A reconsideration of the split- p -orbital(s. p .-o.) method in molecular-orbital theory

A careful study is made of the status of the split-p-orbital method recently introduced by M. J. S. Dewar. There are two chief difficulties. The first, which is associated with a discontinuity in the gradient of the s.p-o., is shown not to make any difference to the calculated energy. But the second, which is associated with a lack of orthogonality between the split-p-function and the orbitals for other electrons (e.g. core electrons) is more serious. A full allowance for this non-orthogonality is possible, but at considerable cost in complexity. Even then, however, there is the chance of ‘sagging’ of the energy when the state being considered is not the lowest one of given spin multiplicity and group symmetry. These conclusions are illustrated by specific numerical application to certain atomic states. The full s.p-o. method is likely to apply more often to molecular states than to atomic ones.