Dual Antiferromagnetic Coupling at ${\mathrm{La}}_{0.67}{\mathrm{Sr}}_{0.33}{\mathrm{MnO}}_3/{\mathrm{SrRuO}}_3$ Interfaces

We have studied the magnetic hysteresis cycle of ${\mathrm{La}}_{0.67}{\mathrm{Sr}}_{0.33}{\mathrm{MnO}}_3/{\mathrm{SrRuO}}_3$ antiferromagnetically coupled bilayers, by magnetometry and polarized neutron reflectometry. A positive exchange bias as well as an unusual asymmetry are observed on the magnetic reversal process of the ${\mathrm{La}}_{0.67}{\mathrm{Sr}}_{0.33}{\mathrm{MnO}}_3$ layer. Through an extended Stoner-Wohlfarth model comprising the magnetic anisotropy of both layers, we give experimental evidence that this asymmetry originates from two different but well-defined antiferromagnetic coupling strengths at the interface between the two magnetic oxides. The possible origin of this dual coupling is discussed in view of our experimental results.

Figure 1

(a) Schematic of LSMO and SRO pseudocube unit cell grown on a cubic STO (001) unit cell. The SRO primary orthorhombic unit cell based on the $[100{]}_{\mathrm{o},\mathrm{SRO}}$, $[010{]}_{\mathrm{o},\mathrm{SRO}}$, and $[001{]}_{\mathrm{o},\mathrm{SRO}}$ axis is also represented by the $a$, $b$, and $c$ axis, respectively. The orientation of the SRO orthorhombic unit cell on STO(001) shown is one of the possible orientations. The magnetic easy axes of LSMO and SRO are shown by large double arrows, as well as the direction of the applied field $B$. Two other arrows around the field represent the LSMO and SRO possible magnetization with their angle $\theta$ and $\phi$ with the field direction. (b) Atomic force microscope image ($5\times 5\mu {\mathrm{m}}^2$) of the $\mathrm{LSMO}/\mathrm{SRO}/\mathrm{STO}$ bilayer.

Figure 2

Major magnetic hysteresis loop of a bilayer LSMO/SRO. Experimental points (dots, red online) are measured by SQUID magnetometry at 80 K, after a field cooled at 0.5 T. The arrows show the cycling of the field from $+1$ to $-1\mathrm{T}$ and back to $+1\mathrm{T}$. Two simulations are presented: (i) one model is using two $J$ exchange constants [gray line (green online)], and (ii) the second model considers a mean exchange constant at the interface (black line).

Figure 3

Minor magnetic hysteresis cycle of a bilayer LSMO/SRO, corresponding to the LSMO switching, while the SRO magnetization is blocked at 45° from the positive direction of the applied field. The experimental points (dots, red online) are measured by vibrating sample magnetometry at 80 K, after a field cooled at 0.5 T. The arrows show the cycling of the field from $+0.15$ to $-0.15\mathrm{T}$ and back to $+0.15\mathrm{T}$. The simulation using two coupling strengths is presented as a line (black online). The moment calculated from the PNR data are shown as large plain dots (blue online) for various fields. The inset at the top is a superposition of the minor and major loops. The inset at the bottom is a zoom of the central part of the hysteresis.

Figure 4

Typical polarized neutron reflectivity measurement. These curves were taken at 22 mT (point 2 in Fig. 3) and at 80 K. Dots are the measured points, and lines are the fit performed on those data. Squares (respectively, circles) show neutrons polarized and measured up (respectively, down), triangles correspond to the spin flip measurement, and neutrons are polarized down and are measured up after reflection.

Figure 5

LSMO magnetization angles simulated (line) and estimated from PNR (dots) in both domains: high exchange coupling $J=-320\mu \mathrm{J}/{\mathrm{m}}^2$ (a) and low exchange $J=-13\mu \mathrm{J}/{\mathrm{m}}^2$ (b), at the same time, as a function of field. The SRO layer is blocked around 45° in all the cases. The numbers are the PNR measured magnetic states.