Perturbation theory and spin coupling constants in substituted compounds. II. Main group tetrahedral and planar trigonal fluorides

Abstract

An analytical LCAO MO theory of the Fermi contact term in reduced nuclear spin coupling constants 1 K(A-F) has been developed for substituted planar trigonal AF 3− k L′ K and tetrahedral AF 4− k L′ K , K = 1 to 3, main group element fluorides. The new model is a perturbation extension of the Pople-Santry (P-S) theory of 1 K(A-F) for unsubstituted polyhedra AF m . The differences between diagonal matrix elements, δα′ s = 〈 s L′| H| s L′〉 − 〈 s F| H| s F〉 and δα′ p = 〈 σ L′| H| σ L′〉 − 〈 σ F| H| σ F〉, have been chosen as perturbations and first-order corrections to the P-S terms (both linear and quadratic in δα′ s and δα′ p ) have been obtained in explicit form. It was shown that both central atoms A and substituents L′ should be divided into two groups displaying quite different regularities of 1 K(A-F). Namely, carbon differs from all other atoms A (B, Si, Ge, etc.) in relative energies of the S A and p F orbitals and H, CH 3, C 6H 5, SiH 3, etc., differ from other substituents L′ (halogens, OR, NR 2, SR, PR 2, etc.) in lack of a lone ns 2 pair. We suggest naming these two kinds of substituents “one-pronged” and “two-pronged,” respectively. It was shown that changes in 1 K(A-F) will be negative and increase monotonically in absolute value as the σ donor ability of the L′ and their numbers k increase only for two-pronged substituents L′ such as halogens. This monotonic increase should be more or less additive for A = B, Si but for A = C noticeable positive deviations from additivity should take place. One-pronged L′ such as H or CH 3 can cause nonmonotonic changes in 1 K(A-F) along the series AF 4− k L′ k , k = 0 to 3 for all atoms A. Moreover, for A = C the resulting changes in 1 K(C-F) may even be substantially positive. Model results agree with experiment and permit a number of predictions to be made. Extended Hückel calculations were performed as an independent check of model conclusions. All of the expected trends were found but only after it was discovered that a key factor was the magnitude of the overlap between the fluorine 2 s orbital and the central atom ns orbital.