Resolving the Hubble Tension with New Early Dark Energy

TL;DR

This work proposes New Early Dark Energy (NEDE), a first‑order phase transition in a dark sector near recombination, to resolve the Hubble tension. It develops a two‑field microscopic model with a trigger field that induces rapid nucleation and bubble percolation, and then constructs an effective instantaneous cosmological description that matches perturbations across the transition surface via Israel’s junction conditions. Parameter inference using Planck, BAO, Pantheon, and large‑scale structure data (with and without the local $H_0$ prior) shows that NEDE can raise $H_0$ and reduce the $H_0$ tension to about $2.5\sigma$ without the local data, and to $\sim4\sigma$ when the local measurement is included, with $f_{\rm NEDE}\sim\mathcal{O}(0.1)$ and a transition redshift around $z_*\,\sim\,5{,}000$, while remaining compatible with CMB and LSS observations. The model predicts a stochastic gravitational‑wave background from bubble collisions and distinctive small‑scale CMB signatures, offering several observational tests. Compared to other early‑dark‑energy proposals, NEDE ties its phenomenology to a concrete microphysics with a trigger mechanism, providing a viable, technically natural mechanism to reconcile early‑ and late‑time measurements of the expansion rate. The findings have significant implications for upcoming CMB and GW surveys and motivate further exploration of the dark‑sector trigger dynamics and potential connections to UV completions.

Abstract

New Early Dark Energy (NEDE) is a component of vacuum energy at the electron volt scale, which decays in a first-order phase transition shortly before recombination [arXiv:1910.10739]. The NEDE component has the potential to resolve the tension between recent local measurements of the expansion rate of the Universe using supernovae (SN) data and the expansion rate inferred from the early Universe through measurements of the cosmic microwave background (CMB) when assuming $Λ$CDM. We discuss in depth the two-scalar field model of the NEDE phase transition including the process of bubble percolation, collision, and coalescence. We also estimate the gravitational wave signal produced during the collision phase and argue that it can be searched for using pulsar timing arrays. In a second step, we construct an effective cosmological model, which describes the phase transition as an instantaneous process, and derive the covariant equations that match perturbations across the transition surface. Fitting the cosmological model to CMB, baryonic acoustic oscillations and SN data, we report $H_0 = 69.6^{+1.0}_{-1.3} \, \textrm{km}\, \textrm{s}^{-1}\, \textrm{Mpc}^{-1}$ $(68 \%$ C.L.) without the local measurement of the Hubble parameter, bringing the tension down to $2.5\, σ$. Including the local input, we find $H_0 = 71.4 \pm 1.0 \, \textrm{km}\, \textrm{s}^{-1}\, \textrm{Mpc}^{-1}$ $(68 \%$ C.L.) and strong evidence for a non-vanishing NEDE component with a $\simeq 4\, σ$ significance.

Figures (17)

• Figure 1: Dimensionless potential $\bar{V}$ for different choices of $\delta$. The vacua become degenerate as $\delta \to 2$. For $\delta \to 0$, the barrier height vanishes and the true vacuum is deepest.
• Figure 2: Illustration of the potential in \ref{['eq:action2']}. At early times, the field is frozen high up the false vacuum valley (above the blue dot) and protected against tunneling to the true vacuum (white dot) by a high potential barrier. Tunneling is turned on when the field rolls past the orange dot. Shortly after that, bubble percolation overcomes the expansion of space and reaches its maximal efficiency at the red dot.
• Figure 3: Result of the two-field shooting method. Each dot corresponds to one numerical integration. Different colors correspond to different choices of $\tilde{\lambda} / \lambda$.
• Figure 4: Sensitivity curve (blue solid) of the Square Kilometre Array (SKA) in terms of the dimensionless energy density of gravitational waves. Below the black dotted line (dark red) we can describe the phase transition as an instantaneous process on cosmological scales, making it accessible to our effective description in Sec. \ref{['sec:cosmo_model']}. On the other hand, pushing $\bar{\beta}$ towards cosmological scales, i.e., $\bar{\beta}^{-1} \to H^{-1}$, might lead to an observable gravitational wave signal. In this limit, we also expect the colliding bubble walls to leave a direct imprint in the CMB (green shaded region). We used $f_\text{NEDE} = 0.14$ and $z_* = 5300$ as of Tab. \ref{['tab:means_NEDE_LCDM']}.
• Figure 5: Background evolution for the best fit model in Tab. \ref{['tab:means_NEDE_LCDM']} (w/ SH$_0$ES) as a function of the scale factor $a$. The transition is characterized by a jump in the effective equation of state parameter (dotted line). $\rho_m = \rho_b + \rho_\text{cdm}$ and $\rho_\text{rad}$ are the matter and radiation density, respectively, and $\rho_\Lambda$ is the cosmological constant contribution.