1: %\documentclass[12pt,reqno]{article}

2: \documentclass[11pt]{article}

3: \usepackage{amsmath}

4: \usepackage{amsfonts}

5: %\usepackage[hypertex]{hyperref}

6: %\usepackage{showkeys}       

7: \usepackage{latexsym}

8: \usepackage[dvips]{graphicx}

9: \usepackage{epsf}

10: 

11: \textwidth 162mm

12: \textheight 230mm

13: \topmargin -7mm

14: \headheight 0mm

15: \oddsidemargin 0mm

16: \evensidemargin 0mm

17: 

18: %\parindent 0em

19: \allowdisplaybreaks

20: %\numberwithin{equation}{section}

21: 

22: %\renewcommand{\Large}{\normalsize}

23: \renewcommand{\baselinestretch}{1.2}

24: %*******************************************

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

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

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

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

29: \newcommand{\nn}{\nonumber} 

30: %*******************************************

31: %

32: \newcommand{\Zom}{\mathbb{Z}}

33: \newcommand{\Rom}{\mathbb{R}}

34: \newcommand{\Com}{\mathbb{C}}

35: 

36: \newcommand{\cC}{\mathcal{C}}

37: \newcommand{\cD}{\mathcal{D}}

38: \newcommand{\cF}{\mathcal{F}}

39: \newcommand{\cG}{\mathcal{G}}

40: \newcommand{\cL}{\mathcal{L}}

41: \newcommand{\cO}{\mathcal{O}}

42: \newcommand{\cR}{\mathcal{R}}

43: \newcommand{\cV}{\mathcal{V}}

44: 

45: \newcommand{\x}{u}

46: 

47: \newcommand{\Mm}{{\bf M}}

48: 

49: \newcommand{\Ff}{F}

50: 

51: \newcommand{\2}{\sqrt{2\,}}

52: \newcommand{\half}{\tfrac{1}{2}}

53: \newcommand{\quart}{\tfrac{1}{4}}

54: 

55: \newcommand{\ab}{\bar{a}}

56: \newcommand{\bb}{\bar{b}}

57: \newcommand{\cb}{\bar{c}}

58: \newcommand{\gb}{\bar{g}}

59: \newcommand{\vb}{\bar{v}}

60: \newcommand{\wb}{\bar{w}}

61: \newcommand{\lab}{\bar{\lambda}}

62: \newcommand{\la}{\lambda}

63: \newcommand{\I}{\mathrm{i}}

64: \newcommand{\e}{\mathrm{e}}

65: \newcommand{\p}{\partial}

66: 

67: \newcommand{\rmd}{{\rm d}}

68: 

69: \newcommand{\mh}{\hat{\mu}}

70: \newcommand{\nh}{\hat{\nu}}

71: \newcommand{\rh}{\hat{\rho}}

72: \newcommand{\sh}{\hat{\sigma}}

73: \newcommand{\hg}{\hat{\gamma}}

74: \newcommand{\hC}{\hat{C}}

75: 

76: \newcommand{\Vf}{\mathcal{F}}

77: \newcommand{\Li}{\mathrm{Li}_3}

78: 

79: \newcommand{\Xb}{\bar{X}}

80: \newcommand{\Fb}{\bar{F}}

81: \newcommand{\Wb}{\bar{W}}

82: \newcommand{\zb}{\bar{z}}

83: 

84: \newcommand{\vr}[1]{{\vec{r}\,}^{#1}}

85: 

86: \newcommand{\paslash}{\ensuremath

87: \raisebox{0.025cm}{\slash}\hspace{-0.25cm}\partial\/} 

88: \newcommand{\dfrakslash}{\ensuremath \slash\hspace{-0.20cm} \mathfrak{d}\/}

89: \newcommand{\Dslash}{\ensuremath \raisebox{0.025cm}{\slash}\hspace{-0.32cm} D}

90: %%%%%%%%%%%

91: \def\slash#1{\rlap{\hbox{$\mskip 1 mu /$}}#1}      % good slash for lower case

92: \def\Slash#1{\rlap{\hbox{$\mskip 3 mu /$}}#1}      % " upper

93: %%%%%%%%%%%

94: %\newcommand{\ga}{\gamma}

95: %\newcommand{\gab}{\bar{\gamma}}

96: %\newcommand{\de}{\delta}

97: %\newcommand{\ep}{\varepsilon}

98: \newcommand{\eps}{\epsilon}

99: \newcommand{\epsb}{\bar{\epsilon}}

100: %\newcommand{\si}{\sigma}

101: 

102: 

103: \renewcommand{\thesection}{\arabic{section}}

104: \renewcommand{\theequation}{\thesection.\arabic{equation}}

105: 

106: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

107: \begin{document}

108: 

109: \begin{flushright} \small

110:  ITP--UU--07/21 \\ SPIN--07/14 

111: \end{flushright}

112: \bigskip

113: 

114: \begin{center}

115:  {\LARGE\bfseries Conifold singularities, resumming instantons  \\[1.2ex]

116:  and non-perturbative mirror symmetry }

117: \\[10mm]

118:  Frank Saueressig and Stefan Vandoren \\[3mm]

119:  {\small\slshape

120:  Institute for Theoretical Physics \emph{and} Spinoza Institute \\

121:  Utrecht University, 3508 TD Utrecht, The Netherlands \\

122:  {\upshape\ttfamily F.S.Saueressig,

123:  S.Vandoren@phys.uu.nl} }

124: \end{center}

125: \vspace{5mm}

126: 

127: \hrule\bigskip

128: 

129: \centerline{\bfseries Abstract} \medskip

130: 

131: We determine the instanton corrected hypermultiplet moduli space in type IIB

132: compactifications near a Calabi-Yau conifold point where the size of 

133: a two-cycle shrinks to zero. We show that D1-instantons resolve the 

134: conifold singularity caused by worldsheet instantons. Furthermore, by 

135: resumming the instanton series, we reproduce exactly the results obtained 

136: by Ooguri and Vafa on the type IIA side, where membrane instantons correct 

137: the hypermultiplet moduli space. Our calculations therefore establish 

138: that mirror symmetry holds non-perturbatively in the string coupling.

139: 

140: \bigskip

141: 

142: \hrule\bigskip

143: %--------------------------------------------------------------------------

144: \section{Introduction}

145: \setcounter{equation}{0}

146: 

147: 

148: 

149: 

150: Recently, Calabi-Yau singularities have played a prominent role in ``bottom-up'' approaches connecting string theory to particle physics. The basic idea behind these constructions is to first locate a set of D-branes giving rise to the desired particle physics model at the singularity and later on perform the embedding into a compact Calabi-Yau threefold (CY$_3$). In this paper we study the simplest type of CY$_3$ singularities, the conifold, and determine the non-perturbative quantum corrections to the effective action for the (bulk) modulus which controls the size of the vanishing cycle. We expect that these type of corrections will become relevant when embedding the singularity into a full-fledged CY$_3$ compactification.

151: 

152: Geometrically, a conifold point is a point in the moduli space where the CY$_3$ becomes singular by developing a set of conical singularities (nodes) with base $S^2 \times S^3$. Locally these nodes can be resolved by either carrying out a deformation, by expanding the node into an $S^3$, or a small resolution expanding the node into an $S^2$. The process of shrinking a set of two-cycles $S^2$ to points and subsequently re-expanding the singularities into $S^3$'s (or vice versa) is called a conifold transition and connects moduli spaces of CY$_3$ with different Hodge numbers \cite{Candelas:1989ug,Aspinwall:1993nu}.

153: 

154: 

155: When approaching a conifold point by degenerating a complex structure ($S^3 \rightarrow 0$) and a K\"ahler structure ($S^2 \rightarrow 0$) in type IIB and type IIA string compactifications, respectively, the vector multiplet moduli space of the low energy effective action (LEEA) develops a logarithmic singularity. 

