Phenomenological implications of modified loop cosmologies: an overview

Bao-Fei Li 1,2∗, Parampreet Singh1†, and Anzhong Wang3‡ 1 Department of Physics and Astronomy, Louisiana State University, Baton Rouge, LA 70803, USA 2 Institute for Theoretical Physics & Cosmology, Zhejiang University of Technology, Hangzhou, 310032, China 3 GCAP-CASPER, Department of Physics, Baylor University, Waco, TX, 76798-7316, USA (Dated: June 25, 2021) In this paper, we first provide a brief review of the effective dynamics of two recently well-studied models of modified loop quantum cosmologies (mLQCs), which arise from different regularizations of the Hamiltonian constraint and show the robustness of a generic resolution of the big bang singularity, replaced by a quantum bounce due to non-perturbative Planck scale effects. As in loop quantum cosmology (LQC), in these modified models the slow-roll inflation happens generically. We consider the cosmological perturbations following the dressed and hybrid approaches and clarify some subtle issues regarding the ambiguity of the extension of the effective potential of the scalar perturbations across the quantum bounce, and the choice of initial conditions. Both of the modified regularizations yield primordial power spectra that are consistent with current observations for the Starobinsky potential within the framework of either the dressed or the hybrid approach. But differences in primordial power spectra are also identified among the mLQCs and LQC. In particular, for mLQC-I, striking differences arise between the dressed and hybrid approaches in the infrared and oscillatory regimes. While the differences between the two modified models can be attributed to differences in the Planck scale physics, the permissible choices of the initial conditions and the differences between the two perturbation approaches have been reported for the first time. All these differences, due to either the different regularizations or the different perturbation approaches , in principle can be observed in terms of non-Gaussianities.

I. INTRODUCTION problem of cosmological fluctuations 1. The second problem of the inflationary paradigm is more severe. It is well known that inflationary space- Despite offering a solution to several fundamental and times are past-incomplete because of the big bang sin- conceptual problems of the standard big bang cosmology, gularity [10], with which it is not clear how to impose including the flatness, horizon, and exotic-relic problems, the initial conditions. This problem gets aggravated for the cosmic inflation in the early universe also provides a low energy inflation in spatially-closed models which are mechanism to produce density perturbations and primor- slightly favored by current observations where the uni- dial gravitational waves (PGWs) [1–3]. The latter arise verse encounters a big crunch singularity and lasts only from quantum fluctuations of spacetimes and produce for a few Planck seconds [11]. not only a temperature anisotropy, but also polarizations Another problematic feature of inflation is that one in the cosmic microwave background (CMB), a smoking often ignores the pre-inflationary dynamics and sets the gun of PGWs. However, the inflationary paradigm is Bunch-Davies (BD) vacuum at a very early time. But, incomplete without the knowledge of key elements from it is not clear how such a vacuum state can be realized quantum gravity. First, it is well-known that the cosmic dynamically in the framework of quantum cosmology [6], inflation is sensitive to the ultraviolet (UV) physics, and considering the fact that a pre-inflationary phase always its successes are tightly contingent on the understanding exists between the Plank and inflation scales, which are of this UV physics [4–6]. In particular, if the inflationary about 103 ∼ 1012 orders of magnitude difference [2, 6]. arXiv:2105.14067v2 [gr-qc] 24 Jun 2021 phase lasts somewhat longer than the minimal period re- While these problems of inflationary paradigm demand a quired to solve the above mentioned problems, the length completion from quantum theory of spacetimes, they also scales we observe today will originate from modes that open an avenue to overcome one of the main obstacles in are smaller than the Planck length during inflation [4]. the development of quantum gravity, which concerns with Then, the treatment of the underlying quantum field the- the lack of experimental evidences. Thus, understanding ory on a classical spacetime becomes questionable, as now inflation in the framework of quantum gravity could offer the quantum geometric effects are expected to be large, valuable guidances to the construction of the underlying and the space and time cannot be treated classically any theory [13–15]. more. This is often referred to as the trans-Planckian

1 It has been conjectured using models in string theory that the ∗E-mail address: [email protected] trans-Planckian problem might never arise [7], which results on †E-mail address: [email protected] severe constraints on various cosmological models (See [8, 9] for ‡E-mail address: Anzhong [email protected] more details). 2

In particular, when applying the techniques of loop µ¯ scheme [19], which is known to be the only possible quantum gravity (LQG) to homogeneous and isotropic choice in LQC, and results in physics that is independent universe, namely loop quantum cosmology (LQC) [16], from underlying fiducial structures used during quantiza- it was shown that, purely due to quantum geometric ef- tion, and meanwhile yields a consistent infrared behav- fects, the big bang singularity is generically resolved and ior for all matter obeying the weak energy condition [70]. replaced by a quantum bounce at which the spacetime Lately, the studies of [66] have been extended to theµ ¯ curvature becomes Planckian [17–20]. The robustness scheme [71]. of the singularity resolution has been shown for a vari- In the two modified LQC models, mLQC-I and mLQC- ety of isotropic and anisotropic spacetimes [21]. Interest- II, since different regularizations of the Lorentzian term ingly, there exists a reliable effective spacetime descrip- were used, the resulting equations become the fourth- tion, which has been used to confirm a generic resolution order and non-singular quantum difference equations, of all strong curvature singularities [22]. Various phe- instead of the second-order difference ones obtained in nomenological implications have been studied using this LQC. In these two models the big bang singularity is also effective spacetime description, whose validity has been generically resolved and replaced by a quantum bounce. verified for isotropic and anisotropic spacetimes [23–27]. In addition, the inflationary phase can naturally take For low energy inflation models with a positive spatial place with a very high probability [72–76]. In addition, curvature, the singularity resolution and a successful on- the dynamics in LQC and mLQC-II is qualitatively sim- set of inflation for classically inadmissible initial condi- ilar in the whole evolution of the universe, while the one tions have been demonstrated [28–30]. in mLQC-I becomes significantly different from LQC (as In the last couple of years, several approaches have well as mLQC-II) in the contracting phase, in which an been proposed, in order to address the impacts of the emergent quasi de Sitter space is present. This implies quantum geometry on the primordial power spectra. that the contracting phase in mLQC-I is purely a quan- These include the approaches of the deformed algebra tum regime without any classical limit 2. [31–33], dressed metric [34–36], and hybrid [37–41] (For An important question now is what are the effects of a recent discussion about similar ideas in anisotropic these models and approaches on the CMB observations. Bianchi I LQC spacetimes see Refs. [42–45] and refer- The answer to this question requires the knowledge of ences therein.). In particular, the last two approaches how the quantum fluctuations propagate on a quantum have been widely studied and found that they are all con- spacetime in LQC and modified loop cosmological mod- sistent with current cosmological observations [46–59]. In els. In particular, in the framework of the dressed metric addition, within the framework of the dressed metric ap- approach the power spectra of the cosmological perturba- proach recently it has been also shown that some anoma- tions for both mLQC-I and mLQC-II models were inves- lies from the CMB data [60–62] can be reconciled purely tigated [78]. In the same framework but restricted only to due to the quantum geometric effects [63, 64]. the mLQC-I model, the power spectra of the cosmological In addition to the standard LQC, in which the perturbations were studied in [79]. More recently, the hy- Lorentzian term of the classical Hamiltonian constraint brid approach was applied to mLQC-I [80–82], for which is first expressed in terms of the Euclidean term in the time-dependent mass of the perturbations was stud- the spatially flat Friedmann-Lemaˆıtre-Robertson-Walker ied in detail [83]. The primordial scalar power spectra (FLRW) universe, and then only the quantization of the obtained in the three models, LQC, mLQC-I and mLQC- Euclidean term is considered, the robustness of the singu- II, were also investigated in the hybrid approach [84], and larity resolution with respect to different quantizations of found that the relative differences in the amplitudes of the classical Hamiltonian constraint in the symmetry re- the power spectra among the three models could be as duced spacetimes have been extensively studied. Two no- large as 2% in the UV regime of the spectra, which is table examples are the so-called modified LQC-I (mLQC- relevant to the current observations. Interestingly, in the I) and modified LQC-II (mLQC-II) models, which were above work, differences in primordial power spectra were first proposed by Yang, Ding and Ma more than a decade found between the hybrid and dressed metric approaches ago [65]. In a recent study, Dapor and Liegener (DL) in the infra-red and oscillatory regimes in mLQC-I. [66, 67] obtained the expectation values of the Hamilto- In this brief review, we shall focus mainly on the states nian operator in LQG using complexifier coherent states that are sharply peaked along the classical trajectories, [68], adapted to the spatially flat FLRW universe. Using so that the description of the “effective” dynamics of the non-graph changing regularization of the Hamilto- the universe becomes available [16], and the questions nian advocated by Thiemann [69], DL obtained an ef- raised recently in [85] are avoided. This includes the stud- fective Hamiltonian constraint, which, to the leading or- ies of the “effective” dynamics of the homogeneous and der, agrees with the mLQC-I model first obtained in [65]. isotropic mLQC-I and mLQC-II models, and their cos- Sometimes, this model has also been referred to as the DL model or Thiemann regularized LQC. Strictly speak- ing, when constructing loops in [66] DL treated the edge length µ as a free parameter, but in [65] it was consid- 2 A similar contracting branch is found in certain anisotropic mod- ered as a specific triad dependent function, the so-called els in the standard regularization of LQC (see for eg. [77]). 3 mological perturbations in the framework of the dressed moment for both of these two modified LQC models, as metric and hybrid approaches. We shall first clarify the that adopted in LQC [36]. In addition, in mLQC-I the issue regarding the ambiguities in the extension of the de Sitter state given by Eq.(3.7) 4 can be imposed in the effective potential for the scalar perturbations across the contracting phase as long as t0 is sufficiently early, so the quantum bounce, and then pay particular attention to universe is well inside the de Sitter phase. On the other the differences among the three models, LQC, mLQC- hand, in mLQC-II and LQC, the BD vacuum can be im- I and mLQC-II, and possible observational signals. It posed in the contracting phase as long as t0 is sufficiently is important to note that initial conditions are another early, so the universe becomes so large that the spacetime subtle and important issue not only in LQC but also in curvature is very small, and particle creation is negligi- mLQCs. This includes two parts: (i) when to impose ble. With these in mind, the power spectra obtained in the initial conditions, and (ii) which kind of initial con- the three models, mLQC-I, mLQC-II and LQC, within ditions one can impose consistently. To clarify this issue, the framework of the dressed metric approach were cal- we discuss it at length by showing the (generalized) co- culated and compared by imposing the initial conditions moving Hubble radius in each model and in each of the in the contracting phase. In particular, the spectra can dressed and hybrid approaches. From this analysis, one be universally divided into three regimes, the infrared, can see clearly what initial conditions can and cannot be intermediate and UV. In the infrared and intermediate imposed at a chosen initial time. regimes, the relative differences in the amplitudes of the The outline of this brief overview is as follows. In Sec. spectra can be as large as 100% between mLQC-I and II we consider the effective dynamics of mLQC-I and mLQC-II (the same is also true between mLQC-I and mLQC-II, and discuss some universal features of their LQC), but in the UV regime such differences get dra- dynamics such as the resolution of big bang singularity. matically reduced, which is no larger than 0.1%. Since In addition, in this section we also show that for states the modes in the UV regime are the relevant ones to the such that the evolution of the homogeneous universe was current observations and also their corresponding power dominated initially at the bounce by the kinetic energy spectra are scale-invariant, so these three models are all ˙2 of the inflaton, that is, φB  V (φB), the post-bounce consistent with observations. evolution between the bounce and the reheating can be always divided universally into three different phases: the In Sec. IV, the cosmological perturbations of mLQC-I bouncing, transition, and slow-roll inflation [cf. Fig. 1]. and mLQC-II are studied within the hybrid approach, During each of these phases the expansion factor a(t) and and the subtleties of the initial conditions are shown in the scalar field φ(t) can be given analytically. In particu- Figs. 8, 9 and 10, where Figs. 8 and 9 are respectively lar, they are given by Eqs.(2.54)-(2.55) and (2.56)-(2.57) for the quadratic and Starobinsky potentials in mLQC-I, during the bouncing phase for mLQC-I and mLQC-II, re- while Fig. 10 is for the Starobinsky potential in mLQC- spectively. In this same section, the probabilities of the II. The case with the quadratic potential in mLQC-II is slow-roll inflation is considered, and shown that it occurs similar to that of mLQC-I, given by Fig. 8. From these generically. This particular consideration is restricted to figures it is clear that imposing the initial conditions now the quadratic potential, but is expected to be also true becomes a more delicate issue, and sensitively depends for other cases. on the potential V (φ) of the inflaton field. First, in the cases described by Figs. 8 and 9, all the modes are oscil- In Sec. III, the cosmological perturbations of mLQC-I lating during the time tp < t < t , so one might intend and mLQC-II are studied. We discuss initial conditions i i to impose the BD vacuum at the bounce. However, for and the subtle issue of the ambiguity in the choice of the t < tp the quantity Ω defined in Eq.(3.9) experiences variables π−2 and π−1 (present in the effective potential), i tot a a a period during which it is very large and negative. As which correspond to the quadratic and linear inverse of a result, particle creation is expected not to be negligi- the momentum conjugate to the scale factor. In addition, ble during this period. Then, imposing the BD vacuum to understand the issue of initial conditions properly, we at the bounce will not account for these effects, and the first introduce the comoving Hubble radius λ2 and then H resulting power spectra shall be quite different from the state clearly how this is resolved in GR [cf. Fig. 2], and case, in which in the deep contracting phase (t  t ) which are the relevant questions in mLQC-I [cf. Fig. 5] B the BD vacuum is imposed for mLQC-II and LQC, and and mLQC-II [cf. Fig. 6]. From these figures it is clear the de Sitter state for mLQC-I. On the other hand, in that the BD vacuum cannot be consistently imposed at the case described by Fig. 10, even if the BD vacuum is the bounce 3, as now some modes are inside the (co- moving) Hubble radius while others not. However, the fourth-order adiabatic vacuum may be imposed at this

