Luttinger's Theorem, Fermi Volume, and Failure Modes

Luttinger’s theorem equates charge density, modulo completely filled bands, with the volume enclosed by the interacting Fermi surface when charge $U(1)$ and translations are preserved and the many-body state satisfies the analyticity or topological hypotheses of the chosen proof. It is a volume statement, not a proof of long-lived Landau quasiparticles.

Required background. Use free Fermi-volume counting and Dyson/self-energy structure. Helpful background. ‘t Hooft anomaly matching supplies a modern symmetry-obstruction perspective.

Luttinger volume count and conventions

For a translationally invariant spin-degenerate system with primitive-cell volume $v_c$,

$$\nu=\frac{g\,v_c V_F}{(2\pi)^d}\pmod{g},$$

where $\nu$ is charge per primitive cell, $g$ is the explicitly counted degeneracy, and $V_F$ is the oriented sum of pocket volumes for one member of the $g$-fold degenerate multiplet in the first Brillouin zone. The modulus $g$ assumes filled bands occur in those degenerate multiplets; without that degeneracy the count is stated flavor by flavor or modulo the appropriate integer. In the continuum, it reduces to $n=gV_F/(2\pi)^d$.

In the diagrammatic proof, the density is written as a frequency integral of $G$ and separated into a winding term plus the Luttinger integral involving $G\,\partial_\omega\Sigma$. Vanishing of the latter relies on a conserving analytic construction and an admissible continuation. The original result is Luttinger 1960, pp. 1153–1163.

Oshikawa’s flux-insertion argument instead compares the many-body crystal momentum before and after threading one flux quantum through a torus. Equating that exact momentum shift with the shift of low-energy charged states yields the same volume modulo reciprocal-lattice momentum Oshikawa 2000, pp. 3370–3373. This route makes translations, conserved charge, topology of the torus, and the assumed low-energy sector visible.

What changes the count

Broken translation symmetry enlarges the unit cell and folds the Brillouin zone; one must recount using the actual cell before claiming violation. Spin polarization changes the degeneracy and may require separate spin-resolved volumes. Superconductivity breaks charge $U(1)$ and has no ordinary electron Fermi surface.

Topological order can absorb part of the flux-insertion momentum in an emergent sector, permitting a fractionalized Fermi liquid with a modified visible count. Green-function zeros can change the sign structure used in some diagrammatic formulations, but a zero is not automatically a universal exception: the proof hypotheses and physical singular surface must be specified. A singular self-energy or ground-state degeneracy can invalidate a step without determining the replacement theorem.

A practical theorem validity check

Before interpreting a “small” surface, check the primitive cell, all pockets and degeneracies, conserved charge, broken symmetries, topological ground-state sectors, and whether the measured object is a pole, a zero, or only a broadened spectral maximum. Only then compare density with volume. This validity check tests theorem compatibility; it does not identify the microscopic mechanism.

Exercises

A commensurate density wave doubles a one-dimensional unit cell. Explain why an apparent halving of the Fermi volume is not by itself a violation.

Solution

Doubling the cell halves the Brillouin zone and changes the filling per cell. The volume relation is defined modulo filled bands in the reduced zone. After folding both the density and all Fermi points into that zone, the apparent factor of two can disappear without invoking topological order.

References

• J. M. Luttinger, “Fermi Surface and Some Simple Equilibrium Properties of a System of Interacting Fermions,” Physical Review 119 (1960) 1153–1163, doi:10.1103/PhysRev.119.1153.
• Masaki Oshikawa, “Topological Approach to Luttinger’s Theorem and the Fermi Surface of a Kondo Lattice,” Physical Review Letters 84 (2000) 3370–3373, doi:10.1103/PhysRevLett.84.3370, Open PDF.