156: These singularities can be attributed to illegally integrating out D3 (D2) branes wrapping the vanishing cycles, which, at the conifold point, give rise to extra massless states \cite{SBH}. On the other hand we can also approach the conifold point by degenerating a K\"ahler structure on the type IIB or a complex structure on the type IIA side. This leads to a logarithmic singularity in the hypermultiplet sector of the LEEA. In this case, however, the theory has no BPS states which could wrap the vanishing cycles and could have an interpretation in terms of four-dimensional particles. One expects, however, that non-perturbative string effects originating from instantons \cite{Becker:1995kb} become important in this regime since the real part of their instanton actions are proportional to the volume of the shrinking cycle so that they are no longer suppressed in the limit where the cycle shrinks to zero. See fig. 1 for a schematic illustration. Indeed, as was shown in \cite{OV} in the context of IIA string theory, spacetime instanton effects survive in the effective field theory even in the rigid limit where gravity decouples, $M_{\rm Pl} \rightarrow \infty$.

157: 

158: More recently, new exact results were obtained in \cite{Robles-Llana:2006is} for IIB strings, in which the contributions coming from worldsheet instantons, D1-instantons and D(-1)-instantons to the effective action were determined. Since these results are obtained at a generic point in the moduli space, we can study the behavior near the conifold point, where a two-cycle shrinks to zero size and

159: gravity is decoupled. 

160: In this limit only worldsheet and D1-instanton corrections survive, and we obtain the resulting IIB hypermultiplet moduli space metric in the neighborhood of the conifold.

161: 

162: Our analysis allows us to perform a non-perturbative test of mirror symmetry, which states that the hypermultiplet moduli spaces in type IIA and type IIB on the mirror, after including all quantum corrections, must be the same. After resumming the instanton series on the IIB side, we determine the mirror map and show that the resulting hyperk\"ahler geometry is exactly the one obtained in \cite{OV}. This provides a nice demonstration of open string mirror symmetry on the hypermultiplet moduli space. 

163: 

164: 

165: 

166: 

167: 

168: 

169: \begin{figure*}[t]

170: \setlength{\unitlength}{1cm}

171: \epsfxsize=1\textwidth

172: \begin{center}

173: \leavevmode

174: \epsffile{Conifold.eps}\,

175: \end{center}

176: \parbox[c]{\textwidth}{\caption{\label{fig}{\footnotesize Illustration of 

177: the CY$_3$ moduli space close to a conifold point. For more information on conifold singularities we refer to \cite{SBH,Greene:1996dh}.

178: }}}

179: \end{figure*}

180: 

181: %--------------------------------------------------------------------------

182: \section{IIA: Summing up membrane instantons}

183: \setcounter{equation}{0}

184: %--------------------------------------------------------------------------

185: %

186: In this section, we review the results of \cite{OV} who studied the

187: geometry of the type IIA hypermultiplet (HM) moduli space near a conifold 

188: singularity associated with vanishing three-cycle ${\cal C}_3$,

189: \begin{equation}

190: \mbox{IIA HM conifold limit:}\quad z = \int_{{\cal C}_3} \Omega 

191: \rightarrow 0\ .

192: \end{equation}

193: To decouple gravity, we consider the combined limit

194: %

195: \be\label{CFlimitIIA}

196: z\rightarrow 0 \, , \quad  \lambda \rightarrow 0 \, , \quad

197: {\rm with} \; \; \;   \frac{|z|}{\lambda} = {\rm finite} \, ,

198: \ee

199: %

200: where $\lambda$ is the string coupling constant. 

201: In this limit, the moduli space becomes a four-dimensional hyperk\"ahler 

202: space which at the classical level, develops a singularity at $z=0$. 

203: The metric can be written as

204: \begin{equation}\label{HK-metric}

205: {\rm d}s^2= \lambda^2 [V^{-1}({\rm d}t-\vec{A}\cdot {\rm d}{\vec y})^2

206: +V |{\rm d}{\vec y}|^2]\ .

207: \end{equation}

208: Here, $\vec {y}=(\x,z/\lambda,\bar z/\lambda)$ with $\lambda$ kept fixed and $\x$ and $t$ are the RR scalars originating from the expansion of the RR 3-form with respect to the harmonic three-forms associated with the vanishing cycle $\cC_3$ and its dual. The metric \eqref{HK-metric} is hyperk\"ahler if

209: \begin{equation}

210: V^{-1}\Delta V =0 \ ,\qquad \vec{\nabla} V = \vec{\nabla} \times

211: \vec{A}\ ,

212: \end{equation}

213: where 

214: \begin{equation}

215: \Delta = \partial_{\x}^2+4\lambda^2\partial_z\partial_{\bar z}\ .

216: \end{equation}

217: Classically, at string tree-level and for large $|z|$, the metric is determined by

218: \begin{equation}\label{cfsol}

219: V = \frac{1}{4\pi}\ln \Big(\frac{1}{z\bar{z}}\Big)\ , \qquad

220: A_\x = \frac{\I}{4 \pi} \ln\left( \frac{z}{\zb} \right) \, , \; A_z = 0 \, , \; A_{\zb} = 0 \, ,

221: \end{equation}

222: and has a logarithmic singularity. Using T-duality, which exchanges vector 

223: multiplets and hypermultiplets, this singularity has a counterpart

224: in the vector multiplet moduli space of the IIB theory compactified

225: on the {\it same} Calabi-Yau threefold, where it corresponds to the

226: appearance of massless black holes \cite{SBH}. In fact, the hyperk\"ahler

227: metric \eqref{HK-metric} is related to the vector multiplet moduli space

228: metric by the rigid c-map \cite{CFG,rcmap}. We demonstrate this in Appendix A. This 

229: will be important for us, since we use a similar mechanism for the mirror

230: theory in the next section.

231: 

232: In \cite{OV} Ooguri and Vafa studied the resolution of the singularity 

233: based on D2-brane instanton contributions. Thereby they focused on the situation where the period with respect to $\cC_3$ (A-cycle, say) vanishes while the dual period (from the B-cycle) remained finite. In this case membrane instantons wrapping the vanishing cycle 

234: generate exponential corrections to the hypermultiplet moduli space of the 

235: form $\exp(-|z|/\lambda)$ with $\theta$-angle $\exp(2 \pi \I \x)$, breaking the shift symmetry in $\x$ to a discrete subgroup. Membrane instantons wrapping the dual B-cycle decouple in the rigid limit \eqref{CFlimitIIA}, so that the shift-symmetry in $t$ is unbroken. The instanton corrected $V$ was then found to be \cite{OV}

236: \be\label{V-sum}

237: V = \frac{1}{4 \pi} \ln\left( \frac{\mu^2}{z \zb} \right) 

238: + \frac{1}{2 \pi} \sum_{m \not = 0} K_0\left(2 \pi \frac{|m z|}{\lambda}

239: \right) 

240: \; {\rm e}^{ 2 \pi \I m \x }\ ,

241: \ee

242: for some constant $\mu$.

243: This instanton sum contains the zero-th order modified Bessel function, 

244: accompanied

245: by theta-angle-like terms set by the RR scalar $\x$. The Bessel function 

246: can further be expanded for large argument, yielding exponentially suppressed

247: terms of the form $\exp[-2\pi(|mz|/\lambda -im\x )]$ together with an infinite

248: power series in $\lambda$ that describe the perturbative fluctuations 

249: around the instantons.

250: 

251: To exhibit the resolution of the singularity, one can perform a 

252: Poisson resummation,

253: \begin{equation}

254: V=\frac{1}{4\pi} \sum_{n=-\infty}^{\infty}\left(\frac{1}{{\sqrt{(\x-n)^2

255: +z\bar{z}/\lambda^2}}}-\frac{1}{|n|}\right)+\rm{const}\ ,

256: \end{equation}

257: which leads to a regular metric at $z=0$.

