Theory of stripe-domain nucleation in garnet films

Abstract

Micromagnetics predicts that domain nucleation in an in‐plane field in mechanically rigid uniaxial films with easy axis normal to the faces should take the form of stripes parallel to the field. This prediction does not agree with observation in some real films. It has been shown previously that the magnetoelastic interaction in deformable films may cause the stripes to be perpendicular to the field when the films are thick enough. Quantitative assessment of this effect is difficult because the magnitudes of the magnetoelastic constants are not accurately known, and because high‐quality epitaxial films are at least partly constrained from deforming. In this paper we discuss another interaction which can cause the stripes to be perpendicular to the field in a [111]‐oriented cubic film, such as bubble‐device garnet. The interaction energy is a dipolar energy brought into play by the effectively trigonal anisotropy. It vanishes when H is along [112̄] and is maximum when H is along [1̄10]. Quantitative evaluation shows that the interaction can account for a large part but not all of the behavior observed in recent experiments.