PPP-VB Model

A higher approximation may be achieved by employing two-parameter OEAOs which result from the admixture of the next nearest neighbor AOs (mixing parameter e ) as well as the nearest neighbor (mixing parameter e ) ones. Such a basis set is referred to as the 61,2 basis, and we can similarly define 61,3, 61,2,3, and other basis sets. However, it turns out that little is gained in tins way [60,61], and that in most applications the 61 basis is perfectly satisfactory. 

We must emphasize that in the PPP-VB calculations referred to above, we employed all the covalent structures for a given system. However, these results remain practically unchanged when only Kekule structures Eire used. Merely in cases when only one Kekule structure exists is it advisable to employ one or more additional non-Kekule structures. 

The primary objective of most applications carried out so far was to assess the performance of the PPP-VB method for diverse alternant and nonaltemant -electron systems of aromatic, nonaromatic or antiaromatic character, both electrically neutral and charged. The main emphasis was on ground states of different spin multiplicity, even though some preliminary calculations were also carried out for excited states. The PPP-VB codes were also employed to provide the approximate three- and four-body connected cluster components for the so-called VB-corrected coupled cluster (CC) approach [71]. In the following, we briefly point out the most important aspects of the PPP-VB method and illustrate them with a few typical results. 

Correlated ground states, basis transferability and the role of ionic structures 

To assess the effectiveness of the PPP-VB method in describing the correlated ground states of various w-electron systems, we have examined a set of 70 relatively small molecules 

---