betaFeedback

WorksPhysics and Astronomy

Thermal damping of spectator scalar dark matter: Relaxing the overclosure and isocurvature bounds

Abstract

A light scalar present during inflation generically acquires a misalignment and can later constitute the dark matter. Since it fluctuates independently of the inflaton, however, it carries isocurvature perturbations. Isocurvature and overclosure bounds together rule out the simplest version of this scenario at high inflationary scales. We show that a weak coupling to the primordial plasma can relax both bounds simultaneously. The plasma depletes the scalar abundance through thermal damping, a dissipative effect arising from the plasma's finite equilibration time, thereby relaxing the overclosure bound. If the damping grows with the oscillation amplitude, it can also drive different initial amplitudes onto a common attractor that depends only on the local radiation temperature. This synchronizes the scalar with the adiabatic clock, thus suppressing its isocurvature. The resulting isocurvature bound on the inflationary scale relaxes by exactly the factor by which damping reduces the abundance, in addition to the relaxation from the larger initial misalignment now allowed. This result holds for any damping rate quadratic in the amplitude, regardless of its temperature dependence.

Introduction

Scalar fields are common in extensions of the Standard Model (SM) [1, 2, 3]. A scalar lighter than the inflationary Hubble scale HinfH_{\mathrm{inf}} undergoes de Sitter fluctuations with amplitude δχ∼Hinf/2π\delta\chi\sim H_{\mathrm{inf}}/2\pi and therefore generically acquires a misalignment XiX_i by the end of inflation [4, 5, 6, 7, 8]. At late times, the resulting scalar condensate may constitute all or part of the dark matter (DM), with an abundance set by XiX_i [5, 9, 10]. Since a spectator scalar fluctuates independently of the inflaton, its density perturbations contain an isocurvature component of relative amplitude δρχ/ρχ∼(Hinf/2π)/Xi\delta\rho_\chi/\rho_\chi\sim(H_{\mathrm{inf}}/2\pi)/X_i, which is subject to stringent cosmic microwave background (CMB) limits [11].

This leads to a tension. Avoiding overclosure favors a smaller initial misalignment XiX_i, whereas suppressing the isocurvature favors a larger one [6, 12, 13]. Together, they bound the scalar mass and HinfH_{\mathrm{inf}}, strongly restricting the viable parameter space [10]. These conclusions, however, rest on the assumption that the spectator scalar evolves in isolation from the primordial plasma after inflation.

If the scalar couples, even weakly, to particles populated by reheating, its post-inflation evolution can change dramatically. The plasma modifies the scalar dynamics in two qualitatively distinct ways. First, equilibrium thermal corrections to the effective potential, such as a thermal mass, change the onset of scalar oscillations and how the condensate subsequently redshifts [14, 15, 16]. Second, the plasma responds to the evolving scalar with a finite relaxation time. The resulting lag behind instantaneous equilibrium generates a dissipative response that drains energy from the scalar field and heats the plasma [17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29]. This thermal damping is not captured by the equilibrium effective potential and can substantially alter both the abundance and the perturbations of the scalar.

In this work we study how these thermal effects reshape the background and perturbation evolutions of a spectator scalar. We illustrate the mechanism in a minimal model in which a real scalar χ\chi controls the mass of a fermion ψ\psi that is in thermal contact with the SM plasma. The thermal population of ψ\psi generates both a thermal effective potential and thermal damping for χ\chi. Although the detailed outcomes are model-dependent, the mechanism applies broadly whenever an oscillating scalar modulates the properties of a species in the thermal bath.

At the background level, thermal damping depletes the scalar condensate and thereby relaxes the usual overclosure bound. Moreover, under suitable conditions, thermal damping can drive the scalar amplitude toward a dynamical attractor on which the thermal-damping rate is comparable to the Hubble rate [18, 19]. A range of initial misalignments then converges onto a common evolution, which the field tracks until it eventually decouples. The relic abundance is consequently set by where the attractor tracking ends rather than by the initial spectator misalignment XiX_i. In some sense, the abundance-setting mechanism resembles thermal freeze-out, but for a coherent condensate that is never thermalized.

At the perturbation level, the same attractor suppresses pre-existing isocurvature. While the scalar is on the attractor, its amplitude is dictated by the local plasma temperature and no longer provides an independent clock. A perturbation along the attractor is @page 1/12 starts_mid=0 ends_mid=1

Introduction

Scalar fields are common in extensions of the Standard Model (SM) [1, 2, 3]. A scalar lighter than the inflationary Hubble scale HinfH_{\mathrm{inf}} undergoes de Sitter fluctuations with amplitude δχ∼Hinf/2π\delta\chi\sim H_{\mathrm{inf}}/2\pi and therefore generically acquires a misalignment XiX_i by the end of inflation [4, 5, 6, 7, 8]. At late times, the resulting scalar condensate may constitute all or part of the dark matter (DM), with an abundance set by XiX_i [5, 9, 10]. Since a spectator scalar fluctuates independently of the inflaton, its density perturbations contain an isocurvature component of relative amplitude δρχ/ρχ∼(Hinf/2π)/Xi\delta\rho_\chi/\rho_\chi\sim(H_{\mathrm{inf}}/2\pi)/X_i, which is subject to stringent cosmic microwave background (CMB) limits [11].

This leads to a tension. Avoiding overclosure favors a smaller initial misalignment XiX_i, whereas suppressing the isocurvature favors a larger one [6, 12, 13]. Together, they bound the scalar mass and HinfH_{\mathrm{inf}}, strongly restricting the viable parameter space [10]. These conclusions, however, rest on the assumption that the spectator scalar evolves in isolation from the primordial plasma after inflation.

If the scalar couples, even weakly, to particles populated by reheating, its post-inflation evolution can change dramatically. The plasma modifies the scalar dynamics in two qualitatively distinct ways. First, equilibrium thermal corrections to the effective potential, such as a thermal mass, change the onset of scalar oscillations and how the condensate subsequently redshifts [14, 15, 16]. Second, the plasma responds to the evolving scalar with a finite relaxation time. The resulting lag behind instantaneous equilibrium generates a dissipative response that drains energy from the scalar field and heats the plasma [17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29]. This thermal damping is not captured by the equilibrium effective potential and can substantially alter both the abundance and the perturbations of the scalar.

In this work we study how these thermal effects reshape the background and perturbation evolutions of a spectator scalar. We illustrate the mechanism in a minimal model in which a real scalar χ\chi controls the mass of a fermion ψ\psi that is in thermal contact with the SM plasma. The thermal population of ψ\psi generates both a thermal effective potential and thermal damping for χ\chi. Although the detailed outcomes are model-dependent, the mechanism applies broadly whenever an oscillating scalar modulates the properties of a species in the thermal bath.

At the background level, thermal damping depletes the scalar condensate and thereby relaxes the usual overclosure bound. Moreover, under suitable conditions, thermal damping can drive the scalar amplitude toward a dynamical attractor on which the thermal-damping rate is comparable to the Hubble rate [18, 19]. A range of initial misalignments then converges onto a common evolution, which the field tracks until it eventually decouples. The relic abundance is consequently set by where the attractor tracking ends rather than by the initial spectator misalignment XiX_i. In some sense, the abundance-setting mechanism resembles thermal freeze-out, but for a coherent condensate that is never thermalized.

At the perturbation level, the same attractor suppresses pre-existing isocurvature. While the scalar is on the attractor, its amplitude is dictated by the local plasma temperature and no longer provides an independent clock. A perturbation along the attractor is therefore adiabatic, i.e., equivalent to a local time shift [30, 31]. Isocurvature perturbations, which correspond to displacements away from the attractor, are thermally damped during the attractor tracking. A small residual isocurvature may survive if the tracking is too brief to erase the primordial spectator mode completely.

Thermal damping can thus relax the overclosure and isocurvature constraints simultaneously, removing the usual tension between them. Furthermore, once the attractor is reached, the relic abundance becomes largely insensitive to the unknown inflationary initial conditions and is instead controlled by the low-energy interactions and late thermal history, which are more accessible to experiment.

This paper is organized as follows. In Section II we introduce the model, review the thermal effects, discuss the thermal-damping attractor, and define the scope of the effective description. In Section III we study the evolution of superhorizon perturbations and the suppression of primordial isocurvature. In Section IV we show how thermal damping relaxes both the overclosure and isocurvature bounds. We conclude in § 3. The appendices contain the derivation of the perturbation equations (Appendix A), the relation between the isocurvature mode used in the main text and the scalar-radiation relative entropy (Appendix B), a discussion of subleading isocurvature sources (Appendix C), and model realizations of various thermalization rates (Appendix § 2.3.4).

Spectator scalar dark matter coupled to a thermal bath

We consider a real spectator scalar χ\chi coupled to a Dirac fermion ψ\psi that exchanges energy with the SM plasma. A minimal effective description is

L⊃−Vvac(χ)−(mψ−gχ)ψˉψ+Lψ↔SM,(1)\mathcal{L} \supset-V_{\mathrm{vac}}(\chi) - (m_\psi-g\chi)\bar{\psi}\psi+\mathcal{L}_{\psi\leftrightarrow\mathrm{SM}}, \tag*{(1)}

where VvacV_{\mathrm{vac}} is the vacuum potential of χ\chi and Lψ↔SM\mathcal{L}_{\psi\leftrightarrow\mathrm{SM}} contains interactions that thermally equilibrate ψ\psi with the SM. The scalar χ\chi modulates the effective fermion ψ\psi mass,

Mψ(χ)=∣mψ−gχ∣.(2)M_\psi(\chi)=\lvert m_\psi-g\chi\rvert. \tag*{(2)}

We focus on a post-reheating epoch during which the fermion ψ\psi is relativistic and remains efficiently thermalized with the SM bath at temperature TT, with an effective thermalization rate

Γth=κTn,(3)\Gamma_{\mathrm{th}}=\kappa T^n, \tag*{(3)}

where the coefficient κ\kappa has mass dimension 1−n1-n and the exponent nn depends on the operators in Lψ↔SM\mathcal{L}_{\psi\leftrightarrow\mathrm{SM}}; we discuss representative portals in Appendix § 2.3.4. The rate Γth\Gamma_{\mathrm{th}} describes how quickly the ψ\psi distribution relaxes toward its instantaneous equilibrium as Mψ(χ)M_\psi(\chi) evolves. We assume that Lψ↔SM\mathcal{L}_{\psi\leftrightarrow\mathrm{SM}} does not have further relevant effects beyond setting Γth\Gamma_{\mathrm{th}}.

Thermal Potential and Damping

The equation of motion of the scalar field χ\chi is

□χ+Vvac′(χ)=g⟨ψˉψ⟩.(4)\Box\chi+V'_{\mathrm{vac}}(\chi)=g\langle\bar{\psi}\psi\rangle. \tag*{(4)}

Here and throughout this paper we use the mostly minus metric signature, (+−−−)(+---). If the fermion ψ\psi adjusted instantaneously to the scalar, its response ⟨ψˉψ⟩\langle\bar{\psi}\psi\rangle would depend only on the instantaneous values of χ\chi and TT. Its effect could then be absorbed into the effective potential of χ\chi. In reality, the ψ\psi distribution relaxes toward equilibrium at the finite rate Γth\Gamma_{\mathrm{th}}. Its response lags slightly behind the evolving χ\chi and has a component proportional to χ˙\dot{\chi}, which gives rise to thermal damping.

