# Globular Cluster Formation from Colliding Substructure

## TL;DR

This paper proposes a novel model where globular clusters form from high-speed collisions of infalling subhalos during galaxy assembly, leading to observable properties consistent with real globular cluster populations.

## Contribution

It introduces a new formation mechanism for globular clusters triggered by colliding subhalos, supported by simulations and theoretical analysis.

## Key findings

- Produces a linear GC-halo mass relation matching observations
- Generates old globular populations with realistic age distributions
- Results in a bimodal metallicity distribution similar to observed galaxies

## Abstract

We investigate a scenario where the formation of Globular Clusters (GCs) is triggered by high-speed collisions between infalling atomic-cooling subhalos during the assembly of the main galaxy host, a special dynamical mode of star formation that operates at high gas pressures and is intimately tied to LCDM hierarchical galaxy assembly. The proposed mechanism would give origin to "naked" globulars, as colliding dark matter subhalos and their stars will simply pass through one another while the warm gas within them clashes at highly supersonic speed and decouples from the collisionless component, in a process reminiscent of the Bullet galaxy cluster. We find that the resulting shock-compressed layer cools on a timescale that is typically shorter than the crossing time, first by atomic line emission and then via fine-structure metal-line emission, and is subject to gravitational instability and fragmentation. Through a combination of kinetic theory approximation and high-resolution $N$-body simulations, we show that this model may produce: (a) a GC number-halo mass relation that is linear down to dwarf galaxy scales and agrees with the trend observed over five orders of magnitude in galaxy mass; (b) a population of old globulars with a median age of 12 Gyr and an age spread similar to that observed; (c) a spatial distribution that is biased relative to the overall mass profile of the host; and (d) a bimodal metallicity distribution with a spread similar to that observed in massive galaxies.