Answers are grounded in this paper and cite their sources.
In this work the influence of shape anisotropy on the magnetic hysteresis and zero-field magnetic domain state of submicron-size particles of Fe was investigated. Arrays of particles having circular (diameter dc=550 or 200 nm), rectangular (900 nm by 300 nm), or elliptical (450 nm by 150 nm) shape were prepared by electron lithography and ion-beam milling of epitaxial Fe (001) films of thicknesses t=50, 30, 15, and 10 nm. The samples were characterized by magnetization measurements and magnetic-force microscopy (MFM). All films have cubic anisotropy, for $t=50,30,$ and 15 nm with the same anisotropy constant as bulk Fe: K₁=(4.3±0.1)×10⁴Jm^-3. For t=10nm the effective anisotropy is increased, K₁=5.7×10⁴Jm^-3, due to surface effects. The effects of the interplay between the magnotocrystalline and shape anisotropies are observed as the lateral extension of the films is decreased. The circular particles with dc=550nm have closed magnetic domain structures with a fourfold symmetry, compatible with the magnetocrystalline anisotropy, for all thicknesses. In the rectangular particles a gradual change is observed as the thickness decreases. For t=50nm a diamond structure comprising three closed-domain substructures is formed in the demagnetized state. The rectangles with $t=30$ and 15 nm are multidomains with the number of closed substructures decreasing to 2 and 2 or 1, respectively. The thickness dependence of the domain structure and an accompanying change of character in the MFM contrast are explained by an increasing amount of N\'eel component in the domain walls with a width that increases with decreasing film thickness. The rectangles with t=10nm are quasisingle domains, forming a flower state. The small circular particles (dc=200nm) and the elliptical ones, both with t=10nm, are considered to be stable single domains in zero field. Judging from the hysteresis curves, magnetization reversal does not occur by coherent rotation in any of the particles.
No takes yet. Share an insight, caveat, or question.
Hanson et al. (2002) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: