1: \documentstyle[aps,epsfig,prl]{revtex}

2: %labels,showkeys,a4

3: \newcommand{\mn}[1]{\marginpar{\tiny#1}}

4: %\newcommand{\pagefinchapitre}{\goodbreak}

5: \def\comment#1{}

6: \def\cm#1{}

7: 

8: \everymath={\displaystyle}

9: 

10: 

11: \newcommand{\lambdap}{\lambda'}

12: \newcommand{\lf}[2]{{#1}/{#2}}

13: \newcommand{\bfOmega}{{\Omega}}

14: \newcommand{\LL}{\Lambda}

15: \newcommand{\x}{{\Lambda}/{M (T=0)}}

16: \newcommand{\N}{N_c}

17: \newcommand{\sla}[1]{{\hspace{1pt}/\!\!\!\hspace{-.5pt}#1\,\,\,}\!\!}

18: \newcommand{\dslash}{\partial\!\!\!/}

19: \newcommand{\sslash}[1]{{#1}\!\!\!/~}

20: \newcommand{\gd}[1]{\gamma_{#1}}

21: \newcommand{\gu}[1]{\gamma^{#1}}

22: \newcommand{\ld}[1]{\lambda_{#1}}

23: \newcommand{\lu}[1]{\lambda^{#1}}

24: \newcommand{\pdd}[1]{\partial_{#1}}

25: \newcommand{\pu}[1]{\partial^{#1}}

26: \newcommand{\p}{\partial}

27: \newcommand{\f}[2]{\frac{#1}{#2}}

28: \newcommand{\tr}{{\rm tr}}

29: \newcommand{\be}{\begin{equation}}

30: \newcommand{\ee}{\end{equation}}

31: \newcommand{\beqn}{\begin{eqnarray}}

32: \newcommand{\eeqn}{\end{eqnarray}}

33: \newcommand{\Tr}{{\rm Tr}}

34: \newcommand{\un}{1\!\!1}

35: \newcommand{\tg}{{\rm tg}}

36: \newcommand{\sh}{{\rm sh}}

37: \newcommand{\ch}{{\rm ch}}

38: \newcommand{\fpi}{$f_{\pi}$}

39: \newcommand{\mpi}{$m_{\pi}$}

40: \newcommand{\s}{\sigma}

41: \newcommand{\st}{\sigma^T}

42: \newcommand{\Sig}{{\Sigma_0}}

43: \newcommand{\SigM}{{M}}

44: \newcommand{\Sigt}{\Sigma^T}

45: \newcommand{\G}{\Gamma}

46: \newcommand{\Gd}{\Gamma^d}

47: \newcommand{\ginv}{g^{-1}}

48: \newcommand{\gdeux}{g^2}

49: \newcommand{\ginvdeux}{g^{-2}}

50: \newcommand{\egrave}{\`e}

51: \newcommand{\ecirc}{\^e}

52: \newcommand{\agrave}{\`a}

53: \newcommand{\ugrave}{\`u}

54: \newcommand{\Vinv}{V^{-1}}

55: \newcommand{\qq}{q^2}

56: \newcommand{\micarre}{m_i^2}

57: \newcommand{\mjcarre}{m_j^2}

58: \newcommand{\diag}{\mbox{diag}}

59: \newcommand{\D}{\mbox{$\Delta$}}

60: \newcommand{\Dtilde}{\mbox{$\tilde{\Delta}$}}

61: \newcommand{\calL}{\mbox{${\cal L}$}}

62: \newcommand{\calZ}{\mbox{${\cal Z}$}}

63: \newcommand{\calD}{\mbox{${\cal D}$}}

64: \newcommand{\psibar}{\bar{\psi}}

65: \newcommand{\etabar}{\bar{\eta}}

66: \newcommand{\ratio}{\f{\N}{2g_0}}

67: \newcommand{\betacrit}{\beta^{\rm cr}}

68: \def\cm#1{}

69: \def\sbf#1{\mbox{\scriptsize{\bf #1}}}

70: \newcommand{\sfrac}[2]{\raisebox{0.095ex}{\footnotesize${\frac{#1}{#2}}$}}

71:  \def\lfrac#1#2{{{{#1}/{#2}}}}

72:  \def\fsz#1{{\footnotesize #1}}

73:  \def\Kappa{K}

74: \newcommand{\sdag}{{\scriptsize \dagger}}

75: \def\comment#1{}

76: %\newcommand{\Tr}{\mbox{Tr\,}}

77: \def\cc{{\rm c.c.}}

78: %\newcommand{\tr}{\mbox{tr\,}}

79: \newcommand{\Sbar}{~\bar{}\!\!S}

80: %\newcommand{\sla}[1]{{\hspace{1pt}/\!\!\!\hspace{-.5pt}#1\,\,\,}\!\!}

81: \def\lsim{\raisebox{-1.1mm}{$\,{\displaystyle\mathop{\sim}^{<}}\,$}}

82: %\renewcommand{\xi}{M}

83: %

84: \title{Nonlinear sigma model

85: approach for chiral

86: fluctuations \\ and symmetry breakdown

87: in Nambu--Jona-Lasinio model.}

88: \author{Egor Babaev\footnote {

89: email: egor@teorfys.uu.se

90: Tel: +46-18-391902, Fax +46-18-533180

91: }}

92: \address{

93: Institute for Theoretical Physics, Uppsala University

94: Box 803, S-75108 Uppsala, Sweden }

95: \begin{document}

96: \maketitle

97: \begin{abstract}

98: This paper is organized in two parts. We start 

99: with an observation that the recent claim

100: that the chiral symmetry in NJL

101: model is necessarily 

102: restored by violent chiral fluctuations at $N_c=3$

103:  (H. Kleinert and B. Van den Bossche, Phys. Lett. B 474

104: 336 (2000) ) appears to be incorrect since 

105: the critical stiffness of the effective

106: nonlinear sigma model used in the above reference is

107: not an universal quantity in 3+1 - dimensions.

108: In the second part we discuss 

109: a modified NJL model, where the critical 

110: stiffness is expressed via an additional cutoff parameter. This

111: model displays a symmetry breakdown 

112: and also under certain

113: conditions the chiral fluctuations give rise to a phase analogous to 

114: pseudogap phase of strong-coupling and low carrier 

115: density superconductors.

116: \end{abstract}

117: %

118: \section{Introduction}

119: %

120: Many concepts of particle physics have a close

121: relation to superconductivity, for example

122: Nambu--Jona-Lasinio model \cite{NJLM}-\cite{ht}

123: was proposed in analogy to BCS theory and is considered to be a low-energy effective

124: theory of QCD. Recently there was made a substantial

125: progress in the theory of superconductivity

126: in systems with strong attraction and low carrier

127: density.

128: Namely

129: it was observed that away from

130: the limits of infinitesimally weak coupling strength

131: or very high carrier density, the BCS-like mean-field theories

132: are qualitatively wrong and these systems

133: possess along with a superconductive phase

134: an additional phase where

135: there exist Cooper pairs but no symmetry is broken

136: due to violent  fluctuations

137: ({\it pseudogap phase}). What may

138: be regarded as an indication of the

139: importance of this concept to particle physics is

140: that recently we have found a formation

141: of the pseudogap phase

142: due to dynamic quantum fluctuations at low $N$

143: %was found 

144: in the chiral Gross-Neveu model \cite{GNM} in $2+\epsilon $

145: dimensions \cite{gn1}. 

146: The chiral GN model at low $N$ exhibit two 

147: phase transition at two characteristic values 

148: of renormalized coupling constant $g$. 

149: At a certain value $g^*$ a gap modulus 

150: forms locally, but there exists as well 

151: another characteristic 

152: coupling value $g_{KT} > g^*$ where the

153: system undergoes a Kosterlitz-Thouless transition into 

154: a quasiordered state. 

155: %For a given $\epsilon$ there 

156: %always exists a characteristic value of $N$, when 

157: %the number of field components drops below this value 

158: %the  system

159: %does not undergo a Kostrelitz-Thouless transition 

160: %irrespective of the magnitude of a renormalized coupling constant

161: %$g$ \cite{gn1}.

162: The region between $g_{KT}$ and $g^*$ is 

163: analogous to the pseudogap phase in superconductors.

164: At very large $N$, $g_{KT}$ merges with $g^*$ thus recovering 

165: BCS-like scenario for the phase transition in the chiral GN model.

166: Recently,  an attempt was made 

167: \cite{kb} to generalize this result to the NJL model that lead 

168: the  authors of \cite{kb} to the

169: conclusion that at $N_c=3$

170: the NJL model does not display a spontaneous breakdown

171: of the chiral symmetry due to  fluctuations.

172:  The paper \cite{kb}

173: is at the moment a subject of numerous

174: controversial discussions. Below we present a "no-go" result

175: that  one can not prove in principle in analogy to \cite{gn1}

176: the necessary

177: restoration of the chiral symmetry at low $N_c$ in the NJL model.

178: As it is discussed in details below, the reason is the nonuniversality of the

179: critical stiffness of the effective $O(4)$ - nonlinear sigma model (NLSM)

180: in $3+1$-dimensions.  This results in

181: the fact that the $3+1$-dimensional

182: theory possesses an additional nonspecified

183: fitting parameter that should be fixed from phenomenological

184: considerations. 

185: This is in contrast to the chiral GN model in $2+\epsilon$-dimensions

186: where 

187: one can prove that the model does not exhibit 

188: a quasi-long range order when the number of field 

189: components  $N$ drops below $8$ \cite{gn1}.

190: 

191: 

192: In the last section 

193: %account the above circumstance 

194: we discuss possibility

195: of formation of a phase analogous to the pseudogap phase

196: in a modified NJL model

197: taking special care of the existence of an

198: additional cutoff  parameter. 

199: %In conclusion we also discuss

200: %the  roles of thermal and zero temperature chiral fluctuations.

201: 

202: %However in this

203: %paper we show that no direct generalization

204: %of result \cite{gn1} to NJL model is possible.

205: As it was mentioned above the subject of the

206: discussion, which is the possibility of the 

207: restoration of chiral symmetry

208: by directional fluctuations in a degenerate valley 

209: of an effective potential while preserving

210: nonzero gap modulus locally, is closely related

211: to the pseudogap phenomena in superconductors.

212: Separation of the  temperatures of the pair formation and

213: of the onset  of the phase coherence (pair condensation)

214: in strong-coupling superconductors

215: is in fact known already for many years (Crossover from BCS

216: superconductivity to -- Bose-Einstein Condensation (BEC)

217: of tightly bound fermion pairs)

218: \cite{Le,N}.

219: Intensive theoretical studies

220: of these phenomena in the past several years

221: (see for example \cite{sc}-\cite{pist}),

222: were sparked by experimental

223: results on

224: underdoped (low carrier density)

225: cuprates that display  a "gap-like" (pseudogap) feature

226: {\it above} $T_c$. 

227: The pseudogap  disappears only at the substantially higher

228: temperature $T^*$.

229: There is experimental evidence that

230: this phenomenon in high-$T_c$ superconductors

231: may be connected with precritical pairing

232: fluctuations above $T_c$.

233: 

234: Due to intimate relation of

235: many problems in particle

236: physics to superconductivity

237: it seems to be natural to guess that

238: pseudogap may become a

239: fruitful concept in high energy physics too.

240: %In the following section

241: %we reviewing two

242: %models that do display pseudogap behavior: i.e.

243: %destruction of long-range or quasi long-range order

244: %due to phase disorder transition rather than pairbreaking.

245: %Then we suggest existence of the same phenomena

246: %in QCD at finite temperature and propose

247: %a nonlinear-sigma model with temperature-dependent

248: %stiffness coefficient as a toy model for QCD that would

249: %display two characteristic temperatures corresponding to

250: %the temperatures of pair formation and pair condensation

251: %of a strong-coupling

252: %and low carrier density superconductors.

253: %In conclusion we show  also failure of the

254: %nonlinear-sigma model argument in \cite{kb}

255: %that lead the authors of \cite{kb} to the

256: %conclusion that NJL does not display

257: %chiral  symmetry breakdown for $N_c=3$.

258: \comment{

259: The paper is organized as follows:

260: In section (II) we review

261: strong-coupling and low carrier density theories

262: of superconductivity and pseudogap phase since

263: with it we can gain more insight into

264: a possible analogous phenomena in QCD.

265: % whether one can study phase

266: %diagram of NJL model at finite temperatures within nonlinear

267: %sigma-model approach.

268: Then in the section (III)

269: we discuss  the appearance of the pseudogap phase

270: in the chiral Gross-Neveu model at low N.

271: Then we show failure of the attempt

272: of generalization of our results on GN model

273: to NJL model, namely 4D O(4) non-linear sigma model

274: approach proposed by Kleinert and Van den Bossche from

275: which authors \cite{kb}came to the

276: conclusion of absence

277: of  the chiral symmetry breakdown in  NJL

278: model at zero temperature.

279:  In conclusion we discuss possibility of construction

280: of a toy model with pseudogap behavior for QCD.}

281: We  start with a brief introduction to this phenomena in superconductors

282: and discuss its possible implications for QCD.

283: 

284: It is a well known fact that the 

285:  BCS theory describes perfectly metallic superconductors.

286: However it failed to describe even qualitatively

287: superconductivity in underdoped High-$T_c$ compounds.

288: One of the most exotic properties of the later materials

289: is the existence of a  pseudogap  in the spectrum of the normal state

290: well above critical temperature. From experimental point

291: of view this  manifests itself

292: as an essential suppression of low frequency spectral weight thus

293: being in contrast to exactly zero spectral weight in the case of

294: superconductive gap. Moreover spectroscopy experiments

295: show that  superconductive gap evolves

296: smoothly by magnitude and wave vector dependence to the pseudogap

297: in the normal state. Except for it NMR and tunneling

298:  experiments indicate

299: existence of incoherent Cooper pairs well above $T_c$. In principle it

300: is easy to guess what is hidden behind this circumstances,

301:  and why BCS theory is incapable to describe it.

302: Let us imagine for a moment that we were able

303: to bind electrons in Cooper pairs infinitely tightly -

304: obviously this implies that characteristic temperature

305: of thermal pair decomposition will be also

306: infinitely high, however this does not imply that     the

307: long-range order will survive at infinitely high temperatures.

308: As it was first observed in \cite{N}, the

309: long-range order will be destroyed in a similar way as it happens say in

310: superfluid ${}^4He$ i.e. tightly bound Cooper pairs,

311: at certain temperature will acquire a nonzero momentum and

312: thus we will have a gas of tightly bound

313: Cooper pairs but no macroscopic occupation of

314: zero momentum level ${\bf q} =0$

315: and with it no long-range order. Thus a phase diagram

316: of a strong coupling superconductor has three regions:

317: \begin{itemize}

318: \item

319: Superconductive phase where there are condensed fermion pairs,

320: \item

321: {\it Pseudogap} phase where there exist

322: fermion pairs but there is no condensate

323: and with it there is no symmetry breakdown and no superconductivity,

324: \item

325:  Normal phase with thermally decomposed Cooper pairs.

326: \end{itemize}

327: Of course, the existence  of bound pairs above critical temperature

328: will result in deviations from Fermi-liquid behavior that

329: makes pseudogap phase to be a very interesting object of

330: study.

331: In order to describe superconductivity in a such system

332: %with a strong attractive interaction

333: theory should incorporate pairs with

334: nonzero momentum. Thus { \it the BCS scenario

335: is invalid for description of spontaneous symmetry breakdown

336: in a system

337: with strong attractive interaction or low carrier density}

338: (see \cite{N}- \cite{pist}). So in principle

339: in a  strong-coupling superconductor the onset of a long range order has

340: nothing to do with a pair formation transition.

341: % Former process is a property

342: %of non-ideal Bose gas of Cooper pairs.

343: Existence of the paired fermions is a necessary but not sufficient condition for

344: symmetry breakdown.

345: BCS limit is a rather exotic case

346: of infinitesimally weak coupling strength and high carrier density

347: when the disappearance of superconductivity

348: can  be {\it approximately} described  as  a pairbreaking transition.

349: Strong-coupling limit is another exotic

350: case when temperatures of the pair decomposition and

351: symmetry breakdown can be arbitrarily separated. There is nothing surprising

352: in it: formally in the case of Bose condensation of ${}^4He$ we can also

353: introduce a characteristic

354: temperature of thermal decomposition of a ${} He$  atom,

355: however this would not mean that this temperature will be somehow

356: related  to  the temperature of the Bose condensation of the

357: gas of these atoms.

358: 

359: Schematic phase diagram of a superconductor

360: is shown in Fig ~1.

361: %\vskip 3cm

362: %

363: \begin{figure}[tb]

364: \input Phases.tps

365: \caption[]{Schematic phase diagram of a superconductor

366: with arbitrary coupling strength. In the strong coupling limit,

367: temperature of superconductive phase transition

368: tends to a plateau value corresponding

369: to temperature of Bose condensation of the gas of tightly

370: bound fermion pairs, whereas characteristic temperature of

371: thermal pair decomposition grows monotonously

372: as a function of the coupling strength.}

373: \label{phases.tps}\end{figure}

374: %

375: One can obtain a

376: pseudogap phase starting from

377: BCS Hamiltonian. This was first done in a pioneering work

378: by  Nozieres and Schmitt-Rink \cite{N}

379: and in a formalism of functional integral by Sa de Melo, Randeria and Engelbrecht

380: \cite{R}.

381: In order to study behavior of $T_c$ one should solve

382: a set of  number and gap equations including fluctuation

383: corrections.

384: In the BCS limit $T_c$ is not affected substantially by the gaussian

385: corrections and superconductive transition

386: can be described by a 

387: mean-field theory and correspondingly $T_c \approx T^*$

388: \footnote{As it was first discussed

389: in the sixties \cite{asl}, even in the BCS superconductors there is a narrow

390: fluctuation region near $T_c$

391: that gives rise e.g. to a so-called paraconductivity effect.

392: In particle physics, it was pointed on this

393: phenomenon by Hatsuda and Kunihiro \cite{ht2,ht}. }.

394: In the opposite limit numerical solution, and analytic

395: perturbative treatment \cite{N,R} shows

396: that the temperature of the superconductive phase transition tends

397: to a constant value

398: $ T_c= \left[{n}/{2\zeta(3/2)}\right]^{2/3}{\pi}/{m}$

399:  that does not depend on the coupling strength

400: and is equal to the temperature of the condensation of the

401: ideal Bose gas of particles of mass $2m$ and density $n/2$,

402: where $m$ and $n$ are the mass and the 

403: density of electrons. Qualitatively the same 

404: behavior is reproduced in a nonperturbative 

405:  nonlinear sigma model

406: approach analogous to discussed in this article \cite{sc}.

407: %The BCS - BEC crossover

408: %was recently studied in details perturbatively

409: %in a  variety of approximations

410: %(see for example \cite{R,H,pist}). 

411: One can observe that although there

412: is no superconductivity in the pseudogap phase

413: it exhibits reach exotic non-Fermi-liquid behavior due to

414: pairing correlations  that makes it as

415: interesting object of theoretical and

416: experimental study as superconductive phase itself.

417: \comment{ In particular,

418: along with specific heat, optical conductivity

419: and tunneling experiments there are following

420: circumstances observed in the pseudogap phase:

421: In experiments on YBCO a significant suppression of

422: in-plane conductivity $\sigma_{ab}(\omega)$

423: was observed at frequencies below 500 ${\rm cm}^{-1}$  beginning

424: at temperatures much above $T_c$.

425: Experiments on underdoped samples revealed

426: deviations from the linear resistivity law. In particular

427: $\sigma_{ab}(\omega=0;T)$

428: increases slightly with decreasing $T$

429: below a certain temperature.

430: NMR and neutrons observations

431: show that below temperatures $T^*$ much higher than $T_c$,

432:  spin susceptibility starts decreasing.

433: }

434: %

435: %The region

436: %of non-Fermi liquid behavior

437: %between $T_c$ and $T^*$  that is called  pseudogap phase,

438: %however 

439: One should stress however that the

440: term "pseudogap", originated in early experimental

441: papers, may seem somewhat misleading since even though

442: a substantial depletion of low-frequency spectral weight is

443: observed in this region experimentally - there is

444: {\it no true gap} in the spectrum.

445: %One should note as well that there is { \it no

446: %proper phase transition at $T^*$}, which

447: %is simply characteristic temperature of crossover

448: %when certain fraction of incoherent Cooper

449: %pairs is decomposed, one should not expect such a process to be a true

450: %phase transition. Even though the position of this temperature

451: %may be reasonable estimated with mean-field methods,

452: %second-order phase transition at $T^*$ is certainly an

453: %artifact of mean-field approximation. Experiments  on specific

454: %heat indicate however certain features at this characteristic

455: %temperature.

456: \section{Chiral fluctuations in the the NJL model at zero temperature}

457: As it was mentioned above, recently it was made an attempt

458: \cite{kb} of generalization to the NJL model

459: the nonlinear-sigma approach  for description of chiral

460: fluctuations proposed in \cite{gn1,sc}.

461: The authors \cite{kb}

462: claimed that at $N_c=3$ the NJL

463: model does not display spontaneous  symmetry breakdown

464: due to chiral fluctuations.

465: We show below that

466: NLSM approach does not allow

467:  to prove that the chiral symmetry is

468: always restored  

469: by  fluctuations in the NJL model at $N_c=3$.

470: Below we also discuss difference  with the chiral GN model 

471: where  NLSM approach 

472: allows one to reach a similar conclusion at low $N$  \cite{gn1}.

473: 

474: 

475: The Lagrangian of the NJL model reads \cite{NJLM}:

476: %

477: \be

478: \calL=\psibar

479: i\dslash

480: \psi+\f{g_0}{2N_c}\left[

481: \left(

482: \psibar\psi

483: \right)^2+\left(

484: \psibar\ld{a}i\gd{5}\psi

485: \right)^2

486: \right].

487: \label{NJLModel}

488: \ee

489: %

490: The three $2\times2$-dimensional

491: matrices $ \ld{a}/2$, generate the fundamental representation

492: of flavor $SU(2)$, and are normalized by

493: $\tr (\ld{a}\ld{b})=2\delta_{ab}$.

494: One can introduce Hubbard - Stratonovich fields

495: $ \s$ and $\pi_a$:

496: %

497: \be

498: \calL=\psibar\left(

499: i\dslash-\s-i\gd{5}\ld{a}\pi_a

500: \right)

501: \psi-\f{\N}{2g_0}\left(

502: \s^2+\pi_a^2

503: \right).

504: \label{hsnjl}

505: \ee

506: %

507: After integrating out quark fields, following to

508: a standard mean-field variation procedure

509: % standard mean-field variation

510: %procedure gives the gap equation:

511: %\be

512: %\tr_{\gamma}\tr_f \left[ G(x,x) {1 \choose i\ld{a}\gd{5}} \right]

513: %= \f{1}{g_0}{\s(x) \choose \pi_a(x)}.

514: %\label{gapp}

515: %\ee

516: one can choose pseudoscalar solution $\pi_a$ to be vanishing

517: and scalar solution $\sigma\equiv M$ to be given by a gap equation:

518: % analogous to

519: %the gap equation in a BCS superconductor:

520: \be

521: \f{1}{g_0}=i(\tr_f 1)(\tr_{\gamma} 1)\int

522: \f{d^Dp}{(2\pi)^D}\f{1}{p^2-M^2}

523: \label{gap1}

524: \ee

525: The momentum integral is regularized by means of a cutoff $\Lambda$.

526: \comment{

527: Mean-field  $N_c\rightarrow \infty$

528: treatment gives an effective action

529: %

530: \be

531:  \Gamma (\rho)=- \Omega [\Delta v( \rho )+v_0]

532: \label{EffectivePotential}

533: \ee

534: %

535: where $\Omega$ is the spacetime volume, and $v_0$

536: is energy density of the symmetric state,

537: whereas

538: %

539: \beqn

540:  &&\Delta v(\rho)=\f{N_c}{2}\Bigg\{

541: \f{1}{g_0}\rho^2-\f{2}{(2\pi)^2}\Bigg[

542: \f{\rho^2\Lambda^2}{2}

543: +\f{\Lambda^4}{2}\ln\left(

544: 1+\f{\rho^2}{\Lambda^2}

545: \right)\nonumber\\

546: &&\mbox{}-\f{\rho^4}{2}\ln\left(

547: 1+\f{\Lambda^2}{\rho^2}

548: \right)

549: \Bigg]

550: \Bigg\}

551: \label{lambdapotential}

552: \eeqn

553: %

554: the mean-field condensation energy at {\it constant}

555: $  \sigma ^2+\pi_a^2\equiv  \rho ^2$.

556: The momentum integral is regularized by means of a cutoff

557: $\Lambda$.

558: The condensation energy is extremal

559: at

560: $ \rho = \SigM$ which solves the {\em gap equation\/}

561: \beqn

562:  \f{1}{g_0}&=&\f{2}{(2\pi)^2}

563: \left[

564: \Lambda^2- \SigM^2\ln\left(

565: 1+\f{\Lambda^2}{\SigM^2}

566: \right)

567: \right].

568: \label{lambdagap}

569: \eeqn    }

570: %

571: The constituent quark mass $\SigM$

572: in the limit $\N\rightarrow \infty$

573: is analogous to superconductive gap

574: in the BCS limit of the theory of superconductivity.

575: 

576: 

577: 

578: 

579: \comment{

580: In the paper \cite{kb} following to our

581: previous considerations of sigma-model approach

582: for description of the symmetry breakdown in 3D

583: superconductors and Gross-Neveu model \cite{sc} \cite{gn1},

584: it was suggested that in order to account for dynamic

585: chiral fluctuation in NJL model at zero temperature one should

586: set up 4D O(4) sigma-model.

587: % that will serve for nonperturbative description

588: %f the onset and disappearance of the

589: %phase coherence in the system with preformed gap

590: %modulus.

591: Authors of \cite{kb} came to conclusion

592: that resulting stiffness of the effective 4D O(4) sigma model

593: is too small  and

594: %for any values cutoff modulus of the gap function and

595: thus effective sigma model is

596: always in disordered phase due to strong dynamic

597: chiral fluctuations in the regime when $N_c=3$.

598: % and with it no symmetry

599: %breakdown occur in NJL model.

600: %These authors claims

601: %that their consideration invalidates finite temperature study

602: %NJL model never display spontaneous symmetry breakdown.

603: %Even though

604: %For instance accuracy of the critical stiffness was estimated

605: %by comparing with lattice simulation and results of the later

606: %depend on the type of the lattice

607: %We also know

608: %from the study of BCS - BEC crossover in

609: %superconductors that lattice sigma model

610: %estimates can differ significantly from the results of

611: %continuum theory

612: %\cite{det}.

613: %In this paper we show that claim that was

614: %

615: %this consideration at zero temperature shows

616: %importance of the dynamic chiral fluctuations

617: %but can not be considered as a proof of the

618: %lack of symmetry breakdown in this model since

619: %in contrast to superconductivity these

620: %calculations can not be verified perturbatively.

621: Let consider a regime of finite number of $N_c$.

622: Then fields start to perform fluctuations

623: around the extremal value $( \sigma ,\pi_a)=( M,0)$.

624: % pointed already in our previous paper on low-N behavior

625: %of chiral Gross-Neveu model \cite{gn}.

626: We can expand action in small

627: deviations from mean-field solution.}

628: %As long as $\N$ can be considered as a large number,

629: %the deviations

630: %$(\s',\pi'_a)\equiv(\s- \SigM,\pi_a)$

631: %are small, and the action can be expanded in powers

632: %of $(  \s',\pi'_a)$.

633: %The quadratic terms in this expansion

634: %define the propagators of the collective

635: %fields $( \s',\pi'_a)$.

636: %while higher expansion terms

637: %define the interactions.

638: At finite $N_c$ one can study fluctuations

639: around the saddle point solution.

640: The quadratic terms of expansion

641: around  the saddle point are:

642: %

643: %

644: \begin{eqnarray}

645: {\cal A}_0[\s',\pi'] = \f{1}{2}\!\int\!\!d^4q\!\left[

646: \left(\!\!

647: \begin{array}{c}

648: \pi'_a(q)\\

649: \s'(q)

650: \end{array}

651: \!\!\right)^T\!\!

652: \left(\!\!

653: \begin{array}{cc}

654: G_{\pi}^{-1}&0\\

655: 0&G_{\s}^{-1}

656: \end{array}

657: \!\!\right)\!\!

658: \left(\!\!

659: \begin{array}{c}

660: \pi'_a(-q)\\

661: \s'(-q)

662: \end{array}

663: \!\!\right)

664: \right],

665: \label{ao}\end{eqnarray}

666: %

667: %

668: where

669: $(\s',\pi'_a)\equiv(\s- \SigM,\pi_a)$

670: and $G_{\s,\pi}^{-1}$

671: are the inverse bosonic propagators.

672: \comment{

673: %

674: \begin{equation}

675: G_{\s}^{-1} =

676: \N\left[ 2\times2^{D/2} \int

677: \f{d^4p_E}{(2\pi)^4}\f{(p_E^2+p_Eq_E - M^2)}

678: {(p_E^2 + M^2)[(p_E+q_E)^2 +M^2]}

679: - \f{1}{g_0}\right];

680: G_{\pi}^{-1} =

681: \N\left[ 2\times2^{D/2} \int

682: \f{d^4p_E}{(2\pi)^4}\f{(p_E^2+p_Eq_E + M^2)}

683: {(p_E^2 + M^2)[(p_E+q_E)^2 +M^2]}

684: - \f{1}{g_0}\right].

685: %\label{@gre1}

686: \end{equation}

687: In the above expression one should introduce

688: a momentum cutoff $\Lambda_2$.} Implementing  a momentum

689: cutoff $\Lambda$, we can write

690: $G_{\pi,\sigma}^{-1}$ for small $q_E$ as:

691: %

692: \beqn

693: \!\!\!\!\!\!G_{\pi}^{-1}\!\approx\!\!-\f{\N}{(2\pi)^2}

694: \!\left[

695: \ln\left(

696: 1\!+\!\f{\Lambda^2}{\SigM^2}

697: \right)

698: \!-\!\f{\Lambda^2}{\Lambda^2\!+\!\SigM^2}\right]

699: \!q_E^2\!\equiv\!\!-Z(\SigM/\Lambda)q_E^2; \ \ \ \ \ \ \ \ \

700:  G_{ \sigma }^{-1}\!\approx\!\!

701: -Z(\SigM/\Lambda)(q_E^2+4\SigM^2).

702: \label{SigPropStiffLambda}

703: \eeqn

704: %

705: In analogy to $3D \ XY$-model approach to

706: strong-coupling superconductivity \cite{sc}

707: the authors of \cite{kb} introduced

708: a unit vector field

709: $n_i\equiv (n_0,n_a)\equiv(\s,\pi_a)/ \rho $

710: and set up an effective nonlinear sigma-model

711: %

712: \be

713: {\cal A}_0[n_i]= \f{\beta}{2}\int d^4x

714: [\partial n_i(x)]^2.

715: \label{@prop}

716: \ee

717: %

718: The prefactor $\beta=M^2 Z(M/\Lambda)$,

719:  that follows from (\ref{ao}),

720: (\ref{SigPropStiffLambda}) 

721: is playing the role of a stiffness

722: of the unit field fluctuations.

723: %

724: %\be

725: %\label{StiffModel}

726: %\ee

727: %

728: %

729: 

730: Now let us observe

731: that from the arguments given in

732: \cite{kb} it does not follow that NJL model necessarily

733: remains in a chirally symmetric phase

734: at $N_c=3$.

735: At first, in contrast to discussed in \cite{gn1}

736:  $2+\epsilon$-dimensional case,

737: one unfortunately, can not made any similar calculations in a

738: closed form in $3+1$-dimensions because this

739: theory is not renormalizable. 

740: %And stiffness coefficient

741: %should depend on two parameters, say $M$ and $\Lambda$.

742: %Whereas stiffness coefficient in renormalizable theory such 

743: %as superconductor can be expressed via one parameter. 

744: It was already observed in \cite{cut}-\cite{bub} that

745: cutoff of meson loops can not be set equal to cutoff

746: for quark loops and thus the $1/N_c$ corrected theory

747: \cite{bub} possesses two

748: independent  parameters that may be adjusted at will. 

749: We present another arguments 

750: of a different nature

751: rooted in a nonuniversality of a 

752: critical stiffness of a NLSM in four dimensions,

753: %codel in $3+1$-dimensions

754: %of more general nature, 

755: that does not allow 

756: in the  framework 

757: of the NLSM approach 

758: to reach the conclusion of \cite{kb}.

759: Our observation 

760: also applies to NLSM

761: description of precritical fluctuations in general systems. 

762: It also allows us to show that the discussed below additional

763: cutoff can not be related to the inverse coherence length 

764: of the radial fluctuations in the effective potential as it was suggested in 

765: \cite{prl,kb}.

766: 

767: Basically the  authors of \cite{kb} by deriving

768: $G_{\pi, \sigma}$ have extracted two 

769: characteristics from the initial system:

770: stiffness of the phase fluctuations in the degenerate minimum

771: of the effective potential

772: and the mass of the radial fluctuation. However knowledge of these

773: characteristics does not allow in principle to judge 

774: if directional fluctuations will destroy

775: long range order or system will possess  a BCS-like

776: phase transition.

777: The reason is that a critical stiffness of the nonlinear sigma

778: model is not an universal quantity in $3+1$-dimensions.

779: % and

780: %is expressed via an additional cutoff $\Lambda_3$

781: %in the gap  equation (\ref{CriticalStiff}). 

782: So in principle knowledge of the stiffness of NJL model 

783: is not sufficient for finding the

784: {\it position} of the phase transition in the effective

785: nonlinear sigma model.

786: The situation is just like in a Heisenberg

787: magnet where the critical temperature depends

788: on the stiffness along with lattice spacing and lattice

789: structure. Thus if one is given only a

790: stiffness coefficient one can not determine

791: temperature of the phase transition\footnote{Due to this

792: reason one can not refer to lattice simulation for 

793: finding the value of the critical stiffiness as it was done in \cite{kb,prl}

794: since implicitly these numerical values contain information 

795: of lattice structure and are not universal.}.

796: The situation is

797: in contrast to 2D case when a position of a

798: KT-transition can be deduced from the stiffness coefficient \cite{bkt}.

799: In two dimensions the critical stiffness of O(2) nonlinear 

800: sigma model is an universal quantity and is given by 

801: $\beta_{KT}=2/\pi$ \cite{bkt}, so comparing 

802: it with the stiffness coefficient 

803: derived from the initial theory 

804: (phase stiffness of the chiral GN model in $D=2$ is

805: $\beta = N/4\pi$ )

806: one can judge if the 

807: system has enough stiffen phase in order to 

808: preserve quasi-long range order 

809: as we have shown in \cite{gn1}

810: \footnote{There is a misleading 

811: statement about three dimensional case in \cite{gn1}.}. I.e. one can determine

812: the number of field components N 

813: %and renormalized

814: %couling strength

815:  that is needed to remain below the

816: position of Kosterliz-Thouless transition.

817: This is 

818: in contrast to discussed here  $D=3+1$ case.

819: %where given stiffness coefficient one can not make 

820: %such evaluations in general { \it in principle}.

821: 

822: %The authors of \cite{kb,prl}

823: %argued that on can 

824: %introduce in the $3+1  \ D$ theory an

825: %universal critical stiffness by relating the stiffness

826: %to the radial.

827: % Obviously 

828: %such a modified theory 

829: %has nothing to do with initial NJL model

830: %where critical stiffness is not universal.

831: %However we show below that these

832: %arguments are not consistent and do 

833: %not lead to a modified theory with an

834: %universal critical stiffness either.

835: 

836: 

837: Let us first recall a procedure how one 

838: can express a critical stiffness

839: of the O(4)-nonlinear sigma model via 

840: an additional parameter:

841: one can relax constrain $n_i^2 =1$

842: and introduce an extra integration over the

843: lagrange multiplier $\lambda$ rewriting (\ref{@prop}) as:

844: %

845: %\be

846: $(\beta/2) \int d^4x

847: \left\{ [\partial n_i(x)]^2+ \lambda \left[ n_i^2(x)-1\right] \right\}$.

848: %\ee

849: %

850: Integration out the $n_i(x)$-fields, yelds:

851: %

852: \be

853: {\cal A}_0[\lambda]=-\beta\int d^4x  \f{\lambda(x)}{2}+\f{N_n}{2}\Tr\ln\left[

854: -\p^2+\lambda(x)

855: \right],

856: \label{newaction}

857: \ee

858: %

859: where $N_n$ is the number of components of $n_i(x)$, and $\Tr$ denotes the

860: functional  trace.

861: This yields a gap equation:

862: %

863: \be

864: \beta=N_n\int \f{d^4k}{(2\pi)^4}\f{1}{k^2+\lambda} .

865: \label{@secge}\ee

866: %

867: The model has  a phase transition at a  critical stiffness

868: that depends on an unspecified  additional cutoff parameter that 

869: should be applied to the gap equation:

870: %

871: \be

872: \betacrit=N_n\int \f{d^4k}{(2\pi)^4}\f{1}{k^2}.

873: \label{CriticalStiff}

874: \ee

875: %

876: For example in the case of magnets the additional cutoff

877: needed in (\ref{CriticalStiff})

878: is naturally related to the lattice spacing.

879: %The authors of \cite{kb} proposed that 

880: %one can fix in the above equation a cutoff 

881: %according to physical considerations.

882: %derived from NJL model

883: %that one can 

884: %use in the above equation a cutoff 

885: %derived from NJL model

886: In the paper \cite{prl} it was proposed a

887: criterion that states that one can relate the inverse

888: coherence length extracted from  radial fluctuations in an

889: effective potential of an initial theory to the cutoff 

890: in the integral (\ref{CriticalStiff}) so that all the parameters

891: in the theory would be expressed from quantities

892: derived from an initial model and thus this 

893: modified model would possess an universal 

894: critical stiffens.

895: However there is no reason

896: for relating the cutoff needed in (\ref{CriticalStiff})

897: to the coherence length of the modulus fluctuations 

898:  and moreover we show that 

899: this procedure leads in general to unphysical consequences.

900: %it does not make the theory consistent due to the following 

901: %circumstance:

902: %Authors of \cite{kb} claim that "since pions in the symmetry

903: %broken

904: %phase are composite they are not "defined" over the 

905: %length scales much shorter than the inverse 

906: %binding energy of the pair wave function which is 

907: %equal to $2 M$ and thus they performed integral

908: %in (\ref{CriticalStiff}) up to cutoff $4M^2$".

909: %However

910: %it does not make the theory consistent since 

911: %according to arguments \cite{kb}

912: %pion fields are not "defined" over the 

913: %length scales much {\it shorter} than the inverse 

914: %binding energy and obviously this 

915: %determines according to arguments \cite{kb}

916: %only {\it upper} boundary of what would

917: %be an universal critical stiffness value.

918: %So the conclusion of absence of symmetry breakdown

919: %in such a modified theory was made by stating that 

920: %stiffness derived from the initial model is  

921: %{ \it smaller} than the {\it maximal

922: %possible} value of would be universal critical stiffness.

923: %Obviously there is no reason for 

924: %concluding from it that NJL model

925: %does not display symmetry breakdown

926: %and such a modified theory 

927: %has nothing to do with initial NJL model.

928: %Moreover

929: %we illustrate below that such a procedure 

930: It was supposed in \cite{prl} 

931: that relation of coherence length to cutoff

932: in the equation (\ref{CriticalStiff})

933: yields an universal criterion 

934: for judgement of 

935: nature of symmetry breakdown in 

936: general physical systems.

937: %leads to a number of unphysical consequences.

938: There is a simple counterexample:

939: in the case of  a

940: strong-coupling superconductor, the 

941:  characteristic nonlinear sigma model

942: that describes 

943: fluctuations in a degenerate valley 

944: of the effective potential is a 3D XY-model.  In

945: the continuous case it is  a free field

946: theory and has no phase transition at all.

947: The phase transition appears only in the lattice theory

948: and of course its temperature

949: depends on the lattice spacing.

950: % However

951: %choosing lattice spacing to be equal to $1/n_{bosons}^{1/3}$

952: %(this amounts to mapping a system of Cooper pairs

953: %to the system of hardcore composite bosons

954: %on the lattice, see previous sections and \cite{sc}) the resulting 

955: %lattice model surprisingly behaves very similar to initial

956: %continual 

957: %a phase diagram of the continual theory.

958: \comment{

959: In the case of NJL model there are however no length scales

960: that can be used to estimate position of the

961: phase transition of the effective nonlinear sigma model.

962: There was made an attempt

963: of finding such a scale in the paper \cite{kb},

964: namely it was suggested that

965: since the pion fields are composite, "they are not

966: defined over length scales much shorter

967: than the inverse binding energy of the pair wave function

968: which is equal to $2M$".

969: Following to this assumption the authors of \cite{kb}

970: performed the integral in Eq.~(\ref{CriticalStiff})

971: up to the cutoff $2M$ proposing it as an estimate

972: for the critical stiffness of the effective sigma model.

973: However, unfortunately, there is no reason to use this scale for

974: such estimate. In fact even if to consider following to \cite{kb}

975: that "pion fields are not defined over the

976: scales {\it shorter} than that"

977: it would be an estimate for upper boundary of the

978: value of the critical

979:  stiffness and thus one could not find from it

980: if the directional fluctuations in NJL can restore

981: chiral symmetry at low $N_c$.

982:  Moreover if

983: do not take it into the account and proceed

984: exactly along the same

985: lines as in \cite{kb} this construction

986: would lead to incorrect result of absence of superconductivity

987: in a strong-coupling superconductor too: 

988: }

989: With increasing coupling strength the 

990: low-temperature phase stiffness of the effective 3D XY model tends

991: to a plateau value

992: $J=n/4m$, where $n$ and $m$ are density and mass

993: of fermions \cite{sc}. Whereas the temperature of the phase transition

994: of the 3D XY-model is 

995: \be

996: T_c^{3D XY} \propto  \frac{n}{m} a,

997: \label{xy}

998: \ee

999:  where

1000: $a$ is the lattice spacing. 

1001: To be careful one should remark that accurate analysis shows

1002: that 

1003: a strong coupling superconductor possesses two characteristic length 

1004: scales: size of the Cooper pairs  that tends to zero with increasing 

1005: coupling strength 

1006: and a coherence length that tends to infinity with increasing coupling

1007: strength as the system evolves towards a weakly nonideal gas

1008: of true composite bosons \cite{R,pist}.

1009: At first if one relates the constant $a$ in (\ref{xy}) to the size of 

1010: the Cooper pairs 

1011: %n the strong coupling limit

1012: %ize of the Cooper pairs becomes smaller

1013: %ith increasing coupling strength 

1014: %\footnote{

1015: %Let us observe that in BCS superconductors

1016: %the Cooper pair size is determined by the

1017: %procedure equivalent to 

1018: %used in \cite{kb,prl} i.e. from the mass of the radial

1019: %fluctuations in the effective potential}

1020: following to the arguments of \cite{kb}

1021: one would come to an incorrect conclusion of absence of the

1022: superconductivity in strong-coupling superconductors

1023: in the way similar as the authors

1024: of \cite{kb} came to a conclusion 

1025: of inexistence of symmetry 

1026: breakdown

1027: in  the NJL model.

1028: This is in a direct

1029: contradiction with 

1030: behavior of the strong coupling superconductors 

1031: (see references menitioned in the Introduction).

1032: Second, if one attempts to relate $a$ in (\ref{xy}) 

1033: to the second length scale

1034: of the theory - namely true coherence length, that 

1035: tends to infinity with increasing coupling strength, 

1036: then one would come to a qualitatively incorrect conclusion too

1037: \cite{rem}. Thus the existence of an universal NLSM-based

1038: fluctuations criterion \cite{prl} appears to be incorrect.

1039: 

1040: So in general the nonlinear sigma model approach 

1041: for precritical fluctuations possesses an additional 

1042: fitting parameter which is the

1043: cutoff in the gap equation (\ref{CriticalStiff})

1044: that can not be related to inverse coherence length 

1045: extracted 

1046: from radial fluctuations in an effective potential.

1047: Thus within the NLSM approach  one can 

1048: not proove if the NJL model displays

1049: necessarily the directional fluctuations driven restoration of the chiral 

1050: symmetry at low $N_c$.

1051: %Let us for a moment do not take

1052: %into account above "no-go"

1053: %arguments and observe one more

1054: %internal controversy 

1055: %in the discussion \cite{kb}, namely the claim 

1056: %that "this approach invalidates studies of

1057: %temperature dependence of symmetry broken phase" .

1058: %Let us follow the lines of \cite{kb} and observe 

1059: %how the system will behave at finite temperatures

1060: %in the framework of the approach \cite{kb}.

1061: \section{Chiral fluctuations at finite temperature and a modified NJL

1062: model with a pseudogap}

1063: The authors of \cite{kb} employed NLSM 

1064: arguments in attempt to show that the NJL model can not serve 

1065: for the study of the chiral symmetry breakdown. We have shown 

1066: above that this conclusion appears to be incorrect

1067: since the critical stiffness

1068: in 3+1-dimensions is not an universal quantity and one 

1069: has an additional fitting parameter.

1070: % that should be chosen

1071: %from phenomenological considerations.

1072: This is an inborn feature of the discussed NLSM approach

1073:  in 3+1 dimensions

1074: % and does not depend

1075: %on whether the theory is renormalizable or not 

1076: (compare with the mentioned above

1077: cutoffs 

1078: discussions in nonrenormalizable models in a different approach

1079: \cite{cut}-\cite{bub}, and also \cite{bl}).

1080: The above circumstance allows one to fix the critical 

1081: stiffness from phenomenological considerations.

1082: However,  we argue below that, what is missed 

1083: in \cite{kb} is that, in principle, the low-$N_c$

1084: fluctuation instabilities, when properly treated, 

1085: have a clear physical meaning.

1086: Moreover we argue that 

1087:  one can employ a NLSM 

1088: for descriptions of the chiral fluctuations,

1089: providing that special care is taken of

1090: the additional cutoff parameter.

1091: It was indeed already discussed in literature that at finite temperatures

1092: the chiral phase transition should be accompanied by 

1093: developed  fluctuations (\cite{ht,ht2} and references therein).

1094: We argue that this 

1095: process at low $N_c$ should give  rise to a 

1096: phase analogous to the pseudogap phase that may be conveniently 

1097: described within a nonlinear sigma model approach. There are

1098: indeed other ways to describe these phenomena \cite{H}.

1099:  However the NLSM

1100: approach  seems to be especially convenient in the case

1101: of a nonrenormalizable theory. The descrition of 

1102: the two-step chiral phase transition and appearence of the 

1103: intermediate phase requires to study the system at the next-to-mean-field

1104: level. Unfortunately the NJL model is not renormalizable 

1105: and does not allow to make any conclusions about 

1106: importance of fluctuations in a closed form \cite{bub}.

1107: On the other hand a pseudogap phase 

1108: is a general feature of fermi system with attraction.

1109: The NLSM construction discussed below, due to

1110: its nonperturbative nature can not be 

1111: regarded as a regular approximation but may be 

1112: considered as a tractable modification of the NJL model that 

1113: has a pseudogap.

1114: One can also find an additional

1115: motivation for  employing these arguments in the fact that 

1116: NLSM allows one

1117: to prove an existence of a phase analogous to pseudogap phase

1118: in the chiral GN model \cite{gn1} which is the closest relative to NJL model.

1119: %but allows to show appearance of a pseudogap rigorously.

1120: Also NLSM approach works well for the description of 

1121: precritical fluctuations in superconductors \cite{sc} - where 

1122: the theory is renormalizable and 

1123: essentially the same results can be obtained perturbatively.

1124: We stress that

1125: these phenomena is a general feature 

1126: of any Fermi system with attraction.

1127: Also, to certain extend similar crossovers 

1128: are known in a large variety of condensed matter systems.

1129: In particular, besides  superconductors

1130: we can mention the

1131: exitonic condensate in

1132: semiconductors, the

1133: Josephson junction arrays,

1134: the itinerant and local-momentum theories

1135: of magnetism and  the ferroelectrics.

1136: 

1137: Let us now consider the chiral fluctuations in NJL model

1138: at finite temperature.

1139: Then

1140: % in the presence of heat resevoir

1141:  following standard dimensional reduction 

1142: arguments \cite{Wil}, the chiral fluctuations should be

1143:  described by a  $3D \ O(4)$-sigma model.

1144: Thus one  has 

1145: the following gap equation for the effective NLSM

1146: (i.e. finite temperature analogue of (\ref{@secge})):

1147: \be

1148: \frac{J_T}{T} = N_n \int \frac{d^3 k}{(2\pi)^3} \frac{1}{k^2+\lambda}

1149: \label{gt}

1150: \ee

1151: The temperature of the phase transition of the three dimensional

1152: classical $O(4)$ sigma

1153: model with stiffness $J_T$ is expressed via an additional parameter

1154: ${\tilde \Lambda}_T$ needed in (\ref{gt}) as :

1155: \be

1156: T_c =  \frac{\pi^2}{2}\frac{J_T}{{\tilde \Lambda}_T}

1157: \label{tc}

1158: \ee

1159: The stiffness of thermal fluctuations 

1160: $J_T$ can be readily extracted from the NJL model.

1161: At finite temperature the inverse bosonic propagator of the collective

1162: field $\pi$ for small $q$ can be written as:

1163: \begin{eqnarray}

1164: G^{-1}_\pi &= & -2^{D/2} N_c \int \frac{d^3p}{(2\pi)^3} \sum_n

1165: \left[ \frac{T}{(p^2+M^2+\omega_n^2)^2}\right] q^2 =

1166: %\ \ \ \ \ \ \ \ \ \

1167: %\ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \

1168: %\ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \  \  \ \ \ \ \ \ \ \ \ \ \

1169: \nonumber \\

1170: & - &  2^{D/2} N_c \int \frac{d^3 p}{ (2 \pi)^3}

1171: \left[ \frac{1}{8}

1172: \frac{1}{(p^2+M^2)^{3/2}}

1173: \tanh

1174: \left( \frac{\sqrt{p^2+M^2}}{2 T}\right)

1175: -\frac{1}{16 T}\frac{1}{p^2+M^2} \cosh^{-2}

1176: \left(

1177: \frac{\sqrt{p^2+M^2}}{2T}

1178: \right)

1179: \right] q^2 = \nonumber \\

1180: & - &  K (T,\Lambda_T, M, N_c) q^2,

1181: \label{st0}

1182: \end{eqnarray}

1183: where $\Lambda_T$ is a momentum cutoff. 

1184: The propagator (\ref{st0}) renders the 

1185: gradient term that allows one to set up an effective 

1186: classical $3D ~ O(4)$-nonlinear sigma model :

1187: \be

1188: E=\frac{J_T (T, \Lambda_T, M, N_c)}{2} \int d^3 x [\partial n_i (x)]^2,

1189: \label{Hei}

1190: \ee

1191: where 

1192: \be

1193: J_T(T, \LL_T, M, N_c) = K(T, \LL_T, M, N_c) ~M^2(T,\Lambda_T)

1194: \label{st1}

1195: \ee

1196: is the stiffness of the thermal  fluctuations in the degenerate 

1197: valley of the effective potential. The temperature -dependent quark

1198: mass $M$ that enters this expression 

1199: is given by a standard mean-field gap equation that 

1200: also should be regularized with the cutoff $\Lambda_T$:

1201: \be

1202: \f{1}{g_0}=2\times2^{D/2}\sum_n

1203: \int \f{d^3p}{(2\pi)^3}\f{T}{p^2+ M^2 +\omega_n^2}.

1204: \label{gap}

1205: \ee

1206: It can be easily seen that

1207: when we approach the temperature $T^*$

1208: where the mass the $M(T)$ becomes zero, the stiffness

1209: $J(T,\Lambda_T, M, N_c)$ also tends to zero.

1210: Formula (\ref{Hei}) defines a generalized Heisenberg model

1211: with a {\it temperature dependent stiffness coefficient}.

1212: Position of the phase disorder transition in a such 

1213: system should be determined self-consistently by

1214: solving the system of the equations for $T_c$ and $M(T_c)$.

1215: Apparently just like 

1216: in a superconductor

1217: with a pseudogap 

1218: the phase transition in a such system is a competition

1219: between a thermal depletion of the gap modulus (this 

1220: rudely corresponds to thermal pairbreaking in a superconductor)

1221: and a process of thermal excitations of 

1222: the directional  fluctuations in the

1223: degenerate minimum of the effective potential.

1224: "BCS" limit corresponds to the situation when $T^*$ merges with

1225: $T_c$ and it is easily seen that this scenario always holds true at

1226:  $N_c \rightarrow \infty$. I.e. at infinite $N$ the mean-field

1227: theory is always accurate just like BCS theory works well in the

1228: weak coupling superconductors. In the framework 

1229: of this NLSM construction, at low $N_c$ the scenario 

1230: of the phase transition

1231: depends on the  choice of $M(0), \Lambda_T $ 

1232: and ${\tilde \Lambda}_T$, that should be fixed from 

1233: phenomenological considerations.

1234: 

1235: %we introduced

1236: %following to \cite{kb}

1237: %a unit vector field

1238: %$n_i

1239: %\equiv (n_0,n_a)

1240: %\equiv(\s,\pi_a)/ M $.

1241: 

1242: 

1243: \section{Conclusion}

1244: %

1245: In the first part of this

1246: paper we presented a no-go result that

1247: within a framework of the nonlinear sigma model approach one

1248: can not answer the question if the chiral

1249: symmetry in NJL model is always restored by quantum

1250: fluctuations at  $N_c=3$. The reason is 

1251: the nonuniversality

1252: of the critical stiffness of $4 D \ O(4) $ nonlinear sigma model.

1253: This, along with the discussed above observations made in

1254: a framework of a different approach in \cite{bub}

1255: resolves numerous controversial discussions

1256: initiated by a recent paper \cite{kb}

1257: where it was argued that 

1258: there is no spontaneous breakdown of the 

1259: chiral symmetry in the NJL model,

1260: which appears to be incorrect.

1261: 

1262: In the second part of the paper we discussed a NLSM

1263: approach for precritical fluctuations 

1264: in a modified NJL model where a critical stiffness is expressed via

1265: an additional cutoff parameter that should be fixed 

1266: from phenomenological consideration. We discuss a

1267: formation in the above model 

1268: of  a phase analogous to the pseudogap phase

1269: in strong-coupling and low-carrier density superconductors.

1270: Appearence of this phase  

1271: may accompany the chiral phase transition in  QCD.

1272: Since the precursor pairing fluctuations

1273: is a general feature of any Fermi system with attraction

1274: and moreover it is a dominating region of a phase

1275: diagram

1276: of strong-coupling and low carrier density superconductors,

1277: the interesting question is to what extend the discussed 

1278: above phase

1279: is developed in QCD and color superconductors.

1280: \comment{

1281: Indication of possible importance of the pseudogap

1282: concept in particle physics is the mentioned above existence of this

1283: phenomenon in the chiral Gross-Neveu model at low $N$.

1284: Even though these results can not be directly generalized to

1285: NJL model, one can guess that in analogy

1286: to 3D XY-model approach to strong-coupling

1287: and low carrier density superconductivity, one

1288: can set up a nonlinear 3D O(4)-sigma

1289: model with temperature depended stiffness

1290: coefficient as a toy model for QCD at finite temperatures

1291: that would possess two characteristic temperatures

1292: corresponding to discussed in this paper $T_c$ and $T^*$.

1293: Speaking about the BCS-BEC crossover,

1294: precritical fluctuations and the pseudogap phase,

1295: it should be noted as well that

1296: in some sense similar phenomena are known

1297: in a large variety of condensed matter systems,

1298: in particular, except for superconductors

1299: we can mention itinerant and local-momentum theories

1300: of magnetism, exitonic condensate in

1301: semiconductors, ferroelectrics and Josephson junction arrays.}

1302: 

1303: %We also observed that due to nonuniversality of the 

1304: %critical stiffness the is {\it no} universal criterion 

1305: %for judgment of the nature 

1306: %of symmetry restoration in 

1307: %general systems as it was proposed in \cite{prl}

1308: \begin{acknowledgments}

1309: The author is grateful to Prof. A. J. Niemi and Dr. V.V. Cheianov

1310: for discussions, to Prof. T. Hatsuda and Prof. D. Blaschke 

1311: for communicating useful references.

1312: \end{acknowledgments}

1313: \begin{thebibliography}{999}

1314: \bibitem{NJLM}

1315:  Y. Nambu and G. Jona Lasinio, Phys.~Rev.~{\bf 122}, 345 (1961);

1316:   {\bf 124}, 246 (1961).

1317: 

1318: 

1319: \bibitem{vdb}

1320: S. P. Klevansky,

1321: Rev. Mod. Phys. {\bf 64}, 649 (1992).

1322: 

1323: \bibitem{ht}

1324: T. Hatsuda and T. Kunihiro,

1325: Phys. Rep. {\bf 247}, 221 (1994).

1326: 

1327: \bibitem{GNM}

1328: {D.~Gross} and {A.~Neveu}, Phys.~Rev.~D {\bf 10}, 3235 (1974)

1329: 

1330: 

1331: \bibitem{gn1}

1332: H.Kleinert and E.Babaev  preprint hep-th/9809112

1333: Phys.Let. B. 438, 311 (1998)

1334: 

1335: \bibitem{kb} H. Kleinert and B. Van den Bossche, Phys. Lett. B 474

1336: 336 (2000); 

1337: 

1338: \bibitem{Le}

1339: A. J. Leggett, in { \em Modern Trends in the Theory of Condensed

1340: Matter}, ed. by A. Pekalski and J. Przystawa, Lecture

1341: Notes in Physics, Vol. 115 (Springer-Verlag, Berlin, 1980),p. 13.

1342: 

1343: \bibitem{N}P. Nozieres and S. Schmitt-Rink,

1344:  J. Low. Temp. Phys. {\bf 59},195 (1985).

1345: 

1346: \bibitem{sc}

1347: E.Babaev and H.Kleinert  Phys. Rev. B 59 12083 (1999)

1348: ( preprint cond-mat/9907138)

1349: 

1350: 

1351: \bibitem{R} C. A. R. S\'a de Melo, M. Randeria, and J.R. Engelbrecht,

1352: Phys. Rev. Lett. {\bf 71}, 3202 (1993).

1353: 

1354: \bibitem{H} 

1355: V.J.Emery, S.A. Kivelson, Nature {\bf 374} 434 (1995);

1356: Phys.Rev.Lett {\bf 74}  3253 (1995).\\

1357: R. Haussmann, Z. Phys. B {\bf 91}, 291 (1993);

1358: Phys. Rev. B {\bf 49}, 12975 (1994).\\

1359: O. Tchernyshyov, Phys. Rev. B {\bf 56}, 3372 (1997).\\

1360: M. Drechsler, W. Zwerger.

1361: Ann.~Phys.(Germany), {\bf 1}, 15 (1992). \\

1362: V.P. Gusynin, V.M. Loktev, S.G. Sharapov

1363: JETP Lett. { \bf 65} 182 (1997); JETP {\bf 88} 685 (1999)\\

1364: S. Stintzing, W. Zwerger. Phys. Rev. { \bf B 56}, 9004 (1997) \\

1365: J.R. Engelbrecht, M. Randeria, C. A. R. S\'a de Melo,

1366: Phys. Rev. B {\bf 55}, 15153 (1997). \\

1367: M. Randeria. in: Bose-Einstein Condensation,

1368: edited by A. Griffin, D. W. Snoke., and S. Stringary.

1369: New York, Cambridge University Press, 1995. p.355-392.

1370: M. Randeria preprint cond-mat/9710223\\

1371: A. Schnell, G. Roepke, P. Schuck Phys.Rev.Lett. {\bf 83} 1926 (1999) 

1372: 

1373: \bibitem{pist} F. Pistolesi and G. C. Strinati,

1374: Phys. Rev. B {\bf 49}, 6356 (1994); Phys. Rev B {\bf 53}, 15168 (1996)

1375: 

1376: \bibitem{asl}

1377: L. G. Aslamazov and A. I. Larkin Sov. Phys. Solid State {\bf 10} 875 (1968) \\

1378: %Sov. Phys. Solid State. (1968)  \\

1379: K. Maki ~Progr. Teor. Phys.

1380: {\bf 39} 879 (1968); 

1381: {\bf 40} 193 (1968); \\ R.S. Thompson Phys. Rev.  {\bf B 1} 327 (1970)

1382: %See also A.V. Svidzinskii "Prostranstvenno-neodnorodnye

1383: %zadachi teorii sverhprovodimosti".Moskva.Nauka (1982).

1384: %(in russian).

1385: 

1386: 

1387: \bibitem{ht2} T. Hatsuda and T. Kunihiro Phys. Rev. Lett.{ \bf 55}

1388: 158   (1985); see also S. Chiku and T. Hatsuda   Phys. Rev. {\bf D57}

1389: 6 (1998); T. Hatsuda, T. Kunihiro and H. Shimizu Phys. Rev. Lett.

1390: { \bf 82} 2840 (1999)

1391: 

1392: 

1393: %\bibitem{ranrev} M. Randeria preprint cond-mat/9710223

1394: %\bibitem{W} E.~Witten, Nucl.~Phys.~B {\bf 145}, 110 (1978).

1395: 

1396: 

1397: %\bibitem{GFCM}

1398: %H.~Kleinert, Gauge Fields in Condensed Matter. World Scientific, 1989

1399: 

1400: 

1401: %\bibitem{kt}

1402: %V. L. Berezinskii.

1403: % Zh.~Eksp.~Teor.~Fiz., 1970, {\bf 59},  907;

1404: %J. Kosterlitz, D. Thouless. J.~Phys. C,  {\bf 6} 1181 (1973).

1405: 

1406: \bibitem{cut}

1407: V. Dmitrasinovic, H.-J. Schulze, R. Tegen and R.H. Lemmer, Ann. Phys. (NY)

1408: {\bf 238} 332 (1995);

1409: 

1410: \bibitem{cut2}

1411: E. Nikolov, W. Broniowski, C. Christov, G. Ripka and K. Goeke, Nucl.

1412: Phys. A {\bf 608} 411 (1996);

1413: 

1414: \bibitem{bub}

1415: M. Oertel, M. Buballa, J. Wambach  Phys.Lett. {\bf B477} 77 (2000)

1416: 

1417: \bibitem{bkt}

1418: V. L. Berezinskii.

1419:  Zh.~Eksp.~Teor.~Fiz. { \bf 59} 907  (1970) \\

1420: J. Kosterlitz, D. Thouless. J.~Phys.,  {\bf C6}, 1181 (1973).

1421: 

1422: \bibitem{prl} H. Kleinert Phys. Rev. Lett. {\bf 84}, 286 (2000)

1423: 

1424: \bibitem{rem}

1425: In fact, in a nonperturbative 

1426: NLSM approach to the  condensation from 

1427: pseudogap phase in strong-coupling superconductors one can reach 

1428: an excellent agreement with the perturbative results \cite{N,R}

1429: if one considers an effective sigma model on the lattice 

1430: with the spacing given by $a \propto 1/n_{b}^{1/3}$,

1431: where $n_b$ is  density of Cooper pairs \cite{sc}. Then 

1432: the model is related to the condensation of 

1433: hard-core composite bosons on a lattice.

1434: % and it reproduces

1435: %surprisingly well the phase diagram of the continual theory \cite{sc}.

1436: 

1437: \bibitem{bl}

1438: D. Blaschke, Yu.L. Kalinovsky, G. Roepke, S. Schmidt, M.K. Volkov

1439: ~ Phys. Rev. {\bf C 53}  2394 (1996);

1440: 

1441: \bibitem{Wil}

1442: R. Pisarski, F.Wilczek Phys. Rev.  {\bf D 29} 338 (1983); \\

1443: F. Wilczek Int. J. Mod. Phys.{ \bf A 7 } 3911 (1992); \\

1444: K. Radjagopal, F. Wilczek Nucl. Phys {\bf D 404}   577 (1993)

1445: 

1446: \end{thebibliography}

1447: \end{document}

1448: %1