1: \documentclass[aps,rmp,eqsecnum,amssymb,12pt]{article}

2: %\usepackage{graphicx}

3: %\usepackage{showkeys}

4: \usepackage{amssymb}

5: \usepackage{amsmath}

6: \usepackage{epsfig}

7: \usepackage{natbib}

8: \usepackage[dvips]{color}

9: 

10: \setlength{\bibsep}{6.2pt}

11: 

12: \addtolength{\oddsidemargin}{-.875in}

13:     \addtolength{\evensidemargin}{-.875in}

14:     \addtolength{\textwidth}{1.75in}

15: 

16:     \addtolength{\topmargin}{-.875in}

17:     \addtolength{\textheight}{1.75in}

18: 

19: \def\beq{\begin{equation}}

20: \def\eeq{\end{equation}}

21: 

22: 

23: \def\beqa{\begin{eqnarray}}

24: \def\eeqa{\end{eqnarray}}

25: 

26: \begin{document}

27: 

28: 

29: \begin{titlepage}

30: 

31: \setcounter{page}{1} \baselineskip=15.5pt \thispagestyle{empty}

32: 

33: \begin{flushright}

34: PUPT-2221\\

35: hep-th/0701064\\

36: \end{flushright}

37: \vfil

38: 

39: \begin{center}

40: {\LARGE Gauge/Gravity Duality and Warped Resolved Conifold}

41: \end{center}

42: \bigskip\

43: 

44: \begin{center}

45: %

46: {\large Igor R. Klebanov and Arvind Murugan}

47: %

48: \end{center}

49: 

50: \begin{center}

51: \textit{Department of Physics and Princeton Center for Theoretical

52: Physics\\ Princeton University, Princeton, NJ 08544}

53: \end{center} \vfil

54: 

55: 

56: \noindent We study supergravity backgrounds encoded through the

57: gauge/string correspondence by the $SU(N) \times SU(N)$ theory

58: arising on $N$ D3-branes on the conifold. As discussed in

59: hep-th/9905104, the dynamics of this theory describes warped

60: versions of both the singular and the resolved conifolds through

61: different (symmetry breaking) vacua. We construct these

62: supergravity solutions explicitly and match them

63: with the gauge theory with different sets of vacuum expectation

64: values of the bi-fundamental fields $A_1, A_2, B_1, B_2$. For the

65: resolved conifold, we find a non-singular $SU(2)\times U(1)\times

66: U(1)$ symmetric warped solution produced by a stack of D3-branes

67: localized at a point on the blown-up 2-sphere. It describes a

68: smooth RG flow from $AdS_5 \times T^{1,1}$ in the UV to $AdS_5

69: \times S^5$ in the IR, produced by giving a VEV to just one field,

70: e.g. $B_2$. The presence of a condensate of 

71: baryonic operator ${\rm det} B_2$ is confirmed using a Euclidean D3-brane

72: wrapping a 4-cycle inside the resolved conifold. 

73: The Green's functions on the singular and resolved

74: conifolds are central to our calculations and

75: are discussed in some detail.

76: 

77: \vfil

78: 

79: \end{titlepage}

80: 

81: %=========================================================================

82: \newpage

83: 

84: \tableofcontents

85: 

86: \pagestyle{headings}

87: 

88: \section{Introduction}

89: \label{sec-intro} The basic AdS/CFT correspondence

90: \cite{Maldacena,GKP,Witten} (see

91: \cite{Aharony:1999ti,Klebanov:2000me} for reviews) is motivated by

92: considering the low energy physics of a heavy stack of D3-branes at

93: a point in flat spacetime. Taking the near-horizon limit of this

94: geometry motivates a duality between type IIB string theory on

95: $AdS_5 \times S^5$ and ${\cal N}=4$ $SU(N)$ supersymmetric

96: Yang-Mills gauge theory. This correspondence was generalized

97: to theories with ${\cal N}=1$ superconformal symmetry in

98: \cite{KW,Morrison} by considering a stack D3-branes, not in flat

99: space, but placed at the tip of a 6d Calabi-Yau cone $X_6$. The near

100: horizon limit in this case turns out to be $AdS_5\times Y_5$ where

101: $Y_5$ is the compact 5 dimensional base of $X_6$ and is a

102: Sasaki-Einstein space.

103: 

104: Among the simplest of these examples is $Y_5 = T^{1,1}$, corresponding

105: $X_6$ being the conifold. It was found that the low-energy

106: gauge theory on the D3-branes at the tip of the conifold is a ${\cal

107: N}=1$ supersymmetric $SU(N)\times SU(N)$ gauge theory with

108: bi-fundamental chiral superfields $A_i$, $B_j$ $(i,j=1,2)$ in

109: $(N,\bar{N})$ and $(\bar{N},N)$ representations of the gauge groups,

110: respectively \cite{KW,Morrison}. The superpotential for this gauge

111: theory is $W \sim {\rm Tr} \det A_i B_j = {\rm Tr}\ (A_1 B_1 A_2 B_2 -

112: A_1 B_2 A_2 B_1)$. The continuous global symmetries of this theory

113: are $SU(2) \times SU(2) \times U(1)_R \times U(1)_B$ where the $SU(2)$

114: factors act on

115: $A_i$ and $B_j$ respectively, $U(1)_B$ is a baryonic

116: symmetry, and $U(1)_R$ is the R-symmetry with $R_A=R_B=\frac{1}{2}$.

117: This assignment ensures that $W$ is marginal, and one can also show

118: that the gauge couplings do not run. Hence this theory is

119: superconformal for all values of gauge couplings and superpotential

120: coupling \cite{KW,Morrison}.

121: 

122:   When the above gauge theory is considered with no vacuum expectation values

123:   (VEV's) for any of the fields,

124:   we have a superconformal theory with the $AdS_5\times T^{1,1}$ dual. In \cite{KW2},

125:   more general vacua of this theory were studied. It was argued that moving the D3-branes off the tip

126:   of the singular conifold corresponds to a symmetry breaking in the gauge theory due to VEV's for

127:   the $A,B$ matter fields such that the VEV of operator

128: \begin{equation}

129:   {\cal U} = {1\over N}

130: {\rm Tr} \left (|B_1|^2 +|B_2|^2-|A_1|^2 -|A_2|^2 \right )

131:   \end{equation}

132:   vanishes.

133:   Further, more general vacua exist for this theory

134:   in which this operator acquires a non-zero VEV.\footnote{

135:   As was pointed out in \cite{KW}, no D-term equation

136: constrains this

137:   operator since the $U(1)$ gauge groups decouple in the infrared.}

138:   It was pointed out in \cite{KW2} that these vacua cannot correspond to D3-branes on the singular

139:   conifold.

140:   Instead, such vacua with ${\cal U} \neq 0$ correspond to D3-branes on the resolved conifold.

141: This ``small resolution'' is a motion along the K\" ahler moduli space

142:   where the singularity of the conifold

143: is replaced by a finite $S^2$. Thus the $SU(N)\times SU(N)$ gauge theory was argued to

144:   incorporate in its different vacua both the singular and resolved conifolds.

145:   On the other hand, the deformation of the conifold, which is a motion along

146: the complex structure moduli space, can be achieved through

147: replacing

148:   the gauge theory by the cascading $SU(N) \times SU(N+M)$ gauge theory (see \cite{KS}).

149: 

150: One of the goals of this paper is to construct the warped SUGRA

151: solutions corresponding to the gauge theory vacua with

152:   ${\cal U} \neq 0$. Our work builds on the earlier resolved conifold solutions

153:   constructed by Pando Zayas and Tseytlin

154:   \cite{PandoZayas}, where additional simplifying symmetries were sometimes imposed.

155:   Such solutions corresponding to D3-branes ``smeared'' over a

156:   region were found to be singular in the IR \cite{PandoZayas}.

157: We will instead look for ``localized'' solutions

158:   corresponding to the whole D3-brane stack located at one point on the (resolved) conifold.

159:   This corresponds to giving VEV's to the fields $A_i, B_j$ which are proportional

160: to $1_{N\times N}$.  We construct the

161: duals of these gauge theory vacua and find them

162: to be completely non-singular. The solution acquires a particularly simple

163: form when the stack is placed at the north pole of the blown up

164: 2-sphere at the bottom of the resolved conifold.

165:   It corresponds to the simplest way to have ${\cal U} \neq 0 $ by setting

166:   $B_2= u 1_{N\times N}$ while

167:   keeping $A_1 = A_2 = B_1 =0 $.

168: 

169:  Following \cite{Balasubramanian,KW2}, we also interpret our solutions

170: %this construction in the gauge theory

171:  as having

172:  an infinite series of VEV's for various operators in addition to

173: ${\cal U}$. For this, we rely on the relation between

174:  normalizable SUGRA modes and gauge theory VEV's in the AdS/CFT dictionary.

175:  When a given asymptotically AdS

176: solution has a (linearized) perturbation that falls off as $r^{-\Delta}$ at large $r$,

177:  it corresponds to assigning a VEV for a certain operator ${\cal O}$ of dimension $\Delta$ in the dual

178:  gauge theory \cite{Balasubramanian,KW2}.

179:  The warp factor produced by a stack of D3-branes on the resolved conifold is related to

180:  the Green's function on

181:  the resolved conifold. This warp factor can be expanded in harmonics and corresponds to a series of

182:  normalizable fluctuations as above, and hence a series of operators in the gauge theory

183:  acquire VEV's.\footnote{In the ${\cal N}=4$ SUSY example,

184: the normalizations of the VEV's have been matched with the size of

185: the SUGRA perturbations around $AdS_5\times S^5$

186:  (see \cite{deHaro,Skenderis1,Skenderis2}).

187:  In this paper we limit ourselves to a more qualitative picture where the

188: precise normalizations

189:  of the VEV's are not calculated.}

190:  For this purpose, we write the harmonics in a convenient set of variables $a_i,b_j$ that makes the

191:  link with gauge theory operators built from $A_i,B_j$

192:  immediate.

193: Due to these symmetry breaking VEV's, the gauge theory flows from the $SU(N)\times

194: SU(N)$ ${\cal N}=1$ theory in

195:    the UV to the $SU(N)$ ${\cal N}=4$ theory in the IR,

196: as one would expect when D3-branes are placed at a

197:    smooth point.  The SUGRA solution is shown to have two asymptotic AdS regions --

198: an $AdS_5 \times T^{1,1}$ region in the UV, and also an $AdS_5   \times S^5$ region

199:    produced in the IR by the localized stack of D3-branes.

200:    This can be considered an example of holographic RG flow.

201: The Green's functions determined here might also have applications to models of

202:    D-brane inflation,

203:    and to computing 1-loop corrections to gauge couplings in gauge theories

204:    living on cycles in the geometry \cite{Giddings,BDKMMM}.

205: 

206: 

207:    When the branes are placed on the blown up 2-sphere at the bottom of the resolved

208: conifold, this corresponds to

209: $A_1= A_2=0$ in the gauge theory.

210: Hence no chiral mesonic operators, such as ${\rm Tr} A_i B_j$, have VEV's,

211: but baryonic operators, such as $\det B_2$, do acquire VEV's. Therefore,

212: such solutions, parametrized by the size of the resolution and position of the stack on the 2-sphere,

213: are dual to a ``non-mesonic'' (or ``baryonic'') branch of the $SU(N)\times SU(N)$ SCFT

214: (see \cite{Benvenuti} for a related discussion). These solutions have a blown up $S^2$.

215: On the other hand, the

216: solutions dual to the baryonic branch of the cascading $SU(N) \times SU(N+M)$

217: gauge theory were constructed in

218: \cite{Butti,Dymarsky} (for an earlier linearized treatment, see \cite{GHK}) and have a blown up

219: $S^3$ supported by the 3-form flux.

220: 

221: The paper is organized as follows.

222:  In Section \ref{sec-conifold}, we review and establish notation for describing the conifold,

223:  its resolution, its symmetries and coordinates that make the symmetries manifest.

224: We also review 

225: the metric of the resolved conifold and the

226:  singular smeared solution found in \cite{PandoZayas}.

227:  In Section \ref{sec-singular}, as a warm up, we study the simple example of moving a stack of D3-branes

228:  away from the tip of the singular conifold. We present the explicit supergravity solution for this

229:  configuration by determining the Green's function on the conifold. We interpret the operators that get

230:  VEV's and note that in general, chiral as well as non-chiral operators get VEV's.

231:  In Section \ref{sec-resolved}, we determine the explicit SUGRA solution corresponding to a heavy stack of

232:  D3-branes at a point on the resolved conifold, again by finding the Green's function on the manifold.

233:  We find a non-singular solution with an $AdS_5 \times S^5$ region and interpret this construction

234:  in gauge theory. We consider a wrapped Euclidean D3-brane to confirm the presence of baryonic VEVs and reproduce the wavefunction of a charged particle in a monopole field from the DBI action as a check on our calculations. We make a brief note on turning on a fluxless NS-NS $B_2$ field on the warped resolved

235:  conifold in Section \ref{sec-Bfield}. In Appendix A we discuss the harmonics on $T^{1,1}$ in

236:  co-ordinates that make

237:  the symmetries manifest. We then classify operators in the gauge theory by symmetry

238:  in an analogous way to enable simple matching of operator VEV's and normalizable fluctuations.

239: 

240: \section{The Conifold and its Resolution}

241: \label{sec-conifold}

242: \label{sec-smeared}

243: 

244:  The conifold is a singular non-compact Calabi-Yau three-fold \cite{Candelas}. Its importance

245: arises from the fact that the generic singularity in a Calabi-Yau

246: three-fold locally looks like the conifold. This is because it is

247: given by the quadratic equation,

248:  \beq

249:  z_1^2 + z_2^2 + z_3^2 + z_4^2 = 0.

250:  \eeq

251:  This homogeneous equation defines a real cone over a 5 dimensional manifold. For the cone to be

252:  Ricci-flat

253: the 5d base must be an Einstein manifold ($R_{\mu \nu} = 4 g_{\mu \nu}$).

254:  For the conifold \cite{Candelas}, the topology of the base can be shown to be $S^2 \times S^3$ and it is

255:  called $T^{1,1}$ with the following Einstein metric,

256: \begin{eqnarray} \label{T11metric}

257:   d\Omega_{T^{1,1}}^2 &=&  \frac{1}{9}  \left( d\psi + \cos \theta_1 d\phi_1 + \cos \theta_2 d\phi_2 \right)^2  \nonumber \\

258:  & &  \quad + \frac{1}{6}  (d\theta_1^2 + \sin^2 \theta_1 d\phi_1^2) + \frac{1}{6} (d\theta_2^2 + \sin^2 \theta_2 d\phi_2^2).

259: \end{eqnarray}

260: The metric on the cone is then $ds^2 = dr^2 + r^2

261: d\Omega_{T^{1,1}}^2$. As shown in \cite{Candelas} and earlier in

262: \cite{Romans}, $T^{1,1}$ is a homogeneous space, being the coset

263: $SU(2)\times SU(2)/U(1)$ and the above metric is the invariant

264: metric on the coset space.

265: 

266: We may introduce two other types  of complex coordinates on the

267: conifold, $w_a$ and $a_i,b_j$, as follows,

268: \begin{eqnarray} \label{coords}

269:  Z = \left( \begin{array}{cc}  z^3 + i z^4 & z^1 - i z^2 \\ z^1 +i z^2 & -z^3 +i z^4 \\ \end{array} \right) \nonumber

270:  = \left(\begin{array}{cc}w_1 & w_3 \\ w_4& w_2 \end{array} \right)

271: = \left( \begin{array}{cc}  a_1 b_1 & a_1 b_2 \\ a_2 b_1 & a_2 b_2 \\ \end{array} \right) \\

272: = r^\frac{3}{2} \left(\begin{array}{cc} - c_1 s_2 \; e^{\frac{i}{2} ( \psi + \phi_1-\phi_2)} &

273:     c_1 c_2 \; e^{\frac{i}{2} ( \psi+ \phi_1 +\phi_2)} \\

274:                   - s_1 s_2 \; e^{\frac{i}{2} ( \psi - \phi_1 -\phi_2)} & s_1 c_2 \;e^{\frac{i}{2}

275:                   ( \psi - \phi_1+\phi_2)} \end{array} \right)

276: \end{eqnarray}

277: where $c_i = \cos \frac{\theta_i}{2},\ s_i = \sin

278: \frac{\theta_i}{2}$ (see \cite{Candelas} for other details on the

279: $w,z$ and angular coordinates.) The equation defining the conifold

280: is now $\det Z = 0$.

281: 

282: The $a,b$ coordinates above will be of particular interest in this

283: paper because the symmetries of the conifold are most apparent in

284: this basis. The conifold equation has $SU(2)\times SU(2) \times

285: U(1)$ symmetry since under these symmetry transformations, \beq

286: \det L Z R^{T} = \det e^{i \alpha} Z = 0. \eeq This is also a

287: symmetry of the metric presented above where each $SU(2)$ acts on

288: $\theta_i,\phi_i,\psi$ (thought of as Euler angles on $S^3$) while

289: the $U(1)$ acts by shifting $\psi$. This symmetry can be

290: identified with $U(1)_R$, the R-symmetry of the dual gauge theory,

291:  in the conformal case.

292: The action of the $SU(2) \times SU(2) \times U(1)_R$ symmetry on

293: $a_i,b_j$ (defined in (\ref{coords})): \beqa SU(2) \times SU(2)

294: \mbox{  symmetry}&:& \quad

295:  \left( a_1 \atop a_2 \right)  \rightarrow L \; \left( a_1 \atop a_2 \right) ,

296: \qquad \left( b_1 \atop b_2 \right)  \rightarrow R \;

297: \left( b_1 \atop b_2 \right)  \\

298:  \mbox{R-symmetry}&:& \quad (a_i,b_j) \rightarrow e^{i \frac{\alpha}{2}} (a_i,b_j)

299: \ ,\eeqa i.e. $a$ and $b$ transform as $(1/2,0)$ and $(0,1/2)$

300: under $SU(2) \times SU(2)$ with R-charge $1/2$ each.

301: We can thus describe the singular conifold as points parametrized by

302: $a,b$ but from (\ref{coords}), we see that there is some redundancy

303: in the $a,b$ coordinates. Namely, the transformation \beq

304: \label{removeredun}

305:   a_i  \rightarrow \; \lambda \; a_i  \quad,\quad  b_j  \rightarrow \; \frac{1}{\lambda} \; b_j  \quad(\lambda \in \mathbf{C})

306: \eeq

307:  give the same $z,w$ in (\ref{coords}).

308:  We impose the constraint $|a_1|^2 + |a_2|^2 - |b_1|^2 -|b_2|^2 = 0$ to fix the magnitude in the

309:  above transformation. To account for the remaining phase, we describe the singular conifold as

310:  the quotient of the $a,b$ space with the above constraint by the relation

311:  $ a \sim e^{i\alpha} a, b \sim e^{-i\alpha} b$.

312: 

313: One simple way to describe the resolution is as the space obtained

314: by modifying the above constraint to, \beq \label{resconstr} |b_1|^2

315: + |b_2|^2 - |a_1|^2 -|a_2|^2 = u^2 \eeq and then taking the

316: quotient, $ a \sim e^{i\alpha} a, b \sim e^{-i\alpha} b$. Then $u$

317: is a measure of the resolution and it can be seen that this space is

318: a smooth Calabi-Yau space where the singular point of the conifold

319: is replaced by a finite $S^2$. The complex metric on this space is

320: given in \cite{Candelas} while an explicit metric, first presented

321: in \cite{PandoZayas}, is: \beqa \label{resolvedmetric}

322:  ds_6^2 &=& \kappa^{-1}(r) dr^2 +  \frac{1}{9} \kappa(r) r^2 \left( d\psi + \cos \theta_1 d\phi_1 + \cos \theta_2 d\phi_2 \right)^2  \nonumber \\

323:  & &  \quad + \frac{1}{6} r^2 (d\theta_1^2 + \sin^2 \theta_1 d\phi_1^2) + \frac{1}{6} (r^2 + 6 u^2) (d\theta_2^2 + \sin^2 \theta_2 d\phi_2^2)

324: \eeqa

325: where

326: \beqa

327:   \kappa(r) = \displaystyle\frac{r^2 + 9 u^2}{r^2 + 6 u^2}\ ,

328: \eeqa 

329: where $r$ ranges from $0$ to $\infty$.

330: Note that the above metric has a finite $S^2$ of radius $u$ at

331: $r=0$, parametrized by $\theta_2,\phi_2$. Topologically, the resolved conifold

332: is an ${\bf R}^4$ bundle over $S^2$.

333: The metric asymptotes to that of the

334: singular conifold for large $r$.

335: 

336: Now we consider metrics produced by D3-branes on the conifold. As a

337: warm-up to the case of the resolved conifold, we consider the

338: example of placing a stack of D3-branes away from the apex of the

339: singular conifold. As in \cite{KW2}, the corresponding supergravity

340: solution is

341: \beqa \label{singmetric} ds^2

342: &=& \sqrt{H^{-1}(y)} \; \eta_{\mu\nu} dx^\mu dx^\nu + \sqrt{H(y)}

343: \left( dr^2 + r^2 d\Omega^2_{T^{1,1}} \right)\, ,\\ \label{f5dil}

344: F_5 &=& (1 +

345: *)dH^{-1}\wedge dx^0\wedge dx^1\wedge dx^2\wedge dx^3 , \qquad \Phi

346: = \mbox{const} \eeqa where $\mu, \nu = 0,1,2,3$ are the directions

347: along the D3-branes. $H(y)$ is a solution of the Green's equation on

348: the conifold

349: \beqa \label{greendef0} \Delta H(r,Z ; r_0, Z_0) =

350: \frac{1}{\sqrt{g}}\partial_m (\sqrt{g} g^{mn} \partial_n H ) &=& - {\cal C}

351: \frac{1}{\sqrt{g}} \delta(r-r_0) \delta^5(Z-Z_0) \ , \\

352: {\cal C} =2\kappa_{10}^2 T_3 N &=& (2 \pi)^4 g_s N (\alpha')^2 \ ,

353: \eeqa 

354: where

355: $(r_0,Z_0)$ is the location of the stack ($Z$ will represent

356: coordinates on $T^{1,1}$) and $T_3= \frac{1}{g_s (2\pi)^3 (\alpha')^2}$ is the

357: D3-brane tension.

358: %Here ${\cal C} =2\kappa_{10}^2 T_3 N=

359: % (2 \pi)^4 g_s N (\alpha')^2$

360: %(see \cite{BDKMMM}).

361: 

362: When the stack of D3-branes is placed at $r_0 = 0$, the solution is $H

363: =L^4/r^4$ where $L^4  = \frac{27 \pi g_s N (\alpha')^2}{4}$.  This reduces the

364: metric to ($z=L^2/r$),

365: \begin{eqnarray} \label{AdST11}

366: ds^2 = \frac{L^2}{z^2} \; (dz^2 + \eta_{\mu\nu} dx^\mu dx^\nu) +

367: L^2 d\Omega^2_{T^{1,1}}

368: \end{eqnarray}

369: This is the $AdS_5 \times T^{1,1}$ background, which is dual to the

370: superconformal $SU(N) \times SU(N)$ theory without any VEV's for the

371: bifundamental superfields. More general locations of the stack,

372: corresponding to giving VEV's that preserve the condition ${\cal

373: U}=0$, will be considered in section 4.

374: 

375: Now consider the case of resolved conifold. With D3-branes placed on

376: this manifold, we get the warped 10-d metric, \beq

377: \label{warpedresolved} ds_{10}^2 = \sqrt{H^{-1}(y)} dx^\mu dx_\mu +

378: \sqrt{H(y)} ds_6^2 \eeq where

379: $ds_6^2$ is the resolved conifold metric (\ref{resolvedmetric}) and

380: $H(y)$ is the warp factor as a function of the transverse

381: co-ordinates $y$, determined by the D3-brane positions. The dilaton

382: is again constant, and $F_5 = (1 +

383: *)dH^{-1}\wedge dx^0\wedge dx^1\wedge dx^2\wedge dx^3$.

384: 

385: In \cite{PandoZayas}, the warped supergravity solution was worked

386: out assuming a warp factor with only radial dependence (i.e no

387: angular dependence on $\theta_2,\phi_2$):

388: \begin{equation} H_{PT}(r)= \frac{2L^4}{9u^2 r^2} - 

389: \frac{2L^4}{81 u^4} \log \left(1 + \frac{9u^2}{r^2}\right )\ .

390: \end{equation}

391:  The small $r$ behavior of

392: $H_{PT}$ is $\sim \frac{1}{r^2}$.

393: This produces a metric singular at $r=0$ since the radius of

394: $S^2(\theta_2,\phi_2)$ blows up and the Ricci tensor is singular.

395:  Imposing the symmetry that $H$ has only radial dependence corresponds not to having a stack of

396:  D3-branes at a point (which would necessarily break the $SU(2)$ symmetry in $\theta_2,\phi_2$)

397:  but rather having the branes smeared out uniformly on the entire two sphere at the origin.

398:  The origin of this singularity is precisely the smearing of the D3-brane charge.

399:  In Section~\ref{sec-resolved}, we confirm this by constructing the solution corresponding to

400:  localized branes and find that there is no singularity.

401: 

402: \section{Flows on the Singular Conifold}

403:  \label{sec-singular}

404: 

405: 

406: 

407: Let us consider the case when the stack of D3-branes is moved away

408: from the singular point of the conifold. Since the branes are at a

409: smooth point on the conifold, we expect the near brane geometry to

410: become $AdS_5 \times S^5$ and thus describe ${\cal N} = 4$ $SU(N)$

411: SYM theory. The warp factor $H(r,Z)$ can be written as an expansion

412: in harmonics on $T^{1,1}$ starting with the leading term $1/r^4$

413: followed by higher powers of $1/r$. Thus, the full solution still

414: looks like $AdS_5 \times T^{1,1}$ at large $r$, but further terms in

415: the expansion of the warp factor change the geometry near the branes

416: to $AdS_5 \times S^5$. Such a SUGRA solution describes the RG flow

417: from the ${\cal N} = 1$ $ SU(N) \times SU(N)$ theory in the UV to

418: the ${\cal N}=4$ $SU(N)$ SYM in the IR. We will confirm this

419: explicitly through the computation of the general Green's function

420: on the conifold. We display the series of perturbations of the

421: metric and interpret these normalizable solutions in terms of

422:  VEVs in the gauge theory for a series of operators using the setup of Appendix A.

423: This was originally studied in \cite{KW2} where a restricted class of chiral

424: operators was considered.

425: 

426: \begin{figure}[t]

427: \centering

428: \includegraphics{Fig1.eps}

429: \caption{A stack of D3-branes warping the singular conifold}

430: \end{figure}

431: 

432: Let us place the stack at a point $(r_0,Z_0)$ on the singular

433: conifold. We rewrite (\ref{greendef0}) as

434: \begin{eqnarray}

435: \label{Green}

436: \label{SGf}

437: \Delta H = {\Delta_r} H + \frac{ \Delta_Z}{r^2} H & = & - \frac{{\cal C}}{\sqrt{g}}\delta( r - r_0)\Pi_i \delta^5(Z_i -Z_{0i}) \nonumber \\

438: & \equiv& - \frac{{\cal C}}{\sqrt{g_r}} \delta( r - r_0) \delta_{A}(Z - Z_0)

439: \end{eqnarray}

440: where $\Delta_r = \frac{1}{\sqrt{g}} \: \partial_r \left(\sqrt{g} \: \partial_r \right)$ is the radial Laplacian, $\Delta_{Z}$ the remaining angular laplacian. In the second line, $g_r$ is defined to have the radial dependence in $g$ and the angular delta function $\delta_{A}(Z - Z_0)$ is defined by absorbing the angular factor $\sqrt{g_5} = \sqrt{g/g_r}$. In this section, we have $ \sqrt{g} = \frac{1}{108} r^5 \sin \theta_1 \sin \theta_2$ and we take $\sqrt{g_r} = r^5$.

441: 

442: The eigenfunctions $Y_I(Z)$ of the angular laplacian on $T^{1,1}$ can be classified by a set $I$

443: of symmetry charges since $T^{1,1}$ is a coset space \cite{Gubser,Ceresole}. The eigenfunctions $Y_I$ are constructed explicitly in the appendix, including using the $a_i,b_j$ coordinates which will facilitate the comparison with the gauge theory below.

444: If we normalize these angular eigenfunctions as,

445: \begin{eqnarray}

446: \label{Ynorm}

447: \int \; Y^*_{I_0}(Z) Y_I(Z) \;\; \sqrt{g_5} \; d^5 \varphi_i & = & \delta_{I_0, I}

448: \end{eqnarray}

449: we then have the complementary result,

450: \begin{eqnarray}

451: \label{deltaexp}

452: \sum_I Y^*_I(Z_0) Y_I(Z) = \frac{1}{\sqrt{g_5}} \; \delta(\varphi_i - \varphi_{0i}) \equiv \delta_{A}(Z - Z_0).

453: \end{eqnarray}

454: 

455: We expand the $\delta_{A}(Z - Z_0)$ in (\ref{Green}) using (\ref{deltaexp}) and see that the Green's function can be expanded as,

456: \begin{eqnarray}

457: \label{Greenexp}

458:   H & = & \displaystyle\sum_I  \; H_I(r, r_0) \; Y_I(Z) \; Y^*_I(Z_0)

459: \end{eqnarray}

460: which reduces (\ref{Green}) to the radial equation,

461: \begin{eqnarray}

462: \displaystyle{ \label{eq2}

463:  \frac{1}{r^5} \frac{\partial}{\partial r} \left( r^5 \frac{\partial}{\partial

464: r} H_I\right) -  \frac{E_I}{r^2} H_I = - \frac{{\cal C}}{r^5}\delta( r - r_0)

465: }

466: \end{eqnarray}

467: where $ \Delta_Z Y_I(Z) = - E_I Y_I(Z)$ (see appendix A for details of $E_I$.)

468: 

469: As is easily seen, the solutions to this equation away from $ r = r_0$ are

470: 

471: $$

472: H_I \; = \; {A_\pm \; r^{c_\pm}}, \textrm{ where } \; \; c_\pm = -2\pm \sqrt{E_I +4}.

473:  $$

474: The constants $A_\pm$ are uniquely determined  integrating (\ref{eq2}) past

475: $r_0$. This determines $H_I$ and we put it all together to get the

476: solution to (\ref{SGf}), the Green's function on the singular

477: conifold

478: \begin{equation} \label{singgreen}

479: H(r ,Z ; r_0, Z_0) =   \sum_I \frac{{\cal C}}{2 \sqrt{E_I +4} } \;

480: Y^*_I(Z_0) Y_I(Z) \; \times

481:  \left\{ \begin{array}{ll}

482:   \displaystyle{\frac{1}{r^4_0} \left(\frac{r}{r_0}\right)^{c_I} }& r \leq r_0 \\

483:  \\

484:   \displaystyle{\frac{1}{r^4} \left(\frac{r_0}{r}\right)^{c_I} } & r \geq r_0 \ ,\\

485: \end{array} \right.

486: \end{equation}

487: where $c_I = c_+$. The term with $E_I=0$ gives $L^4/r^4$ where 

488: \begin{equation}

489: L^4 = \frac{{\cal C} }{4 \mathbf{Vol}(T^{1,1})}=

490: \frac{27 \pi g_s N (\alpha')^2}{4}

491: \ .\end{equation}

492: Since $E_I = 6(l_1(l_1+1) + l_2(l_2+1) - R^2/8)$, there are $(2l_1+1) \times (2l_2+1)$ terms

493: with the same $E_I$ and hence powers of $r$ and factors.

494: Also note that when $l_1 = l_2 = \pm \frac{R}{2}$, $c_I$ is a rational number and

495: these are related to (anti) chiral superfields in the gauge theory.

496: 

497: 

498: We can argue that the geometry near the stack (at $r_0,Z_0$) is actually a long 

499: $AdS_5 \times S^5$ throat.

500: We observe that $H$ must behave as $L^4/y^4$ near the stack 

501: (where $y$ is the distance between $(r,Z)$ and $(r_0,Z_0)$) since it is the solution 

502: of the Green's function and locally, the manifold looks flat and is 6 dimensional.

503: This leads to the usual $AdS_5\times S^5$ throat. We show this explicitly in Appendix B.

504: The complete metric thus describes holographic RG flow from

505:   $AdS_5 \times T^{1,1}$  geometry in the UV to $AdS_5 \times S^5$ in the IR. 

506: Note, however that

507: this background has a conifold singularity at $r=0$.

508: 

509: \subsection*{Gauge theory operators}

510: Let the stack of branes be placed at a point $a_i,b_j$ on the conifold. Then consider assigning the VEVS, $A_i = a_i^* 1_{N\times N}, B_j=b_j^* 1_{N\times N}$, i.e the prescription

511:  \beq \label{singvevs}

512:  Z_0  = \left( \begin{array}{cc}  a_1 b_1 & a_1 b_2 \\ a_2 b_1 & a_2 b_2 \\ \end{array} \right)  \quad \Longleftrightarrow  \quad \begin{array}{cc}

513: A_1 = a_1^* 1_{N\times N}, & A_2 = a_2^* 1_{N\times N}, \\ B_1 = b_1^* 1_{N\times N}, & B_2 = b_2^* 1_{N\times N}.  \end{array}

514:  \eeq

515:  In the appendix, we construct operators ${\cal O}_I$ transforming with the symmetry charges $I$. From the similar construction of the operator ${\cal O}_I$ and $Y_I(Z)$ (compare (\ref{compwvf}) and  (\ref{operatorL})), this automatically leads to a VEV proportional to $Y^*_I(Z_0)$ for the operator ${\cal O}_I$.

516: 

517: Meanwhile, the linearized perturbations

518: of the metric are determined by binomially expanding  $\sqrt{H}$ in (\ref{singmetric})

519: and considering terms linear in $Y_I(Z)$. These are easily seen to be of the form $

520: Y^*_I(Z_0) Y_I(Z) { \left(\frac{r_0}{r}\right)^{c_I} }$.

521:  From its form and symmetry properties, we conclude that it is the dual to the above VEV,

522: \beqa

523:  Y^*_I(Z_0) Y_I(Z) \; \times  \displaystyle{ \left(\frac{r_0}{r}\right)^{c_I} }  \quad \Longleftrightarrow \quad  \langle {\cal O}_I \rangle \propto Y^*_I(Z_0) \; r_0^{c_I}.

524: \eeqa

525: 

526: This is the sought relation between normalizable perturbations and operator VEV's.

527: For a general position of the stack $(r_0,Z_0)$, all $Y_I^*(Z_0)$ are non-vanishing. Being a coset space, we can use the symmetry of $T^{1,1}$, to set the D3-branes to lie at any specific point without loss of generality.

528: For example, consider the choice

529: \beq

530: Z_0 = \left( \begin{array}{cc}  a_1 b_1 & a_1 b_2 \\ a_2 b_1 & a_2 b_2 \\ \end{array} \right)  = \left( \begin{array}{cc} 1 & 0 \\ 0 & 0 \end{array} \right) \qquad \Rightarrow \qquad a_1 = b_1 = 1 , \; a_2 = b_2 = 0.

531: \eeq

532: 

533: Using (\ref{compwvf}) and (\ref{wvfdefn}) for $Y_I$, we find that $Y_I(Z_0) = 0$ unless $m_1 = m_2 = R/2$ and for these non-vanishing $Y_I$ we get,

534: \beq \label{specialfluct}

535: Y_I(Z_0) \sim  a_1^{l_1+\frac{R}{2}} \bar{a}_1^{l_1-\frac{R}{2}} b_1^{l_2+\frac{R}{2}} \bar{b}_1^{l_2-\frac{R}{2}}

536: \eeq

537:  If we give the VEVs $A_1 = B_1 = 1_{N\times N}, A_2 = B_2 = 0$, 

538: we get $\langle TrA_1 B_1\rangle \neq 0$ and all other $\langle TrA_i B_j \rangle = 0$. 

539: In fact, by this assignment, the only gauge invariant operators with non-zero vevs are 

540: the ${\cal O}_I$ with $m_1=m_2 = R/2$. These are precisely the operators dual to 

541: fluctuations $Y_I(Z)$ that have non-zero coefficient $Y_I^*(Z_0)$ as was seen in 

542: (\ref{specialfluct}).

543: 

544:  The physical dimension of this operator (at the UV fixed point) is read off as $c_I$ from the metric fluctuation - a supergravity prediction for strongly coupled gauge theory. (Above, $r_0$ serves as a scale for dimensional consistency.) In \cite{KW2}, the (anti) chiral operators were discussed ($l_1 = l_2 = \pm \frac{R}{2}$) . These have rational dimensions but as we see here, for any position of the stack of D3-branes, other operators (with generically irrational dimensions) also get vevs. For example, the dimension of the simplest non-chiral operator ($I \equiv l_1 =1, l_2 =0, R =0$) is $2$ but when $I \equiv l_1=2,l_2=0, R=0$, ${\cal O}_I$ has dimension $2(\sqrt{10}-1)$. This interesting observation about highly non-trivial scaling dimensions in strongly coupled gauge theory was first made in \cite{Gubser}.

545: 

546:  When operators $A_i,B_j$ get vevs as in (\ref{singvevs}), the $SU(N) \times SU(N)$ gauge

547: group is broken down to the diagonal $SU(N)$. The bifundamental fields

548: $A,B$ now become adjoint fields. With one linear combination of fields having a VEV,

549: we can expand the superpotential

550: $W \sim {\rm Tr} \det A_i B_j =

551: {\rm Tr} (A_1 B_1 A_2 B_2 - A_1 B_2 A_2 B_1)$ of the $SU(N)\times SU(N)$ theory to find that it

552: is of the form ${\rm Tr} (X[Y,Z])$ in the remaining adjoint fields \cite{KW}.

553: This is exactly ${\cal N} = 4$ $SU(N)$ super Yang-Mills, now obtained through symmetry breaking in

554: the conifold theory. This corresponds to the $AdS_5 \times S^5$ throat we found on the

555: gravity side near the source at $r_0,Z_0$.

556: 

557:  Thus we have established a gauge theory RG flow from ${\cal N}=1$ $SU(N) \times SU(N)$ theory in the UV to

558: ${\cal N}=4$ $SU(N)$ theory in the IR. The corresponding gravity dual was constructed and found to be asymptotically $AdS_5 \times T^{1,1}$ (the UV fixed point) but developing a $AdS_5 \times S^5$ throat at the other end of the geometry (the IR fixed point). The simple example is generalized to the resolved conifold in the next section.

559: 

560: \section{Flows on the Resolved Conifold}

561: \label{sec-resolved}

562: In this section we use similar methods to construct the Green's function on the resolved conifold and corresponding

563: warped solutions due to a localized stack of D3-branes.

564: We will work out explicitly

565: the $SU(2)\times U(1)\times U(1)$ symmetric RG flow

566: corresponding to a stack of D3-branes localized on the finite $S^2$ at $r=0$.

567: Such a solution is dual to giving a VEV to just one bi-fundamental field, e.g. $B_2$, which

568: Higgses the ${\cal N} = 1$ $SU(N)\times SU(N)$ gauge theory theory to

569: the ${\cal N}=4$ $SU(N)$ SYM. We also show

570: how the naked singularity found in \cite{PandoZayas} is removed through the localization of the D3-branes.

571: 

572:   The supergravity metric is of the form (\ref{warpedresolved}).

573: The stack could be placed at non-zero $r$ but in this case,

574: the symmetry breaking

575: pattern is similar in character to the singular case discussed above.

576: The essence of what is new to the resolved conifold is best captured with the stack placed at

577: a point on the

578: blown up $S^2$ at $r=0$; this

579: breaks the $SU(2)$ symmetry rotating $(\theta_2,\phi_2)$ down to a $U(1)$.

580: The branes also preserve the $SU(2)$ symmetry rotating $(\theta_1,\phi_1)$ as well as the

581: $U(1)$ symmetry corresponding to the shift of $\psi$.

582: On the other hand, the $U(1)_B$ symmetry is broken because

583: the resolved conifold has no non-trivial three-cycles \cite{KW2}. Thus the warped

584: resolved conifold background has

585: unbroken $SU(2) \times U(1) \times U(1)$ symmetry.

586: 

587: To match this with the gauge theory, we first recall that in the absence of VEV's

588: we have $SU(2) \times SU(2) \times U(1)_R \times U(1)_B$ where the $SU(2)$'s act on

589: $A_i,B_j$ respectively, the $U(1)_R$ is the R-charge ($R_A = R_B = 1/2$) and

590: $U(1)_B$ is the baryonic symmetry, $A \to e^{i \theta} A, B \to e^{-i \theta} B$.

591: As noted above, the VEV $B_2 = u 1_{N\times N}, B_1 = A_i = 0$ corresponds to placing the branes

592: at a point on the blown-up 2-sphere. This clearly leaves one of the $SU(2)$ factors unbroken.

