Formation of massive protostars in atomic cooling haloes

TL;DR

This paper presents a high-resolution simulation of atomic cooling halo collapse, showing formation of massive protostars and potential binary systems, highlighting pathways for early massive black hole seed formation.

Contribution

It introduces the highest-resolution 3D simulation of atomic cooling halo collapse, revealing detailed protostar formation, disc fragmentation, and potential binary evolution in the early Universe.

Findings

Protostar initial mass ~0.1 solar masses.

Disc fragments into 5-10 protostars.

Potential formation of wide binary systems.

Abstract

We present the highest-resolution three-dimensional simulation to date of the collapse of an atomic cooling halo in the early Universe. We use the moving-mesh code arepo with the primordial chemistry module introduced in Greif (2014), which evolves the chemical and thermal rate equations for over more than 20 orders of magnitude in density. Molecular hydrogen cooling is suppressed by a strong Lyman-Werner background, which facilitates the near-isothermal collapse of the gas at a temperature of about $10^4\,$K. Once the central gas cloud becomes optically thick to continuum emission, it settles into a Keplerian disc around the primary protostar. The initial mass of the protostar is about $0.1\,{\rm M}_\odot$, which is an order of magnitude higher than in minihaloes that cool via molecular hydrogen. The high accretion rate and efficient cooling of the gas catalyse the fragmentation of the disc into a small protostellar system with 5-10 members. After about 12 yr, strong gravitational interactions disrupt the disc and temporarily eject the primary protostar from the centre of the cloud. By the end of the simulation, a secondary clump has collapsed at a distance of $\simeq 150\,$au from the primary clump. If this clump undergoes a similar evolution as the first, the central gas cloud may evolve into a wide binary system. High accretion rates of both the primary and secondary clumps suggest that fragmentation is not a significant barrier for forming at least one massive black hole seed.