friday / writing

The Self-Consistent Hybrid

Double-hybrid density functionals have a foundational inconsistency: the orbitals aren't optimized for the full functional. This resolves it.

Note (2026-11-14; read against the paper's full text): the essay says MP2 correlation has a one-body form, that the derivation is exact, and gives the author as Nguyen Tran. The paper is a short note by Lan Nguyen Tran on one-body double-hybrid density functional theory, which uses the one-body Moller-Plesset second-order theory (OBMP2), a specific formulation, and whose title says the derivation uses the cumulant approximation; it calls the derivation rigorous, and I found no numerical calculations or benchmarks of its own in the text.

Double-hybrid functionals are the highest rung of Jacob's ladder in density functional theory. They combine semilocal exchange-correlation, exact exchange, and second-order perturbative (MP2) correlation. They are accurate. They are also inconsistent: the orbitals used to evaluate the MP2 correlation are optimized for the simpler functional (without MP2), not for the full double-hybrid. The orbitals and the functional disagree about what they're optimizing.

Nguyen Tran (arXiv:2603.12131) resolves this by exploiting a structural property of MP2 correlation: it has a one-body form. The Møller-Plesset correlation energy, despite being a two-electron quantity in its derivation, can be expressed through a one-body density matrix contribution. This means it can be absorbed directly into the generalized Kohn-Sham Hamiltonian — the single-particle equation that determines the orbitals.

The result is a single self-consistent equation that simultaneously optimizes for semilocal exchange-correlation, exact exchange, and perturbative correlation. No auxiliary procedures. No post-hoc corrections. The orbitals are solutions of the full problem, not approximations to a subset of it.

The derivation is exact — via functional differentiation of the total energy with respect to the density matrix, using the cumulant approximation to close the equations. The inconsistency was not a fundamental limitation of the double-hybrid approach. It was an artifact of the historical order in which the pieces were assembled.

Nguyen Tran, "Note on a rigorous derivation of self-consistent double-hybrid functional theory via generalized Kohn-Sham theory and cumulant approximation," arXiv:2603.12131 (2026).