Several laboratory experiments have published limits on axionlike particles (ALPs) with feeble couplings to electrons and masses in the kilo-electron-volt to mega-electron-volt range, under the assumption that such ALPs comprise the dark matter. We note that ALPs decay radiatively into photons, and show that for a large subset of the parameter space ostensibly probed by these experiments, the lifetime of the ALPs is shorter than the age of the Universe. Such ALPs cannot consistently make up the dark matter, which significantly affects the interpretation of published limits from GERDA, Edelweiss-III, SuperCDMS, and Majorana. Moreover, constraints from x-ray and γ-ray astronomy exclude a wide range of the ALP-electron coupling, and supersede all current laboratory limits on dark matter ALPs in the 6 keV to 1 MeV mass range. These conclusions are rather model independent, and can only be avoided at the expense of significant fine-tuning in theories where the ALP has additional couplings to other particles.

We show that in theories of axionlike particles (ALPs) coupled to electrons at tree-level, the one-loop effective coupling to photons is process dependent: the effective coupling relevant for decay processes, g_aγ^(D), differs significantly from the coupling appearing in the phenomenologically important Primakoff process, g_aγ^(P). We show that this has important implications for the physics of massive ALPs in hot and dense environments, such as supernovae. We derive, as a consequence, new limits on the ALP-electron coupling, ĝ_ae, from SN 1987A by accounting for all relevant production processes, including one-loop processes, and considering bounds from excess cooling as well as the absence of an associated gamma-ray burst from ALP decays. Our limits are among the strongest to date for ALP masses in the range 0.03 MeV < m_a < 240 MeV. Moreover, we also show how cosmological bounds on the ALP-photon coupling translate into new, strong limits on ĝ_ae at one loop. Our analysis emphasises that large hierarchies between ALP effective couplings are difficult to realise once quantum loops are taken into account.

We investigate the characteristics of the gamma-ray signal following the decay of MeV-scale Axion-Like Particles (ALPs) coupled to photons which are produced in a Supernova (SN) explosion. This analysis is the first to include the production of heavier ALPs through the photon coalescence process, enlarging the mass range of ALPs that could be observed in this way and giving a stronger bound from the observation of SN 1987A. Furthermore, we present a new analytical method for calculating the predicted gamma-ray signal from ALP decays. With this method we can rigorously prove the validity of an approximation that has been used in some of the previous literature, which we show here to be valid only if all gamma rays arrive under extremely small observation angles (i.e. very close to the line of sight to the SN). However, it also shows where the approximation is not valid, and offers an efficient alternative to calculate the ALP-induced gamma-ray flux in a general setting when the observation angles are not guaranteed to be small. We also estimate the sensitivity of the Fermi Large Area Telescope (Fermi-LAT) to this gamma-ray signal from a future nearby SN and show that in the case of a non-observation the current bounds on the ALP-photon coupling g_aγ are strengthened by about an order of magnitude. In the case of an observation, we show that it may be possible to reconstruct the product g_aγ²m_a, with m_a the mass of the ALP.