Systematic Trends in the Coupling Constants of Directly Bonded Nuclei

Abstract

The indirect coupling constant JXN has been observed for the magnetic nuclei in 50 different pairs of directly bonded X–N atoms. A synopsis is given of the reported values along with the corresponding reduced constant KXN = (2π / ℏγXγN)JXN which depends only on the molecular electronic structure. There are three nuclei, N=1H, 13C, and 19F for which KXN is now known for 15 or more different nuclei X, enough that trends are visible in the dependence of KXN upon the position of X in the periodic table. The sign of KXN(positive for H2) changes across the table somewhat between Groups V and VI, the sense of the change for N=19F being the reverse of that for N=1H and 13C. Furthermore, there is a marked increase in the magnitude of KXN with increasing nuclear charge of atom X in each Group, for negative as well as positive coupling constants. The significance of these observed trends is considered. The Ramsey theory for the electron coupling of the nuclear spins includes orbital, spin-dipolar, and contact contributions. For directly bonded atoms, the orbital contribution is zero unless there is multiple bonding, the tendency for which decreases with increasing Z in a given Group. The spin-dipolar contribution increases with increasing Z; however, it is positive, and the values calculated are an order of magnitude smaller than those found experimentally for KXN. A model is presented attributing the observed trends to the contact contribution, which depends upon the nature of the bonding orbitals employed by each atom in the bond. If both atoms employ ns orbitals in the bond, the direct contact interaction term, which is positive, dominates. The bonding of Group VII and, to a lesser degree, Group VI atoms employs primarily p orbitals. In such atoms the contact interaction is indirect, involving polarization of the core s electrons and a change in sign of the term. The contact term with inclusion of such core polarization effects provides a model consistent with the data available. The model is used to predict the signs, in some cases also the magnitudes, of several coupling constants not yet observed. For example, in NF3, OF2, and F2, we expect KNF to be negative, KFF to be positive, and KOF probably to be positive. A number of features are discussed including the coupling in highly ionic bonds such as the Rb, Cs, and H fluorides, and the relationship of the model to nuclear hyperfine interactions in atoms and ions with unpaired spins and in organic free radicals.