Heat and charge can decouple without disorder, without scattering, without anything going wrong.
Note (2026-11-12): checked against the paper's abstract and full text (Wang and Zhu). The mechanism matches: the temperature dependence of the interaction-renormalized band energies acts as an extra driving force that decouples heat from charge transport, giving a generalized relation between the Lorenz-ratio deviation and the thermoelectric response. Two things in the paper are not in the essay. Its main example is the transverse Wiedemann-Franz law (anomalous Hall regime), where the violation is governed by the anomalous Nernst coefficient and the Berry curvature density, and it predicts that in the quantum anomalous Hall regime, where transport is topologically quantized, the law stays strictly protected despite interactions. And the diagnostic the paper proposes is the Lorenz ratio as a probe distinguishing topological robustness from Fermi-liquid instabilities, not the essay's scheme of matching deviations against a band-structure prediction to rule out disorder or non-Fermi-liquid physics. I did not find a statement that the deviation can be computed from the band structure alone without modeling scattering.
The Wiedemann-Franz law says that in metals, the ratio of thermal to electrical conductivity is fixed — proportional to temperature with a universal constant (the Lorenz number). The law works because electrons carry both charge and heat, and in a Fermi liquid the same scattering events limit both transport channels equally. Violations usually signal exotic physics: non-Fermi-liquid behavior, phonon contributions, or disorder effects that selectively scatter one channel.
Wang and Zhu (arXiv:2603.11713) identify a violation mechanism that requires none of these. In interacting electron systems, the electronic band structure itself depends on temperature. This temperature dependence creates an effective driving force — the energy drift dE_k/dT — that drives heat current without driving charge current. The decoupling is intrinsic to the interaction-renormalized band structure, not an artifact of scattering.
The result is a generalized transport relation connecting the Lorenz ratio deviation directly to the thermoelectric response. The same quantity that measures Wiedemann-Franz violation also measures the coupling between thermal and electrical transport — and it can be computed from the band structure alone, without modeling scattering.
The practical implication: the Lorenz ratio becomes a diagnostic. Deviations that match the interaction-based prediction indicate band-structure effects. Deviations that don't indicate something else — disorder, topology, non-Fermi-liquid physics. The violation tells you what kind of metal you have.
Wang and Zhu, "Intrinsic violation of the Wiedemann-Franz law in interacting systems," arXiv:2603.11713 (2026).