1: \global\def\draftcontrol{0}

2: 

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

4:    \def\versionno{ n=2thermo -- draft   }

5: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

6: 

7: \catcode`\@=11

8: 

9: %%%

10: %%%%%% draftcontrol

11: %%%%%%%%%

12: \expandafter\ifx\csname draftcontrol\endcsname\relax\global\def\draftcontrol{0}

13: \fi

14: 

15: %%%

16: %%%%%% draftdate

17: %%%%%%%%%

18: {\count255=\time\divide\count255 by 60

19: \xdef\hourmin{\number\count255}

20: \multiply\count255 by-60\advance\count255 by\time

21: \xdef\hourmin{\hourmin:\ifnum\count255<10 0\fi\the\count255}}

22: \def\draftdate{\number\month/\number\day/\number\year\ \ \ \hourmin }

23: 

24: %%%

25: %%%%%% titlepage (adapted from article.cls)

26: %%%%%%%%%

27: \newcommand\makepapertitle{\par

28:   \begingroup

29:     \renewcommand\thefootnote{\@fnsymbol\c@footnote}%

30:     \def\@makefnmark{\rlap{\@textsuperscript{\normalfont\@thefnmark}}}%

31:     \long\def\@makefntext##1{\parindent 1em\noindent

32:             \hb@xt@1.8em{%

33:                 \hss\@textsuperscript{\normalfont\@thefnmark}}##1}%

34:      \newpage

35:      \global\@topnum\z@   % Prevents figures from going at top of page.

36:      \@makepapertitle

37:      \thispagestyle{empty}\@thanks

38:   \endgroup

39:   \setcounter{footnote}{0}%

40:   \global\let\thanks\relax

41:   \global\let\makepapertitle\relax

42:   \global\let\@makepapertitle\relax

43:   \global\let\@thanks\@empty

44:   \global\let\@author\@empty

45:   \global\let\@date\@empty

46:   \global\let\@title\@empty

47:   \global\let\title\relax

48:   \global\let\author\relax

49:   \global\let\date\relax

50:   \global\let\and\relax

51:   \def\version{\let\version\@version\@gobble}

52: }

53: \def\@makepapertitle{%

54:   \newpage

55:    \ifnum\draftcontrol=1 {}

56:    \version\versionno

57:    \vskip 3em%

58:    \else

59:    \hfill\hbox to 3cm {\parbox{4cm}{\@pubnum}\hss}%

60:    \vskip 3em%

61:    \fi

62:    \begin{center}%

63:    \let \footnote \thanks

64:      {\LARGE {\@title}}%

65:      \vskip 1.5em%

66:      {\normalsize%\large

67:        \lineskip .5em%

68:        \begin{tabular}[t]{c}%

69:          \@author

70:        \end{tabular}\par}%

71:      \vskip 1.5em%

72:      {\@bstract}%

73:      \end{center}%

74:      \vskip 1.5em

75:      \@date%

76:    \par

77: }

78: 

79: \gdef\@pubnum{}

80: %\@latex@error{No \noexpand \pubnum given}\@ehc}

81: \def\pubnum#1{%

82:   \gdef\@pubnum{#1}}

83: 

84: \gdef\@bstract{}

85: \def\Abstract#1{%

86:   \gdef\@bstract{%

87:    \parbox{\textwidth-0pc}{%

88:    \centerline{\bf Abstract}\penalty1000%

89: \kern.2cm%

90: \noindent%\abstractfont \baselineskip=12pt

91: \renewcommand\baselinestretch{1.0}%

92: {#1}}}

93: }

94: 

95: 

96: %%%

97: %%%%%% pagestyle

98: %%%%%%%%% for body of text

99: %%%%%%%%%%%%

100: \def\ps@paper{\let\@mkboth\@gobbletwo%

101:      \ifnum\draftcontrol=1

102:     \def\@oddfoot{\hbox to \textwidth{\tiny \versionno \hfil\tiny\draftdate}%

103:     \hskip -\textwidth \hbox to \textwidth{\hfil\rm\thepage\hfil}}%

104:      \else\def\@oddfoot{\hbox to \textwidth{\hfil\rm\thepage\hfil}}

105:      \fi

106:      \let\@evenfoot\@oddfoot

107: }

108: %%%%%%%%%%%%%%%%%%

109: 

110: %%%

111: %%%%%% front vs body vs coda

112: %%%%%%%%%

113: \def\body{\clearpage

114: %         \pagenumbering{arabic}

115:           \pagestyle{paper}

116:     }

117: \newenvironment{acknowledgments}{%

118: \vskip 3.25ex

119: %\@startsection {section}{1}{\z@}%

120: %                                   {-3.25ex \@plus -1ex \@minus -.2ex}%

121: %                                   {1.5ex \@plus.2ex}%

122: %                                   {\normalfont\normalsize\bfseries}

123: \noindent {\bf Acknowledgments}

124: %\vskip 0in

125: %\noindent

126: }

127: 

128: %%%%%%%%%%%%

129: 

130: %%%

131: %%%%%% definitions for draftmode

132: %%%%%%%%%

133: %%%%%%%%%%%% versioncontrol

134: %%%%%%%%%%%%%%%

135: \def\@version#1{\ifnum\draftcontrol=1

136: \typeout{}\typeout{#1}\typeout{}

137: \vskip3mm\centerline{\hbox{\fbox{\normalsize{\tt DRAFT -- #1 -- }

138:                    {\draftdate}}}}\vskip3mm

139: \fi}

140: \let\version\@version

141: %%%%%%%%%%%% labels in math mode

142: %%%%%%%%%%%%%%%

143: \long\def\eqlabel#1{\ifnum\draftcontrol=1

144:                     \tag@false  % there are some problems with multline without this

145:                     \tag*{(\theequation) \hbox to -0.2cm{\hspace{0cm}\small{#1}\hss}}

146:                     \refstepcounter{equation}

147:                     \edef\@currentlabel{\theequation}

148:                     \ltx@label{#1}          % use old LaTeX \label instead of new definition

149:                                             % of \label in AMSLaTeX.

150:                     \else

151:                     \label{#1}

152:                     \fi

153:                     }

154: %%%%%%%%%%%% citations keys in bibliography (very short version of showkeys.sty)

155: %%%%%%%%%%%%%%%

156: \let\st@bibitem\@bibitem

157: \let\st@lbibitem\@lbibitem

158: \ifnum\draftcontrol=1

159:   \def\@bibitem#1{%

160:     \st@bibitem{#1}\a@@label{#1}\ignorespaces}

161:   \def\@lbibitem[#1]#2{%

162:     \st@lbibitem[#1]{#2}\a@@label{#2}\ignorespaces}

163:   \def\a@@label#1{%

164:     \gdef\a@lab{\smash{\normalfont\small#1}}

165:     \ifvmode

166:       \if@inlabel

167:         \global\setbox\@labels\hbox{%

168:           \llap{\a@lab\let\a@lab\relax

169:                 \kern\@totalleftmargin\kern\marginparsep}%

170:           \box\@labels}%

171:       \fi

172:     \fi}

173: \fi

174: %%%%%%%%%%%%%%%

175: 

176: 

177: %%%

178: %%%%%% start

179: %%%%%%%%%

180: \documentclass[12pt,letterpaper]{article}

181: %%%%%%%%%%%%

182: 

183: %%%

184: %%%%%% load packages

185: %%%%%%%%%

186: \usepackage{amsmath,amssymb,array,calc,epsfig}

187: \usepackage{psfrag,verbatim,bm}

188: \usepackage[nosort]{cite}

189: %%%%%%%%%%%%

190: 

191: %%%

192: %%%%%% relax

193: %%%%%%%%%

194: \ifnum\draftcontrol=1

195: \tolerance=1000

196: \fi

197: %%%%%%%%%%%%

198: 

199: %%%

200: %%%%%% layout

201: %%%%%%%%%

202: \renewcommand\baselinestretch{1.25}

203: \setlength{\paperheight}{11in}

204: \setlength{\paperwidth}{8.5in}

205: \setlength{\textwidth}{\paperwidth-2.4in}     \hoffset= -.3in   % +1in from printer

206: \setlength{\textheight}{\paperheight-2.4in}   \topmargin= -.6in % +1in from printer

207: 

208: %%%%%%%%% section titles

209: \renewcommand\section{\@startsection {section}{1}{\z@}%

210:                                    {-3.5ex \@plus -1ex \@minus -.2ex}%

211:                                    {2.3ex \@plus.2ex}%

212:                                    {\normalfont\large\bfseries}}

213: \renewcommand\subsection{\@startsection{subsection}{2}{\z@}%

214:                                    {-3.25ex\@plus -1ex \@minus -.2ex}%

215:                                    {1.5ex \@plus .2ex}%

216:                                    {\normalfont\normalsize\bfseries}}

217: \renewcommand\subsubsection{\@startsection{subsubsection}{3}{\z@}%

218:                                    {-3.25ex\@plus -1ex \@minus -.2ex}%

219:                                    {1.5ex \@plus .2ex}%

220:                                    {\normalfont\normalsize\it}}

221: \renewcommand\paragraph{\@startsection{paragraph}{4}{\z@}%

222:                                    {-3.25ex\@plus -1ex \@minus -.2ex}%

223:                                    {1.5ex \@plus .2ex}%

224:                                    {\normalfont\normalsize\bf}}

225: 

226: %%%

227: %%%%%% number equations within sections

228: %%%%%%%%%

229: \numberwithin{equation}{section}

230: 

231: %%%

232: %%%%%% macros

233: %%%%%%%%%

234: 

235: %%%%%%%%% standard

236: %%%%%%%%%%%%

237: 

238: \def\ie{{\it i.e.}}

239: \def\eg{{\it e.g.}}

240: 

241: \def\revise#1       {\raisebox{-0em}{\rule{3pt}{1em}}%

242:                      \marginpar{\raisebox{.5em}{\vrule width3pt\

243:                      \vrule width0pt height 0pt depth0.5em

244:                      \hbox to 0cm{\hspace{0cm}{%

245:                      \parbox[t]{4em}{\raggedright\footnotesize{#1}}}\hss}}}}

246: 

247: \newcommand\fnxt[1] {\raisebox{.12em}{\rule{.35em}{.35em}}\mbox{\hspace{0.6em}}#1}

248: \newcommand\nxt[1]  {\\\fnxt#1}

249: 

250: \def\cala         {{\cal A}}

251: \def\calA         {{\mathfrak A}}

252: \def\calAbar      {{\underline \calA}}

253: \def\calb         {{\cal B}}

254: \def\calc         {{\cal C}}

255: \def\cald         {{\cal D}}

256: \def\cale         {{\cal E}}

257: \def\calf         {{\cal F}}

258: \def\calg         {{\cal G}}

259: \def\calG         {{\mathfrak G}}

260: \def\calh         {{\cal H}}

261: \def\cali         {{\cal I}}

262: \def\calj         {{\cal J}}

263: \def\calk         {{\cal K}}

264: \def\call         {{\cal L}}

265: \def\calm         {{\cal M}}

266: \def\caln         {{\cal N}}

267: \def\calo         {{\cal O}}

268: \def\calp         {{\cal P}}

269: \def\calq         {{\cal Q}}

270: \def\calr         {{\cal R}}

271: \def\cals         {{\cal S}}

272: \def\calt         {{\cal T}}

273: \def\calu         {{\cal U}}

274: \def\calv         {{\cal V}}

275: \def\calw         {{\cal W}}

276: 

277: \def\complex      {{\mathbb C}}

278: \def\naturals     {{\mathbb N}}

279: \def\projective   {{\mathbb P}}

280: \def\rationals    {{\mathbb Q}}

281: \def\reals        {{\mathbb R}}

282: \def\zet          {{\mathbb Z}}

283: 

284: \def\del          {\partial}

285: \def\delbar       {\bar\partial}

286: \def\ee           {{\rm e}}

287: \def\ii           {{\rm i}}

288: \def\chain        {{\circ}}

289: \def\tr           {\mathop{\rm Tr}}

290: \def\Re           {{\rm Re\hskip0.1em}}

291: \def\Im           {{\rm Im\hskip0.1em}}

292: \def\id           {{\it id}}

293: 

294: \def\de#1#2{{\rm d}^{#1}\!#2\,}

295: \def\De#1{{\cald}#1\,}

296: 

297: \def\half{{\frac12}}

298: \newcommand\topa[2]{\genfrac{}{}{0pt}{2}{\scriptstyle #1}{\scriptstyle #2}}

299: \def\undertilde#1{{\vphantom#1\smash{\underset{\widetilde{\hphantom{\displaystyle#1}}}{#1}}}}

300: \def\prodprime{\mathop{{\prod}'}}

301: \def\gsq#1#2{%

302:     {\scriptstyle #1}\square\limits_{\scriptstyle #2}{\,}} % Ginsparg square

303: \def\sqr#1#2{{\vcenter{\vbox{\hrule height.#2pt

304:  \hbox{\vrule width.#2pt height#1pt \kern#1pt

305:  \vrule width.#2pt}\hrule height.#2pt}}}}

306: \def\square{%

307:   \mathop{\mathchoice{\sqr{12}{15}}{\sqr{9}{12}}{\sqr{6.3}{9}}{\sqr{4.5}{9}}}}

308: 

309: %%%%%%%%% jtl macros

310: %%%%%%%%%%%%

311: \newcommand{\fft}[2]{{\frac{#1}{#2}}}

312: \newcommand{\ft}[2]{{\textstyle{\frac{#1}{#2}}}}

313: \def\jsquare{\mathop{\mathchoice{\sqr{8}{32}}{\sqr{8}{32}}

314: {\sqr{6.3}{9}}{\sqr{4.5}{9}}}}

315: 

316: 

317: \newcommand{\wn}{\mathfrak{w}}

318: 

319: 

320: %%%%%%%%% paper specific macros

321: %%%%%%%%%%%%

322: 

323: 

324: \def\a{\alpha}

325: \def\b{\beta}

326: \def\w{\omega}

327: \def\r{\rho}

328: \def\dd{\delta}

329: \def\hdd{\hat{\delta}}

330: \def\e{\epsilon}

331: \def\c{\chi}

332: \def\ga{\gamma}

333: \def\ee{\epsilon}

334: \def\hx{\hat{x}}

335: \def\rh{\hat{\rho}}

336: \def\xh{\hat{x}}

337: \def\chih{\hat{\chi}}

338: \def\hh{\hat{h}}

339: \def\aa1{\phi}

340: \def\cc1{\psi}

341: \def\hh{\hat{h}}

342: \def\hr{\hat{\rho}}

343: \def\nm{\nabla_\mu}

344: \def\nn{\nabla_\nu}

345: \def\G{\Gamma}

346: \def\arctanh{{\rm arctanh}}

347: \def\hdd{\hat{\delta}}

348: \def\tQ{\tilde{Q}}

349: \def\k{\kappa}

350: %%%%%%%%%%%%

351: 

352: \catcode`\@=12

353: 

354: \begin{document}

355: 

356: %%%

357: %%%%%% text starts here

358: %%%%%%%%%

359: 

360: \title{Thermodynamics of the $\caln=2^*$ strongly coupled plasma}

361: 

362: \pubnum{%

363: UWO-TH-07/01

364: MCTP-07-04}

365: \date{January 2007}

366: 

367: \author{

368: Alex Buchel$ ^{1,2}$,  Stan Deakin$ ^1$, Patrick Kerner$^{1}$ and James T. Liu$^{3}$\\[0.4cm]

369: \it $ ^1$Department of Applied Mathematics\\

370: \it University of Western Ontario\\

371: \it London, Ontario N6A 5B7, Canada\\[0.2cm]

372: \it $ ^2$Perimeter Institute for Theoretical Physics\\

373: \it Waterloo, Ontario N2J 2W9, Canada\\[0.2cm]

374: \it $^3$Michigan Center for Theoretical Physics\\

375: \it Randall Laboratory of Physics, The University of Michigan\\

376: \it Ann Arbor, MI 48109--1040, USA

377: }

378: 

379: \Abstract{

380: Gauge/string duality is a potentially important framework for

381: addressing the properties of the strongly coupled quark gluon plasma

382: produced at RHIC.  However, constructing an actual string theory dual

383: to QCD has so far proven elusive.  In this paper, we take a partial

384: step towards exploring the QCD plasma by investigating the

385: thermodynamics of a non-conformal system, namely the $\caln=2^*$

386: theory, which is obtained as a mass deformation of the conformal

387: $\caln=4$ gauge theory.  We find that at temperatures of order the

388: mass scale, the thermodynamics of the mass deformed plasma is

389: surprisingly close to that of the conformal gauge theory plasma.  This

390: suggests that many properties of the quark gluon plasma at RHIC may in

391: fact be well described by even relatively simple models such as that

392: of the conformal $\caln=4$ plasma.

393: }

394: 

395: 

396: 

397: 

398: 

399: 

400: \makepapertitle

401: 

402: \body

403: 

404: \version\versionno

405: 

406: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

407: \section{Introduction }

408: 

409: The purpose of this paper is to explore properties of strongly coupled

410: nonconformal gauge theory plasma within the framework of the gauge

411: theory/string theory correspondence of Maldacena \cite{m1,m2}.  We are

412: primarily motivated by recent attempts (see \cite{raj} and references therein)

413: to describe the quark gluon plasma (QGP) produced at RHIC

414: \cite{rhic1,rhic2,rhic3,rhic4} from the dual holographic perspective.

415: Unfortunately, we do not as yet have a controllable string theory dual

416: to QCD. In fact, most applications of the Maldacena correspondence to RHIC

417: physics are discussed in the context of conformal $\caln=4$ supersymmetric

418: $SU(N)$ Yang-Mills (SYM) theory in the 't~Hooft (planar) limit and for

419: large 't~Hooft coupling.  Remarkably, $\caln=4$ SYM plasma

420: \cite{ads1,ads2,ads3,ads4,ads5,ads6,ads7,ads8,ads9} as a model for the

421: RHIC QGP appears to be reasonably good \cite{shu2}. We would like to

422: stress, however, that such an agreement is rather paradoxical.

423: One requires a strong 't~Hooft coupling in order to have a controllable

424: string theory dual, and the QGP does indeed appear to be strongly coupled.

425: However, it is strongly coupled because it is produced at temperatures of

426: order the QCD strong coupling scale, where, naively, conformal invariance

427: is badly broken.  So, why then does the conformal gauge theory 

428: plasma serve as {\it the model} for RHIC physics? 

429: 

430: As a first step towards answering the latter question, we study the

431: thermodynamics of the $\caln=2^*$ ({\it i.e.}~mass deformed $\caln=4$)

432: SYM plasma over a wide range of temperatures and for different mass 

433: deformations%

434: %

435: \footnote{Although there have been previous attempts to study the

436: thermodynamics of strongly coupled nonconformal four dimensional gauge

437: theory plasmas, they have been limited to the high temperature regime,

438: where the theory is almost conformal \cite{kt1,kt2,kt3,bl,aby}. A notable difference is 

439: a recent study of thermodynamics of strongly coupled 

440: $\caln=2$ gauge theory plasma with massive fundamental hypermultiplets \cite{rob}.}.

441: %

442: The expectation here is that the deformation mass scale would provide

443: a model for the QCD strong coupling scale, and the thermodynamics of 

444: the mass deformed plasma at temperatures of order the mass scale would

445: then serve as a more realistic model for the RHIC QGP.

