Abstract
Both the chemical shifts of nuclear magnetic resonances and molecular diamagneticsusceptibilities have their origin in the electronic currents induced by an external magnetic field. For the aromatic hydrocarbons detailed chemical‐shift measurements and extensive measurements of the magnetic anisotropies are available to aid in a study of the validity of various theories connecting magnetic properties with magnetic electron currents. The available experimental data seem to present a definite paradox. The chemical shifts due to ring currents are theoretically directly proportional to the values of Δχ, the anisotropy in the diamagneticsusceptibility; yet the experimental values of Δχ are much larger than those calculated from theory, while the experimental chemical shifts are substantially smaller than the theoretical ones. Hoarau has suggested that only a part, χ L , of the total anisotropy Δχ arises from the π electrons involved in the ring current and that one should add Δχσ , a contribution from the electrons in the σ bonds, and Δχ p , the contribution due to the localized p electrons. The SCF—MO calculations of χ L for benzene give a value of approximately —30×10—6, which is only about half the experimentally observed value of —59.7×10—6. Rough calculations are presented which indicate that the contributions of the σ and localized p electrons are large enough to account for the observed values of Δχ and χ z for the aromatic hydrocarbons. An analysis of the chemical shifts of the aromatic hydrocarbons shows that they are in agreement with a Δχ value for benzene of approximately —31×10—6 in good agreement with the SCF—MO value of χ L . Calculations of the chemical shifts due to localized magnetic anisotropies show that they are quite small and can frequently be neglected. In calculating ring current contributions to the chemical shift in NMR spectroscopy it is recommended that the SCF—MO value of χ L , —30×10—6, be used in place of the classical value, —49.5×10—6.