Unitarity-Safe Models of Non-Minimal Inflation in Supergravity

We show that models of chaotic inflation based on the phi^p potential and a linear non-minimal coupling to gravity, fR=1+cR phi, can be done consistent with data in the context of Supergravity, retaining the perturbative unitarity up to the Planck scale, if we employ logarithmic Kahler potentials with prefactors -p(1+n) or -p(n+1)-1, where -0.035<= n<=0.007 for p=2 or -0.0145<= n<=0.006 for p=4. Focusing, moreover, on a model employing a gauge non-singlet inflaton, we show that a solution to the mu problem of MSSM and baryogenesis via non-thermal leptogensesis can be also accommodated.


I. INTRODUCTION
Inflation established in the presence of a non-minimal coupling between the inflaton φ and the Ricci scalar R is called collectively non-minimal inflation (nMI) [1][2][3][4][5][6][7][8]. Between the numerus models, which may be proposed in this context, universal attractor models (UAMs) [9] occupy a prominent position since they exhibit an attractor towards an inflationary phase excellently compatible with data [10] for c R ≫ 1 and φ < 1 -in the reduced Planck units with m P = 1. UAMs consider a monomial potential of the type V CI (φ) = λ 2 φ p /2 p/2 (1) in conjunction with a strong non-minimal coupling [3,9] f R (φ) = 1 + c R φ q/2 (2) with p = q. The emergence of an inflationary plateau in these models can be transparently shown in the Einstein frame (EF) where the inflationary potential, V attr , takes the form with the exponent in the denominator being related to the conformal transformation employed [1][2][3] to move from the Jordan frame (JF) to EF. However, UAMs are plagued with the consistency of the effective theory for 2 < q ≤ 14/3 [11,12], due to the large c R values needed for the establishment of the inflationary stage with φ ≤ 1. Indeed, for the q values above, the inflationary scale V 1/4 attr turns out to be larger than the Ultraviolet (UV) cutoff scale of the effective theory which thereby breaks down above it. Several ways have been proposed to surpass this inconsistency. E.g., incorporating new degrees of freedom at Λ attr UV [13], or assuming additional interactions [14], or invoking a large inflaton vacuum expectation value (v.e.v) φ as in Refs. [15][16][17][18], or introducing a sizable kinetic mixing in the inflaton sector which dominates overf R [19][20][21][22].
Here we propose a novel solution -applied only in the context of Supergravity (SUGRA) -to the aforementioned problem, by exclusively considering q = 2 in Eq. (2). In this case, the canonically normalized inflaton φ is related to the initial field φ as φ ∼ c R φ at the vacuum of the theory, in sharp contrast to what we obtain for p > 2 where φ ≃ φ. As a consequence, the small-field series of the various terms of the action expressed in terms of φ, does not contain c R in the numerators, preventing thereby the reduction of Λ UV below m P [5,12]. In other words, moving to the JF, no dangerous inflaton-inflaton-graviton interaction appears sincef R,φφ = 0 [5] and so, the theory does not face any problem with the perturbative unitarity. A permanently linearf R can be reconciled with an inflationary plateau, similar to that obtained in Eq. (3), in the context of SUGRA, by suitably selecting the employed Kähler potentials. Indeed, this kind of models is realized in SUGRA using logarithmic or semilogarithmic Kähler potentials [6,7] with the prefactor (−N ) of the logarithms being related to the exponent of the denominator in Eq. (3). Therefore, by conveniently adjusting N we can achieve, in principle, a flat enough EF potential for any p in Eq. (1) but taking exclusively q = 2 in Eq. (2).
As we show in the following, this idea works for p ≤ 4 in Eq. (1) supporting nMI compatible with the present data [10]. For p = 4 we also show that the inflaton may be identified with a gauge singlet or non-singlet field. In the latter case, models of non-minimal Higgs inflation are introduced, which may be embedded in a more complete extension of MSSMcf. Refs. [18,23] -offering a solution to the µ problem [24] and allowing for an explanation of baryon asymmetry of the universe (BAU) [25] via non-thermal leptogenesis (nTL) [26].
Below, we first -in Sec. II -describe the SUGRA set-up of our models and then, in Sec. III, we analyze the inflationary dynamics and predictions. In Sec. IV we concentrate on the case of nMI driven by a Higgs field and propose a possible post-inflationary completion. We conclude in Sec. V.

II. SUPERGRAVITY FRAMEWORK
In Sec. II A we describe the generic formulation of our models within SUGRA, and then we apply it for a gauge singlet and non-singlet inflaton in Sec. II B and II C respectively. Finally, in Sec. II D, we analyze the UV behavior of the models.

A. GENERAL FRAMEWORK
We focus on the part of the EF action within SUGRA related to the complex scalars z α -denoted by the same superfield symbol -which has the form [6] (5a) where R is the EF Ricci scalar curvature, D µ is the gauge covariant derivative, K αβ = K ,z α z * β , and K αβ Kβ γ = δ α γ -here and hereafter subscript of type , z denotes derivation with respect to (w.r.t) the field z. Also, V is the EF SUGRA potential which can be found once we select a superpotential W in Eq. (24) and a Kähler potential K via the formula β α K β and the summation is applied over the generators T a of a gauge group. In the right-hand side (r.h.s) of the equation above we clearly recognize the contribution from the F terms arising from the two first terms and the remaining one (proportional to the squared gauge coupling constant g 2 ) which comes from the D terms. The last one vanishes for a gauge singlet inflaton and can be eliminated during nMI for a gauge non-singlet inflaton, by identifying it with the radial part of a conjugate pair of Higgs superfields -see Sec. II C. In both our scenaria, we employ a "stabilizer" field S placed at the origin during nMI. Thanks to this arrangement, the term 3|W | 2 in V vanishes, avoiding thereby a possible runaway problem, and the derivation of V is facilitated since the non-vanishing terms arise from those proportional to W ,S and W * ,S * -see Secs. II B 2 and II C 2 below.
Conformally transforming to the JF -defining the frame function as −Ω/N = exp (−K/N ) ⇒ K = −N ln (−Ω/N ) , (6) where N > 0 is a dimensionless parameter -we can obtain following Refs. [6,20] the form of S in the Jordan Frame (JF) which is written as where we use the shorthand notation Ω α = Ω ,z α , and Ωᾱ = Ω ,z * ᾱ . We also set V = V Ω 2 /N 2 and Although the choice N = 3 ensures canonical kinetic terms in Eq. (7a), N may be considered in general as a free parameter with interesting consequences not only on the inflationary observables [5,17,20,22,27] but also on the consistency of the effective theory, as we show below.

B. GAUGE-SINGLET INFLATON
Below, in Sec. II B 1, we specify the necessary ingredients (super-and Kähler potentials) which allow us to implement our scenario with a gauge-singlet inflaton. Then, in Sec. II B 2, we outline the derivation of the inflationary potential.

Set-up
This class of models requires the utilization of two gauge singlet chiral superfields, i.e., z α = Φ, S, with Φ (α = 1) and S (α = 2) being the inflaton and a "stabilizer" field respectively. More specifically, we adopt the superpotential which can be uniquely determined if we impose two symmetries: (i) an R symmetry under which S and Φ have charges 1 and 0; (ii) a global U (1) symmetry with assigned charges −1 and 2/p for S and Φ. To obtain a linear non-minimal coupling of Φ to gravity, though, we have to violate the latter symmetry as regards Φ. Indeed, we propose the following set of Kähler potentials Recall that N > 0. From the involved functions the first one allows for the introduction of the linear nonminimal coupling of Φ to gravity whereas the second one assures canonical normalization of Φ without any contribution to the non-minimal coupling along the inflationary path -cf. Refs. [9,21]. On the other hand, the functions F lS with l = 1, 2, 3 offer canonical normalization and safe stabilization of S during and after nMI. Their possible forms are given in Ref. [23]. Just for definiteness, we adopt here only their logarithmic form, i.e., with 0 < N S < 6. Recall [6,28] that the simplest term |S| 2 leads to instabilities for K = K 1 and K 2 and light excitations for K = K 3 − K 5 . The heaviness of these modes is required so that the observed curvature perturbation is generated wholly by our inflaton in accordance with the lack of any observational hint [25] for large non-Gaussianity in the cosmic microwave background. Apart from F R , all the proposed K's contain up to quadratic terms of the various fields. Note also that F R (and F * R ) is exclusively included in the logarithmic part of the K's whereas F − may or may not accompany it in the argument of the logarithm. Note finally that, although quadratic nMI is analyzed in Refs. [4,5,9] too, the present set of K's is examined for first time.

Inflationary Potential
Along the inflationary track determined by the constraints if we express Φ and S according to the parametrization the only surviving term in Eq. (5b) is which, for the K's in Eqs. (9a) -(9e), reads where we define the (inflationary) frame function as As expected, f R coincides withf R in Eq. (2) for q = 2. This form of f R assures the preservation of unitarity up to m P = 1 as explained in Sec. I and verified in Sec. II D. The last factor in Eq. (15) originates from the expression of K SS * for the various K's. Indeed, K αβ along the configuration in Eq. (12) takes the form where and If we set we arrive at a universal expression for V CI which is For n = 0 and p = 2, V CI reduces to V attr in Eq. (3) whereas for p > 2, V CI deviates from V attr although it develops a similar inflationary plateau, for c R ≫ 1, since both numerator and denominator are dominated by a term proportional to φ p as in the case of UAMs. The choice n = 0 is special since, for integer p, it yields integer N in Eqs. (9a) -(9e), i.e., N = p + 1 for K = K 1 and K 2 or N = p for K = K 3 − K 5 . Although integer N 's are more friendly to string theory -and give observationally acceptable results as shown in Sec. III B -, noninteger N 's are also acceptable [17,20,22,23,27] and assist us to cover the whole allowed domain of the observables. More specifically, for n < 0, V CI remains an increasing function of φ, whereas for n > 0, it develops a local maximum In a such case we are forced to assume that hilltop [29] nMI occurs with φ rolling from the region of the maximum down to smaller values. Defining the EF canonically normalized fields, denoted by hat, via the relations d φ dφ = K ΦΦ * = J, θ = Jθφ and ( s, s) = K SS * (s,s) (23) we can verify that the configuration in Eq. (12) is stable w.r.t the excitations of the non-inflaton fields. Taking the limit c R ≫ 1 we find the expressions of the masses squared m 2 z α (with z α = θ and s) arranged in Table I, which approach rather well the quite lengthy, exact expressions taken into account in our numerical computation. We infer that m 2 z α ≫ H 2 CI = V CI /3 for 1 < N < 6 and K = K 1 , K 2 or for 0 < N S < 6 and K = K 3 − K 5 . Therefore m 2 z α are not only positive but also heavy enough during nMI. In Table I we display the masses m 2 ψ ± of the corresponding fermions too. We define ψ S = √ K SS * ψ S and ψ Φ = √ K ΦΦ * ψ Φ where ψ Φ and ψ S are the Weyl spinors associated with S and Φ respectively.
Following the strategy of the previous section, we show below, in Sec. II C 1, how we can establish a model of unitarityconserving nMI driven by a a gauge non-singlet inflaton and then, in Sec. II C 2, we outline the derivation of the corresponding inflationary potential.

Set-up
In this case, as explained below Eq. (5b), we employ a pair of left-handed chiral superfields,Φ and Φ, oppositely charged under a gauge group, besides the stabilizer, S, which is a gauge-singlet chiral superfield. Here, we take for simplicity the group U (1) B−L where B and L denote the baryon and lepton number respectively. We base our construction on the superpotential [30] where λ and M are parameters which can be made positive by field redefinitions. W HI is the most general renormalizable superpotential consistent with a continuous R symmetry [30] under which S and W HI are equally charged whereasΦΦ is uncharged. To obtain nMI -which is actually promoted to Higgs inflation -with a linear non-minmal coupling to gravity we combine W HI with one of the Kähler potentials in Eqs. (9a) -(9e) where the functions F R and F − are now defined as Note that the proposed K's respect the symmetries of W HI . As in the case of Sec. II B 1, F − ensures that the kinetic terms ofΦ and Φ do not enter the expression of f R along the inflationary trough -cf. Refs. [18,[20][21][22]].

Inflationary Potential
Employing the parameterization of S in Eq. (13) and expressing Φ andΦ as follows with 0 ≤ θ Φ ≤ π/2, we can determine a D-flat direction from the conditions Along this path the only surviving term is again given by Eq. (14) which now reads -cf. Eq. (15) whereas f R is again given by Eq. (16). If we take into account the definition of n in Eq. (20) we end up with which can be approached rather well by Eq. (21) for p = 4, when M ≪ 1, and λ 2 replaced with λ 2 /4. To specify φ in the present case we note that, for all K's in Eqs. (9a) -(9e) with F R and F − given in Eq. (25), K αβ along the configuration in Eq. (27) takes the form where with κ given in Eq. (18). Upon diagonalization we obtain the following eigenvalues where κ is given in Eq. (19). Inserting Eqs. (26) and (30) in the second term of the r.h.s of Eq. (5a) we can define the EF canonically normalized fields, as follows where θ ± = θ ± θ / √ 2. We notice that the normalization of φ, s and s coincides with that found in Eq. (23). Note, in passing, that the spinors ψ Φ± associated with the superfields Φ andΦ are similarly normalized, i.e., ψ Φ± = √ κ ± ψ Φ± with To check the stability of inflationary direction in Eq. (27) w.r.t the fluctuations of the non-inflaton fields, we derive the mass-squared spectrum of the various scalars defined in Eqs. (33a) and (33b). Taking the limit c R ≫ 1, we find the approximate expressions listed in Table II which are rather accurate at the horizon crossing of the pivot scale. As in case of Table I, we again deduce that m 2 z α ≫ H 2 HI = V HI /3 for 1 < N < 6 and K = K 1 , K 2 or for 0 < N S < 6 and K = K 3 − K 5 . In Table II we also display the masses, M BL , of the gauge boson A BL and the corresponding fermions. The non-vanishing of M BL signals the fact that U (1) B−L is broken during nMI and so no cosmic string are produced at its end. Finally, the unspecified eigenstates ψ ± are defined as

D. EFFECTIVE CUT-OFF SCALE
The motivation of our proposal originates from the fact that f R in Eq. (16) assures that the perturbative unitarity is retained up to m P = 1 although that the attainment of nMI for φ ≤ 1 requires large c R 's -as expected from the UAMs [3][4][5] and verified in Sec. III below. This outstanding trademark occurs since φ = J φ does not coincide with φ at the vacuum of the theory -contrary to the UAMs [11,12] for p > 2 -, because inserting φ = 0 into Eq. (23) or (33a), we obtain To clarify further this key point we analyze the small-field behavior of our models in the EF. Although the expansions about φ = 0, presented below, are not valid [31] during nMI, we consider the UV cutoff scale, Λ UV , extracted this way as the overall cut-off scale of the theory, since the reheating phaserealized via oscillations about φ -is an unavoidable stage of the inflationary dynamics. We focus on the second term in the r.h.s of (5a) for µ = ν = 0 and we expand it about φ = 0 in terms of φ -see Eq. (23). Our result can be written as This expression agrees with that in Ref. [5] for N = 3(1 + n).
Expanding similarly V CI , see Eq. (21), in terms of φ we have From the expressions above we conclude that Λ UV = m P and so the perturbative unitarity up to m P = 1 is conserved, pro-vided that V 1/4 CI ≪ m P . This prerequisite is readily fulfilled as we see Sec. III B.

III. INFLATION ANALYSIS
In Secs. III A and III B below we examine semi-analytically and numerically respectively, if V CI in Eq. (21) may be consistent with a number of observational constraints. Needless to say, as in Sec. II D, this analysis covers also the case of V HI in Eq. (29) for M ≪ 1, p = 4 and λ 2 replaced by λ 2 /4.

A. SEMI-ANALYTIC RESULTS
The period of slow-roll nMI is determined in the EF by the condition -see, e.g., Ref. [32]: where the slow-roll parameters ǫ and η read and can be derived employing J in Eq. (23) without express explicitly V CI in terms of φ. Since J for K = K 1 , K 3 deviates slightly from that for K = K 2 , K 4 , K 5 -see Eq. (19) -we have a discrimination as regards the expressions of ǫ and η in this two cases. Indeed, our results are For any of the K's above we can numerically verify that Eq. (36a) is saturated for φ = φ f , which is found from the condition Apart from irrelevant constant prefactors, the formulas above for n = 0 and K = K 1 , K 3 reduces to the ones obtained for quadratic nMI -cf. Refs. [4,8].
The number of e-foldings N ⋆ that the pivot scale k ⋆ = 0.05/Mpc experiences during nMI and the amplitude A s of the power spectrum of the curvature perturbations generated by φ can be computed using the standard formulae when k ⋆ crosses the inflationary horizon. Taking into account φ ⋆ ≫ φ f , we can derive N ⋆ . We single out the following cases: • For n = 0 and any K in Eqs. (9a) -(9e), we obtain Note that for n = 0 the formulas below for N ⋆ cannot be reduced to the previous one.
• For n < 0 and K = K 2 , K 4 , K 5 , we obtain (41c) where W k is the Lambert or product logarithmic function [35] with k = 0 and • For n > 0 and K = K 2 , K 4 , K 5 , we obtain Here we are not able to solve the equation above w.r.t φ ⋆ . As a consequence, it is not doable to find an analytical expression for φ ⋆ and the inflationary observables -see below. Therefore, in this portion of parameter space, our last resort is the numerical computation, whose the results are presented in Sec. III B.
In all cases above, there is a lower bound on c R , above which φ ⋆ < 1 and so, our proposal can be stabilized against corrections from higher order terms in the K's. E.g., for n = 0 with N ⋆ ≃ (52−59) for p = 2−4. As shown in Sec. II D these large c R 's do not disturb the validity of perturbative unitarity up to m P . From the right formula in Eq. (40) we can derive a relation between λ and c p/2 R for fixed n, in sharp contrast to UAMs where the same condition implies a relation between λ and c R [4,7,8]. Given that c R assumes large values, we expect that λ increases with p and rapidly (for p ≥ 5 as we find numerically) violates the perturbative bound λ ≤ 2 √ π ≃ 3.5. In particular, our results can be cast as following: • For n = 0 and any K in Eqs. (9a) -(9e), we obtain • For n = 0 and K = K 1 , K 3 , we obtain . Taking into account that n ≪ 1 and f R⋆ is almost proportional to c R for large c R , we can easily convince ourselves that the output above implies that λ/c p/2 R remains constant for fixed n.
• For n < 0 and K = K 2 , K 4 , K 5 , we obtain from which we can again verify that the approximate proportionality of λ on c p/2 R holds. The remaining inflationary observables -i.e., the (scalar) spectral index n s , its running a s , and the scalar-to-tensor ratio r -are found from the relations [32] where the variables with subscript ⋆ are evaluated at φ = φ ⋆ and ξ = V CI, φ V CI, φ φ φ / V 2 CI . Inserting φ ⋆ from Eqs. (41a), (41b) and (41c) into Eq. (36b) and then into equations above we can obtain some analytical estimates. Namely: • For K = K 1 and K 3 we end up with a unified result For n = 0 the above results are also valid for K = K 2 , K 4 and K 5 and yield observables identical with those obtained within UAMs [4,6,7].
• For n < 0 and K = K 2 , K 4 and K 5 we arrive at the following results with negligibly small a s , as we find out numerically. Contrary to our previous results, here a c R dependence arises which complicates somehow the investigation of these models -see Sec. III B.

B. Numerical Results
The conclusions above can be verified and extended for any n numerically. In particular, we confront the quantities in Eq. (40) with the observational requirements [25] where we assume that nMI is followed in turn by a oscillatory phase, with mean equation-of-state parameter w rh ≃ 0 or 1/3 for p = 2 or 4 respectively [25], radiation and matter domination. Also T rh is the reheat temperature after nMI, with energy-density effective number of degrees of freedom g rh * = 228.75 which corresponds to the MSSM spectrumsee Sec. IV C 1. Enforcing Eq. (47a) we can restrict λ/c p/2 R and φ ⋆ and compute the model predictions via Eqs. (44a) and (44b), for any selected p and n. The outputs, encoded as lines in the n s − r 0.002 plane, are compared against the observational data [25,33] in Fig. 1 for K = K 1 and K 3 (left plot) or K = K 2 , K 4 and K 5 (right plot) -here r 0.002 = 16 ǫ(φ 0.002 ) where φ 0.002 is the value of φ when the scale k = 0.002/Mpc, which undergoes N 0.002 = ( N ⋆ + 3.22) e-foldings during nMI, crosses the horizon of nMI. We draw dashed [solid] lines for p = 2 [p = 4] and show the variation of n along each line. We take into account the data from Planck and Baryon Acoustic Oscillations (BAO) and the BK14 data taken by the BICEP2/Keck Array CMB polarization experiments up to and including the 2014 observing season. Fitting the data above [10,33] with ΛCDM+r we obtain the marginalized joint 68% at 95% confidence level (c.l.) with |a s | ≪ 0.01.
From the left plot of Fig. 1 we observe that the whole observationally favored range of n s is covered varying n which, though, remains close to zero signalizing an amount of tuning. In accordance with Eqs. (45a) and (45b), we find the allowed ranges As n varies in its allowed ranges above, we obtain for p = 2 or p = 4 respectively in agreement with Eqs. (43a) and (43b). If we take n = 0, we find n s = 0.963, a s ≃ −6.7 · 10 −4 and r = 0.004 for p = 2 demanding N ⋆ ≃ 53, whereas for p = 4 we get n s = 0.967, a s ≃ −5.6 · 10 −4 and r = 0.005 requiring N ⋆ ≃ 58.6. Therefore, for integer prefactors of the logarithms in Eqs. (9a) and (9c), n s converges towards its central value in Eq. (48) and practically coincides with the prediction of the UAMs [3,7,9]. Fixing, in addition to n = 0 and K = K 1 , φ ⋆ = 1 -i.e. confining the corresponding c R and λ values to their lowest possible values enforcing Eq. (43a) -we illustrate in Fig. 2 the structure of V CI as a function of φ for p = 2 (light gray line) or p = 4 (gray line). More specifically, we find λ = 1.173 · 10 −3 or 0.257 and c R = 75 or 99 for p = 2 or p = 4 respectively. We see that in both cases V CI develops a plateau with magnitude 10 −10 which is similar to that obtained in Starobinsky inflation [16,28] but one order of lower than that obtained from the models analyzed in Refs. [22,23] where r is a little more enhanced. Obviously, the requirement -mentioned in Sec. II D -V 1/4 CI ≪ m P dictated from the validity of the effective theory is readily fulfilled. The values corresponding to φ⋆, φ f are also indicated.
Practically the same observables for n = 0 are shown in the right plot of Fig. 1. In that plot, though, we see that the model predictions are confined to n s 0.974, n may deviate more appreciably from zero (mainly for p = 2) and the maximal possible r is somewhat larger. Moreover, , these predictions depend harder on c R for |n| > 0.01, as expected from Eqs. (46a) and (46b). Therefore, in that regime, we could say that these models are less predictive than those based on K = K 1 and K 3 . Our results below are presented for c R such that φ ⋆ ≃ 1. Namely, for p = 2, we find From the data of both plots of Fig. 1, we remark that r 0.0013. These r values are testable by the forthcoming experiments [34], which are expected to measure r with an accuracy of 10 −3 . The tuning, finally, required for the attainment of hilltop nMI for n > 0 is very low, since φ max ≫ φ ⋆ .
Although λ/c p/2 R is constant for fixed n, the amplitudes of λ and c R can be bounded. This fact is illustrated in Fig. 3 where we display the allowed values c R versus √ λ for p = 4 and K = K 1 (gray lines) or K = K 5 (light gray lines). We take n = 0 (solid lines) and n = −0.004 (dashed lines). As anticipated in Eq. (42) for any n there is a lower bound on c R , above which φ ⋆ ≤ 1 stabilizing thereby the results against corrections from higher order terms -e.g., (ΦΦ) l with l > 1 in Eq. (24). The perturbative bound λ = 3.5 limits the various lines at the other end. We observe that the ranges of the allowed lines are much more limited compared to other models -cf. Refs. [7,21] -and displaced to higher λ values as seen, also, by Eqs. (43a) and (43b). We find that for p = 5, λ corresponding to lowest possible c R violates the perturbative bound and so, our proposal can not be applied for p > 4.

IV. A POST-INFLATIONARY COMPLETION
In a couple of recent papers [18,23] we attempt to connect the high-scale inflationary scenario based on W = W HI in Eq. (24) with the low energy physics, taking into account constraints from the observed BAU, neutrino data and MSSM phenomenology. It would be, therefore, interesting to check if this scheme can be applied also in the case of our present setup where W HI in Eq. (24) cooperates with the K's in Eqs. (9a) -(9e) where F R and F − given in Eq. (25). The necessary extra ingredients for such a scenario are described in Sec. IV A. Next, we show how we can correlate nMI with the generation of the µ term of MSSM -see Sec. IV B -and the generation of BAU via nTL -see Sec. IV C. Hereafter, we restore units, i.e., we take m P = 2.433 · 10 18 GeV.

A. RELEVANT SET-UP
Following the post-inflationary setting of Ref. [23] we consider a B − L extension of MSSM with the field content charged under B − L and R as displayed in Table 1 therein. The superpotential of the model contains W HI in Eq. (24), the superpotential of MSSM with µ = 0 and the following two terms From the terms above, the first one inspired by Ref. [24] helps to justify the existence of the µ term within MSSM, whereas the second one allows for the implementation of (type I) seesaw mechanism (providing masses to light neutrinos) and supports a robust baryogenesis scenario through nTL. Let us note that L i denotes the i-th generation SU ( been rotated in the family space so that the coupling constants λ i are real and positive. This is the so-called [16,23] N c i basis, where the N c i masses, M iN c , are diagonal, real and positive. We assume that the extra scalar fields X β = H u , H d , N c i have identical kinetic terms as the stabilizer field S expressed by the functions F lS with l = 1, 2, 3 in Eqs. (11a) -(11c)see Ref. [23]. Therefore, N S may be renamed N X henceforth. The inflationary trajectory in Eq. (12) has to be supplemented by the conditions and the stability of this path has to be checked, parameterizing the complex fields above as we do for S in Eq. (13). The relevant masses squared are listed in Table III for K = K 1 − −K 5 , with hatted fields being defined as s and s in Eq. (23) and In Table III we see that m 2 iν c > 0 and m 2 (56) From the bounds above the first one depends on φ and assumes its lowest for φ ≃ 0.1. Taking, in addition, n = 0 and, e.g., N X = 2, the equations above imply Similar bounds are obtained in Refs. [16,18,23] and should not be characterized as unnatural, given that the Yukawa coupling constant which provides masses to the up-type quarks, is of the same order of magnitude too at a high scalecf. Ref. [36].

B. SOLUTION TO THE µ PROBLEM OF MSSM
Supplementing with the soft SUSY breaking terms in Sec. IV B 2 the SUSY limit of the SUGRA potential -see Sec. IV B 1 -we can show that our model assists us to understand the origin of µ term of MSSM, consistently with the low-energy phenomenology -see Sec. IV B 3.

SUSY Potential
The SUSY limit V SUSY of V HI in Eq. (5b) is given by where K is found expanding for m P → ∞ the K's in Eqs. (9a) -(9e) with F R and F − defined in Eq. (25). Focusing on the S −Φ − Φ system we obtain (59) from which we can then compute where the matrix M ± reads (60b) After calculating K −1 and substituting it into Eq. (58) we find where the non-diagonal contributions in the F terms proportional to W Φ W * Φ * and WΦW * Φ * are cancelled out and From Eq. (61), we find that the SUSY vacuum lies along the D-flat direction |Φ| = |Φ| with As a consequence, Φ and Φ break spontaneously . Since U (1) B−L is already broken during nMI, no cosmic string are formed. Given, finally, that X β participate in K, in Eq. (59), with terms similar to the one we have for S, we can easily verify that their v.e.vs lie along the direction in Eq. (54). The required λµ values rendering our models for n = 0 compatible with the best-fit points of the CMSSM as found in Ref. [37] with the assumptions in Eq. (69).

Generation of the µ Term of MSSM
The contributions from the TeV scale soft SUSY breaking terms, although negligible during nMI may shift slightly S from zero in Eq. (63). The relevant potential terms are where m α , A λ and a S are soft SUSY breaking mass parameters. Considering V soft together with V SUSY from Eq. (61) we end up with the total low energy potential which, replacing Φ andΦ by their SUSY v.e.vs from Eq. (63), can be rephrased as Here m 3/2 is the gravitino ( G) mass and a 3/2 > 0 a parameter of order unity which parameterizes our ignorance for the dependence of |A λ | and |a S | on m 3/2 -note that the phases of A λ and a S have been chosen so that V tot is minimized and S has been rotated in the real axis by an appropriate R-transformation. The extremum condition for V tot (S) in Eq. (66a) w.r.t S leads to a non-vanishing S as follows which, although similar, is clearly distinguishable from the results obtained in Refs. [16,18,23]. The resulting µ above depends only on n and λ µ since λ/c 2 R is fixed for frozen n by virtue of Eqs. (43a) -(43c). As a consequence, we may verify that any |µ| value is accessible for the λ µ 's allowed by Eq. (57) without any ugly hierarchy between m 3/2 and µ.

Link to the MSSM Phenomenology
The subgroup, Z R 2 of U (1) R -which remains unbroken after the consideration of the SUSY breaking effects in Eq. (64) -combined with the Z f 2 fermion parity yields the well-known R-parity. This symmetry guarantees the stability of the lightest SUSY particle (LSP), providing thereby a well-motivated cold dark matter (CDM) candidate.
The candidacy of LSP may be successful, if it generates the correct CDM abundance [25] within a concrete low-energy framework. In the case under consideration [23] this could be the Constrained MSSM (CMSSM), which employs the following free parameters where signµ is the sign of µ, and the three last mass parameters denote the common gaugino mass, scalar mass and trilinear coupling constant, respectively, defined (normally) at M GUT . The parameter |µ| is not free, since it is computed at low scale by enforcing the conditions for the electroweak symmetry breaking. The values of the parameters in Eq. (68) can be tightly restricted imposing a number of cosmophenomenological constraints from which the consistency of LSP relic density with observations plays a central role. Some updated results are recently presented in Ref. [37], where we can also find the best-fit values of |A 0 |, m 0 and |µ| listed in the first four lines of Table IV. We see that there are four allowed regions characterized by the specific mechanism for suppressing the relic density of the LSP which is the lightest neutralino (χ) -note thatτ 1 ,t 1 andχ ± 1 stand for the lightest stau, stop and chargino eigenstate.
The inputs from Ref. [37] can be deployed within our setting for n = 0, if we identify, e.g., and derive first a 3/2 from Eq. (66b) -see fifth column of Table IV -and then the λ µ values which yield the phenomenologically desired |µ| -ignoring renormalization group effects. The outputs w.r.t λ µ of our computation are listed in the four rightmost columns of Table IV for K = K 1 − K 5 . From these we infer that the required λ µ values vary slightly depending on N and c R required by each K and, besides the ones written in italics, are comfortably compatible with Eq. (57). Therefore, the whole inflationary scenario can be successfully combined with all the allowed regions CMSSM besides thẽ τ 1 − χ coannihilation region for K = K 1 and K 2 . On the other hand, all the CMSSM regions can be consistent with the gravitino limit on T rh , under the assumption of the unstable G, for the T rh values, necessitated for satisfactory leptogenesissee Sec. IV C 2.

C. NON-THERMAL LEPTOGENESIS AND NEUTRINO MASSES
We below specify how our inflationary scenario makes a transition to the radiation dominated era (Sec. IV C 1) and offers an explanation of the observed BAU (Sec. IV C 2) consistently with the G constraint and the low energy neutrino data. Our results are summarized in Sec. IV C 3.

