Abstract
The complete rotational diffusiontensors of the three symmetric top molecules PCl3, PBr3, and POCl3 are obtained from viscosity and halogen NMRrelaxation time data. These tensors allow us to obtain estimates of the average correlation times for each molecule, which in turn are used in Hubbard's isotropic spin—rotation equation to derive NMR values of their spin—rotation interaction constants C eff. These and similarly derived values of C eff for a number of other 31P compounds are compared to values of the spin—rotation interaction constant C σ deduced from an absolute shielding scale based on a recent molecular beam determination of the 31P spin—rotation interaction tensor in PH3. In contrast to previous work the values of C σ are in reasonable agreement with the values of C eff, therefore offering evidence for the validity of Hubbard's relationship. Earlier discrepancies between C σ and C eff appear to be mainly due to poor estimates of the average correlation times, although effects of anisotropy appear to contribute slightly. The results of this paper offer evidence that the so‐called microviscosity approach should be the favored method for estimating average reorientational correlation times in complex rigid molecules and correlation times for reorientation perpendicular to the symmetry axis in symmetric top molecules whenever direct quadrupolar or dielectric relaxation times are not available. The derived rotational diffusiontensors for PCl3, PBr3, and POCl3 are used to discuss a recent extension to Hubbard's isotropic spin—rotation equation which takes account of the anisotropic rotational motions in symmetric top molecules. In favorable circumstances this approach can be used to calculate the entire spin—rotation interaction tensor for a nucleus in a symmetric top molecule.