446: 

447: 

448: The paper is organized as follows. In the next section we review the

449: non-extremal $\caln=2^*$ geometry constructed in \cite{bl}, the holographic

450: renormalization of this theory \cite{hr} and also recall the explicit

451: expressions for the free energy density, energy density and the entropy

452: density \cite{hr,bbs}.  In section~\ref{sec:gtvsp} we discuss the map

453: between finite temperature $\caln=2^*$ gauge theory parameters and the

454: parameters of the dual non-extremal geometry \cite{bpp,bl}. In

455: section~\ref{sec:thtpwf} we review the high-temperature thermodynamics

456: of the finite temperature PW flow \cite{bl,hr,bbs}.  After this, we

457: turn towards a numerical investigation of the flow away from the

458: high-temperature regime. We describe our numerical procedure in

459: section~\ref{sec:np}, and present the results of the analysis in

460: section~\ref{sec:r}.  Finally, we conclude and also outline open

461: problems in section~\ref{sec:cfd}.       

462: 

463: 

464: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

465: \section{Non-extremal $\caln=2^*$ geometry}

466: \label{sec:neg}

467: 

468: The supergravity background dual to finite

469: temperature $\caln=2^*$ gauge theory \cite{bl} is a deformation

470: of the original $AdS_5\times S^5$ geometry induced by a pair of

471: scalars $\alpha$ and $\chi$ of the five-dimensional gauge

472: supergravity. (At zero temperature, such a deformation was

473: constructed by Pilch and Warner \cite{pw}.) According to the

474: general scenario of a holographic RG flow, the asymptotic

475: boundary behavior of the supergravity scalars is related to the

476: bosonic and fermionic mass parameters of the relevant operators

477: inducing the RG flow in the boundary gauge theory. Based on such a

478: relation, and the fact that $\alpha$ and $\chi$ have conformal dimensions

479: two and one, respectively, we call the supergravity 

480: scalar $\a$ a {\it bosonic} deformation, and the supergravity scalar $\chi$  

481: a {\it fermionic} deformation of the D3-brane geometry.

482: 

483: The action of the five-dimensional gauged supergravity including the

484: scalars $\alpha$ and $\chi$ is given by

485: %

486: \begin{equation}

487: \begin{split}

488: S=&\,

489: \int_{\calm_5} d\xi^5 \sqrt{-g}\ \call_5\\

490: =&\frac{1}{4\pi G_5}\,

491: \int_{\calm_5} d\xi^5 \sqrt{-g}\left[\ft14 R-3 (\del\a)^2-(\del\chi)^2-

492: \calp\right]\,,

493: \end{split}

494: \eqlabel{action5}

495: \end{equation}

496: %

497: where the potential%

498: %

499: \footnote{We set the five-dimensional gauged

500: supergravity coupling to one. This corresponds to setting the

501: radius $L$ of the five-dimensional sphere in the undeformed metric

502: to $2$.}

503: %

504: %

505: \begin{equation}

506: \calp=\frac{1}{16}\left[\frac 13 \left(\frac{\del W}{\del

507: \a}\right)^2+ \left(\frac{\del W}{\del \chi}\right)^2\right]-\frac

508: 13 W^2\,

509:  \eqlabel{pp}

510: \end{equation}

511: %

512: is a function of $\alpha$ and $\chi$, and is determined by the

513: superpotential

514: %

515: \begin{equation}

516: W=- e^{-2\alpha} - \frac{1}{2} e^{4\alpha} \cosh(2\chi)\,.

517: \eqlabel{supp}

518: \end{equation}

519: %

520: In our conventions, the five-dimensional Newton's constant is

521: %

522: \begin{equation}

523: G_5\equiv \frac{G_{10}}{2^5\ {\rm vol}_{S^5}}=\frac{4\pi}{N^2}\,.

524: \eqlabel{g5}

525: \end{equation}

526: %

527: The action \eqref{action5} yields the Einstein equations

528: %

529: \begin{equation}

530: R_{\mu\nu}=12 \del_\mu \a \del_\nu\a+ 4 \del_\mu\chi\del_\nu\chi

531: +\frac43 g_{\mu\nu} \calp\,,

532: \eqlabel{ee}

533: \end{equation}

534: %

535: as well as the equations for the scalars

536: %

537: \begin{equation}

538: \jsquare\alpha=\fft16\fft{\del\calp}{\del\alpha}\,,\qquad

539: \jsquare\chi=\fft12\fft{\del\calp}{\del\chi}\,.

540: \eqlabel{scalar}

541: \end{equation}

542: %

543: 

544: To construct a finite-temperature version of the Pilch-Warner

545: flow, we choose an ansatz for the metric respecting rotational

546: but not the Lorentzian invariance

547: %

548: \begin{equation}

549: ds_5^2=-c_1^2(r)\ dt^2 +c_2^2(r)\ \left( d x_1^2 +  d x_2^2 + d x_3^2 \right)

550: +  dr^2\,.\\

551: \eqlabel{ab}

552: \end{equation}

553: %

554: With this ansatz, the equations of motion for the background become

555: %

556: \begin{equation}

557: \begin{split}

558: &\a''+\a'\ \left(\ln{c_1c_2^3}\right)'

559: -\frac 16 \frac{\del\calp}{\del\a}=0\,,\\

560: &\c''+\c'\ \left(\ln{c_1c_2^3}\right)'-\frac 12 \frac{\del\calp}{\del\c}=0\,,\\

561: &c_1''+c_1'\ \left(\ln{c_2^3}\right)'+\frac 43c_1  \calp=0\,,\\

562: &c_2''+c_2'\ \left(\ln{c_1c_2^2}\right)'+\frac 43c_2  \calp=0\,,

563: \end{split}

564: \eqlabel{beom}

565: \end{equation}

566: %

567: where the prime denotes a derivative with respect to the radial

568: coordinate $r$.

569: In addition, there is a first-order constraint

570: %

571: \begin{equation}

572: \left(\a'\right)^2+\frac 13 \left(\c'\right)^2 -\frac 13

573: \calp-\frac 12 (\ln c_2)'(\ln c_1 c_2)' =0\,. 

574: \eqlabel{backconst}

575: \end{equation}

576: %

577: It was shown in \cite{bl} that any solution to \eqref{beom} and

578: \eqref{backconst} can be lifted to a 

579: full ten-dimensional solution of type IIb supergravity. This includes 

580: the metric, the three- and five-form fluxes, the dilaton and the axion.

581: In particular, the ten-dimensional 

582: Einstein frame metric is given by Eq.~(4.12) in \cite{bl}.

583: 

584: For finite temperature flows, we find it convenient to introduce a

585: new radial coordinate $x$ which spans the range from the horizon to

586: the boundary in finite coordinate distance:

587: %

588: \begin{equation}

589: 1-x(r) = \frac{c_1}{c_2}\,, \qquad x\in [0,1]\,.

590: \eqlabel{radgauge}

591: \end{equation}

592: %

593: With this new coordinate, the black brane's horizon is at $x=1$,

594: while the boundary of the asymptotically $AdS_5$ space-time is at

595: $x=0$%

596: %

597: \footnote{This $x$ coordinate is inappropriate for the extremal PW

598: flow, since in that case $x$ is always vanishing.  In this sense,

599: backgrounds with and without horizons (or equivalently physics at

600: finite temperature and zero temperature) may have distinctly different

601: characteristics.}.

602: %

603: The background equations of motion \eqref{beom} become

604: %

605: \begin{equation}

606: \begin{split}

607: &c_2''+4 c_2\ (\alpha')^2-\frac {1}{x-1} c_2'-\frac{5}{c_2} (c_2')^2

608: +\frac 43 c_2\ (\c')^2 =0\,,\\

609: &\alpha''+\frac {1}{x-1} \, \alpha'-

610: \frac{1}{12\, \calp c_2^2 (x-1)}\biggl[     (x-1) \left(6 (\alpha')^2+2(\chi')^2\right) c_2^2\\

611: &\kern16em -3 c_2' c_2-6 (c_2')^2 (x-1) \biggr] \,

612:  \frac{\del\calp}{\del\a} =0\,,\\

613: &\chi''+\frac {1}{x-1} \, \chi'-

614: \frac{1}{4\, \calp c_2^2 (x-1)}\biggl[     (x-1) \left(6 (\alpha')^2+2(\chi')^2\right) c_2^2 \\

615: &\kern16em-3 c_2' c_2-6 (c_2')^2 (x-1) \biggr] \,

616:  \frac{\del\calp}{\del\c} =0\,,

617: \end{split}

618: \eqlabel{beomx}

619: \end{equation}

620: %

621: where the prime now denotes a derivative with respect to $x$.

622: We demand that a physical RG flow should correspond to a background

623: geometry with a regular horizon.  To ensure regularity, it is

624: necessary to impose the following boundary conditions at the horizon:

625: %

626: \begin{equation}

627: x\to 1_- : \qquad

628: \biggl\{\a(x),\ \c(x),\ c_2(x)\biggr\}\longrightarrow \biggl\{\dd_1,\dd_2,\dd_3\biggr\}\,,

629: \eqlabel{boundh}

630: \end{equation}

631: %

632: where the $\dd_i$ are constants.

633: 

634: The boundary conditions at $x= 0$ are determined from the

635: requirement that the solution should approach the $AdS_5$ geometry

636: as $x\to 0_+$:

637: %

638: \begin{equation}

639: x\to 0_+ :  \qquad

640: \biggl\{\a(x),\ \c(x),\ c_2(x)\biggr\}\longrightarrow

641: \biggl\{0,0,\propto x^{-1/4}\biggr\}\,.

642: \eqlabel{bboun}

643: \end{equation}

644: %

645: The three supergravity parameters

646: $\dd_i$ uniquely determine a non-singular RG flow in the dual gauge theory.

647: As we review in section~\ref{sec:gtvsp}, they

648: are unambiguously related to the three physical parameters in

649: the gauge theory:

650: the temperature $T$, and the bosonic and fermionic masses

651: $m_b$ and $m_f$ of the $\caln=2^*$ hypermultiplet components.

652: 

653: A general analytical solution of the system \eqref{beomx} with the boundary

654: conditions  \eqref{boundh}, \eqref{bboun} is unknown.

655: However, it is possible to find an analytical solution in the regime of

656: high temperatures%

657: %

658: \footnote{Non-extremal $AdS_5$ geometry is obtained as a trivial solution for

659: vanishing bosonic and fermionic deformations: $\a=\chi=0$ identically.}

660: %

661: (see section~\ref{sec:thtpwf}).

662: Notice that given \eqref{pp} and \eqref{supp}, one can consistently truncate 

663: the supergravity system \eqref{beomx} to bosonic deformation only, \ie, by

664: setting $\chi=0$ identically. On the other hand, it is inconsistent (beyond

665: the linear approximation) to set bosonic deformation to zero, \ie, to set

666: $\a=0$ identically while keeping the fermionic deformation $\chi\ne 0$. 

667: 

668: In the rest of this section we discuss the asymptotic singularity-free

669: solution of \eqref{beomx} near the boundary $x\to 0_+$ and near the horizon

670: $x\to 1_-$, constraint by the boundary conditions \eqref{bboun} and

671: \eqref{boundh}, respectively. We justify referring to the $\a$ 

672: and $\chi$ scalars as bosonic and fermionic deformations. We also explain

673: the advantages of using the radial coordinate \eqref{radgauge} in numerical

674: integration.  Finally, we recall holographic renormalization of the thermal

675: PW flows.  

676: 

677: In what follows we find it convenient to introduce 

678: %

679: \begin{equation}

680: \begin{split}

681: &c_2\equiv e^A\,,\qquad A\equiv \ln\hdd_3-\frac 14 \ln (2x-x^2)+a(x)\,,\\

682: &\r\equiv e^\a\,.

683: \end{split} 

684: \eqlabel{defc2}

685: \end{equation}

686: %

687: The form of $A(x)$ is chosen to extract the leading asymptotic behavior

688: $c_2\sim x^{-1/4}$ from the metric function $c_2(x)$.  In addition, the

689: introduction of $\rho=e^\a$ is natural, as the bosonic scalar $\a$ enters

690: exponentially in the superpotential \eqref{supp}.  Imposing the boundary

691: conditions \eqref{boundh} and \eqref{bboun}, we see that the new warp factor

692: $a(x)$ and scalar $\r(x)$ satisfy the boundary conditions

693: %

694: \begin{equation}

695: \begin{split}

696: &x\to 0_+:\qquad a(x)\to 0\,,\qquad \r(x)\to 1\,,\\

697: &x\to 1_-:\qquad a(x)\to \ln\frac{\dd_3}{\hdd_3}\,,\qquad \r(x)\to e^{\dd_1}\,,

698: \end{split}

699: \eqlabel{bcax}

700: \end{equation}

701: %

702: where $\hdd_3$ is a new constant.

703: 

704: \subsection{Asymptotics of the thermal PW flow near the boundary}

705: 

706: The most general solution of \eqref{beomx} subject to the boundary conditions \eqref{bboun} and \eqref{bcax} may be expanded as a series around $x=0$:

707: %

708: \begin{equation}

709: \begin{split}

710: \r=\,&1+x^{1/2}\left(\r_{10}+\r_{11}\ \ln x\right)+x\left(\r_{20}+\r_{21}\ \ln x+\r_{22}\ \ln^2 x\right)\kern4em\\

711: &+\dots+ x^{k/2} \left(\sum_{i=1}^k \r_{ki}\ \ln^i x\right)+\cdots\,,

712: \end{split}

713: \eqlabel{bounexp1}

714: \end{equation}

715: %

716: \begin{equation}

717: \begin{split}

718: \chi=\,&\c_0 x^{1/4}\Biggl[

719: 1+x^{1/2}\left(\c_{10}+\c_{11}\ \ln x\right)+x\left(\c_{20}+\c_{21}\ \ln x+\c_{22}\ \ln^2 x\right)\\

720: &+\dots+ x^{k/2} \left(\sum_{i=1}^k \c_{ki}\ \ln^i x\right)+\cdots\Biggr]\,,

721: \end{split}

722: \eqlabel{bounexp2}

723: \end{equation}

724: %

725: \begin{equation}

726: \begin{split}

727: a=\,&x^{1/2}\left(\a_{10}+\a_{11}\ \ln x\right)+x\left(\a_{20}+\a_{21}\ \ln x+\a_{22}\ \ln^2 x\right)\kern5em\\

728: &+\dots+ x^{k/2} \left(\sum_{i=1}^k \a_{ki}\ \ln^i x\right)+\cdots\,.

729: \end{split}

730: \eqlabel{bounexp3}

731: \end{equation}

732: %

733: This solution is characterized by five independent parameters

734: %

735: \begin{equation}

736: \biggl\{\ln\hdd_3;\ \r_{11},\r_{10};\ \c_0,\c_{10}\biggr\}\,,

737: \eqlabel{par}

738: \end{equation}

739: %

740: where we have included $\hdd_3$ from \eqref{defc2}.  The remaining

741: series coefficients are completely determined from the above parameters.

742: 

743: For the holographic renormalization (see subsection \ref{subsec:htftpf}) 

744: we will need the coefficients of the first two subleading terms in 

745: \eqref{bounexp1}--\eqref{bounexp3}.  Explicitly, these are given by

746: %

747: \begin{equation}

748: \begin{split}

749: \r_{22}=&\frac 32 \r_{11}^2\,,\\

750: \r_{21}=&3\r_{10}\r_{11}-8 \r_{11}^2+\frac{26}{9}\c_0^2\r_{11}\,,\\

751: \r_{20}=&24\r_{11}^2+\frac 32 \r_{10}^2-8\r_{10}\r_{11}+\frac{26}{9}\c_0^2\r_{10}-\frac{104}{9}\c_0^2\r_{11}+\frac 13 \c_0^4\,,\kern7.5em

752: \end{split}

753: \eqlabel{rr2}

754: \end{equation}

755: %

756: \begin{equation}

757: \begin{split}

758: \kern-17.8em\c_{11}=&\frac 13 \c_0^2\,,\kern10em

759: \end{split}

760: \eqlabel{cc1}

761: \end{equation}

762: %

763: \begin{equation}

764: \begin{split}

765: \c_{22}=&\frac 92 \r_{11}^2\,,\\

766: \c_{21}=&2\c_0^2\r_{11}+9\r_{10}\r_{11}+\frac{13}{36}\c_0^4-\frac{51}{2}\r_{11}^2\,,\\

767: \c_{20}=&\frac 18-6\c_0\r_{11}+\frac 92\r_{10}^2+\frac{243}{4}\r_{11}^2-\frac{51}{2}\r_{11}\r_{10}+2\c_0^2\r_{10}+\frac{13}{12}

768: \c_0^2\c_{10}-\frac{131}{120}\c_0^4\,,

769: \end{split}

770: \eqlabel{cc2}

771: \end{equation}

772: %

773: \begin{equation}

774: \begin{split}

775: a_{11}=&0\,,\\

776: a_{10}=&-\frac 19 \c_0^2\,,\kern27em

777: \end{split}

778: \eqlabel{aa1}

779: \end{equation}

780: %

781: \begin{equation}

782: \begin{split}

783: a_{22}=&-\frac 12 \r_{11}^2\,,\\

784: a_{21}=&-\r_{10}\r_{11}-\frac 12 \r_{11}^2-\frac{1}{12}\c_0^4\,,\\

785: a_{20}=&-\frac 12 \r_{10}\r_{11}-\frac 14 \c_0^2\c_{10}-\frac 34 \r_{11}^2+\frac{13}{648}\c_0^4-\frac 12 \r_{10}^2\,.\kern10.5em

786: \end{split}

787: \eqlabel{aa2}

788: \end{equation}

789: %

790: As indicated above, the non-extremal $AdS_5$ geometry is obtained by setting $\a\equiv 0$, $\c\equiv 0$, which in asymptotic expansions

791: \eqref{rr2}--\eqref{aa2} corresponds to taking $\r_{11}=\r_{10}=\c_0=0$. 

792: This leads to

793: %

794: \begin{equation}

795: ds_5^2 = (2\pi T)^2 \left(2x-x^2\right)^{-1/2} \left( -\left(1- x\right)^2 dt^2 + d x_1^2 +

796: d x_2^2 + d x_3^2 \right) + \frac{dx^2}{2x-x^2}\,,

797: \eqlabel{d3n}

798: \end{equation}

799: %

800: where the temperature $T$ is given by

801: %

802: \begin{equation}

803: 2\pi T=\hdd_3\,.

804: \end{equation}

805: %

806: Matching with asymptotic extremal $AdS_5$ geometry with a standard Poincare

807: patch radial coordinate $R$ (and with our conventional choice for the radius

808: of curvature $L=2$)

809: %

810: \begin{equation}

811: ds_5^2 = \frac{R^2}{4} \left( - dt^2 + d x_1^2 +

812: d x_2^2 + d x_3^2 \right) + 4\frac{dR^2}{R^2}\,,

813: \eqlabel{ads5}

814: \end{equation}

815: %

816: we identify near the boundary

817: %

818: \begin{equation}

819: x\propto R^{-4}\,,\qquad x\to 0_+\,.

820: \eqlabel{xR}

821: \end{equation}

822: %

823: Given \eqref{xR} and the  asymptotic expansions

824: \eqref{bounexp1}--\eqref{bounexp3}, we identify the conformal weight two

825: supergravity scalar $\r(x)$ as dual to turning on mass for the bosonic

826: components of the $\caln=2^*$ hypermultiplet. Parameters $\r_{11}$ and

827: $\r_{10}$ are coefficients of its non-normalizable and normalizable 

828: modes, correspondingly. Similarly, the conformal weight one supergravity

829: scalar $\c(x)$ can be identified as dual to turning on mass for the

830: fermionic components of the $\caln=2^*$ hypermultiplet. Parameters $\c_{0}$

831: and $\c_{10}$ are coefficients of its non-normalizable and normalizable 

832: modes, correspondingly. We discuss the precise relation of $\r_{11}$ and

833: $\c_0$ to the $\caln=2^*$ gauge theory bosonic and fermionic masses in

834: section~\ref{sec:gtvsp}.  Here, we would simply like to emphasize the

835: following well-known fact \cite{ps}: given $\{\hdd, \r_{11}, \c_0\}$,

836: the coefficients of the normalizable modes of the supergravity scalars

837: $\r(x)$ and $\c(x)$, namely $\{\r_{10},\c_{10}\}$, are uniquely fixed

838: by requiring that the resulting supergravity RG flow is nonsingular in

839: the bulk. This statement is simply the supergravity dual to a gauge

840: theory lore: for a fixed temperature and mass parameters, the bosonic and

841: fermionic condensates are determined uniquely%

842: %

843: \footnote{Strictly speaking this is true in the absence of moduli. 

