Gauged NambuJonaLasinio inflation
Abstract
We investigate the gauged NambuJonaLasinio model in curved spacetime at the large limit and in slowroll approximation.
The model can be described by the renormalization group corrected gaugeHiggsYukawa theory with the corresponding compositeness conditions.
Evaluating the renormalization group (RG) improved effective action, we show that such model can produce CMB fluctuations and find inflationary parameters: spectral index, tensortoscalarratio and running of the spectral index.
We demonstrate that the model can naturally satisfy the Planck 2015 data and maybe considered as an alternative candidate for Higgs inflation.
Keyword: gauged NambuJonaLasinio model, CMB fluctuations
PACS: 04.62.+v 12.60.Rc 98.80.Cq
1 Introduction
The standard model of the particle physics is based on the gauge symmetry and the symmetry breaking. For example, QCD is described as gauge theory. The gauge interaction causes chiral symmetry breaking. Because of the strong coupling constant, low energy phenomena are often approximated by using effective models. For instance, NambuJonaLasinio (NJL) model [1] is a wellknown model for light mesons. Gauged NJL model maybe applied to the description of QCD, as gauge theory, in which the chiral symmetry is broken and meson like composite fields are generated at high energy scale (see, for example [2]). The renormalizability of the model is precisely studied by using the ladder SchwingerDyson equation in flat spacetime [3]. In the leading order of the expansion the model can be described by a gaugeHiggsYukawa type Lagrangian with compositeness conditions [4].
Gauged NJL model is expected to be very important at the early universe^{1}^{1}1Constraints from the high energy collider physics are reviewed in Ref. [5].. The vacuum energy for a fermion field is negative, unlike to the scalar field. After the dynamical symmetry breaking a composite scalar can develop a positive vacuum energy. A primordial potential energy for the composite scalar may induce inflation, i.e. exponential expansion of our universe. An evidence of inflation is observed in the cosmic microwave background radiation (CMB) [6, 7]. In this paper we propose gauged NJL inflation model as an alternative to Higgs inflation. We evaluate the CMB fluctuations in the gauged NJL model and demonstrate that inflationary parameters are consistent with Planck bounds.
The study of NJL model in curved spacetime has been started some time ago (for review, see Ref. [8]). The curvature coupling for the composite scalar field has been obtained in Ref. [9, 10]. In the gauged NJL model it has been shown that the chiral symmetry can be broken for a strong curvature [11]. In Ref. [12] the authors have discussed the contributions from the dynamical symmetry breaking to the expansion of the universe. However, as far as we know there is only preliminary work to argue a contribution from composite fields to the CMB fluctuations. For instance, in Ref. [13] the authors have assumed nonvanishing fermion condensation and evaluated the CMB fluctuations.
This article is organized as follows: In Sec. 2 the gauged NJL model is introduced. The model is rewritten as renormalizationgroup corrected gaugeHiggsYukawa Lagrangian with compositeness conditions. In this way, this looks qualitatively similar to the study of RG improved effective potential in curved spacetime [14] and corresponding RG improved inflation [15]. In Sec. 3 we calculate the inflationary parameters in the gauged NJL model under the slow roll approximation scenario. In Sec. 4 the inflationary parameters are numerically calculated and compared with the Planck data. Then we find a possible parameters range consistent with Planck data for the gauged NJL model. In Sec. 5 we briefly discuss an exit from inflation. Some concluding remarks are given in Sec. 6.
2 Gauged NJL model
We consider fermion flavors with gauge symmetry. It is assumed that the gauge interaction for is strong enough and well described by fourfermion interaction. We employ the gauged NJL model as an effective model of the gauge theory. The Lagrangian is given by
(2.1) 
where are the generators for the flavor symmetry, shows the Lagrangian of the pure gauge theory and is the covariant derivative for the gauge invariant and generally covariant fermion kinetic term. The gauge field for the gauge symmetry is integrated out. We introduce the compositeness scale and use it to define the dimensionless fourfermion coupling . It is noted that the Lagrangian has the flavor symmetry. Thus the model is regarded as a scale up of QCD with QED interaction. The fermion fields, , are confined in baryon like states above the compositeness scale. We consider the model whose compositeness scale is higher than the one for inflation. The composite fields dynamically acquire compositeness scale mass. Only the possible environment to test the model is found at early universe.
According to the auxiliary field method the Lagrangian (2.1) is rewritten as
(2.2) 
Replacing the auxiliary fields and by the solutions of the equations of motion,
(2.3) 
the original Lagrangian (2.1) is reproduced. Starting from the Lagrangian (2.1) at the compositeness scale , one obtains the effective Lagrangian with a gaugeHiggsYukawa form below the compositeness scale through the renormalization group evolution [4],
(2.4) 
Because of the flavor symmetry one can assume that only the composite scalar field, , contributes the vacuum energy and induce exponential expansion of the universe. Below we drop the pseudoscalar field, .
The RG equations for the gauge coupling , the Yukawa coupling , the quartic scalar coupling and the curvature coupling are found in the leading order with respect to a modified expansion [11, 16],
(2.5) 
where RG parameter is defined by with the renormalization scale and a reference scale , and being the positive constants,
(2.6) 
In Ref. [17] it has been shown that the Lagrangian (2.4) is identified with (2.2) under the compositeness conditions in a flat spacetime,
(2.7) 
where is a function of and it has a constant value at the limit, [16]. It is shown that the composite scalar is conformally coupled to curvature [9]. Thus the compositeness condition for is given as,
(2.8) 
Solving the RG equations (2.5) with the compositeness conditions (2.7) and (2.8), we obtain the running couplings,
(2.9)  
(2.10)  
(2.11) 
where is the running fine structure constant for the gauge interaction, .
At the fixed gauge coupling limit, , Eqs. (2.9) and (2.10) reduce to
(2.12)  
(2.13) 
where is defined by with . We also define and at the large limit. At this limit we can explicitly solve the RG equation (2.5) for and find
(2.14) 
where denotes the renormalized fourfermion coupling and is a constant parameter, see Ref. [16]. Below we consider only the fixed gauge coupling limit, , for simplicity.
let us consider CMB fluctuations in the gauged NJL model. For this purpose, we evaluate the RG improved effective potential for the composite field in curved spacetime. We follow the procedure developed in Ref. [11]. The oneloop effective potential for the theory (2.4) up to terms linear in is given by
(2.15) 
To obtain the RG invariant effective potential of the gauged NJL model one introduces the compositeness condition into the effective potential (2.15) through the RG equations. The solution of the RG equation for the effective potential satisfies
(2.16) 
where the barred quantities are the renormalized ones at the scale . The scale is fixed to drop the logarithmic terms in the RG invariant effective potential [16],
(2.17) 
It should be noted that the gravitational couplings are not represented in Eq. (2.16). These terms do not contribute at the linear curvature approximation.
Using the oneloop effective potential (2.15) as boundary condition, we calculate the RG invariant effective potential (2.16) from Eqs. (2.12), (2.13), (2.14) and (2.17). At the limit the RG invariant effective potential is given by
(2.18) 
with
(2.19)  
(2.20)  
(2.21) 
Thus, the gravitational effective action for the composite scalar field with weak curvature in the gauged NJL model is obtained,
(2.22) 
where the reduced Planck mass is set as .
3 Inflationary parameters in gauged NJL model
The gauged NJL model may give us an alternative scenario to induce the exponential expansion of our universe. Here we assume that the typical scale of the inflation is much smaller than the compositeness scale, , and consider the effective potential (2.18) for the composite scalar field, , as an inflaton potential. Then we calculate the inflationary parameters generated by the composite scalar field.
For practical calculations it is more convenient to change the frame into the Einstein frame where the interaction term between the composite scalar and the curvature disappears. This change of the frame is achieved by the Weyl transformation,
(3.1) 
where is the metric tensor in the transformed frame and the Weyl factor, , is an analytic function with respect to the spacetime coordinates. Applying the Weyl transformation (3.1) and setting the conformal factor to be
(3.2) 
one gets the effective action (2.22) in the Einstein frame
(3.3) 
where and are the determinant of the metric tensor and the Ricci scalar in the transformed frame, respectively. In order to obtain the canonical kinetic term we redefine the auxiliary field to satisfy the relation,
(3.4) 
Then the effective action (3.3) reduces to
(3.5) 
where the subscript denotes the Einstein frame and the effective potential is defined by
(3.6) 
The effective potential in the Einstein frame depends on the parameters in the original gauged NJL model through Eqs.(2.19)  (2.21). The quantum effects are encoded via the dependence of these parameters with the compositeness condition. It should be noted that the obtained inflationary parameters are just the same in both the original and Einstein frames [18].
We will not be interesting in the exact solutions for the action (3.5) as it has been discussed in detail for various choices of the potential. We will concentrate here on quaside Sitter phase which eventually describes the inflationary dynamics. In the study of quaside Sitter phase we apply the standard slowroll approximation which is wellknown (see Refs. [19], for instance) to be consistent approximation to describe the quaside Sitter phase.
We consider a locally flat FriedmanRobertsonWalker (FRW) universe defined by the metric,
(3.7) 
where the evolution of the spacetime is described by the temporal development of the scale factor, . In the FRW universe, the equation of motion for and the Freedman equation are given by
(3.8) 
(3.9) 
with the Hubble parameter, , respectively. In the slow roll scenario we assume that the spacetime is slowly varying and impose the conditions
(3.10) 
Then we obtain the quaside Sitter phase,
(3.11) 
where is a constant parameter.
In a standard scenario of the the inflation the CMB fluctuation arises from the primordial quantum fluctuation for the inflaton field. Here we employ the slowroll approximation for the inflationary expansion of the universe. Under the approximation the inflationary parameters can be fully represented by means of the inflaton potential,
(3.12)  
(3.13)  
(3.14) 
The efolding number is given by
(3.15) 
where the integration is performed on the interval from the horizon crossing to the end of the inflation . Since the inflationary parameters have to be small, the field configuration, , can be estimated where one of the inflationary parameters exceeds order one. We noted that the change of the value for gives only a small correction to the efolding number . The dominant contribution comes from the vale at the horizon crossing, , where the derivative of the effective potential is small enough. In the other words the field configuration, is obtained as a function of the efolding number from Eq.(3.15). Then one can evaluate the density fluctuation, , by [19]
(3.16) 
The spectral index, , the tensortoscalar ratio, , and the running of the spectral index, , are represented by using the slowroll parameters [20],
(3.17)  
(3.18)  
(3.19) 
where we omit higherorder terms. Thus the observed CMB constraints for , , and restrict the possible range of the parameters in the gauged NJL model for quaside Sitter universe under discussion..
4 Numerical results
The effective action depends on the finestructure constant , the renormalized fourfermion coupling , the number of fermion species, , and the renormalization scale . In this section we numerically calculate the inflationary parameters , , and with varying the model parameters, , , , , and the efolding number, .
4.1 Small number of fermion species ()
In this subsection we consider a model with a small number of fermion species, . First we set model parameters as and study the gauge coupling dependence. The behavior of the effective potential is drawn in Fig. 2 for and . We plot the points, , and on each lines. A gentler slope of the potential is observed for a weaker gauge coupling. It should be noticed that we have to finetune the field configuration at the horizon crossing to obtain the enough efoldings in the case of the NJL model (See Appendix A). The density fluctuation decreases as the slope becomes gentle. As is shown in Fig. 2, the density fluctuation is monotonically increasing as a function of . The observed density fluctuation is obtained for .
In Table 1 the values for the inflationary parameters , and for and are shown. The spectrum index , tensorscalar ratio , and the running of the spectral index, , are consistent with the observed Planck constraints, , (68% CL, PlanckTT+lowP) and (95% CL, PlanckTT+lowP)[7]. It is also found that the dependence for the inflationary parameters , and is negligible.
0.961  0.0084  0.00077  
0.961  0.0084  0.00077  
0.961  0.0084  0.00077 
Next we discuss the dependence on the renormalized fourfermion coupling, and the renormalization scale, . Since a realistic value for the density fluctuation is obtained for , we set and numerically calculate the inflationary parameters as and vary. We show some typical values in Table 2. The and dependences are also negligible for the density fluctuation, , the spectral index, , the tensortoscalar ratio, , and the running of the spectral index, .
1  0.961  0.0083  0.00076  

1  0.961  0.0083  0.00076  
1  1  0.961  0.0084  0.00077  
1  0.961  0.0083  0.00076  
1  0.961  0.0083  0.00076 
It is found that the spectral index, the tensortoscalar ratio and the running of the spectral index are almost fixed at , , for in , case. These values are consistent with the Planck 2015 data.
4.2 Large number of fermion species
Let us investigate the inflationary parameters at fixed gauge coupling limit, for simplicity. The assumption is valid for . However, we vary the number of the fermion species, and , with a fixed gauge coupling, . Since the extremely small gauge coupling, , is considered the running of the gauge coupling in Eq. (2.5) can be neglected. The effective potential develops a nontrivial extremum for at the small gauge coupling limit. It induces the finetuning problem discussed in App. A. In order to avoid the problem we consider the limited case .
In Figs. 4 and 4 we compare the behavior of the density fluctuation between a small and a large number of fermion species. The density fluctuation grows up near for and cases. Nonnegligible dependence appears for a larger or cases. As is shown in Fig. 4 the obtained density fluctuation for approaches to the one for as the renormalization scale decreases. We note that lines for and cross twice in and a small but a finite difference remains at limit. In Fig. 4 lines for and do not cross in and a larger difference observed at the limit, .
For a larger number of fermion species the dependence appears in the spectral index, , the tensortoscalar ratio, , and the running of the spectral index, in models. We draw the behavior of these inflationary parameters on the interval in Figs. 5 and 6. It should be noticed that we normalize the mass scale by the Planck scale and consider the large limit. Such assumption is suitable for . A similar behavior is observed for and cases. Thus , and seem to depend on the product . At the limit, , the result approaches to the one for a small number of fermion species, and . The obtained trajectories are shifted as the efolding number varies. All the trajectories are consistent with the Planck 2015 observations.
The inflationary parameters depend on the renormalization scale for a larger number of the fermion species. It is observed that the parameters, , and simply approach to the values for a small number of the fermion species at the small limit. Next we fix the model parameters at , , and evaluate the and dependences. The behavior of the density fluctuation, , is plotted as and vary in Figs. 8 and 8, respectively. The range of the density fluctuation is smaller than the one in Fig. 2. The number of the fermion species induces a subdominant contribution to the density fluctuation compared with the gauge coupling. Thus, we can obtain a suitable density fluctuation by tuning the strength of the gauge coupling.
In Figs. 9 and 10 the tensortoscalar ratio, , and the running of the spectral index, , are plotted as functions of the spectral index, , for the efolding number, and . Both the figures show a similar behavior to Figs. 5 and 6. A model with a large number of fermion species at and a small number of fermion species at a small generate nearly equal values of , and . The obtained trajectories are consistent with the Planck 2015 results. The tensortoscalar ratio, , monotonically decreases as and increases. The spectral index, and the running of the spectral index, have maximal values at .
Thus, we demonstrated that realistic gauged NJLmodel under discussion for specific values of parameters is consistent with Planck cosmological data. From another side, the model under discussion is realistic high energy physics model of elementary particles. For numerical calculations, we considered several choices of fermion species so that such choice is quite standard in particle physics. Thus, we arrive to realistic (composite) particle physics model which predicts viable inflation.
5 Exit from inflation
The end of the slow roll era is found by observing the parameters, and . In the present model the parameter exceeds unity at in Fig. 2. Under the slow roll approximation, the deceleration parameter, , is written as
(5.1) 
The deceleration parameter changes the sign at and develops a positive value. Thus the universe turns to decelerated expansion era.
In this paper we set the initial condition to generate the efolding number between 50 and 60. Thus the duration of the inflation can be estimated by
(5.2) 
We numerically find
(5.3) 
From Eqs. (3.8) and (3.9) we obtain
(5.4) 
We define the perturbation by
(5.5) 
with . Inserting Eq. (5.5) into Eq. (5.4), we get
(5.6) 
The solution of this equation is given by
(5.7) 
where we set an initial condition and is a constant parameter. Thus the de Sitter solution (3.11) is unstable. As is shown in Fig. 2, the potential energy extremely decreases at . Then the dominant contribution to derive the expansion of the universe comes from the radiation. Therefore, the standard cosmology is recovered. It should be noted that terms induced by the trace anomaly also unstabilizes the de Sitter solution and then the radiation dominant era realizes [23].
6 Conclusions
In summary, we proposed gauged NJL inflation as an alternative for Higgs inflation. Specifically, the gauged NJL model is studied as the realistic particle physics model for the composite scalar. Applying the auxiliary field method and performing the renormalization group improvement (for review of RG in curved spacetime, see [21]), the model is represented by the gaugeHiggsYukawa theory. We assumed that the gaugeHiggsYukawa theory well describes the model at the scale of inflation at the limit. Then the produced CMB fluctuations have been calculated in the gaugeHiggsYukawa theory with the compositeness conditions at the fixed gauge coupling.
The behavior for inflationary parameters , , and has been numerically evaluated as a function of the couplings, the renormalization scale, the efolding number and the number of the fermion species. It is found that the smallness of the density fluctuation can be realized by tuning the gauge coupling. We proved that the model with a small number of fermion species, satisfies the Planck 2015 data. In this case the spectral index, the tensortoscalar ratio and the running of the spectral index develop almost fixed values, , , for . For a large number of fermion species, these inflationary parameters depend on the renormalization scale . However, the obtained trajectories of the inflationary parameters are consistent with the Planck 2015 data. Therefore, the gauged NJL model naturally generates inflationary parameters consistent with the Planck 2015 observation and may serve as viable inflationary theory.
At the next step, one has to consider the end of inflation and graceful exit. It is clear that higherderivative RG corrected terms (like , etc) which are induced as vacuum polarization beyond the linearcurvature approximation may naturally provide the graceful exit. It is also interesting to consider the reheating process in the gauged NJL model. In fact, this goes beyond this work. However, the preliminary considerations of preheating process indicate reheating maybe successful as it is in usual scalar theories of inflation (in fact, our model is effectively scalar inflationary theory). These questions will be considered elsewhere. Furthermore, NJL model may pretend to the role of dark matter (see, for instance, Ref. [22]), hence presumably unifying inflation with dark matter.
Acknowledgements
The authors are grateful to T. Kugo for useful comments. The work by TI is supported in part by JSPS KAKENHI Grant Number 26400250 and that by SDO is supported in part by MINECO (Spain), project FIS201344881 and by MES (Russia).
Appendix A NJL limit
Let us discuss the slowroll solution in the NJL model. It is obtained by omitting the gauge interaction. Here we consider the NJL limit, , where the gauge interaction is eliminated. At the limit the running couplings (2.9) and (2.10) reduce to
(A.1)  
(A.2) 
From Eqs. (2.14) and (A.1) we obtain
(A.3) 
where we renormalize suitably to obtain a finite expression [16].
Following the similar procedure in Sec. 2 we calculate the RG invariant effective potential at the limit and obtain
(A.4) 
Thus, the gravitational effective action with the weak curvature becomes
(A.5) 
By the Weyl transformation (3.1), one can eliminate the interaction term between the curvature and the auxiliary field . In the Einstein frame the action is simplified as
(A.6) 
with the effective potential
(A.7) 
The denominator comes from the NJL limit of in Eq. (3.2). It suppresses the effective potential for a large . The effective potential (A.7) has a maximum at .
In Fig. 12 the typical behavior of the effective potential (A.7) is shown. It is observed that the horizon crossing should take place near the maximum of the potential to obtain large enough efolding . The efolding is plotted as a function of in Fig. 12. It blows up at . We noted that the dependence gives only a negligible contribution to the efolding. The result introduces a fine tuning problem why . To avoid this finetuning problem one should modify the potential. Since the potential can be easily modified by including terms which are neglected in our approximation, the finetuning problem seems to be not so serious.
References
 [1] Y. Nambu and G. JonaLasinio, Phys. Rev. 122 (1961) 345; Phys. Rev. 124 (1961) 246.

[2]
V. A. Miransky,
Dynamical Symmetry Breaking in Quantum Field Theories,
World Scientific (1993);
M. Harada and K. Yamawaki, Phys. Rept. 381 (2003) 1 [hepph/0302103]. 
[3]
K.I. Kondo, S. Shuto and K. Yamawaki,
Mod. Phys. Lett. A6 (1991) 3385;
K.I. Kondo, M. Tanabashi and K. Yamawaki, Progr. Theor. Phys. 89 (1993) 1249; Mod. Phys. Lett. A8 (1993) 2859.  [4] C. N. Leung, S. T. Love and W. A. Bardeen, Nucl. Phys. B 273 (1986) 649; Nucl. Phys. B 323 (1989) 493.
 [5] C. T. Hill and E. H. Simmons, Phys. Rept. 381 (2003) 235 [Phys. Rept. 390 (2004) 553] [hepph/0203079].
 [6] P. A. R. Ade et al. [BICEP2 and Planck Collaborations], Phys. Rev. Lett. 114 (2015) 101301 [arXiv:1502.00612 [astroph.CO]].
 [7] P. A. R. Ade et al. [Planck Collaboration], arXiv:1502.01589 [astroph.CO]. arXiv:1502.02114 [astroph.CO].
 [8] T. Inagaki, T. Muta and S. D. Odintsov, Mod. Phys. Lett. A 8 (1993) 2117 [hepth/9306023]; Prog. Theor. Phys. Suppl. 127 (1997) 93 [hepth/9711084].
 [9] C. T. Hill and D. S. Salopek, Annals Phys. 213 (1992) 21.
 [10] T. Muta and S. D. Odintsov, Mod. Phys. Lett. A 6 (1991) 3641.
 [11] B. Geyer and S. D. Odintsov, Phys. Lett. B 376 (1996) 260 [hepth/9603172]; Phys. Rev. D 53 (1996) 7321 [hepth/9602110].
 [12] T. Inagaki, Y. Rybalov and X. Meng, Eur. Phys. J. C 71 (2011) 1656 [arXiv:1102.5036 [hepph]].
 [13] S. Iso, K. Kohri and K. Shimada, Phys. Rev. D 91 (2015) 044006 [arXiv:1408.2339 [hepph]].
 [14] E. Elizalde and S. D. Odintsov, Phys. Lett. B 303 (1993) 240 [hepth/9302074]; Phys. Lett. B 321 (1994) 199 [hepth/9311087];Z. Phys. C 64 (1994) 699 [hepth/9401057].

[15]
A. De Simone, M. P. Hertzberg and F. Wilczek,
Phys. Lett. B 678 (2009) 1
[arXiv:0812.4946 [hepph]];
S. Mukaigawa, T. Muta and S. D. Odintsov, Int. J. Mod. Phys. A 13 (1998) 2739 [hepph/9709299];
H. M. Lee, Phys. Lett. B 722 (2013) 198 [arXiv:1301.1787 [hepph]];
N. Okada and Q. Shafi, arXiv:1311.0921 [hepph]; G. Barenboim, E. J. Chun and H. M. Lee, Phys. Lett. B 730 (2014) 81;
Y. Hamada, H. Kawai, K. y. Oda and S. C. Park, Phys. Rev. Lett. 112 (2014) 241301 [arXiv:1403.5043 [hepph]];
R. P. Woodard, Phys. Rev. Lett. 101 (2008) 081301 [arXiv:0805.3089 [grqc]];
I. Oda and T. Tomoyose, Adv. Stud. Theor. Phys. 8 (2014) 551 [arXiv:1404.1538 [hepph]];
J. Ren, Z. Z. Xianyu and H. J. He, JCAP 1406 (2014) 032 [arXiv:1404.4627 [grqc]];
Y. Hamada, H. Kawai and K. y. Oda, JHEP 1407 (2014) 026;
T. Inagaki, R. Nakanishi and S. D. Odintsov, Astrophys. Space Sci. 354 (2014) 2108 [arXiv:1408.1270 [grqc]]; Phys. Lett. B 745 (2015) 105 [arXiv:1502.06301 [hepph]];
E. Elizalde, S. D. Odintsov, E. O. Pozdeeva and S. Y. Vernov, Phys. Rev. D 90 (2014) 084001 [arXiv:1408.1285 [hepth]];
Y. Hamada, H. Kawai, K. y. Oda and S. C. Park, Phys. Rev. D 91 (2015) 053008 [arXiv:1408.4864 [hepph]];
H. J. He and Z. Z. Xianyu, JCAP 1410 (2014) 019 [arXiv:1405.7331 [hepph]];
M. Herranen, T. Markkanen, S. Nurmi and A. Rajantie, Phys. Rev. Lett. 113 (2014) 211102 [arXiv:1407.3141 [hepph]];
M. Herranen, A. Osland and A. Tranberg, arXiv:1503.07661 [hepph].  [16] M. Harada, Y. Kikukawa, T. Kugo and H. Nakano, Prog. Theor. Phys. 92 (1994) 1161 [hepph/9407398].
 [17] W. A. Bardeen, C. T. Hill and M. Lindner, Phys. Rev. D 41 (1990) 1647.
 [18] D. I. Kaiser, Phys. Lett. B 340 (1994) 23 [astroph/9405029]; Phys. Rev. D 52 (1995) 4295 [astroph/9408044].

[19]
A. D. Linde,
Contemp. Concepts Phys. 5 (1990) 1
[hepth/0503203];
D. S. Gorbunov and V. A. Rubakov, “Introduction to the theory of the early universe: Cosmological perturbations and inflationary theory,” Hackensack, USA: World Scientific (2011) 489 p  [20] K. Kohri, Y. Oyama, T. Sekiguchi and T. Takahashi, JCAP 1310 (2013) 065 [arXiv:1303.1688 [astroph.CO]].
 [21] I. L. Buchbinder, S. D. Odintsov, I. L. Shapiro, ”Effective Action in Quantum Gravity,” Bristol, UK: IOP (1992).
 [22] M. Holthausen, J. Kubo, K. S. Lim and M. Lindner, JHEP 1312 (2013) 076 [arXiv:1310.4423 [hepph]].
 [23] K. Bamba, R. Myrzakulov, S. D. Odintsov and L. Sebastiani, Phys. Rev. D 90 (2014) 043505 [arXiv:1403.6649 [hepth]].