Comparison of Origins of Re-Entrant Supercurrents at High In-Plane Magnetic Fields in Planar InAs-Al Josephson Junctions

Abstract

Hybrid superconductor-semiconductor systems with large spin-orbit coupling are important platforms for realizing topological or triplet superconductivity. Planar Josephson junctions made using these materials are predicted to enter the topological state by tuning the phase difference between the two superconductors from 0 to $π$. The 0-$π$ transition can be driven by magnetic field through Zeeman splitting of subbands in the semiconductor. It is expected to manifest as a node, or a re-entrance, in the critical current. Here we present re-entrant switching currents from several InAs/Al planar Josephson junctions in high in-plane magnetic fields. We find that re-entrances in some devices conform with expected signatures for topological or 0-$π$ transitions. However, we show that the data can also be explained in terms of mode interference in the junction in the presence of disorder. We also present simulations of supercurrent interference under in-plane fields that can reproduce re-entrances due to corrugated weak link without invoking the Zeeman effect or topology.

Figures (18)

• Figure 1: Schematic for the three processes leading to re-entrant supercurrents : (a) A trivial Zeeman $\pi$ junction with an antisymmetric order parameter denoted by the black curve, (b) A planar junction with spin-orbit coupling under an in-plane magnetic field leads to a $\pi$ phase difference and the appearance of Majorana Zero Modes at the ends of the junction, (c) A disordered junction which can make the electrons deviate from straight line paths and interfere with each other. The gray and green colors represent the superconductor and the quantum well respectively.
• Figure 2: Supercurrent interference data for Device 1 : (a) SEM Image of the planar InAs/Al Josephson junction. Image was taken before deposition of the dielectric and top gate. (b) Schematic of the device and heterostructure stack, with $x$ and $y$ directions indicated. (c) Differential resistance as a function the in-plane ($B_y$) and out-of plane ($B_z$) fields at DC bias = 0, (d) Differential resistance as a function of the dc current bias and $B_y$ at $B_z = -0.77$ mT (corresponds to the orange arrow in (c)), (e) Differential resistance as a function of $B_y$ and top gate voltage $V_g$ at $B_z = -0.77$ mT. A lock-in excitation of 2 nA at 43 Hz was used for (c), (d) and (e).
• Figure 3: Supercurrent interference data for Device 2: (a) SEM of the junction with two lengths $\approx$ 550 nm and 988 nm, (b) Differential resistance as a function of the in-plane and out-of-plane magnetic fields $B_x$ and $B_z$ at $V_g$ = -2.01 V, (c) Differential resistance as a function of $B_x$ and top gate voltage $V_g$ at $B_z$ = -0.91 mT. A lockin excitation of 1 nA at 49 Hz was used for (b) and (c).
• Figure 4: (a) SEM image of Device 4, (b) Supercurrent interference pattern for Device 4, and (b) Supercurrent interference pattern for Device 5 which looks similar to Device 4 (Supplementary Information Fig. \ref{['Reentrant_Device4']}(a). A lockin excitation of 1 nA at 43 Hz was used for both (b) and (c).
• Figure 5: Results of numerical simulations : (a) Normalized critical current as a function of the out-of-plane and in-plane magnetic fields, (b) Linecut showing a minimum in the critical current at $B_z$ = -1.5 mT (corresponds to the green dashed line in (a)).