1: \documentclass[prl,showpacs,twocolumn]{revtex4}

2: %\documentclass[aps,preprint,showpacs,showkeys]{revtex4}

3: \usepackage{graphicx}

4: %\usepackage{epstopdf}

5: \usepackage{amssymb}

6: \usepackage{bm}

7: 

8: %%%%%%%%%%%%%%%%%%%%%%

9: %

10: % AUTHORS' MACROS BEGIN HERE 

11: %

12: 

13: 

14: %%%%% Simplify some frequently used LaTeX commands %%%%%

15: 

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

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

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

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

20: \newcommand{\nn}{\nonumber}

21: \newcommand{\bn}{\begin{enumerate}}

22: \newcommand{\en}{\end{enumerate}}

23: 

24: \newcommand{\eq}[1]{(\ref{#1})}

25: \newcommand{\newcaption}[1]{\centerline{\parbox{6in}{\caption{#1}}}}

26: 

27: 

28: %%%%%%%%%%%%% Double line letters using amssymb %%%%%%%%%%%

29: 

30: \def\identity{{\rlap{1} \hskip 1.6pt \hbox{1}}}

31: \def\iden{\identity}

32: 

33: \def\IB{\mathbb{B}}

34: \def\IC{\mathbb{C}}

35: \def\ID{\mathbb{D}}

36: \def\IH{\mathbb{H}}

37: \def\IN{\mathbb{N}}

38: \def\IP{\mathbb{P}}

39: \def\IR{\mathbb{R}}

40: \def\IZ{\mathbb{Z}}

41: 

42: %%%%%%%%%%%%%%%% Caligraphic letters %%%%%%%%%%%%%%%%%%

43: 

44: \def\CA{{\cal A}}

45: \def\CB{{\cal B}}

46: \def\CC{{\cal C}}

47: \def\CD{{\cal D}}

48: \def\CE{{\cal E}}

49: \def\CF{{\cal F}}

50: \def\CG{{\cal G}}

51: \def\CH{{\cal H}}

52: \def\CI{{\cal I}}

53: \def\CJ{{\cal J}}

54: \def\CK{{\cal K}}

55: \def\CL{{\cal L}}

56: \def\CM{{\cal M}}

57: \def\CN{{\cal N}}

58: \def\CO{{\cal O}}

59: \def\CP{{\cal P}}

60: \def\CQ{{\cal Q}}

61: \def\CR{{\cal R}}

62: \def\CS{{\cal S}}

63: \def\CT{{\cal T}}

64: \def\CU{{\cal U}}

65: \def\CV{{\cal V}}

66: \def\CW{{\cal W}}

67: \def\CX{{\cal X}}

68: \def\CY{{\cal Y}}

69: \def\CZ{{\cal Z}}

70: 

71: %%%%%%%%%%%%%%%%%% Greek letters %%%%%%%%%%%%%%%%%%%%%%

72: 

73: \def\a{\alpha}

74: \def\b{\beta}

75: \def\g{\gamma}

76: \def\d{\delta}

77: \def\e{\epsilon}

78: \def\ve{\varepsilon}

79: \def\z{\zeta}

80: % eta

81: \def\th{\theta}

82: \def\vth{\vartheta}

83: \def\i{\iota}

84: \def\k{\kappa}

85: \def\l{\lambda}

86: \def\m{\mu}

87: \def\n{\nu}

88: % xi

89: % o

90: % pi

91: \def\vp{\varpi}

92: \def\r{\rho}

93: \def\vr{\varrho}

94: \def\s{\sigma}

95: \def\vs{\varsigma}

96: \def\t{\tau}

97: \def\u{\upsilon}

98: % phi

99: \def\vph{\varphi}

100: % chi

101: % psi

102: \def\w{\omega}

103: %

104: \def\G{\Gamma}

105: \def\D{\Delta}

106: \def\Th{\Theta}

107: \def\L{\Lambda}

108: % Xi

109: % Pi

110: \def\S{\Sigma}

111: \def\Y{\Upsilon}

112: % Phi

113: % Psi

114: \def\O{\Omega}

115: 

116: 

117: %%%%%%%%%%%%%%%%% Mathematical Symbols %%%%%%%%%%%%%%%%%%%%

118: 

119: \def\half{\frac{1}{2}}

120: \def\thalf{{\textstyle \frac{1}{2}}}

121: \def\imp{\Longrightarrow}

122: \def\iff{\Longleftrightarrow}

123: \def\goto{\rightarrow}

124: \def\para{\parallel}

125: \def\vev#1{\langle #1 \rangle}

126: \def\del{\nabla}

127: \def\grad{\nabla}

128: \def\curl{\nabla\times}

129: \def\div{\nabla\cdot}

130: \def\p{\partial}

131: \newcommand{\bra}[1]{\langle{#1}|}

132: \newcommand{\ket}[1]{|{#1}\rangle}

133: 

134: %%%%%%%%%%%%%%%%%%%% Normal font in math %%%%%%%%%%%%%%%%%

135: 

136: \def\Tr{{\rm Tr}}

137: \def\tr{{\rm tr}}

138: \def\det{{\rm det}}

139: 

140: 

141: %%%%%%%%%%%%%%%%%%%%% For this paper only %%%%%%%%%%%%%

142: 

