Disformal interactions in the Dark Sector: From driving Early

Dark Energy to confronting cosmological tensions

Pulkit Bansal,1, ∗ Joseph P. Johnson,2, † and S. Shankaranarayanan1, ‡

1
Department of Physics, Indian Institute of Technology Bombay, Mumbai 400076, India
2
Department of Physical Sciences, Indian Institute of Science
arXiv:2508.17003v1 [astro-ph.CO] 23 Aug 2025

Education and Research Mohali, SAS Nagar, Punjab 140306, India

Abstract
The ΛCDM model faces significant challenges, including an incomplete understanding of the dark
sector and persistent tensions in the Hubble constant and the clustering amplitude. To address
these issues, we propose a general disformal coupling between dark energy (DE) and dark matter
from a field-theoretic action which can generate a rich variety of interactions including confor-
mal and pure-momentum coupling scenarios. Our analysis reveals that a pure disformal coupling
naturally produces a unique interacting Early Dark Sector, wherein the interactions with dark
matter suppress the Hubble friction on the DE scalar field leading to a kinetic-driven cosmological
constant-like behavior at early times followed by its dilution as a−6 and eventually leading to a
potential-driven epoch characteristic of late-time dark energy. In contrast to existing Early Dark
Energy (EDE) models that rely on finely-tuned potentials, the EDE-like behavior, in our frame-
work, is purely a consequence of the disformal coupling paired with the dilution of dark matter,
offering a more fundamental and less ad hoc solution to cosmological tensions. This framework
also predicts a suppression of power in the CMB temperature spectrum on large angular scales,
offering a potential physical explanation for the observed low-ℓ anomaly. By deriving these effects
from a fundamental action, our work provides a unified, testable alternative to ΛCDM that can be
constrained by next-generation cosmological surveys and gravitational wave observations.

∗
[email protected]
†
[email protected]
‡
[email protected]

1
I. INTRODUCTION

The standard cosmological model, Λ-Cold Dark Matter (ΛCDM), has been remarkably
successful in describing a wide array of astronomical observations [1–4]. Despite its successes,
the ΛCDM model faces significant challenges, both theoretical and observational [5]. The
fundamental nature of its two main components, dark energy (DE) and dark matter (DM),
remains unknown [6, 7]. Furthermore, persistent tensions have emerged between different
cosmological datasets [8–11]. The most prominent of these is the Hubble tension [12]: a
statistically significant discrepancy between the value of the Hubble constant (H0 ) measured
from the local distance ladder and the value inferred from Cosmic Microwave Background
(CMB) anisotropies [11]. A related tension for the matter clustering amplitude, σ8 , between
CMB predictions and measurements from weak lensing surveys, which has existed for years,
recently reduced to less than 1σ with the KiDS final data release [13]. These tensions,
coupled with recent findings from the Dark Energy Spectroscopic Instrument (DESI) that
hint at a dynamically evolving dark energy rather than a constant Λ [14], suggest that the
standard model may be incomplete and that new physics is required.

A central issue in moving beyond ΛCDM lies in how the dark sector’s components are
modeled. In standard analyses, dark matter and dark energy are treated as perfect fluids [1–
3]. While computationally convenient and effective for describing background evolution and
large-scale phenomena, the fluid description represents a low-energy, phenomenological ap-
proximation of the underlying physics [15, 16]. A significant limitation is its inherent degen-
eracy: a single fluid equation of state can be realized by multiple, distinct field theories [17].
For instance, both a canonical scalar field and a k-essence scalar field, despite their fun-
damentally different kinetic terms and dynamical properties, can yield a perfect fluid with
an identical equation of state [18]. This degeneracy makes it impossible to discern the true
nature of the cosmic components from fluid models alone.

This limitation becomes especially critical when considering interactions between dark
energy and dark matter whose understanding is crucial for potentially addressing the short-
comings of the ΛCDM model [9, 19, 20]. In the fluid picture, such interactions are typically
introduced ad hoc as source terms in the conservation equations, typically dependent on
the energy densities and possibly the Hubble rate [21–24]. However, without a foundational
field-theoretic underpinning, the true dynamics of such an interaction remain obscure. An

2
interaction term introduced at the fluid level does not reveal whether it arises from a direct
coupling between the fundamental fields, a transformation of the gravitational sector, or an
emergent property of the system.
A field-theoretic description, in contrast, provides a more fundamental and robust frame-
work for investigating DE-DM interactions [25–27]. By starting from an action principle
involving specific scalar fields (or other fundamental fields) for dark energy and dark mat-
ter, interactions are naturally encoded in the Lagrangian. This approach not only provides
a concrete physical origin for the interaction but also allows for a detailed investigation of
the dynamics of these interacting fields. Crucially, a field-theoretic framework can reveal
momentum-dependent interactions or other complex behaviors that are inherently averaged
out or unresolvable in a simple fluid description. Furthermore, it offers the potential to con-
nect cosmological interactions to fundamental particle physics or high-energy gravitational
theories, thereby shedding light on the microphysical origin of dark energy and dark matter
and their observed interplay. Thus, while fluid descriptions are practical, a deeper, field-
theoretic investigation is indispensable for truly understanding the nature and dynamics of
dark sector interactions [25, 27].
This need for a more fundamental description becomes particularly urgent when explor-
ing solutions to the aforementioned cosmological tensions. Among the theoretical models
proposed to address these discrepancies, Early Dark Energy (EDE) models have garnered
significant attention, primarily for their potential to alleviate the Hubble tension [28, 29].
Following its phenomenological success, several field-theoretic models of EDE have been
proposed [30–37] (see Ref. [38] for a recent review). In majority of these models, the typical
EDE scalar field is initially frozen in its potential due to Hubble friction followed by rolling
down and subsequent decay of its energy density faster than matter. This transient dark
energy component in the pre-recombination era alters the expansion history to reconcile con-
flicting observations. However, such models rely on finely-tuned, specific forms of scalar field
potentials and are susceptible to radiative corrections that can disrupt the desired evolution.
Moreover, these approaches often treat early- and late-time dark energy as separate enti-
ties, failing to explore potential cosmological connections between the two. This highlights
the need for a robust and well-motivated, first-principles framework to connect promising
solutions like EDE to the underlying physics of the dark sector.
Motivated by these limitations, in this work, we propose a general action based on dis-

3
formal couplings that provides a fundamental field-theoretic framework for DE-DM inter-
actions. Unlike previous models, our approach is capable of generating a rich variety of
interaction scenarios, including the widely studied conformal couplings [25–27, 39–42] as
well as a unique pure momentum coupling at the linear perturbation level.
Importantly, our framework includes a class of pure disformal couplings that naturally
predicts an interacting Early Dark Sector. In this model, the dynamics of dark energy
resembles that of the phenomenological fluid-EDE models [28], but it fundamentally differs
from typical scalar-field-EDE models [31, 32]. Specifically, the disformal interactions cause
a suppression of Hubble friction, leading to a constant kinetic energy for the DE scalar
field at early times. As the energy density of dark matter dilutes, the coupling weakens,
triggering the dilution of the DE kinetic energy (at a rate of a−6 ), which eventually leads to
a potential-dominated phase characteristic of late-time dark energy.
Compared to existing EDE scenarios, our model does not rely on a finely-tuned potential
since the EDE-like behavior is purely a consequence of the disformal coupling. Additionally,
this model leads to a suppression of power on large angular scales in the CMB temper-
ature spectrum compared to the ΛCDM model. By deriving these interactions from a
first-principles action, our work provides a robust theoretical foundation for exploring the
new physics required to move beyond ΛCDM and resolve existing tensions.
This work is organized as follows. In Sec. II, we consider a general action comprised of
arbitrary scalar field Lagrangians describing dark matter and dark energy including a non-
minimal coupling of the DE field to gravitation and perform a general disformal transfor-
mation. We then establish a field-to-fluid mapping of the dark matter field for the resulting
DE-DM interacting theory in the Einstein frame, which is useful for relating to cosmological
observations. In Sec. III , we discuss the cosmology of the general interacting DE-DM frame-
work and a subclass of the interactions which lead to pure-momentum coupling scenarios.
In Sec. IV, we explore the interaction landscape within our framework by identifying seven
different subclasses of DE-DM interactions (labeled as Models M1-M7 in Table I). We then
perform a detailed investigation of the observational consequences for the background cos-
mology, matter and CMB power spectra for three specific cases of interest: pure-momentum
coupling models (Models M6 and M7) and pure disformal interactions (Model M3). We
then summarize our findings for all models in Sec. IV C (Table II). We conclude and discuss
future directions in Sec. V. The three appendices contain the following details: In Appendix

4
A and B, we investigate the cosmological implications of conformal couplings (Models M1
and M2) and other disformal couplings (Models M4 and M5), respectively. Appendix C
details the implementation of our general interacting DE-DM framework using a custom
Python code and testing its accuracy against the Boltzmann code CLASS [43] for the ΛCDM
model.
In this work, we use the the metric signature (−, +, +, +) and natural units where c =
2
1, MPl = (8πG)−1 . Greek alphabets denote the 4-dimensional space-time coordinates, and
Latin alphabets denote the 3-dimensional spatial coordinates. Overbarred quantities (like
ρ(t), p(t)) are evaluated for the FLRW background, a overdot represents the derivative with
respect to cosmic time t, and H denotes the cosmic Hubble parameter. Unless otherwise
specified, subscript ‘χ’, ‘ϕ’ and ‘X’ denote partial derivatives with respect to χ, ϕ, and the
kinetic term of the scalar field ϕ, respectively. Variables with a Tilde (x̃) correspond to
quantities in the original frame; otherwise, they correspond to the Einstein frame.

II. FIELD-THEORETIC DESCRIPTION AND FLUID MAPPING OF A DISFOR-

MALLY INTERACTING DARK SECTOR

Building on the systematic field-theoretic approach to interacting dark energy-dark mat-

ter developed in our previous works [25, 27], we now generalize the analysis. While our
earlier studies focused exclusively on interactions generated by conformal transformations
from quadratic-Horndeski theories, this work extends that framework to disformal trans-
formations [44]. Disformal transformations offer a wider range of metric possibilities than
their conformal counterparts, primarily because they do not generally preserve the angles
between spacetime curves. This can alter the shape of light cones, depending on the form of
the disformal factor D(ϕ, X). It is important to note, however, that some of these transfor-
mations may lead to unphysical results, such as a change in the metric’s signature [44]. In
this work, we choose the disformal couplings that are physical at all times and length scales.

A. Scalar field interactions from a disformal origin

We consider the following action:

Z  2 
4
p MPl ˜
S̃ = d x −g̃ f (ϕ, X̃)R̃ + L̃ϕ (g̃µν , ϕ) + L̃χ (g̃µν , χ) . (1)
2

5
This action describes two gravitationally interacting scalar fields (ϕ and χ). A non-minimal
coupling exists between the field ϕ and the Ricci scalar described by the arbitrary function
f˜(ϕ, X̃). X̃ = −g̃ µν ∇
˜ µ ϕ∇
˜ ν ϕ/2 is the kinetic term of ϕ. The non-minimal coupling function

f˜(ϕ, X̃) is the most general form and encompasses all the models considered in Ref. [27], as
well as the new cases we will explore in this work.
Our first step is to remove this non-minimal coupling term via a metric transformation.
For this purpose, we perform a general disformal transformation given by [44]:

g̃µν = C(ϕ, X)gµν + D(ϕ, X)ϕµ ϕν , (2)

where C and D are arbitrary functions of ϕ and X, and X ≡ −g µν ∇µ ϕ∇ν ϕ/2 ≡ −ϕµ ϕµ /2.
Under the above disformal transformation, the action (1) transforms to the following form:

√ 2
Z 
MPl
f˜(ϕ, X̃) C(C − 2XD)R + Lϕ (gµν , ϕ)
4
p
S= d x −g
2
√ i
+ C 3/2 C − 2DX L̃χ (C(ϕ, X)gµν + D(ϕ, X)ϕµ ϕν , χ) . (3)

Let us now examine the transformation of each term in action (1) to those in action (3).
Under the disformal transformation (2), the Ricci scalar R̃ transforms into a combination
of R and terms explicitly related to the scalar field ϕ and its derivatives. Consequently, the
transformed Lagrangian for the scalar field ϕ, denoted as Lϕ , will include these additional
contributions from the metric transformation of the Ricci scalar and the original Lagrangian
√
L̃ϕ , along with the Jacobian factor from the metric transformation of −g̃. The Lagrangian
L̃χ remains invariant in form, as the disformal transformation depends only on ϕ, and we
do not perform any redefinition of the scalar field χ.
For the disformal transformation to be invertible and to preserve the metric signature
and causality, the conformal and disformal factors C(ϕ, X) and D(ϕ, X), respectively, must
satisfy the following conditions:

C > 0 , C(C − XCX + 2X 2 DX ) ̸= 0 , C − 2DX > 0 (4)

The non-minimal coupling between the field ϕ and gravitation is effectively removed if the
non-minimal coupling function f˜(ϕ, X̃) satisfies the following condition:

f˜(ϕ, X̃) C(C − 2DX) = 1 ,

p
(5)

6
This transforms the action into the following form in the Einstein frame:
√
 2
√
Z 
4 MPl 3/2
S= d x −g R + Lϕ (gµν , ϕ) + C C − 2DX L̃χ (C(ϕ, X)gµν + D(ϕ, X)ϕµ ϕν , χ) . (6)
2
This action demonstrates a key result: the non-minimal gravitational interaction in the
original frame is mapped to a non-gravitational interaction between the fields ϕ and χ in
the Einstein frame, where the gravitational sector is standard. This systematic process al-
lows for the generation of a wide class of non-gravitational interactions. The precise nature
of these interactions is determined by the initial non-minimal coupling f˜(ϕ, X̃) and the cho-
sen disformal transformation functions, C and D. This highlights that the final Lagrangian
describing the interacting scalar fields is not unique. Therefore, through disformal transfor-
mations, one can systematically generate non-gravitational interactions between the scalar
fields ϕ and χ with arbitrary field descriptions.
Our next objective is to explore the DE-DM interactions generated by a general disformal
coupling. To this end, we extend the field-fluid mapping of our earlier work [25, 27] by
describing the χ-field as a k-essence scalar field [45, 46].

B. Dark sector interactions with a generalized disformal coupling

For the purposes of this study, we shall assume a k−essence description for the scalar field
χ in the original frame [45, 46]. This is motivated by the desire to maintain transparency and
understand the effects of specific interaction types. The k−essence Lagrangian is defined as:
 
L̃χ (g̃µν , χ) = P̃1 χ, Ỹ , (7)

˜ µ χ∇
where P̃1 is an arbitrary function of the field χ and its kinetic term Ỹ = −g̃ µν ∇ ˜ ν χ/2.

The action (6) then transforms to the following form in the Einstein frame:
√
 2
√
Z 
4 MPl 3/2
S = d x −g R + Lϕ (gµν , ϕ) + C C − 2XD × P1 (χ, Z) , (8)
2
where
1 1
Y ≡ − g µν ∇µ χ∇ν χ = − χµ χµ ,
2 2
µ 2 (9)
Y D (χ ϕµ )
Z≡ + .
C 2C (C − 2XD)
Varying the action with respect to the metric gµν will yield the Einstein equations,
1
Gµν = T ,
2 µν
(10)
MPl

7
where, the energy-momentum tensor is given by:
√
(L ) 3/2
√ p CD2 (ϕα χα )2
Tµν = Tµν ϕ + gµν C C − 2DX P1 + C(C − 2DX)χµ χν P1Z + ϕµ ϕν P1Z
(C − 2DX)3/2
r
C C 3/2 D
− D(ϕα χα )(χµ ϕν + ϕµ χν ) P1Z − √ ϕµ ϕν P1
C − 2DX C − 2DX
"r r #
D(C − XD)(ϕα χα )2
 
C C − 2DX
+ CX ϕµ ϕν (2C − 3DX) P1 − Y + P1Z
C − 2DX C (C − 2XD)2
" 3/2 #
C 3/2

1 C α 2
+ DX ϕµ ϕν − √ X P1 + (ϕα χ ) P1Z . (11)
C − 2DX 2 C − 2DX

(L )
The specific form of the energy-momentum tensor Tµν ϕ is directly determined by the
Lagrangian Lϕ which describes the dynamics of the scalar field ϕ. However, the non-
gravitational interactions between ϕ and χ in the Einstein frame leads to violation of the
energy conservation of the individual components. Hence, there is no unique decomposition
of the stress-energy tensor. To enable the mapping of the χ field to a perfect fluid and define
its corresponding four-velocity, we split the total energy-momentum tensor in the following
manner:
" 3/2 #
C 3/2

(ϕ) (L ) 1 C
Tµν = Tµν ϕ + DX ϕµ ϕν − √ X P1 + (ϕα χα )2 P1Z
C − 2DX 2 C − 2DX
"r r #
D(C − XD)(ϕα χα )2
 
C C − 2DX
+ CX ϕ µ ϕ ν (2C − 3DX) P1 − Y + P1Z
C − 2DX C (C − 2XD)2
C 3/2 D
− √ ϕµ ϕν P1 , (12a)
C − 2DX
√
(χ) 3/2
√ p CD2 (ϕα χα )2
Tµν = gµν C C − 2DX P1 + C(C − 2DX)χµ χν P1Z + ϕµ ϕν P1Z
(C − 2DX)3/2
r
C
− D(ϕα χα )(χµ ϕν + ϕµ χν ) P1Z . (12b)
C − 2DX

The interaction between the two scalar fields can be described as [25]:

Qν = ∇µ Tµν
(χ)
= −∇µ Tµν
(ϕ)
. (13)

The field representation yields a detailed, though complex, expression for the interaction
strength Qν . To establish a more direct connection with physical observables, we present
(χ)
Qν using quantities derived from the energy-momentum tensor of the χ field (Tµν ):

T (χ) D(Cα ϕα )ϕν

 
D (χ)
Qν = Cν − − T ϕαβ (14)
2C (C − 2DX) (C − 2DX) αβ

8
 (χ)
" #
D (χ) α β Tαβ q αβ
+ T C ϕ ϕν + (Cµ ϕµ ϕν + 2XCν )
C(C − 2DX) αβ 6
  
1 D (χ) α β 2X (χ) αβ
+ − Tαβ ϕ ϕ + T q (2XDν + Dµ ϕµ ϕν )
(C − 2DX) 2C 3 αβ
  
(χ) α β Dν α β C D (χ)i
− Tαβ D ϕ ϕν + Tαβ ϕ ϕ + √ ∇µ √ Ti ϕν .
2 C − 2DX 3C C − 2DX
(χ)
where T (χ) corresponds to the trace of Tµν . The analysis presented here represents the
most general form of DE-DM interactions arising from a disformal transformation of the
field theoretic action (1), assuming a k−essence description for χ. This is because setting
the disformal factor D = 0 recovers the extended conformal coupling interaction strength
derived in our previous work [27]. The close correspondence of this conformal coupling limit
to the findings in Refs. [25, 27] allows us to establish a physical mapping of the two scalar
fields to the dark sector. Consequently, we will identify ϕ as the dark energy (DE) field and
χ as the dark matter (DM) fluid throughout the rest of this work.

