Zeeman Study of the Jahn-Teller Effect in theT2g3State ofAl2O3:V3+

The fine structure of the zero-phonon ${}_{}{}^3{}_{}{}^{}T_{2g}^{}{}_{}{}^{}$ state of ${\mathrm{Al}}_2$ ${\mathrm{O}}_3$: ${\mathrm{V}}^{3+}$ arising from the trigonal field and spin-orbit coupling is anomalously small. Ligand-field theory predicts a 400-${\mathrm{cm}}^{-1\mathrm{}}$ splitting, while the observed levels, first fully identified by Scott and Sturge, span only 14 ${\mathrm{cm}}^{-1\mathrm{}}$. This has been attributed to a large tetragonal Jahn-Teller effect, and a detailed analysis of the zero-phonon levels was made by Scott and Sturge on the basis of Ham's model calculations for octahedral molecules. We have made new measurements on the 0-0 ${}_{}{}^3{}_{}{}^{}T_{2g}^{}{}_{}{}^{}$ band, including the axial spectrum, which shows one $\sigma$-polarized transition to be magnetic dipole in origin, and the Zeeman effect of the $\pi$ and $\sigma$ spectra at 10°K in fields up to 40 kG parallel and perpendicular to the trigonal axias. The validity of the Ham-Scott-Sturge model has also been further investigated. We have shown that the details of the Scott-Sturge theory require modification, and that it is possible to explain the observed zero-field level pattern within the framework of Ham's general model if, but only if, the assumption of equal quenching of first-order trigonal field and spin-orbit splittings is dropped. In addition, we have obtained values of the first- and second-order trigonal field and spin-orbit coupling energies which reproduce quantitatively the zero-field energies, Zeeman behavior, selection rules, and intensities of all transitions. It is found that the first-order spin-orbit splitting is some four times less quenched than the first-order trigonal-field splitting.