A series of spectra from DyFe/Dy multilayers were obtained to investigate the effect of reduced thickness on DyFe thin films between and . A limiting factor on examining such thin films using Mössbauer spectroscopy is that the signal to noise ratio will be far too large when there is so little iron to obtain a signal from, especially if the iron is non-enriched with the Fe isotope.
To counter this the signal from a number of such films needs to be accumulated. To achieve this we used a number of DyFe films in a multilayer, spaced by layers of dysprosium. Dysprosium is not magnetically ordered at room temperature, with a , and provides good epitaxial growth conditions for the DyFe layers. The substrate, buffer and seed are the same as for the previous thin film samples.
The spectra are shown in Figure 6.6 and the fitting parameters are given in Table 6.5. The spectra have been fitted with four sextets of equal intensity, adopting a model of a low symmetry magnetic easy axis giving four inequivalent iron sites. No particular easy axis direction is assumed in the fits. The relative intensity of lines 2 and 5 to lines 1,3,4 and 6 has been fixed to be equal for all four components but free to vary. All other parameters were free to vary individually.
Comparison between the four dysprosium spaced samples shows little significant variation in hyperfine parameters. The isomer shift and quadrupole splitting values for all sites are identical for the four samples within a fitting error of . The average angle which the spin moments make to the sample plane show no trend with thickness and at most vary by , comparable to the nominal error on the source and sample positioning of .
There are small variations in the hyperfine fields for particular sites but whether these are physical or artifacts of the overlapping nature of four components in a small area is hard to determine. The average hyperfine field is constant across all samples within .
The DyFe sample was also fitted with the same parameter restrictions as the dysprosium spaced samples as a comparison with a ``thick'' film. The results are also shown in Table 6.5. This spectrum was recorded on a different spectrometer to the dysprosium spaced samples so the margins of error when comparing the hyperfine parameters and average moment angle are larger. Again there is no significant variation to be seen.
From these results we can conclude that the mechanisms which were present in the sample are unaffected by reducing the thickness of the layer to at least .
Dr John Bland, 15/03/2003