On the contribution of the fermi contact term to the magnetic field at the nucleus

The dominant source of the effective magnetic field at the nuclei in ferromagnets apparently comes from the Fermi contact contribution of the s electrons in the core. This contribution was investigated for transition element atoms and ions by means of a series of spin (or exchange) polarized Hartree-Fock calculations (calculations where one-electron orbitals of differing m_s are allowed different radial dependence). Calculations were done for both free-atom and crude crystalline environments. The sensitivity of the 3s electrons to 3d behavior was found to contribute to large differences in calculated effective fields. The differences between core polarizations in metals and ions are discussed with the use of a model for metallic iron (i.e., a 3d⁸ configuration) which is more consistent with neutron diffraction and energy band results than the free atom 3d⁶ state (the spin polarized 3d⁸ calculation shows an expanded 3d charge density but an almost unchanged spin density relative to the 3d⁶ results). The resultant (negative) increase in the effective field is not, however, large enough to overcome the positive contributions of the outer electrons (estimated à, la Marshall) so that although a net negative field is found it is not large enough in magnitude to agree with the Mössbauer measurements. A parallel calculation by Goodings and Heine, based on an artificially expanded 3d charge and spin densities, is also found to be similarly deficient. Their model is discussed and shown to be incompatible with neutron experiments and energy band calculations.