593: While $U(1)_R$ and $U(1)_B$ are both broken by the baryonic operator $\det B_2$,

594: their

595: certain $U(1)$ linear combination remains unbroken.

596: Similarly, a combination of $U(1)_B$ and the

597: $U(1)$ subgroup of the other $SU(2)$, that rotates the $B_i$ by phases,

598: remains unbroken.

599: Thus we again have $SU(2) \times U(1) \times U(1)$ as the unbroken symmetry,

600: consistent with the warped resolved conifold solution.

601:   Since the baryon operator $\det B_2$ acquires a VEV while no chiral

602: mesonic operators do (because $A_1=A_2=0$), the solutions found in this section are dual to

603: a ``baryonic branch'' of the CFT (see \cite{Benvenuti} for a discussion of such branches).

604: 

605: 

606: \subsection*{Solving for the warp factor}

607: 

608: 

609: Since the resolution of the conifold preserves the $SU(2)_L\times SU(2)_R\times U(1)_\psi$ symmetry (where $U(1)_\psi$ shifts $\psi$),

610: the equation for Green's function $H$ looks analogous to (\ref{SGf}) for the resolved conifold,

611: \begin{equation}

612: \label{RGf}

613: \frac{1}{r^3(r^2+6 u^2)} \frac{\partial}{\partial r} \left( r^3(r^2+6 u^2) \kappa(r) \frac{\partial}{\partial r}

614: H\right) +  \mathbf{A} H   =

615:  - \frac{{\cal C}}{r^3(r^2 + 6 u^2)}\delta( r - r_0) \;\; \delta_{T^{1,1}}(Z - Z_0)

616: \end{equation}

617: where

618: \begin{equation}

619: \label{reseq2}

620: \mathbf{A} H  =  6 \frac{\Delta_1}{r^2} H  +  6 \frac{\Delta_2}{r^2 + 6 u^2} H+ 9 \frac{\Delta_R}{\kappa(r) r^2}  H

621: \end{equation}

622: and $\Delta_i$ , $\Delta_R$ are defined in the appendix. ($\Delta_i$ are $S^3$ laplacians and

623: $\Delta_R = \partial_\psi^2$. Note that $6 \Delta_1 + 6 \Delta_2 + 9 \Delta_R = \Delta_{T^{1,1}}$).

624: %Again, the normalization factor $C= (2 \pi)^4 g_s N (\alpha')^2$.

625: 

626: 

627:  This form of the ${\bf A}$ is fortuitous and allows us to use the $Y_I$ from the singular conifold,

628: since $Y_I$ are eigenfunctions of each of the three pieces of ${\bf A} $ above.

629: We could solve it for general $r_0$, but $r_0 = 0$ is a particularly simple case that is of primary interest

630: in this paper.

631: 

632: Since (\ref{RGf}) involves the same $ \delta_{T^{1,1}}(Z - Z_0)$ as the singular case,

633: we can expand $H$ again in terms of the angular harmonics and radial functions as $H = \sum_{I}  H_I(r,r_0)

634: Y_I(Z) Y^*_I(Z_0)$ to find the radial equation,

635: \begin{eqnarray}

636:  \label{resradialeqn}

637: &-& \frac{1}{r^3(r^2+6 u^2)} \frac{\partial}{\partial r} 

638: \left( r^3 (r^2 + 6 u^2) \kappa(r) \frac{\partial}{\partial

639: r} H_I\right)  \nonumber \\

640: \quad &+& \left( \frac{6 (l_1(l_1+1) - R^2/4)}{r^2} +

641: \frac{6 (l_2(l_2+1) - R^2/4)}{r^2 + 6 u^2} + \frac{9 R^2/4}{\kappa(r)r^2} \right) H_I =  

642: \frac{{\cal C}}{r^3(r^2+6 u^2)}\delta(r - r_0) .

643: \end{eqnarray}

644: This equation can be solved for $H_I(r)$ exactly in terms of some special functions. If we place the

645: stack at $r_0 = 0$, i.e at location $(\theta_0,\phi_0)$ on the blown up $S^2$, then an additional

646: simplification occurs.

647: The warp factor $H$ must be a singlet under the $SU(2)\times U(1)_\psi$ that rotates $(\theta_1,\phi_1)$ and $\psi$ since

648: these have shrunk at the point where the branes are placed.

649: Hence we only need to solve this equation for $l_1= R = 0$, $l_2=l$.

650: 

651: The two independent solutions (with convenient normalization) to the homogeneous 

652: equation in this case,

653: in terms of the hypergeometric function ${}_2F_1$, are

654: \beqa \label{resradialsoln}

655: H_l^A(r) & = & \frac{2 }{9u^2} \frac{C_\beta}{r^{2+2\beta}} \;\; 

656: {}_2F_1\left(\beta,1+\beta ;1+ 2

657: \beta;-\frac{9 u^2}{r^2}\right)  \nonumber \\

658: H_l^B(r) &\sim & {}_2F_1 \left(1-\beta,1+\beta;2;-\frac{r^2}{9u^2}\right)

659: \eeqa

660: where

661: \beqa \label{normfactor}

662: C_\beta = \frac{(3u)^{2\beta} \Gamma(1+\beta)^2}{\Gamma(1+2 \beta)} \ ,\qquad 

663: \beta = \sqrt{1+(3/2)l(l+1)}\ .

664: \eeqa

665: These two solutions have the following asymptotic behaviors,

666: \beqa

667: \frac{2}{9 u^2 r^2}+ \frac{4 \beta^2}{81 u^4} \ln r + {\cal O}(1)

668:  \quad \stackrel{0\leftarrow r}{\longleftarrow} 

669: \quad &H_l^A(r)& \quad \stackrel{r

670: \rightarrow \infty}{\longrightarrow} \quad \frac{2  C_\beta}{9 u^2 r^{2 + 2\beta}} \\

671: \nonumber \\

672: {\cal O}(1) \quad \stackrel{0\leftarrow r}{\longleftarrow} \quad &H_l^B(r) &\quad

673: \stackrel{r \rightarrow \infty}{\longrightarrow} \quad 

674: {\cal O}\left (r^{- 2 + 2\beta}\right )

675: \eeqa

676: 

677: To find the solution to (\ref{resradialeqn}) 

678: with the $\delta(r-r_0)$ on the RHS, we need to match the 

679: two solutions at $r=r_0$ as well as satisfy the condition on 

680: derivatives obtained by integrating past $r_0$. Since we are interested 

681: in normalizable modes, we use $H_l^A(r)$ for $r > r_0$ and $H_l^B(r)$ for $r<r_0$.

682: Finally, we take $r_0 = \epsilon$ and take the limit $\epsilon \rightarrow 0$ 

683: (since the stack of branes is on

684: the finite $S^2$). We find simply that $H_l(r) = {\cal C} H_l^A(r)$ due to the 

685: normalization chosen earlier in (\ref{resradialsoln}).

686: Putting it all together, we find,

687: \beqa \label{resfullsoln}

688:    H(r,Z;r_0=0,Z_0)  & = & {\cal C} \sum_I Y^{*}_I(Z_0) H_I^A(r) Y_I(Z)

689: \eeqa

690: where only the $l_1=0, R=0$ harmonics

691: contribute since the stack leaves $SU(2) \times U(1)\times U(1)$ symmetry unbroken.

692: In this situation, the $Y_I$ wavefunctions simplify to the usual $S^2$ spherical 

693: harmonics $\sqrt\frac{4 \pi}{\mathbf{Vol}(T^{1,1})} \; Y_{l,m}$.

694: 

695:  Let us take $b_i$ to describe the finite $S^2(\theta_2,\phi_2)$ while $a_j$ are associated with the $S^2$

696: that shrinks to a point. As reviewed in Section \ref{sec-conifold},

697: the resolved conifold can be described with $a,b$ variables governed by the constraint (\ref{resconstr}),

698: where $u$ is the measure of resolution, the radius of the finite $S^2$.

699: The position of the branes on the finite sphere can be parametrized as

700: $b_1 = u \sin \frac{\theta_0}{2} e^{-i \phi_0/2},

701: b_2 = u \cos \frac{\theta_0}{2} e^{i \phi_0/2}$ and $a_1=a_2=0$

702: (since the branes do not break the $SU(2)$ symmetry rotating

703: the $a$'s).

704: Then,

705: \beqa \label{resolvedgreen}

706:     H(r,Z;r_0=0,Z_0=(\theta_0,\phi_0))  = 

707: 4 \pi L^4 \sum_{l, m} H_l^A(r) \; Y^*_{l,m}(\theta_0, \phi_0)

708:  Y_{l,m}(\theta_2, \phi_2).

709: \eeqa

710: Without a loss of generality, we can place

711: the stack of D3-branes at the north pole ($\theta_0 = 0$) of the 2-sphere. 

712: Then (\ref{resolvedgreen}) simplifies further: 

713: only $m=0$ harmonics contribute and we get the explicit 

714: expression for the warp factor which is one of our main results,

715: \beqa \label{simpwarp}

716:     H(r,\theta_2)  =   

717: L^4 \sum_{l=0}^\infty  \; (2 l + 1) H_l^A(r) P_l(\cos \theta_2).

718: \eeqa

719: Now the two unbroken $U(1)$ symmetries are manifest as shifts of $\phi_2$ and $\psi$.

720: 

721: The `smeared' singular solution found in \cite{PandoZayas}

722: corresponds to retaining only the $l=0$ term in this sum. Indeed, we

723: find that \beq \label{zmode}

724: H_0^A(r) =  \frac{2 C_1}{9 u^2 r^{4}} \;\;

725: {}_2F_1\left(1,2 ;3 ;-\frac{9 u^2}{r^2}\right) = \frac{2}{9u^2 r^2}

726: - \frac{2}{81 u^4} \log \left(1 + \frac{9u^2}{r^2}\right) \eeq in

727: agreement with \cite{PandoZayas}. Fortunately,

728: if we consider the full sum over modes appearing in

729: (\ref{resolvedgreenN}), the geometry is no longer singular. 

730: The leading term in the warp factor (\ref{simpwarp}) at small $r$ is

731: \beqa \label{resolvedgreenN}

732: {2L^4\over 9u^2 r^2}  

733: \sum_{l=0}^\infty  \; (2 l + 1) P_l(\cos \theta_2)= {4 L^4\over 9 u^2 r^2}

734: \delta ( 1-\cos \theta_2) 

735: \eeqa

736: This shows that away from the north pole the $1/r^2$ divergence of the warp factor

737: cancels. Similarly, after summing over $l$

738: the term $\sim \ln r$ cancels away from the north pole. This implies that

739: the warp factor is finite at $r=0$ away from the north pole.

740: However, at the north pole it diverges as expected.

741:  Indeed, since the branes are now localized at a smooth point on 

742: the 6-manifold (all points on the resolved conifold are smooth), 

743: very near the source $H$ must again be of the form $L^4/y^4$ where 

744: $y$ is the distance from the source. This is shown explicitly in Appendix B.

745: Writing the local metric in the form $dy^2 + y^2 d\Omega^2_{S^5}$ near the source, we

746: get the $AdS_5 \times S^5$ throat, avoiding the singularity found in \cite{PandoZayas}.

747: 

748: \begin{figure}[t]

749: \centering

750: \includegraphics{Fig2.eps}

751: \caption{A stack of D3-branes warping the resolved conifold}

752: \end{figure}

753: 

754: \subsection*{ Gauge theory operators}

755: With the branes placed at the point $b_1 = u \sin \frac{\theta_0}{2} e^{-i \frac{\phi_0}{2}}, 

756: b_2 = u \cos \frac{\theta_0}{2} e^{i \frac{\phi_0}{2}}$ , $a_1=a_2=0$ on the finite $S^2$, 

757: consider the assignment of VEVs, $B_1 =  u \sin \frac{\theta_0}{2} e^{i \frac{\phi_0}{2}} 1_{N\times N} , 

758: B_2 = u \cos \frac{\theta_0}{2} e^{-i \frac{\phi_0}{2}} 1_{N\times N}, A_1 = A_2 = 0$.

759: 

760: The linearized fluctuations compared to the leading term $1/r^4$ ($l_2 = 0$) are of the form,

761: \beqa \label{resfluctdim}

762: Y_I^*(Z_0) Y_I(Z) r^4 H_I(r)  \rightarrow  Y_I^*(Z_0) Y_I(Z) (\frac{1}{r})^{c_I} \quad ( r \gg a)

763: \eeqa

764: where as earlier, $c_I = 2 \sqrt{1+(3/2)l_2(l_2+1)} - 2$.

765: 

766: With this assignment of VEVs above, operators ${\cal O}_I$ with $l_1

767: = R = 0$ acquire VEVs. In the notation $|l_2,m_2;R>$ of Appendix A,

768: \beqa \label{resoperatorvevs} <{\cal O}_I> &=& <Tr

769: |l_2,m_2;0\rangle_B  > \neq 0. \eeqa For example, when $l_2 =

770: 1,m_2=0$ above, $\langle{\cal O}_I\rangle= \langle Tr B_1 \bar{B}_1

771: - B_2 \bar{B}_2\rangle = u^2( \sin^2 \frac{\theta_0}{2} - \cos^2

772: \frac{\theta_0}{2}) $. By construction of ${\cal O}_I$ and $Y_I$, it

773: is clear that $<{\cal O}_I> \sim Y_I^*(Z_0)$ and is dual to the

774: metric fluctuation above. We can read off the dimensions of these

775: operators as $c_I$

776: % = 2\beta -2, \beta = \sqrt{1+ (3/2)l_2(l_2+1)}$

777: from the large $r$ behavior of the fluctuation (\ref{resfluctdim}).

