Potential curves of the lower nine states of Li2molecule: Accurate calculations with the free complement theory and the comparisons with the SAC/SAC-CI results.

Abstract

The free-complement (FC) theory proposed for solving the Schrödinger equation of atoms and molecules highly accurately was applied to the calculations of the potential curves of the lower nine states of the Li2molecule. The results were compared with the accurate experimental Rydberg-Klein-Rees potential curves available. They overlap completely with each other without any shift everywhere for all the states of Li2. At all the calculated points on the seven potential curves ranging between -14.83 and -15.00 hartree, the average difference was only 0.0583 kcal/mol and the maximum difference was only +0.165 kcal/mol. For the vertical excitation energies from the ground state curve to the seven excited states, the differences between theory and experiment were 0.000 645eV in average and their maximum difference was -0.007 20eV. The potential properties calculated with the FC theory also agreed well with the experimental values. These results show a high potentiality of the FC theory as a highly predictive quantum chemistry theory. For comparison, as an example of the Hartree-Fock based theory popular in modern quantum chemistry, we adopted the symmetry-adapted-cluster (SAC)-configuration-interaction (CI) theory using a highly flexible basis set. While the FC theory gave the absolute agreements with experiments, the SAC-CI potential curves compare reasonably well with experiments only after shifting-down of the SAC-CI curves by 5.727 kcal/mol. The differences in the excitation energies between SAC-CI and experiments were 0.004 28eV on average, and the maximum difference was +0.109 67eV. The SAC-CI results reported in 1985 were less accurate but still reasonable.