258: 

259: In the case of $N$ three-cycles shrinking to zero size, it was argued in 

260: \cite{OV,Greene:1996dh} that this leads to hyperk\"ahler metrics with $\Com^2/Z_N$ singularities. This metric is again of the form \eqref{HK-metric}, with 

261: $V\rightarrow NV$. In the next sections, 

262: we will reproduce all these results from type IIB strings compactified on 

263: the mirror Calabi-Yau, in which $N$ now counts the number of vanishing 

264: two-cycles. Thereby, we perform a non-perturbative test of mirror symmetry.

265: 

266: 

267: %--------------------------------------------------------------------------

268: \section{Conifold singularities in type IIA vector multiplets}

269: \setcounter{equation}{0}

270: %--------------------------------------------------------------------------

271: %

272: 

273: 

274: 

275: To understand the origin of the conifold singularity in the IIB hypermultiplet

276: moduli space, it is insightful to first study its counterpart on the type 

277: IIA vector multiplet side. The two sectors are related by T-duality and, at string tree-level, the IIB hypermultiplet moduli space is obtained from the IIA 

278: vector multiplet moduli space by the c-map \cite{CFG,Ferrara:1989ik}. 

279: At a generic point in the moduli

280: space, the special geometry is determined by a holomorphic prepotential $F(X)$ homogeneous of degree two, which receives perturbative $\alpha^\prime$ corrections from the worldsheet conformal field theory and worldsheet instantons

281: %

282: \be

283: F(X) = F_{\rm cl}(X) + F_{\rm pt}(X) + F_{\rm ws}(X)\ .

284: \ee 

285: %

286: Here

287: %

288: \be\label{VMPP}

289: \begin{split}

290: F_{\rm cl}(X) = & \, \frac{1}{3!} \, \kappa_{abc} \, \frac{X^a X^b X^c}{X^1} \, , \qquad

291: F_{\rm pt}(X) =  \, \I \, \frac{ \zeta(3)}{2 (2 \pi)^3} \, \chi_E \, (X^1)^2 \, , \\

292: F_{\rm ws}(X) = & \,  - \I \,\frac{1}{(2 \pi)^3} \, (X^1)^2 \, \sum_{k_a} \, n_{k_a} \, {\rm Li}_3\left( {\rm e}^{2 \pi \I k_a X^a / X^1} \right) \, ,

293: \end{split}

294: \ee

295: with $\kappa_{abc}, \chi_E$ and $n_{k_a}$ the triple intersection numbers, 

296: Euler number and instanton numbers of the CY$_3$ \cite{Candelas:1990rm} respectively 

297: (see, e.g., \cite{Hori} for additional background).

298: 

299: Microscopically the scalar fields in the vector multiplet sector arise from expanding the K\"ahler form $J$ and the ten-dimensional NS two-form $\hat B$ in terms of harmonic two-forms $\omega_a$ of the CY$_3$ \cite{Bodner:1990zm},

300: %

301: \be

302: \hat B = B_2 + b^a \, \omega_a \; , \quad J = t^a \, \omega_a \; , \quad a = 2, \ldots , h^{(1,1)} + 1 \, .

303: \ee

304: %

305: These fields are combined into complexified K\"ahler moduli

306: %

307: \be

308: z^a = b^a + \I t^a \,  = \frac{X^a}{X^1} \, . 

309: \ee

310: %

311: It is useful to factor out $X^1$ and work with the holomorphic function $f(z)$

312: determined by

313: \begin{equation}

314: F(X) = (X^1)^2 f(z)\ ,

315: \end{equation}

316: which is also computed by the genus zero topological string amplitude.

317: 

318: 

319: We are now interested in the conifold limit of the prepotential given above. In the CY$_3$ geometry the conical singularity is obtained by shrinking the size of a holomorphic two-cycle $\cC_\star$ to zero:

320: %

321: \be\label{gcs}

322: \mbox{geometrical conifold singularity:} \quad t^\star \rightarrow 0 \, . 

323: \ee

324: %

325: 

326: We note, however, that the condition \eqref{gcs} is not sufficient for causing a singularity in the vector multiplet moduli space. Here the singularity arises if the {\it complexified} K\"ahler modulus is taken to zero:

327: %

328: \be\label{mscs}

329: \mbox{moduli space conifold singularity:} \quad z^\star \rightarrow 0 \, . 

330: \ee

331: %

332: This implies that we can avoid hitting the singularity by giving a non-vanishing real part $b^\star$ to the complexified K\"ahler modulus. Thus in the moduli space the conifold singularity is a line of complex codimension one\footnote{This is different from the five-dimensional case \cite{FrankPhD} where such singularities are of real codimension one, so that a generic trajectory moving on the moduli space will not be able to avoid the singularity.}.

333: 

334: We can now take the conifold limit \eqref{mscs} for the prepotentials \eqref{VMPP}. Henceforth we consider the case of a conical singularity where one particular complexified K\"ahler modulus $z^\star = k_a z^a$ (for one particular and fixed vector $k_a$) shrinks to zero\footnote{We will drop the ''$\star$'' in the following.}, while the others are frozen to constant values. By inspection one then finds that the second derivatives of $f$ (determining the metric) arising from $f_{\rm cl}(z)$ and $f_{\rm pt}(z)$ are regular in this limit. Applying the expansion formula \eqref{Li3exp} to the worldsheet instanton contribution, one obtains (we

335: denote $N=n_{k_a}$ for the fixed vector $k_a$)

336: %

337: \be\label{fcfl}

338: f_{\rm ws}(z) = \frac{N}{4 \pi \I} \, z^2 \, \ln(z) + \ldots \, ,

339: \ee

340: %

341: where the dots give rise to regular contributions in the Lagrangian. Computing

342: %

343: \be

344: \p_z f_{\rm ws} = \frac{N}{2 \pi \I} \, z \, \ln(z) + \ldots \, ,

345: \ee

346: %

347: one finds that this is in precise agreement with the singular behavior 

348: found in the IIB vector multiplet sector when 

349: going to the conifold point by shrinking $N$ 

350: Lagrangian three-cycles \cite{SBH}. In this case, however, $z$ is interpreted as a complex structure moduli arising from the periods of the holomorphic three-form of the CY$_3$. 

351: 

352: The qualitative results of this section have already been discussed by Strominger \cite{SBH}, where it was argued that the conifold singularities in the type IIA vector multiplet sector originate from  strong coupling effect involving worldsheet instantons.

353: 

354: 

355: %-----------------------------------------------------------------------

356: \section{Conifold singularities in IIB hypermultiplets}

357: \setcounter{equation}{0}

358: \label{sect:5}

359: %-----------------------------------------------------------------------

360: 

361: In this section, we derive the conifold singularities that arise in the

362: hypermultiplet moduli space of type IIB compactifications. As in 

363: section 2, this singularity can be obtained from the rigid c-map on

364: the vector multiplet sector of the IIA theory. Here, we will rederive it

365: in a different way, starting from a generic point in the (tree-level)  

366: hypermultiplet moduli space, and then taking the conifold limit in which 

367: gravity decouples. 

368: Our description of the hypermultiplet moduli space geometry uses the 

369: conformal tensor

370: calculus combined with methods used in projective superspace. 

371: In this way, $4n$-dimensional quaternion-K\"ahler geometry can be reformulated

372: in terms of $4(n+1)$-dimensional hyperk\"ahler geometry. For some 

373: background material we refer to \cite{deWit:1999fp,deWit:2001dj,

374: Bergshoeff:2004nf,Rocek:2005ij,RSV,deWit:2006gn}.

375: 

376: 

377: The tree-level hypermultiplet 

378: moduli space can be conveniently written down in projective 

379: superspace \cite{Gates:1984nk}, 

380: in terms of a contour integral representation \cite{Hitchin:1986ea} 

381: of the superspace Lagrangian density

382: \begin{equation}

383: \cL(v, \vb, x) = {\rm Im} \oint \, \frac{{\rm d} \zeta}{2 \pi \I \zeta} \, 

384: H(\eta^I(\zeta))\ ,

385: \end{equation}

386: in terms of $h_{1,2}+2$ $N=2$ tensor multipets

387: %

388: \begin{equation}\label{TM}

389: \eta^I(\zeta)=\frac{v^I}{\zeta}+x^I-{\bar v}^I\zeta\ ,

390: \end{equation}

391: %

392: consisting of $N=1$ real linear multiplets $x^I$ and $N=1$ chiral multiplets

393: $v^I$. The Lagrangian density satisfies

394: \begin{equation}

395: {\cal L}_{x^Ix^J}+{\cal L}_{v^I{\bar v}^J}=0\ ,

396: \end{equation}

397: and expresses the fact that the dual hypermultiplets parameterize a 

398: hyperk\"ahler manifold with $h_{1,2}+2$ commuting shift symmetries 

399: \cite{Hitchin:1986ea}. Tensor multiplets can be used because the 

400: hypermultiplet geometries 

401: we need to consider have enough commuting isometries\footnote{This is correct 

402: in the absence of three-brane and five-brane instantons, which are not relevant

403: for the purpose of this paper.}. The scalars of the tensor multiplets 

404: transform as a triplet under $SU(2)$ R-symmetry 

405: %

406: \be\label{SU2vec}

407: \vr{I} = \left[2 \, v^I, \, 2 \, \vb^I, \, x^I \right] \, , \quad \vr{I} \cdot \vr{J} = 2 v^I \vb^J + 2 v^J \vb^I + x^I x^J \, .

408: \ee

409: For a given prepotential $F(X)$ encoding the vector multiplet couplings,

410: the dual tensor multiplet Lagrangian after the (local) c-map can be 

411: obtained by evaluating the following contour integral \cite{Rocek:2005ij}

412: %

413: \be\label{cint}

414: \cL(v, \vb, x) = {\rm Im} \oint \, \frac{{\rm d} \zeta}{2 \pi \I \zeta} \, \frac{ F(\eta^\Lambda)}{\eta^0} \, .

415: \ee

416: %

417: Here $\eta^I = \{ \eta^0 , \eta^\Lambda \}$ with $\eta^0$ being the conformal compensator. The contour integral is taken around one of the roots $\zeta_+$ of $\zeta\eta^0$ and can be evaluated in a gauge invariant way \cite{Neitzke:2007ke}~\footnote{In \cite{sergei} a slightly more complicated formula for $\cL$ has been given, taking into account the logarithmic singularity at $\zeta = 0$. The two expressions, however, only differ by terms linear in $x^I$ and therefore lead to the same Lagrangian.}

418: %

419: \be\label{lagr-cmap}

420: \cL(v, \vb, x) =  - \frac{\I}{2 r^0} \left( F(\eta_+^\Lambda) - \bar{F}(\eta_-^\Lambda) \right) \\

421: =   - \frac{\I}{2 r^0} \left( (\eta_+^1)^2 f(z) - (\eta_-^1)^2 \bar{f}(\zb) \right) \,,

422: \ee

423: %

424: with

425: %

426: \be

427: \eta_+^\Lambda = \eta^\Lambda(\zeta_+) = x^\Lambda - \frac{x^\Lambda}{2} \left( \frac{v^\Lambda}{v^0} + \frac{\vb^\Lambda}{\vb^0} \right) - \frac{r^0}{2} \left( \frac{v^\Lambda}{v^0} - \frac{\vb^\Lambda}{\vb^0}\right) \, ,  

428: \ee

429: $\eta_-=(\eta_+)^*$ and $z^a = \eta_+^a / \eta_+^1$.

430: 

431: {}From the superspace Lagrangian density 

432: ${\cal L}$, one can compute a tensor potential 

433: \cite{deWit:2006gn}

434: \begin{equation}

435: \chi (v,\vb,x)=-{\cal L}(v,\vb,x) + x^I {\cal L}_{x^I} ,

436: \end{equation}

437: where ${\cal L}_{x^I}$ denotes the derivative with respect to $x^I$. Dualizing

438: the tensors to scalars, this potential becomes the hyperk\"ahler potential

439: of the corresponding hyperk\"ahler cone above the quaternion-K\"ahler

440: manifold \cite{deWit:2001dj}. Therefore, this function

441: determines the entire low-energy effective action.

442: Using the homogeneity properties of ${\cal L}$, one can derive the 

443: identity 

444: \begin{equation}

445: \frac{1}{2} \left( \chi_{x^Ix^J} + \chi_{v^I \vb^J} \right) = 

446: {\cal L}_{x^Ix^J}\ .

447: \end{equation}

448: The components ${\cal L}_{x^Ix^J}$ then appear in the kinetic terms of the

449: scalars $x^I$ and $v^I$ in the effective Lagrangian.

450: 

451: Close to the conifold locus $f(z)$ is given by \eqref{fcfl}.

452: Substituting into \eqref{lagr-cmap} yields

453: %

454: \be\label{Lcfl}

455: \cL^{\rm cf}(v, \vb, x) 

456: = - \frac{N}{8 \pi r^0} \left( (\eta_+^1)^2 \, z^2 \, \ln(z) + (\eta_-^1)^2 \, \zb^2 \, \ln(\zb) \right) \,.

457: \ee

458: %

459: Here, only one tensor multiplet 

460: ($v,\vb,x$ with $x=k_ax^a$ etc.) captures the degrees of 

461: freedom, and all others are frozen to constants.

462: As one can explicitly check, this function satisfies

463: %

464: \be\label{dgl1}

465: \left( \p_v \p_{\vb} + \p_x^2 \right) \cL^{\rm cf}(v, \vb, x) = 0 \, .

466: \ee

467: %

468: This is precisely the constraint coming from rigid $N=2$ supersymmetry and 

469: expresses the fact that the geometry is four-dimensional hyperk\"ahler. 

470: This is consistent

471: with the fact that in this limit, gravity is decoupled, and the target 

472: space of hypermultiplets becomes hyperk\"ahler\footnote{The rigid limit in 

473: special K\"ahler geometry was studied in detail in 

474: \cite{Billo:1998yr}. It would be desirable to have a similar study for 

475: hypermultiplets.}.

476: 

477: The function ${\cal L}$ is not yet to be compared with the function $V$

478: appearing in the hyperk\"ahler metric \eqref{HK-metric}. As shown in 

479: \cite{Hitchin:1986ea}, the relation is (in the dilatation gauge $r^0=1$)

480: \begin{equation}

481: V = r^0{\cal L}_{xx}\ .

482: \end{equation}

483: Straightforward computation shows that, up to an additive 

484: constant\footnote{This additive constant contributes to the parameter

485: $\mu$ in \eqref{V-sum} and depends on the particular CY$_3$ under 

486: consideration.},

487: %

488: \be\label{Lxxsing}

489: V = r^0{\cal L}_{xx} = - \frac{N}{4 \pi} \ln(z \zb) \, ,

490: \ee

491: which precisely matches \eqref{cfsol}. 

492: This shows that at string tree-level, mirror symmetry works.

493: 

494: %-------------------------------------------------------------------------------------------

495: \section{IIB: Resummation of D1-instantons}

496: \setcounter{equation}{0}

497: %-------------------------------------------------------------------------------------------

498: %

499: The starting point for including the D1-instantons is the modular invariant tensor potential \cite{Robles-Llana:2006is}

500: %

501: \be \label{TP}

502: \chi_{(1)} =  - \frac{r^0 \tau_2^{1/2}}{(2 \pi)^3}

503: \sum_{ k_a  } n_{k_a} \sideset{}{'} \sum_{m,n} \frac{\tau_2^{3/2}}{|m\tau + n|^3}\,

504:   \big( 1 + 2 \pi |m\tau + n|\, k_a t^a \big) \, \e^{-S_{m,n}}\ ,

505: \ee

506: with instanton action

507: \be

508:   S_{m,n} = 2\pi k_a \big( |m\tau + n|\, t^a - \I m\, c^a - \I n\, b^a

509:   \big)\ .

510: \ee

511: %

512: The primed sum is taken over all integers $(m,n) \in \Zom^2 \backslash (0,0)$.

513: Here we used the notation and conventions as in \cite{Robles-Llana:2006is}, and adapted the normalization in such a way that it is consistent with the prepotential \eqref{VMPP}. The formula \eqref{TP} contains all contributions coming from

514: both worldsheet instantons (sum over $n$) and D1-instantons (sum over $m$), which are the only relevant configurations that survive in the conifold limit\footnote{Reference \cite{Robles-Llana:2006is} also 

515: determined the contributions from

516: D(-1) instantons. They yield exponential corrections of the type

517: $\exp(-|m|\tau_2)$ and therefore vanish in the limit of vanishing string 

518: coupling. Similar arguments show that three-brane and five-brane instantons

519: decouple in the conifold limit \eqref{CFlimitIIB}.}. The instanton action contains the dilaton-axion complex

520: \begin{equation}

521: \tau = \tau_1 + \I \tau_2 = a + \I {\rm e}^{-\phi}\ ,

522: \end{equation}

523: and the string coupling constant is given by $\lambda = {\rm e}^\phi$.

524: Furthermore, the $c^a$ are RR scalars that generate the theta-angle like terms

525: for the D1-instantons. The relation between the ``microscopic'' scalars $\tau, z^a, c^a$ and the scalars appearing in the tensor multiplets \eqref{TM} is given by \cite{Berkovits:1998jh,Neitzke:2007ke,Robles-Llana:2006is}

526: %

527: \be

528: \tau = \frac{1}{(r^{0})^2} \left(\vr{0} \cdot \vr{1} + \I \, | \vr{0} \times \vr{1} | \right) \, , \;

529: z^a = \frac{\eta_+^a}{\eta_+^1} \, , \; c^a = \frac{(\vr{0} \times \vr{1}) \cdot (\vr{1} \times \vr{a})}{|\vr{0} \times \vr{1}|^2} \, .

530: \ee

531: %

532: 

533: Following the discussion in section 4 we compute the function $\cL_{xx}$ arising from \eqref{TP}. The corresponding calculation can be simplified by noting that both $\chi_{(1)}$ and $\cL_{xx}$ are invariant under local $SU(2)$ R-symmetry. Thus we can adopt a particular $SU(2)$ gauge, e.g., setting $x^0 = x^1 = 0, v^0 = \vb^0$ and then taking the derivatives of \eqref{TP} with respect to 

534: $x,v,\vb$. Re-expressing the result in gauge invariant variables we find

535: \be\label{LxxD1}

536: \cL_{xx}(x,v,\vb) = \frac{N}{4 \pi r^0} \sideset{}{'} \sum_{m,n} \frac{1}{|m \tau + n|} 

537: \, \e^{ -2 \pi \, \left( | m \tau + n| \, t - \I m c - \I n b \right) } \, .

538: \ee

539: To compare with the type IIA results obtained by Ooguri and Vafa we also have to take the conifold limit. On the type IIB side this corresponds to

540: %

541: \be\label{CFlimitIIB}

542: t \rightarrow 0 \, , \quad b \rightarrow 0 \, , \quad \tau_2 \rightarrow \infty \, , \quad {\rm with} \; \; \; \tau_2 |b + \I t| = {\rm finite} \, .

543: \ee

544: %

545:  Taking this limit requires resumming the instanton corrections appearing in \eqref{LxxD1}. For this purpose we split the double sum into the contributions coming from worldsheet instantons, $m=0$, and the D1-instantons plus their bound states, $m \not = 0, n \in \Zom$:

546: %

547: \be\label{Lxxsplit}

548: \begin{split}

549: \cL_{xx}(x,v,\vb) = & \, \frac{N}{4 \pi r^0}  \sum_{n \not = 0} \frac{1}{|n|} 

550: \e^{ -2 \pi \left( \, |n| \, t - \I n b \right) } 

551: %\, \\ & \, 

552: +  \frac{N}{4 \pi r^0} \sum_{m \not = 0} \, \sum_{n \in \Zom} \frac{1}{|m \tau + n|} 

553: \, \e^{ -2 \pi \left( \, |m \tau + n| \, t - \I m c - \I n b \right) } \, .

554: \end{split}

555: \ee

556: %

557: The first term can be summed up easily

558: %

559: \be\label{int1}

560: \begin{split}

561: \frac{N}{4 \pi r^0} \, \sum_{n \not = 0} \frac{1}{|n|} \exp

562: \Big\{ -2 \pi \left( |n| t - \I n b \right) \Big\} 

563: & \, = - \frac{N}{4 \pi r^0} \ln\left(1 - {\rm e}^{2 \pi \I z} \right) + \mbox{c.c.} \\

564: & \, \simeq  - \frac{N}{4 \pi r^0} \ln\left( z \zb \right) \, ,

565: \end{split}

566: \ee

567: %

568: where we took the conifold limit $z = b+\I t \rightarrow 0$ in the second line. Observe that this expression precisely reproduces \eqref{Lxxsing}. In order to take the conifold limit in the second line of \eqref{Lxxsplit} we first carry out a Poisson resummation in $n$. Using the results of appendix \ref{SecB.2} we find

569: %

570: \be\label{int2}

571: \begin{split}

572: \frac{1}{4 \pi r^0} & \sum_{m \not = 0} \, \sum_{n \in \Zom} \frac{1}{|m \tau + n|} 

573: \, \e^{ -2 \pi \left( \, |m \tau + n| \, t - \I m c - \I n b \right) } \\

574: & = \frac{1}{2 \pi r^0} \sum_{m \not = 0} \, \sum_{n \in \Zom} 

575: K_0\left( 2 \pi |m \tau_2| \sqrt{t^2 + (b + n)^2} \right) \e^{2 \pi \I m (c - \tau_1 (b+n))} \\

576: & \simeq \frac{1}{2 \pi r^0} \sum_{m \not = 0} K_0(2 \pi \tau_2 |m z|) \; {\rm e}^{ 2 \pi \I m (c - \tau_1 b) } \, .

577: \end{split}

578: \ee

579: %

580: Here we have taken the conifold limit \eqref{CFlimitIIB} in the second step. Note that in this limit the sum over $n$ localizes such that only the $n=0$ part gives a non-zero contribution.

581: 

582: 

583: Combining \eqref{int1} and \eqref{int2}, we then obtain the D1-instanton

584: corrected $\cL_{xx}$ in the conifold limit

585: %

586: \be

587: N^{-1}V=r^0\cL_{xx} = \frac{1}{4 \pi } \ln\left( \frac{1}{z \zb} \right) 

588: + \frac{1}{2 \pi } \sum_{m \not = 0} K_0(2 \pi \tau_2 |m z|) \; {\rm e}^{ 2 \pi \I m (c - \tau_1 b) }

589: \ee

590: %

591: 

592: Comparing this to the instanton corrected function $V$ on the type IIA side 

593: in \eqref{V-sum}, we find perfect agreement if we use the mirror map

594: %

595: \be\label{mirrormap}

596: \lambda = \tau_2^{-1} \,, \qquad z^{\rm IIA} = z^{\rm IIB} \,, \qquad \x = \pm(c - \tau_1 b) \, .

597: \ee

598: %

599: The second relation states that, under mirror symmetry, the complex structure modulus $z^{\rm IIA}$ is equated to the complexified K\"ahler modulus $z^{\rm IIB}$ associated with the vanishing cycles, while the relation between $u$ and $c - \tau_1 b$ is determined up to a sign only. Note that selecting the minus sign, eq. \eqref{mirrormap} is precisely the classical mirror map obtained in \cite{BGHL}. This shows that the classical mirror map does not receive quantum corrections 

600: once the conifold limit is taken. 

601: 

602: \medskip

603: \noindent

604: \textbf{Acknowledgments}

605: 

606: \noindent

607: We thank Cumrun Vafa for suggesting this project and for reading an earlier 

608: draft of this manuscript. We furthermore thank Bernard de Wit, 

609: Daniel Robles-Llana, Martin Ro\v{c}ek, Jan Stienstra and Ulrich Theis for 

610: valuable discussions. This work grew out of the 4th Simons Workshop in 

611: Physics and Mathematics. We thank the YITP and the Department of Mathematics 

612: at Stony Brook University for hospitality.\ FS is supported by the 

613: European Commission Marie Curie Fellowship no.\ MEIF-CT-2005-023966.

614: This work is partially supported by the European Union RTN network

615: MRTN-CT-2004-005104 and INTAS contract 03-51-6346.

616: 

617: 

618: 

619: %-------------------------------------------------------------------------------------------

620: \begin{appendix}

621: %-------------------------------------------------------------------------------------------

622: \section{Conifold singularities from the rigid c-map}

623: \setcounter{equation}{0}

624: %-------------------------------------------------------------------------------------------

625: The leading term in the type IIB vector multiplet prepotential can be determined by monodromy arguments

626: %

627: \be\label{VMPPIIB}

628: f_{\rm ws}(z) = \frac{1}{4 \pi \I} \, z^2 \, \ln(z) + \ldots \, ,

629: \ee

630: %

631: where $z$ is associated with the vanishing period of the holomorphic three-form of the CY$_3$.

632: %

633: We now interpret \eqref{VMPPIIB} as the prepotential underlying a rigid special K\"ahler geometry. We can then use the rigid c-map \cite{CFG,rcmap} to construct the dual hyperk\"ahler metric. For a general (rigid) prepotential $F(X^I)$ the resulting metric reads (up to a trivial rescaling)

634: %

635: \be\label{rcmet}

636: - \tfrac{1}{2} \rmd s^2 =  \I \left(\rmd F_I \, \rmd \bar X^I - \rmd \bar F_I \, \rmd X^I \right)

637: - N^{IJ} \left( \rmd B_I - F_{IK} \rmd A^K \right) \left( \rmd B_J - \Fb_{JL} \rmd A^L \right) \, , 

638: \ee

639: %

640: where $F_I = \p F / \p X^I$ and $N^{IJ}$ being the inverse of

641: %

642: \be

643: N_{IJ} = - \I \left( F_{IJ} -  \Fb_{IJ} \right) \, .

644: \ee

645: %

646: Evaluating \eqref{rcmet} for the prepotential \eqref{VMPPIIB} then yields

647: %

648: \be

649: \begin{split}

650: \rmd s^2 = & \, \frac{1}{2 \pi} \ln(z\zb) \, \rmd z \, \rmd \zb - \frac{\pi}{\ln(z\zb)} 

651: \left( \rmd B -  \frac{1}{2 \pi \I} \ln(z) \, \rmd A \right) \left(  \rmd B + \frac{1}{2 \pi \I} \ln(\zb) \, \rmd A \right) \, .  

652: \end{split}

653: \ee

654: %

655: Changing coordinates 

656: %

657: \be

658: B = 2 \lambda \, t \, , \; A = - 2 \lambda \, \x \, , 

659: \ee

660: %

661: one finds precisely the metric \eqref{HK-metric} obtained from the solution \eqref{cfsol}.

662: %-------------------------------------------------------------------------------------------

663: \section{Polylogarithms and resummation techniques}

664: \setcounter{equation}{0}

665: %-------------------------------------------------------------------------------------------

666: %

667: In this appendix we collect various facts used in the main part of the paper by giving a brief introduction to polylogarithmic functions and Poisson resummation in sections \ref{SecB.1} and \ref{SecB.2}, respectively.

668: %

669: \subsection{Polylogology}

670: \label{SecB.1}

671: %

672: We start by summarizing some properties of polylogarithmic functions, essentially following the appendix B of ref.\ \cite{LMZ}. For $0 < z < 1$, the k-th polylogarithm is defined via the series expansion

673: %

674: \be\label{PLD}

675: {\rm Li}_k (z) = \sum_{n=1}^{\infty} \frac{z^n}{n^k} \, .

676: \ee

677: %

678: It can be analytically continued to a multivalued function on the complex plane. Polylogarithms with different values of $k$ are related by

679: %

680: \be\label{polyrec}

681: z \frac{d}{dz} {\rm Li}_k(z) = {\rm Li}_{k-1}(z) \, .

682: \ee

683: %

684: For $k = 1$ we have

685: %

686: \be

687: {\rm Li}_1(z) = - \log(1-z) \, ,

688: \ee

689: %

690: which we used to sum up \eqref{int1}. From the definition \eqref{PLD} we find

691: %

692: \be

693: {\rm Li}_k(0) = 0 \; , \quad ( k \in \Zom) \quad {\rm and} \quad {\rm Li}_k(1) = \zeta(k) \, , \quad {\rm for} \quad k > 1 \, .

694: \ee

695: %

696: Polylogarithms at values $z$ and $1/z$ are related through the connection formula \cite{ConnectionFormula}, 

697: %

698: \be

699: {\rm Li}_k(z) + (-1)^k \, {\rm Li}_k(1/z) = - \frac{(2 \pi \I)^k}{k!} B_k\left( \frac{\log(z)}{2 \pi \I} \right) \, ,

700: \ee

701: %

702: where $B_k(\cdot)$ are the Bernoulli polynomials. For Li$_3(z)$ this yields

703: %

704: \be\label{CFLi3}

705: {\rm Li}_3(z) - {\rm Li}_3(1/z) = - \tfrac{1}{6} \log^3(z) - \tfrac{\I \pi}{2} \log^2(z) + \tfrac{\pi^2}{3} \log(z) \, .

706: \ee

707: %

708: {}From the point of view of the main part of the paper, it is more natural to work with the variable $x$, $z = {\rm e}^x$. In this case \eqref{CFLi3} becomes

709: %

710: \be\label{CFLix}

711: {\rm Li}_3(\e^x)  = {\rm Li}_3(\e^{-x}) - \tfrac{1}{6} x^3 - \tfrac{\I \pi}{2} x^2 + \tfrac{\pi^2}{3} x \, .

712: \ee

713: %

714: The conifold point corresponds to $x = 0$. At this point the function ${\rm Li}_3(\e^{-x})$ has a logarithmic branch point

715: %

716: \be\label{Li3expans}

717: {\rm Li}_3(\e^{-x}) \simeq q(x) \log(x) + p(x) \quad {\rm for} \; x \rightarrow 0 \, , 

718: \ee

719: %

720: where $q(x)$ and $p(x)$ are power series

721: %

722: \be

723: q(x) = \sum_{j=0}^\infty q_j x^j \; , \qquad p(x) = \sum_{j=0}^\infty p_j x^j \, .

724: \ee

725: %

726: Analytically continuing the ansatz \eqref{Li3expans} to Li$_3(\e^{x})$ using $\log(-x) = \log(x) + \I \pi$ and substituting into the connection formula \eqref{CFLix} we obtain the following expansion for small $x$

727: %

728: \be\label{Li3exp}

729: \Li \left(\e^{-x} \right) = - \frac{1}{2} x^2 \ln(x) + p(x)\ ,

730: \ee

731: %

732: where $p(x) = \zeta(3) - \zeta(2)x + \tfrac{3}{4}x^2 + \frac{1}{12} x^3 + \cO(x^4)$ is polynomial in $x$. With this identity it is then straightforward to determine the conifold limit of the prepotential \eqref{VMPP}.

733: %

734: \subsection{Poisson resummation}

735: \label{SecB.2}

736: %

737: Taking the conifold limit in the D1-brane instanton sector in 

738: section 5 requires a Poisson resummation in the worldsheet instanton number $n$. The technical details of this computation are collected in this appendix.

739: 

740: The basic ingredient for Poisson resummation is the following identity for the Dirac delta-distribution

741: %

742: \be

743: \sum_{n \in \Zom} \delta(y - na) = \frac{1}{a} \sum_{n \in \Zom} {\rm e}^{2 \pi \I n y/a} \, , \; a \in \Rom^+ \, .

744: \ee

745: %

746: Multiplying with an arbitrary function $f(x+y)$ and integrating over $ y \in \Rom$ gives the Poisson resummation formula

747: %

748: \be\label{PRF}

749: \sum_{n \in \Zom} f(x+na) = \frac{1}{a} \sum_{n \in \Zom} \tilde{f}(2 \pi n /a) \, \e^{2 \pi \I n x/a} \, . 

750: \ee

751: %

752: Here $f(x)$ and $\tilde f(k)$ are related by Fourier-transformation

753: %

754: \be

755: \tilde f(k) = \int_{-\infty}^{\infty} {\rm d}x \, f(x) \, \e^{-ikx} \; , \qquad f(x) = \frac{1}{2 \pi} \int_{-\infty}^{\infty} {\rm d}k \, \tilde f(k) \, \e^{ikx} \, . 

756: \ee

757: %

758: We now apply this resummation to the second term in \eqref{Lxxsplit}. Comparing

759: %

760: \be

761: \begin{split}

762: \sum_{n \in \Zom} \frac{1}{\left| m \tau + n \right|} \, \e^{- 2 \pi (| m \tau + n| \, t - \I n b)} 

763: = \sum_{n \in \Zom} \frac{1}{\sqrt{ (m \tau_2)^2 + (n + m \tau_1)^2 }} \, \e^{- 2 \pi ( \sqrt{ (m \tau_2)^2 + (n + m \tau_1)^2} \, t - \I n b)}

764: \end{split}

765: \ee

766: %

767: to the general formula \eqref{PRF} we identify

768: \be\label{B.15}

769: \tilde f(2 \pi n) = \frac{2 \pi}{(\alpha^2 + (2 \pi n + \gamma)^2)^{1/2}} \e^{- (\alpha^2 + (2 \pi n + \gamma)^2)^{1/2} t} \, , 

770: \ee

771: %

772: together with $a = 1, x = b, \alpha = 2 \pi m \tau_2$, and $\gamma = 2 \pi m \tau_1$. The (inverse) Fourier transform of \eqref{B.15} can be found using the following formula for Fourier cosine transformations \cite{IT}:

773: %

774: \be\label{FCT1}

775: \int_0^\infty {\rm d}x \, (x^2 + \alpha^2)^{-1/2} \, {\rm e}^{-\beta  \,(x^2 + \alpha^2)^{1/2}} \cos(xy)  =   K_0\left[\alpha (\beta^2 + y^2)^{1/2} \right] \, .

776: \ee

777: %

778: Substituting the result back into \eqref{PRF} then establishes the identity

779: %

780: \be\label{PR1}

781: \sum_{n \in \Zom} \frac{1}{\left| m \tau + n \right|} \, \e^{- 2 \pi ( | m \tau + n| t - \I n b)} = 2 \sum_{n \in \Zom} K_0\left(2 \pi |m \tau_2| (t^2 + (b+n)^2)^{1/2} \right) \e^{- 2 \pi \I m \tau_1 (b + n)} \, . 

782: \ee

783: %

784: This completes the derivation of the first step in \eqref{int2}.

785: \end{appendix}

786: 

787: 

788: 

789: \begin{thebibliography}{00}

790: 

791: 

792: %\cite{Candelas:1989ug}

793: \bibitem{Candelas:1989ug}

794:   P.~Candelas, P.~S.~Green and T.~H\"ubsch,

795:   {\it Rolling among Calabi-Yau vacua},

796:   Nucl.\ Phys.\  B {\bf 330} (1990) 49.

797:   %%CITATION = NUPHA,B330,49;%%

798: 

799: 

800: 

801: %\cite{Aspinwall:1993nu}

802: \bibitem{Aspinwall:1993nu}

803:   P.~S.~Aspinwall, B.~R.~Greene and D.~R.~Morrison,

804:   {\it Calabi-Yau moduli space, mirror manifolds and spacetime topology 

805:   change in string theory},

806:   Nucl.\ Phys.\  B {\bf 416} (1994) 414, 

807:   \texttt{hep-th/9309097}.

808:   %%CITATION = NUPHA,B416,414;%%

809: 

810: \bibitem{SBH}

811:   A.~Strominger,

812:   {\it Massless black holes and conifolds in string theory},

813:   Nucl.\ Phys.\  B {\bf 451} (1995) 96, {\tt hep-th/9504090};\\

814:   %%CITATION = NUPHA,B451,96;%%

815: B.~R.~Greene, D.~R.~Morrison and A.~Strominger,

816:   {\it Black hole condensation and the unification of string vacua},

817:   Nucl.\ Phys.\  B {\bf 451} (1995) 109, 

818:   \texttt{hep-th/9504145}.

819:   %%CITATION = NUPHA,B451,109;%%

820: 

821: 

822: 

823: 

824: %\cite{Becker:1995kb}

825: \bibitem{Becker:1995kb}

826:   K.~Becker, M.~Becker and A.~Strominger,

827:   {\it Five-Branes, membranes and nonperturbative string theory},

828:   Nucl.\ Phys.\  B {\bf 456} (1995) 130,

829:   \texttt{hep-th/9507158}.

830:   %%CITATION = NUPHA,B456,130;%%

831: 

832: 

833: %\cite{Ooguri:1996me}

834: \bibitem{OV}

835:   H.~Ooguri and C.~Vafa,

836:   {\it Summing up D-instantons},

837:   Phys.\ Rev.\ Lett.\  {\bf 77} (1996) 3296,

838:   {\tt hep-th/9608079}.

839:   %%CITATION = PRLTA,77,3296;%%

840: 

841: 

842: 

843: %\cite{Greene:1996dh}

844: \bibitem{Greene:1996dh}

845:   B.~R.~Greene, D.~R.~Morrison and C.~Vafa,

846:   {\it A geometric realization of confinement},

847:   Nucl.\ Phys.\  B {\bf 481} (1996) 513,

848:   \texttt{hep-th/9608039}.

849:   %%CITATION = NUPHA,B481,513;%%

850: 

851: %\cite{Robles-Llana:2006is}

852: \bibitem{Robles-Llana:2006is}

853:   D.~Robles-Llana, M.~Ro\v{c}ek, F.~Saueressig, U.~Theis and S.~Vandoren,

854:   {\it Nonperturbative corrections to 4D string theory effective actions from

855:   SL(2,Z) duality and supersymmetry},

856:   \texttt{hep-th/0612027}.

857:   %%CITATION = HEP-TH/0612027;%%

858: 

859: %\cite{Cecotti:1988qn}

860: \bibitem{CFG}

861:   S.~Cecotti, S.~Ferrara and L.~Girardello,

862:   {\it Geometry of type II superstrings and the moduli of superconformal field theories},

863:   Int.\ J.\ Mod.\ Phys.\  A {\bf 4} (1989) 2475.

864:   %%CITATION = IMPAE,A4,2475;%%

865: 

866: %\cite{DeJaegher:1997ka}

867: \bibitem{rcmap}

868:   J.~De Jaegher, B.~de Wit, B.~Kleijn and S.~Vandoren,

869:   {\it Special geometry in hypermultiplets},

870:   Nucl.\ Phys.\  B {\bf 514} (1998) 553,

871:   {\tt hep-th/9707262}.

872:   %%CITATION = NUPHA,B514,553;%%

873: 

874: 

875: %\cite{Ferrara:1989ik}

876: \bibitem{Ferrara:1989ik}

877:   S.~Ferrara and S.~Sabharwal,

878:   {\it Quaternionic manifolds for Type II superstring vacua of Calabi-Yau

879:   spaces},''

880:   Nucl.\ Phys.\  B {\bf 332} (1990) 317.

881:   %%CITATION = NUPHA,B332,317;%%

882: 

883: 

884: %\cite{Candelas:1990rm}

885: \bibitem{Candelas:1990rm}

886:   P.~Candelas, X.~C.~De La Ossa, P.~S.~Green and L.~Parkes,

887:   {\it A pair of Calabi-Yau manifolds as an exactly soluble superconformal

888:   theory},

889:   Nucl.\ Phys.\  B {\bf 359} (1991) 21.

