Metastable Dark Energy with Radioactive-like Decay

TL;DR

This work introduces metastable dark energy with a constant intrinsic decay rate $\Gamma$, so the DE density decays as $\rho_{DE}(t)=\rho_{DE}(t_0)\exp[-\Gamma (t-t_0)]$ and the half-life is $t_{1/2}=\ln(2)/\Gamma$. It analyzes three decay channels—exponential decay, DE$\rightarrow$dark matter, and DE$\rightarrow$dark radiation—and studies their impact on the expansion history via FRW evolution, requiring $w_{tot}>-1$. Using data from SNIa, BAO, high-$z$ BAO, and CMB shift parameters with CosmoMC, the authors constrain $\Gamma$ and find that the half-life must exceed the age of the Universe, with DE$\rightarrow$DM providing the best fit by alleviating the CMB–BAO tension; DE$\rightarrow$DR is tightly constrained and DE exponential decay remains viable but not preferred over $\Lambda$CDM. The framework shows that metastable DE can leave observable imprints on the cosmic expansion and high-redshift observables, offering a testable alternative to a true cosmological constant.

Abstract

We propose a new class of metastable dark energy (DE) phenomenological models in which the DE decay rate does not depend on external parameters such as the scale factor or the curvature of the Universe. Instead, the DE decay rate is assumed to be a constant depending only on intrinsic properties of DE and the type of a decay channel, similar to case of the radioactive decay of unstable particles and nuclei. As a consequence, the DE energy density becomes a function of the proper time elapsed since its formation, presumably in the very early Universe. Such a natural type of DE decay can profoundly affect the expansion history of the Universe and its age. Metastable DE can decay in three distinct ways: (i) exponentially, (ii) into dark matter, (iii) into dark radiation. Testing metastable DE models with observational data we find that the decay half-life must be many times larger than the age of the Universe. Models in which dark energy decays into dark matter lead to lower values of the Hubble parameter at large redshifts relative to $Λ$CDM. Consequently these models provide a better fit to cosmological BAO data (especially data from from high redshift quasars) than concordance ($Λ$CDM) cosmology.

Figures (8)

• Figure 1: Observational constraints on standard $\Lambda$CDM. [Left] One dimensional marginalized likelihoods of $\Omega_{\rm 0m}h^2$. [Right] $\lbrace\Omega_{\rm 0m},H_0\rbrace$ contours for different datasets. The color-code convention for the right panel is identical to the left.
• Figure 2: Results for Model I. [Top left] One dimensional marginalized likelihoods of the decay parameter in Eq. \ref{['eq:model-I']}. [Top right] 1D likelihoods of $\Omega_{\rm 0m}h^2$. [Bottom left] 2D contours of $\lbrace\Omega_{\rm 0m},H_0\rbrace$. [Bottom right] 2D contours of $\lbrace\Omega_{\rm DE},\Gamma/H_0\rbrace$. One can see that the tension between CMB data and the $H(2.34)$ data point is somewhat reduced in this model.
• Figure 3: The equation of state of dark energy, $w(z),$ the $Om$ diagnostic $Om(z)$, and the deceleration parameter $q(z)$ (top, middle, bottom) are shown as a function of redshift for model I (Eq. \ref{['eq:model-I']}). Left to right we plot samples within 2$\sigma$ CL's obtained from MCMC chains corresponding to Union-2.1 + BAO, Union-2.1 + BAO + H(2.34) and Union-2.1 + BAO + H(2.34) + CMB respectively. The black lines correspond to the best fit $\Lambda$CDM for the same combination of datasets.
• Figure 4: Results for Model II. [Top left] One dimensional marginalized likelihoods of the decay parameter $\Gamma$ are shown for model II (Eq. \ref{['eq:model-II']}). [Top right] 1D likelihoods of $\Omega_{\rm 0m}h^2$. [Bottom left] 2D contours of $\lbrace\Omega_{\rm 0m},H_0\rbrace$. [Bottom right] 2D contours of $\lbrace\Omega_{\rm DE},\Gamma/H_0\rbrace$. Unlike $\Lambda$CDM, no tension between CMB data and the $H(2.34)$ data point is indicated for Model II.
• Figure 5: Samples of $Om(z)$ and $q(z)$ for model II (Eq. \ref{['eq:model-II']}). The datasets are indicated at the top of each plot. The black lines correspond to the best fit $\Lambda$CDM for the same combination of datasets.