1: \documentclass[12pt] {article}

2: \usepackage{latexsym}

3: \usepackage{amsmath}

4: \usepackage{mathrsfs}

5: \usepackage{amsfonts}

6: \usepackage{graphicx}

7: \usepackage{epsfig}

8: 

9: 

10: \usepackage[totalheight = 23cm, totalwidth = 17cm]{geometry}

11: %\linespread{1.3}

12: 

13: \newcommand{\ud}{\mathrm{d}}

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

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

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

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

18: %\numberwithin{equation}{section}

19: \def\tr{\text{tr}\hspace{1pt}}

20: \def\Tr{\text{Tr}\hspace{1pt}}

21: \def\hsp{\hspace{10pt}}

22: \def\non{\nonumber}

23: \def\jac{\vartheta}

24: 

25: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

26: %NEW COMMANDS%%%%%%%%%%%%%%%%%%%%%%%%

27: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

28: 

29: \newcommand{\ba}{\begin{align}}

30: \newcommand{\ea}{\end{align}}

31: 

32: \newcommand{\A}[1]{A^{(#1)}}

33: \newcommand{\B}[1]{B^{(#1)}}

34: \newcommand{\C}[1]{C^{(#1)}}

35: \newcommand{\G}[1]{G^{(#1)}}

36: \newcommand{\CHI}[1]{\chi^{(#1)}}

37: \newcommand{\GAM}[1]{\Gamma^{(#1)}}

38: \newcommand{\Ge}{\Gamma_{11}}

39: \newcommand{\LAM}[1]{\Lambda^{(#1)}}

40: \newcommand{\PSI}[1]{\Psi^{(#1)}}

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

42: \newcommand{\Int}{\mathop{\rm Int}\nolimits}

43: \newcommand{\nuv}{\underline{\phantom{\mu }\!\!\!\nu}}

44: \newcommand{\wdg}{{\scriptscriptstyle \wedge}}

45: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

46: 

47: 

48: 

49: \begin{document}

50: \begin{titlepage}  

51: %\pagestyle{empty}

52: %\baselineskip=21pt

53: %\vspace{0.01cm}

54: \rightline{\tt gr-qc/0508105}

55: \rightline{CERN-PH-TH/2005-154, ACT-06-05, MIFP-05-19}

56: \vspace{0.01cm}

57: \begin{center}

58: %{\bf {\large Q-Cosmology:}}\\

59: %\vspace{0.03in}

60: {\bf {\large Liouville Cosmology at Zero and Finite Temperatures}}

61: \end{center}

62: \begin{center}

63: \vskip 0.03in

64: {\bf John~Ellis}$^1$,~ {\bf Nikolaos E. Mavromatos}$^{2}$,~ 

65: {\bf Dimitri V. Nanopoulos}$^{3}$ \\ and \\ {\bf Michael Westmuckett}$^{2}$

66: \vskip 0.5cm

67: {\it

68: $^1${TH Division, Physics Department, CERN, CH-1211 Geneva 23, Switzerland}\\

69: $^2${Theoretical Physics, Physics Department, 

70: King's College London, Strand WC2R 2LS, UK}\\

71: $^3${George P. and Cynthia W. Mitchell Institute for Fundamental 

72: Physics, \\ Texas A\&M

73: University, College Station, TX 77843, USA; \\

74: Astroparticle Physics Group, Houston

75: Advanced Research Center (HARC),

76: Mitchell Campus,

77: Woodlands, TX~77381, USA; \\

78: Academy of Athens,

79: Division of Natural Sciences, 28~Panepistimiou Avenue, Athens 10679,

80: Greece}\\

81: } 

82: 

83: \vspace{0.5cm}

84: {\bf Abstract}

85: \end{center}

86: 

87: We discuss cosmology in the context of Liouville strings, characterized by

88: a central-charge deficit $Q^2$, in which target time is identified with

89: (the world-sheet zero mode of the) Liouville field: {\it Q-Cosmology}.  

90: We use a specific example of colliding brane worlds to illustrate the

91: phase diagram of this cosmological framework. The collision provides the

92: necessary initial cosmological instability, expressed as a departure from

93: conformal invariance in the underlying string model. The brane motion

94: provides a way of breaking target-space supersymmetry, and leads to

95: various phases of the brane and bulk Universes.  Specifically, we find a

96: hot metastable phase for the bulk string Universe soon after the brane

97: collision in which supersymmetry is broken, which we describe by means of

98: a subcritical world-sheet $\sigma$ model dressed by a space-like

99: Liouville field, representing finite temperature (Euclidean time). This

100: phase is followed by an inflationary phase for the brane Universe, in

101: which the bulk string excitations are cold. This is described by a

102: super-critical Liouville string with a time-like Liouville mode, whose

103: zero mode is identified with the Minkowski target time. Finally, we

104: speculate on possible ways of exiting the inflationary phase, either by

105: means of subsequent collisions or by deceleration of the brane Universe

106: due to closed-string radiation from the brane to the bulk. While phase

107: transitions from hot to cold configurations occur in the bulk string

108: universe, stringy excitations attached to the brane world remain

109: thermalized throughout, at a temperature which can be relatively high. 

110: %Its maximum value may be constrained by the usual requirement of avoiding

111: %gravitino overproduction. 

112: The late-time behaviour

113: of the model results in dilaton-dominated dark energy and 

114: present-day acceleration of the expansion of the Universe, asymptoting 

115: eventually to zero.

116: %but also past deceleration for redshifts 

117: %$z > z^*$ with $z^* < 1$.                . 

118: 

119: \vspace{0.5cm}

120: \leftline{CERN-PH-TH/2005-154}

121: \leftline{August 2005}

122: \end{titlepage}

123: 

124: 

125: 

126: \section{Introduction} 

127: 

128: Formal developments in string theory~\cite{gsw} over the 

129: past decade, in particular the discovery of a consistent

130: way of studying quantum  domain wall structures (D-branes)~\cite{polchinski},

131: have opened up

132: novel ways of looking at both the microcosmos and the 

133: macrocosmos.

134: In the {\it microcosmos}, there are novel ways of compactification,

135: either via the observation~\cite{add} that extra dimensions

136: that are large compared to the string scale may be consistent with the 

137: foundations of string theory, or by 

138: viewing our four-dimensional world as a brane embedded in 

139: a bulk space-time, allowing for large extra dimensions

140: that might even be infinite in size~\cite{randal},

141: in a manner consistent with a large hierarchy between the Planck scale 

142: and the electroweak or supersymmetry 

143: breaking scale. In this modern approach, 

144: fields in the gravitational (super)multiplet of the (super)string,

145: or more generally those neutral under the Standard Model (SM) group,

146: are allowed to 

147: propagate in the bulk. This is not the case for non-Abelian gauge fields,

148: nor fields charged under the SM group, 

149: which are attached

150: to the brane world. 

151: In this approach, the weakness of  

152: gravity compared to the rest of the interactions is a result of 

153: the large compact dimensions, the compactification 

154: not necessarily being achieved

155: through conventional means, i.e., closing up the 

156: extra dimensions in spatial compact manifolds, 

157: but perhaps also through the involvement of shadow brane worlds 

158: with special reflecting properties, such as orientifolds,

159: which restrict the bulk dimension~\cite{ibanez}.

160: In such approaches the string scale $M_s$ is not necessarily 

161: identical to the four-dimensional

162: Planck mass scale $M_P$. Instead, the

163: two scales are related through the large compactification volume $V_6$:

164: \begin{equation}

165: M_P^2 = \frac{8M_s^2 V_6}{g_s^2}.

166: \label{planckstring}

167: \end{equation}

168: As for the {\it macrocosmos}, 

169: this modern approach has offered new insights 

170: into the cosmic evolution of our Universe. 

171: Novel ways of discussing cosmology

172: in brane worlds have been discovered over the past few years, 

173: which may revolutionize our way of approaching

174: issues such as inflation~\cite{langlois,ekpyrotic} and the present 

175: acceleration of the expansion of the Universe~\cite{emn04}. 

176: 

177: In parallel, mounting experimental evidence from diverse

178: astrophysical sources presents some important puzzles that string theory

179: must address if it is to provide a realistic description of Nature.

180: Observations of distant Type-1a supernovae~\cite{snIa}, as well as 

181: detailed

182: studies of the cosmic microwave background fluctuations by the WMAP

183: satellite~\cite{wmap}, indicate that our Universe is currently in an

184: accelerating epoch, and that 73\% of its energy density consists of dark

185: energy that does not cluster, but is present in `empty space'.

186: These issues are highly significant for string theory, motivating a novel

187: perspective on the treatment and understanding of string dynamics.  If the

188: dark energy turns out to be a true cosmological constant, leading to an

189: asymptotic de Sitter horizon, then the entire concept of the scattering

190: S-matrix, upon which perturbative string theory is built, breaks

191: down~\footnote{This is also true in more general scenarios for the vacuum

192: energy with a de Sitter horizon.}. This would cast doubt on the foundations of

193: string theory, at least as they are conventionally 

194: formulated. On the

195: other hand, even if models for relaxing the vacuum energy are invoked,

196: leading asymptotically to a vanishing vacuum energy density at large

197: cosmic times and consistency with an S-matrix, there is still the open

198: issue of embedding such models in (perturbative) string theory. In

199: particular, the formal question arises how to formulate consistently a

200: world-sheet $\sigma$-model description of strings propagating in such

201: time-dependent space-time backgrounds.

202: 

203: 

204: The standard world-sheet conformal invariance conditions of critical

205: string theory~\cite{gsw}, which are equivalent to target-space equations

206: of motion for the background fields on which the string propagates, are

207: very restrictive, allowing only vacuum solutions of critical strings. The

208: main problem may be illustrated as follows.  Consider the graviton

209: world-sheet $\beta$ function, which is simply the Ricci tensor of the

210: target space-time background to lowest order in $\alpha '$:

211: \begin{equation} 

212: \beta_{\mu\nu} = \alpha ' R_{\mu\nu}~,

213: \label{beta} \end{equation}

214: in the absence of other fields. Conformal invariance would require that

215: $\beta_{\mu\nu}=0$, implying a Ricci-flat background,

216: which is a solution to the {\it vacuum} Einstein equations. {\it A

217: priori}, a cosmological-constant vacuum solution is inconsistent with this

218: conformal invariance in strings, since it has a Ricci tensor $R_{\mu\nu} =

219: \Lambda g_{\mu\nu}$, where $g_{\mu\nu}$ is the metric tensor. The question

220: then arises how to describe cosmological backgrounds for strings that are

221: not vacuum solutions, but require the presence of a matter fluid yielding

222: a non-flat Ricci tensor.

223: 

224: One proposal for obtaining a non-zero cosmological constant in string

225: theory was made in~\cite{fischler}, according to which dilaton tadpoles on

226: higher-genus world-sheet surfaces produce additional modular infinities,

227: whose regularization leads to extra world-sheet structures in the $\sigma$

228: model. Since they do not appear at the world-sheet tree level, they lead

229: to modifications of the $\beta$ function such that the Ricci tensor of the

230: space-time background is now that of an (anti) de Sitter Universe, with a

231: cosmological constant fixed by the dilaton tadpole graph: $J_D >0$ ($J_D <

232: 0$). The problem with this approach is the above-mentioned existence of an

233: asymptotic horizon in the de Sitter case, which prevents the proper

234: definition of asymptotic states, and hence a scattering matrix.  Since the

235: perturbative world-sheet formalism is based on such an S-matrix, there is

236: \emph{a priori} an inconsistency in the approach.

237: 

238: It was proposed in~\cite{aben} that one way out of this dilemma would be

239: to assume a time-dependent dilaton background, with a linear

240: dependence on time in the so-called $\sigma$-model frame. Such

241: backgrounds, even when the $\sigma$-model metric is flat, lead to exact

242: solutions (to all orders in $\alpha '$) of the conformal invariance

243: conditions of the pertinent stringy $\sigma$-model, and so are acceptable

244: solutions from a perturbative viewpoint. It was argued in~\cite{aben} that

245: such backgrounds describe linearly-expanding Robertson-Walker Universes,

246: which were shown to be exact conformal-invariant solutions, corresponding

247: to Wess-Zumino models on appropriate group manifolds.

248: 

249: The pertinent $\sigma$-model action in a background with graviton $G$,

250: antisymmetric tensor $B$ and dilaton $\Phi$ reads~\cite{gsw}:

251: \begin{equation} S_\sigma = \frac{1}{4\pi\alpha '} \int_\Sigma d^2\xi

252: [\sqrt{-\gamma} G_{\mu\nu} \partial_\alpha X^\mu \partial^\alpha X^\nu +

253: i\epsilon^{\alpha\beta} B_{\mu\nu} \partial_\alpha X^\mu \partial_\beta

254: X^\nu + \alpha '\sqrt{-\gamma}R^{(2)}\Phi], \label{sigmamodel}

255: \end{equation} where $\Sigma$ denotes the world-sheet, with metric

256: $\gamma$ and the topology of a sphere, $\alpha$ are world-sheet indices,

257: and $\mu,\nu $ are target-space-time indices. The important point

258: of~\cite{aben} was the role of target time $t$ as a specific dilaton

259: background, linear in that coordinate, of the form

260: \begin{equation}

261: \label{lineardil}

262: \Phi = {\rm const} - \frac{1}{2}Q~t,

263: \end{equation}

264: where $Q$ is a constant and $Q^2 > 0$ is the $\sigma$-model central-charge

265: deficit, allowing this {\it supercritical} string theory to be formulated

266: in some number of dimensions different from the critical number.

267: Consistency of the underlying world-sheet conformal field

268: theory, as well as modular invariance of the string scattering amplitudes,

269: required {\it discrete} values of $Q^2$, when expressed in units of the

270: string length $M_s$~\cite{aben}.

271: This was the first example of a non-critical string cosmology, with the

272: spatial target-space coordinates $X^i$, $i=1, \dots D-1$, playing the

273: r\^oles of $\sigma$-model fields. This non-critical string was not

274: conformally invariant, and hence required Liouville dressing~\cite{ddk}.

275: The Liouville field had time-like signature in target space, since the

276: central charge deficit $Q^2 > 0$ in the model

277: of~\cite{aben}, and its zero mode played the r\^ole of target time.

278: 

279: As a result of the non-trivial dilaton field, the 

280: Einstein term in the effective $D$-dimensional low-energy

281: field theory action is conformally rescaled by $e^{-2\Phi}$.

282: This requires a redefinition 

283: of the $\sigma$-model-frame space-time metric $g_{\mu\nu}^\sigma$ to

284: the `physical' Einstein metric  $g_{\mu\nu}^E$:

285: \begin{equation}

286: g_{\mu\nu}^E = e^{-\frac{4\Phi}{D-2}}G_{\mu\nu}~.

287: \label{smodeinst}

288: \end{equation}

289: Target time must also be rescaled, so that the

290: metric acquires the standard Robertson-Walker (RW) form in the normalized

291: Einstein frame for the effective action:

292: \begin{equation}

293: ds^2_E = -dt_E^2 + a_E^2(t_E) \left(dr^2 + r^2 d\Omega^2 \right),

294: \end{equation}

295: where we show the example of a spatially-flat RW metric for definiteness,

296: and $a_E(t_E)$ is an appropriate scale factor, which is a function of $t_E$

297: alone in the homogeneous cosmological backgrounds we assume throughout.

298: 

299: The Einstein-frame time is related to the 

300: time in the $\sigma$-model frame~\cite{aben} by:

301: \begin{equation}\label{einsttime}

302: dt_E = e^{-2\Phi/(D-2)}dt \qquad \to \qquad t_E = \int ^t e^{-2\Phi(t')/(D-2)}

303: dt'~. 

304: \end{equation} 

305: The linear dilaton background (\ref{lineardil}) yields

306: the following relation between the Einstein and $\sigma$-model frame 

307: times:

308: \begin{equation} 

309: t_E = c_1 + \frac{D-2}{Q}e^{\frac{Q}{D-2}t},

310: \end{equation}

311: where $c_{1,0}$ are appropriate (positive) constants.

312: Thus, a dilaton background (\ref{lineardil}) that is

313: linear in the $\sigma$-model time scales logarithmically with 

314: the Einstein time (Robertson-Walker cosmic time) $t_E$:

315: \begin{equation}\label{dil2}

316: \Phi (t_E) =({\rm const.}') - \frac{D-2}{2}{\rm ln}(\frac{Q}{D-2}t_E).

317: \end{equation} 

318: In this regime, the string coupling~\cite{gsw}: 

319: \begin{equation}

320: g_s = {\rm exp}\left(\Phi(t)\right)

321: \label{defstringcoupl}

322: \end{equation}

323: varies with the cosmic time $t_E$ as 

324: $g_s^2 (t_E) \equiv e^{2\Phi} \propto \frac{1}{t_E^{D-2}}$,

325: thereby implying a vanishing effective string coupling 

326: asymptotically in cosmic time. 

327: In the linear dilaton background of~\cite{aben}, the asymptotic

328: space-time metric in the Einstein frame reads:

329: \begin{equation} \label{metricaben}

330: ds^2 = -dt_E^2 + a_0^2 t_E^2 \left(dr^2 + r^2 d\Omega^2 \right)

331: \end{equation}

332: where $a_0$ a constant.

333: Clearly, there is no acceleration in the expansion of the Universe 

334: (\ref{metricaben}).

335: 

336: The effective low-energy action on the four-dimensional 

337: brane world for the gravitational multiplet 

338: of the string in the Einstein frame reads~\cite{aben}:

339: \begin{equation}

340: S_{\rm eff}^{\rm brane} = \int d^4x\sqrt{-g}\{ R  - 2(\partial_\mu \Phi)^2 

341: - \frac{1}{2} e^{4\Phi}( \partial_\mu b)^2 - \frac{2}{3}e^{2\Phi}\delta c 

342: \},

343: \label{effaction}

344: \end{equation}

345: where, as we discuss below, $b$ is the four-dimensional axion field

346: associated with a four-dimensional representation of the antisymmetric

347: tensor, and $\delta c = C_{\rm int} - c^*$,

348: where $C_{\rm int}$ is the central charge of the conformal world-sheet

349: theory corresponding to the transverse (internal) string dimensions, and

350: $c^*=22 (6)$ is the critical value of this internal central charge of the

351: (super)string theory for flat four-dimensional space-times.  The linear

352: dilaton configuration (\ref{lineardil}) corresponds, in this language, to

353: a background charge $Q$ of the conformal theory, which contributes a term

354: $-3Q^2$ (in our normalization)  to the total central charge. The latter

355: includes the contributions from the four uncompactified dimensions of our

356: world.  In the case of a flat four-dimensional Minkowski space-time, one

357: has $C_{\rm total} = 4 -3Q^2 + C_{\rm int} = 4 - 3Q^2 + c^* + \delta c$,

358: which should equal 26 (10). This implies that $C_{\rm int} = 22 + 3Q^2~(6 

359: +

360: 3Q^2)$ for bosonic (supersymmetric) strings.  

361: 

362: An important result in~\cite{aben} was the discovery of an exact conformal

363: field theory corresponding to the dilaton background (\ref{dil2}) and a

364: constant-curvature (Milne) static metric in the $\sigma$-model frame (or,

365: equivalently, a linearly-expanding Robertson-Walker Universe in the

366: Einstein frame).  The conformal field theory corresponds to a

367: Wess-Zumino-Witten two-dimensional world-sheet model on a group manifold

368: $O(3)$ with appropriate constant curvature, whose coordinates correspond

369: to the spatial components of the four-dimensional metric and antisymmetric

370: tensor fields, together with a free world-sheet field corresponding to the

371: target time coordinate. The total central charge in this more general case

372: reads $C_{\rm total} = 4 - 3Q^2 - \frac{6}{k+2}+ C_{\rm int}$, where $k$

373: is a positive integer corresponding to the level of the Kac-Moody algebra

374: associated with the WZW model on the group manifold. The value of $Q$ is

375: chosen in such a way that the overall central charge $c=26$ and the theory

376: is conformally invariant. Since such unitary conformal field theories have

377: {\it discrete} values of their central charges, which accumulate to

378: integers or half-integers from {\it below}, it follows that the values of

379: the central charge deficit $\delta c$ are {\it discrete} and {\it finite}

380: in number.  From a physical point of view, this implies that the

381: linear-dilaton Universe may either stay in such a state for ever, for a

382: given $\delta c$, or tunnel between the various discrete levels before

383: relaxing to a critical $\delta c =0$ theory. It was argued in~\cite{aben}

384: that, due to the above-mentioned finiteness of the set of allowed discrete

385: values of the central charge deficit $\delta c$, the Universe could reach

386: flat four-dimensional Minkowski space-time, and thus exit from the

387: expanding phase, after a finite number of phase transitions.

388: 

389: The analysis in~\cite{aben} also showed, as we discuss below, that there

390: are tachyonic mass shifts of order $-Q^2$ in the bosonic string

391: excitations, but not in the fermionic ones. This implies the appearance of

392: tachyonic instabilities and the breaking of target-space supersymmetry in

393: such backgrounds, as far as the excitation spectrum is concerned. The

394: instabilities could trigger the cosmological phase transitions, since they

395: correspond to relevant renormalization-group world-sheet operators, and

396: hence initiate the flow of the internal unitary conformal field theory

397: towards minimization of its central charge, in accordance with the

398: Zamolodchikov $c$-theorem~\cite{zam}. As we discuss later on, in

399: semi-realistic cosmological models~\cite{dgmpp} such tachyons decouple

400: from the spectrum relatively quickly. On the other hand, as a result of

401: the form of the dilaton in the Einstein frame (\ref{dil2}), we observe

402: that the dark-energy density for this (four-dimensional)  Universe,

403: $\Lambda \equiv e^{2\Phi}\delta c$, is relaxing to zero with a

404: $1/t_E^{(D-2)}$ dependence on the Einstein-frame time for each of the

405: equilibrium values of $\delta c$.  Therefore, the breaking of

406: supersymmetry induced by the linear dilaton is only an

407: obstruction~\cite{witten}, rather than a spontaneous breaking, in the

408: sense that it appears only temporarily in the boson-fermion mass

409: splittings between the excitations, whilst the vacuum energy of the

410: asymptotic equilibrium theory vanishes.

411: 

412: In~\cite{emn} we went one step beyond the analysis in~\cite{aben}, and

413: considered more complicated $\sigma$-model metric backgrounds that did not

414: satisfy the $\sigma$-model conformal-invariance conditions, and therefore

415: needed Liouville dressing~\cite{ddk} to restore conformal invariance. Such

416: backgrounds could even be time-dependent, living in $(d+1)$-dimensional

417: target space-times. Various mathematically-consistent forms of

418: non-criticality can be considered, for instance cosmic catastrophes such

419: as the collision of brane worlds~\cite{gravanis,brany}. Such models lead

420: to supercriticality of the associated $\sigma$ models describing stringy

421: excitations on the brane worlds.  The Liouville dressing of such

422: non-critical models results in $(d+2)$-dimensional target spaces with two

423: time directions.  An important point in~\cite{emn} was the identification

424: of the (world-sheet zero mode of the) Liouville field with the target

425: time, thereby restricting the Liouville-dressed $\sigma$ model to a

426: $(d+1)$-dimensional hypersurface of the $(d+2)$-dimensional target space 

427: and maintaining the initial target space-time dimensionality. We

428: stress that this identification is possible only in cases where

429: the initial $\sigma$ model is supercritical, so that the Liouville mode

430: has time-like signature~\cite{aben,ddk}. In certain

431: models~\cite{gravanis,brany}, such an identification was proven to be

432: energetically preferable from a target-space viewpoint, since it minimized

433: certain effective potentials in the low-energy field theory corresponding

434: to the string theory at hand.

435: 

436: All such cosmologies require some initial physical reason for the initial

437: departure from the conformal invariance of the underlying $\sigma$ model

438: that describes string excitations in such Universes. The reason could be

439: an initial quantum fluctuation, or, in brane models, a catastrophic cosmic

440: event such as the collision of two or more brane worlds.  Such

441: non-critical $\sigma$ models relax asymptotically to conformal $\sigma$

442: models, which may be viewed as equilibrium points in string theory space,

443: as illustrated in Fig.~\ref{fig:flow}. In some interesting cases of

444: relevance to cosmology~\cite{dgmpp}, which are particularly generic, the

445: asymptotic conformal field theory is that of~\cite{aben} with a linear

446: dilaton and a flat Minkowski target-space metric in the $\sigma$-model

447: frame. In others, the asymptotic theory is characterized by a constant

448: dilaton and a Minkowskian space-time~\cite{gravanis}.

449: Since, as we discuss below, the evolution of the central-charge deficit of

450: such a non-critical $\sigma$ model, $Q^2(t)$, plays a crucial r\^ole in

451: inducing the various phases of the Universe, including an inflationary

452: phase, graceful exit from it, thermalization and a contemporary phase of

453: accelerating expansion, we term such Liouville-string-based cosmologies 

454: {\it Q-Cosmologies}.

455: 

456: \begin{figure}[tb]

457: \begin{center}

458: \includegraphics[width=7cm]{flow2.eps}

459: \end{center}

460: \caption{\it 

461: A schematic view of string theory space, which is an infinite-dimensional

462: manifold endowed with a (Zamolodchikov) metric. The dots denote conformal

463: string backgrounds. A non-conformal string flows (in a two-dimensional

464: renormalization-group sense)  from one fixed point to another, either of

465: which could be a hypersurface in theory space.  The direction of the flow

466: is irreversible, and is directed towards the fixed point with a lesser

467: value of the central charge, for unitary theories, or, for general

468: theories, towards minimization of the degrees of freedom of the system.}

469: \label{fig:flow}

470: \end{figure}

471: 

472: The use of Liouville strings to describe the evolution of our Universe has

473: a broad motivation, since non-critical strings are associated with

474: non-equilibrium situations, as are likely to have occurred in the early

475: Universe. The space of non-critical string theories is much larger than

476: that of critical strings. It is therefore remarkable that the departure

477: from criticality may enhance the predictability of string theory to the

478: extent that a purely stringy quantity such as the string coupling $g_s$

479: may become accessible to experiment via its relation to the present-era

480: cosmic acceleration parameter: $g_s^2 = -q^{0}$~\cite{emn04}.  Another

481: example arises in a non-critical string approach to inflation, if the Big

482: Bang is identified with the collision~\cite{ekpyrotic} of two

483: D-branes~\cite{brany}. In such a scenario, astrophysical observations may

484: place important bounds on the recoil velocity of the brane worlds after

485: the collision, and lead to an estimate of the separation of the branes at

486: the end of the inflationary period.

487: 

488: In such a framework, the identification of target time with a world-sheet

489: renormalization-group scale, the zero mode of the Liouville

490: field~\cite{emn}, provides a novel way of selecting the ground state of

491: the string theory. This is not necessarily associated with minimization of

492: energy, but could simply be a result of cosmic chance. It may be a random

493: global event that the initial state of our cosmos corresponds to a certain

494: Gaussian fixed point in the space of string theories, which is then

495: perturbed into a Big Bang by some relevant (in a world-sheet sense)

496: deformation, which makes the theory non-critical, and hence out of

497: equilibrium from a target space-time viewpoint. The theory then flows, as

498: indicated in Fig.~\ref{fig:flow}, along some specific

499: renormalization-group trajectory, heading asymptotically to some ground

500: state that is a local extremum

501: corresponding to an infrared fixed point of this perturbed world-sheet 

502: $\sigma$-model theory. This approach allows for many `parallel

503: universes' to be implemented, and our world might be just one of these.

504: Each Universe may flow between different fixed points, its trajectory

505: following a perturbation by a different operator.  It seems to us that

506: this scenario is more attractive and specific than the landscape

507: scenario~\cite{sussk}, which has recently been advocated as an framework

508: for parametrizing our ignorance of the true nature of string/M theory.

509: 

510: 

511: 

512: In this article we describe the main features of such non-critical string

513: cosmological models. The structure of the article is as follows:  in the

514: next Section we review briefly the basic properties of Liouville strings

515: at zero temperature, emphasizing the r\^ole of the Liouville mode as

516: target time~\cite{emn}.  We start in Section 2.1 with a comprehensive

517: description of the generic properties of Liouville dressing, and proceed

518: in Section 2.2 to present the basic features of (compactified)

519: cosmological models of non-critical strings upon which we rely later. We

520: give physical reasons for the departure from conformal invariance,in

521: Section 2.3 to discuss inflation in this framework, identifying the Hubble

522: parameter with the central-charge deficit $Q^2$ of the corresponding

523: supercritical $\sigma$ model describing string excitations in the

524: pertinent non-conformal cosmological backgrounds. In Section 2.4 we

525: discuss the late stages of such universes, in particular the r\^ole of the

526: (time-dependent) string coupling of the non-critical string in inducing

527: the present-day acceleration of the Universe. In Section 2.5 we discuss

528: the inclusion of matter in the late-stage evolution of the Universe, 

529: %and

530: %we argue in favour of an era of past deceleration for redshifts $z > 

531: %z^*$,

532: %for some $z^* < 1$. We also pay particular attention to 

533: demonstrating the

534: differences of our non-equilibrium Liouville formalism from standard

535: Friedmann-Robertson-Walker (FRW) Cosmologies. 

536: 

537: In Section 3 we present a concrete example of non-critical strings, that

538: of colliding brane worlds~\cite{gravanis,emw}, where the departure from

539: criticality results from the cosmic collision of the branes. Specific

540: scenarios of this type are discussed in Sections 3.1 and 3.2, relevant

541: aspects of Type-IIA supergravity are discussed in Section 3.3 and

542: compactification issues in Section 3.4. Then, in Section 3.5 we present

543: within this framework some scenarios for supersymmetry breaking at zero

544: temperature, associated with either the presence of moving branes or the

545: existence of magnetic fields in internal manifolds of the compactified

546: space of brane worlds. The latter is compatible with the present value of

547: the dark energy, as inferred from observations.  In these models the dark

548: energy is viewed as a relaxation energy of the brane world, which was

549: excited after the collision.

550: 

551: In Section 4 we present a finite-temperature analysis of the early

552: Universe in the context of Liouville strings. We commence our analysis in

553: Section 4.1 with a review of the hot phase soon after the brane collision,

554: and give estimates of the bulk and brane excitation energies. In Section

555: 4.2 we review the Liouville approach to finite-temperature heterotic

556: strings, whereby the temperature is associated with a space-like Liouville

557: mode in a subcritical string describing the thermal vacuum of the

558: heterotic string~\cite{kounnas}.  In Section 4.3 we discuss the

559: finite-temperature properties of the Type-IIA vacuum which characterises

560: the specific colliding-brane cosmological model mentioned above. In

561: Section 4.4 we discuss metastability properties of the Type-IIA thermal

562: supergravity vacuum, which, in contrast to the heterotic string case, is

563: an unstable vacuum, leading to the exit of our Universe from the hot phase

564: soon after the collision. In Section 4.5 we describe in some detail the

565: various phases of the colliding brane scenario and the associated phase

566: transitions, namely the transition from an initial hot phase to a cold

567: inflationary phase, and its subsequent exit from it.

568: We also pay attention to the fact that in these models the brane world

569: appears thermalized throughout, leading to decelerating brane motion in

570: the bulk, as a result of gravitational radiation leading towards thermal

571: equilibrium between the brane and bulk worlds. This deceleration is the

572: essential mechanism for the exit from the inflationary phase. We also

573: discuss in Section 4.5 several open questions associated with this phase,

574: in the context of our colliding-brane models, such as the possibility of a

575: second collision and some delicate issues concerning nucleosynthesis in

576: these models.

577: 

578: Finally, in Section 5 we present our conclusions and the outlook for

579: future work in Liouville Q-cosmologies.

580: 

581: 

582: \section{Non-Critical Liouville String Q-Cosmologies} 

583: 

584: \subsection{Zero-Temperature Liouville Formalism} 

585: 

586: We consider a $\sigma$-model action deformed by 

587: a family of 

588: vertex operators $V_i$, corresponding to 

589: `couplings' $g^i$, which represent \emph{non-conformal} background space-time

590: fields from the massless string multiplet, such as 

591: gravitons, $G_{\mu\nu}$, antisymmetric tensors, $B_{\mu\nu}$,

592: dilatons $\Phi$, their supersymmetric partners, \emph{etc.}:

593: \begin{equation}

594: S=S_{0}\left( X\right) +\sum_{i}g^{i}\int d^{2}z\,V_{i}\left( 

595: X \right)~,  

596: \label{action}

597: \end{equation}

598: where $S_0$ represents a conformal $\sigma$ model describing 

599: an equilibrium situation.

600: The non-conformality of the background means 

601: that the pertinent $\beta^i$ function $\beta^i \equiv dg^i/d{\rm ln}\mu 

602: \ne 0$, where $\mu$ is a world-sheet renormalization scale. 

603: Conformal invariance would imply restrictions on the 

604: background and couplings $g^i$, corresponding to 

605: the constraints $\beta^i = 0$, which are equivalent to equations

606: of motion derived from a target-space effective action for the corresponding

607: fields $g^i$.  The entire low-energy 

608: phenomenology and model building of critical string theory is based on such

609: restrictions~\cite{gsw}.  

610: 

611: 

612: In the non-conformal case $\beta^i \ne 0$, the theory is in need of

613: dressing by the Liouville field $\phi$ in order to restore conformal

614: symmetry~\cite{ddk}. The field $\phi $ acquires dynamics through the

615: integration over world-sheet covariant metrics in the path integral, and

616: may be viewed as a local dynamical scale on the world sheet~\cite{emn}.  

617: If the central charge of the (supersymmetric)  matter theory is $c_{m}>25

618: (9)$ (i.e., supercritical), the signature of the kinetic term of the

619: Liouville coordinate in the dressed $\sigma$-model is opposite to that of

620: the $\sigma$-model fields corresponding to the other target-space

621: coordinates.  As mentioned previously, this opens the way to the important

622: step of interpreting the Liouville field physically by identifying its

623: world-sheet zero mode $\phi _{0}$ with the target time in supercritical

624: theories~\cite{emn}.  Such an identification emerges naturally from the

625: dynamics of the target-space low-energy effective theory by minimizing the

626: effective potential~\cite{gravanis}.

627: 

628: The action of the Liouville mode $\phi$ reads~\cite{ddk}:

629: \begin{equation}

630: S_L = S_{0}\left( X\right) 

631: + \frac{1}{8\pi} \int_\Sigma d^2\xi \sqrt{\widehat \gamma} 

632: [ \pm (\partial \phi)^2 - QR^{(2)}\phi] + \int_\Sigma d^2\xi 

633: \sqrt{\widehat \gamma} g^i(\phi)V_i(X)~,

634: \label{liouvilleact} 

635: \end{equation} 

636: where ${\widehat \gamma}$ is a fiducial world-sheet metric, and 

637: the plus (minus) sign in front of the kinetic term of the 

638: Liouville mode pertains to subcritical (supercritical) strings. 

639: The dressed couplings $g^i(\phi)$  are obtained by the following procedure: 

640: \begin{equation}\label{dressing}

641: \int d^{2}z\,\,g_{i} \,V_{i}\left(X\right) \rightarrow \int

642: d^{2}z\,\,g_{i}(\phi)\,\,e^{\alpha _{i}\phi }\,V_{i}\left(X\right)~,

643: \end{equation}

644: where $\alpha_i$ is the 

645: ``gravitational'' anomalous dimension. If the original non-conformal 

646: vertex operator 

647: has anomalous scaling dimension $\Delta_i -2$ (for closed strings, 

648: to which we restrict

649: ourselves for definiteness), where $\Delta_i$ is the conformal dimension,

650: and the central charge surplus

651: of the theory is $Q^2 = \frac{c_m - c^*}{3} > 0 $ (for bosonic strings

652: $c^*=25$, for superstrings $c^*=9$), then the condition

653: that the dressed operator is marginal on the world sheet implies the

654: relation:

655: \begin{equation}

656: \alpha_i (\alpha_i + Q) = 2 - \Delta_i~. 

657: \end{equation}

658: Imposing appropriate boundary conditions in the limit 

659: $\phi \to \infty$~\cite{ddk}, the acceptable solution is:

660: \begin{equation}

661: \alpha_i = -\frac{Q}{2} + \sqrt{\frac{Q^2}{4} + 2 - \Delta_i}~.

662: \end{equation}

663: The gravitational dressing is trivial for marginal couplings, 

664: $\Delta_i=2$, as it should be. 

665: This dressing applies also to higher orders in the 

666: perturbative $g^i$ expansion. For instance,

667: at the next order, where the deviation from marginality 

668: in the deformations of the undressed $\sigma$model 

669: is due to the operator product expansion coefficients 

670: $c^i_{jk}$ in the $\beta^i$ function, the Liouville-dressing

671: procedure implies the replacement~\cite{schmid}:

672: \begin{equation}

673: g^i \qquad \to \qquad g^ie^{\alpha_i\phi} + \frac{\pi~\phi}{Q \pm 

674: 2\alpha_i}c^i_{jk}g^jg^ke^{\alpha_i \phi} ,

675: \label{dressing2}

676: \end{equation}

677: in order for the dressed operator to become marginal to this order

678: (the $\pm$ sign originates in (\ref{liouvilleact})). 

679: 

680: 

681: In terms of the Liouville renormalization-group scale, one

682: has the following equation relating Liouville-dressed

683: couplings $g^i$ and $\beta$ functions in the non-critical string case:

684: \begin{equation}

685: {\ddot  g}^i + Q{\dot g}i = \mp\beta^i(g_j)~,

686: \label{liouvilleeq}

687: \end{equation}

688: where the - (+) sign in front of the $\beta$-functions 

689: on the right-hand-side applies to super(sub)critical strings, 

690: the overdot denotes differentiation with respect to the 

691: Liouville zero mode, $\beta^i$ is the world-sheet renormalization-group

692: $\beta$ function (but with the renormalized couplings replaced 

693: by the Liouville-dressed ones as defined by the procedure in

694: (\ref{dressing}), (\ref{dressing2})), 

695: and the minus sign on the right-hand side (r.h.s.) of (\ref{liouvilleeq}) 

696: is due to the time-like signature of the Liouville field.

