V. Sahni, PHYSICAL INTERPRETATION OF DENSITY-FUNCTIONAL THEORY AND OF ITS REPRESENTATION OF THE HARTREE-FOCK AND HARTREE THEORIES, Physical review. A, 55(3), 1997, pp. 1846-1856
In this paper we provide a rigorous physical interpretation of the Koh
n-Sham (KS) density-functional theory electron-interaction energy func
tional E(ee)(KS)[rho] and its functional derivative upsilon(ee)(KS)(r)
= delta E(ee)(KS)[rho]/delta rho(r) based on the original ideas of Ha
rbola and Sahni, and of their extension by Holas and March, The functi
onal, and hence the derivative, incorporate electron correlations due
to the Pauli exclusion principle and Coulomb repulsion as well as thos
e of the correlation contribution to the kinetic energy. The interpret
ation is in terms of a field F(r), which is the sum of two fields whos
e sourer distributions are expectations of Hermitian operators. The fi
rst of these fields epsilon(ee)(r) accounts for Pauli and Coulomb corr
elations. its sourer is the pair-correlation density and it is determi
ned by Coulomb's law. The second Z(tc) (r) accounts for the correlatio
n-kinetic contribution, and its source is the difference between the k
inetic-energy-density tensor for the noninteracting and interacting sy
stems. The corresponding field is the derivative of this tenser. The f
unctional derivative upsilon(ee)(KS)(r)(r) is the work done to move an
electron in the field F(r). Since the field F(r) is conservative, thi
s work done is path independent. The quantum-mechanical electron-inter
action energy E(ee)[rho] and correlation-kinetic energy T-c[rho] compo
nents of E(ee)(KS)[rho] can also be expressed in virial form in terms
of the respective fields epsilon(ee)(r) and Z(tc)(r), which give rise
to them. A similar rigorous physical interpretation of the Kohn-Sham t
heory representation of the Hartree-Fock and Hartree approximations is
also given. If in these representations. the correlation kinetic ener
gy is neglected the equations reduce to the corresponding approximatio
ns of the work formalism of electronic structure. Finally, it is argue
d on physical grounds that the interpretation provided for the ground
state is equally valid for excited states.