C. Field-Fluid mapping of dark matter for the generalized disformal coupling

As mentioned in the introduction, matter fields in cosmology are often described as

a perfect fluid. Although field theory provides a fundamental description, cosmological
observations are often more effectively analyzed using this fluid framework [47, 48]. To
facilitate this connection, we translate the field-theoretic representation of dark matter into
a perfect fluid. The energy-momentum tensor for such a perfect fluid takes the following
form:

Tµν = pDM gµν + (ρDM + pDM )uµ uν . (15)

(DM ) (χ)
To map Tµν (≡ Tµν ) to the perfect fluid tensor, we must consistently define the DM fluid
four-velocity, energy density and pressure. We begin this process by defining the DM fluid
four-velocity in the following manner:
−1/2
D(ϕα χα ) 2D(C − DX)(ϕα χα )2
  
αβ
uµ = − χµ − ϕµ × −g χα χβ + . (16)
(C − 2DX) (C − 2DX)2
This corresponds to mapping the DM fluid energy density ρDM and pressure pDM , via the
following relations:
D(C − DX)(ϕα χα )2
   
p
ρDM = C(C − 2DX) 2 Y + P1Z − C P1 , (17a)
(C − 2DX)2

9
 √
pDM = C 3/2 C − 2DX P1 . (17b)

With these definitions, the DM perfect fluid is then described by the energy-momentum
(DE)
tensor in Eq. (15). Additionally, we can rewrite Tµν in the following form:
 
(DE) (L ) D CX (3C − 4DX)
Tµν = Tµν ϕ
− pDM ϕµ ϕν + −ρDM + pDM ϕµ ϕν
(C − 2DX) 2C (C − 2DX)
 
DX α 2 2XC
+ (ρDM + pDM )(uα ϕ ) − pDM ϕµ ϕν . (18)
2C (C − 2DX)

Consequently, the interaction strength Qν (14) can be rewritten as:

D(Cα ϕα )ϕν
 
(3pDM − ρDM )
Qν = Cν −
2C (C − 2DX)
D h pDM i
+ (ρDM + pDM )(ϕα uα )(Cβ uβ )ϕν + (Cα ϕα ϕν + 2XCν )
C(C − 2DX) 2
1
−pDM XDν − (ρDM + pDM )(ϕα uα )(Dβ uβ )ϕν

+ (19)
(C − 2DX)

1 α 2 β


+ (ρDM + pDM )(ϕα u ) Dν (C − 2DX) − DDβ ϕ ϕν
2C
D DpDM ϕν
− (ρDM + pDM )ϕαβ uα uβ ϕν − ϕµ [Cµ − 2XDµ + 2Dϕµσ ϕσ ]
(C − 2DX) 2(C − 2DX)2
D
+ [pDM ϕµν ϕµ + ϕν ϕµ ∇µ pDM ] .
(C − 2DX)
(DE)
For a pressureless DM fluid, Tµν and Qν reduce to the following forms:

−CX + DX (uα ϕα )2
 
(DE) (Lϕ )
Tµν = Tµν + ρDM ϕµ ϕν , (20)
2C
ρDM
D(Cα ϕα )ϕν − (C − 2DX)Cν + 2D(ϕα uα )(Cβ uβ )ϕν

Qν =
2C(C − 2DX)
− 2C(ϕα uα )(Dβ uβ )ϕν + (uα ϕα )2 (C − 2DX)Dν − DDβ ϕβ ϕν

− 2CDϕαβ uα uβ ϕν .

(21)

This is the first key result of this work, regarding which, we want to discuss the following
points. First, in the special case where D(ϕ, X) = 0, the above expression reduces to the
conformal coupling results [27]. Furthermore, and in contrast to Ref. [27], we find that the
interaction strength Qν remains non-zero for the p = ρ/3 equation of state. This implies
that a dark energy-dark radiation type model can exhibit non-gravitational interactions
of disformal origin [44], a feature absent in conformal and extended conformal couplings.
Consequently, a dark sector one field-one fluid interacting theory with interactions of the

10
form (19) can be mapped to a classical field theoretic description with DE-DM interac-
tions. We also note from Eq. (17) that the DM fluid description cannot distinguish between
quintessence and k-essence scalar fields χ, indicating the non-uniqueness of the interaction
term (19) for a field-fluid mapping of dark matter, consistent with Ref. [27].

III. COSMOLOGY OF THE DISFORMALLY INTERACTING DARK SECTOR

As discussed in Sec. II, the general disformal transformations describe a wide range of
DE-DM interactions. We, now, look at the dynamics of these interactions in a spatially flat
Friedmann-Lemaı̂tre-Robertson-Walker (FLRW) cosmology. In a particular subclass of these
models, we find a unique interaction signature at the linear perturbation level. Specifically,
while the equation of motion for the dark matter density contrast, δDM , simplifies to a
form with no coupling, the velocity perturbation equation contains non-zero terms resulting
directly from the kinetic coupling. This leads to a framework with a pure momentum coupling
at linear order in perturbation theory.
To go about this, we work in the Newtonian gauge, described by the following metric in
cosmic time [49]:

ds2 = − [1 + 2Ψ] dt2 + a2 (t) [1 + 2Φ] dx2 , (22)

where t denotes cosmic time, a(t) denotes the scale factor and Ψ ≡ Ψ(t, x) and Φ ≡ Φ(t, x)
denote the Newtonian potentials. At the background level in this spatially flat FLRW
universe, a pressureless DM perfect fluid obeys the equation of motion:

ρ̇DM + 3HρDM = −Q0 , (23)

where Q0 denotes the zeroth component of the interaction strength Qν in the FLRW back-
ground. Q0 > 0 corresponds to an energy transfer from DM to DE and vice versa. Within
the homogeneous FLRW background, the spatial components of Qν vanish, leading to zero
net momentum coupling at this level. For general disformal interactions, Q0 takes the fol-
lowing form:
ρDM ϕ̇ h i
Q0 = [CX (4DX − C) − 2CDX X − 2CD] ϕ̈ + Cϕ [4DX − C] − 2CDϕ X , (24)
2C[C − 2DX]
where X = ϕ̇2 /2 is the DE kinetic term in the FLRW background. We assume a quintessence
DE scalar field in the rest of this work, for which the energy density and pressure in the

11
FLRW background are given by:
 
2XDX − CX
ρDE = X + V (ϕ) + ρDM X ,
C
pDE = X − V (ϕ) , (25)

where V (ϕ) represents the scalar field potential. It is clear that the explicit kinetic term
dependence of the couplings C and D directly contributes to the DE energy density while
leaving its pressure unchanged. This imposes a constraint on the allowed forms of these
kinetic couplings, as they must be chosen to avoid negative energy density states for DE.
The DE scalar field is governed by the following equation of motion:
h ρDM 2 2 2


ϕ̈ 1 + −CC X + XC X − XCC XX + 4XCD X − 2X C X D X + 2X CD XX
C2    
ρDM X 4DX
+ 1 − (CX − 2XDX ) 2D + 2XDX + CX 1 −
2(C − 2DX) C C
ρDM X
= −3H ϕ̇ − Vϕ − [Cϕ CX − CCϕX − 2XCϕ DX + 2XCDϕX ]
C2
    
ρDM X 4DX
+ 1 − (CX − 2XDX ) Cϕ − 1 − 2XDϕ . (26)
2(C − 2DX) C C
Note that in the limit C → 1, D → 0, the above equation reduces down to the standard
Klein-Gordan equation.
We now examine the modifications to linear perturbations in the matter fields introduced
by the coupling. The perturbed energy density and pressure of the DE field are given by:
  
˙ ˙ ˙ 2
˙ ˙ ˙ 2 C X − 2XDX
δρDE = + ϕδϕ − ϕ Ψ + V ϕ δϕ − ρDM XδDM + ϕδϕ − ϕ Ψ
C
XρDM 

− 2 δϕ C(C ϕX − 2XDϕX ) − C ϕ (C X − 2XDX )
C 
2
˙ ˙ ˙
+(ϕδϕ − ϕ Ψ) C(C − 2D − 2XD ) − C (C − 2XD )


XX X XX X X X

2
δpDE = +ϕ˙ δϕ
˙ − ϕ˙ Ψ − V ϕ δϕ . (27)

Consistent with our background analysis, the kinetic dependence of these interactions modi-
fies the perturbed energy density of the DE field while its pressure remains unchanged. The
equations of motion governing the DM density contrast δDM and velocity perturbation vDM
are modified to the following form:
 
k Q0 δDM − δQ0
δ̇DM + 3Φ̇ − vDM = , (28)
a ρDM

12
 " #   
k Q0 k δϕ k C X − 2XDX ¨ ˙ ˙ ˙ 2
v̇DM + HvDM + Ψ = vDM + + ϕδϕ − ϕδϕ + ϕ Ψ ,
a ρDM a ϕ˙ a 2C

where δQ0 denotes the time (zeroth) component of the perturbed interaction term, repre-
senting energy transfer at the linear level 1 .
The disformal coupling framework allows for a specific subclass of models where the
interaction between dark energy and dark matter vanishes in the background. This occurs
if the coupling functions C(ϕ, X) and D(ϕ, X) in Eq. (24) satisfy the following relations:
 
2DX 1
Cϕ − = XDϕ ,
C 2
 
2DX 1
CX − = D + XDX , (29)
C 2

the interaction terms Q0 as well as δQ0 vanish, which corresponds to no energy exchange
between the DE and DM fluids at the background as well as the perturbed level. From
Eq. (28), we note that the equation of motion governing the DM density contrast δDM
reduces to the form with no coupling. Interestingly, the velocity perturbation equation
contains non-zero terms resulting from the kinetic coupling. Therefore, this subclass of the
general interacting framework leads to pure momentum coupling between the DE and DM
fluids at linear order in perturbation theory.
Within this subclass of models, we can define a momentum coupling function of the form:

2XDX − CX
M T (ϕ, X) = . (30)
2C

The coupled equations of motion governing the DM fluid, then reduce to the following:

ρ̇DM + 3HρDM = 0 ,
k
δ̇DM + 3Φ̇ − vDM = 0 ,
a  
k k ¨ ˙ ˙ ˙ 2
v̇DM + HvDM + Ψ = − M T (ϕ, X) ϕδϕ − ϕδϕ + ϕ Ψ , (31)
a a

At the background level, the DE field ϕ is governed by the following equation:

ϕ̈ [1 + 2ρDM M T + 2ρDM XM TX ] + 3H ϕ̇ + Vϕ + 2ρDM XM Tϕ = 0 . (32)

1
Explicit form of this term as well as the perturbed DE scalar field equation of motion is complex and not

particularly insightful, so it is not provided here.

