Concurrent formation of supermassive stars and globular clusters: implications for early self-enrichment

TL;DR

This paper proposes a model where globular clusters and supermassive stars form simultaneously in high-density environments, explaining observed chemical anomalies through continuous mixing of processed stellar material.

Contribution

It introduces a novel scenario of concurrent GC and SMS formation, addressing the mass budget problem and chemical enrichment in GCs.

Findings

Superlinear relation between processed material and cluster mass.

Continuous SMS wind mixing explains chemical anomalies.

Formation conditions resemble early Universe environments.

Abstract

We present a model for the concurrent formation of globular clusters (GCs) and supermassive stars (SMSs, $>10^3\,{\rm M}_\odot$) to address the origin of the HeCNONaMgAl abundance anomalies in GCs. GCs form in converging gas flows and accumulate low-angular momentum gas, which accretes onto protostars. This leads to an adiabatic contraction of the cluster and an increase of the stellar collision rate. A SMS can form via runaway collisions if the cluster reaches sufficiently high density before two-body relaxation halts the contraction. This condition is met if the number of stars $\gtrsim10^6$ and the gas accretion rate $\gtrsim10^5\,{\rm M}_\odot$/Myr, reminiscent of GC formation in high gas-density environments, such as -- but not restricted to -- the early Universe. The strong SMS wind mixes with the inflowing pristine gas, such that the protostars accrete diluted hot-hydrogen burning yields of the SMS. Because of continuous rejuvenation, the amount of processed material liberated by the SMS can be an order of magnitude higher than its maximum mass. This `conveyor-belt' production of hot-hydrogen burning products provides a solution to the mass budget problem that plagues other scenarios. Additionally, the liberated material is mildly enriched in helium and relatively rich in other hot-hydrogen burning products, in agreement with abundances of GCs today. Finally, we find a super-linear scaling between the amount of processed material and cluster mass, providing an explanation for the observed increase of the fraction of processed material with GC mass. We discuss open questions of this new GC enrichment scenario and propose observational tests.