A method is described for the evaporation of ferromagnetic filmsonto a very thin water soluble polymer coating on glass substrates andtheir subsequent removal onto a liquid glycerin film. Determination ofthe change in magnetic anisotropy on removal provides a measure of thesubstrate constraint contribution to the anisotropy energy. Data aregiven for both Ni-Fe and Ni-Co alloys in the entire range 0 to 100% Ni,deposited at 100°C.
Significant changes in anisotropy upon removal were observed foralmost all Ni-Fe alloys and for Ni-Co alloys in the range 0 to 40% Co.However, essentially a zero percentage change was observed for pure Fe,83% No-Fe and the entire range 0 to 60% Ni in the Ni-Co alloys. Thesedate are in disagreement with present theories of the constraint energyexcept for qualitative agreement in very limited composition ranges.
It has been suggested that the discrepancy between experimental andtheoretical prediction for the anisotropy energy in thin films mayresult from the use of bulk material magnetoelastic constants which areinappropriate for thin films. However, this study of the magneticproperties of epitaxial films strongly suggests the equivalence of themagnetoelastic parameters in thin films and bulk materials. In thisstudy a technique was developed for the preparation of step-free epitaxialfilms of Ni-Fe and Ni-Co alloys deposited at 400°C. The strainsensitivity ΔHk/Ɛ has been measured by ferromagnetic resonance along the[100] and [110] directions in the (001) plane for compositions rangingfrom 44 to 87% Ni in the Ni-Fe alloys and from 70 to 82% Ni in the Ni-Coalloys. The data are in good agreement with the theoretical predictions3(C11-C12)λ100/M and 6C44λ111/M along the [100] and [110] directionsrespectively, evaluated using bulk parameters.
In addition, the strain sensitivity of polycrystalline films depositedat 100°C and 400°C for Ni-Fe alloys and 400°C for Ni-Co alloys hasbeen measured. Contrary to previous investigations, a systematic temperaturedependence is found. Moreover, it is shown that the isotropicmaterial model used previously by others to calculate the strain sensitivityin polycrystalline films is incorrect. However, the apparentproper formulation does not predict experimentally determined result.This discrepancy remains unexplained.
New data for the crystalline anisotropy constant K1 in constrainedepitaxial films of Ni-Fe and Ni-Co are presented. Significant deviations fromvalues in bulk material are observed. However, these deviations arebelieved to result from the substrate constraint and not material differences inherent in thin films.
