SINCE the structure of the sigma phase was established by Bergman and Shoemaker1, it has been found in many binary alloys of transition elements (see, for example, Haworth and Hume-Rothery2). Where the phase boundaries have been established, the width of the range of composition varies from less than 1 per cent in the molybdenum–cobalt system to about 20 per cent in the iron-vanadium system. On the basis of the positions of the phase in various binary alloy phase diagrams, it has been suggested that the sigma phase is some kind of electron compound (see, for example, Kasper3). Greenfield and Beck4 concluded that while the sigma phase might be associated with a particular average number of vacancies in electronic ‘d’ states per atom, no self-consistent scheme could be developed to reconcile both theory and experiment. Nevertheless, the possibility that ‘d’ state vacancies might play a part in stabilizing the structure suggests that magnetic evidence would prove of value. Consequently, in a survey of many binary and ternary sigma phase alloys, Nevitt and Beck5 concluded that all were ferromagnetic, having Curie points in the temperature range 50–200° K. No measurements had been made, however, of the mean magnetic moment per atom (PB) of any ferromagnetic sigma phase, or of its variation, if any, with composition, and so the present work was undertaken, the iron-vanadium system being chosen because of the comparatively wide range of composition over which the sigma phase extends.
Read full abstract