1: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

2: %%%%%%   template.tex for PTPTeX.cls <ver.0.88>  %%%%%

3: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

4: %\documentclass[seceq]{ptptex}

5: %\documentclass[letter]{ptptex}

6: \documentclass[seceq,supplement]{ptptex}

7: %\documentclass[seceq,addenda]{ptptex}

8: %\documentclass[seceq,errata]{ptptex}

9: %\documentclass[seceq,preprint]{ptptex}

10: 

11: \usepackage{graphicx}

12: \usepackage{wrapft}

13: %\usepackage{doublespace}

14: 

15: %%%%% Personal Macros %%%%%%%%%%%%%%%%%%%

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

17: \newcommand{\ee}{\end{equation}\noindent}

18: \newcommand{\bea}{\begin{eqnarray}}

19: \newcommand{\eea}{\end{eqnarray}}

20: \newcommand{\Tr}{\mbox{Tr}}

21: \newcommand{\mD}{\mathcal{D}}

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

23: \newcommand{\lldots}{\ldots\ldots\ldots\ldots\ldots\ldots}

24: \newcommand{\1}{{\bf 1}}

25: \newcommand{\lapprox}{\raisebox{-0.5ex}{$\ 

26: \stackrel{\textstyle<}{\textstyle\sim}\ $}}

27: % rightarrow with subscripts

28: \newcommand{\maprightb}[1]{\smash{\mathop{

29: \hbox to 1cm{\rightarrowfill}}\limits_{#1}}}

30: %

31: 

32: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

33: 

34: %\pubinfo{Vol.~10X, No.~X, Mmmmm YYYY}%Editorial Office will fill in this.

35: %\setcounter{page}{}                  %Editorial Office will fill in this.

36: %\def\ptype{p}                        %Editorial Office will fill in this.

37: %\def\ptpsubject{}                    %Editorial Office will fill in this.

38: %\def\pageinfo{X-X}                   %Editorial Office will fill in this.

39: %-------------------------------------------------------------------------

40: %\nofigureboxrule                     %to eliminate the rule of \figurebox

41: %\notypesetlogo                       %comment in if to eliminate PTPTeX 

42: %---- When [preprint] you can put preprint number at top right corner.

43: %\preprintnumber[3cm]{%<-- [..]: optional width of preprint # column.

44: %KUNS-1325\\PTPTeX ver.0.8\\ August, 1997}

45: %-------------------------------------------------------------------------

46: 

47: \markboth{%     %running head for odd-page (authors' name)

48: Costas G. Strouthos

49: }{%             %running head for even-page (`short' title)

50: %Instruction for Making \LaTeX\ Compuscripts Using \protect\PTPTeX

51: }

52: 

53: 

54: 

55: \title{%        %You can use \\ for explicit line-break

56: Mesons at finite baryon density in $(2+1)d$

57: %\footnote{based on talks at Workshop ``Thermal Field Theory'',

58: %Yukawa Institute, Aug. 9. 2002 and Symposium ``Towards

59: %understanding of finite density systems'', JPS meeting, Sept. 14, 2002}%

60: }

61: 

62: \author{%       %Use \scshape 

63: Costas G. \textsc{Strouthos} 

64: }

65: 

66: \inst{%         %Affiliation, neglected when [addenda] or [errata]

67: \it Department of Physics, Duke University,  Durham, NC 27708-0305, USA\\

68: }

69: 

70: %\publishedin{%         %Write this ONLY in cases of addenda and errata

71: %Prog.~Theor.~Phys.\ \textbf{XX} (19YY), page.}

72: 

73: %\recdate{Mmmmm DD, YYYY}%            %Editorial Office will fill in this.

74: 

75: \abst{

76: We discuss the critical properies of the three-dimensional Gross-Neveu model 

77: at nonzero temperature and nonzero chemical potential. We also present 

78: numerical and analytical results for the in-medium interaction due to scalar

79: meson exchange. Further, we discuss in-medium modifications of mesonic

80: dispersion relations and wavefunctions. 

81: }

82: 

83: \begin{document}

84: 

85: \maketitle

86: 

87: \section{Introduction}\label{sec-intro}

88: Chiral phase transitions and the spectrum of excitations in 

89: strongly interacting matter at finite baryon chemical potential 

90: remain interesting challenges. 

91: Strongly interacting systems are intrinsically non-perturbative

92: and therefore most of our knowledge about the relevant phenomena comes from 

93: lattice simulations. Unfortunately, the complex nature of the determinant 

94: of the QCD Dirac operator at finite chemical potential makes it impossible to use

95: standard simulation techniques to study Fermi surface phenomena in Euclidean simulations. 

