Published online by Cambridge University Press: 10 February 2011
In order to verify the results predicted by the model in Part I of this work [1], permeation experiments were conducted at room and high temperatures on fully-annealed-commerciallypure iron with two kinds of surface treatment, one group with plasma cleaning and presputtering and the other without it. The experimental results show that the diffusivity evaluated by the new model is independent of sample thickness and surface treatment, while the diffusivity evaluated by the time-lag model varies two orders of magnitude. The experimental results confirm that a fine surface treatment yields a low energy barrier for desorption. The energy barrier for either group is higher than the activation energy of diffusion. Consequently, the ratio of drift velocity through surface to that in bulk increases with increasing temperature and makes the time-lag method appropriate at elevated temperatures.