1: \documentclass[prd,aps,floats,floatfix,eqsecnum,nofootinbib]{revtex4}

2: \usepackage{amsmath,amssymb,verbatim,epsfig,graphicx,rotating}

3: \usepackage{psfrag}

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

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

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

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

8: % User-defined commands

9: \begin{document}

10: \title{Single Field Inflation models allowed and ruled out by

11: the three years WMAP data}

12: \author{\bf H. J. de Vega$^{(b,a)}$}\email{devega@lpthe.jussieu.fr}

13: \author{\bf N. G. Sanchez $^{(a)}$}\email{Norma.Sanchez@obspm.fr}

14: \affiliation{$^{(a)}$

15: Observatoire de Paris, LERMA, Laboratoire Associ\'e au CNRS UMR 8112,

16:  \\61, Avenue de l'Observatoire, 75014 Paris, France. %}

17: \\$^{(b)}$ LPTHE, Laboratoire Associ\'e au CNRS UMR 7589,\\

18: Universit\'e Pierre et Marie Curie (Paris VI) et Denis Diderot (Paris VII),\\

19: Tour 24, 5 \`eme. \'etage, 4, Place Jussieu, 75252 Paris, Cedex 05,

20: France.}

21: 

22: \begin{abstract}

23: We study the single field slow-roll inflation models  that

24: better agree with the available CMB and LSS data including the three years WMAP data:

25: new inflation and hybrid inflation. We study these models as effective field 

26: theories in the Ginsburg-Landau context: a trinomial potential turns out to 

27: be a simple and well motivated model.

28: The spectral index $ n_s $ of the adiabatic fluctuations, the ratio $ r $

29: of tensor to scalar fluctuations and the running  index $ d n_s/d 

30: \ln k $ are studied in detail.  We derive explicit formulae for 

31: $ n_s , \; r $ and $ d n_s/d \ln k $ and provide relevant plots. 

32: In new inflation, and for the three years WMAP and 2dF central value $ n_s = 0.95 $,

33: we predict $ 0.03 < r < 0.04 $ and $ -0.00070 < d n_s/d \ln k < -0.00055 $. 

34: In hybrid inflation, and for $ n_s = 0.95 $,  we predict $ r \simeq 0.2 $ and 

35: $ d n_s/d \ln k \simeq -0.001 $ . Interestingly enough, 

36: we find that in new inflation $ n_s $ is {\bf bounded} from above by 

37: $ n_{s~max} = 0.961528 \ldots $ and that $ r $ is a {\bf two} valued function of $ n_s $ 

38: in the interval $ 0.96 < n_s <  n_{s~max} $. In the first branch 

39: we find $ r < r_{max} = 0.114769\ldots $. In hybrid inflation we find a 

40: critical value $ \mu_{0~crit}^2 $ for the mass parameter $ \mu_0^2 $ of the 

41: field $ \sigma $ coupled to the inflaton. 

42: For $ \mu_0^2 < \Lambda_0 \; M_{Pl}^2/192 $, where  $ \Lambda_0 $ is the 

43: cosmological constant, hybrid inflation is ruled out by the WMAP three years

44: data since it yields a blue tilted $ n_s > 1 $ behaviour. 

45: Hybrid inflation for $ \mu_0^2 > \Lambda_0 \; M_{Pl}^2/192 $ fullfills all 

46: the present CMB+LSS data for a large enough initial inflaton amplitude. 

47: Even if chaotic inflation predicts $ n_s $ values 

48: compatible with the data,  chaotic inflation is disfavoured since it predicts 

49: a too high value $ r \simeq 0.27 $ for the ratio of tensor to scalar 

50: fluctuations. The model which  best agrees with the current data and which best 

51: prepares the way to the expected data $ r \lesssim 0.1 $, is the trinomial 

52: potential with negative mass term: new inflation.

53: \end{abstract}

54: 

55: \date{\today}

56: \pacs{98.80.Cq,05.10.Cc,11.10.-z}

57: \maketitle

58: \tableofcontents

59: 

60: \section{Introduction and Results}

61: 

62: Inflation was  introduced  to solve several outstanding

63: problems of the standard Big Bang model \cite{guth} and became an important

64: part of the standard cosmology. At the same time, it provides a natural

65: mechanism for the generation of scalar density fluctuations that

66: seed large scale structure, thus explaining the origin of the

67: temperature anisotropies in the cosmic microwave background (CMB),

68: as well as that of  tensor perturbations (primordial gravitational

69: waves)\cite{mukyotr,libros}. 

70: 

71: \medskip

72: 

73: A distinct aspect of

74: inflationary perturbations is that these are generated by quantum

75: fluctuations of the scalar field(s) that drive inflation. After

76: their wavelength becomes larger than the Hubble radius, these

77: fluctuations are amplified and grow, becoming classical and

78: decoupling from  causal microphysical processes. Upon re-entering

79: the horizon, during the matter era, these classical perturbations

80: seed the inhomogeneities which generate structure upon

81: gravitational collapse\cite{mukyotr,libros}. A great

82: diversity of inflationary models predict fairly generic features:

83: a gaussian, nearly scale invariant spectrum of (mostly) adiabatic

84: scalar and tensor primordial fluctuations, making the inflationary

85: paradigm fairly robust. The gaussian, adiabatic and nearly scale

86: invariant spectrum of primordial fluctuations provide an excellent

87: fit to the highly precise wealth of data provided by the Wilkinson

88: Microwave Anisotropy Probe (WMAP)\cite{WMAP,WMAP3}

89: Perhaps the most striking validation of inflation as a mechanism for generating

90: \emph{superhorizon} (`acausal')  fluctuations is the

91: anticorrelation peak in the temperature-polarization (TE) angular

92: power spectrum at $l \sim 150$ corresponding to superhorizon

93: scales\cite{WMAP}. The confirmation of many of the robust predictions of 

94: inflation by current high precision observations places

95: inflationary cosmology on solid grounds.

96: 

97: \medskip

98: 

99: Amongst the wide variety of inflationary scenarios, single field

100: slow roll models provide an

101: appealing, simple and fairly generic description of inflation. Its

102: simplest implementation is based on a scalar field (the inflaton)

103: whose homogeneous expectation value drives the dynamics of the

104: scale factor, plus small quantum fluctuations. The inflaton

105: potential, is fairly flat during inflation. This flatness not only

106: leads to a slowly varying Hubble parameter, hence ensuring a

107: sufficient number of e-folds, but also provides an explanation for

108: the gaussianity of the fluctuations as well as for the (almost)

109: scale invariance of their power spectrum. A flat potential

110: precludes large non-linearities in the dynamics of the

111: \emph{fluctuations} of the scalar field. 

112: 

113: \medskip

114: 

115: The current WMAP data

116: seem to validate the simpler one-field slow roll

117: scenario \cite{WMAP,WMAP3}. Furthermore, because the potential is flat

118: the scalar field is almost {\bf massless}, and modes cross the horizon

119: with an amplitude proportional to the Hubble parameter. This fact

120: combined with a slowly varying Hubble parameter yields an almost

121: scale invariant primordial power spectrum.  The slow-roll approximation has

122: been recently cast as a $1/N_{efolds}$ expansion \cite{1sN}, where

123: $ N_{efolds} \sim 50 $ is  the number of efolds before the end of inflation

124: when modes of cosmological relevance today first crossed the Hubble

125: radius.

126: 

127: The observational progress permit to start 

128: to discriminate among different inflationary models, placing

129: stringent constraints on them. The upper bound on the ratio $ r $ 

130: of tensor to scalar fluctuations obtained by WMAP \cite{WMAP,WMAP3} 

131: rules out the massless $ \phi^4 $ model and {\bf necessarily} implies 

132: the presence of a {\bf mass term} in the inflaton potential \cite{1sN,WMAP3}.

133: 

134: Besides its simplicity, the trinomial potential is a physically well motivated

135: potential for inflation in the grounds of the Ginsburg-Landau approach

136: to effective field theories (see for example ref.\cite{quir}).

137: This potential is rich enough to describe the physics of inflation and 

138: accurately reproduce the WMAP data \cite{WMAP,WMAP3}. 

139: 

140: The slow-roll expansion plus the WMAP data constraints the inflaton potential

141: to have the form \cite{1sN}

142: \be \label{Vi} 

143: V(\phi) = N_{efolds} \; M^4 \; w(\chi) \; ,

144: \ee  

145: \noindent where $ \phi $ is the inflaton field,

146: $\chi$ is a dimensionless, slowly varying field 

147: \be\label{chiflai} 

148: \chi \equiv \frac{\phi}{\sqrt{N_{efolds}} \;  M_{Pl}}  \; ,

149: \ee 

150: \noindent $ w(\chi) \sim \mathcal{O}(1) $ and 

151: $ M $ is the energy scale of inflation which is determined by the 

152: amplitude of the scalar adiabatic fluctuations \cite{WMAP} to be

153: $$ 

154: M \sim 0.00319 \; M_{Pl} = 0.77 \times 10^{16} {\rm GeV} \; .

155: $$

156: Following the spirit of the Ginsburg-Landau theory of phase transitions,

157: the simplest choice is a quartic trinomial for the inflaton potential

158: \cite{nos,1sN}:

159: \be \label{wxi}

160: w(\chi)= w_0 \pm \frac12 \; \chi^2 + \frac{h}3 \; \sqrt{\frac{y}2} \; \chi^3 +

161: \frac{y}{32} \; \chi^4 \; .

162: \ee

163: where the coefficients $ w_0, \; h $ and $ y $ are dimensionless and of order 

164: one and the signs $ \pm $ correspond to large and small field inflation, 

165: respectively (chaotic and new inflation, respectively).

166: Inserting eq.(\ref{wxi}) in eq.(\ref{Vi}) yields,

167: \be\label{VI}

168: V(\phi)= V_0 \pm \frac{m^2}{2} \; \phi^2 +  \frac{ m

169: \; g }{3} \; \phi^3 + \frac{\lambda}{4}\; \phi^4 \; .

170: \ee

171: where the mass $ m^2 $ and the couplings $ \; g $ and $ \lambda $ are given

172: by the following see-saw-like relations, 

173: \be 

174: m = \frac{M^2}{M_{Pl}} \qquad ,  \qquad g = h \; \sqrt{\frac{y}{2 \; N}} 

175: \; \left( \frac{M}{M_{Pl}}\right)^2  \qquad ,  \qquad \lambda  = 

176: \frac{y}{8 \; N} \left( \frac{M}{M_{Pl}}\right)^4 \label{acoi} 

177: \qquad ,  \qquad  V_0 = N \; M^4 \; w_0 \; . 

178: \ee 

179: where $ N \equiv N_{efolds}$. 

180: Notice that $ y \sim {\cal O}(1) \sim h $ guarantee that $ g  \sim 

181: {\cal O}(10^{-6}) $ and $ \lambda  \sim {\cal O}(10^{-12}) $ without

182: any fine tuning as stressed in ref. \cite{1sN}. That is, the smallness

183: of the couplings  directly follow from the form of the inflaton potential

184: eq.(\ref{Vi}) and the amplitude of the scalar fluctuations that fixes $ M $

185: \cite{1sN}.

186: 

187: \medskip

188: 

189: The small coupling limit $ y \to 0 $ of eqs.(\ref{wxi})-(\ref{VI}) corresponds 

190: to a quadratic potential while the strong coupling limit $ y \to \infty $

191: yields the massless quartic potential. The extreme asymmetric limit

192: $ |h| \to \infty $ yields a massive model without quadratic term. In such limit

193: the product $ |h| \; M^2 $ must be kept fixed since it is determined by the 

194: amplitude of the scalar fluctuations.

195: 

196: \medskip

197: 

198: We study here new inflation with the trinomial potential 

199: eqs.(\ref{wxi})-(\ref{VI}) and hybrid inflation [see below], the two models

200: fulfill the observational constraints. We compute in both scenarios 

201: $ n_s , \; r $ and the running $ d n_s/d \ln k $ as functions of the 

202: parameters of the models, derive explicit formulae for 

203: $ n_s , \; r $ and $ d n_s/d \ln k $ and provide relevant plots. 

204: Moreover, we plot the ratio $ r $ and the running $ d n_s/d \ln k $ 

205: as functions of the scalar index $ n_s $. Since the value of $ n_s $ is now

206: known \cite{WMAP3}-\cite{Teg}, these plots allow us to {\bf predict} the values of 

207: $ r $ and $ d n_s/d \ln k $ for the different inflationary models considered.

208: These predictions and plots are solely produced from theory and not from any 

209: fitting of the data.

210: 

211: \bigskip

212: 

213: The three years WMAP data indicate a red tilted spectrum ($ n_s < 1 $)

214: with a small ratio $ r < 0.28 $  of tensor to scalar fluctuations \cite{WMAP3}.

215: The present data do not permit to find the precise values neither

216: of the ratio $ r $ nor of the running index $ d n_s/d \ln k $, 

217: only upper bounds are obtained \cite{WMAP,WMAP3}. We therefore think

218: that the value of $ n_s $ [eq.(\ref{nswmap})] obtained through a fit of the 

219: data assuming $ r =  d n_s/d \ln k = 0 $ is more precise than the values of 

220: $ n_s $ obtained through fits allowing both $ r $ and $ d n_s/d \ln k $ to vary.

221: Notice that $ n_s = 0.95 $ was independently found from the 2dF data

222: under similar assumptions \cite{2dF}. More precisely, from the three years WMAP 

223: data \cite{WMAP3} as well as ref. \cite{2dF} we take

224: \be \label{nswmapi}

225:  n_s = 0.95 \pm 0.02 \; .

226: \ee

227: We find that for $ n_s = 0.95 $ and any value of the asymmetry $ h $ 

228: [see figs. \ref{nsr} and \ref{nsrun}], new inflation with the trinomial potential 

229: eqs.(\ref{wxi})-(\ref{VI}) predicts

230: $$

231: {\rm trinomial~potential~new~inflation~for~} n_s = 0.95 : \quad 

232: 0.03 < r < 0.04 \quad {\rm and}\quad -0.00070 < d n_s/d \ln k < -0.00055 \; .

233: $$

234: We find for the lower value $ n_s = 0.93 $ of the three years WMAP data band:

235: $$

236: {\rm trinomial~potential~new~inflation~for~} n_s = 0.93 : \quad 

237: 0.003 < r < 0.015 \quad {\rm and}\quad -0.0011 < d n_s/d \ln k < -0.00033 \; .

238: $$

239: Moreover, in new inflation with the trinomial potential, we find that $ n_s $ 

240: is {\bf bounded} from above by  

241: $$ 

242: {\rm new~inflation:} \qquad n_s < n_{s~maximum} = 0.961528 \ldots \; .

243: $$ 

244: For $ n_s = 0.961528 \ldots $ we have in this model $ r =  0.114769\ldots $

245: (see figs. \ref{nsr} and \ref{nsh}). 

246: Interestingly enough, there exists {\bf two} values (two branches) of

247: $ r $ for one value of  $ n_s $ in the interval $ 0.96 < n_s < 0.961528 

248: \ldots $ [see fig. \ref{nsr}]. The value $ r_{max} = 0.114769\ldots $ is the 

249: maximun $ r $ in the first branch. The values $ 0.16 \geq r \geq 

250: 0.114769\ldots $ correspond to a second branch of $ r $ as a function of 

251: $ n_s $ in the interval $ 0.96 < n_s < 0.961528 \ldots $. 

252: In the first branch we have 

253: $$

254:  r_{max} = 0.114769\ldots \; .

255: $$

256: The absolute maximun value $ r_{abs~max} = 0.16 $ belongs to the second branch

257: and corresponds to the quadratic monomial potential obtained from eq.(\ref{wxi}) 

258: at $ y = 0 $.

259: 

260: These predicted values of the ratio $ r $ fullfil the three years WMAP bound

261: including SDSS galaxy survey \cite{WMAP3}

262: \be \label{rwmapi}

263: r < 0.28 \; (95\% ~CL) \; .

264: \ee

265: Moreover, one can see from fig. 14 in ref. \cite{WMAP3} 

266: that $ r < 0.1 ~(68\% ~CL) $ from WMAP$+$SDSS. 

267: 

268: \medskip

269: 

270: Chaotic inflation  with the trinomial potential eq.(\ref{wxi})-(\ref{VI})

271: yields larger values of $ r $ than new inflation

272: for a given value of $ n_s $ \cite{nos}. More precisely, for $ n_s = 0.95 $

273: we find $ r = 0.27 $ for the binomial potential \cite{nos}

274: (the trinomial potential introduces very small changes).

275: 

276: Therefore, although the WMAP value for $ n_s $ [eq.(\ref{nswmapi})]

277: is compatible both with chaotic and new inflation, the WMAP bounds on 

278: $ r $ {\bf clearly disfavour} chaotic inflation. New inflation easily fulfils 

279: the three years WMAP bounds on $ r $ and prepares the way for the expected

280:  data on the ratio of tensor/scalar fluctuations $ r \lesssim 0.1 $.

281: 

282: \bigskip

283: 

284: In the inflationary models of hybrid type, the inflaton is coupled to another 

285: scalar field $ \sigma_0 $ with mass term $ -\mu_0^2 < 0 $

286: through a potential of the type \cite{lin}

287: \bea\label{Vhib1i}

288: &&V_{hyb}(\phi,\sigma_0) = \frac{m^2}{2} \; \phi^2 + \frac{g_0^2}{2} \; 

289: \phi^2 \; \sigma_0^2 + \frac{\mu_0^4}{16 \, \Lambda_0} 

290: \left(\sigma_0^2 -  \frac{4 \, \Lambda_0}{\mu_0^2} \right)^2= \cr \cr

291: &&  =\frac{m^2}{2} \; \phi^2 + \Lambda_0+\frac12 \; (g_0^2 \; 

292: \phi^2-\mu_0^2) \;  \sigma_0^2+\frac{\mu_0^4}{16 \; \Lambda_0} \; 

293: \sigma_0^4\; ,

294: \eea

295: where $ m^2 > 0 , \; \Lambda_0 > 0 $ plays the role of a cosmological constant 

296: and $ g_0^2 $  couples $ \sigma_0 $ with $ \phi $.

