Rotating and Charged Evaporation

Rotation and charge alter both the thermal occupation factor and the greybody transmission problem. Mass, angular momentum, and charge must therefore be evolved together. Bosonic superradiant channels can extract rotation even at low temperature, while charged-particle emission and Schwinger pair production can discharge the hole. Which channel dominates is not universal: it depends on particle masses and charges, partial-wave data, the quantum state, and distance from extremality.

Helpful background. Quantum Fields in Curved Spacetime supplies greybody spectra and superradiance; Einstein Equations supplies Kerr–Newman geometry; Black-Hole Geometry and Mechanics supplies $\Omega_{\mathrm H}$ and $\Phi_{\mathrm H}$; Semiclassical Gravity supplies renormalized fluxes; and Black-Hole Dynamics supplies coupled balance laws and extremality bounds.

Coupled physical flux equations

Let $\chi_{\mathrm H}=\partial_t+\Omega_{\mathrm H}\partial_\phi$ use unit- normalized asymptotic time, and define

$$\Phi_{\mathrm H} =-\left.\chi_{\mathrm H}^{\mu}A_\mu\right|_{\mathrm H} +\left.\chi_{\mathrm H}^{\mu}A_\mu\right|_{\infty}.$$

For an emitted quantum with energy $\omega$, azimuthal number $m$, and charge $q_i$, the horizon-frame energy is

$$\widetilde\omega =\omega-m\Omega_{\mathrm H}-q_i\Phi_{\mathrm H}.$$

The Unruh-state occupation number is filtered by the signed absorption probability $\Gamma_{i\ell m}$. Write the rate in each partial wave as $\mathcal N_{i\ell m}=d^2N_{i\ell m}/(dt\,d\omega)$, including only degeneracies not already counted by the mode sum, as in Luminosity and Mass-Loss Laws. Integrating gives

$$\begin{aligned} \dot M&=-\sum_{i,\ell,m}\int_{\mu_i}^{\infty}d\omega\, \omega\,\mathcal N_{i\ell m}+\dot M_{\mathrm{pair}},\\ \dot J&=-\sum_{i,\ell,m}\int_{\mu_i}^{\infty}d\omega\, m\,\mathcal N_{i\ell m}+\dot J_{\mathrm{pair}},\\ \dot Q&=-\sum_{i,\ell,m}\int_{\mu_i}^{\infty}d\omega\, q_i\,\mathcal N_{i\ell m} +\dot Q_{\mathrm{pair}}. \end{aligned}$$

Here $\mu_i$ is the particle mass. The additional terms denote channels, such as a separately approximated Schwinger process, omitted from the chosen spectra. Discharge changes energy as well as charge; adding only $\dot Q_{\mathrm{pair}}$ while ignoring its corresponding energy balance would be inconsistent. If the full charged-field emission calculation already includes that process, the separate pair terms must be omitted to avoid double counting. These are quasistationary black-hole parameter balances; massive particles are measured through large timelike surfaces, not as flux at future null infinity. Page calculates the physical Kerr partial-wave energy and angular-momentum rates in Page 1976, §§II–V. Hiscock and Weems couple Hawking loss to charged pair production in a specific Reissner–Nordström model Hiscock and Weems 1990, §§II–IV.

Superradiance, discharge, and extremality

For bosons, modes with

$$0\lt\omega\lt m\Omega_{\mathrm H}+q_i\Phi_{\mathrm H}$$

are superradiant. Their absorption probability is negative in the scattering convention, while the thermal denominator changes sign as well; the emitted number flux remains physical. Replacing $\Gamma$ by an unsigned transmission coefficient destroys this balance and can reverse the spin-loss rate.

For this paragraph use geometrized Einstein–Maxwell variables with $G=1$ and Maxwell action $-(16\pi)^{-1}\int\sqrt{-g}F^2d^4x$. Kerr–Newman parameters obey

$$\chi=\frac{J}{M^2}, \qquad q=\frac{Q}{M}, \qquad \chi^2+q^2\leq1.$$

Evolution toward or away from extremality depends on ratios of fluxes, not only on $\dot M$. Light charged particles often make discharge rapid, but a mass threshold can suppress pair creation. Bai and Korwar compute near-extremal charged greybody factors and evolution within a declared semiclassical regime Bai and Korwar 2023, §§2–5, PDF. Very long near-extremal throats can introduce quantum-gravitational dynamics; Brown and collaborators analyze that boundary in a specific framework Brown et al. 2026, §§2–7, PDF. It is a limit on semiclassical extrapolation, not a replacement greybody table.

Declared proxy evolution

The retained computation tests coupled time integration and diagnostics; its coefficients are dimensionless pedagogical proxies, not fits to Page, Hiscock–Weems, or any physical greybody data. All variables and time below are dimensionless; there is no implied physical evaporation timescale. The toy model evolves

$$\begin{aligned} \dot M&=-\frac{0.04}{M^2} \left(1+0.60\chi^2+0.20q^2\right),\\ \dot J&=-\frac{0.12J}{M^3} \left(1+0.40\chi^2\right),\\ \dot Q&=-\frac{0.18Q}{M^2} \left(1+0.30q^2\right). \end{aligned}$$

The initial state is $(M,J,Q)=(1,0.45,0.25)$ and the final dimensionless time is $0.6$. Fixed-step classical RK4 uses 8, 16, 32, and 64 steps. A SciPy DOP853 solution with relative tolerance $10^{-13}$ and absolute tolerance $10^{-15}$ supplies the endpoint comparison. No greybody interpolation is present, so an interpolation refinement is not applicable to this proxy; a physical calculation must refine both time steps and its frequency/parameter tables.

The reference run used CPython 3.9.6, NumPy 2.0.2, SciPy 1.13.1, and IEEE-754 binary64 arithmetic. Download the MIT-licensed benchmark, then use the standalone download-and-run recipe:

curl -fL https://blackhole.org/code/semiclassical-gravity/semiclassical-gravity-benchmarks.py -o semiclassical-gravity-benchmarks.py
python3 semiclassical-gravity-benchmarks.py --case rotating-charged-evaporation

Convergence and balance results

“State error” is the Euclidean norm of the final $(M,J,Q)$ difference from DOP853. “Mass mismatch” compares $M(0)-M(t_f)$ with trapezoidal integration of the proxy mass flux. The extremality column reports the maximum sampled $\chi^2+q^2$.

time-step resolution $N$	$\Delta t$	$M_f$	$J_f$	$Q_f$	state error	mass mismatch	maximum extremality measure
8	7.5000000e-02	9.7210617e-01	4.1497090e-01	2.2329025e-01	1.1524045e-11	2.4902544e-07	0.2650000
16	3.7500000e-02	9.7210617e-01	4.1497090e-01	2.2329025e-01	7.1650052e-13	6.2257142e-08	0.2650000
32	1.8750000e-02	9.7210617e-01	4.1497090e-01	2.2329025e-01	4.4618191e-14	1.5564334e-08	0.2650000
64	9.3750000e-03	9.7210617e-01	4.1497090e-01	2.2329025e-01	2.7560602e-15	3.8910868e-09	0.2650000

The endpoint error falls by approximately a factor of sixteen per step doubling, consistent with fourth-order RK4. The mass mismatch falls by a factor of four because its diagnostic uses second-order trapezoidal quadrature. The largest extremality measure is its initial value $0.265$, so the sampled path remains at least $0.735$ inside the proxy bound.

As an independent benchmark, setting $J=Q=0$ gives the analytic solution

$$M(t)^3=M_0^3-3(0.04)t.$$

The 256-step RK4 neutral run agrees with this exact result to $3.331\times10^{-16}$. That check validates one limiting ODE and the integrator; it does not validate physical emission coefficients.

Failure modes and physical rerun requirements

A production evolution needs frequency- and parameter-resolved greybody tables for every included species, plus a separately validated discharge rate. Refine frequency grids, partial-wave cutoffs, table interpolation, and time steps independently. Monitor the three integrated balance laws, near-horizon regularity, $\chi^2+q^2$, and changes in the dominant channel.

The calculation becomes unreliable if a time step crosses extremality, a superradiant sign is lost, a particle threshold is underresolved, sparse interpolation changes a flux ratio, or near-extremal relaxation is no longer fast compared with evolution. Fixed-background rates must be recomputed on the evolving geometry; treating the proxy coefficients as physical data would be a model error regardless of numerical convergence.

Near extremality, adiabatic validity can therefore break down before a chosen semiclassical cutoff is reached; convergence of the ODE solver cannot repair that approximation failure.

Exercise

Evaluate the proxy’s initial $\dot M$, $\dot J$, and $\dot Q$. Then compute the initial derivatives of $\chi=J/M^2$, $q=Q/M$, and $\chi^2+q^2$. Does the initial trajectory move toward extremality?

Solution

At the initial point, $\chi=0.45$ and $q=0.25$. Substitution gives

$$\dot M=-0.04536, \qquad \dot J=-0.058374, \qquad \dot Q=-0.04584375.$$

Using $\dot\chi=\dot J/M^2-2J\dot M/M^3$ and $\dot q=\dot Q/M-Q\dot M/M^2$ at $M=1$ gives

$$\dot\chi=-0.01755, \qquad \dot q=-0.03450375.$$

Therefore

$$\frac{d}{dt}(\chi^2+q^2) =2\chi\dot\chi+2q\dot q =-0.033046875\lt0.$$

The proxy initially moves away from extremality. This sign follows from its declared coefficients; it is not a universal Kerr–Newman prediction.

References

• Bai, Y., and M. Korwar (2023), “Near-Extremal Charged Black Holes: Greybody Factors and Evolution,” Journal of High Energy Physics 2023(3), 151. Open PDF.
• Brown, A. R., et al. (2026), “The Evaporation of Charged Black Holes,” Journal of High Energy Physics 2026(1), 109. Open PDF.
• Hiscock, W. A., and L. D. Weems (1990), “Evolution of Charged Evaporating Black Holes,” Physical Review D 41, 1142–1151. DOI.
• Page, D. N. (1976), “Particle Emission Rates from a Black Hole. II. Massless Particles from a Rotating Hole,” Physical Review D 14, 3260–3273. DOI.