Massive Seed Black Holes

• Massive seed black holes are early-universe objects with initial masses in the 10³–10⁶ M☉ range that serve as precursors to supermassive black holes.
• They form through various channels—such as direct collapse in atomic-cooling halos, runaway stellar mergers, and gas-rich major mergers—requiring low metallicity and high inflow rates.
• Rapid growth via super-Eddington accretion and regulated feedback in dense protogalactic environments enables these seeds to quickly evolve into quasars and influence galaxy evolution.

Massive seed black holes ("massive seeds") are black holes with initial masses $\gtrsim 10^3\,M_\odot$ formed in the early Universe. These objects represent a crucial phase in the emergence of supermassive black holes (SMBHs, $M_\bullet \gtrsim 10^6\,M_\odot$) detected as high-redshift quasars and in massive galaxy nuclei today. Understanding their formation channels, the conditions enabling their rapid early growth, and their demographic impact remains a central challenge in theoretical astrophysics.

1. Definitions, Mass Ranges, and Theoretical Context

Massive seed black holes are defined as black holes formed at high redshift ($z\sim10$–$30$) with initial masses typically in the range $M_{\rm seed}\sim10^3$–$10^6\,M_\odot$. This distinguishes them from "light seeds" ($M_{\rm seed}\sim10^1$–$10^3\,M_\odot$), which are typically remnants of Population III stars. Key mass ranges for massive seeds are:

• Direct-collapse objects: $M_{\rm seed}\sim10^4$–$10^6\,M_\odot$
• Runaway stellar collisions: $M_\bullet \gtrsim 10^6\,M_\odot$0–$M_\bullet \gtrsim 10^6\,M_\odot$1
• Dynamically heated or merger-assisted fragments: $M_\bullet \gtrsim 10^6\,M_\odot$2–$M_\bullet \gtrsim 10^6\,M_\odot$3 These seeds provide a much shorter logarithmic growth path than light seeds to reach $M_\bullet \gtrsim 10^6\,M_\odot$4 SMBHs by $M_\bullet \gtrsim 10^6\,M_\odot$5–$M_\bullet \gtrsim 10^6\,M_\odot$6 in the context of Eddington-limited or moderately super-Eddington fueling (Das et al., 30 Jan 2026, Regan et al., 2024, Volonteri et al., 2021, Wise, 2023).

Massive seeds are favored as progenitors of early quasars because the available cosmic time between seed formation and quasar assembly is typically $M_\bullet \gtrsim 10^6\,M_\odot$7 Gyr at $M_\bullet \gtrsim 10^6\,M_\odot$8, requiring either massive seeds (to reduce the required number of $M_\bullet \gtrsim 10^6\,M_\odot$9-foldings) or sustained periods of super-Eddington accretion (Wise, 2023, Latif et al., 2016).

2. Formation Channels and Physical Prerequisites

Several astrophysical pathways for massive seed formation are supported:

2.1 Direct Collapse in Atomic-Cooling Halos

• Mechanism: In metal-free halos ($z\sim10$0) above the atomic-cooling mass threshold ($z\sim10$1, $z\sim10$2 K), rapid isothermal inflow at $z\sim10$3–$z\sim10$4 K is triggered by suppression of H$z\sim10$5 cooling, usually by Lyman-Werner flux $z\sim10$6–$z\sim10$7.
• Outcome: Formation of a supermassive ($z\sim10$8–$z\sim10$9) protostar or quasi-star, undergoing general relativistic collapse or quasi-star–mediated black-hole formation (Wise, 2023, Regan et al., 2024, Lupi et al., 2021, Latif et al., 2016).
• Key conditions: Non-fragmenting, low-angular-momentum gas disks; avoidance of prior star formation; high inflow rates $30$0–$30$1; high-density environments or synchronized pairs in overdense regions (Volonteri et al., 2021).
• Seed masses: $30$2–$30$3 (Kimura et al., 14 Apr 2025, Bonoli et al., 2012).

