Electronic structure calculations grounded in the Hohenberg-Kohn- Sham density functional theory predict what atoms, molecules, and solids can exist, and with what properties. The accuracy of these predictions depends on the accuracy of the functionals or rules used to construct the energy from the electron density. Goals of this theoretical research are: (1) to develop, from first principles, more accurate exchange correlation and kinetic energy functionals of the local spin density and generalized gradient types; (2) to apply generalized gradient approximations in calculations of measurable proerties such as the electron spin density at the nucleus of an atom or the energy of a metallic surface; (3) to understand the capabilities and limitations of these approximations in light of radical nonlocality of the exact exchange correlation potential; (4) to refine and test fully nonlocal functionals such as the self-interaction correction or the weighted density approximation; (5) to construct local pseudopotentials describing effective interactions between valence electrons and ionic cores in metals using only the equilibrium density, the valence, and a "smoothness" condition; (6) to study size effects in metals. %%% Theoretical research will be conducted on the foundations of one of the major techniques currently used to determine the properties of atoms, molecules and solids. The results of this work will reinforce the fundamentals of density functional theory and provide guidance for new directions for its computational implementation.
