The revival of galactic cosmic-ray nucleosynthesis?

Because of the roughly linear correlation between Be/H and Fe/H in low-metallicity halo stars, it has been argued that a "primary" component in the nucleosynthesis of Be must be present in addition to the "secondary" component from standard Galactic cosmic-ray nucleosynthesis. In this paper we critically reevaluate the evidence for the primary versus secondary character of Li, Be, and B (LiBeB) evolution, analyzing both the observations and Galactic chemical evolution models. Although it appears that [Be/H] versus [Fe/H] has a logarithmic slope near 1, it is rather the Be-O trend that directly arises from the physics of spallation production. Using new abundances for oxygen in halo stars based on UV OH lines, we find that in Population II stars for which O has been measured, the Be-O slope has a large uncertainty due to systematic effects. Namely, the Be-O logarithmic slope lies in the range 1.3-1.8, rendering it difficult to distinguish from the data between the secondary slope of 2 and the primary slope of 1. The possible difference between the Be-Fe and Be-O slopes is a consequence of the variation in O/Fe versus Fe: recent data suggest that the best-fit O/Fe-Fe slope for Population II is in the range -0.5 to -0.2, rather than zero (i.e., Fe proportional to O) as is often assumed. In addition to this phenomenological analysis of Be and B evolution, we have also examined the predicted LiBeB, O, and Fe trends in Galactic chemical evolution models that include outflow. Based on our results, it is possible that a good fit to the LiBeB evolution requires only the traditional Galactic cosmic-ray spallation and the (primary) neutrino-process contribution to B-11. We thus suggest that these two processes might be sufficient to explain Li-6, Be, and B evolution in the Galaxy, without the need for an additional primary source of Be and B. However, the uncertainties in the data at this time prevent one from reaching a definitive conclusion. Fortunately, several observational tests of this "neoclassical" scenario are available; we note in particular the importance of further observations to secure the O/Fe Population II trend, as well as accurate measurements of B/Be, Li-6/Be, and B-11/B-10 in halo stars.