778: For the $l_2=1$ operator as above, the exact dimension is $2$ (the

779: classical value) because the operator is a superpartner of a conserved

780: current \cite{Itzhaki}. Similarly, the dimension 2 of ${\cal U}$

781: is protected against quantum corrections 

782: because of its relation to a conserved baryonic current \cite{KW2}.

783:   When one expands $H_I(r)$ at large $r$, one finds sub-leading terms in addition to

784: $1/r^{c_I}$ shown above. These terms, which do not appear for the singular conifold,

785: increase in powers of $1/r^2$ and hence describe a series of operators with

786: the same symmetry $I$ but dimension increasing in steps of 2 from $c_I$.

787: These modes appear to correspond to VEV's for the operators

788: $ {\rm Tr}  {\cal O}_I {\cal U}^n $. It would be interesting to 

789: investigate such operators

790: and their dimensions further.

791: 

792: Hence we have an infinite series of operators that get VEV's in

793: the gauge theory dual to the warped resolved conifold.

794: These are in addition to the basic operator ${\cal U}$ which gets

795: a VEV

796: due to the asymptotics of the unwarped resolved conifold

797: metric itself \cite{KW2}. 

798: The operator ${\cal U}$ would get the same VEV of 

799: $u^2$ for any position of the brane on the $S^2$ while the VEV's

800: for the infinite series of operators ${\cal O}_I$ depend on the position.

801: We also note that ${\cal U} = u^2$ is the gauge dual of the

802: constraint $|b_1|^2 + |b_2|^2 - |a_1|^2 - |a_2|^2 = u^2$ 

803: defining the resolved conifold in Section

804: \ref{sec-conifold}.

805: 

806: Lastly, we verify that the gauge theory does flow in the infrared to ${\cal N}=4$ $SU(N)$ SYM.

807: Without loss of generality, we can take the stack of branes to lie on the north pole of the finite sphere

808: ($B_2 = u 1_{N \times N}, B_1 = 0$). As in the singular case, $B_2 = {u} 1_{N\times N}$ breaks the

809: $SU(N) \times SU(N)$ gauge group down to $SU(N)$, all the chiral fields now transforming in the

810: adjoint of this diagonal group. Consider the ${\cal N} =1$ superpotential

811: $W \sim {\rm Tr} \det A_i B_j = Tr (A_1 B_1 A_2 B_2 - A_1 B_2 A_2 B_1)$.

812: When $B_2 \propto 1_{N\times N}$,

813: the superpotential reduces to the $ {\cal N}=4$ form,

814:   \beq

815:   W = \lambda  {\rm Tr} (A_1 B_1 A_2 - A_1A_2 B_1) = \lambda {\rm Tr} (A_1[B_1,A_2]).

816:   \eeq

817: This confirms that the gauge theory flows to the $ {\cal N}=4$ $SU(N)$ SYM theory in the infrared.

818: 

819: 

820: \subsection*{Baryonic Condensates and Euclidean D3-branes}

821: 

822: Here we present a calculation of the baryonic VEV using the dual string theory on

823: the warped resolved conifold background.\footnote{We are indebted to E. Witten

824: for his suggestion that led to the calculation presented in

825: this section.} 

826: A similar question was addressed for the

827: cascading theories on the baryonic branch where the baryonic condensates are

828: related to the action of a Euclidean D5-brane wrapping

829: the deformed conifold \cite{Wittenunpub,BDK}. 

830: In this section we present an analogous construction for the

831: warped resolved conifolds, which are asymptotic to $AdS_5\times T^{1,1}$. 

832: 

833: The objects in $AdS_5\times T^{1,1}$ that are dual to baryonic operators

834: are D3-branes wrapping 3-cycles in $T^{1,1}$ \cite{GK}. Classically, the

835: 3-cycles dual to the baryons made out of the $B$'s are located at fixed

836: $\theta_2$ and $\phi_2$ (quantum mechanically, one has to carry out

837: collective coordinate quantization and finds wave functions of spin $N/2$

838: on the 2-sphere). To calculate VEV's of such baryonic operators, we need

839: to consider Euclidean D3-branes which at large $r$ wrap a 3-cycle at fixed $\theta_2$

840: and $\phi_2$. In fact, the symmetries of the calculation suggest that the

841: smooth 4-cycle wrapped by the Euclidean D3-brane is located at fixed

842: $\theta_2$ and $\phi_2$, and spans the $r$, $\theta_1$, $\phi_1$ and $\psi$ directions. 

843: In other words, the Euclidean D3-brane wraps the ${\bf R}^4$ fiber of the 

844: ${\bf R}^4$ bundle over $S^2$ (recall that the resolved conifold is such a bundle).

845: 

846: The action of the D3-brane will be integrated up to a radial cut-off $r_c$, and

847: we identify $e^{-S(r_c)}$ with the classical field dual to the baryonic operator.

848: The Born-Infeld action is

849: \beq

850: S_{BI}= T_3 \int d^4\xi \sqrt{ g} 

851: \ ,

852: \eeq

853: where $g_{\mu\nu}$ is the metric induced on the D3 world volume. 

854: %Reinstating the normalization factor $L^4 = \frac{27 \pi g_s N (\alpha')^2}{4}$,

855: We find 

856: \beq

857: S_{BI}= 

858: {3N\over 4L^4}\int_0^{r_c} dr r^3 H(r,\theta_2)= 

859: {3N\over 4}\int_0^{r_c} dr r^3  

860: \sum_{l=0}^\infty  \; (2 l + 1) H_l^A(r) P_l(\cos \theta_2)

861: \ .

862: \eeq

863: %where $H(r,\theta_2)$ is the warp factor normalized as in 

864: %(\ref{simpwarp}). 

865: 

866: The $l=0$ term (\ref{zmode}) needs to be evaluated separately since it contains

867: a logarithmic divergence:\footnote{A careful holographic renormalization of

868: divergences for D-brane actions was considered in \cite{Karch}. 

869: We leave a similar construction in the present situation for

870: future work.}

871: \beq

872: \int_0^{r_c} dr r^3 H_0^A (r)= {1\over 4} + {1\over 2}

873: \ln \left (1+ {r_c^2\over 9 u^2} \right )

874: \ .

875: \eeq

876: For the $l>0$ terms the cut-off may be removed and we find

877: a nice cancellation involving the normalization (\ref{normfactor}):

878: \beq

879: \int_0^{\infty} dr r^3 H_l^A (r)= {2\over 3 l(l+1)}\ .

880: \eeq

881: Therefore,

882: \beq

883: \int_0^{\infty} dr r^3 \sum_{l=1}^\infty (2l+1) H_l^A (r) P_l (\cos \theta_2)

884: = {2\over 3}\sum_{l=1}^\infty  {2l+1\over l(l+1)} P_l (\cos \theta_2)

885: ={2\over 3} (-1 -2 \ln [\sin (\theta_2/2)])\ .

886: \eeq

887: This expression is recognized as the Green's function on a sphere.

888: Combining the results, and taking $r_c\gg u$, we find

889: \beq

890: e^{-S_{BI}} = \left ( {3 e^{5/12} u\over r_c }\right )^{3N/4} 

891: \sin^N (\theta_2/2)\ .

892: \eeq

893: %This is not the complete answer since we need to 

894: %include the Chern-Simons term describing

895: %the coupling of the D3-brane to the background $C_4$. Studying the structure of the

896: %self-dual 5-form field

897: %\beq

898: %F_5 = (1+ *) d H^{-1}\wedge dx^0\wedge dx^1\wedge dx^2\wedge dx^3

899: %\eeq 

900: %we find thatg there exists a component of 

901: %$C_4$ multiplying $d\theta_1\wedge d\phi_1\wedge d\psi\wedge dr$.

902: %While the integral of $C_4$ could be evaluated directly, we may use the

903: %holomorphy arguments similar to those in \cite{BDKMMM} to argue that

904: %$e^{-S_{CS}}= e^{-i N\phi_2/2}$, so that

905: %\beq

906: %e^{-S_{BI}-S_{CS}} = \left ( {3 e^{5/12} u\over r_c }\right )^{3N/4} 

907: %\sin^N (\theta_2/2) e^{-iN\phi_2/2}\ .

908: %\eeq

909: 

910: In \cite{GK} it was argued that the wave functions

911: of $\theta_2, \phi_2$, which arise though

912: the collective coordinate quantization of

913: the D3-branes wrapped over the 3-cycle $(\psi,\theta_1,\phi_1)$, correspond to

914: eigenstates of a charged particle on $S^2$ in the presence of a 

915: charge $N$ magnetic monopole. Taking the gauge potential 

916: $A_\phi =N (1+\cos \theta)/2, \ A_\theta=0$ we find that 

917: %The angular factor is recognized as a wave function on $S^2$ corresponding

918: the ground state wave function $\sim \sin^N (\theta_2/2)$ carries 

919: the $J=N/2, m=-N/2$ quantum numbers.\footnote{In

920: a different gauge this wave function would acquire a phase. In the string

921: calculation it comes from the purely imaginary Chern-Simons term in

922: the Euclidean D3-brane action.} 

923: These are the $SU(2)$ quantum numbers of ${\rm det} B_2$.

924: Therefore, the angular dependence of $e^{-S}$ identifies ${\rm det} B_2$

925: as the only operator that acquires a VEV, in agreement with the gauge theory. 

926: 

927: The power of $r_c$ indicates that the operator dimension is $\Delta=3N/4$,

928: which again corresponds to the baryonic operators.

929: The VEV depends on the parameter $u$

930: as $\sim u^{3N/4}$. This is not the same as the classical scaling that would give

931: ${\rm det} B_2= u^N $. The classical scaling is not obeyed because this is an

932: interacting theory where the baryonic operator acquires an anomalous dimension.

933: 

934: The string theoretic arguments presented in this section provide

935: nice consistency checks on the picture developed in this paper, and also

936: confirm that the Eucldean 3-brane can be used to calculate the baryonic condensate.

937: 

938: 

939: 

940: \section{$B$-field on the Resolved Conifold}

941: \label{sec-Bfield}

942:   Our warped resolved conifold solution written with no 

943: NS-NS $B$ field corresponds to a special isolated point in the space 

944: of gauge coupling constants. From \cite{HKO1},

945: the relation between coupling constants and the SUGRA background is known to be,

946: \beqa

947:  \frac{4 \pi^2}{g_1^2} + \frac{4 \pi^2}{g_2^2} &=& \frac{\pi}{g_s e^{\Phi}} \\

948:  \frac{4 \pi^2}{g_1^2} - \frac{4 \pi^2}{g_2^2} &=& \frac{1}{g_s e^{\Phi}} 

949: \left(\frac{1}{2\pi \alpha'} \int_{S^2} B_2 - \pi \right)

950: \eeqa

951: where $\Phi$ is the dilaton. Hence when $B=0$, $g_1$ is infinite.

952: 

953: Since the resolved conifold has a topologically non-trivial two cycle and we could turn on a 

954: $B$-field proportional to the volume of this cycle \cite{KW}:

955: \beqa

956: B_2 \sim \sin \theta_2 d\theta_2 \wedge d\phi_2.

957: \eeqa

958: Such a $B$-field would have no flux, $H=dB_2=0$, while still being non-trivial ($\int_{S^2} B_2 \neq 0$). 

959: Since there is no flux, the rest of the SUGRA solution remains 

960: untouched and we have a description of the gauge theory at generic coupling.

961: 

962: When the resolved conifold is warped by a stack of branes as we have

963: in this paper, the argument of \cite{KW2} continues to hold. A new

964: $AdS_5 \times S^5$ throat branches out at the point where the stack

965: is placed. This modifies the topology by introducing a new

966: non-trivial 5-cycle. However, the earlier two-cycle is untouched and

967: does not become topologically trivial. One way to see this is to

968: note that the new 5-cycle was the trivial cycle that could shrink to

969: a point at the place where the stack is placed. But the finite two

970: cycle of the resolution is topologically distinct from the cycles

971: that shrink here and hence it obviously survives the creation of a

972: new 5-cycle. Hence the fluxless NS-NS $B_2$ field above that naturally

973: exists on such a space can be used to describe the gauge theory at

974: generic coupling.

975: 

976: Had we considered a stack of D3-branes on the deformed conifold, the

977: situation would have been quite different, as emphasized in

978: \cite{KW2}. In that case, a fluxless $B_2$ field cannot be turned

979: on; therefore, there is no simple $SU(N)\times SU(N)$ gauge theory

980: interpretation for backgrounds of the form (\ref{warpedresolved})

981: with $ds_6^2$ being the deformed conifold metric, and $H$ the

982: Green's function of the scalar Laplacian on it. Of course, the

983: deformed conifold with a different warp factor created by self-dual

984: 3-form fluxes corresponds to the cascading $SU(kM)\times SU(k(M+1))$

985: gauge theory \cite{KS,KT}.

986: 

987: \section{Conclusions}

988: \label{sec-conc}

989:  We have constructed the SUGRA duals of the $SU(N)\times SU(N)$ conifold gauge theory with

990:  certain VEV's for the bi-fundamental fields. As discussed in \cite{KW2},

991:  the different vacua of the theory correspond to D3-branes localized on

992:   the singular as well as resolved conifold.

993:   Vacua with ${\cal U} = 0 $ describe the singular conifold with a localized stack of

994:   D3-branes;

995:   vacua with ${\cal U} \neq 0$ instead describe D3-branes localized on the conifold resolved

996:   through blowing up of a 2-sphere.

997: We constructed explicit SUGRA solutions corresponding to

998:   these vacua.

999: In particular, the

1000: solution corresponding to

1001: giving a VEV to only one of the fields in the gauge theory,

1002: $B_2 =u 1_{N\times N}$, while keeping

1003: $ A_i = B_1 = 0$,

1004: corresponds to 

1005: a certain warped resolved conifold. In this case the warp factor

1006: is given by the Green's function with a source at a point on the

1007: blown-up 2-sphere at $r=0$. The baryonic operator $\det B_2$ gets

1008: a VEV while no chiral mesonic operator does. This background is

1009: thus dual to a non-mesonic, or baryonic, branch of the CFT.

1010: To confirm this, we used the action of a Euclidean D3-brane 

1011: wrapping a 4-cycle in the resolved

1012: conifold, to calculate the VEV of the baryonic operator. 

1013: 

1014: The explicit SUGRA solution

1015: was determined and found to asymptote to $AdS_5 \times T^{1,1}$ in

1016: the large $r$ region. When one approaches the blown-up 2-sphere, the

1017: warp factor causes an $AdS_5 \times S^5$ throat to branch off at a

1018: point on the 2-sphere. Our calculation makes use of the explicit

1019: metric on the resolved conifold found in \cite{PandoZayas}. Our

1020: warped solution, with a localized stack of D3-branes, is completely

1021: non-singular in contrast to the smeared-brane solution obtained in

1022: \cite{PandoZayas}.

1023: 

1024:   The Green's functions on the singular and resolved conifolds were determined

1025: in detail

1026:   for the purpose of constructing the SUGRA solutions.

1027: These Green's functions are also useful in

1028:   brane models of inflation where they play a role in computing the one-loop corrections to

1029:   non-perturbative superpotentials (see \cite{Giddings,BDKMMM} for such an application).

1030:   The Green's functions were written using harmonics on $T^{1,1}$ in the $a,b$ variables on the conifold

1031:   (instead of the usual angular variables or the $z,w$ co-ordinates).

1032:   This facilitated the comparison with the explicit gauge theory operators that acquire

1033:   VEVs. 

1034: 

1035:    We see a number of possible extensions of our work. One of them deals with

1036: the AdS/CFT dualities

1037: based on the Sasaki-Einstein spaces $Y^{p,q}$ \cite{Gaunt1,Hanany}.

1038: Calculations similar to ours can be performed for the resolved cones over $Y^{p,q}$ manifolds

1039:    (for recent work, see \cite{ChenLuPope,Oota}). Harmonics in convenient co-ordinates similar to the ones constructed

1040:    here could perhaps be constructed using the bifundamental fields of these quiver gauge

1041:    theories. Again, the basic

1042: non-singular solutions will correspond to

1043:    a stack of branes at a point, and it would be interesting to

1044:    solve for the corresponding warp factors. One could also study

1045:    the resolved cone versions of the solutions found in

1046:    \cite{HEK}, which correspond to cascading gauge theories.

1047:    It is also possible to consider Calabi-Yau cones with blown-up

1048: 4-cycles \cite{PTseyt,Pal,Sfetsos,Hanany,Benini}.

1049: In \cite{Benvenuti}, the gauge theory operator whose VEV

1050: corresponds to blown-up 4-cycles of certain cones was identified.

1051: Perhaps the Green's function could be determined for a stack of branes on such 4-cycles,

1052:    giving the non-singular SUGRA dual of corresponding non-mesonic branches in the gauge theory.

1053: 

1054: \section*{Acknowledgements}

1055: We are indebted to Edward Witten for stimulating discussions that

1056: suggested the way to calculate the baryonic condensate.

1057: We also thank Sergio Benvenuti, David Kutasov, Dmitry Malyshev, Anand

1058: Murugan, Manuela Kulaxizi, Yuji Tachikawa, Justin Vazquez-Poritz 

1059: for useful discussions, and especially Arkady Tseytlin

1060: for discussions and comments on the manuscript. We are also grateful

1061: to Daniel Baumann, Marcus Benna, Anatoly Dymarsky, Juan Maldacena and Liam

1062: McAllister for collaboration on related work, and to Daniel Baumann

1063: for his help in making the figures. This research was supported in

1064: part by the National Science Foundation under Grant No. PHY-0243680.

1065: Any opinions, findings, and conclusions or recommendations expressed

1066: in this material are those of the authors and do not necessarily

1067: reflect the views of the National Science Foundation.

1068: 

1069: \appendix

1070: 

1071: \section{Eigenfunctions of the Scalar Laplacian on $T^{1,1}$}

1072: \label{sec-eigen}

1073: The main emphasis of this Appendix is on writing the harmonics on

1074: $T^{1,1}$ in a way that makes the connection with the dual gauge

1075: theory operators most transparent. The eigenfunctions of the scalar

1076: Laplacian on $T^{1,1}$ have been worked out in \cite{Gubser, Ceresole}. We first review this

1077: calculation and present the harmonics in angular variables on

1078: $T^{1,1}$. This form of the harmonics is useful for some purposes,

1079: such as in \cite{BDKMMM} where it was used to find the potential

1080: generated for a D3-brane moving on the conifold due to a wrapped D7.

1081: We then write the harmonics using the complex $a_i,b_j$ coordinates, generalizing the $z_i$ construction of \cite{KW2}, that makes the connection with the gauge theory manifest. We also construct the operators using $A_i,B_j$ with given symmetry charges, related to the harmonics through the AdS/CFT correspondance.

1082: 

1083: Since $T^{1,1}$ is a product of two $3-$spheres divided by a $U(1)$,

1084: the eigenfunctions are simply products of harmonics on two

1085: $3-$spheres, restricted by the fact that the two spheres share an

1086: angle $\psi$. The laplacian (defined by $\Delta_Z H =

1087: \frac{1}{\sqrt{g}} \partial_m (g^{mn} \sqrt{g}

1088: \partial_n H )$) on $T^{1,1}$ can be written in the following form,

1089: \begin{equation}

1090: {\label{eqA1} \Delta_Z = 6 \Delta_1 + 6 \Delta_2 + 9 \Delta_R }

1091: \end{equation}

1092: where

1093: \begin{equation}

1094: \displaystyle{

1095: \Delta_i = \frac{1}{\sin\theta_i} \partial_{\theta_i}  \;

1096:                (\sin \theta_i \; \partial_{\theta_i} \;\;) +

1097:                \left( \frac{1}{\sin\theta_i} \partial_{\phi_i} - \cot\theta_i

1098: \partial_\psi \right)^2

1099: }

1100: \end{equation}

1101: \begin{equation}

1102: \displaystyle{ \Delta_R = \partial^2_\psi }

1103: \end{equation}

1104: We can solve for the eigenfunctions through separation of variables,

1105: $$\displaystyle{

1106: Y_I(Z) \sim J_{l_1,m_1,R}(\theta_1) \; J_{l_2,m_2,R}(\theta_2) \; e^{i m_1 \phi_1 + i m_2

1107: \phi_2} \; e^{\frac{i R \psi}{2}}   }

1108: $$

1109: This leads to

1110: \begin{equation}

1111: \displaystyle{

1112: \frac{1}{\sin\theta} \partial_{\theta}  \; (\sin \theta \;

1113: \partial_{\theta} \; J_{lmR}(\theta)) -

1114:                \left( \frac{1}{\sin\theta} m - \cot\theta \frac{R}{2}

1115: \right)^2  J_{lmR}(\theta) = - E  J_{lmR}(\theta)

1116: }

1117: \end{equation}

1118: for both sets of angles. When $R=0$, this reduces to the equation for harmonics

1119: on $S^2$. For general integer $R$, this is closely related to the harmonic

1120: equation on $S^3$ in Euler angles $(\theta, \phi, \psi)$. The eigenvalues $E$

1121: are $l(l+1) - \frac{R^2}{4}$ as can be seen by comparing with Laplace's

1122: equation on $S^3$.

1123: 

1124: The solutions for $J_{lmR}$ are,

1125: 

1126: \begin{equation}

1127: \displaystyle{

1128: J_{lmR}^{A}(\theta) =  \sin^m \theta \; \; \cot^{\frac{R}{2}}

1129: \frac{\theta}{2} \;\; {}_2F_1 \left(-l +m, 1+l+m; 1+m -\frac{R}{2}; \sin^2\frac{\theta}{2} \right)

1130: }\end{equation}

1131: 

1132: \begin{equation}

1133: \displaystyle{

1134: J_{lmR}^{B}(\theta) = \sin^{\frac{R}{2}}\theta  \;\; \cot^m

1135: \frac{\theta}{2} \;\; {}_2F_1 \left(-l +\frac{R}{2}, 1+l+\frac{R}{2}; 1-m

1136: +\frac{R}{2}; \sin^2\frac{\theta}{2} \right)

1137: }\end{equation}

1138: 

1139:  Here ${}_2F_1$ is the hypergeometric function. If $m \leq R/2$, solution B is non-singular. If $m \geq R/2$, solution

1140: A is non-singular. (The solutions coincide when $m = R/2$).

1141: 

1142: Putting together these solutions, the spectrum is of the form $$ E_I = 6 \left( l_1(l_1+1) + l_2(l_2+1) - \frac{R^2}{8} \right) $$

1143: with eigenfunctions that transform under $SU(2)_A \times SU(2)_B$ as the spin $(l_1,l_2)$

1144: representation and under the shift of $\psi/2$ (which is $U(1)_R$ in the UV) with charge $R$.

1145:  Here $I$ is a multi-index with the data:

1146: $$

1147: I \; \equiv \; (l_1,m_1), \; (l_2,m_2), \: R

1148: $$

1149: with the following restrictions coming from existence of single valued regular solutions:

1150: \begin{itemize}

1151: \item[-] $l_1$  and $l_2$ both integers or both half-integers

1152: \item[-] $ R \in \mathbf{Z}$  with $ \frac{R}{2} \in \{-l_1, \cdots,

1153: l_1\} $ and $\frac{R}{2} \in \{-l_2, \cdots, l_2\}$

1154: \item[-] $ m_1 \in \{-l_1, \cdots, l_1\}$ and $ m_2 \in \{-l_2, \cdots,

1155: l_2\}$

1156: \end{itemize}

1157: As above $(l_1,l_2),R$ are the $SU(2)\times SU(2)$ spins and R-charge and

1158: $(m_1,m_2)$, the $J_z$ values under the two $SU(2)$s.

1159: 

1160: \subsection*{Harmonics in the $a,b$ basis}

1161: 

1162: In \cite{KW2}, the 'chiral' harmonics were constructed using the complex $z_i$ coordinates. We generalize this to construct harmonics by using the $a_i,b_j$ coordinates which facilitates the comparison with the gauge . We form the eigenfunction $Y_I$ in the $a,b$ basis by tensoring

1163: representations. As we wish to construct

1164: harmonics on the base $T^{1,1}$, we fix the radius $r$ of the conetheory

1165: by setting $ |a_1|^2+|a_2|^2=|b_1|^2+|b_2|^2=1$.

1166: Since we are dealing with commuting functions (or

1167: symmetric tensors), only the highest total spin survives the tensor

1168: product. First we introduce the products, \beqa

1169: \label{symmetricbricks}

1170:  \sqrt{\frac{n!}{(2m)! (n - 2m)!}} a_1^{\frac{n}{2} + m}  a_2^{\frac{n}{2}-m} \;\; \equiv \;\;

1171:    \vert \frac{n}{2}\;,\;\; m \rangle \qquad \left( n \in \mathbf{Z}, m \in \mathbf{Z} - \frac{n}{2} \right) \nonumber \\

1172: \sqrt{\frac{n!}{(2m)! (n - 2m)!}}  \;\; \bar{a}_2^{\frac{n}{2} + m}

1173: \bar{a}_1^{\frac{n}{2}-m} \;\; \equiv \;\;

1174:    \overline{\vert \frac{n}{2}\;,\;\; m \rangle} \qquad \left( n \in \mathbf{Z}, m \in \mathbf{Z} - \frac{n}{2} \right)

1175: \eeqa which are states of definite $SU(2)$ spin $n$ and $R$ charge

1176: $\pm n/2$,

1177:  since the product of $n$ commuting $a$'s and $\bar{a}$'s automatically has only

1178: spin $n/2$ states.

1179: %The combinatorial factors are for normalization.

1180: We combine these to form a state of arbitrary $SU(2)$ spin and $R$

1181: charge using Clebsch-Gordon coefficients, \footnote{We are only

1182: using the `top-spin' Clebsch Gordon coefficients. The notation here

1183: is:

1184: $$

1185: {}_RC^{l_1, m_1}_{  k; \tilde{k} } = \langle l_1,m_1 \vert

1186: \frac{l_1}{2}+ \frac{R}{4} , k \; ; \; \frac{l_1}{2} -\frac{R}{4},

1187: \tilde{k} \rangle \times (-1)^{\frac{l_1}{2} -\frac{R}{4}-

1188: \tilde{k}}

1189: $$ We need this extra $-1$ factor because we tensoring conjugate representations of $SU(2)$ : $J_- a_1 \sim a_2$ but $J_-\bar{a_2} \sim - \bar{a}_1$ }

1190:  by,

1191: \begin{eqnarray}

1192: \label{wvfdefn}

1193:  \vert l_1 , m_1;R/2 \rangle_a & =& \sum_{k,\tilde{k} \atop k + \tilde{k} =

1194: m_1} {}_RC^{l_1, m_1}_{  k \; ;\; \tilde{k} } \vert \frac{l_1}{2}+

1195: \frac{R}{4} , k \rangle \overline{\vert \frac{l_1}{2} -\frac{R}{4},

1196: \tilde{k} \rangle} \nonumber \\

1197: &= & {\displaystyle \left(a_1 a_2\right)^{\frac{l_1}{2}

1198: +\frac{R}{4}} \left(\bar{a}_1\bar{a}_2\right)^{\frac{l_1}{2}

1199: -\frac{R}{4}}} \sum_{k + \tilde{k} = m_1} {}_RC^{l_1, m_1}_{ k \;

1200: ;\; \tilde{k} } \;\; a_1^k a_2^{-k} \bar{a}_2^{\tilde{k}}

1201: \bar{a}_1^{-\tilde{k}}

1202: \end{eqnarray}

1203: where we have introduced $\vert l_1 , m_1;R/2 \rangle_a $ to denote

1204: the wavefunctions with $SU(2)$ spin $(l_1,m_1)$ and $U(1)_R$ charge

1205: $R/2$ constructed from $a_i$ variables.

1206: 

1207: Using the same notation for $b_i$, $\vert l_2 , m_2;R/2\rangle_b$ is

1208: the state with the required symmetry charges. To construct an

1209: eigenfunction $Y_I$ on $T^{1,1}$, we must have equal $R$ charge for

1210: the $a$ and $b$ states above in order to have invariance under the

1211: transformation $a \to e^{i\alpha} a, b \to e^{-i \alpha} b$

1212: explained earlier (see (\ref{removeredun})). Hence, $Y_I$ is simply

1213: a product of the $a$ and $b$ states constructed above, \beqa

1214: \label{compwvf}

1215:  &Y_I& \sim \vert l_1 , m_1;R/2 \rangle_a  \vert l_2 , m_2;R/2 \rangle_b

1216: \ .\eeqa

1217: For example, some of the wavefunctions for $l_1=l_2=1, R=0 $ are :

1218: \begin{eqnarray}

1219:  \;  & a_1\bar{a}_2 \; b_1\bar{b}_2  & (m_1,m_2) = (1,1) \nonumber \\

1220:  \;  &(a_1\bar{a}_1-a_2\bar{a}_2) \; b_1\bar{b}_2 & (m_1,m_2) = (0,1) \nonumber \\

1221:  \; &a_2\bar{a}_1 \; b_2\bar{b}_1 & (m_1,m_2) = (-1,-1) \nonumber

1222: \end{eqnarray}

1223: 

1224: While the harmonics (\ref{compwvf}) are obviously relevant to the

1225: singular conifold, it was also shown in Section \ref{sec-resolved}

1226: that the Laplacian on the resolved conifold (see (\ref{reseq2}))

1227: factors in a form that allows one to use the same angular functions.

1228: This is because the resolution of the conifold preserves the

1229: $SU(2)_L\times SU(2)_R\times U(1)_R$ symmetry.

1230: 

1231: \subsection*{Construction of the dual operators }

1232: 

1233: The above construction of eigenfunctions is useful primarily because

1234: of their one-to-one correspondence with (single trace) operators in

1235: the guage theory. Our stragey is to replace $a_i,b_j$ in the

1236: eigenfunctions by the chiral superfields $A_i,B_j$. However, since

1237: $A_i,B_j$ are non-commuting operators in the gauge theory, we need

1238: to modify the procedure of the previous section to obtain an

1239: operator ${\cal O}_I$ of a given symmetry.

1240: 

1241: We may start with (\ref{symmetricbricks}), and symmetrize the

1242: product of $A_1, A_2$'s (and $\bar{A}_1, \bar{A}_2$'s) by hand (the

1243: gauge index structure seems ill defined but this will be fixed when

1244: the total operator is put together.) So we could now write instead

1245: of (\ref{symmetricbricks}) (with a different normalization factor),

1246: \beqa

1247:    \frac{1}{\sqrt{\frac{n!}{(2m)! (n - 2m)!}}}  \sum_{\frac{n}{2} + m = \sum i \atop \frac{n}{2}-m = \sum j} A_1^{i_1}  A_2^{j_1} A_1^{i_2} \cdots A_2^{j_k} \;\; \equiv \;\;

1248:    \vert \frac{n}{2}\;,\;\; m \rangle \qquad \left( n \in \mathbf{Z}, m \in \mathbf{Z} - \frac{n}{2} \right)

1249: \eeqa

1250:  The same symmetrization applies to $\bar{A}$'s as well.

1251: With this modified definition of $\vert \frac{n}{2},m \rangle$, we

1252: can write down the equation analogous to (\ref{wvfdefn}) with no

1253: change in the form,

1254:  \beqa

1255:  \vert l_1 , m_1;R/2 \rangle_A & =& \sum_{k,\tilde{k} \atop k + \tilde{k} = m_1} {}_RC^{l_1, m_1}_{  k \; ;\; \tilde{k} } \vert \frac{l_1}{2}+ \frac{R}{4} , k \rangle \overline{\vert \frac{l_1}{2} -\frac{R}{4} , \tilde{k} \rangle}

1256:  \eeqa

1257: 

1258:  We make the analogous definitions for $B$.

1259: Finally, we can write down dual operator ${\cal O}_I$ as,

1260:  \beq \label{operatorL}

1261: \displaystyle {\cal O}_I =  {\rm Tr} \left(\vert l_1 , m_1;R/2

1262: \rangle_A  \vert l_2 , m_2;R/2 \rangle_B\right) \eeq

1263: 

1264:  The product of the operators $\vert l_1 , m_1;R/2 \rangle_A$ and $\vert l_2 ,

1265: m_2;R/2 \rangle_B$ is taken in the following way. All the terms are

1266: multiplied out and in each term, one is free to move operators in

1267: the $(N,\bar{N})$ rep of the gauge group (i.e $A,\bar{B}$) past

1268: $(\bar{N},N)$ (i.e $B,\bar{A}$) but no rearrangement among

1269: themselves is allowed. We shuffle them past each other until they

1270: alternate and so we can contract gauge indices properly and take the

1271: trace. It is easy to verify that the numbers of fields of each type

1272: are equal and so there is always one essentially unique way of doing

1273: this. By construction, this operator has the specified symmetry $I$

1274: under the global symmetry group.

1275: 

1276: \section{$AdS_5 \times S^5$ Throats in the IR}

1277: \label{sec-throat}

1278: Here we show explicitly that the Green's function on the resolved conifold

1279: reduces to the form $\frac{1}{y^4}$ near the source as it must

1280: ($y$ here is the physical distance from the source

1281: on the transverse space).

1282: This leads to the usual near-horizon limit when the branes are at a smooth point and

1283: hence an $AdS_5 \times S^5$ throat. This is of course to be expected since close to the source,

1284: we can find coordinates in which the space looks flat at leading order and

1285: hence the Green's function must behave as $\frac{1}{y^4}$. But it is instructive to see

1286: how the series does add up to such a divergence while each individual term has a

1287: different kind of divergence that gives a singular geometry 

1288: in the case of the resolved conifold.

1289: 

1290: We focus on the resolved conifold and consider $\theta_0 = 0 = \phi_0$, i.e set the stack on

1291: the 'north pole' of the finite $S^2$. Also, we approach the singularity by first setting

1292: $\theta_2 = 0$ and taking the $r \rightarrow 0$ limit. 

1293: Now $r$ is the physical distance and from (\ref{resolvedgreenN}), we would like to show that,

1294: \beq \label{RTP}

1295: \sum_{l} (2 l + 1) H_I^A(r) \sim \frac{1}{r^4}  \quad \mbox{ while } 

1296: \quad  H_I^A(r) \sim \frac{1}{r^2} \quad \mbox{ as } r \rightarrow 0

1297: \eeq

1298: 

1299: Consider the regulation of the sum of squares of integers,

1300: \beq

1301: \sum_{n=0}^{\infty} n^2 \to   \sum_{n=0}^{\infty} n^2 R(n \epsilon)

1302: \eeq

1303: where $R(x)$ is a regulator such as $R(x) = e^{-x}$ with the property $R(x) \rightarrow 0$ (fast enough in a sense to be seen below) as $x \rightarrow \infty$. As $\epsilon \to 0$, the sum diverges and this allows one to approximate the sum by an integral in this limit. Further, only $0 \leq n \leq 1/\epsilon$ will contribute. Hence we find,

1304: \beq \label{dimensional}

1305:  \int_0^{\frac{1}{\epsilon}} n^2 R(n \epsilon) dn \;\;\; \sim \;\;\; \frac{1}{\epsilon^3} \int_0^{1} y^2 R(y) dy \quad (\epsilon \to 0)

1306: \eeq

1307: 

1308: Note that the above argument just amounts to dimensional analysis.

1309: To cast the given expression (\ref{RTP}) in the above form with $

1310: H_I^A(r)$ playing the role of a regulator, we note that $H_I^A(r)$ can be approximated for

1311: $r \ll a$ by $(\sqrt{l(l+1)}/r) K_1(\sqrt{l(l+1)}r)$.

1312: \footnote{We mean this in the sense that $K_1(y)$ is the solution to the differential

1313: equation obtained by applying $r \ll a$ to (\ref{resradialeqn}) whose exact solution

1314: was obtained as $H_I^A(r)$. We are interested in how $r$

1315: scales with $l$ to keep $H_I^A(r)$ constant for very small $r$,

1316: since this determines the leading order singularity through essentially 

1317: dimensional analysis in

1318: (\ref{dimensional}).

1319: Approximating $H_I^A$ by $K_1$ is valid in this sense. } Hence we have for $r \ll a$,

1320: \beqa

1321: \sum_{l} (2 l + 1) H_I^A(r) & \sim &  \sum_{l} (2l+1)\frac{\sqrt{l(l+1)}}{r} 

1322: K_1(\sqrt{l(l+1)} r) \sim \frac{1}{r} \int_n (2n+1)n K_1(n r) \\

1323: & \sim & \frac{1}{r} \int_0^{1/r} n^2 K_1(n r) dn \sim \frac{1}{r} \times \frac{1}{r^3} \;\;\; \int_0^{1} y^2 K_1(y) dy

1324: \eeqa

1325: where we have kept track of only the leading order singularity. We have identified 

1326: $R(y) = K_1(y)$ despite the fact $K_1(y) \sim 1/y$ for small y. 

1327: This is allowed here because $\int_0^1 dy y^2 K_1(y)$ converges.

1328: 

1329: Hence we see that indeed, $H(r,\theta_2=0) \sim \frac{1}{r^4}$ near $r=0$ and hence the geometry is non-singular (though each term in the expansion of $H$ behaves as $\frac{1}{r^2}$ giving a singular geometry by itself). The result essentially follows from dimensional analysis in (\ref{dimensional}).

1330: 

1331: \bibliographystyle{unsrt}

1332: \begin{thebibliography}{--}

1333: %\cite{Maldacena:1997re}

1334: \bibitem{Maldacena}

1335:   J.~M.~Maldacena,

1336:   ``The large N limit of superconformal field theories and supergravity,''

1337:   Adv.\ Theor.\ Math.\ Phys.\  {\bf 2}, 231 (1998)

1338:   [Int.\ J.\ Theor.\ Phys.\  {\bf 38}, 1113 (1999)]

1339:   [arXiv:hep-th/9711200].

1340:   %%CITATION = HEP-TH 9711200;%%

1341: 

1342: %\cite{Gubser:1998bc}

1343: \bibitem{GKP}

1344:   S.~S.~Gubser, I.~R.~Klebanov and A.~M.~Polyakov,

1345:   ``Gauge theory correlators from non-critical string theory,''

1346:   Phys.\ Lett.\ B {\bf 428}, 105 (1998)

1347:   [arXiv:hep-th/9802109].

1348:   %%CITATION = HEP-TH 9802109;%%

1349: 

1350: %\cite{Witten:1998qj}

1351: \bibitem{Witten}

1352:   E.~Witten,

1353:   ``Anti-de Sitter space and holography,''

1354:   Adv.\ Theor.\ Math.\ Phys.\  {\bf 2}, 253 (1998)

1355:   [arXiv:hep-th/9802150].

1356:   %%CITATION = HEP-TH 9802150;%%

1357: 

1358: %\cite{Aharony:1999ti}

1359: \bibitem{Aharony:1999ti}

1360:   O.~Aharony, S.~S.~Gubser, J.~M.~Maldacena, H.~Ooguri and Y.~Oz,

1361:   ``Large N field theories, string theory and gravity,''

1362:   Phys.\ Rept.\  {\bf 323}, 183 (2000)

1363:   [arXiv:hep-th/9905111].

1364:   %%CITATION = HEP-TH 9905111;%%

1365: 

1366: 

1367: %\cite{Klebanov:2000me}

1368: \bibitem{Klebanov:2000me}

1369:   I.~R.~Klebanov,

1370:   ``TASI lectures: Introduction to the AdS/CFT correspondence,''

1371:   arXiv:hep-th/0009139.

1372:   %%CITATION = HEP-TH 0009139;%%

1373: 

1374: 

1375: %\cite{Klebanov:1998hh}

1376: \bibitem{KW}

1377:   I.~R.~Klebanov and E.~Witten,

1378:   ``Superconformal field theory on threebranes at a Calabi-Yau  singularity,''

1379:   Nucl.\ Phys.\ B {\bf 536}, 199 (1998)

1380:   [arXiv:hep-th/9807080].

1381:   %%CITATION = HEP-TH 9807080;%%

1382: 

1383: %\cite{Morrison:1998cs}

1384: \bibitem{Morrison}

1385:   D.~R.~Morrison and M.~R.~Plesser,

1386:   ``Non-spherical horizons. I,''

1387:   Adv.\ Theor.\ Math.\ Phys.\  {\bf 3}, 1 (1999)

1388:   [arXiv:hep-th/9810201].

1389:   %%CITATION = HEP-TH 9810201;%%

1390: 

1391: \bibitem{KW2}

1392:   I.~R.~Klebanov and E.~Witten,

1393:   ``AdS/CFT correspondence and symmetry breaking,''

1394:   Nucl.\ Phys.\ B {\bf 556}, 89 (1999)

1395:   [arXiv:hep-th/9905104].

1396:   %%CITATION = HEP-TH 9905104;%%

1397: 

1398: %\cite{Klebanov:2000hb}

1399: \bibitem{KS}

1400:   I.~R.~Klebanov and M.~J.~Strassler,

1401:   ``Supergravity and a confining gauge theory: Duality cascades and

1402:   chiSB-resolution of naked singularities,''

1403:   JHEP {\bf 0008}, 052 (2000)

1404:   [arXiv:hep-th/0007191].

