ar X iv : a st ro - p h / 04 10 66 6 v 1 2 7 O ct 2 00 4 Cosmic abundances : The impact of stellar duplicity

The mass-transfer scenario links chemical peculiarities with stellar duplicity for an increasing number of stellar classes (classical and dwarf barium stars, subgiant and giant CH stars, S stars without technetium, yellow symbiotic stars, WIRRING stars, Abell-35-like nuclei of planetary nebulae...). Despite these successes, the mass-transfer scenario still faces several problems: What is the mass-transfer mode? Why orbital elements of dwarf barium stars do not fully match those of the classical barium stars? What is the origin of the few non-binary stars among dwarf barium stars? 1. The mass-transfer scenario: The origins The impact of the binary nature of a star on its chemical composition seems to have been recognized first in the context of Am and Ap stars in the years 1960– 1968 (Bonsack 1961; Fowler et al. 1965; Conti 1965; Van den Heuvel 1967a,b, 1968a,b,c). Van den Heuvel (1968a) writes: [many properties of] Ap and Am stars (...) can all be explained if these stars were originally the less massive members in spectroscopic binaries in which the primary during its post-mainsequence evolution filled its Roche lobe, transferred a large fraction of its mass towards the secondary and finally evolved into a white dwarf. Current views about the role of duplicity among Am and Ap stars will be shortly reviewed in Sects. 2. and 4.3. Two decades later, the mass-transfer scenario popped up again, this time in relation with barium stars and following the discovery by McClure et al. (1980) of a large frequency of spectroscopic binaries among the family of barium stars. According to McClure, it is not unreasonable to conclude that BaII stars are all binaries with low-mass secondaries consisting of degenerate objects. It is possible that these systems are such that mass has been lost from a more massive evolving star and deposited onto the present primary, the secondary having now evolved to the white dwarf stage. The carbon and s-process elements in this case could be dumped onto the present primary-star atmosphere. (...) The CH stars, which are probably the Population II equivalent of BaII stars, may also, by implication, be binaries. At the time, E. Van den Heuvel was still affiliated with the Astronomical Institute of Brussels Free University! For a review about the s-process of nucleosynthesis and its major astrophysical site, the asymptotic giant branch (AGB) stars, the reader is referred to Lattanzio & Wood (2003)