IL NUOVOCIMENTO Vol. ?, N. ? ? General Aspects of the de Sitter phase 4 0 1 2 3 4 3 Giovanni Imponente ( ) ( ) ( ) and Giovanni Montani ( ) ( ) 0 2 1 ( ) Dipartimento di Fisica, Universitá di Napoli “Federico II”, Napoli – Italy n (2) INFN – Sezione di Napoli a 3 J ( ) ICRA – International Center for Relativistic Astrophysics 4 3 ( ) Dipartimento di Fisica – G9 Universitá di Roma “La Sapienza”, Roma – Italy 1 1 v 5 5 0 Summary.— 1 Wepresent adetailed discussion of theinflationary scenario in thecontext of inho- 0 mogeneous cosmologies. After a review of the fundamental features characterizing 4 theinflationarymodel,asreferredtoahomogeneousandisotropic Universe,wede- 0 velop a generalization in view of including small inhomogeneous corrections in the / c theory. q A second step in our discussion is devoted to show that the inflationary scenario - providesavaluabledynamical“bridge” betweenagenericKasner-likeregimeanda r g homogeneous and isotropic Universein thehorizon scale. : This result is achieved by solving the Hamilton-Jacobi equation for a Bianchi IX v model in the presence of a cosmological space-dependent term. i X Inthisrespect,weconstructaquasi-isotropicinflationarysolution basedontheex- pansion of theEinstein equations up to first-twoorders of approximation, in which r a theisotropyoftheUniverseisduetothedominanceofthescalarfieldkineticterm; thefirst order of approximation corresponds to the inhomogeneous corrections and is driven bythe matterevolution. Weshowhowsuchaquasi-isotropicsolutioncontainsacertainfreedominfixingthe spacefunctionsinvolvedintheproblem. Themainphysicalissueofthisanalysiscor- respondstooutlinetheimpossibilityfortheclassicaloriginofdensityperturbations, duetothe exponential decay of the matter term during thedeSitter phase. PACS 98.80.Bp – Origin and formation of theUniverse. PACS 98.80.Cq – Inflationary Universe. 1. – Introduction ThehomogeneityandisotropyoftheVeryEarlyUniverseevolvingtowardstheinitial singularity shows instability for density perturbations [27]. (cid:13)c SocietaÌtalianadiFisica 1 2 GIOVANNIIMPONENTE and GIOVANNIMONTANI We will discuss how to connect the Mixmaster dynamics and its properties with an inflationary scenario, as well as how such a scheme can be interpolated with a quasi- isotropic solution of the Einstein equations [14, 15, 16], in order to recover the homogeneous and isotropic Universe so far observed. A significant degree of inhomogeneity is manifested at the scale of galaxies, clusters, etc.,neverthelessonsufficientlylargescales(oftheorderof100Mpc)inhomogeneitiesare smoothedoutandisotropyandhomogeneityarereachedlikeintheFriedmann-Lemaitre- Robertson-Walker (FLRW) model. Even the early Universe exhibits a level of isotropy and homogeneity by far higher than now, as testified by the extreme uniformity (of the order of 10−4) of the Cosmic MicrowaveBackground Radiation (CMBR). ThehomogeneityandisotropyoftheUniverseiswelltestedupto10−3 10−2seconds − of its life by the very good agreement between the abundances of light elements as predicted by the Standard Cosmological Model (SCM) and the one actually observed [24];furthermore,reliableindicationssupporttheideathattheUniverseevolvedthrough an inflationary scenario and since that age it reached isotropy and homogeneity on the horizon scale at least [9, 36]. Despite of these evidences favourable to the large scale homogeneity, many relevant features suggestthat in the veryearly stage of evolution it had to be described by much more general inhomogeneous solutions of the Einstein equations. With respect to this we remark the following two issues: (i) the backward instability of the density perturbations to an isotropic and homoge- neousUniverse[27]allowsustoinferthat,whenapproachingtheinitialsingularity, the dynamics evolved as more general and complex models; thus the appearance of an oscillatory regime is expected in view of its general nature, i.e.because its perturbations correspond simply to redefine the spatial gradients involved in the Cauchy problem. (ii) Since the Universe dynamics during the Planckian era underwent a quantum regime, then no symmetry restrictions can be a priori imposed on the cosmological model; in fact, the wave functional of the Universe can provide information on the casual scale at most, and therefore the requirement for a global symmetry to hold would imply a large scale correlation of different horizons. Thus, the quantum evolution of the Universe is appropriately described only in terms of a generic inhomogeneous cosmologicalmodel. In this paper, we address two main topics concerning the Universe evolution: in first place, we show how the inflationary scenario provides the natural framework to link the inhomogeneous oscillatory regime with anisotropic and homogeneous Universe on the horizonscale;then,weinvestigatehowaclassicaloriginofinhomogeneousperturbations is ruled out out by a de Sitter phase in a quasi-isotropic scheme. 2. – The Standard Cosmological Model The most general metric which is spatially homogeneous and isotropic is the FLRW one, which in terms of the comoving coordinates (t,r,θ,φ) reads dr2 (1) ds2 =dt2 R2(t) +r2dθ2+r2sin2θdφ2 − 1 kr2 (cid:18) − (cid:19) GENERALASPECTSOFTHEDESITTERPHASE 3 where the scale factor R(t) is a generic function of time only and, for an appropriate rescaling of the coordinates, the factor k =0, 1 distinguishes the sign of constant spa- ± tial curvature. Such coordinates represent a reference frame participating in the expansion of the Uni- verse: an observer at rest will remain at rest, leaving the effects of the expansion to the cosmic scale factor R(t). The FLRW dynamics is reducedto the time dependence ofthe scalefactor R(t) once solved the Einstein equations with a stress-energy tensor T for all the fields present µν (matter,radiation,etc.) whichthenmustbediagonal,withthespatialcomponentsequal to eachotherforthe homogeneityandisotropyconstraints. Asimple realizationisgiven by the perfect fluid one, characterized by a space-independent energy density ρ(t) and pressure p(t) (2) Tµ =diag(ρ, p, p, p), ν − − − and in this case the 0 0 component of the Einstein equations reads − R˙2 k 8πG (3) + = ρ R2 R2 3 which is the Friedmann equation, while the i i components are − R¨ R˙2 k (4) 2 + + = 8πGp. R R2 R2 − The difference between (3) and (4) leads to R¨ 4πG (5) = (ρ+3p), R − 3 which is solved for R(t) once providedan equation of state, i.e.a relationbetween ρ and p. When the Universe was radiation-dominated, as in the early period, the radiation component provided the greatest contribution to its energy density and for a photon gas we have p=ρ/3. Thepresent-timeUniverseis,onthecontrary,matter-dominated: the“particles”(i.e.the galaxies) have only long-range (gravitational) interactions and can be treated as a pres- sureless gas (“dust”): the equation of state is p=0. Nevertheless,theenergytensorappearingintheEinsteinfieldequationsdescribesthe complete local energy due to all non-gravitational fields, while gravitational energy has a non-localcontribution. An unambiguousformulationforsucha non-localexpressionis foundonlyintheexpressionsusedatinfinityforanasymptoticallyflatspace-time[34,2]. This is due to the property of the mass-energy term to be only one component of the energy-momentum tensor which can be reduced only in a peculiar