Periodic Reporting for period 3 - SuperGate (Gate Tuneable Superconducting Quantum Electronics.)
Periodo di rendicontazione: 2023-03-01 al 2025-02-28
The SuperGate project supports global efforts to reduce the rising energy demands of high-performance computing (HPC) systems. Today’s supercomputers can consume nearly 1 GW—comparable to a small power plant—posing major energy and thermal challenges. SuperGate lays the foundation for greener HPC technology that could significantly lower energy costs.
Our research focused on designing, testing, and optimizing gate-controlled supercurrent (GCS) devices—a new class of superconducting components that function as low-dissipation analogues to CMOS transistors. Unlike CMOS, GCS devices switch states via gate voltage (Vg) but operate with minimal energy loss due to their superconducting nature.
We explored optimal materials and geometries to achieve high operational temperatures and minimal Vg, aligning with CMOS voltage ranges (<5V). This compatibility is essential for integrating GCS devices into existing tech. In parallel, we studied the physical mechanisms behind GCS to improve control and performance.
Additionally, we tested the switching speed of GCS devices and developed the first logic circuit prototype based on them—key steps toward their use in superconducting electronics and quantum computing.
Our research focused on designing, testing, and optimizing gate-controlled supercurrent (GCS) devices—a new class of superconducting components that function as low-dissipation analogues to CMOS transistors. Unlike CMOS, GCS devices switch states via gate voltage (Vg) but operate with minimal energy loss due to their superconducting nature.
We explored optimal materials and geometries to achieve high operational temperatures and minimal Vg, aligning with CMOS voltage ranges (<5V). This compatibility is essential for integrating GCS devices into existing tech. In parallel, we studied the physical mechanisms behind GCS to improve control and performance.
Additionally, we tested the switching speed of GCS devices and developed the first logic circuit prototype based on them—key steps toward their use in superconducting electronics and quantum computing.
Within the first 12 months since launching the project on 1st March 2021, the consortium successfully recruited the research team, established dedicated labs, and acquired essential equipment, including tools for simulating, fabricating, and testing logic circuits and gate-controlled supercurrent (GCS) devices at frequencies up to tens of GHz. Early research identified promising device geometries and materials, such as Ta/InAs core/shell nanowires, that reduce switching voltages in GCS devices. These results were presented at international conferences and led to multiple publications by the end of the first year.
From March 2022 to March 2023, we continued this research, discovering that superconductors with higher atomic numbers offer greater potential for reducing GCS switching voltages. We also observed that devices fabricated via top-down etching lack GCS, unlike similar bottom-up (lift-off) devices. Microstructural analysis linked GCS to microstrain and disorder, which are more pronounced in lift-off processes. Interestingly, GCS could be retrieved in etched devices using highly disordered Nb0.18Re0.82 but only with specific gas mixtures—indicating surface modification also plays a role.
We also established niobium (Nb) as a reliable material for reproducible GCS devices and developed a robust fabrication protocol. This marks a critical step toward future GCS-based technologies.
During this second phase, we conducted initial switching speed tests and simulated circuits combining GCS devices, such as COPY ports and switches. On the theoretical side, we developed new models to explain GCS behaviour and explored control mechanisms. Notably, adding magnetic impurities to the device surface lowered switching voltages.
Our findings suggest multiple coexisting mechanisms could drive GCS, depending on the fabrication method. While prior publications proposed direct field effects, our data point to a leakage-driven mechanism—though not simple quasiparticle heating. Parameters like disorder, microstrain, and surface modification emerged as critical for GCS. We also shifted our focus from minimising gate field strength to optimising gate (Vg) and output (Vout) voltages. In Nb-based devices, we achieved both high reproducibility and increased Vout. Using charge-trapping dielectrics like Al2O₃, we demonstrated the feasibility of non-volatile superconducting memory based on GCS.
Throughout the project, we critically examined electric field effects on superconductivity, both with and without charge injection. We proposed potential surface phases and mechanisms enabling electrostatic control of supercurrent. Our results provide a foundation for superconducting orbitronics, showing that surface electronic states significantly impact supercurrent amplitude. We also introduced magneto-electric mechanisms that may suppress supercurrent without involving leakage currents. A key challenge remains distinguishing effects due to leakage and phonon injection from purely electrostatic influences.
Dynamic response tests showed that GCS devices can function as switches with nanosecond-scale response times, making them suitable for microwave and logic applications. Temperature-dependent measurements linked these times to quasiparticle recombination in the weak link, suggesting potential for even faster switching with optimised materials and geometry. We also developed new methods to study the nonlinear response of metallic superconductors under sub-gap THz pulses, revealing that gate-induced superconductivity changes can occur within a few picoseconds.