1405:   %%CITATION = HEP-TH 0007191;%%

1406: 

1407: %\cite{PandoZayas:2000sq}

1408: \bibitem{PandoZayas}

1409:   L.~A.~Pando Zayas and A.~A.~Tseytlin,

1410:   ``3-branes on resolved conifold,''

1411:   JHEP {\bf 0011}, 028 (2000)

1412:   [arXiv:hep-th/0010088].

1413:   %%CITATION = HEP-TH 0010088;%%

1414: 

1415: %\cite{Balasubramanian:1998sn}

1416: \bibitem{Balasubramanian}

1417:   V.~Balasubramanian, P.~Kraus and A.~E.~Lawrence,

1418:   ``Bulk vs. boundary dynamics in anti-de Sitter spacetime,''

1419:   Phys.\ Rev.\ D {\bf 59}, 046003 (1999)

1420:   [arXiv:hep-th/9805171].

1421:   %%CITATION = HEP-TH 9805171;%%

1422: 

1423: \bibitem{deHaro}

1424:   S.~de Haro, S.~N.~Solodukhin and K.~Skenderis,

1425:   ``Holographic reconstruction of spacetime and renormalization in the  AdS/CFT

1426:   correspondence,''

1427:   Commun.\ Math.\ Phys.\  {\bf 217}, 595 (2001)

1428:   [arXiv:hep-th/0002230].

1429:   %%CITATION = HEP-TH 0002230;%%

1430: 

1431: 

1432: 

1433: %\cite{Skenderis:2006uy}

1434: \bibitem{Skenderis1}

1435:   K.~Skenderis and M.~Taylor,

1436:   ``Kaluza-Klein holography,''

1437:   JHEP {\bf 0605}, 057 (2006)

1438:   [arXiv:hep-th/0603016].

1439:   %%CITATION = HEP-TH 0603016;%%

1440: 

1441: %\cite{Skenderis:2006di}

1442: \bibitem{Skenderis2}

1443:   K.~Skenderis and M.~Taylor,

1444:   ``Holographic Coulomb branch vevs,''

1445:   JHEP {\bf 0608}, 001 (2006)

1446:   [arXiv:hep-th/0604169].

1447:   %%CITATION = HEP-TH 0604169;%%

1448: 

1449: \bibitem{Giddings}

1450:   S.~B.~Giddings and A.~Maharana,

1451:   ``Dynamics of warped compactifications and the shape of the warped

1452:   landscape,''

1453:   Phys.\ Rev.\ D {\bf 73}, 126003 (2006)

1454:   [arXiv:hep-th/0507158].

1455:   %%CITATION = HEP-TH 0507158;%%

1456: 

1457: 

1458: %\cite{Baumann:2006th}

1459: \bibitem{BDKMMM}

1460:   D.~Baumann, A.~Dymarsky, I.~R.~Klebanov, J.~Maldacena, L.~McAllister and A.~Murugan,

1461:   ``On D3-brane potentials in compactifications with fluxes and wrapped

1462:   D-branes,''

1463:   JHEP {\bf 0611}, 031 (2006)

1464:   [arXiv:hep-th/0607050].

1465:   %%CITATION = HEP-TH 0607050;%%

1466: 

1467: 

1468: %\cite{Benvenuti:2005qb}

1469: \bibitem{Benvenuti}

1470:   S.~Benvenuti, M.~Mahato, L.~A.~Pando Zayas and Y.~Tachikawa,

1471:   ``The gauge / gravity theory of blown up four cycles,''

1472:   arXiv:hep-th/0512061.

1473:   %%CITATION = HEP-TH 0512061;%%

1474: 

1475: %\cite{Butti:2004pk}

1476: \bibitem{Butti}

1477:   A.~Butti, M.~Grana, R.~Minasian, M.~Petrini and A.~Zaffaroni,

1478:   ``The baryonic branch of Klebanov-Strassler solution: A supersymmetric

1479:   family of SU(3) structure backgrounds,''

1480:   JHEP {\bf 0503}, 069 (2005)

1481:   [arXiv:hep-th/0412187].

1482:   %%CITATION = HEP-TH 0412187;%%

1483: 

1484: %\cite{Dymarsky:2005xt}

1485: \bibitem{Dymarsky}

1486:   A.~Dymarsky, I.~R.~Klebanov and N.~Seiberg,

1487:   ``On the moduli space of the cascading SU(M+p) x SU(p) gauge theory,''

1488:   JHEP {\bf 0601}, 155 (2006)

1489:   [arXiv:hep-th/0511254].

1490:   %%CITATION = HEP-TH 0511254;%%

1491: 

1492: \bibitem{GHK}

1493:   S.~S.~Gubser, C.~P.~Herzog and I.~R.~Klebanov,

1494:   ``Symmetry breaking and axionic strings in the warped deformed conifold,''

1495:   JHEP {\bf 0409}, 036 (2004)

1496:   [arXiv:hep-th/0405282].

1497:   %%CITATION = HEP-TH 0405282;%%

1498: 

1499: 

1500: %\cite{Candelas:1989js}

1501: \bibitem{Candelas}

1502:   P.~Candelas and X.~C.~de la Ossa,

1503:   ``Comments on conifolds,''

1504:   Nucl.\ Phys.\ B {\bf 342}, 246 (1990).

1505:   %%CITATION = NUPHA,B342,246;%%

1506: 

1507: %\cite{Romans:1984an}

1508: \bibitem{Romans}

1509:   L.~J.~Romans,

1510:   ``New Compactifications Of Chiral N=2 D = 10 Supergravity,''

1511:   Phys.\ Lett.\ B {\bf 153}, 392 (1985).

1512:   %%CITATION = PHLTA,B153,392;%%

1513: 

1514: 

1515: %\cite{Gubser:1998vd}

1516: \bibitem{Gubser}

1517:   S.~S.~Gubser,

1518:   ``Einstein manifolds and conformal field theories,''

1519:   Phys.\ Rev.\ D {\bf 59}, 025006 (1999)

1520:   [arXiv:hep-th/9807164].

1521:   %%CITATION = HEP-TH 9807164;%%

1522: 

1523: 

1524: %\cite{Ceresole:1999zs}

1525: \bibitem{Ceresole}

1526:   A.~Ceresole, G.~Dall'Agata, R.~D'Auria and S.~Ferrara,

1527:   ``Spectrum of type IIB supergravity on AdS(5) x T(11): Predictions on N  = 1 SCFT's,''

1528:   Phys.\ Rev.\ D {\bf 61}, 066001 (2000)

1529:   [arXiv:hep-th/9905226].

1530:   %%CITATION = HEP-TH 9905226;%%

1531: 

1532: 

1533: %\cite{Itzhaki:2002kh}

1534: \bibitem{Itzhaki}

1535:   N.~Itzhaki, I.~R.~Klebanov and S.~Mukhi,

1536:   ``PP wave limit and enhanced supersymmetry in gauge theories,''

1537:   JHEP {\bf 0203}, 048 (2002)

1538:   [arXiv:hep-th/0202153].

1539:   %%CITATION = HEP-TH 0202153;%%

1540: 

1541: 

1542: \bibitem{Wittenunpub}

1543: E. Witten, unpublished, 2004.

1544: 

1545: \bibitem{BDK}

1546:   M.~K.~Benna, A.~Dymarsky and I.~R.~Klebanov,

1547:   ``Baryonic condensates on the conifold,''

1548:   arXiv:hep-th/0612136.

1549:   %%CITATION = HEP-TH 0612136;%%

1550: 

1551: \bibitem{GK}

1552:   S.~S.~Gubser and I.~R.~Klebanov,

1553:   ``Baryons and domain walls in an N = 1 superconformal gauge theory,''

1554:   Phys.\ Rev.\ D {\bf 58}, 125025 (1998)

1555:   [arXiv:hep-th/9808075].

1556:   %%CITATION = HEP-TH 9808075;%%

1557: 

1558: 

1559: \bibitem{Karch}

1560:   A.~Karch, A.~O'Bannon and K.~Skenderis,

1561:   ``Holographic renormalization of probe D-branes in AdS/CFT,''

1562:   JHEP {\bf 0604}, 015 (2006)

1563:   [arXiv:hep-th/0512125].

1564:   %%CITATION = HEP-TH 0512125;%%

1565: 

1566: 

1567: 

1568: %\cite{Herzog:2002ih}

1569: \bibitem{HKO1}

1570:   C.~P.~Herzog, I.~R.~Klebanov and P.~Ouyang,

1571:   ``D-branes on the conifold and N = 1 gauge / gravity dualities,''

1572:   arXiv:hep-th/0205100;

1573:   %%CITATION = HEP-TH 0205100;%%

1574:   ``Remarks on the warped deformed conifold,''

1575:   arXiv:hep-th/0108101.

1576:   %%CITATION = HEP-TH 0108101;%%

1577: 

1578: \bibitem{KT}

1579: I.~R.~Klebanov and A.~A.~Tseytlin,

1580:   ``Gravity duals of supersymmetric SU(N) x SU(N+M) gauge theories,''

1581:   Nucl.\ Phys.\ B {\bf 578}, 123 (2000)

1582:   [arXiv:hep-th/0002159].

1583:   %%CITATION = HEP-TH 0002159;%%

1584: 

1585: 

1586: 

1587: %\cite{Gauntlett:2004zh}

1588: \bibitem{Gaunt1}

1589:   J.~P.~Gauntlett, D.~Martelli, J.~Sparks and D.~Waldram,

1590:   ``Supersymmetric AdS(5) solutions of M-theory,''

1591:   Class.\ Quant.\ Grav.\  {\bf 21}, 4335 (2004)

1592:   [arXiv:hep-th/0402153];

1593:   %%CITATION = HEP-TH 0402153;%%

1594:   %J.~P.~Gauntlett, D.~Martelli, J.~Sparks and D.~Waldram,

1595:   ``Sasaki-Einstein metrics on S(2) x S(3),''

1596:   Adv.\ Theor.\ Math.\ Phys.\  {\bf 8}, 711 (2004)

1597:   [arXiv:hep-th/0403002].

1598:   %%CITATION = HEP-TH 0403002;%%

1599: 

1600: %\cite{Benvenuti:2004dy}

1601: \bibitem{Hanany}

1602:   S.~Benvenuti, S.~Franco, A.~Hanany, D.~Martelli and J.~Sparks,

1603:   ``An infinite family of superconformal quiver gauge theories with

1604:   Sasaki-Einstein duals,''

1605:   JHEP {\bf 0506}, 064 (2005)

1606:   [arXiv:hep-th/0411264].

1607:   %%CITATION = HEP-TH 0411264;%%

1608: 

1609: %\cite{Chen:2006xh}

1610: \bibitem{ChenLuPope}

1611:   W.~Chen, H.~Lu and C.~N.~Pope,

1612:   ``General Kerr-NUT-AdS metrics in all dimensions,''

1613:   Class.\ Quant.\ Grav.\  {\bf 23}, 5323 (2006)

1614:   [arXiv:hep-th/0604125].

1615:   %%CITATION = HEP-TH 0604125;%%

1616: 

1617: %\cite{Oota:2006pm}

1618: \bibitem{Oota}

1619:   T.~Oota and Y.~Yasui,

1620:   ``Explicit toric metric on resolved Calabi-Yau cone,''

1621:   Phys.\ Lett.\ B {\bf 639}, 54 (2006)

1622:   [arXiv:hep-th/0605129].

1623:   %%CITATION = HEP-TH 0605129;%%

1624: 

1625: \bibitem{HEK}

1626:   C.~P.~Herzog, Q.~J.~Ejaz and I.~R.~Klebanov,

1627:   ``Cascading RG flows from new Sasaki-Einstein manifolds,''

1628:   JHEP {\bf 0502}, 009 (2005)

1629:   [arXiv:hep-th/0412193].

1630:   %%CITATION = HEP-TH 0412193;%%

1631: 

1632: 

1633: %\cite{PandoZayas:2001iw}

1634: \bibitem{PTseyt}

1635:   L.~A.~Pando Zayas and A.~A.~Tseytlin,

1636:   ``3-branes on spaces with R x S(2) x S(3) topology,''

1637:   Phys.\ Rev.\ D {\bf 63}, 086006 (2001)

1638:   [arXiv:hep-th/0101043].

1639:   %%CITATION = HEP-TH 0101043;%%

1640: 

1641: %\cite{Pal}

1642: \bibitem{Pal}

1643:   S.~S.~Pal,

1644:   ``A new Ricci flat geometry,''

1645:   Phys.\ Lett.\ B {\bf 614}, 201 (2005)

1646:   [arXiv:hep-th/0501012].

1647:   %%CITATION = HEP-TH 0501012;%%

1648: 

1649: %\cite{Sfetsos:2005kd}

1650: \bibitem{Sfetsos}

1651:   K.~Sfetsos and D.~Zoakos,

1652:   ``Supersymmetric solutions based on Y(p,q) and L(p,q,r),''

1653:   Phys.\ Lett.\ B {\bf 625}, 135 (2005)

1654:   [arXiv:hep-th/0507169].

1655:   %%CITATION = HEP-TH 0507169;%%

1656: 

1657: \bibitem{Benini}

1658:   F.~Benini, F.~Canoura, S.~Cremonesi, C.~Nunez and A.~V.~Ramallo,

1659:   ``Unquenched flavors in the Klebanov-Witten model,''

1660:   arXiv:hep-th/0612118.

1661:   %%CITATION = HEP-TH 0612118;%%

1662: 

1663: 

1664: \end{thebibliography}

1665: \end{document}

1666: 

1667: