材料期刊网

Calculating the S-state splitting of a d5 ion (Fe3+, Mn2+) in a crystal has been a long-standing difficult problem since 1934 when it was discussed by Van Vleck and Penney. Low and Rosengarten have concluded that crystal-field theory is not capable of providing a unified interpretation for the spectrum and for the S-state zero-field splitting of d5 6S ions. In the present paper, by utilizing a method of combined perturbation composed of the electrostatic, crystal-field, and spin-orbit-coupling energies, and by diagonalizing a set of complete energy matrices constructed for a d5 configuration and having a total order of 252, a reasonable interpretation is obtained for the d-d spectrum and the EPR cubic zero-field splitting of Fe3+ ions located at octahedral and tetrahedral sites in yttrium gallium garnet. This shows that crystal-field theory is as well suited for application to d5 ions and to other d(n) ions and that the earlier Low-Rosengarten conclusion is wrong. It is also shown that contributions to the ground-state splitting arising from various excited states of the d5 configuration or from various components of the crystal field do not obey a linear-superposition rule. In the Appendix, a detailed explicit description of the matrices is given.