ABSTRACT
In reality, NiO and MgO are a pretty good diffusion couple to try a marker experiment on. The main problem with marker experiments on compounds is that compounds are brittle. If the diffusion couple is made by a solid-state interdiffusion process at some elevated temperature, then on cooling the total thermal expansion difference between the components must be small. For example, if the difference in expansion coefficient were only ∆α ≅ × − −1 10 7 1K (about 5%), and Young’s modulus E MPa≅ 250 (Kingery et al 1976) cooling a diffusion couple from 1200°C would produce stresses in the couple on the order of σ ε α= ≅ ≅ × ( ) × ≅ × ≅−E TE Pa psi∆ ∆ 1 10 1200 210 10 2 52 10 36 0007 9 7. , close to the fracture strength of these two oxides. Actually, the thermal expansions of NiO and MgO are very close, but both vary with temperature from about 13 10 15 106 6 1× ×− − −to K between 200 and 1200°C (Nielsen and Leipold 1963; Nielsen and Leipold 1965). It would be very fortunate if they matched close enough so that fracture would not occur on cooling. Certainly, there are very few compounds whose thermal expansions are this close. This is one reason that there are not many marker experiments on compounds. Metals are much more forgiving because any stresses introduced by thermal expansion differences are easily relieved by plastic deformation. Some plastic deformation can occur in compounds as well, but only at high temperatures and to a limited extent. Disregarding the experimental difficulties with compound interdiffusion, the models are important for their predicted behavior under different assumptions. However, in the interdiffusion experiments with NiO and MgO single crystals, no markers were used (except for pores at the original interface) (Appel and Pask 1971).
