Rb. Hanson et al., ON THE USE OF [NA FE] AND [ALPHA/FE] RATIOS AND HIPPARCOS-BASED (U, V, W) VELOCITIES AS AGE INDICATORS AMONG LOW-METALLICITY HALO FIELD GIANTS/, The Astronomical journal (New York), 116(3), 1998, pp. 1286-1294
We have examined the [Na/Fe] and [Mg/Fe] ratios in a sample of 68 held
halo giants with -3 less than or similar to [Fe/H] less than or simil
ar to -1. We recalculated the Galactic (U, V, W) velocity components f
or these stars, using Hipparcos proper motions and a new Hipparcos-bas
ed distance scale. We used these data to see how the abundance ratios
may relate to kinematical substructure in the Galactic halo. To isolat
e a set of true halo stars, we eliminated metal-weak thick-disk stars,
about 10% of our sample. The field halo giants show the expected corr
elation of Na and Mg abundances, so we can use Na as a surrogate for M
g and the alpha-elements. The most metal-poor stars show a wider dispe
rsion of [Na/Fe] ratios than do the less metal-poor stars; the differe
nce is most striking for stars on retrograde galactic orbits. Some 20%
of our retrograde giants and 13% of all our halo giants have [Na/Fe]
less than or equal to -0.35 and may be significantly younger than the
oldest halo objects. Halo giants considered ''young'' by this Na abund
ance criterion show a preference for retrograde orbits. Giants in some
globular clusters (e.g., M13) do not exhibit the Mg versus Na correla
tion found among halo held giants. Instead, they have very large [Na/F
e] ratios and widely scattered [Mg/Fe] ratios, probably induced by dee
p mixing, which field halo giants apparently do not experience.