The Saha Equation

The Saha equation tells you, given only a temperature and an electron density, what fraction of a gas has been stripped of its electrons. It is the ionization counterpart of the Boltzmann excitation law: where Boltzmann distributes a fixed atom among its bound levels, Saha distributes a chemical element among its successive ionization stages — neutral, singly ionized, doubly ionized — by treating the reaction $\mathrm{X}\rightleftharpoons\mathrm{X}^{+}+e^{-}$ as a chemical equilibrium and applying the law of mass action. The new ingredient, absent from Boltzmann, is that ionization sets a free electron loose, and that electron's translational phase space must be counted — which is why the result depends on electron density and carries a factor built from the electron's thermal de Broglie wavelength. Everything the equation does follows from one competition. The Boltzmann factor $e^{-\chi/k_BT}$ suppresses ionization, because the binding energy $\chi$ must be supplied to free the electron. Pulling the other way is entropy: a freed electron gains access to an enormous reservoir of momentum states, the quantum concentration $(2\pi m_e k_BT/h^2)^{3/2}$ divided by $n_e$ . When the gas is hot or the electrons dilute, that reservoir is vast and ionization proceeds even when $k_BT\ll\chi$ . This one balance runs from the surfaces of stars — where it orders the entire spectral sequence OBAFGKM and explains why hydrogen lines peak in A stars rather than the hottest ones — to the early universe, where the same formula sets the redshift at which the cosmos recombined and released the microwave background.

Key takeaways

The Saha equation gives the ratio of number densities of two successive ionization stages of an element in thermal equilibrium as a function of temperature $T$ and electron density $n_e$ .
It is the law of mass action applied to $\mathrm{X}\rightleftharpoons\mathrm{X}^{+}+e^{-}$ : each species is an ideal gas with quantum statistical weights and internal partition functions, and the chemical potentials balance.
The heart of it is the quantum concentration $(2\pi m_e k_BT/h^2)^{3/2}=\lambda_{\mathrm{dB}}^{-3}$ multiplying the Boltzmann suppression $e^{-\chi/k_BT}$ — an entropy reward times an energy cost.
Higher $T$ and lower $n_e$ favour ionization; the explicit density (or pressure) dependence is why stellar spectral type is set by both temperature and surface gravity, not temperature alone.
Chained with the Boltzmann excitation law it explains spectral-line strengths — the hydrogen Balmer maximum near $T\approx9500$ K in A stars — and underpins the Harvard classification.
The identical formula, applied to primordial hydrogen, places cosmological recombination near $z\approx1360$ and the last-scattering surface of the CMB near $z\approx1100$ .
Key idea: ionization balance is a chemical equilibrium — the ratio of successive stages equals a Boltzmann suppression $e^{-\chi/k_BT}$ times an entropy-driven phase-space factor $\propto(m_e k_BT)^{3/2}/(n_e h^3)$ counting the translational states available to the freed electron.

The core insight

Ionization balance is a chemical equilibrium: the ratio of successive ionization stages equals a Boltzmann suppression $e^{-\chi/k_BT}$ times an entropy-driven phase-space factor $\propto (m_e k_B T)^{3/2}/(n_e h^3)$ counting the translational states available to the freed electron.

Where this lives today

The Saha equation is not a historical relic; it is a working subroutine inside nearly every model of a star or the early universe. Every stellar-atmosphere code — ATLAS, MARCS, PHOENIX, and the pipelines behind surveys such as APOGEE, GALAH, and Gaia — solves it, chained across successive ionization stages and coupled to Boltzmann excitation, to turn a trial temperature and pressure into the level populations that make each spectral line, then iterates until the synthetic spectrum matches the observed one. That inversion is how the temperatures, gravities, and abundances of millions of stars are measured. In stellar interiors a generalized Saha treatment (with pressure ionization and partial degeneracy) fixes the ionization state that sets the opacity — including the crucial $\mathrm{H^{-}}$ opacity in Sun-like stars, which exists only because Saha delivers the right small population of free electrons from metals. In cosmology it gives the textbook recombination redshift; precision CMB analysis needs the non-equilibrium codes (RECFAST, CosmoRec, HyRec), but those are built as corrections to the Saha equilibrium. Wherever a hot gas must decide how many electrons to keep, Saha is the first equation written down.

The physical picture

Think of ionization as evaporation. The ionization energy $\chi$ is a latent heat that must be paid to unbind an electron, and paying it is exponentially unlikely when the thermal energy is small — the Boltzmann suppression $e^{-\chi/k_BT}$ . But a freed electron, like an evaporated molecule, gains a vast reservoir of new states: it can sit anywhere in the gas with any thermal momentum, and the number of quantum "cells" it can occupy per unit volume is the quantum concentration $n_Q=(2\pi m_e k_BT/h^2)^{3/2}$ , one over the cube of its thermal de Broglie wavelength. Ionization is the tug-of-war between this energy cost and entropy prize, and the winner depends on how crowded the electrons already are — what the division by $n_e$ measures.

Figure

Ionization as a tug-of-war: unbinding costs the energy χ and is punished by the Boltzmann factor e^(−χ/k_BT), but the freed electron wins a huge reservoir of translational states n_Q/n_e, where n_Q ≈ 1.1×10²¹ cm⁻³ at T = 5800 K sets the balance

Because the recombining ion must physically find and capture an electron, the recombination rate scales with $n_e$ while the ionization rate does not — squeeze the gas at fixed temperature and recombination wins. That is why the Saha ratio carries an explicit $1/n_e$ the Boltzmann law never has. The classic demonstration of the two competing factors working against each other is the strength of the hydrogen Balmer lines, which need a neutral atom with an electron already excited to $n=2$ : too cold and no atom is excited, too hot and no atom is neutral. The product peaks sharply, and the second figure shows where.

Figure

The Balmer competition: the Saha neutral fraction N_I/N_total (blue) collapses as hydrogen ionizes, while the Boltzmann n = 2 population (green) climbs; their product, the Balmer line strength (amber), peaks near T ≈ 9500 K — spectral type A0 — not in the hottest stars

The core equations

The Saha equation relates the number densities of two adjacent ionization stages as the mass-action ratio for $\mathrm{X}_i\rightleftharpoons\mathrm{X}_{i+1}+e^{-}$ : the product of the "output" densities (higher ion and electron) over the "input" density equals a temperature-dependent equilibrium constant.

\frac{n_{i+1}\,n_e}{n_i} = \frac{2\,Z_{i+1}}{Z_i}\left(\frac{2\pi m_e k_B T}{h^2}\right)^{3/2} e^{-\chi_i/k_B T} \qquad (1)

where: $n_i,\,n_{i+1}$ — number densities of stages $i$ and $i+1$ ; $n_e$ — free-electron density; $Z_i,\,Z_{i+1}$ — internal partition functions; $m_e$ — electron mass; $k_B,\,h$ — Boltzmann and Planck constants; $T$ — temperature; $\chi_i$ — ionization energy from stage $i$ to $i+1$ ; the leading $2$ is the electron spin weight $g_e=2$ . Reading it: raise $T$ and the exponential relaxes toward one, driving ionization up; raise $n_e$ and the ratio falls, because a crowded electron sea drives recombination. Each ionization stage is a ladder of bound levels, so the "statistical weight" of a stage is its internal partition function — a Boltzmann-weighted count of every level.

Z_i(T) = \sum_j g_{ij}\, e^{-E_{ij}/k_B T} \qquad (2)

where: $g_{ij}$ — degeneracy of level $j$ in ion $i$ ; $E_{ij}$ — excitation energy above that ion's ground state. Reading it: at photospheric temperatures only the lowest levels contribute, so for neutral hydrogen $Z_{\mathrm{I}}\approx2$ (the two spin states of $1s$ ) and for the bare proton $Z_{\mathrm{II}}=1$ , making the prefactor $2Z_{\mathrm{II}}/Z_{\mathrm{I}}=1$ . The constant in front of the exponential is the quantum concentration — the inverse cube of the electron thermal de Broglie wavelength — and recognizing it turns algebra into physics.

\lambda_{\mathrm{dB}} = \frac{h}{\sqrt{2\pi m_e k_B T}} \quad\Rightarrow\quad \left(\frac{2\pi m_e k_B T}{h^2}\right)^{3/2} = \lambda_{\mathrm{dB}}^{-3} \equiv n_Q \qquad (3)

where: $\lambda_{\mathrm{dB}}$ — electron thermal de Broglie wavelength; $n_Q$ — quantum concentration, the translational quantum cells per unit volume. Reading it: the Saha ratio contains $n_Q/n_e$ , so ionization is favoured precisely when the quantum concentration greatly exceeds the electron density — the classical, non-degenerate regime where the freed electron has an abundance of empty states. Dividing equation (1) by $n_e$ isolates the ionization ratio; in stellar atmospheres one usually writes it with the electron pressure $P_e=n_e k_B T$ , the measured variable.

\frac{n_{i+1}}{n_i} = \frac{2\,Z_{i+1}}{Z_i}\,\frac{1}{n_e}\left(\frac{2\pi m_e k_B T}{h^2}\right)^{3/2} e^{-\chi_i/k_B T} = \frac{2\,Z_{i+1}}{Z_i}\,\frac{k_B T}{P_e}\left(\frac{2\pi m_e k_B T}{h^2}\right)^{3/2} e^{-\chi_i/k_B T} \qquad (4)

where: $n_{i+1}/n_i$ — ratio of successive stages; $P_e=n_e k_B T$ — electron partial pressure. Reading it: the $1/n_e$ (or $1/P_e$ ) makes explicit that ionization is not set by temperature alone — a low-gravity giant (small $P_e$ ) shows ions a dwarf of the same $T$ does not, the luminosity effect that lets a spectrum reveal surface gravity. For a pure-hydrogen plasma, charge neutrality collapses the formula onto a single ionization fraction $x$ — the form used in cosmology.

\frac{x^2}{1-x} = \frac{1}{n}\left(\frac{2\pi m_e k_B T}{h^2}\right)^{3/2} e^{-\chi/k_B T} \qquad (5)

where: $x=n_e/n=n_p/n$ — ionization fraction; $n=n_p+n_H$ — total (baryonic) hydrogen density; $\chi=13.6$ eV. Reading it: with $g_p=1$ , $g_e=2$ , $g_H=2$ the statistical prefactor is again exactly $1$ ; the left side rises from $0$ to $\infty$ as $x:0\to1$ , so the equation has a unique root at any $T$ and $n$ .

Derivation ASaha from the law of mass action

The goal is to derive equation (1) from first principles by treating ionization as a chemical reaction in equilibrium. We picture the neutral atom, the ion, and the free electron as three ideal gases sharing one temperature, write the chemical potential of each from statistical mechanics, and impose that the potentials balance. The pay-off is that every piece of Saha — the factor of two, the partition-function ratio, the electron mass alone in the quantum concentration, the Boltzmann exponential — falls out of one line of bookkeeping.

Step 1. Impose chemical equilibrium on $\mathrm{X}_i\rightleftharpoons\mathrm{X}_{i+1}+e^{-}$ : as for any reaction, the chemical potentials balance across the arrow, $\mu_i=\mu_{i+1}+\mu_e$ . Why: the chemical potential is the free-energy cost of adding one particle; when left equals right, the reaction has no net direction and the populations are steady.

