• Axion-like particles
• Dark Matter
• Beyond Standard Model
• Cosmology

In this thesis, we study a model of light dark matter (DM) in which the possible candidate is an axion-like particle (ALP). This is interesting for two reasons: firstly, one can have
light DM within well-motivated models, protected by shift symmetries, where the DM candidate emerges as a Pseudo-Nambu Goldstone boson; the second reason is that it
can be probed experimentally. We propose simple scenarios where quark flavour-violating (FV) couplings generate the observed DM abundance through the Freeze-in of
an ALP with mass in the few hundred keV range. Compared to flavour-diagonal (FD) Freeze-in, this mechanism enhances DM stability, softens stellar cooling constraints
and improves the experimental sensitivity of accelerator-based searches. The main steps of this work concern the discussion of DM production and stability. For the first, we include the associated FD scatterings for each allowed FV decay involving gluons and photons. These processes are infrared (IR), so they do not depend on ultraviolet (UV) physics details. The decay constant has been fixed to reproduce the observed relic density, by solving the Boltzmann equation numerically for the Freeze-in mechanism. In addition, we studied the effect of a non-renormalizable operator that
introduces UV-sensitive processes like those involving the Higgs scattering. The next step concerns stability: to be a possible DM candidate, the ALP must be stable on the
cosmological time scale. The only relevant decay channel is $a \to \gamma \gamma$, which we compute using the Chiral Perturbation Theory. Finally, we fix the constraints on FV
ALP properties using results from astrophysics probes and collider experiments.