Abstract
A model has been developed which can explain observed discrepancies in the determination by different techniques of the activation energy for self-diffusion in plastic crystals. In this model, impurities play an essential role. The basic hypothesis is that an impurity stabilizes the crystal lattice by reducing the self-diffusion around it, to a range of several molecular distances. This hypothesis allows a semi-quantitative fit of published experimental N.M.R. data on perfluorocyclohexane, and the parameters of the model are found to be consistent with previous work on cyclohexane. Moreover, as observed experimentally, it leads to the conclusion that activation energies deduced from N.M.R. linewidth measurements are in general smaller than those deduced from radiotracer data. It is concluded that for infinitely pure samples, all techniques should lead to convergent results for self-diffusion activation energies.