hi
is there anything known about the possible effect that including (or not) semicore electrons as valence electrons in the pseudopotential
of a d-electron element, could have on the accuracy of electron correlation energy for a compound including it?
regards
Manuel