High-Temperature Ferroic Glassy States in ${\mathrm{SrTiO}}_3$-Based Thin Films

Disordered ferroics hold great promise for next-generation magnetoelectric devices because their lack of symmetry constraints implies negligible hysteresis with low energy costs. However, the transition temperature and the magnitude of polarization and magnetization are still too low to meet application requirements. Here, taking the prototype perovskite of ${\mathrm{SrTiO}}_3$ as an instance, we realize a coexisting spin and dipole reentrant glass states in ${\mathrm{SrTiO}}_3$ homoepitaxial films via manipulation of local symmetry. Room-temperature saturation magnetization and spontaneous polarization reach $\sim 10\mathrm{emu}/{\mathrm{cm}}^3$ and $\sim 25\mathrm{\mu }\mathrm{C}/{\mathrm{cm}}^2$, respectively, with high transition temperatures (101 K and 236 K for spin and dipole glass temperatures and 556 K and 1100 K for Curie temperatures, respectively). Our atomic-scale investigation points out an underlying mechanism, where the $\mathrm{Ti}/\mathrm{O}$-defective unit cells break the local translational and orbital symmetry to drive the formation of unusual slush states. This study advances our understanding of the nature of the intricate couplings of ferroic glasses. Our approach could be applied to numerous perovskite oxides for the simultaneous control of the local magnetic and polar orderings and for the exploration of the underlying physics.

Figure 1

(a) $L$-can SXRD pattern and the fitting result. (b) RSM around (103) spots. (c) RBS spectra for the film and bare ${\mathrm{SrTiO}}_3$ crystal. (d) HAADF-STEM image, and (e) corresponding $c$/$a$ map.

Figure 2

(a) RT $M\text{−}H$ curves of the film. (b) RT PNR spectra with an in-plane magnetic field of 1.0 T. The inset is an experimental schematic. (c) $\mathrm{ZFC}/\mathrm{FC}$ $M\text{−}T$ curves of the film. Temperature-dependent (d) $M\text{−}H$ curves and (e) coercive field (${\mu }_0{H}_{\text{c}}$) and $M_{\text{s}}$. (f) ac susceptibility vs temperature curves under different frequencies. (g) Magnetic relaxation curves under $R_{\mathrm{FC}}$ and $R_{\mathrm{ZFC}}$ routes at 10 K and fitting lines.

Figure 3

(a) $P\text{−}E$ loop and leakage current curve. (b) Temperature-dependent XRD patterns of (002) peaks. (c) Temperature–frequency-dependent dielectric spectra and Vogel-Fulcher fitting results. (d) Comparison of transition temperatures in this work with those of the representative disordered ferroics [14, 15, 23, 24].

Figure 4

(a) ABF-STEM image of the film at [100] zone axis. (b) Intensity profiles of $\mathrm{Ti}/\mathrm{O}$ and O atom columns. Colored profiles correspond to the regions selected in (a) with the same color. Black arrows denote the $\mathrm{Ti}/\mathrm{O}$-deficient atomic columns. (c) $c$/$a$ and (d) polar displacement maps overlaid on the contrast-inverted image. Black boxes denote the corresponding $\mathrm{Ti}/\mathrm{O}$-defect columns marked by arrows in (b). (e) O $K$-edge and (f) Ti $L_{3,2}$-edge EELS spectra for the film and substrate. (g) Schematics of Ti $3d$ orbitals. (h) Ti $L_{3,2}$-edge XLD spectra. The inset shows the experimental configuration, where the x-ray polarization orientation is denoted by the yellow arrows. (i) RT Ti $L_{3,2}$-edge XMCD spectra, where ${\mu }^{+}/{\mu }^{-}$ denotes left-hand and right-hand circular x-ray polarization.