Abstract
Magnetic thin films of 3d-4d/5d transition metal alloys such as Fe-Pt, Co-Pt, and Fe-Pd are of technological interest due to their ordered L10 tetragonal intermetallic phase, exhibiting large magnetocrystalline anisotropies of K ~ 10^7 to 10^8 ergs/cm3. Anisotropies of this magnitude are comparable to lanthanide based 3d-4f rare earth permanent magnets, which have become ubiquitous since their development in the 1970s. Despite their prevalence, rare earth magnets are limited by a vulnerability to corrosion, as well as brittleness due to a lack of available slips systems in their complex crystal structures; both causing intractable problems for nanoscale applications. In juxtaposition, the strong hard-magnet properties of 3d-4d/5d magnetic alloys, combined with the ductility and chemical inertness of their ennobled metallic nature, allow these material systems to remain above the thermally induced KV/kBT superparamagnetic limit at the nanometer scale. This combination of properties is ideal for applications in ultra-high-density magnetic storage or micro-electro-mechanical systems. Within this class of materials, Fe-Pd alloys possess comparatively moderate magnetocrystalline anisotropies relative to Co-Pt and Fe-Pt. The Fe-Pd phase diagram, however, exhibits a considerably lower range of order-disorder transition temperatures, rendering the material well-suited for nanostructured magnetic applications by enabling lower processing temperatures. In addition, the higher economic demand for Pt makes Pd-based alternatives of considerable technological interest.
Experimental work to date near the Fe38.5Pd61.5 eutectoid between the chemically ordered L10 and L12 phases of the Fe-Pd system, bounding one side of the technologically relevant L10 phase region, is limited and has left large uncertainties in the experimental phase diagram. The related Co-Pt system has been shown to decompose under bulk conditions into a novel, strain-induced chessboard microstructure at the eutectoid composition between its ordered L10 and L12 intermetallic phases due to coherency strain, making it likely that other 3d-4d/5d material systems may also produce unique strain-induced microstructural behavior. Strain-induced effects are indeed observed for the Fe38.5Pd61.5 thin films presented, but they are of a considerably different nature than the microstructural behavior produced at the Co-Pt eutectoid.
Epitaxial films of Fe38.5Pd61.5 at the L10-L12 eutectoid composition have been grown on MgO (001) oriented substrates by pulsed laser deposition, probing this unexplored region of the binary diagram and solidifying a gap in the experimental record. This thesis advances the scientific understanding of ordered magnetic films by introducing two new strain-induced ordering phenomena. Thin films of Fe38.5Pd61.5 deposited above 600°C are found in a single ordered phase, initially surmised to be L12 due to magnetic properties and the location of the X-Ray Diffraction (XRD) peaks. Careful analysis of peak intensities results in the prediction of anomalous long-range ordering parameters. Quantitative XRD analysis of the films confirms that this is due to a perturbation in the Pd-site occupancy of the non-stoichiometric Fe atoms in the films; resulting in a hybridization of the L10 and L12 ordered structures. This L1' hybridized ordered structure, first postulated by thermodynamic principles to exist for the Au-Cu system†, is believed to be induced by the accommodation of epitaxial strain from the substrate. Classical notations are expanded in discussion of the L1' hybrid phase, defining two independent long-range ordering parameters. Along with its verification, the thermodynamic behavior of this new strain-induced phase is addressed in relation to the equilibrium phase diagram. In addition to this new hybrid phase, Fe38.5Pd61.5 films grown at 550°C have been found to possess a unique two-phase microstructure of prismatic 50 to 100 nm disordered A1 secondary phases with <110> oriented facets, embedded within an ordered L12 matrix. Large strain energies during the early stages of 550°C epitaxial film growth are hypothesized to lead to this two-phase decomposition and the abnormal precipitation of a disordered phase from an ordered state. A L10-L12 two-phase region, as predicted by the equilibrium diagram, remains unobserved.
† W. Shockley, J. Chem. Phys. 6, 130 (1938)