697: Formally, the $\beta^i$ of the r.h.s.\ of (\ref{liouvilleeq})

698: may be viewed as 

699: power series in the (weak) couplings $g^i$. 

700: The covariant (in theory space)

701: ${\cal G}_{ij}\beta^j$ function may be expanded as:

702: \begin{equation}

703: {\cal G}_{ij}\beta^j = 

704: \sum_{i_n} \langle V_i^L V_{i_1}^L \dots V_{i_n}^L \rangle_\phi g^{i_1} 

705: \dots g^{i_n}~,

706: \label{betaexp}

707: \end{equation}

708: where $V_i^L$ indicates Liouville dressing  \`a la (\ref{dressing}),

709: $ \langle \dots \rangle_\phi = \int d\phi d\vec{r}~{\rm exp}(-S(\phi, \vec{r}, g^j))$

710: denotes a functional average including Liouville integration, and 

711: $S(\phi, \vec{r}, g^i)$ is the Liouville-dressed $\sigma$-model

712: action, including the Liouville action~\cite{ddk}.

713: 

714: In the case of stringy $\sigma$ models, 

715: the diffeomorphism invariance of the target space results in the 

716: replacement of (\ref{liouvilleeq} by: 

717: \begin{equation} 

718:   {\ddot g}^i + Q(t){\dot g}^i = \mp{\tilde \beta}^i ,

719: \label{liouvilleeq2}

720: \end{equation} 

721: where the ${\tilde \beta}^i$ are the Weyl anomaly coefficients of the 

722: stringy $\sigma$ model in the background $\{ g^i \}$, which differ 

723: from the ordinary world-sheet renormalization-group $\beta^i$ functions

724: by terms of the form:

725: \begin{equation}

726: {\tilde \beta}^i = \beta^i + \delta g^i 

727: \end{equation}

728: where $\delta g^i$ denote transformations of the 

729: background field $g^i$ under infinitesimal general coordinate

730: transformations, e.g., for gravitons~\cite{gsw} 

731: ${\tilde \beta}^G_{\mu\nu} = 

732: \beta^G_{\mu\nu} + \nabla_{(\mu} W_{\nu)}$, 

733: with $W_\mu = \nabla _\mu \Phi$,

734: and $\beta^G_{\mu\nu} = R_{\mu\nu}$ to order $\alpha '$ 

735: (one $\sigma$-model loop). 

736:  

737: 

738: The set of equations (\ref{liouvilleeq}),(\ref{liouvilleeq2})  defines the

739: \emph{generalized conformal invariance conditions}, expressing the

740: restoration of conformal invariance by the Liouville mode.  The solution

741: of these equations, upon the identification of the Liouville zero mode

742: with the original target time, leads to constraints in the space-time

743: backgrounds~\cite{emn,gravanis}, in much the same way as the conformal

744: invariance conditions $\beta^i = 0$ define consistent space-time

745: backgrounds for critical strings~\cite{gsw}.  It is important to

746: remark~\cite{emninfl, emn} that the equations (\ref{liouvilleeq2}) can be

747: derived from an action.  This follows from general properties of the

748: Liouville renormalization group, which guarantee that the appropriate

749: Helmholtz conditions in the string-theory space $\{ g^i \}$ for the

750: Liouville-flow dynamics to be derivable from an action principle are

751: satisfied.

752: 

753: To be specific, consider the case of the $\sigma$ model (\ref{sigmamodel}),

754: and the ${\cal O}(\alpha ')$ Weyl anomaly coefficients~\cite{aben}, 

755: assuming that the $\sigma$-model target space is a $D$-dimensional 

756: space-time $(X^0,{\vec X})$. The pertinent $\beta$ functions are 

757:  \begin{eqnarray} 

758: && {\tilde \beta}^G_{\mu\nu} = \alpha ' \left( R_{\mu\nu} + 2 \nabla_{\mu} 

759: \partial_{\nu} \Phi 

760:  - \frac{1}{4}H_{\mu\rho\sigma}H_{\nu}^{\rho\sigma}\right)~, \nonumber \\

761: && {\tilde \beta}^B_{\mu\nu} = \alpha '\left(-\frac{1}{2}\nabla_{\rho} H^{\rho}_{\mu\nu} + 

762: H^{\rho}_{\mu\nu}\partial_{\rho} \Phi \right)~, \nonumber \\

763: && {\tilde \beta}^\Phi = \beta^\Phi - \frac{1}{4}G^{\rho\sigma}\beta^G_{\rho\sigma} = 

764: \frac{1}{6}\left( C^{(D)} - 26 \right)~, \\

765: && C^{(D)} = D -\frac{3}{2}\alpha ' [R - \frac{1}{12}H^2 - 4(\nabla \Phi)^2 + 4\Box \Phi ]~,

766: \label{bfunct}

767: \end{eqnarray} 

768: where $\alpha '$ is the Regge slope~\cite{gsw},

769: the Greek indices are D-dimensional,

770: and $H_{\mu\nu\rho}= \partial_{[\mu}B_{\nu\rho]}$ is the

771: field strength of the $B$ field.

772: 

773: Dressing this $\sigma$ model with a Liouville mode  

774: results in the 

775: appropriate equations (\ref{liouvilleeq2}), and it is straightforward

776: to show that these 

777: can be \emph{derived} from the following $(D+1)$-dimensional 

778: action in the $\sigma$-model (string) frame~\cite{schmid}: 

779: \begin{eqnarray}

780: && I^{(D+1)} = \int d\phi d^DX \sqrt{G}\sqrt{|G_{\phi\phi}|} 

781: e^{-2\Phi}\{ C^{(D)} - 25 + \nonumber \\ 

782: && 3G^{\phi\phi}[({\dot \Phi} - 

783: \frac{1}{2}G^{\mu\nu}{\dot G}_{\mu\nu})^2  - \frac{1}{4}

784: G^{\mu\nu}G^{\rho\lambda}\left({\dot G}_{\mu\rho}{\dot G}_{\nu\lambda}

785: + {\dot B}_{\mu\rho}{\dot B}_{\nu\lambda}\right)]\}~,

786: \label{dplus1action} 

787: \end{eqnarray} 

788: where $G,B,\Phi$ are all $D$-dimensional fields, depending

789: in general on $X^0,\phi,{\vec X}$, 

790: and the overdot denotes 

791: differentiation with respect to the Liouville zero mode $\phi$.

792: In the Liouville dressing procedure~\cite{ddk} 

793: we employ in this work one has the 

794: normalization $G_{\phi\phi} = -1$.  This justifies the 

795: presence of only spatial components of the metric and antisymmetric tensor 

796: fields in the terms inside the $[ \dots ]$ in (\ref{dplus1action}). 

797: This action may be schematically represented in the form~\cite{emninfl}:

798: \begin{equation}\label{lambdaaction}  

799: I^{(D+1)} = \int d\phi d^DX e^{-\varphi} \{ C^{(D)}(X) - 25 -3[{\dot \varphi}^2 

800: -\frac{1}{4}\left({\dot \lambda}^I {\cal G}_{IJ} {\dot \lambda}^J\right)]\}~,

801: \end{equation} 

802: where 

803: $\lambda_I = \{ G, B\}$, ${\cal G}_{IJ} =

804: G^{\mu\rho}G^{\nu\lambda}$ is a Zamolodchikov metric

805: in $\lambda_I$ space~\footnote{Note that this is compatible with the

806: definition of this metric in string space as a two-point correlation

807: function of appropriate vertex operators, as explicit ${\cal O}(\alpha ')$ 

808: computations have demonstrated~\cite{mm}.}, 

809: and $\varphi \equiv 2\Phi -{\rm ln}\sqrt{G}$ is a rescaled

810: dilaton~\cite{schmid}, which guarantees the diagonalization of the appropriate 

811: Zamolodchikov metric in the string theory space $(\Phi, \lambda_I)$,

812: with ${\cal G}_{\varphi\varphi} = 1$. 

813: 

814: 

815: 

816: Upon the identification of $\phi$ with the target time $X^0$, the 

817: $(D+1)$-dimensional action is constrained onto a $D$-dimensional

818: hypersurface $(X^0=\phi,~{\vec X})$. In that case the 

819: resulting $D$-dimensional target-space-time action 

820: reads:

821: \begin{eqnarray} 

822: && I^{D} = -\frac{3}{2}\alpha '\int d^DX \sqrt{G} e^{-2\Phi}\left(R +  4(\nabla \Phi)^2 

823: -\frac{1}{12}H^2 

824: - \frac{2}{3\alpha '}(D - 25)\right)  + {\cal I}_\phi~, 

825: \nonumber \\

826: && {\cal I}_\phi \equiv  

827: \int d^DX \sqrt{G} e^{-2\Phi}\left(

828: -3[({\dot \Phi} - \frac{1}{2}G^{\mu\nu}{\dot G}_{\mu\nu})^2 - \frac{1}{4}

829: G^{\mu\nu}G^{\rho\lambda}\left({\dot G}_{\mu\rho}{\dot G}_{\nu\lambda}

830: + {\dot B}_{\mu\rho}{\dot B}_{\nu\lambda}\right)]\}\right).

831: \label{ddimaction}

832: \end{eqnarray}

833: The extra piece, ${\cal I}_\phi$, 

834: as compared with a standard string theory target-space effective action, 

835: describes the non-equilibrium effects associated with the 

836: Liouville flow. 

837: In general, one may have critical target-space dimensionality $D=25$, but 

838: deviations from conformal invariance, due for example

839: to the recoil of brane worlds during collisions or other

840: catastrophic cosmic events, will be the topic 

841: of interest to us here. 

842: The non-critical central-charge term proportional 

843: to $\int d^DX\sqrt{G}e^{-2\Phi}(D-25)$ in the action (\ref{ddimaction}) 

844: may then be written in the form

845: \begin{equation} 

846: \frac{3\alpha '}{2}\int d^DX \sqrt{G} e^{-2\Phi} \frac{2}{\alpha '}Q^2 

847:  \label{centralcharge}

848: \end{equation}

849: in the normalization of the Einstein term appearing in (\ref{ddimaction}), 

850: where $Q^2$ represents the central charge deficit (whatever its origin) 

851: of the appropriate

852: $\sigma$ model: $Q^2 = (C- c^*)/3$, 

853: describing closed-string excitations in an

854: appropriate non-conformal background. 

855: 

856: 

857: \subsection{Cosmological Liouville Models: Generic Features} 

858: 

859: There are many cosmological models that fall in the above category

860: of Liouville strings, notably models  

861: with catastrophic cosmic events such as the collision of

862: two brane worlds~\cite{gravanis,emw}, which we

863: concentrate upon later in this work.

864: When one considers 

865: a brane moving in a bulk space in the presence of other brane

866: worlds~\cite{emw}, there are in general non-trivial 

867: potentials between the moving branes: only the static configuration

868: and certain other special configurations are supersymmetric in target 

869: space, with vanishing ground-state energy~\cite{emw}.

870: When one considers string theory excitations in such Universes,

871: the corresponding $\sigma$ model is \emph{non-critical}. 

872: In the model of \cite{emw}, as we discuss later, 

873: there are various phases for the bulk string Universe, 

874: which involve a passage from subcritical to supercritical strings,

875: due to a change in sign of the pertinent supersymmetry-breaking 

876: potential of the moving brane world.

877: 

878: In such situations, the resulting $\sigma$ model describing low-energy

879: string excitations in the bulk lives in a $(d+1)$-dimensional space-time,

880: $d$ denoting the number of spatial dimensions, and is not conformal. As

881: already discussed, to restore the conformal symmetry required for

882: consistency of the path-integral quantization, one needs Liouville

883: dressing~\cite{ddk}, resulting in equations of the form

884: (\ref{liouvilleeq2}) for the background fields under consideration.

885: Liouville dressing results in a critical string in $(d+2)$ dimensions,

886: with restored conformal symmetry expressed by the vanishing of the

887: $(d+2)$-target-dimensional $\sigma$-model $\beta$ functions. Eventually,

888: dynamics which we review in due course results in the

889: identification of the Liouville mode in supercritical situations with the

890: target time, so the final target-space dimensionality of the dressed

891: $\sigma$ model remains $(d+1)$.

892: 

893: 

894: 

895: One such model was considered in detail in \cite{dgmpp}. The

896: model is based on a specific string theory, namely ten-dimensional

897: Type-0~\cite{type0}, which leads to a non-supersymmetric target-space

898: spectrum as a result of a special projection of the supersymmetric

899: partners out of the spectrum. Nevertheless, the basic properties of its

900: cosmology are sufficiently generic to be extended to the bosonic sector of

901: any other effective low-energy string-inspired supersymmetric field

902: theory. The model also involves flux compactification to four dimensions,

903: which, as was pointed out in \cite{dgmpp}, plays an important r\^ole in

904: ensuring the existence of large stable bulk dimensions.

905: 

906: The ten-dimensional metric configuration considered in~\cite{dgmpp} 

907: was: 

908: \begin{equation}

909: G_{MN}=\left(\begin{array}{ccc}g^{(4)}_{\mu\nu} \qquad 0 \qquad 0 \\

910: 0 \qquad e^{2\sigma_1} \qquad 0 \\ 0 \qquad 0 \qquad

911: e^{2\sigma_2} I_{5\times 5} \end{array}\right)

912: \label{metriccomp}

913: \end{equation}

914: where lower-case Greek indices are four-dimensional space-time

915: indices, and $I_{5\times 5}$ denotes the $5\times 5$ unit matrix.

916: We have chosen two different scales for internal space. The field

917: $\sigma_{1}$ sets the scale of the fifth dimension, while the

918: $\sigma_{2}$ parametrize a flat five-dimensional space. In the

919: context of the cosmological models we deal with here, the

920: fields $g_{\mu\nu}^{(4)}$, $\sigma_{i},~i=1,2$ are assumed to

921: depend on the time $t$ only.

922: Type-0 string theory, as well as its supersymmetric

923: versions appearing in brane models,

924: contains appropriate form fields 

925: with non-trivial gauge fluxes (flux-form fields), 

926: which live in the 

927: higher-dimensional bulk space. In the specific model of~\cite{type0}, just

928: one such field was allowed to be 

929: non-trivial.

930: As was demonstrated in~\cite{dgmpp}, a consistent background choice

931: for the flux-form field has the flux parallel 

932: to the fifth dimension $\sigma_1$. This implies  

933: that the internal space is stabilized 

934: in such a way that this dimension is much larger than the 

935: remaining four $\sigma_2$. This demonstrates the physical

936: importance of the flux fields for large radii of compactification.

937: 

938: Considering the fields to be time-dependent only, i.e., considering

939: spherically-symmetric homogeneous backgrounds, restricting ourselves to

940: the compactification (\ref{metriccomp}), and assuming a Robertson-Walker

941: form of the four-dimensional metric with scale factor $a(t)$, the

942: generalized conformal-invariance conditions and the Curci-Pafutti

943: $\sigma$-model renormalizability constraint~\cite{curci} imply the set of

944: differential equations (\ref{liouvilleeq2}), which were solved numerically

945: in~\cite{dgmpp}. The set of $\{ g^i \}$ contains the graviton, dilaton,

946: tachyon, flux and moduli fields $\sigma_{1,2}$ whose vacuum expectation

947: values control the sizes of the extra dimensions.

948: 

949: 

950: The detailed analysis of \cite{dgmpp} indicated that the moduli 

951: fields $\sigma_i$ freeze quickly to their equilibrium values, so, together 

952: with the

953: tachyon field which also decays to a constant value rapidly, they decouple

954: from the four-dimensional fields at very early stages in the evolution of

955: this string Universe~\footnote{The presence of the tachyonic instability

956: in the spectrum of the model of~\cite{dgmpp} is due to the fact that in

957: Type-0 strings there is no target-space supersymmetry by construction.  

958: In other models with supersymmetry breaking~\cite{gravanis,emw}, due to

959: either thermalization or other instabilities, e.g., brane motion, there

960: are also tachyonic modes reflecting the broken supersymmetric spectrum.  

961: From a cosmological viewpoint such tachyon fields are not necessarily bad

962: features, since they may provide the initial instability leading to

963: cosmic expansion.}. There is an inflationary phase in this scenario and

964: dynamical exit from it. The important point to guarantee the exit is the

965: fact that the central charge deficit $Q^2$ is a time-dependent entity in

966: this approach, which obeys specific relaxation laws determined by the

967: underlying conformal field theory~\cite{dgmpp,gravanis,brany}. In fact,

968: the central charge runs with the local world-sheet renormalization-group

969: scale, the zero mode of the Liouville field, which is

970: identified~\cite{emn} with the target time in the $\sigma$-model frame.  

971: The supercriticality~\cite{aben} $Q^2 > 0$ of the underlying $\sigma$

972: model is crucial, as already mentioned.  Physically, the non-critical

973: string provides a framework for {\it non-equilibrium} dynamics, which may

974: be the result of some catastrophic cosmic event, such as a collision of

975: two brane worlds~\cite{ekpyrotic,gravanis,brany}, or an initial quantum

976: fluctuation~\cite{emninfl,dgmpp}.  It also provides, as we discuss below,

977: a unified mathematical framework for analyzing various phases of string

978: cosmology, from the early inflationary phase, graceful exit from

979: it and reheating, until the current and future eras of accelerated

980: cosmologies. Interestingly, one can constrain string parameters such as

981: the separation of brany worlds at the end of inflation, as well as the

982: recoil velocity of the branes after the collision, by fits to current

983: astrophysical data~\cite{brany}.

984: 

985: \subsection{Liouville Inflation: the General Picture} 

986: 

987: As discussed in~\cite{emninfl,gravanis,brany}, 

988: a constant central-charge deficit

989: $Q^2$ in a stringy $\sigma$ model may be associated with an initial 

990: inflationary phase with

991: \begin{equation}

992: \label{centraldeficit} 

993: Q^2 = 9 H^2 > 0~,

994: \end{equation}

995: where the Hubble parameter $H$ can be fixed in terms of other parameters

996: of the model. One may

997: consider various scenarios for such a departure from criticality. For 

998: example,

999: in the model of~\cite{gravanis,brany} this was due to 

1000: the

1001: collision of two brane worlds. In such a scenario, as we now review

1002: briefly, it is possible to obtain an initial {\it

1003: supercritical } central charge deficit, and hence a time-like Liouville

1004: mode in the theory. For instance, in the specific colliding-brane model

1005: of~\cite{gravanis,brany}, $Q$ (and thus $H$) 

1006: is proportional to the square of the

1007: relative velocity of the colliding branes, $Q \propto u^2$ during the

1008: inflationary era.  As is evident from (\ref{centraldeficit}) and discussed

1009: in more detail below, in a phase of constant $Q$ one obtains an

1010: inflationary de Sitter Universe.

1011: 

1012: However, catastrophic non-critical string scenarios for cosmology, such as

1013: that in~\cite{gravanis}, allow in general for a time-dependent deficit

1014: $Q^2(t)$ that relaxes to zero. This may occur in such a way that, although

1015: during the inflationary era $Q^2$ is (for all practical purposes)

1016: constant, as in (\ref{centraldeficit}), eventually $Q^2$ decreases with

1017: time so that, at the present era, one obtains compatibility with the

1018: current accelerating expansion of the Universe. As already mentioned, such

1019: relaxing quintessential scenarios~\cite{gravanis,dgmpp,brany} have the

1020: advantage of asymptotic states that can be defined properly as $t \to

1021: \infty$, as well as a string scattering $S$-matrix~\footnote{As mentioned

1022: in the Introduction, another string scenario for inducing a de Sitter

1023: Universe envisages generating the inflationary space-time from string

1024: loops (dilaton tadpoles)~\cite{fischler}, but in such models a string

1025: $S$-matrix cannot be properly defined.}.

1026: 

1027: The specific normalization in (\ref{centraldeficit})  is 

1028: due to the 

1029: identification of the time $t$ with

1030: the zero mode of the Liouville field $-\varphi$ of the 

1031: {\it supercritical} $\sigma$ model. The minus 

1032: sign may be understood both mathematically, as

1033: due to properties of the Liouville mode, and physically by the requirement

1034: of the relaxation of the deformation of the space-time following

1035: the distortion induced by the recoil. With this identification, the 

1036: general equation of motion for the couplings $\{ g_i \}$ of the 

1037: $\sigma$-model 

1038: background modes is given by (\ref{liouvilleeq})~\cite{emn}:

1039: \begin{equation}

1040: {\ddot g}^i + Q{\dot g}^i 

1041: = -{\tilde \beta}^i (g) = -{\cal G}^{ij} \partial C[g]/\partial g^j~,

1042: \label{liouveq}

1043: \end{equation}

1044: where the dot denotes a derivative with respect to the Liouville

1045: world-sheet zero mode $\varphi$, i.e., target time, 

1046: and ${\cal G}^{ij}$ is an inverse

1047: Zamolodchikov metric in the space of string theory couplings $\{ g^i 

1048: \}$~\cite{zam}. When

1049: applied to scalar, inflaton-like, string modes, (\ref{liouveq})  would

1050: yield standard field equations for scalar fields in de Sitter

1051: (inflationary)  space-times, provided the normalization

1052: (\ref{centraldeficit}) is valid, implying a `Hubble' expansion

1053: parameter $H=-Q/3$~\footnote{The gradient-flow property of

1054: the $\beta$ functions makes the analogy with the inflationary case even

1055: more profound, with the running central charge $C[g]$~\cite{zam} playing

1056: the r\^ole of the inflaton potential in conventional inflationary field

1057: theory.}.  The minus sign in $Q=-3H$ is due to the fact that, as we

1058: discuss below, one identifies the target time $t$ with the world-sheet

1059: zero mode of $-\varphi$~\cite{emn}.

1060: 

1061: The relations (\ref{liouveq}) replace the conformal invariance conditions

1062: $\beta^i = 0$ of the critical string theory, and express the conditions

1063: necessary for the restoration of conformal invariance by the Liouville

1064: mode~\cite{ddk}. Interpreting the latter as an extra target dimension, the

1065: conditions (\ref{liouveq}) may also be viewed as conformal invariance

1066: conditions of a {\it critical} $\sigma$ model in (D+1) target space-time

1067: dimensions, where D is the target dimension of the non-critical $\sigma$

1068: model before Liouville dressing.  In most Liouville approaches, one treats

1069: the Liouville mode $\varphi$ and time $t$ as independent coordinates.  In

1070: our approach~\cite{emn,dgmpp,gravanis}, however, we take a further step,

1071: basing ourselves on dynamical arguments which restrict this extended

1072: (D+1)-dimensional space-time to a hypersurface determined by the

1073: identification $\varphi = -t$. This means that, as time flows, one is

1074: restricted to this D-dimensional subspace of the full (D+1)-dimensional

1075: Liouville space-time.

1076: 

1077: In the work of~\cite{gravanis,brany} which invoked a brane collision

1078: as a source of departure from criticality, this restriction arose because

1079: the potential between massive particles, in an effective field theory

1080: context, was found to be proportional to ${\rm cosh}(t + \varphi)$, which

1081: is minimized when $\varphi = -t$. However, the flow of the Liouville mode 

1082: opposite

1083: to that of target time may be given a deeper mathematical interpretation.  

1084: It may be viewed as a consequence of a specific treatment of the area

1085: constraint in non-critical (Liouville) $\sigma$ models~\cite{emn},

1086: which involves the evaluation of the Liouville-mode path integral via an

1087: appropriate steepest-descent contour.  In this way, one obtains a

1088: `breathing' world-sheet evolution, in which the world-sheet area starts

1089: from a very large value (serving as an infrared cutoff), shrinks to a very

1090: small one (serving as an ultraviolet cutoff), and then inflates again

1091: towards very large values (returning to an infrared cutoff). Such a

1092: situation may then be interpreted~\cite{emn} 

1093: as a world-sheet `bounce' back to the

1094: infrared, implying that the

1095: physical flow of target time is opposite to that of the world-sheet scale

1096: (Liouville zero mode).

1097: 

1098: We now become more specific.  We consider a non-critical $\sigma$ model in

1099: metric ($G_{\mu\nu}$), antisymmetric tensor ($B_{\mu\nu}$), and dilaton

1100: ($\Phi$) backgrounds. These have the following ${\cal O}(\alpha ')$ 

1101: $\beta$ functions (\ref{bfunct}), 

1102: where $\alpha '$ is the Regge slope~\cite{gsw}:

1103:  \begin{eqnarray} 

1104: && \beta^G_{\mu\nu} = \alpha ' \left( R_{\mu\nu} + 2 \nabla_{\mu} 

1105: \partial_{\nu} \Phi 

1106:  - \frac{1}{4}H_{\mu\rho\sigma}H_{\nu}^{\rho\sigma}\right)~, \nonumber \\

1107: && \beta^B_{\mu\nu} = \alpha '\left(-\frac{1}{2}\nabla_{\rho} H^{\rho}_{\mu\nu} + 

1108: H^{\rho}_{\mu\nu}\partial_{\rho} \Phi \right)~, \nonumber \\

1109: && {\tilde \beta}^\Phi = \beta^\Phi - \frac{1}{4}G^{\rho\sigma}\beta^G_{\rho\sigma} = 

1110: \frac{1}{6}\left( C - 26 \right).

1111: \label{bfunctions}

1112: \end{eqnarray} 

1113: The Greek indices are four-dimensional, including target-space-time

1114: components $\mu, \nu, ...= 0,1,2,3$ on the D3-branes

1115: of~\cite{gravanis}, and $H_{\mu\nu\rho}= \partial_{[\mu}B_{\nu\rho]}$ is the

1116: field strength.

1117: We consider the following representation of the four-dimensional 

1118: field strength in terms of a pseudoscalar (axion-like) field $b$: 

1119: \begin{equation}

1120: H_{\mu\nu\rho} = \epsilon_{\mu\nu\rho\sigma}\partial^\sigma b ~,

1121: \label{axion}

1122: \end{equation}

1123: where $\epsilon_{\mu\nu\rho\sigma}$ is the four-dimensional antisymmetric

1124: symbol. Next, we choose an axion background that is linear in the

1125: time $t$~\cite{aben}:

1126: \begin{equation} 

1127: b = b(t) = \beta t~, \quad  \beta={\rm constant} ,

1128: \label{axion2}

1129: \end{equation}

1130: which yields a constant field strength with spatial indices only: $H_{ijk}

1131: = \epsilon_{ijk}\beta$, $H_{0jk}= 0$.  This implies that such a background

1132: is a conformal solution of the full ${\cal O}(\alpha')$ $\beta$ function

1133: for the four-dimensional antisymmetric tensor. We also consider a dilaton

1134: background that is linear in the time $t$~\cite{aben}:

1135: \begin{equation}

1136: \Phi (t,X) = {\rm const} + ({\rm const})' t .

1137: \label{constdil}

1138: \end{equation}

1139: This background does not contribute to the $\beta$ functions 

1140: for the antisymmetric tensor and metric.

1141: 

1142: Suppose now that only the metric is a non-conformal background, due to 

1143: some initial quantum fluctuation or catastrophic event, such as the 

1144: collision of two branes discussed above, 

1145: which results in an initial central charge deficit $Q^2$ 

1146: (\ref{centraldeficit}) that is constant at early stages after the 

1147: collision. Let 

1148: \begin{equation} 

1149: G_{ij} = e^{\kappa \varphi + Hct}\eta_{ij}~, \quad G_{00}=e^{\kappa '\varphi 

1150: + Hct}\eta_{00},

1151: \label{metricinfl}

1152: \end{equation}

1153: where $t$ is the target time, $\varphi$ is the Liouville mode, 

1154: $\eta_{\mu\nu}$ is the four-dimensional Minkowski metric, 

1155: and $\kappa, \kappa '$ and $c$ are constants to be determined. 

1156: As already discussed, the standard inflationary scenario in 

1157: four-dimensional physics requires $Q = -3H$,

1158: which partially stems from the identification

1159: of the Liouville mode with time~\cite{emn}

1160: \begin{equation}

1161: \varphi = -t.

1162: \label{liouvtime}

1163: \end{equation}

1164: The restriction (\ref{liouvtime}) is 

1165: imposed dynamically~\cite{gravanis}

1166: at the end of our computations. Initially, one should treat

1167: $\varphi, t$ as independent target-space-time components. 

1168: 

1169: The Liouville dressing induces~\cite{ddk} $\sigma$-model terms of the form 

1170: $\int_{\Sigma} R^{(2)} Q \varphi$, where $R^{(2)}$ is the world-sheet curvature.

1171: Such terms provide non-trivial contributions to the dilaton background in 

1172: the (D+1)-dimensional space-time $(\varphi,t,X^i)$:

1173: \begin{equation}

1174: \Phi (\varphi,t,X^i) = \frac{1}{2}Q \,\varphi + ({\rm const})' t + {\rm const}.

1175: \label{seventy}

1176: \end{equation}

1177: If we choose 

1178: \begin{equation}

1179: ({\rm const})'=\frac{1}{2}Q~, 

1180: \label{const=Q}

1181: \end{equation}

1182: then (\ref{seventy}) implies a 

1183: {\it constant} dilaton background during the inflationary era, in which the 

1184: central charge deficit $Q^2$ is constant. We justify physically

1185: the choices (\ref{seventy}) and (\ref{const=Q}) later in the article,

1186: when we discuss a specific example of non-criticality induced by 

1187: the collision of brane worlds. 

1188: 

1189: We now consider the Liouville-dressing equations~\cite{ddk}

1190: (\ref{liouveq})  for the $\beta$ functions of the metric and antisymmetric

1191: tensor fields (\ref{bfunctions}). For a constant dilaton field, the

1192: dilaton equation yields no independent information, apart from expressing

1193: the dilaton $\beta$ function in terms of the central charge deficit, as

1194: usual. For the axion background (\ref{axion2}), only the metric yields a

1195: non-trivial constraint (we work in units with $\alpha ' =1$ for

1196: convenience):

1197: \begin{equation} 

1198: {\ddot G}_{ij} + Q{\dot G}_{ij} = -R_{ij} + \frac{1}{2}\beta^2 G_{ij},

1199: \end{equation}

1200: where the dot indicates differentiation with respect to the (world-sheet

1201: zero mode of the) Liouville mode $\varphi$, and $R_{ij}$ is the

1202: (non-vanishing) Ricci tensor of the (non-critical) $\sigma$ model with

1203: coordinates $(t,{\vec x})$:  $R_{00}=0~, R_{ij}=\frac{c^2H^2}{2}e^{(\kappa

1204: - \kappa ')\varphi}\eta_{ij}$. One should also take into account the

1205: temporal ($t$) equation for the metric tensor (which is identically zero for 

1206: antisymmetric backgrounds):

1207: \begin{equation}

1208: {\ddot G}_{00} + Q{\dot G}_{00} = -R_{00} = 0,

1209: \label{tempgrav}

1210: \end{equation}

1211: where the vanishing of the Ricci tensor stems from the 

1212: specific form of the background (\ref{metricinfl}).

1213: We seek metric backgrounds of Robertson-Walker inflationary 

1214: (de Sitter) form:

1215: \begin{equation}

1216: G_{00}=-1~, \quad G_{ij}=e^{2Ht}\eta_{ij}.

1217: \label{desittermetric}

1218: \end{equation}

1219: Then, using (\ref{desittermetric}), (\ref{metricinfl}),

1220: (\ref{constdil}) and (\ref{axion2}), and imposing

1221: (\ref{liouvtime}) at the end, we observe that there is indeed a consistent

1222: solution with:

1223: \begin{equation}

1224: Q = -3H = - \kappa ',~c=3,~\kappa = H,~\beta^2 = 5H^2,

1225: \label{solution}

1226: \end{equation}

1227: corresponding to the conventional form of inflationary equations for

1228: scalar fields.

1229: 

1230: \subsection{Current Stages of Cosmic Liouville Evolution:

1231: Acceleration, Dark Energy and the String Coupling}

1232: 

1233: In the generic class of non-critical string models of interest in this

1234: work, the $\sigma$ model always asymptotes, for long enough cosmic times,

1235: to the linear dilaton conformal $\sigma$-model field theory

1236: of~\cite{aben}. But it is important to stress that this is only an

1237: asymptotic limit. In this respect, the current era of our Universe may be

1238: viewed as being close to, but still not quite at, the relaxation

1239: (equilibrium) point, in the sense that the dilaton is almost linear in the

1240: $\sigma$-model-frame time, and hence varies logarithmically with the

1241: Einstein-frame time (\ref{dil2}). It is expected that this slight

1242: non-equilibrium will lead to a time-dependence of the unification gauge

1243: coupling and other constants (e.g., the four-dimensional Planck length

1244: (\ref{planckstring})), that characterize the low-energy effective field

1245: theory, mainly through the time-dependence of the string coupling

1246: (\ref{defstringcoupl})  as a result of the time-dependent linear dilaton

1247: (\ref{lineardil}).

1248: 

1249: The asymptotic time regime of the Type-0

1250: cosmological string model of~\cite{dgmpp} was obtained 

1251: analytically, by solving the pertinent equations (\ref{liouvilleeq})

1252: for the various fields. As already mentioned, at late times

1253: the theory becomes four-dimensional, and the only non-trivial

1254: information is contained in the scale factor and the dilaton, 

1255: given that the topological flux field remains conformal in this approach,

1256: and the moduli and initial tachyon fields decouple very fast in the 

1257: initial stages after inflation in this model.

1258: For times long after the initial fluctuations, 

1259: such as the present epoch, when the linear approximation is valid, 

1260: the solution for the dilaton in the $\sigma$-model frame

1261: follows from the equations 

1262: (\ref{liouvilleeq}) and takes the form:

1263: \begin{equation} \label{dilaton} 

1264: \Phi (t) =-{\rm ln}\left[\frac{\alpha A}{F_1}{\rm cosh}(F_1t)\right],

1265: \end{equation}

1266: with $F_1$ a positive constant, $\alpha$ a numerical constant of order one,

1267: and 

1268: \begin{equation}\label{defA2}

1269: A = \frac{C_5 e^{s_{01}}}{\sqrt{2}V_6}~, 

1270: \end{equation}

1271: where $s_{01}$ is the equilibrium

1272: value of the moduli field $\sigma_1$, associated with the large bulk dimension,

1273: and $C_5$ the corresponding flux of the five-form flux field.

1274: Notice the that $A$ is independent of this large bulk dimension.

1275: 

1276: For very large times $F_1 t \gg 1$ (in string units) 

1277: one therefore approaches a 

1278: linear solution for the dilaton 

1279: $\Phi \sim {\rm const} -F_1 t$.

1280: From (\ref{dilaton}), (\ref{defstringcoupl}) and (\ref{planckstring}),   

1281: we thus observe that the asymptotic weakness of 

1282: gravity in this Universe~\cite{dgmpp} is due to the smallness of 

1283: the internal space $V_6$ as compared with the flux $C_5$ of the 

1284: five-form field.

1285: The constant $F_1$ is related to the central charge deficit 

1286: of the underlying the non-conformal $\sigma$ model~\cite{dgmpp}:

1287: \begin{equation}\label{ccd}

1288: Q = Q_0 + \frac{Q_0}{F_1}(F_1 + \frac{d\Phi}{dt})~,

1289: \end{equation}

1290: where $Q_0$ is a constant, and 

1291: the numerical solution of (\ref{liouvilleeq}), studied in \cite{dgmpp},

1292: requires that 

1293: \begin{equation}

1294: Q_0/F_1 =(1 + \sqrt{17})/2 \simeq 2.56~,

1295: \label{f1q0}

1296: \end{equation}

1297: which follows from the dilaton equation of motion. 

1298: This connection of $F_1$ to $Q_0$ 

1299: supports the above-described asymptotic 

1300: conformal theory considerations of \cite{aben}, where the model relaxes to

1301: for large times. In this spirit, we require that 

1302: the value of $Q_0$ to which the central charge deficit (\ref{ccd}) asymptotes

1303: must be, for consistency of the underlying string theory,

1304: {\it one of the discrete values} obtained in \cite{aben}, 

1305: for which factorization (unitarity) of the 

1306: string scattering amplitudes 

1307: occurs. Notice that this asymptotic 

1308: string theory, 

1309: with 

1310: a constant (time-independent) central-charge deficit,  

1311: $Q_0^2 \propto c^*-25 $ (or $c^*-9$ for superstring) 

1312: is considered an {\it equilibrium} situation,

1313: where an $S$-matrix can be defined for specific (discrete)

1314: values of the central charge $c^*$~\cite{ddk,aben}.

1315: 

1316: Defining the Einstein frame time $t_E$ through (\ref{einsttime}),

1317: we obtain in this case 

1318: \begin{equation}\label{einstframe}

1319: t_E=\frac{\alpha A}{F_1^2}sinh(F_1 t)~, \quad F_1t = {\rm ln}\left(\sqrt{1 

1320: + \gamma^2t_E^2} + \gamma t_E\right)~,

1321: \end{equation}

1322: where

1323: \begin{equation}\label{defA} 

1324: \gamma \equiv \frac{F_1^2}{\alpha A}~. 

1325: \end{equation}

1326: In terms of the Einstein-frame time,

1327: one obtains a logarithmic time dependence~\cite{aben} for the 

1328: dilaton (\ref{dilaton})~\footnote{Notice that in this subsection 

1329: we work in $D=4$ space-time dimensions. For higher-dimensional models,

1330: the normalisations given in the Introduction, see (\ref{dil2}),

1331: should be used.}:

1332: \begin{equation} 

1333: \Phi _E = {\rm const} -{\rm ln}(\gamma t_E)~,

1334: \label{einsteindil} 

1335: \end{equation}

1336: For large $t_E$, e.g., present or later cosmological time values, 

1337: one has~\cite{dgmpp,emn04}

1338: \begin{equation}\label{einstmetr} 

1339: a_E(t_E) \simeq \frac{F_1}{\gamma}\sqrt{1 + \gamma^2 t_E^2}~.

