Asymptotic Giant Branch stars at low metallicity: the challenging interplay between mass loss and molecular opacities

Abstract

We investigate the main physical properties of low-metallicity Asymptotic Giant Branch stars, with the aim of quantifying the uncertainties that presently affect the predicted chemical yields of these stars, associated to mass loss and description of molecular opacities. We find that above a threshold mass, M ~ 3.5Msun for Z=0.001, the results are little dependent on the opacity treatment, as long as hot-bottom burning prevents the surface C/O ratio from exceeding unity; the yields of these massive AGB stars are expected to be mostly determined by the efficiency of convection, with a relatively mild dependence on the mass-loss description. A much higher degree of uncertainty is associated to the yields of less massive models, which critically depend on the adopted molecular opacities. An interval of masses exists, say 2.0-3.0Msun, (the exact range depends on mass loss), in which HBB may be even extinguished following the cooling produced by the opacity of C-bearing molecules. The yields of these stars are the most uncertain, the variation range being the largest (up to ~ 2dex) for the nitrogen and sodium yields. For very low-mass models, not experiencing hot-bottom burning (M< 1.5Msun),the description of mass loss and the treatment of the convective boundaries are crucial for the occurrence of the third dredge-up, with sizable consequences on the CNO yields.