143: \def\bp{{\bar{\partial}}}

144: \def\bi{{\bar{i}}}

145: \def\bj{{\bar{j}}}

146: \def\ao{{\alpha_1}}

147: \def\bo{{\beta_1}}

148: \def\at{{\alpha_2}}

149: \def\bt{{\beta_2}}

150: \def\aod{{\dot{\alpha_1}}}

151: \def\bod{{\dot{\beta_1}}}

152: \def\atd{{\dot{\alpha_2}}}

153: \def\btd{{\dot{\beta_2}}}

154: \def\bz{\bar{z}}

155: \def\bw{\bar{w}}

156: \def\vol{\mbox{vol}}

157: \def\Vol{\mbox{Vol}}

158: \def\bra{\langle}

159: \def\ket{\rangle}

160: \def\bw{\bar{w}}

161: \def\tw{\tilde{w}}

162: 

163: 

164: %

165: % AUTHORS' MACROS END HERE

166: % 

167: %%%%%%%%%%%%%%%%%%%%%%

168: 

169: 

170: 

171: \begin{document}

172: 

173: \title{Superconformal field theories from crystal lattices}

174: \author{Sangmin \surname{Lee}}

175: \affiliation{Department of Physics and Astronomy, 

176: Seoul National University, Seoul 151-747, Korea}

177: %\date{\today}

178: 

179: \begin{abstract}

180: We propose a brane configuration for the (2+1)d, $\CN=2$ 

181: superconformal theories (CFT$_3$) arising from M2-branes 

182: probing toric Calabi-Yau 4-fold cones, 

183: using a T-duality transformation of M-theory. 

184: We obtain intersections of M5-branes on a three-torus 

185: which form a 3d bipartite crystal lattice in a way similar to 

186: the 2d dimer models for CFT$_4$. 

187: The fundamental fields of the CFT$_3$ are M2-brane discs localized around 

188: the intersections, and the super-potential terms are identified with the atoms 

189: of the crystal.  

190: The model correctly reproduces the complete BPS spectrum of 

191: mesons and baryons.

192: 

193: \end{abstract}

194: 

195: \pacs{11.25.-w, 04.50.+h, 04.20.Jb}

196: %\keywords{AdS/CFT, Toric-Quiver Dual}

197: 

198: \maketitle

199: 

200: 

201: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

202: 

203: Extending the AdS/CFT correspondence \cite{adscft} to theories 

204: with less than maximal supersymmetry has been an important subject 

205: from early years \cite{kw,ach,mp}. 

206: For four dimensional theories (CFT$_4$), remarkable progress has been 

207: made recently. The discovery of new AdS solutions \cite{gmsw1, clpp} 

208: was accompanied by identification of the corresponding quiver gauge theories 

209: \cite{ms1,bfhms,bk,bfz}. 

210: %

211: It was soon realized that theories arising from D3-branes probing a toric 

212: Calabi-Yau 3-fold (CY$_3$) singularities share common simplifying features. 

213: All essential properties of the CFT$_4$ are encoded in a certain periodic 

214: bipartite graph in two dimensions \cite{hken,fhkvw,fhm,hv,fhkv,cep} 

215: often called a dimer model.

216: The algorithm to go from a given toric CY to a dimer graph and vice versa 

217: has been firmly established \cite{hv,fhkv}. 

218: 

219: Much less is known about three-dimensional theories 

220: arising from M2-branes near CY$_4$ singularities. 

221: There is no qualitative difference on the geometry side; 

222: toric descriptions are known \cite{ms1} for the new supergravity solutions 

223: \cite{gmsw2} and many results in toric geometry 

224: are insensitive to the dimension of the CY \cite{msy1}. 

225: In contrast, conventional field theoretic tools are of little use 

226: because the CFT$_3$ is necessarily strongly coupled.

227: Even the simplest case of M2-branes in flat space remains elusive. 

228: 

229: In this paper, we build upon the intuition from the 2d dimer model for CFT$_4$ 

230: to show that each toric CFT$_3$ can be related to a 3d crystal lattice.  

231: The construction uses T-duality of M-theory on a three-torus 

232: just as a T-duality of IIB string theory is used to 

233: derive the dimer model \cite{fhkvw}.

234: %

235: Our model maps the CFT$_3$ to the world-volume theory of M5-branes 

236: intersecting along the bonds of the crystal lattice. 

237: Although the M5-brane theory is as poorly 

238: understood as the CFT$_3$ itself, we will be able to use our model to 

239: make some quantitative statements.

240: 

241: First, we show that the fundamental excitations of the CFT$_3$ are 

242: M2-brane discs localized around M5-brane intersections. 

243: Second, we show how to evaluate various global charges 

244: of the baryons and mesons and explain how they are related. 

245: Finally, we identify the super-potential terms which yield the F-term conditions.

246: We use them to construct the spectrum of BPS mesons, 

247: which agrees perfectly with the corresponding result from the geometry.

248: 

249: We begin with a stack of $N$ M2-branes near the tip of a CY$_4$ cone $X$. 

250: In the near horizon limit, the $\IR^{1,2}$ world-volume directions 

251: of the M2-branes 

252: and the radial direction of $X$ merge to form an AdS$_4$ 

253: and the base (unit radius section) of $X$ becomes 

254: the internal 7-manifold $Y$. We say $X$ is the cone over $Y$, or $X=C(Y)$. 

255: The cone $X$ being K\"ahler and Ricci-flat 

256: is equivalent to the base $Y$ being Sasakian and Einstein, respectively. 

257: 

258: Our construction assumes that $X$ is toric, so we first recall 

259: some relevant aspects of toric geometry \cite{msy1}. 

260: The toric diagram is a convex polyhedron composed of a set of lattice points 

261: $\{v_I^i\} \in \IZ^4$ ($I=1, \cdots, d \ge 4$). 

262: The CY condition requires that the $v_I$ be on the same $\IZ^3$ subspace. 

263: It is customary to choose a basis to set $v_I^4=1$ for all $I$. 

264: 

265: The toric diagram defines a solid cone 

266: $\D \equiv \{ y_i \in \IR^4 ; (v_I \cdot y) \ge 0 \mbox{ for all }I\}$. 

267: We call the boundary components 

268: $S^I \equiv \D \cap \{v_I\cdot y = 0\}$ the 3-fans. 

269: Two 3-fans meet at a 2-fan and several 2-fans join at a 1-fan. 

270: These fans are graph dual to the original toric diagram in 

271: the sense that each vertex $v_I$ is associated to a 3-fan, 

272: each edge connecting two neighboring vertices, $w_{IJ} = v_I -v_J$, 

273: corresponds to a 2-fan, etc. 

274: The cone $X$ is a $T^4$ fibration over $\D$. 

275: The fiber is aligned with the base in such a way that it shrinks to $T^n$ on 

276: the $n$-fans.

277: 

278: \begin{figure}[htb]

279: \includegraphics[width=4.0cm]{ptn1.eps}

280: \caption{A toric diagram (solid line) is

281: a convex polyhedron with integer-valued vertices in $\IR^3$. 

282: Its graph dual (dashed line) gives the fan diagram.} \label{partition}

283: \end{figure}

284: 

285: 

286: The moduli space of K\"ahler metrics on $X$ is parameterized by the Reeb vector 

287: $b^i \in \IR^4$, which defines the base of the cone by 

288: $Y = X \cap\{b\cdot y = 1/2\}$. 

289: In the basis mentioned above, the CY condition fixes $b^4=4$. 

290: The volume of $Y$ as an explicit function of $v_I$ and $b$ is known. 

291: The Ricci-flat metric is obtained by minimizing the volume with respect to 

292: $b^i$ with the range of $b^i/4$ being precisely the toric diagram. 

293: 

294: For later purposes, we recall that a baryon in the CFT$_3$ is mapped 

295: via AdS/CFT to an M5-brane wrapping a supersymmetric 5-cycle of $Y$ 

296: \cite{bhk}.

297: Each 3-fan $S^I$ dual to a vertex $v_I$ defines a 5-cycle 

298: which is the $T^3$ fibration over $S^I \cap \{b\cdot y = 1/2\}$. 

299: We abuse the notation a bit and use $S^I$ to denote either 

300: the 3-fan or the 5-cycle depending on the context. 

301: The baryons are charged under the global symmetries of the theory. 

302: The four $U(1)$ isometries $F_i$ of $X$ are called flavor symmetries 

303: and there are also $d-4$ baryonic symmetries $Q_a$, where $d$ is the 

304: number of vertices of the toric diagram. 

305: The charges of the the baryons 

306: $F_i[S^I]  \equiv N F_i^I$, $Q_a[S^I] \equiv N Q_a^I$

307: satisfy the toric relations $v_I^i F^I_j = \d^i_j$ and $v_I^i Q_a^I = 0$ \cite{LR}.

308: The $U(1)_R$ charge is the linear combination of the flavor charges 

309: $R = \half b^i F_i$, where $b^i$ is the Reeb vector. 

310: The toric relations and the CY condition ($b^4=4$) implies that $\sum_I R^I = 2$.

311: 

312: Our crystal model follows from a T-duality of M-theory. 

313: It was inspired by a similar derivation of the 2d dimer model 

314: using T-duality of IIB string theory \cite{hken}. 

315: We take the T-duality transformation along $T^3\subset T^4$ aligned with the

316: $y_{1,2,3}$ coordinates.

317: By T-duality, we mean the element $t$ in the $SL(2,\IZ)\times SL(3,\IZ)$ duality 

318: group which acts as $t: \tau \equiv C_{(3)} + i \sqrt{g_{T^3}} \goto -1/\tau$.

319: The stack of $N$ M2-branes turns into a stack of $N$ M5-branes wrapping 

320: the dual $T^3$. 

321: We call them the $T$-branes. 

322: The degenerating circle fibers turn into another M5-brane 

323: extended along the (2+1)d world-volume and a non-trivial 3-manifold 

324: $\S$ in $\IR^3 \times T^3$. We call it the $\S$-brane. 

325: The result is summarized in Table I.

