Critical Thickness in Superconducting ${\mathrm{LaAlO}}_3/\mathrm{KTa}{\mathrm{O}}_3(111)$ Heterostructures

Recently, two-dimensional superconductivity was discovered at the oxide interface between ${\mathrm{KTaO}}_3$ and ${\mathrm{LaAlO}}_3$ (or EuO), whose superconducting transition temperature $T_c$ is up to 2.2 K and exhibits strong crystalline-orientation dependence. However, the origin of the interfacial electron gas, which becomes superconducting at low temperatures, remains elusive. Taking the ${\mathrm{LaAlO}}_3/{\mathrm{KTaO}}_3(111)$ interface as an example, we have demonstrated that there exists a critical ${\mathrm{LaAlO}}_3$ thickness of $\sim 3\mathrm{nm}$. Namely, a thinner ${\mathrm{LaAlO}}_3$ film will give rise to an insulating but not conducting (or superconducting) interface. By in situ transport measurements during growth, we have also revealed that the critical thickness can be suppressed if exposure to oxygen is avoided. These observations, together with other control experiments, suggest strongly that the origination of the electron gas is dominated by the electron transfer that is from oxygen vacancies in the ${\mathrm{LaAlO}}_3$ film to the ${\mathrm{KTaO}}_3$ substrate.

Figure 1

Influence of LAO thickness $d_{\mathrm{LAO}}$ on the transport properties of $\mathrm{LAO}/\mathrm{KTO}(111)$ interfaces. The solid symbols represent the samples grown under normal conditions. Different symbol shapes are used when more than one sample has been fabricated with the same $d_{\mathrm{LAO}}$. The empty symbols represent the samples grown under other conditions in which the growth temperatures were varied from 100 to 800 °C, and the growth oxygen pressures were varied from 0 to $5\times {10}^{-5}\mathrm{mbar}$. The data were taken at room temperature, $T\approx 300\mathrm{K}$.

Figure 2

Comparison of two types of samples. Norm, grown under normal conditions. RT, grown at room temperature in vacuum. (a) Sheet resistance $R_s$ as a function of temperature. (b) A close view of $R_s(T)$ curves in the low-temperature range. (c) Influence of $d_{\mathrm{LAO}}$ (total) on the transport properties of $\mathrm{LAO}/\mathrm{KTO}(111)$ interfaces that were prepared by depositing $x$-nm LAO at RT on norm(1 nm) or, namely, $\mathrm{norm}(1\mathrm{nm})+\mathrm{RT}(x\mathrm{nm})$. Here, $d_{\mathrm{LAO}}(\mathrm{total})=(1+x)\mathrm{nm}$.

Figure 3

In situ transport measurements for depositing LAO at room temperature on norm(1 nm) samples. (a) Evolution of sheet conductance ${\sigma }_s$ with time. The chamber environment in different time zones was as labeled. The red arrows indicate the moment when a laser pulse shoots. Note that the durations of one laser pulse and its induced plume are $\sim 25\mathrm{ns}$ and $\sim 10\mu \mathrm{s}$, respectively. These two timescales are much shorter than the time interval of acquiring data in our measurements ($\sim 0.5\mathrm{s}$). Different timescales were used in different time zones. (b) Depositing 10-nm LAO film under $P({\mathrm{O}}_2)=0.1\mathrm{mbar}$. (c) Depositing LAO on a norm(1 nm) sample that had been annealed at 400 °C in $P({\mathrm{O}}_2)=1\mathrm{bar}$ flow for 24 h. The calibrated growth rate of LAO is $\sim 0.014\mathrm{nm}$ per laser pulse [17].

Figure 4

Low-temperature transport properties of ${\mathrm{Al}}_2{\mathrm{O}}_3/\mathrm{LAO}/\mathrm{KTO}(111)$ samples. The films on the top of KTO substrates are ${\mathrm{Al}}_2{\mathrm{O}}_3/\mathrm{LAO}$ bilayers that were grown subsequently under normal conditions. The thicknesses of the two layers in different samples are as labeled.