Giant branch mixing and the ultimate fate of primordial deuterium in the Galaxy

The observed cosmic abundances of light elements are most consistent with each other, and with the predictions of big bang nucleosynthesis, if, contrary to the usual assumption, galactic chemical evolution reduces (D = He-3)/H with time. Chemical evolution models which accomplish this require that low-mass stars destroy He-3 in the envelope gas that they return to the interstellar medium. A simple argument based on the rates of limiting nuclear reactions shows that the same giant branch mixing process which appears to be needed to explain the observed C-12/C-13 and C/N ratios in 1-2 solar mass stars would indeed also probably destroy He-3 by a large factor in the bulk of the envelope material. The conclusion is that Galactic He-3/H estimates should not be trusted for setting an upper limit on primordial (D = He-3)/H. This removes the strongest lower bound on the cosmic baryon density from big bang nucleosynthesis and the only argument for abundant baryonic dark matter.