96: 

97: In order to understand what ingredients might play a decisive role

98: in more complex systems such as gauge theories, we have studied the

99: simplest non-trivial model simulable with $\mu \neq 0$ using standard algorithms, 

100: namely the three-dimensional Gross-Neveu model (GNM$_3$). 

101: Its Lagrangian in Euclidean metric is written in terms of 

102:  $4N_f$-component spinors $\psi,

103: \bar\psi$ as 

104: \begin{equation}

105: {\cal L}=\bar\psi(\partial{\!\!\!/\,}+m)\psi-{g^2\over{2N_f}}(\bar\psi\psi)^2.

106: \end{equation}

107: In the chiral limit $m=0$ the model has a global Z$_2$ symmetry

108: $\psi\mapsto\gamma_5\psi$, $\bar\psi\mapsto-\bar\psi\gamma_5$.

109: At tree level, the field $\sigma$ has no dynamics; it is trully an

110: auxiliary field. However, it acquires dynamical content by dint of quantum 

111: effects arising from integrating out the fermions. The model is renormalizable in 

112: the $1/N_f$ expansion unlike in the loop expansion \cite{rosen91}. Apart from the obvious 

113: numerical advantages of working with a relatively simple model there are several 

114: other motivations for studying this model.

115: At $T=\mu=0$ for sufficiently strong coupling $g^2$, chiral symmetry is

116: spontaneously broken by a condensate $\langle\bar\psi\psi\rangle\not=0$ leading 

117: to a dynamically generated

118: fermion

119: mass gap given by $M_f=g^2\langle\bar\psi\psi\rangle\gg m$ in the large-$N_f$ approximation.

120: The spectrum of excitations contains both baryons and mesons, i.e. the elementary fermions

121: $f$ and the composite $f\bar{f}$ states. The critical coupling

122: $g_c^2$ at which the gap $M_f/\Lambda_{UV}\to0$,

123: defines an ultra-violet stable fixed point of

124: the renormalisation group at which an interacting continuum limit may be taken.

125: This picture has been verified at next-to-leading order in $1/N_f$

126: and by Monte

127: Carlo simulations with finite $N_f$ \cite{hands93}. 

128: 

129: In the next section we will discuss the issue of chiral symmetry restoration at nonzero 

130: fermion chemical potential. 

131: In section \ref{sec:section3} we will discuss Fermi surface phenomena, such as long range interparticle

132: interaction in medium. We will also present results for dispersion relations (section \ref{sec:section4} 

133: and wavefunctions (section \ref{sec:section5}) of

134: massless particle-hole excitations on or near the Fermi 

135: surface. A study of the dispersion relation of spin-$\frac{1}{2}$ excitations  around the 

136: Fermi surface has shown that there is no mass gap characteristic of a BCS instability \cite{hands03} 

137: and the values of the Fermi liquid parameters (Fermi momentum, Fermi velocity) are in good

138: agreement with the $1/N_f$ predictions. 

139: However, in a different study of a related model, the $(2+1)d$ NJL model, which has an

140: $SU(2)\otimes SU(2)$ chiral symmetry, it was shown that although the BCS instability is also absent 

141: the properies 

142: of the quasi-particle spectrum is determined by physics outside the scope of the 

143: $1/N_f$ expansion \cite{hands02}. 

144: 

145: \section{Chiral phase transitions}

146: The action of the NJL model remains real even after the introduction

147: of nonzero chemical potential $\mu$, which means we can study the

148: physics of the high density regime using standard Monte Carlo

149: techniques.

150: In the presence of a Fermi surface with Fermi momentum $p_F$

151: the creation of $f\bar{f}$ pairs

152: with zero net momentum is energetically suppressed, because one can only create

153: particles with $p>p_F$. So as the fermion number density $\eta(\mu)$ grows

154: the chiral symmetry breaking is suppressed.

155: The large-$N_f$ description of the $\mu \neq 0$ chiral phase transition

156: predicts a first order transition for $T=0$ and a continuous transition for

157: $T>0$ \cite{klimenko}. Furthermore, the critical value of the chemical potential $\mu_c$

158: is equal to the value of the fermion mass at $\mu=0$, which indicates that

159: materialization of the fermion itself drives the symmetry restoration

160: transition.

161: Interactions as expected decrease $\mu_c$ below the mean field result \cite{hands95}.

