Micromagnetic simulations of magnetic nanoparticles for cancer therapy and of Recording heads for two dimensional magnetic recording.

University of Minnesota M.S. thesis. December 2011. Major: Electrical Engineering. Advisor: Randall H. Victora. 1 computer file (PDF); vi, 78 pages.In this work, micromagnetic simulations were used to model magnetic phenomena such as magnetic nanoparticle behavior and magnetic recording processes. These simulations were used to optimize the heating properties of the nanoparticles as well as to design and simulate a new shingled recording head meant for recording densities in the multiple Tbit/inch2 regime. Magnetic hyperthermia for treatment of tumors would benefit greatly from increased heating of the superparamagnetic particles, coupled with reduced incident electromagnetic wave power. Previous micromagnetic simulations based on the Landau–Lifshitz–Gilbert equation with thermal fluctuations showed that the use of incident square waves greatly increases the normalized heat. Experimentally generating a square waveform may produce a trapezoidal waveform due to an inherent rise and fall time. It is found that the normalized heating power given by this wave shows a 30% improvement over the sinusoidal waveform for an anisotropy distribution of 20%. A static magnetic field applied perpendicular to the incident oscillating magnetic waveform was shown analytically to be able to improve the normalized heat. Previous simulations showed that the static field was indeed able to increase the effectiveness of a sinusoidal waveform. This work, using a square waveform, showed that the addition of the static field gives little added benefit to the square waveform’s effectiveness. A shingled recording head design is proposed that makes use of a saturation magnetization gradient to focus flux into the corner of the writing pole. This design yields field gradients of 800 Oe/nm and can write more than 2.5 Tbits/Inch2 of user data onto a 10 nm thick exchange couple composite (ECC) recording layer with 6 nm Voronoi grains. Maximum user densities were predicted to be over 4.66 Tbits/Inch2 with 8 nm grains and 12 Tbits/Inch2 with 5 nm grains using a perfect read and write scheme at a 1:1 bit aspect ratio. Previous work explored the read back process by using finite difference method (FDM) simulations to simulate magnetoresistive head technologies followed by the use of reciprocity theory to recover information previously written to the recording layer. A micromagnetic study was completed to prove that the FDM approach was indeed suitable for these small scale simulations. The study showed that for a conventional recording, the normalized read back waveforms using the FDM approach and using the micromagnetic simulation approach were the same