Kirchhoff's Laws and Thermal Emission

Gustav Kirchhoff, working beside Robert Bunsen in Heidelberg in the 1860s, distilled laboratory spectroscopy into three plain rules. A hot dense body — a solid, a liquid, or an opaque gas — glows with a smooth continuous spectrum. A hot thin gas shines only in bright emission lines at wavelengths fixed by its atoms. And a cool thin gas seen against a hotter continuum stamps dark absorption lines at those same wavelengths. With nothing but these rules, astronomers read the composition and state of the Sun, stars, and nebulae decades before anyone understood the atom. Underneath the three empirical laws lies a single thermodynamic statement — Kirchhoff's law of thermal radiation. For matter in equilibrium with a radiation field, the ratio of what it emits to what it absorbs, $j_\nu/\kappa_\nu$ , is a universal function of temperature and frequency alone: the Planck function $B_\nu(T)$ . Emission and absorption are two faces of the same atomic transitions, and detailed balance forces their ratio to be set by temperature, not the material. Cast into radiative transfer, Kirchhoff's law fixes the source function. When collisions dominate — local thermodynamic equilibrium, LTE — the level populations follow Boltzmann and Saha at the local temperature and the source function collapses to $S_\nu=B_\nu(T)$ . Feed that into the transfer equation and all three spectral types drop out, governed by nothing but optical depth and the run of temperature along the line of sight. This article develops both faces — the observational classification and the law that unifies it — deriving $j_\nu/\kappa_\nu=B_\nu$ two independent ways and following it into stellar atmospheres, the solar Fraunhofer spectrum, and glowing nebulae.

Key takeaways

Kirchhoff's three empirical laws say when a source shows a continuous, emission-line, or absorption-line spectrum, tying each to the density, temperature, and geometry of the gas.
The deeper content is Kirchhoff's law of thermal radiation: in equilibrium $j_\nu/\kappa_\nu=B_\nu(T)$ , the emission-to-absorption ratio is universal, depending only on temperature and frequency.
This makes the source function $S_\nu=j_\nu/(\rho\kappa_\nu)$ equal to $B_\nu(T)$ in LTE, closing the radiative transfer equation.
Whether a line is bright or dark is set by the sign of $B_\nu(T_\mathrm{layer})-I_\nu(0)$ : hotter than background emits, cooler absorbs.
"A good absorber is a good emitter": emissivity equals absorptivity, $\varepsilon_\nu=a_\nu$ , why a matte black object glows and a mirror stays dark.
Key idea: in thermodynamic equilibrium the emissivity and opacity of matter are locked together by $j_\nu/\kappa_\nu = B_\nu(T)$ , so the source function equals the Planck function; the three Kirchhoff spectral types are then just different regimes of optical depth and temperature gradient in the radiative transfer equation.

Where this lives today

Kirchhoff's closure relation is the beating heart of every model stellar atmosphere. Codes such as ATLAS, MARCS, and PHOENIX integrate the transfer equation through a temperature-versus-optical-depth structure, setting $S_\nu=B_\nu(T)$ layer by layer in LTE and solving statistical-equilibrium equations where it breaks down. Fitting the synthetic spectra to observed line depths is how the Gaia-ESO, APOGEE, and GALAH surveys measure temperatures, gravities, and abundances for millions of stars — the chemical cartography of the Galaxy rests on getting the source function right. At low density, the same law explains why H II regions and planetary nebulae glow in forbidden and recombination lines no LTE gas could produce: the clearest proof that $S_\nu\neq B_\nu$ when radiation rules populations. The law even has a live research frontier in its own foundations. Kirchhoff's equality $\varepsilon_\nu=a_\nu$ assumes reciprocity — that the microscopic physics runs the same forwards and backwards — and in a magnetized or otherwise nonreciprocal medium that assumption fails. Since a 2014 proposal by Zhu and Fan, and its experimental confirmation in magneto-optical and Weyl-semimetal emitters around 2023, engineers have built surfaces whose emissivity and absorptivity genuinely differ at the same frequency and angle — a controlled violation of Kirchhoff's law. Such nonreciprocal thermal emitters promise to push photovoltaic and energy-harvesting devices past limits the classical law would impose: a 160-year-old identity turning into a design parameter.

