( see e.g. @xcite ) which we use to transform the differential operators involved in system into operators on the domains @xmath167 and their boundaries , respectively .
each pair @xmath159 induces linear elliptic operators @xmath168 & { \mathcal{a}}(f , h):\mbox{\it{buc}}\,^{2+\alpha}(\omega_+)\to \mbox{\it{buc}}\,^{\alpha}(\omega_+ ) , & { \mathcal{a}}(f , h)v:=\delta(v_+\circ\phi_{f , h}^{-1})\circ\phi_{f , h},\end{aligned}\ ] ] which depend , as bounded operators , analytically on @xmath32 and @xmath153 .
denote by @xmath169 the trace operator with respect to @xmath170 .
we associate problem the following trace operators on @xmath171 : @xmath172 { \mathcal{b}}(f , h)v_+&:=k\mu_+^{-1}{\mathop{\rm tr}\nolimits}_0 ( \langle \nabla(v_+\circ\phi_{f , h}^{-1})|(-f',1)\rangle\circ\phi_{f , h}),\quad v_+\in \mbox{\it{buc}}\,^{2+\alpha}(\omega_+),\end{aligned}\ ] ] which , seen as bounded operators into @xmath173 , depend analytically on @xmath32 and @xmath153 as well .
lastly , we define a boundary operator on @xmath174 given @xmath175 we let @xmath176 whereby @xmath177 is the trace operator with respect to @xmath178
. with this notation one can easily verify that if @xmath179 is a solution of , then @xmath180 solves the following system of equations : @xmath181 { \mathcal{a}}(f ) v_-&=&0&\text{in } & \omega_-,\\[1ex ] { \partial}_t h&=&{\mathcal{b}}_1(f , h)v_+&\text{in } & \gamma_1,\\[1ex ] v_+&=&g\rho_+(1+h)-\gamma_{d}\kappa(h)&\text{on } & \gamma_1,\\[1ex ] v_-&=&b&\text{on}&\gamma_{-1},\\[1ex ] v_+-v_-&=&g(\rho_+-\rho_- ) f+\gamma_{w } \kappa(f)&\text{on } & \gamma_0,\\[1ex ] { \partial}_t f+{\mathcal{b}}(f)v_-&=&0 & \text{on } & \gamma_{0},\\[1ex ] { \partial}_t f+{\mathcal{b}}(f , h)v_+&=&0 & \text{on } & \gamma_{0},\\[1ex ] f(0)&=&f_0,&&\\[1ex ] h(0)&=&h_0 & & \end{array } \right.\ ] ] for all @xmath27 $ ] , where the transformed curvature operator @xmath182 is defined by @xmath183 the notion of solution of is defined analogously to that of .
notice that the parametrisation @xmath67 is left invariant by the transformation above .
in fact , one can see , cf .
* lemma 1.2 ) that each solution of corresponds to a unique solution of .
we introduce now solution operators corresponding to the system . given @xmath184 and @xmath185 , we let @xmath186 denote the solution of the linear , elliptic mixed boundary value problem @xmath187 { \mathcal{b}}(f)v_-&=&q & \text{on } & \gamma_{0},\\[1ex ] v_-&=&p&\text{on } & \gamma_{-1}. \end{array } \right.\ ] ] further on , we define @xmath188 by writing @xmath189 for the unique solution of the problem @xmath190 v_+&=&p&\text{on } & \gamma_1,\\[1ex ] v_+&=&r&\text{on}&\gamma_{0}. \end{array } \right.\ ] ] it is convenient to write @xmath191 where @xmath192 respectively @xmath193 with @xmath194 { \mathcal{s}}_2(f , h)r:&=({\mathcal{a}}(f , h),{\mathop{\rm tr}\nolimits},{\mathop{\rm tr}\nolimits})^{-1}(0,0,r).\end{aligned}\ ] ] the operators @xmath195 and @xmath196 @xmath197 are bounded linear operators and they depend , in the norm topology , analytically on @xmath32 and @xmath153 too .
the key point of our analysis is the following observation .
if @xmath198 is a classical solution of for to the initial data @xmath126 , then it must hold : * @xmath199 and @xmath200 * @xmath201 * @xmath202 * @xmath203 * @xmath204 let us now show that from @xmath205 we can determine the derivative @xmath206 as a function of @xmath32 , @xmath153 , and @xmath2 only .
indeed , we plug @xmath207 into @xmath208 and @xmath208 into @xmath209 to obtain the equation @xmath210\\[1ex ] & \phantom{space=}+{\mathcal{b}}(f , h){\mathcal{s}}_2(f , h)[{\mathop{\rm tr}\nolimits}_0{\mathcal{t}}(f,-{\partial}_t f , b)+g(\rho_+-\rho_- ) f+\gamma_{w}\kappa(f)]=0,\end{aligned}\ ] ] which can be writen equivalently @xmath211 & \phantom{space=}+{\mathcal{b}}(f , h){\mathcal{s}}(f , h , g\rho_+(1+h)-\gamma_{d}\kappa(h),g(\rho_+-\rho_- ) f+\gamma_{w}\kappa(f ) ) .
\end{aligned}\ ] ] the linear operator which is evaluated at @xmath206 is invertible , so that we obtain , by applying its inverse to , an equation expressing the derivative @xmath206 in dependence of @xmath212 and @xmath2 .
indeed , we have : [ l : kl ] the set @xmath213 contains an open neighbourhood @xmath214 of @xmath215 with the property that @xmath216 is an isomorphism for all @xmath217 the proof is based on a continuity argument .
namely , all the operators defined in this section depend analytically on their variables and then so does @xmath218 too .
thus , it suffices to show that @xmath219 is an isomorphism . to do that
, we represent @xmath219 as a fourier multiplication operator .
given @xmath220 we let @xmath221 denote its fourier series expansion .
a fourier series ansatz yields for @xmath222 the following expression @xmath223 for @xmath224 respectively , if @xmath225 then @xmath226 may be expanded as follows @xmath227 for @xmath228 combining these two relations and taking the normal derivative yields that @xmath229 thus @xmath219 is an isomorphism . in virtue of lemma
[ l : kl ] , if the pair @xmath67 maps into @xmath230 we may apply the inverse of @xmath231 to , and get @xmath232 with a nonlinear and nonlocal operator @xmath233 defined by the relation @xmath234 & + { \mathcal{b}}(f , h){\mathcal{s}}(f , h , g\rho_+(1+h)-\gamma_{d}\kappa(h),g(\rho_+-\rho_- ) f+\gamma_{w}\kappa(f ) ) .
\end{aligned}\ ] ] furthermore , from @xmath235 and we obtain that @xmath153 is solution of the equation @xmath236 where the operator @xmath237 is given by @xmath238 & + b_1(f , h){\mathcal{s}}_2(f , h){\mathop{\rm tr}\nolimits}_0{\mathcal{t}}(f,-\phi_1(t , f , h ) , b ) .
\end{aligned}\ ] ] by lemma [ l : kl ] and relations - we found that all the solutions @xmath67 of which are contained in @xmath214 solve the following abstract evolution equation @xmath239 where @xmath240 and we introduced the new variable @xmath241 .
concerning the operator @xmath242 we state : [ t : phi - o ] the operator @xmath242 has the same regularity as @xmath3 , it is analytic in the variable @xmath1 , and if @xmath243 then @xmath244 and @xmath245 @xmath197 are fourier multipliers with symbols @xmath246 and @xmath247 @xmath197 respectively , given by @xmath248\frac{k|m|}{(\mu_++\mu_-)\tanh(|m|)},\\[1ex ] \label{eq : coef_2_f } & \lambda^f_2(m):=\left[\frac{c(\mu_+-\mu_-)+2g\rho_+\mu_-}{\mu_++\mu_-}-g\rho_--\gamma_wm^2\right]\frac{k|m|}{(\mu_++\mu_-)\sinh(|m|)},\\[1ex ] \label{eq : coef_1_h } & \lambda^h_1(m):=-\left[\frac{g\rho_+\mu_-+c\mu_+}{\mu_-+\mu_+}+\gamma_{d}m^2\right]\frac{k|m|}{(\mu_-+\mu_+)\sinh(|m|)},\end{aligned}\ ] ] @xmath249\frac{k|m|}{\mu_+\tanh(|m|)}-\frac{\mu_-}{\mu_+}\frac{\lambda^h_1(m)}{\cosh(m)}.\end{aligned}\ ] ] the regularity assertion is obvious .
that the first order partial derivatives of @xmath250 with respect to @xmath32 and @xmath153 are fourier multipliers follows from - , relations proven in the appendix .
we give now a proof of our first main result .
we verify that the assumptions of ( * ? ? ?
* theorem 8.4.1 ) are fulfilled by @xmath251 theorem [ t : p1 ] is then a consequence of this result . for continuity reasons it suffices in fact to show only that the derivative @xmath9 generates a strongly continuous and analytic semigroup in @xmath252 i.e. @xmath253 for some @xmath254 by using the interpolation properties of the small h " older spaces @xmath255
if @xmath256 @xmath257 we find then all assumptions of ( * ? ? ?
* theorem 8.4.1 ) to be fulfilled . here
@xmath258 denotes the interpolation functor introduced by da prato and grisvard @xcite .
let us first notice that derivative @xmath259 maps @xmath260 continuously into @xmath261 for some @xmath262 .
this property can be verified easily by using ( * ? ? ?
* theorem 3.4 ) , which is a multiplier theorem based on some generalized marcinkiewicz conditions . since by @xmath263 we deduce , in virtue of theorem 1.6.1 and relation ( 2.2.2 ) in @xcite , that @xmath9 generates a strongly continuous and analytic semigroup exactly when @xmath264 and @xmath265 are generators , i.e. @xmath266 & -{\partial}_h\phi_2(0)\in\mathcal{h } ( h^{2 + 2{\mathop{\rm sign}\nolimits}(\gamma_d)+\beta}({\mathbb s } ) , h^{1+\beta}({\mathbb s})).\end{aligned}\ ] ] when considering surface tension effects this property holds independently of the boundary conditions , and when @xmath101 and @xmath267 this is true if @xmath268 respectively , with @xmath269 sufficiently close to @xmath85 in @xmath270 .
we refer to @xcite where the generator property of a fourier multiplier between space of periodic and continuous functions is explicitly verified when knowing its symbol . in virtue of the proof
is completed .
as we mentioned earlier , if we consider a fixed volume of fluid @xmath8 equal to @xmath16 and if @xmath120 , then @xmath119 is the unique flat stationary solution of problem .
moreover , the reduced equation is autonomous since @xmath242 does not depend on time for constant @xmath3 . in order to study the stability properties of this equilibrium ,
as stated in theorem [ t : p2 ] , we shall use the principle of linearised stability , and need therefore to determine the spectrum of the derivative @xmath271 being a generator and taking into consideration that the small hlder space @xmath272 is compactly embedded into @xmath173 , we obtain from ( * ? ? ?
* theorem iii.8.29 ) that its spectrum consists entirely of isolated eigenvalues with finite multiplicity . in virtue of theorems 9.1.2 and 9.1.3 in @xcite
we know that the trivial solution @xmath119 is exponentially stable if the spectrum of @xmath273 is bounded away from the imaginary axis in the left half complex plane , and unstable if the infimum of the real part of all eigenvalues in the right half plane is positive
. one can easily see that if @xmath274 is an eigenvalue of @xmath275 then it must be , for some @xmath276 eigenvalue of the matrix @xmath277 \lambda_2^f(m)&\lambda_2^h(m ) \end{array } \right],\ ] ] where , for @xmath120 , we obtained the simpler expressions for the multiplier symbols : @xmath278\frac{km}{(\mu_++\mu_-)\tanh(m)},\\[1ex ] & \lambda^f_2(m)=-\left[g(\rho_--\rho_+)+\gamma_wm^2\right]\frac{km}{(\mu_++\mu_-)\sinh(m)},\\[1ex ] & \lambda^h_1(m)=-\left[g\rho_++\gamma_{d}m^2\right]\frac{km}{(\mu_-+\mu_+)\sinh(m)},\\[1ex ] & \lambda^h_2(m)=-\left[g\rho_++\gamma_dm^2\right]\frac{km}{\mu_++\mu_-}\frac{\mu_-(\cosh^2(m)-1)+\mu_+\cosh^2(m)}{\mu_+\sinh(m)\cosh(m)}. \end{aligned}\ ] ] thus , the spectrum of @xmath273 consists only of the eigenvalues @xmath279 whereby @xmath280 easily , we see that @xmath281 thus we find our selves in the critical case of stability when @xmath215 is an eigenvalue , which makes it difficult for us to establish the stability properties of the flat solutions .
this is due to the fact that the volume of fluid @xmath8 is preserved by the flow , and this property has not been included yet into our equations .
we do this by introducing a new variable @xmath282 .
then @xmath283 { \partial}_t f&=\phi_1(f , f-{\widetilde}f)=:\psi_1(f,{\widetilde}f ) , \end{aligned}\ ] ] and with this new variable , problem is equivalent to @xmath284 where @xmath285 and @xmath286 .
the trivial solution of corresponds to the solution @xmath287 of , so that we shall study the stability properties of the trivial solution of which , as we shall see , it is more convenient .
this since : [ l : le1 ] it holds that @xmath288 for all @xmath289 in a zero neighbourhood @xmath290 . given @xmath291 as above , let again @xmath292 . setting @xmath293 and @xmath294 , we infer from that @xmath295 which can be reformulated as follows @xmath296 using this relation and , we obtain @xmath297 & + \int_{{\mathbb s}}{\mathcal{b}}_1(f , h){\mathcal{s}}(f , h,0,{\mathop{\rm tr}\nolimits}_0{\mathcal{t}}_1(f)\phi_1(f , h))\ , dx\\[1ex ] & -\int_{{\mathbb s}}{\mathcal{b}}(f , h){\mathcal{s}}(f , h , p , r ) dx-\int_{{\mathbb s}}{\mathcal{b}}_1(f , h){\mathcal{s}}(f , h , p , r)\ , dx.\end{aligned}\ ] ] therefore , in order to prove our claim , it will do if we show that @xmath298 for all @xmath299 and arbitrary @xmath300 defining the harmonic function @xmath301 , we then have @xmath302 & \phantom{space=}-\frac{k}{\mu_+}\int_{{\gamma}(f , h)}{\partial}_n u_+\ , ds=\frac{k}{\mu_+}\int_{{\omega}(f , h)}\delta u_+\ , d(x , y)=0,\end{aligned}\ ] ] where @xmath303 stands for the outward unit normal at @xmath304 i.e. @xmath305 on @xmath306 and @xmath307 on @xmath308 we come now to the proof of our second main result : studying the stability properties of the trivial solution @xmath119 of , under the constrain , is equivalent to study the stability of the trivial solution @xmath287 of , where , by lemma [ l : le1 ] , we have that @xmath309 from we know that every component of the matrix operator @xmath310 { \partial}_f\psi_2(0)&{\partial}_{{\widetilde}f}\psi_2(0 ) \end{array } \right]\ ] ] is a multiplier with symbol @xmath311 , @xmath312 @xmath313 , and @xmath314 respectively , given by @xmath315 & { \widetilde}\lambda^f_2(m)=\lambda_1^f(m)+\lambda_1^h(m)-\lambda_2^f(m)-\lambda_2^h(m ) , & { \widetilde}\lambda^{{\widetilde}f}_2(m)=-\lambda_1^h(m)+\lambda_2^h(m)\end{aligned}\ ] ] for @xmath316 , @xmath317 , and @xmath318 .
a simple computation shows that the spectrum of the frchet derivative @xmath319 , which coincides with its point spectrum from the same reason ( * ? ? ?
* theorem iii.8.29 ) , is given by @xmath320 let us start and estimate the eigenvalues @xmath321 with @xmath280 since @xmath322 $ ] is increasing on @xmath323 we conclude that @xmath324 let now @xmath325 we show that in this case also the other eigenvalues @xmath326 are of negative sign .
indeed , it holds that @xmath327 and , in view of @xmath328}{\mu_+(\mu_++\mu_-)},\ ] ] we conclude that the spectrum of @xmath319 is bounded away from the negative half axis in @xmath329 the assertion stated in theorem [ t : p2 ] @xmath330 follows at once , with the remark that the constant @xmath124 found there may be chosen arbitrarily in the set @xmath331 on the other hand , if @xmath128 one can easily observe that @xmath332 moreover , since @xmath333 as @xmath334 , we conclude theorem [ t : p2 ] @xmath335 this finishes the proof .
this last section is dedicated entirely to the proof of theorem [ t : p3 ] .
if the tupel @xmath336 is a stationary solution of , it must hold that @xmath337 @xmath338 is constant , and @xmath339 \label{eq : fi2 } & \gamma_d \kappa(h)-g\rho_+h = b - u_+.\end{aligned}\ ] ] on @xmath340 equations and , called laplace - young or capillarity equations , have been studied intensively ( see @xcite and the literature therein ) subjected to certain constrains at a fixed rigid boundary .
though , when dealing with periodic solutions , we easily get , cf .
@xcite , that @xmath153 must be constant also in the spatial variable , and if @xmath341 then also @xmath32 is constant .
whence , equations and may have nontrivial solutions @xmath342 only when @xmath132 and @xmath343 .
we are interested to determine only the steady - states @xmath67 of when @xmath120 and which satisfy , i.e. @xmath344 encloses the same volume of fluid as @xmath345 assume by contradiction that @xmath346 for some @xmath347 since @xmath348 we get that @xmath349 on one hand , if @xmath32 is constant , it must hold that @xmath350 which contradicts for @xmath347 on the other hand , the function @xmath351 solves the equation @xmath352 the solutions of this equation are , up to a translation , odd . indeed ,
since @xmath353 is periodic and nonconstant , it must hold @xmath354 for some @xmath355 by translation , we may take @xmath356 the pair @xmath357 is a global solution of the initial value problem @xmath358 q \end{array } \right ) ' = \left ( \begin{array}{c } q\\[1ex ] -\displaystyle\frac{g(\rho_+-\rho_-)}{\gamma_w } p(1+q^2)^{3/2 } \end{array } \right),\quad \left ( \begin{array}{c } p\\ q \end{array } \right)(0)= \left ( \begin{array}{c } 0\\ p'(0 ) \end{array } \right).\ ] ] this is also true for @xmath359 , @xmath360 whence , as we claimed , @xmath353 is odd , so that @xmath361 , which contradicts again @xmath362 and
. consequently , if @xmath363 is a solution of and , then @xmath137 and @xmath138 solves the problem , which implies in turn that @xmath32 has integral mean equal to @xmath215 and an even translation by ( * ? ? ?
* theorem 3.2 ) . combining that particular result with theorem 6.1 in @xcite
we obtain all the claims of theorem [ t : p3 ] excepting the stability assertion .
in the remaining part of this section we prove that the steady - state solution @xmath364 of problem is unstable provided that @xmath365 is sufficiently small .
we rediscover first the global branches @xmath366 @xmath367 at least locally near @xmath368 by applying the theorem on bifurcations from simple eigenvalues , due to crandall and rabinowitz ( * ? ? ?
* theorem 1.7 ) , to the problem @xmath369 where @xmath370 is the mapping defined by .
we shall refer to @xmath371 , @xmath372 as being a trivial solution of . in order to use @xmath373 as a bifurcation argument
we establish first analytic dependence of @xmath370 on @xmath373 and take the restriction @xmath374 where @xmath375 that @xmath370 is well - defined between these spaces follows by using elliptic maximum principles and lemma [ l : le1 ] .
when @xmath376 , we infer from - that all the eigenvalues of @xmath377 are different from zero , thus @xmath377 is an isomorphism .
the implicit function theorem ensures that @xmath378 is not a bifurcation point of the trivial solution .
otherwise , if @xmath379 for some @xmath380 then @xmath381 whereby @xmath382 is the eigenvector corresponding to the eigenvalue @xmath383 we considered that @xmath384 i.e. @xmath385 depends not only on @xmath386 , but also on @xmath387 also , the codimension of the image @xmath388 is one since @xmath389 belongs to @xmath388 if and only if @xmath390 for some @xmath391 moreover , one can easily check that the mixed derivative @xmath392 whereby @xmath393 does not belong to the image of @xmath394 .
we conclude by ( * ? ? ?
* theorem 1.7 ) the existence of a bifurcation curve @xmath395 consisting only of stationary solutions of .
since they correspond all to a volume of fluid @xmath8 equal to @xmath129 it follows that @xmath144 and @xmath396 are , up to a parametrisation , restrictions of the functions obtained in theorem [ t : p3 ] .
the stability properties of the equilibrium @xmath397 for under the constrain , are equivalent with that of the steady - state solution @xmath398 of problem . for our purposes ,
theorem [ t : p3 ] , it suffices in fact to show that @xmath398 is an unstable stationary solution of the abstract cauchy problem @xmath399 where @xmath370 is the restriction . indeed , if @xmath365 small and @xmath400 , then @xmath398 is an unstable solution of since the eigenvalue @xmath401 of @xmath402 is positive . for the stability of the stationary solution @xmath403 ,
when @xmath404 is small and @xmath405 , it is important how the eigenvalue @xmath406 ( which is equal to @xmath215 ) perturbs for small @xmath407 our main tool is the exchange of stability theorem ( * ? ? ?
* theorem 1.16 ) due to crandall and rabinowitz .
the assumptions of this theorem are satisfied by @xmath370 since : * @xmath408 is a fredholm operator of index 0 with a one - dimensional kernel ; * @xmath409 ; * @xmath410 letting @xmath411 denote the inclusion @xmath412 in the terminology of @xcite , @xmath413 , @xmath414 , and @xmath415 mean that @xmath215 is a @xmath416-simple eigenvalue and a @xmath411-simple eigenvalue of @xmath408 . by choosing @xmath417 sufficiently small , we obtain from ( * ? ? ?
* theorem 1.16 ) four continuously differentiable functions @xmath418 , @xmath419 , @xmath420 and @xmath421 such that : @xmath422 & \text{${\partial}_{(f,{\widetilde}f)}\psi({\overline}\gamma_1({\varepsilon}),f_1({\varepsilon}),{\widetilde}f_1({\varepsilon}))w({\varepsilon})=\mu({\varepsilon})w({\varepsilon})$ for $ { \varepsilon}\in(-\delta_1,\delta_1)$},\\[1ex ] & \text{$\lambda({\overline}\gamma_1)=\mu(0)=0 $ , and $ u({\overline}\gamma_1)=w(0)=(\cos(x),\cos(x)).$}\end{aligned}\ ] ] moreover , @xmath423 and @xmath424 since @xmath425 is an eigenvalue of @xmath377 and @xmath426 we get , by continuity , that @xmath427 for all @xmath428 moreover , @xmath429 is positive for @xmath430 , and negative if @xmath431 thus @xmath432 in order to determine the sign of the eigenvalue @xmath433 which is the perturbation of the eigenvalue @xmath215 of @xmath408 , we need to specify the sign of @xmath434 . from theorem [ t :
p3 ] we obtain in view of @xmath435 and @xmath436 , that @xmath434 and @xmath365 have the same sign , thus @xmath437 is a positive eigenvalue , and we are done by ( * ? ? ?
* theorem 9.1.3 ) .
we end this paper with a detailed proof of theorem [ t : phi - o ] .
since the diffeomorphisms used in section 3 to transform the original problem into are given explicitly in terms of @xmath32 and @xmath153 , we obtain by direct computation the following expressions for the elliptic and trace operators defined therein : @xmath438 & -(1+y)\frac{(1+f)f''- 2f'^2}{(1+f)^2}\frac{{\partial}}{{\partial}y},\\[1ex ] { \mathcal{a}}(f , h)=&\frac{{\partial}^2 } { { \partial}x^2}-2\frac{yh'+(1-y)f'}{1 + h - f}\frac{{\partial}^2 } { { \partial}x{\partial}y}+\frac{(yh'+(1-y)f')^2 + 1}{(1+h - f)^2}\frac{{\partial}^2 } { { \partial}y^2}\\[1ex ] & -\left(\frac{yh''+(1-y)f''}{1+h - f}-2\frac{(h'-f')(yh'+(1-y)f')}{(1+h - f)^2}\right)\frac{{\partial}}{{\partial}y},\\[1ex ] { \mathcal{b}}(f)=&\frac{k}{\mu_-}\left(\frac{1+f'^2}{1 + f}{\mathop{\rm tr}\nolimits}_0 \frac{{\partial}}{{\partial}y}-f'{\mathop{\rm tr}\nolimits}_0 \frac{{\partial}}{{\partial}x}\right),\\[1ex ] { \mathcal{b}}(f , h)=&\frac{k}{\mu_+}\left(\frac{1+f'^2}{1+h- f}{\mathop{\rm tr}\nolimits}_0 \frac{{\partial}}{{\partial}y}-f'{\mathop{\rm tr}\nolimits}_0 \frac{{\partial}}{{\partial}x}\right),\\[1ex ] { \mathcal{b}}_1(f , h)=&-\frac{k}{\mu_+}\left(\frac{1+h'^2}{1+h- f}{\mathop{\rm tr}\nolimits}_1 \frac{{\partial}}{{\partial}y}-h'{\mathop{\rm tr}\nolimits}_1 \frac{{\partial}}{{\partial}x}\right).\end{aligned}\ ] ] in the following we take @xmath439 and show that @xmath244 and @xmath440 @xmath197 are fourier multiplication operators .
we shall also determine the symbol of these operators which are the main ingredients when proving the theorems [ t : p1 ] and [ t : p2 ] .
let us start by determining the symbol of the operator @xmath264 .
putting @xmath137 in yields that @xmath441 & -{\mathcal{g}}^{-1}(f,0){\mathcal{b}}(f,0){\mathcal{s}}_2(f,0)[g(\rho_+-\rho_- ) f+\gamma_{w}\kappa(f)]\\[1ex ] & -{\mathcal{g}}^{-1}(f,0){\mathcal{b}}(f,0){\mathcal{s}}_1(f,0)g\rho_+.\end{aligned}\ ] ] we show first that the derivative of @xmath442 with respect to @xmath32 in @xmath215 is a fourier multiplier . indeed , by the chain rule we have , due to @xmath443 that @xmath444m=-{\mathcal{g}}^{-1}(0,0){\partial}_f{\mathcal{g}}(0,0)[f]{\mathcal{g}}^{-1}(0,0)m$ ] for all @xmath445 and @xmath446 .
so , we need to determine @xmath447m=&-{\partial}_f{\mathcal{b}}(0,0)[f]{\mathcal{s}}_2(0,0){\mathop{\rm tr}\nolimits}_0{\mathcal{t}}_1(0)m\\[1ex ] & -{\mathcal{b}}(0,0){\partial}_f{\mathcal{s}}_2(0,0)[f]{\mathop{\rm tr}\nolimits}_0{\mathcal{t}}_1(0)m\\[1ex ] & -{\mathcal{b}}(0,0){\mathcal{s}}_2(0,0){\mathop{\rm tr}\nolimits}_0{\partial}{\mathcal{t}}_1(0)[f]m . \end{aligned}\ ] ] using the expansions found in the proof of lemma [ l : kl ] , we find that @xmath448 and @xmath449 hence @xmath450m=\frac{m\mu_-}{\mu_+}f.\ ] ] concerning the first two terms of we determine first an expansion for @xmath451m$ ] and @xmath452m.$ ] we start with @xmath452m.$ ] elliptic estimates yield that the function @xmath452m$ ] is the solution of dirichlet problem @xmath453{\mathcal{s}}_2(0,0)m =- m(1-y)f''&\text{in } & \omega_+,\\[1ex ] w&=&0&\text{on } & \gamma_1,\\[1ex ] w&=&0&\text{on}&\gamma_{0 } , \end{array } \right.\ ] ] and , as in the proof of lemma [ l : kl ] , we get that @xmath454m=-\sum_{m\in{\mathbb{z}}\setminus\{0\}}m \left(\frac{e^{my}-e^{2m}e^{-my}}{e^{2m}-1}+(1-y)\right){\widehat}f(m)e^{imx}.\ ] ] whence @xmath455{\mathop{\rm tr}\nolimits}_0{\mathcal{t}}_1(0)m=\frac{m\mu_-}{\mu_+}\sum_{m\in{\mathbb{z}}}\frac{m}{\tanh(m)}{\widehat}f(m ) e^{imx}-\frac{m\mu_-}{\mu_+}f.\ ] ] consider now the function @xmath451m$ ] , which is the solution of @xmath456{\mathcal{t}}_1(0)m=\mu_-k^{-1}m(1+y)f''&\text{in } & \omega_-,\\[1ex ] { \partial}_yw&=&-\mu_-k^{-1}{\partial}{\mathcal{b}}(0)[f]{\mathcal{t}}_1(0)m=\mu_-k^{-1}mf&\text{on } & \gamma_0,\\[1ex ] w&=&0&\text{on}&\gamma_{-1}. \end{array } \right.\ ] ] expanding @xmath457 we find that @xmath458 is the solution of the following problem @xmath459 w_m'(0)&=&m\mu_-k^{-1 } { \widehat}f(m),\\[1ex ] w_m(-1)&=&0 , \end{array } \right.\ ] ] which has the solution @xmath460 for all @xmath461 .
whence @xmath462m = m\mu_-k^{-1}(1+y)f,\ ] ] and , with the convention that @xmath463 if @xmath464 , we determine that @xmath465m=\frac{m\mu_-}{\mu_+}\sum_{m\in{\mathbb{z}}}\frac{m}{\tanh ( m)}{\widehat}f(m)e^{imx}.\ ] ] the relations - fuse , in view of lemma [ l : kl ] , to @xmath466 for all @xmath467 and @xmath468 in order to determine the derivative @xmath469 two more steps must be done : we must find the expansions of the derivatives @xmath470 and @xmath471 since @xmath472 , we get that @xmath473m=0,$ ] and proceeding similarly as we did before , yields @xmath474m=\sum_{m\in{\mathbb{z}}\setminus\{0\}}m \left(\frac{e^{my}-e^{2m}e^{-my}}{e^{2m}-1}+(1-y)\right){\widehat}f(m)e^{imx}.\ ] ] combining all these relations , we finally obtain for @xmath475 that the expansions found in the previous subsection are very useful when studying @xmath479 , since by we have @xmath480 & + b_1(f,0){\mathcal{s}}_2(f,0){\mathop{\rm tr}\nolimits}_0{\mathcal{t}}_2(f)c\\[1ex ] & + b_1(f,0){\mathcal{s}}_2(f,0)[\gamma_{w}\kappa(f)+g(\rho_+-\rho_- ) f ] .
\end{aligned}\ ] ] in view of @xmath481 we obtain the following representation @xmath482=\sum_{m\in{\mathbb{z}}}\lambda^f_2(m){\widehat}f(m)e^{imx},\ ] ] with symbol @xmath483 given by .
in order to determine a representation of the frchet derivative @xmath259 we have to investigate first the partial derivatives with respect to @xmath153 of the solution operators defined in section 3 . by , we get for @xmath484 that @xmath485 & -{\mathcal{g}}^{-1}(0,h){\mathcal{b}}(0,h){\mathcal{s}}_1(0,h)[g\rho_+(1+h)-\gamma_{d}\kappa(h)].\end{aligned}\ ] ] differentiating the relation @xmath486 , yields at @xmath137 that @xmath487m=-{\mathcal{g}}^{-1}(0,0){\partial}_h{\mathcal{g}}(0,0)[h]{\mathcal{g}}^{-1}(0,0)m\ ] ] for all @xmath468 we infer from the definition of @xmath218 that @xmath488=-{\partial}_h{\mathcal{b}}(0,0)[h]{\mathcal{s}}_2(0,0){\mathop{\rm tr}\nolimits}_0{\mathcal{t}}_1(0)-{\mathcal{b}}(0,0){\partial}_h{\mathcal{s}}_2(0,0)[h]{\mathop{\rm tr}\nolimits}_0{\mathcal{t}}_1(0).\end{aligned}\ ] ] given @xmath489 the partial derivative @xmath490m$ ] is the solution of the linear dirichlet problem @xmath491{\mathcal{s}}_2(0,0)m =- myh''&\text{in } & \omega_+,\\[1ex ] w&=&0&\text{on } & \gamma_1,\\[1ex ] w&=&0&\text{on}&\gamma_{0 } , \end{array } \right.\ ] ] thus @xmath490m$ ] expands as follows @xmath492m=\sum_{m\in{\mathbb{z}}\setminus\{0\}}m\left(\frac{e^{my}-e^{-my}}{e^{m}-e^{-m}}-y\right){\widehat}h(m)e^{imx},\ ] ] and similarly @xmath493m=-\sum_{m\in{\mathbb{z}}\setminus\{0\}}m\left(\frac{e^{my}-e^{-my}}{e^{m}-e^{-m}}-y\right){\widehat}h(m)e^{imx}.\ ] ] it follows then easily that @xmath494m=\frac{\mu_-\mu_+m}{(\mu_++\mu_-)^2}\sum_{m\in{\mathbb{z}}}\frac{m}{\sinh(m)}{\widehat}h(m)e^{imx}.\end{aligned}\ ] ] by definition , @xmath495 for all @xmath496 , and one can easily check that if @xmath497 , then @xmath498 summarising , we obtain that @xmath499=\sum_{m\in{\mathbb{z}}}\lambda^h_1(m){\widehat}h(m)e^{imx},\ ] ] with symbol @xmath500 given by relation . putting @xmath484 in yields @xmath502\\[1ex ] & -b_1(0,h){\mathcal{s}}_2(0,h){\mathop{\rm tr}\nolimits}_0{\mathcal{t}}_1(0)\phi_1(0,h,0)\\[1ex ] & + b_1(0,h){\mathcal{s}}_2(0,h){\mathop{\rm tr}\nolimits}_0{\mathcal{t}}_2(0)c . \end{aligned}\ ] ] using the relations derived above , finally yields @xmath503=\sum_{m\in{\mathbb{z}}}\lambda^h_2(m){\widehat}h(m)e^{imx},\ ] ] with symbol @xmath504 given by relation . | in this paper we consider the evolution of two fluid phases in a porous medium .
the fluids are separated from each other and also the wetting phase from air by interfaces which evolve in time .
we reduce the problem to an abstract evolution equation .
a generalised rayleigh - taylor condition characterizes the parabolicity regime of the problem and allows us to establish a general well - posedness result and to study stability properties of flat steady - states .
when considering surface tension effects at the interface between the fluids and if the more dense fluid lies above , we find bifurcating finger - shaped equilibria which are all unstable . |
recently , the subject of discontinuous quantum phase transitions has received much attention @xcite . besides the theoretical interest in these transitions
there are many new experimental results showing that they occur in heavy fermions and magnetic transition metal compounds @xcite . in these materials as one approaches a quantum critical point , the transitions change their nature from second to first order @xcite .
there are many mechanisms which can drive a second order transition into a first order one .
for example , in compressible magnets , pressure may produce this change @xcite . in antiferromagnets @xcite and superconductors @xcite
this may be accomplished by an external uniform magnetic field . here
we will be interested in more subtle mechanisms which arise due to a coupling of the order parameter to fluctuations @xcite . in this case
the effects are weaker being associated with what is generally known as _ weak first order transitions_. these mechanisms become more relevant at very low temperatures where quantum fluctuations are important .
the fluctuations we consider are of different types . in the first case
they are soft modes and as a paradigm of this case we study the coupling of the superconductor order parameter to the electromagnetic field .
this has been investigated by halperin , lubensky and ma @xcite but , since we are interested in the fully quantum mechanical case at zero temperature , our problem becomes similar to that treated by coleman and weinberg in particle physics @xcite .
next , we consider the effect of coupling the order parameter of a given phase to fluctuations of a different phase which competes with the former in the same region of the phase diagram .
this problem is particularly relevant for heavy fermion systems where an inhomogeneous state consisting of antiferromagnetic and superconducting regions has been observed @xcite .
this coupling between competing order parameters gives rise to new effects compared with the previous situation .
in this case not only the nature of the quantum transitions may change but the ordered phase itself may become an inhomogeneous mixture with different values for the order parameter .
we obtain general features of the behavior of a system approaching a weak first order quantum phase transition ( wfoqpt ) from finite temperatures . in spite that the correlation length does not diverge as @xmath0 , scaling does apply for these wfoqpt and allows for a general description of this phenomenon .
a well known case of quantum fluctuations inducing a weak first order transition is the coleman -weinberg mechanism @xcite . in the solid - state version ,
we consider a superconductor represented by a complex field @xmath1 coupled to the electromagnetic field @xcite .
the lagrangian density of the model is given by , @xmath2 we are using @xmath3 units and the indices @xmath4 run from @xmath5 to @xmath6 .
( [ sclagrangian ] ) space and time are isotropic and consequently the dynamic critical exponent @xmath7 . for a neutral superfluid @xmath8 the system decouples from the electromagnetic field and has a continuous , zero temperature superfluid - insulator transition at @xmath9 ( see fig .
[ fig1 ] ) .
the effective potential method @xcite yields the quantum corrections to the action given by the lagrangian density of eq .
( [ sclagrangian ] ) . at @xmath10 in the one loop approximation ,
the effective potential close to the transition ( @xmath11 ) is given by @xcite @xmath12 \label{eff1}\ ] ] where @xmath13 comes from renormalization and is completely arbitrary @xcite .
we can take the value of @xmath13 as the minimum of the effective potential @xmath14 . in this case
we have the equation , for @xmath15 , @xmath16 which yields @xmath17 or alternatively for @xmath18 @xmath19 we can use the equation above to remove @xmath18 from the effective potential eq .
( [ eff1 ] ) introducing another free parameter , namely the non - trivial minimum of the effective potential @xmath14 . in this case , eq .
( [ eff1 ] ) reads @xmath20 .
\label{eff2}\ ] ] the change of the parameters above known as dimensional transmutation @xcite can only be done if the effective potential has a non - trivial minimum @xmath21 as is clear from the equations above . since close to the transition @xmath11
, we conclude from eqs .
( [ cond ] ) and ( [ mean phi ] ) that the condition for such a minimum to exist is @xmath22 now let s define the coherence length @xmath23 and the london penetration depth @xmath24 as is usually done for ginzburg - landau models .
we can show that the condition of eq .
( [ lambda q ] ) is equivalent to @xmath25 this is the same condition for the occurrence of weak first order transitions obtained in ref .
however , in general , if @xmath26 , fluctuations of the order parameter can destabilize the coleman - weinberg mechanism and the resulting quantum transitions can be second order ( see ref .
. the condition ( [ lambda_l xi ] ) can be obtained from our results as follows . the ratio @xmath27 , from eq .
( [ xi ] ) and eq .
( [ lambda_l ] ) , is @xmath28 substituting the result of eq .
( [ mean phi ] ) we get @xmath29 this ratio is independent of @xmath30 as expected and it is clear that , if @xmath31 , we have @xmath32 or @xmath33 as expected
. we will discuss more about this condition later in section [ rg ] .
let us consider that @xmath31 ( or equivalently @xmath34 ) and study the minima of the effective potential , eq .
( [ eff2 ] ) .
we find that at a critical value of the mass , given by @xmath35 there is a first order transition to a superconducting state .
notice that the transition in the neutral superfluid ( @xmath36 ) is continuous rather than first order and takes place at @xmath9 .
therefore , quantum fluctuations due to the coupling to soft modes , the photons , lead to symmetry breaking in the normal region of the phase diagram of the neutral superfluid .
this occurs close to the quantum critical point of the chargeless system extending the region where a symmetry broken phase appears in the phase diagram .
the transition occurs at a finite value of the mass @xmath37 and is first order .
the shift of the critical mass depends on a power of the coupling of the order parameter to the soft modes , in the present case the charge of the cooper pairs . introducing a parameter @xmath38 which measures the distance to the transition , we can write the effective potential at @xmath10 as , @xmath39 with the exponent value @xmath40 reflecting the fact the transition is first order @xcite . the associated _ latent heat _ is @xmath41 .
spinodals points at @xmath10 can also be calculated @xcite .
one can distinguish different regimes as explained in the text . ]
the finite temperature case can be studied replacing the frequency integrations in the calculation of the effective potential by sums over matsubara frequencies @xcite .
the generalization of the effective potential to finite temperatures close to the transition can be written as @xcite @xmath42 \label{eff3}\end{aligned}\ ] ] where the integral @xmath43 is given by @xmath44 } .\ ] ] and @xmath45 .
the function @xmath46 can be obtained numerically integrating eq .
( [ i ] ) .
the finite temperature phase diagram is shown in fig .
[ fig1 ] . for completeness
we show the critical line of the neutral superfluid , @xmath47 , which is governed by the shift exponent @xmath48 in @xmath6 ( see ref .
the new line of first order transitions is given by @xmath49 .
we will now consider the system _ sitting _ at the new quantum phase transition point , i.e. , at @xmath50 and decrease the temperature .
this is the equivalent of the non - fermi liquid trajectory for ordinary quantum critical points in metallic systems .
of the zero temperature first order transition .
the different regimes are discussed in the text .
notice the dashed straight line that indicates the @xmath51 behavior in regime ii .
, height=219 ] for high temperatures , @xmath52 , which corresponds to the regime i of fig .
[ fig1 ] and fig .
[ fig2 ] , the function @xmath53 saturates , @xmath54 . in this case
the effective potential , @xmath55 and can be cast in the scaling form , @xmath56 .\ ] ] with @xmath57 @xmath58 .
this scaling form is reminiscent of that for the free energy close to a quantum critical point . in the present case of a discontinuous zero temperature transition ,
the critical exponent @xcite @xmath40 and the characteristic temperature is , @xmath59 @xcite . in this regime
i or scaling regime , along the line @xmath50 shown in fig .
[ fig1 ] , the free energy density has therefore the scaling form @xmath60 and the specific heat as shown in fig .
[ fig2 ] is given by , @xmath61 then the thermodynamic behavior along the line @xmath50 in regime i ( @xmath52 ) is _ the same _ as when approaching the quantum critical point of the neutral superfluid , along the critical trajectory @xmath9 .
the system is unaware of the change in the nature of the zero temperature transition and at such high temperatures charge is irrelevant . when further decreasing temperature along the line @xmath50 there is an intermediate , non - universal regime ( regime ii in figs .
[ fig1 ] and [ fig2 ] ) . in the present case for
@xmath62 the specific heat @xmath63 as can be seen from the dashed straight line shown in the semi - log plot of fig .
[ fig2 ] . finally , at very low temperatures , for @xmath64 and @xmath50 , i.e. , in regime iii of fig . [ fig2 ] , the specific heat vanishes exponentially with temperature , @xmath65
. the gap for thermal excitations is given by the shift @xmath66 of the quantum phase transition . the correlation length which grows along the line @xmath50 with decreasing temperature reaches saturation in regime iii at a value @xmath67 .
then , we can understand the exponential dependence of the specific heat as due to gapped excitations inside superconducting bubbles of finite size @xmath68 .
the gap between the states in the bubbles is @xmath69 as we have found previously .
although the results above have been obtained for a particular model , the behavior in the scaling regime i and iii should be universal and characteristic of any weak first order quantum transition .
notice that the relevant critical exponents which determine the scaling behavior in particular in regime i are those associated with the qcp of the _ uncoupled system _ which in the present case is the neutral superfluid .
weak first order transitions and spontaneous symmetry breaking near quantum phase transitions can also occur due to the coupling of an order parameter to nearly critical fluctuations of another phase .
this provides a new mechanism for wfoqpt besides the coupling to gapless excitations as treated above and in ref .
@xcite for example ( see also refs .
@xcite ) . in this section
we study the coupling , at @xmath10 , of an order parameter to fluctuations of a competing order .
this type of coupling becomes important when two different phases are in competition on the same region of the phase diagram . in this case for some values of the microscopic interactions which are close , there are alternative ground states which interfere with one another .
this is the situation , for example , of several heavy - fermions @xcite and of high - t@xmath70 superconductors @xcite where in general the dispute is among superconductivity and antiferromagnetism .
we obtain the effects on a given phase , of the coupling of its order parameter to fluctuations of the competing order , in the form of quantum corrections .
we show that these can produce spontaneous symmetry breaking and change the nature of the transition even if these fluctuations are non critical . again , due to the weak first order character of the new phase transition we expect to find scaling behavior as described previously .
we consider the case of a heavy - fermion ( hf ) system with a superconducting phase that is close to an antiferromagnetic instability at @xmath10 @xcite as in fig .
[ diagram ] . as before ,
we first describe the superconducting phase by the simplified lorentz invariant free action as in eq .
( [ sclagrangian ] ) but now coupled to an af order parameter instead of the electromagnetic field . in many cases however the effects of dissipation should be taken into account .
then , later on we use another free action for the superconducting field .
this different quadratic form is associated with a dynamic exponent @xmath71 , instead of @xmath7 of the lorentz invariant case @xcite .
our model is of the ginzburg - landau type and contains three real fields .
two fields , @xmath72 and @xmath73 , correspond to the two components of the superconductor order parameter .
the other field @xmath74 , for simplicity represents a one component antiferromagnetic order parameter .
the results are immediately generalized to a three component af vector field , with the unique consequence of changing some numerical factors @xcite .
the free functional associated with the magnetic part represented by the field @xmath74 , the sub - lattice magnetization , takes into account the dissipative nature of the paramagnons near the magnetic phase transition @xcite in the metal and yields the propagator @xmath75 where @xmath76 is a characteristic relaxation time and @xmath77 is related to a local coulomb repulsion @xmath78 and the density of states at the fermi level @xmath79 by @xmath80 the part of the action associated with the potential is given by @xmath81 where the self - interaction of the superconductor field is @xmath82 and that of the antiferromagnet is given by @xmath83 finally , the last term is the ( minimum ) interaction between the relevant fields , @xmath84 this term is the first allowed by symmetry on a series expansion of the interaction .
notice that for @xmath85 , which is the case here , superconductivity and antiferromagnetism are in competition and this term does not break any symmetry of the original model .
however , including quantum fluctuations we show that spontaneous symmetry breaking can occur in the normal phase separating the sc and af phases .
this case is represented schematically in fig .
[ diagram ] and is discussed in the subsequent sections .
we first consider the metal - superconductor transition in hf when the superconducting field is coupled , as in eq .
( [ coupling ] ) , to af fluctuations described by the propagator of eq .
( [ propagatoraf])(fig . [ diagram ] ) .
detailed calculation of the effective potential have already been presented @xcite .
the general result is given by , @xmath86 in eq .
( [ potsc ] ) , @xmath87 is a renormalized coupling but of the same order of the bare coupling @xmath18 .
the new coupling @xmath88 introduced by fluctuations can produce a symmetry breaking in the normal state extending the sc region in the phase diagram at @xmath10 .
the same mechanism turns the transition to weak first order with a small latent heat@xcite . using the same argument as in the preceding section we can introduce again parameters associated with @xmath89 ( coherence length ) and @xmath90 ( london penetration depth ) in this case given by @xmath91 it s difficult to carry out the dimensional transmutation analytically on the result of eq .
( [ potsc ] ) since the equations of minima are hard to solve .
however , we can check numerically that the condition for the existence of minima away from the origin is equivalent to @xmath34 as in the previous case .
it s also interesting to notice that the first order transition is produced by the cubic term in eq .
( [ potsc ] ) and this term is proportional to the _ magnetic mass _ @xmath92 . if magnetic fluctuations were critical , i.e. , @xmath93 , the only effect of the coupling would appear as a term proportional to @xmath94 .
this term is usually neglected since its power is higher than those initially considered in the classical potential and usually insufficient to create new minima around the origin .
therefore , if af fluctuations were critical the effects of the quantum corrections in the transition could be neglected . in order to study the magnetic transition , as mentioned previously we consider two kinds of quadratic forms associated with the free superconducting fields .
the first is the usual lorentz invariant free action that we have discussed before in sec .
[ scandaf ] .
next we work with another free action which takes into account dissipation and is associated with a @xmath71 dynamics .
let s first study the result for the lorentz invariant propagator .
close to the af transition we have @xcite @xmath95 we notice directly from eq .
( [ potaf ] ) that if the superconductor order parameter fluctuations were critical we would obtain the same result of eq .
( [ eff2 ] ) since in this case the parameter @xmath9
. however , sc fluctuations are not critical but close to criticality and the last term of eq .
( [ potaf ] ) may become important .
of course , its relevance depends on the strength of the renormalized coupling @xmath96 and the results considering this term leads to new and interesting changes in the ground state .
we obtain , besides the two finite minima of the coleman - weinberg potential of eq .
( [ eff2 ] ) , two extra minima very close to the origin @xcite .
the normal state associated with these minima has a small moment since the value of the order parameter is related to the sub - lattice magnetization .
an additional first order transition occurs when the other two minima away from the origin become the stable ones as the system moves away from the superconductor instability .
this transition is from a small moment af ( smaf ) to a large moment af ( lmaf ) and occurs even before the continuous mean field transition .
when we move away from the magnetic transition the strength of this new term decreases with the value of @xmath30 and the two new minima move to the origin producing a normal state with vanishing magnetization again .
it s important to note here that the smaf is obtained because the magnetic order parameter couples to fluctuations which are close to criticality but not critical .
critical fluctuations yield the same results of section [ cw - mechanism ] .
now , for many cases of interest , sc fluctuations are better described by a dissipative propagator associated with a @xmath71 dynamics@xcite similar to eq .
( [ propagatoraf ] ) .
this is useful to account for pair breaking interactions as magnetic impurities that can destroy superconductivity @xcite .
it is given by , @xmath97 the parameter @xmath30 is still related with the classical distance from the phase transition and we have a relaxation time @xmath98 .
in general we also have a non - dissipative term ( as in the case studied above ) but this term is neglected since the linear term is the most important in the low frequency region .
calculation of the effective potential is very similar to the previous cases and the result has the form @xmath99 where the renormalized mass is @xmath100 and the renormalized coupling @xmath101 quantum corrections can once again produce a weak first order transition .
we note that in this case the renormalization of the potential obtained is easier since there are no logarithmic terms to renormalize @xcite .
anyway , the result is very similar to that presented in section [ sctransition ] since we have equivalent propagators for the fluctuations . in general the form of the quantum correction depends not only on the effective dimension of the uncoupled qcp but also on the form and specially the dynamics of the soft modes or competing fluctuations . the transformation @xmath102 makes the potential of eq .
( [ veffz=2 ] ) simpler and the analysis of its extrema shows that the transition can be first order depending on the coupling values .
the appearance of smaf phases is not possible in this case .
we note , however , that the couplings are dependent on the renormalization parameter @xmath13 defined in sec .
[ cw - mechanism ] and a renormalization group approach is essential to make the results more reliable . in the next section
we illustrate this with an analysis of the coleman - weinberg case .
in the previous sections we have obtained weak quantum first order transitions from minima generated balancing terms in the effective potential . this is possible when we compare terms of the same order and therefore the minima depend on the values of the couplings .
considering , for example , the superconductor coupled to the electromagnetic field , the terms we have to balance are proportional to @xmath18 and @xmath103 .
new minima away from the origin are produced if @xmath104 and this condition also leads to a relation for the london penetration depth and the coherence length @xmath34 . in this section
we show how we can use renormalization group arguments to generalize the results for any small @xmath18 and @xmath105 and get rid of the condition ( [ conrg ] ) . when deriving eq .
( [ eff1 ] ) we have introduced a new parameter @xmath13 , associated with a renormalization mass and which was completely arbitrary .
the results for the model must be the same for any chosen value of @xmath13 .
now , if we prove that a variation in the value of @xmath13 produces a correspondent variation in the coupling constant @xmath18 , it s always possible to satisfy the condition ( [ conrg ] ) choosing a suitable value of @xmath13 . to prove that , we have to consider eq .
( [ eff1 ] ) .
we want to rewrite this equation with a value @xmath106 , such that , @xmath107 \nonumber\\ v_{eff } & = \frac{1}{2}m^{2}\varphi^{2}+\frac{\lambda}{4!}\varphi^{4}+\frac{3q^{4}}{64\pi^{2}}\varphi^{4}\ln\frac{{m^{\prime}}^{2}}{m^{2 } } + \nonumber\\ & \frac{3q^{4}}{64\pi^{2}}\varphi^{4}\left [ \ln\left ( \frac{\varphi^{2}}{{m^{\prime}}^{2}}\right ) -\frac{25}{6}\right ] \nonumber\end{aligned}\ ] ] which is equivalent to eq .
( [ eff1 ] ) with the reparametrization @xmath108 in this sense the effective potential is always given by the same equation with a suitable parametrization of the coupling .
this argument can be formally stated in an renormalization group approach and we refer to the correspondent section of ref . @xcite . as a result
, a weak first order transition occurs for any small values of @xmath18 and @xmath105 , since we can appropriately choose the renormalization mass .
the only restriction comes from perturbation theory which requires small couplings . for the case of coupling between af and sc order parameters we can also construct a renormalization group . within this approach
we expect to find all the conditions for the occurrence of weak first order transitions .
detailed calculations and results will be published elsewhere .
we have studied the effects on a phase transition of coupling the order parameter of this phase to soft modes or to fluctuations of a competing phase .
since these effects become more important at low temperatures we have mostly considered the zero temperature case .
we have obtained the effects of quantum corrections due to this coupling on the quantum critical point associated with the relevant order parameter .
these corrections were calculated using the effective potential method .
they can produce radical modifications on the nature of the original phase transition .
they can change the nature of the transition , from continuous to discontinuous and also affect the ordered phase itself giving rise to an inhomogeneous phase with two values of the order parameter . whenever the fluctuations change the nature of the transition to weak first order , we can study them using a scaling approach .
we have obtained explicit results for one particular case but that should hold in general . as the system approaches the discontinuous transition from non - zero temperature , we can distinguish three different regimes . in the first , the high temperature regime , the scaling behavior is the same as approaching the quantum critical point of the uncoupled system , i.e. , in the absence of additional fluctuations or soft modes .
further decreasing the temperature there is a non - universal regime which may depend on the particular dynamics of the original system and of the nature of the additional modes .
finally at the lowest temperatures there is a regime which is characteristic of the first order nature of the zero temperature transition . at such low temperatures the correlation length saturates at a finite value .
the thermodynamic properties have a contribution which is thermally activated .
this can be traced to excitations which are gapped due to finite size effects as they occur in finite bubbles of the incipient phase .
the correctness of this interpretation is clear from the fact that the gap for excitations is inversely related to the size of the bubbles which in turn is of the order of the correlation length .
notice also that the former is related to the shift in the zero temperature phase transition with respect to that of the uncoupled system .
the relevant exponents which determine the scaling behavior near the weak first order transition are those associated with the quantum critical point of the uncoupled system , i.e. , of the system in the absence of competing fluctuations or soft modes .
in particular this is the case of the dynamic exponent .
the scaling properties of the new transition are related to the existence of an underlying second order instability which has become discontinuous due to the effects of soft modes or of fluctuations of a competing order parameter .
s. kawasaki , et al.,phys .
lett . * 91 * , 137001 ( 2003 ) ; y. kitaoka , s. kawasaki , t. mito , y. kawasaki , to appear in j. phys .
jpn , 74 , no.1 ( 2005 ) ; w. yu , et al . phys
* 92 * , 086403 ( 2004 ) .
m. uhlarz , c. pfleiderer , and s. m. hayden phys .
lett . * 93 * , 256404 ( 2004);c .
pfleiderer and a. d. huxley phys .
* 89 * , 147005 ( 2002);a .
bianchi , r. movshovich , n. oeschler , p. gegenwart , f. steglich , j. d. thompson , p. g. pagliuso , and j. l. sarrao phys .
lett . * 89 * , 137002 ( 2002);c .
pfleiderer et al . , nature ( london ) * 412 * , 58 ( 2001);n .
kimura , m. endo , t. isshiki , s. minagawa , a. ochiai , h. aoki , t. terashima , s. uji , t. matsumoto , and g. g. lonzarich phys .
lett . * 92 * , 197002 ( 2004 ) the concept of generic scale invariance ( gsi ) has been emphasized recently with respect to systems with continuous broken symmetry @xcite . in the simplest case of a heisenberg ferromagnet in an external magnetic field ( @xmath109 )
gsi is related to the strong coupling fixed point at @xmath110 , @xmath111 which governs the symmetry broken phase .
close to this attractor , @xmath112 and @xmath113 .
the free energy density scales as , @xmath114 $ ] .
then @xmath115 , the magnetization @xmath116 and the susceptibility @xmath117 .
an additional contribution arises at @xmath118 , @xmath111 , namely @xmath119 $ ] .
this yields @xmath120 and @xmath121 for @xmath122 @xcite .
sidorov , m. nicklas , p.g .
pagliuso , j.l .
sarrao , y. bang , a.v .
balatsky , j.d .
thompson phys .
89 , 157004 ( 2002 ) ; p. g. pagliuso , n. o. moreno , n. j. curro , j. d. thompson , m. f. hundley , j. l. sarrao , z. fisk , a. d. christianson , a. h. lacerda , b. e. light and a. l. cornelius . ,
phys . rev . * b66 * , 054433 ( 2002 ) ; m. nicklas , v. a. sidorov , h. a. borges , p. g. pagliuso , c. petrovic , z. fisk , j. l. sarrao , and j. d. thompson , phys.rev . * b67 * , 020506 ( 2003 ) ; r. a. fisher , f. bouquet , n. e. phillips , m. f. hundley , p. g. pagliuso , j. l. sarrao , z. fisk , and j. d. thompson , phys.rev . *
b65 * , 224509 ( 2002 ) . at @xmath10
we associate the expression _ latent heat _ to a latent work . in this case
the latent work is small since it is proportional to @xmath123 in which @xmath66 is the value of the critical mass where the transition occurs .
see a. s. ferreira , m. a. continentino and e. c. marino , sol .
* 130 * , 321 ( 2004 ) . | discontinuous quantum phase transitions besides their general interest are clearly relevant to the study of heavy fermions and magnetic transition metal compounds .
recent results show that in many systems belonging to these classes of materials , the magnetic transition changes from second order to first order as they approach the quantum critical point ( qcp ) .
we investigate here some mechanisms that may be responsible for this change .
specifically the coupling of the order parameter to soft modes and the competition between different types of order near the qcp . for weak first order quantum phase transitions
general results are obtained . in particular
we describe the thermodynamic behavior at this transition when it is approached from finite temperatures .
this is the discontinuous equivalent of the non - fermi liquid trajectory close to a conventional qcp in a heavy fermion material . |
perhaps the most mysterious thing about black holes is their capacity to store and process information . already examining the behavior of black hole entropy @xcite in the classical limit ,
one arrives at the following puzzle @xcite : according to the bekenstein formula , the black hole entropy scales as @xmath1 , where @xmath2 is the gravitational radius and @xmath3 is the planck length , which in terms of newton s coupling @xmath4 and the planck constant @xmath5 is defined as @xmath6 .
this expression diverges in the classical limit , if @xmath2 and @xmath4 are kept fixed .
this is because @xmath7 in this limit .
thus , a quantum physicist , to whom the classical physics is defined as a limit @xmath8 , concludes that the entropy ( and thus the information - storage capacity ) of a classical black hole of a finite mass is infinite . on the other hand , a classical observer would seem to have a very different perspective .
a classical physicist knows that information can only be encoded in the black hole in features that are possible to re - arrange .
but it is well known that classical black holes are essentially featureless .
they are uniquely characterized by their mass , charge and the angular momentum .
consequently , the classical observer would conclude that black holes should carry zero entropy .
thus , it seems that the two observers must disagree on what happens in the @xmath9 limit .
how can infinity and zero be reconciled ?
the above puzzle , along many others , can only be resolved within a microscopic theory , which explicitly identifies the quantum degrees of freedom that act as carriers of the black hole information .
we shall focus on a microscopic theory of this type offered in @xcite , in which the black hole is described as a composite multi - particle state of soft gravitons at a _ quantum critical _ point @xcite .
this system can approximately be described as a bose - einstein condensate ( or a coherent state ) of gravitons of characteristic wavelength @xmath10 and occupation number @xmath11 . according to this picture , the microscopic carriers of black hole entropy and information are collective excitations of the bogolyubov - type @xcite .
the remarkable property of these modes is that around the critical point they become nearly gapless and decouple from the rest of the modes in the @xmath12 limit .
hence , if a system has @xmath0 gapless bogolyubov modes , there are expected to exist @xmath13 nearly - degenerate quantum states that can be labeled by the occupation numbers of these modes .
these states are crowded within a @xmath14 energy gap .
correspondingly , the entropy should scale as @xmath15 .
such a behavior uncovers a very peculiar nature of black hole information storage and in particular addresses the above puzzle . in the classical limit @xmath16 and entropy is indeed infinite , but the energy gap closes to zero . despite the fact that the number of quantum states located within this gap diverges , the required time for resolving their differences also becomes infinite .
this time - delay reconciles the views of the two observers .
the classical black hole appears as featureless for any finite interval of the observation time . in the present paper
we shall focus our attention on improving the understanding of the relation between quantum criticality and physics of information - processing . since , according to the black hole @xmath0-portrait the underlying fundamental ingredient of black hole information - processing is a self - sustained quantum criticality of the cold graviton gas , the natural question to ask is : how important is the precise nature of the constituent bosons ?
of course , for recovering the standard gravitational properties of the system , such as the emergence of the einsteinian metric in a mean field description , the spin-2 nature of the constituents is crucial .
however , one may suspect that the information - processing features that originate from the fact of criticality , may not be so sensitive to the precise nature of the constituents .
if this is true , then a wide class of critical systems must share some of the black hole information - processing abilities .
this would be a very important conclusion because of the following reasons .
first , it would enable us to study black hole properties on much simpler systems , both theoretically and experimentally .
secondly , it would open up a possibility of implementing the black hole information - processing methods in table - top experiments .
indeed , it was noticed already in @xcite that the simplest prototype models of attractive bose - einstein condensates exhibit some striking similarities with the black hole quantum portrait .
for example , the cost of the information - storage per qubit is by a factor of @xmath17 cheaper relative to the energy - cost exhibited by usual ( non - critical ) quantum systems . in the latter systems , for the same amount of information storage , one typically pays the energy price of the inverse size of the system , @xmath18 .
also , the degeneracy of states within the @xmath17 energy gap increases with @xmath0 .
moreover , it was shown that a sharp increase of one - particle entanglement takes place near the critical point @xcite and the scrambling of information becomes maximally - efficient @xcite .
the time - dependent evolution of the critical condensate uncovers a scaling solution in which the condensate is stuck at the critical point throughout the collapse @xcite .
this is the behavior that one would expect if the microscopic foundation of hawking radiation were through the collapse and quantum depletion of the condensate .-
matrix that the above multi - particle picture represents the correct microscopic description of black hole physics .
indeed , the @xmath19 scattering @xmath20-matrix element of two trans - planckian gravitons into @xmath0 soft ones reproduces an entropy suppression factor @xmath21 when the number @xmath0 is given by the quantum critical value @xcite .
similar conclusions were reached in @xcite in a different approach . for other aspects and related work ,
see @xcite . ] in the present paper we would like to take one more step in understanding the universal role of quantum criticality for information - processing .
to this end , we invent a new method for deriving the effective low - energy theory of modes near the critical point .
as we shall see , the resulting effective theory indeed describes nearly - gapless modes with @xmath17-suppressed couplings .
this form immediately reveals the key features of information storage , such as the low energy cost and the long timescale for retrieval as well as the high speed of information scrambling in the over - critical unstable phase .
in our approach we formulate an effective description of quantum criticality in the isomorphic language of a pseudo - goldstone phenomenon in a @xmath22-sigma - model .
the global @xmath22-symmetry of the sigma model corresponds to a continuous unitary transformation among the creation and annihilation operators of the bosons with different momenta @xmath23 .
this identification allows us to map the hamiltonian of the bose - gas onto a potential of the @xmath22-sigma model . in this way , the problem of diagonalization of the hamiltonian becomes isomorphic to finding the symmetry breaking patterns in the sigma model .
the latter is achieved by the straightforward minimization of the potential .
the resulting pattern determines the spectrum of pseudo - goldstone excitations .
the quantum critical points of the bose - gas correspond to the transitions between the different symmetry breaking patterns in the sigma model , where some goldstone modes become gapless .
we show that this transition takes place when the particle number @xmath0 and the strength of the attractive interaction @xmath24 exactly balance each other , @xmath25 .
this approach allows to derive an effective hamiltonian of the gapless excitations in form of the potential energy part of the goldstone chiral lagrangian .
this form reveals that near the critical point the goldstones represent gapless excitations with @xmath17-suppressed self - interactions , @xmath26 where @xmath27 is the goldstone number operator .
analyzing the resulting effective theory , we observe a striking connection with anticipated features of black hole information processing .
namely , the energy cost per information qubit goes to zero as a power of @xmath17 and correspondingly near the critical point the time for information retrieval increases with @xmath0 .
the number of states within the collapsed energy gap increases either as a power law or exponentially with @xmath0 depending on the nature of the attractive coupling . for a non - derivative self interaction
the increase follows a power law , whereas for a gravity - type derivatively - coupled case the degeneracy is increasing exponentially with @xmath0 . on the other hand
, the instability of the system in the over - critical regime implies a positive lyapunov exponent which , as shown in @xcite , is the engine for scrambling of information .
the logarithmic scaling of the scrambling time for black holes was originally conjectured in @xcite .
the observation that critical bose - einstein systems are fast scramblers provides the microscopic understanding of this phenomenon in terms of a lyapunov exponent and the high density of states near the quantum critical point @xcite . in summary , the effective theory of information at quantum critical point developed in the present paper confirms , in accordance with @xcite , that critical multi - boson systems are indeed efficient storers and processors of information .
finally , in identifying the information carriers for black hole entropy , we distinguish two distinct sources of quantum degeneracy .
since degenerate states are labeled by occupation numbers of gapless goldstone modes , we can create degenerate states either by increasing the occupation numbers of a small subset of goldstone modes , or by diversifying the population among many different goldstone modes , while keeping individual occupation numbers small .
the crucial difference among the two types of states is that the information stored in un - diversified large occupation numbers can in principle be resolved in interference experiments even in the classical limit .
irrespectively whether contributions from such states must be counted into black hole entropy , their existence gives a very interesting possibility of detecting the quantum sub - structure in macroscopic experiments .
we start with the system that was used in @xcite as the simplest prototype model for exploring the connection between black hole information processing and quantum criticality . it is a gas of cold bosons with a delta - function - type attractive interaction .
such systems are well known and have in particular been studied in the context of cold atoms @xcite .
we consider a hamiltonian of the following form , @xmath28 where @xmath29 , @xmath30 is the @xmath31-dimensional volume and @xmath32 is the @xmath31-dimensional wave - number vector .
@xmath33 are creation and annihilation operators of bosons of momentum @xmath34 .
they satisfy the usual commutation relation @xmath35 = \delta_{\vec{k}\vec{k'}}$ ] .
@xmath36 is the coupling . rescaling the hamiltonian , we can write @xmath37 . introducing a notation @xmath38 and taking @xmath39 we arrive at the following hamiltonian ,
@xmath40 we restrict ourselves to three levels , @xmath41 .
legitimacy of this approximation shall be justified later .
let us define a triplet operator @xmath42 .
correspondingly , we define the number operators @xmath43 .
we treat the operator @xmath44 as a triplet under a global symmetry group @xmath45 , @xmath46 where @xmath47 is a unitary transformation matrix which keeps invariant the total number operator , @xmath48 thus , in any state with non - zero particle occupation number , @xmath49 the @xmath45-symmetry is spontaneously broken and the order parameter of breaking is @xmath0 .
this fact allows us to map the dynamics of the quantum phase transition in a bose - einstein condensate described by the hamiltonian ( [ hamilton ] ) onto the pseudo - goldstone phenomenon in an @xmath45 sigma model .
the hamiltonian in terms of the triplet @xmath44 can be written as @xmath50 where , @xmath51 is the @xmath45-invariant part of the hamiltonian and has the following form , @xmath52 here , @xmath53 is the lagrange multiplier that fixes the total particle number .
let us first investigate the ground - state of the @xmath45-symmetric hamiltonian .
minimization with respect to @xmath54 gives the following equations , @xmath55 and @xmath56 the vacuum is achieved at @xmath57 and @xmath58 .
the @xmath45-symmetry is spontaneously broken down to @xmath59 and there is a doublet of nambu - goldstone bosons can thus be understood as the breaking of @xmath60 on the hilbert space and gives rise to @xmath61 goldstone modes . ] .
for example , choosing the vacuum at @xmath62 and ignoring @xmath63-corrections to the zero mode occupation , the goldstone doublet is @xmath64 .
the existence of the goldstone boson is the manifestation of the indifference of the system with respect to redistributing occupation numbers among the different levels .
now , the addition of the first three terms in ( [ hamilton1 ] ) breaks the @xmath45 symmetry explicitly and lifts the vacuum degeneracy .
let us investigate the effect of these terms on the state @xmath65 .
notice , this vev continues to be the extremum of the hamiltonian even after the addition of the explicit - breaking terms , but the value of @xmath53 is now shifted to @xmath66 .
the would - be goldstone modes now acquire a non - trivial mass - matrix , @xmath67 this matrix is diagonalized by states @xmath68 and @xmath69 , with eigenvalues @xmath70 and @xmath71 respectively . in this basis the hamiltonian becomes , @xmath72 thus , a would be goldstone doublet is now split into a light mode and its heavy mode partner . since the explicit breaking terms leave unbroken the @xmath73 subgroup of @xmath45 , corresponding to the generator @xmath74 , the mass eigenstates @xmath75 and @xmath76 are also @xmath73-charge eigenstates .
they carry charges equal to @xmath77 and @xmath78 respectively . in the usual language , the charge @xmath79 is a momentum operator .
let us at this point highlight the relation between the goldstone modes @xmath75 and @xmath76 and the usual bogolyubov eigenmodes of the mass matrix .
the latter are obtained by diagonalizing the mass matrix through a canonical transformation ; in our setup , they are given by @xmath80 with @xmath81 in this description , the hamiltonian has two degenerate eigenvalues @xmath82 the modes @xmath75 and @xmath76 relate to the bogolyubov modes through @xmath83 the goldstone mode thus encodes a light direction in configuration space , reachable through occupying both bogolyubov modes simultaneously .
note that a state with bogolyubov occupation number of @xmath63 corresponds to a goldstone configuration with @xmath84 , which is responsible for the @xmath85-difference in the eigenvalues .
the heavy partner , on the other hand , corresponds to a heavy multi - particle direction in configuration space , reachable through populating either one of the bogolyubov modes with @xmath86 particles . in turn
, this explains the @xmath87-independence of the second eigenvalue of . for @xmath88 one of the pseudo - goldstone bosons becomes tachyonic and the vacuum is destabilized .
the physics of this instability is that for @xmath89 the @xmath59 preserving vacuum is no longer energetically favorable and the system flows towards the new ground - state .
this new ground - state can be found by minimizing the full hamiltonian , ( [ hamilton1 ] ) . in order to minimize this hamiltonian ,
let us set @xmath90 .
the only phase - dependent term in the hamiltonian is the second term in ( [ hamilton1 ] ) , which takes the form @xmath91 since there is no other conflicting phase - dependent term in the energy , in the minimum we will have @xmath92 . we can thus set without any loss of generality , @xmath93 another simplifying observation is that in the extremum @xmath94 and @xmath95 must be equal .
this can be seen by extremizing the hamiltonian with respect to @xmath94 and @xmath95 , @xmath96 @xmath97 which show that @xmath94 and @xmath95 can be non - zero only together .
moreover they must be equal , since they satisfy the same equation , with only one positive root .
this can be seen by multiplying ( [ min1 ] ) and ( [ min2 ] ) by @xmath94 and @xmath98 respectively .
the resulting quadratic equations are identical , @xmath99 @xmath100 and have only one positive root .
thus , without any loss of generality we can minimize the hamiltonian for the configuration @xmath101 , which gives @xmath102 since @xmath103 is positive definite , for @xmath104 , the minimum is achieved at @xmath105 . for @xmath89 , the minimum is at @xmath106 not surprisingly , this corresponds to the first fourier modes of the bright soliton solution to the gross - pitaevskii equation in the overcritical regime @xcite .
the energy of the ground - state is given by @xmath107 where the first term only exists for @xmath89 .
we can now understand the critical phenomenon in terms of this goldstone - mode .
the large occupation number serves as an order parameter for a spontaneous breaking of a global @xmath45-symmetry .
this symmetry corresponds to redistribution of the particle occupation numbers between the different momentum states , without changing their total number .
this symmetry is only approximate , and normally the would - be goldstone mode has the mass of the order of the first momentum level ( that is order one in our units ) .
however , the phase transition corresponds to the point where the particle number distribution changes .
this necessarily implies that the corresponding pseudo - goldstone mode must become massless at this point . notice that since @xmath103 parameterizes the occupation number of ( pseudo-)goldstones , @xmath108 , the hamiltonian ( [ hamiltonx ] ) essentially represents the effective action for this mode , @xmath109 thus , the mass of the pseudo - goldstone is given by @xmath110 and the self - coupling is given by @xmath111 . at the critical point ,
the goldstone mass term vanishes and the theory is described by a gapless mode with a self - interaction strength given by @xmath112 , @xmath113 of course , the phenomena uncovered in the goldstone formulation are in one - to - one correspondence to those that are seen in a mean - field and bogolyubov description of the model .
in fact , as hinted before , the minima of the hamiltonian eq .
correspond to the @xmath114 modes of the solutions to the mean - field gross - pitaevskii equation : for @xmath115 we found the homogenous condensate , while the solution for @xmath89 corresponds to a localized bright soliton . at various values of @xmath0 , using the parametrization @xmath116 .
@xmath117 are normalized according to @xmath118 , since as @xmath119 , @xmath120 and thus it s possible to compare them at different @xmath0 . as @xmath0 increases
, the distribution becomes increasingly more peaked at @xmath121 . *
( b ) * @xmath122 for @xmath123 particles compared with the analytic prediction @xmath124 .
, title="fig : " ] at various values of @xmath0 , using the parametrization @xmath116 .
@xmath117 are normalized according to @xmath118 , since as @xmath119 , @xmath120 and thus it s possible to compare them at different @xmath0 .
as @xmath0 increases , the distribution becomes increasingly more peaked at @xmath121 . *
( b ) * @xmath122 for @xmath123 particles compared with the analytic prediction @xmath124 .
, title="fig : " ] in this context , one should also keep in mind that the solution @xmath125 only agrees with the exact quantum mechanical ground state in the limit @xmath126 . at any finite @xmath0 ,
the ground state is a smeared distribution centered around @xmath125 with @xmath103 given by , as can be seen in fig . .
in particular , the ground state at the critical point is characterized by strong entanglement and is therefore not well described by a mean field .
this , however , will not alter our conclusion on state - counting and information processing to be performed in the following sections .
we reemphasize that in the goldstone language , the critical point amounts to destabilization of the @xmath59-invariant vacuum . with the pseudo - goldstone method we have traded the diagonalization of the hamiltonian with the minimization procedure .
substituting the small deformations of the order parameter by the goldstone mode allowed us to derive the effective action for the latter .
this is similar to deriving an effective hamiltonian of a phonon field in a background external magnetic field , which breaks the rotational symmetry of a ferromagnet .
the analogous role in our treatment is assumed by the terms in the hamiltonian that explicitly break @xmath45-symmetry .- color qcd . in this case
our @xmath0 would be mapped on the number of colors , whereas the levels @xmath127 to quark flavors .
the pseudo - goldstone boson then is analogous to pion , which also has @xmath17 suppressed self - coupling . ]
we conclude this section by pointing out that one can continue an analytic treatment of the system into the solitonic regime by performing an @xmath103-dependent canonical rotation on the creation and annihilation operators @xmath128 .
we define the condensate mode @xmath129 and the corresponding orthogonal modes @xmath130\,.\ ] ] the resultant hamiltonian is now always minimized by the configuration @xmath131 .
expanding around this vacuum leads to a quadratic hamiltonian which can be written in matrix form as @xmath132 with @xmath133 canonical diagonalization again leads to two distinct eigenmodes ; for @xmath115 , their energy is @xmath134 . for @xmath89 , one of them
remains exactly gapless , while the other has a frequency @xmath135 .
the former is the mode that corresponds to translations of the soliton , whereas the other is the bogolyubov mode that is only light in the vicinity of the critical point . in the goldstone language ,
the goldstone mode @xmath75 splits into two parts , essentially given by its real and imaginary parts .
one remains gapless and generates translations of the localized ground state ; the other is gapless only at the critical point and obtains a frequency @xmath136 for @xmath89 .
the appearance of the light modes can of course again be related to the breaking of @xmath45 generators . keeping in mind that in the rotated description the symmetry transformations are given by @xmath137 , where @xmath20 is the symplectic matrix that generates the canonical transformation
, we immediately see that the translation generator @xmath138 , which commutes exactly with the hamiltonian for all @xmath103 , does no longer annihilate the groundstate .
the generator that redistributes particle numbers , on the other hand , only commutes with @xmath139 at the critical point .
the above form of the effective hamiltonian displays the role of quantum criticality for information storage and processing . the quantum information in the above system is encoded in the state of the goldstone mode .
the remarkable thing about it is the low energy cost of information - qubit - storage , which is suppressed by powers of @xmath17 relative to the inverse size of the system . this phenomenon is a manifestation of quantum criticality . in the time evolution of the goldstone state
we can distinguish two sources .
one is the interaction governed by a quartic self - interaction hamiltonian ( [ goldeffcritical ! ] ) .
the rate of this process is suppressed by powers of @xmath17 , and correspondingly the time - scale of evolution is very long .
the second source of time evolution is the goldstone mass term that parametrizes the departure from quantum criticality . in the overcritical regime
, this mass is imaginary and results in an exponential growth of the goldstone occupation number .
this instability is described by a lyapunov exponent that , as shown in @xcite , leads to the generation of one - particle entanglement and potentially to the scrambling of information .
an exact numerical diagonalization of the hamiltonian provides a complementary analysis which is valid also at the critical point . using the same technique as in @xcite
, we can verify the above results by comparing the the first bogolyubov state as well as a lowly occupied goldstone state to the exact eigenstates of the hamiltonian .
several quantities are instructive . in fig.[fig :
gsenergy ] we plot the expectation value of the energy in a state with goldstone occupation @xmath140 as a function of @xmath87 .
the solid line corresponds to the analytic result .
the increase of energy around the phase transition is suppressed by @xmath17 and thus not visible . ] as a function of @xmath87 for @xmath141 .
[ fig : gsenergy ] in fig.[fig : time_evol]a we plot the exact time evolution of the bogolyubov state @xmath142 , where @xmath143 is the bogolyubov ground state .
the results are obtained for fixed particle number @xmath141 , while the effective coupling @xmath87 is varied .
we observe an decrease of the frequency as @xmath87 approaches the critical point .
to illustrate this point better , we plot the frequency of oscillations versus @xmath87 for @xmath144 and @xmath145 in fig [ fig : time_evol]b .
we observe the square root behavior , with the frequency scaling with @xmath87 as @xmath146 . and
@xmath147 ( black , solid ) , @xmath148 ( red , dashed ) and @xmath149 ( blue , dotted ) . *
( b ) * frequency of oscillation as a function of @xmath87 at various values of @xmath0 .
the solid line corresponds to the analytic result @xmath150,title="fig : " ] and @xmath147 ( black , solid ) , @xmath148 ( red , dashed ) and @xmath149 ( blue , dotted ) . *
( b ) * frequency of oscillation as a function of @xmath87 at various values of @xmath0 .
the solid line corresponds to the analytic result @xmath150,title="fig : " ]
we can now verify that our starting assumption about the stability and a negligible occupation number of higher momentum modes is valid as long as @xmath151 .
as we shall see , the occupation numbers of modes with momentum number @xmath152 are suppressed by a factor of @xmath153 .
this is a manifestation of the fact that the global @xmath154-symmetry , corresponding to redistribution of occupation numbers among the fist @xmath152 levels , is explicitly broken by the momentum term in the hamiltonian ( [ hamilton ] ) .
we can think of this breaking as the decoupling limit of a _ spontaneous _ breaking by a spurion order parameter , @xmath155 , in the adjoint representation of the @xmath156 group , with the expectation value @xmath157.-subgroup with the generator @xmath79 which is proportional to @xmath155 . as in the case of @xmath45 ,
this charge represents a momentum operator .
] obviously , the kinetic term in the hamiltonian ( [ hamilton ] ) , can then simply be written as an @xmath156-invariant product , @xmath158 .
the spurion @xmath159 can be thought of as an expectation value of the number operator of a `` hidden '' bose - gas of some @xmath160-particles , interacting with @xmath44 by an infinitely weak coupling @xmath161 , s.t .
@xmath162 . in that case
, the momentum operator is replaced by a peculiar @xmath156 invariant interaction term and both occupation numbers spontaneously break one and the same @xmath156-symmetry .
the goldstones bosons are superpositions of @xmath163-s and @xmath164-s .
the orthogonal combinations are pseudo - goldstones .
now , let us take the limit @xmath165 , while simultaneously taking the occupation numbers of @xmath164-particles to infinity in such a way that the products @xmath166 are kept fixed for all @xmath167 .
we arrive at the situation in which in the @xmath163-particle sector , the entire information about the breaking is summed up in the expectation value of the spurion @xmath155 .
correspondingly , all the goldstones in the @xmath163-sector become pseudo - goldstones . with increasing momentum level @xmath168 , the masses of pseudo - goldstones increase whereas their occupation numbers rapidly diminish .
in order to see this , let us compute the occupation numbers of the @xmath169 modes .
the corresponding annihilation operators for @xmath170 and @xmath171 are @xmath172 and @xmath173 , respectively .
the bilinear mass matrix of these modes expanded about the state @xmath174 , has the form @xmath175 this has two eigenvalues , @xmath176 and @xmath177 .
both eigenstates are stable as long as @xmath178 . however , the modes @xmath179 nevertheless can get populated due to the mixing with the lower levels . the occupation number of the lower modes acts as a source for the higher ones .
the occupation numbers of the higher modes can be easily obtained by minimizing the bilinear hamltonian including the effective source terms for @xmath180 .
this source is generated from the interaction term after plugging the expectation values of @xmath181 .
the phases of the expectation values can again safely be set to zero , because all the source terms are negative .
the corresponding effective hamiltonian for @xmath172 and @xmath173 has the form , @xmath182 minimizing this with respect to @xmath180 we see that @xmath183 similarly , we can derive the occupation numbers of higher momentum modes . in general , modes with momentum @xmath168 have occupation numbers @xmath184 , which rapidly approaches zero for @xmath185 .
thus in this regime , our assumption is well justified . in particular
, this regime covers the neighborhood of the phase transition , around which @xmath103 is small .
this behavior can also be confirmed by a numerical study of the system . to this end
, we consider the hamiltonian with a momentum cut - off at @xmath186 .
an instructive quantity is the occupation number @xmath179 of the @xmath186 modes for low lying states around the phase transition . as can be seen in fig .
[ fig : k2 ] , there is no appreciable contribution of the @xmath186 modes to the low lying states at least until @xmath187 .
we can moreover confirm the quadratic behavior .
modes to the lowest lying eigenstates for @xmath188 ( black ) and @xmath189 ( red ) .
the dots correspond to the exact numerical results , while the solid line depicts the analytical behavior . ]
the model considered in the previous sections admits two solutions that become degenerate for @xmath126 at the critical point . in @xmath31 dimensions ,
this number can be increased by a factor of @xmath31 ; this corresponds to the number of gapless bogolyubov modes . for finite @xmath0
, we may define an entropy for a given @xmath0-particle system at the critical point by counting all states with @xmath190 . given the gap of the light modes , @xmath191
, we obtain @xmath192 states within the accountable range of energies .
this implies an entropy @xmath193 . we may attempt to reproduce an entropy that scales with @xmath0 like the bekenstein - hawking entropy of black holes , @xmath194 , by increasing the number of pseudo - goldstone modes at the critical point
this can be achieved , for example , by taking the coupling to be momentum - dependent . in order to see this ,
consider a hamiltonian , @xmath195 where @xmath196 , @xmath30 is the @xmath31-dimensional volume and @xmath32 is the @xmath31-dimensional wave - number vector .
@xmath197 is a fundamental length and sets the cutoff of the effective theory . rescaling the hamiltonian
, we can write @xmath37 , where @xmath198 and @xmath199 .
let us again find the effective bilinear hamiltonian for @xmath200 modes , about the point @xmath201 and @xmath202 .
we obtain @xmath203 thus , at @xmath204 , all modes are critical and there is the same number of massless pseudo - goldstones as the number of momentum modes .
this peculiar behavior can be equivalently understood in a mean field analysis .
minimizing eq . under the constraint of fixed total particle number yields the gross - pitaevskii equation @xmath205 where @xmath53 is again the lagrange multiplier fixing the total particle number @xmath206 . using the ansatz @xmath207 , eq . becomes @xmath208 hence plane waves of arbitrary wavenumber solve the gp equation . by inserting @xmath209 into the hamiltonian , we read off the energy of the plane wave solutions @xmath210 at the critical point , all plane wave solutions become degenerate ; the corresponding modes @xmath211 are the bogolyubov modes . in order to give a lower bound on the entropy of this system for a given @xmath0 , we have to keep in mind that for the validity of the non - relativistic treatment we should only include modes with @xmath212 .
their number is @xmath213 .
this number sets the number of legitimate pseudo - goldstone modes
. however , not all the massless - pseudo - goldstone modes contribute the same weight into the entropy .
this is because their self - couplings have different strengths .
hence exciting different pseudo - goldstone species will contribute into the hamiltonian differently . in order to identify which species give the maximal contribution into the entropy of the system , let us consider the effective hamiltonian for the pseudo - goldstone modes .
it has the following form ( we ignore numerical factors of order one ) , @xmath214 where @xmath215 is the occupation number of pseudo - goldstone of a given momentum number @xmath32 . at the critical point ,
the mass terms vanish and the hamiltonian is given by the quartic couplings , @xmath216 we can obtain a lower bound for the number of states in the allowed range by looking for the @xmath217 that provides the maximal contribution to the partition sum . we need to take into account two competing effects : the larger @xmath218 , the more modes , and thus states , are supplied . on the other hand , the energy cost of a given mode grows with @xmath218 , thereby limiting the maximum allowed occupation .
moreover , we pay a penalty for occupying modes more than once as seen from the quadratic dependence of on @xmath219 .
a conservative lower bound is thus obtained by summing over all states in a shell around the limiting momentum @xmath220 which allows for all goldstone modes of given wavenumber to be occupied exactly once .
lower momenta will yield subdominant contributions due to the reduced number of modes and the related penalty for higher occupation ; higher momentum modes , on the other hand , are too costly to excite .
we obtain the contribution , @xmath221 where we ignored a possible cancellation from the cross - couplings .
the factor @xmath222 comes from the total number of modes .
the maximal contribution to the entropy will come from the species with largest @xmath217 , subject to the condition @xmath223 .
taking into the account that @xmath204 , and assuming @xmath224 , we obtain @xmath225 for the limiting momentum and the corresponding number of modes . we get the following contribution to the entropy from the pseudo - goldstone species , @xmath226 note that we have also evaluated the full partition sum numerically and thereby verified that the wavenumbers with @xmath227 provide the dominant contributions .
if we take into the account the effect of cross - couplings , ignoring possible `` flat directions '' on which the different contributions cancel , we get a more conservative lower bound on the entropy .
the number of cross - couplings for a given momentum level @xmath168 scales as the number of corresponding pseudo - goldstones squared , @xmath228 .
then , equation ( [ variation ] ) changes to @xmath229 this lowers the allowed number of simultaneously - excitable goldstones to @xmath230 independently , we observe the general message that the derivative self - coupling of bosons leads to a dramatic increase of pseudo - goldstone modes .
we finally wish to address the following question . which of the large number of degenerate states are good candidates for carrying the bekenstein - type entropy of a black hole ? in order to answer this question , we shall distinguish two categories of states .
_ 1 ) _ the first category represents states that _ can not _ be resolved semi - classically in any macroscopic measurement .
that is , the quantum information stored in such states becomes unreadable in the infinite-@xmath0 limit .
we shall refer to this category of states as type-@xmath231 .
_ 2 ) _ the second category are states that can be resolved in some macroscopic interference experiments .
that is , the quantum information stored in such states can be read out even for @xmath232 .
we shall correspondingly refer to these states as type-@xmath233 .
what distinguishes these two categories of states microscopically ?
in our picture the states of both categories can be labeled by the occupation numbers of some nearly - gapless quantum degrees of freedom .
as discussed above , these information - carriers can be described as bogolyubov and/or goldstone degrees of freedom .
what distinguishes the two category of states is the scaling behavior of information - carrier occupation numbers in large @xmath0 limit .
the type-@xmath231 category of states refers to those in which the relative occupation number of gapless degrees of freedom vanishes in the large-@xmath0 limit .
that is , none of these information - carrier degrees of freedom are macroscopically occupied .
correspondingly , the type-@xmath231 states can not be resolved in any macroscopic measurements . for type-@xmath233 states ,
some of the goldstone modes can be macroscopically occupied with an occupation number that scales as a non - vanishing fraction of @xmath0 in the @xmath119 limit .
such states can be resolved in macroscopic interference experiments , we connect to the classical limit by performing the counting of states in a coherent state basis . in this basis
, one can immediately write down the matrix elements that describe macroscopic interference experiments capable of resolving differences between the different type-@xmath233 states . ] .
this discussion is more relevant for the derivatively - coupled model ( [ hderivative ] ) , since it exhibits a large diversity of goldstone modes .
however , to keep the discussion as simple as possible , we shall here consider only the model ( [ hnonderivative ] ) .
therein , let us now exemplify which states can be attributed to the type - a and type - b categories . as discussed at the end of section [ sec : critgs ] , at the critical point , two gapless modes emerge .
the appearance of these states is a direct consequence of quantum criticality .
they come from those components of the @xmath59-doublet pseudo - goldstone that are not affected by the explicit breaking .
one of them is only an approximate goldstone , corresponding to one linear superposition of the off - diagonal generators from the quotient @xmath234 .
this goldstone is massless only near the critical point .
in contrast , the other degree of freedom is an exact goldstone corresponding to the spontaneously broken @xmath73-generator @xmath235 of @xmath45 .
this goldstone is massless everywhere at and beyond the critical point .
a counting of states in the goldstone as well as in the bogolyubov language suggests that the number of independent states below the @xmath17 energy gap near the critical point scales at least as @xmath236 .
however , the total number of almost orthogonal states with expectation value of energy below the same gap becomes larger if we also include coherent states in the counting ( see appendix ) .
this is because the coherent states are neither energy nor number eigenstates and involve superpositions of arbitrary energetic modes with arbitrarily large occupation numbers .
correspondingly , they explore a much bigger fraction of the hilbert space than the states constructed out of a finite number of energy or number eigenstates .
only the states with vanishing fractions of goldstone ( or bogolyubov ) occupation numbers belong to the type-@xmath231 category . on the other hand ,
states with macroscopic occupation numbers of goldstones , such as coherent states , belong to the type-@xmath233 category .
the latter states can be resolved in macroscopic measurements , even in @xmath232 limit .
this is a manifestation of the fact that the coherent states are classical " .
given the fact that interference experiments with black holes have never been performed , both of these categories are very interesting , since they both reveal the internal microscopic structure of black holes .
however , the question is whether both of the types of states should be counted as carriers of black hole bekenstein entropy .
_ should bekenstein entropy originate exclusively from the quantum states with small - occupation numbers of many species of gapless bogolyubov / goldstone degrees of freedom , or should the states with large occupation numbers of few species also count ? _ for now , we shall keep both options open .
the purpose of this paper was to emphasize the fundamental role of quantum criticality for information - processing in many - body systems and to establish this phenomenon as a viable candidate for the underlying mechanism for black hole information - processing , along the lines of @xcite .
we showed that the simplest multi - particle systems with critical behavior capture qualitative features of information - processing that is expected for black holes , such as the low energy cost of information storage , large degeneracy of states and scrambling of information . in order to visualize the nature of the critical phase transition in the language of spontaneous symmetry breaking
, we developed a new description in which we mapped the quantum phase transition in attractive bose - gas on a goldstone phenomenon in a sigma model .
the two systems represent two realizations of one and same unitary symmetry that rotates different momentum modes into each other .
this symmetry is broken both spontaneously as well as explicitly , but the explicit breaking vanishes , up to @xmath17 effects , at the critical point , resulting onto the gapless pseudo - goldstone modes .
this mapping allows us to establish a correspondence between the quantum criticality in bose - einstein systems and the change of symmetry breaking pattern in a sigma - model .
correspondingly , we map the entropy - carrying ( nearly)gapless bogolyubov modes of the bose - gas on the ( nearly)gapless goldstone modes of the sigma model
. this mapping also allows to formulate , up to @xmath17-corrections , an effective theory of nearly gapless modes in the regime in which occupation numbers are large .
our findings , within the validity - domain of the description , confirm the results of previous studies @xcite and shed light on the criticality phenomenon from a novel angle .
our studies indicate that the key information - processing properties of black holes are shared by a larger class of the critical systems , including the ones that can be designed in table - top labs .
this opens up a possibility of `` borrowing '' black hole information processing abilities for implementing them in the laboratory systems .
since our findings suggest that critical instability is the key for efficient information processing , it would be interesting to find out , by generalizing our approach , whether other unstable systems , for example , the ones with parametric resonance instabilities @xcite and large classical statistical fluctuations , studied in @xcite , also exhibit some analogous properties of information scrambling and processing .
we would like to thank daniel flassig and mischa panchenko for discussions .
g.d . would like to thank the organizers of the we - heraeus - seminar : cold atoms meet quantum field theory " , where a part of this work was presented , and co - participants for stimulating discussions and comments .
the work of g.d . was supported by the humboldt foundation under alexander von humboldt professorship , the erc advanced grant uv - completion through bose - einstein condensation ( grant no .
339169 ) and by the dfg cluster of excellence
origin and structure of the universe " , fpa 2009 - 07908 , cpan ( csd2007 - 00042 ) and hephacosp - esp00346 . the work of c.g . was supported in part by the humboldt foundation and by grants : fpa 2009 - 07908 , cpan ( csd2007 - 00042 ) and by the erc advanced grant 339169 `` selfcompletion '' .
the work of a.f . was supported by the fct through the grant sfrh / bd / 77473 / 2011 .
the work of n.w . was supported by the swedish research council ( vr ) through the oskar klein centre .
in order to highlight the difference between the above two types of states , it is useful to perform the state - counting in a coherent - state basis . we shall do this on the simplest example of system ( [ hamilton ] ) .
generalization to the derivatively - coupled case is straightforward .
the reason for choosing a coherent - state basis is due to its usefulness for establishing the contact with the classical limit .
since the characteristic property of type-@xmath233 states is precisely their distinguishability in the classical limit , the coherent state picture is more convenient for their description . in particular , in this basis we can immediately write down the matrix elements that describe macroscopic interference experiments capable of resolving differences between the different type-@xmath233 states .
let us evaluate the expectation value of the three - level hamiltonian ( [ hamilton ] ) over a set of coherent states with the fixed total average number @xmath0 .
these states can be labeled by parameter @xmath103 and the three phases @xmath237 respectively .
we shall denote them by @xmath238 .
each such state represents a tensor product @xmath239 of three coherent states , one per each momentum level @xmath152 . these have the following generic form , @xmath240 where @xmath241 .
the mean occupation numbers satisfy @xmath242 and @xmath243 .
the integers @xmath244 label the fock space occupation numbers for each of the three levels . taking the expectation value of the hamiltonian and imposing the minimum constraint on the phases ( [ phaseconstraint ] ) , we arrive ( not surprisingly ) at the effective hamiltonian ( [ hamiltonx ] ) .
obviously , we should only count the coherent states that form an almost - orthogonal set .
the scalar product of the two distinct coherent states @xmath245 and @xmath246 is , @xmath247 thus , the suppression of the matrix element between the two coherent states comes from the two sources : the difference in occupation numbers and difference in phases
. we shall only be interested in the set of states over which the expectation value of the hamiltonian stays within @xmath14 gap .
for such states the value of @xmath103 is bounded from above by @xmath248 .
thus , without taking into the account the variation of phases , we have maximum @xmath249 coherent states that are orthogonal due to difference in the occupation number @xmath103 . in addition
, there is a degeneracy due to variation of the phases . for a fixed @xmath103
, we can perform the two independent variations of phases that preserve the minimum constraint ( [ phaseconstraint ] ) . however , the step of the variation of each independent phase in coherent state counting must be larger than its quantum uncertainty . from the properties of the coherent states
the phases satisfy the quantum uncertainty @xmath250 .
thus , the step in variation of each phase can not be smaller than this uncertainty . for the clarity of counting we should distinguish variation of the different phases .
notice that since the phases are constrained by the minimum condition ( [ phaseconstraint ] ) , there are only two independent phase degrees of freedom , which can be chosen as @xmath251 and @xmath252 .
the states of interest can thus be labeled as @xmath253 .
let us first fix the phase @xmath251 and only vary the phase @xmath254 .
the quantum uncertainly in this phase is @xmath255 and therefore for given @xmath103 there are of order @xmath256 independent coherent states that are orthogonal to each other . this set of states is obtained by variation of the relative phase @xmath254 with the elementary step bounded from below by the minimal quantum uncertainty @xmath257 . now summing over different values of @xmath103 over an interval @xmath258
, we get that , for fixed @xmath251 , the total number of orthogonal coherent states crowded within the @xmath17 energy gap , is given by @xmath259 .
the story with the @xmath251 phase is somewhat different .
this phase is a goldstone boson corresponding to the generator @xmath260 , which is spontaneously broken already in the under - critical regime @xmath261 .
correspondingly , the coherent states that are orthogonal due to variation of @xmath251 exist irrespective of criticality .
the number of such orthogonal states for arbitrary value of @xmath103 is @xmath262 .
of course , they continue to contribute into the degeneracy also at the critical point , but this extra degeneracy has nothing to do with quantum criticality .
d. flassig , a. pritzel and n. wintergerst , `` black holes and quantumness on macroscopic scales , '' phys . rev .
d * 87 * , 084007 ( 2013 ) [ arxiv:1212.3344 ] .
g. dvali , d. flassig , c. gomez , a. pritzel and n. wintergerst , `` scrambling in the black hole portrait , '' phys .
d * 88 * , no .
12 , 124041 ( 2013 ) [ arxiv:1307.3458 [ hep - th ] ] .
foit , n. wintergerst , self - similar evaporation and collapse in the quantum portrait of black holes " , arxiv:1504.04384 [ hep - th ] g. dvali , c. gomez , r.s .
isermann , d. lst , s. stieberger , black hole formation and classicalization in ultra - planckian @xmath263 scattering " , nucl .
b893 ( 2015 ) 187 - 235 , arxiv:1409.7405 [ hep - th ] g. dvali and c. gomez , `` black hole s 1/n hair , '' phys .
b * 719 * ( 2013 ) 419 [ arxiv:1203.6575 [ hep - th ] ] ; g. dvali , c. gomez and d. lst , `` black hole quantum mechanics in the presence of species , '' fortsch . phys .
* 61 * ( 2013 )
768 [ arxiv:1206.2365 [ hep - th ] ] ; g. veneziano , `` quantum hair and the string - black hole correspondence , '' class . quant .
grav . * 30 * ( 2013 ) 092001 [ arxiv:1212.2606 [ hep - th ] ] ; p. binetruy , `` vacuum energy , holography and a quantum portrait of the visible universe , '' arxiv:1208.4645 [ gr - qc ] ; f. berkhahn , s. muller , f. niedermann and r. schneider , `` microscopic picture of non - relativistic classicalons , '' jcap * 1308 * , 028 ( 2013 ) [ arxiv:1302.6581 [ hep - th ] ] .
r. casadio , `` charged shells and elementary particles , '' int .
j. mod .
a * 28 * , 1350088 ( 2013 ) [ arxiv:1303.1274 [ gr - qc ] ] .
r. casadio and a. orlandi , `` quantum harmonic black holes , '' jhep * 1308 * , 025 ( 2013 ) [ arxiv:1302.7138 [ hep - th ] ] ; r. casadio , a. giugno , o. micu and a. orlandi , `` black holes as self - sustained quantum states , and hawking radiation , '' arxiv:1405.4192 [ hep - th ] ; w. mck , `` counting photons in static electric and magnetic fields , '' eur .
j. c * 73 * , 2679 ( 2013 ) [ arxiv:1310.6909 [ hep - th ] ] ; r. casadio , o. micu and f. scardigli , `` quantum hoop conjecture : black hole formation by particle collisions , '' phys .
b * 732 * , 105 ( 2014 ) [ arxiv:1311.5698 [ hep - th ] ] ; f. kuhnel , `` bose - einstein condensates with derivative and long - range interactions as set - ups for analog black holes , '' arxiv:1312.2977 [ gr - qc ] ; f. kuhnel and b. sundborg , `` modified bose - einstein condensate black holes in d dimensions , ''
arxiv:1401.6067 [ hep - th ] ; m. chen and y.c . huang , `` gas model of gravitons with light speed , '' arxiv:1402.5767 [ gr - qc ] ; s. hofmann and t. rug , `` a quantum bound state description of black holes , '' arxiv:1403.3224 [ hep - th ] ; f. kuhnel and b. sundborg , `` decay of graviton condensates and their generalizations in arbitrary dimensions , '' arxiv:1405.2083 [ hep - th ] ; r. casadio , a. giugno , o. micu and a. orlandi , `` black holes as self - sustained quantum states , and hawking radiation , '' arxiv:1405.4192 [ hep - th ] ; r. casadio , o. micu and p. nicolini,``minimum length effects in black hole physics , '' arxiv:1405.1692 [ hep - th ] . l. gruending , s. hofmann , s. mller and t. rug , `` probing the constituent structure of black holes , '' arxiv:1407.1051 [ hep - th ] .
w. mck , on the number of soft quanta in classical field configurations , canadian journal of physics " , 2014 , 92(9 ) : 973 - 975 , arxiv:1306.6245 [ hep - th ] r. casadio , a. giugno , o. micu and a. orlandi , `` black holes as self - sustained quantum states , and hawking radiation , '' arxiv:1405.4192 [ hep - th ] ; r. casadio , a. giugno , a. orlandi , thermal corpuscular black holes " , phys.rev .
d91 ( 2015 ) 124069 , arxiv:1504.05356 [ gr - qc ] | j. berges , a. rothkopf , j. schmidt , non - thermal fixed points : effective weak - coupling for strongly correlated systems far from equilibrium " , phys.rev.lett.101 ( 2008 ) 041603 , arxiv:0803.0131 [ hep - ph ] . | we establish a fundamental connection between quantum criticality of a many - body system , such as bose - einstein condensates , and its capacity of information - storage and processing . for deriving the effective theory of modes in the vicinity of the quantum critical point
we develop a new method by mapping a bose - einstein condensate of @xmath0-particles onto a sigma model with a continuous global ( pseudo)symmetry that mixes bosons of different momenta .
the bogolyubov modes of the condensate are mapped onto the goldstone modes of the sigma model , which become gapless at the critical point .
these gapless goldstone modes are the quantum carriers of information and entropy . analyzing their effective theory
, we observe the information - processing properties strikingly similar to the ones predicted by the black hole portrait .
the energy cost per qubit of information - storage vanishes in the large-@xmath0 limit and the total information - storage capacity increases with @xmath0 either exponentially or as a power law .
the longevity of information - storage also increases with @xmath0 , whereas the scrambling time in the over - critical regime is controlled by the lyapunov exponent and scales logarithmically with @xmath0 .
this connection reveals that the origin of black hole information storage lies in the quantum criticality of the graviton bose - gas , and that much simpler systems that can be manufactured in table - top experiments can exhibit very similar information - processing dynamics . |
microscope technology is used in various fields due to its : ( 1 ) high resolution , ( 2 ) extensive field of view , and ( 3 ) three - dimensional observation .
an optical microscope can improve the resolution power using a lens with high magnification .
however , the field of view becomes narrow .
it is theoretically difficult to solve this problem with an optical microscope .
digital holographic microscopy ( dhm ) attempts to solve this problem using a principle different to that of the optical microscope@xcite .
dhm introduces holographic technology into microscope technology , giving both a high resolution power and extensive field of view , for example ref.@xcite .
furthermore , dhm enables observation of an object three - dimensionally .
although the optical system of dhm is large and expensive , low - cost dhm systems have been reported @xcite . both studies used a digital camera for consumers as the recording device of a hologram . in ref.@xcite ,
the optical system was constructed with a digital camera made by canon , a laser diode , and some lenses .
the cost was approximately $ 1,000 , and it realized a lateral resolution power of 2.8 @xmath1 m .
the system was about 35 cm in size in the optical axis direction .
ref.@xcite is the first paper to suggest using a digital camera for consumers to make low - cost dhm .
however , the system was not conducive to handheld size , and the cost was slightly higher .
in addition , real - time reconstruction was difficult because the system used a digital camera for consumers@xcite . in our study , we succeeded in reducing the cost and the size of dhm .
the cost is less than 20,000 yen ( approximately 240 us dollars at 83 yen / dollar ) , and it is approximately 120 mm @xmath0 80 mm @xmath0 55 mm in size . in addition , by adjusting the recording distance of a hologram , the lateral resolution power at the most suitable distance is 17.5 @xmath1 m .
therefore , this dhm is useful in fieldwork , education , and so forth .
in addition , we can obtain a reconstruction image in real - time by using a web camera and graphics processing unit ( gpu ) acceleration@xcite .
this paper is structured as follows .
section 2 , describes the recording and reconstruction method of a hologram .
section 3 and 4 explain the hardware configuration and the software used .
section 5 evaluates this system .
the final section discusses the study results and proposes further studies .
this study aims to achieve price reduction and downsizing of dhm .
therefore , we adopted the in - line type dhm as the recording setup of holograms , which does not need optical parts , such as lenses@xcite .
the conceptual figure of the in - line type dhm is shown in figure [ fig : in - line.eps ] .
thus , there is no need to divide the object light and reference light unlike in mach - zehnder type dhm , which reduces the cost of lenses .
next , we explain the reconstruction technique . in dhm , computer simulation performs reconstruction from a hologram . here , in order to pursue real - time reconstruction , it is necessary to change the formula of optical propagation into a form that is easy to control by computer .
the angular spectrum method , fresnel diffraction , fraunhofer diffraction , and so forth are approximation techniques .
these are properly used according to conditions , such as propagation distance . in order to realize the downsizing of dhm , which is one of the purposes of this study
, it is necessary to shorten the distance between the object and the hologram side ( charge coupled device ( ccd ) camera ) .
therefore , we adopted the angular spectrum method , which is effective when the distance of the object and the hologram side is short .
the angular spectrum method is shown below .
@xmath2 } , \nonumber\\ & = & \int \int ^{\infty}_{-\infty } a(\xi , \eta ) \exp \bigl\{-2 \pi i ( \xi f_{x } + \eta f_{y } ) \bigr\ } d \xi d \eta.\\ u(f_{x } , f_{y } ) & = & a(f_{x } , f_{y } ) \exp ( -2 \pi i f_{z }
d).\\ u(x , y ) & = & { \mathcal{f}^{-1 } \biggl[u(f_{x } , f_{y } ) \biggl ] } , \nonumber\\ & = & \int \int^{\infty}_{-\infty } u(f_{x } , f_{y } ) \exp \bigl\{2 \pi i ( x f_{x } + y f_{y } ) \bigr\ } d f_{x } d f_{y}.\end{aligned}\ ] ] here , the operators @xmath3}$ ] and @xmath4}$ ] indicate fourier and inverse fourier transforms , and the subscripts @xmath5 , and @xmath6 indicate the horizontal , vertical , and depth components .
( @xmath7 , @xmath8 ) and ( @xmath9 , @xmath10 ) are the coordinates of the hologram side and the observation side , respectively .
@xmath11 is the distance between the hologram side and the observation side , and @xmath12 is the wavelength of the light .
@xmath13 , @xmath14 and @xmath15 are the spatial frequency components of each axis , and @xmath15 is shown as follows .
@xmath16 if we calculate using the angular spectrum method , discrete fourier transform ( dft ) is calculated on a large scale and at high - speed by fast fourier transform ( fft ) .
therefore , it is possible to perform recording of a hologram and reconstruction by computer simulation , and if we use a gpu , the reconstruction is obtained in real - time .
in order to realize low - cost and downsizing of dhm , we build hardware as shown in figure [ fig : dhm - fig.eps ] .
we used a web camera , hd pro webcam c910 made by logicool , as the ccd camera and removed the lens of the web camera for focusing an image on the ccd camera because the lens is not required for dhm application .
the resolution for photography of this web camera is 1920@xmath01080 pixels . using a commercial web camera ,
the cost can be sharply reduced .
the ostcxbc1c1s , which is a high power rgb- light emitting diode ( led ) made by optosupply , is used as the light source .
this led features the ability to allow a 350 ma current to be passed through it .
another feature is having three wavelengths , 470 nm , 525 nm , and 625 nm .
the light of this led is used as the point light source passing through the pinhole .
the diameter of the pinhole is 5@xmath1 m ( the product number of the pinhole made by koyo is 2412 ) .
although the pinhole is generally positioned on the focal side of an object lens , in our study , the perfect lensless optical system is used instead of an object lens for cost reduction . as a substitute
, a pinhole is directly attached to the led . moreover , in order to adjust the distance of the light source , an object , and a ccd camera , variable length supports are used .
the actual dhm system is shown in figure [ fig : hardware.eps ] . when recording a hologram , the actual dhm system of figure [ fig : hardware.eps ] is housed in a shading case .
software for recording a hologram and the reconstruction from a hologram is required .
opencv @xcite , which is the open source library by which intel is developed and released , is used to record the hologram . using opencv , control of a web camera and editing of a recorded hologram can be performed .
moreover , it is necessary to calculate the diffraction of light for reconstruction from a hologram .
therefore , the cwo++ library , which is an open source library for optical calculation , is used@xcite .
the cwo++ library is equipped with several commands used when optical calculation is performed .
therefore , calculation of the angle spectrum method adopted as this research can calculate on cpu or gpu by a simple description .
thus , use of the opencv and cwo++ library is cost free , enables simplification of programming and improvement in the calculation speed .
\(a ) ( b )
the costs of the parts used are shown in table [ tab : cost - parts ] .
as shown in table [ tab : cost - parts ] , it is constituted for less than 20,000 yen ( approximately 250 us dollars at 80 yen / dollar ) .
moreover , the size of dhm is equal to the size of the shading case , and it is 120 mm @xmath0 80 mm @xmath0 55 mm .
.parts used and costs for the dhm system . [ cols="^,^",options="header " , ] [ tab : cost - parts ] in order to measure the lateral resolution power of the dhm system created in this study , the usaf1951 test pattern was used .
in addition , a blue light with the shortest wavelength was used as the light source . to increase the lateral resolution power , we adjusted the positions of the ccd camera and the object using the variable length support . in this system ,
the lateral resolution power was highest when the distance from the light source to the object was 3 mm and the distance from the light source to the ccd camera was 13 mm .
the hologram recorded in this condition is shown in figure [ fig : hologram.eps ] .
furthermore , the reconstruction image obtained from figure [ fig : hologram.eps ] is shown in figure [ fig : reconstruction.eps ] .
figure [ fig : hologram.eps ] and figure [ fig : reconstruction.eps ] show that those have resolved element number 5 of group number 4 . as mentioned above
, the lateral resolution power of this dhm is 17.5 @xmath1 m .
in this study , we reduced the cost of dhm to less than 20,000 yen ( approximately 250 us dollars at 80 yen / dollar ) , and reduced the size to 120 mm @xmath0 80 mm @xmath0 55 mm .
the lateral resolution power of this dhm was 17.5 @xmath1 m .
the downsizing of dhm is useful for fieldwork , and the reduction in the cost of dhm and use of an open source library are of great benefit in educational environments .
moreover , since the led light source of rgb is used in this study , various wavelengths can be controlled .
this is applicable to wavefront recovery using two - wave digital holography@xcite , two - wave laplacian reconstruction@xcite , or two - wave laplacian reconstruction@xcite . in this study , since the angular spectrum method was used , a reconstruction image could be obtained in the same resolution as a hologram . the scaled angular spectrum method @xcite that can calculate diffraction at different sampling rates on a hologram and reconstructed image is useful for this dhm system because it enables observation of a detailed reconstructed - image with a smaller sampling rate on the reconstructed image .
adding such a function of dhm and increasing convenience are future study considerations .
99 u. schnars and w. juptner , `` direct recording of holograms by a ccd target and numerical reconstruction , '' appl.opt .
* 33 * , 179181 ( 1994 ) .
t. g. dimiduk , e. a. kosheleva , d. kaz , r. mcgorty , e. j. gardel , and v. n. manoharan , `` a simple , inexpensive holographic microscope , '' in digital holography and three - dimensional imaging , osa technical digest ( cd ) ( optical society of america , 2010 ) , paper jma38 .
t. shimobaba , j. weng , t. sakurai , n. okada , t. nishitsuji , n. takada , a. shiraki , n. masuda and t. ito , `` computational wave optics library for c++ : cwo++ library , '' comput .
. commun . * 183 * , 11241138 ( 2012 ) .
t. shimobaba , n. masuda and t. ito , `` numerical investigation of holographic wavefront retrieval using laplacian reconstruction , '' digital holography and three - dimensional imaging(dh ) dh2012 * jm3a.51*(2012 ) .
t. shimobaba , n. masuda and t. ito , `` angular spectrum method for different sampling rates on source and destination planes : scaled angular spectrum method , '' digital holography and three - dimensional imaging(dh ) dh2012 * dtu2c.3 * ( 2012 ) . | this study developed handheld and low - cost digital holographic microscopy ( dhm ) by adopting an in - line type hologram , a webcam , a high power rgb light emitting diode ( led ) , and a pinhole .
it cost less than 20,000 yen ( approximately 250 us dollars at 80 yen / dollar ) , and was approximately 120 mm @xmath0 80 mm @xmath0 55 mm in size .
in addition , by adjusting the recording - distance of a hologram , the lateral resolution power at the most suitable distance was 17.5 @xmath1 m .
furthermore , this dhm was developed for use in open source libraries , and is therefore low - cost and can be easily developed by anyone . in this research , it is the feature to cut down cost and size and to improve the lateral resolution power further rather than existing reports .
this dhm will be a useful application in fieldwork , education , and so forth .
keywords dhm , handheld , low - cost , web camera , high power rgb led |
quantum cosmology studies the relation between the observed universe and its boundary conditions in the hope that a natural _ theory _ of the boundary condition might emerge ( see @xcite for an outstanding review of this enterprise . )
assessment of a particular theory requires an understanding of its implications for the present day . to that end , this paper elaborates on work of gell - mann and hartle @xcite and davies and twamley @xcite by examining the observable consequences for the diffuse extragalactic background radiation ( egbr ) of one possible class of boundary conditions , those that are imposed time symmetrically at the beginning and end of a closed universe @xcite , and sketches some of the considerable difficulties in rendering this kind of model credible . assuming such difficulties do not vitiate the consistency of time symmetric boundary conditions as a description of our universe , the principal conclusion is that these boundary conditions imply that the bath of diffuse optical radiation from extragalactic sources be at least twice that due only to the galaxies to our past , and possibly much more . in this sense ,
observations of the egbr are observations of the final boundary condition .
this conclusion will be seen to follow ( section [ sec : limit ] ) because radiation from the present epoch can propagate largely unabsorbed until the universe begins to recollapse ( @xcite , and section [ sec : opacity ] ) , even if the lifetime of the universe is very great . by time symmetry ,
light correlated with the thermodynamically reversed galaxies of the recollapsing phase must exist at the present epoch .
minimal _ predicted excess " egbr in a universe with time symmetric boundary conditions turns out to be consistent with present observations ( section [ sec : observations ] ) , but improved observations and modeling of galactic evolution will soon constrain this minimal prediction very tightly .
in addition , many physical complications with the ansatz that time symmetric boundary conditions provide a reasonable and consistent description of the observed universe will become apparent .
thus this work may be viewed as outlining some reasons why even if very long - lived , our universe is probably not time symmetric .
the plan of the paper is as follows .
section [ sec : motivations ] discusses a model universe that will define the terms of the investigation .
section [ sec : tsbc ] provides some perspective on doing physics with boundary conditions at two times with an eye toward section [ sec : difficulties ] , where some aspects of the reasonableness of two time boundary conditions not immediately related to the extragalactic background radiation are discussed .
section [ sec : opacity ] generalizes and confirms work of davies and twamley @xcite in showing that for processes of practical interest , our future light cone ( flc ) is transparent all the way to the recollapsing era over a wide range of frequencies , even if the universe is arbitrarily long - lived .
section [ sec : limit ] explains why this fact implies a contribution to the optical extragalactic background radiation in a universe with time symmetric boundary conditions in excess of that expected without time symmetry .
in the course of this explanation , some rather serious difficulties will emerge in the attempt to reconcile time symmetric boundary conditions , and a transparent future light cone , into a consistent model of the universe which resembles the one in which we live .
section [ sec : observations ] compares the predictions of section [ sec : limit ] for the optical egbr to models of the extragalactic background light and observations of it .
section [ sec : summation ] is reserved for summation and conclusions .
the possibility that the universe may be time symmetric has been raised by a number of authors @xcite .
of course , what is meant is not _ exact _ time symmetry , in the sense that a long time from now there will be another earth where everything happens backwards .
rather , the idea is that the various observed arrows of time " are directly correlated with the expansion of the universe , consequently reversing themselves during a recontracting phase if the universe is closed .
of central interest is the thermodynamic arrow of entropy increase , from which other time arrows , such as the psychological arrow of perceived time or the arrow defined by the retardation of radiation , are thought to flow ( * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* are some reviews ) .
however , the mere reversal of the universal expansion is insufficient to reverse the direction in which entropy increases @xcite . in order to construct a quantum physics for matter in a recollapsing universe
in which the thermodynamic arrow naturally reverses itself , it appears necessary to employ something like the time neutral generalization of quantum mechanics @xcite in which boundary conditions are imposed near both the big bang _ and _ the big crunch .
these boundary conditions take the form of initial " and final " density operators which , when cpt - reverses of one another , define what is meant here by a time symmetric universe./({\rm tr}{\rho_{\alpha}}{\rho_{\omega}})$ ] , with @xmath0 @xcite . ] in such a model the collection of quantum mechanical histories is time symmetric in the sense that each history in a decohering set ( _ i.e .
_ a set of histories in which relative probabilities may be consistently assigned ) occurs with the same probability as its cpt - reverse @xcite .
( as cp violation is small , there is for many purposes no difference between cpt- and t - symmetry with a t invariant hamiltonian . )
i do not describe these ideas in further detail because very little of the _ formalism _ of generalized quantum theory will be directly applied in this paper , but it is worth mentioning that in order for the resulting time symmetric quantum theory to have non - trivial predictions , the initial and final density operators must not represent pure states .
interesting theories therefore have boundary conditions which are quantum statistical ensembles .
the interest in applying this class of quantum theories ( namely , theories with cpt - related boundary conditions ) in cosmology lies in the idea that the manifest arrow of time we observe is an emergent property of the universe , and not built directly into its structure by asymmetric dynamical laws or an asymmetric choice of boundary conditions .
dynamical laws are believed to be cpt - symmetric , so an asymmetric choice of boundary conditions is usually cited as the explanation for the existence of a definite arrow of time which flows in the same direction throughout the observable part of spacetime @xcite .
however , it is worth investigating whether this assumption is _ required _ of us by making the alternate , apparently natural ansatz that the boundary conditions on a closed universe are ( in a relevant sense ) equivalent at the beginning and end of time , instead of the more usual assumption that the initial condition is somehow special and the final condition one of indifference , " _
i.e. _ determined entirely by the past .
another point of view is that , as noted in @xcite , these alternative choices are in some sense opposite extremes .
it is therefore of interest to determine whether they are distinguishable on experimental grounds , employing time symmetric boundary conditions as a laboratory for testing the sensitivity of physical predictions to the presence of a final boundary condition . for the benefit of those eager to proceed to the definite physical predictions of sections [ sec : opacity ] and [ sec : egbr ] ,
i now specify a model in which they might be expected to arise .
( the cautiousness of this statement is explained in the sequel . )
sections [ sec : tsbc ] , [ sec : amplifications ] , and [ sec : difficulties ] elaborate on the physics expected in a universe with two - time , time symmetric boundary conditions ; here i merely summarize what is required from those sections for a complete statement of the assumptions .
the model of the universe considered here consists in : * a fixed closed , homogeneous and isotropic background spacetime , _ viz . _
a @xmath1 friedmann - robertson - walker ( frw ) universe .
the evolution of the scale factor is determined from einstein s equations by the averaged matter content of the universe .
( inhomogeneities in the matter content can be treated as additional matter fields on this background . )
* boundary conditions imposed on the matter content through a canonical decoherence functional @xmath2 @xcite with cpt - related density operators @xmath3 that describe the state of matter at some small fiducial scale factor , near what would in the absence of quantum gravitational effects be the big bang and big crunch , but outside of the quantum gravity regime .
the matter state described by one of these density operators reflects the presumed state of the early universe , namely matter fields in apparent thermal equilibrium at the temperature appropriate to the fiducial scale factor and the total amount of matter in the universe .
spatial fluctuations should be consistent with present day large scale structure , say being approximately scale invariant and leading to an amplitude of order @xmath4 at decoupling in order to be consistent with recent cobe results @xcite .
possible further conditions on @xmath3 are discussed at the end of this subsection .
the essence that a choice of boundary conditions with these apparently natural characterisics intends to capture is that of a universe in which the cosmological principle holds , which is smooth ( and in apparent thermal equilibrium ) whenever it is small , and which displays more or less familiar behaviour when larger .
most of the conclusions of this investigation really only rely on these general properties , but for the sake of definiteness a fairly specific model which has the right general physical characteristics is offered .
however , as will be repeatedly emphasized in the sequel , in models with boundary conditions at two times , not only does the past have implications for the future , but the future has implications for the past .
therefore , in any attempt to model the _ observed _ universe with time symmetric boundary conditions , we need to make sure it is possible to integrate them into a self - consistent picture of the universe as we see it today .
this means in particular that we must be prepared for the possibility that the early universe in a time symmetric universe may have properties different from those expected in a universe with an initial condition only .
the model boundary conditions sketched above are not intended to be so restrictive as to rule out such differences , and hence must be taken with a grain of salt or two .
the issue is then whether or not these properties are consistent with observation .
indeed , the prediction of an excess " optical egbr ( to be discussed in section [ sec : egbr ] ) in a universe with time symmetric boundary conditions is precisely of this character .
( of course , the most extreme possibility is that a universe burdened with these boundary conditions would look _ nothing at all _ like the universe in which we actually live . )
variations on this theme will recur frequently in the following sections . with these boundary conditions
@xcite , the time neutral generalized quantum mechanics of gell - mann and hartle defines an effective quantum theory for matter in the universe which may be imagined to arise from some other , more fundamental , quantum cosmological theory of the boundary conditions .
( presumably , the fundamental quantum cosmological decoherence functional incorporates the quantum mechanics of the gravitational field as well , a complication that is not addressed in this paper . ) in this connection it is perhaps worth mentioning that it was once claimed @xcite that the no - boundary proposal for the initial condition ( * ? ? ?
* @xcite is an excellent review ) implied just this sort of effective theory in that it appeared to require that the universe be smooth whenever it was small .
thus a fundamental theory of an initial condition _ only _ apparently could be decribed by an effective theory with _ two - time _
boundary conditions .
however , this claim has since been recanted @xcite due to a mathematical oversight .
more generally , the no - boundary wave function does not appear to be a good candidate for a boundary condition imposed time symmetrically at both ends of a closed universe because it is a pure state @xcite , which as noted above yield quantum theories in which essentially all physical information is encoded in the boundary conditions alone .
for contrast , see @xcite , in which it is asserted that the only sensible way to interpret the wheeler - dewitt equation necessitates a boundary condition requiring the universe to be smooth whenever it is small . in order to allow definite predictions to be made
, the final key assumption of the model is that , for a suitable class of physically interesting coarse grainings ( _ e.g . _
coarse grainings defining the domain of classical experience , or local quantum mechanics experiments ) , the probabilities for such coarse grained physical histories unfold near either boundary condition ( relative to the total lifetime of the universe ) in a fashion insensitive to the presence of the boundary condition at the other end , _
i.e. _ as if the other boundary condition were merely the identity . as discussed in section [ sec : tsbc ] , simple stochastic models @xcite suggest that this holds for processes for which the relaxation time " of the process to equilibrium is short compared to the total time between the imposition of the boundary conditions ( see section [ sec : tsbc ] ) , which expectation is rigorously supported in the case of markov processes @xcite .
therefore , for the sake of brevity this predictive ansatz shall often be referred to as the relaxation time hypothesis " ( rth ) .
a more careful statement of the rth would identify the classes of coarse grained histories ( presumably at least those decribing short relaxation time processes at times sufficiently close to , for instance , the big bang ) for which conditional probabilities ( when defined ) are , in a universe with boundary conditions @xmath5 , @xmath6 , supposed to be close to those of a universe with @xmath7 .
it would also define close " and short relaxation time process " more precisely .
( thus , a rigorous statement of what is meant by the rth requires a definite mathematical model . because the concerns of this investigation encompass a variety of complicated physical processes , with the entire universe considered as a single physical system
, i do not attempt that here .
in specific cases the intuitive content of the ansatz ought to be clear enough . ) in order to exploit the rth to its fullest , i shall , when convenient , assume also that the universe is close to the critical density , so that its lifetime is very long .
this plausibly realistic assumption maximizes the possibility that a model of the kind given above accurately describes our universe . with these assumptions
, this model universe might be expected to closely resemble the universe as it is observed today _ if _ most familiar and important physical processes are examples of such `` short relaxation time processes . '' in particular , under the assumption that they are , predictions we expect of a single initial condition @xmath5 only ( _ i.e . _
@xmath8 ) can ( by the rth ) be assumed to hold near the initial condition in the model with cpt - related initial and final conditions .
such predictions should include those regarding inflation , relative particle abundances , and the formation of large ( and small ) scale structure .
because of the cpt - related boundary condition at the big crunch , a qualitatively similar state of affairs is then expected in the recollapsing era , but cpt reversed . as the thermodynamic arrow is caused fundamentally by gravitational collapse driving initially smooth matter away from equilibrium ( * ? ?
? * ; * ? ? ?
* for example ) , the arrow of entropy increase near the big crunch will run in the opposite direction to that near the big bang .
, [ sec : amplifications ] , and [ sec : difficulties ] . ]
observers on planets in the recollapsing phase will find their situation indistinguishable from our own , with all time arrows aligned in the direction of increasing volume of the universe .
it is this interesting ( if unconventional ) state of affairs which leads to the conclusion that observations of the egbr can reveal the presence of a final boundary condition that is cpt - related to the initial , even if the lifetime of the universe is very great .
how this comes about is the topic of section [ sec : egbr ] . before proceeding ,
some points made in the preceeding paragraph require qualification .
( readers whose only interest is the extragalactic background radiation should procede directly to section [ sec : opacity ] . this discussion , and that of the next subsection , are positioned here for the sake of unity and perspective .
) first , the relaxation times " for many important physical processes in a recollapsing universe _ do not _ appear to be short compared to the lifetime of the universe , even if that lifetime is arbitrarily long .
the result described in section [ sec : opacity ] , that light from the present epoch can propagate unabsorbed into the recollapsing phase , is an excellent example of this @xcite
. some physical difficulties with the consistency of time symmetry to which this gives rise will become apparent in section [ sec : amplifications ] .
further examples of such physical difficulties relating to issues such as gravitational collapse , the consequent emergence of a familiar thermodynamic arrow , and baryon decay in a universe with time symmetric boundary conditions are discussed briefly in section [ sec : difficulties ] .
the self - consistency of such models is thus in doubt , with these kinds of complications constituting arguments against the possibility that our universe possesses time symmetric boundary conditions .
that is , it appears that time symmetry is _ not _ consistent with the central predictive ansatz that physics near either boundary condition is practically insensitive to the presence of the other , and it appears likely that a universe with time symmetric , low ( matter ) entropy boundary conditions would look _ nothing like _ the universe in which we actually live .
nevertheless , the strategy of sections [ sec : opacity ] and [ sec : egbr ] is to assume that the time symmetric picture is consistent with the gross characteristics of the observed universe and see what it predicts . because of the prediction of a diffuse optical extragalactic background radiation that is testably different from that in a universe which is not time symmetric , we are provided with a two pronged attack on the hypothesis of time symmetric boundary conditions as a realistic description of our universe : the lack of a self - consistent picture of the observed universe , and observations of the egbr .
second , for completeness it should be mentioned that in the context of the time neutral generalized quantum mechanics assumed here , there are further restrictions on the viability of time symmetric models as a realistic description of our universe which arise from the requirements of decoherence and the emergence of approximately classical behaviour .. ] these topics are discussed by gell - mann and hartle @xcite .
third , it is important to note that the conclusions of section [ sec : egbr ] regarding the egbr depend essentially on the assumed _ global _ homogeneity and isotropy , _ i.e. _ the cosmological principle . "
thus , while inflation ( apparently a generic consequence of quantum field theory in a small universe ) can be taken to be a prediction of the assumed boundary conditions if they are imposed when the universe is sufficiently small , the popular point of view ( somewhat suspect anyway ) that inflation provides an _ explanation _ of homogeneity and isotropy inside the horizon is not a prediction of the effective theory of the universe considered here ; it is an assumption .
however , the other good things inflation does for us would still qualify as predictions .
one further point requires mention . in the conventional picture of inflation ,
the matter in the universe is in a pure state ( say , some vacuum state ) when small .
after inflation and reheating the matter fields appear to be , according to local coarse grainings , in thermal equilibrium , the states of different fields being highly entangled .
furthermore , the correlations required to infer that the complete state is actually pure have been inflated away .
nevertheless , the quantum state is of course still pure @xcite .
as noted previously , pure states are not viable candidates for the initial and final conditions .
therefore , in order to fit the conventional picture of inflation into an effective theory of the kind considered in this paper , the @xmath5 and @xmath6 defined above must be a local description in the sense that they not contain the information required to know that the state from which they were inferred was actually pure , _ i.e. _ they must coarse - grain this information away .
( the alternative is to employ an inflationary scenario in which the universe is not required to be in a pure state . )
the unpleasantness of this restriction could be taken as an argument against the use of cpt - related boundary conditions in a fundamental quantum cosmological theory . finally , for the cognoscenti of generalized quantum mechanics ( * ? ? ?
* for example ) , there is a related observation that is even more interesting .
field theoretic hamiltonians are cpt - invariant , so that if the cpt - related @xmath5 and @xmath6 depend only on the hamiltonian ( as for instance an exactly thermal density operator does ) , @xmath5 and @xmath6 are actually _ equal_. ( the same observation holds for merely t - related boundary conditions if the hamiltonian is t invariant .
however , cpt seems the more relevant symmetry if the hamiltonian and boundary conditions of the considered effective theory of the universe arise from some more fundamental theory . )
this is a potential difficulty for the sample model given above .
a simple extension of a result of gell - mann and hartle @xcite to the case where the background spacetime is an expanding universe shows that the only dynamics allowed with identical initial and final boundary conditions is trivial : for alternatives allowed in sets of histories in which probabilities may be assigned at all , the time dependence of probabilities for alternative outcomes is essentially _ independent of the hamiltonian _ , and is due only to the expansion of the universe . for time
independent hamiltonians , time dependence of probabilities for such projections arises only if the projections are time - dependent in the schrdinger picture . due to the expansion , this will be the case for many projections onto ( otherwise time - independent ) quantities of physical interest in an expanding universe . ] in the present case this may be interpreted as a prediction that a universe which is required to be in thermal equilibrium whenever it is small _ must remain so when large , _ at least in the context of the theoretical structure of gell - mann and hartle .
( classically the expectation is the same , so this result can hardly be written off as an artifact of the formalism . )
now , the evidence suggests that matter in the the early universe _ was _ in local thermal equilibrium .
however , the inhomogeneities in the matter required to generate large scale structure must also be described by the boundary conditions , and , even in the usual case where there is only an initial condition , it is after all gravitational condensation which drives the appearance of a thermodynamic arrow . in the present case ( the model with the two - time boundary conditions characterized above ) ,
the meaning of these observations is that one way to break the cpt invariance of the boundary conditions is for the specification of the deviations from perfect homogeneity and isotropy to be cpt non - invariant , in order that the theory admit interesting dynamics . as a specific example ,
i briefly consider the common circumstance where cosmological matter is given a hydrodynamic description .
scalar - type adiabatic pertubations may be completely specified by , _
e.g. , _ the values of the energy density perturbation and its ( conformal- ) time derivative over a surface of constant conformal time @xcite .
( scalar - type pertubations are the ones of interest , as they are the ones which may exhibit instability to collapse . )
as it is intended here that gravity is treated classically , such a specification might come in the form of a probability distribution for approximately scale invariant metric and energy density fluctuations averaged over macroscopic scales .
if this probability distribution reflects the underlying frw homogeneity and isotropy ( and is thus in particular p - invariant ) , in order to break t - invariance the distribution must distinguish between opposing signs of the ( conformal- ) time derivatives of the energy density perturbations .
an alternative way to break the equality of cpt - related boundary conditions is for the locally observed matter - antimatter asymmetry to extend across the entire surface of constant universal time .
( the recollapsing era will then be antimatter dominated in a cpt symmetric universe , consequently requiring that there be some mechanism to permit baryon decay . )
this possibility may appear more attractive , but it is not without significant complications .
these are addressed briefly at the end of section [ sec : amplifications ] . as noted above
, the time neutral generalized quantum mechanics of gell - mann and hartle @xcite with cpt - related initial and final density operators yields a cpt symmetric ensemble of quantum mechanical histories , in the sense that each history in the collection is accompanied with equal probability by its cpt - reverse .
the _ collection _ of histories ( in each decohering set ) is therefore statistically time symmetric ( sts ) .
now , a set of histories can be statistically time symmetric without any individual history possessing qualitatively similar ( if time - reversed ) characteristics near both ends .
completely time asymmetric histories can constitute a statistically time symmetric set if each history in the collection is accompanied by its time reverse @xcite . with two - time boundary conditions ,
however , more is required .
that is , there is a distinction between a cpt invariant _ set _ of histories , in which each history in a decohering set is accompanied , with equal probability , by its cpt reverse , and the considerably more restrictive notion of a statistically time symmetric universe studied here . in virtue of the boundary conditions at both the beginning and the end of time
, probable histories ( in a set which decoheres with these boundary conditions ) will in general have both initial and final states which to some extent resemble the boundary conditions @xmath3 @xcite . perhaps the clearest way to understand
this is to consider the construction of a classical statistical ensemble with boundary conditions at two times .
the probability for each history in the ensemble may be found in the following way @xcite .
the boundary conditions are given in the form of phase space probability densities for the initial and final states of the system .
pick some initial state .
evolve it forward to the final time .
weight this history by the product of the probabilities that it meets the initial and final conditions , and divide by a normalizing constant so that all the probabilities sum to one .
thus , roughly speaking , in order for a history to be probable in this two - time ensemble both its initial and final state must be probable according to the initial and final probability densities , respectively .
( for a deterministic classical system , two - time statistical boundary conditions are equivalent to a boundary condition at one time constructed in the obvious way . in quantum mechanics
this is no longer true due to the non - commutability of operators at different times .
thus , while the described algorithm is merely a useful heuristic for understanding the implications of two - time boundary conditions _ classically _ , it is more essential quantum - mechanically . )
what is not so clear is that the resulting probable histories look anything like the probable histories in the ensemble with just , say , the initial condition . in general they will not .
classically , this is merely the statement that the probability measure on a space of classical deterministic solutions defined by two - time boundary conditions will in general be quite different than the measure defined by the initial probability density only . for classical systems with stochastic dynamics or for quantum statistical systems , the simple models studied by cocke @xcite , schulman @xcite , and others yield some insight into what is required for evolution near either boundary condition to be influenced only by that boundary condition .
here i merely summarize the intuitively transparent results of this work in the context of time symmetric boundary conditions .
in particular , the emphasis will be on boundary conditions which represent low entropy " initial and final states . in the absence of a final condition
, an initially low entropy state generally implies that evolution away from the initial state displays a
thermodynamic arrow " of entropy increase ( relative to the coarse graining defining the relevant notion of entropy ; see @xcite for a pertinent overview . ) in an ensemble with two - time , time symmetric , low entropy boundary conditions , under what conditions will a familiar thermodynamic arrow " appear near either boundary condition ?
a pertinent observation is that , roughly speaking , in equilibrium all arrows of time disappear .
therefore , if the total time between the imposition of the boundary conditions is much longer than the relaxation time " of the system to equilibrium ( defined as the characteristic time a similar system with an initial condition only takes to relax ) , it might be expected that entropy will increase in the familiar fashion away from either boundary condition . that is , following the coarse - grained evolution of the system forward from the initial condition , entropy increases at the expected rate until the system achieves equilibrium .
the system languishes in equilibrium for a time , and then entropy begins to decrease again until the entropy reaches the low value demanded by the final condition .
thus , this is a system in which the thermodynamic arrow of time " reverses itself , which reversal is enforced by the time symmetric , low entropy boundary conditions .
there is no cause to worry about the coexistence of opposed thermodynamic arrows .
histories for which entropy increases away from either boundary condition are readily compatible with both boundary conditions , and the probable evolutions are those in which , _ near either boundary condition _ , there is a familiar thermodynamic arrow of entropy increase away from the nearest boundary condition .
there , the other boundary condition is effectively invisible .
essentially this is because in equilibrium all states compatible with constraints are equally probable ; the system forgets " its boundary conditions .
on the other hand , if there is not time enough for equilibrium to be reached there must be some reconciliation between the differing arrows of increasing entropy @xcite . as only histories which satisfy the required boundary conditions
are allowed , it can be anticipated that the statistics of physical processes that would ordinarily ( in the absence of the final boundary condition ) lead to equilibration would be affected because histories which were probable with an initial condition only are no longer compatible with the final condition .
the rate of entropy increase is slowed , and in fact the entropy may never achieve its maximum value .
in other words , the approach to equilibrium is suppressed by the necessity of complying with the low entropy boundary condition at the other end .
more probable in this two - time ensemble is that the state _ will continue to resemble the low entropy boundary conditions . _
the entire evolution is sensitive to the presence of _ both _ boundary conditions .
in fact , as noted already , simple models of statistical systems with a boundary condition of low entropy at two times bear out these intuitively transparent expectations @xcite .
moreover , in the case of markov processes schulman has demonstrated the described behaviour _ analytically _ @xcite .
this confluence of intuitive clarity , and , for simple systems , analytic and ( computerized ) experimental evidence will be taken as suggestive that the significantly more complicated physical system considered in this work ( matter fields in a dynamic universe ) behaves in a qualitatively similar fashion when burdened with time symmetric boundary conditions .
( which arise essentially as a result of the requirements of decoherence ) are examples of phenomena with no counterpart in stochastic systems .
see also the discussion of gravitational collapse in section [ sec : collapse ] . ]
the lesson of this section is that the place to look for signs of statistical time symmetry in our universe is in physical processes with long relaxation times , " or more generally in any process which might couple the expanding and recollapsing eras @xcite .
such processes might be constrained by the presence of a final boundary condition .
obvious candidates include decays of long - lived metastable states @xcite , gravitational collapse ( * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* see also @xcite ) and radiations @xcite with great penetrating power such as neutrinos , gravitational waves , and possibly electromagnetic radiation .
thus , there are a variety of tests a cosmological model with both an initial and a final condition must pass in order to provide a plausible description of our universe .
the next two sections discuss the extragalactic background radiation as an example of a physical prediction that is expected to be sensitive to the presence of a final boundary condition ( time symmetrically related to the initial . )
i focus on electromagnetic radiation because it is the most within our present observational and theoretical grasp , but the essence of the discussion is relevant to any similar wave phenomenon . finally , in section [ sec : difficulties ] i offer a few comments on some other issues that need to be addressed in any attempt to describe our universe by a model such as the one sketched in section [ sec : motivations ] .
the aim of this section is to extend arguments of davies and twamley @xcite showing that a photon propagating in intergalactic space is likely to survive until the epoch of maximum expansion ( assuming the universe to be closed ) , no matter how long the total lifetime of the universe .
that is , the future light cone is essentially transparent over a wide range of frequencies for extinction processes relevant in the intergalactic medium ( igm ) , with an optical depth of at most @xmath9 at optical frequencies .
the physics in this result is that , in cases of physical interest , the dilution of scatterers due to the expansion of the universe wins out over the extremely long path length the photon must traverse . as a consequence , the integrated background of light from galaxies in the expanding phase will still be present in the recollapsing phase .
( as explained in section [ sec : egbr ] , it is this fact which implies that there is an excess " egbr in a time symmetric universe that is not associated with galaxies to our past . ) to show this , i compute , for a fairly general class of absorption coefficients , the optical depth between the present epoch and the moment of maximum expansion . for realistic
intergalactic extinction processes this optical depth turns out to be small .
indeed , the opacity of the future light cone turns out to be dominated by collisions " of intergalactic photons with other galaxies , if they are regarded as completely opaque hard spheres . even in the limit that the lifetime of the universe @xmath10 becomes infinite ( @xmath11 from above ) all of the relevant processes yield finite optical depths .
as this is essentially the limit of a flat universe , it is no surprise that extremely simple expressions result .
the results of this section are modest extensions of the work of davies and twamley @xcite . for the intergalactic extinction mechanisms and at the frequencies they consider , the formulae for the opacity derived here give numbers in agreement with their results ( using the same data for the igm , of course . ) the present work is of slightly broader applicability in that the opacity is evaluated for a fairly general class of frequency dependent extinction coefficients ( not just for a few specific processes ) , and its behaviour as the lifetime of the universe becomes arbitrarily long is determined .
it turns out that quite generally , the asymptotic limit is in fact of the order of magnitude of the _ upper _ limit to the opacity in a closed universe .
these results are less general than that of @xcite in that i do not include the effects of a cosmological constant ( which makes the perturbative analysis below significantly more awkward . ) however , a small cosmological constant does not effect the qualitative nature of the conclusions of this section .
optical depth @xmath12 is defined by @xmath13 where @xmath14 , the linear extinction coefficient , is the fractional loss of flux per unit ( proper ) length @xmath15 , and is given microscopically by @xmath16 for incoherent scattering from targets with cross section @xmath17 and proper number density @xmath18 ( this neglects stimulated emission and scattering into the line of sight . )
given @xmath12 , the flux density along the line of sight thus obeys @xmath19 put another way , the probability a photon will propagate a distance @xmath15 without being absorbed is @xmath20 . for further details ,
see for example @xcite . in order to compute @xmath12
we need @xmath21 . in the approximation ( appropriate to the calculation of optical depths between the present and the moment of maximum expansion ) that the universe is exactly described by
closed , dust - filled friedmann - robertson - walker , it turns out to be helpful to trade in the dependence on proper length for time .
the metric is @xmath22 , \ ] ] where @xmath23 is the metric on the unit 3-sphere , and @xmath24 relates the cosmological time ( proper time in the cosmological rest frame ) to the conformal time @xmath25 . for dust the time dependence of the scale factor @xmath26
can be expressed parametrically as @xmath27 so that @xmath28 the lifetime of the universe is then @xmath29 . here
@xmath30 is a constant as the universe expands , @xmath31 being the mass density of the dust .
@xmath32 is related to more familiar cosmological parameters by @xmath33 employing the symmetry of the model to take a photon s path as radial , @xmath34 gives @xmath35 , from which @xmath36 the high symmetry of friedmann - roberston - walker also means that nearly all the relevant physical quantities simply scale as a power of @xmath26 .
thus , as will be seen explicitly in the next subsection , it is necessary to consider only extinction coefficients of the form @xmath37 where @xmath38 is some fiducial scale factor ( conventionally the present one ) and @xmath39 is a number .
the goal is to compute the optical depth between now and the moment of maximum expansion .
now " will be taken to be the time @xmath40 from the big bang to the present .
for absorption coefficients of the form ( [ eq : sigma= ] ) , the optical depth of the future light cone is @xmath41 using ( [ eq : a= ] ) and ( [ eq : dtau= ] ) . here
@xmath42 is the conformal time of the present epoch , and @xmath43 is the dimensionless function @xmath44 for integral and half - integral @xmath39 explicit evaluation of @xmath43 is possible , but not terribly illuminating . in the limit
that the total lifetime of the universe @xmath10 is very long compared to @xmath40 , however , simple expressions for any @xmath39 result .
( this is no surprise as the results must approach those of a flat universe . )
one straightforward procedure involves inverting ( [ eq : t= ] ) to get a power series in @xmath45 for @xmath42 , and using this to evaluate the asymptotic behaviour of @xmath43 as @xmath32 becomes large relative to @xmath40 , which is held fixed .
( @xmath46 corresponds to @xmath47 . ) it is then tedious but straightforward to show that @xmath48{ll } \frac{3}{2p-1 } \left [ 1+\frac{p+1}{10(2p-3 ) } { \left ( \frac{6t_{0}}{m } \right ) } ^{\frac{2}{3 } } \right ] & \mbox{$ p > \frac{1}{2 } \ ; \ p \not= \frac{3}{2 } $ } \\
\frac{3}{2 } \left [ 1-\frac{1}{12}{\left ( \frac{6t_{0}}{m } \right)}^{\frac{2}{3 } } \ln\left(\frac{6t_{0}}{m}\right ) \right ] & \mbox{$p = \frac{3}{2}$}\\
\ln\left(\frac{m}{t_{0}}\right ) \left [ 1-\frac{3}{40}{\left(\frac{6t_{0}}{m}\right ) } ^{\frac{2}{3 } } \right ] & \mbox{$p=\frac{1}{2 } $ } \\ 3\pi \frac{\gamma ( 1 - 2p)}{\gamma^{2 } ( 1-p ) } { \left ( \frac{m}{6t_{0}}\right ) } ^{\frac{1 - 2p}{3 } } - \frac{3}{1 - 2p } &
\mbox{$-\frac{1}{2 } < p < \frac{1}{2}$ } \\ 12{\left ( \frac{m}{6t_{0}}\right ) } ^{\frac{2}{3 } } - \frac{9}{5 } & \mbox{$ p = -\frac{1}{2}$ } \\ 3\pi \frac{\gamma ( 1 - 2p)}{\gamma^{2 } ( 1-p ) } { \left(\frac{m}{6t_{0}}\right ) } ^{\frac{1 - 2p}{3 } } \left [ 1-\frac{p+1}{20}{\left ( \frac{6t_{0}}{m}\right ) } ^{\frac{2}{3 } } \right ] & \mbox{$ p < -\frac{1}{2}$ } \end{array } \right .
\label{eq : g}\ ] ] in each case only the leading order correction in @xmath49 has been retained .
the most important thing to notice is that for @xmath50 , @xmath51 is perfectly finite even as the lifetime of the universe becomes arbitrarily big , and as @xmath49 becomes very small the opacity converges to the value it would have in a flat universe .
it is clear that for @xmath52 the @xmath53 result is a local lower limit on the opacity of the future light cone , as may be verified directly also from the available exact results .
however , for reasonable @xmath39 the corrections to the flat universe result are only a factor of order one .
( in fact , examination of the exact results reveals that the _ maximum _ of @xmath51 as one varies @xmath49 is at most @xmath54 larger than the flat universe result for @xmath55 .
this is a good thing , because unless @xmath56 is fairly close to one , @xmath49 is not a particularly small parameter ! in terms of familiar cosmological parameters , @xmath57 . )
\label{eq : t / m}\ ] ] physically , what s at work is the competition between the slower expansion rates of universes with larger @xmath58 s , which tends to increase the opacity because the scattering medium is nt diluted as rapidly , and the decrease in the opacity due to the shortened time between the present epoch and the moment of maximum expansion . ) and ( [ eq : t / m ] ) that taking the limit @xmath59 holding @xmath40 fixed requires @xmath60 to vary as well , converging to the flat universe relation @xmath61 .
it is possible to repeat the entire analysis holding the observable quantity @xmath60 fixed instead of @xmath40 ( for this purpose the more standard redshift representation is more useful than that in terms of conformal time used above ) , but unsurprisingly the conclusions are the same : the opacity is always finite for @xmath50 ; as @xmath58 approaches one , the opacity approaches the flat universe result ; and the _ maximum _ opacity for reasonable @xmath39 is only a factor @xmath62 times the flat universe result .
the resultant opacities are of course related in these limits via @xmath63 .
similarly , it is possible to perform a related analysis of more complicated extinction coefficients than ( [ eq : sigma= ] ) , for example incorporating the exponential behaviour encountered in free - free absorption ( see ( [ eq : sff ] ) ) or in modeling evolving populations of scatterers with , for instance , a schecter function type profile .
however , these embellishments are not required in the sequel , and the techniques are tedious and fairly ordinary , so space will not be taken to describe them here . ] to summarize , all of the processes relevant to extinction in the intergalactic medium have extinction coefficients that can be bounded above by a coefficient of the form ( [ eq : sigma= ] ) with @xmath50 . using the limiting relationship @xmath63
, we have from ( [ eq : tau= ] ) and ( [ eq : g ] ) the simple result that for these processes , the upper limit to the opacity between the present epoch and the moment of maximum expansion , no matter how long the total lifetime of the universe , is of order @xmath64 ( i have returned to conventional units in this formula . ) in this section i apply the asymptotic formula ( [ eq : tau ] ) for the upper limit to the optical depth of the flc in a long - lived universe to show that if our universe is closed , photons escaping from the galaxy are ( depending on their frequency ) likely to survive into the recollapsing era .
that is , the finite optical depths computed in the previous section are actually small for processes of interest in the intergalactic medium ( igm ) .
for simplicity , i focus on photons softer than the ultraviolet at the present epoch ; the cosmological redshift makes it necessary to consider absorption down to very low frequencies .
it is important to note that in employing standard techniques for computing opacities the effects of the assumed statistical time symmetry of the universe are being neglected . as discussed in section [ sec : tsbc ] and in section [ sec : difficulties ] , when the universe is very large the thermodynamic and gravitational behaviour of matter will begin to deviate from that expected were the universe not time symmetric . due to the manifold uncertainties involved here it is difficult to approach the effects of time symmetric boundary conditions on the opacity of the future light cone with clarity . ) .
this complication is closely related to the difficulty , mentioned in section [ sec : retardation ] , in deriving the retardation of radiation in a universe which is time symmetric and in which the future light cone is transparent . ]
i shall assume they are not such as to increase it .
this is reasonable as the dominant contribution to the opacity comes when the universe is smallest , where in spite of the noted complications the rth is assumed to hold .
what are the processes relevant to extinction of photons in the intergalactic medium ? because the igm appears to consist in hot , diffuse electrons , and perhaps a little dust @xcite , extinction processes to include are thomson scattering , inverse bremsstrahlung ( free - free absorption ) , and absorption by dust .
in addition , absorption by material in galaxies ( treated as completely black in order to gauge an upper limit ) is important .
these processes will treated in turn .
( a useful general reference on all these matters is @xcite . )
the conclusion will be that while absorption by galaxies and thomson scattering are most significant above the radio , none of these processes pose a serious threat to a photon that escapes from our galaxy .
this confirms the results of davies and twamley @xcite , who however did not consider the possibly significant interactions with galaxies .
consequently i will be brief .
some results of davies and twamley regarding absorption mechanisms which may be important when the universe is very large and baryons have had time to decay are quoted at the end of this section .
these do not appear to be significant either .
( for high energy photons compton scattering , pair production , photoelectric absorption by the apparently very small amounts of neutral intergalactic hydrogen , and interactions with cmbr photons will be important , but as none are significant below the ultraviolet i do not discuss them here .
all can be treated by the same methods as the lower energy processes . ) to begin , following davies and twamley @xcite , i quote barcons _
et al . _
@xcite on current beliefs regarding the state of the igm in the form @xmath65 where @xmath66 with the values @xmath67 somewhat preferred by the authors .
in addition , the present upper limit on a smoothly distributed component of neutral hydrogen is about @xmath68 .
thus , the intergalactic medium consists in hot ( but non - relativistic ) electrons , protons , and essentially no neutral hydrogen .
the lack of distortions in the microwave background indicates its relative uniformity , at least to our past . from now on , @xmath18 and @xmath10
simply will be used to refer to the number density and temperature of intergalactic electrons .
finally , very little is known about a possible diffuse component of intergalactic dust @xcite , except that there is probably very little of it .
most dust seems to be clumped around galaxies .
therefore i will ignore possible extinction due to it , subsuming it into the black galaxy " opacity .
davies and twamley @xcite make some estimates for one model for the dust , finding its contribution to the opacity insignificant . at any rate , models for the absorption coefficient due to dust
@xcite all give a cross section @xmath17 that falls with increasing wavelength , @xmath69 with @xmath70 , so that @xmath71 ( neglecting of course a clumping factor expressing the fact that clumping decreases the opacity . ) thus @xmath72 , the dust opacity is bound to be finite , and with a small present density of diffuse dust it is not surprising to find its contribution to be small . before considering the optical depth due to interactions with intergalactic electrons
, i will show that it is reasonable to approximate that most photons escaping our galaxy will travel freely through intergalactic space .
that is , few photons will end up running into another galaxy . drastically overestimating the opacity due to galaxies by pretending that any photon which enters a galaxy or its halo will be absorbed by it ( the black galaxy " approximation ) , and taking the number of galaxies to be constant , @xmath73 where @xmath17 is the cross - sectional area of a typical galaxy and @xmath74 is their present number density .
thus from ( [ eq : tau ] ) , the upper limit on the opacity due to collisions with galaxies is @xmath75 as noted above , this is finite ( even as the lifetime of the universe becomes very large ) because the dilution of targets due to the expansion of the universe is more important than the length of the path the photon must traverse .
notice that assuming target galaxies to be perfectly homogeneously distributed only overestimates their black galaxy " opacity .
volume increases faster than cross - sectional area , so clustering reduces the target area for a given density of material . as galaxy clustering is not insignificant today and
will only increase up to the epoch of maximum expansion even in a time symmetric universe , the degree of overestimation is likely to be significant .
taking @xmath76 , @xmath77 ( where @xmath78 ) , and @xmath79 ( here @xmath80 captures as usual the uncertainty in the hubble constant ) gives the upper limit @xmath81 this can be interpreted as saying that at most about one percent of the lines of sight from our galaxy terminate on another galaxy before reaching the recollapsing era . by time symmetry , neither do most lines of sight connecting the present epoch to its time - reverse .
use of the thomson scattering cross section @xmath82 is acceptable for scattering from non - relativistic electrons for any photon softer than a hard x - ray ( @xmath83 ) .
thus , for the frequencies i will consider , @xmath84 will suffice , giving @xmath85 recalling that @xmath86 is at worst one order of magnitude , it is clear that thomson scattering is not signficant for intergalactic photons @xcite .
it is perhaps worth mentioning that quantum and relativistic effects only tend to decrease the cross section at higher energies . more significant for the purposes of this investigation
is the observation that , at the considered range of frequencies , thomson scattering does not change a photon s frequency , merely its direction .
thus thomson scattering of a homogeneous and isotropic bath of radiation by a homogeneous and isotropic soup of electrons has _ no effect _ as regards the predictions of section [ sec : egbr ] .
even less significant than thomson scattering for frequencies of interest is free - free absorption by the igm @xcite . from , _
@xcite , the linear absorption coefficient for scattering from a thermal bath of ionized hydrogen is @xmath87 in cgs units .
here @xmath88 , the factor @xmath89 contains the effect of stimulated emission , and @xmath90 is a gaunt factor " expressing quantum deviations from classical results .
it is a monotonically decreasing function of @xmath91 which is of order one in the optical ( _ cf . _
@xcite for a general discussion and some references . ) as @xmath92 increases as the universe expands , taking @xmath93 , a constant of order one , will only overestimate the opacity . similarly , following @xcite in dropping the stimulated emission term will yield an upper limit to the free - free opacity . with @xmath94 , @xmath95 when @xmath96 , so stimulated emission will only lead to a noticeable reduction in @xmath97 well below the optical .
