Thermal dark matter at the MeV scale faces stringent bounds from a variety of cosmological probes. Here we perform a detailed evaluation of BBN bounds on the annihilation cross section of dark matter with a mass 1 MeV m χ 1 GeV. For p-wave suppressed annihilations, constraints from BBN turn out to be significantly stronger than the ones from CMB observations, and are competitive with the strongest bounds from other indirect searches. We furthermore update the lower bound from BBN on the mass of thermal dark matter using improved determinations of primordial abundances. While being of similar strength as the corresponding bound from CMB, it is significantly more robust to changes in the particle physics model.

Axion-like particles with masses in the keV-GeV range have a profound impact on the cosmological evolution of our Universe, in particular on the abundance of light elements produced during Big Bang Nucleosynthesis. The resulting limits are complementary to searches in the laboratory and provide valuable additional information regarding the validity of a given point in parameter space. A potential drawback is that altering the cosmological history may potentially weaken or even fully invalidate these bounds. The main objective of this article is therefore to evaluate the robustness of cosmological constraints on axion-like particles in the keV-GeV region, allowing for various additional effects which may weaken the bounds of the standard scenario. Employing the latest determinations of the primordial abundances as well as information from the cosmic microwave background we find that while bounds can indeed be weakened, very relevant robust constraints remain.

In this work, we revise and update model-independent constraints from Big Bang Nucleosynthesis on MeV-scale particles ϕ which decay into photons and/or electron-positron pairs. We use the latest determinations of primordial abundances and extend the analysis in [1] by including all spin-statistical factors as well as inverse decays, significantly strengthening the resulting bounds in particular for small masses. For a very suppressed initial abundance of ϕ, these effects become ever more important and we find that even a pure `freeze-in' abundance can be significantly constrained. In parallel to this article, we release the public code ACROPOLIS which numerically solves the reaction network necessary to evaluate the effect of photodisintegration on the final light element abundances. As an interesting application, we re-evaluate a possible solution of the lithium problem due to the photodisintegration of beryllium and find that e.g. an ALP produced via freeze-in can lead to a viable solution.

The idea of dark matter in the form of primordial black holes has seen a recent revival triggered by the LIGO detection of gravitational waves from binary black hole mergers. In this context, it has been argued that a large initial clustering of primordial black holes can help alleviate the strong constraints on this scenario. In this work, we show that on the contrary, with large initial clustering the problem is exacerbated and constraints on primordial black hole dark matter become overwhelmingly strong.

No abstract