Synthetic Antiferromagnets with Steplike Hysteresis Loops and High-T_C Based on All-Perovskite ${\mathrm{La}}_{0.7}{\mathrm{Sr}}_{0.3}\mathrm{Mn}{\mathrm{O}}_3$ Superlattices

Synthetic antiferromagnets with high Curie temperature (approximately 290 K) are realized among all-perovskite ${\mathrm{La}}_{0.7}{\mathrm{Sr}}_{0.3}\mathrm{Mn}{\mathrm{O}}_3$ $(\mathrm{LSMO})/\mathrm{Ca}{\mathrm{Ru}}_{0.5}{\mathrm{Ti}}_{0.5}{\mathrm{O}}_3$ (CRTO) superlattices, which show antiferromagnetic (AF) interlayer exchange coupling (IEC) rather than traditional interfacial exchange coupling that is usually observed under the Curie temperature of the spacer. The system shows layer-resolved magnetic switching, resulting in sharp steplike hysteresis loops with magnetization plateaus depending on the repetition number of the stacking bilayers. This bilinear IEC can be easily changed to biquadratic IEC by change of orthorhombic $\mathrm{Nd}\mathrm{Ga}{\mathrm{O}}_3(001)$ substrate to cubic $(\mathrm{La}\mathrm{Al}{\mathrm{O}}_3{)}_{0.3}(\mathrm{Sr}{\mathrm{Al}}_{0.5}{\mathrm{Ta}}_{0.5}{\mathrm{O}}_3{)}_{0.7}(001)$ substrate. The strength of AF IEC increases with decrease of temperature, and the modeling results suggest that both magnetic LSMO and the CRTO space layer play a role in this behavior. Remarkably, the LSMO/CRTO system shows controllability of AF-IEC behavior under a moderate magnetic field of hundreds of oersteds. These observations may have potential applications in future oxide spintronic devices.

Figure 1

Characterization of LSMO/CRTO superlattice, [2.8/1.2]₁₀, grown on NGO(001). (a) High-resolution X-ray θ-2θ linear scan. (b) Off-specular reciprocal-space mapping on the (116) main reflection, where the main reflection (116) (denoted as “0”) and the satellites (denoted as “±1”) are narrow in Q_[110] direction and sharply defined even with the thickness fringes. Note that Q_[110] and Q_[001] are coordinates in the reciprocal space and for the present system they are along the in-plane and out-of-plane direction in real space, respectively. The in-plane (out-of-plane) lattice constant can be calculated as $\lambda /{\mathrm{Q}}_{[110]}(3\lambda /{\mathrm{Q}}_{[001]})$. (c) Cross-section high-angle annular dark-field (HAADF) STEM image with elemental mapping measured by spatially resolved EELS. (d) Temperature-dependent magnetization curve. The magnetization is normalized to the value at 350 K, and the paramagnetic signal of the bare NGO(001) substrate is included for comparison. (e) The corresponding field-dependent magnetization loops measured at 200 K. rlu.

Figure 2

Temperature-dependent magnetizations of LSMO/ CRTO superlattices, [2.8/1.2]_N, grown on NGO(001) are normalized to the value at 350 K. The black arrow denotes the reduction in magnetization.

Figure 3

Field-dependent magnetization loops of LSMO/CRTO superlattices, [2.8/1.2]_N, grown on NGO(001) measured at 200 K. The right panel shows the switching process of LSMO layers in decreasing field direction and the red arrows represent the direction of magnetization of individual LSMO layers.

Figure 4

(a) Temperature-dependent magnetization curves of LSMO/CRTO superlattices, [2.8/1.2]_N, grown on LSAT(001) measured at 300 Oe. (b) The corresponding field-dependent magnetization loops measured at 150 K. (c) High-resolution X-ray θ-2θ linear scan for [2.8/1.2]₁₀ grown on LSAT(001).

Figure 5

(a) Field-dependent magnetization loops of LSMO/CRTO superlattices, [2.8/1.2]₁₀, grown on NGO(001) measured at different temperatures. The inset shows the complete loop measured at 150 K. (b) Phase diagram drawn with the values of the field where magnetization begins to drop at each step in the decreasing field direction of loops measured at different temperatures. FM denotes that all LSMO layers are parallel, AF denotes that all adjacent LSMO layers are antiparallel, and IS denotes that two LSMO layers are parallel and the rest of eight LSMO layers are antiparallel. Temperature-dependent IEC strength (J_IEC, blue spheres) and the fitting results obtained with (c) model 1, (d) model 2, and (e) model 3.