Discovery of an electrically-controllable superconducting memory effect

This study demonstrates that bulk single crystals of the triplet superconductor uranium ditelluride (UTe$_2$) exhibit an intrinsic, electrically controllable memory effect where magnetic-field-tuned current pulses switch the material between metastable states of distinct critical currents, offering a promising pathway for ultralow-power superconducting memory in cryogenic computing.

The Gist

Imagine you have a super-fast, super-efficient computer that runs on ice-cold temperatures. To make it work, you need a way to store information (like a 1 or a 0) without using electricity to keep it there. Usually, computers use magnetic fields or electricity to flip switches, but these processes waste energy as heat.

Scientists have been trying to build a computer made entirely of superconductors—materials that conduct electricity with zero resistance. The problem? While we can make superconductors act like diodes (one-way streets for electricity), we haven't been able to make them act like memory (a switch that remembers its state) without using messy, energy-wasting magnetic parts.

This paper reports a breakthrough: they found a way to make a superconductor "remember" its state using only electricity.

The Star of the Show: UTe₂

The material they used is called Uranium Ditelluride (UTe₂). Think of this material as a very special, complex dance floor.

• The Dance Floor: It's a crystal where electrons dance together in pairs (superconductivity).
• The Two Styles: Depending on how strong a magnetic field you apply, the electrons dance in two different styles. Let's call them Style A and Style B.
• The "Gray Area": There is a specific zone where Style A and Style B are fighting for control. This is the "memory zone."

The Magic Trick: The "Pump" and the "Reset"

The scientists discovered that by giving this material a quick, sharp jolt of electricity (a "pulse"), they could force the electrons into a new arrangement that they stay in, even after the electricity stops.

Here is the analogy:
Imagine a room full of people (the electrons) trying to stand in neat, orderly rows (the normal state). This is easy and requires little effort.

1. The Pump (Writing the Memory): If you suddenly shout "FREEZE!" and push everyone hard (a sharp electrical pulse), the people scramble and get stuck in a chaotic, messy pile.
2. The Memory State: Even after you stop shouting, the people stay in that messy pile. It's harder to move them now because they are tangled up. In physics terms, this "messy pile" has stronger pinning forces, meaning it resists moving electricity more effectively. This is the "1" state (High Critical Current).
3. The Probe (Reading the Memory): You can check if they are in the messy pile or the neat rows by sending a gentle current. If they are in the messy pile, the current flows differently.
4. The Erase (Resetting): If you gently guide the people back to their neat rows (a slow, smooth decrease in current), they return to the easy, low-energy state. This is the "0" state.

Why is this a Big Deal?

1. It's "Intrinsic": Previous attempts to make superconducting memory required sticking a magnet or a semiconductor next to the superconductor. It was like trying to make a car run by attaching a horse to the front. This new discovery is like the car engine itself remembering how to run. The memory is built into the material's own nature.
2. Ultra-Low Power: Because superconductors have zero resistance, keeping this "memory" state requires almost no energy. The paper suggests that a tiny version of this could use one billion times less energy than the memory chips in your phone.
3. The "Vortex" Explanation: The scientists believe this happens because of "vortices" (tiny whirlpools of magnetic field) inside the material.
   • In the normal state, the whirlpools are neatly organized.
   • The "pump" jolts them into a chaotic, glass-like mess.
   • This mess is actually better at holding the current in place (higher critical current), creating the memory effect.
   • The "reset" gently smooths the whirlpools back into order.

The Future

This discovery is like finding a new type of switch that works at the speed of light but uses almost no battery. It could revolutionize:

• AI: Training AI models currently eats up massive amounts of electricity. Superconducting memory could make this sustainable.
• Quantum Computers: These computers need to run at near-absolute zero. This memory effect works perfectly at those temperatures, potentially helping us build better quantum hardware.

In short: The scientists found a way to make a superconductor "remember" if it was just jolted or gently cooled down, using the chaotic dance of its internal particles. It's a tiny, invisible switch that could power the super-computers of the future without melting the planet.

Technical Summary

Here is a detailed technical summary of the paper "Discovery of an electrically-controllable superconducting memory effect."

1. Problem Statement

The increasing energy demands of artificial intelligence and data processing necessitate more sustainable computing paradigms. Superconducting electronics offer a path to ultra-low-power computing by eliminating resistive dissipation. While recent advances have achieved superconducting diodes and switching mechanisms, a critical gap remains: the lack of electrically-controllable, non-volatile memory that relies exclusively on superconducting elements.

• Current Limitations: Existing superconducting memory concepts often rely on hybrid ferromagnet-superconductor heterostructures (introducing magnetic dissipation) or complex lithographic vortex-trapping geometries (requiring external fields or high currents).
• Goal: To discover an intrinsic superconducting memory effect that can be written, read, and erased using only electrical stimuli (current pulses) without external magnetic fields or magnetic interfaces.

2. Methodology

The researchers investigated the electrical transport properties of Uranium Ditelluride (UTe$_2$), a candidate for a spin-triplet superconductor known for its multiphase superconductivity.

• Material & Conditions: Bulk single-crystal specimens of UTe$_2$ were studied at ultra-low temperatures (down to 50 mK) under high magnetic fields (up to 30 T).
• Phase Space: Experiments focused on the intermediate regime (termed SC1.5) where two distinct superconducting phases (SC1 and SC2) coexist. This regime occurs when the magnetic field ($B$) is aligned along the hard crystallographic $\hat{b}$-axis, typically between 7 T and 20 T.
• Experimental Protocol:
  • Pump/Probe: The team applied direct current (DC) pulses to perturb the system.
  • Write (Pump): A large current density ($J_{ex}$) was applied and then abruptly switched off (fast relaxation rate).
  • Read (Probe): The current was smoothly swept to measure the Voltage-Current ($V-J$) characteristics and determine the critical current density ($J_c$).
  • Erase (Reset): The system was reset to its equilibrium state by smoothly ramping the current from a high value to zero.
• Tuning Parameters: The memory effect was tuned by varying the amplitude, duration, and relaxation rate of the excitation current.
• Modeling: A modified Resistively-Shunted-Junction (RSJ) model was employed to simulate the dynamics of the superconducting phase, treating the memory effect as a transition between potential energy minima.

3. Key Contributions

• Discovery of Intrinsic Memory: The paper reports the first observation of an intrinsic, electrically-controllable superconducting memory effect in a bulk material without magnetic or semiconducting interfaces.
• Mechanism Identification: The authors propose that the memory effect arises from the competition between two distinct vortex species native to the coexisting SC1 and SC2 phases.
• Non-Equilibrium Control: They demonstrate that the system can be "quenched" into a metastable, high-disorder vortex configuration (high $J_c$) or "annealed" back to an ordered equilibrium state (low $J_c$) purely through the timing and shape of electrical current pulses.

4. Key Results

• Hysteretic $V-J$ Characteristics:
  • In the SC1.5 regime, applying a "pump" pulse (abrupt current drop) shifts the system into a metastable state with a significantly enhanced critical current ($J_c$).
  • The system "remembers" this high-$J_c$ state for extended periods (hours), exhibiting a distinct hysteresis loop in the $V-J$ curve compared to the equilibrium state.
  • A smooth "erase" ramp returns the system to the original low-$J_c$ equilibrium state.
• Tunability: The magnitude of the memory effect (the difference in $J_c$ between states) is controllable by:
  • Amplitude: Higher excitation currents yield larger hysteresis.
  • Duration: Longer pulses increase the effect up to saturation.
  • Relaxation Rate: Faster current turn-off (quenching) is crucial for trapping the system in the memory state; slow relaxation allows the system to anneal back to equilibrium.
• Phase Space Mapping:
  • The memory effect is strictly confined to the SC1.5 coexistence region (approx. 7 T to 20 T at 50 mK).
  • It vanishes in pure SC1 (low field) and pure SC2 (high field >20 T) phases.
  • The effect is highly sensitive to the angle of the magnetic field, disappearing when the field is tilted beyond the SC2 stability range ($\theta \approx 20^\circ$).
• Temperature Dependence: The effect is robust at very low temperatures (50 mK) but diminishes rapidly above 0.6 K, despite the superconducting $T_c$ being 2.1 K. This suggests the mechanism is a low-energy phenomenon distinct from the superconducting gap itself.
• Modeling Validation: The RSJ model, featuring a double-well potential energy landscape ($U(\phi)$), successfully reproduces the experimental data. The "memory state" corresponds to the system being trapped in a local, glassy minimum ($\phi = \pi$) representing a disordered vortex domain structure, while the "erased state" corresponds to the global minimum ($\phi = 0$).

5. Significance and Outlook

• Technological Impact: This discovery enables the potential for ultra-low-power cryogenic memory. The estimated power consumption for reading a memory bit is $\sim 1$ nW for bulk samples, potentially dropping to $\le 1$ pW for mesoscopic devices—orders of magnitude lower than silicon-based memory.
• Cryogenic Computing: Unlike previous superconducting switches that require transitioning to a normal (resistive) state, UTe$_2$ allows read/write operations while remaining superconducting, a critical advantage for energy efficiency.
• Neuromorphic and Quantum Applications: The plasticity of the effect (tunable hysteresis) makes it a candidate for cryogenic neuromorphic computing. Furthermore, the underlying physics involves exotic vortex matter (potentially half-quantum vortices or non-singular vortices) which may host Majorana fermions, linking this memory effect to topological quantum computing.
• Future Directions: The work suggests that other multiphase superconductors (e.g., CeRh$_2$As$_2$, UPt$_3$) may exhibit similar effects, opening a new avenue for engineering superconducting memory elements based on intrinsic vortex dynamics rather than extrinsic heterostructures.