How Do Massive Black Holes Get Their Gas?

We use multi-scale SPH simulations to follow the inflow of gas from galactic scales to <0.1pc, where the gas begins to resemble a traditional Keplerian accretion disk. The key ingredients are gas, stars, black holes (BHs), self-gravity, star formation, and stellar feedback. We use ~100 simulations to survey a large parameter space of galaxy properties and subgrid models for the ISM physics. We generate initial conditions for our simulations of galactic nuclei (<~300pc) using galaxy scale simulations, including both major mergers and isolated bar-(un)stable disk galaxies. For sufficiently gas-rich, disk-dominated systems, a series of gravitational instabilities generates large accretion rates of up to 1-10 M_sun/yr onto the BH (at «0.1pc); sufficient to fuel the most luminous quasars. The BH accretion rate is highly time variable, given fixed conditions at ~kpc. At >~10pc, our simulations resemble the ‘bars within bars’ model, but the gas exhibits diverse morphologies, including spirals, rings, clumps, and bars; their duty cycle is modest, complicating attempts to correlate BH accretion with nuclear morphology. At ~1-10pc, the gravitational potential becomes dominated by the BH and bar-like modes are no longer present. However, the gas becomes unstable to a standing, eccentric disk or a single-armed spiral mode (m=1), driving the gas to sub-pc scales. Proper treatment of this mode requires including star formation and the self-gravity of both the stars and gas. We predict correlations between BHAR and SFR at different galactic nuclei: nuclear SF is more tightly coupled to AGN activity, but correlations exist at all scales.

💡 Research Summary

The authors present a comprehensive suite of multi‑scale smoothed particle hydrodynamics (SPH) simulations that follow the inflow of interstellar gas from galactic (kiloparsec) scales down to sub‑parsec distances where the flow begins to resemble a classic Keplerian accretion disc. About one hundred simulations explore a wide parameter space, varying galaxy mass, gas fraction, merger versus isolated bar‑unstable disks, and sub‑grid prescriptions for star formation, stellar feedback, and the interstellar medium (ISM) physics. Initial conditions for the nuclear (< 300 pc) runs are generated by “zoom‑in” from larger‑scale galaxy simulations, ensuring that the large‑scale tidal torques, merger‑driven inflows, and bar‑driven torques are self‑consistent with the small‑scale dynamics.

The key result is that in gas‑rich, disk‑dominated systems (gas mass fractions ≳ 30 %), a cascade of gravitational instabilities operates continuously from kiloparsec down to a few parsecs. Large‑scale bars and spiral arms generate torques that drive gas inward, but the morphology is highly time‑variable: the classic “bars‑within‑bars” picture appears only intermittently, occupying roughly 10–30 % of the total simulation time. Consequently, a simple one‑to‑one correlation between the presence of a nuclear bar and active galactic nucleus (AGN) activity is unlikely.

At radii ≳ 10 pc the simulations reproduce the familiar multi‑scale bar hierarchy, yet the gas also forms rings, clumps, and transient spirals. The duty cycle of each morphology is modest, and the transition from bar‑driven to other modes is smooth rather than abrupt. When the radius shrinks to ≈ 1–10 pc, the gravitational potential becomes dominated by the central supermassive black hole (SMBH). The bar‑like m = 2 modes fade, and a new, dominant m = 1 instability emerges: a standing eccentric disc or a single‑armed spiral. This mode is fundamentally different because it requires the self‑gravity of both the gas and the newly formed stars, as well as realistic stellar feedback, to grow to non‑linear amplitudes. The m = 1 pattern efficiently removes angular momentum from the gas, funneling it to sub‑parsec scales where it finally settles into a thin, Keplerian disc that can feed the SMBH at rates of 1–10 M⊙ yr⁻¹.

A major insight is the tight coupling between black‑hole accretion rate (BHAR) and nuclear star formation rate (SFR) when examined on comparable spatial scales. Within the innermost ≲ 100 pc, BHAR and SFR rise and fall together, producing a strong, nearly linear correlation that can explain the observed co‑evolution of SMBHs and their host bulges. On larger (kiloparsec) scales, the average gas supply may be steady, yet BHAR remains highly stochastic because it is regulated by the intermittent gravitational torques described above. This scale‑dependent behavior naturally reproduces the mixed observational results on AGN‑SF correlations.

The study therefore extends the classic “bars‑within‑bars” paradigm by demonstrating that (1) the bar cascade is not a persistent feature but a transient, duty‑cycle‑limited phenomenon; (2) the m = 1 eccentric disc is the principal mechanism that bridges the gap between the SMBH’s sphere of influence (∼ 10 pc) and the sub‑parsec accretion disc; and (3) realistic treatment of star formation, stellar feedback, and gas self‑gravity is essential for capturing these processes. By providing quantitative predictions for BHAR‑SFR correlations across multiple radii, the work offers a robust theoretical framework for interpreting current and future high‑resolution observations (e.g., ALMA, JWST, ELT) of galactic nuclei. Future extensions that incorporate radiative feedback, magnetic fields, and direct comparison with observed kinematic maps will further refine our understanding of how massive black holes acquire their fuel.