Perfect Superconducting Diode

• Perfect superconducting diode is defined by nonreciprocal superconducting transport with finite, dissipationless current in one direction and no critical current in the opposite direction.
• Various implementations, including intrinsic bulk, Josephson, vortex-based, and drive-engineered devices, exploit symmetry breaking and phase-space control to approach or achieve perfect efficiency.
• Engineering challenges focus on balancing unidirectionality, field-free operation, and high-temperature functionality to enable practical superconducting electronics and quantum circuit applications.

Searching arXiv for papers on perfect superconducting diodes and closely related superconducting diode mechanisms. arXiv search query: "all:perfect superconducting diode OR ti:superconducting diode effect" A perfect superconducting diode is the ideal limit of nonreciprocal superconducting transport: a finite dissipationless critical current exists in one direction, while the opposite-direction critical current vanishes, so that zero-voltage transport is strictly unidirectional. Across the literature, this limit is variously formulated as $I_{c+}>0$ with $I_{c-}=0$, $|\eta|=1$, $\epsilon=I_c^-/I_c^+=0$, or effectively infinite nonreciprocity $A=|I_c^+/I_c^-|\to\infty$, depending on whether the platform is an intrinsic bulk superconductor, a Josephson device, or a driven nonequilibrium system (Hosur, 24 Dec 2025, Borgongino et al., 11 Apr 2025, Hovhannisyan et al., 29 Aug 2025). Current research spans equilibrium symmetry arguments, finite-momentum and proximity mechanisms, vortex- and self-field-based rectifiers, interferometric Josephson designs, and microwave- or light-driven protocols, with demonstrated operation ranging from millikelvin conventional junctions to cuprates above liquid-nitrogen temperature (Hosur et al., 2022, Wang et al., 29 Sep 2025, Ichikawa et al., 24 May 2026).

1. Definitions and figures of merit

The superconducting diode effect is operationally defined by unequal critical or switching currents for opposite current directions, $I_{c+}\neq |I_{c-}|$, or, in switching experiments, $I_{\rm SW}^+\neq |I_{\rm SW}^-|$ (Mazur et al., 2022, Qi et al., 5 Jan 2025). In the common convention used for Josephson and bulk devices, the diode efficiency is

$\eta=\frac{I_{c+}-|I_{c-}|}{I_{c+}+|I_{c-}|},$

or the same expression multiplied by $100\%$ (Qi et al., 5 Jan 2025, Borgongino et al., 11 Apr 2025). In this convention, $\eta=0$ is reciprocal and $I_{c-}=0$0 is the ideal diode limit.

A second convention, developed for intrinsic bulk superconductors, uses

$I_{c-}=0$1

so that $I_{c-}=0$2 and $I_{c-}=0$3 again denote perfect unidirectionality (Hosur, 24 Dec 2025). Self-field Josephson diodes also use the nonreciprocity factor

$I_{c-}=0$4

with $I_{c-}=0$5 and $I_{c-}=0$6 reported for optimized planar Nb junctions (Hovhannisyan et al., 29 Aug 2025).

The notion of “perfect” depends on the operating manifold. In standard superconducting diodes, the useful states are a zero-voltage branch in one direction and a resistive branch in the other. In the quantum superconducting diode realized in twisted cuprates, the useful operating states are both Cooper-paired states: $I_{c-}=0$7 for one polarity and a quantized Shapiro plateau $I_{c-}=0$8 for the opposite polarity under microwave irradiation (Wang et al., 29 Sep 2025). In interferometric devices, a related regime is the “supercurrent range controller,” where zero-voltage transport exists only inside a finite current interval that excludes zero current (Sun et al., 13 Jul 2025).

2. Symmetry principles and thermodynamic constraints

The standard symmetry requirement is simultaneous breaking of inversion symmetry $I_{c-}=0$9 and time-reversal symmetry $|\eta|=1$0. In proximitized Rashba nanowires, this is expressed through the Edelstein-type term

$|\eta|=1$1

which is odd under both $|\eta|=1$2 and $|\eta|=1$3, and therefore produces nonreciprocal critical currents (Mazur et al., 2022). In field-free cuprate flakes, the same logic is captured by a Lifshitz-invariant contribution to the GL free energy,

$|\eta|=1$4

whose associated current is odd in superfluid momentum and yields $|\eta|=1$5 when the superconducting state already breaks $|\eta|=1$6 and $|\eta|=1$7 at $|\eta|=1$8 (Qi et al., 5 Jan 2025).

A stringent equilibrium constraint was formulated for intrinsic bulk superconductors in terms of a condensation free energy $|\eta|=1$9 as a function of Cooper-pair momentum $\epsilon=I_c^-/I_c^+=0$0, with current density

$\epsilon=I_c^-/I_c^+=0$1

If superconductivity exists on a finite interval $\epsilon=I_c^-/I_c^+=0$2 with $\epsilon=I_c^-/I_c^+=0$3, then exact $\epsilon=I_c^-/I_c^+=0$4 is incompatible with continuity of $\epsilon=I_c^-/I_c^+=0$5: a strictly one-sided current $\epsilon=I_c^-/I_c^+=0$6 would make $\epsilon=I_c^-/I_c^+=0$7 positive, contradicting $\epsilon=I_c^-/I_c^+=0$8 (Hosur, 24 Dec 2025). The same work argues that $\epsilon=I_c^-/I_c^+=0$9 is possible only by tuning to a critical point inside the superconducting phase, where $A=|I_c^+/I_c^-|\to\infty$0 becomes non-analytic. Away from such internal criticality, a finite-order Landau expansion yields a lower bound

$A=|I_c^+/I_c^-|\to\infty$1

with $A=|I_c^+/I_c^-|\to\infty$2 a Chebyshev polynomial and $A=|I_c^+/I_c^-|\to\infty$3 the fraction of the superconducting momentum window carrying positive current (Hosur, 24 Dec 2025).

Those thermodynamic restrictions were stated not to constrain proximity-based routes, Josephson devices, or nonequilibrium driven systems (Hosur, 24 Dec 2025). That distinction now organizes much of the field: perfect intrinsic bulk diodes remain subtle, whereas engineered Josephson, vortex, and driven platforms have already reached or closely approached $A=|I_c^+/I_c^-|\to\infty$4.

3. Intrinsic and proximity-based equilibrium routes

One proximity-based route starts from a metal with asymmetric dispersion $A=|I_c^+/I_c^-|\to\infty$5, proximitized by a conventional $A=|I_c^+/I_c^-|\to\infty$6-wave superconductor. The resulting Bogoliubov–de Gennes Hamiltonian

$A=|I_c^+/I_c^-|\to\infty$7

supports an equilibrium supercurrent and a strongly nonreciprocal current–momentum relation. In this framework, a perfect diode appears when the band-asymmetry scale exceeds the parent superconductor’s critical pair momentum: if $A=|I_c^+/I_c^-|\to\infty$8, then all allowed superconducting states carry current in the same direction and the diode coefficient reaches $A=|I_c^+/I_c^-|\to\infty$9 (Hosur et al., 2022).

A different microscopic proposal uses $I_{c+}\neq |I_{c-}|$0-wave altermagnets. There the normal-state dispersion

$I_{c+}\neq |I_{c-}|$1

drives broad finite-momentum Fulde–Ferrell regimes. The superconducting diode efficiency

$I_{c+}\neq |I_{c-}|$2

can reach $I_{c+}\neq |I_{c-}|$3 in the high-field FF′ regime, where competition between BCS and finite-momentum phases is tied to a topological nodal-to-nodeless transition of the spin-split Fermi surfaces (Chakraborty et al., 2024). This proposal is notable because the large efficiencies are explicitly linked to competition among multiple superconducting states rather than to interface engineering.

A third equilibrium route is the chiral nanotube-based Josephson junction. In that GL theory, an axial magnetic field quantizes the circumferential momentum and induces both an anomalous phase and a phase-independent persistent current. The CPR can be written

$I_{c+}\neq |I_{c-}|$4

with diode efficiency

$I_{c+}\neq |I_{c-}|$5

The central result is that the SDE is independent of the anomalous phase: instead, a non-reciprocal persistent current protected by fluxoid quantization can activate the diode effect, and, in principle, produce perfect diode efficiency even without higher-order pair tunneling processes (Cuozzo et al., 3 Apr 2025).

Experimental field-free intrinsic behavior is presently less extreme but technologically important. In BSCCO flakes without engineered junctions, nominally zero-field SDE was observed up to $I_{c+}\neq |I_{c-}|$6 K, with $I_{c+}\neq |I_{c-}|$7 at $I_{c+}\neq |I_{c-}|$8 K, and interpreted in terms of a superconducting state that already breaks $I_{c+}\neq |I_{c-}|$9 and $I_{\rm SW}^+\neq |I_{\rm SW}^-|$0 through intra-unit-cell loop-current order (Qi et al., 5 Jan 2025). This does not realize perfection, but it demonstrates that high-$I_{\rm SW}^+\neq |I_{\rm SW}^-|$1 superconductors can host intrinsic nonreciprocity without applied field or structural junction asymmetry.

4. Josephson and interferometric engineering

Short hybrid Josephson junctions remain the most controlled platform for studying how microscopic nonreciprocity enters practical devices. In InSb nanowires proximitized by Al, the switching-current asymmetry was defined by

$I_{\rm SW}^+\neq |I_{\rm SW}^-|$2

and reached roughly $I_{\rm SW}^+\neq |I_{\rm SW}^-|$3 in the perpendicular-field, high-supergate regime. The effect is strongest at an angle $I_{\rm SW}^+\neq |I_{\rm SW}^-|$4, interpreted as the spin-orbit field direction in the proximitized leads, and can be enhanced, reshaped, or almost completely suppressed by electrostatic gating (Mazur et al., 2022). Although far from perfect, this work established the gate-tunable Josephson diode as a controllable nanowire element.

Monolithic dc-SQUIDs based on all-Al 3D Dayem nanobridges show a complementary route in which the high harmonic content of the current–phase relation, rather than large screening inductance, produces nonreciprocity. With a bridge asymmetry parameter $I_{\rm SW}^+\neq |I_{\rm SW}^-|$5, the measured flux-tunable rectification efficiency reaches $I_{\rm SW}^+\neq |I_{\rm SW}^-|$6, while the zero-inductance theory predicts a maximum $I_{\rm SW}^+\neq |I_{\rm SW}^-|$7 near $I_{\rm SW}^+\neq |I_{\rm SW}^-|$8 (Greco et al., 2023). This design is significant because its downsizing is not limited by the need for a large SQUID inductance.

Interferometric perfection is obtained most explicitly in the multi-wire SQUID model. For $I_{\rm SW}^+\neq |I_{\rm SW}^-|$9 nanowires obeying a linear CPR, the Meissner phase correlation is

$\eta=\frac{I_{c+}-|I_{c-}|}{I_{c+}+|I_{c-}|},$0

and the total current is a linear function of the individual phase drops (Sun et al., 13 Jul 2025). A $\eta=\frac{I_{c+}-|I_{c-}|}{I_{c+}+|I_{c-}|},$1-invariant perfect superconducting diode emerges when one wire $\eta=\frac{I_{c+}-|I_{c-}|}{I_{c+}+|I_{c-}|},$2 sits at the average position,

$\eta=\frac{I_{c+}-|I_{c-}|}{I_{c+}+|I_{c-}|},$3

and the vorticity configuration satisfies

$\eta=\frac{I_{c+}-|I_{c-}|}{I_{c+}+|I_{c-}|},$4

Under these conditions the negative critical current can be made exactly zero, $\eta=\frac{I_{c+}-|I_{c-}|}{I_{c+}+|I_{c-}|},$5, and the perfect efficiency remains stable against small changes of magnetic field (Sun et al., 13 Jul 2025). The same architecture also supports supercurrent range controllers, where superconductivity exists only inside a finite current window that excludes zero.

These Josephson and interferometric devices demonstrate a recurrent theme: exact one-sided superconducting transport is much easier to obtain once the problem is recast as phase-space engineering in a finite set of weak links, rather than as an equilibrium property of a single homogeneous bulk condensate.

5. Vortex, self-field, and electrothermal diodes

A major class of near-perfect diodes exploits vortex entry, guided motion, and electrothermal switching rather than intrinsic nonreciprocal pairing. In amorphous Mo$\eta=\frac{I_{c+}-|I_{c-}|}{I_{c+}+|I_{c-}|},$6Ge$\eta=\frac{I_{c+}-|I_{c-}|}{I_{c+}+|I_{c-}|},$7, conformal-mapped nanoholes break in-plane inversion symmetry and create an asymmetric pinning landscape. Time-dependent GL plus heat diffusion shows that one current polarity nucleates hot spots and a normal strip while the opposite polarity remains in a superconducting flux-flow state, producing millivolt rectification signals three orders of magnitude larger than conventional flux-quantum diodes (Lyu et al., 2021). The effective critical-current asymmetry is only of order a few percent, so the device is not perfect in the strict sense, but it approximates diode behavior over a finite current window.

