Abstract
The experiments of Walsh, Jeener, and Bloembergen on Mn2+: MgO exhibit a hyperfine-coupling constant A which decreases with increasing temperature, contrary to the contribution from lattice thermal expansion. This paper proposes a mechanism in which excited s-like configurations are admixed into the 3d5 configuration by the orbit-lattice interaction. The hyperfine field resulting from such admixtures is opposite to the T=0 (primarily core polarization) hyperfine field. As the amplitude of phonon vibrations is thermally increased, a net decrease of A is predicted, the temperature dependence of which is in agreement with the results of Walsh et al. An estimate of the magnitude of this mechanism using a near-neighbor point-charge model falls somewhat short of the observed magnitude, but the discrepancy is shown to originate in the evaluation of the orbit-lattice coupling constant. In view of the primitive model currently used to compute these coupling constants, agreement in magnitude of the calculated and observed shift is as good as can be expected.