| In this thesis, we present studies to produce new information about three topics: current perpendicular to the plane magnetoresistance (CPP-MR), spin transfer torque (STT), and antiferromagnetic spintronics.;Large values of CPP-MR interface parameters---specific interface resistance (Area times resistance), 2AR*, and scattering asymmetry, gamma---are desirable for the use of CPP-MR in devices. Stimulated by a nanopillar study by the Cornell Group, we first discovered that Py/Al had an unusually large 2AR*, but a small gamma. In the hope of finding metal pairs with large values of both the interface parameters, the Py/Al studies led us to study the following interfaces: (a) F/Al with F: Py (= Ni84Fe16). Co, Fe, Co91Fe9, and (b) F/N: Py/Pd, Fe/V, Fe/Nb and Co/Pt. None of the metal pairs looks better for CPP-MR devices.;The Cornell group also found that bracketing Al with thin Cu in Py/Al/Py nanopillars, gave an MR similar to Py/Al/Py rather than to Py/Cu/Py. To try to understand this result, we studied the effect of Cu/Al/Cu spacers on ADeltaR = AR(AP) - AR(P) of Py exchange biased spin valves (EBSVs). Here AR(AP) and AR(P) are the specific resistances in the anti-parallel (AP) and parallel (P) configurations of the F layers.;Intriguingly, fixing the Al thickness tAl = 10 nm and varying tCu has no effect on ADeltaR, but fixing tCu = 10 nm and varying t Al significantly affected ADeltaR. These unusual behaviors are probably due to strong Al and Cu intermixing, with probable formation of some fraction of ordered alloys.;Recent calculations predicted that 2AR of Al/Ag interfaces would vary substantially with orientation and with alloying. The latter is a special potential problem, because Al and Ag interdiffuse at room temperature. To compare with the calculations, we determined 2AR of sputtered Al/Ag interfaces with (111) orientation. Our estimate agrees with calculations that assume 4 monolayers of interfacial disorder, consistent with modest intermixing.;To aid in understanding the physics underlying spin transfer torque in F/N/F trilayers, we devised a simple test, where we change the transport properties of only the N-spacer from ballistic to diffusive. To try to minimize effects of systematic errors, we compare ratios of results for diffusive and ballistic transport with calculations. The ratios disagree with predictions of those models that were expected to be most, reliable.;Recently, it was predicted that spintronics effects: magnetoresistance, and spin transfer torque induced magnetization switching, should be observable in systems where ferromagnets are replaced by antiferromagnets. We have done the first experiments to search for those predicted antiferromagnetic spintronics effects. We find effects of large current densities on exchange bias, but no clear evidence for antiferromagnetic giant magnetoresistance. |