890:   %%CITATION = NUPHA,B359,21;%%

891: 

892: 

893: 

894: \bibitem{Hori}

895: K.~Hori, S.~Katz, A.~Klemm, R.~Pandharipande, R.~Thomas, C.~Vafa,

896: R.~Vakil and E.~Zaslow,

897: \emph{Mirror symmetry},

898: Clay mathematics monographs, AMS\@, Providence USA\@, 2003.

899: 

900: %\cite{Bodner:1990zm}

901: \bibitem{Bodner:1990zm}

902:   M.~Bodner, A.~C.~Cadavid and S.~Ferrara,

903:   {\it (2,2) vacuum configurations for type IIA superstrings: N=2 supergravity

904:   Lagrangians and algebraic geometry},

905:   Class.\ Quant.\ Grav.\  {\bf 8} (1991) 789.

906:   %%CITATION = CQGRD,8,789;%%

907: 

908: \bibitem{FrankPhD}

909:   F.~Saueressig,

910:   {\it Topological phase transitions in Calabi-Yau compactifications of

911:   M-theory},

912:   Fortsch.\ Phys.\  {\bf 53} (2005) 5.

913:   %%CITATION = FPYKA,53,5;%%

914: 

915: 

916: %\cite{deWit:1999fp}

917: \bibitem{deWit:1999fp}

918:   B.~de Wit, B.~Kleijn and S.~Vandoren,

919:   {\it Superconformal hypermultiplets},

920:   Nucl.\ Phys.\  B {\bf 568} (2000) 475,

921:   \texttt{hep-th/9909228}.

922:   %%CITATION = NUPHA,B568,475;%%

923: 

924: 

925: %\cite{deWit:2001dj}

926: \bibitem{deWit:2001dj}

927:   B.~de Wit, M.~Ro\v{c}ek and S.~Vandoren,

928:   {\it Hypermultiplets, hyperkaehler cones and quaternion-Kaehler geometry},

929:   JHEP {\bf 0102} (2001) 039, \texttt{hep-th/0101161}.

930:   %%CITATION = JHEPA,0102,039;%%

931: 

932: 

933: %\cite{Bergshoeff:2004nf}

934: \bibitem{Bergshoeff:2004nf}

935:   E.~Bergshoeff, S.~Cucu, T.~de Wit, J.~Gheerardyn, S.~Vandoren and A.~Van Proeyen,

936:   {\it The map between conformal hypercomplex / hyper-Kaehler and

937:   quaternionic(-Kaehler) geometry},

938:   Commun.\ Math.\ Phys.\  {\bf 262} (2006) 411,

939:   \texttt{hep-th/0411209}.

940:   %%CITATION = CMPHA,262,411;%%

941: 

942: %\cite{Rocek:2005ij}

943: \bibitem{Rocek:2005ij}

944:   M.~Ro\v{c}ek, C.~Vafa and S.~Vandoren,

945:   {\it Hypermultiplets and topological strings},

946:   JHEP {\bf 0602} (2006) 062, \texttt{hep-th/0512206}.

947:   %%CITATION = JHEPA,0602,062;%%

948: 

949: 

950: %\cite{Robles-Llana:2006ez}

951: \bibitem{RSV}

952:   D.~Robles-Llana, F.~Saueressig and S.~Vandoren,

953:   {\it String loop corrected hypermultiplet moduli spaces},

954:   JHEP {\bf 0603} (2006) 081, {\tt hep-th/0602164}.

955:   %%CITATION = JHEPA,0603,081;%%

956: 

957: %\cite{deWit:2006gn}

958: \bibitem{deWit:2006gn}

959:   B.~de Wit and F.~Saueressig,

960:   {\it Off-shell N = 2 tensor supermultiplets},

961:   JHEP {\bf 0609} (2006) 062, 

962:   \texttt{hep-th/0606148}.

963:   %%CITATION = JHEPA,0609,062;%%

964: 

965: 

966: 

967: %\cite{Gates:1984nk}

968: \bibitem{Gates:1984nk}

969:   S.~J.~Gates, C.~M.~Hull and M.~Ro\v{c}ek,

970:   {\it Twisted multiplets and new supersymmetric nonlinear sigma models},

971:   Nucl.\ Phys.\  B {\bf 248} (1984) 157.

972:   %%CITATION = NUPHA,B248,157;%%

973: 

974: 

975: %\cite{Hitchin:1986ea}

976: \bibitem{Hitchin:1986ea}

977:   N.~J.~Hitchin, A.~Karlhede, U.~Lindstr\"om and M.~Ro\v{c}ek,

978:   {\it Hyperkahler metrics and supersymmetry},

979:   Commun.\ Math.\ Phys.\  {\bf 108} (1987) 535.

980:   %%CITATION = CMPHA,108,535;%%

981: 

982: 

983: %\cite{Neitzke:2007ke}

984: \bibitem{Neitzke:2007ke}

985:   A.~Neitzke, B.~Pioline and S.~Vandoren,

986:   {\it Twistors and black holes},

987:   \texttt{hep-th/0701214}.

988:   %%CITATION = HEP-TH/0701214;%%

989: 

990: 

991: \bibitem{sergei}

992:   S.~Alexandrov,

993:   {\it Quantum covariant c-map}, {\tt hep-th/0702203}.

994:   %%CITATION = HEP-TH/0702203;%%

995: 

996: %\cite{Billo:1998yr}

997: \bibitem{Billo:1998yr}

998:   M.~Billo, F.~Denef, P.~Fr\'e, I.~Pesando, W.~Troost, A.~Van Proeyen and D.~Zanon, {\it The rigid limit in special Kaehler geometry: From K3-fibrations to  special Riemann surfaces: A detailed case study},

999:   Class.\ Quant.\ Grav.\  {\bf 15} (1998) 2083,

1000:   \texttt{hep-th/9803228}.

1001:   %%CITATION = CQGRD,15,2083;%%

1002: 

1003: 

1004: 

1005: 

1006: 

1007: %\cite{Berkovits:1998jh}

1008: \bibitem{Berkovits:1998jh}

1009:   N.~Berkovits,

1010:   {\it Conformal compensators and manifest type IIB S-duality},

1011:   Phys.\ Lett.\  B {\bf 423} (1998) 265,

1012:   \texttt{hep-th/9801009}.

1013:   %%CITATION = PHLTA,B423,265;%%

1014: 

1015: \bibitem{BGHL}

1016: R.~B\"ohm, H.~G\"unther, C.~Herrmann and J.~Louis,

1017: {\it Compactification of type IIB string theory on Calabi-Yau threefolds},

1018: Nucl.\ Phys.\ \textbf{B569} (2000) 229,

1019: {\tt hep-th/9908007}.

1020: %%CITATION = HEP-TH 9908007;%%

1021: 

1022: 

1023: 

1024: \bibitem{IT}

1025: A.\ Erd\'elyi, et.\ al.\ , {\it Tables of integral transforms}, Vol.\ 1, McGraw-Hill Book Company, Inc., New York, 1954.

1026: 

1027: 

1028: 

1029: 

1030: 

1031: %\cite{Louis:2003gj}

1032: \bibitem{LMZ}

1033:   J.~Louis, T.~Mohaupt and M.~Zagermann,

1034:   {\it Effective actions near singularities},

1035:   JHEP {\bf 0302} (2003) 053, {\tt hep-th/0301125}.

1036:   %%CITATION = JHEPA,0302,053;%%

1037: 

1038: 

1039: \bibitem{ConnectionFormula}

1040: A.~P.~Prudnikov, Yu.~A.~Brychkov and O.~I.~Marikov, 

1041: {\it Integrals and Series}, Vol.\ 3, Gordon and Breach, 1986.

1042: 

1043: 

1044: \end{thebibliography}

1045: \end{document}

1046: 

1047: