Two-Dimensional Superconductivity at the ${\mathrm{LaAlO}}_3/{\mathrm{KTaO}}_3(110)$ Heterointerface

We report on the observation of a $T_c\sim 0.9\mathrm{K}$ superconductivity at the interface between ${\mathrm{LaAlO}}_3$ film and the $5d$ transition metal oxide ${\mathrm{KTaO}}_3(110)$ single crystal. The interface shows a large anisotropy of the upper critical field, and its superconducting transition is consistent with a Berezinskii-Kosterlitz-Thouless transition. Both facts suggest that the superconductivity is two-dimensional (2D) in nature. The carrier density measured at 5 K is $\sim 7\times {10}^{13}{\mathrm{cm}}^{-2}$. The superconducting layer thickness and coherence length are estimated to be $\sim 8$ and $\sim 30\mathrm{nm}$, respectively. Our result provides a new platform for the study of 2D superconductivity at oxide interfaces.

Figure 1

Structural characterization of a 20-nm $\mathrm{LAO}/\mathrm{KTO}(110)$ heterostructure. (a) HAADF-STEM image shows that the LAO film is amorphous. The inset shows the image with a higher magnification and the atomic configuration (colored) of KTO. (b) EDS elemental mapping of the same sample. The top-left corner panel is the HAADF image of the region for EDS mapping. For comparison, the EDS mapping images of different elements (La, Ta, Al, and K) are presented both together and separately. The interface is abrupt, and intermixing is not significant. At the interface, about one monolayer of K is missing, and as a result, a La-rich layer is formed. The dashed line indicates the position of interface.

Figure 2

(a) Schematic view of the samples. (b) A photo of the central Hall bar area. The interface of the Hall bar region (light) is conducting; other regions (dark) are insulating. (c) Dependence of $R_{\text{sheet}}$ on $T$ of the 6- and 20-nm samples measured in a wide temperature range. (d) Dependence of $R_{\mathrm{Hall}}$ on magnetic field ${\mu }_{0}H$. (e) Dependence of $R_{\text{sheet}}$ on $T$ in low temperatures.

Figure 3

Transport behaviors under magnetic field for the 20-nm sample. Dependence of $R_{\text{sheet}}$ on $T$ for field (a) perpendicular and (b) parallel to the interface. Temperature dependence of upper critical field ${\mu }_0{H}_{c2}$, extracted from the 50% normal-state resistance, (c) perpendicular and (d) parallel to the interface. The extrapolated ${\mu }_0{H}_{c2}(0)$’s agree fairly well with that measured in the low-temperature magnetic-field-dependent resistance (Fig. S7).

Figure 4

(a) Temperature-dependent $I\text{−}V$ measurements for the 20-nm sample [the measured bridge is $20\times 100\mu {\mathrm{m}}^2$, the central region of the Hall bar shown in Fig. 2]. Inset: temperature dependence of the linear critical current density $I_c$. (b) $I\text{−}V$ curves on a logarithmic scale. The symbol label is the same as that in (a). The long black line corresponds to $V\sim I^3$ dependence and shows that $0.87<T_{\mathrm{BKT}}<0.88\mathrm{K}$. (c) Temperature dependence of the power-law exponent $\alpha$, as deduced from the fits shown in (b). (d) $R_{\text{sheet}}(T)$ dependence of the same sample [the same data shown in Fig. 2], plotted on a ${[d\ln (R_{\text{sheet}})/dT]}^{-2/3}$ scale. The solid line is the behavior expected for a BKT transition with $T_{\mathrm{BKT}}=0.92\mathrm{K}$.