1340: \end{equation}

1341: At very large (future) times $a(t_E)$ scales linearly

1342: with the Einstein-frame cosmological time $t_E$~\cite{dgmpp},

1343: and hence there is no cosmic horizon. From a field theory

1344: viewpoint, this would 

1345: allow for a proper definition of asymptotic

1346: states and thus a scattering matrix. 

1347: As we mentioned briefly above, however, 

1348: from a stringy point of view,

1349: there are restrictions in the asymptotic values of 

1350: the central charge deficit $Q_0$, and only a  

1351: discrete spectrum of values of $Q_0$ allow for a full stringy S-matrix

1352: to be defined,

1353: respecting modular invariance~\cite{aben}. 

1354: The Universe relaxes asymptotically 

1355: to its ground-state equilibrium situation,

1356: and the non-criticality of the string caused by the initial fluctuation

1357: disappears, yielding a critical (equilibrium) string Universe

1358: with Minkowski metric and a linear-dilaton background. 

1359: This is the generic feature of the models we consider here and 

1360: in \cite{emn04}, allowing the conclusions 

1361: to be extended beyond the Type-0 string theory to incorporate

1362: also target-space supersymmetric strings/brane models, such as 

1363: those in \cite{emw,brany}.

1364: 

1365: An important comment is in order at this point, regarding the 

1366: form of the Einstein metric corresponding to (\ref{einstmetr}):

1367: \begin{equation}\label{einstmetr2}

1368: g_{00}^E=-1, ~~\quad g_{ii} = a_E^2(t_E) 

1369: = \frac{F_1^2}{\gamma^2} + F_1^2t_E^2~.

1370: \end{equation}

1371: Although asymptotically, for $t_E \to \infty$, the above metric

1372: asymptotes to the linearly-expanding Universe (\ref{metricaben}),

1373: the presence of a constant $F_1^2/\gamma^2$  contribution

1374: implies that the solution for large but finite $t_E$, such as the current 

1375: era of the Universe, is different from that of \cite{aben}. 

1376: Indeed, the corresponding 

1377: $\sigma$-model-frame metric (\ref{smodeinst}) is not Minkowski flat,

1378: and in fact the pertinent $\sigma$ model does not correspond 

1379: to a conformal field theory. This should come as no surprise since,

1380: for finite $t_E$ no matter how large, the $\sigma$-model 

1381: theory requires

1382: Liouville dressing. {\it It is only at the end-point of time/flow 

1383: $t_E \to \infty$ that the underlying string theory becomes conformal,

1384: and the system reaches equilibrium.}

1385: 

1386: 

1387: The Hubble parameter of such a 

1388: Universe for large $t_E$ is

1389: \begin{equation}\label{hubble2} 

1390: H(t_E) \simeq \frac{\gamma^2 t_E}{1 + \gamma^2 t_E^2} = 

1391: \frac{F_1^2t_E}{a^2(t_E)}~.

1392: \end{equation}

1393: On the other hand, the Einstein-frame effective four-dimensional 

1394: `vacuum energy density', defined through the running 

1395: central-charge deficit $Q^2$,

1396: upon compactification to four dimensions of the ten-dimensional

1397: expression $2\int d^{10}x \sqrt{-g}e^{2\Phi}Q^2(t_E)$ in the 

1398: Einstein frame, is~\cite{dgmpp}:

1399: \begin{equation} 

1400: \Lambda_E (t_E) = 2e^{2\Phi - \sigma_1 - 5\sigma_2}Q^2(t_E) 

1401: \simeq \frac{2Q_0^2 \gamma^2}{F_1^2 ( 1 + \gamma^2 t_E^2)}

1402: \sim \frac{13.11\gamma^2}{1 + \gamma^2 t_E^2}

1403: \label{cosmoconst2} 

1404: \end{equation}

1405: in the normalization of (\ref{effaction}).

1406: Here we used (\ref{ccd}) for $Q$ at large $t_E$, 

1407: approaching its equilibrium value $Q_0$, and we have also used

1408: (\ref{f1q0}). Thus, the dark energy 

1409: density relaxes to zero for $t_E \to \infty$.

1410: Notice an important feature of the relaxation form (\ref{cosmoconst2}),

1411: namely that the proportionality constants in front 

1412: are such that, for asymptotically large $t_E \to \infty$, 

1413: $\Lambda (t_E \to \infty)$ is independent of 

1414: the equilibrium conformal field theory central charge $Q_0$. 

1415: 

1416: 

1417: Finally, and most importantly for our purposes here, 

1418: the deceleration parameter 

1419: in the same regime 

1420: of $t_E$ becomes:

1421: \begin{equation} 

1422: q(t_E) = -\frac{(d^2a_E/dt_E^2)~a_E}{({da_E/dt_E})^2}

1423: \simeq -\frac{1}{\gamma^2 t_E^2}~.

1424: \label{decel4}

1425: \end{equation}

1426: {\it As is clear from 

1427: (\ref{einsteindil}), (\ref{defstringcoupl}), 

1428: this expression can be identified, up

1429: to irrelevant constant factors which by normalization 

1430: are set to one, 

1431: with the square of the string coupling 

1432: (\ref{defstringcoupl})~\cite{emn04}}:

1433: \begin{equation}

1434: |q(t_E)| = g_s^2~.

1435: \label{important}

1436: \end{equation}

1437: To guarantee the consistency of perturbation theory, one must have $g_s < 

1438: 1$,

1439: which can be achieved in our approach 

1440: if one {\rm defines} the  

1441: {\it present era} by the time regime

1442: \begin{equation}

1443: \gamma^2 \sim \beta^2 t_E^{-2} 

1444: \label{condition}

1445: \end{equation} 

1446: in the Einstein frame.

1447: In view of its relation with the deceleration parameter at late epochs

1448: (\ref{decel4}), $q=-1/\beta^2$, the numerical value of 

1449: $\beta^2$ is determined by 

1450: requiring agreement with the data~\cite{wmap}.

1451: As we discuss below, phenomenologically $\beta^2 = {\cal O}(1)$. 

1452: 

1453: This is compatible with the time $t_E$ being large enough 

1454: (in string units) 

1455: for 

1456: \begin{equation} 

1457: |C_5|e^{-5s_{02}}/F_1^2 \sim |C_5|e^{-5s_{02}}/Q_0^2 \gg 1~, 

1458: \label{largetwe}

1459: \end{equation} 

1460: as becomes clear from (\ref{defA2}),(\ref{defA}), (\ref{f1q0}). 

1461: This condition can be guaranteed

1462: {\it either} for small radii of five of the extra dimensions, 

1463: {\it or} for a large value of the flux $|C_5|$ of the five-form 

1464: of the Type-$0$ string, compared with $Q_0$. We discuss 

1465: in the next subsection concrete examples of non-critical 

1466: string cosmologies, in which the asymptotic value of the 

1467: central charge $Q_0 \ll 1$ in string units.  

1468: Recalling that the relatively large extra dimension

1469: in the direction of the flux

1470: $s_{01}$ decouples from this condition, we thus 

1471: observe that there is the possibility of constructing 

1472: effective five-dimensional models 

1473: with a large uncompactified fifth dimension that are consistent with the 

1474: condition

1475: (\ref{condition}).  

1476: Notice that, in the regime (\ref{condition}) of Einstein-frame times, 

1477: the Hubble parameter

1478: and the cosmological constant continue to be compatible

1479: with the current observations, 

1480: and in fact to depend on $\gamma \sim t_E^{-1}$ in the same way

1481: as in their large-$\gamma t_E$ regime given above 

1482: (\ref{hubble2}),(\ref{cosmoconst2}), but now the string coupling

1483: (\ref{important}) is kept smaller than one and finite, of order

1484: $1/2$, as also suggested  by grand unification phenomenology~\cite{gsw}.

1485: 

1486: We next turn to the equation of state of our Universe.

1487: As discussed in~\cite{dgmpp}, it resembles a  

1488: quintessence model 

1489: with the dilaton playing the r\^ole of the quintessence

1490: field. 

1491: Hence the equation of state

1492: for our Type-$0$ string Universe reads~\cite{carroll}:

1493: \begin{equation}\label{eqnstate} 

1494: w_\Phi = \frac{p_\Phi}{\rho_\Phi}=\frac{\frac{1}{2}({\dot \Phi})^2 - V(\Phi)}

1495: {\frac{1}{2}({\dot \Phi})^2 + V(\Phi)}~,

1496: \end{equation}

1497: where $p_\Phi$ is the pressure and $\rho_\Phi$ is the energy density, 

1498: and $V(\Phi)$ is the effective potential for the dilaton, which in our case

1499: is provided by the central-charge deficit term. 

1500: Here the dot denotes Einstein-frame differentiation. 

1501: In the Einstein frame, in the normalization of (\ref{effaction}),  

1502: the potential $V(\Phi)$ is given by 

1503: \begin{equation}                       

1504: V(\Phi) = \frac{\Lambda_E}{4} \sim \frac{6.56\gamma^2}{2(1 + \gamma^2t_E^{2})}~,

1505: \label{potendilaton}

1506: \end{equation} 

1507: where $\Lambda_E$ is given in (\ref{cosmoconst2})

1508: and we have used (\ref{f1q0}). 

1509: Defining the present era 

1510: by the condition (\ref{condition}), 

1511: we obtain from (\ref{dilaton}),(\ref{einstframe}):

1512: \begin{eqnarray}\label{dilpotkin}  

1513: \frac{1}{2}\left(\frac{d\Phi}{dt_E}\right)^2 = \frac{1}{2}\beta^2 \cdot 

1514: {\rm tanh}^2{\rm ln}\left( \sqrt{1 + \beta^2} + \beta \right) \cdot 

1515: \frac{1}{t_E^2 (1 + \beta^2)}, \qquad V(\Phi) 

1516: \sim \frac{6.56\beta^2}{2(1 + \beta^2)t_E^2}~. 

1517: \end{eqnarray} 

1518: This 

1519: implies a \emph{constant} equation of state (\ref{eqnstate}) in the 

1520: current era: 

1521: \begin{equation} 

1522: w_\Phi (t_E \gg 1) = \frac{{\rm tanh}^2{\rm ln}\left(\sqrt{1 + \beta^2}+ \beta \right)- 6.56}{{\rm tanh}^2{\rm ln}\left(\sqrt{1 + \beta^2}+ \beta \right)

1523: + 6.56}.

1524: \label{eqnstatedilnum}

1525: \end{equation}

1526: We now remark that, if we use as the value of 

1527: $q$ today the one inferred by best fits of FRW cosmology 

1528: to the data on high-redshift supernovae and the CMB~\cite{snIa,wmap}:

1529: \begin{equation}\label{qval}

1530: q_{FRW,data}=-\frac{1}{\beta^2} \simeq -0.57~~({\rm today})~,

1531: \end{equation} 

1532: this corresponds by (\ref{eqnstatedilnum}) to an equation of state with

1533: \begin{equation} 

1534: w_\Phi = -0.82~, 

1535: \label{eqnstatedilnum3}

1536: \end{equation}

1537: which is in the region allowed by the 

1538: data~\cite{wmap,snIa,steinhardt}~\footnote{Although the data at present 

1539: are not sufficient for an accurate measurement of $w(z)$, they

1540: seem to indicate~\cite{steinhardt} negative

1541: values smaller than $-0.6$ for $z \simeq 1$ and $w(z\to 0) \to -1$.}. 

1542: On the other hand, an equation of state $w_\Phi = -0.78$, which is  

1543: the upper bound   

1544: given by the WMAP data~\cite{wmap,steinhardt}, yields   

1545: a current-era 

1546: deceleration (\ref{decel4}),(\ref{condition}) $q \sim -0.25$. 

1547: All such values of $|q|< 1$ imply via (\ref{important})

1548: perturbative values for the string coupling, 

1549: close to the value used frequently in 

1550: string-inspired particle phenomenology: $|q|= g_s^2 \sim 1/2$. 

1551: 

1552: The inclusion of 

1553: matter modifies the situation, and allows for a more complete

1554: expression of the equation of state in terms of the current

1555: acceleration of the Universe for our model. As we discuss

1556: in the next subsection, this is different 

1557: from the analogous relation in conventional FRW cosmologies.

1558: However, 

1559: the function $q(z)$ is not yet measurable with sufficient

1560: accuracy using supernovae alone: in order to infer a precise form of 

1561: $q(z)$ from such measurements,

1562: one has to make certain assumptions about the underlying 

1563: dynamics.

1564: In conventional FRW cosmologies, $q(z)$ 

1565: is expressed in terms of the various 

1566: energy-density components $\Omega_i(z)$ using the 

1567: underlying Einstein cosmology. In a similar vein, in our 

1568: model $q(z)$ can be expressed in terms

1569: of the various energy-density components, in units of the critical

1570: density, using the dynamics encoded in (\ref{eqsmotion}). However,

1571: due to the off-shell Liouville modifications, the critical density 

1572: for a spatially flat Universe and the relation of 

1573: $q(z)$ to the various energy-density components are different from 

1574: those in conventional

1575: FRW models. Thus the above-used values of $\beta^2$ 

1576: should not be taken for 

1577: granted but only as indicative. For us, $\beta^2=1/|q|$ can only be determined

1578: properly after a detailed direct fit of our model to the data.

1579: We discuss these issues in the next Section, where we

1580: show that the above considerations can be made compatible 

1581: with the observations that suggest there was a past epoch of

1582: deceleration at redshifts larger than $0.5$~\cite{deceldata}. 

1583: 

1584: These considerations concern the specific model of~\cite{dgmpp}.

1585: One can be more generic when considering equations of state 

1586: for dilatonic dark energy Liouville models, by simply

1587: requiring that the present era is described by a linear 

1588: dilaton solution (\ref{dil2}), asymptotic to a conformal

1589: field theory with central charge deficit $Q^2$.

1590: In this case, the dilaton potential and kinetic energy are given by 

1591: \begin{eqnarray}\label{genpot}  

1592: && V(\Phi) = \frac{Q^2}{2}e^{2\Phi} = \frac{2}{t_E^2}e^{2\Phi_0}~, 

1593: ~~\Phi = \Phi_0 -{\rm ln}\frac{Qt_E}{2}~, \nonumber \\

1594: && \frac{d\Phi}{dt_E} = \frac{1}{t_E} ,

1595: \end{eqnarray}

1596: where $\Phi_0$ is a constant, denoting the initial value of 

1597: the dilaton field in a generic situation. 

1598: As we have seen previously (\ref{dilaton}),

1599: this constant is 

1600: determined in the model of~\cite{dgmpp} by the values of the flux field 

1601: and the 

1602: frozen moduli. In the general situation, where no microscopic model

1603: is specified, this constant is free to be determined

1604: by phenomenology, as we see below.  

1605: In such a generic situation, the dilatonic dark-energy 

1606: equation of state (in the same normalization of (\ref{effaction}), 

1607: reads:

1608: \begin{equation} 

1609: w_\Phi (t_E \gg 1) \simeq \frac{1 - 4e^{2\Phi_0}}{1 + 

1610: 4e^{2\Phi_0}}~. 

1611: \label{eqnstatedilnum2}

1612: \end{equation}

1613: One can easily obtain 

1614: phenomenologically acceptable values of $w_\Phi$~\cite{wmap}

1615: by adjusting the value of the constant $\Phi_0$.

1616: For instance, for $e^{2\Phi_0} > \frac{7}{4}$ one obtains:

1617: \begin{equation} 

1618: w_\Phi \le -3/4 ,

1619: \label{condition2}

1620: \end{equation} 

1621: in agreement with 

1622: the WMAP value~\cite{wmap,steinhardt}. 

1623: 

1624: Such linear dilaton models can be made compatible with 

1625: perturbative string couplings

1626: $g_s < 1$, as required by string-inspired 

1627: particle-physics phenomenology, provided one chooses the asymptotic central

1628: charge $Q$ in the region 

1629: \begin{equation}

1630: Q t_E > \sqrt{7} .

1631: \label{ccrestriction}

1632: \end{equation}

1633: Notice that the dark energy (\ref{genpot}) 

1634: is independent of the value of the central charge, 

1635: and its magnitude today in such models depends only on the age of the Universe.

1636: The important question in such models is the precise form of the 

1637: scale factor, which should be obtained as a specific solution of the 

1638: appropriate dynamical equations (\ref{liouveq}), whose form depends

1639: on the details of the underlying string theory. 

1640: To be specific, in what follows  

1641: we adopt~\cite{emn04} 

1642: the class of string models that yield 

1643: predictions similar to those in~\cite{dgmpp} 

1644: as best 

1645: describing the current era of our string Universe.

1646: 

1647: \subsection{Inclusion of Matter and Radiation}

1648: 

1649: So far, our model has not included ordinary matter or radiation, as only

1650: fields from the gravitational string multiplet have been included. The

1651: inclusion of ordinary matter is not expected to change the results

1652: significantly, and we conjecture that the fundamental relation

1653: (\ref{important}) will continue to hold, the only difference being that

1654: probably the inclusion of ordinary matter will tend to reduce the string

1655: acceleration, due to the fact that matter, being subjective to attractive

1656: gravity, resists the acceleration of the Universe.

1657: 

1658: 

1659: We now discuss in some detail the formalism that allows the inclusion of

1660: matter in the Liouville framework.  The important thing to notice is that,

1661: in the absence of matter, the Liouville-dressing approach of~\cite{emn},

1662: together with the eventual (dynamical) identification of the Liouville

1663: zero mode with the target time, as explained above, leads to the

1664: generalized conformal invariance conditions (\ref{liouveq}) for the fields

1665: of the gravitational multiplet of the string propagating in a

1666: four-dimensional background~\footnote{Or five-dimensional, in the case of

1667: compactified brane models in a single large bulk dimension.}.  These

1668: \emph{are not} the ordinary equations of motion corresponding to a

1669: four-dimensional gravitational effective action (\ref{effaction}), but

1670: describe the dynamics of an \emph{off-shell relaxation process}.

1671: 

1672: Matter coupling to on-shell dilatonic gravity theory has been considered

1673: in the past, see, e.g.,~\cite{damour,gasperini}, but in an \emph{on-shell}

1674: formalism of critical strings, where the various target-space fields

1675: satisfied classical equations of motion derived from a four-dimensional

1676: action. As explained above, this is distinct from our Liouville cosmology

1677: approach. Moreover, the analysis of~\cite{gasperini}, although dealing

1678: with the possibility of a dilaton playing the r\^ole of a quintessence

1679: field responsible for the current acceleration of the Universe,

1680: nevertheless considers models in which the dilaton as well as its

1681: potential increases to positive infinity, as the cosmic time elapses.  

1682: This is exactly opposite to our situation here and in~\cite{dgmpp,emn04},

1683: where the dilaton $\Phi \to -\infty$ asymptotically. In our situation, for

1684: large cosmic times, the string coupling $e^\Phi \to 0$, and this is the

1685: reason why in present and future eras of the Universe the

1686: string-tree-level approximation is sufficient. This is to be contrasted

1687: with the case of~\cite{gasperini}, where one should include asymptotically

1688: (all) higher loop corrections to the string effective action, which are

1689: not known at present.

1690: 

1691: We now discuss in some detail the proper inclusion of 

1692: matter in our Liouville framework. 

1693: The essential formalism 

1694: is that of~\cite{aben}, in which all physically relevant quantities

1695: should be reduced to the Einstein frame (\ref{smodeinst}) 

1696: and Einstein cosmic time (\ref{einsttime}) 

1697: framework. Examining the four-dimensional 

1698: matter action (including radiation fields),

1699: we observe that in string theory this action couples to the dilaton

1700: field non-trivially, in a way that is specific to the various

1701: matter species, as a result of purely stringy properties of the

1702: effective action~\cite{gsw}. 

1703: A generic $\sigma$-model-frame effective four-dimensional action

1704: with dilaton potential $V(\Phi)$, which could even include higher-string-loop

1705: corrections, has the form:

1706: \begin{eqnarray}

1707: && S^{(4)} = \frac{1}{2\alpha '} \int d^4 x \sqrt{-G} [e^{-\Psi(\Phi)} R(G) 

1708: + Z(\Phi)(\nabla \Phi)^2 + 2\alpha ' V(\Phi)  \dots ] 

1709: - \nonumber \\

1710: && \frac{1}{16\pi}\int d^4x \sqrt{G}\frac{1}{\alpha (\Phi)}F_{\mu\nu}^2 

1711: - I_{\rm m}(\Phi, G, {\rm matter})~, 

1712: \label{matteraction}

1713: \end{eqnarray}

1714: in the notation of \cite{gasperini}, with the various factors

1715: $\Psi, Z, \alpha $ encoding information about higher 

1716: string loop corrections. Also, $F_{\mu\nu}$ denotes 

1717: the radiation field strength and $I_{\rm m}(\Phi, G, {\rm matter})$

1718: represents  matter contributions, which  couple to the dilaton $\Phi$

1719: in a manner dictated by string theory scaling laws~\cite{gsw} 

1720: with shifts of the dilaton field $\Phi \to \Phi + {\rm const}$.

1721: In our situation, where only the string tree level plays a r\^ole 

1722: at late times, the various form factors simplify, e.g., $\Psi (\Phi) = 

1723: 2\Phi, Z(\Phi) = 4 e^{- 2 \Phi}$, \emph{etc.}. However, for purposes of 

1724: generality,

1725: in this section we keep the form (\ref{matteraction}).  

1726: When higher loop corrections 

1727: are important, these factors have a complicated form, for instance 

1728: one has $e^{\Psi (\Phi)} = c_0 e^{-2\Phi} + c_1 + 

1729: c_2e^{2\Phi} + \dots $, with $c_i$ constants, and the powers of the 

1730: square of the string coupling $g_s^2=e^{2\Phi}$ count the numbers of 

1731: closed string loops, as appropriate for the gravitational multiplet. 

1732: For simplicity in this subsection we ignore the four-dimensional

1733: antisymmetric tensor 

1734: field, which, as discussed in~\cite{aben} and mentioned above, 

1735: corresponds to an axion field.

1736: 

1737: According to our discussion in Section 2.1, the action $S^{(4)}$ 

1738: coincides with the $I^{(D)}-{\cal I}_\phi$ 

1739: part of the $D$-dimensional action (\ref{ddimaction}),

1740: obtained from (\ref{lambdaaction}) when $D=4$,

1741: upon the identification of the Liouville mode $\phi$ with the 

1742: target time $X^0$:

1743: \begin{equation}\label{4doffshell}

1744: S^{(4)} = I^{(4)} - {\cal I}_\phi = 

1745: \int d^4X e^{-\varphi} \{ C^{(4)}(X) - 25\}.

1746: \end{equation} 

1747: In contrast to the critical-string case considered 

1748: in~\cite{damour,gasperini},  

1749: the field variations of (\ref{4doffshell}) 

1750: do not yield zero, 

1751: but are such that they compensate the variations of the 

1752: remaining (Liouville) part of (\ref{lambdaaction}), 

1753: in order to yield the 

1754: generalized conformal invariance conditions (\ref{liouvilleeq2}),

1755: augmented by the inclusion of matter fields.

1756: In particular, 

1757: the set of couplings $\lambda^I$ in (\ref{lambdaaction}),

1758: as well as the action $C^{(4)}(X)$ (c.f. (\ref{matteraction})),     

1759: should now include matter fields in addition to the fields of the gravitational 

1760: multiplet of the string. For simplicity, however, we may 

1761: assume that, at least at the late epochs of the Universe

1762: which are of interest to us here, 

1763: the matter couplings are almost conformal, and the 

1764: dominant reason for departure from criticality lies in the

1765: fields of the gravitational multiplet. This is the case,

1766: for instance, in the colliding-brane scenario discussed

1767: in the next Section. This leaves the off-shell 

1768: Liouville part of (\ref{lambdaaction}) in the form 

1769: discussed in the previous Section. This will be understood 

1770: in what follows. 

1771: 

1772: We have: 

1773: \begin{equation} 

1774: \frac{\delta S^{(4)}}{\delta g^i} = -\frac{\delta {\cal I}_\phi}{\delta g^i}~,

1775: \label{vars}

1776: \end{equation}

1777: where $g^i=(\Phi, G, \dots) \equiv (\Phi, \lambda^I)$ and 

1778: we took into account the fact that the action ${\cal I}^{(4)} \equiv

1779: {\cal I}^{(4+1)}|_{\phi =X^0}$ is critical (the identification 

1780: of the Liouville mode with the target time $X^0$ is done after the 

1781: respective variation is taken). 

1782: Near a fixed point 

1783: one has  

1784: $\frac{\delta S^{(4)}}{\delta \lambda^I} = {\ddot g}_I + Q{\dot g}_I$,

1785: with the overdot denoting differentiation with respect 

1786: to the Liouville zero mode. 

1787: When passing to the Einstein frame (\ref{smodeinst}), and 

1788: expressing the time in terms of the cosmic time $t_E$ (\ref{einsttime}),

1789: the left-hand side of (\ref{liouvilleeq2}) in the supercritical

1790: string case for the graviton fields $G_{\mu\nu}$ yields: 

1791: \begin{eqnarray} 

1792: &&{\dot G}_{\mu\nu} = e^{\Phi} \left(2\frac{d\Phi}{dt_E}g_{\mu\nu}^E

1793: + \frac{dg_{\mu\nu}^E}{dt_E} \right)~,

1794: \nonumber \\

1795: && {\ddot G}_{\mu\nu} = 

1796: 2\left(\frac{d\Phi}{dt_E}\right)^2 g_{\mu\nu}^E + 

1797: 2\frac{d^2\Phi}{dt_E^2}g_{\mu\nu}^E + 3 \frac{d\Phi}{dt_E}

1798: \frac{dg_{\mu\nu}^E}{dt_E} + \frac{d^2g_{\mu\nu}^E}{dt_E^2}~,

1799: \nonumber \\

1800: && (0,0)-{\rm component}:~ {\tilde {\cal G}}_{00} \equiv 

1801: {\ddot G}_{00} + Q{\dot G}_{00} = 

1802: -2Q\frac{d}{dt_E}\left(e^\Phi\right) - 2(\frac{d\Phi}{dt_E})^2 - 

1803: 2\frac{d^2\Phi}{dt_E^2}~, ~~(g_{00}^E = -1)~, \nonumber \\

1804: && (i,i)-{\rm component}:~ {\tilde {\cal G}}_{ii} \equiv 

1805: {\ddot G}_{ii} + Q{\dot G}_{ii} = \nonumber \\

1806: && 2a^2(t_E)\left((\frac{d\Phi}{dt_E})^2 + 3\frac{d\Phi}{dt_E}H + 

1807: \frac{d^2\Phi}{dt_E^2} + (1-q)H^2\right) 

1808: + 2Qa^2(t_E)e^\Phi\left(\frac{d\Phi}{dt_E} + H\right)

1809: ~, \nonumber \\

1810: && H \equiv a^{-1}(t_E)\frac{da(t_E)}{dt_E}~, ~~ q \equiv 

1811: -\frac{a(t_E)\frac{d^2}{dt_E^2}a(t_E)}{(\frac{da(t_E)}{dt_E})^2}~,

1812: \label{liouvmatter}

1813: \end{eqnarray}

1814: with $q$ the deceleration parameter (\ref{decel4}). 

1815: The dilaton variation of the function ${\cal I}_\phi$ of (\ref{ddimaction}), 

1816: on the other hand, reads: 

1817: \begin{eqnarray} 

1818: &&\frac{\delta {\cal I}_\phi}{\delta \Phi} \equiv {\cal I}'_\phi = 

1819: 6\int d^{D-1}X~e^{-\varphi}\left({\dot \varphi}^2 - {\ddot \varphi}

1820: + \frac{1}{8}({\dot \lambda}^I)^2\right) = \nonumber \\

1821: && 6V^{(3)}\{ \left(2\frac{d\Phi}{dt_E} + H \right)^2

1822: - \frac{d\Phi}{dt_E}\left(2\frac{d\Phi}{dt_E} + H \right) + 

1823: \frac{dH}{dt_E} + 2\frac{d^2\Phi}{dt_E^2} + 2\left(\frac{d\Phi}{dt_E}\right)^2

1824: + 3H\frac{d\Phi}{dt_E} + \frac{3}{2}H^2\}~, \nonumber \\

1825: \label{iprimephi}

1826: \end{eqnarray}

1827: in the notation of (\ref{lambdaaction}), where $V^{(3)}$ denotes the 

1828: three-dimensional spatial volume.

1829: A complete analysis of matter effects requires solving the equations

1830: emerging from considering the variations (\ref{vars}),(\ref{liouvmatter})

1831: and (\ref{iprimephi}) with respect to the metric field in the Einstein and

1832: cosmic time frames~\cite{aben}. This depends on the specific form of

1833: matter action considered.

1834: 

1835: At this stage we remind the reader 

1836: of a few crucial technical details on the equivalence of the generalized

1837: conformal invariance conditions (\ref{liouveq}) to 

1838: target-space dynamical equations. The Zamolodchikov metric

1839: in theory space,

1840: ${\cal G}_{ij} = z^2{\overline z}^2 \langle V_i(z)V_j(0)\rangle $,

1841: where $\langle \dots \rangle $ denotes a $\sigma$-model 

1842: average including Liouville contributions,  

1843: acts as a link between the 

1844: $\sigma$-model $\beta$ functions and field variations of the

1845: target-space effective actions $S[g]$: 

1846: \begin{equation}\label{effactions}

1847: {\cal G}_{ij}\beta^j = \frac{\delta S[g]}{\delta g^i} ,

1848: \end{equation}

1849: where, in what follows, the $g^i$ 

1850: denote various background target-space fields other than the dilaton,

1851: which is treated separately. 

1852: To order $\alpha '$, standard analysis~\cite{miramontes,emn} 

1853: shows that one can find a renormalization scheme on the 

1854: world sheet in which the Zamolodchikov metric,

1855: in the case of the graviton 

1856: and dilaton backgrounds we are restricting ourselves here,  

1857: becomes near a fixed point 

1858: \begin{equation}

1859: {\cal G}_{ij} = e^{-2\Phi}\left(\delta_{ij} + {\cal O}(g^2)\right)~, \qquad

1860: \delta_{ij} \to \frac{1}{2}\left(-G^{\mu\nu}G^{\alpha\beta} + G^{\mu\alpha}G^{\nu\beta} 

1861: + G^{\mu\beta}G^{\nu\alpha} \right) ,

1862: \label{miramontes}

1863: \end{equation}

1864: where $G_{\mu\nu}$ is a $\sigma$-model-frame target-space metric

1865: (which, in our case, is four-dimensional after appropriate compactification or

1866: restriction on a three-brane). The

1867: exponential dilaton term arises from world-sheet zero-mode

1868: contributions to the $\sigma$-model average at tree level, and

1869: includes (linear) Liouville-zero-mode- ($\phi_0$-) dependent

1870: terms in the non-critical string case. 

1871: This Liouville 

1872: dependence is crucial~\cite{emn} in ensuring the 

1873: following property for the Zamolodchikov metric:

1874: $d{\cal G}_{ij}/d\phi_0 = Q {\cal G}_{ij}$, with $Q^2$ 

1875: the central-charge deficit, 

1876: which guarantees the derivation 

1877: of the Liouville terms on the left-hand side of 

1878: (\ref{liouveq}) from a target-space action, as seen above, and therefore 

1879: the 

1880: canonical quantization of the Liouville-dressed couplings/fields $ g^i$ 

1881: in string-theory space (upon summation over world-sheet topologies). 

1882: 

1883: The form (\ref{miramontes}) implies that a contraction with a Ricci (or

1884: any other symmetric)  tensor in target space, which is contained in the

1885: graviton $\beta$ function to ${\cal O}(\alpha ')$, results in an Einstein

1886: tensor on the right-hand side of (\ref{effactions}), as appropriate for

1887: proper target-space dynamics. In a similar manner, upon contraction with

1888: (\ref{miramontes}), one obtains appropriate Einstein-like tensor

1889: structures for the Liouville modifications apppearing in the left-hand

1890: side of (\ref{liouveq}) for the graviton case.

1891: On the other hand, considering the variation with respect to the 

1892: four-dimensional graviton 

1893: $g^i  \equiv G_{\mu\nu}$, and passing to the Einstein-frame 

1894: (\ref{smodeinst}), we obtain: 

1895: \begin{equation}\label{einsteinframeaction}

1896: \frac{\delta S[g]}{\delta G_{\mu\nu}} = 

1897: e^{-2\Phi}\frac{\delta S[g]}{\delta g^E_{\mu\nu}} ,

1898: \end{equation}

1899: where the precise form of the exponential factor is exclusive to the four 

1900: target space-time dimensions we consider here. As a result 

1901: of (\ref{miramontes}), (\ref{einsteinframeaction}), the 

1902: exponential factors 

1903: cancel out in (\ref{effactions}). The above 

1904: results and 

1905: properties will be understood in what follows.

1906: 

1907: Defining the Einstein-like tensor 

1908: \begin{eqnarray}\label{jcaldef}

1909: {\cal J}_{\mu\nu} \equiv {\tilde {\cal G}}_{\mu\nu} 

1910: - \frac{1}{2}g_{\mu\nu}^E 

1911: \left (g^{\nu\lambda}_E {\tilde {\cal G}}_{\nu\lambda} \right)~,

1912: \end{eqnarray} 

1913: and assuming a normal fluid form for matter or radiation, with stress tensor

1914: $T_\mu^{\nu E} = {\rm diag}\left(-\rho, p\delta_i^j\right)$

1915: in the Einstein frame (\ref{smodeinst}),(\ref{einsttime}),

1916: we then obtain the following gravitational and dilaton equations of motion 

1917: (in units $M_P^2 = 1/8\pi G_N = 2$, where $M_P$ is the four-dimensional 

1918: Planck constant): 

1919: \begin{eqnarray} 

1920: && 6H^2 = \rho + \rho_\Phi + {\cal J}_{00}~, \nonumber \\

1921: && 4\frac{d}{dt_E}H + 6H^2 = -p - p_\Phi - a^{-2}(t_E){\cal J}_{ii}~,\qquad i = 1,2,3, 

1922: \nonumber \\

1923: && \frac{d^2\Phi}{dt_E^2} + 3H\frac{d\Phi}{dt_E} 

1924: +  V'(\phi) + \frac{1}{2}[\Psi'(\Phi)(\rho - 3p) + \sigma + \sigma_\phi ]=0~,\nonumber \\

1925: && \sigma \equiv  -2\frac{1}{V^{(3)}\sqrt{-g_E}}

1926: \frac{\delta (I_{\rm m} + \int (16\pi\alpha (\Phi))^{-1}F^2)}{\delta \Phi}~, 

1927: \nonumber \\

1928: && \sigma_\phi \equiv  -2\frac{1}{V^{(3)}\sqrt{-g_E}}{\cal I}'(\Phi)

1929: = -12 \{ 4\left(\frac{d\Phi}{dt_E}\right)^2  + 6H\frac{d\Phi}{dt_E}

1930: + 2 \frac{d^2\Phi}{dt_E^2} + \frac{dH}{dt_E}+ \frac{5}{2}H^2\}~, 

1931: \nonumber \\

1932: &&\rho_\Phi \equiv \frac{1}{2}\left(\frac{d\Phi}{dt_E}\right)^2 + V(\Phi)~,

1933: \quad p_\Phi \equiv \frac{1}{2}\left(\frac{d\Phi}{dt_E}\right)^2 - V(\Phi)~,

1934: \label{eqsmotion}

1935: \end{eqnarray}

1936: where the prime denotes differentiation

1937: with respect to $\Phi$, 

1938: ${\cal I}_\phi'(\Phi)$ is defined in (\ref{iprimephi}), and we 

1939: use canonically-normalized dilaton fields.

1940: Notice that (\ref{eqsmotion}) differ from the 

1941: corresponding on-shell equations in~\cite{gasperini} by the 

1942: Liouville \emph{out-of--equilibrium} 

1943: contributions ${\cal J}$ and $\sigma_\phi$,

1944: which are exclusive to our treatment~\cite{emn,emninfl,dgmpp}. 

1945: 

1946: The equations (\ref{eqsmotion}) lead, after standard manipulations, 

1947: to the coupled (non-) conservation equations of 

1948: matter and dilaton energy density, 

1949: in the presence of the non-equilibrium contributions:

1950: \begin{eqnarray}\label{eqsmotion2}  

1951: &&\frac{d\rho_\Phi}{dt_E} + 3H(\rho_\Phi + p_\Phi) 

1952: + \frac{1}{2}

1953: \frac{d\Phi}{dt_E}[\Psi'(\Phi)(\rho - 3p) + \sigma + \sigma_\phi ] =0~, 

1954: \nonumber \\ 

1955: && \frac{d\rho}{dt_E} + 3H(\rho + p) 

1956: - \frac{1}{2}\frac{d\Phi}{dt_E}[\Psi'(\Phi)(\rho - 3p) + \sigma + \sigma_\phi ] 

1957: + \left(\frac{d}{dt_E} + 3H\right){\cal J}_{00} + 3Ha^{-2}(t_E){\cal J}_{ii} 

1958: = 0~, \nonumber \\

1959: \end{eqnarray}

1960: with the values of ${\cal J}_{00}$ and ${\cal J}_{ii}$ (common for all

1961: $i=1,2,3$ in our case) given by (\ref{jcaldef}),(\ref{liouvmatter}). 

1962: As we see from (\ref{eqsmotion2}) the covariant 

1963: conservation of the matter stress tensor (the first three terms in the 

1964: second of eqs. (\ref{eqsmotion2})) breaks down, due not only to the presence

1965: of a dilaton field, but also to the off-shell

1966: Liouville contributions given by the ${\cal J}$-dependent terms, 

1967: which express the non-equilibrium nature of the

1968: Liouville cosmology. 

1969: 

1970: To solve (\ref{eqsmotion2}) in the various epochs of the 

1971: Universe, it is convenient first to split the energy density of matter

1972: as well as the function $\sigma$ 

1973: into radiation $\rho_r$, baryonic $\rho_b$ and dark-matter $\rho_d$ 

1974: components, and to use the simple equation of state (\ref{eqnstate})

1975: for the dilaton fluid: 

1976: \begin{eqnarray}\label{step1} 

1977: && \rho = \rho_r + \rho_b + \rho_d \equiv \rho_r  + \rho_m~, \nonumber \\

1978: &&\sigma = \sigma_r + \sigma_b + \sigma_d \equiv \sigma_r + \sigma_m, 

1979: \nonumber \\ 

1980: && p_b = p_d =0, \quad p_r = \frac{1}{3}\rho_r,  \quad 

1981: p_\Phi = w_\Phi \rho_\Phi~.

