Accretion of sub-stellar companions as the origin of chemical abundance inhomogeneities in globular clusters

TL;DR

This paper proposes a novel mechanism where stars in globular clusters accrete sub-stellar companions that merge with them, explaining observed chemical abundance inhomogeneities and multiple populations.

Contribution

It introduces a new model of late pollutant delivery via accretion and merger of sub-stellar companions, expanding understanding of chemical variations in globular clusters.

Findings

Model can reproduce observed abundance variations using AGB and binary ejecta.

Explains correlation of pollution with cluster properties and location.

Requires further investigation into key dynamical processes.

Abstract

Globular clusters exhibit abundance variations, defining `multiple populations', which have prompted a protracted search for their origin. Properties requiring explanation include: the high fraction of polluted stars ($\sim 40{-}90$~percent, correlated with cluster mass), the absence of pollution in young clusters and the lower pollution rate with binarity and distance from the cluster centre. We present a novel mechanism for late delivery of pollutants into stars via accretion of sub-stellar companions. In this scenario, stars move through a medium polluted with AGB and massive star ejecta, accreting material to produce companions with typical mass ratio $q\sim 0.1$. These companions undergo eccentricity excitation due to dynamical perturbations by passing stars, culminating in a merger with their host star. The accretion of the companion alters surface abundances via injected pollutant. Alongside other self-enrichment models, the companion accretion model can explain the dilution of pollutant and correlation with intra-cluster location. The model also explains the ubiquity and discreteness of the populations and correlations of enrichment rates with cluster mass, cluster age and stellar binarity. Abundance variations in some clusters can be broadly reproduced using AGB and massive binary ejecta abundances from the literature. In other clusters, some high companion mass ratios ($q\gtrsim 1$) are required. In these cases, the available mass budget necessitates a variable degree of mixing of the polluted material with the primary star, deviations from model ejecta abundances or mixing of internal burning products. We highlight the avenues of further investigation which are required to explore some of the key processes invoked in this model.