Testing spallation processes with beryllium and boron

The nucleosynthesis of Be and B by spallation processes provides unique insight into the origin of cosmic rays. Namely, different spallation schemes predict sharply different trends for the growth of LiBeB abundances with respect to oxygen. "Primary" mechanisms predict BeB proportional to O and are well motivated by the data if O/Fe is constant at low metallicity. In contrast, "secondary" mechanisms predict BeB proportional to O-2 and are consistent with the data if O/Fe increases toward low metallicity as some recent data suggest. Clearly, any primary mechanism, if operative, will dominate early in the history of the Galaxy. In this paper, we fit the BeB data to a two-component scheme which includes both primary and secondary trends. In this way, the data can be used to probe the period in which primary mechanisms are effective. We analyze the data using consistent stellar atmospheric parameters based on Balmer line data and the continuum infrared flux. Results depend sensitively on Population II O abundances (and O/Fe trends), which have recently seen renewed interest. We explore the implications of these results phenomenologically, using a systematic and consistent compilation and fitting of BeBOFe data. Two-component Be-O fits indicate that primary and secondary components contribute equally at [O/H](eq) = -1.8 for Balmer line data; and [O/H](eq) = -1.4 to -1.8 for IRFM. We apply these constraints to recent models for LiBeB origin. The Balmer line data do not show any evidence for primary production. On the other hand, the IRFM data do indicate a preference for a two-component model, such as a combination of standard GCR and metal-enriched particles accelerated in superbubbles. These conclusions rely on a detailed understanding of the abundance data including systematic effects which may alter the derived O-Fe and BeB-Fe relations.