arXiv:hep-ph/0603244v2 18 Sep 2007 ffiinl rnfrigteeeg,peetn osntre- a not in despite does sult However, preheating energy, [9]. the state(s) transferring efficiently final to the condensate of homogeneous modes to the non-zero oscillations of from of degrees energy dozens two the bosonic about transfer takes are only products It final freedom. particularly the is when it and efficient preheating called 9]. is [8, mechanism non-perturbatively This particles create can inflaton a the in true is 7]. 6, which [2, decayed, scenarios have non-supersymmetric quanta com- inflaton after reaches immediately the plasma equilibrium all the chemical that and assumed kinetic is plete is it sec- couples Usually matter which singlet) the tor. gauge small, to a operators sufficiently non-renormalizable (being are through inflaton fields the if SM cou- justifiable inflaton the many the to over that place plings requires takes This fields inflaton the oscillations. the to inflaton to transfer coupled energy the are non- [5]. if which the valid initially of is considered treatment decay were The one-particle quanta Only inflaton relativistic reheating. of part [4]. (BBN) nucleosynthesis bang big the successful guaran- is a would freedom tee which of phenomenon degrees important infla- SM most the single transferring into Therefore density energy [3]. ton or extensions SM within minimal explained be its cannot matter the to couplings adMdl(M hre.Otnifltoayprdg is Stan- paradigm a inflationary with carry realized Often apriori charges. not (SM) does Model which dard sector, inflaton the to twsas one u httechrn siltosof oscillations coherent the that out pointed also was It relevant most the and first the is decay inflaton The eetn fe naincnet h bevbesector observable the connects inflation after Reheating complete 2 rsrigmse oteifltndcypout hog the VEVs through for products preheating decay blocks inflaton kinematically U the which early to the in masses (VEV) value preserving expectation vacuum slepton large of a combinations is towards invariant reason gauge by The slep classified supersymmetry. s and tial, within that squarks as unlikely expect such is would scalars preheating one new Naively introduces it small. dec as very heating usually are inflaton fields the other case to non-supersymmetric a In metry. o oeswt eksaesprymty hsgenerically this flat supersymmetry, the scale once H weak and with oscillating, start models For directions flat the after o loigteaxlayfil eoetachyonic. become field e auxiliary can the directions allowing flat not supersymmetric non-p Finally through leptogenesis (s)neutrinos. non-thermal no and matter, dark fdneossprymti eis(atclrygraviti (particularly importa relics of the number supersymmetric to are decay dangerous There inflaton of the supersymmetry. where within analysis properly first knowledge our to eateto hsc n srnm,MMse University, McMaster Astronomy, and Physics of Department oiae yRf 1 ] eaayehwteifltndcyreh decay inflaton the how analyze we 2], [1, Refs by Motivated Mguesinglet gauge SM ≃ .INTRODUCTION I. ea fteifltn nsm ae the cases some In inflaton. the of decay ` 10 1 − eiee nttt o hoeia hsc,Wtro,ON Waterloo, Physics, Theoretical for Institute Perimeter 3 − eetn nsprymti ihsaeinflation scale high supersymmetric in Reheating 10 − 1 3 ´ nao hs rgnand origin whose inflaton ODT,Bedmvj1,Cpnae-10 Denmark. Copenhagen-2100, Blegdamsvej-17, NORDITA, e,a hc ieteifltndcy nteprubtv reg perturbative the in decays inflaton the time which at TeV, oze Allahverdi Rouzbeh 1 , 2 n npmMazumdar Anupam and > fitrcin eitdb the by mediated efficiency the interactions of of because quickly very thermalize products was question [2]. last Ref. very how in This addressed is recently question thermalize. next they super- the do minimal quickly Then the (MSSM). of SM freedom into symmetric of Particularly, ask degrees SUSY. to relativistic pertinent within the large then decays is inflaton the it be- the given inflaton, how physics the Therefore, of SM density the scale. energy electroweak of the extensions yond studied widely most 21]. [10, solitons non-topological formation and the topological and [20], of baryogenesis and source non-thermal leptogenesis 19], for [18, non-Gaussianity large genera- of [17], tion isocurvature waves gravity large amplifying [15], [16], moduli of from perturbations and production 14] ranging 9], 13, phenomena [12, [8, gravitinos particles physical which of rich possibility production to a non-thermal rise remains give cosmolog- preheating can realistic [11], potentially model any ical of ingredient sential to plasma. challenging preheated very the of is thermalization it [10]. understand situations evaporation theses surface of non-topological through Irrespective form decays which to solitons fragments condensate inflaton h ea fteifltn hc eut navr low very a in results after which equilibrium i.e. inflaton, quasi-thermal temperature, the a reheat of of decay Universe phase in the The is 2], a thermalization SUSY. [1, in of within Refs. loiters process slow In the painstakingly that general width. mainly out decay is pointed inflaton temperature we reheat the the by Therefore governed [6]. SM the 10 13 naNnSS aei skonta h nao decay inflaton the that known is it case Non-SUSY a In for framework the provides (SUSY) Supersymmetry es- an is reheating perturbative of epoch an Although o) orsnn xiaino superheavy of excitation resonant no nos), e.Tedcywl eoealwdonly allowed become will decay The GeV. n qakfils r eeial displaced generically are fields, squark and e pi yrdiflto l oehrby together all inflation hybrid spoil ven ieto E ssffiinl redshifted. sufficiently is VEV direction os ntecnrr,w on u that out point we contrary, the On tons. rubtv raino h right-handed the of creation erturbative htfltdrcin ntesaa poten- scalar the in directions flat that y i rhaiguls t couplings its unless preheating via ays apn taHbl xaso rate: expansion Hubble a at happens prymtyehne ooi pre- bosonic enhances upersymmetry ies.Te nuesupersymmetry induce They niverse. tcneune:n overproduction no consequences: nt tnadMdlYkw couplings, Yukawa Model Standard tnadMdlprilsi treated is particles Model Standard asteUies ihnsupersym- within Universe the eats aitn N 8 M,Canada. 4M1, L8S ON, Hamilton, 2 Y,Canada. 2Y5, N2L , 3 T R O ∼ asesguebosons gauge massless m.Ti is This ime. (10 3 − NORDITA-2006-5 10 7 e.The GeV. ) of 2
final reheat temperature does not depend on the decay II. INFLATON DECAY IN A NON-SUSY CASE width of the inflaton rather on a thermalization time scale which depends on the vacuum expectation value (VEV) First let us briefly review the initial stage of the infla- of the squarks and sleptons. ton decay which is typically non-perturbative, i.e. pre- heating, in a non-SUSY case. Our focus is on bosonic What was not addressed in Ref. [2] is the initial phase preheating which acts most efficiently in transferring the of the inflaton decay. In this paper we will fill up that energy density from the inflaton oscillations. We consider gap. We would like to know how the inflaton decays: models of large field inflation, such as chaotic inflation, whether perturbatively or non perturbatively. This simple for which bosonic preheating is most pronounced. The question is so relevant that depending on the nature of relevant renormalizable couplings between the inflaton φ the inflaton decay there would be different consequences and a scalar field χ will read from the following potential: all together. For instance production of baryons, cold dark matter, magneto-genesis, electroweak baryogenesis, 1 V = m2 φ2 + σφχ2 + h2φ2χ2 + λχ4 , (1) production of dangerous relics and their abundances de- 2 φ pend on the nature of the primordial plasma. Therefore this is an important topic which relates the early Uni- where we have considered φ and χ to be real. Here σ verse physics to phenomenology. is a coupling which has a [mass] dimension. The only scalar field in the SM is the Higgs doublet. Therefore in a realistic case χ denotes the real and imaginary parts An important fact is the presence of flat directions of the Higgs components 1. Note that the cubic interac- along which the scalar potential identically vanishes in tion term is required for a complete inflaton decay. The the limit of exact SUSY. In the MSSM alone there are quartic self-coupling of χ is required to bound the poten- nearly 300 flat directions [22], which are made up of tial from below along the χ direction. The dimensionless gauge invariant combinations of squarks, sleptons and couplings σ/m and h (as well as λ) are not related to Higgses. These are none but the simplest examples of φ each other, hence either of the cubic or the quartic terms moduli near points of enhanced symmetry. During infla- can dominate at the beginning of inflaton oscillations (i.e. tion more than one MSSM flat directions (orthogonal in when the Hubble expansion rate is H(t) m and the flavor basis) [23] are expected to develop large VEVS, for ≃ φ amplitude of oscillations is φˆ (M )). a review, see [24]. ∼ O P σ h2M : In this regime the h2φ2χ2 term is dom- A large VEV of the MSSM flat directions, during P • inant≪ at the beginning of the inflaton oscillations. and after inflation, spontaneously breaks the SM gauge This case has been studied in detail in first two ref- group, and gives masses to the gauge bosons and gaug- erences of [9]. For a nominal value of the inflaton inos similar to the Higgs mechanism. Many of the flat 13 mass, mφ = 10 GeV, non-perturbative χ produc- directions break the entire SM gauge group [25]. The 1/2 flat direction VEV also induces large SUSY preserving tion with a physical momentum, k < hmφφˆ , masses to (s)quarks, (s)leptons and Higgs/Higgsino fields takes place if h> 10−6. Particle production∼ is par- during and after inflation. As we will see, such large ticularly efficient if h > 3 10−4, and results in masses kinematically prohibit non-perturbative inflaton an explosive transfer of energy× to χ quanta which decay into MSSM fields. The initial stage of inflaton os- ends when re-scatterings destroy the inflaton con- cillations thus produces no significant fraction of MSSM densate 2. The whole process happens over a time particles [2]. The decay will occur much later after the −1 scale 150mφ , which depends logarithmically on flat direction oscillations have started and their VEV has 3 ∼ been sufficiently redshifted. Such a delayed decay of the h . inflaton is typically in the perturbative regime.
Rest of the paper is organized as follows. In Section 1 Since the SM fermions are chiral, the inflaton can only couple to II we briefly discuss inflaton decay in a non-SUSY case. them through dimension-5 operators. The same holds for cou- We highlight inflaton couplings to the MSSM fields in pling to gauge bosons where the inflaton is coupled to gauge Section III, and MSSM flat direction couplings to the field strengths. Such couplings are non-renormalizable and sup- inflaton decay products in Section IV. We then discuss pressed compared to those in Eq. (1), and hence negligible (we the flat direction dynamics during and after inflation in consider the scale of non-renormalizable operator is governed by MP). Section V. In Section VI we explain how the VEV of flat 2 Further note that the h2φ2χ2 term does not produce any signif- directions prevent non-perturbative decay of the inflaton. icant non-Gaussianity if h> 10−5, see the first reference of [18]. We illustrate in Section VII how the inflaton eventually 3 In a non-SUSY case efficient preheating happens over a narrow −4 −3 2 2 2 decays perturbatively, and discuss various cosmological window 3 × 10 ≤ h ≤ 10 . The reason is that the h φ χ term yields a quartic self-coupling for the inflaton at a one-loop consequences in Section VIII. Finally we briefly mention level which is constrained by the COBE normalization of the our conclusion. We have added appendices discussing density perturbations [24, 29]. However in SUSY this correction some minute details for the paper to be self-contained. is canceled out by that from fermionic partner of χ, so in princi- 3
2 2 σ h MP: In this regime the cubic term σφχ importance of gauge invariance was first highlighted in • dominates.≫ This case was recently considered in Refs. [2, 7, 32, 33] 6. Refs. [30, 31], where the the χ field becomes tachy- In almost all known F and/or D-term models of infla- onic during half of each oscillation. For σ > 2 7 tion the inflaton, φ, is considered to be an absolute gauge mφ/MP (which amounts to σ > 10 GeV for singlet. Then the main question arises how the inflaton 13 mφ = 10 GeV) this tachyonic instability trans- couples to the matter. This is one of the most pertinent fers energy from the oscillating condensate very ef- issues which connects inflation to a hot big bang cosmol- ficiently to the χ quanta with a physical momentum ogy. 1/2 k < σφˆ . Particle production ceases when the First note the field content of MSSM which is governed back-reaction∼ from χ self-coupling induces a mass- by the following superpotential: squared > σφˆ. Depending on the size of λ, most of the energy∼ density may or may not be in χ quanta W = λ QH u + λ QH d + λ LH e + µH H , 4 MSSM u u d d e d u d by the time backreaction becomes important [31] . (2) where Hu, Hd, Q, L, u, d, e in Eq. (2) are chiral super- Couple of points to note here. In the borderline regime fields representing the two Higgs fields (and their Hig- σ h2M , the cubic and quartic interaction terms ∼ P gsino partners), left-handed (LH) (s)quark doublets, are comparable. The inflaton decay happens due to a right-handed (RH) up- and down-type (s)quarks, LH combination of resonant and tachyonic instabilities. If (s)lepton doublets and RH (s)leptons respectively. The h m /M and σ m2 /M , the inflaton decays per- ≪ φ P ≪ φ P dimensionless Yukawa couplings λu, λd, λe are 3 3 ma- turbatively via the cubic interaction term. However this trices in the flavor space, and we have omitted the× gauge −6 requires very small couplings: h, (σ/mφ) < 10 . There- and flavor indices. The last term is the µ term, which is fore, unless the inflaton is only gravitationally coupled to a supersymmetric version of the SM Higgs boson mass. other fields, the initial stage of its decay will be generi- cally non-perturbative. There exist two gauge-invariant combinations of only The plasma from the non-perturbative inflaton decay two superfields: eventually reaches full thermal equilibrium, though, at time scales much longer than that of preheating itself [26, HuHd , HuL. (3) 27, 28]. The occupation number of particles is fk 1 in the meantime. This implies that dangerous relics≫ (such as gravitino and moduli) can be produced much The combinations which include three superfields are: more copiously in the aftermath of preheating than in full thermal equilibrium [1, 15, 28]. This is a negative aspect HuQu , HdQd , HdLe , QLd , udd , LLe. (4) of an initial stage of preheating. One usually seeks a late stage of entropy release, in order to dilute the excess of relics. As we shall show, supersymmetry naturally SUSY together with gauge symmetry requires that the provides us a tool to undo preheating completely. inflaton superfield be coupled to these combinations 7. The terms ΦHuHd and ΦHuL have dimension four, and hence are renormalizable. On the other hand, the interac- III. INFLATON COUPLINGS TO MATTER IN tion terms that couple the inflaton to the combinations in SUSY Eq. (4) have dimension five and are non-renormalizable. In following we focus on renormalizable interactions of Inflaton couplings to (MS)SM fields is of utmost im- the inflaton with matter which play the dominant role in 8 portance for (p)reheating. In all the relevant papers, for its decay . instance see Refs. [8, 9], inflaton couplings to matter has not been dealt with carefully. Only toy models have been considered which have no relevance to SM physics 5. The the inflaton couples to SM fermions through non-renormalizable dimension 5 operators, and therefore preheating into SM fermion is unlikely. 