Step 2. Write the ideal-gas chemical potential of each non-degenerate species $s$ , $e^{\mu_s/k_BT}=(n_s/Z_s)(h^2/2\pi m_s k_BT)^{3/2}e^{E_s^{0}/k_BT}$ . Why: this Sackur–Tetrode result has the quantum concentration counting translational states and $Z_s$ counting internal ones — together the phase space one particle commands.

Step 3. Fix the energy zero so freeing the electron to rest costs $\chi_i$ : stage $i$ sits a depth $\chi_i$ below the ion-plus-free-electron system, so combining the three exponentials leaves net energy $-\chi_i$ in the exponent. Why: the ionization energy is the only energy difference in the problem, and this is where it enters as $e^{-\chi_i/k_BT}$ .

Step 4. Substitute into $\mu_i=\mu_{i+1}+\mu_e$ and solve for $n_{i+1}n_e/n_i$ . The ions have nearly equal masses $m_i\approx m_{i+1}$ , so their quantum concentrations cancel and only the electron's survives. Why: the ion barely changes mass on losing one light electron, so the density dependence and de Broglie wavelength in Saha are electron properties.

Step 5. Collect the weights: the electron's spin degeneracy $g_e=2$ gives the leading factor of two, and the ions' internal structure enters only through $Z_{i+1}/Z_i$ , giving equation (1). Why: this separates the universal spin factor from element-specific partition functions, so one master formula serves every element with only $\chi_i$ and the $Z$ 's changed. *The surviving quantum concentration equals $\lambda_{\mathrm{dB}}^{-3}=1.07\times10^{21}\ \mathrm{cm^{-3}}$ at $5800$ K.*

Go deeper · Derivation B · 1 minThe Balmer maximum and the A-star puzzle

Derivation BThe Balmer maximum and the A-star puzzle

The goal is to explain a fact that puzzled early astronomers: hydrogen is the most abundant element in stars, yet its Balmer absorption lines are strongest not in the hottest stars but in the middling A stars near $9500$ K. We track the one level that makes Balmer lines — the $n=2$ level of neutral hydrogen — as the product of a Boltzmann factor (is that level occupied) and a Saha factor (is the atom neutral at all). The two pull in opposite directions as temperature rises, and their product is a sharp peak.

Step 1. Write the fraction of neutral hydrogen in $n=2$ from Boltzmann, $N_2/N_1=(g_2/g_1)\,e^{-(E_2-E_1)/k_BT}$ , with $g_1=2$ , $g_2=8$ , $E_2-E_1=13.6(1-1/4)=10.2$ eV. Why: Balmer absorption starts from $n=2$ , so an atom is useless for Balmer lines unless its electron is already lifted there — rare until $k_BT$ approaches $10.2$ eV.

Step 2. This Boltzmann fraction rises steeply — from negligible at $5000$ K to a few parts in $10^5$ near $10^4$ K — because the $n=2$ level lies high above ground. Why: on its own, hotter would always mean more Balmer absorption, wrongly predicting the strongest hydrogen lines in the hottest O stars.

Step 3. Fold in Saha at photospheric $P_e\approx20\ \mathrm{N\,m^{-2}}$ : as $T$ climbs past $\sim9000$ K the ratio $n_{\mathrm{II}}/n_{\mathrm{I}}$ of equation (4) shoots through unity, so the neutral fraction collapses. Why: above $\sim10^4$ K almost every hydrogen atom is a bare proton with no bound electron and no Balmer line, so ionization erases the lines excitation was strengthening.

Step 4. Multiply: the Balmer strength tracks $N_2/N_{\mathrm{total}}=(N_2/N_{\mathrm{I}})_{\mathrm{Boltz}}\times(N_{\mathrm{I}}/N_{\mathrm{total}})_{\mathrm{Saha}}$ , a maximum near $T\approx9500$ K (A0). Why: the maximum is a competition point, not a resonance, and it moves with electron pressure — which is why line strengths also encode surface gravity. *The Balmer peak falls at $T\approx9500$ K (A0); evaluating equation (5) gives $n_{\mathrm{II}}/n_{\mathrm{I}}=1$ at $T\approx9550$ K for $P_e=20$ Pa, just above the peak.*

Go deeper · Derivation C · 2 minThe redshift of cosmological recombination

Derivation CThe redshift of cosmological recombination

The goal is to use the pure-hydrogen Saha equation (5) to find when the expanding universe first became neutral, releasing the photons we now see as the cosmic microwave background. We insert the cosmic baryon density and photon temperature — each a known function of redshift — set the ionization fraction to one half, and solve for $z$ . The striking result is that recombination waits until the temperature has fallen to a mere $0.3$ eV, far below hydrogen's $13.6$ eV binding, and understanding why exposes the entropy factor at work on the largest scale.

Step 1. Parametrize the epoch by redshift $z$ : the baryon density dilutes as $n=n_0(1+z)^3$ and the temperature cools as $T=T_0(1+z)$ with $T_0=2.725$ K, the baryon-to-photon ratio fixed at $\eta\approx6\times10^{-10}$ . Why: both density and temperature in equation (5) are tied to the single variable $z$ , so the ionization fraction becomes a function of $z$ alone.

Step 2. Ask why recombination is not simply at $k_BT=13.6$ eV. Because photons outnumber baryons by $\eta^{-1}\sim10^{9}$ , even when the mean photon energy is far below $13.6$ eV the enormous number of Wien-tail photons still supplies more than one ionizing photon per atom. Why: this photon reservoir is the cosmological entropy factor — the same $n_Q/n_e$ competition keeping stellar photospheres ionized, here holding the whole universe ionized long past naive energetics.

Step 3. Set $x=1/2$ (so $x^2/(1-x)=1/2$ ) and solve equation (5) numerically with the cosmic baryon density. The root sits at $T\approx3700$ K, $z\approx1360$ , $k_BT\approx0.3$ eV. Why: the universe recombines a factor of forty in energy below the binding energy, purely because of the photon-to-baryon ratio.

Step 4. Push colder and the free-electron density plummets, the Thomson optical depth drops below unity, and photons free-stream to us: last scattering forms near $z\approx1100$ , just after half-ionization. Why: it is set by when the remaining electrons become too sparse to scatter, so it lags half-ionization; the exact value needs the Peebles equations, but Saha fixes the epoch. *Numerical solution of equation (5) gives $x=1/2$ at $z\approx1360$ – $1380$ ( $T\approx3700$ K), with last scattering near $z\approx1090$ .*

Limiting cases

The Saha ratio $n_{i+1}n_e/n_i$ tells its story through four limits.

Case	Result	Meaning
High temperature, $k_BT\gg\chi$	$\dfrac{n_{i+1}}{n_i}\to\dfrac{2Z_{i+1}}{n_e Z_i}\left(\dfrac{2\pi m_e k_BT}{h^2}\right)^{3/2}$	Boltzmann factor $\to1$ ; the gas is fully ionized and the ratio is limited only by phase space and $n_e$ .
Low temperature, $k_BT\ll\chi$	$\dfrac{n_{i+1}}{n_i}\propto e^{-\chi/k_BT}\to0$	Exponential suppression dominates; the lower ionization stage is essentially complete.
High electron density	$\dfrac{n_{i+1}}{n_i}\propto\dfrac{1}{n_e}$	Pressure ionization reversed — a dense gas recombines; at fixed $T$ more electrons force the lower stage.
Half-ionization point	$n_e=\dfrac{2Z_{i+1}}{Z_i}\left(\dfrac{2\pi m_e k_BT}{h^2}\right)^{3/2}e^{-\chi/k_BT}$	The condition $n_{i+1}=n_i$ ; defines the temperature at which a species is 50% ionized for a given $n_e$ .

Common trapReading ionization off temperature alone

The most common Saha mistake is to treat the ionized fraction as a function of temperature only. It is not: the ratio depends on $n_e$ (equivalently $P_e$ ) just as strongly. Raise the pressure at fixed temperature and you push the equilibrium back toward recombination — the luminosity effect that lets a spectroscopist read a star's surface gravity from its ionization balance. Two further traps: do not confuse the internal partition functions $Z$ (which sum bound levels with their degeneracies) with the translational quantum concentration (which counts free-particle momentum cells). And the leading factor of two is the electron spin degeneracy $g_e=2$ , not a stray constant; drop it and every ionization fraction is wrong by a factor of two.

Worked examples

1 · Hydrogen ionization in the solar photosphere (fully worked)

The solar photosphere has $T=5800$ K and an electron pressure $P_e\approx20\ \mathrm{N\,m^{-2}}$ . What fraction of its hydrogen is ionized?

Step 1. Fix the statistical prefactor. For hydrogen $2Z_{\mathrm{II}}/Z_{\mathrm{I}}=2\cdot1/2=\mathbf{1}$ , so the pressure form (4) reduces to $n_{\mathrm{II}}/n_{\mathrm{I}}=(k_BT/P_e)\,n_Q\,e^{-\chi/k_BT}$ .

Step 2. Evaluate the quantum concentration. $n_Q=(2\pi m_e k_BT/h^2)^{3/2}=\mathbf{1.07\times10^{27}\ m^{-3}}$ at $5800$ K, equivalently $\lambda_{\mathrm{dB}}=0.98$ nm — about the size of an atom, so classical counting is valid.

Step 3. Evaluate the pressure factor. $k_BT=1.38\times10^{-23}\times5800=8.01\times10^{-20}$ J, so $k_BT/P_e=8.01\times10^{-20}/20=\mathbf{4.00\times10^{-21}\ m^{3}}$ ; the running product is $4.00\times10^{-21}\times1.07\times10^{27}=4.27\times10^{6}$ .

Step 4. Evaluate the Boltzmann suppression. In eV, $k_BT=8.617\times10^{-5}\times5800=0.500$ , so $\chi/k_BT=13.6/0.500=27.2$ and $e^{-27.2}=\mathbf{1.52\times10^{-12}}$ .

Step 5. Multiply: $n_{\mathrm{II}}/n_{\mathrm{I}}=4.27\times10^{6}\times1.52\times10^{-12}=\mathbf{6.5\times10^{-6}}$ . The Sun's photospheric hydrogen is only one part in $150{,}000$ ionized — essentially all neutral despite the $5800$ K surface, because the $e^{-27.2}$ energy penalty beats the large phase-space reward.

2 · Where does hydrogen reach 50% ionization?

At the same photospheric pressure $P_e=20\ \mathrm{N\,m^{-2}}$ , find the temperature at which hydrogen is half ionized, $n_{\mathrm{II}}=n_{\mathrm{I}}$ .

Step 1. Set the same expression to unity: $1=(k_BT/P_e)\,n_Q(T)\,e^{-\chi/k_BT}$ . The prefactor scales as $T^{5/2}$ , but the exponential $e^{-157{,}800\,\mathrm{K}/T}$ climbs far faster and dominates.

Step 2. Evaluate near the answer. At $9000$ K the ratio is only $0.31$ ; at $9500$ K it is $0.89$ ; at $9600$ K it has passed $1.09$ . The root — 50% ionization — sits at $\mathbf{T\approx9550\ K}$ .

Step 3. Interpret. Between $5800$ K and $9550$ K the ionized fraction climbs from $6.5\times10^{-6}$ to $0.5$ — nearly five orders of magnitude for less than a factor of two in temperature, the signature of the Boltzmann exponential that makes ionization behave almost like a threshold. The half-ionization temperature $T\approx9550$ K sits just above the Balmer maximum at $9500$ K, confirming that the lines fade because the neutral atoms that make them are ionized away.

Physics of the portrait

The Saha equation is where thermodynamics reached up and took hold of the stars. Before it, the variety of stellar spectra was a taxonomist's puzzle; after it, the spectral sequence became a thermometer and the composition of the Sun became computable.

Origins · 1 minMeghnad Saha and how this was discovered

Meghnad Saha

1893–1956 · Indian astrophysicist; thermal ionization and the physics of stellar atmospheres

Meghnad Saha was born in 1893 in a village near Dhaka, the son of a small shopkeeper, and rose through a colonial education system that repeatedly tried to shut him out. Working in Calcutta and then on a fellowship in Europe, he asked in 1920 a question no one had properly posed: if a stellar atmosphere is a hot gas in thermal equilibrium, what sets the degree of ionization of its atoms? His answer married Nernst's chemical thermodynamics of reacting gases to the new quantum statistics, treating ionization $\mathrm{X}\rightleftharpoons\mathrm{X}^{+}+e^{-}$ as a reaction whose equilibrium constant he could compute. The equation explained at a stroke why the spectral sequence OBAFGKM is a temperature sequence — why calcium and iron lines strengthen in cool stars while helium demands the hottest — turning the empirical Harvard classification into physics. Fowler and Milne quickly sharpened the treatment (Milne recasting it in terms of electron pressure), and in 1925 Cecilia Payne used it in her thesis to reach the then-heretical conclusion that stars, and thus the universe, are made overwhelmingly of hydrogen. Saha returned to India, founded research institutes and a physics journal, was elected to the Royal Society, and late in life served in Parliament. His one equation still opens every graduate treatment of stellar atmospheres and cosmological recombination.

History

1859

— Kirchhoff and Bunsen found spectrum analysis, tying dark lines to elements

1885

— Balmer finds the numerical formula for hydrogen's visible lines

1920

— Saha publishes the theory of thermal ionization and applies it to stellar spectra

1923

— Fowler and Milne recast the theory in terms of electron pressure and marginal appearances

1925

— Payne's thesis uses the Saha equation to show stars are mostly hydrogen

Going further

Texts. Carroll & Ostlie, An Introduction to Modern Astrophysics, 2nd ed., §8.1 (the Saha equation and Balmer line strengths) · Ryden, Introduction to Cosmology, 2nd ed., §9.2 (recombination and the Saha equation) · Dodelson, Modern Cosmology, 2nd ed., §3.3 (recombination, Saha and Boltzmann) · Rybicki & Lightman, Radiative Processes in Astrophysics, §9.5 · Mo, van den Bosch & White, Galaxy Formation and Evolution, §3.5 (thermal history and recombination) · Mihalas, Stellar Atmospheres, 2nd ed., Ch. 5. Key numerical benchmarks: $n_Q=\lambda_{\mathrm{dB}}^{-3}=1.07\times10^{21}\ \mathrm{cm^{-3}}$ at $5800$ K; solar-photosphere fraction $n_{\mathrm{II}}/n_{\mathrm{I}}=6.5\times10^{-6}$ at $P_e=20$ Pa; half-ionization temperature $T\approx9550$ K; Balmer maximum $T\approx9500$ K (A0); cosmological values $k_BT\approx0.3$ eV, $z\approx1360$ for half-ionization, last scattering $z\approx1090$ – $1100$ , $\eta\approx6\times10^{-10}$ , all from equation (5).

Prerequisites. Local Thermodynamic Equilibrium · The Boltzmann Excitation Equation · Blackbody Radiation and the Planck Law. Leads to. Spectral Line Formation · Kirchhoff's Laws and Thermal Emission · Opacity and Absorption Coefficients.

Ask the tutor: Ask me to walk through the Saha–Boltzmann competition for any element and temperature, or to solve the recombination redshift.

Why does the Saha equation depend on electron density but the Boltzmann excitation law does not?

Boltzmann only rearranges the bound electrons of a fixed atom among its own levels — no particle is created, so no free-particle phase space enters and the level ratio depends on temperature alone. Saha describes a reaction that releases a free electron, and its reverse, recombination, requires an ion to capture one of the ambient electrons, so that rate scales with $n_e$ . Balancing a rate that depends on $n_e$ against one that does not leaves an explicit $1/n_e$ in the ratio: the denser the electron sea, the easier ions find partners and recombine, so ionization is harder at higher $n_e$ .

What is the quantum concentration, and why does it appear?

The quantum concentration $n_Q=(2\pi m_e k_BT/h^2)^{3/2}$ is the number of translational quantum states per unit volume available to a thermal electron — one over the cube of its de Broglie wavelength, the phase-space "room" a freed electron can spread into. The Saha ratio always contains $n_Q/n_e$ , comparing the states on offer with the electrons competing for them: when $n_Q\gg n_e$ (a hot, dilute gas) each electron has an enormous choice of empty states, a strong entropy incentive to ionize. When $n_e$ approaches $n_Q$ the gas turns quantum-degenerate and the classical Saha equation must be replaced by a degenerate treatment.

Why is the Sun's surface hydrogen almost entirely neutral if it glows at 5800 K?

Because $5800$ K is thermally feeble compared with hydrogen's $13.6$ eV binding: $k_BT$ is only $0.5$ eV, so $\chi/k_BT\approx27$ and the Boltzmann suppression is $e^{-27}\approx1.5\times10^{-12}$ . Even though the phase-space reward $n_Q/n_e$ is large, it cannot overcome twelve orders of magnitude of exponential penalty, and the worked example gives an ionized fraction of only $6.5\times10^{-6}$ . A temperature you would call "hot" is cold on the scale of atomic binding energies; hydrogen does not ionize substantially until it approaches $10^4$ K. The Sun's free electrons come mainly from metals like iron and calcium, whose ionization energies are only a few eV.

When does the Saha equation break down?

It assumes local thermodynamic equilibrium: enough collisions that a single temperature governs the translational, excitation, and ionization states, with ionization and recombination in detailed balance. In tenuous gases — the corona, planetary nebulae, H II regions — collisions are too rare, and ionization is instead driven by a dilute, non-thermal radiation field; there one solves the full photoionization/recombination rate equations, and ionization can far exceed the Saha prediction. It also fails during rapid change, most importantly cosmological recombination, where expansion outruns equilibrium so the electron fraction "freezes out" above the Saha value. Saha remains the benchmark these non-equilibrium treatments are measured against.

How does the Saha equation reveal a star's surface gravity, not just its temperature?

Through the pressure form: ionization depends on $P_e$ as well as $T$ . Two stars at the same temperature but different gravities have different photospheric pressures — a bloated giant has a low-pressure atmosphere, a compact dwarf a high-pressure one. Lower $P_e$ shifts the balance toward ionization, so a giant shows a higher ionization stage more strongly than a dwarf of the same $T$ . Spectroscopists exploit this "luminosity effect": the ratio of ionized to neutral iron lines measures $P_e$ , hence gravity, hence whether the star is a dwarf, giant, or supergiant — the basis of the luminosity class in the MK system.

Why do cosmologists say Saha is only approximate for recombination?

The equation is exact for a gas in equilibrium; the issue is whether equilibrium holds. In a stellar photosphere collisions and radiation keep the gas in near-perfect balance, so Saha is excellent. In the expanding universe, recombination is a race against expansion. Each direct recombination to the ground state emits a photon energetic enough to immediately re-ionize a neighbour, so net recombination proceeds only through slower channels — two-photon decay from the $2s$ level and the redshifting escape of Lyman- $\alpha$ photons. These bottlenecks make recombination lag equilibrium, leaving a residual ionized fraction Saha would set to zero. Saha still nails the epoch — half-ionization near $z\approx1360$ — but the detailed history CMB experiments need requires the non-equilibrium codes built on top of it.