Active electron cooling of graphene

In the emergent field of quantum technology, the ability to manage heat at the nanoscale and in cryogenic conditions is crucial for enhancing device performance in terms of noise, coherence, and sensitivity. Here, we demonstrate the active cooling and refrigeration of the electron gas in a graphene thermal transistor, by taking advantage of nanoscale superconductive tunnel contacts able to pump or extract heat directly from the electrons in the device. Our prototypes achieved a top cooling of electrons in graphene of about 15 mK at a bath temperature of about 450 mK, demonstrating the viability of the proposed device architecture. Our experimental findings are backed by a detailed thermal model that accurately replicated the observed device behavior. Alternative cooling schemes and perspectives are discussed in light of the reported results. Finally, our graphene thermal transistor could find application in superconducting hybrid quantum technologies.

💡 Research Summary

This paper presents the first experimental demonstration of active electron cooling and refrigeration in graphene, a milestone for managing heat dissipation in nanoscale quantum devices operating at cryogenic temperatures.

The core challenge addressed is the critical need to reduce electronic noise and quasiparticle poisoning in quantum technologies, such as qubits and sensitive detectors, by locally lowering the electron temperature. While superconducting tunnel junctions (SINIS coolers) have been used to cool normal metals and superconductors, applying this technique to field-tunable semiconductors like graphene had remained elusive due to difficulties in fabricating high-quality, low-resistance tunnel contacts.

The authors overcame this hurdle by developing a novel hybrid fabrication process. They combined standard techniques for patterning exfoliated monolayer graphene with angle-resolved shadow mask evaporation, a method typically used for all-metallic devices. This allowed them to create a device where a graphene flake is connected via clean galvanic contacts to a metallic source and drain electrode. The source electrode is equipped with superconducting aluminum tunnel coolers, while the drain electrode features similar tunnel junctions acting as sensitive thermometers.

In this architecture, applying a voltage bias (V_cool) to a pair of coolers extracts (or injects) heat from the electrons in the source via the energy-filtering property of superconducting gaps. This temperature change diffuses efficiently through the highly conductive graphene sheet and is detected non-locally by measuring the electron temperature (T_D) at the drain via the thermometer junctions.

The experiments revealed two distinct operational regimes depending on the substrate phonon bath temperature (T_b). At higher T_b (e.g., 448 mK), the device achieved true cooling, where T_D dropped below T_b by up to ~3 mK. At intermediate T_b (e.g., 307 mK), it demonstrated refrigeration, actively reducing T_D from a maximum value by a best value of (15.5 ± 0.5) mK, even though T_D remained above T_b.

A significant achievement of this work is the development of a comprehensive thermal model that accounts for all major heat exchange mechanisms in the complex structure: heat pumping via tunnel coolers, electron-phonon coupling in each element (source, graphene, drain), and thermal diffusion between them. This model not only reproduced the experimental T_D(V_cool) curves with remarkable accuracy but also enabled the extraction of the otherwise inaccessible electron temperatures within the graphene sheet (T_G) and the source electrode (T_S). The model revealed that a ~65 mK cooling at the source translated to a ~15 mK cooling in graphene and a ~3 mK cooling at the drain, demonstrating the efficiency of heat diffusion and the device’s capability as a graphene electron cooler.

The study concludes that non-local electron cooling via superconducting tunnel junctions is a viable and promising strategy for graphene. This breakthrough paves the way for integrating active thermal management into future graphene-based superconducting hybrid quantum technologies, such as coherent electronics, qubits, and radiation detectors, potentially enhancing their performance by reducing noise and quasiparticle-induced decoherence.