case to a four-vector expression which can not be summed in a natural way. Bearing in mind such difficulties, the conservation law Tµν =0 leads to ;ν (6) d(ρR3)= pd(R3) − 4 GIOVANNIIMPONENTE and GIOVANNIMONTANI i.e.the first law of thermodynamics in an expanding Universe which, via the equation of state, leads to a differential relation for ρ(R). If we take its solution together with the Friedmann equation (3) corresponding to k =0, then we find the following behaviours (7a) RADIATION ρ R−4, R t1/2 ∝ ∝ (7b) MATTER ρ R−3, R t2/3 ∝ ∝ whereaslongastheUniverseisnotcurvature-dominated,i.e.forsufficientlysmallvalues of R, the choice of k = 0 is not relevant. The equation of state p = ρ leads to − ρ = const. and R eAt A = const., i.e.a phase of exponential expansion, equivalent ∝ to adding a constant term to the right-hand side of the Einstein equation mimicking a cosmological constant: this is exactly what the inflationary paradigm proposes to overcome the paradoxes of the standard model outlined in the next Section 3. TheHubbleparameterH R˙/Randthecriticaldensityρ 3H2/8πGmakeitpossible c ≡ ≡ to rewrite (3) as k ρ (8) = 1 Ω 1, H2R2 3H2/8πG − ≡ − where Ω is the ratio of the density to the critical one Ω ρ/ρ ; since H2R2 is always c ≡ positive, the relation between the sign of k and the sign of (Ω 1) reads − (9a) k =+1 Ω>1 CLOSED ⇒ (9b) k =0 Ω=1 FLAT ⇒ (9c) k = 1 Ω<1 OPEN − ⇒ 3. – Shortcomings of the Standard Model: Horizon and Flatness Paradoxes Despite the simplicity of the Friedmann solution (in view also of the thermodynamic propertywhichcanbe studiedindetail)someparadoxesoccurwhentakingintoaccount the problem of initial conditions. The observed Universe has to match very specific physical conditions in the very early epoch, but [8] showed that the set of initial data that can evolve to a Universe similar to the present one is of zero measure and the standard model tells nothing about initial conditions. . – Flatness Let us assume that all particle species present in the early Universe have the same temperature as the photon bath, i.e.T =T and are far from any mass threshold. Then i γ the average number of degrees of freedom for the photons and fermions bath g∗ is a constant and T R−1 [24]. The averageenergy density ∝ π2 (10) ρ= g∗T4 30 substituted in the (3) reads 2 T˙ 4π3 (11) +ǫ(T)T2= Gg∗T4, T 45 ! GENERALASPECTSOFTHEDESITTERPHASE 5 where k 2π2g∗ 2/3 (12) ǫ(T) =k , ≡ R2T2 45 S (cid:20) (cid:21) and since S =R3s is the total entropy per comoving volume the entropy density reads 2π2 (13) s= g∗T3. 45 Today ρ ρ , then by taking ρ<10ρ in (8) we have c c ≃ k (14) <9H2. R2 (cid:12) (cid:12) (cid:12) (cid:12) (cid:12) (cid:12) For k = 1 (the case k = 0 is reg(cid:12)aine(cid:12)d in the limit R ), we have for today R > 1H−±1 3 109 years and T 2.7 K; from (12), the p→rese∞nt photon contribution 3 ≈ · γ ≃ to the entropy has the lower bound (15) S >3 1085 γ · expressedinunits ofthe Boltzmannconstantk =1.3806 10−16erg/K.The relativistic B · particles present today together with photons are the three neutrino species and their contribution to the total entropy is of the same order of magnitude (16) S >1086, and finally with (17) ǫ <10−58g∗2/3 | | we gain (18) ρ−ρc = 45 m2P ǫ <3 1059g∗−1/3 mP 2 , ρ 4π3g∗T2| | · T (cid:12) (cid:12) (cid:12) (cid:12) (cid:16) (cid:17) (cid:12) (cid:12) where m is the Pl(cid:12)anck ma(cid:12)ss (~c/G)1/2 =2.1768 10−5g =1.2211 1019 GeV. P · · When taking T = 1017 GeV, all species in the standard model of particle interactions – 8 gluons, W±, Z0, 3 generations of quark and leptons – are relevant and relativistic: then g∗ 100 and finally ≈ ρ ρ (19) − c <10−55. ρ (cid:12) (cid:12)T=1017GeV (cid:12) (cid:12) (cid:12) (cid:12) AflatUniversetodayrequires(cid:12)Ωofthe(cid:12) originaloneclosetounityuptoapartin1055. A little displacement from flatness at the beginning – for example 10−30 – would produce an actual Universe either very open or very closed, so that Ω = 1 is a very unstable condition: this is the flatness problem. 