S. Vauclair et C. Charbonnel, INFLUENCE OF A STELLAR WIND ON THE LITHIUM DEPLETION IN HALO STARS - A NEW STEP TOWARDS THE LITHIUM PRIMORDIAL ABUNDANCE, Astronomy and astrophysics, 295(3), 1995, pp. 715-724
Since the first observations by Spite & Spite (1982) of lithium in hal
o stars, with an abundance one order of magnitude smaller than the abu
ndance derived for young stars, the question has often been addressed
of the primordial lithium abundance. Has lithium been destroyed in hal
o stars, or is the presently observed lithium abundance representative
of the primordial value? If no macroscopic motions occur below the ou
ter convection zone in these stars, microscopic diffusion (gravitation
al and thermal settling) must take place and deplete lithium in a way
which is in contradiction with the observations. Rotation-induced turb
ulence has been invoked in previous papers to prevent microscopic diff
usion. Such a turbulence leads in turn to lithium depletion due to the
mixing of the layers below the convection zone down to the regions wh
ere lithium is destroyed by nuclear reactions. This effect has been in
voked to suggest that the primordial lithium abundance could be that o
f young stars, lithium being depleted by one order of magnitude in hal
o stars. It is however difficult in this framework to account for the
flatness and the small dispersion of the lithium abundances in the ''p
lateau''. In the present paper we introduce a new physical process whi
ch can prevent microscopic diffusion without leading to nuclear destru
ction: small stellar winds. We show that winds of the order of those o
bserved or indirectly derived for PopI stars can lead in PopII stars t
o lithium abundance values in very good agreement with the most recent
observations: the small positive slope presently observed for the lit
hium plateau is reproduced; the dispersion can be due to a range of ma
ss loss rates; the stars observed with no lithium in their spectra can
be those which suffer the largest rates; the light isotope Li-6 is pr
eserved. In this framework, the primordial abundance should be obtaine
d from the upper values of the lithium abundances in the plateau, and
not from their average value. This lead to a primordial abundance of 2
.5 +/- 0.1.