No abstract

Amorphous thin films of NdxCo1−x and NdxFe1−x alloys were prepared over the compositional range 0.08⩽x⩽0.71 by e-beam evaporation. Magnetization and anisotropy of the samples were studied over a wide temperature range with the aid of a force balance magnetometer, Hall effect measurements, and Mössbauer spectroscopy. It was found that the magnetization of the alloys could not be accounted for by a completely collinear alignment (ferromagnetic) of the Nd and transition-metal subnetworks. Mean field analysis of the magnetization data showed a large reduction of the Nd–transition-metal exchange coupling as compared to their Gd analogs. A model was developed which requires that Nd be dispersed in a cone whose axis is parallel to that of the transition-metal subnetwork by strong coupling to randomly oriented local crystal field axes. This dispersion reduces the Nd net moment to 77% of its free-ion moment in NdxCo1−x alloys and to 25% of its free-ion moment in NdxFe1−x alloys. There is evidence that some dispersion also occurs in the Fe subnetwork. Anisotropy in NdxCo1−x alloys was found to be proportional to the square of the Nd subnetwork magnetization, Ku=CM2Nd. C was exponentially dependent on concentration according to the relation C=6 exp(−10x).

The structures of electroplated and vapor-deposited copper films have been studied by x-ray diffraction. While there was no evidence of stacking faults in any of these films, the electroplated films were characterized by presence of twins, large thickness-dependent microstrains, and small particle sizes. The vapor-deposited films, on the other hand, showed no twins, smaller microstrains, and large thickness-dependent particle sizes. Analysis of the electrical resistivity of the electroplated films indicated that the twin resistivity of copper is, at most, half of the grain-boundary resistivity.

A procedure involving the use of Fourier coefficients is suggested for the separation of the el component from an experimental diffraction profile. In contrast to the well-known Rachinger method, the procedure described here does not require experimental data at any predetermined intervals. Furthermore, the separation error, if any, is evenly distributed. As in the Rachinger method, prior knowledge of the ~1-e2 angular separation and the ratio R= Iot2(max)/Ioq(max) is assumed. If, however, R is unknown, then the mathematical analysis can be extended to determine the value of R.

Data on resistivity at room temperature in polycrystalline and single-crystal Permalloy (80 Ni-20 Fe) thin films are analyzed in terms of grain boundary scattering and the dc size effect, respectively.