1982: \end{eqnarray} 

1983: Using (\ref{step1}), one can split the matter evolution equation (second 

1984: of eqs. (\ref{eqsmotion2})) into various components, which for an expanding 

1985: Universe can be cast in the form:

1986: \begin{eqnarray} 

1987: && \frac{d\rho_r}{d\chi} + 4\rho_r 

1988: - \frac{1}{2}\frac{d\Phi}{d\chi}[\sigma_r + \sigma_\phi]+

1989: \frac{d}{d\chi}{\cal J}_{00} + 3({\cal J}_{00} + a^{-2}(t_E){\cal J}_{ii}) =0~, 

1990: \nonumber \\

1991: && \frac{d\rho_A}{d\chi} + 3\rho_A 

1992: - \frac{1}{2}\frac{d\Phi}{d\chi}[\Psi '(\phi)\rho_A + \sigma_A + \sigma_\phi]+

1993: \frac{d}{d\chi}{\cal J}_{00} + 3({\cal J}_{00} + a^{-2}(t_E){\cal J}_{ii}) =0~, 

1994: \quad A=b,d~, \nonumber \\

1995: && \frac{d\rho_\Phi}{d\chi} + 3(1 + w_\Phi)\rho_\Phi 

1996: + \frac{1}{2}\frac{d\Phi}{d\chi}[\Psi '(\phi)\rho_m + \sigma + \sigma_\phi]=0~,

1997: \label{components}

1998: \end{eqnarray} 

1999: with 

2000: \begin{equation}

2001: \chi ={\rm ln}(a/a_{\rm init})=-{\rm ln}(1 + z) + {\rm ln}(a(0)/a_{\rm init}), 

2002: \end{equation}

2003: where $z$ is the redshift, 

2004: $a_{\rm init}$ is an initial scale, and $a(0)$ is the present 

2005: value of the scale factor, evaluated at redshift zero. 

2006: 

2007: Solving the above equations rigorously is a complicated task, and depends

2008: on the details of the matter theory.  In general, one may

2009: relate~\cite{gasperini} the various $\sigma_i$, $i=r,b,d$ with the

2010: corresponding energy densities $\rho_i$, through proportionality factors

2011: that depend on the dilaton field $\Phi$.  However, for our purposes we may

2012: assume that at the current era of the Universe's evolution the dark matter

2013: component dominates over ordinary matter and that the dilaton is

2014: approximated by its logarithmic evolution (\ref{dil2}) in cosmic

2015: Einstein-frame time.  We also assume that the current scale factor is also

2016: approximately given by the expression (\ref{einstmetr}). These assumptions

2017: guarantee that the relation (\ref{important}) between the (square of) the

2018: string coupling and the acceleration of the Universe is valid today.

2019: 

2020: Using the familiar (model-independent) relation of the scale factor 

2021: with the redshift $z$:   

2022: $a(z)= a(0)(1 + z)^{-1}$, we may then determine the region of 

2023: $z$ for which the approximation (\ref{einstmetr}) is consistent.

2024: Recalling that in our model the current era of the Universe  

2025: is defined by the relations (\ref{condition}),(\ref{decel4}), 

2026: with $\beta^2 = -1/q(z=0)$ where $q(z=0)$ is the acceleration of the 

2027: Universe today at $z=0$, it is straightforward to arrive at:

2028: \begin{equation}

2029: a(0) = (F_1/\gamma)(1 + \beta^2)^{1/2}~. 

2030: \label{a0def}

2031: \end{equation}

2032: Making the (wrong) hypothesis 

2033: that the formula for (\ref{einstmetr}) is valid all the way down to $t_E = 0$,

2034: we would then find that $z$ should lie in the region:

2035: \begin{equation}

2036: 0 < z < z_{\rm init}~; \qquad z_{\rm init} = \sqrt{1 + \beta^2} - 1~,  

2037: \label{zregion}

2038: \end{equation}

2039: in order that the form (\ref{einstmetr}) of the space-time metric 

2040: be valid. For $q(z=0)=\beta^{-2} = -0.57$ (c.f., (\ref{qval})), we 

2041: would then have $z_{\rm init} \simeq 0.66$. 

2042:  

2043: We assume $\sigma_d \simeq  \eta  \rho_d$ for the dark matter, 

2044: with $\eta $ an approximately constant 

2045: proportionality factor $\eta  = {\cal O}(1)$ 

2046: for the present and future eras. 

2047: These assumptions lead to simplifications of some of the equations.

2048: With $\Psi \simeq 2\Phi$ (the string tree-level 

2049: approximation), one obtains after some elementary 

2050: algebraic manipulations:

2051: \begin{eqnarray} 

2052: && {\cal J}_{00}  \simeq  -\frac{7H}{t_E} + H^2(1-q) \simeq 

2053: -\frac{6F_1^2}{a^2(t_E)}~,

2054: ~~

2055: {\cal J}_{ii}  \simeq  a^2(t_E)\left(\frac{H}{t_E} -H^2(1-q)\right) 

2056: \simeq 0~, ~i=1,2,3 , \nonumber \\

2057: && \frac{d}{d\chi}{\cal J}_{00} + 3\left({\cal J}_{00} + a^{-2}(t_E){\cal J}_{ii}\right) \simeq 

2058:  -\frac{6F_1^2}{a^2(t_E)} = {\cal J}_{00}~ \rightarrow ~|{\cal J}_{00}| \simeq

2059: \left(\frac{a_{\rm init}}{a(t_E)}\right)^2 = 

2060: \left(\frac{a_{\rm init}}{a(0)}\right)^2(1 + z)^2~, \nonumber \\

2061: \label{jcalvalues}

2062: \end{eqnarray}

2063: for the configuration (\ref{dil2}), (\ref{einstmetr}), (\ref{hubble2}) and

2064: (\ref{decel4}), assumed to characterize the current era of the Universe. 

2065: 

2066: Equation (\ref{components}) yields, then, for the dark matter 

2067: energy density, (which is assumed  to dominate the matter sector: $\rho_m 

2068: \simeq \rho_d$, $\sigma \simeq \sigma_d$):

2069: \begin{equation}\label{darkmatter} 

2070: \frac{d\rho_d}{d\chi} + 3\rho_d + 

2071: \frac{d\rho_\Phi}{d\chi} + 3(1 + w_\Phi)\rho_\Phi + {\cal J}_{00} = 0~.

2072: \end{equation} 

2073: This is a rather complicated equation to solve in 

2074: general~\footnote{Notice that most of the

2075: complications arise from the presence of the 

2076: off-shell Liouville modifications ${\cal J}_{00}$.

2077: In their absence, i.e., in `conventional' dilaton cosmologies, 

2078: one can solve this equation straightforwardly and 

2079: obtain the standard scaling for the various energy-density 

2080: components $\rho_d \sim \rho_d^0 a^{-3}$, 

2081: $\rho_\Phi \sim \rho_\Phi^0 a^{-3(1 + w_\Phi)}$, 

2082: with $\rho_d^0 + \rho_\Phi^0 \simeq 1$ 

2083: in the case of dominant dark matter.

2084: This is no longer true when ${\cal J}_{00} \ne 0$, and, as we shall

2085: see below, one obtains in that case 

2086: a mixed scaling for the matter energy density.}.

2087: Its solution, in conjunction with the rest of eqs.~(\ref{eqsmotion}),

2088: will provide a 

2089: scaling for the energy densities of matter

2090: and dark energy with the scale factor, which is modified in general. 

2091: 

2092: A simplification can occur, however, if one 

2093: concentrates on the epoch of large cosmic time (present era),

2094: and uses the asymptotic behaviour of the dilaton dark energy density 

2095: $\rho_\Phi$, dictated by (\ref{potendilaton}), 

2096: (\ref{dilpotkin})~\footnote{Notice that, despite the $a^{-2}$ scaling of 

2097: $\rho_\Phi$ and the off-shell Liouville term ${\cal J}_{00}$,

2098: none of these contributions is equivalent to a (negative) 

2099: curvature contribution. This is due to the fact that 

2100: the dilaton dark energy and the off-shell Liouville modifications

2101: enter the relevant dynamical equations (\ref{eqsmotion}) in a different

2102: manner than the curvature term. This 

2103: is consistent with the fact that our brane/string Universe is spatially 

2104: flat by construction.}, i.e., $\rho_\Phi\sim \rho^0_{\Phi} a^{-2}(t_E)$. 

2105: Assuming then mixed scaling behaviours

2106: \begin{equation}

2107: \rho_d \sim \rho_{\rm dust}^0a^{-3} + \rho_{\rm exotic}^0 a^{-2}~, 

2108: \label{mixedscaling}

2109: \end{equation}

2110: where the first term is compatible with dust

2111: properties, and the second expresses the 

2112: entanglement with the off-shell Liouville environment, 

2113: and the value (\ref{eqnstatedilnum3}) for the 

2114: equation of state,  

2115: in agreement  

2116: with recent WMAP data~\cite{wmap,steinhardt}, 

2117: we observe that (\ref{darkmatter}) 

2118: is  satisfied, provided  

2119: \begin{equation}

2120:  \rho_{\rm exotic}^0 = 6F_1^2 + 1.46\rho^0_{\Phi}~.

2121: \label{rhoexotic}

2122: \end{equation}

2123: Note that, for the model of \cite{dgmpp}), 

2124: $\rho^0_\Phi$ is in most cases also of 

2125: order $F_1^2$ (c.f., (\ref{dilpotkin}),

2126: which in turn is of the order of the square of the 

2127: asymptotic central charge (\ref{f1q0}). The latter can 

2128: be very small, e.g., of order

2129: $10^ {-60}$ in string units, in 

2130: models~\cite{gravanis} involving compactification on magnetized

2131: tori (c.f. (\ref{sigmamodelq0}), (\ref{horder}) below), which 

2132: guarantees compatibility of the order of the 

2133: magnetically-induced target-space 

2134: supersymmetry breaking with realistic 

2135: phenomenological considerations. In this way, $\rho_{\rm exotic}^0$ 

2136: can be very small, 

2137: and hence both terms in $\rho_d$ may be of comparable magnitude today.

2138: Specifically, from (\ref{dilpotkin}) it follows that 

2139: $\rho_\Phi \simeq 3.78 \frac{F_1^2}{a^2}$, which implies that 

2140: $\rho_\Phi^0 \simeq 3.78F_1^2$. Thus, (\ref{rhoexotic})

2141: would yield in that case: 

2142: \begin{equation}

2143: \rho_{\rm exotic}^0 \simeq 3.25\rho_\Phi^0~.

2144: \label{rhoexotic2}

2145: \end{equation}

2146: Thus, we may write for the (dark) matter energy density today

2147: \begin{equation}

2148: \rho_d^0 \simeq \rho_{\rm dust}^0  + \lambda \rho_\Phi^0~,~~ 0 < 

2149: \lambda ={\cal O}(1-10) .

2150: \label{lambdadef}

2151: \end{equation}

2152: We stress once

2153: more that this mixed scaling in the matter energy density 

2154: is due not only to the entanglement with the dilaton, 

2155: but also to the non-trivial r\^ole of the off-shell Liouville

2156: ${\cal J}$-dependent contributions. 

2157: This is an exclusive

2158: feature of our non-equilibrium Liouville-string 

2159: approach to cosmology~\cite{emn,emn04,dgmpp}, which does not apply in  

2160: conventional on-shell treatments~\cite{damour,gasperini}. 

2161: It has its roots in viewing target time as a 

2162: world-sheet renormalization-group dynamical scale in non-critical

2163: string theory, which is a cornerstone of our approach. 

2164: 

2165: As for the dilaton, a form linear in the $\sigma$-model frame is assumed

2166: throughout, which in turn determines the time dependences of the

2167: quantities $\sigma, \sigma_\phi$ via the respective dilaton equation

2168: (\ref{eqsmotion}).  Since the form of matter action is not in general

2169: fully known in our generic low-energy considerations, and depends on the

2170: details of the underlying microscopic string/brane model, we do not

2171: analyse this equation further here.  The existence of self-consistent

2172: solutions to the graviton equations (\ref{eqsmotion})  including matter

2173: and radiation, inferred by the above analysis, justifies \emph{ a

2174: posteriori} our assumption that the important relation (\ref{important}),

2175: which is based on the solution~\cite{emn04,dgmpp} (\ref{einstmetr2}),

2176: (\ref{einsteindil})  for the space-time metric and the dilaton fields,

2177: survives the inclusion of matter.

2178:  

2179: We now present the formalism for fitting our 

2180: Liouville Cosmology to cosmological data in a rather

2181: model-independent way. Consider the first of the Einstein

2182: equations (\ref{eqsmotion}) for our Liouville cosmology. In the present

2183: era, we may assume the following asymptotic behaviour for the Liouville

2184: part ${\cal J}$ (c.f. (\ref{jcalvalues})): ${\cal J}_{00} \simeq

2185: -6H^2(1-q)$. From the first of eqs.~(\ref{eqsmotion}), we may then

2186: conclude that, as a result of the Liouville out-of-equilibrium

2187: contributions, the \emph{critical} total mass (energy)  density of the

2188: fluid, $\rho_c$ required to have a spatially flat Universe, is no longer

2189: $6H^2$, as in the conventional on-shell Einstein cosmologies, but

2190: \begin{equation}

2191: \rho_c = 6H^2(2 - q).

2192: \label{criticaldens}

2193: \end{equation}

2194: One may then define modified $\Omega'_i$ fractions:

2195: \begin{equation}

2196: \Omega' \equiv \frac{\rho_i}{\rho_c} = \frac{\rho_i}{6H^2(2 - q)}~,~~ 

2197: i={\rm matter},~{\rm dilaton}~\Phi~{\rm dark~energy} ~\emph{etc.}.

2198: \label{newomegas}

2199: \end{equation}

2200: With this definition the first of equations (\ref{eqsmotion}) 

2201: would imply the standard relation for a spatially flat Universe today:

2202: \begin{equation}\label{sumomegas}

2203: \Omega'_{\rm Matter} + \Omega'_\Phi = 1~.

2204: \end{equation}

2205: Notice that the critical density (\ref{criticaldens}) scales with 

2206: $a = a(0)(1 + z)^{-1}$ as:

2207: \begin{equation}

2208: \rho_c(z) = \frac{6F_1^2}{a^2}\left( 2 - \xi ( 1 + z)^2\right)~, 

2209: ~~\xi \equiv \frac{F_1^2}{\gamma ^2 a^2(0)} = \frac{1}{1+ \beta^2}~, 

2210: \beta^2 = -1/q(z=0)~, 

2211: \label{rcscale}

2212: \end{equation}

2213: for the (rather generic) string model of~\cite{dgmpp} used here,

2214: where we took into account (\ref{a0def}). 

2215: The reader should also recall that $0 < z < 0.66$  for the validity of the

2216: approximations leading to the above analysis. 

2217: Thus, we have the following scaling with the redshift:

2218: \begin{eqnarray}

2219: \Omega'_\Phi (z) &=& \frac{\rho_\Phi^0}{6F_1^2 (2 - \xi (1 + z)^2)}

2220:  \equiv {\Omega'_\Phi}^{0}\frac{2-\xi}{2-\xi(1 +z)^2}~,

2221: \nonumber \\

2222: \Omega'_{\rm Matter} & = & \frac{1}{6F_1^2(2 - \xi (1 + z)^2)}\left(

2223: \frac{\rho_{\rm dust}^0}{1+z} + \lambda \rho_\Phi^0 \right)~,~~

2224: \label{omegaprimescaling}

2225: \end{eqnarray}

2226: where 

2227: $\lambda = {\cal O}(1-10)$ depending on model details, and  

2228: ${\Omega'_\Phi}^{0}$denotes the corresponding quantity today, i.e. at $z =0$. 

2229: 

2230: Alternatively, one may use the first of eqs. (\ref{eqsmotion}) 

2231: to express the critical density in terms 

2232: of the various energy-density components instead of the parameter $\xi$.

2233: This allows a more convenient 

2234: model-independent formalism to be used for comparison 

2235: with data. In this way,  

2236: using the mixed scaling  (\ref{mixedscaling}) for the matter sector 

2237: (including the dominant dark matter) of 

2238: our spatially flat Universe,  

2239: $\rho_{\rm Matter} = \rho_{\rm dust}^0 (1 + z)^3 + 

2240: \rho_{\rm exotic}^0(1 + z)^2$ with $\rho_{\rm dust}^0 + \rho_{\rm 

2241: exotic}^0 

2242: + \rho_\Phi^0 = \rho^0_c$, one obtains 

2243: the following scaling of $\Omega_i'$ with the redshift $z$:

2244: \begin{eqnarray}\label{scalings}

2245:  \Omega'_\Phi (z) = \frac{1}{1 + \rho_{\rm Matter}/\rho_\Phi} &=& 

2246: \frac{1}{1 + \frac{{\Omega '}_{\rm dust}^0}{{\Omega '}_\Phi^0}(1 + z) 

2247: + \frac{{\Omega '}_{\rm exotic}^0}{{\Omega '}_\Phi^0}}~, 

2248: \nonumber \\

2249: \Omega'_{\rm Matter} (z) = 1 -  \Omega'_\Phi (z) &=& 

2250: \frac{1}{1 + \frac{1}{\frac{{\Omega '}_{\rm dust}^0}{{\Omega '}_\Phi^0}

2251: (1 + z) + \frac{{\Omega '}_{\rm exotic}^0}{{\Omega '}_\Phi^0}}}~. 

2252: \end{eqnarray} 

2253: These expressions can be used to fit the astrophysical data and derive

2254: values for the cosmological parameters of our Q-cosmology model. 

2255: 

2256: Since the dilaton, matter and radiation energy densities scale

2257: differently, there is a (past) era of this Liouville Universe,

2258: corresponding to redshifts larger than a critical value, $z > z^*$, in

2259: which matter effects dominate over the dilaton dark energy, leading to a

2260: decelerating phase of the Universe.  In fact, such a past early era when

2261: there was deceleration of the Universe was present also in the model

2262: of~\cite{dgmpp}, even in the purely gravitational and moduli sector. Such

2263: a feature is simply pronounced by the inclusion of matter, since the

2264: latter feels the attractive feature of gravity. The

2265: critical $z^*$ is shifted from an early era in the purely gravitational

2266: case of~\cite{dgmpp}) towards the current epoch: $z^* \to \sim {\cal

2267: O}(1)$, as a result of the inclusion of matter in the model. The past

2268: deceleration in our Universe is a feature confirmed by astrophysical

2269: data~\cite{wmap,deceldata}, which indicate a value $z^* \sim 1/2$.  

2270: However, our model is too generic, at this stage, to claim a specific

2271: prediction for $z^*$.

2272: 

2273: \section{Concrete Non-critical String Examples: Colliding Branes}

2274: 

2275: The above considerations are rather generic for models which relax

2276: asymptotically to the linear-dilaton conformal field theory solutions

2277: of~\cite{aben}, and from this point of view are physically interesting. We

2278: have not yet specified the microscopic theory underlying the deviation

2279: from criticality.  For this purpose, one needs specific examples of such

2280: deviations from the conformal invariant points in string theory space. One

2281: such example with physically interesting consequences is provided by a

2282: colliding-brane-world scenario, in which the Liouville string $\sigma$

2283: model describes stringy excitations on the brane worlds for relatively

2284: long times after the collision, so that string perturbation theory is

2285: valid. This Section is devoted to a detailed discussion of such a

2286: scenario~\cite{gravanis,emw}.

2287: 

2288: \subsection{Example I: Colliding Type-IIB Five-Branes}

2289: 

2290: We now concentrate on particular examples of the previous general

2291: scenario~\cite{emninfl}, in which the non-criticality is induced by the

2292: collision of two branes, as seen in Fig.~\ref{infla}. We first discuss the

2293: basic features of this scenario. For our purposes below we assume that the

2294: string scale is of the same order as the four-dimensional Planck scale.

2295: However, this is an assumption which can be relaxed in view of recent

2296: developments in strings with large compactification directions, as was

2297: mentioned in the Introduction.

2298: 

2299: \begin{figure}[htb]

2300: \begin{center}

2301: \epsfxsize=3in

2302: \bigskip

2303: \centerline{\epsffile{colbr2.eps}}

2304: \caption{{\it A scenario 

2305: in which the collision of two Type-II five-branes provides 

2306: inflation and a relaxation model for cosmological vacuum energy.

2307: \label{infla}}}

2308: \end{center} 

2309: \end{figure}

2310: 

2311: Following~\cite{gravanis}, we consider two five-branes of Type-II string

2312: theory, in which the extra two dimensions have been compactified on tori.

2313: On one of the branes (assumed to be the hidden world), the torus is

2314: magnetized with a field intensity ${\cal H}$.  Initially our world is

2315: compactified on a normal torus, without a magnetic field, and the two

2316: branes are assumed to be on a collision course with a small relative

2317: velocity $v \ll 1$ in the bulk, as illustrated in Fig.~\ref{infla}. The

2318: collision produces a non-equilibrium situation, which results in electric

2319: current transfer from the hidden brane to the visible one.  This causes

2320: the (adiabatic) emergence of a magnetic field in our world.

2321: 

2322: The instabilities associated with such magnetized-tori compactifications

2323: are not a problem in the context of the cosmological scenario discussed

2324: here. In fact, as discussed in~\cite{gravanis}, the collision may also

2325: produce decompactification of the extra toroidal dimensions at a rate much

2326: slower than any other rate in the problem. As discussed

2327: in~\cite{gravanis}, this guarantees asymptotic equilibrium and a proper

2328: definition of an $S$-matrix for the stringy excitations on the observable

2329: world. We come back at this issue at the end of this Section.

2330: 

2331: The collision of the two branes implies, for a short period afterwards

2332: while the branes are at most a few string scales apart, the exchange of

2333: open-string excitations stretching between the branes, where their ends

2334: are attached. As argued in~\cite{gravanis}, the exchanges of such pairs of

2335: open strings in Type-II string theory result in an excitation energy in

2336: the visible world. The latter may be estimated by computing the

2337: corresponding scattering amplitude of the two branes, using string-theory

2338: world-sheet methods~\cite{bachas}: the time integral for the relevant

2339: potential yields the scattering amplitude. Such estimates involve the

2340: computation of appropriate world-sheet annulus diagrams, due to the

2341: existence of open string pairs in Type-II string theory. This implies the

2342: presence of `spin factors' as proportionality constants in the scattering

2343: amplitudes, which are expressed in terms of Jacobi $\Theta$ functions. For

2344: the small brane velocities $v \ll 1$ we are considering here, the

2345: appropriate spin structures start at {\it quartic order } in $v$, for the

2346: case of identical branes, as a result of the mathematical properties of

2347: the Jacobi functions~\cite{bachas}. This in turn

2348: implies~\cite{gravanis,emw} that the resulting excitation energy on the

2349: brane world is of order $V = {\cal O}(v^4)$, which may be thought of as an

2350: initial (approximately constant) value of a {\it supercritical}

2351: central-charge deficit for the non-critical $\sigma$ model that describes

2352: stringy excitations in the observable world after the collision:

2353: \begin{equation}

2354: Q^2 = \left(\sqrt{\beta} v^2  + {\cal H}^2\right)^2 > 0,

2355: \label{initialdeficit}

2356: \end{equation} 

2357: where, in the model of~\cite{emw,brany}, the proportionality factor $\beta$, 

2358: computed using string amplitude computations, is of order

2359: \begin{equation} 

2360: \beta \sim 2\sqrt{3} \cdot 10^{-8} \cdot g_s~, 

2361: \label{betaorder}

2362: \end{equation} 

2363: with $g_s $ the string coupling, which is of order $g_s^2 \sim 0.5$ for

2364: interesting phenomenological models~\cite{gsw,ibanez}.  The

2365: supercriticality, i.e., the positive definiteness of the central charge

2366: deficit (\ref{initialdeficit}) of the model, is essential~\cite{aben} for 

2367: a

2368: time-like signature of the Liouville mode and hence its interpretation as

2369: target time.

2370: 

2371: At times long after the collision, the branes slow down and the central 

2372: charge deficit is no

2373: longer constant but relaxes with time $t$.  In the approach

2374: of~\cite{gravanis}, this relaxation has been computed by using world-sheet

2375: logarithmic conformal field theory methods~\cite{kogan}, taking into

2376: account recoil (in the bulk) of the observable-world brane and the

2377: identification of target time with the (zero mode of the) Liouville field.

2378: In that work it was assumed that the final equilibrium 

2379: value of the central-charge deficit was zero, i.e., the theory approached

2380: a critical string. This late-time varying deficit $Q^2(t)$ 

2381: scales with the target time (Liouville mode) as follows (in units 

2382: of the string scale $M_s$):

2383: \begin{equation} 

2384: Q^2 (t) \sim \frac{({\cal H}^2 + v^2)^2}{t^2} .

2385: \label{cosmoconst}

2386: \end{equation} 

2387: Some explanations are necessary at this point.

2388: In arriving at (\ref{cosmoconst}), one identifies 

2389: the world-sheet renormalization group scale ${\cal T} ={\rm ln}(L/a)^2$,

2390: where $(L/a)^2$ is the world-sheet area, which appears in the 

2391: Zamolodchikov $c$-theorem

2392: used to determine the rate of change of $Q$ with ${\cal T}$,  

2393: with the zero mode of a normalized 

2394: Liouville field $\phi_0$, such that $\phi_0 = Q{\cal T}$. This normalization

2395: guarantees a canonical kinetic term for the Liouville field in the 

2396: world-sheet action~\cite{ddk}. Thus, $\phi_0$ is identified

2397: with $-t$, where $t$ is the target time. This will always 

2398: be understood in what follows.

2399: 

2400: 

2401: On the other hand, in other models~\cite{dgmpp}

2402: that we discuss below, 

2403: the asymptotic value of the central-charge 

2404: deficit may not be zero, in the sense that the asymptotic theory 

2405: is that of a 

2406: dilaton field that is linear in time, with a Minkowski metric in the 

2407: $\sigma$-model frame~\cite{aben}. This theory is still a conformal

2408: model, but the central charge is a constant $Q_0$, and in fact the 

2409: dilaton is of the form $\Phi = Q_0 t + {\rm const}$, 

2410: where $t$ is the target time in the $\sigma$-model frame.

2411: Conformal invariance, as already mentioned previously, 

2412: suggests~\cite{aben}

2413: that $Q_0$ takes on 

2414: one of a {\it discrete} set of values, in the way explained

2415: in~\cite{aben}.

2416: In such a case, following the same method as in the $Q_0=0$ case 

2417: of \cite{gravanis},

2418: one arrives at the asymptotic form

2419: \begin{equation} 

2420: Q^2 (t) \sim Q_0^2 + {\cal O}\left(\frac{{\cal H}^2 + v^2)}{t}Q_0\right) 

2421: \label{cosmoconst3}

2422: \end{equation} 

2423: for large times $t$. 

2424: 

2425: \paragraph{}

2426: \begin{figure}[tb]

2427: \begin{center}

2428: \includegraphics[width=4cm]{asymm.eps}

2429: \end{center}

2430: \caption{\it A model for supersymmetric D-particle foam

2431: consisting of two stacks each of sixteen parallel coincident D8-branes, 

2432: with orientifold planes (thick dashed lines) attached to them~\cite{emw}.

2433: The space does not extend beyond the orientifold planes.

2434: The bulk region of ten-dimensional space in which the D8-branes 

2435: are embedded is punctured by D0-branes (D-particles, dark blobs). 

2436: The two parallel stacks are sufficiently far from each other

2437: that any Casimir contribution to the vacuum energy is negligible.

2438: If the branes are stationary,

2439: there is zero vacuum energy, and the configuration 

2440: is a consistent supersymmetric string vacuum. To obtain excitations

2441: corresponding to interesting cosmologies, one should move one (or more) of 

2442: the 

2443: branes from each 

2444: stack, let them collide (Big Bang), bounce back (inflation),

2445: and then relax to their original position, where they collide again

2446: with the remaining branes in each stack (exit from inflation, reheating).}

2447: \label{fig:nonchiral}

2448: \end{figure}

2449: 

2450: \subsection{Example II: Orientifold/Eight-Brane/D-particles 

2451: Colliding-Brane Model}

2452: 

2453: The colliding-brane model of~\cite{gravanis} 

2454: can be extended to incorporate 

2455: proper supersymmetric vacuum configurations 

2456: of string theory ~\cite{emw}. As illustrated in Fig.~\ref{fig:nonchiral},

2457: this model consists of two stacks of D8-branes with the same tension, separated

2458: by a distance $R$. The transverse bulk space is restricted to lie between

2459: two orientifold planes, and is populated by D-particles. It was shown

2460: in~\cite{emw} that, in the limit of static branes and D-particles, this

2461: configuration constitutes a zero vacuum-energy supersymmetric ground state

2462: of this brane theory. 

2463: 

2464: 

2465: The bulk low-energy effective theory in such configurations 

2466: is known to be the ten-dimensional  Type-IIA supergravity,

2467: whose bosonic part is given in the string frame by~\cite{polchinski}:

2468: \begin{equation}

2469: S = S_{NS} + S_R + S_{CS},

2470: \end{equation}

2471: where

2472: \begin{eqnarray}\label{actioniia}

2473: S_{NS} &=& \frac{1}{2\kappa^2_{10}}\int d^{10} x \sqrt{- G }

2474: e^{-2\Phi}\left( R + 4 |\nabla_\mu \Phi|^2 -\frac{1}{2}|

2475: H_3|^2

2476: \right),\\

2477: S_R &=& -\frac{1}{4\kappa^2_{10}}\int d^{10} x \sqrt{-

2478:   G}\left(|F_2|^2+|\tilde F_4|^2  \right),\\  

2479: S_{CS} &=& -\frac{1}{4\kappa^2_{10}}\int B_2 \wedge dC_3 \wedge dC_3 ,

2480: \end{eqnarray}

2481: in standard notation. 

2482: However, in order to incorporate eight-dimensional 

2483: branes, one needs actually a modified version of the Type-IIA

2484: supergravity, which we now proceed to describe briefly, along with its

2485: possible compactifications, since we are eventually interested 

2486: in four space-time dimensional theories of phenomenological interest.

2487: 

2488: \subsection{Dual Formulation of Type-IIA Supergravity}

2489: 

2490: Type-IIA string theory contains all even-dimension D-branes from zero to 

2491: eight dimensions $p$~\cite{polchinski}. D$p$-branes couple to R-R 

2492: $p+1$-forms, but the action

2493: for Type-IIA supergravity only contains 1-form (D0-brane) and 3-form

2494: (D2-brane) gauge potentials - how are the other D-branes incorporated

2495: into the action? In~\cite{Bergshoeff:2001pv}, a \emph{dual}

2496: formulation of Type-IIA supergravity was constructed, which

2497: contains higher-dimensional R-R potentials and hence allows objects

2498: like the D8-brane to be incorporated. The dual Type-IIA supergravity allows 

2499: for the construction of Type-IA supergravity, and has the action:

2500: \begin{align}\label{dualaction}

2501:  S_{\text{bulk}} =

2502:   & - \frac{1}{2\kappa_{10}^2}\int d^{10} x \sqrt{-g}

2503:     \Big\{

2504:     e^{-2\Phi} \big[

2505:     R\big(\omega(e)\big) + 4\big( \partial{\Phi} \big)^{2}

2506:     +\tfrac{1}{2} H \cdot H

2507:     -2\partial^{{\mu}}{\Phi} \CHI{1}_{{\mu}}

2508:         + H \cdot \CHI{3}+ \notag \\

2509:   & +2 \bar{{\psi}}_{{\mu}}{\Gamma}^{{\mu}{\nu}{\rho}}

2510:     {\nabla}_{{\nu}}{\psi}_{{\rho}}

2511:     -2 \bar{{\lambda}}{\Gamma}^{{\mu}}

2512:     {\nabla}_{{\mu}}{\lambda}

2513:     +4 \bar{{\lambda}} {\Gamma}^{{\mu}{\nu}}

2514:     {\nabla}_{{\mu}}{\psi}_{{\nu}}

2515:     \big]

2516:     +   \sum_{n=0,1,2}

2517:      \tfrac{1}{2} \G{2n} \cdot \G{2n}

2518:     + \G{2n} \cdot \PSI{2n}  +  \notag \\

2519:   & - \star \, \big[

2520:       \tfrac{1}{2} \, \G{4} \G{4} B

2521:     - \tfrac{1}{2} \, \G{2} \G{4} B^2

2522:     + \tfrac{1}{6} \, \G{2}{}^2 B^3

2523:     + \tfrac{1}{6} \, \G{0} \G{4} B^3 -\tfrac{1}{8} \, \G{0} \G{2} B^4+ \notag \\

2524:   & +\tfrac{1}{40} \, \G{0}{}^2 B^5

2525:     + {\bf e}^{- B} {\bf G}

2526:     d (\A{5} - \A{7} + \A{9}) \big] \Big\}  +\mbox{ quartic fermionic 

2527: terms}\,,

2528: \end{align}

2529: in conventional notation~\cite{Bergshoeff:2001pv}. However, this dual 

2530: formulation only

2531: describes branes of dimension 4, 6 and 8 because of the problem of

2532: consistently introducing all of the available R-R forms. A

2533: \emph{democratic} formulation was also constructed, which contained all

2534: potentials, but this version has no proper action~\cite{Bergshoeff:2001pv}. 

2535: Since the D-foam model of~\cite{emw} contains D8-branes, O8-planes and

2536: D0-branes, along with fundamental strings when the D0-branes are

2537: between an odd number of D8-branes, neither the dual nor the democratic

2538: action is appropriate. One needs a combined action, which has been

2539: constructed in~\cite{Kallosh:2001zc}: 

2540: \begin{align}\label{kalloshact} 

2541: S =  &   - \frac{1}{2\kappa_{10}^2}\int d^{10} x \sqrt{-G}

2542:     \Big\{  e^{-2\Phi} \big[ R  + 4\big(

2543:   \partial{\Phi}   \big)^{2} +\tfrac{1}{12} (H^{(3)})^2  \big] +

2544:   \tfrac{1}{2} (\G{0})^2  +  \tfrac{1}{2} (\G{2})^2 - \star  \big[

2545:     \G{0}  d \A{9}    \big] \Big\}        \non\\ 

2546:   & -T_8 (n-8) \int d^{10} x

2547:   \left(e^{-\Phi} \sqrt{|G_{(9)}|}+\alpha

2548:   \frac{1}{9!}\epsilon^{(9)} \A{9} \right) \left[\delta(x^9) -

2549:     \delta(x^9-\pi R) \right]  \non\\ 

2550:    -T_2 &\int d^{10} x 

2551:   \left[\left(e^{-\Phi}

2552:     \sqrt{-G_{tt}}-bA_t\right)+\frac{e^{-\phi}}{\sqrt{G_{zz}}}\left(\sqrt{-G_{tt} 

2553:     G_{zz}} - 

2554:     \frac{a}{2!}\epsilon^{\nu\lambda} B_{\nu\lambda}\right)\right] \left[\sum_k N_k

2555:     \delta^8 (\vec x-\vec x_k)  \right]  ,

2556: \end{align}

2557: where $\vec x$ denotes an eight-dimensional vector, 

2558: $\G{n}, A^{(n)}$ are appropriate gauge flux fields, $G$ denotes the

2559: ten-dimensional $\sigma$-model-frame target-space metric, 

2560: $G_{00}, G_{zz}$ are 

2561: the temporal and bulk components of this metric respectively,

2562: and $G_{(9)}$ is a nine-dimensional metric.

2563: The second line describes the D8-branes and orientifold planes

2564: and the third the combined action for the D0-brane and fundamental

2565: strings. These brane-bulk actions describe all of the dynamics relevant to 

2566: the branes of interest.

2567: 

2568: \subsection{Towards Realistic Compactifications}

2569: 

2570: To compactify  such an action, e.g., on $T^4/\mathbb{Z}_2$,

2571: the  fields which survive the orbifold projection must be

2572: determined. An example of this for D6-O6 branes is given 

2573: in~\cite{Villadoro:2005cu}. Once the remaining field content is found,

2574: calculation of the dimensionally-reduced  Bianchi identities then

2575: leads to the lower-dimensional effective potential and  corresponding

2576: superpotential. Instead of a normal Kaluza-Klein dimensional

2577: reduction, Scherk-Schwarz fluxes~\cite{Scherk:1979zr} can be 

2578: added~\cite{Dall'Agata:2005ff}, which have the advantage of allowing a

2579: greater range of vacua: see~\cite{ DeWolfe:2005uu, 

2580: Villadoro:2005cu,Camara:2005dc} and references 

2581: therein~\footnote{As we saw in the generic

2582: analysis of~\cite{dgmpp}, reviewed above, 

2583: such flux fields play an important r\^ole 

2584: in ensuring the stabilisation of large bulk dimensions.

2585: A similar scenario is envisaged for the compactified version

2586: of the eight-brane model of~\cite{emw}.}.     

2587: 

2588: The overall structure of Type~IA string theory/supergravity is

2589: $\mathcal{M}^{\hspace{1pt} 9} \times S^1/\mathbb{Z}_2 I_9\Omega$,

2590: corresponding to an orbifold of the Type-IIA theory with an orientifold

2591: projection in the ninth dimension. A realistic compactification would

2592: result in either a Randall-Sundrum type scenario~\cite{randal}, i.e., a

2593: 3-brane embedded in five dimensions, or a conventional intersecting

2594: D-brane model~\cite{Blumenhagen:2002wn}, with the unusual feature of using

2595: D8-branes instead of D6-branes~\cite{Honecker:2002hp}.

2596: For Randall-Sundrum-II (RS-II) scenarios, it has been 

2597: suggested~\cite{Bergshoeff:2001pv} that a metric product of $AdS_5$ and  some

2598: Euclidean 5-manifold would give 3-branes in 5-dimensional Minkowski

2599: space, with the bulk solution being uplifted 

2600: from~\cite{Freedman:1999gp}~\footnote{We recall that 

2601: the strong-coupling limit of Type-IA string theory is equivalent to the 

2602: solution of Horava and Witten (HW)~\cite{Horava:1996ma, Lukas:1998yy, 

2603: Stelle:1998yv}.}.

2604: 

2605: The range of compactification choices can be summarized as follows:

2606: 

2607: \begin{itemize}

2608: \item{Compactify D8-O8 on $AdS_5 \times \mathcal{M}^5$ to get an 

2609: RS-II scenario~\cite{Bergshoeff:2001pv},}

2610: \item{Compactify on appropriate torii to get an intersecting brane

2611: model~\cite{Honecker:2002hp},}

2612: \item{Compactify on K3 $\times S^d$, giving another intersecting brane

2613: model based upon Calabi-Yau manifolds~\cite{Blumenhagen:2002wn}.}

2614: \end{itemize}

2615: The important requirement is to obtain $D=4$, $\mathcal{N}=1$

2616: supersymmetry on the brane, which in the (compactified) model

2617: of~\cite{emw} would correspond to the static D-brane/D-particle

2618: configuration.  We recall that Type-IIA supergravity has 32

2619: supersymmetries~\cite{polchinski}. In the the model of~\cite{emw}, the

2620: bulk space has $\mathcal{N}=2$, and on the brane there

2621: is~\cite{Bergshoeff:2001pv} $\mathcal{N}=1$. It should be noted that, for

2622: simplicity in this case, we are assuming that all of the branes are

2623: located on the orientifolds, and not in the bulk. Toroidal

2624: compactification of supergravity does not break any supersymmetries, so

2625: compactification on $T^5$ would give $D=5$, $\mathcal{N}=4$ supersymmetry

2626: in the bulk, with $\mathcal{N}=2$ on the brane. Changing this to

2627: $T^5/\mathbb{Z}_2$ breaks half of the supersymmetries resulting in

2628: $\mathcal{N}=2$ in the bulk and $\mathcal{N}=1$ on the brane. More

2629: complicated compactifications would change the precise way in which the

2630: supersymmetries are broken, as in the example suggested

2631: by~\cite{Bergshoeff:2001pv}, where the metric is the product of $AdS_5$

2632: and a Euclidean 5-manifold. This would also result in $D=4$,

2633: $\mathcal{N}=1$ on the brane. Inspired by the analysis on the colliding

2634: five-brane model of~\cite{gravanis}, in which the five-branes were

2635: compactified on magnitized tori to yield three-brane worlds with broken

2636: supersymmetry, as a result of internal magnetic fields, it would be

2637: desirable to discuss similar magnetized compactifications for the

2638: eight-branes of~\cite{emw}. This may be subtle due to the

2639: presence of the orientifolds, but some progress has already been made in 

2640: this

2641: direction~\cite{dudas}.

2642: 

2643: In this work we deal no further with the important issue of

2644: compactification, but postpone a detailed analysis to a future

2645: publication. 

2646: 

2647: \subsection{Supersymmetry Breaking and Vacuum Energy in the \\

2648: Post-Inflationary Era}

2649: 

2650: We now discuss briefly issues related to the supersymmetry breaking that

2651: would result from brane motion in the model.

2652: 

2653: \subsubsection{Supersymmetry Breaking via Internal Magnetic Fields}

2654: 

2655: In the colliding-brane scenario of~\cite{brany}, which 

2656: uses the orientifold configuration of~\cite{emw} shown in

2657: Fig.~\ref{fig:nonchiral}), one may imagine 

2658: that the exit from inflation and the reheating phase

2659: corresponds to a second collision, when the moving brane world

2660: returns to its initial position

2661: and hits the original stack of branes again. In such a case the recoil 

2662: velocity 

2663: of the brane world vanishes, but one may still have a magnetic 

2664: field ${\cal H}$ on the brane world, corresponding to a contribution 

2665: to the four-dimensional 

2666: energy density on the brane of order ${\cal H}^2$, for compactification

2667: radii of the extra dimensions of order $M_s$. One may identify

2668: therefore 

2669: \begin{equation} 

2670: Q_0^2 \sim {\cal H}^2 

2671: \label{sigmamodelq0} 

2672: \end{equation}

2673: in the $\sigma$-model 

2674: frame, leading to a dilaton of the form $\Phi \sim {\cal H} t + {\rm 

2675: const}$.

2676: For consistency with the results of~\cite{aben}, one would then discretize

2677: ${\cal H}$ with one of the values dictated by conformal invariance

2678: in this asymptotic $\sigma$ model. 

2679: 

2680: An important issue that we have already mentioned, but would like to

2681: stress again, is that in the Einstein frame the constants in the

2682: expression for the dilaton are such that the dark-energy density (c.f.,

2683: (\ref{cosmoconst2})  below)  relaxes to zero with the cosmic time $t_E$ in

2684: the Einstein frame (c.f. (\ref{einstframe} below)  as $1/t_E^2$, in a

2685: manner independent of the magnitude of $Q_0$.  In this way, the magnitude

2686: of the supersymmetry breaking in target space induced by the presence of

2687: ${\cal H}$~\cite{bachas,gravanis} may be large enough to be of

2688: phenomenological interest, whilst the observed value of the vacuum energy

2689: may be acceptably small, as we now explain.

2690: 

2691: The reason why the magnetic field ${\cal H}$ in the extra 

2692: dimensions~\cite{gravanis}

2693: breaks target-space supersymmetry~\cite{bachas} is that bosons and 

2694: fermions on the brane worlds couple differently

2695: to ${\cal H}$.  This is nothing other than a Zeeman-type 

2696: energy-splitting effect. 

2697: In our problem, where the magnetic field is turned on

2698: adiabatically, the resulting mass difference between bosonic and fermionic

2699: string excitations is found to be~\cite{gravanis}:

2700: \begin{equation} 

2701: \Delta m^2_{\rm string} \sim 2q_e|{\cal H}|{\rm cosh}\left(\epsilon \varphi + 

2702: \epsilon t\right)\Sigma_{45},

2703: \label{masssplit}

2704: \end{equation} 

2705: where $q_e$ is the electric charge, 

2706: $\Sigma_{45}$ is a standard spin operator in the plane of the torus,

2707: and $\epsilon \to 0^+$ is the regulating parameter 

2708: of the Heaviside operator $\Theta_\epsilon (t) = -i\int_{-\infty}^\infty 

2709: \frac{d\omega}{\omega -i\epsilon}e^{i\omega t}$

2710: appearing in the D-brane recoil formalism~\cite{kogan}.  

2711: The dependence in (\ref{masssplit}) implies that the formalism selects 

2712: dynamically a 

2713: Liouville mode which flows opposite to the target time $\varphi = -t$,

2714: as mentioned earlier, as a result of minimization of the effective 

2715: field-theoretic potential of the various stringy excitations. 

2716: 

2717: 

2718: In the scenario of~\cite{gravanis}, where the 

2719: dilaton remains constant asymptotically in time, 

2720: the mass splitting (\ref{masssplit}) with $\varphi = -t$

2721: is the only contribution to supersymmetry breaking as far

2722: as excitations are concerned. Since in that scenario the

2723: dark energy in target space relaxes to zero asymptotically 

2724: (\ref{cosmoconst}) while the mass splittings remain 

2725: finite, provided ${\cal H}$ remains constant one 

2726: has a supersymmetry {\it obstruction}~\cite{witten},

2727: rather than breaking, on the brane world, 

2728: since the cosmological constant of the vacuum state is 

2729: still zero, as required by a supersymmetric theory, but the excitation

2730: spectrum is not supersymmetric. 

2731: By choosing appropriately 

2732: \begin{equation}

2733: q_e|{\cal H}| \sim 10^{-30} ~\quad {\rm (in~string-scale~units)}, 

2734: \label{horder}

2735: \end{equation}

2736: we may arrange for the 

2737: supersymmetry-breaking/obstruction 

2738: scale to be of the order of a few TeV. Such 

2739: contributions would therefore be significantly 

2740: subdominant, compared with the

2741: velocity contribution, in the expressions (\ref{initialdeficit}), 

2742: (\ref{cosmoconst}) during the inflationary era. (We recall 

2743: that phenomenological analyses

2744: such as those in~\cite{brany,emw}, 

2745: yield recoil velocities as large as   

2746: $v \sim {\cal O}(10^{-1})$ towards the end of inflation.)

2747: However, we note here that if the string scale is itself of the 

2748: order of a few TeV, then $q_e|{\cal H}|$ in 

2749: (\ref{masssplit}) may be chosen of order one in string units

2750: in order to reproduce supersymmetry breaking at TeV scales.

2751: 

2752: There is an issue with the scenario of~\cite{gravanis} concerning the

2753: recoil velocity of the brane worlds after inflation. In~\cite{gravanis} it

2754: was assumed that the brane worlds eventually stop moving in the bulk, as a

2755: result of gravitational radiation, i.e., emission of closed strings from

2756: the brane towards the bulk. This would imply that there were no

2757: velocity-dependent contributions to the supersymmetry breaking in the bulk

2758: asymptotically.  We stress that, if the branes are moving relative to each

2759: other, the asymptotic vacuum energies on the brane world are non-zero, but

2760: depend on some power of the recoil velocity (in the case of identical

2761: recoiling branes this is $v^4$~\cite{emw}),

2762: breaking the (bulk) supersymmetry.

2763: 

2764: On the other hand, in scenarios with an asymptotically linear dilaton

2765: one has~\cite{aben}, as a result of the presence of the background charge 

2766: $Q$, {\it tachyonic shifts} $-Q^2$ in the masses of {\it bosons}, while the

2767: fermion masses remain {\it unaffected}.

2768: Such shifts induce {\it additional} contributions to 

2769: supersymmetry-breaking

2770: mass splittings (\ref{masssplit}) asymptotically~\footnote{Note that, 

2771: since during inflation the dilaton remains constant,

2772: there are no extra shifts in the boson masses due to the central charge 

2773: deficit in that era.}:

2774: \begin{equation} 

2775: \Delta m^2_{\rm susy-br} \sim q_e|{\cal H}| + O\left({\cal H}^2\right),

2776: \label{masssplit2}

2777: \end{equation} 

2778: and in this type of breaking one has~\cite{bachas}

2779: ${\rm Str}m^2 =0$, where ${\rm Str}$ denotes the supertrace.

2780: 

2781: In the scenario of~\cite{brany}, one uses the supersymmetric vacuum

2782: configuration of Fig.~\ref{fig:nonchiral}, where one or more of the branes

2783: of one stack collide with branes of the other stack before returning to

2784: their original position, where they collide for a second time, and

2785: eventually stop. The end of the inflationary era in this framework

2786: corresponds to this {\it second collision}.  We assume for simplicity that

2787: there is only one collision between the Big Bang and the exit from

2788: inflation, where our brane world collides with its original stack of

2789: branes and stops.

2790: %More collisions would correspond to pre-big-bang histories of our 

2791: %Universe. 

2792: This second collision results in a phase transition

2793: and {\it reheats} the Universe, as a result of entropy production due to the 

2794: collision. This second collision is much milder than the initial one, 

2795: because the recoiling brane world 

2796: may lose energy not only via its collisions with D-particles

2797: in the foam, but also due to gravitational radiation, i.e., closed string 

2798: emission in the bulk. The precise mechanism for reheating is still open:

2799: one possible contribution is the gravitational collapse

2800: of the bulk D-particles in the model of~\cite{emw} into black holes,

2801: due to distortions of their populations following the second collision.

2802: Evaporation of such bulky black holes on the brane worlds would

2803: result in Hawking radiation, represented by open string excitations

2804: attached to the brane worlds, thereby contributing to reheating.

2805: 

2806: In such scenarios, the asymptotic value of $Q^2 \sim {\cal H}^2$, since

2807: there is no recoil velocity of the brane after the second collision. For

2808: the order of magnitude of the magnetic field chosen above (\ref{horder}),

2809: such contributions are negligible compared with the Zeeman mass splittings

2810: (\ref{masssplit}). Moreover, for such values of the magnetic field, the

2811: equilibrium central charge (\ref{sigmamodelq0}) is of order $Q_0^2 \sim

2812: 10^{-60}$ (in string units), and the value of $\gamma $ in such a Universe

2813: is compatible with the current-era condition (\ref{condition}), provided

2814: (c.f., (\ref{defA}),(\ref{defA2}), (\ref{largetwe})): $|C_5|e^{-5s_{02}}

2815: \sim 1 $ (in string units), which is a natural value for the flux field in

2816: the model of~\cite{dgmpp}.  This guarantees a present-era vacuum energy

2817: (\ref{cosmoconst}) of the observed order, compatible with a

2818: phenomenologically-viable scenario for supersymmetry-breaking mass

2819: splittings (\ref{masssplit2}). On the other hand, if the string scale is

2820: of the order of a few TeV, then the age of the Universe today is $t_E \sim

2821: 10^{44}$ in string units, and since $Q_0^2 \sim {\cal H}^2 \sim 1$ in

2822: these units, one needs very large five-dimensional fluxes

2823: $|C_5|e^{-5s_{02}} \sim 10^{44}$ to ensure the condition

2824: (\ref{condition}).

2825: 

2826: The basic features of the low-energy limit of the non-supersymmetric

2827: Type-0 string theory that we used in~\cite{brany} and in~\cite{dgmpp},

2828: can be extended appropriately to the supersymmetric brane/orientifold

2829: compactification model of~\cite{emw}, without affecting the basic

2830: characteristics of the model, such as the existence of one large extra

2831: dimension, the presence of flux bulk fields, tachyons and extra moduli

2832: fields which freeze out quickly, and play no r\^ole in the phenomenology

2833: of the Universe in the present era. As far as tachyons are concerned,

2834: these fields existed in Type-0 string theory as a result of the explicit

2835: breaking of supersymmetry due to projecting the partners out of the string

2836: spectrum. In the supersymmetric models of colliding-brane

2837: worlds~\cite{emw,brany}, the motion and collision of the brane worlds

2838: breaks supersymmetry explicitly, both on the branes and in the bulk, as a

2839: result of the non-zero relative velocities. This also results in tachyonic

2840: excitations in the string spectrum, reflecting the instability of the

2841: configuration. This instability is essential in cosmological situations,

2842: such as the one we encounter here. The same analysis as in~\cite{dgmpp}

2843: can then be performed for the bosonic sector of the low-energy field

2844: theory in this case, to demonstrate the existence of solutions of

2845: cosmological relevance, in which the tachyon fields decouple quickly,

2846: leading to a similar late-stage analysis and results like those in

2847: \cite{dgmpp,emn04}, as reviewed in the previous Section.

2848: 

2849: We would like to call the reader's attention to one final point. As

2850: mentioned above, magnetized toroidal compactifications are known to have

2851: Nielsen-Olesen instabilities~\cite{bachas}. It may well

2852: happen~\cite{gravanis}, therefore, that as a result of the collision(s) a

2853: {\it decompactification} process takes place at a rate slower than any

2854: other time scale in our physical Universe, which implies, however, that

2855: the compactification radius $R \to \infty$ asymptotically in cosmic time,

2856: whilst the magnetic field energy ${\cal H}^2R^{p}$, for $p$ compact

2857: dimensions on the brane worlds, remains {\it finite}. This would imply

2858: vanishing magnetic fields asymptotically, and hence restoration of

2859: supersymmetry.

2860: 

2861: However, the compactification on magnetized internal manifolds is not the

2862: only way for supersymmetry to be broken in such cosmologies. As already

2863: mentioned, in the models of~\cite{emw} the motion of the brane world

2864: constitutes another source of breaking of supersymmetry. Moroever, as we

2865: also discuss below, the thermalization of the bulk and brane worlds soon

2866: after the collision could in principle result in yet another (independent)

2867: contribution to supersymmetry breaking. However, the finite recoil

2868: velocity of the colliding brane world and the temperature will be related,

2869: and hence there will be only one independent type of supersymmetry

2870: breaking in the scenario of~\cite{emw}~\footnote{Notice that the model

2871: of~\cite{emw} does not involve compactification, and hence the

2872: considerations on phases with broken supersymmetry pertain to

2873: eight-dimensional brane worlds, moving in the ninth bulk dimension. Upon

2874: subsequent compactification it is possible to have additional sources of

2875: supersymmetry breaking, including the ones associated with possible

2876: internal magnetic fields, as discussed elsewhere.}.

2877: 

2878: \subsubsection{Moving Branes and Supersymmetry Breaking}

2879: 

2880: The colliding-brane scenario can be realized~\cite{brany} in this

2881: framework by allowing (at least one of) the D-branes to move, keeping the

2882: orientifold planes static. One may envisage a situation in which the two

2883: branes collide, at a certain moment in time corresponding to the Big Bang

2884: - a catastrophic cosmological event setting the beginning of observable

2885: time - and then bounce back. The width of the bulk region is assumed to be

2886: long enough that, after a sufficiently long time following the collision,

2887: the excitation energy on the observable brane world - which corresponds to

2888: the conformal charge deficit in a $\sigma$-model

2889: framework~\cite{gravanis,emw} - relaxes to tiny values.

2890: 

2891: It is expected that a ground state-configuration will be achieved when the

2892: branes reach the orientifold planes again (within stringy length

2893: uncertainties of order $\ell_s=1/M_s$, the string scale). In this picture,

2894: since observable time starts ticking after the collision, the question how

2895: the brane worlds started to move is merely philosophical or metaphysical.

2896: The collision results in a kind of phase transition, during which the

2897: system passes through a non-equilibrium phase, in which one loses the

2898: conformal symmetry of the stringy $\sigma$ model that describes

2899: perturbatively string excitations on the branes. At long times after the

2900: collision, the central charge deficit relaxes to zero~\cite{gravanis},

2901: indicating that the system approaches equilibrium again. The dark energy

2902: observed today may be the result of the fact that our world has not yet

2903: relaxed to this equilibrium value. Since the asymptotic ground state

2904: configuration has static D-branes and D-particles, and hence has zero

2905: vacuum energy as guaranteed by the exact conformal field theory

2906: construction of~\cite{emw,brany}, it avoids the fine-tuning problems in

2907: the model of~\cite{gravanis}.

2908: 

2909: 

2910: Thus, the bulk motion of either the D-branes or the

2911: D-particles~\footnote{The latter could arise from recoil effects following

2912: scattering with closed-string states propagating in the bulk.} results in

2913: non-zero `vacuum' (or, rather, `excitation') energy~\cite{emw}, and hence

2914: the \emph{breaking of target-space supersymmetry}, proportional to some

2915: power of the average (recoil) velocity squared, which depends on the

2916: precise string model used to described the (open) stringy matter

2917: excitations on the branes.  Sub-asymptotically, there are several

2918: contributions to the excitation energy of our brane world in this picture.

2919: One comes from the interaction of the brane world with nearby D-particles,

2920: i.e., those within distances at most of order ${\cal O}(\ell_s)$, as a

2921: result of open strings stretched between them.  The other contribution

2922: comes from the collision of the identical D-branes.

2923: 

2924: A detailed analysis, using world-sheet methods for the computation

2925: of the various potentials felt by the D-branes/D-particles in the 

2926: colliding-brane model of~\cite{emw}, yields two types of effective

2927: potentials. One is a potential in the bulk space,

2928: felt by closed-string excitations from the gravitational multiplet

2929: that are allowed to propagate in the bulk. The bulk potential 

2930: is given by: 

2931: \begin{equation}

2932: {\cal V}_{\rm sym} \simeq V_8\frac{(30R-64r)v^4}{2^{13}\pi^9{\alpha '}^5}-

2933: N\left(\frac{v}{\alpha '}\right)^{1/2},

2934: \label{symmpot1}

2935: \end{equation}

2936: where the distances $R, r$ are defined in Fig.~\ref{fig:nonchiral},

2937: $v$ is the recoil velocity of our brane world, and 

2938: $N$ is the number of D-particles near the moving brane 

2939: world, which are the only type of D-particles that contribute

2940: significantly to the potential~\cite{emw}. A symmetric configuration

2941: of branes has been considered in Fig.~\ref{fig:nonchiral} 

2942: for concreteness and simplicity. For a sufficiently dilute

2943: gas of nearby D-particles, one may assume that this latter contribution is

2944: the dominant one. In this case, one may ignore the D-particle/D-brane

2945: contributions to the vacuum energy, and hence apply the

2946: previous considerations on inflation, based on the ${\cal O}(v^4)$ central

2947: charge deficit, with $v$ the velocity of the brane world in the bulk.

2948: 

2949: 

2950: The other type of potential, 

2951: generated in the moving-brane scenario of~\cite{emw}, 

2952: is an effective potential felt by the brane world itself, as 

2953: a result of its interactions with the other branes

2954: and D-particles. This second type of potential

2955: is felt by the open-string excitations whose ends are attached to the 

2956: brane, which constitute the Standard Model matter and radiation, 

2957: living on the brane world. The brane potential is~\cite{emw}:

2958: \begin{equation}

2959: V_{\rm brane} =  -V_8\frac{31(R-2r)v^4}{2^{13}\pi^9{\alpha '}^5}~,

2960: \label{branepotopen1}

2961: \end{equation}

2962: where $r_1 = R - 2r$ denotes the relative separation of the branes

2963: in the symmetric situation of Fig.~\ref{fig:nonchiral} with $r_2~=~r$.

2964: Notice that the potential is negative, 

2965: which expresses the fact that 

2966: the brane world feels an attractive force towards its original stack,

2967: and the configuration is stabilized when $v \to 0$.

2968: In Section 4 we return to a physical interpretation of 

2969: the above potentials, which determine the various phases of 

2970: our early Q-cosmology. 

2971: 

2972: From the point of view of the low-energy bulk action (\ref{actioniia}) and

2973: (\ref{kalloshact}), the bulk potential (\ref{symmpot1})  would correspond

2974: to a non-zero contribution to the scalar potential of the Type-IIA

2975: supergravity theory, proportional to a central charge deficit:

2976: $e^{-2\Phi}Q^2$, expressing the non-criticality of the associated $\sigma$

2977: model describing bulk string excitations. The fact that the potential

2978: (\ref{symmpot1}) changes sign, depending on the value of $r$, will lead to

2979: a rich phase structure, as we discuss in Section 4. However, due to the

2980: fact that we consider here a brane excitation, the system does not sit at

2981: a global minimum of the potential, but rather in a local (metastable)

2982: extremum.  We return to this important point in Section 4, when we discuss

2983: the various phases of the bulk theory. As we show there, the compactified

2984: Type-IIA theory may not be characterized by such a global minimum as a

2985: result of purely stringy properties (lack of certain T-duality

2986: symmetries~\cite{kounnas}).

2987: 

2988: For the effective low-energy of the open-string excitations on the brane,

2989: a similar excitation `vacuum energy' is provided by the potential

2990: (\ref{branepotopen1}), but with subtleties because its value is always

2991: negative. As we discuss in Section 4, this may be interpreted as

2992: thermalization of the brane world, throughout the inflationary period and

2993: its exit phase (this mechanism is an alternative to the usual description

2994: of reheating). Moreover, the presence of matter on the brane world causes

2995: back-reaction onto the space-time, along the lines discussed earlier.

2996: Matter is assumed to satisfy classical equations in an effective

2997: four-dimensional supergravity field theory on the brane world. There are

2998: of course subtleties associated with specific compactification scenarios,

2999: which we do not discuss here.

3000: 

3001: A final comment concerns the r\^ole of the D-particles in the above

3002: models. The presence of these space-time defects, which inevitably cross

3003: the D-branes as the latter move in the bulk, even if the D-particle

3004: defects are static initially, distorts slightly~\cite{papavass} the

3005: inflationary metric on the observable brane world at early times after the

3006: collision, during an era of approximately constant central charge deficit.

3007: However, this effect does not lead to significant qualitative changes.

3008: Moreover, the existence of D-particles on the branes affects the

3009: propagation of string matter on the branes, in the sense of modifying

3010: their dispersion relations by inducing local curvature in space-time, as a

3011: result of recoil following collisions with string matter. However, it was

3012: argued in~\cite{emnequiv} that only photons are susceptible to such

3013: effects in this scenario, due to the specific gauge properties of the

3014: membrane theory at hand.  The dispersion relations for chiral matter

3015: particles, or in general fields on the D-branes that transform

3016: non-trivially under the Standard Model gauge group, are protected by

3017: special gauge symmetries in string theory, and as such are not modified.

3018: 

3019: \section{Finite Temperature in the Liouville Framework} 

3020: 

3021: We now discuss thermalized strings in the context of our Liouville

3022: formalism, and describe the thermal phase diagram of our early Universe.

3023: 

3024: \subsection{Brane Collisions and Hot Universes} 

3025: 

3026: 

3027: In our colliding-brane scenario, each brane collision

3028: thermalizes the string excitation spectrum

3029: on the brane worlds and in the bulk, 

3030: as a result of the conversion of the kinetic energy

3031: of the moving branes into thermal energy. In the 

3032: scenario with two moving colliding branes of~\cite{emw} (c.f., 

3033: Fig.~\ref{fig:nonchiral}), the string excitations may be thermalized

3034: immediately after the collision. 

3035: Indeed, as the detailed computations of~\cite{emw} have shown,

3036: the effective potential of the configuration when the 

3037: two branes lie a distance $r_1 = R-2r$ apart (with the symbols as in 

3038: Fig.~\ref{fig:nonchiral}, restricting ourselves to the 

3039: symmetric case $r_2=r$ for simplicity) is given by (\ref{symmpot1}).

3040: 

3041: We observe from (\ref{symmpot1}) that for a {\it sufficiently dilute gas}

3042: of D-particles, the potential is positive for $r \ll R/2$.

3043: This implies that, for relatively 

3044: long times after the collision when the distance of our brane world

3045: from its original stack satisfies the above constraint $r \ll R/2$,  

3046: closed-string excitations in the bulk

3047: feel this positive vacuum energy, which means that 

3048: the corresponding $\sigma$ model is {\it supercritical}.

3049: It must therefore be dressed

3050: by a time-like Liouville field, which is eventually identified 

3051: with the target time. In fact, in the analysis of~\cite{gravanis},

3052: for reasons associated with the convergence of the 

3053: world-sheet path integral, we considered the initial time coordinate

3054: $X^0$ (before Liouville dressing) as {\it space-like} ({\it Euclidean time}).

3055: This was important, because dressing with a time-like Liouville

3056: field implied a $(D+1)$-dimensional 

3057: target-space metric (in our normalization here)~\cite{gravanis}:

3058: \begin{equation}

3059: ds^2_{D+1} = -2(d\varphi)^2 + (dX^0)^2 + d{\vec x}^2 .

3060: \end{equation}

3061: Upon the identification (\ref{liouvtime}) $\varphi = -t$, 

3062: where now $t=X^0$ is a Euclidean

3063: target time, one obtains a {\it Minkowski-signature} 

3064: $D$-dimensional space-time in a dynamical way. 

3065: Although in~\cite{gravanis} we viewed the use of Euclidean time merely

3066: as a mathematical peculiarity of the world-sheet path integral,

3067: it may be given a physical meaning in the context of 

3068: the colliding-brane scenarios, as follows. 

3069: 

3070: Assuming that the adiabatic analysis of~\cite{emw} is valid soon after the

3071: initial collision, and ignoring again the contributions from D-particles,

3072: assuming them sufficiently dilute, we observe that in that early epoch of

3073: the Universe the potential (\ref{symmpot}) is negative, since in that era

3074: $30R/64 < r < R/2$ (c.f., Fig.~\ref{fig:nonchiral}). The closed-string

3075: excitations find themselves described by a {\it subcritical} $\sigma$

3076: model, which can become critical upon Liouville dressing by a {\it

3077: space-like} Liouville mode. This correspond to {\it thermalization} as a

3078: result of the collision, during which the initial kinetic energy of the

3079: D-branes is transformed into thermal energy. In fact, if we assume that

3080: the initial relative velocity of the D-branes is of the same order as the

3081: recoil velocity, which in~\cite{emw,brany} was estimated to be of order

3082: $10^{-3} < v < 10^{-1}$ in units of $c=1$, then we observe that the

3083: induced temperature $\frac{1}{2}{\cal M} v^2 \sim k_B T$, where ${\cal M}$

3084: is the D-brane mass, could be close to the Hagedorn temperature of the

3085: corresponding string theory, $T \simeq T_H \sim \frac{1}{2\pi

3086: \sqrt{2\alpha '}}$ in order of magnitude~\footnote{Slight differences in

3087: the proportionality factors occur between the various string theories.}.  

3088: Thus, during the collision phase, the branes and the bulk (closed) string

3089: excitations find themselves at a finite (high) temperature.

3090: 

3091: It is interesting to describe the stringy excitations under such

3092: conditions.  In what follows we review the r\^ole of the Liouville

3093: formalism in describing generic strings at finite temperature. We commence

3094: our analysis with the heterotic string case, which is the simplest and

3095: among the most interesting cases for phenomenology.

3096: 

3097: \subsection{Liouville Approach to Finite-Temperature Strings: the Case of

3098: Heterotic Strings}

3099: 

3100: Historically~\cite{kounnas}, there has been interest in obtaining a

3101: description of a hot, stable phase of strings at temperatures beyond the

3102: Hagedorn phase transition at $T_H \sim 1/2\pi \sqrt{\alpha '}$. In our

3103: case, we are interested in the description of strings much below such high

3104: temperatures. However, it is instructive for our purposes to review first

3105: the Hagedorn phase, as studied for heterotic strings in~\cite{kounnas}. We

3106: then return to the brane model of~\cite{emw}, characterized by a bulk

3107: low-energy Type-IIA effective supergravity theory in the next Section.

3108: 

3109: The easiest approach to discussing strings at finite temperature $T$ is to

3110: compactify the time direction on a circle of radius $R =1/\pi T$ and

3111: discuss the mass spectrum of the winding modes of the string. Using

3112: appropriate T-dualities the authors of~\cite{kounnas} have discussed the

3113: instabilities arising from the fact that some of these T-winding modes of

3114: the string become tachyonic above the Hagedorn temperature of a gas of

3115: strings.  This defines the high-temperature phase of strings, and in our

3116: case we could identify it with the epoch soon after the initial collision,

3117: where the separation between the branes is small.

3118: 

3119: The presence of a non-zero temperature leads in general to additional

3120: contributions to supersymmetry breaking beyond the ones discussed

3121: so far. An important result~\cite{rostant} in the context of strings

3122: is that $D$-dimensional superstrings at finite temperature look like 

3123: $(D-1)$-dimensional superstrings with {\it spontaneously} broken 

3124: supersymmetry.  Restricting our attention to the (bulk) closed string 

3125: winding sector,

3126: this observation implies~\cite{kounnas} that 

3127: the corresponding low-energy effective 

3128: supergravity field theory is characterized by a 

3129: a non-zero (negative) value of the 

3130: (global) minimum of its corresponding  scalar potential, 

3131: proportional to the 

3132: square of the gravitino mass:

3133: \begin{equation}

3134: V_{\rm min} = -2m_{3/2}^2\kappa^{-2} =-1/2S\kappa^{-2}~,

3135: \label{minscalar}

3136: \end{equation} 

3137: where $S=e^{-\Phi}$ denotes the dilaton field in the supergravity multiplet,

3138: and $\kappa $ is the (ten-dimensional) gravitational constant.  

3139: From a stringy $\sigma$-model viewpoint, such a minimum

3140: corresponds to the propagation of strings in a space-time

3141: with a tree-level non-constant cosmological term, providing a

3142: runaway-dilaton potential.

3143: A detailed analysis of heterotic superstrings in

3144: high-temperature phases has been performed in~\cite{kounnas}, where

3145: it was shown that there exists a conformal field theory description

3146: of this high-temperature phase, corresponding to a $\sigma$ model 

3147: where the central charge has been lowered by four units.

3148: 

3149: Indeed, the conformal field theory is nothing other than the

3150: strongly-coupled Liouville theory~\cite{ddk}, which is not yet very well

3151: understood. This Liouville conformal theory corresponds in target space to

3152: a {\it subcritical} superstring whose $\sigma$-model-frame metric is the

3153: flat Minkowski one: $G^\sigma_{\mu\nu}=\eta_{\mu\nu}$, with a dilaton

3154: linear in a {\it space-like} coordinate, playing the r\^ole of the

3155: Euclidean compactified time $\Phi = QX^0$, where $Q$ is a background

3156: charge. The space-like nature of the Liouville mode (temperature) is due

3157: to the fact that there is a central-charge {\it deficit} and not a {\it

3158: surplus} as in the time-like Liouville case of~\cite{aben,emn} discussed

3159: in previous Sections. The physical metric, corresponding to a canonically

3160: normalized Einstein term in the effective action, is again given by

3161: (\ref{smodeinst}), and the corresponding target-space effective action by

3162: (\ref{effaction}).  From the cosmological term of this action, and its

3163: identification with the the minimum of the supergravity scalar potential

3164: (\ref{minscalar}), one can compute the central-charge deficit $\delta c

3165: \propto Q^2 <0$ of the conformal theory (paying particular attention to

3166: the appropriate normalization factors~\cite{kounnas}), essentially by

3167: identifying it with the numerical coefficient of $1/S$ in the expression

3168: (\ref{minscalar}):

3169: \begin{equation} 

3170: Q^2 = \frac{\delta c}{8\alpha '} = -\frac{1}{2\alpha '} < 0,

3171: \end{equation}

3172: implying a central charge deficit: $\delta c = -4$ for the superstring.

3173: Taking into account the fact that for flat $\sigma$-model-frame metric

3174: backgrounds $\delta c = D-10$, where $D$ is the space-time dimensionality

3175: of the free superstring, we therefore observe that in the high-temperature

3176: phase the thermalized closed-string system corresponds to a non-critical

3177: superstring living in 5+1 dimensions. In such a phase it was remarked

3178: in~\cite{kounnas} that five-branes condense. Such features may turn out to

3179: be quite important for cosmological model building. For instance, this

3180: would imply that in the original model of colliding five-branes

3181: of~\cite{gravanis}, immediately after the collision one would have

3182: condensation of the five-branes, which does not happen in the model

3183: of~\cite{emw}.

3184: 

3185: An important feature of such finite-temperature

3186: superstrings is the existence of a space-like supersymmetry

3187: at a perturbative level which characterizes the hot phase~\cite{kounnas}.

3188: Indeed, before Liouville dressing, the finite temperature 

3189: contributes to supersymmetry breaking mass shifts 

3190: between the bosonic ($M_B$) and fermionic ($M_F$) 

3191: excitations of the corresponding

3192: supergravity theory:

3193: \begin{equation} 

3194: (M_B)^2_{i\bar j} = (M_F)^2_{i\bar j} - m_{3/2}^2\delta_{i\bar j}

3195: \label{masshift}

3196: \end{equation}

3197: in the mass-matrix notation of~\cite{kounnas}.

3198: After the Liouville dressing by the space-like linear dilaton, the 

3199: fermion masses remain unaffected, but the 

3200: boson masses undergo mass shifts~\cite{aben}. However, this time,

3201: due to the subcriticality of the string, which is to be contrasted

3202: with the case of~\cite{aben}, the mass shifts are not tachyonic,

3203: but real:

3204: \begin{equation}\label{qgravino}

3205: \delta (M_B)^2 = Q^2 = m_{3/2}^2~, \quad \delta (M_F)^2 = 0 

3206: \end{equation}

3207: in our normalization. 

3208: These mass shifts are additive to (\ref{masshift}), which 

3209: implies that {\it at a perturbative level} supersymmetry

3210: is restored in this hot phase of strings. 

3211: In our case, therefore, this means that, at a perturbative level, 

3212: supersymmetry breaking is still be given by the 

3213: magnetic terms as described above. 

3214: 

3215: The supersymmetry is however broken, or rather {\it

3216: obstructed}~\cite{witten} at a {\it non-perturbative} level, due to the

3217: fact that masses in three space dimensions produce conical singularities,

3218: and as such they break supersymmetry at the level of the excitation

3219: spectrum, although the vacuum may still be supersymmetric.  Such

3220: non-perturbative breaking has been discussed explicitly in~\cite{kounnas},

3221: and we do not discuss it further here. We mention, though, that this

3222: corresponds to an instability of the high-temperature phase, because this

3223: breaking of supersymmetry produces tachyonic states.

3224: 

3225: The string must leave this unstable phase and re-enter a phase where such

3226: instabilities eventually disappear, and the string system relaxes to an

3227: equilibrium situation (with unbroken supersymmetry, modulo the effects of

3228: the magnetic field, if compactification on magnetized manifolds is

3229: considered).  We now discuss how this may be understood from a world-sheet

3230: view point, in the context of our cosmological model of colliding branes

3231: presented in~\cite{emw}.  Since the effective low-energy theory in the

3232: bulk in this model is Type-IIA supergravity, we first review a

3233: finite-temperature analysis of this special theory, with the aim of

3234: repeating the analysis of~\cite{kounnas} for this case. There are

3235: important physical differences, however, associated with the lack of

3236: global minima in Type-IIA theories, which we outline in due course.  We

3237: commence our analysis with a finite-temperature study of the effective

3238: Type-IIA supergravity theory, which characterizes the low-energy bulk

3239: dynamics in the model of~\cite{emw}.

3240: 

3241: \subsection{Type-IIA Supergravity at Finite Temperature}

3242: 

3243: As already mentioned, in~\cite{emw} we have a system of D8-branes and

3244: orientifolds, in the configuration known as Type-IA string theory. Between

3245: the two stacks of D8-branes, the bulk space corresponds to Type-IIA

3246: supergravity. When some of the D8-branes move into the bulk, the overall

3247: bulk potential induced by this motion can become negative~\cite{emw},

3248: which can be interpreted as the system moving into a finite-temperature

3249: phase. Finite-temperature field theory is realized by the Euclidean

3250: compactification of the time dimension, and its effects can be calculated

3251: using the Scherk-Schwarz mechanism~\cite{Scherk:1979zr}.