6 GIOVANNIIMPONENTE and GIOVANNIMONTANI The natural time scale for cosmology is the Planck time ( 10−43 sec): in a time of ∼ this order a typical closed Universe would reach the maximum size while an open one wouldbecomecurvaturedominated. TheactualUniversehassurvived1060 Plancktimes without neither recollapsing nor becoming curvature dominated. . – Horizon Neglecting the ǫT2 term, (11) is solved by 4π3 −1/2 m (20) T2 = g∗ P . 45 2t (cid:18) (cid:19) A light signal emitted at t=0 travelled during a time t the physical distance t dt′ (21) l(t)=R(t) =2t R(t′) Z0 in a radiation-dominated Universe with R t1/2, measuring the physical horizon size, ∝ i.e.the linear size of the greatest region causally connected at time t. The distance (21) has to be comparedwith the radius L(t) of the regionwhich will evolve in our currently observed region of the Universe. Conservation of entropy for s T3 gives ∝ 1/3 s 0 (22) L(t)= L , 0 s(t) (cid:18) (cid:19) where s is the present entropy density and L H−1 1010 years is the radius of the 0 0 ∼ ≃ currently observed region of the Universe. The ratio of the volumes provides l3 m 3 (23) =4 10−89g −1/2 P L3 · ∗ T (cid:16) (cid:17) and, as above, for g∗ 100 and T 1017GeV we obtain ∼ ∼ l3 (24) 10−83. L3 ∼ (cid:12)T=1017GeV (cid:12) (cid:12) (cid:12) The currentlyobservableUniverseis composedofseveralregionswhichhavenot beenin causal contact for the most part of their evolution, preventing an explanation about the present days Universe smoothness. In particular, the spectrum of the CMBR is uniform upto10−4. Moreover,wehaveatthetimeofrecombination,i.e.whenthephotonsofthe CMBRlastscattered,theratiol3/L3 105: thepresentHubblevolumeconsistsofabout ∼ 105 causally disconnected regionsat recombinationand no processcould have smoothed out the temperature differences between these regions without violating causality. The particle horizon at recombination subtends an angle of only 0.8◦ in the sky today, while the CMBR is uniform across the sky. GENERALASPECTSOFTHEDESITTERPHASE 7 4. – The Inflationary Paradigm The basic ideas of the theory of inflation rely firstly on the original work by [11], i.e.the old inflation, which provides a phase in the Universe evolution of exponential expansion;thenformulationofnew inflation by[28]introducedthe slow-rollingphasein inflationary dynamics; finally, many models have sprung from the original theory. In [11] is described a scenario capable of avoiding the horizon and flatness problems: both paradoxes would disappear dropping the assumption of adiabaticity and in such a case the entropy per comoving volume s would be related as (25) s =Z3s 0 early where s and s refer to the values at presentand at very early times, for example at 0 early T =T =1017 GeV, and Z is some large factor. 0 With this in mind, the right-handside of (18) is multiplied by a factor Z2 and the value of ρ ρ /ρ would be of the order of unity if c | − | (26) Z >3 1027, · getting rid of the flatness problem. The right-hand side of (22) is multiplied by Z−1: for any given temperature, the length scale of the early Universe is smaller by a factor Z than previously evaluated, and for Z sufficiently large the initial region which has evolved in our observed one would have been smaller than the horizon size at that time. Let us evaluate Z considering that the right-hand side of (22) is multiplied by Z3: if (27) Z >5 1027, · then the horizon problem disappears. Making some ad hoc assumptions, the model accounts for the horizonand flatness paradoxes while a suitable theory needs a physical process capable of such a large entropy production. Asimplesolutionreliesontheassumptionthatatveryearlytimestheenergy densityoftheUniversewasdominatedbyascalarfieldφ(~x,t),i.e.