Magnetic Anisotropy in Antiferromagnetic Corundum-Type Sesquioxides

Evaluations of the magnetic dipolar anisotropy $K_{\mathrm{MD}}$ are made for the antiferromagnetic Cr, Fe, Ti, and V sesquioxides isomorphous to $\alpha -{\mathrm{Al}}_2{\mathrm{O}}_3$. In the first two systems the fine structure anisotropy $K_{\mathrm{FS}}$ is found by substracting $K_{\mathrm{MD}}$ from the total anisotropy $K$ obtained from antiferromagnetic-resonance (AFMR) data. In Shull magnetic configuration type (c), of which ${\mathrm{Cr}}_2$ ${\mathrm{O}}_3$ is the prototype, $K_{\mathrm{MD}}$ was found to be very sensitive to lattice geometry. Owing to the use of inappropriate crystal data, a previous ${\mathrm{Cr}}_2$ ${\mathrm{O}}_3$ $K_{\mathrm{MD}}$ calculation is shown to be in serious error. The ${\mathrm{Cr}}_2$ ${\mathrm{O}}_3$ AFMR results are reinterpreted accordingly. An analysis is also made for $\alpha -{\mathrm{Fe}}_2{\mathrm{O}}_3$ [Shull case (a)] for which some AFMR data have recently become available. The $\alpha -{\mathrm{Fe}}_2{\mathrm{O}}_3$ spin flip near 250 °K is predictable on a thermodynamic basis. The $K_{\mathrm{FS}}$ results in the Cr and Fe cases show little correlation with the corresponding quantities derived from EPR in diluted magnetic systems. The results of variation of lattice parameters are given for possible application to temperature and stress experiments and to diluted crystals. Since the antiferromagnetic properties of the Ti and V oxides currently are not clear, we present the appropriate $K_{\mathrm{MD}}$ calculations for each of the three possible spin configurations.