Inflaton Mass & Decay
When nMI is over, the inflaton continues to roll down towards the SUSY vacuum, Eq. (63). Soon after, it settles into a phase of damped oscillations around the minimum of V HI . The (canonically normalized) inflaton, where J is estimated from Eq. (34), acquires mass given by As we see, m δφ depends crucially on M which is bounded from above by the requirement f R = 1 ensuring the establishment of the conventional Einstein gravity at the vacuum. This bound is translated to an upper bound on the mass M BL that the B − L gauge boson acquires for φ = sgsee Table II. Namely we obtain M BL ≤ 10 14 GeV, which is lower than the value M GUT ≃ 2 · 10 16 GeV dictated by the unification of the MSSM gauge coupling constants -cf.
Refs. [18,23]. However, since U (1) B−L gauge symmetry does not disturb this unification, we can treat M BL = gM as a free parameter with g ≃ 0.5 − 0.7 being the value of the GUT gauge coupling at the scale M BL . During the phase of its oscillations at the SUSY vacuum, δφ decays perturbatively reheating the Universe at a reheat temperature given by [40] T rh = 40/π 2 g rh * where the unusual -cf. Refs. [18,23] -prefactor is consistent with w rh ≃ 0.33 [40] and we set g rh * = 228.75 as in Eq. (47a). The total decay width of δφ is found to be where the individual decay widths are where we assume that m δφ < M 3N c and the relevant coupling constants are defined as follows The decay widths above arise from the lagrangian terms describing respectively δφ decay into a pair of N c j with masses M jN c = λ jN c M , H u and H d and three MSSM (s)particles X, Y, Z involved in a typical trilinear superpotential term W y = yXY Z. Here ψ X , ψ Y and ψ Z are the chiral fermions associated with the superfields X, Y and Z whose scalar components are denoted with the superfield symbols and y = y 3 ≃ (0.4 − 0.6) is a Yukawa coupling constant of the third generation.

Lepton-Number and Gravitino Abundances
For T rh < M iN c , the out-of-equilibrium decay of N c i generates a lepton-number asymmetry (per N c i decay), ε i estimated from Ref. [39]. The resulting lepton-number asymmetry is partially converted through sphaleron effects into a yield of the observed BAU where H u ≃ 174 GeV, for large tan β and m D is the Dirac mass matrix of neutrinos, ν i . The ratio (3/2) is again [40] consistent with w rh = 0.33. The expression above has to reproduce the observational result [25] Y B = 8.64 +0.15 −0.16 · 10 −11 .
The validity of Eq. (76) requires that the δφ decay into a pair of N c i 's is kinematically allowed for at least one species of the N c i 's and also that there is no erasure of the produced Y L due to N c 1 mediated inverse decays and ∆L = 1 scatterings. These prerequisites are ensured if we impose The quantity ε i can be expressed in terms of the Dirac masses of ν i , m iD , arising from the third term of Eq. (24). Employing the (type I) seesaw formula we can then obtain the lightneutrino mass matrix m ν in terms of m iD and M iN c . As a consequence, nTL can be nicely linked to low energy neutrino data. We take as inputs the recently updated best-fit values [44] -cf. Ref. [23] -on the neutrino mass-squared differences, ∆m 2 21 = 7.56 · 10 −5 eV 2 and ∆m 2 31 = 2.55 · 10 −3 eV 2 ∆m 2 31 = 2.49 · 10 −3 eV 2 , on the mixing angles, sin 2 θ 12 = 0.321, sin 2 θ 13 = 0.02155 sin 2 θ 13 = 0.0214 and sin 2 θ 23 = 0.43 sin 2 θ 23 = 0.596 and the CP-violating Dirac phase δ = 1.4π [δ = 1.44π] for normal [inverted] ordered (NO [IO]) neutrino masses, m iν 's. Furthermore, the sum of m iν 's is bounded from above at 95% c.l. by the data [25] i m iν ≤ 0.23 eV. (79) The required T rh in Eq. (76) must be compatible with constraints on the G abundance, Y 3/2 , at the onset of nucleosynthesis (BBN), which is estimated to be [41,42] where we take into account only thermal production of G, and assume that G is much heavier than the MSSM gauginos. Non-thermal contributions to Y 3/2 [38] are also possible but strongly dependent on the mechanism of soft SUSY breaking.
No precise computation of this contribution exists within nMI adopting the simplest Polonyi model of SUSY breaking [43]. It is notable, though, that the non-thermal contribution to Y 3/2 in models with stabilizer field, as in our case, is significantly suppressed compared to the thermal one.
On the other hand, Y 3/2 is bounded from above in order to avoid spoiling the success of the BBN. For the typical case where G decays with a tiny hadronic branching ratio, we obtain [42] an upper bound on T rh , i.e., The bounds above can be somehow relaxed in the case of a stable G.

Results
Confronting with observations Y B and T rh which depend on m δφ , M iN c and m iD 's -see Eqs. (76) and (81) -we can further constrain the parameter space of the our models. In our investigation we follow the bottom-up approach detailed in Ref. [23], according to which we find the M iN c 's by using as inputs the m iD 's, a reference mass of the ν i 's -m 1ν for NO m iν 's, or m 3ν for IO m iν 's -, the two Majorana phases ϕ 1 and ϕ 2 of the PMNS matrix, and the best-fit values for the low energy parameters of neutrino physics mentioned in Sec. IV C 2. In our numerical code, we also estimate [23] the renormalization-group evolved values of the latter parameters at the scale of nTL, Λ L = m δφ , by considering the MSSM with tan β ≃ 50 as an effective theory between Λ L and the soft SUSY breaking scale, M SUSY = 1.5 TeV. We evaluate the M iN c 's at Λ L , and we neglect any possible running Resulting T rh (in GeV) T rh /10 7 2.  Table V. We take n = 0 -to avoid any tuning as regards the inflationary inputs -, λ µ = 10 −6 in accordance with Eq. (57), and M BL = 10 12 GeV. Note that we consider M BL as a free parameter since the unification value -imposed in Refs. [18,23] -is not reconciled with the reappearance of Einstein gravity at low energies, i.e., f R = 1. Setting g = 0.7 in the formula giving M BL in Table II, we obtain M = 1.43 · 10 12 GeV resulting via Eq. (71) to 2.8 ≤ m δφ /10 9 GeV ≤ 4.1 -the variation is due to the choice of K. Although this amount of uncertainty does not cause any essential alteration of the final outputs, we mention just for definiteness that we take throughout K = K 1 corresponding to m δφ = 3.3 · 10 10 GeV. We consider NO (cases A and B), almost degenerate (cases C, D and E) and IO (cases F and G) m iν 's. In all cases, the current limit in Eq. (79) is safely met. This is more restrictive than the 90% c.l. upper bound arising from the effective electron neutrino mass m β in β-decay [45] by various experiments. Indeed, the current upper bounds on m β are comfortably satisfied by the values where the lower and upper bound corresponds to case A and C respectively. The gauge symmetry considered here does not predict any particular Yukawa unification pattern and so, the m iD 's are free parameters. This fact allows us to consider m iD 's which are not hierarchical depending on the generation. Also, it facilitates the fulfilment of Eq. (78b) since m 1D affects heavily M 1N c . Care is also taken so that the perturbativity of λ iN cdefined below Eq. (75c) -holds, i.e., λ 2 iN c /4π ≤ 1. The inflaton δφ decays mostly into N c 1 's -see cases A -E. In all cases Γ δφ→N c i < Γ δφ→H and so the ratios Γ δφ→N c i / Γ δφ introduce a considerable reduction in the derivation of Y B . Namely we obtain  Table V we also display the values of T rh , the majority of which are close to 3 · 10 7 GeV, and consistent with Eq. (81) for m 3/2 1 TeV. These values are in nice agreement with the ones needed for the solution of the µ problem of MSSM -see, e.g., Table IV. Thanks to our non-thermal set-up, successful leptogenesis can be accommodated with T rh 's lower than those necessitated in the thermal regime -cf. Ref. [46].
In order to investigate the robustness of the conclusions inferred from Table V, we examine also how the central value of Y B in Eq. (77) can be achieved by varying vevM BL , or m δφ , and adjusting conveniently m 1D or M 1N c -see Fig. 4-(a) and (b) respectively. We fix again n = 0 and λ µ = 10 −6 . Since the range of Y B in Eq. (77) is very narrow, the 95% c.l. width of these contours is negligible. The convention adopted for the various lines is also depicted. In particular, we use solid, dashed and dot-dashed line when the remaining inputs -i.e. m iν , m 2D , m 3D , ϕ 1 , and ϕ 2 -correspond to the cases A, C and E of Table V, respectively. At the lower limit of these lines nTL becomes inefficient (due to low T rh ) failing to reach the value in Eq. (77). At the other end, these lines terminate at the values of m 1D beyond which Eq. (78b) is violated and, therefore, washout effects start becoming significant. Along the depicted contours, the resulting M 2N c and M 3N c remain close to their values presented in the corresponding cases of Table IV. As regards the other quantities, in all we obtain 0.04 T rh /10 8 GeV 13, (84a) 0.03 m δφ /10 10 GeV 4.64 . (84b) with the lower and upper bound obtained in the limits of the solid line which represent the most ample region of parameters satisfying the imposed requirements. As a bottom line, nTL is a realistic possibility within our setting. It can be comfortably reconciled with the G constraint even for m 3/2 ∼ 1 TeV as deduced from Eqs. (84b) and (81) adopting a sufficiently low M BL .

V. CONCLUSIONS
Motivated by the fact that a strong linear non-minimal coupling of the inflaton to gravity does not cause any problem with the validity of the effective theory up to the Planck scale, we explored the possibility to attain observationally viable nMI (i.e. non-minimal inflation) in the context of standard SUGRA by strictly employing this coupling. We showed that nMI is easily achieved, for p ≤ 4 in the superpotential of Eq. (8), by conveniently adjusting the prefactor (−N ) of the logarithmic part of the relevant Kähler potentials K given in Eqs. (9a) -(9e), where the relevant functions F R and F − are shown in Eq. (10) for a gauge-singlet inflaton. For appropriately selected integer N 's -i.e., setting n = 0 in Eq. (20) -, our models retain the predictive power of well-known universal attractor models -which employ a non-minimal coupling functionally related to the potential -and yield similar results. Allowing for non-integer N values, this predictabil-ity is lost since the observables depend on the adopted K and n in Eq. (20) and may yield any n s in its allowed region and 0.0013 ≤ r ≤ 0.02.
This scheme works also for a gauge non-singlet inflaton employing W HI shown in Eq. (24) and the functions F R and F − in Eq. (25). Embedding these models within a B − L extension of MSSM, we showed that a µ term is easily generated and the baryon asymmetry in the Universe is naturally explained via non-thermal leptogenesis. The B − L breaking scale M BL , though, has to take values lower than the MSSM unification scale. Our scenario can be comfortably tolerated with almost all the allowed regions of the CMSSM with gravitino as low as 1 TeV. Moreover, leptogenesis is realized through the out-of equilibrium decay of the inflaton to the right-handed neutrinos N c 1 and/or N c 2 with masses lower than 2.32·10 10 GeV, and a reheat temperature T rh ≤ 10 9 GeV taking M BL ≤ 10 13 GeV.