Toward Black Hole Stars: supermassive black hole growth in nuclear clusters via stellar-object and gas accretion

Konstantinos Kritos, Joseph Silk

First listed 2026-02-17 | Last updated 2026-03-17

Abstract

Supermassive black hole (SMBH) growth plausibly occurs via runaway astrophysical black hole mergers in nuclear star clusters that form intermediate mass black hole seeds at high redshifts. Such a model yields an order-of-magnitude higher rate of tidal disruption events than that of compact-object captures. Our prediction, normalized to our proposed resolution of SMBH seeding, yields detectable tidal disruption event rates at high redshift. The resulting dense gas cocoons generate compact galactic nuclei, each incorporating a central, massive, black hole star, with comparable masses in gas, stars, and massive black holes within a scale of around a parsec as inferred from the various Little Red Dot spectral signatures.

Short digest

Models SMBH growth in nuclear star clusters where runaway BH mergers seed IMBHs and continued stellar-object plus gas accretion builds compact, parsec-scale nuclei linked to Little Red Dots. Monte Carlo loss-cone calculations show MS-star TDEs dominate early feeding while stellar-mass BH plunges take over later, yielding a TDE rate about an order of magnitude above compact-object captures. The framework predicts detectable high‑z MS TDEs with peak luminosity–timescale tracks overlapping ENTs/ANTs, and a low‑frequency GW background with occasional BH‑EMRIs. Dense gas cocoons produce nuclei with comparable masses in gas, stars, and central “black hole star,” matching LRD spectral inferences.

Key figures to inspect

• Figure 1: Trace how the mass accretion budget shifts from MS TDEs to BH captures with time and SMBH mass; use the ten-realization error bars and the duty-cycle inset to gauge flare visibility windows for MS vs giants.
• Figure 2: Compare model MS‑TDE peak luminosities and decay times to the ENT/ANT regions and Eddington lines; identify which stellar masses and evolutionary epochs place events in the extreme‑flare locus and note the late‑time turnover as turnoff masses drop.
• Figure 3: Read off the GW energy density from BH/NS/WD captures versus frequency and compare to LISA/LGWA sensitivity curves to assess detectability; use the inset strains to connect individual capture channels to the stochastic background and contrast with light/heavy MBH‑seed merger backgrounds.