326: 

327: \begin{table}[htb]

328: \label{brane3}

329: $$

330: \begin{array}{l|ccc|cccccc|cc}

331: \hline

332:            & 0 & 1 & 2 & 3 & 4 & 5 & 6 & 7 & 8 & 9 & 11\\

333: \hline

334: \mbox{M5} & \circ & \circ & \circ & & & & \circ & \circ & \circ \\

335: \mbox{M5} & \circ & \circ & \circ &  \multicolumn{6}{c|}{\Sigma} & & \\

336: \hline

337: \end{array}

338: $$

339: \caption{The brane configuration for the CFT$_3$.  

340: The special Lagrangian 3-manifold $\Sigma$ is locally a product of 

341: a surface in $\IR^3$(345) and a curve in $T^3$(678).}

342: \end{table}

343: 

344: The duality chain shows that $\S$ is locally a product of a surface in $\IR^3$ and 

345: a curve in $T^3$. 

346: The geometry of $\S$ can be read off from the toric data. 

347: In particular, the projection of $\S$ onto $\IR^3$ is a thickened 

348: version of the collection of the 2-fans. This is similar 

349: to the so called amoeba projection used to analyze the 2d dimer model. 

350: Note that a 2-fan dual to an edge $w_{IJ}=(p,q,r)$ is locally a plane 

351: orthogonal to $w_{IJ}$ in $\IR^3$. 

352: Away from the intersections with other 2-fans, 

353: $\S$ is locally the product of this plane and the 

354: $(p,q,r)$ 1-cycle in the $T^3$.  

355: The correlation of the orientation ensures that $\S$ is locally special Lagrangian, 

356: i.e., it is calibrated by $\mbox{Im}\O$, 

357: where $\O$ is the holomorphic three-form 

358: $\O = (dx^3+idx^6)\wedge(dx^4+idx^7)\wedge(dx^5+idx^8)$, 

359: and the pull-back of the K\"ahler form onto $\S$ vanishes. 

360: This is in accordance with the $\CN=2$ supersymmetry on the (2+1)d world-volume.

361: 

362: In analogy with the D5/NS5-brane description of the dimer model, 

363: we expect that the content of the CFT$_3$ should be encoded 

364: in the intersection locus between 

365: the $T$-branes and the $\S$-brane projected onto the $T^3$. 

366: The result is a graph in the $T^3$ which we call the crystal lattice.  

367: As in the 2d dimer model, the graph consists of edges and 

368: vertices which we call {\em bonds} and {\em atoms} 

369: to distinguish them from similar objects in the toric diagram. 

370: 

371: To obtain a bond, we need a collection of 2-fans to form 

372: a closed region in $\IR^3$ containing one or more vertex $v_I$.  

373: When the $(p,q,r)$-cycles of the $T^3$ associated to the 2-fans 

374: join along a line segment, a bond is formed. 

375: Note that while the 2d dimer model allows only a pair of 1-cycles to merge 

376: and form an edge, in our crystal model three or more 1-cycles can merge 

377: along a bond. 

378: 

379: \begin{figure}[htb]

380: \includegraphics[width=4.0cm]{ptn2.eps}

381: \caption{A partition of the toric diagram.} \label{partition2}

382: \end{figure}

383: 

384: To see how several bonds meet at an atom, 

385: consider a partition of the toric diagram (Figure 2). 

386: Each closed region gives a bond as explained above. 

387: The bonds can meet at an atom only if 

388: the regions add up to cover the entire toric diagram \cite{LL}.

389: 

390: 

391: Figures 3 and 4 illustrates the rules with the two simplest examples, 

392: namely, $\IC^4$ and 

393: the cone over the homogeneous space $Q(1,1,1)$. 

394: The latter example has been studied 

395: in the AdS/CFT context in Refs. \cite{ahn1, oh, fab, ahn2, glmw}. 

396: Note that the crystal lattice inherits the discrete symmetry of the toric diagram. 

397: The tetrahedral symmetry of $\IC^4$ is reflected in its 

398: lattice which has the familiar zincblende (GaAs) structure. 

399: The $C(Q(1,1,1))$ cone and its lattice which has the NaCl structure 

400: share the octahedral symmetry. 

401: More examples and details of the general algorithm to determine the crystal structure will be reported in Ref. \cite{LL}.

402: 

403: \begin{figure}[htb]

404: \includegraphics[width=8.0cm]{C4.eps}

405: \caption{The crystal for $\IC^4$ has the zincblende structure.} \label{C4}

406: \end{figure}

407: 

408: 

409: \begin{figure}[htb]

410: \includegraphics[width=8.0cm]{Q111.eps}

411: \caption{The crystal for $C(Q(1,1,1))$ has the NaCl structure.} \label{Q111}

412: \end{figure}

413: 

414: 

415: %\subsection{Baryons and Fundamentals}

416: 

417: We now proceed to study the physical properties of the CFT$_3$ 

418: using our crystal model. First of all, we have to figure out 

419: what the the fundamental excitations are and where in the crystal 

420: they live.

421: Recall that we have a known description of the baryons on the geometry side. 

422: Our strategy is to bring the baryon to the crystal picture 

423: and try to deform it smoothly into a collection of $N$ fundamentals, 

424: thereby identifying the latter. 

425: A similar process in the 2d dimer model was discussed in Ref. \cite{ima}.

426: 

427: The T-duality transformation turns the baryon M5-branes 

428: in the original picture into M2-branes. 

429: Since the projection of the $\S$-brane onto $\IR^3$ realizes 

430: the toric fans, the baryon M2-brane corresponding to a 3-fan 

431: must have a boundary on the 2-fan walls surrounding it. 

432: Topologically, the M2-baryon is a disc. 

433: We can shrink the baryon until we approach 

434: the intersection between the T-branes and the $\S$ brane. 

435: The intersection is a bond in the crystal formed by the 

436: 1-cycles paired with the 2-fans. The boundary of the M2-baryon 

437: is now a small circle on the $\S$ brane localized around the intersection. 

438: We now show that the baryon can be smoothly deformed 

439: into a collection of $N$ M2-branes with its boundary on the T-branes 

440: localized around the intersection. 

441: 

442: Consider a supersymmetric configuration of two M5-branes 

443: intersecting along a (3+1)-dimensional space.

444: Keeping the mutually transverse directions only, 

445: the intersection can be described locally by a holomorphic equation 

446: $zw=0$ with $(z,w)\in \IC^2$. 

447: Suppose we have an M2-brane disc with boundary along a circle around the 

448: origin in the $z$-plane. Infinitesimal deformation of the intersection into 

449: $zw=\e$ shows that the circle on the $z$-plane can be 

450: smoothly deformed into another circle on the $w$-plane. 

451: The argument is easily generalized to multiple branes. 

452: An M5-brane on the $z$-plane intersecting with $N$ M5-branes 

453: on the $w$-plane is described by $z^N w = 0$. 

454: The same argument as above shows that an M2-disc on the $z$-plane 

455: now turns into a stack of $N$ M2-discs on the $w$-plane. 

456: Thus we find that the fundamental excitations of the CFT$_3$ are the 

457: M2-discs whose boundary encircles the bonds of the crystal lattice. 

458: This fundamental M2-disc is the analog of the bi-fundamental  

459: field localized around the D5/NS5 intersection in the 2d dimer model.

460: 

461: The fundamental M2-discs explain why it is difficult, if not impossible, 

462: to write down a field theory Lagrangian for the CFT$_3$ 

463: (See Refs. \cite{oh, ahn1, fab} for previous attempts). 

464: Our model suggests that in order to write write down the field theory, 

465: we first have to understand the quantum theory of M2-discs on the M5-brane world-volume.

466: %

467: The CFT$_3$ is not likely to be an ordinary any gauge theory.  

468: Recall that each face of the 2d dimer graph gives an $SU(N)$ gauge group 

469: in the quiver gauge theory, and the matter fields are 

470: bi-fundamentals connecting adjacent gauge groups. 

471: In contrast, all non-localized M2-discs in our crystal model are equivalent, 

472: and it is not clear whether they represent gauge fields.

473: 

474: 

475: The transition from a baryon to fundamentals also gives 

476: an easy way to find the global charges of the fundamentals, 

477: as the fundamentals simply inherit the charges of the baryons.  

478: If a fundamental M2-disc originates from a region of 

479: a partition shown in Figure 2, its global charges are $1/N$ times 

480: the sums of charges ($F^I_i$, $Q_a^I$) of the 3-fans $S^I$ inside the region. 

481: 

482: The M2 discs localized along the bonds meeting at an atom 

483: can form a spherical membrane. In analogy with 

484: the 2d dimer model, we identify it as a super-potential term of the CFT$_3$.  

485: On general grounds, we expect that the super-potential terms 

486: have $R$-charge two and vanishing flavor charges. 

487: It is indeed true because (i) every atom corresponds to a partition covering the entire toric diagram (ii) the sum of charges over all 3-fans 

488: is either two ($R$-charge) or zero (flavor charges) due to the toric relations.

489: When two atoms are connected by a bond, the M2-disc 

490: contribute to the two super-potential terms with opposite orientation. 

491: If we paint the atoms according to the orientation, then 

492: only atoms with opposite colors can be connected.

493: In other words, the crystal is bipartite. 

494: 

495: The super-potential terms yield the F-term equivalence relations among the 

496: elements of the chiral ring. 

497: As in the 2d dimer models, the relations implies that the 

498: sum of terms corresponding to any two connected vertices 

499: are F-term equivalent to zero. 

500: 

501: Chiral mesons in CFT$_3$ correspond to algebraic (holomorphic polynomial) functions on the CY$_4$ cone $X$. They are labeled by integer 

502: points $m=(m_1,m_2,m_3,m_4)$ in the solid cone $\D$, 

503: which are the momentum quantum numbers 

504: along the $T^4$ fiber of $X$.

505: It follows immediately that the flavor charges of a meson $m$ are $F_i(m) = m_i$. 

506: The $R$-charge of the meson is then given by \cite{msy2, gmsy, zaffnew}

507: \be

508: R(m) 

509: = \frac{1}{2} b^i F_i(m) = \frac{1}{2} (m \cdot b).

510: \ee

511: 

512: Under the T-duality transformation, the mesons are transformed into 

513: closed M2-branes. 

514: The first three components of $m$ define the homology 2-cycle of the 

515: M2-meson along the $T^3$. It is instructive to confirm this fact 

516: in the crystal model.  

