Heusler spin filters for room-temperature spintronic applications

• Online Access: http://hdl.handle.net/2047/D20316398

With current silicon-based technology reaching the limits of scalability, spintronics hopes to extend the life of Moor's law by utilizing the electron spin to improve performance and efficiency of devices. To this end, it is necessary to produce materials with spin-dependent properties that can be utilized to improve current architectures, as well as provide a toolbox of available materials that could be utilized in non-yet-designed systems. In addition to having desirable electronic properties, they should function at or above room temperature and ideally be made with common elements. To fully realize efficient spintronic devices, it is necessary to produce a current where all electrons have the same spin direction, a fully spin-polarized current. The simplest solution would be a half metal, whose density of states contains only one spin at the Fermi energy. However, the spin polarization in a half metal falls off quickly with temperature. To provide a high spin-polarization at room temperature, either the minority gap must be exceptionally large such that the minority channel is fully insulating, or the Curie temperature must be well above room temperature such that the magnetization changes insignificantly between 0~K and 300~K. Spin gapless semiconductors show similar spin-polarizing properties though with the added ability to tune the majority conduction via gating, though they have the same drawbacks with respect to minority gap and magnetization. An alternate method of producing highly spin-polarized currents would use a spin-filter material, a semiconductor that has a different band gap for each spin channel. By utilizing a spin-filter material as a tunneling barrier between non-magnetic metallic contacts, the tunneled current will be spin polarized. Additionally, because the contacts can be nonmagnetic, all the magnetic properties are confined to the barrier itself. Thus, if a zero-moment spin-filter is used, the entire spin injector can be effectively non-magnetic, eliminating the problem of fringing magnetic fields. In this thesis, I have synthesized CrVTiAl in bulk and thin films and studied the properties through a variety of probes, including X-ray diffraction, magneto-transport and magnetometry. CrVTiAl was found to have semiconducting properties consistent with the theoretical predictions. In bulk, CrVTiAl the magnetization was found to change linearly with magnetic field and increase slightly with temperature, indicative of an antiferromagnetic or compensated ferrimagnetic magnetic phase. Rietveld analysis of synchrotron XRD on bulk CrVTiAl samples showed a disordered polycrystalline CrVTiAl phase with minimal secondary phases. The resistivity showed characteristic metallic behavior at low temperature, before reaching a maximum near 400 K and decreasing at high temperature, as a result of thermally activated carriers from a component of ordered CrVTiAl contributing a semiconducting character. In films, the resistivity indicated CrVTiAl to be a spin-gapless semiconductor, which further calculations showed to be a result of the atomic ordering. High-field magnetotransport measurements showed two-carrier transport, indicative of the two spin-channels in the material. Magnetometry showed CrVTiAl to be a compensated ferrimagnet, with T\textsubscript{C} \textgreater~400\degree C. While the spin-gapless behavior is less conducive to the desired spin-filtering properties, the high T\textsubscript{C} and low magnetic moment could potentially make CrVTiAl useful for voltage-tunable spintronics at room temperature. On the spin-filtering front, the results of CrVTiAl are promising, suggesting that further research, particularly computational studies, should be performed on similar materials to investigate whether substituting in elements from the same elemental groups are more likely to form as spin-filters.