Original article can be found at: http://www.nature.com/nature/index.html--Copyright Nature Publishing Group --DOI : 10.1038/nature0345

No abstract

We describe the construction of a database of extremely metal-poor (EMP) stars in the Galaxy. Our database contains detailed elemental abundances, reported equivalent widths, atmospheric parameters, photometry, and binarity status, compiled from papers in the literature that report on studies of EMP halo stars with [Fe=H] Ä 2.5. The compilation procedures for this database were designed to assemble data effectively from electronic tables available from online journals. We have also developed a data retrieval system that enables data searches by various criteria and illustrations to explore relationships between stored variables. Currently, our sample includes 1212 unique stars (many of which are studied by more than one group) with more than 15000 individual reported elemental abundances, covering relevant papers published by 2007 December. We discuss the global characteristics of the present database, as revealed by the EMP stars observed to date. For stars with [Fe=H] Ä 2.5, the number of giants with reported abundances is larger than that of dwarfs by a factor of two. The fraction of carbon-rich stars (among the sample for which the carbon abundance is reported) amounts to 30% for [Fe=H] Ä 2.5. We find that known binaries exhibit different distributions of the orbital period, according to whether they are giants or dwarfs, and also as a function of the metallicity, although the total sample of such stars is still quite small.

We discuss the origin of HE0107-5240 which, with a metallicity of [Fe/H] = −5.3, is the most iron-poor star yet observed. Its discovery has an important bearing on the question of the observability of first generation stars in our Universe. In common with other stars of very small metallicity (−4 [Fe/H] −2.5), HE0107-5240 shows a peculiar abundance pattern, including large enhancements of C, N, and O, and a more modest enhancement of Na. The observed abundance pattern can be explained by nucleosynthesis and mass transfer in a first generation binary star, which, after birth, accretes matter from a primordial cloud mixed with the ejectum of a supernova. We elaborate the binary a metallicity distribution function for first generation (Pop. III) stars currently burning hydrogen and conclude that HE0107-5240 is a first generation object with a surface affected by accreting interstellar matter polluted with heavy elements. Umeda & Nomoto (2003) adjust parameters in a first generation supernova model in such a way as to produce a C/Fe ratio in the supernova ejectum that agrees with the ratio observed in HE0107-5240 and argue that HE0107-5240 is a second generation object formed from the primordial cloud after it has been mixed with the ejectum of the supernova. A similar scenario is presented by Limongi, Chieffi, & Bonifacio (2003) who argue that the HE0107-5240 abundances can be produced by a combination of two types of supernova ejecta: a normal ejectum consisting of ∼ 0.06M ⊙ of iron and an abnormal ejectum consisting only of products of partial helium burning. Though all extant scenarios address important aspects of the problem, further discussion is warranted of the physics of star formation and of the chemical composition expected in a primordial cloud into which matter ejected by a supernova has been mixed. More importantly, the modifications of surface abundances which HE0107-5240 has suffered during its long life should be elucidated. In particular, the possibility of accretion from an evolved first generation companion which has mixed to its surface products of internal nucleosynthesis should be explored. In this paper, we describe results of such an exploration.A major characteristic of models of extremely metal poor stars is that, although the p-p chain reactions are the dominant source of the stellar luminosity and the main driver of evolution during most of the core hydrogen-burning phase, the CNO cycles play an increasingly important role as evolution progresses beyond the main sequence phase. This is because, as temperatures increase, carbon is produced by the highly temperature-sensitive triple-α reaction and because, at high temperatures, only a small abundance of CNO elements is needed for CNO cycle reactions to become the dominant driver of evolution. This character-

The number of known extremely metal-deficient stars has recently increased substantially, stimulating inquiry into the formation and initial chemical evolution of the Milky Way. In order to draw proper inferences from the observations, it is necessary to understand the evolution of these low-mass stars and the modifications in their surface elemental abundances that they have experienced during their long lifetimes. Among the observations to be explained is the fact that the incidence of carbon-enhanced stars increases with decreasing metallicity. We show that low-mass, extremely metal-poor stars evolve into carbon stars along paths that are quite different from those followed by more metal-rich stars of younger populations. This permits us, in principle, to distinguish the brightest survivors of the first generations of stars formed in the universe (Population III carbon stars) from stars belonging to younger populations.