Two possible atomistic mechanisms of grain boundary migration in nickel oxide are examined for two coincidence boundaries using computer simulation techniques. In mechanism I, the boundary migrates by anion vacancy diffusion across it (calculated activation energy 0.64 eV in a (21̄1)/[011] boundary). In mechanism II, ions migrate across the boundary to add onto the opposite crystal lattice without the need for vacancies (activation energy 0.78 eV in a (12̄2)/[011] boundary and control by cation migration). The activation energy for an anion diffusion jump down a (21̄1)/[011] boundary is calculated as 1.57 eV. Anion vacancy supply down the boundary is therefore expected to be rate controlling for any migration mechanism requiring anion vacancies (i.e. mechanism I). There are few literature values for the grain growth activation energy in bulk NiO but values of ~0.9 eV are deduced from published data on grain sizes in NiO scales. 0.9 eV is low in comparison with literature values for a variety of ceramics. Comparison between 0.9 eV and the computer predicted activation energies for defect migration suggests that, unless the effective anion vacancy formation energy at a general grain boundary is considerably lower than expected, it is likely that anion vacancy formation is not required for boundary migration in NiO.
Read full abstract