844: However, we do not expect moduli at finite temperature, and thus unbroken

845: supersymmetry, on the gauge theory side.}.

846: %

847: Thus, for nonsingular finite temperature Pilch-Warner flows, we must have

848: %

849: \begin{equation}

850: \r_{10}=\r_{10}(\hdd_3,\r_{11},\c_0)\,,\qquad

851: \c_{10}=\c_{10}(\hdd_3,\r_{11},\c_0)\,.

852: \eqlabel{condrel}

853: \end{equation}

854: %

855: 

856: \subsection{Asymptotics of the thermal PW flow near the horizon}

857: We now consider the behavior of the solution near the horizon, $x=1$.

858: Notice that the equations \eqref{beomx} are invariant under the transformation

859: %

860: \begin{equation}

861: (1-x)\ \Leftrightarrow\ -(1-x)\,.

862: \end{equation}

863: %

864: It is the straightforward to verify that boundary conditions \eqref{boundh},

865: \ie~the finiteness of $\{a(x), \r(x), \c(x)\}$ near the horizon, implies that

866: $\{a(x), \r(x), \c(x)\}$ are even functions of $(1-x)$ for $|1-x|\ll 1$.

867: It is this fact that justifies our choice of the radial coordinate $x$

868: in \eqref{radgauge}. Indeed, keeping fixed  $\{\hdd_3, \r_{11}, \c_0\}$,

869: for generic values of $\{\r_{10}, \c_{10}\}$, the functions

870: $\{a(x), \r(x), \c(x)\}$  would diverge at the horizon.  Thus, fixing

871: coefficients of the normalizable modes $\{\r_{10}, \c_{10}\}$ from 

872: the requirement of nonsingularity of the holographic RG flow is equivalent

873: to imposing Neumann boundary conditions on $\{a(x), \r(x), \c(x)\}$

874: at the horizon, \ie~as $x\to 1_-$

875: %

876: \begin{equation}

877: \lim_{x\to 1_-} \r'=\lim_{x\to 1_-} \c'=\lim_{x\to 1_-} a'=0\,.

878: \eqlabel{nbc}

879: \end{equation} 

880: %

881: The boundary condition \eqref{nbc} would determine bosonic and fermionic

882: condensate dependence on the hypermultiplet masses and the temperature

883: \eqref{condrel}.

884: 

885: 

886: 

887: \subsection{Holographic thermodynamics of the finite temperature PW flow}

888: \label{subsec:htftpf}

889: 

890: We use the method of holographic renormalization in order to examine the

891: thermodynamics of the $\caln=2^*$ RG flow.  Holographic renormalization of

892: this flow was explained in \cite{hr}, where it was investigated  using the

893: original radial coordinate $r$ of the metric ansatz \eqref{ab}. 

894: In particular, parameterizing

895: %

896: \begin{equation}

897: c_1(r)=e^{A(r)+B(r)}\,,\qquad c_2(r)=e^{A(r)}\,,

898: \eqlabel{recal1}

899: \end{equation}

900: %

901: the results of \cite{hr} indicate that the entropy density $s$, the energy

902: density $\cale$, and the free energy density $\calf$ are given by

903: %

904: \begin{equation}

905: \begin{split}

906: s=&\frac{1}{4G_5}\ \lim_{r\to r_{\rm horizon}} e^{3A}\,,\\

907: \cale=&\frac{1}{8\pi G_5}\ \lim_{r\to\infty}\biggl[-3e^{4A+B}A'+2e^{4A+B}\biggl\{\a_1+\a_3\ \a+\a_4\ \chi^2+\a_5\ \a^2\\

908: &\kern6em

909: +\a_6\ \a\chi^2+\a_8\ \frac{\a^2}{\ln\ee}

910: +\ln\ee\ \a_{10}\ \chi^4+\a_{11}\chi^4\biggr\}\biggr]\,,\\

911: \calf=&\cale-s T=\cale -\frac{1}{8\pi G_5}\ \lim_{r\to\infty}\biggl[e^{4A+B}B'\biggr]\,,

912: \end{split}

913: \eqlabel{efin}

914: \end{equation}

915: %

916: where 

917: %

918: \begin{equation}

919: \ee\equiv \sqrt{-g_{tt}}=e^{A+B}\,,

920: \eqlabel{edef}

921: \end{equation}

922: %

923: and the boundary counterterm coefficients $\a_i$ take the values

924: %

925: \begin{equation}

926: \begin{split}

927: \a_1&=\frac 34\,,\qquad \a_2=\frac{1}{4}\,,\qquad \a_3=0\,,\qquad 

928: \a_4=\frac 12\,,\qquad

929: \a_5=3\,,\qquad \a_6=0\,,\\

930: \a_7&=0\,,\qquad \a_8=-\frac 32\,,

931: \qquad \a_9=- \frac{1}{3}\,,\qquad \a_{10}=-\frac 23\,,\qquad  \a_{11}=\frac 16\,.

932: \end{split}

933: \eqlabel{ai1}

934: \end{equation}

935: %

936: Using the asymptotic expansions \eqref{bounexp1}--\eqref{bounexp3}, and

937: changing the radial coordinate in \eqref{efin} following \eqref{radgauge},

938: we find 

939: %

940: \begin{equation}

941: \begin{split}

942: \calf=-\frac{\hdd_3^4}{32\pi G_5}\biggl(

943: &1+\r_{11}^2 (24-96\ln\hdd_3+24\ln 2)-24 \r_{10}\r_{11}+2\c_0^2\c_{10}\\

944: &+\c_0^4\left(\frac 49-\frac 23\ln 2+\frac 83\ln\hdd_3\right)\biggr)\,,

945: \end{split}

946: \eqlabel{changedf}

947: \end{equation}

948: %

949: \begin{equation}

950: \begin{split}

951: \cale=\calf-\frac{1}{8\pi G_5}\hdd^4_3\,,\qquad sT=\frac {1}{8\pi G_5}\hdd^4_3\,.

952: \kern12em

953: \end{split}

954: \eqlabel{changede}

955: \end{equation}

956: %

957: Finally, the entropy density is given by 

958: %

959: \begin{equation}

960: s=\frac{\hdd_3^3 e^{3a_h}}{4G_5}\,,\qquad a_h=\lim_{x\to 1_-} a(x)\,,

961: \eqlabel{changeds}

962: \end{equation}

963: %

964: and thus we extract the temperature from \eqref{changede}

965: %

966: \begin{equation}

967: T=\frac{\hdd_3}{2\pi}\ e^{-3a_h}\,.

968: \eqlabel{tem}

969: \end{equation}

970: %

971: 

972: Of course, the same value of the Hawking temperature \eqref{tem} can be

973: extracted from the near-horizon geometry, provided one recalls that $sT$ is

974: a renormalization group flow invariant in the supergravity black brane

975: geometries without a chemical potential \cite{blu}. Indeed, from 

976: the relation between components of the Ricci tensor 

977: %

978: \begin{equation}

979: R_{x_1}^{\ x_1}=R_t^{\ t}\,,

980: \eqlabel{rel}

981: \end{equation}

982: %

983: we have a constraint%

984: %

985: \footnote{Eq.~\eqref{rel1} can also be directly derived from \eqref{beom}.}

986: %

987: %

988: \begin{equation}

989: c_2^4\ \left(\frac {c_1}{c_2}\right)'={\rm constant}\,.

990: \eqlabel{rel1}

991: \end{equation}

992: %

993: Evaluating the LHS of \eqref{rel1} near the horizon, we have 

994: %

995: \begin{equation}

996: \lim_{r\to r_{\rm horizon}}\ c_2^4\ \left(\frac {c_1}{c_2}\right)'

997: =8\pi G_5\ sT\,.

998: \eqlabel{rel2}

999: \end{equation}

1000: %

1001: Changing to the radial coordinate $x$ in \eqref{radgauge} and using the

1002: asymptotic expansions \eqref{bounexp1}--\eqref{bounexp3}, the LHS of

1003: \eqref{rel1} near the boundary takes form%

1004: %

1005: \footnote{In principle, one does not have to take a limit in \eqref{rel3}

1006: since this expression is a constant.}

1007: %

1008: %

1009: \begin{equation}

1010: \lim_{r\to \infty}\ c_2^4\ \left(\frac {c_1}{c_2}\right)'=\lim_{x\to 0_+} \left(-\frac{dx}{dr}\right)c_2^4=\hdd_3^4\,.

1011: \eqlabel{rel3}

1012: \end{equation}

1013: %

1014: Comparing \eqref{rel2} with \eqref{rel3} leads to the value of $sT$ presented

1015: in \eqref{changede}.  It is precisely the holographic RG invariance of the

1016: combination $sT$ that guarantees the basic thermodynamic relation 

1017: %

1018: \begin{equation}

1019: \calf=\cale-s T\,.

1020: \eqlabel{basictermo}

1021: \end{equation} 

1022: %

1023: Relation \eqref{basictermo} was verified explicitly for the $\caln=2^*$

1024: thermal RG flow in \cite{bl}; the same relation holds in other thermal

1025: gauge/gravity duals, for example in the thermal cascading RG flow \cite{aby}.

1026: 

1027: 

1028: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

1029: \section{Gauge theory versus supergravity parameters}

1030: \label{sec:gtvsp}

1031: 

1032: The relation between $N=2^*$ gauge theory and the supergravity parameters of

1033: the (thermal) Pilch-Warner flow was established in \cite{bpp,bl}. For

1034: completeness, we review the main points here. We begin with the gauge theory,

1035: then move to the supersymmetric PW flow \cite{bpp}, and finally discuss the

1036: gauge/gravity parameter identification at finite temperature. 

1037: 

1038: \subsection{$N=2^*$ gauge theory}

1039: In the language of four-dimensional $\caln=1$ supersymmetry, the

1040: mass deformed $\caln=4$ $SU(N)$ Yang-Mills theory ($\caln=2^*$) in

1041: $\reals^{3,1}$ consists of a vector multiplet $V$, an adjoint chiral

1042: superfield $\Phi$ related by $\caln=2$ supersymmetry to the gauge

1043: field, and two additional adjoint chiral multiplets $Q$ and $\tilde{Q}$

1044: which form an $\caln=2$ hypermultiplet.  In addition to the usual

1045: gauge-invariant kinetic terms for these fields%

1046: %

1047: \footnote{The classical K\"{a}hler potential is normalized

1048: according to $(2/g_{YM}^2)\tr[\bar{\Phi}\Phi+ \bar{Q}Q+\bar{\tQ}\tQ]$.},

1049: %

1050: the theory has additional interactions and a hypermultiplet mass term

1051: given by the superpotential

1052: %

1053: \begin{equation}

1054: W=\frac{2\sqrt{2}}{g_{YM}^2}\tr([Q,\tQ]\Phi)

1055: +\frac{m} {g_{YM}^2}(\tr Q^2+\tr\tQ^2)\,.

1056: \eqlabel{sp}

1057: \end{equation}

1058: %

1059: When $m=0$, the gauge theory is superconformal with $g_{YM}$

1060: characterizing an exactly marginal deformation. The theory has a

1061: classical $3(N-1)$ complex dimensional moduli space, which is

1062: protected by supersymmetry against (non)-perturbative quantum

1063: corrections.

1064: 

1065: When $m\ne 0$, the $\caln=4$ supersymmetry is softly broken to

1066: $\caln=2$. This mass deformation lifts the $\{Q,\ \tQ\}$ hypermultiplet

1067: moduli directions, leaving the $(N-1)$ complex dimensional Coulomb

1068: branch of the $\caln=2$ $SU(N)$ Yang-Mills theory, parameterized by

1069: expectation values of the adjoint scalar

1070: %

1071: \begin{equation}

1072: \Phi={\rm diag} (a_1,a_2,\cdots,a_N)\,,\quad \sum_i a_i=0\,,

1073: \eqlabel{adsc}

1074: \end{equation}

1075: %

1076: in the Cartan subalgebra of the gauge group.  For generic values

1077: of the moduli $a_i$, the gauge symmetry is broken to that of the

1078: Cartan subalgebra $U(1)^{N-1}$, up to the permutation of individual

1079: $U(1)$ factors. Additionally, the superpotential \eqref{sp} induces

1080: the RG flow of the gauge coupling.  While from the gauge theory

1081: perspective it is straightforward to study this $\caln=2^{*}$ theory

1082: at any point on the Coulomb branch \cite{dw}, the PW supergravity

1083: flow \cite{pw} corresponds to a particular Coulomb branch vacuum.

1084: More specifically, matching the probe computation in gauge theory

1085: and the dual PW supergravity flow, it was argued in \cite{bpp} that

1086: the appropriate Coulomb branch vacuum corresponds to a  linear

1087: distribution of the vevs \eqref{adsc} as

1088: %

1089: \begin{equation}

1090: a_i\in [-a_0,a_0]\,,\qquad a_0^2=\frac{m^2 g_{YM}^2 N}{\pi}\,,

1091: \eqlabel{inter}

1092: \end{equation}

1093: %

1094: with (continuous in the large $N$ limit) linear number density

1095: %

1096: \begin{equation}

1097: \rho(a)=\frac{2}{m^2 g_{YM}^2}\sqrt{a_0^2-a^2}\,,\qquad

1098: \int_{-a_0}^{a_0}da \rho(a)=N\,.

1099: \eqlabel{rho}

1100: \end{equation}

1101: %

1102: Unfortunately, the extension of the $N=2^*$ gauge/gravity

1103: correspondence of \cite{pw,bpp,ejp} for vacua other than \eqref{rho}

1104: is not known.

1105: 

1106: In \cite{bpp,ejp} the dynamics of the gauge theory on the D3 brane

1107: probe in the PW background was studied in detail.  It was shown

1108: in \cite{bpp} that the probe has a one complex dimensional moduli

1109: space, with bulk induced  metric precisely equal to the metric on

1110: the appropriate one complex dimensional submanifold of the $SU(N+1)$

1111: $\caln=2^*$ Donagi-Witten theory Coulomb branch.  This one dimensional

1112: submanifold is parameterized by the expectation value $u$ of the

1113: $U(1)$ complex scalar on the Coulomb branch of the theory where

1114: $SU(N+1)\rightarrow U(1)\times SU(N)_{PW}$. Here the $_{PW}$ subscript

1115: denotes that the $SU(N)$ factor is in the Pilch-Warner vacuum

1116: \eqref{rho}. Whenever $u$ coincides with any of the $a_i$ of the

1117: PW vacuum, the moduli space metric diverges, signaling the appearance

1118: of additional massless states.  An identical divergence is observed

1119: \cite{bpp,ejp} for the probe D3-brane at the {\it enhancon}

1120: singularity of the PW background.  Away from the singularity locus,

1121: $u=a\in [-a_0,a_0]$, the gauge theory computation of the probe

1122: moduli space metric is one-loop exact.  This is due to the suppression

1123: of instanton corrections in the large $N$ limit \cite{bpp,b} of

1124: $\caln=2$ gauge theories.

1125: 

1126: Consider now $\caln=2^*$ gauge theory at finite temperature $T$.

1127: Of course, finite temperature completely breaks supersymmetry. Thus, we

1128: can generalize the thermal $\caln=2^*$ gauge theory by allowing for different

1129: (non-equal) masses $m_b$ and $m_f$ for the bosonic and fermionic components

1130: of the $\caln=2^*$ hypermultiplet $\{Q,\tQ\}$ correspondingly. It is only

1131: when $m_b=m_f=m$, and at zero temperature $T=0$, that we have $\caln=2$

1132: supersymmetry.  Since turning on mass terms for the bosonic or fermionic

1133: components of the hypermultiplet corresponds to deforming the $\caln=4$

1134: supersymmetric Yang-Mills Lagrangian by relevant operators of different

1135: classical dimension (a dimension two operator for bosonic and a dimension

1136: three operator for fermionic mass terms), such deformations will be encoded

1137: in different supergravity modes.  As indicated in section~\ref{sec:neg},

1138: turning on bosonic/fermionic masses corresponds to turning on the

1139: five-dimensional supergravity scalars $\a$/$\c$ correspondingly.

1140: 

1141: We would like to conclude this section with a simple observation.

1142: Turning on bosonic/fermionic masses for the hypermultiplet components

1143: sets a strong coupling scale $\Lambda \propto \max\{m_b,m_f\}$.  In this

1144: case, we expect to find two qualitatively different phases of this gauge

1145: theory, depending on whether $T\gg \Lambda$ or $T\ll \Lambda$.

1146: In the former case we expect the thermodynamics to be qualitatively (and

1147: quantitatively in the limit $T/\Lambda\to \infty$) similar to that of

1148: the $\caln=4$ gauge theory plasma \cite{kp}. On the other hand, when

1149: $T\sim \Lambda$, and with $m_f=0$, we actually expect an instability in the system.

1150: Indeed\footnote{We would like to thank Ofer Aharony for clarifying this.},

1151: turning on only the supergravity scalar $\a$, \ie, setting $m_b\ne 0$ and $m_f=0$,

1152: corresponds to giving positive mass-squared to four out of six $\caln=4$ scalars

1153: (these are the bosonic components of the $\caln=2$ hypermultiplet).

1154: The remaining two $\caln=4$ scalars at the same time obtain a negative mass-squared \cite{pw} ---

1155: they are tachyons at zero temperature. At high enough temperature the thermal corrections 

1156: would stabilize these tachyons. However, as we lower the temperature, we expect

1157: the re-emergence of these tachyons. As argued in \cite{gm}, dynamical

1158: instabilities in thermal systems are reflected in thermodynamic instabilities.

1159: Furthermore, it was argued in general (and demonstrated explicitly with a

1160: concrete example) \cite{binst} that thermodynamic instabilities are reflected

1161: to developing $c_s^2<0$, where $c_s$ is the speed of sound waves in the

1162: thermal gauge theory plasma.  We discuss the supergravity realization of these

1163: phenomena in section~\ref{sec:r}. 

1164: 

1165: 

1166: 

1167: \subsection{Supersymmetric PW flow}

1168: 

1169: It is clear that the following conditions must hold in order to preserve

1170: $\caln=2$ supersymmetry for the mass deformation of $\caln=4$ SYM:

1171: %

1172: \nxt the temperature must be zero: $T=0$; 

1173: \nxt the masses for the bosonic and fermionic components of the $\caln=2$ hypermultiplet $\{Q,\tQ\}$ must be the same: $m_b=m_f=m$.\\

1174: %

1175: The former condition corresponds to a restriction

1176: %

1177: \begin{equation}

1178: c_1(r)=c_2(r)\,,

1179: \eqlabel{t0}

1180: \end{equation}

1181: %

1182: ensuring Lorentz invariance of the metric.  In this case, the supergravity

1183: RG flow cannot be parameterized by the radial coordinate $x$ introduced in

1184: \eqref{radgauge}.  Instead, the supersymmetric flow is easiest to

1185: parameterize in terms of the fermionic scalar $\chi\in [0,+\infty)$,

1186: where $\chi=0$ corresponds to the asymptotic $AdS_5$ boundary, and

1187: $\chi\to +\infty$ to the enhancon location in the bulk \cite{bpp,ejp}.  

