Transfer of Magnetic Order and Anisotropy through Epitaxial Integration of $3d$ and $4f$ Spin Systems

Resonant x-ray scattering at the Dy $M_5$ and Ni $L_3$ absorption edges was used to probe the temperature and magnetic field dependence of magnetic order in epitaxial ${\mathrm{LaNiO}}_3-\mathrm{DySc}{\mathrm{O}}_3$ superlattices. For superlattices with 2 unit cell thick ${\mathrm{LaNiO}}_3$ layers, a commensurate spiral state develops in the Ni spin system below 100 K. Upon cooling below $T_{\mathrm{ind}}=18\mathrm{K}$, Dy-Ni exchange interactions across the ${\mathrm{LaNiO}}_3-\mathrm{DySc}{\mathrm{O}}_3$ interfaces induce collinear magnetic order of interfacial Dy moments as well as a reorientation of the Ni spins to a direction dictated by the strong magnetocrystalline anisotropy of Dy. This transition is reversible by an external magnetic field of 3 T. Tailored exchange interactions between rare-earth and transition-metal ions thus open up new perspectives for the manipulation of spin structures in metal-oxide heterostructures and devices.

Figure 1

(a) Angular dependence within the $a\text{−}b$ plane of the macroscopic saturation magnetization in bulk ${\mathrm{DyScO}}_3$. The black line is the result of the Ising model which has been fitted to the data to determine the orientation of the two Ising axes (red lines). (b) Pictorial representation of the ${\mathrm{DyScO}}_3$ structure [18] with the two Ising axes (only Dy and its nearest coordinating oxygens are shown). One of the Ising axes agrees closely with the axis of the collinear Dy antiferromagnetism determined by RXS in the superlattices [solid green line in (a)]. The dashed blue line is in the plane of the (110) oriented superlattice interface.

Figure 2

(a) Rocking curves around the $\mathbf{q}=(\frac{1}{2}0\frac{1}{2})$ magnetic Bragg peak for various superlattices, with intensity normalized to the number of ${\mathrm{DyScO}}_3$ layers in the superlattice. Measurements were performed at $T=4\mathrm{K}$ and with $\pi$-polarized photons tuned to the Dy $M_5$ edge. The inset shows the photon energy dependence of the $\mathbf{q}=(\frac{1}{2}0\frac{1}{2})$ magnetic Bragg peak intensity across the Dy $M_5$ edge (red circles) compared with the energy-dependent x-ray absorption measured in the total electron yield mode (black line). (b) Measured (points) and calculated (lines) azimuthal-angle ($\psi$) dependence of $I_{\pi }/I_{\sigma }$ at the $\mathbf{q}=(\frac{1}{2}0\overline{\frac{1}{2}})$ magnetic Bragg peak, for photon energies tuned to the Ni $L_3$ and Dy $M_5$ absorption edges [23]. Incident or scattered photons are shadowed by the sample (prohibiting the experiment) in the angular regions indicated with a gray bar. The data point at $\psi =90°$ was measured in an alternate geometry [11]. (c) Temperature dependence of $I_{\pi }/I_{\sigma }$ at the Ni $L_3$ edge for $\psi =30°$ in magnetic fields of $H=0$ (solid squares) and $H=5\mathrm{T}$ (open circles) oriented along the $[10\overline{1}]$ direction. The integrated magnetic scattering intensity at the Dy $M_5$ edge for $\pi$-polarized light is shown in red along with a fit $\propto 1/T$. The inset shows the Ni $L_3$ ratio $I_{\pi }/I_{\sigma }$ at $T=4.4\mathrm{K}$ as a function of $H$.

Figure 3

A depiction of the magnetic structure at the ${\mathrm{LaNiO}}_3\text{−}\mathrm{DySc}{\mathrm{O}}_3$ interface [25] at temperatures above $T_{\mathrm{ind}}$ (center) and at 4 K (right). Since the resonant scattering experiment is not sensitive to a relative phase shift between the Ni and Dy magnetic structures, it is assumed that the Ni-Dy exchange interaction is antiferromagnetic. The lack of an arrow at a Dy site indicates the absence of detected magnetic order. To simplify the sketch, orthorhombic lattice distortions have been omitted and the lengths of the Dy moments are not to scale.