Figure 1 shows high-angle x-ray diffraction spectra for a purely (001)-oriented (Fe40Å/Si14Å)x60 multilayer grown on MgO(001) and a purely (011)-oriented multilayer grown simultaneously on glass. While multiple superlattice satellites are observed around the Fe(002) peak for the film grown on MgO, the film on glass has only one peak corresponding to (011)-textured growth. A single superlattice satellite is typically observed in multilayers on glass on the low-angle side in keeping with previous observations.[6] An estimate of the crystallite sizes in these films can be derived using the Scherrer formula. This analysis gives a coherence length of 165Å for the film on glass and 188Å for the film on MgO.
Since the crystalline coherence lengths of these films are similar, the presence of high-angle satellites in the film on MgO must be due to better layering. The small-angle x-ray scattering data shown in Figure 2 confirm this hypothesis. The multilayer on MgO has 4 low-angle peaks, indicating a moderate degree of composition modulation. The multilayer on glass shows only two relatively broad peaks, indicating larger interfacial roughness and less order in the layering. The low-angle x-ray spectra are consistent with rocking curves which are only about 1° wide for films grown on MgO and Al2O3 but are typically 10° to 15° wide for films on glass. The bilayer periods determined from the low-angle peak positions are (41.0 ± 0.1)Å for the MgO film and (40.9 ± 0.2)Å for the glass film, the same within experimental error.
ø scans of the MgO and Fe [110] peaks for the film on the MgO substrate (not shown) demonstrate that it is oriented in-plane. While theta-2theta scans for (011)-oriented multilayers grown on Al2O3 substrates (not shown) also show multiple high-angle superlattice satellites, the ø scans for this film indicate only weak orientation in-plane.
The shape of the high-angle peaks and their superlattice satellites are described by a well-known theory.[7] Application of this theory to the Fe/Si multilayers is difficult because the iron silicide lattice constant, the thickness of the remaining pure Fe and the thickness of the iron silicide spacer can be estimated only roughly. A precise determination of the silicide lattice constant should make a quantitative analysis of these satellite features possible.