1188: The supersymmetric Pilch-Warner solution is then given by \cite{pw}

1189: %

1190: \begin{equation}

1191: \begin{split}

1192: e^A&=\frac{k \r^2}{\sinh(2\chi)}\,,\\

1193: \r^6&=\cosh(2\chi)+\sinh^2(2\chi)\,\ln\frac{\sinh(\chi)}{\cosh(\chi)}\,,

1194: \end{split}

1195: \eqlabel{pwsolution}

1196: \end{equation}

1197: %

1198: where the single integration constant $k$ is related to the hypermultiplet

1199: mass $m$ according to \cite{bpp}

1200: %

1201: \begin{equation}

1202: k= m L =2 m\,.

1203: \eqlabel{kim}

1204: \end{equation}

1205: %

1206: In order to identify the thermal RG parameters $\{\r_{11},\c_0\}$ with

1207: the gauge theory masses, we need the boundary asymptotics of the PW flow.

1208: This was computed in \cite{bl}.  Introducing 

1209: %

1210: \begin{equation}

1211: \xh\equiv e^{-r/2}\,,

1212: \eqlabel{rx}

1213: \end{equation}

1214: %

1215: we have 

1216: %

1217: \begin{equation}

1218: \begin{split}

1219: \chi&= k\xh \biggl[1+ k^2\xh^2\left(\ft13+\ft{4}{3}\ln(k\xh)\right)

1220: +k^4\xh^4\left(-\ft{7}{90}+\ft{10}{3}\ln(k\xh)+\ft{20}{9}\ln^2(k\xh)\right)

1221: \\

1222: &\qquad+\calo\left(k^6\xh^6\ln^3(k\xh)\right)\biggr]\,,\\

1223: \r&= 1+ k^2\xh^2 \left(\ft13+\ft{2}{3}\ln(k\xh)\right)

1224: +k^4\xh^4\left(\ft{1}{18}+2\ln(k\xh)+\ft23\ln^2(k\xh)\right)+\calo\left(k^6\xh^6\ln^3(k\xh)\right)\,,\\

1225: A&= -\ln (2\xh)- \ft13 k^2\xh^2

1226: -k^4\xh^4\left(\ft{2}{9}+\ft{10}{9}\ln(k\xh)+\ft{4}{9}\ln^2(k\xh)\right)+\calo\left(k^6\xh^6\ln^3(k\xh)\right)\,.

1227: \end{split}

1228: \eqlabel{uvass}

1229: \end{equation}

1230: 

1231: 

1232: \subsection{Thermal PW flow}

1233: \label{subsec:tpwf}

1234: 

1235: We now consider the thermal PW flow with $m_b=m_f=m$.  Here we would like to

1236: identify the parameters $\r_{11}$ and $\c_0$ for such a flow in terms of the

1237: PW quantity $k$ (and thus, following \eqref{kim}, the hypermultiplet mass $m$)

1238: by comparing the boundary asymptotics \eqref{bounexp1}--\eqref{bounexp3} with

1239: \eqref{uvass}.  In order to do so, we first establish the relation between the

1240: radial coordinate $x$ in \eqref{bounexp1}--\eqref{bounexp3} and the radial

1241: coordinate $\hx$ in \eqref{uvass}.  Matching the asymptotic warp factors $e^A$

1242: for the zero temperature PW flow and the thermal PW flow as $x\to 0_+$, we find

1243: %

1244: \begin{equation}

1245: x\sim 8\hdd_3^4\ \hx^4\,,\qquad x\to 0_+\,.

1246: \eqlabel{xxh}

1247: \end{equation}

1248: %

1249: Next, using \eqref{xxh} and matching the leading nontrivial asymptotics

1250: [order $x^{1/2}\ln x$ for $\r(x)$ and order $x^{1/4}$ for $\c(x)$] for

1251: both the thermal PW solution and the supersymmetric PW flow, we find

1252: %

1253: \begin{equation}

1254: \r_{11}=\frac{2^{1/2}\ k^2}{24\ \hdd_3^2}=\frac{\sqrt{2}}{24\pi^2}\ e^{-6a_h}\ \left(\frac mT

1255: \right)^2\,,\qquad \c_0=\frac{2^{1/4}\ k}{2\ \hdd_3}=\frac{1}{2^{3/4}\pi}\ e^{-3a_h}\ \left(\frac mT\right)\,,

1256: \eqlabel{match}

1257: \end{equation} 

1258: %

1259: where we used \eqref{tem}.  We emphasize that the above matching procedure is

1260: well-defined --- it was shown in \cite{bl} that finite temperature effects

1261: modify the asymptotic supersymmetric PW geometry at order $\calo(x)\sim

1262: \calo(\hx^4)$, which is subleading compared to the order at which the

1263: matching \eqref{match} is done. 

1264: 

1265: Motivated by \eqref{match}, we now propose that in the general case,

1266: \ie~for $m_b\ne m_f$, we may independently extract bosonic and fermionic

1267: masses according to

1268: %

1269: \begin{equation}

1270: \r_{11}=\frac{\sqrt{2}}{24\pi^2}\ e^{-6a_h}\ \left(\frac {m_b}{T}

1271: \right)^2\,,\qquad \c_0=\frac{1}{2^{3/4}\pi}\ e^{-3a_h}\ \left(\frac {m_f}{T}\right)\,.

1272: \eqlabel{match2}

1273: \end{equation}

1274: %

1275: In section~\ref{sec:thtpwf} we recall that such an identification leads to a

1276: consistent thermodynamics of the thermal PW flows \cite{hr,bbs} at high

1277: temperatures. Furthermore, we later show in section~\ref{sec:r} that this

1278: identification in fact leads to a consistent thermodynamics at any temperature.

1279: 

1280: 

1281: In the rest of this subsection we derive some useful expressions for the

1282: free energy and the speed of sound waves.  Specifically, we would like to

1283: express the free energy and the speed of sound in terms of the coefficients

1284: of the non-normalizable modes of the supergravity scalars $\r(x),\c(x)$. 

1285: We also present the differential constraint on coefficients of the

1286: normalizable modes of these supergravity scalars following from the first

1287: law of thermodynamics.

1288: 

1289: Since we will constantly refer to the parameters $\r_{11}$ and $\c_0$

1290: related to the bosonic and fermionic masses, we distinguish them from

1291: derived quantities (such as the subleading coefficients) by introducing

1292: the notation

1293: %

1294: \begin{equation}

1295: \xi\equiv \r_{11}\,,\qquad \eta\equiv \c_{0}\,.

1296: \eqlabel{not1}

1297: \end{equation}

1298: %

1299: Notice that because of \eqref{match2}, $\xi$ and $\eta$ are not independent,

1300: but are related through the ratio of bosonic and fermionic masses according

1301: to

1302: %

1303: \begin{equation}

1304: \eta^2=6\ \frac{m_f^2}{m_b^2}\ \xi\,.

1305: \eqlabel{not1a}

1306: \end{equation}

1307: %

1308: 

1309: The nonsingularity of the thermal PW flow at the horizon \eqref{nbc} will determine the coefficients $\{\r_{10},\c_{10}\}$ of the normalizable modes

1310: of the supergravity scalars $\{\r(x),\c(x)\}$ and also the value $a_h$ of

1311: the warp factor $a(x)$ at the horizon in terms of \eqref{not1} and

1312: \eqref{not1a}.  When $m_b\ne0$, we take

1313: %

1314: \begin{equation}

1315: \r_{10}=\r_{10}(\xi)\,,\qquad  \c_{10}=\c_{10}(\xi)\,,\qquad a_h=a_h(\xi)\,.

1316: \eqlabel{not2}

1317: \end{equation}

1318: %

1319: On the other hand, for $m_b=0$, we have instead

1320: %

1321: \begin{equation}

1322: \r_{10}=\r_{10}(\eta)\,,\qquad  \c_{10}=\c_{10}(\eta)\,,\qquad a_h=a_h(\eta)\,.

1323: \eqlabel{not2a}

1324: \end{equation}

1325: %

1326: Notice that a given  pair $\{\xi,\eta\}$ can be unambiguously related

1327: to bosonic and fermionic masses, measured with respect to temperature.

1328: Indeed, following \eqref{match2}

1329: %

1330: \begin{equation}

1331: \left(\frac{m_b}{T}\right)^2=12\sqrt{2}\ \pi^2\ e^{6a_h}\ \xi\,,\qquad \left(\frac{m_f}{T}\right)=2^{3/4}\ \pi\ 

1332: e^{3a_h}\ \eta\,.

1333: \eqlabel{match3}

1334: \end{equation}

1335: %

1336: Solving for $\hdd_3$ from \eqref{tem} and expressing $G_5$ in gauge theory variables \eqref{g5}, we find for the free energy

1337: %

1338: \begin{equation}

1339: \begin{split}

1340: \calf=&-\frac 18\pi^2 N^2 T^4\  e^{12a_h(\xi)}\left(1+24\xi^2\   \ln(\xi^2)\ \left(1-\frac{m_f^4}{m_b^4}\right)

1341: +12\xi\ \frac{m_f^2}{m_b^2}\ \c_{10}(\xi)-24\xi\ \r_{10}(\xi)\right)\\

1342: &+\calf_0\,,

1343: \end{split}

1344: \eqlabel{fres0}

1345: \end{equation}

1346: %

1347: assuming $m_b\ne 0$, and

1348: %

1349: \begin{equation}

1350: \begin{split}

1351: \calf=&-\frac 18\pi^2 N^2 T^4\  e^{12a_h(\eta)}\left(1-\frac 23\ \eta^4  \ln(\eta^4)

1352: +2\eta^2\ \c_{10}(\eta)\right)+\calf_0\,,

1353: \end{split}

1354: \eqlabel{fres00}

1355: \end{equation}

1356: %

1357: if $\xi=0$.  In the above expressions, $\calf_0$ does not 

1358: depend on temperature. Such a constant arises because holographic

1359: renormalization generically%

1360: %

1361: \footnote{Notice that this is not the case when $m_b=m_f$.}

1362: %

1363: leads to $\ln T$ dependence in the free energy, which requires the

1364: introduction of an arbitrary regularization scale $\mu$

1365: %

1366: \begin{equation}

1367: \ln T \ \to \ln \frac T\mu\,.

1368: \eqlabel{muintr}

1369: \end{equation}

1370: %

1371: Changing this scale modifies $\calf_0$, but otherwise has no effect on the

1372: thermodynamics.  This phenomena has been encountered previously; it was

1373: discussed in perturbative gauge theory thermodynamics in \cite{pertg}, and in

1374: the context of gauge/string theory correspondence in \cite{npertg}%

1375: %

1376: \footnote{It appears even in the high temperature thermodynamics of the

1377: thermal PW flow \cite{bbs} (see also section~\ref{sec:thtpwf}).}.

1378: %

1379: Notice that in \eqref{fres0} we wrote $\ln(\xi^2)$ instead of $2\ln (\xi)$.

1380: The reason is that the former expression allows us to study thermodynamics for

1381: $\xi<0$, which according to \eqref{match2} we interpret as $m_b^2<0$.

1382: While introducing tachyonic masses for gauge theory scalars at zero

1383: temperature leads to instability, this is not the case at finite temperature.

1384: Here, the effective mass squared receives thermal corrections of order $T^2$,

1385: which might cure the zero temperature instability. 

1386: 

1387: Besides the free energy, we can also evaluate the entropy density and

1388: the energy density

1389: %

1390: \begin{equation}

1391: s= \frac 12\pi ^2 N^2 T^3\ e^{12 a_h}\,,\qquad \cale=\calf+sT \,.

1392: \eqlabel{seres}

1393: \end{equation}

1394: %

1395: In addition, from the defining equations \eqref{match3}, we can evaluate

1396: the quantities $d\xi/dT$ and $d\eta/dT$:

1397: %

1398: \begin{equation}

1399: \begin{split}

1400: &\frac{d\xi}{dT}=-\frac{m_b^2}{6\sqrt{2}\pi^2 T^3}\ e^{-6 a_h(\xi)}\ \left(6\xi\ \frac{d a_h(\xi)}{d\xi}+1\right)^{-1}\,,

1401: \qquad m_b\ne 0\,;\\

1402: &\frac{d\eta}{dT}=-\frac{m_f}{2^{3/4}\pi T^2}\ e^{-3 a_h(\eta)}\ \left(3\eta\ \frac{d a_h(\eta)}{d\eta}+1\right)^{-1}\,,

1403: \qquad m_b= 0\,.\\

1404: \end{split}

1405: \eqlabel{defder}

1406: \end{equation}

1407: %

1408: Using \eqref{fres0} or \eqref{fres00}, \eqref{seres} and \eqref{defder} the first law of thermodynamics,

1409: %

1410: \begin{equation}

1411: d\calf=-s\ dT\,,

1412: \eqlabel{deff}

1413: \end{equation}

1414: %

1415: reduces to a differential constraint on the quantities%

1416: %

1417: \footnote{Notice that even though, for $\xi=0$, the free 

1418: energy density \eqref{fres00} does not depend on 

1419: $\r_{10}$, one generically expects that $\r_{10}=\r_{10}(\eta)$.}

1420: %

1421: %

1422: \begin{equation}

1423: \begin{split}

1424: &\{\xi,\c_{10}(\xi),\r_{10}(\xi),a_h(\xi)\}\qquad\hbox{for }m_b\ne0\,;\\

1425: &\{\eta,\c_{10}(\eta),\r_{10}(\eta),a_h(\eta)\}\qquad\hbox{for }m_b=0\,.

1426: \end{split}

1427: \end{equation}

1428: %

1429: In particular, the first law in differential form reads

1430: %

1431: \begin{equation}

1432: \begin{split}

1433: 0=&4\xi\left(1-\frac{m_f^4}{m_b^4}\right)+2\left(\r_{10}(\xi)-\xi\frac{d\r_{10}(\xi)}{d\xi}\right)

1434: -\frac{m_f^2}{m_b^2}\left(\c_{10}(\xi)-\xi\frac{d\c_{10}(\xi)}{d\xi}\right)+\frac{da_h(\xi)}{d\xi}\,,

1435: \end{split}

1436: \eqlabel{firstlaw}

1437: \end{equation}

1438: %

1439: for $m_b\ne 0$, and 

1440: %

1441: \begin{equation}

1442: 0=\frac 43 \eta^3+2\eta\ \c_{10}(\eta)-\eta^2\ \frac{d\c_{10}(\eta)}{d\eta}-6\frac{da_h(\eta)}{d\eta}\,,

1443: \eqlabel{firstlawf}

1444: \end{equation}

1445: %

1446: for $m_b=0$.  Using either \eqref{firstlaw} or \eqref{firstlawf}, we

1447: can evaluate the speed of sound entirely as a function of $\xi$ or $\eta$.

1448: Straightforward algebraic manipulations result in 

1449: %

1450: \begin{equation}

1451: c_s^2=\frac{\del P}{\del \cale}=-\frac{\del \calf}{\del \cale}=\frac{1+\Sigma}{3-\Sigma}\,,

1452: \eqlabel{vs2}

1453: \end{equation}

1454: %

1455: where

1456: %

1457: \begin{equation}

1458: \begin{split}

1459: \Sigma\equiv \Sigma(\xi)=&24\xi^2\ \left(\frac{m_f^4}{m_b^4}-1\right)+12\xi\left(\xi\frac{d\r_{10}(\xi)}{d\xi}-\r_{10}(\xi)\right)

1460: \\

1461: &+6\xi\ \frac{m_f^2}{m_b^2}\ \left(\c_{10}(\xi)-\xi\frac{d\chi_{10}(\xi)}{d\xi}\right)\,,\qquad m_b\ne 0\,;\\

1462: \Sigma\equiv \Sigma(\eta)=&\frac 23 \eta^4+\eta^2\ \chi_{10}(\eta)-\frac 12\ \eta^3\ \frac{d\chi_{10}(\eta)}{d\eta}

1463: \,,\qquad m_b=0\,,

1464: \end{split}

1465: \end{equation}

1466: %

1467: and where $P=-\calf$ is the pressure. 

1468: 

1469: In general, one may explore the PW thermodynamics in the entire phase plane

1470: of $m_b$-$m_f$.  However, here we focus on two special cases:

1471: %

1472: \nxt when $m_f=0$ (we call this the {\it bosonic} case)%

1473: %

1474: \footnote{Recall that this corresponds to a consistent truncation of the

1475: supergravity}

1476: %

1477: we have

1478: %

1479: \begin{equation}

1480: \calf_{bosonic}=-\frac{\pi^2N^2 T^4}{8}\ e^{12a_h(\xi)}\  \biggl(1+24\xi^2\ln(\xi^2)-24\xi\ \r_{10}(\xi)\biggr)+\calf_0\,,

1481: \eqlabel{mff}

1482: \end{equation}

1483: %

1484: \begin{equation}

1485: \Sigma_{bosonic}(\xi)=-24 \xi^2-12\xi\ \r_{10}(\xi) +12 \xi^2\ \frac{d\r_{10}(\xi)}{d\xi}\,,

1486: \eqlabel{speedb}

1487: \end{equation} 

1488: and the first law of thermodynamics takes form

1489: %

1490: \begin{equation}

1491: 0=4\xi+2\left(\r_{10}(\xi)-\xi\frac{d\r_{10}(\xi)}{d\xi}\right)

1492: +\frac{da_h(\xi)}{d\xi}\,;

1493: \eqlabel{1stb}

1494: \end{equation}

1495: %

1496: \nxt when $m_f=m_b\equiv m$ (we call this the {\it supersymmetric} case), or

1497: correspondingly from \eqref{not1a}

1498: %

1499: \begin{equation}

1500: \eta^2=6\xi\,,

1501: \eqlabel{susy}

1502: \end{equation}

1503: %

1504: we have 

1505: %

1506: \begin{equation}

1507: \begin{split}

1508: \calf_{susy}=&-\frac{\pi^2N^2 T^4}{8}\ e^{12a_h(\xi)}\  \biggl(1-24\xi\ \r_{10}(\xi)+12\xi\ \c_{10}(\xi)\biggr)+\calf_0\,,

1509: \end{split}

1510: \eqlabel{mfs}

1511: \end{equation}

1512: %

1513: \begin{equation}

1514: \Sigma_{susy}(\xi)=12\xi\left(\xi\frac{d\r_{10}(\xi)}{d\xi}-\r_{10}(\xi)\right)

1515: +6\xi\ \left(\c_{10}(\xi)-\xi\frac{d\chi_{10}(\xi)}{d\xi}\right)\,,

1516: \eqlabel{speeds}

1517: \end{equation} 

1518: %

1519: and the first law of thermodynamics takes form

1520: %

1521: \begin{equation}

1522: 0=2\left(\r_{10}(\xi)-\xi\frac{d\r_{10}(\xi)}{d\xi}\right)

1523: -\left(\c_{10}(\xi)-\xi\frac{d\c_{10}(\xi)}{d\xi}\right)

1524: +\frac{da_h(\xi)}{d\xi}\,.

1525: \eqlabel{1sts}

1526: \end{equation}

1527: %

1528: Notice that in the supersymmetric case there is no $\ln(\xi^2)$ dependence

1529: in the holographic free energy \eqref{fres0}.  As a result one can

1530: unambiguously determine $\calf_0$ directly from \eqref{efin}%

1531: %

1532: \footnote{The latter holographic prescription was formulated in such a way

1533: that the holographic energy of the supersymmetric PW flow vanishes \cite{hr}.}.

1534: %

1535: We find

1536: %

1537: \begin{equation}

1538: \calf_0\equiv\calf_0^{susy}=-\frac{\pi^2N^2 T^4}{8}\ e^{12a_h(\xi)}\ 40\xi^2=-\frac{5N^2}{288\pi^2}\ m^4\,.

