2004ApJ...606..306B

Sixteen Population I solar-type dwarfs have been selected to ascertain the baseline B abundance in the Galactic disk for a range of a factor of 4 in metallicity: from [Fe/H] of -0.5 to +0.1. All the stars selected are undepleted in Be, which ensures that they have also retained their full initial abundance of B. Evaluation of the trend of B with Fe provides a means to study the evolution of B in the Galactic disk. We observed 16 bright stars around the B I 2497 Å line, using the STIS echelle spectrograph on HST. New observations of Li and Be in some stars were made, and previous abundance studies of Li and Be in these stars were reevaluated using revised parameters and a modified spectral synthesis code for consistency with the B measurements. Abundances of B were calculated by spectrum synthesis with the revised MOOG code, which accounts for the increased opacity in the UV due to metals; the LTE B abundances were then corrected for non-LTE effects. Four additional stars with undepleted Be have HST B observations, which increase our sample to 20. For these disk stars there is a shallow slope for B versus Fe and Be versus Fe, such that as Fe increases by a factor of 4, B and Be increase by 1.7 times. The slope for B_LTEversus Fe is 0.31±0.09, for B_NLTE versus Fe 0.40±0.12, and for Be versus Fe 0.38±0.14. We have estimated the effect of additional UV opacity from Mg and find that an increase of 0.3 dex in Mg results in a higher B abundances by 0.1 dex for all the disk stars. Individual stars are not consistently above (or below) the mean in both B and Be, implying that the star-to-star differences are not due to variations in the elemental content of the ``natal'' clouds. We find that the trend of B abundance with [Fe/H] is consistent with the general trend observed in halo stars. If we connect the halo and disk stars, then an increase in the Fe abundance by 10³ is accompanied by increases of 100 times in B and 550 times in Be. However, fitting two separate relations for the disk and the halo stars results in a somewhat steeper slope for Be for the halo stars (1.08±0.07) relative to the disk stars (0.38±0.14). This is the case for B also in LTE, with B_halo(0.90±0.07) versus B_disk(0.32±0.12). However, the NLTE B abundance increases more slowly for halo stars than the Be abundance does; since this is not predicted by light-element synthesis or depletion, we suggest that a full NLTE analysis would be preferable to making the (small) corrections to the LTE abundances. Some of the lowest metallicity stars are thought to have only upper limits on the B abundance; if that is the case, the NLTE B slope is steeper, nearing 1.0. The abundance of B in the disk stars is observed to be a factor of ∼15⁺⁷_–5more than the abundance of Be in these stars, a result consistent with the predictions of Galactic cosmic-ray (GCR) spallation, B/Be=15±5. The upper envelope for Li versus Fe yields Li/B and Li/Be ratios that, when coupled with models and predictions, indicate that 20%-45% of Li might be produced by GCRs. While there is no evidence to support the production of B by neutrino spallation, we cannot rule it out.