The character dependence of interfacial energies in magnesia

The thermodynamic driving force for the formation and evolution of microstructures in polycrystalline materials is provided by the excess free energy per unit area associated with interfaces. Data relating the energy of these interfaces to their character, or crystallographic orientation, are required to understand, predict, and control microstructural evolution. We have developed techniques to extract the population and relative energies for all physically distinct surfaces and grain boundaries and applied the techniques to a polycrystalline magnesia sample. The techniques rely on using microscopic observations to characterize the geometry and crystallography of large quantities of interfaces. From these microscopic measurements, the character of the interfaces can be specified, and at triple junctions, where three interfaces intersect, the local equilibrium condition can be used to calculate the energies from the interfacial geometry and crystallography.; We have evaluated the surface energy anisotropy by making observations of circumferential thermal grooves on the surface of the magnesia sample. By examining 269 circumferential thermal grooves, we find that, for magnesia at 1400°C, the surface energy has a minimum at (100) and a maximum at (111). The relative energies for the low index planes are γ ₁₁₀/γ₁₀₀ = 1.07 ± 0.04 and γ₁₁₁/γ ₁₀₀ = 1.17 ± 0.04. These results are consistent with observations of orientation stability at 1400°C and with previous theoretical predictions and experimental results for magnesia surfaces.; From the magnesia sample, we have also extracted the geometric and crystallographic configuration of 5.4mm2 of grain boundary area and 19094 grain boundary triple junctions in a volume of 0.15 mm3. These data have enabled us to specify the population and relative energy of grain boundaries over all five macroscopic parameters. The population exhibits a strong preference for grain boundaries with ⟨100⟩ plane normals. By comparing the character distribution to several boundary energy models, we find that the predicted energies vary inversely with the observed population. Furthermore, we find a substantial inverse correlation between the distribution and the energies calculated from the triple junction measurements (r_s = −0.77). In general, we find that boundaries with ⟨100⟩ plane normals have reduced energies, which is probably related to the relatively high coordination number of atoms on this plane and its low surface energy.