( actually , methods similar to that employed in section [ sec : transparent ] can be employed to calculate this term , but as @xmath98 will turn out to be insignificant even neglecting it there is no need to go into that here . ) with these approximations , @xmath99 and thus @xmath100 recalling that @xmath101 seem likely physical values , and noting that @xmath102 at worst , taking @xmath103 is not unreasonable for an order of magnitude estimate .
thus @xmath104 ( long radio ) is required to get @xmath105 . since @xmath106 it drops sharply for photons with present frequency above that .
for instance , at 5000@xmath107 and inverse bremsstrahlung is completely negligible .
finally , i mention that davies and twamley @xcite consider what happens if baryons decay in a long lived universe . following the considerations of @xcite , they conclude that the positronium atoms " which will form far in the future ( when the universe is large ) remain transparent to photons with present frequencies in the optical .
this is because the redshifted photons havent enough energy to cause transitions between adjacent ps energy levels .
similarly , if in the nearer future the electrons and protons in the igm recombine to form more neutral hydrogen , this will also be transparent at the considered frequencies .
the goal of this section is to explain why , in a statistically time symmetric universe ( such as one with the cpt - related boundary conditions discussed in section [ sec : motivations ] ) , the optical extragalactic background radiation should be at least twice that expected in a universe which is not time symmetric , and possibly considerably more .
thus , assuming consistency with the rth ( _ i.e . _
the predictive assumption that physics near either boundary condition is practically insensitive to the presence of the other boundary condition , _ cf . _ section [ sec : tsbc ] ) , it is possible to discover _ experimentally _ whether our universe is time symmetric .
section [ sec : observations]compares this prediction with present observations , concluding that the minimal prediction is consistent with upper limits on the observed optical egbr . however , better observations and modeling may soon challenge even this minimal prediction . at optical wavelengths , the isotropic bath of radiation from sources outside our galaxy
is believed to be due almost exclusively to galaxies on our past light cone ( * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* are some good general references ) .
there is no other physically plausible source for this radiation . in a model universe with time
symmetric boundary conditions , however , there must in addition be a significant quantity of radiation correlated with the time - reversed galaxies which will exist in the recollapsing era , far to our future @xcite .
the reason for this is that light from our galaxies can propagate largely unabsorbed into the recollapsing phase no matter how close to open the universe is , as shown in @xcite and in section [ sec : opacity ] .
this light will eventually arrive on galaxies in the recollapsing phase , or , depending on its frequency , be absorbed in the time - reversed equivalent of one of the many high column density clouds ( lyman - limit clouds and damped lyman-@xmath108 systems ) present in our early universe @xcite , in the intergalactic medium , or failing that , at the time - reversed equivalent of the surface of last scattering .
this will appear to observers in the recontracting phase as emission by one of those sources sometime in their galaxy forming era . since future galaxies , up to high time - reversed redshift , occupy only a small part of the sky seen by today s ( on average ) isotropically emitting galaxies , much of the light from the galaxies of the expanding phase will proceed past the recontracting era s galaxies .
thus most of this light will be absorbed in one of the other listed media .
because of the assumption of global homogeneity and isotropy , the light from the entire history of galaxies in the expanding phase will constitute an isotropic bath of radiation to observers at the time - reverse of the present epoch that is _ in addition _ to the light from the galaxies to _ their _ past . by time symmetry
, there will be a similar contribution to our egbr correlated with galaxies which will live in the recollapsing phase , over and above that due to galaxies on our past light cone .
to us this radiation will appear to arise in isotropically distributed sources _ other _ than galaxies .
this picture of a transparent , time symmetric universe is illustrated in figure 1 .
[ fig : egbr ] a lower limit to this excess background can be obtained by considering how much light galaxies to our past have emitted already ( _ cf . _ section [ sec : observations ] ) . according to observers at the time reverse of the present epoch , this background will ( in the absence of interactions )
retain its frequency spectrum and energy density because the size of the universe is the same .
thus , by time symmetry , at a _ minimum _ the predicted optical egbr in a universe with time symmetric boundary conditions is twice that expected in a universe in which the thermodynamic arrow does not reverse .
if much of the luminous matter in galaxies today will eventually be burned into radiation by processing in stars or galactic black holes , the total background radiation correlated with galaxies in the expanding phase could be several orders of magnitude larger , a precise prediction requiring a detailed understanding of the future course of galactic evolution @xcite .
several other processes may also contribute significant excess backgrounds .
these topics are discussed further below .
a number of points in the summary argument above require amplification .
first , however , i summarize the _ minimal _ predictions for the excess " extragalactic background ( _ i.e . _ radiation from non - galactic sources to our past that is correlated with time - reversed galaxies ) in bands for which the future light cone is transparent : * isotropy : the future starlight " should appear in the comoving frame as an approximately isotropic background .
this conclusion depends crucially on the assumed global validity of the cosmological principle .
* energy density : comparable to the present energy density in starlight due to the galaxies on our past light cone .
this assumes the future light cone ( flc ) is totally transparent .
* spectrum : similar to the present spectrum of the background starlight due to galaxies on our past light cone .
again , neglect of further emissions in the expanding phase makes this , by time symmetry , a lower limit in each band .
this conclusion relies on the assumption of a transparent flc in part to the extent that this implies a paucity of standard astrophysical mechanisms for distorting spectra .
thus , at for instance optical frequencies , time symmetry requires an isotropic extragalactic background at least twice that due to galaxies on our past light cone alone .
was of order @xmath109 , mostly due to a liberal ( black galaxy " ) assessment of the rate of interception of photons by galaxies , and i will therefore neglect such losses .
further , it is worth remembering that processes like thomson scattering do not destroy photons or change their frequency , but only scatter them . thus mere scattering processes may introduce isotropically distributed ( via the cosmological principle ) fluctuations in the background , but not change its total energy .
similarly , line or dust absorption usually result in re - radiation of the absorbed photons , conserving the total energy in an isotropic background ( if the size of the universe does nt change much before the photons are re - radiated ) , if not the number of photons with a given energy . ]
the potentially far greater background predicted ( by time symmetry ) if further emissions in the expanding phase are accounted for is a subject taken up in the sequel . before proceeding ,
a comment on the consistency of this picture is in order .
as the excess " radiation is correlated with the detailed histories of future galaxies , the transparency of the future light cone does not appear consistent with the predictive assumption ( the rth ) that physics in the expanding era should be essentially independent of the specifics of what happens in the recontracting phase . at a minimum ,
if the model is to be at all believable it is legitimate to demand that the required radiation appears to us to arise in sources in a fashion consistent with known , or at least plausible , astrophysics .
thus it may be that given a transparent flc , the only viable picture of a time symmetric universe is one in which the radiation correlated with future galaxies _ should " be there anyway , i.e. _ be predicted also in some reasonable model of our universe which is _ not _ time symmetric , and consequently not be excess " radiation at all , but merely optical radiation arising in non - galactic sources during ( or before ) the galaxy forming era .
on the basis of present knowledge this does not describe our universe .
the presence of the radiation required by time symmetry and the transparency of the flc appears to be in significant disagreement with what is known about our galaxy forming era , as will become apparent below .
were it the case for our universe that non - galactic sources provided a significant component of the optical egbr , the difficulties with time symmetric boundary conditions would from a _
practical _ point of view be less severe .
it is true that the non - galactic sources emitting the additional isotropic background would have to do so in just such a way that the radiation contain the correct spatial and spectral correlations to converge on future galaxies at the appropriate rate .
this implies a distressingly detailed connection between the expanding and recontracting phases .
however , if the emission rate and spectrum were close to that expected on the basis of conventional considerations these correlations ( enforced by the time symmetric boundary conditions ) would likely be wholly unobservable in practice , existing over regions that are not causally connected until radiation from them converges onto a future galaxy @xcite , and thus not visible to local coarse grainings ( observers ) in the expanding era . in any event
, the meaning of the transparency of the flc is that starlight is by no means a short relaxation time process . " of course , on the basis of the models discussed in section [ sec : tsbc ] , perhaps the conclusion to draw from this apparent inconsistency between the rth and the transparency of the flc should rather be , that physical histories would unfold in a fashion quite different from that in a universe in which @xmath7 , namely , in such a way that such detailed correlations would never be required in the first place . the very formation of stars might be suppressed ( _ cf . _ section [ sec : difficulties ] ) .
be that as it may , to the extent that the universe to our past is well understood , there are _ no _ sources that could plausibly be responsible for an isotropic optical background comparable to that produced by galaxies .
such an additional background , required in a transparent , time symmetric universe , requires significant , observable deviations from established astrophysics .
this is in direct contradiction with the rth .
thus the entire structure in which an additional background is predicted appears to be both internally inconsistent ( in that it is inconsistent with the postulate which allows predictions to be made in the first place ) , and , completely apart from observations of the egbr discussed in section [ sec : observations ] , inconsistent with what is known about our galaxy forming era .
this should be taken as a strong argument that our universe does not possess time symmetric boundary conditions .
nevertheless , in order to arrive at this conclusion it is necessary to pursue the consequences of assuming the consistency of the model .
the situation may be stated thus : _ either _ our universe is not time symmetric , _ or
_ there is an unexpected contribution to the optical egbr due to non - galactic sources to our past ( and there are indeed detailed correlations between the expanding and recontracting eras ) , _ or _ perhaps our future light cone is not transparent after all .
this latter possibility , perhaps related to the considerable uncertainty regarding the state of the universe when it is large , seems the last resort for a consistent time symmetric model of our universe .
it is now appropriate to justify further some of the points made in arriving at the prediction of an excess contribution to the egbr .
claims requiring elaboration include : i ) most of the light from the expanding era s galaxies wo nt be absorbed by the galaxies of the recontracting phase , and _ vice - versa _ ; ii ) it will therefore be absorbed by something else , and this is inconsistent with the early universe as presently understood ; iii ) a detailed understanding of the old age and death of galaxies , as well as other processes when the universe is large , may lead to a predicted egbr in a time symmetric universe that is orders of magnitude larger than the minimal prediction outlined above .
i deal with these questions in turn .
a photon escaping our galaxy is unlikely to encounter another galaxy before it reaches the time reverse of the present epoch .
in fact , as shown in section [ sec : opacity ] , galaxies between the present epoch and its time reversed equivalent subtend , at most , roughly a mere @xmath110 of the sky ( [ eq : hs ] ) ( neglecting curvature and clumping . ) in light of the present lack of detailed information about our galaxy forming era ( and via time symmetry its time - reverse ) , a photon s fate after that is more difficult to determine . a straightforward extrapolation of the results of section [ sec : opacity ] ( or _ cf .
e.g. _ section 13 of peebles @xcite ) shows that the optical depth for encounters with galaxies of the same size and numbers as today is @xmath111 between a redshift of @xmath112 and today .
again this neglects curvature ( hence overestimating @xmath12 ) and clumping ( which now underestimates @xmath12 . ) assuming the bright parts of galaxies form at @xmath113 , this gives only @xmath114 , and the sky is nt covered with them until @xmath115 .
this however is roughly at the upper limit on how old galaxies are thought to be .
on the other hand , examination of quasar spectra ( out to @xmath113 ) show that most lines of sight pass through many clouds of high column densities of hydrogen called lyman-@xmath108 forest clouds and , at higher densities , damped lyman-@xmath108 systems .
( peebles @xcite is a useful entry point on all of these matters , as is @xcite .
@xcite are the proceedings of a recent conference concerned with these lyman systems . )
the highest density clouds may be young galaxies , but if so galaxies were more diffuse in the past as the observed rate of interception of an arbitrary line of sight with these clouds is a factor of a few or more greater than that based on the assumption that galaxy sizes are constant .
( obviously , this would not be too surprising . ) for instance , for the densest clouds peebles @xcite relates the approximate formula @xmath116 for the observed interception rate per unit redshift of a line of sight with a cloud of column density greater than or equal to @xmath117 ( in units of @xmath118 ) , in a range of redshifts about @xmath119 .
for lyman-@xmath108 forest clouds , @xmath120 , the interception rate is considerably higher .
( for some models see @xcite . )
thus an arbitrary line of sight arriving on our galaxy from a redshift of five , say , is likely to have passed through at least one cloud of column density comparable to a galaxy , and certainly many clouds of lower density . what might this mean for time symmetry ?
( for specificity i shall concentrate on photons which are optical today , say around 5000 .
this band was chosen because at these wavelengths we have the luxury of the coincidence of decent observations , relatively well understood theoretical predictions for the background due to galaxies , the absence of other plausible sources for significant contributions , and a respectable understanding of the intergalactic opacity , including in particular some confidence that the future light cone is transparent . )
photons at 5000 today are at the lyman limit ( 912 ) at @xmath121 , and so are ionizing before that . at these redshifts the bounds on the amount of smoothly distributed neutral hydrogen ( determined by independent measures such as the gunn - peterson test @xcite ) are very low ( _ cf .
_ section [ sec : flc ] ) , presumably because that part of the hydrogen formed at recombination which had not been swept into forming galaxies was ionized by their radiation . before the galactic engines condensed and heated up ,
however , this neutral hydrogen would have been very opaque to ionizing radiation .
similarly , near @xmath122 lyman - limit clouds with @xmath123 are opaque at these frequencies .
the upshot is that most photons from our galaxies which are optical today will make it well past the time reverse of the present epoch , likely ending up in the ( time - reversed ) l-@xmath108 forest or in a young ( to time reversed observers ! ) galaxy by @xmath124 .
( here @xmath125 is the epoch corresponding to the time - reverse of redshift @xmath112 . )
the very few that survive longer must be absorbed in the sea of neutral hydrogen between @xmath126 and ( their ) recombination epoch , @xmath127 .
now , the important point is that on average , galaxies radiate isotropically into the full @xmath128 of sky available to them .
the lesson of the previous paragraph is that most lines of sight from galaxies in the expanding phase will not encounter a high column density cloud until a fairly high time - reversed redshift , @xmath129 , at which point many lines of sight probably _ will _ intersect one of these proto - galaxies or their more diffuse halos . if most photons from our galaxies have not been absorbed by this point , _ this is not consistent with time symmetry _ : the rate of emission of ( what is today ) optical radiation by stars in galaxies could not be time symmetric if the light of the entire history of galaxies in the expanding phase ends up only on the galaxies of the recollapsing phase at high @xmath125 ( due consideration of redshifting effects is implied , of course . )
put another way , time symmetry requires the specific energy density in the backround radiation to be time symmetric .
thus the emission rate in the expanding era must equal ( what we would call ) the absorption rate in the recontracting phase .
if stars in galaxies were exclusively both the sources ( in the expanding phase ) and sinks ( as we would call them in the recontracting phase ) of this radiation , galactic luminosities in the expanding phase would have to track the falling rate of absorption due to photon collisions " with galaxies .
this is absurd . at the present epoch , for example , _ at all frequencies
_ galaxies would ( by time symmetry ) have to be absorbing the diffuse egbr ( a rate for which the upper limit is determined entirely by geometry in the black galaxy " approximation ) at the same rate as their stars were radiating ( a rate that , in a time symmetric universe which resembles our own , one expects to be mostly determined by conventional physics . )
that is , stars would be in radiative equilibrium with the sky !
this may be called the no olber s paradox " argument against the notion that a single class of localized objects could be exclusively responsible for the egbr in a transparent universe equipped with time symmetric boundary conditions .
( it might be thought that this problem would be solved if the time symmetric boundary conditions lead galaxies to radiate preferentially in those directions in which future galaxies lie .
this is not a viable solution , because as noted above , only a small fraction of the sky is subtended by future galaxies up to high time reversed redshift .
the deviations from isotropic emission would be dramatic . ) thus , the option consistent with time symmetry is that most galactic photons which are optical today will ultimately be absorbed in the many ( by time symmetry ) time - reversed lyman-@xmath108 forest clouds or lyman - limit clouds believed to dwell between galaxies , and not in the stars of the time - reversed galaxies themselves .
fortunately for the notion of time symmetry this indeed appears to be the case .
careful studies of the opacity associated with lyman systems @xcite , indicates , within the bounds of our rather limited knowledge , that the light cone between @xmath130 and @xmath131 is essentially totally opaque to radiation that is 5000 at @xmath131 , and that this is due largely to lyman clouds near @xmath122 and in the middle range of observed column densities , @xmath132 or so .
( to be honest , it must be admitted that hard data on just such clouds is very limited @xcite . )
we have now arrived at a terrible conundrum for the notion of time symmetry .
even if one is willing to accept the amazingly detailed correlations between the expanding and recontracting eras that reconciling a transparent future light cone with time symmetry requires , and even if the excess " radiation correlated with the galaxies of the recollapsing era were to be observed , this picture is incompatible with what little is known about the physical properties of the lyman-@xmath108 forest . recalling the minimal prediction above for the excess background required by time symmetry ,
the prediction is that the lyman-@xmath108 forest has produced an amount of radiation at least comparable to that produced by the galaxies to our past .
there is no mechanism by which this is reasonable .
_ there is no energy source to provide this amount of radiation . _
more prosaically , the hydrogen plasma in which the clouds largely consist is observed ( via determination of the line shape , for example ) to be at kinetic temperatures of order @xmath133 , heated by quasars and young galaxies @xcite .
thermal bremsstrahlung is notoriously inefficient , and line radiation at these temperatures is certainly insufficient to compete with nuclear star burning in galaxies ! at for instance
5000today , essentially _ no _ radiation is expected from forest clouds at all , let alone an amount comparable to that generated by galaxies . remembering that by redshifts of 4.5 the lyman forest is essentially totally opaque shortward of 5000 ( observed ) @xcite
, it might have been imagined that an early generation of galaxies veiled by the forest heated up the clouds sufficiently for them to re - radiate the isotropic background radiation required by time symmetry . while it is true that quasars and such are likely sources of heat for these clouds ( * ? ? ?
* for example ) , aside from the considerable difficulties in getting the re - radiated spectrum to resemble that of galaxies , the observed temperatures of the clouds are entirely too low to be compatible with the _ minimum _ amount of energy emission in the bands required .
( a related restriction arises from present day observations of cosmological metallicities , which constrain the amount of star burning allowed to our past .
if observed discrete sources came close to accounting for the required quantity of heavy elements , the contribution of a class of objects veiled completely by the lyman forest would be constrained irrespective of observations of the egbr .
at present direct galaxy counts only provide about 10% of the current upper limits on the extragalactic background light @xcite ( _ cf . _ section [ sec : observations ] ) , the rest conventionally thought to arise in unresolvable galactic sources . consequently , correlating formation of the heavy elements with observed discrete sources does not at present provide a good test of time symmetry . at any rate ,
such a test is likely to be a less definitive constraint on time symmetry because it is possible that a portion of the radiation lighting the lyman forest from high redshift is not due to star burning , but to accretion onto supermassive black holes at the centers of primordial galaxies .
thus the best observational test is the most direct one , comparison of the observed egbr with the contribution expected from galaxies . )
the possibility that somehow the excess radiation does _ not _ come from the lyman-@xmath108 forest , but somehow shines through from other isotropically distributed sources even further in the past , is hardly more appealing .
familiar physics tells us that the forest is totally opaque to radiation that is 5000at @xmath131 .
the conclusion had better be that the universe is not time symmetric , rather than that time symmetry engineers a clear path only for those photons correlated with galaxies in the recollapsing epoch ( and not , say , the light from quasars . ) moreover , even if that were the case , analagous difficulties apply to the vast sea of neutral hydrogen that existed after recombination , totally opaque to ionizing radiation , and again to the highly opaque plasma which constituted the universe _ before _ recombination .
it is possible to conjure progressively more exotic scenarios which save time symmetry by placing the onus on very special boundary conditions which engineer such rescues , but this is not the way to do physics .
the only _ reasonable _ way time symmetry could be rescued would be if it were discovered that for reasons unanticipated here , the future light cone were not transparent after all , thus obviating the need for an excess background radiation with all its attendant difficulties .
otherwise , it is more reasonable to conclude that a universe with time symmetric boundary conditions would not resemble the one in which we actually live .
now that we have seen what kind of trouble time symmetry can get into with only the _ minimal _ required excess background radiation , it is time to make the problems worse .
the background radiation correlated with the galaxies of the recollapsing era was bounded from below , via time symmetry , by including only the radiation that has been emitted by the galaxies to our past .
but as our stars continue to burn , if the future light cone is indeed transparent it is possible a great deal more radiation will survive into the recollapsing era @xcite .
how much more ? to get an idea of what s possible it is necessary to know both what fraction of the baryons left in galaxies will be eventually be burned into starlight , and when . for a rough upper bound , assume that _ all _ of the matter in galaxies today , including the apparently substantial dark halos ( determined by dynamical methods to contribute roughly @xmath134 ) , will eventually be burned into radiation . to get a rough lower bound ,
assume that only the observed luminous matter ( @xmath135 ) will participate significantly , and that only a characteristic fraction of about 4% of _ that _ will not end up in remnants ( jupiters , neutron stars , white dwarfs , brown dwarfs , black holes , _ etc .
_ ) to overestimate the energy density of this background at @xmath136 , assume that all of this energy is released in a sudden burst at some redshift @xmath137 .
then by time symmetry , further star burning will yield a background of radiation correlated with time - reversed galaxies ( expressed as a fraction of the critical density and scaled to @xmath131 ) somewhere in the range @xmath138 ( here i have used the fact that the mass fraction released in nuclear burning as electromagnetic radiation is .007 . ) when @xmath139 the upper limit is two orders of magnitude more than is in the cmbr today and three orders of magnitude more than present observational upper limits on a diffuse optical extragalactic background ( _ cf .
_ section [ sec : observations ] ) .
the lower bound , however , is comparable to the amount of radiation that has already been emitted by galaxies .
thus if the lower bound obtains , the prediction for the optical egbr in a time symmetric universe is only of order three times that due to the galaxies to our past ( if the excess background inferred from continued star burning is not distributed over many decades in frequency , and if most of this burning occurs near @xmath140 . ) as will be seen in section [ sec : observations ] , this may still be consistent with present observational upper limits . on the other hand , if something closer to the upper limit obtains this is a clear death blow to time symmetry .
a more precise prediction is clearly of interest .
this would entail acquiring a detailed understanding of further galactic evolution , integrating over future emissions with due attention to the epoch at which radiation of a given frequency is emitted .
( naturally , this is the same exercise one performs in estimating the egbr due to galaxies to our past @xcite . )
some idea of the possible blueshift ( @xmath141 ) involved comes from estimating how long it will take our galaxies to burn out .
this should not be more than a factor of a few greater than the lifetime of the longest lived stars , so a reasonable ballpark figure is to assume that galaxies will live for only another ten billion years or so . for convenience ,
assume that galaxies will become dark by @xmath142 for some @xmath18 , where @xmath40 is the present age of the universe . to overestimate the blueshift at this time , assume the universe is flat , so that @xmath143 for reasonable @xmath18 s this does not amount to a large ( in order of magnitude ) transfer of energy to the radiation from cosmological recontraction . in a similar fashion to continued burning of our stars ,
any isotropic background produced to our future might by time symmetry be expected to imply an additional contribution to the egbr in an appropriately blueshifted band .
for instance , even if continued star burning does not ( by time symmetry ) yield a background in contradiction with observations of the egbr , it is possible that accretion onto the supermassive black holes likely to form at the centers of many galaxies could ultimately yield a quantity of radiation dramatically in excess of that from star burning alone . in the absence of detailed information about such possibilities
it is perhaps sufficient to note that ignoring possible additional contributions leads to a lower limit on the egbr correlated with sources in the recontracting era , and i will therefore not consider them .
there is one worrying aspect , however . as discussed in some detail by page and mckee @xcite for an approximately @xmath144 universe , and
commented on in a related context in section [ sec : collapse ] , if baryons decay then considerable photons may be produced by for instance the pair annihilation of the resulting electrons and positrons .
should not this , by time symmetry , yield a further contribution to the egbr ?
the answer may well be yes , but there is a possible mechanism which avoids this conclusion .
somehow , with cpt symmetric boundary conditions , the density of baryons must be cpt symmetric .
therefore either baryons do not decay , or they are re - created in precisely correlated collisions .
( in the absence of a final boundary condition , the interaction rate would be too low for ( anti-)baryon recombination to occur naturally . )
the latter ( boundary condition enforced ) possibility appears extraordinary , but if baryon decay occurs in a universe with cpt symmetric boundary conditions , it could be argued that the best electrons and photons for the job would be just those created during baryon decay in the expanding phase , thereby removing this photon background . the no olber s paradox " argument , that most of an isotropically emitting source s light must end up in some homogeneous medium , and not , if time symmetry is to be preserved , equivalent time - reversed point sources , may not apply here if matter is relatively homogeneously distributed when the universe is large .
baryon decay _
might _ smooth out inhomogeneities somewhat before the resulting electrons and positrons annihilate .
( this requires the kinetic energies of the decay products to be comparable to the gravitational binding energy of the relevant inhomogeneity . )
then the picture is no longer necessarily of localized sources emitting into @xmath128 , but of a more homogeneous photon - producing background that might cover enough of the sky to more reasonably secrete the required correlations for reconstruction of more massive particles in the recontracting phase .
nevertheless the extreme awkwardness of this scenario is not encouraging .
the former possibility , clearly more palatable , is that baryons do not decay significantly either because @xmath56 is not so near one after all that they have time enough to do so , or because the presence of the final boundary condition suppresses it . either way , in this ( possibly desperate ) picture there is no additional background due to decaying baryons .
a very similar question relates to the enormous number of particles produced in the last stages of black hole evaporation .
this time , however , the objection that our black holes cover only a small portion of the recontracting era s sky , and consequently their isotropic emissions could not do the job of forming the white holes of the recontracting era ( black holes to observers there ) time symmetrically , would seem to be forceful .
thus if the universe is indeed very long - lived , black hole evaporation may well require yet an additional observable background .
this may not be such a serious difficulty if @xmath145 is not very close to one , however , as the time scales for the evaporation of galactic - scale black holes are quite immense
. further discussion of black holes in time symmetric cosmologies may be had in @xcite .
one last point regarding the predictions described in this section needs to be made .
clearly , a loose end which could dramatically change the conclusions is the condition of matter in the universe when it is very large .
this is uncertain territory , not the least because that is the era in a statistically time symmetric universe when the thermodynamic arrow must begin rolling over . neglecting this confusing complication ( reasonable for some purposes as many interactions are most significant when the universe is small )
, there is not a great deal known about what the far future should look like @xcite .
the study of page and mckee @xcite gives the most detailed picture in the case of a flat universe .
as mentioned at the end of section [ sec : flc ] , davies and twamley @xcite find from this work that interactions of optical ( at @xmath131 ) backgrounds with the electrons produced by baryon decay do not appear to be significant , primarily due to their diffuseness . on the other hand ,
if supermassive black holes ( or any large gravitational inhomogeneities ) appear , interactions with them may induce anisotropies in the future starlight .
however , clumping only decreases the probability a line of sight intersects a matter distribution
. therefore large overdensities probably never subtend enough solid angle to interfere with most lines of sight to the recollapsing era unless gravitational collapse proceeds to the point where it dramatically alters homogeneity and isotropy on the largest scales .
because collapse is rather strongly constrained by time symmetry ( _ cf .
_ section [ sec : collapse ] ) i will not consider this possibility . thus , insofar as the prediction of an excess " background radiation correlated with galaxies in the recollapsing era is concerned , the state of the universe when it is large would does not obviously play a substantial role . nevertheless , given the manifold difficulties cited , the sentiment expressed above is that the best hope a time symmetric model has of providing a realistic description of our universe is that some unforseen mechanism makes the future light cone opaque after all . to summarize , because our future light cone is transparent , neglecting starburning to our future and considering only the contribution to the egbr from stars in our past provides an estimate of the total egbr correlated with galaxies in the recollapsing era that is actually a lower limit on it . as mechanisms for distorting
the spectrum generically become less important as the universe expands ( barring unforseen effects in the far future ) , it is reasonable to take models for the present egbr due to stars in our past as a minimal estimate of the isotropic background of starlight that will make its way to the recollapsing era . by time symmetry we can expect that at the same scale factor in the recollapsing era similar ( but time - reversed ) conditions obtain . as argued above , by time symmetry and geometry this
future starlight " must appear to us as an additional background emanating from homogeneously distributed sources to our past _ other _ than galaxies .
therefore , if the universe has time symmetric boundary conditions which ( more or less ) reproduce familiar physics when the universe is small , and our future light cone is transparent , the optical extragalactic background radiation should be at least twice that expected to be due to stars in our past alone , and possesses a similar spectrum .
if a considerable portion of the matter presently in galaxies will be burned into radiation in our future , by time symmetry the expected background is potentially much larger , and observations of the egbr may already be flatly incompatible with observations . nevertheless ,
in the next section i shall be conservative and stick with the minimal prediction in order to see how it jibes with observations . at optical wavelengths
, it is generally believed that the isotropic background of radiation from extragalactic sources is due entirely to the galaxies on our past light cone @xcite . as shown in the previous section ,
if our universe is time symmetric there must in addition be a significant contribution correlated with the galaxies of the recollapsing era which arises , not in galaxies , but in some homogeneously distributed medium , say for instance the lyman clouds .
the apparent inconsistency of this prediction with what is known about the forest clouds has been discussed above , and may be taken as an argument that our universe is not time symmetric . in this section judgement will be suspended , and the prediction of an excess " egbr at least comparable to , but over and above , that due to galaxies to our past will be compared with experiment .
the conclusion will be that current data are still consistent with time symmetry _ if _ our galaxies will not , in the time left before they die , emit a quantity of radiation that is considerably greater than that which they already have .
a useful resource on both the topics of this section is @xcite .
tyson @xcite has computed how much of the optical extragalactic background is accounted for by resolvable galaxies , concluding that known discrete sources contribute @xmath146 at 4500 . however , because very distant galaxies contribute most of the background radiation it is believed that unresolvable sources provide a significant portion of the egbr . at present
it is not possible to directly identify this radiation as galactic in origin .
however , as understanding of galactic evolution grows so does the ability to model the optical extragalactic background due to galaxies .
these predictions naturally depend on the adopted evolutionary models , what classes of objects are considered , the cosmological model , and so on . as representative samples i quote the results of code and welch
@xcite for a flat universe in which all galaxies evolve , @xmath147 at 5000 , and of cole _
et al . _
@xcite for a similar scenario , @xmath148 also at 5000 .
these figures are to be compared with the results of ( extraordinarily difficult ) observations . as surveyed by mattila @xcite , they give at 5000 an upper limit of @xmath149 as far as i am aware , there has been no direct detection of an optical radiation background of extragalactic origin .
this upper limit represents what is left after what can be accounted for in local sources is removed . comparing these results ,
it is clear that if current models of galactic evolution are reliable , present observations of the extragalactic background radiation leave room for a contribution from non - galactic sources that is comparable to the galactic contribution , but not a great deal more .
these observations therefore constrain the possibility that our universe is time symmetric .
if believable models indicate that further galactic emissions compete with what has been emitted so far , time symmetry could already be incompatible with experiment . a direct detection of the extragalactic background radiation , or even just a better upper limit , could rule out time symmetry on _ experimental _ grounds soon .
( the ideal observational situation would be convergence of all - sky photometry and direct hst galaxy counts , allowing one to dispense with models completely . )
in this section i comment on some issues of a more theoretical nature which must be faced in any attempt to construct a believable model of a time symmetric universe . among other questions , in such a universe the difficulties with incorporating realistic gravitational collapse , and in deriving the fact that radiation is retarded , appear to be considerable . a careful account of the growth of gravitational inhomogeneities from the very smooth conditions when the universe was small is clearly of fundamental importance in any model of the universe , time symmetric or not , not the least because it appears to be the essential origin of the thermodynamic arrow of time from which the other apparent time arrows are thought to flow @xcite .
however , if matter in a closed universe is to be smooth at both epochs of small scale factor then it is incumbent to demonstrate that einstein s equations admit solutions in which an initially smooth universe can develop interesting ( non - linear ) inhomogeneities such as galaxies which eventually smooth out again as the universe recollapses .
this is because the universe is certainly in a quasiclassical domain now , and if it is assumed to remain so whenever the scale factor is large classical general relativity must apply .
laflamme @xcite has shown that in the linear regime there are essentially no perturbations which are regular , small , and growing away from both ends of a closed frw universe , so that in order to be small at both ends a linear perturbation must be too small to ever become non - linear .
but that is not really the point .
interesting perturbations must go non - linear , and there is still no proof of which i am aware that perturbations which go non - linear can not be matched in the non - linear regime so as to allow solutions which are small near both singularities .
put differently , what is required is something like a proof that weyl curvature must increase ( _ cf . _
@xcite ) , _ i.e. _ that given some suitable energy conditions the evolution of gravitational inhomogeneity must be monotonic even in the absence of trapped surfaces , except possibly on a set of initial data of measure zero . while perhaps plausible given the attractive nature of gravity , proof has not been forthcoming .
( it is important for present purposes that genericity conditions on the initial data are not a central part of such a proof , for physically realistic solutions which meet the time symmetric boundary conditions must describe processes in the recollapsing era which from our point of view would look like galaxies disassembling themselves .
reducing such a solution to data on one spacelike hypersurface at , say , maximum expansion , said data will be highly specialized relative to solutions with galaxies which do not behave so unfamiliarly . if such solutions with physically interesting inhomogeneities do exist , the real question here is whether they are generic _ according to the measure defined by the two - time boundary conditions . _ since it ought to be possible to treat this problem classically , in principle this measure is straightforward to construct .
the practical difficulty arises in evaluating the generic behaviour of solutions once they go non - linear .
my own view is that it is highly likely that such solutions remain exceedingly improbable according to a measure defined by a _
generic _ set of ( statistical ) boundary conditions requiring the universe to be smooth when small . as noted ,
laflamme @xcite has already shown that when the initial and final states are required to be smooth , the growth of inhomogeneity is suppressed if only linear perturbations are considered .
unless boundary conditions with very special correlations built in are imposed , probable solutions should resemble their smooth initial and final states throughout the course of their evolution , never developing physically interesting inhomogeneities .
note the concern here is not with occurrences which would be deemed unlikely in a universe with an initial condition only , but occur in a time symmetric universe because of the fate " represented by the final boundary condition , but with occurrences which are unlikely _ even in a universe with ( generic ) time symmetric boundary conditions . _ ) a related question concerns collapsed objects in a time symmetric universe .
page and mckee @xcite have studied the far future of a @xmath144 frw universe under the assumption that baryons decay but electrons are stable .
assuming insensitivity to a final boundary condition their conclusions should have some relevance to the period before maximum expansion in the only slightly over - closed ( and hence very long - lived ) model universes that have mostly been considered here .
as discussed in section [ sec : amplifications ] , if the universe is very long lived it might be imagined that the decay of baryons and subsequent annihilation of the produced electrons and positrons could smooth out inhomogeneities , and also tend to destroy detailed information about the gravitational history of the expanding phase ( by eliminating compact objects such as neutron stars , for instance . )
thus , even though interactions are unlikely to thermalize matter and radiation when the universe is very large @xcite ( _ cf .
_ section [ sec : opacity ] ) , there may be an analagous information loss via the quantum decay of baryons which could serve a similar function . ( for a completely different idea about why quantum mechanics may effectively decouple the expanding and recollapsing eras , see @xcite . ) in any case , if there is no mechanism to eliminate collapsed objects before the time of maximum expansion , then the collapsed objects of the expanding phase are the same as the collapsed objects of the recontracting phase , implying detailed correlations between the expanding and recontracting era histories of these objects which might lead to difficulties of consistency with the rth .
a particular complication is that it is fairly certain that black holes exist , and that more will form as inhomogeneity grows .
the only way to eliminate a black hole is to allow it to evaporate , yet the estimates of page and mckee indicate that it is more likely for black holes to coalesce into ever bigger holes unless for some reason ( a final boundary condition ? ) there is a maximum mass to the black holes which form , in which case they may have time enough to evaporate ( though this requires @xmath56 to be _ exceedingly _ close to one . ) in fact , it may be imperative for a time symmetric scenario that black holes evaporate , else somehow they would have to turn into the white holes of the recollapsing era ( black holes to observers there .
) this is because we do not observe white holes today @xcite .
( a related observation is that in order for the universe to be smooth whenever it is small , black / white hole singularities can not arise @xcite . ) here the evaporation of black holes before maximum expansion would be enforced by the time symmetric boundary conditions selecting out only those histories for which there are no white holes in the expanding era and _ mutatis mutandis _ for the recollapsing era . on the other hand , if evaporating black holes leave remnants they too must be worked into the picture .
again , if the results of the stochastic models with two - time low entropy boundary conditions discussed above are to be taken seriously , the conclusion should probably be that boundary conditions requiring homogeneity when the universe is small suppress histories in which significant gravitational collapse occurs by assigning low probabilities to histories with fluctuations that will go non - linear .
it hardly needs emphasizing that all of these considerations are tentative , and greater clarity would be welcome .
besides gravitational considerations , radiation which connects the expanding and recollapsing eras provides another example of a physical process which samples conditions near both boundary conditions @xcite .
while gravitational radiation and neutrinos are highly penetrating and are likely to provide such a bridge , in neither case are we yet capable of both effectively observing , and accurately predicting , what is expected from sources to our past . therefore the primary focus of this investigation has been electromagnetic radiation . above it was confirmed that modulo the obviously substantial uncertainty regarding the condition of the universe when it is large , even electromagnetic radiation is likely to penetrate to the recollapsing era .
section [ sec : egbr ] was concerned with one relatively prosaic consequence of this prediction if our universe possesses time symmetric boundary conditions . here
i comment briefly on another .
maxwell s equations , the dynamical laws governing electromagnetic radiation , are time symmetric . it is generally believed that the manifest asymmetry in time of radiation _ phenomena _ , that is , that ( in the absence of source - free fields ) observations are described by retarded solutions rather than advanced , is ascribable fundamentally to the thermodynamic arrow of time without additional hypotheses .
( for a contemporary review see @xcite . ) however , if our universe possesses time symmetric boundary conditions then near the big bang the thermodynamic arrow of entropy increase runs oppositely to that near the big crunch .
since radiation can connect the expanding and recollapsing eras , the past light cone of an accelerating charge in the expanding era ends up in matter for which the entropy is increasing , while its future light cone terminates in matter for which entropy is supposed to be _
decreasing_. if the charge radiates into its future light cone this implies detailed correlations in this matter with the motion of the charge which are incompatible with the supposed entropy decrease there ( entropy increase to time reversed observers ) , although it is true that these correlations are causally disconnected to time - reversed observers , and consequently invisible to _ local _ coarse grainings defining a local notion of entropy for such observers .
this state of affairs makes it difficult to decide whether radiation from an accelerating charge ( if its radiation can escape into intergalactic space ) should be retarded ( from the perspective of observers in the expanding era ) or advanced or some mixture of the two .
the conditions under which the radiation arrow is usually derived from the thermodynamic arrow of surrounding matter do not hold .
( notice how this situation is reminiscent of the requirements necessary to derive retardation of radiation in the wheeler - feynman absorber theory " of electrodynamic phenomena @xcite . )
hence the ability of radiation to connect the expanding and recollapsing epochs brings into question the self - consistency of assuming time symmetric boundary conditions on our universe together with physics as usual " ( here meaning radiation which would be described as retarded by observers in both the expanding and recollapsing eras ) near either end .
the retardation of radiation is another important example of a physical prediction which would be expected to be very different in a universe with time symmetric boundary conditions than in one without . once again , if the results of the simple stochastic models are generally applicable , the retardation of radiation should no longer be a prediction in such a universe .
to summarize this section , consideration of gravitational collapse and radiation phenomena reveals that construction of a model universe with time symmetric boundary conditions which resembles our own may be a difficult task indeed .
there are strong suggestions that a model with time symmetric boundary conditions which mimic our own early universe would behave nothing like the universe in which we live .
such a model would most likely predict a universe which remained smooth throughout the course of its evolution , with coupled matter components consequently remaining in the quasi - static equilibrium " appropriate to a dynamic universe .
in spite of the oft - expressed intuitive misgivings regarding the possibility that our universe might be time symmetric , it has generally been felt that if sufficiently long - lived , there might be no way to tell the difference between a time symmetric and time asymmetric universe .
building on suggestions of cocke , schulman , davies , and gell - mann and hartle ( among others ) , this work has explored in some detail one physical process which , happily , belies this feeling : no matter how long our universe will live , the time symmetry of the universe implies that the extragalactic background radiation be _ at least _ twice that due to the galaxies to our past .
this is essentially due to the fact that light can propagate unabsorbed from the present epoch all the way to the recollapsing era .
moreover , geometry and time symmetry requires this excess " egbr to be associated with sources _ other _ than the stars in those galaxies , sources which , according to present knowledge about the era during which galaxies formed , are not capable of producing this radiation ! thus the time symmetry of a closed universe is a property which is _ directly accessible to experiment _ ( present observations are nearly capable of performing this test ) , as well as extremely difficult to model convincingly on the basis of known astrophysics . in addition , the other theoretical obstacles remarked upon briefly in sections [ sec : amplifications]and [ sec : difficulties ] make it difficult to see how a plausible time symmetric model for the observed universe might be constructed .
in particular , any such attempt must demonstrate that in a universe that is smooth whenever it is small , gravitational collapse can procede to an interesting degree of inhomogeneity when the universe is larger @xcite .
this is necessary in order that the universe display a thermodynamic arrow ( consistently defined across spacelike slices ) which naturally reverses itself as the universe begins to recollapse .
furthermore , it appears unlikely that the usual derivation of the retardation of radiation will follow through in a time symmetric universe in which radiation can connect regions displaying opposed thermodynamic arrows . finally , in the context of the time neutral generalized quantum mechanics employed as the framework for this discussion , unless the locally observed matter - antimatter asymmetry extends globally across the present universe , .
] natural choices of cpt - related boundary conditions yield a theory with trivial dynamics if the deviations from exact homogeneity and isotropy are specified in a cpt invariant fashion ( see the last paragraph of section [ sec : motivations ] ) . in sum , were the excess " egbr which has been the primary concern of this investigation to be observed , it would appear necessary to place the onus of explanation of the fact that the final boundary condition is otherwise practically invisible upon very specially chosen boundary conditions which encode the details of physics in our universe .
this would make it difficult to understand these boundary conditions in a natural way . on the dual grounds of theory and experiment
, it therefore appears unlikely that we live in a time symmetric universe .
( a definitive expurgation must await more thorough investigation of at least some of the aforementioned difficulties . )
for the malcontents in the audience , this appendix offers a flat space model explicitly illustrating the no olber s paradox " argument of section [ sec : amplifications ] . as cosmological
redshifting is time symmetric , the complications due to curvature are inconsequential for present purposes .
( curvature may be included in a straightforward fashion , but that and many other embellishments are , out of courtesy , foregone . )
therefore , consider the universe of figure 1 as flat . for convenience ,
relocate the zero of conformal time @xmath150 to be at the moment of maximum expansion .
the specific energy density in radiation obeys a transfer equation @xmath151 here @xmath152 represents sources ( according to expanders ; thus in the recontracting era @xmath152 may be negative ) , and @xmath153 sinks ( same comment ) , of radiation .
time symmetry implies that @xmath154 @xmath155 and @xmath156 now , suppose it is imagined that a time symmetric universe contains only one class of localized , homogeneously and isotropically distributed sources ( _ i.e .
_ galaxies ) in the expanding era , with corresponding time - reversed sinks ( in the language used by expanding era observers ) in the recontracting era ( _ i.e . _ thermodynamically reversed galaxies . ) for isotropically emitting sources , in the black galaxy " approximation the absorption rate ( emission rate to thermodynamically reversed observers ) in the recontracting era can be thought of as being controlled by the amount of radiation from the expanding era which the recontracting era s galaxies intercept .
that is , the thermodynamically reversed observers of the recontracting era would see their galaxies emitting at a rate given ( at most ) by @xmath157 using time symmetry , and where @xmath158 is as in equation ( [ eq : hardspheresigma ] ) .
( this is merely the expression of the fact that in the essentially geometric black galaxy " approximation , galaxies do not care if they are intercepting radiation from " the past or the future . ) but from this it is obvious that @xmath159 and galaxies are in radiative equilibrium with the sky , a situation reminiscent of the historically important olber s paradox ( why is the night sky dark ? " ) thus in a transparent , time symmetric universe in which the night sky is dark , there must be an additional class of sources emitting the radiation which is correlated with the galaxies of the recontracting era .
( in reality , @xmath160 . indeed , @xmath161 using a characteristic galactic luminosity @xmath162 @xcite , @xmath163 ( _ cf . _
( [ eq : hs ] ) ) , ( [ eq : cw ] ) , and taking @xmath164 .
it should not come as a surprise that the energy density in the optical egbr is still increasing ! )
i wish to thank r. antonucci , o. blaes , t. hurt , and r. geller for useful conversations on matters astrophysical , j. t. whelan for discussions , comments , and questions , and j. b. hartle for raising this question and for provocative discussions .
this work was supported in part by nsf grant no .
phy90 - 08502 . m. gell - mann and j. b. hartle ( 1994 ) , _ in _ proceedings of the nato workshop on the physical origins of time asymmetry " , j. j. halliwell , j. perez - mercader , and w. zurek , eds .
( cambridge university press , cambridge ) , and references therein .
l. s. schulman ( 1991 ) , _ physica _ * a177 * , 373 , and l. s. schulman ( 1994 ) , _ in _ proceedings of the nato workshop on the physical origins of time asymmetry " , j. j. halliwell , j. perez - mercader , and w. zurek , eds .
( cambridge university press , cambridge ) .
r. laflamme ( 1993 ) , _ class .
* 10 * , l79 , and r. laflamme ( 1994 ) , _ in _ proceedings of the nato workshop on the physical origins of time asymmetry " , j. j. halliwell , j. perez - mercador , and w. zurek , eds .
( cambridge university press , cambridge ) . | this paper examines an observable consequence for the diffuse extragalactic background radiation ( egbr ) of the hypothesis that if closed , our universe possesses time symmetric boundary conditions . by reason of theoretical and observational clarity , attention is focused on optical wavelengths .
the universe is modeled as closed friedmann - robertson - walker .
it is shown that , over a wide range of frequencies , electromagnetic radiation can propagate largely unabsorbed from the present epoch into the recollapsing phase , confirming and demonstrating the generality of results of davies and twamley @xcite . as a consequence , time symmetric boundary conditions
imply that the optical egbr is at least twice that due to the galaxies on our past light cone , and possibly considerably more .
it is therefore possible to test _ experimentally _ the notion that if our universe is closed , it may be in a certain sense time symmetric .
the lower bound on the excess " egbr in a time symmetric universe is consistent with present observations . nevertheless , better observations and modeling may soon rule it out entirely .
in addition , many physical complications arise in attempting to reconcile a transparent future light cone with time symmetric boundary conditions , thereby providing further arguments against the possibility that our universe is time symmetric .
this is therefore a demonstration by example that physics today can be sensitive to the presence of a boundary condition in the arbitrarily distant future . |
in recent years much attention has been paid to the origin of intermittency in fully developed turbulence .
progress has been made in the somewhat simpler problem of the anomalous scaling in the passive advection of a scalar quantity ( e.g. temperature or the density of a pollutant ) .
fundamental analytical results have been obtained when the advecting velocity field was assumed to be gaussian and delta - correlated in time , the so - called kraichnan model @xcite , and in the context of shell models @xcite . here
we consider the perhaps more realistic situation where the passive scalar shell model is driven by a non - gaussian velocity field with finite correlation time , namely the velocity field of the goy model @xcite . in this case
the model can be analyzed using standard techniques of ( low ) dimensional dynamical systems , e.g. the study of bifurcations , eigenvalue spectra and lyapunov exponents . in absence of convective effects
the passive scalar field @xmath2 is governed by the equation : @xmath3 here @xmath4 is the velocity field , @xmath5 is an external forcing and @xmath6 is the diffusion coefficient . according to the analogue of the k41 theory @xcite for the passive scalar , developed by obukhov and corrsin @xcite , the structure functions @xmath7 ( where @xmath8 ) scale linearly with p , more precisely @xmath9 .
experimentally , however , one observes for @xmath10 @xcite strong deviations from this .
this is usually referred to as anomalous scaling or intermittency .
the deviations seem to be even more pronounced than for the structure functions of the velocity : the passive scalar is said to be more intermittent
. this paper is organized as follows : after having introduced the model , we examine in section iii the scaling of its fixed points and in section iv their stability .
these studies have already been performed for the goy model @xcite , to which the passive scalar model is coupled .
we will review these results for the sake of completeness . in section v we study the full dynamics of the model and investigate its chaotic and intermittent behavior .
shell models appear to capture many properties of fully developed turbulent flows but are easier to study than the navier - stokes equations ( see @xcite for a review ) . in this letter
we study the passive scalar shell model proposed in ref .
the multiscaling of the model is in good agreement with experimental data @xcite .
the goy model has been studied intensively @xcite ; the passive scalar model has attracted somewhat less attention @xcite , @xcite , @xcite .
both models are constructed in fourier space , retaining only the complex modes @xmath11 and @xmath12 as a representative of all modes in the shell of wave number @xmath13 between @xmath14 and @xmath15 .
one uses the following assumptions : ( i ) the dissipation resp .
diffusion is represented by a linear term of the form : @xmath16 resp .
@xmath17 ( ii ) the nonlinear terms of the form @xmath18 resp.@xmath19 with ( iii ) @xmath20 and @xmath21 among the nearest and next - nearest neighbors of @xmath22 and ( iv ) in absence of forcing and damping conservation of volume in phase space and conservation of @xmath23 and @xmath24 .
moreover the scaling laws @xmath25 and @xmath26 form a fixed point of the inviscid unforced equations .
the resulting equations are @xcite : @xmath27 a possible choice for the coefficients is : @xmath28 the boundary conditions are : @xmath29 for the parameters , we choose for example the following standard values : @xmath30 the free parameter @xmath31 is related to a second quadratic invariant which for the canonical value @xmath32 is similar to the helicity @xcite .
these equations determine the evolution of the vector @xmath33 and thus form a @xmath34 dynamical system .
a first step towards a full understanding of the model consists of an investigation of its static properties . in this section
we examine the scaling properties of the fixed points of equations ( [ goy])-([passhell ] ) .
the major problem is to find the static solutions of ( [ goy ] ) , those of ( [ passhell ] ) can then be found easily because @xmath35 is linear in both @xmath36 and @xmath37 .
it was found in @xcite that ( [ goy ] ) in the unforced inviscid limit allows three self - similar static solutions : the trivial fixed point @xmath38 , a `` kolmogorov '' fixed point @xmath39 and a `` flux - less '' fixed point @xmath40 , with @xmath41 and @xmath42 any function of period three in @xmath22 and @xmath43 .
the corresponding fixed points of ( [ passhell ] ) are : @xmath44 , @xmath45 and @xmath46 .
we will here focus on the kolmogorov fixed point which is believed to be the most important for the dynamics @xcite although it was suggested that also the trivial fixed point might play a major role @xcite .
we note that the static solution for @xmath47 can be turned into real form by a change of phase .
following schrghofer et al.@xcite we choose the phases : @xmath48 it can be shown that the static solution of @xmath37 picks up the same phase as that of @xmath36 .
as has been observed @xcite , the dynamics of the system converges to the kolmogorov fixed point for @xmath49 . when increasing @xmath31 the system undergoes a series of hopf bifurcations and becomes chaotic at @xmath50 @xcite .
in order to find the kolmogorov fixed point one can thus vary @xmath31 in small steps and refine the solution with newton s method @xcite . to study the scaling behavior of the static solutions , we apply a much larger number of shells , using the same parameter values as kadanoff et al .
@xcite : @xmath51 the forcing acts in this case on the first shell . for the static solution of @xmath11 we obtain the same result as @xcite , see figure [ ukad ] , where @xmath52 is plotted versus @xmath22 .
it is averaged over a period three to get rid of the well known period three oscillations .
the solution is seen to follow kolmogorov scaling .
= 0.0022 ( 200,200)(0,0 ) ( 0,0)(200,200)=180 ( 10,150)(0,0)@xmath53 ( 165,10)(0,0)@xmath22 surprisingly however , the fixed point for @xmath54 ( with the fixed point of @xmath11 inserted ) deviates strongly from the obukhov scaling for a finite value of the diffusivity @xmath6 , as can be seen in figure [ thstat ] .
there is clearly not a well - defined power law scaling .
this solution also displays the period three behavior . in the diffusive range
the solution looks somewhat noisy , presumably due to the boundary conditions .
it is clear , that there is a `` slow '' bending in the diffusive regime but as @xmath6 approaches zero , the curve becomes more and more flat and we recover the obukhov scaling in the limit @xmath1 .
we thus note that in contrast to the velocity - case where the viscosity only affects the viscous range , for the passive scalar the diffusion seems to act on the whole inertial range , at least for prandtl numbers @xmath55 .
this might have its origin in the linear character of the problem .
= 0.0022 ( 250,250)(0,0 ) ( 0,0)(200,200)=180 ( 10,150)(0,0)@xmath56 ( 160,10)(0,0)@xmath22
now the stability of the kolmogorov - obukhov fixed point is examined in terms of the eigenvalue spectra .
the system eq.([goy]-[passhell ] ) is written as @xmath57 the jacobian matrix is symbolically given by : @xmath58 the matrix @xmath59 will not matter for the eigenvalues of @xmath60 .
the eigenvalues of @xmath61 were studied in @xcite and @xcite .
it is most convenient to look at disturbances of phase and modulus of the velocity variable @xmath62 : @xmath63 and @xmath64 .
one then obtains for the stability of the modulus : @xmath65 the matrices @xmath66 and @xmath67 are the contributions of the dissipation and cascade terms , respectively .
their expressions are : @xmath68 and @xmath69 where the index @xmath70 has been dropped , for convenience .
the linearized equation for the variation of the phase is : @xmath71 thus the stability eigenvalues of modulus and phases differ only in a minus - sign in front of the cascade term .
we obtain similar results as in @xcite . for values of @xmath72
all the eigenvalues of both phase and modulus matrix have negative real part . above this value ,
some of the real eigenvalues of the phase matrix turn complex ( in pairs ) and cross the imaginary axis . at @xmath73
they have become real again , but now positive .
this situation is shown in figure [ ueig ] .
the eigenvalues of the modulus matrix eventually turn positive at @xmath74 . in figure [ ueig ]
we have multiplied both real and imaginary part with a factor @xmath75 @xcite : @xmath76 thus the phase remains unchanged , while the modulus is rescaled . in case
the eigenvalues have constant ratios , they are evenly spaced on the plot .
thus we are able to visualize both the very small and very large eigenvalues .
= 0.0022 ( 200,200)(0,0 ) ( 0,0)(200,200)=180 it is quite trivial to generalize this method to the passive scalar in order to calculate the eigenvalues of @xmath77 and one finds : @xmath78 where @xmath79 is the dissipation term , very similar to the one before @xmath80 and @xmath81 is the contribution of the cascade term given by : @xmath82 we see that @xmath81 does not depend on @xmath37 as the @xmath37-variation will be differentiated out . for the phase disturbance @xmath83 one
gets @xmath84 we note that since @xmath85 and @xmath86 , the matrix @xmath81 is antisymmetric and its eigenvalues will be purely imaginary
. the stability of phase and modulus will thus be the same .
this is actually a consequence of the conservation of @xmath87 . in the model
any cascade term @xmath88 has to be supplemented by a term @xmath89 , since the conservation implies @xmath90 .
the first term gives a contribution to the matrix @xmath81 : @xmath91 , the second @xmath92 which corresponds to @xmath93 .
the matrix is thus antisymmetric and this is not an artefact of the model since it stems from a conservation law also valid in a real system .
one does not expect to find any bifurcations since the diffusion matrix @xmath94 will cause the eigenvalues to have negative real parts , whatever the value of the driving velocity .
indeed one finds that for all values of @xmath31 , the spectrum of eigenvalues of the phase matrix is like shown in figure [ theig ] , where the eigenvalues are again multiplied by @xmath75 .
one observes evenly spaced eigenvalues organized in branches .
the presence of three branches ( in both upper and lower half of the complex plane ) might be caused by the period three of @xmath11 ( in @xmath22 ) .
if one inserts in the matrix @xmath81 a fixed point of an imaginary model without period 3 , one just finds one branch in each half plane .
= 0.0022 ( 200,200)(0,0 ) ( 0,0)(200,200)=180 let s us now consider what the above imply for the dynamics of the model . at time
@xmath95 , starting from the particular state of the scalar @xmath96 we impose a small perturbation : @xmath97 and look how the perturbation evolves in time , subjected to the _ same _ velocity field . for modulus and phase of @xmath98 we have again equations ( [ thmod ] ) and ( [ thphas ] ) , since the matrices @xmath99 and @xmath100 do not depend on @xmath37 .
because the eigenvalues of @xmath101 have negative real parts , the perturbation will damp out , meaning that after some time @xmath102 .
this has been observed already by crisanti et al .
the term intermittency is used in different contexts and a precise mathematical definition does not exist . in turbulence one speaks of intermittency corrections to the kolmogorov or obukhov - corrsin power law . in dynamical systems in general
, intermittency means the presence of quiescent periods randomly interrupted by burst .
it is believed that these phenomena are related : the scaling corrections should have their origin in intermittent behavior ( in space and/or time ) of velocity or energy dissipation .
here we ask ourselves the question whether the more pronounced deviations from classical scaling for the passive scalar are reflected by a more intermittent behavior ( in time ) of the passive scalar field . in order to test this
we invoke a definition of intermittency for dynamical systems proposed by benzi et al .
@xcite .
firstly , the response function @xmath103 is defined as the rate at which a disturbance vector @xmath104 of the system @xmath105 has expanded after a time @xmath106 .
@xmath107 in chaotic systems one typically observes that @xmath108 @xmath109 , where @xmath110 is the maximum lyapunov exponent .
this maximum lyaponov exponent can be calculated by averaging the _ instantaneous _ lyapunov exponents @xmath111 : @xmath112 .
intermittency can be thought of as connected to the fluctuations of the instantaneous lyapunov exponent .
an intermittency parameter @xmath113 is thus defined as the _ variance _ of @xmath114 @xcite : @xmath115 numerically we proceed as follows : the equations of motion ( [ goy]-[passhell ] ) are numerically integrated along the equations for the disturbance vector @xmath116 where @xmath117 and @xmath60 is the jacobian matrix .
we then estimate the response function and the first two cumulants of its logarithm . from time to time
the disturbance vector is renormalized in order to avoid numerical overflow @xcite .
consider now the lyapunov exponents of the full system of velocity and passive scalar field ( [ goy]-[passhell ] ) , with the parameters ( [ standpar ] ) .
we distinguish how the perturbation , initially applied equally on _ both _
systems ( @xmath118 for all @xmath22 ) , evolves independently in each of the two parts of the system .
therefore we introduce response functions that measure the expansions on the velocity part @xmath119 and on the scalar part @xmath120 .
they are defined as : @xmath121 the corresponding maximal lyapunov exponents is defined as : @xmath122 and the analogue for @xmath123 .
this is reminiscent to the `` eulerian lyapunov exponent '' and `` lagrangian lyapunov exponent '' introduced by crisanti et al.@xcite .
however the lagrangian behavior of the particles is only equivalent to the eulerian passive scalar field in the case of vanishing diffusion .
the intermittency parameters are related to the variance of @xmath124 is the same way as before .
we find numerically : @xmath125 where we have denoted the lyapunov exponent for the full system as : @xmath126 . on expects theoretically : @xmath127\ ] ] since the disturbance vector will evolve towards the most expanding direction .
our result is obviously in agreement with this relation .
for the eulerian and lagrangian lyapunov exponent , one has numerically found the generic inequality @xmath128 . our result @xmath129 might be due to the presence of a finite value of diffusivity .
the intermittency parameters @xmath113 differ on the other hand significantly from each other .
we find ( for @xmath73 ) : @xmath130 , @xmath131 and @xmath132 .
thus the passive scalar behaves equally chaotic , but more intermittent , than the velocity . in order to understand this we write the equation for @xmath37 as follows : @xmath133 where a is a matrix depending on @xmath36 , whose eigenvalues @xmath134 fulfil the inequality @xmath135 .
thus , if @xmath36 is constant in time , @xmath37 will converge to its fixed point @xmath136 .
alternatively , if @xmath36 fluctuates as for @xmath73 in the goy model , this `` fixed point '' will fluctuate as well
. moreover the convergence towards it will be irregular since the eigenvalues of @xmath137 vary .
it is therefore natural to expect that the behavior of @xmath37 is more intermittent than that of @xmath36 .
we studied various properties of a shell model for the advection of a passive scalar .
we calculate the eigenvalue spectrum of the kolmogorov fixed point and show that the passive scalar part is stable against perturbations .
this is in fact not very surprising since it follows from conservation of @xmath138 .
the fixed point does on the other hand not follow the obukhov - corrsin scaling .
this suggests that the scaling properties of the passive scalar are much more sensitive , than the velocity , to non - inertial properties such as the dissipation and perhaps also the forcing . in experiments the scaling zones of the passive scalar
are less clearly observed and some `` nonuniversality '' ( with respect to forcing ) seems to have been observed in @xcite , but we of course can not pretend that the mechanism observed here will be found also in real systems . we have measured the relation between lyapunov exponents and intermittency parameters of the passive scalar and velocity .
we find : @xmath129 and @xmath139 .
thus the passive scalar behaves more intermittently which is in agreement with its more pronounced deviations from linear scaling .
it would be interesting to establish these relations from the equations of motion .
it could also be interesting to investigate the relation between coherent structures in the goy model @xcite and those in the passive scalar model .
j.k . is grateful to the niels bohr institute for warm hospitality .
discussions with hugues chat , ted janssen , anal lematre and paolo muratore - ginanneschi are kindly acknowledged .
we thank detlef lohse for providing us with the code to extract the kolmogorov fixed point .
10 present address : + service de physique de letat condens , centre detudes de saclay , 91191 gif - sur - yvette , france m.h .
jensen , g. paladin , and a. vulpiani , phys .
a * 45 * 7214 ( 1992 ) r. benzi , g. paladin , g. parisi , and a. vulpiani , j. phys .
a * 18 * , 2157 ( 1985 ) r.h .
kraichnan , phys .
fluids * 11 * , 945 ( 1968 ) ; phys . rev .
lett . * 72 * , 1016 ( 1994 ) .
k. gawdzki and a. kupiainen , phys .
rev . lett . * 75 * , 3834 ( 1995 ) .
m. chertkov , g. falkovich , i kolokolov and v. lebedev , phys . rev .
e * 52 * , 4924 ( 1995 ) .
a. wirth and l. biferale phys .
e * 54 * , 4982 ( 1996 ) ; r. benzi , l. biferale , a. wirth phys .
78 * , 4926 ( 1997 ) e. b. gledzer , sov .
. dokl . * 18 * , 216 ( 1973 ) .
m. yamada and k. ohkitani , j. phys .
japan * 56 * , 4210(1987 ) ; prog .
phys . * 79*,1265(1988 ) .
kolmogorov , dokl .
nauk sssr , * 30*,301 ; * 32 * , 16 ( 1941 ) a.m. obukhov , izv .
nauk sssr , ser .
. geofiz . * 13*,58 ( 1949 ) ; s. corrsin , j. appl .
* 22 * , 469 ( 1951 ) g. ruiz - chavarria , c. baudet , and s. ciliberto , physica d * 99 * , 369 ( 1996 ) l. kadanoff , d. lohse , j. wang , and r. benzi , phys .
fluids * 7 * , 617 ( 1995 ) .
n. schrghofer , l. kadanoff , and d. lohse , physica d * 88 * , 40 ( 1995 ) .
l. kadanoff , d. lohse , and n. schrghofer , physica d * 100 * , 165 ( 1997 ) .
t. bohr , m. h. jensen , g. paladin , and a. vulpiani , _ dynamical systems approach to turbulence _
( cambridge university press , cambridge ) ( 1998 ) m. h. jensen , g. paladin , and a. vulpiani , phys .
rev . a * 43 * , 798 ( 1991 ) .
r. benzi , l. biferale , and g. parisi , physica d * 65 * , 163 ( 1993 ) .
d. pisarenko , l. biferale , d. courvasier , u. frisch , and m. vergassola , phys .
fluids a * 65 * , 2533 ( 1993 ) .
l. biferale , a. lambert , r. lima , and g. paladin .
physica d * 80 * , 105 ( 1995 ) .
gilson and t. dombre , phys .
lett . * 79 * , 5002 ( 1997 ) f. okkels and m.h .
jensen , phys .
e * 57 * , 6643 ( 1998 ) j. kockelkoren , f. okkels , and m.h .
jensen , j. stat phys .
* 93 * , 833 ( 1998 ) a. crisanti , m.h .
jensen , g. paladin , and a. vulpiani , j. phys .
a * 26 * , 6943 ( 1993 ) k.h .
andersen and p. muratore - ginanneschi , chao - dyn/9902007 , to be published in physica d. g. benettin , l. galgani , a. giorgilli , and j .-
strelcyn , meccanica * 15 * , 22 ( 1980 ) a. crisanti , m. falcioni , g. paladin , and a. vulpiani , riv .
nuovo cimento * 14 * , 1 ( 1991 ) j .- f .
pinton , f. plaza , l. danaila , p. le gal , and f. anselmet , physica d * 122 * , 187 ( 1998 ) | we investigate the fixed points of a shell model for the turbulent advection of passive scalars introduced in @xcite .
the passive scalar field is driven by the velocity field of the popular goy shell model .
the scaling behavior of the static solutions is found to differ significantly from obukhov - corrsin scaling @xmath0 which is only recovered in the limit where the diffusivity vanishes , @xmath1 . from the eigenvalue spectrum we show that any perturbation in the scalar will always damp out , i.e. the eigenvalues of the scalar are negative and are decoupled from the eigenvalues of the velocity .
furthermore we estimate lyapunov exponents and the intermittency parameters using a definition proposed by benzi et al .
@xcite .
the full model is as chaotic as the goy model , measured by the maximal lyapunov exponent , but is more intermittent . 1 |
a recent entry in wikipedia , an internet - based encyclopedia , defines the quantum zeno effect as follows : _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ the quantum zeno effect is a quantum mechanical phenomenon first described by george sudarshan and baidyanaith misra of the university of texas in 1977 .
it describes the situation that an unstable particle , if observed continuously , will never decay .
this occurs because every measurement causes the wavefunction to `` collapse '' to a pure eigenstate of the measurement basis @xcite .
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ this definition is close to the original language of misra and sudarshan @xcite , but is not sufficiently general to describe the many situations that are considered to be examples of the quantum zeno effect .
it is true that the quantum zeno effect describes the situation in which the decay of a particle can be prevented by observations on a sufficiently short time scale .
however , the quantum zeno effect is much more general , since it describes the situation in which the time evolution of _ any _ quantum system can be slowed by sufficiently frequent `` observations . ''
the references to observations and to wavefunction collapse tend to raise unnecessary questions related to the interpretation of quantum mechanics .
actually , all that is required is that some interaction with an external system disturb the unitary evolution of the quantum system in a way that is effectively like a projection operator .
finally , the word `` never '' describes a limiting case .
a slowing of the time evolution , as opposed to a complete freezing , is generally regarded as a demonstration of the quantum zeno effect .
graph of the number of citations per year to misra and sudarshan @xcite and to itano _
et al _ @xcite.,width=480 ]
the 1977 article `` the zeno s paradox in quantum theory '' by misra and sudarshan @xcite studied the evolution of a quantum system subjected to frequent ideal measurements .
they showed that , in the limit of infinitely frequent measurements , a quantum system would remain in its initial state .
applied to the case of an unstable particle whose trajectory is observed in a bubble chamber or film emulsion , this result seemed to imply that such a particle would not decay , in contradiction to experiment . in this case , the resolution to the apparent paradox lies in the fact that the interactions between the particle and its environment that lead to the observed track are not sufficiently frequent to modify the particle s lifetime .
the time distribution of literature citations to misra and sudarshan @xcite is shown in fig .
[ citesfig ] .
the total number of citations listed in the web of science database in october 2006 was 535 .
the graph shows a relatively low but steady number of citations per year for about a decade , followed by a large increase that continues for over a decade , possibly peaking about 25 years after the original publication date .
the great increase in the rate of citations in recent years is partially due to the increased interest in quantum information processing , where the quantum zeno effect may find practical applications .
energy - level diagram for the ihbw demonstration of the quantum zeno effect.,width=384 ]
the quantum zeno effect can be derived in an elementary way by considering the short - time behavior of the state vector @xcite .
( the treatment of misra and sudarshan @xcite is more general since it involves the density matrix . )
let @xmath0 be the state vector at time @xmath1 .
if @xmath2 is the hamiltonian , in units where @xmath3 , then the state vector at time @xmath4 is @xmath5 , and the survival probability is @xmath6 = @xmath7 .
if @xmath4 is small enough , it should be possible to make a power series expansion : @xmath8 so that the survival probability is @xmath9,\ ] ] where @xmath10 many quantum systems have states whose survival probability appears on ordinary time scales to be a decreasing exponential in time .
this is inconsistent with the quadratic time dependence of eq .
( [ quadratic ] ) and implies that in such cases eq .
( [ quadratic ] ) holds only for very short times .
consider the survival probability @xmath11 , where the interval @xmath12 $ ] is interrupted by @xmath13 measurements at times @xmath14 . ideally , these measurements are instantaneous projections and the initial state @xmath0 is an eigenstate of the measurement operator . in that case , the survival probability is @xmath15^n , \label{survival_n}\ ] ] which approaches 1 as @xmath16 .
it is important to note that at this level there should be nothing controversial or problematic about the existence of the quantum zeno effect .
the quantum zeno effect should be observed as long as the physical system can be made to display the behavior shown in eq .
( [ survival_n ] ) . for a given system
, it may be difficult or impossible to make measurements quickly enough for the quadratic time dependence of the survival probability to be observed , so that , as a practical matter , the quantum zeno effect can not be observed .
it should be noted that the semantic arguments over terms such as `` measurement '' or `` observation '' can be avoided if we accept that a `` measurement '' is an operation that interrupts the unitary time evolution governed by @xmath2 in such a way as to yield eq .
( [ survival_n ] ) as a good approximation .
that is , the `` measurement '' should effectively act as a projection operator .
according to this view , it is not necessary that the `` measurements '' be recorded by a macroscopic apparatus or that they be instantaneous .
timing of the radio frequency and optical fields applied to the beryllium ions in the ihbw experiment , width=384 ]
the experiment of itano , heinzen , bollinger , and wineland ( ihbw ) @xcite was based on a proposal of cook @xcite for observing the quantum zeno effect in a three - level atom ( see fig . [ 3levelfig ] ) .
levels 1 and 2 are stable on the time scale of the experiment .
level 3 decays to level 1 with the emission of a photon . in the experiment of ihbw ,
levels 1 and 2 were two of the hyperfine sublevels of the ground @xmath17 state of the be@xmath18 ion .
level 3 was a sublevel of the @xmath19 excited state that decayed only to level 1 .
the experiment was carried out with a sample of about 5000 be@xmath18 ions confined by electric and magnetic fields in a penning trap .
the steps in the experiment were as follows : 1 .
the ions were prepared in level 1 by optical pumping with the laser beam .
2 . a resonant radio frequency ( rf )
magnetic field was applied for the interval required to drive the ions to level 2 .
3 . during the time that the rf pulse was applied
, a variable number @xmath13 of equally spaced short laser pulses was applied to the ions ( see fig . [ pulsefig ] ) .
the laser ( resonant with the 1-to-3 transition ) was turned on , and the induced fluorescence was recorded .
probability of making the 1-to-2 transition as a function of the number @xmath13 of optical `` measurement '' pulses.,width=480 ] the intensity of the laser - induced fluorescence at the end of the experiment was proportional to the population of level 1 . if there are no optical pulses during the long rf pulse , the population of level 2 as a function of the time @xmath4 that the rf pulse is applied is @xmath20,\label{rabiflop}\ ] ] where @xmath21 is proportional to the amplitude of the rf field .
if the duration of the rf pulse is chosen to be @xmath22 ( a pi - pulse ) , then all of the population is transferred from level 1 to level 2 .
if @xmath13 equally - spaced laser pulses of negligible duration are applied during the rf pi - pulse , the population of level 2 at time @xmath23 is @xmath24,\label{simplecalc}\ ] ] which approaches 0 as @xmath13 goes to infinity .
figure [ datafig ] compares the data to theory .
the solid bars represent the transition probability as a function of @xmath13 according to the simplified calculation of eq .
( [ simplecalc ] ) .
the bars with horizontal stripes represent the data .
the bars with diagonal stripes represent a calculation that takes into account the finite duration of the laser pulses and optical pumping effects .
the data are in reasonably good agreement with the simplified calculation and in better agreement with the improved calculation .
the decrease in @xmath25 as @xmath13 increases demonstrates the quantum zeno effect .
a variation of the experiment was carried out by initializing the ion in level 2 and then applying the rf field and the laser pulses . in this case , the transition from level 2 to level 1 was inhibited as @xmath13 increased .
this is another example of the quantum zeno effect . in this case
, the inhibition of the transition is accompanied by the _ absence _ of laser - induced fluorescence .
recently , the quantum zeno effect was observed for an unstable quantum system by fischer _ et al _ @xcite . the quantum zeno effect for induced transitions and for unstable systems
are not fundamentally different , since they both follow from the general arguments of misra and sudarshan @xcite , but it has been difficult to observe in the latter case , because of the short times over which the decay is nonexponential .
et al _ were able to create an artificial system ( atoms tunneling from a standing - wave light field ) in which the interactions could be controlled so as to observe the desired effects .
as can be seen by the history of citations ( fig .
[ citesfig ] ) , the publication of the ihbw experiment @xcite generated considerable interest . initially , some of the responses were critical in one way or another . some ( e. g. , ref .
@xcite ) objected to the use of the term `` wavefunction collapse '' in describing the experiment .
the authors responded that the concept of wavefunction collapse was not essential , and that any interpretation of quantum mechanics that yielded the same prediction of the experimental results should be regarded as valid @xcite .
some objected to the fact that photons were not actually observed during the intermediate `` measurements , '' in the sense of having the scattered photons registered by a detector , so that the experiment did not actually demonstrate the quantum zeno effect @xcite .
however , the results are predicted to be the same whether or not the intermediate measurements are made .
it is enough that the measurements _ could _ have been made .
as long as the laser interactions act effectively as projection operators , so that the algebra of eqs .
( [ quadratic])([survival_n ] ) is followed , the experiment should be regarded as a demonstration of the quantum zeno effect .
it should be noted that none of the criticisms were directed at the execution of the experiment itself , only at the interpretation . for the most part , the citations to ref .
@xcite simply accept it as a demonstration of the quantum zeno effect .
in fact , it is cited in quantum mechanics textbooks @xcite and popular science books @xcite .
while misra and sudarshan originally used the term `` quantum zeno paradox , '' as did peres @xcite and others , the more recent work usually uses the term `` quantum zeno effect , '' perhaps because the effect no longer seems paradoxical .
some authors distinguish between the quantum zeno _ paradox _ and the quantum zeno _ effect _ , but they do so in differing ways .
pascazio and namiki @xcite call the situation in which the frequency of measurements is finite and the evolution is slowed the quantum zeno _ effect , _ and the limiting case in which the frequency of measurements is infinite and the evolution is frozen the quantum zeno _
paradox_. block and berman @xcite call the inhibition of spontaneous decay the quantum zeno _ paradox _ and the inhibition of induced transitions ( as in the ihbw experiment ) the quantum zeno _ effect_. in ref .
@xcite , home and whitaker reserve the term quantum zeno _ paradox _ for a negative - result experiment involving observations with a macroscopic apparatus .
this definition of the quantum zeno paradox seems to exclude most , if not all , feasible experiments . in this context , the ihbw experiment is not regarded as an example of the quantum zeno paradox because a local interaction is present between the laser field and the atoms , and also because the electromagnetic field , containing zero or a few scattered photons , is not regarded as a macroscopic observation apparatus .
they regard the type of experiments where the time evolution of a quantum system is affected by a direct interaction , for example with an external field , as examples of the quantum zeno _
effect_. however , in a later publication @xcite the same authors treat the terms quantum zeno _ paradox _ and quantum zeno _ effect _ as synonymous and restrict both to nonlocal negative - result experiments involving a macroscopic observation apparatus . experiments that do not meet these criteria would not be examples of _ either _ the quantum zeno paradox _ or _ the quantum zeno effect , according to their later definition .
several variations on the general theme of quantum zeno effects have been described .
soon after the ihbw experiment was carried out , peres and ron @xcite showed that a partial quantum zeno effect results if the measurements are too weak to completely destroy the coherence of the state of the measured system .
a modification of the ihbw experiment was proposed in which the measurement laser pulses are weakened .
et al _ @xcite showed that a related effect , damped oscillations of the state populations , can occur if the duration of the experiment is extended , while weak measurements are made . some , including kofman and kurizki @xcite and facchi _ et al _ @xcite have shown that the decay of an unstable quantum system can be accelerated by frequent observations
this is called the quantum anti - zeno effect or the inverse quantum zeno effect . as is the case for the quantum zeno effect
, the observations must take place before the decay becomes exponential . unlike the quantum zeno effect , which follows from rather general arguments , e. g. eqs .
( [ quadratic])([survival_n ] ) , the possibility of observing a quantum anti - zeno effect depends on the details of the system .
the experiment of fischer _
et al _ @xcite demonstrated the quantum anti - zeno effect as well as the quantum zeno effect .
an interesting generalization of the quantum zeno effect is the concept of quantum zeno dynamics @xcite .
frequent measurements can confine the evolution of a quantum system to a subspace of the hilbert space rather than simply to the initial state . compared to the ordinary quantum zeno effect
, the difference is that the measurements distinguish not between the initial _ state _ and all other states but between a _
subspace _ and the rest of the hilbert space .
this form of quantum zeno effect may find application in quantum information processing .
as already noted , the recent increase in the rate of citations to the articles of misra and sudarshan @xcite and ihbw @xcite is partially related to increased interest in quantum information processing . in this context
, there have been various proposals to use the quantum zeno effect to preserve quantum systems from decoherence .
beige _ et al _ @xcite have proposed an arrangement of atoms inside an optical cavity capable of carrying out quantum logic operations with low error rates within a decoherence - free subspace of the hilbert space .
states outside the decoherence - free subspace are coupled strongly to the environment .
the quantum zeno effect then leads to effective confinement of the system to the decoherence - free subspace , which is an example of a quantum zeno subspace .
franson _ et al _ @xcite have proposed use of the quantum zeno effect to suppress errors in a linear optics implementation of quantum computation . in this implementation , the presence of two photons in the same mode indicates an error .
the presence of a strong two - photon absorber in an optical fiber takes the role of the `` observer '' and suppresses the errors .
other proposed applications of the quantum zeno effect to error prevention in quantum computation are discussed in refs .
@xcite .
quantum `` bang - bang '' control and related dynamical decoupling techniques @xcite utilize frequent , pulsed interactions to effectively prevent decoherence of a quantum system by confining the dynamics to a subspace .
this is not exactly the quantum zeno effect , since the interactions are unitary , but the results are mathematically similar to those for the quantum zeno effect .
dhar _ et al _ @xcite have discussed the `` super - zeno effect , '' which preserves a state ( or more generally , keeps a quantum system within a subspace of the hilbert space ) with a set of pulsed interactions unequally spaced in time .
the timing of these interactions can be arranged so as to be more efficient than can be done with the same number of equally spaced interactions ( ordinary quantum zeno effect ) .
also , it should be noted that the pulsed interactions are unitary kicks , as in the so - called `` bang - bang control '' @xcite , and not observations in the usual sense .
the 1977 publication of misra and sudarshan stimulated a great deal of theoretical and experimental work that has enhanced our understanding of the time development of quantum systems , such as the short - time nonexponential decay of unstable quantum systems .
the results of the ihbw experiment , published in 1990 , was a clear confirmation of the existence of the quantum zeno effect for the case of the inhibition of an induced transition .
interest in the quantum zeno effect continues to be high , partially due to the possibility of practical applications in quantum information processing . | as of october 2006 , there were approximately 535 citations to the seminal 1977 paper of misra and sudarshan that pointed out the quantum zeno paradox ( more often called the quantum zeno effect ) .
in simple terms , the quantum zeno effect refers to a slowing down of the evolution of a quantum state in the limit that the state is observed continuously .
there has been much disagreement as to how the quantum zeno effect should be defined and as to whether it is really a paradox , requiring new physics , or merely a consequence of `` ordinary '' quantum mechanics .
the experiment of itano , heinzen , bollinger , and wineland , published in 1990 , has been cited around 347 times and seems to be the one most often called a demonstration of the quantum zeno effect .
given that there is disagreement as to what the quantum zeno effect _ is _ , there naturally is disagreement as to whether that experiment demonstrated the quantum zeno effect .
some differing perspectives regarding the quantum zeno effect and what would constitute an experimental demonstration are discussed . |
qcd hard processes are described within the improved parton model , thanks to the collinear factorization . in that kinematical regime , only a few quasi on - shell quarks and gluons per hadron are relevant for the dynamics , and thus hadrons are appropriately described by parton distribution functions
. however , this picture breaks down in the high - energy limit or equivalently at low bjorken-@xmath0 . indeed , due to the enhancement of soft gluon _
bremsstrahlung _ , the gluon occupation number reaches non - perturbatively large values at low enough @xmath0 .
gluons in this regime are better described as a semiclassical field rather than as a dilute gas of partons .
the growth of the gluon occupation number is then tamed by the non - linearity of the yang - mills equations .
this is the phenomenon of gluon saturation.@xcite this phenomenon is taken into account in various formalisms which aim at describing dense - dilute and/or dense - dense observables . in the former case ,
the low-@xmath0 gluons of a hadron or nucleus target are probed by a dilute projectile .
typical examples of that case are deep inelastic scattering ( dis ) observables at low-@xmath0 and forward physics in hadron - hadron or hadron - nucleus collisions .
by contrast , dense - dense observables involve the collision of the low-@xmath0 gluon field of two hadrons or nuclei against each other . the inclusive @xmath1-particle production at mid - rapidity in heavy ion or hadron - hadron collision at high - energy enters in that category , as well as the @xmath1-point correlators of the energy - momentum released at mid - rapidity in such collisions . in this contribution ,
i focus on the paradigmatic example of dis , whereas dense - dense processes are discussed by venugopalan in his contribution.@xcite note that both for dense - dilute and dense - dense observables , the high - energy leading logs are resummed by the same b - jimwlk equations@xcite acting on wilson line operators , first derived without explicit reference to the projectile .
hence , there is a factorization property valid at leading logarithmic accuracy for several very different observables . in the large @xmath2 limit
, the b - jimwlk equations applied to a dipole of wilson lines operators reduce to the much simpler bk equation.@xcite both the b - jimwlk equations and the bk equation reduce to the bfkl equation@xcite in the dilute limit for the target .
recently , balitsky and chirilli have calculated@xcite the evolution equation resumming the next to leading logarithms ( nll ) in the particular case of dipoles of wilson lines operators ( and thus the nll bk equation ) , as well as the nlo impact factor@xcite ( or high - energy coefficient - functions ) for dis structure functions .
these calculations show that the high - energy factorization remains valid at nlo / nll accuracy , at least in the case of dis structure functions .
this high - energy factorization generalizes the dipole factorization@xcite , valid at lo only .
hence , high - energy factorization is now a good candidate for a factorization property valid to all orders , analog to collinear and tmd factorizations .
however , the nll bk equation is plagued by problems inherited from its linearized analog , the nll bfkl equation.@xcite in both equations , the nlo contributions to the kernel are large , especially in the collinear and anti - collinear kinematical regimes , signaling a breakdown of the weak coupling expansion in most of the available phase - space .
hence , in order to obtain stable and reliable solutions of the nll evolution equations , one needs to perform appropriate resummations , already done in the bfkl case.@xcite in the next section , the derivation of the high - energy factorization is outlined in the context of light - front quantization , and the underlying assumptions are discussed in detail . in the section [ sec : largecorr ] , the large corrections appearing in nll bfkl and nll bk equations and their physical origin are discussed .
progresses towards the resummation of these corrections to the nll bk equation are also reviewed . in the section [ sec : kinconst ] ,
preliminary results about the resummations of the large corrections of kinematical origin are introduced , with emphasis on the factorization scheme dependence of these effects .
the high - energy factorization for dis can be obtained both from covariant or light - front approaches to qcd .
the formalism used by balitsky and chirilli and discussed in their respective contributions,@xcite is fully covariant and thus suitable to perform systematically higher order calculations . as an alternative ,
let us use a light - front formalism , which allows to make the physical content of the calculations more transparent at leading order or at leading logarithmic accuracy .
this formalism is essentially the qcd version of the one constructed by bjorken , kogut and soper for qed.@xcite let us choose a frame such that the projectile moves mainly along the @xmath3 direction and the target mainly along @xmath4 , and such that the target is much more boosted than the projectile . in the high - energy ( or low @xmath0 ) limit , the target is more and more lorentz contracted along @xmath3 , so that the @xmath3 time interval around @xmath5 needed for the projectile to pass through the target becomes arbitrarily short .
hence , the forward scattering process related to dis by the optical theorem happens in three steps * the projectile propagates in the vacuum form @xmath6 to @xmath7 , * the projectile interacts with the target at @xmath5 , * the projectile propagates in the vacuum form @xmath5 to @xmath8 , up to contributions suppressed by powers of the lorentz @xmath9 factor of the target .
the formalism of light - front wave functions allows to describe the state of the projectile just before or just after the collision .
a light - front wave function is indeed the overlap of the physical state of the projectile with a given fock state at @xmath5 . in the dis case ,
the projectile is in principle the incoming lepton , but as in covariant calculations , one can factor out the leptonic tensor , and take effectively the projectile to be the exchanged virtual photon , which can be transverse or longitudinal . at lo , only the quark - antiquark fock component of the virtual photon projectile contributes to dis , so that the hadronic contribution to the projectile state writes formally @xmath10 with a sum over the phase space and quantum numbers of the quark and antiquark . at nlo , the quark - antiquark - gluon fock component contributes as well , so that @xmath11 in the high - energy limit , the various partons present in a fock state of the projectile scatter independently on the target , up to power suppressed corrections .
indeed , the time interval for the interaction between the target and the projectile goes to zero when boosting more and more the target , and also the typical time scale for interactions within the projectile is lorentz dilated when boosting the projectile . following the color glass condensate ( cgc ) formalism@xcite ,
let us model the target by a random classical gluon field , obeying some classical statistical distribution .
i will discuss the validity of this assumption later on .
then , we need to know how a parton scatters in the high - energy limit on a classical gluon field shrinking to a shockwave at @xmath5 .
when passing through such a shockwave , the transverse position of a parton has no time to change .
hence , it is convenient to write the fock state basis in a mixed representation where the transverse position rather than the transverse momentum of the partons is specified , in addition to their light - cone momentum @xmath12 .
both the probability of helicity flip of the parton and the probability of loss of a non - negligible fraction of his light - cone momentum @xmath12 , by scattering on the gluon shockwave , are power suppressed in the high - energy limit .
hence , the scattering of independent partons on the classical gluon shockwave reduces to the so - called eikonal scattering , which is diagonal in the fock state basis except with respect to color indices .
a parton is thus only color rotated when passing through the shockwave field .
this color rotation is given by a wilson line in the appropriate representation , defined along the trajectory of the parton through the target gluon field , with fixed transverse position @xmath13 , noted for example @xmath14 in the case of a quark . in particular , the s - matrix for the forward scattering of a free quark - antiquark color singlet dipole ( at transverse positions @xmath15 and @xmath16 ) on a classical gluon shockwave is the identity matrix in fock space times the factor @xmath17 using fierz identities , the analog s - matrix factor for a quark - antiquark - gluon color singlet can be written as @xmath18 \ : , \ ] ] where @xmath19 is the transverse position of the gluon , so that @xmath20\ : .\ ] ] all in all , the projectile - target total cross section writes at lo as @xmath21 \ : , \label{lo_cross_section}\ ] ] with a different wave function @xmath22 whether the virtual photon projectile is transverse or longitudinal , and where @xmath23 represents the statistical average over the target gluon field . and at nlo
, one expects the total cross section to be @xmath24 + 2 \int \textrm{d}ps_{\textrm{q}\bar{\textrm{q}}\textrm{g } } \ ; \ ; \left|\psi^{\textrm{nlo}}_{\textrm{q}\bar{\textrm{q}}\textrm{g}}\right|^2 \ ; \ ; \big[1-\big < s_{\mathbf{x}_0 \mathbf{x}_1 \mathbf{x}_2}\big >
\big ] \ : , \ ] ] and so on at higher orders . using the probability conservation relation @xmath25 the total cross section at nlo can be rewritten as @xmath26\nonumber\\ & & + \frac{2 n_c^2}{n_c^2\!-\!1 } \int \textrm{d}ps_{\textrm{q}\bar{\textrm{q}}\textrm{g } } \ ; \ ; \left|\psi^{\textrm{nlo}}_{\textrm{q}\bar{\textrm{q}}\textrm{g}}\right|^2 \ ; \ ; \big[\big <
s_{\mathbf{x}_0 \mathbf{x}_1}\big>-\big < s_{\mathbf{x}_0 \mathbf{x}_2}\ , s_{\mathbf{x}_2 \mathbf{x}_1}\big > \big]\label{nlo_cross_section } \ : , \end{aligned}\ ] ] separating the lo and nlo contributions from each other
. however , this is not the end of the story .
indeed , the operators constructed from light - like wilson lines such as @xmath27 suffer from rapidity divergences , coming from gluons of large @xmath12 and/or vanishing @xmath28 belonging to the field of target . and the usual logarithmic divergence of soft gluon _
bremsstrahlung _ shows up in the nlo contribution as @xmath29 these two divergences are both regulated when one avoids double counting of gluons , thanks to a longitudinal cut - off .
for example , let us take a cut - off @xmath30 in @xmath12 , and consider gluons with @xmath31 only as part of the projectile , in order to regulate the integration with the weight , and gluons with @xmath32 only as part of the target semi - classical field , so that @xmath33 and @xmath34 become @xmath30-dependent .
the evolution with @xmath30 of @xmath35 , @xmath36 and similar expectation values is obtained by integrating out qcd quantum corrections slice by slice in @xmath12 . at leading log accuracy ,
one gets @xmath37\label{eqbal1}\ ] ] where @xmath38 , and similar other coupled equations forming balitsky s infinite hierarchy,@xcite involving expectations values of operators with an arbitrary number of wilson lines .
the jimwlk equation@xcite is equivalent to this full hierarchy , and gives the evolution with @xmath30 of the target gluon field probability distribution @xmath39 .
the fact that the quantum corrections to the statistical distribution of gluons fields can be absorbed by evolution of the statistical distribution is a crucial self - consistency check for the cgc formalism .
alternatively , the same equation can be derived@xcite by extracting the soft - gluon logarithmic divergence coming from the photon light - cone wave function in , and requiring the dis cross section to be independent of the cut - off @xmath30 , playing the role of a longitudinal factorization scale .
the consistency of the two approaches shows the validity of the high - energy factorization formulae and for dis once the cut - off @xmath30 is applied , since the b - jimwlk evolution properly resums the high - energy leading logs . in the large @xmath2 limit , one has @xmath40 , so that becomes a closed equation , called balitsky - kovchegov ( bk ) equation.@xcite when using the nlo factorization formula , by consistency , one should probably resum not only the high - energy leading logs , but also the next - to - leading ones
this is performed by the nll extension of the equation , derived recently.@xcite let us come back to the main assumption of the cgc formalism , _
i.e. _ that the target can be modeled by a random classical gluon field .
first , we have seen the self - consistency of this assumption thanks to the jimwlk equation .
second , when the gluon occupation number is large , it is large compared to commutators , so that the physics becomes effectively classical and the cgc assumption is justified . due to lorentz contraction ,
the valence partons of a relativistic large nucleus are densely packed , so that the cgc formalism applies to large nuclei as soon as relativistic effects are important .
moreover , the b - jimwlk evolution leads to a more and more packed target , so that all this formalism is always valid at high enough energy , whatever is the target , even if in intermediate stages of the high - energy evolution , the target is dilute .
finally , the dilute limit ( or two - gluon approximation ) of the equation and of its nll extension correctly reproduces the ll and nll bfkl equations respectively , so that at this accuracy , no information about the dilute regime is lost when taking the semi - classical target approximation .
however , this approximation for the target field might prevent to reproduce the full nnll bfkl equation in the dilute limit .
the bfkl equation receives large corrections at nll,@xcite giving a pathological behavior to its solutions and signaling a breakdown of the corresponding formalism .
soon after this discovery , it has been realized that these issues come entirely from the collinear and anti - collinear kinematical regimes and not from the regge kinematical regime in which the bfkl equation is derived .
the solution is then to perform resummations@xcite in the collinear and anti - collinear regimes , in order to extend the validity of the nll bfkl toward these regimes , which are in principle govern by dglap physics .
several successful resummations schemes have been proposed , which differ only by inessential higher order contributions .
the general idea behind these schemes is to add appropriate nnll and higher order terms to the nll bfkl equation in such a way that the ( n)lo dglap equation is reproduced in the collinear and anti - collinear limits .
the physical origin of the large corrections has also been understood .
they can be classified as follows . *
running coupling scale : the nll bfkl kernel contains terms which are the beginning of the expansion of the running coupling . when the coupling is taken to run , but not with the correct scale , some of these terms stay in the nll kernel instead of being resummed into the running coupling .
if the running coupling scale differs in the collinear or anti - collinear regime from the one dictated by dglap physics , the left - over terms in the kernel are particularly large .
* non - singular part of dglap splitting function : the ll bfkl equation contains the singular part of the lo dglap gluon splitting function both in the collinear and anti - collinear regimes , but not the non - singular part . the non - singular part shows up in an expanded way , term after term when going to nll bfkl and further . this expanded form appears to be pathological , and one should instead include the full non - singular part together with the singular part of the splitting function in a modified version of the ll bfkl equation . *
kinematical constraint : the last large corrections , which are parametrically the largest ones , come from a bad treatment of kinematics when taking the high - energy limit too strictly
. they can be resummed by imposing some kinematical constraint@xcite in the ll bfkl equation , or by shifting collinear and/or anticollinear poles of the kernel in mellin space.@xcite the corrections of the third type depend on the factorization scheme , in particular on the precise definition of the evolution variable for the bfkl equation .
they are however identical in any yang - mills theory .
by contrast , the corrections of the first two types are high - energy factorization scheme independent .
but they vary from theory to theory .
in particular , they are both absent in @xmath41 sym theory
. since the linearization of the nll bk equation gives the nll bfkl equation , those large corrections are also present in the nll bk equation .
the nonlinearity related to gluon saturation in the bk equation can not help stabilize the solutions in the presence of those large corrections , as shown by toy - model numerical simulations.@xcite hence , it is necessary to perform collinear resummations also for the nll bk equation before using it for phenomenology .
however , the resummation schemes developed for nll bfkl can not be used directly for the nll bk equation , because of the nonlinearity of the latter and of the switch from transverse momentum to transverse position space .
the terms related to running coupling in the nll bk equation have been analyzed in detail , and an appropriate running coupling prescription is available.@xcite concerning the non - singular part of the gluon splitting function , only a very rough treatment is available,@xcite so that more work on this issue is desirable .
let us now discuss the kinematical issues in the bk context .
the need for the kinematical constraint in the evolution equation was first discovered in a covariant formalism , in a quite technical calculation aiming at the description of other effects like angular ordering.@xcite however , the origin of the kinematical constraint and of the related large nll corrections can be understood very clearly in light - front perturbation theory , as shown in ref . .
the easiest derivation of the ll bfkl and bk equations is performed by calculating the leading logarithmic contributions to the squared wave functions @xmath42 for the quark antiquark and @xmath1 gluons fock component of the projectile in the multi - regge kinematics.@xcite in light - front perturbation theory , the kinematical issues are only due to the usual implementation of the multi - regge kinematics on the energy denominators , coming with each additional gluon . in this context , the multi - regge kinematics is the assumption that the @xmath12 of the partons are strongly ordered , to the closest to the target , @xmath43 .
@xmath44 is the momentum of the antiquark , and @xmath45 the one of the virtual photon . ]
@xmath46 and that simultaneously their transverse momentum are of the same order @xmath47 if one assumes only the @xmath12 ordering , then for each diagram , the energy denominator associated with the emission of the gluon @xmath48 , with @xmath49 , reduces to @xmath50 then , in the strict multi - regge kinematics - , one can approximate each energy denominator by its last term @xmath51 that approximation is crucial to the derivation of the ll bfkl and bk equations . when the high - energy factorization is formulated with a cut - off in @xmath12 , as presented in section [ he_fact_dis ] ,
the @xmath12-ordering of the gluon cascades generated by the bfkl and bk equations is automatic .
however , the integration over transverse momentum or position in the kernel of the ll bfkl and bk equations is not restricted , so that it contains contributions badly violating and thus outside the multi - regge kinematics .
more precisely , once the ordering is satisfied , the approximation is good typically for gluon cascades in the multi - regge kinematics or in the anti - collinear kinematics , _
i.e. _ with @xmath52 growing with @xmath48 , but not in the deep collinear regime , _
i.e. _ with @xmath52 decreasing fast enough with @xmath48 . from this discussion
, ones predicts that the @xmath12-ordered bfkl and bk equations are less reliable in the deep collinear regime , which is especially relevant for dis , than in the rest of the available phase - space .
this observation explains why the mellin transform of the nll kernel of the @xmath12-ordered bk equation@xcite contains a large and unphysical collinear triple pole , but no anti - collinear triple pole .
all this discussion is however factorization scheme dependent .
there are indeed other ways to regularize the rapidity divergence of wilson line operators and the soft divergence .
for example , one can replace the cut - off in @xmath12 by a cut - off in @xmath28 . in that case , the bfkl and bk equations implement automatically the @xmath28-ordering @xmath53 so that approximations of the type are good . however , this time , the approximation of the energy denominators by the expression is not always valid .
in fact , we are in the inverse situation : the approximation of energy denominators by the @xmath28 of the last gluon is in general valid except in the deep anti - collinear regime .
hence , a spurious anti - collinear triple pole is expected in the mellin transform of the @xmath28-ordered nll bfkl and bk kernels , but no collinear triple pole .
in fact , whatever factorization scheme is chosen , the approximation of the energy denominators by the @xmath28 of the last gluon is valid if and only if both the @xmath12 and @xmath28-ordering - are satisfied , whereas the assumption is unnecessary .
moreover , one can show that if either or is not satisfied , no leading logarithmic contribution is generated.@xcite hence , the bfkl and bk equations should actually resum precisely the gluon cascades ordered both in @xmath12 and @xmath28,@xcite whereas the usual assumption is unnecessary and misleading . in any factorization scheme ,
imposing both @xmath12 and @xmath28 ordering allows to resum into a modified ll equation the largest unphysical collinear or anti - collinear contributions appearing in the kernel at higher logarithmic orders , corresponding in mellin space to triple poles at nll , quintuple poles at nnll and so on . only the @xmath12-ordering or the @xmath28-ordering ,
not both , can be made automatic by a choice of high - energy factorization scheme .
the other one has to be included as a modification of the ll kernel , by restricting the integration over transverse momentum or position to the domain allowing to satisfy both and .
for the bfkl equation in momentum space , this restriction of the integration domain is precisely the kinematical constraint found in ref . .
so far , we have shown the necessity of the kinematical constraint in the bfkl and bk evolutions by an analysis in transverse momentum space .
however , evolutions equations including gluon saturation like the bk and jimwlk equations are more easily written in transverse position space .
hence , one has to translate the kinematical constraint to a restriction in position space . a first implementation of the kinematical constraint in position space for the bfkl and bk equations has been proposed in ref .
i have recently clarified@xcite some issues related to this first implementation , in particular by doing a more careful treatment of virtual corrections and by analyzing systematically the factorization scheme dependence of the kinematical constraint . in the end , one gets a modified bk evolution equation with memory effects and a restricted phase - space for gluon emission which gets progressively wider in the course of the evolution.@xcite
this manuscript has been authored under the contract no . # de - ac02 - 98ch10886 with the u.s .
department of energy .
0 l. v. gribov , e. m. levin and m. g. ryskin , phys .
rept . * 100 * , 1 ( 1983 ) .
a. h. mueller and j. w. qiu , nucl .
b * 268 * , 427 ( 1986 ) .
l. d. mclerran and r. venugopalan , phys .
d * 49 * , 2233 ( 1994 ) .
l. d. mclerran and r. venugopalan , phys .
d * 49 * , 3352 ( 1994 ) .
l. d. mclerran and r. venugopalan , phys .
d * 50 * , 2225 ( 1994 ) .
r. venugopalan , _ these proceedings_. i. balitsky , nucl .
b * 463 * , 99 ( 1996 ) .
j. jalilian - marian , a. kovner , a. leonidov and h. weigert , nucl .
b * 504 * , 415 ( 1997 ) .
j. jalilian - marian , a. kovner , a. leonidov and h. weigert , phys .
d * 59 * , 014014 ( 1999 ) .
j. jalilian - marian , a. kovner and h. weigert , phys .
d * 59 * , 014015 ( 1999 ) .
a. kovner , j. g. milhano and h. weigert , phys .
d * 62 * , 114005 ( 2000 ) .
h. weigert , nucl .
a * 703 * , 823 ( 2002 ) .
e. iancu , a. leonidov and l. d. mclerran , nucl .
a * 692 * , 583 ( 2001 ) .
e. iancu , a. leonidov and l. d. mclerran , phys .
b * 510 * , 133 ( 2001 ) .
e. ferreiro , e. iancu , a. leonidov and l. mclerran , nucl .
a * 703 * , 489 ( 2002 ) .
y. v. kovchegov , phys .
d * 60 * , 034008 ( 1999 ) .
y. v. kovchegov , phys .
d * 61 * , 074018 ( 2000 ) .
l. n. lipatov , sov .
j. nucl .
* 23 * , 338 ( 1976 ) [ yad .
* 23 * , 642 ( 1976 ) ] .
e. a. kuraev , l. n. lipatov and v. s. fadin , sov .
jetp * 45 * , 199 ( 1977 ) [ zh . eksp . teor .
fiz . * 72 * , 377 ( 1977 ) ] .
i. i. balitsky and l. n. lipatov , sov .
j. nucl .
* 28 * , 822 ( 1978 ) [ yad .
* 28 * , 1597 ( 1978 ) ] .
i. balitsky and g. a. chirilli , phys .
d * 77 * , 014019 ( 2008 ) .
i. balitsky and g. a. chirilli , nucl .
b * 822 * , 45 ( 2009 ) .
i. balitsky and g. a. chirilli , phys . rev .
d * 83 * , 031502 ( 2011 ) .
n. n. nikolaev and b. g. zakharov , z. phys .
c * 49 * , 607 ( 1991 ) . v. s. fadin and l. n. lipatov , phys .
b * 429 * , 127 ( 1998 ) .
m. ciafaloni and g. camici , phys .
b * 430 * , 349 ( 1998 ) .
m. ciafaloni , d. colferai and g. p. salam , phys .
d * 60 * , 114036 ( 1999 ) .
m. ciafaloni , d. colferai , g. p. salam and a. m. stasto , phys .
d * 68 * , 114003 ( 2003 ) .
g. altarelli , r. d. ball and s. forte , nucl .
b * 742 * , 1 ( 2006 ) .
i. balitsky , _
these proceedings_. j. d. bjorken , j. b. kogut and d. e. soper , phys .
d * 3 * , 1382 ( 1971 ) .
m. ciafaloni , nucl .
b * 296 * , 49 ( 1988 ) .
b. andersson , g. gustafson and j. samuelsson , nucl .
b * 467 * , 443 ( 1996 ) .
j. kwiecinski , a. d. martin and p. j. sutton , z. phys .
c * 71 * , 585 ( 1996 ) .
g. p. salam , jhep * 9807 * , 019 ( 1998 ) .
e. avsar , a. m. stasto , d. n. triantafyllopoulos and d. zaslavsky , arxiv:1107.1252 [ hep - ph ] .
i. balitsky , phys .
d * 75 * , 014001 ( 2007 ) .
e. gotsman , e. levin , u. maor and e. naftali , nucl .
a * 750 * , 391 ( 2005 ) .
l. motyka and a. m. stasto , phys .
d * 79 * , 085016 ( 2009 ) .
a. h. mueller , nucl .
b * 415 * , 373 ( 1994 ) .
g. beuf , _ to appear_. | the high - energy factorization and the associated b - jimwlk or bk evolution equations are presented , using the example of dis structure functions . the necessity of taking gluon saturation into account is discussed , and also the various approximations underlying high - energy factorization .
the appearance of large nll corrections in such a framework with or without gluon saturation is recalled , and their physical origin is explained .
finally , old and new results are presented about the resummation of some of those large corrections , related to kinematical approximations . |
high - tc cuprate superconductors harbor many fundamental questions , which challenge elaborate numerical analysis on superconductivity , magnetism and other properties .
specifically , there is growing realization that various instabilities can exist along with superconductivity @xcite , and the relation between various charge instabilities and superconductivity in the cuprates is now being intensively studied @xcite . also , some experiments suggest a spontaneous breakdown of the four - fold symmetry of electronic states in the tetragonally structured cuprates , which is viewed as a kind of electronic nematicity"@xcite
. there are some explanations for the nematicity , e.g. in the context of fluctuating stripe orders @xcite .
pomeranchuk instability , a spontaneous breaking of four - fold symmetry of the fermi surface without lattice distortion , is evoked as another possible candidate for nematicity in cuprate superconductors @xcite .
the presence of pomeranchuk instability in two - dimensional lattice models has been suggested in refs . , where the forward scattering was found to develop to induce pomeranchuk instability .
subsequently , properties of this instability were studied primarily in mean - field models ( f - model " ) , where the electrons interact only via forward scattering @xcite . for the two - dimensional ( 2d ) hubbard model on the square lattice , a representative model for cuprates ,
the existence of this instability is yet to be fully clarified microscopically .
functional renormalization group ( frg ) calculations suggest that the superconducting fluctuation is stronger than pomeranchuk instability @xcite , while other numerical renormalization - group approaches suggest pomeranchuk instability to be stronger around van hove fillings@xcite .
gutzwiller wave functions combined with an efficient diagrammatic expansion technique ( de - gwf ) obtained a ground state with a coexistence of the nematic order and superconductivity in 2d hubbard model@xcite .
also , the dynamical cluster approximation ( dca ) and cellular dynamical - mean - field theory ( cdmft ) showed large responses against small distortions of the lattice@xcite , from which the possibility of spontaneous symmetry breaking is suggested to occur at lower temperatures or for larger cluster sizes . while these results suggest that the 2d repulsive hubbard model has a strong tendency toward the pomeranchuk instability , whether or not this instability has higher transition temperature ( @xmath2 ) than that of superconductivity ( @xmath3 ) has to be elaborated .
more importantly , the relation between the pomeranchuk instability and superconductivity ( e.g. , whether they are cooperative or competing ) is an intriguing question .
while the mean - field study for the phenomenological model suggests that they are competing with each other with @xmath3 suppressed in the coexistence region @xcite , the relation should be clarified by going beyond mean - field approaches .
given the situation , we study in the present paper superconductivity and pomeranchuk instability in the intermediate correlation regime by evoking flex+dmft@xcite , a diagrammatic extension of the dynamical mean field theory ( dmft ) , which takes into account the spin and charge fluctuation effects on top of the dmft local self - energy and can reproduce the dome structure in @xmath3 @xcite .
the advantages of this method are first , there is no finite - size effect in contrast to dca and cdmft , which should be important for capturing small fermi surface deformations , and second , we can calculate finite - temperature regions in contrast de - gwf to capture the effect of this nematicity on the superconducting @xmath3 .
this also enables us to systematically examine the relation between superconductivity and pomeranchuk instability when the electron band filling and the second and further neighbor hoppings ( @xmath6 ) are varied . after confirming the existence of pomeranchuk instability near the van hove filling consistently with the previous works
, we shall study the superconducting phase , which indicates that the symmetry of the gap function is changed from the ordinary @xmath0-wave pairing to ( @xmath7)-wave @xcite .
interestingly , the fermi surface distortion can enhance the superconductivity in the overdoped ( or strongly frustrated ) regime with larger @xmath6 .
we shall explain this @xmath3 enhancement by a perturbation picture for small fermi surface distortions , and also by the random phase approximation ( rpa ) in the weak - coupling regime .
we also reveal that the pomeranchuk instability temperature @xmath2 is more sensitive to @xmath8 , hence the fermi surface warping , than the superconducting @xmath3 .
this contrasts with the previous mean - field calculations @xcite that showed almost the same filling dependence for the two transition temperatures .
this should come from the fact that the present formalism takes account of the filling dependence of the pairing interaction beyond mean - field levels .
the result also implies that the region for stronger superconducting fluctuations and that for stronger pomeranchuk instability can be separated .
we consider the standard repulsive hubbard model on the square lattice with a hamiltonian , @xmath9 where @xmath10 creates an electron with wave - vector @xmath11 and spin @xmath12 , @xmath13 is the on - site coulomb interaction , and @xmath14 . in the presence of
second - neighbor ( @xmath15 ) and third - neighbor ( @xmath16 ) hopping parameters , the 2d band dispersion is given as @xmath17 where @xmath18 is the nearest - neighbor hopping ( the unit of energy hereafter ) , and @xmath19 the chemical potential .
we basically adopt @xmath20 , @xmath21 , which are determined to fit the band calculation for a typical hole - doped single - layer cuprate , hgba@xmath22cuo@xmath23 @xcite . for the numerical procedure ,
we employ flex+dmft method , which is a kind of diagrammatic extension of dmft , where the fluctuation exchange approximation ( flex ) and the dmft are combined , with a double self - consistency loop .
this kind of scheme has been considered in ref . , and has recently been formulated through luttinger - ward functional with applications to superconducting states in ref .. the latter can describe a @xmath3 dome against the band filling along with a spectral weight transfer .
these are a virtue of flex+dmft that corrects the overestimated local - flex self - energy in a filling - dependent manner . in flex+dmft ,
the self - energy is calculated through the flex self - energy and dmft self - energy as @xmath24 the flex self - energy @xmath25 is given as @xmath26 g(k^{\prime } ) .
\label{flex - selfenergy}\end{aligned}\ ] ] here @xmath27 and @xmath28 are the number of @xmath29-points and inverse temperature , respectively , @xmath30 with @xmath31 the matsubara frequency for fermions , @xmath32 green s function , and @xmath33 is the irreducible susceptibility . the local part of the flex self - energy , @xmath34 , is computed by replacing green s function @xmath35 with the local one , @xmath36 , in eqs .
( [ flex - selfenergy ] ) and ( [ chi ] ) . for calculating the dmft self - energy , @xmath37
, we need to solve the impurity problem in dmft . here
, we employ the modified iterative perturbation theory ( modified ipt ) as an impurity solver . in this method ,
the original ipt is modified for systems without particle - hole symmetry @xcite , thus applicable to frustrated or non - half - filled cases .
this is not computationally expensive , which enables us to scan over various parameter regions .
we have checked , by using alps library@xcite , that the continuous - time quantum monte carlo ( ct - qmc ) impurity solver@xcite gives basically similar results even away from half - filling in the intermediate - coupling regime [ see fig .
[ fig : eta_u4](a ) ] . after obtaining green s function ,
we plug it into the linearized eliashberg equation , @xmath38 where @xmath39 is the anomalous self - energy , while @xmath40 is the effective pairing interaction , and @xmath41 the eigenvalue of eliashberg s equation .
superconducting @xmath3 is determined as the temperature at which @xmath42 . in the right hand side of eq .
( [ eliashberg ] ) , we have neglected the local dmft vertex contribution , whose validity is discussed in appendix [ appendix ] . to allow the pomeranchuk instability to occur ,
we introduce a seed to deform the fermi surface in the initial input for green s function .
while we linearize the anomalous part of green s function as being infinitesimal , we can deal with finite pomeranchuk order parameters , so that we can discuss superconductivity for finite distortions in this formalism .
for @xmath43 . in ( a ) , the circles and squares correspond to the results of flex+dmft using the modified ipt and ct - qmc as a dmft impurity solver , respectively .
( c ) fermi surface [ as represented by the color - coded spectral weight @xmath44 with @xmath45 , for @xmath46 ( @xmath47 ; left ) and @xmath48 ( @xmath49 ; right ) . ] and band filling @xmath50 for @xmath43 .
we show the superconducting @xmath3 with undistorted fermi surface ( green circles ) , superconducting @xmath3 with fermi surface distortion ( purple circles ) , and pomeranchuk @xmath2 ( black squares ) .
the dotted line represents @xmath2 when we ignore the superconductivity .
the yellow arrow indicates the van hove filling in the interacting system .
( b)density of states at the filling indicated by the yellow arrow in ( a ) for @xmath51 . ]
the pomeranchuk order parameter @xmath52 can be defined , for the originally four - fold cosine bands , as @xmath53 and we display the result against temperature and band filling in fig .
[ fig : eta_u4](a),(b ) , respectively .
we can see that the order parameter starts to grow continuously with decreasing temperature , which indicates a second - order phase transition .
if we turn to the filling dependence , we observe the order parameter abruptly growing around the edges of the pomeranchuk phase , indicative of transferring to the first - order phase transition consistently with the previous work @xcite .
hereafter , we focus on the filling region around the peak of @xmath2 , where the transition is second order .
the peak sits at about @xmath54 for the value of the parameter considered ( @xmath55 ) . if we look at the fermi surface in fig .
[ fig : eta_u4](c ) for @xmath56 , we can see that the fermi surface , identified as the ridges in the spectral function @xmath57 , indeed becomes distorted at lower temperatures , @xmath58 . the phase diagram is displayed in fig . [
fig : phase_eta_u4](a ) , where we can see that the pomeranchuk instability temperature is peaked around @xmath54 for the present parameter set .
an yellow arrow indicates the van hove filling in the interacting system at which the spectrum is peaked at the fermi energy , see fig .
[ fig : phase_eta_u4](b ) , for which we have obtained the spectral function with the pad approximation , and confirmed the peak position does not change for @xmath59 .
the fact that the pomeranchuk instability tends to be strong around this filling is consistent with the previous results @xcite .
the peak in the pomeranchuk dome does not precisely coincide with the van hove filling , which may be an effect of the asymmetric density of states @xcite as in fig .
[ fig : phase_eta_u4](b ) .
by contrast , the present result shows that the superconducting @xmath3 is a monotonically decreasing function of the hole doping near the van hove filling .
( left panels ) and the spectral weight @xmath57 for @xmath60 ( right ) are shown for @xmath61 ( a ) , @xmath62 ( b ) , and @xmath63 ( c ) .
the symbols are the same as in fig .
[ fig : phase_eta_u4](a ) , and yellow allows indicate respective van hove fillings in the interacting system . ] in the previous mean - field calculations @xcite which ignore the filling dependence of the effective pairing interaction , both @xmath3 and @xmath2 are peaked around the van hove filling .
thus the present result indicates that the filling dependence of the effective interaction has an important effect of rendering distinction of optimal doping levels between pomeranchuk @xmath2 dome and superconducting @xmath3 dome . to confirm this difference ,
let us systematically vary the second- and third - neighbor hopping parameters @xmath64 in fig .
[ fig : tp - dependence ] , which change the fermi surface warping as well as the van hove filling .
we can see that , for a fixed @xmath65 , the change in the parameters shifts the distance of the filling from the van hove filling as represented by the blurring of the spectral function around @xmath66 .
the left of fig .
[ fig : tp - dependence ] plots the phase diagrams for three typical cases with different fermi surface warping .
we find that the pomeranchuk @xmath2 drastically changes along with the van hove filling ( yellow allows ) , while the superconducting @xmath3 is much less sensitive .
we can thus conclude that , despite both of superconductivity and pomeranchuk instabilities being fermi surface instabilities affected by the spectral weight at the fermi energy , the pomeranchuk instability is much more sensitive to the fermi surface shape ( distance from the van hove filling ) .
this implies that we can separate the dominant regions for the two instabilities by changing the position against the van hove filling ( dominated by @xmath67 ) .
now , an intriguing issue is how superconductivity behaves in the presence of the pomeranchuk fermi surface distortion .
we actually plot in fig .
[ fig : phase_compare](a ) the superconducting order parameter , where the pairing symmetry is seen to be distorted from the ordinary @xmath0-wave to @xmath68(extended)@xmath69-wave . here
, an interesting observation is that the superconducting @xmath3 can be enhanced by the pomeranchuk distortion of the fermi surface .
indeed , if we go back to fig .
[ fig : phase_eta_u4 ] , we have also plotted the superconducting @xmath3 when the four - fold fermi surface is artificially imposed below pomeranchuk @xmath2(green dots ) .
we can see the @xmath3 with the distorted fermi surface ( purple dots ) is higher . to identify the origin of this enhancement
, we can go back to the pairing interaction to compare between the cases of pomeranchuk - distorted and the four - fold - imposed fermi surfaces .
figure [ fig : phase_compare](b ) plots the difference of the two cases for the same parameters ( @xmath55 , @xmath54 and @xmath70 ) .
we can see that the pomeranchuk instability distorts the pairing interaction , with a @xmath0-wave - like sign reversal .
with @xmath45 ( left panel ) , which can be decomposed into a @xmath0-wave part and an ( extended ) @xmath69-wave ( four - fold symmetric ) part ( right ) .
( b ) difference in the pairing interaction with the fermi surface distortion ( @xmath71 ) and without ( @xmath72 ) , for @xmath73 .
] to pin - point the origin of the distortion effect on the superconducting @xmath3 , we can consider the perturbational effect for small distortions , based on a general linearized gap equation , @xmath74 where @xmath75 , while @xmath76 is the kernel , given as @xmath77 in flex+dmft ( as seen by multiplying @xmath78 to both sides in eq .
( [ eliashberg ] ) ) . if we consider small @xmath0-wave - like distortions [ as displayed in fig .
[ fig : phase_compare](b ) ] for this kernel , @xmath79 the first - order perturbation for the maximum eigenvalue @xmath80 satisfies @xmath81 where the @xmath82 is the eigenvector for the @xmath80 , and @xmath83 identically vanishes due to the @xmath0-wave nature of the @xmath84 , so that the leading term is the second - order one , @xmath85 where @xmath86 is the index for the eigenvalue @xmath87 and eigenvector @xmath88 of the kernel matrix @xmath89 .
since this expression is positive - definite , small @xmath0-wave deformations of the kernel in the linearized gap equation always enhance the superconducting @xmath3 .
this explains the @xmath3 enhancement in fig .
[ fig : phase_eta_u4](a ) , and can provide a new pathway for enhancing superconducting @xmath3 in terms of fermi surface distortion .
however , it should be difficult to achieve purely @xmath0-wave like distortions for the kernel , and the terms having some other symmetries should in general arise even from purely @xmath0-wave distortions of the fermi surface .
we can elaborate this by introducing a parameter @xmath90 where @xmath90 represents either ( 1 ) spontaneous distortion of the electronic states @xmath91 or ( 2 ) small @xmath0-wave modulation of the hamiltonian @xmath92 .
then we can expand the interaction kernel in @xmath90 , which gives , up to the second - order , @xmath93 and the effect on the eigenvalue @xmath41 reads @xmath94 we can see that whether @xmath3 can be enhanced depends on the second term on the right - hand side of eq .
( [ second - order ] ) . from this
we expect that the enhancement tends to occur when the second - largest eigenvalue is close to the largest one , for which the first term on the right - hand side of eq .
( [ second - order ] ) becomes dominant .
calculated within rpa between the four - fold symmetric fermi surface @xmath95 and the distorted fermi surface @xmath96 with red ( black ) dots representing the case of @xmath97 ( @xmath98 ) for @xmath99 .
the horizontal axis corresponds to the band filling from @xmath100 to @xmath101 , while the vertical axis is @xmath102 , with @xmath103 ( which includes the parameter set used in the fig .
[ fig : tp - dependence ] ) . ]
we can in fact check this argument in the weak - coupling case . to obtain qualitative tendencies ,
we have performed the rpa calculation in relatively high temperature for various values of parameters to compare the four - fold symmetric case with the distorted fermi surface by making the nearest - neighbor hopping anisotropic , @xmath104 , by hand with the first line in eq .
( [ dispersion ] ) becoming @xmath105 . in the rpa , we ignore the self - energy effect in the eliashberg eq .
( [ eliashberg ] ) . when we compare the eigenvalue under a distortion @xmath106 with @xmath107 for the symmetric case , the result in fig .
[ fig : compare_weak - nematic ] for @xmath108 shows that we do have a region ( marked with red circles for @xmath109 ) in which the distortion enhances the eigenvalue .
this effect tends to occur away from half - filling , and for larger values of distant - neighbor hopping @xmath110 ( i.e. , more frustrated cases ) .
thus we can confirm that the enhancement of the superconductivity by small distortions indeed occurs in the weak - coupling limit where we can ignore the self - energy effect .
it has been known that the gap symmetry ( for the leading eigenvalue ) tends to be changed for higher doping or more frustrated cases@xcite .
the present result suggests that the @xmath3 enhancement arising from the distortion tends to occur around the boundary for the gap symmetry to change where the leading and sub - leading eigenvalues are close to each other .
this is also consistent with the above result for the @xmath15 dependence in flex+dmft ( fig .
[ fig : tp - dependence ] ) , where the enhancement of @xmath3 occurs for @xmath111 .
we also notice that the structure of eq .
( [ second - order ] ) is reminiscent of the pseudo jahn - teller effect , in which a jahn - teller - like distortion occurs without degeneracies due to the second - order effect of the distortion @xcite . in this context
we can also recall a well - known property that , if the eigenvalues are degenerate ( e.g. for @xmath112-pairing ) , @xmath3 can be enhanced by the strain effect @xcite . from these
, the present result may also be viewed as a possibility for this kind of @xmath3 enhancement revealed even for the ( non - degenerate ) @xmath0-wave regime in 2d square lattice hubbard model .
we have employed flex+dmft approach to study the interplay of pomeranchuk instability and superconductivity in correlated electron systems .
we have revealed that the superconductivity with the distorted fermi surface has the symmetry of the gap function changed from @xmath0-wave to @xmath7 , consistent with the previous studies @xcite .
we have found that the fermi surface distortion can enhance the superconducting @xmath3 in the overdoped regime .
we have explained this enhancement in terms of the perturbation for small distortions , and also with rpa in the weak - coupling regime .
the pomeranchuk @xmath2 is found to be much more sensitive to fermi surface warping and the position of the van hove filling than the superconducting transition temperature . if we comment on the method , flex+dmft , which is an improvement over flex or dmft , still overestimates the non - local self - energy effect . for more accurate estimates , other methods ( e.g. diagrammatic expansion in two - particle level as in d@xmath113a @xcite or dual fermion method @xcite )
will be needed .
also , we have assumed here translationally invariant systems , while the study of the interplay between superconductivity and charge instabilities involving finite wave vectors will be another interesting future work .
the present work was supported by jsps kakenhi grant number jp26247057 , and by impact program of council for science , technology and innovation , cabinet office , government of japan ( no . 2015- pm12 - 05 - 01 ) ( ha ) , by the advanced leading graduate course for photon science ( mk ) , and by jsps kakenhi grant number jp25800192 ( nt ) .
accoding to the formulation in ref . , we should consider the local anomalous self - energy @xmath115 coming from the dmft functional .
then the linearized eliashberg equation becomes @xmath116 \notag \\ & \times |g(k^{\prime})|^2 \delta(k^{\prime}),\end{aligned}\ ] ] where @xmath117 is the functional derivative of the local anomalous self - energy .
while this term can be ignored for studying pure @xmath0-wave pairing as in the previous paper @xcite , we can examine its effect on the @xmath7 pairing .
we consider this effect to be small , because the additional term is an extended @xmath69-wave ( nonlocal ) pairing rather than the ordinary @xmath69-wave , so that a cancellation should occur in the momentum summation . to check this along the argument in the main text
, we can calculate the lower bound of the maximal eigenvalue when @xmath114 is considered without calculating @xmath114 directly . from the eigenvector @xmath118 of eq .
( [ eliashberg ] ) , we extract the part of the gap function that is not affected by @xmath114 as @xmath119 then a quantity , @xmath120 gives the lower bound for the maximal eigenvalue when @xmath114 is considered .
we have actually confirmed that the difference between @xmath121 and @xmath41 ( without @xmath114 ) is very small , @xmath122 .
thus we can conclude the effect of the dmft vertex @xmath114 does not significantly change the the result for the @xmath3 enhancement .
99 b. keimer , s. a. kivelson , m. r. norman , s. uchida and j. zaanen , nature * 518 * , 179 ( 2015 ) .
t. wu , _ et al .
_ , nature * 477 * , 191 ( 2011 ) . e. h. da silva neto , _ et al .
_ , science * 343 * , 393 ( 2014 ) .
r. comin , _ et al .
_ , science * 343 * , 390 ( 2014 ) .
w. tabis , _ et al .
_ , nature commun . * 5 * , 5875 ( 2014 ) .
y. ando , k. segawa , s. komiya , a. n. lavrov , phys .
lett . * 88*,137005 ( 2002 ) .
v. hinkov , _ et al .
_ , science * 319 * , 597 ( 2008 ) .
kivelson , e. fradkin , and v. j. emery , nature ( london ) * 393 * , 550 ( 1998 ) .
h. yamase and w. metzner , phys .
b * 73 * , 214517 ( 2006 ) . c. j. halboth and w. metzner , phys . rev .
lett . * 85*,5162 ( 2000 ) .
h. yamase and h. khono , j. phys .
* 69 * , 332 ( 2000 ) .
h. yamase , v. oganesyan and w. metzner , phys .
b * 72 * , 35114 ( 2005 ) .
p. jakubczyk , w. metzner and h. yamase , phys .
lett . * 103*,239904 ( 2009 ) . c. honerkamp , m. salmhofer and t. m. rice , eur
j. b * 27 * , 127 ( 2002 ) .
v. hankevych , i. grote and f. wegner , phys .
b * 66 * , 094516 ( 2002 ) .
j. kaczmarczyk , t. schickling and j. bunemann , arxiv:1512.06688 .
s. su and t. a. maier , phys .
b * 84 * , 220506(r ) ( 2011 ) .
s. okamoto , d. senechal , m. civelli and a. m. s. tremblay , phys .
b * 82 * , 180511(r ) ( 2010 ) .
h. yamase and w. metzner , phys .
b * 75 * , 155117 ( 2007 ) .
j. gukelberger , l. huang , and p. werner , phys .
b * 91 * , 235114 ( 2015 ) .
m. kitatani , n. tsuji and h. aoki , phys .
b * 92 * , 085104 ( 2015 ) .
k. nishiguchi , k. kuroki , r. arita , t. oka , and h. aoki , phys .
b * 88 * , 014509 ( 2013 ) ; k. nishiguchi , phd thesis univ .
of tokyo ( 2013 ) .
similar values are also obtained in h. sakakibara , h. usui , k. kuroki , r. arita , and h. aoki , phys .
lett . * 105*,057003 ( 2010 ) .
j. p. hague , m. jarrell , and t. c. schulthess , phys .
b * 69 * , 165113 ( 2004 ) .
h. kajueter and g. kotliar , phys .
lett . * 77*,131 ( 1996 ) .
a. n. rubtsov , v. v. savkin , and a. i. lichtenstein , phys .
b * 72 * , 035122 ( 2005 ) .
e. gull , a. j. millis , a. i. lichtenstein , a. n. rubtsov , m. troyer , and p. werner , rev .
phys . * 83 * , 349 ( 2011 ) .
b. bauer , _ et al .
_ , j. stat .
p05001 ( 2011 ) .
a. f. albuquerque , _ et al . _ , j. magn . magn . mater . *
310 * , 1187 ( 2007 ) .
a. t. rmer , _ et al .
_ , phys . rev .
b * 92 * , 104505 ( 2015 ) .
u. opik , and m. h. l. pryce , proc .
london ser .
a * 238 * , 425 ( 1957 ) . m. b. walker , and p. contreras , phys .
b * 66 * , 214508 ( 2002 ) .
a. toschi , a. a. katanin , and k. held , phys .
b * 75 * , 045118 ( 2007 ) .
a. n. rubtsov , m. i. katsnelson , and a. i. lichtenstein , phys .
b * 77 * , 033101 ( 2008 ) . | interplay of pomeranchuk instability ( spontaneous symmetry breaking of the fermi surface ) and @xmath0-wave superconductivity is studied for the repulsive hubbard model on the square lattice with the dynamical mean field theory combined with the fluctuation exchange approximation ( flex+dmft ) .
we show that the four - fold symmetric fermi surface becomes unstable against a spontaneous distortion into two - fold near the van hove filling , where the symmetry of superconductivity coexisting with the pomeranchuk distorted fermi surface is modified from the @xmath0-wave pairing to @xmath1-wave . by systematically shifting the position of van hove filling with varied second- and third - neighbor hoppings , we find that the transition temperature @xmath2 of pomeranchuk instability is more sensitively affected by the position of van hove filling than the superconducting @xmath3 .
this implies that the filling region for strong pomeranchuk instability and that for strong superconducting fluctuations can be separated , and pomeranchuk instability can appear even if the peak of @xmath4 is lower than the peak of @xmath5 .
an interesting observation is that the fermi surface distortion can enhance the superconducting @xmath3 in the overdoped regime , which is explained with a perturbation picture for small distortions . |
the invariant theory of @xmath0 arrays , or trilinear forms on a 3-dimensional vector space , has been studied by many authors ; the earliest reference we have been able to find is the work of aronhold @xcite ( 1850 ) .
a classification of the forms , and an application of classical invariant theory to this problem , appears in chanler and thrall @xcite ( 1938 - 39 ) .
these trilinear forms and their invariants are also closely tied to the classical study of the hesse configuration of 12 lines in the projective plane , which was addressed with a modern view by artebani and dolgachev @xcite ( 2009 ) .
see dolgachev @xcite ( 2012 ) for a comprehensive view .
the first results using the representation theory of lie groups were obtained by vinberg @xcite ( 1976 ) , who embedded the semisimple lie algebra @xmath4 into an exceptional lie algebra of type @xmath5 , and deduced that the algebra of invariants is freely generated by homogeneous polynomials of degrees 6 , 9 and 12 , which we denote respectively by @xmath6 , @xmath7 and @xmath8 .
strassen ( * ? ? ?
* theorem 4.6 ) ( 1983 ) studied @xmath9 arrays ( @xmath10 odd ) , showed that the complement of the set of arrays of maximal border rank is a hypersurface , and determined its equation . for @xmath11 , this polynomial is ( up to a scalar multiple ) the fundamental invariant of degree 9 , which vanishes if and only if the array has rank @xmath12 ; see table [ ranktable ] below .
littelmann @xcite ( 1989 ) classified the irreducible representations of semisimple lie groups for which the algebra of invariants is free .
gelfand et al .
( 1992 ) introduced a general theory of hyperdeterminants ( modernizing a study initiated by cayley @xcite ( 1845 ) almost 150 years earlier ) , and determined their degrees ( * ? ? ?
* corollary 3.9 ) ; in particular , the hyperdeterminant @xmath13 for @xmath0 arrays is an invariant homogeneous polynomial of degree 36 , and hence can be expressed in terms of the fundamental invariants as follows , for some @xmath14 : @xmath15 ( we note that @xmath6 and @xmath7 are uniquely determined up to nonzero scalar multiples , but @xmath8 is only determined up to adding a scalar multiple of @xmath16 . )
ng @xcite ( 1995 ) described the orbits of @xmath17 acting on the projective space @xmath18 , derived explicit matrix representations for the singular forms , and proved that the group quotient is a projective variety .
nurmiev @xcite ( 2000 ) obtained an implicit description of all three fundamental invariants in terms of convolutions of volume forms , and classified the normal forms ( orbit representatives ) .
briand et al .
18 , 52 , 53 ) ( 2004 ) , motivated by a problem from quantum computing , used cayley s @xmath19-process from classical invariant theory to find explicit forms of the fundamental invariants in terms of concomitants .
duff and ferrara @xcite ( 2007 ) presented an analogy between supersymmetric black holes in 5 dimensions and the bipartite entanglement of three qutrits from quantum information theory , where the common symmetry comes from vinberg s embedding @xmath20 .
explicit computations of these invariants have also been attempted classically .
for instance @xmath21 has been attributed to aronhold , @xmath22 was computed by strassen @xcite ( 1983 ) as a determinant of a certain commutator , and ottaviani @xcite ( 2007 ) gave a determinantal formula . in theory schlfli s method may be used to compute the @xmath23 hyperdeterminant , see gelfand et al .
@xcite ( 1994 ) ; this computation turns out to require a large amount of memory , but it may still be useful for evaluating the hyperdeterminant and was used by one of us @xcite ( 2012 ) . in general , these computations become very large very quickly ; see huggins et al .
@xcite ( 2008 ) for explicit results on the @xmath24 hyperdeterminant .
domokos and drensky ( * ? ? ?
* remark 2(ii ) ) ( 2012 ) computed the single defining relation of the algebra of invariants for the action of @xmath25 on triples of @xmath26 matrices , and provided an alternative proof of vinberg s result .
allman et al .
@xcite ( 2013 ) gave a generalization of cayley s hyperdeterminant for @xmath27 arrays to a covariant for the action of @xmath28 on @xmath29 arrays .
two of the present authors recently used computer algebra @xcite ( 2013 ) to determine the fundamental invariants of degrees 6 , 9 and 12 in terms of orbit sums for the symmetry group @xmath30 .
each invariant is a homogeneous polynomial in the variables @xmath31 for @xmath32 : * the polynomial @xmath6 of degree 6 from ( * ? ? ?
* table 5 ) , which is a linear combination of 8 symmetric orbits , has 1152 terms ; * the polynomial @xmath7 of degree 9 from ( * ? ? ?
* table 6 ) , which is the sum of 14 alternating orbits , has 9216 terms ; * the polynomial @xmath8 of degree 12 from ( * ? ? ?
* tables 7 - 11 ) , which is a linear combination of 235 symmetric orbits , has 209061 terms . in this paper
we determine the coefficients @xmath33 in equation .
a similar explicit computation of the lroth invariant in terms of fundamental invariants was recently achieved by basson et al .
@xcite ( 2013 ) , although their task was more difficult : the algebra of invariants for homogeneous polynomials of degree 4 in three variables is not freely generated .
interest in lroth quartics was revived a few years ago by ottaviani and sernesi @xcite ; see ottaviani @xcite ( 2012 ) for the computational aspects . in [ gaussianelimination ]
we recall a method for generating arrays of hyperdeterminant 0 which applies to any format @xmath1 . in [ hdsection ]
we repeatedly generate pseudorandom @xmath0 arrays @xmath34 of hyperdeterminant 0 , substitute their entries for the variables in the fundamental invariants , and set the resulting polynomial to 0 ; in this way , we obtain linear equations satisfied by the coefficients @xmath33 in . in [ nurmiev ]
we apply our results to nurmiev s classification of normal forms .
a similar computation for the so - called trifocal variety was carried out by one of us for low degree covariants in @xcite ( 2012 ) . in [ ranksection ]
, we test and determine which invariants vanish on which ranks . our maple worksheet for the computations described in the present paper is available as an ancillary file with the ` arxiv ` version .
we recall that a @xmath35-dimensional array @xmath36 is said to have rank 1 if it is the outer product of nonzero vectors @xmath37 in the sense that @xmath38 .
the array @xmath39 has rank @xmath40 if @xmath40 is the least number such that @xmath39 can be written as a sum of @xmath40 arrays of rank 1 .
the hyperdeterminant @xmath41 of a @xmath35-dimensional array of size @xmath42 with @xmath43 for @xmath44 is ( by definition ) the polynomial of minimal degree vanishing on the dual variety of the arrays of rank 1 .
( limits of arrays of rank 1 still have rank @xmath45 , so the set of rank 1 tensors is closed in projective space . )
if we assume that @xmath46 is primitive in the sense that its coefficients are integers with no common factors , then @xmath46 is unique up to a sign . using basic multilinear algebra , we can apply a multilinear change of coordinates , so that the arrays in the dual variety have zeros on the edges emanating from one corner and arbitrary scalars in the other positions . with this description
, we can easily generate pseudorandom arrays of hyperdeterminant 0 of any size @xmath42 . to remove dependence on our choice of coordinates
, we can apply pseudorandom changes of basis along each of the @xmath35 directions .
however , by definition of invariant , these changes of basis will not affect the values of the fundamental invariants .
so for the purposes of the present paper , this last step will not be necessary . in this section
we review how to perform a multidimensional generalization ( i.e. to tensors ) of gaussian elimination for matrices , using a projective point of view .
let @xmath47 be vector spaces over @xmath48 , with @xmath49 for @xmath50 .
the corresponding projective spaces @xmath51 have dimensions @xmath52 ; @xmath53 consists of all lines through the origin in @xmath54 .
for a nonzero vector @xmath55 the corresponding element of @xmath53 will be denoted @xmath56 $ ] ; this is the line with direction vector @xmath57 .
the tensor product @xmath58 has dimension @xmath59 , and the projective space @xmath60 has dimension @xmath61 .
an element of @xmath58 is a sum of simple tensors @xmath62 .
the * segre embedding * is the analogue of the tensor product in the setting of projective spaces : @xmath63 , \dots , [ a_k ] \ , ) \mapsto [ \ , a_1 \otimes \cdots \otimes a_k \ , ] .\ ] ] after a choice of basis , an array of size @xmath64 can be identified with an element of @xmath58 .
we denote the image of the segre embedding by @xmath65 .
consider an array of rank 1 in @xmath65 , which by definition has the form @xmath66 $ ] for nonzero @xmath55 . a parametrized curve in @xmath65 through the point @xmath66 $ ] has the form @xmath67 $ ] where @xmath68 for @xmath50 .
we differentiate with respect to @xmath69 and then set @xmath70 to obtain @xmath71 \xrightarrow{\ ; t = 0 \ ; } \big [ \ , \sum_{i=1}^k a_1 \otimes \cdots \otimes a'_i(0 ) \otimes \cdots \otimes a_k \ , \big].\ ] ] the cone over the tangent plane to @xmath65 at the point @xmath66 $ ] is the span of these tangent vectors : @xmath72 we convert this sum of vector spaces into a direct sum by separating the common 1-dimensional subspace : @xmath73 \qquad\qquad\qquad \bigoplus_{i=1}^k ( a_1 \otimes \cdots \otimes a_i \otimes \cdots \otimes a_k ) / \mathbb{c}(a_1 \otimes \cdots \otimes a_k ) .
\end{array}\ ] ] we want to construct the projective dual variety @xmath74 . for an irreducible projective variety @xmath75 , the * projective dual variety *
@xmath76 is the zariski closure of the set of all tangent hyperplanes ; i.e. hyperplanes @xmath77 \subset { \mathbb{p}}(v^\vee)$ ] such that @xmath78 for a smooth point @xmath79 \in x$ ] . by a result of gelfand
@xcite , the dual of the segre variety is a hypersurface if and only if @xmath80 for @xmath44 .
projective duality preserves irreducibility , so in this case the dual of the segre variety is a hypersurface defined by a single equation @xmath81 , where the homogeneous polynomial @xmath46 is known as the hyperdeterminant of type @xmath42 ; again see @xcite .
a hyperplane @xmath77 $ ] in projective space @xmath82 can be identified with a normal line @xmath83 in the dual projective space @xmath84 .
the condition that @xmath85 contains the tangent plane @xmath86 is equivalent to the condition that the corresponding line @xmath83 annihilates @xmath86 in the natural pairing of a variety with its dual . to work in terms of coordinates
, we choose bases : @xmath87 the general element of @xmath88 then has the form @xmath89 by equation , the general element of @xmath86 has the following form where @xmath90 : @xmath91 we evaluate the dual vector at the tangent vector and obtain @xmath92 where the subscript @xmath93 of @xmath94 is in position @xmath95 .
this must vanish for all @xmath96 and @xmath97 which gives the necessary and sufficient condition that @xmath98 when @xmath99 ( or all @xmath35 ) of the subscripts @xmath100 equal 1 .
let @xmath101 be a @xmath35-dimensional array .
the following are equivalent : * the hyperdeterminant @xmath46 of @xmath39 is 0 . * the projectivization of @xmath39 lies in the dual @xmath74 of the segre variety .
* @xmath39 is conjugate by a multilinear change of basis in @xmath102 to an array in which @xmath98 when @xmath99 of the subscripts equal 1 . in the familiar case @xmath103 and @xmath104 ,
a square matrix @xmath39 is singular if and only if there exist nonzero vectors @xmath105 and @xmath106 for which @xmath107 and @xmath108 .
for @xmath109 we choose a basis of @xmath54 which has @xmath57 as its first vector .
with respect to these bases , the matrix @xmath39 is equivalent to a matrix with 0s in its first row and column , and no condition on the remaining entries : @xmath110\ ] ] thus @xmath111 if and only if there exist invertible operators @xmath112 for @xmath109 such that @xmath113 has this form .
[ maintheorem ] the @xmath0 hyperdeterminant has the following explicit form in terms of the fundamental invariants @xmath6 , @xmath7 and @xmath8 as given in @xcite : @xmath114 using the computer algebra system maple , we generate a pseudorandom @xmath0 array @xmath39 with entries in the field with @xmath115 elements ( the smallest prime @xmath116 ) .
we then set the @xmath117 entry equal to 0 whenever at least two of the indices @xmath118 equal 1 .
we compute the values modulo @xmath119 of the fundamental invariants @xmath2 and then of the invariant monomials in equation . if any value is 0 , we repeat the process until all values are nonzero . @xmath120 \qquad \left [ \begin{array}{rrrrrrr } 1 & 0 & 0 & 0 & 0 & 0 & 0 \\ 0 & 1 & 0 & 0 & 0 & 0 & 0 \\ 0 & 0 & 1 & 0 & 0 & 0 & 7818 \\ 0 & 0 & 0 & 1 & 0 & 0 & 2189 \\ 0 & 0 & 0 & 0 & 1 & 0 & 1252 \\ 0 & 0 & 0 & 0 & 0 & 1 & 3756 \end{array } \right]\ ] ]
we perform this procedure 10 times and obtain the first matrix in table [ ptable ] ; this is the coefficient matrix modulo @xmath119 of a homogeneous linear system in the variables @xmath33 in equation .
if we had done this computation using rational arithmetic , and had generated enough pseudorandom arrays , the matrix would have a nullspace of dimension 1 with the hyperdeterminant as basis . hence using modular arithmetic , the nullity is at least 1 .
in fact , we find that the nullity is exactly 1 ; the row canonical form ( rcf ) of the matrix modulo @xmath119 is the second matrix in table [ ptable ] , and the vector @xmath121 $ ] is a basis of the nullspace . using the maple procedure ` iratrecon ` for rational reconstruction
, we find that the simplest rational numbers corresponding to these residue classes modulo @xmath119 are @xmath122 = \left [ \ ; 0 , \ ; 0 , \ ; -\frac{1}{32 } , \ ; \frac{1}{32 } , \ ; -\frac98 , \ ; -\frac{27}{8 } , \ ; 1 \
; \right]\ ] ] we multiply by @xmath123 to obtain the coefficients in the statement of this theorem .
we will confirm this calculation using rational arithmetic in section [ ranksection ] .
nurmiev @xcite classified the normal forms of @xmath0 arrays over @xmath48 with respect to the action of the lie group @xmath124 , and in nurmiev @xcite described the orbit closure poset .
the @xmath0 array @xmath39 is called * semisimple * if its @xmath125-orbit is closed , and * nilpotent * if the closure of its @xmath125-orbit contains the zero array .
the variety where all the invariants vanish is called the nullcone @xcite , and it coincides exactly with the variety of nilpotent arrays .
it is immediate that tensors of rank @xmath126 are nilpotent .
every @xmath0 array @xmath39 can be written uniquely in the form @xmath127 where @xmath128 is semisimple , @xmath129 is nilpotent , and @xmath130 = 0 $ ] where the lie bracket is taken from vinberg s embedding of the vector space of @xmath0 arrays into a simple lie algebra of type @xmath5 .
nurmiev defines three basic semisimple arrays ; we use the notation @xmath131 for the array with 1 in position @xmath117 and 0 in the other positions : @xmath132 nurmiev shows that every semisimple array is equivalent to ( i.e. lies in the @xmath125-orbit of ) an array of the normal form @xmath133 for each type of orbit , non - vanishing of an invariant at a particular point is always a certainty , so evaluating at pseudorandom points is sufficient . to be certain of vanishing for all points on a parametrized orbit
, we must check the expression for the invariant restricted to the parametrized normal form . evaluating the three fundamental invariants and
the hyperdeterminant on @xmath134 gives the following results : @xmath135 i_9 & = - ( a_1 - a_2)(a_1 - a_3)(a_2 - a_3 ) ( a_1 ^ 2 + a_1a_2 + a_2 ^ 2)(a_1 ^ 2 + a_1a_3 + a_3 ^ 2)(a_2 ^ 2 + a_2a_3 + a_3 ^ 2 ) , \\[4pt ] i_{12 } & = a_1 ^ 3a_2 ^ 9 + a_1 ^ 9a_2 ^ 3 + a_1 ^ 9a_3 ^ 3 + a_1 ^ 3a_3 ^ 9 + a_2 ^ 3a_3 ^ 9 + a_2 ^ 9a_3 ^ 3 - 4a_1 ^ 6a_2 ^ 6 - 4a_1 ^ 6a_3 ^ 6 - 4a_2 ^ 6a_3 ^ 6 \\[4pt ] & \qquad + 2a_1 ^ 3a_2 ^ 3a_3 ^ 6 + 2a_1 ^ 3a_2 ^ 6a_3 ^ 3 + 2a_1 ^ 6a_2 ^ 3a_3 ^ 3 , \\[4pt ] \delta & = - 4a_1 ^ 3a_2 ^ 3a_3 ^ 3 ( a_1 + a_2 + a_3)^3 ( a_1 ^ 2 + a_2 ^ 2 + a_3 ^ 2 + 2a_1a_2 - a_1a_3 - a_2a_3)^3 \,\times \\[4pt ] & \qquad ( a_1 ^ 2 + a_2 ^ 2 + a_3 ^ 2 - a_1a_2 + 2a_1a_3 - a_2a_3)^3 ( a_1 ^ 2 + a_2 ^ 2 + a_3 ^ 2 - a_1a_2 - a_1a_3 + 2a_2a_3)^3 \,\times \\[4pt ] & \qquad ( a_1 ^ 2 + a_2 ^ 2 + a_3 ^ 2 - a_1a_2 - a_2a_3 - a_1a_3)^3 . \end{array } \quad\right\}\ ] ] we note that @xmath7 vanishes whenever two of the coefficients @xmath136 are equal , and that @xmath46 vanishes whenever any coefficient is 0 or the sum @xmath137 is 0 . according to the nurmiev classification ,
the normal forms belong to five families .
the coefficients of the semisimple part satisfy @xmath138 the nilpotent part is 0 , so the values of the invariants on this family are as in .
the coefficients of the semisimple part satisfy @xmath139 the possible nilpotent parts are @xmath140 it turns out that the values of the fundamental invariants and the hyperdeterminant do not depend on the nilpotent part , and for the second family we find that : @xmath141 the coefficients of the semisimple part satisfy @xmath142 the possible nilpotent parts are @xmath143 we find that the values of the fundamental invariants and the hyperdeterminant do not depend on the nilpotent part ; for all normal forms in this family we have @xmath144 and @xmath145 .
the coefficients of the semisimple part satisfy @xmath146 the possible nilpotent parts are @xmath147 the values of the fundamental invariants and the hyperdeterminant do not depend on the nilpotent part ; for all normal forms in this family we have @xmath148 this family contains 24 nonzero nilpotent arrays ( the semisimple part is 0 ) ; see nurmiev ( * ? ? ? * table 4 , p. 723 ) . in every case all the fundamental invariants and the hyperdeterminant are 0 . to put this another way ,
we have verified that all nilpotent orbits consist of singular arrays ( from the point of view of hyperdeterminants ) .
these results on the evaluation of the invariants on the normal forms can be used to corroborate results in the second paper of nurmiev @xcite on the structure of the poset of the closures of nilpotent orbits .
in this section we determine which invariants vanish on arrays of each possible rank . in principle this could be deduced from our previous computations by determining the rank of each of nurmiev s normal forms , but we prefer to argue directly . kruskal s theorem @xcite states that every @xmath0 array over @xmath149 or @xmath48 has rank at most 5 .
to parametrize an array of rank @xmath150 ( strictly speaking , @xmath151 ) we take the outer product of three vectors : @xmath152 . to make the array pseudorandom , we assign pseudorandom values to each of the coordinates of each of the vectors in the outer product . then
to generate parametrized ( respectively pseudorandom ) arrays of rank @xmath153 we take the sum of @xmath40 parametrized ( respectively pseudorandom ) arrays of rank 1 .
let @xmath154 denote the set of arrays of rank @xmath155 and let @xmath156 denote its ( zariski ) closure .
it is well known that @xmath157 for all @xmath40 , and obviously @xmath158 ; however it can be that @xmath159 .
the commonly used example is @xmath160 , which is in @xmath161 ( and @xmath162 ) but not in @xmath163 .
we also note that since the zero - locus of an invariant is both ( zariski ) closed and invariant under the group action , if an invariant vanishes at one point of an orbit , then it does so for all points in the orbit , and for all points in the orbit closure .
we use this fact implicitly in what follows . for @xmath163 , we may consider the representative @xmath164^{\otimes 3 } + [ 0,1,0]^{\otimes 3}$ ] .
all other arrays of rank 2 are in the closure of the orbit of this array .
our tests find that all three fundamental invariants vanish on these representatives .
similarly , for @xmath162 we may consider the representative @xmath164^{\otimes 3 } + [ 0,1,0]^{\otimes 3 } + [ 0,0,1]^{\otimes 3}$ ] .
all other arrays of rank 3 are in the closure of the orbit of this array .
we find that @xmath21 does not vanish on this representative , but @xmath165 and @xmath46 do .
experimental results for 10000 pseudorandom arrays with entries in @xmath166 and rank @xmath167 showed that @xmath6 vanishes in only 188 cases , but @xmath7 and @xmath8 vanish in all 10000 cases . for @xmath168 ,
the value of @xmath21 was already non - zero for rank @xmath169 , so it must not vanish identically on rank 4 arrays . it was already determined by strassen @xcite that @xmath22 vanishes on arrays of rank @xmath12 .
experimental results for 10000 pseudorandom arrays with entries in @xmath166 and rank @xmath12 showed that @xmath6 vanishes in only 4 cases , @xmath7 vanishes in all 10000 cases , and @xmath8 vanishes in 573 cases .
in fact , just one non - vanishing result is sufficient to indicate that @xmath8 does not vanish identically on the locus of rank 4 arrays .
a similar test also shows that @xmath46 does not vanish identically for rank 4 arrays .
the non - vanishing of @xmath8 and @xmath46 on @xmath168 can also be deduced from strassen s result that @xmath170 is an irreducible hypersurface defined by @xmath22 and that neither @xmath8 nor @xmath46 are multiples of @xmath22 .
strassen showed that his degree 9 invariant was non - zero for rank 5 arrays , thus proving that his invariant was non - trivial .
the closure of rank 5 arrays is all of the ambient projective space , so no non - trivial invariant vanishes identically on all rank 5 arrays .
experimental results for 10000 pseudorandom arrays with entries in @xmath166 and rank @xmath171 showed that @xmath6 never vanishes , @xmath7 vanishes in 47 cases , and @xmath8 vanishes in 23 cases .
again , just one non - vanishing example is sufficient to prove that the invariant is not identically zero on @xmath172 .
we summarize these results in table [ ranktable ] .
we can use this method of pseudorandom arrays to reduce the size of the matrix entries when we perform the computation of [ hdsection ] using rational arithmetic .
there are 7 nonzero vectors of dimension 3 with entries in @xmath174 , and hence @xmath175 nonzero @xmath0 arrays of rank 1 .
we use a pseudorandom number generator to choose 5 of these arrays ; their sum @xmath176 is an array of rank @xmath171 , which has rank 5 with sufficiently high probability ; in this case , all three fundamental invariants will have nonzero values on @xmath65 .
we set @xmath177 whenever at least two of @xmath118 equal 1 , to ensure that @xmath65 has hyperdeterminant 0 .
following the same procedure as in [ hdsection ] , we obtain the matrix in table [ char0matrix ] .
the vector of equation is a basis for the nullspace , confirming theorem [ maintheorem ] with rational arithmetic .
r. basson , r. lercier , c. ritzenthaler , j. sijsling : an explicit expression of the lroth invariant .
_ proceedings of issac 13 : the 38th international symposium on symbolic and algebraic computation_. pages 3136 .
acm , new york , 2013 .
i. m. gelfand , m. m. kapranov , a. v. zelevinsky : _ discriminants , resultants , and multidimensional determinants_. reprint of the 1994 edition .
modern birkhuser classics .
birkhuser boston , inc . ,
boston , ma , 2008 .
isbn : 978 - 0 - 8176 - 4770 - 4 g. ottaviani : symplectic bundles on the plane , secant varieties and lroth quartics revisited . _ vector bundles and low codimensional subvarieties : state of the art and recent developments _ , pages 315352 . _ quad .
_ , 21 , dept .
, seconda univ .
napoli , caserta , 2007 . | we briefly review previous work on the invariant theory of @xmath0 arrays .
we then recall how to generate arrays of arbitrary size @xmath1 with hyperdeterminant 0 .
our main result is an explicit formula for the @xmath0 hyperdeterminant as a polynomial in the fundamental invariants @xmath2 for the action of the lie group @xmath3 .
we apply our calculations to nurmiev s classification of normal forms for @xmath0 arrays . |
the role of mass loss in the evolution of the most massive stars with initial masses of @xmath060150 m@xmath1 is still poorly understood , partly because the most luminous stars are so rare and each one seems unique at some level , and partly because accurate physical parameters are difficult to determine .
consequently , the connection between spectral characteristics and evolutionary state is often mangled .
one persistent mystery concerns the placement of the very luminous late - type h - rich wn stars ( referred to here as `` wnh stars '' , as justified below ) in the evolutionary sequence of very massive stars ( see crowther et al .
1995a ; hamann et al .
2006 ; langer et al .
because of their high mass - loss rates and consequent emission - line spectra , they are often discussed alongside or confused with core - he burning wolf - rayet ( wr ) stars ( for reviews of wr stars , see abbott & conti [ 1987 ] and crowther [ 2007 ] ) , and it is sometimes suggested that their h content indicates that they represent the early phases of core - he burning . in this interpretation , the wnh phase occurs immediately after or sometimes instead of the luminous blue varibale ( lbv ) phase , marking the beginning of the core - he burning wr stages ( see for example , schaller et al .
1993 ; meynet et al .
1994 ; maeder & meynet 1994 ) .
a different suggestion has been made based on the very high luminosity of some of the wnh stars and especially based on their membership in very young massive star clusters like r136 in 30 doradus , ngc 3603 , and the carina nebula ( e.g. , de koter et al .
1997 ; drissen et al.1995 ; crowther et al . 1995a ; moffat & seggewiss 1979 ) .
namely , these authors suggest that it may be the case instead that the wnh stars are essentially core - h burning stars that precede the lbv phase .
this suggestion has also been made based on the continuity in spectral types from o3 v to o3 if * to wnl+h as noted , for example , by walborn ( 1971 , 1973 , 1974 ) , walborn et al . ( 2002 ) , lamers & leitherer ( 1993 ) , drissen et al .
( 1995 ) and crowther et al .
( 1995a ) .
this continuity has led to the suggestion that the most luminous stars may even _ skip _ the lbv phase altogether ( crowther et al .
i.e. that they evolve directly from o stars to wr stars ( conti 1976 ) . however , this direct path is not taken in all cases because it is contradicted by the existence of very luminous lbvs such as @xmath2 carinae and the pistol star .
langer et al . (
1994 ) have discussed the two possibilities above from the perspective of evolution models , and they suggest instead that both may be true i.e. , that a star may look like a wnh star both before _ and _ after the lbv phase .
langer et al . ( 1994 ) investigated a 60 m@xmath1 initial mass model , and as we will see later , the distinction between stars of 60 and 100 m@xmath1 may be important in terms of their evolutionary path .
our main conclusion in this paper is that available evidence strongly favors the latter interpretation above that wnh stars are pre - lbvs , at least for the highest luminosities above log(l / l@xmath1)=5.86.0 .
as discussed below , this is based on the fact that their distributions of stellar mass are distinct from h - poor stars and that their hydrogen mass fractions are generally higher than lbvs ( although there is a caveat to this last point depending on initial mass , as discussed below ) .
we also argue that one might _
expect _ them to be pre - lbvs if feedback from mass loss on the main - sequence is accounted for in a simple way . in
2 we justify the new naming convention `` wnh '' , in 3 we examine the masses of wnh stars , and in 4 we examine their hydrogen mass fractions compared to lbvs . then in 5 we take a look at how mass loss on the main - sequence might naturally lead to a wnh - like phase , and how that higher mass - loss phase might drive the star toward the eddington limit to become an lbv . in the last few sections we discuss further related implications of the scenario where wnh stars are pre - lbvs .
we propose the new designation `` wnh '' for luminous wr - like stars of the nitrogen sequence exhibiting evidence for hydrogen in their spectra . in current practice , these are commonly referred to collectively as `` wnl stars '' , with individual objects classified as wn6 ha , wn7h , etc .
( see smith et al .
1994 ; smith et al . 1996
; crowther et al .
. we do not propose to change this more specific spectral classification scheme for individual stars .
however , we do advocate that the common usage of `` wnl stars '' as a group should be changed to `` wnh stars '' for two main reasons , one being a practical matter and the other having to do with their evolutionary state and questionable relationship to other wr stars .
the terms wnl , wne , wcl , and wce were first used by vanbeveren & conti ( 1980 ) as a shorthand for `` late '' and `` early '' type wr stars of the nitrogen and carbon sequences .
its use was quickly adopted as a convenience by other authors .
later on , the wnl term came to mean late - type wn stars with hydrogen as most of them have this element present .
however , observations have firmly established that there are both late - type wn stars _ without _ hydrogen ( the _ bona fide _ wn stars ) , and , more problematic , _ early_-type wn stars with hydrogen
. a number of early - type ( wn3 ha , wn4 ha , wn5 ha ) wn stars with hydrogen have been identified recently in the smc ( foellmi et al.2003a ; foellmi 2004 ) , and some in the lmc as well ( foellmi et al.2003b ) .
one such wne star with hydrogen is the famous object hd5980 , an eclipsing massive binary where the hydrogen - rich component had been classified as wn6h , wn11 , lbv , or b1.5 ia at different times during its lbv - like outburst in the 1990s ( koenigsberger 2004 ) , but had been classified as wn4h before that .
wn5h is arguably on the border between early- and late - type , and several of these have been seen in 30 doradus and westerlund 1 ( crowther & dessart 1998 ; crowther et al .
other galactic examples are wr3 ( wn3 ha ) , wr10 ( wn5 ha ) , wr48c ( wn3h+wc4 ) , wr49 ( wn5h ) , wr109 ( wn5h ) , wr128 ( wn4h ) , and wr152 ( wn3h ) ( marchenko et al . 2004 ; hamann et al .
