Extreme mass ratio inspiral rates: dependence on the massive black hole mass

We study the rate at which stars spiral into a massive black hole (MBH) due to the emission of gravitational waves (GWs), as a function of the mass M of the MBH. In the context of our model, it is shown analytically that the rate approximately depends on the MBH mass as M^{-1/4}. Numerical simulations confirm this result, and show that for all MBH masses, the event rate is highest for stellar black holes, followed by white dwarfs, and lowest for neutron stars. The Laser Interferometer Space Antenna (LISA) is expected to see hundreds of these extreme mass ratio inspirals per year. Since the event rate derived here formally diverges as M->0, the model presented here cannot hold for MBHs of masses that are too low, and we discuss what the limitations of the model are.

💡 Research Summary

The paper investigates how often stars spiral into a massive black hole (MBH) by emitting gravitational waves (GWs), a phenomenon known as an extreme mass‑ratio inspiral (EMRI). The authors develop an analytical framework that links the EMRI event rate Γ to the MBH mass M, and they validate the theory with extensive numerical simulations.

First, the authors describe the dynamical pathway that brings a star into the “loss cone,” the region of phase space where the star’s orbit will intersect the MBH’s capture radius. Two‑body relaxation continuously scatters stars in energy and angular momentum, allowing some to drift into the loss cone. Once inside, the star can either plunge directly into the MBH or, if the gravitational‑wave inspiral time t_GW becomes shorter than the relaxation time t_relax, it will gradually lose orbital energy through GW emission and slowly spiral inward. The boundary where t_GW = t_relax defines a critical radius r_c. Inside r_c, GW emission dominates and the star’s orbit shrinks on a timescale much shorter than the diffusion time.

Assuming a steady‑state Bahcall‑Wolf cusp (stellar density ρ(r) ∝ r⁻⁷⁄⁴) around the MBH, the authors calculate the stellar number density n(r_c), the velocity dispersion σ(r_c), and the critical radius r_c as functions of M. The scaling relations are: r_c ∝ M³⁄⁸, n(r_c) ∝ M⁻¹⁄², and σ(r_c) ∝ M¹⁄⁴. The EMRI rate is then estimated as the flux of stars crossing the loss cone and reaching r_c, which yields

Γ ∝ n(r_c) σ(r_c) r_c² ∝ M⁻¹⁄⁴.

Thus, the event rate declines only weakly with increasing MBH mass.

To test this prediction, the authors perform Monte‑Carlo simulations that simultaneously model two‑body relaxation and GW back‑reaction for MBH masses ranging from 10⁴ to 10⁷ M⊙. They consider three representative stellar populations: stellar‑mass black holes (≈10 M⊙), white dwarfs (≈0.6 M⊙), and neutron stars (≈1.4 M⊙). The simulations confirm the M⁻¹⁄⁴ scaling and reveal a clear hierarchy in EMRI rates: black holes produce the highest rates, followed by white dwarfs, with neutron stars being the least frequent. The hierarchy originates from the stronger GW emission of more massive compact objects, which shortens t_GW and pushes more of them into the EMRI regime.

Applying the sensitivity curve of the Laser Interferometer Space Antenna (LISA), the authors estimate that for MBHs in the range 10⁵–10⁶ M⊙, LISA could detect several hundred EMRIs per year. Black‑hole EMRIs dominate the detectable population because they generate the strongest signals and have the highest intrinsic rates.

A notable theoretical issue is that the derived Γ ∝ M⁻¹⁄⁴ formally diverges as M → 0, implying an unphysical infinite rate for arbitrarily small black holes. The authors identify two reasons for this breakdown. First, the Bahcall‑Wolf cusp assumption fails for low‑mass MBHs, where the stellar density in the core is depleted and the power‑law profile no longer holds. Second, the loss‑cone formalism presumes a sufficiently large reservoir of stars; for very small MBHs the total number of stars bound within the sphere of influence becomes too low for a diffusion‑driven description to be valid. Consequently, the simple scaling cannot be extrapolated to MBH masses below a few × 10⁴ M⊙.

The paper concludes by emphasizing that the M⁻¹⁄⁴ scaling provides a robust, first‑order estimate of EMRI rates across the astrophysically relevant MBH mass range, and that the numerical results give confidence in the predicted LISA detection yields. At the same time, the authors stress the need for more sophisticated models that incorporate realistic stellar mass functions, mass segregation, and the breakdown of the cusp at low MBH masses, in order to refine rate predictions and to understand the transition to regimes where the current framework no longer applies.