Kirchhoff's law of thermal emission says that at thermal equilibrium, a body's emissivity equals its absorptivity at every wavelength and angle. An object that absorbs 80% of incoming light at some wavelength must also emit 80% of blackbody radiation at that wavelength. This has been a cornerstone of thermal physics since 1860.
Reader note (2026-11-17): Checked against the full text of arXiv:2603.23538 (Cui and Raman; v1 of March 2026), which I had earlier held as a weak read. The authors, the III-V metasurface and the overlap of the mode's optical spin density with the magneto-optical material match. Three points differ. First, the work is theory validated against full-wave finite-element simulations; I found no fabricated device in it, and the paper itself cites earlier experimental demonstrations that mostly needed 1 to 3 tesla and earlier numerical work showing near-complete violation below 0.5 tesla, so the essay's showing how to break the law almost completely overstates what is new in this paper: its contribution is the closed-form perturbation theory for the resonance frequency shift, which it says earlier phenomenological theory did not give. Second, the field value: the abstract says a contrast of 0.8 at 0.1 tesla, while the body reports a contrast exceeding 0.8 at 0.05 tesla and the difference above 0.9 at 0.1 tesla; the essay follows the abstract. Third, the paper says the first-order theory agrees with the simulations at low field and begins to deviate by about 0.2 tesla, so the essay's theory predicts which designs work before fabrication holds for small fields and for the simulated structures compared; the paper reports a two-orders-of-magnitude difference in frequency shift between two structures at the same field. The essay's through-claim, that the law was never about the material, is the essay's own.
Cui and Raman (arXiv: 2603.23538) show how to break it almost completely. Their perturbation theory identifies the mechanism: magneto-optic materials break Lorentz reciprocity, which is the microscopic reversibility that underwrites Kirchhoff's law. In a carefully designed III-V semiconductor metasurface, a magnetic field of just 0.1 Tesla produces a nonreciprocal emissivity contrast of 0.8 — meaning the difference between absorption and emission at a given wavelength is nearly as large as physically possible.
The key insight: the violation depends on the overlap between the optical mode's spin density and the magneto-optical material. This is a geometric condition. Some photonic structures amplify the effect enormously; others are nearly immune to it. The magnetic field sensitivity varies by orders of magnitude depending on the structure's design, and the theory predicts which designs work before fabrication.
The through-claim: the law was never about the material. It was about the symmetry. Kirchhoff's law holds whenever the underlying physics is time-reversal symmetric. Break that symmetry — with a magnetic field, in the right geometry — and the law breaks with it. The material can absorb light it refuses to emit, or emit light it refuses to absorb. The 165-year-old constraint dissolves the moment reciprocity does.
Cui & Raman, 2603.23538. Thermal physics / photonics / magneto-optics / Kirchhoff's law.