Time Resolved Kerr Microscopy of Materials and Devices for Magnetic Data Storage Applications
Thesis or dissertation
University of Exeter
I have required 18 months embargo when I submitted this thesis for review.
Reason for embargo
We are currently working on writing a paper which is based on one of the chapters within this thesis.
Time resolved scanning Kerr microscopy (TRSKM) has been used to study a number of different magnetic systems. Firstly, partially built hard disk writer structures, with a multilayered yoke formed from 4 repeats of a NiFe(~1 nm)/CoFe(50 nm) bilayer, and with three coil windings underneath, were studied by TRSKM with unipolar driving pulses. Dynamic images of the in-plane magnetization suggest an underlying closure domain equilibrium state. This state is found to be modified by application of a bias magnetic field and also during pulse cycling, leading to different magnetization rotation and relaxation behaviour within the tip region. Studies of a further three yokes with the same stack structure, but with only one coil winding at different positions beneath the yoke, yielded dynamic images of “flux beaming” in a channel parallel to the driving field. The magnetic contrast was strongest when the active coil was located near the centre of the yoke, while relaxation after removal of the excitation was most complete when the active coil was located near the confluence region. These results confirm the need for a multi-turn coil to ensure effective flux propagation along the entire length of the yoke. Furthermore, a structure with a NiFe/CoFe/Ru/NiFe/CoFe synthetic antiferromagnetic (SAF) yoke was studied as a bipolar current pulse with 1MHz repetition rate was delivered to the coil. The component of magnetization parallel to the symmetry axis of the yoke was compared at the pole and above a coil winding in the centre of the yoke. The two responses are in phase as the pulse rises, but the pole piece lags the yoke as the pulse falls. The Kerr signal is smaller within the yoke than within the confluence region during pulse cycling. This suggests funneling of flux into the confluence region. Dynamic images acquired at different time delays showed that the relaxation is faster in the centre of the yoke than in the confluence region, perhaps due to the different magnetic anisotropy in these regions. Although the SAF yoke is designed to support a single domain to aid flux conduction, no obvious flux beaming was observed, suggesting the presence of a more complicated domain structure. The SAF yoke writer hence provides relatively poor flux conduction but good control of rise time compared to single layer and multi-layered yokes studied previously. Secondly, vortex dynamics within arrays of square ferromagnetic nano-elements have been studied using TRSKM with coherent microwave excitation. It is shown that TRSKM can be used to detect vortex gyration in square nanomagnets with a lateral size (250nm) that is smaller than the diameter (300nm) of the focused laser beam. In an array with large element separation and negligible dipolar interaction, TRSKM images acquired at a fixed point in the microwave cycle reveal differences in the phase of the dynamic response of individual nanomagnets. While some variation in phase can be attributed to dispersion in the size and shape of elements, the circulation and polarization of the vortex are also shown to influence the phase. In an array with element separation smaller than the optical spot size, strong magneto optical response was observed within small clusters of elements. Micromagnetic simulations performed for 2 x 2 arrays of elements show that a certain combination of circulation and polarization values is required to generate the observed magneto-optical contrast. Thirdly, polar TRSKM has been used to directly observe magnetostatically coupled transverse domain walls (TDWs) in a pair of closely spaced, curved nanowires (NWs). Kerr images of the precessional response revealed a minimum in the Kerr signal due to the TDW in the region of closest NW separation. When the TDWs were ejected from the NW pair, the minimum in the Kerr signal was no longer observed. By imaging this transition, the static decoupling field was estimated to lie between 38 and 48 Oe, in good agreement with a simple micromagnetic model. This work provides a novel technique by which DC and microwave assisted decoupling fields of TDWs may be explored in NW pairs of different width, separation, and curvature. Fourth, time resolved magneto-optical Kerr effect and phase modulated X-ray ferromagnetic resonance measurements have been performed on a CoO/Py bilayer for different temperatures, RF frequency, and CoO thickness. Kerr hysteresis loops did not show any evidence of exchange bias for temperatures between 200K and 330K for any thickness of CoO, but the coercivity was found to increase with increasing CoO thickness and decreasing temperature. Magneto-optical FMR and XFMR data showed some asymmetry with respect to the sign of the bias field, but the amplitude of the signals decreased rapidly with decreasing temperature. The results are consistent with the appearance of frustrated antiferromagnetic order within the CoO during field cooling.
PhD in Physics