297: 

298: \medskip

299: 

300: The initial conditions are chosen such that $\sigma_0$ and $ \dot\sigma_0$ 

301: are very small (but not identically zero) and therefore 

302: inflation is driven by the cosmological constant $ \Lambda_0 $ 

303: plus the initial value of the inflaton $  \phi(0) $.

304: The inflaton field $ \phi(t) $ decreases with time while the scale factor 

305: $ a(t) $ grows exponentially with time. The  field $ \sigma_0 $ has an

306: effective classical mass square

307: \be\label{masefei}

308: m_{\sigma}^2 = g_0^2 \; \phi^2 - \mu_0^2 \; .

309: \ee

310: Since  the inflaton field $ \phi $ decreases with time, $ m_{\sigma}^2 $ 

311: becomes 

312: negative at some moment during inflation. At such moment, spinodal (tachyonic) 

313: unstabilities appear and the field $ \sigma_0 $ starts to grow exponentially. 

314: Inflation stops when both fields $ \phi $ and $ \sigma_0 $ are comparable with 

315: $ \dot \phi $ and  $ \dot \sigma_0 $ and close to their vaccum values.

316: 

317: We find that the time when the effective mass of the field $ \sigma_0 $ 

318: eq.(\ref{masefei}) becomes negative depends on the

319: values of $ \mu_0^2 $ and $ g_0^2 \; \phi^2(0) $. For low values of 

320: $ \mu_0^2 $ the field  $ \sigma_0 $ starts to grow

321: close to the end of inflation. On the contrary, for higher values of 

322: $ \mu_0^2 $ the field  $ \sigma_0 $ starts to grow

323: well before the end of inflation.  This is explained by the fact

324: that the scale of time variation of $ \sigma_0 $ goes as

325: $ \mu_0^{-1} $.  $ \sigma_0 $ evolves slowly for small $ \mu_0 $ and

326: fastly for large $ \mu_0 $ [see figs. \ref{a}-\ref{pe}].

327: 

328: {\bf Only} at $ \Lambda = 0 $ hybrid inflation becomes chaotic inflation 

329: with the monomial potential $ \frac{m^2}{2} \; \phi^2 $. For any value of 

330: $ \Lambda > 0 $ even very small, the features of hybrid inflation remain.

331: 

332: We compute $ n_s , \; r $ and  $ d n_s/d \ln k $ for hybrid inflation

333: as functions of the parameters in the potential eq.(\ref{Vhib1i}) and

334: the initial value of the inflaton field 

335: [see figs. \ref{nsr005A}-\ref{nsrunA}]. 

336: 

337: The results of our extended numerical investigation of hybrid inflation can

338: be better expressed in terms of the dimensionless variables

339: $$

340: \Lambda \equiv \frac{2 \, \Lambda_0}{M^4 \; N_{efolds}} \quad , \quad

341: {\hat \chi} \equiv \frac{\phi}{\sqrt{N_{efolds} \; \Lambda} \;  M_{Pl}} 

342: \quad {\rm and} \quad {\hat \mu}^2 \equiv \frac{\mu_0^2 \;  

343: M_{Pl}^2 \; N_{efolds}}{2 \; \Lambda_0} \; .

344: $$

345: We depict in figs. \ref{nsr005A}-\ref{nsrunA} the observables 

346: $ n_s , \; r $ and the running index 

347: $ d n_s /d \ln k $ as  functions of $ \Lambda $ and $ n_s $.

348: We present a complete picture for hybrid inflation covering {\bf two} 

349: different, blue tilted and red tilted, regimes. We find that for all the 

350: observables, the shape of the curves depends crucially on the mass parameter 

351: $ {\hat \mu}^2 $ of the $\sigma$ field and the (rescaled) initial amplitude 

352: $ {\hat \chi}(0) $ of the inflaton field.

353: 

354: We find a blue tilted spectrum ($ n_s > 1 $) for 

355: $ {\hat \mu}^2 <  {\hat \mu}^2_{crit}

356: \simeq 0.13 $ while for   $ {\hat \mu}^2 >  {\hat \mu}^2_{crit} $ we can have

357: either $ n_s > 1 $ or  $ n_s < 1 $ depending on the initial conditions:

358: for $ {\hat \chi}(0) > {\hat \chi}(0)_{crit} $ we have $ n_s > 1 $,

359: and for $ {\hat \chi}(0) < {\hat \chi}(0)_{crit} $ we have $ n_s < 1 $. 

360: The value of $ {\hat \chi}(0)_{crit} $ grows with $ {\hat \mu}^2 $:

361: for $ {\hat \mu}^2 = 0.5 $,  we find  $ {\hat \chi}(0)_{crit} = 2.7 $

362: and for $ {\hat \mu}^2 = 1.7 $, we find   $ {\hat \chi}(0)_{crit} = 5.8 $.

363: 

364: We see that $ n_s > 1 $ happens when the cosmological constant 

365: $ \Lambda_0 $ is large enough compared with $  \mu_0^2 \;  M_{Pl}^2 \; 

366: N_{efolds}$. More precisely,  $ {\hat \mu}^2 <  {\hat \mu}^2_{crit} $ for

367: $ \Lambda_0 > 192 \; \mu_0^2 \;  M_{Pl}^2 $ using $ N_{efolds} = 50 $.

368: That is, for $ \Lambda_0 < 192 \; \mu_0^2 \;  M_{Pl}^2$ we have either

369: red or blue tilted spectrum as explained above.

370: 

371: \medskip

372: 

373: For large $\Lambda, \;   n_s -1, \; r $ and $ d n_s/d \ln k $ always tend 

374: asymptotically to zero whatever be $ {\hat \mu}^2 $ and $ {\hat \chi}(0) $.

375: 

376: \medskip

377: 

378: We see from our calculations that all blue tilted values of  

379: $ (n_s, r) $ in the domain $ 1 < n_s < 1.15 , \; 0 < r < 0.2 $ 

380: can be realized by the hybrid inflation model eq.(\ref{Vhib1i}).

381: However, at the light of the three years WMAP data ref. \cite{WMAP3}

382: the blue tilted regime in hybrid inflation $ \mu^2 <  \mu^2_{crit} $ 

383: is ruled out.

384: 

385: The situation is totally different in the red tilted regime $  {\hat  \mu}^2 >

386: {\hat \mu}^2_{crit} \simeq 0.13 $ in hybrid inflation. 

387: The possible values of $ (n_s, r) $ for such regime of hybrid inflation are 

388: in the upper-right quadrant as shown in fig. \ref{bordenuevo}.

389: 

390: Hybrid inflation in the red tilted regime $ \mu^2 >  \mu^2_{crit} $

391: and $ {\hat \chi}(0) <  {\hat \chi(0)}_{crit} $ 

392: fulfills the three years WMAP value for $ n_s $ [see eq.(\ref{nswmap})] as

393: well as the bound on the ratio $ r $ [eq.(\ref{rwmap})].

394: We can read from fig. \ref{nsrY} and \ref{nsrunY} that 

395: $$ 

396: \mu^2 >  \mu^2_{crit} \; \; {\rm hybrid ~ inflation:} \quad 

397: 0.2 > r > 0.14 \quad {\rm and } \quad -0.001 < d n_s /d \ln k < 0 \quad

398: {\rm for} \quad  0.952 <  n_s < 0.97  \; .

399: $$

400: Notice that hybrid inflation in the red tilted regime yields a too large

401: ratio $ r > 0.2 $ for $ n_s < 0.95 $.

402: 

403: \medskip

404: 

405: At the central three years WMAP value  $ n_s = 0.95 $ both new and hybrid 

406: inflation

407: are allowed. However, for $ n_s < 0.95 $ hybrid inflation is in trouble 

408: ($ r > 0.2 $)

409: while for  $ n_s > 0.962 $ new inflation is excluded.

410: 

411: \medskip

412: 

413: The potential which best agree with the present red tilted spectrum 

414: and which best prepares the way to the expected data 

415: (a small $ r \lesssim 0.1 $) is the trinomial  potential eqs.(\ref{wxi})-(\ref{VI}) 

416: with negative mass term, that is 

417: small field (new) inflation. Hybrid inflation with a trinomial potential

418: can also reproduce the present data in the red tilted regime 

419: $ \mu^2 >  \mu^2_{crit} $ and $ {\hat \chi}(0) <  {\hat \chi(0)}_{crit} $.

420: 

421: \medskip

422: 

423: All calculations presented in this paper stem from the inflaton potential

424: in the slow roll approximation (dominant order in $ 1/N \simeq 1/50 $).

425: They do not use observational data as input. The analytical formulas 

426:   and plots provided in the paper allow to read directly the predicted

427:   values of $ r $ and $ dn_s/d \ln k $ as functions of $ n_s $.  In order to make illustrative

428:   predictions, we take the value $ n_s = 0.95 \pm 0.02$, as a judicious

429:   choice. The reader can see directly from the plots presented here

430:  our predictions for $ r $ and $ d n_s /d \ln k $ for

431:  future observational values of $ n_s $.

432: 

433: \section{The Inflaton Potential and the $ 1/N_{efolds} $ Slow Roll Expansion}

434: 

435: The description of cosmological inflation is based on an isotropic

436: and homogeneous geometry, which assuming flat spatial sections is

437: determined by the invariant distance

438: \be

439: ds^2= dt^2-a^2(t) \; d\vec{x}^2 \label{FRW} \; .

440: \ee

441: The scale factor obeys the Friedman equation

442: \be \label{ef}

443: \left[ \frac{1}{a(t)} \; \frac{da}{d t} \right]^2 =

444: \frac{\rho( t)}{3 \; M^2_{Pl}}   \;  ,

445: \ee

446: where $M_{Pl}= 1/\sqrt{8\pi G} = 2.4\times 10^{18}\,\textrm{GeV}$.

447: 

448: In single field inflation the energy density is dominated by a

449: homogeneous scalar \emph{condensate}, the inflaton, whose dynamics

450: is described by an  \emph{effective} Lagrangian

451: \be\label{lagra}

452: {\cal L} = a^3({ t}) \left[ \frac{{\dot

453: \phi}^2}{2} - \frac{({\nabla \phi})^2}{2 \;  a^2({ t})} - V(\phi) \right]

454: \; .

455: \ee

456: \noindent The inflaton potential $ V(\phi) $ is a slowly varying

457: function of $ \phi $ in order to permit a slow-roll solution for

458: the inflaton field $ \phi(t) $.

459: 

460:  We showed in ref. \cite{1sN} that combining the WMAP data with the

461:  slow roll expansion yields an inflaton potential of the form 

462: \be \label{V} 

463: V(\phi) = N \; M^4 \; w(\chi)  \; ,

464: \ee  

465: \noindent where $\chi$ is a dimensionless, slowly varying field 

466: \be\label{chifla} 

467: \chi = \frac{\phi}{\sqrt{N} \;  M_{Pl}}  \; ,

468: \ee 

469: $w(\chi) \sim \mathcal{O}(1)~,~N\sim 50$ is the number of efolds

470: since the cosmologically relevant modes exited the horizon till the 

471: end of inflation and $ M $ is the energy scale of inflation

472: 

473: \medskip

474: 

475: The dynamics of the rescaled field $ \chi $ exhibits the slow

476: time evolution in terms of the \emph{stretched}

477: dimensionless time variable, 

478: \be \label{tau} 

479: \tau =  \frac{t \; M^2}{M_{Pl} \; \sqrt{N}}  \; .

480: \ee 

481: The rescaled variables $ \chi $ and $ \tau $ change slowly with time. 

482: A large change in the field amplitude $\phi$ results in a small change 

483: in the $ \chi $ amplitude, a change in $\phi \sim  M_{Pl}$ results in a 

484: $\chi$ change $\sim 1/\sqrt{N}$. The form of the potential, eq.(\ref{V}) 

485: and the rescaled dimensionless inflaton field   eq.(\ref{chifla}) and time 

486: variable $ \tau $ make manifest

487: the slow-roll expansion as a consistent systematic expansion in powers of 

488: $1/N$ \cite{1sN}.  

489: 

490: \medskip

491: 

492: The inflaton mass around the minimum is given by a see-saw formula

493: $$

494: m = \frac{M^2}{M_{Pl}}  \sim 2.45 \times 10^{13} \, \textrm{GeV} \; .

495: $$

496: The Hubble parameter when the cosmologically relevant modes exit the horizon

497: is given by

498: $$

499: H  = \sqrt{N} \; m \, {\cal H} \sim 1.0 \times 10^{14}\,\textrm{GeV}

500: = 4.1 \; m \; ,

501: $$

502: where we used that $  {\cal H} \sim 1 $. As a result, $ m\ll M $ and 

503: $ H \ll M_{Pl} $. A Ginsburg-Landau realization of the inflationary potential 

504: that fits the amplitude of the CMB anisotropy 

505: remarkably well, reveals that the

506: Hubble parameter, the inflaton mass and non-linear couplings are

507: see-saw-like, namely  powers of the ratio $M/M_{Pl}$ multiplied by

508: further powers of $ 1/N $. Therefore, the smallness of the couplings is not 

509: a result of fine tuning but a {\bf natural} consequence of the form of

510: the potential and the validity of the effective field theory

511: description and slow roll. The quantum expansion in loops is

512: therefore a double expansion on $ \left(H/M_{Pl}\right)^2 $ and $

513: 1/N $. Notice that graviton corrections are  also at least of

514: order $ \left(H/M_{Pl}\right)^2 $ because the amplitude of tensor

515: modes is of order $H/M_{Pl}$ . We showed that the form of the potential which 

516: fits the WMAP data and is consistent with slow roll 

517: eqs.(\ref{V})-(\ref{chifla}) implies the small values for the inflaton 

518: self-couplings \cite{1sN}. 

519: 

520: \medskip

521: 

522: The equations of motion in terms of the dimensionless  rescaled field 

523: $ \chi $ and the slow time variable $ \tau $ take the form,

524: \bea \label{evol} 

525: &&  {\cal H}^2(\tau) = \frac13\left[\frac1{2\;N} 

526: \left(\frac{d\chi}{d \tau}\right)^2 + w(\chi) \right] \quad , \cr \cr

527: && \frac1{N} \;  \frac{d^2

528: \chi}{d \tau^2} + 3 \;  {\cal H} \; \frac{d\chi}{d \tau} + w'(\chi) = 0 \quad .

529: \eea 

530: The slow-roll approximation follows by neglecting the

531: $\frac1{N}$ terms in eqs.(\ref{evol}). Both

532: $w(\chi)$ and $h(\tau)$ are of order $N^0$ for large $N$. Both

533: equations make manifest the slow roll expansion as an expansion in

534: $1/N$.

535: 

536: The number of e-folds $ N[\chi] $ since the field $ \chi $ exits the horizon 

537: till the end of inflation (where $ \chi $ takes the value $ \chi_{end} $) 

538: can be computed in close form from eqs. (\ref{evol}) in the slow-roll 

539: approximation (neglecting $ 1/N $ corrections)

540: \be \label{Nchi}

541: \frac{N[\chi]}{N} = -\int_{\chi}^{\chi_{end}}  \;

542: \frac{w(\chi)}{w'(\chi)} \; d\chi \;  \leqslant 1 \; .

543: \ee

544: 

545: \medskip

546: 

547: The amplitude of adiabatic scalar perturbations is expressed as

548: \cite{libros,WMAP,1sN,barrow,hu}

549: \be \label{ampliI}

550: |{\Delta}_{k\;ad}^{(S)}|^2  = \frac{N^2}{12 \, \pi^2} \;

551: \left(\frac{M}{M_{Pl}}\right)^4 \; \frac{w^3(\chi)}{w'^2(\chi)} \; .

552: \ee

553: The spectral index $ n_s $,  its running and the ratio of tensor to scalar 

554: fluctuations are expressed as

555: \bea \label{indi}

556: &&n_s - 1 = -\frac3{N} \; \left[\frac{w'(\chi)}{w(\chi)} \right]^2

557: +  \frac2{N}  \; \frac{w''(\chi)}{w(\chi)} \quad , \cr \cr 

558: &&\frac{d n_s}{d \ln k}= - \frac2{N^2} \; \frac{w'(\chi) \;

559: w'''(\chi)}{w^2(\chi)} - \frac6{N^2} \; \frac{[w'(\chi)]^4}{w^4(\chi)}

560: + \frac8{N^2} \; \frac{[w'(\chi)]^2 \; w''(\chi)}{w^3(\chi)}\quad , \cr \cr 

561: &&r = \frac8{N} \; \left[\frac{w'(\chi)}{w(\chi)} \right]^2 \quad .

562: \eea

563: In eqs.(\ref{Nchi})-(\ref{indi}) the field $ \chi $ is computed at horizon 

564: exiting. We choose $ N[\chi] = N = 50 $.

565: 

566: Since, $ w(\chi) $ and  $ w'(\chi) $ are of order one, we

567: find from eq.(\ref{ampliI})

568: \be\label{Mwmap}

569: \left(\frac{M}{M_{Pl}}\right)^2 \sim \frac{2

570: \, \sqrt{3} \, \pi}{N} \; |{\Delta}_{k\;ad}^{(S)}| \simeq  1.02

571: \times 10^{-5} \; .

572: \ee

573: where we used $ N \simeq 50 $ and the WMAP

574: value for $ |{\Delta}_{k\;ad}^{(S)}| = (4.67 \pm 0.27)\times

575: 10^{-5} $ \cite{WMAP}. This fixes the scale of inflation to be

576: $$

577: M \simeq 3.19 \times 10^{-3} \; M_{PL} \simeq 0.77

578: \times 10^{16}\,\textrm{GeV} \; .

579: $$

580: This value pinpoints the scale of

581: the potential during inflation to be at the GUT scale suggesting a

582: deep connection between inflation and the physics at the GUT

583: scale in cosmological space-time.

584: 

585: \medskip

586: 

587: We see that $|n_s -1|$ as well as  the ratio $ r $ turn out to be of order 

588: $ 1/N_{efolds} $. This nearly scale invariance is a natural property of

589: inflation which is described by a quasi-de Sitter space-time geometry.

590: This can be understood intuitively as follows: 

591: the geometry of the universe is scale invariant during de Sitter stage 

592: since the metric takes in conformal time the form 

593: $$

594: ds^2 = \frac1{(H \; \eta)^2}\left[ (d \eta)^2 - (d \vec x)^2 \right] \; .

595: $$

596: Therefore, the primordial power generated is scale invariant except

597: for the fact that inflation is not eternal and lasts for $N_{efolds}$.

598: Hence, the  primordial spectrum is scale invariant up to $ 1/N_{efolds} $ 

599: corrections. The values $ n_s = 1, \; r = 0 $ and $ d n_s/d \ln k

600: = 0 $ correspond to a critical point as discussed in ref.\cite{1sN}.

601: This a gaussian fixed point around which the inflation model hovers

602: in the renormalization group sense with an almost scale invariant spectrum 

603: of scalar fluctuations during the slow roll stage.

604: 

605: \medskip

606: 

607: The WMAP results favoured single inflaton models and among them new and hybrid

608: inflation emerge to be preferable than chaotic inflation \cite{nos}.

609: 

610: We analyze in the subsequent sections new inflation and hybrid

611: inflation in its simple physical realizations within the 

612: Ginzburg-Landau approach (the trinomial potential)\cite{nos}.

613: 

614: \section{Spectral index $n_s$, ratio $r$ and running index 

615: $ \frac{d n_s}{d \ln k} $ for New Inflation with the Trinomial Potential}

616: 

617: We consider here the trinomial potential investigated in ref.\cite{nos}

618: \be\label{VN}

619: V(\phi)= V_0 - \frac{m^2}{2} \; \phi^2 + \frac{ m \; g }{3} \; \phi^3 + 

620: \frac{\lambda}{4}\; \phi^4 \; . 

621: \ee

622: where $ m^2 > 0 $ and $ g $ and $ \lambda $ are dimensionless couplings.

623: 

624: The corresponding dimensionless potential $ w(\chi) $ takes the form

625: \be\label{trino}

626: w(\chi) = -\frac12 \; \chi^2 + \frac{h}3 \; \sqrt{\frac{y}2} \; \chi^3 +

627: \frac{y}{32} \; \chi^4 + \frac2{y} \; F(h)  \; ,

628: \ee

629: where the quartic coupling $ y $ is dimensionless as well as 

630: the asymmetry parameter $ h $. The couplings in eq.(\ref{VN}) and 

631: eq.(\ref{trino}) are related by,

632: \be 

633: g = h \; \sqrt{\frac{y}{2 \; N}} \; \left( \frac{M}{M_{Pl}}\right)^2  

634: \qquad ,  \qquad 

635: \lambda  = \frac{y}{8 \; N} \; \left( \frac{M}{M_{Pl}}\right)^4 

636: \label{aco} \; . 

637: \ee 

638: and the constant $ F(h) $ is related to $ V_0 $ by

639: $$

640: \frac2{y} \; F(h) = \frac{V_0}{N \; M^4 } \; .

641: $$

642: The constant $ F(h) $ ensures that 

643: $ w(\chi_+) =  w'(\chi_+) = 0 $ 

644: at the absolute minimum $ \chi = \chi_+ =\sqrt{\frac8{y}} (\Delta + |h|) $ 

645: of the potential $ w(\chi) $. Thus, inflation does not run eternally.

646: $ F(h) $ is given by

647: $$

648: F(h) \equiv \frac83 \, h^4 + 4 \, h^2 + 1 + \frac83 \, |h| \, \Delta^3 

649: \quad , \quad \Delta \equiv \sqrt{h^2 + 1} \; .

650: $$

651: The parameter $ h $ reflects how asymmetric is the potential.

652: Notice that  $ w(\chi) $ is invariant under the changes

653: $ \chi \to - \chi , \;  h \to - h $. Hence, we can restrict

654: ourselves to a given sign for $ h $. Without loss of 

655: generality, we choose $ h < 0 $ and shall

656: work with positive fields $ \chi $.

657: 

658: \begin{figure}[p]

659: \begin{turn}{-90}

660: \centering

661: \psfrag{ufa}{$n_s  ~ vs. ~ \log y$}

662: \includegraphics[width=10cm,height=14cm]{ns}

663: \end{turn}

664: \caption{New Inflation. $ n_s $ as a function of $ \log y $ for the asymmetry 

665: of the potential $ |h| = 0, 1, 3, 7 $ and $ 16 $, $ y $ being the 

666: dimensionless quartic coupling. The $ y \to 0 $ limiting value $ n_s = 

667: 1 - \frac{2}{N} = 0.96 $ is $h$-independent and corresponds to the monomial 

668: potential $ \frac12 \; m^2 \; \phi^2 $.}

669: \label{ns}

670: \end{figure}

671: \begin{figure}[p]

672: \begin{turn}{-90}

673: \centering

674: \includegraphics[width=10cm,height=14cm]{run}

675: \end{turn}

676: \caption{New Inflation. The running $ d n_s/d \ln k $ as a 

677: function of $ \log y $ for the 

678: asymmetry of the potential $ |h| = 0, 1, 3, 7 $ and $ 16 $,

679: $ y $ being the dimensionless quartic coupling.The $ y \to 0 $ limiting value 

680: $ -\frac2{N^2}= -0.0008 $ is $h$-independent and corresponds to the monomial 

681: potential $ \frac12 \; m^2 \; \phi^2 $.}

682: \label{frun}

683: \end{figure}

684: 

685: Notice that $ y \sim {\cal O}(1) \sim h $ guarantee that $ g  \sim 

686: {\cal O}(10^{-6}) $ and $ \lambda  \sim {\cal O}(10^{-12}) $ without

687: any fine tuning as stressed in ref. \cite{1sN}. 

688: 

689: New inflation is obtained by choosing the initial field $ \chi $ in 

690: the interval $ (0,\chi_+) $. The inflaton  $ \chi $ slowly rolls down the slope

691: of the potential from its initial value till the absolute minimum of the 

692: potential $ \chi_+ $.

693: 

694: Computing the number of efolds from eq.(\ref{Nchi}), we find the field 

695: $ \chi $ at horizon crossing related to the parameters $ y $ and $ h $.

696: It is convenient to define the field variable $z$:

697: $$

698: z \equiv \frac{y}8 \; \chi^2 \; .

699: $$

700: We obtain by inserting eq.(\ref{trino}) for $ w(\chi) $ into  eq.(\ref{Nchi})

701: and setting $ N[\chi] = N $,

702: \bea\label{ntrino}

703: && y  = z - 2 \; h^2 -1 - 2  \; |h|  \; \Delta + \frac43 \; 

704: |h|  \; \left( |h| + \Delta - \sqrt{z} \right) + \cr \cr

705: &&+\frac{16}{3} \; |h| \;  (\Delta + |h| ) \; \Delta^2  \; 

706: \log\left[\frac12 \left(1 +  \frac{\sqrt{z} -  |h|}{\Delta}\right)\right] - 

707: 2 \, F(h) \, \log\left[\sqrt{z} \; (\Delta - |h|)\right] \; .

708: \eea

709: $ z $ turns to be a monotonically decreasing function of $ y $:

710: $ z $ decreases from $ z = z _+ = (\Delta + |h|)^2$ till $ z = 0 $ when

711: $ y $ increases from $ y = 0 $ till $ y = \infty $.

712: When  $ \sqrt{z} \to 

713: \sqrt{z_+} , \; y $ vanishes quadratically,

714: $$

715: y \buildrel{z \to z_+}\over= 2 \; 

716: \left(\sqrt{z} - \sqrt{z_+}\right)^2 + {\cal O} 

717: \left(\left[\sqrt{z} - \sqrt{z_+}\right]^3\right) \; .

718: $$

719: We obtain in analogous way from eqs.(\ref{ampliI}) and (\ref{indi}) the 

720: spectral index, its running, the ratio $r$ and the amplitude of adiabatic 

721: perturbations,

722: \bea

723: && n_s=1 - 6 \,  \frac{y}{N} \, \frac{z \; (z + 2  \, h  \, 

724: \sqrt{z} -1)^2}{\left[F(h)  -2 \, z+ \frac83  \, h  \,  z^{3/2} +

725: z^2\right]^2} +  \frac{y}{N} \, \frac{ 3 \, z+ 4 \, h  \, \sqrt{z} -1}{

726: F(h)- 2 \, z+ \frac83  \, h  \,  z^{3/2} + z^2} \quad , \\ \cr\cr

727: \label{nstrino}

728: &&\frac{d n_s}{d \ln k}= - \frac2{N^2} \; \sqrt{z} \; y^2 \; 

729: \frac{(z + 2  \, h  \, \sqrt{z} -1)(h + \frac32 \; \sqrt{z})}{

730: \left[F(h)  -2 \, z+ \frac83  \, h  \,  z^{3/2} +z^2 \right]^2} \cr\cr

731: &&- \frac{24}{N^2} \;  y^2 \; z^2 \; \frac{(z + 2  \, h  \, \sqrt{z} -1)^4}{

732: \left[F(h)  -2 \, z+ \frac83  \, h  \,  z^{3/2} +z^2 \right]^4} \cr\cr

733: && + \frac8{N^2} \;  y^2 \; z \; \frac{ (3 \, z+ 4 \, h  \, \sqrt{z} -1)

734: (z + 2  \, h  \, \sqrt{z} -1)^2}{\left[F(h)  -2 \, z+ \frac83  \, h  \,  

735: z^{3/2} +z^2 \right]^3} \quad , \label{run} \\ \cr \cr

736: && r = 16 \,   \frac{y}{N} \, \frac{z \; (z + 2  \, h  \, \sqrt{z} -1

737: )^2}{\left[F(h)  -2 \, z+ \frac83  \, h  \,  z^{3/2} +z^2

738: \right]^2}  \quad  , \label{rtrino}\\ \cr\cr

739: &&|{\Delta}_{k\;ad}^{(S)}|^2  = \frac{N^2}{12 \, \pi^2} \; 

740: \left(\frac{M}{M_{Pl}}\right)^4 \;

741: \frac{\left[F(h)- 2 \, z+ \frac83  \, h  \,  z^{3/2} + 

742: z^2\right]^3}{y^2 \; z \; (z + 2  \, h  \, \sqrt{z} -1)^2} \; .\label{dtrino}

743: \eea

744: 

745: \section{New Inflation with the Trinomial potential confronted to the 

746: three years WMAP data}

747: 

748: We plot $ n_s $, its running and $ r $ in figs. \ref{ns}, \ref{frun} 

749: and \ref{r} as functions of $ \log y $ for various values of the asymmetry 

750: of the potential $ h , \; y $ being the dimensionless quartic coupling.

751: Figs. \ref{nsr} and \ref{nsrun} depict $ r $ and the running 

752: $ d n_s/d \ln k $ as functions of $ n_s $ for various values of the 

753: asymmetry $ h $. 

754: 

755: \begin{figure}[p]

756: \begin{turn}{-90}

757: \centering

758: \includegraphics[width=9cm,height=14cm]{r}

759: \end{turn}

760: \caption{New Inflation. $ r $ as a function of $ \log y $ for the asymmetry 

761: of the potential $ |h| = 0, 1, 3, 7 $ and $ 16, \; y $ being the dimensionless 

762: quartic coupling. The absolute maximun value $ r = \frac8{N} = 0.16 $ is 

763: reached for

764: $ y = 0 $ and all $h$ and corresponds to the monomial potential 

765: eq.(\ref{maslim}).}

766: \label{r}

767: \end{figure}

768: 

769: \begin{figure}[p]

770: \begin{turn}{-90}

771: \centering

772: \psfrag{"nsr1.dat"}{$h = 0$} 

773: \psfrag{"nsr2.dat"}{$|h| = 0.15$} 

774: \psfrag{"nsr3.dat"}{$|h| = 0.4$} 

775: \psfrag{"nsr4.dat"}{$|h| = 0.7$} 

776: \psfrag{"nsr5.dat"}{$|h| = 20$} 

777: \psfrag{r  vs.  ns}{$r$ vs. $n_s$}

778: \includegraphics[width=10cm,height=14cm]{nsr3}

779: \end{turn}

780: \caption{New Inflation. $ r $ as a function of $ n_s $ for the asymmetry 

781: of the potential $ |h| = 0, 0.15, 0.4, 0.7 $ and $ 20 $. For a given  $ n_s $,

782: $ r $ monotonically and slowly decreases with increasing $ |h| $.

783: $ r = r(n_s) $ is not too sensitive to $ h $. The maximun value of $ n_s $

784: is $ n_s^{maximum} = 0.961528\ldots $ and the corresponding $ r $ is

785: $ r_{max} =  0.114769\ldots $. The maximun value of  $ r $ is 

786: $ r_{abs~max} = 0.16 $ and corresponds to the quadratic potential setting 

787: $ y = 0 $ in eq.(\ref{trino}). For $ n_s = 0.95 $ 

788: (the three years WMAP value), we find $ 0.03 < r < 0.04 $.}

789: \label{nsr}

790: \end{figure}

791: 

792: \begin{figure}[p]

793: \begin{turn}{-90}

794: \centering

795: \psfrag{"nsrun1.dat"}{$h = 0$} 

796: \psfrag{"nsrun2.dat"}{$|h| = 0.15$} 

797: \psfrag{"nsrun3.dat"}{$|h| = 0.4$} 

798: \psfrag{"nsrun4.dat"}{$|h| = 0.7$} 

799: \psfrag{"nsrun5.dat"}{$|h| = 20$} 

800: \includegraphics[width=10cm,height=14cm]{nsrun93}

801: \end{turn}

802: \caption{New Inflation. The running $ d n_s/d \ln k $ as a function of $ n_s $ 

803: for the asymmetry of the potential $ |h| = 0, 0.15, 0.4, 0.7 $ and $ 20 $.

804: The running turns out to be always {\bf negative} in new inflation.

805: For $ n_s < 0.96 $, the running $ d n_s/d \ln k $ decreases with increasing

806: $ |h| $. The opposite happens for  $ n_s > 0.96 $. In the last case the 

807: dependence on $ h $ is weak. We find $ d n_s/d \ln k = -0.00077\ldots $ at the 

808: branch point $ n_s = 0.961\ldots $ for all values of $|h|$. 

809: The point $ n_s = 1 - \frac{2}{N} = 0.96, \;  \frac{d n_s}{d \ln k} 

810: = -\frac2{N^2}= -0.0008 $ is reached for 

811: all values of $ h $ and corresponds to the monomial potential 

812: eq.(\ref{maslim}). For $ n_s = 0.95 $ (the three years WMAP value), 

813: we find $  -0.00070 < d n_s/d \ln k < -0.00055 $.}

814: \label{nsrun}

815: \end{figure}

816: 

817: \begin{figure}[p]

818: \begin{turn}{-90}

819: \centering

820: \psfrag{maxima of n_s vs. |h|}{maxima of $n_s$ vs. $|h|$}

821: \includegraphics[width=9cm,height=14cm]{nsh}

822: \end{turn}

823: \caption{New Inflation. Maxima of $ n_s $ plotted vs. the asymmetry 

824: of the potential $ |h| $. The limiting value for large $ |h| $ is 

825: $ n_s^{maximum} = 0.961528\ldots $}

826: \label{nsh}

827: \end{figure}

828: 

829: We see that generically $ n_s < 1 $ and  $ d n_s/d \ln k < 0 $ for new 

830: inflation for all values of the couplings. 

831: 

832: In new inflation we have the absolute upper bound 

833: \be \label{cotar}

834: {\rm new~inflation:} \qquad r \leq  r_{abs~max} = \frac{8}{N} = 0.16 \quad  ,

835: \ee

836: which is attained by the quadratic monomial potential obtained from 

837: eq.(\ref{trino}) at $ y = 0 $. On the contrary, in chaotic inflation $ r $ 

838: is bounded as

839: $$

840: {\rm chaotic~inflation:} \quad 0.16 = \frac8{N} < r < \frac{16}{N}= 0.32 \quad .

841: $$

842: This bound holds for all values the asymmetry parameter $ h $.

843: The lower and upper bounds for $ r $ are saturated by the quadratic and quartic

844: monomials, respectively.

845: 

846: We see from fig. \ref{ns} that $ n_s $ exhibits a single maximun 

847: $ n_{s~maximum}(h) $ as a function of the quartic coupling $ y $ for fixed 

848: asymmetry $ h $. In fig. \ref{nsh} we plot $ n_{s~maximum}(h) $ as a function

849: of $ h $.  $ n_{s~maximum}(h) $ monotonically increases with $ |h| $ and

850: rapidly reaches its limiting value $ n_{s~maximum} = 0.961528 \ldots $.

851: The corresponding value for $ r $ is $ r =  0.114769\ldots $.

852: Values $ n_s >  n_{s~maximum}  = 0.961528 \ldots $ cannot be described by 

853: new inflation with the trinomial potential eqs.(\ref{VN})-(\ref{trino}).

854: 

855: We see from fig.  \ref{nsr} and  \ref{nsrun} that both $ r $ and the running 

856: $ d n_s/d \ln k $ are {\bf two}-valued functions of $ n_s $ in the interval 

857: $ 0.96 < n_s < 0.961528 \ldots $. That is, for each  $ n_s $ in this range 

858: there are  {\bf two} possible values for $ r $ and for the running 

859: $ d n_s/d \ln k $.

860: Therefore, we can cover the whole range of values $ 0.96 < n_s < 0.961528 

861: \ldots $ choosing the lower branch for $ r $. We find for this branch 

862: $$ 

863: r <  r_{max} \equiv 0.114769\ldots \; . 

864: $$

865: This maximun value $ r_{max} $ is well below the absolute

866: maximun in new inflation $ r_{abs~max} = 0.16 $ [eq.(\ref{cotar})] which

867: belongs to the second branch.

868: 

869: \medskip

870: 

871: The plots of the ratio $ r $ and the running $ d n_s/d \ln k $ as a 

872: function of $ n_s $ show that these quantities are not very sensitive

873: to the asymmetry $ h $ for a given value of $ n_s $.

874: 

875: \bigskip

876: 

877: The three years WMAP \cite{WMAP3} data as well as ref. \cite{2dF}

878: yield for $ n_s $ the value (see also refs. \cite{SDSS} and \cite{Teg})

879: \be \label{nswmap}

880:  n_s = 0.95 \pm 0.02 \; .

881: \ee

882: For $ n_s = 0.95 $ and any value of the asymmetry $ h $ [see fig. \ref{nsr}],

883: new inflation with the trinomial potential eqs.(\ref{VN})-(\ref{trino})

884: yields 

885: $$

886: {\rm new~inflation:} \quad 0.03 < r < 0.04 \quad {\rm and}

887: \quad -0.00070 < d n_s/d \ln k < -0.00055 \; .

888: $$

889: New inflation with the trinomial potential always yield $ n_s $ below

890: the maximun value $ n_{s~maximum} = 0.961528 \ldots $. 

891: For $ n_s = 0.961528 \ldots $ we have in this model $ r =  0.114769\ldots $.

892: These values of the ratio $ r $ fullfil the three years WMAP bound

893: including SDSS galaxy survey \cite{WMAP3}

894: \be \label{rwmap}

895: r < 0.28 \; (95\% ~CL) \; .

896: \ee

897: Moreover, one can see from fig. 14 in ref. \cite{WMAP3} 

898: that $ r < 0.1 ~(68\% ~CL) $ from WMAP$+$SDSS. 

899: 

900: \medskip

901: 

902: Chaotic inflation  with the trinomial potential eq.(\ref{VN})-(\ref{trino})

903: yields larger values of $ r $ than new inflation

904: for a given value of $ n_s $ \cite{nos}. More precisely, 

905: we find $ r = 0.27 $ for $ n_s = 0.95 $ for the binomial potential in

906: chaotic inflation \cite{nos}

907: (the trinomial potential introduces very small changes).

908: 

909: Therefore, although the WMAP value for $ n_s $ [eq.(\ref{nswmap})]

910: is compatible both with chaotic and new inflation, the WMAP bounds on 

911: $ r $ {\bf clearly disfavour} chaotic inflation. New inflation easily fulfils 

912: the three years WMAP bounds on $ r $.

913: 

914: \medskip

915: 

916: The present data do not permit to find the precise values neither

917: of the ratio $ r $ nor of the running index $ d n_s/d \ln k $; 

918: only upper bounds are obtained \cite{WMAP,WMAP3}. We therefore think

919: that the value of $ n_s $ [eq.(\ref{nswmap})] obtained through a fit of the data

920: assuming $ r = 0 $ is more precise than the values of $ n_s $ obtained

921: through fits allowing both $ r $ and $ d n_s/d \ln k $ to vary.

922: Notice that $ n_s = 0.95 $ was independently found from the 2dF data

923: under similar assumptions \cite{2dF}.

924: 

925: Ref. \cite{WMAP3} reports fits  

926: yielding negative values for $ d n_s/d \ln k $ of the order $ \sim -0.05 $.

927: Notice that the order of magnitude of the running $ d n_s/d \ln k $ 

928: is just fixed by the fact that it is a second order quantity in slow-roll:

929: $ \sim \frac1{N^2} \sim 0.0004 $. Still, the negative sign of the running

930: reported by ref.\cite{WMAP3} agrees with the sign prediction of new inflation 

931: with the trinomial potential [see fig. \ref{frun} and \ref{nsrun}]. 

932: 

933: \medskip

934: 

935: In summary, new inflation with the trinomial potential 

936: eq.(\ref{VN})-(\ref{trino}) predicts $ 0.03 < r < 0.04 $ and $ -0.00070 < 

937: d n_s/d \ln k < -0.00055 $ for $ n_s = 0.95 $. For $ 0.93 <  n_s < 0.962 $

938: it predicts $  0.01 < r < 0.115 $ and $ -0.001 < d n_s/d \ln k < -0.0003 $

939: [see figs. \ref{nsr} and \ref{nsrun}].

940: 

941: \section{Limiting Cases of the Trinomial Potential in New Inflation}

942: 

943: Let us now consider the limiting cases: the shallow limit  ($ y \to 0 $),

944:  the steep limit $ y \to \infty $ and the extremely asymmetric limit $ |h| \to 

945: \infty $ of the trinomial potential for new inflation 

946: eqs.(\ref{VN})-(\ref{trino}).

947: 

948: \subsection{The shallow limit $ y \to 0 $ of the Trinomial Potential}

949: 

950: In the shallow  limit $ y \to 0 , z $ tends to $ z = z_+ = (\Delta + |h|)^2 $,

951:  which is the minimum of $ y $ in eq.(\ref{ntrino}).

952: We find from eqs.(\ref{ntrino})-(\ref{dtrino}), 

953: \bea\label{cotrino}

954: && n_s \buildrel{y \to 0}\over= 1 - \frac{2}{N} \simeq 0.96  

955: \quad , \quad \frac{d n_s}{d \ln k} \buildrel{y \to 0}\over= 

956: -\frac2{N^2}\simeq -0.0008 \quad , \cr \cr

957: && r \buildrel{y \to 0}\over= \frac{8}{N}  \simeq 0.16 

958: \quad , \quad 

959: |{\Delta}_{k\;ad}^{(S)}|^2  \buildrel{y \to 0}\over= \frac{N^2}{3 \, \pi^2} \; 

960: \left(\frac{M}{M_{Pl}}\right)^4 \; \Delta(\Delta+|h|) \; ,

961: \eea

962: which coincide with $ n_s , \; \frac{d n_s}{d \ln k}$ and $ r $ 

963: for the monomial quadratic potential. That is, the  $ y \to 0 $ limit

964: is $h$-independent except for $|{\Delta}_{k\;ad}^{(S)}|$.

965: For fixed $ h $ and $ y \to 0 $ the inflaton potential eq.(\ref{trino}) 

966: becomes purely quadratic:

967: \be \label{maslim}

968: w(\chi) \buildrel{y \to 0}\over= \Delta ( \Delta + |h| ) \; (\chi -\chi_+)^2 +

969: {\cal O}(\sqrt{y} )  \; ,

970: \ee

971: where $ \chi_+ \equiv \sqrt{\frac8{y}} \; ( \Delta + |h| ) $.

972: Notice that the amplitude of scalar adiabatic fluctuations eq.(\ref{cotrino})

973: turns out to be proportional to the square mass of the inflaton in this 

974: regime which we read from eq.(\ref{maslim}): $ 2 \; \Delta ( \Delta + |h| ) $. 

975: The shift of the inflaton field by $ \chi_+ $ has no observable consequences.

976: 

977: The numerical values in eq.(\ref{cotrino}) are in agreement with

978: figs. \ref{ns}-\ref{r} in the  $ y \to 0 $ limit. For $ h = 0 $ we recover the

979: results of the monomial potential. 

980: 

981: \subsection{The  steep limit $ y \to \infty $ of the Trinomial Potential}

982: 

983: In the  steep limit $ y \to \infty, \;  z $ tends to zero for new inflation.

984: We find from eq.(\ref{ntrino})

985: \be\label{trikG}

986: y  \buildrel{z \to 0}\over=- F(h) \; \log z

987: -q(h) -1 + {\cal O}(\sqrt{z}) \quad  ,

988: \ee

989: where

990: $$

991: q(h) \equiv 2 \,  F(h) \log[\Delta-|h|] -

992: \frac23 \; \left( h^2 + |h| \; \Delta \right) 

993: \left\{ 8 \, \Delta^2 \, 

994: \log\left[\frac12 \left(1 - \frac{|h|}{\Delta}\right)\right] - 1 \right\}

995: \; ,

996: $$

997: $ q(h) $ is a monotonically increasing function of the asymmetry 

998: $ |h| : \; 0 \leq q(h) < \infty $ for $ 0 < |h| < \infty $.

999: 

1000: Then, eqs.(\ref{nstrino})-(\ref{rtrino}) yield,

1001: \bea\label{nsrtrikG}

1002: &&n_s \buildrel{y \gg 1}\over=1 - \frac{y}{N \; F(h)}

1003: \quad , \quad r \buildrel{y \gg 1}\over= 

1004: \frac{16 \; y}{N \; F^2(h)} \, e^{- \frac{y+1+q(h)}{F(h)}}\quad ,

1005: \cr \cr

1006: && \frac{d n_s}{d \ln k}\buildrel{y \gg 1}\over= -\frac{2 \; y^2 \; |h|}{N^2 \;

1007:  F^2(h)} \, e^{- \frac{y+1+q(h)}{2 \; F(h)}}\quad , \cr \cr

1008: && |{\Delta}_{k\;ad}^{(S)}|^2  \buildrel{y \gg 1}\over= 

1009: \frac{N^2}{12 \, \pi^2} \; \left(\frac{M}{M_{Pl}}\right)^4 \; 

1010: \frac{F^3(h)}{y^2} \; e^{\frac{y+1+q(h)}{F(h)}} \quad .

1011: \eea

1012: In the $ h \to 0 $ limit we recover from eqs.(\ref{trikG})-(\ref{nsrtrikG})

1013: the results for new inflation with a purely quartic potential:

1014: we have $ F(0) = 1 $ and $ q(0) = 0 $ and eq.(\ref{nsrtrikG}) becomes,

1015: \bea\label{nsrtrikGh0}

1016: &&n_s \buildrel{y \gg 1, h \to 0}\over=1 - \frac{y}{N}

1017: \quad , \quad r \buildrel{y \gg 1, h \to 0}\over= 

1018: \frac{16 \; y}{N} \, e^{-y-1}\quad ,

1019: \cr \cr

1020: && \frac{d n_s}{d \ln k}\buildrel{y \gg 1, h \to 0}\over= 

1021: -\frac{2 \; y^2 \; |h|}{N^2} \, e^{-y-1}\quad , \cr \cr

1022: && |{\Delta}_{k\;ad}^{(S)}|^2  \buildrel{y \gg 1, h \to 0}\over= 

1023: \frac{N^2}{12 \, \pi^2} \; \left(\frac{M}{M_{Pl}}\right)^4 \; 

1024: \frac{e^{y+1}}{y^2} \;  \quad .

1025: \eea

1026: The behaviour in eqs.(\ref{nsrtrikG}) is in agreement with 

1027: figs. \ref{ns}-\ref{r} in the  $ y \to +\infty $ limit.

1028: 

1029: \subsection{The extremely asymmetric limit $ |h| \to \infty $ 

1030: of the Trinomial Potential}

1031: 

1032: Eqs.(\ref{ntrino})-(\ref{dtrino}) have a finite limit for $ |h| \to \infty 

1033: $ with $ y $ and $ z $ scaling as $ h^2 $. Define,

1034: $$

1035: Z \equiv \frac{z}{h^2} \quad , \quad Y \equiv \frac{y}{h^2} \; .

1036: $$

1037: Then, we find for $ |h| \to \infty $ 

1038: from eqs.(\ref{ntrino})-(\ref{dtrino}) keeping $ Z $ and $ Y $ fixed,

1039: \bea \label{hgra}

1040: && Y = Z - \frac43 \; \sqrt{Z} - 4 - \frac43 \; \log\frac{Z}4 +

1041: \frac{16}{3 \; \sqrt{Z}}\; , \cr \cr

1042: && n_s = 1 - 6 \; \frac{Y}{N} \; \frac{Z^2 \; (\sqrt{Z} - 2)^2}{

1043: [\frac{16}3 - \frac83 \; Z^{\frac32} + Z^2 ]^2 } + \frac{Y}{N} \;

1044: \frac{ 3 \; Z - 4 \; \sqrt{Z} }{\frac{16}{3}  - \frac83 \; Z^{\frac32} + Z^2 }

1045: \; , \\ \cr

1046: &&\frac{d n_s}{d \ln k}= - \frac2{N^2} \;  Y^2 \; Z \; 

1047: \frac{(\sqrt{Z} -2)(\frac32 \; \sqrt{Z}-1)}{

1048: \left[\frac{16}3  - \frac83 \,  Z^{3/2} + Z^2 \right]^2} \cr\cr

1049: &&- \frac{24}{N^2} \;  Y^2 \; Z^4 \; \frac{(\sqrt{Z} -2)^4}{

1050: \left[\frac{16}3  - \frac83 \,  Z^{3/2} + Z^2 \right]^4} 

1051: + \frac8{N^2} \;  Y^2 \; Z^{\frac52} \; 

1052: \frac{(3 \,\sqrt{Z} - 4)(\sqrt{Z} -2)^2}{

1053: \left[\frac{16}3  - \frac83 \,  Z^{3/2} + Z^2 \right]^3}

1054: \quad , \label{runh} \\ \cr \cr

1055: && r = 16 \; \frac{Y}{N} \; \frac{Z^2 \; (\sqrt{Z} - 2)^2}{

1056: [\frac{16}{3} - \frac83 \; Z^{\frac32} + Z^2 ]^2 }  \; ,  \cr \cr

1057: && |{\Delta}_{k\;ad}^{(S)}|^2  = \frac{N^2 \; h^2}{12 \, \pi^2} \; 

1058: \left(\frac{M}{M_{Pl}}\right)^4 \;

1059: \frac{[\frac{16}{3}  - \frac83 \; Z^{\frac32} + Z^2]^2}{Y^2 \; Z^2  

1060: \; (\sqrt{Z} - 2)^2} \; . \label{amplihg}

1061: \eea

1062: We have $ 0 \leq Z \leq 4 $ for $ +\infty \geq Y \geq 0 $. 

1063: In the  $ |h| \to \infty $ limit the inflaton potential takes the form

1064: $$

1065: W(\chi) \equiv \lim_{|h| \to \infty} \frac{w(\chi)}{h^2} =

1066: \frac{32}{3 \; Y} - \frac13 \; \sqrt{\frac{Y}{2}} \; \chi^3 

1067: + \frac{Y}{32} \; \chi^4 \; .

1068: $$

1069: This is a broken symmetric potential without quadratic term.

1070: Notice that the cubic coupling has dimension of a mass in eq.(\ref{VN})

1071: and hence this is {\bf not} a massless potential contrary to the quartic 

1072: monomial $ \chi^4 $. In addition, eq.(\ref{amplihg}) shows that for large 

1073: $ |h| $ one must keep the product $ |h| \; M^2 $ fixed

1074: since it is determined by the amplitude of the adiabatic perturbations.

1075: We see from eq.(\ref{amplihg}) that $ {\tilde M} \equiv \sqrt{|h|} \; M $ becomes 

1076: the energy scale of inflation in the $ |h| \to \infty $ limit.

1077: $ {\tilde M} \sim 10^{16}$GeV according to the observed value of 

1078: $ |{\Delta}_{k\;ad}^{(S)}|/N $ displayed in eq.(\ref{Mwmap}), while

1079: $ M $ and $ m $,

1080: $$

1081: M = \frac{\tilde M}{\sqrt{|h|}} \buildrel{ |h| \to \infty 

1082: }\over= 0 \quad ,  \quad

1083: m = \frac{M^2}{M_{Pl}} = \frac{ {\tilde M}^2}{|h| \; M_{Pl}} \buildrel{ |h| \to \infty 

1084: }\over= 0 \quad .

1085: $$

1086: vanish as  $ |h| \to \infty $.

1087: 

1088: \medskip

1089: 

1090: The curves in figs. \ref{ns}-\ref{nsh}

1091: for high values of $ |h| $ are well described by eq.(\ref{hgra}).

1092: 

1093: \section{Hybrid Inflation}

1094: 

1095: In the inflationary models of hybrid type, the inflaton is coupled to another 

1096: scalar field $ \sigma_0 $ with mass term $ -\mu_0^2 < 0 $

1097: through a potential of the type \cite{lin}

1098: \bea\label{Vhib1}

1099: &&V_{hyb}(\phi,\sigma_0) = \frac{m^2}{2} \; \phi^2 + \frac{g_0^2}{2} \; 

1100: \phi^2 \; \sigma_0^2 + \frac{\mu_0^4}{16 \, \Lambda_0} 

1101: \left(\sigma_0^2 -  \frac{4 \, \Lambda_0}{\mu_0^2} \right)^2= \cr \cr

1102: &&  =\frac{m^2}{2} \; \phi^2 + \Lambda_0+\frac12 \; (g_0^2 \; 

1103: \phi^2-\mu_0^2) \;  \sigma_0^2+\frac{\mu_0^4}{16 \; \Lambda_0} \; 

1104: \sigma_0^4\; ,

1105: \eea

1106: where $ m^2 > 0 , \; \Lambda_0 > 0 $ plays the role of a cosmological constant 

1107: and $ g_0^2 $  couples $ \sigma_0 $ with $ \phi $.

1108: 

1109: \medskip

1110: 

1111: The initial conditions are chosen such that $\sigma_0$ and $ \dot\sigma_0$ 

1112: are very small (but not identically zero) and therefore one can consider 

1113: initially,

1114: \be\label{vsig0}

1115: V_{hyb}(\phi,0) =\frac{m^2}{2} \; \phi^2 + \Lambda_0 \; .

1116: \ee

1117: One has then inflation driven by the cosmological constant $ \Lambda_0 $ 

1118: plus the initial value of the inflaton $  \phi(0) $.

1119: The inflaton field $ \phi(t) $ 

1120: decreases with time while the scale factor $ a(t) $ grows exponentially 

1121: with time. We see from eq.(\ref{Vhib1}) that

1122: \be\label{masefe}

1123: m_{\sigma}^2 = g_0^2 \; \phi^2 - \mu_0^2  \; ,

1124: \ee

1125: plays the role of a effective classical mass square for the field $ \sigma_0 $.

1126: The initial value of $ m_{\sigma}^2 $ depends on the initial conditions

1127: but is typically positive.

1128: In any case, since  the inflaton field $ \phi $ decreases

1129: with time, $ m_{\sigma}^2 $ will be necessarily negative at some 

1130: moment during inflation.

1131: At such moment, spinodal (tachyonic) unstabilities appear and the

1132: field $ \sigma_0 $ starts to grow exponentially. 

1133: Inflation stops when both fields $ \phi $ and $ \sigma_0 $ reach their

1134: vaccum values. A matter dominated regime follows.

1135: 

1136: \bigskip

1137: 

1138: Normally, the field $ \sigma_0 $

1139: is negligible when the relevant cosmological scales cross out the 

1140: horizon. Hence, $ \sigma_0 $ does not affect the spectrum of density and

1141: tensor fluctuations except through the number of efolds. 

1142: Hence, hybrid inflation is a single-field inflationary model

1143: as long as $ \phi $ solely contributes to the spectrum of density and tensor 

1144: fluctuations. However, we find specific regions of parameters 

1145: $ (g_0, \mu_0 , \Lambda_0 ) $ and initial 

1146: conditions where both fields  $ \phi $ and $ \sigma_0 $ contribute to the 

1147: cosmologically relevant fluctuations making hybrid inflation a two-field 

1148: inflationary model. We do not consider such  regions of parameters here which 

1149: are outside the scope of this paper.

1150: 

1151: \bigskip

1152: 

1153: In terms of the dimensionless fields and couplings, the potential  

1154: $ V_{hyb}(\phi,\sigma_0) $  eq.(\ref{Vhib1}) reads

1155: \bea

1156: && w(\chi,\sigma) = \frac12 \; \chi^2 

1157: + \frac{\mu^4}{8 \; \Lambda}\left(\sigma^2 - \frac{2 \, \Lambda}{\mu^2} 

1158: \right)^2 + \frac12 \; g^2 \; \sigma^2 \; \chi^2 = \cr \cr

1159: &&= \frac12 \; \chi^2 + \frac12 \; ( g^2 \; \chi^2-\mu^2 ) \; \sigma^2  +

1160: \frac12 \;\Lambda + \frac{\mu^4}{8 \; \Lambda} \; \sigma^4 \; .

1161: \eea

1162: where 

1163: $$ \sigma(\tau) \equiv \frac{\sigma_0(t)}{\sqrt{N} \; M_{Pl}} \quad , \quad

1164:  g^2 \equiv g_0^2 \; \frac{N \; M_{Pl}^2}{m^2}\quad , \quad

1165: \mu^2 \equiv \frac{\mu_0^2}{m^2} \quad  \mbox{and}  

1166: \quad \Lambda \equiv \frac{2 \, \Lambda_0}{M^4 \; N} \; .

1167: $$

1168: The evolution equations for this potential in dimensionless variables 

1169: take the form

1170: \bea\label{ecmovh}

1171: && {\cal H}^2(\tau) = \frac1{6} \; \left[ \frac1{N} \; {\dot \chi}^2 + \chi^2 

1172: + \frac{\mu^4}{4 \; \Lambda}\left(\sigma^2 - \frac{2 \, \Lambda}{\mu^2} 

1173: \right)^2 + g^2 \; \sigma^2 \; \chi^2

1174: \right] \; , \cr \cr

1175: && 

1176: \left[\frac1{N} \; \frac{d^2}{d \tau^2}  + 3 \,  {\cal H} \,\frac{d}{d \tau}

1177: + 1 + g^2 \; \sigma^2 \right] \chi = 0 \; , \label{fihib} \\

1178: && 

1179: \left[\frac1{N} \; \frac{d^2}{d \tau^2}  + 3 \,  {\cal H} \,\frac{d}{d \tau}

1180:  - \mu^2+ g^2 \; \chi^2 +\frac{\mu^4}{2 \; \Lambda} \; \sigma^2 \right]

1181: \sigma(\tau) = 0  \; . \nonumber

1182: \eea

1183: Since the field $\sigma$ is chosen initially very small, it

1184: can be neglected and we can approximate the evolution equations 

1185: (\ref{fihib}) as

1186: \be \label{ecmslo}

1187: 3 \,  {\cal H} \, {\dot \chi} + \chi = 0 \quad , \quad

1188:  {\cal H}^2(\tau) = \frac1{6} \; \left[ \chi^2 + \Lambda  \right] \; .

1189: \ee

1190: The number of efolds from the time $\tau$ till the end of inflation

1191: is then given by eq.(\ref{Nchi}),

1192: \be\label{Nhib}

1193: N(\tau) = N \; \int_{\tau}^{\tau_{end}}  {\cal H}(\tau) \; d\tau =  

1194: -  \int_{\chi(\tau)}^{\chi_{end}} \frac{w(\chi)}{w'(\chi)} \; d\chi

1195: = \frac{N}4 \left[ \chi^2(\tau)- \chi^2_{end} \right]

1196: + \frac{N}2 \; \Lambda \; \log\frac{\chi(\tau)}{\chi_{end}}  \; ,

1197: \ee

1198: $ \chi_{end} $ is the inflaton field at the end of inflation.

1199: 

1200: \medskip

1201: 

1202: We see that the inflaton field and its dynamics only appear in $ N(\tau) $ 

1203: eq.(\ref{Nhib}) through the value of $ \chi_{end} $ where inflation stops. The

1204: value of $ \chi_{end} $ follows by solving eqs.(\ref{ecmovh}) and 

1205: depends on the initial conditions as well as on the parameters 

1206: $ g, \; \mu $ and $ \Lambda $.

1207: 

1208: \medskip

1209: 

1210: The spectral indices are given by eqs.(\ref{indi}) and

1211: the amplitude of adiabatic perturbations by eq.(\ref{ampliI}).

1212: By using the potential eq.(\ref{vsig0}) in dimensionless variables we find,

1213: \bea\label{indhi}

1214: && w(\chi,0) = \frac12 \, ( \chi^2 + \Lambda) \quad , \cr \cr

1215: && |{\Delta}_{k\;ad}^{(S)}|^2 = \frac{N^2}{96 \, \pi^2 } 

1216: \left( \frac{M}{M_{Pl}} 

1217: \right)^4 \; \frac{(\chi^2 + \Lambda)^3}{\chi^2} \quad ,  \quad 

1218: r = \frac{32}{N} \; \frac{\chi^2}{(\chi^2 + \Lambda)^2}  \; ,  \\ \cr

1219: &&n_s = 1 + \frac4{N} \; \frac{\Lambda - 2 \; \chi^2}{(\chi^2 + \Lambda)^2}

1220: \quad ,  \quad \frac{d n_s}{d \ln k}= \frac{32}{N^2} \; 

1221: \frac{\chi^2(2 \,\Lambda-\chi^2)}{(\chi^2 + \Lambda)^4} 

1222: \label{nshib} \; , 

1223: \eea

1224: where $ \chi $ is the inflaton at the moment of the first horizon crossing.

1225: 

1226: \begin{figure}[p]

1227: \begin{turn}{-90}

1228: \centering

1229: \includegraphics[width=10cm,height=14cm]{a}

1230: \end{turn}

1231: \caption{Hybrid inflation. The logarithm of the scale factor 

1232: (the number of efolds) as a function of time. The chosen parameters in 

1233: eq.(\ref{ecmovh}) are $ \Lambda = 4 \; N = 200 \;  , \; \mu^2 = 1.7 \; 

1234: \Lambda, \; \phi(0) = 2.3 \; \sqrt{\Lambda} $. 

1235: A stage of slow-roll quasi de Sitter

1236: inflation takes place (till $ \tau \simeq 39 $ in this example) followed by

1237: a matter dominated era.}

1238: \label{a}

1239: \end{figure}

1240: 

1241: 

1242: \begin{figure}[p]

1243: \begin{turn}{-90}

1244: \centering

1245: \includegraphics[width=10cm,height=14cm]{h}

1246: \end{turn}

1247: \caption{ Hybrid inflation. The Hubble parameter $ h $ and its inverse $ 1/h $ 

1248: as a function of time. Same parameters as in fig. \ref{a}.  $ h $ slowly

1249: decreases with time in the slow-roll quasi de Sitter stage 

1250: (till $ \tau \simeq 39 $ in this example) followed by 

1251: $ h \simeq 2/[3 \; \tau ] $ in the  matter dominated era.}

1252: \label{h}

1253: \end{figure}

1254: 

1255: \begin{figure}[p]

1256: \begin{turn}{-90}

1257: \centering

1258: \psfrag{"fi.dat"}{$ \chi $ vs. time } 

1259: \psfrag{"fip.dat"}{$ \dot \chi $ vs. time }

1260: \includegraphics[width=10cm,height=14cm]{fi}

1261: \end{turn}

1262: \caption{Hybrid inflation. The inflaton field $ \chi $ and its time derivative

1263: as a function of time. Same parameters as in fig. \ref{a}.

1264: $ |\dot \chi | \ll | \chi | $ during the slow-roll inflationary stage.

1265: Inflation stops when $  \chi \sim {\dot \chi} \sim 0 $ 

1266:  (at $ \tau \simeq 39 $ in this example).}

1267: \label{fi}

1268: \end{figure}

1269: 

1270: \begin{figure}[p]

1271: \begin{turn}{-90}

1272: \centering

1273: \psfrag{"s.dat"}{$ \sigma $ vs. time } 

1274: \psfrag{"sp.dat"}{$ \dot \sigma $ vs. time }

1275: \includegraphics[width=10cm,height=14cm]{si}

1276: \end{turn}

1277: \caption{Hybrid inflation. The field sigma $ \sigma $ and its time derivative

1278: as a function of time for hybrid inflation. Same parameters as in fig. \ref{a}.

1279: The fields  $ \sigma $ and  $ \dot\sigma $ start with small values and grow 

1280: exponentially fast when $ m_{\sigma}^2 < 0 $ [eq.(\ref{masefch})]

1281: (at $ \tau \simeq 4 $ in this example). }

1282: \label{si}

1283: \end{figure}

1284: 

1285: 

1286: \begin{figure}[p]

1287: \begin{turn}{-90}

1288: \centering

1289: \psfrag{"pe.dat"}{pressure/energy density vs. time } 

1290: \includegraphics[width=10cm,height=14cm]{pe}

1291: \end{turn}

1292: \caption{Hybrid inflation.The equation of state pressure/energy density

1293: as a function of time. Same parameters as in fig. \ref{a}.

1294: The equation of state clearly shows the two stages:

1295: pressure/energy $ = - 1 $ during inflation followed by oscillations with zero average

1296: pressure in the matter dominated era.}

1297: \label{pe}

1298: \end{figure}

1299: 

1300: \begin{figure}[p]

1301: \begin{turn}{-90}

1302: \centering

1303: \psfrag{"nsr11.dat"}{${\hat \chi}(0)=0.01$}

1304: \psfrag{"nsr9.dat"}{$ {\hat \chi}(0)=0.05 $}

1305: \psfrag{"nsr1.dat"}{$ {\hat \chi}(0)=0.1 $} 

1306: \psfrag{"nsr5.dat"}{$ {\hat \chi}(0)=0.2 $}

1307: \psfrag{"nsr8.dat"}{$ {\hat \chi}(0)=0.3 $}

1308: \psfrag{"nsr4.dat"}{$ {\hat \chi}(0)=0.4 $} 

1309: \includegraphics[width=10cm,height=14cm]{nsr005A}

1310: \end{turn}

1311: \caption{Hybrid inflation. The ratio $ r $ vs. $ n_s $ for $ \mu^2 =

1312: 0.05  \; \Lambda  <  \mu^2_{crit} \; , \; g^2 = \frac 14 $ and 

1313: $ 0.01 \; \sqrt{\Lambda}\leq \chi(0)  

1314: \leq 0.4 \; \sqrt{\Lambda} $. 

1315: Notice that here $ r $ has a maximun as a function of $ n_s $.

1316: In addition, $ n_s > 1 $ for all values of $ \Lambda $ and $ {\hat \chi}(0) $

1317: since  $ \mu^2 < \mu^2_{crit} \simeq 0.13 \; \Lambda$. $ r $ increases with 

1318: ${\hat \chi}(0)$ for this range of ${\hat \chi}(0)$.}

1319: \label{nsr005A}

1320: \end{figure}

1321: 

1322: \section{Inflaton Dynamics in Hybrid Inflation}

1323: 

1324: In figs. \ref{a}-\ref{pe} we display the numerical solution of the 

1325: equations of motion (\ref{ecmovh}) as functions of time for 

1326: $ \Lambda = 4 \; N = 200 \; , \; \mu^2 = 1.7 \; \Lambda, \; 

1327: \phi(0) = 2.3 \; \sqrt{\Lambda} $. We see first a stage of slow-roll

1328: quasi-de Sitter inflation till $ \tau \simeq 39 $ in this example. 

1329: Namely, $ | {\dot \phi} | \ll | \phi | $ and $ h $ are practically constant

1330: during this lapse. In this slow-roll stage the equations of motion 

1331: (\ref{ecmslo}) can be integrated in close form with the solution

1332: \be \label{solslo}

1333: \sqrt{\frac2{3 \; \Lambda}} (\tau - \tau_0) = {\rm Arg~ tanh}\frac1{\sqrt{

1334: 1 + \frac{\chi^2}{\Lambda} }} - \sqrt{1 + \frac{\chi^2}{\Lambda} } \; ,

1335: \ee

1336: which defines $ \chi = \chi(\tau) $ and where $ \tau_0 $ is an integration

1337: constant. Notice that $ \chi(\tau) $ is a monotonically decreasing function

1338: of time since $  {\dot \chi} = -\chi/[3 \; h] < 0 $ [eq.(\ref{ecmslo})].

1339: 

1340: When $ \chi \ll  \Lambda $, eq.(\ref{solslo})

1341: approximates by

1342: $$

1343: \chi(\tau) \simeq \chi_0 \; e^{-\sqrt{\frac2{3 \; \Lambda}} \; \tau} \; ,

1344: \quad  \chi \ll  \Lambda \quad \; ,

1345: $$

1346: while in the opposite limit $ \chi \gg \Lambda $

1347: from eq.(\ref{solslo})  we have,

1348: $$

1349: \chi(\tau) \simeq \chi_1 - \sqrt{\frac23} \; \tau  \; ,

1350: \quad  \chi \gg  \Lambda \quad \; .

1351: $$

1352: Here, $ \chi_0 $ and $ \chi_1 $ are integration constants.

1353: 

1354: \medskip

1355: 

1356: We have verified that eq.(\ref{solslo}) as well as eq.(\ref{Nhib}) provide

1357: an excellent approximation to the numerical solution of eqs. (\ref{fihib}). 

1358: 

1359: The number of efolds during inflation is about

1360: $280$ in the example depicted in figs. \ref{a}-\ref{pe}, 

1361: larger than the required minimun of about $60$ efolds.

1362: This stage is followed by a matter dominated era. 

1363: We choose a very small initial  amplitude for the 

1364: sigma field and its time derivative. The sigma field stays very small till

1365: its effective mass square [eq.(\ref{masefe})] becomes negative and

1366: spinodal unstabilities show up.

1367: At this moment, ($ \tau \simeq 4 $ in this example) the sigma field as well as

1368: its time derivative start to increase exponentially fast till the growth

1369: of the non-linear term $ +\frac{\mu^4}{2 \; \Lambda} \; \sigma^2 $ 

1370: in the last equation in eq.(\ref{ecmovh}) shuts off the unstabilities.

1371: 

1372: Inflation stops at the moment when both $ \chi $ and $ \sigma $ are comparable 

1373: with $ \dot \chi $ and  $ \dot \sigma $ ($ \tau \simeq 39 $ in this example). 

1374: At this time, both $ \chi $ and 

1375: $ \sigma $ are very close to their vaccum values $ \chi_{vac} = 0 $ and

1376: $ \sigma_{vac} = \frac{\sqrt{2 \; \Lambda}}{\mu} $. 

1377: That is, when the kinetic terms become relevant, the energy is no more 

1378: dominated by the vaccum energy. At the same time, the slow roll 

1379: approximation ceases to be valid. 

1380: 

1381: The time when the effective mass of the field $ \sigma $ 

1382: [see eq.(\ref{masefe})]

1383: \be \label{masefch}

1384: m_{\sigma}^2 = m^2 \; ( g^2 \; \chi^2 - \mu^2)

1385: \ee

1386: becomes negative and $ \sigma $ starts to grow depends on the

1387: values of $ \mu^2 $ and $ g^2 \; \chi^2(0) $. For low values of 

1388: $ \mu^2 $: (typically for $ \mu^2 < 0.08 \; \Lambda $ when 

1389: $ \chi^2(0) < \Lambda $, and $ \mu^2 < 0.2 \; \Lambda $ when

1390: $ \chi^2(0) < 2. \; \Lambda $), the field  $ \sigma $ starts to grow

1391: close to the end of inflation. On the contrary, for higher values of 

1392: $ \mu^2 $ (typically for $ \mu^2 > 0.08 \; \Lambda $ when

1393: $ \chi^2(0) < \Lambda $, and  $ \mu^2 > 0.23 \; \Lambda $ when 

1394: $ \chi^2(0) < 2. \; \Lambda $), the field  $ \sigma $ starts to grow

1395: well before the end of inflation.  This is explained by the fact

1396: that the scale of time variation of $ \sigma $ goes as

1397: $ \mu^{-1} $;  $ \sigma $ evolves slowly for small $ \mu $ and

1398: fastly for large $ \mu $.

1399: 

1400: \medskip

1401: 

1402: $ \dot \sigma $ exhibits a peak

1403: around the point where $ m_{\sigma}^2 $ changes sign and then returns 

1404: to a very small value while $ \sigma $ slowly approaches its vaccum

1405: value. This evolution is depicted in fig. \ref{si}.

1406: 

1407: \medskip

1408: 

1409: In the example depicted in figs. \ref{a}-\ref{pe} the effective mass square 

1410: [eq.(\ref{masefch})] of the $ \sigma $ field changes sign at $ \tau \sim 4 $ 

1411: well before $ \chi $ reaches its vaccum value (zero) and inflation ends. 

1412: This follows from the choice of a large value for $ \mu^2 $  in figs. 

1413: \ref{a}-\ref{pe}. For smaller values of $ \mu^2 $, $ m_{\sigma}^2 $  

1414: [eq.(\ref{masefch})] flips its sign later when the inflaton $ \chi $ is 

1415: much smaller. 

1416: 

1417: \medskip

1418: 

1419: In order to compute the observables  $ n_s , \; r $ and $ d n_s /d \ln k $ 

1420: from eq.(\ref{indhi}) we need the value of the inflation field

1421: $ \chi $ at $50$ efolds before the end of inflation.

1422: We thus integrated numerically eqs.(\ref{ecmovh}) till the end of

1423: inflation and then extracted the value of $ \chi $ at $50$ efolds before.

1424: We define the end of inflation as the point where the ratio

1425: pressure over energy reaches $10\%$. This gives $ \tau \equiv \tau_{end}

1426: \simeq 34 $ for the example in fig. \ref{pe}.

1427: 

1428: \medskip

1429: 

1430: At $ \Lambda = 0 $ hybrid inflation becomes chaotic inflation with

1431: the monomial potential $ \frac12  \; \chi^2 $. We want to stress that

1432: only {\bf at} $ \Lambda = 0 $ hybrid inflation becomes chaotic inflation.

1433: For any  value of $ \Lambda > 0 $ (even very small) 

1434: the features of hybrid inflation remain. The time $ \tau_{end} $

1435: gets longer and longer for  $ \Lambda \to 0^+ $.

1436: 

1437: \section{Spectral index $n_s$, ratio $r$ and running index $ \frac{d n_s}{d \ln k} $

1438: in Hybrid Inflation}

1439: 

1440: We see from eqs.(\ref{nshib}) that the field $ \chi $ naturally scales

1441: as $ \sqrt{\Lambda} $. It is then convenient to introduce the rescaled

1442: field and the rescaled mass

1443: \be

1444: {\hat \chi} \equiv \frac{\chi}{\sqrt{\Lambda}} \quad , \quad 

1445: {\hat \mu}^2 \equiv  \frac{\mu^2}{\Lambda} \; .

1446: \ee

1447: Then, eqs.(\ref{nshib}) take the form

1448: \bea\label{indhiR}

1449: && w(\chi,0) = \frac{\Lambda}2 \, ( {\hat \chi}^2 + 1) \quad , \cr \cr

1450: && |{\Delta}_{k\;ad}^{(S)}|^2 = \frac{N^2 \; \Lambda^2}{96 \, \pi^2 } 

1451: \left( \frac{M}{M_{Pl}} 

1452: \right)^4 \; \frac{( {\hat \chi}^2 + 1)^3}{ {\hat \chi}^2} \quad ,  \quad 

1453: r = \frac{32}{N\; \Lambda} \; 

1454: \frac{ {\hat \chi}^2}{({\hat \chi}^2 + 1)^2}  \; ,  \\ \cr

1455: &&n_s = 1 + \frac4{N \; \Lambda} \; 

1456: \frac{1 - 2 \; {\hat \chi}^2}{({\hat \chi}^2 + 1)^2}

1457: \quad ,  \quad \frac{d n_s}{d \ln k}= \frac{32}{N^2\; \Lambda^2} \; 

1458: \frac{{\hat \chi}^2(2 -{\hat \chi}^2)}{({\hat \chi}^2 + 1)^4} 

1459: \label{nshibR} \; . 

1460: \eea

1461: Notice that $ (n_s - 1) $ may have either sign according to eq.(\ref{nshibR}).

1462: Hybrid inflation is usually associated with red tilted spectrum ($ n_s > 1 $).

1463: However, both regimes, $ n_s > 1 $ {\bf and} $ n_s < 1 $ are

1464: realized by hybrid inflation.

1465: 

1466: \medskip

1467: 

1468: \begin{figure}[p]

1469: \begin{turn}{-90}

1470: \centering

1471: \psfrag{"nsr3.dat"}{${\hat \chi}(0)=2.4$}

1472: \psfrag{"nsr6.dat"}{${\hat \chi}(0)=0.8$}

1473: \psfrag{"nsr7.dat"}{${\hat \chi}(0)=0.5$}

1474: \psfrag{"nsr10.dat"}{${\hat \chi}(0)=1.0$}

1475: \psfrag{"nsr.dat"}{$  {\hat \chi}(0)=0.6 $}

1476: \includegraphics[width=10cm,height=14cm]{nsr005B}

1477: \end{turn}

1478: \caption{Hybrid inflation. The ratio $ r $ vs. $ n_s $ for $ \mu^2 =

1479: 0.05  \; \Lambda   <  \mu^2_{crit} \; , \; g^2 = \frac 14 $ and 

1480: $ 0.5 \; \sqrt{\Lambda}\leq \chi(0)  

1481: \leq 2.4 \; \sqrt{\Lambda} $. 

1482: Notice that here $ r $ has a maximun as a function of $ n_s $.

1483: In addition, $ n_s > 1 $ for all values of $ \Lambda $ and $ \chi(0) $ since  

1484: $ \mu^2 < \mu^2_{crit} \simeq 0.13 \; \Lambda$. $ r $ decreases with 

1485: ${\hat \chi}(0)$ for this range of ${\hat \chi}(0)$.}

1486: \label{nsr005B}

1487: \end{figure}

1488: 

1489: \begin{figure}[p]

1490: \begin{turn}{-90}

1491: \centering

1492: \psfrag{"nsr4.dat"}{${\hat \chi}(0)=0.7$}

1493: \psfrag{"nsr5.dat"}{${\hat \chi}(0)=0.9$}

1494: \psfrag{"nsr6.dat"}{${\hat \chi}(0)=1.1$}

1495: \psfrag{"nsr7.dat"}{${\hat \chi}(0)=1.3$}

1496: \psfrag{"nsr8.dat"}{${\hat \chi}(0)=1.5$}

1497: \psfrag{"nsr9.dat"}{${\hat \chi}(0)=1.7$}

1498: \psfrag{"nsr10.dat"}{${\hat \chi}(0)=1.9$}

1499: \psfrag{"nsr11.dat"}{${\hat \chi}(0)=2.1$}

1500: \psfrag{"nsr12.dat"}{${\hat \chi}(0)=2.3$}

1501: \includegraphics[width=10cm,height=14cm]{nsr13}

1502: \end{turn}

1503: \caption{Hybrid inflation. The ratio $ r $ vs. $ n_s $ for $ \mu^2 =

1504: 0.13 \; \Lambda \simeq \mu^2_{crit} \; , \; g^2 = \frac 14 $ and 

1505: $ 0.7 \; \sqrt{\Lambda}\leq \chi(0) \leq 2.3 \; \sqrt{\Lambda} $. 

1506: Notice that here $ n_s > 1 $ for all values of $ \Lambda $ and $ \chi(0) $.}

1507: \label{nsrh}

1508: \end{figure}

1509: 

1510: \begin{figure}[p]

1511: \begin{turn}{-90}

1512: \centering

1513: \psfrag{"nsrun4.dat"}{${\hat \chi}(0)=0.7$}

1514: \psfrag{"nsrun5.dat"}{${\hat \chi}(0)=0.9$}

1515: \psfrag{"nsrun6.dat"}{${\hat \chi}(0)=1.1$}

1516: \psfrag{"nsrun7.dat"}{${\hat \chi}(0)=1.3$}

1517: \psfrag{"nsrun8.dat"}{${\hat \chi}(0)=1.5$}

1518: \psfrag{"nsrun9.dat"}{${\hat \chi}(0)=1.7$}

1519: \psfrag{"nsrun10.dat"}{${\hat \chi}(0)=1.9$}

1520: \psfrag{"nsrun11.dat"}{${\hat \chi}(0)=2.1$}

1521: \psfrag{"nsrun12.dat"}{${\hat \chi}(0)=2.3$}

1522: \includegraphics[width=10cm,height=14cm]{nsrun13}

1523: \end{turn}

1524: \caption{Hybrid inflation. The running $ d n_s/ d \ln k $ vs. $ n_s $ for 

1525: $ \mu^2 = 0.13 \; \Lambda \simeq \mu^2_{crit} \; , \; g^2 = \frac 14 $ and 

1526: $ 0.7 \; \sqrt{\Lambda}\leq \chi(0)  \leq 2.3 \; \sqrt{\Lambda} $. 

1527: Notice that here $  d n_s/ d \ln k > 0 $ for all values of $ \Lambda $ and 

1528: $ \chi(0) $. It exhibits a shape similar to $ r $ vs. $ n_s $ 

1529: [see fig. \ref{nsr005B}].}

1530: \label{nsrunh}

1531: \end{figure}

1532: 

1533: \begin{figure}[p]

1534: \begin{turn}{-90}

1535: \centering

1536: \psfrag{"ns7.dat"}{${\hat \chi}(0)=2.7$}

1537: \psfrag{"ns10.dat"}{${\hat \chi}(0)=3.3$}

1538: \psfrag{"ns12.dat"}{${\hat \chi}(0)=3.7$}

1539: \psfrag{"ns14.dat"}{${\hat \chi}(0)=4.1$}

1540: \psfrag{"ns16.dat"}{${\hat \chi}(0)=4.5$}

1541: \psfrag{"ns17.dat"}{${\hat \chi}(0)=4.7$}

1542: \psfrag{"ns19c.dat"}{${\hat \chi}(0)=5.1$}

1543: \psfrag{"ns20c.dat"}{${\hat \chi}(0)=5.3$}

1544: \psfrag{"ns21c.dat"}{${\hat \chi}(0)=5.7$}

1545: \includegraphics[width=10cm,height=14cm]{nsZ}

1546: \end{turn}

1547: \caption{Hybrid inflation. The index $ n_s  $ vs. $ \Lambda $ for $ \mu^2 =

1548: 1.7 \; \Lambda > \mu^2_{crit} , \; g^2 = \frac 14 $ and 

1549: $ 2.7 \; \sqrt{\Lambda}\leq \chi(0) \leq 5.7 \; \sqrt{\Lambda} \simeq 

1550: \chi(0)_{crit} $. Notice that here $ n_s < 1 $ for all values of $ \Lambda $ 

1551: and this range of $ \chi(0) $.}

1552: \label{nsY}

1553: \end{figure}

1554: 

1555: Whether $ n_s > 1 $ or $ n_s < 1 $ for a given set of parameters

1556: $ \Lambda,  \; \mu^2 , \; g $ and the initial conditions

1557: is a {\bf dynamical} question that can only be answered after

1558: evolving the fields till the end of inflation according to

1559: eqs.(\ref{ecmovh}). As we see in eq.(\ref{nshib}) the question is whether twice

1560: $ \chi^2 $ at horizon exit is larger or smaller than $ \Lambda $. Recall that

1561: horizon exit happens about $50$ efolds before the end of inflation and that

1562: $ \chi^2 $ monotonically decreases during inflation. Even if initially

1563: $ 2 \; \chi^2 > \Lambda $, it can be very well that 

1564: $ 2 \; \chi^2 < \Lambda $ at horizon exit. 

1565: This depends on  how many total efolds 

1566: $ N_T \geq 60 $ we have; since horizon exit happens $ (N_T - 50) \geq 10 $ 

1567: efolds after the beginning of inflation, the larger is  $ (N_T - 50) $

1568: the smaller can be $ \chi^2 $ at horizon exit. 

1569: 

1570: We vary the parameters $ \Lambda,  \; \mu^2 $ and the initial conditions

1571: always keeping the total number of efolds $ N_T $ during inflation larger

1572: or equal to $60$. We keep $ g^2 = \frac14 $ since this parameter is less

1573: relevant than the others. We explored the parameters region where $ r < 0.2 $

1574: and $ 0.95 < n_s < 1.15 $. 

1575: 

1576: \subsection{Red tilted and blue tilted regimes in Hybrid Inflation}

1577: 

1578: Extended numerical investigation showed that there exists a {\bf critical 

1579: value} of $ \mu^2 , \; \mu^2_{crit} \simeq 0.13 \;  \Lambda  $ such that 

1580: $ n_s > 1 $  provided $ \mu^2 < \mu^2_{crit} $.

1581: 

1582: \medskip

1583: 

1584: For $ \mu^2 > \mu^2_{crit} $ we find both regimes,  $ n_s > 1 $ {\bf and} 

1585: $ n_s < 1 $. This property is valid for all initial values of the inflaton

1586: compatible with the restrictions $ N_T \geq 60 $ and one-inflaton fluctuations.

1587: Otherwise, if the field $ \sigma $ is relevant at horizon exit we should also

1588: include its contribution to the density fluctuations. Such calculation

1589: is beyond the scope of the present work where we concentrate on single inflaton

1590: fluctuations.

1591: 

1592: The larger is $ \mu^2 $, the earlier inflation ends, the earlier horizon

1593: exit happens and the large is $ \chi $ at horizon exit. 

1594: That is, increasing $ \mu^2 $ decreases $ n_s $. This explains why we 

1595: necessarily find $ n_s < 1 $ for $ \mu^2 > \mu^2_{crit} $.

1596: 

1597: \medskip

1598: 

1599: For $ \mu^2 > \mu^2_{crit} $ we find that $ n_s > 1 $ for

1600: $ \chi(0) >  \chi(0)_{crit} $. That is, increasing $ \chi(0) $, increases 

1601: $ n_s $. This is so because the larger is $ \chi(0) $, the larger is $ N_T $ since 

1602: $ N_T \sim  \chi(0)^2 $  [see eqs.(\ref{ecmslo}) and (\ref{Nhib})]. 

1603: Then, the larger is $ (N_T - 50) $ the smaller is $ \chi $ at horizon exit 

1604: and the larger is $ n_s $. 

1605: 

1606: \medskip

1607: 

1608: In all cases, (both $ \mu^2 > \mu^2_{crit} $ and  $ \mu^2 < \mu^2_{crit} $)

1609: for $ \Lambda \to \infty $ we always find 

1610: $ n_s \to 1, \; r \to 0 $ and $ \frac{d n_s}{d \ln k}  \to 0 $.

1611: 

1612: \medskip

1613: 

1614: Figs. \ref{nsr005A}-\ref{nsrunA} show the observables 

1615: $ n_s , \; r $ and the running index 

1616: $ d n_s /d \ln k $ as  functions of $ \Lambda $ and $ n_s $ for 

1617: $  \mu^2 = 0.05 \; \Lambda , \;  \mu^2 = 0.13 \; \Lambda $ and 

1618: $ \mu^2 = 1.7  \;  \Lambda $. A complete picture for 

1619: hybrid inflation emerges covering {\bf two} different, blue tilted and 

1620: red tilted,  regimes. We find that for all the observables, the shape of 

1621: the curves depends crucially on the mass parameter $ {\hat \mu}^2 $ of the 

1622: $\sigma$ field and the (rescaled) initial amplitude 

1623: $ {\hat \chi}(0) $ of the inflaton field.

1624: 

1625: We find three regimes according to the value of $ {\hat \mu}^2 $:

1626: 

1627: \begin{itemize}

1628: 

1629: \item $ {\hat \mu}^2 < 0.075 $. Here we always have $ n_s > 1 $ and $ r $ 

1630: has one maximun as a function of $ n_s $ (or $ \Lambda $). 

1631:  $ n_s $ monotonically decreases with $ \Lambda $.

1632: Figs. \ref{nsr005A} and \ref{nsr005B} show $ r $ vs. $ n_s $ for

1633: $  \mu^2 = 0.05 \; \Lambda $ and various values of $ {\hat \chi}(0) $.

1634: $ r $ displays a maximun as a function of $ n_s $.

1635: In addition, $ r $ grows with  $  {\hat\chi}(0) $ for  $  {\hat\chi}(0) < 0.5 $

1636: while it decreases with  $ {\hat\chi}(0) $ for  $  {\hat\chi}(0) > 0.5 $.

1637: The running  $ \frac{d n_s}{d \ln k} $ behaviour is qualitatively

1638: similar to the behaviour of $ r $ above described. 

1639: The running $ \frac{d n_s}{d \ln k} $ is here positive and grows when 

1640: $ n_s $ grows.

1641: 

1642: \item $ 0.075 < {\hat \mu}^2 < {\hat \mu}^2_{crit} \simeq 0.13 $. 

1643: Here we always have $ n_s > 1 $ and $ r $ monotonically grows with $ \Lambda $.

1644: $ n_s $ monotonically decreases with $ \Lambda $ for $ {\hat \mu}^2 < 0.1 $

1645: while it exhibits a maximun as a function of $ \Lambda $ for  $ {\hat \mu}^2 

1646: > 0.1 $. Figs. \ref{nsrh} and \ref{nsrunh} depict  $ r $ and

1647: the running $ \frac{d n_s}{d \ln k} $  vs. $ n_s $, respectively,

1648: for $ {\hat \mu}^2 = 0.13 \simeq {\hat \mu}^2_{crit} $. 

1649: We see that the running  $ \frac{d n_s}{d \ln k} $ behaviour 

1650: is qualitatively similar to the one of $ r $.

1651: 

1652: \item $ {\hat \mu}^2 >  0.13 \simeq {\hat \mu}^2_{crit} $.

1653: Here $ n_s > 1 $ for  $ {\hat \chi}(0) > {\hat \chi}(0)_{crit} $

1654: and  $ n_s < 1 $ for  $ {\hat \chi}(0) < {\hat \chi}(0)_{crit} $.

1655: 

1656: The value of $ {\hat \chi}(0)_{crit} $ grows with $ {\hat \mu}^2 $:

1657: for $ {\hat \mu}^2 = 0.5 $,  we find  $ {\hat \chi}(0)_{crit} = 2.7$

1658: and for $ {\hat \mu}^2 = 1.7 $, we find   $ {\hat \chi}(0)_{crit} = 5.8$.

1659: 

1660: For $ {\hat \chi}(0) < {\hat \chi}(0)_{crit} , \; 

1661: n_s $ monotonically increases with $ \Lambda $ with values $ n_s < 1 $. 

1662: For $ {\hat \chi}(0) > {\hat \chi}(0)_{crit} , \; n_s $ shows an absolute 

1663: {\bf maximum}, which is always $ n_{s~max} > 1 $. The highest $n_s$ values 

1664: concentrate and narrow in the small $ \Lambda $ region. 

1665: It must be noticed that for each curve, [each $ {\hat \chi}(0) $],  

1666: $ n_s $ can take values $ n_s >1 $ {\bf and}

1667: $ n_s < 1 $: even if $ n_{s~max} > 1 , \; n_s $ can be below unit in the two 

1668: sides of the curve, [see figs. \ref{nsY} and \ref{nsA}].

1669: 

1670: The value $ n_s = 1 $,  is reached  asymptotically for large $\Lambda$ from 

1671: $ n_s < 1 $ for both $ {\hat \chi}(0) < {\hat \chi}(0)_{crit} $ and  

1672: $ {\hat \chi}(0) > {\hat \chi}(0)_{crit} $. In addition, the value $ n_s = 1 $

1673: with $ 0.2 > r > 0.04 $ is found for a variety of values of $ \Lambda $

1674: and  $ {\hat \chi}(0) > {\hat \chi}(0)_{crit} $ as we see from figs. \ref{nsA}, 

1675: \ref{rA} and \ref{nsrA}.

1676: 

1677: \end{itemize}

1678: 

1679: \subsection{The ratio $r$ in Hybrid Inflation}

1680: 

1681: The ratio $r$ in figs. \ref{rY} and \ref{rA} exhibits an  

1682: {\bf oscillatory} pattern and two different regimes:

1683: 

1684: 

1685: For  $ {\hat \chi}(0) < {\hat \chi}(0)_{crit} $, (fig. \ref{rY}) 

1686: $ r $ decreases monotonically 

1687: reaching very small values for large $\Lambda$. For  

1688: $ {\hat \chi}(0) < {\hat \chi}(0)_{crit} $, $ r $ 

1689: does not feature any oscillation. 

1690: 

1691: For $ {\hat \chi}(0) > {\hat \chi}(0)_{crit} , \; r $ decreases 

1692: with $ \Lambda $, (fig. \ref{rA}) $ r $ has an absolute {\bf minimum} 

1693: $ r_{min} $ and then grows till a {\bf maximum}, $ r_{max} $. The oscillations 

1694: show up and concentrate with growing amplitude for small 

1695: $ \Lambda $ for high $ {\hat \chi}(0) $,  $ r_{min} $ and  $ r_{max} $ 

1696: shift towards the 

1697: smaller $\Lambda$ with increasing $ {\hat \chi}(0) $;  $r_{min}$ decreases, and

1698:  $r_{max}$ increases, for increasing $  {\hat \chi}(0)  $.

1699: The convexity of the curve for small $ \Lambda $ increases

1700: for decreasing $ {\hat \chi}(0) $.

1701: 

1702: For $ {\hat \chi}(0) > {\hat \chi}(0)_{crit} $,

1703: each curve [each ${\hat \chi}(0)$] shows for $ r $ a 

1704:  {\bf oscillatory} behavior with three clear parts: (1) the asymptotic part of 

1705: monotonically decreasing $ r $ for large $ \Lambda $ at the right of 

1706: $r_{max}$; (2) the increasing part at the left of $r_{max}$; (3) the sharp 

1707: decreasing part for small $\Lambda$ at the left of $r_{min}$.

1708: In the minima, $r_{min}$ can be extremely small for small 

1709: $\Lambda$, which is a  {\bf new feature} in hybrid inflation.

1710: 

1711: In the asymptotic regime of large $\Lambda$, $r$ does not feature 

1712: any oscillation. All curves [for all ${\hat \chi}(0) $] coalesce into $ r = 0 $

1713: for $ \Lambda \to \infty $.

1714: 

1715: There are three distinct regimes: small $\Lambda$,  intermediate

1716: $\Lambda$ and large $\Lambda$. 

1717: The new oscillatory behavior for high and intermediate 

1718: $ {\hat \chi}(0) > {\hat \chi}(0)_{crit} $

1719: is in the region of small and intermediate 

1720: $\Lambda$. The monotonically decreasing behavior for low 

1721: ${\hat \chi}(0) < {\hat \chi}(0)_{crit} $ 

1722: is in the asymptotic region of large $\Lambda$.

1723: 

1724: \medskip

1725: 

1726: The highest values of $ r $ appear for small $\Lambda$ whatever be the 

1727: hybrid regime; for such high values of $ r $ both regimes 

1728: $ {\hat \chi}(0) < {\hat \chi}(0)_{crit} $ 

1729: and $ {\hat \chi}(0) > {\hat \chi}(0)_{crit} $

1730: superpose. From such high values, $r$ decreases  sharply till its minimun 

1731: $r_{min}$ in the small $\Lambda$ region for 

1732: ${\hat \chi}(0) > {\hat \chi}(0)_{crit}$;

1733: or $r$ decreases monotonically reaching asymptotically 

1734: the large $ \Lambda $ regime for ${\hat \chi}(0) < {\hat \chi}(0)_{crit}$.

1735: The low values of $ r $ for low $ \Lambda $ are a {\bf totally  new} 

1736: feature in hybrid inflation.

1737: 

1738: \subsection{The running index $ d n_s /d \ln k $ in Hybrid Inflation}

1739: 

1740: The curves of the running index $ d n_s /d \ln k $ figs. \ref{nsrunh},

1741: \ref{nsrunY} and \ref{nsrunA} show {\bf new} features in two different regimes.

1742: 

1743: \medskip

1744: 

1745: For $ {\hat \chi}(0) > {\hat \chi}(0)_{crit} $, the running index  

1746: $ d n_s /d \ln k $ shows a similar shape as $r$ and is essentially 

1747: positive. It oscillates with at least one {\bf maximum} 

1748: and one or two {\bf minima} and three different components: (1) the 

1749: asymptotic part of monotonically decreasing running with increasing 

1750: $\Lambda$, at the right of the maximum, going 

1751: to zero in this regime;  (2) the increasing running with $\Lambda$, 

1752: which is a new feature in hybrid inflation, and (3) the sharp 

1753: decreasing of the running till its minimum value for small $\Lambda$. 

1754: The highest running appears

1755: for small $\Lambda$ as in the known hybrid regime. The lower running 

1756: values for small $\Lambda$, as well as the  {\bf oscillations} for small and 

1757: intermediate $\Lambda$, are totally {\bf new}. \\

1758: 

1759: For low  $ {\hat \chi}(0) < {\hat \chi}(0)_{crit} $ the running index

1760: does not exhibit any oscillation.

1761: Both for $ {\hat \chi}(0) < {\hat \chi}(0)_{crit} $ and  

1762: ${\hat \chi}(0) > {\hat \chi}(0)_{crit}$

1763: $ d n_s /d \ln k $ grows with $ \Lambda $ till it reaches its maximun

1764: and then decreases monotonically with $ \Lambda $.  $ d n_s /d \ln k $

1765: vanishes asymptotically for large $\Lambda$ without any oscillation. 

1766: This is a totally {\bf new} feature for hybrid  inflation. 

1767: 

1768: \medskip

1769: 

1770: Thus, hybrid inflation describes both $ d n_s /d \ln k >0$ and 

1771: $ d n_s /d \ln k <0$. It must be noticed that $ d n_s /d \ln k < 0$ can reach very low 

1772: values for {\bf small} $\Lambda$ which is a  {\bf totally new} feature 

1773: in hybrid inflation.

1774: 

1775: \medskip

1776: 

1777: In summary, the {\bf new} features for the running 

1778: index in hybrid inflation are: both positive and negative running, increasing 

1779: of the running with $\Lambda$, transition from positive to negative running 

1780: passing through zero running when $\Lambda$ grows.

1781:  

1782: \subsection{$r$ vs. $n_s$. Confrontation of Hybrid Inflation to the three

1783: years WMAP data}

1784: 

1785: $r$ vs. $n_s$ in figs. \ref{nsrY}-\ref{nsrA} depicts a oscillatory behavior 

1786: clearly showing the two regimes: ${\hat \chi}(0) > {\hat \chi}(0)_{crit}$ 

1787: corresponding  mainly to $ n_s > 1$, although it also covers a small portion 

1788: of $n_s < 1$, and ${\hat \chi}(0) < {\hat \chi}(0)_{crit}$ for which $n_s$ 

1789: is entirely red tilted. \\

1790: 

1791: \medskip

1792: 

1793: All curves end [${\hat \chi}(0) < {\hat \chi}(0)_{crit}$], or start 

1794: [${\hat \chi}(0) > {\hat \chi}(0)_{crit}$], 

1795: at $n_s = 1$, which is the {\bf fixed} point for all ${\hat \chi}(0)$,

1796: with three different behaviors: \\

1797: 

1798: (1) the sharp decreasing of $r$ in the range $n_s <1$, approaching $n_s =1$ 

1799: as the end point, this is for ${\hat \chi}(0) < {\hat \chi}(0)_{crit}$, 

1800: in which $r$ can take high values for small $n_s$, ($n_s$ near $0.95$). \\

1801: 

1802: (2) the monotonically decreasing of $r$ with $n_s$ at the right of $r_{max}$, 

1803: for  ${\hat \chi}(0) > {\hat \chi}(0)_{crit}$, in 

1804: which $r$ vanishes asymptotically for `high' $n_s$ , $(n_s > 1.07)$.\\

1805: 

1806: (3) The {\bf new} hybrid behavior for high ${\hat \chi}(0) > 

1807: {\hat \chi}(0)_{crit}$ in between 

1808: the above two regimes, in which $r$ shows a maximum and a minimum between 

1809: two sharp decreasing and increasing `arms', lying at $n_s < 1$ and $n_s >1$ 

1810: respectively. 

1811:  $r_{min}$ decreases and $r_{max}$ increases as increasing ${\hat \chi}(0) $. 

1812: All $r_{max}$ lie in the red tilted regime $n_s<1$. All $r_{min}$ 

1813: lie in the blue tilted regime $n_s >1$.

1814: 

1815: \medskip

1816: 

1817: For ${\hat \chi}(0) > {\hat \chi}(0)_{crit}$, all curves go towards 

1818: $ n_s=1, \; r=0 $. Most of each curve lies in the $ n_s >1 $ 

1819: region, $ r $ monotonically decreases with $n_s$ in the range $ n_s <1 $

1820: from  the maximun $r_{max}$  going  towards $ r = 0 $ for $ n_s \to 1 $.

1821: For $ {\hat \chi}(0) < {\hat \chi}(0)_{crit} $ the curves pill up in the 

1822: $ n_s<1 $ region with almost the same slope, $r$ sharply decreases in this 

1823: region. 

1824: 

1825: The curves for ${\hat \chi}(0) < {\hat \chi}(0)_{crit}$ pill up in the 

1826: $ n_s<1 $ region with almost the same slope, 

1827: $r$ sharply decreases in this region. 

1828: 

1829: \begin{figure}[p]

1830: \begin{turn}{-90}

1831: \centering

1832: \psfrag{"ns21.dat"}{${\hat \chi}(0)=5.7$}

1833: \psfrag{"ns21a.dat"}{${\hat \chi}(0)=5.9$}

1834: \psfrag{"ns22.dat"}{${\hat \chi}(0)=6.1$}

1835: \psfrag{"ns23.dat"}{${\hat \chi}(0)=6.5$}

1836: \psfrag{"ns24.dat"}{${\hat \chi}(0)=6.9$}

1837: \psfrag{"ns25.dat"}{${\hat \chi}(0)=7.3$}

1838: \psfrag{"ns26.dat"}{${\hat \chi}(0)=7.7$}

1839: \psfrag{"ns27.dat"}{${\hat \chi}(0)=8.1$}

1840: \psfrag{"ns28.dat"}{${\hat \chi}(0)=8.7$}

1841: \psfrag{"ns30.dat"}{${\hat \chi}(0)=10.7$}

1842: \includegraphics[width=10cm,height=14cm]{nsAR}

1843: \end{turn}

1844: \caption{Hybrid inflation. The index $ n_s $ vs. $ \Lambda $ for $ \mu^2 =

1845: 1.7 \; \Lambda  > \mu^2_{crit} , \; g^2 = \frac 14 $  and 

1846: $ \chi(0)_{crit} \simeq 5.7 \; \sqrt{\Lambda}\leq \chi(0) \leq 10.7 

1847: \; \sqrt{\Lambda} $. Notice that here we have both $ n_s > 1 $ and  

1848: $ n_s < 1 $ depending on the values of $ \Lambda $ and $ \hat \chi(0) $.

1849: All curves reach asymptotically $ n_s = 1 $ for $ \Lambda \to \infty $.}

1850: \label{nsA}

1851: \end{figure}

1852: 

1853: \begin{figure}[p]

1854: \begin{turn}{-90}

1855: \centering

1856: \psfrag{"r7.dat"}{${\hat \chi}(0)=2.7$}

1857: \psfrag{"r10.dat"}{${\hat \chi}(0)=3.3$}

1858: \psfrag{"r12.dat"}{${\hat \chi}(0)=3.7$}

1859: \psfrag{"r14.dat"}{${\hat \chi}(0)=4.1$}

1860: \psfrag{"r16.dat"}{${\hat \chi}(0)=4.5$}

1861: \psfrag{"r17.dat"}{${\hat \chi}(0)=4.7$}

1862: \psfrag{"r19c.dat"}{${\hat \chi}(0)=5.1$}

1863: \psfrag{"r20c.dat"}{${\hat \chi}(0)=5.3$}

1864: \psfrag{"r21c.dat"}{${\hat \chi}(0)=5.7$}

1865: \includegraphics[width=10cm,height=14cm]{rZ}

1866: \end{turn}

1867: \caption{Hybrid inflation. The ratio $ r $ vs. $ \Lambda $ for $ \mu^2 =

1868: 1.7 \; \Lambda > \mu^2_{crit} , \; g^2 = \frac 14 $ and $ 2.7 \; 

1869: \sqrt{\Lambda}\leq \chi(0) \leq 5.7 \; \sqrt{\Lambda}\simeq \chi(0)_{crit} $. 

1870: $ r $ decreases monotonically with $ \Lambda $ in this regime 

1871: $ \chi(0) < \chi(0)_{crit} $ and asymptotically vanishes for $ \Lambda \to \infty $.}

1872: \label{rY}

1873: \end{figure}

1874: 

1875: \begin{figure}[p]

1876: \begin{turn}{-90}

1877: \centering

1878: \psfrag{"r21.dat"}{${\hat \chi}(0)=5.7$}

1879: \psfrag{"r22.dat"}{${\hat \chi}(0)=6.1$}

1880: \psfrag{"r23.dat"}{${\hat \chi}(0)=6.5$}

1881: \psfrag{"r24.dat"}{${\hat \chi}(0)=6.9$}

1882: \psfrag{"r25.dat"}{${\hat \chi}(0)=7.3$}

1883: \psfrag{"r26.dat"}{${\hat \chi}(0)=7.7$}

1884: \psfrag{"r27c.dat"}{${\hat \chi}(0)=8.1$}

1885: \psfrag{"r28.dat"}{${\hat \chi}(0)=8.7$}

1886: \psfrag{"r30.dat"}{${\hat \chi}(0)=10.7$}

1887: \includegraphics[width=10cm,height=14cm]{rAR}

1888: \end{turn}

1889: \caption{Hybrid inflation. The ratio $ r $ vs. $ \Lambda $ for $ \mu^2 =

1890: 1.7 \; \Lambda  > \mu^2_{crit} , \; g^2 = \frac14 $

1891: and $ \chi(0)_{crit} \simeq  5.7 \; \sqrt{\Lambda}\leq \chi(0)  

1892: \leq 10.7 \; \sqrt{\Lambda} . \;  r $ shows an oscillatory pattern in this

1893: regime $ \chi(0)  \geq \chi(0)_{crit} $ and asymptotically vanishes for 

1894: $ \Lambda \to \infty $ with no oscillations. The oscillations show up and concentrate

1895: with increasing amplitude for small $ \Lambda $.}

1896: \label{rA}

1897: \end{figure}

1898: 

1899: \begin{figure}[p]

1900: \begin{turn}{-90}

1901: \centering

1902: \psfrag{"nsr7.dat"}{${\hat \chi}(0)=2.7$}

1903: \psfrag{"nsr10.dat"}{${\hat \chi}(0)=3.3$}

1904: \psfrag{"nsr12.dat"}{${\hat \chi}(0)=3.7$}

1905: \psfrag{"nsr14.dat"}{${\hat \chi}(0)=4.1$}

1906: \psfrag{"nsr16.dat"}{${\hat \chi}(0)=4.5$}

1907: \psfrag{"nsr17.dat"}{${\hat \chi}(0)=4.7$}

1908: \psfrag{"nsr19.dat"}{${\hat \chi}(0)=5.1$}

1909: \psfrag{"nsr20.dat"}{${\hat \chi}(0)=5.3$}

1910: \psfrag{"nsr21.dat"}{${\hat \chi}(0)=5.7$}

1911: \includegraphics[width=10cm,height=14cm]{nsrZ}

1912: \end{turn}

1913: \caption{Hybrid inflation. The ratio $ r $ vs. $ n_s $ for $ \mu^2 =

1914: 1.7 \; \Lambda > \mu^2_{crit} , \; g^2 = \frac 14 $ and $ 2.7 \; 

1915: \sqrt{\Lambda}\leq \chi(0) \leq 5.7 \; \sqrt{\Lambda} \simeq \chi(0)_{crit} $. 

1916: Notice that here $ n_s < 1 $ for all values of $ \Lambda $ and this range 

1917: of $ \hat  \chi(0) $. We see that $ 0.2 > r > 0.14 $ for the interval

1918: $ 0.952 <  n_s < 0.97$.}

1919: \label{nsrY}

1920: \end{figure}

1921: 

1922: \begin{figure}[p]

1923: \begin{turn}{-90}

1924: \centering

1925: \psfrag{"nsr21.dat"}{${\hat \chi}(0)=5.7$}

1926: \psfrag{"nsr22.dat"}{${\hat \chi}(0)=6.1$}

1927: \psfrag{"nsr23.dat"}{${\hat \chi}(0)=6.5$}

1928: \psfrag{"nsr24.dat"}{${\hat \chi}(0)=6.9$}

1929: \psfrag{"nsr25.dat"}{${\hat \chi}(0)=7.3$}

1930: \psfrag{"nsr26.dat"}{${\hat \chi}(0)=7.7$}

1931: \psfrag{"nsr27.dat"}{${\hat \chi}(0)=8.1$}

1932: \psfrag{"nsr28.dat"}{${\hat \chi}(0)=8.7$}

1933: \psfrag{"nsr30.dat"}{${\hat \chi}(0)=10.7$}

1934: \psfrag{"nsr33.dat"}{${\hat \chi}(0)=13.7$}

1935: \includegraphics[width=10cm,height=14cm]{nsrAR}

1936: \end{turn}

1937: \caption{Hybrid inflation. The ratio $ r $ vs. $ n_s $ for $ \mu^2 =

1938: 1.7 \; \Lambda  > \mu^2_{crit} , \; g^2 = \frac 14 $ and 

1939: $ \chi(0)_{crit}  \simeq 5.7 \; \sqrt{\Lambda}\leq \chi(0)  \leq 13.7 

1940: \; \sqrt{\Lambda} $. Notice that we have here both $ n_s > 1 $ and  

1941: $ n_s < 1 $ depending on the values of $ \Lambda $ and $ \hat \chi(0) $.

1942: All curves end [$ {\hat \chi(0)} <  {\hat \chi(0)}_{crit} $] or start

1943:  [$ {\hat \chi(0)} > {\hat \chi(0)}_{crit} $] at $ n_s = 1 $ which is the fixed 

1944: point for all values of $ {\hat \chi(0)} $.}

1945: \label{nsrA}

1946: \end{figure}

1947: 

1948: \begin{figure}[p]

1949: \begin{turn}{-90}

1950: \centering

1951: \psfrag{"nsrun7.dat"}{${\hat \chi}(0)=2.7$}

1952: \psfrag{"nsrun10.dat"}{${\hat \chi}(0)=3.3$}

1953: \psfrag{"nsrun12.dat"}{${\hat \chi}(0)=3.7$}

1954: \psfrag{"nsrun14.dat"}{${\hat \chi}(0)=4.1$}

1955: \psfrag{"nsrun16.dat"}{${\hat \chi}(0)=4.5$}

1956: \psfrag{"nsrun17.dat"}{${\hat \chi}(0)=4.7$}

1957: \psfrag{"nsrun19.dat"}{${\hat \chi}(0)=5.1$}

1958: \psfrag{"nsrun20.dat"}{${\hat \chi}(0)=5.3$}

1959: \psfrag{"nsrun21.dat"}{${\hat \chi}(0)=5.7$}

1960: \includegraphics[width=10cm,height=14cm]{nsrunZ}

1961: \end{turn}

1962: \caption{Hybrid inflation. The running $ d n_s/ d \ln k  $ vs. $ n_s $ for 

1963: $ \mu^2 = 1.7 \; \Lambda  > \mu^2_{crit}, \; g^2 = \frac 14 $ and 

1964: $ 1.9 \; \sqrt{\Lambda}\leq \chi(0)  \leq 5.7 \; \sqrt{\Lambda} \simeq 

1965: \chi(0)_{crit} $. Notice that we have here both positive and negative running

1966: $  d n_s/ d \ln k $ depending on the values of $ \Lambda $ and $ \hat \chi(0) $.

1967: Most values of the running are negative in this regime. For $ 0.952 <  n_s < 0.97 $,

1968: we have $ -0.001 < d n_s /d \ln k < 0 $.} \label{nsrunY}

1969: \end{figure}

1970: 

1971: \begin{figure}[p]

1972: \begin{turn}{-90}

1973: \centering

1974: \psfrag{"nsrun21.dat"}{${\hat \chi}(0)=5.7$}

1975: \psfrag{"nsrun22.dat"}{${\hat \chi}(0)=6.1$}

1976: \psfrag{"nsrun23.dat"}{${\hat \chi}(0)=6.5$}

1977: \psfrag{"nsrun24.dat"}{${\hat \chi}(0)=6.9$}

1978: \psfrag{"nsrun25.dat"}{${\hat \chi}(0)=7.3$}

1979: \psfrag{"nsrun26.dat"}{${\hat \chi}(0)=7.7$}

1980: \psfrag{"nsrun27.dat"}{${\hat \chi}(0)=8.1$}

1981: \psfrag{"nsrun28.dat"}{${\hat \chi}(0)=8.7$}

1982: \psfrag{"nsrun30.dat"}{${\hat \chi}(0)=10.7$}

1983: \psfrag{"nsrun33.dat"}{${\hat \chi}(0)=13.7$}

1984: \includegraphics[width=10cm,height=14cm]{nsrunAR}

1985: \end{turn}

1986: \caption{Hybrid inflation. The running $  d n_s/ d \ln k $ vs. $ n_s $ 

1987: for $ \mu^2 = 1.7 \; \Lambda  > \mu^2_{crit}, \; g^2 = \frac 14 $ 

1988: and $ \chi(0)_{crit}  \simeq 5.7 \; \sqrt{\Lambda} \leq \chi(0)  

1989: \leq 13.7 \; \sqrt{\Lambda} $. Notice that we have here both positive and 

1990: negative running $  d n_s/ d \ln k $ depending on the values of 

1991: $ \Lambda $ and $ \hat \chi(0) $.}

1992: \label{nsrunA}

1993: \end{figure}

1994: 

1995: \begin{figure}[p]

1996: \begin{turn}{-90}

1997: \centering

1998: \psfrag{"nsr93h20"}{border for new inflation}

1999: \psfrag{"nsr93h0"}{border for new inflation}

2000: \psfrag{"nsr33"}{border for hybrid inflation}

2001: \includegraphics[width=10cm,height=14cm]{borde93}

2002: \end{turn}

2003: \caption{Regions described in the $ (r, n_s)$-plane by new and single field

2004: hybrid inflation for $ n_s < 1 $. The hybrid inflation border corresponds to 

2005: $ \mu^2 = 1.7 \; \Lambda $. For $ n_s > 1 $, all values of $ (r, n_s) $ 

2006: can be described by hybrid inflation (at least for $ r < 0.2, \; n_s < 1.15 $).

2007: The excluded region cannot be described by single field inflation 

2008: (neither hybrid inflation, nor new inflation). Two or more fields inflation

2009: could describe such regions.}

2010: \label{bordenuevo}

2011: \end{figure}

2012: 

2013: All the blue tilted values of $ (n_s, r) $ in the domain $ 1 < n_s < 1.15 , 

2014: \; 0 < r < 0.2 $ are realized by hybrid inflation. 

2015: 

2016: \medskip

2017: 

2018: The red tilted regime in hybrid inflation can only be realized

2019: for $  \mu^2 >  \mu^2_{crit} \simeq 0.13 \; \Lambda $.

2020: Moreover, the possible values of  $ (n_s, r) $ are in the upper-right quadrant 

2021: as shown in fig. \ref{bordenuevo}.

2022: 

2023: \medskip

2024: 

2025: We see at the light of the three years WMAP data ref. \cite{WMAP3}

2026: that hybrid inflation in the blue tilted regime $ n_s > 1 $ is ruled out 

2027: [eq.(\ref{nswmap})] \cite{WMAP3}. That is, hybrid inflation in the regime 

2028: $ {\hat\mu^2} < {\hat\mu^2_{crit}} $ is ruled out as well as hybrid inflation

2029: in the regime $ {\hat\mu^2} >  {\hat\mu^2_{crit}} $ with 

2030: $ {\hat \chi}(0) > {\hat \chi}(0)_{crit} $.

2031: 

2032: \medskip

2033: 

2034: Hybrid inflation in the red tilted regime $ {\hat \mu^2} > {\hat\mu^2_{crit}} $

2035: and $ {\hat \chi}(0) < {\hat \chi}(0)_{crit} $

2036: fulfills the three years WMAP value for $ n_s $ eq.(\ref{nswmap}), as

2037: well as the bound on the ratio $ r $ eq.(\ref{rwmap}).

2038: We can read from fig. \ref{nsrY} that 

2039: $$ 

2040: 0.2 > r < 0.14  \quad {\rm for} \quad  0.952 <  n_s < 0.97 \; .

2041: $$

2042: In addition, we find in fig. \ref{nsrunY}

2043: negative values for the running in this range, that is:

2044: $$ 

2045: -0.001 < d n_s /d \ln k < 0 \quad {\rm for} \quad 0.952 <  n_s < 0.97 \; .

2046: $$

2047: There are clearly two regions which are not covered by hybrid inflation 

2048: with only one inflaton field, neither by new inflation as shown 

2049: in fig. \ref{bordenuevo}.

2050: 

2051: \bigskip

2052: 

2053: Simple single field  inflation models have been recently studied within a 

2054: numerical approach \cite{BST}. In ref. \cite{PE} cosmological

2055: data are fitted with the help of a Markov Chain Monte Carlo analysis.

2056:  

2057: \medskip

2058: 

2059: There is an interplay between the bounds of neutrino masses and the sign

2060: of $ (n_s - 1) $. A non-zero neutrino mass decreases the power in the 

2061: small scales (large wavenumbers $k$). The same happens if $ n_s $ becomes 

2062: smaller than unit.

2063: Therefore, {\it if} $  n_s < 1 $, the power spectrum permits more stringent

2064: tests of the neutrino masses \cite{fuku}. The effect of neutrino masses

2065: for small scales $ n_s > 1 $ can be cancelled by a spectral index $ n_s > 1 $.

2066: 

2067: \bigskip

2068: 

2069: {\bf Acknowledgment:} we thank Daniel Boyanovsky for useful discussions.

2070: 

2071: \begin{thebibliography}{99}

2072: \bibitem{guth} D. Kazanas, ApJ 241, L59 (1980);

2073: A. Guth, Phys. Rev. \textbf{D23}, 347 (1981);

2074: K. Sato, MNRAS, {\bf 195}, 467 (1981).

2075: 

2076: \bibitem{mukyotr} V. F. Mukhanov , G. V. Chibisov, Soviet Phys.

2077: JETP Lett. \textbf{33}, 532 (1981). 

2078: S. W. Hawking, Phys. Lett. \textbf{B115}, 295

2079: (1982). A. H. Guth , S. Y. Pi, Phys. Rev. Lett.

2080: \textbf{49}, 1110 (1982). A. A. Starobinsky, Phys. Lett.

2081: \textbf{B117}, 175 (1982). J. M. Bardeen, P. J. Steinhardt , M. S.

2082: Turner, Phys. Rev. \textbf{D28}, 679 (1983). 

2083: V. F. Mukhanov, H. A. Feldman , R. H.

2084: Brandenberger, Phys. Rept. \textbf{215}, 203 (1992).

2085: 

2086: \bibitem{libros}

2087: P. Coles and F. Lucchin, {\em Cosmology}, John Wiley,

2088: Chichester, 1995. A. R. Liddle and D. H. Lyth,

2089: {\em Cosmological Inflation and Large Scale Structure}, Cambridge University

2090: Press, 2000. S. Dodelson, {\em Modern Cosmology}, Academic Press, 2003.

2091: D. H. Lyth , A. Riotto, Phys. Rept. \textbf{314}, 1 (1999).

2092: \bibitem{WMAP} C. L. Bennett \emph{et.al.} (WMAP collaboration),

2093: Ap. J. Suppl. \textbf{148}, 1 (2003).

2094: 

2095: A. Kogut  \emph{et.al.} (WMAP collaboration),

2096: Ap. J. Suppl. \textbf{148}, 161 (2003).

2097: 

2098: D. N. Spergel \emph{et. al. }(WMAP collaboration),

2099: Ap. J. Suppl. \textbf{148}, 175 (2003).

2100: 

2101: H. V. Peiris \emph{et.al.} (WMAP collaboration), Ap. J.

2102: Suppl.\textbf{148}, 213 (2003).

2103: 

2104: \bibitem{WMAP3}

2105: D. N. Spergel \emph{et. al. }(WMAP collaboration),

2106: astro-ph/0603449.

2107: 

2108:  L. Page, \emph{et. al. }(WMAP collaboration),

2109: astro-ph/0603450.

2110: 

2111: G. Hinshaw,\emph{et. al. }(WMAP collaboration),

2112: astro-ph/0603451.

2113: 

2114:  N. Jarosik, \emph{et. al. }(WMAP collaboration),

2115: astro-ph/0603452.

2116: 

2117: \bibitem{2dF} A. G. S\'anchez et al., Mon. Not. Roy. Astron. Soc. {\bf 366}, 

2118: 189 (2006).

2119: 

2120: \bibitem{SDSS} U. Seljak et al.,  Phys. Rev. D71, 103515 (2005).

2121: 

2122: \bibitem{Teg} M. Tegmark wt al.,  Phys. Rev. D69, 103501 (2004).

2123: 

2124: \bibitem{1sN} D. Boyanovsky, H. J. de Vega, N. G. S\'anchez,

2125: Phys. Rev. {\bf D 73}, 023008 (2006).

2126: 

2127: \bibitem{nos} D. Cirigliano,  H. J. de Vega, N. G. Sanchez,

2128: Phys. Rev. {\bf D 71}, 103518 (2005).

2129: 

2130: \bibitem{quir} L. D. Landau, E. M. Lifshits, `Physique Statistique',

2131: chapter 14, Mir Ellipses, Paris, 1994.

2132: H. Leutwyler, Ann. Phys. 235, 165 (1994), hep-ph/9409423.

2133: S. Weinberg, hep-ph/9412326 and `The Quantum Theory of Fields', vol.

2134: 2, Cambridge University Press, Cambridge, 2000.

2135: 

2136: \bibitem{BST} L. A. Boyle, P. J. Steinhardt, N. Turok, Phys. Rev. Lett., 

2137: {\bf 96}, 111301, (2006).

2138: 

2139: \bibitem{PE} H. V. Peiris, R. Easther, astro-ph/0603587.

2140: 

2141: \bibitem{fuku} M. Fukugita, hep-ph/0511068.

2142: 

2143: \bibitem{barrow} A. R. Liddle, P. Parsons , J. D. Barrow, Phys.

2144: Rev. \textbf{D50}, 7222 (1994).

2145: 

2146: \bibitem{hu} See for example,  W. Hu and  S. Dodelson,

2147: Ann. Rev. Astron. Ap. \textbf{40}, 171 (2002); J. Lidsey, A.

2148: Liddle, E. Kolb, E. Copeland, T. Barreiro and M. Abney, Rev. of

2149: Mod. Phys. {\bf 69}, 373, (1997). W. Hu, astro-ph/0402060.

2150: 

2151: \bibitem{lin} A. Linde,  Phys. Rev. D49, 748 (1994).

2152: J. Garc\'{\i}a Bellido,  A. Linde,  Phys. Rev. D57, 6075 (1998).

2153: E. J. Copeland, A. R. Liddle, D. H. Lyth, E. D. Stewart, 

2154: D. Wands, Phys. Rev. D49, 6410 (1994). 

2155: \end{thebibliography}

2156: 

2157: \end{document}