The low-temperature specific heats of some nickel-based iron and copper alloys

Abstract

Measurements have been made of the specific heats over the range 1.2 to 4.2°K of a series of face-centred cubic alloys of nickel with iron (nine alloys) and nickel with copper (seven alloys). The low-temperature specific heat of the ferromagnetic alloys may be represented by the equation C = γT + βT 3 + αT 3/2 , the successive terms corresponding to the electronic, lattice and spin-wave contributions. If the third term is appreciable the standard plot of C/T against T 2 will show a concavity towards the T 2 axis. Such plots are given for all the alloys. The best values of γ, β , and α , and the standard errors, are obtained by the analytical and computer methods previously described. The results are presented in two tables, and particular features of them and comparisons with other results are shown graphically. (For the paramagnetic nickel—copper alloys the third term does not apply, and a fairly satis­factory fit is obtained by replacing it by a constant, ϵ .) Although the results for the variation of γ with composition are believed to be more reliable than those previously given, the general character of the curves is similar. The values of β for Ni and the Ni-rich alloys agree closely with those deduced from the elastic constants, but over the whole range of composi­tion, especially for the Ni-Cu alloys, there is more scatter than was expected. This is tenta­tively attributed to residual inhomogeneities or clustering effects. For the Ni-Fe alloys a spin-wave contribution is definitely indicated of the same order of magnitude as that deduced by other methods. For the Ni-Cu alloys the presence of a spin-wave contribution is less certain, partly owing to the greater scatter in the values for β .