Remnant Thermodynamic Constraints

A finite-mass remnant can store arbitrarily large information only if it admits an unbounded number of distinguishable internal states at essentially fixed exterior energy, size, and charges, while its couplings avoid catastrophic production and still permit the required formation or release of information. Those assumptions pull in opposite directions. Semiclassical thermodynamics exposes the tension but does not prove a universal no-remnant theorem.

Helpful background. Quantum Fields in Curved Spacetime supplies particle production; Einstein Equations supplies exterior charges; Black-Hole Geometry and Mechanics supplies horizon area; Semiclassical Gravity supplies entropy and backreaction constraints; Black-Hole Information treats information capacity; and Quantum Gravity studies possible microscopic remnant spectra.

Capacity at fixed exterior data

Suppose a remnant has mass $m_r$, radius $R_r$, and $N_r$ orthogonal internal states indistinguishable by long-range charges. Its information capacity is

$$S_r=\log N_r.$$

For a pure joint state of radiation $R$ and remnant $r$, the Schmidt decomposition gives $S(R)=S(r)\le\log N_r$. Thus the remnant needs unbounded capacity if the radiation still has unbounded fine-grained entropy when only the remnant remains to purify it. Arbitrarily large progenitor entropy alone is insufficient: information might already have escaped in radiation correlations. Under the remnant-storage premise and bounded $m_r,R_r$, the external effective description is often modeled as an unbounded collection of states with the same energy and charges; whether those states behave as local particle species is an additional assumption.

Ordinary black-hole thermodynamics would instead associate a compact horizon of area $A_r$ with entropy $A_r/(4G)$. A weak-gravity entropy bound would estimate $S\lesssim2\pi m_rR_r$. Neither expression permits unbounded $S_r$ at fixed $m_r,R_r$. A remnant proposal can reject their applicability—for example, by postulating a large interior volume, nonlocal encoding, a superselection sector, or no ordinary horizon—but that rejection is a new microscopic assumption, not a semiclassical derivation.

The structured objections and proposed evasions are reviewed in Chen, Ong, and Yeom 2015, §§ 2–4, PDF. The review is useful as a map of mechanisms; it is not positive evidence that any one spectrum is realized.

Production and coupling tension

In an effective field theory, a species label enters loop sums and production amplitudes. If each of $N_r$ remnant states couples with comparable amplitude $\mathcal A$ to ordinary fields, an inclusive rate behaves schematically as

$$\Gamma_{\rm inclusive}\sim \sum_{i=1}^{N_r}|\mathcal A_i|^2 \sim N_r|\mathcal A|^2.$$

An infinite degeneracy then threatens divergent pair production, vacuum polarization, or thermal abundance. A fixed nonzero suppression common to infinitely many states still leaves a divergent sum. State-dependent couplings must instead give a summable sequence, $\sum_i|\mathcal A_i|^2\lt\infty$, or another controlled restriction must limit the accessible states. The production problem in a low-energy internal-state description is analyzed directly in Giddings 1995, § 2, Eqs. (2.2)–(2.7), and § 4, PDF. Merely saying that remnants are Planckian does not derive the required suppression.

The opposite problem appears if all couplings vanish. A completely decoupled remnant is difficult to form from ordinary collapse, to carry conserved energy and charge consistently, or to return information in a metastable scenario. A viable model needs state-dependent couplings that are weak enough to control inclusive production yet strong enough to match its formation history and any late decay.

Charge provides both a label and a constraint. An exactly conserved discrete or gauge charge can stabilize a state, but infinitely many states with the same measured charges retain the degeneracy problem. Infinitely many distinct long-range charges would be observationally different species and would alter the low-energy theory. Stability, capacity, and coupling cannot be checked independently.

Thermodynamic consistency tests

Canonical test. A gas of remnant species at inverse temperature $\beta$ contributes roughly

$$Z_r\sim N_r e^{-\beta m_r}.$$

At fixed finite $m_r$, unbounded $N_r$ makes this one-particle contribution diverge. A finite spatial box discretizes momenta but does not cure an infinite internal degeneracy at the same energy. Interactions that lift levels, a restriction to an accessible sector, or a derived cutoff could change the trace, but each changes an assumption and must be justified. This is a diagnostic, not a theorem about every gravitational system, where the proposed canonical ensemble may not exist.

Formation-versus-decay test. Time-reversal or crossing intuition relates formation and decay amplitudes, but gravity, horizons, and coarse-graining complicate that relation. A proposal that asserts exponentially easy formation and exactly forbidden decay must identify the asymmetry. Conversely, exponentially slow information release requires a lifetime long enough that accumulated abundance and interactions remain acceptable.

Species-scale test. Many light species lower the gravitational effective cutoff. Treating $N_r$ internal states as local species may therefore invalidate the effective theory before one computes production. A proposal can evade that conclusion only by explaining why the states are not local independent fields below the cutoff.

Results and open questions through 2026-08-19

The literature through 2026-08-19 establishes a collection of conditional constraints and model-dependent evasions. It remains unresolved whether an ultraviolet-complete theory realizes a remnant satisfying all of them.

Evidence or argument	Result	Scope and limitation
Chen, Ong, and Yeom 2015, §§ 3.1 and 3.3, pp. 43–45 and 48–49, PDF	Arbitrary progenitor information at fixed $m_r,R_r$ requires unbounded state density	Assumes information is stored in distinguishable remnant states rather than nonlocal correlations
Chen, Ong, and Yeom 2015, § 3.1, pp. 43–45, PDF	Comparable couplings make rates or $Z_r$ grow with $N_r$	Local effective-field-theory treatment may fail for extended interiors or superselection sectors
Russo, Susskind, and Thorlacius 1992, §§ II–IV, PDF	The one-loop equations exhibit a curvature singularity; their large-mass backreaction regime loses validity as the receding apparent horizon approaches it, before evaporation completes	Model-specific; it neither constructs nor excludes four-dimensional remnants
Chen, Ong, and Yeom 2015, §§ 2–4, PDF	Catalogues stable, metastable, charged, large-interior, and other proposals with objections and responses	Review of possibilities, not a universal endpoint calculation

The conservative position treats unbounded degeneracy at fixed exterior data as incompatible with an ordinary local low-energy description. The competing permissive position allows remnants to be nonlocal, extended internally, or governed by selection rules absent from naive species counting. The verdict depends on state density, locality, charge sectors, form factors, lifetime, ensemble, and gravitational cutoff.

The comparison axes are therefore explicit: finite versus unbounded state density, local versus nonlocal degrees of freedom, universal versus suppressed couplings, stable versus metastable lifetime, and fixed versus running gravitational cutoff. A proposal must occupy the same point on every axis before two production estimates can be compared.

What remains open

No semiclassical stress tensor determines the microscopic level density, and no thermodynamic inequality used here survives every proposed violation of locality or geometry. Equally, naming a large interior does not compute its states or couplings. A credible model must specify the Hilbert space, measured charges, density of states, formation amplitude, inclusive production rate, lifetime, decay channels, and regime of the exterior mean metric.

The conclusions could change after an ultraviolet-complete model computes both the remnant spectrum and inclusive production, a no-go theorem controls nonlocal or large-interior evasions, or an independent semiclassical solution fixes the final mass and lifetime; a demonstrated observational or cosmological abundance bound for the same coupling assumptions could also change the conclusion.

Exercise

Let $N_r=e^{S_i}$ remnant states of mass $m_r$ each have production amplitude $|\mathcal A_i|=e^{-aS_i}|\mathcal A_0|$. For which $a$ does the inclusive rate remain bounded as $S_i\to\infty$?

Solution

The inclusive rate scales as $N_r|\mathcal A_i|^2=e^{(1-2a)S_i}|\mathcal A_0|^2$. It diverges for $a\lt 1/2$, stays constant for $a=1/2$, and vanishes for $a\gt 1/2$. This counting only states the required suppression; a viable remnant model must derive that form factor and reconcile it with formation and information release.

References

• Chen, P., Ong, Y. C., and Yeom, D.-h. (2015). “Black Hole Remnants and the Information Loss Paradox.” Physics Reports 603, 1–45. Open PDF.
• Giddings, S. B. (1995). “Why Aren’t Black Holes Infinitely Produced?” Physical Review D 51, 6860–6869. DOI. Open PDF.
• Russo, J. G., Susskind, L., and Thorlacius, L. (1992). “Black Hole Evaporation in 1+1 Dimensions.” Physics Letters B 292, 13–18. DOI. Open PDF.