The role of epitaxial strain for the in-plane magnetic anisotropies is studied for epitaxial Fe(001) thin films (0.9–60 nm) deposited by molecular-beam epitaxy at room temperature on ${\text{Al}}_{0.48}{\text{In}}_{0.52}\text{As}\left(001\right)$ layers. The correlation between structural and magnetic properties has been investigated using ex situ x-ray diffraction (XRD) experiments and magneto-optical Kerr effect (MOKE) measurements. Fe grows in a body-centered-cubic (bcc) structure with epitaxial relationship $\text{Fe}\left(001\right)⟨100⟩\parallel {\text{Al}}_{0.48}{\text{In}}_{0.52}\text{As}\left(001\right)⟨100⟩$. The mismatch in lattice parameter between pure Fe and ${\text{Al}}_{0.48}{\text{In}}_{0.52}\text{As}$ is of 2.3% and the iron films are under an in-plane tensile biaxial strain. The films remain pseudomorphous up to 4.4 nm and then progressively relax when increasing the Fe coverage. All the Fe layers are ferromagnetic at room temperature and show an in-plane magnetization with a fourfold anisotropy (with $⟨100⟩$ directions as easy axes) superimposed to a twofold anisotropy (with [110] direction as easy axis) which probably originates from anisotropic bonding at the $\text{Fe}/{\text{Al}}_{0.48}{\text{In}}_{0.52}\text{As}$ interface. The fourth-order anisotropy constant of the magnetic films shows nonmonotonous changes with Fe coverage. We show that this unusual evolution can be reproduced within the Néel’s pair model in which we have considered high-order Néel parameters and included the strain and interface alloying effects. From our analysis we find that this behavior is due mainly to the in-plane strain effect in the films through the bulk magnetoelastic coupling and a fourfold surface anisotropy term whose strength decreases with the film thickness. This surface magnetic anisotropy induced by the broken symmetry at the interfaces favors the $⟨110⟩$ directions as easy axis while the bulk magnetoelastic anisotropy induced by an in-plane tensile biaxial strain favors the $⟨100⟩$ directions as easy axis. We find that the surface magnetoelastic anisotropy energy contribution to the in-plane magnetic anisotropy energy is much smaller than the other contributions.

• Received 4 July 2008

DOI:https://doi.org/10.1103/PhysRevB.78.134401