Computational Studies of Dynamical Phenomena in Nanoscale Ferromagnets

Mississippi State University, Mississippi State MS

Investigators

Abstract

This is an award to Mississippi State University with a subaward to Florida State University. It is a renewal of an ongoing research program on state of the art computational studies of nanomagnetism. As such, it captures the spirit of recent NSF initiatives on ITR and NSE, in that its impact will be felt on new algorithms for complex interacting systems, understanding magnetism on the nanoscale, and the design of new ultrahigh-density magnetic strorage systems. One technological motivation of the research is the ongoing effort to increase the information storage density of magnetic recording media. It is likely that ultrahigh-density magnetic recording media that can store one bit of information on one single-domain, nanoscale magnetic particle may become available in the next few years. To achieve this goal with acceptable data integrity and read/write speeds at room temperature, it is necessary to improve the scientific understanding of magnetization-reversal dynamics in magnetic nanoparticles, molecules, and ultrathin films, at nonzero temperatures. This will require large-scale numerical simulations of realistic models of technologically important magnetic materials. This research project will further develop novel simulation algorithms for hysteresis and thermally driven magnetization reversal in models of nanoscale magnets. The materials objective of the project is to improve understanding of dynamical phenomena in real nanoscale ferromagnetic materials at nonzero temperature over a large range of time scales. Previous work will be extended by conducting simulations of a wider range of real materials through increased emphasis on quantum mechanical systems and less strongly anisotropic materials. These generalizations will include continuous spin models with finite spin anisotropy, systems with defects and quenched disorder, three-dimensional models and quantum spin models of magnetic molecules. In order to study these models, novel algorithms capable of covering a wide range of time scales will be adapted and developed, in particular the Projective Dynamics and Monte Carlo with Absorbing Markov Chains algorithms previously developed on this project. The studies of continuous spin models will also use Langevin Micromagnetics methods for finite-temperature simulations developed during the previous grant period. Particular emphasis will be given to developing methods that enable these simulations to cover a wide range of time scales. It is also proposed to use quantum density matrices to predict experimental EPR line widths for magnetic molecules and to provide ab initio transition probabilities for kinetic Monte Carlo and Langevin simulations. The suitability of these algorithms for various parallelization paradigms will be studied and they will be implemented on scalable parallel computers. The scaling properties of parallel simulation algorithms will be studied using mappings to non-equilibrium interface-growth problems that were discovered on this project. These algorithmic and parallelization methods constitute the computational objectives of the research. %%% This is an award to Mississippi State University with a subaward to Florida State University. It is a renewal of an ongoing research program on state of the art computational studies of nanomagnetism. As such, it captures the spirit of recent NSF initiatives on ITR and NSE, in that its impact will be felt on new algorithms for complex interacting systems, understanding magnetism on the nanoscale, and the design of new ultrahigh-density magnetic strorage systems. ***