Current-Gated Orthogonal Superconducting Transistor

TL;DR

The paper addresses nonreciprocal superconducting transport by exploiting intrinsic anisotropy in 2D superconductors to create a transverse superconducting diode effect under a DC bias. Using Ginzburg-Landau theory and self-consistent mean-field calculations, it shows that a bias along a misaligned direction converts anisotropy into a transverse nonreciprocal response and can drive a bias-controlled unidirectional superconductivity (USC) above a threshold. It generalizes the mechanism to high-$T_c$ cuprates with fourfold anisotropy and demonstrates two practical devices: a tunable supercurrent range controller and a bias-controlled half-wave rectifier, all guided by simple angle-dependent design rules. The results expand material platforms and enable bias-tunable, symmetry-breaking-free nonreciprocal superconducting electronics using a multi-terminal CGOST architecture.

Abstract

Nonreciprocal charge transport in superconductors enables rectification but is usually limited to the longitudinal direction. In this work, we show that a direct current bias injected off principal axes in two-dimensional anisotropic superconductors converts anisotropy into transverse nonreciprocity, enabling supercurrent diode effect measurement. This is demonstrated within both a Ginzburg-Landau framework and self-consistent mean-field calculations. When the control bias exceeds its critical value, the transverse dissipationless currents can only flow unidirectionally. This mechanism motivates the design of a multi-terminal current-gated orthogonal superconducting transistor (CGOST) and yields simple, bias direction angle-dependent design rules for device optimization. As direct applications, we propose a tunable supercurrent range controller and a half-wave rectifier based on the CGOST. Our findings open new avenues for developing nonreciprocal superconducting electronic devices.

Figures (6)

• Figure 1: (a) Schematic of CGOST, the bias current $j_{bias}$ is applied along the direction of the black arrow (control terminal), and we measure the transverse critical currents $j_{mc\pm}$ (function terminal) in the direction vertical to the bias. (b) The working mechanism of CGOST is based on a two-fold anisotropic superconductor. The distribution of critical current shows an elliptical pattern. The critical currents with positive ($j_{mc1+}$) and negative ($j_{mc2+}$, $j_{mc1-}$ and $j_{mc2-}$) transverse components are labeled in red and blue, respectively. When $j_{\parallel}$ is within the range of the ellipse, $j_{m}$ can be either positive or negative ($j_{mc1-}\leq j_{m}\leq j_{mc1+}$), but when $j_{\parallel}$ is greater than the critical current along the bias direction $j_{c+}(\theta)$, $j_{m}$ can only be negative ($j_{mc2-}\leq j_{m}\leq j_{mc2+}<0$).
• Figure 2: (a) The evolution of $j_{mc+}$ and $j_{mc-}$ (labeled as red and blue lines respectively) with increasing $j_{bias}$, $j_{mc+}$ becomes negative when $j_{bias} > j_{bias,c1}=1.217et$. Here we set $\theta = \pi/4$. (b) The diode efficiency $\eta(j_{bias})$ for three different $\theta$. Parameters used here: $\alpha = -0.5t, \beta = 0.1t, m_x = 4/t, m_y = 1/t$. $t$ is an arbitrary unit.
• Figure 3: (a) Schematic illustration of anisotropic Fermi surface with $s$-wave pairing. (b) Schematic illustration of an isotropic Fermi surface with nematic $p$-wave pairing. (c) Similar to Fig. \ref{['fig:fig2']}(a), the results of the square lattice model ($t_x \neq t_y$) with $s$-wave pairing, $j_{mc+}$ becomes negative when $j_{bias} > j_{bias,c1} = 0.0034et$. Parameter used here: $\theta = \pi/4, t_x = 0.5t_y = t$, $\mu = 2.5t$, $U = 1.5t$. (d) The results of the square lattice model ($t_x = t_y$) with $p$-wave pairing, $j_{mc-}$ becomes positive when $j_{bias} > j_{bias,c1} = 0.0086et$. Parameter used here: $\theta = \pi/4, t_x = t_y = t$, $\mu = 2.5t$, $U = 3.25t$.
• Figure 4: (a) The black curve shows the envelope of the critical momenta $\bm{q}_c(\theta)$. The blue curve shows the envelope of the critical current $j_{c\pm}(\theta)$. Both are derived from the GL analysis of the $d_{x^2-y^2}$-wave superconductor. (b) The evolution of $j_{mc+}$ (solid lines) and $j_{mc-}$ (dashed lines) from the GL analysis (in red) and MF calculations (in blue), $j_{bias}$ is rescaled to $[0,1]$ in arbitrary units $j_0$. Parameter used in GL analysis: $\theta = \pi/8, \alpha_0 = -t, \beta_0 = 2t, K_{1}^{d} = 0.5t, K_{2}^{d} = 4t, K_{3}^{d} = t$, then $j_{bias,c1} = j_{c+}(\theta) = 0.7096et, j_{bias,c1}/j_{0}=0.8870$ with $j_{0}=0.8et$, and $Q_{USC} = 0.054$. Parameter used in MF calculation: $\theta = \pi/8, t_x = t_y = t, \mu = 2.5t, U = 3t$, then $j_{bias,c1} = j_{c+}(\theta) = 0.0112et$, $j_{bias,c1}/j_{0}=0.8615$ with $j_{0}=0.013et$ and $Q_{USC} = 0.076$.
• Figure 5: (a) Circuit schematic of the bias-controlled supercurrent rectifier. The input AC current $I_{in} = I_0 \sin(\omega t)$, the load resistance $R_1 = R_2 = 2R_0$. $R_d$ is the resistance of the CGOST. The rectifier is set to its working regime, so the bias current $j_{bias}=j_{bias,c1}$ and the amplitude of the input current $I_{0} \leq j_{mc1}$. (b) The evolution of $R_d$ (black solid line) with the input AC current $I_{in}$ (blue dashed line). (c) The evolution of the output voltage $U_{out}$ (red solid line) and the output current $I_{out}$ (blue solid line) with the input AC current $I_{in}$ (blue dashed line), $U_0 = I_0R_0$. (d) The quality factor $Q_{rec}(\theta)$ of different effective mass $m_x$ and $m_y$.