Observing order in chemical disorder

By Tom Sharp on September 14, 2016

One of the investigated candidates for a new-rare-earth free permanent magnetic material is FeNi with L1₀ structure. A recent paper in Journal of Physics: Condensed Matter sheds a fascinating insight on chemical disorder in these systems. The author Andreas Frisk explains below.

Materials with the L1₀ structure are special as permanent magnetic materials since they possess a magnetic anisotropy due to their chemical anisotropy composed of monolayer (ML) stacking of the two constitution elements. Determining if a material has this chemical ordering can be difficult, and in our article we explain how two types of disorder affect the X-ray diffraction (XRD) pattern.

The L1₀ structure can be viewed as tetragonal strained face centred cubic (fcc). Determining if a material has a strained fcc structure is easily done with standard XRD techniques. But for a material to have the L1₀ structure it must also have the chemical ordering. The usual approach to determine the chemical order is to measure the 001 diffraction reflection, which is forbidden for a chemically disordered fcc, but allowed for L1₀ due to the difference in atomic form factors of the constituent elements. In other words, this reflection is caused by the chemical order, and would be lost if there was chemical disorder.

In our article we describe a new way to synthesize the FeNi L1₀ phase in thin film form by monolayer deposition on Si(001) using sputtering. We did not succeed in achieving a perfectly chemically ordered L1₀ phase, but we saw something very interesting in the diffraction patterns: a split 001 reflection.

This split we can explain as originating from order within the chemical disorder. During the deposition we aimed to deposit single monolayers of each element but the deposition rates differed slightly from the intended ones. This resulted in a composition modulation throughout the stack of layers, with a measurable periodicity, dependent on the difference from the nominal composition, figure 1b. We used a simple model to simulate a diffraction pattern from such a structure and it corresponded qualitatively well with the measured pattern, figure 2.

Figure 1: Different cases of disorder in L10 material grown by ML deposition. a) perfect chemical order, b) composition modulation due to a 25% too high deposition rate of Ni, c) perfectly ordered grains phase shifted 1ML, d) a combination of the effects. Frisk et al. J. Phys. Cond. Mat. 28 406002 © IOP Publishing, All Rights Reserved.

There was actually another contribution to the disorder, which we believe is the main source of the small measured magnetic anisotropy: the surfaces we grow the films on were not flat, Figure 1c. If there is an atomic step on a surface there will be two parts of the film shifted relative to each other by 1ML, and the combined diffraction from two perfectly chemically-ordered parts will be zero. It is interesting to note that if this would be the only cause of chemical disorder we would not have seen any 001 diffraction at all, even if our ML deposition was perfect. However, because there was already some disorder in our films, the different parts could never be perfectly out of phase, and thus the 001 diffraction was not killed, figure 2 c.

Figure 2: Simulated XRD patterns from the structures in Figure 1, 001 reflection at 1.75 Å-1, 002 reflections at 3.5 Å-1. a) and b) corresponds to a) and b) in Figure 1. c) corresponds to d) in Figure 1: Case c) in Figure 1 would not have any 001 reflection at all. Frisk et al. J. Phys. Cond. Mat. 28 406002 © IOP Publishing, All Rights Reserved.

These effects are important to keep in mind when doing XRD on L1₀ materials in general, especially the extinction of the 001 reflection. This will occur if the material is perfectly ordered within grains which are shifted 1ML relative to each other. Thus, one may have achieved a high degree of L1₀ phase without seeing it in XRD measurements.

Andreas Frisk is a PhD student in experimental materials physics at the Department of Physics and Astronomy, Uppsala University, Sweden. He is doing work on anisotropy in magnetic thin films deposited by magnetron sputtering. Besides the work presented here he has done studies on combinatorial sputtering of amorphous thin films. His two main areas of interest are the control of thin film growth by sputtering and X-ray diffraction techniques, especially energy-dependent scattering.

Professor Bengt Lindgren is professor emeritus in experimental physics at the Department of Physics and Astronomy, Uppsala University, Sweden.

Dr Spyridon D Pappas is a postdoc at the Department of Physics and Astronomy, Uppsala University, Sweden. His research interests are currently focused on the study of magnetization dynamics in continuous magnetic thin films and patterned magnetic materials, by means of time-resolved techniques.

Dr Erik Johansson is a Material Chemist at the Power Device department at ABB, Corporate Research, Sweden. He is interested in magnetic materials for electromagnetic applications.

Professor Gabriella Andersson is an experimental physicist at the Department of Physics and Astronomy, Uppsala University, Sweden. She specializes in tuning of properties in magnetic thin film materials, with special emphasis on anisotropy and domain structure.