The Reluctant Superconductor

The Meissner effect is the definition of superconductivity. A superconductor expels magnetic flux from its interior — currents circulate on the surface and cancel the applied field inside. This diamagnetic response is not a secondary feature. It is the test. If a material shows the Meissner effect, it is a superconductor. If it doesn't, it isn't.

Zhang and colleagues (arXiv:2603.25807, March 2026) imaged the Meissner effect in a rhombohedral graphene superconductor by mapping nanotesla-scale fringe fields in real space. They confirmed superconductivity. But the screening was almost negligible — the sample expelled roughly 100 parts per million of the applied magnetic field. A conventional superconductor expels all of it. This one expels almost none.

The superconductor is reluctant because it is also a magnet. Superconductivity in this material emerges during a continuous quantum phase transition into a canted spin ferromagnet. The same electrons that form Cooper pairs for superconductivity are simultaneously developing magnetic order. The two states — one that expels fields, one that generates them — coexist in the same electron system at the same temperature.

The superfluid stiffness — the energy cost of phase fluctuations in the superconducting order parameter — depends on temperature in a way that violates the predictions of BCS theory. In standard superconductors, stiffness drops exponentially near zero temperature as quasiparticle excitations freeze out. Here the drop is not exponential. And the zero-temperature stiffness is linearly proportional to the critical temperature, a relationship seen in cuprate high-temperature superconductors but not in conventional materials.

The material passed the test. It shows the Meissner effect, so it is a superconductor. But it barely passed. The magnetic order competing for the same electrons leaves almost nothing for flux expulsion. The superconducting state exists at the margin of what the definition requires — a phase that is technically present but functionally overwhelmed by its competitor. The measurement had to resolve nanotesla signals to see it at all.

The structural observation: the boundary between superconductor and not-superconductor is not a wall. It is a continuum, and this material sits at the edge — superconducting in principle, barely superconducting in practice, and interesting precisely because the competition between orders leaves the superconductivity almost undetectable.