Evaporation Endpoints and Remnants

An initially macroscopic black hole may disappear, leave a long-lived remnant, approach an extremal state, or enter a quantum regime for which none of those classical nouns is adequate. At the 2026-08-20 evidence cutoff, controlled calculations do not select a universal endpoint for a neutral four-dimensional hole. This is a strong nonselection statement: Hawking’s large-mass approximation ends before the question is decided, while remnant and bounce models add microscopic assumptions whose consistency must be tested individually.

Helpful background. Endpoint Regimes and Remnant Questions explains the semiclassical breakdown scale and separates charged, regular, and information-bearing endpoint proposals.

Four outcomes that must not be conflated

Complete evaporation means that no long-lived localized object remains and requires a rule for the last quantum-gravity regime. A remnant is a state with bounded exterior mass but potentially large internal information capacity. An extremal endpoint is tied to a conserved charge or angular momentum and need not behave like a neutral Planck-mass remnant. A bounce or transition replaces the interior geometry or creates a new asymptotic region. These alternatives have different conservation laws, degeneracies, couplings, and observational signatures.

Assessment at 2026-08-20	Meaning
Question status	Open
Evidence status	Contested; controlled evidence does not select a generic endpoint
Confidence	High in nonselection, not in any particular outcome
Domain	Initially macroscopic four-dimensional holes, with charge and spin stated separately
What could change the conclusion	A UV-complete dynamical endpoint or model-independent no-go result

The finite question is not whether one can write a nonsingular metric. It is whether a theory evolves specified collapse data through the endpoint while preserving its constraints, quantum state, energy accounting, and stated microscopic interpretation. A regular line element is only one ingredient.

A starting project: separate an endpoint from its boundary condition

Choose endpoint research when you are comfortable identifying where an effective description loses control, not merely following a metric to its last coordinate value. First read the linked endpoint background and the self-consistent evaporation project. Then study Russo, Susskind, and Thorlacius 1992, §§2–3, open PDF, extracting the collapse solution, apparent-horizon curve, singularity curve, and proposed matching prescription as four separate ingredients.

Reproduce the horizon–singularity intersection and identify which evolution equations determine it. Follow the subsequent matching carefully and record the additional boundary information and outgoing energy contribution. Draw a causal diagram from the derived curves and label the region in which the semiclassical description is being used. A parameter variation may illustrate dependence within this model; changing the endpoint prescription must be labelled as a different boundary problem, with the constraints rechecked.

Deliver a compact derivation, a reproducible causal-curve plot, and a table distinguishing determined quantities from boundary choices. The calculation needs modest symbolic or numerical tools and an adviser familiar with two-dimensional dilaton gravity. It does not need a full numerical-relativity cluster. A proposed generic remnant model, in contrast, also needs a microscopic state space and couplings before its production or information capacity can be assessed.

A candidate extension is to test the consistency of one specified alternative boundary prescription, including energy balance and the domain where the effective equations apply. Search for existing analyses before presenting that as a new question. Pivot if the conclusion relies entirely on evolution past the singular field map or on an unspecified ultraviolet rule. Explaining precisely what the model leaves undecided is a defensible finite result; selecting an endpoint by hand is not evidence that nature selects it.

Where semiclassical evolution stops answering

Hawking’s leading calculation establishes particle creation on a slowly evolving or fixed classical background Hawking 1975, §§2–4. For a neutral Schwarzschild hole, the temperature grows as the mass falls, so curvature and fractional backreaction eventually approach the cutoff of the approximation. Extrapolating $dM/dt\propto-M^{-2}$ to zero mass is therefore not a controlled derivation of disappearance.

Regular-black-hole geometries demonstrate that curvature singularities can be avoided by suitable effective stress tensors or modified dynamics, but they do not by themselves derive that stress tensor from a quantum state. Frolov’s analysis is particularly useful because it exposes inner-horizon, lifetime, and mass-inflation issues instead of treating regularity as sufficient Frolov 2016, §§I–II, VII.

Results and open questions at 2026-08-20

Evidence	Result	What it establishes	What it leaves open
Fixed-background Hawking calculation	Thermal flux during the large-hole regime	Mass loss and growing temperature under stated approximations	The Planckian endpoint and final quantum state
RST two-dimensional model	Exact modified-model evolution to the apparent-horizon/singularity meeting, followed by a boundary-conditioned continuation	Bulk evolution and endpoint matching can be formulated sharply in a solvable model	Control of the imposed continuation or universality in four dimensions
Regular and limiting-curvature models	Explicit nonsingular causal structures	Logical possibilities and consistency conditions	Derivation from a UV-complete microscopic theory
Remnant production arguments	Potential tension between huge degeneracy and pair production	A serious constraint on couplings and state counting	A universal theorem excluding every remnant
Current remnant reviews	Broad taxonomy of mechanisms	Which assumptions generate which endpoint	Empirical or mathematical selection among them

The RST solution illustrates both the value and limitation of exact models: the counterterm-modified bulk evolution is exact until the apparent horizon meets the singularity, while the subsequent matching requires boundary data Russo, Susskind, and Thorlacius 1992, §§3–4. It shows that endpoint questions can be posed sharply, not that the chosen continuation is controlled through the matching or supplies the four-dimensional answer.

Remnant capacity and production

If a remnant of mass $m_\star$ must encode arbitrarily much information from arbitrarily massive progenitors, it appears to require an unbounded set of internal states at approximately fixed exterior energy. Treating each as an independently coupled species suggests catastrophic production. Giddings showed why that inference is not automatic: production amplitudes depend on form factors and how internal states couple, so naive degeneracy counting is not a model-independent rate calculation Giddings 1995, §§I–IV.

The burden nevertheless remains substantial. A viable remnant proposal should give a Hilbert-space count, formation amplitude, decay law, coupling to ordinary probes, and cosmological abundance. It must also explain how the exterior mass, information capacity, and species renormalization coexist. Ong’s review catalogs many mechanisms, including generalized-uncertainty and regular-hole scenarios, but its examples are evidence for possibilities rather than a dynamical selection theorem Ong 2024, §§2–6.

Adverse evidence and failed routes

The strongest challenge to the statement “remnants are already excluded” is the missing universal relation between internal degeneracy and production coupling. In the other direction, a regular metric or minimum mass does not supply the microscopic state map; that is a failure of the proposed inference, not a null result from a defined search. Both slogans skip the decisive calculation.

Other failed shortcuts include identifying an extremal charged endpoint with a neutral remnant, reading a lower-dimensional boundary condition as a universal law, and inferring unitarity from finite curvature. A white-hole transition can move information without demonstrating a stable amplitude or acceptable lifetime. Conversely, rapid decay avoids overclosure only if it actually returns the encoded information in a controlled channel.

What would change the assessment

A UV-complete calculation starting from a specified macroscopic state and producing a unique late-time density matrix, causal structure, and spectrum would sharply change the verdict. It would need controlled matching across the semiclassical breakdown region and checks of conservation, constraints, and sensitivity to ultraviolet data. A model-independent theorem could instead exclude a broad endpoint class by proving inconsistency with unitarity, locality assumptions, or finite-energy production.

Observation could constrain a specified remnant population, but only after fixing formation abundance, lifetime, mass distribution, and couplings. A null search does not exclude objects that the assumed production channel never creates. A claimed detection must distinguish remnants from ordinary compact relics or dark-sector particles.

Research checklist

• State whether the endpoint is neutral, charged, rotating, extremal, or none of these.
• Give its exterior mass, lifetime, degeneracy, information capacity, and coupling model.
• Track Bondi or ADM energy and any conserved gauge charges.
• Show where semiclassical curvature and fluctuation bounds fail or remain valid.
• Derive rather than stipulate the effective stress tensor producing a regular geometry.
• Test inner-horizon stability and sensitivity to the ultraviolet boundary condition.
• Translate observational limits using an explicit formation and abundance model.

Source and data map

The large-hole premise traces to Hawking’s primary calculation; the exactly solvable modified-model bulk evolution and boundary-conditioned continuation are illustrated by RST; regular-model limitations follow Frolov; and the principal production objection and its caveat trace to Giddings. The recent remnant review is a taxonomy, not neutral adjudication. No endpoint simulation was rerun for this assessment. A useful reproduction must expose initial data, evolution equations, constraints, curvature-to-cutoff ratios, energy balance, state count, and coupling assumptions.

References

• Frolov, V. P. “Notes on Nonsingular Models of Black Holes.” Physical Review D 94, 104056 (2016). doi:10.1103/PhysRevD.94.104056.
• Giddings, S. B. “Why Aren’t Black Holes Infinitely Produced?” Physical Review D 51, 6860–6869 (1995). doi:10.1103/PhysRevD.51.6860.
• Hawking, S. W. “Particle Creation by Black Holes.” Communications in Mathematical Physics 43, 199–220 (1975). doi:10.1007/BF02345020.
• Ong, Y. C. “The Case for Black Hole Remnants: A Review.” arXiv:2412.00322v2 (2024). arXiv.
• Russo, J. G., L. Susskind, and L. Thorlacius. “End Point of Hawking Radiation.” Physical Review D 46, 3444–3449 (1992). doi:10.1103/PhysRevD.46.3444.