A Model for Clumpy Self-Enrichment in Globular Clusters

TL;DR

This paper introduces a theoretical model explaining how globular clusters self-enrich through internal clump formation and supernova feedback, matching observed metallicity spreads and mass-metallicity relations.

Contribution

The model provides a new explanation for internal metallicity variations in GCs based on clump formation and supernova enrichment, aligning with observations.

Findings

Model reproduces observed mass-metallicity relation.

Predicts metallicity spreads consistent with Milky Way GCs.

Internal metallicity spread correlates with self-enrichment level.

Abstract

Detailed observations of globular clusters (GCs) have revealed evidence of self-enrichment: some of the heavy elements that we see in stars today were produced by cluster stars themselves. Moreover, GCs have internal subpopulations with different elemental abundances, including, in some cases, in elements such as iron that are produced by supernovae. This paper presents a theoretical model for GC formation motivated by observations of Milky Way star forming regions and simulations of star formation, where giant molecular clouds fragment into multiple clumps which undergo star formation at slightly different times. Core collapse supernovae from earlier-forming clumps can enrich later-forming clumps to the degree that the ejecta can be retained within the gravitational potential well, resulting in subpopulations with different total metallicities once the clumps merge to form the final cluster. The model matches the mass-metallicity relation seen in GC populations around massive elliptical galaxies, and predicts metallicity spreads within clusters in excellent agreement with those seen in Milky Way GCs, even for those whose internal abundance spreads are so large that their entire identity as a GC is in question. The internal metallicity spread serves as an excellent measurement of how much self-enrichment has occurred in a cluster, a result that is very robust to variation in the model parameters.