Conventional superconductivity begins with a normal metal. Electrons in the metal pair up at low temperatures through phonon-mediated attraction, forming Cooper pairs that flow without resistance. The starting point matters: the normal metal provides a Fermi sea of well-defined quasiparticles that can be destabilized into the superconducting state. Unconventional superconductors — the copper oxides, the iron pnictides — have resisted explanation partly because the assumption persists: superconductivity should emerge from something that looks approximately like a normal metal.
Published in Nature, Kim et al. combined scanning tunneling microscopy with theoretical modeling on twisted graphene moiré systems and showed that superconductivity in these materials does not arise from an ordinary metal at all. It emerges from an already strongly correlated state with broken symmetry — a spiral ordering of the valley degree of freedom, multiple energy gaps with temperature and magnetic field dependence, and evidence of a many-body Kondo resonance. The parent state is exotic before superconductivity even appears.
The structural insight is about the assumption embedded in the question. Asking “how does the normal metal become superconducting?” presupposes that the starting point is normal. If the starting point is already broken — already correlated, already ordered — then the transition to superconductivity is not a symmetry-breaking event from a symmetric state. It is a reorganization within an already broken landscape. The mechanism is different because the question was wrong. The parent state is not a metal with latent pairing instability. It is an interacting many-body system that happens to also superconduct.
This is the first direct microscopic connection between the correlated insulating states and superconductivity in moiré systems. The two were known to be neighbors in the phase diagram. Now they are shown to be parent and child.