517: When an M2-meson wraps a 2-cycle, its charge can be computed 

518: from its intersections with the bonds in the crystal, 

519: as the meson is a bound state of the fundamental M2-discs. 

520: Since the bonds are made of $(p,q,r)$-cycles corresponding to 

521: the edges $w_{IJ}$ of the toric diagram, we may equally well compute the  

522: charges from the intersection of the meson with the $(p,q,r)$-cycles.

523: Following Ref. \cite{buttinew}, we use the fact that $\sum_I F_i^I = 0$ 

524: ($i \ne 4$) to assign variables $F_i^{IJ}=-F_i^{JI}$ to the edges $w_{IJ}$, 

525: such that 

526: $F_i^I = \sum_{J} F_i^{IJ}$, where the sum runs over the neighboring vertices. 

527: Then we find that the charge of the M2-meson is indeed given by 

528: \be

529: F_i (m) &=& \sum_{(IJ)} (m\cdot w_{IJ}) F_i^{IJ} 

530: = \sum_{(IJ)} m \cdot(v_I-v_J) F_i^{IJ} 

531: \nn \\ 

532: &=& \sum_I (m \cdot v_I) F^I_i  = m_i . 

533: \ee

534: The last component of $m$, contributing $2m_4$ to the $R$-charge, 

535: measures how many times the M2-meson wraps a super-potential term. 

536: The F-term condition discussed above 

537: guarantees that all the M2-meson with the same value of $m$ 

538: are F-term equivalent, regardless of the precise way they wrap the 2-cycle. 

539: So there is a unique meson for each value of $m$. 

540: This perfectly matches the known results on the BPS spectrum of chiral mesons from the geometry side \cite{msy2, gmsy, zaffnew, counting}. 

541: 

542: Some remarks are in order. Both for the 2d dimer and 3d crystal models, 

543: the derivation based on T-duality connect objects on the geometry side 

544: such as baryons and mesons with their counterparts on the CFT in a smooth 

545: manner. It realizes in a concrete way the common lore that the AdS/CFT correspondence is a open/closed duality. We also note that the derivation of the 

546: $\b$-deformed geometry of the toric CY cones discussed in Ref. \cite{lunin} 

547: tacitly assumed a model for CFT$_3$ of the type we presented here. 

548: 

549: There are several issues that deserve further investigation.  

550: It is well-known that the 2d dimer model for a given CY$_3$ is not unique 

551: and different models are related by Seiberg duality. It would be interesting 

552: to see whether similar phenomena occurs in the crystal model. 

553: The analysis in this paper was restricted to the topological aspects of the model. 

554: It would be important to extend it to include dynamical issues 

555: such as non-BPS meson spectrum, moduli space of vacua, 

556: %

557: marginal deformations other than the $\b$-deformation \cite{lunin, ahn2, glmw},  and volume minimization \cite{msy1} and $\t_{RR}$-minimization \cite{trr}.

558: Finally, it would be interesting to look for a brane configuration 

559: for the $\CN=3$ CFT$_3$ \cite{billo, ly}. 

560: 

561: 

562: \begin{acknowledgments}

563: We thank Surjay Lee, Jaemo Park and Soo-Jong Rey for useful discussions 

564: and comments on the manuscript.

565: We are especially grateful to Amihay Hanany for a series of lectures 

566: at the APCTP/CQueST/KIAS Winter School on String 

567: Theory in February 2006.

568: %

569: This work was supported by the Faculty Grant of Seoul National University and 

570: the KOSEF Basic Research Program, grant No. R01-2006-000-10965-0.

571: \end{acknowledgments}

572: 

573: 

574: 

575: \begin{thebibliography}{99}

576: 

577: 

578: %%% AdS/CFT %%%

579: 

580: \bibitem{adscft}

581: J.~M.~Maldacena, 

582: %%``The large $N$ limit of superconformal field theories and supergravity," 

583: Adv. Theor. Math. Phys. {\bf 2}, 231(1998)  [Int. J. Theor. Phys. {\bf 38}, 1113 (1999)].

584: %[arXiv:hep-th/9711200].

585: 

586: %%% Old SE %%%

587: 

588: \bibitem{kw}

589: I.~R.~Klebanov and E.~Witten,

590: % %``Superconformal field theory on threebranes at a Calabi-Yau singularity,Ó 

591: Nucl.\ Phys.\ B {\bf 536}, 199 (1998). %[arXiv:hep-th/9807080].

592: 

593: %\bibitem{gub}

594: %S. S. Gubser, % %``Einstein manifolds and conformal field theories,"

595: %Phys. Rev. D {\bf 59}, 025006 (1999). %[arXiv:hep-th/9807164].

596: 

597: \bibitem{ach}

598:   B.~S.~Acharya, J.~M.~Figueroa-O'Farrill, C.~M.~Hull and B.~J.~Spence,

599:   %``Branes at conical singularities and holography,''

600:   Adv.\ Theor.\ Math.\ Phys.\  {\bf 2}, 1249 (1999)

601:  . %[arXiv:hep-th/9808014].

602: 

603: 

604: \bibitem{mp}

605: D.~R.~Morrison and M.~R.~Plesser,

606:   %``Non-spherical horizons. I,''

607:   Adv.\ Theor.\ Math.\ Phys.\  {\bf 3}, 1 (1999)

608:  . %[arXiv:hep-th/9810201].