4 To be distinguished from the BD vacuum described by Eq.(3.4) we refer to the state described by Eq.(3.7) as the de Sitter state. The difference between them is due to the term i/(kη), which 3 It should be noted that anisotropies rise during the contracting is not negligible in the deep contracting phase of the de Sitter phase and generically dominate the earliest stages of the post- background, as now |kη| could be very small. For more details, bounce of the homogeneous universe [42–45]. So, cautions must see the discussions given in Sec. III.A, especially the paragraph be taken, when imposing initial conditions at the bounce. after Eq.(3.9). 4 chosen at the bounce, it may not be quite different from there exists only one degree of freedom, the scale factor the one imposed in the deep contracting phase, as now a(t), where N(t) is the lapse function and can be freely in the whole contracting phase all the modes are oscil- chosen, given the freedom in reparametrizing t, and qab(t) lating, and particle creation is not expected to be impor- denotes the 3-dimensional (3D) spatial metric of the hy- tant up to the bounce. To compare the results from the persurface t = Constant. In this paper, we shall use three different models, in this section the second-order the indices a, b, c, ... to denote spatial coordinates and adiabatic vacuum conditions are chosen in the contract- i, j, k, ... to denote the internal su(2) indices. Repeated ing phase, which is expected not to be much different indices will represent sum, unless otherwise specified. from the de Sitter state for mLQC-I and the BD vacuum In full GR, the gravitational phase space consists of i a for mLQC-II and LQC, as long as t0  tB in all the the connection Aa and density weighted triad Ei . In the cases described by Figs. 8 - 10. The ambiguities of the present case, the 3D spatial space M has a R3 topology, −2 −1 choice of πa and πa also occur in this approach, but as from which we can introduce a fiducial cell V and re- far as the power spectra are concerned, different choices strict all integrations to this cell, in order to avoid some lead to similar conclusions [82]. So, in this section only artificial divergences and have a well-defined symplectic the so-called prescription A is considered. Then, similar structure. Within this cell, we introduce a fiducial flat 2 conclusions are obtained in this approach regarding the metric ˚qab via the relation qab(t) = a (t) ˚qab, and then an a i differences among the amplitudes of the power spectra associated constant orthogonal triad ˚ei and a cotriad ˚ωa. i a in the three different models. In particular, the relative Then, after symmetry reduction Aa and Ei are given by, differences can be as large as 100% between mLQC-I and i −1/3 i a −2/3p a mLQC-II/LQC, but in the UV regime such differences Aa = c vo ˚ωa,Ei = |p| vo ˚q ˚ei , (2.2) are reduced to about 2%. A remarkable feature between 2/3 2 4 the two different approaches is also identified: in the in- where |p| = vo a , κ = 8πG/c , vo denotes the volume frared and oscillatory regimes, the power spectrum in of the fiducial cell measured by ˚qab, ˚q is the determinant mLQC-I is suppressed as compared with its counterpart of ˚qab, and γ is the Barbero-Immirzi parameter whose in LQC in the hybrid approach. On the other hand, value can be set to γ ≈ 0.2375 using black hole thermo- in the dressed metric approach, the power spectrum in dynamics in LQG [86]. For classical solutions, symmetry mLQC-I is largely amplified in the infrared regime where reduced connection c is related to time derivative of scale its magnitude is as large as of the Planck scale [78, 79]. factor as c = γa˙, where an over dot denotes a derivative The main reason for such differences is that the effective with respect to t for the choice N = 1. a mass in the hybrid approach is strictly positive near the The physical triad and cotriad are given by ei = −1/2 1/3 a i 1/2 −1/3 i bounce, while it is strictly negative in the dressed metric (sgn p) |p| vo ˚ei and ωa = (sgn p) |p| vo ˚ωa, approach for states that are initially dominated by the where (sgn p) arises because in connection dynamics the kinetic energy of the inflaton [78, 79, 82]. phase space contains triads with both orientations. In The review is concluded in Sec. V, in which we sum- the following we choose this orientation to be positive marize the main conclusions and point out some open and volume of the fiducial cell to be vo = 1. The vari- questions for future studies. ables c and p satisfy the communication relation, κγ {c, p} = . (2.3) 3 II. EFFECTIVE QUANTUM DYNAMICS IN MODIFIED LQCS Then, the gravitational part of the Hamiltonian is a sum of the Euclidean and Lorentzian terms, To facilitate our following discussions, let us first H = H(E) − (1 + γ2)H(L) , (2.4) briefly review the standard process of quantization car- grav grav grav ried out in LQC, from which one can see clearly the sim- where, with the choice N = 1, these two terms are given, ilarities and differences among the three models, LQC, respectively, by mLQC-I and mLQC-II. 1 Z EajEbk H(E) = d3x  F i √ , (2.5) grav 2κ ijk ab q A. Quantum Dynamics of LQC 1 Z EajEbk H(L) = d3x Kj Kk √ , (2.6) grav κ [a b] q The key idea of LQC is to use the fundamental vari- ables and quantization techniques of LQG to cosmologi- 6 k where q = det(qab) = a ˚q, Fab is the field strength of the cal spacetimes, by taking advantage of the simplifications i i connection Aa, and Ka is the extrinsic curvature, given, that arise from the symmetries of these spacetimes. In respectively, by the spatially-flat FLRW spacetime, k k k i j 2 k i j Fab ≡ 2∂[aAb] + ij AaAb = c ij ˚ωa ˚ωb , 2 2 2 a b ds = −N (t)dt + qab(t)dx dx eb 2 2 2 a b Ki ≡ K eb = i q˙ − 2D N
= ±a˙ ˚ωi . (2.7) ≡ −N (t)dt + a (t)δabdx dx , (2.1) a ab i 2N ab (a b) a 5

Upon quantization, ambiguities can arise due to dif- where ferent treatments of the Euclidean and Lorentzian terms 3πKG in the Hamiltonian constraint. In particular, LQC takes C+(v) ≡ |v + 2| ||v + 1| − |v + 3|| , the advantage that in the spatially-flat FLRW universe 8 − + o + − the Lorentzian part is proportional to the Euclidean part, C (v) ≡ C (v − 4),C (v) ≡ C (v) + C (v). (2.14) (L) −2 (E) Hgrav = γ Hgrav, (2.8) This is the main result of LQC [19], which shows that: (i) It is a second order quantum difference equation with so that, when coupled to a massless scalar field, the clas- uniform discreteness in volume, rather than a simple sical Hamiltonian can be rewritten as [19, 87] differential equation, a direct consequence of the dis- crete nature of loop quantum geometry. (ii) It provides H ≡ Hgrav + HM the evolution of the quantum cosmological wavefunction 1 Z EajEbk = − d3x  F i √ + H , (2.9) Ψ(v, φ), in which the scalar field serves as a clock. Thus, 2κγ2 ijk ab q M once an initial state Ψ(v, φ0) is given at the initial mo- √ ment φ , the study of the quantum dynamics of LQC where H = p2 / 2 q
, with p being the momentum 0 M φ φ can be carried out. It is found that, instead of a big conjugate of φ. bang singularity, a quantum bounce is generically pro- The elementary operators in the standard LQC are duced, a result confirmed through extensive numerical the triads5 p and elements of the holonomies given by q √ simulations [23–27] and an exactly solvable model [20]. \iµc/¯ 2 2 e of c, whereµ ¯ = ∆lpl/|p| with ∆ ≡ 4 3πγ, and Using this model, one can compute the probability for 2 the quantum bounce which turns out to be unity for an ∆lpl being the minimum non-zero eigenvalue of the area arbitrary superposition of wavefunctions [89]. operator,√ and the Planck length lpl is defined as lpl ≡ For the states sharply peaked around a classical so- ~G. However, it is found that, instead of using the eigenket |pi of the area operator p as the basis, it is more lution, we can obtain “effective” Friedmann and Ray- convenient to use the eigenket |vi of the volume operator chaudhuri (FR) equations, by using the geometric quan- vˆ ≡ sgn(p)|p|3/2
, where tum mechanics in terms of the expectation values of the ˆ ˆ operators (b, v,ˆ φ, pcφ), 8πγ 3/2 |v| vˆ |vi = l3 |vi , e\iµc/¯ 2 |vi = |v + 1i , ˙ 6 K pl b = {b, H} , v˙ = {v, H} , (2.15) ˙ (2.10) φ = {φ, H} , p˙φ = {pφ, H} , (2.16) √  p √ −1 with K ≡ 2 2 3 3 3 . Let Ψ(v, φ) denote the which take the same forms as their classical ones, but wavefunction in the kinematical Hilbert space of the grav- all the quantities now represent their expectation values, itational field coupled with the scalar field φ, we have D E AI ≡ AcI . Then, it was found that the effective Hamil- φˆΨ(v, φ) = φΨ(v, φ), tonian is given by [19], ∂ pˆφΨ(v, φ) = −i~ Ψ(v, φ), 3 1/2 2 1 3/2 2 ∂φ Heff. = − |p| sin (¯µc) + |p| p , (2.17) 8πγ2µ¯2G 2 φ |p\|−3/2Ψ(v, φ) = B(v)Ψ(v, φ), (2.11) which can also be expressed in terms of v and b via the where relations v = |p|3/2 and b = c/p|p|. Then, from Eqs. !3/2 (2.15) and (2.16) one can find that the “effective” FR 6 B(v) ≡ B(v), equations are given by, 8πγl2 pl   2 8πG ρ 33 3 H = ρ 1 − , (2.18) 1/3 1/3 3 ρc B(v) ≡ K|v| |v + 1| − |v − 1| . (2.12) 2  ρ  H˙ = −4πG (ρ + P ) 1 − , (2.19) Then, the equation satisfied by selecting the physical ρc states HˆΨ(v, φ) = 0 can be cast in the form, where 1 h ∂2Ψ(v, φ) = C+(v)Ψ(v + 4, φ) − Co(v)Ψ(v, φ) v˙ a˙ 3 φ B(v) H ≡ = , ρ ≡ , 3v a c 8πλa2γ2G − i + C (v)Ψ(v − 4, φ) , (2.13) H ∂H ρ ≡ M ,P ≡ − M , (2.20) v ∂v

3 2 and v = voa . Since H cannot be negative, from 5 For a modification of LQC based on using gauge-covariant fluxes, Eq.(2.18) we can see that we must have ρ ≤ ρc, and 2 see [88]. at ρ = ρc we have H = 0, that is, a quantum bounce 6 occurs at this moment. When ρ  ρc, the quantum grav- which once substituted back into Eq. (2.6) lead to an ex- (L) (E) ity effects are negligible, whereby the classical relativistic pression of Hgrav different from that of Hgrav in the stan- limit is obtained. dard LQC (see [65] for more details). Correspondingly, For a scalar field φ with its potential V (φ), we have one is able to obtain the following “effective” Hamilto- nian [65, 72], " 2 # pφ HM ≡ Hφ = v + V (φ) . (2.21) 2v2 3v n (γ2 + 1) sin2(2λb)o HI = sin2(λb) − eff. 8πGλ2 4γ2 Then, from Eq.(2.16) we find that + HM . (2.25) p Hence, the Hamilton’s equations take the form, φ˙ = φ , (2.22) v 3v sin(2λb)n o p˙φ = −vV,φ(φ). (2.23) v˙ = (γ2 + 1) cos(2λb) − γ2 , (2.26) 2γλ In the rest of this review, we shall consider only the 3 sin2(λb)n o b˙ = γ2 sin2(λb) − cos2(λb) states that are sharply peaked around a classical solution, 2γλ2 so the above “effective” descriptions are valid, and the questions raised recently in [85] are avoided. −4πGγP, (2.27) where P represents the pressure defined in Eq.(2.20). B. Effective Dynamics of mLQC-I Once the matter Hamiltonian HM is specified, together with the Hamiltonian constraint, As mentioned in the introduction, an important open issue in LQC is its connection with LQG [90]. In particu- H ≈ 0, (2.28) lar, in LQC the spacetime symmetry is first imposed (in the classical level), before the quantization process is car- Eqs.(2.26) and (2.27) uniquely determine the evolution ried out. However, it is well-known that this is different of the universe. Using the non-graph changing regular- from the general process of LQG [15], and as a result, dif- ization of the Hamiltonian [69], the expectation values ferent Hamiltonian constraints could be resulted, hence of the Hamiltonian operator yield the same “effective” resulting in different Planck scale physics. Though the Hamiltonian of Eq.(2.25) to the leading order [66]. question of ambiguities in obtaining the Hamiltonian in It has been shown in detail that the big bang singu- LQG is still open, based on some rigorous proposals by larity is generically replaced by a quantum bounce when I Thiemann [69], various attempts have been carried out, the energy reaches its maximum ρc [65, 66, 72–74], where in order to obtain deeper insights into the question. ρ One of the first attempts to understand this issue was ρI ≡ c , (2.29) made in [65], in which the Euclidean and Lorentzian c 4(1 + γ2) terms given by Eqs.(2.5) and (2.6) are treated differently, by closely following the actual construction of LQG. To and the universe is asymmetric with respect to the be more specific, in the full theory [15], the extrinsic cur- bounce, in contrast to LQC. vature in the Lorentzian term (2.6) can be expressed in To write Eqs.(2.26)-(2.28) in terms of H, ρ and P , it terms of the connection and the volume as was found that one must distinguish the pre- and post- 1 bounce phases [72]. In particular, before the bounce, the Ki = {Ai , {H(E) ,V }}, (2.24) modified FR equations take the form [72], a κγ3 a grav

  p   8πGαρ  ρ  1 − 2γ2 + 1 − ρ/ρI ρ H2 = Λ 1 − 1 + c , (2.30) I     I  3 ρc 2 p I ρc 4γ 1 + 1 − ρ/ρc  p  a¨ 4παG 2 − 3γ2 + 2 1 − ρ/ρI ρ = − (ρ + 3P − 2ρ ) + 4πGαP c Λ    I a 3 2 p I ρc (1 − 5γ ) 1 + 1 − ρ/ρc     2  2 2 p I p I 4παGρ 2γ + 5γ 1 + 1 − ρ/ρc − 4 1 + 1 − ρ/ρc ρ −   , (2.31)   2  I 3 2 p I ρc (1 − 5γ ) 1 + 1 − ρ/ρc 7 where In addition, we also have 1 − 5γ2 γ2ρ α ≡ , ρ ≡ c . (2.32) γ2 + 1 Λ (1 + γ2)(1 − 5γ2) Gα 1 − 1 ' 0.32 > . (2.35) I G γ≈0.2375 8 As ρ  ρc, Eqs.(2.30) and (2.31) reduce, respectively, to

2 8παG Finally, we want to emphasize that the minimal energy H ≈ (ρ + ρΛ) , (2.33) 3 density of this branch, for which the Hubble rate van- a¨ 4παG ishes, turns out to be negative which can be shown as ≈ − (ρ + 3P − 2ρΛ) . (2.34) 3 a 3 ρmin = − 8πGλ2 ≈ −0.023. As a result, the necessary condition to generate a cyclic universe in mLQC-I is the These are exactly the FR equations in GR for an ordinary violation of the weak energy condition which is in con- matter field coupled with a positive cosmological con- trast to the cyclic universes in LQC where the energy stant ρΛ, and a modified Newton’s constant, Gα ≡ αG. density is always non-negative [91]. For γ ≈ 0.2375, we have ρΛ ≈ 0.03ρpl, which is of the same order as the one deduced conventionally in quan- In the post-bounce phase (t > tB), from Eqs.(2.26)- tum field theory for the vacuum energy in our universe. (2.28) we find that [72],

" p !2 # 8πGρ  ρ  γ2 ρ/ρI H2 = 1 − 1 + c , (2.36) I 2 p I 3 ρc γ + 1 1 + 1 − ρ/ρc   I p a¨ 4πG 4πGρ 7γ2 + 8
− 4ρ/ρ + 5γ214 + 8
1 − ρ/ρI ρ = − (ρ + 3P ) +  c c    2  I a 3 3 2 p I ρc (γ + 1) 1 + 1 − ρ/ρc  p  3γ2 + 2 + 2 1 − ρ/ρI ρ +4πGP c , (2.37)    I 2 p I ρc (γ + 1) 1 + 1 − ρ/ρc

from which we obtain a modified Newtonian constant Gα, while in the post- bounce the same limits leads to Eqs.(2.40) and (2.41) 4Gπ(P + ρ)  ρ r ρ  H˙ = 2γ2 + 2 − 3γ2 1 − − 1 . but now with the precise Newtonian constant G. 2 I I (1 + γ ) ρc ρc (2.38) Therefore, regardless of the matter content, the super- inflation (starting at the bounce) will always end at ρ = ρs, where C. Effective Dynamics of mLQC-II ρI  p  ρ = c 4 − 8γ2 − 9γ4 + 3γ2 8 + 16γ2 + 9γ4 , s 8 (2.39) In LQG, the fundamental variables for the gravita- for which we have H˙ (ρs) = 0. tional sector are the su(2) Ashtekar-Barbero connection I i a In the classical limit ρ/ρc  1, Eqs. (2.36) and (2.37) Aa and the conjugate triad Ei . When the Gauss and spa- reduce, respectively, to tial diffeomorphism constraints are fixed, in the homoge- neous and isotropic universe the only relevant constraint 8πG H2 ≈ ρ, (2.40) is the Hamiltonian constraint, from which we obtain the 3 FR equations, as shown in the previous section. The a¨ 4πG ≈ − (ρ + 3P ) , (2.41) Hamiltonian in mLQC-II arises from the substitution a 3 whereby the standard relativistic cosmology is recovered. i It is remarkable to note that in the pre-bounce phase i Aa I Ka = , (2.42) the limit ρ/ρc  1 leads to Eqs.(2.33) and (2.34) with γ 8 in the Lorentzian term (2.6). Then, the following effec- D. Universal Properties of mLQC-I/II Models tive Hamiltonian is resulted [65],      II 3v 2 λb 2 2 λb To study further the evolution of the universe, it is Heff. = − sin 1 + γ sin 2πGλ2γ2 2 2 necessary to specify the matter content HM . For a sin- +H , (2.43) gle scalar field with its potential V (φ), the correspond- M ing Hamiltonian takes the form (2.21). As a result, the from which we find that the corresponding Hamilton’s Hamilton’s equations of the matter sector are given by equations are given by, Eqs.(2.22) and (2.23). 3v sin(λb)n o The effective quantum dynamics of LQC, mLQC-I, and v˙ = 1 + γ2 − γ2 cos (λb) , (2.44) γλ mLQC-II were studied in detail recently in [73] for a single scalar field with various potentials, including the 6 sin2 λb
n λb o ˙ 2 2 2 chaotic inflation, Starobinsky inflation, fractional mon- b = − 2 1 + γ sin − 4πGγP γλ 2 odromy inflation, non-minimal Higgs inflation, and infla- = −4πGγ(ρ + P ). (2.45) tion with an exponential potential, by using dynamical It can be shown that the corresponding (modified) FR system analysis. It was found that, while several fea- equations now read [73], tures of LQC were shared by the mLQC-I and mLQC-II models, others belong to particular models. In partic-    2  2 8πGρ 2 ρ (γ + 1)ρ ular, in the pre-bounce phase, the qualitative dynamics H = 1 + γ 1 − 2 , (2.46) 3 ρc ∆ ρc of LQC and mLQC-II are quite similar, but are strik- ingly different from that of mLQC-I. In all the three a¨ 4πG 4πGP ρ  2
 = − (ρ + 3P ) − 3 γ + 1 − 2∆ models, the non-perturbative quantum gravitational ef- a 3 ∆ρ c fects always result in a non-singular post-bounce phase, 2  2  4πGρ 2 2
4γ ρ in which a short period of super-inflation always exists − 2 7γ − 1 + 5γ − 3 (∆ − 1) − , 3∆ ρc ρc right after the bounce, and is succeeded by the conven- (2.47) tional inflation. The latter is an attractor in the phase space for all the three models. p 2 where ∆ ≡ 1 + 1 + γ ρ/ρc. From these equations we 6 can see that now the quantum bounce occurs when ρ = Moreover, similar to LQC [51, 52] , for the initially II kinetic energy dominated conditions, ρc , at which we have H = 0 anda ¨ > 0, where II 2 ρc = 4(γ + 1)ρc, (2.48) 1 ˙2 φB  V (φB), (2.52) which is different from the critical density ρc in LQC as 2 I well as the one ρc in mLQC-I. Therefore, the big bang singularity is also resolved in this model, and replaced by it was found that the evolution of the universe before the II a quantum bounce at ρ = ρc , similar to LQC and mLQC- reheating is universal. In particular, in the post-bounce I, despite the fact that the bounce in each of these models phase (between the quantum bounce and the reheating), occurs at a different energy density. However, in contrast the evolution can be uniquely divided into three phases: to mLQC-I, the evolution of the universe is symmetric bouncing, transition and slow-roll inflation, as shown in with respect to the bounce, which is quite similar to the Fig. 1 for the Starobinsky potential, standard LQC model. In addition, similar to the other two cases, now the 3m2  √ 2 bounce is accompanied by a phase of super-inflation, i.e. V (φ) = 1 − e− 16πG/3φ . (2.53) 32πG H˙ > 0, which ends at H˙ (ρs) = 0, but now ρs is given by, ρ   c 2 p 2 4 4 2 For other potentials, similar results can be obtained, as ρs = 2 3(γ + 1) 1 + 2γ + 9γ + 9γ + 10γ − 3 . 8γ long as at the bounce the evolution of the universe is (2.49) II dominated by the kinetic energy of the inflaton w(φB) ' For γ = 0.2375, we find ρs = 0.5132ρc . II 1 [74, 102]. When ρ  ρc , the modified FR equations (2.46)-(2.47) reduce to, In each of these three phases, the evolutions of a(t) and φ(t) can be well approximated by analytical solutions. In 8πG H2 ≈ ρ, (2.50) 3 a¨ 4πG ≈ − (ρ + 3P ) , (2.51) a 3 6 In LQC, this universality was first found for the quadratic and which are identical to those given in GR. Therefore, in Starobinsky potentials [51, 52] (see also [92]), and later was shown that they are also true for other potentials, including this model, the classical limit is obtained in both pre- the power-law potentials [93, 94], α-attractor potentials [95, 96], and post-bounce when the energy density ρ is much lower Monodromy potentials [97], warm inflation [98], Tachyonic infla- II than the critical one ρc . tion [99] and even in Brans-Dicke LQC [100, 101]. 9

a that the best fitting is provided by [74],

A = C = 1.2,B = 6,D = 2. (2.55) 108 For the mLQC-II model, during the bouncing phase a(t) and φ(t) are given by 105   2  1/6 II Aγ 2 a(t) = 1 + 24πGρc 1 + t , 100 1 + Bt r   2  II Cγ t mpl arcsinh 24πGρc 1 + t 104 106 1+Dt 1 100 φ(t) = φ ± , ϕ B r  Cγ2  1.0 12πG 1 + 1+Dt

0.5 (2.56)

but now with, 4 6 t 1 100 10 10

-0.5 A = 2.5,B = 7,C = D = 2. (2.57)

-1.0 In the transition and slow-roll inflationary phases, the functions a(t) and φ(t) were given explicitly in [74]. -1.5 For the initially potential energy dominated cases, w ϕ 1 φ˙2  V (φ ), (2.58) 1.0 2 B B

0.5 it was found that such universalities are lost. In partic- ular, for the Starobinsky potential, the potential energy dominated bounce cannot give rise to any period of infla- 4 6 t 1 100 10 10 tion for both mLQC-I and mLQC-II models, quite similar to what happens in LQC [47, 48]. -0.5

-1.0 E. Probabilities of the Slow-roll Inflation in mLQC-I/II Models FIG. 1: The evolution of the scale factor a(t), the scalar field φ(t), and the equation wφ of state of the scalar field (from To consider the probability of the slow-roll inflation in top to bottom) in the post-bounce phase are depicted and the modified LQC models, let us start with the phase compared among the three modes, LQC (red solid curves), space S of the modified Friedmann and Klein-Gordon mLQC-I (blue dotted curves) and mLQC-II (green dot-dashed equations, which now is four-dimensional (4D), and con- curves), with the Starobinsky potential. In the last panel, wφ sists of the four variables, (v, b) and (φ, p ), from the ˙2 ˙2 φ is defined via wφ ≡ P (φ)/ρ(φ) = [φ − 2V (φ)]/[φ + 2V (φ)]. gravitational and matter sectors, respectively. Using the The initial condition for the simulation is chosen at the bounce with φ = −1.6 m , φ˙ > 0 [74]. canonical commutation relations, the symplectic form on B pl B the 4D phase space is given by [103–110], dv ∧ db particular, during the bouncing phase, they are given by Ω = dpφ ∧ dφ + . (2.59) 4πGγ   Aγ2  1/6 However, after taking the effective Hamiltonian con- a(t) = 1 + 24πGρI 1 + t2 , c 1 + Bt straint into account, r   Cγ2  ( 2 ) m arcsinh 24πGρI 1 + t pφ pl c 1+Dt C = 16πG H (v, b) + + vV (φ) ' 0, (2.60) grav 2v φ(t) = φB ± r ,  2  12πG 1 + Cγ 1+Dt where “ ' ” means that the equality holds only on Γ,¯ (2.54) we can see that the 4D phase space S reduces to a three- dimensional (3D) hypersurface Γ.¯ for mLQC-I model, where the parameters A, B, C and On the other hand, the phase space S is isomorphic to D are fixed through numerical simulations. It was found a 2-dimensional (2D) gauge-fixed surface Γˆ of Γ,¯ which is 10 intersected by each dynamical trajectory once and only where I(E) is the interval on the φB-axis, which cor- once [105]. From the FR equations, it can be shown that responds to the physically distinct initial conditions in for both mLQC-I and mLQC-II the variable b satisfies which the event E happens, and D is the total measure the equation [72–74], Z φmax n h io1/2 D ≡ −2 HˆA (b ) + V (φ) dφ. (2.69) b˙ = −4πGγ(ρ + P ). (2.61) grav B φmin For any given matter field that satisfies the weak energy Once the probability is properly defined, we can calcu- condition [111], we have ρ + P > 0, so the function b is late it in different models of the modified LQCs. In LQC monotonically decreasing. Then, a natural parameteri- [105], the calculations were carried out for the quadratic zation of this 2D surface is b = constant, say, b0. Hence, potential. In order to compare the results obtained in dif- using the Hamiltonian constraint (2.60) we find ferent models, let us consider the same potential. Then, for the mLQC-I model it was found that [74] n h io1/2 pA = v −2 HˆA + V (φ) , (2.62) φ grav r 1 p2γ2 + 1 sin (λbI ) = , sin (2λbI ) = , B 2γ2 + 2 B γ2 + 1 where A = I, II, and 1 1 I I 2 I I 2 pφ = v (2ρc − 2V ) , dω = (2ρc − 2V ) dφ, (2.70) ˆA −1 A Hgrav ≡ v Hgrav(v, b0). (2.63) so that the probability for the desired slow-roll not to On the other hand, from Eqs.(2.25) and (2.43) we find happen is, that HˆA = HˆA (b ) = constant on Γ.ˆ Thus, we find grav grav 0 0.917 R dωI I −5.158 −5 A 2 P (not realized) . I ' 1.12 × 10 , (2.71) p φmax A φ v V,φ R I dp = dv − dφ. (2.64) −φI dω φ Γˆ max v pφ I 5 where φmax = 3.49 × 10 mpl. Inserting this expression into Eq.(2.59), we find that the In mLQC-II, following a similar analysis, it can be pulled-back symplectic structure Ωˆ reads shown that the probability for the desired slow-roll not to happen is [74], n h io1/2 Ωˆ A = −2 HˆA (b ) + V (φ) dφ ∧ dv, (2.65) grav 0 II −6 Γˆ P (not realized) . 2.62 × 10 . (2.72) from which we find that the Liouville measure dµˆL on Γˆ Note that in LQC the probability for the desired slow roll is given by inflation not to occur is [105],

1/2 LQC −6 A n h ˆA io P (not realized) . 2.74 × 10 , (2.73) dµˆL = −2 Hgrav(b0) + V (φ) dφdv. (2.66) which is smaller than that for mLQC-I and slightly larger A Note that dµˆL does not depend on v, so that the in- than the one for mLQC-II. However, it is clear that the tegral with respect to dv is infinite. However, this diver- desired slow-roll inflation is very likely to occur in all the gency shall be cancelled when calculating the probabil- models, including the two modified LQC ones. ity, as it will appear in both denominator and numerator. Therefore, the measure for the space of physically distinct solutions can be finally taken as III. PRIMORDIAL POWER SPECTRA OF MODIFIED LQCS IN DRESSED METRIC n h io1/2 APPROACH A ˆA dω = −2 Hgrav(b0) + V (φ) dφ, (2.67) As mentioned above, in the literature there exist sev- so that the 2D phase space Γˆ is further reduced to an eral approaches to investigate the inhomogeneities of the interval φ ∈ (φmin, φmax). It should be noted that such a universe. Such approaches can be generalized to the defined measure depends explicitly on b0, and its choice in modified LQC models, including mLQC-I and mLQC-II. principle is arbitrary. However, in loop cosmology there In this section we shall focus ourselves on cosmological exists a preferred choice, which is its value at the quan- perturbations in the framework of mLQCs following the tum bounce b0 = b (tB) [105]. With such a choice, the dressed metric approach [34–36], while in the next sec- probability of the occurrence of an event E becomes tion we will be following the hybrid approach [37–41]. We shall restrict ourselves to the effective dynamics, as Z 1/2 1 n h ˆA io we did with the homogeneous background in the last sec- P (E) = −2 Hgrav(bB) + V (φ) dφ, D I(E) tion. Such investigations in general include two parts: (a) (2.68) the initial conditions; and (b) the dynamical evolutions 11

00 2 of perturbations. In the framework of effective dynamics, vk + k vk = 0, which has the solution, the latter is a second-order ordinary differential equation in the momentum space, so in principle once the initial αk −ikη βk ikη vk ' √ e + √ e , (3.2) conditions are given, it uniquely determines the cosmo- 2k 2k logical (scalar and tensor) perturbations. However, the initial conditions are a subtle issue, which where αk and βk are two integration constants, and must is true not only in LQC but also in mLQCs. This is satisfy the normalized condition, mainly because that in general there does not exist a ∗ 0 ∗0 preferred initial time and state for a quantum field in vkvk − vk vk = −i. (3.3) an arbitrarily curved space-time [12]. If the universe is If we further require the vacuum to be the minimum en- sufficiently smooth and its evolution is sufficiently slow, ergy state, a unique solution exists, which is given by so that the characteristic scale of perturbations is much α = 1, β = 0, that is, larger than the wavelengths of all the relevant modes, a k k well justified initial state can be defined: the BD vacuum. 1 −ikη This is precisely the initial state commonly adopted in lim vk ' √ e , (3.4) kaH 2k GR at the beginning of the slow-roll inflation, in which all the relevant perturbation modes are well inside the which is often referred to as the BD vacuum [112]. comoving Hubble radius [112] [cf. Fig.2]. However, in LQC/mLQCs, especially near the bounce, ln LH the evolution of the background is far from “slow”, and its geometry is also far from the de Sitter. In particu- lar, for the perturbations during the bouncing phase, the wavelengths could be larger, equal, or smaller than the −1 corresponding characteristic scale, as it can be seen, for ln k example, from Fig. 5. Thus, it is in general impossi- ble to assume that the universe is in the BD vacuum at the bounce [36, 49–52]. Therefore, in the following let us first elaborate in more details about the subtle issues regarding the initial conditions. t it * t end ln a

FIG. 2: The evolution of the comoving Hubble radius ln(LH ) A. Initial Conditions for Cosmological vs ln(a) in GR, where t denotes the moment of the onset Perturbations i of the slow-roll inflation, t∗ the horizon crossing time of a mode with the wavenumber k, and tend the moment that the The initial conditions for cosmological perturbations slow-roll inflation ends. in fact consists of two parts: when and which? How- ever, both questions are related to each other. In LQC Consider the de Sitter space as the background, we 00 00 2 literature, for cosmological perturbations, two different have a(η) = 1/(−ηH), and z /z = a /a = 2/LH , where moments have been chosen so far in the dressed and hy- LH ≡ 1/(aH) = −η is the corresponding comoving Hub- brid approaches: (i) the remote past in the contracting ble radius. Then, Eq.(3.1) reads, phase [78] and (ii) the bounce [36, 49, 50]. To see which   conditions we need to impose at a given moment, let us 00 1 2 vk + 2 − 2 vk = 0, (3.5) first recall how to impose the initial conditions in GR, λ LH in which the scalar perturbations are governed by the equation, where λ (≡ 1/k) denotes the comoving wavelength. The above equation has the following exact solutions,  00  00 2 z     vk + k − vk = 0, (3.1) αk −ikη i βk ikη i z vk = √ e 1 − + √ e 1 + . (3.6) 2k kη 2k kη where k denotes the comoving wave number, and z ≡ It is clear that on scales much smaller than the comoving ˙ aδφ/H, with δφ being the scalar field perturbations, φ = Hubble radius (λ  L ), v is oscillating with frequency ¯ H k φ(t)+δφ(t, x). A prime denotes a derivative with respect k and constant amplitude, given by Eq.(3.2). Then, set- to the conformal time η, while an over dot denotes a ting (αk, βk) = (1, 0) we find that Eq. (3.6) reduces derivative with respect to the cosmic time t, where dη = to dt/a(t). The standard choice of the initial sate is the   Minkowski vacuum of an incoming observer in the far 1 −ikη i vk = √ e 1 − . (3.7) past, k  aH [cf. Fig. 2]. In this limit, Eq.(3.1) becomes 2k kη 12

Note that if the initial time ti is chosen sufficiently small, conditions can be imposed consistently. In particular, i.e., ti  tend or |kη|  1, all the modes are inside the if at this moment some modes are inside the comoving comoving Hubble radius LH [cf. Fig. 2], and the BD Hubble radius and others are not, it is clear that in this vacuum (3.4) becomes a natural choice. case imposing the BD vacuum at the bounce will lead However, on the scales much larger than the comoving to inconsistencies. In addition, there also exist the cases 2 Hubble radius (λ  LH ), the k term is negligible com- in which particle creation in the contracting phase is not pared to the squeezing term, z00/z, and as a result, the negligible, then it is unclear how a BD vacuum can be fluctuations will stop oscillating and the amplitude of vk imposed at the bounce, after the universe is contracting starts to increase, yielding for such a long time before the bounce. Thus, in these cases other initial conditions need to be considered, such vk ' z(η). (3.8) as the fourth-order adiabatic vacuum [36, 49–52]. With the above in mind, in the following we turn to As shown in Fig. 2, if the initial time ti is chosen to be sufficiently early, all the currently observed modes consider power spectra of the cosmological perturbations. ∗ kph ∈ (0.1, 1000) × k0 will be well inside the comoving Hubble radius at t = ti, so the mode function vk can be well approximately given by Eq.(3.4), which is the well- B. Power Spectra of Cosmological Perturbations ∗ known zeroth order adiabatic state, where k0 = 0.002 −1 Mpc and kph(t) ≡ k/a(t) [113]. Since the evolutions of the effective dynamics of the In modified LQC models, the mode function vk satisfies homogeneous backgrounds for mLQC-I and mLQC-II are the following modified equation, different, in this subsection let us first consider the case of mLQC-I and then mLQC-II. To compare the results with 00 2 vk + Ωtot(η, k)vk = 0, (3.9) those obtained in LQC, at the end of this subsection, we 2 also discuss the LQC case. where Ωtot(η, k) depends on the homogeneous back- ground and the inflation potential V (φ), so it is model- dependent. Therefore, the choice of the initial conditions 1. mLQC-I will depend on not only the modified LQC models to be considered but also the moment at which the initial conditions are imposed. For mLQC-I, the power spectrum of the cosmologi- cal scalar perturbations was first studied in [79] for the One of the main reasons to choose the remote past in 2 the contracting phase as the initial time for perturba- quadratic φ potential, and re-examined in [78]. In the tions is that at such time either the background is well terminology used in [79], it was found that the corre- described by the de Sitter space (mLQC-I) or the ex- sponding mode function vk(≡ qk/a) satisfies Eq.(3.9) pansion factor a(t) becomes so large that the curvature with of the background is negligible (mLQC-II and LQC), so a00 24πGφ˙2 imposing the BD vacuum for mLQC-II and LQC and the Ω2 = k2 − + A , r ≡ , tot a − ρ de Sitter state given by Eq.(3.7) for mLQC-I at this mo- 2  √  ment is well justified. It should be noted that the reason A− ≡ a V (φ)r − 2V,φ(φ) r + V,φφ(φ) , (3.10) to refer to the state described by Eq.(3.7) as the de Sitter ˙2 state is the following: In the slow-roll inflation, the homo- where r = 24πGφ /ρ and V (φ) denotes the scalar field geneous and isotropic universe is almost de Sitter, as the potential. Hubble parameter H ≡ a/a˙ is almost a constant, so we It should be noted that, when generalizing the classical have a(η) ' 1/(−Hη). For ti  tend we have a(η)  1, expression of the function A− defined in Eq.(3.10) to its and |ηk| ' |Hη|  1, so the choice (αk βk) = (1, 0) corresponding quantum mechanics operator, there exists will lead Eq.(3.7) directly to Eq.(3.4) at the onset of the ambiguities. In fact, classically A− only coincides with 2 slow-roll inflation [cf. Fig. 2]. However, in the deep ΩQ in the expanding phase. The latter is a function contracting phase of the same de Sitter space, now the of the phase space variables which is explicitly given by universe is very large, that is, a(η)  1, so we must [36, 59, 78, 79], have |Hη|  1 and |ηk| ' |k/H |a−1(η), which can be Λ p2 p4 very small in sufficiently early times of the contracting 2 φ φ pφ 2 ΩQ = 3κ 4 − 18 6 2 − 12a V,φ + a V,φφ, (3.11) phase, so the terms i/(kη) in Eq.(3.6) now cannot be ne- a a πa πa glected. To distinguish this case from the one described by Eq.(3.4), In this review we refer the state described where πa is the moment conjugate to a, and given by one by Eq.(3.7) with the term i/(kη) not being negligible as of Hamilton’s dynamical equations, the de Sitter state, while the state described by Eq.(3.4) 3a2 is still called the BD vacuum state, or simply the BD π = − H, (3.12) a 4πG vacuum. On the other hand, if the initial time is chosen to be with the choice of the lapse function N = 1. Therefore, at the bounce, cautions must be taken on what initial πa < 0 (πa > 0) corresponds to H > 0 (H < 0). At the 13

quantum bounce we have H(tB) = 0, so that πa(tB) = 0. 2 -10 Then, ΩQ defined by Eq.(3.11) diverges at the bounce. 6.×10 2 � Hence, from the Friedmann equation, H = (8πG/3)ρ, Ω+ 5.×10 -10 we find that � Ω- s 4.×10 -10 � 1 2πG 1 2πG Ω = , = ± , (3.13) 2 4 4 -10 πa 3a ρ πa 3a ρ 3.×10

where “-” corresponds to H > 0, and “+” to H < 0. 2.×10 -10 Then, a direct generalization leads to [78], 1.×10 -10 -1.0 -0.5 0.0 0.5 1.0 ( 2 A−,H ≥ 0, t ΩQ = (3.14) A+,H ≤ 0, 2 2 FIG. 3: The potential terms Ω+ and Ω− are compared with their smooth extension Ω2 across the bounce in mLQC-I for where 1 2 2 the quadratic potential V (φ) = 2 m φ [78]. 2  √  A± ≡ a V (φ)r ± 2V,φ(φ) r + V,φφ(φ) . (3.15)

It should be noted that in [79] only the function A− was chosen over the whole process of the evolution of the 10 universe. The same choice was also adopted in [59, 64]. |Ω�| � In addition, A defined by Eq.(3.13) is not√ continuous 0.01 |Ω��� | across the bounce, as the coefficient 2V,φ(φ) r in gen- |�� /�| eral does not vanish at the bounce. In [51, 52, 55] A− 2 10-5 appearing in Eq.(3.10) was replaced by U(φ)(≡ Ω+) over the whole process of the evolution of the homogeneous universe, where 10-8

2 2  2  Ω± ≡ a f V (φ) ± 2fV,φ(φ) + V,φφ(φ) , (3.16) -8 -6 -4 -2 0 and f ≡ (24πG/ρ)1/2 φ˙. Another choice was introduced in [78], which was mo- 0.001 |Ω�| � tivated from the following considerations. The functions |Ω��� | 2 Ω defined above are not continuous across the bounce, |�� /�| ± 10-5 quite similar to A±. However, if we introduce the quan- tity Ω2 as,

10-7 2 2  2  Ω = a f V (φ) + 2Θ(b)fV,φ(φ) + V,φφ(φ) , (3.17)

to replace A− in Eq.(3.10), it could be continuous across 10-9 the bounce by properly choosing Θ(b). In particular, the 7 variable b(t) satisfies Eq.(2.61) [72–74] , from which we 50 100 500 1000 5000 104 can see that b(t) is always a monotonically decreasing 2 2 function for any matter that satisfies the weak energy FIG. 4: The potential terms Ω and Ωeff are compared with condition [111]. Moreover, one can construct a step-like the curvature term a00/a in mLQC-I near the bounce and function of b with the bounce as its symmetry axis [72– the preinflationary regime for the quadratic potential V (φ) = 1 2 2 74]. Therefore, if we define Θ(b) as 2 m φ [78].

Θ(b) = 1 − 2 1 + γ2
sin2 (λb) , (3.18) In addition to the above choices, motivated by the hy- it behaves precisely as a step function, so that Ω2 −2 brid approach [37–41], the following replacements for πa smoothly connects Ω2 across the bounce, as shown in −1 ± and πa in Eq.(3.11) were also introduced in [78], Fig. 3. 1 64π2G2λ2γ2 2 → 4 , (3.19) πa 9a D 7 1 8πGλγΘ(b) It is interesting to note that Eq.(2.61) holds not only in mLQC-I, → − , (3.20) but also in LQC and mLQC-II. 2 1/2 πa 3a D 14 where tB), the universe is asymptotically de Sitter, and only very near the bounce (about several Planck seconds), the 2
2 2 2 D ≡ 1 + γ sin (2λb) − 4γ sin (λb) . (3.21) Hubble parameter H which is negative in the pre-bounce regime suddenly increases to zero at the bounce. Then, 2 2 Such obtained ΩQ was referred to as Ωeff in [78], and the universe enters a very short super-acceleration phase 2 2 ˙ in Fig. 4, we show the differences among Ω ,Ωeff and H > 0 (super-inflation) right after the bounce, which 00 I ˙ the quantity a /a, from which one can see that the term lasts until ρ ' ρc/2, where H(t)ρ' 1 ρI = 0. Afterwards, 00 2 c a /a dominates the other two terms over the whole range ˙2 4
for a kinetic energy dominated bounce φB  V (φB), it t/tpl ∈ −8, 10 . takes about 104 ∼ 106 Planck seconds before entering the To study the effects of the curvature term, let us first slow-roll inflation [72–74]. Introducing the quantity, introduce the quantity, a 1 λ2 ≡ = − , (3.25) 00 1/2 H 00 a  a − aA− meff kI = ≈ 1.60, (3.22) B a t=t where meff is the effective mass of the modes, from B Eq.(3.10) we find that 2 2 which is much larger than other two terms Ω and Ωeff ,  2 2 2 −10 2 > 0, λH > λ , where Ω (tB) = 1.75 × 10 and Ωeff (tB) = 0.006. 1 1  2 2 Ω2 = − = < 0, 0 < λ2 < λ2, (3.26) Therefore, the differences between Ω and Ωeff near the tot 2 2 H λ λH  2 bounce are highly diluted by the background. On the > 0, λH < 0, 2 2 other hand, in the post-bounce phase, Ω and Ωeff coin- 4 where λ (≡ k−1) denotes the comoving wavelength of the cide after t/tpl ' 10 , while near the onset of the inflation their amplitudes first become almost equal to that of the mode k, as mentioned above. Note that such a defined 2 curvature term, and then quickly exceeds it during the quantity λH becomes negative when the effective mass 2 slow-roll inflation, as we can see from Fig. 4. is positive. In Fig. 5, we plot λH schematically for the From Fig. 4 we can also see that the difference between quadratic and the Starobinsky potentials with the mass −6 −6 2 2 of the inflaton set to 1.21×10 mpl and 2.44×10 mpl Ω and Ωeff lies mainly in the region near the bounce. However, as the curvature term a00/a overwhelmingly respectively. The initial conditions for the background dominates in this region, it is usually expected that the evolution are set as follows: for the quadratic potential, impact of the different choices of Ω2 on the power spec- the inflaton starts with a positive velocity on the right trum might not be very large [36, 59, 79]. However, in wing of the potential and for the Starobinsky potential [78], it was found that the relative difference in the mag- the inflaton is released from the left wing of the potential nitude of the power spectrum in the IR and oscillating with a positive velocity. For both potentials, the initial regimes could be as large as 10%, where the relative dif- conditions are set at the bounce which is dominated by ference is defined as, the kinetic energy of the inflaton field. The same mass parameters and similar initial conditions are also used

P1 − P2 in the following figures where the comoving Hubble ra- E ≡ 2 . (3.23) P + P dius is plotted schematically. In Fig. 5, the moments tH 1 2 00 00 and ti are defined, respectively, by a (tH ) = a (ti) = 0, 2 2 so t represents the beginning of the inflationary phase, However, the power spectra obtained from Ω and Ω± i are substantially different even in the UV regime due to and during the slow-roll inflation (Region III), we have 2 λ2 ≈ L2 /2 ' 1/(2a2H2), which is exponentially de- the (tiny) difference between Ω± at the bounce, see Fig. H H 14 given in [78]. In fact, the difference is so large that the creasing, and all the modes observed today were inside power spectrum calculated from Ω2 is essentially already the comoving Hubble radius at t = ti. Between the ± 2 2 ruled out by current observations. times tH and ti, λH is negative, and Ωtot is strictly pos- With the clarification of the ambiguities caused by the itive. Therefore, during this period the mode functions quantum mechanical generalization of the function A− are oscillating, while during the epoch between tB and −2 −2 defined in Eq.(3.10), now let us turn to the issue of the tH , some modes (k > kB ) are inside the comoving −2 −2 initial conditions, for which we consider only two repre- Hubble radius, and others (k < kB ) are outside it −1 sentative potentials, the quadratic and Starobinsky, given right after the bounce, where kB ≡ λB (tB). In the con- explicitly by tracting phase, when t  tB, the universe is quasi-de 2 2 2 Sitter and λH ' 1/(2a H ) increases exponentially to-  1 2 2 m φ , quadratic, ward the bounce t → tB, as now a(t) is decreasing ex-  2 √ V = 2  2 ponentially. However, several Planck seconds before the 3m 1 − e− 16πG/3 φ , Starobinsky.  32πG bounce, the universe enters a non-de Sitter state, during 2 (3.24) which λH starts to decrease until the bounce, at which In mLQC-I, the evolution of the effective (quantum) a characteristic Planck scale kB (≡ 1/λH ) can be well homogeneous universe is asymmetric with respect to the defined. Therefore, for t  tB, all the modes are out- bounce [66, 72–74]. In particular, before the bounce (t < side the comoving Hubble radius. Then, following our 15 previous arguments, if the initial moment is chosen at λ2 Η t0  tB, the de Sitter state seems not to be viable. How- ever, when t  t we have a(η) ' 1/(−η|H |), where 0 B Λ −2 2 −1 HΛ = −[λ(1 + γ )] and kB 2 I II III Ω2 ' k2 − , (3.27) tot η2 k −2 t for which Eq.(3.9) has the exact solutions given by t B t H i t Eq.(3.6). Therefore, at sufficient early times, choosing αk = 1, βk = 0 leads us to the de Sitter state (3.7). From the above analysis it is clear that this is possible precisely because of the isometry of the de Sitter space, which is sufficient to single out a preferred state, the de Sitter state [79]. 2 With the exact solution (3.7) as the initial condi- FIG. 5: Schematic plot of λH defined by Eq.(3.25) vs t for tions imposed at the moment t0 ( tB) in the con- mLQC-I in the dressed metric approach for the quadratic and 00 00 tracting phase, it was found that the power spectrum the Starobinsky potentials, where a (tH ) = 0 and a (ti) = 0 of the cosmological scalar perturbations can be divided with ti being the starting time of the inflationary phase. Dur- 2 2 into three different regimes: (i) the ultraviolet (UV) ing the slow-roll inflation, we have λH ≈ LH/2. In the con- tracting phase t < tB , the universe is initially de Sitter and (k > kmLQC-I); (ii) intermediate (ki < k < kmLQC-I); and 2 2 p we still have λH ≈ LH/2, but now it increases exponentially (iii) infrared (k < ki), where kmLQC-I ≡ aB RB/6 and p toward bounce t → tB , as the universe in this phase is expo- ki = ai Ri/6, and RB and Ri are the curvatures given nentially contracting. However, several Planck seconds before at the bounce and the beginning of the slow-roll inflation, the bounce, the universe enters a non-de Sitter state, during 2 respectively [cf. Fig. 5]. During the infrared regime, which λH starts to decrease until the bounce. The qualitative the power spectrum increases as k increases, while in behavior of the comoving Hubble radius is the same for the the intermediate regime it is oscillating very fast and quadratic and the Starobinsky potentials. Different potentials the averaged amplitude of the power spectrum is de- will change the the values of tH and ti correspondingly. creasing as k increases. In the UV regime, the spec- trum is almost scale-invariant, which is consistent with ticular, it is symmetric with respect to the bounce and in the current observations. There exists a narrow band, the initially kinetic energy dominated case at the bounce 0.1 × k∗ < k < k , in which the quantum gravita- 0 mLQC-I the solutions can be well approximated by Eq.(2.56) in tional effects could be detectable by current or forthcom- the bouncing phase [73, 74], similar to that of LQC [16]. ing cosmological observations [79]. Within the dressed When considering the cosmological perturbations, sim- metric approach, one of the most distinctive features of ilar ambiguities in the choices of π−2 and π−1 in Eq.(3.11) the power spectrum in mLQC-I is that its magnitude a a exist. In particular, for the choice of Eq.(3.17) now the in the IR regime is of the Planck scale [78, 79]. This function Θ(b) is replaced by is because those infrared modes are originally outside the Hubble horizon in the contracting phase and thus λb Θ(b) = cos , (3.28) their magnitudes are frozen as they propagate across the 2 bounce and then into the inflationary phase. Consider- ing that the contracting phase is a quasi de Sitter phase which behaves also like a step function across the bounce with a Plank-scale Hubble rate, the magnitude of the IR and picks up the right sign in both contracting and ex- modes is thus also Planckian [78]. panding phases, so it smoothly connects Ω± defined by It should be noted that if the initial conditions are Eq.(3.16). 2 imposed at the bounce, from Fig. 5 we can see clearly On the other hand, Ωeff is obtained from Eq.(3.11) by that some modes are inside the comoving Hubble radius, the replacements, and some are not. In addition, in the neighborhood of 1 4π2γ2λ2 the bounce, the background is far from de Sitter. So, it → , (3.29) 2 4 2 is impossible to impose either the BD vacuum or the de πa 9a sin (λb/2) D Sitter state at the bounce. In this case, one of the choices 1 −2πγλ cos (λb/2) → , (3.30) 2 1/2 of the initial conditions is the fourth-order adiabatic vac- πa 3a sin (λb/2) D uum, similar to that in LQC [36, 49–52]. but now with λb 2. mLQC-II D ≡ 1 + γ2 sin2 . (3.31) 2

2 2 2 In mLQC-II, the evolution of the effective homoge- Such obtained Ω ,Ω± and Ωeff are quite similar to neous universe is different from that of mLQC-I. In par- those given by Figs. 3 and 4 in mLQC-I. In particular, 16 at the bounce, we have k ∈ (5 × 10−6, 50) were studied and found that the rel- ative difference in the magnitude of the power spectra 2 −10 2 Ω (tB) = 1.59 × 10 , Ωeff (tB) = 0.265, is around 30% in the IR regime and less than 10% in the intermediate regime. In the UV regime, the relative a00 1/2 kII = ≈ 6.84, (3.32) difference can be as small as 0.1% or even less. B a t=tB that is, the curvature term a00/a still dominates the evo- 3. LQC lution near the bounce. To consider the effects of the ambiguities in the choice 2 −2 −1 8 λ of πa and πa in Eq.(3.11) , power spectra of the cos- H mological perturbations were also studied in the frame- −2 2 k work of LQC in [78]. In this case, Ω is obtained from

−2 Eq.(3.17) with kB III’ III Θ(b) = cos(λb), (3.33) t B while Ω2 is obtained from Eq.(3.11) by the replace- p p t eff t i t H H t i t ments,

1 16π2G2γ2λ2 → , (3.34) 2 4 2 πa 9a sin (λb) 1 −4πγλ cos (λb) → . (3.35) 2 2 FIG. 6: Schematic plot of λH defined by Eq.(3.25) vs t for πa 3a sin (λb) mLQC-II in the dressed metric approach for the quadratic 00 p 00 2 and the Starobinsky potentials, where a (tH ) = a (tH ) = 0 As shown explicitly, the term Ω+ is always negligible 00 p 00 00 and a (ti ) = a (ti) = 0, and t = ti denotes the starting comparing with the curvature term a /a in the expres- p 2 time of the inflationary phase, while t = ti is the end time sion of Ωtot defined in Eq.(3.10) by replacing A− with of the deflationary phase in the contracting branch. During Ω2 . So, from Eq.(3.25) we find that 2 2 2 + the slow-roll inflation, we have λH ≈ LH/2. In particular, λH is decreasing (increasing) exponentially in Region III (Region 1 a λ2 = ' , (3.36) III’). The corresponding effective mass near the bounce is al- H a00/a − Ω2 a00 ways negative. Similar behavior also happens in LQC in the + dressed metric approach [51]. The bounce is dominated by the 2 00 p during the bouncing phase t ∈ (tB, ti), and λH ' a/a kinetic energy of the scalar field, which leads to tH ≈ −tH . p was shown schematically by Fig. 18 in [52], which is quite However, in general we find that ti 6= −ti due to the effects of the potential energy of the scalar field far from the bouncing similar to Fig. 6 given above for mLQC-II. point. The comoving Hubble radius has the same qualitative As a result, the initial states of the linear perturba- behavior for the quadratic and the Starobinsky potentials, tions can be either imposed in the contracting phase at a P p p while the values of ti (ti ) and tH (tH ) depend on the type of moment t0  ti as the BD vacuum, or at the bounce as the potentials and the initial conditions. the fourth-order adiabatic vacuum [36]. However, it was shown analytically that such two conditions lead to the To see how to impose the initial conditions, let us in- same results [52]. 2 To compare the results obtained from the three dif- troduce the quantity λH defined by Eq.(3.25) but now 2 2 ferent models, in [78] the fourth-order adiabatic vacuum A− will be replaced either by Ω or Ωeff . The details 2 was chosen even in the contracting phase for LQC. Here, here are not important, and λH is schematically plotted in Fig. 6, from which we can see that if the initial con- we cite some of the results in Fig. 7. In particular, it was ditions are chosen to be imposed at the bounce, the BD found that the relative difference in the amplitudes of the power spectra of the scalar perturbations due to the vacuum (as well as the de Sitter state) is still not avail- 2 2 able, and the fourth-order adiabatic vacuum is one of the choice of Ω or Ωeff is about 10% in the infrared regime, possible choices, similar to the LQC case. However, if the about 100% in the intermediate regime, and about 0.1% initial conditions are imposed in the contracting phase at in the UV regime. Since only modes in the UV regime p can be observed currently, clearly this difference is out of t0  ti , the universe becomes very large a(t)  1 and can be practically considered as flat, then the BD vacuum can be chosen.

Certainly, one can choose different initial conditions. 8 In particular, the fourth-order adiabatic vacuum was cho- In the framework of LQC, such effects were also studied in [36, 51, 52, 55, 78]. In particular, in [59, 64, 79] the function A− sen even in the contracting phase in [78]. With such a defined in Eq.(3.15) was chosen over the whole process of the 2 2 choice, the power spectra from Ω and Ωeff in the region evolution of the universe. 17

PS IV. PRIMORDIAL POWER SPECTRA OF 1 △△△△ MODIFIED LQCS IN HYBRID APPROACH △△ △△ △△ △ △△△△△ △△ △ △ As in the previous section, in this section we also △△ mLQC-I 0.001 △△△ consider the three different models, LQC, mLQC-I, and △△ mLQC-II, but now in the hybrid approach, and pay par- △ △ △△ ticular attention to the differences of the power spectra △△ -6 △ 10 △ △ among these models. Since the scalar perturbations are △△ △ the most relevant ones in the current CMB observations, △ △△△ in the following we shall mainly focus on them, and such △ △ △△△ △△△△△△△△△△△△△△△△ studies can be easily extended to the tensor perturba- 10-9 △ △ k tions. -4 10 0.001 0.010 0.100 1 10 PS 10-4 □ A. mLQC-I

□ □ LQC 10-5 ○ mLQC-II Power spectra of the cosmological scalar and tensor □ ○ □ □□ perturbations for the effective Hamilton in mLQC-I were 10-6 □○○ ○ □○ recently studied in the hybrid approach [82–84]. In par- ○ □ □○ ○ ○ ○ ○ ○ ticular, the mode function vk of the scalar perturbations -7 ○ □ ○□ 10 ○□ □ □ ○ ○□ ○□ satisfies the differential equation [84], ○○□ ○□□ ○ ○○○○○○○○□ □○□ □□□ □□□□□□□□ ○ ○ ○ □□ 00 2 -8 ○ ○□ □
10 □ vk + k + s vk = 0, (4.1) ○ □ ○○ ○ ○□○□□○□○□○□○□○□ □○□ ○ ○ where k -5 -4 10 10 0.001 0.010 0.100 1 10 4πGp2 p2 ! s = φ 19 − 24πGγ2 φ 4/3 2 FIG. 7: The figure shows the results of the scalar power spec- 3v πa tra from three models presented in [78] when the potential   2 2/3 16πGγpφ 16πG term is given by Ωeff. The inflationary potential is chosen to +v V,φφ + V,φ − V be the quadratic potential and the e-foldings of the inflation- πa 3 ary phases in all three models are chosen to be 72.8. The first 4πG = − a2 (ρ − 3P ) + U, (4.2) panel shows the scalar power spectrum in mLQC-I which is 3 characterized by its unique infrared regime. In the second panel, we compare the scalar power spectra from LQC and which is the effective mass of the scalar mode, with mLQC-II.  2 V,φ U ≡ a V,φφ − 12 πa 64a6V (φ)  3V (φ) 1  + ρ − 2 . (4.3) the sensitivities of the current and forhcoming observa- πG 4πG πa tions [114]. Note that in [84], instead of πa, the symbol Ω was used. However, comparing the power spectra obtained from In addition, the cosmological tensor perturbations are the three different models, even with the same choice of also given by Eqs.(4.1) and (4.2) but with the vanishing −2 −1 πa and πa , it was found that the relative difference potential U = 0. Then, we immediately realize that in among LQC, mLQC-I and mQLC-II are significant only the hybrid approach quantum mechanically there are also −2 −1 in the IR and oscillating regimes, while in the UV regime, ambiguities in the replacements πa and πa , as men- all three models give quite similar results. In particular, tioned in the last section. So far, two possibilities were with the same regularization of πa the difference can be as considered [82, 83]. One is given by the replacements, large as 100% throughout the IR and oscillating regimes, while in the UV regime it is about 0.1%. 1 1 1 ΛI 2 → 2 , → 2 , (4.4) πa ΩI πa ΩI For the tensor perturbations, the potential term ΩQ vanishes identically, so no ambiguities related to the in Eq.(4.2), where choice of πa exist. But, due to different models, the dif- v2γ2 n γ2 + 1 o ferences of the power spectra of the tensor perturbations Ω2 ≡ − sin2 (λb) − sin2 (2λb) , can be still very large in the IR and oscillating regimes I λ2 4γ2 among the three models, although they are very small in sin(2λb) Λ ≡ v . (4.5) the UV regime, see, for example, Fig. 12 given in [78]. I 2λ 18

p This is the case referred to as prescription A in [83]. bounce. In addition, when t  ti , the background is well The other possibility is obtained by the replacement of described by the de Sitter space, so the de Sitter state Eqs.(3.34) and (3.35), which was referred to as Prescrip- can be imposed in the deep contracting phase. However, tion B [83], and showed that the two prescriptions lead imposing the BD vacuum at the bounce will clearly lead to almost the same results. So, in the rest of this section to different power spectra at the end of the slow-roll in- we restrict ourselves only to prescription A. flation from that obtained by imposing the de Sitter state Then, for the case in which the evolution of the homo- in the deep contracting phase. This is because, when the p− geneous universe was dominated by kinetic energy at the background is contracting to about the moments t ' ti , bounce, the effective mass becomes so large and negative that the mode function vk will be modified significantly, in com- ˙2 p φB  2V (φB), (4.6) parison with that given at t0 ( ti ), or in other words, particle creation now becomes not negligible during the it was shown that the effective mass is always positive at contracting phase. Then, other initial conditions at the the bounce [83]. In fact, near the bounce we have [58], bounce may need to be considered.

4πG 2 s = − a (ρ − 3P ) + U(η) λ 2 3 H 8πG ' a2ρ > 0. (4.7) 3 −2 k Note that in writing the above expression, we have used the fact that during the bouncing phase we have III’ III w ≡ P/ρ ' 1, and |U(η)|  1. On the other hand, in φ tB the pre-bounce phase, when t  t the background is a B tp t contracting de Sitter spacetime, so we have [83], i i t

4πG s = − a2 (ρ − 3P ) + U(η) 'U(η) ' 5a2V , 3 ,φφ HΛ(t−tB ) a ' aBe , (4.8) 2 p FIG. 8: Schematic plot of λH defined by Eq.(4.10) in mLQC- where HΛ ≡ − 8παGρΛ/3. Thus, the effective mass I for the quadratic potential in the hybrid approach, where p remains positive in the pre-bounce phase, as long as s(ti) = s(ti ) = 0, and t = ti is the starting time of the V,φφ(t  tB) > 0. This is the case for both quadratic inflationary phase. During the slow-roll inflation, we have 2 2 and Starobinsky potentials. In fact, from (3.24), we find λH ≈ LH/2 (Region III). In the contracting phase, the back- that ground is asymptotically de Sitter. The evolution of the uni- verse is asymmetric with respect to the bounce. In particu- ( 2 2 p p− m , quadratic, lar, λH is strictly negative for ti < t < ti, while for t ' ti √ √ 2 V,φφ =  4 πG/3 φ −8 πG/3 φ the “generalized” comoving Hubble radius λH becomes pos- m2 2 − e e , Starobinsky. 2 itive and large. However, as t decreases, λH becomes nega- (4.9) tive again. Although the values of ti and tH depend on the For the case that satisfies the condition (4.6) initially at initial conditions for the background evolution, for example, 4 the bounce, we find that φ(t) becomes very negative at when φB = 1.27 mpl at the bounce, ti ≈ 7.55 × 10 tpl and t  t for the Starobinsky potential, so V (t  t ) is tH ≈ −21.85 tpl, the qualitative behavior of the comoving B ,φφ B Hubble radius is robust with respect to the choice of the ini- positive even in this case. Then, the quantity defined by tial conditions as long as the bounce is dominated by the 1 kinetic energy of the scalar field. λ2 ≡ − , (4.10) H s has similar behavior in the post-bounce phases for the case in which the evolution of the homogeneous universe was dominated by kinetic energy at the bounce, but has B. mLQC-II different behaviors in the pre-bounce phases, depending specifically on the potentials considered. In Figs. 8 and 9 we show the comoving Hubble ra- Similar to LQC, the homogeneous universe of mLQC- dius for the quadratic and Starobinsky potentials, respec- II is symmetric with respect to the bounce, and is well p described by the analytical solutions given by Eqs.(2.56) tively. From these figures it is clear that for ti < t < ti, 2 and (2.57) for the states that are dominated by kinetic λH is strictly negative, which implies the effective mass s is positive in this regime. Hence, all the modes assume energy at the bounce. the oscillatory behavior as the modes inside the Hub- In this model, the cosmological perturbations are also ble horizon, and we may impose the BD vacuum at the given by Eqs.(4.1)-(4.3) but now with the replacement 19

λ 2 chosen in the distant past, not only for the Starobinsky H potential but also for other potentials. Due to the oscil- lating behavior of the mode function over the whole con- −2 k tracting phase, imposing the BD vacuum at the bounce is expected not to lead to significant difference in the power III’ III spectra from that in which the same condition is imposed in the deep contracting phase. p p t t t t i t 2 H i B λ H

k −2

2 III FIG. 9: Schematic plot of λH defined by Eq.(4.10) for the t Starobinsky potential and mLQC-I in the hybrid approach, B p p where s(ti) = s(t ) = s(t ) = 0, and t = ti is the starting i H t i t time of the inflationary phase. During the slow-roll inflation, 2 2 we have λH ≈ LH/2 (Region III). In the contracting phase, the background is asymptotically de Sitter. The evolution of the universe is asymmetric with respect to the bounce. In 2 p particular, λH is strictly negative for ti < t < ti, while for p− t ' ti it becomes positive and large. However, as t decreases, 2 2 FIG. 10: Schematic plot of λ2 defined by Eq.(4.10) for the λH becomes negative again. The qualitative behavior of λH H does not change with the choice of the initial conditions as Starobinsky potential and mLQC-II in the hybrid approach, long as the inflaton initially starts from the left wing of the where s(ti) = 0, and t = ti is the starting time of the inflation- 2 2 potential at the kinetic-energy-dominated bounce. However, ary phase. During the slow-roll inflation, we have λH ≈ LH/2 the exact values of ti and tH depend on the initial conditions. (Region III) since the contribution from the potential is in p p general less than a00/a. Again the qualitative behavior of λ2 For example, when φB = −1.32 mpl, tH = −7.88 tpl, ti = H 5 −4.11 tpl and ti = 4.90 × 10 tpl. remains the same as long as the inflaton starts from the left wing of the potential with a positive velocity and the bounce is initially dominated by the kinetic energy of the inflaton [84], field.

1 1 1 Λ → , → II , (4.11) π2 Ω2 π Ω2 a II a II C. LQC where The evolution of the homogeneous universe of standard 4v2 λb  λb Ω2 ≡ sin2 1 + γ2 sin2 , LQC model is also symmetric with respect to the bounce, II λ2 2 2 and is well described by the analytical solutions given sin (λb) in [51, 52] for the states that are dominated by kinetic Λ ≡ v . (4.12) II λ energy at the bounce. In this model, the cosmological perturbations are also In this case, it can be shown that the effective mass given by Eqs.(4.1)-(4.3) but now with the replacement defined by Eq.(4.2) is always positive in the neighborhood [84], of the bounce, but far from the bounce, the properties 2 1 1 1 ΛLQC of λH depend on the potential in the pre-bounce phase, 2 → 2 , → 2 , (4.13) similar to mLQC-I. πa ΩLQC πa ΩLQC 2 In Fig. 10, we plot λH for the Starobinsky potential, while for the quadratic one, it is quite similar to the cor- where responding one in mLQC-I, given by Fig. 8. From Fig. v sin (λb) v sin (2λb) 10 we can see that λ2 now is negative not only near the Ω ≡ , Λ ≡ . (4.14) H LQC λ LQC 2λ bounce but also in the whole contracting phase, so that all the modes are oscillating for t < ti. Then, one can In this case, it can be shown that the effective mass choose the BD vacuum at the bounce. It is remarkable defined by Eq.(4.2) is always positive for the states that that for the quadratic potential, this is impossible [cf. are dominated by kinetic energy at the bounce [55, 58], 2 Fig. 8]. and the quantity λH defined by Eq.(4.10) is negative near Moreover, as t → −∞, the expansion factor becomes the bounce. Again, similar to the mLQC-II case, the very large, and the corresponding curvature is quite low, modes are oscillating near the bounce. However, in the 2 so to a good approximation, the BD vacuum can also be contracting phase the behavior of λH sensitively depends 20

2 PS on the inflation potentials. For the Starobinsky one, λH -6 behaves similar to that described by Fig. 10, so the BD 10 □○○ □○□ □○ ○○○ ○ vacuum can be imposed either in the deep contracting □□□○□□○○□ ○○□ -7 □○△ □○ □ ○ □ ○○□ ○ 10 △ □○ □ □ ○□ □ ○ △ △ □○○○□○□□○□ ○ phase or at the bounce, and such resulted power spectra △□○ △△△△ △△ △□ □ ○○○ △ △ △△ △ △ △ □□□ □ △ △ △△ △ △ ○○□○○○○ are expected not to be significantly different from one 10-8 △ △ △△△ △ ○△ □□ ○ □○ △ △ △□ □ ○ □ △△ □ ○ another. But for the quadratic potential the situation is △ △ △△□□□ ○○○○ △ △△○△□△△□□△△□□△○△□○△□○△□○△□○△□○□△○□△○□△○□△○□△○□△○□△ quite different, and a preferred choice is to impose the 10-9 □○□○□○□ □○ △ ○ BD vacuum in the deep contracting phase (t0  tB). -10 □○ □○ □ LQC 10 □○ □○ △ mLQC-I □○ △△△ △ -11 □○ △ △ 10 □○ △ ○ mLQC-II □○ △ D. Primordial Power Spectra □○ △ ○ △ △ k -4 △ 10 0.001 0.010 0.100 1 10 As it can be seen that one of the preferred moments to impose the initial conditions for the cosmological per- FIG. 11: The primordial power spectra of the cosmolog- turbations in all these three models is a moment in the ical scalar perturbations in the hybrid approach with the contracting phase t0 < tB. In this phase, we can impose Starobinsky potential, respectively, for LQC, mLQC-I, and mLQC-II. The mass of the inflaton field is set to 2.44 × the BD vacuum state as long as the moment is sufficiently −6 10 mpl. The background evolution is chosen so that the earlier, t0  tB. Certainly, other initial conditions can also be chosen. In particular, in [84] the second-order adi- pivot mode is k∗ = 5.15 in all three models. The initial states are the second-order adiabatic states imposed in the contract- abatic vacuum conditions were selected, but it was found ing phase at the moment t with t  t [84]. that the same results can also be obtained even when 0 0 B the BD vacuum or the fourth-order adiabatic vacuum is imposed initially. The nth-order adiabatic vacuum conditions can be ob- In addition, in the following only the Starobinsky po- tained as follows: Let us first consider the solution, tential given in Eq.(3.24) will be considered, as it repre- sents one of the most favorable models by current obser- 1 R η vations [113]. Let us turn to consider the power spectrum −i Wk(¯η)dη¯ νk = √ e . (4.15) of the scalar perturbations in each of the three models. 2Wk Similar results can be also obtained for the tensor per- Then, inserting it into (4.1), one can find an iterative turbations. In particular, it was found that the scalar equation for Wk. In particular, it can be shown that the power spectra in these three models can be still divided (0) into three distinctive regimes: the infrared, oscillatory zeroth-order solution is given by Wk = k, while the second and fourth order adiabatic solutions are given by, and UV, as shown in Fig. 11. In the infrared and oscillatory regimes, the relative dif- p (2) p (4) f(s, k) ference between LQC and mLQC-I can be as large as W = k2 + s, W = . (4.16) k k 4|k2 + s| 100%, while this difference reduces to less than 1% in the UV regime. This is mainly because LQC and mLQC-I 00 Here f(s, k) = 5s02+16k4(k2+3s)+16s2(3k2+s)−4s (s+ have the same classical limit in the post-bounce phase, k2). It should be noted that, in order to compare directly and as shown in Figs. 5 and 8, the effective masses in with observations, it is found convenient to calculate the both approaches tend to be the same during the infla- power spectrum of the comoving curvature perturbation tionary phase. Rk, which is related to the Mukhanov-Sasaki variable via However, it is interesting to note that in the infrared ˙ the relation Rk = νk/z, with z = aφ/H. Then, its power and oscillatory regimes, the power spectrum in mLQC-I spectrum reads is suppressed in comparison with that of LQC, which has been found only in the hybrid approach. As a matter of P k3 |ν |2 fact, in the dressed metric approach, the power spectrum P = νk = k . (4.17) Rk z2 2π2 z2 in mLQC-I is largely amplified in the infrared regime, and its magnitude is of the Planck scale as depicted in Fig. In addition, the power spectrum is normally evaluated at 7 [78, 79]. The main reason might root in the distinctive the end of inflation, at which all the relevant modes are behavior of the effective masses in the two approaches, well outside the Hubble horizon [cf. Fig. 2]. as shown explicitly in Figs. 5 and 8. It should be also noted that the above formula is only On the other hand, the difference of the power spectra (2) (4) applicable to the case where Wk and/or Wk remains between LQC and mLQC-II is smaller than that between real at the initial time. This is equivalent to require k2 + LQC and mLQC-I. In particular, in the infrared regime, (2) (4) s ≥ 0 for Wk and f(s, k) ≥ 0 for Wk . Since the it is about 50%. The large relative difference (more than effective mass s in general depends on t, it is clear that 100%) of the power spectra between mLQC-I and mLQC- the validity of (4.16) depends not only on the initial states II also happens in the infrared and oscillatory regimes, but also on the initial times. while in the UV regime it reduces to about 2%. 21

−5 −6 −6 To summarize, in the hybrid approach the maximum occur are . 1.12×10 , . 2.62×10 , and . 2.74×10 relative differences of the power spectra among these for mLQC-I, mLQC-II and LQC, respectively. three different models always happen in the infrared and When dealing with perturbations, another ambiguity oscillatory regimes, while in the UV regime, the differ- rises in the replacement of the momentum conjugate πa ences reduce to no larger than 2%, and all the three of the expansion factor a in the effective potential of the models predict a scale-invariant power spectrum, and is scalar perturbations. This ambiguity occurs not only in consistent with the current CMB observations. However, the dressed metric approach [cf. Eq.(3.11)] but also in in the hybrid approach, the power spectrum in mLQC-I the hybrid approach [cf. Eq.(4.3)], as it is closely related is suppressed in the infrared and oscillatory regimes. The to the quantization strategy used in LQG/LQC, because latter is in a striking contrast to the results obtained from now only the holonomies (complex exponentials) of πa the dressed metric approach, which might be closely re- are defined as operators. Several choices have been pro- lated to the fact that the effective masses in these two posed in the literature [36, 59, 78, 79, 82–84, 125]. In approaches are significantly different, especially near the Secs. III and IV, we have shown that for some choices bounce and in the prebounce stage. the effects on the power spectra are non-trivial, while for others the effects are negligible. However, even with the same choice, the relative differences in the amplitudes V. CONCLUSIONS AND OUTLOOK of the power spectra among the three different models can be as large as 100% in the infrared and intermedi- In the past two decades, LQC has been studied exten- ate regimes of the spectra, while in the UV regime the sively, and several remarkable features have been found relative differences are no larger than 2%, and the cor- [16], including the generic resolution of the big bang sin- responding power spectra are scale-invariant. Since only gularity (replaced by a quantum bounce) in the Planck- the modes in the UV regime are relevant to the current ian scale, the slow-roll inflation as an attractor in the observations, the power spectra obtained in all the three post-bounce evolution of the universe, and the scale- models are consistent with current observations [113]. invariant power spectra of the cosmological perturba- However, the interactions between the infrared and UV tions, which are consistent with the current CMB obser- modes appearing in non-Gaussianities might provide an vations. Even more interestingly, it was shown recently excellent window to observe such effects. This was ini- that some anomalies from the CMB data [60–62] can be tially done in LQC [57–59], and lately generalized to reconciled purely due to the quantum geometric effects bouncing cosmologies [64]. It should be noted that in in the framework of LQC [63, 64]. [64], the expansion factor a(t) near the bounce was as- sumed to take the form, Despite of all these achievements, LQC is still plagued with some ambiguities in the quantization procedure. In 2
n a(t) = aB 1 + bt , particular, its connection with LQG is still not estab- lished [90], and the quantization procedure used in LQC where b and n are two free parameters. For example, owing to symmetry reduction before quantization can re- for LQC we have n = 1/6 and b = RB/2, where RB sult in different Hamiltonian constraints than the one of is the Ricci scalar at the bounce [51, 52]. But, it is LQG. clear that near the bounce a(t) takes forms different from Motivated by the above considerations, recently vari- the above expression for mLQC-I/II, as one can see from ous modified LQC models have been proposed, see, for Eqs.(2.54)-(2.57). Thus, it would be very interesting to example, [115–124] and references therein. In this brief study such effects in mLQC-I/II, and look for some ob- review, we have restricted ourselves only to mLQC-I and servational signals. mLQC-II [65–67], as they are the ones that have been Moreover, initial conditions are another subtle and im- extensively studied in the literature not only the dynam- portant issue not only in LQC but also in mLQCs. As ics of the homogeneous universe [72–76, 80, 81], but also a matter of fact, the initial conditions consist of two the cosmological perturbations [78, 79, 82–84]. parts: the initial time, and the initial conditions. Dif- In these two modified LQC models, it was found that ferent choices of the initial times lead to different choices the resolution of the big bang singularity is also generic of the initial conditions, or vice versa. To clarify these is- [72–76]. This is closely related to the fact that the area sues, in Sections III and IV we have discussed it at length operator in LQG has a minimal but nonzero eigenvalue by showing the (generalized) comoving Hubble radius in [15, 16], quite similar to the eigenvalue of the ground state each model as well as in each of the two approaches, of the energy operator of a simple harmonic oscillator in dressed metric and hybrid. From these analyses, we have quantum mechanics. This deep connection also shows shown clearly which initial conditions can and cannot be that the resolution of the big bang singularity is purely imposed at a given initial time. due to the quantum geometric effects. In addition, simi- In addition, when the universe changes from contrac- lar to LQC, the slow-roll inflation also occurs generically tion to expansion at the bounce, particle and entropy in both mLQC-I and mLQC-II [74]. In particular, when creations are expected to be very large, and it is crucial the inflaton has a quadratic potential, V (φ) = m2φ2/2, to keep such creations under control, so that the basic the probabilities for the desired slow-roll inflation not to assumptions of the models are valid, including the one 22 that the cosmological perturbations are small and can the quantum geometric effects, and the instability prob- be treated as test fields propagating on the quantum ho- lem is automatically out of the question. So, it would mogeneous background, as assumed in both the dressed be very interesting to study bouncing cosmologies in the metric and hybrid approaches. framework of LQC/mLQCs. The first step in this direc- Yet, different initial conditions also affect the ampli- tion has already been taken [122], and more detailed and tudes and shapes of the primordial power spectra, and it extensive analyses are still needed. would be very interesting to investigate the consistency of such obtained spectra with current observations, in par- ticular the possible explanations to the anomalies found in the CMB data [60–62], and the naturalness of such Acknowledgements initial conditions. On the other hand, bouncing cosmologies, as an al- We are grateful to Robert Brandenberger and David ternative to the cosmic inflation paradigm, have been Wands for valuable comments on the manuscript and extensively studied in the literature, see, for example, helpful discussions. We thank Javier Olmedo and Tao [126] and references therein. However, in such classi- Zhu for various discussions related to works presented in cal bounces, exotic matter fields are required in order to this manuscript. B.-F.L and P.S. are supported by NSF keep the bounce open. This in turn raises the question grant PHY-1454832. B.-F.L is also partially supported of instabilities of the models. On the other hand, quan- by the National Natural Science Foundation of China tum bounces found in LQC/mLQCs are purely due to (NNSFC) with the Grants No. 12005186.

[1] A. Guth, Inflationary universe: A possible solution to 12, 038 (2002); R. Easther, W. H. Kinney, and H. the horizon and flatness problems, Phys. Rev. D23, 347 Peiris, Observing trans-Planckian signatures in the cos- (1981); A.A. Starobinsky, A new type of isotropic cos- mic microwave background, J. Cosmol. Astropart. Phys. mological models without singularity, Phys. Lett. B91, 05 (2005) 009; A. E. Joras and G. Marozzi, Trans- 99 (1980); K. Sato, First-order phase transition of a Planckian physics from a nonlinear dispersion relation, vacuum and the expansion of the universe, Mon. Not. Phys. Rev. D79, 023514 (2009); A. Ashoorioon, D. R. Astron. Soc. 195, 467 (1981). Chialva and U. Danielsson, Effects of Nonlinear Dis- [2] S. Dodelson, Modern Cosmology (Academic Press, New persion Relations on Non-Gaussianities, J. Cosmol. As- York, 2003); V. Mukhanov, Physics Foundations of tropart. Phys. 06, 034 (2011); M.G. Jackson and K. Cosmology (Cambridge University Press, Cambridge, Schalm, Model Independent Signatures of New Physics 2005)); S. Weinberg, Cosmology (Oxford University in the Inflationary Power Spectrum, Phys. Rev. Lett. Press, Oxford, 2008). 108, 111301 (2012); R.H. Brandenberger and J. Mar- [3] H. Kodama, M. Sasaki, Cosmological Perturbation The- tin, Trans-Planckian issues for inflationary cosmology, ory, Progr. Theoret. Phys. Suppl. 78 (1984) 1; V.F. Class. Quantum. Grav. 30, 113001 (2013). Mukhanov, H.A. Feldman, R.H. Brandenberger, Theory [5] C. Kiefer, and M. Kramer, Quantum Gravitational of Cosmological Perturbations, Phys. Rep. 215 (1992) Contributions to the CMB Anisotropy Spectrum, Phys. 203; K.A. Malik, Cosmological Perturbations in an In- Rev. Lett. 108, 021301 (2012); L.M. Krauss and F. flationary Universe, arXiv:astro-ph/0101563; K.A. Ma- Wilczek, Using Cosmology to Establish the Quantiza- lik and D. Wands, Cosmological perturbations, Phys. tion of Gravity, Phys. Rev. D89, 047501 (2014); R.P. Rept. 475 (2009) 1; L. Senatore, Lectures on Inflation, Woodard, Perturbative Quantum Gravity Comes of Age, in New Frontiers in Fields and Strings, Proceedings of arXiv:1407.4748. the 2015 Theoretical Advanced Study Institute in El- [6] L. McAllister and E. Silverstein, String Cosmology: ementary Particle Physics, Boulder, Colorado, 1 - 26 A Review, Gen. Relativ. Grav. 40, 565 (2007); C.P. June 2015, Edited by J. Polchinski, P. Vieira and O. Burgess, M. Cicoli, and F. Quevedo, String Inflation DeWolfe (World Scientific, Singapore, 2017) pp. 447- After Planck 2013, JCAP 1311 (2013) 003; D.F. Cher- 543. noff, S.-H. H. Tye, Inflation, String Theory and Cos- [4] R.H. Brandenberger, Inflationary Cosmology: Progress mology, arXiv:1412.0579; D. Baumann and L. McAl- and Problems, arXiv:hep-th/9910410; J. Martin and lister, Inflation and String Theory (Cambridge Mono- R.H. Brandenberger, Trans-Planckian problem of in- graphs on Mathematical Physics, Cambridge University flationary cosmology, Phys. Rev. D63, 123501 (2001); Press, 2015); M. Cicoli, Recent developments in string J.C. Niemeyer and R. Parentani, Trans-Planckian dis- model-building and cosmology, arXiv:1604.00904; E. Sil- persion relation and scale invariance of inflationary per- verstein, TASI lectures on cosmological observables and turbations, Phys. Rev. D64, 101301(R) (2001); Corley- string theory, arXiv:1606.03640. Jacobson dispersion relation and trans-Planckian in- [7] A. Bedroya and C. Vafa, Trans-Planckian Censor- flation, Phys. Rev. D65, 103514 (2002); Dependence ship and the Swampland, JHEP 2009, 123 (2020) of the spectra of fluctuations in inflationary cosmol- [arXiv:1909.11063]. ogy on trans-Planckian physics,Phys. Rev. D68, 063513 [8] A. Bedroya, R. Brandenberger, M. Loverde and (2003); L. Bergstorm and U.H. Danielsson, Can MAP C. Vafa, Trans-Planckian Censorship and Inflation- and Planck map Planck physics?, J. High Energy Phys. ary Cosmology, Phys. Rev. D101, 103502 (2020) 23

[arXiv:1909.11106]. ture of the big bang, Phys. Rev. Lett. 96, 141301 (2006). [9] R. Brandenberger, Trans-Planckian Censorship Conjec- [18] A. Ashtekar, T. Pawlowski and P. Singh, Quantum Na- ture and Early Universe Cosmology, arXiv:2102.09641. ture of the Big Bang: An Analytical and Numerical In- [10] A. Borde and A. Vilenkin, Eternal inflation and the ini- vestigation. I., Phys. Rev. D73, 124038 (2006). tial singularity, Phys. Rev. Lett. 72, 3305 (1994); A. [19] A. Ashtekar, T. Pawlowski and P. Singh, Quantum na- Borde, A.H. Guth, and A. Vilenkin, Inflationary Space- ture of the big bang: improved dynamics, Phys. Rev. times Are Incomplete in Past Directions, Phys. Rev. D74, 084003 (2006). Lett. 90, 151301 (2003). [20] A. Ashtekar, A. Corichi. and P. Singh, Robustness of key [11] A, Linde, On the problem of initial conditions for infla- features of loop quantum cosmology, Phys. Rev. D77, tion, Found Phys 48, 1246(2018); A, Linde, Inflationary 024046 (2010). Cosmology after Planck 2013, arXiv: 1402.0526. [21] K. Giesel, B.-F. Li and P. Singh, Towards a reduced [12] N. D. Birrell and P.C.W. Davies, Quantum fields phase space quantization in loop quantum cosmology in curved space (Cambridge University Press, Cam- with an inflationary potential, Phys. Rev. D102, 126024 bridge, 1982); R.M. Wald, Quantum Field Theory (2020). om Curved Spacetime and Black Hole Thermodynam- [22] P. Singh, Are loop quantum cosmos never singular?, ics (The University of Chicago Press, Chicago, 1994); Class. Quant. Grav. 26, 125005 (2009); P. Singh, Loop V.F. Mukhanov, and S. Winitzki, Introduction to Quan- quantum cosmology and the fate of cosmological singu- tum Effects in Gravity (Cambridge University Press, larities, Bull. Astron. Soc. India 42, 121 (2014). Cambridge, 2007); L. Parker and D. Toms, Quantum [23] P. Singh, Glimpses of Space-Time Beyond the Singular- Field Theory in Curved Spacetime: Quantized Fields ities Using Supercomputers, Comput. Sci. Eng. 20, 26 and Gravity (Cambridge University Press, Cambridge, (2018). 2009). [24] P. Diener, B. Gupt and P. Singh, Numerical simula- [13] S. Weinberg, in General Relativity, An Einstein Cen- tions of a loop quantum cosmos: robustness of the quan- tenary Survey, edited by S.W. Hawking and W. Israel tum bounce and the validity of effective dynamics, Class. (Cambridge University Press, Cambridge, 1980); S. Car- Quant. Grav. 31, 105015 (2014). lip, Quantum Gravity in 2+1 Dimensions, Cambridge [25] P. Diener, B. Gupt, M. Megevand and P. Singh, Nu- Monographs on Mathematical Physics (Cambridge Uni- merical evolution of squeezed and non-Gaussian states versity Press, Cambridge, 2003); C. Kiefer, Quantum in loop quantum cosmology, Class. Quant. Grav. 31, Gravity (Oxford Science Publications, Oxford Univer- 165006 (2014). sity Press, 2007). [26] P. Diener, A. Joe, M. Megevand and P. Singh, Numer- [14] M.B. Green, J.H. Schwarz and E. Witten, Superstring ical simulations of loop quantum Bianchi-I spacetimes, Theory: Vol.1 & 2, Cambridge Monographs on Math- Class. Quant. Grav. 34, 094004 (2017). ematical Physics (Cambridge University Press, Cam- [27] I. Agullo, A. Ashtekar and B. Gupt, Phenomenology bridge, 1999); J. Polchinski, String Theory, Vol. 1 & 2 with fluctuating quantum geometries in loop quantum (Cambridge University Press, Cambridge, 2001); C. V. cosmology, Class. Quantum Grav. 34, 074003 (2017). Johson, D-Branes, Cambridge Monographs on Math- [28] J. L. Dupuy and P. Singh, Hysteresis and beats in loop ematical Physics (Cambridge University Press, Cam- quantum cosmology, Phys. Rev. D101, 086016 (2020). bridge, 2003); K. Becker, M. Becker, and J.H. Schwarz, [29] L. Gordon, B.-F. Li and P. Singh, Quantum gravita- String Theory and M-Theory (Cambridge University tional onset of Starobinsky inflation in a closed universe, Press, Cambridge, 2007). Phys. Rev. D103, 046016 (2021). [15] A. Ashtekar and J. Lewandowski, Background indepen- [30] M. Motaharfar and P. Singh, On the role of dissipa- dent quantum gravity: A status report, Class. Quantum tive effects in the quantum gravitational onset of warm Grav. 21, R53 (2004); T. Thiemann, Modern Canoni- Starobinsky inflation in a closed universe, arXiv:gr- cal Quantum General Relativity Cambridge University qc/2102.09578. Press, Cambridge, 2007); C. Rovelli, Quantum gravity [31] M. Bojowald, G. M. Hossain, M. Kagan, S. Shankara- (Cambridge University Press, Cambridge, 2008); M. Bo- narayanan, Anomaly freedom in perturbative loop quan- jowald, Canonical Gravity and Applications: Cosmol- tum gravity, Phys. Rev. D78, 063547 (2008). ogy, Black Holes, and Quantum Gravity (Cambridge [32] T. Cailleteau, A. Barrau, J. Grain, F. Vidotto, Consis- University Press, Cambridge, 2011); R. Gambini and J. tency of holonomy-corrected scalar, vector and tensor Pullin, A First Course in Loop Quantum Gravity (Ox- perturbations in Loop Quantum Cosmology, Phys. Rev. ford University Press, Oxford, 2011); A. Ashtekar and D86, 087301 (2012). J. Pullin, Loop Quantum Gravity - the first 30 years [33] T. Cailleteau, J. Mielczarek, A. Barrau, J. Grain, (World Scientific, Singapore, 2017). Anomaly-free scalar perturbations with holonomy cor- [16] A. Ashtekar and P. Singh, Loop quantum cosmology: a rections in loop quantum cosmology, Class. Quant. Grav. status report, Class. Quantum Grav. 28 (2011) 213001; 29, 095010 (2012). A. Ashtekar and A. Barrau, Loop quantum cosmology: [34] I. Agullo, A. Ashtekar and W. Nelson, Quantum gravity From pre-inflationary dynamics to observations, Class. extension of the inflationary scenario, Phys. Rev. Lett. Quantum Grav. 32 (2015) 234001; M. Bojowald, Quan- 109, 251301 (2012). tum cosmology: a review, Rep. Prog. Phys. 78 (2015) [35] I. Agullo, A. Ashtekar and W. Nelson, Extension of 023901; I. Agullo and P. Singh, Loop Quantum Cosmol- the quantum theory of cosmological perturbations to the ogy, in Loop Quantum Gravity: The First 30 Years, Planck era, Phys. Rev. D87, 043507 (2013). edited by A. Ashtekar and J. Pullin (Wald Scientific, [36] I. Agullo, A. Ashtekar and W. Nelson, The pre- Singapore, 2017). inflationary dynamics of loop quantum cosmology: con- [17] A. Ashtekar, T. Pawlowski and P. Singh, Quantum na- fronting quantum gravity with observations, Class. 24

Quantum Grav. 30, 085014 (2013). 98 (2018). [37] M. Fern´andez-M´endez, G. A. Mena Marug´anand J. [57] T. Zhu, A. Wang, K. Kirsten, G. Cleaver, and Q. Sheng, Olmedo, Hybrid quantization of an inflationary uni- Primoridial non-Gaussianity and power asymmetry with verse, Phys. Rev. D86, 024003 (2012). quantum gravitational effects in loop quantum cosmol- [38] M. Fern´andez-M´endez, G. A. Mena Marug´anand J. ogy, Phys. Rev. D97, 043501 (2018). Olmedo, Hybrid quantization of an inflationary model: [58] Q. Wu, T. Zhu, and A. Wang, Nonadiabatic evolution of The flat case, Phys. Rev. D88, 044013 (2013). primordial perturbations and non-Gaussianity in hybrid [39] L. Castell´oGomar, M. Fern´andez-M´endez,G. A. Mena approach of loop quantum cosmology Phys. Rev. D98, Marugan and J. Olmedo, Cosmological perturbations in 103528 (2018). hybrid loop quantum cosmology: Mukhanov-Sasaki vari- [59] I. Agullo, B. Bolliet and V. Sreenath, Non-Gaussianity ables, Phys. Rev. D90, 064015 (2014). in loop quantum cosmology, Phys. Rev. D97, 066021 [40] L. Castell´o Gomar, M. Mart´ın-Benito, G. A. Mena (2018). Marug´an, Gauge-invariant perturbations in hybrid [60] Y. Akrami et al. (Planck collaboration), Planck 2018 re- quantum cosmology, JCAP 1506 (2015) 045. sults. I. Overview and the cosmological legacy of Planck, [41] F. Ben´ıtezMart´ınezand J. Olmedo, Primordial tensor A&A, 641 (2020) A1. modes of the early Universe, Phys. Rev. D93, 124008 [61] Y. Akrami et al. (Planck collaboration), Planck 2018 (2016). results. VII. Isotropy and Statistics of the CMB, A&A, [42] B. Gupt, P. Singh, Quantum gravitational Kasner tran- 641 (2020) A7. sitions in Bianchi-I spacetime, Phys. Rev. D86, 024034 [62] D.J. Schwarz, C.J. Copi, D. Huterer, G.D. Starkman, (2012). CMB Anomalies after Planck, Class. Quantum Grav. [43] B. Gupt, P. Singh, A quantum gravitational inflationary 33 (2016) 184001. scenario in Bianchi-I spacetime, Class. Quantum Grav. [63] A. Ashtekar, B. Gupt, D. Jeong, and V. Sreenath, Al- 30, 145013 (2013). leviating the Tension in the Cosmic Microwave Back- [44] I. Agullo, J. Olmedo and V. Sreenath, Predictions for ground Using Planck-Scale Physics, Phys. Rev. Lett. the Cosmic Microwave Background from an Anisotropic 125, 051302 (2020); Cosmic tango between the very Quantum Bounce, Phys. Rev. Lett. 124, 251301 (2020). small and the very large: Addressing CMB anomalies [45] I. Agullo, J. Olmedo and V. Sreenath, Observational through Loop Quantum Cosmology, arXiv:2103.14568. consequences of Bianchi I spacetimes in loop quantum [64] I. Agullo, D. Kranas, and V. Sreenath, Anomalies in cosmology, Phys. Rev. D102, 043523 (2020). the CMB from a cosmic bounce, Gen. Relativ. Grav. 53 [46] I. Agullo and N. A. Morris, Detailed analysis of the pre- (2021) 17; Large scale anomalies in the CMB and non- dictions of loop quantum cosmology for the primordial Gaussianity in bouncing cosmologies, Class. Quantum power spectra, Phys. Rev. D92, 124040 (2015). Grav. 38, 065010 (2021). [47] B. Bonga and B. Gupt, Inflation with the Starobin- [65] J. Yang, Y. Ding and Y. Ma, Alternative quantization sky potential in loop quantum cosmology, Gen. Relativ. of the Hamiltonian in loop quantum cosmology II: In- Grav. 48, 1 (2016). cluding the Lorentz term, Phys. Lett. B682 (2009) 1. [48] B. Bonga and B. Gupt, Phenomenological investigation [66] A. Dapor and K. Liegener, Cosmological Effective of a quantum gravity extension of inflation with the Hamiltonian from full Loop Quantum Gravity Dynam- Starobinsky potential, Phys. Rev. D93, 063513 (2016). ics, Phys. Lett. B785 (2018) 506. [49] A. Ashtekar and B. Gupt, Quantum Gravity in the Sky: [67] A. Dapor and K. Liegener, Cosmological Coherent State Interplay between fundamental theory and observations, Expectation Values in LQG I. Isotropic Kinematics, Class. Quantum Grav. 34, 014002 (2017). Quant. Grav. 35 (2018) 135011. [50] A. Ashtekar and B. Gupt, Initial conditions for cosmo- [68] T. Thiemann, Gauge Field Theory Coherent States logical perturbations, Class. Quant. Grav. 34, 035004 (GCS): I. General Properties, Class. Quantum Grav. (2017). 18 (2001) 2025; T. Thiemann, O. Winkler, Gauge Field [51] T. Zhu, A. Wang, K. Kirsten, G. Cleaver, Q. Sheng, Theory Coherent States (GCS): II. Peakedness Prop- Universal features of quantum bounce in loop quantum erties, Class. Quantum Grav. 18 (2001) 2561; T. Thie- cosmology, Phys. Lett. B773 (2017) 196. mann, Complexifier coherent states for quantum general [52] T. Zhu, A. Wang, K. Kirsten, G. Cleaver, Q. Sheng, Pre- relativity, Class. Quantum Grav. 23 (2006) 2063. inflationary universe in loop quantum cosmology, Phys. [69] T. Thiemann, Quantum Spin Dynamics (QSD), Class. Rev. D96, 083520 (2017). Quantum Grav. 15, 839 (1998); Quantum Spin Dynam- [53] D. Mart´ın de Blas and J. Olmedo, Primordial power ics (QSD) II, Class. Quantum Grav. 15, 875 (1998); spectra for scalar perturbations in loop quantum cosmol- K. Giesel, T. Thiemann, Algebraic Quantum Gravity ogy, JCAP 1606, 029 (2016). (AQG) I. Conceptual Setup, Class. Quantum Grav. 24, [54] L. Castell´oGomar, D. Mart´ın de Blas, G. A. Mena 2465 (2007). Marug´anand J. Olmedo, Hybrid loop quantum cos- [70] A. Corichi and P. Singh, Is loop quantization in cosmol- mology and predictions for the cosmic microwave back- ogy unique?, Phys. Rev. D78, 024034 (2008). ground, Phys. Rev. D96, 103528 (2017). [71] M. Assanioussi, A. Dapor, K. Liegener and [55] B. E. Navascues, D. M. de Blas and G. A. M. Marugan, T. Pawlowski, Emergent de Sitter epoch of the quantum Time-dependent mass of cosmological perturbations in Cosmos from Loop Quantum Cosmology, Phys. Rev. the hybrid and dressed metric approaches to loop quan- Lett. 121, 081303 (2018) [arXiv:1801.00768]; Emergent tum cosmology, Phys. Rev. D97, 043523 (2018). de Sitter epoch of the loop quantum cosmos: A detailed [56] B. Elizaga Navascu´es, D. Mart´ın de Blas and G. A. analysis, Phys. Rev. D100, 084003 (2019); Challenges Mena Marug´an, The vacuum state of primordial fluc- in recovering a consistent cosmology from the effective tuations in hybrid loop quantum cosmology, Universe 4, dynamics of loop quantum gravity, Phys. Rev. D100, 25

106016 (2019); K. Liegener and P. Singh, Some physical 205008 (2013). implications of regularization ambiguities in SU(2) [90] J. Engle, Relating loop quantum cosmology to loop gauge-invariant loop quantum cosmology, Phys. Rev. quantum gravity: Symmetric sectors and embeddings, D100, 124049 (2019). Class. Quant. Grav. 24, 5777 (2007); J. Brunnemann [72] B.-F. Li, P. Singh, A. Wang, Towards cosmological dy- and C. Fleischhack, On the configuration spaces of namics from loop quantum gravity, Phys. Rev. D97, homogeneous loop quantum cosmology and loop quan- 084029 (2018). tum gravity, arXiv:0709.1621; J. Brunnemann and [73] B.-F. Li, P. Singh, A. Wang, Qualitative dynamics and T. A. Koslowski, Symmetry Reduction of Loop Quan- inflationary attractors in loop cosmology, Phys. Rev. tum Gravity, Class. Quant. Grav. 28, 245014 (2011). D98, 066016 (2018). [91] B.-F. Li and P. Singh, Singularity resolution may be for- [74] B.-F. Li, P. Singh, A. Wang, Genericness of pre- bid a cyclic evolution. To appear. inflationary dynamics and probability of the desired [92] A. Bhardwaj, E.J. Copeland, J. Louko, Inflation in Loop slow-roll inflation in modified loop quantum cosmologies, Quantum Cosmology, Phys. Rev. D99, 063520 (2019). Phys. Rev. D100, 063513 (2019). [93] M. Shahalam, M. Sharma, Q. Wu, A. Wang, Pre- [75] S. Saini, P. Singh, Generic absence of strong singulari- inflationary dynamics in loop quantum cosmology: ties and geodesic completeness in modified loop quantum Power-law potentials, Phys. Rev. D96 (2017) 123533. cosmologies, Class. Quantum Grav. 36 (2019) 105014. [94] M. Shahalam, Preinflationary dynamics of power-law [76] S. Saini, P. Singh, Von Neumann stability of modified potential in loop quantum cosmology, Universe 4 ( 2018) loop quantum cosmologies, Class. Quantum Grav. 36 87. (2019) 105010. [95] M. Shahalam, M. Sami, A. Wang, Pre-inflationary dy- [77] N. Dadhich, A. Joe and P. Singh, Emergence of product namics of α-attractor in loop quantum cosmology, Phys. of constant curvature spaces in loop quantum cosmology, Rev. D98, 043524 (2018). Class. Quantum Grav. 32 (2015) 185006. [96] M. Shahalam, M.A. Ajmi, R. Myrzakulov, A. Wang, [78] B.-F. Li, P. Singh, A. Wang, Primordial power spectrum Revisiting pre-inflationary universe of family of α- from the dressed metric approach in loop cosmologies, attractor in loop quantum cosmology, Class.Quant.Grav. Phys. Rev. D100, 086004 (2020). 37, 195026 (2020). [79] I. Agullo, Primordial power spectrum from the Dapor- [97] M. Sharma, M. Shahalam, W. Qiang, A. Wang, Prein- Liegener model of loop quantum cosmology, Gen. Rela- flationary dynamics in loop quantum cosmology: Mon- tiv. Grav. 50 (2018) 91. odromy Potential, JCAP, 11 (2018) 003. [80] A. Garc´ıa-Quismondo and G. A. Mena Marug´an, [98] K. Xiao and S.-Q. Wang, Pre-inflation dynamical behav- Martin-Benito-Mena Marugan-Olmedo prescription for ior of warm inflation in loop quantum cosmology, Mod. the Dapor-Liegener model of loop quantum cosmology, Phys. Lett. A35(2020) 2050293. Phys. Rev. D99, 083505 (2019). [99] K. Xiao, Tachyonic inflation in loop quantum cosmol- [81] A. Garc´ıa-Quismondo and G. A. Mena Marug´an, ogy, Eur. Phys. J. C79 (2019) 1019. Dapor-Liegener formalism of loop quantum cosmology [100] M. Sharma, T. Zhu, A. Wang, Background dynamics of for Bianchi I spacetimes, Phys. Rev. D101, 023520 pre-inflationary scenario in Brans-Dicke loop quantum (2020). cosmology, Commun. Theor. Phys. 71 (2019) 1205. [82] L. Castell´oGomar, A. Garc´ıa-Quismondo and G. A. [101] W.-J. Jin, Y.-G. Ma, T. Zhu, Pre-inflationary dynamics Mena Marug´an, Primordial perturbations in the Dapor- of Starobinsky inflation and its generalization in Loop Liegener model of hybrid loop quantum cosmology, Phys. Quantum Brans-Dicke Cosmology, JCAP 02 (2019) 010. Rev. D102, 083524 (2020). [102] K. Xiao, Tachyon field in loop cosmology, Phys. Lett. [83] A. Garc´ıa-Quismondo, G. A. Mena Marug´anand G. B811 (2020) 135859. S´anchez P´erez, The time-dependent mass of cosmolog- [103] P. Singh, K. Vandersloot and G. V. Vereshchagin, Non- ical perturbations in loop quantum cosmology: Dapor- singular bouncing universes in loop quantum cosmology, Liegener regularization, Class. Quantum Grav. 37, Phys. Rev. D74 (2006) 043510. 195003 (2020). [104] X. Zhang and Y. Ling, Inflationary universe in loop [84] B.-F. Li, J. Olmedo, P. Singh, and A. Wang, primordial quantum cosmology, JCAP, 08 (2007) 012. scalar power spectrum from the hybrid approach in loop [105] A. Ashtekar and D. Sloan, Probability of Inflation in cosmologies, Phys. Rev. D102, 126025 (2020). Loop Quantum Cosmology, Gen. Relativ. Gravit. 43, [85] W. Kami´nski, M. Kolanowski, and J. Lewandowski, 3619 (2011); A. Ashtekar and D. Sloan, Loop quantum Dressed metric predictions revisited, Class. Quantum cosmology and slow roll inflation, Phys. Lett. B 694, Grav. 37 (2020) 095001. 108 (2011) [86] K.A. Meissner, Black hole entropy in Loop Quantum [106] A. Corichi and A. Karami, On the measure problem in Gravity, Class. Quantum Grav. 21 (2004) 5245. slow roll inflation and loop quantum cosmology, Phys. [87] A. Ashtekar, M. Bojowald, and J. Lewandowski, Math- Rev. D83 104006 (2011). ematical structure of loop quantum cosmology, Adv. [107] A. Corichi and D. Sloan, Inflationary Attractors and Theor. Math. Phys. 7, 233 (2003). their Measures, Class. Quant. Grav. 31, 062001 (2014) [88] K. Liegener and P. Singh, Gauge-invariant bounce from [108] S. Bedic and G. Vereshchagin, Probability of inflation loop quantum gravity, Class. Quant. Grav. 37, 085015 in Loop Quantum Cosmology, Phys. Rev. D99, no. 4, (2020); New Loop Quantum Cosmology Modifications 043512 (2019) from Gauge-covariant Fluxes, Phys. Rev. D 100, 124048 [109] L. Linsefors and A. Barrau, Duration of inflation and (2019). conditions at the bounce as a prediction of effective [89] D. A. Craig and P. Singh, Consistent probabilities isotropic loop quantum cosmology, Phys. Rev. D87 in loop quantum cosmology, Class. Quant. Grav. 30, (2013) 123509. 26

[110] L. Chen and J.-Y. Zhu, Loop quantum cosmology: The F. Cianfrani, Quantum Reduced Loop Gravity: a realis- horizon problem and the probability of inflation, Phys. tic Universe, Phys. Rev. D92, 084065 (2015); E. Alesci Rev. D92 (2015) 084063. and F. Cianfrani, Quantum-Reduced Loop Gravity: Cos- [111] S.W. Hawking, G.F.R. Ellis, The Large Scale Structure mology, Phys. Rev. D87, 083521 (2013). of Spacetime (Cambridge University Press, Cambridge, [118] J. Engle and I. Vilensky, Deriving loop quantum cosmol- 1973). ogy dynamics from diffeomorphism invariance, Phys. [112] D. Baumann, TASI Lectures on Inflation, Rev. D98, 023505 (2018); Uniqueness of minimal loop arXiv:0907.5424. quantum cosmology dynamics, Phys. Rev. D100, 121901 [113] C. L. Bennett et al., 4-Year COBE DMR Cosmic Mi- (2019). crowave Background Observations: Maps and Basic Re- [119] E. Wilson-Ewing, The loop quantum cosmology bounce sults, Astrophys. J. 464, L1 (1996); K. M. Gorski et as a Kasner transition, Class. Quantum Grav. 35, al., Power Spectrum of Primordial Inhomogeneity De- 065005 (2018). termined from the 4-Year COBE DMR Sky Maps, ibid., [120] K. Giesel, B.-F. Li and P. Singh, Towards a reduced 464, L11 (1996); E. Komatsu et al. (WMAP Collab- phase space quantization in loop quantum cosmology oration), Seven-Year Wilkinson Microwave Anisotropy with an inflationary potential, Phys. Rev. D102, 126024 Probe (WMAP) Observations: Cosmological Interpreta- (2020). tion, Astrophys. J. Suppl. Ser. 192, 18 (2011); D. Larson [121] K. Giesel, B.-F. Li and P. Singh, Revisiting the Bardeen et al. (WMAP Collaboration), Seven-Year Wilkinson and Mukhanov-Sasaki equations in the Brown-Kuchar Microwave Anisotropy Probe (WMAP) Observations: and Gaussian dust models, arXiv:2012.14443. Power Spectra and WMAP-Derived Parameters, ibid., [122] B.-F. Li, S. Saini, and P. Singh, Primordial power spec- 192, 16 (2011); P. Ade et al. (PLANCK Collaboration), trum from a matter-Ekpyrotic bounce scenario in loop Planck 2015 results. XX. Constraints on inflation, A&A quantum cosmology, arXiv:2012.10462. 594, A20 (2016); N. Aghanim, et al. (PLANCK Collab- [123] M. Han and H. Liu, Effective Dynamics from Coher- oration), Planck 2018 results. X. Constraints on infla- ent State Path Integral of Full Loop Quantum Gravity, tion, A&A 641, A10 (2020). Phys. Rev. D101, 046003 (2020); Semiclassical limit [114] K.N. Abazajian et al., Inflation Physics from the Cos- of new path integral formulation from reduced phase mic Microwave Background and Large Scale Struc- space loop quantum gravity, Phys. Rev. D102, 024083 ture, Astropart. Phys. 63, 55 (2015); K. Abazajian, (2020); M. Han, H. Li, and H. Liu, Manifestly Gauge- et al., CMB-S4 Decadal Survey APC White Paper, Invariant Cosmological Perturbation Theory from Full arXiv:1908.01062. Loop Quantum Gravity, arXiv:2005.00883. [115] E. Wilson-Ewing, Testing loop quantum cosmology, [124] S. Schander, T. Thiemann, Quantum Cosmologi- Comptes Rendus Physique 18, 207 (2017). cal Backreactions I: Cosmological Space Adiabatic [116] F. Gerhardt, D. Oriti, and E. Wilson-Ewing, The sepa- Perturbation Theory, arXiv:1906.08166; J. Neuser, rate universe framework in group field theory conden- S. Schander, T. Thiemann, Quantum Cosmological sate cosmology, Phys. Rev. D98, 066011 (2018): D. Backreactions II: Purely Homogeneous Quantum Cos- Oriti, The universe as a quantum gravity condensate, mology, arXiv:1906.08185; S. Schander, T. Thie- Comptes Rendus Physique 18, 235 (2017); D. Oriti, mann, Quantum Cosmological Backreactions III: De- L. Sindoni, and E. Wilson-Ewing, Bouncing cosmolo- parametrised Quantum Cosmological Perturbation The- gies from quantum gravity condensates, Class. Quantum ory, arXiv:1906.08194. Grav. 34, 04LT01 (2017); B. Baytas, M. Bojowald, S. [125] G. A. Mena Marug´an,J. Olmedo, and T. Pawlowski, Crowe, Equivalence of models in loop quantum cosmol- Prescriptions in loop quantum cosmology: A compara- ogy and group field theory, Universe 5 (2019) 41. tive analysis, Phys. Rev. D 84, 064012 (2011). [117] J. Olmedo, E. Alesci, Power spectrum of primordial per- [126] R. Brandenberger and P. Peter, Bouncing Cosmologies: turbations for an emergent universe in quantum reduced Progress and Problems, Found Phys. 47 (2017) 797; D. loop gravity, JCAP 04 (2019) 030; E. Alesci, G. Botta, Wand, Duality invariance of cosmological perturbation F. Cianfrani, and S. Liberati, Cosmological singularity spectra, Phys. Rev. D60, 023507 (1999). resolution from quantum gravity: the emergent-bouncing universe, Phys. Rev. D96, 046008 (2017); E. Alesci and