Modelling of thermal desorption of hydrogen from metals

The thermal desorption technique can be used in principle to determine the trapping characteristics of different microstructural trap sites in metals provided there are adequate models to fit to the experimental data. A brief review of models of thermal desorption is presented which indicates that there are limitations in the assumptions made or in the scope of existing models. A more rigorous mathematical model has now been developed which accounts for diffusion, detrapping, and retrapping at one or more type of trap site and which allows for varying trap occupancy. The effect of material and experimental variables on the thermal desorption spectrum has been evaluated and the validity of simple models of desorption assessed. The simpler analytical models, such as the detrapping model of Lee and Lee, in which diffusion is neglected relative to detrapping, do not inspire confidence and are applicable only under very limiting circumstances; for example, in low alloy steels at very low hydrogen contents. It is recommended that thermal desorption measurements be made at progressively decreasing values of initial hydrogen content until the simple analysis yields a consistent value for the trapping parameters. This experimental approach is applicable also to models of thermal desorption which account for diffusion using an effective diffusivity, since trap occupancy is neglected in these. The more rigorous model described herein can be used to determine the binding energy of the traps directly which, together with the density of trap sites, is the most important parameter with respect to hydrogen assisted cracking. The height of the energy barrier to trapping, at constant value of the binding energy, is shown to have only a modest effect on the thermal desorption spectrum compared with the impact of binding energy and of density of trap sites.