X-ray diffraction study of the (magneto-)structural transition in FeRh thin layers
L. Horák
Faculty of Mathematics and Physics, Charles University in Prague, Ke Karlovu 5, 121 16 Praha 2
horak@karlov.mff.cuni.cz
The Fe50Rh50 material exhibits a reversible magneto-structural transition between a room-temperature antiferromagnetic (AFM) and a high-temperature ferromagnetic (FM) phase approximately at 350 K [1]. At room temperature, the magnetic moment located at the Fe atoms are AFM ordered while there are no localized magnetic moments at the Rh atoms [2]. After the transition, the lattice is discretely expanded and there are localized magnetic moment FM ordered at the Fe atoms, and at the Rh atoms as well [2].
However, the principally first-order transition itself displays continuous behaviour with the presence of the structurally transited but non-magnetic phase [3]. Moreover, the transition shows the temperature/field hysteresis in the lattice parameter and in the magnetic net moment [4].
Most probably, the initial growth of the ferromagnetic phase is stimulated by the defects located at the surface and/or the interface with the substrate [4]. Further, in the thin FeRh layers, the presence of the stable residual FM volume is being observed [5]. It was shown that this volume can be located at the layer/substrate or layer/capping interface [5]. Such defected interfaces can be rich of seeds for the formation of ferromagnetic FeRh regions.
Using High-Resolution x-ray diffractometry (HR-XRD), we have measured several samples of FeRh thin layers with various thicknesses. Benefiting from the different lattice parameter of the AFM and the FM phase, we determined from the measured peak-intensity the volume of both phases in the samples. From the temperature dependent measurement (heating and cooling), we reconstructed the hysteresis loop of the FM/AFM volume (figure 1).
At room temperature, the measured curves were evidently asymmetric indicating the presence of the strong AFM-phase peak and the weak FM-phase peak (see figure 2). We were looking for a rich-Fe interfacial layer that could be a source of that residual FM phase. From the x-ray reflectivity, it follows that our FeRh thin layers themselves can be described by a model of a single homogeneous layer as the measured x-ray reflectivity was successfully fitted by this model. The reflectivity proved that there is no observable FeRh sub-layer with different stoichiometry at the interface with the substrate and/or with the capping layer.
On the other hand, the thickness oscillations were present in the diffraction curves for all temperatures. Their frequency corresponded very well to the thickness obtained from the x-ray. Surprisingly, the width of the possible FM-phase peak was as the same as its width above the transition temperature. Later analysis, when we simulated and fitted measured curves (figure 2), supported the suspicion that the residual FM volume (although very small) has in the out-of-plane direction the dimension being equal to the film thickness, i.e., it is grown from the very bottom to the top of the thin layer.
Besides the findings that the residual ferromagnetic volume can be concentrated at the layer/substrate (or cap) interface [5], we found in our samples that this volume can be laterally spread in the layer in a form of thin columns, but already having their final thickness. Just these columns could be the seeds for the emerging FM phase during the heating.
Figure 1. Evolution of the AFM and FM volume during the heating&cooling loop.
Figure 2. Measured diffraction curves for different temperatures (heating and cooling loop). The experimental data (coloured) are fitted with the simulation (black solid curve).