Thermal Friction as a Solution to the Hubble Tension

TL;DR

The paper proposes a dissipative axion model with a dark non-Abelian gauge sector that overdamps the axion and sources a dark-radiation bath, yielding an early dark-energy–like component that peaks near matter–radiation equality. This background-level behavior can mimic the phenomenology of EDE and reconcile CMB-based H0 inferences with local measurements, without resorting to finely tuned scalar potentials. The authors compare the DA's background evolution to standard EDE, arguing that key CMB background observables such as the sound horizon are scarcely altered and identify two main parameters governing the effect. They discuss the model’s UV motivation, potential extensions to perturbations, and broader cosmological applications beyond the Hubble tension.

Abstract

A new component added to the standard model of cosmology that behaves like a cosmological constant at early times and then dilutes away as radiation or faster can resolve the Hubble tension. We show that a rolling axion coupled to a non-Abelian gauge group exhibits the behavior of such an extra component at the background level and can present a natural particle-physics model solution to the Hubble tension. We compare the contribution of this bottom-up model to the phenomenological fluid approximation and determine that CMB observables sensitive only to the background evolution of the Universe are expected to be similar in both cases, strengthening the case for this model to provide a viable solution to the Hubble tension.

Figures (2)

• Figure 1: The fractional energy densities $\Omega_i = \rho_i / \rho_{\rm crit}$ of the different components in the DA and those in a $\Lambda$CDM universe, where $\rho_{\rm crit}$ is the critical density today. The total DA contribution (green) is a sum of its sub-components. At early times ($z \gg z_d$), the energy density $\Omega_{\phi}$ in the scalar field (blue) is roughly constant and the dark radiation component $\Omega_{\text{dr}}$ (yellow) is subdominant. At intermediate times $(z_{\text{peak}} < z < z_d)$, the dark radiation $\Omega_{\text{dr}}$ transitions to become dominant as $\Omega_{\phi}$ drops. Shortly after $T_{\text{dr}}$ reaches a maximum, the total fractional DA energy density peaks at redshift $z_{\text{peak}}$.
• Figure 2: We compare the fractional early dark energy density of the full temperature dependent DA model [$\Upsilon(z) \propto T^3_{\text{dr}}$, solid green] with the semi-analytical approximations in equation \ref{['fDA']} and \ref{['rhoda']}, treating the friction as constant [$\Upsilon(z) \approx \Upsilon_0$ dashed green] and the EDE fluid approximation of an oscillating scalar field from Poulin et. al. Poulin:2018cxd (purple). This plot uses the $n=2$ EDE best-fit parameters [$z_{\text{c}} = 5345, f_{\text{EDE}}(z_{\text{c}}) =0.044$ which corresponds to $z_{\text{peak}} = 3322,\, f_{\text{EDE}}(z_{\text{peak}}) =0.060$] and dissipative axion parameters $\frac{\Upsilon(z_{\text{peak}})}{m^2} = 1.3 *10^{36}\, \text{GeV}^{-1}$$(\frac{\Upsilon_0}{m^2} = 5.7 *10^{36}\,\ \text{GeV}^{-1})$, and $\frac{1}{2}m^2 \phi^2_0 = 0.55 \,\ \text{eV}^4$$(\frac{1}{2}m^2 \phi^2_0 = 0.21 \,\ \text{eV}^4)$ for the temperature dependent (independent) DA model.