13
Additionally, the DE field energy density and pressure at the background and linear level
are given by:

ρDE = X + V (ϕ) + 2 ρDM X M T (ϕ, X) ,

pDE = X − V (ϕ) ,
 
2
δρDE = ˙ ˙ ˙    
ϕδϕ − ϕ Ψ 1 + 2ρDM M T + 2ρDM XM T X + 2XρDM M T ϕ δϕ + M T δDM + V ϕ δϕ ,
2
δpDE = ϕ˙ δϕ
˙ − ϕ˙ Ψ − V ϕ δϕ . (33)

The unique dynamics of this subclass of models, where energy exchange vanishes while a
non-trivial momentum coupling persists, showcases the rich phenomenology enabled by the
general disformal transformations. As we will demonstrate in the rest of this work, the
general disformal framework provides a powerful tool for exploring new scenarios that can
potentially address current cosmological tensions. Interestingly, the separation of energy
and momentum coupling offers a promising way to resolving discrepancies, such as those
related to the Hubble constant and the growth of large-scale structure, by providing new
dynamics that depart from standard interacting dark energy models.

IV. COSMOLOGICAL IMPLICATIONS OF THE DISFORMAL DARK SECTOR

We now turn to a detailed assessment of the cosmological implications of the DE-DM

interactions arising from our general disformal coupling framework. Our analysis will focus
on the effects of these interactions on the background cosmology, large-scale structure (LSS)
formation, and the cosmic microwave background (CMB) observations. Specifically, we
will examine the evolution of the DE field, the Hubble parameter, the linear matter power
spectrum, and the unlensed CMB temperature power spectrum.
For our numerical analysis, we fix the standard cosmological parameters {Ωb h2 , Ωcdm h2 ,
θs , τreio , ns , As } to the Planck 2018 constraints on the ΛCDM model [50]. The normalization
of the DE scalar field potential, V0 , is computed to match the present-day value of ΩDE
derived from these fixed parameters. The results presented in this work are valid for the
following DE potential:

V (ϕ) = V0 ϕn e−λϕ , (34)

14
 Evolution of
Model Form of coupling Modified scalar field equations
DE energy density
Cϕ
C(ϕ) = c0 ϕcf0 ecf1 ϕ ϕ̈ + 3H ϕ̇ + Vϕ + ρDM = 0 ,
M1 2C ρDE ∝ a−2 → a−3 → a0
D=0 ρDE = X + V
3H ϕ̇ + Vϕ
ϕ̈ +  2  = 0,
C(X) = c0 eck1 X CX XCX XCXX
M2 1 + ρDM − + − ρDE ∝ a−3 → a−6 → a0
2C  2C 2  C
D=0 XCX
ρDE = X + V − ρDM
C

3H ϕ̇ + Vϕ
ϕ̈ + = 0,
C = c0 ρDM d0
M3 1+ ρDE ∝ a0 → a−6 → a0
[c0 − 2d0 X]
D = d0
ρDE = X + V
Dϕ XρDM
3H ϕ̇ + Vϕ +
c0 − 2DX
C = c0 ϕ̈ + = 0,
M4 ρDM D ρDE ∝ a0 → a−6 → a0
1+
D(ϕ) = d0 ϕ df0 df1 ϕ
e [c0 − 2DX]
ρDE = X + V
3H ϕ̇ + Vϕ
ϕ̈ +   = 0,
4XDX +2X 2 DXX
c0
C = c0 1 + ρDM  
M5 + (c0 +2X
2 D )(D+XD )
X X ρDE ∝ a−3 → a−6 → a0
dk1 X c0 (c0 −2DX)
D(X) = d0 e
2X 2 DX
 
ρDE = X + V + ρDM
C

3H ϕ̇ + Vϕ
M T (X) = ϕ̈ + = 0,
M6 1 + 2ρDM (M T + XM TX ) ρDE ∝ a−3 → a−6 → a0
m0 X mk0 emk1 X ρDE = X + V + 2ρDM X M T (X)

m1 ϕ̈ + 3H ϕ̇ + Vϕ = 0 ,
M7 M T (X) = ρDE ∝ a−3 → a0
X ρDE = X + V + 2ρDM m1

TABLE I. Modified cosmological background evolution of the DE scalar field resulting from various
sub-classes of our general interacting DE-DM framework assuming a pressureless DM fluid.

with n ≤ 1, which encompasses linear, inverse, constant, and exponential potentials. We

present the numerical results for inverse DE scalar field potential (n = −1, λ = 0) and set
the initial field value to ϕi = 10. This ensures that the dynamics of the uncoupled model
remain close to the ΛCDM model. For all models, we have evolved the system over the

15
redshift range z ∈ [108 , 0] .

To provide a general overview of the interaction landscape within our framework, we

have classified different models based on power-law and exponential dependencies of the
coupling functions C and D on the DE field ϕ and its kinetic term X. These are presented
in Table I. Models M1 and M2 correspond to purely conformal couplings; M3, M4, and
M5 correspond to disformal couplings; while M6 and M7 are examples of the unique pure-
momentum coupling scenario.

We will now discuss these models in detail in the subsequent subsections. Our primary in-
terest lies in the disformally interacting model M3 and the pure-momentum coupling models
(M6, M7) due to their unique features. A brief discussion of the cosmological implications
of purely conformal interactions, which have been investigated by the current authors in
Refs. [25–27], is provided in Appendix A. In Appendix B, we provide a detailed discus-
sion of the cosmology of scalar-field-dependent and kinetic-term dependent pure disformal
coupling models (M4, M5).

A. Pure-Momentum Coupling between DE and DM fluids

We first turn our attention to Models M6 and M7, that exhibit a pure-momentum coupling
between the dark matter and dark energy fluids. While it is not straightforward to find
analytical forms for the coupling functions C(ϕ, X) and D(ϕ, X) that satisfy the conditions
in Eq. (29) and simultaneously yield a simple form for the momentum coupling function
M T (ϕ, X) as defined in Eq. (30), we consider a few specific examples. A detailed analysis
of these coupling functions lies beyond the scope of this work and is deferred to a future
investigation.

For our analysis, we treat the power-law and exponential forms of the coupling functions
assumed for Model M6 in Table I as a toy model. In contrast, Model M7 can be constructed
from a simpler form of C and D, where C(ϕ, X) is a constant and D(ϕ, X) ∝ 1/X .

At the background level, the absence of an energy exchange means the DM fluid evolves
identically to its uncoupled counterpart. The modified equations of motion for the DE scalar
field and the form of its energy density ρDE for Models M6 and M7 are provided in Table I .

16
FIG. 1. Modified evolution of the energy densities of different components (top), DE equation of
state (center) and the Hubble parameter (bottom) due to pure-momentum DE-DM interactions.
Left: Model M6 and Right: Model M7 from Table I .

1. Cosmological evolution within Model M6

We start by discussing the cosmological evolution in Model M6. For numerical analysis,
we assume a constant momentum coupling function, M T (ϕ, X) = m0 . It is important to
note that similar results are obtained for more general power-law or exponential couplings.
As shown in the left panels of Fig. 1 , the background evolution of the DE energy density
(ρDE ), its equation of state (ωDE ), and the Hubble parameter (H) are significantly modified.
As discussed in Sec. III , the non-zero momentum coupling modifies the description of
the DE scalar field by suppressing Hubble friction at early times. This suppression results
in a constant kinetic energy for the DE field. The DE energy density is given by

ρDE = X + V + 2ρDM X M T (ϕ, X) . (35)

17
 ρDE = 2ρDM X M T (X) + X + V
I II III

Early times: Intermediate times: Late times:

ρDM K.E.
Hubble friction is suppressed Kinetic energy decays Potential-dominated
decays decays
=⇒ Constant Kinetic Energy due to Hubble friction late-time DE

I dominates II dominates III dominates

ρDE ∝ a−3 : matter-like ρDE ∝ a−6 : free scalar ρDE ∝ a0 : Cosmological

evolution field-like evolution constant-like evolution

FIG. 2. Flow chart detailing the evolution of DE scalar field ϕ and its energy density ρDE from
early till late times due to Model M6-like pure-momentum coupling between DE and DM fluids.
Note that one obtains a functionally identical evolution of the DE field for kinetic-energy (K.E.)
dependent conformal and disformal coupling models M2 and M5 from Table I .

