High-temperature superconductivity has been one of physics' hardest problems for four decades. The central question: what pairs the electrons? In conventional superconductors, lattice vibrations (phonons) do the pairing — atoms vibrate, creating an attractive interaction between electrons that allows resistance-free current. In high-temperature superconductors, the pairing mechanism has resisted identification despite thousands of papers, multiple Nobel Prizes worth of experimental infrastructure, and some of the most sophisticated computational methods in physics.
The standard approach treats the atomic lattice as classical scaffolding. The Born-Oppenheimer approximation — developed in 1927 — separates nuclear and electronic motion: electrons are quantum, nuclei are classical. Nuclei vibrate, but their positions are well-defined. The phonon description follows from this separation. Every first-principles calculation of superconducting properties uses some version of this assumption.
Zhu, Zeng, and Li (2602.03576) abandoned it. They treated the nuclear positions as fully quantum — delocalized, correlated, exhibiting many-body entanglement. In hydrogen sulfide (H₃S) under extreme pressure — a room-temperature superconductor at 150 GPa — the result was striking. A new phase emerged: “lattice quantum disorder,” where atomic positions are fundamentally uncertain, not just thermally smeared.
The lattice quantum disordered phase coincides exactly with the peak superconducting temperature. The pressure at which nuclear quantum delocalization is maximized is the pressure at which superconductivity is strongest. Not nearby. Not correlated. Coincident.
Decades of theory focused on electrons because the approximation made the nuclei invisible. Not wrong — the Born-Oppenheimer approximation works extraordinarily well for most purposes. But when an approximation works well enough for long enough, it stops being recognized as an approximation and becomes the floor you stand on. Nobody questions the floor. The lattice was scaffolding: essential to the structure, invisible to the inhabitants.
The finding isn't that the Born-Oppenheimer approximation fails. It's that the failure mode is invisible from within the approximation. The lattice quantum disordered phase doesn't appear in any phonon calculation because phonons presuppose classical nuclear positions. The answer was in the variable that had been approximated out of existence.
This is a general pattern. When the simplification that enables your field becomes invisible, the phenomena it hides become undiscoverable — not because they're complex, but because the tools to see them don't exist within the standard framework. The scaffolding determines what the building can look like, and after enough time, nobody remembers the scaffolding is there.