Contents lists available at ScienceDirect

Physics Letters B

www.elsevier.com/locate/physletb

Primordial blackholes and gravitational waves for an inflection-point model of inflation v-/

CrossMark

Sayantan Choudhurya*, Anupam Mazumdar

a Physics and Applied Mathematics Unit, Indian Statistical Institute, 203 B.T. Road, Kolkata 700 108, India b Consortium for Fundamental Physics, Physics Department, Lancaster University, LA1 4YB, UK

ARTICLE INFO

ABSTRACT

Article history:

Received 5 February 2014

Received in revised form 24 April 2014

Accepted 27 April 2014

Available online 1 May 2014

Editor: M. Trodden

In this article we provide a new closed relationship between cosmic abundance of primordial gravitational waves and primordial blackholes that originated from initial inflationary perturbations for inflection-point models of inflation where inflation occurs below the Planck scale. The current Planck constraint on tensor-to-scalar ratio, running of the spectral tilt, and from the abundance of dark matter content in the universe, we can deduce a strict bound on the current abundance of primordial blackholes to be within a range, 9.99712 x 10-3 < &PBHh2 < 9.99736 x 10-3.

© 2014 The Authors. Published by Elsevier B.V. This is an open access article under the CC BY license

(http://creativecommons.org/licenses/by/3.0/). Funded by SCOAP3.

1. Introduction

In the Einstein general relativity (GR) the primordial blackholes (PBHs) with a small mass can be created during the radiation epoch due to over density on length scales R ~ 1 /kPBH, which is typically much smaller than the pivot scale at which the relevant perturbations re-enter the Hubble patch for the large scale structures, [1-3]. Typically the regions with a mass less than the size of the Hubble radius can collapse to form PBHs, i.e. M < yMH ~ y(4n/3)pH-3(t) ^ 2 x 105y(t/1 s)M0, where p is the energy density of the radiation epoch, H is the Hubble radius, M0 ~ 1033 g, and y < 0.2 is the numerical factor during the radiation era which depends on the dynamics of gravitational collapse [2]. For instance, an economical way would be to create PBH abundance from an initial primordial inflationary fluctuations which had already entered the Hubble patch during the radiation era, but whose amplitude had increased on small scales due to the running in the spectral index, ns [4,5].1

An interesting observation was first made in Ref. [8] and in Refs. [9-14], that a sub-Planckian inflaton field can create a significant primordial gravitational waves (PGWs) provided the last 50-60 e-foldings of inflation are driven through the inflection-point,

* Corresponding author.

1 A word of caution - GR is not an ultraviolet (UV) complete theory. An UV completion of gravity may naturally lead to ghost free and asymptotically free theory of gravity, as recently proposed in Refs. [6,7]. In such a class of theory it has been shown that mini-blackhole with a mass less than the Planck mass, i.e. 10-5 g does not have a singularity and nor does it have a horizon [6].

such that the tensor to scalar ratio saturates the Planck constrain, r < 0.12 [15]. One requires a marginal running in the power spectrum which is now well constrained by the Planck+WMAP9 combined data. A valid particle physics model of inflation can only occur below the cut-off scale of gravity, see for a review on particle physics models of inflation [16]. It would be interesting to study the implications of the running of the spectral tilt, as, for both PGWs and PBHs.

Formation of the significant amount of PBHs on a specific mass scale is realized iff the power spectrum of primordial fluctuations has amplitude 10-2-10-1 on the corresponding scales [17]. In such a physical situation the second-order effects in the cos-mological perturbation are expected to play a significant role in the present set up. Also such non-negligible effects generate tensor fluctuations to produce PGWs from scalar-tensor modes via terms quadratic in the first-order matter and metric perturbations [18-20]. Most importantly, their amplitude may well exceed the first-order tensor perturbation generated by quantum fluctuation during inflation in the present set up as the amplitude of density fluctuations required to produce PBHs is large.

The aim of this paper is to provide a unique link between the current abundance of PBHs, i2PBH(no) = pPBH/pc, and the abundance of primordial gravitational waves £2qw = Pgw/Pc in our universe originated from the primordial fluctuations, where no is the present conformal time and Pc denotes the critical energy density of the universe. With the help of Planck data, we will be able to constrain a concrete bound on i2PBHh2.

http://dx.doi.org/10.1016/j.physletb.2014.04.050

0370-2693/© 2014 The Authors. Published by Elsevier B.V. This is an open access article under the CC BY license (http://creativecommons.org/licenses/by/3.0/). Funded by SCOAP3.

Fig. 1. In (a), we show the scalar spectral index ns(k), and in (b), we show the running of the scalar spectral index as(k), with respect to the momentum scale k. The black dotted line corresponds to kA = 0.056 Mpc-1 for lA = 2500, the blue dotted line corresponds to kL = 4.488 x 10-5 Mpc-1 for lL = 2, and in all the plots violet dashed dotted line represents the pivot scale of momentum at kt = 0.002 Mpc-1 for lt ~ 80 at which PS(kt) = 2.2 x 10-9, nS(kt) = 0.96, aS(kt) = -0.02. In both (a) and (b) the PBH formation scale kPBH is lying within the region bounded by the blue dotted line and the violet dashed dotted line. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)

2. PBH formation

expanded around the pivot scale (k*) as:

Let us first start the discussion with the amplitude of the scalar power spectrum, which is defined at any arbitrary momentum scale lying within the window, kL < k < kA, by:

/ k \ns-1+^ k)+KS ln2(k*)+••• Ps(k) = Ps(k*)l -j , (1)

where the parameters ns, as and Ks are spectral tilt, running and running of running of the tilt of the scalar perturbations defined in the momentum pivot scale k*. Also note that the PBH formation scale is lying within the window, kL < kPBH < k* < kA. In the realistic situations the upper and lower bound of the momentum scale is fixed at, kA = 0.056 Mpc-1 for lA = 2500 and ki = 4.488 x 10-5 Mpc-1 for lL = 2 as explicitly depicted/shown in Fig. 1.

Within this window, we need to modify the power law parameterization of the power spectrum by incorporating the effects of higher order Logarithmic corrections in terms of the non-negligible running, and running of the running of the spectral tilt as shown in Eq. (1), which involves higher order slow-roll corrections in the next to leading order of effective field theory of inflation. This is important to consider since Planck-WMAP9 combined data have already placed interesting constraints on ns(k*) = 0.9603 ± 0.0073 (within 2a C.L.), as(k*) = 0.0134 ± 0.0090, and

ks(k*) = 0.020+0.016 (within 1a-1.5a C.L.) [15].

Further using Eq. (1), spectral tilt, running of the tilt, and running of the running of the tilt for the scalar perturbations can be written at any arbitrary momentum scale within the widow, ki=2 < k < k(=2500, as:

... .d ln Ps (k) ns (k) = 1 +

ns(kpBH )

as (k) =

Ks(k) =

d ln k

= ns(k*)+ as(k*)ln dns(k)

Ks (k*)

,, . = as(k*) + ks(k*) ln. d ln k \ k*

das(k) d2ns (k)

'Ks (k*) + ■

d ln k d ln k2

At the scale of PBH formation, k = kPBH, the value of the tilt, and running of the running of tilt for the scalar perturbations can be

j-i s . ,, ,, fkpBH\ , Ks(k*) 2/kpBH\ ,

-ns (k*) + as (k*) lnl — I + ——In21 — I +

as (kpBH) = as(k*) + ks (k*) ln

Ks(kpBH) ~ Ks(k*) +----,

(6) (7)

provided the expansion is valid when, kPBH < k*. Throughout the article we will fix the pivot scale to be the same as that of the Planck, k* ~ 0.002 Mpc-1. However, it is important to note that the physics is independent of the choice of the numerical value of k*. The ••• represent higher order slow-roll corrections appearing in the expansion. Here the pivot scale of momentum k* is a normalization scale and is of the order of the UV regularized scale of the momentum cut-off of the power spectrum beyond which the logarithmically corrected power law parameterization of the primordial power spectrum for scalar modes does not hold good. More precisely, k* be a floating momenta in the present context. Additionally, we have used another restriction on the momentum scale, ki=2 < kPBH < k* < k(=2500. For more details see Eq. (1) mentioned later. The initial PBHs mass, MPBh, is related to the Hubble mass, M, by:

Mpbh = m y = — yph , (8)

at the Hubble entry, where the Hubble parameter h is defined in terms of the conformal time, n. The PBH is formed when the density fluctuation exceeds the threshold for PBH formation given by the Press-Schechter theory [21]

(2) f (> M) = 2y j d&P(0; M(kPBH)).

Here P(0; M(kPBH)) is the Gaussian probability distribution function of the linearized density field 0 smoothed on a scale, kPBH = 1/R, by [22]:

P(0; kPBH) =

(kpBH)

2S0 (kpBH )

where the standard deviation is given by

£& (kpBH ) =

[ dk ( k2 f exp(-kkkBHH

P @(k).

Here it is important to note that the fine details of our conclusions might change while taking into N body simulations into account [23-26]. For a generic class of inflationary models, linearized smooth density field 0(k), and the corresponding power spectrum P0(k) can be written as:

0(k) =

Rc (k),

P of =

5 V aUJ 4 (1 + w)2

25 (1 + 5w)2 VaU)

Ps (J),

where w represents the effective equation of state parameter after the end of inflation. Assuming that the inflaton decays into the relativistic species instantly, we may be able to fix w & 1/3, for a radiation dominated universe. Additionally, Rc (k) characterizes the curvature perturbation, and PS denotes the amplitude of the scalar power spectrum.

Now substituting Eq. (12) and Eq. (1) in Eq. (11), for kPBH = 1 /R, we can express £0(kPBH) as:

£& (kPBH )

= 2 (1 + w)VPJf)/ 2 = 5 (1 + 5 w ) \au)

dk k2 — exp --3— k, V k2PBH

ns+2 + ln(f) + Ks ln2(f) + ■■

1 (1 + w)VPjf)/ k, 5 (1 + 3 w ) yaH,

x Va + Bns(J,) + Cas(k,) + DKs(k,) +----

where we have reparametrized the integral in terms of the regulated UV (high) and IR (low) momentum scales. The cut-offs (kA and kL ) are floating momenta to collect only the finite contributions. The technique we employ here has a similarity to the cut-off regularization scheme, which is being introduced in such a fashion that after taking the physical limit, (kA ^ w, kL ^ 0), the result returns to the original infinite integral.

Here the UV and the IR cut-offs must satisfy the constraint condition, kL ^ kPBH ^ k, ^ ka, for which the integral appearing in the expression for the standard deviation can be regularized. In Eq. (13), A, B, C and D are all momentum dependent coefficients which are explicitly mentioned in Appendix A, see Eq. (23). Moreover, at the Hubble exit an additional constraint, k, = aU, will have to be satisfied in order to do the matching of the long and short wavelength perturbations.

Hence, substituting the explicit expressions for PS, ns, as and Ks in presence of the higher order corrections at the pivot scale k,, the simplified expression for the regularized standard deviation in terms of the leading order slow-roll parameters can be written as:

£0 (Jpbh )

(1 + w)

cy (J,)

8V6n M2pl(1 + § w )

1 + T (CE + 7 )nv (J, ) -

-— £v (k,) + ^ f ) + ■

2Ce + 11 jcy (k,)

where CE = 4(ln2 + yE) — 5, and yE = 0.5772 is the Euler-Mascheroni constant. Here the (ev, nv ) are slow roll pa-

rameters for a given inflationary potential V ($). It is important to mention that the results obtained in this paper are inflation centric - true only for inflection point models of inflation.

3. PBH and GW for sub-PIanckian model of inflation

For a successful inflation, the potential should be flat enough, and for a generic inflationary potential around the vicinity of the VEV 00, where inflation occurs, we may impose the flatness condition such that, V"(fo) ^ 0. This yields a simple flat potential which has been imposed in many well motivated particle physics models of inflation with an inflection-point [27]:

V(0) = a + № - 00) + Y(0 - M3 + k(0 - 00)4 + ■

where a denotes the height of the potential, and the coefficients /), y , k determine the shape of the potential in terms of the model parameters. Note that at this point, we do not need to specify any particular model of inflation for the above expansion of V (0). However, not all of the coefficients are independent once we prescribe the model of inflation here. This only happens if the VEV of the inflaton 00 < Mp must be bounded by the cut-off of the particle theory, where the reduced Planck mass Mp = 2.4 x 1018 GeV.

We are assuming that in 4 dimensions Mp puts a natural cutoff here for any physics beyond the Standard Model. The another assumption we have made here is that the range of flatness of the potential of inflation \A$\ = \<pcmb — O(10—1 Mp) < Mp, for which the model is fully embedded within a particle theory such as that of gauge invariant flat directions of minimal supersymmet-ric Standard Model (MSSM), or MSSM x U(1)b—l [27]. Here 0cmb and are the inflaton field at the Hubble crossing and the end of inflation respectively.

Both these sub-Planckian constraints lead to the observed tension of the low power of the CMB at low-l with the high-l, which leads to a negative curvature of the power spectrum at small k and a positive curvature of the power spectrum at large k, see [28]. For an inflection-point model of inflation this will provide an improved constraint on PBH formation and GW waves via running in the power spectrum (see Fig. 2 for the details).

The fractional density of PBH formation can be calculated as:

f (> M) = y erf

*j2£o (jpbh )

In general the mass of PBHs is expected to depend on the amplitude and the shape of the primordial perturbations. The relation between the PBH formation scale (kPBH = 1/R ) and the PBH mass can be expressed as:

kPBH =

VY ( m pbh 5.54 x 10-24 V 1 g

J \3.36

Moreover, we can express the fractional density of PBH formation in terms of the PBH abundance at the present epoch, no, as [18]:

= 1014 x f( MPBg)!^

Fig. 2. QGWh2 and MPBH have been plotted in a logarithmic scale where left and right wedge-shaped red colored curves indicate power spectra of GWs from two different peaked fluctuations corresponding to the set of values for (QPBHh2, MPBH) =

5,30MG) (left) and (10-1,1022 g) (right) for relativistic degrees of freedom 228.75, from Eq. (18) and Eq. (19). The green and yellow solid line shows

an envelope curve, corresponding to high V*1/4 = 1.96 x 1016 GeV, and low scale V*/4 = 6.48 x 108 GeV, sub-Planckian inflationary models obtained by varying the scale of PBH formation kPBH and the tensor-to-scalar ratio r, which depend on the frequency of GW in a logarithmic scale. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)

The recent observations from Planck puts an upper bound on the amplitude of primordial gravitational waves via tensor-to-scalar ratio, r(k*) = PT/PS. This bounds the potential energy stored in the inflationary potential, i.e. V* < (1.96 x 1016 GeV)4(r(k*)/0.12) [15].

With the help of Eqs. (10), (1), (14), (17), (18), we can link the GW abundance at the present time:

ßGwh2(no) =

= ( ZeikpBH) W Mpbh \1/2 ßpBHh2 f -1

v 10-2 J V1020 g/ 1-7 X1021

In order to realize inflation below the Planck scale, i.e. Mp, we need to observe the constraint on the flatness of the potential, i.e. , via the tensor-to-scalar ratio as recently obtained in Ref. [14]2:

|A0| Mp

r(k*) nv (k*)

- 1 +•

Note that it is possible to saturate the Planck upper bound on tensor-to-scalar ratio, i.e. r(k*) < 0.12 for small field excursion characterized by, 00 < Mp and \A$\ < Mp. Additionally, ••• contain the higher order terms in the slow roll parameters which only dominate in the next to leading order.

Collecting the real root of the tensor-to-scalar ratio, r, in terms of the field displacement \A$\ from Eq. (20), at the leading order, we can derive a closed constraint relationship between QGW and QPBH at the present epoch, for inflection point inflationary models:

ßGwh2 <

6 x 10-18 ( Mpbh \ 2 O2BHßpBHh2 Y V102°g/ erf( )

\¡20pBH

where we introduce a new parameter OPBH, which can be expressed in terms of the inflationary observables as mentioned in

2 There was a very minor numerical error in earlier computation in Refs. [10,13], which we have now corrected it in Ref. [14]. But the final conclusion remains unchanged due to such correction.

Appendix A, see Eq. (23). For definiteness, we also identify the PBH mass with the horizon mass when the peak scale is within the sub-hubble region. An interesting observation can be made here; typically the PBH mass can also be related to the peak frequency of the GWs produced from the collapse of an over-densed region of the space-time to form PBHs, see [18], by following the linearized gravitational wave fluctuations up to the second order [19]. The peak frequency is given by: fGWW' = kp/V3na0, where kp is the peak value of the momentum scale, and a0 is the scale factor at the present epoch. Therefore, the final expression yields for frequency and the amplitude [18]:

fGwewk=0.03 Hz

Mpbh V1/2 102°gy

AGWk = 6 X 10-8

where S0(kPBH &kp) is explicitly defined in Eq. (14), see Ref. [18].

It is important to note that the space-based laser interferometers are sensitive to GWs with frequency range 10-5 Hz < f < 10 Hz, which covers the entire mass range of the PBHs, 1020 g < MPBH < 1026 g, which comes from the stringent Dark Matter constraint, QDMh2 = 0.1196 ± 0.0031 from Planck+WMAP9 combined dataset within 1.5a C.L. [28].

LISA [29] can probe up to its best sensitivity QGWh2 ~ 10-11 at GW frequency f ~ 10-2 Hz corresponding to the PBH mass MPBH ~ 1021 g, DECIGO/BBO [30] and the ultimate-DECIGO [31] are designed to probe up to QGWh2 ~ 10-13 and QGWh2 ~ 10-17, respectively at the peak frequency f ~ 10-1 Hz with PBH mass MPBH ~ 1019 g in its near future run [32,33]. On the other hand the sensitivity of LIGO [34] is too low at present and in the near future to detect the primordial GWs. This implies that for LIGO the abundance of the PBHs are constrained within the PBH mass MPBH < 7 x 1014 g with effective GW frequency fGW > 10 Hz which cannot be observed at the present epoch.

Constraints from all of these GW detectors represented by convex lines with different color codes in logarithmic scale in Fig. 2. We have also shown the variation of GW abundance for low (green) and high (yellow) scale sub-Planckian models by varying PBH mass (MPBH) and tensor-to-scalar ratio (r) using Eq. (21) and Eq. (19) in Fig. 2. Additionally, we have shown the two wedge-shaped curves shown in red represented by (QPBHh2, MPBH) = (10-5, 30M0) (left) and (10-1,1022 g) (right) for relativistic degrees of freedom g* = 228.75. The appearance of the sharp peaks in the left and right wedge shaped red curves signify the presence of maximum value of the GW abundances at the present epoch corresponding to the peak frequency given by Eq. (22). Each wedge shaped curves accompany smooth peaks, this is due to the resonant amplification procedure when the peak width for fluctuation, Ap ^ kp/2. If the peak width exceeds such a limit then the frequency of the fluctuations will increase and we get back the peak for sharp fluctuation in the right side for each of the wedge shaped curve.

In Fig. 3, we have shown the behavior of the PBH abundance with running of the spectral tilt within the Planck 2a C.L. (black region) of spectral-tilt [15]. We have explicitly shown the 1a allowed constraint on the running of the spectral tilt by the green shaded region which additionally puts a stringent constraint on the PBH abundance within a tiny region 9.99712 x 10-3 < S2PBHh2 < 9.99736 x 10-3. Note that if we incorporate Planck + WMAP9 + high L + BICEP2 data [35], our results would modify, although the physics behind the mechanism would remain be the same. We would like to revisit the problem in future with a more detailed study.

fîpBH" vs as plot

0.009975

0.009974

0.009973

0.009972

0.009971

0.009970

Planck 1 cr allowed region

-0.025

-0.020

-0.015

-0.010

-0.005

Fig. 3. We have plotted S2PBHh2 with respect to the running of the tilt, as. The black colored band shows Planck allowed region for ns within the range 0.955 < ns < 0.970 at 2a C.L. within the threshold interval 0.2 < ©th < 0.8 of Press-Schechter theory. The green band for the 1a allowed region by Planck from the constraints on the running of the spectral tilt (as) within the range —0.022 <as < —4 x 10—3. This additionally puts a stringent constraint on the PBH abundance within a region 9.99712 x 10—3 < QPBHh2 < 9.99736 x 10—3. Here the red straight line is drawn for the central value of the spectral-tilt ns = 0.962 and 0th = 0.5. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)

4. Conclusion

vn I<pbh 4 k.

ln2(f k

k*) ln(k* - I.I ) 'F»

2 ' 2 ' 2

1 1 2 ' 2

3 3 2 ' 2

kPBH ■

3 3 3 2 ' 2 ' 2

kPBH ■

\fn IkPBH

T2 ~k**~ k 2k**

k ' kpBH

lnM I k

-< k* ^

i) pFq 1 1 1 2 ' 2 ' 2

1111 2 ' 2 ' 2 ' 2

1 1 2 ' 2

3 3 3 2 ' 2 ' 2

3 3 2 ' 2

kPBH -I

3 3 3 3 2 ' 2 ' 2 ' 2

k2 kA k2

kPBH -I

Opbh =

5vA(1 + w)(8.17 x 10-3)2

12\f2.n (1 + 3 w )

2Bnv (k*) ,

x <1 +---+

1 + 5 A +

B fr(k*)\ 3B

sööäI012 - Aev(k*)

C 2 2D 3

- 5A^ ( ) + 5A a3 (k*) + '

where pFq represents generalized hypergeometric function. References

To summarize, we have shown that it is possible to establish a generic relationship between PBH and GW abundance for a sub-Planckian model of inflation with a flat potential, where inflation is driven near an inflection-point. For such a class of model it is possible to predict QcWh2 and £2PBHh2 with the help of this new expression given by Eq. (21). We have used important constraints arising from various GW detectors, which we have shown in Fig. 2, and the PBH abundance with running of the spectral tilt in Fig. 3.

Acknowledgements

A.M. would like to thank Andrew Liddle for helpful discussions. S.C. thanks Council of Scientific and Industrial Research, India for financial support through Senior Research Fellowship (Grant No. 09/093(0132)/2010). A.M. is supported by the Lancaster-Manchester-Sheffield Consortium for Fundamental Physics under STFC grant ST/J000418/1.

Appendix A

The momentum dependent co-efficients appearing in Eq. (13) and OPBH appearing in Eq. (21) are given by:

Vn kPBH 2 k*

Vn kPBH 2 k*

' k ' J<PBH. k ' kPBH

1 + ln

1 1 2 ' 2

3 3 2 ' 2

k2 kA k2

ka kL '

[1] S. Hawking, Mon. Not. R. Astron. Soc. 152 (1971) 75;

B.J. Carr, S.W. Hawking, Mon. Not. R. Astron. Soc. 168 (1974) 399.

[2] B.J. Carr, Astrophys. J. 201 (1975) 1.

[3] B.J. Carr, J.H. Gilbert, J.E. Lidsey, Phys. Rev. D 50 (1994 ) 4853, arXiv:astro-ph/9405027.

[4] D. Lyth, K.A. Malik, M. Sasaki, I. Zaballa, J. Cosmol. Astropart. Phys. 0601 (2006), 011, arXiv:astro-ph/0510647.

[5] M. Drees, E. Erfani, J. Cosmol. Astropart. Phys. 1104 (2011) 005, arXiv:1102. 2340 [hep-ph];

M. Drees, E. Erfani, J. Cosmol. Astropart. Phys. 1201 (2012) 035, arXiv:1110. 6052 [astro-ph.CO].

[6] T. Biswas, E. Gerwick, T. Koivisto, A. Mazumdar, Phys. Rev. Lett. 108 (2012) 031101, arXiv:1110.5249 [gr-qc].

[7] T. Biswas, A. Conroy, A.S. Koshelev, A. Mazumdar, Class. Quantum Gravity 31 (2013) 015022.

[8] S. Hotchkiss, A. Mazumdar, S. Nadathur, J. Cosmol. Astropart. Phys. 1202 (2012) 008, arXiv:1110.5389 [astro-ph.CO].

[9] S. Choudhury, A. Mazumdar, S. Pal, J. Cosmol. Astropart. Phys. 1307 (2013) 041, arXiv:1305.6398 [hep-ph].

[10] S. Choudhury, A. Mazumdar, Nucl. Phys. B 882 (2014) 386, arXiv:1306.4496 [hep-ph].

[11] S. Choudhury, A. Mazumdar, E. Pukartas, J. High Energy Phys. 1404 (2014) 077, arXiv:1402.1227 [hep-th].

[12] S. Choudhury, J. High Energy Phys. 04 (2014) 105, arXiv:1402.1251 [hep-th].

[13] S. Choudhury, A. Mazumdar, arXiv:1403.5549 [hep-th].

[14] S. Choudhury, A. Mazumdar, arXiv:1404.3398 [hep-th].

[15] P.A.R. Ade, et al., Planck Collaboration, arXiv:1303.5082 [astro-ph.CO].

[16] A. Mazumdar, J. Rocher, Phys. Rep. 497 (2011) 85, arXiv:1001.0993 [hep-ph].

[17] M. Kawasaki, T. Takayama, M. Yamaguchi, J.'i. Yokoyama, Phys. Rev. D 74 (2006) 043525, arXiv:hep-ph/0605271;

J.'i. Yokoyama, Phys. Rev. D 58 (1998) 083510, arXiv:astro-ph/9802357; J.'i. Yokoyama, Phys. Rep. 307 (1998) 133.

[18] R. Saito, J. Yokoyama, Phys. Rev. Lett. 102 (2009) 161101, arXiv:0812.4339 [astro-ph];

R. Saito, J.'i. Yokoyama, Prog. Theor. Phys. 123 (2010) 867, arXiv:0912.5317 [astro-ph.CO];

R. Saito, J.'i. Yokoyama, Prog. Theor. Phys. 126 (2011) 351, Erratum.

[19] K.N. Ananda, C. Clarkson, D. Wands, Phys. Rev. D 75 (2007) 123518, arXiv:gr-qc/0612013.

[20] S. Matarrese, O. Pantano, D. Saez, Phys. Rev. Lett. 72 (1994) 320, arXiv:astro-ph/9310036.

[21] W.H. Press, P. Schechter, Astrophys. J. 187 (1974) 425.

[22] A.M. Green, A.R. Liddle, K.A. Malik, M. Sasaki, Phys. Rev. D 70 (2004) 041502, arXiv:astro-ph/0403181.

[23] C.G. Lacey, S. Cole, Mon. Not. R. Astron. Soc. 271 (1994) 676, arXiv:astro-ph/9402069.

[24] Volker Springel, et al., Nature 435 (2005) 629-636.

[25] J.F. Navarro, C.S. Frenk, S.D.M. White, Astrophys. J. 490 (1997) 493, arXiv:astro-ph/9611107.

[26] P. Dayal, J.S. Dunlop, U. Maio, B. Ciardi, arXiv:1211.1034 [astro-ph.CO]; R. Salvaterra, A. Ferrara, P. Dayal, arXiv:1003.3873 [astro-ph.CO].

[27] R. Allahverdi, K. Enqvist, J. Garcia-Bellido, A. Mazumdar, Phys. Rev. Lett. 97 (2006) 191304, arXiv:hep-ph/0605035;

R. Allahverdi, K. Enqvist, J. Garcia-Bellido, A. Jokinen, A. Mazumdar, J. Cosmol. Astropart. Phys. 0706 (2007) 019, arXiv:hep-ph/0610134;

R. Allahverdi, A. Kusenko, A. Mazumdar, J. Cosmol. Astropart. Phys. 0707 (2007) 018 arXiv:hep-ph/0608138;

A. Mazumdar, S. Nadathur, P. Stephens, Phys. Rev. D 85 (2012) 045001, arXiv:1105.0430 [hep-th].

[28] P.A.R. Ade, et al., Planck Collaboration, arXiv:1303.5076 [astro-ph.CO].

[29] http://lisa.nasa.gov/.

[30] S. Phinney, et al., The Big Bang observer: direct detection of gravitational waves from the birth of the Universe to the present, in: NASA Mission Concept Study, 2004.

[31] N. Seto, S. Kawamura, T. Nakamura, Phys. Rev. Lett. 87 (2001) 221103, arXiv: astro-ph/0108011.

[32] http://www.srl.caltech.edu/~shane/sensitivity/.

[33] H. Kudoh, A. Taruya, T. Hiramatsu, Y. Himemoto, Phys. Rev. D 73 (2006) 064006, arXiv:gr-qc/0511145.

[34] http://www.ligo.caltech.edu/.

[35] P.A.R. Ade, et al., B1CEP2 Collaboration, arXiv:1403.3985 [astro-ph.CO].