Heavy Neutral Leptons (HNLs) are strongly motivated by theory due to their capability of simultaneously explaining the observed phenomena of dark matter, neutrino oscillations and the baryon asymmetry of the Universe. The existence of such particles would affect the expansion history of the Universe and the synthesis of primordial abundances of light elements. In this work we review, revise and extend the phenomenology of HNLs during the Big Bang Nucleosynthesis (BBN) epoch for masses up to 1 GeV. This is of great importance, as BBN is able to provide complementary bounds to those from upcoming and proposed laboratory experiments. To this end we have developed a high-precision Boltzmann code that simulates BBN in the presence of HNLs and takes into account all relevant HNL decay channels, as well as subsequent interactions of decay products (thermalization and decay showers), dilution due to QCD phase transition, active neutrino oscillations and corrections to the weak reaction rates. We present robust bounds on the lifetime and mixing angles of HNLs for masses $3\,\mathrm{MeV}\leq m_N \leq 1\,\mathrm{GeV}$ and show that BBN is able to constrain HNL lifetimes down to $0.03 - 0.05$ s, depending on the mixing pattern. Moreover, combining our results with current experimental searches, we can exclude HNLs that mix purely with electron neutrinos up to ${\sim}$450 MeV and those that mix purely with muon neutrinos up to ${\sim}$360 MeV, in both cases for lifetimes up to at least a few tens of seconds. Finally, we compare the BBN constraints with those obtained from Cosmic Microwave Background observations and explore how our results will be improved by a number of upcoming and proposed laboratory experiments.

The Lyman-$\alpha$ forest is a powerful tool to constrain warm dark matter models (WDM). Its main observable -- flux power spectrum -- should exhibit a suppression at small scales in WDM models. This suppression, however, can be mimicked by a number of thermal effects related to the instantaneous temperature of the intergalactic medium (IGM), and to the history of reionization and of the IGM heating ("pressure effects"). Therefore, to put robust bounds on WDM one needs to disentangle the effect of free-streaming of dark matter particles from the influence of all astrophysical effects. This task cannot be brute-forced due to the complexity of the IGM modelling. In this work, we model the sample of high-resolution and high-redshift quasar spectra (Boera et al 2018) assuming a thermal history that leads to the smallest pressure effects while still being broadly compatible with observations. We explicitly marginalize over observationally allowed values of IGM temperature and find that (thermal) WDM models with masses above 1.9 keV (at 95% CL) are consistent with the spatial shape of the observed flux power spectrum at $z=4-5$. Even warmer models would produce a suppression at scales that are larger than observed, independently of assumptions about thermal effects. This bound is significantly lower than previously claimed bounds, demonstrating the importance of the knowledge about the reionization history and of the proper marginalization over unknowns.

The observed Lyman-$\alpha$ flux power spectrum (FPS) is suppressed on scales below $\sim~ 30~{\rm km~s}^{-1}$. This cutoff could be due to the high temperature, $T_0$, and pressure, $p_0$, of the absorbing gas or, alternatively, it could reflect the free streaming of dark matter particles in the early universe. We perform a set of very high resolution cosmological hydrodynamic simulations in which we vary $T_0$, $p_0$ and the amplitude of the dark matter free streaming, and compare the FPS of mock spectra to the data. We show that the location of the dark matter free-streaming cutoff scales differently with redshift than the cutoff produced by thermal effects and is more pronounced at higher redshift. We, therefore, focus on a comparison to the observed FPS at $z>5$. We demonstrate that the FPS cutoff can be fit assuming cold dark matter, but it can be equally well fit assuming that the dark matter consists of $\sim 7$ keV sterile neutrinos in which case the cutoff is due primarily to the dark matter free streaming.