Specific Heat of Dilute Alloys

Recently Zimmerman has observed that the addition of Mn to Cu produces a large contribution to the specific heat which, at low temperatures, is linear in temperature and independent of Mn concentration. It is shown that: (1) this remarkable result can be explained in terms of the well-known Ruderman-Kittel-Yosida spin-spin coupling via conduction electrons; (2) the specific heat results of Beck et al. on FeV and FeCr alloys are probably of essentially the same origin as those of Zimmerman on Cu Mn; (3) that there are serious objections to the mechanism of antiferromagnetism postulated by Overhauser and used by him to explain the specific heat results. In contrast to the Overhauser theory, no new concepts are involved and it is suggested that the large specific heat comes from a small fraction of Mn spins which, because of the random nature of the alloy, happen to be in small effective fields and therefore not strictly aligned. The theory depends on two plausible assumptions which have not, at present, been proved rigorously valid.