Title: How much metals did the first stars provide to the ultra-faint dwarfs?

Abstract: Numerical simulations of dwarf galaxies have so far failed to reproduce the observed metallicity-luminosity relation, down to the ultra-faint dwarfs (UFDs). We address this issue exploring how the first generations of metal-free stars (Pop III) could help increase the mean metallicity of those faint galaxies. We run zoom-in chemo-dynamical simulations of nineteen halos extracted from a cosmological box and follow down to redshift 0. Models are validated not only on the basis of galaxy global properties, but also the stellar abundance ratios. We identify the necessary conditions for the formation of first stars in mini-halos and derive constraints on the metal ejection schemes. The impact of Pop III stars on the final metallicity of UFDs is evaluated by considering different IMFs, the influence of pair-instability supernovae (PISNe) and their energetic feedback, as well as the metallicity threshold marks the transition from first stars to the formation of low-mass long-lived stars. The inclusion of Pop III stars does increase the global metallicity of UFDs, though insufficient to resolve the tension with observations. PISNe with progenitor masses above 140Msun do allow to further increase the metal content of UFDs. However, as PISNe are rare and sometimes absent in the faintest UFDs, they have a limited impact on the global faint end of the metallicity-luminosity relation. Despite a limited number of spectroscopically confirmed members in UFDs, that makes the metallicity distribution of some UFDs uncertain, our analysis reveals this is the metal-rich tail that is missing in the models. The remaining challenges are thus both observational and numerical: i) to extend high resolution spectroscopy data samples and confirm the mean metallicity of the faintest UFDs, ii) to explain the presence of chemically enriched stars in galaxies with very short star formation histories.