Cosmological bounds on sub-MeV mass axions

TL;DR

This work extends cosmological constraints on axions into the sub-MeV regime by analyzing how axion decays modify the baryon-photon ratio, neutrino dilution, and the CMB’s radiation content and spectrum. A Boltzmann treatment of axion production, decoupling, and decay via Primakoff/Compton/inverse-decay processes is combined with a modified BBN calculation to track the impact on deuterium yields, and with CMB analyses to constrain N_eff and μ-distortions. The authors find robust bounds: m_a > 300 keV from BBN, m_a > 3 keV from N_eff, and m_a > 8.7 keV from μ-distortions (for δ=1), collectively excluding the 0.7 eV–300 keV window for hadronic axions. These cosmological bounds complement laboratory and stellar limits, highlighting precision cosmology’s power to probe light, decaying particles.

Abstract

Axions with mass greater than 0.7 eV are excluded by cosmological precision data because they provide too much hot dark matter. While for masses above 20 eV the axion lifetime drops below the age of the universe, we show that the cosmological exclusion range can be extended from 0.7 eV till 300 keV, primarily by the cosmic deuterium abundance: axion decays would strongly modify the baryon-to-photon ratio at BBN relative to the one at CMB decoupling. Additional arguments include neutrino dilution relative to photons by axion decays and spectral CMB distortions. Our new cosmological constraints complement stellar-evolution limits and laboratory bounds.

Figures (8)

• Figure 1: Axion decoupling and recoupling ($\delta=1$, $C_e=1/6$). Thick solid line: Freeze-out of Primakoff process. Medium solid line: Coupling and freeze-out of the Compton process. Thin solid line: Recoupling of inverse decay. In the yellow shaded region, axions are in thermal equilibrium, where the lighter yellow region is only relevant if the Compton process is effective. The dashed diagonal line denotes $m_a=3T$. The vertical lines delimit the $e^+e^-$ annihilation epoch.
• Figure 2: Axion number density $n_a$ after $e^+e^-$ annihilation from numerically solving the Boltzmann equation until $T=m_e/10$. The equilibrium density $n_a^{\rm eq}$ is defined in terms of the photon temperature. Solid line: only Primakoff process (hadronic axions). Dashed line: Primakoff and Compton.
• Figure 3: Photon density increase in our modified cosmology as expressed by $B_2/B_1$ for $\delta=1$. Solid and dashed lines stand for hadronic and non-hadronic ($C_e=1/6$) axions respectively. The thin lines show the value if we assume that no entropy is generated in axion decay.
• Figure 4: Deuterium yield $y_D$ as a function of $m_a$ for $\delta=1$ (Hadronic axions). The width of the red band represents the $1\sigma$ uncertainty of the CMB determination of $\eta$. To the right of the break, axions are treated as being in LTE, to the left they are assumed to be decoupled during BBN. The grey bands represent the 95% and 99% range for the observed D abundance. It is derived from 7 high redshift Ly-$\alpha$ clouds with individual results shown in the right panel as a function of the hydrogen column density Pettini:2008mq.
• Figure 5: 2D marginal 68%, 95% and 99% contours in the $m_\nu$--$N_{\rm eff}$ plane, where $m_\nu$ is the individual neutrino mass (not the often-used sum over masses).