Post-AGB stars in the SMC as tracers of stellar evolution: the extreme s-process enrichment of the 21 micrometer star J004441.04-732136.4

TL;DR

This study examines the unique s-process enrichment in the post-AGB star J004441.04-732136.4 in the SMC, revealing discrepancies between observed chemical abundances and theoretical nucleosynthesis models, challenging current stellar evolution theories.

Contribution

It provides detailed abundance analysis of a rare 21 micrometer post-AGB star in the SMC and compares observations with nucleosynthesis predictions, highlighting significant model-observation discrepancies.

Findings

J004441.04-732136.4 is highly s-process enriched.

Observed chemical abundances do not match model predictions.

The star's pulsational properties are difficult to explain with current models.

Abstract

By virtue of their spectral types, favourable bolometric corrections as well as their constrained distances, post-AGB stars in external galaxies offer unprecedented tests to AGB nucleosynthesis and dredge-up predictions. We focus here on one object J004441.04-732136.4, which is the only known 21 micrometer source of the SMC. Spectral abundance results reveal J004441.04-732136.4 to be one of the most s-process enriched objects found up to date, while the photospheric C/O ratio of 1.9 +- 0.7, shows the star is only modestly C-rich. J004441.04-732136.4 also displays a low [Fe/H] = -1.34 +- 0.32, which is significantly lower than the mean metallicity of the SMC. From the SED, a luminosity of 7600 +- 200 solar luminosities is found, together with E(B-V) = 0.64 +- 0.02. According to evolutionary post-AGB tracks, the initial mass should be approximately 1.3 solar masses. The photometric variability shows a clear period of 97.6 +- 0.3 days. The detected C/O as well as the high s-process overabundances (e.g. [Y/Fe] = 2.15, [La/Fe] = 2.84) are hard to reconcile with the predictions. The chemical models also predict a high Pb abundance, which is not compatible with the detected spectrum, and a very high 12C/13C, which is not yet constrained by observations. The predictions are only marginally dependent on the evolution codes used. We show that our theoretical predictions match the s-process distribution, but fail in reproducing the detected high overabundances and predict a high Pb abundance which is not detected. Additionally, there remain serious problems in explaining the observed pulsational properties of this source.