When ψ\psi remains close to thermal equilibrium, the right-hand side of Eq. (4) can be evaluated as [18, 19]

g⟨ψˉψ⟩=mχ,th2(mψg−χ)−Γ~χ˙,mχ,th=cmgT.(5)g\langle\bar{\psi}\psi\rangle=m_{\chi,\mathrm{th}}^2\left(\frac{m_\psi}{g}-\chi\right)-\widetilde{\Gamma}\dot{\chi},\qquad m_{\chi,\mathrm{th}}=c_m gT . \tag*{(5)}

The first term is the equilibrium response, characterized by the thermal mass mχ,thm_{\chi,\mathrm{th}}, where cm=1/6c_m=1/\sqrt{6} for a Dirac fermion ψ\psi. The second term is the non-equilibrium lag giving rise to thermal damping with coefficient Γ~\widetilde{\Gamma}. For the homogeneous background χˉ\bar{\chi}, the equation of motion reduces to

χˉ¨+(3H+Γ~)χˉ˙+Mχ2(χˉ−χth)=0,(6)\ddot{\bar{\chi}}+(3H+\widetilde{\Gamma})\dot{\bar{\chi}}+M_\chi^2(\bar{\chi}-\chi_{\mathrm{th}})=0, \tag*{(6)}

with Mχ2=Vvac′′+mχ,th2M_\chi^2=V_{\mathrm{vac}}''+m_{\chi,\mathrm{th}}^2 the effective mass of χ\chi, χth=(mχ,th2/Mχ2)mψ/g\chi_{\mathrm{th}}=(m_{\chi,\mathrm{th}}^2/M_\chi^2)m_\psi/g the ψ\psi-induced shift of the effective potential minimum, and Γ~\widetilde{\Gamma} the thermal damping coefficient evaluated on the background χˉ\bar{\chi}.

We assume χ\chi has a large amplitude ≫mψ/g\gg m_\psi/g (which also implies ≫χth\gg\chi_{\mathrm{th}}) and oscillates rapidly compared to thermal and Hubble damping (Mχ≫1,HM_\chi\gg1,H).1 The background χˉ\bar{\chi} is then approximately sinusoidal

χˉ=X(t)cos⁡θ(t),θ˙(t)≈Mχ(t),(7)\bar{\chi}=X(t)\cos\theta(t),\qquad\dot{\theta}(t)\approx M_\chi(t), \tag*{(7)}

with a slowly varying envelope XX, ∣X˙∣/X≪Mχ\lvert\dot{X}\rvert/X\ll M_\chi, that evolves as [18, 19]

X˙X=−12(3H+Γ~+M˙χMχ),Γ~≡⟨Γ~χ˙2⟩⟨χ˙2⟩.(8)\frac{\dot{X}}{X}=-\frac{1}{2}\left(3H+\widetilde{\Gamma}+\frac{\dot{M}_\chi}{M_\chi}\right),\qquad\widetilde{\Gamma}\equiv\frac{\langle\widetilde{\Gamma}\dot{\chi}^{2}\rangle}{\langle\dot{\chi}^{2}\rangle}. \tag*{(8)}

Here ⟨⋯ ⟩\langle\cdots\rangle denotes an average over an oscillation period, and Γ~\widetilde{\Gamma} is the χ˙2\dot{\chi}^{2}-weighted oscillation-averaged damping rate. In this regime, the Γ~\widetilde{\Gamma} takes the form [18, 19]

Γ~=cγg4X2ΓthΓth2Γth2+4Mχ2.(9)\widetilde{\Gamma}=c_\gamma\frac{g^4X^2}{\Gamma_{\mathrm{th}}}\frac{\Gamma_{\mathrm{th}}^2}{\Gamma_{\mathrm{th}}^2+4M_\chi^2}. \tag*{(9)}

with cΥ=3cm2/2π2c_{\Upsilon}=3c_m^2/2\pi^2, a dimensionless coefficient determined by the number of degrees of freedom of ψ\psi.

In what follows we focus, for simplicity, on an epoch during radiation domination, with approximately constant relativistic degrees of freedom, in which the potential of χ\chi is dominated by its thermal mass:

T˙T≈−H,Mχ≈mχ,th=cmgT.(10)\frac{\dot{T}}{T}\approx-H,\qquad M_{\chi}\approx m_{\chi,\mathrm{th}}=c_m gT. \tag*{(10)}

Then we can use the temperature TT as a clock and rewrite the envelope equation, Eq. (8), as

ξ=ξfree+Υ~2H,(11)\xi=\xi_{\mathrm{free}}+\frac{\tilde{\Upsilon}}{2H}, \tag*{(11)}

where we defined the logarithmic slope of the envelope

ξ≡dln⁡Xdln⁡T,ξfree≡ξ∣Υ~=0=1.(12)\xi\equiv\frac{d\ln X}{d\ln T},\qquad\xi_{\mathrm{free}}\equiv\left.\xi\right|_{\tilde{\Upsilon}=0}=1. \tag*{(12)}

Without damping, the amplitude scales as X∝Tξfree=TX\propto T^{\xi_{\mathrm{free}}}=T under thermal mass. Thermal damping increases the exponent by Υ~/2H\tilde{\Upsilon}/2H.

Thermal-Damping Attractor

Since thermal damping weakens as the amplitude decreases, Υ~∝X2\tilde{\Upsilon}\propto X^2, it is self-limiting. When XX is sufficiently large that Υ~≫H\tilde{\Upsilon}\gg H, XX decreases rapidly, and Υ~\tilde{\Upsilon} drops to ∼H\sim H in a fraction of a Hubble time [18, 19]. Whether it then remains at Υ~∼H\tilde{\Upsilon}\sim H or continues to drop toward Υ~≪H\tilde{\Upsilon}\ll H depends on the temperature dependence of Υ~\tilde{\Upsilon} at fixed amplitude, which is controlled by Γth(T)\Gamma_{\mathrm{th}}(T) and Mχ(T)M_{\chi}(T).

To see which occurs, consider the amplitude XΥ(T)X_{\Upsilon}(T) at which Υ~=ctrackH\tilde{\Upsilon}=c_{\mathrm{track}}H for some constant ctrackc_{\mathrm{track}}. From Eq. (9),

XΥ(T)=ctrackHΓthg2cΥΓth2+4Mχ2Γth2.(13)X_{\Upsilon}(T)=\frac{\sqrt{c_{\mathrm{track}}H\Gamma_{\mathrm{th}}}}{g^2\sqrt{c_{\Upsilon}}}\sqrt{\frac{\Gamma_{\mathrm{th}}^2+4M_{\chi}^2}{\Gamma_{\mathrm{th}}^2}}. \tag*{(13)}

Differentiating with respect to TT at fixed ctrackc_{\mathrm{track}} and using H∝T2H\propto T^2 gives its slope

ξΥ≡dln⁡XΥdln⁡T=1−12∂ln⁡Υ~∂ln⁡T∣X.(14)\xi_{\Upsilon}\equiv\frac{d\ln X_{\Upsilon}}{d\ln T}=1-\frac{1}{2}\left.\frac{\partial\ln\tilde{\Upsilon}}{\partial\ln T}\right|_X. \tag*{(14)}

Requiring X=XΥX=X_{\Upsilon} to solve the envelope equation, Eq. (11), fixes ctrack=2(ξΥ−ξfree)c_{\mathrm{track}}=2(\xi_{\Upsilon}-\xi_{\mathrm{free}}), and so

ctrack=−∂ln⁡Υ~∂ln⁡T∣X=n−8(n−1)Mχ2Γth2+4Mχ2.(15)c_{\mathrm{track}}=-\left.\frac{\partial\ln\tilde{\Upsilon}}{\partial\ln T}\right|_X=n-\frac{8(n-1)M_{\chi}^2}{\Gamma_{\mathrm{th}}^2+4M_{\chi}^2}. \tag*{(15)}

The amplitude XΥ∝T1+ctrack/2X_{\Upsilon}\propto T^{1+c_{\mathrm{track}}/2} acts as a dynamical attractor if the undamped amplitude X∝TX\propto T decreases more slowly than XΥX_{\Upsilon} as TT goes down, i.e., if

ctrack>0(XΥ is an attractor).(16)c_{\mathrm{track}}>0\qquad(X_{\Upsilon}\text{ is an attractor}). \tag*{(16)}

An amplitude above XΥX_{\Upsilon} is damped faster than the Hubble rate and is driven down to it, while an amplitude below XΥX_{\Upsilon} catches up with it, so XΥX_{\Upsilon} is an attractor from both sides.2 In the fast thermalization (Γth≫Mχ\Gamma_{\mathrm{th}}\gg M_{\chi}) and slow thermalization (Γth≪Mχ\Gamma_{\mathrm{th}}\ll M_{\chi}) regimes, ctrackc_{\mathrm{track}} simplifies to

ctrack={n,Γth≫Mχ,2−n,Γth≪Mχ.(17)c_{\mathrm{track}}= \begin{cases} n, & \Gamma_{\mathrm{th}}\gg M_{\chi},\\ 2-n, & \Gamma_{\mathrm{th}}\ll M_{\chi}. \end{cases} \tag*{(17)}

So XΥX_{\Upsilon} is an attractor for n>0n>0 in the fast-thermalization regime and for n<2n<2 in the slow-thermalization regime.

Our description applies after reheating, once ψ\psi is thermalized and χ\chi has started to oscillate. The approach to the attractor depends on where the amplitude XX starts in this regime, which is set by inflation and reheating. If XX starts above XΥX_{\Upsilon}, it drops to XΥX_{\Upsilon} within a fraction of a Hubble time. If XX starts below XΥX_{\Upsilon}, it redshifts freely as X∝TX\propto T, while XΥ∝T1+ctrack/2X_{\Upsilon}\propto T^{1+c_{\mathrm{track}}/2}. The ratio X/XΥX/X_{\Upsilon} therefore grows by ectrack/2e^{c_{\mathrm{track}}/2} per e-fold, and the two meet after Napproach≃(2/ctrack)ln⁡(XΥ/X)∣TstartN_{\mathrm{approach}}\simeq(2/c_{\mathrm{track}})\ln(X_{\Upsilon}/X)|_{T_{\mathrm{start}}} e-folds, where TstartT_{\mathrm{start}} is the temperature at which our description first applies. We denote the temperatures at the start and end of tracking by TinT_{\mathrm{in}} and ToutT_{\mathrm{out}}, and the number of e-folds of tracking by

Ntrack≡ln⁡TinTout.(18)N_{\mathrm{track}}\equiv\ln\frac{T_{\mathrm{in}}}{T_{\mathrm{out}}}. \tag*{(18)}

As stated earlier, we work in the regime where the thermal mass dominates, the amplitude is large X≫mψ/gX\gg m_{\psi}/g, and the exponent nn is fixed. Continued tracking requires ctrack>0c_{\mathrm{track}}>0 and that ψ\psi remains relativistic and thermalized, T≳max⁡(mψ,gX)T\gtrsim\max(m_{\psi},gX) and Γth≳H\Gamma_{\mathrm{th}}\gtrsim H. Tracking ends when one of these conditions ceases to hold. Along the attractor, the relativistic condition T≳max⁡(mψ,gX)T\gtrsim\max(m_{\psi},gX) is less strict compared to X≫mψ/gX\gg m_{\psi}/g. The thermalization requirement Γth≳H\Gamma_{\mathrm{th}}\gtrsim H can be violated with time only for the n>2n>2 case where Γth/H∝Tn−2\Gamma_{\mathrm{th}}/H\propto T^{n-2} decreases. In that case, however, ctrackc_{\mathrm{track}} changes sign first, because at that point Γth∼Mχ≫H\Gamma_{\mathrm{th}}\sim M_{\chi}\gg H. The tracking therefore ends on its own only through ctrackc_{\mathrm{track}} changing sign. Setting Eq. (15) to zero gives

Tout=(2cmgκ1−2n)1n−1,for n>2.(19)T_{\mathrm{out}}=\left(\frac{2c_m g}{\kappa}\sqrt{1-\frac{2}{n}}\right)^{\frac{1}{n-1}},\qquad\text{for }n>2. \tag*{(19)}

This endpoint is relevant only for n>2n>2, because only then does ctrackc_{\mathrm{track}} change sign as Γth/Mχ\Gamma_{\mathrm{th}}/M_{\chi} evolves. For 0<n<20<n<2, ctrackc_{\mathrm{track}} never changes sign on its own. When n=2n=2, Γth/Mχ\Gamma_{\mathrm{th}}/M_{\chi} decreases as the temperature decreases, and ctrack=2Γth2/(Γth2+4Mχ2)c_{\mathrm{track}}=2\Gamma_{\mathrm{th}}^2/(\Gamma_{\mathrm{th}}^2+4M_{\chi}^2) decreases with it. but stays positive. For n=1n = 1, Γth/Mχ\Gamma_{\mathrm{th}}/M_{\chi} is constant and ctrack=1c_{\mathrm{track}} = 1. Tracking thus does not end on its own for 0<n≤20 < n \le2, and it can persist for many ee-folds. It can instead end at a temperature T∗T_{*} where nn changes in a way such that ctrack<0c_{\mathrm{track}} < 0. We discuss an example of this in Section IV.

Our analysis must be extended if one of its assumptions fails before tracking ends, namely if the amplitude drops to X∼mψ/gX \sim m_{\psi}/g, if the thermal mass falls below the bare mass, or if ψ\psi becomes non-relativistic. We restrict our analysis to the parameter space where tracking ends first.

The Scope of This Work

A complete theory would specify the origin of the scalar background and its perturbations, their evolution through reheating, and the later dynamics that determine the relic abundance and observable perturbations, possibly with different degrees of freedom and effective descriptions at different stages of this evolution. We instead restrict our analysis to a finite window from a few ee-folds before the onset of attractor tracking at TinT_{\mathrm{in}} to the end of tracking at ToutT_{\mathrm{out}}. Throughout this interval, we assume that the thermal effects described above dominate the dynamics of the scalar χ\chi. We do not require this minimal description, or even the same degrees of freedom, to hold during inflation or after ToutT_{\mathrm{out}}. This lets us study the thermal mechanism without committing to a single effective theory valid from inflation to the present day.

Within this window, we assume that the EFT remains valid up to a cutoff Λ\Lambda above TinT_{\mathrm{in}} and that the scalar mass is protected from ∼gΛ\sim g\Lambda mass corrections. The Coleman-Weinberg potential generated by ψ\psi, ΔVCW(χ)=−[Mψ4(χ)/16π2][ln⁡Mψ2(χ)/μ2−3/2]\Delta V_{\mathrm{CW}}(\chi) = -[M_{\psi}^{4}(\chi)/16\pi^{2}][\ln M_{\psi}^{2}(\chi)/\mu^{2} - 3/2] is negligible compared to the thermal potential as long as gX≪TgX \ll T, i.e., as long as ψ\psi is relativistic. After ToutT_{\mathrm{out}}, this potential and other vacuum contributions, such as the bare mass, may come to dominate.

We take the scalar background and its superhorizon perturbations shortly before TinT_{\mathrm{in}} as initial data, without specifying their earlier origin. As long as the initial conditions lie within the basin of attraction and tracking lasts sufficiently long, different initial conditions converge toward a common temperature-dependent evolution, and memory of the earlier evolution is erased up to a small residual. Our robust background result is that thermal damping decreases the scalar amplitude relative to the undamped evolution. As we show in the next section, convergence onto the thermal-damping attractor also suppresses isocurvature. Once suppressed, isocurvature cannot be regenerated on superhorizon scales by later evolution.

Superhorizon Perturbations

Evolution of Superhorizon Perturbations

BBN requires ψ\psi, and hence the thermal effects on χ\chi, to be absent at the latest by T∼MeVT \sim\mathrm{MeV}. Hence, modes with k≲aH∣T∼MeV∼104 Mpc−1k \lesssim aH|_{T\sim\mathrm{MeV}} \sim10^{4}\,\mathrm{Mpc}^{-1} are superhorizon during the entire phase in which thermal effects are significant.3 The equation of motion for superhorizon χ\chi perturbations δχ\delta\chi is

δχ¨+(3H+Γˉ)δχ˙+(Mχ2+dΓˉdχˉχˉ˙)δχ=S.(20)\delta\ddot{\chi} + (3H+\bar{\Gamma})\delta\dot{\chi} + \left(M_{\chi}^{2}+\frac{\mathrm{d}\bar{\Gamma}}{\mathrm{d}\bar{\chi}}\dot{\bar{\chi}}\right)\delta\chi= S . \tag*{(20)}

For X≫mψ/gX \gg m_{\psi}/g, the source is S=SΦ+Seq+SΓS = S_{\Phi} + S_{\mathrm{eq}} + S_{\Gamma} with

SΦ=4Φ˙χˉ˙−2Φ(Mχ2χˉ+12Γˉχˉ˙).(21)S_{\Phi} = 4\dot{\Phi}\dot{\bar{\chi}} - 2\Phi\left(M_{\chi}^{2}\bar{\chi}+\frac{1}{2}\bar{\Gamma}\dot{\bar{\chi}}\right) . \tag*{(21)}
Seq=−2δTTmχ,th2χˉ,(22)S_{\mathrm{eq}} = -2\frac{\delta T}{T}m_{\chi,\mathrm{th}}^{2}\bar{\chi} , \tag*{(22)}
SΓ=−δTT∂Γˉ∂ln⁡T∣χˉχˉ˙.(23)S_{\Gamma} = -\frac{\delta T}{T}\left.\frac{\partial\bar{\Gamma}}{\partial\ln T}\right|_{\bar{\chi}}\dot{\bar{\chi}} . \tag*{(23)}

SΦS_{\Phi}, SeqS_{\mathrm{eq}}, and SΓS_{\Gamma} arise, respectively, from gravitational driving by the metric perturbation (∝Φ,Φ˙\propto\Phi, \dot{\Phi}), the temperature dependence of the (equilibrium) thermal effective potential, and the temperature dependence of thermal damping. We provide a derivation of the perturbation equation in Appendix A. By the fluctuation-dissipation relation, thermal damping should in general be accompanied with a stochastic thermal noise term in the equation of motion [40]. As we discuss in Appendix C, this noise has a blue spectrum and is negligible for the superhorizon modes considered here, so we omit it from Eq. (20).

Adiabatic and Isocurvature Modes

The adiabatic mode of δχ\delta\chi corresponds to a local time shift t→t+δtt \to t+\delta t of the background solution χˉ\bar{\chi} (Eq. (7))

δχad=χˉ˙δt=X(−Mχsin⁡θ+X˙Xcos⁡θ)δt.(24)\delta\chi_{\mathrm{ad}} = \dot{\bar{\chi}}\delta t = X\left(-M_{\chi}\sin\theta+ \frac{\dot{X}}{X}\cos\theta\right)\delta t . \tag*{(24)}

The first and second terms stem from phase and amplitude perturbations, respectively. In the limit (∣X˙∣/X)/Mχ≪1(|\dot{X}|/X)/M_{\chi} \ll1, the adiabatic mode δχad\delta\chi_{\mathrm{ad}} is predominantly in the phase direction.

Any perturbation of χ\chi that is not of the form of Eq. (24) is isocurvature. For (∣X˙∣/X)/Mχ≪1(|\dot{X}|/X)/M_{\chi} \ll1, the isocurvature mode is naively just the amplitude perturbation δX\delta X. However, Eq. (24) shows that a part X˙δt=XξδT/T\dot{X}\delta t = X\xi\delta T/T of the amplitude perturbation, with ξ\xi defined in Eq. (12), is itself adiabatic. Subtracting this adiabatic contamination from the amplitude perturbation, we find that the fractional amplitude perturbation not accounted for by the local radiation clock is

I≡δXX−ξδTT,(25)I \equiv\frac{\delta X}{X} - \xi\frac{\delta T}{T}, \tag*{(25)}

which is gauge-invariant and vanishes identically for a pure time shift. The isocurvature variable II can be understood as the fractional amplitude perturbation on uniform temperature slices, I=δX/X∣δT=0I = \left.\delta X/X\right|_{\delta T=0}. To connect to observables, one can convert II into the relative isocurvature between χ\chi and radiation (see Appendix B),

Sχr=−3H(δρχρ˙χ−δρradρ˙rad)=3I1+ξ,(26)S_{\chi r} = -3H\left(\frac{\delta\rho_\chi}{\dot{\rho}_\chi} - \frac{\delta\rho_{\mathrm{rad}}}{\dot{\rho}_{\mathrm{rad}}}\right) = \frac{3I}{1+\xi}, \tag*{(26)}

when the thermal mass dominates the scalar potential. This confirms that I=0I = 0 is equivalent to vanishing relative entropy between χ\chi and radiation.

Erasure of Isocurvature

The time evolution of II follows from time averaging Eq. (20). For Υ~∝X2\widetilde{\Upsilon} \propto X^2 and Mχ≈mχ,thM_\chi\approx m_{\chi,\mathrm{th}} one obtains (see Appendix A for a derivation)

I˙+12∂Υ~∂ln⁡X∣TI=−[ξ˙+Υ~(ξ~−ξΥ)]δTT,(27)\dot{I} + \frac{1}{2}\left.\frac{\partial\widetilde{\Upsilon}}{\partial\ln X}\right|_T I = -\left[\dot{\xi} + \widetilde{\Upsilon}(\widetilde{\xi}-\xi_{\Upsilon})\right]\frac{\delta T}{T}, \tag*{(27)}

with ξΥ\xi_{\Upsilon} defined in Eq. (14). The above equation implies that II is damped at the rate (1/2)∂Υ~/∂ln⁡X∣T=Υ~(1/2)\left.\partial\widetilde{\Upsilon}/\partial\ln X\right|_T = \widetilde{\Upsilon}, for Υ~\widetilde{\Upsilon} as given in Eq. (9). Note that the amplitude dependence of the thermal damping, Υ~∝X2\widetilde{\Upsilon} \propto X^2, is crucial for the damping of II. It can be checked with Eq. (11) and Eq. (14) that the source term vanishes identically on any thermal-mass-dominated background solution,

ξ˙+Υ~(ξ~−ξΥ)=0.(28)\dot{\xi} + \widetilde{\Upsilon}(\widetilde{\xi}-\xi_{\Upsilon}) = 0. \tag*{(28)}

Thus, an adiabatic perturbation with I=0I = 0 remains adiabatic, as expected from Weinberg’s adiabatic theorem [30]. The variable II is consequently damped as

I(t)=I(ti)e−∫titdt′ Υ~(t′).(29)I(t) = I(t_i)e^{-\int_{t_i}^{t} \mathrm{d}t'\,\widetilde{\Upsilon}(t')}. \tag*{(29)}

This holds during the approach to the attractor, during tracking, and after decoupling, when Υ~/H→0\widetilde{\Upsilon}/H \to0 and II freezes. Thus the same thermal damping that depletes the background χ\chi also erases II. In fact, II is suppressed by exactly the same factor as the squared background amplitude X2X^2 and hence the abundance Ωχ∝X2\Omega_\chi\propto X^2 at a given time (temperature)

II(ti)=X2Xfree2=ΩχΩχfree,(30)\frac{I}{I(t_i)} = \frac{X^2}{X_{\mathrm{free}}^2} = \frac{\Omega_\chi}{\Omega_\chi^{\mathrm{free}}}, \tag*{(30)}

where Xfree=XΥ~=0X_{\mathrm{free}} = X_{\widetilde{\Upsilon}=0} is the amplitude that χ\chi would have in the absence of thermal damping, starting from the same X(ti)X(t_i).

Isocurvature is present when χ\chi perturbations carry information beyond a single clock. This could arise, for instance, from the inflationary spectator initial condition δXi∼Hinf/2π\delta X_i \sim H_{\mathrm{inf}}/2\pi. After reheating, the universe then contains two local clocks: the thermal bath temperature TT and the scalar amplitude XX. Since the thermal damping rate, and hence the energy exchange between χ\chi and the thermal bath, is a function of both XX and TT, Υ~(X,T)\widetilde{\Upsilon}(X,T), it can drive patches with the same temperature TT but different scalar amplitudes XX toward one another. Since Υ~∝X2\widetilde{\Upsilon} \propto X^2, patches with larger amplitudes are damped faster, and their amplitudes converge toward a common value. The role of the attractor is to sustain Υ~∼H\widetilde{\Upsilon} \sim H for many ee-folds, so that the suppression in Eq. (29) accumulates. After many ee-folds of attractor tracking, the amplitude becomes so close to Xγ(T)X_\gamma(T) that it is essentially a function of TT alone. The system is then effectively single-clock.

A nonzero residual isocurvature can nevertheless survive if tracking is too short to erase the primordial perturbation. In addition, there are at least two effects not captured by the separate-universe picture that may source isocurvature at subleading order. First, gradient terms generate a correlated isocurvature suppressed by (k/aH)2(k/aH)^2. Second, thermal fluctuations, expected from the fluctuation-dissipation relation, generate uncorrelated white noise, i.e. a dimensionless power spectrum ∝k3\propto k^3, on scales larger than the thermal correlation length ∼T−1\sim T^{-1}. We discuss these further in Appendix C. Both subleading effects are negligible on CMB scales, but could serve as interesting targets for future probes of small-scale structures. The perturbation variable II does not describe all possible departures from a common time shift. Perturbations of the oscillation phase θ\theta of the form δθiso≡δθ−Mχδt\delta\theta_{\mathrm{iso}} \equiv\delta\theta- M_\chi\delta t are not thermally damped (see Appendix A). However, small phase perturbations do not contribute at linear order to the oscillation-averaged energy density ⟨ρχ⟩=Mχ2X2/2\langle\rho_\chi\rangle= M_\chi^2X^2/2, and hence do not source energy-density isocurvature.

Thermal damping relaxes the overclosure-isocurvature tension

The standard tension

Consider first spectator scalar dark matter without thermal effects. If the scalar is light during inflation, mχ≪Hinfm_\chi\ll H_{\mathrm{inf}}, it acquires isocurvature perturbation uncorrelated with the adiabatic mode, with power spectrum Piso≈4(Hinf/2πXi)2\mathcal{P}_{\mathrm{iso}} \approx4(H_{\mathrm{inf}}/2\pi X_i)^2. The Planck constraint on uncorrelated isocurvature is commonly expressed as Piso/(PR+Piso)<0.04\mathcal{P}_{\mathrm{iso}}/(\mathcal{P}_{\mathcal{R}}+\mathcal{P}_{\mathrm{iso}}) < 0.04 at the usual CMB pivot scale,

ToscT_{\mathrm{osc}} is the temperature at which H=mχH = m_\chi, which amounts to

Xi≤1012 GeV(mχeV)−1/4.(32)X_i \leq10^{12}\ \mathrm{GeV}\left(\frac{m_\chi}{\mathrm{eV}}\right)^{-1/4}. \tag*{(32)}

Together, the two constraints leave no viable parameter space for Hinf≳107 GeV(mχ/eV)−1/4H_{\mathrm{inf}} \gtrsim10^7\ \mathrm{GeV}\left(m_\chi/\mathrm{eV}\right)^{-1/4}. In the rest of this section we follow the isocurvature evolution for a concrete benchmark and show that thermal damping can revive this parameter space.

In Figure 1 we plot XX, II, and Υ~/H\widetilde{\Upsilon}/H for g=4×10−5g = 4 \times10^{-5}, κ=10−3\kappa= 10^{-3}, n=1n = 1 (n=5n = 5) at T≳T∗T \gtrsim T_\ast (T≲T∗T \lesssim T_\ast) with T∗=3×106 GeVT_\ast= 3 \times10^6\ \mathrm{GeV}, mψ=0.1 GeVm_\psi= 0.1\ \mathrm{GeV}. We take an inflationary Hubble scale Hinf=1012 GeVH_{\mathrm{inf}} = 10^{12}\ \mathrm{GeV}, and assume instantaneous reheating with a reheating temperature TRH=8.4×1014 GeVT_{\mathrm{RH}} = 8.4 \times10^{14}\ \mathrm{GeV}. We display the evolution for three initial amplitudes Xi=3×1016,3×1015,1014 GeVX_i = 3 \times10^{16}, 3 \times10^{15}, 10^{14}\ \mathrm{GeV}. The trajectories with the two larger XiX_i merge into a single curve, and consequently their relic abundance is independent of XiX_i. Isocurvature is suppressed during attractor tracking and remains constant after tracking ends at Tout=1.3×106 GeVT_{\mathrm{out}} = 1.3 \times10^6\ \mathrm{GeV}.

Figure 1. Evolution of the oscillation amplitude XX (top), the fractional amplitude perturbation on isothermal slices as a proxy for isocurvature I≡δX/X−(dln⁡X/dln⁡T)δT/TI \equiv\delta X/X - (\mathrm{d}\ln X/\mathrm{d}\ln T)\delta T/T (middle), and the oscillation-averaged thermal damping rate to Hubble rate ratio Υ~/H\widetilde{\Upsilon}/H (bottom). Here, g=4×10−5g = 4 \times10^{-5}, κ=10−3\kappa= 10^{-3}, T∗=3×106 GeVT_* = 3 \times10^6\,\mathrm{GeV}, mψ=0.1 GeVm_\psi= 0.1\,\mathrm{GeV}, and Hinf=1012 GeVH_{\mathrm{inf}} = 10^{12}\,\mathrm{GeV}, corresponding to TRH=8.4×1014 GeVT_{\mathrm{RH}} = 8.4 \times10^{14}\,\mathrm{GeV} for instant reheating. We use Eq. (33) and solve Eqs. (8) and (29) with Ii=Hinf/(2πXi)I_i = H_{\mathrm{inf}}/(2\pi X_i). The field remains frozen until Mχ=3HM_\chi= 3H at Tosc=4×1012 GeVT_{\mathrm{osc}} = 4 \times10^{12}\,\mathrm{GeV}. In the top panel, the orange dashed curve is XΥX_\Upsilon from Eq. (13) with ctrackc_{\mathrm{track}} replaced by n=dln⁡Γth/dln⁡Tn = \mathrm{d}\ln\Gamma_{\mathrm{th}}/\mathrm{d}\ln T and in the bottom panel, it is ctrackc_{\mathrm{track}} from Eq. (15). In the gray region, gX>TgX > T, ψ\psi is non-relativistic. The two larger XiX_i redshift freely, X∝TX \propto T, until they reach XΥX_\Upsilon and then track it, while the smallest XiX_i is only slightly thermally damped. Tracking ends at Tout=1.3×106 GeVT_{\mathrm{out}} = 1.3 \times10^6\,\mathrm{GeV}, where ctrack=0c_{\mathrm{track}} = 0. Afterwards Υ~/H\widetilde{\Upsilon}/H falls as T3T^3 and II freezes, suppressed by (3×10−5,3×10−3,0.74)(3 \times10^{-5}, 3 \times10^{-3}, 0.74) for the three XiX_i (3×1016 GeV,3×1015 GeV,1014 GeV)(3 \times10^{16}\,\mathrm{GeV}, 3 \times10^{15}\,\mathrm{GeV}, 10^{14}\,\mathrm{GeV}), the same factors by which their abundances Ωχ∝X2\Omega_\chi\propto X^2 are depleted relative to the free evolution. Throughout, X≫mψ/gX \gg m_\psi/g (dotted).

with PR≈5×10−5\sqrt{\mathcal{P}_{\mathcal{R}}} \approx5 \times10^{-5} [11]. This simplifies to

Xi≳1012 GeV(Hinf107 GeV).(31)X_i \gtrsim10^{12}\,\mathrm{GeV}\left(\frac{H_{\mathrm{inf}}}{10^7\,\mathrm{GeV}}\right). \tag*{(31)}

In the absence of thermal effects, avoiding overclosure imposes Xi≲(2ρDM,0/mχ2)1/2(g∗Tosc3/g∗,0T03)1/2X_i \lesssim(2\rho_{\mathrm{DM},0}/m_\chi^2)^{1/2}(g_{*}T_{\mathrm{osc}}^3/g_{*,0}T_0^3)^{1/2}, where

A benchmark

We take tracking to occur while the thermalization rate Γth=κTn\Gamma_{\mathrm{th}} = \kappa T^n scales with n=1n = 1. In this case, Γth/mχ,th\Gamma_{\mathrm{th}}/m_{\chi,\mathrm{th}} is temperature independent and the tracking does not end by itself. We end it by letting the exponent switch at a temperature T∗T_\ast to n=5n = 5, for which tracking ends by itself when Γth\Gamma_{\mathrm{th}} falls below ∼Mχ\sim M_\chi:

Γth={κT,T≳T∗,κT∗(TT∗)5,T≲T∗.(33)\Gamma_{\mathrm{th}} = \begin{cases} \kappa T, & T \gtrsim T_\ast, \\ \kappa T_\ast\left(\dfrac{T}{T_\ast}\right)^5, & T \lesssim T_\ast. \end{cases} \tag*{(33)}

This arises naturally, for instance, if ψ\psi interacts with the SM plasma via a vector mediator VV with mass mV∼T∗m_V \sim T_\ast and SM coupling αV\alpha_V. For T≫mVT \gg m_V the mediator is effectively massless and Γth∼αV2T\Gamma_{\mathrm{th}} \sim\alpha_V^2 T and for T≪mVT \ll m_V the mediator can be integrated out and Γth∼αV2T5/mV4\Gamma_{\mathrm{th}} \sim\alpha_V^2 T^5/m_V^4. This structure is the same as for SM neutrinos, whose weak interaction rate changes from ∼αW2T\sim\alpha_W^2 T to ∼GF2T5\sim G_F^2 T^5 at around the electroweak scale.

Right below T∗T_\ast, the ratio Γth/Mχ\Gamma_{\mathrm{th}}/M_\chi is κ/(cmg)\kappa/(c_m g). If κ/(cmg)≪1\kappa/(c_m g) \ll1, then tracking ends immediately at T∗T_\ast. Otherwise, Γth/Mχ\Gamma_{\mathrm{th}}/M_\chi decreases as T4T^4 until it falls below unity. Accounting for both cases, tracking ends at

Tout∼T∗min⁡[1,(12cmg51κ)1/4].(34)T_{\mathrm{out}} \sim T_\ast\min\left[1,\left(\sqrt{\frac{12c_m g}{5}}\frac{1}{\kappa}\right)^{1/4}\right]. \tag*{(34)}

For the benchmark of Figure 1, κ/(cmg)≈60\kappa/(c_m g) \approx60, giving Tout≈0.4T∗=1.2×106 GeVT_{\mathrm{out}} \approx0.4T_\ast= 1.2 \times10^6\ \mathrm{GeV}, consistent with the figure.

Isocurvature evolution

The evolution of the isocurvature can be divided into four stages: generation of spectator fluctuations during inflation, onset of oscillations and approach to the attractor, attractor tracking, and late-time free evolution after departure from the attractor. Throughout the last three stages, Eq. (29) applies, so the net suppression is given by e−∫dt Υ~e^{-\int dt\,\widetilde{\Upsilon}}. The stages differ only in how the integral accumulates.

During inflation, χ\chi field acquires fluctuations δX∼Hinf/2π\delta X \sim H_{\mathrm{inf}}/2\pi, giving an initial fractional field perturbation δXi/Xi∼Hinf/(2πXi)\delta X_i/X_i \sim H_{\mathrm{inf}}/(2\pi X_i) uncorrelated with the adiabatic (inflaton) perturbation. Since the radiation bath is not yet present, the isocurvature mode II defined in Eq. (25) in terms of the radiation temperature is not applicable. Nevertheless, it is clear that the inflaton and the spectator χ\chi provide two independent clocks, and δXi\delta X_i is an isocurvature perturbation. The initial condition for II for the subsequent evolution is therefore Ii=δXi/XiI_i = \delta X_i/X_i.

After reheating, the thermal population of ψ\psi generates a thermal mass and thermal damping of χ\chi. The χ\chi field begins to oscillate at the temperature ToscT_{\mathrm{osc}} where Mχ∼HM_\chi\sim H. If at that time XiX_i is well above XYX_Y then XX is driven rapidly onto the attractor XYX_Y within a fraction of a Hubble time at ToscT_{\mathrm{osc}}. During the approach, the isocurvature amplitude decays as I∝X2I \propto X^2, as per Eq. (30). By the end of the approach, X≈XY(T)X \approx X_Y(T), and the initial spectator value IiI_i has decayed to I(Tin)≈Ii(XY/Xi)Tosc2I(T_{\mathrm{in}}) \approx I_i(X_Y/X_i)^2_{\mathrm{T_{\mathrm{osc}}}}. If instead XX starts below XYX_Y, damping is negligible until the two meet after Napproach∼(2/ctrack)ln⁡(XY/X)∣ToscN_{\mathrm{approach}} \sim(2/c_{\mathrm{track}})\ln(X_Y/X)|_{T_{\mathrm{osc}}} ee-folds (see Section II), and I(Tin)=IiI(T_{\mathrm{in}}) = I_i. Every trajectory plotted in Figure 1 starts below the attractor.

Once on the attractor, Γ~=ctrackH\widetilde{\Gamma} = c_{\mathrm{track}}H, and II falls exponentially in the number of ee-folds as e−∫ToutTindln⁡T ctracke^{-\int_{T_{\mathrm{out}}}^{T_{\mathrm{in}}}d\ln T\,c_{\mathrm{track}}}, which is Tout/TinT_{\mathrm{out}}/T_{\mathrm{in}} for ctrack=1c_{\mathrm{track}} = 1. Deep into the tracking, the evolution of χ\chi is dictated by TT alone. The two initially independent clocks, TT and XX, have been driven toward each other, and the system is effectively single-clock. Unless the tracking period is short, the relative isocurvature between scalar and radiation Sχr∝IS_{\chi r} \propto I is essentially erased. At the end of tracking, XX stops chasing XYX_Y, Γ~/H\widetilde{\Gamma}/H drops below unity, and thermal damping shuts off. Thereafter, II, and with it SχrS_{\chi r}, is constant. The prediction of our scenario is therefore the standard spectator isocurvature mode uncorrelated with the adiabatic perturbation, but with its amplitude suppressed by a calculable factor,

ϵ≡I(Tout)Ii≈XY2(Tout)[Xi(Tout/Tosc)]2.(35)\epsilon\equiv\frac{I(T_{\mathrm{out}})}{I_i} \approx\frac{X_Y^2(T_{\mathrm{out}})}{\left[X_i(T_{\mathrm{out}}/T_{\mathrm{osc}})\right]^2}. \tag*{(35)}

which follows from Eq. (30), since Xfree(Tout)=XiTout/ToscX_{\mathrm{free}}(T_{\mathrm{out}}) = X_iT_{\mathrm{out}}/T_{\mathrm{osc}}. For cmg/κ≪1c_m g/\kappa\ll1, as is the case for the benchmark of Figure 1, this boils down to ϵ≈0.2κ1/4g−7/4MPlT∗1/2/Xi2\epsilon\approx0.2\kappa^{1/4}g^{-7/4}M_{\mathrm{Pl}}T_\ast^{1/2}/X_i^2. Thus the isocurvature is more suppressed for a larger XiX_i, as reflected in Figure 1.

Since the suppression is the same for every mode that is superhorizon at ToutT_{\mathrm{out}} (kk-independent), the original spectral shape is retained and only the normalization changes. Even without an attractor, any epoch with Γ~≳H\tilde{\Gamma} \gtrsim H suppresses the isocurvature by e−∫dt Γ~e^{-\int\mathrm{d}t\,\tilde{\Gamma}}. The attractor makes this suppression large and the final abundance largely insensitive to XiX_i.

Relaxing the tension

By Eq. (30), the χ\chi energy density ρχ=Mχ2X2/2\rho_\chi= M_\chi^2 X^2/2 is suppressed by the same factor ϵ\epsilon, relative to undamped evolution from the same XiX_i, as the isocurvature. Thermal mass alone already raises ToscT_{\mathrm{osc}} and can reduce ρχ,0∝Xi2/(MplTosc)\rho_{\chi,0} \propto X_i^2/(M_{\mathrm{pl}}T_{\mathrm{osc}}) relative to the pure bare-mass misalignment case. Thermal damping depletes χ\chi further and relaxes the overclosure constraint even more. Due to the attractor, the relic abundance is now independent of XiX_i and HinfH_{\mathrm{inf}}, and is instead set by where tracking ends. Since Piso∝ϵ2P_{\mathrm{iso}} \propto\epsilon^2, the isocurvature limit relaxes to Hinf≲ϵ−1×3×107 GeV (Xi/1012 GeV)H_{\mathrm{inf}} \lesssim\epsilon^{-1} \times3 \times10^7\ \mathrm{GeV}\ (X_i/10^{12}\ \mathrm{GeV}) if χ\chi constitutes the full DM. The tension is thus relaxed and the parameter space at high HinfH_{\mathrm{inf}} is reopened.

As found in Eq. (30), the isocurvature suppression factor ϵ≡Iout/Ii\epsilon\equiv I_{\mathrm{out}}/I_i is also the abundance depletion factor ϵ=ΩDM/Ωχfree(Xi)\epsilon= \Omega_{\mathrm{DM}}/\Omega_\chi^{\mathrm{free}}(X_i), where Ωχfree(Xi)=ΩχΥ=0(Xi)\Omega_\chi^{\mathrm{free}}(X_i) = \Omega_\chi^{\Upsilon=0}(X_i) is the abundance the same initial amplitude would produce without thermal damping (Υ=0\Upsilon= 0), and we again assume that χ\chi constitutes the DM at late times. One can thus rewrite the relaxed isocurvature bound in a form that is largely independent of the details of the thermal damping4

Hinf≲3×107 GeV(Xi1012 GeV)(Ωχfree(Xi)ΩDM).(36)H_{\mathrm{inf}} \lesssim3 \times10^7\ \mathrm{GeV}\left(\frac{X_i}{10^{12}\ \mathrm{GeV}}\right)\left(\frac{\Omega_\chi^{\mathrm{free}}(X_i)}{\Omega_{\mathrm{DM}}}\right). \tag*{(36)}

Thus, the isocurvature bound is relaxed by the factor by which the undamped scenario would overclose the universe, Ωχfree(Xi)/ΩDM≫1\Omega_\chi^{\mathrm{free}}(X_i)/\Omega_{\mathrm{DM}} \gg1.

A non-quadratic vacuum potential changes the relation between X(Tout)X(T_{\mathrm{out}}) and the final abundance, but the fractional reduction of the χ\chi energy density due to thermal damping, relative to the undamped evolution, is robust. Once suppressed, the isocurvature cannot be regenerated by the subsequent single-clock evolution.

Conclusion and Outlook

Light spectator scalars are subject to stringent overclosure and CMB isocurvature bounds, which assume that the scalar evolves essentially in isolation after inflation. In this work, we have studied how thermal effects reshape the post-inflation evolution of the spectator scalar and its superhorizon perturbations. A thermal population of particles coupled to the scalar produces both a thermal effective potential and thermal damping. Both effects must be accounted for in predicting the relic abundance and perturbations of the scalar.

Our central result is that amplitude-dependent thermal damping suppresses not only the relic abundance, but also isocurvature perturbations. When thermal damping increases with the oscillation amplitude, patches with larger amplitudes are damped faster, causing sufficiently large amplitudes in different patches to converge. If, in addition, the damping at fixed amplitude grows with decreasing temperature, the amplitudes converge onto an attractor, on which the amplitude is a unique function of the local temperature. Sustained tracking of the attractor dynamically converts an initially two-clock system into an effectively single-clock one. Thermal damping therefore relaxes the overclosure and isocurvature bounds simultaneously. In fact, we found in Eq. (36) that the isocurvature bound on the inflationary scale HinfH_{\mathrm{inf}} relaxes by exactly the same factor by which the overclosure bound is relaxed. This relation holds whenever the oscillation-averaged thermal damping rate scales with the oscillation amplitude XX as Γ~∝X2\tilde{\Gamma} \propto X^2, independently of its temperature dependence.

Although we have illustrated the mechanism in a specific model, the qualitative features discussed here are not tied to it. A thermal mass and thermal damping are generic consequences of coupling a scalar to a thermal bath. Whenever thermal damping increases with the amplitude, the resulting negative feedback can lead to a dynamical attractor that sets the relic abundance and erases pre-existing isocurvature. The basin of attraction, the slope of the attractor, and its endpoint depend on the particle content and interactions, but the main conclusion, that a thermal-damping attractor relaxes both the overclosure and isocurvature bounds applies broadly.

We have not carried out a systematic study of the parameter space. Doing so for concrete, well-motivated models is a natural next step. Such a study should also address the fate of ψ\psi particles, which must disappear before BBN, the late evolution of χ\chi, and the naturalness of the bare mass of χ\chi selected by the observed DM abundance. Concrete portals would also connect the cosmology to complementary experimental probes. Depending on the portal, the interactions responsible for thermal damping can be tested in beam-dump experiments, through stellar cooling, or in searches for ultra-light dark matter. It would also be interesting to revisit scalar-induced gravitational waves and other signatures of spectator scalars in light of the enlarged parameter space.

The damping of χ\chi heats the radiation in different patches of the universe by different amounts, since patches with larger initial amplitudes XiX_i dissipate more energy. As in the curvature scenario [41, 42, 43], this converts part of the primordial isocurvature IiI_i into a curva ture perturbation, δζ∼(ρχ/ρrad)Ii\delta\zeta\sim(\rho_\chi/\rho_{\rm rad})I_i. The ratio ρχ/ρrad\rho_\chi/\rho_{\rm rad} is bounded by its value at the onset of χ\chi oscillation, ρχ/ρrad∼(Xi/Mpl)2\rho_\chi/\rho_{\rm rad} \sim(X_i/M_{\rm pl})^2. This conversion is negligible for Xi≪MplX_i \ll M_{\rm pl} in cases studied here. It can, however, become efficient and source the dominant adiabatic perturbation if χ\chi makes up a sizable fraction of the energy density of the universe when it is damped, in which case χ\chi acts as a curvaton. Since the attractor ties the surviving condensate to the local temperature, such a curvaton might survive as the dark matter with adiabatic perturbations. Exploring this possibility is an interesting direction for future work.

Finally, thermal damping need not always suppress isocurvature. We have found that isocurvature is damped at the rate (1/2)∂Υ~/∂ln⁡X∣T(1/2)\left.\partial\widetilde{\Upsilon}/\partial\ln X\right|_{T}. If the damping instead decreases with amplitude, patches with larger amplitudes are damped more weakly, potentially amplifying rather than erasing isocurvature. Such amplitude dependence can arise when the damping is mediated by off-shell particles or on-shell particles that are stable but pair-annihilate [17, 23, 25]. It would be interesting to investigate this opposite case in detail.

Acknowledgements

We thank Itamar Allali, Dawid Brzeminski, Ngan H. Nguyen, Mudit Rai for useful discussion. E.H.T. is supported by NSF Physics Frontier Center grant PHY-2020275, NSF grant PHY-2310429, Simons Investigator Award No. 824870, the Gordon and Betty Moore Foundation Grant GBMF7946. The work of S.V. was supported by the Kavli Institute for Cosmological Physics at the University of Chicago. This work was performed in part at the Aspen Center for Physics, which is supported by National Science Foundation grant PHY-2210452.

Perturbation Equations

Derivation

We work in Newtonian gauge with mostly minus metric signature and neglect anisotropic stress, so the two scalar metric potentials are equal,

ds2=(1+2Φ)dt2−a2(1−2Φ)∣dx∣2(37)ds^2 = (1+2\Phi)dt^2-a^2(1-2\Phi)|d\mathbf{x}|^2 \tag*{(37)}

with −g=a3(1−2Φ)\sqrt{-g}=a^3(1-2\Phi). This yields

□χ=(1−2Φ)(χ¨+3Hχ˙)−4Φ˙χ˙−1a2∇2χ.(38)\Box\chi=(1-2\Phi)(\ddot{\chi}+3H\dot{\chi})-4\dot{\Phi}\dot{\chi}-\frac{1}{a^2}\nabla^2\chi. \tag*{(38)}

Thus the equation of motion of the scalar field χ=χˉ+δχ\chi=\bar{\chi}+\delta\chi to first order in the metric perturbation Φ\Phi is

χ¨−1a2∇2χ+3Hχ˙+Vvac′=g⟨ψˉψ⟩+4Φ˙χ˙+2Φ(g⟨ψˉψ⟩−Vvac′).(39)\ddot{\chi}-\frac{1}{a^2}\nabla^2\chi+3H\dot{\chi}+V'_{\rm vac} =g\langle\bar{\psi}\psi\rangle+4\dot{\Phi}\dot{\chi}+2\Phi\left(g\langle\bar{\psi}\psi\rangle-V'_{\rm vac}\right). \tag*{(39)}

Close to thermal equilibrium, the response of ψ\psi can be separated into equilibrium, dissipative, and stochastic contributions. We omit the stochastic part here and return to it in Appendix C. In the local rest frame of the radiation fluid, with 4-velocity uμ=(1/g00,vi/a)u^\mu=(1/\sqrt{g_{00}},v^i/a), such that uμ∇μχ≈(1−Φ)χ˙u^\mu\nabla_\mu\chi\approx(1-\Phi)\dot{\chi}, it reads [44, 45]

g⟨ψˉψ⟩=mχ,th2(mψg−χ)−(1−Φ)Yχ˙.(40)g\langle\bar{\psi}\psi\rangle=m_{\chi,\mathrm{th}}^2\left(\frac{m_\psi}{g}-\chi\right)-(1-\Phi)\mathcal{Y}\dot{\chi}. \tag*{(40)}

Its perturbation is

gδ⟨ψˉψ⟩=2mχ,th2(mψg−χˉ)δTT−mχ,th2δχ−Yδχ˙−∂Y~∂χˉ∣Tχˉ˙δχ−∂Y~∂ln⁡T∣χˉχˉ˙δTT+Yˉχ˙Φ.(41)g\delta\langle\bar{\psi}\psi\rangle=2m_{\chi,\mathrm{th}}^2\left(\frac{m_\psi}{g}-\bar{\chi}\right)\frac{\delta T}{T}-m_{\chi,\mathrm{th}}^2\delta\chi-\mathcal{Y}\dot{\delta\chi} -\left.\frac{\partial\widetilde{\mathcal{Y}}}{\partial\bar{\chi}}\right|_T\dot{\bar{\chi}}\delta\chi-\left.\frac{\partial\widetilde{\mathcal{Y}}}{\partial\ln T}\right|_{\bar{\chi}}\dot{\bar{\chi}}\frac{\delta T}{T}+\bar{\mathcal{Y}}\dot{\chi}\Phi. \tag*{(41)}

The evolution equation for the superhorizon perturbation δχ\delta\chi, Eq. (20) is obtained by perturbing Eq. (39), substituting the above, keeping the first-order terms in Φ\Phi, δχ\delta\chi, and δT\delta T, and taking the superhorizon k→0k\to0 limit. Here we take Vvac′′=Vvac′′χˉV''_{\rm vac}=V''_{\rm vac}\bar{\chi} and drop terms suppressed by mψ/(gX)m_\psi/(gX).

Phase and Amplitude

Any superhorizon perturbation δχ\delta\chi can be decomposed into a ∝cos⁡θ\propto\cos\theta piece (in phase with χˉ=Xcos⁡θ\bar{\chi}=X\cos\theta) and a ∝−sin⁡θ\propto-\sin\theta piece (nearly in phase with χˉ˙=−Xθ˙sin⁡θ+X˙cos⁡θ\dot{\bar{\chi}}=-X\dot{\theta}\sin\theta+\dot{X}\cos\theta),

δχ=X(ΔXcos⁡θ−Δθsin⁡θ),(42)\delta\chi=X\left(\Delta_X\cos\theta-\Delta_\theta\sin\theta\right), \tag*{(42)}

where ΔX=δX/X\Delta_X=\delta X/X, Δθ=δθ\Delta_\theta=\delta\theta, and we assume that the envelopes of both modes vary slowly, ∣Δ˙X/ΔX∣,∣Δ˙θ/Δθ∣≪MX|\dot{\Delta}_X/\Delta_X|,|\dot{\Delta}_\theta/\Delta_\theta|\ll M_X. Inserting the above ansatz into the δχ\delta\chi-evolution equation, multiplying by either cos⁡θ\cos\theta or sin⁡θ\sin\theta, and then time averaging using ⟨cos⁡2θ⟩=⟨sin⁡2θ⟩=1/2\langle\cos^2\theta\rangle=\langle\sin^2\theta\rangle=1/2, ⟨sin⁡θcos⁡θ⟩=0\langle\sin\theta\cos\theta\rangle=0, ⟨sin⁡2θcos⁡2θ⟩=1/8\langle\sin^2\theta\cos^2\theta\rangle=1/8, ⟨cos⁡4θ⟩=3/8\langle\cos^4\theta\rangle=3/8, ⟨sin⁡θcos⁡3θ⟩=⟨sin⁡3θcos⁡θ⟩=0\langle\sin\theta\cos^3\theta\rangle=\langle\sin^3\theta\cos\theta\rangle=0 yields

Δ˙X+12∂Y~∂ln⁡X∣TΔX=−⟨Ssin⁡θ⟩MXX=2Φ˙−12Y~Φ−12∂Y~∂ln⁡T∣XδTT,(43)\begin{aligned} \dot{\Delta}_X+\frac{1}{2}\left.\frac{\partial\widetilde{\mathcal{Y}}}{\partial\ln X}\right|_T\Delta_X =-\frac{\langle S\sin\theta\rangle}{M_X X} \\ &=2\dot{\Phi}-\frac{1}{2}\widetilde{\mathcal{Y}}\Phi-\frac{1}{2}\left.\frac{\partial\widetilde{\mathcal{Y}}}{\partial\ln T}\right|_X\frac{\delta T}{T}, \tag*{(43)} \end{aligned}
Δ˙θ=−⟨Scos⁡θ⟩MXX=MXΦ+mχ,th2MXδTT.(44)\begin{aligned} \dot{\Delta}_\theta=-\frac{\langle S\cos\theta\rangle}{M_X X} \\ &=M_X\Phi+\frac{m_{\chi,\mathrm{th}}^2}{M_X}\frac{\delta T}{T}. \tag*{(44)} \end{aligned}

Note that the phase perturbation Δθ\Delta_\theta is not thermally damped.

Adiabatic and Isocurvature

The cos⁡θ\cos\theta-sin⁡θ\sin\theta decomposition discussed above is convenient for deriving the time-averaged evolution of δχ\delta\chi modes, but it does not coincide with the cosmological adiabatic-isocurvature split.

An intuitive way to separate adiabatic and isocurvature modes is to picture it geometrically [46]. We depict the background trajectory χˉ(t)\bar{\chi}(t) in the complex plane by writing χˉ=Re⁡[Xeiθ(t)]\bar{\chi}=\operatorname{Re}[Xe^{i\theta(t)}]. Adiabatic and isocurvature perturbations are, respectively, those along and orthogonal to the χˉ(t)\bar{\chi}(t) trajectory on the complex plane. If XX were constant, the trajectory would be circular, and it is then clear that the phase perturbation (δθ=Δθ)(\delta\theta=\Delta_\theta) is adiabatic and the amplitude perturbation (δX/X=ΔX)(\delta X/X=\Delta_X) is isocurvature. For nonzero X˙\dot{X}, the trajectory becomes a spiral and its tangent has both angular and radial components. The phase perturbation (δθ=Δθ)(\delta\theta=\Delta_\theta) remains, to leading order, the adiabatic, tangent mode with a subleading, O[(∣X˙∣/X)/Mχ]O[(|\dot{X}|/X)/M_\chi] isocurvature contamination. The amplitude perturbation, on the other hand, receives an adiabatic contribution that is not small. Since ∣X˙/X∣≪Mχ|\dot{X}/X|\ll M_\chi implies that the background motion is dominantly in the phase direction, even a small tilt in the χˉ\bar{\chi} trajectory away from the phase direction contributes an adiabatic piece X˙δt/X\dot{X}\delta t/X to the radial mode ΔX=δX/X\Delta_X=\delta X/X. In a radiation-dominated universe, the time shift is δt=δT/T˙\delta t=\delta T/\dot{T}. The isocurvature is the residual amplitude that is not explained by a time shift, ΔX−X˙(δT/T˙)/X\Delta_X-\dot{X}(\delta T/\dot{T})/X.

Therefore one can, up to an arbitrary normalization, choose the following adiabatic (AA) and isocurvature (II) modes

A≡Δθ,I≡ΔX−ξδTT,(45)A\equiv\Delta_\theta,\qquad I\equiv\Delta_X-\xi\frac{\delta T}{T}, \tag*{(45)}

which evolve as

A˙=MχΦ+mχ,th2MχδTT,(46)\dot{A}=M_\chi\Phi+\frac{m_{\chi,\mathrm{th}}^2}{M_\chi}\frac{\delta T}{T}, \tag*{(46)}
I˙+12∂ln⁡Γ~∂ln⁡X∣TI=2Φ˙−ddt(ξδTT)−12Γ~Φ˙−(12∂Γ~∂ln⁡X∣Tξ+12∂Γ~∂ln⁡T∣X)δTT≈−[ξ˙+Γ~(ξ−ξΥ)]δTT.(47)\begin{aligned} \dot{I}+\frac{1}{2}\frac{\partial\ln\widetilde{\Gamma}}{\partial\ln X}\bigg|_T I &=2\dot{\Phi}-\frac{d}{dt}\left(\xi\frac{\delta T}{T}\right)-\frac{1}{2}\widetilde{\Gamma}\dot{\Phi}\\ &\quad-\left(\frac{1}{2}\frac{\partial\widetilde{\Gamma}}{\partial\ln X}\bigg|_T\xi+\frac{1}{2}\frac{\partial\widetilde{\Gamma}}{\partial\ln T}\bigg|_X\right)\frac{\delta T}{T}\\ &\approx-\left[\dot{\xi}+\widetilde{\Gamma}(\xi-\xi_\Upsilon)\right]\frac{\delta T}{T}. \tag*{(47)} \end{aligned}

The amplitude dependence of Γ~\widetilde{\Gamma} is what damps the II. For Γ~∝X2\widetilde{\Gamma}\propto X^2, ∂Γ~/∂ln⁡X=2Γ~\partial\widetilde{\Gamma}/\partial\ln X=2\widetilde{\Gamma}. In the last step of the I˙\dot{I} equation, we assume Γ~∝X2\widetilde{\Gamma}\propto X^2 and the spectator limit (valid deep in radiation domination) where the radiation continuity equation δρ˙r+4Hδρr=4Φ˙ρr\dot{\delta\rho}_r+4H\delta\rho_r=4\dot{\Phi}\rho_r reduces to d(δT/T)/dt=Φ˙d(\delta T/T)/dt=\dot{\Phi}, with Φ˙≈0\dot{\Phi}\approx0 and δT/T≈−Φ/2\delta T/T\approx-\Phi/2; we also used Eq. (14).

The structure of these equations reflects Weinberg’s adiabatic theorem [30]. The phase equation is solved exactly by the time shift δt=Φt=−(2δT/T)t\delta t=\Phi t=-(2\delta T/T)t for Mχ≈mχ,thM_\chi\approx m_{\chi,\mathrm{th}}. Consequently the phase isocurvature Δθ−Mχδt\Delta_\theta-M_\chi\delta t is conserved on superhorizon scales. The source of II vanishes on any background solution (see Section III), so an initially zero II will remain zero. The gravitational source 2Φ˙2\dot{\Phi} of I˙\dot{I} switches on only as a mode crosses the horizon, where Φ∼(k/aH)2Φ\Phi\sim(k/aH)^2\Phi. This generates the blue-tilted isocurvature of order (k/aH)2(k/aH)^2 correlated with the adiabatic one; see Appendix C.

-Radiation Relative Isocurvature

The coarse-grained continuity equations for the oscillation-averaged energy densities ρχ=Mχ2X2/2\rho_\chi=M_\chi^2X^2/2 and ρr∝T4\rho_r\propto T^4 of the χ\chi and radiation fluids read

ρ˙χ+3H(1+wχ)ρχ=Qχ,(48)\dot{\rho}_\chi+3H(1+w_\chi)\rho_\chi=Q_\chi, \tag*{(48)}
ρ˙r+4Hρr=Qr,(49)\dot{\rho}_r+4H\rho_r=Q_r, \tag*{(49)}

where

wχ=−13HM˙χMχ,Qr=−Qχ=Γ~ρχ.(50)w_\chi=-\frac{1}{3H}\frac{\dot{M}_\chi}{M_\chi},\qquad Q_r=-Q_\chi=\widetilde{\Gamma}\rho_\chi. \tag*{(50)}

We work in the spectator limit, in which ρχ≪ρr\rho_\chi\ll\rho_r and Γ~ρχ≪4Hρr\widetilde{\Gamma}\rho_\chi\ll4H\rho_r, and so χ\chi does not backreact appreciably on the radiation and the expansion. The oscillation-averaged density perturbations are

δρχρχ=2δXX+δ(Mχ2)Mχ2=2I+2(ξ+mχ,th2Mχ2)δTT,(51)\frac{\delta\rho_\chi}{\rho_\chi}=2\frac{\delta X}{X}+\frac{\delta(M_\chi^2)}{M_\chi^2}=2I+2\left(\xi+\frac{m_{\chi,\mathrm{th}}^2}{M_\chi^2}\right)\frac{\delta T}{T}, \tag*{(51)}
δρradρrad=4δTT.(52)\frac{\delta\rho_{\mathrm{rad}}}{\rho_{\mathrm{rad}}}=4\frac{\delta T}{T}. \tag*{(52)}

We define the curvature perturbation on uniform-ρα\rho_\alpha slices ζα\zeta_\alpha as the e-fold shift between flat slicing and uniform-ρα\rho_\alpha slicing ζα=−Hδρα/ρ˙α\zeta_\alpha=-H\delta\rho_\alpha/\dot{\rho}_\alpha. The total curvature perturbation is the weighted sum ζ=∑α(ρ˙α/ρ˙)ζα\zeta=\sum_\alpha(\dot{\rho}_\alpha/\dot{\rho})\zeta_\alpha, which in the limit that the energy density of χ\chi is negligible, reduces to ζ≃ζr\zeta\simeq\zeta_r. The relative isocurvature between χ\chi and radiation is Sχr≡3(ζχ−ζr)S_{\chi r}\equiv3(\zeta_\chi-\zeta_r) and can be expressed as

Sχr=−3H(δρχρ˙χ−δρrρ˙r)=3Iξ+mχ,th2/Mχ2,(53)S_{\chi r}=-3H\left(\frac{\delta\rho_\chi}{\dot{\rho}_\chi}-\frac{\delta\rho_r}{\dot{\rho}_r}\right)=\frac{3I}{\xi+m_{\chi,\mathrm{th}}^2/M_\chi^2}, \tag*{(53)}

where we have used ρ˙χ/ρχ=−2H[mχ,th2/Mχ2+ξ]\dot{\rho}_\chi/\rho_\chi=-2H[m_{\chi,\mathrm{th}}^2/M_\chi^2+\xi]. This further reduces to Sχr=3I/(1+ξ)S_{\chi r}=3I/(1+\xi) when the potential of χ\chi is thermal-mass dominated.

In the interacting-fluid formalism of Ref. [47], the erasure of SχrS_{\chi r} is driven by the perturbed energy transfer, δQχ=−δ(Γ~ρχ)\delta Q_\chi=-\delta(\widetilde{\Gamma}\rho_\chi), which has a non-adiabatic component stemming from the XX dependence of Γ~\widetilde{\Gamma}. If Γ~\widetilde{\Gamma} depended on TT alone, δQχ\delta Q_\chi would be completely adiabatic.

Subleading Isocurvature Sources

We find in the main text that thermal-damping attractor tracking exponentially suppresses the spectator isocurvature mode, resulting in a nearly adiabatic final perturbation. This conclusion holds in the separate-universe (deep superhorizon) limit and after neglecting stochastic thermal fluctuations of the radiation. Spatial gradients and thermal noise can produce additional isocurvature perturbations, but both have spectra that are strongly blue and are negligible on CMB scales. Below we denote the temperature below which thermal damping becomes negligible as ToutT_{\mathrm{out}}.

Gradient Effects

Reinstating the gradient terms (k/a)2δχ(k/a)^2\delta\chi and Φ˙∼(k2/a2H2)HΦ\dot{\Phi} \sim(k^2/a^2H^2)H\Phi in Eq. (20) produces corrections to resulting δχ\delta\chi. These effects lead parametrically to [48, 49]

Sχr∼(kaH)2ζ.(54)S_{\chi r} \sim\left(\frac{k}{aH}\right)^2 \zeta. \tag*{(54)}

The resulting isocurvature is correlated with the curvature perturbation ζ\zeta. The mode crossing the horizon at the end of tracking is kout=aH∣Toutk_{\mathrm{out}} = aH|_{T_{\mathrm{out}}}, which evaluates to kout=1×104 Mpc−1(Tout/MeV)k_{\mathrm{out}} = 1 \times10^4\,\mathrm{Mpc}^{-1}(T_{\mathrm{out}}/\mathrm{MeV}), where we took g∗(Tout)=10.75g_*(T_{\mathrm{out}}) = 10.75. Note that Tout≳MeVT_{\mathrm{out}} \gtrsim\mathrm{MeV} by construction. In the extreme case where tracking persists until the onset of BBN, Tout∼1 MeVT_{\mathrm{out}} \sim1\,\mathrm{MeV}, the modes k∼2×104 Mpc−1k \sim2 \times10^4\,\mathrm{Mpc}^{-1} remain superhorizon throughout the damping epoch. At the CMB pivot scale k∗=0.05 Mpc−1k_* = 0.05\,\mathrm{Mpc}^{-1}, for instance, one has (k∗/kout)2∼2×10−11(Tout/MeV)−2(k_*/k_{\mathrm{out}})^2 \sim2 \times10^{-11}(T_{\mathrm{out}}/\mathrm{MeV})^{-2}. Therefore the corresponding suppression in the isocurvature power spectrum, (k∗/kout)4(k_*/k_{\mathrm{out}})^4, is typically more severe than 10−2210^{-22}, far beyond the current observational reach. Nevertheless it is a prediction of our scenario that could serve as an interesting target for probes of higher-kk modes, such as Ref. [50].

Thermal Noise

The fluctuation-dissipation relation implies that thermal damping must be accompanied with stochastic fluctuations. The perturbation equation Eq. (20) should therefore include a thermal noise source η\eta. A cosmological scale typically contains an enormous number of independent thermal-correlation domains, which add up incoherently and largely cancel, up to Poisson mis-cancellations. More formally, at scales larger than the thermal correlation length, the noise is approximately Gaussian with correlator ⟨η(t,x)η(t′,x′)⟩=(2ΥT/a3)δ(t−t′)δ(3)(x−x′)\langle\eta(t,\mathbf{x})\eta(t',\mathbf{x}')\rangle= (2\Upsilon T/a^3)\delta(t-t')\delta^{(3)}(\mathbf{x}-\mathbf{x}'), which corresponds to white noise ⟨ηk(t)ηk′(t′)⟩=(2ΥT/a3)(2π)3δ(3)(k+k′)δ(t−t′)\langle\eta_k(t)\eta_{k'}(t')\rangle= (2\Upsilon T/a^3)(2\pi)^3\delta^{(3)}(\mathbf{k}+\mathbf{k}')\delta(t-t') in the kk-space. Its effect is highly suppressed by the smallness of k/aHk/aH on observationally relevant scales [51].

Models

We have parametrized the thermalization rate of ψ\psi as Γth=κTn\Gamma_{\mathrm{th}} = \kappa T^n. If the scalar primarily perturbs the number density of ψ\psi then the ψ\psi thermalization rate Γth\Gamma_{\mathrm{th}} is set by number-changing processes that restore chemical equilibrium of ψ\psi with the primordial plasma. If it instead mainly distorts the momentum distribution at fixed number density, Γth\Gamma_{\mathrm{th}} is set by processes that restore kinetic equilibrium. In either case, the relevant rate is the slowest one needed to restore equilibrium.

The n=1n = 1 case can be realized, for instance, with a light vector mediator (mV≪T)(m_V \ll T) coupled to both (Vμψˉγμψ)\left(V_\mu\bar{\psi}\gamma^\mu\psi\right) and SM fermions (VμψˉSMγμψSM)\left(V_\mu\bar{\psi}_{\mathrm{SM}}\gamma^\mu\psi_{\mathrm{SM}}\right); the same model realizes n=5n = 5 if the mediator is heavy (mV≫T)(m_V \gg T). If the dominant number-changing process is 2↔22 \leftrightarrow2 annihilation ψψˉ↔SM SM\psi\bar{\psi} \leftrightarrow\mathrm{SM\,SM}, then the thermalization rate is parametrically Γth∼nψσψ↔SM∼T3σψ↔SM\Gamma_{\mathrm{th}} \sim n_{\psi}\sigma_{\psi\leftrightarrow\mathrm{SM}} \sim T^3\sigma_{\psi\leftrightarrow\mathrm{SM}}. If thermalization is mediated by an operator of mass dimension dd, suppressed by Λ4−d\Lambda^{4-d}, dimensional analysis gives Γth∼T2d−7/Λ2d−8\Gamma_{\mathrm{th}} \sim T^{2d-7}/\Lambda^{2d-8}, which corresponds to n=2d−7n = 2d - 7 and κ∼Λ8−2d\kappa\sim\Lambda^{8-2d}. For d=4,5,6,7,…d = 4, 5, 6, 7, \ldots, this gives n=1,3,5,7,…n = 1, 3, 5, 7, \ldots; d=4d = 4 is the light vector mediator case above, d=5d = 5 includes the fermionic Higgs portal (ψˉψ)(H†H)/Λ(\bar{\psi}\psi)(H^\dagger H)/\Lambda, and d=6d = 6 includes four-fermion operators and the heavy limit of the vector mediator above. Thermalization via relativistic 2→22 \to2 annihilation therefore naturally produces positive odd values of nn. The case n=−1n = -1 can be realized with SψˉψS\bar{\psi}\psi, for which the (inverse-)decay rate is Γth∝Mψ2/T\Gamma_{\mathrm{th}} \propto M_\psi^2/T.

Some of these example thermalization rates Γth\Gamma_{\mathrm{th}} introduce dependence on MψM_\psi, and hence on XX. In those cases the attractor analysis must be redone with the modified XX dependence.

References

References

  1. [1]R. Essig et al., in Snowmass 2013: Snowmass on the Mississippi (2013) arXiv:1311.0029 [hep-ph].
  2. [2]D. Antypas et al., (2022), arXiv:2203.14915 [hep-ex].arxiv.org/abs/2203.14915
  3. [3]P. W. Graham, H. Ramani, O. Simon, and E. H. Tanin, (2025), arXiv:2511.09614 [hep-ph].arxiv.org/abs/2511.09614
  4. [4]A. A. Starobinsky and J. Yokoyama, Phys. Rev. D 50, 6357 (1994), arXiv:astro-ph/9407016.DOI
  5. [5]E. W. Kolb and A. J. Soltz, Rev. Mod. Phys. 96, 045005 (2024), arXiv:2312.09042 [astro-ph.CO].arxiv.org/abs/2312.09042
  6. [6]P. W. Graham and A. Scherlis, Phys. Rev. D 98, 035017 (2018), arXiv:1805.07362 [hep-ph].DOI
  7. [7]M. P. Hertzberg, M. Tegmark, and F. Wilczek, Phys. Rev. D 78, 083507 (2008), arXiv:0807.1726 [astro-ph].DOI
  8. [8]K. Kainulainen, S. Nurmi, T. Tenkanen, K. Tuominen, and V. Vaskonen, JCAP 06, 022 (2016), arXiv:1601.07733 [astro-ph.CO].DOI
  9. [9]T. Markkanen, A. Rajantie, and T. Tenkanen, Phys. Rev. D 98, 123532 (2018), arXiv:1811.02586 [astro-ph.CO].DOI
  10. [10]T. Tenkanen, Phys. Rev. Lett. 123, 061302 (2019), arXiv:1905.01214 [astro-ph.CO].DOI
  11. [11]Y. Akrami et al. (Planck), Astron. Astrophys. 641, A10 (2020), arXiv:1807.06211 [astro-ph.CO].DOI
  12. [12]D. J. H. Chung, E. W. Kolb, A. Riotto, and L. Senatore, Phys. Rev. D 72, 023511 (2005), arXiv:astro-ph/0411468.
  13. [13]A. D. Linde, Phys. Lett. B 259, 38 (1991).
  14. [14]S. E. Henrich, Y. Mambrini, and K. A. Olive, (2026), arXiv:2610.07331 [hep-ph].arxiv.org/abs/2610.07331
  15. [15]M. L. Bellac, Thermal Field Theory, Cambridge Monographs on Mathematical Physics (Cambridge University Press, 2011).
  16. [16]M. Quiros, in ICTP Summer School in High-Energy Physics and Cosmology (1999) pp. 187–259, arXiv:hep-ph/9901312.
  17. [17]E. H. Tanin and E. D. Stewart, JCAP 11, 019 (2017), arXiv:1708.04865 [hep-ph].DOI
  18. [18]A. Banerjee, H. Nguyen, and E. H. Tanin, (2025), arXiv:2512.05188 [hep-ph].arxiv.org/abs/2512.05188
  19. [19]A. Banerjee, N. H. Nguyen, and E. H. Tanin, (2025), arXiv:2509.05238 [hep-ph].
  20. [20]A. Hosoya, M.-a. Sakagami, and M. Takao, Annals Phys. 154, 229 (1984).
  21. [21]J. Yokoyama and A. D. Linde, Phys. Rev. D 60, 083509 (1999), arXiv:hep-ph/9809409.DOI
  22. [22]J. Yokoyama, Phys. Rev. D 70, 103511 (2004), arXiv:hep-ph/0406072.DOI
  23. [23]K. Mukaida and K. Nakayama, JCAP 01, 017 (2013), arXiv:1208.3399 [hep-ph].DOI
  24. [24]K. Mukaida and K. Nakayama, JCAP 03, 002 (2013), arXiv:1212.4985 [hep-ph].DOI
  25. [25]K. Mukaida, K. Nakayama, and M. Takimoto, JHEP 12, 053 (2013), arXiv:1308.4394 [hep-ph].DOI
  26. [26]K. Mukaida, K. Nakayama, and M. Takimoto, JCAP 06, 013 (2014), arXiv:1401.5821 [hep-ph].DOI
  27. [27]M. Bastero-Gil, A. Berera, and R. O. Ramos, JCAP 09, 033 (2011), arXiv:1008.1929 [hep-ph].
  28. [28]D. Bödeker and J. Nienaber, Phys. Rev. D 106, 056016 (2022), arXiv:2205.14166 [hep-ph].DOI
  29. [29]M. Drewes and J. U. Kang, Nucl. Phys. B 875, 315 (2013), [Erratum: Nucl.Phys.B 888, 284–286 (2014)], arXiv:1305.0267 [hep-ph].arxiv.org/abs/1305.0267
  30. [30]S. Weinberg, Phys. Rev. D 67, 123504 (2003), arXiv:astro-ph/0302326.DOI
  31. [31]J. Allali, P. Chakraborty, J. Fan, and M. Reece, Phys. Rev. D 114, 043517 (2026), arXiv:2510.07371 [hep-ph].arxiv.org/abs/2510.07371
  32. [32]B. Batell and A. Ghalsasi, Phys. Rev. D 107, L091701 (2023), arXiv:2109.04476 [hep-ph].arxiv.org/abs/2109.04476
  33. [33]B. Batell, A. Ghalsasi, and M. Rai, JHEP 01, 038 (2024), arXiv:2211.09132 [hep-ph].arxiv.org/abs/2211.09132
  34. [34]B. Batell, A. Ghalsasi, S. Ghosh, and M. Rai, (2026), arXiv:2603.18132 [hep-ph].arxiv.org/abs/2603.18132
  35. [35]L. D. McLerran, E. Mottola, and M. E. Shaposhnikov, Phys. Rev. D 43, 2027 (1991).
  36. [36]K. V. Berghaus, P. W. Graham, and D. E. Kaplan, JCAP 03, 034 (2020), [Erratum: JCAP 10, E02 (2023)], arXiv:1910.07525 [hep-ph].DOI
  37. [37]L. Ji, D. E. Kaplan, S. Rajendran, and E. H. Tanin, Phys. Rev. D 105, 015025 (2022), arXiv:2109.05285 [hep-ph].
  38. [38]A. Papageorgiou, P. Quiléz, and K. Schmitz, JHEP 01, 169 (2023), arXiv:2206.01129 [hep-ph].DOI
  39. [39]K. Choi, S. H. Im, H. J. Kim, and H. Seong, JHEP 02, 180 (2023), arXiv:2206.01462 [hep-ph].DOI
  40. [40]M. Bastero-Gil, A. Berera, I. G. Moss, and R. O. Ramos, JCAP 05, 004 (2014), arXiv:1401.1149 [astro-ph.CO].
  41. [41]T. Moroi and T. Takahashi, Phys. Rev. D 64, 023515 (2001), [Erratum: Phys.Lett.B 539, 303–303 (2002)], arXiv:hep-ph/0110096.DOI
  42. [42]D. H. Lyth and D. Wands, Phys. Lett. B 524, 5 (2002), arXiv:hep-ph/0110002.DOI
  43. [43]K. Enqvist and M. S. Sloth, Nucl. Phys. B 626, 395 (2002), arXiv:hep-ph/0109214.
  44. [44]M. Bastero-Gil, A. Berera, and R. O. Ramos, JCAP 07, 030 (2011), arXiv:1106.0701 [astro-ph.CO].
  45. [45]M. Mirbabayi and A. Gruzinov, JCAP 02, 012 (2023), arXiv:2205.13227 [astro-ph.CO].DOI
  46. [46]C. Gordon, D. Wands, B. A. Bassett, and R. Maartens, Phys. Rev. D 63, 023506 (2000), arXiv:astro-ph/0009131.DOI
  47. [47]K. A. Malik and D. Wands, JCAP 02, 007 (2005), arXiv:astro-ph/0411703.DOI
  48. [48]I. Holst, W. Hu, and L. Jenks, Phys. Rev. D 109, 063507 (2024), arXiv:2311.17164 [astro-ph.CO].DOI
  49. [49]A. Stebbins, (2023), arXiv:2311.17379 [astro-ph.CO].arxiv.org/abs/2311.17379
  50. [50]P. W. Graham and H. Ramani, Phys. Rev. D 110, 075012 (2024), arXiv:2404.01378 [hep-ph].DOI
  51. [51]J.-M. Bae, H. R. Mohapatra, M. E. D. Stewart, and H. Zoe, JCAP 04, 038 (2025), arXiv:2501.06580 [astro-ph.CO].arxiv.org/abs/2501.06580

Paper details

vlnc:W_s4x5496n
Version
v1 · latest
Published
Fresh — added to the Hub within the last 30 days
Added
License
CC BY 4.0
Usage

views · 0

downloads · 0

A view counts once per browser per day, a download each time the PDF is requested.

  • Physics and Astronomy

Not recorded for this paper

View statement
Looking for related papers…
Contents