Detailed Balance and the Planck Spectrum

Detailed balance and the Planck spectrum are woven through modern physics far beyond blackbody radiation. The 2.725 K cosmic microwave background is the most perfect blackbody ever measured — its spectrum matches Planck's law to better than one part in $10^4$ across four decades of the COBE-FIRAS measurement — and tiny departures from a perfect Planck curve encode the physics of the early universe. In astrophysics, the assumption of local thermodynamic equilibrium (detailed balance holding locally even when the star as a whole is not in equilibrium) is the workhorse that lets radiative-transfer codes model stellar atmospheres and infer temperatures, densities, and abundances. In the laboratory, blackbody sources are primary radiometric standards, and the Stefan–Boltzmann law calibrates everything from infrared thermometers to spacecraft thermal design. And the detailed-balance argument itself remains a template: the same logic that forced stimulated emission out of thermodynamic consistency now bounds the efficiency of solar cells (the Shockley–Queisser limit is a detailed-balance argument) and LEDs. Einstein's cavity full of atoms turned out to be a machine for deriving the deepest constraints on how light and matter can exchange energy.

Detailed balance is a stronger statement than "nothing changes." Equilibrium only requires that the total rate into each level equal the total rate out. Detailed balance requires more: that every individual process be balanced by its exact reverse. For a two-level atom in a radiation bath this means the rate of upward transitions (absorption) must equal the combined rate of downward transitions (spontaneous plus stimulated emission), atom by atom, at the transition frequency. That is the entire input. Now watch the machine turn. The absorption and stimulated-emission rates both scale with how much light is present, the spectral energy density $\rho(\nu)$; the spontaneous rate does not. Solving the balance equation for $\rho$ therefore gives a specific function of temperature — but the atoms are also Boltzmann-distributed, $N_1/N_2 = (g_1/g_2)e^{h\nu/k_BT}$, which is fixed independently. The only way these two constraints can agree at every temperature is if the coefficients take particular values, and those values make $\rho$ collapse to exactly the Planck form. The spectrum is not assumed; it is squeezed out. The deepest part is what the ratio $A/B$ turns out to be. It equals $8\pi h\nu^3/c^3$, which is $h\nu$ times the density of electromagnetic modes per unit volume per unit frequency, $8\pi\nu^2/c^3$. That mode density is the same quantity statistical mechanics counts when it derives Planck's law by stuffing photons into cavity modes (owned by aq-sm-photon-gas-planck-radiation). Two utterly different arguments — one balancing atomic rates, one counting field modes — meet at the same number. Spontaneous emission, in this light, is just stimulated emission driven by the one photon of zero-point field that every mode carries. The Planck curve that results has two revealing edges. At low frequency it becomes the classical Rayleigh–Jeans law, which rises as $\nu^2$ without limit — the "ultraviolet catastrophe" that classical physics could not escape. At high frequency it becomes the exponentially-suppressed Wien law. Between them sits a peak that marches to shorter wavelengths as the temperature rises — the Wien displacement law that turns a heated poker from red to white to blue, that puts the Sun's peak in the green at 502 nm, and that places the cosmic microwave background's peak at a millimetre.

Detailed balance sets the total upward rate equal to the total downward rate for a two-level atom in a radiation field. The absorption and stimulated-emission terms are proportional to the field; the spontaneous term is not.

where: $N_1,N_2$ — lower/upper populations; $A_{21}$ — spontaneous rate; $B_{12},B_{21}$ — absorption/stimulated coefficients; $\rho(\nu)$ — spectral energy density. Read it as a balance sheet: absorption lifts atoms up; spontaneous and stimulated emission bring them back down; in equilibrium the two sides are equal at the transition frequency. Solving Eq. (1) for the field, and inserting the thermal (Boltzmann) population ratio, gives the radiation density the atoms are consistent with.

where: $g_1,g_2$ — degeneracies. This must equal the Planck law at every temperature. That is only possible if the coefficient of $e^{h\nu/k_BT}$ in the denominator is exactly one, and the prefactor is exactly the mode density times $h\nu$ — which forces the two Einstein relations.

where: the second relation is $h\nu$ times the free-space mode density $8\pi\nu^2/c^3$. These two identities (owned in detail by aq-qm-spontaneous-emission-einstein-coefficients) are what detailed balance demands. Substituting them back into Eq. (2) yields the Planck spectrum itself.

where: $\bar n$ — mean number of photons per mode (Bose–Einstein). The factorisation is the whole story: energy density = (how many modes) × (energy per photon) × (how many photons per mode). Its two limits reveal the classical failure and its quantum cure.

where: the Rayleigh–Jeans form (left) is the classical equipartition result — $k_BT$ per mode — which diverges as $\nu^2$ (the ultraviolet catastrophe); the Wien form (right) is the exponentially-suppressed high-frequency tail. Integrating the full curve gives the total energy density and the familiar radiation laws.

where: the Wien displacement constant comes from maximising the wavelength form (root $x=4.965$); the Stefan–Boltzmann constant from integrating $u$ over all frequencies. Both are pure combinations of $h$, $c$, and $k_B$ — no atomic detail survives, because the spectrum is a property of the radiation field, not of any particular atom.

Example 1 — the colour of the Sun. Treat the Sun as a blackbody at its effective temperature $T=5772\ \mathrm{K}$ and find the wavelength at which its spectrum peaks.

Step 1. Use the Wien displacement law $\lambda_{\max}T = b = 2.898\times10^{-3}\ \mathrm{m\,K}$, with $b=hc/(4.965k_B)$.

Step 2. Solve: $\lambda_{\max} = b/T = 2.898\times10^{-3}/5772 = 5.02\times10^{-7}\ \mathrm{m}$.

Step 3. Convert: $\lambda_{\max} = \mathbf{502\ nm}$ — green light, remarkably close to the wavelength at which the human eye is most sensitive. The Sun's radiation peaks at 502 nm, in the green, and human vision evolved to peak there too. (Its frequency spectrum peaks near 880 nm — a reminder to match the displacement constant to the variable.) The same law, applied to the 2.725 K cosmic microwave background, gives a peak at $\lambda_{\max}=b/2.725=\mathbf{1.06\ mm}$ — microwaves, the reason it is called the microwave background.

Example 2 — why stimulated emission cannot be dropped. Show that omitting the stimulated-emission term $B_{21}$ from the balance equation destroys the Planck spectrum.

Step 1. Without $B_{21}$, balance reads $B_{12}\rho N_1 = A_{21}N_2$, so $\rho = (A_{21}/B_{12})(N_2/N_1) = (A_{21}/B_{12})(g_2/g_1)e^{-h\nu/k_BT}$.

Step 2. This is the Wien form $\rho\propto e^{-h\nu/k_BT}$ — it has no $-1$ in a denominator, so it fails at low frequency where the real spectrum must approach Rayleigh–Jeans.

Step 3. Restoring $B_{21}$ puts the $B_{21}\rho N_2$ term back, which becomes the $-1$ in the denominator of Eq. (2) and rebuilds the full Planck curve. The $-1$ in the Planck denominator is stimulated emission. Einstein realised in 1917 that a purely absorption-plus-spontaneous model gives only the Wien tail; a third process, stimulated emission, is thermodynamically required to reproduce the full spectrum. He predicted the effect that would later make the laser possible, purely by demanding detailed balance.

Detailed balance is a principle, not a person, and it deserves to be foregrounded as one of the most powerful reasoning tools in physics. Its content is deceptively simple: in true thermodynamic equilibrium, every microscopic process runs at exactly the same rate as its reverse — not merely that the totals balance, but that each forward-and-back pair is individually balanced. This is a consequence of microscopic reversibility (time-reversal symmetry of the underlying dynamics), and it is far stronger than equilibrium alone. Einstein's genius was to see that this principle, applied to atoms and photons, is enough to derive the Planck spectrum without any statistical mechanics at all — and to notice that consistency demanded a process, stimulated emission, that no one had yet observed. The same principle underlies the fluctuation–dissipation theorem, Kirchhoff's law of thermal radiation, and the Onsager reciprocal relations. It is how equilibrium constraints pin down physics that then holds far from equilibrium.

Texts. Foot, Atomic Physics (2005), §1.3 · Loudon, The Quantum Theory of Light, 3rd ed., §1.5 · Demtröder, Laser Spectroscopy, Vol. 1, §2.3 · Saleh & Teich, Fundamentals of Photonics, §13.1 · Hecht, Optics, 5th ed., §13.1.

Provenance (verbatim facts preserved from the base). The verifier reproduces: (1) solving the two-level balance for $\rho(\nu)$ and equating to the Planck form to obtain $g_1B_{12}=g_2B_{21}$ and $A_{21}/B_{21}=8\pi h\nu^3/c^3$ (Foot §1.3; Loudon §1.5); (2) the Rayleigh–Jeans and Wien limits (Demtröder §2.3); (3) the mode density $8\pi\nu^2/c^3$ (Saleh & Teich §13.1); (4) the Wien displacement root $x=4.965$, $b=2.898$ mm·K, Sun 502 nm; (5) Stefan–Boltzmann $\sigma=2\pi^5k_B^4/15h^3c^2=5.670\times10^{-8}$. All constants are CODATA; the argument states no atom-specific value since $A_{21}$ is not fixed by it. Historical attribution: Einstein, Phys. Z. 18, 121 (1917). The mode-counting route is owned by aq-sm-photon-gas-planck-radiation and the A/B relation by aq-qm-spontaneous-emission-einstein-coefficients, both recalled here by name.

Builds on: Einstein A and B Coefficients · Spontaneous Emission · Stimulated Emission · Resonant Absorption. Leads to: Radiative Lifetime and Natural Linewidth · Oscillator Strength · The Electric Dipole Approximation.