LITHIUM IN HALO STARS FROM STANDARD STELLAR EVOLUTION

We have constructed a grid of theoretical evolutionary sequences of models for low-metallicity stars (0.0001 ≤ Z ≤ 0.001) from the pre-main-sequence phase to the giant branch. These continuous tracks are especially designed to analyze the history of surface Li abundance during their evolution. We examine several mechanisms that could alter the surface Li abundance of stars at different phases of evolution. In particular, we consider nuclear burning at the base of the convection zone (occurring primarily during the pre-main sequence and early main sequence), diffusion by gravitational settling (main sequence and turnoff), convective dredge-up (early post-main sequence), and dilution (post-main sequence). The relative importance of each process is very sensitive to stellar mass, chemical composition, and ratio of mixing length to pressure scale height, α. To apply these models to a detailed analysis of the available data on Li abundances measured in halo stars, we construct Li isochrones for different ages, values of α, and compositions. We show that when particular care is taken to preserve numerical accuracy, the standard theory of stellar evolution agrees closely with the observations of extreme halo stars that satisfy simultaneously the criteria of low metallicity ([Fe/H] ≤ -1.3) and high space velocity (VLSR ≥ 160 km s-1). The best agreement between these stars and our models is for ages in the range 16-20 Gyr and α= 1.5, in good general agreement with globular cluster data. Lithium isochrones with these parameter values simultaneously reproduce the (depleted) main-sequence cool star observations, the (essentially undepleted) main-sequence Spite Li plateau, and the (depleted and undepleted) subgiant observations. Thus, our Li isochrones support the conclusion of Spite and Spite that the level of the Li plateau essentially reflects that of the initial Galactic Li abundance. Using χ2 fitting at the 2 σ confidence level of our Li isochrones with parameters in the ranges 10 Gyr ≤ age ≤ 20 Gyr, 1.1 ≤ α ≤ 1.5, 0.0001 ≤ Z ≤ 0.001, and allowing up to (0.2,0.4) dex uncertainty in the Li abundance of (plateau, cool) stars and more uncertainty for upper limits, we derive for the pre-Galactic Li abundance 12 + log(NLi/NH) = 2.17-0.13+0.04 for standard Li isochrones, and 2.31-0.17+0.05 for Li isochrones that include diffusion, leading to an overall range of 2.04-2.36. Uncertainties in the 4He abundance, opacities, and nuclear destruction rate of 7Li do not affect this range. This range agrees with the prediction of and reinforces the self-consistency of standard big bang nucleosynthesis. We have argued that mass loss is not likely to have been a cause for significant Li depletion in Population II stars. The principal remaining theoretical uncertainty resides in the role of rotation and possible interaction of diffusion with rotationally induced motions. Isochrones for 6Li, also of cosmological importance, are presented.