Axions at the meV Crossroads: Theory, Cosmology, Astrophysics, and Experiments

Abstract

The meV mass range has emerged as a focal point in axion physics, where advances in theory, cosmology, astrophysics, and experimental techniques converge. Axions in this mass range are theoretically well motivated, can arise in ultraviolet-complete models, and can have significant cosmological impacts as dark matter or dark radiation. In parallel, their efficient production in stellar and supernova environments provides powerful astrophysical probes. Here, we provide a comprehensive overview of meV axions across these domains, highlighting both established results and open questions. We discuss the theoretical underpinnings of meV axions, their cosmological and astrophysical signatures, and the diverse experimental strategies -- ranging from helioscopes and haloscopes to quasiparticle systems and large-volume Cherenkov detectors -- that aim to explore this regime. The convergence of these approaches emphasizes the pivotal role of the meV mass range for axion discovery in the coming years, identifying meV axions as a key probe for testing beyond-Standard-Model physics. This review document is the direct outcome of the discussions at the dedicated workshop "The meV Mass Axion Frontier: Challenges and Opportunities", held at Laboratori Nazionali di Frascati (IT) on 27--28 October 2025, and organized by the EU funded COST Action "Cosmic WISPers in the Dark Universe: Theory, astrophysics, and experiments" (CA21106, https://www.cost.eu/actions/CA21106). Its aim is to provide an overview of current efforts in meV axion research, their motivations, and the research goals that animate the community involved in this search.

Figures (14)

• Figure 1: QCD axion mass, at random points along a ray from the tip of the SKC, for different values of $h^{1,1}$ in type IIB CY models. Left. Probability distribution on the QCD axion mass colour coded by $h^{1,1}$. Distributions move towards meV for $h^{1,1}\gtrsim 200$ (reproduced from Gendler:2024adn). Right. QCD axion mass as a function of $h^{1,1}$. Coloured dots denote the average in the bin, the line denotes a power law assuming $\mathcal{V}\sim (h^{1,1})^5$. The horizontal solid line is the current limit from SN1987A, and the dotted line is the projection for the next galactic supernova (reproduced from Gendler:2023kjt).
• Figure 2: Left. Value of the initial misalignment angle required to obtain different fractions of axion cold dark matter. Right. Upper bound on the CDM abundance of meV axions in the LMA scenario as a function of the inflationary rate $H_I$. Figure reproduced from Ref. IAXO:2019mpb.
• Figure 3: Abundance of $N_{\rm DW}=1$ post-inflationary axions from strings as fraction of the observed CDM. Different colors correspond to different assumptions about the spectral index, $q \approx 1$ (blue) and $q \gg 1$ (orange) evaluated at the physical value of $\ell\sim 70$, and the latter includes the effect of nonlinearities discussed in Ref. Gorghetto:2020qws. The colored regions also show the systematic uncertainties associated with the extrapolation of the simulation results estimated in Ref. Saikawa:2024bta.
• Figure 4: Predicted $\Delta N_{\rm eff}$ from thermal axion production as a function of $f_a$, compared with current and future CMB sensitivities. Left. KSVZ axion framework, with the corresponding $m_a$ shown on the upper horizontal axis DEramo:2021lgb. Right. Axion couplings to SM fermions with unit Wilson coefficient, based on the full phase-space analysis of Ref. DEramo:2024jhn.
• Figure 5: Hillas plot of the astrophysical sources as axion-photon converters. The colored lines identify the probability of axion-photon conversion for massless axions, while the gray lines identify the axion masses at which coherence in the axion-photon conversion is lost. Adapted from Ref. Fiorillo:2025gnd.