Researchers Suppress Superconductivity in Magic-Angle Graphene, Solving Decade-Old Mystery

A decade of debate about how magic-angle graphene becomes superconducting has ended with an experiment that simply turns the effect off. Researchers at the University of Manchester's National Graphene Institute, working with teams from Singapore, St. Louis, Philadelphia, Antwerp, and Japan, have demonstrated that the superconductivity vanishes entirely when they screen the electrical interactions between electrons in the material. The finding settles a fundamental question that has divided the field since magic-angle graphene's superconducting properties were first discovered: does the material owe its behavior to electrons pushing against each other, or to vibrations in the atomic lattice coupling them together?

Magic-angle graphene is made by taking two sheets of graphene and rotating one relative to the other by about 1.1 degrees—a tiny angle that produces outsized effects. The material has become one of the most intensely studied quantum systems of the past decade, precisely because its superconductivity seemed to emerge from nowhere, defying easy explanation. Two competing theories emerged. One held that electron-electron interactions—the electromagnetic repulsion between negatively charged particles—somehow drove the pairing that creates superconductivity, a mechanism analogous to what happens in high-temperature cuprate superconductors. The other camp argued for a more conventional explanation: that vibrations in the crystal lattice, called phonons, were doing the work, coupling electrons into pairs in the way phonons do in ordinary superconductors like lead or mercury.

Previous attempts to distinguish between these mechanisms had placed screening layers near the magic-angle graphene to weaken the Coulomb repulsion between electrons. But those layers sat several nanometers away, separated by an insulating spacer. The result was underwhelming: the correlated-insulator states in the material could be suppressed, but superconductivity remained largely intact, or shifted by only 2 to 3 percent. The screening was too distant to matter much.

The Manchester team, led by Julien Barrier with corresponding authors André Geim and Alexey Berdyugin, took a different approach. Instead of inserting a dielectric spacer, they stacked two twisted graphene bilayers directly on top of each other, separated by less than a nanometer. The two layers remained electronically decoupled because they were twisted at a large angle relative to each other, but the geometry allowed one bilayer—tuned to a small twist angle to create a high density of states—to act as an electrically tunable screening layer positioned immediately adjacent to the magic-angle bilayer. As the researchers increased the carrier density in the screening layer, both the superconductivity and the correlated-insulator states in the adjacent magic-angle graphene were progressively suppressed. The superconducting critical temperature fell by more than an order of magnitude. At sufficiently high screening densities, superconductivity vanished entirely—a far larger effect than any previous screening experiment had achieved.

This result contradicts what conventional phonon-mediated superconductivity would predict. If phonons were responsible for the pairing, screening the Coulomb repulsion between electrons should leave phonon-coupled pairing unchanged or even slightly enhance it. Instead, the opposite happened: screening destroyed superconductivity. When the researchers compared their experimental data against theoretical models of several candidate mechanisms, they found the observed suppression consistent with pairing driven by electron-electron interactions—possibly through electron-plasmon coupling—rather than by phonons. The authors are careful to note that their work does not identify a single definitive unconventional mechanism; several remain consistent with the data. But the broad conclusion is clear: electron-electron interactions, not atomic vibrations, drive the superconductivity in magic-angle graphene.

Barrier explained the technical challenge: "To make a difference, we had to solve two issues. First, to build a device in which the screening layer sits extremely close, a fraction of a nanometer, to the superconducting graphene while remaining electronically separate. Second, we had to make that screening layer tunable. To this effect, we used a twisted graphene bilayer in atomic contact to the magic-angle graphene." Berdyugin reflected on the surprise of the result: "When we switched on the screening, we were surprised to find that superconductivity was completely suppressed. This provides clear experimental evidence that superconductivity in this system originates from strong electron-electron interactions." He added that the sub-nanometer screening technique, working over distances as short as 0.3 nanometers, could help clarify other debated phenomena in materials with strong electronic interactions.

Geim, who won the Nobel Prize for his work on graphene, placed the finding in a larger context. "Personally, I am interested only in high-temperature superconductivity—preferably at room temperature or above," he said. "This study was done at temperatures so low that even helium turns liquid. But unless we understand what makes superconductivity work, we are unlikely ever to reach room-temperature superconductivity, let alone make this remarkable phenomenon commercially useful. Our study takes only a tiny step—but still a step—in that direction, helping to nail down the mechanism of exotic superconductivity in graphene." The work does not solve the larger puzzle of room-temperature superconductivity, but it removes one layer of uncertainty from the path toward it.