Simulations and prototypes confirmed the feasibility of single logic ports using GCS. However, more complex circuits—like ring oscillators or electro-optical modulator interfaces—remain out of reach due to current limitations. Further engineering is needed to reduce gate voltage thresholds and boost output voltages.
To date, the SuperGate project has produced 76 articles (62 published, 14 under review or pre-print). Notably, we published a review paper summarising the GCS research done worldwide, analysing proposed physical mechanisms, and highlighting the technology’s broad application potential.
From March 2022 to March 2023, we continued this research, discovering that superconductors with higher atomic numbers offer greater potential for reducing GCS switching voltages. We also observed that devices fabricated via top-down etching lack GCS, unlike similar bottom-up (lift-off) devices. Microstructural analysis linked GCS to microstrain and disorder, which are more pronounced in lift-off processes. Interestingly, GCS could be retrieved in etched devices using highly disordered Nb0.18Re0.82 but only with specific gas mixtures—indicating surface modification also plays a role.
We also established niobium (Nb) as a reliable material for reproducible GCS devices and developed a robust fabrication protocol. This marks a critical step toward future GCS-based technologies.
During this second phase, we conducted initial switching speed tests and simulated circuits combining GCS devices, such as COPY ports and switches. On the theoretical side, we developed new models to explain GCS behaviour and explored control mechanisms. Notably, adding magnetic impurities to the device surface lowered switching voltages.
Our findings suggest multiple coexisting mechanisms could drive GCS, depending on the fabrication method. While prior publications proposed direct field effects, our data point to a leakage-driven mechanism—though not simple quasiparticle heating. Parameters like disorder, microstrain, and surface modification emerged as critical for GCS. We also shifted our focus from minimising gate field strength to optimising gate (Vg) and output (Vout) voltages. In Nb-based devices, we achieved both high reproducibility and increased Vout. Using charge-trapping dielectrics like Al2O₃, we demonstrated the feasibility of non-volatile superconducting memory based on GCS.
Throughout the project, we critically examined electric field effects on superconductivity, both with and without charge injection. We proposed potential surface phases and mechanisms enabling electrostatic control of supercurrent. Our results provide a foundation for superconducting orbitronics, showing that surface electronic states significantly impact supercurrent amplitude. We also introduced magneto-electric mechanisms that may suppress supercurrent without involving leakage currents. A key challenge remains distinguishing effects due to leakage and phonon injection from purely electrostatic influences.
Dynamic response tests showed that GCS devices can function as switches with nanosecond-scale response times, making them suitable for microwave and logic applications. Temperature-dependent measurements linked these times to quasiparticle recombination in the weak link, suggesting potential for even faster switching with optimised materials and geometry. We also developed new methods to study the nonlinear response of metallic superconductors under sub-gap THz pulses, revealing that gate-induced superconductivity changes can occur within a few picoseconds.
Simulations and prototypes confirmed the feasibility of single logic ports using GCS. However, more complex circuits—like ring oscillators or electro-optical modulator interfaces—remain out of reach due to current limitations. Further engineering is needed to reduce gate voltage thresholds and boost output voltages.
To date, the SuperGate project has produced 76 articles (62 published, 14 under review or pre-print). Notably, we published a review paper summarising the GCS research done worldwide, analysing proposed physical mechanisms, and highlighting the technology’s broad application potential.
Our research lays the groundwork for future hybrid supercomputers, where energy-efficient logic based on GCS technology interfaces with CMOS. To enable this, GCS devices must achieve lower operating voltages and higher output voltages—key for interconnection and complex circuit design.
We filed four patents during the project, with potential to advance superconducting logic and memory circuits. GCS-based memories address a longstanding challenge in superconducting electronics and may be essential for building fully superconducting HPC prototypes.
A start-up, Digital Superconducting Quantum Machine, was founded by SuperGate members and selected for the European Innovation Council’s TechtoMarket (T2M) program.
While the GCS effect may not be purely electric-field driven and remains at a low technological readiness level, these exploitation outcomes highlight the strong potential and impact of SuperGate’s innovations.
We filed four patents during the project, with potential to advance superconducting logic and memory circuits. GCS-based memories address a longstanding challenge in superconducting electronics and may be essential for building fully superconducting HPC prototypes.
A start-up, Digital Superconducting Quantum Machine, was founded by SuperGate members and selected for the European Innovation Council’s TechtoMarket (T2M) program.
While the GCS effect may not be purely electric-field driven and remains at a low technological readiness level, these exploitation outcomes highlight the strong potential and impact of SuperGate’s innovations.