White Dwarf Cooling and Crystallization

A white dwarf is the last chapter in the life of every star born with less than about eight solar masses — the exposed, still-glowing core that remains once nuclear burning has stopped for good. Nothing inside it makes new energy. It is held up not by heat but by electron degeneracy pressure, a quantum stiffness that persists all the way down to absolute zero, so the star does not contract and reheat as an ordinary gas sphere would. It simply leaks the heat it was born with into the dark, growing fainter and redder over billions of years. That single fact — a fixed heat reservoir draining through a fixed-size body — turns a white dwarf into a clock: measure how bright it is now, know the rate at which it loses energy, and read off how long it has been cooling. Two threads run through the subject, and this article braids them. The first is a division of labour: the degenerate electrons supply the pressure but store almost no heat, while the classical ion gas stores essentially all the heat but little pressure. The electrons, superb heat conductors, flatten the interior to a nearly uniform temperature, banishing the entire temperature drop to a thin, non-degenerate surface blanket that throttles the escape of energy. Leon Mestel captured this in 1952 in a relation of startling economy, cooling age ∝ L⁻⁵ᐟ⁷. The second thread is what happens when the ion gas gets cold enough to feel its own electrostatics: it freezes. The Coulomb plasma crystallizes into a lattice, releasing latent heat that stalls the cooling and leaves a fingerprint in the census of faint white dwarfs — the physics the Gaia satellite made visible across the sky in 2019.

Key takeaways

A white dwarf has no fusion; its light is powered entirely by the slow radiation of thermal energy stored in its ion core, while degenerate electron pressure holds the structure fixed.
Degenerate electrons conduct heat so well that the interior is nearly isothermal — the whole temperature gradient sits in a thin non-degenerate envelope that acts as the insulating blanket setting the luminosity.
Mestel's model gives cooling age ∝ L⁻⁵ᐟ⁷, so a white dwarf spends the vast majority of its life very faint, and old remnants pile up at low luminosity.
When the ion coupling reaches Γ ≈ 175 the plasma crystallizes into a body-centred-cubic lattice, releasing latent heat (∼ k_B T per ion) and raising the heat capacity to the Dulong–Petit value 3k_B — a pause in the cooling curve.
Below the Debye temperature the lattice heat capacity collapses as T³, the reservoir empties, and cooling accelerates toward the coolest, oldest remnants.
The observed white-dwarf luminosity function ends at a finite faint cutoff that dates the local Galactic disk to ∼ 8–10 Gyr — white-dwarf cosmochronology.
Key idea: a white dwarf is an isothermal degenerate electron frame wrapped in a thin insulating envelope; the ions hold the heat, the envelope throttles its escape, and crystallization latent heat punctuates the cooling — turning the fading remnant into a Gyr-scale clock.

The core insight

A white dwarf is an isothermal degenerate electron reservoir wrapped in a thin insulating envelope; the ions hold the heat, the envelope throttles its escape, and crystallization latent heat punctuates the cooling — turning the fading remnant into a Gyr-scale clock.

Where this lives today

The most striking modern confirmation came in 2019, when Tremblay and collaborators used the Gaia mission — which measured parallaxes and colours for hundreds of thousands of white dwarfs — to expose a pile-up, a bright over-density on the cooling sequence in the colour–magnitude diagram, sitting exactly where theory says C/O cores should be crystallizing and releasing latent heat. A phase transition inside stellar interiors was, for the first time, read directly off a map of the sky. That sharpened cosmochronology: the local luminosity function terminates at a finite faint cutoff — no white dwarf has had time to cool past it — and locating it dates the Galactic disk to roughly 8–10 Gyr, independent of turnoff fitting and of cosmology. Deep Hubble imaging pushes the technique into globular clusters, where the end of the cooling sequence yields ages of 11–13 Gyr and a hard lower bound on the age of the Universe. Asteroseismology of pulsating white dwarfs (the ZZ Ceti class) measures the rate of period change Ṗ/P, tracking the cooling rate directly. The current frontier is a second delay mechanism — settling and distillation of trace species such as ²²Ne — invoked to explain white dwarfs that appear anomalously old.

