The Acoustic Branch
Walk through reciprocal space and the phonon branches reveal themselves in a pattern that is both universal and deeply material-specific. The acoustic branches are the ones you always find: three of them, emerging from ω = 0 at the Γ point, with linear dispersion for small wavevectors. The optical branches are the ones that tell you something about the internal structure of the unit cell: they appear at finite frequency when the unit cell contains more than one atom, and they form flat or weakly dispersing bands that are the hallmark of localized vibrations.
The acoustic branches are named for what they are in the long-wavelength limit: sound waves. Longitudinal acoustic (LA) phonons correspond to compressions and rarefactions — atoms moving parallel to the direction of propagation, creating regions of high and low density. Transverse acoustic (TA) phonons correspond to shear deformations — atoms moving perpendicular to the direction of propagation, displacing planes of atoms relative to each other. In a crystal with N atoms per unit cell, there are always three acoustic branches: one LA and two TA. The two transverse branches are degenerate at the Γ point in a cubic crystal but generally split in lower-symmetry systems, which is why measuring the acoustic dispersion is a sensitive probe of crystal symmetry.
The dispersion of the acoustic branches is not perfectly linear. Near Γ, ω ≈ v_s |q|, where v_s is the sound velocity for the branch in question. But as you move away from Γ toward the Brillouin zone boundary, the dispersion curves off. The linear approximation assumes a continuous medium, but the crystal is discrete, and that discreteness matters. At the zone boundary, the phonon wavelength is on the order of the lattice spacing, the Bragg condition is satisfied, and the wave becomes a standing wave. The dispersion relation flattens. The group velocity dω/d|q| goes to zero. The phonon cannot propagate further.
This flattening is not a mathematical artifact — it has measurable consequences. The density of states diverges at the zone boundary because dω/dq → 0, producing van Hove singularities in the phonon density of states. These singularities show up as peaks in inelastic neutron scattering cross sections and as features in the specific heat that the Debye model cannot reproduce. They are fingerprints of the Brillouin zone structure.
The optical branches tell a different story. In a diatomic chain with alternating masses M and m, the optical branch at Γ corresponds to the two atoms in the unit cell vibrating against each other. The center of mass stays still. The frequency is finite even at q = 0 because the relative motion is resisted by the interatomic spring. This is why optical phonons are active in infrared and Raman spectroscopy: the relative displacement of the two atoms creates an oscillating dipole (or changes the polarizability, in the case of Raman), and photons can couple to that oscillation. The optical branches are often flatter than the acoustic branches because the vibration is more localized — it does not involve the collective motion of the whole crystal.
The distinction between acoustic and optical is not just semantic. Acoustic phonons dominate thermal transport because they have the highest group velocities and the longest mean free paths. Optical phonons tend to be low-velocity, short-lived, and heavily scattered. They contribute to the heat capacity but not much to the thermal conductivity. In Raman and infrared spectroscopy, you are predominantly seeing optical phonons. In neutron and X-ray scattering, you see everything — acoustic, optical, and the transitions between them.
The branching structure itself — the way the phonon dispersion folds into branches as you fill the Brillouin zone — is determined by the number of atoms per unit cell and the symmetry of the lattice. For a crystal with N atoms per primitive cell, there are always 3N branches: 3 acoustic and 3N−3 optical. This is a robust topological feature. It follows from the fact that the dynamical matrix is a 3N × 3N Hermitian matrix, and every Hermitian matrix has as many real eigenvalues as its dimension. The acoustic branches always include a zero-frequency solution because uniform translation costs no energy. That is the Goldstone mode associated with spontaneous breaking of translational symmetry. The acoustic phonons are the Goldstone bosons of the crystal.