No abstract

No abstract

The magnetization reversal process of Co nanowires was investigated using a moving mesh technique. The nucleation and expansion of reversed domains is calculated by solving the Gilbert equation of motion for different damping constants. The adaptive finite element method reduces the total CPU time by more than a factor of 4 as compared to a uniform mesh. Two different domain wall types are observed. For a wire diameter of d=10 nm transverse walls occur and gyromagnetic precession limits the domain wall velocity. The domain wall velocity increases from 50 to 520 m/s as the Gilbert damping constant increases from α=0.05 to α=1 at an applied field of 500 kA/m. For a diameter greater than 20 nm vortex walls are formed. The vortex mobility increases with decreasing damping constant. Thus velocities up to 2000 m/s are reached for a wire diameter of 40 nm, α=0.05, and an applied field of 250 kA/m.

We study the low-temperature isothermal magnetic hysteresis of cubical and spherical nanoparticles with ferromagnetic (FM) core -antiferromagnetic (AF) shell morphology, in order to elucidate the sensitivity of the exchange bias effect to the shape of the particles and the structural imperfections at the core-shell interface. We model the magnetic structure using a classical Heisenberg Hamiltonian with uniaxial anisotropy and simulate the hysteresis loop using the Metropolis Monte Carlo algorithm. For nanoparticles with geometrically sharp interfaces, we find that cubes exhibit higher coercivity and lower exchange bias field than spheres of the same size. With increasing interface roughness, the shape-dependence of the characteristic fields gradually decays and eventually, the distinction between cubical and spherical particles is lost for moderately rough interfaces. The sensitivity of the exchange bias field to the microstructural details of the interface is quantified by a scaling factor (b) relating the bias field to the net moment of the AF shell (H eb = bMAF + Ho). Cubical particles exhibit lower sensitivity to the dispersed values of the net interfacial moment.

No abstract