Metal-ceramic ion-beam mixing: A quest for general principles

One of the most important features regarding ion-beam mixing of metals-on-oxides is that discussed recently by McHargue et al. They showed that for metals on Al2O3, ZrO2, YPO4, and Si3N4 the tendency was that mixing would not be expected in most instances using the enthalpy rule and that experimentally one normally observes either no detectable mixing, or what is termed segregation or agglomeration, or only ballistic mixing (the latter, of course, is independent of thermodynamic considerations). A notable exception was metals on SiO2 and SiC. There is thus a fundamental difference from the mixing of metals-on-metals, where there are only very few examples of systems which do not mix, and where the small but universal ballistic component is commonly supplemented by a much larger component due to bombardment-induced residual defects ('defect-enhanced motion'). By referring to the extensive defect calculations of Catlow, Mackrodt and coauthors it emerges that the motion of the basic defects is significantly different. With all metals the interstitial is mobile at 300 K, while the vacancy is sometimes mobile or, at worst, becomes mobile with only a moderate increase in temperature. Furthermore, such changes as occur at higher temperatures invariably occur in the sense expected. With oxides, however, both cation and anion vacancies typically show motional temperatures exceeding similar to 600 K. The notable exception is the anion vacancy of LaAlO3, UO2 and cubic-ZrO2. The situation with interstitials is less straightforward. Nevertheless, one of the several conclusions to be reached is that the failure to observe extensive mixing with appropriate metals-on-oxides is due to the lack of a contribution from defect-enhanced motion, this being in turn due to the target temperature being too low. The existence of an activation energy may also be involved with metals-on-oxides, but we are unable to comment on this possibility.