I. Ab-initio effective potentials for use in molecular calculations. II. The Sternheimer correction, perturbation theory, and approximate wavefunctions. III. The theoretical determination of the Li_2 B^1π_u potential energy curve.

Citation

Kahn, Luis R. (1972) I. Ab-initio effective potentials for use in molecular calculations. II. The Sternheimer correction, perturbation theory, and approximate wavefunctions. III. The theoretical determination of the Li_2 B^1π_u potential energy curve. Dissertation (Ph.D.), California Institute of Technology. doi:10.7907/P3JP-H317. https://resolver.caltech.edu/CaltechTHESIS:11202009-150548056

Abstract

Part I: We have investigated the efficacy of ab-initio effective potentials in replacing the core electrons of atoms for use in molecular calculations. The effective potentials are obtained from ab initio GI calculations on atoms and are unique and local. We find that the use of these effective potentials to replace the core orbitals of such molecules as LiH, Li_2, BH, or LiH_2, leads to wavefunctions in excellent agreement with all-electron ab-initio results. The use of such effective potentials should allow ab-initio quality wavefunctions to be obtained for systems too large for the ab-initio consideration of all the electrons. Part II: We have investigated the Sternheimer correction for the calculation of the nuclear quadrupole coupling constants and its relation to the approximate nature of the zero-order wave- function. The first-order perturbed Hartree-Fock equations, and some approximations to them, are solved for the 2^2P state of Li, and the resulting Sternheimer type corrections are compared with Sternheimeris approximate calculations and with results from non-perturbation theory approaches. PART III: The Li_2 B^1π_u potential energy curve has been calculated with a Multi-Configuration SCF (MCSCF) wavefunction. Several different types of wavefunctions and basis sets have been examined and their accuracy determined. The most accurate wavefunction used predicts a binding energy of 0.3015 e.V. (84% of the experimental value of 0.362 e.V.), and predicts a potential hump of 0.0724 e.V. with its maximum in the vicinity of 10.6 Bohr. It is argued that the theoretical value of the hump is an upper bound to the experimental value.

Thesis Files