3252: 

3253: When one of the D8-branes from each stack moves into the bulk, there are

3254: two potentials which must be taken into account. First there is the bulk

3255: potential (\ref{symmpot}), describing the overall energy of the system,

3256: where the potential is positive as long as the distance between the brane

3257: and its originating stack, $r$, is less than $15R/32$. Secondly there is

3258: the potential on the moving brane itself (as we are dealing with a

3259: symmetric case, we consider the left-hand brane). This case is more

3260: complex and will be discussed later on.

3261: 

3262: As stressed above, the important result when considering-finite

3263: temperature supergravity is that $D$-dimensional superstrings at finite

3264: temperature look like ($D-1$)-dimensional superstrings with spontaneously

3265: broken supersymmetry~\cite{rostant,kounnas}. Thus, spontaneous

3266: supersymmetry breaking via the Scherk-Schwarz mechanism is equivalent to

3267: considering the system at finite temperature.  The Scherk-Schwarz

3268: mechanism~\cite{Scherk:1979zr} works by generalizing the standard

3269: dimensional reduction procedure~\cite{Antoniadis:1997wu}, in which all of

3270: the fields are taken to be independent of the compact coordinates.

3271: Instead, the fields are given a specific dependence on the internal

3272: coordinates of the compact manifold, namely twisting the boundary

3273: conditions of the compact dimensions by a global symmetry of the action.

3274: This twist induces a shift in the mass terms of the lower-dimensional

3275: fields.

3276: 

3277: In finite-temperature QFT~\cite{Atick:1988si}, bosons are

3278: periodic and fermions anti-periodic in the compact Euclidean time

3279: dimension~\cite{kounnas}:

3280: \begin{equation}

3281: \Phi(t+ 2L \pi R) = (-)^{La} \Phi(t),

3282: \end{equation}

3283: where for a $2\pi$ rotation $L=1$, and  $a=0,1$ for bosons and

3284: fermions respectively. The modular 

3285: invariance of Type~II string theory requires further constraints to be

3286: placed on the periodicity conditions~\cite{rostant, kounnas,Atick:1988si}:

3287: \begin{equation}

3288: (-)^{aL+bn}

3289: \end{equation}

3290: where $m,n$ are winding numbers and $a$ and $b$ are the fermionic spin

3291: structures along the world-sheet torus. The resulting shift in the

3292: lattice momenta along the compact coordinate is~\footnote{The following 

3293: discussion is taken from~\cite{kounnas}.}:

3294: \begin{equation}\label{momenta shift}

3295: p_{L,R} = \frac{m+a/2}{R}\pm \frac{nR}{2},

3296: \end{equation}

3297: with an additional sign factor of $(-)^{ab}$ which reverses the GSO

3298: projection in the odd winding-number sectors. By redefining $m$, $a$

3299: can be identified with the $D$-dimensional helicity operator ${\mathcal Q} =

3300: \mathbb{Z}+a/2$, so that $(D-1)$-dimensional thermal states are mapped

3301: to a supersymmetric theory on $S^1$ without the momentum shift

3302: (\ref{momenta shift}). The helicity vector $\vec{{\mathcal Q}} = 

3303: ({\mathcal Q}_L , {\mathcal Q}_R)$ is

3304: defined in terms of the left- and right-moving string helicities, and

3305: the vector $\vec{e}=(1,1)$ for the Type II string. The inner product

3306: is Lorentzian, $\vec{A} \cdot \vec{B} = A_L B_L - A_R B_R$. 

3307: 

3308: Thus there is a mapping between the $(D-1)$-dimensional supersymmetric

3309: theory with quantum numbers $(n,m,{\mathcal Q})$ and the $(D-1)$-dimensional

3310: thermal theory, which results in the quantum numbers of the

3311: supersymmetric theory being shifted:

3312: \begin{equation}

3313: \left( \begin{array}{c}

3314:  n \\ m \\ {\mathcal Q} 

3315: \end{array}\right)

3316: \longrightarrow 

3317: \left( \begin{array}{c}

3318:  n \\ m + {\mathcal Q} \cdot e +\frac{1}{2}ne\cdot e \\ {\mathcal Q} -ne

3319: \end{array}\right).

3320: \end{equation}

3321: Clearly, all of the previously massless fermionic states  with $n = m = 0$

3322: have their masses shifted to non-zero values, which means  that

3323: supersymmetry is broken, with a supersymmetry-breaking  mass 

3324: \begin{equation}

3325: m_{3/2} = \frac{{\mathcal Q} \cdot e}{R}.

3326: \end{equation}

3327: In the case of Type-IIA string theory, the

3328: vector product ${\mathcal Q} \cdot e = 1/2$~\cite{kounnas}, giving

3329: a supersymmetry-breaking mass of  $m_{3/2} = 1/(2R)$. As already

3330: noted, a $D$-dimensional theory at finite temperature is equivalent to

3331: a $(D-1)$-dimensional theory with broken supersymmetry, so the radius

3332: $R$ can be identified with the temperature of the system, $2\pi R=

3333: T^{-1}$, giving   

3334: \begin{equation}

3335: \label{gravitino}

3336: m_{3/2} = \pi T,

3337: \end{equation}

3338: from which it is clear that supersymmetry is restored when

3339: $R\rightarrow \infty$, i.e., at zero temperature.

3340: 

3341: 

3342: 

3343: \subsection{Effective Potentials in Type-IIA Supergravity} 

3344: 

3345: The spontaneously broken $(D-1)$-dimensional theory can be used in certain

3346: cases to determine the value of the gravitino mass, via minimization of

3347: the scalar potential. As discussed in~\cite{kounnas}, for the case they

3348: considered of $D=5$ heterotic theory at temperature, a global minimum was

3349: found which could be used to calculate $m_{3/2}$ in terms of the dilaton

3350: field. For the case of Type-IIA strings, a similar analysis was performed,

3351: but the form of the scalar potential was such that there was no global

3352: minimum.

3353: 

3354: This is understandable in view of the duality symmetries which occur

3355: at finite temperature. At finite temperature the heterotic string

3356: possesses a duality which relates the original Hagedorn temperature to

3357: an upper Hagedorn temperature, above which the tachyon 

3358: disappears~\cite{Atick:1988si, O'Brien:1987pn}:

3359: \be

3360: R \rightarrow \alpha^\prime/R, \hspace{20pt} T \rightarrow (4\pi^2

3361:   \alpha^\prime T)^{-1}.

3362: \ee 

3363: This temperature duality of the heterotic string is directly related

3364: to a duality in the scalar potential found by~\cite{kounnas}, the

3365: existence of which appears to determine the existence of the global

3366: minimum. For the five-dimensional Type-IIA string theory considered

3367: in \cite{kounnas}, there is no such duality, thus no

3368: global minimum. These considerations, however, concern the 

3369: compactified theories~\footnote{The spontaneously broken supersymmetric 

3370: $9$-dimensional effective theory,

3371: representing the ten-dimensional supergravity at finite temperature, 

3372: has a scalar superpotential proportional, as usual, to the 

3373: square of the gravitino mass.}.

3374: In the compactified Type-IIA case, the scalar potential assumes the form:

3375: \begin{equation}

3376: V = -\frac{1}{S} \cdot a(Z, \Omega, \dots) \propto -m_{3/2}^2,

3377: \label{scpot}

3378: \end{equation} 

3379: where $S = e^{-\Phi}$ is the dilaton, and the (positive) function 

3380: $a (Z, \Omega, \dots )$, with $Z$, $\Omega, \dots $ appropriate

3381: moduli fields in the supergravity multiplet, 

3382: is given in~\cite{kounnas} for the D=5 case. 

3383: As discussed in~\cite{kounnas}, minimization with respect to the 

3384: $\Omega$ field leads to a runaway potential in the $Z$ direction,

3385: thereby leading to the absence of a global minimum, in accordance with the 

3386: above-mentioned duality argument. 

3387: 

3388: The absence of a global minimum is not an unwelcome situation for the

3389: cosmological model of~\cite{emw}, where the collision of branes causes an

3390: excitation of the brane world, which no longer sits at its stable minimum

3391: and becomes metastable. The excitation energy is determined in this case

3392: by the bulk potential (\ref{symmpot}), which in turn is identified with

3393: the central charge deficit of an appropriate non-critical $\sigma$ model,

3394: describing (perturbative) string (bulk) excitations.

3395: 

3396: 

3397: 

3398: 

3399: \subsection{Colliding-Brane Scenario, Non-Critical Strings and Effective 

3400: Potentials in Type-IIA Theories} 

3401: 

3402: We now examine the previous case in some detail, with the aim of

3403: understanding from a world-sheet framework the various hot and cold phases

3404: of the theory.

3405: 

3406: \subsubsection{Thermal Type-IIA Phase following the Collision}

3407: 

3408: We return to the colliding brane scenario described in~\cite{emw},

3409: in particular in the phase shortly after the first collision in the 

3410: configuration of Fig.~\ref{fig:nonchiral}, 

3411: when the relative separation $r_1$ of the colliding branes is 

3412: \begin{equation}\label{negativepotregion}

3413: r_1 \le \frac{R}{16} .

3414: \end{equation}

3415: In this region the bulk effective potential (\ref{symmpot}) is negative.

3416: 

3417: In the colliding-brane scenario~\cite{emw}, we are dealing essentially

3418: with a \emph{non-equilibrium} situation.  The bulk potential

3419: (\ref{symmpot}), therefore, should \emph{not} be viewed as indicating a

3420: minimum value of a superpotential of the low-energy supergravity theory in

3421: the bulk. Indeed, as we discussed in the previous Section, the effective

3422: potential of Type-IIA supergravity does not have a global minimum.

3423: Instead, we view the potential (\ref{symmpot}) as a \emph{non-equilibrium}

3424: excitation energy of the vacuum due to the collision of the brane worlds.

3425: From the point of view of the low-energy effective theory this is a

3426: \emph{metastable vacuum} (local minimum), which is potentially interesting

3427: in the cosmological context considered here.

3428: 

3429: Following the analysis in~\cite{emw}, we may associate 

3430: the negative potential (\ref{symmpot}) with a central charge

3431: deficit $Q^2 = C - c^* < 0$ of a \emph{subcritical} 

3432: $\sigma$ model describing (perturbative) string 

3433: bulk excitations in this hot phase. 

3434: The analysis of~\cite{emw} assumed configurations of the bulk D-particles 

3435: that were sufficiently 

3436: dilute that the dominant contribution to the 

3437: central-charge deficit, identified as the ten-dimensional

3438: energy density corresponding to the potential (\ref{symmpot}),

3439: is: 

3440: \begin{equation}\label{ccd2}

3441:  |Q^2| \simeq 1.2 \cdot 10^{-8} v^4 g_s^2,

3442: \end{equation}

3443: where $g_s$ is the string coupling, and 

3444: $v$ is the brane-world recoil velocity, which

3445: is constrained by WMAP~\cite{wmap}

3446: data to be at most of order~\cite{emw,brany}: $v \le 0.8$

3447: for the symmetric model of colliding branes of~\cite{emw}, as

3448: depicted in Fig.~\ref{fig:nonchiral}.

3449: This last relation is obtained upon compactifying (formally) 

3450: the model into one large dimension along the ninth (bulk) direction,

3451: and five small directions of order $\sqrt{\alpha '}$)~\footnote{The 

3452: compactification issue is a non-trivial one in our case, and the resulting 

3453: four-dimensional

3454: supergravity may present complications. For our purposes here we only

3455: present generic qualitative arguments, postponing a detailed compactification

3456: analysis for a future publication.}. 

3457: 

3458: 

3459: Since, according to~\cite{rostant}, $D$-dimensional strings at finite

3460: temperature are equivalent to $(D-1)$-dimensional strings with

3461: spontaneously broken supersymmetry, we may view the effective target-space

3462: supergravity theory corresponding to the hot phase of the colliding branes

3463: of~\cite{emw} as living in 9 target dimensions, and corresponding to the

3464: effective action of a non-critical string with an anti-de-Sitter

3465: (negative) cosmological constant whose magnitude is given by (\ref{ccd}).

3466: The pertinent nine-target-dimensional 

3467: $\sigma$-model theory needs Liouville dressing~\cite{ddk}

3468: to restore conformal symmetry, but with a 

3469: \emph{space-like} Liouville mode. The pertinent

3470: dressed $\sigma$ model is characterized 

3471: by a flat Minkowski target-space metric $G_{\mu\nu}=\eta_{\mu\nu}$ 

3472: and a background dilaton linear in 

3473: the Liouville coordinate~\cite{aben}, which is viewed as a 

3474: Euclidean time $X^0_E$: 

3475: \begin{equation}\label{lindil}

3476: \Phi = -\frac{1}{2}QX^0_E~.

3477: \end{equation} 

3478: The corresponding target space of the dressed theory is again 

3479: ten-dimensional: $(9,X_E^0)$, and the corresponding effective

3480: action in the $\sigma$-model frame is given by 

3481: \begin{equation} 

3482: S_{\rm \sigma-{\rm frame}} = \int_0^\beta dX^0_ E~d^{9}X 

3483: \sqrt{G} e^{-2\Phi}\left( R - Q^2

3484: +  4(\nabla _\mu \Phi)^2 + \dots \right) ,

3485: \end{equation}

3486: where $\beta = 1/2\pi T$ is the inverse temperature, which should

3487: be compared with the appropriate parts of (\ref{actioniia}).

3488: We see that the difference from (\ref{actioniia}) is the 

3489: presence of a dark energy term proportional to $e^{-2\Phi}Q^2$,

3490: which plays the r\^ole of a non-zero contribution to the

3491: appropriate scalar potential, and is responsible for supersymmetry

3492: breaking. Additional contributions/modifications 

3493: will result from compactification, but for the purposes

3494: of this Section we restrict ourselves to the 

3495: uncompactified thermal case. 

3496: 

3497: The equations of motion obtainable from this action are equivalent to the 

3498: conformal invariance conditions of the Liouville-dressed 

3499: ten-target-dimensional stringy $\sigma$ model.

3500: The dilaton equation (equivalently the vanishing of the 

3501: ten-euclidean-dimensional dilaton $\beta$ function) reads~\cite{gsw}:

3502: \begin{equation} 

3503: R + 4 (\partial _\mu \Phi)(\partial ^\mu \Phi) - 4\Box \Phi = Q^2 ,

3504: \end{equation}

3505: from which we see that the linear-dilaton background~\cite{aben}

3506: (\ref{lindil}) in a flat $\sigma$-model-frame target metric 

3507: satisfies this equation, as expected from the fact that 

3508: the Liouville dressing restores the conformal symmetry. 

3509: This implies that this background is at least a \emph{local minimum}

3510: of the action. Upon compactification of the type IIA theory, we know from 

3511: the work of~\cite{kounnas} that there may be no global minimum, thereby 

3512: making the above-mentioned extremum of the action a metastable vacuum.

3513: This is a welcome fact, because this will lead to the cosmological

3514: evolution of our brane world, and its eventual exit from this hot phase.

3515: 

3516: One may go one step further, and derive a relation between the temperature

3517: of the hot phase and the recoil velocity by requiring a perturbative

3518: space-like supersymmetry between bosonic and fermionic degrees of freedom,

3519: as in the heterotic string case (\ref{masshift}),(\ref{qgravino}). Indeed,

3520: since we \emph{have postulated} that the string theory describing the

3521: excitations in the bulk of this situation is a \emph{subcritical

3522: Liouville} string~\cite{emn,aben}, we know from the generic analysis

3523: of~\cite{aben} that in such a non-critical string the bosonic masses will

3524: acquire a shift by $Q^2$, as compared with the $Q=0$ case, while the

3525: fermion masses remain unshifted (c.f., (\ref{qgravino})). We now

3526: \emph{require} that there should be no supersymmetry breaking at the

3527: perturbative level in the bulk theory, exactly as happens in the heterotic

3528: string case. We base this postulate on duality symmetries between the

3529: heterotic and Type-IIA theories. It means that the finite-temperature mass

3530: shift of the gravitino (\ref{masshift})  should compensate the Liouville

3531: shift (\ref{qgravino}). This would result to a restoration of a bulk

3532: space-like supersymmetry, at the perturbative level. From the point of

3533: view of the original model of~\cite{emw}, this supersymmetry restoration

3534: would be compatible with the anti-de-Sitter nature of the bulk geometry in

3535: the regime where the effective potential (\ref{symmpot}) is negative.

3536: 

3537: From (\ref{gravitino}) and (\ref{qgravino}), then, we may determine a 

3538: relationship between the central charge deficit $Q^2$ 

3539: of the Liouville $\sigma$ model, describing bulk string excitations,

3540: and the temperature $T$. Furthermore, as we mentioned above, 

3541: the analysis of~\cite{emw} relates the central charge deficit to the brane 

3542: recoil velocity $v$ (\ref{ccd2}). The result of such an analysis is 

3543: therefore:

3544: \begin{equation} 

3545: m_{3/2}^2 = \pi^2 T^2 = Q^2 \simeq 1.2 \cdot 10^{-8} v^4 g_s^2.

3546: \label{qt}

3547: \end{equation}

3548: From the point of view of the 

3549: spontaneously-broken nine-dimensional target-space theory,

3550: this gravitino mass is proportional to the scalar potential at a 

3551: local minimum. Upon compactification of the theory, this minimum 

3552: is not a global one, as can be seen by an analysis 

3553: similar to that of~\cite{kounnas}, 

3554: mentioned previously. The metastable vacuum state 

3555: of the thermal vacuum of the compactified Type~IIA theory 

3556: can then be found by solving the appropriate dilaton 

3557: equation. The cleanest method is to use the equation of motion 

3558: in the Einstein frame (\ref{smodeinst}), 

3559: where the gravitational curvature term

3560: in the effective action has a canonical normalization.

3561: The pertinent dilaton equation in a conformally flat target-space

3562: background reads:

3563: \begin{equation}\label{eqbox}

3564: 4\Box \Phi = -Q^2e^{2\Phi}.

3565: \end{equation} 

3566: The solution of such equations (of the compactified theory), together with

3567: the Einstein equations, determines the metastable 

3568: thermal Type-IIA vacuum corresponding 

3569: to our case, with an excitation energy proportional to $Q^2 < 0$, 

3570: given in (in magnitude) by (\ref{ccd2}). Notice that a dilaton 

3571: of the form (\ref{dil2}) in the Euclidean-time Einstein frame 

3572: satisfies the above equation for $D = 4$ uncompactified dimensions. 

3573: The non-trivial issue in the higher-dimensional case of~\cite{emw}

3574: is to find, upon compactification, the dependence of a 

3575: dilaton satisfying (\ref{eqbox})

3576: on the radii of the compact dimensions/moduli~\cite{kounnas}. 

3577: We do not consider this issue further here.

3578: 

3579: We now remark that the equality 

3580: (\ref{qt}) allows us to determine a recoil-velocity 

3581: dependence of the temperature of the early phase of the brane Universe 

3582: after the collision, in the adiabatic situation considered here:

3583: \begin{equation}\label{tv2}

3584: T \sim 10^{-4} 2\sqrt{2} g_s v^2/(2\pi \sqrt{2\alpha '})~\le~ 

3585: 1.28 \cdot 10^{-4}~T_H,

3586: \end{equation}

3587: where $T_H = 1/2\pi\sqrt{2\alpha '}$ is the respective Hagedorn temperature,

3588: and we assumed standard weakly-coupled strings with $g_s^2 \sim 1/2$.

3589: The fact that the temperature turns out to be proportional to $v^2$ 

3590: is in agreement with the arguments given above on the 

3591: transformation of most of the (non-relativistic) 

3592: kinetic energy of the colliding branes

3593: into thermal energy, in the adiabatic approximation we use here.

3594: 

3595: We see from (\ref{tv2}) that this early phase of the brane Universe, soon

3596: after the collision, could be characterized by quite a high temperature,

3597: up to $10^{13}$ GeV, if we accept that a typical string scale corresponds

3598: to an energy of $10^{18}$~GeV~\cite{gsw}. Of course, the above estimate

3599: has been obtained by saturating the upper bounds for the recoil velocity

3600: that fit the WMAP data~\cite{emw,brany}, and in practice one may have

3601: somewhat lower temperatures. In fact, lower temperatures may be required

3602: in order to avoid massive gravitino overproduction. Such constraints would

3603: restrict further the upper bound on the recoil velocities in the

3604: (compactified version of the) model of~\cite{emw}.

3605: 

3606: However, despite the perturbative supersymmetry restoration,

3607: one would have non-perturb- ative thermal instabilities, 

3608: for the same reason as in the heterotic case examined 

3609: above~\cite{kounnas}, associated with supersymmetry obstruction.

3610: Such non-perturbative instabilities would result in 

3611: the presence of tachyonic states in the string spectrum,

3612: which could provide the initial cosmological instability.

3613: As discussed in~\cite{dgmpp}, it seems to be a generic feature

3614: of such tachyonic states to decouple quickly in the cosmological

3615: Liouville evolution. In addition to these non-perturbative

3616: instabilities, compactification of Type-IIA theories

3617: leads to extra instabilities, due to the above-mentioned

3618: lack of a global minimum in the low-energy effective scalar

3619: potentials arising from thermal supersymmetry breaking.

3620: The metastable nature of the hot phase of the Type-IIA vacuum 

3621: leads to an exit from this phase, which is succeeded by a 

3622: cold inflationary phase that we now proceed to discuss. 

3623: 

3624: \subsubsection{Inflationary Phase}

3625:  

3626: Some time after the initial collision,

3627: the recoiling D-brane world's bulk potential (\ref{symmpot})

3628: becomes {\it positive}. 

3629: From a conformal field theory point of view, and in 

3630: the adiabatic approximation we assumed in~\cite{emw},

3631: this phase might be described by {\it an analytic continuation}

3632: of the above linear-space-like dilaton solution:

3633: \begin{equation}

3634: Q \to iQ~, \quad X_E^0 \to i~t .

3635: \label{analcont}

3636: \end{equation} 

3637: The corresponding $\sigma$-model-frame metric, which in the 

3638: hot phase was a flat Minkowski metric, becomes now

3639: conformally flat:

3640: \begin{equation}

3641: G_{\mu\nu} = e^{-2\Phi}\eta_{\mu\nu}~,~~\Phi = \frac{1}{2}Q~t  .

3642: \label{cflat}

3643: \end{equation}

3644: In the model of~\cite{emw} this could be the full ten-dimensional metric,

3645: although appropriate compactifications can restrict the indices to 

3646: the four dimensions relevant for

3647: a three-brane, as seen in Fig.~\ref{infla}, which represents

3648: the inflationary model~\cite{brany}

3649: reviewed in Section 2.3 above. The normalizations in (\ref{cflat}) 

3650: pertain to the four-dimensional case of the three-branes, and would

3651: change for higher-dimensional branes~\cite{emw}.

3652: 

3653: It is a curiosity that, setting $Q=-3H$, the physical (Einstein-frame)  

3654: dilaton and metric fields (\ref{smodeinst}) of the hot phase, which remain

3655: {\it real} under (\ref{analcont}), become equivalent to the corresponding

3656: Liouville-undressed fields (\ref{cflat}) (c.f.,

3657: (\ref{metricinfl}),(\ref{seventy})) when one sets the Liouville field

3658: $\varphi = 0$. However, this is only a coincidence, since in the

3659: inflationary phase it is the $\sigma$-model-frame metric that acquires the

3660: conformally-flat form. As a $\sigma$-model-frame metric, (\ref{cflat}) is

3661: \emph{not} conformal invariant, since its one-loop $\beta$ function is

3662: non-vanishing: $\beta_{\mu\nu}^G = R_{\mu\nu} = Q^2 G_{\mu\nu} \ne 0$. The

3663: central charge deficit $Q^2$ in this case is given by the potential

3664: (\ref{symmpot})  in the regime in which is positive, which is treated as a

3665: constant in the adiabatic approximation.

3666: 

3667: Because of the positive central charge deficit, the string system requires

3668: Liouville dressing by a {\it time-like} field, $\varphi$, which is an

3669: extra time-like coordinate, in addition to $t$.  The eventual

3670: identification (\ref{liouvtime}), which in this scenario is dictated by

3671: dynamical reasons~\cite{gravanis}, ensures that there is only one time

3672: variable in the formalism, and leads to an eventual constant dilaton

3673: during the inflationary phase. This phase in which the Universe cools down

3674: is nothing other than the inflationary phase, described in Section 2.3

3675: above. The analytic continuation procedure (\ref{analcont})  describes

3676: simply a phase transition of the bulk superstrings from the hot phase to a

3677: cold inflationary one, within the colliding-brane system of

3678: Fig.~\ref{fig:nonchiral}.

3679: 

3680: We recall that, as a result of the non-perturbative breaking of

3681: target-space supersymmetry in the hot phase, there are tachyonic states in

3682: the spectrum, which trigger the initial cosmological instability. However,

3683: as discussed in~\cite{dgmpp}, such states decouple relatively quickly in

3684: the cosmic evolution.

3685: 

3686: \subsubsection{Exit from the Inflationary Phase: Reheating and Possible

3687: Subsequent Collision(s)}

3688: 

3689: In a similar vein, one may discuss the phase transition associated with

3690: the second collision, and the subsequent reheating of the Universe.  

3691: However, the physics of reheating is not understood at a satisfactory

3692: level in this framework. In the context of the model of~\cite{emw}, one

3693: has to understand technical details associated with internal magnetic

3694: fields in the respective orientifold compactification~\cite{dudas}, as

3695: well as issues with the potential felt by open-string excitations on a

3696: brane world. These issues still raise many open questions, but, for

3697: completeness, we now present some relevant speculations.

3698: 

3699: The potential $V_{\rm brane}$ felt by our brane world in the 

3700: model of~\cite{emw},

3701: in the configuration of Fig.~\ref{fig:nonchiral}, is by itself negative

3702: even during the inflationary phase~\cite{emn}, 

3703: \begin{equation}

3704: V_{\rm brane} =  -V_8\frac{31(R-2r)v^4}{2^{13}\pi^9{\alpha '}^5}~,

3705: \label{branepotopen}

3706: \end{equation}

3707: where $r_1 = R - 2r$ denotes the relative separation of the branes

3708: in the symmetric situation of Fig.~\ref{fig:nonchiral} with $r_2~=~r$.

3709: This negative value can be understood by recalling that 

3710: the brane world feels an attractive force towards its original stack,

3711: and the configuration is stabilized when $v \to 0$.

3712: 

3713: An issue arises at this point, concerning the boundary conformal field

3714: theory of open-string excitations, with their ends attached to the brane.

3715: In view of this negative potential, one may think of dressing the

3716: open-string $\sigma$ model with a {\it space-like} Liouville mode to

3717: restore conformal symmetry, already during the inflationary phase.  In

3718: view of the corresponding situation in the closed string

3719: sector~\cite{kounnas}, discussed above, one is tempted to take the view

3720: that such a negative brane potential represents some sort of

3721: thermalization of open string excitations on the brane world during the

3722: inflationary era.

3723: 

3724: Indeed, for $r < R/2$, i.e., very soon after the initial brane collision

3725: {\it both} the brane and bulk Universes are {\it hot} and in thermal

3726: equilibrium. As the two bouncing brane worlds of Fig.~\ref{fig:nonchiral}

3727: move further apart, the closed-string excitations in the bulk cool down,

3728: since the available space becomes larger and their collisions rarer,

3729: whilst the brane Universe remains initially hot, because the inflationary

3730: expansion is only a `mirage' due to the brane motion.

3731: 

3732: The order of magnitude of the brane potential (\ref{branepotopen}) is the

3733: same as the bulk one (\ref{symmpot}), which implies that, after the exit

3734: from the inflationary phase, our brane Universe remains thermalized with a

3735: temperature at intermediate energy scales $\sim 10^{13}$ GeV, according to

3736: the calculation above~\footnote{Such temperatures may characterize

3737: no-scale supergravity models~\cite{noscale}.}. This may provide an

3738: alternative to conventional reheating scenarios in the following sense:

3739: although the bulk cools down significantly, and the Universe exits from

3740: inflation, undergoing an appropriate phase transition, expressed by the

3741: change in sign of the bulk potential (\ref{symmpot}), the brane world

3742: remains thermalized after the exit from inflation at temperatures of order

3743: $10^{13}$ GeV. This is simply a result of the initial collision, without

3744: the need for other reheating mechanisms.

3745: 

3746: The usual constraints on gravitino overproduction in spontaneously-broken

3747: supergravity models, such as those pertaining to the brane Q-cosmologies

3748: of~\cite{emw}, restrict the allowed temperature to values much smaller

3749: than $10^{13}$~GeV. This in turn implies an upper bound on the brane

3750: recoil velocities, according to the discussion following (\ref{tv2}).  

3751: However, one should bear in mind that in brane models the produced

3752: gravitino will escape in the bulk, since it is an excitation of the closed

3753: superstring multiplet, and therefore these constraints may not be so

3754: strict as in conventional supergravity cosmologies.

3755: 

3756: This approach may provide an explicit realization of the ideas in the

3757: ekpyrotic scenario for inflation and reheating of the

3758: Universe~\cite{ekpyrotic}. Immediately after inflation there is a

3759: difference in temperature between the bulk and the brane worlds.  As time

3760: passes, this difference in temperature will cause significant closed

3761: string (gravitational)  emission from the brane to the bulk, in order to

3762: equilibrate the situation with a common (low) temperature in both brane

3763: and bulk worlds.

3764: 

3765: Due to energy conservation, this causes non-adiabatic motion, with the

3766: brane world decelerating towards an eventually zero velocity. This would

3767: correspond to the exit phase from the inflationary epoch, given that the

3768: central charge deficit of the pertinent stringy $\sigma$ model would

3769: vanish asymptotically and, according to the discussion in Section 2.3, the

3770: space-time metric would tend to that of a static flat Minkowski

3771: space-time. In some models, however, e.g., those with internal magnetic

3772: flux contributions, such as the Type-0 models discussed previously, the

3773: asymptotic state may be that of a linear dilaton, leading to a linearly

3774: expanding Einstein-frame Universe. In addition, as a result of brane

3775: recoil effects also discussed above, one would have contributions to the

3776: dark energy, relaxing asymptotically either to zero or to a constant

3777: contribution (set, for instance, by the internal magnetic field

3778: contributions in magnetized compactifications~\cite{bachas}), as in

3779: (\ref{cosmoconst}) or (\ref{cosmoconst3}), (\ref{sigmamodelq0}).  In all

3780: such models, current-era cosmology can be made compatible with

3781: observations by fixing the various stringy

3782: parameters~\cite{dgmpp,brany,emn04}.

3783: 

3784: \subsubsection{Open Issues: a Second Collision? Nucleosynthesis?

3785: ...} 

3786: 

3787: There are several open issues regarding the fate of the inflationary

3788: Universe.  It is an open, and certainly model-dependent question whether

3789: the gravitational radiation from the brane to the bulk makes the brane

3790: world stop before the second collision takes place (in which case the

3791: latter will never occur).  Indeed, if the brane gravitational radiation

3792: causes a significant reduction of the brane velocity $v$ before the second

3793: collision with the stack of D-branes in Fig.~\ref{fig:nonchiral}, it is

3794: possible that the velocity contribution to the bulk potential

3795: (\ref{symmpot}) diminishes significantly, in such a way that the

3796: D-particle term overcomes the positive $v^4$ term. In such a case the bulk

3797: potential becomes unstable (negative) and the bulk string system may

3798: thermalize in the way described earlier. The thermalization may also have

3799: a conformal field theory description, with the bulk background central

3800: charge deficit being given by $-Nv^{1/2}$.  If the second collision takes

3801: place before the brane stops due to radiation, there may be a local

3802: disturbance in the population of the D-particles near the brane worlds

3803: during the (second)  collision, such that $N$ is significantly higher than

3804: before, when the brane was moving in the bulk. Such local disturbances may

3805: even cause the massive D-particles to collapse forming black holes, whose

3806: Hawking evaporation leads to additional thermal contributions to the brane

3807: world after the second collision.

3808: 

3809: It is an open issue whether the potential energy of these re-thermalized

3810: bulk closed strings becomes of similar order as the brane potential

3811: (\ref{branepotopen}).  If such were the case, one would reach thermal

3812: equilibrium between brane and bulk worlds, the gravitational radiation

3813: towards the bulk could counterbalance the breaking up of bulk closed

3814: strings on the brane, and the brane world would stop decelerating before

3815: the second collision take place. The brane world could then either move

3816: again adiabatically in the bulk with a very small velocity, in which case

3817: there could still be a long time before it starts accelerating again due

3818: to the influence of the other branes, either until the second collision

3819: takes place, or until it stops.  The low (equilibrium) temperature in

3820: either case could be identified with the CMB temperature of the present

3821: era of the Universe.

3822: 

3823: \begin{figure}[tb]

3824: \begin{center}

3825: \epsfxsize=3.5in

3826: \bigskip

3827: \centerline{\epsffile{colbr.eps}}

3828: \caption{{\it A nucleosynthesis-friendly Liouville cosmology scenario,  

3829: according to which the second collision of a  moving brane world 

3830: with the 

3831: (left) stack of branes in Fig.~\ref{fig:nonchiral}

3832: occurs after the phase where the brane almost stops due to radiation.

3833: This scenario provides 

3834: a relaxation model for the cosmological vacuum energy (central-charge 

3835: deficit of the Liouville $\sigma$ model), 

3836: which passes first through a metastable phase where it almost vanishes 

3837: (up to magnetic field contributions) during the nucleosynthesis era,

3838: and then raises again,  as a result of the second collision,  

3839: but at a much lesser height than in the initial collision.

3840: \label{fig:nucleosynthesis}}}

3841: \end{center} 

3842: \end{figure}

3843: 

3844: 

3845: 

3846: Another open issue concerns the mechanism for \emph{nucleosynthesis} in

3847: such a scenario.  Nucleosynthesis requires a delicate balance between the

3848: expansion of the Universe and the rate of nuclear reactions for the

3849: formation of the light elements, which appears to work very well in

3850: scenarios with a negligible cosmological constant.  It may therefore be

3851: desirable that the reduction in brane velocity due to radiation occurs

3852: around the nucleosynthesis era, so that in such a case the brane Universe

3853: has only a very small vacuum energy. For instance, in the class of models

3854: with compactified branes in magnetized internal manifolds~\cite{gravanis}

3855: it could be that only the magnetic field supersymmetry-breaking

3856: contributions to the vacuum energy are present during the nucleosynthesis

3857: era. At the end of nucleosynthesis a second collision of the brane world

3858: with the stack of branes in Fig.~\ref{fig:nonchiral} takes place,

3859: resulting in an increase of the central-charge deficit (vacuum energy),

3860: but at a much lesser height than in the initial collision (due to the much

3861: smaller velocities involved), as seen in Fig.~\ref{fig:nucleosynthesis}).  

3862: Eventually the central charge relaxes again to zero asymptotically,

3863: providing the vacuum energy in the present epoch.

3864: 

3865: The question then arises as to what precisely causes the current energy

3866: density of the Universe in such a case.  So far we have argued that a

3867: linear (in the string frame) dilaton may act as a quintessence field, in

3868: accordance with current cosmological phenomenology.  However, it is our

3869: opinion that, in order to answer this question completely, one should also

3870: incorporate in the above discussion the {\it recoil} fluctuations on the

3871: brane world, which echo the initial brane collision.  As mentioned above,

3872: such effects would provide {\it positive} contributions to the present-era

3873: central charge deficit of the corresponding stringy $\sigma$ model, for

3874: asymptotically long times after the initial collision. The recoil

3875: contributions depend on the recoil velocity of the branes during the

3876: (adiabatic)  bouncing inflationary phase, but they diminish with the

3877: cosmic time, relaxing towards either zero or some other small positive

3878: (equilibrium) value, determined for instance by the internal magnetic

3879: field (c.f., (\ref{cosmoconst}) or (\ref{cosmoconst3}),

3880: (\ref{sigmamodelq0}) respectively). These recoil contributions may be

3881: responsible for parts of the dark energy density of the observable

3882: Universe, which exceed those due to the linear-dilaton quintessence.  

3883: They could also be in accordance with current astrophysical

3884: observations~\cite{snIa,wmap}. Such recoil contributions may overcome any

3885: negative thermal contributions in the bulk, so that the bulk energy never

3886: becomes negative, in contrast to the negative brane energy for non-zero

3887: velocities.

3888: 

3889: These and other related issues are currently under investigation, and we

3890: hope to be able to report some more complete results soon.

3891: 

3892: \section{Conclusions and Outlook} 

3893: 

3894: We have examined in this work various cosmological models based on

3895: non-critical Liouville strings - Q-Cosmologies - with various asymptotic

3896: configurations of the dilaton, and have speculated on the inflationary

3897: phase, on the possibility of exit from it and reheating, as well as the

3898: large-times eras of the Universe (current and future). A particularly

3899: interesting case from a physical point of view is that of a linear dilaton

3900: that is asymptotically linear in cosmic time, which is known to correspond

3901: to a true conformal field theory~\cite{aben}.  In such a model we have

3902: observed that the string coupling is identified (up to irrelevant

3903: constants of order one)~\cite{emn04} with the deceleration parameter of

3904: the Universe, through equation (\ref{important}). We have argued that the

3905: present-era phenomenology of the model, including matter, is compatible

3906: with the astrophysical data in a quite natural way, for suitable values of

3907: the adjustable parameters in the model.

3908: 

3909: We stress once more the importance of being

3910: non-critical in order to arrive at (\ref{important}). In critical strings,

3911: which usually assume the absence of a four-dimensional dilaton, such a

3912: relation cannot be obtained, and the string coupling is not directly

3913: measurable with cosmological data. The logarithmic variation with the 

3914: cosmic time of the dilaton

3915: field at late times implies a slow variation of the string coupling

3916: (\ref{important}), ${\dot g_s}/g_s = 1/t_E \sim 10^{-60}$ in the present

3917: era, and hence a corresponding variation of the gauge coupling constants.

3918: However, this variation is too small to be seen currently.

3919: 

3920: The use of Liouville strings to describe the evolution of our Universe is

3921: natural, since non-critical strings are associated with non-equilibrium

3922: situations which undoubtedly occurred in the early Universe.  We have

3923: discussed in this framework the phase diagram of a Liouville cosmological

3924: string model of two colliding-brane worlds. We have seen that, immediately

3925: after the collision, the bulk string Universe passes through a hot,

3926: metastable phase, before entering an inflationary cold phase. On the other

3927: hand, the brane Universe (our world) remains thermalized throughout the

3928: two phases, at a relatively high temperature, causing gravitational

3929: radiation from the brane to the bulk, which tends to equilibrate the

3930: temperature, which eventually decelerates the motion of the brane world in

3931: the bulk. From the point of view of an observer on the brane, however, the

3932: brane Universe may at present seem to be accelerating, with the

3933: acceleration provided by the dilaton field of the string multiplet, as

3934: mentioned above.

3935: 

3936: Exit from the inflationary phase is still an unresolved issue, although

3937: scenarios have been conjectured, involving for instance a second collision

3938: of the brane world of the model of~\cite{emw} with the stack of D-branes

3939: in Fig.~\ref{fig:nonchiral}. This could provide extra contributions to the

3940: reheating of the brane world, as a result of the gravitational collapse of

3941: D-particle populations to form bulk black holes, which subsequently emit

3942: Hawking radiation.

3943: 

3944: There are many phenomenological tests of this class of cosmologies that

3945: can be performed, which the generic analysis presented here is not

3946: sufficient to encapsulate. Tensor perturbations in the cosmic microwave

3947: background radiation is one of them. The emission of gravitational degrees

3948: of freedom from the hot brane to the cold bulk, during the inflationary

3949: and post-inflationary phases is something to be investigated in detail. A

3950: detailed knowledge of the dependence of the equation of state on the

3951: redshift is something that needs to be looked at in the context of

3952: specific models. The constant equation of state obtained here is only an

3953: asymptotic feature of an era where the gravitational sector dominates.  

3954: Moreover, issues regarding the delicate balance of the expansion of the

3955: Universe and nucleosynthesis, which requires a very low vacuum energy,

3956: must be resolved in specific, phenomenologically semi-realistic models,

3957: after proper compactification to three spatial dimensions, in order that

3958: the conjectured cosmological evolution has a chance of success.

3959: 

3960: Finally, the compactification issue \emph{per se} is a most important part

3961: of a realistic stringy cosmology. In our discussion above, we have

3962: presented a rather simplified compactification on magnetized internal

3963: manifolds, in Type-II five-brane models, which also provided

3964: phenomenologically realistic ways of breaking target-space supersymmetry

3965: in cold Universes, compatible with the very small value of the vacuum

3966: energy that has been reported in the Universe today. However, in the

3967: context of the model of~\cite{emw}, involving eight-branes and

3968: orientifolds, such compactifications may present subtleties that require

3969: extra attention~\cite{dudas}.

3970: 

3971: We hope to be able to report on these and other related issues in future

3972: work. We are far from claiming a detailed understanding in this framework

3973: of several important facts of modern cosmology, such as the Universe's

3974: current acceleration, dark energy, the various phase transitions in the

3975: past history of the cosmos, {\it etc.}. Nevertheless, we believe that 

3976: Liouville

3977: strings are probably the only viable way, in the context of string theory,

3978: to discuss rigorously cosmological string backgrounds, especially those

3979: involving accelerated Universes and, in general, dark-energy contributions

3980: to the Universe's energy budget.

3981: 

3982: In this last respect, we stress once more that the non-equilibrium

3983: Liouville approach to cosmology advocated in this article is based

3984: exclusively on the treatment of target time as an irreversible dynamical

3985: renormalization-group scale on the world sheet of the Liouville string

3986: (the zero mode of the Liouville field itself). This irreversibility is

3987: associated with fundamental properties of the world-sheet renormalization

3988: group, which lead in turn to the loss of information carried by

3989: two-dimensional degrees of freedom with world-sheet momenta beyond the

3990: ultraviolet cutoff~\cite{zam} of the world-sheet theory. This fundamental

3991: microscopic time irreversibility may have other important consequences,

3992: associated with fundamental violations of CPT

3993: invariance~\cite{emninfl,emn,mavromatosdecoh} in both the early Universe

3994: and the laboratory, providing other tests of these ideas.

3995: 

3996: 

3997: \section*{Acknowledgments}

3998: 

3999: We thank V. A. Mitsou for useful discussions about the cosmological data. 

4000: N.E.M. thanks the University of Valencia and IFIC for hospitality during

4001: the final stages of this work. The work of D.V.N. is supported by D.O.E.

4002: grant DE-FG03-95-ER-40917. That of M.W. is supported by an EPSRC (U.K.)  

4003: Research Studentship.

4004: 

4005: \begin{thebibliography}{cc} 

4006: 

4007: \bibitem{gsw} M.B. Green, J.H. Schwarz and E. Witten, {\it Superstring 

4008: Theory}, Vols. I \& II (Cambridge University Press, 1987).

4009: 

4010: \bibitem{polchinski} J.~Polchinski,

4011: {\it String theory}, Vol. 2 (Cambridge University Press, 1998);

4012: J.~H.~Schwarz,

4013: %``Some properties of type I' string theory,''

4014: arXiv:hep-th/9907061.

4015: %%CITATION = HEP-TH 9907061;%%

4016: 

4017: 

4018: \bibitem{add} 

4019: N.~Arkani-Hamed, S.~Dimopoulos and G.~R.~Dvali,

4020: %``The hierarchy problem and new dimensions at a millimeter,''

4021: Phys.\ Lett.\ B {\bf 429}, 263 (1998)

4022: [arXiv:hep-ph/9803315];

4023: %%CITATION = HEP-PH 9803315;%%

4024: I.~Antoniadis, N.~Arkani-Hamed, S.~Dimopoulos and G.~R.~Dvali,

4025: %``New dimensions at a millimeter to a Fermi and superstrings at a TeV,''

4026: Phys.\ Lett.\ B {\bf 436}, 257 (1998)

4027: [arXiv:hep-ph/9804398].

4028: %%CITATION = HEP-PH 9804398;%%

4029: 

4030: 

4031: \bibitem{randal} L.~Randall and R.~Sundrum,

4032: %``An alternative to compactification,''

4033: Phys.\ Rev.\ Lett.\  {\bf 83}, 4690 (1999)

4034: [arXiv:hep-th/9906064].

4035: %%CITATION = HEP-TH 9906064;%%

4036: 

4037: 

4038: 

4039: \bibitem{ibanez} see, for instance: 

4040: L.~E.~Ibanez, R.~Rabadan and A.~M.~Uranga,

4041: %``Sigma-model anomalies in compact D = 4, N = 1 type IIB orientifolds and

4042: %Fayet-Iliopoulos terms,''

4043: Nucl.\ Phys.\ B {\bf 576}, 285 (2000)

4044: [arXiv:hep-th/9905098];

4045: L.~E.~Ibanez,

4046: %``Mirage gauge coupling unification,''

4047: arXiv:hep-ph/9905349;

4048: %%CITATION = HEP-PH 9905349;%%

4049: %%CITATION = HEP-TH 9905098;%%

4050: C.~A.~Scrucca and M.~Serone,

4051: %``Gauge and gravitational anomalies in D = 4 N = 1 orientifolds,''

4052: JHEP {\bf 9912}, 024 (1999)

4053: [arXiv:hep-th/9912108];

4054: %%CITATION = HEP-TH 9912108;%%

4055: D.~M.~Ghilencea and G.~G.~Ross,

4056: %``A note on the RG flow in (N = 1, D = 4) supergravity and applications  to

4057: %Z(3) orbifold / orientifold compactification,''

4058: Nucl.\ Phys.\ B {\bf 595}, 277 (2001)

4059: [arXiv:hep-ph/0006318], and references therein.

4060: %%CITATION = HEP-PH 0006318;%%

4061: 

4062: 

4063: \bibitem{langlois} D.~Langlois,

4064: %``Brane cosmology: An introduction,''

4065: Prog.\ Theor.\ Phys.\ Suppl.\  {\bf 148}, 181 (2003)

4066: [arXiv:hep-th/0209261], and references therein.

4067: %%CITATION = HEP-TH 0209261;%%

4068: 

4069: \bibitem{ekpyrotic} J.~Khoury, B.~A.~Ovrut, P.~J.~Steinhardt and N.~Turok,

4070: %``The ekpyrotic universe: Colliding branes and the origin of the hot big  bang,''

4071: Phys.\ Rev.\ D {\bf 64}, 123522 (2001)

4072: [arXiv:hep-th/0103239].

4073: %%CITATION = HEP-TH 0103239;%%

4074: 

4075: \bibitem{emn04} J.~R.~Ellis, N.~E.~Mavromatos and D.~V.~Nanopoulos,

4076:   %``Cosmic acceleration and the string coupling,''

4077:   Phys.\ Lett.\ B {\bf 619}, 17 (2005)

4078:   [arXiv:hep-th/0412240] and

4079:   %%CITATION = HEP-TH 0412240;%%

4080:   arXiv:gr-qc/0503120 (awarded First Prize in the Gravity Research 

4081: Foundation Essay Competition for 2005).

4082:   %%CITATION = GR-QC 0503120;%%

4083: 

4084: 

4085: \bibitem{snIa} A.~G.~Riess \textit{et al.} [Supernova Search Team

4086: Collaboration], 

4087: %``Observational Evidence from Supernovae for an Accelerating Universe and a

4088: %Cosmological Constant,''

4089: Astron.\ J.\ \textbf{116}, 1009 (1998) [arXiv:astro-ph/9805201]; 

4090: %%CITATION = ASTRO-PH 9805201;%%

4091: B.~P.~Schmidt {\it et al.}  [Supernova Search Team Collaboration],

4092: %``The High-Z Supernova Search: Measuring Cosmic Deceleration and Global Cur

4093: %vature of the Universe Using Type Ia Supernovae,''

4094: Astrophys.\ J.\  {\bf 507}, 46 (1998)

4095: [arXiv:astro-ph/9805200];

4096: %%CITATION = ASTRO-PH 9805200;%%

4097: S.~Perlmutter {\it et al.}  [Supernova Cosmology Project Collaboration],

4098: %``Measurements of Omega and Lambda from 42 High-Redshift Supernovae,''

4099: Astrophys.\ J.\  {\bf 517}, 565 (1999)

4100: [arXiv:astro-ph/9812133];

4101: %%CITATION = ASTRO-PH 9812133;%%

4102: J.~P.~Blakeslee {\it et al.}  [Supernova Search Team Collaboration],

4103: %``Discovery of Two Distant Type Ia Supernovae in the Hubble Deep Field North

4104: %with the Advanced Camera for Surveys,''

4105: Astrophys.\ J.\  {\bf 589}, 693 (2003)

4106: [arXiv:astro-ph/0302402];

4107: %%CITATION = ASTRO-PH 0302402;%%

4108: A.~G.~Riess {\it et al.}  [Supernova Search Team Collaboration],

4109: %``The Farthest Known Supernova: Support for an Accelerating Universe and a

4110: %Glimpse of the Epoch of Deceleration,''

4111: Astrophys.\ J.\  {\bf 560}, 49 (2001)

4112: [arXiv:astro-ph/0104455].

4113: %%CITATION = ASTRO-PH 0104455;%%

4114: 

4115: 

4116: 

4117: \bibitem{wmap}D.~N.~Spergel {\it et al.}  [WMAP Collaboration],

4118: %``First Year Wilkinson Microwave Anisotropy Probe (WMAP) Observations:

4119: %Determination of Cosmological Parameters,''

4120: Astrophys.\ J.\ Suppl.\  {\bf 148}, 175 (2003)

4121: [arXiv:astro-ph/0302209].

4122: %%CITATION = ASTRO-PH 0302209;%%

4123: 

4124: 

4125: 

4126: \bibitem{fischler} W.~Fischler and L.~Susskind,

4127: %``Dilaton Tadpoles, String Condensates And Scale Invariance,''

4128: Phys.\ Lett.\ B {\bf 171}, 383 (1986);

4129: %%CITATION = PHLTA,B171,383;%%

4130: %``Dilaton Tadpoles, String Condensates And Scale Invariance. 2,''

4131: Phys.\ Lett.\ B {\bf 173}, 262 (1986).

4132: %%CITATION = PHLTA,B173,262;%%

4133: 

4134: 

4135: 

4136: 

4137: 

4138: \bibitem{aben} I.~Antoniadis, C.~Bachas, J.~R.~Ellis and D.~V.~Nanopoulos,

4139: %``Cosmological String Theories And Discrete Inflation,''

4140: Phys.\ Lett.\ B {\bf 211}, 393 (1988);

4141: %%CITATION = PHLTA,B211,393;%%

4142: %``An Expanding Universe In String Theory,''

4143: Nucl.\ Phys.\ B {\bf 328}, 117 (1989);

4144: %%CITATION = NUPHA,B328,117;%%

4145: %``Comments On Cosmological String Solutions,''

4146: Phys.\ Lett.\ B {\bf 257}, 278 (1991).

4147: %%CITATION = PHLTA,B257,278;%%

4148: 

4149: \bibitem{ddk} F.~David,

4150: %``Conformal Field Theories Coupled To 2-D Gravity In The Conformal Gauge,''

4151: Mod.\ Phys.\ Lett.\ A {\bf 3}, 1651 (1988);

4152: %%CITATION = MPLAE,A3,1651;%%

4153: J.~Distler and H.~Kawai,

4154: %``Conformal Field Theory And 2-D Quantum Gravity Or Who's Afraid Of Joseph

4155: %Liouville?,''

4156: Nucl.\ Phys.\ B {\bf 321}, 509 (1989);

4157: %%CITATION = NUPHA,B321,509;%%

4158: J.~Distler, Z.~Hlousek and H.~Kawai,

4159: %``Superliouville Theory As A Two-Dimensional, Superconformal Supergravity

4160: %Theory,''

4161: Int.\ J.\ Mod.\ Phys.\ A {\bf 5}, 391 (1990);

4162: %%CITATION = IMPAE,A5,391;%% \\

4163: see also: N.~E.~Mavromatos and J.~L.~Miramontes,

4164: %``Regularizing The Functional Integral In 2-D Quantum Gravity,''

4165: Mod.\ Phys.\ Lett.\ A {\bf 4}, 1847 (1989);

4166: %%CITATION = MPLAE,A4,1847;%%

4167: E.~D'Hoker and P.~S.~Kurzepa,

4168: %``2-D Quantum Gravity And Liouville Theory,''

4169: Mod.\ Phys.\ Lett.\ A {\bf 5}, 1411 (1990).

4170: %%CITATION = MPLAE,A5,1411;%%

4171: 

4172: 

4173: 

4174: \bibitem{zam} A.~B.~Zamolodchikov,

4175: %``'Irreversibility' Of The Flux Of The Renormalization Group In A 2-D Field

4176: %Theory,''

4177: JETP Lett.\  {\bf 43}, 730 (1986)

4178: [Pisma Zh.\ Eksp.\ Teor.\ Fiz.\  {\bf 43}, 565 (1986)].

4179: %%CITATION = JTPLA,43,730;%%

4180: 

4181: \bibitem{dgmpp} G.~A.~Diamandis, B.~C.~Georgalas, N.~E.~Mavromatos and E.~Papantonopoulos,

4182: %``Acceleration of the universe in type-0 non-critical strings,''

4183: Int.\ J.\ Mod.\ Phys.\ A {\bf 17}, 4567 (2002)

4184: [arXiv:hep-th/0203241];

4185: %%CITATION = HEP-TH 0203241;%%

4186: G.~A.~Diamandis, B.~C.~Georgalas, N.~E.~Mavromatos, E.~Papantonopoulos and I.~Pappa,

4187: %``Cosmological evolution in a type-0 string theory,''

4188: Int.\ J.\ Mod.\ Phys.\ A {\bf 17}, 2241 (2002)

4189: [arXiv:hep-th/0107124].

4190: %%CITATION = HEP-TH 0107124;%%

4191: 

4192: \bibitem{witten} E.~Witten,

4193: %``Is supersymmetry really broken?,''

4194: Int.\ J.\ Mod.\ Phys.\ A {\bf 10}, 1247 (1995)

4195: [arXiv:hep-th/9409111].

4196: %%CITATION = HEP-TH 9409111;%%

4197: 

4198: 

4199: 

4200: \bibitem{emn} J.~R.~Ellis, N.~E.~Mavromatos and D.~V.~Nanopoulos,

4201: %``String theory modifies quantum mechanics,''

4202: Phys.\ Lett.\ B {\bf 293}, 37 (1992)

4203: [arXiv:hep-th/9207103]; 

4204: %``A microscopic Liouville arrow of time,''

4205: {\it Invited review for the special Issue of J.\ Chaos Solitons Fractals}, 

4206: Vol.\ 10, (eds. C. Castro amd M.S. El Naschie, 

4207: Elsevier Science, Pergamon 1999) 345   

4208: [arXiv:hep-th/9805120];

4209: %``How large are dissipative effects in noncritical Liouville string

4210: %theory?,''

4211: Phys.\ Rev.\ D {\bf 63}, 024024 (2001)

4212: [arXiv:gr-qc/0007044].

4213: %%CITATION = GR-QC 0007044;%%

4214: %%CITATION = HEP-TH 9207103;%%

4215: 

4216: 

4217: 

4218: 

4219: \bibitem{gravanis} E.~Gravanis and N.~E.~Mavromatos,

4220: %``Vacuum energy and cosmological supersymmetry breaking in brane worlds,''

4221: Phys.\ Lett.\ B {\bf 547}, 117 (2002)

4222: [arXiv:hep-th/0205298];

4223: %%CITATION = HEP-TH 0205298;%%

4224: N.~E.~Mavromatos,

4225: %``Quantum gravity, cosmology, (Liouville) strings and Lorentz invariance,''

4226: arXiv:hep-th/0210079 (published in {\it 

4227: Beyond the Desert, Oulu 2002 (Finland)} (ed. H.V. Klapdor-Kleingrothaus, 

4228: IoP 2003)), 3. 

4229: %%CITATION = HEP-TH 0210079;%%

4230: 

4231: 

4232: 

4233: 

4234: 

4235: \bibitem{brany} J.~Ellis, N.~E.~Mavromatos, D.~V.~Nanopoulos and A.~Sakharov,

4236: %``Brany Liouville inflation,''

4237: arXiv:gr-qc/0407089, New J. Phys. {\bf 6}, 171 (2004). 

4238: %%CITATION = GR-QC 0407089;%%

4239: 

4240: 

4241: 

4242: \bibitem{sussk} L.~Susskind,

4243: %``The anthropic landscape of string theory,''

4244: arXiv:hep-th/0302219.

4245: %%CITATION = HEP-TH 0302219;%%

4246: 

4247: 

4248: \bibitem{emw} J.~Ellis, N.~E.~Mavromatos and M.~Westmuckett,

4249: %``A supersymmetric D-brane model of space-time foam,''

4250: Phys.\ Rev.\ D {\bf 70}, 044036 (2004)

4251: [arXiv:gr-qc/0405066];

4252: %%CITATION = GR-QC 0405066;%%

4253: %``Potentials between D-Branes in a Supersymmetric Model of Space-Time Foam,''

4254:  Phys.\ Rev.\ D {\bf 71}, 106006 (2005)

4255: [arXiv:gr-qc/0501060].

4256:   %%CITATION = GR-QC 0501060;%%

4257: 

4258: \bibitem{kounnas} I.~Antoniadis and C.~Kounnas,

4259: %``Superstring phase transition at high temperature,''

4260: Phys.\ Lett.\ B {\bf 261}, 369 (1991);

4261: %%CITATION = PHLTA,B261,369;%%

4262: I.~Antoniadis, J.~P.~Derendinger and C.~Kounnas,

4263: %``Non-perturbative temperature instabilities in N = 4 strings,''

4264: Nucl.\ Phys.\ B {\bf 551}, 41 (1999)

4265: [arXiv:hep-th/9902032].

4266: %%CITATION = HEP-TH 9902032;%%

4267: 

4268: 

4269: 

4270: \bibitem{schmid} C.~Schmidhuber,

4271:   %``Exactly marginal operators and running coupling constants in 2-D gravity,''

4272:   Nucl.\ Phys.\ B {\bf 404}, 342 (1993)

4273:   [arXiv:hep-th/9212075];

4274:   %%CITATION = HEP-TH 9212075;%%

4275:  C.~Schmidhuber and A.~A.~Tseytlin,

4276:   %``On string cosmology and the RG flow in 2-d field theory,''

4277:   Nucl.\ Phys.\ B {\bf 426}, 187 (1994)

4278:   [arXiv:hep-th/9404180].

4279:   %%CITATION = HEP-TH 9404180;%%

4280: 

4281: 

4282: \bibitem{emninfl} J.~R.~Ellis, N.~E.~Mavromatos and D.~V.~Nanopoulos,

4283: %``A String scenario for inflationary cosmology,''

4284: Mod.\ Phys.\ Lett.\ A {\bf 10}, 1685 (1995)

4285: [arXiv:hep-th/9503162].

4286: %%CITATION = HEP-TH 9503162;%%

4287: 

4288: 

4289: 

4290: \bibitem{mm}  N.~E.~Mavromatos and J.~L.~Miramontes,

4291:   %``Effective Actions From The Conformal Invariance Conditions Of Bosonic Sigma

4292:   %Models With Graviton And Dilaton Backgrounds,''

4293:   Phys.\ Lett.\ B {\bf 201}, 473 (1988).

4294:   %%CITATION = PHLTA,B201,473;%%

4295: 

4296: \bibitem{type0} I. Klebanov and A.A. Tseytlin, Nucl. Phys. {\bf B546}, 155

4297: (1999); Nucl. Phys. {\bf B547}, 143 (1999).

4298: 

4299: 

4300: \bibitem{curci} G.~Curci and G.~Paffuti,

4301: %``Consistency Between The String Background Field Equation Of Motion And The

4302: %Vanishing Of The Conformal Anomaly,''

4303: Nucl.\ Phys.\ B {\bf 286}, 399 (1987).

4304: %%CITATION = NUPHA,B286,399;%%

4305: 

4306: 

4307: 

4308: 

4309: 

4310: \bibitem{carroll} S.~M.~Carroll,

4311: %``The cosmological constant,''

4312: Living Rev.\ Rel.\  {\bf 4}, 1 (2001)

4313: [arXiv:astro-ph/0004075].

4314: %%CITATION = ASTRO-PH 0004075;%%

4315: 

4316: 

4317: 

4318: \bibitem{steinhardt} A.~Upadhye, M.~Ishak and P.~J.~Steinhardt,

4319:   %``Dynamical dark energy: Current constraints and forecasts,''

4320:   arXiv:astro-ph/0411803.

4321:   %%CITATION = ASTRO-PH 0411803;%%

4322: 

4323: \bibitem{deceldata} A.~G.~Riess {\it et al.}  [Supernova Search Team Collaboration],

4324:   %``Type Ia Supernova Discoveries at z>1 From the Hubble Space Telescope:

4325:   %Evidence for Past Deceleration and Constraints on Dark Energy Evolution,''

4326:   Astrophys.\ J.\  {\bf 607}, 665 (2004)

4327:   [arXiv:astro-ph/0402512].

4328:   %%CITATION = ASTRO-PH 0402512;%%

4329: 

4330: 

4331: \bibitem{damour}  See for instance:   

4332: T.~Damour and A.~M.~Polyakov,

4333:   %``String theory and gravity,''

4334:   Gen.\ Rel.\ Grav.\  {\bf 26}, 1171 (1994)

4335:   [arXiv:gr-qc/9411069];

4336: L.~Amendola,

4337:   %``Scaling solutions in general non-minimal coupling theories,''

4338:   Phys.\ Rev.\ D {\bf 60}, 043501 (1999)

4339:   [arXiv:astro-ph/9904120].

4340:   %%CITATION = ASTRO-PH 9904120;%%

4341:   %%CITATION = GR-QC 9411069;%%

4342: R.~Bean and J.~Magueijo,

4343:   %``Dilaton-derived quintessence scenario leading naturally to the  late-time

4344:   %acceleration of the universe,''

4345:   Phys.\ Lett.\ B {\bf 517}, 177 (2001)

4346:   [arXiv:astro-ph/0007199].

4347:   %%CITATION = ASTRO-PH 0007199;%%

4348: 

4349: \bibitem{gasperini} M.~Gasperini, F.~Piazza and G.~Veneziano,

4350:   %``Quintessence as a run-away dilaton,''

4351:   Phys.\ Rev.\ D {\bf 65}, 023508 (2002)

4352:   [arXiv:gr-qc/0108016].

4353:   %%CITATION = GR-QC 0108016;%%

4354: 

4355: \bibitem{miramontes} N.~E.~Mavromatos and J.~L.~Miramontes,

4356:   %``Effective Actions From The Conformal Invariance Conditions Of Bosonic Sigma

4357:   %Models With Graviton And Dilaton Backgrounds,''

4358:   Phys.\ Lett.\ B {\bf 201}, 473 (1988).

4359:   %%CITATION = PHLTA,B201,473;%%

4360: 

4361: 

4362: 

4363: \bibitem{bachas} C.~Bachas,

4364: %``A Way to break supersymmetry,''

4365: arXiv:hep-th/9503030.

4366: %%CITATION = HEP-TH 9503030;%%

4367: 

4368: \bibitem{kogan} I.~I.~Kogan and N.~E.~Mavromatos,

4369: %``World-Sheet Logarithmic Operators and Target Space Symmetries in String Theory,''

4370: Phys.\ Lett.\ B {\bf 375}, 111 (1996)

4371: [arXiv:hep-th/9512210];

4372: %%CITATION = HEP-TH 9512210;%%

4373: I.~I.~Kogan, N.~E.~Mavromatos and J.~F.~Wheater,

4374: %``D-brane recoil and logarithmic operators,''

4375: Phys.\ Lett.\ B {\bf 387}, 483 (1996)

4376: [arXiv:hep-th/9606102];

4377: %%CITATION = HEP-TH 9606102;%%

4378: J.~R.~Ellis, N.~E.~Mavromatos and D.~V.~Nanopoulos,

4379: %``D Branes from Liouville Strings,''

4380: Int.\ J.\ Mod.\ Phys.\ A {\bf 12}, 2639 (1997)

4381: [arXiv:hep-th/9605046];

4382: %%CITATION = HEP-TH 9605046;%%

4383: N.~E.~Mavromatos and R.~J.~Szabo,

4384: %``Matrix D-brane dynamics, logarithmic operators and quantization of  noncommutative spacetime,''

4385: Phys.\ Rev.\ D {\bf 59}, 104018 (1999)

4386: [arXiv:hep-th/9808124];

4387: %%CITATION = HEP-TH 9808124;%%

4388: %``D-brane dynamics and logarithmic superconformal algebras,''

4389: JHEP {\bf 0110}, 027 (2001)

4390: [arXiv:hep-th/0106259].

4391: %%CITATION = HEP-TH 0106259;%%

4392: 

4393: \bibitem{Bergshoeff:2001pv}

4394:   E.~Bergshoeff, R.~Kallosh, T.~Ortin, D.~Roest and A.~Van Proeyen,

4395:   %``New formulations of D = 10 supersymmetry and D8 - O8 domain walls,''

4396:   Class.\ Quant.\ Grav.\  {\bf 18}, 3359 (2001)

4397:   [arXiv:hep-th/0103233].

4398:   %%CITATION = HEP-TH 0103233;%%

4399: 

4400: \bibitem{Kallosh:2001zc}

4401:   R.~Kallosh, S.~Prokushkin and M.~Shmakova,

4402:   %``Domain walls with strings attached,''

4403:   JHEP {\bf 0107}, 023 (2001)

4404:   [arXiv:hep-th/0107097].

4405:   %%CITATION = HEP-TH 0107097;%%

4406: 

4407: 

4408: 

4409: \bibitem{Villadoro:2005cu}

4410:   G.~Villadoro and F.~Zwirner,

4411:   %``N = 1 effective potential from dual type-IIA D6/O6 orientifolds with

4412:   %general fluxes,''

4413:   JHEP {\bf 0506}, 047 (2005)

4414:   [arXiv:hep-th/0503169].

4415:   %%CITATION = HEP-TH 0503169;%%

4416: 

4417: 

4418: \bibitem{Scherk:1979zr}

4419:   J.~Scherk and J.~H.~Schwarz,

4420: %``How To Get Masses From Extra Dimensions,''

4421:   Nucl.\ Phys.\ B {\bf 153}, 61 (1979).

4422:   %%CITATION = NUPHA,B153,61;%%

4423: 

4424: 

4425: 

4426: \bibitem{Dall'Agata:2005ff}

4427:   G.~Dall'Agata and S.~Ferrara,

4428:   %``Gauged supergravity algebras from twisted tori compactifications with

4429:   %fluxes,''

4430:   Nucl.\ Phys.\ B {\bf 717}, 223 (2005)

4431:   [arXiv:hep-th/0502066].

4432:   %%CITATION = HEP-TH 0502066;%%

4433: 

4434: \bibitem{DeWolfe:2005uu}

4435:   O.~DeWolfe, A.~Giryavets, S.~Kachru and W.~Taylor,

4436:   %``Type IIA moduli stabilization,''

4437:   arXiv:hep-th/0505160.

4438:   %%CITATION = HEP-TH 0505160;%%

4439: 

4440: \bibitem{Camara:2005dc}

4441:   P.~G.~Camara, A.~Font and L.~E.~Ibanez,

4442:   %``Fluxes, moduli fixing and MSSM-like vacua in a simple IIA orientifold,''

4443:   arXiv:hep-th/0506066.

4444:   %%CITATION = HEP-TH 0506066;%%

4445: 

4446: 

4447: 

4448: 

4449: \bibitem{Blumenhagen:2002wn}

4450:   R.~Blumenhagen, V.~Braun, B.~Kors and D.~Lust,

4451:   %``Orientifolds of K3 and Calabi-Yau manifolds with intersecting D-branes,''

4452:   JHEP {\bf 0207}, 026 (2002)

4453:   [arXiv:hep-th/0206038].

4454:   %%CITATION = HEP-TH 0206038;%%

4455: 

4456: \bibitem{Honecker:2002hp}

4457:   G.~Honecker,

4458:   %``Intersecting brane world models from D8-branes on  (T(2) x T(4)/Z(3))/Omega

4459:   %R(1) type IIA orientifolds,''

4460:   JHEP {\bf 0201}, 025 (2002)

4461:   [arXiv:hep-th/0201037].

4462:   %%CITATION = HEP-TH 0201037;%%

4463: 

4464: \bibitem{Freedman:1999gp}

4465:   D.~Z.~Freedman, S.~S.~Gubser, K.~Pilch and N.~P.~Warner,

4466:   %``Renormalization group flows from holography supersymmetry and a

4467:   %c-theorem,''

4468:   Adv.\ Theor.\ Math.\ Phys.\  {\bf 3}, 363 (1999)

4469:   [arXiv:hep-th/9904017].

4470:   %%CITATION = HEP-TH 9904017;%%

4471: 

4472: 

4473: \bibitem{Horava:1996ma}

4474:   P.~Horava and E.~Witten,

4475:   %``Eleven-Dimensional Supergravity on a Manifold with Boundary,''

4476:   Nucl.\ Phys.\ B {\bf 475}, 94 (1996)

4477:   [arXiv:hep-th/9603142].

4478:   %%CITATION = HEP-TH 9603142;%%

4479: 

4480: 

4481: \bibitem{Lukas:1998yy}

4482:   A.~Lukas, B.~A.~Ovrut, K.~S.~Stelle and D.~Waldram,

4483:   %``The universe as a domain wall,''

4484:   Phys.\ Rev.\ D {\bf 59}, 086001 (1999)

4485:   [arXiv:hep-th/9803235].

4486:   %%CITATION = HEP-TH 9803235;%%

4487: 

4488: 

4489: \bibitem{Stelle:1998yv}

4490:   K.~S.~Stelle,

4491:   %``Domain walls and the universe,''

4492:   arXiv:hep-th/9812086.

4493:   %%CITATION = HEP-TH 9812086;%%

4494: 

4495: 

4496: 

4497: \bibitem{dudas} E.~Dudas and C.~Timirgaziu,

4498: %``Internal magnetic fields and supersymmetry in orientifolds,''

4499: arXiv:hep-th/0502085, and references therein.

4500: %%CITATION = HEP-TH 0502085;%%

4501: 

4502: 

4503: 

4504: 

4505: \bibitem{papavass} N.~E.~Mavromatos and J.~Papavassiliou,

4506:   %``Super heavy dark matter anisotropies from D-particles in the early

4507:   %universe,''

4508:   Int.\ J.\ Mod.\ Phys.\ A {\bf 19}, 2355 (2004)

4509:   [arXiv:hep-th/0307028].

4510:   %%CITATION = HEP-TH 0307028;%%

4511: 

4512: 

4513: \bibitem{emnequiv} J.~R.~Ellis, N.~E.~Mavromatos, D.~V.~Nanopoulos and A.~S.~Sakharov,

4514: %``Space-time foam may violate the principle of equivalence,''

4515: Int.\ J.\ Mod.\ Phys.\ A {\bf 19}, 4413 (2004)

4516: [arXiv:gr-qc/0312044].

4517: %%CITATION = GR-QC 0312044;%%

4518: 

4519: 

4520: \bibitem{rostant} C.~Kounnas and B.~Rostand,

4521: %``Coordinate Dependent Compactifications And Discrete Symmetries,''

4522: Nucl.\ Phys.\ B {\bf 341}, 641 (1990).

4523: %%CITATION = NUPHA,B341,641;%%

4524: 

4525: 

4526: 

4527: \bibitem{Antoniadis:1997wu}

4528:   I.~Antoniadis and M.~Quiros,

4529: %``Supersymmetry breaking in M-theory,''

4530:   Nucl.\ Phys.\ Proc.\ Suppl.\  {\bf 62}, 312 (1998)

4531:   [arXiv:hep-th/9709023].

4532:   %%CITATION = HEP-TH 9709023;%%

4533: 

4534: 

4535: \bibitem{Atick:1988si}

4536:   J.~J.~Atick and E.~Witten,

4537:  % ``The Hagedorn Transition And The Number Of Degrees Of Freedom Of

4538:   Nucl.\ Phys.\ B {\bf 310}, 291 (1988).

4539:   %%CITATION = NUPHA,B310,291;%%

4540: 

4541: 

4542: \bibitem{O'Brien:1987pn}

4543:   K.~H.~O'Brien and C.~I.~Tan, 

4544:   %``Modular Invariance Of Thermopartition Function And Global Phase  Structure  

4545:   %Of Heterotic String,''

4546:   Phys.\ Rev.\ D {\bf 36}, 1184 (1987).

4547:   %%CITATION = PHRVA,D36,1184;%%

4548: 

4549: 

4550: 

4551: 

4552: 

4553: \bibitem{noscale} E.~Cremmer, S.~Ferrara, C.~Kounnas and D.~V.~Nanopoulos,

4554:   %``Naturally Vanishing Cosmological Constant In N=1 Supergravity,''

4555:   Phys.\ Lett.\ B {\bf 133}, 61 (1983);

4556:   %%CITATION = PHLTA,B133,61;%%

4557:   J.~R.~Ellis, A.~B.~Lahanas, D.~V.~Nanopoulos and K.~Tamvakis,

4558:   %``No - Scale Supersymmetric Standard Model,''

4559:   Phys.\ Lett.\ B {\bf 134}, 429 (1984);

4560:   %%CITATION = PHLTA,B134,429;%%

4561:  for reviews see: H.~P.~Nilles,

4562:   %``Supersymmetry, Supergravity And Particle Physics,''

4563:   Phys.\ Rept.\  {\bf 110}, 1 (1984);

4564:   %%CITATION = PRPLC,110,1;%%

4565: A.~B.~Lahanas and D.~V.~Nanopoulos,

4566:   %``The Road To No Scale Supergravity,''

4567:   Phys.\ Rept.\  {\bf 145}, 1 (1987).

4568:   %%CITATION = PRPLC,145,1;%%

4569: 

4570: \bibitem{mavromatosdecoh}  N.~E.~Mavromatos,

4571:   %``CPT violation and decoherence in quantum gravity,''

4572:   arXiv:gr-qc/0407005, Lect. Notes Phys. (Springer) in press, and 

4573: references therein. .  

4574:   %%CITATION = GR-QC 0407005;%%

4575: 

4576: \end{thebibliography}

4577: 

4578: \end{document}

4579: 

4580: \bibitem{CY} T.~Hertog, G.~T.~Horowitz and K.~Maeda,

4581:   %``Negative energy density in Calabi-Yau compactifications,''

4582:   JHEP {\bf 0305}, 060 (2003)

4583:   [arXiv:hep-th/0304199].

4584:   %%CITATION = HEP-TH 0304199;%%

4585: 

4586: \bibitem{KKmodes} M.~Minamitsuji, M.~Sasaki and D.~Langlois,

4587:   %``Kaluza-Klein gravitons are negative energy dust in brane cosmology,''

4588:   Phys.\ Rev.\ D {\bf 71}, 084019 (2005)

4589:   [arXiv:gr-qc/0501086].

4590:   %%CITATION = GR-QC 0501086;%%

4591: 

4592: 

4593: 

4594: 

4595: The predictions from our model can be compared with the cosmic microwave

4596: background data. Such analyses are usually performed in terms of the

4597: conventional $\Omega_i$ and their ratios to the conventional critical

4598: density $6H^2$. One can easily adapt these analyses to our model, using

4599: now the quantities $\Omega'_i$ (\ref{newomegas}). In the conventional FRW

4600: cosmology the position of the first acoustic peak determines the total

4601: $\Omega_{\rm total}$, and the data favour a value close to one, thereby

4602: providing evidence for a spatially flat FRW Universe. Although a detailed

4603: analysis for our model is still pending, it is to be expected that the

4604: same flat-space limit would be obtained from a fit to the data in our

4605: case. However, we postpone a detailed analysis of this point for future

4606: work.