Consequently, the DE field behaves like a dark matter fluid at early times. As the DM
energy density decays, the Hubble friction becomes dominant causing the DE kinetic energy
to dilute at a rate proportional to a−6 . Eventually, the decaying kinetic energy of the DE
field becomes subdominant to its potential energy, which drives the late-time cosmological
constant-like evolution of dark energy. This evolution is clearly seen in the DE equation
of state, ωDE , shown in the left-center panel of Fig. 1 . Fig. 2 provides a schematic of this
cosmological evolution. A similar background evolution is observed in other models with
kinetic couplings, such as the conformal and disformal models M2 and M5, respectively.

The transition redshift, zt , from matter-like to the free scalar field-like evolution of dark
energy is determined by the coupling strength, m0 . A larger m0 shifts zt to later times. We
find that a transition redshift in the range of roughly 102 < zt < 104 leads to a larger value
for the Hubble constant, H0 , as illustrated in the bottom-right panel of Fig. 1 . The degree
of modification to the cosmological evolution in this model is determined by both m0 and
the initial kinetic energy, Xi , of the DE scalar field. The value of Xi is crucial because it
sets the amplitude of the DE’s contribution to the total energy budget from the onset of the
matter-dominated epoch, which significantly affects H0 . In Fig. 1 , we have dynamically set

18
FIG. 3. Modified linear matter power spectrum (left), CMB temperature power spectrum (right)
and relative deviation from their uncoupled counterparts due to pure-momentum exchange, corre-
sponding to Model M6 from Table I , between the DE and DM fluids.

Xi for different values of m0 such that the evolution of the DE field starts with the same
initial energy density, ρDE (z = zi ) .

We now look at the impact on linear perturbations. The linear matter and CMB temper-
ature power spectra for this model are plotted in Fig. 3 . We find that the model predicts a
power suppression over large scales in the matter power spectrum, but an enhanced struc-
ture growth over small scales, leading to a larger value of σ8 . On the other hand, the CMB
temperature spectrum shows a power enhancement on large angular scales (low multipoles,
ℓ), while power is suppressed on small angular scales (high multipoles). For large coupling
strengths, the low-ℓ power enhancement in the CMB spectrum changes sign around the mul-
tipole range of ℓ = 65 − 72 . As we will discuss in detail in Sec. IV B, the Planck 2018 results
demonstrated that the best-fit amplitude for the low-ℓ spectrum (ℓ < 30) is approximately
10% lower than for the high-ℓ spectrum (ℓ ≥ 30) [50].

19
FIG. 4. Modified linear matter power spectrum (left), CMB temperature power spectrum (right)
and relative deviation from their uncoupled counterparts due to pure-momentum exchange, corre-
sponding to Model M7 from Table I , between the DE and DM fluids.

2. Cosmological evolution within Model M7

We now examine Model M7, which assumes a specific form for the momentum coupling
function: M T (ϕ, X) = m1 /X . The cosmological background evolution for this model is
shown in the right panels of Fig. 1 .
Unlike Model M6, this specific form of coupling does not modify the DE scalar field’s
equation of motion from its standard Klein-Gordan form (cf. Table I). As a result, kinetic
energy of the the DE field simply redshifts away due to Hubble friction. Consequently, for
a non-zero value of the constant m1 , the DE field initially evolves like a DM fluid and then
transitions to a potential-dominated, cosmological constant-like behavior. This is evident
from the evolution of the equation of state, ωDE , which transitions from 0 to −1, as plotted
in the right-center panel of Fig. 1 .
The value of m1 determines the transition redshift, zt , and the impact on the Hubble con-
stant. We find that smaller values of m1 , which cause an earlier transition (at high redshift),
do not produce a visible modification of the background dynamics. Only values of m1 large
enough to cause a transition at zt < 10 lead to a noticeable deviation. Interestingly, unlike

20
M6, this model leads to smaller values of H0 , which further exacerbates the existing tension
between the CMB-inferred value of the present-day expansion rate and local measurements.
The linear matter and CMB TT power spectra for this model are plotted in Fig. 4 .
Similar to Model M6, we observe an enhanced growth of structures on small scales, leading
to a larger value of σ8 . However, we do not find any power suppression on large scales
in the matter power spectrum, unlike M6. The CMB temperature power spectrum shows
similar characteristic deviations to Model M6: enhanced power on large angular scales and
suppressed power on small angular scales, with a change in the sign around ℓ = 70 .

B. Pure disformal coupling, and a unified Early and late dark energy

We now examine Model M3 from Table I, which assumes a pure disformal coupling where
the functions C and D are constants — C = c0 (=1, here) and D = d0 . For these interactions,
the DE field ϕ is governed by the following equation of motion:
3H ϕ̇ + Vϕ
ϕ̈ +   = 0. (36)
ρDM d0
1+
c0 − 2Xd0
In Fig. 5, we have plotted the background evolution of the DE energy density (top left), DM
energy density relative to the non-interacting case (top right), DE effective equation of state
(bottom left) and the Hubble parameter relative to the non-interacting case (bottom right),
for different coupling strengths of the disformal DE-DM interactions. As seen from Fig. 5,
for a non-zero coupling constant d0 , the disformal coupling suppresses Hubble friction at
early times. This leads to ϕ̈ ≈ 0 and a constant kinetic energy for the DE field, causing it
to behave like a kinetic energy-dominated cosmological constant during this epoch. As the
universe expands, the DM energy density dilutes, and the coupling weakens. The DE kinetic
energy remains constant until the effects of the coupling become subdominant to Hubble
friction. Consequently, the DE field starts evolving like a free scalar field: its kinetic energy
decays proportionally to a−6 according to the standard Klein-Gordon equation. Eventually,
the decaying kinetic energy becomes subdominant to its potential energy, and the DE field
behaves like a potential energy-dominated cosmological constant at late times. A schematic
of this modified evolution of the DE energy density, ρDE , is provided in Fig. 6 .
This early-time cosmological constant-like evolution of dark energy closely resembles that
of Early Dark Energy (EDE) models, which are strong candidates for resolving the Hubble

21
FIG. 5. Modified cosmological background evolution of the DE energy density (top left), DM energy
density relative to the non-interacting case (top right), DE effective equation of state (bottom left)
and the relative change in Hubble parameter compared to the non-interacting case (bottom right),
for different coupling strengths of disformal DE-DM interactions (Model M3 in Table I).

tension [28, 29]. However, previously explored EDE models typically rely on finely-tuned
potentials to maintain a Hubble-frozen state at early times, followed by a transition where
the EDE field rolls down and oscillates about the minimum of its potential, causing its energy
density to decay at a rate determined by the shape of its potential [31, 32]. Eventually, the
field becomes completely subdominant at late times. In contrast, the early-time behavior
in our model is a direct consequence of the disformal coupling, which suppresses Hubble
friction and leads to a constant kinetic energy. Furthermore, the transition from a constant
to decaying DE kinetic energy (as a−6 ) is triggered by the dilution of dark matter. This
unique evolution is evident from the DE effective equation of state transition: ωDE,ef f =

22
Suppressed Hubble drag: Free scalar field: Potential energy-dominated
Coupling −6 K.E.
Kinetic energy-dominated ρDE redshifts away as a Cosmological constant
weakens decays
Cosmological Constant (Kinetic energy-dominated) (Late-time dark energy)

FIG. 6. Flow chart detailing the evolution of the DE field energy density ρDE from early till late
times due to disformal interactions with the DM fluid (Models M3 and M4 in Table I).

−1 → 1 → −1 for different coupling strengths (d0 ), as plotted in Fig. 5 . Furthermore, our

model uses a single scalar field for both Early Dark Energy and Late-time Dark Energy
without requiring a fine-tuned potential.
The strength of the coupling, quantified by d0 , determines the transition redshift, zt ,
from the constant kinetic energy phase (ρDE ≈ K.E.) to the decaying phase (ρDE ∝ a−6 ).
A larger d0 shifts this transition to later times (or smaller redshifts), as shown in Fig. 5 .
Similar to EDE models, this transition epoch contributes to the total energy budget, leading
to a significant increase in the Hubble constant if the transition occurs around the redshift
range 104 > zt > 102 or a reduction if it happens at late times (zt < 10). This behavior is
shown in Fig. 5 (bottom right). In contrast to standard EDE models, the increase in the
past expansion rate in our model results from the increased energy density of both the DM
fluid and the DE field. Since this behavior arises from disformal interactions between the
two dark sector components, it is more appropriate to describe this cosmological evolution
as a disformally interacting Early Dark Sector.
Beyond d0 , the amplitude of the contribution of the DE field during the transition and
its impact on H0 are determined by the initial kinetic energy, Xi . Therefore, this class of
interacting models introduces two additional parameters: {d0 , Xi }.

1. Impact on linear perturbations

We now turn to the impact on linear perturbations, which are illustrated in Fig. 7 . We
find that the interactions suppress the growth of matter density fluctuations for smaller val-
ues of d0 , leading to a reduced σ8 . Conversely, larger coupling strengths enhance structure
growth on small scales, corresponding to a larger σ8 . Notably, modes that enter the horizon
when z ≫ zt undergo enhanced structure growth, while modes that remain super-horizon
during this period undergo suppressed growth. The scale (k) of the transition from sup-

23
FIG. 7. Modified linear matter power spectrum, CMB temperature spectrum and relative deviation
from their uncoupled counterparts due to disformal DE-DM interactions (Model M3 in Table I).

pressed to enhanced structure growth in the matter power spectrum is strongly dependent
on the transition redshift zt , while the amplitude of this deviation from the non-interacting
scenario depends on the initial kinetic energy configuration Xi of the DE field.