162: Work by Stephanov \cite{stephanov96} suggests that any non-zero density simulation which

163: incorporates a real path integral measure proportional to $\det(MM^{\dagger})$

164: is doomed to failure due to the formation of a light baryonic pion from a quark $q$

165: and a conjugate quark $q^c$.

166: The NJL model however, does not exhibit such a pathology, because the realization of the

167: Goldstone mechanism in this model is fundamentally different from that in QCD.

168: In the NJL model the Goldstone mechanism is realized by a pseudoscalar

169: channel pole formed from disconnected diagrams and the connected diagrams

170: yield a bound state of mass $\approx 2 m_f$ \cite{hands99}. This implies the absence

171: of a light $qq^c$ state.

172: 

173: As expected our simulations of GNM$_3$

174: with $N_f=4$ \cite{strouthos01} did not provide any evidence for the

175: existence of a nuclear liquid-gas transition at $\mu<\mu_c$.

176: It was shown in \cite{buballa} that in the standard four-fermi models there is no saturation density

177: for stable matter. In order to get the saturation features the authors of

178: \cite{buballa} introduced a chirally invariant scalar-vector interaction term which

179: cures the binding problem.

180: Furthermore, our results showed that the second order nature of the

181: $T \neq 0$, $\mu=0$ transition remains stable down to low $T$ and large

182: $\mu$.

183: \begin{figure}[hbt]

184: \begin{center}

185: \begin{minipage}{ 0.48\linewidth}

186: \includegraphics[width=1.0 \linewidth]{./sigma.eps}

187: \end{minipage}

188: \hspace{2mm}

189: \begin{minipage}{ 0.48\linewidth}

190: \includegraphics[width=1.0 \linewidth]{./density.eps}

191: \end{minipage}

192: \end{center}

193: \caption{

194: Chiral order parameter (left) and fermion number density (right) vs. chemical potential 

195: at various values of $T$.

196: }

197: \label{fig:1}

198: \end{figure}

199: In Fig.\ref{fig:1} (left) we plot the normalized order parameter $\Sigma(T,\mu)/\Sigma(T,0)$

200: ($\Sigma\equiv\langle\sigma\rangle$)

201: as a function of $\mu/m_f$ at different values of $T$. It is clear from the shapes of these

202: curves that the transition becomes sharper as we decrease the temperature. In Fig.\ref{fig:1}

203: (right)

204: we plot the normalized fermion number density as a function of the chemical potential

205: at different values of $T$. In the limit $T \rightarrow 0$ we see that the fermion

206: density is strongly suppressed before the transition and then jumps discontinuously.

207: By performing detailed finite size scaling analysis

208: which allowed us to distinguish between second order and weak first order transitions

209: we showed that the tricritical point lies on the section

210: of the phase boundary defined by $T/T_c \leq0.23$, $\mu/\mu_c \geq0.97$ \cite{strouthos01}.

211: It was also shown some time ago \cite{kogut98} that the second order transitions 

212: belong to the $2d$ Ising universality class in accordance with the dimensional 

213: reduction scenario.

214: 

215: \section{The sigma propagator in medium}

216: \label{sec:section3}

217: In this section we examine the auxiliary scalar propagator $D_{\sigma}$ in the

218: presence of non-zero baryon density.

219: Although at tree level $\sigma$ is non-propagating, the leading order

220: $1/N_f$ expansion at $\mu=0$ predicts that it acquires dynamics through

221: quantum corrections due to virtual $q\bar{q}$ pairs, resulting in a propagator

222: of the form \cite{rosen91, hands93}

223: \begin{equation}

224: D_\sigma(k^2)={1\over

225: g^2}{{(4\pi)^{d\over2}}\over{2\Gamma(2-{\textstyle{d\over2}})}}

226: {M_f^{4-d}\over{(k^2+4M_f^2)F(1,2-{\textstyle{d\over2}};{3\over2};

227: -{k^2\over{4M_f^2}})}}.

228: \label{eq:Dsigma}

229: \end{equation}

230: Immediately we see the difference between this model

231: and QCD.

232: For $k^2\ll M_f^2$ the hypergeometric function $F\approx1$, implying that

233: to this order the $\sigma$ resembles a weakly-bound meson of mass

234: $M_\sigma=2M_f$; however,

235: the hypergeometric function

236: in the denominator indicates a strongly interacting $q\bar q$

237: continuum immediately above the threshold $2M_f$.

238: This implies that if truly bound, its binding energy is

239: $O(1/N_f)$ at best

240: (to our knowledge there have so far been no analytic calculations), implying

241: little if any separation between pole and threshold. A recent study using the maximum entropy 

242: method has shown evidence for a nonzero binding energy in the $\sigma$ channel for 

243: finite $N_f=4$ \cite{allton02}.

244: Since all residual interactions

245: are subleading in $1/N_f$, we surmise that all other mesons

246: are similarly weakly bound states of massive fermions, and hence effectively

247: described by a two-dimensional ``non-relativistic quark model''.

248: 

249: Similarly, at nonzero chemical potential $\mu$ to leading order in the $1/N_f$

250: expansion, we have 

251: \begin{eqnarray}

252: D_\sigma^{-1}(k)&=&1-\Pi(k;\mu)\nonumber\\

253: &=&g^2\left[{1\over\Sigma_0}\int_{q,\mu=0}

254: \mbox{tr}{1\over{iq{\!\!\! /\,}+\Sigma_0}}

255: +\int_{q,\mu>\mu_c}\mbox{tr}{1\over{iq{\!\!\! /\,}+\mu\gamma_0}}

256: {1\over{i(q{\!\!\!/\,}-k{\!\!\!/\,})+\mu\gamma_0}}

257: \right],

258: \end{eqnarray}

259: where $\Pi$ is the virtual fermion-antifermion vacuum polarization bubble.

260: We have used the gap equation at $\mu=0$ to express $1/g^2$ in terms of the zero

261: density gap $\Sigma_0$ and assumed that the gap vanished in the

262: integral defining $\Pi$ for $\mu>\mu_c$.

263: The functional form of the propagator in the Hard Dense Loop (HDL) approximation

264: ($k_0, |\vec{k}| \ll \mu$) in $d+1$ dimensions is given by \cite{hands03} 

265: \begin{eqnarray}

266: D_\sigma^{-1}(k_0,\vec k)=

267: {{g^2\mu^{d-3}}\over{(4\pi)^{d\over2}\Gamma(\textstyle{d\over2})(3-d)}}

268: \Biggl[4{{(3-d)}\over{(d-1)}}\mu^{3-d}(\mu^{d-1}-\mu_c^{d-1})+\;\;\;\;\;\;\;

269: \;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\;\label{eq_Dsigma}\\

270: k_0^2+\vert\vec k\vert^2

271: \biggl(1-{{(3-d)(k_0^2+\vert\vec k\vert^2)^{1\over2}}\over

272: {k_0+(k_0^2+\vert\vec

273: k\vert^2)^{1\over2}}}\biggr)

274: +i{{k_0\vert\vec k\vert^2}\over{4\mu}}

275: \biggl(1+{{\vert\vec k\vert^2}\over

276: {[k_0+(k_0^2+\vert\vec

277: k\vert^2)^{1\over2}]^2}}\biggr)

278: +O\Bigl({k^4\over\mu^2}\Bigr)\Biggr].\nonumber

279: \label{htl}

280: \end{eqnarray}

281: An interesting observation is that in the static limit $k_0=0$

282: the momentum dependence of eq.~(\ref{htl}) vanished for the value $d=2$

283: implying complete screening for the static potential due to $\sigma$

284: exchange. Furthermore, for $\vec{k}=\vec{0}$ $D_{\sigma}$ is proportional

285: to a conventional zero-momentum boson propagator with mass

286: \begin{equation}

287: M_\sigma=2\sqrt{\mu(\mu-\mu_c)},

288: \label{eq:msigma}

289: \end{equation}

290: implying that just above the transition there is a tightly bound state.

291: In Fig.~\ref{fig:2} we show the results for the sigma correlator obtained 

292: from numerical simulations with $N_f=4$ and $1/g^2=0.75$ on 

293: $48 \times 32^2$ lattices.  The value of the critical chemical potential $\mu_c=0.17(1)$. 

294: \smallskip

295: \bigskip

296: \begin{figure}[]

297: \begin{center}

298: \begin{minipage}{ 0.43\linewidth}

299: \includegraphics[width=1.0 \linewidth]{./sigma_corr_mon.eps}

300: \end{minipage}

301: \hspace{2mm}

302: \begin{minipage}{ 0.43\linewidth}

303: \includegraphics[width=1.0 \linewidth]{./msigma_mon.eps}

304: \end{minipage}

305: \end{center}

306: \caption{

307: The zero momentum $\sigma$ timeslice correlator for 4 values of $\mu$ (left)

308: and the $M_{\sigma}$ as a function of $\mu$ (right); the line is the leading order 

309: $1/N_f$ prediction.

310: }

311: \label{fig:2}

312: \end{figure}

313: The numerical results for $M_{\sigma}$ fall $20-30\%$ lower than the leading order 

314: prediction, which is consistent with a correction of $O(N_f^{-1})$.