2006 ; van der hucht 2001 ) .
clearly , the phenomenon of hydrogen in wn stars is not limited to late types , rendering `` wnl '' an inappropriate designation as it refers mainly to wn7/wn8 ( moffat & seggewiss 1979 ) .
the shorthand `` wnh '' would encompass both early- and late - type wn stars with hydrogen .
second , the designation `` wnh '' is also useful to emphasize the apparent fact that while they exhibit wr - like spectral features because of their strong winds ( they are sometimes referred to as `` o - stars on steroids '' ) , the wnh stars are distinct from core - he burning wr stars in several observed characteristics and most probably in their evolutionary state .
the assertion that they represent a distinct evolutionary state is justified in the following sections , where we show that like lbvs , the wnh stars are consistently more massive than wn and wc stars , and that they have h mass fractions more in line with lbvs .
we argue that their observed masses , mass - loss rates , luminosities , wind speeds , and other properties can all be understood naturally if they represent the later stages of core - h burning of very massive stars with ages @xmath02 myr , making them immediate precursors of lbvs , or perhaps , quiescent lbvs at lower luminosities .
lastly , the high mass - loss rates seen in the wnh phase might be a _
necessary condition _ to push the star toward the eddington limit , again arguing that wnh stars are the best candidates for the immediate precursors of high luminosity lbvs .
we reiterate that we intend the term `` wnh '' only as collective shorthand for the class of wn - like stars that show hydrogen in their spectra , because we believe them to be distinct from h - free wn stars . while this new terminology is arguably imperfect
because one star or another might challenge categorization or blurr the boundaries , the term is needed simply to avoid confusion , and it makes a clear distinction that is also useful for purposes of discussion . because they show a range in h mass fraction ( see below ) , the wnh stars constitute a somewhat heterogeneous class and may even overlap with the lbvs . in that sense , though , the term `` wnh '' has a usefulness similar to that of `` lbv '' , since the lbv class contains stars labeled variously as s doradus variables , @xmath2 car variables , @xmath3 cygni variables , p cygni stars , ofpe / wn9 stars , etc . , yet we still refer to them all collectively as `` lbvs '' .
the term `` lbv '' has been useful for broadly discussing stars that we think share a common evolutionary phase , even though it is not always agreed whether an individual object is a _ bona - fide _ lbv or not .
the term `` wnh '' is also critical to distinguish these stars from h - free wn stars , because massive stars are the progenitors of various types of core - collapse supernovae ( sn ) .
the single most important observable trait in classifying a sn spectrum is the presence or absence of hydrogen , making it a type ii or a type ib / c , respectively .
the `` h '' in `` wnh '' therefore serves as a clear reminder that hydrogen is present , and that these are _ not _ to be confused with the h - free wolf - rayet stars that are the likely progenitors of type ib supernovae and possibly also long - duration gamma ray bursts .
the more specific classification criteria of wn stars are not relevant or useful in this context as they mainly describe the temperature and ionization level of the wind .
observationally , one can not determine if the progenitor of a sn is wne or wnl , but in pinciple , one can tell the difference between wn and wnh .
figure 1 collects mass estimates for wr stars from the literature , including masses measured both spectroscopically and in binary systems .
it compares present masses of wnh stars to h - poor wn and wc stars ( spectroscopic masses are unreliable for wc stars ) .
the spectroscopic masses are taken mostly from hamann et al .
( 2006 ) , who derived the masses using the mass - luminosity relation for helium stars from langer ( 1989 ) .
they noted that this relation might not be adequate for wnh stars , but they also note that when masses measured in binary systems are available for the same stars for which they estimated a spectroscopic mass , the two methods show no wild disagreement . for example , the spectroscopic mass they derive for the wnh star wr22 in the carina nebula is 74 m@xmath1 , compared to 72 m@xmath1 determined from the binary orbit by rauw et al.(1996 ) . for the wnh component . ]
thus , while the exact values for the masses in the top panel of figure 1 might be somewhat suspect , the main trend of the separation between wn and wnh stars is not .
masses measured for wr stars in binary systems are more reliable , but there are not many of them available , shown in the bottom panel of figure 1 .
the results for the distribution of masses for different wr types measured spectroscopically and in binaries show very good general agreement .
the clear lesson to be learned from figure 1 is that wnh stars have systematically higher masses than h - poor wn and wc stars . with binaries and spectroscopic masses , the mass distribution of h - poor wn stars clearly peaks at 1520 m@xmath1 , while wnh stars are spread more evenly over a large range of masses mostly above 30 m@xmath1 .
there are only a few cases of mass overlap between the wn stars and a small fraction of the wnh stars
. this may be because of wn stars in binary systems where h is present , or it may signify a real evolutionary overlap . note that the overlap is more prevalent in the spectroscopic masses where unrecognized binaries are more likely to contaminate the sample , and where the general shape of the wnh mass distribution in the 1025 m@xmath1 range matches that for wn stars , as one might expect if they are not true wnh stars .
all the h - poor wr stars have relatively low masses below 30 m@xmath1 . estimated by moffat et al .
( 1990 ) , because this system has not been studied in enough detail yet to determine if h is present or not .
( see also karteshiva 2002 , who estimate its mass as only 812 m@xmath1 . ) ] the wnh star masses peak around 50 m@xmath1 ( for the more numerous spectroscopic masses ) , and masses measured in binaries extend to very large masses as high as more than 120 m@xmath1 for the wnh star r145 in 30 doradus ( moffat 2006 ; private comm . ) .
overall , then , figure 1 makes a strong case that wnh stars are in fact a separate and distinct group of stars from the h - poor wr stars .
it is obviously logical to assume that wnh stars are more massive than wn and wc stars because they have not yet shed their h envelopes , and that they may eventually become h - poor wr stars . in that case
wnh stars must be at a significantly earlier evolutionary stage , which is consistent with the fact that they are preferentially seen in massive clusters within giant h ii regions like the carina nebula , ngc 3603 , and 30 doradus , whereas h - free wn stars are not .
on the other hand , wnh stars may be consistently more massive than wn or wc stars simply because they evolve from stars with higher initial masses and follow different evolutionary paths .
furthermore , the wide mass discrepancy between wnh and h - poor wr stars and the fact that the two populations hardly overlap at all provides a vital clue to their evolutionary states .
it argues strongly _ against _ the notion that wnh stars are simply the initial stages of the core - he burning wr phase , where they are still in the process of shedding the last remaining layers of their hydrogen envelopes , gradually transitioning into h - free wn stars .
( besides , shedding the required 2080 @xmath4 in @xmath510@xmath6 yr in this transition requires mass - loss rates higher than observed for wnh stars . ) instead , the discontinuity in mass of 10 s of solar masses between wnh and wn stars suggests that some other intermediate stage must quickly remove the large mass remaining in the hydrogen envelopes of the wnh stars before they can become wn stars .
the obvious choice for the subsequent removal of that mass is the violent eruptions of lbvs ( smith & owocki 2006 ) .
in other words , based on their masses , wnh stars are likely to be pre - lbvs , not post - lbv stars .
interestingly , unlike the h - poor wr stars , lbvs exhibit a range of masses ( with current masses of roughly 30100 m@xmath1 ) that does overlap with that of the wnh stars , arguing that the two are closely related .
we also note a systematic difference in the masses of wc and wn stars in figure 1 .
the wc star mass distribution peaks in the 515 m@xmath1 range , whereas the wn distribution peaks at higher masses in the 1520 m@xmath1 range . one might expect this if wc stars are more evolved descendants of wn stars ( paczynski 1973 ) , as the products of further mass stripping by the powerful wn stellar wind ( conti 1976 ) or sudden events like that inferred for the progenitor of sn 2006jc ( smith et al . 2008 ;
pastorello et al . 2007 ; foley et al.2007 ) .
a prediction of this trend is that sne of type ib should result from more massive progenitors than type ic sne .
if not , then an interesting mystery needs to be solved . in any case , the disparity in mass between the wnh stars and all the other wr types ( fig .
1 ) supports the main thrust of this paper , and is obviously relevant for interpreting the progenitors of types ib and ic sne .
figure 2 shows the hydrogen mass fraction , @xmath7 , as a function of stellar luminosity for wnh stars compared to lbvs .
this is obviously an important quantity for unraveling the evolutionary relationship between these two classes of stars . a similar plot and its implications have already been discussed by langer et al .
( 1994 ) , but we update it here with several additional values from the literature .
this updated information is most relevant at the highest luminosities , as there were only a few data points above log ( l / l@xmath1)=6.0 in the similar plot by langer et al .
the behavior of those high - luminosity objects is the focus here . excluding @xmath2 car and the pistol star , the lbvs seem to cluster around @xmath80.35 , while the wnh stars occupy a large range of @xmath7 values from 0.05 to 0.5 , above and below the values for lbvs .
however , the way that this range of @xmath7 values for wnh stars compares to those of the lbvs changes with luminosity , and it depends on which set of wnh luminosities is adopted . in cases
when both crowther et al .
( 1995a ) and hamann et al .
( 2006 ) analyzed the same target stars , values for @xmath7 generally agree to within a few percent if they differ at all .
the disagreement is in the stellar luminosities , with the luminosities from hamann et al .
( 2006 ) being systematically higher by roughly 0.5 dex for the same target stars .
this may be due to the fact that hamann et al .
_ assumed _ that wnh stars had relatively high luminosities . shifting the wnh stars horizontally in figure 2 significantly impacts our interpretation of the relative evolutionary status of wnh stars compared to lbvs .
figures 2@xmath9 and 2@xmath10 suggest the following different implications , respectively : \1 ) figure 2@xmath9 presents a suggestive picture that the relationship between wnh stars and lbvs is dependent on luminosity ( and hence , on initial mass ) . below log ( l / l@xmath1)=5.8 ( the upper limit for rsgs and for h - free wr stars )
, we see that the wnh stars are generally more evolved with lower hydrogen content than lbvs . at an intermediate luminosity range of log ( l / l@xmath1)=5.86.0 , the case is less clear ; wnh stars bracket the @xmath7 values for lbvs , suggesting that they could be both pre- and post - lbvs , as in the scenario suggested by langer et al .
( 1994 ) for a star of initial mass 60 m@xmath1 corresponding to p cygni . at high luminosities above log ( l / l@xmath1)=6.0 ,
however , no wnh stars are seen to have smaller hydrogen mass fractions than lbvs
they generally lie above the lbvs , indicating that they should be pre - lbvs .
\2 ) with the higher wnh luminosites in figure 2@xmath10 ( hamann et al .
2006 ) , on the other hand , one would conclude that the lbvs are intermixed with and are essentially indistinguishable from wnh stars in their abundances , except that lbvs occupy a narrower range .
this would imply that wnh stars can be both pre- and post - lbvs , regardless of luminosity .
given the comments above , this is a fair possibility at lower luminosities , but seems less likely for the high - luminosity stars above log ( @xmath11/@xmath12)@xmath135.8 . at this time
, we can not be certain which of these luminosity estimates to trust , and this highlights the need for accurate estimates of @xmath11 and @xmath7 in order to understand the evolution of massive stars .
there are a few reasons to favor the somewhat lower luminosities of crowther et al .
( 1995a ) and the corresponding interpretation implied by figure 2@xmath9 .
first , for the carina nebula wnh stars wr22 , 24 , and 25 ( shown with bold triangles in figs . 2@xmath9 and 2@xmath10 ) , the lower luminosities of crowther are in much closer agreement with the spectroscopically similar wnh stars in r136 and ngc 3603 ( de koter et al .
1997 ; drissen et al .
1995 ) , shown with squares in figure 2 . also , the higher luminosity for wr25 from hamann et al .
( 2006 ) makes this star more luminous than @xmath2 car , making it difficult to understand why @xmath2 car is so much more unstable and more evolved with a larger n abundance even though they are members of the same star cluster .
second , the implication in figure 2@xmath9 that wnh stars precede the lbv phase at high luminosities ( above log l / l@xmath1=5.86.0 ) is in much better agreement with the conclusions drawn from considering their current ( high ) masses discussed in the previous section , as well as the discussion of the mass - loss rates to follow . in figure 2@xmath10 , on the other hand , the occurrence of many high luminosity wnh stars with @xmath7 values significantly lower than lbvs is at odds with the impression that they typically have higher stellar masses than lbvs .
this argument is admittedly somewhat circular .
therefore , we can not consider the measured abundances as giving any _ definite _ answer to the relative evolutionary states of wnh stars and lbvs we can only say that the scenario in figure 2@xmath9 fits together in a more consistent way with the idea that high - luminosity wnh stars are pre - lbvs , while the implication from figure 2@xmath10 that wnh stars are both pre- and post - lbvs would present an unresolved puzzle ( which may nevertheless be true ) .
one last reason to favor figure 2@xmath9 concerns the environments in which these stars are found .
the most luminous wnh stars with higher h content than lbvs reside in massive giant h ii regions like 30 dor , carina , and ngc 3603 .
the wnh stars with lower h conent than lbvs ( and at lower @xmath11 ) do not reside in such massive young regions ( like most of the lbvs , incidentally ) .
this argues for a lower initial mass and larger age for the lower-@xmath11 wnh stars . in any case , further work on the h abundances of wnh stars is needed .
of course , factors other than initial mass may influence the apparent values of @xmath7 as well , such as the star s rotation rate and consequent level of mixing during core - h burning phases ( maeder & meynet 2000 ) .
this is beyond the scope of our consideration here , but it obviously may be important .
our primary goal in this paper is to determine how the wnh stars fit into the evolutionary sequence of massive stars .
when does the wnh phase `` turn on '' , how long does it last , how much mass is lost from the star , and what are the preceding and subsequent evolutionary phases ? in the previous sections , we showed that wnh stars are considerably more massive than h - poor wr stars , arguing that they are more massive because they have not yet shed their h envelopes and that they therefore represent an earlier evolutionary phase , before the heavy mass loss encountered as an lbv .
we also showed that the most luminous wnh stars tend to be more h - rich than lbvs , arguing again that they are pre - lbvs . another way to attack
the problem is to ask when in the lifetime of a massive star we should _ expect _ the wnh phase to occur , given some initial mass , luminosity , and mass - loss rate .
these expectations can then be compared with the measured masses , mass - loss rates , luminosities , and other properties of wnh stars . to this end , in the following discussion we consider what the _ expected _ properties of luminous o - type stars should be as they reach the end of core h burning .
we are primarily interested in _ the rate at which the mass loss rate grows from an initial value during core - h burning .
_ we consider the generic effect of mass loss on the stellar properties , and on the evolution of the mass - loss rate itself through a `` feedback '' effect .
the parameterization of mass loss discussed below is quite simple , and is not claimed to be an adequate substitute for renewed calculations of stellar evolution codes .
rather , it is meant only to illustrate the principle that the effect of mass loss can account for the properties of wnh stars if they are luminous and massive stars near the end of core h burning .
our results argue that renewed efforts to calculate the evolution of massive stars with lower mass - loss rates are essential in light of recent observational estimates of lower mass - loss rates due to wind clumping .
our arguments here strengthen the case that mass - loss rates of o - type stars need to revised downward from the `` standard '' observed rates ( de jager et al . 1988 ;
nieuwenhuijzen & de jager 1990 ) due to the observational effects of clumping , and more in line with ( but probably even lower than ) the theoretically expected values of vink et al .
( 2001 ) . in figure 3
we consider the expected properties of stars with initial masses of 120 , 85 , and 60 m@xmath1 as a function of time on the core - h burning main - sequence .
the prescribed inputs here are the initial mass , the initial value for the mass - loss rate , and the bolometric luminosity @xmath14 throughout the core - h burning lifetime .
we adopt @xmath14 from the solar metallicity models of schaller et al.(1992 ) , which include mass loss .
although it would be better to calculate new stellar evolution models self - consistently instead of adopting a luminosity from an existing model with different mass - loss rates , our approach here is a demonstrative first step . in any case , the adopted luminosities are sufficient to illustrate the main point of this paper , which is that the expected climb of the mass loss with time can account for the apparent properties of the wnh stars . with these prescribed conditions , figure 3 shows the time evolution of the mass - loss rate , the stellar mass , the eddington factor @xmath15 , where @xmath16 is the classical eddington limit due to electron scattering opacity , and the star s surface escape velocity . at each time step , each quantity is calculated iteratively from the previous time step .
for instance , following the initial mass , the mass at each subsequent time step is calculated by simply reducing the mass by @xmath17 . for reasons that will become obvious later
, we consider only the core - h burning main sequence lifetime and not core - he burning phases .
the purpose of figure 3 is to illustrate how these quantities change during the core - h burning lifetime of the star , in response to the choice of an initial value for the mass - loss rate .
we are especially interested in the way that the mass - loss rate grows due to previous mass loss
what we refer to below as mass loss `` feedback '' .
essentially all the observable properties of subsequent post - ms phases and the type of supernova eventually seen depend critically on the adopted @xmath18 . during core - h burning ,
stellar evolution calculations have typically assumed mass - loss rates adopted from observed `` standard '' values such as those given by de jager et al.(1988 ) or nieuwenhuijzen & de jager ( 1990 ) , as was done in schaller et al .
( 1992 ) and subsequent studies , heger et al .
( 2003 ) , eldridge & tout ( 2004 ) , and several others .
however , adopting these mass - loss rates is arguably not the best treatment of mass loss if one is interested in asking what mass - loss rates to expect as the star evolves , since this method simply prescribes them ( not to mention the fact that @xmath19 values need to be revised downward due to clumping ; see below ) .
stars at the same position in the hr diagram can have different masses , mass - loss rates , and other properties , arguing for a different approach .
this may be one of the reasons that the wnh phase is often assumed to be associated with later evolutionary phases ; for example , following the end of core - h burning , schaller et al .
( 1992 ) _ impose _ a wnh phase with a constant @xmath19=4@xmath2010@xmath21 @xmath4 yr@xmath22 until the h envelope is removed .
here we take a different approach . instead of prescribing mass - loss rates throughout the star s evolution ,
we consider the effect that the mass - loss rate for a line - driven stellar wind from a hot massive star should change during its lifetime , responding to changes in the star s luminosity with time , as well as to changes in its mass . during core - h burning
, a star s luminosity gradually climbs as the core contracts ( fig .
3 ) , providing one mechanism that will act to increase the mass - loss rate since the wind is radiatively driven .
simultaneously , the star s wind is removing mass from the star s surface , lowering the star s mass considerably , which also acts to increase the mass loss since the star has a shallower gravitational potential well .
in essence , both these effects conspire to raise the star s proximity to the classical eddington limit , since @xmath15 is proportional to l / m .
this increase in mass - loss rate accelerates the growth of @xmath23 , making the problem worse .
essentially , this behavior introduces a very important _ feedback loop _ that is currently not included in stellar evolution calculations .
this feedback effect can be treated in a simple way , sufficient for our limited purposes here .
the cak theory of radiatively driven winds ( castor et al . 1975 ) provides a prescription for how the mass - loss rate should vary with the star s
luminosity @xmath11 and the eddington factor @xmath23 ( and hence , the star s mass ) .
following owocki ( 2003 ) , for example , the dependence can be written as @xmath24 where @xmath3 is the usual cak power index .
for illustrative purposes in figure 2 , we adopt @xmath3=0.5 as is common in line - driven winds of o - type stars .
equation ( 1 ) then simplifies to @xmath25 which is the mass - loss rate dependence that we adopt in figure 3 .
when this feedback is included , we see that a star s mass - loss rate will climb steadily and substantially , even during the core - h burning phase alone .
given an initial mass - loss rate at @xmath26=0 , one can then calculate the mass - loss rate and the stellar mass at subsequent times , self consistently , given @xmath14 , as long as it is safe to assume that the wind is line - driven .
this assumption will break down as the eddington factor climbs near @xmath23=1 and the mass - loss rate skyrockets , when a continuum - driven wind may take over ( smith & owocki 2006 ; owocki et al .
2004 ) . given the simplistic treatment and the fact that we extrapolate from a single initial value , it is reassuring that our values of @xmath18 are not too different from the predicted values of vink et al .
( 2001 ; dotted line in the top panel of fig .
3 ) , especially at the later wnh stages that are the focus here . in figure 3 ,
the initial mass - loss rates we have adopted are the moderately - clumped rates given by repolust et al .
( 2004 ) appropriate for o - type dwarfs with the adopted luminosity for each of the three initial masses considered .
these mass - loss rates and the degree of clumping in stellar winds is a larger issue than we can address here ( see for example , puls et al . 2006 ; fullerton et al . 2006 ;
bouret et al . 2005 ; smith & owocki 2006 ; smith 2007a ; eversberg et al .
1998 ; lepine & moffat 1999 ) . as a result of the climbing mass - loss rate , we see that @xmath23 climbs significantly as well , ramping up at the end of core - h burning
. the implications of this for triggering the lbv instability are discussed below in 7 .
another result of the ramping - up of @xmath19 is that more of the mass ultimately shed from the star is lost later in its life , as is also the case if lbv eruptions are a dominant mode of mass loss ( smith & owocki 2006 ) .
this will significantly impact other properties such as the mass of the he core produced , as well as the angular momentum evolution of the star and its rotational mixing .
lower mass - loss rates will cause the wind to shed less angular momentum . in turn
, faster rotation will likely enhance axisymmetric / bipolar mass loss in later evolutionary phases ( cranmer & owocki 1995 ; owocki et al .
1996 , 1998 ; owocki & gayley 1997 ; maeder & meynet 2000 ) .
such effects wo nt be discussed in detail below , but they need to be reinvestigated in future stellar evolution calculations .
if figure 3 gives the expected behavior of @xmath18 for massive stars , we can then compare it with the observed mass - loss rates of wnh stars to deduce where they might fit in .
figure 4 is the same as figure 3 , but it includes some rough ranges of observed values corresponding to wnh stars for comparison .
we see here that for these very luminous stars , the mass - loss rate only climbs to values seen in wnh stars after about 2 myr from the beginning of the star s life . for the luminous wnh stars , typical mass - loss rates are well above 10@xmath21 @xmath4 yr@xmath22 ( crowther et al .
this favors the interpretation that the wnh stars have ages of 23 myr .
it is also consistent with the cohabitation of 3 wnh stars in the same region alongside @xmath2 car , whose late evolutionary stage points to an age of roughly 2.53 myr .
an independent check on the ages of wnh stars is given by their observed wind speeds ( bottom panel in fig .
we see that the observed wnh wind speeds of several hundred to 2000 km / s are not consistent with zams massive stars , nor with the very fast winds of h - free wn and wc stars . instead
, these values are only reached late in a star s core - h burning lifetime ( after roughly 2 myr ) as a natural consequence of mass loss and the corresponding lowered @xmath27 as the star gets pushed to higher values of @xmath23 .
thus , both the observed mass - loss rates and wind speeds of wnh stars would seem to favor ages above 2 myr .
caveat : one could of course assume that wnh stars are indeed very young that they are somehow born with such high mass - loss rates but what should we expect the star s subsequent evolution to look like in that case ?
this is akin to adopting higher mass - loss rates for o stars ( homogeneous winds instead of clumped winds ; see puls et al.2006 ) , and this introduces severe problems as discussed next .
if wnh stars do indeed reside near the end of core - h burning with ages @xmath132 myr , then their measured masses and luminosities may provide us with important clues to the mass - loss rates on the main sequence and the degree of clumping , which is a major issue in stellar evolution .
for example , if wnh stars have relatively high masses when they reach the ends of their core - h burning lives , then the mass - loss rates throughout core - h burning ca nt be very high .
in fact , some very high masses have been observed for wnh stars . as shown in figure 1 ,
several examples exist of wnh stars with masses of 80120 m@xmath1 measured in binary systems .
if figures 3 and 4 are accurate representations of the trend of @xmath19 on the main sequence , then we can see that the high masses of wnh stars make sense if they occur at the end of core - h burning _ but only if the winds are moderately clumped_. remember that in figure 3 we assumed that the initial mass - loss rates were those of repolust et al .
( 2004 ) , which correspond to conservative wind clumping factors of @xmath05 , reducing the mass - loss rates by factors of @xmath02 compared to the `` standard '' mass - loss rates and radio continuum diagnostics of the mass - loss rates . ]
derived from h@xmath3 and radio observations with the assumption of homogeneous winds ( de jager et al .
1988 ; nieuwenhuijzen & de jager 1990 ) . in this case , the early phases with lower mass - loss rates as an o - type star are nearly irrelevant to the star s total mass loss .
the mass - loss rate increases later as the star gradually moves into the wnh phase and on into the lbv phase , indicating that most of a star s mass is lost late in its lifetime at a quickened pace .
these moderately - clumped mass - loss rates are just about as high as they can be in order for the most massive stars to reach the eddington limit at the end of core - h burning .
let s turn this question around and view the problem from another perspective : if we still include the feedback effect of mass loss described earlier , as we arguably should , then what will happen
if the stellar winds are _ not _ clumped and we adopt initial mass - loss rates that are higher ?
figure 5 shows the same type of evolution with feedback ( solid lines ) as in figure 3 , with the only difference being that we started with higher initial mass - loss rates .
these initial rates ( dots ) adopt the prescription for the `` standard '' values for homogeneous winds given by nieuwenhuijzen & de jager ( 1990 ) .
these are the values that have most commonly been used in stellar evolution calculations ( e.g. , heger et al .
2003 ) , although sometimes rates even a factor of 2 higher than these are used to match observed statistics of massive stars ( meynet et al .
1994 ) .
the results of adopting the higher mass - loss rates in figure 5 are quite dramatic ( the same calculations without the feedback effect are shown with dashed lines in fig . 5 , for comparison ) .
figure 5 shows that with these high initial mass - loss rates , mass - loss feedback quickly drives the star s eddington factor up to @xmath23=1 in only 1.3 , 2 myr and 3 myr for the 120 , 85 , and 60 m@xmath1 models , respectively , likely triggering a very early lbv - like phase with catastrophic mass loss .
however , this is certainly in conflict with observations , at least for the 85 to 120 m@xmath1 models , since aside from the one case of @xmath2 carinae the stars in massive young clusters like r136 , ngc3603 , and carina typically do not have very massive shells of recently ejected material around them .
thus , we consider it unlikely that very luminous stars are born with mass - loss rates as high as those seen in wnh stars or with the mass - loss rates corresponding to homogeneous stellar winds .
also , the fact that massive stars reach the end of core - h burning as wnh stars with relatively high masses can not be explained by models in figure 5 with high mass - loss rates because too much mass has already been lost , whereas the high masses of wnh stars arise naturally for moderately - clumped winds .
this is the sensitive nature of the feedback effect that even a small reduction by a factor of @xmath02 in the initial mass - loss rate can have dramatic consequences later in a star s life .
furthermore , there s also the case of lbvs like @xmath2 car to consider : it is the most luminous and most evolved member of a rich region containing over 65 o - type stars , as well as the 3 well - known wnh stars ( see smith 2006 for a review , including details of the age differences among clusters in the region ) .
it is fair to assume that the current lbv phase of @xmath2 car is not only a post - ms phase , but probably also a post - wnh phase , since its ejecta are more nitrogen rich than the wnh stars in carina .
it is also safe to assume that @xmath2 car has advanced further in its evolution sooner than the wnh stars of the same age in this region simply because it is more luminous and started with a higher initial mass .
now , @xmath2 car is seen today surviving as a very massive star of around 100 m@xmath1 or more ( allowing for a hypothetical @xmath030 @xmath4 companion star ) , and we measure a total of something like 20 - 35 m@xmath1 in its circumstellar material ejected in only the last few thousand years ( the homunculus , plus more extended outer material ; see smith et al.2003 , 2005 ) .
that means @xmath2 car began its lbv phase and _ ended _ its ms and/or wnh phase with more than 120 m@xmath1 still bound to the star . if there really is an upper limit of about 150 m@xmath1 to the mass of stars ( figer 2005 ) , then this mass budget demands that the o - star and wnh winds were indeed quite meager before reaching this phase .
similarly , there s the pistol star to consider as well , which is also a post - ms object and has a present - day mass that probably exceeds 100 m@xmath1 ( figer et al .
1998 ) . in conclusion ,
then , the relatively high masses of wnh stars , the high masses of the most luminous lbvs , and the intuition that we should include the feedback effect of mass loss all argue that line - driven winds of massive stars _ must _ be clumped .
this argument is independent of the many spectroscopic clues that these winds are clumped ( e.g. , puls et al .
2006 ; bouret et al .
2005 ; fullerton et al . 2006 ;
eversberg et al . 1998 ; lepine & moffat 1999 ) .
we find that the mass - loss rate reductions due to clumping must be at least a factor of 2 compared to `` standard '' rates for homogeneous winds , consistent with the conservative factors adopted by repolust et al.(2004 ) .
it is encouraging that this amount of mass - loss reduction brings the mass - loss rates of o stars into better agreement with theoretical predictions for line - driven winds of o - type stars ( e.g. , vink et al .
2001 ) , although those may still be to high .
unfortunately , it is difficult to reliably determine the age of a wnh star directly from observations of its environment .
one can easily deduce cluster ages from observations that are either too low or too high by 1 myr or more : the essential problem is that the lifetimes of these very massive stars are so short that they are often comparable to the uncertainty in the age of their parent cluster .
in addition , that parent cluster or association may have a real age spead that makes the problem even worse .
thus , identifying a wnh star in a young 12 myr cluster like r136 or ngc3603 ( if their ages are that low ) does not necessarily mean that particular wnh star itself has an age of 12 myr . in any case
, the validity of ascribing a single - valued age to a given cluster is highly debatable .
since we ca nt really trust direct estimates of the ages for individual wnh stars based on their environments at least not at the precision of @xmath281 myr we must try to infer relative ages in other ways . in this paper
we have demonstrated that there are four main reasons to favor the interpretation that luminous wnh stars are pre - lbvs : \1 ) the stellar masses of wnh stars are systematically higher than h - poor wn stars , and more specifically , they have very different distributions ( fig . 1 ) .
namely , h - free wr stars are all heavily clustered around 1520 m@xmath1 , whereas wnh stars seem to be spread evenly from 20 m@xmath1 all the way up to the most massive stars known well above 100 m@xmath1 .
this is clearly in conflict with the notion that wnh stars are in the process of continuously becoming wn stars through _ their own
_ stellar winds .
instead , the strong discontinuity in mass distribution between the wnh and wn stars argues for an intervening phase of episodic mass loss that quickly sheds 10 s of solar masses and removes essentially all the remaining h envelope .
this intervening rapid mass - loss phase is almost certainly the lbv phase ( regardless of what the interpretation for the cause of the lbv phase might be ; i.e. inherent instability of single stars vs. binary mergers , etc . ) . unlike for h - poor wn stars , the stellar masses for lbvs overlap
quite well with the wnh stars .
\2 ) at high luminosities , the hydrogen mass fractions , @xmath7 , for wnh stars tend to be higher than for lbvs ( fig .
2@xmath9 ) .
this requires that they are less evolved .
we noted that this case is not definitive , since the higher wnh luminosities from hamann et al .
( 2006 ) paint a somewhat different picture than the lower luminosities of crowther et al .
( 1995a ) .
however , the properties of wnh stars in the cores of ngc 3603 and 30 doradus would seem to favor the interpretation that they are pre - lbvs based on their abundances ( de koter et al .
1997 ; drissen et al .
1995 ) .
\3 ) we showed that if one includes feedback due to mass loss on the main sequence , then starting with a moderately - clumped initial mass - loss rate , one naturally
_ expects _ the star s mass - loss rate to climb after about 2 myr to values that would make it appear as a wnh star , near the end of core - h burning and before the lbv phase ( figs . 3 & 4 ) .
as we showed earlier , if massive stars are born with the high mass - loss rates of wnh stars , then the subsequent evolution does not make sense ( fig .
\4 ) conversely , if the mass - loss rates are lower , then the effect of feedback is not so severe .
interestingly , the relative _ rate _ at which @xmath19(t ) climbs with the lower initial mass - loss rates of clumped winds ( fig .
3 ) more closely matches that of the _ observed _ mass - loss rates , even though the observed rates are offset to higher values .
therefore , simply lowering the observed `` standard '' values of nieuwenhuijzen & de jager ( 1990 ) by a factor of 23 gives a fairly accurate match to the expected mass - loss rates with feedback and clumping .
this provides a powerful , self - consistent argument that the mass - loss rates are in fact lowered due to clumping in stellar winds .
one of the interesting results of figure 3 is that this steady march toward increased mass - loss rates from feedback on the main sequence also provides a natural explanation for the apparent continuity in observed spectral traits from o3 v @xmath29 o3 if * @xmath29 wnh noted previously ( walborn 1971 , 1973 , 1974 ; walborn et al .
2002 ; walborn & blades 1997 ; conti 1976 ; melnick 1985 ; massey & hunter 1998 ; lamers & leitherer 1993 ; drissen et al . 1995 ; crowther et al.1995a ; etc . ) , and onward from wnh @xmath29 lbv as well .
this sequence is known to show intermediate stages , such as hot slash stars like melnick 42 in 30 dor and weak - lined wnh stars like wr25 ( e.g. , walborn et al .
1992 ) , attributed mainly to changes in wind density during stellar evolution .
we argue that no special circumstances like pulsationally - enhanced mass loss , rapid rotation , binary mergers , or unusual abundances are needed to account for the presence of wnh stars in massive young clusters with ages of around 23 myr it is a natural outcome of intially moderate mass loss on the main sequence that gradually grows more severe later on ( fig .
there are no h - free wr stars with luminosities above log ( l / l@xmath1)=5.8 ( fig . 2 ) , and if we favor the results shown in figure 2@xmath9 , there are not even any wnh stars with lower hydrogen mass fractions than lbvs for luminosities above log ( l / l@xmath1)=6.0 .
one possibility is that the lbv mass loss is so extreme that giant eruptions can completely remove the remaining h envelope to expose the bare he core .
thus , in figure 2 , a star would effectively move instantaneously from the position of an lbv like ag car to a lower - luminosity wr star .
however , the clear absence of h - free wr stars above log ( l / l@xmath1)=5.8 is puzzling in that case , as is the general dearth of h - free wn stars in giant h ii regions .
the sudden removal of the outer layers at the end of core - h burning should not much affect the luminosity of the he core that remains behind , so where are these luminous h - free stars ?
one way out of this predicament , hinted at by figure 2@xmath9 , could be if the more massive stars explode before shedding their h envelopes . in fact
, there is mounting evidence that some massive stars may explode during the lbv phase before ever making it all the way to the h - free wr stage ( see the discussions in smith & owocki 2006 ; smith 2007a , and gal - yam et al .
some examples are the recent type iin event sn 2006gy , which may have been the explosion of a very massive star like @xmath2 carinae ( smith et al .
2007 ) , the type iin event sn 2006gl , whose putative progenitor star identified by gal - yam et al . ( 2007 ) had photometric properties consistent with an lbv , the variable radio properties of some sne ( kotak & vink 2006 ; for other interpretations , however , see soderberg et al . [ 2006 ] and ryder et al .
[ 2006 ] ) , the sn1987a - like nebula around the lbv star hd168625 ( smith 2007b ) , plus many other type iin sne with dense environments .
there are also some he - rich stars ( perhaps lbv / wn transition stars ) that appear to have suffered lbv - like mass ejections shortly before a type ib / c sn explosion the clearest example being sn 2006jc ( see foley et al . 2007 ; pastorello et al . 2007 ;
smith et al .
2008 ) , which was actually _ observed _ to have an lbv - like event 2 yr prior to the sn explosion .
for the low - luminosity lbvs like hd168625 or r71 , explosion as an lbv is not necessarily a problem or even a surprise because those lower - l stars are likely to be in a post - rsg phase ( see smith et al .
for the high - luminosity lbvs above log ( l / l@xmath1)=5.8 , however , it presents a serious challenge to our current paradigms of stellar evolution .
obviously , we have stopped short of a full calculation of stellar evolution models that would take into account the way that the core luminosity may respond to mass loss .
however , in exploring the feedback effect , we adopted the core luminosities from the models of schaller et al .
( 1992 ) , which used relatively high mass - loss rates as noted earlier . therefore , with the lower initial mass - loss rates we argue for here ( and a consequent higher core luminosity ) , the feedback effect we propose could be even more extreme .
nevertheless , we only set out to demonstrate the principle of the feedback effect and that it can lead to high mass - loss rates of wnh stars if they are in the latter part of core - h burning for initially very massive o - type stars .
renewed efforts to calculate full stellar evolution models are encouraged .
our simple analysis does not account for chemical mixing and possible effects of rotation on that parameter , which is still a central problem in the evolution of massive stars .
it is quite possible that rotationally - enhanced mixing could lead stars with identical masses to evolve on somewhat different paths , hence surface abundances could differ .
thus , the exact time of `` onset '' of the wnh phase might vary from one star to another even for stars of the same initial mass based on the initial rotation rate and efficiency of the mixing .
therefore , our quoted age of @xmath02 myr for the onset of the wnh phase , which is imprecise to begin with , is not meant to be definitive .
it is well known that lbvs can undergo minor outbursts ( i.e. , `` s dor eruptions '' ) where they may change their mass - loss rate while remaining at constant bolometric luminosity , causing a corresponding change in apparent temperature and wind speed due to a pseudo photosphere .
more relevant here is that they can also suffer giant eruptions with violent sudden mass loss , like those seen in @xmath2 car and p cygni ( see , e.g. , davidson et al .
1989 ) , where they may eject @xmath010 @xmath4 in a single burst ( smith et al . 2003
; smith & owocki 2006 ) , and these event are likely to repeat .
there is a suspicion in the hot star community that the most luminous objects ( e.g. @xmath2 car or the pistol star ) might have fewer but more severe outbursts , while less luminous lbvs could repeat these episodes many times with each individual event less violent than for @xmath2 car ( again see , e.g. , contributions in davidson et al .
the mass lost each time , the frequency , and the total number of such outbursts as a function of luminosity or intial mass , combined with the potential regulating / perturbing effect of close binaries , are parameters that are badly needed for modeling the late evolution of massive stars , but we are still far from accurate empirical prescriptions of this mass loss behavior .
therefore , we must remain skeptical of the predictions of stellar evolution codes beyond the end of core - h burning for very massive stars .
how do the wnh stars fit into this scenario ?
could a wnh become an lbv and thence return to the wnh stage again ?
this is thought to be the case , specifically , for some ofpe / wn9 stars , since the lbvs ag car and r127 both look like ofpe / wn9 stars in their quiescent state ( stahl 1986 ; stahl et al .
perhaps something similar is happening for hd5980 in the smc .
can we determine if such lbv / wnh transition stars are occurring relatively early or late in a broader lbv phase ?
how does the initial stellar composition affect the wnh / lbv scenario we propose here ? does the presence of a number of early wn stars in the smc that are also wnh tell us something ?
we have shown that the masses , mass - loss rates , and abundances of luminous wnh stars are distinct from h - free wr stars , that they can be explained naturally if they are in the late stages of core - h burning , and that this can be understood as a direct consequence of mass - loss feedback during core - h burning .
we treat this feedback very simplistically as the dependence of the mass - loss rate on the luminosity and eddington factor associated with conventional line - driven wind theory , as explained in 5 .
we have ignored additional effects such as rotationally or pulsationally enhanced mass loss , but those effects may become necessary if o - star mass - loss rates are indeed reduced much below the values corresponding to moderate clumping factors ( repolust et al .
2004 ) that we have adopted here .
the feedback causes a steady climb in the star s mass - loss rate . just as this o - star mass loss provides a sort of feedback that leads to the higher mass - loss rates of wnh stars ,
so too does the higher mass - loss rate of the wnh phase enable further instability in later phases .
namely , with increased mass loss , the wnh wind lowers the stellar mass even further as the luminosity continues to climb at an ever faster pace .
this runaway eventually pushes the star to the classical eddington limit ( @xmath23=1 ) as shown in figure 3 . at some point along the way , perhaps at @xmath300.9 , an opacity modified eddington limit ( e.g. , lamers & fitzpatrick 1988 ; appenzeller 1986 ; and many contributions in davidson et al .
1989 ) presumably takes over and runaway mass - loss will ensue , marking the beginning of the lbv phase .
exactly when this is initiated relative to the brief transition from core h to he burning is unclear , and would seem to vary with the initial mass of the star and with metallicity . for the most massive stars ,
the sharp increase in luminosity right at this transition pushes the star past @xmath23=1 .
however , down near 60 m@xmath1 , for instance , it would appear that the star can linger on well into core he burning before actually triggering the lbv instability .
again , this is very interesting from the point of view of lbvs being potential type iin sn progenitors .
another consequence of lower mass - loss rates throughout core - h burning is that the star will suffer less angular momentum loss .
this , in turn , makes rotation have an even more important effect in the later stages as an lbv , as discussed by langer and others ( langer 1998 ; langer et al .
1994 ; glatzel 1998 ; maeder & meynet 2000 ) .
perhaps this rotation is critical for triggering the lbv instability for some range of lbvs , since models for initial masses below 85 m@xmath1 do not quite reach @xmath311 by the end of core - h burning ( fig . 3 ) . in any case , what happens to the structure of the star after the onset of the lbv phase has not yet been adequately addressed in stellar evolution calculations , because short duration eruptions and explosions of lbvs seem to be an integral part of exceeding the eddington limit . we have benefitted from numerous discussions about stellar winds and wnh stars with paul crowther , gloria koenigsberger , tony moffat , stan owocki , joachim puls , and jorick vink .
we are particularly grateful to p. crowther , t. moffat , nolan walborn , and an anonymous referee for detailed comments on the manuscript .
ns acknowledges support from nasa through stsci and jpl , and psc appreciates support from the nsf . | we propose the new designation `` wnh '' for luminous wolf - rayet ( wr ) stars of the nitrogen sequence with hydrogen in their spectra .
these have been commonly referred to as wnl stars ( wn7h , for example ) , but this new shorthand avoids confusion because there are late - type wn stars without hydrogen and early - type wn stars with hydrogen . clearly differentiating wnh stars from h - poor / h - free wn stars is critical when discussing them as potential progenitors of type ib / c supernovae and gamma ray bursts the massive wnh stars are _ not _ likely type ib / c supernova progenitors , and are distinct from core - he burning wr stars in several respects .
we show that the stellar masses of wnh stars are systematically higher than for _ bona fide _ h - poor wr stars ( both wn and wc ) , with little overlap .
also , the hydrogen mass fractions of the most luminous wnh stars are higher than those of luminous blue variables ( lbvs ) .
these two trends favor the interpretation that the most luminous wnh stars are still core - h burning , preceding the lbv phase ( at lower luminosities the wnh stars are less clearly distinguished from lbvs ) . while on the main sequence , a star s mass is reduced due to winds and its luminosity slowly rises , so the star increases its eddington factor , which in turn strongly increases the mass - loss rate , pushing it even closer to the eddington limit . accounting for this feedback from mass loss ,
we show that observed masses and mass - loss rates of wnh stars are a natural and expected outcome for very luminous stars approaching the end of core - h burning ( ages @xmath02 myr ) .
feedback from the strong wnh wind itself plays a similar role , enabling the eruptive instability seen subsequently as an lbv .
altogether , for initial masses above 4060 m@xmath1 , we find a strong and self - consistent case that luminous wnh stars are pre - lbvs rather than post - lbvs .
the steady march toward increased mass - loss rates from feedback also provides a natural explanation for the continuity in observed spectral traits from o3 v to o3 if * to wnh noted previously . |
dirac @xcite considered magnetic monopoles and derived a quantization rule for magnetic charge @xmath6 : @xmath7 , where @xmath8 is an integer ( below we put @xmath9 ) .
zwanziger and schwinger @xcite generalized this condition for dyons , which carry both electric ( @xmath10 ) and magnetic charges : @xmath11 .
this formula may be derived heuristically by the quantization of the angular momentum ( half - integer or integer ) of the electromagnetic field in a system consisting of two dyons @xcite .
e. purcell and n. ramsey in ref .
@xcite and n. ramsey @xcite discussed the possibility of there being elementary particle and nuclear edms due to the existence of magnetic monopoles .
later , e. witten @xcite showed that t hooft - polyakov magnetic monopoles carry small electric charges due to cp violating interactions . in the case of the @xmath12-term ( the interaction @xmath13 ) the electric charge of the monopole is @xmath14 .
v. sokolov developed nonrelativistic lagrangian and hamiltonian formalisms for the interaction of electric and magnetic charges which do not involve `` strings '' @xcite .
it is possible that monopoles may appear in particle - antiparticle pairs in which the point - like positive and negative magnetic charges could be very hard or impossible to separate . recall that a `` supercritical '' electric charge @xmath15 with @xmath16 must be screened down to the value @xmath17 due to spontaneous @xmath18 pair production .
the strength of the interaction in this case ( @xmath19 ) is still smaller than @xmath20 ( using the quantization rule for magnetic charge ) .
also , nambu showed that in the standard electroweak model the classical solution is a monopole - antimonopole pair connected by a @xmath21-field string @xcite .
this could explain the absence of free monopoles .
however even in this case one can still search for the effects of virtual monopole - antimonopole pairs and this will be discussed in the present paper . in ref .
@xcite it was pointed out that an electric dipole moment ( edm ) of an atom ( or any quantum system ) can be induced by the interaction of the electrons with a radial magnetic field ( @xmath22 ) .
the existence of such a field would contradict both time reflection invariance ( t ) and gauss s law ( @xmath23 ) .
the limit on such a field can be of interest by itself as a very precise test of electrodynamics at small distances since the accuracy of edm measurements is very high now .
it was also pointed out in @xcite that a radial magnetic field can be produced due to a monopole - antimonopole pair contribution to the magnetic moment ( if there is no magnetic string ) and a time invariance violating interaction which polarizes nucleon spins along the radial direction .
the mechanisms of radial magnetic field and edm creation can be much simpler if the magnetic monopole has an electric charge or is subject to the strong interaction .
the simplest example would be for a magnetic charge to be captured by an atomic nucleus and hence produce a `` coulomb '' magnetic field @xmath24 .
it is easy to find the exact solution of this problem ( an electron in the field of the dyon ) in the nonrelativistic case .
all stationary states in this problem possess an edm .
another example is a dyon - antidyon system with nonzero orbital angular momentum .
similarly to the way in which orbiting electric charges produce a magnetic moment , orbiting magnetic charges produce an edm @xcite
. the most interesting possibility would be for the radial magnetic field and edm to be induced by the virtual production of dyon - antidyon pairs
. the mechanism could be the following .
the electric field of an atomic nucleus polarizes the vacuum of dyon pairs and creates corrections to the radial electric field , @xmath25 .
the ratio of the magnetic to the electric field produced by these dyons is @xmath26 .
thus dyon vacuum polarization produces the radial magnetic field @xmath27 .
the interaction of atomic electrons with this magnetic field produces an atomic edm .
a similar mechanism could produce 35 orders of magnitude bigger p , t - odd effects in diatomic polar molecules .
we also consider the contribution of this mechanism to the neutron edm in the constituent quark model .
note that the electric field can be replaced by the strong field if the monopole interacts strongly .
moreover , it seems that `` chromodyons '' , which could exist in a generalization of qcd , could produce similar effects to that of `` electromagnetic '' dyons . thus the problem considered in the present work
could be related to the recent ideas about the role of monopole condensate in quark confinement ( see e.g. the recent review in @xcite and references therein ) . in this paper
we calculate the possible effects of time invariance violation in atoms , molecules and the neutron produced by a radial magnetic field .
we also estimate the field produced by the polarization of the dyon vacuum according to the mechanism discussed above .
an accurate calculation of the monopole effects requires the solution of numerous complicated problems such as the large value of the magnetic charge , `` strings '' , the finite size of the classical monopole solution , etc .
we stress that in the present work we are mainly trying to avoid these problems rather than to solve them .
in fact we explore an approach using simple heuristic arguments and perturbation theory which allows us to estimate and compare the values of t , p - odd effects in different quantum systems .
we must add that our attempt to use the results of two - vector potential theory for the dyon electrodynamics @xcite ( see also @xcite ) has lead to the conclusion that the magnetic field due to dyon vacuum polarization seems to vanish within this theory .
( this contradicts the simple and natural picture discussed above !
note however that the theory @xcite includes dyons only and the introduction of the usual charges into this theory leads to serious complications . )
an alternative approach is to use the theories with t hooft - polyakov magnetic monopoles .
we have not done any calculations within these theories .
our more simple `` heuristic '' calculation of the radial magnetic field due to dyon vacuum polarization ( see below ) is not based on a complete consistent theory and strictly speaking does not prove the existence of the effect .
however , the calculations of the effects of this field in sections [ sc][se ] apply to any contact radial magnetic field ; they are not restricted to fields produced by dyon vacuum polarization .
thus , the results in these sections can be used in general by using the equations in terms of @xmath28 ( @xmath28 is defined below ) .
let us start from an estimate of the radial magnetic field which could be produced due to polarization of the dyon vacuum by an electric charge .
the correction to the electrostatic potential @xmath29 due to the vacuum polarization of spin @xmath30 particles can be found in any textbook on quantum electrodynamics ( see e.g. @xcite ) : @xmath31 where @xmath10 and @xmath32 are the electric charge and mass of the dyons .
( in fact , @xmath33 in eq .
( [ e1 ] ) is the threshold of production of a dyon pair .
in the case of a bound pair ( connected e.g. by a @xmath21-string ) we could substitute the mass of the two - dyon system instead . ) the magnetic field due to the dyons can be expressed in terms of the corresponding electric field : @xmath34 , where @xmath6 is the magnetic charge .
if the dyons are heavy then the potential can be written as @xmath35 , where the constant can be found by the integration of eq .
( [ e1 ] ) over @xmath36
. thus polarization of the dyon vacuum produces the following radial magnetic field around the point - like charge @xmath37 : @xmath38 with @xmath39 note that t - invariance in this case can be restored if there is one more dyon with the same mass but the opposite sign of the product @xmath40 .
we can obtain the radial magnetic field of the nucleus by replacing @xmath41 by the proton density distribution , @xmath42 : @xmath43 where @xmath44 is a radial unit vector and @xmath45 .
here we took into account the fact that the nuclear density varies in a small interval around the nuclear radius @xmath46 .
consider now the interaction between an atomic electron and the contact radial magnetic field . a radial magnetic field can not be described by a nonsingular vector potential @xmath47 .
therefore , we will avoid using the vector potential .
let us start from the nonrelativistic problem .
the interaction between the spin magnetic moment and the magnetic field does not contain the vector potential : @xmath48 where @xmath49 are pauli matrices .
for the electron @xmath50 , with @xmath51 . note that @xmath28 in the above equation is only given by eq .
( [ e2ex2 ] ) for the dyon vacuum polarization mechanism , and can otherwise be considered as a more general parameter .
the orbital contribution to the interaction seems to vanish due to a cancelation between the contributions of the radial magnetic field of the nucleus and that of the electron .
let us first calculate the force acting on the electron from the radial magnetic field of the nucleus , in the latter s rest frame : @xmath52 where @xmath53 is the electron s velocity . to calculate the second force acting on the electron , we first calculate the force acting on the nucleus from the magnetic field of the electron @xmath54 .
since the electron is in motion , this magnetic field will be transformed into an electric field in the rest frame of the nucleus : @xmath55 ( to terms of order @xmath56 ) .
let us assume for now that the radial magnetic field around a particle is proportional to its electric charge .
this is obviously true for the magnetic field due to the dyon vacuum polarization mechanism in eqs .
( [ e2ex1 ] ) and ( [ e2ex2 ] ) , however this discussion is intended to be more general .
( an argument for this proportionality that is based on angular momentum quantization will be given below . ) under this assumption the force on the nucleus from the electric field ( [ electric ] ) is exactly equal to @xmath57 in eq .
( [ force ] ) . according to newton
s third law the corresponding force acting on the electron from the nucleus is @xmath58 , i.e. the net force acting on the electron is @xmath59 .
one can say that a cancelation occurs between the force @xmath57 acting on the electron from the vacuum monopole distribution near the nucleus and the force @xmath60 acting on the vacuum monopole distribution near the electron due to the electric field of the nucleus .
there is a second argument in favor of the cancelation of the orbital contribution .
as is known @xcite , the radial magnetic field of a magnetic charge and the radial electric field of an electric charge together produce a nonzero angular momentum around the axis connecting the charges ( @xmath61 ) . in quantum mechanics angular momentum
is quantized .
this requirement implies the quantization of the product of the magnetic and the electric charges . for two dyons
this condition is @xmath62 , where @xmath63 is the projection of the angular momentum @xmath64 onto the axis connecting these particles and @xmath65 is a unit vector along this axis .
recall that the system s orbital angular momentum is orthogonal to the connecting axis and does not contribute to @xmath66 .
we see that for @xmath67 the electromagnetic field angular momentum is zero due to the cancelation between the contributions of these two dyons .
a similar situation arises for the short - range induced magnetic field .
if both the radial magnetic fields of the electron and the point - like nucleus are proportional to their charges then the angular momentum of the field will be zero due to a cancelation between the two contributions .
( actually , the nucleus is not point - like , therefore , strictly speaking , we should consider an electron - quark system here . )
we stress that in the absence of such a cancelation there would be a problem : the size of the magnetic field region is very small ( @xmath68 ) , and so we can not satisfy the condition of angular momentum quantization for nonzero @xmath66 ( this provides an argument for the proportionality of the radial magnetic fields to the electric charges ) .
now there is a relation between the orbital contribution to the edm of the two - particle system and @xmath66 .
the angular wave function for the system is the wigner @xmath69-function , @xmath70 , where @xmath6 is the projection of @xmath71 onto the z - axis ( compare e.g. with ref .
@xcite , where the charge - monopole solution was found ) .
the orbital contribution to the edm in such an angular state is proportional to the angular integral @xmath72 $ ] .
this relation is especially simple for @xmath73 since the @xmath69-function in this case coincides with the usual @xmath74 angular function which gives an ( orbital ) edm of zero .
thus , we can consider the zero value of the electromagnetic field angular momentum in the electron - nucleus case to be an argument for the absence of an orbital contribution to the electric dipole moment .
we must stress that there is no such cancelation between the contributions of the magnetic moments of the nucleus and the electron ( see eq .
( [ nonrel ] ) ) since these magnetic moments are very different .
for example , the nuclear magnetic moment can be zero . in the relativistic case
the interaction of a magnetic moment @xmath75 with a magnetic field can be expressed in terms of the magnetic field only using the following well known identity ( see e.g. @xcite ) : @xmath76 where the first term in the r.h.s .
is an orbital contribution to the electromagnetic current and the second term is a spin one and @xmath77 .
taking into account the fact that @xmath78 we obtain ( from @xmath79 ) a relativistic expression for the interaction of a magnetic moment @xmath75 with the radial magnetic field ( [ e3 ] ) : @xmath80 expression ( [ e5 ] ) can also be obtained from the quadratic form of the dirac equation .
in this section we calculate the contribution of dyons to the atomic edm , as well as their effect on molecules .
the interaction ( [ e5 ] ) mixes atomic electron states of opposite parity , mostly @xmath81 and @xmath82 orbitals which are large at the nuclear surface .
@xmath83 \nonumber \\
g_s g_p & \approx & -n_0 z^2 \alpha^2 \nonumber \\ n_o & = & \frac{z r_r}{a^3 ( \nu_s \nu_p)^{\frac{3}{2 } } } \nonumber \\ \label{e5b } r_r & = & \left [ \frac{2}{\gamma(2\gamma+1 ) } \left ( \frac{a}{2zr } \right)^{1-\gamma } \right]^2\end{aligned}\ ] ] we use expressions for the electron wave functions at the nuclear surface from @xcite . in the notation used in this book @xmath84 is the electron energy , @xmath85 is the bohr radius , @xmath86 is the gamma function , @xmath87 and @xmath0 is the electron spinor . using @xmath88
we obtain the matrix element of the interaction of the electron magnetic moment with the radial magnetic field @xmath89 ( [ e5 ] ) : @xmath90 \ ; .\ ] ] note that the first term in the brackets is much larger in heavy atoms and the second term is necessary for the correct nonrelativistic limit as @xmath91 .
( it is interesting that the `` relativistic enhancement factor '' in this case can exceed several hundred . )
the electric dipole moment of an atom with one external electron ( e.g. tl , cs , fr ) generated by the interaction ( [ e6 ] ) can be calculated using perturbation theory in @xmath92 : @xmath93 where @xmath94 is the perturbed wave function , @xmath95 is the radial integral ( @xmath96 ) and @xmath97 is the matrix element of @xmath92 between the @xmath98 and @xmath99 orbitals ( @xmath100 and @xmath101 in cs , @xmath102 , @xmath103 in tl ) . there is a simpler way to obtain numerical results for the atomic edm : to use existing calculations of the electron edm enhancement factor in atoms and molecules @xcite or calculations of the atomic edm produced by the t , p - odd electron - nucleon interaction @xmath104 , where @xmath105 is a dimensionless interaction constant ( @xmath106 or @xmath107 ) @xcite .
comparison of the matrix element of the effective interaction between the electron edm @xmath108 and the atomic electric field @xmath109
( @xmath110 @xcite ) with expression ( [ e6 ] ) as well as a similar comparison with the matrix element of the t , p - odd interaction @xmath111 give the following substitutions in the expressions for the atomic and molecular edm ( and linear energy shifts in external electric fields ) : @xmath112 or @xmath113 @xmath114 to avoid confusion note that the limits are usually presented for @xmath115 . using equation ( [ e8 ] ) and the results of numerical atomic and molecular calculations of t , p - odd effects induced by the electron edm or the interaction @xmath111 ( see e.g. @xcite )
one can easily calculate the contribution of dyons to the atomic edm and the constants , @xmath0 of the t , p - odd interactions @xmath1 and @xmath2 for molecules ( here @xmath3 is the electron and @xmath4 the nuclear spin and @xmath5 is the molecular axis ) .
see tables [ t1 ] and [ t2 ] .
note that for the atoms and molecules with closed electron shells ( hg , xe , tlf ) the effect is proportional to the hyperfine interaction .
for these we present rough estimates based on the expression for @xmath106 in eq .
( [ e8 ] ) .
however a more accurate calculation is possible using the approach of ref .
@xcite .
one can also use eq .
( [ e8 ] ) to calculate the limit on @xmath116 from the known limits on the electron edm @xmath108 and @xmath106 .
at present the best limit follows from the measurement of the edm of the tl atom @xcite : @xmath117 \cdot 10^{-24 } e\cdot\mbox { cm}$ ] or @xmath118 \cdot 10^{-27 } e\cdot\mbox{cm}$ ] . using eq .
( [ e8 ] ) or table [ t1 ] we obtain the following limits for the dyon mass @xmath32 and the radial magnetic field produced by a particle with charge @xmath37 ( @xmath119 ) : @xmath120 ^ 2 } \cdot [ 0.35 \pm 0.23 \pm 0.20 ] \ ; , \ ] ] i.e. @xmath121 . according to dirac
@xcite @xmath122 and if the dyon charge @xmath123 the dyons would be very heavy .
the situation is different if the product @xmath40 is proportional to the strength of the t - violating interaction .
recall that according to @xcite @xmath124 and the present limit is @xmath125 @xcite . in this case
and so @xmath127 is not necessarily large .
note also that according to @xcite the mass of a monopole - antimonopole pair ( connected by a @xmath21-string ) in the standard model is in the tev range .
another way to search for t - violation is by neutron edm measurement . we can easily estimate the neutron edm using eqs . ( [ e2ex1])([e7 ] ) and the constituent quark model in a 3d - oscillator potential .
the magnetic moment of the neutron can be reproduced if we assume that the quarks have magnetic moments @xmath128 , masses @xmath129 and are in 1s states with the total spin of the d - quarks equal to 1 ( su(6 ) model ) .
the matrix element @xmath130 in eq .
( [ e7 ] ) can be easily estimated using the values of the oscillator wave functions at @xmath131 , @xmath132 and @xmath133 .
@xmath92 in this case is the interaction between the magnetic moment of one quark and the radial magnetic field produced by the two other quarks ( with total charge @xmath134 ) .
the sum over the excited states in eq .
( [ e7 ] ) is saturated by the nearest state @xmath135 in the oscillator model and the result can be expressed in terms of one parameter , @xmath136 , which is in fact the size of the nucleon ( @xmath137 , where @xmath138 is the @xmath139-meson mass ) .
after summation over the 3 quarks the edm of the neutron is : @xmath140 of course this is only an order of magnitude estimate which is however good enough to say that the neutron edm is @xmath141 times smaller than that of a heavy atom with nonzero electron angular momentum ( see table [ t1 ] ) .
the electron edm will be even smaller since it should be proportional to @xmath142 and @xmath143 .
a neutron edm can also be made up of `` intrinsic '' quark edms .
this quark edm ( as well as the electron s edm ) could appear due to radiative corrections with dyon loops .
1a corresponds to the vacuum transformation of a homogeneous electric field to a magnetic field ( @xmath144 term ) and should be omitted .
diagram 1b is forbidden by the furry theorem ( which however must be checked for dyons ) .
other diagrams such as that in fig .
1c are of a higher order in the electromagnetic interaction . they are not necessarily small ( since @xmath40 could be @xmath145 ) but they hardly can exceed the lower order contribution of fig .
1d calculated in eq .
( [ e10 ] ) .
we stress once more that a similar mechanism can possibly produce an edm of nucleons and nuclei if we consider `` chromomagnetic '' monopoles instead of the `` usual '' monopoles .
all that is required is to replace electric ( @xmath10 ) and magnetic ( @xmath6 ) charges in the expression for @xmath146 by `` chromoelectric '' ( color ) and `` chromomagnetic '' charges .
an atomic edm in this case can be generated by the interaction between electrons and nuclear t , p - odd moments .
one of us ( vvf ) is deeply thankful to m. yu . kuchiev for stimulating discussions .
he is also grateful to i.b .
khriplovich , b. marciano , v. soni , e. shuryak , o. sushkov and participants of the program `` physics beyond the standard model '' at the institute of nuclear theory ( university of washington ) for valuable discussions and references .
this work is supported by the australian research council and by the national science foundation through a grant for the institute for theoretical atomic and molecular physics at harvard university and the smithsonian astrophysical observatory .
dirac , proc .
( london ) * a133 * , 60 ( 1931 ) .
d. zwanziger , phys . rev .
* 176 * , 1480 ; 1489 ( 1968 ) . j. schwinger , phys . rev .
* 144 * , 1087 ( 1966 ) ; * 173 * , 1536 ( 1968 ) .
saha , indian j. phys .
* 10 * , 145 ( 1936 ) .
m. fierz , helv .
* 17 * , 27 ( 1944 ) .
wilson , phys . rev . * 75 * , 309 ( 1949 ) .
sokolov , yad .
* 23 * , 628 ( 1976 ) [ sov .
* 23 * , 330 ( 1976 ) ] .
sokolov , yad .
* 26 * , 427 ( 1977 ) [ sov . j. nucl
* 26 * , 224 ( 1977 ) ] .
purcell and n.f .
ramsey , phys .
rev . * 78 * , 807 ( 1950 ) .
n.f . ramsey , phys . rev . * 109 * , 225 ( 1958 ) .
e. witten , phys .
b * 86 * , 283 ( 1979 ) . y. nambu , nucl
b * 130 * , 505 ( 1977 ) .
flambaum , phys .
b * 320 * , 211 ( 1994 ) .
collins , phys . today * 48 * ( 3 ) , 17 ( 1995 ) . c. panagiotakopoulos , nucl .
b * 212 * , 118 ( 1983 ) .
d. zwanziger , phys .
d * 3 * , 880 ( 1971 ) .
brandt , f. neri and d. zwanziger , phys .
d * 19 * , 1153 ( 1979 ) .
berestetski , e.m . lifshitz and l.p .
pitaevski , _ relativistic quantum theory _
( pergamon press , oxford , 1971 ) .
khriplovich , _ parity nonconservation in atomic phenomena _
( gordon and breach , new york , 1991 ) .
flambaum , yad . fiz . * 24 * , 383 ( 1976 ) [ sov .
* 24 * , 199 ( 1976 ) ] .
kozlov and l.n .
labzowsky , j. phys .
b * 28 * , 1933 ( 1995 ) .
kozlov and v.f .
ezhov , phys . rev . a * 49 * , 4502 ( 1994 )
hinds , c.e .
loving and p.g.h .
sandars , phys .
b * 62 * , 97 ( 1976 ) .
v.v . flambaum and i.b .
khriplovich , zh .
89 * , 1505 ( 1985 ) [ sov .
jetp * 62 * , 872 ( 1985 ) ] .
flambaum , doctor of science thesis , novisibirsk , 1987 ( unpublished ) .
d. cho , k. sangster and e.a .
hinds , phys .
a * 44 * , 2783 ( 1991 ) .
commins , s.b .
ross , d. demille and b.c .
regan , phys .
a * 50 * , 2960 ( 1994 ) .
et al . _ , phys .
b * 234 * , 191 ( 1990 ) .
jacobs , w.m .
klipstein , s.k .
lamoreaux , b.r .
heckel and e.n .
fortson , phys .
lett . * 71 * , 3782 ( 1993 ) . | a radial magnetic field can induce a time invariance violating electric dipole moment ( edm ) in quantum systems .
the edms of the tl , cs , xe and hg atoms and the neutron that are produced by such a field are estimated .
the contributions of such a field to the constants , @xmath0 of the t , p - odd interactions @xmath1 and @xmath2 are also estimated for the tlf , hgf and ybf molecules ( where @xmath3 ( @xmath4 ) is the electron ( nuclear ) spin and @xmath5 is the molecular axis ) .
the best limit on the contact monopole field can be obtained from the measured value of the tl edm .
the possibility of such a field being produced from polarization of the vacuum of electrically charged magnetic monopoles ( dyons ) by a coulomb field is discussed , as well as the limit on these dyons .
an alternative mechanism involves chromomagnetic and chromoelectric fields in qcd . |
since the experimental realization of solitons in bose - einstein condensates @xcite , intense interest in their properties has emerged @xcite .
recently , exact solitonic solutions of the time - dependent gross - pitaevskii equation that describes the behavior of the condensate , have been obtained @xcite .
we have shown in @xcite , that exact solutions of the gross - pitaevskii equation can not be obtained for general external potential and interatomic interaction .
for the gross - pitaevskii equation to be solved exactly , the external potential and the interatomic interaction must be related . as an example on such a relation , the external potential in the gross - pitaevskii equation solved in ref .
@xcite is an expulsive harmonic potential and the interatomic interaction has an exponential prefactor that grows with time with a rate that is related to the strength of the harmonic potential .
our investigation in @xcite has provided a mathematical proof for the existence of such a correlation between the external potential and the interatomic interaction . from an experimental point of view
, it would be more interesting to find exact solutions for the external potentials and interatomic interactions used in the experiments .
for instance , it would be more interesting to have exact solutions for the realistic case of an upright harmonic external potential and constant interatomic interaction rather than for an inverted harmonic potential and exponentially growing interatomic interaction strength as in @xcite .
it turns out , however , that for the gross - pitaevskii equation to be solved exactly with an upright harmonic potential , the interatomic interaction will have to be oscillatory and divergent at regular times @xcite . in this paper
, we use the darboux transformation method @xcite to investigate the possibility of obtaining exact solutions of a time - dependent gross - pitaevskii equation with strengths of the harmonic external potential and interatomic interaction that are general functions of time .
it turns out that some special cases of such a general form of harmonic potential and interatomic interaction can be easily realized experimentally by controlling the strength of the magnetic field that provides the harmonic trap and using feshbach management of interatomic interaction @xcite .
one interesting special case corresponds to an oscillating strength of the harmonic potential and practically constant strength of the interatomic interaction .
the amplitude and frequency of the oscillating external potential can be used to delay the _ escape _ of the soliton from the harmonic potential or even to _ trap _ it at the center .
the rest of the paper is organized as follows . in the next section
, we present the general form of the gross - pitaevskii equation to be solved . in section [ darboux ]
we use the darboux transformation method to derive the exact solutions . in section [ results ] , we present and discuss the properties of the solutions in the special case of an oscillating external potential .
we end in section [ conc ] with a summary of our main conclusions .
when the confinement of the bose - einstein condensate is much larger in , say the @xmath0 and @xmath1 ( transverse ) directions , compared to the confinement in the @xmath2 direction , the system can be considered effectively one - dimensional along the @xmath2 direction .
the three - dimensional gross - pitaevskii equation can then be _ integrated _ over the @xmath0 and @xmath1 directions to result in the following one - dimensional gross - pitaevskii equation @xcite @xmath3 here @xmath4 is the @xmath5-wave scattering length , and @xmath6 , where @xmath7 and @xmath8 are the characteristic frequencies of the harmonic trapping potential in the @xmath2 and transverse directions , respectively . in the last equation ,
length is scaled to @xmath9 , time to @xmath10 , and @xmath11 to @xmath12 , where @xmath13 is the characteristic length of the harmonic trap in the transverse direction .
the dimensionless general functions @xmath14 and @xmath15 are introduced to account for the time - dependencies of the strengths of the trapping potential and the interatomic interaction . for eq .
[ gp1 ] to be solved exactly using the darboux transformation method , the functions @xmath14 and @xmath15 must be parametriclly related to each other throw a general function @xmath16 @xcite . in this case , the gross - pitaevskii equation , that we derive exact solutions for , takes the form @xmath17x^2\psi(x , t)-2a e^{2c_1+\lambda\int{g(t)dt}}|\psi(x , t)|^2\psi(x , t ) \label{gp2},\ ] ] where @xmath18 is an arbitrary constant .
in the darboux transformation method , we start by finding a linear system of equations for an _ auxiliary _ field @xmath19 such that eq .
( [ gp2 ] ) is its _ consistency condition _ @xcite .
we find that eq .
( [ gp2 ] ) corresponds to the consistency condition of the following linear system @xcite : @xmath20 @xmath21 where , + @xmath22 , @xmath23 , @xmath24 , @xmath25 , + + @xmath26 , + and @xmath27 .
+ the constants @xmath28 and @xmath29 are arbitrary constants .
the subscripts @xmath2 and @xmath30 denote partial derivatives with respect to @xmath2 and @xmath30 , respectively .
the function @xmath31 is related to the wave function through @xmath32 .
equation ( [ gp2 ] ) is obtained from the consistency condition @xmath33 . the linear system of 8 equations , eqs .
( [ psi_x ] ) and ( [ psi_t ] ) , reduces to an equivalent system of 4 equations with nontrivial solutions by making the following substitutions : @xmath34 , @xmath35 , and @xmath36 . to be able to solve this reduced linear system we need to know an exact ( _ seed _ ) solution of eq .
( [ gp2 ] ) .
following our previous approach of finding seed solutions @xcite , we find the following seed solution to eq .
( [ gp2 ] ) @xmath37}~ , \label{seed}\ ] ] where @xmath38 , @xmath39 and @xmath40 are arbitrary constants and @xmath41 .
the darboux transformation can now be applied to the linear system to generate a new solution of eq .
( [ gp2 ] ) as follows @xcite @xmath42 solving the linear system ( [ psi_x ] ) and ( [ psi_t ] ) using the seed solution eq .
( [ seed ] ) and then substituting for @xmath43 , @xmath44 , and @xmath45 in the last equation , we obtain the following new exact solution to eq .
( [ gp2 ] ) : @xmath46\left.\frac{}{}\right\ } \label{gpsol},\end{aligned}\ ] ] where + @xmath47/2\lambda$ ] , + @xmath48 , + @xmath49 , + @xmath50 , + @xmath51 , + @xmath52 , + @xmath53 , + @xmath54 , + @xmath55 , + @xmath56 , + @xmath57 $ ] , + @xmath58 $ ] , + @xmath59 , + @xmath60 , + the subscripts @xmath61 and @xmath62 denote real and imaginary parts , respectively , and @xmath63 and @xmath64 are arbitrary constants .
further reduction and simplification of this new general solution as performed in @xcite will not be considered here .
furthermore , we do not attempt to obtain all classes of solitonic solutions . instead , the focus in this paper will be on the effect of the parameters of the oscillating trapping potential on the dynamics of single and multiple solitons which are described by eq .
( [ gpsol ] ) .
in this section we derive from the general solution found in the previous section solitonic solutions for an oscillating harmonic trapping potential . to this end , we take @xmath65 where @xmath66 , @xmath67 , @xmath68 , and @xmath69 are constants . using the last equation and @xmath60 to substitute for @xmath70 in eq .
( [ gp2 ] ) , the gross - pitaevskii equation takes the form @xmath71x^2 - 2ae^{2c_1+(\alpha_1+\alpha_2\sin{\omega t+\delta})/2}|\psi(x , t)|^2\right\}\psi(x , t ) \label{gp3}.\ ] ] it should be noted that the specific choice of @xmath72 leads to the gross - pitaevskii equation of ref .
@xcite with an expulsive harmonic potential . simulating the time dependencies of the harmonic potential and the interatomic interaction of eq .
( [ gp3 ] ) may be difficult experimentally , but if we choose the parameters such that @xmath73 , the interatomic interaction can be considered practically as constant , and the strength of the harmonic potential will be oscillating as @xmath74
. it should be noted , however , that even in this limit , the contribution of the positive @xmath75 term will result in that the trapping potential will spend more time being an inverted parabola than being an upright parabola . over a large enough time interval , this will lead to expelling the solitons away from the center of the trap .
this is indeed the picture that we get when we plot the density of a single - soliton solution as shown in fig .
it is clear from this figure that the soliton is being expelled out from the center of the condensate at @xmath76 .
the oscillation in the trajectory of the soliton is due to the oscillating trapping potential .
the discontinuous appearing of the soliton s peak density is due to the interaction with the background .
this trajectory can be also extracted from the general solution eq .
( [ gpsol ] ) by considering the term @xmath77 in the denominator . at the peaks of the oscillation in fig .
( [ fig1 ] ) , this is the dominant term that determines the position of the peak of the soliton .
specifically , the peaks are given by the condition @xmath78 .
using this condition to plot @xmath2 vs. @xmath30 in fig .
[ fig2 ] , we obtain a curve that is identical to the soliton trajectory in fig .
[ fig1 ] . the mean slope of this curve is proportional to @xmath79 .
hence , the rate at which the soliton leaves the center of the trapping potential can be delayed by choosing the parameters and the arbitrary constants such that @xmath79 is small .
for the special case of @xmath80 the center of mass of the soliton will be located at @xmath76 at all times .
the oscillating trapping potential will result only in oscillations in the width and peak density of the solitons .
the solitons in this case remain trapped at the center of the trapping potential .
we show this case in figs .
[ fig3]-[fig6 ] where we see a multi - soliton solution with central soliton being _ pinned _ at @xmath76 and off - central ones oscillating around their initial positions . in fig .
[ fig4 ] , we use the same parameters as in fig . [ fig3 ] but with a doubled frequency @xmath68 .
this shows clearly that the oscillations of the off - centered solitons are due to the oscillation in the trapping potential . in fig .
[ fig5 ] , we show that for some values of the parameters the dynamics of the peak soliton density can be drastic such that the soliton disappears in the background and reappears at regular discrete times .
figure [ fig6 ] is similar to fig .
[ fig3 ] but with a larger number of solitons .
we have found exact single and multi solitonic solutions of a time - dependent gross - pitaevskii equation with time - dependent amplitudes of harmonic trapping potential and interatomic interaction .
we considered the interesting special case of sinusoidally oscillating strength of the harmonic trapping potential .
we focused on the effect of the frequency and amplitude of the oscillating trapping potential on the dynamics of the solitons .
the parameters can be chosen such that the soliton is trapped at the center of the condensate . for a typical @xmath81rb condensate with @xmath82 atoms and trapping frequency of order 100 hz
, the thomas - fermi size of the condensate is @xmath83 @xcite .
this means that the width of the solitons of figs .
[ fig3]-[fig4 ] are of the order of the size of a nonsolitonic condensate while the width of the solitons in figs [ fig5]-[fig6 ] is smaller than the size of the nonsolitonic condensate .
j. denschlag _
et al . _ ,
science * 287 * , 97 ( 2000 ) .
anderson , p.c .
haljan , c.a .
regal , d.l .
feder , l.a .
collins , c.w .
clark , and e.a .
cornell , phys .
* 86 * , 2926 ( 2001 ) .
l. khaykovich _
et al . _ ,
science * 296 * , 1290 ( 2002 ) .
strecker , g.b .
partridge , a.g .
truscott , and r.g .
hulet , nature * 417 * , 150 ( 2002 ) .
s. inouye , m.r .
andrews , j. stenger , h .- j .
miesner , d.m .
stamper - kurn , and w. ketterle , nature ( london ) * 392 * , 151 ( 1998 ) ; s.l .
cornish , n.r .
claussen , j.l .
roberts , e.a .
cornell , and c.e .
weiman , phys .
85 , 1795 ( 2000 ) . .
the values of the parameters used in this plot are : @xmath84 , @xmath85 , @xmath96 , @xmath97 , @xmath98 , @xmath99 , @xmath100 , @xmath93 , @xmath101 , @xmath95 .
the value of @xmath40 is the solution of @xmath80 with respect to @xmath40.,width=377 ] | we derive exact solitonic solutions of the one - dimensional time - dependent gross - pitaevskii equation with time - dependent strengths of the harmonic external potential and the interatomic interaction .
the time - dependence of the external potential and interatomic interaction are given in terms of a general function of time .
for an oscillating strength of the external potential , the solutions correspond to breathing single and multiple solitons .
the amplitude and frequency of the oscillating potential can be used to control the dynamics of the center of mass of the solitons . for certain values of these parameters ,
the solitons can be _ trapped _ at the center of the harmonic potential . |
the discovery of a higgs boson with mass of about 125 gev in 2012 @xcite at the large hadron collider ( lhc ) has initiated a new era of excitement in particle physics aimed at the understanding of the full nature of electroweak symmetry breaking . in particular
, the question of whether this new boson is solely responsible for electroweak symmetry breaking and the origin of mass is a central thrust of the physics program at the lhc .
the standard model ( sm ) of particle physics has been successful at explaining a wide range of experimental phenomena . however , there many open questions it fails to answer .
similar to the introduction of the higgs mechanism to explain the su(2)@xmath10u(1 ) symmetry breaking of the sm , several extensions to the sm model have been proposed to address its incompleteness .
for example , models of extra dimensions , such as the randall - sundrum ( rs ) model @xcite , were introduced to solve the hierarchy problem by postulating that gravity is weaker compared to the other fundamental forces because it can propagate in extra spatial dimensions .
while the extent of these new physics models is vast and their physics motivation often different , a common theme is the manifestation of new tev scale neutral gauge bosons ( @xmath11 ) that can be probed at the lhc . at the cern lhc ,
the atlas @xcite and cms @xcite experiments have an extensive physics program to search for @xmath11 .
a widely used model in these searches is the sequential standard model ( seqsm ) , which predicts a spin-1 neutral boson ( @xmath12 ) with sm - like couplings @xcite .
results of direct searches for high mass dilepton resonances in proton - proton collisions at @xmath13 , 8 , and 13 tev exclude @xmath12 masses below 3.75 and 2.1 tev in the @xmath7 and @xmath6 channels , respectively @xcite .
the strategy pursued in those analyses is a simple and robust dilepton selection , targeting production via drell - yan ( dy ) processes of order @xmath14 ( i.e. @xmath15 ) , with high signal acceptance that produces a `` bump '' ( narrow in @xmath7 and broad in @xmath6 ) in the reconstructed invariant mass spectrum of lepton pairs that sits above a smooth and steeply falling background distribution .
the focus of this paper is to propose new searches for @xmath11 bosons at the lhc using events produced through vector boson fusion ( vbf ) processes ( figure [ fig : feyn ] ) .
the tagging of events produced though vbf processes has been proposed by some of the present authors as an effective experimental tool for dark matter ( dm ) and electroweak supersymmetry ( susy ) searches at the lhc @xcite .
in particular , it has been shown in @xcite that vbf processes become increasingly more important for probing large mass scales due to collinear logarithmic enhancements in the production cross - sections .
this characteristic is particularly interesting in light of the recent observation of a @xmath16 excess ( local significance ) near @xmath17 gev in the high - mass diphoton searches at atlas and cms .
although the significance of this bump has decreased with the 13 tev data acquired this year @xcite , it has motivated exploring the effectiveness of the vbf topology for tev scale @xmath11 searches .
additionally , the vbf topology provides effective reduction of the qcd multijet background in susy and dm searches @xcite , which is an important characteristic which can also be utilized in @xmath11 searches which suffer from large qcd multijet backgrounds , such as @xmath18 .
although there are many ways that a new heavy neu- particle and two forward jets.,scaledwidth=50.0% ] tral gauge boson may arise from extensions to the sm , in order to allow for comparisons with current results from atlas / cms and to highlight the usefulness of the vbf topology in @xmath11 searches with large qcd backgrounds as well as those with clean signatures , we use the seqsm as a benchmark and consider @xmath11 decays to @xmath6 and @xmath7 . in the vbf @xmath19 study , we focus on the final state where both @xmath20 leptons decay to hadrons ( @xmath21 ) since it provides the largest branching fraction ( 42% ) and is expected to produce better sensitivity compared to final states with semi - leptonic decays of @xmath20 leptons .
the dominant sources of background in these studies are production of top quark pairs ( @xmath22 ) and @xmath23/@xmath24 bosons with associated jets ( referred to as @xmath23+jets , @xmath24+jets , and more generally as @xmath25+jets ) .
the @xmath23+jets background is characterized by two real prompt leptons from the @xmath23 boson decay in addition to two jets , either from initial state radiation ( processes of order @xmath26 ) or from pure electroweak processes .
the w+jets background is only important in the @xmath6 study and satisfies the selection criteria when a jet is misidentified as a @xmath21 and two additional jets ( e.g. initial state radiation ) satisfy the vbf criteria .
background from @xmath22 events is usually accompanied by one or two b quark jets , in addition to real prompt leptons .
signal and background samples were generated with madgraph ( v2.2.3 ) @xcite , considering proton - proton beams with @xmath27 tev .
pythia ( v6.416 ) @xcite was used for the hadronization process , while delphes production cross - section as a function of mass .
the vbf @xmath11 production cross - section excees the production via the dy process at approximately @xmath28 tev .
, scaledwidth=45.0% ] ( v3.3.2 ) @xcite was used to simulate detector effects ( cms configuration was used ) .
the @xmath23/@xmath24(@xmath29/@xmath30)+jets background events were generated with up to four associated jets , inclusive in @xmath31 and @xmath32 .
the mlm algorithm @xcite was used for jet matching and jet merging , which optimizes two variables ( xqcut and qcut ) related to the jet definition .
the xqcut variable defines the minimal distance between partons at madgraph level .
the qcut variable defines the minimum energy spread for a clustered jet in pythia . in order to determine appropriate xqcut and qcut values ,
the distribution of the differential jet rate was required to smoothly transition between events with @xmath33 and @xmath34 jets .
the jet matching and merging studies resulted in an optimized xqcut of 20 and qcut of 40 . at the madgraph level
, leptons were required to have a @xmath35 gev and @xmath36 , while jets were required to have a minimum @xmath37 gev and @xmath38 .
the signal samples were produced considering only pure electroweak production of @xmath11 and two associated jets : @xmath39 with @xmath40 . as mentioned above
, we considered the seqsm as our benchmark scenario for historical reasons and allowed decays to either @xmath6 or @xmath7 .
we scanned @xmath41 ranging from 1 tev to 5 tev , in steps of 250 gev .
figure [ fig : xsections ] shows the @xmath12 production cross - section as a function of mass for both `` standard '' dy - like production , which are consistent with the values used by atlas and cms in @xcite , and pure electroweak vbf production .
the vbf @xmath11 production cross - section surpasses that of the dy process at approximately @xmath42 tev , making vbf processes l c ccriterion & @xmath43 & @xmath44 + + @xmath45 & @xmath46 & @xmath46 + @xmath47 & @xmath48 gev & @xmath49 gev + @xmath50 & @xmath51 & @xmath51 + @xmath52 & @xmath53 & @xmath53 + @xmath54 & @xmath55 & @xmath55 + n(b - jets ) & 0 & 0 + @xmath56 & @xmath57 gev &
+ + @xmath58 & 50 gev & 30 gev + @xmath59 & @xmath60 & @xmath60 + @xmath61 & 50 gev & 30 gev + @xmath62 & @xmath60 & @xmath60 + @xmath63 & @xmath64 & @xmath64 + @xmath65 & @xmath53 & @xmath53 + @xmath66 & @xmath67 & @xmath67 + @xmath68 & @xmath69 tev & @xmath70 tev + important for searches involving tev scale @xmath11 masses .
for completeness , it is noted the cross - sections for vbf higgs production in @xcite have also been reproduced as a cross - check of the madgraph calculations for vbf @xmath12 .
the vbf topology is characterized by two high @xmath3 forward jets ( @xmath71 ) , with large pseudorapidity gap , located in opposite hemispheres of the detector , and tev scale dijet invariant masses .
similar to the vbf susy and vbf dm searches in @xcite , an advantage of this topology is the reduction of the qcd multijet background particularly important in searches with @xmath20 leptons suffering from relatively high jet@xmath72 misidentification rates .
however , a vbf @xmath11 search differs fundamentally from those analyses in the following ways : @xmath73 since susy particles must be produced in pairs under the assumption of r - parity conservation , the vbf processes considered in @xcite occur through t - channel diagrams , while vbf @xmath11 proceeds via s - channel ( figure [ fig : feyn ] ) , resulting in lower average @xmath3 for the jets ; @xmath74 vbf @xmath75 contains leptons with higher @xmath3 ( tev scale ) .
these features mean different search philosophies , including e.g. the @xmath21 identification strategy and the choice of the experimental trigger .
for example , while the vbf susy searches profit from the use of a vbf dijet + @xmath76 trigger to select low @xmath3 leptons without bias , the high @xmath3 leptons from @xmath11 decays allow the use of single lepton or dilepton triggers to maintain unbiased vbf dijet and @xmath76 cuts targeting the lower jet @xmath3 and @xmath76 values typical of s - channel vbf @xmath11 production . while discovery of either type of new physics would be an incredible achievement , the difference between t - channel and s - channel provides distribution of the highest @xmath3 jet for the sum of sm backgrounds and the vbf @xmath18 signal benchmark point with @xmath77 tev.,scaledwidth=50.0% ] a clean approach to reveal the difference between the two models .
the event selection criteria used in both the @xmath43 and @xmath44 channels is summarized in table [ tab : selections ] . in the @xmath43 channel ,
we select events with two oppositely charged @xmath21 candidates ( @xmath78 ) , each with @xmath79 gev and @xmath80 2.5 , that must be well separated in @xmath81 space by requiring @xmath82 .
the requirement on @xmath83 bounds the selected @xmath21 candidates to be within the tracker acceptance region of the detector .
contamination from @xmath22 events is largely suppressed by vetoing events containing jets , with @xmath84 gev , identified as b quarks . to further reduce possible contamination from qcd multijet and @xmath85 + jets backgrounds
, events must have @xmath86 gev . in table
[ tab : selections ] we refer to the aforementioned criteria as central selections . in order to impose a stringent vbf selection aimed at the reduction of qcd multijet and background processes with two jets from initial state radiation ( e.g. @xmath87 order @xmath88 ) , a minimum @xmath3 threshold of 50 gev
is used for the leading and sub - leading jets .
the selected jets and @xmath21 candidates must be well separated by requiring @xmath89 .
the @xmath90 distribution of the highest @xmath3 jet for the signal and sum of sm backgrounds is shown in figure [ fig : jeteta ] .
figure [ fig : deta ] displays the difference in pseudorapidity , @xmath91 , between the two leading jets . the requirement that the two jets be in opposite hemispheres of the detector is imposed with @xmath92 and @xmath93 .
finally , the invariant mass of the dijet pair , @xmath68 , must be greater than 1 tev .
similar event selection criteria is used for the @xmath44 between the two leadings jets for the sum of sm backgrounds and the vbf @xmath18 signal benchmark point with @xmath77 tev.,scaledwidth=50.0% ] channel .
nevertheless , this channel has significantly lower background contamination in comparison to the @xmath43 channel .
this is due to a much smaller jet@xmath94 misidentification rate ( at least one order of magnitude smaller ) .
the low jet@xmath94 misidentification rate allows for lower @xmath3 thresholds on @xmath95 and jet candidates .
muons and jets are required to have @xmath96 gev .
the less stringent vbf selection in the @xmath44 channel allows us to regain signal acceptance .
although we have assumed the atlas and cms dimuon triggers allow for 30 gev cuts on @xmath97 ( mostly to allow for observation of a @xmath23 mass peak ) , it should be noted that the average transverse momentum of muons from @xmath11 decays is @xmath98/2 , independent of the @xmath11 boost from the vbf selection , and thus the muon @xmath3 thresholds can be increased without affecting the projected signal significance .
other sets of topological variables were considered , such as the cos@xmath99 and @xmath100 variables used in the atlas and cms searches to reduce the w+jets background and targeting the back - to - back nature of the lepton pair @xcite . however , since the @xmath11 produced in vbf processes is boosted in order to balance the momentum of the vbf jets , those variables are sub - optimal in these studies . on the other hand ,
the w+jets background is greatly reduced owing to the vbf criteria .
the reconstructed dilepton mass is proposed as the main variable to discriminate against known sm backgrounds .
an enhancement in the tails of the mass distribution would indicate the presence of new physics at the lhc . while the reconstructed dimuon mass @xmath101 in vbf @xmath102 shows up as a narrow bump
, the visible ditau mass ( i.e. @xmath103 ) in vbf @xmath18 does not pr- distributions in dy and vbf production of @xmath18 with @xmath77 tev.,scaledwidth=50.0% ] oduce a narrow `` peak '' since the @xmath6 system decays into neutrinos .
for this reason , the reconstructed invariant mass distribution in the @xmath43 channel , @xmath104 , utilizes the reconstructe @xmath76 and is defined based on equation [ eq : mt ] .
@xmath105 figure [ fig : dyandvbfditaumass ] shows the @xmath104 distributions in dy and vbf production of @xmath18 with @xmath77 tev .
since the neutrinos from the @xmath106 decay are back - to - back in dy processes , there is a cancellation of missing momentum which does not allow the reconstruction of the true mass , on average ( i.e. @xmath107 tev in @xmath108 ) . on the other hand , and as noted above , since the @xmath11 produced in vbf processes is boosted in order to balance the momentum of the vbf jets , the neutrinos are not back - to - back and thus results in improved reconstructed mass scale and resolution . figure [ fig : massditau ] shows the expected @xmath104 background and signal distributions using events satisfying the @xmath43 cuts in table [ tab : selections ] .
the signal is overlaid on top of the stacked backgrounds .
figure [ fig : massdimuon ] shows a similar distribution , except for the @xmath7 channel .
the last bin in these distributions represent the overflow bin .
the bulk of the background distribution resides at low mass , while the signal dominates in the tails of the distribution .
distribution for the main backgrounds and three chosen signal benchmark points , after applying the final event selection criteria.,scaledwidth=50.0% ]
similar to the atlas and cms searches , we utilize a shape based analysis of the reconstruced mass distribution , using the rootfit @xcite toolkit , to construct a binned likelihood following the test statistic based on the profile likelihood ratio .
reasonable systematic uncertainty is considered in order to calculate projected significance . however , since the background rates are small at high mass values where the signal dominates , the poisson uncertainty on the yields dominate the total uncertainty . the dominant sources of systematics are expected to be the uncertainty on @xmath21 identification ( 6% based on @xcite ) , vbf selection efficiency ( 20% based on @xcite ) , and the uncertainty due to the variations in the yields and mass shapes arising from the choice of parton distribution function ( 15% based on @xcite ) .
based on these considerations , a 25% total systematic uncertainty on the signal and background yields is a reasonable choice .
systematic uncertainties are incorporated via nuisance parameters following the frequentist approach .
a local p - value is calculated as the probability under a background only hypothesis to obtain a value of the test statistic as large as that obtained with a signal plus background hypothesis . the significance @xmath109 is then determined as the value at which the integral of a gaussian between @xmath109 and @xmath110 results in a value equal to the local p - value . figure [ fig : results ] shows the expected signal significance , considering an integrated luminosity of 30 fb@xmath8 .
the proposed @xmath11 searches using the vbf topology can provide expected significances greater than 5@xmath9 ( 3@xmath9 ) for @xmath12 masses up distribution for the main backgrounds and three chosen signal benchmark points , after applying the final event selection criteria.,scaledwidth=50.0% ] to 2.2 ( 2.9 ) tev and 3.4 ( 4.0 ) tev in the @xmath6 and @xmath7 channels , respectively .
using 1.69@xmath9 to define an exclusion bound at 95% confidence level , the projected exclusion bounds are @xmath111 tev and @xmath112 tev in the @xmath43 and @xmath44 channels .
these projected exclusion bounds can be compared to the most recent atlas and cms results , which currently exclude @xmath12 masses below 2.1 tev and 3.75 tev in @xmath18 and @xmath102 channels , respectively .
the proposed searches for tev scale neutral gauge bosons produced via vbf processes can increase the current atlas and cms bounds on @xmath41 by about 70% in the @xmath6 decay mode and 25% in @xmath7 decay mode . we stress that while it s clear the @xmath7 channel has better discovery potential for models with universal couplings of the @xmath11 boson to the leptons , the improved reach in the @xmath6 decay mode is particularly important in models with enhanced couplings to third - generation fermions ( e.g. topcolor - assisted technicolor @xcite ) and can be derived from figure [ fig : results ] by appropriately rescaling the vbf @xmath11 cross - section .
the main result of this paper is that probing heavy neutral gauge bosons @xmath11 produced through vbf processes can be a key methodology to complement current searches for tev scale particles at the lhc . in the context of the seqsm , it is shown that the vbf production cross - section surpasses that of the more standard dy process at approximately @xmath28 tev , which makes and vbf @xmath102 searches , using a shape based statistical analysis of the reconstructed mass distribution , as a function of @xmath41 .
the assumed integrated luminosity is 30 fb@xmath8 .
a total systematic uncertainty of 25% has been considered on the signal and background yields.,scaledwidth=50.0% ] @xmath11 searches with the vbf topology particularly important for the long term search strategies at the lhc . to highlight the power of the vbf topology in @xmath11 searches with large qcd backgrounds as well as those with clean signatures , we consider @xmath11 decays to @xmath6 and @xmath7 and
show that the requirement of a dilepton pair combined with two high @xmath3 forward jets with large separation in pseudorapidity and with large dijet mass is effective in reducing sm backgrounds .
the expected exclusion bounds ( at 95% confidence level ) are @xmath111 tev and @xmath112 tev in the @xmath43 and @xmath44 channels , respectively , assuming 30 fb@xmath8 of 13 tev data from the lhc .
the use of the vbf topology to search for massive neutral gauge bosons increases the discovery reach at the lhc , with expected significances greater than 5@xmath9 ( 3@xmath9 ) for @xmath12 masses up to 2.2 ( 2.9 ) tev and 3.4 ( 4.0 ) tev in the @xmath43 and @xmath44 channels . we have developed this paper for the seqsm benchmark .
thus , the sensitivities we expect are derived under the assumption of seqsm values for the couplings @xmath113 and @xmath114 .
our calculations can be generalized to other models by varying the couplings .
the event rate for @xmath11 detection at the lhc scales as @xmath115@xmath116 , and so larger couplings lead to greater sensitivity , while smaller couplings lead to less sensitivity .
we thank the constant and enduring financial support received for this project from the faculty of science at universidad de los andes ( bogot , colombia ) , the administrative department of science , technology and innovation of colombia ( colciencias ) , the physics & astronomy department at vanderbilt university and the us national science foundation .
this work is supported in part by nsf award phy-1506406 .
tw thanks doe for partial support via grant de - sc0011981 .
g. aad _ et al . _ [ atlas collaboration ] , `` observation of a new particle in the search for the standard model higgs boson with the atlas detector at the lhc , '' phys .
b * 716 * , 1 , 1 - 29 ( 2012 ) 10.1016/j.physletb.2012.08.020 [ arxiv:1207.7214 [ hep - ex ] ] .
v. khachatryan _ et al . _
[ cms collaboration ] , `` observation of a new boson at a mass of 125 gev with the cms experiment at the lhc , '' phys .
b * 716 * , 1 , 30 - 61 ( 2012 ) 10.1016/j.physletb.2012.08.021 [ arxiv:1207.7235 [ hep - ex ] ] .
g. aad _ et al . _ [ atlas collaboration ] , `` a search for high - mass resonances decaying to @xmath117 in @xmath118 collisions at @xmath119 tev with the atlas detector , '' jhep * 07 * , 157 ( 2015 ) .
10.1007/jhep07(2015)157 [ arxiv:1502.07177 [ hep - ex ] ] .
b. dutta , a. gurrola , w. johns , t. kamon , p. sheldon and k. sinha , `` vector boson fusion processes as a probe of supersymmetric electroweak sectors at the lhc , '' phys .
d * 87 * , no . 3 , 035029 ( 2013 ) doi:10.1103/physrevd.87.035029 [ arxiv:1210.0964 [ hep - ph ] ] .
b. dutta , t. ghosh , a. gurrola , w. johns , t. kamon , p. sheldon , k. sinha and s. wu , `` probing compressed sleptons at the lhc using vector boson fusion processes , '' phys .
d * 91 * , 055025 ( 2015 ) 10.1103/physrevd.91.055025 [ arxiv:1411.6043 [ hep - ph ] ] .
b. dutta , w. flanagan , a. gurrola , w. johns , t. kamon , p. sheldon , k. sinha , k. wang and s. wu , `` probing compressed top squarks at the lhc at 14 tev , '' phys . rev .
d * 90 * , 095022 ( 2014 ) 10.1103/physrevd.90.095022 [ arxiv:1312.1348 [ hep - ph ] ] .
m. aaboud _ et al . _ [ atlas collaboration ] , `` search for resonances in diphoton events at @xmath120=13 tev with the atlas detector , '' jhep * 1609 * , 001 ( 2016 ) doi:10.1007/jhep09(2016)001 [ arxiv:1606.03833 [ hep - ex ] ] .
v. khachatryan _ et al . _
[ cms collaboration ] , `` search for supersymmetry in the vector - boson fusion topology in proton - proton collisions at @xmath121 tev , '' jhep * 1511 * , 189 ( 2015 ) doi:10.1007/jhep11(2015)189 [ arxiv:1508.07628 [ hep - ex ] ] .
v. khachatryan _ et al .
_ [ cms collaboration ] , `` search for dark matter and supersymmetry with a compressed mass spectrum in the vector boson fusion topology in proton - proton collisions at @xmath119 tev , '' cms - sus-14 - 019 .
j. alwall _ et al .
_ , `` the automated computation of tree - level and next - to - leading order differential cross sections , and their matching to parton shower simulations , '' jhep * 1407 * , 079 ( 2014 ) doi:10.1007/jhep07(2014)079 [ arxiv:1405.0301 [ hep - ph ] ] .
j. alwall _ et al .
_ , `` comparative study of various algorithms for the merging of parton showers and matrix elements in hadronic collisions , '' eur .
j. c 53 , 473 ( 2008 ) doi:10.1140/epjc / s10052 - 007 - 0490 - 5 , arxiv:0706.2569 . v. khachatryan _ et al . _
[ cms collaboration ] , performance of @xmath20-lepton reconstruction and identification in cms , " j. instrum . * 07 * , p01001 ( 2012 ) doi:10.1088/1748 - 0221/7/01/p01001 [ arxiv:109.6034 [ hep - ex ] ] . | new massive resonances are predicted in many extensions to the standard model of particle physics and constitutes one of the most promising searches for new physics at the lhc .
we present a feasibility study to search for new heavy neutral gauge bosons using vector boson fusion ( vbf ) processes , which become especially important as the lhc probes higher collision energies . in the context of the sequential standard model ( seqsm ) , which predicts a neutral spin-1 @xmath0 boson with the same couplings to the sm particles as the z boson
, we find that the vbf @xmath0 production cross - section surpasses that of the drell - yan process at tev scale masses .
we consider the @xmath1 and @xmath2 decay modes to show that the requirement of a dilepton pair combined with two high @xmath3 forward jets with large separation in pseudorapidity and with large dijet mass is effective in reducing sm backgrounds , providing expected exclusion bounds ( at 95% confidence level ) of @xmath4 tev and @xmath5 tev in the @xmath6 and @xmath7 channels , respectively , assuming 30 fb@xmath8 of 13 tev data from the lhc .
the use of the vbf topology to search for massive neutral gauge bosons increases the discovery reach at the lhc , with expected significances greater than 5@xmath9 ( 3@xmath9 ) for @xmath0 masses up to 2.2 ( 2.9 ) tev and 3.4 ( 4.0 ) tev in the @xmath6 and @xmath7 channels . |
value transformation ( cvt ) is one of the first and foremost transformation in the field of manipulating the strings of bits and other transformations of similar nature such as extreme value transformation ( evt ) [ 2 ] , 2- variable boolean operation ( 2-vbo ) [ 4 ] , integral value transformations ( ivts ) [ 5 ] and so on came after that .
all these transformations have lots of applications in the field like pattern formations [ 2 , 4 ] , solving round rabin tournament problem [ 6 ] , collatz - like functions [ 5 ] and so forth . in [ 7 ] with the help of cellular automata , cvt has been used for efficient hardware design of some basic arithmetic operations .
as cvt is one of the important transformations in this area of study so further properties of cvt should have been thoroughly developed for some of its other future scope . in the field of data structure for the organization of non - linear data
we have seen many tree structures like binary tree , avl tree , b tree , b+ tree etc . and many of their applications .
in this paper , we have designed a new tree structure named as cvt - xor tree in the domain of cvt and xor . for this tree
two fundamental logics of cvt and xor from [ 3 ] are used : ( 1 ) in [ 3 ] it is proved that addition of any two non - negative integers expressed as binary numbers is same as addition of their cvt and their xor values .
this result is also shown to be true for any base of the number system .
( 2 ) it has also been proved that in a successive addition of cvt and xor of any two non - negative integers , the maximum number of iterations required to get either cvt=0 or xor=0 is equal to the length of the bigger integer expressed as a binary string .
organization of this paper is as follows : in section 2 some preliminaries of cvt and xor along with a recursive algorithm for addition of two numbers are discussed .
section 3 deals with the fundamentals of cvt - xor tree and two approaches for the construction of cvt - xor tree .
first part in section 4 discusses the different properties of cvt - xor tree and in second part some other significant properties of cvt - xor tree from the p , d and f matrices are enumerated .
section 5 deals with the conclusion of this paper along with some future research direction .
cvt and xor are two transformations defined on a pair of non - negative integers expressed in binary .
interested reader can refer [ 1 , 2 , 3 ] for the definition of cvt which we are omitting for shortening the paper size . some of the properties which we will be using in this paper are enumerated as follows : a ) : : cvt is always an even number and length is ( n+1 ) bits .
b ) : : it has been proved in [ ] that addition of two non - negative integers x and y is equal to the sum of their cvt and xor values i.e. x+y = cvt(x , y ) + xor(x , y ) and also the recurrence scheme always converges to ( 0 , x+y ) in at most ( n+1 ) iterations where n is the maximum number of bits required to represent the bigger number . * _ algorithm 1 : _ * recursive algorithm for addition of two non - negative integers : @xmath0 @xmath1 @xmath2 x @xmath3 @xmath4 @xmath5 as at the time of execution every recursive algorithm forms an activation record in a tree form so we are motivated to construct the tree for the above recurrence algorithm named as cvt - xor tree and analyzed some of its significant features .
if we start from a pair of numbers as root node of the form ( 0 , n ) and back tracking from it using the iterative process x+y
= cvt(x , y)+xor(x , y ) [ ] we come across many intermediate nodes of the form ( cvt , xor ) pair then we will get an m - ary tree . for a given number n
, each node in the tree is represented by a pair or co - ordinate of the form ( x , y ) such that x+y = n .
first co - ordinate i.e. x is the resultant cvt of two numbers and second co - ordinate i.e. y is the resultant xor of two numbers whose sum is n. as ( cvt ( 0 , n ) , xor ( 0 , n ) ) = ( 0 , n ) so and this forms a self - loop at the root node . thus a cvt - xor tree is like an m - ary tree with a loop only at the root node along with several non - leaf and leaf nodes .
depending of n value we have two types of tree .
if n is even , then the tree is called even cvt - xor tree and if n is odd , then the tree is called odd cvt - xor tree . *
_ illustration 1 : _ * let an even integer n=40 then ( 0 , 40 ) is the root node and other leaf and non - leaf nodes in the form ( x , y ) ( such that x+y=40 ) are shown in fig 1 .
as 20 is an even number , so pair ( x , y ) is in the form of either ( even , even ) or ( odd , odd ) . * _ illustration 2 : _ * let an even integer n=25 then ( 0 , 25 ) is the root node and other leaf and non - leaf nodes in the form ( x , y ) ( such that x+y=25 ) are shown in fig 2 . as 25 is an odd number , so pair ( x , y ) is in the form of either ( even , odd ) or ( odd , even ) . using section 2.(b )
the maximum height of the tree is and is the number of significant bits required to represent n. total number of nodes in cvt - xor tree is n+1 which is discussed in section 4 .
depending on the values of n either even or odd two types of trees can be analyzed named as even cvt - xor tree or odd cvt - xor tree .
as cvt of two numbers is always an even number so the first co - ordinate in the node is always even for all internal nodes .
therefore for these pairs , second co - ordinate should be an even number when n is even and is an odd number when n is odd .
but when n is even the leaf nodes can be any pair of the form either ( even , even ) or ( odd , odd ) otherwise it is of the form ( even , odd ) or ( odd , even ) .
following algorithm 2 shows whether a given node ( x , y ) is leaf node or not . *
_ algorithm 2 : _ * algorithm to check whether a given node is leaf or not @xmath6print ( x , y ) is root node@xmath7print ( x , y ) is a leaf node in x+y cvty - xor tree @xmath8print ( x , y ) may be a leaf node or an internal node@xmath9further checking is required @xmath10 + finding the structure of the tree for a given number n two different approaches are discussed below : the root is always of the form ( 0 , n ) this is consider in level-0 and using iterative process to find the nodes in level-1 we will search for the pair of integers x and y such that x+y = n , cvt ( x , y)=0 and xor(x , y)=n except node ( 0 , n ) .
all the nodes in this process are kept in level-1 .
this process is continuing until all the leaf nodes are found i.e. the nodes with no predecessor .
this process starts from all the leaf nodes whose sum is n. all the leaf nodes for the number n can be found easily . for each such pairs
we apply cvt and xor operation and keep those nodes such that their pair sum is equal to n. many nodes may converge towards the single node after cvt and xor operations .
this process is continued until we reach to the root node from all the leaf nodes which are in the form of ( 0 ,
n ) . in the cvt - xor tree
if we backtrack from the root node to the leaf node we find different paths to reach to the leaf nodes at different depths .
it seems that parent child relationship in the cvt - xor tree totally depends on characterization or bit patterns of the node pairs in the path . to understand these dynamics
we have analyzed some of the properties of cvt - xor tree discussed in the following section .
the numbers of nodes or vertices in the cvt - xor tree is n+1 .
number of non - negative pair wise integer partitions of n is n+1 of the form ( 0 , n ) , ( 1 , n-1 )
( n , 0 ) . from section 2(b ) all these pairs must converges to ( 0 , n ) so all these n+1 number of nodes are connected and will form a single cvt - xor tree . hence the total numbers of nodes in the cvt - xor tree is n+1 . * _ illustration 3 : _ * let an even integer n=18 , we have 19 pairs as ( 0 , 18 ) , ( 18 , 0 ) , ( 1 , 17 ) , ( 2 , 16 ) , ( 3 , 15 ) , ( 4 , 14 ) , ( 5 , 13 ) , ( 6 , 12 ) , ( 7 , 11 ) , ( 8 , 10 ) , ( 9 , 9 ) , ( 10 , 8) , ( 11 , 7 ) , ( 12 , 6 ) , ( 13 , 5 ) , ( 14 , 4 ) , ( 15 , 3 ) , ( 16 , 2 ) , ( 17 , 1 ) in cvt - xor tree shown in fig 3 . in cvt - xor tree if the pair ( a , b ) is in depth - d , then its symmetric pair ( b , a ) is also in depth - d except the root node .
let , cvt(a , b)=p and xor(a , b)=q ; where p and q are the parent node ( p , q ) for one child pair ( a , b ) . + we know that , cvt(a , b)=cvt(b , a ) and xor(a , b)=xor(b , a ) + then cvt(a , b)=cvt(b , a)=p and xor(a , b)=xor(b , a)=q + that implies both the pairs ( a , b ) and ( b , a ) are the child nodes whose parent node is ( p , q ) i.e. nodes ( a , b ) and ( b , a ) are two predecessor .
+ therefore , ( a , b)@xmath11 ( p , q ) and ( b , a)@xmath12 .
so they must belong to the same level i.e. in same depth . + according to the cvt properties , if the xor value is zero in @xmath13 iteration , then the cvt value is zero in @xmath14 iteration .
therefore ( n , 0 ) @xmath11 ( 0 , n ) and ( 0 , n ) is the root node in cvt - xor tree paradigm , so if ( 0 , n ) is in level-0 then the node ( n , 0 ) will be always in level-1 in the cvt - xor tree .
* _ illustration 4 : _ * from fig .
3 , the pairs ( 2 , 16 ) and ( 16 , 2 ) are in lavel-1 similarly pairs ( 8 , 10 ) and ( 10 , 8) are in level-2 and so on .
but the root node ( 0 , 18 ) and its symmetric pair ( 18 , 0 ) are in level-0 and level-1 respectively . for any pair ( p , q ) where @xmath15 is an even number and @xmath16 is either even or odd and if @xmath17i ; @xmath18 such that @xmath19 , then the pair ( p , q ) is the leaf node or contradictory ( even , even ) pair .
the theorem statement demands two binary variable @xmath20 and @xmath21 such that @xmath22 and @xmath23 for some i , which is impossible .
so the such pair ( p , q ) must be a ( even , even ) leaf node . * _ illustration 5 : _ * as per the definition of cvt , we know that cvt of any two numbers is always an even number .
therefor there is no predecessor for an odd pair . but some ( even , even ) contradictory pairs are also in leaf node . fig .
4 shows that pair ( 3 , 5 ) and ( 5,3 ) are ( odd , odd ) leaf nodes and ( 6 , 2 ) is a leaf node although the first co - ordinate is even because as third bit and second bit of 6 ( 110 ) and 2 ( 10 ) from lsb position are 1 and 1 respectively .
so the pair ( 6 , 2 ) is a contradictory leaf node . if the pair ( p , q ) where the cvt value @xmath15 and the xor value @xmath16 is a non - leaf node in the cvt - xor tree and if @xmath24 and @xmath25 for m such i positions where @xmath26 and @xmath27 ; then the number of successors of the ( p , q ) node=@xmath28 .
these nodes can be generated by substituting m column positions either @xmath29 or @xmath30 and keeping other ( n - m ) column values fixed .
as m column values has to be changed by @xmath29 or @xmath30 in 2 different ways , so using fundamental principle of counting the total successor pairs will be @xmath28 .
* _ illustration 6 : _ * let an example of ( even , even ) non - leaf node ( 2 , 6 ) shown below : + @xmath31 + here two position of a and b i.e. ( * , * * ) in two columns can be filled up @xmath32 ways .
so according to the theorem 4 such predecessor of ( 2 , 6 ) are ( 1 , 7)=(001,111 ) , ( 7 , 1)=(111 , 001 ) , ( 3 , 5)=(011 , 101 ) and ( 5 , 3)=(101 , 011 ) shown in fig .
4 . let ( p , q)@xmath11(m , n ) , if ( p , q ) and ( m , n ) are two odd - odd pair ( i.e. leaf node ) and immediate even - even pair respectively , then the second last lsb bit of m must be always 1 i.e. @xmath33 and @xmath34 @xmath35 where @xmath36 and @xmath37 .
* + 1st part : * as binary representation of two odd numbers must leading to last lsb bit 1(for both p & q ) , therefore according to cvt definition second last lsb bit of m must be 1 as m = cvt(p , q ) .
+ * 2nd part : * we will proof this part through the principle of method of induction .
+ for n=1 , m=4 - 2=2 , binary representation of 2 is 10 i.e. second last lsb bit is 1 .
so formula is true for n=1 .
+ let the formula be true for n i.e. @xmath38 we have to prove that this is true for m = n+1 i.e. @xmath39 + @xmath40 + so the formula is also be true for m = n+1 . therefor the formula is true for all values greater than equal 1
now we are ready to write the algorithm to check whether a given input node ( x , y ) is a leaf node or not .
algorithm 2 is now modified using theorem 3 & 4 by which a given node whether belong to the leaf or non - leaf can be verified easily is as follows : * _ algorithm 2 : _ * algorithm to verify whether a given node is leaf or not @xmath41and @xmath42where x+y = n is an even number@xmath43print ( x , y ) is root node@xmath44print ( x , y ) is a ( odd , odd ) and ( even , even ) leaf node @xmath9respectively @xmath8print ( x , y ) is a non - leaf node @xmath45 any even integer ( e ) in the form of @xmath46 will be always a cvt value 2 for all possible ( odd , odd ) pairs ( say ( x , y ) ) of e ; where e = x+y . notes the two facts as following : 1 . for every pair ( x , y ) of e such that if e is an n bits number then the maximum number of bit for x and y will be n-1 .
every number in the form m= ( 2k -1 ) is always the addition result of two complementary numbers ( say p and q ) where one of them will be even and another one is odd .
we can represent e as : @xmath47 + so to get e we have to add only 1 ( or binary 1 ) to p ( if p is even ) or q ( if q is even ) such that p+q = x+y .
+ now we get two odd numbers @xmath48 and @xmath49 such that e = m+1= @xmath50 having both lsb position are binary1 others bits position are complementary . therefore according to the cvt definition there
cvt value must be 2 .
now question can be asked that what is the total number of leaf nodes and internal nodes in cvt - xor tree whose is root is ( 0 , n ) ? for answering this and other similar questions
we have defined three different matrices namely cvt - xor depth matrix , cvt - xor parent matrix , and cvt - xor frequency matrix to store depth information , parent information and frequency information respectively . 1 .
the diagonal from ( n , 0 ) to ( 0 , n ) in the cvt - xor depth matrix gives the following information regarding the cvt - xor tree whose root is the ( 0 , n ) ) 1 .
if dm ( i , j)=h then ( i , j ) vertex is in depth - h in the cvt - xor tree .
2 . no . of nodes in depth - h = no .
of times h present on the diagonal of the cvt - xor depth matrix ( frequency of the occurrence of h ) 3 .
average height of the tree= ( sum of all the diagonal value)/n+1 4 .
if the binary pits pattern of n such that only the lsb bit is 0 i.e. number is in the form ( 0 , 2k-2 ) then for such even integer n=2 , 6 , 14 , 32, ,2k-2 ) , the maximum value through the diagonal from ( n , 0 ) to ( n , 0 ) is 2 this implies the maximum height of the tree is always 3 . 5 . if the binary pits pattern of n such all bits are 0 i.e. number is in the form ( 0 , 2k-1 ) then for such odd integers n= 3 , 7 , 15 , 31 2k-1 ) , the maximum value through the diagonal from ( n , 0 ) to ( n , 0 ) is 1 this implies the maximum height of the tree is always 2 . once we will find the number of nodes in different depth , link information ( ) among the nodes in cvt - xor tree can be found form the cvt - xor parent matrix .
the diagonal from ( n , 0 ) to ( 0 , n ) in the cvt - xor parent matrix gives the following information regarding the cvt - xor tree whose root is the ( 0 ,
n ) ) 1 . if pm ( i , j)=(p , q ) this implies ( p , q ) is the parent of node ( i , j ) .
if ( p , q ) is repeats m times in the pm matrix then there is m such similar ( i , j ) whose parent is ( 0 , n ) 3 .
the frequency matrix ( fm ) help us to find the following : let , fm ( i , j ) = m , where m indicate the number of predecessor or child nodes of the node ( i , j ) .
if i is odd then the corresponding value of m is 0 , this indicate if the first co - ordinate of the node is odd and the node has no child or predecessor i.e. the node is leaf node .
2 . through the diagonal value we can find what the leaf nodes are for a given n. below in fig . 8 shown the cvt value for all ( odd , odd ) pairs . according to the theorem 5
, cvt value for all ( odd , odd ) pairs in this figure contains all the numbers from the set s. where 2 is the smallest number and the difference between two consecutive number is 4 i.e. @xmath51 .
it is also showing the beautiful fractal studying in [ 1 ] . if we traverse through a diagonal path from ( k , 10 ) ( where k is any odd number and gradually decreasing k ) , ( k-2 , 3 ) , ... , ( 1 , k ) figure ( fig .
9 ) like palindrome triangle is obtained .
this figure demonstrate that for any arbitrary given positive integer all the possible cvt values where bold cvt values refer to the prime pair solution .
this paper shows a new tree structure called cvt - xor tree along with two different approaches for the construction of the tree .
we also have seen different properties of the tree like number of nodes in any depth , characterization of child , parent , internal and leaf nodes .
some other significant properties also can be obtained easily from the three different matrices along with an easy construction process of cvt - xor tree from the matrices . in future
this work will help us in further studies in the domain of cvt - xor to characterize the pattern formation by different cvt - xor trees where it is conjectured that there exists always a path from root(zero , even integer ) to leaf ( prime pair ) signifying the goldach conjecture - the unfinished work remaining , how to distinguish this kind of paths from the rest .
we can further explore the tree structure by considering three numbers combination such that x+y+z = n and to study the different properties of such trees .
the authors would like to thank ... 1 s. pal , s. shaoo and b. k. nayak , _ properties of carry value transformation_,international journal of mathematics and mathematical sciences volume 2012 , article i d 174372 , 10 pages http://dx.doi.org/10.1155/2012/174372 , 2012 p. pal choudhury , sk .
s. hassan , s. shaoo and b. k. nayak , _ act of cvt and evt in the formation of number theoretic fractals_,international journal of computational cognition ( http://www.ijcc.us ) , vol .
1 , march 2011 .
p. pal choudhury , s. shaoo and b. k. nayak , _ theory of carry value transformation and its application in fractal formation _
, ieee international advance computing conference , doi : 10.1109/iadcc.2009.4809146 , 2009 .
p. pal choudhury , hassan sk .
s.,sahoo s. and b. k. nayak , _ theory of rule 6 and its application to round robin tournament _ , arxiv 0906.5450v1 , cs.dm , cs.gt , int .
journal of computational cognition , vol . 8 , no .
33 - 37 , sep . 2010 .
p. pal choudhury , s. shaoo and m. chakraborty , _ implementation of basic arithmetic operations using cellular automata _ ,
ieee computer society , isbn:978 - 0 - 7695 - 3513 - 5,http://doi.ieeecomputersociety.org/10.1109/icit.2008.18 pp : 79 - 80,dec .
17 , 2008 to dec . 20 , 2008 | cvt and xor are two binary operations together used to calculate the sum of two non - negative integers on using a recursive mechanism . in this present
study the convergence behaviors of this recursive mechanism has been captured through a tree like structure named as cvt - xor tree . we have analyzed how to identify the parent nodes , leaf nodes and internal nodes in the cvt - xor tree .
we also provide the parent information , depth information and the number of children of a node in different cvt - xor trees on defining three different matrices .
lastly , one observation is made towards very old mathematical problem of goldbach conjecture .
carry value transformation ( cvt ) , exclusive or ( xor ) , cvt - xor tree , fractal , goldbach conjecture . |
recently a new model of restricted @xmath0-gravity has been proposed in @xcite .
this proposal is based on the idea of a mild breaking of the diffeomorphism invariance of the four- dimensional @xmath0 gravity .
recall that @xmath5-gravity is described by the action @xmath6 + s_{matter } \ , \ ] ] where @xmath7 is the reduced planck mass and @xmath8 is a dimensionless constant .
this action is invariant under the full four - dimensional diffeomorphism by construction . in order to find the restricted form of @xmath0-gravity
, it is useful to introduce the @xmath9-decomposition of the metric @xmath10 @xcite @xmath11 where we have defined @xmath12 as the inverse of the induced metric @xmath13 on the cauchy surface @xmath14 at each time @xmath15 , @xmath16 and we denote @xmath17 .
the four - dimensional scalar curvature in @xmath9 formalism has the form @xmath18 where the extrinsic curvature of the spatial hypersurface @xmath14 at the time @xmath15 is defined as @xmath19 with @xmath20 being the covariant derivative determined by the metric @xmath13 , and where the de witt metric is defined as @xmath21 with inverse @xmath22 which obeys the relation @xmath23 further , @xmath24 is the future - pointing unit normal vector to the hypersurface @xmath14 , which is written in terms of the adm variables as @xmath25 in order to formulate the restricted @xmath0-gravity we break the full diffeomorphism invariance of the @xmath0-gravity by performing the replacement @xmath26 where @xmath27 is a dimensionless parameter that controls the breaking of the diffeomorphism invariance of the action .
even if the shift ( [ i-2 ] ) seems to be very mild , it turns out that it has significant impact on the hamiltonian structure of this theory as was shown in @xcite .
careful analysis performed there showed that in order to have a consistent theory from the hamiltonian analysis point of view , it is necessary to include terms which depend on the spatial derivative of the lapse @xmath28 into the action .
further , the breaking of the diffeomorphism invariance suggests the possibility to define theory with the generalized de witt metric @xcite @xmath29 in summary , we proposed in @xcite the extended form of the restricted @xmath0-gravity when we performed the replacement @xmath30 where @xmath31 are the corresponding coupling constants .
we showed there that this theory is consistent from the hamiltonian analysis point of view since the structure of constraints is the same as in case of non - projectable hoava - lifshitz gravity @xcite , for the hamiltonian analysis , see @xcite and also @xcite .
we also analyzed cosmological solutions of the restricted @xmath0-gravity and we found new solutions whose properties depend on the value of the parameter @xmath27 .
it is important to emphasize that some of these solutions can not be found in the diffeomorphism invariant @xmath0-gravity . due to the fact that there are new cosmological solutions ,
it is natural to investigate them in more details .
the aim of this paper is to focus on the analysis of the cosmological fluctuations of the restricted @xmath0-gravity .
as the first step we determine the background equations from the restricted @xmath0-gravity action when we focus on i a time - dependent ansatz . then we analyze the fluctuations above this background . due to the fact that the scalar , vector and tensor fluctuations decouple at the quadratic order we can analyze each kind of the fluctuations separately .
we show that the restricted @xmath0-gravity differs from the standard @xmath0-gravity in the scalar sector , while the vector and tensor sectors have the same properties as in the case of diffeomorphism invariant @xmath0-gravity .
more precisely , we show that with suitable choice of parameters the restricted @xmath0-gravity allows to explain recent observation data that predict the possibility of the existence of a non - zero anisotropic stress ( see for example @xcite . ) . this paper is organized as follows . in the next section ( [ second ] )
we briefly review the equations of motion of the restricted @xmath0-gravity evaluated on the time - dependent cosmological solution . in section ( [ third ] ) we analyze the fluctuations above this background and in section ( [ fourth ] ) we discuss our results with relation to the recent phenomenological observations .
finally , in conclusion ( [ fifth ] ) we outline our results .
in this section we derive the equations of motion for the restricted @xmath0-gravity when we presume the friedmann - robertson - walker ( frw ) form of the background . recall that the restricted @xmath32-gravity action has the form @xmath33+s_{matter } \ , \ ] ] where @xmath34 is defined in ( [ repla ] ) .
the matter part contributes to the gravitational equations of motion through the stress - energy tensor which is defined as @xmath35 our goal is to analyze the spatially homogeneous and isotropic universe so that the metric ansatz is the frw metric which has the form @xmath36 where @xmath37 is the lapse and @xmath38 is the scale factor . as usual ,
the hubble parameter is defined as @xmath39 .
note that we can not set @xmath40 from the beginning due to the restricted form of the diffeomorphism .
but , since for the background equations all quantities depend only on time , one can use redefinition of time in to set @xmath41 in the background equations .
we also presume matter in the form of a prefect fluid which means that the stress - energy tensor has the form @xmath42 since we presume that the full diffeomorphism is broken in the gravitational sector only we find that the matter action is diffeomorphism invariant . as a result
the stress - energy tensor is conserved in the sense @xmath43 on the condition that the matter satisfies the equation of motion .
now for the background ( [ 1 - 1 ] ) we obtain @xmath44 so that the conservation of the stress - energy tensor implies the standard conservation equation @xmath45 using ( [ 1 - 1 ] ) and the general relations in the adm formalism , which are discussed in ref .
@xcite , one can show that @xmath46 the generalized friedmman equation for the model has been determined in @xcite as @xmath47 where @xmath48 note that we have defined prime as derivative with respect to the argument of @xmath49 . as a result ,
@xmath50 are dimensionless quantities .
finally , we note that there is still another equation after performing the time derivative of eq . and then use eq .
@xmath51 after this brief review of the background equations we now switch to the main topic of this paper which is the analysis of the cosmological perturbations .
the goal of this section is to derive equations for the perturbed frw space - time in the arnowitt - deser - misner(adm ) formalism . as a check of the validity of our approach
, we note that for @xmath52 and @xmath53 the results derived in this paper should reduce to the corresponding results of the standard @xmath54 gravity , see for example the review @xcite . to begin with , we emphasize that we are writing our equations in the newtonian gauge which is defined by @xmath55 .
we use the newtonian gauge for the following reason : as it has been argued in @xcite , one of the advantages of the newtonian gauge is that the physical fields which are defined by the gauge fixing , coincide with the gauge - invariant variables .
note also that it is an easy task to move from the newtonian gauge to other gauges @xcite .
finally , note that due to the fact that the spatial section of the metric ( [ 1 - 1 ] ) is flat , it is natural to use the fourier decomposition of the perturbation where the fourier components of a general perturbation @xmath56 are defined as @xmath57 where @xmath58 .
further , we also decompose @xmath59 and @xmath60 into the homogeneous and perturbed parts as @xmath61 where @xmath62 over any quantity shows the unperturbed part of that quantity . in the case of the fluctuations of the matter we use the following parameterization for the perturbed energy - momentum tensor @xmath63 where @xmath64 is the potential for the spatial velocity of the fluid . + before we proceed to the study of the perturbed equations it is instructive to discuss the role of diffeomorphism symmetry in the cosmological context .
+ for a full diffeomorphism invariant model , as for example the einstein - hilbert action or @xmath0gravity action , there is no a priori preferred coordinate system .
of course the symmetry can be broken by imposing other restrictions on the model .
for example , in the cosmological context , it has been shown that the only coordinate system(observer ) that is compatible with the assumption of homogeneity and isotropy of space is described by the frw metric @xcite .
the uniqueness of frw metric helps us to choose it as the background metric for our model in this paper .
but , the uniqueness of the metric is broken if we consider universe which is not homogeneous . the deviation from homogeneity
is considered as the perturbed frw metric .
for example , the scalar metric perturbations can be parameterized as @xcite @xmath65dx^{i}dx^{j } \ , \end{aligned}\ ] ] where @xmath66,@xmath67 , @xmath68 , @xmath69 are @xmath70scalars .
+ it is an unnecessary and senseless task to insert the above expression into general equations of motion in order to obtain the equations of motion for these @xmath70scalars .
the reason for this statement is that , similar to the model with the full diffeomorphism symmetry , here we confront with the so - called gauge problem .
the problem arises when we note that the model is invariant under the coordinate transformation as @xmath71 , @xmath72 , where @xmath73 . under this transformation
the components of the metric ( [ note - add ] ) transform as @xcite @xmath74 therefore , even if we tried to obtain the equations for the components of eq .
, we would have some solutions which are unphysical in the sense that they can be derived by application of the transformation ( [ note - add1 ] ) on some particular solutions . in order to avoid the above problem , i.e. the gauge problem
, one can choose a specific gauge ( coordinate ) or use the gauge - invariant quantities .
in the process of the gauge fixing , values for @xmath75 and @xmath76 are explicitly specified such that we have not any residual symmetry for the solution . _ physically , gauge fixing means that a specific spatial coordinate has been chosen_. + in the following we will work with the newtonian gauge .
the newtonian gauge is defined as the perturbed metric for observers who are immobile at the hypersurfaces of constant time . also , the normal vector of the hypersurfaces are the same as the worldlines of the observers at any time .
thus , the newtonian gauge is described by @xcite @xmath77dx^{i}dx^{j } \ .\end{aligned}\ ] ] the newtonian gauge is the preferred gauge for the late - time cosmology as was argued in @xcite .
therefore , we will use it to compare the results of our model with the plank observations .
finally we should emphasize that it is still possible to construct the gauge invariant quantities form the newtonian gauge variables @xcite .
+ to proceed further , we note that it is convenient to parameterize the scalar metric perturbations in the newtonian gauge as @xmath78 where @xmath79 and @xmath80 are space and time dependent . using these definitions in ( [ repla ] )
we easily find @xmath81 as a check , we note that for @xmath82 the above relation reduces to the background form of @xmath83 given in eq .. now by performing the linearization of the expression ( [ rper ] ) , we obtain fourier component of @xmath84 @xmath85 where @xmath86 .
further , inserting ( [ fl1 ] ) and ( [ defnfluc ] ) into the equation ( [ consten ] ) and by performing the corresponding linearization , we obtain two equations @xmath87 and @xmath88 where @xmath89 .
note that these equations have the same forms as the corresponding equations for the usual @xmath90 gravity . in order to derive the remaining equations , we proceed in the following way . to begin with ,
we insert into the action . as we pointed out we work in the newtonian gauge .
then in order to vary the action with respect to the shift , it is sufficient to consider the terms which are proportional to @xmath91 and @xmath92 . for example , if we define @xmath93 as a variation with respect to the shift then for @xmath94 we obtain @xmath95 using this result and also using the following relations @xmath96 and after some integration by parts we obtain @xmath97 \partial_{i}n_{i}(1 + 2\eta\bar{f})-4 a \eta m_{p}^{2}\int d^{4}x a h n_{i}\partial_{i}\delta f\\ & -2\eta\upsilon m_{p}^{2}\int d^{4}x a[n_{i}\partial_{i}\dot{\delta f}-\partial_{i}\phi n_{i}\dot{\bar{f}}-3h n_{i}\partial_{i}\delta f]+ s_{matter } \ . \end{split}\ ] ] further on , performing the variation with respect to @xmath98 and using the fourier components of the perturbations we obtain @xmath99 -\frac{1}{2m_{p}^{2}}\delta q \ .\ ] ] in case of the variation of the action with respect to @xmath79 we set @xmath100 in the action and then expand the action up to the second order in @xmath79 and @xmath80 .
this procedure , after some integration be parts and using eq .
, leads to @xmath101 \ \end{split}\ ] ] so that we easily find the fourier form of the equation of motion for @xmath79 @xmath102\delta f -3\upsilon\eta\dot{\bar{f}}(2h\phi+\dot{\psi})\\ & + 3\eta\upsilon h\dot{\delta f}-\frac{\delta\rho}{2m_{p}^{2 } } = ( 3a\phi h^{2}+\frac{k^{2}}{a^{2}}\psi + 3ah\dot{\psi}-\gamma_{1}\frac{k^{2}}{2 a^{2}}\phi)(1 + 2\eta\bar{f } ) \ . \end{split}\ ] ] for reasons that will become clear later we derive @xmath103 in two different ways . in the first case we use and . then in order to eliminate @xmath104 in these formula , we use . collecting these pieces together
we obtain an equation @xmath105\delta f\\ & + \big[\dot{\phi}h+2a\dot{h}\phi+a\ddot{\psi}+3a\phi h^{2}+3ah\dot{\psi}+\frac{k^2}{3a^{2}}(\psi - a\phi)\\ & \hspace{.15cm}+(1-a)\frac{k^{2}}{3ha^{2}}\dot{\psi}- \gamma_{1}\frac{k^{2}}{6a^{2}}(\frac{\dot{\phi}}{h\phi}+1)\phi\big](1 + 2\eta \bar{f } ) \ , \end{split}\ ] ] where we also used the fact that @xmath106 .
+ the second way how to derive the relation for @xmath103 is to use eqs . , .
we again use eq . in order to eliminate @xmath104 , so that we obtain @xmath107 \dot{\bar f}-2\eta a h\dot{\delta f}+2\eta\upsilon\phi\ddot{\bar f}-\eta\upsilon\ddot{\delta f}\\ & + \eta\big[(3\upsilon-2a)\dot{h}+(9\upsilon-6a)h^{2}\big ] \delta f\\ & + \big[\dot{\phi}h+2\dot{h}\phi+\ddot{\psi}+3\phi h^{2}+3h\dot{\psi}\big](1 + 2\eta \bar{f } ) \ . \end{split}\ ] ] the right - hand side of eqs . and are the same if @xmath108\big[1 + 2\eta \bar{f}\big]\\ & = 2\eta\frac{k^{2}}{3a^{2}}(2\upsilon - a ) \delta f+\eta\big[2(1-a)h\phi+2(1-a)\dot{\psi } + \gamma_{1}\frac{k^{2}}{3ha^{2}}\phi\big]\dot{\bar{f } } \ . \end{split}\ ] ] as is clear , the equations are simplified by taking @xmath109 , which results in @xmath110 . as a check , note that for @xmath111 and @xmath112 these equations match the corresponding relations for the usual @xmath113 gravity .
in this section we focus on the metric tensor perturbations @xmath114 . recall that the perturbed line element has the form @xmath115dx^{i}dx^{j},\ ] ] where @xmath116 . in order to derive equations in this sector
we use the fact that @xmath117 does not contain the metric tensor perturbations up to the second order .
further , using the traceless condition on @xmath114 it is easy to show that the metric perturbations do not appear in @xmath118 .
thus , it turns out that the terms in which are proportional to @xmath27 and @xmath119 do not contribute to the metric tensor perturbations . as a result
the analysis of this sector is very similar to the analysis of tensor fluctuations in standard @xmath0 gravity .
explicitly , the second order action for these modes has the form @xmath120\big[a \gamma_{ij}\partial^{2}\gamma_{ij}+a^{3}\dot{\gamma}_{ij}^{2 } \big ] \ .\ ] ] then in order to avoid the ghost instability we must impose the following condition @xmath121 performing variation of with respect to @xmath114 and using the following fourier representation @xmath122 where @xmath123 and @xmath124 , leads to the second order differential equation for @xmath125 @xmath126+(\frac{k}{a})^{2}\gamma_{k}^{s}=0\ ] ] which has the same form as in the standard @xmath0-gravity . in this section
we perform an analysis of vector metric perturbations . as in the case of the prefect fluid we consider the following form of the perturbed stress - tensor in the vector sector @xmath127
then from @xmath128 it follows that @xmath129 as in case of the metric perturbations the favorite gauge in this sector is the so - called vector gauge which is defined by @xmath130 where @xmath131 .
again , from the above definition and the condition on @xmath132 it turns out that the terms in which are proportional to @xmath27 do not contribute in this sector . then the second order action takes the following form @xmath133 then it is easy to see that the equation for @xmath132 in the fourier space has the form @xmath134 which coincides with the equation for the vector perturbation in the standard @xmath0-gravity .
in this section we explore some phenomenological implications of our results . in this section
we take @xmath110 that corresponds to @xmath109 .
note that this is a natural presumption , since it is expected that @xmath119 should approach to @xmath135 in the low energy regime @xcite .
it is important to emphasize that the einstein - hilbert gravity predicts @xmath136 . in the more general case
when @xmath137 , it is useful to define an anisotropic stress , @xmath138 as @xmath139 clearly a non - vanishing anisotropic stress is the signature of modification of general relativity @xcite . further , it is also convenient to define @xmath140 so that the einstein - hilbert gravity predicts @xmath141 . + but , from the planck cmb temperature data ,
the current value for @xmath142 reported as @xcite @xmath143 which is shown by yellow and gray colors in fig . 1 .
on the other hand if we consider the weak lensing data , we have @xcite @xmath144 which is shown by the gray and blue colors in fig
+ let us now return to our model and consider the late - time cosmology when we can neglect the radiation and the cold dark matter . explicitly , we consider two cases : * _ _ case i _ _ : in this case we presume that the cosmological constant @xmath145 is non - zero while @xmath146
. + the energy density of the cosmological constant is @xmath147 so that implies @xmath148 note that this result is the same as in the case of einstein - hilbert gravity with the cosmological constant .
+ since for the cosmological constant we have @xmath149 , from eqs .
and we have + [ des23 ] @xmath150 + if we combine these two equations together , we obtain @xmath151 that has solution @xmath152 where @xmath153 and @xmath154 are arbitrary constants .
it is clear that in order to avoid instability we have to require that @xmath155 .
+ using now eqs . , and , we obtain @xmath156 \ ( \hspace{.2 cm } \text{for } \ \gamma_{1 } \neq2)\ ] ] and @xmath157 for the present universe we have @xmath158 and it is easy to see that we can choose the parameters in the above equations to obtain a consistent result with the reported data for @xmath159 . note also that the above relations show that if we take @xmath160 , we have @xmath161 * _ _ case ii _ _ : in the second case , we neglect the cosmological constant but @xmath162 so that we have only two independent equations which determine @xmath79 and @xmath80 . one of these equations is obtained by multiplying eq . with @xmath163 and then using the corresponding result to eliminate some terms in eq .. this procedure gives the following equation(for @xmath110 ) @xmath164(1 + 2\eta \bar{f})=-6\eta\upsilon[\dot{\psi}+h\phi]\dot{\bar{f } } + 6\eta\upsilon(\frac{k^{2}}{3a^{2}}-\dot{h})\delta f \ .\ ] ] further , eq . takes the following form for @xmath110 @xmath165\big[1 + 2\eta \bar{f}\big]\\ = 2\eta(2\upsilon-1)\delta f+\frac{\eta\gamma_{1}}{h}\phi\dot{\bar{f}}.\ ] ] in the de sitter space , which is a very good approximation for the late - time cosmology , eqs . and simplify considerably @xmath166(1 + 2\eta \bar{f})= 2\eta\upsilon\delta f\ ] ] and @xmath167\big[1 + 2\eta \bar{f}\big]\\ = 2\eta(2\upsilon-1)\delta f.\ ] ] if we combine these two equations , we obtain the following results + [ des3 ] @xmath168 + note that for @xmath169 , the first equation implies @xmath170 + space of parameters ( @xmath171,@xmath27 ) for the model .
the yellow and gray regions show the values for @xmath159 from the planck cmb temperature data @xcite .
the gray and blue regions show the values for @xmath159 from the combination of the planck cmb temperature data and the weak lensing @xcite .
solid line represents the prediction of the model in the de sitter space for @xmath172 .
] + finally when we insert and ( [ dotphihphi ] ) into eq . or eq . , we obtain the equations for the remaining variables that can be easily solved .
equations are interesting results since they give a relation between the parameters of the model and observation .
if we take @xmath172 limit , it follows that + [ des4 ] @xmath173 + note that , except for @xmath169 , in this limit ( i.e @xmath172 ) the above relations do not depend on the form and dynamics of @xmath49 in the action . as shown in fig .
1 , for @xmath172 one can choose @xmath27 to reconcile the model with the current observations .
this short note is devoted to the analysis of the cosmological fluctuations of the restricted @xmath0-theory of gravity .
we have determined the background equations of motion and then we have carefully analyzed the fluctuations above this background solution .
we show that the vector and tensor fluctuations have the same dynamics as in case of the standard @xmath0-gravity while the scalar sector possesses new interesting possibilities which depend on the values of coupling constants . in more details , we show that it is possible to choose the values of these parameters so that the predictions of the restricted @xmath0-gravity are in agreement with recent observation data .
in fact , we showed previously in @xcite that the cosmological solutions found in the restricted @xmath0-gravity are in agreement with observation and our current analysis of fluctuations confirms this fact as well .
this is by itself an interesting result that suggests that it is indeed plausible to consider theories with the restricted diffeomorphism invariance as interesting alternatives to the fully diffeomorphism invariant theories of gravity .
99 p. a. r. ade _ et al .
_ [ planck collaboration ] , _ `` planck 2015 results
constraints on inflation , '' _
arxiv:1502.02114 [ astro-ph.co ] .
m. chaichian , a. ghalee and j. kluson , _
`` restricted f(r ) gravity and its cosmological implications , '' _ phys .
d * 93 * ( 2016 ) 104020 [ arxiv:1512.05866 [ gr - qc ] ] .
a. de felice and s. tsujikawa , _
`` f(r ) theories , '' _ living rev . rel .
* 13 * ( 2010 ) 3 [ arxiv:1002.4928 [ gr - qc ] ] .
r. l. arnowitt , s. deser and c. w. misner , _ `` the dynamics of general relativity , '' _ gen .
* 40 * ( 2008 ) 1997 doi:10.1007/s10714 - 008 - 0661 - 1 [ gr - qc/0405109 ] .
e. gourgoulhon , _ `` 3 + 1 formalism and bases of numerical relativity , '' _ arxiv : gr - qc/0703035 .
p. horava , _
`` membranes at quantum criticality , '' _ jhep * 0903 * ( 2009 ) 020 [ arxiv:0812.4287 [ hep - th ] ] .
p. horava , _ `` quantum gravity at a lifshitz point , '' _ phys.rev .
d * 79 * ( 2009 ) 084008 [ arxiv:0901.3775 [ hep - th ] ] .
d. blas , o. pujolas and s. sibiryakov , _
`` consistent extension of horava gravity , '' _ phys .
* 104 * ( 2010 ) 181302 [ arxiv:0909.3525 [ hep - th ] ] .
d. blas , o. pujolas and s. sibiryakov , _
`` models of non - relativistic quantum gravity : the good , the bad and the healthy , '' _ jhep * 1104 * ( 2011 ) 018 [ arxiv:1007.3503 [ hep - th ] ]
. d. blas , o. pujolas and s. sibiryakov , _
`` comment on ` strong coupling in extended horava - lifshitz gravity ' , '' _ phys .
b * 688 * ( 2010 ) 350 [ arxiv:0912.0550 [ hep - th ] ] .
j. kluson , _ `` note about hamiltonian formalism of healthy extended horava - lifshitz gravity , '' _ jhep * 1007 * ( 2010 ) 038 [ arxiv:1004.3428 [ hep - th ] ]
. w. donnelly and t. jacobson , _ `` hamiltonian structure of horava gravity , '' _ phys .
d * 84 * ( 2011 ) 104019 [ arxiv:1106.2131 [ hep - th ] ] .
s. mukohyama , r. namba , r. saitou and y. watanabe , _
`` hamiltonian analysis of nonprojectable hooava - lifshitz gravity with @xmath174 symmetry , '' _ phys .
d * 92 * ( 2015 ) 2 , 024005 [ arxiv:1504.07357 [ hep - th ] ] .
m. chaichian , j. kluson and m. oksanen , _
`` nonprojectable horava - lifshitz gravity without the unwanted scalar graviton , '' _ phys .
d * 92 * ( 2015 ) no.10 , 104043 doi:10.1103/physrevd.92.104043 [ arxiv:1509.06528 [ gr - qc ] ] .
j. kluson , _ `` horava - lifshitz f(r ) gravity , '' _ jhep * 0911 * ( 2009 ) 078 [ arxiv:0907.3566 [ hep - th ] ] .
j. kluson , s. nojiri , s. d. odintsov and d. saez - gomez , _
`` u(1 ) invariant @xmath175 horava - lifshitz gravity , '' _ eur . phys .
j. c * 71 * ( 2011 ) 1690 [ arxiv:1012.0473 [ hep - th ] ] .
j. kluson , _ `` note about hamiltonian formalism of modified @xmath175 hoava - lifshitz gravities and their healthy extension , '' _ phys .
d * 82 * ( 2010 ) 044004 [ arxiv:1002.4859 [ hep - th ] ] .
j. kluson , _
`` new models of f(r ) theories of gravity , '' _ phys.rev .
d * 81 * ( 2010 ) 064028 [ arxiv:0910.5852 [ hep - th ] ] .
s. carloni , m. chaichian , s. nojiri , s. d. odintsov , m. oksanen and a. tureanu , _
`` modified first - order horava - lifshitz gravity : hamiltonian analysis of the general theory and accelerating frw cosmology in power - law f(r ) model , '' _ phys .
d * 82 * ( 2010 ) 065020 [ phys . rev .
d * 85 * ( 2012 ) 129904 ] [ arxiv:1003.3925 [ hep - th ] ] .
m. chaichian , s. nojiri , s. d. odintsov , m. oksanen and a. tureanu , _ `` modified f(r ) horava - lifshitz gravity : a way to accelerating frw cosmology , '' _ class .
* 27 * ( 2010 ) 185021 [ class . quant .
* 29 * ( 2012 ) 159501 ] [ arxiv:1001.4102 [ hep - th ] ] .
m. chaichian , m. oksanen and a. tureanu , _ `` hamiltonian analysis of non - projectable modified f(r ) horava - lifshitz gravity , '' _ phys .
b * 693 * ( 2010 ) 404 [ phys . lett .
b * 713 * ( 2012 ) 514 ] [ arxiv:1006.3235 [ hep - th ] ] .
s. weinberg , _ gravitation and cosmology _ , john wiley & sons , new york , ( 1972 ) . v. f. mukhanov , h. a. feldman , r. h. brandenberger , _ theory of cosmological perturbations_,*215 * , 203 , ( 1992 ) .
p. a. r. ade _ et al .
_ [ planck collaboration ] , _ `` planck 2015 results .
dark energy and modified gravity , '' _
arxiv:1502.01590 [ astro-ph.co ] . e. di valentino , a. melchiorri , j. silk , _ `` cosmological hints of modified gravity ? '' _ arxiv:1509.07501 [ astro-ph.co ] . | we investigate the metric perturbations of the restricted @xmath0 theory of gravity in the cosmological context and explore the phenomenological implications of this model .
we show that it is possible to construct a restricted model of gravity , in which the background equations are the same as the equations of motion which are derived from the einstein - hilbert action with the cosmological constant term .
we argue that the deviation from the einstein - hilbert model emerges in the perturbed equations , for which we have a non - vanishing anisotropic stress .
further , with the help of the results of planck data for the modified gravity we obtain constraints on the parameters of the model .
0.4 cm * cosmological perturbations in restricted * @xmath0-gravity m. chaichian,@xmath1 a. ghalee,@xmath2 j. kluso,@xmath3 [ multiblock footnote omitted ] + @xmath1__department of physics , university of helsinki , p.o .
box 64 , + fi-00014 helsinki , finland _ _
+ @xmath2__department of physics , tafresh university , tafresh 39518 79611 , iran _ _ + @xmath4__department of theoretical physics and astrophysics , faculty of science , + masaryk university , kotlsk 2 , 611 37 , brno , czech republic _
_ 0.8 cm # 1 # 1x^(#1 ) # 1\{#1 } |
scaling to large numbers of quantum bits ( qubits ) is important for the development of a future quantum information processor .
several pathways toward this goal have been proposed based on using trapped atomic ions as qubits @xcite .
one particular approach envisions processing , transporting , and sympathetic cooling of ions inside a large array of traps @xcite .
recent experiments with various trap architectures have demonstrated the key ingredients for scalable ion loading , trapping , processing , and transport @xcite .
current implementations of quantum information processors require a reduction of gate errors to be efficiently scalable @xcite , necessitating stringent qubit control . in most trapped - ion quantum information experiments ,
control is accomplished via laser - based techniques @xcite , but other approaches are being investigated , e.g. , those based on magnetic fields @xcite .
these methods reduce the required laser overhead and may enable a higher level of control and integration than laser - based control .
while the highest entangled - state fidelity is currently obtained with laser - based operations @xcite , magnetic - field based control may eventually provide sufficiently low errors for fault tolerant operations @xcite .
for single - qubit operations performed with microwave magnetic near fields , errors per computational gate of @xmath3 and better have been reported @xcite .
although the theoretical threshold for fault tolerance depends strongly on the assumptions of the model under study , an operational error of 10@xmath4 or below may be a reasonable starting point in a future fault - tolerant architecture .
reaching this level of precision for two - qubit operations requires improvements on the state - of - the - art errors by approximately two orders of magnitude . as with conventional computers , microfabrication may be advantageous in realizing a future quantum processor , such as those based on surface - electrode ion traps @xcite ( for a review see @xcite ) .
in addition to reducing the overall volume of a processor for a given number of qubits , a reduction in scale could allow for tighter confinement with trap frequencies scaling approximately as @xmath5 if all other parameters are held constant . here
@xmath6 denotes a characteristic length scale of the trap , often characterized by the smallest ion - electrode distance .
higher trap frequencies can enable shorter durations for transport , separation and recombination , and two - qubit operations mediated by ion motion . in the last two decades , typical trap length scales have been reduced from centimeters to tens of micrometers .
this increases the oscillating magnetic near - fields ( scaling as @xmath7 ) , originating from oscillating currents running in one or multiple trap electrodes .
fast and precise single - qubit rotations ( carrier transitions ) have been demonstrated in this way @xcite .
in addition , the corresponding field gradients ( proportional to @xmath5 ) can become large enough to couple the ions spin with their motional state ( sideband transitions ) on experimentally accessible timescales @xcite . for trapped ions close to their motional ground states ,
the ratio of sideband - transition rates to carrier - transition rates driven by oscillating near fields and their corresponding field gradients is on the order of @xmath8 , where @xmath9 is the harmonic - oscillator ground - state extent @xcite . for typical experimental parameters
, @xmath9 is on the order of @xmath10 nm , and therefore @xmath8 ranges between @xmath11 and @xmath12 for the smallest traps today .
further reduction in trap size would increase this ratio , enabling faster spin - motional entanglement .
for the same reason , near - field gradients are typically not useful for larger ( @xmath13 m ) trap structures .
the main practical limitation to size reduction has been `` anomalous '' motional heating , caused by electric - field noise at the positions of the ions in excess of johnson noise from the resistances of the electrodes and of their filters @xcite .
anomalous noise fields have been phenomenologically modeled as originating from potential fluctuations of surface patches or dipoles with typical dimensions much smaller than @xmath6 , which yields a noise spectral density that scales as @xmath14 @xcite .
as @xmath6 is decreased , the resulting heating grows faster than trap frequencies , oscillating near fields , and their corresponding gradients .
consequently , this form of heating appears to prevent scaling to even smaller trap dimensions .
however , recent studies have demonstrated a substantial reduction of anomalous heating by cooling the electrodes @xcite and by treatment of the electrode surfaces @xcite .
such methods may enable the further reduction in trap size . in the experiments described here ,
we investigate building blocks for oscillating magnetic near - field quantum control of trapped ions @xcite ; a similar approach is pursued in @xcite . to test the basic scheme , we integrated three microwave current - carrying electrodes into a room - temperature surface - electrode trap with @xmath15 m . frequencies of the microwave currents are tuned to near resonance with hyperfine transitions in @xmath0mg@xmath1 ( nuclear spin @xmath16 ) .
the resulting magnetic field at the position of the ions determine the carrier coupling , while the magnetic - field gradient couples the hyperfine levels ( spin states ) to the motional degrees of freedom @xcite .
we previously reported on fast - carrier transitions and generation of two - ion entanglement @xcite ; the entanglement was produced with a two - qubit gate operation @xcite suitable for scalable quantum information processing . in this report , we give a more detailed account of our experimental setup and the procedures for calibrating and optimizing the operation of the system . in sec .
[ sec : trap ] , we discuss design , fabrication , and performance of the surface - electrode trap and describe the laser beams for loading , state preparation , and detection .
section [ sec : mwsetup ] describes the microwave system that generates and shapes the signals used for coherent single - qubit and two - qubit operations . in sec .
[ sec : procedure ] , we give an experimental sequence for calibrating and studying entangling - gate operations . the experimental procedure to adjust the microwave field at the ion s position for driving sideband transitions ,
a prerequisite for entangling gates @xcite , is described in sec .
[ sec : oscb ] .
section [ sec : mm ] details a technique for rf micromotion detection by observing transitions in microwave magnetic - field gradients . in sec .
[ sec : radialmode ] , we describe a procedure to optimize the orientation of the radial - motional modes relative to the microwave field gradient components and a static homogeneous quantization field . finally , in sec .
[ sec : limits ] we discuss the factors currently limiting two - qubit gate fidelity @xcite .
the surface - electrode trap is designed as an asymmetric `` five - wire '' trap @xcite .
it consists of two radio - frequency ( rf ) , six control , and three microwave electrodes ; see fig .
[ fig : chip ] for a view of the trap assembly ( fig .
[ fig : chip]a , b ) and the central region of the trap electrodes ( fig .
[ fig : chip]c ) .
mt , located in the @xmath17-@xmath18 plane , is shown .
the @xmath0mg@xmath1 ions are trapped at a distance @xmath19 m above the surface , at the position labeled trap center . ]
the trap chip is fabricated in the nist boulder cleanroom .
we choose aluminum nitride ( aln ) as a substrate for its high thermal conductivity of approximately @xmath20 w/(m@xmath21k ) , which aids dissipation of heat generated by the microwave currents in the trap electrodes .
adhesion ( ti , 10 nm ) and seed ( cu , 60 nm ) layers are deposited on a @xmath22 mm diameter aln wafer by electron beam evaporation . a 14 @xmath23 m thick layer of photoresist ( shipley spr220 - 7 ) is spin - coated onto the wafer , exposed through a patterned chromium mask by contact lithography , and developed .
the photoresist pattern defines @xmath24 m wide gaps between electrodes in the central region of the trap ; the gap width is increased to @xmath25 m at distances @xmath26 mm away from the trap center .
gold is then electroplated , in a 140 ml sulfite gold solution ( transene tsg-250 @xcite ) , up to a @xmath27 m thickness onto individual approximately @xmath28 mm dies .
the use of a small electroplating volume allows us to replace the solution for each electroplating process , thus reducing the possibility of contamination . after removing the photoresist with acetone ,
the seed and adhesion layers are wet etched , and a total of nine ( approximately @xmath29 mm ) trap chips are obtained by dicing . to provide good thermal contact ,
the chip is bonded with a thin layer of silver - filled vacuum compatible epoxy ( epo - tek@xmath30 h21d @xcite ) to a solid copper support ( fig .
[ fig : chip]a ) .
this support structure also holds a printed circuit board for in - vacuum filtering of the control potentials .
we use low - pass rc filters with @xmath31k@xmath32 ( anaren , part no . : h2b15081001f2lo @xcite ) and @xmath33pf ( novacap , part no . : 0504n821j101p
@xcite ) . in addition , the printed circuit board is used for connecting both rf electrodes to a resonant quarter - wave step - up transformer ( @xmath34 , when loaded with the trap ) @xcite .
connections between all trap chip electrodes and the filter board are made with gap - welded gold ribbons with a @xmath35m@xmath36 m cross section and length @xmath37 mm .
each microwave electrode is connected to a microstrip line on the filter board , which is soldered to a sma jack on the input end .
the other end of each microwave electrode is shorted to ground at the edge of the chip via gap - welded gold ribbons . in - vacuum
coaxial cables connect the vacuum feedthroughs with the sma connectors .
all of the assembly steps are performed in a cleanroom environment to minimize the presence of contamination , e.g. , dust particles on the surface of the trap chip , which can generate unwanted conduction paths and stray electric fields , possibly enhanced when exposed to uv laser beams .
an rf peak voltage @xmath38 v to @xmath39 v at @xmath40 mhz is applied to both rf electrodes , providing the @xmath41-@xmath18 radial confinement of the ions at a distance @xmath19 m above the surface .
the rf frequency is stabilized to the center of the quarter - wave resonance of the resonator with a feedback loop to the frequency modulation input port of the rf generator .
the six control electrodes ( c1 to c6 ) are biased between @xmath42 v and @xmath43 v to provide confinement along the trap axis .
the potentials are applied by six digital - to - analog converters ( dacs ; for details see sec .
[ sec : mwsetup ] ) .
single - ion motional frequencies range between @xmath44 mhz and @xmath45 mhz in the axial direction , while two non - degenerate frequencies between @xmath46 mhz and @xmath47 mhz aligned along adjustable directions ( sec .
[ sec : radialmode ] ) in the radial @xmath41-@xmath18 plane are used .
we calculate the trapping potential @xmath48 in the radial and axial directions by means of the gapless plane approximation @xcite ; @xmath48 is generated by combining the rf pseudopotential and the control potential .
( in units of mev ) as a function of distance from the trap center ( rf null and axial minimum ) , which is located @xmath49 m above the surface
. applied potentials are given in the text .
( a ) potential in radial plane at @xmath50 . the dot marks the trap center .
the vectors indicate the calculated motional mode orientation . here
@xmath51 denotes the angle between the lf - radial mode vector and the @xmath41 axis .
( b ) potential in the axial direction at @xmath52 .
the contribution from the axial rf pseudopotential ( see text ) is scaled by a factor of ten for visibility . ] with @xmath53 v and @xmath54 v , fig .
[ fig : trappot]a shows the calculated @xmath55 in the radial plane around the trap center at @xmath56 , while fig .
[ fig : trappot]b illustrates @xmath57 in axial direction .
the overall trap depth , limited by a saddle point in the @xmath41-@xmath18 plane ( fig .
[ fig : trappot]a ) , is @xmath58 mev for this choice of potentials .
we denote the two radial center - of - mass modes by their relative frequencies : low - frequency ( lf ) and high - frequency ( hf ) mode ; their orientations are indicated by arrows in fig .
[ fig : trappot]a .
the orientation of these modes is determined by the control potentials while the rf pseudopotential is cylindrically - symmetric in a small volume around the trap center . due to the tapered geometry of the rf electrodes
, we find a weak axial rf pseudopotential , with a curvature corresponding to a motional frequency @xmath59 khz .
the axial confinement , however , is dominated by the applied control potentials and its frequency is simulated to be @xmath60 mhz
. additionally , we notice the presence of significant stray potentials , which must be taken into account in the simulations to reproduce the observed motional - mode directions and frequencies ( sec .
[ sec : radialmode ] ) .
we employ a set of `` shim '' fields , designed through simulations to produce orthogonal displacements of the ion(s ) along @xmath41 , @xmath17 , and @xmath18 while approximately preserving the control - field curvatures , for micromotion nulling ( sec .
[ sec : mm ] ) and to obtain oscillating magnetic - field maps ( sec .
[ sec : oscb ] ) . due to the limited dac resolution
[ sec : mwsetup ] ) , the ion position can be controlled to a precision of @xmath61 nm for typical trap frequencies @xmath62 mhz and @xmath63 mhz .
a static magnetic field of magnitude @xmath64 mt is aligned parallel to the trap surface at an angle of @xmath65 with respect to the @xmath18 axis ( fig .
[ fig : chip]c ) and provides the ions internal - state quantization axis . at this field strength ,
the ground state @xmath66s@xmath67 @xmath68 to @xmath69 hyperfine transition frequency @xmath70 ghz is first - order insensitive to magnetic field changes , while the quadratic frequency deviation is @xmath71 khz/(mt)@xmath66 ; we use the @xmath72 and @xmath73 states as the qubit . here
, @xmath74 is the total angular momentum and @xmath75 is the projection of the angular momentum along the magnetic field axis .
figure [ fig : level ] illustrates the ground - state hyperfine structure , highlighting the qubit states .
mg@xmath1 ( nuclear spin @xmath16 ) at @xmath64 mt . the microwave driven hyperfine transitions used to prepare , detect , and control the qubit states are depicted as arrows . the cooling and detection laser beam is indicated as a ( blue ) wavy arrow .
( inset ) illustration of the qubit states ( bold lines ) and their radial motional states ( gray ) .
the `` red '' and `` blue '' sideband transitions as well as the qubit transition are indicated with arrows . ]
we track @xmath76-field drifts every 20 to 30 minutes with a ramsey experiment on the @xmath77 to @xmath78 transition ( excited with microwave fields ) , which has a magnetic - field sensitivity of @xmath79 mhz / mt and adjust a computer - controlled correction coil current accordingly .
we observe day - to - day @xmath76-field variations of less than @xmath80 mt .
two overlapping @xmath81-polarized laser beams for doppler cooling on the @xmath66s@xmath82p@xmath83 cycling transition propagate parallel to the magnetic field .
a first beam ( bdd ) , approximately @xmath84 mhz detuned from this transition , performs initial doppler cooling and optical pumping .
a second beam ( bd ) is red detuned from this transition by approximately half the natural line width ( @xmath85 mhz ) and is used for final doppler cooling and state preparation into the @xmath77 level of the @xmath66s@xmath67 ground state state for preparation instead of the @xmath86 state used in @xcite . ] .
state preparation and cooling takes @xmath87 ms ; during the final @xmath35s , only the bd beam is applied . for state detection , the bd beam
is tuned to resonance for approximately @xmath88s , discriminating the @xmath77 state from the other hyperfine ground states through resonance fluorescence .
the fluorescence photons are detected by a photon - multiplier tube ( pmt ) detector . in most experiments ,
the @xmath77 population is transferred via two microwave preparation pulses to @xmath73 to initialize the qubit . to detect the @xmath73 state ,
two pulses are used to transfer the @xmath73 population to the @xmath77 ( bright ) state , while the @xmath72 population is `` shelved '' into the @xmath89 ( dark ) state ( fig .
[ fig : level ] ) . during the detection period ,
the pmt registers @xmath90 counts for the bright state and @xmath91 counts for the dark state on average per ion .
stray light accounts for @xmath92 counts on average .
the bdd and bd beams are both focused into the vacuum chamber with a 10-cm focal length lens , and their powers are adjusted to be @xmath93w and @xmath94w , respectively ; both have waists half width ] of @xmath95 m at the position of the ions . to create ions , a third continuous - wave ( cw ) laser beam ( denoted pi : @xmath96 nm , power @xmath97 mw , and beam waist @xmath95 m )
is directed antiparallel to the cooling beams and is used to photoionize neutral mg atoms inside the trapping region .
an oven tube , holding enriched @xmath0 mg , is mounted approximately @xmath98 cm away from the trap center . after allowing the oven to heat up for 30 s ,
the pi is applied for a few seconds until ion(s ) are loaded .
the oven current is adjusted to a loading rate of less than one ion per second , so the pi beam can be turned off manually after the desired number of ions are detected on a ccd camera . at a background pressure of @xmath99
pa , the storage period of a single ion is observed to be up to a few hours , with doppler cooling applied periodically every one or two milliseconds .
typical trapping lifetimes for two ions are on the order of 30 min .
to implement microwave control of the ion(s ) spin and motional states , we built a system assembled mostly from commercially available discrete microwave components ( fig .
[ fig : mwsetup ] ) .
( bsb and rsb ) around @xmath100 ghz , within a range of about @xmath101 mhz .
it also enables a signal between @xmath102 ghz and @xmath103 ghz applied to mw2 for hyperfine state control and single - qubit operations .
the gray components are computer controlled . ] for hyperfine state manipulation and single - qubit operations , magnetic fields from microwave currents in electrode mw2 drive transitions at frequencies between @xmath102 ghz to @xmath103 ghz , spanning all hyperfine transitions in the @xmath66s@xmath67 manifold .
as noted above , we use a series of two transfer pulses ( carrier @xmath104-pulses ) to transfer population from @xmath77 to @xmath105 via @xmath78 , to initialize the qubit state .
we have demonstrated @xmath104-times for qubit spin flips as short as @xmath106 ns @xcite ; however , here we adjust the power of the microwave signals to levels corresponding to @xmath104-times on the order of a few microseconds , limited by the timing resolution of our data acquisition system ( see below ) . to drive `` blue '' ( bsb ) and `` red '' ( rsb ) sidebands of the qubit transition ( fig .
[ fig : level ] ) , we apply currents simultaneously to mw1 , mw2 , and mw3 at @xmath107 or @xmath108 . the amplitudes and phases of the three signals
are adjusted to minimize the oscillating magnetic field at the position of the ion(s ) , while maximizing the field gradient ( sec .
[ sec : oscb ] ) .
this control is enabled by the six phase shifters and six attenuators .
the microwave setup is integrated with a data acquisition system , controlled by a field - programmable gate array ( fpga ) with a timing resolution of @xmath109 ns @xcite .
the entire system comprises 16 digital ( ttl ) output channels , 24 dac channels , eight direct - digital synthesizer ( dds ) modules , and two digital ( ttl ) inputs for registering pmt counts .
each dds can generate signals at frequencies up to @xmath110 mhz with a frequency resolution of @xmath111 hz and phase resolution of @xmath112 .
the dacs have an update rate of @xmath113 khz and can be set between @xmath114 v and @xmath43 v with a resolution of @xmath115 mv .
we use six of these channels to apply potentials to the trap control electrodes and 14 channels to control microwave components .
in addition , a fast , two - channel dac is used to generate arbitrary waveforms with a @xmath116 mhz update rate , a voltage range from @xmath114 v to @xmath10 v , and a resolution of @xmath115 mv .
this dac is programmed via usb and triggered by the data acquisition fpga .
we use three dds modules as sources for three frequency octupling modules , each consisting of three doubling stages in series with amplifiers and filters .
two of these generate the rsb and bsb signals , which pass through individual pulse shaping and control stages ( phase and amplitude control ) .
the pulse envelopes for the bsb and rsb signals are produced with a multiplier ( analog devices adl5391
@xcite ) used to shape pulse amplitudes and controlled by one channel of the fast dac .
in addition , a voltage - controlled attenuator adjusts the overall power of a pulse . after this pulse - shaping stage the signals are split and directed to three individual voltage - controlled phase shifters ( @xmath117/v ) and attenuators ( @xmath118 db / v ) , to control the power and phase of the signals delivered to the individual microwave electrodes . before the rsb and bsb signals
are combined onto the inputs of each microwave electrode , they are amplified by high - power amplifiers ( mini - circuits zhl-30w-252-s+
following three high - power combiners , three coaxial cables connect the microwave setup with the vacuum feedthroughs .
the third octupling module generates frequencies to drive the hyperfine carrier transitions , and its signal is combined with one of the sideband lines attached to electrode mw2 .
its power is adjusted with fixed attenuators to yield carrier @xmath104-pulse durations of a few microseconds . on a daily basis , we calibrate the frequencies and @xmath104-pulse durations for the transfer and single - qubit pulses . from day to day , we observe fractional variations of @xmath119 for the pulse durations .
due to our quantization field calibrations and adjustments , the hyperfine - transition frequency values stay constant within a statistical uncertainty of approximately @xmath120 khz .
we use a single ion to calibrate experiments and set parameters for the two - qubit gate . after an ion
is loaded into the trap , micromotion along the projection of the detection laser beam in the radial plane ( @xmath18 direction ) is minimized by measuring the variation in laser fluorescence from the bd beam @xcite as a function of the `` horizontal''-shim field ( fig .
[ fig : rffield ] ) . since the trap rf quadrupole is tilted in the radial plane by @xmath121 from the @xmath41 axis ( fig .
[ fig : rffield ] ) , the horizontal - shim field is chosen to push along a vector rotated by @xmath122 from the @xmath18 axis .
a push along this axis will result in a maximal variation of the horizontal micromotion amplitude . .
the arrows indicate the horizontal ( red ) and the vertical ( gray ) micromotion component .
only the horizontal component can be detected by the detection laser beam .
the red and gray lines indicate the two shim directions used for the compensation of horizontal and vertical micromotion . ]
we verify this compensation field with a nulling scheme using microwave magnetic - field gradients ( described in sec .
[ sec : mm ] ) .
this technique can also be used to minimize the micromotion in the vertical ( @xmath17 ) direction .
we find that the trapping fields are perturbed by stray fields , with corresponding potential @xmath123 .
the exact origin of these fields is unknown , but they are consistent with a contact potential patch , and can be compensated for as described in appendix [ sec : trapperform ] . following micromotion nulling , we calibrate the microwave magnetic fields used for driving the radial sidebands of the qubit transition ( sec .
[ sec : oscb ] ) . from these experiments
, we extract the orientation and the strength of the microwave magnetic - field gradient . as a last calibration step to set up the two - qubit gate experiment
, we align the radial - mode vectors relative to the microwave magnetic field by rotating the modes with an additional control field ( sec .
[ sec : radialmode ] ) . to implement motional sideband transitions at frequencies @xmath124 ,
a gradient of the microwave magnetic field is necessary .
however , any oscillating field at the ion(s ) position will cause off - resonant carrier transitions and ac zeeman shifts @xcite , both of which will inhibit precise control .
thus , the magnetic field amplitude should be suppressed as much as possible at the position of the ion .
we realize this experimentally by adjusting the phases and amplitudes of the three microwave currents to minimize ac zeeman shifts detected by the ion . the experimental sequence is shown in fig .
[ fig : gradmap]a . and @xmath125 .
the spin echo @xmath104-pulse makes the result insensitive to magnetic - field fluctuations slower than the time scale of an individual experiment ( typically @xmath126s ) .
we observe minimal phase pickup when the oscillating magnetic field is minimized at the position of the ion .
( b ) extension to measure the ac zeeman shift as a function of displacement @xmath127 ( see text ) .
( c ) experimental data mapping out the ac zeeman shift in the radial @xmath41-@xmath18 plane by use of the superposition state of @xmath128 and @xmath129 .
the fitted isolines are obtained from the quadrupole oscillating magnetic field component @xmath130 along @xmath131 through its resulting ac zeeman shift @xmath132 , which is proportional to @xmath133 .
the angle @xmath134 is the rotation angle of the microwave quadrupole field , but the directions of minimum and maximum values of @xmath135 are along angles @xmath136 and @xmath137 ( fig . [
fig : mwfields]a ) . ]
the nulling procedure relies on the measurement of ac zeeman shifts on superpositions of selected hyperfine states @xmath138 and @xmath125 . in a coordinate system with unit vectors
@xmath139 , @xmath140 and @xmath141 , the magnetic - field components oscillating at @xmath142 can be written as : @xmath143 where @xmath130 and @xmath144 are the strengths of the oscillating magnetic - field components parallel and perpendicular to the quantization field @xmath145 . here @xmath146 is
the rotation angle of the perpendicular magnetic - field component polarization ellipse with the @xmath147 axis and @xmath148 is a phase . for @xmath149 , the oscillating magnetic field in the plane
perpendicular to @xmath145 is linearly polarized . for @xmath150 ,
the perpendicular field component has pure circular polarization .
the @xmath130 field couples levels with @xmath151 , whereas the @xmath144 field drives transitions with @xmath152 .
these fields induce an ac zeeman shift @xmath153 on the @xmath154 transition , where the coefficients may depend on @xmath155 and @xmath156 .
note that we implicitly include the bloch - siegert shifts ( energy shift resulting from the counter - rotating part of the field ) in the calculation of the coefficients in @xmath132 .
first consider @xmath157 and @xmath158 .
a @xmath144 component applied near @xmath159 is red detuned from the @xmath160 and @xmath161 transitions , thereby increasing the energy of @xmath157 and decreasing the energy of @xmath158 .
from the clebsch - gordan coefficients for these transitions , we determine @xmath162 at @xmath163 .
a @xmath130 component near @xmath159 is red detuned from the @xmath164 transition and also increases the @xmath165 transition frequency : @xmath166 at @xmath163 .
both @xmath167 and @xmath168 are positive , of similar magnitude for all @xmath156 and depend only weakly on @xmath142 around @xmath159 within the useful range of motional sideband frequencies , so this choice of @xmath138 and @xmath125 can be used to minimize both @xmath144 and @xmath130 .
another useful choice is @xmath157 and @xmath169 . here ,
the dominant ac zeeman shift results from @xmath130 coupling to the @xmath170 transition : @xmath171 , and @xmath172 at @xmath173 . this transition can be used to minimize @xmath130 even further due to its large @xmath168 value , by use of currents in electrodes mw2 and mw3 ( those electrodes contribute dominantly to @xmath130 due to their geometry ) .
we produce the data of fig .
[ fig : gradmap]c using this transition and an experimental sequence presented in fig .
[ fig : gradmap]b . with @xmath174 at position @xmath175 in the coordinate system of fig .
[ fig : chip]c [ not the frame of eq .
( [ eq : oscbfieldwrb ] ) ] we expect to first order @xmath176 here @xmath177 and @xmath134 characterize the strength and orientation of the quadrupole for small displacements
@xmath178 m from the field null point . by fitting the ac zeeman shift resulting from the projection @xmath130 of @xmath179 onto @xmath145 to the experimental data of fig .
[ fig : gradmap]c , we extract @xmath180 t / m and @xmath181 . from this fit
, we determine a small shift of the center of the magnetic quadrupole with respect to the trap center by @xmath182 m , limited by the sensitivity of our nulling procedure .
we detect phase accumulations over a time @xmath183s ; we can not measure phase shifts smaller than @xmath184 during that time .
this limits the ac zeeman shift sensitivity to @xmath91 khz and leads to finite fields at the trap center .
the experimental day - to - day variations for the nulled configuration are @xmath185 db for the relative powers of mw2 and mw3 with respect to mw1 , and @xmath186 for the relative phases of mw1 and mw3 with respect to mw2 .
numerical simulations of the microwave currents in the three electrodes by use of ansys hfss software @xcite provide a value of @xmath187 t / m for a current amplitude of @xmath188 a in conductor mw1 , while the amplitudes and phases of currents in mw2 and mw3 are adjusted to give a nulled field configuration at the position of the trap rf null .
we find the following current amplitude ratios @xmath189 and phase shifts @xmath190 in mw2 and mw3 relative to the current in mw1 for the nulled configuration : @xmath191 and @xmath192 .
a comparison with the experimental current ratios is difficult because of slightly varying attenuations in the signal paths leading to the three microwave electrodes .
however , from simulations , we extract an angle @xmath193 , close to the value fitted to the experimental data ( also sec .
[ sec : mm ] ) . a similar level of agreement between simulation and experiment is found in @xcite . in our experiments ,
current densities are simulated to be as high as @xmath194 near the edges of the conductor mw1 , with a total current of @xmath195 a in mw1 , for the above stated experimental settings , producing a gradient of @xmath196 t / m .
we estimate the corresponding power dissipation in all electrodes due to all currents from the bulk resistance of gold ( @xmath197 m ) in the central region of the chip ( @xmath198 m along @xmath17 ) to be @xmath199 mw .
currents in microwave electrodes induce currents in neighboring electrodes , here illustrated for rf1 and c2 . ] in fig .
[ fig : mwdensity ] the current density distribution in electrode mw1 and the induced - current densities in the neighboring electrodes rf1 and c2 are shown .
the simulated magnetic and electric microwave fields for the nulled configuration are plotted in fig .
[ fig : mwfields ] .
the influence of the microwave electric fields on the trapping potential is discussed in appendix [ sec : mwtrap ] . ) . the estimated @xmath200 nm displacement for typical trap parameters is shown in gray , scaled by a factor @xmath201 for visibility . ] in the presence of excess rf micromotion , the microwave magnetic near - field gradient can also drive micromotion - induced sidebands on carrier transitions @xcite .
this effect can be used to null micromotion along any direction in the radial @xmath41-@xmath18 plane ( refer to the frame of fig . [
fig : chip]b ) .
therefore , it may have an advantage over certain optical schemes that are sensitive to micromotion only along the direction of a laser beam ( or along the direction of the effective @xmath202-vector for stimulated - raman transitions ) @xcite .
for example , in a surface - electrode trap , the direction perpendicular to the surface can be difficult to probe with laser beams , because scattered light from the trap surface may interfere with detection and/or cause electrode charging . the micromotion can be written as @xmath203 . in the presence of a microwave magnetic field gradient of frequency @xmath124 and in the coordinate frame of fig .
[ fig : chip]c , the ion in its rest frame experiences fields with frequencies @xmath204 and amplitude @xmath205 , cf
( [ eq : oscbfield ] ) : @xmath206 for @xmath207 , where @xmath208 is the frequency of a transition between states in the hyperfine manifold , the ion experiences an oscillating field at frequency @xmath208 .
the resulting rabi rate is proportional to @xmath209 and therefore to @xmath210 , the micromotion amplitude . to minimize the excess micromotion , we move the ion in the radial plane ( along the shim directions illustrated in fig .
[ fig : rffield ] ) to a position that minimizes the rabi rate of the micromotion sidebands .
the oscillating field @xmath211 has projections @xmath130 and @xmath144 on @xmath212 .
therefore , we can sense micromotion along any direction in the @xmath213 plane , since it is always detectable by either a @xmath214 or @xmath215 transition ( or both ) . in the experiment
, we use the @xmath216 and @xmath217 transitions , with a magnetic - dipole matrix element @xmath218 and @xmath219 , respectively , where @xmath220 denotes the bohr magneton . to estimate the sensitivity of this nulling method , we assume @xmath221 t / m and a transition matrix element for @xmath208 of @xmath222 . to further simplify the estimate , the rotation angles of the microwave - magnetic quadrupole ( @xmath223 , fig .
[ fig : mwfields]a ) and rf - electric quadrupole ( @xmath121 , fig .
[ fig : rffield ] ) are assumed to be oriented such that in the rest frame of the ion , we achieve a maximum oscillating field component that drives the transition ( @xmath224 ) .
for the given misalignment , this approximation is good to within @xmath225 for the estimated sensitivity .
note that in the experiments both components of the microwave - field gradient can be used , and the misalignment does not limit the sensitivity of this method .
the rabi rate for the micromotion sideband - induced carrier transition is then given by @xcite : @xmath226 for our parameters , and assuming a @xmath184 rotation detection sensitivity on the micromotion sideband during a drive time @xmath227ms , a micromotion amplitude of @xmath228 nm can be detected .
in addition to detecting micromotion , we use this method to probe characteristics of the microwave - magnetic and rf - electric field by measuring the micromotion sideband rabi rate as a function of position .
figure [ fig : mmmap ] shows the rabi rate on the red micromotion sideband of the @xmath216 transition versus ion position . transition as a function of ion position .
the isolines result from a model fit to the data , from which @xmath229 and @xmath230 are extracted from eq .
( [ ref : mmfield ] ) . ] using these data together with the calculated shape of the rf quadrupole , we extract the parameters @xmath231 and @xmath134 of the microwave quadrupole field , finding @xmath229 and @xmath230 .
this map was taken with a different realization of the trap described in sec .
[ sec : oscb ] but of the same design , so imperfections in construction and small variations in the electrode impedance may have affected the measurement .
differences in impedance affect the strength of the gradient , so the agreement in @xmath231 values is probably fortuitous .
however , we find again a value for @xmath134 that is close to the simulated value @xmath232 ( sec . [
sec : oscb ] ) . after determining the characteristics of the microwave quadrupole
, we align the orientation of the radial motional modes to the microwave field gradient to optimize sideband rabi rates .
since the qubit is controlled with a @xmath233 transition , its motional sidebands are driven by @xmath234 ( fig .
[ fig : mwfields]a ) . to overlap the hf - mode vector with @xmath235
, we use an additional control potential @xmath236 with potentials @xmath237 v applied to the control electrodes , which leads to a mode orientation @xmath238 in the absence of any additional control potential . in the experiments , the total potential @xmath239 breaks the symmetry of the rf pseudopotential , and sets the orientation of the radial modes . the value of the scaling coefficient @xmath240 is used to adjust the orientation of the radial modes . for the given total potential @xmath241 including the stray potential ( appendix [ sec : trapperform ] ) , negative
@xmath240 rotates the lf - mode vector toward @xmath242 and for positive values it rotates toward @xmath243 .
furthermore , @xmath236 is designed such that the ion position remains unchanged while all trap frequencies change as a function of the scaling .
a comparison of the simulated frequency changes with the actual experimental data can be used to determine and adjust the orientation of the radial modes . in fig .
[ fig : radialcontrol]a such a comparison between simulated and measured mode frequencies shows good agreement . extracted from the simulation . at an angle @xmath244 the high - frequency mode is aligned along @xmath245 and couples maximally to the microwave - field gradient . ]
figure [ fig : radialcontrol]b illustrates the corresponding mode - angle values extracted from the simulation .
note that for the simulation we had to include @xmath246 to find agreement with the experimental data , details are given in appendix [ sec : trapperform ] .
when choosing @xmath244 , we observe the maximal coupling of the microwave - field gradient to the hf mode , while the lf mode couples very weakly to the microwave field .
we find a corresponding rabi rate for driving the hf sideband transition of the motional ground state of @xmath247 khz , with the microwave - field gradient adjusted to @xmath196 t / m .
internal - to - motional state coupling as well as an entangling two - qubit gate based on the simultaneous application of detuned red and blue sideband drives @xcite was presented in @xcite . using this gate , entangled states with a fidelity of approximately @xmath248 were produced . in these experiments we used an out - of - phase radial mode of two ions , because it is less sensitive to electric - field noise than the corresponding center - of - mass mode .
the out - of - phase motional heating was measured to be @xmath249 quanta / ms , while the center - of - mass motional heating rate was approximately @xmath250 quanta / ms ( for radial mode frequencies around @xmath251 mhz ) .
heating corresponding to an increase of a single quantum of motion on the mode used for an entangling - gate operation effectively decoheres the gate operation @xcite . since in @xcite
the entangling gate durations were approximately @xmath252s , even the out - of - phase mode heating can cause substantial gate errors . to investigate this effect for our experimental conditions , we numerically solve the master equation given in @xcite .
we determine that this effect alone limits the fidelity to approximately @xmath253 .
furthermore , we observed drifts of the motional sideband frequencies in the experiments , which may be caused by mode frequency drifts and/or varying ac zeeman shifts from drifts of the residual magnetic and electric microwave fields .
we estimate the frequency variation during the gate pulse to be less than @xmath254 khz from these effects , while a typical detuning for the gate is around 4.5 khz .
this systematic effect can contribute an additional @xmath255 loss of fidelity .
in summary , we have given details of the surface - electrode trap and microwave electronics used for the microwave near - field quantum control experiments described in @xcite .
we have described the experimental procedures used to control , shape and calibrate the microwave fields that implement motional - sideband transitions using the near - field gradients .
we have introduced a novel technique for rf micromotion detection .
furthermore , we presented and experimentally verified a method to rotate the radial mode axes . finally , we investigated limitations in fidelity of the two - qubit gates in @xcite through simulations , and examined the impact of leading error sources .
we determined that the motional heating during the gate pulse is the dominant source of infidelity .
a significant reduction ( by a factor of 100 or more ) of the heating rate should lead to entangled state fidelities higher than @xmath256 .
a reduction of ambient heating may be achieved by surface treatments of the electrode structure as described in @xcite and/or cryogenic cooling of the electrodes @xcite .
in addition , higher microwave currents can increase the gate speed and thereby can reduce the effect of motional heating .
the corresponding increased heat load may be mitigated by reducing electrode resistance through low temperature operation .
we thank m. j. biercuk , j. j. bollinger , and a. p. vandevender for experimental assistance , and d. h. slichter and r. jrdens for helpful comments on the manuscript .
we also thank d. hanneke and j. p. home for useful discussions and p. treutlein for advice on microfabrication techniques .
we received valuable assistance with microwave current simulations from d.t.c .
allcock and j. schbel .
this work was supported by iarpa , aro contract # eao139840 , onr , darpa , sandia national laboratories and the nist quantum information program .
contribution of nist , not subject to u. s. copyright .
to compare the trapping potential calculations from sec .
[ sec : trap ] to the experiment , we first determine the electric field strengths at the position of the ion to compensate the rf micromotion ( sec . [
sec : procedure ] ) .
we then measure the mode frequencies , and find the orientation of the radial mode vectors ( sec .
[ sec : radialmode ] ) , characterized by @xmath51 ( fig .
[ fig : trappot ] ) .
we find a stray field @xmath257 v / m and a stray field curvature tensor , given by the hessian matrix of @xmath246 at the center of the trap , @xmath258 after including @xmath246 into the potential simulations , we find reasonably good agreement ( motional frequencies and mode angles agree to within @xmath259 or better ) between the simulations and the measurements ( sec .
[ sec : radialmode ] ) for several different settings of rf voltage and control potentials .
the stray potential might be caused by work function variations on the electrode surface @xcite ( e.g. , from varying mg contamination ) or stray charges localized in the central region of the trap chip .
one way of modeling modeling the stray effects is a rectangular patch with size @xmath260 m and @xmath261 m located on the chip surface , whose center is displaced by @xmath262 m , @xmath263 m relative to the trap center and that is biased to @xmath264 v ( typical work function differences can exceed one volt ) .
we observe temporal variations in the stray potential effects , perhaps due to uncontrolled charging of the electrodes which may be caused by laser light impinging on the electrodes and/or changing mg coverage of the chip surface during loading of the trap .
in particular , the applied compensation fields vary within a few tens of volts per meter on a day - to - day basis , and the axial - mode frequency can vary more than @xmath101 khz over the course of a few weeks with many loading cycles .
we estimate the deviation of ion displacements from simulated values due to stray - electric fields to be less than a few percent for typical displacements from the rf null of @xmath265 m at radial - mode frequencies of @xmath266 mhz .
we do not include this effect in our analysis of the experiments described in sec .
[ sec : oscb ] and sec .
[ sec : mm ] .
however , for the calibrations described in sec .
[ sec : radialmode ] , we take the stray - electric field into account because of its significant impact on the radial motional - mode vector directions .
in addition to the oscillating magnetic field @xmath267 , an electric field @xmath268 ( at the microwave frequency ) is also present ( fig [ fig : mwfields]b ) .
if the null points for @xmath268 and @xmath267 fields differ , the pseudopotential produced by @xmath268 can push the ions away from the main trap rf pseudopotential null . from numerical simulations , we determine that the ion position in a @xmath269 mhz trap is shifted by @xmath270 nm when @xmath271 t / m ( fig .
[ fig : mwfields]b ) .
this displacement is much smaller than the precision for positioning the ion with the shim fields ( sec .
[ sec : trap ] ) and therefore this shift can not be resolved .
another systematic effect is caused by the curvature of the microwave pseudopotential .
it modifies the mode frequency ( based on simulations ) by about @xmath254 khz , depending on the mode direction .
this can be a significant shift for the two - qubit gate .
however , when calibrating the sideband frequencies this effect is included , since we probe the sidebands with the microwave drive , which inherently includes the microwave pseudopotential produced by the gate pulse . to experimentally check for this shift , we measure the mode frequencies by applying a oscillating potential ( coherent excitation ) to one of the control electrodes , while the ion is in the @xmath272 state
we then detect the ion fluorescence as a function of excitation frequency : a decrease in the resulting fluorescence indicates that a mode of the ion s motion has been resonantly excited @xcite .
we compare the measured mode frequencies with and without the microwave fields while applying the excitation pulse . with the microwaves applied ,
we observe shifts of the radial - mode frequencies of @xmath97 khz .
the observed shift is about seven times higher than what we expect from our estimate ; this discrepancy is not understood . the mode frequency shift ( for a @xmath269 mhz trap ) for the ion in the @xmath272 state due to the spatially varying ac zeeman shift
is estimated to be less than @xmath273 mhz and can be neglected here .
we checked for motional excitations when simultaneously applying two tones near @xmath274 , as required for the two - qubit gate ( sec .
[ sec : limits ] ) . for pulse lengths up to @xmath87
ms ( four times longer than typical gate times ) we observe no change in the motional state within an uncertainty of approximately @xmath275 quanta .
however , when we apply the two tones at @xmath276 , we record coherent motional excitations corresponding to a few tens of quanta after @xmath35s .
the origin of the observed excitation may be the gradients of the two microwave pseudopotentials ( with difference frequency @xmath277 ) .
we estimate this effect to be negligible for our experimental conditions .
however , we can not rule out that the excitation is caused directly by an electric field at frequency @xmath277 .
this field may originate from mixing of the two microwave tones due to nonlinear behavior of one or more of the microwave elements .
57ifxundefined [ 1 ] ifx#1 ifnum [ 1 ] # 1firstoftwo secondoftwo ifx [ 1 ] # 1firstoftwo secondoftwo `` `` # 1''''@noop [ 0]secondoftwosanitize@url [ 0 ]
+ 12$12 & 12#1212_12%12@startlink[1]@endlink[0]@bib@innerbibempty @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop ( )
@noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) _ _ , @noop ph.d .
thesis , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop ( ) @noop * * , ( ) @noop * * , ( ) | microwave near - field quantum control of spin and motional degrees of freedom of @xmath0mg@xmath1 ions can be used to generate two - ion entanglement , as recently demonstrated in ospelkaus _
et al . _
[ nature @xmath2 , 181 ( 2011 ) ] . here
, we describe additional details of the setup and calibration procedures for these experiments .
we discuss the design and characteristics of the surface - electrode trap and the microwave system , and compare experimental measurements of the microwave near - fields with numerical simulations .
additionally , we present a method that utilizes oscillating magnetic - field gradients to detect micromotion induced by the ponderomotive radio - frequency potential in linear traps . finally , we discuss the present limitations of microwave - driven two - ion entangling gates in our system . |
the current paradigm for star formation is that most stars form in stellar clusters rather than in taurus - like star - forming regions . since binary systems are more frequent than single stars ,
the study of multiple systems in clusters is expected to provide us with some insight on the stellar formation process itself .
for instance , it has been suggested that the gravitational collapse of molecular clouds leads to a binary fraction close to 100% , which would decrease over time to about 60% , the value observed on the main sequence ( ghez _ et al .
the physical processes leading to such an evolution and the timescale over which they act still need to be asserted . alternatively , the impact of local physical conditions in the molecular cloud , such as the magnetic field strength or gas temperature , on the output of the star formation process has still to be ascertain and might lead to different binary frequencies in various environments ( durisen & sterzik 1994 ) .
observing stellar clusters with a wide range of physical parameters allows to perform a direct comparison between binary properties and model predictions .
clusters with ages ranging from one to several hundred million years are known within 500pc from the sun with some hosting many massive stars while other containing only low - mass stars .
the stellar density can vary over more than an order of magnitude from cluster to cluster , as well as the physical conditions of the parent molecular cloud . in this contribution
, we review the observed properties of visual binaries in clusters and compare them with cluster properties and with model predictions regarding the formation and evolution of stellar clusters .
the advent of several high - angular imaging techniques makes the binary frequency a relatively straightforward property to derive in stellar populations .
for instance , @xmath0 imaging ( prosser _ et al .
_ 1994 ) , speckle interferometry ( leinert _ et al .
_ 1993 , patience _
et al . _ 1998 ) and adaptive optics ( bouvier _ et al . _
1997 ) have been used , mostly in the near - infrared in order to facilitate the detection of relatively faint and cold low - mass companions .
the low - mass populations of extremely young clusters ( the trapezium cluster , ic348 ) as well as of more evolved ones ( @xmath1per , the pleiades , the hyades , prsepe ) have been studied , which span an age range from @xmath21 to several 100 myr .
the range of semi - major axes probed by these surveys usually extends from a few 10 to @xmath22000 au .
the hyades cluster , which is the closest to the sun , is the only one for which visual binaries could be resolved down to about 5au .
most surveys miss the faintest companions so that a statistical completeness correction has to be applied in order to estimate the true binary frequency from the observed one .
this correction can be large , especially at small separations for adaptive optics surveys due to the separation - dependent detection limit inherent to this technique .
the correction factor also depends upon assumptions regarding the distribution of mass ratios and/or orbital periods of the binary population , which obviously increases the statistical uncertainties on the derived binary frequency . with these limitations ,
all studied clusters are found to exhibit a visual binary frequency consistent with that of field g - dwarfs ( duquennoy & mayor 1991 ) , to within 10% or less ( e.g. prosser _ et al .
_ 1994 , bouvier _ et al .
_ 1997 , duchne _ et al .
this result does not seem to depend upon the age of the cluster between @xmath21 myr and 600 myr .
one possible exception might be the hyades cluster which appears to host more tight ( 550au ) binaries than the field population does , though statistical uncertainties could account for this slight excess ( patience _ et al .
it hence appears that the fraction of binaries in very young clusters , at an age of about a million years , is already similar to that measured for field dwarfs and that it does not significantly evolve as the clusters dynamically relax on the zams .
while the overall binary frequency integrated over a range of semi - major axes already provides clues to the properties of binary populations in clusters , one would ideally wish to compare the distribution of their orbital periods .
one is thus led to convert the observed angular separations into orbital periods assuming an average system mass and applying a statistical correction factor to take into account projection effects . over a limited range of orbital periods
, the distribution of semi - major axes in cluster binaries again appears to be consistent with that of field g - dwarf binaries ( duquennoy & mayor , 1991 ) .
the comparison between cluster and field binaries can be further investigated by considering not only visual binaries but spectroscopic binaries as well .
this has been done for zams clusters , where the low - mass stars have already been searched for short - period companions . within statistical uncertainties ,
the results reinforce the similarity between cluster and field binaries , as shown for the pleiades in figure1 . in order to improve the statistics
, we combined the results obtained for several zams clusters .
this yields the orbital period distribution shown in figure2 for visual binaries . while the distribution of period in cluster binaries agrees with that of field binaries over the orbital range @xmath3@xmath4days , a paucity of wide ( sep @xmath5 300 au ) cluster binaries appears relative to the field at the 2.5@xmath6 level .
such a deficit can not be related to a detection bias , since the distant companions of wide binaries are the most easily detected .
the uncompleteness correction which was applied here affects only the first orbital bin of the distribution .
this difference between field and cluster binaries is reinforced by the evidence that the orbital period distribution peaks at a smaller value in clusters than in the field ( patience _ et al .
the mode of the distribution is 5au for cluster binaries and 40au for field binaries , the former being estimated under the assumption that all clusters , including the hyades , can be combined into a single sample .
furthermore , scally _
( 1999 ) have shown that there are no very wide binaries ( sep @xmath5 2000 au ) in the trapezium cluster .
hence , even though clusters at various evolutionary stages appear to host a roughly similar fraction of low - mass binary systems , which itself is similar to the field binary fraction , the distribution of orbital periods appears to exhibit small but significant differences between clusters and field binary populations .
the mass of the primaries and secondaries of binary systems on the zams can be reliably estimated from their near - infrared flux using main sequence mass - luminosity relationships . in order to increase the statistical significance of the derived mass ratio distribution ,
the combined results for three clusters observed with similar set - ups are shown in figure3 . the observed distribution appears to be essentially flat over the range @xmath71 .
this agrees with the @xmath8-distribution derived by duquennoy & mayor ( 1991 ) for low - mass field binaries if the limited sensitivity of our observations , which prevents us from detecting tight systems with low mass ratios , is taken into account .
contrary to main sequence binaries , young low - mass systems frequently exhibit significant continuum excesses at near - infrared wavelengths and this may affect the derivation of mass ratios from near - ir flux ratios .
however , it is usually assumed on a statistical basis that both stars in a young binary system exhibit similar excesses , and that the overall flux ratio distribution can be safely transformed into a mass ratio distribution .
( 1998 ) s models were used to get the mass - luminosity relationship relevant for such young stars . both in the taurus star - forming region ( leinert _ et al .
_ 1993 ) and in the trapezium cluster ( prosser _ et al .
_ 1994 ) , the observed @xmath8-distribution is consistent with being flat down to @xmath9 , and therefore not different from the zams and field estimates .
current results thus suggest that the mass ratio distribution is similar for all low - mass binary populations studied so far , independent of their evolutionary status and environmental conditions ( field or clusters ) .
no peak is found in the distribution towards @xmath10 nor towards low @xmath8 .
furthermore , the zams clusters mass ratio distribution does not seem to vary with separation for visual binaries ( patience _ et al .
_ 2000 ) .
studies conducted so far have mostly focussed on the low - mass binary population of clusters .
more recently , however , the investigation of massive binaries has started , with the additional difficulty of a huge contrast between the bright primary and the low - mass companion which is a major limitation of current massive binary surveys .
the recent _ hst _ study of thirteen ob stars in the trapezium cluster by preibisch _
( 1999 ) already revealed eight visual companions with separations ranging from 15 to 450au .
the binary frequency of high - mass stars in this cluster thus appears significantly higher than that of their low - mass counterparts ( prosser _ et al .
_ 1994 ) , and the difference is further enhanced when completeness corrections are taken into account .
similarly , an adaptive optics survey of about fifty ob stars in the distant cluster ngc6611 revealed eight companions in the separation range 2002000au ( duchne 2000 ) , indicative of a high binary fraction compared to g - type field dwarfs if the orbital period distribution is the same for high - mass and low - mass binaries .
a study of young , mostly isolated , herbig aebe stars conducted by bouvier & corporon ( 2000 ) yields an uncorrected binary frequency which is significantly higher than that of lower mass main sequence stars for both be and ae pms stars .
the mass - ratio distribution of massive binaries can be estimated from the flux ratios measured by imaging surveys .
the @xmath8-distribution derived for massive binaries with b0b7 primaries is shown in figure4 .
the sample includes members of the trapezium cluster and ngc 6611 as well as isolated herbig be stars , with a semi - major axis ranging from 100 to 2000au .
the mass ratio distribution is strongly peaked towards small values : only about 25% of the binaries
have @xmath11 .
et al . _ ( 2000 ) obtained a similar trend for a and f stars in various zams clusters .
this distribution is dramatically different from that of low - mass binaries , indicating that the primary s mass is a relevant parameter when investigating the formation and evolution of binary systems .
the observed properties of young binary systems hold clues to their formation mechanism and their early evolution . in the following , we consider both classes of models , binary formation and dynamical evolution of clusters , and compare their predictions with observations .
the fact that all clusters studied so far exhibit a similar fraction of visual binaries regardless of their age rules out a long - term decrease of the binary frequency past the first million years after gravitational collapse .
in addition , since similarly - aged loose associations and clusters display different binary frequencies , it is tempting to conclude that environmental effects play a key role in fixing the binary content of a star forming region .
@xmath12-body simulations of the dynamical evolution of stellar clusters were ran by kroupa _
( 1999 ) and compared with observations of the trapezium cluster .
they found that close encounters between cluster members could efficiently disrupt binaries in less than a million years while the number of binaries would remain constant afterwards .
this results from the fact that the number of encounters falls off dramatically after one relaxation time .
if the initial low - mass binary fraction is of the order of 100% at the end of the protostellar collapse , such a scenario would naturally explain the high binary frequency observed in ttauri associations : in these regions , the stellar density is several orders of magnitudes too low to generate numerous enough close - by interactions so as to impact on the initial binary content ( clarke & pringle 1991 ) .
in contrast , frequent gravitational encounters in dense clusters would rapidly erode the initially high binary fraction and lead to a binary frequency similar to that of the field after a million years of evolution or so .
the observed properties of binary populations described in section2 further support the hypothesis that dynamical evolution of clusters may largely drive the binary properties . in dynamical models , wide binaries are preferentially disrupted , because their binding energy is small compared to the average kinetic energy of the cluster ( see the review by p. kroupa in this volume ) .
this prediction agrees with the observed distribution of orbital periods for binaries in zams clusters which exhibits a paucity of wide systems , as shown in figure5 .
the higher binary fraction found for massive stars is also in qualitative agreement with this model since the potential well of massive primaries is deeper than that of solar - mass stars , making it more difficult to disrupt more massive binaries . the `` softening boundary '' which delineates hard and soft systems , the latter being prone to disruption ,
is thus moved towards wider separations .
a number of issues remain however to be addressed before accepting this paradigm .
for instance , dynamical models predict that low-@xmath8 systems are preferentially disrupted , introducing a non - negligible modification in the shape of this distribution .
this constrasts with the derivation of similar mass ratio distributions for low - mass binaries in clusters and associations .
quantitatively , kroupa ( 1995 ) estimated that 70% of the initial binaries must have a mass ratio @xmath13 in order to account for the observed distribution of field binaries .
this is unlike the @xmath8-distribution observed for ttauri associations .
another puzzling result is the fact that all clusters seem to have very similar binary frequency and orbital period distributions . hence , binary properties appear to be weakly dependent on small differences in initial conditions .
dyamical models have focussed on reproducing the binary population of the trapezium cluster and it remains to be seen whether the same results would be obtained for slightly less dense clusters such as ic348 .
according to these models , wide binaries found in the solar neighbourhood can not have formed in clusters , since they would have been rapidly disrupted .
one thus has to assume that they were born in loose associations that have subsequently dissipated in the field .
however , the field binary population does not appear to be a simple combination of cluster and association members : a weighted mixture of these 2 populations that reproduces the observed fraction of tight and wide binaries also yields a large excess of systems in the 550au separation range compared to field binaries .
the issue of the efficiency of star formation in clusters as opposed to associations is still open .
if the dynamical evolution of clusters is responsible for the observed binary properties , then the formation process itself must yield a large initial binary fraction .
furthermore , the masses of the components must be taken at random from the initial mass function , as suggested by kroupa _
we note that the observations of high - mass binaries are qualitatively consistent with random pairing , with a limited number of @xmath11 pairs , though this result is still preliminary due to the small sample size
. we will now discuss whether binary formation mechanisms do predict such properties for initial binaries , prior to cluster s evolution .
the formation of a stellar cluster probably results from the fragmentation of the parent molecular cloud into a large number of individual cores . in this environment
, wide binary systems can form either by capture or through the fragmentation of a small collapsing core .
models considering capture in small-@xmath12 ( @xmath14 ) subclusters predict that the two most massive members are usually associated in a binary system , while the other stars are ejected ( mcdonald & clarke 1993 ) .
this leads to a mass ratio distribution which either peaks towards unity or is roughly flat , depending on the value of @xmath12 .
however , taking into account the dissipative effect of protostellar disks during close encounters leads to a secondary mass distribution that mimics the imf ( mcdonald & clarke 1995 ) .
furthermore , the model predicts a higher binary frequency for more massive stars .
similar simulations made by sterzik & durisen ( 1998 ) additionally predict that the mass ratio distribution is more heavily weighted towards low-@xmath8 values in higher mass systems .
all these predictions are consistent with the observations , indicating that stars in clusters may form in compact subclusters .
however , such a scenario does not end up with a 100% initial binary frequency , since some single stars are ejected from each subcluster .
capture during the relaxation of small-@xmath12 subgroups appears to be the model that best fits current observations , and especially the random pairing of secondary masses .
the alternative model of scale - free fragmentation ( clarke 1996 ) predicts that binary properties do not depend on the primary mass , which conflicts with the observations .
binary properties could also result , at least partly , from the accretion of residual material onto two gravitationally bound seeds ( bate & bonnell 1997 , bate 2000 ) .
however , these models predict a larger number of @xmath15 binaries for tighter systems as well as for more massive binaries , which is not verified in the observational trends .
finally , binaries formed through disk fragmentation should preferentially have small separations and low mass ratios , a prediction which contrasts with the apparent lack of a dependence of the mass ratio distribution upon separation .
the advent of high - angular resolution imaging techniques have offered the opportunity to systematically survey binaries in young clusters , thus complementing spectroscopic studies .
we reviewed some statistical properties of visual binaries such as their frequency , orbital period distribution and mass ratio distribution ; both low - mass and high - mass stars were considered .
observational results can be summarized as follows : the low - mass binaries in cluster have similar properties as main sequence ( field ) binaries , except for a significant deficit of systems with a separation larger than 300au .
high - mass stars host many more companions than do solar - type stars , and most massive binaries have a small mass ratio .
these observational constraints are consistent with the results of numerical simulations describing the early dynamical evolution of stellar clusters provided that the initial binary frequency is of the order of 100% for low - mass stars .
disk - assisted capture in small subclusters is the formation model that best approaches the initial conditions required to account for the observations . in both classes of models , however , several issues still need to be investigated and additional simulations over a wider range of physical conditions have to be run before these scenarios can be considered as offering an acceptable new paradigm for the star and binary formation process . in parallel , future observations of young embedded clusters will help to derive the properties of protobinaries at an even earlier stage of evolution than discussed in this review , which will undoubtly lead to new clues to their formation . | in this review , we summarize the observed statistical properties of visual binaries in various young stellar clusters . the binary frequency , orbital period distribution and mass ratio distribution are considered for populations of both low - mass and high - mass stars .
these properties are then compared to numerical models of the dynamical evolution of stellar clusters and to the predictions of binary formation mechanisms .
_ keywords _ : visual binaries statistical properties young stellar clusters dynamical evolution binary formation mechanisms
# 1_#1 _ # 1_#1 _ = # 1 1.25 in .125 in .25 in |
although a large number of physical systems admit studies of their basic properties using different types of time - independent formalisms , in many important cases one has to explicitly take into account the time dependence and to apply the appropriate approach , i.e. one of available time - dependent analytical or numerical methods .
this has to be done even for systems with time - independent potentials if their dynamical properties and time evolution is studied .
however , for describing systems in genuinely time - dependent external potentials or for rotating systems , using of such approaches is always necessary .
this applies equally to classical and quantum systems , and various methods have been developed to address relevant physical problems . for quantum systems a number of general methods is available , ranging from time - dependent perturbation theory and variational perturbation theory , to specialized approaches such as the density matrix renormalization group ( dmrg ) @xcite , the density functional theory ( dft ) @xcite , and the density matrix functional theory ( dmft ) @xcite .
in addition to these generic schemes , several specific numerical methods have been developed for enabling a quantitative description of quantum systems that have attracted a significant research interest .
this includes studies of ultra - cold quantum gases @xcite and optical lattices @xcite , whose comprehensive and highly tunable features make them an important example of feynman s quantum simulator @xcite . in such numerical approaches
@xcite usually a second - order algorithm in the propagation time is used , which is basically the same as in the time - independent case . however ,
also higher order schemes have been derived , including fourth @xcite and higher - order expansions @xcite .
the main goal of this paper is to develop a formalism which enables a systematic improvement in the convergence order of numerical algorithms for general time - dependent quantum systems . in our previous paper @xcite ,
we have extended the earlier established approach @xcite for obtaining a high - order short - time expansion of transition amplitudes for time - independent potentials to the important time - dependent case . in that paper , we have first calculated a short - time expansion of a generic transition amplitude using the path integral formalism to fourth order in the propagation time , which has then served the important purpose : to introduce the proper functional form of the ansatz for the ideal discretized effective potential , as well as its short - time double series expansion in both the propagation time and the discretized velocity .
using both the forward and backward schr " odinger equation for transition amplitudes , we have then derived the appropriate equation for the ideal effective potential , which represents an important generalization of the equation derived in ref .
@xcite for the time - independent case . using the double series expansion ansatz for the effective potential ,
we have been able to set up and solve recursive relations for the effective potential for one - dimensional systems , and to numerically verify that higher - order analytic approximative results for transition amplitudes obtained in this way , indeed , give the correct convergence order for a number of models . in this paper
we further develop and generalize the time - dependent approach introduced in ref .
first , in sec . [ sec : eff ]
we briefly review the time - dependent effective action approach in the path - integral formalism of quantum mechanics @xcite , as well as the main results of our previous paper @xcite .
then we extend the recursive approach for calculating the short - time expansion of the effective potential to quantum systems with many degrees of freedom in sec .
[ sec : mb ] . in this section ,
the generalized approach is also numerically verified for the case of a simple time - dependent multi - component harmonic oscillator system , and possible relevant physical applications in the realm of ultracold quantum gases are briefly indicated . in sec .
[ sec:1dz ] we present another important extension of the time - dependent formalism , and set up a specific recursive relation for calculating diagonal transition amplitudes , which are necessary for a numeric high - precision calculation of partition functions . in this case , a simplified set of recursive relations is obtained , which is numerically verified for several model potentials .
finally , sec .
[ sec : rt ] illustrates how the developed imaginary - time formalism can be transformed into a real - time one , and its applicability is numerically demonstrated by treating several models .
we also show how the real - time evolution can be described using the time - dependent effective action approach , and how the associated numerical errors can be assessed and controlled in typical applications .
in this section we give a brief overview of the time - dependent effective action approach established in ref .
this formalism considers a non - relativistic quantum multi - component system in @xmath0 spatial dimensions with a hamiltonian of the form @xmath1 where @xmath2 denotes the number of particles in the system and the @xmath3 dimensional vectors @xmath4 and @xmath5 contain positions and momenta of all particles , while the parenthetic subscript @xmath6 denotes the particle number .
for such a system we consider the calculation of the transition amplitudes @xmath7 where the vectors @xmath8 and @xmath9 describe the positions of all particles at the initial and final time @xmath10 and @xmath11 , respectively .
the above transition amplitude is a coordinate - space matrix element of the evolution operator , which describes the propagation of the system ( [ eq : hamiltonian ] ) from @xmath10 to @xmath11 , and is defined by the time - ordered exponential @xmath12 in the path - integral formalism , the transition amplitude can be expressed by a coordinate - space path integral @xmath13\right\}\ , , \ ] ] where the integration is defined over all possible trajectories @xmath14 .
this usually involves the discretization of the trajectories , which is usually performed by dividing the time evolution from @xmath10 to @xmath11 into @xmath15 equal time steps . in the above equation
, @xmath16 denotes the action for a given trajectory @xmath14 : @xmath17=\int_{t_a}^{t_b } dt\ , \left\ { \frac{1}{2}\dot{\mathbf q}^2(t)-v(\mathbf q(t ) , t)\right\}\ , .\ ] ] the common step at this point is to switch to the imaginary - time formalism , which is usually applied in numerical simulations @xcite , due to problems which may arise from the oscillatory nature of the integrand in the real - time approach .
we will do so as well in the next two sections , but in sec .
[ sec : rt ] we will switch the developed formalism back to the real time and demonstrate how it can be used for studying the dynamics and real - time evolution of quantum systems . after wick rotation to the imaginary time ,
the transition amplitude in the path - integral formalism is expressed as @xmath18}\ , , \ ] ] where the minkowski action is now replaced by its imaginary - time counterpart , the euclidean action @xmath19=\int_{t_a}^{t_b } dt\ , \left\ { \frac{1}{2}\dot{\mathbf q}^2(t)+v(\mathbf q(t ) , t)\right\}\ , , \ ] ] which is actually the energy functional for the system . as was shown previously for time - independent potentials @xcite , as well as for the time - dependent ones in ref .
@xcite , the exact imaginary - time transition amplitudes can be expressed in the form @xmath20 where @xmath21 stands for the ideal discretized action and depends on the coordinate mid - point @xmath22 , the discretized velocity @xmath23 , the time interval @xmath24 , and the time mid - point @xmath25 .
note that we have used the convention @xmath26 , and we have restricted ourselves to particles with unitary masses @xmath27 .
the ideal discretized action further reads @xcite @xmath28 where @xmath29 is the ideal discretized effective potential , which also depends on the time mid - point @xmath30 due to the explicit time dependence of the potential @xmath31 .
this represents the major difference in the formalism compared to the previously developed one in ref .
@xcite for the time - independent case . in order to determine a partial differential equation for the ideal effective potential @xmath29 , we have derived in sec .
4 of our previous paper @xcite the forward and the backward schrdinger equation for time derivatives of the transition amplitude with respect to the initial and final time , and have found that they obey @xmath32 where @xmath33 stands for the coordinate - space hamilton operator @xmath34 , in which momentum and position operators are replaced by their coordinate - space representations at @xmath9 ( and similarly for @xmath35 ) .
when we insert the ansatz ( [ eq : aexacta ] ) together with ( [ eq : aexactb ] ) into the equations ( [ eq : dtba ] ) and ( [ eq : dtaa ] ) , we obtain the corresponding partial differential equation for the effective potential in the form @xmath36 where @xmath37 , i.e. @xmath38 and @xmath39 .
as expected , the form of this equation is the same as eq .
( 29 ) from ref .
@xcite , and if the potential @xmath31 does not depend on time , we immediately obtain the previously derived result .
the above equation is the starting point for calculating the effective potential .
naturally , it can be solved analytically only for exactly solvable models . however , if the propagation time is short , we can perform a short - time expansion of the effective potential and set up appropriate equations for the coefficients in this expansion . in our previous paper @xcite this
was done for one - dimensional systems , and it was shown that we have to use a double expansion of the effective potential in both @xmath40 and @xmath41 .
the reason for this is the fact that the propagation in the imaginary time is equivalent to diffusion , and therefore , on the average , we expect the diffusion relation @xmath42 to hold .
this allows us to effectively couple the two expansion parameters and to establish a unique counting of powers in @xmath40 for all terms in the expansion for @xmath29 .
however , if the diffusion relation is not applicable , the two expansions can be considered as independent , and the whole approach can still be applied , with an independent counting of powers in @xmath40 and in @xmath41 . in the next section
we will analytically derive a systematic short - time expansion of the effective potential @xmath29 for quantum many - body systems , which yields a significant improvement in the convergence of numerically calculated transition amplitudes and partition functions for systems in time - dependent potentials . as we see from eq .
( [ eq : aexactb ] ) , if the effective potential @xmath29 is calculated to order @xmath43 , we get the effective action correct to order @xmath44 . if we take into account the normalization factor in the expression ( [ eq : aexacta ] ) , the corresponding error in the calculation of the short - time amplitude is given by @xmath45 where subscript @xmath46 denotes that we use the effective action of that order in @xmath40 .
therefore , we see that the analytical calculation of higher - order effective actions is beneficial , since it provides an analytic approximation for transition amplitudes which yield high - precision results of the desired order in numerical calculations .
now we focus on the development of the recursive formalism for calculating the effective potential for the case of a general multi - component non - relativistic quantum system of @xmath2 particles in @xmath0 dimensions by extending the one - dimensional calculations of sec . 5 in our previous paper @xcite .
note that such a formalism is needed even for studies of single - particle systems in two or three spatial dimensions , which have more than one degree of freedom .
for example , such time - independent many - body effective actions of level @xmath47 have already been used for a numerical study of fast - rotating bose - einstein condensates @xcite , as well as a high - precision calculation of the energy spectra and eigenfunctions of several two - dimensional models @xcite .
the presented extension of the many - body formalism will allow studies of such systems in external time - dependent potentials , as well as the investigation of the formation and evolution of vortices @xcite and other dynamical phenomena .
we also plan to study collective oscillation modes of bose - einstein condensates with a parametrically modulated interaction @xcite . to develop the time - dependent many - body formalism , we solve the partial differential equation ( [ eq : weq ] ) for the effective potential @xmath29 by using a multi - dimensional many - particle generalization of the double power expansion used in eq .
( 33 ) of our previous paper @xcite for one - component systems , which has the form @xmath48 here we have introduced the contractions @xmath49 @xmath50 in such a way that they correspond to the case of the time - independent potential @xcite . in the above relations we assume
the einstein convention that summation over repeated indices is performed .
introducing such contractions of tensorial coefficients
@xmath51 significantly simplifies in this case the analytic derivation and also provides a key ingredient for implementing the many - body recursion relations in symbolic calculations using e.g. mathematica software package @xcite .
otherwise , the task to explicitly symmetrize the coefficients would amount to a complexity of the algorithm which scales with the number of possible permutations @xmath52 and which would not be feasible even for a very moderate number of particles .
using scalar quantities , which are obtained by contracting the coefficients with the discretized velocity @xmath41 , efficiently solves this problem . the multiple - component version of the expression from the right - hand side of eq .
( [ eq : weq ] ) has the form @xmath53 where @xmath54 represents @xmath55-th partial derivative of the potential with respect to the time @xmath30 . after inserting the above expression into the partial differential equation ( [ eq : weq ] ) for the effective potential we straightforwardly obtain recursive relations for even- and odd - power contractions
@xmath56 and @xmath57 : @xmath58 the diagonal contractions can easily be calculated in a closed form as in the single - particle one - dimensional case , yielding @xmath59 thus , the recursion relations ( [ eq : receven ] ) and ( [ eq : recodd ] ) can be solved up to a given order @xmath46 together with ( [ eq : wmm ] ) and ( [ eq : wm12 m ] ) by using a similar procedure as before .
here we give the solution up to order @xmath60 , which generalizes the previously given solution for the simple case @xmath61 , obtained in ref . @xcite .
for @xmath62 we only have the naive @xmath63 term , i.e. @xmath64 while @xmath65 yields the first non - trivial even - power terms @xmath66 which are sufficient to construct @xmath67 effective action . the next term we calculate
is the odd - power contraction for @xmath65 , i.e. @xmath68 which contains the explicit time derivative of @xmath31 .
for @xmath69 we obtain the next order of even - power terms : @xmath70 these terms , together with the previously calculated ones , are sufficient to construct level @xmath71 effective action . in order to be able to complete @xmath60 effective action derivation
, we still need to calculate odd - power contractions corresponding to @xmath69 @xmath72 as well as even - power contractions for @xmath73 , @xmath74 this concludes the calculation of level @xmath60 effective action for a general many - body quantum system . as before ,
the obtained results automatically reduce to the already known effective actions for the time - independent potentials @xcite if we set all time - derivatives of the potential to zero . furthermore , the many - body results ( [ nai])([mbb ] ) reduce for the special case @xmath61 to the previous time - dependent results @xcite .
the outlined procedure continues in the same way for higher levels @xmath46 .
we have automatized this procedure and implemented it in our speedup code @xcite using the mathematica software package @xcite for symbolic calculus . in order to numerically verify the developed expressions for the case of multi - component quantum systems
, we will calculate transition amplitudes of a set of time - dependent harmonic oscillators , @xmath75 which represents the archetypical model for many physical phenomena .
this exactly solvable model allows us to compare analytical expressions obtained from recursive relations and to verify that using level @xmath46 effective action leads to values of transition amplitudes which are correct up to order @xmath76 . as we can see in fig .
[ fig : mc ] for a system of @xmath77 time - dependent oscillators , the respective scaling is perfect .
the middle and bottom plots illustrate another important feature of the short - time expansion for multi - component systems : as the number of components of the system @xmath78 increases , the exponent @xmath79 may become zero or negative for a given effective action level @xmath46 .
this leads to the peculiar behavior observed in the middle and bottom plots for small values of @xmath46 , with the deviation of the amplitude being constant ( @xmath80 , @xmath63 in the middle plot , @xmath81 , @xmath67 in the bottom plot ) or even increasing ( @xmath81 , @xmath63 in the bottom plot ) when @xmath40 is decreased .
thus , this prevents the calculation of the transition amplitude with high accuracy , which is , in principle , expected to be possible by decreasing @xmath40 .
as we see , to enable such high - accuracy calculations , one has to use an effective action with sufficiently high level @xmath46 .
the important contribution of the presented approach lies in the fact that it offers a systematic formalism for deriving such higher order expressions for a general quantum multi - component system . from the exact values as a function of propagation time @xmath40 for a multi - component system ( [ eq : mctdho ] ) of time - dependent harmonic oscillators :
( top left ) @xmath82 oscillators , with @xmath83 , @xmath84 ; ( top right ) @xmath80 oscillators , with @xmath85 , @xmath86 , @xmath87 , @xmath88 ; ( bottom ) @xmath81 oscillators , with @xmath85 , @xmath86 , @xmath89 , @xmath90 , @xmath91 , @xmath92 .
each plot gives results for transition amplitudes calculated using the effective action levels @xmath93 , from top to bottom.,title="fig:",width=302 ] from the exact values as a function of propagation time @xmath40 for a multi - component system ( [ eq : mctdho ] ) of time - dependent harmonic oscillators : ( top left ) @xmath82 oscillators , with @xmath83 , @xmath84 ; ( top right ) @xmath80 oscillators , with @xmath85 , @xmath86 , @xmath87 , @xmath88 ; ( bottom ) @xmath81 oscillators , with @xmath85 , @xmath86 , @xmath89 , @xmath90 , @xmath91 , @xmath92 .
each plot gives results for transition amplitudes calculated using the effective action levels @xmath93 , from top to bottom.,title="fig:",width=302 ] from the exact values as a function of propagation time @xmath40 for a multi - component system ( [ eq : mctdho ] ) of time - dependent harmonic oscillators : ( top left ) @xmath82 oscillators , with @xmath83 , @xmath84 ; ( top right ) @xmath80 oscillators , with @xmath85 , @xmath86 , @xmath87 , @xmath88 ; ( bottom ) @xmath81 oscillators , with @xmath85 , @xmath86 , @xmath89 , @xmath90 , @xmath91 , @xmath92 .
each plot gives results for transition amplitudes calculated using the effective action levels @xmath93 , from top to bottom.,title="fig:",width=302 ] as in the case of single - particle one - dimensional systems , computational complexity of higher - order effective actions might significantly increase for higher values of @xmath46 ( typically exponentially , see e.g. fig .
5 in ref .
this increase strongly depends on the type of potential , but we expect to obtain significant computational benefits at least for moderate values of @xmath46 .
the optimal value of the level @xmath46 to be used in practical applications can always be found through a small - scale numerical complexity study , by measuring relative increase in cpu times as a function of @xmath46 . using this data , as well as the known scaling of monte carlo errors and cpu times on the time step and number of samples , optimal value of @xmath46
is obtained by minimizing the cpu time for a required precision of transition amplitudes . at the end of this section , we emphasize that the obtained discretized effective actions can be used for solving a plethora of non - equilibrium many - body quantum problems within the exact diagonalization @xcite or path integral monte carlo approach , including the continuous - space worm algorithm @xcite .
for instance , in typical experimental setups with ultracold quantum gases harmonic or anharmonic confining potentials are generically switched on and off , thus generating natural non - equilibrium situations .
as so far mainly quenched potentials have been considered , it would certainly be rewarding to study in a systematic way how the time scale , upon which a potential is switched off , influences the observed time - of - flight absorption pictures .
another upcoming research field is the investigation of the phenomenon of parametric resonance in bose - einstein condensates .
a first experiment , where the s - wave scattering length of @xmath94li atoms has been modulated periodically with the help of a feshbach resonance , has recently been performed @xcite . in order to understand the observed resonance spectrum both analytical methods from nonlinear dynamics @xcite and numerical methods as the presented fast converging path - integral approach have to be combined .
now we turn our attention to the special case of the velocity independent part of the effective potential .
it determines the diagonal amplitudes , for which the discretized velocity @xmath41 is equal to zero .
the efficient and precise calculation of diagonal transition amplitudes is essential for many quantum statistical problems , since it provides a direct method to obtain partition functions , and can be used to calculate energy spectra , density profiles , and other relevant physical quantities .
therefore , we will derive a new set of recursive relations for the coefficients which determine the velocity - independent part of the effective potential . for simplicity
, we will present the derivation for one - dimensional systems , where the effective potential for @xmath95 can be written as @xmath96 such recursive relations for coefficients @xmath97 will turn out to be much simpler than the full set of recursions obtained in the previous paper @xcite . in order to derive equations determining the coefficients @xmath98 , we have to perform the limit @xmath99 in the partial differential equation ( [ eq : weq ] ) for the effective potential @xmath29 .
this is nontrivial , since the equation contains derivatives with respect to @xmath100 .
therefore , we have to re - examine both schrdinger equations ( [ eq : dtba ] ) and ( [ eq : dtaa ] ) , and express them using the variables @xmath101 , @xmath100 , @xmath40 , and @xmath30 . after a change of variables , we get the following system of equations for the transition amplitude : @xmath102a(x , \bar { x } ; \varepsilon , \tau)&=&0\ , , \\
\label{eq : dt } \left[\partial_\tau
-\frac{1}{2}\partial\bar\partial+v_+ -v_- \right]a ( x , \bar { x } ; \varepsilon , \tau ) & = & 0\ , .\end{aligned}\ ] ] if we take the derivative with respect to @xmath100 of the second equation , use it to express the term @xmath103 , and insert it into the derivative of the first equation with respect to @xmath101 , we obtain the partial differential equation @xmath104 in which it is easier to perform the required @xmath99 limit . in the terms that do not contain derivatives with respect to @xmath100
, we can just set @xmath95 and replace the transition amplitude @xmath105 with @xmath106 . in the remaining terms
the limit has to be performed more carefully .
the terms containing combinations @xmath107 and their derivatives are the simplest , and we obtain @xmath108 where the prime in the last expression denotes the derivative with respect to @xmath101 . for the terms @xmath109 and @xmath110 we have to explicitly use the full double power expansion for the effective potential @xmath111 \ , , \ ] ] perform the differentiation and take the limit afterwards .
this yields the results @xmath112 where dots now represent derivatives with respect to the time argument @xmath30 of the coefficient @xmath113 and the effective potential @xmath114 .
as we see , the odd - power coefficients @xmath115 can not be eliminated altogether , although we are considering the diagonal amplitudes , for which we have @xmath95 .
this is due to the derivatives with respect to @xmath100 . from this
we can deduce that we will need again two recursion relations to determine all needed coefficients , despite the fact that in the end we will use only the even - power ones .
therefore , we will use eq .
( [ eq : dt ] ) to derive the second recursive relation for the coefficients .
in order to do so , we still need to calculate the term @xmath116 in the considered limit @xmath99 : @xmath117 inserting all calculated @xmath99 terms into equations ( [ eq : dt ] ) and ( [ eq : a0comb ] ) , as well as using the expansion ( [ eq : w0double ] ) , we finally obtain the following coupled system of recursive relations for the coefficients @xmath98 and @xmath115 : @xmath118 the above recursive relations are solved in a similar way as before . in order to obtain the level @xmath46 diagonal effective action @xmath114 , we need to take into account the terms in the expansion with @xmath119 .
the recursions for @xmath120 and @xmath121 are easily solved starting from @xmath62 up to a desired level @xmath122 .
although we get in this way only their first derivatives with respect to @xmath101 , the coefficients themselves can be calculated by direct symbolic integration , and all solutions can be expressed in a closed form .
the explicit calculation of the coefficients to high orders yields the same results we have obtained in the previous section . to order @xmath63
, we only have the trivial equation @xmath123 that gives the well - known boundary condition @xmath124 . to order @xmath67 we have @xmath125 which yields @xmath126 and @xmath127 . to order @xmath71
we first calculate the odd - power coefficient , @xmath128 which leads to the result @xmath129 .
the even - power coefficient @xmath130 is obtained from @xmath131 which finally yields @xmath132 these results coincide with the results obtained in our earlier paper @xcite from the general recursion relations .
we similarly proceed to calculate higher - order coefficients .
this is easily automated in symbolic calculus software packages like mathematica @xcite , and we have implemented the derived recursions as a part of our speedup @xcite code .
the main advantage of this approach is that we have been able to derive recursion relations involving only the lower - level coefficients @xmath133 and @xmath134 , thus not requiring the calculation of all even- and odd - power coefficients , which are needed for the general case . at the end of this section
, we note that for time - independent potential @xmath31 the recursive relation ( [ eq : cmrec ] ) reduces to the previously known result @xcite , while the second recursion ( [ eq : cm12rec ] ) yields the expected result @xmath135 .
note that our recursive approach thus allows to calculate higher orders of the seminal wigner expansion @xcite . as a function of propagation time @xmath40 for : ( left ) time - dependent harmonic oscillator ( [ eq : hogrosche ] ) , calculated analytically for @xmath136 from top to bottom ; ( right ) forced harmonic oscillator ( [ eq : fho ] ) with @xmath137 and @xmath138 , calculated analytically for @xmath139 from top to bottom .
the dashed lines on both graphs are proportional to @xmath140 and @xmath141 , and demonstrate the perfect scaling of the corresponding level @xmath63 and @xmath142 results.,title="fig:",width=306 ] as a function of propagation time @xmath40 for : ( left ) time - dependent harmonic oscillator ( [ eq : hogrosche ] ) , calculated analytically for @xmath136 from top to bottom ; ( right ) forced harmonic oscillator ( [ eq : fho ] ) with @xmath137 and @xmath138 , calculated analytically for @xmath139 from top to bottom .
the dashed lines on both graphs are proportional to @xmath140 and @xmath141 , and demonstrate the perfect scaling of the corresponding level @xmath63 and @xmath142 results.,title="fig:",width=306 ] fig .
[ fig : diag ] illustrates practical advantages of using velocity - independent effective actions for the numerical calculation of diagonal transition amplitudes .
the plot on the left gives the deviations of diagonal amplitudes calculated with different levels @xmath46 of the effective potential @xmath114 for the grosche - rescaled @xcite harmonic oscillator , @xmath143 while the plot on the right gives the analogous results for the forced harmonic oscillator @xmath144 where @xmath145 denotes the frequency of the external driving field .
both models are exactly solvable , and the obtained @xmath40-scaling to exceedingly high orders @xmath46 demonstrates the correctness of the analytically derived results .
the presented approach has so far been developed within the imaginary - time framework , which is useful in many practical applications .
however , in order to study the more relevant real - time dynamics of quantum systems , we have to switch back to the real - time formalism .
one possibility would be to make all calculations in imaginary time , and then to try to perform an inverse wick rotation , which might be difficult due to the oscillatory nature of the integrand in calculating the real - time propagator .
another , much more straight - forward possibility is to derive a new set of recursive relations within the real - time formalism . in this section
we will briefly outline such a procedure .
reverting the imaginary - time formalism into a real - time one is achieved by replacing the variable @xmath146 representing the imaginary time with @xmath147 in all expressions , where now @xmath148 represents the real time .
this includes also the replacement of the time - interval @xmath40 with @xmath149 , and the time - midpoint @xmath30 with its real - time counterpart @xmath150 . instead of ( [ eq : aexacta ] )
, the short - time transition amplitude is now expressed as @xmath151 and the real - time version of the ideal effective action is defined as @xmath152 which represents the counterpart of eq .
( [ eq : aexactb ] ) .
following the same procedure outlined in sec .
[ sec : eff ] , we arrive at the real - time counterpart of eq .
( [ eq : weq ] ) for the effective potential , @xmath153 where the subscript r is dropped for simplicity .
further derivation of real - time recursion relations is a straight - forward task . for brevity
, we will not give the explicit form of the recursion relations , but their mathematica implementation is available from the speedup code web page @xcite .
we illustrate the applicability of this formalism for studying the real - time dynamics within the space - discretized approach @xcite . if we discretize the continuous space and replace it with a grid defined by a discretization step @xmath154 , all quantities are only defined on a discrete set of coordinates @xmath155 , where @xmath156 is a vector of @xmath78 integer numbers .
matrix elements of the evolution operator , @xmath157 represent real - time transition amplitudes @xmath158 , which can be calculated using the real - time effective action approach .
if the initial state of the system @xmath159 is represented by a vector @xmath160 whose elements are @xmath161 , its dynamics can be calculated by a simple matrix multiplication @xmath162 , i.e. @xmath163 therefore , since the matrix elements of the evolution operator can be accurately calculated using the effective action approach , we are able to study real - time dynamics of the system starting from any desired initial state .
note that , although we rely on the short - time expansion of transition amplitudes , we are not limited to study only a short - time evolution , since the above matrix multiplication can be repeatedly performed .
for any given propagation time @xmath164 , we can divide the evolution to @xmath15 sub - intervals of length @xmath165 , which now correspond to short - time evolution matrix elements @xmath166 .
in addition to this , using the higher - order effective actions makes it possible to perform high - accuracy calculations of @xmath167 , and , correspondingly , to eliminate the associated numerical errors for all practical purposes .
calculated using the space - discretized method @xcite and the effective action approach with : ( left ) @xmath63 , @xmath168 and ( right ) @xmath142 .
both graphs display the time dependence @xmath169 of the absolute value of the wave function at @xmath170 , and solid line represents the exact solution .
the harmonic frequency was @xmath137 , the time - interval for propagation was @xmath171 , and the initial state was set to the ground state of the harmonic oscillator with @xmath172.,title="fig:",width=306 ] calculated using the space - discretized method @xcite and the effective action approach with : ( left ) @xmath63 , @xmath168 and ( right ) @xmath142 .
both graphs display the time dependence @xmath169 of the absolute value of the wave function at @xmath170 , and solid line represents the exact solution .
the harmonic frequency was @xmath137 , the time - interval for propagation was @xmath171 , and the initial state was set to the ground state of the harmonic oscillator with @xmath172.,title="fig:",width=306 ] to demonstrate this , we show in fig .
[ fig : rt1 ] the time evolution of a harmonic oscillator @xmath173 calculated using the described method with effective action levels @xmath63 , @xmath168 and @xmath142 .
as we can see , using the propagation interval @xmath171 , the naive @xmath63 action can be used only for short propagations times , while we are able to reproduce very accurately the long - time evolution of the system with higher @xmath46 levels . in order to further quantitatively assess numerical errors of the obtained results , we use the following integral measure , @xmath174 where @xmath175 represents the exact time evolution of the wave function and @xmath176 is the approximate time evolution calculated using level @xmath46 effective action . the semi - log plot in fig .
[ fig : rt2 ] gives the @xmath46-dependence of the above - defined integral measure and demonstrates that it obeys the expected power law , i.e. @xmath177 in this case , leading to a much smaller error when the propagation interval is reduced from @xmath178 to @xmath171 .
this graph also shows that errors due to the repeated matrix multiplication accumulate linearly with the number of time steps . ) for numerically calculated time evolution of the harmonic oscillator as a function of the effective action level @xmath46 for different values of the propagation time @xmath146 .
the parameters are the same as in fig .
[ fig : rt1].,width=453 ] the study of errors presented in fig .
[ fig : rt2 ] is very instrumental in optimizing numerical parameters in practical applications .
if we compare errors , which correspond to the same total evolution time @xmath179 and are calculated for a propagation in one time - step @xmath180 and in @xmath181 steps @xmath182 , we can see that decreasing @xmath40 substantially reduces errors .
this is easily understood , since errors are proportional to @xmath76 and , therefore , introducing @xmath15 time steps is expected to reduce errors by a factor of @xmath183 .
however , the fact that the matrix multiplication will have to be repeated @xmath15 times introduces an additional factor of @xmath15 , thus leading to the total reduction factor of @xmath184 . )
calculated using the space - discretized method @xcite and the effective action approach with @xmath63 , @xmath67 , and @xmath185 .
the graph displays the time dependence @xmath169 of the absolute value of the wave function at @xmath170 .
the time - dependent harmonic frequency is given by @xmath186 , time - interval for propagation was @xmath171 , and the initial state was set according to eq .
( [ eq : tdhopsi0]).,width=453 ] as a final example , we calculate the time evolution of the time - dependent harmonic oscillator with the potential @xmath187 with the frequency @xmath186 for @xmath63 and @xmath185 .
[ fig : rt3 ] displays the time evolution of the absolute value of the wave function at @xmath170 with the propagation interval @xmath171 , and the initial state set to @xmath188 as expected , the naive @xmath63 effective action can only be used for very short propagation times , while @xmath67 action gives accurate results for longer propagation times @xmath189 .
a moderate level @xmath185 effective action represents an even further substantial improvement and can be used to accurately study much longer propagation times , as can be seen from the inset in fig .
[ fig : rt3 ] .
we emphasize that the presented approach might be especially of interest for path integral monte carlo ( pimc ) calculations of the dynamics of quantum systems , in conjunction e.g. with the multilevel blocking method @xcite .
dynamical pimc calculations are quite difficult due to the dynamical sign problem , and the availability of highly accurate propagators allows use of relatively small number of trotter slices , thus substantially improving the numerical convergence .
in this paper we have presented a significant extension of the approach introduced in the preceding paper @xcite , which has established a recursive procedure for calculating the short - time transition amplitudes for one - dimensional quantum systems in time - dependent potentials .
this approach is generalized here to non - relativistic many - body quantum systems with many degrees of freedom . in parallel to the approach for time - independent potentials @xcite , we have introduced an ideal effective potential for time - dependent systems , derived the appropriate equation using the forward and backward schr " odinger equation for the transition amplitude , and set up an efficient system of recursive relations , which can be analytically solved to high orders in the short propagation time .
furthermore , we have implemented a symbolic calculation scheme for higher - order effective actions in the speedup code @xcite .
the analytically derived results are verified by studying several models and a list of possible applications of the presented method to relevant many - body quantum systems has been briefly outlined .
in addition to this , we have also studied velocity - independent part of the effective action , which is relevant for calculating the diagonal amplitudes and partition functions .
we have obtained a new , simpler set of recursion relations , which determine the diagonal effective action , and have numerically verified that it yields the correct systematic increase in the convergence of diagonal amplitudes for several models .
finally , we have also looked at how the developed formalism can be transformed from its original , imaginary - time setup to the real - time one . we have derived the real - time counterparts of equations for the effective potential and applied the higher - order real - time effective actions to a numerical study of the time evolution of several models using the space - discretized approach @xcite .
we have demonstrated that the presented approach can be successfully used both in the real - time and in the imaginary - time formalism , and that in both cases we obtain analytically derived improved convergence of numerically calculated transition amplitudes and other quantities .
we point out that the presented approach can contribute to improving path integral monte carlo calculations of the dynamics of quantum systems in conjunction e.g. with the multilevel blocking method , since higher - order propagators enable use of fewer number of trotter slices .
we thank hagen kleinert for several useful suggestions . this work was supported in part by the ministry of science and technological development of the republic of serbia , under project no .
on171017 and bilateral projects pi - bec and nad - bec , funded jointly with the german academic exchange service ( daad ) .
numerical simulations were run on the aegis e - infrastructure , supported in part by fp7 project egi - inspire , hp - see and prace-1ip .
a. bogojevi ' c , a. bala , and a. beli ' c , phys .
lett . * 94 * , 180403 ( 2005 ) .
a. bogojevi ' c , a. bala , and a. beli ' c , phys .
b * 72 * , 064302 ( 2005 ) . a. bogojevi ' c , i. vidanovi ' c , a. bala , and a. beli ' c , phys . lett .
a * 372 * , 3341 ( 2008 ) .
r. p. feynman , rev .
phys . * 20 * , 367 ( 1948 ) .
r. p. feynman and a. r. hibbs , _ quantum mechanics and path integrals _ ( mcgraw - hill , new york , 1965 ) .
r. p. feynman , _ statistical mechanics _
( benjamin , new york , 1972 ) .
h. kleinert , _ path integrals in quantum mechanics , statistics , polymer physics , and financial markets _ , 5th ed .
( world scientific , singapore , 2009 ) . | based on a previously developed recursive approach for calculating the short - time expansion of the propagator for systems with time - independent potentials and its time - dependent generalization for simple single - particle systems , in this paper we present a full extension of this formalism to a general quantum system with many degrees of freedom in a time - dependent potential .
furthermore , we also present a recursive approach for the velocity - independent part of the effective potential , which is necessary for calculating diagonal amplitudes and partition functions , as well as an extension from the imaginary - time formalism to the real - time one , which enables to study the dynamical properties of quantum systems .
the recursive approach developed here allows an analytic derivation of the short - time expansion to orders that have not been accessible before , using the implemented speedup symbolic calculation code .
the analytically derived results are extensively numerically verified by treating several models in both imaginary and real time . |
ground states of deformed exotic nuclei , unstable isotopes with extreme isospin values , display many interesting and unique properties .
the description of observed phenomena , as well as the prediction of new and unexpected properties , present an exciting challenge for modern theoretical advances . for neutron - rich nuclei , a fascinating example of the modification of the effective single - nucleon potential
is the observed suppression of shell effects , the disappearance of spherical magic numbers , and the resulting onset of deformation and shape coexistence .
fine examples are @xmath3 mg ( extreme quadrupole deformation ) and @xmath4s ( shape coexistence ) .
isovector quadrupole deformations are expected at the neutron drip - lines , and possible low - energy collective isovector modes have been predicted .
a very interesting phenomenon is a possible formation of deformed halo structures in weakly bound nuclei . in much heavier neutron - rich systems ,
the modification of shell structure could produce an enhancement of stability for superheavy elements @xmath5 .
proton - rich nuclei present the opportunity to study the structure of systems beyond the drip - line .
the phenomenon of ground - state proton radioactivity is determined by a delicate interplay between the coulomb and centrifugal terms of the effective potential .
proton decay rates indicate that the region @xmath6 contains strongly deformed nuclei at the drip - lines .
the lifetimes of deformed proton emitters provide direct information on the shape of the nucleus . and while an impressive amount of experimental data on deformed exotic nuclei has been published recently ,
relatively few pertinent theoretical studies have been reported .
for relatively light exotic nuclei , as for example @xmath3 mg , shell model predictions for the onset of deformation and shape coexistence have been remarkably successful .
however for medium - heavy and heavy systems , the only viable approach at present are large scale self - consistent mean - field calculations ( hartree - fock , hartree - fock - bogoliubov , relativistic mean - field ) .
the problem is of course that , in addition to the self - consistent mean - field potential , pairing correlations have to be included in order to describe ground - state properties of open - shell nuclei . and
while for strongly bound systems pairing can be included in the simple bcs scheme in the valence shell , exotic nuclei with extreme isospin values require a careful treatment of the asymptotic part of the nucleonic densities , and therefore a unified description of mean - field and pairing correlations .
the nonrelativistic hartree - fock - bogoliubov ( hfb ) and relativistic hartree - bogoliubov ( rhb ) models have been very successfully applied in the description of a variety of nuclear structure phenomena in spherical exotic nuclei on both sides of the valley of @xmath7- stability . for deformed exotic nuclei
however , so far pairing has been only described in the bcs approximation .
the bcs scheme presents only a poor approximation for nuclei with small separation energies , i.e. with the fermi level close to the particle continuum .
the inclusion of deformation in hfb or rhb models with finite range pairing still presents considerable difficulties , and it has not been possible yet to obtain reliable results in coordinate space calculations . in the present work we report on the first application of the relativistic hartree - bogoliubov theory with finite range pairing interaction to the structure of deformed nuclei in the n=28 region of neutron - rich nuclei . in the framework of the relativistic hartree - bogoliubov model , the ground state of a nucleus @xmath8 is represented by the product of independent single - quasiparticle states .
these states are eigenvectors of the generalized single - nucleon hamiltonian which contains two average potentials : the self - consistent mean - field @xmath9 which encloses all the long range particle - hole ( _ ph _ ) correlations , and a pairing field @xmath10 which sums up the particle - particle ( _ pp _ ) correlations .
the single - quasiparticle equations result from the variation of the energy functional with respect to the hermitian density matrix @xmath11 and the antisymmetric pairing tensor @xmath12 .
the relativistic model was formulated in ref .
@xcite . in the hartree approximation for the self - consistent mean field ,
the relativistic hartree - bogoliubov ( rhb ) equations read @xmath13 where @xmath14 is the nucleon mass and @xmath15 is the single - nucleon dirac hamiltonian @xcite @xmath16 the chemical potential @xmath17 has to be determined by the particle number subsidiary condition in order that the expectation value of the particle number operator in the ground state equals the number of nucleons .
the column vectors denote the quasi - particle spinors and @xmath18 are the quasi - particle energies .
the dirac hamiltonian contains the mean - field potentials of the isoscalar scalar @xmath19-meson , the isoscalar vector @xmath20-meson , the isovector vector @xmath11-meson , as well as the electrostatic potential .
the rhb equations have to be solved self - consistently with potentials determined in the mean - field approximation from solutions of klein - gordon equations .
the equation for the sigma meson contains the non - linear @xmath19 self - interaction terms . because of charge conservation only the third component of the isovector @xmath11-meson contributes .
the source terms for the klein - gordon equations are calculated in the _ no - sea _ approximation . in the present version of the model
we do not perform angular momentum or particle number projection . the pairing field @xmath21 in ( [ equ.2.2 ] )
is an integral operator with the kernel @xmath22 where @xmath23 denote quantum numbers that specify the dirac indices of the spinors , @xmath24 are matrix elements of a general relativistic two - body pairing interaction , and the pairing tensor is defined @xmath25 in the applications of the rhb theory to spherical nuclei @xcite we have used a phenomenological non relativistic interaction in the pairing channel , the pairing part of the gogny force @xmath26 with the set d1s @xcite for the parameters @xmath27 , @xmath28 , @xmath29 , @xmath30 and @xmath31 @xmath32 .
this force has been very carefully adjusted to the pairing properties of finite nuclei all over the periodic table .
in particular , the basic advantage of the gogny force is the finite range , which automatically guarantees a proper cut - off in momentum space . the fact that it is a non - relativistic interaction has negligible influence on the results of rhb calculations .
of course , in order to have a consistent formulation of the model , the matrix elements in the pairing channel @xmath33 should be derived as a one - meson exchange interaction by eliminating the mesonic degrees of freedom in the model lagrangian @xcite . however , although recently results have been reported in applications to nuclear matter , a microscopic and fully relativistic pairing force , derived starting from the lagrangian of quantum hadrodynamics , still can not be applied to nuclei . therefore also in the present calculations we use the effective gogny interaction in the pairing channel .
the eigensolutions of eq .
( [ equ.2.2 ] ) form a set of orthogonal and normalized single quasi - particle states .
the corresponding eigenvalues are the single quasi - particle energies .
the self - consistent iteration procedure is performed in the basis of quasi - particle states .
the resulting quasi - particle eigenspectrum is then transformed into the canonical basis of single - particle states , in which the rhb ground - state takes the bcs form .
the transformation determines the energies and occupation probabilities of the canonical states .
the self - consistent solution of the dirac - hartree - bogoliubov integro - differential eigenvalue equations and klein - gordon equations for the meson fields determines the nuclear ground state . for systems with spherical symmetry , i.e. single closed - shell nuclei ,
the coupled system of equations has been solved using finite element methods in coordinate space @xcite , and by expansion in a basis of spherical harmonic oscillator @xcite . for deformed nuclei the present version of the model
does not include solutions in coordinate space .
the dirac - hartree - bogoliubov equations and the equations for the meson fields are solved by expanding the nucleon spinors @xmath34 and @xmath35 , and the meson fields in terms of the eigenfunctions of a deformed axially symmetric oscillator potential @xcite . in the present calculations the number of oscillator shells in the expansion is 12 for fermion wave functions , and 20 for the meson fields .
of course for nuclei at the drip - lines , solutions in configurational representation might not provide an accurate description of properties that crucially depend on the spatial extension of nucleon densities , as for example nuclear radii . in the present study we are particularly interested in the deformed n=28 region : shell effects and shape coexistence phenomena . for a description of general trends
coordinate space solutions are not essential , solutions in the oscillator basis should present a pertinent approximation .
the region of neutron - rich @xmath36 nuclei presents many interesting phenomena : the average nucleonic potential is modified , shell effects are suppressed , large quadrupole deformations are observed as well as shape coexistence , isovector quadrupole deformations are predicted at drip - lines .
the detailed knowledge of the microscopic structure of these nuclei is also essential for a correct description of the nucleosynthesis of the heavy ca , ti and cr isotopes . in the present application of the rhb theory we study the structure of exotic neutron rich - nuclei with @xmath37 , and in particular the light @xmath1 nuclei .
we are interested in the influence of the spherical shell @xmath1 on the structure of nuclei below @xmath38ca , in deformation effects that result from the @xmath39 core breaking , and in shape coexistence phenomena predicted for these @xmath40-soft nuclei .
the structure of exotic neutron - rich nuclei with @xmath36 has been extensively studied by werner et al .
@xcite in the framework of the self - consistent mean - field theory : skyrme hartree - fock model and relativistic mean - field approach .
skyrme - hf calculations were performed by discretizing the energy functional on a three - dimensional cartesian spline collocation lattice .
the skyrme interaction siii @xcite was used . for the rmf model the dirac equation for the baryons and the klein - gordon equations for the meson fields
were solved using the basis expansion methods .
the nl - sh parameter set @xcite was used for the mean - field lagrangian . in both models the pairing correlations were included using the bcs formalism in the constant gap approximation with a strongly reduced pairing interaction .
hf and rmf results were compared with the predictions of the finite - range droplet ( frdm ) and the extended thomas - fermi with strutinsky integral ( etfsi ) mass models , as well as with available experimental data . in particular , the onset of deformations around @xmath1 , the stability of the heaviest si , s , ar and ca isotopes , and the isovector deformations were investigated .
calculations confirmed strong deformation effects caused by the the @xmath39 core breaking , especially in the rmf model .
almost all deformed nuclei were found to be @xmath40-soft , with the deformation dependence of the single - particle spectrum favoring prolate shapes for @xmath41 and oblate for @xmath42 .
the results obtained with the two models , hf and rmf , were found to be similar , although also important differences were calculated .
for example , the equilibrium shape of the @xmath1 nucleus @xmath4s .
rmf calculations predict a well deformed prolate ground state with @xmath43 for this nucleus , while according to the hf model , @xmath4s is @xmath40-soft with a small quadrupole deformation @xmath44 .
the frdm and etfsi models predict spherical and oblate ( @xmath45 ) ground states , respectively .
another detailed analysis of the ground - state properties of nuclei in the light mass region @xmath46 in the framework of the rmf approach is reported in ref .
the calculations were also performed in an axially deformed configuration using the nl - sh effective interaction . pairing was included in the bcs approximation , but a somewhat different prescription was used to calculate the neutron and proton pairing gaps in regions where nuclear masses were not known @xcite .
nuclei at the stability line were considered , as well as those close to the proton and neutron drip - lines .
results of calculations were compared with available empirical data and predictions of mass models .
a very interesting result is the predicted ground - state deformation for @xmath4s : oblate ( @xmath47 ) , with a highly prolate shape ( @xmath48 ) about 30 kev above the ground state .
this result is at variance with the calculations of refs .
@xcite , although in both cases the same effective interaction was used for the mean - field lagrangian .
the pairing gaps used in the two calculations however were different .
therefore it appears that , for nuclei which are very @xmath40-soft , the pairing interaction determines the shape transition to the deformed intruder configurations .
results of skyrme - hf and rmf calculations clearly indicate that mean - field and pairing correlations should be described in a unified self - consistent framework : the nonrelativistic hartree - fock bogoliubov theory @xcite or the relativistic hartree - bogoliubov model that was described in the previous section .
only fully self - consistent hfb or rhb models correctly describe the virtual scattering of nucleonic pairs from bound states to the positive energy particle continuum @xcite .
the correct representation of pairing correlations is an essential ingredient for microscopic models of the structure of neutron - rich nuclei .
the input parameters for our model are the coupling constants and masses for the effective mean - field lagrangian , and the effective interaction in the pairing channel . in the present calculation we use the nl3 effective interaction @xcite for the rmf lagrangian .
we have used the nl3 force in most of the applications of rhb theory to spherical nuclei : in the analysis of light neutron - rich nuclei in refs .
@xcite , in the study of ground - state properties of ni and sn isotopes @xcite , and in the description of proton - rich nuclei with @xmath49 and @xmath50 .
properties calculated with nl3 indicate that this is probably the best rmf effective interaction so far , both for nuclei at and away from the line of @xmath7-stability .
nl3 has also been used in calculations of binding energies and deformation parameters of rare - earth nuclei @xcite . for the pairing field
we employ the pairing part of the gogny interaction , with the parameter set d1s @xcite . in fig .
[ figa ] the two - neutron separation energies are plotted @xmath51 for the even - even nuclei @xmath52 and @xmath53 .
the values that correspond to the self - consistent rhb ground - states ( symbols connected by lines ) are compared with experimental data and extrapolated values from ref .
@xcite ( filled symbols ) . except for mg and si , the nuclei that we consider are not at the drip - lines .
the theoretical values reproduce in detail the experimental separation energies , except for @xmath38cr . in general
we have found that rhb model binding energies are in very good agreement with experimental data when one of the shells ( proton or neutron ) is closed , or when valence protons and neutrons occupy different major shells ( i.e. below and above @xmath54 and/or @xmath55 ) .
the differences are more pronounced when both protons and neutrons occupy the same major shell , and especially for the @xmath56 nuclei . for these nuclei
additional correlations should be taken into account , and in particular proton - neutron pairing could have a strong influence on the masses .
proton - neutron short - range correlations are not included in our model . in fig .
[ figb ] the calculated neutron radii are displayed .
the model of course predicts an increase of the calculated radii with the number of neutrons .
it is interesting however that cr , ti , ca , ar , and to a certain extent s isotopes , display very similar neutron radii in this region . only for the two more exotic chains of si and mg isotopes , i.e. more neutron - rich , a substantial increase of the calculated radii is observed . very important for our present investigation
is the effect that the @xmath57 spherical shell closure produces on the neutron radii of cr , ti and ca isotopes . on the other hand , no effect of shell closure
is observed for s , si and mg .
this is already an indication that the @xmath57 shell effects are suppressed in neutron - rich nuclei .
the predicted mass quadrupole deformations for the ground states of @xmath1 nuclei are shown in the upper panel of fig .
we notice a staggering between prolate and oblate configurations , and this indicates that the potential is @xmath40-soft .
the absolute values of the deformation decrease as we approach the @xmath55 closed shell . starting with ca ,
the @xmath1 nuclei are spherical in the ground state .
the calculated quadrupole deformations are in agreement with theoretical results reported by werner et al .
@xcite ( prolate for @xmath41 , oblate for @xmath42 ) , and with available experimental data : @xmath58 for @xmath4s @xcite , and @xmath59 for @xmath60ar @xcite . experimental data ( energies of @xmath61 states and @xmath62 values ) do not determine the sign of deformation , i.e. do not differentiate between prolate and oblate shapes . in the lower panel of fig .
[ figc ] we display the average values of the neutron pairing gaps for occupied canonical states @xmath63 } \delta_{[\omega \pi ] } v_{[\omega \pi]}^2}\over { \sum_{[\omega \pi ] } v_{[\omega \pi]}^2 } } , \label{ang}\ ] ] where @xmath64}^2 $ ] are the occupation probabilities , and @xmath65}$ ] are the diagonal matrix elements of the pairing part of the rhb single - nucleon hamiltonian in the canonical basis .
@xmath66 provides an excellent quantitative measure of pairing correlations .
the calculated values of @xmath67 mev correspond to those found in open - shell ni and sn isotopes @xcite .
the spherical shell closure @xmath1 is strongly suppressed for nuclei with @xmath68 , and only for @xmath69 neutron pairing correlations vanish .
it is also interesting to notice how in the self - consistent rhb model the calculated gaps reflect the staggering of ground - state quadrupole deformations . for @xmath4s in fig .
[ figd ] we plot the neutron canonical pairing gaps @xmath65}$ ] as function of canonical single - particle energies .
the gaps are displayed for canonical states that correspond to the self - consistent ground state .
the dashed line denotes the position of the fermi energy .
the pairing gaps are more or less constant for deeply bound states and display a sharp decrease at the fermi surface .
this is related to the volume character of the gogny interaction in the pairing channel .
the values of the @xmath65}$ ] are around 2.2 mev in the volume , and the average value at the fermi surface is around 1.8 mev . the results of skyrme+hf and rmf calculations of refs .
@xcite have shown that neutron - rich si , s and ar isotopes can be considered as @xmath40-soft , with deformations depending on subtle interplay between the deformed gaps @xmath41 and @xmath70 , and the spherical gap at @xmath1 .
because of cross - shell excitations to the @xmath71 , @xmath72 and @xmath73 shells , the @xmath1 gap appears to be broken in most cases . in the rmf analysis of ref .
@xcite a careful study of the phenomenon of shape coexistence was performed for nuclei in this region .
it was shown that several si and s isotopes exhibit shape coexistence : two minima with different deformations occur in the binding energy .
the energy difference between the two minima is of the order of few hundred kev . for @xmath4s
this difference was found to be only 30 kev . in the fully microscopic and self - consistent rhb model
, we have the possibility to analyze in detail the single - neutron levels and to study the formation of minima in the binding energy . in figs .
[ fige]-[figg ] we display the single - neutron levels in the canonical basis for the @xmath1 nuclei @xmath74si , @xmath4s , and @xmath60ar , respectively .
the single - neutron eigenstates of the density matrix result from constrained rhb calculations performed by imposing a quadratic constraint on the quadrupole moment .
the canonical states are plotted as function of the quadrupole deformation , and the dotted curve denotes the position of the fermi level . in the insert we plot the corresponding total binding energy curve as function of the quadrupole moment .
for @xmath74si the binding energy displays a deep oblate minimum ( @xmath75 ) .
the second , prolate minimum is found at an excitation energy of @xmath76 mev .
shape coexistence is more pronounced for @xmath4s .
the ground state is prolate deformed , the calculated deformation in excellent agreement with experimental data @xcite .
the oblate minimum is found only @xmath77 kev above the ground state .
finally , for the nucleus @xmath60ar we find a very flat energy surface on the oblate side .
the deformation of the ground - state oblate minimum agrees with experimental data @xcite , the spherical state is only few kev higher .
it is also interesting to observe how the spherical gap between the @xmath78 orbital and the @xmath71 , @xmath72 orbitals varies with proton number .
while the gap is really strongly reduced for @xmath74si and @xmath4s , in the @xmath42 isotone @xmath60ar the spherical gap is @xmath79 mev .
of course from @xmath38ca the @xmath1 nuclei become spherical . therefore the single - neutron canonical states in figs .
[ fige]-[figg ] clearly display the disappearance of the spherical @xmath1 shell closure for neutron - rich nuclei below @xmath42 . in order to illustrate the importance of the correct description of pairing correlations , in figs .
[ figh]-[figk ] we compare results of fully self - consistent rhb calculations with those obtained by a simplified rmf+bcs approach that was also used in refs .
binding energies , neutron and proton rms radii , and ground - state quadrupole deformations are compared for chains of mg , si , s and ar isotopes . in both models
the nl3 effective interaction has been used for the mean - field lagrangian .
the d1s gogny interaction in the pairing channel of the rhb calculation , the constant gap approximation in the bcs scheme . since for many of these nuclei the experimental odd - even mass differences are not known , the proton and neutron gaps are calculated following the prescription of m " oller and nix @xcite @xmath80 from figs . [ figh]-[figk ]
we notice that , although the calculated binding energies are almost identical with a possible exception for mg at the drip - line , the other quantities display significant differences .
rmf+bcs systematically predicts larger rms radii , especially for protons . for neutron
radii rmf+bcs calculations produce interesting kinks at @xmath1 for mg , si and ar , which suggest shell closure . in the case of ar isotopes ,
a strong shift is also calculated for the proton rms radius at n=28 .
these discontinuities are not found in rhb results , which display a uniform increase of radii with neutron number .
the two models predict similar ground state quadrupole deformations , except for the very @xmath40-soft ar isotopes .
for @xmath81ar the differences are rather pronounced , the ground - states of @xmath4ar is prolate in rmf+bcs , slightly oblate in rhb calculations . in conclusion , in the present work
we have performed a detailed analysis of the deformed @xmath1 region of neutron - rich nuclei in the framework of the relativistic hartree - bogoliubov theory with finite range pairing interaction .
this is the first application of the rhb model to the structure of deformed nuclei . in particular , we have investigated the suppression of the spherical @xmath1 shell gap for neutron - rich nuclei , and the related phenomenon of shape coexistence .
the nl3 effective interaction has been used for the mean - field lagrangian , and pairing correlations have been described by the pairing part of the finite range gogny interaction d1s .
two - neutron separation energies , rms radii , and ground - state quadrupole deformations have been compared with available experimental data and results of previous skyrme hartre - fock and relativistic mean - field + bcs calculations .
we have shown that the present version of the rhb model produces results in excellent agreement with experimental data , both for binding energies and quadrupole deformations .
our results confirm the strong deformations caused by @xmath2 neutron excitations .
constrained rhb calculations have been used to construct the binding energy curves as function of quadrupole deformation .
pronounced shape coexistence is found to be a characteristic of neutron - rich nuclei in this region .
@xmath36 nuclei become extremely @xmath40-soft just before the closure of the @xmath55 shell . for @xmath82 single neutron spectra in the canonical basis of rhb display a strong reduction of the gap between the @xmath78 orbital and the @xmath71 , @xmath72 levels .
the suppression of the spherical @xmath1 shell closure favors deformed ground states .
the importance of the full rhb description of pairing correlations has been illustrated by a comparison with ground - state properties of mg , si , s and ar isotopes calculated in the rmf+bcs scheme with constant gap approximation . this work has been supported in part by the bundesministerium fr bildung und forschung under project 06 tm 875 , and by the bulgarian national foundation for scientific research under project @xmath83 - 527 .
h. kucharek and p. ring , z. phys .
a * 339 * , 23 ( 1991 ) . p. ring , progr .
. phys . * 37 * , 193 ( 1996 ) . t. gonzalez - llarena , j.l .
egido , g.a .
lalazissis , and p. ring , phys .
b * 379 * , 13 ( 1996 ) .
w. pschl , d. vretenar , g.a .
lalazissis , and p. ring , phys .
lett . * 79 * , 3841 ( 1997 ) .
lalazissis , d. vretenar , w. pschl , and p. ring , nucl . phys . *
a632 * , 363 ( 1998 ) .
lalazissis , d. vretenar , w. pschl , and p. ring , phys . lett .
* b418 * , 7 ( 1998 ) .
d. vretenar , w. pschl , g.a .
lalazissis , and p. ring , phys .
c * 57 * , r1060 ( 1998 ) .
lalazissis , d. vretenar , and p. ring , phys .
c * 57 * , 2294 ( 1998 ) .
d. vretenar , g.a .
lalazissis , and p. ring , phys .
c * 57 * , 3071 ( 1998 ) .
j. f. berger , m. girod and d. gogny ; nucl .
a428 * , 32 ( 1984 ) .
gambhir , p. ring , and a. thimet , ann .
( n.y . ) , * 198 * , 132 ( 1990 )
. t. r. werner , j. a. sheikh , w. nazarewicz , m. r. strayer , a. s. umar and m. misu , phys . lett . * b335 * , 259 ( 1994 ) . t. r. werner , j. a. sheikh , m .. misu , w. nazarewicz , j. rikovska , k. heeger , a. s. umar and m. r. strayer , nucl . phys . * a597 * , 327 ( 1996 ) .
m. beiner , h. flocard , n. van giai and p. quentin , nucl . phys . *
a238 * , 29 ( 1975 ) .
sharma , m.a .
nagarajan , and p. ring ; phys . lett . *
b312 * 377 ( 1993 ) .
g. a. lalazissis , a. r. farhan and m. m. sharma , nucl .
* a628 * , 221 ( 1998 ) .
p. m " oller and j. r. nix , nucl . phys . *
a536 * , 20 ( 1992 ) . p. ring , and p. schuck , the nuclear many - body problem ( springer - verlag , heidelberg 1980 )
. j. dobaczewski , w. nazarewicz , t. r. werner , j. f. berger , c. r. chinn , and j.decharg ' e , phys .
c * 53 * , 2809 ( 1996 ) .
g. a. lalazissis , j. knig and p. ring ; phys . rev . *
c55 * , 540 ( 1997 ) .
lalazissis and p. ring , phys .
b * 427 * , ( 1998 ) 225 .
g. audi and a. h. wapstra , nucl . phys . *
a595 * , 409 ( 1995 ) .
t. glasmacher et al .
, phys . lett . *
b395 * , 163 ( 1997 ) .
t. glasmacher , nucl . phys . *
a630 * , 278c ( 1998 ) .
h. scheit et al . , phys .
lett . * 77 * , 3967 ( 1996 ) . | ground - state properties of neutron - rich @xmath0 nuclei are described in the framework of relativistic hartree bogoliubov ( rhb ) theory .
the model uses the nl3 effective interaction in the mean - field lagrangian , and describes pairing correlations by the pairing part of the finite range gogny interaction d1s .
two - neutron separation energies and ground - state quadrupole deformations that result from fully self - consistent rhb solutions , are compared with available experimental data .
the model predicts a strong suppression of the spherical @xmath1 shell gap for neutron - rich nuclei : the @xmath2 core breaking results in deformed ground states .
shape coexistence is expected for neutron - rich si , s and ar isotopes .
= 16 true cm = 23.5 true cm = cmbx10 scaled2 |
today , several different types of isolated neutron stars which have comparably very different physical properties are known . these are radio pulsars , dim radio - quiet neutron stars ( drqnss ) , dim thermal neutron stars ( dtnss ) , anomalous x - ray pulsars ( axps ) and soft gamma - ray repeaters ( sgrs ) .
about 1700 radio pulsars have been observed up to date @xmath1 , some of which have also been detected in other bands of the electromagnetic spectrum .
radio pulsar is relatively the most well known and understood object both theoretically and observationally among all the classes of isolated neutron stars .
drqnss , dtnss and axps - sgrs are new types of isolated neutron stars ( point x - ray sources ) .
all axps - sgrs , which should be considered as a single class , and some drqnss and dtnss , which are cooling neutron stars with short ( p@xmath2few 100 milliseconds ) and long ( p@xmath2few seconds ) rotation periods respectively , show themselves as x - ray pulsars .
axps and sgrs ( which have also long rotation periods similar to dtnss and which experience x - ray and @xmath3-ray bursts ) may be in early phases of at least some of the dtnss based on field decay models ( see fig.1 ) .
recently , ankay et al .
@xmath4 have shown the possibility that radio - quiet strong x - ray pulsar j1846 - 0258 may be in a phase preceding axp - sgr and/or dtns phases based on a detailed comparison between the observational characteristics of these sources .
none of the drqnss , dtnss or axps / sgrs has been detected at radio frequencies with the exception axp xte j1810 - 197 from the direction of which radio emission at 1.4 ghz has been detected @xmath5 , but this radio emission may also be produced by a possible pulsar wind nebula ( pwn ) around this axp .
there are only a few observationally known drqnss .
all of them are physically connected to galactic supernova remnants ( snrs ) which have ages less than about 20 kyr @xmath6 .
there are two basic observable quantities of isolated pulsars : the rotation period p and the time rate of change of the rotation period p. time rate of change of the rotational kinetic energy of a rigid body is @xmath7 where i is the moment of inertia and @xmath8=2@xmath9/p is the rotational velocity .
the component of the dipole magnetic field of pulsar which is perpendicular to the rotation axis also depends on p and p : @xmath10 where c is the speed of light in free space and r is the radius of pulsar .
if we assume that the time rate of change of the rotational velocity of pulsar can be expressed as a power law : @xmath11 where k is a proportionality constant and the power n is called the braking index , then the real age of pulsar can be represented in terms of p and p as : @xmath12\ ] ] where p@xmath13 is the initial rotation period of pulsar .
if p@xmath13 is much less than p : @xmath14 magneto - dipole radiative power of pulsars is @xmath15 where b@xmath16 is the strength of the dipole magnetic field at the magnetic pole and
@xmath17 is the angle between the rotation axis and the magnetic axis @xmath18 .
if ( eqn.(1 ) ) is equal to l ( eqn.(6 ) ) ( i.e. if the net torque on pulsar is equal to the magneto - dipole radiation torque ) , then the braking index turns out to be n=3 ( assuming also that eqn.(3 ) can be applied to express the rotational evolution of pulsar ) . in such a case , the characteristic age ( @xmath0 ) of pulsar is defined using the n=3 condition in eqn.(5 ) : @xmath19 so , when n=3 and p@xmath13@xmath20p , @xmath0 is approximately equal to the real age of pulsar . when n is greater ( less ) than 3 , @xmath0 is greater ( less ) than the real age .
so , if there are extra torques on pulsar which spin it down in addition to the effect of the magneto - dipole radiation torque , the pulsar will evolve with larger p values compared to the case of pure magneto - dipole radiation torque . on the other hand ,
if there is b - decay , the p values the pulsar has throughout the evolution will be smaller .
note that the braking index n should be adapted as the average braking index ( @xmath21 ) of pulsar when considering long time intervals in pulsar s lifetime , because the value of instantaneous ( i.e. considering very short time intervals compared to the lifetime of pulsar ) braking index may change in time in a complicated way in general .
although , none of the drqnss has been detected at radio frequencies , this does not necessarily mean that they have no radio emission .
there are basically two selection effects ( other than the background radiation which is effective only on the sources located in the galactic central directions , see ankay et al .
@xmath22 ) which can prevent detection of the radio emission : the beaming fraction and the luminosity function . based on radio pulsar observations , there exists a strong evidence that beaming fraction decreases in time possibly because of the angle between the rotation axis and the magnetic axis ( @xmath17 ) decreasing during the evolution and the large number of the observed radio pulsars with possibly small values of @xmath17 supports this interpretation @xmath23 .
indeed , the existence of b - decay ( where b is the component of the dipole magnetic field perpendicular to the rotation axis ) for a large sample of radio pulsars was shown by guseinov et al .
@xmath24 by comparing the characteristic ages of radio pulsars with their kinematic ages ( i.e. their distances from the galactic plane ) .
guseinov et al .
@xmath24 assumed an exponential decay with a characteristic decay time @xmath25=3@xmath2610@xmath18 yr .
the cause of b - decay in the case of radio pulsars can be a temporal decrease in @xmath17 , but the possibility of a decay in the dipole magnetic field itself can not be totally excluded ( see e.g geppert & rheinhardt @xmath27 on the possibility of magnetic field decay in neutron stars ) . the other selection effect is related to radio luminosity versus number distribution of radio pulsars ( luminosity function ) . based on the observational data ,
most of the radio pulsars must have low radio luminosity at birth and the radio luminosity , which is only a small fraction of the magneto - dipole radiation produced by the pulsar , does not change significantly in time @xmath28 .
as mentioned above , all the known drqnss are connected to galactic snrs with ages @xmath292@xmath2610@xmath30 yr .
most of the galactic snrs are shell - type @xmath31 and many of these snrs lack detected point sources or pwne in them .
direct detection of neutron stars in many of the snrs may not be possible because of the selection effects and the large distances .
on the other hand , pwn is not seen around pulsars which have rate of rotational energy loss @xmath3210@xmath33 erg / s @xmath34 .
the only known drqns with measured p and p values is 1e 1207.4 - 5209 and it has =2@xmath2610@xmath35 erg / s .
this explains why there is no pwn around 1e 1207.4 - 5209 . for 1e 1207.4 - 5209
, there are 2 problems based on its position on the p - p diagram ( fig.1 ) : first of all , its present b value is much less than the magnetic field measured from its 0.7 kev proton cyclotron line detected by both _
chandra _ and _ xmm - newton_. secondly , the @xmath0 value of this drqns is much larger than the age of the snr which it is physically connected to .
below , we will present a b - decay model in order to clarify these problems and to explain the possible evolution of 1e 1207.4 - 5209 . in section 2 , the observational data on 1e 1207.4 - 5209
are represented . in section 3 ,
an exponential b - decay model is introduced as a possible evolutionary scenario for this drqns . in the last section ,
discussions and conclusions are represented .
drqns 1e 1207.4 - 5209 is the only known isolated neutron star which have x - ray absorption lines clearly observed in its spectrum .
the possible proton cyclotron line ( 0.7 kev ) of 1e 1207.4 - 5209 and its harmonics at 1.4 and 2.1 kev ( and possibly at 2.8 kev ) detected by _ xmm - newton _ ( 260 ks observation @xmath36 ) corresponds to a magnetic field @xmath21.6@xmath2610@xmath37 g @xmath36 .
sanwal et al .
@xmath38 also pointed out the possibility that the 0.7 and 1.4 kev lines ( _ chandra _ observations @xmath260 ks ) could be formed by transitions of singly ionized helium in a magnetic field @xmath2(1.4 - 1.7)@xmath2610@xmath37 g. we will adapt b@xmath39=1.6@xmath2610@xmath37 g as the present surface dipole magnetic field of 1e 1207.4 - 5209 .
note that , if the cyclotron line is interpreted in terms of electrons , the corresponding magnetic field value ( @xmath28@xmath2610@xmath27 g @xmath36 ) is much smaller than its present b value .
assuming conventional values of mass and radius for 1e 1207.4 - 5209 and using its measured p=424 ms and [email protected]@xmath2610@xmath41 ss@xmath42 values , one finds [email protected]@xmath2610@xmath43 g. this leads to a very small angle between the magnetic axis and the rotation axis : @[email protected] . on the other hand , snr g296.5 + 10.0 , which is physically connected to 1e 1207.4 - 5209 , has age t@xmath46 10 kyr @xmath47 and the characteristic age of 1e 1207 - 5209 @xmath48480 kyr .
the difference between t@xmath49 and @xmath0 leads to a very large average braking index unless the initial rotation period ( p@xmath13 ) of the pulsar is comparable to its present p value .
if 1e 1207.4 - 5209 has evolved with n=3 since it was born , it should have p@xmath50420 ms ( see eqn.4 ) .
a neutron star born with such a large rotation period is a significant problem for theories on core - collapse supernovae and formation of pulsars ( see ardeljan et al .
@xmath51 , moiseenko et al .
@xmath52 and heger et al .
@xmath53 on the necessity of rapid rotation for newborn neutron stars ) , though this possibility may not be totally excluded .
we can explain the huge difference between t@xmath49 and @xmath0 and the large discrepancy between b@xmath39 and b by a simple b - decay model .
we will assume an exponential decay similar to guseinov et al .
@xmath24 who showed the existence of b - decay for a large sample of radio pulsars by comparing the kinematic ages with the characteristic ages .
assuming conventional values of mass and radius for the pulsar and using eqn.(2 ) , the exponential b - decay can be written as : @xmath54 for t@xmath55@xmath25 .
here , p and p are the present values of the pulsar , b@xmath13 is the initial value of the magnetic field component perpendicular to the rotation axis , @xmath25 is the characteristic decay time , and t is the real age of pulsar . if the initial angle @xmath56=90@xmath45 ( i.e. if b@xmath13=b@xmath39 assuming no field decay ) , then @xmath572.4@xmath2610@xmath5 yr for t = t@xmath4610@xmath30 yr .
we can also put an upper limit on @xmath25 : if the pulsar had evolved with n=3 , it should have had [email protected]@xmath2610@xmath58 g ( assuming that the pulsar s age is 10@xmath30 yr and p@xmath59p , see fig.1 ) .
using b@xmath13=2@xmath2610@xmath58 g in eqn.(8 ) gives @xmath574.8@xmath2610@xmath5 yr .
so , @xmath57(2.4 - 4.8)@xmath2610@xmath5 yr corresponding to b@xmath50(16 - 2)@xmath2610@xmath58 g. in this analysis , we assume that a possible evolution with @xmath603 when the pulsar is very young compared to its present age will have a negligible influence on its evolution on the p - p diagram . in the b - decay model introduced above
, the characteristic age can be written as a function of @xmath25 and t as : @xmath61 with the condition t@xmath55@xmath25 .
using eqns .
( 7),(8 ) and ( 9 ) , we have found the positions on the p - p diagram ( shown by light squares in fig.1 ) corresponding to different evolutionary tracks based on the b - decay model . in fig.1 , the points ( light squares ) a ( t=5 kyr ) , b ( t=10 kyr ) and c ( t=15 kyr ) are on the evolutionary track with b@xmath13=2@xmath2610@xmath58 g and @xmath25=2.5 kyr .
the points d ( t=10 kyr ) , e ( t=15 kyr ) and f ( t=20 kyr ) correspond to the evolutionary track with b@xmath13=2@xmath2610@xmath58 g and @xmath25=5 kyr .
the points g ( t=5 kyr ) , h ( t=10 kyr ) , i ( t=15 kyr ) and j ( t=10 kyr ) , k ( t=15 kyr ) , l ( t=20 kyr ) are on the evolutionary tracks with b@xmath13=4@xmath2610@xmath58 g , @xmath25=2.5 kyr and b@xmath13=4@xmath2610@xmath58 g , @xmath25=5 kyr , respectively .
the present position of 1e 1207.4 - 5209 on the p - p diagram is in between the last two tracks ( b@xmath13=4@xmath2610@xmath58 g , @xmath25=2.5 kyr and b@xmath13=4@xmath2610@xmath58 g , @xmath25=5 kyr , see fig.1 ) .
we can conclude that b@xmath504@xmath2610@xmath58 g and @xmath573.75 kyr for this drqns in the b - decay model .
the corresponding initial value of the angle between the rotation axis and the magnetic field axis is @[email protected] .
if the present surface dipole magnetic field of 1e 1207.4 - 5209 is about 1.6@xmath2610@xmath37 g based on resonant proton cyclotron scattering measurements and since b@xmath63=2.5@xmath2610@xmath43 g , there may be a decay in the angle @xmath17 between the rotation and the magnetic axes ( but not a decay in the magnetic field itself ) considering the huge difference between the characteristic age of 1e 1207.4 - 5209 and the age of its snr .
this can be explained by a simple exponential b - decay model .
characteristic time of the exponential decay must be about 2.5 - 5 kyr depending on the initial value of b which must be in the interval b@xmath13=(16 - 2)@xmath2610@xmath58 g. such a rapid decrease in @xmath17 leads to a short lifetime ( possibly about ( 1 - 2)@xmath2610@xmath30 yr taking into consideration the ages of the snrs physically connected to drqnss ) for this type of neutron star as an x - ray pulsar and this may also be the reason for the lack of detected radio emission from such neutron stars as the radio beam width can be narrower compared to the x - ray beam width .
after about ( 1 - 2)@xmath2610@xmath30 yr pass since the birth of such a pulsar , it can be detected as a cooling neutron star ( but not as an x - ray or radio pulsar ) if it is located close enough to the sun .
this result can be used to explain the lack of detected point sources in at least some of the galactic snrs ( especially the distant ones ) as the lifetime of an snr can be as long as 10@xmath64 yr . the huge difference between the characteristic age of 1207.4 - 5209 and the age of its snr may in principle be explained by assuming that this drqns was born with an initial rotation period very close to its present rotation period .
but such a long initial rotation period ( p@xmath13@xmath40420 ms ) contradicts the estimations based on existing theories on formation of neutron stars by core - collapse supernovae ( which predict one order of magnitude smaller p@xmath13 values ) . on the other hand ,
the huge difference between the surface dipole magnetic field ( about 1.6@xmath2610@xmath37 g obtained from the cyclotron line measurement ) of 1e 1207.4 - 5209 and its b value ( about 2.5@xmath2610@xmath43 g ) can not be explained by assuming a constant and very small initial angle ( @xmath56@[email protected] ) between the rotation and magnetic axes , if the initial rotation period is much less than the present rotation period ( i.e. if the average braking index is much greater than 3 , see eqn.4 and eqn.5 ) .
there may be basically two different magnetic fields in neutron stars : the magnetic field formed by superconducting entrainment currents which is parallel to the rotation axis and the relic magnetic field ( which may be homogeneous inside and in the form of a dipole field on the surface of the neutron star ) which is formed when the progenitor star collapses leading to supernova explosion @xmath65 .
the relic magnetic field axis can have an arbitrary angle ( @xmath17 ) with respect to the rotation axis and the resultant magnetic field will be a superposition of the two magnetic fields . as the axis of the relic magnetic field approaches the rotation axis , the magnitude of the resultant surface magnetic field ( i.e. the total magnetic field on the surface which is the superposition of the relic field and the field formed by the superconducting entrainment currents ) increases and its component which is perpendicular to the rotation axis decreases in time . since 1e 1207.4 - 5209
has most probably very high b@xmath39 and comparably very low b ( perpendicular component of the resultant surface magnetic field ) and as this neutron star is observed as a pulsar , b@xmath39 must be comparable to the resultant surface magnetic field which is increasing in time ( or to the relic magnetic field as the magnetic field formed by the entrainment currents must be on the order of 10@xmath43 g on average @xmath65 ) .
on the other hand , the actual low b value of 1e 1207.4 - 5209 as compared to the case of n=3 ( see fig.1 ) can be explained by the exponential decay as this perpendicular component decreases in time based on the discrepancy between the age values .
the increase in the magnitude of the resultant surface magnetic field of 1e 1207.4 - 5209 must be very small if the relic field is actually much higher than the entrainment field , so that the b@xmath39 and hence the magnitude of the present resultant surface magnetic field must be comparable to the magnitude of the initial resultant surface magnetic field
. on the other hand , the decrease in the perpendicular component of the resultant surface magnetic field must be very large ( about one order of magnitude ) in a very short time interval ( @xmath66 yr ) compared to radio pulsar lifetimes . in the case of radio pulsars , the increase in the magnitude of the resultant surface magnetic field must be relatively large ( up to a factor of about 1.5 ) based on the known b values of radio pulsars and the prediction on the average strength of the entrainment field , and the decrease in the perpendicular component of the field must exist on a much longer timescale .
comparing the magnetic field values of 1e 1207.4 - 5209 with the b values of most of the radio pulsars together with the age discrepancy for 1e 1207.4 - 5209 snr g296.5 + 10.0 pair ( which also exists for several other pulsar
snr pair but without any clearly detected cyclotron line ) may be an evidence for the existence of two different magnetic fields for neutron stars ; the relic field and the entrainment field .
the interaction between the superconducting entrainment currents and the relic field may be the reason for the decrease in @xmath17 .
such a possible interaction between the two fields throughout the evolution of neutron stars with different initial intrinsic physical conditions has yet to be examined .
the rate of temporal decrease in @xmath17 ( and hence the evolution of the perpendicular component of the total surface dipole magnetic field ) may depend on both the relic and hence the resultant magnetic field , the mass and the equation of state of the neutron star ( it may also depend on the initial value of @xmath17 ) . the changes in p observed for 1e 1207.4 - 5209 may also be explained by oscillations in the magnetic dipole axis as it approaches the rotation axis .
such changes in p must also exist in other 1207-like isolated radio - quiet x - ray pulsars . as a comparison
please note that the magnetic field values of accreting x - ray pulsars in x - ray binaries obtained from cyclotron line measurements are on the order of 10@xmath43 g @xmath67 , that is comparable to the predicted surface magnetic fields due to the entrainment currents @xmath65 , and the b values of recycled millisecond pulsars are several orders of magnitude smaller ( see bisnovatyi - kogan & komberg @xmath68 who gave the first reliable explanation for the magnetic field decay in x - ray binaries by accretion and the formation of recycled millisecond pulsars ) .
the surface magnetic field in the case of x - ray binaries most probably decreases because of the plasma falling upon the surface of the neutron star during the accretion process @xmath69 .
a decrease in the resultant surface field because of accretion and maybe also a decrease in the value of @xmath17 ( but on a much longer timescale compared to the 1e 1207.4 - 5209 case ) may explain the measured conventional 10@xmath43 g magnetic field values of x - ray pulsars in binary systems and the very low perpendicular components of the resultant surface field of recycled millisecond pulsars . as a last remark we would like to note that the birth rate of 1207-like x - ray pulsars must be about 20 - 30% of the total supernova rate ( excluding type - ia supernovae as this type of explosion most probably does not lead to formation of a neutron star ) taking into consideration that the lifetime of such sources as x - ray pulsars must be about 2@xmath2610@xmath30 yr because of rapid b - decay ( see guseinov et al .
@xmath64 on the birth rates of different types of isolated neutron stars including dim radio quiet neutron stars some of which seem to be 1207-like objects based on the existing observational data ) .
* acknowledgments * we would like to thank the anonymous referee for his / her constructive and helpful contributions .
this work is supported by boazii university .
\1 . atnf australia telescope national facility pulsar catalogue ( 2006 ) , + http://www.atnf.csiro.au / research / pulsar / psrcat/. + 2 . a. ankay , s. sahin , g. karanfil and e. yazgan , int . j. mod
d * 14 * , 1075 ( 2005 ) .
j. p. halpern , e. v. gotthelf , r. h. becker , d. j. helfand and r. l. white , astrophys .
j. * 632 * , l29 ( 2005 ) .
o. h. guseinov , a. ankay and s. o. tagieva , astrophys .
space sci . * 289 * , 23 ( 2004 ) .
o. h. guseinov , a. ankay and s. o. tagieva , int . j. mod .
phys d * 14 * , 643 ( 2005 ) .
v. m. lipunov , astrophysics of neutron stars , springer ( 1992 ) .
a. ankay , o.h .
guseinov and s.o .
tagieva , astronomical and astrophysical transactions * 23 * , 503 ( 2004 ) .
t. m. tauris and r. n. manchester , mon . not .
. soc . * 298 * , 625 ( 1998 ) .
o. h. guseinov , a. ankay and s. o. tagieva , int . j. mod
d * 13 * , 1805 ( 2004 ) .
u. geppert and m. rheinhardt , astronomy and astrophysics * 392 * , 1015 ( 2002 ) .
o. h. guseinov , e. yazgan , s. o. tagieva and s. ozkan , rev .
. astron . astrof . * 39 * , 267 ( 2003 ) .
d. a. green , a catalogue of galactic supernova remnants ( 2006 april version ) , http://www.mrao.cam.ac.uk / surveys / snrs/. + 13 .
o. h. guseinov , a. ankay and s. o. tagieva , serb .
j. * 167 * , 93 ( 2003 ) .
o. h. guseinov , a. ankay and s. o. tagieva , serb .
j. * 168 * , 55 ( 2004 ) .
o. h. guseinov , a. ankay and s. o. tagieva , serb .
j. * 169 * , 65 ( 2004 ) .
. o. h. guseinov , a. ankay , s. o. tagieva and m. o. taskin , int .
d * 13 * , 197 ) ( 2004 ) .
o. h. guseinov , a. ankay and s. o. tagieva , astro - ph/0401114 ( 2004 ) .
a. de luca , p. caraveo , s. mereghetti , m. moroni , g. f. bignami and r. mignani , young neutron stars and their environments , iau symposium no.218 , edited by fernando camilo and bryan m. gaensler , p.273 ( 2004 ) .
a. de luca , s. mereghetti , p. caraveo , m. moroni , r. mignani and g. f. bignami , astron .
astrophys . * 418 * , 625 ( 2004 ) .
d. sanwal , g. g. pavlov , v. e. zavlin and m. a. teter , astrophys .
j. * 574 * , l61 ( 2002 ) .
n. v. ardeljan , g. s. bisnovatyi - kogan and s. g. moiseenko , mon . not .
359 * , 333 ( 2005 ) .
s. g. moiseenko , g. s. bisnovatyi - kogan , n. v. ardeljan , mon . not .
. soc . * 370 * , 501 ( 2006 ) .
a. heger , s. e. woosley , n. langer and h. c. spruit , stellar rotation , proceedings of iau symposium no.215 , edi . by andre maeder and philippe eenens , p.591 ( 2004 ) .
a. heger , s. e. woosley and h. c. spruit , astrophys .
j. * 626 * , 350 ( 2005 ) .
d. m. sedrakian and d. blaschke , astrophysics * 45 * , 166 ( 2002 ) .
d. m. sedrakian and k. m. shahabasyan , `` the magnetic field of pulsars '' in neutron stars , supernovae and supernova remnants , edi . by efe yazgan , askin ankay and oktay h. guseinov , nova science publishers ( to be published in 2006 ) .
w. coburn , w. a. heindl , r. e. rothschild , d. e. gruber , i. kreykenbohm , j. wilms , p. kretschmar and r. staubert , astrophys .
j. * 580 * , 394 ( 2002 ) .
g. s. bisnovatyi - kogan and b. v. komberg , sov .
astron . * 18 * , 217 ( 1974 ) .
g. s. bisnovatyi - kogan and b. v. komberg , sov .
astron . lett .
* 2 * , 130 ( 1976 ) .
+ 30 . g. s. bisnovatyi - kogan and a. m. fridman , astron .
zh . * 46 * , 721 ( 1969 ) .
p. r. amnuel and o. h. guseinov , astron .
no.524 ( 1969 ) .
g. s. bisnovatyi - kogan , `` evolutions of neutron stars and their magnetic fields '' in neutron stars , supernovae and supernova remnants , edi .
by efe yazgan , askin ankay and oktay h. guseinov , nova science publishers ( to be published in 2006 ) .
rotation period ( p ) versus temporal change of rotation period ( p ) diagram of various types of isolated pulsars .
small dots represent radio pulsars .
symbols cross and star denote axps - sgrs and dtnss , respectively ( upper limits on p for 3 dtnss are shown by arbitrary arrows ) . the actual position of 1e 1207.4 - 5209
is shown by a dark square and its position for n=3 case ( see text ) is displayed by a plus sign .
constant lines of b , and @xmath0 are denoted by b11-b15 , e29-e41 and t3-t9 , respectively .
some points on the evolutionary tracks found from the exponential b - decay model ( see text ) are displayed as light squares : a , b , c ( b@xmath13=2@xmath2610@xmath58 g , @xmath25=2.5 kyr ) ; d , e , f ( b@xmath13=2@xmath2610@xmath58 g , @xmath25=5 kyr ) ; g , h , i ( b@xmath13=4@xmath2610@xmath58 g , @xmath25=2.5 kyr ) ; j , k , l ( b@xmath13=4@xmath2610@xmath58 g , @xmath25=5 kyr ) . | dim radio - quiet neutron star ( drqns ) 1e 1207.4 - 5209 is one of the most heavily examined isolated neutron stars .
wide absorption lines were observed in its spectrum obtained by both xmm - newton and chandra x - ray satellites .
these absorption lines can be interpreted as a principal frequency centered at 0.7 kev and its harmonics at 1.4 , 2.1 and possibly 2.8 kev .
the principal line can be formed by resonant proton cyclotron scattering leading to a magnetic field which is two orders of magnitude larger than the perpendicular component of the surface dipole magnetic field ( b ) found from the rotation period ( p ) and the time rate of change in the rotation period ( p ) of 1e 1207.4 - 5209 . besides , age of the supernova remnant ( snr )
g296.5 + 10.0 which is physically connected to 1e 1207.4 - 5209 is two orders of magnitude smaller than the characteristic age ( @xmath0=p/2p ) of the neutron star .
these huge differences between the magnetic field values and the ages can be explained based on a b - decay model .
if the decay is assumed to be exponential , the characteristic decay time turns out to be several thousand years which is three orders of magnitude smaller than the characteristic decay time of radio pulsars represented in an earlier work .
the lack of detection of radio emission from drqnss and the lack of point sources and pulsar wind nebulae in most of the observed snrs can also be partly explained by such a very rapid exponential decay .
the large difference between the characteristic decay times of drqnss and radio pulsars must be related to the differences in the magnetic fields , equation of states and masses of these isolated neutron stars .
+ + keywords : neutron star ; pulsar : 1e 1207.4 - 5209 ; evolution . |
white dwarfs ( wd ) are the typical endpoint of the evolution of most main sequence stars .
because nuclear reactions do not occur in their deep interiors , the evolution of wds can be considered as a relatively simple and well understood gravothermal cooling process @xcite .
actually , the evolutionary cooling times are now very accurate ( e.g. * ? ? ?
* ) , providing a reliable way of measuring the wd cooling age from the temperature and surface gravity measured observationally .
wds are hence very useful tools in astronomy .
for example , the wd luminosity function ( lf ) is an important statistical instrument which has been used not only to derive an accurate age of the galactic disk in the solar neighborhood @xcite but also to constrain the local star formation rate @xcite .
moreover , the wd mass function ( mf ) has been successfully employed over the years as a tool to test the theory of stellar evolution , offering information on stellar mass loss .
this function has also been used to study how the evolution of close binaries is able to produce low - mass wds @xcite , thus helping to asses the contribution of the distinct evolutionary scenarios in producing the current population of galactic binaries in which one of the components is a wd .
finally , the wd age function ( af ) is also a valuable tool for constraining the wd formation history @xcite . with the advent of modern , large - scale surveys such as the sloan digital sky survey ( sdss ; * ?
* ) or the super - cosmos sky survey @xcite , the size of the current observational wd samples has increased dramatically .
this has allowed producing more accurate lfs , mfs and afs @xcite . however , the major drawback of most studies is the complicated target selection algorithms , which incorporate observational biases that are almost impossible to quantify in numerical simulations that aim at reproducing the ensemble properties of the observed samples .
significant observational efforts have also allowed to unveil the population of wds within 20pc of the sun @xcite .
this sample is considerably less numerous than those mentioned above , however it is claimed to be reasonable complete and can therefore be considered as a volume ( rather than magnitude ) limited sample , which suffers effectively from no selection biases .
the lamost ( large sky area multi - object fiber spectroscopic telescope ) spectroscopic survey of the galactic anti - center ( lss - gac ; * ? ? ?
* ; * ? ? ?
* ) follows a well - defined selection criteria aiming at providing spectra for stellar sources of all colours in the galactic anti - center ( including wds ) so that they can be studied in a statistically meaningful way .
lss - gac started operations in 2011 and will provide a significantly larger sample of wds than those within 20pc of the sun . in this paper
we derive preliminary observed lf , mf and af of wds identified within the first data release of lss - gac , and use a state - of - the - art monte carlo population synthesis code adapted to the characteristics of the survey to simulate the wd population in the galactic anti - center .
we apply the lss - gac selection criteria to the simulated samples , carefully evaluate all possible observational biases , and derive synthetic lfs , mfs and afs .
this exercise allows us to perform a meaningful comparison between the outcome of simulations and the observational data .
lamost is a quasi - meridian reflecting schmidt telescope located at xinglong observing station in the hebei province of china @xcite .
the effective aperture of lamost is about 4 meters .
lamost is exclusively dedicated to obtain optical spectroscopy of celestial objects .
each `` spectral plate '' refers to a focal surface with 4,000 auto - positioned optical fibers to observe spectroscopic plus calibration targets simultaneously , equally distributed among 16 fiber - fed spectrographs .
each spectrograph is equipped with two ccd cameras of blue and red channels that simultaneously provide blue and red spectra of the 4,000 selected targets , respectively .
the lss - gac is a major component of the lamost galactic survey @xcite . by selecting targets uniformly and randomly in ( @xmath4 , @xmath5 ) and ( @xmath4 , @xmath6 ) hess diagrams @xcite ,
the lss - gac main survey aims at collecting @xmath73,700 9,000 low resolution ( @xmath8 ) spectra for a statistically complete sample of @xmath9million stars of all colours down to a limiting magnitude of @xmath1017.8mag ( 18.5mag for limited fields ) , distributed in a contiguous sky area of over 3,400 deg@xmath11 centered on the galactic anti - center ( @xmath12 , @xmath13 ) .
the simple yet non - trivial target selection of the survey makes possible to study the underlying stellar populations for any given type of target , such as wds .
in addition to the main survey , the lss - gac also includes a survey of the m31/m33 area , targeting hundreds of thousands of objects in the vicinity fields of m31 and m33 , and a survey of very bright ( vb ) plates , targeting over a million of randomly selected very bright stars ( @xmath14mag ) in the northern hemisphere during bright / grey nights .
targets for the lss - gac main survey are selected for spectroscopic follow - up from the xuyi schmidt telescope photometric survey of the galactic anti - center ( xstps - gac ) catalogue @xcite .
the xstps - gac survey was initiated in october 2009 and completed in march 2011 .
photometry in sdss @xmath15 , @xmath4 and @xmath16 bands was acquired with the xuyi 1.04/1.20 m schmidt telescope located at the xuyi station of the purple mountain observatory , p.r.china .
the xstps - gac surveyed an area of about 5,400deg@xmath11 , from @xmath17 to 9@xmath18 and @xmath19 to + 60@xmath20 to cover the lss - gac main survey footprint , plus an extension of @xmath21deg@xmath11 to the m31 , m33 region to cover the lss - gac m31/m33 survey area .
the resulting catalogues contain about 100 million stars down to a limiting magnitude of @xmath22mag ( @xmath23 ) , with an astrometric calibration accuracy of 0.1arcsec @xcite and a global photometric calibration accuracy of @xmath242 per cent ( * ? ? ?
* yuan et al . , in prep . ) .
for the lss - gac main survey , bright ( b ) , medium - bright ( m ) and faint ( f ) plates are designated to target sources of brightness 14.0@xmath25 16.3mag , 16.3 @xmath2617.8mag and 17.8 @xmath27mag , respectively . during the pilot ( oct .
2011 jun . 2012 ) and the first year regular ( oct .
2012 jun .
2013 ) surveys of lamost , a total of 1,042,586 [ 750,867 ] spectra of a signal to noise ratio ( s / n ) @xmath28 10 at 7450 [ s / n @xmath28 10 at 4650 ] have been collected , including 439,560 [ 225,522 ] spectra from the main survey .
most of the stars are from the b plates .
the raw data were reduced with the lamost 2d pipeline ( version 2.6 ) ( e.g. * ? ? ?
* ) , following the standard procedures of bias subtraction , cosmic - ray removal , 1d spectral extraction , flat - fielding , wavelength calibration , and sky subtraction .
the lamost spectra are recorded in two arms , 3,700
5,900 in the blue and 5,700 9,000 in the red .
the blue- and red - arm spectra are processed independently in the 2d pipeline and joined together after flux calibration .
flux calibration for the lss - gac observations is performed using an iterative algorithm developed at peking university by @xcite , achieving an accuracy of about 10 per cent for the whole wavelength ranges of blue- and red - arm spectra . a detailed description of the target selection algorithm , survey design , observations , data reduction , and value - added catalogs of the first lss - gac data release is presented in @xcite . in the current work we only consider spectra obtained by the main lss - gac survey . a list including the names of the plates employed for the main survey is provided in tablea1 ( see appendix ) .
in this section we describe our methodology for identifying wds within the lss - gac spectroscopic data base .
we do this following two independent but complementary routines .
the first identifies wds by @xmath29-template fitting all lss - gac spectra , the second selects wds by applying a well - defined colour cut to the xstps - gac photometric catalogue .
we also estimate the spectroscopic completeness of the lss - gac wd sample .
we use the @xmath29-template fitting method described by @xcite to identify lss - gac wds .
this routine was originally developed to search for sdss wd - main sequence binaries ( wdms ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
however its mathematical prescription can be easily implemented to the lss - gac data for identifying single wds . as the first step
, a given wd template is used to fit the entire lss - gac spectroscopic data base and the resultant reduced @xmath30 values are recorded .
those @xmath30 , together with the overall s / n ratios of the lss - gac spectra are represented in a 2d map and an equation of the form , is defined so that all systems below this curve ( @xmath32 ) are considered as wd candidates , where @xmath33 and @xmath34 are free parameters defined for each template .
an example is illustrated in fig.[f - chi ] .
the form of eq.(1 ) is defined to account for systematic errors between template and observed spectra , which become more important the larger the s / n is .
the particular shape of the curve is carefully evaluated for each template as a compromise between the number of excluded / selected spectra . in other words , we aim at selecting as many wd candidate spectra as possible of similar spectral features as the templates used ( note that , strictly speaking , the only free parameter we are considering in the fitting process is the normalization factor ) .
however , the number of selected targets can not be too large , as otherwise the automatization becomes useless .
the spectra of all wd selected candidates are then visually inspected and systems that are not wds are simply excluded .
this exercise is repeated for all wd templates considered . in this case ,
our templates are the result of adding artificial gaussian noise to 45 da ( hydrogen - rich ) carefully selected wd model spectra provided by @xcite that cover a broad range of wd effective temperatures ( 6,000k-60,000k ) and surface gravities ( 7 @xmath359dex ) .
the model spectra were binned to the lss - gac resolving power .
four of our considered wd templates are shown in fig.[f - temp ] .
obviously , using only da wd templates restricts our search to hydrogen - rich wds .
however , as we show below ( section[s - compl ] ) , the number of non - da wds currently observed by the lss - gac is rather small .
moreover , it has to be stressed that the presence of a small fraction of non - da wds does not seem to influence significantly the shape of the lf @xcite
. as already mentioned , lss - gac spectra are the result of combining a blue and a red optical spectrum obtained from two separate arms . given that single wds are blue objects , in principle ,
the blue arm spectra alone should allow us to efficiently identify such stars . to investigate this
, we separately applied the above described routine to all blue - arm lss - gac spectra as well as to all combined ( blue- and red - arm ) spectra and compared the results .
we found that the number of identified wds differed significantly .
specifically , we identified @xmath36 per cent more objects using the combined , blue plus red , spectra .
moreover , all wds identified using only the blue - arm spectra were included in the list of targets found using the combined lss - gac spectra .
the overall shape of the continuum spectrum is hence required for efficiently selecting single wds .
we found this effect to be most important when selecting wds against hot ( a ) main sequence stars in which the strengths of the balmer lines are similar .
this was also the case when selecting wds amongst low s / n ( @xmath37 ) spectra .
for these two reasons , we decided to only consider the combined lss - gac spectra of s / n ratio above 5 in this work .
this excluded 15,442 lss - gac spectra for which the default 2d pipeline ( section[s - gac ] ) failed to deliver usable red- and/or blue - arm spectra .
it also excluded 622,112 spectra of s / n ratios below 5 .
this requirement is further supported by the fact that reliable stellar parameters ( necessary for obtaining reliable lf , mf and af ) can not be derived for low s / n ratio wd spectra @xcite .
thus , the total number of lss - gac spectra considered in this work was 306,600 , of which 20,029 were selected as wd candidates by our @xmath29-template fitting method . a visual inspection revealed that 94 of those are genuine wds , 81 of da type ( see table[tab : numbers ] ) .
we also identified 26 wdms binaries . with the exception of j080357.21
+ 334138.6 , these binaries are already included in the lamost wdms binary catalogue by @xcite . in the previous section we searched for da wds by @xmath29-template fitting all lss - gac spectra . here
we complement this with an independent strategy based on xstps - gac @xmath15 , @xmath4 and @xmath16 photometry ( section[s - gac ] ) .
this exercise relies simply on applying a well - defined cut to the xstps - gac @xmath38 and @xmath39 colours ( see table1 of @xcite and fig.[f - colcut ] of this paper ) .
@xcite provide two additional colour cuts , using the @xmath40 and @xmath41 magnitudes respectively , however these are not used here because the xstps - gac does not provide magnitudes in those filters .
16,824 sources fall within the area defined by our cut and visual inspection of spectra of those targets confirms 78 as wds , of which 70 are das see table[tab : numbers ] .
the vast majority of the remaining targets are single a , f main sequence stars whose colours overlap with those of cool wds .
the contamination by quasars is negligible as these sources are generaly too faint to be observed by the lss - gac survey .
63 ( 57 ) of these 78 ( 70 ) wds ( das ) are already discovered following the above described @xmath29-template fitting method .
thus , the colour method adds 15 new objects ( 13 of which are das ) to our sample .
the total number of lss - gac wds thus raises to 105 , and among them we classify 90 as da wds ( table[tab : numbers ] ) .
the complete lss - gac da wd sample is provided in tablea2 ( see appendix ) .
two important conclusions can be drawn from the above exercise .
first , the @xmath29-template fitting method failed to identify @xmath42 per cent of the whole lss - gac da wd sample . visual inspection of the spectra of those systems revealed that they were either of low s / n ratios ( @xmath43 in the blue - arm spectra ) , or were subject to artifacts of a bad flux calibration / merging of the two - arm ( blue plus red ) spectra , or a combination of both .
although those effects clearly affect the identification of da wds when applying the @xmath29-template fitting method , they do not influence the colour selection of da wds .
secondly , the da wd colour cut we applied missed @xmath44 per cent of the whole da wd sample identified .
a closer inspection revealed the @xmath15 magnitudes of these targets to be unreliable .
this caused those objects to fall far from the colour locus expected for da wds ( see fig.[f - colcut ] ) .
.[tab : numbers ] number of wds ( including das , dbs , etc . ) and das alone identified by the @xmath29-s / n template fitting and colour methods , in common by the two methods , independently by each method , and missed by both methods . the total number of lss - gac wds and das are provided in the last row . [
cols="<,^,^",options="header " , ] | modern large - scale surveys have allowed the identification of large numbers of white dwarfs .
however , these surveys are subject to complicated target selection algorithms , which make it almost impossible to quantify to what extent the observational biases affect the observed populations . the lamost ( large sky area multi - object fiber spectroscopic telescope ) spectroscopic survey of the galactic anti - center ( lss - gac ) follows a well - defined set of criteria for selecting targets for observations .
this advantage over previous surveys has been fully exploited here to identify a small yet well - characterised magnitude - limited sample of hydrogen - rich ( da ) white dwarfs .
we derive preliminary lss - gac da white dwarf luminosity and mass functions .
the space density and average formation rate of da white dwarfs we derive are @xmath0 pc@xmath1 and @xmath2 pc@xmath1 yr@xmath3 , respectively .
additionally , using an existing monte carlo population synthesis code we simulate the population of single da white dwarfs in the galactic anti - center , under various assumptions .
the synthetic populations are passed through the lss - gac selection criteria , taking into account all possible observational biases .
this allows us to perform a meaningful comparison of the observed and simulated distributions .
we find that the lss - gac set of criteria is highly efficient in selecting white dwarfs for spectroscopic observations ( 80 - 85 per cent ) and that , overall , our simulations reproduce well the observed luminosity function .
however , they fail at reproducing an excess of massive white dwarfs present in the observed mass function .
a plausible explanation for this is that a sizable fraction of massive white dwarfs in the galaxy are the product of white dwarf - white dwarf mergers .
( stars : ) white dwarfs ; stars : luminosity function , mass function |
the case for cosmic ray acceleration at supernova remnant ( snr ) shocks has found increasing support in observational evidence , both from the thermal and non - thermal components of the emission from these objects .
the production of relativistic electrons has been clearly established from the non - thermal x - ray emission detected from several shell - type snrs , e.g. , sn 1006 @xcite , rx j1713.7 - 3946 @xcite , and vela jr .
@xcite . additionally , the dynamical properties of some snrs suggest that a significant fraction of the explosion energy is going into the acceleration of ions to relativistic speeds , like in the case of 1e 0102.2 - 7219 @xcite .
x - ray measurements of the relative positions of the forward and reverse shocks , and contact discontinuity in tycho s snr also indicate modification of the remnant dynamics by the acceleration of cosmic - ray ions @xcite .
@xmath0-ray emission in the tev range from some snrs , observed with ground - based telescopes , seems to confirm that particles are indeed being accelerated to energies approaching the _ knee _ of the cosmic ray spectrum @xcite .
however , there is growing debate as to whether the origin of this high energy emission is hadronic or leptonic in nature , as reviewed in @xcite .
pion production , followed by @xmath3 decay , is a natural result of the interaction of relativistic ions , like those generated through diffusive shock acceleration at snr shocks , with dense material @xcite .
hence , snrs known to be interacting with molecular clouds present particularly interesting targets for the detection of @xmath0-ray emission of hadronic origin , as suggested by @xcite and @xcite .
the detection of hydroxyl ( oh ) masers at 1720 mhz is a very useful tool for finding this type of interactions , as they indicate the presence of shocked clouds of dense molecular material @xcite .
observations in the mev to gev range are especially important in order to identify signatures of @xmath1-decay , because the photon spectrum from this process has a distribution that peaks approximately at 70 mev @xcite .
the energetic gamma - ray experiment telescope ( egret ) , on board the compton gamma - ray observatory , found a host of discrete @xmath0-ray sources in the gev band possibly associated with galactic snrs @xcite .
however , due to poor spatial resolution and other instrumental constraints , it has proven impossible to determine unambiguously the origins of these egret sources . the large area telescope ( lat ) , onboard the _ fermi gamma - ray telescope _ has opened a new window for studying snrs in the gev energy range , due to its improved sensitivity and resolution .
for example , the detection of an extended lat source coincident with snr w51c ( g49.2 - 0.7 ) , was reported by @xcite , and their study suggests that @xmath1-decay emission is the dominant component of the @xmath0-ray flux observed in its direction .
the observation of oh maser ( 1720 mhz ) emission in the direction of this snr , reported by @xcite , suggests that it is interacting with a cloud of dense material .
snrs g349.7 + 0.2 , ctb 37a , 3c 391 and g8.70.1 , are known to be interacting with molecular clouds , as evidenced by oh masers observed in their directions @xcite . in this paper
we report the detection of four lat sources coincident with these snrs , and on the results of our detailed study of @xmath0-ray emission in their regions . after a description of how the lat data are analyzed in this work in section 2
, we present the results of the observations and put them in context for each individual object in section 3 .
a discussion of the results is given in section 4 .
the lat @xmath0-ray detector is the main instrument on board the _ fermi gamma - ray space telescope _ , which was launched on 2008 june 11 .
the lat is a pair - conversion , imaging , large field - of - view ( fov ) @xmath0-ray telescope , that covers a wide energy range , from 20 mev and up to 300 gev .
high precision tracking of electrons and positrons resulting from pair conversion allow the lat to reconstruct the direction of @xmath0-rays incident on the detector , and measurements of the energy of the subsequent electromagnetic showers are performed in its calorimeter .
the fov is @xmath4 2.3 sr and the point - spread function ( psf ) varies largely with photon energy , and improves at higher energies .
the converter - tracker device has 16 planes of high-_z _ material , which are divided in two regions .
the _ front _
section contains the first 12 layers which are thinner and optimize the psf , and the _ back _ region has 4 thicker planes , which maximize the effective area at the expense of a lower angular resolution relative to the _ front _ section .
further details about the instrument and data calibration are described in @xcite .
the lat , while operating on `` sky survey '' mode , observes the entire sky every two orbits ( approximately 3 hr ) . in this work ,
15 months of lat data ( from 2008 august until 2009 november ) are analyzed .
only events belonging to the _ diffuse _ class , which reduces the residual background rate as explained in detail in @xcite , have been selected for this study .
the updated instrument response functions ( irfs ) used are called `` pass6 version 3 '' , which were developed using in - flight data and consider pile - up and coincidence effects not included in the pre - launch analysis @xcite .
the systematic uncertainties of the effective area , for the irf used , are energy dependent : 10% at 100 mev , decreasing to 5% at 560 mev , and increasing to 20% at 10 gev @xcite .
additionally , only events coming from zenith angles smaller than 105@xmath5 are selected in order to reduce the contribution from terrestrial albedo @xmath0-rays @xcite . the @xmath0-ray data in the direction of g349.7 + 0.2 , ctb 37a , 3c 391 and g8.70.1 are analyzed using the fermi science tools .
in order to study spatial distribution , position and spectral characteristics , the maximum likelihood fitting technique is employed using _
gtlike_. the likelihood statistical approach is used on lat data because of the low detection rates and the extent of the psf , and it estimates the best model parameters by maximizing the joint probability for the data given the emission model @xcite .
the emission models used in _ gtlike _
place individual sources , with particular spectral properties , at fixed positions , and include a galactic diffuse component resulting from the interactions of cosmic rays with interstellar material and photons , and an isotropic one that accounts for the extragalactic diffuse and residual backgrounds . in this work
, the mapcube file gll_iem_v02.fit is used to describe the @xmath0-ray emission from the milky way , and the isotropic component is modeled using the isotropic_iem_v02.txt table .
spatial analysis is performed using @xmath0-ray data in the energy range 2 - 200 gev , and converted in the _ front _ section , in order to improve the angular resolution . the 68% containment radius angle for normal incidence
_ front_-selected photons in this energy band is @xmath6 . to analyze the neighborhoods of the snrs studied , count maps of the @xmath7 region centered on the region of interest ( roi )
are constructed , and smoothed by a gaussian kernel to match the psf of the events selected . in order to determine the detection significance and the precise position of the sources associated with the objects of interest , galactic and isotropic backgrounds
are modeled and test statistic maps are constructed using _ gttsmap_. the test statistic is the logarithmic ratio of the likelihood of a point source being at a given position in a grid , to the likelihood of the model without the additional source , @xmath8 .
the resolution of these maps , set by the size of the grid used , is @xmath9 . the study of the spectral energy distribution ( sed ) characteristics of each of the sources associated with g349.7 + 0.2 , ctb 37a , 3c 391 , and g8.70.1 , is performed through likelihood analysis with _
gtlike_. events converted in both _
front _ and _ back _ sections , and in the energy range 0.2 - 51.2 gev , are included in the study . the lower energy bound is selected to avoid source confusion , to avoid the rapidly changing effective area of the instrument at low energies , and because of the large uncertainty below 0.2 gev related to the galactic diffuse model used .
the spectral analysis is performed using _ gtlike _ to model the flux at each energy bin and estimate , through the maximum likelihood technique , the best - fit parameters .
background sources from the 3 month _
fermi _ lat bright gamma - ray source list @xcite , and detected pulsars from the 11 month _ fermi _ lat pulsar catalogue @xcite , within a @xmath10 radius of the sources of interest are modeled .
sources detected in the test statistic map of each field with significance greater than @xmath11 are also included in the background model .
additionally , the analysis was repeated using all background sources in the recently released one - year _ fermi _ lat first source catalog ( 1fgl ) in each region , with no significant effect on the results .
the normalization of the galactic diffuse component is left free at each energy step . as an addition to the statistical uncertainties associated with the likelihood approach , and the systematic errors related to the irf as mentioned above ,
the uncertainty of the underlying galactic diffuse level is considered .
the treatment of this error in the evaluation of the systematic uncertainties is performed by artificially changing the normalization of the galactic background by @xmath12 , similarly to the analysis in @xcite .
+ g349 + 0.2 has one of the highest radio and x - ray surface brightnesses in the milky way .
its radio image is characterized by a small diameter ( approximately @xmath13 ) partial shell , somewhat enhanced toward the southern and eastern limbs @xcite .
five oh maser emission spots ( at 1720 mhz ) were detected in the direction of this snr , which suggests that it is interacting with a dense molecular cloud @xcite .
the velocity of these maser lines places the regions of shocked material at a distance @xmath14 kpc .
the characteristics of the emission detected from molecular transitions of @xmath15co , @xmath16co , cs , hco@xmath17 , h@xmath18co , hcn and so , at similar velocities to those of the oh masers , are further evidence of molecular cloud - shock interaction @xcite .
+ in the x - ray band , the morphology of the remnant observed with _
asca _ and _ chandra _ is similar to that from the radio , and the high photon absorption of its x - ray spectrum ( @xmath19 @xmath20 ) is consistent with the distance estimate from maser emission @xcite . in the analysis of the _ chandra _ observations ,
@xcite find that the overall x - ray emission of the remnant is best - fit by a nonequilibrium ionization plasma model with enhanced si abundance , and temperature @xmath21 kev .
the fit is significantly improved with the addition of a softer component of equilibrium thermal plasma with solar abundances and temperature @xmath22 kev .
the si overabundance suggests that the harder spectral component is emission from the supernova ejecta , and the softer component , with solar abundances , is related to shocked circumstellar material . from the volume emission measure these authors estimate the hydrogen density for the soft component to be @xmath23 @xmath24 , where @xmath25 is the volume filling factor , and assuming that the distance is 22 kpc and @xmath26 .
the _ chandra _
observations also revealed x - ray emission from a point source , cxou j171801.0 - 372617 , inside the snr shell , with x - ray characteristics similar to the compact central object ( cco ) class , such as the source in cas a , and unlike those for energetic pulsars @xcite .
the _ fermi _ lat count map around g349.7 + 0.2 is shown in figure [ fig : im]a , overlaid with the contours of the _ chandra _ x - ray map .
a bright unresolved @xmath0-ray source is coincident with the position of the remnant , and with that of the source designated as 1fgl j1717.9 - 3729 in the one - year _ fermi _ lat first source catalog .
the significance of the detection , obtained from the evaluation of the test statistic in the region , is @xmath27 .
the residual test statistic map shows no evidence that the source is spatially extended .
the spectrum of the @xmath0-ray emission is shown in figure [ fig : spec]a , and its shape suggests that the spectral index steepens above a few gev . the best - fit model for the data yields a spectral index of @xmath28 .
the addition of an exponential cutoff is found to marginally improve the fit , the cutoff energy is estimated to be @xmath29 gev , and the spectral index is @xmath30 .
table [ tab : specs ] summarizes the characteristics of the best fit spectral model of the emission from this source , as well as its position .
the luminosity in the 0.1 - 100 gev band of this source is estimated to be @xmath31erg s@xmath32 , at a presumed distance of 22 kpc .
the snr ctb 37a ( g348.5 + 0.1 ) is located in a complicated region composed of three different remnants ( figure [ fig : im]b ) .
its radio emission overlaps that of a second snr , g348.5 - 0.0 on the east @xcite , and a third remnant , ctb 37b , lies approximately 20@xmath33 to the north .
the radio morphology of ctb 37a appears as a well - defined partial shell towards the north and east , and an extended outbreak to the south .
hi absorption measurements by @xcite , toward both ctb 37a and ctb 37b , place constraints on the distance to the ctb 37a / b complex , which is estimated to lie between 6.6 kpc and 11.4 kpc , using a distance to the galactic center of 8.5 kpc . [ cols="<,^,^,^,^,^,^,^,^,^,^ " , ] various oh masers have been detected toward the interior of ctb 37a at 1720 mhz @xcite .
eight of these have velocities close to the upper velocity limit for this object given by the neutral hydrogen measurements ( -65 km s@xmath32 ) . at this velocity
, @xcite observed , in the co ( @xmath34 ) transition at 115 ghz , three molecular clouds coincident with the maser spots and estimated the distance to the clouds to be 11.3 kpc .
@xcite detected extended thermal x - ray emission from inside the ne shell of the remnant , using observations with _
chandra _ and _ xmm - newton_. this study found consistent fit results from the detectors on both x - ray observatories , with plasma temperature t@xmath35 kev and absorption @xmath36 @xmath20 .
additionally , these authors observed an extended non - thermal x - ray source , cxou j171419.8 - 383023 , toward the nw of the ctb 37a complex ( see figure [ fig : im]b ) , which they suggest might be associated with the emission of a pulsar wind nebula ( pwn ) . a vhe ( @xmath37100 gev ) @xmath0-ray source , hess j1714 - 385 , is coincident with ctb 37a ( figure [ fig : im]b ) , and it has been suggested this is a signature of the interaction between the remnant s shock wave and the surrounding clouds of dense material @xcite . in figure [
fig : im]b , the smoothed gev map of the region around ctb 37a is shown .
the @xmath0-ray emission in the _ fermi _ lat range appears as an unresolved source whose centroid lies within the eastern radio shell of ctb 37a , with detection significance 17@xmath2 .
this source is coincident with 1fgl j1714.5 - 3830 .
the residual test statistic map gives no indication that this source is extended beyond the psf of the instrument .
the spectrum of this source is displayed in figure [ fig : spec]b , together with the best - fit models .
the simple power - law model for the data yields a spectral index @xmath38 .
however , the fit is slightly improved by including an exponential cutoff at @xmath39 gev , with spectral index @xmath40 .
table [ tab : specs ] summarizes the result of the analysis of this source .
the luminosity in the 0.1 - 100 gev band of this source is estimated to be @xmath31erg s@xmath32 , at a presumed distance of 11.3 kpc .
3c391 ( g31.9 + 0.0 ) is a radio bright snr with an irregular spatial structure
. observations with the very large array ( vla ) in the radio band uncovered an elongated morphology , which extends from northwest to southeast @xcite .
the shell surrounding the radio emission is enhanced toward the nw , and does not extend to the se , where it appears that the snr has broken out into a lower density region .
@xcite observed two oh maser spots coincident with 3c 391 at 1720 mhz , with velocities ( 105 and 110 km s@xmath32 ) that agree well with the neutral hydrogen absorption velocity @xcite .
the hi absorption measurements indicate a distance of at least 7.2 kpc , and the absence of absorption at negative velocities suggests an upper limit of 11.4 kpc , using a distance to the center of the galaxy of 8.5 kpc .
the case for the interaction of 3c 391 with a molecular cloud , as suggested by the maser emission in the direction of the snr , is further strengthened by the observations of co ( @xmath34 ) emission in the region @xcite .
the co maps show evidence that 3c 391 resides at the edge of a molecular cloud located to the nw of the snr , and that the `` breakout '' morphology is produced as the snr blast wave reaches lower density regions outside the cloud .
observations with _ asca _ @xcite and _ chandra _ @xcite show that the x - ray emission associated with this object is thermal and clumpy in nature , and its brightness peaks well inside the radio shell .
the plasma temperature derived from the x - ray observations is @xmath41 kev , and the absorption column density is found to increase from se to nw ( from 0.6 to 4.1 @xmath42 ) , which is consistent with the existence of molecular cloud to the northwest @xcite .
these authors also discovered an unresolved x - ray source on the nw border , with spectral characteristics similar to those of cco class of objects . the smoothed photon count map from the _
fermi_-lat data is shown in figure [ fig : im]c , and presents a strong case for an association between a gev source and 3c 391 .
the unresolved @xmath0-ray source is detected with a 13@xmath2 significance and the peak of the test statistic map is 4@xmath33 from the northwestern radio shell and the position of the closest maser spot .
this source is coincident with 1fgl j1849.0 - 0055 .
the sed of the emission from this source is best described by a power law model , with spectral index @xmath43 .
the sed is shown in figure [ fig : spec]c , and spectral characteristics are presented in table [ tab : specs ] .
the luminosity of this source in the 0.1 - 100 gev band is @xmath44erg s@xmath32 , at a presumed distance of 8 kpc .
the rather large snr g8.70.1 , together with nine discrete hii regions , forms the w30 complex .
g8.70.1 was identified as a galactic snr on the basis of the observation of a 45@xmath33 diameter area of extended non - thermal radio emission @xcite .
@xcite detected a single oh maser at 36 km s@xmath32 , on the eastern edge of g8.70.1 .
this relatively bright emission at 1720 mhz does not coincide with any of the compact radio sources , and hence they associate it with the snr , and using the rotation curve of @xcite they estimate a distance to the remnant of @xmath45 kpc .
an extended @xmath0-ray source , hess j1804 - 216 , has been detected to the southwest of the remnant ( figure [ fig : im]d ) , which could be further evidence of shock interaction with clouds of dense material @xcite .
several other sources in the field have been suggested as possible associations with the source hess j1804 - 216 , and they are shown in figure [ fig : im]d @xcite .
@xcite detected diffuse thermal x - ray emission from the northern half of the remnant in the 0.1 - 2.4 kev band with _
rosat _ , and using sedov analysis estimated the distance to the remnant to be @xmath46 kpc @xmath47 kpc .
the estimated plasma temperature from these x - ray observations is @xmath48 kev , the absorption column density is @xmath49 @xmath20 , and the derived ambient density is @xmath50 .
these authors also determined the distance to the hii regions associated with remnant to be @xmath51 kpc , using the kinematic measurements of @xcite and more recent galactic rotation models . the @xmath0-ray photon count map in the 2 - 200 gev range of the crowded vicinity of the w30 complex is shown in figure [ fig : im]d .
the @xmath0-ray emission in the region appears as a source centered on a position inside the radio boundary of g8.70.1 .
this detection is coincident with 1fgl j1805.2 - 2137 .
the residual test statistic map does not rule out the possibility that this source is extended .
the significance of the detection is @xmath52 .
the spectrum of this source , shown in figure [ fig : spec]d , seems to steepen above 2 gev , and flatten at @xmath53 gev . for a single power - law model
the best - fit spectral index is @xmath54 .
the spectral characteristics of this region are summarized in table [ tab : specs ] .
the luminosity in the 0.1 - 100 gev band of this source is @xmath55erg s@xmath32 , at a presumed distance of 4.5 kpc .
the detection with the _ fermi_-lat of unresolved @xmath0-ray sources in the direction of g349.7 + 0.2 , ctb 37a , and 3c 391 , as well as a possibly extended region of emission coincident with g8.70.1 is reported in this work .
the best - fit positions and spectral characteristics of these sources are summarized in table [ tab : specs ] .
the shapes of the spectra of the sources coincident with g349.7 - 0.5 , 3c 391 and g8.70.1 do not seem consistent with what has been observed from pulsars in @xmath0-rays .
@xcite found the sed of pulsars in the _ fermi _ lat band to be best described by power - law distributions , with exponential cutoffs in the energy range 1 - 6 gev ( the only exception , psr j1833 - 1034 with cutoff energy @xmath56 8 gev , was reported to be poorly measured ) .
the spectral characteristics of the sources coincident with 3c 391 and g8.70.1 do not suggest an exponential cutoff at all , and the spectrum of g349.7 + 0.2 is very modestly improved by a cutoff at @xmath57 gev , which is very high for a @xmath0-ray pulsar . additionally , there are no pulsars from the australian telescope national facility ( atnf ) pulsar catalogue coincident with this source @xcite
. the spectrum of the source coincident with ctb 37a is well fit by a power law with an exponential cutoff at energy @xmath58 gev , and hence does not rule out a pulsar hypothesis .
however , there are no atnf pulsars within 20@xmath33 radius of this source .
given the lack of evidence for significant contribution by a pulsar , and the known presence of maser emission for each remnant , an appropriate hypothesis appears to be that the @xmath0-ray sources are products of snr interactions with molecular clouds . to investigate properties of the expected spectra
, we have considered the scenario in which the @xmath0-rays are produced by @xmath1 decay of accelerated hadrons . in order to reproduce the observed spectra ,
we have used a model based on that from @xcite using a scaling factor of 1.85 for helium and heavier nuclei @xcite , and adopted a power law model with an exponential cutoff for the energy distribution of the protons : where @xmath60 is the proton energy cutoff .
table [ tab : dens ] presents sets of representative model parameters that adequately reproduce the observed @xmath0-ray spectra ( see figure 2 ) . while the spectra of snrs 3c 391 and g8.70.1 are well matched by models with proton distributions with spectral indices @xmath61 and high - energy cutoff , @xmath62 , the spectra of g349.7 + 0.2 and ctb 37a require harder proton distributions , @xmath63 , and @xmath64 .
the model provides an estimate of the density of the material the snr shock wave is impacting , given reasonable assumptions about the shock compression ratio , @xmath65 , and the fraction @xmath66 of the total supernova explosion energy @xmath67 that is converted into cosmic ray energy .
here we assume @xmath68 , @xmath69 , and @xmath70 erg ; we use distances derived from observations at other wavelengths for each snr . in table
[ tab : dens ] , estimates of the densities of the @xmath0-ray emitting material are compared to the densities derived from x - ray observations in other studies @xcite .
object & & @xmath71&@xmath60&_n_@xmath72 & _ n_@xmath73&_d + & & & ( tev)&(@xmath74&(@xmath74 & ( kpc ) + g349.7 + 0.2&&1.7&0.16&65&2.5&22 + ctb 37a&&1.7&0.08&37& ... &11.3 + 3c 391&&2.4&100&28&1&8 + g8.7 - 0.1&&2.45&100&18&0.03&4.5
+ + + _ the values obtained from the @xmath1-decay model fits to the @xmath0-ray spectra of the snrs are much higher than those from x - ray observations .
it is possible that , upon interacting with dense ambient material , instabilities in the postshock flow result in considerable clumping of material .
the denser regions would then not be expected to yield significant x - ray emission , although radiative shocks would then be expected to produce optical emission from these regions .
if the filling fraction of the clumped material is large , then the densities inferred for the x - ray emitting material are underestimates , potentially leading to more reasonable values for the density contrast between clump and interclump regions .
alternatively , the energetic particles that escape the particle acceleration region at the snr shock could stream ahead of the shock and encounter clouds of molecular material , where the much higher density could dramatically enhance the @xmath0-ray emission @xcite .
this scenario has been proposed as a possible explanation of the observed @xmath0-ray fluxes in the direction of some galactic snrs , including rx j1713.7 - 3946 @xcite and w28 @xcite .
@xcite model the interaction of particles accelerated by the diffusive shock of an snr with arbitrary matter distributions outside the remnant blast wave , including an appropriate treatment of the escape and diffusion of these particles .
these authors , as well as other works @xcite , show that clouds of dense material in close proximity to snrs can have an important , and even dominant contribution to the @xmath0-ray flux from these regions .
it is important to note , however , that the escaping particles are predominantly from the high - energy portion of the spectrum ; the observed spectra , which extend to low energies , may be problematic for such a scenario .
an alternative scenario is that the @xmath0-ray emission coincident with these sources is of leptonic origin .
bremsstrahlung and inverse compton ( ic ) radiation from the nonthermal electron population have been proposed as production mechanisms of @xmath0-rays in snrs ( e.g. * ? ? ?
* ) , but these models present certain difficulties as well .
non - thermal bremsstrahlung can be the dominant emission process in this band for electron - to - proton ratios , @xmath75 , greater than @xmath76 @xcite . however , cosmic - ray abundance ratios suggest @xmath77 , and models for @xmath0-ray emission from other snrs ( e.g. , rx j1713.7 - 3946 ) favor even smaller values @xcite , making the bremsstrahlung - dominated model appear less likely .
the ic - dominated scenario , on the other hand , is difficult to reconcile with the observed characteristics of the sources studied in this work .
the expected spectral index for ic emission ( @xmath78 ) is considerably harder than the values obtained here ( @xmath79 ) . in summary , using observations with the _
fermi_-lat , we have revealed @xmath0-ray emission from four snrs known to be interacting with molecular clouds . in each case
the inferred density is high , assuming a hadronic origin for the @xmath0-rays , consistent with the presence of dense material indicated by maser emission .
however , estimated densities of the ambient material derived under simple geometric assumptions are in considerable excess of those derived from the hot x - ray gas observed in the snrs .
more detailed modeling of the x - ray and @xmath0-ray emission under a variety of scenarios for the ambient medium is required to address the emission mechanism more fully .
these results are consistent with the scenario presented by @xcite for snr w51c , and further studies of other such systems and their gev - scale @xmath0-ray emission with the _ fermi_-lat are of particular interest .
the authors thank yasunobu uchiyama and stefan funk for their careful review of this work , and don ellison , keith bechtol , joshua lande , andrei bykov , luke drury , and felix aharonian for helpful discussions .
this work was partially funded by nasa contract nas 8 - 03060 and fermi grant nnx09at68 g . p.o.s .
acknowledges the kitp in santa barbara , where work on this project was begun while participating in a kitp program . | we report the detection of @xmath0-ray emission coincident with four supernova remnants ( snrs ) using data from the large area telescope on board the _ fermi gamma - ray space telescope_. g349.7 + 0.2 , ctb 37a , 3c 391 , and g8.70.1 are snrs known to be interacting with molecular clouds , as evidenced by observations of hydroxyl ( oh ) maser emission at 1720 mhz in their directions .
snr shocks are expected to be sites of cosmic ray acceleration , and clouds of dense material can provide effective targets for production of @xmath0-rays from @xmath1-decay .
the observations reveal unresolved sources in the direction of g349.7 + 0.2 , ctb 37a , and 3c 391 , and a possibly extended source coincident with g8.70.1 , all with significance levels greater than 10@xmath2 . |
unconventional superconductors usually refer to the superconductors those can not be understood within the conventional bardeen - cooper - schrieffer ( bcs ) theory .
notable examples are high-@xmath4 cuprates @xcite , heavy fermion superconductors @xcite , organic superconductors @xcite and iron based superconductors @xcite .
the unconventional superconductivity is usually driven by strong electron - electron interaction , and the electron pairing mechanism usually has a magnetic origin . in the past decades , identifying the precise gap symmetry of unconventional superconductors has attracted great theoretical and experimental efforts since such efforts may lead to important progress in seeking the microscopic pairing mechanism .
many unconventional superconductors are believed to have a @xmath1-wave energy gap , which is different from that of isotropic @xmath5-wave superconductors .
however , it is not an easy task to determine the precise @xmath1-wave gap symmetry .
a powerful and frequently used approach is to probe the angular dependence of various observable quantities , such as upper critical field @xcite , specific heat @xcite , and thermal conductivity @xcite . in this paper , we are mainly interested in the behaviors of in - plane upper critical field @xmath0 in heavy fermion superconductors .
this issue has recently been addressed with the aim to identify the precise gap symmetry of some heavy fermion compounds , especially cecoin@xmath6 @xcite and cecu@xmath3si@xmath7 @xcite . despite the intensive theoretical and experimental efforts ,
it remains unclear whether the gap symmetry of these compounds is @xmath8-wave or @xmath9-wave .
these two gaps are different from each other primarily in the positions of gap nodes . in principle , their positions can be clarified by measuring the angular dependence of @xmath0 . unfortunately , experimental studies have not yet reached a consensus on this issue . in the case of cecoin@xmath2 ,
currently most of experiments suggest that the gap symmetry should be @xmath8-wave@xcite , however there is currently still an experimental discrepancy in the concrete angular dependence of @xmath0 in cecoin@xmath2 : some experiments find that the maxima of @xmath0 are along the [ 100 ] direction @xcite , whereas other experiment observes the maxima along the [ 110 ] direction @xcite .
this discrepancy is still a open puzzle which need to be resolved@xcite . in the case of cecu@xmath3si@xmath3 , many earlier experiments suggest a @xmath10-wave gap@xcite .
nevertheless , a recent measurement @xcite observes the maxima of @xmath0 along the @xmath11 $ ] direction , which is argued to infer a @xmath9-wave gap according the corresponding theoretical analysis@xcite .
apparently , more research efforts are called for to solve these puzzles , which have motivated us to revisit this issue more systematically .
now suppose an external magnetic field is introduced to a superconductor . in principle , this field can couple to the charge and spin degrees of freedom of the electrons via the orbital and zeeman mechanisms respectively .
the former mechanism is described by the minimal coupling between the momentum of electrons and the vector potential , and can lead to the well - known abrikosov mixed state in type - ii superconductors .
the latter mechanism , usually called pauli paramagnetic or pauli limiting effect , is known to be important in some heavy fermion compounds @xcite .
which one of these two effects plays a dominant role is determined by a number of physical factors .
when both of them are important , novel and interesting properties may emerge . since the middle of 1990s , the in - plane
@xmath0 has been applied to identify the gap symmetry in layered unconventional superconductors @xcite .
early theoretical calculations have showed that the in - plane @xmath0 exhibits a fourfold oscillation in @xmath1-wave superconductors @xcite .
the presence of such a fourfold oscillation has already been verified in many unconventional superconductors , including high-@xmath4 cuprate superconductors @xcite , luni@xmath3b@xmath3c @xcite , heavy fermion compounds cecoin@xmath6 @xcite and cecu@xmath3si@xmath7 @xcite .
+ [ fig : shaped ] in the early calculations of won _ et .
_ @xcite and takanaka _ et .
@xcite who solely considered the orbital effect , @xmath0 is found to exhibit its maxima along the antinodal directions where the @xmath1-wave superconducting gap is maximal .
the subsequent analysis of weickert _ et .
@xcite includes both the orbital and pauli paramagnetic effects , but still finds the maxima of @xmath0 along the antinodal directions .
a similar conclusion is drawn in a recent work @xcite , where the authors also show that increasing the pauli effect reduces the difference in @xmath0 between nodal and antinodal directions .
there seems to be a priori hypothesis in the literature that a larger gap necessarily leads to a larger magnitude of @xmath0 , which means @xmath0 and @xmath1-wave gap should always have their maxima and minima at exactly the same azimuthal angles . if such a hypothesis is correct , it would be straightforward to identify the precise gap symmetry .
for instance , if the experimentally observed @xmath0 displays its maxima along the [ 100 ] direction , the gap possesses a @xmath8 symmetry . on the other hand , if the maxima are observed along the direction [ 110 ] , the gap symmetry should be @xmath9-wave . to make a comparison
, we show the angular dependence of @xmath8- and @xmath9-wave gaps in fig .
[ fig : shaped ] .
it is necessary to emphasize that the above hypothesized connection between in - plane @xmath0 and @xmath1-wave gap , though intuitively reasonable , is actually not always correct . when there is only orbital effect , the maxima of @xmath0 and @xmath1-wave gap are along the same directions in all cases . in the presence of pauli paramagnetic effect ,
however , there is indeed no guarantee that such a connection is valid . in order to clarify the detailed connection between the precise gap symmetry and the angular dependence of @xmath0
, we will consider the influence of the interplay of orbital and pauli effects on @xmath0 more systematically .
this problem is important because in - plane @xmath0 has recently played a significant role in the determination of the gap symmetries of cecoin@xmath2 and cecu@xmath3si@xmath3 . in this paper , motivated by the recent progress and the existing controversy
, we analyze the angular dependence of in - plane @xmath0 and its connection with the @xmath1-wave gap symmetry by considering the interplay of orbital and pauli effects in the contexts of heavy fermion compounds . after carrying out systematical calculations
, we will show that the maxima of angle - dependent @xmath12 are not always along the antinodal directions when both the orbital and pauli effects are important . the concrete fourfold oscillation pattern of @xmath12 is determined by a number of physical parameters , including temperature @xmath13 , critical temperature @xmath14 , gyromagnetic ratio @xmath15 , fermion velocity @xmath16 , and two parameters that characterize the shape of the underlying fermi surface .
each of these parameters can strongly affect the angular dependence of @xmath0 . among the above six relevant parameters , the temperature @xmath13 is particularly interesting , due to that in any given compound @xmath17 is the only free parameter and all the other parameters
are fixed at certain values .
if we vary temperature @xmath13 but fix all the rest parameters , @xmath12 is found to exhibit its maxima along the nodal directions at lower temperatures and along the antinodal directions at higher temperatures .
this means the angle - dependent @xmath12 is shifted by @xmath18 as temperature increases across certain critical value .
our results can be used to clarify the aforementioned experimental puzzle about the angular dependence of in - plane @xmath0 . since @xmath12 shifts by @xmath18 as some of the relevant parameters are changed
, the seemingly contradictory experimental results reported in refs.@xcite may be well consistent .
on the other hand , since the concrete behavior of @xmath12 is very sensitive to the specific values of several parameters , one should be extremely careful when judging the gap symmetry by measuring @xmath0 . in sec.[sec
: derive ] , we derive the equation for @xmath0 after including both the orbital and pauli paramagnetic effects . in sec.[sec : numresults ] , we present numerical results for @xmath0 in three cases , i.e. , pure orbital effect , pure pauli paramagnetic effect , and interplay of both orbital and pauli effects .
we show that @xmath0 displays complicated angle dependence due to interplay of orbital and pauli effects . in sec.[sec : discussion ] , we discuss the physical implications of our results and make a comparison with some relevant experiments .
heavy fermion compounds are known to have a layered structure , which is analogous to cuprates . however , the inter - layer coupling is not as weak as that in cuprates .
it is convenient to consider a rippled cylinder fermi surface , schematically shown in fig .
[ fig : fs ] .
the fermion velocity has three components @xmath19 , where @xmath20 denote the two components in the superconducting plane . here
, we use @xmath21 to represent the inter - layer hoping parameter and @xmath22 the unit size along @xmath23-direction , and then write the dispersion as @xcite @xmath24 introducing a constant magnetic field @xmath25 to the system leads to fruitful behaviors . for type - ii superconductors ,
the field @xmath25 weaker than lower critical field @xmath26 can not penetrate the sample due to the meissner effect .
as @xmath25 exceeds @xmath26 and further increases , the superconducting pairing is gradually destructed by the orbital effect .
the superconductivity is entirely suppressed once @xmath25 reaches the upper critical field @xmath0 , which can be obtained by solving the corresponding linearized gap equation . in some superconductors ,
the pauli paramagnetic effect can also close the gap by breaking spin singlet pairs , and may even be more important than the orbital effect @xcite . in order to make a general analysis
, we consider both of these two effects in the following .
to proceed , it is useful to rewrite the in - plane magnetic field @xmath27 in terms of a vector potential @xmath28
. let us choose the @xmath29-axis as @xmath30-coordinate and @xmath31-axis as @xmath32-coordinate , and then write down a vector potential @xmath33 where @xmath34 denotes the angle between a - axis and field @xmath27 . for conventional @xmath5-wave superconductors ,
the gap is isotropic and the upper critical field @xmath0 is certainly @xmath34-independent . in the case of @xmath1-wave superconductors
, however , the gap is strongly anisotropic , thus @xmath0 becomes @xmath34-dependent .
now the field @xmath27 takes the form @xmath35 one can write the generalized derivative operator as @xmath36 following the general methods presented in refs .
@xcite , we obtain the following linearized gap equation : @xmath37 \right\}\delta(\mathbf{r } ) , \label{eqn : gapl}\end{aligned}\ ] ] where @xmath38 . the function @xmath39 is @xmath40 here we do not include landau level mixing @xcite for simplicity , which will not affect our conclusion . for the chosen fermi surface ,
the fermi velocity vector is @xcite @xmath41 the fermi velocity component along the @xmath22-axis is @xmath42 .
the two - component in - plane velocity vector has a constant magnitude @xmath43 , defined as @xmath44 , where @xmath45 with the fermi momentum @xmath46 being related to the fermi energy @xmath47 by @xmath48 .
the shape of rippled cylinder fermi surface is characterized by a velocity ratio @xmath49 , where @xmath50 and @xmath51 .
as will shown below , both @xmath52 and @xmath53 can strongly affect the behavior of @xmath0 .
moreover , we define @xmath54 , where @xmath55 is bohr magneton and @xmath15 is the gyromagnetic ratio .
the orbital effect of magnetic field is reflected in the factor @xmath56 , whereas the pauli paramagnetic effect is reflected in the factor @xmath57 .
the concrete behavior of @xmath0 is determined by the interplay of these two effects . in eq.([eqn : gapl ] ) , the influence of gap symmetry is reflected in the function @xmath58 . for isotropic @xmath5-wave pairing , @xmath59 ; for @xmath10-wave pairing , @xmath60 ; for @xmath9-wave pairing , @xmath61 .
-dependent @xmath0 at two representative temperatures @xmath62 and @xmath63.,width=297 ]
although the linearized gap equation eq.(4 ) is formally general and valid in many superconductors , its solution is determined by a number of physical effects and associated parameters .
for instance , the behavior of @xmath0 may be strongly influenced by the concrete shapes of the fermi surface .
the fermi surface has different spatial dependence in various superconductors , which naturally leads to different forms of fermion dispersion and fermi velocity @xmath64 .
such a difference certainly affects the equation of @xmath0 . for spherical fermi surface , fermi velocity
@xmath64 depends on the azimuthal angle @xmath65 within the basal plane and the angle between @xmath23-axis and @xmath64 .
therefore , the equation of @xmath0 contains the integrations over these two variables @xcite . for cylindrical fermi surface ,
the direction of vector @xmath64 solely depends on the azimuthal angle @xmath65 , so there is only the integration over angle @xmath65 in the equation of @xmath0 @xcite . for rippled fermi surface ,
the direction of @xmath64 depends on the azimuthal angle @xmath65 and the coordinate @xmath66 along @xmath23-axis , then the integrations over @xmath65 and @xmath66 enter into the equation of @xmath0 , as shown in eq .
( [ eqn : gapl ] ) .
in addition , there are two independent parameters @xmath52 and @xmath53 which can characterize the rippled fermi surface in eq .
( [ eqn : gapl ] ) .
notice that once @xmath67 , the rippled cylindrical fermi surface reduces to the cylindrical fermi surface
. the influence of fermi surface on @xmath0 is rarely studied in the literature . in this paper , we adopt rippled fermi surface and show that @xmath0 can exhibit different behaviors under different parameters .
-dependence of @xmath0 with @xmath68 , @xmath69 , @xmath70 , and @xmath71.,width=297 ] to facilitate analytical computation , we can choose the direction of field @xmath27 as a new @xmath72-axis and define @xmath73 in the coordinate frame spanned by @xmath74 , we have a new velocity vector @xmath75 and a new generalized derivative operator @xmath76 , \label{eqn : pidef}\end{aligned}\ ] ] where @xmath77 , \\ a_{0 } & = & \frac{1}{\sqrt{2eh}}\left[-i\partial_{x}\cos\theta - i\partial_{y}\sin\theta \right],\end{aligned}\ ] ] which satisfy @xmath78 = 1 , [ a_{\pm},a_{0 } ] = 0.\end{aligned}\ ] ] in the following analysis , we take @xmath8-wave pairing as an example and assume that @xmath60 . the results in the case of @xmath9-wave pairing can be obtained analogously .
it is easy to examine that the qualitative conclusion will be not changed . after averaging over the ground state @xmath79 on both sides of eq.([eqn : gapl ] ) and inserting the @xmath8-wave gap @xmath80 , we obtain the following integral equation for @xmath0 , @xmath81 \nonumber \\ & & \times \exp\left[-\rho u^2 \left(\lambda^2\gamma^2\sin^{2}(\chi)\right.\right
. \nonumber \\ & & \left.\left.\left.+\left(1+\lambda\cos(\chi)\right)\sin^{2}(\varphi ) \right)\right ] \right\},\label{eqn : hc2expression}\end{aligned}\ ] ] where @xmath82 , @xmath83 and @xmath84 .
one can analyze the detailed behavior of @xmath0 , especially its dependence on various physical parameters , systematically by solving this integral equation .
this will be done in the next section .
-dependence of @xmath0 with @xmath68 , @xmath85 , @xmath86 , and @xmath87.,width=345 ]
in this section , we first present the numerical solutions of eq .
( [ eqn : hc2expression ] ) , then discuss the physical implications of the results , and finally compare our results with some recent experiments . from eq .
( [ eqn : hc2expression ] ) , we know the behavior of @xmath12 is determined by six physical parameters : @xmath88 among this set of parameters , @xmath52 and @xmath53 are related to the shape of rippled cylinder fermi surface .
we notice that the influence of these two parameters are rarely investigated in previous works on @xmath0 .
the critical temperature @xmath4 and the gyromagnetic ratio @xmath15 will be taken as varying parameters .
the detailed behavior of @xmath0 can be clearly seen from its angular dependence .
in addition , it is also interesting to analyze the difference of @xmath0 between its values obtained at @xmath89 and @xmath90 : @xmath91 since the maxima and minima of @xmath0 always appear at these two angles .
@xmath0 exhibits its maxima at @xmath92 if @xmath93 and at @xmath90 if @xmath94 . in order to demonstrate the influence of the orbital effect and that of the pauli paramagnetic effect on the angular dependence of in - plane @xmath0
, we find it helpful to consider three cases separately : pure orbital effect ; pure pauli effect ; interplay of orbital and pauli effects . at @xmath62 and @xmath63 with @xmath95 , @xmath96 , @xmath85 , @xmath86 , and @xmath97 .
the fourfold oscillation patterns are apparently different at low and high temperatures.,width=288 ] first , we consider only the orbital effect by setting the gyromagnetic factor @xmath98 . in this case
, the factor @xmath99 appearing in eq.([eqn : hc2expression ] ) is equal to unity , @xmath100 .
we assume that @xmath101 , @xmath96 , @xmath85 , and @xmath71 , which are suitable parameters in heavy fermion compounds . after carrying out numerical calculations
, we plot the angular dependence of @xmath12 in fig .
[ fig : hc2tnp ] at two representative temperatures @xmath62 and @xmath63 .
it is easy to see from fig .
[ fig : hc2tnp ] that @xmath12 exhibits a fourfold oscillation pattern .
moreover , the maxima of @xmath0 are always along the antinodal directions for any values of the relevant parameters , which means the angular dependence of orbital effect - induced @xmath0 is exactly the same as that of @xmath1-wave gap .
this is consistent with the original theoretical predictions of won _ et .
_ @xcite and takanaka _ et .
an important feature that needs to be emphasized is that the positions of peaks are temperature independent , as clearly manifested in both fig .
[ fig : hc2tnp ] and fig .
[ fig : hc2tcpnp ] .
the above properties can also be elaborated by the detailed @xmath17-dependence of @xmath0 and @xmath102 are presented in fig .
[ fig : hc2tcpnp ] .
@xmath0 is an monotonously decreasing function of parameter @xmath17 , valid for all values of @xmath34 .
this is easy to understand since the magnitude of the superconducting gap always decreases monotonously with growing temperature .
moreover , the difference @xmath102 is negative for all values of @xmath17 .
-dependence of @xmath0 with @xmath95 , @xmath69 , @xmath70 , @xmath71 , and @xmath87.,width=288 ] we next consider the effects of pure pauli paramagnetic effect by setting @xmath103 , which leads to @xmath104 which is completely independent of @xmath34 .
the @xmath17-dependence of @xmath0 is shown in fig .
[ fig : paulionly ] .
different from pure orbital effect , @xmath0 is not a monotonous function : it rises initially with growing @xmath17 , but decreases as @xmath17 is larger than certain critical value @xmath105 , which is roughly @xmath106 under the chosen set of parameters .
we now turn to the general and interesting case in which both the orbital and pauli paramagnetic effects are important . this case
is broadly believed to be realized in several heavy fermion compounds , such as cecoin@xmath6 and cecu@xmath3si@xmath7 . as aforementioned ,
the concrete behaviors of @xmath34-dependent @xmath0 are influenced by a number of parameters . in order to illustrate the numerical results and their physical implications , we vary one particular parameter while fixing all the rest parameters . in most of the following calculations ,
the gyromagnetic factor is taken to be @xmath87 .
the influence of various values of @xmath15 will be analyzed separately .
-dependence of @xmath0 with @xmath107 , @xmath69 , @xmath85 , @xmath71 , and @xmath97.,width=288 ] as shown in fig .
[ fig : hc2 t ] , under the currently chosen parameters , the maxima of @xmath0 locates along the antinodal directions at a relatively higher temperature @xmath63 .
this behavior is very similar to that in the case of pure orbital effect .
however , at a relatively lower temperature @xmath62 , the maxima of @xmath0 are along the nodal directions where the @xmath108-wave gap vanishes .
two conclusions can be immediately drawn : @xmath0 does not always exhibit its maxima at the angles where the superconducting gap reaches its maximal value ; the fourfold oscillation curves of @xmath0 is shifted by @xmath18 as temperature grows in the range of @xmath109 . from fig .
[ fig : hx2tcp](a ) , we see that @xmath0 first arises with growing @xmath17 and then decreases rapidly once @xmath17 exceeds a critical value .
apparently , such a non - monotonous @xmath17-dependence of @xmath0 is a consequence of the interplay of both orbital and pauli paramagnetic effects . on the other hand ,
the difference @xmath110 shown in fig .
[ fig : hx2tcp](b ) is positive for small values of @xmath17 but becomes negative for larger values of @xmath17 .
addition to temperature @xmath17 , the concrete angular dependence of @xmath0 is also strongly influenced by a number of other physical parameters , including critical temperature @xmath4 , fermion velocity @xmath16 , gyromagnetic factor @xmath15 , and two fermi surface factors @xmath52 and @xmath53 .
indeed , different values of these parameters can lead to very different behaviors of @xmath0 . in the following ,
we show how @xmath0 and @xmath102 are changed as these parameters are varying . to simplify the analysis
, we vary one particular parameter and fix all the other parameters in each figure .
-dependence of @xmath0 with @xmath111 , @xmath68 , @xmath85 , @xmath71 , and @xmath97.,width=288 ] @xmath4 : first , we consider the influence of critical temperature @xmath4 on @xmath0 and @xmath102 , and show the results in fig .
[ fig : hc2tc ] .
it is well - known that @xmath4 of heavy fermion compounds is actually quite low , especially when compared with cuprates and iron - based superconductors . to cover all possible heavy fermion compounds , we assume @xmath14 varies in the range of @xmath112 $ ] .
all the other parameters are fixed .
@xmath0 rises monotonously with growing @xmath4 , which is obviously owing to the monotonous increase of the superconducting gap .
moreover , if @xmath14 is smaller than some critical value , @xmath102 is negative , which means the maxima of @xmath0 are along the antinodal directions . for larger @xmath14 ,
@xmath102 becomes positive and the maxima of @xmath0 are shifted to the nodal directions .
apparently , @xmath4 has important impacts on the concrete angular dependence of @xmath0 . in passing , we point out that the maxima of @xmath0 will be shifted back to the antinodal directions for even higher @xmath4 ( not shown in the figure ) .
@xmath16 : we then consider the influence of fermion velocity @xmath16 on @xmath0 and @xmath102 , and show the results in fig .
[ fig : hc2v0 ] . in the limit @xmath113 ,
the orbital effect is actually ignored and the pauli effect entirely determines @xmath0 .
in such a limit , @xmath0 is angle independent , so @xmath114 . for finite @xmath16
, @xmath0 becomes angle dependent and exhibits fourfold oscillation , as a consequence of the interplay between orbital and pauli effects .
as @xmath16 is growing , @xmath0 first increases and then decreases , which indicates that the enhancement of orbital effect does not necessarily suppress @xmath0 once the pauli paramagnetic effect is present .
however , as already discussed earlier , @xmath0 deceases monotonously with growing @xmath115 when the pauli effect is completely neglected .
furthermore , @xmath110 is negative for both small and large values of @xmath115 , but is positive for intermediate values of @xmath115 .
therefore , the concrete angular dependence of @xmath0 is very sensitive to the values of fermion velocity .
-dependence of @xmath0 with @xmath62 , @xmath69 , @xmath95 , @xmath70 , and @xmath86.,width=291 ] @xmath15 : we next consider the influence of the gyromagnetic factor @xmath15 , which characterizes the effective strength of pauli paramagnetic effect . the dependence of @xmath0 and @xmath102 on @xmath15 is given in fig .
[ fig : hc2 g ] .
first of all , taking @xmath98 simply leads to the known results obtained in the case of pure orbital effect presented in sec.[subsec : pureorbital ] .
second , @xmath0 decreases monotonously with growing @xmath15 .
an immediate indication of this behavior is that increasing the pauli paramagnetic effect always tends to suppress @xmath0 in the presence of orbital effect . finally , it is easy to observe that @xmath102 is negative if @xmath15 takes very small values and positive when @xmath15 becomes larger than certain critical value
. therefore , the gyromagnetic factor @xmath15 also plays a crucial role in the determination of the concrete angle dependence @xmath0 .
@xmath52 : @xmath52 represents the ratio of inter layer coupling @xmath116 and the fermi energy @xmath117 . if @xmath118 , the corresponding @xmath119 , then the rippled cylindrical fermi surface reduce to the cylindrical fermi surface . the dependence of @xmath0 and @xmath102 on @xmath52 is as depicted in fig .
[ fig : hc2lambda ] .
@xmath0 deceases monotonously with the growing @xmath52 . for given values of other parameters shown in fig .
[ fig : hc2lambda ] , the maxima of @xmath0 are along the nodal directions for small @xmath52 , but along the antinodal directions for large values of @xmath52 .
@xmath53 : @xmath53 represents the ratio of two momentum @xmath46 and @xmath120 , @xmath22 is the unit cell size along third direction . the dependence of @xmath0 and @xmath102 on @xmath53 is shown in fig . [
fig : hc2gamma ] .
@xmath0 deceases monotonously with the growing @xmath53 . for given values of other parameters shown in fig .
[ fig : hc2gamma ] , the maxima of @xmath0 are along the nodal directions for small @xmath53 , but along the antinodal directions for large values of @xmath53 .
fixing @xmath121 , @xmath122 , @xmath68 , @xmath71 and @xmath97.,width=288 ] from all these results , we see that both the magnitudes and the detailed angular dependence of in - plane @xmath0 are significantly influenced by a number of physical parameters .
a particularly interesting feature is the fourfold oscillation pattern of angle dependent @xmath0 can be shifted by @xmath18 if one varies any one of these parameters .
@xmath0 may exhibit its maxima along either nodal or antinodal directions , depending on the specific values of relevant parameters , which is apparently in sharp contrast in the naive notion that @xmath0 always displays the same angle dependence of the @xmath1-wave superconducting gap . as aforementioned , in the last several years the
in - plane @xmath0 has been widely investigated with the aim to identify the precise superconducting gap symmetry in two heavy fermion compounds cecoin@xmath6 and cecu@xmath3si@xmath7 . in this subsection
, we make a comparison between our theoretical analysis and some recent experiments of @xmath0 .
our results are valuable to theoretical and experimental research of @xmath0 in two main aspects .
-dependence of @xmath0 with @xmath107 , @xmath69 , @xmath95 , @xmath70 , and @xmath87.,width=288 ] first , one should be very careful when fitting theoretical calculations with experimental data . in the current literature
, it is often taken for granted that the in - plane @xmath0 always exhibits exactly the same angular dependence as that of the superconducting gap . in other words , the maxima of in - plane @xmath0 are believed to be always along the antinodal directions where the @xmath1-wave gap is maximal . according to this seemingly correct relationship
, the superconducting gap symmetry is simply identified as @xmath108-wave ( @xmath9-wave ) if @xmath12 is found to exhibit its maxima at @xmath123 ( @xmath124 ) . however , as showed in our extensive calculations , such a relationship is not always correct . in a pauli - limited _
d_-wave superconductor , the maxima of @xmath0 may be along either the nodal or the antinodal direction , depending on the specific values of a number of physical parameters , as a consequence of the delicate interplay between orbital and pauli effects .
inaccurate and even incorrect conclusions might be drawn if some of these parameters are not properly chosen . in order to identify the precise gap symmetry of cecoin@xmath6 or cecu@xmath3si@xmath3
, one should first choose suitable values for all the relevant parameters before probing the angular dependence of @xmath0 and deducing the gap symmetry from experimental data . among the above six relevant parameters , the temperature @xmath17 is particularly interesting , because in any given compound @xmath17 is the only free parameter and all the other parameters
are fixed at certain values .
our extensive calculations show that there is always a @xmath18 difference between @xmath0 and @xmath1-wave gap at small @xmath17 and that @xmath0 and @xmath1-wave gap always exhibit exactly the same angular dependence once @xmath17 exceeds certain critical value , provided that the gyromagnetic factor @xmath15 is sufficiently large .
it appears that the impact of pauli effect on @xmath0 is much more important at low temperatures than at high temperatures .
if one attempts to deduce the precise gap symmetry by fitting experiments of @xmath0 , it would be better to measure @xmath0 at a series of very different temperatures .
otherwise , incorrect results might be obtained .
second , our results may help to resolve some current controversies with regard to the precise gap symmetry of heavy fermion compounds .
the gap symmetry of cecoin@xmath2 has been investigated extensively by means of various experimental techniques .
settai _ et . al . _
@xcite reported that the maxima of in - plane @xmath0 are along [ 100 ] direction through de haas - van alphen oscillation signal at @xmath125mk .
the cantilever magnetometer measurements at @xmath126mk of murphy _ et .
@xcite observed that the maxima of @xmath0 are along [ 110 ] direction .
bianchi et .
@xcite measured the specific heat and found the maxima of @xmath0 along [ 100 ] direction at temperatures higher than @xmath127k .
after measuring the magnetic field dependence of electric resistivity at @xmath128mk , weickert _ _ et .
al.__@xcite revealed that the maxima of @xmath0 are along [ 100 ] direction .
obviously , there seems to be a discrepancy among the experimental results about the detailed angular dependence of @xmath0 , which is considered as an open puzzle @xcite and complicates the search for the precise gap symmetry .
according to our results , however , probably such a discrepancy does not exist at all , since the maxima of @xmath0 may be along either [ 100 ] or [ 110 ] direction when some of the relevant physical parameters are moderately changed .
in particular , the maxima of @xmath0 can shift by @xmath18 as the temperature increases beyond certain critical value .
notice that the measurements of ref .
@xcite were performed at a temperature as low as @xmath126mk .
there is a good possibility that the position of the maxima of @xmath0 are shifted from [ 110 ] direction at low temperatures to [ 100 ] direction at higher temperatures .
although this possibility needs to be further examined , it should be safe to say that the seemingly contradictory experimental results about in - plane @xmath0 of cecoin@xmath2 may be well consistent with each other .
more careful and more systematical research are required to completely solve this problem .
in this paper , we have performed a detailed and systematical analysis of the unusual behaviors of in - plane upper critical field @xmath0 in the contexts of pauli - limited heavy fermion compounds .
we show that the concrete angular dependence of @xmath0 is determined by a delicate interplay of the orbital and pauli paramagnetic effects .
the most interesting result is that @xmath0 does not necessarily exhibit the same fourfold oscillation pattern as the @xmath1-wave superconducting gap , which is often taken for granted in the literature . for certain values of a series of physical parameters
, @xmath0 may display its maxima along the nodal directions where the superconducting gap vanishes .
we also have compared our theoretical analysis with some current measurements of in - plane @xmath0 in two heavy fermion compounds cecoin@xmath6 and cecu@xmath3si@xmath7 .
the theoretical results presented in this paper impose an important restraint on the determination of the precise gap symmetry of pauli limited @xmath1-wave heavy fermion superconductors by means of measuring the in - plane @xmath0 .
one has to be extremely careful when trying to deduce the gap symmetry from experiments of @xmath0 .
a. d. bianchi , m. kenzelmann , lisa , debeer - schmitt , j. s. white , e. m. forgan , j. mesot , m. zolliker , j. kohlbrecher , r. movshovich , e. d. bauer , j. l. sarreo , z. fisk , c. petrovic , m. r. eskildsen , science * 319 * , 177 ( 2008 ) . | the angular dependence of in - plane upper critical field @xmath0 is widely used to identify the precise gap symmetry of unconventional @xmath1-wave superconductors .
apart from the well studied orbital effect of external magnetic field , @xmath0 is also believed to be strongly influenced by the pauli paramagnetic effect in some heavy fermion compounds . after calculating @xmath0 in the presence of both the orbital and pauli effects
, we find that its concrete angular dependence is determined by the subtle interplay of these two effects .
an interesting and unexpected result is that @xmath0 may exhibit its maximal values along the nodal or antinodal directions , depending on the specific values of a number of physical parameters .
we perform a systematical analysis on how the fourfold oscillation pattern of @xmath0 is affected by these parameters , and then apply our results to understand the recent experiments of @xmath0 in two heavy fermion compounds cecoin@xmath2 and cecu@xmath3si@xmath3 . |
continuous time markov processes with discrete variable have been used to model phenomena over a wide range of fields , such as biology , ecology and stoichiometry @xcite . among them , the system with infinite - state - level represented by a birth - death process have been well studied , however one often encounters the situation where the state - level is finite , such as spike activity of neurons or chemical reaction systems .
recently , a phase - transition like phenomenon caused by the stochastic noise as represented by the autocatalytic reaction system and the foraging behavior of ants has attracting attentions @xcite , where the qualitative behavior of the system changes drastically as the system size becomes small .
these stochastic processes are described by the master equation , however it is in general difficult to analyze the master equation directly .
therefore , one has to resort to an approximation scheme to analyze the system . a well - known approximation scheme to the master equation is the fokker - planck approximation @xcite .
this approximation is able to capture the noise - induced bimodality in the model of foraging behavior of ants @xcite .
however , the fokker - planck approximation is a continuous approximation to intrinsically discrete variables , and it was pointed out that the method may fail to capture some kind of feature of the stochastic system @xcite .
another method , which retains discreteness of the system , is to use the creation and annihilation operators for boson , initiated by doi and peliti @xcite . in this formalism , various statistics can be represented by the boson coherent state path integral , and this formalism gives a framework to apply various methods developed in quantum mechanics .
in fact , the doi - peliti formalism was applied to a biochemical reaction system and models of spike activity of neurons @xcite . however , the doi - peliti formalism is originally the method to treat the infinite level systems such as growth of bacteria or variations of animal populations , and it seems to be not appropriate to treat a finite - level system using this formalism .
in fact , thermodynamic limit was taken in the application to the spike activity of neurons , on the assumption that the number of neuron in each population is infinite @xcite .
such a thermodynamic limit may again lose some features of the stochastic system , and the finite - size effect could not be estimated correctly .
therefore , an approximation scheme which captures the finite - size effect appropriately is desired .
therefore in this paper , we propose a method to treat discrete stochastic processes with finite - state - level .
for this purpose , we use a representation of su(2 ) lie algebra .
we found that by employing an appropriate normalization of state vectors in hilbert space the master equation of the system can be rewritten as a time evolution equation for the state vector corresponding to the probability generating function .
then , the generating function is expressed as a propagator of the spin coherent state .
we derive the spin coherent state path integral representation for the generating function .
a similar formulation was shown previously in ref .
@xcite , however we use a different representation of su(2 ) algebra .
we study the linear susceptible - infected - susceptible ( sis ) epidemic model with arbitrary time - dependent transition probabilities as a simplest application to demonstrate the effectiveness of our method .
we employ two approach , wei - norman theorem and path integral representation , to calculate the generating function .
this paper is organized as follows . in sec .
[ sec : algebra ] , we show how the discrete stochastic system with finite - state - level can be re - expressed by using a representation of su(2 ) algebra . in sec .
[ sec : application ] , we study the stochastic linear sis epidemic model as a simplest application of our method . sec .
[ sec : conclusion ] is devoted to conclusion .
the system of our interest consists of single population with @xmath0 individuals .
each individual can take two discrete states .
for example , the two states correspond to firing and quiescent states in the model of spike activity of neurons , and susceptible and infected states in epidemic model .
thus , @xmath0 individuals are divided into two classes .
the state of the system at a given time @xmath2 is represented by a discrete variable @xmath3 with @xmath1 state - level , which takes @xmath4 .
time - evolution of the system is assumed to be a continuous time markov process with asynchronous updating , that is , only one individual changes the state during infinitesimal time interval . given a state transition probability from @xmath5 to @xmath6 per unit time @xmath7 ,
the probability with the state @xmath5 at time @xmath2 is described by the following master equation ,
@xmath8 expressing the excitation and de - excitation probabilities for each individual as @xmath9 and @xmath10 respectively , the master equation is expressed as @xmath11 f(n-1 ) p(n-1,t)-(n - n)f(n ) p(n , t ) \notag \\ & + ( n+1)g(n+1 ) p(n+1,t ) -n g(n ) p(n , t),\end{aligned}\ ] ] where we have defined @xmath12 .
it is in general difficult to analyze the master equation directly .
therefore , we first consider to rewrite the master equation to more tractable form .
in fact , by using a representation of su(2 ) algebra , the master equation can be rewritten as a time evolution equation for the state vector .
though su(2 ) algebra is very familiar in the field of quantum mechanics , we once again illustrate the algebra here for the purpose of emphasizing the difference to the previous study @xcite and quantum mechanical systems .
su(2 ) algebra is defined by the three generators satisfying the following commutation relations , @xmath13 = 2 j_z , \\ & [ j_z , j_{\pm } ] = \pm j_{\pm } , \end{split}\ ] ] where @xmath14 and @xmath15 are called raising and lowering operators .
next , we introduce the lowest - weight state @xmath16 , which has a normalization @xmath17 and satisfies the action @xmath18 .
the parameter @xmath19 takes an integer or a half - integer . here
, we define the following @xmath1-dimensional orthogonal vectors @xmath20 , @xmath21 it should be noted that the inner - product of this state is not normalized to one , unlike that in quantum mechanics . this normalization is crucial that one applies the su(2 ) algebra to the stochastic systems .
then , the action of the operators @xmath22 and @xmath23 to the state @xmath20 is respectively given by @xmath24 here , we define the state vector corresponding to the probability distribution @xmath25 as @xmath26 as we will see , this state is nothing but the probability generating function .
then , using the action of the operators eq .
( [ eq : operation ] ) , we can rewrite the master equation to a time - evolution equation of this state vector @xmath27 : @xmath28 @xmath29 where @xmath30 is the identity operator and we have defined the hamiltonian @xmath31 of the time evolution . the formal solution of the time evolution equation eq . ( [ eq : time evolution ] ) is given by @xmath32 where @xmath33 denotes the time - ordered product and @xmath34 is a time - evolution operator satisfying the following differential equation @xmath35 in this section , we discuss how various statistics of the probability distribution can be calculated from the state vector @xmath36 . in quantum mechanics , the expectation value of a physical quantity is given by an inner - product of the corresponding operator with respect to the state vector .
on the other hand in stochastic system , the statistics is given by an inner - product of the corresponding operator with respect to the projection vector , as defined in the following .
first , noting the orthonormalization relation @xmath37 , the normalization condition of the probability distribution is calculated by the following inner - product , @xmath38 where @xmath39 is the projection vector . then given the state vector @xmath27 , the expectation value of certain function @xmath40 is given by the inner - product of the corresponding operator @xmath41 as @xmath42 for example , the probability generating function @xmath43 , which corresponds to @xmath44 , is calculated as @xmath45 where the operator @xmath46 is defined by the taylor expansion , @xmath47 . in the same way , the moment generating function , which corresponds to @xmath48 ,
is calculated as @xmath49 seeing from a different perspective , the above calculation indicates that the state vector @xmath27 is nothing but the probability generating function .
in fact , if we define the unnormalized spin coherent state as @xmath50 then , the coherent state representation of the state vector @xmath27 is given by @xmath51 this is the definition of the probability generating function . here , we mention the difference between our su(2 ) method and that in the previous study . in the previous study , a non - hermitian representation of the su(2 )
operators was used to realize the operator action eq .
( [ eq : operation ] ) @xcite , instead of the modification of the state normalization . that is , the lowering operator is not hermite conjugate to the raising operator , @xmath52 .
this non - hermiticity makes a complicated situation that the coherent state @xmath53 is not adjoint to the coherent state @xmath54 . on the other hand in our formalism
, the hermiticity is retained .
therefore , it is anticipated that in our formalism one can readily apply the methods developed in quantum mechanics and calculation would become concise .
this is demonstrated in the following sections .
let us consider the case of the binomial distribution : @xmath55 the binomial distribution is represented by the state vector @xmath56 as @xmath57 where in the last line we have set @xmath58 .
that is , the binomial distribution can be represented by an unnormalized spin coherent state .
then when the initial state at time @xmath59 is given by the binomial distribution , the probability generating function @xmath60 of the system at an arbitrary time @xmath61 is given as an equivalent form to the spin coherent state propagator as follows , @xmath62 the case where the initial state has a specific number can be treated in a similar way .
in fact , since the specific number state can be represented as @xmath63 , one can express the specific number state as a derivative of the binomial state , @xmath64 therefore , the generating function @xmath65 with the initial state @xmath20 at time @xmath61 can be expressed as @xmath66 since the probability generating function is expressed as a propagator of the spin coherent state , this generating function can be expressed as a path integral form dividing the time - evolution operator ( see appendix [ sec : path integral ] ) @xmath67 \bigr\}.\ ] ] @xmath68 $ ] is the action functional @xmath69 = j \ln ( 1+x z(t ) ) + j \ln ( 1+\bar{z}(0 ) y ) + \int_0^t \mathrm{d } t \biggl\ { j \frac { \dot{\bar{z } } z - \bar{z } \dot{z } } { 1+\bar{z } z } + h(\bar{z } , z ) \biggr\},\ ] ] where the overdot denotes time derivative and the variables @xmath70 and @xmath71 have the following boundary conditions , @xmath72 hamiltonian @xmath73 is given by a matrix element of the operator @xmath31 with respect to the unnormalized spin coherent state @xmath74 where @xmath75 .
specific matrix elements of lower powers of operators are calculated in appendix [ sec : path integral ] .
in this section , we apply our formalism to the simplest stochastic system with finite - state - level .
this is the linear sis epidemic model . in this model
, each individual can take either susceptible state @xmath76 or infected state @xmath77 .
we take the state of the system @xmath3 as the number of infected individuals .
therefore , the number of susceptible individuals is @xmath78 .
transition probabilities of this model is given as follows , @xmath79 that is , each susceptible individual is infected with a probability @xmath80 per unit time , and transitions to the infected state . on the other hand , an individual in the infected state recovers from the infection disease with a probability @xmath81 per unit time , and transitions to the susceptible state .
immunity is not taken into account upon recover from the infection .
thus , recovered individuals return immediately to the susceptible state .
we allow the transition probabilities @xmath80 and @xmath81 to have arbitrary time - dependences for generality .
a natural specific assignment of the recover probability may be to set @xmath82 as time - independent .
the infection probability @xmath80 may have periodicity to reflect the seasonal epidemic . then
, the corresponding master equation is given by @xmath83 it should be noted that this is a linear model , where the allowed infection process is @xmath84 , and the process @xmath85 have not been considered .
thus , our model is different to the usual sis epidemic model @xcite .
the hamiltonian corresponding to the master equation is given as a linear combination of the su(2 ) operators , @xmath86 here , we analyze the linear sis model by an algebraic method .
a similar attempt has been already made in the previous study , however the analysis has been restricted to a specific initial state or a specific transition probability @xcite .
therefore , we make a complete analysis here . from the algebraic property of this system , the time - evolution operator @xmath34 can be expressed as a decomposed exponential form ( see appendix [ wei - norman ] ) @xcite : @xmath87 where each function @xmath88 is respectively given by @xmath89 @xmath90 @xmath91 @xmath92 @xmath93 we first consider the case of the binomial initial state at initial time @xmath59 , @xmath94 where @xmath95 . by using formula of su(2 ) operators ( see appendix [ sec : operator ] ) , the probability generating function @xmath60 at time @xmath61 is calculated as follows , @xmath96 } e^{(x + g_1 ) j_- } |0 \rangle \notag \\ = & ( 1+y)^{-n } e^{g_0}e^ { -j g_2 } \bigl[1 + ( x+g_1 ) ( y + g_3 ) e^{g_2 } \bigr]^{n},\end{aligned}\ ] ] where in the third and fourth line we have used the ordering formulae eqs .
( [ eq : order z+ ] ) , ( [ eq : order -+ ] ) , and used the decomposition formula eq .
( [ eq : decomposition ] ) in the last line . using the explicit expressions for @xmath97
, we finally arrive at the following expression for the probability generating function , @xmath98^n \notag \\ \equiv & \bigl[e^ { d } ( w + \theta_i ) x + 1- e^ { d } ( w + \theta_i ) \bigr]^n.\end{aligned}\ ] ] this is the probability generating function of the binomial distribution with the rate parameter given by @xmath99 .
thus , when the initial distribution is given by the binomial distribution , probability distribution retains the shape of the binomial distribution , and time - dependence in the transition probabilities only affects its rate parameter to be time - dependent .
this result seems plausible , reflecting the fact that the hamiltonian eq .
( [ eq : sis hamiltonian ] ) takes the most general form of the coherence - preserving hamiltonian in quantum mechanics , where the minimum uncertainty of the quantum fluctuation is retained in the course of a time - evolution . the moment
generating function can also be calculated in the same way , changing @xmath100 in the probability generating function to @xmath101 , @xmath102^n.\ ] ] next , we consider the case where the initial state is a specific number state @xmath103 . in this case , the generating function is obtained by differentiating that of the binomial initial state @xmath104^n \bigr|_{y=0 } \notag \\ = & \bigl[e^ { d } ( 1 + w ) x + 1- e^ { d } ( 1 + w ) \bigr]^m \bigl[e^ { d } w x + 1- e^ { d } w \bigr]^{n - m } \notag \\ \equiv & \bigl [ \theta_1 x + 1- \theta_1 \bigr]^m \bigl [ \theta_0 x + 1- \theta_0 \bigr]^{n - m},\end{aligned}\ ] ] where we have defined two rate parameters of the binomial states , @xmath105 and @xmath106 .
the generating function consists of the product of that of the binomial distributions .
if a sufficient time passes and information on the initial state vanishes , the probability generating function converges to that of a single binomial distribution .
@xmath107^{n},\ ] ] now we calculate the probability distribution explicitly . this is done by dividing cases according to the magnitude relation of @xmath108 .
the results are the following : @xmath109 , \\ & \!\!\!\!\!\!\!\ !
\sum_{l=0}^{n - m } { m \choose n - l } \theta_1^{n - l } ( 1-\theta_1)^{m - n+l } { n - m \choose l } \theta_0^{l } ( 1-\theta_0)^{n - m - l } , \ : [ n - m < n \le m ] , \\ & \!\!\!\!\!\!\!\ ! \sum_{l=0}^{m } { m \choose l } \theta_1^{l } ( 1-\theta_1)^{m - l } { n - m \choose
n - l } \theta_0^{n - l } ( 1-\theta_0)^{n - m - n+l } , \;\;\;\;\;\;\;\ ; [ m < n ] .
\end{dcases } \end{split}\ ] ] in this section , we derive the probability generating function by evaluating the path integral directly to check the consistency of our formalism . here
, we consider the case with time - independent transition probability for simplicity .
the generating function has a following path integral form , @xmath67 \bigr\}.\ ] ] the matrix element of the hamiltonian @xmath73 in the action corresponding to eq .
( [ eq : sis hamiltonian ] ) is given by ( see appendix [ sec : path integral ] ) , @xmath110 now we consider the `` classical '' trajectories of the variables @xmath70 , @xmath71 of this system , which correspond to the extremal trajectories of the action .
these variables obey the following hamilton s equations respectively , @xmath111 this the riccati differential equation , therefore the solutions satisfying the boundary conditions @xmath112 and @xmath113 are respectively given by , @xmath114\end{aligned}\ ] ] @xmath115\end{aligned}\ ] ] the `` classical '' action is obtained by inserting the `` classical '' solutions into the action functional , @xmath116 e^{-wt } + 1 \biggr\ } + 2j\ln(1+y)\end{aligned}\ ] ] using the `` classical '' action , we obtain the `` classical '' approximation to the probability generating function , @xmath117 e^{-wt } + 1 \biggr\}^{n}.\end{aligned}\ ] ] this is precisely the generating function of the binomial distribution .
thus , we have confirmed that the `` classical '' approximation gives the exact result in the linear sis model , @xmath118 , and that our path integral representation gives the correct result .
in this paper , we propose the method to analyze a discrete stochastic process with finite - state - level . by using a representation of su(2 ) algebra , the master equation of the system
is rewritten to a time evolution equation for the state vector corresponding to the probability generating function .
then , we found that the generating function can be expressed as a propagator of the spin coherent state .
furthermore , the generating function has a path integral representation of the spin coherent state . also , we found the correspondence between coherent state and the probability distribution . that is ,
the generating function of the binomial distribution is expressed by the spin coherent state . as a simplest application of our method ,
a linear sis epidemic model with arbitrary time - dependent transition probabilities was analyzed .
the generating function of the system was calculated using underling algebraic property of the system or a path integral representation .
we considered two initial states , the binomial state and the specific number state . after a sufficient time passed
, the generating function converges to that of the binomial distribution in both cases , where the time - dependent transition probabilities only affect the rate parameter of the binomial distribution .
the model analyzed in this paper is the too simple system in which the interactions between each individual are ignored , and the result is little bit trivial .
however , our results indicate that our formalism is suited for analyzing the finite - level system and gives basis for applying the method developed in quantum mechanics to more complicated stochastic systems .
one example of such an approximation scheme is the wentzel - kramers - brillouin ( wkb ) approximation to the propagator using a spin coherent state path integral @xcite .
it is of significance to see how such an approximation scheme captures the finite - size effect .
in addition , it is of interest to further investigate the relationship between probability distributions and generalized coherent states .
the relationship will be helpful in analyzing the corresponding stochastic systems as illustrated in this paper .
in this appendix , we derive the spin coherent state path integral representation to the probability generating function in detail for the purpose of specifying our notation . in the path integral
formalism , one first divides the time evolution operator @xmath34 into that of the short - time interval , @xmath119 then , inserting the @xmath120 over - completeness relations in terms of spin coherent state between the @xmath0 short - term time evolution operators , @xmath121 the probability generating function can be expressed as product of @xmath0 short - term propagators as follows , @xmath122 \notag \\ & \langle z_n |\;\!\!|\bigl(1+\frac { \hat{h } t}{n } \bigr ) |\;\!\!| z_{n-1 } \rangle \cdots \langle z_k |\;\!\!| \bigl(1+\frac { \hat{h } t}{n } \bigr ) |\;\!\!| z_{k-1 } \rangle \cdots \langle z_1 |\;\!\!| \bigl(1+\frac { \hat{h } t}{n } \bigr ) |\;\!\!| z_0\rangle \notag \\ = & ( 1+y)^{-n } \lim_{n \rightarrow \infty } \biggl [ \prod_{i=1}^{n-1 } \frac{2j+1}{\pi } \int \frac{\mathrm{d } \bar{z}_i \mathrm{d } z_i}{(1+\bar{z}_i z_i)^{2 } } \biggr ] ( 1+\bar{z}_n z_n)^j ( 1+\bar{z}_0 z_0)^j \notag \\ & \prod_{k=1}^n \frac{(1+\bar{z}_k z_{k-1})^{2j } } { ( 1+\bar{z}_k z_k)^j ( 1+\bar{z}_{k-1 } z_{k-1 } ) ^j } \bigl(1+\frac{h(\bar{z}_k , z_{k-1 } ) t}{n } \bigr),\end{aligned}\ ] ] where @xmath123 is a matrix element of the hamiltonian , and we have set @xmath124 , @xmath125 . here ,
if we assume @xmath126 as a small quantity of order @xmath127 , the following approximation can be made , @xmath128 re - expressing the above @xmath129 terms into the exponential function , we finally arrive at the following path integral representation to the probability generating function , @xmath130 \bigr\},\ ] ] where the integration measure is defined by @xmath131.\ ] ] the action functional is given by @xmath69 = j \ln ( 1+x z(t ) ) + j \ln ( 1+\bar{z}(0 ) y ) + \int_0^t \mathrm{d } t \biggl\ { j \frac { \dot{\bar{z } } z - \bar{z } \dot{z } } { 1+\bar{z } z } + h(\bar{z } , z ) \biggr\},\ ] ] where the variables @xmath70 and @xmath71 have the boundary conditions @xmath112 and @xmath113 .
the hamiltonian @xmath73 is defined as a matrix element of the operator @xmath31 .
for example , matrix elements lower powers of operators are calculated as follows .
the first - order operators are respectively given by @xmath132 the second - order operators are respectively given by @xmath133 @xmath134 `` classical '' trajectories , which correspond the extremal trajectories of the action , obey the following hamilton s equations , @xmath135
time evolution operator @xmath34 obeys the following differential equation , @xmath136 where we have set the coefficients @xmath137 and the operators @xmath138 as , @xmath139 wei - norman theorem states that the solution of the above differential equation takes a product of exponential operators @xcite , @xmath140 where @xmath97 satisfy the initial conditions @xmath141 .
the functions @xmath97 can be derived by differentiating eq .
( [ eq : wei - norman ] ) with respect to time @xmath2 . in fact ,
the time derivative of the time - evolution operator eq .
( [ eq : wei - norman ] ) takes the following form , @xmath142 with the aid of the baker - campbell - hausdorff formula , @xmath143 where the adjoint operator @xmath144 is defined as @xmath145 \notag \\ = & h_i h_j -h_j h_i.\end{aligned}\ ] ] using the su(2 ) adjoint operators , @xmath146 we obtain @xmath147 comparing the above equation to the original equation eq .
( [ eq : unitary evolution ] ) , we find that @xmath148 are the solutions of the following differential equations @xmath149 the differential equation for @xmath150 is the riccati differential equation .
therefore , the solutions are respectively given by , @xmath151 @xmath152 @xmath153 @xmath154 where @xmath155 @xmath156 and we have taken the integration constants so as to satisfy the initial conditions @xmath141 .
operators satisfying the su(2 ) algebra @xmath157=\pm j_{\pm}$ ] , @xmath158= 2 j_z$ ] have a following decomposition formula @xcite , @xmath159 where the parameter constants are respectively given by @xmath160 @xmath161 ^ 2},\ ] ] @xmath162 su(2 ) operators have the exchanging formulae .
in fact , using the formula for exponential adjoint operators eqs .
( [ eq : adjoint1 ] ) , ( [ eq : adjoint2 ] ) , exponential operators can be exchanged as follows , @xmath164 @xmath165 buice m.a . ,
cowan j.d . ,
chow c.c . : systematic fluctuation expansion for neural network activity equation .
neural computation * 22 * , 377 - 426 ( 2010 ) doi , m. : second quantization representation for classical many - particle system . j. phys . a * 9 * , 1465 - 1477 ( 1976 ) doi , m. : stochastic theory of diffusion - controlled reaction . j. phys .
a * 9 * , 1479 - 1495 ( 1976 ) | we propose a formalism to analyze discrete stochastic processes with finite - state - level @xmath0 . by using an @xmath1-dimensional representation of su(2 )
lie algebra , we re - express the master equation to a time - evolution equation for the state vector corresponding to the probability generating function .
we found that the generating function of the system can be expressed as a propagator in the spin coherent state representation .
the generating function has a path integral representation in terms of the spin coherent state .
we apply our formalism to a linear susceptible - infected - susceptible ( sis ) epidemic model with time - dependent transition probabilities .
the probability generating function of the system is calculated concisely using an algebraic property of the system or a path integral representation .
our results indicate that the method of analysis developed in the field of quantum mechanics is applicable to discrete stochastic processes with finite - state - level . |
in this article we review recent heavy flavour physics results from the tevatron @xmath2 collider at fermilab , where heavy flavour refers to the top quark as well as the bottom quark . after a brief historical overview
we summarize the status of top quark physics at cdf and d in sec .
[ top_sec ] .
in particular we discuss the measurement of the top production cross section and the top quark mass .
section [ b_sec ] is devoted to recent @xmath0 physics results at a hadron collider , where we concentrate on @xmath0 hadron lifetimes and latest time dependent @xmath1 mixing results .
a brief look to the future , summarizing the prospects of top quark physics as well as @xmath0 physics and @xmath3 violation in run ii starting in 1999 , is given in sec .
[ outlook_sec ] .
we conclude with section [ conclude_sec ] . in 1977
the bottom quark was discovered as a resonance in the dimuon mass spectrum in 400 gev proton - nucleus collisions at fermilab @xcite .
soon after the discovery the so - called @xmath4 resonances at a mass of about 9.5 were confirmed in @xmath5 collisions at the doris storage ring at desy @xcite . from the narrow resonance observed at @xmath6
the pluto and dasp detectors were able to determine the electronic width @xmath7 , which implied a charge assignment of @xmath8 for the @xmath9 quark .
it took a few years until decays of @xmath0 mesons , bound states of a @xmath9 quark and a light quark , were observed in 1983 by the cleo collaboration @xcite .
the weak isospin of the @xmath9 quark was first extracted from the forward - backward asymmetry @xmath10 in @xmath11 at petra @xcite , where the measurement of @xmath12% at @xmath13 gev was found to be consistent with the standard model prediction of @xmath14% assuming @xmath15 for the weak isospin of the @xmath9 quark .
this indicated that the bottom quark has a partner and it can be counted as the first evidence for the existence of the top quark .
finally , concluding this brief historical overview with a link between top and @xmath0 physics , @xmath16 mixing was first observed by the argus collaboration in 1987 @xcite .
the measurement of a large mixing parameter @xmath17 was the first indication of a large top quark mass .
why are we discussing the history of @xmath9 quark physics ?
a certain pattern can be observed after the discovery of a new quark .
this pattern appears to repeat itself after the discovery of the top quark .
it looks like right after the discovery of the @xmath9 quark an attempt was made to confirm the discovery in a different environment like @xmath18 collisions .
then the fundamental quantities of the newly discovered quark , like its charge and isospin , were determined .
after sufficient statistics is accumulated rare phenomena like @xmath19 oscillations were searched for . nowadays , @xmath0 physics
is fully explored at @xmath5 colliders and rare phenomena like @xmath20 transitions or @xmath21 penguin decays are also studied @xcite .
this raises the question of why @xmath0 physics is studied at a @xmath2 hadron collider which is a much more difficult environment to study low p_t @xmath22physics .
we shall discuss this issue further on , in section [ b_sec ] .
the experimental search for the top quark begun soon after the discovery of the @xmath9 quark . between 1979 and 1984 measurements of @xmath23 , the ratio of the cross sections @xmath24 hadrons ) to @xmath25 ,
were performed at the petra @xmath18 storage ring up to a centre - of - mass energy of 46.8 gev .
the value of @xmath23 was found to be consistent with standard model predictions without a top quark contribution setting a lower bound on the top quark mass @xcite of @xmath26 .
later searches at @xmath5 colliders were also negative and limits on m_top @xmath27 of half of the centre - of - mass energy were set .
these limits from tristan @xcite as well as slc @xcite and lep @xcite are listed in table [ top_limits ] . with the advantage of higher mass regions being accessible ,
the search for the top quark was soon dominated by @xmath2 colliders , first at the sps collider ( @xmath28 gev ) at cern and then at the tevatron ( @xmath29 tev ) at fermilab .
initial results reported in 1984 by the ua1 collaboration @xcite at the sps seemed to be consistent with the production of a top quark of mass @xmath30 in @xmath31 .
the results were based on the observation of 12 isolated lepton plus 2-jet events with an expected background of approximately 3.5 events in 200 @xmath32b@xmath33 . however , these first results were not supported by a subsequent ua1 analysis @xcite with a higher statistics data sample setting a lower limit on m_top @xmath27 of @xmath34 .
more sensitive searches were performed later on by ua2 @xcite and with the start of the tevatron .
cdf @xcite and d @xcite increased the limit on the mass of the top quark to finally @xmath35 in 1994 ( see table [ top_limits ] ) .
.historical overview of searches for the top quark and limits on the top quark mass .
[ cols="^,>,<,^,^",options="header " , ] from a fit to the like - sign lepton fraction as a function of @xmath36 the mixing frequency @xmath37 is extracted as shown in fig .
[ bmixemu]b ) .
the fit includes components for direct and sequential @xmath9 decays , @xmath38 , and fake events . in about 16% of the events with a secondary vertex around one lepton
, a secondary vertex is also found around the other lepton .
these events enter the like - sign fraction distribution twice and we allow for a statistical correlation between the two entries .
the final fit result is @xmath39 where the dominant systematic error arises from the uncertainty in the sample composition . for this analysis @xmath0 mesons
are reconstructed through their semileptonic decays @xmath41 ( see also sec .
[ partblife ] ) .
the analysis starts with the single lepton trigger data and reconstructs @xmath42 meson candidates in a cone around the trigger electron or muon in the following channels : @xmath43 tracks with impact parameters significantly displaced from the primary vertex are selected in order to decrease combinatorial backgrounds .
the signals are identified as peaks in the invariant mass spectra of the charm decay products as shown in fig .
[ bmixdlep ] for the different @xmath42 decay modes .
( 145,70)(0,0 ) ( 10,60)*(a ) * ( 44,60)*(b ) * ( 10,25)*(c ) * ( 44,25)*(d ) * ( 87,61)*(e ) * = 7.0 cm = 7.0 cm the @xmath42 candidates are intersected with the lepton to find the @xmath0 decay vertex .
since the @xmath0 meson is not fully reconstructed , its @xmath44 can not be directly obtained .
the boost of the @xmath0 meson is determined from the observed decay products and a @xmath45 correction is applied as obtained from a monte carlo simulation . in order to tag the @xmath0 flavour at production
, we use a ` same side tagging ' ( sst ) algorithm , which exploits the correlation between the @xmath0 flavour and the charge of tracks from either the fragmentation process or @xmath46 resonances @xcite . in this analysis
no attempt is made to differentiate the sources of correlated pions .
to study the correlation between the flavour of the @xmath0 meson and the charged particles produced in association with it , we consider all tracks that are within an @xmath47-@xmath48 cone of radius @xmath49 centered around the direction of the @xmath0 candidate . since the @xmath0 meson is only partially reconstructed , we approximate this direction with the momentum sum of the lepton and charm hadron
. the tracks considered as tags should be consistent with the hypothesis that they originate from the fragmentation chain or the decay of @xmath46 mesons , i.e. that they originate from the primary vertex of the event .
all tracks with transverse momentum @xmath50
gev/@xmath51 are therefore required to satisfy @xmath52 , where @xmath53 is the distance of closest approach of the track trajectory to the estimated @xmath0 production position , and @xmath54 is the estimated error on this quantity .
string fragmentation models indicate that the velocity of the fragmentation particles , that we seek for our tag , is close to the velocity of the @xmath0 meson .
similarly , pions from @xmath46 decays should also have a velocity that is close to the velocity of the @xmath0 meson .
in particular , the relative - transverse momentum ( @xmath55 ) of the particle with respect to the combined momentum of the @xmath0 momentum plus particle momentum , should be small .
of the candidate tracks , we select as the tag the track that has the minimum component of momentum @xmath55 orthogonal to the momentum sum of that track , the lepton , and the @xmath56 meson . the efficiency for finding such a tag is about @xmath57 for this algorithm .
since we know the flavour of the @xmath0 meson at decay from the @xmath58 signature , we compare the number of right - sign ( @xmath59 ) correlations to the number of wrong - sign ( @xmath60 ) tags as a function of @xmath36 . for the @xmath61 meson
we expect the asymmetry @xmath62 to be : @xmath63 where @xmath56 is the dilution of the same side tagging algorithm .
@xmath56 is also often expressed in terms of the mistag fraction @xmath64 as @xmath65 . in our analysis
we fit for both @xmath37 and @xmath56 . to obtain the asymmetry for @xmath61 and @xmath66 mesons , we correct for the fact that each @xmath58 signal has contributions from both neutral and charged @xmath0 mesons via @xmath67 decays .
we correct for this cross talk by performing a fit bin by bin in @xmath36 .
the inputs to the fit are the raw asymmetries as measured in each sample for a given @xmath36 bin , and parameters describing the @xmath67 composition in semileptonic @xmath0 decays .
we fit the corrected @xmath61 asymmetry as a function of @xmath36 to a cosine convoluted with the @xmath36 resolution function , and extract the mixing frequency @xmath37 as well as the dilution @xmath56 of the same side tagging algorithm as shown in fig .
[ bmix : petar_owen]a ) .
we also determine the asymmetry for the charged @xmath0 meson , which is flat in @xmath36 as expected .
we measure @xmath68 and find @xmath69 .
we also determine the effective tagging efficiency to be @xmath70 .
the dominant systematic error arises from the uncertainty in the fraction of @xmath67 in semileptonic @xmath0 decays .
( 145,70)(0,0 ) ( 63,60)*(a ) * ( 96,54)*(b ) * = 7.0 cm = 7.5 cm in summary , we have reported two measurements of @xmath37 at cdf using @xmath58 events with a same side tag as well as @xmath71 dilepton events .
there are more time dependent @xmath1 oscillation measurements in preparation at cdf , like a mixing analysis using a soft lepton tag ( see sec .
[ toplepjet ] ) and a jet charge tag as shown in fig .
[ bmix : petar_owen]b ) .
the result from this analysis is @xmath72 since there is an event overlap of about 10% between this analysis and the @xmath71 analysis , a combined cdf average has not yet been determined , but cdf s mixing results start to become competitive with the lep measurements @xcite .
in run i the luminosity of the tevatron was limited by the antiproton current .
the fermilab accelerator complex is undergoing an upgrade to produce an order of magnitude higher luminosities in the tevatron . the largest change will be to replace the main ring with the new main injector , which will be housed in a new tunnel . the main injector will provide higher proton intensity onto the antiproton production target , and larger aperture for the antiproton transfer into the tevatron . after the completion and comissioning of the main injector ,
the tevatron is scheduled to deliver luminosity again in summer of 1999 .
the centre - of - mass energy will then be at 2.0 tev .
luminosities of @xmath73 @xmath74s@xmath33 will be reached yielding an integrated luminosity of 2 @xmath75b@xmath33 delivered to the collider experiments within two years .
the physics projections for run ii presented here assume 2 @xmath75b@xmath33 of integrated luminosity .
in addition to the tevatron upgrade the cdf and d experiments will undergo major detector upgrades , which are described in detail elsewhere @xcite .
in run ii both collider detectors will be equipped with new and improved silicon vertex detectors which will for example enhance the efficiency to detect top decays .
in addition , the tevatron s centre - of - mass energy will be at 2.0 tev , which will increase the top quark production cross section by 40% .
the estimated yield per experiment will be about 160 dilepton events , 990 lepton plus 4-jet events with one or more @xmath9 tags and for the top mass measurement about 240 lepton plus 4-jet events where both @xmath9 jets are tagged .
this will result in a statistical uncertainty on the top quark mass measurement of 1 - 2 .
the top production cross section will be known with a precision of better than 10% .
in addition , single top production can be studied , top polarization measurements can be performed , and rare top decays will be searched for .
a more detailed overview of top quark physics in run ii can be found in ref .
@xcite in this section we concentrate on reviewing the future of @xmath0 physics at cdf by summarizing the prospects of measuring @xmath3 violation in run ii .
the @xmath0 physics goal is to measure the @xmath3 asymmetry in b^0 j / k^0_s @xmath76 and b^0 ^+ ^- @xmath77 determining 2 @xmath78 and 2 @xmath79 , respectively . cdf plans to also look for @xmath3 violation in @xmath80 and @xmath81 probing 2 @xmath82 .
cdf has the advantage of being an existing experiment that took plenty of data in run i. we can use these data to study the ingredients for a future @xmath3 violation measurement eg . in b^0 j
/ k^0_s @xmath76 .
one input is the knowledge of the expected number of @xmath83 events , which can be extrapolated from the @xmath83 yield in cdf s current data .
the second ingredient is the knowledge of the @xmath0 flavour at production . for this purpose cdf studies several @xmath0 flavour tagging methods at a hadron collider environment .
the figure of merit to compare flavour tagging algorithms is the effective tagging efficiency d^2 @xmath84 , where @xmath85 is the efficiency of how often a flavour tag is applicable and @xmath56 is the dilution @xmath86 . here ,
@xmath59 and @xmath60 are the numbers of right and wrong sign tags , while @xmath64 is the mistag fraction ( see also sec .
[ sec : sst ] ) . figure [ runik0s ] shows the invariant @xmath83 mass distribution from cdf s current data corresponding to about 110 @xmath87b@xmath33 .
about 240 signal events with a signal - to - noise ratio better than 1:1 are observed .
this is currently the world s largest sample of b^0 j / k^0_s @xmath76 and serves as proof that b^0 j / k^0_s @xmath76 decays can be fully reconstructed in a hadron collider environment in a well understood way .
( 145,70)(0,0 ) = 7.0 cm several @xmath0 flavour tagging methods are studied with cdf data .
one is referred to as ` same side tagging ' ( sst ) which is described in more detail in sec . [ sec : sst ] .
this method exploits charge correlations between @xmath0 mesons and charged particles produced in the fragmentation of the @xmath9 quarks .
such correlations are expected to arise from particles produced in the fragmentation chain and from decays of the @xmath88 @xmath0 mesons ( the @xmath46 mesons ) @xcite .
another way of tagging the flavour of a @xmath0 meson at production is to exploit the flavour of the other @xmath0 meson in the event .
this can be done through a @xmath0 semileptonic decay ( lepton tagging ) or by counting the charge of the other @xmath9 jet ( jet charge tagging ) .
cdf has preliminary results on the effective tagging efficiency of these three methods . using a high statistics sample of partially reconstructed @xmath0 mesons from @xmath89 ( @xmath66 signature ) and @xmath90 ( @xmath61 signature ) combinations ,
charge correlations between the @xmath0 candidate and tracks in its vicinity are studied as detailed in sec .
[ sec : sst ] . in the case of charged @xmath0 mesons
we measure @xmath91 ( @xmath92 ) for the same side tagging algorithm .
for the neutral @xmath0 meson we obtain @xmath93 with @xmath94 . the proof that this tagging method actually works is shown in fig .
[ bmix : petar_owen]a ) , where the asymmetry of right minus wrong sign tags normalized to the sum of both is plotted versus the @xmath36 of the @xmath0 meson .
the top plot is for @xmath66 candidates and shows a flat behavior as expected .
the bottom plots shows the same distribution for @xmath61 mesons , and exhibits an oscillatory behaviour as expected from @xmath19 mixing .
this means applying same side tagging to a sample of partially reconstructed @xmath0 mesons results in a measurement of time dependent @xmath61 mixing .
we also studied same side tagging with a sample of fully reconstructed @xmath95 decays .
the results of @xmath96 ( @xmath97 ) and @xmath98 ( @xmath99 ) for @xmath66 and @xmath61 mesons , respectively , are somewhat limited by statistics .
cdf also studied opposite side lepton tagging and obtains an effective tagging efficiency of @xmath100 for soft muon tagging and @xmath101 for soft electron tagging .
the tagging algorithms are similar to the ones used in the top quark search as described in sec .
[ toplepjet ] .
the preliminary result for the effective tagging efficieny of jet charge tagging is @xmath102 , which makes use of information from the silicon vertex detector .
the proof that jet charge and soft lepton tagging actually works is shown in fig .
[ bmix : petar_owen]b ) , where the time dependent @xmath0 mixing measurement using these two tags is displayed . combining the existing measurements on @xmath0 flavour tagging from run i results in an effective tagging efficiency of @xmath103 . for the measurement of 2
@xmath78 in b^0 j / k^0_s @xmath76 cdf expects about 15,000 @xmath83 events .
this event number will be obtained with a lower muon trigger threshold of @xmath105 compared to about 2.0 in run i , improved muon coverage , and by also triggering on @xmath106 .
the effective tagging efficiencies are expected to improve with the upgraded detector .
we expect @xmath107 for lepton tagging due to a better coverage for the lepton identification .
for same side tagging we expect @xmath107 from a cleaner selection of fragmentation tracks with svx ii .
finally , for jet charge tagging we expect @xmath108 from an improved purity of the algorithm with 3-dimensional vertexing and the extented coverage of svx ii .
considering the overlap of all three tags by combining them , we expect a total d^2 @xmath84 of about 5.5% resulting in an uncertainty on 2 @xmath78 of @xmath109 .
the key to measure the @xmath3 asymmetry in b^0 ^+ ^- @xmath77 is to trigger on this decay mode in hadronic collisions .
cdf plans to do this with a three level trigger system . on level 1 two
oppositely charged tracks with @xmath111 found with a fast track processor yield an accept rate of about 16 khz .
this will be reduced to about 20 hz on level 2 using impact parameter information ( @xmath112 m ) . on level 3
the full event information is available further reducing the trigger rate to about 1 hz . with this trigger
we expect about 10,000 b^0 ^+ ^- @xmath77 events in 2 @xmath75b@xmath33 .
assuming the same effective tagging efficiency of @xmath113 we expect an uncertainty on 2 @xmath79 of @xmath114 . backgrounds from @xmath115 and @xmath116 decays can be extracted from the untagged signal by making use of the invariant mass distribution as well as cdf s d@xmath117/d@xmath118 capability in the central tracking chamber .
the @xmath3 asymmetry in 2 @xmath82 completes the test of the unitarity triangle .
the angle @xmath120 can be probed via the decay @xmath121 and @xmath122 , where both the mixed and the unmixed amplitudes can decay to the same final @xmath123 state . the interference between both amplitudes results in the weak phase @xmath120 , but also in a relative qcd phase which is expected to be small but a priori unknown .
this measurement requires an all hadronic trigger with similar requirements as for b^0 ^+ ^- @xmath77 .
we expect an overall efficiency times acceptance of @xmath124 . because of
rapid @xmath125 oscillations a time dependent analysis is required but only a small sample of tagged events is expected at cdf in run ii .
the @xmath3 asymmetry in 2 @xmath82 can also be explored via @xmath81 . in this case
the decays are self tagging and a time integrated analysis can be performed , but the theoretical uncertainties are large .
thus , performing a measurement of 2 @xmath82 through both modes will be a challenge at cdf in run ii .
in this article we have reviewed recent heavy flavour physics results from the tevatron @xmath2 collider at fermilab .
we summarized the status of top quark physics at cdf and d .
the top production cross section has been measured to be @xmath126 and the top quark mass is know with a precision of @xmath127 we also discussed recent @xmath0 physics results from the cdf collaboration .
we summarized cdf s @xmath0 hadron lifetime measurements , which are very competitive with the lep and slc results , and discussed latest time dependent @xmath1 mixing results from cdf .
we also reviewed future prospects of top and @xmath0 physics at the tevatron . in run
ii , starting in 1999 , both experiments will measure the top quark mass with a statistical uncertainty of 1 - 2 . the top production cross section will be known with a precision of better than 10% .
the future prospects of @xmath0 physics at cdf will concentrate on the discovery of @xmath3 violation in the @xmath0 system .
cdf expects to measure 2 @xmath78 in b^0 j / k^0_s @xmath76 with a precision of @xmath109 .
cdf will also search for @xmath3 violation in b^0 ^+ ^- @xmath77 and expects to measure 2 @xmath79 with a precision of @xmath114 .
a measurement of 2 @xmath82 will be challenging in run ii .
it is a pleasure to thank all friends and colleagues from the cdf and d collaboration for their excellent work and their help in preparing this talk .
special thanks go to l. galtieri , j. lys , m. shapiro , and b. winer
. a constant source of inspiration and support is my wife ann , who i would like to thank for her continuous understanding about the life of a physicist .
i would like to thank the cdf and d technical support staff at all cdf and d institutions for their hard work and dedication .
i also thank the fermilab accelerator division for their hard and successful work in commissioning the machine for this physics run .
the cdf and d experiments are supported by the u.s .
department of energy ; the u.s . national science foundation ; the istituto nazionale di fisica nucleare , italy ; the ministry of science , culture , and education of japan ; the natural sciences and engineering research council of canada ; the national science council of the republic of china ; the commissariat lenergie atomique in france , the ministry for atomic energy and the ministry of science and technology policy in russia , cnpq in brazil , the departments of atomic energy and science and education in india , colciencias in colombia , cona - cyt in mexico , the ministry of education , research foundation and kosef in korea , and the a. p. sloan foundation .
i gratefully acknowledge the organizers of this inspiring meeting which was a genuine pleasure to attend .
i am sorry for the late submission of this manuscript but as albert einstein already pointed out , time is relative : _ `` put your hand on a hot stove for a minute , and it seems like an hour . sit with a pretty girl for an hour , and it seems like a minute . that s relativity . '' _
herb et al .
, phys.rev.lett .
* 39 * ( 1977 ) 252 .
berger et al .
( pluto ) , phys.lett .
* b76 * ( 1978 ) 243 ; + c.w .
darden et al .
( dasp ) , phys.lett . * b76 * ( 1978 ) 246 .
s. behrends et al .
( cleo ) , phys.rev.lett .
* 50 * ( 1983 ) 881 .
w. behrends et al .
( jade ) , phys.lett . *
b146 * ( 1984 ) 437 . h. albrecht et al .
( argus ) , phys.lett .
* b192 * ( 1987 ) 245 .
particle data group , phys.rev .
* d54 * ( 1996 ) 1 .
barber et al .
( mark - j ) , phys.lett .
* b85 * ( 1979 ) 463 , + phys.rev.lett .
* 44 * ( 1980 ) 1722 ; + ch .
berger et al .
( pluto ) , phys.lett .
* b86 * ( 1979 ) 413 , + phys.rev.lett .
* 45 * ( 1980 ) 1533 ; + w.
bartel et al . (
jade ) , phys.lett . *
b88 * ( 1979 ) 171 , * b89 * ( 1979 ) 136 , + phys.lett .
* b99 * ( 1981 ) 277 ; + r. brandelik et al .
( tasso ) , phys.lett .
* b133 * ( 1983 ) 499 ; + b. adeva et al .
( mark - j ) , phys.rev.lett .
* 50 * ( 1983 ) 799 , * 51 * ( 1983 ) 433 , + phys.lett .
* b152 * ( 1985 ) 439 , phys.rev . *
d34 * ( 1986 ) 681 ; + m. althoff et al .
( tasso ) , phys.lett .
* b138 * ( 1984 ) 441 , + z.phys . *
c22 * ( 1984 ) 307 ; + h.j .
behrend et al .
( cello ) , phys.lett . *
b144 * ( 1984 ) 297 . h. yoshida et al .
( venus ) , phys.lett .
* b198 * ( 1987 ) 570 ; + h. sagawa et al . (
amy ) , phys.rev.lett . * 60 * ( 1988 ) 93 ; + i. adachi et al . ( topaz ) , phys.rev.lett . *
60 * ( 1988 ) 97 ; + s. igarashi et al .
( amy ) , phys.rev.lett . * 60 * ( 1988 ) 2359 ; + k.
abe et al .
( venus ) , phys.lett .
* b234 * ( 1990 ) 382 .
abrams et al .
( markii ) , phys.rev.lett . * 63 * ( 1989 ) 2447 .
akrawy et al .
( opal ) , phys.lett .
* b236 * ( 1990 ) 364 ; + d. decamp et al .
( aleph ) , phys.lett . *
b236 * ( 1990 ) 511 ; + p. abreu et al .
( delphi ) , phys.lett .
* b242 * ( 1990 ) 536 ; + o. adriani et al .
( l3 ) , phys.rep . * 236 * ( 1993 ) 1 .
g. arnison et al .
( ua1 ) , phys.lett . *
b147 * ( 1984 ) 493 .
c. albajar et al .
( ua1 ) , z.phys .
* c37 * ( 1988 ) 505 .
t. akesson et al .
( ua2 ) , z.phys . *
c46 * ( 1990 ) 179 .
f. abe et al .
( cdf ) , phys.rev.lett .
* 64 * ( 1990 ) 174 .
f. abe et al .
( cdf ) , phys.rev.lett .
* 68 * ( 1992 ) 447 , + phys.rev .
* d45 * ( 1992 ) 3921 .
s. abachi et al .
( d ) , phys.rev.lett .
* 72 * ( 1994 ) 2138 .
f. abe et al .
( cdf ) , phys.rev.lett .
* 73 * ( 1994 ) 225 .
f. abe et al .
( cdf ) , phys.rev .
* d50 * ( 1994 ) 2966 .
f. abe et al .
( cdf ) , phys.rev.lett .
* 74 * ( 1995 ) 2626 .
s. abachi et al .
( d ) , phys.rev.lett .
* 74 * ( 1995 ) 2632 .
f. abe et al .
( cdf ) , nucl.instr.methods * a271 * ( 1988 ) 387 , and references therein .
d. amidei et al .
( cdf ) , nucl.instr.methods * a350 * ( 1994 ) 73 .
s. abachi et al .
( d ) , nucl.instr.methods * a338 * ( 1994 ) 185 , and references therein .
e. laenen , j. smith , and w.l .
van neerven , nucl.phys . *
b369 * ( 1992 ) 543 , + phys.lett .
* b321 * ( 1994 ) 254 .
p. nason , s. dawson , and r.k .
ellis , nucl.phys . *
b303 * ( 1988 ) 607 .
berger and h. contopanagos , phys.rev . *
d54 * ( 1996 ) 3085 .
s. catani et al . , phys.lett . * b378 * ( 1996 ) 329 . n. cabibbo , phys.rev.lett .
* 10 * ( 1963 ) 531 ; + m. kobayashi , t. maskawa , prog.theor.phys .
* 49 * ( 1973 ) 652 . c. albajar et al .
( ua1 ) , phys.lett . *
b186 * ( 1987 ) 247 ; + phys.lett .
* b256 * ( 1991 ) 221 ; erratum : ibid , * b262 * ( 1991 ) 497 .
s. abachi et al .
( d ) , phys.rev.lett . *
74 * ( 1995 ) 3548 ; + phys.lett .
* b370 * ( 1996 ) 239 .
bigi , nuovo cim .
* a109 * ( 1996 ) 713 .
m. neubert , int.j.mod.phys .
* a11 * ( 1996 ) 4173 .
richman , proceedings of the 28th international conference on high energy physics , 25 - 31 july 1996 , warsaw , poland .
m. gronau , a. nippe and j. rosner , phys.rev . *
d47 * ( 1993 ) 1988 . l. gibbons , proceedings of the 28th international conference on high energy physics , 25 - 31 july 1996 , warsaw , poland . the cdf collaboration , fermilab - pub-96/390-e ( 1996 ) .
the d collaboration , pac - report , sept .
report from the tev2000 study group , fermilab - pub-96/082-e ( 1996 ) . | we review the status of heavy flavour physics at the fermilab tevatron collider by summarizing recent top quark and @xmath0 physics results from cdf and d .
in particular we discuss the measurement of the top quark mass and top production cross section as well as @xmath0 meson lifetimes and time dependent @xmath1 mixing results .
an outlook of perspectives for top and @xmath0 physics in run ii starting in 1999 is also given . |
a subject of great current interest is that of a microscopic description of the ferroelectric transition . it is known that at the transition a divergence of the dielectric constant @xmath2 at @xmath3 occurs , which in most cases is understood as corresponding to the fact that the frequency of an infrared peak goes to zero . on the other hand ,
the infrared frequencies are usually studied via a linear analysis of phonon dispersion relations , while the nonlinear contribution to the dynamics should be relevant at the transition to the ferroelectric phase , as should be near any phase transition .
so the problem arises of how the infrared peaks should be described in a fully nonlinear setting .
a description can actually be given through the study of the time autocorrelation of the polarization due to the ions , which can be computed in the classical approximation via molecular dynamics simulations .
computations of this type were indeed performed successfully in the case of @xcite , with results that agree with the experimental data in a surprisingly good way . at the moment , as will be explained below , we are unable to study ferroelectrics through molecular dynamics .
so in this paper we limit ourselves to quartz , which is not a ferroelectric material , but however presents a divergence in the dielectric constant at the temperature of the @xmath0@xmath1 transition .
we compute here its refractive index curves in the infrared region .
our main concerns are the dynamical properties of the system , particularly at the @xmath0@xmath1 transition , and the quantitative agreement between calculated and experimental spectra in the infrared for @xmath0 quartz .
as is well known @xcite , linear analysis shows that quartz is doubly refractive and that the peaks in its refractive index curves correspond to the frequencies of the active normal modes .
such qualitative results are confirmed by the present molecular dynamics simulations for @xmath0 quartz at high temperatures , even at the transition to the @xmath1 phase , notwithstanding the high nonlinearity of the system .
this quantitative agreement of the nonlinear results with those of the linear analysis , in particular for the values of the frequencies , is quite surprising , in view of the large nonlinear contributions , and probably is due to some deep reason not yet fully understood . on the other hand , both the linear analysis and our nonlinear study fail in perfectly reproducing the experimental data , as some systematic deviations are observed .
we tried several procedures for choosing the parameters , both of the linear model and of the nonlinear one , in order to find a better agreement with the experimental curves .
but there was no way of reducing the relative error below a threshold of the order of 6% .
this fact , too , requires an explanation .
these are the two main results of the present work .
.[tab:1 ] geometric parameters for @xmath0 and @xmath1 quartz ( see text ) . [ cols="^,^,^,^",options="header " , ] | this paper is concerned with theoretical estimates of the refractive index curves for quartz , obtained by the kubo formul in the classical approximation , through md simulations for the motions of the ions .
two objectives are considered .
the first one is to understand the role of nonlinearities in situations where they are very large , as at the @xmath0@xmath1 structural phase transition .
we show that on the one hand they do nt play an essential role in connection with the form of the spectra in the infrared . on the other hand they play an essential role in introducing a chaoticity which involves a definite normal mode .
this might explain why that mode is raman active in the @xmath0 phase , but not in the @xmath1 phase .
the second objective concerns whether it is possible in a microscopic model to obtain normal mode frequencies , or peak frequencies in the optical spectra , that are in good agreement with the experimental data for quartz . notwithstanding a lot of effort , we were unable to find results agreeing better than about 6% , as apparently also occurs in the whole available literature .
we interpret this fact as indicating that some essential qualitative feature is lacking in all models which consider , as the present one , only short
range repulsive potentials and unretarded long range electric forces . *
pacs * : 78.20.ci , 42.70.ce , 63.20.ry |
the use of radiation - tolerant cmos active pixel sensors ( aps ) @xcite for direct detection in transmission electron microscopy ( tem ) , opens new opportunities for fast imaging with high sensitivity @xcite .
one of the key figures of merit for a tem imaging sensor is the point spread function ( psf ) , which affects both the imaging resolving capabilities of the sensor and the absolute image contrast @xcite .
the psf of a pixellated aps used in direct detection depends on several parameters of which the most important are the pixel size , the electron multiple scattering in the sensor and the charge carrier diffusion in the active layer .
this letter discusses the psf achieved using an aps in two different regimes . in traditional bright field illumination ,
the electron flux is such that each pixel is illuminated by one or more electrons per acquisition frame . in this regime
the point spread function has a contribution from the lateral charge spread due to charge carrier diffusion in the active volume .
since the epitaxial layer of a cmos aps is nearly field - free , charge carriers reach the collection diode through thermal diffusion , with collection times of @xmath3 .
the typical cluster size for a cmos aps with 10 @xmath1 m pixels is 4 - 5 pixels for 300 kev electrons and about 45 % of the charge is collected on the central pixel of the cluster . with bright field illumination , at high rate , the signal recorded on each individual pixel is the superposition of the charge directly deposited by a particle below the pixel area with that collected from nearby pixels through diffusion , multiple scattering and backscattering from the bulk si . if the electron rate is kept low enough so that individual electron clusters can be reconstructed , a new regime of operation becomes available .
the electron impact position is reconstructed by calculating the centre - of - gravity of the observed pulse heights on the pixels in the cluster . for pixel detectors with @xmath4 pixel
pitch this technique makes it possible to obtain an @xmath5 point resolution in tracking applications for accelerator particle physics .
the same technique can now be adopted for imaging , provided electron fluxes are low enough so that the detector occupancy is @xmath6 0.05 , and individual clusters can easily be resolved . under these conditions
the psf is expected to depend only on the detector pixel size and cluster s / n ( determining the single point resolution ) and on multiple scattering .
the image is reconstructed by combining a large number of frames , each giving 2 - 5 @xmath1 m accuracy for 0.01 - 0.05 % of the field of view , provided the frame rate is much faster than the dynamics being observed on the sample . in this paper
we refer to this technique as `` cluster imaging '' .
the principle of cluster imaging for tem has been tested using the team1k cmos monolithic pixel sensor , developed as part of the the multi - institutional team ( transmission electron aberration - corrected microscope ) project @xcite .
the team1k sensor and its performance will be discussed in details in a forthcoming paper on this journal @xcite .
the chip is produced in the ams 0.35 @xmath1 m cmos - opto process and has a 1,024@xmath01,024 pixel imaging area with pixels arrayed on a 9.5 @xmath1 m pitch .
the detector employed in this test has been back - thinned to 50 @xmath1 m and mounted on a carrier circuit board which is cut out below its active area to minimise back - scattering effects .
a 75 @xmath1m - diameter au wire is mounted along the pixel rows on top of the sensor at a distance of @xmath72 mm above its surface .
tests have been conducted with the sensor installed at the bottom of a fei titan microscope column at the national center for electron microscopy ( ncem ) .
the energy deposited by electrons in the sensor active layer and the lateral charge spread are simulated with the geant-4 program @xcite using the low - energy electromagnetic physics models @xcite .
the cmos pixel sensor is modelled according to the detailed geometric structure of oxide , metal interconnect and silicon layers .
electrons are incident perpendicular to the detector plane and tracked through it .
charge collection in the sensor is simulated with pixelsim , a dedicated digitisation module @xcite .
the simulation has the diffusion parameter @xmath8 , used to determine the width of the charge carrier cloud , free .
its value is extracted from data by a @xmath9 fit to the pixel multiplicity in the clusters of 300 kev electrons , where multiple scattering is lower .
we find @xmath8 = ( 14.5@xmath10 ) @xmath1 m , which is compatible with the value obtained for 1.5 gev @xmath11s in a prototype sensor with 20@xmath020 @xmath1m@xmath2 pixels produced in the same cmos process @xcite .
the response to single electrons is characterised in terms of the cluster size .
we operate with a flux of @xmath12 50 @xmath11 mm@xmath13 frame@xmath14 which allows us to resolve individual electrons .
electron hits are reconstructed as clusters of pixels . a clustering algorithm with two thresholds
is used @xcite .
first the detector is scanned for `` seed '' pixels with pulse height values over s / n threshold set to 3.5 .
seeds are sorted according to their pulse heights and the surrounding neighbouring pixels are added to the cluster , if their s / n exceeds 2.5 . [
cols="^,^ " , ] for cluster imaging , the electron flux is reduced to @xmath12 50 e@xmath15 mm@xmath13 frame@xmath14 and 20000 subsequent frames are acquired at each energy .
figure [ fig : plotwire ] shows the image of the wire obtained in the two regimes . with cluster imaging , due to the continuous spatial sampling , the information is no longer discretised by the pixel pitch as it is with bright field illumination .
charge diffusion , which limits the psf in bright field illumination , now rather improves the cluster resolution by spreading the charge on more pixels and the relevant parameter for psf becomes the ratio of pixel pitch to signal - to - background , which determines the cluster spatial resolution .
we also observe that the image contrast improves by a factor of three for cluster imaging compared to bright field illumination , as it is apparent in figure [ fig : plotwire ] .
the lsf is extracted by a @xmath9 fit , as discussed above .
we measure ( [email protected] ) @xmath1 m at 300 kev , ( [email protected] ) @xmath1 m at 200 kev and ( [email protected] ) @xmath1 m at 80 kev .
these results show that cluster imaging improves the point spread function by a factor of two compared to traditional bright field illumination , for the team1k detector parameters .
we perform the study on simulation describing the detector and wire geometry , for both operation regimes and compare the results to the measured lsf values .
results are summarised in figure [ fig : plotpsf ] .
simulation reproduces well the data both in absolute terms and in the observed scaling with the electron energy .
we also note that the cluster imaging lsf closely follows the simulation predictions for single point resolution .
the team cmos aps has demonstrated a psf for direct detection under bright field illumination comparable to its pixel size .
image reconstruction by clustering at the impact point of individual electrons at low flux allows to reduce the psf down to ( [email protected] ) @xmath1 m and improves the image contrast . in order to practically exploit this method in tem
it is necessary to ensure that the sensor can be read out at a rate of several hundred frames s@xmath14 and cluster reconstruction can be performed in real time .
this work was supported by the director , office of science , of the u.s .
department of energy under contract no.de-ac02-05ch11231 .
the team project is supported by the department of energy , office of science , basic energy sciences . | a cluster imaging technique for transmission electron microscopy with a direct detection cmos pixel sensor is presented .
charge centre - of - gravity reconstruction for individual electron clusters improves the spatial resolution and thus the point spread function .
data collected with a cmos sensor with [email protected] @xmath1m@xmath2 pixels show an improvement of a factor of two in point spread function to 2.7 @xmath1 m at 300 kev and of a factor of three in the image contrast , compared to traditional bright field illumination . ,
, , monolithic active pixel sensor , transmission electron microscopy ; |
liquid bulks exhibit short range order , extending to a few molecular diameters only . in certain liquids , however , a free surface induces ordering of the near - surface molecules into well - defined layers that typically extend into the bulk to depths of the order of the bulk correlation length . for liquid crystals this can be several tens of molecular lengths
; however , for more typical atomic liquids layering extends only 3 or 4 molecular diameters.@xcite for all pure metals and non - dilute alloys , atoms within each surface - induced layer were found to exhibit only liquid - like short range order in the surface - parallel directions .
this general observation differs strikingly from our recent discovery of a two - dimensional crystalline surface phase in the au@xmath0si@xmath1 liquid eutectic alloy.@xcite prior to this discovery observation of laterally ordered phases in metallic liquids has been limited to binary dilute systems at concentrations close to the phase boundary for coexistence with the 3d solid crystal . in those cases , the dilute , higher melting point component segregated to the surface to form a complete surface monolayer.@xcite
the observation of a surface - crystalline monolayer in a non - dilute system , which is far from the phase boundaries of the bulk solid phase , is , therefore , quite unusual and very unexpected .
the formation of a crystalline surface phase in au@xmath0si@xmath1 alloy is also accompanied by a surface layering enhancement of 6 - 7 well - defined atomic layers , manifested by an @xmath830-fold enhancement of the intensity of the bragg - like layering peak in the x - ray specular reflectivity , as compared to that observed for all other metallic liquids measured to date . at t=@xmath3c
the x - ray reflectivity curve was found to undergo an abrupt change to a different profile , in which the intensity of the reflectivity peak was reduced by a factor of about 5 , indicating a first - order surface phase transition .
here we present evidence revealing that at @xmath3c the quasi-2d crystalline layer undergoes a phase transition into a different crystalline surface phase .
interest in the au@xmath0si@xmath1 eutectic alloy is due to a number of factors .
ausi has an unusually deep eutectic point : typically the melting point for the eutectic composition is lower by 100 - 200@xmath9c than the melting points of its individual pure components . for ausi the eutectic point ( 359@xmath9c )
is many hundreds of degrees below the melting points of au ( 1063@xmath9c ) or si ( 1412@xmath9c).@xcite the reasons for such an unusually deep eutectic point are still not fully understood and attract , therefore , considerable attention . remarkably , ausi also does not form thermodynamically stable crystalline intermetallic compounds for any temperature or composition and this has been speculated to be related to the deep eutectic point . upon cooling below the eutectic point , au@xmath0si@xmath1 either phase separates into pure au and pure si regions , or ,
if quenched rapidly enough it forms an amorphous solid .
ausi was the first discovered metallic glass@xcite and amorphous ausi can be produced with a variety of techniques , including splat - quenching and evaporation.@xcite in addition to the fundamentally important questions of its unusual bonding properties , ausi eutectic has found a variety of practical applications .
both au and si are crucial to the electronics industry , and thin gold wires are commonly used to interconnect silicon - based devices .
the low melting point of the eutectic alloy , the eutectic s deep and narrow shape in the phase diagram , and the lack of intermetallic crystalline compounds have resulted in ausi being used as a low - melting solder for micro- and nano - electro - mechanical systems ( mems and nems).@xcite the same properties are employed to manufacture high - purity si nanowires and other nanostructures from ausi alloys through a vapor - solid - liquid growth mechanism .
@xcite finally , extensive work has been done over the last 20 years on thin solid films of ausi , among other metal - semiconductor compounds , in connection with schottky barrier diode devices . despite of the large amount of research performed on solid bulk and thin solid films of amorphous ausi phases ,
remarkably little is known about the properties of the liquid ausi eutectic alloy , in particular about atomic structure and the compositional variations in the near - surface region .
the gibbs segregation rule predicts a significant enhancement ( 65 - 85 at .
% ) of si in the surface monolayer,@xcite which would result in a local composition that is far from the eutectic composition of 18 at .
the high brightness of the x - ray beams generated by modern synchrotron sources , in combination with the high - precision liquid surface reflectometers available at some at these sources,@xcite make possible detailed , atomic resolution , investigations of the near - surface structure of liquids .
these methods are employed here to determine the surface structure of the ausi eutectic over an extended temperature range from its melting point up .
x - ray measurements were carried out at chemmatcars , aps , argonne national laboratory , and at beamline x22b , nsls , brookhaven national laboratory , using x - rays of a wavelengths @xmath10 and @xmath11 , respectively .
a number of measurements were carried out on each of the two different samples used , yielding highly reproducible data .
each ausi sample was prepared from 100 g au@xmath0si@xmath1 ingots of 99.99@xmath12 purity ( goodfellow ) melted under uhv conditions ( p@xmath13 torr ) in a mo sample pan .
the resultant liquid sample had a diameter of 20 mm , a thickness of @xmath145 mm and a surface radius of curvature of @xmath1420 m. macroscopic surface oxides were scraped off by an in - vacuum mo scraper .
several hours of ar@xmath15 ion sputtering removed the remaining microscopic oxide traces . the sample temperature was electronically controlled to @xmath140.01 @xmath16c .
pickup of acoustic noise and mechanical vibrations was eliminated by placing the uhv sample chamber on an active vibration isolation unit mounted on the liquid surface reflectometer .
it is important to describe the experimental steps that were undertaken to guarantee that the signatures of the crystalline surface phase observed in the x - ray diffraction and reflectivity measurements are indeed due to the intrinsic surface structure of the ausi alloy , rather than being an experimental artifact , such as the formation of an oxide or chemical impurities segregated at the surface .
the presence of even minute concentrations of surface impurities would have been detected by surface - sensitive probes , such as x - ray photoemission spectroscopy ( xps ) and x - ray fluorescence ( xf ) measurements performed in grazing incidence geometry .
however , no signature of any impurity was , in fact , detected by these techniques .
since these surface probes are capable of detecting even a small fraction of a monolayer of another species , the crystalline monolayer reported here is not likely to be due to impurities .
reproducibility of the data for two ausi samples make an artifact due to contamination even less likely .
additional evidence for lack of contamination or oxidation at the surface comes from ar@xmath17 ion sputtering that was applied to the surface during the experiments .
measurements performed while sputtering , immediately after the end of the sputtering session ( typically lasting several hours ) , and a long time ( 24 hrs ) after the sputtering has ended , have produced identical and highly reproducible results .
re - formation of a sputtered - off surface layer by surface segregation of impurities from the bulk or oxidation of the surface under uhv conditions ( partial pressure of oxygen and water was less than 10@xmath18 torr according to in - situ residual gas analysis of the chamber s contents ) should occur within something of the order of an hour ( or more ) , which would have stood out clearly during the series of measurements that were carried out .
details of the x - ray scattering geometry of the experiments have been previously described in detail.@xcite the x - ray reflectivity , @xmath19 , is the fraction of the intensity of the incident x - rays reflected by a surface , for a grazing angle of incidence @xmath20 .
@xmath21 is the surface - normal wavevector transfer , and @xmath22 @xmath23 is the critical @xmath24 for total external reflection in our case.@xcite for an ideal case of sharply terminated , static , structureless interface , where the electron density profile normal to the surface is a step function , reflectivity is described by the fresnel law : @xmath25 where @xmath26 is the critical wavevector , @xmath27 is the classical electron radius and @xmath28 is the electron density at the surface .
reflectivity from a _ real _ liquid surface deviates from @xmath29 in two significant ways .
first , the free surfaces are never ideal , and exhibit a monotonic or a non - monotonic variation of the electron density across the interface between the ( constant ) value of the vapour and the ( constant ) value of the bulk .
second , at any finite temperature the surface of every liquid is roughened by thermally excited capillary waves .
these waves partially destroy the constructive interference between x - rays reflected from different points on the surface , leading to a debye - waller - like reduction of the specularly reflected intensities , the magnitude of which is determined by the liquid s surface tension , temperature , and the portion of the capillary waves spectrum sampled by the reflectometer.@xcite for @xmath30 , @xmath19 is well approximated by : @xmath31 where @xmath32 is a surface roughness term due to the capillary waves @xcite , and @xmath33 is the surface structure factor , given by:@xcite @xmath34 since @xmath29 is a universal function , depending only on the measured @xmath35 , and @xmath32 is obtained from measurements of the off - specular diffuse scattering,@xcite @xmath36 can be determined , within a phase factor , @xcite from the measured @xmath19 through eq .
( [ eq : fresnel2 ] ) .
this , in turn , allows a determination of the laterally - averaged surface - normal electron density profile @xmath37 from eq .
( [ eq : structure ] ) by a fit to a physically - motivated model for @xmath38 . for liquid metals
a surface - induced layering is observed .
similar to bragg diffraction from a crystal , constructive interference between x - rays reflected from these well - defined surface - induced atomic layers leads to a significant enhancement of the reflected signal at @xmath39 where @xmath40 is the spacing of the atomic layers . in the case of
the standard layering observed for most liquid metals , the electron density is typically described by a semi - infinite sum of equidistantly spaced gaussians , normalized to the bulk density @xmath41 : @xmath42 \label{eq : rho}\ ] ] the widths of the gaussians are defined by @xmath43 , where @xmath44 and @xmath45 are adjustable fit parameters .
this form for @xmath46 produces a gradual increase in the gaussian width with distance @xmath47 below the surface , and a corresponding decrease in their height .
since they are equally spaced , this leads to their gradual merging , yielding eventually a constant density for large @xmath47 .
thus , @xmath45 is inversely related to the decay length for the surface layering . in all liquid metals studied to date the typical layering extends to 3 - 4 well - defined layers before density oscillations decay to @xmath48 .
the density profile that had to be used to fit the measured ausi reflectivity curves included n additional layer - representing gaussians ( where usually n=4 or 5 ) , with adjustable densities ( @xmath49 ) , positions ( @xmath50 ) and widths ( @xmath51 ) : @xmath52 + \nonumber\\ + \frac{d/\sigma _ n}{\sqrt{2\pi}}\sum _ { n=0 } ^\infty \exp \left [ -(z - z_n - nd)^{2}/2\sigma _
n ^2 \right ] \label{eq : rho_ausi}\end{aligned}\ ] ] finally , the properties of the thermally excited capillary waves are governed by the competition between the thermal energy , described by the temperature @xmath53 and the restoring force , provided by the surface tension @xmath54 .
capillary waves induce surface roughness that gives rise to diffuse scattering away from the specular condition , thus reducing @xmath19 .
detailed analysis of the lineshape of the off - specular diffuse scattering intensity allows one to determine the term @xmath55 in eq .
[ eq : fresnel2].@xcite in practice any specular reflectivity measurement involves integration of the off - specular diffuse scattering contributions over the finite resolution function of the detector .
details of the capillary wave theory and the resolution effects on capillary contribution are described elsewhere.@xcite 371 @xmath5c ( red symbols ) and t@xmath4371 @xmath5c ( black symbols ) along with their corresponding model fits ( lines ) discussed in the text .
the reflectivity expected for a standard 3 - 4 surface layers , observed in all other metallic liquids , is shown in a dashed blue line .
the solid blue line shows the fresnel reflectivity curve , expected for an ideally flat and smooth liquid ausi alloy s surface . ]
the as - measured and the fresnel - nomalized x - ray reflectivity curves are shown at five different temperatures in fig . [ fig : ausi_fig1 ] and fig .
[ fig : ausi_fig2 ] , respectively .
two features stand out immediately .
the first feature is the bimodal temperature dependence , with a sharp division between the shapes of the low - temperature ( t@xmath56371 @xmath5c , red symbols ) and the high - temperature ( t@xmath4371 @xmath5c , black symbols ) @xmath57 , but with practically no shape variation within each temperature range .
the second feature is that @xmath57 in _ both _ regimes are much higher than that expected on the basis of the conventional layering that was observed in almost all liquid metals and alloys studied to date .
the low-@xmath53 @xmath57 , in particular , exhibits an @xmath830-fold increase over the standard layering peak .
the transition temperature between the two regimes , @xmath58 @xmath5c , was determined by measuring @xmath57 at a fixed @xmath24 while varying @xmath53 , and found to be hysteresis - free to within our @xmath59 @xmath5c @xmath53-resolution @xcite . from the liquid ausi surface for different temperatures .
notation as in fig .
[ fig : ausi_fig2 ] .
note the logarithmic @xmath57 scale . ] and [ fig : ausi_fig2 ] . ]
the model fits shown in lines in figs .
[ fig : ausi_fig1 ] and [ fig : ausi_fig2 ] , yield the electron density profiles plotted in fig . [ fig : ausi_fig3 ] , along with that expected from conventional layering in ausi .
it should be noted that some specific details of the density profile can not be determined uniquely from such fits , due primarily to the relatively large number of the adjustable parameters in using density model in eq .
[ eq : rho_ausi ] .
this model includes n=4 - 5 adjustable layers in addition to 3 - 4 layers employed in fitting conventional surface layering .
specifically , the precise positions , relative intensities and widths of the layers are correlated , albeit weakly , and fits of similar quality could be obtained by correlated variations in the values of several of these parameters within restricted ranges of values .
some of these ambiguities can be lifted by taking into account the physical details of the density model - for example , a layer can not be narrower than the diameter of the atom , and the density can not rise above that of a close - packed layer of atoms . despite this uncertainty , the main physical features of the fits , such as the existence of at least 6 - 7 well - defined , atomically sharp , almost periodically ordered layers are essential for reproducing the measured @xmath57 shown in figs . [
fig : ausi_fig1 ] and [ fig : ausi_fig2 ] .
these features are , therefore , strongly supported by the measured curves . as fig .
[ fig : ausi_fig3 ] demonstrates , both low- and high-@xmath53 density profiles show a similarly well - ordered surface - normal structure .
however , the high-@xmath53 profile exhibits a higher degree of disorder , reflected in fewer and lower layering peaks in fig .
[ fig : ausi_fig3 ] . the increased disorder in the high-@xmath53 phase results in a lower intensity of the corresponding quasi - bragg layering peak of the high-@xmath53 phase @xmath57 curves in fig .
[ fig : ausi_fig2 ] as compared to that of the low-@xmath53 phase @xmath57 .
both low- and high-@xmath53 layering peaks in fig .
[ fig : ausi_fig2 ] are significantly broader than the conventional layering peak , shown in a dash - dot - dot line .
this effect is likely due to a slight aperiodicity of the layering , that is also apparent in the density profiles in fig .
[ fig : ausi_fig3 ] . as the electron densities of au and of si are , respectively , much higher and much lower than the average density of the alloy , @xmath48 , the @xmath60 relative density of the top surface layer in fig .
[ fig : ausi_fig3 ] implies a si enrichment relative to the bulk s composition . using the fitted density profile and assuming closed - packed homogeneous layers of au and si atoms of known diameters ,
a simple calculation yields a @xmath61 atomic % of si atoms in the topmost surface layer , in good agreement with the 65 - 85 atomic % of si predicted by the gibbs adsorption rule for an ideal binary solution of ausi.@xcite the surface - parallel ordering was explored by grazing - incidence x - ray diffraction ( gixd).@xcite x - rays incident on the surface below the critical angle @xmath62 generate an evanescent wave which travels along the surface , and has an amplitude which decays exponentially with depth below the surface , with a 1/e decay length of @xmath63 .
thus , the x - rays sample only the top few atomic layers . scattering by surface - parallel order present in these layers gives rise to gixd patterns .
such patterns measured for the low-@xmath53 and the high-@xmath53 phases are shown in fig .
[ fig : ausi_fig4 ] .
in addition to the broad peak arising from the short - range order of the underlying bulk liquid , one can clearly see a number of sharp diffraction peaks indicating existence of long - range in - plane order .
this is the first , and so far only , liquid metal or non - dilute alloy to show surface - parallel order . all previous gixd measurements on liquid metals and non - dilute alloys revealed only a broad liquid - like peak , indicating that despite the increased order in the surface - normal direction , i.e. the surface - induced layering , the order within each layer in the surface - parallel directions remained liquid - like . as observed in fig .
[ fig : ausi_fig4 ] , the gixd patterns found above and below @xmath64 @xmath5c are not identical , implying a change in the lateral order of the surface .
this confirms that the sharp structural transition detected in the reflectivity measurements , discussed above , is indeed a first - order solid - solid surface phase transition .
we now discus the order in the two phases , as revealed by the gixd measurements .
indexing of the gixd pattern for the low-@xmath53 surface phase has been previously reported@xcite within a 2d rectangular @xmath65 @xmath66 unit cell , with a au@xmath67si@xmath68 stoichiometry .
this composition is in quantitative agreement with the electron density derived from the @xmath57 fit .
the unit cell dimensions , which determine the positions of the peaks observed in the gixd pattern , were determined using the crysfire @xcite powder indexing software .
the automatic exhaustive search for the crystallographic structure allowed indexing all observed peaks in a single two - dimensional unit cell .
the simplest bravais lattice symmetry was rectangular , and attempts to fit the data by an oblique bravais lattice invariably resulted in a primitive unit cell with @xmath69 .
a nearly perfect match was obtained for 22 bragg peaks that were within our measurable range of gixd measurements , even though some peaks ( for example ( 23 ) and ( 31 ) , or ( 22 ) and ( 13 ) pairs ) could not be resolved with resolution provided by the soller slit setup .
positions of atoms within the unit cell were refined using the gsas , @xcite powdercell @xcite and carine @xcite software packages .
the initial composition was varied from pure au to pure si , including au@xmath67si , au@xmath70si , ausi , ausi@xmath70 stoichiometries . while during the diffraction pattern refinement the number of atoms within the unit cell was defined roughly by the requirement of close packing , in principle it is also conceivable that various atoms of quasi-2d crystalline structure are displaced up or down , and are not strictly in the same horizontal ( liquid surface - parallel ) plane .
in fact , such `` puckering '' may also be one of the reasons why formation of crystalline 2d layer does not result in the nucleation of a 3d bulk phase .
the distribution of the atoms along the surface - normal direction can be obtained , in principle , from a detailed analysis of the intensity distribution along the bragg rods of the different gixd peaks@xcite , a complete measurement of which has not been done in the present study .
extensive attempts to index the high-@xmath53 gixd pattern failed to produce a single unit cell which reproduces all observed peaks without producing additional strong peaks that were not observed experimentally .
however , in fig .
[ fig : ausi_fig5](b ) we show that the pattern is reproduced well by two coexisting oblique two - dimensional unit cells , shown in the inset .
phase a produces 10 gixd peaks shown in fig .
[ fig : ausi_fig5](b ) in red .
it has a unit cell of @xmath71 @xmath66 and @xmath72 , containing 4 au atoms and 2 si atoms .
phase b ( 4 gixd peaks shown in blue ) has a smaller unit cell of @xmath73 @xmath66 and @xmath74 , containing 1 au atom and 2 si atoms .
@xmath23 . peaks ( 23 ) and ( 31 ) appear as a single peak with the low - resolution soller slit setup , but are well resolved in the pattern measured with the crystal analyzer . ]
the gixd data discussed so far was collected with a soller slit analyzer in front of the detector , providing an in - plane resolution of @xmath75 @xmath23.with this resolution , the widths of the diffraction peaks were resolution - limited .
the gixd peaks were remeasured using a double - bounce bonse - hart si(111 ) analyzer , @xcite having a resolution of @xmath76 @xmath23 .
high - resolution scans of selected peaks are shown in fig .
[ fig : ausi_fig6 ] .
application of the debye - scherrer formula to the linewidths of low- and high - resolution gixd lines , as well as stability of gixd peak intensity during sample rotation imply micron - size crystallites .
the intensities of the gixd peaks in the high-@xmath53 phase were found to fluctuate with time .
this could be assigned to fluctuations in the orientation of the smaller number of larger ( l@xmath77 m ) crystallites in this phase .
these fluctuations would then bring a randomly varying number of crystallites into and out of the bragg condition , resulting in peak intensity fluctuations .
the thickness of the quasi-2d crystalline phase can be estimated from the ratio of the integrated intensities of the sharp diffraction peaks to that of the broad liquid bulk peak .
this ratio , 0.18 , should be approximately equal to the ratio of the thickness of the crystalline phase to the decay length @xmath78 of the evanescent wave . for the @xmath79
obtained above , this yields a @xmath80 thick crystalline phase , or about a single atomic layer .
@xmath23 , collected with soller slits .
bragg rod is oriented along surface - normal direction @xmath24 , indicative of a two - dimensional powder of crystallites aligned flat against the liquid surface .
a dashed line shows theoretically expected trajectory for a three - dimensional powder cone ( @xmath81=const ) . ]
a more precise estimate of the thickness of the crystalline phase comes from a measurement of the bragg rods , @xcite which are the intensity distributions along @xmath24 at the gixd peak positions .
the bragg rod of the combined ( 23 ) and ( 31 ) gixd peaks , observed with the low - resolution setup at @xmath82 @xmath23 , was measured by employing a fixed incidence angles @xmath83 , and performing a series of @xmath84 scans at several ( fixed ) exit angles @xmath85 relative to the surface .
as observed in fig .
[ fig : ausi_fig7 ] , the bragg rod is oriented along surface - normal ( @xmath24 ) direction , indicating that it originates in a 2d crystal and not in a 3d powder .
had the gixd peak originated in a 3d powder , a circular `` powder ring '' , shown in a dashed line in fig .
[ fig : ausi_fig7 ] , would have been observed in the @xmath86 plane .
this powder ring is the locus of the ewald sphere for which the modulus of the scattering vector @xmath87 is a constant value @xmath88 .
clearly , the measured rod does not follow this trajectory , and thus does not originate in a 3d powder .
the intensity distribution along the bragg rod ( @xmath24 direction ) is shown in fig .
[ fig : ausi_fig8 ] for the highest - intensity gixd peak of the low-@xmath53 phase observed at low resolution at @xmath82 @xmath23 .
had this rod originated in a single , ideally flat planar monolayer the intensity distribution along @xmath24 would have followed the form of the square of the atomic scattering factor of au .
in fact , the atomic scattering factor of au does not vary by more than a few percent over the 1.5 @xmath23 range shown in fig .
[ fig : ausi_fig8 ] .
however , even for an intrinsically sharp density profile , the surface roughening due to capillary waves would have induced an exponential fall - off in the bragg rod intensity around @xmath89 @xmath23 , consistent with the fall - off observed in fig .
[ fig : ausi_fig8 ] .
it is also possible that the fall - off near 1.5 @xmath23 arises from a combination of the capillary roughening and a destructive interference between waves diffracted by the 2d order in two atomic layers .
the condition for a destructive interference of waves diffracted from two layers separated by a distance @xmath90 , @xmath91 , yields , in our case , a distance of @xmath92 @xmath23 .
this is about 2/3 of the mean layering distance ( i.e. the distance between atomic layers measured along surface normal ) observed in the reflectivity measurements discussed above .
moreover , this @xmath90 is also consistent with the @xmath93 thickness of the crystalline layer derived at the end of the previous section from the intensity ratio of the sharp gixd peaks and the broad liquid peak of the bulk .
a similar estimation of the layer thickness from the intensity ratios for the high-@xmath53 phase is complicated by the intensity fluctuations of the gixd peaks , mentioned above .
nevertheless , the crystalline layer thickness of that phase seems to be comparable to that of the low-@xmath53 phase . clearly , bragg rod measurements to larger @xmath24 values , and for additional gixd peaks , are required to elucidate the details of the 2d surface - parallel order in both low- and high-@xmath53 phases , and , in particular , their surface - normal variation .
, for the combined ( 23 ) and ( 31 ) gixd peaks centered at @xmath94=2.653 @xmath23 , measured with the low - resolution soller slits setup ]
detailed reasons for the formation of quasi-2d crystalline surface layers in the eutectic ausi alloy have not yet been established .
surface melting ( also known as pre - melting ) , i.e. a formation of a quasi - liquid layer at the surface of a crystal below its bulk melting point is a common and well - documented behavior for a wide range of materials from metals and semiconductors to dielectrics.@xcite the existence of such a liquid - like layer is the main reason for ice being slippery @xcite as has been first proposed in 1859 by michael faraday.@xcite upon approaching the bulk melting point from below the crystalline structure at the surface starts disordering because of the increased entropy due to the lower number of near - neighbor atoms at the surface .
this coexistence of a disordered surface with an ordered bulk is , thus , quite common .
by contrast , the reverse effect of the coexistence of a long - range ordered surface with a disordered ( liquid ) bulk at a high temperature , as observed here , is quite unusual .
surface freezing , i.e. long - range crystalline in - plane ordering of a surface layer in coexistence with an underlying liquid bulk , has been previously reported in systems composed of long chain molecules.@xcite several theoretical explanations have been suggested for this effect , all of which ( but one ) are based on the highly anisotropic shape of the molecules and their large , @xmath95 , lengths.@xcite a different type of surface ordering effect involves crystallization of the higher melting point dilute chemical component in a binary metallic alloy , when it is segregated to the surface , for example the crystallization of a surface - adsorbed monolayer of pb and of tl in ga@xmath96pb@xmath97 @xcite and ga@xmath98tl@xmath99 @xcite alloys , respectively .
this phenomenon differs substantially in several respects from the crystalline phase reported here for au@xmath0si@xmath1 in several respects .
as our ( unpublished ) studies of the freezing of a surface - segregated phase in a non - dilute alloy ( gabi ) show , the frozen layer ( bi in this case ) is a macroscopically thick crystalline surface film .
this is not surprising , since once the surface monolayer of pure bi crystallizes , it serves as a nucleating center for further crystal growth .
second , the crystalline surface phase in eutectic ausi is not the pure chemical species , as is the case in gapb and gatl , and in fact is not a phase that is stable in the bulk at any composition or temperature . finally , the solid - solid surface phase transition observed in ausi 12 @xmath16c above the eutectic point , combined with the stability of the high - temperature crystalline surface phase up to at least 430 @xmath16c , 70 @xmath16c above the melting point , are difficult to reconcile within the basic thermodynamics of the ausi phase diagram .
an intriguing question posed by this study is the underlying reason for the formation of the crystalline monolayer phase at the surface when such crystalline phase are not stable in the bulk .
the unusually deep eutectic observed in ausi , combined with the relative ease of formation of a supercooled liquid and of glassy phases @xcite , implies an increase resistance to , and a substantial degree of frustration against , crystal formation in this alloy .
the existence of supercooling in liquid metals dates to observations by turnbull.@xcite based on these results , frank proposed @xcite the existence of a nucleation barrier due to the icosahedral short - range order of the liquid .
this prediction was recently confirmed by neutron @xcite and x - ray @xcite scattering studies on electrostatically levitated supercooled metals and alloys . even though it is impossible to construct a long - range periodically ordered 3d bulk phase based on close packing of polytetrahedral structures due to the associated five - fold symmetry
, this topological frustration may be lifted at quasi-2d surfaces and interfaces .
for example , x - ray studies of liquid - solid interfaces have demonstrated that liquid five - fold symmetry can be observed in monatomic liquid pb when it is aligned by the crystalline si substrate.@xcite packing configuration consisting of two different kinds of atoms with substantial size mismatch , as is the case of ausi eutectic , results in a far more complex topological problem compared to monatomic liquids .
however , one can draw an interesting parallel between the low - temperature crystalline phase unit cell structure shown in the inset of fig.[fig : ausi_fig5](a ) and the proposed local structure in binary metallic glasses that is based on the idea of overlapping , or interpenetrating clusters@xcite ( compare the inset of fig.[fig : ausi_fig5 ] ( a ) with fig . 3 in ref .
@xcite ) . because atoms within the surface layer have the additional freedom of moving slightly up or down ( puckering , buckling or relaxation type effects ) , it is plausible that an otherwise metastable 2d crystalline structure based on the interpenetrating cluster model can be stabilized at the free surface , even if long - range order can not be extended into the 3d bulk . aside from these purely topological considerations ,
surface and bulk electronic properties of the ausi eutectic alloy are likely to play a crucial role in the formation of the crystalline structure as well .
changes in local chemical composition due to gibbs adsorption effects are expected to influence the short - range order as well as the electronic behavior in the near surface region .
amorphous ausi alloys vary from semiconductors for si - rich compositions , characterized by continuous random network of covalently bonded si atoms , low packing density and a low atomic coordination number ( 4 to 5 ) , to metallic - like au - rich ausi glasses of a high density due to random hard - sphere packing ( high coordination number of 8 - 9).@xcite since si is expected to have a lower surface tension than that of au,@xcite the chemical composition of liquid ausi alloys evolves from si - rich at the surface to au - rich in the bulk , resulting in a non - trivial evolution of the electronic and the structural properties.@xcite thin _ solid _ films composed of metal - semiconductor alloys have been an active area of research because of their interest for schottky diode devices.@xcite the crystalline surface phase reported here does not resemble those of pure au , pure si or any of the reported metastable bulk intermetalics.@xcite because of the metastable nature of solid gold silicides , details of reported intermetallic structures strongly depend on the methods of the film deposition method , substrate surface type and preparation , annealing temperature and duration , as well as other variables .
an additional difficulty is the accurate characterization of sub - nanometer thick layers sandwiched between a solid substrate and a thick film of pure au or si . however , crystalline phases with unit cell dimensions @xmath100 @xmath66 , similar to our low - t phase , were reported in thin au films deposited on a si(111 ) surface.@xcite while no reliable thicknesses could be deduced for these phases from the low - energy electron diffraction ( leed ) and auger electron spectroscopy techniques that were used , sputtering studies indicate that the thickness of crystalline gold silicide phases is in the range of 2 - 9 .@xcite our investigations of other eutectic binary alloys involving au and elements of the iv semiconductor group , namely ausn @xcite and auge ( unpublished data ) do not exhibit any evidence for surface freezing effects akin to those reported here for ausi . for these alloys
the layering peak in the x - ray reflectivity signal is considerably weaker than that for ausi , and is consistent with a simple layering model comprising 3 - 4 atomic layers .
their gixd shows no sharp diffraction peaks that would indicate the existence of a long - range crystalline in - plane order in the near - surface region of the liquid alloy .
it is notable that the au@xmath101ge@xmath102 eutectic shows nearly the same melting point ( @xmath103c ) as the au@xmath0si@xmath1 eutectic .
melting point of pure ge , 938 @xmath16c , is however much lower than that of si ( 1410 @xmath16c ) .
the size mismatch between au and si is much greater than that between au and ge . while both auge and ausi exhibit no stable intermetalic compounds , au and si are not mutually soluble in solid form , but solid ge is weakly soluble in solid au.@xcite these properties imply that the mismatch between atomic size is greater and the bonding of unlike species weaker in ausi ( evidenced by strong covalent si - si bond @xcite ) than in auge .
further insight into the detailed reasons for the formation of these unusual quasi-2d crystalline surface phases in coexistence with the underlying liquid bulk might be ascertained from theoretical and computer simulations of the evolution of the electronic , chemical and structural properties of the near - surface region of the ausi eutectic , as well as future experimental studies of the atomic structure of related liquid metal alloys . because of the close proximity of the eutectic temperatures and au compositions of ausi and auge alloys , one of the very promising approaches is the study of the evolution of surface order in liquid ternary eutectic ausige alloys upon variation of ge composition from zero up .
finally , we note that the recent electron microscopy studies of the atomic structure of liquid metals at their interface with a solid@xcite , detailed investigation of the chemistry of the vapor - liquid - solid@xcite and the vapor - solid - solid @xcite si and ge nanowire growth mechanisms , as well as direct visualization of alternative crystallization pathways in nanoscopic metallic droplets of auge alloys @xcite are related to the present study and our results might provide a better basis for understanding this rich phenomena .
this work was supported by the u.s .
department of energy ( doe ) grant de - fg02 - 88-er45379 and the u.s .- israel binational science foundation , jerusalem .
work at brookhaven national laboratory is supported by u.s .
doe contract de - ac02 - 98ch10886 .
chemmatcars sector 15 is principally supported by nsf / doe grant che0087817 .
the advanced photon source is supported by the u.s .
doe contract w-31 - 109-eng-38 .
authors would like to acknowledge jeff gebhardt and frank westferro for their contribution in making and testing of the crystal analyzer .
discussions with dr stefan mechler ( harvard)are gratefully acknowledged .
surface tension values of au and si at 370 @xmath5c were extrapolated to 1270 - 1300 mn / m for au and 860 - 1000 mn / m for si .
such large uncertainty is due to a large temperature range over which extrapolation from melting points of pure metals was carried out ( 700 @xmath5c and 1030 @xmath5c for au and si , respectively ) .
application of guggenheim equation , assuming interaction parameter set to zero results in 65 - 85 at .
@xmath12 of si in the near - surface monolayer , in contrast with 19 at . @xmath12 of si in the bulk .
see , e.g. , a. guinier , x - ray diffraction in crystals , imperfect crystals , and amorphous bodies ( dover publication , new york , 1994 ) , pp .
121 - 125 ; b. e. warren , x - ray diffraction ( dover , new york , 1990 ) , pp .
251 - 253 .
x. z. wu , b. m. ocko , e. b. sirota , s. k. sinha , and m. deutsch , science * 261 * , 1018 ( 1993 ) ; x. z. wu , e. b. sirota , s. k. sinha , b. m. ocko , and m. deutsch , phys .
* 70 * , 958 ( 1993 ) ; e. sloutskin , x.z .
wu , t. b. peterson , o. gang , b. m. ocko , e. b. sirota , and m. deutsch , phys .
e * 68 * , 031605 ( 2003 ) .
f. a. m. leermakers and m. a. cohen - stuart , phys .
lett . * 76*,82 ( 1996 ) ; e. b. sirota , x. z. wu , b. m. ocko , and m. deutsch , phys .
lett . * 79 * , 531 ( 1997 ) ; a. v. tkachenko and y. rabin , phys .
lett . * 76 * , 2527 ( 1996 ) ; a. j. colussi , m. r. hoffmann , and y. tang , langmuir * 16 * , 5213 ( 2000 ) ; m. kawamata and t. yamamoto , j. phys .
. jpn . * 66 * , 2350 ( 1997 ) ; p. smith , r. m. lynden - bell , j. c. earnshaw , and w. smith , mol .
phys * 96 * , 249 ( 1999 ) .
s. e. donnelly , r. c. birtcher , c. w. allen , i. morrison , k. furuya , m. song , k. mitsuishi , and u. dahmen , science * 296 * , 507(2002 ) .
s. h. oh , y. kauffmann , c. scheu , w. d. kaplan , and m. rhle , science * 310 * , 661 ( 2005 ) .
w. sutter and e. a. sutter , nat . mater .
* 6 * , 363 ( 2007 ) . | a two dimensional crystalline layer is found at the surface of the liquid eutectic au@xmath0si@xmath1 alloy above its melting point @xmath2c . underlying this crystalline layer we find a layered structure , 6 - 7 atomic layers thick .
this surface layer undergoes a first - order solid - solid phase transition occurring at @xmath3c .
the crystalline phase observed for t@xmath4371 @xmath5c is stable up to at least 430 @xmath5c .
grazing incidence x - ray diffraction data at t@xmath4371 @xmath5c imply lateral order comprising two coexisting phases of different oblique unit cells , in stark contrast with the single phase with a rectangular unit cell found for low - temperature crystalline phase @xmath6c@xmath7c . |
egb 6 ( pn g221.5 + 46.3 ) is a low - surface - brightness , high - galactic - latitude , angularly large ( @xmath3 ) planetary nebula ( pn ) , discovered by one of us in 1978purely by chance while examining palomar observatory sky survey ( poss ) prints .
the pn appears close to spherical , but its southwestern edge has a bright rim , possibly indicating an interaction with the interstellar medium .
the object was included in a list of faint nebulae found on poss prints by ellis et al .
( 1984 , hereafter egb ) .
egb noted that the poss photographs show a blue 16th - mag star near the center of the nebula .
this star had also been detected independently by green et al.(1986 ) during the palomar - green survey for high - latitude blue objects , and was designated pg 0950 + 139
. follow - up spectroscopy by egb revealed that this candidate planetary - nebula nucleus ( pnn ) has strong [ ] emission lines , making it elementary that it is the central star .
subsequently , fleming et al.(1986 ) classified the absorption - line spectrum of the pnn as that of a hot da white dwarf ( wd )
. however , liebert et al .
( 1989 , hereafter l89 ) showed that is seen in absorption in the optical spectrum , and later gianninas et al .
( 2010 , hereafter g10 ) found absorption lines of heavy elements in the far ultraviolet .
thus the spectral type of the central star is daoz .
based on a non - lte model - atmosphere analysis , g10 derived stellar parameters of @xmath4 k and @xmath5 .
the [ ] emission from the pnn is too strong to come from the surrounding large , faint , ancient pn 5007 emission from the large pn has been detected serendipitously in the sdss spectra of two faint galaxies that happen to lie behind egb 6 : yuan & liu ( 2013 ) .
acker et al .
( 1992 ) list the relative intensities of h@xmath6 , h@xmath7 , and 5007 . ] .
observations presented by l89 showed that the nebular lines arise from a compact emission knot ( cek ) , which is unresolved and appears to coincide with the pnn in ground - based images .
moreover , the electron density of the cek is remarkably high , about @xmath8 , according to an emission - line analysis by dopita & liebert ( 1989 , hereafter dl89 ) .
in addition to the hot wd and associated cek , there is a compact near - infrared ( nir ) source located near the egb 6 nucleus .
this was first revealed through nir photometry of the star by zuckerman et al .
( 1991 ) , and confirmed by fulbright & liebert ( 1993 , hereafter fl93 ) .
these studies showed that the @xmath9 @xmath10 , and @xmath11 fluxes exceed those expected from the hot pnn .
both sets of authors concluded that the nir excess is consistent with the presence of an m dwarf companion star .
subsequent photometry presented by de marco et al .
( 2013 ) and douchin et al .
( 2015 , hereafter d15 ) found no significant excess at the @xmath2 band , but based on the absolute magnitude of the excess at sdss @xmath12 and at @xmath13 , and an adopted distance , they inferred that the nir source is a star of spectral type m3 - 5 v. however , there had been no direct spectroscopic confirmation that an m dwarf is present , and it should be remembered `` how dangerous it always is to reason from insufficient data '' ( conan doyle 1892 ) .
indeed , de marco et al .
( 2013 ) noted the curious incident that the @xmath14 color is redder than expected for an m3 - 5 spectral type , and miszalski et al .
( 2011 ) pointed out that the nir source was too luminous for an unreddened star of the observed color .
yet another component of the system was discovered in _
spitzer space telescope _
observations by chu et al .
( 2011 , hereafter c11 ) , who showed that the nucleus of egb 6 has a strong mid - infrared ( mir ) excess .
c11 surveyed 71 hot wds ( including 35 that are pnni ) with the _ spitzer _ multi - band imaging photometer ( mips ) at @xmath15 m , detecting nine of them at fluxes two or more orders of magnitude above those expected from the hot wds .
follow - up observations were obtained by c11 with the _ spitzer _ infrared array camera ( irac ) in its four bands from 3.6 to @xmath16 m .
of the nine @xmath15 m sources , egb 6 is the apparently brightest of all in the four irac bands , and second brightest at @xmath15 m .
c11 modeled its mir excess as arising from two cool dust shells , with blackbody temperatures of about 500 k and 150 k. an extensive discussion of the astrophysical mysteries of the egb 6 central system , with additional literature references , was published recently by liebert et al .
( 2013 , hereafter l13 ) .
l13 included results from broad - band and emission - line imaging and grism spectroscopy with the _ hubble space telescope _
( _ hst _ ) obtained between 1991 and 1995 , but which had not been presented in detail previously .
hst _ images and spectra revealed the startling result that the cek is not centered on the pnn , but is instead a separate object with a stellar profile , lying @xmath17 away from the hot nucleus .
this corresponds to a projected linear separation of @xmath1118 au , for a nominal distance of about 725 pc . the longest - wavelength broad - band _ hst _ images available to l13 were in the f785lp filter of the original wide field / planetary camera ( wf / pc1 ) and the f814w filter of the wide field planetary camera 2 ( wfpc2)that is , filters covering roughly the ground - based @xmath2 band .
these images revealed a faint , resolved point - source companion of the central wd , which was argued to be the dm star responsible for the nir excess .
remarkably , the location of the @xmath2-band stellar source coincides with that of the cek .
thus egb 6 raises several astrophysical puzzles , including how to explain the existence and survival of a compact dense [ ] -emitting nebula apparently associated with a cool m dwarf , located at least 118 au from the source of ionizing radiation .
egb 6 may not be entirely unique , however : frew & parker ( 2010 ) and miszalski et al .
( 2011 ) discuss several additional examples of otherwise normal pne with compact , unresolved , high - density nebulae at their centers , comprising a class of `` egb 6-like '' pnni .
the peculiar emission - line object tol 26 ( ctio 1230@xmath18275 ) , having a high - density , compact nebula ( hawley 1981 ) , may be related to this class . in this paper , we discuss new _ hst _ images of egb 6 , and present previously unpublished ground - based photometric and spectroscopic monitoring of the object , obtained in a search for time variability in the stellar and nebular fluxes .
we also present egb 6 s nir spectrum , and discuss the spectral - energy distribution ( sed ) of the nucleus .
we end with some scenarios to explain the origin of this puzzling object .
as noted in 1 , the _ hst _ data discussed by l13 were taken between 1991 and 1995 .
two decades later , in late 2013 , we obtained new _
_ images , for the purposes of confirming the conclusions of l13 , verifying that the cek is physically associated with the pnn by showing that they have the same proper motions , and obtaining nir images of the source .
table 1 presents details of our new _
hst _ observations .
all data were obtained with the wide field planetary camera 3 ( wfc3 ) , using its uvis channel in the near - ultraviolet ( nuv ) and optical , and its ir channel in the nir . for the uvis imaging
, we used a three - point dither pattern in each of seven different filters : five broad - band filters , extending from the nuv to the _ hst _ equivalent of the @xmath2 band , and two narrow - band filters covering the emission lines of [ ] 5007 and h@xmath6 .
the uvis frames were taken with a @xmath19-pixel subarray , yielding a field of view of @xmath20 .
for the ir imaging , we used a four - point dither pattern and the f160w filter , with a bandpass similar to the ground - based @xmath10 band .
the ir subarray was also @xmath19 pixels , giving an angular size of @xmath21 . for our analysis , we downloaded the default pipeline drizzle - combined images in each filter from the mikulski archive for space telescopes ( mast ) .
these frames have cosmic rays removed , are corrected for geometric distortions , and have pixel values corresponding to counts ( electrons ) per second .
figure 1 shows pictorial representations of the wfc3 images . in the nuv ( f225w and f275w ) and optical @xmath22 ( f336w ) bands ,
only the hot daoz central star is seen . the @xmath23-band image ( f555w )
shows the companion cek faintly , but only because several strong emission lines ( principally h@xmath7 and [ ] 49595007 ) lie within the broad filter bandpass .
the narrow - band images in [ ] 5007 ( f502n ) and h@xmath6 ( f656n ) clearly resolve the stellar - appearing cek from the central star .
the widths of the filter bandpasses for f502n and f656n , according to the wfc3 _ instrument handbook _ ( dressel 2015 ) , are 65 and 18 , respectively .
the much wider bandpass of the [ ] filter compared to that of h@xmath6 explains why the cek appears fainter than the central wd in f502n , but brighter in f656n ( in addition to the daoz wd having h@xmath6 in absorption ) . at the @xmath2 band ( f814w ) ,
a faint source is seen at the position of the cek .
we measure this source to be about 3.6 mag fainter than the wd .
unfortunately , we have no spectroscopy of egb 6 that covers the entire @xmath2 band . however , the detected source is almost certainly too bright to be due to emission lines from the cek .
there are very few nebular emission lines in the f814w bandpass ; see , for example , the @xmath2-band spectrum of the pn ic 2165 in fig . 4 of dufour et al .
ic 2165 has a central star with parameters similar to that of the hot component of egb 6 , according to henry et al.(2015 ) , so a similar nebular emission spectrum is expected .
moreover , dl89 calculated a photoionization model for the egb 6 cek , leading to predictions of emission - line strengths , including lines in the f814w bandpass ( their table 1 ) .
using either the line fluxes measured in ic 2165 ( scaled to the h@xmath6 flux in egb 6 ) , or the model predictions from dl89 , as input to the wfc3 exposure time calculator , we find that the emission lines can account for no more than @xmath110% of the signal seen in f814w .
thus we confirm the argument in l13 that a cool source of faint @xmath2-band light lies at the location of the cek . at the @xmath10 band ( ir - channel f160w )
we see a single source with a stellar profile .
although the wd is comparable in flux to the cool source in the @xmath10 band ( see 4 below ) , the ir channel has considerably larger pixels ( @xmath24 ) compared to those of the uvis channel ( @xmath25 ) , so we expect at most a small elongation of the image .
however , the two components are well resolved in ground - based nir active - optics ( ao ) data , as discussed below in 3.4 .
we used routines from the iraf software package to carry out aperture photometry on the wfc3 images .
the count rates within apertures with a radius of @xmath26 ( thus including the central star , cool companion , and cek ) were first determined with the phot task .
these were then converted to absolute fluxes , using the photometric zero - points given by the photflam keywords in the image headers , and scaling these values ( which are for apertures of infinite radius ) to @xmath26 apertures , using ratios derived from the information given at the wfc3 website at space telescope science institute ( stsci ) .
the resulting @xmath27 values are given in the right - hand columns in table 1 for the broad - band filters .
for the narrow - band filters , we calculated the emission - line fluxes from the cek for [ ] 5007 and h@xmath6 by multiplying @xmath27 by the filter bandwidths given above , and then scaling by the magnitude differences between the central star and the cek measured in the two images .
the resulting monochromatic line fluxes are given in the final column of table 1 , for the [ ] 5007 and h@xmath6 lines .
the line fluxes are consistent with all of the emission being from the knot .
the separation and position angle ( pa ) of the cek with respect to the pnn were measured by l13 in the _ hst _ frames taken between 1991 and 1995 . for convenience
, these measures are reproduced in the first three lines of table 2 . in order to determine whether our new observations reveal any appreciable relative motion , we measured the positions of the cek and pnn , using the centroiding task in the imexamine package in iraf .
we then employed the stsdas xy2rd routine to convert the @xmath28 positions to j2000 right ascension and declination , from which separation and pa can be determined .
the average values from the 2013 frames in [ ] and h@xmath6 are given in the last line of table 2 .
these results show that there has been no significant change in the separation and pa over the approximately two - decade interval covered by our observations .
independent measurements of the absolute proper motion of the central star are available from the ppmxl ( roeser et al .
2010 ) and ucac4 ( zacharias et al.2013 ) catalogs , and from data release 7 of the sloan digital sky survey ( sdss ) as quoted by girven et al .
these sources give proper motions , @xmath29 in @xmath30 , of @xmath31 , @xmath32 , and @xmath33 , respectively .
taking the mean of these values , we find that the total absolute motion of the central star from the epoch of the first _ hst _ observation to that of the most recent one is @xmath34 . since the measured separation of the pnn and cek has changed by no more than a few mas , this is a strong indication that they are physically associated .
there has , however , been no detectable orbital motion .
this is slightly surprising , since the nominal period , for a true separation of @xmath1118 au , would be of order 1450 years , and thus about 1/60th of the period would have elapsed between the two sets of _ hst _ images .
this would give a change of @xmath35 in pa for a face - on circular orbit , which we do not see .
however , there is a wide range of orbital parameters for a gravitationally bound system that are consistent with our observations .
we turn now to ground - based photometry and spectroscopy of the egb 6 central object .
we discuss absolute photometry , monitoring programs aimed at searching for photometric and/or spectroscopic variations that might shed light on the nature of the central source , and an archival but previously unpublished nir spectroscopic observation .
we obtained absolute photometry of the central source in egb 6 on three photometric nights between 1991 and 1998 , using 0.9-m telescopes at kitt peak national observatory ( kpno ) and cerro tololo inter - american observatory ( ctio ) , and calibrated to the johnson - kron - cousins _ bvri _ system via observations of standard stars from landolt ( 1992 , 2009 ) .
these previously unpublished data are presented in table 3 , along with results quoted from the literature sources indicated in the last column .
in addition , the last two data entries from 2016 are from observations with the asiago 0.67/0.92-m schmidt telescope , and likewise calibrated to landolt standards .
the first two lines in table 3 are from photoelectric observations in 1978 and 1982 , with errors of approximately @xmath360.05 mag .
the remaining data were obtained with ccds , were reduced with standard aperture - photometry routines , and have typical errors usually of about @xmath360.0050.010 mag .
averages of the ccd results , and the errors of the means , are given at the bottom of table 3 .
the @xmath37 and @xmath23 magnitudes of the central object have been essentially constant within the errors from 1991 to 2016 , and , with slightly larger uncertainty , since the photoelectric observations in 19781982 .
the @xmath38 magnitude is also nearly constant , but with perhaps a slightly larger scatter .
there does appear to be variability at the @xmath2 band , at the level of several hundredths of a magnitude , even during a single observing run with identical equipment .
based primarily on the two observations in 1991 and 1994 , compared with subsequent measures , there is a slight suggestion of a slow secular fading of the @xmath2 magnitude , but this could simply be due to the randomness of the short - timescale variability .
our measured magnitude difference of 3.6 between the companion and the wd in the _ hst _ f814w image ( discussed in 2.1 ) , along with an @xmath2 magnitude of the total flux of 16.31 , imply that the apparent magnitude of the companion is @xmath39 .
we obtained differential ccd photometry during six different observing runs on 0.9-m telescopes at kpno ( 1990 december , 1991 november , 1998 march ) and ctio ( 1991 january , 1992 may , 1994 march ) .
these results have likewise not been published previously . in nearly all cases
only a single set of _ bvri _ observations was made per night during these runs , as they were part of a larger monitoring survey of many pnni in search of variable central stars ( e.g. , bond et al .
1992 ; ciardullo & bond 1996 ) .
the total number of observations of egb 6 was 57 in @xmath37 and @xmath23 , 50 in @xmath38 , and 53 in @xmath2 .
the seeing in these frames generally ranged between @xmath40 and @xmath41 .
the frames were bias - subtracted and flat - fielded using standard iraf tasks .
differential photometry was carried out using the daophot point - spread - function ( psf)-fitting routines within iraf .
each frame s psf was determined from three bright , isolated field stars located near the egb 6 central star .
we then calculated the magnitude difference between egb 6 and the sum of intensities of these three comparison stars .
the formal precisions for differential magnitudes determined in this way for a source as faint as egb 6 are about @xmath360.01 mag in @xmath37 , @xmath23 , and @xmath38 , and about @xmath360.02 mag in @xmath2 . unfortunately , for the purpose of long - term monitoring
, this is a cumbersome batch of data .
not only were the six runs made with five different combinations of ccd and filter sets , but the pnn ( @xmath42 ) is considerably bluer than the comparison stars ( whose combined light has @xmath43 ) .
thus our differential magnitudes are sensitive to changes in the bandpass effective wavelengths among the different instrumental setups . during three of the six observing runs , observations of landolt standard stars allowed us to determine the color terms , and thus correct our data to the standard system . for the other three runs ( kpno 1990 december , kpno 1991 november , and ctio 1992 may ) , we simply forced the mean differential magnitudes to equal the means from the three calibrated runs .
( this would mean that we might miss any systematic long - term changes in brightness , but the absolute photometry discussed above indicates this is not the case , except possibly in @xmath2 . )
figure 2 plots the resulting magnitudes against date of observation .
the zero - points in each filter have been set to reproduce the mean calibrated magnitudes given in table 3 , but with @xmath38 and @xmath2 offset for clarity by the amounts indicated in the figure .
the photometric scatter in @xmath37 , @xmath23 , and @xmath38 is roughly twice that suggested by the photon - statistical errors quoted above , but this is not surprising because some of the data were taken in non - photometric conditions , and because of small systematic errors due to flat - field illumination and other instrumental effects .
the scatter in @xmath2 does , however , appear to exceed the expected amount ; since the absolute photometry discussed in 3.1 also suggested variability , we believe that the source is indeed a short - term variable at the @xmath2 bandpass . we searched for a periodic signal using a periodogram analysis , but did not find any evidence for one . however , the observing cadence of one observation per night was not particularly suitable for this purpose .
an intensive campaign of monitoring would be useful for further investigation .
photometry of egb 6 in the @xmath23 band is also available from late 2005 to late 2013 from the catalina real - time transient survey ( crts ) ( drake et al .
no significant variability at @xmath23 , apart from a few outliers , is seen in more than 700 observations .
typical error bars for the crts data points are about @xmath360.07 mag .
there are also about 20 photographic observations from the digital access to a sky century @ harvard ( dasch ) project ( grindlay et al.2012 ) , going back to 1914 , again showing no convincing evidence for long - term variability at blue wavelengths .
the star did nothing in the night - time .
as discussed in 1 , it had been believed that there is an m dwarf companion to the central wd star of egb 6 .
moreover , the system is suspected of variability in the @xmath2 band ( and in the nir
see below ) .
this suggested the possibility that the cool component could be a dm flare star or other type of variable .
if so , the optical spectrum during a flare might exhibit enhanced balmer emission as well as emission at h and k. moreover , the discussion in l13 raised the possibility that the electron density in the cek could be changing , for example if the dense compact nebula is dispersing .
in this case , the nebular forbidden emission - line spectrum might be variable . with these motivations
, we monitored the central source by arranging to obtain optical spectra with the 1.5-m telescope at ctio operated by the smarts consortium .
the observations were conducted by chilean service observers on 47 nights between 2004 january 31 and 2012 january 24 .
we used the rc - focus spectrograph equipped with a ccd camera , and grating 26 in first order , covering either 35325300 or 36605440 , at a spectral resolution of 4.3 .
exposure times each night were @xmath44 s. a short exposure on a hear lamp was taken before each set of stellar observations for wavelength calibration .
the ccd images were bias - subtracted and flat - fielded , combined for cosmic - ray removal , and then the spectrum was extracted and wavelength - calibrated , all using standard iraf routines .
the spectra were normalized to a flat continuum , and smoothed with a three - point boxcar kernel .
the top panel in figure 3 shows two examples of the spectra , obtained near the beginning and end of the eight - year observing interval , in order to illustrate the quality of the individual data .
this panel also shows a mean spectrum made by combining all 47 observations .
the strongest emission lines are marked .
the bottom panel in figure 3 shows the same combined spectrum , with the vertical scale expanded so as to show the photospheric absorption features from the wd more clearly .
we see no strong evidence for spectroscopic variability in these data .
enhanced balmer emissions , and the appearance of h and k in emission , never occurred .
using standard iraf tasks , we combined the spectra into seasonal averages , and determined the equivalent widths ( ews ) of the emission lines during each season .
we also measured the ews in the combined spectrum of all 47 observations .
the results are presented in table 4 .
the uncertainties for the seasonal averages are of order a few percent for the strong lines , and up to about 12% for the weaker ones . in the next - to - last line of table 4 we list the ews published by l89 , based on the averages of seven spectra obtained with various telescopes over the interval 1978 to 1987 .
we see no evidence for large changes in the emission - line spectrum over the entire range from 1978 to 2012 .
( there is a slight suggestion that [ ] 5007 and h@xmath7 have weakened , but at least for the latter this could be a systematic effect due to the blend with the photospheric absorption feature , combined with the relatively low spectral resolution of the smarts spectra . )
as this paper was being completed , we were able to obtain spectrograms of egb 6 on 2016 april 3 , using the asiago astrophysical observatory 1.22-m telescope .
the exposures were @xmath45 s. equivalent widths derived from these spectra are given in the final entry in table 4 .
they are generally consistent with the values in the rest of the table , showing again that any secular changes in the emission - line spectrum since the late 1970s have been small .
the new observation provides marginal evidence for a slow strengthening of [ ] 4363 , which if so would suggest an increase in the electron density .
there is further evidence that h@xmath7 has weakened , and [ ] 49595007 have strengthened .
continued monitoring of the spectrum is desirable . in a search of the gemini observatory archive
, we found that observations of the central source in egb 6 had been obtained with the 8-m gemini north nir spectrograph ( gnirs ; elias et al .
2006 ) on 2012 december 29 ( program gn-2012b - q-60 ; pi : j. bilkov ) .
the observing sequence was a standard abba pattern of @xmath46 s exposures on egb 6 and @xmath47 s exposures on hr 4041 , a neighboring a0 v star to be used for telluric correction .
gnirs was operated using ao in its cross - dispersed mode with the @xmath48 grating and a @xmath49 slit .
wavelength coverage was 0.952.5 @xmath50 m , at resolution @xmath51 .
in the @xmath10-band acquisition images , we noticed that egb 6 was partially resolved into two point sources , at a separation consistent with the _ hst _ observations described above .
the spectrograph slit , perhaps fortuitously , was oriented such that it lay almost exactly along the nearly east - west orientation of the two sources ; thus the spectra of the two sources are spatially separated .
visual examination shows that the hot wd dominates the continuum emission blueward of @xmath1@xmath52 m . at longer wavelengths , the cool source brightens and the wd weakens . in the @xmath10 band ,
the two sources have nearly equal fluxes .
the cool companion dominates redward of there .
figure 4 presents close - ups of four sections of the two - dimensional spectrum .
the first panel on the left shows the region around the 10830 emission line from the cek .
the emission is spatially offset from the continuum of the wd , consistent with our findings from the narrow - band _ hst _ images discussed in 2.1 and in l13 .
the second panel is centered on the region around paschen-@xmath7 at 12818 ; this emission line is likewise offset from the wd continuum , and coincides instead with the weakly visible continuum of the cool source .
the third panel lies near the center of the @xmath10 band ; now the continua of the wd and cool source have similar brightnesses .
in the final panel , near the center of the @xmath11 band , the spectrum of the cool source is brighter than that of the wd . after removing electronic artifacts from the images using the cleanir.py routine provided by the gemini observatory , we processed the data using standard iraf tasks for long - slit spectroscopy .
consecutive images were first flat - fielded , then subtracted to remove sky emission , and finally stacked . due to the semi - resolved nature of the sources , we elected to extract the combined spectrum of both objects using an aperture with a width of 12 pixels .
we attempted to extract separate spectra of the two sources , but were only partially successful , due to the varying flux ratio between the sources and across spectral orders . ] .
the spectrum for hr 4041 was processed in a similar manner .
an argon arc - lamp spectrum taken immediately after the science observations provided wavelength calibration . to remove telluric absorption lines and correct for instrumental response
, we divided the combined spectrum of the two egb 6 sources by the spectrum of hr 4041 , after first interpolating across its intrinsic hydrogen absorption lines .
this ratio spectrum was then multiplied by the @xmath27 vs. wavelength relation for a 9500 k blackbody ( the approximate effective temperature of hr 4041 ) .
the zero - point flux level of the final spectrum was normalized to match the available absolute nir photometry ( tabulated below in table 5 ) .
figure 5 plots the resulting spectrum of the wd plus companion source as a black line .
we saw no absorption features in the spectrum at its modest snr , so we applied an 11-point boxcar smoothing .
several prominent emission lines from the cek are labeledm , which we have been unable to identify .
inspection of the spectrum image indicates that it appears to be real , and is spatially associated with the companion source . ] . to recover the contribution of the cool companion , we subtracted a 93,230 k blackbody , representing the wd component , shown as a blue curve .
the resulting companion spectrum is plotted as a red line .
we stared at it in astonishment : the companion spectrum is _ not _ that of the m dwarf that has been claimed by earlier authors ( 1 ) , including our own previous report in l13 .
instead , longward of about @xmath53 m , there is a smooth continuum , consistent with that of a cool blackbody .
there appears to be a broad bump from approximately 1.0 to @xmath54 m , but this feature may be of doubtful reality since the intrinsic signal in this part of the spectrum ( before conversion to @xmath27 ) is quite weak and noisy .
( if the bump is real , we have no obvious explanation for it ; for example , it is much too narrow to be attributed to a blackbody contribution , and there are no features in late - type stellar spectra with such a structure . )
overall , the companion s energy distribution can be fitted approximately by a blackbody of about 1850 k , plotted as a green line in figure 5 , or somewhat cooler if we were to discount the 1.0@xmath54 m bump .
thus the gemini spectrum has clearly revealed the spatial location of the nir excess that was discovered more than two decades ago .
the nir source is a point - like companion of the central star , which coincides with the unresolved cek seen in the _ hst _ images .
it is not a dm star , but a considerably cooler source .
in table 1 we had presented the continuum fluxes ( @xmath27 ) derived from our 2013 _ hst _ broad - band photometry . in table 5
we collect additional broad - band photometry of the egb 6 central source and accompanying cek from the following archival and literature sources : ( 1 ) fuv and nuv magnitudes from the _ galaxy evolution explorer _ ( _ galex _ ) ; ( 2 ) a @xmath22 magnitude from d15 ; ( 3 ) mean _ bvri _ magnitudes from our table 3 ( which included data from d15 ) ; ( 4 ) _ ugriz _ photometry from the sdss ; ( 5 ) the mean of the _ jhk _ magnitudes listed by fl93 ; ( 6 ) the 2mass @xmath55 magnitudes ; ( 7 ) the _ wise _ @xmath56 through @xmath57 magnitudes ; and ( 8) fluxes from 3.6 to @xmath15 m from the _ spitzer _ irac and mips instruments .
the footnotes to table 5 give literature or archive references for each of these sources of data .
column 4 gives effective wavelengths for each bandpass , with the literature sources given in another footnote .
the magnitudes have then been converted to @xmath27 values in the fifth column , using zero - points referenced in the final footnote .
we also make use of _ herschel space observatory _ imaging of egb 6 , available at the _ herschel _ science archive .
the target was observed ( program ot1_ksu_2 ; pi : k. su ) with the photodetector array camera and spectrometer ( pacs ; poglitsch et al .
2010 ) and the spectral and photometric imaging receiver ( spire ; griffin et al .
2010 ) on 2011 november 27 and 2011 october 23 , respectively .
the pacs observations consist of images at 70 and 160 @xmath50 m , but the central source in egb 6 was only detected at 70 @xmath50 m .
egb 6 was likewise undetected in the spire data ( 250 , 350 , and 500 @xmath50 m ) .
we reduced the pacs data using the _ herschel _ interactive processing environment ( hipe ; ott 2010 ) software version 9.0.0 , following the standard pipeline .
we measured the flux density at 70 @xmath50 m using aperture photometry with a radius of 6@xmath580 , and then applying the aperture corrections suggested by the nasa _
herschel _ science center . since the error maps created by the hipe are unreliable , we estimated the flux error by taking the standard deviation of the background measurements in five arbitrarily distributed 6@xmath580 apertures placed in the high - coverage area of the observations .
we also estimated an upper limit to the non - detection at 160 @xmath50 m in a similar manner and report the measured standard deviation multiplied by three .
the final results are flux densities of @xmath59 mjy at 70 @xmath50 m , and an upper limit of 15 mjy at 160 @xmath50 m .
the 70 @xmath50 m value , converted to @xmath27 , is given in the final line of table 5 .
in addition to the broad - band photometry presented in table 5 , and the gemini nir spectrum discussed in 3.4 , there are flux - calibrated space - based spectroscopic observations available from the following archival sources : ( 1 ) three fuv spectra ( 11501970 ) obtained with the _ international ultraviolet explorer _
( _ iue _ ) in 1987 ( pi : j. holberg ) and 1989 ( pi : j.l . ) ; ( 2 ) two spectra , covering 11402500 and 22203300 , obtained with the faint object spectrograph ( fos ) on _ hst _ in 1992 ( pi : h. shipman ) ; and ( 3 ) a mir spectrum ( 5.2537.3 @xmath50 m ) obtained with the infrared spectrograph ( irs ) on _ spitzer _ in 2009 ( pi : k. su ) .
we downloaded these data from mast ( for _ iue _ and fos ) and from the cornell atlas of _ spitzer _ irs sources ( for irs ; lebouteiller et al .
2011 ) . in figure 6
we plot @xmath60 values based on tables 1 and 5 , and the spectroscopic fluxes , against wavelength .
( this figure is an update of one presented in a conference poster by su et al .
the legend in the figure identifies the sources of the plotted photometry and spectra . from the fuv through the nuv and optical , up to about the @xmath38 band ,
all of the data are in excellent agreement , in spite of having been obtained over a wide interval of dates .
we thus verify again that egb 6 is non - variable over this spectral range .
we had already noted evidence for variability in the ground - based @xmath2 band in 3.2 .
figure 6 shows that even larger discrepancies start to appear as we move to longer wavelengths .
the nir and mir data were obtained over a range of dates : in chronological order ( 1 ) the ground - based @xmath61 observations were made in 1991 ( fl93 ) , with variations on a timescale of one day noted ; ( 2 ) the 2mass data , especially discrepant with fl93 at @xmath11 , are a combination of survey observations obtained from 1997 to 2001 ; ( 3 ) the four _ spitzer _ irac data points which are systematically bright by @xmath10.75 mag relative to the rest of the observations were obtained in 2007 , and the mips datum in 2008 ; ( 4 ) the _ spitzer _ irs spectrum is from 2009 ; ( 5 ) the four _ wise _ points are from survey observations made over seven months in 2010 ; ( 6 ) the _ herschel _ 70 @xmath50 m datum was obtained in 2011 ; and ( 7 ) the gemini nir spectrum is from 2012 . because of the variability of the egb 6 nucleus in the nir and mir , and the fact that the data plotted in figure 6 were obtained at a range of different epochs , it is doubtless premature to attempt a general , static model to explain the sed
however , as a rough guide to the astrophysical parameters , here we will simply fit the sed of the combined light of the nucleus and companion with four components , represented by blackbodies .
we fixed the temperature of the hot wd at 93,230 k from the atmospheric analysis ( 1 ) .
the optical colors as well as the _ iue _ and _ hst_/fos spectra show that the wd is very lightly reddened ; we adopted @xmath62 , giving a good fit to the sed from the fuv to the optical .
since the total reddening in the direction of egb 6 is @xmath63 , according to the reddening maps of schlafly & finkbeiner ( 2011 ) , it appears that the hot central star suffers no significant reddening within the system .
we arbitrarily omitted the four high _ spitzer _ irac points , because the remaining data are reasonably self - consistent , although still showing some discordances due to variability , and then performed a @xmath64 fit of four blackbodies to the remaining data .
the blackbodies were corrected for the adopted reddening , using the formulae of cardelli et al .
( 1989 ) with @xmath65 .
the resulting seds of these blackbodies are shown as dashed black lines in figure 6 , and their sum as a solid red line .
the mir data are fit reasonably well by two cool blackbodies , corresponding to dust with temperatures of 385 and 175 k. this result is in fairly good agreement with c11 , who found 500 and 150 k for these components , but using only the _ spitzer _
irac and mips data .
however , these two blackbodies , combined with the rayleigh - jeans tail of the wd , fail to account for the excess flux in the nir , whose presence as a separate source is clearly revealed by the gemini spectrum .
this nir excess has been attributed to an m3m5 dwarf ( @xmath663000 kemamajek / eem_dwarf_ubvijhk_colors_teff.txt ] ) in previous studies , as summarized in 1 and in l13 and references therein .
however , our @xmath64 fit , as well as the gemini spectrum ( 3.4 and figure 5 ) , require a much lower temperature for this component , around 1850 k.
to summarize the results presented here and in l13 : the hot wd nucleus of the old pn egb 6 is accompanied by a compact source at a projected separation of @xmath1118 au .
this companion object emits both forbidden emission lines from a dense nebula , and nir flux with a blackbody - like spectrum at a temperature of @xmath11850 k. the nir flux varies on a timescale possibly as short as a few days , but the nebular emission spectrum has not shown any large variability .
in addition , there is cooler dust , with temperatures of 175385 k , whose physical location within the system remains unknown from the available observations .
the flux from at least the 385 k component of this dust is also time - variable .
l13 discussed two scenarios that might explain the emission - line knot seen in egb 6 , based on their conclusion that it is physically associated with an m dwarf : ( 1 ) the cek is a stable region of compressed gas where winds from the wd and dm star collide , or ( 2 ) the cek is a photoionized remnant envelope or accretion disk around the dm star , which ( as first suggested by zuckerman et al .
1991 ) was captured during the epoch of rapid mass outflow from the wd progenitor that created the large , faint surrounding pn a few times @xmath67 yr ago .
our new finding that there is _ not _ an exposed m dwarf in the system now makes the colliding - wind scenario unlikely . a combination of four blackbodies wd , nir source , and the two cool mir components does fit the entire sed reasonably well ( again , apart from the four discrepantly high _ spitzer _ irac points ) , as depicted by the red line in figure 6 .
l13 estimated a distance to egb 6 of @xmath68 pc , based on the absolute magnitude of the wd component derived from its atmospheric parameters ; the large uncertainty in @xmath69 is due to relatively large uncertainties in the @xmath70 and @xmath71 of the wd , a consequence of emission - line contamination of the photospheric balmer lines in the spectra analyzed by g10 .
however , frew et al .
( 2016 ) have developed a statistical distance indicator for pne based on an h@xmath6 surface - brightness vs.physical radius relation . applying this calibration to the surrounding large , faint pn of egb 6 , frew et al .
find a similar distance but with a smaller uncertainty : @xmath72 pc .
if the nir source were a star with an effective temperature of @xmath11850 k , it would be an early l - type dwarf , with a @xmath11-band absolute magnitude of roughly + 11.3 . at a distance of about 725 pc ( the unweighted average of the above two estimates ) , this l dwarf would have an apparent magnitude of @xmath73 ; however , the observed @xmath11 magnitudes ( table 5 ) are in the range 15.616.1
. moreover , the observed nir spectrum does not have the strong molecular bands seen in l - type dwarfs ; as described in 3.4 , it has a smooth blackbody - like spectrum longward of @xmath11.2 @xmath50 m .
clearly the nir source is not a stellar photosphere . at a separation of at least @xmath1118 au from the wd ,
the nir source is much too warm to be heated by the wd .
the nir source is more plausibly a dusty envelope or disk , which reradiates the luminosity of a heavily obscured low - mass star . the physical radius of the 1850 k nir source , for a 725 pc distance , is about @xmath74 .
its radiated luminosity is @xmath75 .
if this luminosity is due entirely to reprocessed radiation from an enshrouded star , it corresponds to that of an m3.5 v dwarf , using the mamajek table cited above . in this picture
, the companion source is an m dwarf after all , but one that is hidden from view by surrounding dust yr .
this is based on an assumed mass of @xmath76 and the `` montreal '' wd cooling tracks at http://www.astro.umontreal.ca/@xmath77bergeron/coolingmodels ( e.g. , tremblay et al .
2011 ) . ] . in the remainder of this section we discuss two alternative pictures : one in which the nir and mir sources have separate locations in the system , and another in which they are both emitted by the obscured companion .
we first consider a scenario in which the source of the mir excess in egb 6 does not coincide with the nir and emission - line companion , but is located elsewhere in the binary system .
over the past decade , it has been discovered that a significant fraction of hot wds and pnni are associated with nir and/or mir excesses , indicating the presence of warm and/or cool dust .
su et al .
( 2007 , hereafter s07 ) , using _
observations , detected a mir flux from the central star of the helix nebula ( ngc 7293 ) , a well - known and very nearby pn .
this mir source exhibits a thermal continuum , with a blackbody temperature of @xmath1120 k , attributed to an optically thin debris disk of cool dust surrounding the hot wd at separations of 35150 au .
s07 suggest that this dust arose from collisions of kuiper - belt objects or breakup of oort - cloud comets , due to the dynamical perturbation induced by the sudden mass loss from the central star when the pn was ejected .
_ hst _ imaging of the helix pnn ( ciardullo et al .
1999 ) detected no companion stars with spectral types earlier than m8 ( for projected separations greater than 65 au ) to m5 ( for separations as close as @xmath111 au ) .
s07 noted moreover that the helix pnn shows no nir excess ; thus there is no compelling argument that it has a binary companion .
although the _ hst _ and nir observations of egb 6 clearly associate the nir excess ( and the emission - line source ) with a companion object , there is no direct evidence that the mir excess is located at the position of the companion . thus the helix results may suggest a scenario in which the mir excess in egb 6 is likewise due to a debris disk encircling the hot wd nucleus , and is not directly associated with a companion star searches for ir excesses associated with hot wds and pnni have been made by c11 ( see 1 ) , bilkov et al .
( 2012 ) , and clayton et al .
( 2014 , hereafter c14 ) .
( hoard et al .
2013 carried out a wider - scale search of the all - sky _ wise _ archive for dust excesses around a large sample of wds over a broad range of temperatures , but excluded known pnni and binaries , including egb 6 . )
c11 found nine cases of mir excesses out of a sample of 71 hot wds ; seven of the nine wds are pn nuclei , including egb 6 itself .
dust temperatures are typically 120190 k , but four of the pnni
which c11 call `` egb 6-like'' also exhibit nir excesses implying additional , warmer dust components with temperatures of 5001200 k. bilkov et al.searched the _
spitzer _ archive and found additional examples ; overall , about 18% of pnni have associated dust disks . this is a significantly higher fraction than the @xmath113% incidence of nir excesses due to warm dust around cool ( @xmath7825,000 k ) wds ( farihi et al . 2009 ; c14 and references therein ) .
a further difference is that the warm - dust disks around the cool wds lie very close to the star , typically within the tidal - disruption radius for asteroidal bodies .
as discussed by c14 , this suggests that the dust around pnni is recently formed and relatively transitory , in addition to lying at much greater distances from the central stars as demanded by the low dust temperatures .
these authors note that eight out of 13 pnni with dust disks detected at 8 or @xmath15 m are known or suspected to have binary companions .
this raises the possibility that a binary interaction with the asymptotic - giant - branch ( agb ) wind aids the formation of a dusty disk around the mass - losing star in at least some systems , if not in the helix nebula .
stone et al .
( 2015 , hereafter sml15 ) have discussed these phenomena from a theoretical standpoint .
they propose that the cool dust debris at large separations from pnni arises from the response of a pre - existing oort - cloud analog to the sudden ejection of the pn and the `` natal kick '' received by the wd .
some of the comets would be placed in orbits bringing them close enough to the wd for evaporation or even tidal disruption , leading for formation of a spherical , optically thin , cool dust cloud around the wd .
their model does not directly account for the warmer dust detected in the nir
but , at least in the case of egb 6 , this dust appears to be associated with a companion rather than a cloud around the pn nucleus . for egb 6 itself ,
modeling by sml15 finds that the @xmath15 m flux implies a dust mass of @xmath79 .
now we consider an alternative scenario in which the egb 6 nucleus is the immediate descendant of a symbiotic - like binary similar to the mira ( o ceti ) system . mira itself , a prototypical mass - losing long - period variable star , is accompanied by a companion , mira b , at a separation ( in 1995 ) of @xmath80 ( karovska et al .
1997 ) , corresponding to a projected linear separation of @xmath150 au .
mira b is hidden by an optically thick accretion disk of material captured from the wind of mira a ( e.g. , ireland et al .
2007 , hereafter i07 , and references therein ) .
i07 argue on both theoretical and observational grounds that the underlying mira b star is a late - type dwarf , although many other authors ( e.g. , sokoloski & bildsten 2010 and references therein ) cite evidence for it being a wd .
based on the statistics of wide binaries , i07 predict that binary systems in which an accretion disk is formed around a companion during the agb wind phase are relatively common about one in five among solar - type stars . once the agb star becomes a wd and its slow , dusty stellar wind dies out ,
the companion s accretion disk will also eventually disappear . citing alexander et al .
( 2006 ) for the timescale for viscous evolution of such accretion disks , i07 conclude that `` systems like mira should produce clear observational signatures of an accretion disk around the secondary for at least a few times @xmath81 yr after the primary becomes a wd . ''
wind accretion from agb stars in wide binaries has been studied theoretically by a number of authors , including soker & rappaport ( 2000 ) , perets & kenyon ( 2013 , hereafter pk13 ) , and huarte - espinosa et al .
pk13 state that `` for separations of 3100 au @xmath82 wind - fed disks have surface - density and temperature profiles similar to those observed in low - mass protoplanetary disks . ''
pk13 even raise the possibility of planet formation within such wind - created accretion disks .
if we attribute both the nir and mir flux to the companion object in egb 6 , then in fact it does have several features in common with young stellar objects ( ysos ) , in particular those of class 0/i ( the youngest protostars , with both an obscuring envelope and an accretion disk ) .
these traits include : ( 1 ) significant nir and mir excesses ; ( 2 ) large - amplitude variability in the nir and mir ( up to 2 mag in extreme cases , e.g. , carpenter et al .
2001 ; morales - caldern et al.2011 ) ; and ( 3 ) occasionally h , he , and other species in emission , including [ ] ( e.g. , van loon et al .
these phenomena arise from combinations of accretion , disk morphology , shocks , starspots , and ionizing radiation from nearby bright stars .
these give rise to a wide variety of nir and mir seds and spectra of ysos ( e.g. , furlan et al .
2011 ) . for a direct comparison of the nir / mir sed of the egb 6 companion with those of ysos , in figure 7 we plot its position in color - magnitude and color - color diagrams for combinations of the 2mass @xmath13 and _ wise _ magnitudes .
the contribution of the hot wd to the egb 6 magnitudes has been removed .
we also plot data for samples of ysos in the taurus and upper scorpius star - forming regions , taken from esplin et al .
( 2014 , hereafter e14 ) and luhman & mamajek ( 2012 ) , respectively .
the @xmath56 magnitudes for egb 6 and upper scorpius have been adjusted to the 140 pc distance of the taurus region .
figure 7 illustrates that the infrared luminosity and colors of the egb 6 companion are consistent with those observed in ysos . in this interpretation ,
the nir / mir excess , sed , and spectrum of egb 6 are attributed to a companion object , embedded in an accretion disk and envelope
. the star itself is obscured , and the disk and envelope , which have a wide range of effective temperatures , provide the ir excess .
the emission lines of h and possibly he would be produced by accretion shocks from in - falling material ( e.g. , calvet & hartmann 1992 ) , while the forbidden lines are formed in an outflow or wind .
variations in the accretion rate could be responsible for the observed infrared variability of the system . in figure 7 , we highlight ( filled green circles ) a particular yso , haro 6 - 39 in the taurus region .
we chose this source because , as the figure shows , it has very similar colors to egb 6 ( but is somewhat more luminous ) .
e14 report that nir spectra of haro 6 - 39 show no absorption features , and it has h and he emission ( again , a signature of accretion ) .
thus this yso has several properties remarkably similar to those of the egb 6 companion .
the direction of evolution of ysos in figure 7 is from right to left , i.e. , from young , very red class 0/i protostars on the right , through class ii and then to the least obscured and bluest class iii objects on the left .
the egb 6 companion has the characteristics of a relatively young object
consistent with our interpretation of a recent accretion process but with a very different origin from the true ysos in taurus and scorpius .
egb 6 does have a luminosity , as represented by the @xmath56 magnitude , that is relatively low compared to most ysos at the same @xmath83 color , which may reflect its different formation process .
there is an extensive literature on much closer binary systems among post - agb stars .
there is substantial evidence for interactions with the companion stars during the agb mass - losing phase .
recently , for example , a dusty circumstellar disk has been directly imaged around the post - agb binary iras 08544@xmath184431 by hillen et al .
they have moreover detected a likely compact accretion disk around the binary companion in this system .
however , with an orbital period of only 499 days , the companion is at least two orders of magnitude closer to the mass - losing star than in the case of egb 6 .
to summarize , we have considered two scenarios to explain the astrophysical puzzles presented by egb 6 .
( 1 ) one interpretation is that the mir excess is due to a large - scale , optically thin dust disk surrounding the wd , making it an analog of the cool dust disks seen around a significant fraction of pnni . in this case , the faint companion star is surrounded only by warm dust , located very close to the companion , and having an effective temperature of @xmath11850 k , which accounts for the nir excess flux .
evaporation from the outer surface , and photoionization due to illumination by the nearby hot wd , produce the compact emission - line nebula . to account for the optically thick dust shell around the companion star , we could speculate that it was produced by pulverized rocky bodies that passed very close to the m dwarf .
in fact , the inferred radius of this shell , @xmath84 , is quite close to the tidal - disruption radius of a low - mass star .
( 2 ) alternatively , the central hot wd has a companion m dwarf ( or possibly a second wd ) , which captured an optically thick accretion disk during the mass - ejection episode that produced the surrounding large pn a few times @xmath67 yr ago . in this picture
, the system is descended from a binary similar to the present - day mira that is , a mass - losing agb star accompanied by a distant main - sequence ( or wd ) companion .
both the nir and mir excesses of egb 6 are due to this thick accretion disk .
its properties , including short - timescale variability , are similar to those of ysos in star - forming regions .
a portion of the accretion disk is evaporating rather than falling onto the companion , or there are possibly jets being ejected .
photoionization of the ejecta by the uv flux of the pnn produces the forbidden - emission - line source .
a key uncertainty in distinguishing between these possibilities is the currently unknown physical location of the mir source .
high - resolution imaging , or even just high - precision astrometry , with the james webb space telescope might be able to make this distinction .
another useful program would be simultaneous nir and mir photometric monitoring of the egb 6 nucleus .
if both the nir and mir fluxes vary in tandem , it would argue strongly that they are located close together ; but if the nir variations are not associated with variations in the mir , it could indicate physically separate locations
. also useful would be to apply investigations as comprehensive as the ones described here to a larger sample of pnni with ir excesses . such studies would help us understand empirically whether all of the egb 6-like central stars have binary companions , or whether these phenomena can arise even from single pn nuclei . as the fictional sherlock holmes pointed out , `` it is a capital mistake to theorize before you have all the evidence . ''
we thank the stsci director s discretionary research fund for supporting our participation in the smarts consortium , and fred walter for scheduling the 1.5-m queue observations .
we especially appreciate the excellent work of the ctio / smarts service observers who obtained the direct images and spectra during many long clear tololo nights : claudio aguilera , s. gonzlez , manuel hernndez , rodrigo hernndez , alberto miranda , alberto pasten , and jos velsquez .
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 the nsf .
it also makes use of data products from the wide - field infrared survey explorer , which is a joint project of the university of california , los angeles , and the jet propulsion laboratory / california institute of technology , and neowise , which is a project of the jet propulsion laboratory / california institute of technology .
wise and neowise are funded by nasa . based in part on observations obtained at the gemini observatory , which is operated by the association of universities for research in astronomy , inc .
, under a cooperative agreement with the nsf on behalf of the gemini partnership : the national science foundation ( united states ) , the national research council ( canada ) , conicyt ( chile ) , ministerio de ciencia , tecnologa e innovacin productiva ( argentina ) , and ministrio da cincia , tecnologia e inovao ( brazil ) .
cclcc 2013 dec 6 & wfc3/uvis & f225w & 15 & @xmath85 + @xmath86 & @xmath86 & f275w & 15 & @xmath87 + @xmath86 & @xmath86 & f336w & 15 & @xmath88 + @xmath86 & @xmath86 & f502n & 360 & @xmath89 + @xmath86 & @xmath86 & f555w & 15 & @xmath90 + @xmath86 & @xmath86 & f656n & 720 & @xmath91 + @xmath86 & @xmath86 & f814w & 30 & @xmath92 + @xmath86 & wfc3/ir & f160w & 20.47 & @xmath93 + lccl 1991.9198 & @xmath94 & @xmath95 & liebert et al .
( 2013 ) + 1993.0942 & @xmath96 & @xmath97 & @xmath86 + 1995.7768 & @xmath98 & @xmath99 & @xmath86 + 2013.9314 & @xmath100 & @xmath101 & this paper + lccccl 1978 apr 10 & 16.05 & @xmath102 & @xmath103 & @xmath103 & l89 + 1982 apr 24 & 16.00 & @xmath104 & @xmath103 & @xmath103 & egb ; l89 + 1991 jan 9 & 15.992 & @xmath105 & @xmath106 & @xmath107 & this paper + 1994 may 20 & 15.995 & @xmath108 & @xmath109 & @xmath110 & @xmath86 + 1998 mar 20 & 16.018 & @xmath111 & @xmath103 & @xmath112 & @xmath86 + 2007 nov 2 & 16.001 & @xmath113 & @xmath114 & @xmath115 & de marco et al . 2013 + 2007
nov 5 & 15.997 & @xmath116 & @xmath117 & @xmath118 & @xmath86 + 2011 mar 11 & 15.991 & @xmath116 & @xmath103 & @xmath119 & douchin et al . 2015 + 2011
mar 16 & 16.004 & @xmath120 & @xmath103 & @xmath121 & @xmath86 + 2016 mar 14 & 15.992 & @xmath122 & @xmath123 & @xmath124 & this paper + 2016 mar 17 & 15.989 & @xmath125 & @xmath126 & @xmath127 & @xmath86 + mean & & 15.998 & @xmath113 & @xmath128 & @xmath113 & + error & @xmath360.003 & @xmath360.003 & @xmath360.004 & @xmath360.010 & + lcccccc 2004.12004.3 & 2 & 2.0 & 0.9 & 2.1 & 5.8 & 18.7 + 2004.92005.4 & 6 & 2.3 & 1.0 & 2.3 & 5.5 & 16.8 + 2005.92006.4 & 6 & 2.2 & 1.0 & 2.6 & 5.8 & 17.8 + 2007.02007.3 & 5 & 2.1 & 0.9 & 2.2 & 5.5 & 17.2 + 2007.92008.5 & 6 & 1.7 & 1.2 & 2.2 & 5.5 & 16.6 + 2008.92009.3 & 7 & 2.0 & 1.2 & 2.4 & 5.5 & 17.3 + 2010.02010.4 & 6 & 1.8 & 1.2 & 2.4 & 5.8 & 17.4 + 2011.02011.2 & 5 & 1.8 & 1.2 & 2.6 & 5.9 & 17.9 + 2011.92012.1 & 4 & 1.5 & 1.3 & 2.6 & 6.4 & 17.8 + mean 1.5-m spectrum & 47 & 1.9 & 1.1 & 2.4 & 5.7 & 17.4 + 19781982 & & 1.7 & 1.1 & 2.9 & 5.4 & 17.9 + 2016.25 asiago 1.22-m & & 1.7 & 1.7 & 1.9 & 6.4 & 18.4 + lcccc fuv & 13.75 & galex & 0.1516 & @xmath129 + nuv & 14.58 & galex & 0.2267 & @xmath130 + @xmath22 & 14.465 & d15 & 0.3597 & @xmath131 + @xmath37 & 15.688 & table 3 & 0.4386 & @xmath132 + @xmath23 & 15.998 & table 3 & 0.5491 & @xmath133 + @xmath38 & 16.132 & table 3 & 0.6500 & @xmath134 + @xmath2 & 16.308 & table 3 & 0.7884 & @xmath135 + @xmath136 & 15.199 & d15 & 0.3586 & @xmath137 + @xmath138 & 15.660 & d15 & 0.4716 & @xmath139 + @xmath140 & 16.216 & d15 & 0.6165 & @xmath141 + @xmath142 & 16.563 & d15 & 0.7475 & @xmath143 + @xmath12 & 16.906 & d15 & 0.8922 & @xmath144 + @xmath13 & 16.38 & fl93 & 1.237 & @xmath145 + @xmath10 & 16.08 & fl93 & 1.645 & @xmath146 + @xmath11 & 15.65 & fl93 & 2.212 & @xmath147 + @xmath13 & 16.518 & 2mass & 1.241 & @xmath148 + @xmath10 & 15.945 & 2mass & 1.651 & @xmath149 + @xmath150 & 16.099 & 2mass & 2.165 & @xmath151 + @xmath1521 & 14.651 & allwise & 3.37 & @xmath153 + @xmath1522 & 13.636 & allwise & 4.62 & @xmath154 + @xmath1523 & 9.520 & allwise & 12.08 & @xmath155 + @xmath1524 & 7.170 & allwise & 22.19 & @xmath156 + irac 1 & @xmath82 & c11 & 3.550 & @xmath157 + irac 2 & @xmath82 & c11 & 4.493 & @xmath158 + irac 3 & @xmath82 & c11 & 5.731 & @xmath159 + irac 4 & @xmath82 & c11 & 7.872 & @xmath160 + mips 1 & @xmath82 & c11 & 23.68 & @xmath161 + pacs & @xmath82 & herschel & 70 & @xmath162 + | egb 6 is a faint , large , ancient planetary nebula ( pn ) .
its central star , a hot daoz white dwarf ( wd ) , is a prototype of a rare class of pn nuclei associated with dense , compact emission - line knots .
the central star also shows excess fluxes in both the near- ( nir ) and mid - infrared ( mir ) . in a 2013 paper
, we used _ hubble space telescope _ ( _ hst _ ) images to show that the compact nebula is a point - like source , located @xmath0 ( @xmath1118 au ) from the wd .
we attributed the nir excess to an m dwarf companion star , which appeared to coincide with the dense emission knot .
we now present new ground - based nir spectroscopy , showing that the companion is actually a much cooler source with a continuous spectrum , apparently a dust - enshrouded low - luminosity star .
new _
hst _ images confirm common proper motion of the emission knot and red source with the wd .
the @xmath2-band , nir , and mir fluxes are variable , possibly on timescales as short as days .
we can fit the spectral - energy distribution with four blackbodies ( the wd , a @xmath11850 k nir component , and mir dust at 385 and 175 k ) .
alternatively , we show that the nir / mir sed is very similar to that of class 0/i young stellar objects .
we suggest a scenario in which the egb 6 nucleus is descended from a wide binary similar to the mira system , in which a portion of the wind from an agb star was captured into an accretion disk around a companion star ; a remnant of this disk has survived to the present time , and is surrounded by gas photoionized by uv radiation from the wd . |
the most noted power law in economics is perhaps the pareto law , first observed by vilfredo pareto @xcite more than a century ago .
it was found that in an economy the higher end of the distribution of wealth @xmath1 follows a power - law @xmath2 and @xmath3 is an exponent ( now known as the pareto exponent ) which pareto estimated to be @xmath4 .
for the last hundred years the changes in value @xmath5 3/2 in time and across the various capitalist economies seem to be small .
this implies that there is inequality in the wealth distribution , and only a few persons hold the majority of wealth . in 1931
, gibrat @xcite suggested that while pareto s law is valid only for the high wealth range , the middle wealth range is given by the probability density @xmath6 where @xmath7 is a mean value and @xmath8 is the variance .
the factor @xmath9 is also know an as the gibrat index , and a small gibrat index corresponds to a uneven wealth distribution .
an unequal wealth distribution is associated not only to these functions , but also to any other one which is not of the form a dirac @xmath10-function .
the problem of the appearance of inequalities seems therefore to be rather general and not necessarily related to a particular shape
e.g. the power - law form of the pareto law of the wealth distribution , despite distributions can vary from case to case assuming qualitatively different shapes .
in fact wealth distribution has always been a prime concern of economics .
classical economists such as adam smith , thomas malthus and david ricardo were mainly concerned with factor wealth distribution , that is , the distribution of wealth between the main factors of production , land , labor and capital .
modern economists have also addressed this issue , but have been more concerned with the distribution of wealth across individuals and households .
important theoretical and policy concerns include the relationship between wealth inequality and economic growth .
wealth inequality metrics or wealth distribution metrics are techniques used by economists to measure the distribution of wealth among the participants in a particular economy , such as that of a specific country or of the world in general .
these techniques are typically categorized as either absolute measures or relative measures , and in the literature one of the most important debates is on the issue of measuring inequality .
while one type deals with the objective sense of inequality , usually employing some statistical measure of relative variation of wealth , the other type deals with some indices that try to measure inequality in terms of some normative notion of social welfare for a given total of wealth . _
absolute criteria_. absolute measures define a minimum standard , and then calculate the number ( or percent ) of individuals below this threshold .
these methods are most useful when determining the amount of poverty in a society .
examples include the poverty line , which is a measure of the level of wealth necessary to subsist in a society .
it varies from place to place and from time to time , depending on the cost of living and people s expectations . _
relative criteria_. relative measures compare the wealth of one individual ( or group ) with the wealth of another individual ( or group ) .
these measures are most useful when analyzing the scope and distribution of wealth inequality .
examples include the gini coefficient , which is a summary statistic used to quantify the extent of wealth inequality depicted in a particular lorenz curve .
the gini coefficient is a number between 0 and 1 , where 0 corresponds with perfect equality ( where everyone has the same wealth ) and 1 corresponds with perfect inequality ( where one person has all the wealth , and everyone else has zero wealth ) .
however , it has to be noted that `` wealth '' is here understood differently respect to its common meaning : it represents the total amount of goods and services that a person receives , and thus there is not necessarily money or cash involved .
services like public health and education are also counted in , and often expenditure or consumption ( which is the same in an economic sense ) is used to measure wealth .
thus , it is not clear how wealth should be defined .
there is also the question that `` should the basic unit of measurement be households or individuals ? ''
the gini value for households is always lower than for individuals because of wealth pooling and intra - family transfers .
the metrics will be biased either upward or downward depending on which unit of measurement is used . these and many other criticisms need to be addressed for the proper use of inequality measures in a well - explained and consistent way . in the attempt to answer some such basic questions and provide a foundation to the complex issues related to the appearance of wealth inequalities ,
various authors have independently formulated minimal models of wealth exchange @xcite which , while being general enough to catch some universal features of economic exchanges , are simple enough to be simulated numerically in detail and studied analytically . in these models a set of agents @xmath11 ,
representing individuals or companies whose state is defined by the respective wealth @xmath12 , interact with each other from time to time by exchanging ( a part of ) their wealths .
such exchanges are defined by laws depending on the @xmath13 and also contain some random elements , as detailed below .
a striking analogy was recognized and actually motivated the introduction of some of these models between the statistical mechanics of molecule collisions in a gas and these minimal models of economy , which are therefore referred to sometimes as _
kinetic wealth - exchange models_. such an analogy , rather than for its peculiarity , should be noticed since it signals a possible universal statical mechanism in action in the dynamical evolution of systems composed by single units , from gas composed of molecules colliding with each other exchanging their energy to economic societies in which single units interact by exchanging wealth .
this analogy is being analysed here in more details than done previously @xcite , and represents the goal of the investigations presented here .
we begin by recalling the main features of kinetic wealth - exchange models in sec .
[ features ] , concentrating on an example of model with a fixed _ saving propensity _ @xmath14 . in sec .
[ variational ] it is shown how the fact that for a saving propensity @xmath15 one obtains an equilibrium @xmath0-distribution @xmath16 of order @xmath17 , instead of the boltzmann law @xmath18 obtained for @xmath19 , actually strengthens the kinetic analogy between economy systems and a gas in @xmath20 dimensions , with @xmath20 a known function of @xmath14 @xcite : through a general variational approach based on the boltzmann entropy it is shown that the @xmath0-distribution @xmath16 of order @xmath17 is the equilibrium canonical distribution of a system with a quadratic hamiltonian @xmath21 and @xmath22 degrees of freedom @xmath23 .
the analogy is further discussed in sec .
[ kinetic ] , this time through a complementary microscopic approach based on the analysis of the dynamics of particle collisions in an @xmath24-dimensional space . through mechanical considerations
only based on momentum and energy conservation , it is shown that collision dynamics in @xmath24 dimension can be recast in the form of the evolution laws of kinetic wealth - exchange models , both for @xmath19 and in the case with saving propensity @xmath15 , corresponding to a number of effective dimensions @xmath25 .
finally , in sec . [ conclusion ] , conclusions are drawn .
simple social models of wealth exchange have been shown to well reproduce many features of real wealth distribution .
for instance , the exponential law @xmath26 observed at intermediate values of wealth is reproduced by a many - agent model system composed of @xmath27 agents , who are assumed to exchange wealth in pairs at each iteration , according to the wealth - conserving evolution equations @xcite @xmath28 here @xmath29 is a uniform random number in @xmath30 , @xmath31 and @xmath32 are the labels of two agents chosen randomly at each iteration , and @xmath33 and @xmath34 represent the corresponding wealths before and after a trade , respectively . more general versions of this model assign a ( same ) saving propensity @xmath15 to all agents , which represents the minimum fraction of wealth saved during a trade @xcite .
as an example , in the model of ref .
@xcite the evolution law is @xmath35 it is to be noticed that while the total wealth is still conserved during a trade , @xmath36 , only a fraction @xmath37 of the initial total wealth is reshuffled between the two agents during the trade .
these models relax toward an equilibrium distribution well fitted by a @xmath0-distribution @xmath16 , as also noted by angle @xcite , @xmath38 where the scaling parameter is @xmath39 and the parameter @xmath40 , as shown below , represents an effective dimension of the system and is explicitly given by @xcite @xmath41 as the saving propensity @xmath14 varies in @xmath42 , the effective dimension @xmath20 continuously assumes the values in the interval @xmath43 . a comparison between results of numerical simulation of wealth exchange models and the @xmath0-distribution is shown in fig
. [ realdata3 ] , together with some real data for uk . the the zero - saving propensity model of eqs .
( [ sp1 ] ) is described by the exponential curve for @xmath19 .
the main difference of the curves with @xmath15 respect to the exponential distribution is the appearance of a peak : while the average wealth is unchanged ( the system is closed and one still has @xmath44 , where @xmath27 is the total number of agents ) the number of agents with a a wealth close to the average value increase or , in other words , the wealth distribution becomes more fair for larger @xmath14 s and therefore larger values of the effective dimension @xmath24 .
eventually , as shown in ref .
@xcite , @xmath45 for @xmath46 ( @xmath47 ) thus becoming a perfectly fair distribution . as a consequence , measures of the inequality of the wealth distribution , such as the gini coefficient ,
decrease for increasing @xmath14 and tend to zero for @xmath46 . in the rest of the paper
we discuss the statistical mechanical interpretation of the equilibrium @xmath0-distribution defined by eqs .
( [ f - global ] ) and ( [ n ] ) .
it may seem that in going from the exponential wealth distribution obtained for @xmath19 to the @xmath0-distribution corresponding to a @xmath15 the link between wealth - exchange models and kinetic theory has been lost .
in fact , the @xmath0-distribution @xmath48 represents but the maxwell - boltzmann equilibrium kinetic energy distribution for a gas in @xmath24 dimensions , as shown in the appendix of ref .
@xcite . here
a more general derivation of the @xmath0-distribution @xmath48 is presented , in which it is shown , solely using a variational approach based on the boltzmann entropy , that @xmath48 is the most general canonical equilibrium distribution of an @xmath24-dimensional system with a hamiltonian quadratic in the coordinates .
entropy - based variational approaches have been suggested ( e.g. by mimkes , refs .
@xcite ) to be relevant in the description and understanding of economic processes .
for instance , the exponential wealth distribution observed in real data and obtained theoretically in the framework of the dragulescu - yakovenko model discussed above , can also be derived through entropy considerations . considering a discrete set of @xmath27 economic subsystems which can be in one of @xmath49 possible different states labeled with @xmath50 and characterized by a wealth @xmath51
, one can follow a line similar to the original boltzmann s argumentation for the states of a mechanical system @xcite , by defining the total entropy as @xmath52 where @xmath53 represents the occupation number of the @xmath32-th state ; by variation of @xmath54 respect to the generic @xmath53 , with the constraints of conservation of the total number of systems @xmath55 and wealth @xmath56 , one obtains the canonical equilibrium distribution @xmath57 where @xmath58 defines the temperature of the economic system . here
we repeat the same argumentation for an ensemble of systems described by a continuous distribution @xmath1 , to show that this approach not only can reproduce the exponential distribution , but also the @xmath0-distribution obtained in the framework of wealth - exchange models with a saving parameter @xmath15 , a natural effective dimension @xmath59 being associated to systems with @xmath15 .
the representative system is assumed to have @xmath24 degrees of freedom @xmath23 and a homogeneous quadratic hamiltonian , that for convenience is written in the rescaled form @xmath60 where @xmath61 is the distance from the origin in the @xmath24-dimensional @xmath62-space . as a mechanical example
, one can think of the @xmath24 momentum cartesian components @xmath63 and the corresponding kinetic energy function @xmath64 , where @xmath65 is the particle mass ; or of the cartesian coordinates @xmath23 of an isotropic harmonic oscillator with elastic constant @xmath66 and potential energy @xmath67 . it can be checked e.g. using the stirling approximation for the factorial function that in the limit of large occupation numbers , the discrete version ( [ w ] ) of the boltzmann distribution becomes proportional to @xmath68 , where the continuous variable @xmath69 replaces the discrete label @xmath32 .
for an @xmath24-dimensional system , the entropy will be proportional to @xmath70 $ ] , where @xmath71 is the probability distribution in the @xmath24-dimensional space .
then the constraints on the conservation of the total number of systems reads @xmath72 const and that on total wealth is @xmath73 . in the end , using the lagrange method , the equilibrium distribution density can be derived by functional variation respect to @xmath71 of the effective functional @xmath74 = \int dq_1\dots\int dq_n f_n(q_1,\dots , q_n)\times \nonumber \\
\times\ { \ln[f_n(q_1,\dots , q_n ) ] + \mu + \beta x(q ) \ } \ , , \end{aligned}\ ] ] where @xmath75 and @xmath76 are two lagrange multipliers . before carrying out explicitly the variation it is convenient to simplify the problem exploiting the invariance of the hamiltonian , only depending on the variable @xmath62 , under rotations in the @xmath62-space .
the equilibrium probability density is expected to depend solely on the distance @xmath62 as well , i.e. , @xmath77 .
this allows the reformulation of the variational principle as an effective one - dimensional problem in the variable @xmath62 , by transforming cartesian to polar coordinates , so that @xmath78 where the integration in @xmath79 is over the total solid angle in the @xmath24-dimensional space .
after integration of the @xmath80 angular variables , one is left with @xmath81 = \ !
\int_{0}^{+\infty } \!\!\!\ ! dq \ , f_1(q ) \left [ \ ! \ln\left ( \frac{f_1(q)}{\sigma(n ) q^{n-1 } } \right ) \!+\ ! \mu \!+\ !
\beta x(q ) \ !
\right ] \,\end{aligned}\ ] ] where @xmath82 represents the surface of a hyper - sphere in @xmath24 dimensions @xcite with unit radius and we have expressed the probability density @xmath77 in the @xmath24-dimensional space in terms of the reduced probability density @xmath83 for the variable @xmath62 , @xmath84 finally , moving from @xmath62 to the energy variable @xmath85 , the corresponding probability density is @xmath86 where @xmath87 is obtained by inverting eq .
( [ x ] ) . in terms of the new variable @xmath88 and distribution @xmath1 , the effective functional ( [ func4 ] ) becomes @xmath89 = \ !
\int_{0}^{+\infty } \!\!\!\!\ ! dx \ ,
f(x ) \left[\ ! \ln\left ( \frac{f(x)}{\sigma(n ) x^{n/\,2 - 1 } } \right ) \!+\ ! \mu \!+\ !
\beta x \right ] \!,\end{aligned}\ ] ] variation of this functional , @xmath90/\delta f(x ) = 0 $ ] , leads to the equilibrium @xmath0-distribution ( [ f - global ] ) with dimensionless variable @xmath91 and index @xmath92 . as a simple example of application of this formula , one can obtain the maxwell - boltzmann probability density in three dimensions @xmath93 for the kinetic energy @xmath94 letting @xmath95 . in turn
this suggests the interpretation of the parameter @xmath20 of wealth - exchange models as an effective dimension of the system and of the lagrange multiplier @xmath96 as the effective temperature , as it is in fact consistently recovered according to the equipartition theorem , @xmath97
the deep analogy between kinetic wealth - exchange models of closed economy systems , where agents exchange wealth at each trade , and kinetic gas models , in which energy exchanges take place at each particle collisions , can be further investigated and justified by studying the microscopic dynamics of interacting particles . in this section
we make more rigorous an argumentation only mentioned in ref .
@xcite . in one dimension
, particles undergo head - on collisions , in which they can exchange the total amount of energy they have , i.e. a fraction @xmath98 of it .
alternatively , one can say that the minimum fraction of energy that a particle saves in a collision is in this case @xmath99 . in the framework of wealth - exchange models
, this case corresponds to the model of dragulescu and yakovenko mentioned above @xcite , in which the _ total _ wealth of the two agents is reshuffled during a trade . in an arbitrary ( larger ) number of dimensions ,
however , this does not take place , unless the two particles are travelling exactly along the same line in opposite verses .
on average , only a fraction @xmath100 of the total energy will be lost or gained by a particle during a collision , that is most of the collisions will be practically characterized by an energy saving parameter @xmath15 .
this corresponds to the model of chakraborti and chakrabarti @xcite , in which there is a fixed maximum fraction @xmath101 of wealth which can be reshuffled .
consider a collision between two particles in an @xmath24-dimensional space , with initial velocities represented by the vectors @xmath102 and @xmath103 .
for the sake of simplicity , the masses of the two particles are assumed to be equal to each other and will be set equal to 1 , so that momentum conservation implies that @xmath104 where @xmath105 and @xmath106 are the velocities after the collisions and @xmath107 is the momentum transferred .
conservation of energy implies that @xmath108 which , by using eq .
( [ v1 ] ) , leads to @xmath109 introducing the cosines @xmath110 of the angles @xmath111 between the momentum transferred @xmath107 and the initial velocity @xmath112 of the @xmath31-th particle ( @xmath113 ) , @xmath114 where @xmath115 and @xmath116 , and using eq .
( [ v2 ] ) , one obtains that the modulus of momentum transferred is @xmath117 from this expression one can now compute explicitly the differences in particle energies @xmath12 due to a collision , that are the quantities @xmath118 . with the help of the relation ( [ v2 ] ) one obtains @xmath119 comparison with eqs .
( [ sp1 ] ) for the kinetic model of dragulescu and yakovenko clearly shows their equivalence consider that also here the @xmath110 s are in the interval @xmath30 and , furthermore , they can be considered as random variables , if a hypothesis of molecular chaos is made concerning the random initial directions of the two particles entering the collision .
however , the @xmath110 s are not uniformly distributed and their most probable value drastically depends on the space dimension : the greater the dimension @xmath24 , the more unlikely it becomes that the corresponding values @xmath120 assume values close to 1 and the more probable that instead they assume a small value close to @xmath121 .
this can be seen by computing their average over the incoming directions of the two particles or , equivalently , on the orientation of the initial velocity @xmath112 of one of the two particles and of the momentum transferred @xmath107 . in @xmath24 dimensions ,
the cartesian components of a generic velocity vector @xmath122 are related to the corresponding hyper - spherical coordinates the velocity modulus @xmath123 and the @xmath124 angular variables @xmath125 through the following relations , @xmath126 using these transformations to express the initial velocity @xmath127 of the first particle and the momentum transferred @xmath107 in terms of their respective moduli @xmath128 and @xmath129 and angular variables @xmath130 and @xmath131 , the expression ( [ cos ] ) for the cosine @xmath132 becomes @xmath133 [ \sin\theta_1 \cos\theta_2 ] + \nonumber\\ & & \!\!\!\!\dots \nonumber\\ & & \!\!\!\!=\ ! [ \sin\phi_1 \sin\phi_2 \dots \cos\phi_n ] [ \sin\theta_1 \sin\theta_2 \dots \cos\theta_n ] .\end{aligned}\ ] ] the average of the square cosine @xmath134 is performed by first taking the square of ( [ v5 ] ) , integrating over the angular variables , considering that the @xmath24-dimensional volume element is given by @xmath135^{{j-1 } } d\varphi_{j } \
, , \ ] ] and finally by dividing by the total solid angle
. in the integration only the squared terms survive , obtained from the square of ( [ v5 ] ) , since all the cross - terms are zero after integration
they contain at least one integral of a term of the form @xmath136 which averages to zero . by direct integration over the angle @xmath137 and @xmath138
, it can be shown that in @xmath24 dimensions @xmath139 this means that the center of mass of the distribution for @xmath132 , considered as a random variable due to the random initial directions of the two particles , shifts toward smaller and smaller values as @xmath24 increases .
the @xmath121 dependence of @xmath140 can be compared with the wealth - exchange model with @xmath15 .
there a similar relation is found between the average value of the corresponding coefficients @xmath141 or @xmath142 in the evolution equations ( [ sp2 ] ) for the wealth exchange and the effective dimensions @xmath20 : since @xmath143 is a uniform random number in @xmath30 , then @xmath144 and from eq .
( [ n ] ) one finds @xmath145 .
the appearance of wealth inequalities in the minimal models and examples of closed economy systems considered above appears to reflect a general statistical mechanism taking place for a wide class of stochastic exchange law
besides closed economy models in which the state of the @xmath27 units @xmath146 is characterized by the values @xmath13 of a certain quantity @xmath88 ( e.g. wealth or energy ) exchanged when units interact with each other .
the mechanism involved seems to be quite general and leads to equilibrium distributions @xmath1 with a broad shape . in the special but important case of systems with a homogeneous quadratic hamiltonian or equivalently with evolution laws linear in the quantities @xmath12 and @xmath24 ( effective ) degrees of freedom , the canonical equilibrium distribution is a @xmath0-distribution @xmath16 of order @xmath147 .
the corresponding distribution for the closed economy model with a fixed saving propensity @xmath14 has the special property that it becomes a fair ( dirac-@xmath10 ) distribution when @xmath46 or @xmath148 .
the possibility for single units to exchange only a fraction of their wealth during a trade corresponding from a technical point of view to a wealth dynamics in a space with larger effective dimensions @xmath24 seems to be the key element which makes the wealth distribution less inequal .
j. angle , the surplus theory of social stratification and the size distribution of personal wealth , in : proceedings of the american social statistical association , social statistics section , alexandria , va , 1983 , p. 395 .
m. patriarca , a. chakraborti , k. kaski , g. germano , kinetic theory models for the distribution of wealth : power law from overlap of exponentials , in : a. chatterjee , s.yarlagadda , b. k. chakrabarti ( eds . ) , econophysics of wealth distributions , springer , 2005 , p. 93 .
arxiv.org : physics/0504153 | we discuss the equivalence between kinetic wealth - exchange models , in which agents exchange wealth during trades , and mechanical models of particles , exchanging energy during collisions .
the universality of the underlying dynamics is shown both through a variational approach based on the minimization of the boltzmann entropy and a complementary microscopic analysis of the collision dynamics of molecules in a gas . in various relevant cases
the equilibrium distribution is the same for all these models , namely a @xmath0-distribution with suitably defined temperature and number of dimensions .
this in turn allows one to quantify the inequalities observed in the wealth distributions and suggests that their origin should be traced back to very general underlying mechanisms : for instance , it follows that the smaller the fraction of the relevant quantity ( e.g. wealth or energy ) that agents can exchange during an interaction , the closer the corresponding equilibrium distribution is to a fair distribution .
+ + presented to the international workshop and conference on : statistical physics approaches to multi - disciplinary problems , january 07 - 13 , 2008 , iit guwahati , india |
molecular abundances and excitation are sensitive probes of physical processes taking place in molecular clouds ( e.g. , * ? ? ?
chemistry is pivotal both in controlling gas cooling and ionization balance as well as in revealing changing cloud physical conditions resulting from shocks and feedback from massive star formation and galactic structure .
millimeter telescopes are now capable of surveying many molecular transitions in other galaxies , opening up chemical studies to the rest of the universe .
single dish millimeter line surveys show that spectra of external galaxies are extremely rich in molecular lines ( e.g. * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ) . for nearby galaxies , giant molecular cloud ( gmc )
scale chemistry is within reach of current millimeter interferometers . in the first paper of this series ( meier & turner 2005 ; hereafter * ? ? ?
* ) , we exploited this richness to carry out the first chemical imaging survey of the nucleus of an external galaxy with spatial resolution high enough to separate individual gmcs .
images of eight astrochemically important species were obtained in the nearby spiral ic 342 , revealing strong chemical differentiation among the gmcs .
a principal component analysis demonstrated that the chemical variation is strongly correlated with galactic structure such as bar and arm features , as well as the presence and evolutionary state of star forming regions @xcite . in this paper
we extend the analysis to the nucleus of a second nearby spiral galaxy , maffei 2 .
we have two main goals for this survey .
the first is to test the generalizability of the results discovered in @xcite .
maffei 2 and ic 342 are at the same distance and their nuclear morphologies are similar .
both have two molecular spiral arms terminating on a central ring , with the central ring being the site of intense , young , massive star formation .
it is reasonable to expect that the two nuclei should exhibit similar chemistries if the results of @xcite for ic 342 are applicable to barred spiral nuclei in general .
the second goal is to investigate the chemistry of maffei 2 with particular attention focused on its unique features .
@xcite find that in ic 342 , shocks appear to be an important influence on the nuclear chemistry and speculate that the shocks are associated with orbital resonances in a bar potential .
the bar in ic 342 is weak @xcite ; however , and there remain other explanations for the nuclear morphology @xcite .
it is important to follow up with an object that has a well established , strong bar .
kinematic modeling leaves little doubt that the nucleus of maffei 2 hosts a strong , highly defined nuclear bar of 320 pc radius ( meier et al .
2008 , hereafter * ? ? ?
* ; * ? ? ?
because of the strength of the bar , we expect that shock chemistry should be more important than in ic 342 .
maffei 2 also has a somewhat more intense and extended starburst than ic 342 @xcite , so photon - dominated region ( pdr ) effects may be more prominent than they are in ic 342 .
to pursue these goals , we obtain aperture synthesis observations of eleven important molecular transitions over the central [email protected] kpc of the nuclear bar in maffei 2 , for comparison with co isotopologues @xcite , hcn(10 ) @xcite and cs @xcite .
table [ molp ] lists molecular parameters for each transition . with the owens valley radio observatory ( ovro )
millimeter interferometer we observed three lines , hco@xmath1(10 ) , hnco(5@xmath54@xmath6 ) and sio(10 ; @xmath7 = 1 ) .
hco@xmath1(10 ) and sio(10 ; @xmath7 = 1 ) were observed between 1999 january 28 and 1999 march 29 along with the hcn(10 ) dataset @xcite .
hnco(5@xmath54@xmath6 ) was observed along with @xmath8co(10 ) and c@xmath9o(10 ) between 1998 october 19 and 1999 january 5 ( table [ obst ] ) .
the ovro interferometer consisted of six 10.4 meter antennas equipped with sis receivers @xcite .
transitions were observed in l , h and uh array configurations except for the hnco , which was observed in l and h , providing spatial resolutions of @xmath10 ( table [ obst ] ) .
data were calibrated using the in - house ovro calibration package , mma .
phase calibration was done by observing the point source 0224 + 671 every 20 minutes .
absolute flux calibration was based on observations of neptune or uranus with 3c 273 , 3c 84 and 3c 454.3 as supplementary flux calibrators .
based on the derived fluxes and flux histories of these secondary flux calibrators we estimate that the absolute fluxes are good to 10 - 15% .
two pointings were mosaiced to cover the nuclear bar and therefore maps are corrected from the primary beam response .
the eight remaining transitions were observed with the ten element berkeley - illinois - maryland association ( bima ) array @xcite between 2002 september 08 and 2003 march 18 ( table [ obst ] ) .
the eight spectral lines were observed in two sets of spectrometer configurations .
c@xmath3h(n = 1 0 , j = @xmath11@xmath12 ) , c@xmath3h(n = 1 0 , j = @xmath12@xmath12 ) , sio(21;@xmath13 ) , hnc(10 ) and hc@xmath2n(109 ) were observed together , as were ch@xmath2oh(@xmath14 ) , c@xmath15s(21 ) and n@xmath3h@xmath1(10 ) .
data were obtained in b , c and d configurations for all transitions giving beamsizes of @xmath16 depending on frequency and sensitivity .
calibration and data reduction was done in miriad .
phase calibration was also done by observing the same point source 0224 + 671 ( 0228 + 673 ) interlaced with the source every 24 minutes .
absolute flux calibration was done by observing w 3(oh ) . based on the derived fluxes and flux histories of 0224 + 671 we estimate that the uncertainties in absolute flux are 10 - 15% , similar to the ovro data .
the @xmath17 fwhm primary beam of the bima antenna is large enough such that only one pointing was required .
no corrections for primary beam attenuation have been applied .
image analysis was done with the nrao aips package . in making the integrated intensity maps emission greater
than 1.2@xmath18 was included .
continuum emission has not been subtracted from the maps since the 3 mm continuum peak is below 1@xmath18 .
the sb galaxy maffei 2 is one of the nearest and brightest extragalactic molecular line sources , at a distance of only 3.3 mpc . a strongly barred and somewhat asymmetric galaxy ,
maffei 2 has undergone a recent interaction with a small companion that has driven a large quantity of gas into the nucleus , building up a compact central bulge seen in the nir @xcite .
this bar - like compact bulge , has further driven the molecular gas into a central molecular zone of radius @xmath19 pc that is barred and highly inclined @xcite .
this nuclear bar , containing m(h@xmath20)@xmath21 of gas @xcite , is structurally reminiscent of a miniature version of large scale gaseous bars @xcite ( fig .
[ fintro ] ) . using 2 cm and 3 mm radio / mm continuum to trace ionizing radiation
it is seen that the intersections of bar arms and central ring are sites of intense localized star formation ( sfr @xmath4 0.04 m@xmath22 yr@xmath23 per gmc or a total of @xmath24 ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
. averaged over the inner @xmath25 radius , gas consumption timescales decrease below 100 myrs , making maffei 2 s nucleus a starburst .
a central molecular ring of radius @xmath26 ( 110 pc ) is clearly resolved in gas kinematics , although the ring is less apparent in integrated intensity images @xcite .
gmcs d , e , and f ( cloud designations following mth08 are shown in figs . [ fintro ] ) are located along the eastern side of the nuclear ring .
bar arms ( including gmcs b and c in the north and g and h in the south ) extend off the leading edges of the ring .
the bright gmcs of the central ring , typically @xmath46080 pc in extent , and with masses of a few @xmath27 , delineate the sites where gas flowing in along the bar arms piles up at the inner ring .
dense gas traced by hcn preferentially picks out the arm - ring intersections .
gas not incorporated into the dense component at the starburst locations is tidally sheared into a moderate density , smoothly distributed ring @xcite .
massive star formation is largely absent in the western side of the ring @xcite . based on hcn ,
the fraction of dense gas is significantly lower towards the bar ends .
densities derived from co tend to be fairly constant ( n@xmath28 ) over the inner bar , with the clouds associated with young star formation slightly warmer ( @xmath29 30 40 k ) than the others ( t@xmath30 k ) .
gas clouds on the quiescent western side of the nuclear ring are slightly cooler yet @xcite .
column densities are determined assuming optically thin emission , and lte : @xmath31 where @xmath32 , @xmath33 and @xmath34 are the line strength , degeneracy and upper energy of each state , respectively , and @xmath35 is the excitation temperature associated with the transition . given that we have mapped only one transition of each species , corrections for background radiation and opacity have been ignored .
column densities are sensitive to @xmath35 through the partition function , @xmath36 , and the energy of the upper state .
the asymmetric tops ( hnco and ch@xmath2oh ) are more sensitive to temperature changes than are the linear rotors .
changes in gas density also affect excitation , particularly for molecules with high critical densities ( hnc , n@xmath3h@xmath1 and hc@xmath2n ) .
densities have been at least partially constrained by observation in maffei 2 @xcite .
we adopt a t@xmath37 = 10 k for the molecular transitions here .
this is similar to what is derived from the @xmath8co data @xcite and is the same as assumed for ic 342 @xcite , permitting easy comparison .
however t@xmath37 = 10 k may slightly underestimate excitation at the massive star forming regions . measured line intensities , peak temperatures , line centroids and line widths for the gmcs are reported in table [ intt ] .
fractional abundances ( @xmath38@xmath39 ) calculated from the intensities given the above assumptions are listed in table [ abut ] , based on molecular parameters in table [ molp ] .
fractional abundances also require an @xmath40 column density , n(h@xmath3 ) .
n(h@xmath3 ) is most easily obtained from the co(1 - 0 ) brightness and an empirical galactic conversion factor , @xmath41 . however , @xmath41 overpredicts n(h@xmath3 ) in nearby galaxy centers , including maffei 2 , by factors of a few @xcite .
an alternative measure of n(h@xmath3 ) can be obtained from the co isotopologues , which tend to be optically thin .
we use the @xmath8co(1 - 0 ) integrated intensity and [ h@xmath3/@xmath8co ] = @xmath42 ( [ @xmath43co/@xmath8o ] = 60 and [ co / h@xmath3 ] = @xmath45 ) when calculating the h@xmath3 column densities ( see * ? ? ? * for a detailed discussion of n(h@xmath3 ) and its uncertainties in maffei 2 ) .
fractional abundances for the species below are estimated to be uncertain to at least a factor of three , although the relative column densities that is , the relative spatial distributions within the nucleus of the different molecules should be more reliable .
@xcite has a detailed discussion of the uncertainties associated with the abundance estimates .
table [ molp ] records for reference , the change in derived column densities if t@xmath37 is changed from 10 k to 50 k , more typical of the large beam gas _
kinetic _ temperatures .
we now discuss the morphologies of each species .
spectra taken at selected locations across the nucleus with 8@xmath0 apertures are shown for each transition in fig .
[ fspec ] .
[ fintio ] displays maps of the transitions detected with ovro and fig .
[ fintib ] displays those detected with bima . in the subsequent analysis
we also include 3 mm continuum emission and co isotopologues from @xcite .
the millimeter continuum in maffei 2 is dominated by free - free emission rather than dust emission @xcite . _
sio [ morph_sio]silicon monoxide : _ the j = 21 rotational line of both the @xmath7 = 0 and 1 vibrational states were observed .
the @xmath7 = 1 , j = 2 1 transition is a maser in the galaxy @xcite , but we do not detect it in maffei 2 .
the thermal @xmath7 = 0 transition of sio is tentatively detected toward gmcs b , f and g with peak temperatures over the 8@xmath0 aperture of 35 mk ( fig .
[ fintib ] ) .
the integrated intensity map is noisy but consistent in morphology to hnco and ch@xmath2oh , extended along the southern arm .
implied sio abundances reach @xmath46 .
this abundance is lower than estimated for ic 342 and the nearby starburst nuclei , ngc 253 and m 82 @xcite , but typical of diffuse galactic clouds @xcite away from molecular outflows @xcite ; it is much larger than what is found for galactic dark clouds ( e.g. , * ? ? ?
c@xmath3h [ morph_c2h]ethynyl : _ we detect both the j = @xmath47 and @xmath48 fine structure components of the n = 10 line .
c@xmath3h is brightest towards the northern star forming ring ( gmc d+e ) , peaking at 0.11 k , and weakens significantly along the bar arms ( fig .
[ fintib ] ) . towards gmcs
f and g , c@xmath3h avoids the column density peaks traced in @xmath8co(10 ) .
line widths are significantly broader than those seen in the other weaker lines , so the hyperfine structure of each fine structure component is being marginally resolved here .
evidence for the most blueshifted n = 1 0 , j = @xmath48 , f = 1 0 hyperfine component is seen towards the brightest gmcs ( see fig .
[ fspec ] ) .
there is a plume of emission along the minor axis in the @xmath47 line . from the relative intensities of the fine structure lines we can determine the c@xmath3h opacity . the ratio of the brightness temperature of the two fine structure transitions , i(@xmath47)/i(@xmath48 ) ranges from 1.2 - @xmath42.8 , with lowest values toward the regions with the brighter emission .
for optically thin , lte excitation a this ratio should be 2.0 . therefore c@xmath3h
opacities are significant along the central ring , with implied values for the @xmath47 transition of @xmath49 1.5 . along the arms @xmath501 except at regions around gmc g away from the @xmath8co(10 ) column density peaks where @xmath51 .
hence , despite its lower signal - to - noise , the @xmath48 transition probably represents a closer picture of the true c@xmath3h abundance distribution .
fractional abundances assuming c@xmath3h is optically thin peak at @xmath52 .
if we account for opacity , the abundance rises to @xmath53 .
these are similar to those seen towards the pdr site in ic 342 @xcite and at the high end of the range seen in galactic cores @xcite . _
[ morph_hco]formyl ion : _
hco@xmath1(10 ) is the brightest line of this sample , and is imaged at high resolution ( @xmath54 ) .
the emission is dominated by gmcs d+e and f , with line peaks of 2.0 k ( 0.58 k over the @xmath55 aperture ) ( fig .
[ fintio ] ) . like hcn(10 ) @xcite , hco@xmath1(10 ) is primarily confined to the central ring region .
there is a second blueshifted velocity component at @xmath4 - 135 km s@xmath23 ( lsr ) along the northern portion of the bar .
hco@xmath1(10 ) abundances in maffei 2 are between 2 - 8@xmath56 , but these ( along with those of hcn and hnc ) should be considered lower limits since these very bright transitions may have significant optical depths ( see section [ hcohcn ] ) .
these agree well with the hco@xmath1 abundances in most galactic dense clouds ( eg . * ? ? ?
* ; * ? ? ?
* ) and single - dish extragalactic measurements @xcite . _ hnc [ morph_hnc]hydrogen isocyanide : _ the j = 10 line of the linear molecule hnc is bright over the whole nuclear bar , with peak temperatures @xmath40.25 k ( fig .
[ fintib ] ) . at this resolution ,
the morphology matches closely both @xmath8co(10 ) and hcn(10 ) , being dominated by the bright clouds , d + e and f , but hnc emission is seen from every cloud including the non - nuclear bar gmc , a. towards the northern side of the bar hnc emission is especially bright at higher blueshifted velocities , similar to the component seen in hco@xmath1(10 ) .
hnc abundances are @xmath57 , in agreement with ic 342 @xcite .
these abundances are intermediate between typical massive star forming cores and galactic dark clouds @xcite . _
[ morph_hc3n]cyanoacetylene : _ we mapped the j = 109 rotational line of hc@xmath2n .
this molecule has the largest electric dipole moment and the highest upper energy state of the ( thermal ) sample ( table [ molp ] ) .
it is thus not surprising that we observe hc@xmath2n to be faint and spatially confined ( fig .
[ fintib ] ) . to improve s / n , the hc@xmath2n map was smoothed to a slightly larger beam size , but even then peaks at only @xmath4 50 mk towards two unresolved locations at the intersection of the ring and bar arms ( c and f ) .
abundances are @xmath58 , similar to those in ic 342 @xcite , somewhat higher than in cold galactic clouds but much lower than the localized galactic hot core values ( @xmath59 ; eg .
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
n@xmath3h@xmath1 [ morph_n2h]diazenylium : _ in the galaxy , n@xmath3h@xmath1 is regarded as a tracer of quiescent gas .
we observed the j = 10 rotational state of this linear ion , which has hyperfine splitting much smaller than the observed linewidths .
as is the case in ic 342 , n@xmath3h@xmath1 is bright and extended across the nucleus ( fig .
[ fintib ] ) .
emission peaks towards gmc f with a peak brightness temperature reaching @xmath40.15 k. gmcs d and e , the closest clouds to the northern star formation region , show weak emission in n@xmath3h@xmath1 : @xmath60 reaches @xmath61 at gmc f and drops by nearly a factor of three towards the 3 mm continuum peak .
the highest abundances are similar to what is seen in ic 342 and dense , quiescent cores in the galactic disk @xcite . _
c@xmath15s [ morph_c34s]carbon monosulfide : _ this is an isotopologue of the dense gas tracer cs .
the j = 21 rotational transition of c@xmath15s is tentatively detected only towards gmcs a , c and f , at @xmath62mk ( fig . [ fintib ] ) .
the tentative detections imply @xmath63 abundances of at most a few @xmath64 , typical of galactic dense cores and diffuse / translucent clouds ( for a c@xmath65s / c@xmath15s isotopic abundance of 23 ; eg .
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
in the gmcs near the northern star forming region abundances are an order of magnitude lower . evidently c@xmath15s(21 )
is not strongly enhanced anywhere across the nucleus , and especially not at the massive star formation sites .
limits on the cs / c@xmath15s isotopologue ratio assuming optically thin cs ranges from 4 to @xmath6614 based on the cs(2 - 1 ) line of @xcite .
the tentative detections of cs / c@xmath15s @xmath4 4 suggests opacities significantly greater than unity in cs(21 ) . _
ch@xmath2oh [ morph_ch3oh]methanol : _ we observed the low energy , blended set of @xmath67e , @xmath68e , @xmath68a+ and @xmath69e thermal transitions of ch@xmath2oh ( hereafter labeled the @xmath14 transition ) .
we can not resolve these transitions in this group , given the source linewidths .
the @xmath14 transition of ch@xmath2oh is very bright across the nucleus , with peak temperatures reaching 0.28 k ( fig .
[ fintib ] ) . of the species observed in this paper
only hco@xmath1(10 ) is brighter .
remarkably though , the morphology of ch@xmath2oh is completely different from what is seen in most of the other lines , including @xmath8co(10 ) and hcn(10 ) .
ch@xmath2oh peaks strongly towards the southern nuclear bar end ( g and h ) and the southern nuclear bar arm / ring intersection region ( f ) .
by contrast , emission is barely detected near gmc d+e .
the emission favors the right side of the southern arm , which is the trailing , upstream side . at the bar ends methanol abundances peak at @xmath70 . towards gmc d abundances drop by a factor of 5 .
the peak arm abundances are a factor of two higher than those of ic 342 @xcite .
galactic methanol abundances can reach these levels in small localized hot cores regions ( eg . * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
ch@xmath2oh velocities appear consistently blueshifted with respect to most of the other transitions . in the north
the velocity shift is @xmath4 -10 km s@xmath23 , growing to @xmath4 -40 km s@xmath23 along the southern bar end .
the shift towards the northern cloud is not significant , however the shift seen towards gmcs f and g appears real .
this may result from an increased population in the higher excitation blueshifted lines that make up the quadruplet or somewhat different kinematics due to the presence of shocks ( see section [ barshocks ] ) .
_ hnco [ morph_hnco]isocyanic acid : _ we observed the k@xmath71 = 0 transition of the j = 54 rotational state ( @xmath72 ) of this prolate , slightly asymmetric top .
in the higher resolution ovro maps , hnco peaks at gmcs f and g , with faint emission along the bar connecting the two clouds ( fig . [ fintio ] ) .
when smoothed to match the resolution of the bima data , hnco has a very similar morphology to ch@xmath2oh . over the 8@xmath73 aperture used for the spectra in fig .
[ fspec ] , hnco has a peak brightness of 0.25 k , whereas over the 40 pc resolution achieved in the ovro data peak brightnesses reach 1.0 k. no emission is detected towards gmc d in the high resolution map .
hnco abundances reach @xmath38@xmath74 along the southern nuclear bar .
these abundances are typical of those observed on one to two orders of magnitude smaller scales in galactic massive dense cores @xcite and are even a factor of 3 larger than the hnco enhanced shock regions in ic 342 @xcite .
ch@xmath2oh and hnco in maffei 2 have some of the most pronounced morphological differences from co yet seen in extragalactic sources .
these two species illustrate the importance of high spatial resolution interferometric observations of chemical species .
_ other transitions : [ morph_others ] _ table [ othermolt ] presents the limits ( 2@xmath18 ) for other potentially detectable transitions within the observed bandwidth .
these include the @xmath8c substituted isotopologues of hnc(10 ) and hc@xmath2n(109 ) and two transitions of the radical hco
. none of these transitions are detected . in the case of
the @xmath8c isotopologues only the hn@xmath8c(10 ) limits are at all constraining ( table [ ratt ] ) . towards the central ring and southern
bar end 1@xmath18 hnc(10)/hn@xmath8c(10 ) line ratio lower limits are @xmath661016 , which in these locations are consistent with or slightly above the corresponding co isotopic line ratios @xcite .
therefore the hnc(10 ) opacities are lower than @xmath43co(10 ) opacities . to facilitate the exposition
, we omit the transitions for each species where possible in the remainder of the text .
we discuss transition dependent effects specifically where important .
the maps give a picture of the molecular cloud astrochemistry in the nuclear region of maffei 2 . to interpret these maps
, we need to establish similarities and differences between the molecules .
the correlations reveal trends in the chemistry , which in turn , can suggest the physical driving forces responsible for these characteristics and how they change across the nucleus . as in mt05
, we quantify morphological correlations amongst the images via a principal component analysis ( pca ) .
pca is a common technique ( eg . * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ) used to reduce the dimensionality of a dataset .
the pca simplifies the picture of molecular distribution , reducing a large amount of information to a few images , providing an excellent framework within which to study the complex variations in molecular properties in maffei 2 . for a general description of pca applied to multi - transition molecular maps ,
see @xcite or for specific applications to extragalactic molecular maps , see @xcite .
the pca is dependent on the particular set of data given to it ; here , we are analyzing the correlations of 3 mm lines of heavy molecules , so the results we obtain pertain to the cool and relatively dense cloud population in maffei 2 . to calculate the principal components for maffei 2
, the line maps were remade to the same geometry and beamsize ( that of hnc ; table [ obst ] ) , normalized and mean - centered .
thirteen maps , including @xmath8co(1 - 0 ) , hcn(1 - 0 ) and 3 mm continuum from @xcite , plus the ten transitions of the nine species discussed in section [ morph_spec ] , were sampled at 1@xmath73 intervals over the central @xmath75 , making up a 27664 element data - space .
the algorithm used to calculate the principal components is that of @xcite .
the results of the pca are tabulated in tables [ pcacort ] and [ pcat ] .
table [ pcacort ] shows the matrix of pair - wise correlation coefficients between all of the thirteen maps .
table [ pcat ] records the projection of each map onto the seven modeled pcs . fig .
[ fpcamaps ] displays the maps of the first three ( most important ) principal components ( pcs ) .
these three pcs account for 80 % of the variation present in the data set .
the projections of each map onto the first three pcs are shown vectorially in fig .
[ fpcavect ] . inspection of figure [ fpcavect ] reveals that pc 1 is most strongly correlated with all maps .
it represents a mix of the @xmath8co and hcn maps , implying that pc 1 is a good density - weighted column density map of the low excitation molecules in maffei 2 .
this was also found in ic 342 ; however , here @xmath8co projects along pc 1 more closely than hcn , hence pc 1 is nearer to a pure column density map than its analogue in ic 342 .
all maps except the low s / n ratio c@xmath15s show significant correlation with pc1 .
pc 2 characterizes the dominant variation seen once the distribution of column density has been accounted for .
pc 2 distinguishes two major distinct groups of molecules , those peaking towards the norther star forming region and those peaking towards the southern nuclear bar end . species with large positive projections on pc 2 ( star formation peakers ) include hco@xmath1 , hcn , hnc , c@xmath3h and of course 3 mm continuum , which is free - free emission . species with large negative projections ( southern bar peakers ) include hnco , ch@xmath2oh and to a lesser extent n@xmath3h@xmath1 and sio .
pc 3 has a quadrupolar morphology dominated by sio and c@xmath15s .
pc 4 and beyond are at best marginally significant so we do not discuss them in detail .
the correlation matrix resulting from the pca ( table [ pcacort ] ) shows in more detail the correlations between various species .
the following trends are noted . _
( 1 ) hcn , hnc , hco@xmath1 and 3 mm continuum are all extremely tightly correlated , indicating a close association with massive star formation .
_ at the resolution of the pca hco@xmath1 and hcn are effectively perfectly correlated . at three times higher spatial resolution
subtle differences become apparent ( fig .
[ fdenserat ] ) .
hcn is the map most tightly correlated with the 3 mm continuum .
therefore , the ( massive ) star formation rate versus hcn line luminosity relation seen in single - dish surveys ( e.g. * ? ? ?
* ) persists to at least 100 pc scales in maffei 2 , as noted in @xcite .
however , the correlation between hc@xmath2n(109 ) and 3 mm continuum is weaker than found for ic 342 . _ ( 2 ) ch@xmath2oh and hnco are very well correlated with each other and are strongly anti - correlated with the massive star formation .
_ this leaves little doubt that on these gmc spatial scales the chemistries of these two species are either directly coupled or are mutually coupled to a specific chemical mechanism .
moreover these species show little correlation with 3 mm continuum and hence star formation is not relevant to establishing the morphologies of these species . nor are these species correlated with the standard dense gas tracers , hcn , hco@xmath1 and hc@xmath2n .
[ fpv ] displays the position - velocity diagrams of @xmath8co , c@xmath3h and hnco .
the strong differences between the starburst tracers and ch@xmath2oh or hnco carry over to the kinematics .
the parallelogram originating from the central ring is nearly absent in ch@xmath2oh and hnco but remains pronounced in c@xmath3h .
_ ( 3 ) sio and especially c@xmath15s are less tightly correlated with the other species .
_ this may in part be due to their low s / n ratio and that some of the structure that appears in these maps are noise .
however c@xmath15s was also anomalous in ic 342 @xcite .
the pronounced variations in spatial morphology observed for the different molecules across the central half kiloparsec of maffei 2 are caused by either differences in emissivity ( excitation ) or abundances ( chemistry ) . in this section
we discuss the possible influences on the molecular emission that could lead to the observed variations .
the co / hcn ratio is widely used to study the relative fraction of dense gas in galaxies .
@xcite used the co / hcn ratio to identify the sites of high dense gas fraction and concluded that the @xmath76 orbital intersections and the eastern ( star forming ) portion of the central ring are the locations of the high dense gas fraction .
the fraction of dense gas drops towards the western ring and moving out along the bar arms . to study the nature of the dense component directly , line ratios between dense gas tracers are required , since co does not trace dense gas . line ratios between bright , high critical density species , such as hcn , hco@xmath1 , hnc , hc@xmath2n and cs are useful diagnostics of the physical and chemical properties of the dense component of the ism .
much observational and theoretical effort has been dedicated to understanding these diagnostic species in nearby starbursts ( e.g. * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ) . unfortunately interpreting these results can be difficult since most of these studies have been done at low spatial resolution , either with single - dish telescopes or in distant sources .
these ratios mix many interstellar environments into one beam . now that we have access to an inventory of dense gas tracers at 30 - 100 pc resolution , we investigate changes in the hcn / hco@xmath1 , hcn / hnc , hco@xmath1/n@xmath3h@xmath1 , hnc / hco@xmath1 and hcn / cs line ratios on gmc scales across the nucleus of maffei 2 . fig .
[ fdenserat ] displays four of these line ratios .
line ratios are tabulated in table [ ratt ] . organized variations in all ratios are observed on these scales and are discussed in terms of the co / hcn ratio .
the hcn / hco@xmath1 ratio relates two of the brightest lines that trace dense gas .
it has become a popular diagnostic of gas with densities intermediate between co and hcn since hco@xmath1 has a critical density nearly an order of magnitude lower than hcn ( e.g. , * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
hco@xmath1 also has a formation route that maintains significant abundances in relatively diffuse gas ( see below ) , and it is an ion making it more sensitive to collisions with electrons ( e.g. * ? ? ?
* ) ( see table [ molp ] ) . as a result hco@xmath1
can remain abundant and excited in diffuse gas @xcite .
theoretical modeling finds that hcn / hco@xmath1 is a strong function of density for a given column density in the presence of ionizing radiation , with ratios of unity for densities of @xmath410@xmath77 @xmath78 ( e.g. * ? ? ?
* ; * ? ? ?
there is , however , the complication of opacity . despite its lower critical density hco@xmath1 has a higher einstein a@xmath79 , implying higher radiative efficiencies than hcn ( and hnc ) . for a given abundance per line width ,
hco@xmath1 exhibits a line opacity nearly twice that of hcn and hnc . for extragalactic studies
it is usually assumed that hcn abundances are significantly higher than hco@xmath1 .
however , for the lte abundances calculated in @xmath80 [ abuintro ] and @xcite , hco@xmath1 would have approximately twice the observed abundance of hcn , contrary to what is commonly assumed .
these abundances have large uncertainties but the two lines have similar energy levels . the similarity in excitation requirements of the two lines
makes it difficult to reverse this situation purely by changing gas temperatures .
if the lte abundances are correct then hco@xmath1 line opacities must be at least as large as the presumably large hcn line opacities .
if opacities in both lines are greater than unity the diagnostic power of hcn / hco@xmath1 is limited . towards the nuclear bar in maffei 2 the overall morphologies of hco@xmath1 and hcn are similar .
however , detailed differences in the two transitions are revealed in their ratio .
hcn / hco@xmath1 ranges from 0.62.2 over 30 pc scales in maffei 2 .
_ these values cover much of the range sampled in the single - dish surveys of many galaxies _ @xcite .
both hcn @xcite and hco@xmath1 fall off more rapidly with galactocentric distance than does @xmath8co(10 ) .
hco@xmath1 decreases more slowly than hcn in the north ( gmc b ) but more rapidly along the southern bar arm .
hcn is elevated relative to hco@xmath1 on the active eastern side of the central ring , gmc g and a localized cloud just south of gmc f , while hco@xmath1 is enhanced , relatively speaking , in the quiescent , western ring region .
we detect clouds with co / hcn vs. hcn / hco@xmath1 values that reside in the `` agn dominated '' portion of parameter space set by global single - dish studies where clearly there is no agn @xcite .
similar statements apply for `` star formation dominated '' parameters . as a result interpretation of single - dish values
must be done with great care .
the enhancement of hco@xmath1 towards the western ring and hcn towards the northeastern ring makes sense in the conventional critical density context , since one expects the denser gas to be more closely associated with the massive star formation .
however , the elevated hcn / hco@xmath1 ratio towards the southern bar end is unusual .
a decreasing hcn / co ratio suggests the fraction of dense gas in the south is significantly lower , which should favor hco@xmath1 .
that this is not seen implies some other excitation mechanism or chemistry is in play in the gas of the southwestern arm . from the distribution of shock tracers ( section [ gasgrain ] ) , gmc g must be one of the clouds most effected by shocks .
it is tempting , therefore , to suggest that the southern bar end in maffei 2 is an example of a location of elevated hcn abundances in shocks . in the galaxy , bipolar molecular outflows
do show hcn much more strongly elevated than hco@xmath1 ( e.g. * ? ? ?
this scenario seems less favored when the other dense gas tracers are incorporated . towards gmc g , both the hco@xmath1/n@xmath3h@xmath1 and hco@xmath1/hnc ratios are anomalously low , while the hcn / hnc ratio remains roughly constant relative to the rest of the ( non - star forming ) nuclear disk .
this argues for a depletion / destruction of hco@xmath1 rather than an enhancement of hcn .
the decrease in hco@xmath1 intensity here may stem from the shock in a more indirect way .
the passage of a shock alters the chemical state of the molecular gas , including a dramatic increase in gas - phase water , which is effective at destroying hco@xmath1 @xcite .
the main difficulty with this mechanism is that h@xmath3o is also expected to destroy n@xmath3h@xmath1 , which is not observed .
an alternate and perhaps simpler explanation for the elevated hcn / hco@xmath1 ratio and the brightness of high density tracers like hnco and n@xmath3h@xmath1 at the southern bar end is that while the dense gas fraction decreases along the southern bar arm , the density of the dense component does not . in this context
the correlation between the dense gas tracers and co / hcn towards the star forming ring would imply that the dense gas fraction and dense gas density are coupled there .
the difference between the two environments could be a result of the different dynamical environments at each location . towards the southern bar end
one can envision a situation where the shocks ( section [ barshocks ] ) produce a low filling factor , high density component that explains the behavior of both ratios .
the fact that the hnco / ch@xmath2oh ratio remains high gives additional direct evidence that densities of the shocked gas are quite high towards gmc g ( section [ relshocks ] ) . in summary
, while the co / hcn ratio suggests a rapidly decreasing dense gas fraction along the arms @xcite , the hcn / hco@xmath1 ratio suggests that the actual density does not fall significantly
. clearly not all the properties of the dense component are traced in the co / hcn ratio .
this further emphasizes the fact that the _ co / hcn measures the fraction of the gas that is dense , not necessarily the actual density of the gas .
_ ratios that sample only dense gas , such as hcn / hco@xmath1 , are needed to determine the physical conditions of the dense component .
the hcn / hnc line ratio is another dense gas ratio of increasing popularity , in which high values indicate hot gas and/or pdr conditions .
essentially identical a@xmath79 , collision coefficients , upper energy states and frequencies make the ratio particularly insensitive to changing excitation
. dissociative recombination of hcnh@xmath1 dominates the formation of both hcn and hnc in cool , dense clouds , producing both species in equal amounts ( e.g. * ? ? ?
so in these environments the abundances of the two species are locked together .
additionally both transitions likely have significant opacities .
it is therefore not surprising that this ratio remains near unity over a wide range of conditions .
when gas temperatures get large hnc can be converted to hcn via hnc @xmath81 h @xmath82 hcn @xmath81 h , while in diffuse gas / pdr - like environments , the reaction n + ch@xmath3 @xmath82 hcn @xmath81 n becomes a major additional hcn source ( e.g. * ? ? ?
* ; * ? ? ?
[ fdenserat ] ( table [ ratt ] ) displays the hcn / hnc intensity ratio map .
indeed the ratio remains fairly constant across the nucleus with values of @xmath41.1 @xmath830.3 .
variations in the ratio are only at the 30 % level over most of the nucleus , with the highest values on the star forming eastern ring and south of gmc f , which suggests mild enhancement by pdrs .
ratios significantly below unity on these small spatial scales are not seen anywhere across the nucleus .
hence there is no evidence for anomalous enhancement of hnc relative to hcn that has been found in some ulirgs @xcite . in ic 342 ,
the hcn / hnc intensity ratio is also fairly constant , with weak enhancement seen towards the pdr gmc @xcite . in summary , at least in maffei 2 and its less active cousin , ic 342 , the hcn / hnc line ratio has limited diagnostic power .
its diagnostic power becomes stronger for molecular gas with extreme temperatures and density , such as lirgs and ulirgs , where the chemistry is driven away from dissociative recombination of hcnh@xmath1 .
both hcn and hco@xmath1 ( and to a lesser degree hc@xmath2n ) are observed to maintain large abundances in pdrs , shocks and other energetic regions . by contrast , in the galaxy n@xmath3h@xmath1 is observed to avoid pdr regions , maintaining high abundances only in dense , quiescent , dark cloud - like conditions @xcite . in dense , quiescent gas both n@xmath3h@xmath1 and hco@xmath1 form from proton transfer reactions of h@xmath84 with n@xmath3 and co , respectively : @xmath85 @xmath86 given the reasonable assumption that the rate coefficients for these two reactions , @xmath87 and @xmath88 , are approximately equal @xcite , then @xmath38(n@xmath3h@xmath89)/@xmath38(hco@xmath1 ) @xmath90 @xmath38(n@xmath3)/@xmath38(co ) .
however , only hco@xmath1 maintains a relatively rapid formation route in diffuse or pdr gas ( c@xmath1 + oh @xmath82 co@xmath1 + h followed by h@xmath3 + co@xmath1 @xmath82 hco@xmath1 + h ) .
no corresponding route exists for n@xmath3h@xmath1 in pdrs @xcite .
this explains the observed behavior of the hco@xmath1/n@xmath3h@xmath1 line ratio ( fig .
[ fdenserat ] and table [ ratt ] ) . towards the pdr regions associated with the active star formation sites ,
hco@xmath1 can maintain appreciable abundances while n@xmath3h@xmath1 is readily destroyed .
the fact that this ratio , which reflects dense gas properties , is modified by the starburst pdrs implies that towards this region ( gmc e and the eastern ring ) pdrs penetrate the dense gas . _ the strong correlation between high hco@xmath1/n@xmath3h@xmath1 ratios and pdr tracers like c@xmath3h , suggests that this ratio may be a powerful new probe of * dense * pdrs in external galaxies . _ along the arms , @xmath38(n@xmath3h@xmath1)/@xmath38(hco@xmath1 ) abundances are typically @xmath40.1 - 0.15 , with values reaching up to @xmath40.33 towards the southern bar end .
if a standard co abundance of @xmath38(co ) @xmath90 10@xmath91 is adopted , @xmath38(n@xmath3 ) abundances of @xmath92 are implied .
if metallicities are solar ( e.g. * ? ? ?
* ) and all n is in the form of n@xmath3 , then @xmath38(n@xmath3 ) @xmath93 .
evidently , in the more quiescent dense clouds of the nucleus of maffei 2 , at least a quarter of all nitrogen is in molecular form . using the above @xmath38(n@xmath3 ) estimation together with table [ abut ] gives @xmath38(n@xmath3)/@xmath38(n@xmath3h@xmath1 ) @xmath94 for these regions .
following the analysis of @xcite , @xmath38(n@xmath3)/@xmath38(n@xmath3h@xmath1 ) provides indirect constraints on the cosmic ray ionization rate , @xmath95 . adopting @xmath38(n@xmath3)/@xmath38(n@xmath3h@xmath1 ) @xmath96 and the extremely conservative lower limit , @xmath97 = @xmath98 @xmath99n@xmath3h@xmath1 ) + x(hco@xmath1 ) @xmath100 , figure 9 of @xcite indicates @xmath101 s@xmath23 @xmath102 .
less conservatively , for @xmath103 typical of galactic star forming regions ( e.g. * ? ? ?
* ) , @xmath104 s@xmath23 @xmath102 .
therefore , @xmath95 is at least the galactic value if the density of the gas that the n@xmath3h@xmath1 and hco@xmath1 emission originates in is @xmath105 @xmath78 .
these constraints are comparable to though somewhat stronger than observed for ic 342 .
@xcite has published a map of cs(21 ) towards the nucleus of maffei 2 with similar spatial resolution to the data presented here .
we make use of this map to compute the cs / c@xmath15s and cs / hcn ratios towards the gmcs .
the cs / hcn intensity ratio is fairly constant across the nucleus except towards the southern arm ( gmc g ) , where it is a factor of two lower .
if strong chemical evolution has not altered the @xmath65s/@xmath15s ratio , the c@xmath15s emission implies the regions of high cs opacity are along the southern arm and not at the cs peak near the massive star formation .
this is unexpected . for a @xmath65s/@xmath15s ratio of 23 ,
lte line ratios imply cs opacities of @xmath46 along the southern arm . comparing the observed peak temperature of c@xmath15s gas to optically thin emission with t@xmath37 = 10 k yields a filling factor , @xmath106 .
given the differences in beam size this filling factor is intermediate between @xmath8co and hcn @xcite . towards the gmcs d+e the larger cs / c@xmath15s ratio implies that cs opacities here must be less than 1.4 . since the intensity of the main isotopologue peaks towards this region , gas temperatures or filling factors are much larger here .
so in order to explain the morphology of both cs and c@xmath15s , it is required that cs be warmer and more diffuse towards gmcs d+e when compared to the gmc g. in summary cs appears compatible with the interpretation based on hcn / hco@xmath1 that the southern bar end has high density but low dense gas fraction .
hc@xmath2n has a critical density that is comparable or slightly higher than hcn but very likely has low opacity due to its lower abundance .
moreover its levels are more closely spaced and so population is distributed over a wider range of levels , making any one level more sensitive to changes in excitation .
hc@xmath2n originates primarily from the arm - ring intersection regions similar to hcn and hco@xmath1 , although perhaps peaks slightly farther from the center of the galaxy than does hcn . using the kinematics to further resolve the emission suggests that hc@xmath2n is tracing the inner portion of the arm rather than the @xmath76 intersection region
evidently , densities increase along the arms as the central ring is approached .
unlike hcn , hc@xmath2n weakens just before the strong star formation is reached at gmc d + e and just north of gmc f. potentially this results from warmer temperatures depopulating the j = 10 level towards the starburst .
ionizing radiation from massive stars impinging on nearby molecular clouds changes their chemical composition .
large abundances of ionized carbon in pdrs drive a rich hydrocarbon chemistry on cloud surfaces .
one of the simplest species from such regions is c@xmath3h .
c@xmath3h is formed from rapid dissociative recombination of c@xmath3h@xmath107 coming from reactions between c@xmath1 + ch@xmath2 @xcite .
_ c@xmath3h chemistry is particularly closely tied to ionized carbon and is an excellent probe of these pdrs . _
its usefulness as an extragalactic pdr tracer was demonstrated in @xcite .
after accounting for opacity effects ( see @xmath80[morph_spec ] for c@xmath3h(10 ; @xmath48 ) for an optically thin c@xmath3h transition ) c@xmath3h pervades much of the molecular clouds in the northern star forming region .
lower level c@xmath3h emission extends over the rest of the nucleus but avoids the cloud cores , instead favoring the side of the dense cloud facing star formation ( traced here by 2 cm radio continuum ) ( see fig .
[ cchoutf ] ) .
this is particularly clearly seen west of gmc f and northeast of gmc g. as discussed in @xmath80[hcon2h ] , the dramatic increase in the hco@xmath1/n@xmath3h@xmath1 line ratio across the central ring shows not only that pdrs are prevalent near the young , star forming region but that the pdrs penetrate much of the dense gas there .
given the distinct region of altered chemistry , constraints can be placed on the range of influence of the current burst of star formation . for the northern star forming region ,
the region of elevated hco@xmath1/n@xmath3h@xmath1 ratio is 80 pc diameter in extent .
we assume that the volume of the region emitting strong pdr emission reflects the volume of molecular gas engulfed within hii regions that have expanded into the ambient medium due to over - pressure .
this should be reasonable on gmc scales .
for an ionized gas sound speed of 10 km s@xmath23 , n@xmath108 s@xmath23 and a uniform density of @xmath109 @xmath78 typical of the central ring @xcite , a standard expanding hii region would reach a radius of 40 pc in @xmath47 myr . while this analysis is sensitive to the exact density distribution , it is interesting that this timescale for dynamical evolution of gmcs due to pressure is an order of magnitude longer than characteristic chemical timescales and consistent with the timescale of massive star evolution . _
c@xmath3h emission appears to trace a molecular outflow along the minor axis ( fig .
[ cchoutf ] ) . _
the northwest extension of this bipolar flow is redshifted by @xmath4 15 km s@xmath23 and the southeast blob blueshifted by @xmath11015 km s@xmath23 with respect to the velocity of the corresponding in - plane material .
these velocity shifts have the correct sign for outflow given maffei 2 s geometry ( northern arm being nearer to us , and the arms trailing ) .
there is tentative evidence for the flow in hnc and sio ( and possibly co(10 ) ; * ? ? ?
* ) but the molecular outflow is not obvious in any of the other species .
this is understandable given the strength of c@xmath3h in the outflow reflects its relatively low critical density and its ability to maintain high abundances in the presence of ionizing radiation .
the molecular outflow in maffei 2 does not appear to follow the distribution of ionized gas traced by 2 cm .
this is different from what is seen in the much more pronounced outflow in m82 @xcite .
the 2 cm radio continuum emission appears to trace the walls of maffei 2 s outflow ( fig .
[ cchoutf ] ; * ? ? ?
. the c@xmath3h emission must trace entrained molecular material being driven out directly .
the explanation for the 2 cm emission along the edges remains unclear but may be due to a path length effect of a cone of ionized gas surrounding the molecular flow . since evidence is seen for a small minor axis outflow it is possible to obtain a second estimate of the characteristic age by assuming that the gas traced in c@xmath3h is free to flow ballistically along the minor axis .
the extent of the outflow is @xmath83200 pc . for an inclination angle of 67@xmath111 and the measured line of sight velocity
, the time of flight corresponds to 6 myr .
this is a very young outflow .
the agreement with the characteristic in - plane timescale is excellent , suggesting that the current starburst may have an age of @xmath112 6 - 7 myr .
elevated gas - phase abundances of species whose formation is sensitive to the grain processing , such as ch@xmath2oh , hnco , or sio , are important probes of hot core gas and shocks .
as noted in @xcite , and illustrated even more clearly here , these species do not trace hot cores .
typical galactic hot cores , even significant collections of them can not produce the required emission necessary to explain the intensities on these large scales .
moreover , we detect no spatial connection between ch@xmath2oh and hnco , and star formation , if anything , we see an anticorrelation .
we can , therefore , safely conclude that the above three species are not tracing hot core emission or outflows from young stellar objects but must originate in large scale shocks .
inter - comparisons can be used to identify the responsible shock source @xcite and place constraints on shock properties such as shock strength , type of shock and liberated grain / mantle composition @xcite . in this dataset
three shock tracers , ch@xmath2oh , hnco and sio , are mapped .
each is sensitive to somewhat different shock properties .
it is expected that elevated gas - phase sio results from release of si from destruction of silicate grain cores ( e.g. * ? ? ?
* ) . that the grain core must be disrupted requires strong shocks capable of sputtering grain cores .
high observed abundances of gas - phase ch@xmath2oh likely come from direct liberation of ch@xmath2oh that has formed on grain surfaces by complete hydrogenation of co ( e.g. * ? ? ?
* ; * ? ? ?
liberation of mantles does not require shocks as strong as those that effect the grain core .
thus the relative enhancement of sio versus ch@xmath2oh will depend on the strength of the shock .
grain core destruction requires shock velocities of @xmath11325 km s@xmath23 whereas shock velocities of only @xmath113 10 - 15 km s@xmath23 are required for significantly elevated gas - phase ch@xmath2oh abundances ( e.g. * ? ? ?
* ; * ? ? ?
it is reasonably well established that hnco is a mantle species along with ch@xmath2oh @xcite .
but its formation mechanism remains unclear .
the most commonly proposed mechanism is partial hydrogenation of ocn on grain surfaces followed by direct liberation ( e.g. * ? ? ?
other suggested mechanisms include acid - base reactions between co and nh@xmath2 and uv photolysis @xcite .
thus if hnco forms on grain mantles and is liberated directly then its unsaturated nature implies that it originates in grain mantles with different chemical composition from ch@xmath2oh ( however see @xcite for a possible formation mechanism in more saturated ices ) .
unsaturated ices are favored when gas accretes in atomic hydrogen - poor environment whereas saturated ices favor atomic hydrogen - rich accretion ( e.g. * ? ? ?
* ; * ? ? ?
hydrogen - poor conditions reflect accretion at high densities when almost all hydrogen is in molecular form .
if it is assumed that the shock simply injects molecules into the gas phase , the hnco / ch@xmath2oh abundance ratio is then sensitive to gas physical conditions during ice mantle formation .
chemistry is not the only effect that influences the relative abundance of these three species .
excitation is also important
. the excitation requirements of sio(21 ) , ch@xmath2oh(@xmath14 ) and hnco(@xmath72 ) transitions differ ( see table [ molp ] ) .
the critical density of the hnco transition is an order of magnitude higher than sio and ch@xmath2oh .
furthermore the energy of the upper state in hnco(@xmath72 ) is more than twice that of sio and higher than two of the three main transitions that make up the ch@xmath2oh line .
therefore , ch@xmath2oh and sio lines will be favored over hnco in lower density gas .
also in the presence of ionizing radiation each species dissociates at different rates .
hnco is photodissociated at a rate twice that of ch@xmath2oh and @xmath430 times that of sio @xcite . thus considering
the distribution of ch@xmath2oh , hnco , and sio can give insights into both the chemical and physical conditions present in the shocked gas and potentially tell us about nature of the shocks themselves .
below we discuss the overall structure of the shock tracers in the context of nuclear bar models , then we analyze the subtle differences between each shock tracer to further constrain shock properties . in maffei 2 the outer bar arms
dominate the overall distribution of the shock tracers ch@xmath2oh , hnco , and sio , whereas emission from the star forming ring is minimal .
this is very different from the morphology of co , hcn and the pdr tracers .
the bar ends are especially bright in the shock tracers .
( the faintness of shock tracers along the northern arm is presumably due to the fact that there is much less molecular gas there , particularly at what would be the location of the northern bar end . )
it is perhaps surprising that it is the bar ends and not the inner arms that have the largest fraction of the gas influenced by shocks .
this provides constraints on the nature of gas flows in maffei 2 .
different models exist for describing the structure of gas bars , each predicting different shock strengths and locations .
the models fall into two broad classes , orbit - crowding models and hydrodynamical flows .
orbit - crowding models treat the gas as distinct clouds orbiting on essentially ballistic orbits set up by the potential ( e.g. * ? ? ?
* ; * ? ? ?
the arm morphology of the bar is then determined by where orbits converge . in hydrodynamical models
the gas is treated as a fluid that responds to the stellar potential and is susceptible to hydrodynamical forces such as pressure and viscosity . in such models , gas flows along streamlines with shocks occurring where the streamlines collide ( e.g. * ? ? ?
* ; * ? ? ?
both models predict similar morphological distributions but rather distinct shock properties ( e.g. * ? ? ?
* ) . in orbit - crowding models
the bar arms result from molecular clouds moving along converging orbits . in such models
only weak shocks are expected because the bulk of the velocities are parallel to both clouds ( with the possible exception of arm / nuclear ring intersections ) .
moreover such models do not isolate the bar ends as locations of strongly convergent orbits and so it is not expected that these location would be sites of vigorous shock activity .
hydrodynamical models , on the other hand , show gas flow streamlines that change directions dramatically .
the places where flows converge are accompanied by strong changes in velocity , and hence strong shocks . in hydro models ,
gas drifts slowly inward along bar - aligned @xmath114 stellar orbits ( or in the case of maffei 2 , from radial inflow along the outer bar ) until the gas reaches the innermost , cuspy @xmath114 orbit ( of the nuclear bar ) .
the cusps of these orbits lie at the bar ends .
there the gas is compressed , shocks and plunges nearly radially inward along the bar arm until it reaches the much further in , perpendicular to the bar @xmath115 orbitals , where it is again shocked , loses angular momentum and transitions onto the @xmath115 orbitals .
some high angular momentum gas sprays off the inner @xmath115 orbits , swinging out and colliding with the arm gas flow ( e.g. * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
the amount , location and relative velocity of gas in the spray region depends on bar parameters . _ in summary , the strength of the emission from the shock tracers ch@xmath2oh , hnco , and sio along the bright bar ends favors the hydrodynamical bar models , as orbit - crowding models do not predict strong cloud collisions at the bar end whereas they are sites of shocks in hydrodynamical models . _ what about changes in shock properties across the nuclear bar ?
[ hncoch3oh ] displays the hnco / ch@xmath2oh , sio / ch@xmath2oh and sio / hnco peak temperature ratios smoothed to 8@xmath73 and sampled at the gmcs .
given the significant difference in the native resolution of the three datasets and the fact that they were obtained with two different interferometers , only the clearest trends are discussed .
as demonstrated by the principal component analysis , the morphologies of the three species are strongly correlated , implying that we have a robust measure of the locations of shocked gas .
_ so similar are the maps that only very subtle differences between hnco , ch@xmath2oh and sio appear , suggesting that shock conditions are uniform across the nucleus . _ with the possible exception of the uncertain sio ratios towards gmc b , the arm clouds ( c , d , f , g ) have effectively constant hnco / ch@xmath2oh ( @xmath116 ) , sio / ch@xmath2oh ( @xmath117 ) and sio / hnco ( @xmath118 ) ratios . only towards gmc e and the off arm spray regions ( gmc h and just south of the western ring ) are different ratios observed . towards gmc e ( the northern star forming region ) , hnco / ch@xmath2oh is lower and the sio ratios are higher relative to the rest of the arm . in the off arm regions hnco / ch@xmath2oh ratios
are depressed while the sio / ch@xmath2oh ratio is elevated and sio / hnco are elevated ( spray region ) or flat ( gmc h ) . from these ratios ,
three main points are concluded .
1 ) while gmc g ( the bar end ) is the brightest source in the nuclear bar , indicating that the largest fraction of gas along this line of sight is shocked , evidently shock strengths are not enhanced here .
sio to mantle species have the same basic ratio at the bar end as it has over the rest of the bar .
\2 ) towards gmc e , the decreased intensity of hnco relative to ch@xmath2oh and sio is consistent with what is expected from photodissociation . from section [ pdr ] , we know that the dense gas here is permeated by uv photons for the young star formation . the shock tracer ratios further confirm this .
because of this , the absence ( or faintness ) of shock tracers towards the central ring and in particular the northern @xmath76 orbit intersection can not be taken as evidence of absence of shocks here .
\3 ) the spray region upstream ( counterclockwise ) from the main bar arms favor ch@xmath2oh and to a lesser degree sio relative to hnco .
it is possible that because gas is more diffuse in this region grain mantles are h - richer than in the denser arm .
however , we favor a simpler explanation associated with changing excitation .
the spray regions have lower density hence the higher critical density hnco transition is expected to be weaker relative to the lower critical density , ch@xmath2oh and sio transitions . that the @xmath72 transition of hnco preferentially favors the dense clouds relative to more extended ch@xmath2oh explains why larger beam single - dish multi - line observations derive an excitation temperature for hnco that is nearly a factor of @xmath119 larger than derived for the ch@xmath2oh transitions in maffei 2 @xcite .
note that the excitation temperature of both species is quite low ( t@xmath37 @xmath12015 k ) , implying that the shock regions remain cool or the densities are low enough for these transitions to be sub - thermal .
_ in summary , we conclude that there is little evidence for strongly varying grain chemistry across the shocked regions on 8@xmath73 scales .
the small level of variation seen is consistent with changing physical conditions ( radiation field and gas density ) . _
the overall nuclear structures of maffei 2 and ic 342 are similar .
both have extended bar - like arms that terminate on a central ring that is actively forming stars .
the molecular gas morphology of ic 342 is consistent with that of a gaseous bar @xcite .
however , because of the very small central ring and ambiguous kinematics , other conclusions including unbarred spiral arms coupled with massive star feedback remain feasible @xcite . on
the other hand maffei 2 is clearly a nuclear bar based on both its morphology and kinematics ( e.g. * ? ? ?
* ; * ? ? ?
given the potential dynamical similarities of the two nuclei a comparison of the chemistry of the two can provide important insights on the generalizability of the the overall chemical picture developed in @xcite .
the statistically robust principal component analysis demonstrates that the chemistry of the two nuclei are remarkably consistent .
maffei 2 can be defined chemically as a highly inclined copy of ic 342 ( fig .
[ fgenchem ] ) . in both systems
the first principal component characterizes the density - weighted column density , and peaks along the inner arms and central ring .
the second principal component separates the main chemical forces responsible for controlling observed abundances , namely shocks versus pdrs .
pdr species dominate close to the young star forming region while the shock tracers peak along the arm , with preference for the cuspy bar ends . as a result ,
the bar driven star formation / chemical scenario outlined in @xcite must apply also to maffei 2 s nuclear bar .
however the chemical picture is not precisely identical , and these differences give added insight into the relative differences in evolutionary phase of the two nuclei .
the main difference between the two galaxies is in the distribution of n@xmath3h@xmath1 . in ic 342
the pdr tracers are confined , appearing to be primarily cloud surface features , lining the inner ring .
significant n@xmath3h@xmath1 ( and hnco ) seen near the young star forming region in ic 342 implies that its main star forming region influences the dense gas only locally , on scales of 50 pc or less . _
whereas in maffei 2 , the n@xmath3h@xmath1/hco@xmath1 ratio and photodissociation of hnco clearly demonstrate that the star formation has begun to dramatically alter the chemical state of the dense molecular clouds in its immediate environment , not just the surfaces .
_ evidently the central molecular ring in maffei 2 is structurally different ( lower clump density ) and that towards gmc e star formation is somewhat more evolved than the main burst in ic 342 .
this is consistent with the steeper radio spectrum towards gmc e @xcite .
the crude estimates of timescale required to alter the chemistry of the dense component is @xmath4 7 myrs .
this is a reasonable timescale compared to stellar evolution of the burst in maffei 2 and the kinematics of the outflow .
however , this timescale is still very short ; short enough that even this evolved part of the burst is in its early stages .
the star formation extending to the north and south of gmc e is likely as young as the embedded star formation in ic 342 . in ic 342
the peak temperature ratios between sio , ch@xmath2oh and hnco , are similar to those seen in maffei 2 .
the pdr gmc in ic 342 has slightly lower hnco / ch@xmath2oh and elevated sio / hnco ratios .
this is exactly the behavior exhibited by the pdr - modified gmc e in maffei 2 , lending strong support for the influence of photodissociation on shock tracer brightness .
the shock chemistry dominated cloud in ic 342 has slightly elevated hnco relative to the other species , making its set of ratios a match within the uncertainties of gmc g. this agreement suggests that both the grain composition and shock properties are the same for these structurally similar locales .
however , since brightnesses are sensitive to radiative transfer effects , one must be careful over - interpreting ratios if gas excitation is changing rapidly .
maffei 2 and ic 342 do have rather similar co excitation conditions @xcite . in summary
, we find that when the dynamical environments are similar , as they are in ic 342 and maffei 2 , a consistent chemical picture begins to emerge .
gas chemistry is confirmed as a powerful method for isolating the physical mechanisms that influence gas conditions in spiral nuclei . in future papers in this series
, we will broaden the range of galactic environments studied to assess how far chemistry can be pushed as a diagnostic of sub - beam physics .
this paper presents 2@xmath0 - 10@xmath0 imaging of transitions from nine separate molecules across the nuclear bar in maffei 2 obtained with the bima and ovro millimeter interferometers .
the results are compared with the first paper in the series on ic 342 to test the generalizability of the chemistry discovered there and to constrain differences in structure between the two nuclei .
our main conclusions are as follows : 1 .
morphologies of the observed transitions exhibit both dramatic ( e.g. , c@xmath3h vs. ch@xmath2oh ) and subtle ( e.g. , hcn vs. hco@xmath1 ) spatial differences indicative of the varying dominance of shocks vs. radiation fields as drivers of the chemistry in individual clouds .
a principal component analysis reveals that hcn , hnc , hco@xmath1 and 3 mm continuum are all extremely tightly correlated , indicating a direct connection to massive star formation .
ch@xmath2oh and hnco are tightly correlated with each other and are strongly anti - correlated with the tracers of massive star formation ; based on their spatial distribution , it appears that these two molecules are formed by liberation of grain mantles in the presence of bar shocks .
4 . variations in the ratios amongst the popular @xmath121 3 mm dense gas tracers , hcn , hco@xmath1 , hnc , cs and hc@xmath2n , are at or less than a factor of three , unlike the order of magnitude variations exhibited by the chemistry - sensitive molecules .
this indicates that variations in excitation in the cool , dense gas are much less dramatic than variations in dense gas chemical conditions .
localized variations in the ratios hcn / hco@xmath1 , hcn / hnc , hcn / cs and hcn / co can be explained by variations in density .
the dense gas ratios indicate that the highest density gas is found at the @xmath76 orbit intersections , the lowest density gas is along the non star - forming western portion of the ring , and intermediate density gas is found near the star formation .
although high density gas is found throughout the full extent of the nuclear bar arms , the dense gas fraction falls at larger radii from the nucleus . 6 .
there is no evidence for x - ray dissociation regions or agn influences on the gas in the center of maffei 2 , specifically , as traced by the hcn / hco@xmath1 ratio .
the strong correlation between high hco@xmath1/n@xmath3h@xmath1 ratios and the pdr tracer molecule , c@xmath3h , suggests that this ratio may be a powerful new probe of dense pdrs in external galaxies .
given the high critical density of both hco@xmath1 and n@xmath3h@xmath1 in comparison to c@xmath3h , the elevated ratio demonstrates that pdr gas towards the strong star forming regions permeate the dense clumps .
when combined with the morphology of c@xmath3h this ratio may provide useful insights into the relative density of pdr gas .
c@xmath3h emission reveals a molecular outflow along the minor axis that has entrained molecular gas .
wrapped around the outside of this outflow is ionized gas traced in 2 cm radio continuum . assuming ballistic motion , the morphology and kinematics are consistent with a very young age of 6 - 7 myrs for the outflow .
the bar induced orbital shocks in maffei 2 are traced in three species , hnco , ch@xmath2oh and sio .
all three species support the conclusion that the nuclear bar end is the site of the largest fraction of shocked gas .
that this location exhibits such strong shock properties argues in favor of hydrodynamical bar models .
we appreciate the support and assistance of the faculty , staff and postdocs at bima and ovro during the observations . we l. e. snyder and e. c. sutton for helpful discussions .
the referee is acknowledged for a thorough and very helpful report .
dsm acknowledges support from nsf ast-0506669 , ast-1009620 and the national radio astronomy observatory which is operated by associated universities , inc .
, under cooperative agreement with the national science foundation
. additional support for this work is provided by nsf grant ast-0506469 to jlt .
aalto , s. , polatidis , a. g. , httemeister , s. & curran , s. j. 2002 , , 381 , 783 aladro , r. , martn - pintado , j. , martn , s. , mauersberger , r. , & bayet , e. 2011a , , 525 , a89 aladro , r. , martn , s. , martn - pintado , j. , mauersberger , r. , henkel , c. ochaa , b & amo - baladrn , a 2011b , arxiv:1109.6236 anders , e. & grevesse , n. 1989 , , 53 , 197 athanassoula , e. 1992 , mnras , 259 , 328 baan , w. a. , henkel , c. , loenen , a. f. , baudry , a. , & wiklind , t. 2008 , , 477 , 747 bachiller , r. , liechti , s. , walmsley , c. m. , & colomer , f. 1995 , , 295 , l51 bachiller , r. & perez gutierrez , m. 1997 , , 487 , l93 bayet , e. , aladro , r. , martn , s. , viti , s. , & martn - pintado , j. 2009 , , 707 , 126 benson , p. j. , caselli , p. , & myers , p. c. 1998 , , 506 , 743 bergin , e. a. , melnick , g. j. , & neufeld , d. a. 1998 , , 499 , 777 blake , g. a. , sutton , e. c. , masson , c. r. & phillips , t. g. 1987 , , 315 , 621 byrd , g. , rautiainen , p. , salo , h. , buta , r. , & crocher , d. a. 1994 , , 108 , 476 caselli , p. , hartquist , t. w. , & havnes , o. 1997 , , 322 , 296 caselli , p. , walmsley , c. m. , terzieva , r. , & herbst , e. 1998 , , 499 , 234 charnley , s. b. , tielens , a. g. g. m. , & rodgers , s. d. 1997 , , 482 , l203 chin , y .-
, henkel , c. , whiteoak , j. b. , langer , n. , & churchwell , e. b. 1996 , , 305 , 960 chung , h. s. , osamu , k. , & masaki , m. 1991 , jkas , 24 , 217 combes , f. , & gerin , m. 1985 , , 150 , 327 costagliola , f. et al .
2011 , , 528 , a30 crosthwaite , l. p. , turner , j. l. , hurt , r. l. , et al .
2001 , , 122 , 797 dayou , f. , & balana , c. 2006 , , 459 , 297 de vicente , p. , martn - pintado , j. , neri , r. , & colom , p. 2000 , , 361 , 1058 dickinson , a. s. , phillips , t. g. , goldsmith , p. f. , percival , i. c. , & richards , d. 1977 , , 54 , 645 drdla , k. , knapp , g. r. , & van dishoeck , e. f. 1989 , , 345 , 815 everitt , b. s. , & dunn , g. 2001 , applied multivariate data analysis ( london : oxford univ .
press ) , 48 fingerhut , r. l. , lee , h. , mccall , m. l. , & richer , m. g. 2007 , , 655 , 814 flower , d. r. 1999 , , 305 , 651 gao , y. & solomon , p. m. 2004 , , 606 , 271 garcia - burillo , s. , martn - pintado , j. , fuente , a. & neri , r. 2000 , , 355 , 499 garca - burillo , s. , martn - pintado , j. , fuente , a. , & neri , r. 2001 , , 563 , l27 graci - carpio , j. , garca - burillo , s. , planesas , p. , & colina , l. 2006 , , 640 , l135 graci - carpio , j. , garca - burillo , s. , planesas , p. , fuente , a. , & usero , a. 2008 , , 479 , 703 greaves , j. s. , ohishi , m. , & nyman , l .- a .
1996 , , 307 , 898 green , s. 1975 , , 201 , 366 green , s. 1986 , nasa tech . memo . , nasa tm 87791 green , s. & chapman , s. 1978 , , 37 , 169 green , s. & thaddeus , p. 1974
, , 191 , 653 hasegawa , t. i. , & herbst , e. 1993 , mnras , 263 , 589 helfer , t. t. & blitz , l. 1993 , , 419 , 86 henkel , c. , baan , w. a. & mauersberger , r. 1991 , a&ar , 3 47 heyer , m. h. , & schloerb , f. p. 1997
, , 475 , 173 hirota , t. , yamamoto , s. , mikami , h. , & ohishi , m. 1998 , apj , 503 , 717 huggins , p. j. , carlson , w. j. , & kinney , a. l. 1984 , , 133 , 347 hurt , r. l. , merrill , k. m. , gatley , i. , & turner , j. l. 1993 , , 105 , 121 hurt , r. l. , & turner , j. l. 1991 , , 377 , 434 hurt , r. l. , turner , j. l. , & ho , p. t. p. 1996
, , 466 , 135 httemeister , s. , mauersberger , r. & henkel , c. 1997 , , 326 , 59 imanishi , m. , nakanishi , k. , tamura , y. , oi , n. , & kohno , k. 2007 , , 134 , 2366 ishiguro , m. , kawabe , r. , morita , k .-
, et al . 1989 , , 344 , 763 juneau , s. , narayanan , d. t. , moustakas , j. , shirely , y. l. , bussman , r. s. , kennicutt , r. c. , jr .
, & vanden bout , p. a. 2009 , , 707 , 1217 kalenskii , s. v. , dzura , a. m. , booth , r. s. , winnberg , a. , & alakoz , a. v. 1997 , , 321 , 311 kalenskii , s. v. & sobolev , a. m. 1994 , astl , 20 , 91 keane , j. v. , tielens , a. g. g. m. , boogert , a. c. a. , schutte , w. a. , & whittet , d. c. b. 2001 , , 376 , 254 kim , s. s. , saitoh , t. r. , jeon , m. , figer , d. , merritt , d. & wada , k. 2011 , , 735 , l11 kohno , k. , matsushita , s. , vila - vilar , b. , okumura , s. k. , shibatsuka , t. , okiura , m , ishizuki , s. kawabe , r. 2001 , the central kiloparsec of starbursts and agn : the la palma connection , 249 , 672 krips , m. , neri , r. , garca - burillo , s. , martn , s. , combes , f. , graci - carpio , j. & eckart , a. 2008 , , 677 , 262 krzanowski , w. j. 1988 , principles of multivariate analysis ( oxford : clarendon ) kuno , n. , nakanishi , k. , sorai , k. , & shibatsuka , t. 2008 , , 60 , 475 lapinov , a. v. , schilke , p. , juvela , m. , & zinchenko , i. i. 1998 , , 336 , 1007 lindberg , j. e. , aalto , s. , costagliola , f. , prez - beaupuits , j - p .
, monje , r. & muller , s. 2011 , , 527 , a150 lucas , r. , & liszt , h. 1996 , , 307 , 237 lucas , r. & liszt , h. s. 2000 , , 358 , 1069 lucas , r. , & liszt , h. s. 2002 , , 384 , 1054 martn - pintado , j. , bachiller , r. & fuente , a. 1992 , a&a , 254 , 315 martn , s. , martn - pintado , j. , & mauersberger , r. 2009 , , 694 , 610 martn , s. , mauersberger , r. , martn - pintado , j. , henkel , c. , & garca - burillo , s. 2006 , , 164 , 450 martn , s. , requena - torres , m. a. , martn - pintado , j. , & mauersberger , r. 2008 , , 678 , 245 meier , d. s. & turner , j. l. 2001 , , 551 , 687 meier , d. s. & turner , j. l. 2005 , , 618 , 259 meier , d. s. , turner , j. l. & hurt , r. l. 2000 , , 531 , 200 meier , d. s. , turner , j. l. , & hurt , r. l. 2008 , , 675 , 281 meier , d. s. , turner , j. l. , & schinnerer , e. 2011 , , 142 , 32 meijerink , r. , spaans , m. , & israel , f. p. 2007
, , 461 , 793 millar , t. j. , herbst , e. & charnley , s. b. 1991 , , 369 , 147 minier , v. & booth , r. s. 2002 , , 387 , 179 morata , o. , estalella , r. , lopez , r. , & planesas , p. 1997
, , 292 , 120 morris , m. , turner , b. e. , palmer , p. , & zuckerman , b. 1976 , , 205 , 82 murtagh , f. , & heck , a. 1987 , multivariate data analysis ( dordrecht : d. reidel ) nguyen - q - rieu , jackson , j. m. , henkel , c. , truong , b. & mauersberger , r. 1992 , , 399 , 521 nummelin , a. , bergman , p. , hjalmarson , . , friberg , p. , irvine , w. m. , millar , t. j. , ohishi , m. , & saito , s. 2000 , , 128 , 213 nyman , l .- a .
1984 , , 141 , 323 padin , s. , scott , s. l. , woody , d. p. , scoville , n. z. , seling , t. v. , finch , r. p. , ciovanine , c. j. , & lowrance , r. p. 1991
, pasp , 103 , 461 papadopoulos , p. p.
2007 , , 656 , 792 pickett , h. m. , r. l. poynter , e. a. cohen , m. l. delitsky , j. c. pearson , & h. s. p. muller , 1998 , `` submillimeter , millimeter , and microwave spectral line catalog , '' j. quant .
spectrosc . &
rad . transfer 60 , 883 - 890 .
piner , b. g. , stone , j. m. , & teuben , p. j. 1995 , , 449 , 508 pirogov , l. , zinchencko , i. , lapinov , a. , myshenko , v. & shulga , v. 1995 , , 109 , 333 pottage , j. t. , flower , d. r. , & davis , s. l. 2004 , , 352 , 39 regan , m. w. , sheth , k. , & vogel , s. n. 1999 , , 526 , 97 requena - torres , m. a. , martn - pintado , j. , rodrguez - franco , a. , martn , s. , rodrguez - fernndez . n. j. & de vincente , p. 2006 , , 455 , 971 roberge , w. g. , jones , d. , lepp , s. , & dalgarno , a. 1991 , , 77 , 287 rodriguez - fernandez , n. j. , combes , f. , martn - pintado , j. , wilson , t. l. , & apponi , a. 2006 , , 455 , 963 rodrguez - fernndez , n. j. , tafalla , m. , gueth , f. , & bachiller , r. 2010 , , 516 , a98 schilke , p. , walmsley , c. m. , pineau des forets , g. , roueff , e. , flower , d. r. , & guilloteau , s. 1992 , , 256 , 595 schinnerer , e. , bker , t. , meier , d. s. , & calzetti , d. 2008 , , 684 , l21 schutte , w. a. & greenberg , j. m. 1997 , , 317 , l43 scoville , n. z. , carlstrom , j. , padin , s. , sargent , a. , scott , s. & woody , d. 1994 , astronomy with millimeter and submillimeter wave interferometry , iau colloquium 140 , asp conference series , vol .
59 , 1994 , m. ishiguro and j. welch , eds . , p.10 snell , r. l. , narayanan , g. , yun , m. s. , heyer , m. , chung , a. , irvine , w. m. , erickson , n. r. & liu , g. 2011 , , 141 , 38 snyder , l. e. , & buhl , d. 1975 , , 197 , 329 solomon , p. m. , downes , d. , & radford , s. j. e. 1992 , , 387 , l55 sternberg , a. & dalgarno , a. 1995 , , 99 , 565 tideswell , d. m. , fuller , g. a. , millar , t. j. , & markwick , a. j. 2010 , , 510 , a85 tsai , c .- w . , turner , j. l. , beck , s. c. , crosthwaite , l .p . , ho , p. t. p. & meier , d. s. 2006 , , 132 , 2383 turner , b. e. 1991 , , 76 , 617 turner , b. e. 1995a , , 444 , 708 turner , b. e. 1995b , , 449 , 635 turner , b. e. 1998a , , 501 , 731 turner , b. e. 1998b , , 495 , 804 turner , b. e. , pirogov , l. , & minh , y. c. 1997 , , 483 , 235 turner , b. e. , terzieva , r. , & herbst , e. 1999 , apj , 518 , 694 turner , j. l. , & ho , p. t. p. 1983
, , 268 , l79 turner , j. l. , & ho , p. t. p. 1994
, , 421 , 122 turner , j. l. , & hurt , r. l. 1992 , , 384 , 72 usero , a. , garca - burillo , s. , fuente , a. , martn - pintado , j. , & rodrguez - fernndez , n. j. 2004 , , 419 , 897 usero , a. , garca - burillo , s. , martn - pintado , j. , fuente , a. , & neri , r. 2006 , , 448 , 457 vanden bout , p. a. , loren , r. b. , snell , r. l. , & wootten , a. 1983 , , 271 , 161 van dishoeck , e. f. , & blake , g. a. 1998 , , 36 , 317 wall , j. v. & jenkins , c. r. 2003 , in `` practical statistics for astronomers '' ( cambridge univ . press : cambridge ) , chp .
4.5 walter , f. , weiss , a. & scoville , n. 2002 , , 580 , l21 wang , y. , jaffe , d. t. , evans , n. j. , hayashi , m. , tatematsu , k. , & zhou , s. 1993 , , 419 , 707 watanabe , n. , & kouchi , a. 2002 , , 571 , l173 watt , g. d. 1983 , , 205 , 321 welch , w. j. et al .
1996 , , 108 , 93 womack , m. , ziurys , l. m. & wyckoff , s. 1992b , , 387 , 417 woodall , j. , agndez , m. , markwick - kemper , a. j. , & millar , t. j. 2007 , , 466 , 1197 wootten , a. , bozyan , e. p. , garrett , d. b. , loren , r. b. & snell , r. l. 1980 , , 239 , 844 wootten , a. , evans , n. j. , snell , r. , & vanden bout , p. 1978
, , 225 , l143 yamada , m. , wada , k. , & tomisaka , k. 2007 , , 671 , 73 zinchenko , i , henkel , c. , & mao , r. q. 2000 , , 361 , 1079 ziurys , l. m. , friberg , p. & irvine , w. m. 1989 , , 343 , 201 lccccccc sio(2 1 ; @xmath122)&3.098 & /21.712/ & 2 & 1787 & / & & + sio(2 1 ; @xmath13)&3.098 & /21.712/ & 2 & 6.25 & 5.27/5.66 & 1.86(12)&2.94 + c@xmath3h(1 0 ; @xmath47 ) & 0.8&/43.675/&1 & 4.19 & 5.13/6.24 & 2.32(13)&3.01 + c@xmath3h(1 0 ; @xmath48 ) & 0.8&/43.675/&1 & 4.19 & 5.13/6.24 & 4.64(13)&3.01 + hco@xmath123(1 0 ) & 3.888&/44.594/&1 & 4.28 & 5.15/5.80 & 9.51(11)&3.36 + hnc(1 0 ) & 3.05&/45.332/&1 & 4.35 & 6.39/5.87 & 1.51(12)&2.94 + hc@xmath2n(10 9)&3.72 & /4.5491/&10 & 24.02 & 5.86/6.87 & 6.73(12)&0.723 + n@xmath3h@xmath1(1 0)&3.40 & /46.587/&1 & 4.47 & 5.63/5.85&1.16(12 ) & 3.00 + c@xmath15s(2 1)&1.96 & /24.104/&2 & 6.94&5.65/6.17&4.06(12 ) & 2.56 + ch@xmath2oh(2@xmath124 @xmath125)&0.89 & 127.484/24.680/23.770 & 4,3,4 & 6.98 & 4.83/6.39&1.58(13 ) & 5.96 + hnco(@xmath126 @xmath127)&1.60 & 912.711/11.071/10.911 & 5 & 15.83 & 6.29/5.99 & 9.27(12 ) & 3.15 + lcccccccc bima : & & & & & & & & + sio(21;v=0 ) & 0902 - 0303 & b , c , d & 86.847 & 140 - 1000 & 21.57&@xmath128;83 & 1.68 & 10 + c@xmath3h(10;3/21/2 ) & 0902 - 0303 & b , c , d & 87.317 & 140 - 1000 & 21.46 & @xmath129;-84 & 2.75 & 11 + hnc(10 ) & 0902 - 0303 & b , c , d & 90.664 & 140 - 1000 & 20.67 & @xmath130;81 & 2.59 & 11 + hc@xmath2n(109 ) & 0902 - 0303 & b , c , d & 90.979 & 140 - 1000 & 20.60 & @xmath131;83 & 1.77 & 7.2 + n@xmath3h@xmath1(10 ) & 0902 - 0303 & b , c , d & 93.174 & 130 - 1100 & 20.12 & @xmath132;-85 & 2.53 & 9.0 + c@xmath15s(21 ) & 0902 - 0303 & b , c , d & 96.413 & 130 - 1100 & 19.43 & @xmath133;-86 & 2.51 & 11 + ch@xmath2oh(@xmath134@xmath125 ) & 0902 - 0303 & b , c , d & 96.741 & 130 - 1100 & 19.37 & @xmath133;-86 & 2.48 & 15 + ovro : & & & & & & & & + sio(21;v=1 ) & 0199 - 0399 & l , h , uh &
86.243 & 300 - 410 & 13.90 & @xmath135;-56 & 36.5 & 9.5 + hco@xmath1(10 ) & 0199 - 0399 & l , h , uh & 89.189 & 300 - 410 & 13.44&@xmath136;-24 & 30.5 & 6.7 + hnco(@xmath126@xmath127 ) & 1098 - 0199 & l , h & 109.906 & 240 - 430 & 10.92 & @xmath137;-84 & 17.3 & 6.3 + lccccc a & sio(10;v=1 ) & @xmath13819 & @xmath13888 & & + ( 02:41:57.17 ) & sio(10;v=0)&@xmath1381.0 & @xmath13823 & & + ( 59:36:42.4 ) & c@xmath3h(@xmath11@xmath12)&@xmath1383.3 & @xmath13834 & & + & c@xmath3h(@xmath12@xmath12)&@xmath1383.3 & @xmath13834 & & + & hco@xmath1(10 ) & 16@xmath837 & 270@xmath8338 & -34@xmath835 & 21@xmath8310 + & hnc(10 ) & @xmath1382.9 & 45@xmath8316 & -100@xmath8311 & 120@xmath8325 + & hc@xmath2n(109 ) & @xmath1381.0 & @xmath13826 & & + & n@xmath3h@xmath1(10 ) & @xmath1381.9 & @xmath13830 & & + & c@xmath15s(21 ) & @xmath1382.2 & @xmath13828 & & + & ch@xmath2oh(@xmath134@xmath125 ) & @xmath1383.2 & 46@xmath8319 & -100@xmath836 & 45@xmath8315 + & hnco(@xmath126@xmath127 ) & & & & + b & sio(10;v=1 ) & @xmath13819 & @xmath13888 & & + ( 02:41:55.75 ) & sio(10;v=0)&@xmath1381.0 & 35@xmath8312 & -67@xmath8313 & 55@xmath8330 + ( 59:36:29.6 ) & c@xmath3h(@xmath11@xmath12)&@xmath1383.3 & @xmath13834 & & + & c@xmath3h(@xmath12@xmath12)&@xmath1383.3 & @xmath13834 & & + & hco@xmath1(10 ) & 61@xmath837 & 470@xmath8338 & -84@xmath833 & 68@xmath837 + & hnc(10 ) & 9.1@xmath831 & 150@xmath8316 & -134@xmath833 & 65@xmath837 + & hc@xmath2n(109 ) & @xmath1381.0 & @xmath13826 & & + & n@xmath3h@xmath1(10 ) & 5.1@xmath831 & 60@xmath8315 & -98@xmath836 & 50@xmath8310 + & ch@xmath2oh(@xmath134@xmath125 ) & 10@xmath832 & 53@xmath8319 & -87@xmath838 & 48@xmath8320 + & c@xmath15s(21 ) & @xmath1382.2 & @xmath13828 & & + & hnco(@xmath126@xmath127 ) & @xmath1394.8 & 50@xmath8329 & -112@xmath8320 & 110@xmath8350 + c & sio(10;v=1 ) & @xmath13819 & @xmath13888 & & + ( 02:41:55.48 ) & sio(10;v=0)&@xmath1381.0 & 24@xmath8312 & -80@xmath8320 & 83@xmath8347 + ( 59:36:24.8 ) & c@xmath3h(@xmath11@xmath12)&11@xmath832 & 100@xmath8317 & -88@xmath836 & 93@xmath8315 + & c@xmath3h(@xmath12@xmath12)&3.9@xmath832 & 48@xmath8317 & -76@xmath8317 & 90@xmath8340 + & hco@xmath1(10 ) & 82@xmath837 & 520@xmath8338 & -78@xmath833 & 83@xmath836 + & hnc(10 ) & 24@xmath831 & 200@xmath8316 & -120@xmath834 & 120@xmath838 + & hc@xmath2n(109 ) & [email protected] & 42@xmath8313 & -88@xmath837 & 65@xmath8316 + & n@xmath3h@xmath1(10 ) & 9.6@xmath831 & 110@xmath8315 & -88@xmath835 & 77@xmath8310 + & ch@xmath2oh(@xmath134@xmath125 ) & 5.0@xmath832 & 100@xmath8319 & -99@xmath834 & 48@xmath838 + & c@xmath15s(21 ) & @xmath1382.2 & @xmath13828 & & + & hnco(@xmath126@xmath127 ) & @xmath1394.8 & 130@xmath8329 & -89@xmath835 & 42@xmath8310 + d & sio(10;v=1 ) & @xmath13819 & @xmath13888 & & + ( 02:41:55.14 ) & sio(10;v=0)&[email protected] & @xmath13823 & & + ( 59:36:20.7 ) & c@xmath3h(@xmath11@xmath12)&17@xmath832 & 110@xmath8317 & -82@xmath839 & 130@xmath8322 + & c@xmath3h(@xmath12@xmath12)&11@xmath832 & 80@xmath8317 & -77@xmath8317 & 120@xmath8335 + & hco@xmath1(10 ) & 180@xmath837 & 580@xmath8338 & -74@xmath833 & 110@xmath837 + & hnc(10 ) & 41@xmath831 & 250@xmath8316 & -109@xmath834 & 140@xmath838 + & hc@xmath2n(109 ) & [email protected] & 50@xmath8313 & -80@xmath836 & 67@xmath8315 + & n@xmath3h@xmath1(10 ) & 9.4@xmath831 & 110@xmath8315 & -87@xmath834 & 82@xmath838 + & ch@xmath2oh(@xmath134@xmath125 ) & 5.5@xmath832 & 95@xmath8319 & -101@xmath835 & 57@xmath8312 + & c@xmath15s(21 ) & @xmath1382.2 & @xmath13828 & & + & hnco(@xmath126@xmath127 ) & @xmath1394.8 & 86@xmath8329 & -90@xmath8311 & 82@xmath8315 + e & sio(10;v=1 ) & @xmath13819 & @xmath13888 & & + ( 02:41:55.08 ) & sio(10;v=0 ) & @xmath40.77 & 22@xmath8312 & -2.0@xmath8326 & 220@xmath8360 + ( 59:36:18.4 ) & c@xmath3h(@xmath11@xmath12)&19@xmath832 & 97@xmath8317 & -73@xmath8312 & 160@xmath8328 + & c@xmath3h(@xmath12@xmath12)&13@xmath832 & 77@xmath8317 & -72@xmath8316 & 130@xmath8338 + & hco@xmath1(10 ) & 130@xmath837 & 520@xmath8338 & -73@xmath834 & 130@xmath838 + & hnc(10 ) & 44@xmath831 & 220@xmath8316 & -104@xmath834 & 150@xmath8310 + & hc@xmath2n(109 ) & [email protected] & 39@xmath8313 & -80@xmath838 & 67@xmath8318 + & n@xmath3h@xmath1(10 ) & 6.6@xmath831 & 85@xmath8315 & -83@xmath835 & 92@xmath8312 + & ch@xmath2oh(@xmath134@xmath125 ) & 6.7@xmath832 & 70@xmath8319 & -94@xmath838 & 75@xmath8318 + & c@xmath15s(21 ) & @xmath1382.2 & @xmath13828 & & + & hnco(@xmath126@xmath127 ) & @xmath1384.8 & @xmath13858 & & + f & sio(10;v=1 ) & @xmath13819 & @xmath13888 & & + ( 02:41:54.86 ) & sio(10;v=0 ) & @xmath1381.0 & 35@xmath8312 & 17@xmath838 & 50@xmath8318 + ( 59:36:09.6 ) & c@xmath3h(@xmath11@xmath12)&9.4@xmath832 & 63@xmath8317 & 10@xmath8312 & 130@xmath8350 + & c@xmath3h(@xmath12@xmath12)&@xmath1383.3 & @xmath446 & & + & hco@xmath1(10 ) & 85@xmath837 & 450@xmath8338 & -6.9@xmath834 & 83@xmath838 + & hnc(10 ) & 36@xmath831 & 250@xmath8316 & -0.9@xmath833 & 67@xmath8317 + & hc@xmath2n(109 ) & [email protected] & 31@xmath8313 & 0.1@xmath8313 & 130@xmath8330 + & n@xmath3h@xmath1(10 ) & 15@xmath831 & 150@xmath8315 & 5.4@xmath836 & 70@xmath8310 + & ch@xmath2oh(@xmath134@xmath125 ) & 15@xmath832 & 190@xmath8319 & -43@xmath833 & 68@xmath835 + & c@xmath15s(21 ) & 5.0@xmath831 & @xmath430 & -46@xmath8326 & 180@xmath8362 + & hnco(@xmath126@xmath127 ) & 22@xmath832 & 190@xmath8329 & -1.6@xmath833 & 40@xmath837 + lccccc g & sio(10;v=1 ) & @xmath13819 & @xmath13888 & & + ( 02:41:54.12 ) & sio(10;v=0 ) & [email protected] & 52@xmath8317 & 28@xmath834 & 32@xmath8313 + ( 59:35:59.6 ) & c@xmath3h(@xmath11@xmath12)&6.1@xmath832 & 110@xmath8317 & 12@xmath834 & 55@xmath8310 + & c@xmath3h(@xmath12@xmath12)&4.1@xmath832 & 51@xmath8317 & 27@xmath8312 & 72@xmath83 28 + & hco@xmath1(10 ) & @xmath415@xmath837 & 200@xmath8338 & 15@xmath837 & 62@xmath8315 + & hnc(10 ) & 17@xmath831 & 200@xmath8316 & 20@xmath832 & 55@xmath835 + & hc@xmath2n(109 ) & @xmath1381.5 & @xmath13826 & & + & n@xmath3h@xmath1(10 ) & 10@xmath831 & 140@xmath8315 & 23@xmath833 & 57@xmath837 + & ch@xmath2oh(@xmath134@xmath125 ) & 25@xmath832 & 300@xmath8319 & -27@xmath832 & 55@xmath833 + & c@xmath15s(21 ) & 3.3@xmath832 & 47@xmath8319 & 40@xmath836 & 50@xmath8315 + & hnco(@xmath126@xmath127 ) & 33@xmath832 & 320@xmath8329 & 21@xmath832 & 43@xmath833 + h & sio(10;v=1 ) & @xmath13819 & @xmath13888 & & + ( 02:41:53.76 ) & sio(10;v=0 ) & @xmath1381.0 & @xmath13823 & & + ( 59:36:04.7 ) & c@xmath3h(@xmath11@xmath12)&@xmath1383.3 & 60@xmath8317 & 5.8@xmath839 & 78@xmath8322 + & c@xmath3h(@xmath12@xmath12)&@xmath1383.3 & @xmath13834 & & + & hco@xmath1(10 ) & @xmath13815 & 150@xmath8338 & 6.2@xmath836 & 35@xmath8313 + & hnc(10 ) & 7.8@xmath831 & 130@xmath8316 & 10@xmath835 & 58@xmath8317 + & hc@xmath2n(109 ) & @xmath1381.0 & @xmath13826 & & + & n@xmath3h@xmath1(10 ) & @xmath1381.9 & 73@xmath8315 & 20@xmath836 & 57@xmath8310 + & ch@xmath2oh(@xmath134@xmath125 ) & 8.4@xmath832 & 200@xmath8319 & -32@xmath832 & 60@xmath835 + & c@xmath15s(21 ) & @xmath1382.2 & @xmath13828 & & + & hnco(@xmath126@xmath127 ) & @xmath1394.8 & 110@xmath8329 & 11@xmath834 & 32@xmath8310 + lccccccccccc a & 2.8(21 ) & @xmath1386.8(-10)&@xmath1385.4(-8 ) & & 5.4(-9 ) & @xmath1381.6(-9 ) & @xmath1382.4(-9 ) & @xmath1387.9(-10 ) & @xmath1383.2(-9 ) & @xmath1381.8(-8 ) & + b & 9.7(21 ) & @xmath1382.0(-10 ) & @xmath1381.5(-8 ) & & 6.1(-9 ) & 1.4(-9 ) & @xmath1386.9(-10 ) & 6.1(-10 ) & @xmath1389.2(-10 ) & 1.6(-8 ) & @xmath1381.6(-8 ) + c & 1.9(22 ) & @xmath1381.0(-10 ) & 1.8(-8 ) & 1.8(-8 ) & 4.2(-9 ) & 1.9(-9 ) & 1.0(-9 ) & 5.8(-10 ) & @xmath1384.9(-10 ) & 4.2(-9 ) & @xmath1382.4(-8 ) + d & 2.9(22 ) & 1.6(-10 ) & 2.2(-8 ) & 4.2(-8 ) & 6.0(-9 ) & 2.1(-9 ) & 7.6(-10 ) & 3.8(-10 ) & @xmath1383.0(-11 ) & 3.0(-9 ) & @xmath41.1(-9 ) + e & 3.0(22 ) & @xmath44.7(-11 ) & 2.5(-8 ) & 4.8(-8 ) & 4.1(-9 ) & 2.2(-9 ) & 4.7(-10 ) & 2.6(-10 ) & @xmath1383.0(-10 ) & 3.7(-9 ) & @xmath1381.5(-9 ) + f & 2.3(22 ) & @xmath1398.3(-11 ) & 9.6(-9)&9.6(-9 ) & 3.5(-9 ) & 2.3(-9 ) & 1.2(-9 ) & 7.4(-10 ) & 8.7(-10 ) & 1.0(-8 ) & 8.7(-9 ) + g & 2.0(22 ) & @xmath41.1(-10 ) & 1.2(-8 ) & 2.3(-8 ) & 7.2(-10 ) & 1.3(-9 ) & @xmath1385.0(-10 ) & 6.0(-10 ) & @xmath46.5(-10 ) & 2.0(-8 ) & 1.6(-8 ) + h & 8.9(21 ) & @xmath1382.1(-10 ) & @xmath1381.7(8 ) & @xmath1382.0(-9 ) & @xmath1381.6(-9 ) & 1.3(-9 ) & @xmath1387.5(-10 ) & @xmath1382.5(-10 ) & @xmath1381.0(-9 ) & 1.5(-8 ) & @xmath1385.1(-9 ) + lcccc hco & @xmath140 @xmath141 ; @xmath142 ; 1 1 & 86.777 & @xmath13820 & + hco & @xmath140 @xmath141 ; @xmath142 ; 0 1 & 86.806 & @xmath13820 & + hn@xmath8c & 1 0 & 87.091 & @xmath13822 & + hcc@xmath8cn / hc@xmath8c@xmath3n & 10 9 & 90.593/90.602 & @xmath13822 & + + & & 89.222(5 ) & 70@xmath8315 & c + & & 90.912(5 ) & 10@xmath836 & f + lcccccc a & & & & @xmath660.79 & @xmath661.4 & @xmath1391.6 + b & [email protected] & & [email protected] & @xmath664.5 & @xmath661.4 & [email protected] + c & [email protected] & [email protected] & [email protected] & @xmath666.6 & @xmath666.0 & [email protected] + d & [email protected] & [email protected] & [email protected] & @xmath6610 & @xmath6614 & [email protected] + e & [email protected] & [email protected] & [email protected] & @xmath669.4&@xmath6613 & [email protected] + f & [email protected] & [email protected] & [email protected] & @xmath6616&@xmath44.1 & [email protected] + g & [email protected] & [email protected] & @xmath42.2 & @xmath6611&@xmath43.8 & [email protected] + h & [email protected] & & & @xmath664.4&@xmath661.0 & @xmath1382.4 + lcccccccccccccc @xmath8co&1.00 & & & & & & & & & & & & + 3mm&0.88 & 1.00 & & & & & & & & & & & + c@xmath3h(@xmath11 ) & 0.85 & 0.77 & 1.00 & & & & & & & & & & + hcn & 0.93 & 0.92 & 0.84 & 1.00 & & & & & & & & + hco@xmath1 & 0.89 & 0.88 & 0.81 & 0.98 & 1.00 & & & & & & & & + hnc & 0.95 & 0.87 & 0.84 & 0.97 & 0.94 & 1.00 & & & & & & & + ch@xmath2oh & 0.68 & 0.47 & 0.49 & 0.43 & 0.34 & 0.51 & 1.00 & & & & & & + hnco & 0.52 & 0.27 & 0.32 & 0.21 & 0.12 & 0.33 & 0.86 & 1.00 & & & & & + n@xmath3h@xmath1 & 0.86 & 0.71 & 0.67 & 0.77 & 0.72 & 0.82 & 0.71 & 0.58 & 1.00 & & & & + hc@xmath2n & 0.54 & 0.53 & 0.41 & 0.62 & 0.60 & 0.62 & 0.32 & 0.16 & 0.63 & 1.00 & & & + c@xmath15s & 0.20 & 0.24 & 0.17 & 0.18 & 0.20 & 0.19 & 0.14 & 0.095 & 0.20 & 0.15 & 1.00 & & + sio & 0.38 & 0.23 & 0.34 & 0.34 & 0.30 & 0.38 & 0.35 & 0.30 & 0.35 & 0.31&-0.023 & 1.00 & + c@xmath3h(@xmath12 ) & 0.64 & 0.60 & 0.70 & 0.66 & 0.63 & 0.63 & 0.27 & 0.18 & 0.45 & 0.24 & 0.008 & 0.32 & 1.00 + lccccccc @xmath8co & 0.35 & -0.045 & -0.0043 & 0.11 & -0.042 & 0.15 & -0.064 + 3mm&0.32 & 0.15 & -0.10 & 0.12 & -0.086 & 0.19 & -0.50 + c@xmath3h(@xmath11 ) & 0.31 & 0.097 & 0.086 & 0.24 & 0.12 & 0.020 & 0.81 + hcn & 0.34 & 0.21 & 0.00 & -0.012 & -0.057 & 0.16 & -0.086 + hco@xmath1 & 0.33 & 0.28 & -0.029 & -0.024 & -0.055 & 0.21 & -0.038 + hnc & 0.35 & 0.11 & 0.0015 & -0.031 & -0.054 & 0.16 & 0.038 + ch@xmath2oh & 0.23 & -0.55 & -0.03 & 0.10 & -0.091 & 0.0061 & -0.031 + hnco & 0.17 & -0.65 & -0.01 & 0.17 & -0.050 & -0.10 & -0.061 + n@xmath3h@xmath1 & 0.32 & -0.19 & -0.089 & -0.12 & -0.20 & -0.061 & 0.070 + hc@xmath2n & 0.23 & 0.067 & -0.13 & -0.68 & -0.29 & -0.52 & 0.086 + c@xmath15s & 0.080 & -0.0027 & -0.81 & 0.0092 & 0.56 & -0.11 & -0.0031 + sio & 0.16 & -0.19 & 0.43 & -0.51 & 0.66 & 0.23 & -0.074 + c@xmath3h(@xmath12 ) & 0.24 & 0.19 & 0.33 & 0.37 & 0.29 & -0.70 & -0.23 + egnv .
% & 59&13&8.3&6.6&5.6&2.7&1.5 + | we present 2@xmath0 - 10@xmath0 imaging of eleven transitions from nine molecular species across the nuclear bar in maffei 2 .
the data were obtained with the bima and ovro interferometers .
the ten detected transitions are compared with existing co isotopologues , hcn , cs and millimeter continuum data .
dramatic spatial variations among the mapped species are observed across the nuclear bar .
a principle component analysis is performed to characterize correlations between the transitions , star formation and molecular column density .
the analysis reveals that hcn , hnc , hco@xmath1 and 3 mm continuum are tightly correlated , indicating a direct connection to massive star formation .
we find two main morphologically distinct chemical groups , ch@xmath2oh , sio and hnco comprising the grain chemistry molecules , versus hcn , hnc , hco@xmath1 and c@xmath3h , molecules strong in the presence of star formation .
the grain chemistry molecules , hnco , ch@xmath2oh and sio , trace hydrodynamical bar shocks .
the near constancy of the hnco / ch@xmath2oh , sio / ch@xmath2oh and sio / hnco ratios argue that shock properties are uniform across the nucleus .
hcn / hco@xmath1 , hcn / hnc , hcn / cs and hcn / co ratios are explained primarily by variations in density .
high hco@xmath1/n@xmath3h@xmath1 ratios are correlated with the c@xmath3h line , suggesting that this ratio may be a powerful new dense photon - dominated region ( pdr ) probe in external galaxies .
c@xmath3h reveals a molecular outflow along the minor axis .
the morphology and kinematics of the outflow are consistent with an outflow age of 6 - 7 myrs . |
angular correlations among jets produced in association with scalar bosons and massive gravitons have been the subject of several studies @xcite .
it has been found that experimental determination of such correlations can give us important information about the spin and cp - properties of these particles .
for example , by studying the distribution of azimuthal angle difference between two tagging jets ( @xmath12 ) in the gluon fusion ( gf ) production of a scalar ( @xmath13 ) or pseudo - scalar ( @xmath14 ) higgs with two jets , it has been shown that the cp - odd and cp - even higgs can clearly be discriminated at the lhc .
it has also been shown that for jets produced with massive gravitons ( g ) , the @xmath12 distribution is flat while the sum of the azimuthal angles show a characteristic distribution @xcite .
it was further demonstrated analytically in ref .
