KTaO3-Based Supercurrent Diode

TL;DR

The paper demonstrates a geometrically programmable supercurrent diode effect (SDE) at the LaAlO$_3$/KTaO$_3$ interface by patterning reconfigurable superconducting weak links with conductive AFM lithography. SDE arises from the interplay of Meissner screening currents and asymmetric vortex surface barriers at the device edges, and its polarity can be reversed by shifting the WL position; rectification efficiencies up to about 13% are achieved under modest out-of-plane fields. Time-dependent Ginzburg–Landau simulations quantitatively reproduce the key features, validate the vortex-based mechanism, and reveal how measurement configurations and edge geometry shape the I$_c^\ ext{±}$ and η(B) patterns. In addition, two devices (D, E) exhibit magnetic-field–enhanced superconductivity with a slanted M-shaped I$_c$(B) dependence, suggesting exchange scattering by local moments and possible Weber blockade, while backgate tuning modulates ΔB without destroying the SDE. Overall, the LAO/KTO platform offers a versatile testbed for exploring 2D vortex dynamics and engineering nonreciprocal, low-dissipation superconducting devices for future quantum circuits.

Abstract

The supercurrent diode effect (SDE), characterized by nonreciprocal critical currents, represents a promising building block for future dissipationless electronics and quantum circuits. Realizing SDE requires breaking both time-reversal and inversion symmetry in the device. Here we use conductive atomic force microscopy (c-AFM) lithography to pattern reconfigurable superconducting weak links (WLs) at the LaAlO3/KTaO3 (LAO/KTO) interface. By deliberately engineering the WL geometry at the nanoscale, we realize SDE in these devices in the presence of modest out-of-plane magnetic fields. The SDE polarity can be reversed by simply changing the WL position, and the rectification efficiency reaches up to 13% under optimal magnetic field conditions. Time-dependent Ginzburg-Landau simulations reveal that the observed SDE originates from asymmetric vortex motion in the inversion-symmetry-breaking device geometry. This demonstration of SDE in the LAO/KTO system establishes a versatile platform for investigating and engineering vortex dynamics, forming the basis for engineered quantum circuit elements.

Figures (20)

• Figure 1: Supercurrent diode effect in KTO WLs A-C. (a) Layout of the reference Device A. Device A is created by cutting a 2D channel (dark green, width $w=400\,\text{nm}$) into the left and right halves by the red rectangle, and then bridging them with a nanowire (light green path) which serves as the WL. The gap $g$ created by the cutting corresponds with the length of the weak link: $l_{WL}=g$, which is estimated to be $\approx200\,\text{nm}$ (see Supplementary Note S2). The WL is centered in the vertical direction. We define $y$ to be the vertical distance between the center of WL to the bottom edge of the 2D channel, which equals $w/2=200\,\text{nm}$. I+, I-, V+ and V- indicate the current source, current drain and the two voltages leads used in the following four-terminal $I-V$ measurements. Positive bias current ($I>0$) flows from the right to the left. Positive magnetic field ($B>0$) points into the sample plane. (b) Layout of Device B, where WL is placed close to the bottom edge of the 2D channel ($w=400\,\text{nm}$, $y=32\,\text{nm}$). (c) Layout of Device C, where WL is placed close to the top edge of the 2D channel ($w=400\,\text{nm}$, $y=368\,\text{nm}$). (d) $I-V$ measurements of Device A at $B=-500\,\text{Oe}$ (top) and $B=+500\,\text{Oe}$ (bottom). The red curve is the $V$ vs $I$ curve under positive current ($I>0$) while the blue curve is $V$ vs $|I|$ curve under negative current ($I<0$). The arrows indicate the current sweep directions, while switching currents $I_{c+}$, $|I_{c-}|$ and retrapping currents $I_{r+}$, $|I_{r-}|$ are labeled. (e) $I-V$ measurements of Device B, where obvious mismatch between $I_{c+}$ and $|I_{c-}|$ can be observed. At $B=-500\,\text{Oe}$, $I_{c+}>|I_{c-}|$ while at $B=+500\,\text{Oe}$, $I_{c+}<|I_{c-}|$. (f) $I-V$ measurements of Device C. At $B=-500\,\text{Oe}$, $I_{c+}<|I_{c-}|$ while at $B=+500\,\text{Oe}$, $I_{c+}>|I_{c-}|$. Note: all plots in this figure were taken at $T=50$ mK with a backgate voltage $V_{bg}=-30$ V applied on Devices A-C.
• Figure 2: Magnetic field sweep of Devices A through C. Panels (a), (b), and (c) show intensity plots of differential resistance $dV/dI$ versus $I$ versus $B$ for Devices A, B, and C, respectively. We note that in these plots, current $I$ sweeps from $I=0$ to $|I|>0$ to capture the switching behavior from superconducting state to normal state. Panels (d), (e), and (f) display the extracted switching currents $I_{c\pm}$ as a function of $B$ for Devices A, B, and C. Panels (g), (h), and (i) show the extracted diode efficiency $\eta$ as a function of $B$ for Devices A through C. On the top left corner of each panel, the label consists of a letter that indicates the corresponding Device, and a number that points to the measurement configuration (mapping in Figure \ref{['config']}). All measurements were performed at $T=50$ mK with a backgate voltage $V_{\rm bg}=-30$ V.
• Figure 3: Supercurrent diode D and E with slanted M-shaped $I_c$ vs $B$ pattern. (a) Device layout. Device D and E were patterned together in one run by c-AFM lithography. Both WLs are put in the $w=400\,\text{nm}$ 2D channel with WL D(E) positioned at $y=36\,\text{nm}$ ($y=364\,\text{nm}$). The measurement configuration for Device D(E) is indicated in black(orange) $I+/I-/V+/V-$ labels, as two leads are shared between them. (b)(c) $I$-$V$ measurements of Device D(E) at $B=\pm500\,\text{Oe}$. (d) $dV/dI$ vs $I$ vs $B$ intensity plot of Device D. (e) $I_{c\pm}$ vs $B$ of Device D, which follows a slanted "M" pattern. Black dashed lines label the two $I_{c+}$ maxima at $B=+400\,\text{Oe}$ and at $B=-380\,\text{Oe}$, while $I_{c+}(B=0)$ is lower than either of these maxima. (f) $\eta$ vs $B$ relation of Device D. (g)(h)(i) $dV/dI$ vs $I$ vs $B$ intensity plot, $I_{c\pm}$ vs $B$ and $\eta$ vs $B$ relations of Device E. In panel (h), the two $I_{c+}$ maxima occur at $B=+540\,\text{Oe}$ and at $B=-620\,\text{Oe}$. All plots in this figure taken at $T=50$ mK with backgate grounded $V_{\rm bg}=0$ V on Devices D & E.
• Figure 4: Simulation of KTO WL using time-dependent Ginzburg-Landau theory. (a) Device geometry used in TDGL simulation. The 2D channel (blue) has width $w=400\,\text{nm}$ and length $l=1200\,\text{nm}$. On its top and bottom edges there are current source and drain, indicated by the orange and green bars respectively. Positive current $I>0$ flows from the top to the bottom. The narrow constriction (WL) has width $w_{WL}=50\,\text{nm}$ and length $l_{WL}=200\,\text{nm}$, located near the left edge of the 2D channel. Phase difference and voltage between the two black dots are output by TDGL calculation. We define positive magnetic field ($B>0$) to be pointing into the sample plane, same as the experimental setup. (b) Current density $\textbf{K}(x,y)$ calculated under the condition $B=-2000\,\text{Oe}$ and $I=+150\,\text{nA}$. Color scale indicates magnitude of $\textbf{K}$ while white arrows indicate its direction. The two blue arrows point to the two locations with relatively low surface barrier for vortex entry. (c) $\textbf{K}(x,y)$ calculated at $B=-2000\,\text{Oe}$ and $I=-150\,\text{nA}$. (d) Evolution of phase difference $\Delta \phi$ and voltage $V$ as a function of time. Black dashed line: time-averaged voltage under $I=+150\,\text{nA}$ bias. (e) $V$ vs $I$ vs $B$ intensity plot from pyTDGL simulation. (f) $I_{c\pm}$ vs $B$ relation extracted from the simulated $I$-$V$ curves. (g) $\eta$ vs $B$ relation extracted from (f).
• Figure S1: Retrapping currents of Devices A-C vs magnetic field. (a)(b)(c) Intensity plots of differential resistance $dV/dI$ vs $I$ vs $B$ of Devices A, B and C. In these plots, current $I$ sweeps from $|I|>0$ to $I=0$, as indicated by the black arrows above each plot. (d)(e)(f) Extracted retrapping currents $I_{r\pm}$ of Devices A-C. These plots are from the same dataset as main text Figure 2, which was taken at $T=50$ mK with a backgate voltage $V_{\rm bg}=-30$ V applied on Devices A-C.