The low-end of the black hole mass function at cosmic dawn

Understanding the formation and growth of supermassive black holes (SMBHs) at high redshift represents a major challenge for theoretical models. In this work we investigate the early evolution of the first SMBHs by constraining their distribution in mass and luminosity at $z > 4$. In particular, we focus on the poorly explored low-mass end of the nuclear black hole (BH) distribution down to $z \simeq 4$, and explore its connection with the nature of the first BH seeds and the processes governing their mass growth. To this aim, we have developed CAT (Cosmic Archaeology Tool), a new semi-analytic model that describes the formation of the first stars and black holes in a self-consistent way and follows the co-evolution of nuclear BHs and their host galaxies for a representative population at $z > 4$. We find that current observational constraints favour models where the growth of BH seeds is Eddington limited and occurs at the Bondi-Hoyle-Lyttleton rate or where super-Eddington accretion occurs via a slim disk during gas rich galaxy mergers. The main difference between these two model variants lies at the low-end of the predicted mass and luminosity functions at $4 \le z \le 6$, where a clear gap appears in the first model, reflecting the stunted growth of light BH seeds formed as remnants of the first stars. Detecting this signature will be extremely challenging even for the future generation of space observatories, such as JWST, Athena and Lynx.