Beyond the Goldilocks Zone: Identifying Critical Features in Massive Black Hole Formation

Summary

Abstract

Mone and collaborators use statistical and machine-learning methods to determine which properties are most useful for identifying halos that may form massive black-hole seeds. They compare 35 direct-collapse candidates from the Renaissance Simulations with about 4,000 other atomic-cooling halos. The strongest differences are found in the halos’ central density and inward gas flow, while environmental properties such as Lyman-Werner radiation and distance to the nearest galaxy are less important.

1. Introduction

• Direct collapse may produce heavy black-hole seeds that can grow into the supermassive black holes observed in the early universe.
• Earlier models often placed candidates in a narrow “Goldilocks zone” close enough to a galaxy for radiation but far enough away to avoid metals.
• This paper tests whether internal halo properties provide better clues than the surrounding environment.

2. Methods

• The sample contains 35 metal-poor, starless candidate halos and about 4,000 non-candidates with similar total masses.
• The authors examine 18 central, overall halo, and environmental properties using statistical tests and machine-learning rankings.
• The tested properties include density, gas inflow, temperature, growth rate, rotation, metallicity, radiation, and distance to nearby galaxies.

3. Results

• Candidate halos have denser cores and stronger inward gas flows than non-candidates of similar total mass.
• After excluding the properties used to define candidates, density and radial mass influx are consistently ranked as the most useful features.
• Lyman-Werner radiation, tidal forces, overdensity, and distance to the nearest galaxy have comparatively little influence on candidacy.

4. Discussion

• A halo’s internal structure and growth history appear more important than whether it occupies one particular environment near another galaxy.
• Candidate halos grow by roughly a factor of 12 during their final 130 million years, while non-candidates approximately double in mass.
• These results could improve black-hole seeding models used in larger simulations that cannot resolve the full collapse process.

5. Conclusion

• Direct-collapse candidates form a statistically distinct population from other atomic-cooling halos.
• Future seeding models should focus on central density, gas inflow, and halo growth instead of relying mainly on total mass or radiation.
• Higher-resolution simulations are still needed to confirm which candidates actually produce massive stars or black holes.

Best Figures

Figure 1

Figure 4