Thermodynamics and Statistical Physics Bridge

Temperature and entropy enter black-hole physics through a precise thermodynamic structure, not through analogy alone. Before using horizon laws, one must keep state variables, ensembles, Legendre transforms, fluctuations, and stability conditions straight. This bridge checks those operations in ordinary systems and then transfers them to one sharply stated black-hole setting; it is not a replacement for statistical mechanics or semiclassical gravity.

Required background. Calculus and Linear Algebra for Relativity supplies constrained differentiation, logarithms, eigenvalues, and traces.

Helpful background. Quantum States, Observables, and Measurement develops the recommended quantum-mechanics capability; the Quantum Mechanics and Field Theory Diagnostic checks density operators and Hamiltonian evolution.

You should leave able to use thermodynamic potentials, ensembles, entropy, response functions, and fluctuation formulas with their assumptions explicit, and to test equilibrium reasoning in a black-hole example.

Thermodynamics and statistical-physics diagnostic

1. Starting from $dE=T\,dS-p\,dV+\mu\,dN$, define the Helmholtz free energy $F=E-TS$. Find $dF$, its natural variables, and the derivative that returns the entropy.
2. A two-level system has energies $0$ and $\Delta\gt0$. In the canonical ensemble, find $Z$, the excited-state probability, and the mean energy $U$.
3. Show that a canonical ensemble obeys $\operatorname{Var}(E)=T^2C_V$ in units with $k_B=1$. What stability condition follows for an ordinary canonical equilibrium?
4. Let $\rho_\beta=e^{-\beta H}/Z$ and $F_\beta(\rho)=\operatorname{tr}(\rho H)-T S(\rho)$. Express $F_\beta(\rho)-F_\beta(\rho_\beta)$ in terms of quantum relative entropy and state the conclusion.

Hints and solutions

Hints for 1

Differentiate $F=E-TS$ and substitute the fundamental relation before canceling terms.

Solution to 1

Direct differentiation gives

$$dF=dE-T\,dS-S\,dT =-S\,dT-p\,dV+\mu\,dN.$$

Thus $F=F(T,V,N)$ for this simple system and

$$S=-\left(\frac{\partial F}{\partial T}\right)_{V,N}, \quad p=-\left(\frac{\partial F}{\partial V}\right)_{T,N}, \quad \mu=\left(\frac{\partial F}{\partial N}\right)_{T,V}.$$

The held-fixed variables are part of each statement; omitting them can change the response being computed.

Hints for 2

Sum $e^{-\beta E_i}$ over the two levels. Then use either the normalized weights or $U=-\partial_\beta\ln Z$.

Solution to 2

The partition function and normalized excited-state probability are

$$Z=1+e^{-\beta\Delta}, \qquad p_1=\frac{e^{-\beta\Delta}}{1+e^{-\beta\Delta}} =\frac{1}{e^{\beta\Delta}+1}.$$

Therefore

$$U=\Delta p_1=\frac{\Delta}{e^{\beta\Delta}+1}.$$

$U\to0$ as $T\to0^+$ and $U\to\Delta/2$ as $T\to\infty$, providing useful limiting checks.

Hints for 3

Use $U=-\partial_\beta\ln Z$, $\partial_\beta^2\ln Z=\langle E^2\rangle-\langle E\rangle^2$, and $d\beta/dT=-1/T^2$.

Solution to 3

Differentiating the mean energy yields

$$\frac{\partial U}{\partial\beta} =-\frac{\partial^2\ln Z}{\partial\beta^2} =-\operatorname{Var}(E).$$

At fixed volume and other external parameters,

$$C_V=\frac{\partial U}{\partial T} =\frac{\partial U}{\partial\beta}\frac{d\beta}{dT} =\frac{\operatorname{Var}(E)}{T^2}.$$

Hence $C_V\geq0$ for a well-defined canonical ensemble. A system with negative heat capacity cannot be a stable unconstrained canonical equilibrium of this form; long-range gravitating systems require particular care about ensemble and boundary conditions.

Hints for 4

Insert $\ln\rho_\beta=-\beta H-\ln Z$ into $D(\rho\Vert\rho_\beta)=\operatorname{tr}[\rho(\ln\rho-\ln\rho_\beta)]$.

Solution to 4

Using $S(\rho)=-\operatorname{tr}(\rho\ln\rho)$ gives

$$T D(\rho\Vert\rho_\beta) =\operatorname{tr}(\rho H)-T S(\rho)+T\ln Z.$$

For the Gibbs state, $F_\beta(\rho_\beta)=-T\ln Z$. Therefore

$$F_\beta(\rho)-F_\beta(\rho_\beta) =T D(\rho\Vert\rho_\beta)\geq0.$$

At fixed $T$ and Hamiltonian, the Gibbs state minimizes the Helmholtz free-energy functional. The conclusion assumes a normalized state and finite relative entropy; it does not say that every gravitating system admits a canonical Gibbs state.

Interpret the result

• Secure: potentials are paired with their natural variables, ensemble averages are normalized, response functions include what is held fixed, and stability claims name the ensemble.
• Uncertain: a sign or derivative needs one hint, but the correct ensemble and physical limit are identified. Repeat that problem with a changed level spacing or controlled variable.
• Not yet: entropy and energy are varied independently without a fundamental relation, $Z$ is left unnormalized, fluctuation formulas are used across ensembles without qualification, or negative heat capacity is declared impossible for every isolated gravitating system.

This is a low-stakes bridge. Practice a particular thermodynamic operation before returning to the gravity route that needs it.

Focused review

State functions and Legendre transforms

A thermodynamic potential is useful only after its natural variables and constraints are fixed. Legendre transforms exchange an extensive variable for its conjugate intensive variable; they do not remove physical degrees of freedom. Review States, Processes, and the Laws of Thermodynamics and Thermodynamic Potentials and Stability before retrying item 1.

Ensembles and partition functions

The canonical ensemble fixes $T$ and external controls while allowing energy exchange. Its partition function normalizes the state and generates energy cumulants. Statistical Ensembles and Partition Functions develops the ordinary framework; equivalence with a microcanonical description requires a thermodynamic limit and stability assumptions rather than following automatically.

Entropy, fluctuations, and response

Entropy is a property of a specified state or statistical model. Susceptibilities are derivatives taken under specified controls, and fluctuation identities inherit the ensemble used to derive them. Review these connections with Entropy, Relative Entropy, and Information and Fluctuations, Response, and Susceptibilities.

Density operators

For quantum systems, probabilities are encoded by a positive unit-trace operator and equilibrium by a Gibbs state only when the Hamiltonian, temperature, and trace are well defined. Quantum Statistical Mechanics and Density Operators develops that construction.

Transfer task: Schwarzschild response

In four asymptotically flat dimensions, consider a stationary, neutral, nonrotating Schwarzschild black hole in the semiclassical regime. With $c=\hbar=k_B=1$, take

$$A=16\pi G^2M^2, \qquad S_{\rm BH}=\frac{A}{4G}, \qquad T_H=\frac{1}{8\pi GM}.$$

Verify the first law $dM=T_H\,dS_{\rm BH}$, compute the heat capacity $C=dM/dT_H$, and interpret its sign without assuming that the hole is a stable canonical subsystem in an infinite heat bath.

Hint 1

First write $S_{\rm BH}$ solely as a function of $M$. Compare $dS_{\rm BH}/dM$ with $1/T_H$.

Hint 2

Differentiate $T_H(M)$ and invert. Then compare with the canonical fluctuation identity from diagnostic 3.

Solution

The entropy is

$$S_{\rm BH}=4\pi GM^2,$$

so

$$\frac{dS_{\rm BH}}{dM}=8\pi GM=\frac{1}{T_H}.$$

Consequently $dM=T_H\,dS_{\rm BH}$ for this one-parameter family. The temperature derivative is

$$\frac{dT_H}{dM}=-\frac{1}{8\pi GM^2},$$

and therefore

$$C=\frac{dM}{dT_H}=-8\pi GM^2\lt0.$$

As the black hole loses energy, its temperature rises. The negative heat capacity is compatible with a microcanonical description of an isolated gravitating system, but it conflicts with the nonnegative energy variance required by an ordinary stable canonical ensemble. Thus an asymptotically flat Schwarzschild black hole cannot simply be treated as a stable canonical subsystem in an infinite heat bath. A finite cavity or anti-de Sitter boundary conditions can change the equilibrium and stability problem; those are different systems and must be analyzed separately. The semiclassical origin and domain of $T_H$ are developed in Black-Hole Temperature and Thermodynamic Potentials.

Common errors

• Taking a thermodynamic derivative without stating which variables are fixed.
• Moving between microcanonical and canonical conclusions as though long-range interactions guaranteed ensemble equivalence.
• Interpreting negative heat capacity as an algebra mistake rather than a warning about canonical stability.
• Using the area law outside its stated gravitational theory and semiclassical regime.

Return to your route

After the Schwarzschild transfer problem, return to Relativistic Astrophysics, Black-Hole Thermodynamics and Semiclassical Gravity, or Black-Hole Information.