Background: The Eight-Year Pairing Mystery
In March 2018, Yuan Cao and Pablo Jarillo-Herrero at MIT twisted two graphene layers by 1.1 degrees, producing flat bands and superconductivity at 1.7 K — a Nature cover story that named "magic-angle graphene." For eight years, one fundamental question remained: why do electrons pair? Conventional superconductors rely on lattice vibrations (phonons) as glue; magic-angle graphene, with its flat bands, strong correlations, and compressed phase diagram, looked convincingly "unconventional," like the cuprates. Neither camp could settle the debate.
Why Previous Screening Experiments Fell Short
The decisive idea sounds simple: place a tunable metal next to the superconducting layer to screen Coulomb interactions. If pairing is phonon-mediated, screening should make pairing easier and raise Tc. If pairing is driven by electron-electron interactions, screening removes the glue and Tc should collapse.
But Coulomb screening decays exponentially with distance, acting only over 1–2 nanometers. Previous attempts with graphite gates, separated by 6–10 nm of hBN (2020) or 3 nm (2021), left superconductivity essentially intact — the 2021 experiment even saw Tc rise 2–3%, misleadingly read as a conventional signature.
The Manchester Solution: Twist Instead of a Substrate
The Barrier team at Manchester's National Graphene Institute built a "twisted four-layer" device: a 1.15-degree magic-angle bilayer (superconducting) directly stacked — no spacer — against a 0.46-degree bilayer (screening layer), with an approximately 10-degree relative twist that misaligns the moiré momenta. Electrons do not tunnel between them, but Coulomb interactions pass through at atomic distance. hBN encapsulates only the outer surfaces.
When the screening layer was doped metallic, the superconducting dome continuously shrank, the critical current went to zero, and above a doping of about 2×10¹² cm⁻² superconductivity vanished entirely even at 35 mK — a Tc change of more than an order of magnitude. Correlated insulating states died simultaneously. Normal-state transport was unaffected, and control experiments ruled out parallel shorting, displacement field, and proximity effects as explanations.
The Verdict
The paper concludes carefully: "the screening behavior rules out conventional (BCS-like) Cooper pairing and supports an electronic pairing mechanism." Kaxiras, writing the accompanying APS Viewpoint, adds nuance: phonons may still play a stabilizing role, but are "unlikely to play a dominant role in pairing." Which specific electronic mechanism — plasmonic models can reproduce the collapse — remains open. The victory here belongs to clean elimination.
Notably, the 2021 result's 2–3% Tc rise at 3 nm spacing now appears to be an artifact of screening strength decaying to near zero over that distance. A parallel Nature Physics study by Gao et al. reached the same conclusion via a different route. And in the spirit of laboratory honesty: only two devices in this batch achieved magic angle, and only one superconducted — a yield the field has endured for eight years.
Geim himself told reporters he only cares about room-temperature superconductivity — "preferably at room temperature or above" — adding that "Rome wasn't built in a day." This twisted four-layer device is far from that dream, but it leaves behind a powerful tool: a screening knob that can be twisted flush against correlated materials. Which unconventional superconductor gets turned next?
References: