Disorder generally suppresses superconductivity. Impurities scatter electrons, breaking the phase coherence that allows Cooper pairs to carry current without resistance. Anderson's theorem protects conventional superconductors against weak non-magnetic disorder, but beyond that threshold, more mess means less superconductivity. This is textbook.
Ślebarski and Maśka (2026) study quasiskutterudites — tin-based cage compounds doped with substitutional impurities — and find the opposite. Increasing disorder creates locally superconducting regions with critical temperatures T_c* that exceed the bulk transition temperature T_c. Upper critical field measurements reveal two distinct branches: one for the bulk phase, one for the local phase, providing direct experimental evidence for a percolative superconducting state. The enhancement peaks where entropy is maximized — where disorder is strongest.
The mechanism is a trade-off between two effects that disorder produces simultaneously. Locally, impurities modify the electronic environment in ways that strengthen the pairing interaction — the potential well deepens around each impurity site. Globally, the same impurities destroy the long-range phase coherence needed for bulk superconductivity. The result is an archipelago: isolated superconducting islands, each stronger than the continent they replaced, separated by normal metal. As disorder increases further, the islands percolate — connect through random proximity — creating a new kind of global superconductivity that is messier but locally more robust than the clean original.
The general principle: fragmentation can strengthen the fragments. When a global order depends on properties that compete with local optimization, disrupting the global order releases the local potential. The fragments are not pieces of a broken whole. They are a different state — one that could only emerge when the coherence that was holding them together was removed.