6 In Refs. [34] the inflaton belonged to a gauge sector which can ple one could expect a rather broader range of parameter space carry SM charges based on the ideas of assisted inflation [35]. In within SUSY. which case the inflaton couplings to matter are governed by the 4 In Ref. [31], the authors attempted to motivate the cubic cou- usual Yukawas. However it is hard to construct realistic models pling from SUSY. However they missed vital ingredients which without gauge singlets. exist in a realistic case, such as the strength of the SM couplings, 7 It is possible that the inflaton mainly decays to another singlet contributions from SUSY flat directions, etc. In this paper we (for example, the RH neutrinos) superfield, see the discussion in wish to note that neither the couplings nor the interactions are section XI B taken arbitrarily. 8 We note that terms representing gauge-invariant coupling of the 5 Fermionic preheating has been discussed in Refs. [44], but the inflaton to the gauge fields and gauginos are also of dimension importance of SM gauge invariance was grossly neglected. In fact five, and hence preheating into them will be suppressed. 4
A. Two choices of renormalizable couplings irrelevant during inflaton oscillations. We have neglected the inflaton coupling to the fermionic partners of χ as The simplest case is when the inflaton is coupled to we focus on the bosonic preheating here. However our matter via superpotential terms of the form: analysis will follow similarly to the fermionic case and the same conclusions hold for fermionic preheating too. 2gΦHuHd , 2gΦHuL . (5) In addition to the terms in Eq. (9) there are also 2 2 2 2 2 2 the self- and-cross-couplings, g /4 χ1 χ2 +αχ1χ2, where g can be as large as (1). The factor of 2 − O arising from the superpotential andD -terms respectively as we shall see, leads to convenience in field redefini- (α is a gauge fine structure constant). Therefore even in tions. Besides the SM gauge group the MSSM La- the simplest SUSY set up the scalar potential is more grangian is also invariant under a discrete Z2 symme- involved than the non-SUSY case given in Eq. (1), which try namely “R-parity”. This symmetry assigns the num- can alter the picture of preheating presented in the lit- 3B+L+2S ber R = ( 1) to the component fields where erature [37, 38]. Note however that these terms become − B, L, S denote the baryon number, lepton number and important after particle production has started. Here we spin respectively. This amounts to +1 for the SM fields focus on the terms in Eq. (9) which are relevant for par- and 1 for their supersymmetric partners. As a result ticle creation from the very beginning of the oscillations. − the lightest supersymmetric particle (LSP) will be stable A remarkable feature in Eq. (9) is that SUSY naturally and can account for dark matter in the Universe. This relates the strength of cubic φχ2 and quartic φ2χ2 inter- is one of the most remarkable cosmological features of actions. We re-emphasize that the cubic term is required MSSM. for complete decay of the inflaton field. This is a natural consequence of SUSY which holds so long as the inflaton mass is larger than the soft SUSY breaking masses. B. Preserving R-parity
Preserving R-parity at the renormalizable level further IV. FLAT DIRECTION COUPLINGS TO INFLATON DECAY PRODUCTS constrains inflaton couplings to matter. Note that HuHd is assigned +1 under R-parity, while HuL has the oppo- site assignment 1. Therefore only one of the couplings Consider a MSSM flat direction, ϕ, with the corre- − in Eq. (5) preserves R-parity: ΦHuHd if RΦ = +1, and sponding superfield denoted by ϕ (only for flat directions ΦHuL if RΦ = 1 (such as models where the RH sneu- we are denoting the superfield and the field with the same trino plays the role− of the inflaton [36]). Therefore the notation). For a brief discussion on MSSM flat directions, renormalizable inflaton coupling to matter can be repre- see Appendices XI A, XI C. Note that the ϕ and X super- sented as fields are linear combinations of the MSSM superfields, see Eq. (8), and hence are coupled through the MSSM 2gΦHuΨ (6) superpotential in Eq. (2). The couplings are nothing but the (MS)SM Yukawas. Then the MSSM superpotential where can be recast in the following form:
Ψ = Hu if RΦ = +1 , W λ H ϕΣ + λ ΨϕΣ + ..., (10) ⊃ 1 u 1 2 2 Ψ = L if RΦ = 1 . (7) 9 − where Σ1,2 are some MSSM superfields . For example Taking into account of the inflaton superpotential mass consider the case where ϕ is a flat direction classified by term: (mφ/2) ΦΦ, and after defining the udd monomial, and Ψ = Hd. In this case Σ1,2 are Q superfields and λ1,2 correspond to λu and λd respectively, (Hu Ψ) see Eq. (2). After using Eq. (8) we find: X1,2 = ± , (8) √2 λ1 λ2 W XϕΣ1 + XϕΣ2 . (11) and with the help of Eqs. (24,25), see appendix XI A, ⊃ √2 √2 we find the renormalizable part of the potential which is relevant for the inflaton decay into MSSM scalars is given This results in: by: λ2 + λ2 1/2 V λ2 ϕ 2χ2 , λ 1 2 , (12) 1 2 2 2 2 2 1 2 ⊃ | | ≡ 8 V m φ + g φ χ gmφφχ , (9) ⊃ 2 φ ± √2 where we have again considered the real part of χ. where χ denotes the scalar component of X1,2 superfields, and we have only considered the real parts of the inflaton, φ, and χ field. Further note that the cubic interaction 9 term appears with different signs for χ1 and χ2, but this is Note that Σ1 6= Ψ and Σ2 6= Hu, since ϕ is a non-gauge-singlet. 5
Let us determine the strength of flat direction cou- accumulate (in a coherent state) along ϕ and form a pling to χ, which is denoted in Eq. (12) by λ. Note that condensate with a large VEV, ϕ0. Because inflation the first generation of (s)leptons and (s)quarks have a smoothes out all gradients, only the homogeneous con- Yukawa coupling (10−6 10−5), while the rest of densate mode survives. However, the zero point fluctua- the SM Yukawa couplings∼ O are− 10−3. The MSSM flat tions of the condensate impart a small, and in inflation- directions can be grouped in 6≥ categories mentioned in ary models a calculable, spectrum of perturbations on Appendix XI C, out of which the condensate [24]. If the higher-order superpotential term is forbidden, Only 11 directions: 3 udds, 6 QdLs, 1 LLddd and due to an R symmetry (or a set of R symmetries) [42], • LLe − − 1 have couplings to MSSM particles/sparticles then we naturally have, ϕ0 MP [39]. On the other such that λ< (10−5), ∼ O hand, ϕ0 MP will be possible if non-renormalizable superpotential≪ terms are allowed. As shown in [22], in the Rest of the flat directions have λ 3 10−4. • ≥ × absence of any R-symmetry, all the MSSM flat directions are lifted by higher-order terms with n 9. If a flat direction is lifted at the superpotential level≤ n, the VEV V. FLAT DIRECTION POTENTIAL that it acquires during inflation cannot exceed:
n−3 1/(n−2) The flat directions are lifted by soft SUSY break- ϕ0 HIM , (14) ing mass term, m (TeV), Hubble induced cor- ∼ 0 ∼ O rections and superpotential corrections of type: W where HI is the expansion rate of the Universe in the n n−3 ∼ λnϕ /nM [39] inflationary epoch. After inflation, H(t) t−1, the flat direction stays 2(n−1) at a relatively larger VEV∝ due to large Hubble friction 2 2 2 2 ϕ V m0 + cH H ϕ + λn | | , (13) term, note that the Hubble expansion rate gradually ⊃ | | M 2(n−3) decreases but it is still large compared to m0. When with n 4. Here M is the scale of new physics which in- H(t) m0, the condensate along the flat direction starts duces the≥ non-renormalizable terms, typically the Planck oscillating≃ around the origin with an initial amplitude − 1/(n−2) scale M = MP or the grand unification scale M = MGUT. ϕ m M n 3 . From then on ϕ is redshifted in ∼ 0 P h i Note that cH can have either sign. If cH > 1, the flat by the Hubble expansion H for matter dominated and direction mass is > H. It therefore settles at∼ the origin H3/4 for radiation dominated∝ Universe. during inflation and remains there 10 . Since ϕ =0 at ∝ all times, the flat direction will have no interestingh i con- sequences in this case. However there is a large class of VI. NO PREHEATING IN SUSY theories which predicts c < 0 and also c +1. Neg- H H ≪ ative cH may arise naturally if the inflaton and MSSM In order to understand the preheating dynamics it is flat directions have positive higher order couplings in important to take into account of χ coupling to the infla- † † the K¨ahler potential, i.e. Φ Φϕ ϕ [39], such that all ton φ, as well as to the MSSM flat direction, ϕ, which is the eigenvalues of the K¨ahler matrix is positive definite. displaced away from its minimum (towards large VEVs) There is no symmetry which prohibits such couplings. during inflation. The governing potential can be obtained Moreover string theory, which we believe will provide from Eqs. (9,12) the true low energy effective theory, also generically pre- dicts no-scale type K¨ahler potential based on Heisenberg 1 2 2 2 2 2 g 2 2 2 2 V = m φ + g φ χ + mφφχ + λ ϕ χ . (15) symmetry [40], which at tree level gives no correction to 2 φ √2 the flat direction mass, i.e. cH = 0. However a Hub- ble induced mass term is generated at a loop level, be- As mentioned in the previous section, we generically have λ 3 10−4, and g can be as large as (1). cause MSSM superpotential (as well as D-terms) breaks ≥ × ∼ O the Heisenberg symmetry, which induces a calculable but After mode decomposition of the field χ, the energy small contribution, i.e. c 10−2 [41]. Moreover, even of the mode with momentum k, denoted by χk, is given H ≤ by [9]: starting with cH > 0 at a high scale, it is possible that cH quickly changes sign due to loop corrections from large 1/2 2 2 2 2 2 Hubble-induced SUSY breaking terms [43]. ωk = k +2g φ + √2gmφ φ +2λ ϕ . (16) h i h i h i In the absence of cH > 1, the flat direction remains flat during inflation as∼ the Hubble expansion rate is We have frozen the expansion of the Universe. Including HI m0. Therefore quantum fluctuations are free to the expansion will not change our conclusions anyway. ≫ First note that during inflation the inflaton VEV is large, −6 i.e. φ >MP. Therefore if g > 10 the inflaton induces a largeh i mass g φ >H for χ during inflation. h i I 10 This has a similar origin as a supergravity inflationary η-problem, As a result, χ, quickly settles down to the minimum, see [3]. i.e. χ = 0, even if it is initially displaced, and remains h i 6
there. Therefore, ϕ, does not receive any mass correc- It turns out from Eqs. (16,17) that the energy of tions from its coupling to χ during inflation. Note that mode χk changes adiabatically at all times if Eq. (18) the VEV of the flat direction, ϕ, induces a large mass, is satisfied at the beginning of inflaton oscillations, i.e. 12 λϕ0, to the χ field during inflation. H = mφ . Hence there will be no resonant production At the end of inflation, i.e. when H(t) m , the in- of χ quanta, provided that ≃ φ flaton starts oscillating with frequency mφ and an initial −6 −1 1/2 amplitude (MP). Note that for g > 10 the quartic ϕ0 > λ (gMPmφ) . (19) inflaton couplingO takes over the cubic one in Eq. (15). This surmounts to a kinematical blocking of preheating In the interval m0 H(t) mφ the flat direction VEV slides very slowly≤ because≤ of the under damped by inducing a piece (which is virtually constant at time motion due to large Hubble friction term, the flat di- scales of interest) to the mass of inflaton decay products rection effectively slow rolls. Non-perturbative produc- due to their couplings to a flat direction which has a large tion of χ quanta will occur if there is a non-adiabatic VEV. 2 Once gφˆ mφ, the Hubble expansion rate becomes: time-variation in the energy, i.e. that dωk/dt > ωk. The 2∼ 2 inflaton oscillations result in a time-varying contribution∼ H(t) mφ/gMP, and the cubic interaction term φχ ∼ 13 to ωk, while the flat direction coupling to χ yields a vir- takes over the quartic one. For mφ = 10 GeV and −6 tually constant piece. g > 10 this happens when the expansion rate is still 1 TeV. Eq. (19) implies that λ2ϕ2 m2 , and hence Obviously the piece induced by the flat direction VEV ≫ 0 ≫ φ weakens the non-adiabaticity condition 11. Indeed time- the flat direction VEV totally dominates the mass of χ variation of ωk will be adiabatic at all times when the cubic term has taken over. Therefore there will be no tachyonic instability in the mass of χ which would dω k <ω2 , (17) otherwise occur during half of each inflaton oscillation dt k and could lead to efficient particle production [31]. provided that Now let us find the range of VEVs for which a MSSM flat direction can satisfy the condition given in Eq. (19) 2 2 λ ϕ >gφmˆ φ , (18) and shut-off preheating. We choose a nominal value of h i 13 the inflaton mass mφ = 10 GeV. As discussed in the where φˆ is the amplitude of the inflaton oscillations. We previous subsection, we have λ 3 10−4 for all flat find the most conservative bound by considering the most directions but few exceptional ones.≥ The× required values 2 −7 optimistic situation for preheating: of ϕ0 are depicted in Table. 1 for λ = 10 and λ = 1)−1. The three categoric values of inflaton coupling g = The largest possible amplitude for the inflaton os- • 1, 3 10−4, 10−6 illustrate the distinctive regimes where cillations, φˆ (M ). × ∼ O P inflaton decay would take place in the absence of flat The largest possible coupling to the inflaton, g direction VEV: • ∼ (1). −4 O 3 10 g 1: In this case preheating via • × ≤ ≤ 2 2 2 We remind that in the absence of flat direction VEV the quartic interaction term g φ χ would be effi- preheating would enter an explosive stage when H(t) cient [9]. −2 −2 ∼ 10 mφ, at which time φˆ 10 MP [9]. Therefore, to prevent efficient preheating,∼ it will be actually suf- ficient to satisfy Eq. (17) at this time rather than the g =1 g =3 10−4 g = 10−6 very beginning of inflaton oscillations. Moreover, as men- 2 −7 × 17 16 tioned earlier, having g (1) also implies a large χ λ 10 ϕ0 MP ϕ0 > 3 10 ϕ0 > 10 ∼ O 2 ≃ −1 ∼ 16 × 14 13 self-coupling in supersymmetry. Then one in addition λ 10 ϕ0 > 10 ϕ0 > 3 10 ϕ0 > 10 ≃ × expects preheating to be considerably suppressed due to self-interactions [37, 38]. Nevertheless we want to find the Table 1: The flat direction VEV, inflaton coupling g, strongest bound on the flat direction VEV which shuts-off and two values of generic flat direction (Yukawa) cou- non-perturbative particle production at all times and for plings λ within MSSM are illustrated. The VEVs are the largest coupling to the inflaton. The natural conclu- denoted in GeV. For the VEVs above the quoted num- sion is that there will be no preheating in more realistic bers preheating is kinematically blocked. situations.
12 11 Note that in an absence of 2λ2hϕi2 term in Eq. (16), the adi- In the absence of resonant particle production the amplitude ˆ abaticity condition would be violated every time the inflaton of the inflaton oscillations is redshifted like φ ∝ H(t) due to the Hubble expansion. While, for H > m0, we have hϕi ∝ would cross the origin leading to a copious production of χ par- − 1/2 H(t)1/(n 2) , see Eq. (14). Hence the RH side of Eq. (18) be- ˆ ˆ ticles with momentum k ∼< “gφmφ” [8, 9] (φ is the amplitude comes increasingly larger than its LH side, and the adiabaticity of the inflaton oscillations). condition will be satisfied more comfortably as time goes by. 7
10−6 g < 3 10−4: In this case preheating via the slides down in a non-renormalizable potential preheating • quartic≤ term× would not be efficient, however, the will remain shut-off. cubic interaction term gm /√2 φχ2 could lead Once H(t) m 1 TeV, soft SUSY breaking mass φ ≃ 0 ∼ to efficient particle production [31]. term in the potential takes over and the flat direction starts oscillating around its origin with an initial ampli- g < 10−6: In this case inflaton would decay in the 1/n−2 tude: ϕ m M n−3 . However this happens at • perturbative regime from the beginning, thus no in 0 P time scales∼ hierarchically longer than those relevant for resonant and/or tachyonic particle production. the preheating phenomena. It is important to note that preheating is always shut- off for a sub-Planckian flat direction VEV. Even in the VII. INFLATON LIFE TIME most extreme case, i.e. the largest inflaton coupling g (1) and for a flat direction coupling λ 10−4, the∼ O ∼ required VEV for the flat direction is ϕ0 MP. Much Since preheating is kinematically shut-off, the inflaton smaller VEVs, ϕ 1013 GeV, are required≃ for moderate simply oscillates with a decreasing amplitude due to the 0 ≥ values of g and/or λ (1) (which is the case for flat Hubble expansion rate for H(t) < mφ. The Universe is directions including a∼ sizeable O component of Q and/or u therefore dominated by the inflaton oscillations, which 2 2 from the third generation). for mφφ potential act as a non-relativistic matter, im- plying that φˆ H(t). ∝ Once H(t) m0, the flat direction also starts oscillat- A. Additional observations ing and, due≃ to Hubble damping, ϕ H. We remind h i ∝ 2 that for gφˆ mφ the cubic interaction term φχ is dom- Some words on fermionic preheating, see Refs. [44]. inant and production≪ of χ can only occur in a narrow mo- Preheating of superheavy fermions can be much more mentum band peaked around k = mφ/2 [8, 9]. Note that efficient than bosonic preheating [20]. However within at the onset of flat direction oscillations φˆ = m0MP/mφ, SUSY this is not the case as the symmetry between and hence gφˆ m even if g = 1. Therefore the inflaton bosons and fermions implies similar equations for the mo- φ decay will be kinematically≪ forbidden until λ ϕ
for m0 (1) TeV. This underlines the fact that, re- A. No gravitino problem gardless∼ of O how large its coupling to χ is, the inflaton will not decay until after the flat direction has started The foremost consequence of NO preheating is that oscillating, and even then its decay will be strictly per- there will be no copious production of dangerous relics turbative. The expression in Eq. (21) for the inflaton from scatterings in the aftermath of inflaton decay. First lifetime is very robust and practically independent from of all note that there will be no preheat plasma with large how exactly flat direction oscillations terminate (see the occupation numbers fk 1, which could lead to disas- discussion below). trous production of relics≥ from scatterings [1, 28]. On As pointed out in Refs. [1, 2], when the inflaton com- the other hand, since the inflaton decays perturbatively, pletely decays, the decay products of MSSM fields do the MSSM fields are scarcely populated [2]. The chal- not thermalize promptly. The flat direction VEV also lenge is to enhance the number density of particles which gives masses to gauge bosons and gauginos which slow mainly happens via 2 3 scatterings. As pointed out down various scattering processes, i.e. 2 2 and 2 3, in Refs. [1, 2] the production→ of dangerous relics such as ↔ → required for a complete thermalization. The Universe un- gravitinos is very much suppressed in the initial plasma. dergoes a bout of quasi-thermal phase where the plasma Large VEVs of the flat directions do not modify the obtains (near) kinetic equilibrium, but not chemical equi- non-thermal production of gravitinos during the coherent librium. Eventually when the flat direction starts oscil- oscillations of the inflaton [12, 13, 14]. However this will lating and its VEV decreases the Universe reaches full not be a threat for BBN. Helicity 3/2 are not produced thermal equilibrium. In a model independent case, the copiously at the first point, while± 1/2 fermions thus reheat temperature is governed by the thermalization produced are mainly inflatinos, which± decay along with rate, i.e. T (Γ M )1/2, and can be as low as TeV [2]. the inflaton (thus long before BBN) [14]. R ∼ th P We note that the VEVs required for kinematical blocking Another source is direct production of gravitinos via of preheating, see Table. 1, are sufficiently large to also the channel inflaton inflatino + gravitino (if kine- delay complete thermalization of the Universe [2]. matically open). If the→ inflaton decay to matter occurs In some cases the coherent oscillations of the flat di- very late, this process would have a large branching ratio rection can fragment the homogeneous mode into a non- and could lead to overproduction of gravitinos [48, 49]. topological solitons [45], known as Q-balls. In which case However this will not be a problem because delayed infla- the flat direction VEV vanishes outside the Q-balls, while ton decay (due to kinematical blocking) occurs at Hd, see still being large inside them. Then, since the Q-balls oc- Eqs. (20,21), in which case the direct gravitino produc- cupy a very tiny fraction of the space, the inflaton decay tion from inflaton decay is under control. Similar analysis will become kinematically allowed as soon as Q-balls are holds for moduli fields also since their interaction rates formed. In principle flat direction oscillations might also are also suppressed similar to the case of gravitinos. decay rapidly via preheating due to their gauge interac- tions and initial condition ϕ m 14. An important 0 ≫ 0 point is that the time scale for fragmentation of flat di- B. No non-thermal leptogenesis rection oscillations, or their non-perturbative decay via preheating, lies within the same order as the RH side of Eq. (21). Therefore, irrespective of the fate of flat direc- Leptogenesis is a scheme for creating a lepton asymme- tion oscillations, Eq. (21) provides a robust lower bound try, which is then partially converted into baryon asym- on the inflaton lifetime. metry via SM sphalerons, for a review see [50]. The prospect for thermal leptogenesis is severely hampered within SUSY due to late thermalization [2]. This moti- vates the case for non-thermal leptogenesis (for example, see [53]). VIII. CONSEQUENCES OF NO-PREHEATING Note that the RH (s)neutrinos can obtain large masses through their coupling to Hu and L superfields (for de- The absence of preheating and a delayed perturbative tails see Section XI B). If flat directions including Hu decay of the inflaton has interesting cosmological conse- and/or L develop a large VEV then the resonant ex- quences which we briefly discuss here. citation of the heavy (s)neutrinos is kinematically for- bidden. If flat directions which do not include Hu or L develop a VEV, for instance udd, then the RH (s)neutrinos obtain large VEV dependent masses from 14 This does not happen for flat directions which have a non-zero A- the non-renormalizable superpotential through effective term (either from higher-order superpotential or K¨ahler potential Yukawa couplings [32]. In either case the resonant pro- terms). The A-term in this case triggers out-of-phase oscillations duction of (s)neutrinos is unlikely for the same reason as of the real and imaginary parts of the flat direction (with compa- rable amplitudes). Then the mass of particles which are coupled we discussed above. This seriously severs the prospect to the flat direction will not experience a non-adiabatic variation, for non-thermal leptogenesis from on-shell superheavy and hence no preheating [46, 47]. (s)neutrinos. 9
C. No creation of superheavy WIMPS never take place. More importantly it implies that infla- tion will never end! Therefore, unlike chaotic inflation, In Ref. [9] it was advocated that it is possible to ex- the absence of preheating in hybrid models has a neg- cite superheavy a weakly interacting massive particle ative consequence. Indeed for a graceful exit from the 1/2 (WIMP) from non-perturbative inflaton decay. In most inflationary phase one needs to have λϕ0
Note however that a large subset of MSSM flat direc- 3- Directions which include two or more d from dif- tions survive. In particular those which are only made ferent generations: udd and LLddd . These direc- −3 up of Q, u, e superfields will not be affected at all . Inter- tions have Yukawa couplings 10 to Hd. estingly these directions are coupled through the MSSM ≥ 4- Directions which include Q and u from different superpotential in Eq. (2) to both of the Hu and Ψ super- generations: QuLe and QuQue . These directions fields, whether Ψ = Hd or Ψ = L, see Eq. (6). As we will have a coupling 10−3 to H . see, this is very important for shutting off preheating. ≥ u 5- Directions which include Q and d from different QdL B. Inflaton couplings to SM gauge singlets generations: . These directions have Yukawa couplings 10−3 to H . ≥ d Inflaton being a gauge singlet need not directly couple 6- Directions which include two L from different gen- to the (MS)SM sector, but instead can do so through an- erations: LLe. These directions have Yukawa cou- −3 other SM gauge singlet. Phenomenologically the best plings 10 to Hd. motivated example is when the inflaton is coupled to ≥ the RH (s)neutrinos. Note that the couplings of RH Note that the requirement that more than one genera- (s)neutrinos to the (MS)SM sector can explain the origin tion of squarks and/or sleptons be involved comes as a of light neutrino masses via see-saw mechanism [52]. direct consequence of F- and D-flatness [22]. Now let us Assuming that the RH (s)neutrinos obtain masses from consider the Ψ = Hd and Ψ = L cases separately, see some other source, the relevant part of the superpotential Eqs. (6,7). will be given by: −3 −3 1 1 1 Ψ = Hd. In this case λ1 10 and/or λ2 10 for W mφΦΦ + gΦNN + hNHuL + MN NN . (28) ≥ ≥ ⊃ 2 2 2 all of the MSSM flat directions, see Eq. (10), implying − Here Φ, N, L, H stand for the inflaton, the RH neutrino, that λ 3 10 4. u ≥ × the lepton doublet, and the Higgs (which gives mass to the top quark) superfields, respectively. Also, m and φ Ψ = L. In this case λ 10−3, hence λ 3 10−4, M denote inflaton and RH (s)neutrino masses, respec- 1 N for flat directions listed in≥ 1, 2, 4. There are≥ exceptional× tively. Here h denotes the 3 3 neutrino Yukawa matrix. × flat directions for which λ 10−3: For simplicity, we have omitted all indices in h matrix ≪ and superfields, and work in the basis where MN is diag- udd: there are three such directions (with u be- • onal. The inflaton coupling to the RH (s)neutrinos can longing to the first generation) for which λ1 be quite large: g (1). −5 −5 ∼ ∼ O (10 ) and λ2 = 0, hence λ (10 ). Any flat direction that includes Hu and/or L can in- O ∼ O duce a large mass to N through renormalizable couplings. QdL: there are 6 such directions (with Q belonging • −5 Moreover note that N can also couple to MSSM fields to the first generation) for which λ1 (10 and −5 ∼ O via non-renormalizable superpotential terms the same as λ2 = 0, hence λ (10 ). ∼ O Eq. (27) with Φ replaced by N. In the presence of large LLddd flat direction VEVs, non-renormalizable interactions in : there is one such direction (where the two • L Ψ Eq. (27) can lead to large effective couplings between are orthogonal to . As mentioned earlier, ψ N and MSSM fields [32]. For a reasonable flat direc- cannot acquire a large VEV since its coupling to −4 the inflaton induces a mass HI.) for which λ1 = tion VEV, one can have λeff 3 10 . Then non- ≫ perturbative decay of the inflaton≥ to× RH (s)neutrinos will λ2 = 0, hence λ = 0. be kinematically forbidden similar to our earlier analysis LLe: there is one such direction (where the two L as in Section IV. • are orthogonal to Ψ) for which λ1 = 0 and λ2 (10−5), hence λ (10−5). ∼ O ∼ O C. Categorizing flat directions The exceptional directions constitute a small subset of all MSSM flat directions: 11 out of nearly 300. More- A close inspection to the MSSM flat directions shows over, considering multi-dimensionality of the space of all that they belong to one of the following groups (for in- flat directions, it is very unlikely that a VEV grows ex- stance see [24]): actly along one of the exceptional directions during in- flation. To elucidate consider flat directions represented 1- Directions which include two or more Q from dif- by the udd monomial. When different generations are ferent generations: QQQL. These directions have taken into account, this monomial represents a space of Yukawa couplings 10−3 to both H and H both. ≥ u d complex dimension 9 [22]. While, exceptional directions 2- Directions which includes two or more u from dif- span a three-dimensional subspace. Therefore even with ferent generations: uude and uuuee. These direc- a probabilistic argument it is very hard to imagine that tions have Yukawa couplings 10−3 to H . a non-zero VEV will be confined to this subspace of flat ≥ u 12
−3 −4 directions. For the bulk of 9-dimensional space the super- implying that λ1 > 10 and λ 3 10 . field, u, has comparable components from all generations, ∼ ≥ ×
[1] R. Allahverdi and A. Mazumdar, “Quasi-thermal uni- [15] G. F. Giudice, A. Riotto and I. I. Tkachev, JHEP 0106, verse and its implications for gravitino production, baryo- 020 (2001) [arXiv:hep-ph/0103248]. genesis and dark matter,” arXiv:hep-ph/0505050. [16] F. Finelli and R. H. Brandenberger, Phys. Rev. Lett. [2] R. Allahverdi and A. Mazumdar, “Supersymmetric ther- 82, 1362 (1999) [arXiv:hep-ph/9809490]. F. Finelli and malization and quasi-thermal universe: Consequences for R. H. Brandenberger, Phys. Rev. D 62, 083502 (2000) gravitinos and leptogenesis,” arXiv:hep-ph/0512227. [arXiv:hep-ph/0003172]. [3] D. H. Lyth and A. Riotto, “Particle physics models of in- [17] S. Y. Khlebnikov and I. I. Tkachev, Phys. Rev. D 56, flation and the cosmological density perturbation,” Phys. 653 (1997) [arXiv:hep-ph/9701423]. Rept. 314, 1 (1999) [arXiv:hep-ph/9807278]. [18] K. Enqvist, A. Jokinen, A. Mazumdar, T. Multa- [4] For a review, see: K. A. Olive, G. Steigman and T. P. maki and A. Vaihkonen, Phys. Rev. Lett. 94, 161301 Walker, Phys. Rept. 333, 389 (2000). (2005) [arXiv:astro-ph/0411394]. K. Enqvist, A. Jokinen, [5] A. Dolgov and A. D. Linde, Phys. Lett. B 116, 329 A. Mazumdar, T. Multamaki and A. Vaihkonen, JCAP (1982); L. F. Abbott, E. Farhi and M. Wise, Phys. Lett. 0503, 010 (2005) [arXiv:hep-ph/0501076]. K. Enqvist, B 117, 29 (1982). A. Jokinen, A. Mazumdar, T. Multamaki and A. Vaihko- [6] S. Davidson and S. Sarkar, JHEP 0011; 012 (2000) nen, JHEP 0508, 084 (2005) [arXiv:hep-th/0502185]. [arXiv:hep-ph/0009078]. R. Allahverdi, Phys. Rev. D N. Barnaby and J. M. Cline, Phys. Rev. D 73, 62, 063509 (2000) [arXiv:hep-ph/0004035]. R. Allahverdi 106012 (2006) [arXiv:astro-ph/0601481]. N. Barnaby and M. Drees, Phys. Rev. D 66, 063513 (2002) and J. M. Cline, Phys. Rev. D 75, 086004 (2007) [arXiv:hep-ph/0205246]. [arXiv:astro-ph/0611750]. [7] P. Jaikumar and A. Mazumdar, Nucl. Phys. B 683, 264 [19] A. Jokinen and A. Mazumdar, arXiv:astro-ph/0512368. (2004) [arXiv:hep-ph/0212265]. [20] G. F. Giudice, M. Peloso, A. Riotto and I. I. Tkachev, [8] J. Traschen and R. Brandenberger, Phys. Rev. D 42, JHEP 9908, 014 (1999) [arXiv:hep-ph/9905242]. 2491 (1990). [21] I. I. Tkachev, Phys. Lett. B 376, 35 (1996) [9] L. Kofman, A. D. Linde and A. A. Starobinsky, Phys. [arXiv:hep-th/9510146]. S. Khlebnikov, L. Kofman, Rev. Lett. 73, 3195 (1994) [arXiv:hep-ph/9405187]; L. A. D. Linde and I. Tkachev, Phys. Rev. Lett. 81, 2012 Kofman, A. D. Linde and A. A. Starobinsky, Phys. Rev. (1998) [arXiv:hep-ph/9804425]. I. Tkachev, S. Khleb- D 56, 3258 (1997) [arXiv:hep-ph/9704452]; Y. Shtanov, nikov, L. Kofman and A. D. Linde, Phys. Lett. B 440, J. H. Traschen and R. H. Brandenberger, Phys. Rev. D 262 (1998) [arXiv:hep-ph/9805209]. 51, 5438 (1995) [arXiv:hep-ph/9407247]; D. Boyanovsky, [22] T. Gherghetta, C. Kolda and S. P. Martin, Nucl. Phys. H. J. de Vega and R. Holman, arXiv:hep-ph/9701304; B 468, 37 (1996) [arXiv:hep-ph/9510370]. D. Cormier, K. Heitmann and A. Mazumdar, Phys. Rev. [23] K. Enqvist, A. Jokinen and A. Mazumdar, JCAP 0401, D 65, 083521 (2002) [arXiv:hep-ph/0105236]. 008 (2004) [arXiv:hep-ph/0311336]. [10] K. Enqvist, S. Kasuya and A. Mazumdar, Phys. Rev. [24] K. Enqvist and A. Mazumdar, Phys. Rept. 380, 99 Lett. 89, 091301 (2002) [arXiv:hep-ph/0204270]. K. En- (2003) [arXiv:hep-ph/0209244]. M. Dine and A. Kusenko, qvist, S. Kasuya and A. Mazumdar, Phys. Rev. D 66, Rev. Mod. Phys. 76, 1 (2004) [arXiv:hep-ph/0303065]. 043505 (2002) [arXiv:hep-ph/0206272]. [25] K. Enqvist, A. Jokinen and A. Mazumdar, JCAP 0411, [11] For example, see: K. Jedamzik, Class. Quant. Grav. 19, 001 (2004) [arXiv:hep-ph/0404269]. 3417 (2002) [arXiv:astro-ph/0112226]. [26] G. N. Felder and L. Kofman, Phys. Rev. D 63, 103503 [12] A. L. Maroto and A. Mazumdar, Phys. Rev. Lett. 84, (2001) [arXiv:hep-ph/0011160]. 