Behavior of [S/Fe] in Very Metal-Poor Stars from the S I 1.046 micron Lines Revisited

TL;DR

This study revisits sulfur-to-iron ratios in very metal-poor stars using near-infrared spectral lines, finding a moderate and consistent [S/Fe] ratio around +0.2 to +0.5, challenging previous claims of higher ratios.

Contribution

It provides a non-LTE analysis of S I lines at 1.046 microns in very metal-poor stars, clarifying the behavior of [S/Fe] at low metallicities and correcting earlier overestimations.

Findings

[S/Fe] is moderately supersolar (~+0.2 to +0.5) across the studied metallicity range.

The flat [S/Fe] trend contradicts earlier suggestions of higher ratios (~+0.8) at [Fe/H] ~ -3.

Previous overestimations of [S/Fe] in some metal-poor stars are attributed to observational or analytical issues.

Abstract

With an aim of establishing how the [S/Fe] ratios behave at the very low metallicity regime down to [Fe/H] ~ -3, we conducted a non-LTE analysis of near-IR S ~ {\sc i} triplet lines (multiplet 3) at 10455-10459 A for a dozen of very metal-poor stars (-3.2 <[Fe/H] < -1.9) based on the new observational data obtained with IRCS+AO188 of the Subaru Telescope. It turned out that the resulting [S/Fe] values are only moderately supersolar at [S/Fe] ~ +0.2-0.5 irrespective of the metallicity. While this "flat" tendency is consistent with the trend recently corroborated by Spite et al. (2011, A&A, 528, A9) based on the S I 9212/9228/9237 lines (multiplet 1), it disaffirms the possibility of conspicuously large [S/Fe] (up to ~+0.8) at [Fe/H] ~ -3 that we once suggested in our first report on the S abundances of disk/halo stars using S I 10455-10459 lines (Takeda & Takada-Hidai 2011, PASJ, 63, S537). Given these new observational facts, we withdraw our previous argument, since we consider that [S/Fe]'s of some most metal-poor objects were overestimated in that paper; the likely cause for this failure is also discussed.