The chemical signatures of the first stars: from near- to far-field cosmology

The first stars, also called Population III (Pop III) stars, formed about 200-400 Myr after the Big Bang in primordial composition gas clouds dwelling in low-mass dark matter structures, or mini-halos. Pop III stars played a crucial role in shaping the early phases of the evolution of the Universe producing the first ionizing photons and the first chemical elements heavier than helium, the so-called metals. Cosmological simulations and indirect studies predict that Pop III stars were likely more massive than present-day Population II/I (Pop II/I) stars, with masses possibly up to ∼ 1000 M⊙. However, different studies find a variety of mass ranges for Pop III stars. This implies that the Pop III initial mass function (IMF) is still unknown although likely biased towards higher masses with respect to present-day stars. If this is the case, then the majority of Pop III stars evolved as supernovae (SNe) in few Myr. However, also the explosion energy of Pop III SNe is unknown, probably distributed between 10^50 and 10^53 erg. The amounts of photons and metals dispersed in primordial galaxies strongly depend on the IMF of Pop III stars and on the explosion energy distribution function of their SNe. Knowing these functions is the key to understand how the Universe evolved. The metals ejected by Pop III SNe promptly enriched the interstellar medium (ISM), and eventually the intergalactic medium (IGM). Second-generation stars then formed in metal-enriched galaxies with masses distributed as present-day stars. Those with masses ≤ 0.8M⊙ can survive until today. Thus, the metals ejected by Pop III SNe are now retained on the surfaces of these long-lived stellar descendants of Pop III stars and in distant gas clouds that now we can observe as high-redshift absorption systems in the spectra of further quasars. Thus, we can use the chemical fingerprints left by Pop III SNe in these different environments to infer their properties. Second-generation stars are expected to be metal-poor because they formed in environments enriched solely by Pop III stars. The most metal-poor environments close enough for us to observe individual stars are the halo of the Milky Way and some of its satellite galaxies. In particular, stars that are iron-poor, rich in carbon and other alpha-elements but with low barium abundance, the so-called CEMP-no stars, are considered to be direct descendants of Pop III SNe. The use of high-redshift absorption systems as tracers of the metals produced by Pop III stars is a more recent field of study. Although their spectra reveal the abundances of only a few elements, they provide complementary insights to those obtained from metal-poor stars. Indeed, second-generation stars form from the dense gas of the ISM, whereas the gas probed by absorption systems can span a range of densities, from the ISM to the CGM. The aim of this Thesis is to characterize the chemical fingerprints of Pop III SNe, to identify, among the observed metal-poor stars and high-redshift absorbers, which have been predominantly enriched by Pop III SNe, and to finally infer the properties of their pollutants. To achieve my objectives I developed a parametric model, with which I predicted the chemical abundances of environments solely enriched by Pop III SNe by exploring their full possible range of initial masses and explosion energies. For the first time, furthermore, I’ve included in my simple and general parametric model the subsequent pollution by normal Pop II SNe. This allowed me to investigate for how long the Pop III chemical signature is preserved. The first key issue I addressed in my Thesis is whether all metal-poor stars in the Galactic halo are second-generation stars. I demonstrated that only CEMP-no stars with [C/Fe]> +2.5 form in environments polluted purely by low-to-normal energy Pop III SNe. These results are in agreement with chemical evolution models of ultra-faint dwarf galaxies and semi-analytical models coupled with N-body simulations of the Milky Way assembly. Conversely, I showed that most halo stars with [C/Fe] ≲ 0 have been predominantly imprinted by normal Pop II SNe. Still, I demonstrated that (rare) second-generation stars polluted by energetic hypernovae (10 − 100 M⊙, E ≥ 5×10^51 erg) or Pair-Instability SNe (PISNe, 140−260M⊙, E = 10^52−10^53 erg) can also have subsolar [C/Fe]. For this reason, the identification of hypernovae- and PISNe-descendants is difficult using C and Fe abundances and the abundances of other elements (such as Zn) should be exploited. With this strategy, we identified the descendant of a primordial hypernova in the satellite dwarf galaxy Sculptor, by comparing the observed abundance pattern with the predictions of my model. On the contrary, PISNe-descendants have not been identified yet, nor in dwarf galaxies, nor in the Galactic halo, although a recent claim. Indeed, this star has been re-observed by our group, and we found that its abundances are no more consistent with a PISN enrichment, which can only account for <10% of its metals. However, PISNe descendants are predicted to be very rare: <0.1% of the total stellar population in the Galactic halo, thus we might need to observe more stars to uncover one of them. In anticipation of future spectroscopic surveys that will dramatically increase the number of stars with measured chemical abundances, we prepared the PISN-explorer tool, to analyze large amount of data and pinpoint PISN descendants using my theoretical predictions. While we are still not observing PISNe fingerprints in local stars, these might have been found in high-z absorption systems observed with JWST. Inspired by this discovery, I developed new chemical diagnostics to uncover environments enriched by PISNe and distinguish them from those enriched by less massive Pop III and/or by normal Pop II SNe. In this publication, we also analyze higher resolution spectra of two of the three putative PISN-enriched absorbers, finding that they have instead normal chemical abundances. However, by comparing my novel chemical diagnostics with the abundances of high-z absorbers present in literature, I pinpointed a peculiar z ≈ 3 absorber which might has been enriched by a primordial PISN. Yet, we would need the abundance of other elements, such as Zn, to confirm its nature. This is one of the science cases for the ELT spectrograph ANDES, which is expected to observe with unprecedented resolution high-z absorbers. In conclusion, during my PhD Thesis I developed a general parametric model that I used to interpret the chemical abundances of Pop III-enriched regions of different natures. Very soon, my predictions will become publicly available as an online fitting tool and will be usable by the community to interpret the huge amount of data from upcoming instruments and surveys for metal-poor stars and high-z absorbers.