Copper sulfur: the death of a superconductor claim

In July 2023 a small South Korean team posted preprints describing a lead-copper apatite, LK-99, that they claimed was a superconductor at room temperature and ambient pressure. A video of a flake levitating over a magnet travelled around the world within days, and laboratories on several continents spent the following weeks trying to reproduce the result. By August, groups at the Institute of Physics of the Chinese Academy of Sciences had already pointed at a suspicious suspect: copper sulfide, a common impurity in the synthesis, which undergoes its own sharp resistivity change near the temperature where LK-99 seemed to turn superconducting. On 24 November 2023 that team published the full account in Matter, titled ‘First-order transition in LK-99 containing Cu2S’, and the claim was, in the careful language of materials science, closed.

The mechanism is a phase transition, not superconductivity. Pure copper(I) sulfide, Cu2S, shifts from a hexagonal high-temperature phase to a monoclinic low-temperature phase near 385 kelvin, about 112 degrees Celsius. Across that transition its resistivity drops abruptly, exactly the signature the Korean team had read as the onset of superconductivity. The Chinese team showed that the resistivity of their LK-99 samples tracked the Cu2S transition, that the samples contained copper sulfide in precisely this temperature range, and that no zero resistivity, the defining requirement of a superconductor, appeared anywhere in the data. A first-order structural transition in an impurity can mimic a superconducting transition in its electrical behaviour while being an entirely different physical phenomenon.

The elegance of the work is in the control. Rather than argue about what LK-99 might be, the authors measured the impurity phase directly, compared its transition temperature with the electrical anomalies reported in the original claim, and showed that the bend in the resistance-versus-temperature curve followed the copper sulfide transition rather than the material of interest. A parallel replication study posted to arXiv in November reached the same conclusion: the anomalies in LK-99 are attributable to phase transitions in the Cu2S impurity, and LK-99 itself is not a room-temperature ambient-pressure superconductor.

The episode is worth recording as a case study in how a materials claim dies. The original evidence was thin even at the start: partial levitation rather than full Meissner repulsion, a resistive drop that was not a zero-resistance plateau, and samples whose purity the authors had not fully characterized. What was surprising was not that it failed but how quickly it failed. Within six weeks, the worldwide community had isolated the impurity, identified its transition, and reproduced the mimicry, which is roughly how long a credible claim of that size should take to test.

What the paper leaves standing is more interesting than what it knocks down. Apatite-based chemistry with copper substituted into lead sites remains a legitimate playground for condensed matter, and the synthesis route the Korean team used is now better characterized than it was. The field’s next room-temperature superconductor claim will, one hopes, arrive with zero-resistance measurements on pure samples, not with a video of a flake. For the moment, the room-temperature superconductor is again something no one has, and the known culprit for the summer’s excitement is a 385-kelvin phase transition in copper sulfide.