Boson Cloud Atlas: Direct mass measurements of superradiance clouds near black holes

Ultralight scalars emerge naturally in several motivated particle physics scenarios and are viable candidates for dark matter. While laboratory detection of such bosons is challenging, their existence in nature can be imprinted on measurable properties of astrophysical black holes (BHs). The phenomenon of superradiance can convert the BH spin kinetic energy into a bound cloud of scalars. In this letter, we propose a new technique for directly measuring the mass of a dark cloud around a spinning BH. We compare the measurement of the BH spin obtained with two independent electromagnetic techniques: continuum fitting and iron K$α$ spectroscopy. Since the former technique depends on a dynamical observation of the BH mass while the latter does not, a mismatch between the two measurements can be used to infer the presence of additional extended mass around the BH. We find that a precision of $\sim 1%$ on the two spin measurements is required to exclude the null hypothesis of no dark mass around the BH at a 2$σ$ confidence level for dark masses about a few percent of the BH mass, as motivated in some superradiance scenarios.

💡 Research Summary

The paper proposes a novel, purely electromagnetic method to directly measure the mass of a boson cloud that may form around a rotating black hole (BH) via the superradiance (SR) instability. Ultralight scalar (ULS) particles, motivated by many extensions of the Standard Model, can extract rotational energy from a Kerr BH and accumulate in hydrogen‑like bound states. The resulting cloud can reach a few percent of the BH mass, thereby altering the effective gravitational mass felt by orbiting matter.

Two widely used X‑ray techniques for measuring BH spin are examined: (i) continuum fitting (CF) and (ii) iron Kα reflection spectroscopy (Kα). CF determines the innermost stable circular orbit (ISCO) radius from the thermal disk spectrum, but it requires an independent dynamical mass estimate (M_dyn) obtained from orbital motion of a companion star, broad‑line region clouds, or tidal‑disruption debris. Kα, by contrast, measures the dimensionless ISCO radius r_ISCO directly from the relativistically broadened Fe Kα line profile, and therefore yields a spin estimate that does not depend on any mass measurement.

If an extended mass M_c (the boson cloud) surrounds the BH, the dynamical mass used in CF becomes M_dyn = M + M_c, while the Kα spin remains unbiased because it does not involve M_dyn. Consequently the two spin estimates, χ_CF and χ_Kα, will differ by an amount that encodes the cloud’s mass fraction ζ ≡ M_c/M. The authors derive a simple relation ζ = g(χ_Kα)/g(χ_CF) − 1, where g(χ) is the Kerr mapping from spin to ISCO radius. Propagating measurement errors Δχ_Kα and Δχ_CF yields an uncertainty Δζ, allowing one to assess the statistical significance of a non‑zero ζ.

Theoretical modeling of the SR process is presented in the non‑relativistic regime. The key parameter is α = GMµ (µ is the boson mass). When the boson Compton wavelength matches the BH horizon scale, the SR condition ω ≤ mΩ_+ is satisfied and the bound‑state mode |nℓm⟩ grows exponentially with rate Γ_nℓm. The authors consider the three lowest‑energy states (|211⟩, |322⟩, |433⟩) and use an improved analytic expression for Γ that remains accurate for α up to ~0.8 and spin χ up to the astrophysical maximum χ_0 ≈ 0.998. Energy extraction stops when the cloud saturates, at which point the cloud mass fraction ζ reaches its maximum, typically a few percent of the BH mass for realistic astrophysical ages (Myr–Gyr).

To connect theory with observations, the paper explores two representative BH masses: a stellar‑mass BH (M = 10 M_⊙) and a supermassive BH (M = 10⁵ M_⊙). For a range of boson masses (µ ≈ 10⁻¹⁶–10⁻¹⁰ eV) the authors compute the evolution of α, χ, and ζ over system ages of 5 Myr (young X‑ray binaries) and 1 Gyr (old low‑mass X‑ray binaries). The results are displayed in Figure 1, which shows contours of ζ/Δζ = 1, 2, 5 in the (χ, σ_χ) plane, where σ_χ is the assumed symmetric spin measurement error for both techniques. The figure also overlays real data points for several well‑studied X‑ray binaries (4U 1543‑475, XTE J1550‑56, GRO J1655‑40, LMC X‑1, GRS 1915+105), illustrating that current spin uncertainties (typically a few percent) are insufficient to detect clouds with ζ ≈ 0.1. However, if σ_χ can be reduced to ≈10⁻² (1 % precision), a 2σ detection of a cloud comprising ~10 % of the BH mass becomes feasible for a broad range of spins. For supermassive BHs, a slightly looser precision (σ_χ ≈ few × 10⁻²) would already allow detection across much of the parameter space.

The authors discuss systematic uncertainties inherent to each technique. CF is sensitive to disk thickness, magnetic pressure, deviations from the Novikov‑Thorne profile, and uncertainties in distance, inclination, and the dynamical mass tracer. Kα spectroscopy suffers from ambiguities in coronal geometry, ionization gradients, high‑density plasma effects, and possible emission from inside the ISCO. Mitigating these systematics to the sub‑percent level is identified as a critical challenge.

In summary, the paper demonstrates that a precise, dual‑method spin comparison can serve as a direct probe of boson clouds, offering a complementary approach to gravitational‑wave searches for continuous signals from cloud annihilation or to stellar‑dynamics constraints near Sgr A*. Achieving ≲1 % spin measurement precision—potentially within reach of upcoming X‑ray missions such as XRISM, Athena, and Lynx—could either reveal the presence of ultralight boson clouds or place stringent limits on their existence, thereby informing both particle physics models of dark matter and the astrophysics of black‑hole spin evolution.