A graphene Cooper pair splitter can use a temperature difference—not an applied electrical bias—to generate current between two quantum dots linked by a superconductor. In an experiment using graphene quantum dots and an aluminum superconducting lead, researchers observed a nonlocal Seebeck effect and found that gate voltages could tune the contributions of two distinct transport processes: Cooper pair splitting and elastic cotunneling.
The result is relevant to engineers and technical decision-makers following quantum-device development because it demonstrates a way to control coherent electron transport through thermal drive. It is a foundational physics result, not an industrial power-generation technology. Its practical significance lies in the device architecture, measurement approach, and potential to support experiments on electron entanglement and heat transport.
How a graphene Cooper pair splitter works
The device consists of two graphene quantum dots connected to a common aluminum superconductor. Each dot is also connected to a normal graphene lead, and independently controlled side gates adjust the dots’ energy levels. A separate resistive heater creates a temperature difference between the left and right sides of the structure.
In a conventional Seebeck effect, a temperature gradient across a material produces an electrical response. Here, the measured currents include both local thermoelectric contributions within each dot and smaller nonlocal contributions involving transport through the shared superconducting region. The experiment isolates the nonlocal component by comparing current measurements as the gate voltage on the opposite dot is varied.
The distinction between local and nonlocal current matters: a current detected at one dot may depend on the temperature and electronic transmission of the other dot. That dependence provides evidence that the two sides are coupled through the superconducting structure rather than acting as independent thermoelectric devices.
Two competing paths for nonlocal transport
Two coherent processes contribute to the nonlocal current:
- Elastic cotunneling (EC): An electron moves from one normal lead to the other through the superconductor, without entering the superconductor as part of a Cooper pair.
- Cooper pair splitting (CPS): The two electrons forming a Cooper pair are separated, with one transferred toward each normal lead.
Both processes depend on the energy-dependent transmission probabilities of the quantum dots. The side gates shift the dots’ resonant energy levels, changing how readily electrons at different energies pass through each dot. Because CPS and EC respond differently to those energy-dependent transmissions, adjusting the gate voltages changes their relative contributions to the thermoelectric signal.
The researchers compared currents measured in the left and right dots. Their sum and difference provide combinations associated with CPS and EC, respectively. The observed gate-voltage patterns and current magnitudes agreed with the coherent transport model, supporting the interpretation that both mechanisms contribute to the nonlocal thermoelectric response.
What the measurements establish
The experiment showed that a thermal gradient can drive a nonlocal current in a graphene quantum-dot–superconductor–quantum-dot structure. The nonlocal currents were smaller than the dominant local thermoelectric currents, making their separation and interpretation an important part of the measurement. The gate-dependent sign and magnitude changes, together with comparison to theoretical predictions, helped distinguish the nonlocal contribution.
The device used a graphene heater, two graphene-based superconducting junctions as thermometers, and a substrate that carried heat from the heater toward the splitter. The heater was electrically isolated from the splitter, helping separate thermal driving from direct electrical bias. Thermometer calibration was used to estimate the temperature conditions at the device.
Local thermoelectric behavior also provided useful context. At higher drive levels, the measured local current broadly followed the quantum-dot conductance, and a coherent model described much of the response. At lower temperatures, however, some additional sign changes were not captured by that model. An incoherent transport model, which accounts for thermally generated voltages and their circuit back-action, helped describe those features. This distinction underscores that local and nonlocal signals can involve different transport regimes.
Why thermal control matters for quantum-device research
Using a temperature gradient as the drive offers a way to examine how heat flow and coherent charge transport interact in a hybrid normal-metal–superconductor system. Since the two dots can be tuned separately, the architecture allows researchers to explore conditions where CPS or EC makes a stronger contribution. That tunability makes the platform useful for testing theoretical ideas about the relationship between entanglement and heat transport.
The work does not, by itself, establish a ready-to-deploy source of entangled electrons or a practical thermoelectric generator. Rather, it demonstrates a controllable mechanism that could support future experiments. Further work would be needed to assess its suitability for applications and to establish how reliably particular operating regimes can be prepared and verified.
Practical implications
For teams developing quantum measurement systems, the study highlights several design considerations: independently tunable quantum dots, careful thermal-gradient measurement, electrical isolation of the heater, and analysis that separates local signals from nonlocal transport. It also shows why models may need to account for both coherent transport and thermal or circuit-related effects, depending on operating conditions.
More broadly, the graphene Cooper pair splitter demonstrates how thermal drive can be used to investigate and tune electron transport in a superconducting hybrid device. Its immediate value is as a research platform for studying nonlocal thermoelectricity, Cooper pair splitting, and elastic cotunneling—not as an established industrial component. For quantum-technology developers, it offers a useful example of how material choice, gate control, and thermal engineering can be combined to probe complex transport behavior.