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

2: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

3: %%NRC, Canadian Journal of Physics

4: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

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

6: 

7: \documentclass[genTeX]{nrc1}

8: \usepackage[pctex32]{graphicx}

9: %\usepackage[figuresright]{rotating}

10: %\usepackage{cite}

11: \newcommand{\beq}{\begin{equation}}

12: \newcommand{\eeq}{\end{equation}}

13: \newcommand{\beqa}{\begin{eqnarray}}

14: \newcommand{\eeqa}{\end{eqnarray}}

15: \newcommand{\roughly}[1]{\mathrel{\raise.3ex\hbox{$#1$\kern-0.85em

16: \lower1ex\hbox{$\sim$}}}}

17: \newcommand{\gsim}{\roughly>}

18: 

19: \setcounter{page}{1}

20: \volyear{83}{2005}

21: \journal{Can. J. Phys.}

22: \received{June 5, 2005}

23: \accepted{?????}

24: \begin{document}

25: 

26: \title{Aspects of brane-antibrane inflation}

27: %

28: \author[James M.\ Cline]{James M.\ Cline}

29: \address{Department of Physics, McGill University, Montr\'eal, Qc 

30: H3A 2T8, Canada. \email{jcline@physics.mcgill.ca}}

31: 

32: \shortauthor{Cline}

33: 

34: \maketitle

35: %

36: \begin{abstract}

37: I describe a dynamical mechanism for solving the fine-tuning

38: problem of brane-antibrane inflation.  By inflating with stacks of

39: branes and antibranes, the branes can naturally be trapped at a

40: metastable minimum of the potential.  As branes tunnel out of this

41: minimum, the shape of the potential changes to make the minimum

42: shallower.  Eventually the minimum disappears and the remaining

43: branes roll slowly because the potential is nearly flat.  I show that

44: even with a small number of branes, there is a good chance of getting

45: enough inflation.  Running of the spectral index is correlated

46: with the tilt in such a way as to provide a test of the model by

47: future CMB experiments.

48: \\\\PACS Nos.: 11.25.Wx, 98.80.Cq 

49: \end{abstract}

50: %

51: 

52: 

53: \section{Introduction}

54: Significant progress has been made recently in obtaining inflation

55: from specific string theory constructions, notably in type IIB

56: string theory in which moduli are stabilized by fluxes.  Aspects

57: of racetrack inflation, brane-antibrane inflation, and the problem

58: of reheating and cosmic string production in the latter, have been

59: reviewed elsewhere \cite{Cline}, and I will not repeat that material

60: here.  Instead I focus on a newer idea, multibrane inflation

61: \cite{CS}, which provides a simple and novel 

62: solution to the problem of fine-tuning of the inflaton potential in 

63: the KKLMMT \cite{KKLMMT} proposal for brane-antibrane inflation.

64: It is a straightforward generalization of the KKLMMT construction

65: shown in figure \ref{fig0}, where branes are attracted to antibranes

66: which live in the bottom of a Klebanov-Strassler \cite{KS} warped throat.

67: 

68: \begin{figure}

69: \centerline{\includegraphics[width=0.5\hsize,angle=0]{KS.eps}}  

70: \caption{The KKLMMT brane-antibrane inflation setup in a Calabi-Yau

71: space stabilized by fluxes and containing a warped throat.}

72: \label{fig0}

73: \end{figure}

74: 

75: 

76: 

77: The Lagrangian for the inflaton field $\psi$, which is the 

78: brane-antibrane separation, is

79: \beq

80: \label{pot1}

81: 	{\cal L} =  {1\over (2\sigma - \psi^2)^2}

82: \left(  6\sigma\dot\psi^2 - {\epsilon\, \tau\over  

83: 1+ {\epsilon\over (\psi-\psi_0)^4}}\right)

84: \eeq

85: where $\epsilon$ is the fourth power of the warp factor at the

86: bottom of the Klebanov-Strassler throat, $\tau$ is the unwarped 3-brane

87: tension, $\sigma$ is the K\"ahler modulus of the Calabi-Yau manifold,

88: and $\psi_0$ is the position of the antibrane at the bottom of the

89: throat.  It was noted in \cite{BCSQ} that, depending on the values of

90: the parameters $\tau$, $\epsilon$, $\psi_0$, the inflaton may be

91: trapped in a metastable minimum near $\psi=0$.  For inflation,

92: one needs to tune the parameters to avoid this minimum, because

93: otherwise the inflaton does not roll, but rather undergoes old inflation with

94: its infamous graceful exit problem.  A tuning at the level of 1 part

95: in 1000 was needed to obtain sufficient inflation.

96: 

97: In this discussion, it was assumed that there is just a single brane

98: and antibrane driving inflation.  However in string theory it is

99: natural to consider stacks of coincident branes or antibranes.  If we

100: make this simple generalization, then a qualitatively new effect can

101: occur.  The potential (\ref{pot1}) generalizes to 

102: \beq

103: \label{potN}

104: 	{\cal L} \to {1\over (2\sigma - \sum_i\psi_i^2)^2}

105: \left( \sum_i 6\sigma\dot\psi_i^2 - {N \epsilon\, \tau\over  

106: 1+ \sum_i {\epsilon\over (\psi_i-\psi_0)^4}}\right)

107: \eeq

108: For some limited range of the parameter 

109: $b\equiv b\equiv {\sqrt{2\sigma\epsilon}\over \psi_0^3}$, this 

110: potential has the interesting property that for large a number $N$ of

111: branes, there is a metastable minimum near $\psi=0$, but the minimum

112: disappears as $N$ decreases.  This behavior is illustrated in figure

113: \ref{fig1}.  Therefore if one starts with a large enough number of branes,

114: then the successive tunneling of single branes from the metastable

115: minimum will cause the potential to automatically tune itself to

116: become increasingly flat, until the point at which the branes

117: can start to slowly roll and drive inflation.  During the tunneling

118: phase, ``old inflation'' is occurring, which by itself is not a 

119: successful inflationary scenario, but this is unimportant for us

120: since the last phase of old inflation is followed by inflation driven

121: by the slowly rolling branes. 

122: 

123: \begin{figure}

124: \centerline{\includegraphics[width=0.5\hsize,angle=0]{multipot-fig.eps}}  

125: \caption{Evolution of brane-antibrane potentials as number of trapped

126: branes decreases due to tunneling.}

127: \label{fig1}

128: \end{figure}

129: 

130: 

131: One way to understand the origin of this behavior

132: of the potential is to consider its curvature at the origin.

133: There are competing effects which depend

134: differently on $N$: one comes from the brane-antibrane Coulomb-like

135: potential, and the other comes from a supergravity effect, with the

136: entire flat-space potential being divided by the Calabi-Yau volume,

137: $2\sigma - N\psi^2$, as modified by the presence of the branes.

138: The curvature of the potential at $\psi=0$ is proportional to

139: \beq

140: V'' \sim {1\over {4\sigma}}-5/2\,{\frac {\epsilon}{{{ \psi_0}}^{2} \left( {{

141:  \psi_0}}^{4}+N\epsilon \right) }}+4\,{\frac {N{\epsilon}^{2}}{{{

142: \psi_0}}^{2} \left( {{ \psi_0}}^{4}+N\epsilon \right) ^{2}}}

143: \eeq

144: Clearly this is always positive for large enough $N$.  But when $N=0$,

145: it is proportional to 

146: \beq

147: V'' \sim {1\over {4\sigma}}-5/2\,{\frac {\epsilon}{{{ \psi_0}}^{6}}}

148: \eeq

149: which will be negative if $b^2\equiv 2\sigma\epsilon/\psi_0^6 > 0.2$ 

150: Thus the curvature of the potential

151: changes sign as a function of $N$. This

152: turns out to be just one condition on $b$ which must be satisfied in

153: order to get the right behavior.  

154: We will show how to get around 

155: this restriction later.	

156: 

157: We have explored the parameter space of $\sigma$ and $\epsilon$

158: (the size of the internal manifold and the warping in the inflationary

159: throat) to determine how many branes are needed, and how much

160: inflation results.  The correlation between the number of $e$-foldings

161: and the number of branes is shown in figure \label{fig2},  which was obtained

162: by uniformly sampling the parameter space $-5 < \log_10\epsilon < -2$

163: and $1 < \log_10(2\sigma) < 4$.  This corresponds to a moderately

164: warped inflationary throat, and large Calabi-Yau volumes ($2\sigma$ is

165: the size of the manifold in string units), which is required for

166: consistency of the low-energy supergravity description.  The

167: figure shows the actual points as well as an empirical formula which

168: explains the complicated structure rather well.  The origin of the

169: steep lines of points is due to the fact that, if $N$ could be

170: noninteger, it would be possible to potentials with perfectly flat

171: spots (where $V'$ and $V''$ vanish simultaneously) leading to very

172: long periods of inflations.  The lines are sequences of points in

173: which the actual value of $N$ becomes increasingly close to the 

174: ideal value.  

175: 

176: 

177: \begin{figure}

178: \centerline{\includegraphics[width=0.7\hsize,angle=0]{NevNc.eps}}  

179: \caption{Number of e-foldings of inflation versus critical number

180: of branes needed for a flat potential, from scanning over the

181: $\sigma$-$\epsilon$ parameter space.}

182: \label{fig2}

183: \end{figure}

184: 

185: It is interesting to note that, even though we consider the

186: possibility of large numbers of branes, there is a high probability of

187: getting enough inflation $N_e \gsim 60$ even with a relatively small

188: number of branes, of order 10.  This is shown in figure \ref{fig3},

189: which is a close-up of the small-$N$ region of figure \ref{fig2}.

190: Nevertheless, brane stacks of up to 100 or even 1000 could be used

191: to satisfy the tadpole conditions of the flux compactification

192: scenario, given a Calabi-Yau manifold with large enough Euler

193: characteristic \cite{Shamit}.  Examples with $\chi = 10^5$ are known.

194: 

195: \begin{figure}

196: \centerline{\includegraphics[width=0.7\hsize,angle=0]{Nezoom.eps}}  

197: \caption{Close-up of small-$N$ region of figure \ref{fig2}.}

198: \label{fig3}

199: \end{figure}

200: 

201: I come back to the fine-tuning issue.  If we had no restriction

202: on the combination $b\equiv\sqrt{2\sigma\epsilon}/\psi_0^3$, this

203: mechanism would be a complete solution to the problem of fine-tuning.

204: The fact that $b$ must lie in a narrow range makes it less compelling.

205: However, we have found that by adding corrections to the potential 

206: which are generically expected to be present, the restriction on 

207: $b$ can be removed altogether.  These take the form of corrections

208: to the superpotential and K\"ahler potential.   The potential 

209: generalizes to 

210: \beq

211: 	V \to 

212: {N\delta(\psi^2-\psi_0^2)^2  - {N \epsilon\, \tau\over  

213: 1+ N{\epsilon (\psi-\psi_0)^{-4}}} \over 

214: (2\sigma - N(\psi^2+d\psi^4))^2}

215: \eeq

216: Then using, for example,

217: $d = -0.5\psi_0^{-2}$ and $\delta = 0.1\epsilon\tau\psi_0^{-4}$,

218: we get the required self-flattening behavior of the potential for

219: any value of $b$.

220: 

221: The question of whether this theory of inflation can be distinguished

222: from others by the data is difficult.  In our scan of parameter

223: space, we find a wide range of predictions for properties of the

224: scalar and tensor perturbation spectra relevant for the Cosmic

225: Microwave Background.  For example, applying the COBE normalization

226: to the spectrum implies that a range of energy scales

227: for the potential is allowed, $V^{1/4} < 10^{-3} M_p$, with the

228: inequality being saturated only for the rare cases with $N\sim 1$

229: branes.  For this reason, a measurable tensor component in the CMB is

230: not a generic prediction of the model.  Moreover we find a wide range

231: of spectral tilts,  $0.93 < n_s < 1.15$, although the majority of

232: realizations has $n_s < 1$.  Thus there is no smoking-gun prediction

233: for the spectral index in this model.  

234: 

235: However, there is an interesting test which could allow future

236: CMB experiments to rule out the models under consideration.  There

237: is presently a hint in the WMAP data of strong running of the spectral

238: index, $dn_s/d\ln k \sim -0.1$.  Our models were not able to produce

239: such a large amount of running, and the exact value awaits

240: experimental confirmation because the errors are still large.  But

241: if upcoming measurements were able to confirm even a much smaller

242: level of running,  $dn_s/d\ln k \sim -0.01$, it would effectively 

243: rule out the model because of the very definite correlation between

244: $dn_s/d\ln k$ and  the spectral index itself, $n_s$, as shown in

245: figure \ref{fig4}.  The predicted points fall along the solid line.

246: Although many of these points lie within the experimentally allowed

247: region \cite{Seljak}, those with significant running do not.  

248: 

249: \begin{figure}

250: \centerline{\includegraphics[width=0.7\hsize,angle=0]{comp.eps}}  

251: \caption{Running of spectral index versus spectral index in multibrane

252: model.}

253: \label{fig4}

254: \end{figure}

255: 

256: In summary, we have presented a novel mechanism for solving the

257: fine-tuning problem of obtaining a flat inflaton potential, within

258: a popular string theoretic model of inflation.  The self-flattening

259: feature of the brane-antibrane potential does not naturally occur in

260: field theory models of inflation, but it appears quite generically

261: in the present stringy context.  It is also encouraging that at least

262: one experimental test can potentially falsify the model.  

263: 

264: 

265: \begin{thebibliography}{9}

266: \bibitem{Cline}

267:   J.~M.~Cline,

268:   ``Inflation from string theory,''

269:   arXiv:hep-th/0501179.

270:   %%CITATION = HEP-TH 0501179;%%

271: 

272: 

273: \bibitem{CS}

274:   J.~M.~Cline and H.~Stoica,

275:   ``Multibrane inflation and dynamical flattening of the inflaton potential,''

276:   arXiv:hep-th/0508029.

277:   %%CITATION = HEP-TH 0508029;%%

278: 

279: \bibitem{KKLMMT}

280:   S.~Kachru, R.~Kallosh, A.~Linde and S.~P.~Trivedi,

281:   ``De Sitter vacua in string theory,''

282:   Phys.\ Rev.\ D {\bf 68}, 046005 (2003)

283:   [arXiv:hep-th/0301240].

284:   %%CITATION = HEP-TH 0301240;%%

285: 

286: \bibitem{KS}

287:   I.~R.~Klebanov and M.~J.~Strassler,

288:   ``Supergravity and a confining gauge theory: Duality cascades and

289:   chiSB-resolution of naked singularities,''

290:   JHEP {\bf 0008}, 052 (2000)

291:   [arXiv:hep-th/0007191].

292:   %%CITATION = HEP-TH 0007191;%%

293: 

294: 

295: 

296: \bibitem{BCSQ}

297:   C.~P.~Burgess, J.~M.~Cline, H.~Stoica and F.~Quevedo,

298:   ``Inflation in realistic D-brane models,''

299:   JHEP {\bf 0409}, 033 (2004)

300:   [arXiv:hep-th/0403119].

301:   %%CITATION = HEP-TH 0403119;%%

302: 

303: \bibitem{Shamit}

304: S.\ Kacrhu, private communication

305: 

306: \bibitem{Seljak}

307:   U.~Seljak {\it et al.},

308:   ``Cosmological parameter analysis including SDSS Ly-alpha forest and  galaxy

309:   bias: Constraints on the primordial spectrum of fluctuations,  neutrino mass,

310:   and dark energy,''

311:   Phys.\ Rev.\ D {\bf 71}, 103515 (2005)

312:   [arXiv:astro-ph/0407372].

313:   %%CITATION = ASTRO-PH 0407372;%%

314: 

315: 

316: \end{thebibliography}

317: \end{document}

318: 

319: