Torque Measurements of the Anisotropy of Energy and Magnetization of Nickel

An analysis of experimental methods and the design of an appropriate torquemeter made it possible to improve by several orders of magnitude the accuracy of anisotropy measurements. The experiments performed on nickel between 100°K and the Curie point allowed, for instance, the determination of the eighth‐order anisotropy constant $K_3$ within a few percent. The anisotropy constants $K_1$ , $K_2$ , and $K_3$ are found to be negative in this whole temperature range. An explanation is given of the main discrepancies in published data between the numerous determinations of anisotropy constants by static methods. The influences of both anisotropy of energy and anisotropy of magnetization are separated and measured as functions of applied magnetic field and temperature in the above‐mentioned range. This interpretation of torque measurements has been confirmed by a direct determination of anisotropic magnetization of nickel at room temperature, and an absolute measurement of the saturation magnetization of nickel at 20°C has given $\text{55.01±0.05\hspace{0.17em}}\mathrm{emu}/{\mathrm{cm}}^3$ , a value higher by more than one percent than that usually quoted. Experimental evidence is given that the determinations of anisotropy constants from static measurements and from ferromagnetic resonance experiments may lead to fundamentally different results. Finally, a new phenomenological model of ferromagnetic anisotropy is proposed, introducing an effective field to describe spin‐orbit interaction. This model seems to give a better agreement with experiments than the previous ones.