@xcite that these angular correlations arise from the quantum interference of different helicity states of the intermediate vector bosons , which can either be a weak boson or a gluon .
another critical observation made in ref .
@xcite is that the amplitudes for @xmath15 jets production can be approximated by those of the vector boson fusion ( vbf ) sub - processes when a specific set of kinematic cuts are applied , where the two tagging jets are demanded to be in the opposite hemispheres of the detector and also to be well separated in rapidity ( the so called vbf cuts ) , and in addition their transverse momenta are subjected to a slicing cut , restricting them to be sufficiently low as compared to the h / a / g masses .
we note in passing that another complementary method of determining the spin and cp information of new particles is from their decay patterns ( for example , the decay of a higgs boson to a pair of gauge bosons ) , which has also been widely discussed in the literature @xcite . in this paper
, we study heavy quark pair @xmath16 ( @xmath17 ) production at the lhc in association with two jets , and study the azimuthal angle correlations among these jets .
our goal is to predict the jet angular correlations for the sm processes by using the same approximation , so that the experimental technique to measure the new particle properties via initial state radiation patterns can be established using the ample sm processes .
we first calculate the helicity amplitudes for gluon fusion processes initiated by @xmath18 initial states at the lhc leading to @xmath19 jets production . here
, we work in the limit in which the intermediate gluons are collinear to the initial partons .
we work out the angular correlations using the formalism of ref .
@xcite , and then verify our approximate analytical results in several kinematic regions by comparing them with the exact parton level matrix elements in the tree level .
this paper is organized as follows . in section 2
we briefly review the helicity formalism used to obtain our analytical results . in section 3
we give the spin summed matrix - element squared for @xmath20 jets production via gluon fusion in the collinear limit for the exchanged gluons , and then analyze the results in two different kinematic limits , namely that of the @xmath3 production threshold ( @xmath21 ) in which the invariant mass of the @xmath22 pair , @xmath23 , is very close to @xmath24 , and that of the relativistic limit where @xmath25 is much higher than @xmath26 , for @xmath27 and @xmath28 . in section 4
we show the exact numerical results for these angular correlations at parton level , where the expected azimuthal angle correlations are reproduced once appropriate final state selection cuts are applied .
finally , we summarize our findings in section 5 .
the helicity amplitude formalism for vector boson fusion processes has been discussed in detail in ref .
@xcite in the context of production and decay of a heavy particle in association with two jets .
we follow the same notations and conventions in this paper . the inclusive process , @xmath29 can proceed , via gluon fusion , through the sub - processes @xmath30 , where @xmath31 is one of the three possible initial states @xmath32 , @xmath33 is a @xmath28-channel intermediate gluon and @xmath34 stands for a quark or anti - quark of any flavour : [ eq1 ] @xmath35 in addition to gluon fusion , the above sub - processes receive contributions from all other diagrams at the same order in perturbation theory to make a gauge - invariant physical amplitude .
however , in this and the next section , we shall consider only the gluon fusion diagrams .
as mentioned earlier , it has been demonstrated in ref .
@xcite that after applying the vbf selection cuts , the gf contribution can be made to dominate the total cross - section .
we shall also show the validity of this approach by first determining analytically the distributions predicted by the gf diagrams in various kinematic limits , and then comparing them with the exact matrix elements . to begin with ,
we define a common set of kinematic variables for the @xmath36 sub - processes in eqn .
[ eq1 ] as @xmath37 the helicities of the quarks or anti - quarks have the value @xmath38 while for on - shell gluons they take the values @xmath39 .
the helicity amplitudes for this process can be written as @xmath40 where the external quark or gluon currents are denoted by @xmath41 , and the gluon propagator is given by @xmath42 for the conserved currents @xmath43 we note in passing that the conditions in eqn .
[ eq : curr - consv ] are satisfied not only for massless quark currents but also for the gluonic currents , in the processes ( 2b ) and ( 2c ) in the light - cone gauge @xcite .
the real gluon approximation to the amplitudes ( eqn .
[ eq : ampl ] ) is obtained by replacing the off - shell @xmath44 amplitudes by the on - shell , and hence gauge - invariant @xmath45 amplitudes @xmath46 while keeping the four - momenta of the @xmath16 system and the orientation of the colliding virtual gluon momenta in the @xmath47 rest - frame . the full amplitudes for the process ( 3 )
are now approximated as @xmath48 where the incoming current amplitudes are given by @xmath49 it has been shown in ref .
@xcite that the helicity dependent phases of the above current amplitudes reproduce azimuthal angle correlations among the jets while their magnitudes give the well - known dglap splitting functions when squared and summed over helicities . following ref .
@xcite , we define the reduced current amplitudes @xmath50 as follows @xmath51 where @xmath52 and the coupling factor @xmath53 is given by @xmath54 for quark or anti - quark currents , and @xmath55 for gluon currents . here , @xmath56 is the qcd coupling constant , while we suppress the colour indices in the helicity amplitudes of eqn .
[ eq : ampl ] for brevity .
the reduced current amplitudes @xmath50 have been computed and tabulated in ref .
@xcite , to which we refer the reader for details .
however , it might be worth recalling some of the salient features of these reduced current amplitudes .
first of all , the reduced amplitudes for anti - quark currents are same as those of quark currents . and since we neglect light quark masses , by chirality conservation , the helicity - flip amplitudes will be zero for quark currents , i.e. , @xmath57 , while this is not the case for gluon current amplitudes .
the existence of helicity - flip amplitudes in gluon currents give rise to certain special features in the amplitudes for @xmath19 jets involving them .
it is useful to look into the current amplitudes from the point of view of parton branching @xcite , whereby the outgoing partons are emitted collinearly . in the current amplitudes @xmath58 or @xmath59 ,
if @xmath60 is the energy fraction of the initial parton that is carried away by the off - shell gluon , then one can write the current amplitudes in terms of the energy fractions of the two t - channel gluons , @xmath61 , @xmath62 , and the azimuthal angles of the outgoing partons @xmath63 and @xmath64 , in the frame in which the t - channel gluons are taken as propagating along the @xmath65-axis .
let us first fix the reference frame in which we shall be working . as mentioned above
, we take this frame to be the centre of mass frame of the two t - channel gluons producing the @xmath16 pair , which is also the rest frame of the @xmath16 pair . in this frame
, the two gluons can be taken to be propagating in opposite directions along the @xmath60-axis with energy @xmath66 each . therefore , the scattering process
, @xmath67 can be assumed to be taking place in the @xmath68 plane . here ,
@xmath69 denote the colour indices for the gluons and @xmath70 denote the colour indices for the quark and the anti - quark .
the four momenta of the gluons , @xmath71 and @xmath72 are given by @xmath73 where @xmath74 denotes the velocity of @xmath71 in the @xmath16 rest frame and is given by @xmath75 , and @xmath76 is the invariant mass of the @xmath16 system .
the helicity amplitudes for the process @xmath45 can be expressed in the following form : @xmath77_{i,{\overline{i } } } \mathcal{\hat{n}}_{\lambda_1 \lambda_2}^{\sigma \bar{\sigma}}\right]~. \label{eq - amp - ggqq}\ ] ] one can further simplify eqn .
[ eq - amp - ggqq ] by noting the following relation @xmath78 finally , the helicity amplitudes @xmath79 , for different helicity combinations of the heavy quarks and gluons , are given by @xmath80
the squared matrix element ( averaged over initial state and summed over final state colour ( c ) and spin ( s ) ) can be obtained by using eqns .
[ eq - gen - form ] , [ eq - amp - ggqq ] , [ eq - hel - amp ] , and the current amplitudes @xmath81 as tabulated in ref .
we find that for all of the different collision processes initiated by @xmath82 and @xmath2 , the squared amplitude can be cast in the following form : @xmath83 \label{eq - msq}\end{aligned}\ ] ] where the colour factor @xmath84 comes from the quark and gluon current amplitudes , averaged over initial colour , and is given by @xmath85 and @xmath86 for @xmath82 and @xmath2 sub - processes respectively .
the co - efficients @xmath87 s in the above expression are functions of @xmath88 , the energy fractions of the initial partons carried away by the t - channel intermediate gluons . for notational convenience ,
let us also define @xmath89 . here
@xmath63 and @xmath64 are the azimuthal angles of the two outgoing partons ( jets ) in the @xmath16 rest frame whose polar axis is chosen along the virtual gluon momentum as defined in sub - section [ sec : gg - qq ] . in the collinear limit , these azimuthal angles reduce to the standard ones in the lab frame with polar axis along the beam direction .
for definiteness , we take the heavy quark q ( q = b or t ) momentum to have zero azimuthal angle .
the general expressions for @xmath90 , which determines the total cross - section , for the @xmath0 , @xmath1 and @xmath2 initiated sub - processes are as follows : @xmath91 & = \frac{(1+\overline{z_1}^2)}{z_1 ^ 2 } \frac{(1+\overline{z_2}^2)}{z_2 ^ 2 } \nonumber \\
f_0[qg ] & = \frac{(1+\overline{z_1}^2)}{z_1 ^ 2}\frac{(1+z_2 ^ 4+\zb_2 ^ 4)}{z_2 ^ 2 \zb_2 } \nonumber
\\ f_0[gg ] & = \frac{(1+z_1 ^ 4+\zb_1 ^ 4)}{z_1 ^ 2 \zb_1 } \frac{(1+z_2 ^ 4+\zb_2 ^ 4)}{z_2 ^ 2 \zb_2}~. \label{eq - f0}\end{aligned}\ ] ] the function @xmath92 is found to be the same for the @xmath1 and @xmath2 initiated processes , and is given by , @xmath93 & = \frac{\overline{z_1}}{z_1 ^ 2 } \frac{(1+\overline{z_2}^2)}{z_2 ^ 2 } \nonumber \\
f_1[qg / gg ] & = \frac{\zb_1}{z_1 ^ 2 } \frac{(1+z_2 ^ 4+\zb_2 ^ 4)}{z_2 ^ 2 \zb_2}~. \label{eq - f1}\end{aligned}\ ] ] similarly , the function @xmath94 is the same for the @xmath0 and @xmath1 initial states , and is as follows : @xmath95 & = \frac{\overline{z_2}}{z_2 ^ 2 } \frac{(1+\overline{z_1}^2)}{z_1 ^ 2 } \nonumber \\ f_2[gg ] & = \frac{\zb_2}{z_2 ^ 2}\frac{(1+z_1 ^ 4+\zb_1 ^ 4)}{z_1 ^ 2 \zb_1}~. \label{eq - fqg}\end{aligned}\ ] ] the @xmath88-dependent co - efficients of the terms proportional to @xmath96 and @xmath97 , namely @xmath98 and @xmath99 , which will be seen to determine the azimuthal correlations in the threshold and the relativistic limits respectively , are found to be equal to each other ( i.e. , @xmath100 ) , and are also remarkably the same for all three sub - processes , @xmath101 = \frac{\zb_1}{z_1 ^ 2 } \frac{\zb_2}{z_2 ^ 2}~. \label{eq - fgg}\ ] ] we see that there is an enhancement in all the amplitudes in the @xmath102 and @xmath103 limits , i.e. , when the t - channel intermediate gluons become soft .
furthermore , in the amplitudes involving gluon currents ( i.e. , the @xmath1 and @xmath2 initiated processes ) , there is an additional singularity as @xmath104 ( i.e. , @xmath105 ) .
this happens because not only the intermediate gluons can become soft , but now an outgoing gluon jet can become soft as well . since the form of @xmath98 and @xmath99 are identical for all of the @xmath82 and @xmath2 collision processes , and these functions
determine the angular distributions in the two kinematic limits to be considered next , the only difference that will arise is from @xmath90 .
we shall see that the ratio of @xmath106 or @xmath99 to @xmath90 determines how strong the azimuthal correlations will be .
we now explore the two different kinematic limits in which one obtains characteristic strong azimuthal angle correlations that can be probed at the lhc .
the first one is when the @xmath16 pair is produced near the threshold , i.e. , the invariant mass @xmath107 , and therefore , @xmath108 .
we can then express the squared matrix element in eqn .
[ eq - msq ] in this simple form : @xmath109 \label{eq : thres}\ ] ] from this equation , we can see that the differential distribution in @xmath110 will show a strong suppression at @xmath111 , while it will show an enhancement at @xmath112 ( we use the convention in which the azimuthal angles are in the interval @xmath113 $ ] ) .
note that , in eqn .
[ eq : thres ] , in the soft - limit for the t - channel gluons ( @xmath114 ) , the combination @xmath115 . in the next section
we shall see that by studying the threshold production of top quark pair in @xmath116 jets events at the lhc , we do observe these features in the corresponding differential distribution for @xmath110 .
we would like to point out that in the production of a cp - odd scalar boson in association with 2 jets one expects a similar distribution in @xmath110 .
the second kinematic region of interest is the relativistic limit in which the @xmath16 pair invariant mass is very high , such that the masses of the heavy quarks can be ignored . in this limit , as @xmath117 , @xmath118 .
therefore , we can express the squared matrix element in eqn .
[ eq - msq ] as follows : @xmath119 \label{eq : relat}\ ] ] from the above expression we see that in the relativistic limit , the differential distribution in the sum of azimuthal angles , @xmath120 , will show a strong suppression at @xmath121 , while it will show an enhancement at @xmath122 . in the next section , we shall study @xmath19 jets ( @xmath17 ) processes at the lhc by demanding a high invariant mass for the @xmath123 pair , and shall see that we are able to observe these features in the differential distribution for @xmath120 .
we now discuss the results for exact numerical evaluation of the differential distributions for @xmath110 and @xmath120 in the @xmath116 jets and @xmath124 jets processes . here , in addition to the gf diagrams , all other diagrams at the tree level have also been taken into account . using suitably devised cuts ,
we show that we are able to observe the angular correlations in different kinematic regions , as obtained using only the gf contribution in the previous section .
we also compare the distributions and total cross - sections obtained by using the approximate matrix elements ( presented in section [ analytic ] , eqns .
[ eq - msq]-[eq - fgg ] ) with the exact ones . for exploring the threshold limit , we concentrate on the process @xmath116 jets , with @xmath125 close to @xmath24 . for the relativistic limit , in which @xmath117
, we consider both the production of @xmath124 jets and @xmath116 jets with a high invariant mass of the bottom or top pair .
we present our results for lhc with a centre of mass energy of @xmath126 .
the exact matrix elements were calculated with the help of madgraph 5 @xcite at parton level .
subsequently , both the exact and approximate matrix elements were integrated over the relevant four - body phase - space using the bases @xcite monte - carlo integration package . note that , we do not include the effects of parton shower in this study .
we have used the cteq6l1 @xcite parton distribution functions with lo running of @xmath127 , and @xmath128 . for the factorization scale
we choose the minimum value of transverse momenta demanded for the tagging jets ( @xmath129 ) .
the strong coupling constant @xmath127 is evaluated at the corresponding minimum transverse energy of the final state partons or heavy quarks , i.e. , for the @xmath116 jets process , we take @xmath130 , while for @xmath124 jets it is taken to be @xmath131 . in our analysis , @xmath132 is taken as @xmath133 ( see cut-3 below ) .
the values of top and bottom quark masses used are @xmath134 and @xmath135 respectively @xcite .
to start with , we impose a set of selection cuts on the final - state partons ( jets ) : * cut-1 * : @xmath136 , @xmath137 , @xmath138 , @xmath139 here @xmath140 , @xmath141 and @xmath142 denote the transverse momenta , energy and pseudorapidity of the jets respectively , while @xmath143 defines the separation between the jets in the pseudorapidity - azimuthal angle plane .
we have subjected the tagging jets to a slicing @xmath144 cut , and have also demanded a high value for the minimum energy of the jets . since we want to isolate the contribution of the gf diagrams to the full process , we also need to impose the so - called vbf cuts , which require the two tagging jets to reside in the opposite hemispheres of the detector and to be well separated in pseudorapidity .
the vbf cuts imposed in our study are as follows : * cut-2 * : @xmath145 , @xmath146 .
in addition to the above cuts on the tagging jets , we also impose a set of cuts on the @xmath28 and @xmath147 quarks .
we demand a minimum transverse energy ( @xmath148 ) for the @xmath147 quarks , such that @xmath149 is always higher than the tagging jet @xmath144 s .
also , the @xmath28 and @xmath147 quarks are required to be in the central region of the detector .
these cuts are summarized below : * cut-3 * : @xmath150 , @xmath151 in figure [ fig : fig-1 ] , we show the differential distributions obtained using the exact matrix elements for both the azimuthal angle difference of the two tagging jets , @xmath110 , and the sum of the azimuthal angles @xmath120 in @xmath116 jets events , for the sub - processes initiated by @xmath152 and @xmath2 .
in addition , we show the sum of all sub - process contributions , too . in order to obtain the azimuthal angle distributions in the threshold limit
, we need to ensure that the invariant mass of the @xmath3 system , @xmath153 , should be very close to @xmath154 .
therefore , in this figure , we have imposed an invariant mass cut of @xmath155 .
as explained in sub - section 3.1 , we indeed see a distribution resembling the behaviour predicted by eqn .
[ eq : thres ] .
in particular , we see the expected suppression at @xmath111 , and enhancement at @xmath112 . on the otherhand , the @xmath156 distribution is observed to be rather flat , as expected in this region of @xmath153 .
this agreement between the analytic approximation and the exact numerical evaluation shows the usefulness of the vbf cuts in selecting the gf contribution to the total cross - section . in order to estimate the feasibility of studying the angular distributions in the @xmath116 jets process at the @xmath126 lhc
, we show the cross - sections after cuts 1 , 2 and 3 , and various values of the @xmath153 cut in table [ mtt ] .
as we can infer from this table , the total cross - section from all sub - processes taken together is considerable , such that with an expected integrated luminosity of @xmath157 , one can have sufficient number of events to study the azimuthal angle correlations at the 8 tev run of lhc .
.[mtt ] cross - sections of various sub - processes contributing to @xmath116 jets at 8 tev lhc after cuts 1 , 2 and 3 and different values of the @xmath125 cut .
we also show the total @xmath116 jets cross - section after these cuts . [ cols="<,^,^,^,^",options="header " , ] in figure [ fig : fig-4 ] , we show the normalized differential distributions for the @xmath124 jets process , where we present the sum of all sub - process contributions coming from @xmath82 and @xmath2 . here , we have shown the effect of variation in the @xmath158 cut .
we also compare the normalized distributions obtained by using the approximate and the exact matrix elements , and they are found to agree to a good accuracy , as in the @xmath116 jets case .
we show in table [ bb - comp ] the ratio of the total cross - section obtained by using the exact and approximate matrix elements for two different choices of the slicing @xmath144 cuts on the tagging jets . here also , the approximate matrix element is found to overestimate the total cross - section , and the difference between the two is reduced once the tagging jet @xmath144 s are demanded to be very small , thereby approaching the collinear limit for the t - channel gluons .
in this paper , we have studied the azimuthal angle correlations of jets produced in association with a top or bottom quark pair at the lhc .
we have presented the helicity amplitudes for the process @xmath19 jets using only the gluon fusion diagrams and employing the on - shell gluon approximation to the matrix element .
this gives us a general expression for the azimuthal angle distributions for processes initiated by @xmath152 and @xmath2 initial states .
subsequently , we have explored this matrix element in two different kinematic limits .
the first one is the threshold production of the @xmath16 pair , in which the invariant mass @xmath76 is very near to @xmath159 . in this limit
we find a strong angular correlation in the distribution of the difference of azimuthal angles of the tagging jets , @xmath110 .
the other kinematic limit considered is the relativistic limit , in which @xmath76 is very much higher than @xmath159 . in this limit , while the distribution in @xmath110 is expected to be rather flat , there is a strong correlation in the sum of the azimuthal angles , @xmath120 .
we performed an exact numerical evaluation of these angular distributions in @xmath116 jets and @xmath124 jets processes at the lhc , including all the diagrams at the tree level , and verified our analytical approximations using a suitably devised set of kinematic selection cuts .
the vbf selection cuts are found to efficiently select the gf contribution to these processes .
a comparison between the normalized differential distributions obtained by using the exact and the approximate matrix elements also shows a very good agreement between the two .
the total cross - sections are overestimated by the approximate matrix elements by about a factor of 2 , while they approach the exact cross - sections as the @xmath144 s of the tagging jets are demanded to be smaller , thereby achieving the collinear limit for the t - channel intermediate gluons , in which our approximation approaches the exact matrix elements .
we expect these azimuthal correlations calculated at the leading order to survive even after including higher order corrections , as has been previously demonstrated in the case of higgs production with multiple hard jets @xcite .
one particular future direction that can be explored is to extend our parton level results by using a matrix elements matched with parton shower method .
we would like to stress that the study of such angular correlations among jets produced with a top or bottom quark pair is not only interesting in itself , it has useful implications in the study of vbf processes for determining the spin and cp properties of new particles which can be discovered at the lhc . in particular , the experimental technique to measure such correlations between tagging jets
can be established first by using these sm processes which have sufficiently high cross - sections .
we expect that these studies can be performed with the already accumulated data at the 8 tev run of the lhc .
sm is grateful to junichi kanzaki for many a help regarding the use of the bases integration package .
sm would also like to acknowledge the warm hospitality of the theory center at kek , japan , and the recapp centre at harish - chandra research institute , india , where major portions of this work were carried out .
this work is partially supported by world premier international research center initiative ( wpi initiative ) , mext , japan , and by grant - in - aid for scientific research ( no.-20340064 ) from jsps .
t. plehn , d. l. rainwater and d. zeppenfeld , phys .
* 88 * ( 2002 ) 051801 ; v. del duca , w. kilgore , c. oleari , c. schmidt and d. zeppenfeld , phys .
* 87 * ( 2001 ) 122001 ; v. del duca , w. kilgore , c. oleari , c. schmidt and d. zeppenfeld , nucl . phys .
b * 616 * ( 2001 ) 367 ; v. hankele , g. klamke and d. zeppenfeld , arxiv : hep - ph/0605117 ; k. odagiri , jhep * 0303 * ( 2003 ) 009 ; v. del duca _
et al . _ ,
jhep * 0610 * ( 2006 ) 016 ; m. r. buckley and m. j. ramsey - musolf , jhep * 1109 * ( 2011 ) 094 ; s. y. choi , m. m. muhlleitner and p. m. zerwas , phys . lett .
b * 718 * ( 2013 ) 1031 ; j. frank , m. rauch and d. zeppenfeld , arxiv:1211.3658 [ hep - ph ] ; c. englert , d. goncalves - netto , k. mawatari and t. plehn , arxiv:1212.0843 [ hep - ph ] .
k. hagiwara , q. li and k. mawatari , jhep * 0907 * , 101 ( 2009 ) .
n. cabibbo , a. maksymowicz , phys .
* 137 * ( 1965 ) b438 ; j. r. dellaquila , c. a. nelson , phys .
d * 33 * ( 1986 ) 80 ; j. r. dellaquila , c. a. nelson , phys .
d * 33 * ( 1986 ) 93 ; c. a. nelson , phys .
d * 37 * ( 1988 ) 1220 .
v. n. gribov and l. n. lipatov , sov .
j. nucl .
* 15 * ( 1972 ) 438 ; l. n. lipatov , sov .
j. nucl .
phys . * 20 * ( 1975 ) 94 ; g. altarelli and g. parisi , nucl .
b * 126 * ( 1977 ) 298 ; y. l. dokshitzer , sov .
jetp * 46 * ( 1977 ) 641 .
j. pumplin , d. r. stump , j. huston , h. l. lai , p. nadolsky and w. k. tung , jhep * 0207 * ( 2002 ) 012 . k. nakamura _ et al . _
[ particle data group collaboration ] , j. phys .
g g * 37 * , 075021 ( 2010 ) .
j. r. andersen , v. del duca and c. d. white , jhep * 0902 * ( 2009 ) 015 ; | angular correlation of jets produced in association with a massive scalar , vector or tensor boson is crucial in the determination of their spin and cp properties .
we study jet angular correlations in events with a high mass bottom quark pair or a top quark pair and two jets at the lhc , whose cross - section is dominated by the virtual gluon fusion sub - processes when appropriate kinematic selection cuts ( vector - boson fusion cuts ) are applied .
we evaluate helicity amplitudes for sub - processes initiated by @xmath0 , @xmath1 and @xmath2 collisions in the limit where the intermediate gluons are collinear to the initial partons .
we first obtain a general expression for the azimuthal angle correlations among the dijets and @xmath3 or @xmath4 , in terms of the @xmath5 or @xmath4 helicity amplitudes in the real gluon approximation of the full matrix elements , and find simple analytic expressions in the two kinematic limits , the production of a heavy quark pair near the threshold , and in the relativistic limit where the invariant mass of the heavy quark pair is much larger than the quark mass . for @xmath6 jets we find strong azimuthal angle correlations which are distinct from those expected for events with a cp - even or odd scalar boson which may decay into a @xmath4 pair . for @xmath7 jets
we find that the angular correlations are similar to that of a cp - odd scalar@xmath8 jets near the threshold @xmath9 , while in the relativistic limit they resemble the distribution for @xmath6 jets .
these correlations in the standard qcd processes will help establish the experimental technique to measure the spin and cp properties of new particles produced via gluon fusion at the lhc .
kek - th-1597 + ipmu 13 - 0025 + kaoru hagiwara@xmath10 , satyanarayan mukhopadhyay@xmath11 + _ @xmath10kek theory center and sokendai , + tsukuba 305 - 0801 , japan .
+ @xmath11kavli ipmu ( wpi ) , university of tokyo , + kashiwa 277 - 8583 , japan .
_ + |
the gamma - ray large area space telescope ( glast ) is the next nasa general gamma - ray astrophysics mission , which is scheduled to be launched into low earth orbit in february , 2007 , for 510 years of operation .
it will consist of two instruments : the large area telescope ( lat ) and the glast burst monitor ( gbm ) . a product of a nasa / doe / international collaboration
, the lat will be a pair conversion telescope covering the @xmath020 mev to @xmath1300 gev energy band
. the lat will be @xmath230 times more sensitive than _
cgro s _ egret .
the gbm will detect and localize bursts , and extend glast s burst spectral sensitivity to the @xmath010 kev to @xmath125 mev band . consisting of 12 nai(tl ) ( 101000 kev ) and 2 bgo ( 0.1525 mev ) detectors
, the gbm will monitor @xmath18 sr of the sky , including the lat s field - of - view ( fov ) .
bursts will be localized to @xmath3 ( 1@xmath4 ) by comparing the rates in different detectors . during most of the mission glast
will survey the sky by rocking @xmath5 above and below the orbital plane around the zenith direction once per orbit .
the first year will be devoted to a sky survey while the instrument teams calibrate their instruments . during subsequent years
guest investigators may propose pointed observations , but continued survey mode is anticipated because it will usually be most efficient . both the gbm and the lat will have burst triggers . when either instrument triggers , a notice with a preliminary localization will be sent to the ground through tdrss and then disseminated by gcn .
additional data will be sent down through tdrss for an improved rapid localization on the ground .
`` final '' positions will be calculated from the full downlinked data .
all positions will be disseminated as gcn notices , and additional information ( e.g. , fluences and durations ) will be sent out as gcn circulars .
using its own and the gbm s observations , the lat will decide autonomously whether to slew to the burst location for a 5 hour followup pointed observation .
the threshold will be higher for gbm - detected bursts outside the lat s fov .
the gbm s nai and bgo detectors will provide the number of counts detected in 8 energy bands every 16 ms ; the gbm will trigger off these rates .
rate triggers test whether the increase in the number of counts in an energy band @xmath6 and time bin @xmath7 is statistically significant ; the expected number of non - burst photons in the @xmath6@xmath7 bin is estimated by accumulating counts before the time bin being tested . building on our experience with the batse trigger , we are performing trade studies to optimize the sensitivity of these triggers . here
we present the results of our studies focusing on the choice of @xmath8 and @xmath7 .
we consider two @xmath7 hierarchies@xmath7 spaced by factors of @xmath102 ( e.g. , 16 ms , 32 ms , 64 ms ... ) or @xmath104 ( e.g. , 16 ms , 64 ms , 256 ms ... )and three time bin registrations non - overlapping bins ( e.g. , bins separated by 1024 ms for @xmath7=1024 ms bins ) , half - step bins ( e.g. , bins separated by 512 ms for @xmath11=1024 ms bins ) , and all possible bins ( e.g. , bins every 16 ms ) .
varying @xmath7 can maximize the signal - to - noise ratio , while more time registrations permit the bin to be centered over the peak flux , also maximizing the signal - to - noise ratio . to test these 6 triggers
we applied them to the 64 ms resolution lightcurves of the 25 brightest batse bursts ; for each lightcurve we chose 10 starting times at random . note that the gbm rates will have 16 ms resolution .
the most sensitive trigger would have @xmath7 spaced by @xmath102 and every possible time bin .
the next most sensitive trigger would have @xmath7 spaced by @xmath102 and bins spaced every half step .
these triggers would test different numbers of bins11264 bins vs. 3070 bins tested every 16.384 s. besides the increased computational burden , the risk of a false trigger increases as the number of bins tested increases , but the false trigger probability is not proportional to the number of bins because the bins are not independent .
our simulations indicate this is a @xmath05% effect for the same false trigger rate the trigger threshold should be raised by a few percent as the number of time bins tested increases .
triggering on the counts accumulated in different @xmath6 can tailor the detector sensitivity to hard or soft bursts .
the gbm will be able to trigger on more than one @xmath8 , and therefore we seek the set that will maximize the sensitivity for both hard and soft bursts , although hard bursts are a priority since their spectra are more likely to extend into the lat s energy band ( of course one must be careful of one s theoretical prejudices ) . for the study of detector sensitivity to different types of bursts and for comparisons between detectors ,
the @xmath12-@xmath13 plane is useful@xcite , where @xmath12 is the peak photon flux in a fiducial energy band ( here 11000 kev ) and @xmath13 is the energy of the peak of @xmath14 .
burst spectra are also characterized by asymptotic low and high energy power laws with spectral indices @xmath15 and @xmath16 , respectively ; the dependence of @xmath12 on @xmath15 and @xmath16 is not as great as on @xmath13 . for a given set of spectral indices
the detector sensitivity ( the threshold value of @xmath12 at a give @xmath13 ) is a curve in the @xmath12-@xmath13 plane . to calculate these sensitivity curves
we need both the number of counts a detector will detect in the nominal @xmath8 band for a given burst spectrum and the number of background counts in this @xmath8 . a code has been developed to calculate these numbers for each gbm detector for a burst in any direction relative to the spacecraft .
currently the code uses response matrices for the flux directly incident on the detectors ( without scattering off the spacecraft or the earth , but with obscuration by other parts of the observatory ) , and a model of the background on orbit .
we used this code to calculate the sensitivity along the normal to the lat for @xmath7=1.024 s assuming at least two detectors trigger at @xmath17 .
we calculated the sensitivity curves for a variety of @xmath8 .
the extremes of our spectral index sets were @xmath15=0 , @xmath18 and @xmath19 , @xmath20 .
the first set is similar to the spectra sometimes observed early in a burst ; its high energy tail would be more easily detected by the lat .
the second set is a spectrum without a high energy tail . for @xmath15=0 , @xmath18 ( left plot ) and @xmath19 , @xmath20 ( right plot ) .
the solid curves are for ( left to right within plot ) @xmath8=5100 , 50300 and 1001000 kev and the dashed for @xmath8=51000 and 501000 kev .
lower curves are more sensitive.,title="fig : " ] for @xmath15=0 , @xmath18 ( left plot ) and @xmath19 , @xmath20 ( right plot ) .
the solid curves are for ( left to right within plot ) @xmath8=5100 , 50300 and 1001000 kev and the dashed for @xmath8=51000 and 501000 kev .
lower curves are more sensitive.,title="fig : " ] figure 1 shows the sensitivity for two sets of @xmath8 . to compare the gbm and batse burst distributions we would like to include @xmath8=50300 kev , which was batse s primary trigger band .
as can be seen , @xmath8 with a low energy cutoff of @xmath250 kev is optimal for high energy sensitivity because it does not include the large low energy background .
conversely , @xmath8 should always extend to the highest energy possible because of the low background at high energy .
we find that triggering a single bgo detector with @xmath21 increases the sensitivity for @xmath22 mev for @xmath15=0 , @xmath18 .
there is no increase in sensitivity for @xmath23 .
figure 2 compares the @xmath7=1 s sensitivity for the gbm ( solid ) and batse ( dot - dashed ) with the burst intensity ( dashed ) that when the spectrum is extrapolated to the lat energy band would result in 25 detected photons per second .
the burst is assumed to be on the lat normal , @xmath19 , @xmath18 , and the gbm triggers on @xmath8=5100 and 50300 kev .
the @xmath7 calculations suggest that spacing @xmath7 by factors of @xmath102 ( i.e. , 16 ms , 32 ms , 64 ms ... ) and staggering the bins by half a timestep ( e.g. , the 1024 ms bins are accumulated every 512 ms ) would be particularly efficient given the number of time bins that would be tested . choosing two triggers with @xmath8 starting at 5 kev and 50 kev and extending to the detector s high energy cutoff would provide good high and low energy sensitivity . using @xmath8=50300 kev would reproduce the batse trigger , but would reduce the @xmath24500 kev sensitivity for the hardest bursts ( which are more likely to have lat flux ) ;
this can be mitigated by adding @xmath6=1001000 kev .
future trade studies will focus on the methodology for measuring the background .
glast will slew more frequently than _ cgro _ did , and therefore the background rates in glast s different detectors will probably vary more rapidly .
we will need to suppress spurious triggers resulting from variations in the background without a major reduction in the gbm s sensitivity . | the glast burst monitor ( gbm ) will detect and localize bursts for the glast mission , and provide the spectral and temporal context in the traditional 10 kev to 25 mev band for the high energy observations by the large area telescope ( lat ) .
the gbm will use traditional rate triggers in up to three energy bands , and on a variety of timescales between 16 ms and 16 s. address = code 661 , nasa / goddard space flight center , greenbelt , md 20771 address = national space science and technology center , huntsville , al 35805 address = national space science and technology center , huntsville , al 35805 address = nis-2 , los alamos national laboratory , los alamos , nm 87545 address = national space science and technology center , huntsville , al 35805 |
the aim of carmenes ( @xcite ) is to perform high - precision measurements of stellar radial velocities with long - term stability .
the fundamental science objective is to carry out a survey of late - type main sequence stars ( with special focus on moderately active stars of spectral type m4v and later ) with the goal of detecting low - mass planets in their habitable zones . for stars
later than m4-m5 ( @xmath1m@xmath2 ) , a radial velocity precision of 1 m / s ( per measurement ; @xmath3 ) will permit the detection of super - earths of 5 m@xmath4 and smaller inside the entire width of the habitable zone with 2@xmath3 radial - velocity amplitudes ( i.e. , @xmath5 m / s ) .
for a star near the hydrogen - burning limit and a precision of 1 m / s , a planet as small as our own earth in the habitable zone could be detected .
in addition , the habitable zones of all m - type dwarfs can be probed for super - earths .
the carmenes survey will be carried out with the 3.5-m telescope on calar alto , using at least 600 clear nights in the 2014 - 2018 time frame .
we plan to survey a sample of 300 m - type stars for low - mass planet companions ( see also @xcite ) .
this will provide sufficient statistics to assess the overall distribution of planets around m dwarfs : frequency , masses , and orbital parameters .
the seemingly low occurrence of jovian planets should be confirmed , and the frequency of ice giants and terrestrial planets should be established along with their typical separations , eccentricities , multiplicities , and dynamics .
the project is now at cruise speed towards its final phases of final design ( fd ) and assembly integration and verification ( aiv ) , with most of the funding needed to build the instrument already secured . in this contribution
we will also describe how this project was born and the first historical steps taken to consolidate the consortium and reach the current situation .
we believe that the details of this process can be of interest to researchers who find themselves at the first stages of an instrumental project in spain .
temporal line of the carmenes project . past
phases are shown in black and current and future phases in red .
, width=340 ] the time frame for the first stages of the project must be set during 2005 , when one of us ( pja ) was already part of the scientific committee of the project nahual ( principal investigator ( pi ) e. l. martn ; @xcite ) , to build and exploit a high - resolution , near - infrared ( nir ) spectrograph for the gran telescopio de canarias ( gtc ) .
the responsibilities in that committee were to represent the interests of the line of research of asteroseismology and magnetic activity in cool stars . for scientific and economic reasons , e. l. martn , r. garrido and p. j. amado decided , at the end of 2006 , to present a similar instrument to a call for proposals for the 3.5-m telescope of the centro hispano - alemn de calar alto ( caha ) in almera , spain ,
for which pja would be the pi .
the first step ( february 2007 ) as pi of such a project was to send a `` letter of interest '' to the iaa s director , which was immediately forwarded to the csic , informing of our intention to design , build and exploit a common - user , fiber - fed , high - resolution spectrograph in the nir for the 3.5-m telescope at caha .
the main science case of this project was the search for exoplanets around m - type stars and asteroseismology of pulsating stars with planets .
the researchers involved in the project , at that time , were only those mentioned above .
the second step was , of course , to raise funds to start supporting the work of our engineers at the instituto de astrofsica de andaluca ( iaa - csic ) for a conceptual design ( cd ) and of the researchers at the iac and the iaa for the preparation of the science .
to this aim , we decided to apply for funds to the 6th framework programme era - space technologies applications and research for the regions and medium - sized countries ( era - star regions ) aimed at boosting space research at regions or small countries throughout europe known to have space research capabilities .
an application to this programme needed of a small consortium to be set up of at least two proposers of two different countries / regions eligible for funding ( with space research capabilities ) .
andaluca , in the south of spain , and madrid both belonged to this program .
therefore , the universidad complutense de madrid ( ucm ) together with the iaa and two more european nodes , austria and lombardy ( a region of italy ) , were involved in what it would become the first carmenes ( although the project did not have that name at the time ) consortium .
two months later ( april 2007 ) other groups working in different research lines at the iaa were also involved to apply for funds to a programme of the andalusian regional government .
our possibilities of success had seem to be larger at that time , considering that caha is sited on andalusian territory .
eventually , these two proposal did not succeed but engineers and researchers at the iaa kept working on what it was thought to be very interesting project .
the milestone in all this process arrived when the iaa , as co - manager of the caha observatory , issued an open call for ideas for new instrumentation for the observatory in march 2008 . to this call
, seven instruments were presented , among which was carmenes and other proposals from german institutions based on designs for high - resolution single - object and multi - object spectrographs in the visible wavelength range with similar science cases .
pja came up with the name `` carmenes '' for the project just the night before his presentation during the workshop that was held at the iaa , a few months later , to manage the results of the call .
more information on the reason for this name and the meaning of it can be found at http://carmenes.caha.es/. layout of the instrument design .
two spectrographs ( channels ) , one in the vis and one in the nir wavelength ranges will be thermo - mechanically stabilized in the 3.5-m telescope s coude room and attached to the telescope with fibres through a front - end adaptor . , width=340 ] from there , things happened very quickly . contacts with our current german colleagues were initiated from the iaa to suggest a merging of our projects . that increased the probabilities of being selected to be funded for a cd by caha .
this collaboration started the second era of the carmenes project .
eventually , a unified proposal for a `` modular '' instrument with three high - resolution channels ( spectrographs ) was presented in november 2008 .
the three spectrographs were one single - object and one multi - object , both in the visible and one single - object in the nir , the latter being the core , non - downscopable part , of the project .
the consortium was ( is ) based on the parity between the germany and spain and its structure will be explained below in sect .
[ consortium ] . in january 2009 , carmenes was selected as one out of the two instruments to be funded for a cd by the observatory . to that moment
all the instruments built for , and in operation at , caha had been proposed by german institutions and , therefore , carmenes becomes , to date , the first instrument proposed , accepted and co - led by a spanish institution . during the next few months we developed the cd , which we presented in october 2009 to a cd review board composed by an international team of experienced researchers and senior engineers .
the caha executive committee , based on the recommendations from this panel , took the decision of selecting carmenes as the next - generation instrument for the caha 3.5-m telescope , culminating with success the work carried out during a period of time of almost three years .
since then , the project has evolved in a normal way , if for normal we understand the extremely tight time line this `` schedule - driven '' project must keep to have the instrument working at the telescope in 2014 . in any case , we went through a `` not - too - normal '' situation during 2009 and 2010 in which germany and spain had started negotiations on the agreement for the operation of the observatory for the period 2014 - 2018 , which eventually finished extending the current agreement ( 2004 - 2013 ) for five more years and having carmenes as the key project for the observatory . in july 2011 , we passed successfully a preliminary design review and at the end of this year ( 2012 ) we will go through the final design review ( fdr ) . for a brief summary of the carmenes temporal line see fig . [ fig1 ] .
finally , in 2014 the instrument will be handed to and accepted by caha observatory to be used by the spanish and german communities through both open and guaranteed time .
the guaranteed time will sum up to a total of 600 nights in five years ( with a possible buffer of another 150 nights ) , and will be used by the consortium to exploit its main science case .
this guaranteed program is similar to that of the harps instrument and similar high - impact results are expected .
as mentioned above we are in the final design phase towards the fdr at the end of the year and the assembly , integration and verification ( aiv ) phase during next year .
the carmenes instrument consists now of two separate chelle spectrographs covering the wavelength range from 0.55 to 1.7 @xmath0 m at a spectral resolution of @xmath6 , fed by fibers from the cassegrain focus of the telescope .
the spectrographs are housed in vacuum tanks providing the temperature - stabilized environments necessary to enable a 1 m / s radial velocity precision employing a simultaneous calibration with emission - line lamps ( see fig . [ fig2 ] ) . due to strict timeline
the project is subjected to , we decided to have a fdr for the optics well in advance of the full instrument fdr to be able to start the procurement of the optical elements , some of which have long - lead times .
the main advantages of this project with respect to similar ones are : 1 ) the possibility of acquiring simultaneous near - infrared and visible observations , 2 ) the requirement for both high resolution and wide spectral coverage will be achieved , 3 ) having long guaranteed time at the 3.5-m telescope for the completion of the project and 4 ) the project is now leading the development of this type of instrumentation and will , therefore , lead its exploitation and this research topic for the next few years .
schematics of the carmenes consortium structure .
, width=340 ] the carmenes consortium is managed on the basis of german - spanish parity .
it was officially created in early 2009 with the aim of designing , building and exploiting the instrument .
the consortium is currently made up of eleven centres and universities in germany and spain , including the observatory : iaa ( granada ) , lsw ( heidelberg ) , mpia ( heidelberg ) , cab ( madrid ) , hs ( hamburg ) , iac ( tenerife ) , iag ( gttingen ) , ice ( barcelona ) , tls ( tautenburg ) and ucm ( madrid ) .
the member institutions of carmenes are centres of the german max - planck - gesellschaft and the spanish consejo superior de investigaciones cientficas , as well as some german and spanish universities and research institutes . a schematic representation of the consortium structure is given in fig . [ fig3 ] .
the pis represent the spanish and german scientific communities .
two project managers , together with two system engineers , two project sientists , ( two posts each , one for a spanish and one for a german representative ) , the caha representative and the mpia representative , together with the pis make up the core management team ( cmt ) .
the cmt informs the carmenes project supervisory board which is composed of the pis , the directors of the institutions co - managing the observatory ( iaa and mpia ) and the observatory s director . to date
, there are over 100 researchers and engineers in the consortium .
the carmenes instrument will require a total investment of 7.5 meur .
the cash cost is of around 5.0 meur , and the value of manpower contributed in - kind by the members of the carmenes consortium is 2.5 meur .
the amount totaling at least 5 meur will be contributed by both caha partners , csic and mpg , and the partners of the carmenes consortium .
most of the funding to be raised by the spanish partners of the consortium is already available through funded projects from the ministry of research and the local government of the junta de andaluca . as a sort of conclusion
, we can say that , if we consider the ratio between the relatively low cost of this instrument and the highly interesting , competitive and with potentially high - impact - result scientific case , the instrument can be defined as a `` very good instrument '' .
pja acknowledges support from the ministry of research through grants aya2010 - 14840 and aya2011 - 30147-c03 - 01 . | carmenes ( calar alto high - resolution search for m dwarfs with exo - earths with near - infrared and optical echelle spectrographs ) is a next generation instrument being built for the 3.5 m telescope at the calar alto observatory by a consortium of eleven spanish and german institutions .
conducting a five - year exoplanet survey targeting 300 m dwarfs with the completed instrument is an integral part of the project .
the carmenes instrument consists of two separate chelle spectrographs covering the wavelength range from 0.55 to 1.7 @xmath0 m at a spectral resolution of r = 82000 , fed by fibers from the cassegrain focus of the telescope .
the spectrographs are housed in vacuum tanks providing the temperature - stabilized environments necessary to enable a 1 m / s radial velocity precision employing a simultaneous calibration with emission - line lamps . |
scanning tunneling microscopy ( stm ) has proven to be an invaluable tool for studying single molecules and their potential applications in nanoscale devices .
in addition to measuring conductance properties , the electroluminescence spectrum , induced by the tunneling current can be studied.@xcite electroluminescence has been experimentally observed on clean metal surfaces,@xcite
films,@xcite nanoparticles,@xcite and single molecules.@xcite additionally , electroluminescence has been observed in lithographically defined three - terminal cdse single - nanocrystal junctions.@xcite in the case of single molecules , the simultaneous measurement of the electrical and optical behavior has the potential to greatly enhance our understanding of nanoscale junctions , and , through detailed analysis of the vibrational spectrum , to provide valuable insight into the conformational structure of single molecules in a junction.@xcite [ cols="^,^ " , ] the measured and calculated quantum yield as a function of the bias voltage are shown in fig .
[ fig : h2tbpp_qc ] .
the difference in absolute magnitude between the measurement and the calculation is due to the unknown detection efficiency in the measurement , which is taken to be a 100 % in the calculation . in both the measurement and the calculation ,
the quantum yield is zero until an excited state becomes available in the bias voltage window.@xcite it then increases until a maximum is reached when all vibrationally excited states have become available .
as noted above , the quantum yield depends on @xmath0 , and not on @xmath1 .
since going from positive to negative bias effectively means switching @xmath0 and @xmath1 , the maximum value of the quantum yield changes by a factor of @xmath2 , and is therefore directly proportional to the asymmetry in the electronic coupling . in the measurement ,
the quantum yield drops when the bias exceeds 3.5 v , which is attributed to damage to the molecules in ref . .
this can , however , also be explained by the appearance of another charge state in the bias window , which changes the quantum yield by a factor of @xmath3 ( see sec .
[ sec : method ] ) . with approximately equal couplings to the source and drain ( @xmath4 ) ,
the quantum yield is reduced by about 50% [ as is the case in fig .
[ fig : h2tbpp_qc](b ) ] .
however , with very asymmetric , couplings ( @xmath5 ) , as is the case in the measurement of qiu _
et al._,@xcite the reduction is expected to be unobservable . calculated current dependence of h@xmath6tbpp at positive and negative bias .
the current is varied by changing the electronic coupling to the stm tip . ]
dong _ _ et al.__@xcite do not report on the current dependence of the photon count.@xcite however , in the case of symmetric coupling , a linear dependence of the photon count on the current , by varying @xmath1 with the stm tip , is still expected at positive bias ( see fig . [
fig : h2tbpp_current ] ) . at negative bias , on the other hand
, varying the position of the stm tip equals changing @xmath0 , resulting in a _ decrease _ in the photon count with increasing current , albeit at a lower rate .
this is a direct result of the dependence of the quantum yield on @xmath0 via @xmath7 ( see sec .
[ sec : method ] ) .
high - energy part of the calculated electroluminescence spectrum of h@xmath6tbpp at 2.5 v as a function of the vibrational relaxation time .
@xmath8 k for clarity , the other parameters are the same as in fig . [
fig : h2tbpp_spectrum](b ) .
note the appearance of peaks at around 600 nm , to the left of the main peak , when the vibrational relaxation time becomes comparable to the electron transmission rate . ]
since vibrational relaxation is several orders of magnitude faster than photo - emission ( which is on the order of 10@xmath9 s in the measurements and calculations under discussion ) , it is assumed in standard fluorescence measurements that the molecule is always in the vibrational ground state before emitting a photon , although at finite temperatures the lowest vibrational excitations can still be occupied , _
e.g. _ , as observed in anti - stokes scattering . in transport measurements
, however , even the higher vibrationally excited states are continuously repopulated by new electrons from the leads .
when the electron transmission rate becomes on the order of the photo - emission rate , vibrational relaxation will not always take place before the electron leaves the excited state .
it is therefore the ratio between the vibrational relaxation rate and the electron transmission rate , not just the photo - emission rate , that determines the emission spectrum .
this can be seen in fig .
[ fig : h2tbpp_relaxation ] , where we have plotted the high energy part of the emission spectrum of h@xmath6tbpp for different values of the vibrational relaxation time .
peaks start appearing to the _ left _ of the main peak , _
i.e. _ , at higher energies , as soon as the vibrational relaxation rate becomes comparable to the electron transmission rate ( at around 10@xmath10 s ) .
this effect may be present in the electroluminescence spectrum measured by dong _ _ et al.__@xcite
[ see fig .
[ fig : h2tbpp_spectrum](a ) ] , where such a peak can be seen at around 40 nm to the left of the main peak .
in conclusion , we have shown that a computationally efficient method all calculations have been performed on commodity hardware based on the rate - equation formalism is able to obtain good agreement with the measured electroluminescence spectra of two different porphyrin derivatives .
we reproduce the bias voltage and current dependence of the spectra for both symmetric and asymmetric couplings and provide an explanation for the suppression of the quantum yield at high bias .
although in the measurement of dong _ _ et al.__@xcite this suppression may have been caused by damage to molecules,@xcite a significant reduction in the quantum yield is expected in general for every ( nearly ) symmetrically coupled single - molecule junction in the sequential tunneling regime .
in addition , we have shown that vibrational relaxation rates become important when they are comparable to the electron transmission rate , giving rise to peaks in the spectra at higher energies higher than the homo - lumo gap . however
, a detailed study of this effect requires a more sophisticated model for vibrational relaxation , since a single relaxation time for all vibrational modes is only a sufficiently accurate approximation when relaxation is either much faster or much slower than all other rates .
we thank ferry prins for discussions and chris verzijl for help with the calculations .
financial support was obtained from stichting fom ( project 86 ) , from the eu fp7 program under the grant agreement `` single '' , and from the division of chemistry and the division of materials research of the nsf through the northwestern university mrsec . | we have combined _ ab initio _ quantum chemistry calculations with a rate - equation formalism to analyze electroluminescence spectra in single - molecule junctions , measured recently by several groups in scanning tunneling microscope setups . in our method , the entire vibrational spectrum is taken into account .
our method leads to good quantitative agreement with both the spectroscopic features of the measurements and their current and voltage dependence .
moreover , our method is able to explain several previously unexplained features .
we show that in general , the quantum yield is expected to be suppressed at high bias , as is observed in one of the measurements .
additionally , we comment on the influence of the vibrational relaxation times on several features of the spectrum . |
the violation of charge - parity ( cp ) in charm decay requires two amplitudes with different strong and weak phases that interfere to produce cp violating effects .
there are three distinct types of cp violation .
( 1 ) cp violation from a non - vanishing relative phase between the mass and width components of the mixing matrix usually called `` indirect '' ; ( 2 ) direct cp violation due to the two decay amplitudes having different weak phases ; ( 3 ) interference between decays with and without mixing .
the cp conserving phase shift is usually generated by qcd final state interactions ( fsi ) . in the standard model
, the relative weak phase is typically between tree level and penguin amplitudes .
extensions to the standard model introduce additional amplitudes with weak phases that can contribute to cp violation . in the standard model ,
cp violation in the charm sector is small and @xmath0 mixing is highly suppressed , so at current experimental sensitivities searches for cp violation in charm is for physics beyond the standard model .
most cp violation results are from the fnal fixed target experiments e791@xcite and focus@xcite , and the cleo@xcite experiment and search for direct cp violation .
the cp violation asymmetry is defined as @xmath1 a few results from cleo , babar@xcite and belle@xcite experiments consider cp violation in mixing .
99 e. m. aitala _ et al .
_ [ e791 collaboration ] , eur .
j. directc * 1 * , 4 ( 1999 ) .
l. frabetti _ et al . _
[ e-687 collaboration ] , nucl .
instrum .
a * 320 * , 519 ( 1992 ) ; p. frabetti _ et al . _
fermilab - proposal-0831 - 1 .
y. kubota _ et al . _
[ cleo collaboration ] , nucl .
instrum .
a * 320 * , 66 ( 1992 ) ; t. s. hill , nucl .
instrum .
a * 418 * , 32 ( 1998 ) .
b. aubert _ et al . _ [ babar collaboration ] , nucl .
instrum .
a * 479 * , 1 ( 2002 ) .
[ belle collaboration ] , nucl .
instrum .
a * 479 * , 117 ( 2002 )
. j. m. link _ et al . _
[ focus collaboration ] , phys .
lett . * 88 * , 041602 ( 2002 ) [ erratum - ibid . *
88 * , 159903 ( 2002 ) ] .
s. kopp _ et al .
_ [ cleo collaboration ] , phys . rev .
d * 63 * , 092001 ( 2001 ) .
d. m. asner _
[ cleo collaboration ] , submitted to phys .
d , arxiv : hep - ex/0311033 .
e. m. aitala _ et al .
_ [ e791 collaboration ] , phys . lett .
b * 403 * , 377 ( 1997 ) .
s. malvezzi , aip conf .
proc . * 549 * 569 ( 2002 ) .
g. brandenburg _ et al .
_ [ cleo collaboration ] , phys . rev .
lett . * 87 * , 071802 ( 2001 ) . v. v. frolov _ et al . _
[ cleo collaboration ] , arxiv : hep - ex/0306048 .
d. pedrini , _ int .
conf on frontier science _ , frascati ( italy ) , october 2002 .
j. m. link _ et al . _
[ focus collaboration ] , phys .
b * 491 * , 232 ( 2000 ) [ erratum - ibid .
b * 495 * , 443 ( 2000 ) ] .
s. e. csorna _ et al . _
[ cleo collaboration ] , phys . rev .
d * 65 * , 092001 ( 2002 ) .
b. aubert _ et al . _ [ babar collaboration ] , phys .
lett . * 91 * , 121801 ( 2003 ) .
g. bonvicini _ et al . _
[ cleo collaboration ] , phys .
d * 63 * , 071101 ( 2001 ) .
j. bartelt _ et al .
_ [ cleo collaboration ] , phys .
d * 52 * , 4860 ( 1995 ) .
r. a. briere _ et al .
_ , `` cleo c and cesr
c '' , clns-01 - 1742 ( 2001 ) . | searches for cp violation in the charm sector from the e791 , focus , cleo , babar and belle experiments are presented .
most analyses consider cp violation in two - body or quasi - two - body decays .
preliminary results from cleo and focus using dalitz - plot analyses are also presented .
= 11.6pt |
in a metric space @xmath2 the hausdorff dimension of a compact set @xmath3 is defined as the supremum of the @xmath4 such that the @xmath5-dimensional hausdorff measure @xmath6 of @xmath7 is infinite .
the hausdorff dimension takes values in the non - negative reals and extends the elementary integer - valued topological dimension in the sense that for a large class of ` reasonable ' sets these two values coincide .
sets with non - coinciding hausdorff and topological dimensions are called ` fractal ' , a term coined by mandelbrot in his original study of such sets @xcite .
hausdorff introduced this generalised dimension in @xcite and its subsequent extensive use in geometric measure theory is developed by federer @xcite and falconer @xcite .
the fact that the hausdorff dimension satisfies @xmath8 for the cartesian product of sets was proved in full generality in @xcite ( and later summarised in @xcite 7.1 ` product formulae ' ) after some partial results : the inequality was proved in @xcite with the restriction that @xmath9 for some @xmath10 and was extended to a larger class of sets in @xcite .
the paper @xcite also provides an example for which there is a strict inequality in the product formula and again this is summarised in @xcite 7.1 . [ [ section ] ] in this paper
we prove similar product inequalities for the upper and lower box - counting dimensions which are less familiar generalisations of dimension ( treated briefly in @xcite ; see @xcite for a more detailed exposition ) and have applications to dynamical systems ( see , for example , @xcite ) .
our main result is an example analogous to that in @xcite which demonstrates that the box - counting product inequalities can be strict . in a metric space @xmath2
the upper and lower box - counting dimensions of a compact set @xmath3 are defined by @xmath11 respectively , where @xmath12 is the smallest number of sets with diameter at most @xmath13 which form a cover ( called a @xmath13-cover ) of @xmath7 .
essentially , if @xmath12 scales like @xmath14 as @xmath15 then these quantities capture @xmath16 which gives an indication of how many more sets are required to cover @xmath7 as the length - scales decrease and so encodes how ` spread out ' the set @xmath7 is at small length - scales .
these limits are unchanged if we replace @xmath12 with one of many similar quantities ( discussed by falconer in @xcite 3.1 ` equivalent definitions ' ) . of these quantities
we will also make use of the largest number of disjoint closed balls of diameter @xmath13 with centres in @xmath7 , which we denote @xmath17 .
[ [ section-1 ] ] we first recall the standard ( see , for example , @xcite or @xcite ) proof of the box - counting product inequalities when @xmath7 and @xmath18 are compact sets in metric spaces @xmath2 and @xmath19 respectively , although the inequality is less familiar ( robinson provides a proof in @xcite ) .
we endow the product space @xmath20 with the usual euclidean metric @xmath21 , but the proof can be adapted for a variety of product metrics ( see @xcite ) .
[ product inequalities theorem ] for compact sets @xmath3 and @xmath22 the box - counting dimensions of the product set @xmath23 satisfy the inequalities @xmath24 suppose @xmath25 and @xmath26 are @xmath13-covers of @xmath7 and of @xmath18 respectively then the set of products @xmath27 is a cover of @xmath23 with a total of @xmath28 elements and the diameter of each @xmath29 is no greater than @xmath30 .
as there exist @xmath13-covers of @xmath7 and @xmath18 consisting of @xmath12 and @xmath31 elements respectively this construction gives a @xmath30-cover of @xmath23 consisting of @xmath32 elements , hence @xmath33 next , if both @xmath34 and @xmath35 are sets of centres of disjoint balls with diameter @xmath13 then the balls with radius @xmath13 centred on the @xmath28 points @xmath36 , are also disjoint .
as there exist sets of disjoint balls of diameter @xmath13 with centres in @xmath7 and @xmath18 consisting of @xmath17 and @xmath37 elements respectively the above construction gives @xmath38 disjoint balls of diameter @xmath13 with centres in @xmath23 , hence @xmath39 from and we see that @xmath40\label{limsup of sum}\\ & \leq \limsup_{\delta\searrow 0 } \frac{\log\left(n\left(f,\delta\right)\right)}{-\log\delta -\log\sqrt{2 } } + \limsup_{\delta\searrow 0 } \frac{\log\left(n\left(g,\delta\right)\right)}{-\log\delta -\log\sqrt{2}}\notag\\ & = \dim_{b}\left(f\right ) + \dim_{b}\left(g\right ) .
\notag\intertext{from \eqref{dimlb def } and \eqref{maximal disjoint inequality } we have } \dim_{lb}\left(f\times g\right ) & = \liminf_{\delta\searrow 0}\frac{\log\left(m\left(f\times g,\delta\right)\right)}{-\log\delta}\notag\\ & \geq \liminf_{\delta\searrow 0}\left[\frac{\log\left(m\left(f,\delta\right)\right)}{-\log\delta } + \frac{\log\left(m\left(g,\delta\right)\right)}{-\log\delta}\right]\label{liminf of sum}\\ & \geq \liminf_{\delta\searrow 0}\frac{\log\left(m\left(f,\delta\right)\right)}{-\log\delta } + \liminf_{\delta\searrow 0}\frac{\log\left(m\left(g,\delta\right)\right)}{-\log\delta } \notag \\ & = \dim_{lb}\left(f\right ) + \dim_{lb}\left(g\right ) . \notag\ ] ] it is known that there are sets with unequal upper and lower box - counting dimension ( see exercise 3.8 of @xcite or @xcite 3.1 ) , however if these values coincide for a set @xmath7 we define their common value as the box - counting dimension of @xmath7 .
if sets @xmath7 and @xmath18 have well - defined box - counting dimension then the box - counting dimension of their product is also well behaved . if @xmath41 and @xmath42 then @xmath43 from the inequalities and we have @xmath44 but from the definition of the box - counting dimension @xmath45
so we must have equality throughout the above .
[ [ section-2 ] ] in the following construction both sets @xmath7 and @xmath18 have non - coinciding upper and lower box - counting dimensions so that as @xmath15 the box - counting functions @xmath46 and @xmath47 oscillate between two values .
further , by ensuring that these functions oscillate with different phases ( see figure [ figure - product graph ] ) we can produce strict inequalities after and and so yield strict inequality in both product formulas , that is @xmath48 and @xmath18 constructed in the next section ( explicitly computed here for small @xmath13 ) oscillate between @xmath49 and @xmath50 . the @xmath51-axis is scaled so that the slow oscillation can be graphed , and this oscillation continues as @xmath52 , that is as @xmath15 .
the differing phases guarantee that the sum of these functions does nt approach either @xmath53 or @xmath54 . ] to this end we construct variations of the cantor middle - third set from the initial interval @xmath55 $ ] except at each stage we use one of the three generators @xmath56 or @xmath57 which remove the middle @xmath58 or @xmath59 of each interval respectively .
note that if we exclusively use the @xmath60 generator we produce the usual cantor middle - third set , which has lower and upper box - counting dimension @xmath49 and if we exclusively use the @xmath57 generator we produce a similar cantor set with lower and upper box - counting dimension @xmath50 . by switching generators at certain stages of our construction
we can cause @xmath46 to oscillate between these values , providing that we apply the generators a sufficiently large number of times .
[ [ section-3 ] ] to simplify notation , we choose a sequence of integers @xmath61 which increases sufficiently quickly that @xmath62
we construct two sets @xmath7 and @xmath18 using the following iterative procedure : for @xmath7 apply the following generator at the @xmath63 stage @xmath64 and for @xmath18 apply the following generator at the @xmath63 stage @xmath65 let @xmath66 be the number of times @xmath60 has been applied and @xmath67 the number of times @xmath57 has been applied in the construction @xmath7 by stage @xmath68 . with this notation @xmath69
has been applied @xmath70 times by stage @xmath68 .
similarly define @xmath71 and @xmath72 for the construction of @xmath18 .
clearly these functions are non - decreasing and we refrain from writing them explicitly except to note that @xmath73 and @xmath74 denote the sets at the @xmath75 stage of the construction of @xmath7 and @xmath18 by * @xmath76 , which consists of @xmath77 intervals of length @xmath78 and * @xmath79 , which consists of @xmath77 intervals of length @xmath80 so that @xmath7 and @xmath18 are defined by @xmath81 and @xmath82 .
note that for every @xmath68 the endpoints of the intervals in @xmath76 and @xmath79 are in @xmath7 and @xmath18 respectively as the generators only remove the middle of each interval .
[ n = m=2j prop ] for @xmath13 such that @xmath83 we have @xmath84 .
we refer to those @xmath13 in the range as length - scales corresponding to stage @xmath68 in the construction of @xmath7 .
clearly every @xmath85 is a length scale corresponding to exactly one stage @xmath86 and @xmath87 as @xmath15 .
we refer to length - scales corresponding to the construction of @xmath18 in an analogous fashion . for @xmath13 in the range
the obvious cover consisting of all intervals in @xmath76 gives @xmath88 .
the opposite inequality comes from the fact that a set with diameter @xmath13 in this range intersects at most one @xmath89 stage interval @xmath90 but can not cover both @xmath63 stage subintervals of @xmath90 ( see figure [ figure - 2jballsneeded ] ) so at least @xmath91 elements are needed to form a cover of @xmath7 . + next ,
@xmath13 in the range is less than the length of the intervals in @xmath92 so balls of diameter @xmath13 centred on the end points of the intervals of @xmath92 are disjoint and have centres in @xmath7 ( see figure [ figure - 2jdisjointballscover ] ) .
this gives two disjoint balls for each interval , hence @xmath93 .
for the opposite inequality suppose for a contradiction that @xmath94 , then at least one of the @xmath95 intervals in @xmath92 contains the centres of least three disjoint balls with centres in @xmath7 .
let @xmath90 be such an interval .
at the next stage of the construction @xmath90 is split into two sub - intervals , one of which contains the centres of at least two of these three disjoint balls .
however , this @xmath63 stage subinterval has length no greater than @xmath13 so two closed balls of diameter @xmath13 centred in this interval can not be disjoint ( see figure [ figure - 2jdisjointballsmax ] ) , which is a contradiction . and @xmath76 ( shown in black ) to illustrate that each set ( shown as grey ellipses ) with diameter @xmath96 ( i.e. less than the length of the intervals of @xmath92 ) can neither intersect two intervals of @xmath92 nor cover two intervals of @xmath76 . ]
( shown in black ) to illustrate that balls ( shown in grey ) with diameter @xmath96 ( i.e. less than the length of the intervals of @xmath92 ) with centres the endpoints of the intervals @xmath92 are disjoint , giving two disjoint balls for each interval of @xmath92 . ]
( shown in black ) which contains the centres of three balls ( shown in grey ) with diameter @xmath97 ( i.e. greater than the the length of the intervals of @xmath76 ) with centres in @xmath7 . as the centres are also contained in @xmath76 (
also shown in black ) at least one interval of @xmath76 contains the centres of two of these balls .
consequently , the distance between the centres is at most the length of an interval of @xmath76 which is at most @xmath13 which is the sum of the radii of the closed balls so the two balls are not disjoint . ]
adapting the argument we can prove a similar proposition for the set @xmath18 .
[ n = m=2j prop g ] for @xmath13 such that @xmath98 we have @xmath99 . by taking logarithms
we obtain the following more useful form of propositions [ n = m=2j prop ] and [ n = m=2j prop g ] @xmath100+f_{7}\left(j-1\right)\left[\log\left(7\right)-\log\left(5\right)\right ] + \left(j-1\right)\log\left(5\right)\\ < -\log\delta \leq f_{3}\left(j\right)\left[\log\left(3\right)-\log\left(5\right)\right]+f_{7}\left(j\right)\left[\log\left(7\right)-\log\left(5\right)\right ] + j\log\left(5\right)\\ \rightarrow \log\left(n\left(f,\delta\right)\right ) = \log\left(m\left(f,\delta\right)\right ) = j\log\left(2\right)\end{gathered}\ ] ] and @xmath101+g_{7}\left(j-1\right)\left[\log\left(7\right)-\log\left(5\right)\right ] + \left(j-1\right)\log\left(5\right)\\ < -\log\delta \leq g_{3}\left(j\right)\left[\log\left(3\right)-\log\left(5\right)\right]+g_{7}\left(j\right)\left[\log\left(7\right)-\log\left(5\right)\right ] + j\log\left(5\right)\\ \rightarrow \log\left(n\left(g,\delta\right)\right ) = \log\left(m\left(g,\delta\right)\right ) = j\log\left(2\right)\end{gathered}\ ] ] [ [ section-4 ] ] the essential feature of our construction is that the sets @xmath7 and @xmath18 at some length - scales look like the cantor middle - third set , while at other length - scales look like the cantor middle-@xmath102 set .
further , this ` local ' behaviour is maintained over sufficient length - scales that the box - counting limits of @xmath7 and @xmath18 at these length - scales approach the box - counting dimensions of the relevant cantor set , which we will establish in the following section .
we conclude this section by proving that at any length - scale the sets @xmath7 and @xmath18 do not both look like the middle - third set nor do they both look like the middle-@xmath102 set .
[ delta not in both length - scales upper ] no @xmath13 is both a length - scale corresponding to some stage in @xmath103 $ ] in the construction of @xmath7 for some @xmath104 and a length - scale corresponding to some stage in @xmath105 $ ] in the construction of @xmath18 for some @xmath106 . assume for a contradiction that @xmath13 is such a length - scale , that is @xmath107 and @xmath108 we first demonstrate that this could only hold if @xmath109 . from
and we have @xmath110 and @xmath111 respectively , which in turn yield @xmath112 and @xmath113 .
taking logarithms we get @xmath114 where the final inequalities follow from the growth property .
we conclude that @xmath115 and @xmath116 , which implies that @xmath117 . with this restriction ,
if @xmath13 is such a length - scale then the lower bound from and the upper bound from imply @xmath118 for clarity we suppress the argument of the functions and take logarithms which yields @xmath119\left(f_{3}-g_{3}\right ) + \left[\log\left(5\right)-\log\left(7\right)\right]\left(f_{7}-g_{7}\right ) + \log\left(5\right)\left(k_{6n+3}-k_{6n+2}\right ) < 0.\ ] ] the bounds in and give @xmath120 consequently we drop the positive first term of which implies @xmath121\left(k_{6n+3 } - k_{6n-2}\right ) + \log\left(5\right)\left(k_{6n+3}-k_{6n+2}\right ) & < 0 \intertext{which is that } \left[2\log\left(5\right)-\log\left(7\right)\right]k_{6n+3 } + \left[\log\left(7\right)-\log\left(5\right)\right]k_{6n-2 } - \log\left(5\right)k_{6n+2 } & < 0\end{aligned}\ ] ] after dropping the positive middle term and rearranging we get @xmath122}k_{6n+2 } \intertext{so that } k_{6n+3 } & < \frac{\log\left(7\right)}{\log\left(3\right)}k_{6n+2}\end{aligned}\ ] ] however , by condition the right hand side is less than @xmath123 giving the required contradiction .
consequently , at any length - scale the sets @xmath7 and @xmath18 do not both look like the cantor middle - third set .
this lemma gives the following useful corollary : [ subsequence corollary upper ] every sequence @xmath124 with @xmath125 either contains a subsequence @xmath126 with each @xmath127 corresponding to some stage @xmath128 $ ] in the construction of @xmath7 or contains a subsequence @xmath129 corresponding to some stage @xmath130 $ ] in the construction of @xmath18 .
if the sequence @xmath124 did not contain such a subsequence , then there is a @xmath131 such that each @xmath132 is neither a length - scale corresponding to some stage @xmath133 $ ] in the construction of @xmath7 nor a length - scale corresponding to some stage @xmath134 $ ] in the construction of @xmath18 .
consequently , @xmath135 is a length scale corresponding to a stage @xmath136 $ ] in the construction of @xmath7 and also a length - scale corresponding to a stage @xmath137 $ ] which , from lemma [ delta not in both length - scales upper ] , is contradictory .
the corresponding lemma that @xmath7 and
@xmath18 do not both look like the cantor middle-@xmath102 set at any length - scale and the corresponding corollary for subsequences are proved in a similar way .
[ delta not in both length - scales lower ] no @xmath13 is both a length - scale corresponding to some stage in @xmath138 $ ] in the construction of @xmath7 for some @xmath104 and a length - scale corresponding to some stage in @xmath139 $ ] in the construction of @xmath18 for some @xmath106 , [ subsequence corollary lower ] every sequence @xmath124 with @xmath125 either contains a subsequence @xmath126 with each @xmath127 corresponding to some stage @xmath140 $ ] in the construction of @xmath7 or contains a subsequence @xmath129 corresponding to some stage @xmath141 $ ] in the construction of @xmath18 .
in order to establish the box - counting dimensions of @xmath7 and @xmath18 we need the following proposition on the behaviour of the generator - counting functions at the limit : from we have @xmath145 which converges to 1 as @xmath146 by . by we have @xmath147 for @xmath148
so @xmath149 which converges to 0 as @xmath146 by .
similarly , @xmath150 so @xmath151 the remaining results follow analogously .
first , we show that @xmath153 .
writing @xmath86 for the stage associated with the length - scale @xmath13 we have from @xmath154 + f_{7}\left(j_{\delta}-1\right)\left[\log\left(7\right)-\log\left(5\right)\right ] + \left(j_{\delta}-1\right)\log\left(5\right)}\\ & \leq \frac{j_{\delta}\log\left(2\right)}{f_{3}\left(j_{\delta}-1\right)\left[\log\left(3\right)-\log\left(5\right)\right ] + \left(j_{\delta}-1\right)\log\left(5\right ) } \intertext{and as $ f_{3}\left(j_{\delta}-1\right ) < j_{\delta}-1 $ } & < \frac{j_{\delta}\log\left(2\right)}{\left(j_{\delta}-1\right)\left[\log\left(3\right)-\log\left(5\right)\right ] + \left(j_{\delta}-1\right)\log\left(5\right ) } = \frac{j_{\delta}\log\left(2\right)}{\left(j_{\delta}-1\right)\log\left(3\right)}\\\end{aligned}\ ] ] which converges to @xmath49 as @xmath15 .
similarly we can show @xmath155 next , if we take the sequence @xmath156 with @xmath157 + f_{7}\left(k_{6n+1}\right)\left[\log\left(7\right)-\log\left(5\right)\right]+ k_{6n+1}\log\left(5\right)\ ] ] we have from that @xmath158 .
consequently , @xmath159 + f_{7}\left(k_{6n+1}\right)\left[\log\left(7\right)-\log\left(5\right)\right]+ k_{6n+1}\log\left(5\right)}\\ & = \frac{\log\left(2\right)}{\frac{f_{3}\left(k_{6n+1}\right)}{k_{6n+1}}\left[\log\left(3\right)-\log\left(5\right)\right]+\frac{f_{7}\left(k_{6n+1}\right)}{k_{6n+1}}\left[\log\left(7\right)-\log\left(5\right)\right]+ \log\left(5\right)}\\ & \rightarrow \frac{\log\left(2\right)}{\log\left(3\right)-\log\left(5\right ) + 0 + \log\left(5\right ) } = \frac{\log\left(2\right)}{\log\left(3\right)}\end{aligned}\ ] ] as @xmath160 by the convergence results so that @xmath161 + a similar sequence gives the corresponding inequality for @xmath18 . for all @xmath163
the implication gives @xmath164 + f_{7}\left(j_{\delta}\right)\left[\log\left(7\right)-\log\left(5\right)\right]+j_{\delta}\log\left(5\right)}\\ & \geq \frac{j_{\delta}\log\left(2\right ) } { f_{7}\left(j_{\delta}\right)\left[\log\left(7\right)-\log\left(5\right)\right]+j_{\delta}\log\left(5\right)}\\ & > \frac{j_{\delta}\log\left(2\right ) } { j_{\delta}\left[\log\left(7\right)-\log\left(5\right)\right]+j_{\delta } \log\left(5\right)}=\frac{\log\left(2\right)}{\log\left(7\right)}\end{aligned}\ ] ] next , if we take the sequence @xmath156 with @xmath165 + f_{7}\left(k_{6n+2}\right)\left[\log\left(7\right)-\log\left(5\right)\right]+ k_{6n+2}\log\left(5\right)\ ] ] we have @xmath166 + f_{7}\left(k_{6n+2}\right)\left[\log\left(7\right)-\log\left(5\right)\right]+ k_{6n+2}\log\left(5\right)}\\ & = \frac{\log\left(2\right)}{\frac{f_{3}\left(k_{6n+2}\right)}{k_{6n+2}}\left[\log\left(3\right)-\log\left(5\right)\right]+\frac{f_{7}\left(k_{6n+2}\right)}{k_{6n+2}}\left[\log\left(7\right)-\log\left(5\right)\right]+ \log\left(5\right)}\\ & \rightarrow \frac{\log\left(2\right)}{0 + \log\left(7\right)-\log\left(5\right ) + \log\left(5\right ) } = \frac{\log\left(2\right)}{\log\left(7\right)}\end{aligned}\ ] ] as @xmath160 by the convergence results .
hence @xmath167 , and similarly @xmath168 .
whilst the above lemmas demonstrate that for @xmath7 and @xmath18 there are sequences of length - scales @xmath156 with @xmath170 equal to @xmath49 or equal to @xmath50 we now show that for a large class of sequences ( in fact the very sequences that corollaries [ subsequence corollary upper ] and [ subsequence corollary lower ] produce ) this limit , if it exists , is bounded by @xmath171 .
[ deltan n(f , deltan ) converge to log2log5 ] suppose @xmath156 is a sequence such that each length - scale @xmath172 corresponds to the construction of @xmath7 at some stage @xmath173 $ ] , then if the limit exists @xmath174 essentially , these stages are sufficiently far from the range @xmath175 $ ] where @xmath60 is applied so that the set @xmath7 does not look like the cantor middle - third set at these stages .
the proof relies on the fact that by stage @xmath176 the generator @xmath60 has not been applied for at least the last @xmath177 stages . for each @xmath104 from @xmath178+f_{7}\left(j_{n}-1\right)\left[\log\left(7\right)-\log\left(5\right)\right ]
+ \left(j_{n}-1\right)\log\left(5\right)}\\ & \leq \frac{j_{n}\log\left(2\right)}{f_{3}\left(j_{n}-1\right)\left[\log\left(3\right)-\log\left(5\right)\right]+ \left(j_{n}-1\right)\log\left(5\right ) } \intertext{from \eqref{f constant on intervals } we have $ f_{3}\left(j_{n}-1\right)=f_{3}\left(k_{6n+1}\right)$ so that } & = \frac{j_{n}\log\left(2\right)}{f_{3}\left(k_{6n+1}\right)\left[\log\left(3\right)-\log\left(5\right)\right]+ \left(j_{n}-1\right)\log\left(5\right)}\\ & = \frac{\log\left(2\right)}{\frac{f\left(k_{6n+1}\right)}{j_{n}}\left[\log\left(3\right)-\log\left(5\right)\right]+\log\left(5\right)-\frac{1}{j_{n}}\log\left(5\right)}. \intertext{next , as $ j_{n}>k_{6n+2}$ } & \leq \frac{\log\left(2\right)}{\frac{f_{3}\left(k_{6n+1}\right)}{k_{6n+2}}\left[\log\left(3\right)-\log\left(5\right)\right]+\log\left(5\right)-\frac{1}{j_{n}}\log\left(5\right)}\\ & \rightarrow \frac{\log\left(2\right)}{\log\left(5\right)}\end{aligned}\ ] ] as @xmath160 by the convergence result .
[ deltan n(g , deltan ) converge to log2log5 ] suppose @xmath179 is a sequence such that each length - scale @xmath180 corresponds to the construction of @xmath18 at some stage @xmath181 $ ] , then if the limit exists @xmath182 [ deltan m(f , deltan ) converge to log2log5 ] suppose @xmath156 is a sequence such that each length - scale @xmath172 corresponds to the construction of @xmath7 at some stage @xmath183 $ ] , then if the limit exists @xmath184 [ deltan m(g , deltan ) converge to log2log5 ] suppose @xmath179 is a sequence such that each length - scale @xmath180 corresponds to the construction of @xmath18 at some stage @xmath185 $ ] , then if the limit exists @xmath186 we have from that @xmath188\ ] ] so it is sufficient to show that the right hand side is no greater than @xmath189 .
suppose that @xmath124 is a sequence with @xmath125 such that the limits @xmath190 and @xmath191 exist .
corollary [ subsequence corollary upper ] guarantees that this sequence either contains a subsequence @xmath126 satisfying the hypothesis of lemma [ deltan n(f , deltan ) converge to log2log5 ] or contains a subsequence @xmath129 satisfying the hypothesis of lemma [ deltan n(g , deltan ) converge to log2log5 ] so at least one of @xmath192 and @xmath193 converges to @xmath171 . using the upper box - counting dimension from lemma [ upper box lemma ] to bind the other term yields @xmath194 a bound which also hold for the original sequence @xmath124 . as @xmath124 was an arbitrary convergent sequence , @xmath195 \leq \frac{\log\left(2\right)}{\log\left(3\right ) } + \frac{\log\left(2\right)}{\log\left(5\right)}\ ] ] from we have @xmath198\ ] ] so it is sufficient to prove that the right hand side is no less than @xmath199 .
suppose that @xmath124 is a sequence with @xmath125 such that the limits @xmath190 and @xmath191 exist . in a similar fashion
to theorem [ strict upper box product inequality ] , corollary [ subsequence corollary lower ] and lemmas [ deltan m(f , deltan ) converge to log2log5 ] and [ deltan m(g , deltan ) converge to log2log5 ] guarantee that at least one of @xmath200 and @xmath201 is no less than @xmath171 so @xmath202 \geq \frac{\log\left(2\right)}{\log\left(5\right ) } + \frac{\log\left(2\right)}{\log\left(7\right)}\ ] ] [ ackref ] i am greatly indebted to my phd supervisor , james robinson , for highlighting that there was no example of a strict product inequality in the box - counting literature and to the generosity of isabelle harding and matt gibson in providing accommodation whilst much of this work was completed . 9 , `
the measure of product and cylinder sets ' _ j. lond .
_ 20 ( 1945 ) 110120 . ,
` a correction to a paper on the dimension of cartesian product sets ' _ proc .
_ 49 ( 1953 ) 437440 . ,
_ the geometry of fractal sets _ ( cambridge university press , 1985 ) . , _ fractal geometry _
( john wiley & sons , england , 2003 ) . , _ geometric measure theory _ ( springer 1969 ) . ,
` dimension und ueres ma ' , _ mathematische annalen _ 79 ( 1918 ) 157179 . ,
` the dimension of cartesian product sets ' , _ proc . cam .
_ 50 ( 1954 ) 198202 . ,
les objets fractals - forme , hasard et dimension _ , ( flammarion , paris , 1975 ) , _ infinite dimensional dynamical systems _ ( cambridge university press , 2001 ) . , _ dimensions , embeddings , and attractors ( in press ) _ ( cambridge university press , 2010 ) . | it is known that the upper box - counting dimension of a cartesian product satisfies the inequality @xmath0 whilst the lower box - counting dimension satisfies the inequality @xmath1 .
we construct cantor - like sets to demonstrate that both of these inequalities can be strict . |
the observation of the cutoff in the spectrum of ultra - high energy cosmic rays ( uhecrs ) @xcite as predicted by @xcite provides compelling evidence for the shortening of the uhecr propagation length at high energies .
the highest energy events then must have come from relatively close sources ( within @xmath2 mpc ) . at these
length scales the matter in the universe is distributed inhomogeneously , being organized into clusters and superclusters .
one should , therefore , expect the flux of highest - energy cosmic rays to be anisotropic . in astrophysical scenarios , it is natural to assume that the number of sources within @xmath2 mpc is large , and that these sources trace the distribution of matter . under these assumptions ,
the anisotropy at earth depends only on the nature and size of uhecr deflections .
measurement of the anisotropy , therefore , gives direct experimental access to parameters that determine the deflections , notably to the uhecr charge composition and cosmic magnetic fields .
several investigations of anisotropy in arrival directions of uhecrs have been previously undertaken . at small angular scales ,
correlations with different classes of putative sources were claimed ( e.g. @xcite ) . at larger angular scales and energies below 10 eev possible
anisotropy towards the galactic center was reported in @xcite , but not supported by more recent studies @xcite . at higher energies
, @xcite found evidence against an isotropic flux above 40 eev through correlations with the supergalactic plane , but this was not confirmed by other authors @xcite .
finally , using the pierre auger observatory ( pao ) data , @xcite have found correlations between uhecr arrival directions and the large - scale structure of the universe which are incompatible with an isotropic flux ( see , however , @xcite ) . in this paper , we analyze the data accumulated by the hires experiment for anisotropy associated with the large - scale structure of the universe .
the hires experiment has been described previously @xcite .
it studied ultrahigh energy cosmic rays from @xmath3 ev to @xmath4 ev using the fluorescence technique .
hires operated two fluorescence detectors located atop desert mountains in west - central utah .
the data set used in this study consists of events observed by both detectors , analyzed jointly in what is commonly called `` stereo mode '' . in this mode
the angular resolution in cosmic rays pointing directions is about @xmath5 degrees , and the energy resolution is about 10% .
the hires experiment operated in stereo mode between december , 1999 , and april , 2006 . at the highest energies hires has the largest data set in the northern hemisphere .
large number of events , good angular resolution ( better than the bending angles expected from galactic and extragalactic magnetic fields ) and the wide energy range covered make the hires data particularly suitable for anisotropy studies .
the exact data set used in this study was described previously in @xcite .
we consider here a generic model that assumes many sources within @xmath2 mpc tracing the distribution of matter , which we refer to as the `` matter tracer '' model .
we also assume that deflections of uhecr do not exceed the angular size of the nearby structures , that is 10 - 20@xmath6 . in this regime , both regular and random deflections in magnetic fields
can be modeled with a one - parameter distribution , for which we take a gaussian distribution centered at zero with width @xmath7 .
this width is treated as a free parameter , whose value we aim to constrain from the data .
constraints on @xmath8 may then be used to obtain information on the strength of galactic and extragalactic magnetic fields . in keeping with our assumption of small deflections , we assume a proton composition in this study , which is consistent with the @xmath9 analysis based on the same dataset ( for confirmation see @xcite ) .
the hires data is compared to model predictions using the `` flux sampling '' test put forward by @xcite .
this test has good discrimination power at small statistics and is insensitive to the details of deflections .
the comparison is performed at three different threshold energies that have been used in previous studies : 10 eev , 40 eev , and 57 eev @xcite .
an _ a priori _ significance of 5% , corresponding to a confidence level ( cl ) of 95% , is chosen for this work .
the paper is organized as follows . in section [ section : modeling ]
we discuss the modeling of uhecr arrival directions .
section [ section : data ] concerns the hires data used in the analysis , while section [ section : fluxsampling ] describes the flux sampling method .
we present our results in section [ sec : results ] and conclude in section [ sec : conclusions ] .
_ galaxy catalog _ the distribution of matter in the local universe is modeled with the 2 micron all - sky redshift survey ( 2mrs ; @xcite ) galaxy catalog , using galaxies as samplers of the underlying matter density field .
the 2mrs is a flux - limited sample of galaxies , that is , the sample containing all galaxies with observed magnitude in the @xmath10 band @xmath11 .
it contains spectroscopically measured redshifts for all but a few galaxies .
a number of cuts have been applied to the galaxy sample .
first , the galactic plane , where the sample is incomplete , has been excluded from the sample by removing all galaxies with @xmath12 .
second , objects with @xmath13 mpc are removed because such objects should be treated on an individual basis . finally , the catalog is cut at 250 mpc because the sample becomes too sparse .
the resulting sample provides an accurate statistical description at smearing angles @xmath14 . the flux from sources beyond 250 mpc is taken to be isotropic .
a total of 15508 galaxies remain in the hires field of view after the cuts . to compensate for observational selection effects in the ( flux - limited ) 2mrs catalog , weights @xmath15
are assigned to the galaxies with the sliding - box method as described in @xcite .
_ energy loss _ uhecr fluxes are affected by energy loss due to redshift and interactions with the cosmic microwave background ( cmb ) . to account for the resulting flux suppression ,
the integral flux , @xmath16 , from a single source is expressed as follows : @xmath17 where @xmath18 is the threshold energy , @xmath19 is the source distance , @xmath20 stands for the integral injection spectrum at the source , and @xmath21 represents the flux fraction that remains after interactions and redshift .
we take an injection spectrum @xmath22 extending to very high energies .
the function @xmath21 is determined using a numerical propagation code based on the continuous loss approximation that is described in @xcite .
energy loss due to interactions with the extragalactic background light is neglected . in figure
[ fig : ffunc ] , top panel , the fraction @xmath21 is shown as a function of distance for the different energies considered in this work .
the strength of the isotropic flux component that is added to account for sources beyond 250 mpc also depends on uhecr energy loss . using the computer code described in the previous paragraph
, we estimate the fraction @xmath23 of total flux contributed by sources within 250 mpc to be 0.4 , 0.7 , and 1.0 for threshold energies @xmath24 eev , 40 eev , and 57 eev , respectively ( see figure 1 , bottom panel ) .
_ deflections _ uhecr protons ( as well as nuclei ) are deflected by galactic and intergalactic magnetic fields .
these deflections are taken into account by an angular smearing procedure , which replaces the point - source flux , @xmath25 , by a flux density distribution : @xmath26 where @xmath27 represents the probability density that an uhecr is deflected by @xmath28 , the angle between galaxy @xmath29 and the line of sight .
this procedure also accounts for the detector s angular resolution and prevents unphysical fluctuations due to the use of a catalog of point sources . in the absence of detailed knowledge on the structure of galactic and extragalactic magnetic fields ,
we adopt a simple gaussian probability density distribution with characteristic angle , @xmath7 .
this angle is treated as a free model parameter .
the gaussian distribution is a fair approximation when the deflections are small . for large deflections , details on the structure of the galactic and extragalactic magnetic fields
become important .
accounting for these details goes beyond the scope of the present study .
_ exposure _ the hires exposure is modeled using our monte carlo simulation of the experiment @xcite .
the aim of this simulation was to create a set of monte carlo events that would be in all essences identical to the actual data . in making the simulation we put in the properties of cosmic ray air showers as measured by previous experiments @xcite .
we used cosmic ray showers generated by the corsika and qgsjet programs @xcite and simulated the generation of fluorescence light ( see references in @xcite ) and its propagation through the atmosphere ( see references in @xcite ) .
a complete simulation of the optics and electronics ( trigger and front - end electronics ) of our detectors was performed .
the result was an excellent simulation of our experiment as evidenced by the very good agreement between data and simulated events in the distribution of all kinematic variables , e.g. zenith angle , impact parameter to detector , etc . by assigning monte carlo events times of occurrence taken from the actual on - time of the experiment
we are able to calculate the exposure on the sky very accurately .
_ model flux maps _ the integral uhecr flux from a given direction is expressed as follows : @xmath30 where @xmath29 enumerates galaxies in the 2mrs sample , @xmath31 denotes the weight assigned to galaxy @xmath29 in the catalog , @xmath32 is the deflection probability distribution , and @xmath33 is the uhecr flux arising from sources beyond 250 mpc . the probability to observe a cr from a given direction is proportional to the product of flux and exposure .
we denote this probability as @xmath34 where @xmath35 stands for exposure . in figure
[ fig : skymap : struct ] , the distribution of @xmath36 over the sky is shown for three different threshold energies .
the contrast in the flux distributions becomes more pronounced with increasing energy .
also shown are the arrival directions of uhecrs in the hires data to which the model flux has to be compared .
hammer projection ( galactic coordinates ) of @xmath36 ( flux times exposure ) with threshold energies 10 eev ( top panel ) , 40 eev ( middle ) , and 57 eev ( bottom ) .
darker gray indicates a higher value ; the bands are chosen such that each band contains 1/5 of the total flux ( weighted with exposure ) .
excluded regions , viz . the galactic plane ( @xmath12 ) and the region outside the hires field of view , are shown in white .
white dots indicate hires events .
all maps are produced with @xmath37.,width=302 ] hammer projection ( galactic coordinates ) of @xmath36 ( flux times exposure ) with threshold energies 10 eev ( top panel ) , 40 eev ( middle ) , and 57 eev ( bottom ) .
darker gray indicates a higher value ; the bands are chosen such that each band contains 1/5 of the total flux ( weighted with exposure ) .
excluded regions , viz . the galactic plane ( @xmath12 ) and the region outside the hires field of view ,
are shown in white .
white dots indicate hires events .
all maps are produced with @xmath37.,width=302 ] hammer projection ( galactic coordinates ) of @xmath36 ( flux times exposure ) with threshold energies 10 eev ( top panel ) , 40 eev ( middle ) , and 57 eev ( bottom ) .
darker gray indicates a higher value ; the bands are chosen such that each band contains 1/5 of the total flux ( weighted with exposure ) .
excluded regions , viz .
the galactic plane ( @xmath12 ) and the region outside the hires field of view , are shown in white .
white dots indicate hires events .
all maps are produced with @xmath37.,width=302 ]
the data set used in this study was described previously in @xcite , including selection criteria and a correction to the energy scale .
our sample of the 2mrs catalog does not cover the region near the galactic plane with @xmath12 .
we therefore removed cosmic ray events with @xmath12 from the analysis .
the resulting sample contains : * 309 events with @xmath38 eev ; * 27 events with @xmath39 eev ; * 10 events with @xmath40 eev .
the arrival directions of these events are shown as white dots in fig .
[ fig : skymap : struct ] .
the compatibility of a model flux map with the set of uhecr arrival directions is quantified by the flux sampling method introduced by @xcite .
the idea of the method is as follows . to any set of arrival directions one
associates the set of flux values that are obtained by sampling a given flux map ( such as the map shown in fig .
[ fig : skymap : struct ] ) , i.e. by extracting the flux values at corresponding points on the sphere .
the two - dimensional distribution of arrival directions thus translates into a one - dimensional distribution of flux values .
if the reference model is true , this flux distribution will tend to high values since events fall preferentially into regions where the model flux ( times exposure ) is high .
if , on the other hand , the reference model is not true , the flux distribution is more uniform because the correlation between arrival directions and regions of high model flux is ( partly ) destroyed . by comparing the flux distribution to a model flux distribution , the compatibility between a set of arrival directions and model predictions can be quantified .
this comparison is performed by the kolmogorov - smirnov ( ks ) test , which yields a test statistic @xmath41 .
the relevant statistical quantities , in particular powers and @xmath42-values , are computed from the distribution of this test statistic .
note that this test does not involve any additional parameters like bin size .
the ability of the test to discriminate between models may be quantified in terms of the statistical power , @xmath43 , i.e. the probability to rule out , at a given confidence level , the reference model when an alternative model is true .
within numerical uncertainties , the statistical power is equal to the fraction of event sets generated under the alternative model that leads to rejecting the reference model .
figure [ fig : power - nvsths ] shows the number of events required for a power @xmath44 ( i.e. , a 50% probability ) to rule out ( at 95% cl ) the matter tracer model when the true flux is isotropic .
the number of events increases with increasing smearing angle and decreasing energy : the decreasing flux contrasts in the matter tracer model call for an increase in statistics to achieve the same discriminatory power .
observe that the event numbers indicated in figure [ fig : power - nvsths ] are of the same order as the data analyzed in this work .
we thus expect that there is sufficient data to obtain meaningful constraints at 95% cl .
the level of compatibility between data and model predictions is quantified by a @xmath42-value , which represents the model probability of obtaining a measurement that is at least as extreme as the actual measurement . with our _ a priori _ choice of significance , a @xmath42-value smaller than @xmath45 rules out the model .
the probability of falsely excluding the model is then @xmath46% , translating to a cl of @xmath47% .
figure [ fig : pvalues ] shows the @xmath42-values obtained by the flux sampling method for the hires data and predictions of the matter tracer model .
the smearing angle , @xmath7 , is treated as a free parameter .
that is , at each value of @xmath7 and each threshold energy a flux map is generated and compared to the hires data as described above .
the results can be summarized as follows : * for the threshold energies of 40 eev and 57 eev , the tests show disagreement between data and the matter tracer model for @xmath48 . within this parameter range ,
a source distribution tracing the distribution of matter is excluded at a 95% confidence level .
* for the threshold energy of 10 eev , the test shows agreement between data and the matter tracer model .
the incompatibility between data and matter tracer model is illustrated by the non - correlation between the observed arrival directions and regions of high model flux shown in the two lower panels of figure [ fig : skymap : struct ] .
we have also tested the data for compatibility with an isotropic model flux and found no disagreement , at 95% cl , for any of the three tested threshold energies ( the data with threshold energy 57 eev are marginally consistent with an isotropic flux ) . at energy threshold @xmath51 eev a correlation between the arrival directions of uhecrs and the location of agns contained in the 12@xmath52 edition of the vron - cetty & vron catalog
was reported by the pao @xcite .
this correlation was found to be maximal for @xmath53 , where @xmath54 denotes the maximum angular distance between uhecrs and agns . in the northern hemisphere ,
correlation with agn was not confirmed by the hires experiment @xcite . since agns are tracers of the distribution of matter in the universe , the pao result is suggestive of a more general correlation between uhecrs and the local structure of the universe on an angular scale of a few degrees .
the methods presented in this paper allow a check on the existence of such correlations in the hires data .
the results presented in the previous section disfavor a correlation between uhecrs and the local structure of the universe . in fact , the flux sampling test yields @xmath42-values smaller than @xmath55 for the matter tracer model with @xmath56 , with a @xmath42-value of @xmath57 for @xmath58 .
( note that @xmath7 is not in @xmath59 correspondence with @xmath54 ; both quantities are however representative of the angular scale of the problem ) .
focusing on the case of @xmath60 in more detail , figure [ fig : ddist ] shows the distribution of the test statistic @xmath41 for the matter tracer model and for an isotropic flux for this smearing angle .
the vertical line shows the value @xmath61 obtained for the hires data .
this demonstrates the strong incompatiblity between hires data and the matter tracer model for smearing angle @xmath58 and threshold energy @xmath62 eev .
to summarize , we have confronted the stereo data collected by the hires experiment with predictions of the matter tracer model , a generic model of cosmic ray origin and propagation .
the model assumes a large number of cosmic ray sources within @xmath2 mpc whose distribution traces that of matter , and relatively small deflections characterized by a single parameter , the typical deflection angle @xmath0 .
we have found that the hires data with energy thresholds @xmath63 eev and @xmath62 eev are incompatible with the matter tracer model for @xmath64 at 95% cl . with an energy threshold @xmath24
eev the hires data are compatible with the matter tracer model .
at all three energy thresholds , the data are compatible with an isotropic flux at 95% cl . in the present analysis
we have treated the deflections as random and gaussian , which is only appropriate for small deflection angles and limited number of events .
the actual deflections are expected to contain a coherent component due to the galactic magnetic field . with the accumulation of uhecr events by pao in the southern hemisphere and by telescope array in the northern hemisphere ,
our analysis will become sensitive to the nature of deflections and thus , with proper modifications of the statistical procedure , may give direct access to the parameters of cosmic magnetic fields .
this work is supported by the national science foundation under contracts nsf - phy-9321949 , nsf - phy-9322298 , nsf - phy-9974537 , nsf - phy-0071069 , nsf - phy-0098826 , nsf - phy-0140688 , nsf - phy-0245328 , nsf - phy-0307098 , and nsf - phy-0305516 , by department of energy grant fg03 - 92er40732 , by the bsp under iuap vi/11 , by the fnrs contract 1.5.335.08 and by the iisn contract 4.4509.10 .
we gratefully acknowledge the contribution from the technical staffs of our home institutions and thank the university of utah center for high performance computing for their contributions .
the cooperation of colonels e. fisher , g. harter , and g. olsen , the us army and the dugway proving ground staff is appreciated . | stereo data collected by the hires experiment over a six year period are examined for large - scale anisotropy related to the inhomogeneous distribution of matter in the nearby universe .
we consider the generic case of small cosmic - ray deflections and a large number of sources tracing the matter distribution . in this matter tracer model
the expected cosmic ray flux depends essentially on a single free parameter , the typical deflection angle @xmath0 .
we find that the hires data with threshold energies of 40 eev and 57 eev are incompatible with the matter tracer model at a 95% confidence level unless @xmath1 and are compatible with an isotropic flux .
the data set above 10 eev is compatible with both the matter tracer model and an isotropic flux . |
throughout this paper we work in the piecewise linear category .
we shall study knots and links in the three - dimensional euclidean space @xmath0 . for the standard definitions and results of knots and links
, we refer to @xcite , @xcite , @xcite , @xcite , @xcite , @xcite , @xcite and @xcite . for the purpose of defining a numerical invariant of knots and links ,
we need to prepare following terms .
a link diagram is _ ordered _ if an order is given to its components .
a link diagram is _ based _ if a basepoint ( different from the crossing points ) is specified on each component .
a link diagram is _ oriented _ if an orientation is specified on each component .
let @xmath1 be a link , and let @xmath2 be a based ordered oriented link diagram of @xmath1 .
the _ descending diagram _ of @xmath2 is obtained as follows , and will be denoted by @xmath3 .
beginning at the basepoint of the first component of @xmath2 and proceeding in the direction specified by the orientation , change the crossings as necessary so that each crossing is first encountered as an over - crossing .
continue this procedure with the remaining components in the sequence determined by the ordering , proceeding from the basepoint in the direction determined by the orientation , changing crossings so that ultimately every crossing is first encountered as an over - crossing .
the result is the descending diagram @xmath3 obtained from @xmath2 .
an example is shown in figure [ example1 ] .
note that @xmath3 is a diagram of a trivial link . | we introduce a new numerical invariant of knots and links from the descending diagrams . it is considered to live between the unknotting number and the bridge number . |
let @xmath7 be the set of permutations of @xmath8=\{1,\ldots , n\}$ ] . given a list of @xmath9 distinct integers @xmath10 ,
the _ standardization _ of @xmath11 , written @xmath12 , is the unique permutation of @xmath13 $ ] that is order - isomorphic to @xmath11 .
we obtain @xmath12 by replacing the smallest element among @xmath14 with @xmath15 , the second smallest with @xmath16 , and so on .
we say that @xmath17 _ contains @xmath18 consecutively _ if there exists an index @xmath19 $ ] such that @xmath20 ; otherwise , we say that @xmath1 _ avoids @xmath5 consecutively_. in this context , we call @xmath5 a _ pattern _ and we write @xmath21 to denote the set of permutations in @xmath7 that avoid @xmath5 as a consecutive pattern .
the systematic study of consecutive patterns originated in @xcite and has expanded in various directions since .
although much of the literature on pattern avoidance deals with classical pattern avoidance , whereby @xmath17 avoids @xmath18 only if @xmath22 for all length @xmath9 subsequences @xmath23 , consecutive pattern avoidance has received significant attention in the last 15 years ; see @xcite for a survey .
except for some references in the introduction , we speak only of consecutive pattern avoidance in this paper , even if not explicitly stated . in this paper
we connect prior work on consecutive pattern avoidance to ongoing developments at the interface of pattern avoidance and random permutations .
classical enumerative studies of pattern avoidance admit a straightforward probabilistic interpretation , as the problem of enumerating permutations avoiding a consecutive pattern @xmath18 is equivalent to finding the probability that a uniform random permutation avoids @xmath5 .
this observation makes a variety of combinatorial and probabilistic techniques available for studying pattern avoidance .
so far , these connections have been mostly confined to the equivalence between enumeration and uniform random permutations .
below we bring tools from analytic combinatorics and probability theory to bear on the behavior of pattern avoidance for a widely studied class of non - uniform random permutations . a straightforward and natural way to generate
a class of non - uniform distributions is by exponential tilting of the uniform distribution with respect to some statistic @xmath24 on the space of permutations . in this generic setting , we assign probability @xmath25 to each @xmath26 .
any such model specializes to the uniform distribution upon setting @xmath27 .
for example , taking @xmath28 equal to the number of cycles in @xmath1 yields the well known ewens sampling formula @xcite .
previous authors @xcite consider certain questions of pattern avoidance , such as analogs of wilf equivalence , in a setting equivalent to .
owing to its natural importance in the context of ranking statistics , we study pattern avoidance probabilities under the setting of with @xmath28 taken to be the number of inversions in @xmath1 . an _ inversion _ in @xmath29 is a pair of indices @xmath30 with @xmath31 and @xmath32 . for a real - valued parameter @xmath4 ,
the _ mallows distribution _ with parameter @xmath0 on @xmath7sometimes denoted mallows(@xmath0)assigns probability @xmath33_q!}\ ] ] to each @xmath17 , where @xmath3 is the number of inversions in @xmath1 , @xmath8_q=1+q+\dots+q^{n-1}$ ] , @xmath8_q!=[1]_q [ 2]_q \dots [ n]_q$ ] , and @xmath34_q!}{[m]_q![n]_q!}$ ] are the usual @xmath0-analogues with @xmath0 treated as a real - valued parameter .
mallows s distribution @xcite is a canonical statistical model for ranking data , with many useful statistical properties stemming from its exponential family structure endowed from and other probabilistic properties ; see , e.g. , ( * ? ? ?
* chapters 6 and 7 ) for more discussion of its statistical applications .
more recent work on the mallows distribution reveals connections to a range of topical problems in statistical physics @xcite , markov chain monte carlo @xcite , the longest increasing subsequence problem @xcite , cycle lengths in random permutations @xcite , and quasi - exchangeable sequences @xcite .
others have recently studied characteristics of classical pattern avoidance for uniform random permutations @xcite , with the former two references tying characteristics of certain pattern avoiding permutations to brownian excursion , a fundamental probabilistic object .
perarnau @xcite uses probabilistic techniques to give an alternate proof of the fact that , among consecutive patterns of a given length , the monotone pattern is avoided by the largest number of permutations of length @xmath35 , for sufficiently large @xmath35 . a proof of this fact using singularity analysis of generating function was previously given in @xcite . in @xcite ,
two of the authors used the technique of poisson approximation to estimate pattern avoidance probabilities for sufficiently large patterns in random permutations from the mallows distribution .
the method in @xcite breaks down for fixed patterns of small size . our analysis below builds on these other recent developments at the interface of probability and combinatorics in several distinct ways .
much of the contents of sections [ section : gj]-[section : compare ] extend the techniques from singularity analysis and analytic combinatorics developed in @xcite . in section [ section : bounds ] , we use probabilistic techniques to obtain bounds on the growth rate for arbitrary patterns , generalizing the outcomes in @xcite . in section [ section : stein ] , we conclude our discussion by studying complementary behavior of pattern avoidance in mallows permutations using the probabilistic technique of stein s method . whereas the techniques of sections [ section : gj]-[section : bounds ] zero in on the growth rate of the probability of pattern avoidance , our analysis in section [ section : stein ] considers the asymptotic distribution of the number of occurrences of a given pattern in a random permutation from the mallows distribution , which also gives the rate of convergence to the gaussian distribution , providing a complementary perspective to the results of sections [ section : growth]-[section : compare ] .
the results of section [ section : stein ] are in the same vein as other work of @xcite , who studied the asymptotic distribution of the number of occurrences of patterns in uniform random permutations , and janson @xcite , who studied the behavior of random permutations drawn uniformly from the set of 132-avoiding permutations . for @xmath4 and @xmath18 , we write @xmath36 to denote the probability that a random permutation from distribution avoids @xmath5 and @xmath37 to denote the corresponding generating function . by definition @xmath38_q!},\ ] ] so that @xmath39_q!}\ ] ] is the @xmath0-exponential generating function for @xmath5-avoiding permutations with respect to the inversion number . in fact , the same argument works for any family of subsets @xmath40 : the ordinary generating function for the probability that a permutation in @xmath7 belongs to @xmath41 equals the @xmath0-exponential generating function for permutations in @xmath41 with respect to the inversion number . the definition of @xmath42 can be generalized in order to consider not only permutations that avoid @xmath5 but also all permutations with respect to the number of occurrences of @xmath5 .
let @xmath43 denote the number of occurrences of @xmath5 in @xmath1 as a consecutive pattern , and let @xmath44_q!}\ ] ] so that @xmath45 .
the mallows distribution possesses several nice properties that are amenable to the study of pattern avoidance .
one immediately useful observation is that the probability of a permutation @xmath17 under the mallows distribution with parameter @xmath46 equals @xmath47_{q^{-1}}!}=\frac{q^{\binom{n}{2}-\operatorname{inv}(\pi)}}{[n]_{q}!}=\frac{q^{\operatorname{inv}(\pi^r)}}{[n]_{q}!},\ ] ] that is , the probability of the permutation @xmath48 under the mallows distribution with parameter @xmath0 .
the same holds by replacing @xmath49 with @xmath50 ; whence , @xmath51 two other crucial properties of the mallows distribution , whose combinatorial equivalent is given in lemma [ lem : product ] , are weak dissociation and consecutive homogeneity .
weak dissociation says that for any non - overlapping sets of indices @xmath52 and @xmath53 , i.e. , @xmath54 , the probability that a given pattern of size @xmath55 occurs at the indices in @xmath56 is independent of the probability that any other pattern of size @xmath57 occurs at the indices of @xmath58 .
consecutive homogeneity says that for @xmath59 from the mallows(@xmath60 distribution on @xmath7 , @xmath61 , and any @xmath18 , the pattern @xmath62 is distributed according to mallows(@xmath0 ) on @xmath63 .
these two properties appear repeatedly throughout our analysis below .
we organize the rest of the paper as follows . in section [ section : gj ]
, we generalize the cluster method of goulden and jackson @xcite to keep track of the inversion number . in section [
section : growth ] , we show that the growth rate , defined as @xmath64 , exists for all patterns @xmath5 and all @xmath4 . in section [ section : monotone ] , we apply the method of section [ section : gj ] to approximate the growth rate of pattern avoidance probabilities for monotone patterns ( with a precise description for certain values of @xmath0 ) . in section [ section:132 ]
, we treat the pattern @xmath65 , for which we can determine the growth rate in all cases , and generalize the argument to all non - overlapping patterns starting with a 1 . in section [ section :
compare ] , we compare the growth rates between the different patterns of length 3 . in section [ section : bounds ] , we prove upper and lower bounds for the growth rate of arbitrary patterns . in section [ section : stein ] , we apply stein s method to @xmath9-dependent random variables to obtain a central limit theorem governing the number of occurrences of a given pattern , with explicit error rates .
the proof of the following lemma uses the well - known fact that if @xmath66 is the set of words consisting of @xmath9 zeros and @xmath35 ones , then @xmath67 [ lem : product ] let @xmath68 , @xmath69 , and @xmath70 then @xmath71_q!}\right ) \left(\sum_{\pi\in b } \frac{q^{\operatorname{inv}(\pi)}}{[n]_q!}\right ) = \sum_{\pi\in c } \frac{q^{\operatorname{inv}(\pi)}}{[m+n]_q!}.\ ] ] given @xmath72 and @xmath73 , there are @xmath74 permutations @xmath75 such that @xmath76 and @xmath77 , obtained by choosing which @xmath9 of the values @xmath78 will be placed in the first @xmath9 entries . such a choice can be encoded by a word @xmath11 of length @xmath79 where @xmath80 if the value @xmath81 appears in the first @xmath9 entries and @xmath82 otherwise .
we then have @xmath83 .
denoting by @xmath66 the set of words consisting of @xmath9 zeros and @xmath35 ones , we have @xmath84 the result now follows from equation .
lemma [ lem : product ] can be deduced by interpreting the sums as probabilities and appealing to the weak dissociation and consecutive homogeneity properties of the mallows@xmath85 distribution .
see ( * ? ? ?
* section 5 ) for a more detailed account of these properties .
lemma [ lem : product ] allows us to generalize the cluster method of goulden and jackson @xcite in order to be able to keep track of the inversion number . for @xmath18
, we say that @xmath86 is a _
@xmath55-cluster of length @xmath35 with respect to @xmath5 _ if * @xmath17 , * @xmath87 , * @xmath88 for all @xmath89 , and * @xmath90 for all @xmath89 .
one can think of a @xmath55-cluster as a permutation @xmath1 with @xmath55 marked occurrences of @xmath5 that start at the positions @xmath91 , each marked occurrence overlaps the next one , and the first and last marked occurrences are all the way at the beginning and at the end of @xmath1 , respectively .
for example @xmath92 is a @xmath93-cluster with respect to @xmath94 and @xmath95 is a @xmath93-cluster with respect to @xmath96 .
note that @xmath1 may have additional occurrences of @xmath5 aside from the marked ones .
the number of inversions of the cluster is defined to be @xmath3 .
let @xmath97 denote the number of @xmath55-clusters of length @xmath35 with respect to @xmath5 having @xmath81 inversions .
for example , @xmath98 for any @xmath18 and @xmath99 because of the cluster @xmath100 .
let @xmath101_q!}\ ] ] be the corresponding @xmath0-exponential generating function .
recall the definitions of @xmath102 and @xmath103 from equations and .
the following is a generalization of the cluster method that keeps track of the inversion number .
[ thm : cluster ] for every @xmath18 , @xmath104 in particular , setting @xmath105 , @xmath106 by _ concatenation _ of two permutations @xmath107 and @xmath108 we mean any permutation @xmath109 such that @xmath110 and @xmath111 . by _
permutation with some marked occurrences of @xmath5 _ we mean a permutation @xmath1 together with a set @xmath112 of indices so that @xmath1 has an occurrence of @xmath5 starting at each @xmath113 , and possibly other occurrences as well .
every permutation with marked occurrences of @xmath5 can be decomposed uniquely as a concatenation of clusters with respect to @xmath5 and single entries ( those that do not belong to any marked occurrence ) .
it follows that the generating function @xmath114 counts permutations @xmath1 with marked occurrences of @xmath5 , where the exponent of @xmath115 is the length of @xmath1 , the exponent of @xmath116 is the number of marked occurrences of @xmath5 in @xmath1 ( which is also an additive parameter ) , and the exponent of @xmath0 is the total number of inversions of @xmath1 .
indeed , by lemma [ lem : product ] , the product of the @xmath0-exponential generating functions of two sets of permutations @xmath117 and @xmath118 keeps track of the number of inversions of all the possible concatenations of a permutation from @xmath117 with a permutation from @xmath118 .
for example , in the expansion of @xmath119 , the term @xmath120 corresponds to all possible concatenations consisting of a cluster , followed by a single entry , followed by another cluster . to this generating function ,
a permutation @xmath17 with a total of @xmath57 occurrences of @xmath5 contributes a term @xmath121 for each of the @xmath122 ways to mark @xmath55 of these @xmath57 occurrences , for every @xmath123 .
thus , the contribution of @xmath1 to @xmath124 is @xmath125_q!}=q^{\operatorname{inv}(\pi)}u^\ell \frac{z^n}{[n]_q!},\ ] ] which agrees with the contribution of @xmath1 to @xmath126 . setting @xmath105 ,
this contribution is @xmath127_q!}$ ] if @xmath1 avoids @xmath5 , and @xmath128 otherwise .
it follows that @xmath129 .
[ thm : monotone ] @xmath130_q ! } - \sum_{j\ge0 } \frac{z^{jm+1}}{[jm+1]_q ! } \right)^{-1}\ ] ] by the second part of theorem [ thm : cluster ] , it is enough to show that @xmath131_q ! } - \sum_{j\ge0 } \frac{z^{jm+1}}{[jm+1]_q!}.\ ] ] clusters with respect to @xmath132 have no inversions , since the underlying permutation is of the form @xmath133 .
thus , @xmath134_q!}.\ ] ] to simplify the alternating sum in @xmath55 , note that @xmath55-clusters of length @xmath35 with respect to @xmath132 are in bijection with sequences @xmath135 with @xmath136 , @xmath137 , and @xmath138 for all @xmath89 , which are in bijection with compositions of @xmath139 with @xmath140 parts , each of size at most @xmath141 .
it follows that the ordinary generating function of @xmath55 clusters can be expressed as @xmath142 and so @xmath143 we conclude that @xmath144_q!}=-\sum_{j\ge1}\left(\frac{z^{jm}}{[jm]_q!}-\frac{z^{jm+1}}{[jm+1]_q!}\right),\ ] ] from where equation follows .
a pattern @xmath18 is said to be _ non - overlapping _ ( also called _ minimally overlapping _ in the literature ) if no permutation of length @xmath145 has two occurrences of @xmath5 .
for example , @xmath65 is non - overlapping .
[ thm:1b ] let @xmath146 be a non - overlapping pattern with @xmath147 and let @xmath148 .
then @xmath149_q!}\right)^{-1}.\ ] ] since @xmath5 is non - overlapping , all @xmath55-clusters with respect to @xmath5 have length @xmath150 .
let @xmath151 be the polynomial that counts @xmath55-clusters according to their number of inversions .
clearly @xmath152 , so let us assume that @xmath153 . in every @xmath55-cluster @xmath1 , the positions of the entries @xmath154 are forced , since they must belong to the occurrence of @xmath5 that starts in position @xmath15 , namely @xmath155 .
on the other hand , the remaining @xmath156 entries in this occurrence of @xmath5 can take any of the remaining @xmath157 values .
such a choice can be encoded by a word @xmath11 of length @xmath157 where @xmath80 if the value @xmath158 appears in @xmath155 and @xmath82 otherwise .
once this word @xmath11 is chosen , @xmath1 is uniquely determined by the relative order of the entries @xmath159 , which , up to standardization , form an arbitrary @xmath160-cluster @xmath161 with respect to @xmath5 .
additionally , the number of inversions of @xmath1 equals the number of inversions among the first @xmath9 entries ( which is @xmath6 ) , plus the number of inversions among the entries @xmath159 ( which is @xmath162 ) , plus the number of inversions between one of the first @xmath9 entries and one of the remaining entries ( which is @xmath163 ) .
it follows that @xmath164 where @xmath161 ranges over all @xmath160-clusters with respect to @xmath5 and @xmath11 ranges over all @xmath165-valued words of length @xmath157 having @xmath156 zeros . using equation
, we have @xmath166 by induction on @xmath55 , we get @xmath167 by equation and using the above argument , @xmath168_q!}=\sum_{k , i}c_\sigma(i , k , k(m-1)+1)q^i(u-1)^k\frac{z^{k(m-1)+1}}{[k(m-1)+1]_q!}\\ = \sum_{k}a_k(u-1)^k\frac{z^{k(m-1)+1}}{[k(m-1)+1]_q!}=\sum_{k}\prod_{j=1}^{k-1}\binom{j(m-1)+m - b}{m - b}_q\frac{q^{k\operatorname{inv}(\sigma)}(u-1)^k z^{k(m-1)+1}}{[k(m-1)+1]_q!}.$$\end{gathered}\ ] ] the result follows now from theorem [ thm : cluster ] .
in analogy to the fact that @xmath169 exists for every @xmath5 , we can prove the following result about the mallows distribution .
recall from section [ sec : intro ] that we write @xmath36 for the probability that a random permutation of length @xmath35 from the mallows distribution with parameter @xmath0 avoids @xmath5 .
[ thm : rho ] for every @xmath4 and every pattern @xmath5 , the limit @xmath170 exists .
first we show that for every @xmath171 , @xmath172 this is a consequence of lemma [ lem : product ] . indeed ,
taking @xmath173 , @xmath174 , and @xmath175 we have that @xmath176 , since a permutation that avoids @xmath5 must avoid @xmath5 in the first @xmath9 and in the last @xmath35 entries .
thus , @xmath177_q!}\le \sum_{\pi\in c } \frac{q^{\operatorname{inv}(\pi)}}{[m+n]_q!}=\left(\sum_{\pi\in a } \frac{q^{\operatorname{inv}(\pi)}}{[m]_q!}\right ) \left(\sum_{\pi\in b } \frac{q^{\operatorname{inv}(\pi)}}{[n]_q!}\right)=p_m(\sigma , q ) p_n(\sigma , q).\ ] ] applying fekete s lemma ( * ? ? ?
* lemma 1.6 ) , it follows that @xmath178 exists .
let @xmath179 be a random permutation from the mallows(@xmath0 ) distribution on @xmath7 and let @xmath180 for @xmath181 .
for @xmath182 , let @xmath183 be the event that @xmath184 , that is , the segment of @xmath1 starting at position @xmath89 avoids @xmath5 .
then we have @xmath185 by removing those events @xmath186 that overlap with the first @xmath35 and last @xmath9 locations of @xmath1 . by the weak dissociation property of the mallows distribution
, @xmath187 implies that @xmath188 and @xmath183 are independent , that is , @xmath189 .
thus , @xmath190 where the last equality is a consequence of the consecutive heterogeneity property of the mallows distribution .
this proves . to deduce that @xmath178 exists
we now apply fekete s lemma as before .
the alternate proofs of theorem [ thm : rho ] illustrate the interplay between combinatorial and probabilistic techniques on display throughout the paper .
the quantity @xmath191 defined in theorem [ thm : rho ] is called the _ growth rate _ of @xmath5 , and it is a function of @xmath0 . from the definition it follows that @xmath192 for all @xmath5 and @xmath0 .
it is clear from equation that @xmath193 . to draw plots of @xmath191 for @xmath194
, it will be convenient to use the change of variables @xmath195 , or equivalently @xmath196 . with this transformation ,
@xmath197 ranges between @xmath198 and @xmath15 , and the values @xmath199 for @xmath197 correspond to @xmath200 for @xmath0 , respectively .
additionally , the symmetry between @xmath0 and @xmath201 resulting from reversing the pattern corresponds to the symmetry between @xmath197 and @xmath202 , and thus the graph for @xmath203 is obtained by reflecting the graph for @xmath5 with respect to the vertical axis . using a well - known fact of singularity analysis ( * ? ? ?
* theorem iv.7 ) , @xmath204 equals the radius of convergence of @xmath42 as a function of a complex variable @xmath115 . additionally , if this radius is finite , then @xmath42 has a real singularity at @xmath205 , by pringsheim s theorem (
* theorem iv.6 ) and the fact that this generating function has non - negative coefficients . by theorem [ thm : cluster ] , defining @xmath206 allows us to write @xmath207 .
in particular , the smallest singularity of @xmath42 , when it exists , is either a zero or a singularity of @xmath208 . in some cases there is no such singularity , that is , @xmath42 has infinite radius of convergence .
in this section we study the growth rate @xmath191 when @xmath209 for any @xmath210 .
( when @xmath211 , there is exactly 1 permutation of @xmath35 , namely @xmath212 , that avoids @xmath213 , yielding @xmath214 . )
note that @xmath215 , so our results apply to the monotone decreasing pattern as well . by the arguments in section [ section : growth ]
, @xmath216 equals the smallest positive singularity of the generating function @xmath217 given in theorem [ thm : monotone ] , for @xmath218 as defined in .
let @xmath219_q!}$ ] , so that @xmath220_q ! } - \frac{z^{jm+1}}{[jm+1]_q!}\right)=\sum_{j\ge0 } ( a_{jm}-a_{jm+1}).\ ] ] note that @xmath221_q}$ ] .
we consider two cases depending on whether @xmath0 is larger or smaller than @xmath15 . * case @xmath222 . * in this case , @xmath8_q=1+q+\dots+q^{n-1}$ ] goes to infinity as @xmath223 .
thus , for any fixed @xmath115 , we have @xmath224 for @xmath35 large enough .
it follows that the series converges , and thus @xmath216 is the smallest positive zero of @xmath218 , which can be easily approximated by truncating the sum , since the factorial in the denominator makes the terms quickly go to zero .
figure [ fig : monotone_x01 ] shows the plots of @xmath225 for @xmath226 and @xmath227 , with the horizontal axis rescaled by @xmath196 , so that @xmath228 .
the values of @xmath225 have been obtained by truncating the series at @xmath229 and finding its zero numerically .
with @xmath230 for @xmath231 ( red ) , @xmath232 ( brown ) and @xmath233 ( green).,height=226 ] if we now let @xmath9 go to infinity , and we write the smallest positive zero of @xmath218 as @xmath234 , then we have that @xmath235_q!}-\frac{1}{[m+1]_q!}+ o\left(\frac{1}{[m+2]_q!}\right)$ ] , yielding @xmath236_q!}+\frac{1}{[m+1]_q!}+o\left(\frac{1}{[m+2]_q!}\right).\ ] ] * case @xmath237 . * in this case , @xmath238 as @xmath223 .
thus , the radius of convergence of @xmath218 is @xmath239 . for @xmath0 above a certain threshold ,
the function @xmath218 has a zero @xmath240 with @xmath241 . in this case , the smallest positive such zero can be estimated by truncating the series as before , giving the growth rate @xmath242 .
however , for small values of @xmath0 , the function @xmath218 does not have a positive zero smaller than @xmath239 , so we use a different method to approximate its smallest zero @xmath240 .
we first approximate the tail of @xmath218 as follows .
first note that @xmath243_q!=\frac{\prod_{i=1}^n(1-q^i)}{(1-q)^n}.\ ] ] as @xmath223 , the numerator approaches a constant , whose reciprocal we denote by @xmath244 using the approximation @xmath245_q!}\approx c_q(z(1-q))^n,\ ] ] we get , for large @xmath246 , @xmath247 and so @xmath248 we can find the smallest positive zero of the right - hand side of numerically , obtaining an approximation of the smallest zero @xmath240 of @xmath218 . using this method for @xmath231
, we have computed an approximation of @xmath249 as the reciprocal of the smallest positive zero of the right - hand side of for @xmath250 .
the resulting plot appears in figure [ fig:123_x_neg_tail ] .
we note that the smallest zero satisfies @xmath241 roughly when @xmath251 ( equivalently , @xmath252 with the rescaling @xmath196 ) .
however , for smaller values of @xmath0 , our method still yields an approximation of @xmath249 . for @xmath253 obtained by computing @xmath254 where @xmath240 is the smallest positive zero of @xmath255 , approximated by with @xmath250 .
the horizontal axis has been rescaled by @xmath196.,height=226 ] in the limit as @xmath256 , the mallows distribution assigns probability one to monotone increasing permutations , and so @xmath257 , which agrees with figure [ fig:123_x_neg_tail ] .
for comparison purposes , figure [ fig:123_x_all_tail ] shows the plots of @xmath258 for @xmath259 and @xmath260 .
as @xmath35 goes to infinity , these plots approach @xmath249 . in general , @xmath261 can be obtained by computing the coefficient of @xmath262 in the expression for @xmath263 in theorem [ thm : monotone ] .
( turquoise ) and @xmath264 ( orange ) as functions of @xmath197 , overlaid with @xmath249 ( red ) found using our method.,height=226 ] let us finish this section by analyzing the behavior of the growth rate @xmath225 for fixed @xmath237 as @xmath9 goes to infinity . for large @xmath9 , we can use the approximation @xmath265 writing the smallest positive zero of @xmath218 as @xmath234 , then we have that @xmath266 the possibility that @xmath267 can be easily ruled out because it gives a contradition with the fact that @xmath268 as @xmath269 . thus , using that @xmath270 , we can write @xmath271 from where we get @xmath272
in this section we study the growth rates @xmath191 when @xmath273 is a non - ovelapping pattern with @xmath147 .
the generating function @xmath274 for such a pattern is given in theorem [ thm:1b ] , and @xmath204 equals the smallest positive singularity of this generating function .
let @xmath275_q \prod_{j=1}^k [ 2j]_q},\ ] ] so that , by theorem [ thm:1b ] , @xmath276_q\right ) \frac{q^{k}(-1)^k z^{2k+1}}{[2k+1]_q ! } = 1-z-\sum_{k\ge1 } ( -1)^k h_k.\ ] ] we consider the quotient @xmath277_q}{[2k+1]_q[2k]_q},\ ] ] and distinguish two cases depending on whether @xmath0 is larger or smaller than @xmath15 .
* case @xmath222 . * in this case , @xmath278 as @xmath279 , and so the series @xmath280 is convergent .
it follows that @xmath281 is the smallest positive zero of @xmath280 , and it can be approximated by truncating the sum .
figure [ fig:132_x ] shows a plot of @xmath282 with the rescaling @xmath196 . for @xmath222 ,
@xmath282 has been found by taking the reciprocal of the smallest zero of a truncation of @xmath283 . as a function of @xmath197 , obtained by taking the reciprocal of the smallest positive zero of the series for @xmath283 truncated at @xmath284.,height=226 ]
if we write @xmath234 , we have that @xmath285 is equivalent to @xmath286_q \prod_{j=1}^k [ 2j]_q}.\ ] ] as @xmath287 , the dominant term of @xmath8_q$ ] is @xmath288 , and so @xmath289_q \prod_{j=1}^k [ 2j]_q}=o(q^{-k(k+1)})$ ] .
it follows that @xmath290_q[3]_q}+o(q^{-6})=\frac{q}{[2]_q[3]_q}+\frac{3q^2}{[2]_q^2[3]_q^2}+o(q^{-6})=q^{-2}-2q^{-3}+o(q^{-4})\ ] ] and @xmath291 equivalently , as @xmath256 , @xmath292 * case @xmath237 . * in this case , @xmath293 as @xmath279 .
thus , for @xmath294 , the series @xmath280 is convergent .
we will show that , for every @xmath237 , this series has a zero with @xmath295 ; therefore , this zero equals @xmath281 and can be approximated by truncating the series as before .
in fact , we will show that @xmath280 always has a zero with @xmath296 , which implies the above statement because @xmath297 for @xmath298 .
first , note that for @xmath299 , equation gives @xmath300_q}{[2k+1]_q[2k]_q}<\frac{3q}{[2k+1]_q}\le \frac{3q}{[3]_q}\le1\ ] ] for every @xmath301 and @xmath302 $ ] , where in the last inequality we have used the fact that @xmath303 for all @xmath302 $ ] , which follows from the fact that @xmath304 . by truncating an alternating series whose terms decrease in absolute value
, we have that , for any odd @xmath305 , @xmath306_q \prod_{j=1}^k [ 2j]_q}=:p_n(z)\ ] ] for @xmath299 .
since @xmath307 , it follows that @xmath308 , where @xmath309 is the smallest positive zero of @xmath310 . finally , by truncating at @xmath311 ( any larger odd @xmath305 works too ) , we obtain that @xmath312 , as can be seen by checking that @xmath313 for all @xmath298 .
thus , we have proved that @xmath314 as desired .
the above argument shows that for @xmath222 , the value of @xmath282 can be approximated by taking the reciprocal of the smallest zero of a truncation of @xmath283 , as we have done in figure [ fig:132_x ] .
an interesting feature of this graph is that the minimum of @xmath282 is attained for some value of @xmath0 strictly between @xmath128 and @xmath315 . a numerical approximation of this value is @xmath316 , which gives a growth rate of @xmath317 . as @xmath256 , equation gives the approximation @xmath318_q[3]_q } - \frac{q^2 ( 1+\eps)^5}{[2]_q[4]_q[5]_q } + o(q^3)= \frac{q}{[2]_q[3]_q}+\frac{3q^2}{[2]_q^2[3]_q^2}-\frac{q^2}{[2]_q[4]_q[5]_q}+o(q^3)=q+o(q^3),\ ] ] and @xmath319 note that some care must be taken when expanding out expressions containing @xmath0-analogues in powers of @xmath0 .
the arguments in the previous subsection generalize to non - overlapping patterns that start with a @xmath15 . letting @xmath18
be a non - overlapping pattern with @xmath147 and @xmath320 , theorem [ thm:1b ] gives @xmath321 , where we define @xmath322_q!}.\ ] ] the quotient of consecutive terms can be simplified as @xmath323_q\cdots [ k(m-1)+2]_q}\\ & = \frac{[k(m-1)+m - b]_q\cdots [ k(m-1)+2]_q[k(m-1)+1]_q \ , q^{\operatorname{inv}(\sigma)}z^{m-1}}{[m - b]_q!\ , [ k(m-1)+m]_q\cdots [ k(m-1)+3]_q[k(m-1)+2]_q}\\ & = \frac{[k(m-1)+1]_q\ , q^{\operatorname{inv}(\sigma)}z^{m-1}}{[m - b]_q ! \prod_{i=0}^{b-1}[k(m-1)+m - i]_q}.\end{aligned}\ ] ] * case @xmath222 . * in this case , noting that @xmath324 we see that @xmath325 as @xmath279 .
it follows that the series @xmath326 is convergent , and that @xmath204 is the smallest positive zero of @xmath326 , which can be approximated by truncating the sum , analogously to what we did for the pattern @xmath65 . in figure
[ fig:1b_m4 ] , the parts of the graphs of @xmath191 for @xmath222 ( equivalently , @xmath327 ) have been computed in this way truncating at @xmath328 .
* case @xmath237 . * in this case , @xmath329_q!}\ ] ] as @xmath279 .
thus , for @xmath330_q!}{(1-q)^{b-1 } q^{\operatorname{inv}(\sigma)}}\right)^{\frac{1}{m-1}},\ ] ] the series @xmath326 is convergent .
if this series has a positive zero @xmath331 satisfying , then this zero equals @xmath204 and can be approximated by truncating the series as we did for the pattern @xmath65 .
however , even without the assumption that the smallest positive zero of @xmath326 satisfies , we can use the following method to estimate this zero .
first , we write as @xmath332_q!\left([m - b]_q!\right)^{k-1}}\prod_{j=1}^{k-1}\prod_{i=1}^{m - b}[j(m-1)+i]_q.\ ] ] as in the case of monotone patterns , we use the approximation @xmath333_q!\approx \frac{1}{c_q ( 1-q)^{k(m-1)+1}}\ ] ] as @xmath279 , with @xmath334 as defined in equation . similarly ,
we can approximate @xmath335_q\approx\frac{1}{d_{q , m , b}(1-q)^{(k-1)(m - b)}}\ ] ] as @xmath279 , with @xmath336 being the constant defined by @xmath337 thus , for large @xmath55 , we can write @xmath338_q!\ , z}{d_{q , m , b}}\left(\frac{q^{\operatorname{inv}(\sigma)}(1-q)^{b-1}z^{m-1}}{[m - b]_q ! } \right)^k.\ ] ] and for @xmath246 large enough , @xmath339_q!\ , z}{d_{q , m , b}}\frac{\left(-\frac{q^{\operatorname{inv}(\sigma)}(1-q)^{b-1}z^{m-1}}{[m - b]_q ! } \right)^k}{1+\frac{q^{\operatorname{inv}(\sigma)}(1-q)^{b-1}z^{m-1}}{[m - b]_q!}}.\ ] ] in figure [ fig:1b_m4 ] , @xmath191 for @xmath237 ( equivalently , @xmath340 ) has been approximated for the three non - overlapping patterns of length 4 starting with a 1 as the reciprocal of the smallest positive zero of with @xmath341 .
with @xmath196 for @xmath342 ( red ) , @xmath343 ( purple ) , and @xmath344 ( blue ) .
the intersection of these curves with the @xmath345-axis coincides for @xmath346 and @xmath347 ( see @xcite for an explanation ) but it is slightly lower for the pattern @xmath348 .
we also point out that the curve for @xmath347 is not perfectly symmetric with respect to the @xmath345-axis.,height=226 ]
[ comparison : theorem ] for @xmath222 and every @xmath35 , we have @xmath349 and so @xmath350 for @xmath222 .
equivalently , for @xmath351 and every @xmath35 , we have @xmath352 , and so @xmath353 for @xmath351 . in @xcite , the authors give an injection @xmath354 defined as follows . if @xmath1 contains both @xmath355 and @xmath65 , let @xmath356 , otherwise , let @xmath357 be the permutation obtained by traversing @xmath1 from left to right and replacing each occurrence of @xmath355 with one of @xmath65 , by switching the entries playing the role of 2 and 3 .
it is clear that @xmath358 for every @xmath359 , and so @xmath360 for @xmath222 .
since @xmath361 is an injection , we have @xmath362 for @xmath222 , and so @xmath363 dividing by @xmath8_q!$ ] on both sides , we get that @xmath364 for @xmath222 .
the other statements now follow immediately . in section [ section : stein ] , we extend our discussion of pattern avoidance , which specializes to the setting in which a pattern occurs exactly zero times , by considering the distribution of the number of times a pattern occurs in a random permutation .
for example , by looking just at the expected number of occurrences of the patterns @xmath65 and @xmath365 in a random permutation of size @xmath35 from the mallows distribution with parameter @xmath4 , we obtain @xmath366_q!$ ] and @xmath367_q!$ ] , respectively . from this observation
, we see that there are more occurrences _ on average _ of @xmath365 than @xmath65 when @xmath368 and vice versa for @xmath369 .
it is natural to conjecture that the probability of avoiding @xmath365 ought to be smaller than the probability of avoiding @xmath65 when @xmath370 , and vice versa when @xmath253 .
the corresponding statement in the limit , at the level of growth rates , can be verified by looking at figure [ fig:132_x ] , which shows that for @xmath371 ( equivalently , @xmath340 ) , we have @xmath372 , since the transformation @xmath373 corresponds to the reflection @xmath374 in the picture . there are , however , interesting characteristics other than the growth rate @xmath191 of the pattern avoidance probability , such as the asymptotic distribution of the number of occurrences of a pattern and its rate of convergence .
these properties rely on calculations other than the marginal probability @xmath36 needed to obtain the expected number of occurrences , and therefore reveal more intricate structural features of pattern avoiding permutations .
we illustrate this further in our comparison of patterns @xmath347 , @xmath375 and @xmath376 in sections [ section : bounds ] and [ section : stein ] . increasing the value of @xmath0 in the mallows distribution
gives higher probability to permutations with more inversions .
it is therefore natural to conjecture that the probability of avoiding the pattern @xmath132 increases with @xmath0 .
this conjecture is supported by our numerical evidence in figures [ fig : monotone_x01 ] , [ fig:123_x_neg_tail ] , and [ fig:123_x_all_tail ] . [ conj:123increasing ] if @xmath377 , then @xmath378 and @xmath379 .
this conjecture would follow if we could show that the sequence @xmath380 is weakly increasing in @xmath55 .
note that , using inversion tables to represent permutations , the above denominator equals the number of sequences @xmath381 with @xmath382 for all @xmath81 and @xmath383 , whereas the numerator equals the number of such sequences that additionally do not contain three consecutive entries with @xmath384 . a more probabilistic approach , which appears to suffer from similar difficulties , is to obtain a coupling between the two bumping processes driving the mallows@xmath85 distribution in order to establish stochastic dominance between the number of occurrences of @xmath132 in mallows(@xmath0 ) and mallows(@xmath385 ) distributions for @xmath377 ; see @xcite for the relevant definitions .
we have been unable to establish the conjecture using either of these techniques , and so we leave this as an open problem .
our techniques above allow us to approximate the growth rate for those patterns whose generating function is sufficiently tractable to analyze . in the absence of explicit generating functions ,
we establish upper and lower bounds for the growth rate using different techniques . in this section
, we generalize many of the results in @xcite to bound @xmath191 , for arbitrary @xmath18 and @xmath4 , using suen s inequality and a version of the lovsz local lemma .
the analysis in @xcite specializes to @xmath387 only . for any permutation @xmath59 and any pattern @xmath18
, we can mark the occurrences of @xmath5 in @xmath1 by mapping @xmath1 to a @xmath165-valued sequence @xmath388 , with @xmath389 if @xmath1 is a random permutation of @xmath35 drawn from the mallows(@xmath0 ) distribution , then the associated sequence of markings @xmath388 is also random and @xmath390 is the random variable counting the number of occurrences of @xmath5 in @xmath1 . for each @xmath391
, we define @xmath392 to be the event that @xmath5 occurs in position @xmath89 of a mallows(@xmath0 ) permutation .
we write @xmath393 to denote the complement of @xmath183 , and @xmath394 .
we define the _ dependency graph _ @xmath395 of @xmath396 to be the graph with vertex set @xmath397 and edge set @xmath398 with an edge between nodes @xmath81 and @xmath89 if and only if @xmath399 . by the weak dissociation property of the mallows distribution , if any two disjoint subsets @xmath400 are such that there are no edges between the nodes in @xmath112 and @xmath401 in the dependency graph @xmath395 , then the sets of events @xmath402 and @xmath403 are independent .
[ prop : generic : upper ] fix any @xmath404 , @xmath405 , and @xmath4 .
then we have @xmath406_q!}\right)^{1/m}.\ ] ] in particular , the right - hand side of equation is the same for all patterns @xmath405 with the same number of inversions .
we follow the same routine as in @xcite .
we define the index set @xmath407 so that @xmath408 by the weak dissociation property , events @xmath188 and @xmath183 are independent for @xmath409 and @xmath410 , and by consecutive homogeneity @xmath411_q!$ ] for every @xmath89 . combining this with the fact that @xmath412 , we have @xmath413_q!}\ ] ] and @xmath414_q!}\right ) \leq \left(1 - \frac{q^{\operatorname{inv}(\sigma)}}{[m]_q!}\right)^{n / m}. \qedhere\ ] ] suen
s inequality provides a notable improvement to the above upper bound in many instances by taking into account interactions between dependent events .
let @xmath398 denote the set of edges in the previously defined dependency graph @xmath395 .
we define @xmath415 and @xmath416 the following is an improvement to suen s original inequality due to janson ( * ? ? ?
* theorem 2 ) .
[ suen : inequality ] let @xmath417 be a finite family of indicator random variables with dependency graph @xmath395 and define @xmath418 .
let @xmath419 , @xmath420 .
then @xmath421 we are thus able to produce an improved upper bound for @xmath191 using the overlap set of @xmath5 , defined below .
[ overlap ] for every @xmath405 and @xmath422 , we define the _ overlap set of @xmath5 of size @xmath423 _ as the set of permutations @xmath424 such that @xmath425 , denoted by @xmath426 .
we similarly define @xmath427 as the _ overlap set of @xmath5_. in the language of section [ section : gj ] , each @xmath428 corresponds to a @xmath16-cluster @xmath429 of @xmath5 , whose only marked occurrences of @xmath5 overlap in exactly @xmath423 positions and occur at the beginning and end of @xmath161 .
[ prop : improved ] fix any @xmath404 , @xmath405 and @xmath4 , and define @xmath430_q!}.\ ] ] then @xmath431_q ! } + \exp\left ( 4(m-1 ) \frac{q^{\operatorname{inv}(\sigma)}}{[m]_q!}\right)\ \sum_{s=1}^{m-1}t(s,\sigma , q ) \right).\ ] ] for fixed @xmath404 , @xmath432 , and @xmath4 , with @xmath433 , let @xmath434 be the markings defined in for a mallows(@xmath60 permutation and let @xmath435 be as defined above .
we have @xmath436_q ! } , \qquad \p(a_j \cap a_i ) = \e x_i x_j,\ ] ] whence @xmath437 and also @xmath438_q!}.\ ] ] we then apply theorem [ suen : inequality ] by noting that @xmath439 and @xmath440 as @xmath223 , and also @xmath441 and @xmath442 the lower bound is a bit more delicate , and uses a version of the lovsz local lemma ( * ? ? ?
* lemma 2.1 ) particularly suited to this type of problem . [ lll ]
let @xmath443 be events in some probability space , and let @xmath444 be a sequence of numbers in @xmath445 .
for each @xmath446 , suppose there is an integer @xmath447 satisfying @xmath448 such that @xmath449 then , for any @xmath450 , we have @xmath451 [ lower : bound ] fix any @xmath404 , @xmath405 , @xmath4 , and assume that @xmath452_q ! } - \sqrt{1-(4m-2)\frac{q^{\operatorname{inv}(\sigma)}}{[m]_q!}+\frac{q^{2\operatorname{inv}(\sigma)}}{[m]_q!^2}}\right)\ ] ] is such that @xmath453_q ! } e^{f(q , m ) } \in ( 0,1)$ ] . then @xmath454_q ! } \exp\left(\frac{1}{2}\left(1 - \frac{q^{\operatorname{inv}(\sigma)}}{[m]_q !
} - \sqrt{1-(4m-2)\frac{q^{\operatorname{inv}(\sigma)}}{[m]_q!}+\frac{q^{2\operatorname{inv}(\sigma)}}{[m]_q!^2}}\right)\right).\ ] ] in order to apply theorem [ lll ] in our setting , we first choose @xmath455 , so that the left - hand side of equation is @xmath456_q!},\ ] ] as in the proof of proposition [ prop : generic : upper ] . then , by symmetry of the events in index @xmath81 , it suffices to find a single @xmath457 such that @xmath458_q ! } \leq z ( 1-z)^{m-1}.\ ] ] following @xcite , we consider @xmath115 of the form @xmath459_q ! } e^{f(q , m)}$ ] , for some positive function @xmath460 . then is equivalent to @xmath461_q!}e^{f(q , m ) }
\right)^{m-1 } , \\ \nonumber e^{-\frac{f(q , m)}{m-1 } } \leq & \ 1 - \frac{q^{\operatorname{inv}(\sigma)}}{[m]_q ! } e^{f(q , m ) } , \\
\label{before : inequality}-\frac{f(q , m)}{m-1 } \leq & \ \log\left(1 - \frac{q^{\operatorname{inv}(\sigma)}}{[m]_q ! } e^{f(q , m)}\right).\end{aligned}\ ] ] at this point we utilize the inequality @xmath462 , valid for all @xmath463 , and so we must assume that @xmath453_q ! } e^{f(q , m ) } < 1 $ ] .
equation will then hold assuming the following holds : @xmath464_q ! } e^{f(q , m)}}{1-\frac{q^{\operatorname{inv}(\sigma)}}{[m]_q ! } e^{f(q , m)}}.\ ] ] rearranging , this is equivalent to @xmath465_q!}\right ) \geq ( m-1 ) \frac{q^{\operatorname{inv}(\sigma)}}{[m]_q!}.\ ] ] note that @xmath466 for @xmath467 , and so the above inequality will be satisfied if the following holds : @xmath468_q ! }
\right ) \geq ( m-1)\frac{q^{\operatorname{inv}(\sigma)}}{[m]_q!}.\ ] ] since this is quadratic in @xmath460 , by solving using equality in place of inequality in , we can obtain the region in which the inequality is satisfied .
the two solutions are @xmath469_q ! } \pm \sqrt{1-(4m-2)\frac{q^{\operatorname{inv}(\sigma)}}{[m]_q!}+\frac{q^{2\operatorname{inv}(\sigma)}}{[m]_q!^2}}\right).\ ] ] since the coefficient of @xmath470 is negative in , we know that inequality is satisfied for all values in between the solutions . since both solutions are positive , we take the smaller solution to obtain the optimal value @xmath471_q ! } \exp\left(\frac{1}{2}\left(1 - \frac{q^{\operatorname{inv}(\sigma)}}{[m]_q ! } - \sqrt{1-(4m-2)\frac{q^{\operatorname{inv}(\sigma)}}{[m]_q!}+\frac{q^{2\operatorname{inv}(\sigma)}}{[m]_q!^2}}\right)\right).\ ] ] this completes the proof .
there is a minor technical issue in ( * ? ? ?
* proof of theorem 5 ) , which does not affect the validity of ( * ? ? ?
* theorem 5 ) as stated , since the resulting bounds are stated in terms of asymptotic estimates as @xmath9 tends to infinity .
however , it does impact the final form and proof of our proposition [ lower : bound ] , since we are adapting the argument and using it in its preasymptotic form .
specifically , the argument used to obtain @xmath472 , see @xcite , is flawed .
it seems that instead of using the inequality @xmath473 , as we have done following equation , the reverse inequality @xmath474 was used .
we have verified numerically that for @xmath475 , inequality fails for @xmath27 and this value of @xmath115 . rather than rely on the analytic lower bound obtained in proposition [ lower : bound ] , we can instead compute the optimal value of @xmath115 in numerically for each @xmath5 and @xmath0 to obtain a better estimate , as we have done in the examples of the next section .
we next compute the explicit bound implied by suen s inequality in proposition [ prop : improved ] for some specific patterns .
see figure [ fig : bounds ] for plots of this improved bound , together with the bounds implied by propositions [ prop : generic : upper ] and [ lower : bound ] .
+ + first , we consider the pattern @xmath476 of length @xmath477 with @xmath93 inversions previously encountered in section [ section : generalizations ] .
the upper and lower bounds given in propositions [ prop : generic : upper ] and [ lower : bound ] apply to all other patterns of length @xmath477 with @xmath93 inversions , namely , @xmath375 , @xmath376 , @xmath478 , @xmath479 , @xmath480 .
the upper bound from proposition [ prop : improved ] , however , depends on the set of overlaps @xmath481 , which we have presented in table [ ov:1432 ] along with the number of inversions for each pattern .
the quantity @xmath482 , with @xmath483 defined in equation , is then @xmath484_q ! } = : t_{1432}(q),\ ] ] allowing us to compute the upper bound for all @xmath4 : @xmath485_q ! } + t_{1432}(q)\ e^{\frac{12 q^3}{[4]_q!}}\right).\ ] ] @xmath486 @xmath487 [ ov:1342][ov:1243][ov:2341][ov:2413 ] we next consider 1342 , which has length 4 and 2 inversions .
the other patterns of length 4 with 2 inversions are @xmath94 , @xmath488 , @xmath489 , and @xmath490 . as in the previous example
, we compute @xmath491 along with the number of inversions for each element in the set , summarized in table [ ov:1342 ] .
we have @xmath492_q ! } = : t_{1342}(q),\ ] ] and so @xmath493_q!}+t_{1342}(q)\ e^{\frac{12 q^2}{[4]_q!}}\right).\ ] ] the pattern 1243 demonstrates that the condition in proposition [ lower : bound ] is essential , and that sometimes it is best to evaluate the relevant quantities numerically , as the interplay between @xmath0 and @xmath9 can be quite complicated when attempting to solve inequality , which does not always have a solution . in this case ,
pattern @xmath348 has length @xmath477 with exactly @xmath15 inversion , which is also the case for patterns @xmath346 and @xmath494 . the overlap set is much smaller than in previous examples , consisting of just the four elements given in table [ ov:1243 ] .
we have @xmath495_q ! } = : t_{1243}(q),\ ] ] and so @xmath496_q ! }
+ t_{1243}(q)\ e^{\frac{12 q}{[4]_q!}}\right).\ ] ] in particular , this is the first case covered so far where there is a range of values of @xmath0 for which inequality has no solution .
figure [ fig:2413:bounds ] plots the bounds for the pattern @xmath376 , which has 3 inversions and , therefore , shares the universal upper and lower bounds of figure [ fig:1432:bounds ] , but which differs from pattern 1432 with its own custom improved upper bound . figure [ fig:5:bounds ] plots the various universal bounds for patterns of length 5 with inversions @xmath497 .
the symmetry in these bounds for @xmath498 reflects the symmetry in @xmath191 .
figure [ fig:10:5:bounds ] showcases the versatility and computational efficiency of the generic bounds by plotting the bounds for any pattern of length 10 with exactly 5 inversions .
finally , we revisit the monotonically increasing pattern @xmath96 , which is the only pattern of length 4 with 0 inversions . in this case
, there is also a large range of values of @xmath0 for which inequality has no solution . in addition , suen s inequality rises above 1 for @xmath0 small enough , whereas the universal bound is more useful .
figure [ fig:1234:bounds ] plots these bounds , and figure [ fig:12m : bounds ] plots the bounds for the monotonically increasing patterns @xmath499 for @xmath500 . , with @xmath501 , as described in the caption of figure [ fig : bounds ] . ]
+ + + for any @xmath502 with exactly 5 inversions.,height=226 ]
for any permutation @xmath59 and @xmath18 , recall the definition of @xmath503 as the indicator variables defined in and @xmath504 as the random variable counting the number of occurrences of @xmath5 in @xmath1 from the mallows(@xmath0 ) distribution on @xmath7 . when concerned only with the binary event of whether @xmath1 avoids @xmath5 ,
the only relevant information in @xmath503 is whether the event @xmath505 occurs , and our analysis above focuses on the asymptotic behavior of @xmath506 . by theorem [ thm : rho ] , this probability decays exponentially fast in @xmath35 , implying that for all practical purposes @xmath36 is negligible for even moderately large values of @xmath35 . as a complement of our prior analysis
, we now consider how far a permutation strays from avoiding a given pattern by studying the distribution of the number of occurrences of @xmath5 in a random permutation from the mallows(@xmath0 ) distribution .
other authors , e.g. , @xcite , have studied similar questions using different techniques and under different assumptions .
computing the distribution of the number of occurrences , however , is complicated by the dependence among the locations at which a pattern is allowed to occur . for a simple example ,
notice that the pattern @xmath507 can occur at most once in a permutation of length 4 , meaning that the variables @xmath508 indicating occurrence starting in positions 1 and 2 , respectively , are negatively correlated .
on the other hand , the pattern @xmath355 can occur 0 , 1 , or 2 times in a permutation of length 4 .
complications due to this dependence , however , are mitigated by the weak dissociation property of the mallows(@xmath0 ) distribution , according to which @xmath509 and @xmath510 are independent as long as @xmath511 . in probabilistic terms
, @xmath512 is an _
@xmath513-dependent _ sequence , for which much is already known , including quantitative bounds on convergence rates to a central limit theorem ; see @xcite .
the consecutive homogeneity property of the mallows(@xmath0 ) distribution implies that @xmath509 and @xmath510 have the same distribution .
both the weak dissociation and consecutive homogeneity properties of the mallows(@xmath0 ) distribution play a critical role , implicitly or explicitly , throughout all of our above analysis , including the probabilistic proof that the growth rate always exists ( theorem [ thm : rho ] ) and the bounds obtained in section [ section : bounds ] , as when deriving the improved upper bounds in proposition [ prop : improved ]
. these properties also play a role in the following theorem , which applies to the entire distribution of @xmath514 and , therefore , says something about its behavior in high probability regions that are most likely to occur .
[ theorem : stein ] fix any @xmath404 , @xmath18 , @xmath4 , @xmath433 , and let @xmath514 denote the number of times @xmath5 occurs in a random permutation of length @xmath35 from the mallows@xmath85 distribution . for @xmath516 , define @xmath517_q ! } ; \qquad a_{n}(\sigma , q ) : = ( n - m+1)\ , \mu(\sigma , q ) ; \qquad t(s,\sigma , q ) : = \sum_{\tau \in { \small\ov}_s(\sigma ) } \frac{q^{\operatorname{inv}(\tau)}}{[2m - s]_q!};\ ] ] @xmath518 @xmath519 then we have @xmath520 where @xmath521 is a gaussian random variable with mean 0 and variance 1 , that is , @xmath522 fix any @xmath523 , @xmath405 , @xmath4 , and @xmath433 .
we use the notation @xmath524 , @xmath525 , and @xmath526 . by linearity of expectation
, it immediately follows that @xmath527 . for @xmath528
, we use the decomposition of the variance of @xmath529 by @xmath530 where @xmath531 is the covariance of @xmath509 and @xmath510 . by the weak dissociation property of the mallows(@xmath0 ) distribution , @xmath532 whenever @xmath533 , and so we need only consider pairs @xmath81 and @xmath89 for which @xmath534 . using definition [ overlap ] , we have @xmath535_q ! } , \qquad 1 \leq s \leq m-1 , \quad 1 \leq i \leq n- 2 m + 1 + s.\ ] ] since @xmath509 is an indicator random variable , we also have @xmath536 for all @xmath537 by the consecutive homogeneity property of the mallows distribution .
whence , @xmath538 where we used in the last equality .
next , we define the random variables @xmath539 and @xmath540 we have @xmath541 for @xmath537 , and by our choice of @xmath528 above we have @xmath542 .
we also have , since @xmath509 is an indicator random variable , that @xmath543 , and so @xmath544 by ( * ? ? ?
* theorem 2.6 ) , since the collection @xmath545 is @xmath513-dependent , it satisfies @xmath541 for all @xmath546 , and since @xmath542 , we obtain the following bound for @xmath66 : @xmath547 inequality now follows since @xmath548 for all @xmath537 .
theorem [ theorem : stein ] is stated in terms of an inequality that is valid for all finite values of parameters with explicitly defined constants , rather than as a limit theorem .
a distinct advantage of stein s method is that it yields explicit preasymptotic bounds , from which quantitative statements can be derived for all finite values of the parameters .
theorem [ theorem : stein ] also yields many corollaries .
we assume the notation from theorem [ theorem : stein ] in the rest of this section .
fix any @xmath404 , @xmath405 , and @xmath4 .
assume @xmath549 i.e. , it is _
strictly positive_. then @xmath550 converges in distribution to a normal random variable with mean 0 and variance 1 as @xmath35 tends to infinity , with rate @xmath551 . for fixed @xmath404 , @xmath405 , and @xmath4
, we have @xmath552 where @xmath553 is a function which does not depend on @xmath35 .
similarly , we have @xmath554 where @xmath555 and @xmath556 do not depend on @xmath35 .
note that @xmath557 equals the variance of a sum of random variables , which must always nonnegative .
thus , by assuming that @xmath558 we have @xmath559 and also @xmath560 thus , asymptotically as @xmath35 tends to infinity , the right - hand side of equation is @xmath561 , which by is also @xmath562 .
fix any @xmath404 , @xmath405 , and take @xmath27 .
assume @xmath563 let @xmath564 denote the number of permutations of @xmath35 with exactly @xmath55 occurrences of @xmath5 .
for any @xmath565 , define @xmath566 .
then _ for each _ @xmath433 and @xmath565 , we have @xmath567 in particular , as @xmath568 and @xmath35 tend to infinity , suppose there is an @xmath569 such that also @xmath570 , then @xmath571 in a random sample of size @xmath572 generated from the mallows@xmath85 distribution using @xmath573 , plotted with the cumulative normal distribution function implied by theorem [ theorem : stein ] , for values of @xmath574 , as indicated below each plot . , title="fig : " ] in a random sample of size @xmath572 generated from the mallows@xmath85 distribution using @xmath573 , plotted with the cumulative normal distribution function implied by theorem [ theorem : stein ] , for values of @xmath574 , as indicated below each plot .
, title="fig : " ] in a random sample of size @xmath572 generated from the mallows@xmath85 distribution using @xmath573 , plotted with the cumulative normal distribution function implied by theorem [ theorem : stein ] , for values of @xmath574 , as indicated below each plot .
, title="fig : " ] + .2 in in a random sample of size @xmath572 generated from the mallows@xmath85 distribution using @xmath573 , plotted with the cumulative normal distribution function implied by theorem [ theorem : stein ] , for values of @xmath574 , as indicated below each plot .
, title="fig : " ] in a random sample of size @xmath572 generated from the mallows@xmath85 distribution using @xmath573 , plotted with the cumulative normal distribution function implied by theorem [ theorem : stein ] , for values of @xmath574 , as indicated below each plot .
, title="fig : " ] in a random sample of size @xmath572 generated from the mallows@xmath85 distribution using @xmath573 , plotted with the cumulative normal distribution function implied by theorem [ theorem : stein ] , for values of @xmath574 , as indicated below each plot .
, title="fig : " ] in this section , we compare the patterns @xmath347 , @xmath375 , and @xmath376 , each of which has 3 inversions .
thus , we expect _ on average _ the same number of these patterns in a random permutation , that is , @xmath575 , for all @xmath4 and @xmath576 .
however , the variance of the number of occurrences , denoted @xmath577 , differs for each pattern @xmath5 . to see why this is the case , we note that the only difference between patterns in the formula for @xmath578 in equation is in the terms @xmath483 , @xmath422 , where @xmath483 is the probability that in a random permutation of length @xmath579
, the pattern @xmath5 occurs at the start and at the end ( with possibly more occurrences allowed ) .
we showed previously in section [ sect:1432 ] and in table [ ov:1432 ] that for pattern @xmath347 we have @xmath580 , whence @xmath581 and also @xmath582_q!}.\ ] ] in fact , it is clear ( see ( * ? ? ?
* lemma 9 ) ) that , unless @xmath18 is monotone , @xmath583 .
thus , we need only consider @xmath483 for @xmath584 in this section .
continuing for pattern @xmath375 , we have @xmath585 , so @xmath586 ; also , by table [ ov:2341 ] , which enumerates the elements of @xmath587 , we have @xmath588_q!}.\ ] ] at this point , it is apparent that when @xmath27 , the variances of @xmath589 and @xmath590 are in fact the same , since @xmath591 for all @xmath592 . in addition , since the reversal of @xmath347 is @xmath375 , by we also have @xmath593 for all @xmath4 , which is apparent by the symmetry about the vertical axis in figure [ stddev : compare ] between the two corresponding curves . finally , the pattern 2413 is distinct from the previous two patterns in the sense that @xmath594 . by table
[ ov:2413 ] , we have @xmath595_q!},\ ] ] @xmath596_q!}.\ ] ] we have found that @xmath597 and @xmath598 cross at the point @xmath599 , corresponding to a value of @xmath600 , which gives @xmath601 ( see figure [ stddev : compare ] ) . as @xmath35 grows , this intersection appears to achieve a limiting value of @xmath602 , corresponding to @xmath603 , and taking @xmath35 large we have @xmath604 in addition , we observe numerically that the curve @xmath597 always stays strictly above @xmath605 for @xmath4 .
p. baldi , y. rinott , and c. stein .
a normal approximation for the number of local maxima of a random function on a graph .
_ probability , statistics , and mathematics _ , pages 5981 . academic press , boston , ma p. erds and l. lovsz .
problems and results on 3-chromatic hypergraphs and some related questions . in a.
hajnal , r. rado , and v. t. ss , eds
. infinite and finite sets ( to paul erds on his 60th birthday ) .
north - holland . | we use various combinatorial and probabilistic techniques to study growth rates for the probability that a random permutation from the mallows distribution avoids consecutive patterns .
the mallows distribution behaves like a @xmath0-analogue of the uniform distribution by weighting each permutation @xmath1 by @xmath2 , where @xmath3 is the number of inversions in @xmath1 and @xmath0 is a positive , real - valued parameter .
we prove that the growth rate exists for all patterns and all @xmath4 , and we generalize goulden and jackson s cluster method to keep track of the number of inversions in permutations avoiding a given consecutive pattern . using singularity analysis , we approximate the growth rates for length-3 patterns , monotone patterns , and non - overlapping patterns starting with 1 , and we compare growth rates between different patterns .
we also use stein s method to show that , under certain assumptions on @xmath0 , the length of @xmath5 , and @xmath6 , the number of occurrences of a given pattern @xmath5 is well approximated by the normal distribution . _
keywords : _ consecutive pattern , permutation , mallows distribution , inversion , growth rate , cluster method , stein s method _ mathematics subject classification : _ 05a05 , 60c05 , 05a15 , 05a16 , 62e17 , 62e20 , 05a30 |
human beings are extraordinarily complex agents .
remarkably , in spite of that complexity a number of striking statistical regularities are known to describe individual and societal human behavior @xcite .
these regularities are of enormous practical importance because of the influence of individual behaviors on social and economic outcomes .
even though the analysis of social and economic data has a long and illustrious history , from smith @xcite to pareto @xcite and to zipf @xcite , the recent availability of digital records has made it much easier for researchers to _ quantitatively _ investigate various aspects of human behavior . in particular , the availability and omnipresence of e - mail communication records is attracting much attention @xcite .
recently , barabsi studied the e - mail records of users at a university and reported two patterns in e - mail communication @xcite : the time interval between two consecutive e - mails sent by the same user , which we will denote as the interevent time @xmath1 , and the time interval between when a user sends an e - mail and when a recipient sends an e - mail back to the original sender , which we will denote as the waiting time @xmath2 , follow power - law distributions which decay in the tail with exponent @xmath0 .
additionally , barabsi proposed a priority queuing model that reportedly captures the processes by which individuals reply to e - mails , thereby predicting the probability distribution of @xmath2 . here , we demonstrate that the empirical results reported in ref .
@xcite are an artifact of the data analysis .
we perform a standard bayesian model selection analysis that demonstrates that the interevent times are well - described by a single log - normal while the waiting times are better described by the superposition of two log - normals .
our analysis rejects beyond any doubt the possibility that the data could be described by truncated power - law distributions .
we also critically evaluate the priority queuing model proposed by barabsi to describe the observed waiting time distributions .
we show that neither the assumptions nor the predictions of the model are plausible .
we thus conclude that the description of human e - mail communication patterns remains an open problem .
the remainder of this paper is organized as follows . in section
ii , we describe the preprocessing of the data .
we then analyze the distribution of interevent times ( section [ sect : tau ] ) and the distribution of waiting times ( section [ sect : tauw ] ) .
finally , in section [ sect : pqm ] we investigate the priority queuing model of ref .
we consider here the database investigated by barabsi @xcite , which was also the focus of an earlier paper by eckmann et al . @xcite .
this database consists of e - mail records for 3,188 e - mail accounts at a university covering an 83-day period .
each record comprises a sender identifier , a recipient identifier , the size of the e - mail , and a time stamp with a precision of one second . before describing our analysis of the data , we first note some important features of the data which impact the analysis .
the first important fact is that the data were gathered at an e - mail server , not from the e - mail clients of the individual users .
it is quite possible that some users have e - mail clients , like microsoft outlook , which permit users to send multiple e - mails at once regardless of when the e - mails were composed .
moreover , servers may parse long recipient lists into several shorter lists @xcite . for this reason , e - mails to multiple recipients
were occasionally recorded in the server as _ multiple _ e - mails .
each of these duplicate e - mails was then sent in rapid succession to a different subset of the list of recipients in the actual e - mail .
both the client - side and server - side uncertainties introduce artifacts in the time series of interevent times for each user as it could appear that a user is sending several e - mails over a very short time interval . to minimize these uncertainties , we preprocessed the data in order to focus on _ actual _ human behavior .
first , we identify sets of e - mails sent by a user that have the _ exact _ same size but whose time stamp differs by at most five seconds . ] .
we then remove all but the first e - mail from the time series of e - mails sent , while adjusting the list of recipients to the first e - mail to include all recipients in the removed e - mails .
an additional important fact to note is that some of the e - mail accounts do not belong to `` typical '' users .
for example , user 1962 only sent 5 e - mails while receiving 2,284 e - mails .
this individual s e - mail use is too infrequent to provide useful information on human dynamics .
meanwhile user 4099 sent 9,431 e - mails while receiving no e - mails .
although it can not be confirmed due to the anonymous nature of the data , this e - mail account was in all likelihood used for bulk e - mails , implying that it can not provide information on _ human _ e - mail usage . to avoid having our analysis distorted ,
we first restrict our attention to users which sent at least 11 e - mails over the 83-day experiment , yielding a minimum of 10 interevent times .
our reasoning is that users sending fewer e - mails do not use e - mail regularly enough to allow us to truly infer patterns of human dynamics .
this procedure excludes 1,976 of the 3,188 original e - mail accounts .
next we examine the ratio of the number of e - mails received to the number of e - mails sent to determine what constitutes a `` typical '' user .
this ratio is well - described by a log - normal distribution , and we use this fact to consider only those users in our study who are within three standard deviations from the mean .
this added constraint excludes an additional 46 users .
we thus focus here on the 1,152 users who fulfill the above criteria ( fig .
[ parsing ] ) .
reference @xcite reports that the probability distribution of time intervals @xmath1 between consecutive e - mails sent by an individual follows a power - law @xmath3 with @xmath4 .
a basic examination of barabsi s results , however , quickly reveals a number of issues . 1 . figure 2a of ref .
@xcite features three bins corresponding to interevent times @xmath5 seconds , an unphysical interval ( fig .
[ fig2a]a ) .
the events in those bins in fact account for 9% of all events
figure 2a of ref .
@xcite features at least one bin confined to interevent times @xmath6 second ( fig .
[ fig2a]b c ) while the data have a precision of one second @xcite . .
two of us sent about twenty e - mails trying to minimize the time interval between consecutive e - mails .
to be as fast as possible , we sent replies to an e - mail already in our inboxes .
additionally , we did not even write any text or read the e - mail to which we were responding .
we find that humans need at least 3 seconds to send consecutive e - mails . *
b * , reproduction of fig.2a of ref .
@xcite obtained with vistametrix @xcite and , * c * , the same data with the boundaries of the bins clearly marked .
we assumed that the data points in fig .
2a of ref .
@xcite were placed in the middle of the bin .
note that there is a bin recording data for @xmath6 second , whereas the data have a resolution of one second .
the shaded bins indicate values with @xmath5 second , which contain 9% of all events for the unidentified user . ]
we next quantitatively compare the plausibility of our log - normal hypothesis with the plausibility of the power - law hypothesis of ref .
@xcite for interevent times . to simplify the analysis
, we do not consider @xmath1 , but its logarithm .
if a random variable @xmath1 is log - normally distributed , then @xmath7 follows a gaussian distribution , whereas if @xmath1 is distributed according to a power - law with exponent @xmath8 , then @xmath9 is uniformly distributed in the interval @xmath10 $ ] .
specifically , for @xmath11 the distribution of @xmath7 is @xmath12 for three users in the database .
the top panels show data and gaussian model predictions for the entire range of @xmath9 , whereas the bottom panels show data and power - law model for intermediate values of @xmath9 . *
b * , scatter plot of @xmath13 , the ratio of the two @xmath14 values , for all available users depending on the number of consecutive e - mails sent for each user .
the larger circles colored green , yellow , and red correspond to the data shown in ( a ) .
note that there are 2 users for which the ratio of @xmath14 is greater than @xmath15 .
those users are indicated by the arrows .
* c * , recursively calculated posterior probability of accepting the power - law model .
we use bayesian model selection to recursively calculate the posterior probability of selecting the power - law distribution for two different prior probabilities : @xmath16 ( solid black ) and @xmath17 ( red dashed ) . in both cases , the posterior probability of selecting the power - law model _ vanishes _ after considering 932 of the 1,016 users meeting our preprocessing criteria . ]
barabsi @xcite has argued that the power - law model is meant to describe only `` intermediate '' @xmath1 values falling between 100 and 10,000 seconds .
since some users have a smaller range of @xmath1 values than that interval , we test the agreement of the predictions of the power - law model only with data in the interval @xmath18 $ ] , where @xmath19 and @xmath20 . to properly specify the power - law distribution
, we must have at least two data points in [ @xmath21 $ ] .
this constraint leads to the exclusion of an additional 136 users .
we then use the kolmogorov - smirnov ( ks ) test @xcite as a measure of the plausibility of a model given the user s data . specifically , we compare the distribution of the logarithm of the interevent times for a given user to two candidate models : a gaussian distribution and a uniform distribution .
importantly , _ there is absolutely no fitting in our analysis_. the parameters of the gaussian distribution , @xmath22 and @xmath23 , are simply the sample average and standard deviation of @xmath9 , while the uniform distribution is completely specified by @xmath24 and @xmath25 .
figure [ ks_sent]d displays the ratio of the two ks probabilities versus number of e - mails sent for all users with at least two data points in the interval @xmath26 $ ] . in order to determine which of the two models provides a more accurate description of the empirical data , we use the results of the ks test as inputs in a bayesian model selection analysis @xcite .
bayes rule states that @xmath27 where @xmath28 is the posterior probability of selecting model @xmath29 given an observation @xmath30 , @xmath31 is the probability of observing @xmath30 given a model @xmath29 , and @xmath32 is the prior probability of selecting model @xmath29 . assuming no prior knowledge about the correctness of the power - law and log - normal models , one would select @xmath33log - normal@xmath34power - law@xmath35 for each model . however , to eliminate any bias on our part , we perform the bayesian model selection analysis for two cases : ( _ i _ ) no prior knowledge , @xmath33log - normal@xmath34power - law@xmath35 , and ( _ ii _ ) the power - law model is far more likely to be correct , @xmath33power - law@xmath36 .
the availability of data for multiple users enables us to perform this analysis recursively to obtain posterior probabilities of selecting each model given the available data .
concretely , the analysis of the interevent times @xmath30 from user @xmath37 updates the posterior probabilities of the two models @xmath28 using eq .
( [ bayes ] ) .
these updated posterior probabilities are then used as prior probabilities for the next user @xmath38 . when all of the users have been included
, this analysis reveals the posterior probability of the model given all of the available data .
the bayesian model selection analysis demonstrates that the likelihood of the truncated power - law model being a good description of the data vanishes to zero when all data is considered ( fig .
[ ks_sent]c ) .
before we present our analysis of the waiting times , we must note that the database collected by eckmann et al .
@xcite and analyzed by barabsi @xcite is not particularly well - suited for identifying the waiting times for replying to an e - mail .
the data merely records that an e - mail was sent by user a to user b at time @xmath39 .
the data does not specify whether the e - mail from a to b is , in fact , a reply to a prior message .
imagine the following scenario : user a sends an e - mail to user b. three days later , user b sends an unrelated e - mail to user a. barabsi s approach @xcite , which we follow , is to classify this e - mail as a reply to the e - mail sent by user a three days earlier . as this case illustrates ,
_ the analysis of waiting times is significantly less reliable than that of interevent times . _ reference @xcite reports that the probability distribution of time intervals @xmath2 between receiving a message from a sender and sending another e - mail to that sender follows a power - law distribution @xmath40 with @xmath4 .
a cursory analysis of this result again reveals several problems .
1 . figure 2b of ref .
@xcite features three bins corresponding to waiting times @xmath41 seconds , an unphysical interval ( fig . [ fig2b]a ) .
the events in those bins account for 1% of all events
figure 2b of ref .
@xcite features two bins confined to waiting times @xmath42 second ( fig .
[ fig2b]b c ) while the data have a precision of one second @xcite . .
two of us sent about twenty replies to an e - mail already in our inboxes . to minimize the time required to do this
, we did not read the e - mail to which we were responding but simply wrote `` yes '' at the top of our reply and then clicked send .
we find that 6 seconds is the smallest waiting time feasible for a human . *
b * , reproduction of fig.2a of ref .
@xcite obtained with vistametrix @xcite and , * c * , the same data but with the boundaries of the bins clearly marked .
we assumed that the data points in ( b ) are placed in the middle of the bin .
note that there are two bins recording data for @xmath43 second , while the data have a resolution of one second .
the shaded bins indicate values with @xmath44 seconds consisting of 1% of the data . ]
we characterize the actual distribution of waiting times @xmath2 following the same procedure outlined in section [ sect : tau ] .
after parsing the data , we are left with 724 users which have sent at least 10 response e - mails over 83 days and have at least two waiting times in the interval @xmath45 seconds .
we then perform ks tests and bayesian model selection to determine whether the waiting times are better described by a power - law or log - normal distribution .
the bayesian model selection analysis demonstrates that the likelihood of the truncated power - law model being a good description of the data vanishes to zero when all data is considered ( fig .
[ ks_waiting ] ) . for three users in the database .
the top panels show data and power - law predictions over intermediate values of @xmath9 whereas the middle and bottom panels depict gaussian and double gaussian model predictions for the entire range of @xmath9 . * b**c * , scatter plot of the ratio of the two @xmath14 values for all available users depending on the number of e - mails sent for each user .
the larger circles highlighted in green , yellow , and red correspond to the data shown in ( a ) .
users outside the domain are indicated by the arrows .
* d**e * , recursively calculated posterior probability of accepting the power - law model for in comparison with the log - normal and double log - normal models .
we use bayesian model selection to recursively calculate the posterior probability of selecting the power - law distribution for two different prior probabilities : @xmath46 ( solid black ) and @xmath47 ( red dashed ) .
the posterior probability of selecting the power - law model vanishes after considering 140 and 49 of the 724 users for the log - normal and double log - normal comparisons , respectively . ]
analysis of the data for the users with the largest number of replies suggests that @xmath2 may actually be better described by a superposition of two log - normal peaks : the first peak which contains most of the probability mass typically corresponds with waiting times of an hour , and the second peak typically corresponds with waiting times of two days .
this finding prompted us to investigate whether the superposition of two log - normals would provide a better description of the data than a single log - normal .
the probability function in this case has the functional form : @xmath48\ , , \label{eqn : dln}\ ] ] where @xmath49 and @xmath50 are the means the two peaks , @xmath51 and @xmath52 are the standard deviations of the two peaks , and @xmath53 is the probability mass in the first peak . distributions .
after considering all of the 724 available users , the posterior probability of the power - law and log - normal vanishes . ] in order to conduct the ks tests and bayesian model selection , we must first estimate the parameters of the double log - normal distribution , eq .
( [ eqn : dln ] ) . unlike the earlier analyses , it is not possible to estimate the parameters of the distribution without performing a fit of eq .
( [ eqn : dln ] ) to the data .
we perform maximum likelihood estimation @xcite to determine the best estimate parametrization of eq .
( [ eqn : dln ] ) ; see appendix [ sect : mle ] for details .
after determining the parameters of the double log - normal distribution , we conduct ks tests and bayesian model selection as before , and we find that a double log - normal has a posterior probability of one when compared with the power - law model ( fig . [ ks_waiting ] ) .
in fact , if we consider all three candidate models simultaneously , we still find that the posterior probability of the double log - normal is one ( fig .
[ threeway ] ) .
recently , barabsi and co - workers @xcite have reinterpreted the definition of the waiting times introduced in ref .
barabsi and co - workers note that the actual waiting time should not be counted from the time the original e - mail was sent , but from the time the original e - mail was first read .
this appears perfectly logical , but the database under investigation does not provide us with information on when the user _ actually _ first read the e - mail .
in fact , as we explained earlier the database does not even provide information that would enable one to decide whether an e - mail is a reply to a previous message or whether it is a totally unrelated message . nonetheless , it is worthwhile to analyze in greater detail the manner in which the authors of refs .
@xcite measure the waiting time @xmath2 since they characterize it as an improvement over the original method @xcite . at time @xmath54
user a sends an e - mail to user b. at time @xmath55 , user b sends an e - mail . at time
@xmath56 , user b sends an e - mail to user a. the `` real '' waiting time is now defined as @xmath57 , instead of @xmath58 .
note that @xmath59 still is _ not _ the actual time when the user actually first read the e - mail .
, @xmath60 ) in figs .
1a b of ref . @xcite . * a * , reproduction of the empirical probability density ( open circles ) and purported model solution ( red line ) from fig .
1a of ref . @xcite using vistametrix @xcite . to match the model with the empirical data , the authors of ref .
@xcite claim that @xmath61 is actually @xmath62 ( arrow ) .
moreover , the authors of ref .
@xcite do not use the actual model solution from ref .
@xcite ( blue dashed line ) as claimed .
* b * , probability function for the empirical waiting times , the purported model prediction , and the actual model solution .
even if the purported model solution was correct , it is visually apparent that it does not match the large gap in waiting times between @xmath63 and @xmath64 seconds . ]
we find three major problems with the reported predictive ability of the priority queuing model to capture the peak , the power - law regime , and the exponential cut - off of the waiting time distributions .
first , we are troubled that the `` agreement '' for the peak at @xmath62 is obtained by making the transformation @xmath62 if @xmath65 , instead of @xmath61 as would be expected from the definition . in other words , to match the peak at @xmath62 , barabsi and co - workers state that @xmath66 . moreover , we are surprised that barabsi and co - workers claim to use the exact model solution to predict the empirical waiting times . unlike fig .
1a of ref .
@xcite , the exact probability density has a large , discontinuous drop at @xmath62 @xcite .
when we compare the data presented in ref .
@xcite with the actual solution , it is clear that the model does not , in fact , match the empirical data ( fig .
[ reply_critique]a )
. finally , there are no waiting time values for @xmath2 between 1 and 60 seconds whereas the priority queuing model predicts a smooth continuous decrease of the probability density function in that region . while the difference between the two functions is difficult to discern in the plot of ref .
@xcite , the difference is actually quite marked ( fig . [ reply_critique]b ) .
we also examined the priority queuing model presented to explain the reported power - law in e - mail communication @xcite .
this model is defined as follows .
an individual has a priority queue with @xmath67 tasks .
each task is assigned a priority @xmath68 drawn from a uniform distribution @xmath69 $ ] . at each unit time step , the user executes either the highest - priority task with probability @xmath70 or a randomly selected task with probability @xmath71 .
the executed task is then removed from the queue and a new task with priority @xmath68 , again drawn from @xmath72 , is added to the queue . for the sake of comparison of the model predictions with the empirical data ,
barabsi surmised that a user s queue consists of e - mails which require a response .
the model thus predicts the time @xmath2 that a message spends in the user s inbox prior to response . and ref .
@xcite , the interevent times are distributed with a heavy - tail . ]
we first address the deficiencies in the model s assumptions .
first , humans can only process a handful of pieces of information at any time @xcite
. however , many users of e - mail hold tens , hundreds , or even thousands of e - mails in their inbox which may require action .
it is therefore unrealistic to expect any user to account for each task s priority or to be able to carefully determine the absolute ( or even the relative ) priority of such a large number of tasks .
secondly , the priority queuing model does not account for the heterogeneities in interevent times revealed by our analysis in section [ sect : tau ] and reported in ref .
@xcite . in the priority queuing model ,
tasks are executed at each time step which means that the distribution of interevent times can be described by dirac s delta function ( fig .
[ modeltime ] ) . in queue after @xmath73 tasks have been executed where @xmath74 . the author of ref .
@xcite , considers the case of ( d ) @xmath75 and @xmath76 after @xmath77 tasks have been executed ( red dashed line ) . in this case , however , the model has not even reached steady - state . the important thing
to notice is that after a short transient time , the priorities become uniformly distributed in the very small interval @xmath78 $ ] denoted with the gray shading .
when new tasks are added to the queue from a uniform distribution on the interval @xmath79 $ ] , the vast majority of new tasks are executed immediately .
this feature of the priority queuing model is not representative of human behavior . ]
the priority queuing model also suffers from several unrealistic predictions .
first , the time for the model to reach steady - state increases as @xmath80 @xcite .
this means that for the case considered in ref .
@xcite ( @xmath75 , @xmath76 ) , the time to reach steady - state is on the order of @xmath81 tasks .
if a user operating according to those parameter values sends 100 e - mails a day ( a very large number of e - mails ) , it would take him 100,000 days @xmath82 300 years to reach steady - state .
it is also worthwhile to note that the results for the model in ref .
@xcite were not even obtained for steady - state , implying that the data is actually a mixture of different stages of the relaxation process of the model .
second , after reaching steady - state , the dynamics of the model become quite anomalous .
the priorities of the tasks in the user s queue converge to a uniform distribution @xmath83 $ ] ( fig .
[ x_evo ] ) , while new tasks arrive with a priority drawn from @xmath84 $ ] .
thus , in the limit @xmath85 , an e - mail user has a queue consisting of extremely low priority tasks and consequently performs new tasks with probability @xmath85 immediately upon arrival .
this results in a peak at @xmath86 that accounts for nearly all of the probability mass ( fig .
[ trans_ss_tauw ] ) .
clearly this situation is not representative of e - mail activity , let alone human behavior as ref .
@xcite claims .
( red ) and steady - state for times @xmath87 ( black ) .
notice that the likelihood that a task is executed after spending more than one unit of time on the queue is vanishingly small . ]
more fundamentally , the priority queuing model predicts a distribution of waiting times that decays as a power - law with an exponent @xmath88 @xcite whereas the actual data clearly rejects that description ; cf . section [ sect : tauw ] .
in fact , a superposition of two log - normals , one corresponding to waiting times of less than a day and another corresponding to a waiting time of several days provides an excellent description of the empirical data .
more importantly , that description agrees with the common experience of e - mails users : one replies to e - mails within the day , if not , within the next few days , and if not then , never .
here , we have quantitatively analyzed human e - mail communication patterns . in particular , we have found that the interevent times are well - described by a log - normal distribution while the waiting times are well - described by the superposition of two log - normal distributions .
we have simultaneously rejected the hypothesis that either quantity is adequately described by a truncated power - law with exponent @xmath0 .
we have also critically examined the priority queuing model proposed by barabsi to match the empirically observed waiting time distributions .
after detailed analysis , we conclude that neither the assumptions nor the predictions of the model are plausible .
we note that the model does not match the empirically observed waiting time distribution , and we therefore contend that the theoretical description of human dynamics is an open problem .
barabsi and coworkers have also examined the dynamics of letter writing @xcite , web browsing , library loans , and stock broker transactions @xcite .
they argue that these processes also follow power - law distributions and are consequences of similar priority queuing processes .
our analysis demonstrates that care must be taken when describing data with fat - tails , particularly when the apparent scaling exponent is close to one and the probability distribution is concave .
in maximum likelihood estimation , the likelihood function @xmath89 for distribution model @xmath29 given the data @xmath90 is @xmath91 where @xmath92 is the number of data points in the sample , the @xmath93 are the empirical data points and @xmath94 is the probability density function for the candidate model @xmath29 evaluated at each empirical data point .
we then maximize the likelihood @xmath89 to find the parametrization of the model distribution @xmath95 that best approximates the data .
for the double log - normal model , @xmath96double log - normal@xmath97 . to find the best estimate of the five parameters @xmath98
, we obtain preliminary estimates for @xmath99 and @xmath100 from the mean and standard deviation of @xmath90 and subsequently maximize likelihood @xmath89 to find the appropriate double log - normal parameters . in practice , however , one typically performs a minimization of @xmath101 @xcite . | following up on barabsi s recent letter to _ nature _ [ * 435 * , 207211 ( 2005 ) ] , we systematically investigate the time series of e - mail usage for 3,188 users at a university .
we focus on two quantities for each user : the time interval between consecutively sent e - mails ( interevent time ) , and the time interval between when a user sends an e - mail and when a recipient sends an e - mail back to the original sender ( waiting time ) .
we perform a standard bayesian model selection analysis that demonstrates that the interevent times are well - described by a single log - normal while the waiting times are better described by the superposition of two log - normals .
our analysis rejects the possibility that either measure could be described by truncated power - law distributions with exponent @xmath0 .
we also critically evaluate the priority queuing model proposed by barabsi to describe the distribution of the waiting times .
we show that neither the assumptions nor the predictions of the model are plausible , and conclude that a theoretical description of human e - mail communication patterns remains an open problem . |
the x - shooter instrument @xcite of the vlt is an echelle slit spectrograph that covers the very wide wavelength range from 0.3 to 2.5 by three arms called uvb ( 0.30 to 0.59 ) , vis ( 0.53 to 1.02 ) , and nir ( 0.99 to 2.48 ) .
this allows one to study various airglow bands simultaneously , as demonstrated by @xcite , who investigated 25 ( @xmath1-@xmath25 ) bands between 0.58 and 2.24 with @xmath1 from 2 to 9 . since x - shooter is an instrument for observing astronomical targets , airglow studies have to rely on archival , unprocessed spectra , which differ in terms of line of sight , exposure time , and slit - dependent spectral resolution . for this study , we have used the same data set that was selected and processed by @xcite .
the extended analysis of these data , therefore , focuses on the @xmath2 and @xmath3 bands ( see sect . [
sec : lines ] ) .
@xcite reduced archival data from october 2009 ( start of the archive ) to march 2013 using version v2.0.0 of the eso public pipeline @xcite .
the resulting wavelength - calibrated two - dimensional ( 2-d ) spectra were then collapsed into a 1-d spectrum by applying a median along the spatial direction .
for the flux calibration , response curves were created based on data of spectrophotometric standard stars @xcite corrected for molecular absorption @xcite and atmospheric extinction @xcite . due to the instability of the flat - field calibration lamps for the correction of the pixel - to - pixel sensitivity variations in the 2-d echelle spectra , the fluxes of long wavelengths beyond 1.9 including the bands and are relatively uncertain . for 77% of the data ,
this could be improved by means of spectra of standard stars with known spectral shapes , i.e. rayleigh jeans slopes . from the processed data ,
a sample of 343 vis- and nir - arm spectra was selected .
the selection criteria were a minimum exposure time of 3 , a maximum slit width of 1.5@xmath26 corresponding to @xmath27 for the vis arm and 3300 for the nir arm , and a difference in start , end , and length of the exposure between vis and nir arm of less than 5 .
the latter was necessary since the observations in the different x - shooter arms can be performed independently .
moreover , 21% of the combined vis- and nir - arm spectra were rejected due to unreliable @xmath0 in at least one of the 25 investigated bands .
this can be caused by low data quality , an erroneous data reduction , or critical contaminations by residuals of the observed astronomical object .
36 spectra of the final sample could only be used up to 2.1 , which excludes , because a so - called @xmath28-blocking filter was applied @xcite .
the 343 selected spectra show a good coverage of nighttimes and seasons and , therefore , appear to be representative of cerro paranal . for more details , see @xcite . during evening ( solid lines ) and morning twilight ( dashed lines ) for the latitude range from 23 to 26s and the years from 2009 to 2013 .
the plot shows profiles for five different solar zenith distances @xmath29 from 94 to 110 .
the vers decrease with increasing @xmath29.,width=302 ] in as a function of the longitude distance to cerro paranal @xmath30 in degrees for the altitudes 85 ( long dashed ) , 90 ( solid ) , and 95 ( short dashed ) and bin sizes of 4 and 1 .
the averages were derived from saber data for the latitudes between 23 to 26s , the years from 2009 to 2013 , and zenith distances of at least 100 .
the final saber sample selection criterion @xmath31 is visualised by vertical bars.,width=302 ] our analysis of ver and temperature profiles is based on version 2.0 of the limb sounding products of the saber multi - channel photometer on timed @xcite .
we used ver profiles from the 1.27 channel of @xmath3 and the two -related channels centred on 1.64 and 2.06 .
we took the `` unfiltered '' data products , which are corrected vers that also consider the emission of the targeted molecular band(s ) outside the filter @xcite . for @xmath3 ,
the correction factor is 1.28 @xcite .
finally , we used the saber kinetic temperature @xmath32 profiles . in version 2.0 , they are based on the limb radiance measurements in the 4.3 and 15 broad - band channels and a complex retrieval algorithm involving non - lte radiative transfer calculations @xcite . for our analysis , we could use the ver and @xmath32 profiles between 40 and 110 because only for this altitude range all data are available .
the lower and upper limits are related to the vers and @xmath32 , respectively .
we mapped all profiles ( having a vertical sampling of about 0.4 ) to a regular grid with a step size of 0.2 . to cover a similar period as the x - shooter sample ( see sect .
[ sec : xshooter ] ) , we only considered saber data of the years 2009 to 2013 .
the x - shooter data set is a nighttime sample with a minimum solar zenith distance @xmath29 of 104 . for this reason
, we only selected saber data with a minimum @xmath29 of 100 . with this limit
, it is guaranteed that the @xmath3 main emission peak is located in the mesopause region , i.e. it is typical of nighttime conditions
. the fast change of the @xmath3 ver profile at twilight is illustrated in fig .
[ fig : vero2a_twilight ] .
as discussed in sect .
[ sec : intro ] , the daytime population related to ozone photolysis with a maximum around the stratopause decays after sunset .
the late nighttime profile remains relatively constant until the middle atmosphere is illuminated by the sun at dawn . since airglow emissions and temperatures significantly depend on geographic location ( e.g. * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ) due to differences in the solar irradiation and activity of tides , planetary and gravity waves , we selected only saber data within the range from 23 to 26s .
cerro paranal is located at 24.6s . with this criterion and
the exclusion of 14 profiles which were suspicious in terms of the ver peak positions for the two channels , the resulting sample includes 27796 measurements .
we further restricted the data set by also considering longitudinal variations .
figure [ fig : th_lon ] shows the saber @xmath32 for the altitude levels 85 , 90 , and 95 as a function of longitudinal distance @xmath33 to cerro paranal ( 70.4w ) . on average , there are variations of several and changes of the variability pattern depending on altitude .
for this reason , we limited the sample by the criterion @xmath34 , which resulted in 1685 profiles . a further reduction of the longitude range is not prudent because then there could be too few profiles for representative results for time and/or seasonal averages .
we also checked the results presented in the subsequent sections with respect to the longitude limit ( see sect .
[ sec : tempcorrect ] ) .
indeed , the agreement of x - shooter and saber variations does not significantly improve with more restrictive limits . also note that the saber tangent point can easily move by more than 100 during a profile scan .
moreover , the timed yaw cycle of 60 days @xcite causes large gaps in the distribution of local times and days of year . for odd months ( e.g. march ) , the period around midnight is well covered at cerro paranal .
even months show a better evening / morning coverage .
therefore , limitations in the correlation of x - shooter and saber data are unavoidable .
this study mainly focuses on the quantification of non - lte contributions to @xmath0 .
the following section describes the different steps of this analysis , which relies on x - shooter - related and @xmath0 measurements and the saber - related correction of temperature differences by different band - specific emission profiles . the wavelength range , spectral resolution , and quality of the x - shooter spectra ( sect .
[ sec : xshooter ] ) allowed us to derive @xmath0 from line intensities of 25 bands , @xmath2 , and @xmath3 .
[ [ sec : oh ] ] the @xmath0 related to are based on the line selection and intensity measurements of @xcite .
the selection only includes lines originating from rotational levels @xmath35 of both electronic substates @xmath11 and @xmath12 to minimise the non - lte effects that increase with @xmath5 @xcite .
moreover , it focuses on @xmath36-branch lines ( related to an increase of the total angular momentum by 1 ) since these lines are well separated in the x - shooter spectra .
the @xmath37 doublets ( e.g. * ? ? ?
* ) are not resolved . due to reasons like blending by airglow lines of other bands or strong absorption in the lower atmosphere ( e.g. by water vapour ) , the number of suitable @xmath36-branch lines varied between 2 and 8 for the 25 considered bands . in the case of
, an additional @xmath38-branch line was measured in order to have a minimum of three lines for quality checks .
the measured intensities of all the lines were corrected for molecular absorption in the lower atmosphere by means of a transmission curve for cerro paranal computed with the radiative transfer code lblrtm @xcite by @xcite .
the absorption by water vapour was adapted based on data for the precipitable water vapour ( pwv ) retrieved from standard star spectra @xcite .
for the line shape , doppler - broadened gaussian profiles related to a temperature of 200 were assumed .
the measurement of @xmath0 is based on the assumption of a boltzmann distribution of the @xmath5-related populations , which are derived from @xmath39 , i.e. the measured intensity @xmath40 divided by the @xmath5-dependent statistical weight @xmath41 ( number of sublevels ) times the einstein-@xmath16 coefficient for the line . in this case , the natural logarithm of this population of a hyperfine structure level of @xmath5 , which we call @xmath42 , is a linear function of the level energy @xmath43 .
the line slope is then proportional to the reciprocal of @xmath0 @xcite .
@xcite used the molecular parameters @xmath43 , @xmath16 , and @xmath41 from the hitran2012 database @xcite , which are based on @xcite for the investigated lines @xcite , and from @xcite .
we will mostly focus on the hitran - based parameters since they appear to provide more reliable @xmath0 .
@xcite corrected the resulting @xmath0 for the non - lte effects caused by the different band - dependent line sets ( see sect .
[ sec : intro ] ) by introducing a reference line set consisting of the three @xmath44- ( @xmath11 ) and @xmath45-branch lines ( @xmath12 ) with @xmath46 .
the temperature changes for the transition from the actual to the reference line set were calculated based on the sample mean @xmath0 of the 16 most reliable bands .
the resulting set - specific @xmath47 were then applied to the temperatures for the individual observations . for this study
, we used a different reference line set only including the three @xmath44-branch lines .
the corresponding results are easier to interpret since the mixing of @xmath44- and @xmath45-branch lines causes a negative non - lte effect , i.e. @xmath0 decreases by the lower relative population of the @xmath12 state , which has higher @xmath43 than @xmath11 .
we calculated the correction to the new reference set based on the four very reliable bands , , , and ( cf . * ? ? ?
this choice is not critical since the slight @xmath1 dependence of the energy differences for the rotational levels has only a minor impact on the resulting @xmath48 .
the @xmath0 only based on the @xmath44 branch are @xmath49 higher than those given by @xcite if the molecular parameters are taken from hitran . in the case of molecular data from @xcite
, @xmath47 is @xmath50 .
if we focused on the fainter @xmath45-branch lines , the corrections would be @xmath51 and @xmath52 , respectively .
our study is based on mean @xmath0 derived from bands with the same @xmath1 , which show very similar intensity and @xmath0 variations .
we took the @xmath1-specific @xmath0 from @xcite ( corrected for the different reference line set ) for each of the 343 sample spectra .
note that the @xmath8 mean values of 36 spectra with reduced wavelength range do not include ( see sect .
[ sec : xshooter ] ) .
the effect on the sample mean @xmath0 is negligible .
the very faint band is only considered by a fixed @xmath47 for @xmath9 derived from the sample
mean ( see * ? ? ?
excluding errors in the molecular parameters , the uncertainties of the @xmath0 ( @xmath1 ) based on the sample mean intensities are between 1.0 for @xmath9 and 2.0 for @xmath53 .
these values are mainly based on the @xmath0 variation for fixed @xmath1 and are related to hitran data . for molecular data from @xcite ,
the corresponding uncertainties range from 1.8 for @xmath54 to 3.6 for @xmath53 .
the @xmath0 uncertainties for individual observations are higher than the stated values due to measurement errors depending on the signal - to - noise ratio
. for the 189 vis - arm sample spectra taken with the standard slit width of 0.9@xmath26 .
the original spectrum was corrected for solar absorption lines in the continuum ( green areas ) .
pairs of @xmath56 and @xmath57 lines with even upper rotational levels @xmath5 from 0 to 18 are marked by circles .
note that @xmath56 ( 0 ) does not exist .
the filled symbols indicate lines that were used for the derivation of @xmath0.,width=302 ] @xmath0 under the assumption of a boltzmann distribution for the different rotational levels .
the abscissa shows the hitran - based energy of a given level in @xmath58 .
the ordinate displays the natural logarithm of the hitran - based population of a hyperfine - structure level relative to 1 rayleigh second . resulting from @xmath56 and @xmath57 line intensities , the mean populations of the even rotational levels with @xmath5 from 0 to 18 for the full x - shooter sample
are shown .
filled symbols mark reliable data points that were used for the derivation of @xmath0 ( see fig .
[ fig : spec_o2b01 ] ) , which is illustrated by the solid regression line.,width=302 ] as discussed in sect .
[ sec : intro ] , @xmath2 at @xmath14 is the most popular non- molecular band for @xmath0 retrievals in the mesopause region .
it can also be measured in the x - shooter spectra . with a minimum resolving power of 5400 in the vis arm ,
the individual lines of @xmath2 are not separated . however , in the @xmath36 branch , blended lines originate from the same upper rotational level of @xmath17(@xmath59 ) ( see fig .
[ fig : spec_o2b01 ] ) .
this allows @xmath0 to be directly measured without the requirement for @xmath0-dependent synthetic spectra and data fitting , which is the standard approach ( e.g. * ? ? ?
* ; * ? ? ?
consequently , the systematic uncertainties of the @xmath0 measurement are expected to be significantly lower .
the two blended sub - branches are @xmath56 and @xmath57 ( e.g. * ? ? ?
* ) , where the rotational level @xmath60 increases by 1 and the total angular momentum @xmath61 changes by 0 or @xmath62 depending on the spin orientations in the lower state @xmath19(@xmath63 ) . due to the symmetry of the molecule , the upper state rotational levels @xmath5 have to be even .
the selection rules also forbid @xmath56 ( 0 ) , i.e. there is only a single @xmath36-branch line for @xmath64 . for the @xmath0 determination ,
the intensities of the @xmath36-branch line pairs with @xmath5 from 0 to 18 were measured ( see fig . [
fig : spec_o2b01 ] ) .
the lines with @xmath5 higher than 18 are already relatively faint and strongly blended with lines .
before the line intensities could be determined , the continuum had to be measured and subtracted . since an important part of the continuum
is caused by zodiacal light ( sunlight scattered by interplanetary dust particles ) and/or scattered moonlight ( e.g. * ? ? ?
* ) , the strong solar absorption line at 866.2 is critical for the continuum window between the @xmath65 and 8 lines ( see fig . [
fig : spec_o2b01 ] ) . for this reason ,
the solar flux atlas of @xcite was used to calculate a continuum correction , which required the adaption of the solar transmission to the corresponding resolution of the x - shooter spectra and the determination of the variable fractional contribution of a solar - type spectrum to the continuum at @xmath2 .
the latter was derived from the depth of the absorption in the x - shooter data compared to the solar spectrum and is important since there is also an airglow continuum related to reactions involving @xcite .
after the removal of the solar lines , the median continuum level was measured in nine fixed windows , which were optimised for the setups with the lowest spectral resolution , i.e. a slit width of 1.5@xmath26 .
the range between the @xmath66 and 6 lines could not be used for the continuum determination due to significant contributions by lines .
the resulting continuum fluxes were then linearly interpolated and subtracted from the x - shooter spectra . finally , the intensities of the line pairs were integrated within limits which were adapted to the spectral resolution . since the @xmath36 branch of @xmath2 is not affected by molecular absorption ( cf .
[ sec : oh ] ) , the corresponding correction could be neglected .
as the observations were carried out at different zenith distances , the intensities were corrected to be representative of the zenith by applying the equation of @xcite ( see * ? ? ?
the resulting sample mean populations of the rotational levels @xmath64 to 18 are shown in fig .
[ fig : y_ep_o2b01 ] . similar to described in sect .
[ sec : oh ] and @xcite , they correspond to @xmath39 , where @xmath40 and @xmath16 are sums for the line pairs .
the molecular parameters were taken from hitran2012 @xcite , which is essentially based on @xcite for @xmath2 .
most @xmath42 values show a good correlation with @xmath43 .
the outliers can be explained by contamination from lines .
this result supports the assumption that non - lte effects are not relevant for the rotational level population .
the radiative lifetime of about 210 derived from the hitran data appears to be sufficiently long .
the optimal line set for the calculation of @xmath0 turned out to consist of @xmath57 ( 0 ) and the @xmath36-branch line pairs with @xmath5 from 4 to 14 .
for the mean populations shown in fig .
[ fig : y_ep_o2b01 ] , the corresponding systematic @xmath0 uncertainty is only 1.8 .
this line set could be applied to 326 spectra . due to instrumental issues or contamination by astronomical objects ,
15 fits were significantly better by using only six instead of seven lines / line pairs . in two cases ,
only five emission features were considered .
the optimal line sets were determined by an automatic iterative procedure . in order to avoid systematic differences depending on the line set ,
the @xmath0 based on less than seven data points were corrected .
for this , the reliable spectra were used to derive the differences between the @xmath0 for the full and each reduced line set .
the mean amount of the correction was 0.5 .
we applied further corrections to the measured @xmath0 .
first , we investigated possible systematic errors related to the widths of our continuum windows , which also had to be suitable for the low - resolution data . for 23 reliable high - resolution spectra taken with the 0.7@xmath26 slit
, we could significantly increase the number of continuum pixels .
the resulting temperatures were about 0.4 lower than those obtained based on the standard windows .
as a consequence , we decreased the @xmath0 of the whole sample by the same amount .
second , we corrected the influence of the slit width on @xmath0 by adapting the resolution of the 23 0.7@xmath26 spectra to those taken with a 0.9@xmath26 , 1.2@xmath26 , or 1.5@xmath26 slit and deriving the @xmath0 differences .
this resulted in temperature increases of about 0.2 , 0.4 , and 1.0 , respectively .
all these corrections are relatively small , which makes us confident that our @xmath0 for @xmath2 are robust . for the 177 nir - arm sample spectra taken with the standard slit width of 0.9@xmath26 .
@xmath68 and @xmath69 lines with odd upper rotational levels @xmath5 from 13 to 23 are marked by circles and diamonds , respectively .
the filled symbols indicate lines that were used for the derivation of @xmath0.,width=302 ] @xmath0 .
the plot is similar to fig .
[ fig : y_ep_o2b01 ] . resulting from @xmath68 ( circles ) and @xmath69 ( diamonds ) line intensities ,
the mean populations of the odd rotational levels with @xmath5 from 13 to 23 for the full x - shooter sample are shown .
filled symbols mark reliable data points that were used for the derivation of @xmath0 ( see fig . [
fig : spec_o2a00 ] ) , which is illustrated by the solid regression line.,width=302 ] intensity relative to mean after midnight as a function of the time since sunset for the x - shooter sample ( dots ) .
the solid line shows the best fit of a constant and an exponential function to the data .
the plot also displays the corresponding fit results for the saber sample without longitude restriction and lower intensity integration limits of 40 ( dashed ) and 80 ( dash dotted ) . for the three fits ,
the resulting radiative lifetimes @xmath15 of the daytime population are shown in the legend.,width=302 ] the x - shooter nir arm covers the strong @xmath3 band at 1.27 ( see fig . [ fig : spec_o2a00 ] ) , which contains nine branches related to the dominating magnetic dipole transitions ( e.g. * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ) . with the minimum resolving power of 3300 in the x - shooter nir arm ,
the lines of two branches are sufficiently separated to allow individual intensity measurements .
this could not be performed by @xcite , who have done the only known @xmath3 @xmath0 measurements ( see sect . [
sec : intro ] ) .
the suitable branches are @xmath68 and @xmath69 ( see fig .
[ fig : spec_o2a00 ] ) , which have the same upper levels but differ by a change of @xmath60 by @xmath70 and @xmath71 , respectively . to fulfil the selection rules for @xmath72
, these @xmath73 have to be accompanied by an opposite change of the spin @xmath74 by 1
. only transitions with odd @xmath5 originate from @xmath18 . for the line measurements
, we focused on intermediate @xmath5 between 13 and 23 , which are sufficiently strong and not too close to the highly blended branches in the band centre .
the line intensities were measured in a similar way as described in sect .
[ sec : o2b ] for @xmath2 .
the effect of solar lines on the continuum could be neglected . however , since the lower state @xmath19(@xmath59 ) of @xmath3 is the ground state , the resulting line intensities had to be corrected for the strong self - absorption at low altitudes .
the procedure was similar to the approach for described in sect .
[ sec : oh ] and @xcite .
since the transmission curves for cerro paranal with a resolving power of about @xmath75 calculated by @xcite were not accurate enough for @xmath3 , we recalculated the absorption with the maximum resolving power of @xmath76 inherent to the radiative transfer code lblrtm @xcite . for a nominal temperature of 190k for the @xmath3 emission layer
, we obtained zenith transmissions between 0.14 and 0.80 for the measured @xmath68 lines and between 0.57 and 0.96 for the corresponding @xmath69 lines .
the use of a fixed temperature for the doppler broadening calculation has only a minor impact on the resulting @xmath0 .
we performed a second iteration of the transmission correction using the measured @xmath0 .
the mean @xmath0 change compared to the first iteration was only about @xmath77 .
figure [ fig : y_ep_o2a00 ] shows the sample mean populations for the upper states of the measured @xmath3 lines .
the populations were calculated by means of the molecular parameters of the hitran2012 database @xcite , which are based on several studies @xcite .
most lines indicate a tight linear relation of @xmath42 and @xmath43 .
the good agreement of the populations derived from @xmath68 and @xmath69 lines with the same @xmath5 is also convincing .
the few outliers can be explained by blends with lines from other branches and lines .
this result supports the expected good thermalisation of the rotational level populations of @xmath18(@xmath59 ) due to the long radiative lifetime .
the hitran data indicate an einstein-@xmath16 coefficient of @xmath78 ( cf .
[ sec : intro ] ) , which corresponds to @xmath79 .
we can directly constrain @xmath15 from our data . since the daytime production of @xmath18 by ozone photolysis stops at sunset ,
the subsequent decrease of the @xmath3 intensity is an indicator of @xmath15 ( e.g. * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
since there are also @xmath18 losses by collisions , the derived lifetime is only a lower limit .
figure [ fig : into2a_time ] shows the @xmath3 intensities normalised to the mean after midnight as a function of the time after sunset for our x - shooter data set .
the steep intensity decrease at the beginning of the night is evident .
we fit this decrease by an exponential function plus 1 .
the latter reflects the population related to the nighttime @xmath18 excitation , which is considered to be constant .
this assumption is reasonable since the average nighttime variations are expected to be much weaker than those of the daytime - related population . the resulting lower lifetime limit is @xmath80 , which is within the range of other ground - based measurements @xcite .
we can also derive @xmath15 from our saber data set ( sect .
[ sec : saber ] ) . for the latitude range from 23 to 26s , @xmath29 greater than 100 , and a minimum height of 40
, we obtained about 48 , which is close to the x - shooter value .
the saber best - fit curve in fig .
[ fig : into2a_time ] is shifted to earlier times due to the sparse twilight sampling of the x - shooter data with a minimum @xmath29 of 105 .
this could also contribute to the small difference of the @xmath15 values .
as illustrated by fig .
[ fig : vero2a_twilight ] , the decay of the daytime population is faster at lower altitudes , where collisions are more frequent .
therefore , the resulting @xmath15 depends on the considered altitude range . for a lower limit of 80
, we obtained about 60 ( see fig .
[ fig : into2a_time ] ) . also based on saber data
, @xcite derived an einstein-@xmath16 coefficient consistent with the theoretical value for a height of 85 .
as shown by figs . [ fig : spec_o2a00 ] and [ fig : y_ep_o2a00 ] , the optimal line set for the @xmath0 retrieval consists of @xmath69 ( 13 ) , @xmath68 ( 15 ) , @xmath69 ( 15 ) , @xmath68 ( 17 ) , @xmath69 ( 17 ) , and @xmath68 ( 21 ) . in principle ,
the two lines with @xmath81 could also have been included since the @xmath0 derived from the mean populations in fig .
[ fig : y_ep_o2a00 ] only differs by @xmath77 from the result for our standard line set .
however , they are relatively faint , which causes significant uncertainties if the @xmath0 are obtained from individual spectra . using the mean populations and the extended line set , the systematic @xmath0 uncertainty from the regression analysis is about @xmath82 .
the resulting @xmath0 were corrected for different systematic effects . for this purpose ,
the 47 spectra taken with the narrow 0.4@xmath26 and 0.6@xmath26 slits were used .
first , we checked the quality of the crucial self - absorption correction .
this was achieved by multiplying the optical depth of all lines by a line - independent but variable factor and recalculating @xmath0 to find the results with the lowest regression uncertainties .
for the 47 test spectra , this procedure resulted in an optimal factor of @xmath83 and a @xmath0 decrease by @xmath84 .
this small correction confirms the reliability of our radiative transfer calculations .
then , similar corrections as for @xmath2 ( sect .
[ sec : o2b ] ) were performed . with continuum windows optimised for the narrow slits instead of the wide 1.5@xmath26 slit
, we obtained a @xmath85 of @xmath86 .
the effect of the spectral resolution on the line measurements resulted in @xmath0 corrections of @xmath87 , @xmath88 , and @xmath89 for the 0.9@xmath26 , 1.2@xmath26 , and 1.5@xmath26 slits . taking
the 0.6@xmath26 slit as a reference , the @xmath0 of the three spectra obtained with a 0.4@xmath26 slit were also corrected and increased by 3.5 .
this relatively large correction is probably related to the starting resolution of underlying faint lines . in general , the applied corrections are small and partly cancel each other out .
the systematic @xmath0 uncertainty slightly increases from 2.1 to 2.2 when the uncertainties of the @xmath0 corrections are also considered . in order to compare the @xmath0 derived from the discussed bands ( sect .
[ sec : lines ] ) , the differing height distributions of the emission have to be considered since the measured bands probe different parts of the mesopause temperature profile .
since the ground - based x - shooter data do not provide this information , we obtained ver and temperature profiles representative of cerro paranal from the satellite - based saber data @xcite .
[ [ sec : ohprof ] ] as described in sect .
[ sec : saber ] , saber has two -related photometric channels centred on 1.64 and 2.06 @xcite , which essentially cover the , and , bands , respectively ( see * ? ? ?
the saber profiles can not be used directly since the emission height depends on @xmath1 @xcite , which ranges from 2 to 9 for our x - shooter data ( sect .
[ sec : oh ] ) .
for this reason , we had to derive @xmath1-related profiles from the saber data .
it is reasonable to assume that the @xmath1-dependent emission peak altitudes are equidistant ( cf .
* ) ( see also sect . [
sec : correl ] ) . even if there are significant deviations from this assumption , the maximum errors should only be of the order of a few 100 , which is still relatively small compared to the typical fwhm of 8 to 9 of the ver profiles ( e.g. * ? ? ?
then , the @xmath1-dependent emission peak altitudes can be derived by means of effective @xmath1 for the two saber channels .
we achieved this by calculating line spectra from the vibrational level populations and @xmath0 given by @xcite for hitran data and convolving them with the saber filter curves @xcite . in this way
, we obtained fractional contributions of the different bands and , finally , the effective @xmath1 , which are 4.57 and 8.29 for the 1.64 and 2.06 channels , respectively .
although we neglected the higher relative populations at high @xmath5 compared to those related to @xmath0 ( e.g. * ? ? ?
* ) and did not consider the variability of the relative populations @xcite , the resulting values are sufficiently robust for our purpose .
the corresponding altitude uncertainty should be well below 100 . from @xmath90 and the emission peak altitudes @xmath91 of both channels , we derived a mean altitude difference of 0.37 for @xmath92 and our cerro paranal saber sample consisting of 1685 measurements ( sect .
[ sec : saber ] ) .
this is in good agreement with data from other studies @xcite . since @xmath91 is not very accurate due to a step size of 0.2 ( sect .
[ sec : saber ] ) , we used the centre of the range limited by the interpolated half maximum positions @xmath93 as a more robust measure to calculate the @xmath1-dependent profile altitudes . in this case , the mean @xmath94 altitude difference is 0.39 with a standard deviation @xmath95 of 0.15 . and the mean profile of the 2.06 saber channel peaks higher than the 1.64 one.,width=302 ] the ver profiles of the two channels ( see fig . [
fig : rver_all ] ) show slightly different shapes . for the cerro paranal saber sample ,
the mean fwhm are 8.1 and 9.1 for the 1.64 and 2.06 channels , respectively .
the corrresponding @xmath95 are 1.9 and 2.1 . for a better comparability of the @xmath1-dependent profiles and the related temperatures
( see sect .
[ sec : tempcorrect ] ) , we used merged profiles . for all saber measurements , we shifted both profiles normalised to the same integrated intensity to the calculated @xmath93 for each @xmath1 and averaged them .
the resulting mean profiles for the cerro paranal saber sample are shown in fig .
[ fig : rver_all ] .
they have a fwhm of 8.6 with a @xmath95 of 1.9 . the average @xmath93 range from 86.2 ( @xmath53 ) to 89.0 ( @xmath8 ) with a scatter between 1.3 and 1.4 .
the profiles are asymmetric .
the mean @xmath93 are about 0.6 ( @xmath96 ) higher than the corresponding @xmath91 . for the effective height @xmath97 , i.e. the ver - weighted altitude
, the average difference is 1.2 .
the calculation of @xmath97 excluded altitudes with negative vers and @xmath98 ( see also sect . [
sec : tempcorrect ] ) .
saber does not provide ver profiles for @xmath2 .
for this reason , we assumed a simple gaussian profile , which appears to be justified by the relatively small measured asymmetries ( see * ? ? ? * ) .
based on values from the literature ( see sect . [
sec : intro ] ) , where the measurements are usually based on @xmath2 ( e.g. * ? ? ?
* ) , peak height and fwhm were set to 94.5 and 9 km ( see fig . [
fig : rver_all ] ) . since our @xmath2 profile is fixed , we checked the influence of variations of @xmath91 and fwhm on the derived temperatures ( see sect .
[ sec : tempcorrect ] ) . for this analysis
, we assumed the same variations of these parameters as described in sect .
[ sec : ohprof ] for . as a result
, we found for individual profiles an average @xmath95 of 2.0k .
however , for the sample mean ( where transient waves vanish ) , @xmath95 is only 0.5k .
since we are mainly interested in average properties , the exact choice of @xmath91 and fwhm for @xmath2 is not critical for our study .
saber has a dedicated channel for @xmath3 ( see sect .
[ sec : saber ] ) .
this was one of the reasons for selecting saber for this study since this band shows strong variations of the emission profile . in fig .
[ fig : vero2a_twilight ] , this was illustrated for twilight conditions . as our x - shooter data set is a nighttime sample ,
we show the average emission profiles for the cerro paranal saber sample at five different time intervals in fig .
[ fig : rvero2a_nbina ] . for all periods ,
most of the emission originates close to the emission layer ( see also fig . [
fig : rver_all ] ) , which is crucial for a reliable @xmath0 comparison ( see sect . [
sec : tempcorrect ] ) .
however , the emission peak indicates a clear transition from about 84 to 90 in the first half of the night .
the early - night peak is obviously caused by a daytime ozone ( and its photolysis ) maximum in the upper mesosphere , whereas the late - night peak appears to be related to atomic oxygen recombination and subsequent collisions ( e.g. * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ) ( see also sect . [
sec : intro ] ) .
ver profiles scaled to the emission peak for the cerro paranal saber sample .
mean profiles for five different nocturnal local time ( lt ) periods are shown ( see legend ) . for comparison ,
the @xmath3-based reference profile is also displayed ( light solid line).,width=302 ] in for ( @xmath9 ) and @xmath3 for the cerro paranal saber sample .
the mean values for five different nighttime lt periods are shown by big circles of different colours connected by lines ( see legend ) .
the colours of the individual data points correspond to those of the related mean values.,width=302 ] the nocturnal variations of @xmath3 are illustrated in fig .
[ fig : heff_nbina ] . here , the effective heights @xmath97 are given compared with those of ( @xmath9 ) .
the calculation of the @xmath3 @xmath97 excluded altitudes with negative vers , which essentially appear in the lower mesosphere ( see fig . [
fig : rvero2a_nbina ] ) .
additional height limits as for ( see sect . [
sec : ohprof ] ) were not applied because of the strong profile variations .
the general profile limits of 40 and 110 ( sect .
[ sec : saber ] ) are not an issue for the nighttime .
figure [ fig : heff_nbina ] shows a relatively large scatter of 2.4 in the @xmath3 @xmath97 at the beginning of the night , which is not seen for ( @xmath9 ) ( @xmath96 ) . in the course of the night , the deviation ( 4.4 ) and scatter become smaller until a convincing linear relation with ( @xmath99 ) is established .
the correlation coefficient for the last period is @xmath100 , which is significantly higher than @xmath101 for the first period .
the increase of @xmath97 for @xmath3 is partly compensated by a simultaneous rise of the layer of 1.1 between the first and last period . the smallest discrepancy and scatter of @xmath102 and @xmath103 for @xmath3
are found for the fourth nighttime interval .
the result for the discrepancy will change with a different ( @xmath1 ) .
the analysis of the @xmath97 data confirms our expectation that an and @xmath3 @xmath0 comparison is possible with only minor corrections for a wide range of nighttimes ( see sect . [
sec : tempcorrect ] ) .
a correction is not negligible even for equal @xmath97 since the profile shapes and fwhm can be very different ( see fig .
[ fig : rver_all ] ) .
the average emission profile fwhm for @xmath3 and are 12.9 and 8.7 , respectively .
there is still a difference of 3.6 ( corresponding to 11.7 and 8.0 ) for the fourth nighttime period , which shows the best agreement in terms of @xmath97 .
the emission profile widths for @xmath3 and decrease during the night .
the difference is 4.6 and 2.3 , respectively , if the fwhm of the first and last period are compared .
the emission profiles of @xmath3 and the negligible non - lte contributions to the corresponding @xmath0 make this band ideal for an analysis of the -related @xmath0 .
however , the strong profile variations do not allow one to assign a reference altitude for the @xmath3-related temperatures .
therefore , we also introduce a reference profile , which is based on @xmath3 profiles observed in the second half of the night .
since the corresponding median @xmath91 and fwhm are 89.8 and 10.9 , respectively , we selected 90.0 and 11.0 as reference values . observed saber ver profiles with similar properties
were then averaged .
as the cerro paranal saber sample is too small for this , we used the larger sample without longitude restriction ( see sect . [
sec : saber ] ) .
we considered all 297 profiles with @xmath91 between 88 and 90 and fwhm between 10.8 and 11.2 . to avoid any broadening of the averaged profile ,
the limits for the fwhm were more restrictive . in the case of @xmath91 , this effect
could be prevented by shifting the profiles to the reference height before the averaging .
figures [ fig : rver_all ] and [ fig : rvero2a_nbina ] show the resulting reference profile , which is not significantly affected by the decaying daytime @xmath18 population .
the profile is slightly asymmetric with @xmath104 and @xmath105 .
the final step of the preparation of @xmath0 data for a comparison is the correction of differences in the emission profiles .
this requires some knowledge of the true temperature profile in the mesopause region .
saber provides this kind of data ( @xmath32 ) , which are based on -related measurements and non - lte radiative transfer calculations ( see sect . [
sec : saber ] ) . for the derivation of the sample - averaged temperature differences for slightly deviating emission profiles ,
the @xmath32 uncertainties are probably small .
however , we also want to directly compare x - shooter @xmath0 and saber @xmath32 . for this reason ,
the temperature uncertainties have to be known in detail . according to @xcite and j. m. russell iii ( personal communication , 2015 ) , the total uncertainties for single profiles are about 5 at 90 and tend to increase with altitude .
since we are interested in average properties , we can neglect the statistical errors , which leaves systematic uncertainties of 3 to 4 in the relevant altitude range
. this may further decrease if it is considered that systematic errors can also partly cancel out @xcite .
for this reason , we assume a relevant uncertainty of 2.5 as input for sect . [
sec : results ] . as an additional check
, we compared the preferred version 2.0 data to those of version 1.07 , which is possible for saber data taken until the end of 2012 .
the old @xmath32 show deviations from the new ones , which are between @xmath106 for ( @xmath53 ) and @xmath107 for @xmath2 . assuming that the new @xmath32 are more correct , the changes are consistent with the assumed errors . to obtain effective temperatures
@xmath108 from the saber @xmath32 for the different bands , we used a similar procedure as described for @xmath97 ( sects .
[ sec : ohprof ] and [ sec : o2aprof ] ) , i.e. the @xmath32 of the different altitudes were weighted depending on the related vers for each emission layer
. altitudes with negative vers were rejected .
moreover , weights were set to zero for altitude distances of more than 15 from the @xmath93 .
the difference between the resulting @xmath108 of a given band and a desired reference band @xmath109 indicates the change of the effective true temperature by deviating altitudes , widths , and shapes of the corresponding emissions profiles . for the derivation of possible non - lte effects
, these @xmath109 have to be subtracted from the measured @xmath0 .
the corrections can not directly be applied since both quantities originate from different data sets , i.e the cerro paranal saber sample with 1685 profiles ( sect .
[ sec : saber ] ) and the x - shooter sample with 343 spectra ( sect .
[ sec : xshooter ] ) .
for this reason , we needed to obtain saber - specific parameters such as @xmath108 and @xmath97 for each x - shooter spectrum .
we achieved this by smoothing the saber data in the two dimensions day of year ( doy ) and time .
as a smoothing filter , we used a 2-d gaussian with @xmath95 of half a month ( 15.2 days ) and half an hour .
the gaussians were calculated across the new year boundary if required . for each x - shooter data point
, we could then compute weights for the different saber data points and , finally , weighted averages for the saber - related parameters .
a suitable smoothing procedure is crucial since both data sets have large gaps . in particular ,
the saber data for our selected area around cerro paranal only cover a very narrow range of times for a certain doy ( see sect . [
sec : saber ] ) .
it is desirable to average as many saber data as possible to obtain reliable mean values .
however , if this requires to add data taken at very different conditions , this might cancel out variability patterns that are required for realistic corrections .
the saber data set is already a compromise since it includes data that were taken several 100 away from cerro paranal .
hence , we checked the influence of the smoothing procedure on the resulting weighted parameters .
focusing on the x - shooter sample mean @xmath108 for the eight , two , and one reference emission layers ( see sect . [
sec : profiles ] ) , we found a mean change of about 0.1 for a bisection of the gaussian @xmath95 .
a doubling of the limiting periods resulted in a corresponding mean shift of about 0.2 .
moreover , a different smoothing function ( e.g. an exponential function with a flat top for distances smaller than @xmath95 ) had little influence on the results . finally , the introduction of the solar activity ( measured by the solar radio flux ) as a third dimension did not significantly change the resulting @xmath108 either
therefore , we conclude that the details of the smoothing procedure for the saber data are not critical for our analysis . the main contributions to the weighted parameter averages mainly originate from a relatively small fraction of the saber data . within the @xmath110 perimeter of the 2-d gaussian around the cerro paranal data points , the number of saber profiles ranges from 0 to 41 for our applied procedure .
the mean is 12 .
if a weight sum is used ( defining weight @xmath111 in the centre of the gaussian ) , values between 4 and 49 with a mean of 23 are found .
these numbers suggest that profile - to - profile variations limit the accuracy of the resulting averages .
this is confirmed by typical @xmath108 mean errors between 1.3 ( @xmath3 ) and 1.9 ( @xmath2 ) for individual x - shooter observations .
these uncertainties are comparable to the systematic @xmath0 measurement errors ( see sect . [
sec : lines ] ) .
however , for sample averages , the statistical @xmath108-related errors strongly decrease .
the full sample mean errors are only about 0.1 .
the uncertainties are further reduced if the @xmath109 for the @xmath0 correction are taken into account . in this case , the individual mean errors range from 0.6 ( @xmath3 ) to 1.3 ( @xmath2 ) for the reference profile at 90 described in sect . [
sec : o2aprof ] .
the uncertainties for the sample mean are negligible .
the most critical source of error is probably a significant deviation of the @xmath108 climatologies from those related to the cerro paranal x - shooter data .
this is relevant since the set of saber profiles used covers a relatively wide area comprising 3 in latitude and 20 in longitude ( see sect .
[ sec : saber ] ) . moreover , there are large gaps in the local time coverage for a given month . in order to investigate this , we used the five nighttime periods first introduced in fig .
[ fig : rvero2a_nbina ] . for the ( @xmath1 ) and two bands , we then measured the sample - averaged difference of the x - shooter - related @xmath0 and the saber - related @xmath108 for each time interval .
finally , mean values and standard deviations were calculated for the five data points .
this provides us with a rough estimate of the agreement of the x - shooter and saber nocturnal variability patterns .
the resulting standard deviations range from 0.7 for ( @xmath8 ) to 2.1 for ( @xmath53 ) .
the average for all emission layers is 1.5 .
we can repeat this calculation by taking the temperature differences for two bands , which is needed for the emission profile correction . for the emission of @xmath3 as reference profile
, the standard deviations now range from 0.2k for ( @xmath112 ) to 3.2k for @xmath2 , which reflects the relatively low effective height of the reference layer .
our preferred reference profile at 90 can not be used in this context since @xmath0 measurements do not exist . for this reason
, we calculated the standard deviations for all possible band combinations and performed a regression analysis depending on the related height differences @xmath113 .
then , the climatology - related uncertainties can be estimated only based on differences of emission altitudes .
this approach assumes that @xmath97 is the major driver for the changes in the variability pattern , which is likely since the widths of the emission layers do not differ a lot .
moreover , a linear change of the variability with @xmath113 is required , which is more difficult to fulfil if the height differences are relatively large .
significant variations of the non - lte effects could influence the fit ( see sect . [
sec : results ] ) . for our x - shooter sample ,
we obtain a slope of 0.37 and an offset of 0.47 .
the latter includes influences like fwhm differences .
the fit is robust since the uncertainties of both values are lower by more than an order of magnitude . for a change from a band - specific emission profile to our reference profile , the linear fit results in systematic errors in the temperature correction by the different climatologies for the saber and x - shooter data between 0.7 for ( @xmath8 ) and 2.0 for @xmath2 . for the discussion of seasonal variations in sect .
[ sec : seasvar ] , we also performed the whole analysis by using double month periods , which have a similar length as the timed yaw cycle of about 60 days @xcite .
independent of the month grouping starting with december / january or january / february , we obtained slightly higher errors between 1.1 for ( @xmath8 ) and 2.3 for @xmath2 .
our estimates might be lower than the real systematic uncertainties due to the low bin number , which is limited by the small data sets and gaps in the time coverage ( cf .
[ sec : seasvar ] ) .
the relatively large distances between cerro paranal and the tangent points where the saber profiles were taken are an important reason for the deviating variability . with the approach described above
, we can check the robustness of our area selection limits . focusing on the @xmath34 criterion ( sect .
[ sec : saber ] ) , we also studied 5 and 20 limits . for the correction of the differences between the emission profiles of the measured bands and the reference profile ,
we find average temperature uncertainties of 1.2 , 1.3 , and 1.4 for the 5 , 10 , and 20 limits , respectively .
the averages were derived from the results for the eight and two emission layers .
the errors show the expected increase with growing longitude interval .
however , the changes are small enough that the selection of the limit does not need to be very accurate .
a tight longitude range is not recommended since this would significantly shrink the saber sample and , hence , increase the statistical errors . in from saber data vs. the correlation coefficient @xmath114 for the correlation of the @xmath0 ( filled symbols ) , @xmath108 for the full x - shooter sample ( open symbols ) , and @xmath108 for the cerro paranal saber sample ( crosses ) of ( @xmath1 ) and @xmath3 with those of @xmath2 .
the dash - dotted ( @xmath0 , x - shooter sample ) , long - dashed ( @xmath108 , x - shooter sample ) , and dotted line ( @xmath108 , saber sample ) show fits of the data , which are connected by short - dashed lines.,width=302 ] in the same way as described in sect .
[ sec : tempcorrect ] for @xmath108 , the band - specific @xmath97 ( sect .
[ sec : ohprof ] ) were derived for each x - shooter observation .
it is possible to check the reliability of these data by comparing average @xmath97 with correlation coefficients @xmath114 for the correlation of the corresponding @xmath0 with the @xmath0 for a reference band derived from the same data set ( sect .
[ sec : lines ] ) . assuming a gradual change of the temperature variability pattern with height
, the @xmath0-based @xmath114 should be a function of the altitude difference of the emission layers used for the correlation analysis .
the required altitude - dependent temperature variations can be caused by waves like tides ( e.g. * ? ? ?
. a complication could be @xmath1-dependent variations of the non - lte contributions to the measured @xmath0 ( see sect .
[ sec : results ] ) . as @xmath114 is sensitive to measurement uncertainties , we only selected the most reliable band for each @xmath1 , i.e. we used , , , , , , , and .
the results for the latter band are only based on 240 instead of 343 spectra due to a lack of wavelength coverage and issues with the flux calibration ( see sect . [
sec : xshooter ] ) . figure [ fig : heff_r ] shows the mean @xmath97 vs. @xmath114 for the x - shooter sample with @xmath2 as a reference band ( @xmath115 ) .
there is a striking linear relation for the data , which supports our assumption of equidistant emission layers ( sect .
[ sec : ohprof ] ) .
the slope of a linear fit is @xmath116 .
the deviation of may be explained by the different sample and the measurement uncertainties .
@xmath3 is very close to the fit and best agrees with .
this is another argument for using this band for temperature comparisons with data .
@xmath2 deviates by about 2 from the extrapolated fit , which is most probably caused by a non - linear relation of @xmath97 and @xmath114 for the large emission height differences .
the mesopause temperature minimum , which is close to the @xmath2 emission layer ( see sect . [
sec : results ] ) , might have an impact .
our interpretations are supported by the @xmath114 that we derived from the saber - related @xmath108 for the x - shooter sample ( fig .
[ fig : heff_r ] ) .
there is an almost perfect linear relation of @xmath97 and @xmath114 for these data , which is in good agreement with the found equidistance of the layers and the position of @xmath3 in relation to the bands .
the position of @xmath2 deviates in a similar way from a linear fit of the data with a slope of @xmath117 .
the correlations for @xmath108 are significantly stronger than for @xmath0 .
the relatively low signal - to - noise ratio of the @xmath2 data compared to those of the other bands can have a small impact .
however , a direct correlation for e.g. @xmath112 and 8 revealed @xmath118 for @xmath0 and @xmath119 for @xmath108 .
possible variations of the @xmath0 non - lte contributions do not appear to be sufficiently strong since the @xmath114 difference for @xmath3 is in good agreement with those of ( cf .
[ sec : noctvar ] ) .
therefore , the best explanation for the @xmath114 discrepancies is the weighted averaging of the saber measurements to derive representative values for the x - shooter data set ( sect . [
sec : tempcorrect ] ) .
this effect can be estimated by calculating @xmath114 for the @xmath108 of the original cerro paranal saber sample with 1685 profiles ( sect .
[ sec : saber ] ) . as shown in fig .
[ fig : heff_r ] , the @xmath114 values for these data with a slope of @xmath120 are very close to those for @xmath0 , i.e. the @xmath114 differences are indeed related to the averaging procedure .
the position of @xmath3 deviates from the trend for , which is probably caused by differences in the x - shooter and saber sample properties .
the latter also contributes to the remaining differences between the data points .
moreover , the limited vertical resolution of the saber profiles can cause higher @xmath114 for @xmath108 than for @xmath0 . the sampling of about 0.4 ( sect .
[ sec : saber ] ) is similar to the height difference of adjacent ( @xmath1 ) emission layers ( sect .
[ sec : ohprof ] ) .
therefore , small - scale variations can not be probed by saber .
after the derivation of @xmath0 corrected for the emission layer differences ( sect . [
sec : analysis ] ) , we can finally compare these temperatures and quantify the -related non - lte effects depending on @xmath1 and time .
measurements and their uncertainties ( excluding errors related to the molecular parameters where hitran was used ) for from @xmath53 to 9 ( cf . fig .
[ fig : heff_r ] ) using the first three @xmath44 lines ( circles ) , @xmath3 ( diamond ) , and @xmath2 ( square ) .
the effective emission heights were derived by means of saber data .
the bars in height direction mark the profile width at half maximum .
the cross is related to the saber @xmath32 for the @xmath3-based reference profile .
open symbols indicate the @xmath32 results for the emission profiles of the bands measured by x - shooter . for these data ,
we assume temperature uncertainties of 2.5 .
the solid curve shows the mean @xmath32 profile for the x - shooter sample.,width=302 ] figure [ fig : trot_h ] gives an overview of the mean @xmath0 and their uncertainties for the eight ( @xmath1 ) , @xmath2 , and @xmath3 derived from the full x - shooter sample of 343 spectra ( see sect .
[ sec : lines ] ) .
the @xmath0 are plotted in comparison to the corresponding saber - based @xmath97 ( sect .
[ sec : profiles ] ) projected onto the x - shooter sample ( sect .
[ sec : tempcorrect ] ) .
the fwhm of the emission profiles are indicated by vertical bars .
the average @xmath3 profile is the widest with about 12.7 .
the figure shows a large overlap of the and @xmath3 emission profiles , as already demonstrated by fig .
[ fig : rver_all ] . however , there are large differences in the measured @xmath0 , which range from 188.1 for @xmath53 to 202.5 for @xmath9 . as discussed in @xcite and sect .
[ sec : intro ] , the maximum at @xmath9 is typical of the @xmath0 ( @xmath1 ) pattern and can be explained by the contribution of a nearly nascent population , which shows a particularly hot rotational level population distribution for this @xmath1 .
the mean @xmath0 of @xmath3 is at the lower end of the @xmath0 distribution with @xmath121 .
the @xmath2 @xmath0 of @xmath122 is even lower .
these temperatures can be compared to the saber - based @xmath108 for the x - shooter sample with an estimated uncertainty of about 2.5 ( sect .
[ sec : tempcorrect ] ) .
while the @xmath0 and @xmath108 for each band agree within the errors , this is not the case for all the bands .
the @xmath1 dependence of the temperature is also very different .
the @xmath108 only range from 190.5 to 190.7 , i.e. the temperature gradient is almost zero . in fig .
[ fig : trot_h ] , we also display the mean @xmath32 profile calculated for the x - shooter sample .
indeed , the mean temperature profile is relatively flat in the altitude range of the emission .
together with the large profile overlap , this explains the almost absent temperature gradient .
this implies that the contribution of real temperature variations to the characteristic @xmath0 ( @xmath1 ) pattern is small .
non - lte effects have to dominate .
only the relatively low @xmath0 of @xmath2 can be explained by the proximity of the mesopause temperature minimum . -based reference profile and the full x - shooter sample .
the @xmath1-dependent @xmath0 ( circles ) are based on measurements of the first three @xmath44 lines and hitran - based molecular parameters @xcite .
the systematic errors related to the latter are not included in the plotted error bars .
hitran - based molecular parameters were also used for the derivation of the @xmath0 from @xmath3 ( diamond ) and @xmath2 ( square ) .
the cross shows the -based saber @xmath32 for the reference profile and the full x - shooter sample .
the mean and the uncertainties derived from the three non- temperatures are marked by horizontal lines.,width=302 ] figure [ fig : trot_h ] also shows the @xmath97 , fwhm , and @xmath108 of the @xmath3-based reference profile introduced in sect .
[ sec : o2aprof ] .
@xmath108 is @xmath123 .
the different @xmath0 can be corrected to be representative of this profile that peaks at 90 and has a fwhm of 11 .
this can be done by subtracting the difference between the @xmath108 related to a given band and the reference profile from the @xmath0 for the selected band ( see sect . [
sec : tempcorrect ] ) .
the resulting temperatures are shown in fig .
[ fig : trotref_hit ] . for @xmath3 and @xmath2
, we obtained @xmath124 and @xmath125 , respectively .
the uncertainties are a combination of the measurement uncertainties already shown in fig .
[ fig : trot_h ] and the temperature correction errors discussed in sect .
[ sec : tempcorrect ] .
the two -related temperatures and the -based saber @xmath108 agree very well within their errors .
since these three temperatures are not affected by non - lte effects , we can combine them , which results in an average temperature of @xmath126 .
@xcite published a mean temperature profile based on sodium lidar measurements carried out at so jos dos campos in brazil ( 23s , 46w ) from 2007 to 2010 .
for our reference emission and their temperature profile , this implies a @xmath108 of 188 , which is in remarkable agreement with our measurement despite their assumed absolute errors of about 5 , differences in the time coverage ( no january data ) , and the small but probably non - negligible differences in latitude and longitude . for the latter effect ,
see fig .
[ fig : th_lon ] .
the upper part of the lidar - based temperature profile is also in good agreement with the saber one in fig .
[ fig : trot_h ] .
the lower part shows some discrepancies due to a different height and strength of the secondary minimum , which is at about 88 . in the saber case
, there is only a nearly constant temperature down to about 81 , which causes a lower onset of the steep temperature increase that is already visible in the @xcite profile above 80 .
the comparison with satellite and lidar data shows that the @xmath0 from @xmath3 and @xmath2 line measurements in x - shooter spectra provide reliable temperatures in the mesopause region .
the @xmath0 indicate significant non - lte contributions in comparison to our lte reference temperature of about 188 .
the temperatures range from @xmath127 for @xmath53 to @xmath128 for @xmath7 . in the latter case
, the difference is about 13 . with the estimated errors , the @xmath129 of @xmath9 , 9 , and 6 are highly significant ( @xmath130 ) .
the situation for @xmath53 and 3 is unclear .
the bands and , which by far show the lowest @xmath0 @xcite , are also affected by relatively large systematic uncertainties .
is partly absorbed by water vapour ( more than any other band considered ) and is partly affected by line blending and close to the upper wavelength limit of the x - shooter vis arm .
therefore , it can not be excluded that the true temperatures of @xmath53 and 3 are similar to those of higher @xmath1 .
the plotted temperatures are only valid if the @xmath0 are derived from the first three @xmath44-branch lines .
for example , the resulting @xmath131 and the significance of the non - lte effects would increase if @xmath44 ( 4 ) was also used . in this case
, the @xmath0 would be higher by about 1 for all @xmath1 .
the change to the first four @xmath45-branch lines would even increase the @xmath0 by about 11 .
moreover , the discussed @xmath0 are based on einstein coefficients and level energies from the hitran database @xcite , which are related to @xcite .
the results change if we use the molecular parameters of @xcite ( fig .
[ fig : trotref_vdl ] ) .
the @xmath0 are more uncertain and tend to be lower than those based on hitran data . if we neglect the very uncertain data point for @xmath53 ( see * ? ? ?
* ) , the mean shift is 4.2 .
hence , only the @xmath129 of 10 for @xmath9 remains significant with more than @xmath132 .
although the @xmath0 based on @xcite are less reliable than those based on @xcite , these results demonstrate that the uncertainties of the molecular parameters limit the accuracy of @xmath129 that can currently be achieved .
nevertheless , @xmath129 of more than 10 for @xmath9 are plausible if the first three @xmath44-branch lines are used .
finally , the results show that the consideration of emission profile differences does not significantly change the characteristic @xmath0 ( @xmath1 ) pattern found by @xcite and @xcite .
it can only be explained by non - lte effects .
-dependent non - lte contributions to @xmath0 based on the first three @xmath44 lines and hitran data @xcite for five nighttime lt periods derived from the full x - shooter sample .
the lte base temperature for each @xmath1 was derived from the corresponding @xmath3 @xmath0 corrected for the difference in the ver profiles .
the uncertainties of the absolute @xmath129 are similar to those given in fig .
[ fig : trotref_hit ] ( excluding the unknown uncertainties of the molecular parameters ) . comparisons of @xmath133 for the different night - time periods and/or @xmath1 are more reliable .
the errors should be of the order of 1 , i.e. very likely smaller than 2.,width=302 ] except for ( @xmath53 ) as base for the calculation of @xmath133 .
the errors for comparisons of different nighttime periods and/or @xmath1 are probably smaller than those of fig .
[ fig : deltrot_vp_nbina ] since temperature correction uncertainties are less critical .
the absolute @xmath133 are less reliable.,width=302 ] @xcite studied the nocturnal variations of the @xmath0 by means of five nighttime periods , which we also used in figs .
[ fig : rvero2a_nbina ] and [ fig : heff_nbina ] .
the analysis resulted in hitran - based @xmath0 differences ( i.e. maximum minus minimum ) between 4.8 for @xmath134 and 7.0 for @xmath135 . in order to separate non - lte from temperature profile variations
, we used the observed @xmath3 profiles as reference for an emission profile correction .
we did not take the reference profile at 90 , which was the basis for the discussion in sect .
[ sec : tempcomp ] , since this reduces the uncertainties related to the temperature correction from about 1.2 to 0.9 on average .
at the beginning / end of the night , the uncertainties are higher for high / low @xmath1 .
the resulting @xmath131 are displayed in fig .
[ fig : deltrot_vp_nbina ] . compared to the uncorrected @xmath0 ,
the time - related variation for fixed @xmath1 significantly decreased .
the @xmath129 differences range from 0.6 for @xmath112 to 3.6 for @xmath8 .
hence , the average nocturnal variations of the non - lte contributions to the @xmath0 are significantly lower than their actual values . moreover , at least the low @xmath1 indicate that the @xmath0 variation is obviously dominated by the variability of the real temperature profile .
however , it seems that only one third of the @xmath0 ( @xmath8 ) variability can be explained in this way .
figure [ fig : deltrot_vp_nbina ] indicates a nocturnal trend of increasing @xmath129 .
the trend appears to be stronger for higher @xmath1 .
the significance is about @xmath132 for the differences between the first and last interval and is limited by the statistical mean errors with an average of about 1.1 and at least part of the temperature correction errors due to the changing @xmath3 emission profile .
@xcite discovered a significant increase of the @xmath0 differences of adjacent intermediate @xmath1 for the second compared to the first nighttime period . in fig .
[ fig : deltrot_vp_nbina ] , this causes a relatively strong @xmath129 increase for @xmath135 but a decrease for @xmath136 and 7 . in order to better evaluate the reliability of these trends
, we also calculated @xmath131 based on ( @xmath53 ) instead of @xmath3 emission profiles .
although the absolute @xmath129 are less trustworthy and possible weak non - lte @xmath0 variations for @xmath53 are neglected , this approach is justified for the comparisons of nighttime intervals and @xmath1 since it can further reduce the uncertainties related to the temperature correction .
the emission profile differences between ( @xmath53 ) and the other bands are much smaller and less variable than those related to @xmath3 . the resulting fig .
[ fig : deltrot_vp_nbina_bv2 ] does not show a @xmath131 decrease between the first and second nighttime period . however , significant increases appear to be limited to @xmath135 , 8 , and 9 . for @xmath136 and 7
, the major change happens between the third and fourth period , i.e. after midnight .
the last period indicates a decrease of @xmath131 for most @xmath1 . for @xmath137 ,
the @xmath129 increases between the first and fourth period vary between 3.6 ( @xmath138 ) and 4.3 ( @xmath9 ) .
the significance is higher than for the results based on @xmath3 as reference band .
the found @xmath129 variations appear to be in agreement with those of the emission layer height ( see sect . [
sec : intro ] ) . up to the fourth nighttime period , the sample - averaged @xmath97 increases by 0.9 ( @xmath8 ) to 1.7 ( @xmath53 ) .
afterwards the emission layers sink again except for @xmath8 .
see fig .
[ fig : heff_nbina ] for the @xmath97 trend related to @xmath9 . as already discussed by @xcite based on @xmath0 differences and vibrational level populations , a higher emission layer is in a lower density environment , which decreases the frequency of thermalising collisions by ( e.g. * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ) and , consequently , can increase the @xmath0 non - lte effects .
the saber data show that the @xmath97 increase is more related to a ver decrease at low altitudes than an increase in the upper part of the emission layer .
this trend mainly depends on changes in the distribution of atomic oxygen ( which is required for the production ) by chemical reactions and atmospheric dynamics @xcite . in order to better estimate the effect of the ver profile change on the @xmath0 non - lte effects , we calculated the fraction of emission above a certain altitude as a function of the nighttime period . for this
, we used the original ver profiles of the two saber channels at 1.64 and 2.06 corresponding to @xmath139 and 8.3 , respectively ( sect .
[ sec : ohprof ] ) . for a minimum altitude of 95
, we found an increase from 4.7% ( first period ) to 7.3% ( fourth period ) for @xmath139 and from 9.4% to 12.0% for @xmath140 .
the differences are similar for both saber channels but the absolute values are higher for @xmath141 . the latter effect and the relative change of the emission fractions with time enhance with increasing minimum altitude .
these results show the large impact of a relatively small rise of the emission layer on the highest parts of the ver profile . to fully understand fig .
[ fig : deltrot_vp_nbina_bv2 ] , it will be necessary to model @xmath0 depending on @xmath1 and altitude , which has not been done so far . while there is some knowledge on the nascent population distribution over the rotational levels of the different @xmath142 @xcite
, it is not clear how this distribution is modified by collisional quenching . except for showing results for the first ( upper panel ) and the second half of the night ( lower panel ) separately.,title="fig:",width=302 ] except for showing results for the first ( upper panel ) and the second half of the night ( lower panel ) separately.,title="fig:",width=302 ] figure [ fig : trot_h_nh ] illustrates the effect of the changing emission layers and temperature profile on the measured @xmath0 by comparing @xmath0 , @xmath108 , and @xmath32 for the first and second half of the night .
the plots are similar to fig .
[ fig : trot_h ] discussed in sect .
[ sec : tempcomp ] .
the @xmath0 show a wider spread in the second half of the night .
this can partly be explained by a significant change of the temperature profile .
the saber data reveal an average nocturnal trend from a single deep mesopause temperature minimum at the beginning to a double minimum , where the secondary trough is almost as deep as the primary one .
the average altitudes of both minima are 98 to 99 and between 80 and 85 .
observations of a secondary minimum are not unusual @xcite and can be explained by tidal perturbations and tide gravity wave interactions @xcite .
the appearance of a secondary mesopause minimum causes a temperature inversion at the main emission altitudes .
the comparison of the second with the first half of the night , therefore , shows a @xmath32 increase above and decrease below 85 .
the @xmath108 are higher in the second period and this increase is stronger for higher @xmath1 due to the different temperature gradients .
the amounts are 1.9 for @xmath53 and 3.5 for @xmath54 . nevertheless , a @xmath0 increase with @xmath1 remains , which is disturbed by the odd / even @xmath1 dichotomy , as indicated by figs .
[ fig : deltrot_vp_nbina ] and [ fig : deltrot_vp_nbina_bv2 ] .
the -related @xmath0 follow the changes of the temperature profile convincingly .
as expected , the emission height , profile width , and temperature for @xmath3 are much more similar to those for the reference profile ( sect .
[ sec : o2aprof ] ) in the second half of the night . except for curves related to the four meteorological seasons instead of nighttime periods .
the uncertainties for comparisons of @xmath133 for the different seasons and/or @xmath1 are somewhat higher .
2 and more can not be excluded.,width=302 ] @xcite also studied the @xmath0 ( @xmath1 ) of the four meteorological seasons .
the x - shooter data indicated higher temperatures for the equinoxes than for the solstices irrespective of @xmath1 .
the maximum differences were between 7.4 for @xmath53 and 3.3 for @xmath9 .
after applying a temperature correction based on @xmath3 , these differences decreased to 4.7 and 1.2 , as shown in fig .
[ fig : deltrot_vp_seasa ] .
these values imply stronger non - lte effects near the equinoxes than near the solstices .
moreover , the effects seem to be stronger for lower @xmath1 .
this is contradictory to the results and explanations discussed in sect .
[ sec : noctvar ] .
the apparent trend is most probably caused by an inefficient temperature correction .
although the reduction of the seasonal variations by the conversion from @xmath0 to @xmath129 was even the largest for low @xmath1 ( @xmath143 vs. @xmath82 , see above ) , it was obviously insufficient to obtain reliable seasonal variations of the @xmath0 non - lte contributions .
this is probably related to the timed yaw cycle of 60 days , which causes large gaps in the seasonal coverage for fixed local times ( sects .
[ sec : intro ] and [ sec : saber ] ) .
therefore , the true seasonal variations could be stronger than those used for the temperature correction . at cerro paranal and similar latitudes , studying seasonal variations
is more challenging than an analysis of the nocturnal variations since the combination of semi - annual and the weaker annual oscillations for @xmath0 @xcite is more complex than the night trends @xcite , which are dominated by the solar migrating diurnal tide @xcite .
hence , the limitations of our saber and x - shooter data sets do not allow us to conclude on seasonal variations of the @xmath0 non - lte effects .
based on the vlt / x - shooter data set of @xcite with 343 spectra , we could compare their measurements of rotational temperatures @xmath0 from 25 bands covering @xmath53 to 9 with new results for the bands @xmath2 and @xmath3 .
for the latter band , it was just the second time that @xmath0 could be derived and the first time that it could be done with spectroscopically resolved lines . since @xmath3 emits at similar altitudes as and the rotational level populations appear to be consistent with a boltzmann distribution , the accurate @xmath0 for @xmath3 could be used to quantify the non - lte contributions to the @xmath0 . to correct for the different emission profiles and the corresponding ver - weighted @xmath108 , we used , @xmath3 , and -based @xmath32 profile data from timed / saber suitable for cerro paranal .
the correction is particularly important for the beginning of the night when the @xmath3 emission peaks several below the nighttime average due to a decaying daytime population from ozone photolysis . from the x - shooter data
set , we obtained an effective population lifetime of about 46 , which is in good agreement with saber and other data .
the required emission profiles for the different @xmath1 were derived from the two saber channels under the assumption that emission peaks are equidistant for adjacent @xmath1 .
this approach is supported by a comparison of the effective emission heights @xmath97 with temperature correlation coefficients for the x - shooter and saber samples , which showed nearly linear relations . for an @xmath3-based reference profile at 90 , we compared the corrected temperature measurements and found that the results for @xmath2 ( where a fixed emission profile at 94.5 was assumed ) , @xmath3 , and the saber @xmath32 agree very well within the errors .
the sample mean temperature at the reference altitude was @xmath126 .
except for the uncertain @xmath53 and 3 , the resulting -related temperatures were significantly higher .
for the first three @xmath44-branch lines and hitran molecular data , a maximum difference of about 13 was achieved for @xmath9 . the characteristic @xmath0 ( @xmath1 ) pattern found by @xcite and @xcite showing an odd / even @xmath1 dichotomy
could also be seen in the corrected temperature data .
the amount of non - lte contributions to the @xmath0 @xmath129 strongly depends on the selected lines .
a change of the line set affects all bands in a similar way .
apart from measurement and correction errors , the still uncertain molecular parameters limit the accuracy of the derived @xmath129 to a few .
an analysis of the nocturnal variations of @xmath129 with five nighttime periods revealed that a significant fraction of the @xmath0 variability is caused by changes in the emission and mesopause kinetic temperature profiles .
however , fractions above 50% could only be found for low @xmath1 . at cerro paranal
, the residual non - lte effects tend to increase during the night and to be more variable at higher @xmath1 .
the largest @xmath129 differences were found for the first ( before 21:00lt ) and fourth period ( between 01:00 and 03:00lt ) , and @xmath137 .
they amount to about 4 with an uncertainty of about 2 .
the amplitude of the variations appears to be only a minor fraction of the total amount of the non - lte effects .
the observed trends could be explained by the nocturnal rise of the emission layer , which we see in the saber data up to the fourth nighttime period , and a corresponding reduction of the thermalising collisions due to the lower effective density .
higher @xmath1 could be more affected since the related emission peaks at higher altitudes .
we also studied possible seasonal variations of @xmath129 .
however , the limitations of the saber and x - shooter data sets did not allow us to find convincing trends . in agreement with @xcite ,
non - lte effects are critical for absolute mesopause temperature estimates from @xmath0 and comparisons of bands with different @xmath1 .
dynamical studies based on @xmath0 derived from bands with high @xmath1 could also be significantly affected . in this respect ,
only low @xmath1 appear to be sufficiently safe .
this project made use of the eso science archive facility .
x - shooter spectra from different observing programmes of the period from october 2009 to march 2013 were analysed .
moreover , we thank the saber team for providing the data products used in this paper .
we thank the two anonymous referees for their detailed and very helpful comments .
this publication is supported by the austrian science fund ( fwf ) .
s. noll and s. unterguggenberger receive funding from the fwf project p26130 .
w. kausch is funded by the project is538003 ( hochschulraumstrukturmittel ) provided by the austrian ministry for research ( bmwfw ) .
adler - golden , s. : kinetic parameters for nightglow modeling consistent with recent laboratory measurements , j. geophys .
res . , 102 , 1996919976 , doi : http://dx.doi.org/10.1029/97ja01622[10.1029/97ja01622 ] , 1997 . baker , d. j. and stair jr . , a. t. : rocket measurements of the altitude distributions of the hydroxyl airglow , phys .
scripta , 37 , 611622 , doi : http://dx.doi.org/10.1088/0031 - 8949/37/4/021[10.1088/0031 - 8949/37/4/021 ] , 1988 .
baker , d. j. , thurgood , b. k. , harrison , w. k. , mlynczak , m. g. , and russell , j. m. : equatorial enhancement of the nighttime mesospheric infrared airglow , phys .
scripta , 75 , 615619 , doi : http://dx.doi.org/10.1088/0031 - 8949/75/5/004[10.1088/0031 - 8949/75/5/004 ] , 2007 .
beig , g. , keckhut , p. , lowe , r. p. , roble , r. g. , mlynczak , m. g. , scheer , j. , fomichev , v. i. , offermann , d. , french , w. j. r. , shepherd , m. g. , semenov , a. i. , remsberg , e. e. , she , c. y. , lbken , f. j. , bremer , j. , clemesha , b. r. , stegman , j. , sigernes , f. , and fadnavis , s. : review of mesospheric temperature trends , rev .
geophys . , 41 ,
rg1015 , doi : http://dx.doi.org/10.1029/2002rg000121[10.1029/2002rg000121 ] , 2003 .
beig , g. , scheer , j. , mlynczak , m. g. , and keckhut , p. : overview of the temperature response in the mesosphere and lower thermosphere to solar activity , rev .
geophys . , 46 , rg3002 , doi : http://dx.doi.org/10.1029/2007rg000236[10.1029/2007rg000236 ] , 2008 .
charters , p. e. , macdonald , r. g. , and polanyi , j. c. : formation of vibrationally excited by the reaction @xmath144 , appl
. optics , 10 , 17471754 , doi : http://dx.doi.org/10.1364/ao.10.001747[10.1364/ao.10.001747 ] , 1971 .
clemesha , b. , simonich , d. , and batista , p. : sodium lidar measurements of mesopause region temperatures at 23s , adv .
space res . , 47 , 11651171 , doi : http://dx.doi.org/10.1016/j.asr.2010.11.030[10.1016/j.asr.2010.11.030 ] , 2011 . clough , s. a. , shephard , m. w. , mlawer , e. j. , delamere , j. s. , iacono , m. j. , cady - pereira , k. , boukabara , s. , and brown , p. d. : atmospheric radiative transfer modeling : a summary of the aer codes , j. quant .
ra . , 91 , 233244 , doi : http://dx.doi.org/10.1016/j.jqsrt.2004.05.058[10.1016/j.jqsrt.2004.05.058 ] , 2005 .
evans , w. f. j. , hunten , d. m. , llewellyn , e. j. , and vallance jones , a. : altitude profile of the infrared atmospheric system of oxygen in the dayglow , j. geophys .
res . , 73 , 28852896 , doi : http://dx.doi.org/10.1029/ja073i009p02885[10.1029/ja073i009p02885 ] , 1968 . friedman , j. s. and chu , x. : nocturnal temperature structure in the mesopause region over the arecibo observatory ( 18.35n , 66.75w ) : seasonal variations , j. geophys .
res . , 112 ,
d14107 , doi : http://dx.doi.org/10.1029/2006jd008220[10.1029/2006jd008220 ] , 2007 .
gamache , r. r. and goldman , a. : einstein a coefficient , integrated band intensity , and population factors : application to the @xmath145@xmath19 band , j. quant .
ra . , 69 , 389401 , doi : http://dx.doi.org/10.1016/s0022 - 4073(00)00072 - 8[10.1016/s0022 - 4073(00)00072 - 8 ] , 2001 .
gamache , r. r. , goldman , a. , and rothman , l. s. : improved spectral parameters for the three most abundant isotopomers of the oxygen molecule , j. quant .
ra . , 59 , 495509 , doi : http://dx.doi.org/10.1016/s0022 - 4073(97)00124 - 6[10.1016/s0022 - 4073(97)00124 - 6 ] , 1998 .
gao , h. , xu , j. y. , chen , g. m. , yuan , w. , and beletsky , a. b. : global distributions of and ( 1.27 ) nightglow emissions observed by timed satellite , sci . china ser .
e , 54 , 447456 , doi : http://dx.doi.org/10.1007/s11431 - 010 - 4236 - 5[10.1007/s11431 - 010 - 4236 - 5 ] , 2011 .
gelinas , l. j. , hecht , j. h. , walterscheid , r. l. , roble , r. g. , and woithe , j. m. : a seasonal study of mesospheric temperatures and emission intensities at adelaide and alice springs , j. geophys .
113 , a01304 , doi : http://dx.doi.org/10.1029/2007ja012587[10.1029/2007ja012587 ] , 2008 . goldman , a. , schoenfeld , w. g. , goorvitch , d. , chackerian jr . ,
c. , dothe , h. , mlen , f. , abrams , m. c. , and selby , j. e. a. : updated line parameters for @xmath146ii-@xmath146ii ( @xmath1 , @xmath25 ) transitions , j. quant .
ra . , 59 , 453469 , doi : http://dx.doi.org/10.1016/s0022 - 4073(97)00112-x[10.1016/s0022 - 4073(97)00112-x ] , 1998 .
greer , r. g. h. , llewellyn , e. j. , solheim , b. h. , and witt , g. : the excitation of ( @xmath147 ) in the nightglow , planet .
space sci .
, 29 , 383389 , doi : http://dx.doi.org/10.1016/0032 - 0633(81)90081 - 7[10.1016/0032 - 0633(81)90081 - 7 ] , 1981 .
greer , r. g. h. , murtagh , d. p. , mcdade , i. c. , dickinson , p. h. g. , and thomas , l. : eton 1 a data base pertinent to the study of energy transfer in the oxygen nightglow , planet .
space sci .
, 34 , 771788 , doi : http://dx.doi.org/10.1016/0032 - 0633(86)90074 - 7[10.1016/0032 - 0633(86)90074 - 7 ] , 1986 .
hays , p. b. , kafkalidis , j. f. , skinner , w. r. , and roble , r. g. : a global view of the molecular oxygen night airglow , j. geophys .
res . , 108 , 4646 , doi : http://dx.doi.org/10.1029/2003jd003400[10.1029/2003jd003400 ] , 2003 . howell , c. d. , michelangeli , d. v. , allen , m. , yung , y. l. , and thomas , r. j. : sme observations of ( @xmath145 ) nightglow an assessment of the chemical production mechanisms , planet .
space sci .
, 38 , 529537 , doi : http://dx.doi.org/10.1016/0032 - 0633(90)90145-g[10.1016/0032 - 0633(90)90145-g ] , 1990 .
kaufmann , m. , gusev , o. a. , grossmann , k. u. , martn - torres , f. j. , marsh , d. r. , and kutepov , a. a. : satellite observations of daytime and nighttime ozone in the mesosphere and lower thermosphere , j. geophys .
res . , 108 , 4272 , doi : http://dx.doi.org/10.1029/2002jd002800[10.1029/2002jd002800 ] , 2003 . kausch , w. , noll , s. , smette , a. , kimeswenger , s. , barden , m. , szyszka , c. , jones , a. m. , sana , h. , horst , h. , and kerber , f. : molecfit : a general tool for telluric absorption correction , ii
. quantitative evaluation on eso - vlt / x - shooter spectra , astron .
astrophys . , 576 ,
a78 , doi : http://dx.doi.org/10.1051/0004 - 6361/201423909[10.1051/0004 - 6361/201423909 ] , 2015 .
lafferty , w. j. , solodov , a. m. , lugez , c. l. , and fraser , g. t. : rotational line strengths and self - pressure - broadening coefficients for the 1.27- , @xmath148 , @xmath149 band of , appl .
optics , 37 , 22642270 , doi : http://dx.doi.org/10.1364/ao.37.002264[10.1364/ao.37.002264 ] , 1998 .
leshchishina , o. , kassi , s. , gordon , i. e. , rothman , l. s. , wang , l. , and campargue , a. : high sensitivity crds of the @xmath150-@xmath151 band of oxygen near 1.27 : extended observations , quadrupole transitions , hot bands and minor isotopologues , j. quant .
ra . , 111 , 22362245 , doi : http://dx.doi.org/10.1016/j.jqsrt.2010.05.014[10.1016/j.jqsrt.2010.05.014 ] , 2010 . lipatov , k. v. : empirical model of variations in the ir atmospheric system of molecular oxygen : 2 . emitting layer height , geomagn .
aeron . , 53 , 104112 , doi : http://dx.doi.org/10.1134/s001679321301012x[10.1134/s001679321301012x ] , 2013 .
lpez - gonzlez , m. j. , lpez - moreno , j. j. , lpez - valverde , m. a. , and rodrigo , r. : behaviour of the infrared atmospheric band in the middle atmosphere during evening twilight and at night , planet .
space sci .
, 37 , 6172 , doi : http://dx.doi.org/10.1016/0032 - 0633(89)90069-x[10.1016/0032 - 0633(89)90069-x ] , 1989 .
lpez - gonzlez , m. j. , lpez - moreno , j. j. , and rodrigo , r. : the altitude profile of the infrared atmospheric system of in twilight and early night derivation of ozone abundances , planet .
space sci .
, 40 , 13911397 , doi : http://dx.doi.org/10.1016/0032 - 0633(92)90094 - 5[10.1016/0032 - 0633(92)90094 - 5 ] , 1992 .
lowe , r. p. , leblanc , l. m. , and gilbert , k. l. : windii / uars observation of twilight behaviour of the hydroxyl airglow , at mid - latitude equinox , j. atmos .
phys . , 58 , 18631869 , doi : http://dx.doi.org/10.1016/0021 - 9169(95)00178 - 6[10.1016/0021 - 9169(95)00178 - 6 ] , 1996 .
marsh , d. r. , smith , a. k. , mlynczak , m. g. , and russell , j. m. : saber observations of the meinel airglow variability near the mesopause , j. geophys .
res . , 111 ,
a10s05 , doi : http://dx.doi.org/10.1029/2005ja011451[10.1029/2005ja011451 ] , 2006 .
mcdade , i. c. , llewellyn , e. j. , greer , r. g. h. , and murtagh , d. p. : eton 6 a rocket measurement of the o2 infrared atmospheric ( 0 - 0 ) band in the nightglow , planet .
space sci .
, 35 , 15411552 , doi : http://dx.doi.org/10.1016/0032 - 0633(87)90079 - 1[10.1016/0032 - 0633(87)90079 - 1 ] , 1987 .
melo , s. m. l. , lowe , r. p. , and takahashi , h. : the nocturnal behavior of the hydroxyl airglow at the equatorial and low latitudes as observed by windii : comparison with ground - based measurements , j. geophys .
res . , 104 , 2465724666 , doi : http://dx.doi.org/10.1029/1999ja900291[10.1029/1999ja900291 ] , 1999 . meriwether , j. w. and gerrard , a. j. : mesosphere inversion layers and stratosphere temperature enhancements , rev .
geophys . , 42 ,
rg3003 , doi : http://dx.doi.org/10.1029/2003rg000133[10.1029/2003rg000133 ] , 2004 .
mertens , c. j. , mlynczak , m. g. , lpez - puertas , m. , wintersteiner , p. p. , picard , r. h. , winick , j. r. , gordley , l. l. , and russell iii , j. m. : retrieval of mesospheric and lower thermospheric kinetic temperature from measurements of 15 earth limb emission under non - lte conditions , geophys .
lett . , 28 , 13911394 , doi : http://dx.doi.org/10.1029/2000gl012189[10.1029/2000gl012189 ] , 2001
. mertens , c. j. , schmidlin , f. j. , goldberg , r. a. , remsberg , e. e. , pesnell , w. d. , russell iii , j. m. , mlynczak , m. g. , lpez - puertas , m. , wintersteiner , p. p. , picard , r. h. , winick , j. r. , and gordley , l. l. : saber observations of mesospheric temperatures and comparisons with falling sphere measurements taken during the 2002 summer macwave campaign , geophys .
lett . , 31 ,
l03105 , doi : http://dx.doi.org/10.1029/2003gl018605[10.1029/2003gl018605 ] , 2004 .
mlynczak , m. g. , martin - torres , f. j. , crowley , g. , kratz , d. p. , funke , b. , lu , g. , lopez - puertas , m. , russell iii , j. m. , kozyra , j. , mertens , c. , sharma , r. , gordley , l. , picard , r. , winick , j. , and paxton , l. : energy transport in the thermosphere during the solar storms of april 2002 , j. geophys .
res . , 110 ,
a12s25 , doi : http://dx.doi.org/10.1029/2005ja011141[10.1029/2005ja011141 ] , 2005 .
mlynczak , m. g. , marshall , b. t. , martin - torres , f. j. , russell iii , j. m. , thompson , r. e. , remsberg , e. e. , and gordley , l. l. : sounding of the atmosphere using broadband emission radiometry observations of daytime mesospheric ( @xmath153 ) 1.27 emission and derivation of ozone , atomic oxygen , and solar and chemical energy deposition rates , j. geophys .
res . , 112 ,
d15306 , doi : http://dx.doi.org/10.1029/2006jd008355[10.1029/2006jd008355 ] , 2007 .
modigliani , a. , goldoni , p. , royer , f. , haigron , r. , guglielmi , l. , franois , p. , horrobin , m. , bristow , p. , vernet , j. , moehler , s. , kerber , f. , ballester , p. , mason , e. , and christensen , l. : the x - shooter pipeline , spie proc .
ser . , 7737 , 773728 , doi : http://dx.doi.org/10.1117/12.857211[10.1117/12.857211 ] , 2010 .
moehler , s. , modigliani , a. , freudling , w. , giammichele , n. , gianninas , a. , gonneau , a. , kausch , w. , lanon , a. , noll , s. , rauch , t. , and vinther , j. : flux calibration of medium - resolution spectra from 300 nm to 2500 nm : model reference spectra and telluric correction , astron .
astrophys . , 568 ,
a9 , doi : http://dx.doi.org/10.1051/0004 - 6361/201423790[10.1051/0004 - 6361/201423790 ] , 2014 .
mulligan , f. j. and galligan , j. m. : mesopause temperatures calculated from the o@xmath154(@xmath150 ) twilight airglow emission recorded at maynooth ( 53.2n , 6.4w ) , ann .
geophys . , 13 , 558566 , doi : http://dx.doi.org/10.1007/s00585 - 995 - 0558 - 1[10.1007/s00585 - 995 - 0558 - 1 ] , 1995 . nair , h. and yee , j .- h . :
( @xmath150 , @xmath59 ) chemical loss coefficients determined from saber sunset measurements , geophys .
lett . , 36 ,
l15829 , doi : http://dx.doi.org/10.1029/2009gl039335[10.1029/2009gl039335 ] , 2009 .
newman , s. m. , lane , i. c. , orr - ewing , a. j. , newnham , d. a. , and ballard , j. : integrated absorption intensity and einstein coefficients for the @xmath155 transition : a comparison of cavity ringdown and high resolution fourier transform spectroscopy with a long - path absorption cell , j. chem .
phys . , 110 , 1074910757 , doi : http://dx.doi.org/10.1063/1.479018[10.1063/1.479018 ] , 1999 . nikoukar , r. , swenson , g. r. , liu , a. z. , and kamalabadi , f. : on the variability of mesospheric oh emission profiles , j. geophys .
res . , 112 ,
d19109 , doi : http://dx.doi.org/10.1029/2007jd008601[10.1029/2007jd008601 ] , 2007 .
noll , s. , kausch , w. , barden , m. , jones , a. m. , szyszka , c. , kimeswenger , s. , and vinther , j. : an atmospheric radiation model for cerro paranal .
i. the optical spectral range , astron .
astrophys . , 543 ,
a92 , doi : http://dx.doi.org/10.1051/0004 - 6361/201219040[10.1051/0004 - 6361/201219040 ] , 2012 .
noll , s. , kausch , w. , kimeswenger , s. , unterguggenberger , s. , and jones , a. m. : oh populations and temperatures from simultaneous spectroscopic observations of 25 bands , atmos .
phys . , 15 , 36473669 , doi : http://dx.doi.org/10.5194/acp-15 - 3647 - 2015[10.5194/acp-15 - 3647 - 2015 ] , 2015 .
ohoyama , h. , kasai , t. , yoshimura , y. , kimura , h. , and kuwata , k. : initial distribution of vibration of the radicals produced in the @xmath156@xmath157(x@xmath158@xmath159@xmath160)@xmath156 reaction .
chemiluminescence by a crossed beam technique , chem .
phys . , 118 , 263266 , doi : http://dx.doi.org/10.1016/0009 - 2614(85)85312 - 4[10.1016/0009 - 2614(85)85312 - 4 ] , 1985 . patat , f. , moehler , s. , obrien , k. , pompei , e. , bensby , t. , carraro , g. , de ugarte postigo , a. , fox , a. , gavignaud , i. , james , g. , korhonen , h. , ledoux , c. , randall , s. , sana , h. , smoker , j. , stefl , s. , and szeifert , t. : optical atmospheric extinction over cerro paranal , astron .
astrophys . , 527 ,
a91 , doi : http://dx.doi.org/10.1051/0004 - 6361/201015537[10.1051/0004 - 6361/201015537 ] , 2011 .
pendleton jr . , w. , espy , p. , baker , d. , steed , a. , and fetrow , m. : observation of meinel ( 7 , 4 ) @xmath36 ( @xmath161 ) transitions in the night airglow , j. geophys .
res . , 94 , 505510 , doi : http://dx.doi.org/10.1029/ja094ia01p00505[10.1029/ja094ia01p00505 ] , 1989 .
pendleton jr . , w. r. , espy , p. j. , and hammond , m. r. : evidence for non - local - thermodynamic - equilibrium rotation in the nightglow , j. geophys .
res . , 98 , 1156711579 , doi : http://dx.doi.org/10.1029/93ja00740[10.1029/93ja00740 ] , 1993 . pendleton jr . , w. r. , baker , d. j. , reese , r. j. , and oneil , r. r. : decay of ( @xmath162 ) in the evening twilight airglow : implications for the radiative lifetime , geophys .
lett . , 23 , 10131016 , doi : http://dx.doi.org/10.1029/96gl00946[10.1029/96gl00946 ] , 1996 . perminov , v. i. and lipatov , k. v. : empirical model of variations in the emission of the infrared atmospheric system of molecular oxygen : 3
. temperature , geomagn .
aeron . , 53 , 389396 , doi : http://dx.doi.org/10.1134/s0016793213030158[10.1134/s0016793213030158 ] , 2013 .
perminov , v. i. , semenov , a. i. , and shefov , n. n. : on rotational temperature of the hydroxyl emission , geomagn .
aeron . , 47 , 756763 , doi : http://dx.doi.org/10.1134/s0016793207060084[10.1134/s0016793207060084 ] , 2007 .
reisin , e. r. , scheer , j. , dyrland , m. e. , sigernes , f. , deehr , c. s. , schmidt , c. , hppner , k. , bittner , m. , ammosov , p. p. , gavrilyeva , g. a. , stegman , j. , perminov , v. i. , semenov , a. i. , knieling , p. , koppmann , r. , shiokawa , k. , lowe , r. p. , lpez - gonzlez , m. j. , rodrguez , e. , zhao , y. , taylor , m. j. , buriti , r. a. , espy , p. j. , french , w. j. r. , eichmann , k .- u . ,
burrows , j. p. , and von savigny , c. : traveling planetary wave activity from mesopause region airglow temperatures determined by the network for the detection of mesospheric change ( ndmc ) , j. atmos .
sol .- terr .
phy . , 119 , 7182 , doi : http://dx.doi.org/10.1016/j.jastp.2014.07.002[10.1016/j.jastp.2014.07.002 ] , 2014
. remsberg , e. e. , marshall , b. t. , garcia - comas , m. , krueger , d. , lingenfelser , g. s. , martin - torres , j. , mlynczak , m. g. , russell iii , j. m. , smith , a. k. , zhao , y. , brown , c. , gordley , l. l. , lopez - gonzalez , m. j. , lopez - puertas , m. , she , c .- y . ,
taylor , m. j. , and thompson , r. e. : assessment of the quality of the version 1.07 temperature - versus - pressure profiles of the middle atmosphere from timed / saber , j. geophys .
res . , 113 ,
d17101 , doi : http://dx.doi.org/10.1029/2008jd010013[10.1029/2008jd010013 ] , 2008 .
rezac , l. , kutepov , a. , russell iii , j. m. , feofilov , a. g. , yue , j. , and goldberg , r. a. : simultaneous retrieval of @xmath163 and vmr from two - channel non - lte limb radiances and application to daytime saber / timed measurements , j. atmos .
sol .- terr .
phy . , 130 , 2342 , doi : http://dx.doi.org/10.1016/j.jastp.2015.05.004[10.1016/j.jastp.2015.05.004 ] , 2015 . rothman , l. s. , gordon , i. e. , barbe , a. , chris benner , d. , bernath , p. f. , birk , m. , boudon , v. , brown , l. r. , campargue , a. , champion , j .- p . , chance , k. , coudert , l. h. , dana , v. , devi , v. m. , fally , s. , flaud , j .- m . , gamache , r. r. , goldman , a. , jacquemart , d. , kleiner , i. , lacome , n. , lafferty , w. j. , mandin , j .- y . , massie , s. t. , mikhailenko , s. n. , miller , c. e. , moazzen - ahmadi , n. , naumenko , o. v. , nikitin , a. v. , orphal , j. , perevalov , v. i. , perrin , a. , predoi - cross , a. , rinsland , c. p. , rotger , m. ,
imekov , m. , smith , m. a. h. , sung , k. , tashkun , s. a. , tennyson , j. , toth , r. a. , vandaele , a. c. , and vander auwera , j. : the hitran 2008 molecular spectroscopic database , j. quant .
ra . , 110 , 533572 , doi : http://dx.doi.org/10.1016/j.jqsrt.2009.02.013[10.1016/j.jqsrt.2009.02.013 ] , 2009 . rothman , l. s. , gordon , i. e. , babikov , y. , barbe , a. , chris benner , d. , bernath , p. f. , birk , m. , bizzocchi , l. , boudon , v. , brown , l. r. , campargue , a. , chance , k. , cohen , e. a. , coudert , l. h. , devi , v. m. , drouin , b. j. , fayt , a. , flaud , j .- m . , gamache , r. r. , harrison , j. j. , hartmann , j .- m . , hill , c. , hodges , j. t. , jacquemart , d. , jolly , a. , lamouroux , j. , le roy , r. j. , li , g. , long , d. a. , lyulin , o. m. , mackie , c. j. , massie , s. t. , mikhailenko , s. , mller , h. s. p. , naumenko , o. v. , nikitin , a. v. , orphal , j. , perevalov , v. , perrin , a. , polovtseva , e. r. , richard , c. , smith , m. a. h. , starikova , e. , sung , k. , tashkun , s. , tennyson , j. , toon , g. c. , tyuterev , v. g. , and wagner , g. : the hitran2012 molecular spectroscopic database , j. quant . spectrosc . ra . , 130 , 450 , doi : http://dx.doi.org/10.1016/j.jqsrt.2013.07.002[10.1016/j.jqsrt.2013.07.002 ] , 2013 .
russell iii , j. m. , mlynczak , m. g. , gordley , l. l. , tansock jr .
, j. j. , and esplin , r. w. : overview of the saber experiment and preliminary calibration results , spie proc .
ser . , 3756 , 277288 ,
doi : http://dx.doi.org/10.1117/12.366382[10.1117/12.366382 ] , 1999 .
scheer , j. and reisin , e. r. : refinements of a classical technique of airglow spectroscopy , adv .
space res . ,
27 , 11531158 , doi : http://dx.doi.org/10.1016/s0273 - 1177(01)00150 - 8[10.1016/s0273 - 1177(01)00150 - 8 ] , 2001 . schmidt , c. , hppner , k. , and bittner , m. : a ground - based spectrometer equipped with an ingaas array for routine observations of rotational temperatures in the mesopause region , j. atmos . sol .- terr .
phy . , 102 , 125139 , doi : http://dx.doi.org/10.1016/j.jastp.2013.05.001[10.1016/j.jastp.2013.05.001 ] , 2013 .
semenov , a. i. , shefov , n. n. , and medvedeva , i. v. : empirical model of variations in the continuum emission in the upper atmosphere .
1 . intensity , geomagn .
aeron . , 54 , 488499 , doi : http://dx.doi.org/10.1134/s0016793214040148[10.1134/s0016793214040148 ] , 2014 .
shepherd , m. g. , evans , w. f. j. , hernandez , g. , offermann , d. , and takahashi , h. : global variability of mesospheric temperature : mean temperature field , j. geophys .
res . , 109 ,
d24117 , doi : http://dx.doi.org/10.1029/2004jd005054[10.1029/2004jd005054 ] , 2004 .
shiokawa , k. , otsuka , y. , suzuki , s. , katoh , t. , katoh , y. , satoh , m. , ogawa , t. , takahashi , h. , gobbi , d. , nakamura , t. , williams , b. p. , she , c .- y . ,
taguchi , m. , and shimomai , t. : development of airglow temperature photometers with cooled - ccd detectors , earth planets space , 59 , 585599 , doi : http://dx.doi.org/10.1186/bf03352721[10.1186/bf03352721 ] , 2007 .
smette , a. , sana , h. , noll , s. , horst , h. , kausch , w. , kimeswenger , s. , barden , m. , szyszka , c. , jones , a. m. , gallenne , a. , vinther , j. , ballester , p. , and taylor , j. : molecfit : a general tool for telluric absorption correction .
i. method and application to eso instruments , astron .
astrophys . , 576 ,
a77 , doi : http://dx.doi.org/10.1051/0004 - 6361/201423932[10.1051/0004 - 6361/201423932 ] , 2015 .
takahashi , h. , clemesha , b. r. , and batista , p. p. : predominant semi - annual oscillation of the upper mesospheric airglow intensities and temperatures in the equatorial region , j. atmos .
phys . , 57 , 407414 , 1995 .
takahashi , h. , gobbi , d. , batista , p. p. , melo , s. m. l. , teixeira , n. r. , and buriti , r. a. : dynamical influence on the equatorial airglow observed from the south american sector , adv .
space res . , 2 , 817825 , doi : http://dx.doi.org/10.1016/s0273 - 1177(97)00680 - 7[10.1016/s0273 - 1177(97)00680 - 7 ] , 1998 .
vallance jones , a. and gattinger , r. l. : the seasonal variation and excitation mechanism of the 1.58@xmath164@xmath165-@xmath166 twilight airglow band , planet .
space sci .
, 11 , 961974 , doi : http://dx.doi.org/10.1016/0032 - 0633(63)90123 - 5[10.1016/0032 - 0633(63)90123 - 5 ] , 1963 .
van der loo , m. p. j. and groenenboom , g. c. : erratum : `` theoretical transition probabilities for the meinel system '' [ j. chem
126 , 114314 ] , j. chem .
phys . , 128 , 159902 , doi : http://dx.doi.org/10.1063/1.2899016[10.1063/1.2899016 ] , 2008 . vernet , j. , dekker , dodorico , s. , kaper , l. , kjaergaard , p. , hammer , f. , randich , s. , zerbi , f. , groot , p. m. , hjorth , j. , guinouard , i. , navarro , r. , adolfse , t. , albers , p. w. , amans , j .- p . , andersen , j. j. , andersen , m. i. , binetruy , p. , bristow , p. , castillo , r. , chemla , f. , christensen , l. , conconi , p. , conzelmann , r. , dam , j. , de caprio , v. , de ugarte postigo , a. , delabre , b. , di marcantonio , p. , downing , m. , elswijk , e. , finger , g. , fischer , g. , flores , h. , franois , p. , goldoni , p. , guglielmi , l. , haigron , r. , hanenburg , h. , hendriks , i. , horrobin , m. , horville , d. , jessen , n. c. , kerber , f. , kern , l. , kiekebusch , m. , kleszcz , p. , klougart , j. , kragt , j. , larsen , h. h. , lizon , j .- l . , lucuix , c. , mainieri , v. , manuputy , r. , martayan , c. , mason , e. , mazzoleni , r. , michaelsen , n. , modigliani , a. , moehler , s. , mller , p. , norup srensen , a. , nrregaard , p. , proux , c. , patat , f. , pena , e. , pragt , j. , reinero , c. , rigal , f. , riva , m. , roelfsema , r. , royer , f. , sacco , g. , santin , p. , schoenmaker , t. , spano , p. , sweers , e. , ter horst , r. , tintori , m. , tromp , n. , van dael , p. , van der vliet , h. , venema , l. , vidali , m. , vinther , j. , vola , p. , winters , r. , wistisen , d. , wulterkens , g. , and zacchei , a. : x - shooter , the new wide band intermediate resolution spectrograph at the eso very large telescope , astron . astrophys . , 536 ,
a105 , doi : http://dx.doi.org/10.1051/0004 - 6361/201117752[10.1051/0004 - 6361/201117752 ] , 2011 .
von savigny , c. , mcdade , i. c. , eichmann , k .- u . , and burrows , j. p. : on the dependence of the oh@xmath167 meinel emission altitude on vibrational level : sciamachy observations and model simulations , atmos .
, 12 , 88138828 , doi : http://dx.doi.org/10.5194/acp-12 - 8813 - 2012[10.5194/acp-12 - 8813 - 2012 ] , 2012 .
von zahn , u. , hffner , j. , eska , v. , and alpers , m. : the mesopause altitude : only two distinctive levels worldwide ? , geophys .
lett . , 23 , 32313234 , doi : http://dx.doi.org/10.1029/96gl03041[10.1029/96gl03041 ] , 1996 . wallace , l. , hinkle , k. h. , livingston , w. c. , and davis , s. p. : an optical and near - infrared ( 29589250 ) solar flux atlas , astrophys . j. suppl .
s. , 195 , 6 , doi : http://dx.doi.org/10.1088/0067 - 0049/195/1/6[10.1088/0067 - 0049/195/1/6 ] , 2011 .
washenfelder , r. a. , toon , g. c. , blavier , j .- f .
, yang , z. , allen , n. t. , wennberg , p. o. , vay , s. a. , matross , d. m. , and daube , b. c. : carbon dioxide column abundances at the wisconsin tall tower site , j. geophys . res . , 111 ,
d22305 , doi : http://dx.doi.org/10.1029/2006jd007154[10.1029/2006jd007154 ] , 2006 .
watanabe , t. , nakamura , m. , and ogawa , t. : rocket measurements of o2 atmospheric and meinel bands in the airglow , j. geophys .
res . , 86 , 57685774 , doi : http://dx.doi.org/10.1029/ja086ia07p05768[10.1029/ja086ia07p05768 ] , 1981 .
xu , j. , smith , a. k. , jiang , g. , gao , h. , wei , y. , mlynczak , m. g. , and russell iii , j. m. : strong longitudinal variations in the nightglow , j. geophys .
res . , 37 ,
l21801 , doi : http://dx.doi.org/10.1029/2010gl043972[10.1029/2010gl043972 ] , 2010 .
xu , j. , gao , h. , smith , a. k. , and zhu , y. : using timed / saber nightglow observations to investigate hydroxyl emission mechanisms in the mesopause region , j. geophys .
res . , 117 ,
d02301 , doi : http://dx.doi.org/10.1029/2011jd016342[10.1029/2011jd016342 ] , 2012 .
yee , j .- h . ,
crowley , g. , roble , r. g. , skinner , w. r. , burrage , m. d. , and hays , p. b. : global simulations and observations of ( @xmath168 ) , ( @xmath169 ) and mesospheric nightglow emissions , j. geophys .
res . , 102 , 1994919968 , doi : http://dx.doi.org/10.1029/96ja01833[10.1029/96ja01833 ] , 1997 . | rotational temperatures @xmath0 derived from lines of the same band are an important method to study the dynamics and long - term trends in the mesopause region near 87 . to measure realistic temperatures , the rotational level populations have to be in local thermodynamic equilibrium ( lte ) .
however , this might not be fulfilled , especially at high emission altitudes . in order to quantify possible non - lte contributions to the @xmath0 as a function of the upper vibrational level @xmath1
, we studied a sample of 343 echelle spectra taken with the x - shooter spectrograph at the very large telescope at cerro paranal in chile .
these data allowed us to analyse 25 bands in each spectrum .
moreover , we could measure lines of @xmath2 , which peaks at about 94 to 95 , and @xmath3 with an emission peak at about 90 .
the latter altitude is reached in the second half of the night after a rise of several because of the decay of a daytime population of excited .
since the radiative lifetimes for the upper levels of the two bands are relatively long , the derived @xmath0 are not significantly affected by non - lte contributions .
these bands are well suited for a comparison with if the differences in the emission profiles are corrected . for different sample averages
, we made these corrections by using emission , @xmath3 emission , and -based temperature profile data from the multi - channel radiometer saber on the timed satellite .
the procedure relies on differences of profile - weighted saber temperatures . for an @xmath3-based reference profile at 90
, we found a good agreement of the with the saber - related temperatures , whereas the temperatures , especially for the high and even @xmath1 , showed significant excesses with a maximum of more than 10 for @xmath4 .
the exact value depends on the selected lines and molecular parameters .
we could also find a nocturnal trend towards higher non - lte effects , particularly for high @xmath1 .
the amplitude of these variations can be about 2 or less , which tends to be significantly smaller than the total amount of the non - lte contributions .
the variations revealed may be important for dynamical studies based on @xmath0 derived from bands with high @xmath1 . [ sec : intro ] studies of long - term trends as well as short - term variations of the temperature in the mesopause region are often based on so - called rotational temperatures @xmath0 derived from emission lines of the hydroxyl ( ) airglow ( e.g. * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ) , which is mostly caused by the chemical reaction of hydrogen ( ) and ozone ( ) @xcite at a typical height of 87 ( e.g. * ? ? ?
* ) .
ground - based instruments usually measure the intensities of a few lines with different rotational upper levels @xmath5 of a single band ( e.g. * ? ? ?
* ; * ? ? ?
* ) .
assuming a boltzmann distribution of the populations at the different rotational levels , a temperature can be derived , which is reliable if the frequency of thermalising collisions is high compared to the natural radiative lifetime of a state .
the latter ranges from about 58 to 5 for the relevant upper vibrational levels @xmath6 to 9 of the electronic ground state @xcite . for the peak emission height
, this appears to be sufficiently long to establish the assumed boltzmann distribution @xcite
. however , at higher altitudes with lower air density , collisions may be too infrequent , especially for the higher @xmath1 .
hence , the measured @xmath0 can deviate from the true temperatures due to the contribution of a population of molecules , where the condition of a local thermodynamic equilibrium ( lte ) for the rotational levels is not fulfilled .
comparisons of @xmath0 from various bands differing in @xmath1 show clear signatures of such non - lte effects @xcite .
the measured @xmath0 tend to increase with @xmath1 and even @xmath1 appear to indicate higher @xmath0 than the adjacent odd @xmath1 .
for this reason , the highest @xmath0 are found for bands with @xmath7 . as discussed by @xcite , this
can only be explained by a significant impact of a nearly nascent vibrational and rotational level distribution related to the hydrogen ozone reaction , which mostly populates @xmath8 and 8 @xcite .
the differences in the results for even and odd @xmath1 can then be understood by the higher initial @xmath0 for @xmath9 @xcite and the high probability of radiative transitions with @xmath10 @xcite .
the @xmath0 non - lte contributions also appear to vary with time , as observed by @xcite . at cerro
paranal ( 24.6s ) , they tend to enhance during the night when data of several years are averaged .
this is probably caused by a rising emission layer , which can be inferred from a decrease of the measured intensity @xcite .
apart from @xmath1 , the measured @xmath0 are also strongly affected by the choice of lines .
higher @xmath5 tend to show stronger non - lte contributions @xcite .
for this reason , only lines of the first three or four rotational levels are usually used for @xmath0 derivations ( e.g. * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ) . however , even this restriction is not sufficient to obtain line - independent @xmath0 .
the electronic ground level has two substates , @xmath11 and @xmath12 .
population differences of several per cent were found by @xcite for rotational levels of the same substate and for comparisons between the two substates .
the approach to estimate temperatures in the mesopause region by means of airglow lines with different rotational upper levels can also be applied to bands of molecular oxygen ( ) . however , long ground - based time series are only available for ( @xmath13 ) at about @xmath14 ( e.g. * ? ? ?
* ; * ? ? ?
* ) , which is sufficiently bright , hardly blended by lines , and not affected by self - absorption in the lower atmosphere like the stronger band .
since @xmath2 remains unresolved with the standard instruments in use , a simulation of the temperature dependence of the band shape is required ( e.g. * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ) .
in contrast to , the radiative lifetime @xmath15 of @xmath2 is relatively long . using the einstein-@xmath16 coefficients from the hitran2012 molecular line database @xcite
, @xmath15 is about 210s .
this suggests that @xmath2 @xmath0 are not significantly affected by non - lte contributions .
the @xmath17 state is essentially populated by atomic oxygen recombination and subsequent collisions @xcite , which results in an emission profile with a peak at about 94 to 95 ( derived from the band ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ) , i.e. @xmath2 is sensitive to a different part of the mesopause region compared to . with full widths at half maximum ( fwhm ) of 8 to 9 @xcite for both kinds of emission , there is only a minor overlap of their profiles . for a comparison with ,
transitions from the @xmath18 state of to the ground state @xmath19 are more promising since the measured nighttime emission peak heights of the band at 1.27 are close to those of @xcite . during most of the night ,
they are at about 90 . at the beginning of the night
, the effective emission heights can be several lower .
this behaviour is caused by the contribution of a daytime - related @xmath18 excitation by ozone photolysis to the nighttime - related atomic oxygen recombination @xcite .
the former is relevant for the night since the spontaneous radiative lifetime of @xmath18 is about 75min due to an einstein-@xmath16 coefficient of about @xmath20 for the band @xcite .
the daytime emission peaks at about 50 @xcite and effectively rises in altitude after sunset up to the nighttime peak due to the faster decay of the @xmath18 population by collisional quenching at lower altitudes .
although non - lte effects are negligible , only @xcite studied @xmath0 based on the bright @xmath3 band .
their sample consists of ground - based observations with a fourier transform spectrometer at maynooth , ireland ( 53.2n ) , and covers 91 twilight periods over 18 months .
the @xmath0 measurements required simulations of the band for different ambient temperatures due to the strong line blending at their resolution .
moreover , @xmath3 is strongly affected by self - absorption , which required careful radiative transfer calculations .
these efforts might explain the lack of @xmath3 @xmath0 in the literature . as an alternative , @xmath3 can be used .
although this band is fairly weak and partly blended with strong lines , it was studied by @xcite using spectroscopic data taken at zvenigorod , russia ( 55.7n ) , from 2010 to 2011 .
the band emission had to be modelled to retrieve @xmath0 . in this paper
, we will compare @xmath0 derived from 25 bands , @xmath2 , and @xmath3 to quantify the non - lte effects related to the measurements .
for this purpose , we will use 343 observations from the x - shooter echelle spectrograph @xcite at the very large telescope ( vlt ) of the european southern observatory ( eso ) at cerro paranal in chile ( 2635 m , 2438@xmath21s , 7024@xmath21w ) .
the data were already discussed by @xcite .
as x - shooter covers the wavelength range from 0.3 to 2.5 with a resolving power of at least @xmath22 , the different and bands can be studied in parallel and without the need of band simulations for the @xmath0 retrieval .
the resolution is sufficient to measure individual lines even in @xmath3 .
we will not directly compare the derived @xmath0 since the studied bands probe different parts of the mesopause temperature profile due to the differences in the altitude distributions of the emission .
in particular , the significant changes in the emission profile of @xmath3 @xcite have to be corrected .
for this reason , we will use volume emission rate ( ver ) and temperature profiles from satellite observations to derive temperature corrections depending on the ver differences of the bands to be compared .
for this purpose , data of the ten - channel broad - band radiometer saber @xcite on the timed satellite launched in 2001 are most suited .
the limb sounder measures profiles of the crucial @xmath3 band and two -related channels at 1.64 and 2.06 , which probe different upper vibrational levels @xmath1 ( @xmath23 and @xmath24 ; see * ? ? ?
* ) .
the latter is important since the emission peak heights depend on @xmath1 @xcite .
the temperature profiles are derived from the saber channels under the consideration of non - lte effects @xcite .
timed has a slowly - precessing polar orbit , which allows a 24 coverage at low and mid - latitudes within about 60 days @xcite .
since this results in only two narrow local time intervals for a given site and day of year , our study will focus on average x - shooter and saber data for certain periods .
the paper is structured as follows .
we will start with the description of the x - shooter and saber data used for this study ( sect .
[ sec : data ] ) .
then , we will describe line measurements and @xmath0 retrieval from the x - shooter data , the derivation of saber - based profiles suitable for the measured airglow bands , and the correction of the corresponding @xmath0 for the differences in the ver profiles ( sect .
[ sec : analysis ] ) . in sect .
[ sec : results ] , we will discuss the resulting estimates of the non - lte effects related to the @xmath0 and their change with nighttime and season .
finally , we will draw our conclusions ( sect .
[ sec : conclusions ] ) . |
strong interactions are hard to tackle and different tools or limits in the parameter space of a given asymptotically free theory have been proposed .
the large number of colors ( @xmath3 ) limit is a relevant example @xcite . but
how large should @xmath3 be for the counting scheme to be not only qualitatively but also quantitatively interesting for qcd ?
in this note we try to answer this question in terms of the nature of the low energy hadronic spectrum .
the physical situation where the large @xmath3 approximation seems closest to direct experimental check is perhaps low energy @xmath0 ( or meson meson ) scattering .
large @xmath3 predicts the low energy amplitude to be a sum of tree diagrams contact terms and resonance pole diagrams . with the two reasonable additions : truncation of number of resonances and some kind of regularization to give the resonances finite widths
, this corresponds to the way in which the analyses of experimentally derived partial wave amplitudes are usually performed .
we shall proceed in this manner with the additional requirement that all pre - regularized tree amplitudes be obtained from a chiral invariant ( modulo usual symmetry breaking terms ) lagrangian .
the most crucial partial wave for our considerations turns out to be the scalar , isotopic spin zero channel .
recently there has been renewed interest @xcite- @xcite in the low energy behavior of this channel .
especially , many physicists now believe in the existence of the light , broad @xmath4 resonance , sigma in the 500 - 600 mev region .
the sigma is also very likely the tip of an iceberg " which may consist of a nonet of light four quark - type scalars @xcite mixing @xcite with another heavier @xmath5 - type scalar nonet as well as a glueball .
however there has been some controversy as to the underlying nature of these low lying scalar objects . in the present context , it was argued @xcite that a four quark state in the leading large @xmath3 limit would choose to appear as two non - interacting @xmath6 states .
the necessity of such a state to explain the scattering then suggests a critical value of @xmath3 , below which a simple analysis of @xmath7 scattering up to a given energy ( taken as 1 gev for definiteness ) and based on the large @xmath3 approximation is not accurate . to obtain this critical value
, we make a reanalysis of meson - meson scattering as a function of the number of colors .
the @xmath3 dependences can be easily incorporated in the parameters ( such as the pion decay constant or the vector meson @xmath8 coupling to two pions ) of the theory and of the states whose large @xmath3 counting is known . the criterion for
an accurate description will be whether the unitarity bounds are satisfied up to the energy of at least 1 gev .
a variety of different approaches to meson - meson scattering have been developed in the literature and all are able to provide a reasonable fit to the data .
an approach to describing @xmath0 , @xmath4 scattering based on a conventional non - linear chiral lagrangian of pseudoscalar and vector fields augmented by scalar fields introduced in a chiral invariant manner was discussed in and will be adoped here .
this will have an advantage , for our present purpose , of not forcing the amplitudes to be unitary .
thus unitarity can be considered as a criterion for an accurate theory . in the physical @xmath9 case
, the experimental data from threshold to a bit beyond 1 gev can be fit by computing the tree amplitude from this lagrangian and choosing parameters to yield an approximate unitarization .
the following ingredients are present : i ) the current algebra " contact term , ii ) the vector meson exchange terms , iii ) the unitarized @xmath10(560 ) exchange terms and iv ) the unitarized @xmath11(980 ) exchange terms .
although the @xmath8(770 ) vector meson certainly is a crucial feature of low energy physics , a fit can be made @xcite in which the contribution ii ) is absent .
this results in a somewhat lighter and broader sigma meson , in agreement with other approaches which neglect the effect of the @xmath8 meson .
we note that the @xmath12 contribution is needed to keep the @xmath4 scattering amplitude within the unitarity bounds @xcite . in @xcite
it was noted that the @xmath12 contribution is also needed to provide the correct background which flips the sign of the @xmath13 resonance ( the ramsauer - townsend effect ) . for the purpose of checking the above strong interaction calculation
, the meson - meson scattering was also calculated in a general version of the _ linear _
su(3 ) sigma model , which contains both @xmath10(560 ) and @xmath13 candidates @xcite .
the procedure adopted was to calculate the tree amplitude and then to unitarize , without introducing any new parameters , by identifying the tree amplitude as the k - matrix amplitude .
this also gave a reasonable fit to the data , including the characteristic ramsauer - townsend effect of the @xmath13 resonance contribution . at a deeper and more realistic level of description in the linear sigma model framework ,
one expects two different chiral multiplets - @xmath6 as well as @xmath14 - to mix with each other .
a start on this model seems encouraging @xcite .
although we shall adopt the non - linear realization approach , our results are expected to be robust .
we will also focus , for simplicity , on the su(2 ) flavor limit of qcd . in section ii we study the real parts of the @xmath0 scattering amplitudes for the @xmath4 and @xmath15 channels as functions of the number of colors .
the contributions of the current algebra and the vector meson @xmath8 exchange pieces are both taken into account .
we show that chiral symmetry , crossing symmetry and the unitarity bounds are all satisfied in the leading order in @xmath3 for @xmath16 .
however the unitarity bound will not hold for lower values of @xmath3 .
section iii is dedicated to the low energy scalar ( the @xmath10 ) which is needed to unitarize the @xmath17 amplitude for @xmath18 .
we first analyze the @xmath4 pion - pion scattering amplitude as a function of the number of colors . in terms of the conventional amplitude , @xmath19 the @xmath20 amplitude is @xmath21 .
one gets the @xmath22 state by projecting out the correct partial wave .
we focus , following , on the real part of the scattering amplitude since the imaginary part can be determined , using the unitarity relation , from the real part .
of course , unitarity requires that the real part of the elastic scattering amplitude itself must satisfy the constraint @xmath23 .
this condition is crucial to obtain the critical number of colors in the following analysis .
the current algebra contribution to the conventional amplitude is @xmath24 where the pion decay constant , @xmath25 depends on @xmath3 as @xmath26 so that @xmath27 mev .
furthermore @xmath28 mev is independent of @xmath3 . in fig .
[ ca ] we plot the current algebra contribution to the real part of the @xmath4 partial wave amplitude , @xmath17 for increasing values of the number of colors @xmath3 .
we observe that the unitarity bound is violated for a value of @xmath29 which increases linearly with @xmath3 .
thus the net effect of increasing @xmath3 is to delay the onset of the unitarity bound violation . for @xmath30
the unitarity bound is satisfied up to about 1 gev .
thus the current algebra contribution alone indicates that @xmath3 should be of the order 13 to lead to a unitary chiral theory .
we now demonstrate that this result is strongly modified by the presence of the well established @xmath31 companion of the pion the @xmath8 vector meson .
the amplitude we consider now is obtained by adding to the current algebra contribution the following vector meson @xmath32 contribution : @xmath33 \ , \label{arho}\end{aligned}\ ] ] where @xmath34 is the @xmath35 coupling constant .
also , @xmath36 mev is independent of @xmath3 and @xmath37 it should be noted that the first term in eq .
( [ arho ] ) , which implies the existence of an additional non - resonant contact interaction other than the current algebra contribution , is required by the chiral symmetry when we include the @xmath8 vector meson contribution in a chiral invariant manner .
although the vector meson contribution to the @xmath4 partial wave amplitude is due only to the crossed channel , as noted in ref .
@xcite , adding the @xmath8(770 ) for the 3 color case greatly decreases the amount of unitarity violation .
however for three colors this contribution is not sufficient to keep the scattering amplitude completely within the unitarity bounds and a broad state ( the sigma ) is needed @xcite .
here we follow a different route and investigate the unitarity bound restoration in the absence of the sigma as a function of @xmath3 .
we compute analytically the relevant partial wave projections . in fig .
[ carho ] we plot the real part of the @xmath4 amplitude ( i.e. @xmath17 ) due to current algebra plus the @xmath8 contribution for increasing values of @xmath3 .
since in this channel the vector meson is never on shell we suppress the contribution of the width in the vector meson propagator in eq . ( [ arho ] ) .
we observe that the unitarity bound ( i.e. , @xmath38 ) is satisfied for @xmath16 till well beyond the 1 gev region .
however unitarity is still a problem for @xmath39 and @xmath40 colors . at energy scales larger than the one associated with the vector meson
clearly other resonances are needed @xcite but we shall not be concerned about those here .
it is also interesting to note that these considerations are essentially unchanged when the pion mass ( i.e. explicit chiral symmetry breaking in the lagrangian ) is set to zero .
we also note that the curves in fig .
[ carho ] are multiples of each other , so all of the extrema coincide as well as the points where the curves cross zero .
this is so since the @xmath3 dependent scattering amplitude which is the sum of the current algebra and vector meson contributions obeys , by construction : @xmath41 where @xmath42 is defined replacing @xmath25 and @xmath43 with the @xmath3 independent quantities @xmath44 and @xmath45 scaling property in eq .
( [ nc scaling ] ) is true for @xmath46 and @xmath47 where the width in the vector meson propagator in eq .
( [ arho ] ) vanishes . ] .
one might be worried that the critical value , @xmath2 determined by studying the @xmath4 channel would not hold for other channels .
this is not so since @xmath4 is actually the worst channel with respect to unitarity bound violation at the low energies of present interest .
indeed channels with larger @xmath48 have angular momentum suppression factors which delay the onset of such a violation .
let us consider as a relevant example the @xmath15 channel .
figure [ ca11 ] shows that , even including just the current algebra term , the unitarity bound is not an issue for 3 colors till energies well above the @xmath8 mass . partial wave amplitude due to current algebra plotted for the following increasing values of @xmath3 ( from left to right ) , @xmath49 .
the curve which first violates the unitarity bound corresponds to @xmath9.,title="fig:",width=264 ] the combined contributions of the vector meson @xmath8 term and the current algebra term in this channel as a function of the number of colors is shown in fig .
[ carho11 ] . for this channel
the non zero @xmath8 width has been of course taken into acount .
the already narrow @xmath8 becomes more and more narrow as we increase @xmath3 while the @xmath3 effects are not very relevant concerning unitarity bounds .
near the vector resonance we observe practically a breit - wigner consistent with the well known cancellation between the current algebra contribution and part of the non resonant vector meson contribution in chiral models .
it is amusing that the vector meson plays a major role in helping satisfy the unitarity bound for large @xmath3 not in the direct channel but in the crossed one .
it is also partial wave amplitude due to the current algebra @xmath50 terms , plotted for the following increasing values of @xmath3 , @xmath51 .
@xmath9 is the curve with the broadest @xmath8 meson.,title="fig:",width=264 ] interesting that the critical number of colors emerges from the relation between @xmath52 , @xmath25 and @xmath53 while explicit chiral symmetry breaking has been seen to play a negligible role . notice that in the most crucial @xmath4 channel the vector meson width can be deleted to good accuracy and plays no role in satisfying the unitarity bound .
thus apart from the truncation of higher mass resonant states the calculation of this partial wave is exactly that of the leading large @xmath3 approximation .
we also note that the @xmath54 partial wave amplitude computed in this way obeys the unitarity bound to well above 1 gev @xcite .
we have thus demonstrated that for @xmath16 and two flavor qcd the low energy theory is crossing symmetric and unitary up to well beyond the one gev region considering just the pion and the @xmath8 vector meson .
this supports the usual expectation that when @xmath3 is large _ enough _ the meson - meson scattering theory is well represented by including just the relevant @xmath31 resonances at the tree level .
in order to explain low energy @xmath0 scattering for the physical value @xmath9 ( and for the cases @xmath55 ) using the attractive notion of tree diagram dominance involving near by resonances , it thus seems necessary to include a scalar singlet resonance like the light sigma which is likely to be of four quark " type and hence not to contribute in the leading large @xmath3 approximation . as mentioned before ,
many models @xcite-@xcite have recently been proposed which successfully employ the sigma in one way or another .
some discussions of @xmath3 dependence have also been very recently given @xcite .
using the approach of refs .
@xcite it is easy to make a quick estimate of the mass of the needed sigma particle .
a unitarization procedure is needed when the current algebra plus the @xmath8 amplitude first starts violating the unitarity bound .
denote this point as @xmath56 .
the @xmath10 pole structure is such that the real part of its amplitude is positive for @xmath57 and negative for @xmath58 . identifying the squared sigma mass roughly with @xmath56
will then give a negative contribution where the real part of the amplitude exceeds @xmath59 .
in the case when only the current algebra term is included we get @xmath60 this shows that the squared mass of the sigma meson needed to restore unitarity for @xmath61 increases roughly linearly with @xmath3 .
this estimate gets modified a bit when we include the vector meson ( see fig . [ masse ] ) , yielding @xmath62 , where @xmath63 is to be obtained from fig . [ carho ] .
for which unitarity bounds are first violated as function of @xmath3 and normalized to @xmath64 .
the dashed line corresponds to the pure current algebra contribution while the solid line to current algebra @xmath50 contribution.,title="fig:",width=226 ] in the case of three colors we then get @xmath65 mev which is close to the best fit to the @xmath0 scattering data determined in refs . @xcite .
finally , we briefly discuss the possible four quark nature of the sigma .
most naively , the needed particle is rather light while a usual constituent quark @xmath6 bound state should be a somewhat heavy p - wave state .
of course we have been working in an effective lagrangian framework which does not directly tell us anything about the inner structure of the sigma meson .
indirectly one can examine the three flavor generalization and observe that the lightest nine scalars found in the present manner have the right flavor quantum numbers to make up a usual nonet .
however the characteristics of this nonet , such as usual vs. inverted mass ordering and the value of the mixing angle between the two isoscalars , turn out @xcite to favor a four quark interpretation .
still the four quarks may be of @xmath66 type @xcite , meson - meson molecule type @xcite or any linear combination .
for our present purpose it is enough to know that they are not of @xmath6 type .
we studied the dependence on number of colors of @xmath0 scattering .
we showed the existence of a critical number of colors for and above which , chiral theory is unitary at leading order in the @xmath1 expansion to beyond the one gev range .
the critical number of colors is @xmath2 and is insensitive to the explicit breaking of chiral symmetry .
for @xmath67 the @xmath31 resonances are sufficient to keep the low energy theory unitary . below this critical value
the @xmath10 meson is needed to keep the amplitude within the unitarity bounds .
this is a light , broad object which is unlikely to be of @xmath6 nature .
it is a pleasure to thank a. abdel rehim , d. black , a. h. fariborz , a. mcsy , s. nasri and k. tuominen for discussions and for careful reading of the manuscript .
the work of f.s . is supported by the marie curie fellowship under contract mcfi-2001 - 00181 .
the work of j.s . is supported in part by the u. s. doe under contract no .
de - fg-02 - 85er 40231 .
see the dedicated conference proceedings , s. ishida et al possible existence of the sigma meson and its implication to hadron physics " , kek proceedings 2000 - 4 , soryyushiron kenkyu 102 , no . 5 , 2001 .
additional points of view are expressed in the proceedings , d. amelin and a.m. zaitsev `` hadron spectroscopy '' , ninth international conference on hadron spectroscopy , protvino , russia(2001 ) .
e. van beveren , t. a. rijken , k. metzger , c. dullemond , g. rupp and j. e. ribeiro , z. phys .
c * 30 * , 615 ( 1986 ) ; e. van beveren and g. rupp , eur .
j. c * 10 * , 469 ( 1999 ) . see also j. j. de swart , p. m. maessen and t. a. rijken , talk given at u.s .
/ japan seminar : the hyperon - nucleon interaction , maui , hi , 25 - 28 oct 1993 [ arxiv : nucl - th/9405008 ] .
d. morgan and m. r. pennington , phys .
d * 48 * , 1185 ( 1993 ) .
a. a. bolokhov , a. n. manashov , v. v. vereshagin and v. v. polyakov , phys .
d * 48 * , 3090 ( 1993 ) .
see also v. a. andrianov and a. n. manashov , mod .
lett . a * 8 * , 2199 ( 1993 ) .
extension of this string - like approach to the @xmath68 case has been made in v. v. vereshagin , phys .
d * 55 * , 5349 ( 1997 ) and in a. v. vereshagin and v. v. vereshagin , phys .
d * 59 * , 016002 ( 1999 ) .
n. n. achasov and g. n. shestakov , phys .
d * 49 * , 5779 ( 1994 ) .
r. kamnski , l. leniak and j. p. maillet , phys .
d * 50 * , 3145 ( 1994 ) .
f. sannino and j. schechter , phys .
d * 52 * , 96 ( 1995 ) .
n. a. tornqvist , z. phys .
c * 68 * , 647 ( 1995 ) and references therein .
in addition see n. a. trnqvist and m. roos , phys .
lett . *
76 * , 1575 ( 1996 ) ; n. a. tornqvist , talk given at 7th international conference on hadron spectroscopy ( hadron 97 ) , upton , ny , 25 - 30 aug 1997 and at eurodaphne meeting , barcelona , spain , 6 - 9 nov 1997 [ arxiv : hep - ph/9711483 ] ; phys .
b * 426 * , 105 ( 1998 ) .
r. delbourgo and m. d. scadron , mod .
phys . lett .
a * 10 * , 251 ( 1995 ) ; see also d. atkinson , m. harada and a. i. sanda , phys . rev .
d * 46 * , 3884 ( 1992 ) . g. janssen , b. c. pearce , k. holinde and j. speth , phys .
d * 52 * , 2690 ( 1995 ) .
m. svec , phys .
d * 53 * , 2343 ( 1996 ) .
s. ishida , m. ishida , h. takahashi , t. ishida , k. takamatsu and t. tsuru , prog .
* 95 * , 745 ( 1996 ) ; s. ishida , m. ishida , t. ishida , k. takamatsu and t. tsuru , prog .
phys . * 98 * , 621 ( 1997 ) .
see also m. ishida and s. ishida , talk given at 7th international conference on hadron spectroscopy ( hadron 97 ) , upton , ny , 25 - 30 aug 1997 [ arxiv : hep - ph/9712231 ] .
m. harada , f. sannino and j. schechter , phys .
d * 54 * , 1991 ( 1996 ) .
m. harada , f. sannino and j. schechter , phys .
lett . * 78 * , 1603 ( 1997 ) .
d. black , a. h. fariborz , f. sannino and j. schechter , phys . rev .
d * 58 * , 054012 ( 1998 ) .
d. black , a. h. fariborz , f. sannino and j. schechter , phys .
d * 59 * , 074026 ( 1999 ) .
j. a. oller , e. oset and j. r. pelaez , phys .
lett . *
80 * , 3452 ( 1998 ) .
see also k. igi and k. i. hikasa , phys .
d * 59 * , 034005 ( 1999 ) .
a. v. anisovich and a. v. sarantsev , phys .
b * 413 * , 137 ( 1997 ) .
v. elias , a. h. fariborz , f. shi and t. g. steele , nucl . phys .
a * 633 * , 279 ( 1998 ) .
v. dmitrasinovic , phys .
c * 53 * , 1383 ( 1996 ) .
p. minkowski and w. ochs , eur .
j. c * 9 * , 283 ( 1999 ) .
s. godfrey and j. napolitano , rev .
phys . * 71 * , 1411 ( 1999 ) .
l. burakovsky and t. goldman , phys .
d * 57 * , 2879 ( 1998 ) .
a. h. fariborz and j. schechter , phys .
d * 60 * , 034002 ( 1999 ) .
d. black , a. h. fariborz and j. schechter , phys .
d * 61 * , 074030 ( 2000 ) .
see also v. bernard , n. kaiser and u. g. meissner , phys .
d * 44 * , 3698 ( 1991 ) .
d. black , a. h. fariborz and j. schechter , phys .
d * 61 * , 074001 ( 2000 ) .
l. s. celenza , s. f. gao , b. huang , h. wang and c. m. shakin , phys .
c * 61 * , 035201 ( 2000 ) .
d. black , a. h. fariborz , s. moussa , s. nasri and j. schechter , phys .
d * 64 * , 014031 ( 2001 ) .
r. l. jaffe , phys .
d * 15 * , 267 ( 1977 ) ; phys .
d * 15 * , 281 ( 1977 ) .
in addition to @xcite and @xcite above see t. teshima , i. kitamura and n. morisita , j. phys .
g * 28 * , 1391 ( 2002 ) ; f. close and n. tornqvist , _ ibid . _
* 28 * , r249 ( 2002 ) and a. h. fariborz , arxiv : hep - ph/0302133 . see section v of @xcite above .
see @xcite above and also m. s. chanowitz and w. kilgore , phys .
b * 322 * , 147 ( 1994 ) .
j. r. pelaez , arxiv : hep - ph/0306063 ; m. uehara , arxiv : hep - ph/0308241 .
black et al @xcite above ; v. cirigliano , g. ecker , h. neufeld and a. pich , jhep * 0306 * , 012 ( 2003 ) ; j. a. oller , arxiv : hep - ph/0306031
. j. d. weinstein and n. isgur , phys .
* 48 * , 659 ( 1982 ) . | we study the dependence on the number of colors of the leading @xmath0 scattering amplitude in chiral dynamics .
we demonstrate the existence of a critical number of colors for and above which the low energy @xmath0 scattering amplitude computed from the simple sum of the current algebra and vector meson terms is crossing symmetric and unitary at leading order in a truncated and regularized @xmath1 expansion .
the critical number of colors turns out to be @xmath2 and is insensitive to the explicit breaking of chiral symmetry . below this critical value
, an additional state is needed to enforce the unitarity bound ; it is a broad one , most likely of four quark " nature . |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.