2.2 Runaway Stellar Mergers and Dense Cluster Evolution

• Mechanism: Formation of dense nuclear star clusters ($30$4–$30$5, $30$6 pc), leading to core-collapse and runaway physical mergers among massive stars, producing a very massive star, which collapses to a black hole if $30$7.
• Post-collapse: Efficient dynamical loss-cone refilling (non-resonant and resonant relaxation) enables continued mass growth by swallowing cluster stars, reaching seed masses $30$8–$30$9 (Yajima et al., 2015, Schleicher et al., 2018, Schleicher et al., 2018, Boekholt et al., 2018).
• Requirement: Cluster core-collapse time $M_{\rm seed}\sim10^3$0 Myr (massive star lifetimes) (Yajima et al., 2015).
• Parameter dependence: Gas supply, accretion prescription, and dynamical environment set the efficiency and final seed mass.

2.3 Gas-rich Major Mergers (Direct-Collapse Mergers)

• Mechanism: Major mergers of gas-rich disk galaxies ($M_{\rm seed}\sim10^3$1) without pre-existing massive BHs drive intense ($M_{\rm seed}\sim10^3$2) multiscale inflow into nuclear disks ($M_{\rm seed}\sim10^3$3 pc) and central clouds ($M_{\rm seed}\sim10^3$4 pc), leading to rapid supermassive star (and then BH) formation (Bonoli et al., 2012).
• Seed masses: $M_{\rm seed}\sim10^3$5.
• Environmental preference: These events are common at $M_{\rm seed}\sim10^3$6 ($M_{\rm seed}\sim10^3$7\% of favorable mergers), decline to 20\% at $M_{\rm seed}\sim10^3$8.
• Clustering: Hosts typically reside in lower-than-average local galaxy density (Bonoli et al., 2012).

2.4 Collisional and Accretional Runaways in Metal-free Star Clusters

• Mechanism: In gas-rich Pop III clusters ($M_{\rm seed}\sim10^3$9–$10^6\,M_\odot$0, $10^6\,M_\odot$1 pc, $10^6\,M_\odot$2–$10^6\,M_\odot$3), high protostellar accretion rates inflate stellar radii, boosting collisional cross sections and resulting in runaway mergers forming central objects of $10^6\,M_\odot$4–$10^6\,M_\odot$5 (Boekholt et al., 2018, Schleicher et al., 2018).
• Advantage: Unlike atomic-direct-collapse, no extreme UV background or zero metallicity on global scales is required; only the assembly of compact, massive cluster progenitors.

2.5 Primordial Black Holes and Exotic Channels

• PBHs: Massive PBHs formed in the early Universe could seed galaxies, especially for $10^6\,M_\odot$6–$10^6\,M_\odot$7, although their abundance is constrained to $10^6\,M_\odot$8 by dark matter and galaxy statistics (Das et al., 30 Jan 2026, Volonteri et al., 2021).

3. Early Growth: Accretion Dynamics, Feedback, and Environmental Constraints

Early growth rates of massive seeds depend on local gas density, bulge assembly, and radiative or mechanical feedback:

• Super-Eddington Episodes: In protogalactic bulges with $10^6\,M_\odot$9, $M_{\rm seed}\sim10^1$0 and metal-poor gas, $M_{\rm seed}\sim10^1$1 seeds accrete at $M_{\rm seed}\sim10^1$2–$M_{\rm seed}\sim10^1$3, achieving a tenfold mass increase ($M_{\rm seed}\sim10^1$4 Myr) before feedback limits inflow (Inayoshi et al., 2021).
• Feedback Regulation: Ionizing and radiation pressure confine the accretion to phases where the ionized bubble remains trapped inside the bulge's core radius; outflows modulate but do not quench equatorial inflow.
• Halo Mass Threshold: Efficient rapid growth requires host halos $M_{\rm seed}\sim10^1$5 ($M_{\rm seed}\sim10^1$6 K) at $M_{\rm seed}\sim10^1$7–$M_{\rm seed}\sim10^1$8. Lighter halos cannot confine the ionized region, preventing sustained accretion above $M_{\rm seed}\sim10^1$9 (Inayoshi et al., 2021).
• Super-critically Accreting Stellar-mass Seeds: With efficient photon trapping and slim-disk solutions, even stellar-mass ($10^3\,M_\odot$0) seeds embedded in dense clumps can reach $10^3\,M_\odot$1 in $10^3\,M_\odot$2 yr, provided sufficient gas supply and low radiative efficiency $10^3\,M_\odot$3 (Lupi et al., 2015).

4. Demographics, Occupation Fractions, and Evolution

• Number Density: Typical models predict a massive-seed abundance $10^3\,M_\odot$4 at $10^3\,M_\odot$5 and $10^3\,M_\odot$6 for the most massive seeds ($10^3\,M_\odot$7) (Regan et al., 2024, Kimura et al., 14 Apr 2025).
• Occupation Fraction: By $10^3\,M_\odot$8, halos with $10^3\,M_\odot$9 virtually all host a massive seed, but the occurrence in dwarf or low-mass systems is limited by metallicity enrichment and star-formation feedback (Bellovary et al., 2011).
• Host Galaxy Properties: Massive-seed formation at high $M_{\rm seed}\sim10^4$0 is anti-biased with respect to galaxy clustering (favoring isolated, first-time mergers or overdense regions, depending on the channel) (Bonoli et al., 2012, Lupi et al., 2021).
• Present-day Relics: In merger-poor halos, massive seeds formed at $M_{\rm seed}\sim10^4$1 and isolated from subsequent growth could persist as intermediate-mass black holes in local low-mass galaxies or globular clusters (Pacucci et al., 2017).

5. Constraints from Feedback, Environmental Factors, and Assembly Physics

• Feedback-limited Growth: Growth of a massive seed in isolation is capped by feedback; the "transition radius" marks the region where accretion is halted by radiative or mechanical outflows. The maximum seed mass is $M_{\rm seed}\sim10^4$2 for high-$M_{\rm seed}\sim10^4$3 halos $M_{\rm seed}\sim10^4$4 (Pacucci et al., 2017).
• Impact of Strong X-rays: Enhanced X-ray backgrounds, as implied by recent 21-cm constraints, catalyze H$M_{\rm seed}\sim10^4$5 formation, suppressing the direct-collapse pathway, but $M_{\rm seed}\sim10^4$6 seeds still form in regions with baryonic streaming motions, reaching number densities $M_{\rm seed}\sim10^4$7 (Kimura et al., 14 Apr 2025).
• Dynamical Friction and Sinking: Seeds with $M_{\rm seed}\sim10^4$8 may fail to reach galaxy centers within a Hubble time in clumpy $M_{\rm seed}\sim10^4$9 galaxies unless embedded in dense nuclear star clusters or formed in situ within the bulge (Ma et al., 2021).
• Enhancement by Halo Mergers: Fly-by or minor mergers of atomic-cooling halos can double or triple central seed masses, yielding $10^6\,M_\odot$0 per event, compared to isolated collapse (Prole et al., 2024).
• Continuum of Seed Masses: The seed mass function is not strictly bimodal but spans a continuum with $10^6\,M_\odot$1, $10^6\,M_\odot$2–$10^6\,M_\odot$3, reflecting a spectrum of formation mechanisms and environmental prerequisites (Regan et al., 2024).

6. Observational Signatures and Constraints

• Mass Ratios: Early rapid growth yields high birth ratios $10^6\,M_\odot$4–$10^6\,M_\odot$5, above the local mean ($10^6\,M_\odot$6), consistent with "overmassive" black holes observed in some $10^6\,M_\odot$7 AGN and the "Little Red Dots" population at $10^6\,M_\odot$8 (Inayoshi et al., 2021, Kimura et al., 14 Apr 2025, Das et al., 30 Jan 2026).
• Luminous Transients: Super-Eddington accreting seeds at $10^6\,M_\odot$9 are predicted to be detectable with $M_\bullet \gtrsim 10^6\,M_\odot$00–$M_\bullet \gtrsim 10^6\,M_\odot$01 at $M_\bullet \gtrsim 10^6\,M_\odot$02m (Inayoshi et al., 2021).
• Gravitational Waves: Mergers of massive seeds ($M_\bullet \gtrsim 10^6\,M_\odot$03–$M_\bullet \gtrsim 10^6\,M_\odot$04) at high $M_\bullet \gtrsim 10^6\,M_\odot$05 are prime targets for LISA and future GW detectors, with model-predicted rates tied directly to the underlying seed mass function (Lupi et al., 2021, Yajima et al., 2015).
• Occupation Fractions and $M_\bullet \gtrsim 10^6\,M_\odot$06–$M_\bullet \gtrsim 10^6\,M_\odot$07 Relation: Massive-seed models produce increased scatter and potential flattening in the $M_\bullet \gtrsim 10^6\,M_\odot$08–$M_\bullet \gtrsim 10^6\,M_\odot$09 relation at $M_\bullet \gtrsim 10^6\,M_\odot$10 km/s, and a population of BH-free dwarfs, both distinct from light-seed models (Natarajan, 2011, Natarajan, 2011).

7. Current Uncertainties, Outstanding Challenges, and Prospects

• Rarity and Environmental Specificity: The environmental prerequisites for massive seed formation (low metallicity, high inflow rates, suppression of fragmentation) are rare, constraining the overall number density and limiting their contribution to the global SMBH census (Regan et al., 2024, Bonoli et al., 2012, Kimura et al., 14 Apr 2025).
• Role of Mergers: Hierarchical mergers are necessary both for the centralization of seeds via dynamical friction in massive potentials and for overcoming feedback-limited mass caps; isolated seeds rarely reach SMBH scales (Pacucci et al., 2017, Ma et al., 2021).
• Continuum Versus Bimodality: Simulations increasingly favor a continuous seed mass distribution, blending light- and massive-seed channels according to local environmental statistics (Regan et al., 2024).
• Observational Discrimination: Key diagnostics include tracing $M_\bullet \gtrsim 10^6\,M_\odot$11 at high $M_\bullet \gtrsim 10^6\,M_\odot$12, the local occupation fraction in dwarfs, the GW event rate for intermediate-mass binaries, and direct detection of low-luminosity AGN at $M_\bullet \gtrsim 10^6\,M_\odot$13 (Kimura et al., 14 Apr 2025, Volonteri et al., 2021).
• Cosmological Implications: Formation and early growth of massive seeds are robust to standard and many non-standard cosmologies (e.g., $M_\bullet \gtrsim 10^6\,M_\odot$14CDM, $M_\bullet \gtrsim 10^6\,M_\odot$15CDM, braneworld), with heavy initial seed masses ($M_\bullet \gtrsim 10^6\,M_\odot$16) enabling SMBH assembly by $M_\bullet \gtrsim 10^6\,M_\odot$17 across scenarios (Das et al., 30 Jan 2026).

In summary, massive seed black holes constitute a rare but astrophysically crucial pathway to high-redshift SMBH assembly. Their formation demands special cosmic environments, rapid inflows, and efficient fueling before feedback and metal-enrichment curtail their growth. Future observational facilities and next-generation simulations advancing environmental realism and multi-physics feedback are expected to decisively constrain their prevalence and role in cosmic SMBH demographics (Volonteri et al., 2021, Regan et al., 2024, Kimura et al., 14 Apr 2025, Inayoshi et al., 2021).