Covariant Local Matrix Density Approximation through Spectral Reconstruction in Multistate Density Functional Theory

Abstract A covariant local matrix density approximation (LMDA) is introduced within multistate density functional theory based on subspace invariance and the spectral decomposition of the matrix containing state and transition densities. The exchange-correlation matrix functional is constructed through spectral reconstruction of this matrix density, whereby local exchange-correlation functionals are incorporated as spectral-channel functionals within a covariant matrix-functional formalism. The resulting formulation preserves exact normalization of the exchange-correlation matrix hole, recovers ground-state density functional theory in the single-state limit, and preserves spin-multiplet degeneracy through covariance of the spin matrix functional under spin rotations. Combined with multistate self-consistent-field (MSSCF) optimization, the present framework enables fully variational calculations of interacting ground and excited states. Applications to atomic excitations, H2 dissociation, ethylene torsion, and cyclobutadiene automerization reveal that the resulting MSDFT/LMDA approach captures essential multistate and strong-correlation physics, including static correlation and spin symmetry. The key insight is that the central obstacle in extending density functional theory to coupled electronic states is not necessarily the lack of appropriate scalar exchange-correlation approximations, but the lack of a covariant matrix-functional formalism in which such functionals can operate. The present MSDFT/LMDA provides a direct realization of this matrix-functional formalism.