The Debye Model: Why Heat Capacity Falls as T³

A puzzle in the 19th-century data

Dulong and Petit noticed in 1819 that most solids share the same molar heat capacity, about 3R ≈ 25 J/(mol·K), once you're at room temperature. Classical physics explains this neatly with the equipartition theorem: each atom vibrates in three dimensions, each dimension acts like a spring storing both kinetic and potential energy, and each of those six quadratic energy terms contributes ½k_B of heat capacity — six halves make 3k_B per atom, 3R per mole. It is a clean, parameter-free prediction, and it works — until you cool the sample down. Every measured solid's heat capacity falls well below 3R at low temperature and heads to zero as T → 0, in flat contradiction with a theory that has no temperature dependence at all.

Einstein's fix, and where it still falls short

Einstein's 1907 answer was to quantize the vibrations: each atom oscillates like a quantum harmonic oscillator with a single frequency f, and a mode can only hold energy in units of hf. At low temperature, k_BT is much smaller than the quantum hf, so the mode is almost always sitting in its ground state and simply cannot absorb a small amount of extra heat — it is frozen out. This correctly predicts that heat capacity falls at low temperature and gets Dulong-Petit right at high temperature. The one thing it gets wrong is the shape of that fall: the Einstein model drops off exponentially, far faster than real measurements, because it assumes every atom vibrates at one identical frequency.

Debye's continuum: a spectrum, not a single note

Debye's 1912 refinement was to stop pretending each atom vibrates independently and instead treat the whole crystal as an elastic continuum that supports sound waves. Those waves come in a whole spectrum of frequencies, from long, slow wavelengths spanning the entire sample up to a maximum frequency set by the fact that a wave cannot have a wavelength shorter than about twice the spacing between atoms — you cannot resolve a wiggle finer than the grid it lives on. Counting the number of standing-wave modes below a frequency f the way you would count modes of light in a box, then quantizing each one as its own oscillator, gives a density of states that rises as f² and is cut off sharply at the Debye frequency f_D.

g(f) df  =  (9N/f_D^3) f^2 df        for 0 < f < f_D   (density of modes)
Theta_D  =  h f_D / k_B              (the Debye temperature)

C_V(T)   =  9N k_B (T/Theta_D)^3  INTEGRAL[0, Theta_D/T]  x^4 e^x / (e^x - 1)^2  dx

low T  (T << Theta_D):   C_V  ~  T^3
high T (T >> Theta_D):   C_V  ->  3N k_B      (Dulong-Petit)

Why T-cubed, specifically

At low temperature only modes with hf ≲ k_BT can be thermally excited at all, so effectively only modes below some cutoff frequency proportional to T contribute. Because the density of modes itself grows as f², the number of "live" modes below that cutoff scales as T³, and each of those live modes contributes roughly k_B to the heat capacity by the classical equipartition-like argument — multiply the two together and the whole heat capacity scales as T³. This is the celebrated Debye T-cubed law, and it matches real low-temperature calorimetry data for insulators and simple metals (once you separately account for the electron gas's own linear-in-T contribution) far better than the single-frequency Einstein picture.

The Debye temperature as a material fingerprint

The one free parameter, the Debye temperature Θ_D = hf_D/k_B, packages together the speed of sound in the material and the density of atoms — stiffer bonds and lighter atoms push the maximum vibrational frequency, and therefore Θ_D, higher. Diamond, with its extremely stiff covalent bonds and light carbon atoms, has a Debye temperature above 2200 K, meaning it stays near the classical Dulong-Petit regime all the way up to very high temperatures and drops off its heat capacity only very close to absolute zero. Soft, heavy metals like lead have a Debye temperature under 100 K, so their heat capacity is already deep into the classical plateau at ordinary room temperature. In modern usage, phonons — the quantized lattice vibrations Debye's counting scheme describes — are the standard language for heat capacity, thermal conductivity and electron-phonon coupling in solids far beyond Debye's original 1912 derivation.

Frequently asked questions

Try it live

Everything above runs in your browser — open Debye Model and change the parameters while it is running. Nothing is installed, nothing is uploaded, the whole model lives in one tab.