1539: \eqlabel{fosusy}

1540: \end{equation}

1541: %

1542: 

1543: 

1544: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

1545: \section{Thermodynamics of the high temperature Pilch-Warner flow}

1546: \label{sec:thtpwf}

1547: 

1548: We now recall that the differential equations \eqref{beomx} describing the

1549: finite temperature PW renormalization group flow admit a perturbative

1550: analytical solution at high temperature \cite{bl,hr,bbs}. In this regime,

1551: the appropriate expansion parameters are

1552: %

1553: \begin{equation}

1554: \dd_1\propto \left(\frac{m_b}{T}\right)^2\ll1\,,

1555: \qquad \dd_2\propto \frac{m_f}{T}\ll 1\,.

1556: \eqlabel{larget}

1557: \end{equation}

1558: %

1559: In this case, the metric function $A(x)$ is given to first nontrivial order

1560: in $\dd_1$, $\dd_2$ by

1561: %

1562: \begin{equation}

1563: \begin{split}

1564: A(x)=&\ln\dd_3-\frac 14\ \ln \left(2x-x^2\right)+\dd_1^2\ A_1(x)+\dd_2^2\ A_2(x)\,,\\

1565: \a(x)=&\dd_1\ \a_1(x)\,,\\

1566: \c(x)=&\dd_2\ \c_2(x) \,,

1567: \end{split}

1568: \eqlabel{hightsol}

1569: \end{equation}

1570: %

1571: where

1572: %

1573: \begin{equation}

1574: \a_1=\left(2x -x^2\right)^{1/2}\ _2 F_1\left(\ft 12,\ft 12; 1; (1-x)^2\right)\,,

1575: \eqlabel{orko1}

1576: \end{equation}

1577: %

1578: \begin{equation}

1579: \c_2=(2x-x^2)^{3/4}\ _2F_1\left(\ft 34, \ft 34; 1; (1-x)^2\right)\,,

1580: \eqlabel{orko2}

1581: \end{equation}

1582: %

1583: \begin{equation}

1584: \begin{split}

1585: A_1=&4\int_x^1\ \frac{(z-1)dz}{(2z-z^2)^2}

1586: \left(\ga_1-\int_z^1dy\left(\frac{\del\a_1}{\del y}

1587: \right)^2\frac{(2y-y^2)^2}{y-1}\right)\,,\\

1588: A_2=&\frac 43\int_x^1\ \frac{(z-1)dz}{(2z-z^2)^2}

1589: \left(\ga_2-\int_z^1dy\left(\frac{\del\c_2}

1590: {\del y}\right)^2\frac{(2y-y^2)^2}{y-1}\right)\,.\\

1591: \end{split}

1592: \eqlabel{expsol}

1593: \end{equation}

1594: %

1595: The constants $\ga_i$ are fine-tuned to satisfy the boundary

1596: conditions, and are given by

1597: %

1598: \begin{equation}

1599: \ga_1=\frac{8-\pi^2}{2\pi^2}\,,\qquad \ga_2=\frac{8-3\pi}{8\pi}\,.

1600: \eqlabel{gai}

1601: \end{equation}

1602: %

1603: 

1604: Comparing \eqref{orko1}, \eqref{orko2}  and \eqref{bounexp1}, \eqref{bounexp2}

1605: we find the relations

1606: %

1607: \begin{equation}

1608: \begin{split}

1609: &\r_{11}=-\frac{\sqrt{2}}{\pi}\ \dd_1\,,\qquad \r_{10}=\frac{3\sqrt{2}\ \ln 2 }{\pi}\ \dd_1\,,\\

1610: &\c_0=\frac{2^{1/4}\sqrt{\pi}}{\Gamma\left(\frac 34\right)^2}\ \dd_2\,,\qquad \c_{10}=

1611: -\frac{2^{1/2}\Gamma\left(\frac 34\right)^4}{\pi^2} \,.

1612: \end{split}

1613: \eqlabel{rel6}

1614: \end{equation}

1615: %

1616: Also notice that 

1617: %

1618: \begin{equation}

1619: \begin{split}

1620: 2\pi T=&\dd_3\left(1+\dd_1^2\left(2-\frac{d^2A_1}{dx^2}\bigg|_{x=1}\right)+\dd_2^2\left(\frac 12

1621: -\frac{d^2A_2}{dx^2}\bigg|_{x=1}\right)\right)\\

1622: =&\dd_3\left(1+\frac{16}{\pi^2}\ \dd_1^2+\frac{4}{3\pi}\ \dd_2^2\right) \,.

1623: \end{split}

1624: \eqlabel{rel7}

1625: \end{equation}

1626: %

1627: The parameters $\{\dd_1,\dd_2\}$ can be related to the

1628: $\left\{m_b/T,m_f/T\right\}$ parameters of the dual gauge theory via

1629: %

1630: \begin{equation}

1631: \begin{split}

1632: \dd_1=&-\frac{1}{24\pi}\ \left(\frac{m_b}{T}\right)^2\,,\\

1633: \dd_2=&\frac{\left[\Gamma\left(\ft 34\right)\right]^2}{2\pi^{3/2}}

1634: \ \frac{m_f}{T}\,.

1635: \end{split}

1636: \eqlabel{ddphys}

1637: \end{equation}

1638: 

1639: 

1640: To leading order in mass deformation, the free energy \eqref{fres0} takes

1641: form \cite{hr,bbs}

1642: %

1643: \begin{equation}

1644: \begin{split}

1645: \calf=-\frac 18 \pi^2 N^2 T^4\left[ 1-\frac{192}{\pi^2}\ \ln(\pi T)\

1646: \delta_1^2-\frac{8}{\pi}\ \delta_2^2\right]\,.

1647: \end{split}

1648: \eqlabel{pert}

1649: \end{equation}

1650: %

1651: One can further evaluate the entropy density of the non-extremal PW geometry,

1652: %

1653: \begin{equation}

1654: s=\frac 12 \pi^2 N^2 T^3\left(1-\frac{48}{\pi^2}\ \delta_1^2-\frac{4}{\pi}\ \delta_2^2\right)\,,

1655: \end{equation}

1656: %

1657: and verify that the first law of thermodynamics \eqref{deff} is satisfied.

1658: 

1659: 

1660: 

1661: 

1662: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

1663: \section{Numerical procedure and results}

1664: \label{sec:np}

1665: 

1666: We are mainly interested in the thermodynamics of the PW flow away

1667: from the high temperature limit.  In this case, since analytic solutions

1668: to the equations of motion \eqref{beomx} are unavailable, we resort to

1669: numerical integration.  In this section we describe the numerical procedure

1670: and some of the consistency checks we have performed on the numerical

1671: results.  Extracting the actual thermodynamics from these solutions will be

1672: taken up in the following section.

1673: 

1674: The numerical procedure is conceptually the same for both the bosonic and 

1675: supersymmetric mass deformations of the non-extremal $AdS_5$ geometry.

1676: In practice, however, it is much easier to implement for the

1677: bosonic deformation

1678: %

1679: \begin{equation}

1680: m_f=0\,,\qquad m_b\ne 0\,,

1681: \eqlabel{mbos}   

1682: \end{equation}

1683: %

1684: and so we will describe the details of this first.  We then outline the

1685: modifications required for handling the supersymmetric mass deformation

1686: %

1687: \begin{equation}

1688: m_f=m_b\equiv m\,,

1689: \eqlabel{msus}

1690: \end{equation}

1691: %

1692: and comment on thermal flows corresponding to generic mass deformations. 

1693: Finally, we discuss the validity of the supergravity approximation. 

1694: 

1695: \subsection{Bosonic deformation}

1696: \label{subsec:bd}

1697: 

1698: Simplicity of treating the bosonic mass deformation stems from the

1699: fact that thermal PW flows allow for a consistent truncation with the

1700: supergravity scalar $\chi(x)$ identically set to zero.  The relevant

1701: equations can thus be obtained from 

1702: %

1703: \eqref{beomx} with

1704: \begin{equation}

1705: \chi(x)\equiv 0\,,\qquad \a\equiv \ln\r\,,\qquad \ln c_2(x)\equiv A(x)\equiv \ln\hdd_3-\frac 14\ln(2x-x^2)+a(x)\,.

1706: \eqlabel{bose0}

1707: \end{equation}

1708: %

1709: The boundary conditions (which guarantee that the flow is singularity-free)  

1710: are given by \eqref{bcax} and \eqref{nbc}

1711: \begin{equation}

1712: \begin{split}

1713: &x\to 0_+\qquad a(x)\to 0\,,\qquad \r(x)\to 1\,; \\

1714: &x\to 1_-\qquad a'(x)\to 0\,,\qquad \r'(x)\to 0\,.

1715: \end{split}

1716: \eqlabel{bose1}

1717: \end{equation}

1718: Asymptotic expansions for $\{\r(x),a(x)\}$ as $x\to 0_+$ are given by \eqref{bounexp1}, \eqref{bounexp3}. 

1719: The first couple terms are presented in \eqref{rr2}-\eqref{aa2}, where one has to set $\c_0=0$

1720: corresponding to $\chi(x)\equiv 0$, see \eqref{bounexp2}. 

1721: Notice that numerical integration of $\r(x)$ and $a(x)$ from $x=0_+$ is uniquely determined 

1722: specifying the non-normalizable  $\r_{11}$ and normalizable $\r_{10}$ mode coefficients.

1723: However, a generic choice of a pair $\{\r_{11},\r_{10}\}$ would produce a solution that does not 

1724: satisfy the horizon boundary condition (the $x\to 1_-$ boundary condition  in \eqref{bose1}).

1725: Solving Sturm-Liouville problem with boundary conditions \eqref{bose1} would determine

1726: \begin{equation}

1727: \r_{10}\equiv\r_{10}(\r_{11})\qquad \Longleftrightarrow \qquad \r_{10}\equiv \r_{10}(\xi) \,,

1728: \eqlabel{bose2} 

1729: \end{equation}  

1730: where on the RHS we used conventions of subsection \ref{subsec:tpwf}.

1731: For a flow satisfying \eqref{bose2} we can extract

1732: \begin{equation}

1733: a_h(\xi)=\lim_{x\to 1_-} a(x) \,.

1734: \eqlabel{bose3}

1735: \end{equation}

1736: To summarize, each  non-singular bosonic deformation flow would generate a triplet of numbers

1737: \begin{equation}

1738: \{\xi, \r_{10}(\xi), a_h(\xi)\}\,,

1739: \eqlabel{bose4}

1740: \end{equation}

1741: which can be used to evaluate  the free energy $\calf_{bosonic}$ \eqref{mff}, the speed of sound $c_{s,bosonic}^2$ \eqref{vs2} 

1742: and \eqref{speedb}, and verify (numerically) the first law of thermodynamics \eqref{1stb}. 

1743: 

1744: We use the following algorithm to generate triplets \eqref{bose4}.   

1745: First, we choose a mass deformation parameter $\r_{11}=\xi$ which we keep 

1746: fixed during the iteration procedure. The iteration starts by choosing a trial value of $\r_{10}$. 

1747: Given $\{\r_{11}, \r_{10}\}$ we integrate numerically%

1748: %

1749: \footnote{We used $Mathematica$ for numerical integration as well as our

1750: own $C$ code based on fifth-order Runge-Kutta.  Both procedures (up to

1751: controllable numerical errors) produced equivalent results.}

1752: %

1753: \eqref{beomx} (with \eqref{bose0}) from 

1754: $x_{initial}\ll 1$ to $x_{final}$ (such that $(1-x_{final})\ll 1$) using the power series \eqref{bounexp1},

1755: \eqref{bounexp2} to specify\footnote{In practice we truncated the power series to $k=3$ terms (inclusive), 

1756: and used $x_{initial}=10^{-9}$, $x_{final}=1-10^{-6}$.} 

1757: \begin{equation}

1758: \rho\left(x_{initial}\right)\,, \qquad \r'\left(x_{initial}\right)\,, \qquad a\left(x_{initial}\right)\,,\qquad a'\left(x_{initial}\right)\,.

1759: \end{equation}

1760: For a generic value of $\r_{10}$, 

1761: \begin{equation}

1762: \lim_{x\to 1_{-}}\rho\left(x\right)\to \pm \infty\,,\qquad \lim_{x\to 1_{-}}\a\left(x\right)\to -\infty\,.

1763: \end{equation} 

1764: We find that for a given value $\r_{11}=\xi$ there is  (generically) a unique $\r_{10}=\r_{10}(\xi)$, such that $\r(x)$ and $a(x)$ 

1765: are finite as we go to the horizon, $x\to 1_{-}$, \ie, 

1766: \begin{equation}

1767: \lim_{x\to 1_-}{\frac{d\r}{dx}(x; \r_{11}=\xi,\hr_{10}=\r_{10}(\xi))}=0\,,

1768: \eqlabel{hit}

1769: \end{equation}

1770: while 

1771: \begin{equation}

1772: \lim_{x\to 1_-}{\frac{d\r}{dx}(x; \r_{11}=\xi,\r_{10}\ne \r_{10}(\xi))}=\pm \infty\,.

1773: \eqlabel{nothit}

1774: \end{equation}

1775: Now, the iterative algorithm is clear: we use Newton's shooting method of determining $\r_{10}(\xi)$ for a fixed $\r_{11}=\xi$. 

1776: 

1777: Results of the numerical analysis are presented in Figs.~\ref{figb1}-\ref{figb3}. 

1778: From Fig.~\ref{figb1} notice that for $\r_{11}> \r_{crit}\approx 0.03582$ there is no solution to Sturm-Liouville problem with boundary 

1779: condition \eqref{bose1}. Using \eqref{match3} we find that corresponding to $\r_{crit}$,

1780: \begin{equation}

1781: \frac{m_b}{T}\bigg|_{crit}\approx2.29(9)\,.

1782: \eqlabel{mbcrit}

1783: \end{equation} 

1784: We find that for each value of $\xi \in (0,\r_{crit})$ there are two sets $\{\r_{10}(\xi),a_h(\xi)\}$ which are represented 

1785: by red/blue segments of the plots in  Fig.~\ref{figb1}. For each $\xi<0$ we find a single set  $\{\r_{10}(\xi),a_h(\xi)\}$ 

1786: represented by black segments of the plots in Fig.~\ref{figb1}.

1787: In section \ref{sec:r} we show that red/blue values of the bosonic condensate $\r_{10}(\xi)$ correspond to

1788: different phases of the $\caln=2^*$ plasma.  We also show that there is $\xi=\rho^*<\r_{crit}$ such that 

1789: for all $\xi>\r^*$,   $\caln=2^*$ plasma becomes unstable both thermodynamically and dynamically, in agreement 

1790: with general  arguments of \cite{binst}.

1791: 

1792: 

1793: 

1794: 

1795: \begin{figure}[t]

1796:   \hspace*{-20pt}

1797: \psfrag{rho11}{\raisebox{1ex}{\footnotesize\hspace{-0.8cm}$\qquad \xi$}}

1798: \psfrag{ah}{\raisebox{0.0ex}{$a_h(\xi)$}}

1799: \psfrag{rc}{\raisebox{0.0ex}{$\r_{crit}$}}

1800: \psfrag{rho10}{\raisebox{0.5ex}{$\r_{10}(\xi)$}}

1801:   \includegraphics[width=3.4in]{rhob.eps}

1802: %  \hspace*{-10pt}

1803:   \includegraphics[width=3.4in]{ahb.eps}%

1804: %  \hspace*{-20pt}

1805: %  \vspace*{-10pt}

1806:   \caption{

1807: Altogether we obtained 3227 triplets $\{\xi,\r_{10}(\xi),a_h(\xi)\}$.  

1808: The left plot represents $\r_{10}(\xi)$, and the right plot represents $a_h(\xi)$. The critical value of $\xi$ is at

1809: $\r_{crit}\approx 0.03582(2)$. Red segments of the plots correspond to $\xi\in[0,\r_{crit})$ such that $\r_{10}(0)=a_h(0)=0$.

1810: Blue segments of the plots do not pass through the origin and  correspond to  $0<\xi <\r_{crit}$. Black segments 

1811: of the plots correspond to $\xi<0$. 

1812: }

1813: \label{figb1}

1814: \end{figure}

1815: 

1816: Figs.~\ref{figb2}-\ref{figb3} represent $\r_{10}'(\xi)$, $a_h'(\xi)$, and a numerical verification of the first law of thermodynamics 

1817: \eqref{1stb}. Since our numerical data is rather densely spaced\footnote{Mostly we have $\Delta\xi=10^{-4}$.} in $\xi$, we use 

1818: \begin{equation}

1819: f'(\xi)=\frac{df(\xi)}{d\xi}=\frac{f(\xi+\Delta\xi)-f(\xi-\Delta\xi)}{2\Delta\xi}\,,

1820: \eqlabel{derf}

1821: \end{equation}

1822: as a numerical approximation for a derivative of a function $f(\xi)$.

1823: Notice that  

1824: \begin{equation}

1825: \lim_{\xi\to \r_{crit}-0}\frac{d\r_{10}(\xi)}{d\xi}=\mp \infty\,,\qquad \lim_{\xi\to \r_{crit}-0}\frac{da_h(\xi)}{d\xi}=\mp \infty\,,

1826: \eqlabel{rdercr}

1827: \end{equation}

1828: with minus/plus signs for red/blue segments of the curves in Fig.~\ref{figb2}/Fig.~\ref{figb3} correspondingly.

1829: Also, from Fig.~\ref{figb3}

1830: \begin{equation}

1831: \lim_{\xi\to +0}\frac{d\r_{10}(\xi)}{d\xi}=+\infty\,,\qquad \lim_{\xi\to +0}\frac{da_h(\xi)}{d\xi}=+\infty\,.

1832: \eqlabel{rder0}

1833: \end{equation}

1834: We find that 

1835: \begin{equation}

1836: \r_{10}'(\xi)\bigg|_{\xi=+2\times 10^{-4}}\approx -2.10(6)\,,

1837: \eqlabel{der}

1838: \end{equation}

1839: which is rather close to a predicted value from the high temperature analysis of section \ref{sec:thtpwf} (see \eqref{rel6})

1840: \begin{equation}

1841: \r_{10}'(\xi)\bigg|_{\xi=0}^{high\ temperature}=-3\ln 2\approx -2.07944\,.

1842: \eqlabel{predr10}

1843: \end{equation}

1844: A better agreement is achieved if we fit the first 200 points of the red segment of  $\r_{10}(\xi)$ 

1845: on the left plot in Fig.~\ref{figb1} with a 10th order polynomial in $\xi$. We find in this case 

1846: \begin{equation}

1847: \r_{10}'(\xi)\bigg|_{\xi=0}^{fit}\approx -2.07972\,.

1848: \eqlabel{fitr10}

1849: \end{equation}

1850: The green points on the right plots in Figs.~\ref{figb2}-\ref{figb3} demonstrate the cancellation of $a'_h(\xi)$ with a combination

1851: $$

1852: 4\xi+2\left(\r_{10}(\xi)-\xi \frac{d\r_{10}(\xi)}{d\xi}\right)\,,

1853: $$

1854: as required by the first law of thermodynamics, see \eqref{1stb}. We find that such a cancellation is achieved with an accuracy of better 

1855: than $10^{-4}$. Clearly, this provides a rather impressive check on both the validity of the holographic renormalization 

1856: of the thermal PW flows explained in \cite{hr}, and the numerical procedures developed in this paper.  

1857: 

1858: 

1859: 

1860: \begin{figure}[t]

1861:   \hspace*{-20pt}

1862: \psfrag{rho11}{\raisebox{1ex}{\footnotesize\hspace{-0.8cm}$\qquad \xi$}}

1863: \psfrag{dah}{\raisebox{0.0ex}{$a_h'(\xi)$}}

1864: \psfrag{rc}{\raisebox{0.0ex}{$\r_{crit}$}}

1865: \psfrag{drho10}{\raisebox{0.5ex}{$\r_{10}'(\xi)$}}

1866: \includegraphics[width=3.4in]{drhob.eps}

1867: %  \hspace*{-10pt}

1868:   \includegraphics[width=3.4in]{dahb.eps}%

1869: %  \hspace*{-20pt}

1870: %  \vspace*{-10pt}

1871:   \caption{

1872: The left plot represents $\r_{10}'(\xi)$ for both the red and the black segments of the left plot in Fig.~\ref{figb1}.

