Electron transport in ferromagnetic nanostructures
Doctor of Philosophy
As the size of a physical system decreases toward the nanoscale, quantum mechanical effects such as the discretization of energy levels and the interactions of the electronic spins become readily observable. To understand what happens within submicrometer scale samples is one of the goals of modern condensed matter physics. Electron transport phenomena drew a lot of attention over the past two decades or so, not only because quantum corrections to the classical transport theory, but also they allow us to probe deeply into the microscopic nature of the system put to test. Although a significant amount of research was done in the past and thus extended our understanding in this field, most of these works were concentrated on simpler examples. Electron transport in strongly correlated systems is still a field that needs to be explored more thoroughly. In fact, experimental works that have been done so far to characterize coherence physics in correlated systems such as ferromagnetic metals are far from conclusive. One reason ferromagnetic samples draw such attention is that there exist correlations that lead to excitations (e.g. spin waves, domain wall motions) not present in normal metals, and these new environmental degrees of freedom can have profound effects on decoherence processes. In this thesis, three different types of magnetic samples were examined: a band ferromagnetism based metallic ferromagnet, permalloy, a III-V diluted ferromagnetic semiconductor with ferromagnetism from a hole-mediated exchange interaction, and magnetite nanocrystals and films. The first observation of time-dependent universal conductance fluctuations (TD-UCF) in permalloy is presented and our observations lead to three major conclusions. First, the cooperon contribution to the conductance is suppressed in this material. This is consistent with some theoretical expectations, and implies that weak localization will be suppressed as well. Second, we see evidence that domain wall motion leads to enhanced conductance fluctuations, demonstrating experimentally that domain walls can act as coherent scatterers of electrons. Third, the temperature dependence of the fluctuations is surprisingly strong, suggesting that the dominant decoherence mechanism in these wires is different than that in similar normal metal nanostructures. The first observation of TD-UCF in diluted magnetic semiconductors (DMS) is also presented. In contrast to analogous measurements on permalloy samples, we find a surprising suppression of TD-UCF noise in this material at low temperatures, independent of field orientation. We believe this implies that the suppression is not due to an orbital effect, and therefore some of the fluctuations originate with time-varying magnetic disorder. The temperature dependence of the TD-UCF implies either an unusual fluctuator spectrum or a nonstandard dephasing mechanism. Measurements of UCF as a function of magnetic field allow an order of magnitude estimate of the coherence length at 2 K of approximately 50 nm in this material. The last samples examined were magnetite nanocrystals and films. Magnetite has been used in technologies for millennia, from compasses to magnetoelectronic devices, although its electronic structure has remained controversial for seven decades, with a low temperature insulator and a high temperature "bad metal" separated by the Verwey transition at 120 K. A new electrically driven insulator-metal transition below the Verwey temperature in both magnetite films and nanocrystals was observed. The possibility that this was a thermal effect was tested through various methods, and we have shown that the transition is in fact truly electrically driven. This electrically driven transition also showed a great deal of rigidity against external magnetic field and high gate voltages.
Condensed matter physics