Supergiants and their shells in young globular clusters

TL;DR

This paper proposes a scenario where second-generation stars in young globular clusters form in shells around cool supergiants, explaining observed abundance anomalies through simulations of massive star evolution and star formation in these shells.

Contribution

It introduces a novel formation mechanism for second-generation stars in globular clusters via shell formation around supergiants, supported by detailed stellar evolution simulations.

Findings

Shells around supergiants are gravitationally unstable and can form new stars.

The composition of shell-stars matches observed abundance anomalies.

A top-heavy initial mass function or truncated mass distribution can resolve the mass-budget issue.

Abstract

Anomalous surface abundances are observed in a fraction of the low-mass stars of Galactic globular clusters, that may originate from hot-hydrogen-burning products ejected by a previous generation of massive stars. We present and investigate a scenario in which the second generation of polluted low-mass stars can form in shells around cool supergiant stars within a young globular cluster. Simulations of low-Z massive stars (M$_{\rm i}\sim$ 150$-$600 M$_{\odot}$) show that both core-hydrogen-burning cool supergiants and hot ionizing stellar sources are expected to be present simulaneously in young globular clusters. Under these conditions, photoionization-confined shells form around the supergiants. We find that the shell is gravitationally unstable on a timescale that is shorter than the lifetime of the supergiant, and the Bonnor-Ebert mass of the overdense regions is low enough to allow star formation. Since the low-mass stellar generation formed in this shell is made up of the material lost from the supergiant, its composition necessarily reflects the composition of the supergiant wind. We show that the wind contains hot-hydrogen-burning products, and that the shell-stars therefore have very similar abundance anomalies that are observed in the second generation stars of globular clusters. Considering the mass-budget required for the second generation star-formation, we offer two solutions. Either a top-heavy initial mass function is needed with an index of $-$1.71..$-$2.07. Alternatively, we suggest the shell-stars to have a truncated mass distribution, and solve the mass budget problem by justifiably accounting for only a fraction of the first generation. Even without forming a photoionizaton-confined shell, the cool supergiant stars predicted at low-Z could contribute to the pollution of the interstellar medium of the cluster from which the second generation was born.