Constraints of the physics of low-mass AGB stars from CH and CEMP stars

TL;DR

This study uses homogeneous elemental abundance data from CH and CEMP stars to constrain the physics of low-mass AGB stars, revealing limitations of current models and suggesting the need for additional processes like the i-process.

Contribution

It provides the largest homogeneous abundance dataset for CH stars and compares observations with AGB nucleosynthesis models, highlighting discrepancies and the potential role of the i-process.

Findings

Standard models fit only part of the observed data.

Rotation and convective boundary treatments improve model fits at certain metallicities.

Observed high [hs/ls] ratios in CEMP stars suggest additional nucleosynthesis processes.

Abstract

We analyze a set of published elemental abundances from a sample of CH stars which are based on high resolution spectral analysis of ELODIE and SUBARU/HDS spectra. All the elemental abundances were derived from local thermodynamic equilibrium analysis usingmodel atmospheres, and thus, they represent the largest homogeneous abundance data available for CH stars up to date. For this reason, we can use the set to constrain the physics and the nucleosynthesis occurring in low mass AGB stars. CH stars have been polluted in the past from an already extinct AGB companion and thus show s-process enriched surfaces. We discuss the effects induced on the surface AGB s-process distributions by different prescriptions for convection and rotation. Our reference theoretical FRUITY set fits only part of the observations. Moreover, the s-process observational spread for a fixed metallicity cannot be reproduced. At Fe/H]>-1, a good fit is found when rotation and a different treatment of the inner border of the convective envelope are simultaneously taken into account. In order to increase the statistics at low metallicities, we include in our analysis a selected number of CEMP stars and, therefore, we compute additional AGB models down to [Fe/H]=-2.85. Our theoretical models are unable to attain the large [hs/ls] ratios characterizing the surfaces of those objects. We speculate on the reasons for such a discrepancy, discussing the possibility that the observed distribution is a result of a proton mixing episode leading to a very high neutron density (the so-called i-process)