1873: The right plot represents $a_h'(\xi)$ for both the red and the black segments of the right plot in Fig.~\ref{figb1}.

1874: The green points on the right plot represent numerical verification of the first law of the thermodynamics, 

1875: where $a_h'(\xi)$ must be canceled with an appropriate combination of $\{\xi,\r_{10}(\xi),\r'(\xi)\}$, see \eqref{1stb}.  

1876: }

1877: \label{figb2}

1878: \end{figure}

1879: 

1880: 

1881: 

1882: \begin{figure}[t]

1883:   \hspace*{-20pt}

1884: \psfrag{rho11}{\raisebox{1ex}{\footnotesize\hspace{-0.8cm}$\qquad \xi$}}

1885: \psfrag{dah}{\raisebox{0.0ex}{$a_h'(\xi)$}}

1886: \psfrag{rc}{\raisebox{0.0ex}{$\r_{crit}$}}

1887: \psfrag{drho10}{\raisebox{0.5ex}{$\r_{10}'(\xi)$}}

1888: \includegraphics[width=3.4in]{drhobd.eps}

1889: %  \hspace*{-10pt}

1890:   \includegraphics[width=3.4in]{dahbd.eps}%

1891: %  \hspace*{-20pt}

1892: %  \vspace*{-10pt}

1893:   \caption{

1894: The left plot represents $\r_{10}'(\xi)$ for  the blue segment of the left plot in Fig.~\ref{figb1}.

1895: The right plot represents $a_h'(\xi)$ for the blue segment of the right plot in Fig.~\ref{figb1}.

1896: The green points on the right plot represent numerical verification of the first law of the thermodynamics, 

1897: where $a_h'(\xi)$ must be canceled with an appropriate combination of $\{\xi,\r_{10}(\xi),\r'(\xi)\}$, see \eqref{1stb}. 

1898: }

1899: \label{figb3}

1900: \end{figure}

1901: 

1902: 

1903: \subsection{Supersymmetric deformation}

1904: Albeit more complicated, the Sturm-Liouville problem for the supersymmetric mass deformation is conceptually 

1905: similar to the one for the bosonic mass deformation studied in previous subsection. Here we have to use all 

1906: equations in \eqref{beomx}. 

1907: As for the bosonic deformation, we rewrite these equations using $\{\r(x),a(x)\}$ such that 

1908: \begin{equation}

1909: \a\equiv \ln\r\,,\qquad \ln c_2(x)\equiv A(x)\equiv \ln\hdd_3-\frac 14\ln(2x-x^2)+a(x)\,.

1910: \eqlabel{susy0}

1911: \end{equation}

1912: The boundary conditions (which guarantee that the flow is singularity-free)  

1913: are given by \eqref{bcax} and \eqref{nbc}

1914: \begin{equation}

1915: \begin{split}

1916: &x\to 0_+\qquad a(x)\to 0\,,\qquad \r(x)\to 1\,,\qquad \c(x)\to 0\,; \\

1917: &x\to 1_-\qquad a'(x)\to 0\,,\qquad \r'(x)\to 0\,,\qquad \c'(x)\to 0\,.

1918: \end{split}

1919: \eqlabel{susy1}

1920: \end{equation}

1921: For the bosonic mass deformation we have a one-dimensional Sturm-Liouville problem: given $\r_{11}$, 

1922: the boundary condition at the horizon determines $\r_{10}$; the pair $\{\r_{11}\equiv \xi,\r_{10}(\xi)\}$ 

1923: further determines the value of the $a(x)$ at the horizon, $a_h(\xi)$.  For the supersymmetric 

1924: mass deformation the Sturm-Liouville problem is two-dimensional: given $\r_{11}\equiv \xi$, 

1925: the boundary condition \eqref{susy0} would determine

1926: \begin{equation}

1927: \r_{10}\equiv \r_{10}(\xi)\,,\qquad \c_{10}\equiv \c_{10}(\xi)\,.

1928: \eqlabel{susy2}

1929: \end{equation}

1930: For a flow satisfying \eqref{susy2} we can extract 

1931: \begin{equation}

1932: a_h(\xi)=\lim_{x\to 1_-} a(x)\,.

1933: \eqlabel{susy3}

1934: \end{equation} 

1935: Thus, each non-singular supersymmetric mass deformation flow would generate a set of four 

1936:  numbers 

1937: \begin{equation}

1938: \{\xi,\r_{10}(\xi),\chi_{10}(\xi),a_h(\xi)\}\,,

1939: \eqlabel{susy4}

1940: \end{equation}

1941: which can be used to evaluate the free energy $\calf_{susy}$ \eqref{mfs}, the speed of sound 

1942: $c_{s,susy}^2$ \eqref{vs2} and \eqref{speeds}, and verify (numerically) the first law of thermodynamics 

1943: \eqref{1sts}.

1944: 

1945: 

1946: 

1947: 

1948: 

1949: \begin{figure}[t]

1950:   \hspace*{-20pt}

1951: \psfrag{rho11}{\raisebox{1ex}{\footnotesize\hspace{-0.8cm}$\qquad \xi$}}

1952: \psfrag{ah}{\raisebox{0.0ex}{$a_h(\xi)$}}

1953: \psfrag{chi10}{\raisebox{0.5ex}{$\c_{10}(\xi)$}}

1954: \psfrag{rho10}{\raisebox{0.5ex}{$\r_{10}(\xi)$}}

1955:   \includegraphics[width=3.4in]{rhos.eps}

1956: %  \hspace*{-10pt}

1957:   \includegraphics[width=3.4in]{chis.eps}%

1958: %  \hspace*{-20pt}

1959: %  \vspace*{-10pt}

1960:   \caption{

1961: Altogether we obtained 2712 sets $\{\xi,\r_{10}(\xi),\chi(\xi),a_h(\xi)\}$.  

1962: The left plot represents $\r_{10}(\xi)$, and the right plot represents $\chi_{10}(\xi)$. 

1963: }

1964: \label{figs1}

1965: \end{figure}

1966: 

1967: 

1968: Since the Sturm-Liouville problem is technically more complex here, we restrict our discussion to 

1969: $\xi\ge 0$, or following \eqref{match2} $m^2\ge 0$. Results of the numerical analysis are presented in 

1970: Figs.~\eqref{figs1}-\eqref{figs2}. Fitting the first 200 points of $\r_{10}(\xi)$ and $\c_{10}(\xi)$

1971: with a 10th order polynomial in $\xi$ we find 

1972: \begin{equation}

1973: \r'_{10}(\xi)\bigg|_{\xi=0}^{fit}=-2.069(2)\,,\qquad \c_{10}\bigg|_{\xi=0}^{fit}=-0.3219(7)\,,

1974: \eqlabel{fits}

1975: \end{equation}

1976: which should be compared with the high temperature predictions\footnote{We reduce the accuracy compare to the 

1977: bosonic mass deformation case in order to speed-up computations. For a selected set of $41$ sets 

1978: \eqref{susy4} we improve accuracy by roughly a factor of $10^2$. More accurate data does not provide 

1979: noticeable quantitative difference.} \eqref{rel6}

1980: \begin{equation}

1981: \begin{split}

1982: \r'_{10}(\xi)\bigg|_{\xi=0}^{high\ temperature}&=-3\ln 2\approx -2.07944\,,\\ 

1983: \c_{10}(\xi)\bigg|_{\xi=0}^{high\ temperature}&=-\frac{2^{1/2}\Gamma\left(\frac 34\right)^4}{\pi^2}\approx 

1984: -0.3231(1)\,.

1985: \end{split}

1986: \eqlabel{predsusy}

1987: \end{equation}

1988: We do not find a signature for the existence of $\r_{crit}$ --- it appears for supersymmetric 

1989: deformation the boundary value problem \eqref{susy1} always have a solution\footnote{Recall for a bosonic 

1990: deformation $\r_{crit}\approx -0.0358(2)$ which is well inside the set of $\xi$ considered in Fig.~\ref{figs1}.}.

1991: 

1992: 

1993: 

1994: \begin{figure}[t]

1995:   \hspace*{-20pt}

1996: \psfrag{rho11}{\raisebox{1ex}{\footnotesize\hspace{-0.8cm}$\qquad \xi$}}

1997: \psfrag{ah}{\raisebox{0.0ex}{$a_h(\xi)$}}

1998: \psfrag{chi10}{\raisebox{0.5ex}{$\c_{10}(\xi)$}}

1999: \psfrag{rho10}{\raisebox{0.5ex}{$\r_{10}(\xi)$}}

2000:   \includegraphics[width=3.4in]{ahs.eps}

2001: %  \hspace*{-10pt}

2002:   \includegraphics[width=3.4in]{firsts.eps}%

2003: %  \hspace*{-20pt}

2004: %  \vspace*{-10pt}

2005:   \caption{

2006: The left plot represents $a_h(\xi)$ for the supersymmetric mass deformation.

2007: The red/blue/black points on the right plot represent $\a_h'(\xi)$, $2(\r_{10}-\xi \r'_{10}(\xi))$

2008: and $-(\c_{10}(\xi)-\xi\c_{10}'(\xi))$ correspondingly. The green points on the right plot 

2009: represent numerical verification of the first law of the thermodynamics, see \eqref{1sts}.

2010: }

2011: \label{figs2}

2012: \end{figure}

2013: 

2014: Fig.~\ref{figs2} represent $a_h(\xi)$ (the left plot) and a numerical verification of the first law of thermodynamics 

2015: \eqref{1sts} (the right plot):

2016: \begin{equation}

2017: \begin{split}

2018: \biggl\{0\biggr\}_{green}=&\biggl\{2\left(\r_{10}(\xi)-\xi\frac{d\r_{10}(\xi)}{d\xi}\right)\biggr\}_{blue}+

2019: \biggl\{-\left(\c_{10}(\xi)-\xi\frac{d\c_{10}(\xi)}{d\xi}\right)\biggr\}_{black}\\

2020: &+\biggl\{\frac{da_h(\xi)}{d\xi}

2021: \biggr\}_{red}\,,

2022: \end{split}

2023: \eqlabel{1stscolor}

2024: \end{equation}

2025: where subscripts refer to the color of the points on the right plot in Fig.~\ref{figs2}.

2026: We find that \eqref{1stscolor} holds with an accuracy of $10^{-3}$.

2027: 

2028: 

2029: 

2030: \subsection{Generic deformation}

2031: Though in this paper we study two special cases of the thermal PW flows --- one corresponding to 

2032:  a bosonic mass deformation where the supergravity scalar $\c(x)\equiv 0$, and the other one 

2033: corresponding to a supersymmetric mass deformation where the ratio of the coefficients

2034: of the non-normalizable modes of the supergravity scalars $\{\r(x),\c(x)\}$ is the same as for the 

2035: supersymmetric PW flow  --- the numerical methods developed here can be applied for a generic 

2036: mass deformation as well. In fact, the Sturm-Liouville problem one has to solve in a generic case is exactly the same 

2037: as in the supersymmetric case where the  boundary conditions are given by \eqref{susy1}. The only difference from the  

2038: supersymmetric mass deformation case is the (fixed) relation between the  coefficients of the non-normalizable modes of the 

2039: supergravity scalars $\{\r(x),\c(x)\}$ given by \eqref{not1a}

2040: $$

2041: \eta^2=6\ \frac{m_f^2}{m_b^2}\ \xi\,.

2042: $$

2043: 

2044: Recall (see Fig.\ref{figb1}) that when $m_f=0$ there is a critical value $\r_{crit}$, such that for 

2045: $\xi> \r_{crit}$ there is no nonsingular thermal PW flow. On the other hand (see Fig.\ref{figs1}), we do not find the  evidence for 

2046: $\r_{crit}$ for the supersymmetric mass deformation $m_f=m_b$. Thus we expect that 

2047: \begin{equation}

2048: \r_{crit}\equiv \r_{crit}\left(\Delta\equiv \frac{m_f^2}{m_b^2}\right)\,,

2049: \eqlabel{rcritdep} 

2050: \end{equation}

2051: such that $\r_{crit}\to \infty$ as $\frac{m_f^2}{m_b^2}\to \Delta^*\le 1$. 

2052: As we show in subsection \ref{subsec:tpwb} the existence of $\r_{crit}$ for $m_f=0$ implies destabilization of the 

2053: thermal plasma at low temperatures. What we appear to be finding is that this particular instability 

2054: is cured by turning on $m_f\ne 0$. 

2055: Verifying the latter conjecture is beyond the scope of this paper.  However, we would like to point out a different type of instability 

2056: in a system closely related to the one studies here which is eliminated by  turning on sufficiently large $m_f$. 

2057: Consider thermal PW flow with the flat $R^3$ in \eqref{ab} compactified on a three torus\footnote{Since a 

2058: three torus is locally the same as $R^3$,

2059: such a compactification does not change any of the above computations.} $T^3$. 

2060: This represents holographic dual to thermal $\caln=2^*$ plasma on finite volume $T^3$. 

2061: Analysis identical to that in \cite{bninst} shows  that this system is generically non-perturbatively unstable 

2062: with respect to a spontaneous creation of the $D3\overline{D3}$ brane pairs\footnote{Similar instability 

2063: was first identified in \cite{mm}.}. This instability is suppressed provided\footnote{See 

2064: eq.(4.16) of \cite{bninst}

2065: where one has to remove $(-2)$ from the inequality (coming from the curvature of the 3-space) and replace $\r_{11}$

2066: of \cite{bninst} with $4\r_{11}$ due to a different choice of a radial coordinate.} 

2067: \begin{equation}

2068: \begin{split}

2069: \c_0^2\ge 6\r_{11}\ge 0\qquad \Longleftrightarrow\qquad \eta^2\ge 6\xi\ge 0\qquad \Rightarrow\qquad  \Delta\equiv \frac{m_f^2}

2070: {m_b^2}\ge 1     \,.

2071: \end{split}

2072: \eqlabel{qn1}

2073: \end{equation}

2074: So, for the nonperturbative instability due to $D3\overline{D3}$ spontaneous brane pair creation for the thermal PW flow  

2075: compactified on $T^3$ we find that $\Delta^*=1$.

2076: 

2077: 

2078: \subsection{Validity of the supergravity approximation}

2079: \label{subsec:vsa}

2080: 

2081: \begin{figure}[t]

2082:   \hspace*{-20pt}

2083: \psfrag{rho11}{\raisebox{1ex}{\footnotesize\hspace{-0.8cm}$\qquad \xi$}}

2084: \psfrag{k}{\raisebox{0.0ex}{$\k(\xi)$}}

2085: \psfrag{rc}{\raisebox{0.5ex}{$\r_{crit}$}}

2086:   \includegraphics[width=3.4in]{cuvb.eps}

2087: %  \hspace*{-10pt}

2088:   \includegraphics[width=3.4in]{cuvs.eps}%

2089: %  \hspace*{-20pt}

2090: %  \vspace*{-10pt}

2091:   \caption{

2092: The left plot represents ratio of thermal PW geometry horizon curvature to 

2093: its curvature at the boundary for bosonic mass deformation. The right plot represents 

2094: this ratio for the supersymmetric mass deformation. 

2095: }

2096: \label{figcuv}

2097: \end{figure}

2098: 

2099: 

2100: 

2101: Our analysis of the holographic dual to thermal $\caln=2^*$ gauge theory plasmas are done in the supergravity 

2102: approximation to type IIb string theory. In the planar limit considered here the string loop corrections

2103: are suppressed and the supergravity approximation  is valid provided curvature invariants of the thermal 

2104: PW flow geometry \eqref{ab} are small. We find that along the flow from the boundary to the horizon

2105: the absolute value of the Ricci scalar 

2106: of the five dimensional thermal PW geometry \eqref{ab} increases. As a quantitative criteria for such 

2107: an increase we propose the ratio of the scalar curvature at the horizon to the scalar curvature near the boundary

2108: \begin{equation}

2109: \kappa\equiv\frac{R^{horizon}}{R^{boundary}}=\frac{\calp^{horizon}}{\calp^{boundary}}=-\frac 43\ \calp^{horizon}\,,

2110: \eqlabel{valid} 

2111: \end{equation}

2112: where we used \eqref{ee} and the fact that $\calp^{boundary}=-\frac 34$.

2113: Values\footnote{In order to evaluate $\calp^{horizon}$ besides $a_h(\xi)$ we retained 

2114: horizon values of the supergravity scalars $\{\r(x),\c(x)\}$ while solving the Sturm-Liouville problem 

2115: \eqref{bose1} or \eqref{susy1}. } of $\kappa(\xi)$ for both bosonic and supersymmetric mass deformations are presented in Fig.~\ref{figcuv}.

2116: Notice that for the 'blue' phase 

2117: \begin{equation}

2118: \lim_{\xi\to 0+}\k^{bosonic}_{blue}(\xi)=+\infty\,,

2119: \eqlabel{sugrabreak}

2120: \end{equation}

2121: and thus supergravity approximation for the blue phase of the bosonic mass deformation for small 

2122: values of $\xi$ is not valid. 

2123: 

2124: 

2125: 

2126: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

2127: \section{$\caln=2^*$ plasma at finite temperature}

2128: \label{sec:r}

2129: 

2130: In this section we discuss thermodynamics of the $\caln=2^*$ gauge theory plasma for two special cases:

2131: \nxt $bosonic$ deformation, when $m_f=0$ and $m_b\ne 0$;

2132: \nxt $supersymmetric$ deformation, when $m_f=m_b=m$.

2133: 

2134: 

2135: \subsection{Thermal PW flows with $m_f=0$, $m_b\ne 0$}

2136: \label{subsec:tpwb}

2137: 

2138: 

2139: \begin{figure}[t]

2140:   \hspace*{-20pt}

2141: \psfrag{mbT2}{\raisebox{2ex}{\footnotesize\hspace{-0.5cm}$\qquad  \mu\equiv \frac{m_b^2}{T^2}$}}

2142: \psfrag{mc2}{\raisebox{0.0ex}{$\mu_{crit}$}}

2143: \psfrag{Fred}{\raisebox{0.5ex}{$\calf_{bosonic}/\left(\frac 18 \pi^2 N^2 T^4\right)\ \qquad\ $}}

2144: \psfrag{vsb2}{\raisebox{0.5ex}{$3\ \times\ c_{s,bosonic}^2$}}

2145:   \includegraphics[width=3.4in]{freeb.eps}

2146: %  \hspace*{-10pt}

2147:  \includegraphics[width=3.4in]{soundb1.eps}%

2148: %  \hspace*{-20pt}

2149: %  \vspace*{-10pt}

2150:   \caption{

2151: The left plot represents bosonic mass deformation free energy $\calf_{bosonic}$ as 

2152: a function of $\mu\equiv \frac{m_b^2}{T^2}$. The critical mass is at $\mu_{crit}\approx 5.28(5)$. 

2153: The right plot represents speed of sound squared $c_{s,bosonic}^2$ for 

2154: bosonic mass deformation as a function of $\mu$.   Recall that the speed of sound in the conformal plasma 

2155: $c_{s,CFT}^2=\frac 13$.

