General-relativistic instability in hylotropic supermassive stars

TL;DR

This paper develops analytical models to accurately predict the onset of general-relativistic instability in supermassive stars, crucial for understanding black hole seed formation via direct collapse.

Contribution

It introduces hylotropic models that precisely determine the conditions for GR instability in supermassive stars, linking core mass and total mass to collapse outcomes.

Findings

GR instability occurs at total mass >10^5 Msun with core mass >10^4 Msun.

Lower core mass leads to larger total mass needed for instability.

Supermassive stars in primordial haloes collapse below 5×10^5 Msun.

Abstract

The formation of supermassive black holes by direct collapse would imply the existence of supermassive stars (SMSs) and their collapse through the general-relativistic (GR) instability into massive black hole seeds. However, the final mass of SMSs is weakly constrained by existing models, in spite of the importance this quantity plays in the consistency of the direct collapse scenario. We estimate the final masses of spherical SMSs in the whole parameter space relevant for these objects. We build analytical stellar structures (hylotropes) that mimic existing numerical SMS models accounting for full stellar evolution with rapid accretion. From these hydrostatic structures, we determine ab initio the conditions for GR instability, and compare the results with the predictions of full stellar evolution. We show that hylotropic models predict the onset of GR instability with high precision. The mass of the convective core appears as a decisive quantity. The lower it is, the larger is the total mass required for GR instability. Typical conditions for GR instability are a total mass >10^5 Msun with a core mass >10^4 Msun. If the core mass remains below 10^4 Msun, total masses in excess of 10^6-10^7 Msun can be reached. Our results confirm that spherical SMSs forming in primordial, atomically cooled haloes collapse at masses below 500 000 Msun. On the other hand, accretion rates in excess of 1000 Msun/yr, leading to final stellar masses >10^6 Msun, are required for massive black hole formation in metal-rich gas. Thus, the different channels of direct collapse imply distinct final masses for the progenitor of the black hole seed.