ρ =ρ +ρ +ρ +... φ rad mat with ρ ρ and hence ρ ρ . φ rad,mat,etc φ ≫ ≃ The quantum field theory Lagrangiandensity for such a field is ∂µφ∂ φ µ (28) = V(φ) L 2 − and the corresponding stress-energy tensor (29) Tµ =∂µφ∂ φ δµ ν ν −L ν by which for a spatially homogeneous and isotropic Universe the form of a perfect fluid leads to [26] φ˙2 1 (30a) ρ = +V(φ)+ 2φ, φ 2 2R2∇ φ˙2 1 (30b) p = V(φ) 2φ. φ 2 − − 6R2∇ 8 GIOVANNIIMPONENTE and GIOVANNIMONTANI Spatial homogeneitywould induce a slow variationofφ with position, hence the spatial gradients are negligible and the ratio ω =p/ρ reads φ˙2 +V(φ) pφ 2 (31) . ρφ ≃ φ˙2 V(φ) 2 − If the field is at a minimum of the potential, φ˙ = 0, and (31) becomes an equation of state (32) p = ρ φ φ − giving rise to a phase of exponential growth of R eHt, the inflationary or de Sitter ∝ phase. The field evolution is very different when in vacuum or in a thermal bath and such a coupling can be summarized by adding a term (1/2)λT2φ2 to the Lagrangian. The − potential V(φ) is replaced by the finite-temperature effective potential 1 (33) V (φ)=V(φ)+ λT2φ2. T 2 In the old inflation, V(φ) appearing in (33) has the form of a Georgi-Glashow SU(5) theory 1 1 1 (34) V(φ)= φ4 (A+B)φ3+ ABφ2 4 − 3 2 with A > 2B > 0, and possesses a local minimum at φ = 0 and a global minimum at φ = A, separated by a barrier with a maximum at φ = B. The temperature-dependent term (1/2)λT2φ2 leaves the local minimum unchanged and raises the global minimum − as well as the maximum – the former by a larger amount than the latter. At sufficiently high temperature V (φ) has only one global minimum at φ = 0 and, as T long as T decreases, a second minimum develops at φ = σ(T), with V(0) < V(σ), and φ=0 is still the true minimum of the potential. At a certain critical temperature T the two minima are exactly degenerate as V(0) = c V(σ) and at temperatures below T , V(σ)T , the field φ i c is trapped in the minimum at φ=0 (false vacuum) with constant energy density, given by (30a) V (0) T4. The temperature lowers with the Universe expansion up to the T ≃ C criticalvalue T : the scalarfield begins dominating the Universe anda secondminimum c of the potential develops at φ=σ. The inflationary phase is characterized by (35) R(t) eHt ∝ where the Hubble parameterH R˙/R is a constant; φ=0 becomes a metastable state, ≡ since there exists a more energetically favourable one at φ=σ (true vacuum). GENERALASPECTSOFTHEDESITTERPHASE 9 The potential barrier lying within cannot prevent a non-vanishing probability per unit time that the field performs a first order phase transition via quantum tunnelling to the true vacuum state, proceeding along by bubble nucleation: bubbles of the true vacuum phase are created expanding outward at the speed of light in the surrounding “sea” of false vacuum, until all the Universe has undergone the phase transition. If the rate of bubble nucleation is low, the time to complete the phase transition can be verylongifcomparedtotheexpansiontimescale: whenthetransitionends,theUniverse has cooled to a temperature T many orders of magnitude lower than T . f c On the contrary, when approaching the true vacuum, the field φ begins to oscillate aroundthispositiononatime scaleshortifcomparedto the expansionone,releasingall its vacuum energy in the form of φ-particles, the quanta of the φ field. The oscillations are damped by particles decay and, when the corresponding products thermalize, the Universe is reheated to a temperature T of the order of T . This represents the release r c of the latent heat associated with the phase transition after which the scalar field is no longer the dominant component of the Universe: inflation comes to an end and the standard FLRW cosmology is recovered. The non-adiabatic reheating process releases an enormous amount of entropy whose density is increased by a factor of the order of (T /T )3 (T /T )3, while R remains r f c f ≃ nearly constant and the entropy is increased by a factor (T /T )3 too. c f FlatnessandhorizonparadoxesaresolvediftheUniversesuper-coolsof28ormoreorders ofmagnitudeduringinflation. Evenifthislooksverydifficulttoachieve,itisenoughthat the transition takes place in about a hundred Hubble times: the inflationary expansion is adiabatic and the entropy density s R−3. Since s T3, then T R−1 e−Ht and ∼ ∝ ∝ ∝ finally T (36) c =eH∆t, T f where ∆t is the duration of the de Sitter phase. From the requirement (T /T ) > 1028 c f it follows (37) ∆t>50H−1. The critical temperature is estimated to be of order of the energy typically involved for Grand Unification Theory spontaneous symmetry-breaking phase transition, 1014 GeV, and 8πG 8πG T4 (38) H2 = ρ= V(0) c H−1 10−34s. 3 3 ≃ m2 → ≃ P The inflation removes the discussed paradoxes if the transition takes place in a time t 10−32s, nevertheless leaving some open problems regarding its dynamics: ≈ - a - intheoldscenarioinflationneverends,duetosmallnessofthetunnellingtransition rate, so that the nucleation of true vacuum bubbles is rare; - b - the energy released during the reheating is stored in the bubbles kinetic energy so that the reheating proceeds via bubble collisions which remain too rare due to 10 GIOVANNIIMPONENTE and GIOVANNIMONTANI low transition rate to produce sufficient reheating: the phase transition is never completed; - c - such a discontinuous process of bubble nucleation via quantum tunnelling should produce a lot of inhomogeneities which aren’t actually observed. . 41. New Inflation: the Slow Rolling Model. – In 1982, both [28] and [1] proposed a variant of Guth’s model, now referred to as new inflation or slow-rolling inflation, in order to avoid the shortcomings of the old inflation. Their original idea considered a different mechanism of symmetry breaking, the so-called Coleman-Weinberg (CW) one. The potential of the CW model for a gauge boson field with a vanishing mass reads as Bσ4 φ2 1 (39) V(φ)= +Bφ4 ln , 2 σ2 − 2 (cid:20) (cid:18) (cid:19) (cid:21) where B is connected to the fundamental constantsof the theory andis 10−3, while σ ≃ gives the energy associated with the symmetry breaking process and is 2 1015 GeV. ≃ · The finite-temperature effective potential is obtained as above in (33) by adding a term of the form (1/2)λT2φ2. Expression (39) can be generalized by adding a mass term of the form (1/2)m2φ2. Defining a temperature-dependent “mass” − (40) m m2+λT2, T ≡ − p the temperature-dependent potential becomes Bσ4 φ2 1 1 (41) V (φ)= +Bφ4 ln + m2φ2. T 2 σ2 − 2 2 T (cid:20) (cid:18) (cid:19) (cid:21) The quantity m 2 can be used to parametrize the potential (41): T 1. whenm 2 >0,the pointφ=0is aminimumofthe potential,while whenm2 <0 T T it is a maximum; 4σ2 2. when m 2 < 1.5σ2, a second minimum develops for some φ¯ > 0; initially T e ≃ this minimum is higher than the one at 0, but when m becomes lower than a T certain value m∗ (0

General Aspects of the de Sitter phase

Similar General Aspects of the de Sitter phase