2156: }

2157: \label{figfreeb}

2158: \end{figure}

2159: 

2160: \begin{figure}[t]

2161:   \hspace*{80pt}

2162: \psfrag{mbT2}{\raisebox{2ex}{\footnotesize\hspace{-0.5cm}$\qquad  \mu\equiv \frac{m_b^2}{T^2}$}}

2163: \psfrag{mc2}{\raisebox{0.0ex}{$\mu_{crit}$}}

2164: \psfrag{ms2}{\raisebox{0.0ex}{$\ \ \mu^*$}}

2165: \psfrag{Fred}{\raisebox{0.5ex}{$\calf_{bosonic}/\left(\frac 18 \pi^2 N^2 T^4\right)\ \qquad\ $}}

2166: \psfrag{vsb2}{\raisebox{0.5ex}{$3\ \times\ c_{s,bosonic}^2$}}

2167: %  \includegraphics[width=3.4in]{freeb}

2168: %  \hspace*{-10pt}

2169:  \includegraphics[width=3.4in]{soundb2.eps}%

2170: %  \hspace*{-20pt}

2171: %  \vspace*{-10pt}

2172:   \caption{

2173: More detailed plot of the speed of sound squared $c_{s,bosonic}^2$ for 

2174: bosonic mass deformation as a function of $\mu$. The speed of sound vanishes at $\mu^*\approx 5.40(9)$.

2175: Notice that though $\mu^*>\mu_{crit}$, corresponding to $\mu^*$ value of $\xi^*\approx 0.035(1)$ is less than 

2176: $\r_{crit}\approx 0.03582(2)$,

2177: see Fig.~\ref{figb1}.    Recall that the speed of sound in the conformal plasma $c_{s,CFT}^2=\frac 13$.

2178: }

2179: \label{figsoundb}

2180: \end{figure}

2181: 

2182: As explained in subsection \ref{subsec:bd}, for the bosonic deformation we observe two phases (`red' phase and 

2183: `blue' phase) of the 

2184: thermal gauge theory plasma for mass deformation parameter $\xi \in (0,\r_{crit})$ 

2185: characterized by two different values for the bosonic condensate $\r_{10}(\xi)$. 

2186: These are the red and the blue sets of points in Fig.~\ref{figb1}.

2187: For large values of the bosonic deformation parameter ($\xi>\r_{crit}$ in Fig.~\ref{figb1})

2188: thermal $\caln=2^*$ gauge theory plasma with bosonic deformation discussed does not exist.  

2189: The set of black points in Fig.~\ref{figb1} corresponds to $\xi<0$, which following \eqref{match3}

2190: corresponds to $m_b^2<0$. As we argued in subsection \ref{subsec:vsa}, we can trust supergravity approximation for 

2191: the red set of points (everywhere) and for black points over the range we studies (corresponding to $m_b^2/T^2<-15$).

2192: The supergravity approximation for the blue set of point is much less reliable\footnote{The smallest absolute value 

2193: horizon curvature for the blue points is the same as the largest horizon curvature for the red 

2194: points.}, especially for small values of $\xi$.

2195: 

2196: We use \eqref{match3}, \eqref{vs2}, \eqref{mff} and \eqref{speedb} to evaluate the free energy density and the speed of sound 

2197: for the bosonic mass deformation. For the free energy density we set an overall (arbitrary) constant part of the free 

2198: energy\footnote{Such ambiguity is present both in thermal quantum field theory and in the framework of the 

2199: holographic renormalization of the string theory duals to thermal gauge theories.} 

2200: \begin{equation}

2201: \calf_{0}^{bosonic}=0\,.

2202: \eqlabel{f00}   

2203: \end{equation}

2204: Notice that while in the supersymmetric case one can determine $\calf_{0}^{susy}$ unambiguously (see \eqref{fosusy})

2205: by requiring the $\calf_{susy}(T=0)=0$, we can not do so for the bosonic deformation: as we describe shortly, 

2206: the bosonic deformation becomes unstable both thermodynamically and dynamically at low temperatures.

2207: Results of the analysis are presented in Figs.~\ref{figfreeb}-\ref{figsoundb}.

2208: 

2209: The left plot in Fig.~\ref{figfreeb} describes the free energy density for the bosonic deformation $\calf_{bosonic}$

2210: relative (up to a sign) to the free energy density of the conformal gauge theory plasma

2211: \begin{equation}

2212: \calf_{CFT}=-\frac 18\ \pi^2 N^2 T^4\,,

2213: \eqlabel{fcft}

2214: \end{equation} 

2215: as a function of 

2216: \begin{equation}

2217: \mu\equiv \frac{m_b^2}{T^2}\,.

2218: \eqlabel{defmu}

2219: \end{equation}

2220: Notice that the blue phase has always a larger (or equal) free energy than the red phase

2221: \begin{equation}

2222: \calf_{bosonic,red}(\mu)\le \calf_{bosonic,blue}(\mu)\,,

2223: \end{equation}

2224: with equality achieved only at $\mu=\mu_{crit}$ corresponding to $\r_{crit}$ according to \eqref{match3}, \eqref{defmu}.

2225:  For the red phase

2226: \begin{equation}

2227: \frac{\calf_{bosonic,red}(\mu=0)}{\ft 18\pi^2 N^2 T^4}=-1\,,

2228: \eqlabel{red0}  

2229: \end{equation}

2230: in agreement with the fact that for $m_b=0$ (or at very high temperatures $T\to \infty$)

2231: the free energy density must be that of the conformal gauge theory. 

2232: 

2233: The right plot in Fig.~\ref{figfreeb} and a more detailed plot in Fig.~\ref{figsoundb}

2234: present the speed of sound squared for the thermal $\caln=2^*$ gauge theory plasma, relative to the speed of sound 

2235: in the conformal gauge theory plasma

2236: \begin{equation}

2237: c_{s,CFT}^2=\frac 13\,,

2238: \eqlabel{cscft}

2239: \end{equation}

2240: for bosonic deformation as a function of $\mu$.

2241: Although bosonic deformation phase discussed here does not exist for $\xi>\r_{crit}$, 

2242: bosonic phase actually extends to values of $\mu_{crit}<\mu \le \mu^*$ (see Fig.~\ref{figsoundb}). 

2243: The reason for this is that relation between $\xi$ and $\mu$ as given by \eqref{match3} involves 

2244: a factor $e^{6a_h(\xi)}$ and is not monotonic around  $\xi\sim \r_{crit}$. A value of 

2245: $\xi^*$ corresponding to $\mu^*$ is actually less than $\r_{crit}$: 

2246: \begin{equation}

2247: \mu^* \approx 5.40(9)\ >\ \mu_{crit}\approx 5.28(5)\qquad \Rightarrow\qquad \xi^*\approx 0.035(1)\ <\ \r_{crit}\approx 0.03582(2) \,.

2248: \eqlabel{msv}

2249: \end{equation}

2250: We find that the speed of sound for the blue phase is superlunimal, which is yet another indication that this 

2251: phase is unphysical. The speed of sound diverges for the blue phase as one approach $\mu_{crit}$

2252: either from above or below. The speed of sound for the red phase at $\mu=0$ is that as in the CFT plasma;

2253: it vanishes at $\mu=\mu^*$ and becomes purely imaginary for $\xi^*<\xi< \r_{crit}$ (still for the red phase

2254: and well within supergravity approximation) 

2255: \begin{equation}

2256: \begin{split}

2257: &c_{s,bosonic,red}^2(\mu=0)=\frac 13\,,\qquad c_{s,bosonic,red}^2(\mu=\mu^*)=0\,,\\

2258: &c_{s,bosonic,red}^2\left(\mu\equiv 12\sqrt{2}\pi^2e^{6a_h(\xi)}\xi\right)< 0\,,\qquad \xi^*<\xi<\r_{crit}\,.

2259: \end{split}

2260: \eqlabel{csbos}

2261: \end{equation}

2262: We would like to stress that conclusions \eqref{csbos} are very robust and in fact, immediately follow from the 

2263: existence of $\r_{crit}$ for the bosonic condensate $\r_{10}(\xi)$. 

2264: Indeed, existence of $\r_{crit}$, \ie,\ impossibility to determine $\r_{10}(\xi)$ for $\xi>\r_{crit}$,

2265: along with the assumption for the smoothness of $\r_{10}(\xi)$, implies that $\rho_{10}(\xi)$ is not a single valued 

2266: function of $\xi$ around $\r_{crit}$; moreover, 

2267: \begin{equation}

2268: \lim_{\xi\to \r_{crit}-0}\r_{10}'(\xi)=\mp \infty\,,

2269: \eqlabel{bin1}

2270: \end{equation}

2271: with the minus sign in the red phase and the plus sign in the blue phase (precisely what is presented in Fig.~\ref{figb1}).

2272: But then from \eqref{speedb}

2273: \begin{equation}

2274: \lim_{\xi\to \r_{crit}-0}\Sigma_{bosonic}(\xi)=\mp \infty\,,

2275: \eqlabel{bin2}

2276: \end{equation}

2277: again with the minus sign in the red phase and the plus sign in the blue phase.

2278: Thus using \eqref{vs2} we see that in the red phase 

2279: \begin{equation}

2280: \lim_{\xi\to \r_{crit}-0}\biggl(3\ \times\ c_{s,bosonic}^2(\xi)\biggr)=-3\,,

2281: \eqlabel{bin3}

2282: \end{equation}

2283: (in agreement with Figs.~\ref{figfreeb}-\ref{figsoundb}), and for the blue phase 

2284: \begin{equation}

2285: \lim_{\xi\to \xi_{\infty}\mp 0}\biggl(3\ \times\ c_{s,bosonic}^2(\xi)\biggr)=\pm \infty\,,\qquad \Sigma_{bosonic,blue}(\xi_{\infty})=3\,,

2286: \eqlabel{bin4}

2287: \end{equation}

2288: where $0<\xi_{\infty}<\r_{crit}$.  Since in the red phase at $\xi=0$ the speed of sound squared is $\frac 13>0$

2289: and given \eqref{bin3} there is a guaranteed value of $\xi^*<\r_{crit}$ at which the speed of sound vanishes, 

2290: and for $\xi^*<\xi<\r_{crit}$ the speed of sound is purely imaginary. 

2291: As emphasized in \cite{binst}, 

2292: a thermal system with $c_s^2<0$ is unstable simultaneously thermodynamically (it has a negative specific heat)

2293: and dynamically (amplitude of small pressure/energy density fluctuations exponentially grows).  

2294: 

2295: Finally, we do not see thermodynamic instability in thermal $\caln=2^*$ plasma with bosonic mass deformation for 

2296: $m_b^2<0$, see  Fig.~\ref{figfreeb}. In fact, we find that for negative values of $\mu$ the speed of sound 

2297: slowly (and monotonically) decreases --- it reaches about $90\%$ of the conformal plasma speed of sound \eqref{cscft}

2298: at\footnote{This is outside the range of the right plot in Fig.~\ref{figfreeb}.} $\mu\approx -35$. 

2299: Of course, this does not exclude the fact that thermal $\caln=2^*$ plasma with $m_b^2<0$ might have dynamical instability

2300: without  accompanying thermodynamics instability \cite{binst}.

2301: 

2302: 

2303: \subsection{Thermal PW flows with $m_f=m_b\equiv m$}

2304: 

2305: \begin{figure}[t]

2306:   \hspace*{-20pt}

2307: \psfrag{mT}{\raisebox{2ex}{\footnotesize\hspace{-1.0cm}$\qquad \frac{m}{T}$}}

2308: \psfrag{Fs}{\raisebox{0.5ex}{$f_{susy}\equiv -\calf_{susy}/\left(\frac 18 \pi^2 N^2 T^4\right)\ \qquad\ $}}

2309: \psfrag{lnf}{\raisebox{0.5ex}{$\ln\left(f_{susy}\right)$}}

2310:   \includegraphics[width=3.4in]{frees1.eps}

2311: %  \hspace*{-10pt}

2312:  \includegraphics[width=3.4in]{frees2.eps}%

2313: %  \hspace*{-20pt}

2314: %  \vspace*{-10pt}

2315:   \caption{

2316: The left plot represents supersymmetric mass deformation free energy $\calf_{susy}$ as 

2317: a function of $\frac{m}{T}$. The right plot represents $\ln(f_{susy})$ as a function of 

2318: $\frac{m}{T}$ (black points) and a straight line fit (red curve) for the last 2500 points

2319: of the distribution.

2320: } 

2321: \label{figfrees}

2322: \end{figure}

2323: 

2324: 

2325: \begin{figure}[t]

2326:   \hspace*{80pt}

2327: \psfrag{mT}{\raisebox{2ex}{\footnotesize\hspace{-0.5cm}$\frac{m}{T}$}}

2328: \psfrag{vs2}{\raisebox{0.5ex}{$3\ \times\ c_{s,susy}^2$}}

2329: %  \includegraphics[width=3.4in]{freeb}

2330: %  \hspace*{-10pt}

2331:  \includegraphics[width=3.4in]{sounds.eps}%

2332: %  \hspace*{-20pt}

2333: %  \vspace*{-10pt}

2334:   \caption{

2335: The speed of sound squared $c_{s,susy}^2$ for 

2336: supersymmetric mass deformation as a function of $\frac{m}{T}$.

2337: }

2338: \label{figsounds}

2339: \end{figure}

2340: 

2341: Bosonic and fermionic condensates $\{\r_{10}(\xi), \chi_{10}(\xi)\}$ 

2342: for thermal $\caln=2^*$ gauge theory plasma with supersymmetric mass deformation are presented in Fig.~\ref{figs1}. 

2343: As we argued in subsection \ref{subsec:vsa}, supergravity approximation here is valid of the whole range of deformation 

2344: parameter $\xi$ considered. We use \eqref{match3}, \eqref{vs2}, \eqref{mfs}, \eqref{speeds} and \eqref{fosusy} 

2345: to evaluate the free energy density and the speed of sound for the supersymmetric mass deformations.  

2346: Results of the analysis are presented in Figs.~\ref{figfrees}-\ref{figsounds}.

2347: We do not see any signature of the phase transition\footnote{This would show up as the free energy 

2348: $\calf_{susy}\left(\frac mT\right)$ changing sign for a particular 

2349: value of $\frac mT$. }, or the thermodynamic instability. 

2350: 

2351: Probably the most unexpected result of the analysis is that strongly coupled thermal $\caln=2^*$ plasma 

2352: for supersymmetric mass deformation behaves much closer to a conformal plasma than one would naively expect. 

2353: The right plot on Fig.~\ref{figfrees} presents the logarithm of the ratio  of the $\caln=2^*$ plasma free energy density

2354: to the free energy density of the conformal plasma \eqref{fcft}  

2355: \begin{equation}

2356: \ln(f_{susy})\equiv \ln \frac{\calf_{susy}}{\calf_{CFT}} \,,

2357: \eqlabel{fsusydef}

2358: \end{equation}

2359: as a function of $\frac mT$. 

2360: We did a straight line line fit (solid red line on the right plot on Fig.~\ref{figfrees}) for the low temperature\footnote{In practice, 

2361: we define low temperatures as the subset of  the mass deformation parameters $\frac mT>\sim 1$.} 

2362: \begin{equation}

2363: \ln(f_{susy})\bigg|^{FIT}\left(x\right)=0.118(8) - 0.138(1)\ x\,.

2364: \eqlabel{fitf}

2365: \end{equation}

2366: Thus, to a good approximation the free energy of the thermal $\caln=2^*$ plasma can be approximated by 

2367: \begin{equation}

2368: \calf_{susy}^{approx}\ \approx\ 1.1(3)\times \calf_{CFT}\ e^{-\frac{m}{7.2(4)\ T}}\ \propto\ - T^4\ e^{-\frac{m}{7\ T}}\,,

2369: \eqlabel{fapprox}

2370: \end{equation}

2371: so that noticeable deviation from conformality occurs at $T\sim \frac 17 m$.

2372: 

2373: 

2374: It is instructing to compare \eqref{fapprox} with the zero coupling result for the $\caln=2^*$ plasma.

2375: At zero coupling we have\footnote{We neglect $(-1)$ in the large $N$ limit.} $N^2$ massless $\caln=2$  vector multiples and $N^2$ massive $\caln=2$ 

2376: hypermultiplets. The free energy density of the massless states is just $\frac 23 \calf_{CFT}$ in \eqref{fcft}, while the 

2377: free energy density of a mass $m$ spin $s$ particle is given by \cite{dj,a} 

2378: \begin{equation}

2379: \calf^{s}=-\frac{(2s+1)m^2T^2}{2\pi^2}\sum_{\ell=1}^{\infty} \frac{\eta^{\ell+1}}{\ell^2} K_2\left(\frac{\ell m}{T}\right)\,,

2380: \eqlabel{frees}

2381: \end{equation}

2382: where $\eta=\pm 1$ for bosons/fermions correspondingly. As low temperatures $\frac mT\gg 1$ we can use asymptotic expansion 

2383: for the Bessel function $K_2(x)$ 

2384: \begin{equation}

2385: K_2(x)\approx \sqrt{\frac{\pi}{2x}}\ e^{-x}\,,\qquad x\gg 1\,,

2386: \eqlabel{k2ex}

2387: \end{equation}

2388: to conclude that 

2389: \begin{equation}

2390: \calf^{s}\approx -\frac{(2s+1)m^{3/2}T^{5/2}}{(2\pi)^{3/2}}\ e^{-\frac{m}{T}}\,,\qquad \frac{m}{T}\gg 1\,.

2391: \eqlabel{freea}

2392: \end{equation}

2393: Altogether we find  that the free energy of the thermal $\caln=2^*$ plasma for supersymmetric mass deformation at zero coupling is 

2394: \begin{equation}

2395: \calf_{\caln=2^*}=-\frac{\pi^2 N^2T^4}{12}\left(1+\frac{24\sqrt{2}}{\pi^{7/2}}\ \left(\frac mT\right)^{3/2}\ e^{-\frac mT}\right)\,,

2396: \qquad  \frac mT\gg 1\,.

2397: \eqlabel{free0c}

2398: \end{equation}

2399: From \eqref{free0c}, we find that the free energy density is roughly $60\%$ of the high temperature result at $T\sim m$, 

2400: while the corresponding number for the strongly coupled plasma \eqref{fitf} is about $98\%$.

2401: 

2402: The hydrodynamic properties and the jet quenching of strongly coupled nonconformal gauge theory plasma are 

2403: typically parameterized in terms of \cite{bbs,bjet} $\delta$

2404: \begin{equation}

2405: \delta\equiv \left(\frac 13-c_s^2\right) \,.

2406: \end{equation}

2407: For the strongly coupled $\caln=2^*$ thermal plasma at\footnote{The temperature of the 

2408: quark-gluon plasma at RHIC is of order the QCD deconfinement temperature \cite{shu}.} $T\sim 1.5 m$ we find (see Fig.~\ref{figsounds}) 

2409: \begin{equation}

2410: \delta^{\caln=2^*}|_{T\approx 1.5 m}\approx 2\times 10^{-3}\,,

2411: \eqlabel{delta}

2412: \end{equation}

2413: which further emphasizes the striking similarity between strongly coupled $\caln=2^*$ plasma and 

2414: conformal $\caln=4$ gauge theory plasma. 

2415:   

2416: 

2417: 

2418: 

2419: 

2420: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

2421: \section{Conclusion and future directions}

2422: \label{sec:cfd}

2423: 

2424: 

2425: 

2426: \begin{figure}[t]

2427:   \hspace*{-20pt}

2428: \psfrag{mT}{\raisebox{2ex}{\footnotesize\hspace{-1.0cm}$\qquad \frac{m}{T}$}}

2429: \psfrag{mbT}{\raisebox{2ex}{\footnotesize\hspace{-1.0cm}$\qquad \frac{m_b}{T}$}}

2430: \psfrag{es}{\raisebox{0.5ex}{$\cale_{susy}/\left(\frac 38 \pi^2 N^2 T^4\right)\ \qquad\ $}}

2431: \psfrag{eb}{\raisebox{0.5ex}{$\cale_{bosonic}/\left(\frac 38 \pi^2 N^2 T^4\right)\ \qquad\ $}}

2432: \psfrag{ms}{\raisebox{0ex}{\footnotesize\hspace{0.0cm}$\sqrt{\mu^*}$}}

2433:  \includegraphics[width=3.4in]{energyb.eps}

2434: %  \hspace*{-10pt}

2435:  \includegraphics[width=3.4in]{energys.eps}%

2436: %  \hspace*{-20pt}

2437: %  \vspace*{-10pt}

2438:   \caption{

2439: The left plot represents bosonic mass deformation energy density $\cale_{bosonic}$ as 

2440: a function of $\frac{m_b}{T}$. The value $\sqrt{\mu^*}$  (see Fig.~\ref{figsoundb}) denotes the ratio $\frac{m_b}{T}$

2441: for which the speed of sound vanishes.

2442: The right plot represents supersymmetric mass deformation energy density $\cale_{susy}$ as 

2443: a function of $\frac{m}{T}$.

2444: } 

2445: \label{figenergy}

2446: \end{figure}

2447: 

2448: 

2449: In this paper we presented detailed analysis of the thermodynamics of the strongly coupled $\caln=2^*$

2450: gauge theory plasma. We considered two special cases of the mass deformations: bosonic mass deformation 

2451: (where only the bosonic components of the $\caln=2$ hypermultiplet get nonzero mass) and supersymmetric  

2452: mass deformation (with bosonic and fermionic components of the hypermultiplet getting the same mass).

2453: We argued that both our theoretical framework and the numerical procedures are under control,

2454: given a highly nontrivial consistency check of the first law of thermodynamics for the extracted data.

2455: We also argued that the supergravity approximation is under control, as the curvature invariants of the 

2456: dual geometry remain small. 

2457: 

2458: The most important result of the analysis is the striking similarity between thermodynamics of the 

2459: mass deformed $\caln=2^*$ gauge theory plasma at temperatures of order the mass scale  

2460: and the thermodynamics of the conformal $\caln=4$ SYM plasma. We would like to reemphasize the latter fact by presenting the plots 

2461: for the energy energy for various mass deformations, see Fig.~\ref{figenergy}.

2462: 

2463: We could like to concluded with some open problems.

2464: \nxt

2465: In is important to generalize our analysis to other examples of the strongly coupled nonconformal gauge theory 

2466: plasma. Most notably the cascading gauge theory plasma \cite{kt1,kt2,kt3,aby}.

2467: \nxt It is interesting to evaluate hydrodynamic properties and jet quenching in $\caln=2^*$ gauge theory 

2468: plasma in the low temperature regime. 

2469: \nxt The speed of sound waves in $\caln=2^*$ gauge theory plasma for the bosonic mass deformation was 

2470: found to vanish for certain value of $\frac{m_b}{T}$. It is interesting to study 

2471: hydrodynamics of the gauge theory plasma in the regime of the vanishing speed of sound and determine 

2472: the critical exponent of the bulk viscosity.   

2473: 

2474: 

2475: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

2476: \section*{Acknowledgments}

2477: 

2478: We would like to thank Jaume Gomis, Chris Herzog, Pavel Kovtun, Gerry McKeon

2479: and Andrei Starinets for interesting discussions. 

2480: We would like to thank Ofer Aharony and Rob Myers for valuable comments on the manuscript. 

2481: AB's research at Perimeter

2482: Institute is supported in part by the Government of Canada through

2483: NSERC and by the Province of Ontario through MEDT.  AB gratefully

2484: acknowledges further support by an NSERC Discovery grant.  JTL's research

2485: was supported in part by the US Department of Energy under grant

2486: DE-FG02-95ER40899 and the National Science Foundation under grant

2487: PHY99-07949.  JTL acknowledges the hospitality of the KITP, where this

2488: work was initiated.

2489: 

2490: 

2491: 

2492: 

2493: 

2494: 

2495: 

2496: 

2497: 

2498: 

2499: 

2500: 

2501: 

2502: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

2503: \begin{thebibliography}{99}

2504: 

2505: \bibitem{m1}

2506:   J.~M.~Maldacena,

2507:   ``The large N limit of superconformal field theories and supergravity,''

2508:   Adv.\ Theor.\ Math.\ Phys.\  {\bf 2}, 231 (1998)

2509:   [Int.\ J.\ Theor.\ Phys.\  {\bf 38}, 1113 (1999)]

2510:   [arXiv:hep-th/9711200].

2511:   %%CITATION = HEP-TH 9711200;%%

2512: 

2513: \bibitem{m2}

2514:   O.~Aharony, S.~S.~Gubser, J.~M.~Maldacena, H.~Ooguri and Y.~Oz,

2515:   ``Large N field theories, string theory and gravity,''

2516:   Phys.\ Rept.\  {\bf 323}, 183 (2000)

2517:   [arXiv:hep-th/9905111].

2518:   %%CITATION = HEP-TH 9905111;%%

2519: 

2520: \bibitem{raj}

2521:   H.~Liu, K.~Rajagopal and U.~A.~Wiedemann,

2522:   ``Wilson loops in heavy ion collisions and their calculation in AdS/CFT,''

2523:   arXiv:hep-ph/0612168.

2524:   %%CITATION = HEP-PH 0612168;%%

2525: 

2526: 

2527: 

2528: \bibitem{rhic1}

2529:   K.~Adcox {\it et al.}  [PHENIX Collaboration],

2530:   ``Formation of dense partonic matter in relativistic nucleus nucleus

2531:   collisions at RHIC: Experimental evaluation by the PHENIX  collaboration,''

2532:   Nucl.\ Phys.\ A {\bf 757}, 184 (2005)

2533:   [arXiv:nucl-ex/0410003].

2534:   %%CITATION = NUCL-EX 0410003;%%

2535: 

2536: \bibitem{rhic2}

2537:   B.~B.~Back {\it et al.},

2538:   ``The PHOBOS perspective on discoveries at RHIC,''

2539:   Nucl.\ Phys.\ A {\bf 757}, 28 (2005)

2540:   [arXiv:nucl-ex/0410022].

2541:   %%CITATION = NUCL-EX 0410022;%%

2542: 

2543: \bibitem{rhic3}

2544:   I.~Arsene {\it et al.}  [BRAHMS Collaboration],

2545:   ``Quark gluon plasma and color glass condensate at RHIC? The perspective

2546:   from the BRAHMS experiment,''

2547:   Nucl.\ Phys.\ A {\bf 757}, 1 (2005)

2548:   [arXiv:nucl-ex/0410020].

2549:   %%CITATION = NUCL-EX 0410020;%%

2550: 

2551: \bibitem{rhic4}

2552:   J.~Adams {\it et al.}  [STAR Collaboration],

2553:   ``Experimental and theoretical challenges in the search for the quark  gluon

2554:   plasma: The STAR collaboration's critical assessment of the  evidence from

2555:   RHIC collisions,''

2556:   Nucl.\ Phys.\ A {\bf 757}, 102 (2005)

2557:   [arXiv:nucl-ex/0501009].

2558:   %%CITATION = NUCL-EX 0501009;%%

2559: 

2560: \bibitem{ads1}

2561:   S.~S.~Gubser, I.~R.~Klebanov and A.~W.~Peet,

2562:   %``Entropy and Temperature of Black 3-Branes,''

2563:   Phys.\ Rev.\ D {\bf 54}, 3915 (1996)

2564:   [arXiv:hep-th/9602135].

2565: 

2566: \bibitem{ads2}

2567:   S.~S.~Gubser, I.~R.~Klebanov and A.~A.~Tseytlin,

2568:   ``Coupling constant dependence in the thermodynamics of N = 4  supersymmetric

2569:   Yang-Mills theory,''

2570:   Nucl.\ Phys.\ B {\bf 534}, 202 (1998)

2571:   [arXiv:hep-th/9805156].

2572:   %%CITATION = HEP-TH 9805156;%%

2573: 

2574: 

2575: \bibitem{ads3}

2576:   G.~Policastro, D.~T.~Son and A.~O.~Starinets,

2577:   ``The shear viscosity of strongly coupled N = 4 supersymmetric Yang-Mills

2578:   plasma,''

2579:   Phys.\ Rev.\ Lett.\  {\bf 87}, 081601 (2001)

2580:   [arXiv:hep-th/0104066].

2581:   %%CITATION = HEP-TH 0104066;%%

2582: 

2583: \bibitem{ads4}

2584:   A.~Buchel, J.~T.~Liu and A.~O.~Starinets,

2585:   ``Coupling constant dependence of the shear viscosity in N = 4

2586:   supersymmetric Yang-Mills theory,''

2587:   Nucl.\ Phys.\ B {\bf 707}, 56 (2005)

2588:   [arXiv:hep-th/0406264].

2589:   %%CITATION = HEP-TH 0406264;%%

2590: 

2591: \bibitem{ads5}

2592:   P.~Benincasa and A.~Buchel,

2593:   ``Transport properties of N = 4 supersymmetric Yang-Mills theory at finite

2594:   coupling,''

2595:   JHEP {\bf 0601}, 103 (2006)

2596:   [arXiv:hep-th/0510041].

2597:   %%CITATION = HEP-TH 0510041;%%

2598: 

2599: \bibitem{ads6}

2600:   H.~Liu, K.~Rajagopal and U.~A.~Wiedemann,

2601:   ``Calculating the jet quenching parameter from AdS/CFT,''

2602:   Phys.\ Rev.\ Lett.\  {\bf 97}, 182301 (2006)

2603:   [arXiv:hep-ph/0605178].

2604:   %%CITATION = HEP-PH 0605178;%%

2605: 

2606: \bibitem{ads7}

2607:   C.~P.~Herzog, A.~Karch, P.~Kovtun, C.~Kozcaz and L.~G.~Yaffe,

2608:   %``Energy loss of a heavy quark moving through N = 4 supersymmetric

2609:   %Yang-Mills plasma,''

2610:   JHEP {\bf 0607}, 013 (2006)

2611:   [arXiv:hep-th/0605158].

2612: 

2613: 

2614: \bibitem{ads8}

2615:   H.~Liu, K.~Rajagopal and U.~A.~Wiedemann,

2616:   ``An AdS/CFT calculation of screening in a hot wind,''

2617:   arXiv:hep-ph/0607062.

2618:   %%CITATION = HEP-PH 0607062;%%

2619: 

2620: \bibitem{ads9}

2621:   S.~Caron-Huot, P.~Kovtun, G.~D.~Moore, A.~Starinets and L.~G.~Yaffe,

2622:   ``Photon and dilepton production in supersymmetric Yang-Mills plasma,''

2623:   arXiv:hep-th/0607237.

2624:   %%CITATION = HEP-TH 0607237;%%

2625: 

2626: \bibitem{shu2}

2627:   E.~V.~Shuryak,

2628:   ``Strongly coupled quark-gluon plasma: The status report,''

2629:   arXiv:hep-ph/0608177.

2630:   %%CITATION = HEP-PH 0608177;%%

2631: 

2632: \bibitem{kt1}

2633:   A.~Buchel,

2634:   ``Finite temperature resolution of the Klebanov-Tseytlin singularity,''

2635:   Nucl.\ Phys.\ B {\bf 600}, 219 (2001)

2636:   [arXiv:hep-th/0011146].

2637:   %%CITATION = HEP-TH 0011146;%%

2638: 

2639: \bibitem{kt2}

2640:   A.~Buchel, C.~P.~Herzog, I.~R.~Klebanov, L.~A.~Pando Zayas and A.~A.~Tseytlin,

2641:   ``Non-extremal gravity duals for fractional D3-branes on the conifold,''

2642:   JHEP {\bf 0104}, 033 (2001)

2643:   [arXiv:hep-th/0102105].

2644:   %%CITATION = HEP-TH 0102105;%%

2645: 

2646: \bibitem{kt3}

2647:   S.~S.~Gubser, C.~P.~Herzog, I.~R.~Klebanov and A.~A.~Tseytlin,

2648:   ``Restoration of chiral symmetry: A supergravity perspective,''

2649:   JHEP {\bf 0105}, 028 (2001)

2650:   [arXiv:hep-th/0102172].

2651:   %%CITATION = HEP-TH 0102172;%%

2652: 

2653: 

2654: 

2655: 

2656: 

2657: \bibitem{bl}

2658: A.~Buchel and J.~T.~Liu,

2659: ``Thermodynamics of the N = 2* flow,''

2660: JHEP {\bf 0311}, 031 (2003)

2661: [arXiv:hep-th/0305064].

2662: 

2663: 

2664: \bibitem{aby}

2665:   O.~Aharony, A.~Buchel and A.~Yarom,

2666:   ``Holographic renormalization of cascading gauge theories,''

2667:   Phys.\ Rev.\ D {\bf 72}, 066003 (2005)

2668:   [arXiv:hep-th/0506002].

2669:   %%CITATION = HEP-TH 0506002;%%

2670: 

2671: \bibitem{rob}

2672:   D.~Mateos, R.~C.~Myers and R.~M.~Thomson,

2673:   ``Thermodynamics of the brane,''

2674:   arXiv:hep-th/0701132.

2675: 

2676: 

2677: 

2678: \bibitem{hr}

2679:   A.~Buchel,

2680:   ``N = 2* hydrodynamics,''

2681:   Nucl.\ Phys.\ B {\bf 708}, 451 (2005)

2682:   [arXiv:hep-th/0406200].

2683:   %%CITATION = HEP-TH 0406200;%%

2684: 

2685: 

2686: \bibitem{bbs}

2687: P.~Benincasa, A.~Buchel and A.~O.~Starinets,

2688: ``Sound waves in strongly coupled non-conformal gauge theory plasma,''

2689:   Nucl.\ Phys.\ B {\bf 733}, 160 (2006)

2690:   [arXiv:hep-th/0507026].

2691: 

2692: 

2693: 

2694: \bibitem{bpp}

2695:   A.~Buchel, A.~W.~Peet and J.~Polchinski,

2696:   ``Gauge dual and noncommutative extension of an N = 2 supergravity

2697:   solution,''

2698:   Phys.\ Rev.\ D {\bf 63}, 044009 (2001)

2699:   [arXiv:hep-th/0008076].

2700:   %%CITATION = HEP-TH 0008076;%%

2701: 

2702: 

2703: \bibitem{pw}

2704: K.~Pilch and N.~P.~Warner,

2705: ``N = 2 supersymmetric RG flows and the IIB dilaton,''

2706: Nucl.\ Phys.\ B {\bf 594}, 209 (2001)

2707: [arXiv:hep-th/0004063].

2708: 

2709: 

2710: 

2711: \bibitem{dw}

2712:   R.~Donagi and E.~Witten,

2713:   ``Supersymmetric Yang-Mills Theory And Integrable Systems,''

2714:   Nucl.\ Phys.\ B {\bf 460}, 299 (1996)

2715:   [arXiv:hep-th/9510101].

2716:   %%CITATION = HEP-TH 9510101;%%

2717: 

2718: 

2719: \bibitem{ps}

2720:   J.~Polchinski and M.~J.~Strassler,

2721:   ``The string dual of a confining four-dimensional gauge theory,''

2722:   arXiv:hep-th/0003136.

2723:   %%CITATION = HEP-TH 0003136;%%

2724: 

2725: 

2726: 

2727: \bibitem{blu}

2728:   A.~Buchel and J.~T.~Liu,

2729:   ``Universality of the shear viscosity in supergravity,''

2730:   Phys.\ Rev.\ Lett.\  {\bf 93}, 090602 (2004)

2731:   [arXiv:hep-th/0311175].

2732:   %%CITATION = HEP-TH 0311175;%%

2733: 

2734:  

2735: 

2736: \bibitem{ejp}

2737:   N.~J.~Evans, C.~V.~Johnson and M.~Petrini,

2738:   ``The enhancon and N = 2 gauge theory/gravity RG flows,''

2739:   JHEP {\bf 0010}, 022 (2000)

2740:   [arXiv:hep-th/0008081].

2741:   %%CITATION = HEP-TH 0008081;%%

2742: 

2743: 

2744: \bibitem{b}

2745:   A.~Buchel,

2746:   ``Comments on fractional instantons in N = 2 gauge theories,''

2747:   Phys.\ Lett.\ B {\bf 514}, 417 (2001)

2748:   [arXiv:hep-th/0101056].

2749:   %%CITATION = HEP-TH 0101056;%%

2750: 

2751: 

2752: \bibitem{kp}

2753:   S.~S.~Gubser, I.~R.~Klebanov and A.~W.~Peet,

2754:   ``Entropy and Temperature of Black 3-Branes,''

2755:   Phys.\ Rev.\ D {\bf 54}, 3915 (1996)

2756:   [arXiv:hep-th/9602135].

2757:   %%CITATION = HEP-TH 9602135;%%

2758: 

2759: \bibitem{gm}

2760:   S.~S.~Gubser and I.~Mitra,

2761:   ``Instability of charged black holes in anti-de Sitter space,''

2762:   arXiv:hep-th/0009126.

2763:   %%CITATION = HEP-TH 0009126;%%

2764: 

2765: \bibitem{binst}

2766:   A.~Buchel,

2767:   ``A holographic perspective on Gubser-Mitra conjecture,''

2768:   Nucl.\ Phys.\ B {\bf 731}, 109 (2005)

2769:   [arXiv:hep-th/0507275].

2770:   %%CITATION = HEP-TH 0507275;%%

2771: 

2772: 

2773: \bibitem{pertg}

2774:   J.~O.~Andersen, E.~Petitgirard and M.~Strickland,

2775:   ``Two-loop HTL thermodynamics with quarks,''

2776:   Phys.\ Rev.\ D {\bf 70}, 045001 (2004)

2777:   [arXiv:hep-ph/0302069].

2778:   %%CITATION = HEP-PH 0302069;%%

2779: 

2780: \bibitem{npertg}

2781:   A.~Buchel,

2782:   ``Higher derivative corrections to near-extremal black holes in type IIB

2783:   supergravity,''

2784:   Nucl.\ Phys.\ B {\bf 750}, 45 (2006)

2785:   [arXiv:hep-th/0604167].

2786:   %%CITATION = HEP-TH 0604167;%%

2787: 

2788: \bibitem{bninst}

2789:   A.~Buchel,

2790:   ``Gauge theories on hyperbolic spaces and dual wormhole instabilities,''

2791:   Phys.\ Rev.\ D {\bf 70}, 066004 (2004)

2792:   [arXiv:hep-th/0402174].

2793:   %%CITATION = HEP-TH 0402174;%%

2794: 

2795: \bibitem{mm}

2796:   J.~M.~Maldacena and L.~Maoz,

2797:   ``Wormholes in AdS,''

2798:   JHEP {\bf 0402}, 053 (2004)

2799:   [arXiv:hep-th/0401024].

2800:   %%CITATION = HEP-TH 0401024;%%

2801: 

2802: \bibitem{dj}

2803:   J.~I.~Kapusta and C.~Gale,

2804:   ``Finite-Temperature Field Theory: Principles and Applications,''

2805:   Cambridge Monographs on Mathematical Physics  (2006).

2806:   

2807: 

2808: \bibitem{a}

2809: A.~O.~Starinets, ``Effective potentials in quantum field theory at finite 

2810: temperature'', Cand. Sci. Dissertation (1994), in Russian.

2811: 

2812: \bibitem{bjet} 

2813:   A.~Buchel,

2814:   ``On jet quenching parameters in strongly coupled non-conformal gauge

2815:   theories,''

2816:   Phys.\ Rev.\ D {\bf 74}, 046006 (2006)

2817:   [arXiv:hep-th/0605178].

2818:   %%CITATION = HEP-TH 0605178;%%

2819: 

2820: \bibitem{shu}

2821:   E.~V.~Shuryak,

2822:   ``What RHIC experiments and theory tell us about properties of  quark-gluon

2823:   plasma?,''

2824:   Nucl.\ Phys.\ A {\bf 750}, 64 (2005)

2825:   [arXiv:hep-ph/0405066].

2826:   %%CITATION = HEP-PH 0405066;%%

2827: 

2828: 

2829: 

2830: \end{thebibliography}

2831: 

2832: 

2833: \end{document}

2834: 

2835: