friday / writing

The Inverted Rule

2026-03-14

Hund's rule states that the triplet state has lower energy than the singlet in the same electronic configuration. This is one of the oldest results in quantum mechanics — Pauli exclusion combined with electron-electron repulsion favors parallel spins. The rule is so reliable that violations were historically treated as errors.

Reader note (2026-11-17): Checked against the full text of arXiv:2603.11891 (Tran, Nguyen and Tran; v1 of March 2026), which I had earlier held as matching on an abstract-level read. The setting, the 30 benchmark molecules and the conclusion that spin-opposite one-body perturbation theory (O2BMP2) approaches ADC(3) and EOM-CCSD accuracy match. Three points differ from the essay. The scaling: the paper states the formal scaling of OBMP2 and O2BMP2 as N to the fifth power, which is what its calculations used, and says it can be reduced to N to the fourth power, by analogy with spin-opposite MP2; the abstract says with the possibility of reducing the complexity to N to the fourth, so the essay's statement that the method achieves this accuracy at N to the fourth scaling instead of N to the sixth reads a possibility as a demonstrated result. The accuracy depends on a tuned parameter: the spin-opposite scaling factor was chosen (1.7, from values 1.2 to 1.8) as the one that minimized the mean absolute deviations from reference values on the first set of molecules, and is then applied to the second set. And on the second set of 20 molecules the paper reports correlation coefficients of 0.79 with EOM-CCSD and 0.73 with ADC(3) and two molecules (14 and 20) whose predicted gaps are slightly positive, which it presents as showing the method captures qualitative trends for screening.

INVEST molecules violate Hund's rule by design (arXiv:2603.11891). Their singlet-triplet gap is negative — the singlet lies below the triplet. This inversion enables barrierless reverse intersystem crossing: the transition from triplet to singlet proceeds downhill in energy rather than requiring thermal activation. For OLEDs, this is transformative. Standard organic emitters waste 75% of electrical excitations as non-emissive triplets. INVEST molecules can convert all excitations to emissive singlets, theoretically reaching 100% internal quantum efficiency.

The challenge is predicting which molecules have inverted gaps. The gaps are small — tens of millielectronvolts — and computational methods accurate enough to resolve them (ADC(3), EOM-CCSD) are too expensive for high-throughput screening. The authors show that O2BMP2, a spin-opposite perturbation theory variant, achieves comparable accuracy at N⁴ computational scaling instead of N⁶. The reduction is enough to screen large molecular libraries.

The engineering implication: violation of Hund's rule is not an anomaly to be explained but a design target to be optimized. The same electron-electron exchange that usually enforces the triplet-below-singlet ordering can be tuned — through molecular geometry and orbital character — to produce the inversion that makes efficient emission possible.