On the other hand, the effect of the coupling remains significant over large angular scales
(or low multipoles) but diminishes over small scales (higher multipoles) in the CMB power
spectrum. Interestingly, the coupling leads to a power suppression in the TT spectrum
compared to the non-interacting quintessence scenario. Specifically, we find a significant
suppression of power on large angular scales, which gradually fades as ℓ ∼ 100 . These
modifications can be attributed to the effect of coupling on the contributions of the Sachs-
Wolfe and Integrated Sachs-Wolfe (ISW) effects.

The observed CMB exhibits a well-known tension with the standard ΛCDM model
— a suppression of power at large angular scales (low multipoles, ℓ), particularly in the
quadrupole moment (ℓ = 2). While a low quadrupole could be attributed to cosmic vari-
ance, a more detailed analysis by the Planck collaboration revealed a broader suppression of
power. The Planck 2018 results demonstrated that the best-fit amplitude for the low-ℓ spec-
trum (ℓ < 30) is approximately 10% lower than that for the high-ℓ spectrum (ℓ ≥ 30) [50].

24
 Naturally, a number of theoretical models have been proposed to address the low-ℓ power
suppression by modifying the physics of the early universe. These models often propose a
departure from the standard inflationary slow-roll conditions. For instance, some models
suggest a pre-inflationary era or a slow-roll violation occurring approximately 60 e-folds be-
fore the end of inflation. Such a deviation from standard inflation can alter the primordial
power spectrum on the largest scales, corresponding to the low-ℓ modes, leading to a sup-
pression of power where the spectrum would otherwise be nearly scale-invariant [51–53]. In
contrast, our disformal interaction model provides a completely different explanation. By
modifying the late-time dynamics of the dark sector, it produces a suppression of power on
large angular scales, offering a distinct alternative to early-universe scenarios. A detailed
analysis of how the disformal coupling affects individual contributions of the CMB spectra
as well as parameter estimation for best fits and compatibility with datasets will be carried
out in a future work.
This discussion of Model M3, with its constant coupling functions, represents the simplest
form of a pure disformal coupling. Model M4, which incorporates a scalar field dependent
disformal coupling (D ≡ D(ϕ)), yields similar results and therefore does not require a
separate, detailed discussion here. Similarly, Model M5, which assumes a kinetic term
dependence (D ≡ D(X)), leads to a background evolution similar to Model M6 discussed
in the Sec. IV A. For completeness, we discuss the numerical evolution of the background,
matter and CMB power spectra for Models M4 and M5 in Appendix B.

C. Summary of results

To provide a comprehensive overview of the key results, we summarize the characteristics

and main findings for all models discussed in this work in Table II . This table consolidates
the key characteristics of each model at both the background FLRW and linear perturbation
levels, highlighting the unique features and cosmological implications of the different coupling
types. Specifically, we compare their effects on the background evolution of dark energy, the
Hubble parameter, the matter power spectrum, and the CMB temperature power spectrum.
This allows for a direct comparison of the distinct phenomenologies arising from purely
conformal, disformal, and pure-momentum couplings, providing a clear roadmap to the
potential of each model in addressing current cosmological tensions.

25
Model Form of coupling Effect on ρDE and H0 Effect on P(k), σ8 and CℓTT

Large scales: P(k) decreases (↓)

C(ϕ) = c0 ϕcf0 ecf1 ϕ ρDE ∝ a−2 → a−3 → a0 Small scales: P(k) increases (↑) =⇒ σ8 ↑
M1
D=0 Value of H0 increases (↑) ℓ ≲ 30 : CℓTT ↑ (Moderate changee
ℓ ≳ 30 : CℓTT ↓ over low and high ℓ)

C = c0 EDE-like evolution: Large scales: P(k) ↓

0 zt −6 0
M3 ρDE ∝ a − →a →a Small scales: Low d0 : P(k)↓ =⇒ σ8 ↓
D = d0 a b c
(KECC → FF → PECC ) High d0 : P(k)↑ =⇒ σ8 ↑
C = c0 Transition epoch zt d : ℓ ≲ 70: CℓTT ↓ (Large suppressione at low ℓ)
M4 zt ≲ 10 : H0 ↓ ℓ ≳ 70: CℓTT ↓ (Small suppressione at high ℓ)
D(ϕ) = d0 ϕdf0 edf1 ϕ
zt ≫ 10 : H0 ↑
All scales: P(k) ↑ =⇒ σ8 ↑
C(X) = c0 eck1 X
M2 ℓ ≲ 80 − 100 : CℓTT ↑
D=0
(Large enhancemente at low ℓ)
Large scales: P(k) ↓
z Small scales: Low d0 : P(k) ↓ =⇒ σ8 ↓
ρDE ∝ a−3 −
→t
a−6 → a0
High d0 : P(k) ↑ =⇒ σ8 ↑
C = c0 (DM → FFb → PECCc )
M5 Low d0 − ℓ ≲ 70 : CℓTT ↓ (Large suppressione )
D(X) = d0 edk1 X Transition epoch zt d :
ℓ ≳ 70 : CℓTT ↓ (Small suppressione )
zt ≲ 10 : H0 ↓
High d0 − ℓ < 10 and ℓ > 100 : CℓTT ↓
zt ≫ 10 : H0 ↑
10 ≲ ℓ ≲ 100 : CℓTT ↑
Large scales: P(k) ↓
M T (X) = Small scales: P(k) ↑ =⇒ σ8 ↑
M6
m0 X mk0 emk1 X ℓ < 70 : CℓTT ↑ (Large increasee at low ℓ)
ℓ ≳ 70 : CℓTT ↓ (Small suppressione at high ℓ)
ρDE ∝ a−3 → a0 All scales: P(k) ↑ =⇒ σ8 ↑
m1
M7 M T (X) = (DM → PECCc ) ℓ < 70 : CℓTT ↑ (Large increasee at low ℓ)
X
H0 ↓ ℓ ≳ 70 : CℓTT ↓ (Small suppressione at high ℓ)

a
KECC denotes Kinetic-Energy-dominated Cosmological Constant
b
FF denotes Free Field: Kinetic-energy–driven scalar field, governed by the free Klein-Gordan equation
c
PECC denotes Potential-Energy-dominated Cosmological Constant
d
zt depends on the coupling strengths {ck1 , d0 , df0 , df1 , dk1 , m0 , mk0 , mk1 }: Coupling ↑ =⇒ zt ↓
e
For reference, Large change: ≳10%; Moderate change: ∼ 5-6%; Small change: ≲ 1-2% deviation in CℓTT

TABLE II. Table summarizes the key characteristics of each model at both the background FLRW
and linear perturbation levels, highlighting the unique features and cosmological implications of
the different coupling types, including modifications relative to the non-interacting scenario.

26
V. CONCLUSIONS AND DISCUSSIONS

In this work, we have provided a first-principles, field-theoretic framework for DE-DM

interactions using a general disformal coupling. By moving beyond the simple fluid descrip-
tion, our approach offers a rich set of scenarios like conformal, disformal and pure-momentum
coupling models with distinct cosmological behaviors. Our analysis of these models at both
the background and linear perturbation levels reveals compelling connections to the major
tensions faced by the standard ΛCDM model.
Our most significant finding is related to Models M3 and M4, which feature a pure disfor-
mal coupling. These models naturally lead to an interacting Early Dark Sector where the
dark energy field’s evolution exhibits a unique three-stage behavior: an early-time cosmo-
logical constant-like phase driven by a constant kinetic energy, a subsequent decay phase at
a rate proportional to a−6 , and a late-time potential-driven cosmological constant-like phase
accounting for late dark energy. This evolution, purely driven by the disformal coupling and
the dynamics of dark matter, offers a physically motivated alternative to the existing EDE
models, which often require finely-tuned scalar field potentials. The ability of these models
to increase the CMB-inferred Hubble constant, H0 , by a transition occurring around the
epoch of recombination, offers a promising pathway to alleviating the Hubble tension.
Furthermore, these models exhibit distinct signatures in the linear matter and CMB power
spectra. Interacting Early Dark Sector models with a high transition redshift (zt ≳ 105 )
suppress structure growth, while those with a lower transition redshift lead to enhanced
small-scale structure growth. On the other hand, they produce a crucial power suppression
on large angular scales in the CMB temperature spectrum. This suppression is consistent
with the observed low-ℓ CMB anomaly from Planck, demonstrating that our framework
offers a potential resolution for this persistent observational puzzle. We contrast these
results with other models in our framework, such as the pure-momentum coupling Models
M6 and M7, which predict enhanced power at low-ℓ, making it disfavored by current data.
The unique signatures of the disformally interacting Early Dark Sector provide a clear
avenue for future observational tests. Upcoming high-precision surveys from telescopes like
the James Webb Space Telescope (JWST), along with ground-based projects such as the
Dark Energy Spectroscopic Instrument (DESI), the Euclid mission, and the Vera C. Rubin
Observatory’s LSST, will provide unprecedented data. These observations will be able to

27
probe the subtle modifications to the expansion history and structure growth, offering a
critical test for our disformal coupling framework and its potential to resolve cosmic tensions.

To date, these observations have primarily relied on electromagnetic signals, which have
provided critical insights but are subject to limitations in probing certain aspects of cosmic
evolution. A promising new avenue to distinguish our disformally coupled dark energy
models from ΛCDM lies in the detection of gravitational wave (GW) memory [54, 55]. This
phenomenon, a permanent spacetime displacement caused by passing gravitational waves,
is sensitive to the underlying cosmological model. In Refs. [54, 55], the authors have shown
that the integrated GW memory from high-redshift sources carries a distinctive signature,
with amplification factors potentially reaching 100 in certain cosmological backgrounds.
This places it within the detection capabilities of next-generation observatories like Cosmic
Explorer and the Einstein Telescope. Because our disformal coupling framework uniquely
modifies the expansion history and the effective dark energy equation of state—especially at
early times—it imparts a distinct signature on GW memory, offering a powerful, independent
probe to test our models against competing dark energy scenarios and potentially resolve
the current cosmological tensions.

ACKNOWLEDGMENTS

The work is supported by the SERB-Core Research Grant (Project SERB/CRG/2022/002348).

Appendix A: Models M1 and M2: Purely conformal DE-DM interactions

In this appendix, we examine purely conformal interactions between DE and DM, specif-
ically Models M1 and M2 from Table I . The cosmological background evolution for these
models is shown in Fig. 8 , while the linear matter and CMB temperature power spectra are
shown in Figs. 9, 10 . For numerical analysis, we assume an exponential dependence of the
couplings C(ϕ) and C(X) on the field ϕ and its kinetic term X, respectively.

28
FIG. 8. Modified background evolution of the DE energy density (top), its effective equation of
state (center) and the deviation of the Hubble parameter from the non-interacting case (bottom)
due to purely conformal DE-DM interactions. Left: Scalar field coupling corresponding to Model
M1, Right: Kinetic coupling corresponding to Model M2 from Tables I and II .

1. Model M1: Scalar Field-Dependent Conformal Coupling

Model M1 corresponds to a conformal scalar field-dependent coupling (C ≡ C(ϕ), D = 0),

a class of models that have been extensively studied in the literature [25–27, 39–42]. We find
that for a sufficiently large coupling strength (on the order of 10−1 ), the effective equation
of state of the DE field evolves as ωDE,eff ≈ −1/3 → 0 → −1 .

Unlike the disformal coupling models discussed in Sec. IV B , this model does not generate
interactions localized in time. In other words, the DE-DM coupling is non-trivial from the
onset of matter domination and persists until the present time. This is evident from the

29
FIG. 9. Linear matter power spectrum (left), CMB temperature spectrum (right) and relative
deviation from the uncoupled scenario due to scalar field dependent conformal DE-DM interactions
(Model M1 in Tables I and II).

evolution of the Hubble parameter relative to the uncoupled scenario, as plotted in the
bottom-left panel of Fig. 8 . These couplings generally increase the value of the Hubble
constant and enhance the growth of matter density perturbations on small scales, leading to
a larger σ8 value compared to the uncoupled case. In the CMB TT spectrum, we find a power
enhancement on large angular scales (low multipoles) and a suppression on small angular
scales (high multipoles). Notably, the deviation in the CMB TT power at the acoustic peaks
is relatively large compared to that observed in the disformal interaction models.

2. Model M2: Kinetic Energy-Dependent Conformal Coupling

Model M2 corresponds to a purely kinetic energy-dependent conformal coupling (C ≡

C(X), D = 0). The cosmological evolution in this model shares similarities with the pure-
momentum coupling Model M6, discussed in Sec. IV A .
At high redshifts, a non-zero kinetic coupling suppresses Hubble friction, leading to a
constant kinetic energy for the DE field. As the DM energy density decays and the coupling
weakens, the DE kinetic energy begins to dilute at a rate of a−6 . In this model, the DE

30
FIG. 10. Linear matter power spectrum (left), CMB temperature spectrum (right) and relative
deviation from the uncoupled scenario due to kinetic dependent conformal DE-DM interactions
(Model M2 in Tables I and II).

energy density takes the form:

ρDE = X + V − ρDM X ck1 . (A1)

For the DE field to avoid negative energy density states, the coupling strength ck1 must be
negative. Therefore, dark energy evolves like a DM fluid at early times, a free scalar field
with a dominating but decaying kinetic energy at intermediate times, and a potential-driven
cosmological constant at late times. The value of the constant ck1 determines the localized,
non-trivial impact on the Hubble parameter, as shown in the bottom-right panel of Fig. 8 .
Unlike Model M6, the presence of background energy transfer means the DM fluid does not
simply decay proportionally to a−3 .
Similar to the pure-momentum coupling models (M6 and M7), the conformal kinetic
coupling leads to a power enhancement over low multipoles and power suppression over
larger multipoles in the CMB TT spectrum. Additionally, it leads to an enhanced growth
of matter density fluctuations. However, for lower values of the coupling strength, the
linear-level energy coupling in this model results in enhanced growth compared to the pure-
momentum coupling scenario, which did not show any visible effect on the evolution of linear

31
perturbations.

Appendix B: Models M4 and M5: Field and kinetic dependent disformal couplings

In this appendix, we discuss the characteristic modifications introduced by pure disformal

couplings that depend on the scalar field (D ≡ D(ϕ)) and the kinetic energy (D ≡ D(X)),
corresponding to Models M4 and M5. For our numerical analysis, we have assumed an
exponential dependence on the scalar field ϕ for Model M4 and on the kinetic term X for
Model M5, respectively:

M4 : C = 1, D(ϕ) = d0 edf1 ϕ , (B1)

M5 : C = 1, D(X) = d0 edk1 X . (B2)

For both models, we set Xi d0 ≈ 0.02, where Xi denotes the initial kinetic energy of the DE
field. The background evolution of the energy densities, the linear matter power spectrum,
and the CMB TT spectrum for the two models are shown in Fig. 11.

1. Model M4: Scalar Field-Dependent Disformal Coupling

For Model M4, we have set df1 = 1.0 and varied d0 . As evident from the modified scalar
field equation (in Table I) and the left-panel plots in Fig. 11, the cosmological evolution is
functionally identical to that of Model M3. The key difference lies in the value of d0 required
to produce a similar evolution, which is shifted by several orders of magnitude (four orders
for df1 = 1.0) compared to Model M3 due to the exponential coupling. The resulting EDE-
like evolution of the DE scalar field and the modifications to the matter and CMB power
spectra are identical to the results obtained for M3.

2. Model M5: Kinetic Energy-Dependent Disformal Coupling

For Model M5, we have set dk1 = 15 d0 . Similar to Models M3 and M4, the kinetic
energy of the DE field remains constant at early times. The disformal kinetic coupling in
M5 directly contributes to the DE energy density, which takes the form:

ρDE = X + V + 2ρDM X 2 dk1 d0 edk1 X . (B3)

32
FIG. 11. Evolution of the energy densities of different components (top), linear matter power
spectrum (center) and CMB temperature spectrum (bottom) in the field-dependent (Model M4)
and kinetic energy-dependent (Model M5) disformal DE-DM interactions from Tables I and II .

33
To avoid negative energy density states for the DE field, we assume d0 > 0 and dk1 > 0 .
As in the pure-momentum coupling Model M6 and the conformal Model M2, the DE field
evolves in three distinct stages: like a dark matter fluid at early times, a free-scalar field at
intermediate times, and a potential-driven cosmological constant at late times. The linear
matter and CMB power spectra are illustrated in the right-panel plots of Fig. 11 .
Similar to Models M3 and M4, we find a suppressed structure growth for smaller values
of d0 and an enhanced growth on small scales for larger values of d0 . However, we observe
that the CMB spectrum in M5 is slightly enhanced compared to Models M3 and M4.

Appendix C: Comparing the numerical accuracy of our code with CLASS

FIG. 12. Deviation in our constructed linear matter power spectrum and the CMB temperature
power spectrum for the ΛCDM model with the results computed using the Boltzmann code CLASS.

In this appendix, we provide an overview of the custom Python code developed for the
numerical analysis of our disformal coupling framework. We also compare its accuracy for
the ΛCDM model with the results from the well-established Boltzmann code, CLASS [43].
For the background evolution, our code solves the coupled Friedmann equations and the
equations of motion for the dark matter fluid and the dark energy scalar field using for-
ward shooting. The dark energy scalar field’s potential normalization, V0 , is dynamically
calculated to match the background parameters from the Planck 2018 constraints [50]. For
recombination history, we solve the Saha and Peebles equations, using a corrected three-level
atom approach to improve accuracy. We acknowledge that this method does not reach the
high precision of state-of-the-art recombination codes like HyRec [56] or CosmoRec [57],
which account for a detailed multi-level atom and radiative-transfer effects. We model hy-
drogen reionization as well as first and second helium reionization using the tanh model

34
[58]. For linear perturbations, our code works in the conformal Newtonian gauge, employ-
ing the tight-coupling approximation for photons and baryons at early times and the full
Boltzmann system thereafter. The Boltzmann hierarchy for temperature, polarization and
neutrino multipoles is stopped at lmax = 46 and we have used the cut-off method discussed
in [59] to prevent any unphysical reflection of power from lmax to lower multipoles. For
computing the CMB spectra, we have used the line-of-sight integration approach.
For our comparison, we calibrated the cosmological parameters to the Planck 2018 con-
straints, specifically using the following values:

h = 0.6732117, ωb = 0.0223828, ωcdm = 0.1201075, Neff = 3.046,

T0,CMB = 2.7255 K, YHe = 0.2454006, τreio = 0.0543084,

ns = 0.9660499, As = 2.100549 × 10−9 . (C1)

Fig. 12 shows the deviation of our code’s linear matter power spectrum and CMB temper-
ature power spectrum for the ΛCDM model from the results obtained using CLASS. Our
matter power spectrum agrees with CLASS within 0.1%, the σ8 value agrees within 0.01%,
and the CMB TT spectrum agrees within 1%. The primary deviation is located in the
small-scale regime around the acoustic peaks.
While the Planck collaboration’s analysis requires a higher level of theoretical precision,
our code’s accuracy is sufficient for the present work. We have focused on analyzing the
relative modifications to the power spectra introduced by our general disformal coupling
framework, rather than on providing absolute predictions. For future directions, including
MCMC analysis for parameter estimation and compatibility tests with datasets, we plan to
utilize publicly available, state-of-the-art Boltzmann solvers.

[1] T. Padmanabhan, Theoretical Astrophysics: Volume 3, Galaxies and Cosmology, Theoretical

Astrophysics (Cambridge University Press, 2000).
[2] V. Mukhanov, Physical Foundations of Cosmology (Cambridge University Press, 2005).
[3] S. Weinberg, Cosmology, Cosmology (OUP Oxford, 2008).
[4] P. J. E. Peebles, Rev. Mod. Phys. 92, 030501 (2020).
[5] L. Perivolaropoulos and F. Skara, New Astron. Rev. 95, 101659 (2022), arXiv:2105.05208
[astro-ph.CO].

35
 [6] D. Scott, Proc. Int. Sch. Phys. Fermi 200, 133 (2020), arXiv:1804.01318 [astro-ph.CO].
[7] P. J. E. Peebles, Phil. Trans. Roy. Soc. Lond. A 383, 20240021 (2025).
[8] A. G. Riess et al., Astrophys. J. 826, 56 (2016), arXiv:1604.01424 [astro-ph.CO].
[9] E. Di Valentino, O. Mena, S. Pan, L. Visinelli, W. Yang, A. Melchiorri, D. F. Mota, A. G.
Riess, and J. Silk, Class. Quant. Grav. 38, 153001 (2021), arXiv:2103.01183 [astro-ph.CO].
[10] N. Schöneberg, G. Franco Abellán, A. Pérez Sánchez, S. J. Witte, V. Poulin, and J. Lesgour-
gues, Phys. Rept. 984, 1 (2022), arXiv:2107.10291 [astro-ph.CO].
[11] M. Kamionkowski and A. G. Riess, Ann. Rev. Nucl. Part. Sci. 73, 153 (2023),
arXiv:2211.04492.
[12] E. Abdalla et al., JHEAp 34, 49 (2022), arXiv:2203.06142 [astro-ph.CO].
[13] A. H. Wright et al., (2025), arXiv:2503.19441 [astro-ph.CO].
[14] G. Gu et al. (DESI), (2025), arXiv:2504.06118 [astro-ph.CO].
[15] I. Prigogine, J. Geheniau, E. Gunzig, and P. Nardone, Gen. Rel. Grav. 21, 767 (1989).
[16] E. Di Valentino et al. (CosmoVerse Network), Phys. Dark Univ. 49, 101965 (2025),
arXiv:2504.01669 [astro-ph.CO].
[17] T. Padmanabhan, Phys. Rev. D 66, 021301 (2002), arXiv:hep-th/0204150.
[18] E. J. Copeland, M. Sami, and S. Tsujikawa, Int. J. Mod. Phys. D 15, 1753 (2006), arXiv:hep-
th/0603057.
[19] B. Wang, E. Abdalla, F. Atrio-Barandela, and D. Pavon, Rept. Prog. Phys. 79, 096901 (2016),
arXiv:1603.08299 [astro-ph.CO].
[20] B. Wang, E. Abdalla, F. Atrio-Barandela, and D. Pavón, Rept. Prog. Phys. 87, 036901
(2024), arXiv:2402.00819 [astro-ph.CO].
[21] W. Yang, H. Li, Y. Wu, and J. Lu, JCAP 10, 007 (2016), arXiv:1608.07039 [astro-ph.CO].
[22] R. J. F. Marcondes, R. C. G. Landim, A. A. Costa, B. Wang, and E. Abdalla, JCAP 12, 009
(2016), arXiv:1605.05264 [astro-ph.CO].
[23] L. Feng and X. Zhang, JCAP 08, 072 (2016), arXiv:1607.05567 [astro-ph.CO].
[24] A. A. Costa, X.-D. Xu, B. Wang, and E. Abdalla, JCAP 01, 028 (2017), arXiv:1605.04138
[astro-ph.CO].
[25] J. P. Johnson and S. Shankaranarayanan, Phys. Rev. D 103, 023510 (2021), arXiv:2006.04618
[gr-qc].
[26] J. P. Johnson, A. Sangwan, and S. Shankaranarayanan, JCAP 01, 024 (2022),

36
 arXiv:2102.12367 [astro-ph.CO].
[27] P. Bansal, J. P. Johnson, and S. Shankaranarayanan, Phys. Rev. D 111, 024071 (2025).
[28] T. Karwal and M. Kamionkowski, Phys. Rev. D 94, 103523 (2016), arXiv:1608.01309 [astro-
ph.CO].
[29] V. Poulin, T. L. Smith, T. Karwal, and M. Kamionkowski, Phys. Rev. Lett. 122, 221301
(2019), arXiv:1811.04083 [astro-ph.CO].
[30] S. Alexander and E. McDonough, Phys. Lett. B 797, 134830 (2019), arXiv:1904.08912 [astro-
ph.CO].
[31] T. L. Smith, V. Poulin, and M. A. Amin, Phys. Rev. D 101, 063523 (2020), arXiv:1908.06995
[astro-ph.CO].
[32] P. Agrawal, F.-Y. Cyr-Racine, D. Pinner, and L. Randall, Phys. Dark Univ. 42, 101347
(2023), arXiv:1904.01016 [astro-ph.CO].
[33] F. Niedermann and M. S. Sloth, Phys. Rev. D 103, L041303 (2021), arXiv:1910.10739 [astro-
ph.CO].
[34] J. Sakstein and M. Trodden, Phys. Rev. Lett. 124, 161301 (2020), arXiv:1911.11760 [astro-
ph.CO].
[35] T. Karwal, M. Raveri, B. Jain, J. Khoury, and M. Trodden, Phys. Rev. D 105, 063535 (2022),
arXiv:2106.13290 [astro-ph.CO].
[36] E. McDonough, M.-X. Lin, J. C. Hill, W. Hu, and S. Zhou, Phys. Rev. D 106, 043525 (2022),
arXiv:2112.09128 [astro-ph.CO].
[37] M.-X. Lin, E. McDonough, J. C. Hill, and W. Hu, Phys. Rev. D 107, 103523 (2023),
arXiv:2212.08098 [astro-ph.CO].
[38] V. Poulin, T. L. Smith, and T. Karwal, Phys. Dark Univ. 42, 101348 (2023), arXiv:2302.09032
[astro-ph.CO].
[39] L. Amendola, Phys. Rev. D 62, 043511 (2000), arXiv:astro-ph/9908023.
[40] C. G. Boehmer, G. Caldera-Cabral, R. Lazkoz, and R. Maartens, Phys. Rev. D 78, 023505
(2008), arXiv:0801.1565 [gr-qc].
[41] J. Beyer, S. Nurmi, and C. Wetterich, Phys. Rev. D 84, 023010 (2011), arXiv:1012.1175
[astro-ph.CO].
[42] M. Carrillo González and M. Trodden, Phys. Rev. D 97, 043508 (2018), [Erratum: Phys.Rev.D
101, 089901 (2020)], arXiv:1705.04737 [astro-ph.CO].

37
[43] J. Lesgourgues, (2011), arXiv:1104.2932 [astro-ph.IM].
[44] J. D. Bekenstein, Phys. Rev. D 48, 3641 (1993), arXiv:gr-qc/9211017.
[45] C. Armendariz-Picon, V. F. Mukhanov, and P. J. Steinhardt, Phys. Rev. Lett. 85, 4438
(2000), arXiv:astro-ph/0004134.
[46] C. Armendariz-Picon, V. F. Mukhanov, and P. J. Steinhardt, Phys. Rev. D 63, 103510 (2001),
arXiv:astro-ph/0006373.
[47] E. Calabrese, M. Migliaccio, L. Pagano, G. De Troia, A. Melchiorri, and P. Natoli, Phys.
Rev. D 80, 063539 (2009).
[48] M. Kopp, C. Skordis, D. B. Thomas, and S. Ilić, Phys. Rev. Lett. 120, 221102 (2018),
arXiv:1802.09541 [astro-ph.CO].
[49] V. F. Mukhanov, H. A. Feldman, and R. H. Brandenberger, Phys. Rept. 215, 203 (1992).
[50] N. Aghanim et al. (Planck), Astron. Astrophys. 641, A6 (2020), [Erratum: Astron.Astrophys.
652, C4 (2021)], arXiv:1807.06209 [astro-ph.CO].
[51] C. R. Contaldi, M. Peloso, L. Kofman, and A. D. Linde, JCAP 07, 002 (2003), arXiv:astro-
ph/0303636.
[52] S. Shankaranarayanan and L. Sriramkumar, Phys. Rev. D 70, 123520 (2004), arXiv:hep-
th/0403236.
[53] J. White, Y.-l. Zhang, and M. Sasaki, Phys. Rev. D 90, 083517 (2014), arXiv:1407.5816
[astro-ph.CO].
[54] I. Chakraborty, S. Jana, and S. Shankaranarayanan, Phys. Rev. D 111, 083548 (2025),
arXiv:2402.18083 [gr-qc].
[55] I. Chakraborty, S. Jana, and S. Shankaranarayanan, Phys. Rev. D 112, L021503 (2025),
arXiv:2505.09096 [gr-qc].
[56] Y. Ali-Haı̈moud and C. M. Hirata, Phys. Rev. D 83, 043513 (2011).
[57] J. Chluba and R. M. Thomas, Mon. Not. Roy. Astron. Soc. 412, 748 (2011), arXiv:1010.3631
[astro-ph.CO].
[58] A. Lewis, Phys. Rev. D 78, 023002 (2008), arXiv:0804.3865 [astro-ph].
[59] C.-P. Ma and E. Bertschinger, Astrophys. J. 455, 7 (1995), arXiv:astro-ph/9506072.

38