In the last decade, the high sensitivity of the KECK telescope and the HST has allowed numerous observations of the Be and B abundances in halo stars having very low metallicities, down to , i.e. (e.g. Molaro, et al., 1997; Duncan, et al., 1997; Garcia-Lopez, et al., 1998). These observations show a clear proportionality between the Be, B and Fe abundances. Considering that the light elements are secondary nuclei synthesized by spallation from C and O nuclei (Reeves, et al., 1970), it had been expected instead that their abundance would increase as the square of the metallicity (Vangioni-Flam, et al., 1990). This secondary behavior follows directly from the standard Galactic Cosmic Ray Nucleosynthesis (GCRN) scenario (Meneguzzi, et al., 1971), in which most of the Be and B are produced by energetic protons and particles interacting with C and O nuclei accumulated in the ISM (direct spallation).
The most natural way to account for the unexpected constancy of the Be/Fe and B/Fe ratios is to assume that Be and B nuclei are mainly produced by reverse spallation, i.e. by energetic C and O nuclei accelerated shortly after their release into the interstellar medium (ISM), and interacting with ambient H and He nuclei (Duncan, et al., 1992; Cassé, et al., 1995). This makes the production rates independent of the ambient metallicity, and the amount of light elements (L elements) in the Galaxy therefore increases jointly with the most abundant metals (M elements), namely C, O and Fe. This behavior is refered to as primary, and has been shown to follow naturally from the assumption that most of the supernova (SN) explosions occur in OB associations. This is the heart of the so-called superbubble models, whose main lines we recall in Sect. 2.
Some recent observations have suggested that [O/Fe] continues to decrease at the lowest metallicities, rather than reaching a "plateau" value which is the same for all metal-poor stars. These observations are still controversial (Fulbright & Kraft, 1999), although they could be accounted for by allowing for different `mixing times' for the freshly ejected Fe and O nuclei in the ISM (Ramaty, et al., 1999). They have raised the question whether a primary process for Be and B production in the early Galaxy is still needed (Fields & Olive, 1999). According to both energetics considerations (Ramaty, et al., 1999; Parizot & Drury, 2000) and a detailed analysis of the available data (Fields, et al., 2000), it is now widely agreed that the answer is yes, at least at a metallicity lower than a so-called transition metallicity, , say for . In this paper, we concentrate on the very early Galaxy, when the primary behavior dominates, and discuss the implications of the superbubble model for the distribution of the (light elements)/(metals) ratios (namely Be/O and B/O, or L/M for short) in very metal-poor stars.
© European Southern Observatory (ESO) 2000
Online publication: April 17, 2000