A more systematic GL design uses a central superconducting film flanked by two current-carrying side wires that generate a tailored inhomogeneous field. In this geometry, diode efficiencies were defined separately for flux-flow onset and for complete normal-state switching: $\eta=\frac{I_{c+}-|I_{c-}|}{I_{c+}+|I_{c-}|},$8 Numerical optimization gives an ideal superconducting half-wave rectifier with efficiencies surpassing $\eta=\frac{I_{c+}-|I_{c-}|}{I_{c+}+|I_{c-}|},$9, and identifies optimal edge fields around $100\%$0 for the NSD and $100\%$1 for the FFD (Cadorim et al., 2024). This route is conceptually close to perfection because it directly engineers one half-cycle to remain superconducting while the opposite half-cycle is driven normal by vortices and hot spots.

The strongest experimental near-perfect result in this class comes from planar Nb Josephson junctions with geometric self-field asymmetry. The optimization conditions are

$100\%$2

so that the central maximum of one $100\%$3 branch coincides with the first minimum of the opposite branch (Hovhannisyan et al., 29 Aug 2025). In the optimized device D2 at $100\%$4 K, the forward critical current is $100\%$5, while the reverse current is below the resolution limit, $100\%$6, yielding $100\%$7 and $100\%$8. The same device rectifies $100\%$9 GHz radiation without an observable threshold, demonstrating near-ideal optical nonreciprocity (Hovhannisyan et al., 29 Aug 2025).

Electrical programmability was added in a nanoscale electrothermal-switch diode based on NbN nanowires. There a gate-controlled hotspot dynamically breaks inversion symmetry and generates two coexisting nonreciprocal regimes: a nonreciprocal superconducting-to-normal transition with efficiencies up to $\eta=0$0, and a ratchet-like vortex regime with efficiencies up to $\eta=0$1 (Li et al., 14 Apr 2026). The diode can be switched on, off, or reversed in polarity in situ by a small gate current, enabling electrically reconfigurable full-wave and half-wave rectification. Editable vortex-based nonreciprocity was also realized in LaAlO$\eta=0$2/KTaO$\eta=0$3, where c-AFM edge writing changes the polarity and magnitude of the SDE, producing efficiencies above $\eta=0$4 and rectification signals exceeding $\eta=0$5 mV in a single nonvolatile device (Wang et al., 10 Nov 2025).

6. High-temperature, quantum, and driven perfect diodes

High operating temperature has become a defining axis of progress. In single BSCCO flakes, nominally zero-field SDE persists up to $\eta=0$6 K with $\eta=0$7 at $\eta=0$8 K and stability beyond two hundred sweeping cycles (Qi et al., 5 Jan 2025). In twisted BSCCO artificial Josephson junctions, a small perpendicular magnetic field induces a vortex-based Josephson diode effect for all studied twist angles, with a record asymmetry of $\eta=0$9 at $I_{c-}=0$00 K and operation up to $I_{c-}=0$01 K (Ghosh et al., 2022). Twisted cuprate Josephson junctions then pushed the “perfect” regime into the quantum Josephson domain: after current training and under microwave irradiation, a quantum superconducting diode with perfect efficiency was realized up to $I_{c-}=0$02 K, above liquid-nitrogen temperature, while the useful states remained Cooper-paired Shapiro-step states rather than superconducting-to-normal switching states (Wang et al., 29 Sep 2025).

Drive engineering provides a distinct route to exact $I_{c-}=0$03. In a conventional Al–InAs–Al Josephson junction, a biharmonic current

$I_{c-}=0$04

breaks spatio-temporal symmetries and yields effective critical currents

$I_{c-}=0$05

At $I_{c-}=0$06 and suitable amplitudes, one of the effective critical currents vanishes, producing $I_{c-}=0$07 over a broad frequency range from $I_{c-}=0$08 to several GHz, with temperature resilience up to $I_{c-}=0$09 mK (Borgongino et al., 11 Apr 2025). This route is platform-independent and requires neither exotic materials nor static magnetic asymmetry.

Light-driven nonequilibrium control generalizes the same idea to intrinsic superconductors. In the TDGL formulation,

$I_{c-}=0$10

with monochromatic or multi-frequency $I_{c-}=0$11, perfect SDE arises by dynamically reshaping the allowed superconducting $I_{c-}=0$12-window and generating nonlinear dc photocurrents (Ichikawa et al., 24 May 2026). Monochromatic light produces perfect SDE in systems already lacking inversion and time-reversal symmetry, while multi-frequency light produces perfect SDE even in centrosymmetric systems by breaking the dynamical symmetry $I_{c-}=0$13. This result explicitly identifies symmetry engineering of the drive as a general principle for unidirectional superconducting transport (Ichikawa et al., 24 May 2026).

