Population II stars and the Spite plateau - Stellar evolution models with mass loss

Aims. We aim to determine the constraints that observed chemical abundances put on the potential role of mass loss in metal poor dwarfs. Methods. Self-consistent stellar evolutionary models that include all the effects of atomic diffusion and radiative accelerations for 28 chemical species were computed for stellar masses between 0.6 and 0.8M� . Models with an initial metallicity of Z0 = 0.00017 and mass loss rates from 10 −15 Myr −1 to 10 −12 Myr −1 were calculated. They were then compared to previous models with mass loss, as well as to models with turbulent mixing. Results. For models with an initial metallicity of (Fe/H)0 = −2.31, mass loss rates of about 10 −12 Myr −1 lead to surface abundance profiles that are very similar to those obtained in models with turbulence. Both models have about the same level of agreement with observations of galactic-halo lithium abundances, as well as lithium and other elemental abundances from metal poor globular clusters such as NGC 6397. In this cluster, models with mass loss agree slightly better with subgiant observations of Li abundance than those with turbulence. Lower red giant branch stars instead favor the models with turbulence. Larger differences between models with mass loss and those with turbulence appear in the interior concentrations of metals. Conclusions. The relatively high mass loss rates required to reproduce plateau-like lithium abundances appear unlikely when com- pared to the solar mass-loss rate. However the presence of a chromosphere on these stars justifies further investigation of the mass-loss rates.