New cosmological bounds on hot relics: axions and neutrinos

Axions, if realized in nature, can be copiously produced in the early universe via thermal processes, contributing to the mass-energy density of thermal hot relics. In light of the most recent cosmological observations, we analyze two different thermal processes within a realistic mixed hot-dark-matter scenario which includes also massive neutrinos. Considering the axion-gluon thermalization channel we derive our most constraining bounds on the hot relic masses $m_a < 7.46$ eV and $sum m_ u< 0.114$ eV both at 95% CL; while studying the axion-pion scattering, without assuming any specific model for the axion-pion interactions and remaining in the range of validity of the chiral perturbation theory, our most constraining bounds are improved to $m_a<0.91$ eV and $sum m_ u< 0.105$ eV, both at 95% CL. Interestingly, in both cases, the total neutrino mass lies very close to the inverted neutrino mass ordering prediction. If future terrestrial double beta decay and/or long baseline neutrino experiments find that the nature mass ordering is the inverted one, this could rule out a wide region in the currently allowed thermal axion window. Our results therefore strongly support multi-messenger searches of axions and neutrino properties, together with joint analyses of their expected sensitivities.