The physical picture

Send a beam through a slab of gas. At each frequency two things happen along every step: some photons are stolen from the beam (rate set by the opacity $\kappa_\nu$ ) and some are added (rate set by the emissivity $j_\nu$ ). Put that slab inside an isothermal oven and demand equilibrium — the radiation must stay a blackbody whatever the walls are made of — and the two rates are forced into a fixed ratio. That balance is the content of $j_\nu/\kappa_\nu=B_\nu(T)$ : the same transitions that make a gas an efficient thief of photons make it an equally efficient source. The first figure follows one ray through the slab.

Figure

A beam of background intensity I_ν(0) crosses an LTE slab at temperature T and optical depth τ_ν. The slab absorbs a fraction and re-emits its own Planck glow, so the emergent ray is I_ν = I_ν(0) e^(−τ_ν) + B_ν(T)(1 − e^(−τ_ν)). At τ_ν = 1 the mix is I_ν = 0.37 I_ν(0) + 0.63 B_ν(T)

Whether a line is bright or dark is a contest between the line-forming layer and whatever lies behind it. On the Sun the deep, hot photosphere supplies a near-Planck continuum at $T\approx5772\ \mathrm{K}$ ; a cooler layer above emits only its weaker $B_\nu(T_\mathrm{cool})$ in the lines while absorbing the brighter continuum from below, so at line centre the emergent intensity sinks and the line goes dark. Flip the geometry — a hot thin gas with empty space behind it, a nebula — and the same transitions glow in emission. The three Kirchhoff laws are one transfer equation read in three regimes; temperature gradient and optical depth are the only knobs. The second figure is what a spectroscopist measures: a strong absorption line whose core sinks to a computable fraction of the continuum.

Figure

The quantitative punchline. A strong solar line forms its core in a layer at T ≈ 4700 K sitting over a photospheric continuum at T ≈ 5772 K. By the Eddington–Barbier relation the residual intensity at line centre is r = B_ν(4700 K) / B_ν(5772 K) = 0.38 at λ = 589 nm, a line depth of 62 percent

"A good absorber is a good emitter" is the everyday form of the law. Heat a lump of soot and it glows near full Planck brightness; heat a polished mirror alongside it and it barely radiates, because it barely absorbs. Emissivity equals absorptivity, $\varepsilon_\nu=a_\nu$ , at every frequency — and only frequency by frequency, the loophole engineers exploit and the trap novices fall into.

The core equations

A pencil of radiation crossing $\mathrm{d}s$ of matter loses intensity to absorption and gains it from emission — the transfer equation in its rawest form.

\mathrm{d}I_\nu = -\kappa_\nu \rho\, I_\nu\, \mathrm{d}s + j_\nu \rho\, \mathrm{d}s \qquad (1)

where: $I_\nu$ — specific intensity at frequency $\nu$ ; $\kappa_\nu$ — mass absorption coefficient (opacity), cm² g⁻¹; $j_\nu$ — mass emission coefficient, erg s⁻¹ g⁻¹ Hz⁻¹ sr⁻¹; $\rho$ — mass density; $\mathrm{d}s$ — path length. The first term drains the beam in proportion to what is there; the second refills it independently of $I_\nu$ . Dividing by $\kappa_\nu\rho$ and measuring path in optical depth, $\mathrm{d}\tau_\nu=\kappa_\nu\rho\,\mathrm{d}s$ , strips the equation to one dimensionless combination — the source function — so only the ratio of emission to absorption matters.

\frac{\mathrm{d}I_\nu}{\mathrm{d}\tau_\nu} = -I_\nu + S_\nu, \qquad S_\nu \equiv \frac{j_\nu}{\kappa_\nu} \qquad (2)

where: $\tau_\nu$ — optical depth, the number of mean free paths along the ray; $S_\nu$ — source function, the ratio of emissivity to opacity. Intensity relaxes toward $S_\nu$ over one optical depth, so where the medium is thick, $I_\nu$ reports $S_\nu$ . Kirchhoff's law of thermal radiation closes the equation by naming the source function. In equilibrium the ratio is not a property of the gas but a universal function of temperature and frequency.

\frac{j_\nu}{\kappa_\nu} = B_\nu(T) \qquad \Longrightarrow \qquad S_\nu = B_\nu(T) \qquad (3)

where: $B_\nu(T)$ — the Planck function at the local temperature $T$ . In LTE the radiation is fed by a source that is locally a blackbody, whatever atoms and ions carry the opacity. The universal function is Planck's, quoted as the value the source function takes in equilibrium.

