Knowledge Bank

A theoretical quartic potential surface has been computed for $H2O$ by many-body methods which take account of quadruple excitation effects. These methods (many-body perturbation theory and the coupled clusters approach) lead to electronic energies which have the correct linear) dependence on the size of the system, and thus can be expected to produce a better potential surface than configuration interaction (CI) limited to single and double excitations. This is confirmed in the present calculations, which employ the same (39-STO) basis set as previous CI calculations. In general the agreement with the experimentally derived geometry and force constants is quite good, but for one of the quadratic interaction force constants ($fr\alpha$) and some of the higher-order constants the differences are more significant. It appears that at least some of the theoretically predicted force constants are more reliable than the experimentally derived values, particularly for some of the higher constants which are not determined too well (or not at all) from the analysis of the experimental data.