315: 

316: \section{Mesonic dispersion relations}

317: \label{sec:section4}

318: In this section we investigate mesonic states by measuring the connected contribution 

319: to the correlation functions 

320: ${\cal C}_\Gamma(\vec k,x_0)=\sum_{\vec x}\langle j_\Gamma(\vec 0,0)

321: j_\Gamma^\dagger(\vec

322: x,x_0)\rangle e^{-i\vec k.\vec x}$ where the bilinears $j_\Gamma(x)$

323: are defined with

324: scalar, pseudoscalar or vector quantum numbers.

325: In terms of staggered fermion fields $\chi,\bar\chi$ the operators are

326: \begin{equation}

327: j_{\bf 1}(x)=\bar\chi_x\chi_x;\;\; 

328: j_{\gamma_5}(x)=\varepsilon_x\bar\chi_x\chi_x;\;\;

329: j_{\gamma_i}(x)={\eta_{ix}\over2}[\bar\chi_x\chi_{x+\hat\imath}+

330:                           \bar\chi_{x+\hat\imath}\chi_x],

331: \end{equation}

332: where $\eta_{1x}=(-1)^{x_0}$, $\eta_{2x}=(-1)^{x_0+x_1}$ and

333: $\varepsilon_x=(-1)^{x_0+x_1+x_2}$. As before we study the timeslice

334: correlators in each channel as a function of spatial momentum $\vec

335: k\parallel\hat x$.

336: \bigskip

337: \begin{figure}[hbt]

338: \begin{center}

339: \begin{minipage}{ 0.53\linewidth}

340: \includegraphics[width=1.2 \linewidth]{./psdisperse_mon.eps}

341: \end{minipage}

342: \end{center}

343: \caption{

344: Pseudoscalar correlator ${\cal C}_{\gamma_5}(\vert\vec k\vert,x_0)$

345: at 4 different $\mu$

346: for momenta $\vert\vec k\vert=0$ (filled circles),

347: ${\pi\over16}$ (filled squares), ${\pi\over8}$ (filled diamonds),

348: ${3\pi\over16}$ (filled up triangles), ${\pi\over4}$ (filled down triangles),

349: ${5\pi\over16}$ (open circles), ${3\pi\over8}$ (open squares), ${7\pi\over16}$

350: (open diamonds) and  ${\pi\over2}$ (open up triangles).

351: }

352: \label{fig:psdisperse}

353: \end{figure}

354: It was shown \cite{hands03} that in free field theory for $\mu>\mu_c$ the the $x_0 \rightarrow 0$ 

355: limit of ${\cal C}_\Gamma(\vec k,x_0)$

356: is dominated by a continuum of particle-hole pairs at or near the Fermi surface, which effectively 

357: cost zero energy to excite. 

358: The generic result is that for

359: $\vert\vec k\vert\leq2\mu$ the decay is algebraic, with

360: \begin{equation}

361: {\cal C}_\Gamma(\vec k,x_0)\propto \left\{

362: \begin{array}{ll}

363: x_0^{-2}         & \mbox{$\vert\vec k\vert\ll\mu$} \\

364: x_0^{-{3\over2}} & \mbox{$\vert\vec k\vert\simeq2\mu$}.

365: \end{array}

366: \right.  

367: \end{equation}

368: Only once $\vert\vec k\vert>2\mu$ does it become kinematically impossible to

369: excite a pair with zero energy, resulting in exponential decay:

370: \begin{equation}

371: {\cal C}_\Gamma(\vec k,x_0)\propto x_0^{-{3\over2}}\exp\Bigl(-(\vert\vec

372: k\vert-2\mu)x_0\Bigr).

373: \end{equation}

374: The sequence of plots in Fig.~\ref{fig:psdisperse} is in qualitative agreement

375: with these findings. For each $\mu$ there is a particular value of

376: $\vert\vec k\vert$, highlighted with a solid line in the plots, for which the

377: temporal falloff is particularly slow, corresponding to

378: $\vert\vec k\vert\approx2\mu$ (e.g. for $\mu=0.4$ the slow falloff

379: occurs for $\vert\vec k\vert={\pi\over4}\simeq0.785$).

380: For $\vert\vec k\vert$

381: larger than this value the decay

382: is much steeper, although only for $\mu=0.2$ does it resemble an

383: exponential form.

384: Because of the technical difficulties in treating correlators with power-law

385: decays in a finite volume we have made no attempt to

386: fit the numerical data for ${\cal C}_{\gamma_5}(\vert\vec k\vert,x_0)$

387: to a functional form.

388: 

389: Light states in the $\rho$ channel have long been of interest.

390: Here we discuss tentative results in the vector channel, corresponding 

391: to the quantum numbers of the $\rho$ meson. 

392: In this case for $\vert\vec k\vert>0$

393: we can distinguish

394: between ${\cal C}_{\gamma_\parallel}(\vec k)$,

395: in which the component of the vector is parallel

396: to $\vec k$, and ${\cal C}_{\gamma_\perp}(\vec k)$.

397: In Fig.~\ref{fig:rhodisp} we plot

398: the correlator for several $\vert\vec k\vert$ values at $\mu=0.6$ in each case.

399: For ${\cal C}_{\gamma_\perp}$

400: the curves are qualitatively very similar to those of

401: Fig.~\ref{fig:psdisperse} at $\mu=0.6$, with a distinguished momentum

402: $\vert\vec k\vert={3\pi\over8}$.

403: The correlator ${\cal C}_{\gamma_\parallel}$ is much smaller in

404: magnitude, and is consistent with exponential rather than algebraic decay. The

405: lines are fits of the form ${\cal C}_{\gamma_\parallel}(\vert\vec k\vert,x_0)=

406: A(e^{-Ex_0}+e^{-E(L_t-x_0)})$.

407: For small $\vert\vec k\vert$ the resulting $E(\vert\vec k\vert)$ 

408: resembles that of a massless

409: pole, ie. $E=\beta_0\vert\vec k\vert$, with velocity $\beta_0\approx0.5$.

410: It should be mentioned, however, that although ${\cal C}_{\gamma_\parallel}$ and

411: ${\cal C}_{\gamma_\perp}$ still differed, no evidence for a massless pole

412: was seen in the data at $\mu=0.8$.

413: A potential problem for this picture is that the extracted velocity

414: $\beta_0\approx0.5$ is less than

415: $\beta_F\approx1$, implying that quasiparticles would experience damping via

416: \v Cerenkov radiation of zero sound. Clearly further work exploring the

417: systematic

418: effects of varying $\mu$, $g^2$, $\vec k$ and volume

419: will be needed for a more complete understanding to emerge.

420: \smallskip

421: \bigskip

422: \begin{figure}[]

423: \begin{center}

424: \begin{minipage}{ 0.69\linewidth}

425: \includegraphics[width=1.0 \linewidth]{./rhodisp_mon.eps}

426: \end{minipage}

427: \end{center} 

428: \caption{Vector correlators ${\cal C}_{\gamma_\perp}(\vert\vec k\vert,x_0)$

429: (left) and

430: ${\cal C}_{\gamma_\parallel}(\pi-\vert\vec k\vert,x_0)$ (right) at $\mu=0.6$.

431: The symbols have the same

432: meaning as in Fig.~\ref{fig:psdisperse}.}

433: \label{fig:rhodisp}

434: \end{figure}

435: 

436: \section{Mesonic wavefunctions and Friedel oscillations}

437: \label{sec:section5}

438: In this section we expose another characteristic of the Fermi surface 

439: by studying the spatial correlations between 

440: fermions and antifermions in various mesonic channels, probed via the

441: wavefunction $\Psi(\vec x)$ defined by 

442: \begin{equation} 

443: \Psi_\Gamma(\vec x)=\lim_{x_0\to\infty}\Psi_\Gamma^{-1}(\vec x=\vec0) 

444: \sum_{\vec y}\langle{\cal G}_q(\vec 0,0;\vec

445: y,x_0)\Gamma{\cal G}_{\bar q}(\vec 0,0;\vec y+\vec x,x_0)\Gamma\rangle,

446: \end{equation}

447: where as usual 

448: $\Gamma$ projects out the quantum numbers of the channel of interest.

449: In the GNM$_3$ and related models $\Psi(\vec x)$ is technically 

450: much easier to define and measure than in QCD-like theories where they are 

451: gauge-dependent.

452: Meson wavefunctions in the GNM$_3$ model have been studied at $T,\mu=0$ in

453: \cite{strouthos02}, where further technical details are given.

454: 

455: It is easy to show \cite{hands03} that in the large-$N_f$ approximation the PS wavefunction in 

456: the chirally broken phase is given by 

457: \begin{equation}

458: \lim_{x_0\to\infty}C(\vec y;x_0)\sim {M\over x_0}e^{-2Mx_0}\exp\left(-

459: {{\vert\vec y\vert^2M}\over4x_0}\right).

460: \end{equation}

461: The general profile of the wavefunction is a gaussian with width increasing as

462: $\surd x_0$.

463: In the chirally symmetric phase we get, 

464: \begin{equation}

465: \lim_{x_0\to0}C(\vec y;x_0)\sim{\mu\over x_0}e^{-2\mu x_0}J_0(\mu\vert\vec

466: y\vert).

467: \label{eq:friedel}

468: \end{equation}

469: The wavefunction profile no longer changes with  $x_0$, but instead oscillates

470: with a spatial frequency determined by $\mu$, which to this order may be

471: identified with the Fermi momentum $k_F$. The oscillations observed in $C(\vec y;x_0)$ 

472: are characteristic

473: of a sharp Fermi surface and are reminiscent of oscillations in

474: either the density-density correlation function, or the screened inter-particle

475: potential, in degenerate systems generically known as {\em Friedel

476: Oscillations\/} \cite{fri}.

477: \bigskip

478: \smallskip

479: \begin{figure}[hbt]

480: \begin{center}

481: \begin{minipage}{ 0.46\linewidth}

482: \includegraphics[width=1.0 \linewidth]{./mu0.4.eps}

483: \end{minipage}

484: \hspace{2mm}

485: \begin{minipage}{ 0.46\linewidth}

486: \includegraphics[width=1.0 \linewidth]{./mu0.6.eps}

487: \end{minipage}

488: \end{center}

489: \caption{

490: Scalar wavefunction at $\mu=0.4$ (left) and $\mu=0.6$ (right).

491: }

492: \label{wvfn}

493: \end{figure}

494: In Fig.~\ref{wvfn} we plot the PS wavefunctions at $1/g^2=0.75$ for 

495: $\mu=0.4, 0.6$. The simulation data are connected by solid lines. As $\mu$, 

496: and hence $k_F$ increases the oscillations decrease in wavelength in accordance 

497: with the theoretical expectation and provide a graphic illustration of the

498: presence of a sharp Fermi surface. The dashed lines show measurements taken with the same

499: lattice parameters but with the

500: interaction switched off. The disparity with the interacting theory is

501: small, though increasing with $\vert\vec x\vert$,

502: showing that the free field description of

503: the oscillations is qualitatively correct. A possible explanation

504: is the infinite value predicted for the

505: Debye mass in section \ref{sec:section3}, implying that interactions between static

506: quarks at non-zero $\vert\vec x\vert$ are completely screened.

507: Also shown is the

508: theoretical form (eq.~\ref{eq:friedel})

509: $\Psi(x)=J_0(k_F x)$ with $k_F\equiv\mu$, showing good agreement with the data

510: for small $x\lapprox2k_F^{-1}$. Unfortunately it appears hard to

511: obtain more quantitative information, such as an independent fit for

512: the Fermi momentum $k_F$, because $\vert J_0(kx)\vert$

513: decays only as $x^{-{1\over2}}$. This means that fits should

514: not only include the backwards-propagating signal $J_0(k_F(L_s-x))$ but also

515: image contributions $J_0(k_F(nL_s-x))$ -- our attempts to find a

516: satisfactory fit were unsuccessful. The figures therefore

517: simply show both the ``forwards'' $J_0(\mu x)$

518: and ``forwards-and-backwards'' $J_0(\mu x) + J_0(\mu(L_s-x))$ forms, showing

519: that neither gives a satisfactory description of the data over the full range

520: of $x$ and $\mu$.

521: We also checked that there are no significant differences among

522: the various

523: mesonic channels, implying that in contrast to the situation at $\mu=0$

524: \cite{strouthos02}, effects due to eg. $\sigma$ exchange are very hard to detect.

525: 

526: \section{Summary}

527: The four-fermion models remain the only interacting fermionic field 

528: theories both simulable by standard lattice methods at $\mu \neq 0$ 

529: and displaying a Fermi surface, thought to be an essential figure

530: of dense quark matter. We have shown that the $\mu \neq 0$, $T=0$ 

531: transition is strongly first order and a tricritical point exists 

532: in the $(\mu,T)$ plane at very low $T$. 

533: 

534: We also discussed in-medium effects on the character of mesons. 

535: The large-$N_f$ auxiliary boson propagator has a branch-cut in the 

536: complex-$k$ plane which is modified to an isolated pole in the high 

537: density phase. Lattice simulations with $N_f=4$ have verified this 

538: with reasonable precision. We have also shown that the connected 

539: (flavor non-singlet) diagrams instead of showing exponential falloff

540: with Euclidean time, they generically decay algebraically, signalling 

541: the presence of massless particle-hole excitations. 

542: We also presented tentative evidence for a massless pole in the 

543: vector channel, which is possibly a manifestation of zero sound.

544: Furthermore, 

545: a graphic confirmation of the presence of a sharp Fermi surface

546: was shown by the oscillatory behavior in mesonic wavefunctions, 

547: which resemble the Friedel oscillations in many-body physics. 

548: 

549: \section*{Acknowledgements}

550: The work presented here was done in collaboration with Simon Hands \cite{hands03}, 

551: John Kogut \cite{strouthos01, hands03}

552: and Thao Tran \cite{hands03}.

553: 

554: \begin{thebibliography}{99}

555: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

556: % Some macros are available for the bibliography:

557: %   o for general use

558: %      \JL : general journals          \andvol : Vol (Year) Page

559: %   o for individual journal 

560: %      \PR  : Phys. Rev.               \PRL : Phys. Rev. Lett.

561: %      \NP  : Nucl. Phys.              \PL  : Phys. Lett.

562: %      \JMP : J. Math. Phys.           \CMP : Commun. Math. Phys.

563: %      \PTP : Prog. Theor. Phys.       \JPSJ: J. Phys. Soc. Jpn.

564: %      \JP  : J. of Phys.              \NC  : Nouvo Cim.

565: %      \IJMP: Int. J. Mod. Phys.       \ANN : Ann. of Phys.

566: % Usage:

567: %   \PR{D45,1990,345}            ==> Phys.~Rev.\ {\bf D45} (1990), 345

568: %   \JL{Phys.~Lett.,A30,1981,56} ==> Phys.~Lett.\ {\bf A30} (1981), 56

569: %   \andvol{B123,1995,1020}      ==> {\bf B123} (1995), 1020

570: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

571: \bibitem{rosen91}

572: B. Rosenstein, B.J. Warr and S.H. Park, Phys. Rep. {\bf 205} (1991), 59.

573: 

574: \bibitem{hands93}

575: S.J. Hands, A. Kocic and K.B. Kogut, Ann. Phys. {\bf 224} (1993), 29.

576: 

577: \bibitem{hands03}

578: S.J. Hands, J.B. Kogut, C.G. Strouthos and T.N. Tran, {\bf D68} (2003), 016005.

579: 

580: \bibitem{hands02}

581: S.J. Hands, B. Lucini and S.E. Morrison, Phys.~Rev.\ {\bf D65} (2002), 036004.

582: 

583: \bibitem{klimenko}

584: K.G. Klimenko, Z. Phys. {\bf C37} (1988), 457. \\

585: B. Rosenstein, B.J. Warr and S.H. Park, Phys. Rev. {\bf D39} (1989), 3088. \\

586: S.J. Hands, A. Kocic and J.B. Kogut, Nucl. Phys. {\bf B390} (1993), 355.

587: 

588: \bibitem{hands95}

589: S. Hands, S. Kim and J.B. Kogut, Nucl. Phys. {\bf B442} (1995), 364.

590: 

591: \bibitem{stephanov96}

592: M.A. Stephanov, Phys. Rev. Lett. {\bf 76} (1996), 4472.

593: 

594: \bibitem{hands99}

595: I. Barbour, S. Hands, J.B. Kogut, M.-P. Lombardo and S. Morrison, Nucl. Phys. {\bf B557}

596: (1999), 327.

597: 

598: \bibitem{strouthos01}

599: J.B. Kogut and C.G. Strouthos, Phys. Rev. {\bf D63} (2001), 054502.

600: 

601: \bibitem{buballa}

602: M. Buballa, Nucl. Phys. {\bf A609} (1996), 519.

603: 

604: \bibitem{kogut98}

605: J.B. Kogut, M.A. Stephanov and C.G. Strouthos, Phys. Rev. {\bf D58} (1998), 096001.

606: 

607: \bibitem{allton02}

608: C.R. Allton, J.E. Clowser, S.J. Hands, J.B. Kogut and C.G. Strouthos, Phys. Rev. {\bf D66}

609: (2002), 094511.

610: 

611: \bibitem{strouthos02}

612: S.J. Hands, J.B. Kogut and C.G. Strouthos, Phys. Rev. {\bf D65} (2002), 202302.

613: 

614: \bibitem{fri}

615: A.L. Fetter and J.D. Walecka, {\em Quantum Theory of Many-Particle Systems}, 

616: (McGraw-Hill, New York, 1971).

617: 

618: \end{thebibliography}

619: 

620: \end{document}

621: