1: % correcciones finales a las pruebas de PRD incluidas
2: \documentclass[aps,superscriptaddress,tightenlines,nofootinbib]{revtex4}
3: \usepackage{graphicx,amssymb,amsbsy,bm,amsmath}
4: \usepackage{hyperref}
5: \newcommand{\be}{\begin{equation}}
6: \newcommand{\ee}{\end{equation}}
7: \newcommand{\bea}{\begin{eqnarray}}
8: \newcommand{\eea}{\end{eqnarray}}
9: \newcommand{\bk}{\ensuremath{\mathbf{k}}}
10: \newcommand{\vx}{\ensuremath{\vec{x}}}
11: \newcommand{\OChi}{\ensuremath{\chi_{c}}}
12: \newcommand{\vk}{\ensuremath{\vec{k}}}
13: \newcommand{\vq}{\ensuremath{\vec{q}}}
14: \newcommand{\bp}{\ensuremath{\mathbf{p}}}
15: \newcommand{\bq}{\ensuremath{\mathbf{q}}}
16: \newcommand{\ba}{\ensuremath{\mathbf{a}}}
17: \newcommand{\bnu}{\ensuremath{\bar{\nu}}}
18: \newcommand{\bhk}{\ensuremath{\widehat{\mathbf{k}}}}
19: \newcommand{\bhp}{\ensuremath{\widehat{\mathbf{p}}}}
20: \newcommand{\bhq}{\ensuremath{\widehat{\mathbf{q}}}}
21: \begin{document}
22: \title{New Inflation vs. Chaotic Inflation, higher degree potentials
23: and the Reconstruction Program in light of WMAP3}
24: \author{D. Boyanovsky}
25: \email{boyan@pitt.edu} \affiliation{Department of Physics and
26: Astronomy, University of Pittsburgh, Pittsburgh, Pennsylvania 15260,
27: USA} \affiliation{Observatoire de Paris, LERMA. Laboratoire
28: Associ\'e au CNRS UMR 8112.
29: \\61, Avenue de l'Observatoire, 75014 Paris, France.}
30: \affiliation{LPTHE, Universit\'e Pierre et Marie Curie (Paris VI) et
31: Denis Diderot (Paris VII), Laboratoire Associ\'e au CNRS UMR 7589,
32: Tour 24, 5\`eme. \'etage, 4, Place Jussieu, 75252 Paris, Cedex 05,
33: France}
34: \author{H. J. de Vega}
35: \email{devega@lpthe.jussieu.fr} \affiliation{LPTHE, Universit\'e
36: Pierre et Marie Curie (Paris VI) et Denis Diderot (Paris VII),
37: Laboratoire Associ\'e au CNRS UMR 7589, Tour 24, 5\`eme. \'etage, 4,
38: Place Jussieu, 75252 Paris, Cedex 05,
39: France}\affiliation{Observatoire de Paris, LERMA. Laboratoire
40: Associ\'e au CNRS UMR 8112.
41: \\61, Avenue de l'Observatoire, 75014 Paris, France.}
42: \affiliation{Department of Physics and Astronomy, University of
43: Pittsburgh, Pittsburgh, Pennsylvania 15260, USA}
44: \author{C. M. Ho} \email{cmho@phyast.pitt.edu}
45: \affiliation{Department of Physics and Astronomy, University of
46: Pittsburgh, Pittsburgh, Pennsylvania 15260, USA}
47: \author{N. G. Sanchez}
48: \email{Norma.Sanchez@obspm.fr} \affiliation{Observatoire de Paris,
49: LERMA. Laboratoire Associ\'e au CNRS UMR 8112.
50: \\61, Avenue de l'Observatoire, 75014 Paris, France.}
51: \date{\today}
52: \begin{abstract}
53: The CMB power spectra are studied for different \emph{families} of
54: single field new and chaotic inflation models in the effective field
55: theory approach to inflation. We implement a systematic expansion
56: in $ 1/N_e $ where $ N_e\sim 50 $ is the number of e-folds before
57: the end of inflation. We study the dependence of the observables ($
58: n_s, \; r $ and $ dn_s/d\ln k $) on the degree of the potential ($
59: 2 \, n $) and confront them to the WMAP3 and large scale structure
60: data: This shows in general that fourth degree potentials ($ n=2 $)
61: provide the best fit to the data; the window of consistency with the
62: WMAP3 and LSS data narrows for growing $ n $. New inflation
63: yields a good fit to the $ r $ and $ n_s $ data in a wide range of
64: field and parameter space. Small field inflation yields $ r<0.16 $
65: while large field inflation yields $ r>0.16 $ (for $ N_e=50 $). All
66: members of the new inflation family predict a small but negative
67: running $ -4 \, (n+1)\times 10^{-4} \leq dn_s/d\ln k \leq -2\times
68: 10^{-4} $. (The values of $ r, \; n_s, \; dn_s/d\ln k $ for arbitrary
69: $ N_e $ follow by a simple rescaling from the $ N_e=50 $ values).
70: A reconstruction program is carried out suggesting quite
71: generally that for $ n_s $ consistent with the WMAP3 and LSS data
72: and $ r<0.1 $ the {\bf symmetry breaking scale} for new inflation is
73: $ |\phi_0| \sim 10~M_{Pl} $ while the {\bf field scale} at Hubble
74: crossing is $ |\phi_{c}| \sim M_{Pl} $. The family of chaotic
75: models feature $ r \geq 0.16 $ (for $ N_e=50 $) and only a {\it
76: restricted subset} of chaotic models are consistent with the
77: combined WMAP3 bounds on $ r, \; n_s, \; dn_s/d\ln k $ with a
78: narrow window in field amplitude around $ |\phi_{c}| \sim 15~M_{Pl}
79: $. We conclude that a measurement of $ r<0.16 $ (for $ N_e =50 $)
80: distinctly rules out a large class of chaotic scenarios and favors
81: small field new inflationary models. As a general consequence, {\bf
82: new inflation} emerges more favoured than chaotic inflation.
83: \end{abstract}
84:
85: \pacs{98.80.Cq,05.10.Cc,11.10.-z}
86:
87: \maketitle
88:
89: \section{Introduction}
90: Inflation provides a simple and robust mechanism to solve several
91: outstanding problems of the standard Big Bang model
92: \cite{infla,libros} becoming a leading paradigm in cosmology.
93: Superhorizon quantum fluctuations amplified during inflation provide
94: an explanation of the origin of the temperature anisotropies in the
95: cosmic microwave background (CMB) and the seeds for large scale
96: structure formation\cite{fluc}, as well as of tensor perturbations
97: (primordial gravitational waves). Although there is a diversity of
98: inflationary models, most of them predict fairly generic features: a
99: gaussian, nearly scale invariant spectrum of (mostly) adiabatic
100: scalar and tensor primordial fluctuations \cite{fluc}. These
101: features provide an excellent fit to the highly precise data
102: provided by the Wilkinson Microwave Anisotropy Probe (WMAP)
103: \cite{WMAPa,peiris,WMAP3,WMAP3b} which begins to constrain
104: inflationary models.
105:
106: The combination of CMB \cite{peiris,WMAP3} and large scale structure
107: data \cite{SDSS,2dF} yield fairly tight constraints for the two
108: dimensional marginalized contours of the tensor to scalar ratio $r$
109: and the scalar index $n_s$. While $n_s=1$ was excluded at the $
110: 95\%CL $ in \cite{2dF} a most notable result that stems from the
111: analysis of WMAP3 data is a confirmation that
112: a scale invariant Harrison-Zeldovich
113: spectrum is excluded at the $3\sigma$ level \cite{WMAP3}. A
114: combination of data from WMAP3 and large scale surveys distinctly
115: favor $n_s <1$ \cite{sanchez}. These latest bounds on the index of
116: the power spectrum of scalar perturbations, and emerging bounds on
117: the ratio of tensor to scalar fluctuations $ r $ begin to offer the
118: possibility to discriminate different inflationary models. For
119: example, the third year WMAP data disfavors the predictions for
120: the scalar index and the tensor to scalar ratio from a monomial
121: inflationary potential $ \lambda \; \phi^4 $ showing them to lie
122: outside the $ 3\sigma $ contour, but the simple monomial $ m^2 \;
123: \phi^2/2 $ yields a good fit to the data \cite{WMAP3} and predicts a
124: tensor to scalar ratio $ r\sim 0.16 $ within the range of
125: forthcoming CMB observations.
126:
127: Current and future CMB observations in combination
128: with large scale structure surveys will yield tight constraints on
129: the inflationary models, this motivates the exploration of clear predictions
130: from the models and their confrontation with the data.
131:
132: In distinction with the approach followed in
133: \cite{peiris,WMAP3,hoki} or studies of specific
134: models\cite{salama}, or statistical analysis combined with WMAP3 and
135: LSS data\cite{kinkol,fipele}, we study the
136: predictions for the power spectra of scalar fluctuations and the
137: tensor to scalar ratio for \emph{families} of new and chaotic
138: inflationary models in the framework of the method presented in
139: ref.\cite{1sN}. This method relies on the effective field theory
140: approach combined with a systematic expansion in $ 1/N_e $ where $ N_e
141: \sim 50 $ is the number of e-folds before the end of inflation. The
142: family of inflationary models that we study is characterized by
143: effective field theories with potentials of the form
144: \bea
145: V(\phi) &=& V_0 -\frac12 \; m^2 \; \phi^2 + \frac{\lambda}{2 \, n}\;
146: \phi^{2 \, n} \quad , \quad {\rm broken ~ symmetry}\label{nuevito}\\
147: V(\phi) & = & \frac12 \; m^2 \; \phi^2 + \frac{\lambda}{2 \, n} \;
148: \phi^{2 \, n} \quad , \quad {\rm unbroken ~ symmetry} \; ,
149: \label{caoco}
150: \eea
151: with $ n=2,3,4\cdots $. For broken symmetry models
152: with potentials of the form (\ref{nuevito}) there are two distinct
153: regions: small and large field, corresponding to values of the
154: inflaton field smaller or larger than the symmetry breaking scale respectively.
155:
156: We implement the systematic expansion in $ 1/N_e
157: $ where $ N_e\sim 50 $ is the number of e-folds before the end of
158: inflation when wavelengths of cosmological relevance crossed the
159: Hubble radius\cite{1sN}. The $ 1/N_e $ expansion is a powerful and
160: systematic tool that allows to re-cast the slow roll hierarchy as
161: expansion in powers of $ 1/N_e $ \cite{1sN}. This expansion
162: allows us to implement a reconstruction program \cite{reco} which
163: yields the scale of the inflaton field when modes of cosmological
164: relevance today crossed the Hubble radius during inflation, and in
165: the case of new inflation models also yields the {\it scale of
166: symmetry breaking}.
167:
168: \medskip
169:
170: We study the dependence of the observables ($ n_s, \; r $ and $
171: dn_s/d\ln k $) on the degree of the inflaton potential ($ 2 \, n $)
172: for new and chaotic inflation and confront them with the WMAP3 data.
173: This study shows in general that fourth degree potentials ($ n=2 $)
174: provide the best fit to the data. We find that new inflation fits
175: the data on an appreciable wider range of the parameters while
176: chaotic inflation does this in a much narrow range. Therefore,
177: amongst the families of inflationary models studied, new inflation
178: emerges as a leading contender in comparison with chaotic inflation.
179: The present analysis confirms the statement that within the
180: framework of effective field theories with polynomial potentials,
181: new inflation is a preferred model reproducing the present data
182: \cite{ciri,pre,heclast}.
183:
184: \medskip
185:
186: {\bf Main results of this article:}
187:
188: \begin{itemize}
189:
190: \item{The region in inflaton field space which is consistent with the
191: marginalized WMAP3 data can be explored in an expansion in
192: $n_s-1+2/N_e$. }
193:
194: \item{ We find that the point $ n_s=1 -2/N_e= 0.96, \; r=8/N_e=0.16 $ which is in the region
195: allowed by the WMAP3 analysis\cite{WMAP3} belongs both to new inflation models
196: as a limiting point and to the simple chaotic inflation monomial, $ m^2 \, \phi^2/2 $.
197: (i) This point describes a
198: region in field and parameter space that separates small fields from
199: large fields, and (ii) is a degeneracy point for the family of
200: models describing both chaotic and new inflation. }
201: \item{ For all members $ n = 2, \; 3, \; 4, ...$ of the new
202: inflation family, the small field region yields $ r<0.16 $ while the
203: large field region yields $ r>0.16 $. All members of the new inflation family
204: predict a small but negative running:
205: $$
206: -4 \; (n+1)\times 10^{-4} \leq dn_s/d\ln k \leq -2\times 10^{-4} \; .
207: $$
208: This new inflation family features a {\it large window} of
209: consistency with the WMAP and LSS data for $ n = 2 $ that narrows
210: for growing $ n $. If forthcoming data on tensor modes pinpoints the
211: tensor to scalar ratio to be $ r<0.1 $, we predict that the {\it
212: symmetry breaking scale} for these models is $ \phi_0 \sim
213: 10\,M_{Pl} $ and that the scale of the field at which modes of
214: cosmological relevance today cross the Hubble radius is $ \phi_{c}
215: \sim M_{Pl} $.}
216:
217: \item{Chaotic inflationary models all yield a tensor to scalar
218: ratio $ r \geq 0.16 $, where the minimum value $ r=0.16 $
219: corresponds to small amplitude of the inflaton and coincides with
220: the value obtained from the monomial $ m^2\phi^2/2 $. The combined
221: marginalized data from WMAP3 \cite{WMAP3} yields a very small window
222: of field amplitude, around $|\phi_{c}|\sim 15~M_{Pl}$ within which
223: chaotic models are allowed by the data. These regions become
224: progressively smaller for larger $n$. Some small regions in field
225: space consistent with the WMAP3 data feature peaks in the running of
226: the scalar index but in the region consistent with the WMAP3 data in
227: chaotic inflation the running is again negligible ($ \sim 10^{-3}
228: $). If future observations determine a tensor to scalar ratio $
229: r<0.16 $, this by itself will {\bf rule out} a large family of
230: chaotic inflationary models.}
231: \end{itemize}
232: \section{Effective field theory, slow roll and $ 1/N_e $ expansions}
233:
234: In the absence of a fundamental microscopic description of
235: inflation, an effective field theory approach, when
236: combined with the slow roll expansion provides a robust paradigm for
237: inflation with predictive power. The reliability of the effective
238: field theory description hinges on a wide separation between the
239: Hubble and Planck scales, and is validated by the bound from
240: temperature fluctuations $ H/M_{Pl} < 10^{-5} $ \cite{1sN}.
241:
242: The slow roll
243: expansion relies on the smallness of a hierarchy of the
244: dimensionless ratios \cite{libros,barrow,reco},
245: \be
246: \epsilon_v = \frac{M^2_{Pl}}{2} \; \left[\frac{V^{'}(\phi)}{V(\phi)} \right]^2
247: \quad , \quad \eta_v = M^2_{Pl} \; \frac{V^{''}(\phi)}{V(\phi)}
248: \quad , \quad
249: \label{etav} \xi_v = M^4_{Pl} \; \frac{V'(\phi) \;
250: V^{'''}(\phi)}{V^2(\phi)}\;.
251: \ee
252: The effective field theory expansion in $H/M_{Pl}$ and the slow roll
253: expansion are independent, the latter can be interpreted as an
254: adiabatic expansion \cite{1sN} wherein the derivatives of the
255: inflationary potential are small.
256:
257: The CMB data is consistently described within the slow roll
258: expansion with inflationary potentials of the form \cite{libros,1sN}
259: \be
260: V(\phi) = M^4 \; v\left(\frac{\phi}{M_{Pl}}\right) \; .\label{potential}
261: \ee
262: Within the slow roll approximation the number
263: of e-folds before the end of inflation for which the value of the
264: field is $ \phi_{end} $ is given by
265: \be
266: N\left[\phi(t)\right] = -\frac{1}{M^2_{Pl}} \int_{\phi(t)}^{\phi_{end}} V(\phi)~
267: \frac{d\phi}{dV}~d\phi ~~ .
268: \ee
269: It proves convenient to introduce $ N_e $ as the typical number of
270: e-folds before the end of inflation
271: during which cosmologically relevant wavelengths cross the Hubble
272: radius during inflation, and $ \phi_c $ as the value of the inflaton
273: field corresponding to $ N_e $
274: \be
275: N_e = -\frac{1}{M^2_{Pl}} \int_{\phi_{c}}^{\phi_{end}} V(\phi)~ \frac{d\phi}{dV}~d\phi \,.
276: \label{Ne}
277: \ee
278: The precise value of $ N_e $ is certainly near $ N_e=50 $ \cite{libros,Ne}.
279: We will take the value $ N_e=50 $ as a reference
280: baseline value for numerical analysis, but from the explicit expressions
281: obtained in the systematic $ 1/N_e $ expansion below, it becomes a
282: simple rescaling to obtain results for arbitrary values of $ N_e $
283: [see eq. (\ref{etav2}) below].
284:
285: \medskip
286:
287: The form of the potential eq.(\ref{potential}) and the above
288: definition for the number of e-folds, suggests to introduce the
289: following rescaled field variable \cite{1sN}
290: \be \label{chi}
291: \phi = \sqrt{N_e} M_{Pl}\, \chi
292: \ee
293: where the rescaled field $\chi$ is
294: dimensionless. Furthermore, it is also convenient to scale $ N_e $
295: out of the potential and write
296: \be
297: V(\phi) = N_e \; M^4 \; w(\chi) \; .
298: \ee
299: With this definition, eq. (\ref{Ne}) becomes
300: \be \label{condi}
301: 1= - \int^{\chi_{end}}_{\chi_{c}} \frac{w(\chi)}{w'(\chi)}\,d\chi
302: \ee
303: where the prime stands for derivative with respect to $ \chi $,
304: $ \OChi $ is the value of $ \chi $ corresponding to $ N_e $ e-folds before the end of
305: inflation, and $ \chi_{end} $ is the value of $ \chi $ at the end of inflation.
306:
307: We emphasize that there is \emph{no} dependence on $ N_e $ in the
308: expression (\ref{condi}), therefore $ \chi_c,\chi_{end} $ only depend
309: on the coupling and the degree $ n $. The slow roll parameters
310: then become,
311: \be \label{etav2}
312: \epsilon_v = \frac{1}{2 \, N_e} \;
313: \left[\frac{w'(\chi_c)}{w(\chi_c)} \right]^2 \quad , \quad \eta_v =
314: \frac{1}{N_e} \; \frac{w''(\chi_c)}{w(\chi_c)} \;
315: \quad , \quad
316: \xi_v = \frac{1}{N^2_e} \; \frac{w'(\chi_c) \; w'''(\chi_c)}{w^2(\chi_c)} \; .
317: \ee
318: It is clear from eqs.(\ref{condi}) and (\ref{etav2}) that
319: during the inflationary stage when wavelengths of cosmological
320: relevance cross the horizon, it follows that $ w(\chi_c),
321: w'(\chi_c) \sim \mathcal{O}(1) $ leading to the slow roll
322: expansion as a consistent expansion in $ 1/N_e $ \cite{1sN}.
323:
324: The connection between the slow roll expansion
325: and the expansion in $ 1/N_e $ becomes more explicit upon introducing
326: a \emph{stretched} dimensionless time variable $\tau$ and a
327: dimensionless Hubble parameter $h$ as follows \cite{1sN}
328: \be
329: \tau = \frac{M^2 \,t }{\sqrt{N_e} \; M_{Pl}} ~~;~~ h =
330: \frac{M_{Pl} \; H}{\sqrt{N_e} \; M^2} \label{slowvars} \; .
331: \ee
332: In terms of $ \tau , \; \chi $ and $ h $ the Friedmann equation and the
333: equation of motion for the inflaton become,
334: \bea
335: h^2(\tau) = \frac{1}{3}\left[
336: \frac{1}{2 \, N_e}\left(\frac{d\chi}{d\tau}\right)^2+w(\chi)\right] \; ,
337: \label{fried} \cr \cr
338: \frac{1}{N_e} \; \frac{d^2\chi}{d\tau^2}+ 3 \; h \;
339: \frac{d\chi}{d\tau} + w'(\chi) = 0 \; . \label{inflaeq}
340: \eea
341: This equation of motion can be solved in a systematic expansion in
342: $ 1/N_e $. The definition (\ref{chi}) also makes manifest that $ \chi $
343: is a \emph{slowly varying field}, since a change $ \Delta \phi \sim M_{Pl} $
344: in the inflaton field implies a \emph{small} change of the
345: dimensionless field $ \Delta \chi \sim 1/\sqrt{N_e} $.
346:
347: In terms of these definitions, the CMB observables can be written
348: manifestly in terms of the $1/N_e$ expansion. The amplitude of
349: scalar perturbations is given by
350: \be
351: \Delta^2_{\mathcal{R}} =
352: \frac{N^2_e}{12\pi^2} \; \left( \frac{M}{M_{Pl}}\right)^4\,
353: \frac{w^3(\chi_c)}{\left[ w'(\chi_c)\right]^2} \; \, ,
354: \label{ampli}
355: \ee
356: and the spectral index $ n_s $, the ratio of
357: tensor to scalar perturbations $ r $ and the running of $ n_s $
358: with the wavevector $ dn_s / d\ln k $ become
359: \bea
360: n_s & = & 1-6 \;
361: \epsilon_v+ 2 \; \eta_v \label{ns} \quad , \quad r = 16 \;
362: \epsilon_v \\ \cr \frac{dn_s}{d\ln k } & = & -\frac{2}{N^2_e}\left\{
363: \frac{w'(\chi_c)w'''(\chi_c)}{w^2(\chi_c)}+3 \left[
364: \frac{w'(\chi_c)}{w(\chi_c)}\right]^4-
365: 4\,\frac{\big[w'(\chi_c)\big]^2\,w''(\chi_c)}{w^3(\chi_c)}\right\}
366: \label{dns} \; .
