Abstract
Magnetic patterns with different shape and structure were fabricated by e-beam lithography and lift-off techniques to investigate their special properties. It includes four parts of experiments: 1. Switching field and the contact angle dependence in giant magnetoresistance (GMR) spin-valve wires with five different shaped reservoirs. 2. The depth of the trench and switching field dependence in the permalloy (Py) wires. 3. Numbers of steady-states in half-ring chains spin-valve with various linewidth. 4. The thickness dependence of hard layer in the pseudo spin-valve elements. For the spin-valve wires with different shaped reservoirs, the contact angle of the wires and reservoirs provides different injection of magnetic domain wall. The switching field increases with the decreasing of contact angle. The magnetization reversal processes obtained experimentally were consistent and illustrated by using the micromagnetic simulation program, OOMMF. The dependence between the depth of trench and the switching field in Py wires was observed by using a real-time magnetic force microscopy (MFM) and electrical measurement. The coercive and switching fields were increased with increasing the depth of trenches. The MFM images also clearly illustrated the magnetization reversal and magnetic domain structure in the different trench samples. In the half-ring chains spin-valves with different linewidth, domain walls are significantly created at the corners between the two half rings. In the MR measurements, the domain walls at the different linewidth corners provide the difference of the switching field due to different pinning force. Numbers of steady states were obtained and attributed to the different switching fields. The experiment of current induce magnetization reversal was also observed in the samples. The critical current densities at zero magnetic field were 3×107 A/cm2 on Py and 2×108A/cm2 on Co samples, respectively. Furthermore, in the current perpendicular to plane (CPP) GMR spin valves, the thickness of the soft layer was fixed and the hard layer was varied. The MR ratio versus the thickness of hard layer shows nonmonotonic dependence. The maximum valve of MR ratio was acquired when the thickness of hard layer was around 21 nm. From the current induce magnetization reversal study, the lowest critical current was acquired around the same thickness, 21 nm, of the hard layer, which was close to the spin diffusion length. Finally, we obtained experimentally the result of maximum efficiency (MR ratio) and minimum consumption (critical current) in a spintronic device. The product of MR ratio and critical current density is a constant, ~ 4.5±0.5×107 A/cm2.