| Summary: | Ferromagnetism in 3d metals is re-examined in a simple band model. It is shown that the molecular field model cannot account for the low values of coercive field in ferromagnetic pure metals, and that the standard Stoner theory of band ferromagnetism incorrectly evaluates the total electronic energy and cannot predict reasonable Curie temperatures. A simple band model for magnetism is formulated, yielding a criterion for ferromagnetism even in absence of a Hubbard term, involving only the density of states (DOS) at the Fermi level, its derivative and the filling of the 3d band. By introducing a double-peaked DOS, one may explain the occurrence of ferromagnetism in bcc Fe, hcp-fcc Co and fcc Ni, the stabilization of fcc-hcp or bcc structures across all 3d elements, the occurrence of antiferromagnetism in chromium, and derive reasonable Curie temperatures. ‘Re-entrant’ ferromagnetism is predicted at ultrahigh temperatures, suggesting an alternate origin for the geomagnetic field.
|
|---|