The physical picture

Picture a hot brick wrapped in a thick blanket. The brick is the degenerate isothermal core; it holds nearly all the heat because the ions inside behave like an ordinary gas, each carrying ³⁄₂ k_B T. The electrons are different: being degenerate, they are like a rigid quantum scaffold that conducts heat superbly but carries almost none of its own, because only the sliver of electrons within k_B T of the Fermi surface can change their energy. That conduction is what makes the deep interior nearly the same temperature everywhere. The blanket is the thin, non-degenerate envelope, and it — not the core — sets how fast the star leaks. The first figure shows both the layered structure and the temperature profile that collapses almost entirely across it.

Figure

A white dwarf is a hot isothermal degenerate core in a thin insulating envelope. The core sits near T_c ≈ 10⁷ K almost uniformly because degenerate electrons conduct heat superbly; the entire temperature drop to T_eff ≈ 10⁴ K happens across an envelope only ∼ 1% of the radius, and that blanket fixes the leak rate L ∝ M T_c^3.5

As the brick cools the leak slows, so the last stages take by far the longest — this is why the coolest white dwarfs are the oldest and pile up at low luminosity. The whole story is told by the specific heat of the ions. While they form a classical gas the heat capacity is ³⁄₂ k_B per ion and cooling follows Mestel's law. When neighbouring ions feel each other's Coulomb repulsion more strongly than their thermal jostling — quantified by the coupling parameter Γ — the plasma freezes into a lattice at Γ ≈ 175. Freezing, like water turning to ice, releases latent heat that briefly refills the reservoir and stalls the cooling, while the heat capacity jumps to the Dulong–Petit value 3k_B. Then, far below the Debye temperature, the lattice's quantum modes freeze out, the heat capacity collapses as T³, and the reservoir empties. The second figure follows a cooling track through these three regimes.

Figure

A white-dwarf cooling track: luminosity versus cooling age. The early decline follows Mestel's law with ion heat capacity 3/2 k_B, so age ∝ L^−5/7. Near log L/L_sun ≈ −4 the core reaches Γ = 175 (T_cryst ≈ 2.6×10⁶ K for a carbon core) and crystallizes, releasing latent heat that stalls cooling for ∼ 1 Gyr; below the Debye temperature the heat capacity falls as T³ and cooling accelerates

The core equations

Everything begins with a bookkeeping statement: since a white dwarf makes no new energy, whatever it radiates comes out of the reservoirs it already has — the internal thermal energy being drained, plus any energy liberated by phase changes.

L = -\frac{\mathrm{d}E_{\rm thermal}}{\mathrm{d}t} \;+\; L_{\rm latent} \;+\; L_{\rm grav} \qquad (1)

where: $L$ — surface (photon) luminosity, with neutrino losses added while the core is very hot; $E_{\rm thermal}$ — internal thermal energy, stored almost entirely in the ions; $L_{\rm latent}$ — power released as the ion plasma crystallizes; $L_{\rm grav}$ — gravitational energy from residual contraction and chemical phase separation. Reading: the light you see is the reservoir emptying, plus two punctuation marks that briefly slow the drain. The size of that reservoir is set by the ions, not the electrons — the single most important structural fact of the subject. Each classical ion carries ³⁄₂ k_B T, and the number of ions is the mass divided by the mass per ion.

E_{\rm thermal} = \frac{3}{2}\,\frac{k_B T}{A m_u}\,M \qquad (2)

where: $T$ — the nearly uniform interior temperature; $A$ — mean ion mass number ( $A = 12$ for a carbon core, $\approx 16$ for oxygen); $m_u$ — atomic mass unit; $M$ — stellar mass, so $M/(A m_u)$ is the ion count; $k_B$ — Boltzmann constant. Reading: for a 0.6 M⊙ carbon core at $T = 10^7$ K this is $\sim 1.2\times10^{47}$ erg — the entire fuel tank for a life lasting billions of years. The pressure that supports the star, meanwhile, comes from the degenerate electrons and barely depends on temperature. This decoupling of pressure (electrons) from heat content (ions) is what lets the star cool at essentially fixed radius.

P = P_e^{\rm deg} + n_{\rm ion}k_B T,\qquad P_e^{\rm deg}\propto \rho^{5/3}\ \text{(non-relativistic)} \qquad (3)

where: $P_e^{\rm deg}$ — degenerate electron pressure, which holds the structure; $n_{\rm ion}k_B T$ — the small ideal-gas ion pressure that carries the temperature dependence; $\rho$ — mass density. Reading: because $P_e^{\rm deg}$ dwarfs the thermal term and is nearly $T$ -independent, the mechanical structure is frozen while the temperature falls. Two competing conductivities decide where the temperature gradient lives. In the degenerate interior electron conduction overwhelms radiative transport and erases any gradient; the insulating job falls to the outer layers.

K_{\rm cond}\gg K_{\rm rad}\;\Rightarrow\; \nabla T \approx 0\ \text{(isothermal core)} \qquad (4)

where: $K_{\rm cond}$ — electron thermal conductivity, enormous for a degenerate gas; $K_{\rm rad}$ — radiative conductivity; $\nabla T$ — interior temperature gradient, driven to nearly zero. Reading: the deep star is one isothermal reservoir, and the entire drop from $\sim 10^7$ K to the $\sim 10^4$ K surface sits in the thin non-degenerate envelope. Integrating that thin envelope with a Kramers opacity ties the leak rate to the interior temperature — Mestel's luminosity–temperature relation, whose defining feature is a steep power of $T$ .

L = C\,M\,T^{3.5},\qquad \kappa = \kappa_0\,\rho\,T^{-3.5} \qquad (5)

where: $C$ — a constant collecting $\kappa_0$ , the radiation constant, and envelope factors; $\kappa$ — Kramers opacity; $T$ — the isothermal core temperature (not the surface $T_{\rm eff}$ ). Reading: the luminosity is exquisitely sensitive to interior temperature — a factor-of-two cooling of the core cuts $L$ by $2^{3.5} \approx 11$ . Combining the reservoir (2) with the leak (5) in the balance (1) gives the cooling ODE, which integrates to Mestel's celebrated age law.

\frac{\mathrm{d}T}{\mathrm{d}t} = -\frac{L}{\tfrac{3}{2}(k_B/A m_u)M} \;\Longrightarrow\; t_{\rm cool}\;\propto\;\left(\frac{M}{L}\right)^{5/7} \qquad (6)

where: $t_{\rm cool}$ — time since the star became a white dwarf. Reading: at fixed mass the age scales as $L^{-5/7}$ , so each further decade of dimming multiplies the age by $10^{5/7} \approx 5.2$ — the origin of the faint pile-up. At low temperature the ion gas stops being ideal. The relevant number is the Coulomb coupling parameter, the ratio of a neighbour's electrostatic energy to the thermal energy; past a critical value the plasma freezes.

\Gamma = \frac{(Ze)^2}{a\,k_B T},\qquad a=\left(\frac{3}{4\pi n_{\rm ion}}\right)^{1/3},\qquad \Gamma_{\rm melt}\approx 175 \qquad (7)

where: $Z$ — ion charge number; $a$ — ion-sphere radius; $n_{\rm ion}$ — ion number density; $\Gamma_{\rm melt}\approx175$ — the Monte-Carlo value at which a body-centred-cubic Coulomb lattice becomes stable. Reading: as the star cools $\Gamma$ climbs, and crossing $\Gamma \approx 175$ marks a genuine first-order phase transition inside the star. That transition liberates latent heat, roughly $k_B T$ per ion, as freezing sweeps outward from the dense centre.

L_{\rm latent} \approx \frac{q\,k_B T}{A m_u}\,\frac{\mathrm{d}M_{\rm cryst}}{\mathrm{d}t},\qquad q\sim 0.5\text{–}1 \qquad (8)

where: $q$ — latent heat per ion in units of $k_B T$ , of order unity; $M_{\rm cryst}$ — crystallized mass, growing from the centre. Reading: integrated over the star this is a sizeable fraction ( $\sim 0.3$ – $0.7$ ) of the remaining thermal budget, which is why crystallization visibly delays the cooling.

Derivation AMestel cooling and the age-luminosity relation

The goal is to get from a bare energy-conservation statement to Mestel's law, cooling age ∝ (M/L)^{5/7}, using only three ingredients already assembled: the ion thermal reservoir E = (3/2)(k_B/A m_u) M T, the Kramers-envelope leak L = C M T^{3.5}, and the fact that here the ions behave as an ideal gas so nothing else stores or releases heat. We substitute, separate variables, integrate, and re-express the interior temperature through the observable luminosity, checking the steep exponent that makes faint white dwarfs old.

Step 1. Write energy conservation with only the ion reservoir changing: $L = -\mathrm{d}E/\mathrm{d}t = -\tfrac{3}{2}(k_B/A m_u)\,M\,\mathrm{d}T/\mathrm{d}t$ . Why: the electrons store negligible heat (their specific heat is suppressed by $T/T_F \ll 1$ ) and the radius is fixed, so the only reservoir emptying is the classical ion gas.

Step 2. Insert the envelope relation $L = C M T^{3.5}$ to close the equation in $T$ alone: $\tfrac{3}{2}(k_B/A m_u)\,M\,\mathrm{d}T/\mathrm{d}t = -C M T^{3.5}$ . Why: the envelope is what converts an interior temperature into an observable luminosity, so it is the bridge between the reservoir and the leak.

Step 3. Separate and integrate, $\int T^{-3.5}\,\mathrm{d}T = -\tfrac{2 A m_u C}{3 k_B}\int \mathrm{d}t$ , which gives $t_{\rm cool} = \tfrac{3 k_B}{5 A m_u C}\,T^{-5/2}$ , dropping the hot-start boundary term because $T_0^{-5/2}\ll T^{-5/2}$ once the star has faded. Why: the initial hot phase contributes negligibly to the age, so the clock effectively starts from an arbitrarily hot core and the age is set by the present temperature.

Step 4. Eliminate the unobservable core temperature using $T = (L/CM)^{2/7}$ , so $t_{\rm cool} \propto T^{-5/2} \propto (M/L)^{5/7}$ . Why: this is the whole payoff — a directly measurable luminosity now carries the age, with a steep enough exponent that the last decade in $L$ costs more time than all the brighter phases combined. *Why the exponent is exactly −5/7: the integral $\int T^{-3.5}\,\mathrm{d}T = -\tfrac{2}{5}T^{-5/2}$ gives the $-5/2$ power in $T$ , and since $3.5 = 7/2$ this translates into the $L^{-5/7}$ scaling; a decade of dimming lengthens the age by $10^{5/7} \approx 5.2$ .*

Go deeper · Derivation B · 2 minCrystallization: latent heat, the heat-capacity jump, and the cooling delay

Derivation BCrystallization: latent heat, the heat-capacity jump, and the cooling delay

The goal is to show why the cooling curve develops a pause. As the core cools the Coulomb coupling Γ = (Ze)^2/(a k_B T) rises, and Monte-Carlo simulations of the one-component plasma find a first-order liquid-to-bcc transition at Γ ≈ 175. We locate where this happens inside the star, tally the latent heat it feeds back into the luminosity, note the rise of the ion heat capacity to the Dulong-Petit value, and add the gravitational contribution from chemical settling — accounting for every term in the energy balance (1) that slows the drain.

Step 1. Set $\Gamma = \Gamma_{\rm melt}$ to find the crystallization temperature $T_{\rm cryst} = (Ze)^2/(a\,k_B\,\Gamma_{\rm melt}) \propto n_{\rm ion}^{1/3} \propto \rho^{1/3}$ . Why: because $a\propto n_{\rm ion}^{-1/3}$ , denser cores reach $\Gamma=175$ at a higher temperature, so more massive white dwarfs crystallize earlier and warmer than lighter ones — a mass-ordered prediction observers can test.

Step 2. Add the latent heat as a source term: each freezing ion sheds $\sim q\,k_B T$ with $q\sim0.5$ – $1$ , so $\Delta E_{\rm latent}\approx q\,k_B T\,M/(A m_u)$ , a fraction $q/1.5 \approx 0.3$ – $0.7$ of the thermal energy still present at that temperature. Why: this extra energy must be radiated before the star can cool further, so it stretches the time spent near the crystallization luminosity — the pause.

Step 3. Account for the heat-capacity jump $c_V:\ \tfrac{3}{2}k_B \to 3k_B$ (Dulong–Petit) as the ions go from three translational degrees of freedom to three vibrational modes with both kinetic and potential energy. Why: a larger heat capacity means more energy must be removed per degree of cooling, further slowing the descent right after freezing begins.

Step 4. Add gravitational release from chemical separation, $L_{\rm grav}$ from oxygen settling: in a C/O mixture oxygen freezes out preferentially and sinks, lowering the star's gravitational energy and adding up to $\sim 1$ Gyr of extra delay. Why: the transition is not just a phase change of a pure substance but a partial distillation, and the sinking of the denser species is an independent, sometimes dominant, energy source.

Step 5. Collect the observable consequence: the extra energy sources produce a bump or plateau in the differential white-dwarf luminosity function, detected in field and cluster samples. Why: the same delay that stalls one star piles up many stars at the crystallization luminosity, converting a microphysical phase transition into a feature in a population-wide histogram. *As a check: for a carbon core at $\rho = 2\times10^6\ \mathrm{g\ cm^{-3}}$ the ion-sphere radius $a \approx 1.34\ \text{Å}$ gives $T_{\rm cryst} \approx 2.6\times10^6\ \mathrm{K}$ at $\Gamma_{\rm melt} \approx 175$ .*

Limiting cases

The character of the cooling changes qualitatively as the ion Coulomb coupling $\Gamma=(Ze)^2/(a k_B T)$ grows while the star cools:

Case	Result	Meaning
Ideal ion gas (Γ ≲ 1)	$c_V^{\rm ion}=\tfrac{3}{2}k_B,\quad t\propto L^{-5/7}$	Classic Mestel regime: ions are a nearly ideal Boltzmann gas; cooling follows the power law and the age climbs steeply as $L$ falls.
Strong-coupled liquid (1 ≲ Γ ≲ 175)	non-ideal $c_V$ corrections	The Coulomb liquid stores extra energy; ages depart from pure Mestel and need tabulated equations of state.
Crystallization (Γ ≈ 175)	$L_{\rm latent}\sim \frac{k_B T}{A m_u}M\,\frac{\mathrm{d}m_{\rm cryst}}{\mathrm{d}t},\ c_V\to 3k_B$	First-order transition releases $\sim k_B T$ per ion; heat capacity peaks at Dulong–Petit, producing a delay and a bump in the luminosity function.
Debye regime (T ≪ Θ_D)	$c_V^{\rm ion}\propto \left(\frac{T}{\Theta_D}\right)^{3}$	Lattice quantum modes freeze out; the specific heat collapses, the reservoir empties, and the star cools rapidly toward invisibility.

Common trapThe heat is stored in the electrons because they set the pressure

It is not. Pressure and heat content are supplied by different populations, and conflating them is the most common error in the subject. The degenerate electrons carry the pressure that holds the star up, but their specific heat is suppressed by the tiny factor T/T_F, so they store almost no thermal energy — only the sliver within k_B T of the Fermi surface can absorb heat. Essentially all the thermal energy sits in the classical ion gas at ³⁄₂ k_B T per ion, and later in the ion lattice; that is precisely why crystallization latent heat matters and why the mean ion mass A appears in every cooling formula. A second, related trap: the cooling age is not simply E/L at the present luminosity with L held constant. Because L falls steeply as the core cools, most of a white dwarf's life is spent extremely faint, and treating L as constant badly underestimates the time in the final, dimmest decade.

Worked examples

1 · When does a carbon core crystallize (confidence builder)

A 0.6 M⊙ white dwarf has a carbon core (Z = 6, A = 12) with central density ρ_c ≈ 2×10⁶ g cm⁻³. At what interior temperature does the centre reach the crystallization threshold Γ = 175, and where does that place it on the cooling track?

Step 1. Find the ion number density and ion-sphere radius. With $n_{\rm ion}=\rho_c/(A m_u) = (2\times10^6)/(12\times1.66\times10^{-24}) = 1.0\times10^{29}\ \mathrm{cm^{-3}}$ , the ion-sphere radius is $a=(3/4\pi n_{\rm ion})^{1/3} = \mathbf{1.34\times10^{-8}\ cm} = 1.34\ \text{Å}$ . This is the mean spacing that each ion's Coulomb field must bridge.

Step 2. Set $\Gamma=175$ and solve for temperature: $T_{\rm cryst}=(Ze)^2/(a\,k_B\,\Gamma_{\rm melt})$ . The numerator is $(6\times4.803\times10^{-10})^2 = 8.30\times10^{-18}$ erg cm; the denominator is $(1.34\times10^{-8})(1.38\times10^{-16})(175) = 3.23\times10^{-22}$ . Dividing, $T_{\rm cryst}=\mathbf{2.6\times10^{6}\ K}$ .

Step 3. Interpret. A core temperature of $2.6\times10^6$ K corresponds through the envelope relation to a luminosity near $\log(L/L_\odot)\approx -4$ , deep in the faint half of the cooling sequence — so crystallization is a late-life event, its latent heat released when the star is already dim. A denser (more massive) white dwarf hits $\Gamma=175$ at higher $T$ and so crystallizes earlier and brighter, exactly the mass ordering Gaia sees. T_cryst ≈ 2.6×10⁶ K; denser cores freeze sooner.

2 · The reservoir and the steepness of the clock

Estimate the thermal energy of the same 0.6 M⊙ carbon core at T = 10⁷ K, the naive cooling time if it radiates at L = 10⁻³ L⊙, and then use Mestel scaling to find how much longer it takes to fade to 10⁻⁴ L⊙.

Step 1. Reservoir. The ion count is $N_{\rm ion}=M/(A m_u)=(0.6\times1.989\times10^{33})/(12\times1.66\times10^{-24}) = 6.0\times10^{55}$ ions. The thermal energy is $E=\tfrac{3}{2}k_B T N_{\rm ion}=1.5\times(1.38\times10^{-16})(10^{7})(6.0\times10^{55}) = \mathbf{1.2\times10^{47}\ erg}$ .

Step 2. Naive timescale. Radiating at $L=10^{-3}L_\odot=3.83\times10^{30}$ erg s⁻¹, the order-of-magnitude cooling time is $E/L = (1.2\times10^{47})/(3.83\times10^{30}) = 3.2\times10^{16}\ \text{s} \approx \mathbf{1\ Gyr}$ . Right for a mid-life white dwarf, but it treats $L$ as constant, which it is not.

Step 3. Correct with Mestel scaling. Because $t\propto L^{-5/7}$ , dropping the luminosity by one decade multiplies the accumulated age by $(10)^{5/7}=\mathbf{5.2}$ . So reaching $10^{-4}L_\odot$ takes about $5.2$ times as long as reaching $10^{-3}L_\odot$ — the star spends far more time in its faint final decade than in all its brighter history, which is why the coolest white dwarfs are the oldest. E ≈ 1.2×10⁴⁷ erg, naive age ≈ 1 Gyr, and each further decade of dimming costs 5.2× more time.

Physics of the portrait

White-dwarf cooling is one of those rare corners of astrophysics where a single, mostly-analytic paper by one person still frames how the field thinks — Leon Mestel's 1952 treatment. The portrait belongs to him.

Origins · 1 minLeon Mestel and how this was discovered

Leon Mestel

1927–2017 · stellar structure, magnetohydrodynamics, and white-dwarf cooling theory

Leon Mestel was a British astrophysicist of the golden age of stellar-structure theory, who spent most of his career at the University of Sussex. In 1952, working from Cambridge, he published "On the theory of white dwarf stars" (MNRAS 112, 583) and in a few pages laid down the skeleton the field still uses. His decisive insight was the division of roles: the degenerate electrons supply the pressure but almost none of the heat capacity, while the classical ions hold the thermal reservoir, and the electrons' huge conductivity forces the interior to be isothermal so the luminosity is throttled entirely by a thin non-degenerate envelope. Coupling that Kramers envelope to the draining ion reservoir gave him the compact law that cooling age scales as L⁻⁵ᐟ⁷ — the result that turned white dwarfs into clocks and made the faint end of their luminosity function a cosmochronometer. His later work ranged across magnetic stars, star formation, and cosmical magnetohydrodynamics, but the cooling law bears his name. The refinements that followed — Coulomb corrections, crystallization latent heat, Debye cooling, phase separation — are all corrections to the Mestel backbone, not replacements for it, the surest sign of a foundational idea.

History

1926

— Fowler explains white dwarfs as matter supported by degenerate electron pressure

1952

— Mestel derives the cooling theory, giving age ∝ L⁻⁵ᐟ⁷

1960

— Salpeter, Kirzhnits and Abrikosov predict crystallization of the ion Coulomb lattice

1988

— Winget and collaborators date the Galactic disk from the luminosity-function cutoff

2019

— Gaia reveals the crystallization pile-up on the observed cooling sequence

Going further

Texts. Carroll & Ostlie, An Introduction to Modern Astrophysics, 2nd ed., §16.6–16.7 (white dwarfs and their cooling) · Mestel, L. 1952, MNRAS 112, 583 (the original cooling theory) · Fontaine, Brassard & Bergeron 2001, PASP 113, 409 (the standard cooling and cosmochronology review) · Shapiro & Teukolsky, Black Holes, White Dwarfs, and Neutron Stars, ch. 3–4 (degenerate structure and the Coulomb plasma) · Mo, van den Bosch & White, Galaxy Formation and Evolution (stellar ages as galactic-evolution constraints) · Ryden, Introduction to Cosmology (stellar ages as a bound on the age of the Universe). Key numerical anchors: the integral $\int T^{-3.5}\,\mathrm{d}T = -\tfrac{2}{5}T^{-5/2}$ underlies the $-5/7$ exponent; $\Gamma_{\rm melt} \approx 175$ is the Monte-Carlo value for a body-centred-cubic one-component plasma; the latent heat $\sim k_B T$ per ion and the Dulong–Petit $c_V = 3k_B$ are standard solid-state results. For a carbon core at $\rho = 2\times10^6\ \mathrm{g\ cm^{-3}}$ , $a = 1.34\ \text{Å}$ and $T_{\rm cryst} \approx 2.6\times10^6\ \mathrm{K}$ ; the 0.6 M⊙ reservoir at $10^7\ \mathrm{K}$ is $E \approx 1.2\times10^{47}\ \mathrm{erg}$ ; the per-decade age factor is $10^{5/7} \approx 5.2$ . Disk ages of 8–10 Gyr and globular-cluster ages of 11–13 Gyr come from standard cooling surveys; the Gaia crystallization pile-up is reported in Tremblay et al. 2019 (Nature 565, 202).

Prerequisites. Electron Degeneracy Pressure · White Dwarfs · The Chandrasekhar Mass Limit · Equation of State of Dense Matter. Leads to. Neutron Stars · Neutron-Star Structure and Interior.

Ask the tutor: Ask me to walk through Mestel's integral line by line, or to find when a white dwarf's core crystallizes from its mass and composition.

Why do the ions, and not the electrons, control the cooling time?

Because pressure and heat storage are different jobs done by different particles. The degenerate electrons supply the pressure that holds the star up, but a degenerate gas stores almost no thermal energy: only the electrons within about k_B T of the Fermi energy can change their state, a fraction of order T/T_F, which is tiny in a white dwarf. The ions, by contrast, are a classical gas with the full ³⁄₂ k_B T apiece, so they hold essentially the entire thermal reservoir. When the star radiates it is the ion reservoir that empties, and later the ions that freeze and release latent heat — which is why the mean ion mass A appears in every cooling formula.

What is the coupling parameter Γ and why does 175 matter?

Γ is the ratio of the electrostatic energy between neighbouring ions to their thermal energy, Γ = (Ze)²/(a k_B T), with a the ion-sphere radius set by the density. When Γ is small the thermal jostling dominates and the ions behave as a gas or liquid; when Γ is large the Coulomb repulsion wins and the ions lock into a lattice to minimize their mutual energy. Monte-Carlo simulations of a one-component plasma find that the liquid becomes unstable to a body-centred-cubic crystal at Γ ≈ 175 — the freezing point of white-dwarf matter. Because it is a first-order transition it releases latent heat, the physical origin of the cooling delay.

Why does crystallization slow the cooling instead of speeding it up?

Two reasons, both energetic. First, freezing is exothermic: as each ion locks into the lattice it releases latent heat of order k_B T, which must be radiated away before the star can cool further, so it lingers at the crystallization luminosity. Second, the crystallized ions have a larger heat capacity — the Dulong–Petit value 3k_B rather than the gas value ³⁄₂ k_B — because a lattice mode carries both kinetic and potential energy, so more energy must be removed per degree. On top of both, in a carbon–oxygen mixture the heavier oxygen freezes out first and sinks, releasing gravitational energy. All three push the same way: a pause, sometimes lasting a gigayear, that shows up as a bump in the luminosity function.

If cooling slows at crystallization, why does it later speed up in the Debye regime?

Because the same lattice that stored extra energy eventually runs out of modes to excite. Well above the Debye temperature Θ_D the lattice has its full Dulong–Petit heat capacity 3k_B per ion, but as T falls below Θ_D the quantized vibrational modes freeze out — there is not enough thermal energy to excite the higher-frequency phonons. Quantum statistics then drives the heat capacity down as (T/Θ_D)³, so the ions store less and less energy. With the reservoir shrinking fast there is little left to radiate, the surface temperature drops quickly, and the star's luminosity falls off the bottom of the observable range — one reason the very oldest white dwarfs are so hard to find.

How exactly does a white dwarf work as a clock, and how reliable is it?

The clock works because cooling is monotonic and calculable: a white dwarf only ever gets fainter, and theory relates its present luminosity and colour to the time since it formed. Measure a white dwarf's luminosity and effective temperature, compare against a cooling model that includes the envelope physics, crystallization, and phase separation, and read off its cooling age; add the progenitor's pre-white-dwarf lifetime for a total age. Applied to a population, the sharpest signal is the faint cutoff of the luminosity function, which dates the onset of white-dwarf formation. Reliability is limited by microphysics — envelope opacities, the exact latent heat, and phase-separation delays — but the method is powerful because its assumptions (degeneracy, isothermal core, thin envelope) are so simple, and it cross-checks main-sequence turnoff ages.

Do all white dwarfs cool the same way, or does composition matter?

Composition matters at several points. The mean ion mass A sets the reservoir per unit mass and appears directly in the cooling law, so a helium-core and a carbon–oxygen white dwarf of the same mass cool at different rates. The core density, set by the total mass, fixes when the centre reaches Γ = 175, so more massive white dwarfs crystallize earlier and warmer. The mixture matters too: a carbon–oxygen or oxygen–neon mixture separates chemically as it freezes, the denser species settling inward and releasing gravitational energy. And the thin hydrogen or helium surface layer controls the envelope opacity and hence the leak rate. So the clock must be calibrated per composition — which is why detailed cooling grids, not a single Mestel curve, are used for precision ages.