4607: 

4608: Using the definition (\ref{newomegas}) 

4609: and taking into account the

4610: equation of state for matter $p=0$ and dilaton $p_\Phi = w_\Phi \rho_\Phi$,

4611: the second of eqs.~(\ref{eqsmotion}) can be expressed (for the current epoch) 

4612: in terms of $\Omega'$ as:

4613: \begin{equation}\label{omega-w}

4614: \frac{1}{3}\left(\frac{1 - 2q}{2 - q}\right) + w_\Phi \Omega'_\Phi = 0.

4615: \end{equation}

4616: Using this equation we can determine $\Omega'_\Phi$ in our model,

4617: in the presence of matter. If we use the value (\ref{eqnstatedilnum3}) 

4618: for $w_\Phi$ 

4619: and the present-era value (\ref{qval}) for the deceleration parameter,

4620: we infer from (\ref{omega-w}) the following present-day 

4621: value for $\Omega'_\Phi$:

4622: \begin{equation}\label{omegaprimetoday}

4623: {\Omega'_\Phi}^0 \simeq 0.34.

4624: \end{equation}

4625: On the other hand, for $q=-0.25$ and $w_\Phi \sim -0.78$, 

4626: one obtains ${\Omega'_\Phi}^0 \simeq 0.28$. 

4627: The density fractions are different from the dark energy content inferred 

4628: in a conventional FRW model, because

4629: of the different critical density traceable to the

4630: off-shell Liouville modifications in (\ref{eqsmotion}). 

4631: The value (\ref{omegaprimetoday}) is a prediction of our model, 

4632: and needs to be confronted 

4633: directly with the data, by appropriately fitting all the cosmological 

4634: data within our Liouville cosmology

4635: framework, taking into account the above-mentioned differences in the 

4636: underlying dynamics. 

4637: 

4638: We now comment on the matter content of the Liouville

4639: models which satisfy (\ref{rhoexotic2}).

4640: Recall first that this relation stems from 

4641: assuming the form (\ref{einstmetr}) for the 

4642: space-time metric in the presence of 

4643: matter. Referrings to (\ref{rhoexotic2}) and (\ref{sumomegas}), we then 

4644: observe

4645: that the dust-type matter in our models \emph{must} have negative energy

4646: density. In fact, taking into account (\ref{omegaprimetoday}), we obtain

4647: ${\Omega'}^0_{\rm exotic} \sim 3.25 {\Omega'}^0_\Phi > 0$  

4648: and ${\Omega'}^0_{\rm dust} \simeq -1.32{\Omega'}^0_\Phi < 0$.

4649: Negative-energy fields are not uncommon 

4650: in (relativistic) quantum field theories,

4651: especially in string/brane-inspired models after compactification.

4652: For instance, negative-energy four-dimensional configurations may

4653: arise from Calabi-Yau compactification in string theory~\cite{CY}, 

4654: and Kaluza-Klein graviton modes may behave as 

4655: negative-energy dust in brane models~\cite{KKmodes}. 

4656: %Thus, we may view our effective $Q$-cosmologies with negative-energy 

4657: %dust,

4658: %which appears necessary in order to accommodate the metric 

4659: %(\ref{einstmetr})

4660: %in a non-trivial matter environment, as pertaining to such microscopic

4661: %models. 

4662: However, the bulk of our matter is characterized by a positive energy 

4663: density, $\rho_{\rm exotic}$, which

4664: scales with the scale factor of the Universe in the same way as

4665: the dilaton dark-energy contribution. Thus there is no problem with 

4666: positive-energy theorems or the stability of our space-time configuration. 

4667: 

4668: However, one may also consider a more general class of Q-cosmologies,

4669: where the back-reaction effects of matter change the form of the 

4670: metric (\ref{einstmetr}) and the off-shell Liouville contributions 

4671: ${\cal J}_{\mu\nu}$, in such a way that only positive energy-density 

4672: matter appears. For instance, assume for the sake of simplicity that

4673: the back-reaction effects of matter in the space-time geometry 

4674: bring mild changes as compared to the case of~\cite{dgmpp},

4675: in the sense that the $a^{-2}$ scaling of the 

4676: dilaton energy density and  

4677: the off-shell Liouville terms is maintained, but the 

4678: relative numerical coefficients change in 

4679: such a way that (\ref{rhoexotic2}) is now replaced by 

4680: the more general relation 

4681: \begin{equation}

4682: \rho_{\rm exotic}^0 \sim \omega \rho_\Phi^0~,

4683: \label{rhoexotic3}

4684: \end{equation}

4685: where $\omega > 0$ is a phenomenological parameter, which may take a value

4686: such that ${\Omega'}_{\rm dust}^0 > 0$.

4687: As we see later, this guarantees

4688: the existence of a previous deceleration era of the Universe at redshifts

4689: near $z = 1$, as suggested by the cosmological data~\cite{deceldata}. 

4690: 

4691: We note that, using (\ref{sumomegas}) and ignoring radiation, which is 

4692: negligible as long as $1 + z \ll 10^3$, we may rewrite (\ref{omega-w}) 

4693: for the present era in the form

4694: \begin{equation}\label{q-w}

4695: q = \frac{\Omega'_M + \Omega'_\Phi ( 1 + 6w_\Phi)}{2 + 3w_\Phi 

4696: \Omega'_\Phi},

4697: \end{equation} 

4698: where $\Omega'_M$ includes dark matter, which dominates the matter 

4699: contributions. Eq.~(\ref{q-w}) should be compared with the 

4700: corresponding equation in conventional flat FRW quintessence 

4701: cosmologies with $\Omega_M + \Omega_{quint} = 1$ (again 

4702: ignoring radiation), which reads: 

4703: $q = \frac{1}{2}\Omega_M + \frac{1 + 

4704: 3w_{\rm quint}}{2}\Omega_{quint}$~\footnote{This conventional equation 

4705: could have been derived from our 

4706: second equation (\ref{eqsmotion}), if one used the conventional

4707: critical value $\rho_c = 6H^2$ and assumed 

4708: that $\Omega_M + \Omega_\Phi = 1$. However, this is not valid in our case.

4709: The formal reason for this is that in our model

4710: ${\cal J}_{ii} \simeq 0$ for large times, and so 

4711: there is no Liouville modification in this component of the Einstein 

4712: equation.}.

4713: 

4714: Notice that if one uses the critical density for a given redshift $z$

4715: at late epochs, (\ref{criticaldens}), so as to construct 

4716: the $z$-dependent $\Omega'_i (z)$, then the relations (\ref{sumomegas}) and 

4717: (\ref{q-w}) are exact, and valid for every $z$ at late epochs.

4718: This is due to the fact that our model predicts a constant

4719: equation of state (\ref{eqnstatedilnum}) at late eras.

4720: 

4721: However, as becomes clear from (\ref{sumomegas}),(\ref{q-w}), there is no

4722: deceleration for $z < 0.66$, i.e., in the entire regime of $z$ for which

4723: our analytic treatment is valid. This was expected of course from

4724: (\ref{decel4}), which is based on (\ref{einstmetr}).  Moreover, we notice

4725: that if one attempts to extrapolate (\ref{q-w}) outside this regime of

4726: $z$, one encounters an unphysical \emph{discontinuity} (divergence)  in

4727: the value of $q(z)$ for $z=z^*$ such that $1 + 6w_\Phi \Omega'_\Phi (z^*)=

4728: -3$, after which $q(z)$ becomes positive. However, it can readily be seen

4729: that $z^* \sim 0.9$, i.e., outside the regime of the validity of our

4730: approximations. In conventional FRW cosmologies the fit to the data yield

4731: $z^* \sim 0.5$~\cite{deceldata}. To determine whether such a $z^*$ exists

4732: or not in our Q-cosmology framework one should actually abandon the form

4733: (\ref{einstmetr}) and most of the current formalism, and attempt to

4734: determine the correct scale factor by solving the appropriate equations

4735: (\ref{eqsmotion})  for $z > 0.66$. This depends on the details of the

4736: matter action, as well as the underlying string theory structures.

4737: 

4738: Our approximation before was based on scale factors of the form

4739: (\ref{einstmetr}) and a linear dilaton, assumptions that are not valid for

4740: $z > 0.66$ as we have seen above. Above such redshifts the correct scale

4741: factor should be determined by the back-reaction of matter, which is

4742: dominant, and is determined in principle by solving the appropriate

4743: equations of motion (\ref{eqsmotion}). However, in such regimes the form

4744: (\ref{jcalvalues}) may not be valid. This would eliminate the fake

4745: discontinuity of $q(z)$ at $z^*$, which is only an artefact of the

4746: incorrect extrapolation of our current formalism to the regime $z > 0.66$.  

4747: It may not be impossible that in such a case a $z^*$ would be close to the

4748: value $0.5$ suggested by the present data~\cite{deceldata}.

4749: In this respect, it should also be noted that the constant equation of

4750: state (\ref{eqnstatedilnum}) used above is only an approximation, which is

4751: likely to characterize accurately only an asymptotic era when the

4752: gravitational sector dominates. In the presence of matter, in particular

4753: in eras with $z > z_0$, $z_0 \le 0.66$, where our approximations in this

4754: work for the scale factor (\ref{einstmetr}) and the (logarithmic in cosmic

4755: time) dilaton may fail, one would most likely find a $z$-dependent $w(z)$

4756: for the dark energy sector, which in our model can only be predicted with

4757: accuracy if details of the underlying string/brane theory are taken into

4758: account.

4759: 

4760: It then becomes apparent 

4761: that such a 

4762: $w(z)$ above $z_0$ is also capable of eliminating 

4763: the above-mentioned unphysical singularity

4764: in $q(z)$. Indeed, the root of that singularity is 

4765: the presence

4766: of the term $(2-q)$ in the denominator of (\ref{omega-w}).

4767: Assuming for $z > z_0$ the relation  

4768: \begin{equation}

4769: -w(z)(2-q(z)) \simeq \gamma (z)>0~, 

4770: \label{waccel}

4771: \end{equation}

4772: with the function $\gamma (z)$ of such a form that

4773: $w(z)$ is slightly increases monotonically with $z > 0$,

4774: as $q(z)$ approaches zero from negative values, as seems to 

4775: be the case indicated by data~\cite{steinhardt}~\footnote{For instance, 

4776: as an indicative example, 

4777: assume the form $w(z) = -|w_0|$ for $z < z_0$ and 

4778: $w(z) = -|w_0|  + |w_1|(z-z_0)$ for $z > z_0$, and also 

4779: $q(z) = -|q_0| + |q_1| z$ and  

4780: $\gamma (z) = \gamma_0 + \gamma_1 z$. For sufficiedntly small

4781: $q_1,\gamma_1, w_1$, and keeping first-order terms in the expansion, 

4782: the function $\gamma (z)$,  for which (\ref{waccel}) is valid, reads:

4783: $\gamma (z) = (2 + |q_0|)(|w_0| + |w_1|z_0)  

4784: - [(|w_0|+ |w_1|z_0)|q_1| + (2 + |q_0|)|w_1|]z$,

4785: for $z > z_0$.} and insisting on the 

4786: (approximate) validity of the form 

4787: of the off-shell Liouville 

4788: terms ${\cal J}_{00} \simeq 6H^2(1-q(z))$ and ${\cal 

4789: J}_{ii} \simeq 0$ 

4790: (which, though, may oversimplifications, as already mentioned),

4791: we then obtain from the first two equations of (\ref{eqsmotion}):

4792: \begin{equation}

4793: q = \frac{1}{2}\left(1 - 3\gamma (z) \Omega'_\Phi (z)\right).

4794: \label{qzz0}

4795: \end{equation}

4796: Notice that there is no 

4797: singularity in this expression for $q(z)$, which is $< 1/2$, 

4798: due to the fact that $\gamma (z) \Omega '_\Phi > 0$.

4799: 

4800: 

4801: 

4802: To determine $z^*$ from (\ref{qzz0}), it is convenient to use the

4803: expression $\Omega'_\Phi (z) = \frac{1}{1 + \rho_{\rm Matter}/\rho_\Phi}$

4804: obtained from the first of eqs. (\ref{eqsmotion}). As a simplifying

4805: assumption, we may maintain the scaling (\ref{mixedscaling}) for the

4806: matter sector (including the dominant dark matter): $\rho_{\rm Matter} =

4807: \rho_{\rm dust}^0 (1 + z)^3 + \rho_{\rm exotic}^0(1 + z)^2$, found above

4808: for $z < z_0$, also for redshifts $z > z_0$, with $\rho_{\rm dust}^0 +

4809: \rho_{\rm exotic}^0 + \rho_\Phi^0 = 1$, for our spatially flat Universe.

4810: Then we obtain a past deceleration era for $z$ such that:

4811: \begin{equation}

4812: \frac{\rho_{\rm dust}^0}{\rho_\Phi^0}(1 + z) + 

4813: \frac{\rho_{\rm exotic}^0}{\rho_\Phi^0} > 3\gamma (z)> 0,

4814: \label{rhogamma}

4815: \end{equation}

4816: from which we obtain the following expression for the critical $z^*$: 

4817: \begin{equation}

4818: z^* = (3\gamma (z^*) + 1)\frac{{\Omega'}_\Phi^0}{{\Omega'}_{\rm dust}^0} - 

4819: \frac{1}{{\Omega'}_{\rm dust}^0},

4820: \label{zstarcorrect}

4821: \end{equation}

4822: where we used (\ref{sumomegas}). 

4823: There are natural values of the various parameters

4824: for which $z^* \simeq 0.5$ as required by the data.

4825: In such a case, the present era for which the approximations 

4826: (\ref{einstmetr}) and logarithmic dilaton are valid applies to

4827: $z< z_0$ with $z_0 < 0.5$, which is consistent with our formalism 

4828: above. 

4829: 

4830: In fact, we can fit the various parameters with the current

4831: data~\cite{steinhardt}, which indicate a $z$-dependent $w(z)$ for the dark

4832: energy sector.  In this respect, recall that the present-era analysis

4833: earlier in this work indicated ((\ref{omegaprimetoday})) the value

4834: ${\Omega'}_\Phi^0 = 0.34$. We have also seen previously

4835: ((\ref{rhoexotic3})) that ${\Omega'}_{\rm exotic}^0/{\Omega'}_\Phi =

4836: \omega $, with $\omega $ a phenomenological parameter that guarantees the

4837: positivity of the dust energy density. From (\ref{sumomegas}) we have:

4838: ${\Omega'}_{\rm dust}^0/{\Omega'}_\Phi^0 = \frac{1}{{\Omega'}_\Phi^0} - 1

4839: - \omega = 1.94 - \omega >0$, which imposes appropriate constraints on the

4840: allowed range of $\omega$ if we restrict our attention to models in which

4841: the dust has positive energy density~\footnote{We note at this stage that,

4842: for models in which ${\Omega'}_{\rm dust}^0 < 0$ (c.f.

4843: (\ref{rhoexotic2})), there is no era of past deceleration in the model,

4844: for the phenomenologically relevant values of the cosmological parameters

4845: we use here.}. Then, from (\ref{zstarcorrect}) we have: 

4846: \begin{equation}\label{phidust}

4847: z^* \simeq \frac{3\gamma (z^*)- 1.94}{1.94 - \omega}~. 

4848: \end{equation}

4849: We next notice that in our

4850: simplified model above, $2w(z^*) = -\gamma (z^*)$ 

4851: (c.f. (\ref{waccel})). The WMAP data seem to 

4852: indicate that $w(z^* \sim 1/2) \simeq -0.75$, which fixes $\gamma (z^*) 

4853: \simeq 1.5$. 

4854: Then, from (\ref{phidust}) we obtain: $z^* \sim 2.56/(1.94 - \omega)$.

4855: Unfortunately, 

4856: for $1.94 > \omega > 0$, guaranteeing positivity of the 

4857: energy for the dust-type matter, 

4858: this yields $z^* > 1.3$ for past deceleration, inconsistent 

4859: with the value $z^* \simeq 0.5$ indicated by data~\cite{deceldata}. 

4860: On the other hand, if we insist on $z^* = 0.5$, then from (\ref{phidust}) 

4861: we obtain $\gamma (z^*) \simeq 0.97 -\omega/6$, corresponding to

4862: an equation of state  $w(z^*=0.5) \simeq -0.485 + \omega/12 $, 

4863: which is much smaller than that indicated by the WMAP data at 

4864: such redshifts~\cite{steinhardt}. 

4865: 

4866: However, this failure to match precisely the data may be attributed to the

4867: various approximations made, which essentially extrapolated most of the

4868: analysis valid in the present era ($z < 0.66$)  to earlier epochs,

4869: something that may not be entirely correct. In our opinion, the most

4870: appropriate approach is not to assume any specific form for the scale

4871: factor, but rather to consider specific microscopic theories, and solve

4872: the generalised conformal invariance equations (\ref{liouveq}), after

4873: compactification, in the same way as in the analysis of~\cite{dgmpp}, but

4874: now including the appropriate matter.  It may then not be impossible to

4875: find consistency of all the current data with the theoretical predictions

4876: of Q-cosmology.  We leave such detailed studies for future work.

4877: 

4878: The above, rather generic predictions, are not entirely model-independent.

4879: They are based on some string models with specific forms of flux

4880: compactification~\cite{dgmpp}.  In general, one needs detailed microscopic

4881: models with specified compactifications before definite conclusions can be

4882: reached regarding the critical redshift for the deceleration/acceleration

4883: transition and/or the energy-density budget of the Liouville Universe.

4884: Moreover, the reader should also bear in mind that there are

4885: non-equilibrium modifications to the corresponding equations

4886: (\ref{eqsmotion}) underlying the dynamics of our system, whose value

4887: depends on the details of the microscopic string/brane theory at hand. In

4888: our analysis above, we have estimated such contributions as being of order

4889: $-6H^2(1-q)$ but the precise numerical coefficient (6)  is in general

4890: different for different string models.

4891: 

4892: It is clear from the above discussion that details of the underlying

4893: string theory are important for making definite predictions concerning the

4894: various cosmological parameters underlying a Q-cosmology model.  In view

4895: of these theoretical uncertainties, as well as the different Friedmann

4896: equation and the modified critical density of a flat Universe within our

4897: framework, one must perform a new fit to these data using the details of

4898: our model in order to obtain accurate information on quantities such as

4899: $z^*$ or $\Omega'_i$, which will then guide theoretical model building.  

4900: This is the most appropriate way to deal with our generic Q-Cosmology

4901: framework at present.  We plan to make such a study in a future

4902: publication.

4903: 

4904: Notice also that, as discussed in detail in the framework of~\cite{dgmpp},

4905: there is a phase in the past history of that Universe in which the vacuum

4906: energy density of the dilaton field drops towards zero. This is due to the

4907: fact that both the kinetic energy of the dilaton field, which approaches a

4908: constant as one travels back in cosmic time~\footnote{In fact, as we have

4909: seen in previous Sections, and shall discuss further in Section 3 in the

4910: context of specific models, the dilaton remains constant during the

4911: inflationary era, and starts becoming non-trivial (linear)  only much

4912: after the exit from that phase.}, and the central-charge deficit $Q^2$,

4913: approach zero.  The central charge increases again for even earlier eras

4914: to approach a constant value during inflation~\cite{dgmpp}.  This makes

4915: the dilaton in the model a perfect tracking field for quintessence, which

4916: operates in the current era of the Universe but decouples in the past.

4917: 

4918: The era with vanishing vacuum energy is a metastable one~\cite{dgmpp},

4919: which may be appropriate for nucleosynthesis. As we discuss later on, due

4920: to the delicate balance between nuclear reaction rates and the expansion

4921: of the Universe that are needed for the formation of the elements, a

4922: vanishing (or at least very small compared to matter) vacuum energy

4923: density may be required for the nucleosynthesis era. In Section 3 we

4924: describe specific models with such features.  After the creation of

4925: matter, the dilaton picks up a logarithmic dependence on cosmic time in

4926: the Einstein frame (i.e., a linear dilaton in the string frame), which

4927: survives asymptotically in time, eventually becoming the dominant cause

4928: for the acceleration of the Universe. In the scenario~\cite{emn04}, which

4929: we have checked above, this linear-dilaton behaviour already describes the

4930: present era of the Universe, in agreement with current phenomenology.  In

4931: Section 3 we present a rather detailed mathematical description of the

4932: various phases of such a Universe.

4933: 

4934: \paragraph{}

4935: \noindent \emph{A Cosmological \underline{Constant} in the asymptotic regime of the Colliding 

4936: Branes Model} 

4937: \paragraph{} 

4938: Before proceeding to an analysis of thermal effects in the early Universe

4939: we would like first to discuss the r\^ole of the D-particles in the bulk 

4940: of the model of ref. \cite{emw} in inducing a present-day (almost) 

4941: \emph{constant} contribution to the dark energy. 

4942: 

4943: The starting point is the effective action 

4944: (\ref{kalloshact}). The action consists of a field theoretic 

4945: part, described by the first line, while the other two

4946: lines describe the coupling to D8/Orientifold planes and 

4947: D0/fundamental string(F1) sources. Our focus will be on this 

4948: last D0/F1 term.                              

4949: As a result of the existence of only one component 

4950: of the metric tensor inside the determinant $\sqrt{-G_{tt}}$, 

4951: where $G_{\mu\nu}$ denotes

4952: the $\sigma$-model-frame metric, we observe that upon 

4953: considering cosmological spatially-homogeneous backgrounds, 

4954: performing appropriate compactification to four space-time dimensions, 

4955: which we shall not discuss in detail here, 

4956: and passing 

4957: onto the Einstein frame (\ref{smodeinst}) with metric $g_{\mu\nu}^E$, $\mu,\nu =0,\dots 3$, 

4958: the dilaton factors disappear, yielding 

4959: effective action terms of the form

4960: \begin{equation}\label{t2} 

4961: S \ni -T_2 \int dt R \sqrt{-g_{00}^E}\sum_k N_k 

4962: \end{equation}

4963: where $T_2 = 1/4\pi \alpha '$ is the string tension,  

4964: $R$ is the bulk distance between the 

4965: two orientifold planes in the model 

4966: of \cite{emw} (c.f. fig. \ref{fig:nonchiral}), and 

4967: $\sum_k N_k$ denotes the total number of D-particles on a 

4968: four-dimensional (eight-dimensional compactified) brane 

4969: in the model of ref. \cite{emw}, 

4970: representing our world.

4971: We then represent 

4972: $R\sum_k N_k = \int \sqrt{-g} d^{9}X n $, where $n$ is the density of 

4973: D-particle defects in the bulk of our 

4974: configuration depicted in fig. \ref{fig:nonchiral}, 

4975: after compactification to four dimensions, with the compact and bulk spaces 

4976: having been stabilised.

4977: Taking into accout that in the cosmological backgrounds we consider 

4978: we work in the gauge $g_{00}=-1$, 

4979: we may then write 

4980: (\ref{t2}) in the form of a vacuum energy contribution

4981: in the Einstein frame of the string, 

4982: \begin{equation}\label{t2vac}

4983: S \ni -\int d^{4}x \sqrt{-g^E} \rho 

4984: \end{equation} 

4985: where $\rho$ represents now a contribution to the four-dimensional

4986: vacuum energy density. Notice that in this expression 

4987: \emph{there is no}  dilaton scaling factor.

4988: This should be contrasted with the dark-energy contribution of  

4989: the central charge deficit $Q^2$, 

4990: expressing

4991: non-criticality of the $\sigma$ model pertaining to the gravitational

4992: degrees of freedom of the string theory at hand. 

4993: Indeed this contributes to the Dark energy a term of the form

4994: (\ref{centralcharge}), which, when transformed to the Einstein frame, 

4995: scales non-trivially 

4996: with the dilaton.  In four dimensions it reads

4997: \begin{equation} 

4998: -\int d^DX \sqrt{-g^E} e^{2\Phi}Q^2 

4999:  \label{centralcharge2}

5000: \end{equation}

5001: The term (\ref{centralcharge2}) should be added to the 

5002: field theoretic part (first line) of the action (\ref{kalloshact}),

5003: when one considers motion of branes, as discussed in \cite{emw} and above. 

5004: The motion breaks target-space supersymmetry, and hence contributes

5005: to a non-zero dark energy. 

5006: 

5007: In what follows we shall present a scenario in which $\rho$ in (\ref{t2vac}) 

5008: remains almost constant at present, 

5009: after compactification to four dimensions. This will appear as 

5010: a current-era 

5011: cosmological \emph{constant} 

5012: contribution to the gravitational 

5013: vacuum energy of our effective field theory (\ref{kalloshact}).

5014: 

5015: 

5016: 

5017: \paragraph{}

5018: \begin{figure}[tb]

5019: \begin{center}

5020: \includegraphics[width=4cm]{bouncebrane.eps}

5021: \end{center}

5022: \caption{\it 

5023: A scenario for obtaining an approximately constant 

5024: contribution to the dark energy at the present era

5025: in the model of supersymmetric D-particle foam of fig. \ref{fig:nonchiral}.

5026: The picture illustrates the second collision 

5027: of our brane world with the left stack of the branes.

5028: The collision is very mild, due to the very small velocities involved,

5029: as a result of a significant loss in 

5030: brane kinetic energy in the form of closed-string

5031: radiation from the brane to the bulk, during the time 

5032: between the two collisions. 

5033: This second collision, therefore, 

5034: essentially corresponds to a bounce of the brane world, involving 

5035: a minimum distance from the stack of order

5036: $v^{1/2}\ell_s$, with $v \ll 1$ the brane velocity. There is a 

5037: higher concentration of D-particles near the stack as compared to

5038: their population in the bulk. This is the result of the 

5039: initial disturbance of the bulk space-time due to the original collision

5040: of the branes.} 

5041: \label{fig:secondcollision}

5042: \end{figure}

5043: 

5044: To this end

5045: we shall concentrate on the late time

5046: regime, and \emph{assume} that the Brane world 

5047: is very near the stack of the D-branes  

5048: it emerged from initially (say the stack on the left hand side of fig. 

5049: \ref{fig:nonchiral}). 

5050: What we envisage is the following: long after the collision

5051: the brane world approaches the left stack of branes, where it emerged from 

5052: initially, but with a velocity much much smaller than the initial one

5053: after the collision.

5054: As we shall discuss

5055: in section 3, such small velocities long after the initial collision,

5056: is the result of the fact that  

5057: the brane world has lost a significant amount of 

5058: kinetic energy due to closed-string radiation in the 

5059: bulk, as a result of temperature differences between brane and 

5060: bulk worlds. 

5061: Thus, the second collision

5062: can take place at extremely mild conditions, 

5063: and one can talk about a ``bounce''

5064: rather than a catastrophic event. The brane world may approach 

5065: the stack up to a minimum

5066: distance, below the string length~\cite{emw}, 

5067: $r/\ell-s \sim v^{1/2} \ll 1$ and then  bounce back. 

5068: The regime of almost constant contribution to the four-dimensional 

5069: vacuum  energy will be precisely the regime after the bounce,

5070: where the brane moves extremely slow. Eventually, the brane world will stop,

5071: as a result of gravitational ``friction'' by the other D-particles 

5072: in the bulk, and hence the vacuum energy due to the motion of brasnes 

5073: will vanish asymptotically (equilibrium). 

5074: 

5075: 

5076: In this phase, one may also assume that 

5077: there is a  

5078: sufficiently large number $N$ of D-particles 

5079: near the brane world, as a result of the depletion of bulk D-particles 

5080: from the region of the initial collision (middle region in the symmetric

5081: scenario of fig. \ref{fig:nonchiral}), due to the disturbance 

5082: caused by the colliding branes.

5083: Under such conditions, then, 

5084: the ${\cal O}(v^4)$ term in (\ref{symmpot1})  can be ignored

5085: in front of (\ref{d0d8}) term. 

5086: 

5087: According to 

5088: the considerations of ref. \cite{emw}, 

5089: the D-particles near the brane world 

5090: will exert 

5091: repulsive forces 

5092: on the brane. Each D-particle/D8 brane interaction

5093: contributes a potential (in units of the string scale $\alpha '$):

5094: \begin{equation}\label{d0d8} 

5095: {\cal V}_{D0/D8} = -\frac{2\pi v^2}{r^3}   

5096: \end{equation} 

5097: where $r$ denotes the bulk distance of the D-particle from the D8 domain 

5098: wall. Notice that this is the dominant D-particle contribution to the 

5099: bulk potential energy of the configuration. 

5100: D-particles which are far from the brane, contribute potentials

5101: which scale like $r^7$, and hence are subdominant. 

5102: 

5103: As mentioned previosuly, 

5104: there is a minimum distance in this framework~\cite{emw}, 

5105: which is smaller

5106: than the conventional string scale minimum distance, $r \sim v^{1/2}$,

5107: for which the contribution (\ref{d0d8}) for $N$ D-particles 

5108: near the brane world yield the ${\cal O}(v^{1/2})$ terms in 

5109: (\ref{symmpot1}). 

5110: 

5111: We now notice that the total force corresponding to potentials 

5112: (\ref{d0d8}) due 

5113: to the presence of $N$ D-particles near the brane (i.e. at distances 

5114: at most $v^{1/2}$ from the brane, which from the quantum point of 

5115: view are considered as lying on the D8 brane)

5116: reads:

5117: \begin{equation}

5118: F = -\frac{d}{dr}{\cal V}_{D0/D8} = -N\frac{6\pi}{r^4}v^2|_{r\sim v^{1/2}} \sim

5119: -6\pi N = -6\pi n_8 V_8 

5120: \label{d0d8N}

5121: \end{equation}

5122: where $n_8$ is the number density of the D-particles on the brane,

5123: and $V_8$ is the eight-dimensional brane volume.

5124: We next assume a non-relativistic adiabatic motion of the bulk D8 brane 

5125: such that the force 

5126: \begin{equation}\label{force}

5127: F = \rho V_8 \frac{dv}{dt_E}~, 

5128: \end{equation} 

5129: where $M_{\rm total} = \rho_8 V_8$ is the total mass, with $\rho$ the D-brane

5130: mass (energy) density, and $t_E$ is the Einstein-frame cosmic time 

5131: (\ref{einsttime}), which constitutes the physical time for an observer

5132: living on the brane, who is in a comoving frame with the cosmic

5133: expansion on the brane world.  

5134: 

5135: From (\ref{d0d8}) and (\ref{force}), assuming appropriate compactification to 

5136: four dimensions, with stabilised internal and bulk spaces,  

5137: we obtain a deceleration:

5138: \begin{equation}\label{deceleration} 

5139: \frac{dv}{dt_E} = - 6\pi \left(\frac{n}{\rho}\right) ~\to~~ 

5140: v= v_0 -6\pi \left(\frac{n}{\rho}\right)t_E   

5141: \end{equation}

5142: where $v_0$ is the recoil velocity at the second collision,

5143: which we may assume to be vanishing $v_0 \simeq 0$.

5144: The negative value of the velocity denotes the motion of the 

5145: brane away from the left stack of fig. \ref{fig:nonchiral}, and in this 

5146: sense the D-particles accelerate the brane Universe away from the 

5147: orientifold plane. Hence

5148: \begin{equation}\label{vel}

5149: |v| \simeq 6\pi \left(\frac{n}{\rho}\right) t_E ~\ll ~1

5150: \end{equation}

5151: The current era of our Universe is defined in such a way

5152: that (\ref{vel}) is valid. This can be arranged 

5153: by appropriate adjustments of the ratio $n/\rho$ in the model. 

5154: The quantity  

5155: $n = n_8$ now denotes a three spatial dimensional 

5156: density of D-particles on the compactified three-brane world, 

5157: taking into account  

5158: that upon compactitication the 

5159: number density of D-particles did not change, due to their 

5160: point-like nature. We also ignore wrapped strings and other branes

5161: around the compact dimensions, whose r\^ole however may be non trivial.

5162: We postpone a detailed discussion on compactification for 

5163: a future publication. 

5164: 

5165: We notice that as the 

5166: brane world moves away from the stack with velocity $v$ the 

5167: number of D-particles crossing the D8 brane, and, according to 

5168: the discussion in \cite{emw}, `hooked' onto it via the 

5169: creation of a fundamental string to form a $D0F1$ configuration

5170: increases as: 

5171: \begin{equation}\label{dndt}

5172: \frac{dN}{dt_E} = \frac{d}{dt_E}\left(\rho V_8 \right) = 

5173: n_{\rm bulk} V_8 |v| = \frac{{\cal N}_{\rm total}}{R} |v| 

5174: \end{equation} 

5175: where 

5176: $R$ is the distance between the orientifolds, which defines

5177: the size of the large bulk dimension in our orientifold 

5178: compactification, 

5179: $n_{\rm bulk}$ is the bulk number density of D-particles, 

5180: and ${\cal N}_{\rm total}$ denotes the total number of D-particles

5181: in the configuration of \cite{emw} (fig. \ref{fig:nonchiral}). 

5182: This number is assumed finite and constant, as we do not consider

5183: creation of D-particles. Such a process does not take place 

5184: in our contruction due to global flux conservation. 

5185: 

5186: We have from (\ref{dndt}) in a cosmological background

5187: in which the brane four dimensions expand with a scale factor

5188: $a(t_E)$ in the Eisntein frame, while the compact and bulk dimensions

5189: remain frozen (i.e. $V_8 \sim a^3$, $H \equiv a^{-1}da/dt_E$):

5190: \begin{equation}\label{drdt} 

5191: \frac{d\rho}{dt_E} \sim a^{-3}\left(- 3(H/a)\rho + 

5192: a^3\frac{{\cal N}_{\rm total}}{R} |v|\right) 

5193: \end{equation}

5194: Using (\ref{vel}) we finally have:

5195: \begin{equation}\label{drdt2} 

5196: \frac{d\rho}{dt_E} \sim a^{-3}\left(- 3(H/a)\rho + 

5197: a^3~6\pi\frac{{\cal N}_{\rm total}}{R} 

5198: \left(\frac{n}{\rho}\right)t_E\right) 

5199: \end{equation}

5200: Assuming that we are in the large (asymptoptic) cosmic time regime

5201: (\ref{condition}),

5202: for which the scale factor is given by (\ref{einstmetr2}), and the 

5203: Hubble parameter by 

5204: (\ref{hubble2}), we have: $Ha^2 \sim F_1^2t_E $, with $F_1 \sim Q_0/2$

5205: $Q_0^2$ the asymptoptic central charge of the $\sigma$ model 

5206: describing bulk excitations. In the model considered here,

5207: with compactified dimensions in magnetized manifold

5208: with an internal magnetic field intensity ${\cal H}$,

5209: we have $Q_0 \sim {\cal H}^2$ (c.f. (\ref{sigmamodelq0})). 

5210: Note that the ratio $n/\rho$ remains constant under the cosmic expansion,

5211: and therefore we may consider it as an adjustable parameter of our model.

5212: 

5213: 

5214: 

5215: In this regime of cosmic times then, which is supposed to characterise

5216: the present-day Universe we may have an approximately constant

5217: $\rho$ today, if we require the right hand side of (\ref{drdt2}) to vanish.

5218: This implies

5219: \begin{equation}\label{constrho} 

5220: \rho \simeq \rho_0 = 24\pi~\frac{{\cal N}_{\rm total}}{{\cal H}^2~R} 

5221: \left(\frac{n}{\rho}\right)    

5222: \end{equation}

5223:  where we require that the value of $\rho_0$ is the one compatible 

5224: with the value of the 

5225: cosmological constant claimed to have been 

5226: observed today~\cite{wmap} ($\sim 10^{-120}$ in Planck units

5227: in four dimensions). Noting that in our model 

5228: we do not have really an upper bound on the size $R$ of the bulk

5229: dimension~\footnote{In fact, in orientifold

5230: models with $n$-branes in each stack the upper bound of $R$ obtained

5231: from considerations of massive type II supergravity~\cite{emw} 

5232: is proportional

5233: to $1/n-8$, and in our construction we use eight branes in each stack},

5234: we observe that the smallness of $\rho$ can be easily achieved fr 

5235: large bulk dimensions, but also for tiny values of $n/\rho$. 

5236: In particular, the last requirement

5237: leads (\ref{vel}) to very small values of the velocity, thereby

5238: proving the self-consistency of our skmall velocities

5239: approximation in the large cosmic time regime. 

5240:  

5241: The constant contribution $\rho_0$ is of comparable magnitude

5242: with the (\ref{centralcharge2}) dilaton-dependent

5243: quintessence like vacuum energy today for $\Phi \sim {\rm const} -{\rm ln}(\gamma t_E)$

5244: (c.f. (\ref{cosmoconst2}). 

5245: 

5246: It should be noted that, although, at present era the D-particle

5247: contributions lead to an \emph{approximately constant} vacuum energy

5248: contribution (\ref{t2vac}),

5249: nevertheless eventually our brane world will completely stop,

5250: as a result of gravitational friction from the bulk

5251: D-particles, due to their repulsive forces. 

5252: This will lead to an eventual restoration of bulk supersymmetry.

5253: In that limit any dark energy contribution in (\ref{symmpot1}), (\ref{d0d8})

5254: tends to 

5255: zero, thereby leading to the absence of an asymptotic 

5256: cosmic horizon in this model.

5257: Even if the branes have been compactified to magnetized tori,

5258: the associated Nielesen-Olesen instabilities will

5259: lead to de-compactification, as discussed in \cite{gravanis},

5260: and eventual cancellation of the magnetic-field induced 

5261: supersymmetry breaking.  

5262: This situation is compatible with the string theory model, since it allows

5263: the construction of an asymptotic scattering matrix. 

5264: 

5265: 

5266: 

5267: 

5268: 

5269: 

5270: We have also seen that if the second collision takes place

5271: after the nucleoysnthesis era, one may receive 

5272: approximately constant contributions (\ref{t2vac}), (\ref{constrho})

5273: today, as a result of a special choice of the parameters of 

5274: the (compactified) model of \cite{emw}. Such contributions

5275: seem to be preferred by current data~\cite{wmap}. 

5276: