Graphene that superconducts harder when the magnetic field rises

Ordinary graphite, the material in pencil lead, is about as boring as a crystalline solid gets. Carbon atoms in flat sheets, stacked loosely, a conductor, a semimetal, the kind of material students use to learn about band structure. Zoom in, though, and the sheets can be stacked in different ways, and one of those ways, the rhombohedral stacking, has over the last few years turned out to host a surprising zoo of electronic states. A paper in Nature on 29 June, from Long Ju’s group at MIT with collaborators at Basel, Florida State and a Japanese national materials lab, adds a new entry to that zoo: a family of superconducting states in rhombohedral graphene, several of which get stronger rather than weaker when a magnetic field is turned on.

The experiment is a study in controlled delicacy. The team exfoliated flakes of natural graphite, searched for the rhombohedral four- and five-layer stacks, and shaped them into devices that could be tuned by gate voltage. In earlier work the group had found superconductivity when they doped electrons into the stack. In this new study they did the opposite, sweeping the electron density downward, and combined that with magnetic fields applied both parallel and perpendicular to the graphene plane, in a cryogenic set-up built with the Basel collaborators.

The results are strange. At four separate electron densities, the resistance dropped to zero: four distinct superconducting states. That alone would be a result, since superconductivity in a single material usually comes in one form. But three of the four states survived a parallel magnetic field of around nine tesla, roughly 180,000 times the Earth’s field, when a magnetic field of that strength should rip the Cooper pairs apart. Worse, or better, depending on your taste: at a particular density, switching the field to a perpendicular orientation did not destroy superconductivity. It enhanced it. The transition temperature climbed from about 55 millikelvin to around 90 millikelvin, and the critical current, the amount of supercurrent the material could carry before going normal, went up by another 50 to 60 per cent.

Magnetic fields normally kill superconductivity by pulling electron spins out of their paired, opposite-alignment configuration. The MIT team’s proposal is that in rhombohedral graphene, at the relevant densities, the electron pairs are not spin-opposite at all. They are aligned. A magnetic field then pulls both spins the same way, preserving rather than breaking the pair. That would explain why a parallel field, which acts on the spins without disrupting the orbital motion, lets the pairs survive at high field, and why a carefully aligned perpendicular field can even stabilize a more ordered pairing. It is a proposal, not yet a theory, and the authors are clear that the microscopic mechanism is still open.

What the paper establishes is the family, not the mechanism. Rhombohedral graphene is now a material that hosts chiral superconductivity, fractionalized charge and a small menagerie of field-tolerant superconducting states, all in the same four or five atomic layers of carbon. That is a remarkable amount of physics for a material you can scratch off the end of a pencil. The tunability is the point: gate voltage, magnetic field and density, each a knob that brings out a different state.

The caveats are the standard ones for this corner of condensed matter. The temperatures are millikelvin, far below anything anyone would call applied, and the samples are small, fragile and painstakingly found rather than grown. Nobody is making a superconducting quantum computer out of rhombohedral graphene next year. But a simple element, stacked in a slightly unusual way and tuned with a few knobs, is now showing superconductivity that strengthens in a magnetic field. That is the kind of result that sends theorists back to the whiteboard, and the whiteboard is where new physics usually begins.