B_\nu(T) = \frac{2h\nu^3}{c^2}\,\frac{1}{e^{h\nu/k_BT}-1} \qquad (4)

where: $h$ — Planck constant; $c$ — speed of light; $k_B$ — Boltzmann constant; $T$ — thermodynamic temperature. It rises as $\nu^2$ in the Rayleigh–Jeans limit and falls exponentially in the Wien limit, peaking near $h\nu\approx2.82\,k_BT$ . Integrating equation (2) gives the formal solution: the far-boundary intensity attenuated across the full optical depth, plus the source function summed over every layer with its own attenuation to the observer.

I_\nu(\tau_\nu) = I_\nu(0)\,e^{-\tau_\nu} + \int_0^{\tau_\nu} S_\nu(t)\,e^{-(\tau_\nu - t)}\,\mathrm{d}t \qquad (5)

where: $I_\nu(0)$ — incident intensity at the far boundary (background source); $t$ — running optical-depth variable. This is the master relation; every Kirchhoff law is a special case. For a uniform LTE slab the source function is the constant $B_\nu(T)$ , and the integral collapses to a closed form — the working equation for line and continuum formation in one layer.

I_\nu = I_\nu(0)\,e^{-\tau_\nu} + B_\nu(T)\left(1 - e^{-\tau_\nu}\right) \qquad (6)

where: $T$ — uniform slab temperature. As $\tau_\nu\to0$ the slab is transparent, $I_\nu\to I_\nu(0)$ ; as $\tau_\nu\to\infty$ the background is buried and $I_\nu\to B_\nu(T)$ , a blackbody. At the microscopic level a line's opacity and emissivity are built from Einstein coefficients, with the absorption coefficient carrying the stimulated-emission term with a minus sign — the "negative absorption" behind lasers and masers.

j_\nu^{\text{line}} = \frac{h\nu_0}{4\pi}\,n_u A_{ul}\,\phi(\nu),\qquad \kappa_\nu^{\text{line}}\rho = \frac{h\nu_0}{4\pi}\big(n_l B_{lu} - n_u B_{ul}\big)\phi(\nu) \qquad (7)

where: $A_{ul},B_{ul},B_{lu}$ — Einstein coefficients for spontaneous emission, stimulated emission, and absorption; $n_u,n_l$ — upper- and lower-level number densities; $\nu_0$ — line-centre frequency; $\phi(\nu)$ — normalized profile, $\int\phi\,\mathrm{d}\nu=1$ . Their ratio, worked out below, is exactly $B_\nu(T)$ once populations are thermal.

The derivations

Derivation AKirchhoff's law from thermodynamic equilibrium in a cavity

The goal is to prove that the ratio of emission to absorption is a universal function of temperature and frequency, forced to equal the Planck function independent of what the matter is made of. The argument is Kirchhoff's own and almost embarrassingly economical: put any gas element inside a sealed isothermal cavity, demand the equilibrium blackbody field, and require the element neither heat nor cool. Those three statements pin the ratio, with no fact about atomic structure.

Step 1. Place an element of arbitrary matter inside a cavity whose walls are held at temperature $T$ and let it equilibrate. Along any ray the transfer equation (1) gives no net change, because in equilibrium the field is stationary: $\mathrm{d}I_\nu/\mathrm{d}s=-\kappa_\nu\rho I_\nu+j_\nu\rho=0$ . Why: if the intensity changed along any direction, a temperature difference could be tapped to run a perpetual motion machine, which thermodynamics forbids.

Step 2. Identify the equilibrium field. Inside an isothermal cavity the radiation is the blackbody field, isotropic and universal, so $I_\nu=B_\nu(T)$ everywhere. Why: its independence from wall material is a second-law result — two cavities at the same $T$ joined by a filter would otherwise transfer heat between equal temperatures.

Step 3. Substitute and solve. Putting $I_\nu=B_\nu(T)$ into Step 1 gives $j_\nu\rho=\kappa_\nu\rho\,B_\nu(T)$ , and dividing out the density leaves $j_\nu/\kappa_\nu = B_\nu(T)$ , i.e. $S_\nu=B_\nu(T)$ . Why: the emission needed to replace what a layer absorbs from a blackbody field is proportional to that intensity, with the opacity as the constant.

Step 4. Read off universality. Because $B_\nu(T)$ contains only $T$ and $\nu$ , the ratio $j_\nu/\kappa_\nu$ is the same for every material in equilibrium: a strong absorber is necessarily a strong emitter. Why: this is "emissivity equals absorptivity," and why a blackbody is the brightest possible thermal emitter at its temperature. *Verified: the balance $j_\nu=\kappa_\nu\rho B_\nu$ reduces to $S_\nu=B_\nu$ with no residual material dependence; the derivation below reconstructs the same result from Einstein coefficients, an independent route.*

Go deeper · Derivation B · 1 minThe same law from Einstein coefficients and detailed balance

Derivation BThe same law from Einstein coefficients and detailed balance

The aim here is to rebuild Kirchhoff's law from the atom outward, showing that the thermodynamic result of the previous derivation is reproduced exactly by the microscopic rates of absorption, spontaneous emission, and stimulated emission once the populations are thermal. This route exposes precisely which assumption does the work — the Boltzmann populations — and therefore when the law must fail, while forcing the stimulated-emission term into the open.

Step 1. Balance the rates for a two-level atom bathed in mean intensity $\bar J_\nu$ . In statistical equilibrium the upward and downward rates match: $n_l B_{lu}\bar J_\nu = n_u A_{ul} + n_u B_{ul}\bar J_\nu$ , the last term stimulated emission. Why: every absorption promotes an atom and every emission demotes one, so a steady state requires this equality.

Step 2. Solve for the field that sustains the balance: $\bar J_\nu = \dfrac{A_{ul}/B_{ul}}{(n_l B_{lu})/(n_u B_{ul}) - 1}$ . In genuine equilibrium this must be the blackbody field $B_\nu(T)$ . Why: as in the cavity argument, the atoms equilibrate only with a field that is itself the equilibrium field.

Step 3. Impose thermal populations. Boltzmann gives $n_u/n_l=(g_u/g_l)e^{-h\nu/k_BT}$ , so $(n_l B_{lu})/(n_u B_{ul})=(g_l B_{lu})/(g_u B_{ul})\,e^{h\nu/k_BT}$ . Matching to $B_\nu=(2h\nu^3/c^2)/(e^{h\nu/k_BT}-1)$ for all $T$ demands the Einstein relations $g_l B_{lu}=g_u B_{ul}$ and $A_{ul}/B_{ul}=2h\nu^3/c^2$ . Why: exponential and prefactor must line up term by term, and since the coefficients are atomic constants, the relations hold always.

Step 4. Return to the transfer coefficients (7): their ratio is $S_\nu=j_\nu/(\kappa_\nu\rho)=\dfrac{n_u A_{ul}}{n_l B_{lu}-n_u B_{ul}}$ , and inserting the Einstein relations and Boltzmann populations gives back $S_\nu=B_\nu(T)$ . Why: the stimulated term $-n_u B_{ul}$ builds the Planck denominator $e^{h\nu/k_BT}-1$ ; drop it and the law breaks. *Verified: with $g_lB_{lu}=g_uB_{ul}$ , $A_{ul}=(2h\nu^3/c^2)B_{ul}$ , and $n_u/n_l=(g_u/g_l)e^{-h\nu/k_BT}$ , the ratio collapses to $B_\nu(T)$ identically — matching the cavity derivation.*

Go deeper · Derivation C · 1 minEmission versus absorption lines from an isothermal slab

Derivation CEmission versus absorption lines from an isothermal slab

The purpose of this derivation is to show that the single isothermal-slab solution (6) already contains all three Kirchhoff spectral types, making the emission-or-absorption decision explicit as the sign of one bracket. Rather than three rules to memorize, there is one contrast formula whose sign flips with whether the slab is brighter or fainter than its background — the engine behind the solar-line figure and every stellar-atmosphere calculation.

Step 1. Start from the slab solution $I_\nu = I_\nu(0)e^{-\tau_\nu} + B_\nu(T)(1-e^{-\tau_\nu})$ for a uniform LTE layer of temperature $T$ backlit by $I_\nu(0)$ . Why: this is the closed form of the master equation (5) with the constant $S_\nu=B_\nu(T)$ that LTE guarantees.

Step 2. Rearrange into contrast form by adding and subtracting $I_\nu(0)$ : $I_\nu = I_\nu(0) + \big(B_\nu(T)-I_\nu(0)\big)\big(1-e^{-\tau_\nu}\big)$ . The factor $(1-e^{-\tau_\nu})\ge0$ grows with $\tau_\nu$ and peaks at line centre. Why: this isolates the sign of the line as the sign of the bracket $B_\nu(T)-I_\nu(0)$ .

Step 3. Read the emission case. If $B_\nu(T)>I_\nu(0)$ — the slab brighter than what lies behind it, or nothing behind it — the bracket is positive and $I_\nu$ rises above the continuum most where $\tau_\nu$ peaks: a bright emission line. Why: a hot thin gas against the dark adds photons to an empty beam, so highest opacity means highest emission.

Step 4. Read the absorption case. If $B_\nu(T_\mathrm{slab})<I_\nu(0)$ — a cool layer over a hotter continuum — the bracket is negative and $I_\nu$ dips below the continuum at line centre: a dark Fraunhofer line. Why: the cool layer replaces bright background photons with its own fainter emission, most efficiently where it is most opaque.

Step 5. Read the null case. If $B_\nu(T)=I_\nu(0)$ the bracket vanishes and no line forms at any $\tau_\nu$ : a gas in equilibrium with its own field shows a featureless continuum — why deep isothermal layers emit pure blackbody spectra. Why: with source and background equally bright, nothing distinguishes opaque from transparent frequencies. *Verified: $I_\nu(0)=0$ with $\tau_\nu\ll1$ gives $I_\nu\approx\tau_\nu B_\nu$ (emission grows with opacity); $B_\nu<I_\nu(0)$ with large line $\tau_\nu$ drives $I_\nu\to B_\nu$ at line centre (saturated absorption).*

Limiting cases

The master relation (5) with $S_\nu=B_\nu(T)$ reproduces each Kirchhoff law, and the Planck function itself has two well-known limits that shape how the lines appear at radio versus optical frequencies:

Case	Result	Meaning
Optically thick, LTE	$\tau_\nu \gg 1 \Rightarrow I_\nu \to B_\nu(T)$	A hot dense source in equilibrium emits the smooth Planck continuum — Kirchhoff's first law.
Optically thin, hot, no background	$\tau_\nu \ll 1,\ I_\nu(0)=0 \Rightarrow I_\nu \approx \tau_\nu B_\nu(T)$	Intensity peaks where $\kappa_\nu$ (hence $\tau_\nu$ ) peaks — bright emission lines. Kirchhoff's second law.
Cool layer over hot continuum	$I_\nu = I_\nu(0)e^{-\tau_\nu} + B_\nu(T_\ell)(1-e^{-\tau_\nu}),\ T_\ell<T_\mathrm{back}$	At line cores $\tau_\nu$ is large, so $I_\nu$ falls toward $B_\nu(T_\ell)<I_\nu(0)$ — dark lines. Kirchhoff's third law.
Non-LTE, radiation-dominated gas	$S_\nu \neq B_\nu(T)$	Populations set by radiation, not collisions; the law fails and lines need statistical equilibrium.
Rayleigh–Jeans ( $h\nu\ll k_BT$ )	$B_\nu \to 2\nu^2 k_BT/c^2$	Intensity tracks $T$ linearly — radio astronomers quote line contrast as a brightness temperature.
Wien ( $h\nu\gg k_BT$ )	$B_\nu \to (2h\nu^3/c^2)e^{-h\nu/k_BT}$	Exponential sensitivity to $T$ makes optical/UV line depths a steep thermometer of the temperature drop.

Common trapThe misconception: that S_ν = B_ν always, so any hot gas glows as a blackbody

Kirchhoff's law $j_\nu/\kappa_\nu=B_\nu(T)$ is not a universal truth about matter and light — it is a statement about matter in local thermodynamic equilibrium, where collisions are frequent enough to hold the level populations at their Boltzmann and Saha values. Derivation B makes the dependence explicit: the result $S_\nu=B_\nu$ rests entirely on inserting thermal populations. In a low-density, radiation-dominated gas the populations are set by the radiation field itself, the source function departs from $B_\nu$ , and a hot thin gas shows bright emission lines rather than a smooth continuum. This is not a rare edge case — it is the normal state in nebulae, coronae, the interstellar medium, and the outer layers of every star, where non-LTE spectral synthesis exists precisely because $S_\nu\neq B_\nu$ . Never assume the source function is Planckian without checking that collisions win.

Worked examples

1 · A selective coating keeps a spacecraft cold (confidence builder)

A flat plate in full sunlight ( $S=1361\ \mathrm{W\,m^{-2}}$ ) absorbs on one face and radiates from both. Compare a black plate ( $a=\varepsilon=1$ ) with a selective white coating that absorbs only $a_\mathrm{s}=0.2$ of sunlight but emits $\varepsilon_\mathrm{IR}=0.9$ in the thermal infrared. If "a good absorber is a good emitter," how can the two numbers differ — and what temperature does each plate reach?

Step 1. Resolve the paradox first. Kirchhoff's law equates emissivity and absorptivity at the same frequency, $\varepsilon_\nu=a_\nu$ . Sunlight arrives in the visible; the plate radiates in the far infrared, near room temperature. These are different frequencies, so $a_\mathrm{s}=0.2$ and $\varepsilon_\mathrm{IR}=0.9$ violate nothing. The equality is per-frequency, and that is the entire design freedom.

Step 2. Balance absorbed against emitted power: $a_\mathrm{s}S$ on the sunlit face, $\varepsilon_\mathrm{IR}\sigma T^4$ from each of two faces. Equating, $T=\left(\dfrac{a_\mathrm{s}S}{2\varepsilon_\mathrm{IR}\sigma}\right)^{1/4}$ , with $\sigma=5.67\times10^{-8}\ \mathrm{W\,m^{-2}\,K^{-4}}$ .

Step 3. Black plate ( $a_\mathrm{s}=\varepsilon_\mathrm{IR}=1$ ): $T=\left(\dfrac{1361}{2\times5.67\times10^{-8}}\right)^{1/4}=(1.20\times10^{10})^{1/4}=\mathbf{331\ K}$ , a toasty $+58\,°\mathrm{C}$ .

Step 4. Selective coating: $T=\left(\dfrac{0.2\times1361}{2\times0.9\times5.67\times10^{-8}}\right)^{1/4}=(2.67\times10^{9})^{1/4}=\mathbf{227\ K}$ , a chilly $-46\,°\mathrm{C}$ . Same geometry, same Sun, over a hundred kelvin colder — because the law leaves visible and infrared emissivities free to differ. Selective surfaces are Kirchhoff's law read frequency by frequency, and why real spacecraft wear white.

2 · Why the same sodium gas gives a dark line in the Sun and a bright line in a nebula

Sodium's D line sits at $\lambda=589\ \mathrm{nm}$ ( $\nu=5.09\times10^{14}\ \mathrm{Hz}$ ). (a) In the Sun the core forms in a layer at $T_\ell\approx4700\ \mathrm{K}$ over a continuum at $T_\mathrm{c}\approx5772\ \mathrm{K}$ ; find the residual intensity at line centre. (b) Put the same sodium in a hot thin nebula at $10^4\ \mathrm{K}$ with empty space behind it and $\tau_\nu=0.1$ ; what does the spectrum show?

Step 1. Use the Eddington–Barbier relation: the emergent intensity samples the source function at optical depth unity, $I_\nu\approx B_\nu\big(T(\tau_\nu=1)\big)$ . The continuum reaches $\tau=1$ deep and hot ( $T_\mathrm{c}$ ); at line centre the huge line opacity pushes $\tau=1$ up into the cool layer ( $T_\ell$ ), so the residual is $r=B_\nu(T_\ell)/B_\nu(T_\mathrm{c})$ .

Step 2. With two Planck functions at the same $\nu$ , the prefactor $2h\nu^3/c^2$ cancels and only the denominators remain: $r=\dfrac{e^{h\nu/k_BT_\mathrm{c}}-1}{e^{h\nu/k_BT_\ell}-1}$ . Evaluate $h\nu/k_B=(6.626\times10^{-34})(5.09\times10^{14})/(1.381\times10^{-23})=2.44\times10^{4}\ \mathrm{K}$ .

Step 3. So $h\nu/k_BT_\mathrm{c}=24\,430/5772=4.23$ and $h\nu/k_BT_\ell=24\,430/4700=5.20$ , giving $r=\dfrac{e^{4.23}-1}{e^{5.20}-1}=\dfrac{67.8}{179.6}=\mathbf{0.38}$ . The core sits at 38 percent of the continuum — a line depth of 62 percent, a strong dark Fraunhofer line, as drawn in the second figure. Note the Wien-limit steepness: a drop of under 1100 K more than halves the intensity.

Step 4. Now the nebula. With no background, $I_\nu(0)=0$ , equation (6) gives $I_\nu\approx\tau_\nu B_\nu=0.1\,B_\nu(10^4\,\mathrm{K})$ , a positive spike at line centre where $\tau_\nu$ peaks — a bright emission line. The identical transition gives a dark line against the Sun's hot continuum and a bright line against the dark of space: the spectrum is set by the background, not the gas. (In a real H II region the populations are radiatively controlled, so LTE $B_\nu$ is only scaffolding — but the sign of the effect is exactly this geometry.)

Physics of the portrait

Kirchhoff's law is where thermodynamics, spectroscopy, and the future quantum theory of radiation first touched. The man who stated it turned a chemist's flame test into the tool that would weigh the Sun's elements — and, without meaning to, wrote down the boundary condition Planck would spend forty years solving.

Origins · 1 minGustav Kirchhoff and how this was discovered

Gustav Kirchhoff

1824–1887 · spectroscopy and the law of thermal emission

Gustav Robert Kirchhoff is remembered by every student for his two circuit rules, formulated at twenty-one, but his deepest mark is on radiation. Working with the chemist Robert Bunsen at Heidelberg from 1859, he showed that each element imprints a fixed pattern of spectral lines, and that the dark Fraunhofer lines in the Sun coincide with the bright lines of terrestrial elements — proving the Sun contains sodium and iron and founding astrophysical spectroscopy at a stroke. In the same year he proved the law that bears his name: for any body in thermal equilibrium the ratio of emissive to absorptive power is a universal function of temperature and wavelength alone. To make the statement exact he defined the perfect absorber — coining "black body" (schwarzer Körper) in 1862 — and posed its universal emission function as a challenge. That challenge stood four decades: Wien found the high-frequency form, Rayleigh and Jeans the low, and only in 1900 did Max Planck supply the full curve, at the cost of quantizing energy. Kirchhoff's law was thus the question whose answer opened the quantum era, and it remains the closure relation $S_\nu=B_\nu$ on which the theory of stellar atmospheres is built.

History

1859

— Kirchhoff states the law of thermal emission: emissive over absorptive power is universal

1860

— Kirchhoff and Bunsen found spectral analysis and identify solar sodium; discover caesium

1861

— the Kirchhoff–Bunsen method yields a second new element, rubidium

1862

— Kirchhoff coins "black body" for the ideal perfect absorber and emitter

1900

— Planck supplies the universal function B_ν(T), quantizing energy to do it

Going further

Texts. Carroll & Ostlie, An Introduction to Modern Astrophysics, 2nd ed., Ch. 5 (Kirchhoff's laws, the source function, radiative transfer) · Rybicki & Lightman, Radiative Processes in Astrophysics, Ch. 1 (Kirchhoff's law, detailed balance, Einstein coefficients) · Mihalas, Stellar Atmospheres, 2nd ed., Ch. 2 (source function, LTE and its breakdown) · Chandrasekhar, Radiative Transfer, Ch. 1 (formal transfer theory) · Ryden, Introduction to Cosmology, Ch. 2 (thermal radiation background).

Provenance. The empirical laws and transfer framing follow Carroll & Ostlie Ch. 5; the cavity and Einstein-coefficient derivations follow Rybicki & Lightman Ch. 1; the LTE discussion follows Mihalas Ch. 2. A verifier should confirm: (1) the cavity argument giving $j_\nu/\kappa_\nu=B_\nu$ with no material dependence; (2) the Einstein relations $g_lB_{lu}=g_uB_{ul}$ , $A_{ul}=(2h\nu^3/c^2)B_{ul}$ and the resulting $S_\nu=B_\nu$ ; (3) the slab solution and its contrast form; (4) the residual $r=(e^{4.23}-1)/(e^{5.20}-1)=0.38$ at 589 nm; (5) the spacecraft temperatures 331 K and 227 K; (6) $T_{\rm eff,\odot}\approx5772$ K (IAU nominal). All equations checked against the cited texts.

Prerequisites. Blackbody Radiation and the Planck Law · Specific Intensity and Radiative Flux · The Equation of Radiative Transfer · Opacity and Absorption Coefficients. Leads to. Emission and the Source Function · Local Thermodynamic Equilibrium · Spectral Line Formation · The Saha Ionization Equation.

Ask the tutor: Ask me to walk through why the very same sodium atoms carve a dark line into the Sun's spectrum but blaze in emission across a nebula, straight from the one contrast formula.

What is the difference between Kirchhoff's three empirical laws and "Kirchhoff's law"?

They share a name but differ. The three empirical laws (continuous, emission-line, absorption-line spectra) are phenomenological rules from 1859 describing what spectrum a source shows given its density and geometry. "Kirchhoff's law of thermal radiation" is the theorem $j_\nu/\kappa_\nu=B_\nu(T)$ underneath them: feed $S_\nu=B_\nu$ into the transfer equation and the three rules fall out as limits of optical depth and temperature gradient. The three laws are the observations; the law is the explanation.

Why does the source function equal the Planck function only in LTE?

Because the equality is inherited from the level populations. Derivation B shows $S_\nu=B_\nu$ follows only after inserting Boltzmann populations, which LTE guarantees by assuming collisions are frequent enough to thermalize the gas locally. When density drops and collisions become rare, the radiation field sets the populations, they depart from Boltzmann, and the source function departs from $B_\nu$ . LTE is therefore not a statement about the radiation being Planckian — it can be far from it — but about the matter being collisionally thermalized, the condition the law needs.

If emission and absorption are locked together, how can a gas cool by radiating?

Kirchhoff's law equates the ratio $j_\nu/\kappa_\nu$ to $B_\nu(T)$ ; it does not say emission equals absorption. A gas cools when it emits more than it absorbs, which happens whenever the local field is weaker than the blackbody field at the gas temperature — an optically thin cloud with cold space around it. There $j_\nu$ still obeys the law, but the absorbed energy $\kappa_\nu\rho I_\nu$ is small because $I_\nu\ll B_\nu$ , so the net is a loss. Only in full equilibrium, $I_\nu=B_\nu$ everywhere, do the two cancel and the gas neither heats nor cools.

Why does the same spectral line appear dark in the Sun but bright in a nebula?

Because the contrast is $\big(B_\nu(T_\mathrm{gas})-I_\nu(0)\big)(1-e^{-\tau_\nu})$ , and only the sign of the bracket decides bright versus dark. In the Sun the line forms in a cool layer over a hotter, brighter continuum, so $B_\nu(T_\mathrm{gas})<I_\nu(0)$ and it dips: absorption. In a nebula the same sodium sits in hot thin gas with empty space behind it, $I_\nu(0)=0$ , so the bracket is positive and it spikes: emission. The transition is identical; the background flips — Kirchhoff's second and third laws, read off one formula.

What actually is the stimulated-emission term doing in the absorption coefficient?

In equation (7) the line opacity is proportional to $n_l B_{lu}-n_u B_{ul}$ , not just $n_l B_{lu}$ . The subtracted piece is stimulated emission: an upper-level atom struck by a photon is induced to emit a second identical one, adding to the beam as negative absorption. It is no small correction — it is the term that builds the Planck denominator $e^{h\nu/k_BT}-1$ in Derivation B, so dropping it breaks the law. When the upper level is overpopulated the net opacity goes negative and you have a maser or laser — the masers seen around evolved stars and star-forming regions.

Can Kirchhoff's law ever be violated on purpose?

The classical law assumes reciprocity, and in a nonreciprocal medium it genuinely fails. Apply a magnetic field, or use a magneto-optical or Weyl-semimetal material, and absorption and emission stop being time-reversal symmetric, so a surface can absorb strongly where it emits weakly — $\varepsilon_\nu\neq a_\nu$ . This does not break thermodynamics: the second law constrains only the total balance over all directions, not the directional equality Kirchhoff assumed. Since 2014 nonreciprocal emitters have become an active field, routing thermal photons to raise the efficiency ceiling of thermophotovoltaic devices.