Micromagnetic study of heat assisted magnetic recording using renormalized media cells

Abstract

Perpendicular magnetic recording is currently very near to its physical limit, making it difficult for researchers to keep the pace of the growth of areal density of hard disk drives. Heat-assisted magnetic recording (HAMR) is considered to be the next generation technology for magnetic recording beyond 1 Tb/in<super>2</super>. Complete understanding of HAMR processes is necessary to optimize the design parameters. In this thesis, current state-of-the-art modelling methods are developed aiming at HAMR recording. First a simple torque-based method for calculating the transient behavior of temperature-dependent magnetic anisotropy is introduced. By using this method several physical quantities at finite temperature including effective anisotropy, anisotropy field, and their fluctuations are obtained. A composite grain that includes a high Curie temperature soft layer can reduce the anisotropy fluctuations. Then a new scheme for the simulation of HAMR that systematically includes fluctuating material properties above a predefined length scale, while retaining magnetostatic interactions is introduced. Renormalized media parameters, M_s, K_u, A_ex and &alpha, suitable for useful length scales, are found numerically. These renormalized parameters are then used to model the Voronoi-cell-composed medium in the HAMR simulation. Transition jitters are obtained under various conditions. The results show that moderate maximum temperature of the heat spot, intergranular exchange coupling, media thickness of at least 10 nm, nonzero canting angle of the head field, relatively low head velocity, and large head-field strength are helpful for a successful recording. This scheme of HAMR simulation is used to find the dependencies of recording performance on the grain size and damping. The simulated results are used to compare with an experimental demonstration. Finally, composite FeRh/FePt for HAMR media is investigated with micromagnetic simulation. It is found to potentially lower recording temperature, while retaining high anisotropy field gradient. The transition width is predicted to depend on the media cooling rate. The thickness of the FeRh layer and the applied field can significantly affect the switching time of FePt layer, and therefore alter recording performance. Applied field magnitudes and angles are identified that allow successful switching within 100 pS. It is shown that by using up to 15 nm of FeRh with 6 nm of FePt, the jitter for 5.6 nm grains can be nearly equal to the grain-size limited value, for head velocities as high as 20 m/s.