Early cosmic star formation in the Milky Way environment

Understanding how the first stars formed and gave rise to the first galaxies is one of the major challenges of modern Cosmology. In standard Λ Cold Dark Matter (ΛCDM) cosmological models the formation of dark matter (DM) structures occurs hierarchically from smaller DM halos (the so-called “mini-halos”, with virial temperatures T_vir < 104K and total mass ~ 10^6 M⊙ collapsing at redshift z ≈ 20) which gradually merge to form larger halos (T_vir > 104K). Prior to the formation of the stars, the gas in the primordial Universe was made mostly by H and He, with traces of Li and Be. Indeed, heavier elements (the so-called metals) are formed in the interior of stars and released into the surrounding gas through stellar winds and supernova (SN) explosions. Hence, the first stars, referred to as Population III (Pop III) stars, have formed in DM halos out of metal-free gas. Numerical simulations that study the formation of the first stars and look at different physical processes do not agree in the final mass of Pop III stars: while some simulations predict them to be much more massive than present-day stars, from ∼10s to 100s M⊙ (or even ∼1000 M⊙), others show that they can also have subsolar masses. Hence the initial mass function (IMF) of Pop III stars is still largely unknown. Being likely massive, Pop III stars do not survive until today, however they are the first sources of chemical enrichment, producing metals and dust. Heavy elements provide efficient cooling channels for the formation of subsequent less massive stars, the so called Population II (Pop II) stars, which could survive to present-day. Although it has been demonstrated that the presence of heavy elements can trigger the transition to the first low-mass stars at a given metallicity Z, the role of metal line-cooling and dust cooling in driving this transition is still debated. The duration of Pop III star formation and their impact on cosmic evolution depend on the complex interplay of different feedback processes, other than the chemical one. The ultraviolet (UV) radiation quenches star formation by photo-dissociating H2 molecules (i.e. the main coolant at high-z) and gradually (re)ionize H in the intergalactic medium (IGM), reducing/suppressing gas infall in mini-halos. When exploding, Pop III stars also eject gas from the halo through SN-driven outflows. Thus the duration of the Pop III epoque is unknown due the complex interplay between these feedback effects. Up to now no metal-free star has been detected. Yet a promising way to investigate the properties of Pop III stars is by studying the ancient most metal-poor stars observed in our own Milky Way (MW) and its environment. Measuring the chemical abundances in their photosphere and assuming that they are the same of their natal cloud, polluted by Pop III stars, allows to constrain the nucleosynthetic products of Pop III stars. In addition, the number distribution of stars as function of their [Fe/H], the so-called Metallicity Distribution Function (MDF), provides constraints on the physical processes regulating star formation at high-z. Large surveys provide the Galactic halo MDF and chemical abundances of most metal-poor stars, showing an increasing [C/Fe] with decreasing [Fe/H]. These observations could give important information on the nature of the first stars, on the physical processes driving the transition to the first low-mass stars, and on feedback effects regulating star formation at high-z. To unveil the potential of these observations, adequate theoretical models must be adopted to connect high-z star formation with the MW local relics. Here we use two complementary tools: (i) GAMETE, a code which follows star formation in a large sample of semi-analytical DM merger trees (plausible MW formation histories, which is unknown) thus allowing to rapidly explore different model parameters and to statistically quantify the errors induced by different histories, although no spatial information is provided; (ii) GAMESH, a pipeline where GAMETE is applied to a DM merger tree from an N-body simulation, and combined with the radiative transfer code CRASH. This approach cannot account for different MW formation histories, but allows to follow in detail the build-up of the MW accounting for the spatial distribution of the star forming progenitors and for the ionizing radiation they produce locally. The outline of the Thesis is the following: In Chapter 1 we describe the theory of formation of DM halos where first stars form. In particular we focus on the physics leading to the formation of Pop III stars and to the transition to low-mass stars, describing the main feedback processes which could affect the star formation at high-z. We also present available observations for ancient metal-poor stars in the MW Galactic halo. In Chapter 2 we investigate whether current observations of the Galactic halo MDF can provide constraints on the physics of the Pop III/II transition and some indications on the mass of Pop III stars. To this end we use GAMETE to follow the chemical enrichment (metals and dust) across the MW formation. We explore different mass ranges and chemical yields of Pop III stars and compare simulated and observed MDF. This part of the work has been published in de Bennassuti, Schneider, Valiante & Salvadori (2014), MNRAS, 445, 3039. In Chapter 3, we investigate the role of star formation in mini-halos and its effect in shaping the Galactic halo MDF, providing robust data-driven constraints on the PopIII IMF. To this aim we use an improved version of GAMETE, to self-consistently describe the physical processes regulating star-formation in mini-halos: the poor sampling of the Pop III IMF and the effect of UV radiation. We study the effect of this new physics and of the IMF of Pop III stars on the MDF and on the properties of C-enhanced and C-normal stars. This part of the work is published in de Bennassuti, Salvadori, Schneider, Valiante (2017), MNRAS, 465, 926. In Chapter 4 we study the interplay between different feedback processes along the MW formation. To this aim, we apply GAMESH to a low-resolution N-body simulation and we account for the radiative transfer of ionizing photons to follow the inhomogeneous reionization and heating of the IGM. This part of the work has been published in Graziani, Salvadori, Schneider, Kawata, de Bennassuti, Maselli (2015), MNRAS, 449, 3137. In Chapter 5, we study the history of the dark and luminous MW progenitors and their role in shaping the properties of the MW. This is done by applying GAMESH to a higher resolution simulation which allows a more detailed investigation of the MW properties, also providing a larger statistics of mini-halos and satellite galaxies. This part of the work will be published in a forthcoming paper. Finally, in Chapter 6 we present the main conclusions of the work.