Platform	Key reported performance	Regime
BSCCO flake (Qi et al., 5 Jan 2025)	field-free SDE up to $I_{c-}=0$14 K; $I_{c-}=0$15 at $I_{c-}=0$16 K	intrinsic high-$I_{c-}=0$17
Twisted BSCCO AJJ (Ghosh et al., 2022)	JDE up to $I_{c-}=0$18 K; asymmetry $I_{c-}=0$19 at $I_{c-}=0$20 K	vortex-based Josephson
Twisted cuprate QSD (Wang et al., 29 Sep 2025)	perfect diode efficiency up to $I_{c-}=0$21 K	microwave-driven quantum diode
Conventional JJ with biharmonic drive (Borgongino et al., 11 Apr 2025)	$I_{c-}=0$22 from Hz to GHz; up to $I_{c-}=0$23 mK	spatio-temporal symmetry breaking
Planar Nb JJ (Hovhannisyan et al., 29 Aug 2025)	$I_{c-}=0$24, $I_{c-}=0$25; threshold-free $I_{c-}=0$26 GHz rectification	self-field optical diode

These examples show that “perfect superconducting diode” now refers to several experimentally distinct limits: exact one-way critical currents in driven Josephson junctions, effectively resolution-limited one-way transport in self-field Nb devices, and quantized one-way Josephson states in twisted cuprates.

7. Open problems and circuit relevance

The literature now distinguishes three partially independent targets: exact unidirectionality, field-free operation, and high-temperature scalability. Exact unidirectionality has been reached most cleanly in driven and interferometric Josephson settings (Borgongino et al., 11 Apr 2025, Sun et al., 13 Jul 2025), while near-perfect static transport has been realized in self-field planar Nb junctions (Hovhannisyan et al., 29 Aug 2025). Field-free operation has been demonstrated in high-$I_{c-}=0$27 BSCCO flakes, but with moderate efficiency (Qi et al., 5 Jan 2025). High-temperature perfect efficiency has been achieved in twisted-cuprate quantum diodes, though under microwave irradiation and after current training (Wang et al., 29 Sep 2025).

Several engineering tensions recur across platforms. Strong nonreciprocity often competes with large gap, large forward critical current, and low dissipation; this tradeoff is explicit in proximitized nanowire Josephson diodes, where stronger nonreciprocity tends to coincide with finite-momentum physics, reduced induced gap, and greater sensitivity to disorder and field alignment (Mazur et al., 2022). Vortex-based devices offer high temperature and large rectified voltages, but their diode action can depend on metastable pinning landscapes and thermal relaxation (Lyu et al., 2021, Wang et al., 10 Nov 2025). Drive-based perfect diodes avoid some equilibrium restrictions, yet introduce their own constraints through adiabaticity, RF power delivery, phase noise, and system-level power balance (Borgongino et al., 11 Apr 2025).

The circuit relevance is already concrete. Gate-tunable Josephson diodes have been proposed for on-chip gyrators and circulators and as elements of innovative superconducting circuits and computation devices (Mazur et al., 2022). Biharmonic-drive perfect diodes are explicitly positioned as efficient logic gates, ultra-fast switches, and dynamic half-wave supercurrent rectifiers (Borgongino et al., 11 Apr 2025). Planar Nb optical diodes point toward wireless sub-THz signal processing (Hovhannisyan et al., 29 Aug 2025). Electrothermal-switch diodes already realize electrically reconfigurable full-wave and half-wave rectification in a lithography-compatible architecture (Li et al., 14 Apr 2026). High-$I_{c-}=0$28 cuprate implementations further suggest a route to superconducting electronics and quantum circuits operating near liquid-nitrogen temperature (Qi et al., 5 Jan 2025, Ghosh et al., 2022, Wang et al., 29 Sep 2025).

A plausible implication is that no single mechanism presently dominates the path to perfection. Instead, the field is converging on a layered taxonomy. Bulk intrinsic perfect diodes remain constrained by general thermodynamic arguments unless criticality intervenes (Hosur, 24 Dec 2025). Near-perfect and exact one-way transport are already available in extrinsic, finite-size, and driven systems (Hovhannisyan et al., 29 Aug 2025, Borgongino et al., 11 Apr 2025). High-temperature operation is now compatible with both strong and perfect diode action in cuprates, albeit via distinct microscopic mechanisms (Ghosh et al., 2022, Wang et al., 29 Sep 2025). The practical perfect superconducting diode is therefore emerging less as a single universal object than as a family of symmetry-engineered superconducting elements, each optimized for a different combination of directionality, temperature, programmability, and circuit function.