609: 

610: %%% New SE %%%

611: 

612: \bibitem{gmsw1}

613: J. P. Gauntlett, D. Martelli, J. Sparks and D. Waldram,

614: %``Sasaki-Einstein metrics on $S^2 \times S^3$ ," 

615: Adv.\ Theor.\ Math.\ Phys.\ {\bf 8}, 711 (2004). %[arXiv:hep-th/0403002].

616: 

617: \bibitem{clpp}

618: M. Cvetic, H. Lu, D. N. Page and C. N. Pope, 

619: %``New Einstein-Sasaki spaces in five and higher dimensions," 

620: Phys.\ Rev.\ Lett.\ {\bf 95},

621: 071101 (2005). %[arXiv:hep-th/0504225].

622: 

623: %%% New Quiver %%% 

624: 

625: \bibitem{ms1}

626: D.~Martelli and J.~Sparks,

627:   %``Toric geometry, Sasaki-Einstein manifolds and a new infinite class of

628:   %AdS/CFT duals,''

629:   Commun.\ Math.\ Phys.\  {\bf 262}, 51 (2006). %[arXiv:hep-th/0411238].

630: 

631: \bibitem{bfhms}

632: S. Benvenuti, S. Franco, A. Hanany, D. Martelli and J. Sparks, 

633: %``An infinite family of superconformal quiver gauge theories with Sasaki-Einstein duals," 

634: JHEP {\bf 0506}, 064 (2005).

635: %[arXiv:hep-th/0411264].

636: 

637: \bibitem{bk}

638: S.~Benvenuti and M.~Kruczenski,

639: %``From Sasaki-Einstein spaces to quivers via BPS geodesics: $L^{p,q|r}$ ," 

640: JHEP {\bf 0604} 033 (2006). %[arXiv:hep-th/0505206].

641: 

642: \bibitem{bfz}

643: A. Butti, D. Forcella and A. Zaffaroni, 

644: %``The dual superconformal theory for $L^{p,q,r}$ manifolds," 

645: JHEP {\bf 0509}, 018 (2005). %[arXiv:hep-th/0505220].

646: 

647: %%% Tiling-Dimer %%% 

648: 

649: \bibitem{hken}

650: A.~Hanany and K.D.~Kennaway, 

651: %``Dimer models and toric diagrams,"

652: arXiv:hep-th/0503149.

653: 

654: \bibitem{fhkvw}

655: S. Franco, A. Hanany, K. D. Kennaway, D. Vegh and B. Wecht,

656: %``Brane dimers and quiver gauge theories," 

657: JHEP {\bf 0601} (2006) 096.

658: %[arXiv:hep-th/0504110].

659: 

660: \bibitem{fhm}

661: S. Franco, A. Hanany, D. Martelli, J. Sparks, D. Vegh and B. Wecht,

662:  %``Gauge theories from toric geometry and brane tilings,"

663: JHEP {\bf 0601}, 128 (2006).

664: %[arXiv:hep-th/0505211].

665: 

666: \bibitem{hv}

667: A. Hanany and D. Vegh, 

668: %``Quivers, tilings, branes and rhombi,"

669: arXiv:hep-th/0511063.

670: 

671: \bibitem{fhkv}

672: B. Feng, Y. H. He, K. D. Kennaway and C. Vafa, 

673: %``Dimer models from mirror symmetry and quivering amoebae,"

674: arXiv:hep-th/0511287.

675: 

676: \bibitem{cep}

677:  F. Canoura, J. D. Edelstein, L. A. Pando Zayas, A. V. Ramallo and D. Vaman,

678: %``Supersymmetric branes on AdS$5 \times Y_{p,q}$ and their field theory duals," 

679: JHEP {\bf 0603}, 101 (2006).

680: %[arXiv:hep-th/0512087].

681: 

682: 

683: \bibitem{gmsw2}

684: J.~P.~Gauntlett, D.~Martelli, J.~F.~Sparks and D.~Waldram,

685:   %``A new infinite class of Sasaki-Einstein manifolds,''

686:   Adv.\ Theor.\ Math.\ Phys.\  {\bf 8}, 987 (2006). 

687:   %[arXiv:hep-th/0403038].

688: 

689: \bibitem{msy1} 

690:   D.~Martelli, J.~Sparks and S.~T.~Yau,

691:   %``The geometric dual of a-maximisation for toric Sasaki-Einstein

692:   %manifolds,''

693:   Commun.\ Math.\ Phys.\  {\bf 268}, 39 (2006).

694:   %arXiv:hep-th/0503183.

695:   

696: 

697: \bibitem{bhk}

698: D. Berenstein, C. P. Herzog and I. R. Klebanov, 

699: %``Baryon spectra and AdS/CFT correspondence," 

700: JHEP {\bf 0206}, 047 (2002). 

701: %[arXiv:hep-th/0202150].

702: 

703: 

704: 

705: \bibitem{LR}

706:   S.~Lee and S.-J.~Rey,

707:   %%``Comments on anomalies and charges of toric-quiver duals,''

708:   JHEP {\bf 0603}, 068 (2006).

709:   %[arXiv:hep-th/0601223].

710: 

711: \bibitem{LL} 

712: S. Lee and S. Lee, to appear.

713: 

714: 

715: \bibitem{oh}

716:   K.~Oh and R.~Tatar,

717:   %``Three dimensional SCFT from M2 branes at conifold singularities,''

718:   JHEP {\bf 9902}, 025 (1999).

719:   %[arXiv:hep-th/9810244].

720:   %%CITATION = HEP-TH 9810244;%%

721: 

722: \bibitem{ahn1}

723:   C.~Ahn,

724:   %``N = 2 SCFT and M theory on AdS(4) x Q(1,1,1),''

725:   Phys.\ Lett.\ B {\bf 466}, 171 (1999).

726:   %[arXiv:hep-th/9908162].

727: 

728: \bibitem{fab}

729:   D.~Fabbri, P.~Fr\'e, L.~Gualtieri, C.~Reina, A.~Tomasiello, A.~Zaffaroni, A.~Zampa,

730:   % ``3D superconformal theories from Sasakian seven-manifolds: New  nontrivial

731:   %evidences for AdS(4)/CFT(3),''

732:   Nucl.\ Phys\ B {\bf 577}, 547 (2000).

733:   %[arXiv:hep-th/9907219].

734: 

735: \bibitem{ahn2}

736:   C.~Ahn and J.F.~Vazquez-Poritz,

737:   %``Marginal deformations with U(1)**3 global symmetry,''

738:   JHEP~{\bf 0507},~032~(2005).

739:   %[arXiv:hep-th/0505168].

740:   %%CITATION = HEP-TH 0505168;%%

741: 

742: \bibitem{glmw}

743:   J.~P.~Gauntlett, S.~Lee, T.~Mateos and D.~Waldram,

744:   %``Marginal deformations of field theories with AdS(4) duals,''

745:   JHEP {\bf 0508}, 030 (2005).

746:   %[arXiv:hep-th/0505207].

747:   

748: 

749: \bibitem{ima}

750:   Y.~Imamura,

751:   %``Global symmetries and 't Hooft anomalies in brane tilings,''

752:   arXiv:hep-th/0609163.

753:   %%CITATION = HEP-TH 0609163;%%

754: 

755: \bibitem{msy2}

756:   D.~Martelli, J.~Sparks, S.~T.~Yau,

757:   %``Sasaki-Einstein manifolds and volume minimisation,''

758:   arXiv:hep-th/0603021.

759: 

760: \bibitem{gmsy}

761:   J.~P.~Gauntlett, D.~Martelli, J.~Sparks and S.~T.~Yau,

762:   %``Obstructions to the existence of Sasaki-Einstein metrics,''

763:   arXiv:hep-th/0607080.

764: 

765: \bibitem{zaffnew}

766:   A.~Butti, D.~Forcella and A.~Zaffaroni,

767:   %``Deformations of conformal theories and non-toric quiver gauge theories,''

768:   arXiv:hep-th/0607147.

769:   %%CITATION = HEP-TH 0607147;%%

770: 

771: \bibitem{buttinew}

772:   A.~Butti,

773:   %``Deformations of toric singularities and fractional branes,''

774:   arXiv:hep-th/0603253.

775:   %%CITATION = HEP-TH 0603253;%%

776: 

777: \bibitem{counting}

778:   S.~Benvenuti, B.~Feng, A.~Hanany and Y.~H.~He,

779:   % ``Counting BPS operators in gauge theories: Quivers, syzygies and

780:   %plethystics,''

781:   arXiv:hep-th/0608050.

782:   %%CITATION = HEP-TH 0608050;%%

783: 

784: 

785:  \bibitem{lunin}

786:   O.~Lunin and J.~M.~Maldacena,

787:  %  ``Deforming field theories with U(1) x U(1) global symmetry and their

788:   %gravity duals,''

789:   JHEP {\bf 0505}, 033 (2005).

790:   %[arXiv:hep-th/0502086].

791: 

792: 

793: \bibitem{trr}

794: E.~Barnes, E.~Gorbatov, K.~Intriligator, M.~Sudano and J.~Wright,

795: %``The exact superconformal $R$-symmetry minimizes $\tau_{RR}$,''

796:   Nucl.\ Phys.\ B {\bf 730}, 210 (2005).

797:  %[arXiv:hep-th/0507137].

798: 

799: 

800: \bibitem{billo}

801:   M.~Billo, D.~Fabbri, P.~Fre, P.~Merlatti and A.~Zaffaroni,

802:   % ``Rings of short N = 3 superfields in three dimensions and M-theory on

803:   %AdS(4) x N(0,1,0),''

804:   Class.\ Quant.\ Grav.\  {\bf 18}, 1269 (2001).

805:   %[arXiv:hep-th/0005219].

806: 

807: \bibitem{ly}

808:   K.~M.~Lee and H.~U.~Yee,

809:   %``New AdS(4) x X(7) geometries with N = 6 in M theory,''

810:   arXiv:hep-th/0605214.

811:   %%CITATION = HEP-TH 0605214;%%

812: 

813: 

814: 

815: 

816: \end{thebibliography}

817: 

818: 

819: \end{document}

820: