Transverse magnetization in Cu/Ni/Cu epitaxial nanorings

Abstract

The micromagnetic structure in epitaxial (001)-oriented Cu/Ni(14 nm)/Cu rings fabricated by electron beam and focused ion beam lithographies with external diameter of 3 μm and linewidths between 100 and 500 nm is presented. We found that a state with radial orientation of the magnetization prevails at remanence. The evaluation of the magnetoelastic, magnetocrystalline and magnetostatic energies shows that a value as low as 1.5 •10 −3 for the anisotropic relaxation of the in-plane strain components is enough to induce an effective radial easy magnetization direction

Strain-induced magnetization reorientation in epitaxial Cu/Ni/Cu rings

“…The domain structure in this film with a weak out-of-plane magnetic anisotropy and effective in-plane magnetic anisotropy consist of stripes or maze structures with M pointing up and down with respect to the film surface. 9 Strain relaxation occurs normal to the ring edges and favors the in-plane orientation of M along the radial direction of the ring, as we observed previously in a narrow ring 9 and stripes. 7 For wide rings this radially magnetized structure coexists with perpendicular stripe domains in the less-relaxed regions of the ring away from the edges, whereas for ring widths below ≈400 nm a fully radial domain configuration is present.…”

“…The stripe width was about 200 nm, a value similar to that obtained for the unpatterned thin film. 6,9 However, for the region at the edge of the ring, the pattern consisted of alternating dark and bright contrast sectors that we interpret as the stray field due to inplane domains. These configurations are sketched in Fig.…”

“…In a previous work 9 we showed that a difference between radial and tangential strain as low as 1.5 × 10 −3 is capable, through the ME energy contribution, to overcome the shape anisotropy and favor the transverse (i.e., radial) orientation of M in a narrow Ni ring. The ME contribution is given by the expression: 9 e mel = (B 1 cos 2 2φ + B 2 sin 2 2φ) ,…”

“…A bilayer film of Cr/Al was deposited by electron-beam evaporation on the resist and lifted off to leave ring-shaped Cr/Al hardmasks, and ion beam etching was used to transfer the pattern into the Cu/Ni/Cu layer in a procedure reported elsewhere. 9 An energy-dispersive x-ray spectroscopy map (not shown) performed on a milled ring showed no trace of nickel outside the masked area and signal from the Cr/Al layers on the ring region, indicating that the mask was not completely eroded and protected the Cu/Ni/Cu structures during the ion etching process. This procedure is analogous to the process used in previous work to obtain arrays of epitaxial nanowires.…”

“…9 For the case of a ring divided into N domains with identical size, it can be demonstrated, see the appendix, that e ms is given by volume and surface terms, respectively, …”