Abundances in the most metal-poor stars: nucleosynthesis in the early Milky Way

GEPI, UMR 8111, l'Observatoire de Paris, 5 pl. J. Janssen, 92195 Meudon Cedex, France

Abstract

The detailed abundance patterns in extremely metal-poor (EMP) stars are examined, using the results of the First Stars ESO-VLT Large Program on large samples of dwarfs (located close to the main sequence turnoff) and giant stars (Cayrel et al. 2004; Bonifacio et al. 2007 and subsequent papers). Abundances measured in giants are compared to those measured in turnoff stars, and to comparable quality data in the literature, to demonstrate that the systematics involved in most abundance ratios are very small. A few exceptions are discussed: a reanalysis of Na and Al in EMP samples is performed using NLTE computations. In addition, giants with luminosities brighter than the RGB bump are shown to exhibit the sign of surface pollution by mixing of the products of H-burning, affecting the measured abundances of C, N and Li. 
The implications of a few key abundance ratios (among those considered to be robust and representative of the early galactic gas) are examined in terms of the early Milky-Way enrichment processes and the possible nucleosynthesis contributors (PISNs, SNII, hypernovae). The very low dispersion found around the mean trends is also discussed in terms of chemical enrichment processes.