Strain adaption in epitaxial Fe-Rh nanostructures

Nanostructured magnetic materials continuously attract tremendous interest in both science and technology and their applications are found, for example, in the area of magnetic data storage. One versatile, and technologically mature, route to design and tailor modern magnetic materials is the growth of thin films as epitaxial heterostructures. The precise control of the associated epitaxial strain, in other words an elastic deformation, can profoundly alter the intrinsic magnetic properties of the coherently grown layers. However, going beyond the elastic regime, strain may also create lattice instabilities, reshaping the material’s energy landscape, and possibly promoting novel, metastable phases and/or the spontaneous formation of magnetic nanostructures. In this thesis, the iron-rhodium (Fe-Rh) binary alloy system is identified as a promising candidate for epitaxial strain-tailoring due to the presence of both, fragile and competing magnetic exchange interactions and inherent lattice instabilities. The close entanglement of these properties may lead to unprecedented strain adaption mechanisms accompanied by beneficially modified magnetic characteristics. Indeed, in the course of this work, it was discovered that epitaxial layers of chemically disordered equiatomic FeRh grown on tungsten (W) buffer layers show a novel strain adaption behavior. Essentially, the strain triggers a lattice instability, which in turn drives the film from a tetragonal into an orthorhombic structure, featuring a 90° domain pattern, reminiscent of adaptive martensites. The structural changes have a profound impact on magnetism, suppressing ferromagnetic (FM) order and eventually resulting in a spin glas (SG) configuration at low temperatures. A study of the thickness dependence evidenced a gradual evolution from a tetragonally distorted lattice to the adaptive orthorhombic structure in films with increasing thickness, representing a coherent release of epitaxial stress and hence decrease of the system’s elastic energy. Then, the compositional limit of the lattice instability was explored. Beyond a Fe content of 72 at.-% it is no longer possible to accommodate the strain by adoption of the orthorhombic phase. Instead a fully relaxed, bcc phase is found, which shows its expected FM ground state. The influence of the epitaxial misfit was investigated by using tungsten-vanadium (W-V) alloy buffer layers. Here, a distinct dependence of the final structure on the growth temperature was revealed. FeRh films grown at ambient temperatures on buffer layers with decreasing mismatch (compared to pure W) develop a structure similar to the orthorhombic phase, which features an extra directional tilting in order to match to the W-V lattice constant. However, at elevated growth temperatures the films spontaneously segregate on the lateral scale, into two different phases: the orthorhombic and a partially ordered body-centered-tetragonal (bct) phase, reducing the elastic energy in each individual phase. Most importantly, these two phases are in different magnetic states, the first being paramagnetic (PM) and the second ferromagnetic (FM). The resulting nano-scale arrangement of the PM and FM phase can be described as a strain-induced self-assembled magnetic nanostructure. Hence, the newly found lattice instability is directly placed into an application related context, potentially allowing the bottom up down-scaling of the magnetic bit size in FeRh based data storage media. The findings of this work extend the concepts of strain-induced or strain-engineered magnetic nanostructures from the purely nano-twinned structures to a two-phase-adaption mechanism. On a more general note, the presented studies highlight the fascinating properties of hitherto unknown metastable phases in the Fe-Rh system, which may motivate similar studies in related binary or ternary alloy systems, likewise offering advantageous prerequisites