We review some recent results concerning the probability that two electrons will be found close together in any interacting electronic system, and why this probability is usually well approximated by local (LSD) and semilocal spin density functional theories. The success of these approximations for the energy in "normal" systems is explained by the usual sum rule arguments on the system- and spherically-averaged exchange-correlation hole density <n xc (u)>, coupled with the nearly correct, but not exact, behavior of these approximations as the interelectronic separation u → 0. We argue that the accuracy of the LSD on-top hole density in "normal" systems is due to its accuracy in the noninteracting, weakly-interacting, and strongly-interacting limits.
Discussion(0)
No comments yet. Be the first to comment.