1655 (2000) [arXiv:hep-ph/9904206]. [27] R. Micha and I. I. Tkachev, arXiv:hep-ph/0301249. [13] R. Kallosh, L. Kofman, A. D. Linde and A. Van Proeyen, R. Micha and I. I. Tkachev, Phys. Rev. D 70, 043538 Phys. Rev. D 61, 103503 (2000) [arXiv:hep-th/9907124]; (2004) [arXiv:hep-ph/0403101]. R. Kallosh, L. Kofman, A. D. Linde and A. Van [28] D. I. Podolsky, G. N. Felder, L. Kofman and M. Peloso, Proeyen, Class. Quant. Grav. 17, 4269 (2000) Phys. Rev. D 73, 023501 (2006) [arXiv:hep-ph/0507096]. [arXiv:hep-th/0006179]; A. L. Maroto and J. R. Pelaez, [29] V. F. Mukhanov, H. A. Feldman and R. H. Branden- Phys. Rev. D 62, 023518 (2000) [arXiv:hep-ph/9912212]; berger, Phys. Rept. 215, 203 (1992). G. F. Giudice, A. Riotto and I. I. Tkachev, JHEP [30] N. Shuhmaher and R. Brandenberger, Phys. Rev. D 73, 9911, 036 (1999) [arXiv:hep-ph/9911302]; G. F. Giu- 043519 (2006) [arXiv:hep-th/0507103]. dice, I. I. Tkachev and A. Riotto, JHEP 9908, [31] J. F. Dufaux, G. Felder, L. Kofman, M. Peloso and 009 (1999) [arXiv:hep-ph/9907510]; M. Bastero-Gil D. Podolsky, arXiv:hep-ph/0602144. and A. Mazumdar, Phys. Rev. D 62, 083510 (2000) [32] R. Allahverdi, R. Brandenberger and A. Mazumdar, [arXiv:hep-ph/0002004]. Phys. Rev. D 70, 083535 (2004) [arXiv:hep-ph/0407230]. [14] R. Allahverdi, M. Bastero-Gil and A. Mazumdar, Phys. [33] K. Enqvist, A. Mazumdar and M. Postma, Phys. Rev. D 64, 023516 (2001) [arXiv:hep-ph/0012057]. Rev. D 67, 121303 (2003) [arXiv:astro-ph/0304187]; H. P. Nilles, M. Peloso and L. Sorbo, Phys. Rev. Lett. A. Mazumdar and M. Postma, Phys. Lett. B 87, 051302 (2001) [arXiv:hep-ph/0102264]. H. P. Nilles, 573, 5 (2003) [Erratum-ibid. B 585, 295 (2004)] M. Peloso and L. Sorbo, JHEP 0104, 004 (2001) [arXiv:astro-ph/0306509]. R. Allahverdi, Phys. Rev. D [arXiv:hep-th/0103202]. 70, 043507 (2004) [arXiv:astro-ph/0403351]. 13
[34] R. Brandenberger, P. M. Ho and H. c. Kao, JCAP S. Pascoli, arXiv:hep-ph/0504102. 0411, 011 (2004) [arXiv:hep-th/0312288]. A. Jokinen [50] W. Buchm¨uller, R. D. Peccei and T. Yanagida, and A. Mazumdar, Phys. Lett. B 597, 222 (2004) arXiv:hep-ph/0502169. [arXiv:hep-th/0406074]. [51] A. Mazumdar, Phys. Lett. B 580, 7 (2004) [35] A. R. Liddle, A. Mazumdar and F. E. Schunck, Phys. [arXiv:hep-ph/0308020]. Rev. D 58, 061301 (1998) [arXiv:astro-ph/9804177]; [52] P. Minkowski, Phys. Lett. B 67, 421 (1977). M. Gell- E. J. Copeland, A. Mazumdar and N. J. Nunes, Phys. Mann, P. Ramond and R. Slansky, in Supergravity, eds. Rev. D 60, 083506 (1999) [arXiv:astro-ph/9904309]; P. van Nieuwenhuizen and D. Z. Freedman (North Hol- A. Mazumdar, S. Panda and A. Perez-Lorenzana, Nucl. land, 1979). T. Yanagida, Proceedings of Workshop on Phys. B 614, 101 (2001) [arXiv:hep-ph/0107058]. Unified Theory and Baryon number in the Universe, eds. [36] H. Murayama, H. Suzuki, T. Yanagida and J. Yokoyama, O. Sawada and A. Sugamoto (KEK, Tsukuba, 1979). R. Phys. Rev. Lett. 70, 1912 (1993); H. Murayama, H. N. Mohapatra and G. Senjanovic, Phys. Rev. Lett. 44, Suzuki, T. Yanagida and J. Yokoyama, Phys. Rev. D 50, 912 (1980). 2356 (1994) [arXiv:hep-ph/9311326]. [53] G. Lazarides and Q. Shafi, Phys. Lett. B 258, 305 [37] R. Allahverdi and B. A. Campbell, Phys. Lett. B 395, (1991). R. Allahverdi and A. Mazumdar, Phys. Rev. 169 (1997) [arXiv:hep-ph/9606463]. D 67, 023509 (2003) [arXiv:hep-ph/0208268]. T. Dent, [38] T. Prokopec and T. G. Roos, Phys. Rev. D 55, 3768 G. Lazarides and R. Ruiz de Austri, Phys. Rev. D (1997) [arXiv:hep-ph/9610400]. 69, 075012 (2004) [arXiv:hep-ph/0312033]. T. Dent, G. [39] M. Dine, L. Randall and S. Thomas, Phys. Rev. Lett. Lazarides and R. Ruiz de Austri, arXiv:hep-ph/0503235. 75, 398 (1995) [arXiv:hep-ph/9503303]. M. Dine, L. Ran- A. Mazumdar, Phys. Rev. Lett. 92, 241301 (2004) dall and S. Thomas, Nucl. Phys. B 458, 291 (1996) [arXiv:hep-ph/0306026]. R. Allahverdi, B. Dutta and [arXiv:hep-ph/9507453]. A. Mazumdar, Phys. Rev. D 67, 123515 (2003) [40] A. B. Lahanas and D. V. Nanopoulos, Phys. Rept. 145, [arXiv:hep-ph/0301184]. H. Murayama and T. Yanagida, 1 (1987). Phys. Lett. B 322, 349 (1994) [arXiv:hep-ph/9310297]. [41] M. K. Gaillard, H. Murayama and K. A. Olive, Phys. K. Hamaguchi, H. Murayama and T. Yanagida, Phys. Lett. B 355, 71 (1995) [arXiv:hep-ph/9504307]. Rev. D 65, 043512 (2002) [arXiv:hep-ph/0109030]. Z. [42] M. Dine and N. Seiberg, Nucl. Phys. B 160, 243 (1985). Berezhiani, A. Mazumdar and A. P´erez-Lorenzana, Phys. [43] R. Allahverdi, M. Drees and A. Mazumdar, Phys. Rev. Lett. B 518, 282 (2001) [arXiv:hep-ph/0107239]. R. Al- D 65, 065010 (2002) [arXiv:hep-ph/0108225]. lahverdi and M. Drees, Phys. Rev. D 69, 103522 (2004) [44] P. B. Greene and L. Kofman, Phys. Lett. B 448, [arXiv:hep-ph/0401054]. R. Allahverdi and M. Drees, 6 (1999) [arXiv:hep-ph/9807339]. J. Baacke, K. Heit- Phys. Rev. D 70, 123522 (2004) [arXiv:hep-ph/0408289]; mann and C. Patzold, Phys. Rev. D 58, 125013 [54] L. Boubekeur, S. Davidson, M. Peloso and L. Sorbo, (1998) [arXiv:hep-ph/9806205]. A. L. Maroto and Phys. Rev. D 67, 043515 (2003) [arXiv:hep-ph/0209256]. A. Mazumdar, Phys. Rev. D 59, 083510 (1999) [55] G. N. Felder, J. Garcia-Bellido, P. B. Greene, L. Kof- [arXiv:hep-ph/9811288]. man, A. D. Linde and I. I. Tkachev, Phys. Rev. Lett. [45] A. Kusenko, Phys. Lett. B 405, 108 (1997) 87, 011601 (2001) [arXiv:hep-ph/0012142]. G. N. Felder, [arXiv:hep-ph/9704273]. A. Kusenko, Phys. Lett. B L. Kofman and A. D. Linde, Phys. Rev. D 64, 123517 404, 285 (1997) [arXiv:hep-th/9704073]. A. Kusenko (2001) [arXiv:hep-ph/0106179]. and M. E. Shaposhnikov, Phys. Lett. B 418, 46 (1998) [56] P. B. Greene, L. Kofman, A. D. Linde and A. [arXiv:hep-ph/9709492]. K. Enqvist and J. McDonald, A. Starobinsky, Phys. Rev. D 56, 6175 (1997) Phys. Lett. B 425, 309 (1998) [arXiv:hep-ph/9711514]. [arXiv:hep-ph/9705347]. K. Enqvist and J. McDonald, Nucl. Phys. B 538, 321 [57] R. Kallosh, A. D. Linde, D. A. Linde and L. Susskind, (1999) [arXiv:hep-ph/9803380]. Phys. Rev. D 52, 912 (1995) [arXiv:hep-th/9502069]. [46] R. Allahverdi, R. H. A. Shaw and B. A. Campbell, Phys. [58] F. Buccella, J. P. Derendinger, S. Ferrara, and C. A. Lett. B 473, 246 (2000) [arXiv:hep-ph/9909256]. Savoy, Phys. Lett. B 115, 375 (1982). [47] M. Postma and A. Mazumdar, JCAP 0401, 005 (2004) [59] I. Affleck, M. Dine, and N. Seiberg, Nucl. Phys. B 241, [arXiv:hep-ph/0304246]. 493 (1984). [48] H. P. Nilles, K. A. Olive and M. Peloso, Phys. Lett. B [60] I. Affleck, M. Dine, and N. Seiberg, Nucl. Phys. B 256, 522, 304 (2001) [arXiv:hep-ph/0107212]. 557 (1985). [49] R. Allahverdi, A. Jokinen and A. Mazumdar, Phys. Rev. [61] M. A. Luty, and W. Taylor, Phys. Rev. D 53, 3399 D 71, 043505 (2005) [arXiv:hep-ph/0410169]. R. Al- (1996). [arXiv:hep-th/9506098]. lahverdi, S. Hannestad, A. Jokinen, A. Mazumdar and