367: \eea
368: The virtue of the $ 1/N_e $ expansion is that we can choose a
369: reference baseline value for $ N_e $, say $ N_e=50 $ for numerical study,
370: and use the scaling with $ N_e $ of the slow roll parameters given by
371: eqs.(\ref{etav2}), (\ref{ns}) and (\ref{dns}) to obtain their values
372: for arbitrary $ N_e $, namely
373: \bea
374: \epsilon_v[N_e] & = &
375: \epsilon_v[50] \; \frac{50}{N_e} \quad , \quad
376: \eta_v[N_e] = \eta_v[50] \; \frac{50}{N_e} \quad , \quad
377: r[N_e] = r[50] \; \frac{50}{N_e} \quad , \nonumber \\
378: \frac{dn_s}{d\ln k }[N_e] & = & \frac{dn_s}{d\ln k }[50] \; \left(\frac{50}{N_e}\right)^2
379: \quad , \quad n_s[N_e] = n_s[50]+\left(1- n_s[50]\right) \; \frac{N_e-50}{N_e}
380: \; . \label{resca}
381: \eea
382: In what follows we will take $N_e=50$ as representative of the
383: cosmologically relevant case, however, the simple scaling relations
384: (\ref{resca}) allow a straightforward extrapolation to other
385: values.
386:
387: The combination of WMAP and SDSS (LRG) data yields \cite{WMAP3}
388: \bea
389: && n_s = 0.958\pm 0.016~~~~(\mathrm{assuming~} r=0 \mathrm{~with ~no~ running})\\
390: && r < 0.28 ~ (95\% \, CL) ~~\mathrm{no ~running}
391: \\&& r < 0.67 ~ (95\% \, CL) ~~\mathrm{with ~ running}
392: \label{wmapvals} \; .
393: \eea
394: The running must be very small and of the order $ \mathcal{O}(1/N^2_e)\sim 10^{-3} $
395: in slow roll for generic potentials\cite{1sN}. Therefore, we can
396: safely consider $ dn_s/d\ln k =0 $ in our analysis. Figure 14 in
397: ref.\cite{WMAP3} and figure 2 in ref.\cite{heclast} show that the
398: preferred value of $n_s$ slowly grows with the preferred value of
399: $ r $ for $ r>0 $. We find approximately that
400: \be
401: \frac{\Delta n_s}{\Delta r} \simeq 0.12 \label{ley}
402: \ee
403: Therefore, for $ r\sim 0.1 $
404: the central value of $ n_s $ shifts from $ n_s = 0.958 ~(r=0) $ to $ n_s
405: = 0.97~ (r=0.1) $ as can be readily gleaned from the quoted figures
406: in these references.
407:
408: As a simple example that
409: provides a guide post for comparison let us consider first the
410: monomial potential
411: \be
412: V(\phi) = \frac{\lambda}{2 \, n} \, \phi^{2
413: \, n}\,. \label{quadpot }
414: \ee
415: The case $ n=1 $ yields a satisfactory
416: fit to the WMAP data \cite{peiris,WMAP3}. For these potentials it
417: follows that,
418: \be
419: w(\chi) = \frac{\chi^{2 \, n}}{2 \, n}~~;~~ M^4 =
420: \lambda\,N_e^{n-1}\,M^{2 \, n}_{Pl}\, \,, \label{potchi}
421: \ee
422: inflation ends at $ \chi_{end} =0 $, and the value of the
423: dimensionless field $ \chi_c $ at $ \; N_e $ e-folds before the end
424: of inflation is
425: \be \label{chi50}
426: |\chi_c| = 2 \; \sqrt{n} \; .
427: \ee
428: These results lead to \cite{libros}
429: \bea
430: &&\epsilon_v = \frac{n}{2
431: \, N_e} \quad , \quad \eta_v = \frac{2 \, n-1}{2 \, N_e}\\&& n_s-1
432: = -\frac{n+1}{N_e} \label{nsmono} \quad , \quad
433: r = \frac{8 \, n}{N_e} \quad , \quad \frac{dn_s}{d\ln k } =
434: -\frac{n+1}{N^2_e}\,.
435: \eea
436: Taking $ N_e=50 $ as a baseline, these yield
437: \be
438: n_s-1 = -2 \, (n+1)\times 10^{-2}~\Big(\frac{50}{N_e}\Big) \quad , \quad r = 0.16
439: \, n ~\Big(\frac{50}{N_e}\Big) \quad , \quad
440: \frac{dn_s}{d\ln k }=-4 \, (n+1)\times 10^{-4}~\Big(\frac{50}{N_e}\Big)^2 \; . \label{indicesmono}
441: \ee
442:
443: \section{Family of models}
444:
445: We study the CMB observables $ n_s,r,dn_s/d\ln k $ for families of
446: new inflation and chaotic models determined by the following
447: inflationary potentials.
448: \bea
449: V(\phi) & = & V_0 -\frac{1}{2} \;
450: m^2 \; \phi^2 + \frac{\lambda}{2 \, n}\; \phi^{2 \, n} \quad , \quad {\rm
451: broken ~ symmetry} \label{rota}
452: \\ \cr V(\phi) & = & \frac{1}{2} \; m^2 \; \phi^2 +
453: \frac{\lambda}{2 \, n} \; \phi^{2 \, n} \quad , \quad {\rm unbroken ~
454: symmetry} \; . \label{cao}
455: \eea
456: Upon introducing the rescaled field
457: $ \chi $ given by eq. (\ref{chi}) we find that these potentials can be
458: written as
459: \be
460: V(\phi) = N_e \; m^2 \; M^2_{Pl} \; w(\chi) \; ,
461: \ee
462: where we recognize that
463: \be
464: M^4 = m^2 \; M^2_{Pl} \; \, .
465: \ee
466: Then, the family of potentials eqs.(\ref{rota})-(\ref{cao}) are
467: \bea
468: && w(\chi) = w_0-\frac{1}{2} \; \chi^2 + \frac{g}{2 \, n} \;
469: \chi^{2 \, n}\quad , \quad {\rm broken ~ symmetry} \label{neww} \\ \cr
470: && w(\chi) = \frac12 \; \chi^2 + \frac{g}{2 \, n} \; \chi^{2 \, n} \quad
471: , \quad {\rm unbroken ~ symmetry} \; \, .\label{newcao1}
472: \eea
473: where
474: $ w_0 $ and $ g $ are dimensionless and related to $ V_0 $ and $
475: \lambda $ by
476: \be
477: V_0 = w_0 \; N_e \; m^2 \; M^2_{Pl} \quad ,
478: \quad \lambda = \frac{m^2 \; g}{M^{2 \, n-2}_{Pl} \; N^{n-1}_e} \; .
479: \ee
480: New inflation models described by the dimensionless potential
481: given by eq. (\ref{neww}) feature a minimum at $ \chi_0 $ which is
482: the solution to the following conditions
483: \be
484: w'(\chi_0)=w(\chi_0)=0
485: \; .\label{condchi0} \ee These conditions yield, \be \label{coupdef}
486: g = \frac{1}{\chi^{2 \, n-2}_0} \quad , \quad w_0 = \frac{\chi^2_0}{2 \, n}
487: \; \left(n-1\right)\; ,
488: \ee
489: $ \chi_0 $ determines the scale of
490: symmetry breaking $ \phi_0 $ of the inflaton potential upon the
491: rescaling eq.(\ref{chi}), namely
492: \be
493: \phi_0 = \sqrt{N_e} \; M_{Pl}
494: \; \chi_0 \; \,. \label{fi0}
495: \ee
496: It is convenient to introduce the
497: dimensionless variable
498: \be
499: \label{defx} x = \frac{\chi}{\chi_0}
500: \ee
501: Then, from eqs.(\ref{coupdef}) and (\ref{defx}), the family of
502: inflation models eq.(\ref{neww})-(\ref{newcao1}) take the form
503: \bea
504: &&w(\chi) = \frac{\chi^2_0}{2 \, n} \; \left[ n \; (1-x^2) +
505: x^{2 \, n}-1\right] \quad , \quad {\rm broken ~ symmetry} \; ,
506: \label{newchi}\\
507: && w(\chi) = \frac{\chi^2_0}{2 \, n} \left[n \;
508: x^2+x^{2 \, n}\right] \quad , \quad {\rm unbroken ~ symmetry} \; .
509: \label{caochi}
510: \eea
511: In terms of the variable $ x $, the small and
512: large field regions for the potential eq.(\ref{newchi}) correspond to $
513: x<1 $ and $ x>1 $, respectively. The form of the dimensionless
514: potentials $ w(x) $ for both families are depicted in fig.
515: \ref{fig:newinflapot}. The left panel in this figure shows the new
516: (small field) or chaotic (large field) behavior of the broken
517: symmetry potential eq.(\ref{newchi}) depending on whether the
518: initial value of the inflaton is smaller or larger than the minimum
519: of the potential $ x = 1 $.
520:
521: \begin{figure}[h]
522: \begin{center}
523: \includegraphics[height=3in,width=3in,keepaspectratio=true]{newinflapotential.eps}
524: \includegraphics[height=3in,width=3in,keepaspectratio=true]{caoticpotential.eps}
525: \caption{Left panel: broken symmetry potential for for $ n=2 $, small
526: and large field cases. Right panel: unbroken symmetry potential
527: for $ n=2 $, large field inflation. } \label{fig:newinflapot}
528: \end{center}
529: \end{figure}
530:
531: \section{Broken Symmetry models.}
532:
533: Inflation ends when the inflaton field arrives to the minimum of the
534: potential. For the new inflation family of models eq.(\ref{newchi})
535: inflation ends for \be \chi_{end} = \chi_0 \; . \ee In terms of the
536: dimensionless variable $ x $, the condition eq.(\ref{condi}) becomes
537: \be
538: \frac{2 \, n}{\chi^2_0} = I_n(X) \label{chi0I} \quad , \quad
539: X=\frac{\OChi}{\chi_0} \equiv x_{c} \; , \ee where \be \label{IXn}
540: I_n(X) = \int_{X}^1 \frac{dx}{x} \; \frac{n \; (1-x^2) + x^{2 \, n}
541: - 1 }{1- x^{2 \, n-2}}= \int_{X}^1 \frac{n-\sum_{m=0}^{n-1}
542: x^{2m}}{\sum_{m=0}^{n-2} x^{2m}} \; \frac{dx}{x} \; .
543: \ee
544: This integral can be computed in closed form in terms of hypergeometric
545: functions \cite{gr} which can be reduced to a finite sum of
546: elementary functions\cite{pru}.
547:
548: For a fixed given value of $ X $, the value of $ \chi_0 $ and
549: therefore of the dimensionless coupling $ g $ is determined by the
550: equation (\ref{chi0I}). Once we obtain this value, the CMB
551: observables eqs. (\ref{ns})-(\ref{dns}) are obtained by evaluating
552: the derivatives of $ w(\chi) $ at the value $ \chi = \OChi = \chi_0
553: \; X $ with the corresponding value of the coupling $ g $. Thus, a
554: study of the range of possible values for $ n_s, \; r, \; dn_s/d\ln
555: k $ is carried out by exploring the relationship between these
556: spectral indices as a function of $ X $.
557: For this study we choose the baseline value $ N_e=50 $ from which the
558: indices can be obtained for arbitrary value of $N_e$ by the relation
559: (\ref{resca}).
560:
561: While the dependence of $ \chi_0 $ and $ g $ upon the variable $ X $ must
562: in general be studied numerically, their behavior in the
563: relevant limits, $ X \rightarrow (0, \; 1) $ for small field inflation and $ X >> 1 $
564: for large field inflation can be derived from eqs.(\ref{chi0I})-(\ref{IXn}).
565:
566: For small field inflation and $ X \rightarrow 0 $ the lower
567: limit of the integral dominates leading to
568: \be
569: \chi^2_0 \buildrel{X
570: \to 0}\over= \frac{2 \, n}{n-1} \; \frac1{\log\frac1{X}} \quad ,
571: \quad g \buildrel{X \to 0}\over= \left( \frac{n-1}{2 \, n} \;
572: \log\frac1{X}\right)^{n-1} \; , \label{smallX}
573: \ee
574: thus, as $ X \rightarrow 0 $ these are \emph{strongly coupled theories}. This
575: result has a clear and simple interpretation: for $ N_e =50 $ to be
576: the number of e-folds between $ x=X $ and $ x=1 $ the coupling $ g $
577: must be large and the potential must be steep, otherwise there would
578: be many more e-folds in such interval.
579:
580: For small field inflation and $ X \rightarrow 1^{-} $ the integral $ I_n(X) $
581: obviously vanishes and
582: \be\label{Xten1}
583: \chi^2_0 \buildrel{X \to 1^{-}}\over=\left( \frac2{1-X}\right)^2
584: \left[ 1 + \frac{2\, n - 1}9 \; (X-1) + {\cal O}(X-1)^2 \right]
585: \quad , \quad g \buildrel{X \to 1^{-}}\over= =\left[\frac12 \; (1-X)\right]^{2\,(n-1)}
586: \rightarrow 0\; ,
587: \ee
588: thus, as $ X \rightarrow 1^{-} $, these are a \emph{weakly coupled family of models}.
589:
590: \medskip
591:
592: For large field inflation and $ X \gg 1 $, the integral $ I_n(X) $ is dominated by the
593: term with the highest power, namely $ x^{2 \, n} $, leading to the behavior
594: \be\label{Xgran}
595: \chi^2_0 \buildrel{X \gg 1}\over= \frac{4 \, n}{X^2}
596: \quad , \quad g \buildrel{X \gg 1}\over= \left( \frac{X^2}{4 \, n} \right)^{n-1} \; ,
597: \ee
598: which leads to a strongly coupled regime.
599: The results of a numerical analysis are depicted in fig. \ref{fig:couplingsni}.
600:
601: \begin{figure}[ht!]
602: \begin{center}
603: \includegraphics[height=3.5in,width=3.5in,keepaspectratio=true]{coupnew.eps}
604: \caption{The coupling $ g $ as a function of $ X $, for the degrees
605: of the new inflation potential $ n=2, \; 3, \; 4 $. For $ X \to 1 ,
606: \; g $ vanishes as $ \left[\frac{1-X}2\right]^{2\,(n-1)} $. The
607: point $ X = 1, \; g = 0 $ corresponds to the monomial $ m^2 \;
608: \phi^2/2 $. $ g $ increases both for $ X \to 0 $ and for large $ X $
609: as, $ \left( \log\frac1{X}\right)^{n-1} $ and $ \left( \frac{X^2}{4
610: \, n} \right)^{n-1} $, respectively.} \label{fig:couplingsni}
611: \end{center}
612: \end{figure}
613:
614: Before we proceed with a numerical study of the CMB indices and the
615: tensor to scalar ratio, we can extract interesting and relevant
616: information by focusing on the region $ X \sim 1 $ which as
617: discussed above corresponds to a weakly coupled family for broken
618: symmetry potentials. This is the region near the \emph{minimum} of
619: the potential and the integral $ I_n(X) $ can be evaluated simply by
620: expanding $ w(\chi) $ and its derivative near the minimum. To
621: leading order in $ (X-1) $ the condition (\ref{chi0I}) leads to
622: eq.(\ref{Xten1}) and
623: \be \label{quadcond}
624: (\chi_c-\chi_0)^2 = 4
625: \quad {\rm or} \quad |\chi_c-\chi_0| = 2 \; .
626: \ee
627: This is precisely
628: eq.(\ref{chi50}) for $ n = 1 $ upon the shift $ \chi_c \to
629: \chi_c-\chi_0 $. Namely, eq.(\ref{quadcond}) is the condition
630: eq.(\ref{chi50}) for the quadratic monomial potential with minimum
631: at $ \chi = \chi_0 $ instead of $ \chi = 0 $ as in
632: eq.(\ref{potchi}). This is clearly a consequence of the fact that
633: near the minimum $ X = 1 $ the potential is quadratic, therefore for
634: $ X \sim 1 $ the quadratic monomial is an excellent approximation to
635: the family of higher degree potentials and more so because $ g \sim
636: 0 $. For $ X\sim 1 $ we find to leading order in $ (X-1) $ the
637: values:
638: \be
639: \epsilon_v =\eta_v = 0.01 ~\Big(\frac{50}{N_e}\Big)
640: \quad , \quad n_s = 0.96+0.04~\Big(\frac{N_e-50}{N_e}\Big) \quad ,
641: \quad r = 0.16~\Big(\frac{50}{N_e}\Big) \quad , \quad
642: \frac{dn_s}{d\ln k } = - 0.0008~\Big(\frac{50}{N_e}\Big)^2 \; .
643: \label{quadvals}
644: \ee
645: The fact that the potential eq.(\ref{newchi})
646: is quadratic around the minimum $ X = 1 $ explains why we have in
647: this limit identical results for new inflation with the potential
648: eq.(\ref{newchi}) and chaotic inflation with the monomial
649: potential $ m^2 \; \phi^2/2 $.
650:
651: As observed in \cite{WMAP3} these values of $ r, \; n_s $ for $N_e
652: \sim 50$ yield a good fit to the available CMB data.
653:
654: The results of the numerical analysis for $ \epsilon_v, \; \eta_v,
655: \; n_s, \; r $ and $ dn_s/d\ln k $ for the baseline value $N_e=50$
656: are depicted in fig. \ref{fig:epsilonv}, \ref{fig:NsNI},
657: \ref{fig:rni} and \ref{fig:dnsni} respectively.
658:
659: \begin{figure}[h]
660: \begin{center}
661: \includegraphics[height=3 in,width=3 in,keepaspectratio=true]{epsilonvni.eps}
662: \includegraphics[height=3in,width=3in,keepaspectratio=true]{etavni.eps}
663: \caption{Slow roll parameters as a function of $ X $ for $ N_e=50 $.
664: Left panel $ \epsilon_v $, right panel $ \eta_v $, for new inflation
665: with the degrees of the potential $ n=2,3,4 $. The results for
666: arbitrary values of $ N_e $ are obtained by multiplying by the factor
667: $ \frac{50}{N_e} $.} \label{fig:epsilonv}
668: \end{center}
669: \end{figure}
670:
671: \begin{figure}[h!]
672: \begin{center}
673: \includegraphics[height=3.5in,width=3.5in,keepaspectratio=true]{nsni.eps}
674: \caption{Scalar spectral index $ n_s $ for the degrees of the
675: potential $ n=2,3,4 $ for new inflation with $ N_e=50 $. The
676: vertical line delimits the smallest value of $n_s$ (for
677: $ r=0 $) \cite{WMAP3}. The grey dot at $ n_s=0.96, \; X=1 $
678: corresponds to the value for the monomial potential $ n = 1 , \; m^2
679: \; \phi^2/2 $. Notice that the small field behavior is $ n $
680: independent. For arbitrary $ N_e $ the result follows directly
681: from the $ N_e=50 $ value by using eq.(\ref{resca}).} \label{fig:NsNI}
682: \end{center}
683: \end{figure}
684:
685:
686: \begin{figure}[h!]
687: \begin{center}
688: \includegraphics[height=3.5 in,width=3.5 in,keepaspectratio=true]{rni.eps}
689: \caption{Tensor to scalar ratio $ r $ vs. $ X $
690: for the degrees of the potential $ n=2, \; 3, \; 4 $ for
691: new inflation with $ N_e=50 $.
692: The horizontal dashed line corresponds to the upper limit $ r=0.28~(95\% CL) $
693: from WMAP3 without running. The vertical dashed line
694: determines the minimum value of $ X , \; X \sim 0.2 $, consistent
695: with the WMAP limits for $ n_s $ as in fig. \ref{fig:NsNI}.
696: The grey dot at $ X=1, \; r=0.16 $ corresponds to the value for the
697: monomial potential $ m^2 \; \phi^2/2 $. The small field limit is
698: nearly independent of $ n $. For arbitrary $ N_e $ the result follows directly
699: from the $ N_e=50 $ value by using eq. (\ref{resca}).}
700: \label{fig:rni}
701: \end{center}
702: \end{figure}
703:
704: \begin{figure}[h!]
705: \begin{center}
706: \includegraphics[height=3.5 in,width=3.5 in,keepaspectratio=true]{dnsni.eps}
707: \caption{Running of the scalar index $ dn_s/d\ln k $ vs. $ X $ for
708: the degrees of the potential $ n=2,3,4 $ respectively for new
709: inflation with $ N_e=50 $. The small field behavior is independent
710: of $ n $. For arbitrary $ N_e $ the result follows directly
711: from the $ N_e=50 $ value by using eq. (\ref{resca}).}
712: \label{fig:dnsni}
713: \end{center}
714: \end{figure}
715:
716: The vertical dashed line in fig. \ref{fig:NsNI} at $ X \sim 0.2 $
717: determines the lower limit of $ X $ for which $ n_s $ is consistent
718: with the WMAP data for $r=0$. For large values of $ X , \; n_s $
719: approaches asymptotically the values for the monomial potentials $
720: \phi^{2 \, n} $ given by eq. (\ref{indicesmono}). For the larger
721: degrees $ n $, the asymptotic behavior of $ n_s $ and $ r $ settles
722: at larger values of $ X $, this is a consequence of the larger
723: region in which the coupling is small as observed in fig.
724: \ref{fig:couplingsni} for larger degrees $ n $. The horizontal
725: dashed lines with vertical downward arrows in fig. \ref{fig:rni}
726: determines the upper bound from WMAP \cite{WMAP3} given by eq.
727: (\ref{wmapvals}) \emph{without} running, since from fig.
728: \ref{fig:dnsni} the values of $ dn_s/d\ln k $ for these models are
729: negligible. The vertical dashed line with the right-pointing arrow
730: in fig. \ref{fig:rni} determines the values of $ X $ for which $ n_s
731: $ are consistent with the WMAP data (see also fig. \ref{fig:NsNI}).
732:
733: \medskip
734:
735: From these figures we see that unlike the case of a pure monomial potential
736: $ \lambda \, \phi^{2 \, n} $ with $ n\geq 2 $, there is a {\bf large} region of field
737: space within which the new inflation models given by eq.(\ref{rota}) are {\bf consistent}
738: with the bounds from marginalized WMAP3 data and the combined WMAP3 + LSS
739: data \cite{WMAP3}.
740:
741: \medskip
742:
743: Fig. \ref{fig:rvnsni} displays $ r $ vs $ n_s $ for the values $
744: n=2,3,4 $ in new inflation and indicate the trend with $n$.
745: While $ r $ is a monotonically increasing function of $ X, \; n_s $ features a
746: \emph{maximum} as a function of $ X $, hence $ r $ becomes a {\bf
747: double-valued} function of $ n_s $. The grey
748: dot at $ r=0.16, \; n_s =0.96 $ corresponds to the monomial
749: potential $ m^2 \; \phi^2/2 $ for $ N_e=50 $. Values below the grey
750: dot along the curve in fig. \ref{fig:rvnsni} correspond to small
751: fields $ X<1 $ while values above it correspond to large fields $
752: X>1 $. We see from figs. \ref{fig:rni} and \ref{fig:rvnsni} that
753: \emph{large fields systematically lead to larger values of} $ r $.
754: Models that fit the WMAP data to $ 95\%~CL $ are within the tilted
755: box in fig. \ref{fig:rvnsni}. The tilt accounts for the growth of the preferred value of
756: $ n_s $ with $ r $ \cite{WMAP3} according to eq. (\ref{ley}).
757:
758: \begin{figure}[h]
759: \begin{center}
760: \includegraphics[height=3.5 in,width=3.5 in,keepaspectratio=true]{rvnsni.eps}
761: \caption{Tensor to scalar ratio $ r $ vs. $ n_s $
762: for degrees of the potential $ n=2, \, 3, \, 4 $ respectively for new inflation with $ N_e=50 $.
763: $ r $ turns to be a {\bf double-valued} function
764: of $ n_s $ exhibiting a maximum value for $ n_s $.
765: The values inside the box between the dashed lines
766: correspond to the WMAP3 marginalized region of the
767: $(n_s,r)$ plane with $ (95\% CL): \; r < 0.28, \; 0.942+0.12 ~ r
768: \leq n_s \leq 0.974+0.12 ~ r $, see eq.(\ref{ley}). The grey dot
769: corresponds to the values for the monomial potential $ m^2\phi^2/2 $
770: and the value $ X=1: \; r = 0.16, \; n_s = 0.96 $.}
771: \label{fig:rvnsni}
772: \end{center}
773: \end{figure}
774:
775: \begin{figure}[h]
776: \begin{center}
777: \includegraphics[height=3.5 in,width=3.5 in,keepaspectratio=true]{ndnsni.eps}
778: \caption{Running of the scalar index $ dn_s/d\ln k $ vs. $ n_s $ for degrees of the potential
779: $ n=2, \, 3, \, 4 $ respectively for new inflation with $ N_e=50 $.
780: The values for arbitrary $ N_e $ follow directly
781: from the $ N_e=50 $ value by using eq. (\ref{resca}).}
782: \label{fig:ndnsni}
783: \end{center}
784: \end{figure}
785:
786: \medskip
787:
788: Fig. \ref{fig:ndnsni} displays the running of the scalar index
789: vs. $ n_s $ for the different members of the family of new inflation
790: showing clearly that running is all but negligible in the entire range of values
791: consistent with the WMAP data. This was expected since the running in
792: slow-roll is of the order $ \sim \frac1{N_e^2} \simeq 4 \times 10^{-4} $
793: [see eq.(\ref{dns})] \cite{1sN}.
794:
795: We note that $ dn_s/d\ln k $ is a
796: monotonically \emph{decreasing} function of $ X $ approaching
797: asymptotically the values for the monomials $ \phi^{2 \, n} $ given by
798: eq. (\ref{indicesmono}). $ X \sim 0.2$ which is the {\it minimum
799: value} of $ X $ consistent with the bounds from WMAP on $ n_s $ (see
800: fig. \ref{fig:NsNI}).
801:
802: \subsection{Field reconstruction }
803:
804: The above analysis suggests to study the \emph{inverse problem},
805: namely, for a given member of the family labeled by $ n $, we may ask
806: what is the value $ \phi_{c} $ of the field at Hubble crossing and what is the scale
807: $ \phi_{0} $ of symmetry breaking of the potential which are consistent with the
808: CMB+LSS data. This is tantamount to the program of reconstruction
809: of the inflaton potential advocated in ref. \cite{reco} and is
810: achieved as follows: eq.(\ref{chi0I}) yields
811: $ \chi_0=\chi_0[X] $ from which we obtain $ \chi_c = \chi_0 \; X $.
812: These results are then input into the
813: expression for $n_s$ by evaluating the potential $ w(\chi) $ and its
814: derivatives at the value of
815: $ \chi_c $. This yields $ n_s=n_s[\chi_c] $ which is then
816: inverted to obtain $ \chi_c= \chi_c[n_s] $ and thus $ \phi_{c} $.
817:
818: In the region $ X \sim 1 $ corresponding to the weakly coupled case, this
819: reconstruction program can be carried out as a systematic series in
820: \be \label{delta}
821: \Delta \equiv X -1 = \frac{\OChi}{\chi_0}-1 \; ,
822: \ee
823: by expanding the inflationary potential and its derivatives
824: in a power series in $ x $ around $ x=1 $ in the integrand of $ I_n(X) $ [eq.
825: (\ref{IXn})]. For $ X=1 $ the value of the scalar index $ n_s $ is
826: determined by the simple monomial $ m^2 \; \phi^2/2 $ which from eq.
827: (\ref{nsmono}) for $ n=1 $ is given by $ n_s-1 = -2/N_e $.
828: Therefore, in terms of $ n_s $, the actual expansion parameter is $ n_s-1+2/N_e $.
829:
830: We obtain $ n_s $ to first order in $ \Delta $ from
831: eqs.(\ref{etav2}), (\ref{ns}), (\ref{newchi}), (\ref{Xten1}) and
832: (\ref{delta}) with the result,
833: $$
834: n_s-1 = -\frac2{N_e}\left[ 1 + \frac{2 \, n - 1}{18} \; \Delta + {\cal O}(\Delta^2)\right]
835: $$
836: then, by inverting this equation we find:
837: \be
838: \Delta(n_s,n) = X-1 = -\frac{9
839: N_e}{2 \, n-1}\left( n_s-1+\frac{2}{N_e}\right) +
840: \mathcal{O}\left(\left[ n_s-1+\frac{2}{N_e}\right]^2\right) \; , \label{delns}
841: \ee
842: and from eqs. (\ref{Xten1}) and (\ref{delns}) we find,
843: \be
844: \chi_0(n_s,n) =
845: \frac{2 \; (2 \, n-1)}{9 \, N_e \left| n_s-1+\frac{2}{N_e}\right|}\left[ 1- \frac{N_e}{2}
846: \left( n_s-1+\frac{2}{N_e}\right) \right] +
847: \mathcal{O}\left(\left[ n_s-1+\frac{2}{N_e}\right]\right)
848: \label{chi0s}
849: \ee
850: The leading order ($ \propto 1/\Delta $) of this result for $ \chi_0(n_s,n) $
851: can be simply cast as eq.(\ref{quadcond}):
852: this is recognized as the condition to have $ 50 $ e-folds for the quadratic
853: monomial centered in the broken symmetry minimum [see discussion below eq.
854: (\ref{quadcond})].
855:
856: \medskip
857:
858: Finally, the value of the (dimensionless) field $ \chi_c $ at Hubble
859: crossing is determined
860: from $ \chi_c(n_s,n) = \chi_0 \; [1+\Delta(n_s,n)] $ from which we obtain
861: \be
862: \chi_c = \frac{2 \; (2 \, n-1)}{9 \, N_e \left|
863: n_s-1+\frac{2}{N_e}\right|}\left[ 1- \frac{(2 \, n+17) \; N_e}{2 \;
864: (2 \, n-1)}\left( n_s-1+\frac{2}{N_e}\right)
865: \right] + \mathcal{O}\left(\left[ n_s-1+\frac{2}{N_e}\right]\right) \; .
866: \label{chi50s}
867: \ee
868: The coupling constant $ g $ can be also expressed in terms of $ n_s $
869: in this regime with the result,
870: $$
871: g = \left[ \frac{9 \, N_e \left| n_s-1+\frac{2}{N_e}\right|}{2 \; (2 \, n-1)}\right]^{2 \, n - 2}\to 0 \; ,
872: $$
873: which exhibits the weak coupling character of this limit.
874:
875: This analysis shows that the region in field space that
876: corresponds to the region in $n_s$ that \emph{best fits the WMAP
877: data} can be systematically reconstructed in an expansion in
878: $ n_s-1+2/N_e $. This is yet another bonus of the $1/N_e$ expansion.
879: Although the above analysis can be carried out to an arbitrary
880: order in $ n_s-1+2/N_e $, it is more convenient to perform a numerical
881: study of the region outside from $ X\sim 1 $ to find the values
882: of
883: $ \chi_c $ and $ \chi_0 $ as a function of $ n_s $ for fixed values of $ n, \; N_e $. Figures
884: \ref{fig:chinsniSF} and \ref{fig:chinsniLF} display $ \chi_c, \; \chi_0 $ as a function of
885: $ n_s $ with $ N_e=50 $ for different values of $ n $ for the small field region $ X<1 $ and the
886: large field region $ X>1 $
887: respectively. The point $ X=1 $ is a degeneracy point and corresponds
888: to the quadratic monomial as discussed above.
889:
890: \begin{figure}[h]
891: \begin{center}
892: \includegraphics[height=2.5in,width=2.5in,keepaspectratio=true]{chinsni2.eps}
893: \includegraphics[height=2.5in,width=2.5in,keepaspectratio=true]{chinsni3.eps}
894: \includegraphics[height=2.5in,width=2.5in,keepaspectratio=true]{chinsni4.eps}
895: \caption{Reconstruction program for broken symmetry potentials with $ N_e=50 $,
896: small field case $ X<1 . \; \chi_{50}\equiv \chi_c $ and $ \chi_0$
897: vs. $ n_s $ for degrees of the potential
898: $ n=2,3,4 $, respectively. These values of $ \chi_c, \; n_s $ correspond to the region $ r<0.16 $.}
899: \label{fig:chinsniSF}
900: \end{center}
901: \end{figure}
902:
903: \begin{figure}[h]
904: \begin{center}
905: \includegraphics[height=2.5in,width=2.5in,keepaspectratio=true]{chinsnilf2.eps}
906: \includegraphics[height=2.5in,width=2.5in,keepaspectratio=true]{chinsnilf3.eps}
907: \includegraphics[height=2.5in,width=2.5in,keepaspectratio=true]{chinsnilf4.eps}
908: \caption{Reconstruction Program for broken symmetry potentials with $ N_e=50 $,
909: large field case $ X>1 $. $\chi_{50}\equiv \chi_c $ and $ \chi_0 $
910: vs. $ n_s $ for degrees of the potential
911: $ n=2,3,4 $, respectively. These values of $ \chi, \; n_s $ correspond to the region $ r>0.16 $.}
912: \label{fig:chinsniLF}
913: \end{center}
914: \end{figure}
915:
916: Finally, the values for the dimensionful field $ \phi $ are given
917: by $ \phi_{c} = \sqrt{N_e} \; M_{Pl} \; \chi_c , \; \phi_0 =
918: \sqrt{N_e} \; M_{Pl} \; \chi_{0} $. Figures \ref{fig:rvnsni},
919: \ref{fig:chinsniSF} and \ref{fig:chinsniLF} lead to the conclusion
920: that for the range of CMB parameters $ r < 0.1 $ and $ n_s \leq
921: 0.96 $, the typical value of the {\it symmetry breaking scale} is $
922: \phi_0 \sim 10 \; M_{Pl} $ and the value of the inflaton field at
923: which cosmologically relevant wavelengths crossed the Hubble radius
924: during new inflation is $ \phi_{c} \sim M_{Pl} $ with a weak
925: dependence on $ n $. For $ 0.1 < r < 0.16 $ we have $ |\phi_{c} -
926: \phi_0| \sim 15 \; M_{Pl} $.
927:
928: We obtain for the coupling $ g $ in the $ X \to 0 $ limit which is a strong
929: coupling regime [see eq.(\ref{smallX})] where $ n_s \ll 1 $,
930: $$
931: g = \left[ \frac{N_e}4 \; \left( 1 - \frac1{n} \right) (1 - n_s) \right]^{n-1} \; .
932: $$
933: Finally, we have the $ X \to \infty $ limit which is also a strong coupling
934: limit [see eq.(\ref{Xgran})] where $ n_s \to 1 - (n+1)/N_e $ and we find,
935: \bea
936: &&\chi^2_0 = (4 \, n)^2 \left[\frac{N_e(1-n_s) -(n + 1)}{4 \, n (n-1) + 3 \, \left(1
937: +\frac1{n-2} \right)}\right]^{\frac1{n-1}} \to 0 \cr \cr
938: && g = \frac{4 \, n (n-1) + 3 \, \left(1 +\frac1{n-2} \right)}{(4 \, n)^{n-1} \;
939: \left[N_e(1-n_s) - (n + 1)\right]} \to \infty \; .
940: \eea
941:
942: \section{Chaotic inflation models.}
943:
944: We now turn to the study of the family of chaotic inflationary
945: potentials given by eq. (\ref{caochi}). Taking that the end of
946: inflation corresponds to $ x=0 $, the condition eq.(\ref{condi}) now
947: becomes
948: \be
949: \frac{2 \, n}{\chi^2_0} = J_n(X) = \int^{X}_0 \frac{n
950: +x^{2 \, n-2}}{1+x^{2 \, n-2}}~ x~{dx}\,.\label{chis}
951: \ee
952: Again, this integral can be computed in closed form in terms of
953: hypergeometric functions \cite{gr} which can be reduced to a finite
954: sum of elementary functions\cite{pru}. For general values of $ X $
955: the integral will be studied numerically, but the small $ X $ region
956: can be studied by expanding the integrand in powers of $ x^{2 \,
957: n-2} $, with the result
958: \be
959: 1=\frac{\chi^2_0\,X^2}{4}\left[1-\frac{n-1}{n^2} \; X^{2 \, n-2}+
960: \mathcal{O}(X^{4 \, n-4}) \right] \; . \label{chi0cao}
961: \ee
962: For small $ X $ and recalling that $ X = \chi_c/\chi_0 $ this relation yields
963: \be \label{xi50c} |\chi_c| =2\left[1+\frac{n-1}{2 \, n^2} \; X^{2 \,
964: n-2}+\mathcal{O}(X^{4 \, n-4}) \right] \ee which is again, at
965: dominant order the relation for the quadratic monomial potential
966: eq.(\ref{chi50}) for $ n = 1 $. This must be the case because the
967: small field limit is dominated by the quadratic term in the
968: potential. For small fields, $ \chi_0 \approx 2/X $ and the
969: coupling $ g $ vanishes as,
970: \be
971: g(X) \buildrel{X \to 0}\over=
972: \left(\frac{X}{2}\right)^{2 \, n-2} \; .
973: \ee
974: The coupling as a function of $ X $ is shown in fig. \ref{fig:couplingscao}.
975:
976: \begin{figure}[h]
977: \begin{center}
978: \includegraphics[height=3in,width=3in,keepaspectratio=true]{coupcao.eps}
979: \caption{Coupling $ g $ as a function of $ X $ for $ N_e=50 $, for
980: $n=2,3,4$ for chaotic inflation. $ g $ turns to be a monotonically increasing
981: function of $ X $. $ g $ vanishes for $ X \to 0 $ as $ \left(\frac{X}{2}\right)^{2 \, n-2} $
982: in sharp contrast with new inflation where $ g $ strongly increases for $ X \to 0 $.}
983: \label{fig:couplingscao}
984: \end{center}
985: \end{figure}
986:
987: The dependence of $ \epsilon_v, \; \eta_v $ in the full range of
988: $ X $ for several representative values of $ n $ is studied numerically:
989: these results are displayed in fig. \ref{fig:epsilonvcao}.
990: In the small $ X $ regime, we obtain from eqs. (\ref{etav2}), (\ref{caochi}) and (\ref{chi0cao})
991: the expressions,
992: \bea
993: && \epsilon_v =
994: \frac{1}{2 \, N_e}\left[1+ \frac{(2 \, n-1)(n-1)}{n^2} \;
995: X^{2 \, n-2}+\mathcal{O}\left(X^{4 \, n-4}\right)\right] \label{evsX} \\
996: &&\eta_v = \frac{1}{2 \, N_e}\left[1+ \frac{(2 \, n-1)(n^2-1)}{n^2}
997: X^{2 \, n-2}+\mathcal{O}\left(X^{4 \, n-4}\right)\right]
998: \label{etavsX}
999: \eea
1000: As $ X\rightarrow 0 , \; \epsilon_v $ and $
1001: \eta_v $ tend to the result from the quadratic monomial potential,
1002: namely $ \epsilon_v=\eta_v=1/[2 \, N_e] $ as must be the case because the
1003: quadratic term dominates the potential for $ X\ll 1 $.
1004:
1005: Figures \ref{fig:Nscao}, \ref{fig:rcao} display $ n_s, \; r $ as
1006: functions of $ X $ for $ N_e=50 $ respectively. For $ X\rightarrow 0, \; n_s \rightarrow
1007: 0.96 $ and $ r \rightarrow 0.16 $ which are the values from the
1008: quadratic monomial $ m^2 \; \phi^2/2 $.
1009:
1010: For $ X\gg 1 $, the values of $ n_s, \; r $
1011: for the monomial potentials $ \phi^{2 \, n} $ are attained asymptotically,
1012: namely, (for $ N_e = 50 $): $ n_s-1 = -2 \, (n+1)\times 10^{-2} , \; r = 0.16 \, n $.
1013:
1014: Comparing figs. \ref{fig:Nscao}, \ref{fig:rcao} to those for the
1015: new inflation case, (figs. \ref{fig:NsNI} and \ref{fig:rni}), we note
1016: that the range in which the chaotic family provide a
1017: good fit to the WMAP data is \emph{very much smaller} than for new inflation.
1018: Fig. \ref{fig:Nscao} shows that in chaotic inflation
1019: {\it only} for $ n=2 $ the range of $ n_s $ is
1020: allowed by the WMAP data in a fairly extensive range of values of
1021: $ X $, whereas for $ n=3, \; 4 $ (and certainly larger), there is a
1022: \emph{relatively small} window in field space for $ X<1 $ which satisfies the
1023: data for $ n_s $ and $ r $ simultaneously.
1024:
1025: The tensor to scalar ratio $ r $ in chaotic inflationary models is
1026: {\it larger} than $ 0.16 $ for all values of $ X $, approaching asymptotically
1027: for large $ X $ the value $ r = 0.16 \, n $ associated to the monomial potentials $ \phi^{2 \, n} $.
1028:
1029: \begin{figure}[h]
1030: \begin{center}
1031: \includegraphics[height=3in,width=3in,keepaspectratio=true]{epsicao.eps}
1032: \includegraphics[height=3in,width=3in,keepaspectratio=true]{etacao.eps}
1033: \caption{ Left panel $ \epsilon_v $, right pannel $ \eta_v $ as a
1034: function of $ X $ for $ N_e=50 $, for chaotic inflation with degrees of the potential $ n=2,3,4 $.
1035: The small $ X $ behavior is $ n $ independent.} \label{fig:epsilonvcao}
1036: \end{center}
1037: \end{figure}
1038:
1039: \begin{figure}[h]
1040: \begin{center}
1041: \includegraphics[height=3.5in,width=3.5in,keepaspectratio=true]{nscao.eps}
1042: \caption{Scalar spectral index $ n_s $ for degrees of the potential
1043: $ n=2,3,4 $ respectively for chaotic inflation with $ N_e=50 $. For
1044: $ X \to 0, \; n_s $ reaches for all $ n $ the value $ n_s = 0.96 $
1045: corresponding to the monomial potential $ m^2\phi^2/2 $. }
1046: \label{fig:Nscao}
1047: \end{center}
1048: \end{figure}
1049:
1050: \begin{figure}[h]
1051: \begin{center}
1052: \includegraphics[height=3 in,width=3 in,keepaspectratio=true]{rcao.eps}
1053: \caption{Tensor to scalar ratio $ r $ vs. $ X $
1054: for degrees of the potential $ n=2,3,4 $ respectively for chaotic inflation with $ N_e=50 $.
1055: The horizontal dashed lines with the downward arrows delimit
1056: the region of $95\%~CL$ given by WMAP3 with no running $ r < 0.28 $ \cite{WMAP3}.}
1057: \label{fig:rcao}
1058: \end{center}
1059: \end{figure}
1060:
1061: Fig. \ref{fig:dnscao} displays $ dn_s/d\ln k $ as a function of
1062: $ X $, while the running is again negligible, it is {\it strikingly
1063: different} from the new inflation case. \emph{Again} this figure, in
1064: combination with those for $ n_s $ and $ r $ as functions of $ X $
1065: distinctly shows that {\bf only} $ n=2 $ in chaotic inflation is compatible with the
1066: bounds from the WMAP data, while for $ n=3,4 $ only a {\it small window}
1067: for $ X<1 $ is allowed by the data. We see from fig. \ref{fig:dnscao} that
1068: $ dn_s/d\ln k $ takes negative as well as positive values for chaotic inflation,
1069: in contrast with new inflation where $ dn_s/d\ln k < 0 $.
1070:
1071: \begin{figure}[h]
1072: \begin{center}
1073: \includegraphics[height=3.5 in,width=3.5 in,keepaspectratio=true]{dnscao.eps}
1074: \caption{Running of the scalar index $ dn_s/d\ln k $ vs. $ X $ for degrees of the potential
1075: $ n=2, \; 3, \; 4 $ respectively for chaotic inflation with $ N_e=50 $.
1076: The $ X \to 0 $ behavior is $ n $ independent. $ dn_s/d\ln k $ features a maximun value
1077: that gets stronger with increasing $ n $. For chaotic inflation $ dn_s/d\ln k $ takes negative
1078: as well as positive values, in contrast with new inflation where $ dn_s/d\ln k < 0$.}
1079: \label{fig:dnscao}
1080: \end{center}
1081: \end{figure}
1082:
1083: The fact that the combined bounds on $ n_s, \; r $ and $ dn_s/d\ln k
1084: $ from the WMAP3 data \cite{WMAP3} provide much more stringent
1085: constraints on chaotic models is best captured by displaying $ r $
1086: as a function of $ n_s $ in fig. \ref{fig:rvsncao}. The region
1087: allowed by the WMAP data lies within the tilted box delimited by the
1088: vertical and horizontal dashed lines that represent the $ 95\% CL $
1089: band \cite{WMAP3}.
1090:
1091: \begin{figure}[h]
1092: \begin{center}
1093: \includegraphics[height=3 in,width=3 in,keepaspectratio=true]{rvsncao.eps}
1094: \caption{Tensor to scalar ratio $ r $ vs. $ n_s $
1095: for degrees of the potential $ n=2, \; 3, \; 4 $ respectively
1096: for chaotic inflation with $ N_e=50 $.
1097: The range of $ 95\%~CL $ as determined by WMAP3 \cite{WMAP3} is within
1098: the tilted box delimited by :
1099: $ r < 0.28, \; 0.942+0.12 \, r \leq n_s \leq 0.974+0.12 \, r $, see
1100: eq.(\ref{ley}).}
1101: \label{fig:rvsncao}
1102: \end{center}
1103: \end{figure}
1104:
1105: A complementary assessment of the
1106: allowed region for this family of effective field theories is
1107: shown in fig. \ref{fig:ndnscao} which distinctly shows that {\bf only}
1108: the $ n=2 $ case of chaotic inflation is allowed by the WMAP3 data.
1109:
1110: \begin{figure}[h]
1111: \begin{center}
1112: \includegraphics[height=3.5 in,width=3.5 in,keepaspectratio=true]{ndnscao.eps}
1113: \caption{Running of the scalar index $ dn_s/d\ln k $ vs. $ n_s $ for degrees of the potential
1114: $ n=2, \, 3, \, 4 $ respectively for chaotic inflation with $ N_e=50 $. }
1115: \label{fig:ndnscao}
1116: \end{center}
1117: \end{figure}
1118:
1119: \subsection{Field reconstruction }
1120:
1121: The reconstruction program proceeds in the same manner as in the
1122: case of new inflation: the first step is to obtain $ \chi_0(X) $
1123: from eq.(\ref{chis}). Then $ \epsilon_v $ and $ \eta_v $ are
1124: obtained as a function of $ X $ which yields $ n_s(X) $. Inverting
1125: this relation we find $ X=X(n_s) $ and finally $ \chi_c(n_s) =
1126: \chi_0 \; X(n_s) $. While this program must be carried out
1127: numerically, we can gain important insight by focusing on the small
1128: $ X $ region and using eq.(\ref{chi0cao}).
1129:
1130: From eqs. (\ref{ns}), (\ref{evsX}) and (\ref{etavsX}) we find
1131: \be
1132: n_s-1+\frac{2}{N_e} = X^{2 \, n-2} \; \frac{(2 \,
1133: n-1)(n-1)(n-2)}{n^2\,N_e}+\mathcal{O}\Big(X^{4n-4}\Big)\label{expacao}
1134: \ee
1135: As $ X \rightarrow 0 $ it follows that $ n_s \rightarrow 1-2/N_e
1136: $ which is the value for the scalar index for the quadratic monomial
1137: potential $ m^2\phi^2/2 $. However, for $n>2$ this limit is
1138: approached from above, namely for $ n>2 $ it follows that
1139: $ n_s>1-2/N_e $. The small $ X $ region corresponds to small departures
1140: of $ n_s $ from the value determined by the quadratic monomial $
1141: 1-2/N_e $ but always larger than this value for $ n>2 $. In the small
1142: field limit we reconstruct the value of $ \chi_c $ in an expansion in
1143: $ n_s-1+2/N_e $. The leading order in this expansion is obtained by
1144: combining eqs. (\ref{xi50c}) and (\ref{expacao}), we obtain
1145: \be
1146: \label{chifif} |\chi_c|
1147: =2\Bigg[1+\frac{\big(n_s-1+\frac{2}{N_e}\big)N_e}{2 \, (2 \, n-1) \,
1148: (n-2)}\Bigg]+ {\cal O}\left( \left[n_s-1+2/N_e \right]^2\right)
1149: \ee
1150: Obviously, this leading order term is singular at $ n=2 $, this is a
1151: consequence of the result eq.(\ref{expacao}) which entails that for
1152: $ n=2 $ the expansion must be pursued to higher order, up to $
1153: X^{4n-4} $.
1154:
1155: We find from eqs. (\ref{ns}), (\ref{evsX}) and (\ref{etavsX}) for $
1156: n = 2 $,
1157: \be \label{neq2}
1158: n_s-1+\frac{2}{N_e} = - \frac{17}{24 \,
1159: N_e} \; X^4 + \mathcal{O}\Big(X^6\Big) \quad , \quad n=2 \; ,
1160: \ee
1161: therefore,
1162: \be\label{chin2}
1163: |\chi_c| =2\Bigg[1+\sqrt{\frac{3
1164: \, N_e}{136}\left(1 -\frac2{N_e}-n_s \right)}\,\Bigg] + {\cal
1165: O}\left(n_s-1+\frac2{N_e} \right)\quad , \quad n=2 \; .
1166: \ee
1167: We see that the derivative of $ \chi_c $ with respect to $ n_s $ is
1168: singular for $ n=2 $ at $ n_s=1-\frac2{N_e} $. We note that for
1169: $ n=2 $ there is a sign change with respect to the cases $ n>2 $ and
1170: $ n_s-1+2/N_e \leq 0 $ as determined by eq. (\ref{neq2}).
1171:
1172: Fig. \ref{fig:chinscao} shows $ \chi_c $ as a function of $ n_s $
1173: for $ n=2, \; 3, \;4 $ for $ N_e=50 $. The case $ n=2 $ clearly
1174: shows the singularity in the derivative $ \partial \chi_c/
1175: \partial n_s $ at $ n_s = 1-2/N_e = 0.96 $ [see eq.(\ref{chin2})].
1176:
1177:
1178: Combining fig. \ref{fig:chinscao} with fig. \ref{fig:rvsncao} it is
1179: clear that there is a small window in field space within which
1180: chaotic models provide a good fit to the WMAP3 data, for $N_e=50$ we
1181: find:
1182: \bea \label{regiochi50}
1183: && n =2 ~: ~~0.95 \lesssim n_s \leq 0.960
1184: ~,~2.0 \leq |\chi_c| \lesssim 2.25 \nonumber \\&& n =3~: ~~0.96
1185: \leq n_s \lesssim 0.965 ~,~2.0 \leq |\chi_c| \lesssim 2.15 \nonumber \\
1186: && n =4~: ~~0.96 \leq n_s \leq 0.975 ~,~2.0 \leq |\chi_c| \lesssim
1187: 2.10 \; .
1188: \eea
1189: Restoring the dimensions via eq. (\ref{chi}) these
1190: values translate into a narrow region of width $\Delta \phi \lesssim
1191: 1.5\,M_{Pl}$ around the scale $|\phi_c| \sim 15~M_{Pl}$.
1192:
1193: Therefore, the joint analysis for $ n_s, \; r, \; dn_s/d\ln k $
1194: distinctly reveals that: (i) chaotic models favor {\it larger}
1195: values of $ r $ thus, larger tensor amplitudes, and (ii) chaotic models feature
1196: {\it smaller} regions in field space consistent with the CMB and
1197: large scale structure data. {\it Only} the case $ n=2 $ features a
1198: larger region of consistency with the combined WMAP3 data.
1199:
1200: \begin{figure}[h]
1201: \begin{center}
1202: \includegraphics[height=3.5 in,width=3.5 in,keepaspectratio=true]{chi50cao.eps}
1203: \caption{Reconstruction program for chaotic inflation with $ N_e=50 $,
1204: $ \chi_c $ vs. $ n_s $ for $ n=2, \, 3, \, 4 $ respectively. }
1205: \label{fig:chinscao}
1206: \end{center}
1207: \end{figure}
1208:
1209: \section{Conclusions}
1210:
1211: The fact that current observations of the CMB and LSS are already
1212: placing constraints on inflationary models which will undoubtedly
1213: become more stringent with forthcoming observations, motivates a
1214: study of the predictions for the CMB power spectra from different
1215: inflationary scenarios. We perform a systematic study of
1216: \emph{families} of single field new and chaotic inflation slow roll
1217: models characterized by effective field theories with potentials of
1218: the form
1219: \bea
1220: V(\phi) &=& V_0 -\frac{1}{2} \; m^2 \; \phi^2 +
1221: \frac{\lambda}{2 \, n}\; \phi^{2n} \quad , \quad {\rm broken ~ symmetry}
1222: \label{nuev}\\V(\phi) & = & \frac{1}{2} \; m^2 \; \phi^2 +
1223: \frac{\lambda}{2 \, n} \; \phi^{2 \, n} \quad , \quad {\rm unbroken ~
1224: symmetry}\label{chau} \; .
1225: \eea
1226: Unlike the approach followed in \cite{peiris,WMAP3} based on the
1227: inflationary flow equations\cite{hoki}, or more recent
1228: studies which focused on specific inflationary
1229: models \cite{salama}, or on statistical analysis of models \cite{kinkol},
1230: we implement an expansion in $ 1/N_e $ where $ N_e \sim
1231: 50 $ is the number of e-folds before the end of inflation when
1232: wavelengths of cosmological relevance today cross the Hubble radius
1233: during inflation. We provide an analysis of the dependence of CMB
1234: observables ($ n_s, \; r $ and $ dn_s/d\ln k $) with $ n $ and
1235: establish the region in field space within which these families
1236: provide a good agreement with the WMAP3 data combined with large
1237: scale surveys.
1238:
1239:
1240: For new inflation models with potentials eq.(\ref{nuev}) there are two distinct
1241: regions corresponding to values of the inflaton field smaller
1242: (small field) or larger (large field) than the symmetry breaking
1243: scale. For this family we find a {\bf wide range} in the $(n_s,r)$
1244: plane in which the different members $ n=2, \; 3, \; 4...$ are
1245: allowed by the data both for small and large fields with negligible
1246: running of the scalar index:
1247: $$
1248: -4 \; (n+1)\times 10^{-4} \leq dn_s/d\ln k \leq -2\times 10^{-4} \; .
1249: $$
1250: For $ N_e=50 $ the values $ n_s=0.96,r=0.16 $ which are those
1251: determined by the simple monomial potential $ m^2 \; \phi^2/2 $
1252: determine a divide and a degeneracy point in the field and parameter
1253: space. Small field regions yield $ r< 0.16 $ while large field
1254: regions correspond to $ r>0.16 $.
1255:
1256: The $ 1/N_e $ expansion also provides a powerful tool to
1257: implement a {\it reconstruction program} that allows to extract the
1258: value of the field $ N_e $ e-folds before the end of inflation, and in
1259: the case of new inflationary models, the symmetry breaking scale.
1260:
1261: We find that the region of field space favored by the WMAP3 data can
1262: be explored in a systematic expansion in $n_s-1+2/N_e$. An analytic
1263: and numerical study of this region lead us to conclude that if
1264: forthcoming data on tensor modes favors $ r<0.16 $ then
1265: new inflation {\bf is favored}, and we {\bf predict} for $ r < 0.1 $ that
1266: (i) the {\bf symmetry breaking scale} is
1267: $$
1268: \phi_0 \sim 10~M_{Pl} \; ,
1269: $$
1270: and (ii) the value of the field when cosmologically relevant
1271: wavelengths cross the Hubble radius is $ |\phi_c|\sim M_{Pl} $ .
1272:
1273: The family of chaotic inflationary models characterized by the
1274: potentials eq.(\ref{chau}) feature tensor to scalar ratios $ r \geq
1275: 0.16 $ (for $ N_e=50 $), with the minimum, $ r=0.16 $ obtained in
1276: the limit of small inflaton amplitude and corresponds to the
1277: monomial potential $ m^2 \; \phi^2/2 $ which is again a degeneracy
1278: point for this family of models.
1279:
1280: The combined marginalized data from WMAP3 \cite{WMAP3} yields a very
1281: small window within which chaotic models are allowed by the data,
1282: the largest region of overlap with the $ (r,n_s) $ WMAP3 data
1283: corresponds to $ n=2 $ and the width of the region decreases with
1284: larger $n$. The typical scale of the field at Hubble crossing for
1285: these models is $|\phi_c|\sim 15~M_{Pl}$ (for $N_e = 50$). Some
1286: small regions in field space consistent with the WMAP3 data feature
1287: peaks in the running of the scalar index but in the region
1288: consistent with the WMAP3 data in chaotic inflation the running is
1289: again negligible. If future observations determine a tensor to
1290: scalar ratio $ r<0.16 $, such bound will, all by itself, {\bf rule
1291: out} the large family of chaotic inflationary models of the form
1292: (\ref{chau}) for any $ n $.
1293:
1294: \begin{acknowledgments} D. B. and C. M. Ho thank the US NSF for support under grant
1295: PHY-0242134, LPTHE and the Observatoire de Paris and LERMA for
1296: hospitality during this work. D.B. thanks A. Kosowsky for
1297: illuminating conversations.
1298: \end{acknowledgments}
1299:
1300: \begin{thebibliography}{99}
1301: \bibitem{infla} D. Kazanas, Astrophys. J \textbf{241}, L59 (1980).
1302: A. Guth, Phys. Rev. \textbf{D23}, 347 (1981); astro-ph/0404546 (2004).
1303: K. Sato, Mon. Not. R. Astron. Soc. \textbf{195}, 467 (1981).
1304:
1305: \bibitem{libros} E. W. Kolb and M. S. Turner, {\em The Early Universe}
1306: Addison Wesley, Redwood City, C.A. 1990.
1307: P. Coles and F. Lucchin, {\em Cosmology}, John Wiley, Chichester,
1308: 1995.
1309: A. R. Liddle and D. H. Lyth,
1310: {\em Cosmological Inflation and Large Scale Structure}, Cambridge
1311: University Press, 2000.
1312: S. Dodelson, {\em Modern Cosmology}, Academic Press, 2003.
1313: D. H. Lyth , A. Riotto, Phys. Rept. \textbf{314}, 1 (1999).
1314:
1315: \bibitem{fluc} V. F. Mukhanov , G. V. Chibisov, Soviet Phys.
1316: JETP Lett. \textbf{33}, 532 (1981); V. F. Mukhanov, H. A. Feldman ,
1317: R. H. Brandenberger, Phys. Rept. \textbf{215}, 203 (1992).
1318: S. W. Hawking, Phys. Lett. \textbf{B115}, 295 (1982).
1319: A. H. Guth , S. Y. Pi, Phys. Rev. Lett. \textbf{49}, 1110 (1982).
1320: A. A. Starobinsky, Phys. Lett. \textbf{B117}, 175 (1982).
1321: J. M. Bardeen, P. J. Steinhardt , M. S. Turner, Phys. Rev.
1322: \textbf{D28}, 679 (1983).
1323:
1324: \bibitem{WMAPa} C. L. Bennett \emph{et.al.} (WMAP collaboration),
1325: Ap. J. Suppl. \textbf{148}, 1 (2003).
1326: A. Kogut \emph{et.al.} (WMAP collaboration), Ap. J. Suppl.
1327: \textbf{148}, 161 (2003).
1328: D. N. Spergel \emph{et. al. }(WMAP collaboration), Ap. J. Suppl.
1329: \textbf{148}, 175 (2003).
1330:
1331: \bibitem{peiris} H. V. Peiris \emph{et.al.} (WMAP collaboration), Ap. J.
1332: Suppl.\textbf{148}, 213 (2003).
1333:
1334: \bibitem{WMAP3} D. N. Spergel \emph{et. al. }(WMAP collaboration), astro-ph/0603449.
1335:
1336: \bibitem{WMAP3b} L. Page, \emph{et. al. }(WMAP collaboration), astro-ph/0603450.
1337: G. Hinshaw,\emph{et. al. }(WMAP collaboration), astro-ph/0603451.
1338: N. Jarosik, \emph{et. al. }(WMAP collaboration),
1339: astro-ph/0603452.
1340:
1341:
1342: \bibitem{SDSS} M. Tegmark \emph{et.al.} Phys. Rev
1343: \textbf{D69},103501, (2004)
1344: D. J. Eisenstein \emph{et.al.} ApJ. \textbf{633}, 560 (2005)
1345: U. Seljak et al., Phys. Rev. D71, 103515 (2005).
1346:
1347: \bibitem{2dF} A. G. S\'anchez \emph{et. al.}, Mon. Not. Roy. Astron. Soc. {\bf 366},
1348: 189 (2006). S. Cole \emph{et.al.} Mon. Not. Roy. Astron. Soc. {\bf 362}, 505
1349: (2005).
1350:
1351: \bibitem{sanchez} A. G. Sanchez and C. M. Baugh, astro-ph/0612743.
1352:
1353: \bibitem{hoki} M. B. Hoffman and M. S. Turner, Phys. Rev.
1354: \textbf{64}, 023506 (2001). W. H. Kinney, Phys. Rev. \textbf{D66},083508 (2002).
1355:
1356: \bibitem{salama} C. Savage, K. Freese and W. H. Kinney, Phys. Rev. \textbf{D 74}, 123511
1357: (2006); L. Alabidi and D. H. Lyth, astro-ph/0510441,
1358: astro-ph/0603539; J. Martin and C. Ringeval, JCAP \textbf{0608}
1359: (2006) 009.
1360:
1361: \bibitem{kinkol} W. H. Kinney, E. W. Kolb, A. Melchiorri and A.
1362: Riotto, Phys.Rev. D74 (2006) 023502.
1363:
1364: \bibitem{fipele} F. Finelli, M. Rianna, N. Mandolesi, JCAP
1365: \textbf{0612}, 006 (2006). H. Peiris, R. Easther, JCAP \textbf{0610}, 017
1366: (2006); JCAP \textbf{0607}, 002 (2006); R. Easther, H. Peiris, JCAP
1367: \textbf{0609}, 010 (2006). S. M. Leach, A. R Liddle, Phys.Rev. \textbf{D68}, 123508
1368: (2003); Mon.Not.Roy.Astron.Soc. \textbf{341}, 1151 (2003).
1369:
1370: \bibitem{1sN}
1371: D. Boyanovsky, H. J. de Vega and N. G. Sanchez, Phys. Rev.
1372: \textbf{D73}, 023008 (2006). See also ref. \cite{mangano}.
1373:
1374: \bibitem{reco} J. Lidsey, A.
1375: Liddle, E. Kolb, E. Copeland, T. Barreiro and M. Abney, Rev. of Mod.
1376: Phys. {\bf 69}, 373, (1997).
1377:
1378: \bibitem{Ne} W. H. Kinney, A. Riotto, JCAP \textbf{0603}, 0.11
1379: (2006); M. Giovannini, arXiv:astro-ph/0703730.
1380:
1381: \bibitem{ciri} D. Cirigliano, H. J. de Vega and N. G. Sanchez, Phys.
1382: Rev. \textbf{D71}, 103518 (2005);
1383:
1384: \bibitem{pre} H. J. de Vega
1385: and N. G. Sanchez, Phys. Rev. \textbf{D74}, 063519 (2006).
1386:
1387: \bibitem{heclast} C. Destri, H. J. de Vega, N. G. Sanchez, arXiv:astro-ph/0703417
1388:
1389: \bibitem{barrow} A. R. Liddle, P. Parsons , J. D. Barrow, Phys.
1390: Rev. \textbf{D50}, 7222 (1994).
1391:
1392: \bibitem{mangano} G. Mangano, G. Miele, C. Stornaiolo, Mod.Phys.Lett.A10, 1977
1393: (1995).
1394:
1395: \bibitem{gr} I. S. Gradshteyn and I. M. Ryshik, Table of Integrals, Series
1396: and Products, Academic Press, 1980.
1397:
1398: \bibitem{pru} A. P. Prudnikov, Yu. A. Brichkov, O, I. Marichev, Integrals and Series,
1399: vol. 3, Nauka, Moscow, 1986.
1400:
1401: \end{thebibliography}
1402:
1403:
1404: \end{document}
1405: