11.05.06 · stat-mech-physics / quantum-stats

Photon gas, phonon gas, and the Debye model of solids

shipped3 tiersLean: none

Anchor (Master): Landau & Lifshitz, *Statistical Physics, Part 1*, 3e (Pergamon, 1980), §64-65; Pathria & Beale (2021), §7.3-7.4; Born & Huang, *Dynamical Theory of Crystal Lattices* (Oxford, 1954); Kittel (2005), §4-5; Ziman, *Electrons and Phonons* (Oxford, 1960)

Intuition Beginner

Two of the most useful gases in physics contain no matter at all. The first is the photon gas — light bouncing around inside a hot oven, or the radiation that fills the early universe. The second is the phonon gas — quantised sound waves rattling around inside a solid block of diamond or copper. Both are made of particles that have no rest mass, both have no fixed total number, and both follow the same Bose-Einstein statistics with chemical potential set to zero. Different physical context, identical mathematical structure.

A photon's energy is set by its frequency through Planck's relation $E = h f$ . A phonon's energy is also $E = h f$ , but instead of the frequency being the colour of light, it is the pitch of a sound wave travelling through the crystal. In both cases, doubling the temperature lets each oscillation mode hold more energy, and the average occupation of a mode of frequency $f$ is given by the Bose-Einstein formula $n (f) = 1/ (e^{h f / k_{B} T} - 1)$ . This is the same formula that Planck wrote down in 1900 to fix the ultraviolet catastrophe of blackbody radiation.

The big difference between the two gases is at high frequency. Light can have any frequency at all, so the photon spectrum extends to infinity; you just have to be hot enough to populate the high-frequency modes. Sound in a solid cannot extend to arbitrarily high frequency, because there are only as many vibrational modes as there are atoms in the crystal. If a piece of copper has $N$ atoms, it has exactly $3 N$ sound modes — three for each atom, one in each spatial direction. The highest sound frequency that can fit in the crystal is called the Debye frequency $f_{D}$ , and the corresponding temperature is the Debye temperature $T_{D} = h f_{D} / k_{B}$ .

Why does this matter? Because the Debye temperature controls how hot a solid has to be before it absorbs heat in the classical way. Diamond is extraordinarily stiff, with $T_{D} \approx 2230$ K — far above room temperature — so at room temperature most of diamond's vibrational modes are frozen out and diamond has a very small heat capacity. Lead is soft, with $T_{D} \approx 105$ K — well below room temperature — so at room temperature all of lead's modes are populated and lead behaves classically, with heat capacity $3 R$ per mole. Silicon sits in between at $T_{D} \approx 645$ K.

Peter Debye in 1912 worked out the precise formula. At low temperature the heat capacity of a solid scales as $C_{V} \propto T^{3}$ , which is one of the most striking signatures of quantum mechanics in everyday matter. At high temperature it approaches the classical Dulong-Petit value $3 N k_{B}$ . The cubic-in-temperature law has been measured in diamond, silicon, copper, lead, argon, and every other crystalline solid ever tested. It is wrong only when superconductivity intervenes (the electronic contribution becomes exponentially small below $T_{c}$ and the phonon $T^{3}$ piece dominates the residual specific heat).

One-sentence takeaway: a solid behaves like a box of light when sound waves are quantised, and the resulting Debye $T^{3}$ law together with the photon-gas Stefan-Boltzmann $T^{4}$ law are the same Planck spectrum applied to different dispersion relations.

Visual Beginner

Imagine a hollow oven held at temperature $T$ . Light fills the cavity, with shorter wavelengths becoming more populated as $T$ rises. A simple sketch of the energy density per unit frequency shows the Planck curve: low at small frequency, rising to a peak near $f_{peak} \approx 6 \times 1 0^{10} T$ Hz, then falling back to zero. Heat the oven, the peak shifts to higher frequency and the area under the curve grows as $T^{4}$ .

The phonon picture is similar but with two key differences. First, the curve has a hard cutoff at the Debye frequency $f_{D}$ — beyond this, there are no more modes available, because the crystal has only $3 N$ of them in total. Second, the area under the curve approaches a fixed maximum equal to the classical value $3 N k_{B} T$ at high temperature, instead of growing without bound. At low temperature, however, the cubic rise of the mode density at small frequency produces the famous Debye $T^{3}$ heat-capacity law.

Worked example Beginner

Estimate the Debye temperature of silicon from its measured sound speed, and predict its low-temperature heat capacity.

Silicon has $N / V \approx 5.0 \times 1 0^{28}$ atoms per cubic metre and a mean sound speed (average over longitudinal and transverse modes) of $v_{s} \approx 6, 400$ m/s. The Debye formula gives $ω_{D} = v_{s} (6 π^{2} N / V)^{1/3}$ .

Step 1. Compute $(6 π^{2} N / V)^{1/3}$ : $$ 6 \pi^2 \times 5.0 \times 10^{28} \approx 2.96 \times 10^{30},\text{m}^{-3}, $$ so $(6 π^{2} N / V)^{1/3} \approx 1.44 \times 1 0^{10} m^{- 1}$ .

Step 2. Multiply by the sound speed: $$ \omega_D \approx 6{,}400 \times 1.44 \times 10^{10} \approx 9.2 \times 10^{13},\text{rad/s}. $$

Step 3. Convert to Debye temperature: $T_{D} = ℏ ω_{D} / k_{B}$ . With $ℏ = 1.055 \times 1 0^{- 34}$ J·s and $k_{B} = 1.381 \times 1 0^{- 23}$ J/K, $$ T_D \approx 1.055 \times 10^{-34} \times 9.2 \times 10^{13} / 1.381 \times 10^{-23} \approx 700,\text{K}. $$ The measured value for silicon is $T_{D} \approx 645$ K. The simple Debye estimate sits within $10%$ of experiment, capturing the right magnitude despite ignoring all band-structure detail.

What this tells us: a one-line application of the Debye formula gets the heat-capacity scale of a real solid right to within $10%$ . The residual difference arises from the difference between longitudinal and transverse sound speeds and from the curvature of the actual phonon dispersion near the zone boundary.

Repeating the same exercise for diamond ( $v_{s} \approx 18, 000$ m/s, $N / V \approx 1.76 \times 1 0^{29}$ m $^{- 3}$ , predicted $T_{D} \approx 2300$ K, measured $T_{D} \approx 2230$ K), lead ( $v_{s} \approx 1, 200$ m/s, $N / V \approx 3.3 \times 1 0^{28}$ m $^{- 3}$ , predicted $T_{D} \approx 100$ K, measured $T_{D} \approx 105$ K), and aluminium ( $v_{s} \approx 5, 100$ m/s, $N / V \approx 6.0 \times 1 0^{28}$ m $^{- 3}$ , predicted $T_{D} \approx 420$ K, measured $T_{D} \approx 428$ K) gives the same level of agreement.

At room temperature $T = 300$ K, lead has $T / T_{D} \approx 2.86$ , which is well above the classical limit; its heat capacity is essentially the Dulong-Petit value $3 N k_{B}$ . Silicon has $T / T_{D} \approx 0.46$ , intermediate between cubic and classical. Diamond has $T / T_{D} \approx 0.13$ at room temperature, well into the low-temperature regime; its heat capacity is far below the classical value. This is why diamond feels cold to the touch even at room temperature: heat flows quickly into the crystal because its thermal conductivity is high, but the crystal itself stores little thermal energy per atom because most modes are frozen.

Check your understanding Beginner

Exercise (easy, multiple choice).

The Debye model predicts that the heat capacity of a solid at temperatures well below the Debye temperature is:

A. Constant (Dulong-Petit law)

B. Proportional to $T$ (linear, like the Sommerfeld electronic contribution)

C. Proportional to $T^{3}$ (cubic)

D. Exponentially small, like the Einstein model

Hint

The Debye mode density grows as $ω^{2}$ at low frequency, and the total energy involves integrating $ω^{3} / (e^{ℏ ω / k_{B} T} - 1)$ , which after rescaling gives a fixed integral times a factor of $T^{4}$ .

Answer

Option C. $C_{V} = (12 π^{4} /5) N k_{B} (T / T_{D})^{3}$ , cubic in $T$ . The linear-in- $T$ option B describes the electronic Sommerfeld contribution in a metal; the exponentially small option D describes the older Einstein model, which assumed all modes have the same frequency and therefore predicted exponential suppression of the heat capacity at low temperature. The Einstein prediction fails badly against experimental data; Debye 1912 fixed it.

Exercise (easy, true/false).

A photon gas and a phonon gas both follow Bose-Einstein statistics with chemical potential equal to zero, because in both cases the number of particles is not conserved — photons can be created or destroyed by absorption and emission, and phonons can be created or destroyed by lattice excitation and relaxation.

Hint

If a particle number is not conserved, the system minimises free energy by adjusting $N$ freely, which forces the chemical potential — the cost of adding one particle — to equal zero at equilibrium.

Answer

True. Both photons and phonons are bosonic excitations with no conserved particle number. The equilibrium number is fixed by temperature alone: doubling $T$ creates more photons or phonons until the system reaches the new equilibrium. This is the precise statistical-mechanical reason $μ = 0$ for both gases. By contrast, atomic Bose gases have a conserved particle number and a non-zero chemical potential.

Formal definition Intermediate+

Both the photon gas and the phonon gas are ideal gases of massless bosons in thermal equilibrium at temperature $T$ . The chemical potential vanishes because the particle number is not conserved: photon emission and absorption (or phonon creation and annihilation by lattice excitation) drive the system to the equilibrium $N (T)$ determined by $T$ alone. The mean occupation of a mode of angular frequency $ω$ is the Planck function $$ n(\omega) = \frac{1}{e^{\hbar\omega/k_B T} - 1}. $$

Definition (photon gas). Electromagnetic radiation in a box of volume $V$ at temperature $T$ in thermal equilibrium with the walls. The single-mode dispersion is $ω = c k$ (linear, massless), with two transverse polarisations per wavevector $k$ . The mode density per unit angular frequency is $$ g_{\rm ph}(\omega) = \frac{V \omega^2}{\pi^2 c^3}, $$ where the factor of 2 from polarisation is already included.

Definition (phonon gas, Debye model). Quantised lattice vibrations of a crystal of $N$ atoms in volume $V$ . In the Debye approximation, the dispersion is treated as linear $ω = v_{s} k$ up to a cutoff $ω_{D}$ , where $v_{s}$ is the angle-averaged sound speed defined by $3/ v_{s}^{3} = 1/ v_{L}^{3} + 2/ v_{T}^{3}$ (one longitudinal mode, two transverse modes). The total mode count is constrained to $3 N$ : $$ 3 N = \int_0^{\omega_D} g_{\rm Debye}(\omega),d\omega, \qquad g_{\rm Debye}(\omega) = \frac{3 V \omega^2}{2 \pi^2 v_s^3}. $$ Solving for $ω_{D}$ gives the Debye cutoff frequency $$ \omega_D = v_s \big(6 \pi^2 N / V\big)^{1/3}, $$ and the Debye temperature $T_{D} = ℏ ω_{D} / k_{B}$ .

Definition (Debye function). The dimensionless function $$ D(x) = \frac{3}{x^3}\int_0^x \frac{y^3}{e^y - 1},dy $$ controls the phonon-gas internal energy through $U (T) = 3 N k_{B} T D (T_{D} / T)$ — equivalently, $U (T) = 9 N k_{B} T (T / T_{D})^{3} \int_{0}^{T_{D} / T} y^{3} / (e^{y} - 1) d y$ . Limits: $D (x) \to 1$ as $x \to 0$ (classical Dulong-Petit limit, $T ≫ T_{D}$ ); $D (x) \to (π^{4} /5) / x^{3}$ as $x \to \infty$ (low-temperature Debye limit, $T ≪ T_{D}$ ).

Photon-gas thermodynamics

Computing the internal energy of a photon gas in a box, using the mode density $g_{ph} (ω)$ and the Bose occupation $n (ω)$ : $$ \frac{U_{\rm ph}}{V} = \int_0^\infty \hbar\omega,n(\omega),\frac{\omega^2}{\pi^2 c^3},d\omega = \frac{(k_B T)^4}{\pi^2 (\hbar c)^3}\int_0^\infty \frac{x^3}{e^x - 1},dx. $$ The dimensionless integral evaluates to $π^{4} /15$ via $\int_{0}^{\infty} x^{3} / (e^{x} - 1) d x = 6 ζ (4) = π^{4} /15$ , giving the Stefan-Boltzmann law $$ u(T) = a T^4, \qquad a = \frac{\pi^2 k_B^4}{15 \hbar^3 c^3} = 7.566 \times 10^{-16},\text{J/(m}^3,\text{K}^4\text{)}. $$ Photon pressure is one-third of the energy density: $P V = U /3$ , the equation of state of an ultrarelativistic gas. Entropy is $S = (4/3) U / T$ , and the total photon number scales as $N (T) \propto V T^{3}$ (with prefactor $2 ζ (3) / π^{2}$ per cubic thermal wavelength).

Phonon-gas thermodynamics

For the phonon gas the integral is cut off at $ω_{D}$ , giving the Debye internal energy $$ U_{\rm Debye}(T) = 9 N k_B T \left(\frac{T}{T_D}\right)^3 \int_0^{T_D/T} \frac{y^3}{e^y - 1},dy = 3 N k_B T,D(T_D/T). $$

In the high-temperature limit $T ≫ T_{D}$ , $D \to 1$ and $U \to 3 N k_{B} T$ , so $C_{V} \to 3 N k_{B}$ — the Dulong-Petit law. In the low-temperature limit $T ≪ T_{D}$ , the upper limit can be replaced by $\infty$ and the Debye integral becomes $\int_{0}^{\infty} y^{3} / (e^{y} - 1) d y = π^{4} /15$ , giving $$ U_{\rm Debye}(T) \to \frac{3 \pi^4}{5} N k_B T \left(\frac{T}{T_D}\right)^3, \qquad C_V \to \frac{12 \pi^4}{5} N k_B \left(\frac{T}{T_D}\right)^3. $$ The famous Debye $T^{3}$ law is the universal low-temperature heat-capacity scaling of insulators. In metals, this adds to the Sommerfeld linear-in- $T$ electronic term, producing $C_{V} = γ T + β T^{3}$ with $β = (12 π^{4} /5) N k_{B} / T_{D}^{3}$ .

Counterexamples to common slips Intermediate+

The photon gas is not an ideal classical gas. Stefan-Boltzmann $u \propto T^{4}$ replaces the classical Rayleigh-Jeans $u \propto T$ that fails at high frequency (the ultraviolet catastrophe). Total photon number $N \propto T^{3}$ rather than fixed — photons can be created or destroyed at the walls.
The Debye approximation is not exact. It replaces the actual phonon dispersion (with its first Brillouin zone, optical branches, anisotropies, and acoustic-branch curvature) by an isotropic linear dispersion up to a sphere of radius $k_{D} = ω_{D} / v_{s}$ . The cutoff is artificial; real dispersion deviates at the zone boundary, and the agreement is only asymptotic.
Phonons cannot Bose-condense at equilibrium. Unlike a conserved-particle Bose gas where $μ$ can be tuned negative, the phonon $μ = 0$ is fixed and the equilibrium $N (T)$ is determined entirely by temperature. No analogue of the BEC critical line exists.
The Einstein 1907 model is a low-quality approximation. Einstein assumed all modes share a single frequency $ω_{E}$ , producing $C_{V} \propto (ω_{E} / T)^{2} exp (- ω_{E} / T)$ at low $T$ — exponentially small, contradicting the measured cubic behaviour. Debye 1912 corrected this by replacing the single frequency with a continuum from 0 to $ω_{D}$ , restoring the $T^{3}$ law.
The Stefan-Boltzmann radiation constant $a$ differs from the surface emission constant $σ$ . The relation is $a = 4 σ / c$ , where $σ = π^{2} k_{B}^{4} / (60 ℏ^{3} c^{2}) = 5.670 \times 1 0^{- 8}$ W/(m $^{2}$ ·K $^{4}$ ) is the Stefan-Boltzmann surface constant relevant to blackbody emission per unit area, and $a$ is the volume energy density. The factor of 4 comes from the angle-averaged outgoing flux.

Einstein vs. Debye: side-by-side comparison

The Einstein model (Einstein 1907 Annalen 22, 180) was the first quantum theory of solid heat capacities. It assumed each atom in a crystal vibrates independently at a single frequency $ω_{E}$ , giving $$ C_V^{\rm Einstein} = 3 N k_B \left(\frac{T_E}{T}\right)^2 \frac{e^{T_E/T}}{(e^{T_E/T} - 1)^2}, \qquad T_E = \hbar\omega_E / k_B. $$ At high temperature both Einstein and Debye approach $3 N k_{B}$ ; the Dulong-Petit limit is recovered correctly by either model. The difference appears at low temperature. The Einstein model predicts exponential suppression $C_{V} \propto (T_{E} / T)^{2} exp (- T_{E} / T)$ — but measurements on diamond, silicon, and metals showed a much slower $T^{3}$ falloff. Debye 1912 replaced the single Einstein frequency with a distribution $g (ω) \propto ω^{2}$ extending from 0 to $ω_{D}$ , which restores the experimentally observed $T^{3}$ scaling. The Einstein model survives today as the description of optical phonon branches (which have a finite gap and exponentially small low- $T$ contribution); the Debye model describes the acoustic phonons (which are gapless and dominate at low $T$ ).

Key theorem with proof Intermediate+

Theorem (Debye $T^{3}$ law). For a Debye solid of $N$ atoms in volume $V$ at temperature $T ≪ T_{D}$ , the lattice heat capacity is $$ C_V = \frac{12 \pi^4}{5} N k_B \left(\frac{T}{T_D}\right)^3 + O(T^5). $$ The same Sommerfeld-style asymptotic expansion that controls the photon-gas Stefan-Boltzmann $T^{4}$ energy density also controls the Debye phonon $T^{3}$ heat capacity, with the cutoff inactive at low temperature.

Proof. Start with the Debye internal energy $$ U(T) = \int_0^{\omega_D} \hbar\omega,n(\omega),g_{\rm Debye}(\omega),d\omega = \frac{3 V \hbar}{2 \pi^2 v_s^3}\int_0^{\omega_D} \frac{\omega^3}{e^{\hbar\omega/k_B T} - 1},d\omega. $$ Substitute $y = ℏ ω / k_{B} T$ , so $ω = (k_{B} T /ℏ) y$ and $d ω = (k_{B} T /ℏ) d y$ . The upper limit becomes $y_{D} = T_{D} / T$ , where $T_{D} = ℏ ω_{D} / k_{B}$ . Then $$ U(T) = \frac{3 V (k_B T)^4}{2 \pi^2 (\hbar v_s)^3}\int_0^{T_D/T} \frac{y^3}{e^y - 1},dy. $$ Using the Debye normalisation $ω_{D}^{3} = 6 π^{2} v_{s}^{3} N / V$ , equivalently $V / v_{s}^{3} = 6 π^{2} N / ω_{D}^{3}$ , the prefactor becomes $$ \frac{3 V (k_B T)^4}{2 \pi^2 (\hbar v_s)^3} = \frac{3 \cdot 6 \pi^2 N (k_B T)^4}{2 \pi^2 (\hbar\omega_D)^3} = \frac{9 N (k_B T)^4}{(\hbar\omega_D)^3} = 9 N k_B T \left(\frac{T}{T_D}\right)^3. $$ Therefore $$ U(T) = 9 N k_B T \left(\frac{T}{T_D}\right)^3 \int_0^{T_D/T} \frac{y^3}{e^y - 1},dy. $$

For $T ≪ T_{D}$ , the upper limit $T_{D} / T \to \infty$ . The integral converges to $$ \int_0^\infty \frac{y^3}{e^y - 1},dy = \Gamma(4),\zeta(4) = 6 \cdot \frac{\pi^4}{90} = \frac{\pi^4}{15}. $$ The correction from extending the upper limit to infinity rather than $T_{D} / T$ is exponentially small in $T_{D} / T$ , of order $exp (- T_{D} / T)$ , and contributes to the $O (T^{5})$ remainder via the next term in the expansion of $1/ (e^{y} - 1)$ at large $y$ .

Therefore the low-temperature internal energy is $$ U(T) = \frac{9 \pi^4}{15} N k_B T \left(\frac{T}{T_D}\right)^3 + O(T^5) = \frac{3 \pi^4}{5} N k_B T \left(\frac{T}{T_D}\right)^3 + O(T^5). $$

Differentiating with respect to $T$ : $$ C_V = \frac{\partial U}{\partial T} = \frac{3 \pi^4}{5} N k_B \left[1 + 3\right] \left(\frac{T}{T_D}\right)^3 + O(T^5) = \frac{12 \pi^4}{5} N k_B \left(\frac{T}{T_D}\right)^3 + O(T^5), $$ which is the Debye $T^{3}$ law. The numerical coefficient $12 π^{4} /5 \approx 233.78$ . $□$

Bridge. The Debye $T^{3}$ law builds toward 11.05.03 blackbody Stefan-Boltzmann radiation as the structurally identical phenomenon for a different dispersion: the same dimensionless integral $\int_{0}^{\infty} y^{3} / (e^{y} - 1) d y = π^{4} /15$ appears in both, and the central insight is that the low-frequency mode density $g (ω) \propto ω^{2}$ controls the leading temperature scaling — $T^{4}$ for the photon energy density and $T^{3}$ for the phonon heat capacity. This is exactly the same Planck spectrum applied to two different physical contexts. Putting these together with 11.05.01 Bose-Einstein occupation $n (ω) = 1/ (e^{ℏ ω / k_{B} T} - 1)$ , the Debye result generalises Planck's blackbody calculation to a finite-mode lattice, and the bridge is between cavity electromagnetism (mode count infinite) and crystal vibrations (mode count $3 N$ ). The Debye temperature $T_{D}$ identifies the cutoff with the typical interatomic distance via $ω_{D} / v_{s} = (6 π^{2} N / V)^{1/3} \sim 1/ a$ , appears again in 12.13.01 bosonic Fock space where phonons live as quanta in the harmonic-oscillator ladder, and recurs in the BCS theory of superconductivity where the cutoff $ω_{D}$ enters the exponential prefactor of the critical temperature.

Exercises Intermediate+

Exercise 3 (medium, symbolic).

Derive the photon-gas equation of state $P V = U /3$ from the relation between energy and momentum for a massless particle.

Hint

For a single mode of frequency $ω$ in a cubical box of side $L$ , the energy is $ℏ ω = ℏ c ∣ k ∣$ with $k = (n_{1}, n_{2}, n_{3}) π / L$ . Differentiate energy with respect to $V$ at fixed mode occupation to get the contribution to pressure.

Answer

For a mode $k$ in volume $V = L^{3}$ , $∣ k ∣ \propto 1/ L \propto V^{- 1/3}$ , so $ω \propto V^{- 1/3}$ and $\partial ln ω / \partial ln V = - 1/3$ . The pressure from a single occupied mode is $$ P_{\rm mode} = -\partial E / \partial V = -\hbar\omega \cdot \partial \ln\omega / \partial \ln V \cdot 1/V = (1/3)\hbar\omega/V. $$ Summing over modes with the Bose-Einstein occupation $n (ω)$ : $$ P = \frac{1}{3 V} \sum_{\mathbf k, \rm pol} \hbar\omega,n(\omega) = U/(3V), $$ i.e., $P V = U /3$ . This is the equation of state of an ultrarelativistic ideal gas (any massless gas with linear dispersion). Contrast with the non-relativistic ideal gas $P V = 2 U /3$ . The same factor of $1/3$ governs ultrarelativistic gauge-boson gases in the early universe before electroweak symmetry breaking.

Exercise 5 (medium, symbolic).

Derive Wien's displacement law $ω_{peak} / T = const$ for the peak of the Planck spectrum of a photon gas.

Hint

The energy density per angular frequency is $u (ω, T) = (ℏ ω^{3} / π^{2} c^{3}) / (e^{ℏ ω / k_{B} T} - 1)$ . Differentiate with respect to $ω$ at fixed $T$ , set to zero, and solve numerically for $x = ℏ ω / k_{B} T$ .

Answer

Setting $\partial u / \partial ω = 0$ gives the transcendental equation $3 (1 - e^{- x}) = x$ , whose solution is $x \approx 2.821$ . Therefore $ω_{peak} = 2.821 \cdot k_{B} T /ℏ$ , so $ω_{peak} / T \approx 3.69 \times 1 0^{11}$ rad/(s·K). Alternatively, in terms of wavelength $λ$ via $u_{λ} d λ = u_{ω} d ω$ , the peak shifts and the Wien displacement constant is $λ_{peak} T = 2.898 \times 1 0^{- 3}$ m·K. The constancy of $ω_{peak} / T$ across all temperatures is a manifestation of the self-similar scaling of the Planck spectrum under $ω \to λω$ , $T \to λ T$ .

Exercise 7 (hard, symbolic).

Derive the leading high-temperature correction to the Debye heat capacity, showing that for $T ≫ T_{D}$ , $C_{V} = 3 N k_{B} [1 - (1/20) (T_{D} / T)^{2} + O ((T_{D} / T)^{4})]$ .

Hint

Use the Bernoulli-number expansion $1/ (e^{y} - 1) = 1/ y - 1/2 + y /12 - y^{3} /720 + \dots$ in the Debye integral, integrate term by term up to $ω_{D}$ , and expand $D (T_{D} / T)$ in powers of $T_{D} / T$ .

Answer

Expand $y^{3} / (e^{y} - 1) = y^{2} - y^{3} /2 + y^{4} /12 - y^{6} /720 + \dots$ . Integrating from $0$ to $x = T_{D} / T$ : $$ \int_0^x \frac{y^3}{e^y - 1},dy = \frac{x^3}{3} - \frac{x^4}{8} + \frac{x^5}{60} - \frac{x^7}{5040} + \cdots $$ Then $D (x) = (3/ x^{3}) [x^{3} /3 - x^{4} /8 + x^{5} /60 - \dots] = 1 - (3/8) x + (1/20) x^{2} - \dots$ . Multiplying by $3 N k_{B} T$ : $$ U(T) = 3 N k_B T - \frac{9}{8} N k_B T_D + \frac{3}{20} N k_B T_D^2 / T - \cdots $$ Differentiating in $T$ gives $$ C_V = 3 N k_B - \frac{3}{20} N k_B (T_D/T)^2 + O((T_D/T)^4). $$ Note the $(3/8) N k_{B} T_{D}$ is a $T$ -independent contribution to $U$ (the zero-point energy of the phonon modes); it does not contribute to $C_{V}$ . The leading $- (1/20) (T_{D} / T)^{2}$ correction explains the small undershoot of measured high-temperature heat capacities below the Dulong-Petit value.

Exercise 8 (hard, symbolic).

Compute the zero-point energy of the Debye phonon gas in a solid of $N$ atoms, and show that it produces a finite zero-point pressure.

Hint

Each mode of angular frequency $ω$ contributes $ℏ ω /2$ to the ground-state energy. Sum over the Debye spectrum.

Answer

The zero-point energy is $$ E_{\rm zp} = \int_0^{\omega_D} \frac{\hbar\omega}{2},g_{\rm Debye}(\omega),d\omega = \frac{3 V \hbar}{4\pi^2 v_s^3}\int_0^{\omega_D} \omega^3,d\omega = \frac{3 V \hbar \omega_D^4}{16 \pi^2 v_s^3}. $$ Using $ω_{D}^{3} = 6 π^{2} v_{s}^{3} N / V$ , this simplifies to $E_{zp} = (9/8) N ℏ ω_{D} = (9/8) N k_{B} T_{D}$ . For diamond ( $T_{D} = 2230$ K, $N_{A}$ atoms), $E_{zp}^{mol} = (9/8) R T_{D} \approx 2.09 \times 1 0^{4}$ J/mol, or about $0.22$ eV per atom — significant on the scale of thermal energies but small compared to chemical bond energies (4-8 eV per bond).

Zero-point pressure: $P_{zp} = - \partial E_{zp} / \partial V$ at fixed $N$ . With $ω_{D} \propto V^{- 1/3}$ , $E_{zp} \propto V^{- 1/3}$ and so $P_{zp} = (1/3) E_{zp} / V$ . For diamond at typical densities, $P_{zp} \sim 10$ GPa — a large lattice-stabilising pressure that anharmonic corrections must balance against the electronic bonding to reach the equilibrium lattice constant.

Exercise 9 (hard, symbolic).

Derive the Grüneisen relation $α V = γ C_{V} κ_{T}$ between thermal expansion coefficient $α$ , isothermal compressibility $κ_{T}$ , heat capacity $C_{V}$ , and Grüneisen parameter $γ = - d ln ω / d ln V$ .

Hint

Start from the Helmholtz free energy of harmonic phonons $F (V, T) = \sum_{k} [ℏ ω_{k} /2 + k_{B} T ln (1 - e^{- ℏ ω_{k} / k_{B} T})]$ , compute $P = - (\partial F / \partial V)_{T}$ , and combine with $α = (\partial V / \partial T)_{P} / V$ and $κ_{T} = - (1/ V) (\partial V / \partial P)_{T}$ .

Answer

The phonon free energy is $F_{ph} (V, T) = \sum_{k} ℏ ω_{k} (V) /2 + k_{B} T \sum_{k} ln (1 - e^{- ℏ ω_{k} (V) / k_{B} T})$ . Differentiate in $V$ : $$ P_{\rm ph} = -\partial F_{\rm ph}/\partial V = -\sum_k (\partial\omega_k/\partial V)[\hbar/2 + \hbar n(\omega_k)] = \sum_k \gamma_k \cdot u_k(T) / V, $$ where $γ_{k} = - d ln ω_{k} / d ln V$ is the mode Grüneisen parameter and $u_{k} (T) = ℏ ω_{k} [1/2 + n (ω_{k})]$ is the mode energy. Defining a thermal-average Grüneisen parameter $γ$ as the heat-capacity-weighted mean of $γ_{k}$ : $$ \gamma = \sum_k \gamma_k c_{V,k} / \sum_k c_{V,k}, $$ the pressure can be written $P_{ph} \cdot V = γ \cdot U_{ph} (T)$ + zero-point piece. Differentiating with respect to $T$ at fixed $V$ : $$ (\partial P/\partial T)V = \gamma C_V/V. $$ Combined with the Maxwell-relation triple product $α = κ_{T} (\partial P / \partial T)_{V}$ : $$ \alpha = \gamma C_V \kappa_T / V. $$ This is the Grüneisen relation (Grüneisen 1908 Annalen 26, 393). For most solids, $γ \approx 1$ - $3$ . For metals at low temperature, the relation extends with an electronic contribution: $\alpha V = (\gamma_{\rm ph} C_V^{\rm ph} + \gamma{\rm el} C_V^{\rm el}) \kappa_T$.

Exercise 10 (hard, symbolic).

Sketch the Peierls 1929 derivation of the high-temperature thermal conductivity $κ \propto 1/ T$ for a phonon gas via Umklapp processes.

Hint

Phonon-phonon scattering processes split into Normal (N) processes preserving crystal momentum $ℏ k_{1} + ℏ k_{2} = ℏ k_{3}$ and Umklapp (U) processes with $k_{1} + k_{2} = k_{3} + G$ for a non-zero reciprocal lattice vector $G$ . Only U-processes degrade heat current; their rate is proportional to the number of phonons with momentum near the zone boundary, which scales as $e^{- T_{D} / (2 T)}$ at low $T$ and as $T$ at high $T$ .

Answer

The thermal conductivity in kinetic theory is $κ = (1/3) C_{V} v_{s} ℓ$ , where $ℓ$ is the phonon mean free path. At high temperature, U-processes dominate because the phonon population at the zone boundary is large; the U-process rate is proportional to the number of phonons with $ℏ ω ≳ k_{B} T_{D}$ , which by Bose-Einstein occupation is approximately $T / T_{D}$ at $T ≫ T_{D}$ (using $1/ (e^{x} - 1) \approx 1/ x$ for $x ≪ 1$ ).

The phonon mean free path is therefore $ℓ \propto 1/ T$ . The heat capacity in the same regime is $C_{V} \to 3 N k_{B}$ , independent of $T$ . Combining: $κ \propto C_{V} v_{s} ℓ \propto (const) (const) (1/ T) \propto 1/ T$ . This is the Peierls 1929 result (Peierls 1929 Annalen 3, 1055), and it explains the observed $κ (T) \propto 1/ T$ falloff in pure crystals above their Debye temperature.

At low temperature $T ≪ T_{D}$ , U-processes are exponentially suppressed because they require thermally-activated phonons near the zone boundary: $Γ_{U} \propto e^{- T_{D} / (2 T)}$ . The mean free path can become enormous, limited only by boundary scattering or impurities (the "Casimir regime"). Callaway 1959 Phys. Rev. 113, 1046 gave a complete relaxation-time treatment combining N and U processes; the result is the Callaway model that fits experimental $κ (T)$ data on Ge, Si, and diamond from millikelvin to $1000$ K with two free parameters.

Advanced results Master

The photon-phonon-Debye framework rests on a sequence of named results connecting cavity electromagnetism, crystal vibrations, anharmonic corrections, transport phenomena, and modern condensed-matter and cosmological applications.

Theorem 1 (Planck 1900). Electromagnetic radiation in thermal equilibrium with a blackbody at temperature $T$ has spectral energy density $$ u(\omega, T),d\omega = \frac{\hbar\omega^3}{\pi^2 c^3},\frac{1}{e^{\hbar\omega/k_BT} - 1},d\omega, $$ integrating to the Stefan-Boltzmann law $u (T) = a T^{4}$ with $a = π^{2} k_{B}^{4} / (15 ℏ^{3} c^{3})$ . Planck 1900 Verh. Dtsch. Phys. Ges. 2, 237 introduced the quantum of action $ℏ$ to fit experimental blackbody data; his discrete-energy hypothesis $E = n ℏ ω$ launched quantum mechanics. The unit 11.05.03 develops the full blackbody framework; the photon gas of the present unit is the same physical system with explicit Bose-Einstein occupation derivation.

Theorem 2 (Einstein 1907). The heat capacity of a solid with all $3 N$ vibrational modes at frequency $ω_{E}$ is $$ C_V^{\rm Einstein}(T) = 3 N k_B \left(\frac{T_E}{T}\right)^2 \frac{e^{T_E/T}}{(e^{T_E/T} - 1)^2}. $$ At high temperature $C_{V} \to 3 N k_{B}$ (Dulong-Petit); at low temperature $C_{V} \to 3 N k_{B} (T_{E} / T)^{2} exp (- T_{E} / T)$ , exponentially small. Einstein 1907 Annalen 22, 180 was the first quantum theory of solids, demonstrating that quantisation suppresses the classical equipartition result at low $T$ . Experimental data on diamond, silicon, and metals showed slower-than-exponential falloff, motivating Debye's 1912 refinement.

Theorem 3 (Debye 1912). The heat capacity of a Debye solid of $N$ atoms with Debye temperature $T_{D} = ℏ ω_{D} / k_{B}$ is $$ C_V(T) = 9 N k_B \left(\frac{T}{T_D}\right)^3 \int_0^{T_D/T}\frac{y^4 e^y}{(e^y - 1)^2},dy. $$ Limits: $C_{V} \to 3 N k_{B}$ for $T ≫ T_{D}$ (Dulong-Petit); $C_{V} \to (12 π^{4} /5) N k_{B} (T / T_{D})^{3}$ for $T ≪ T_{D}$ (Debye $T^{3}$ law). Debye 1912 Annalen 39, 789 introduced the linear-dispersion approximation with cutoff, predicting the universal $T^{3}$ scaling that has been confirmed in essentially every crystalline solid measured.

Theorem 4 (Born-von Karman 1912). The exact spectrum of lattice vibrations in a periodic crystal consists of $3 p$ branches per primitive cell (with $p$ atoms in the primitive cell), of which $3$ are acoustic (gapless) and $3 p - 3$ are optical (gapped). At small $∣ k ∣$ , the acoustic branches are linear $ω = v_{s} ∣ k ∣$ ; at the zone boundary the dispersion deviates from linearity, and the optical branches have a finite frequency gap at $k = 0$ . Born-von Karman 1912 Z. Phys. 13, 297 introduced the periodic boundary conditions and the exact dispersion calculation for a 1D diatomic chain. The full theory, developed in Born-Huang Dynamical Theory of Crystal Lattices (Oxford 1954), provides the exact lattice-dynamics framework that the Debye model approximates.

Theorem 5 (Grüneisen 1908). In a solid, the thermal expansion coefficient $α$ , the heat capacity $C_{V}$ , the isothermal compressibility $κ_{T}$ , and the average Grüneisen parameter $γ = - d ln ω / d ln V$ obey the Grüneisen relation $$ \alpha = \gamma,C_V,\kappa_T / V. $$ The thermal expansion is non-zero only because the lattice potential is anharmonic ( $d ω / d V \neq = 0$ ); harmonic crystals do not expand. Grüneisen 1908 Annalen 26, 393 derived the relation empirically; the modern derivation via the quasi-harmonic approximation (Leibfried-Ludwig 1961, Born-Huang 1954 §6) confirms it. Typical values: $γ \approx 1.5$ - $2.5$ for most solids, $γ \approx 1.0$ for diamond, $γ \approx 2.7$ for sodium.

Theorem 6 (Peierls 1929). In a phonon gas, only Umklapp processes (involving a non-zero reciprocal lattice vector $G$ in the momentum conservation $k_{1} + k_{2} = k_{3} + G$ ) degrade heat current. At high temperature $T ≳ T_{D}$ , the U-process scattering rate scales as $Γ_{U} \propto T$ , the phonon mean free path as $ℓ \propto 1/ T$ , and the thermal conductivity as $$ \kappa(T) \propto 1/T \quad (T \gtrsim T_D). $$ At low temperature, U-processes are exponentially suppressed as $Γ_{U} \propto e^{- T_{D} / (2 T)}$ , giving the Casimir-regime mean free path limited by sample size or impurities. Peierls 1929 Annalen 3, 1055 established the N/U distinction and the complete Boltzmann transport theory of phonons; Callaway 1959 PR 113, 1046 gave the modern relaxation-time treatment that fits Ge, Si, and diamond data across the full temperature range.

Theorem 7 (BCS phonon-mediated superconductivity, Bardeen-Cooper-Schrieffer 1957). In a metal with an attractive phonon-mediated interaction $V_{att}$ between electrons of opposite spin near the Fermi surface, the system is unstable to Cooper pairing at critical temperature $$ T_c = 1.14,\Theta_D \exp!\big[-1/(g(E_F) V_{\rm att})\big], $$ with $Θ_{D}$ the Debye temperature setting the cutoff of the phonon spectrum. BCS 1957 Phys. Rev. 108, 1175 showed that the Debye cutoff $ω_{D}$ directly enters the prefactor of $T_{c}$ . The isotope effect $T_{c} \propto M^{- 1/2}$ (Maxwell 1950; Reynolds-Serin-Wright-Nesbitt 1950) confirms the phonon-mediated mechanism: heavier isotopes have lower $ω_{D}$ , hence lower $T_{c}$ . Detailed material-by-material predictions follow McMillan 1968 PR 167, 331 and Allen-Dynes 1975 PRB 12, 905.

Theorem 8 (CMB blackbody spectrum, Mather et al. 1994). The cosmic microwave background is a photon gas at temperature $T_{0} = 2.7255 \pm 0.0006$ K, with measured spectrum agreement to the Planck blackbody form better than $1 0^{- 4}$ across the FIRAS frequency range. Mather et al. 1994 ApJ 420, 439 reported COBE/FIRAS data establishing the most precise blackbody spectrum ever measured. The energy density $u = a T^{4} \approx 4.17 \times 1 0^{- 14}$ J/m $^{3}$ , photon number density $n_{γ} = 2 ζ (3) (k_{B} T)^{3} / (π^{2} ℏ^{3} c^{3}) \approx 4.11 \times 1 0^{8}$ m $^{- 3}$ . Before redshift $z \approx 3400$ , photon energy density exceeded matter energy density and the universe was radiation-dominated.

Theorem 9 (Photonic BEC, Klaers et al. 2010). Photons in a dye-filled optical microcavity acquire an effective non-vanishing chemical potential through thermalisation with the dye molecules, and can Bose-condense at room temperature when the photon density exceeds a critical value. Klaers et al. 2010 Nature 468, 545 reported the first observation. Unlike the equilibrium photon gas (where $μ = 0$ and BEC is excluded), the microcavity system has a conserved photon number set by the cavity loss rate, producing an effective $μ > 0$ and a true thermodynamic BEC phase transition. Magnon BEC (Demokritov et al. 2006 Nature 443, 430) is the analogous result for spin waves driven out of equilibrium.

Theorem 10 (Phononic crystals and metamaterials). Artificial periodic structures with engineered lattice constants in the millimetre-to-micron range can exhibit phonon bandgaps in the GHz-to-THz range, controlling thermal conductivity and acoustic propagation. Kushwaha-Halevi-Dobrzynski-Djafari-Rouhani 1993 PRL 71, 2022 introduced phononic crystals; modern thermoelectric materials such as filled skutterudites (CoSb $_{3}$ , LaCoSb $_{4}$ ) and clathrates (Ba $_{8}$ Ga $_{16}$ Ge $_{30}$ ) exploit Einstein-mode optical phonons of "rattler" atoms to suppress lattice thermal conductivity, raising the dimensionless figure of merit $Z T = S^{2} σ T / κ$ above unity for waste-heat recovery applications (Slack 1995; Snyder-Toberer 2008 Nature Mater. 7, 105).

Synthesis. The photon-phonon-Debye framework is the foundational reason that two physically distinct systems — cavity electromagnetism and crystal lattice vibrations — share an identical statistical-mechanical description: both are gases of massless bosons with $μ = 0$ , and the central insight is that the low-frequency mode density $g (ω) \propto ω^{2}$ controls all leading-order thermodynamic scalings, generalising Planck's 1900 blackbody calculation into the universal Debye $T^{3}$ heat-capacity law of solids. The bridge is between the infinite-mode photon gas of Stefan-Boltzmann $u = a T^{4}$ and the finite-mode phonon gas of Debye $C_{V} \propto T^{3}$ , with the Debye cutoff $ω_{D}$ identifying the typical interatomic spacing.

This is exactly the same Planck-occupation framework applied to two dispersion relations: linear $ω = c k$ (massless, unbounded) for photons, linear $ω = v_{s} k$ (massless, bounded by $ω_{D}$ ) for acoustic phonons. Putting these together with 11.05.01 Bose-Einstein occupation and 11.05.05 the Fermi-gas counterpart, the trio (Bose photons, Bose phonons, Fermi electrons) accounts for the entire low-temperature thermodynamics of insulators and metals: $C_{V} = γ T + β T^{3}$ with $γ$ from Sommerfeld electrons and $β = (12 π^{4} /5) N k_{B} / T_{D}^{3}$ from Debye phonons. The pattern recurs in nuclear matter (Debye-like phonon modes of nuclei in neutron-star crusts), in cold-atom optical lattices (engineered phonon dispersion), in cosmology (CMB photon gas + relic neutrino gas, both with $μ = 0$ and $u \propto T^{4}$ ), in superconductivity (BCS gap exponent depending on $ω_{D}$ ), in thermoelectrics (Einstein-mode rattlers in skutterudites), and in topological phononics (Weyl phonons, phonon Hall effect), all unified by the same Planck-Bose framework.

Full proof set Master

Proposition 1 (Stefan-Boltzmann constant). The photon-gas energy density at temperature $T$ is $u = a T^{4}$ with $a = π^{2} k_{B}^{4} / (15 ℏ^{3} c^{3}) \approx 7.566 \times 1 0^{- 16}$ J/(m $^{3}$ ·K $^{4}$ ).

Proof. Insert the photon mode density $g (ω) = V ω^{2} / (π^{2} c^{3})$ and Bose-Einstein occupation $n (ω) = 1/ (e^{ℏ ω / k_{B} T} - 1)$ into the energy formula: $$ U = \int_0^\infty \hbar\omega,n(\omega),g(\omega),d\omega = \frac{V\hbar}{\pi^2 c^3}\int_0^\infty \frac{\omega^3}{e^{\hbar\omega/k_BT} - 1},d\omega. $$ Substitute $y = ℏ ω / k_{B} T$ : $$ U = \frac{V (k_B T)^4}{\pi^2 (\hbar c)^3}\int_0^\infty \frac{y^3}{e^y - 1},dy = \frac{V (k_B T)^4}{\pi^2 (\hbar c)^3}\cdot \Gamma(4),\zeta(4) = \frac{V (k_B T)^4}{\pi^2 (\hbar c)^3}\cdot 6 \cdot \frac{\pi^4}{90} = \frac{\pi^2 V (k_B T)^4}{15 (\hbar c)^3}. $$ The energy density is $u = U / V = a T^{4}$ with $a = π^{2} k_{B}^{4} / (15 ℏ^{3} c^{3})$ . Numerically $a \approx 7.566 \times 1 0^{- 16}$ J/(m $^{3}$ ·K $^{4}$ ). The surface-emission constant relates by $σ = a c /4 \approx 5.670 \times 1 0^{- 8}$ W/(m $^{2}$ ·K $^{4}$ ). $□$

Proposition 2 (photon-gas equation of state). A photon gas at temperature $T$ satisfies $P V = U /3$ , the ultrarelativistic equation of state.

Proof. The free energy of a photon gas with $μ = 0$ is $$ F = k_B T \cdot V \int_0^\infty \frac{\omega^2}{\pi^2 c^3} \ln(1 - e^{-\hbar\omega/k_BT}),d\omega. $$ Substitute $y = ℏ ω / k_{B} T$ and integrate by parts: $$ F = -\frac{V(k_BT)^4}{3\pi^2(\hbar c)^3}\int_0^\infty \frac{y^3}{e^y - 1},dy = -\frac{\pi^2 V(k_BT)^4}{45(\hbar c)^3} = -aT^4 V/3 = -U/3. $$ Since $F = - P V$ for an open system at $μ = 0$ , $P V = U /3$ . Equivalently, $P = a T^{4} /3$ , giving a temperature-pressure relation that contrasts with the classical ideal gas $P V = N k_{B} T$ . Photon radiation pressure $P_{rad} \approx 2.5 \times 1 0^{- 6}$ Pa at $T = 300$ K rises to $\sim 0.1$ Pa at $T = 1000$ K — appreciable at stellar interior temperatures and dominant inside very massive stars. $□$

Proposition 3 (Debye cutoff frequency). For a Debye solid of $N$ atoms in volume $V$ with isotropic sound speed $v_{s}$ , the Debye cutoff frequency is $ω_{D} = v_{s} (6 π^{2} N / V)^{1/3}$ .

Proof. The Debye approximation replaces the actual phonon dispersion with linear $ω = v_{s} k$ up to a cutoff sphere in $k$ -space, with the cutoff fixed by the total-mode constraint $3 N$ . The mode count (three polarisations included) is $$ 3 N = 3 \cdot \frac{V}{(2\pi)^3} \cdot \frac{4\pi k_D^3}{3} = \frac{V k_D^3}{2\pi^2}. $$ Solving for $k_{D}$ : $k_{D} = (6 π^{2} N / V)^{1/3}$ . Converting via $ω_{D} = v_{s} k_{D}$ : $$ \omega_D = v_s,(6\pi^2 N/V)^{1/3}. $$ The Debye temperature is $T_{D} = ℏ ω_{D} / k_{B}$ . For copper with $N / V = 8.5 \times 1 0^{28}$ m $^{- 3}$ and angle-averaged $v_{s} \approx 3, 570$ m/s, the calculation gives $T_{D} \approx 468$ K versus the experimental $343$ K — the discrepancy reflects the anisotropy of cubic-crystal elasticity that the Debye isotropic approximation ignores. $□$

Connections Master

Blackbody radiation, Planck spectrum, Stefan-Boltzmann, Wien displacement 11.05.03. The photon-gas half of the present unit is a direct restatement of the blackbody framework with the Bose-Einstein occupation $n (ω) = 1/ (e^{ℏ ω / k_{B} T} - 1)$ explicit. The Stefan-Boltzmann constant $a = π^{2} k_{B}^{4} / (15 ℏ^{3} c^{3})$ , the Wien displacement constant $λ_{peak} T = 2.898 \times 1 0^{- 3}$ m·K, and the cosmic microwave background measurement $T_{0} = 2.7255$ K (Mather et al. 1994 ApJ 420, 439) all originate there. The phonon-gas mechanics here generalise to a bounded-mode bosonic system with cutoff $ω_{D}$ , where the Planck integral becomes the Debye function.
Bose-Einstein distribution 11.05.01. Supplies the underlying quantum occupation $n (ω) = 1/ (e^{ℏ ω / k_{B} T} - 1)$ for massless bosons with $μ = 0$ . Both photons and phonons inherit this distribution because their particle number is not conserved — photon absorption/emission at the cavity walls and phonon creation/annihilation by lattice excitation drive $μ = 0$ . The unit 11.05.01 derives the distribution from grand-canonical maximisation; the present unit applies it to specific dispersion relations and mode counts.
Bose-Einstein condensation 11.05.04. Provides the contrast: a Bose gas of conserved particles with $μ \to 0^{-}$ as $T \to T_{c}$ produces a macroscopic ground-state population. Photons and phonons cannot condense at equilibrium because $μ = 0$ is fixed and the equilibrium $N (T)$ is determined by $T$ alone — there is no critical line. The Klaers et al. 2010 Nature 468, 545 photon BEC in a dye-filled microcavity escapes this by enforcing photon-number conservation through cavity confinement, generating an effective $μ > 0$ that does support condensation. Magnon BEC (Demokritov et al. 2006 Nature 443, 430) is the analogous result for spin waves.
Bosonic Fock space and second quantisation 12.13.01. Phonons live as quanta of harmonic-oscillator ladder operators on the bosonic Fock space, with $H = \sum_{k, s} ℏ ω_{k, s} (b_{k, s}^{†} b_{k, s} + 1/2)$ summing over wavevector $k$ and polarisation branch $s$ . The zero-point energy $\sum ℏ ω /2$ produces the $9 N k_{B} T_{D} /8$ ground-state lattice energy computed in Exercise 8, and is the same algebraic structure as the cavity-mode photon Fock space and the electromagnetic-field zero-point energy underlying the Casimir effect.
Hawking radiation 13.06.04. The black-hole horizon emits a thermal photon-gas spectrum at the Hawking temperature $T_{H} = ℏ c^{3} / (8 π GM k_{B})$ , with Stefan-Boltzmann luminosity $L = 4 π R_{s}^{2} σ T_{H}^{4}$ where $R_{s} = 2 GM / c^{2}$ is the Schwarzschild radius. The Hawking framework applies the same photon-gas thermodynamics here to a horizon-localised radiation field, producing one of the deepest connections between thermodynamics, quantum field theory, and gravitation — the entropy $S_{BH} = A / (4 G ℏ)$ identifies horizon area with thermodynamic entropy.

Historical & philosophical context Master

Max Planck 1900 ^[Planck1900] introduced the quantum of action $h$ in Verh. Dtsch. Phys. Ges. 2, 237 to resolve the ultraviolet catastrophe of classical blackbody radiation. His discrete-energy hypothesis $E_{n} = n ℏ ω$ fit the experimental data of Rubens-Kurlbaum 1900 across the infrared and visible spectrum, launching quantum mechanics. The Stefan-Boltzmann law $u = a T^{4}$ had been derived empirically by Stefan 1879 and theoretically by Boltzmann 1884 from thermodynamic arguments; Planck supplied the spectral form $u (ω, T)$ that integrates to it.

Albert Einstein 1907 ^{[Einstein1907]} applied Planck's quantisation hypothesis to the heat capacity of solids in Annalen der Physik 22, 180, treating each atomic vibration as a quantum oscillator at frequency $ω_{E}$ and producing the first quantum theory of crystal heat capacities. The Einstein model correctly recovered the Dulong-Petit limit $3 N k_{B}$ at high temperature and predicted exponential suppression at low temperature. Experimental measurements by Nernst and collaborators 1908-1911 showed that the actual low-temperature falloff was much slower than exponential, motivating a more sophisticated treatment.

Peter Debye 1912 ^[Debye1912] resolved the discrepancy in Annalen der Physik 39, 789 by replacing Einstein's single frequency with a continuous distribution $g (ω) \propto ω^{2}$ extending from 0 to a cutoff $ω_{D}$ fixed by the total-mode count $3 N$ . The resulting Debye $T^{3}$ heat capacity at low temperature reproduced the experimental data quantitatively for diamond, aluminium, copper, and lead, establishing the Debye temperature $T_{D}$ as a single material-specific parameter characterising the lattice vibration spectrum. The same year, Max Born and Theodore von Karman 1912 ^{[BornvonKarman1912]} developed the exact lattice-dynamics framework in Z. Phys. 13, 297, introducing periodic boundary conditions on a finite crystal and deriving the full phonon dispersion $ω (k)$ from the interatomic-spring constants.

The Grüneisen relation $α V = γ C_{V} κ_{T}$ , derived by Eduard Grüneisen 1908 ^{[Grueneisen1908]} in Annalen der Physik 26, 393, connected the thermal expansion of solids to anharmonic corrections to the lattice potential. Felix Bloch 1928 ^[Bloch1928] extended the framework to electron-phonon coupling in Z. Phys. 52, 555, deriving the resistivity formula $ρ \propto T^{5}$ at $T ≪ T_{D}$ that bears his name and underlies modern metal-transport theory. Rudolf Peierls 1929 ^{[Peierls1929]} established the Normal-versus-Umklapp distinction in phonon scattering in Annalen der Physik 3, 1055, showing that only U-processes degrade heat current and providing the foundation of the modern Boltzmann transport theory of phonons. Callaway 1959 ^{[Callaway1959]} gave the modern relaxation-time treatment in Phys. Rev. 113, 1046 that fits experimental thermal-conductivity data from millikelvin to $1000$ K with two free parameters.

Bardeen-Cooper-Schrieffer 1957 ^[BCS1957] in Phys. Rev. 108, 1175 showed that the Debye cutoff $ω_{D}$ controls the exponential prefactor of the superconducting critical temperature via $T_{c} \approx 1.14 Θ_{D} exp [- 1/ (g (E_{F}) V_{att})]$ . The Maxwell-Reynolds isotope effect $T_{c} \propto M^{- 1/2}$ confirmed the phonon-mediated mechanism. McMillan 1968 PR 167, 331 and Allen-Dynes 1975 PRB 12, 905 extended the framework to strong-coupling superconductors with detailed phonon spectra.

The modern era of the photon-phonon framework includes the COBE/FIRAS measurement (Mather et al. 1994) of the cosmic microwave background as a photon gas at $T_{0} = 2.7255$ K, the Klaers et al. 2010 ^[Klaers2010] observation in Nature 468, 545 of out-of-equilibrium photonic BEC in a dye-filled microcavity, the Demokritov et al. 2006 magnon BEC, and the rapid development of phononic crystals and engineered thermoelectrics (skutterudites, clathrates) that exploit Einstein-mode "rattler" phonons to suppress lattice thermal conductivity for waste-heat recovery. The photon-phonon-Debye scaffold of 1900-1912 remains the universal starting point for every quantitative theory of crystal thermodynamics, blackbody radiation, electromagnetic vacuum fluctuations, and lattice transport in modern condensed matter and cosmology.

Bibliography Master

@article{Planck1900,
  author = {Planck, Max},
  title = {Zur {T}heorie des {G}esetzes der {E}nergieverteilung im {N}ormalspectrum},
  journal = {Verhandlungen der Deutschen Physikalischen Gesellschaft},
  volume = {2},
  pages = {237--245},
  year = {1900},
}

@article{Einstein1907,
  author = {Einstein, Albert},
  title = {Die {P}lancksche {T}heorie der {S}trahlung und die {T}heorie der spezifischen {W}{\"a}rme},
  journal = {Annalen der Physik},
  volume = {22},
  pages = {180--190},
  year = {1907},
}

@article{Grueneisen1908,
  author = {Gr{\"u}neisen, Eduard},
  title = {{\"U}ber die thermische {A}usdehnung fester {K}{\"o}rper},
  journal = {Annalen der Physik},
  volume = {26},
  pages = {393--400},
  year = {1908},
}

@article{Debye1912,
  author = {Debye, Peter},
  title = {Zur {T}heorie der spezifischen {W}{\"a}rmen},
  journal = {Annalen der Physik},
  volume = {39},
  pages = {789--839},
  year = {1912},
}

@article{BornvonKarman1912,
  author = {Born, Max and von K{\'a}rm{\'a}n, Theodore},
  title = {{\"U}ber {S}chwingungen in {R}aumgittern},
  journal = {Physikalische Zeitschrift},
  volume = {13},
  pages = {297--309},
  year = {1912},
}

@article{Bloch1928,
  author = {Bloch, Felix},
  title = {{\"U}ber die {Q}uantenmechanik der {E}lektronen in {K}ristallgittern},
  journal = {Zeitschrift f\"ur Physik},
  volume = {52},
  pages = {555--600},
  year = {1928},
}

@article{Peierls1929,
  author = {Peierls, Rudolf},
  title = {Zur kinetischen {T}heorie der {W}{\"a}rmeleitung in {K}ristallen},
  journal = {Annalen der Physik},
  volume = {3},
  pages = {1055--1101},
  year = {1929},
}

@article{BCS1957,
  author = {Bardeen, J. and Cooper, L. N. and Schrieffer, J. R.},
  title = {Theory of superconductivity},
  journal = {Physical Review},
  volume = {108},
  pages = {1175--1204},
  year = {1957},
}

@article{Callaway1959,
  author = {Callaway, J.},
  title = {Model for lattice thermal conductivity at low temperatures},
  journal = {Physical Review},
  volume = {113},
  pages = {1046--1051},
  year = {1959},
}

@article{Klaers2010,
  author = {Klaers, Jan and Schmitt, Julian and Vewinger, Frank and Weitz, Martin},
  title = {Bose-{E}instein condensation of photons in an optical microcavity},
  journal = {Nature},
  volume = {468},
  pages = {545--548},
  year = {2010},
}

@book{BornHuang1954,
  author = {Born, Max and Huang, Kun},
  title = {Dynamical Theory of Crystal Lattices},
  publisher = {Oxford University Press},
  year = {1954},
}

@book{AshcroftMermin1976,
  author = {Ashcroft, N. W. and Mermin, N. D.},
  title = {Solid State Physics},
  publisher = {Holt-Saunders},
  year = {1976},
}

@book{LandauLifshitz1980,
  author = {Landau, L. D. and Lifshitz, E. M.},
  title = {Statistical Physics, Part 1},
  edition = {3},
  publisher = {Pergamon},
  year = {1980},
}

@book{PathriaBeale2021,
  author = {Pathria, R. K. and Beale, P. D.},
  title = {Statistical Mechanics},
  edition = {4},
  publisher = {Elsevier},
  year = {2021},
}

@book{Kittel2005,
  author = {Kittel, Charles},
  title = {Introduction to Solid State Physics},
  edition = {8},
  publisher = {Wiley},
  year = {2005},
}

@book{Ziman1960,
  author = {Ziman, J. M.},
  title = {Electrons and Phonons},
  publisher = {Oxford University Press},
  year = {1960},
}

Prerequisites

11.05.01
11.05.03
11.05.04

Tier anchors

beginner: Eisberg & Resnick, *Quantum Physics of Atoms, Molecules, Solids, Nuclei, and Particles*, 2e (Wiley, 1985), Ch. 11; Ashcroft & Mermin, *Solid State Physics* (Saunders, 1976), Ch. 23
intermediate: Reif, *Fundamentals of Statistical and Thermal Physics* (McGraw-Hill, 1965), §10; Ashcroft & Mermin (1976), Ch. 22-23; Pathria & Beale, *Statistical Mechanics*, 4e (Elsevier, 2021), §7.3-7.4; Kittel, *Introduction to Solid State Physics*, 8e (Wiley, 2005), §5
master: Landau & Lifshitz, *Statistical Physics, Part 1*, 3e (Pergamon, 1980), §64-65; Pathria & Beale (2021), §7.3-7.4; Born & Huang, *Dynamical Theory of Crystal Lattices* (Oxford, 1954); Kittel (2005), §4-5; Ziman, *Electrons and Phonons* (Oxford, 1960)

References

Planck 1900 — Zur Theorie des Gesetzes der Energieverteilung im Normalspectrum · Verh. Dtsch. Phys. Ges. 2, 237-245
Einstein 1907 — Die Plancksche Theorie der Strahlung und die Theorie der spezifischen Wärme · Annalen der Physik 22, 180-190
Grueneisen 1908 — Über die thermische Ausdehnung fester Körper · Annalen der Physik 26, 393-400
Debye 1912 — Zur Theorie der spezifischen Wärmen · Annalen der Physik 39, 789-839
Born & von Karman 1912 — Über Schwingungen in Raumgittern · Z. Phys. 13, 297-309
Bloch 1928 — Über die Quantenmechanik der Elektronen in Kristallgittern · Z. Phys. 52, 555-600
Peierls 1929 — Zur kinetischen Theorie der Wärmeleitung in Kristallen · Annalen der Physik 3, 1055-1101
BCS 1957 — Theory of superconductivity · Phys. Rev. 108, 1175-1204
Callaway 1959 — Model for lattice thermal conductivity at low temperatures · Phys. Rev. 113, 1046-1051
Born & Huang 1954 — Dynamical Theory of Crystal Lattices · Oxford University Press, full monograph
Ashcroft & Mermin 1976 — Solid State Physics · Ch. 22-23 phonons, Debye model, anharmonicity
Landau & Lifshitz 1980 — Statistical Physics, Part 1, 3e · §64-65 photon gas, phonon gas, Debye specific heat
Pathria & Beale 2021 — Statistical Mechanics, 4e · §7.3-7.4 photons and phonons
Klaers et al. 2010 — Bose-Einstein condensation of photons in an optical microcavity · Nature 468, 545-548

Estimated time

beginner: 15m
intermediate: 45m
master: 75m