Why a room-temperature superconductor claim is not yet a material
A reader asked why a room-temperature superconductor claim is not yet a material. The short answer: a claim is one measurement, on one sample, in one lab. A material is a recipe that other people can follow, and get a sample they can identify, that shows the two signatures of superconductivity: a resistance that falls to exactly zero, and a magnetic field pushed out of its inside. Every claim of superconductivity at room temperature and ordinary pressure so far has failed somewhere on that road. As of 2026, none has reached the end of it.
What a superconductor has to show
The first signature is that the resistance does not just get small. It goes to zero. A current started in a ring of superconductor keeps flowing with no battery, because nothing in the metal takes energy out of it. The trouble is that "zero" is a measurement, and every meter has a floor. A sample whose resistance has dropped a thousandfold can read zero on a meter that is not sensitive enough to tell it apart from copper.
The second signature is the Meissner effect. Cool a superconductor through its transition while it sits in a magnetic field, and it pushes the field out of its interior. A perfect conductor that was not a superconductor would only trap whatever field it already held. That expulsion is why a magnet can float steadily above a cooled superconducting disc, which has its own page here.
Then come the cross-checks. A magnetic field weakens superconductivity, so in a field the drop in resistance should move to a lower temperature. In the conventional kind, where vibrations of the crystal pair the electrons up, swapping an element for a heavier isotope should shift the transition too, because heavier atoms vibrate more slowly. And the check that matters most: a second lab, making its own sample of the same compound, has to see the same thing.
Where the real numbers sit
Temperatures here are in kelvin, counted up from absolute zero; a room at 20 °C is about 293 K. For thirty-three years the best superconductor at ordinary pressure was a ceramic of mercury, barium, calcium, copper and oxygen called Hg-1223, which loses its resistance at 133 K (−140 °C). Squeezed to about 31 billion pascals, around 300,000 atmospheres, the same compound reached 164 K.
Higher temperatures have come only from crushing hydrogen-rich compounds between the tips of two diamonds. In 2015 Alexander Drozdov, Mikhail Eremets and colleagues in Mainz measured hydrogen sulfide squeezed to well over a million atmospheres, where, they argued, it turns into a compound with three hydrogens per sulfur. It lost all resistance at 203 K, the drop moved down in a magnetic field, and the deuterium version had a lower transition, as the theory of conventional superconductors predicts. In a paper posted in late 2018 and published in 2019, the same group reported lanthanum hydride at about 250 K (−23 °C) under about 170 billion pascals, 1.7 million atmospheres. The hydride results have had critics. At the end of 2024 fifteen physicists who had never worked directly on hydrides reviewed the evidence and concluded it is real, largely because the sulfur hydride transition came out the same in Mainz, in measurements on a Mainz sample at the US national magnet laboratory by a team otherwise unconnected to Mainz, and in a sample made by a different route in Bristol.
One room-temperature claim at high pressure is still open. In 2025 Yinggang Song, Yanming Ma and colleagues at Jilin University reported a hydride of lanthanum and scandium that lost its resistance at up to 298 K, room temperature, under about 260 billion pascals, 2.6 million atmospheres. In some runs its resistance reached zero, the drop moved down in a magnetic field, and its crystal structure matched one the same group had predicted. In 2026 Dmitrii Semenok, Viktor Struzhkin and colleagues tried to make it from a lanthanum-scandium alloy at similar pressures and found no trace of superconductivity between 245 and 300 K. They suggest the way the starting metal is prepared may matter. Until a second lab sees it, it is a claim.
In March 2026 Liangzi Deng, Ching-Wu Chu and colleagues at the University of Houston moved the ordinary-pressure record for the first time since 1993. They squeezed Hg-1223 to between 10 and 30 billion pascals, cooled it to 4.2 K, and let the pressure off fast. The squeezed form, with its higher transition, stayed behind, and the best samples kept a transition as high as 151 K at ordinary pressure: 18 K better than before, and still about 140 K short of a room. The gain is fragile. Warming a sample back up to room temperature lowered its transition by a few kelvin.
Case one: LK-99
In July 2023 Sukbae Lee, Ji-Hoon Kim and colleagues at a Seoul start-up posted two preprints saying that LK-99, a lead phosphate crystal with some of its lead swapped for copper, superconducts at ambient pressure up to at least 400 K (127 °C). They reported a sharp drop in resistance, and a sample that levitated over a magnet, which they took to be the Meissner effect. Labs around the world tried to make it within weeks.
The clue was the temperature of the drop. Making LK-99 by the published recipe leaves copper sulfide behind as an impurity, and copper sulfide changes its crystal structure near 385 K. When it does, its resistance falls by a factor of a thousand to ten thousand. Shilin Zhu, Jianlin Luo and colleagues in Beijing measured LK-99 containing copper sulfide and found a sharp drop that looked like superconductivity, but the resistance did not reach zero, and the drop came at a different temperature on cooling than on warming. A superconducting transition in zero field comes at the same temperature both ways. A change of crystal structure has that lag. Then Pascal Puphal, Bernhard Keimer and colleagues at the Max Planck Institute in Stuttgart grew pure single crystals of the compound with no copper sulfide in them. The crystals were transparent and highly insulating, and their magnetism from 2 K to 800 K showed no sign of superconductivity or of any transition at all.
Interactive Run the three checks yourself on two samples: a real superconductor, and a sample whose drop comes from an impurity changing its crystal structure. Pick the sample, the direction of the temperature sweep, the magnetic field and how fine the meter is. Every control is a button or slider, so Tab and the arrow keys work too.
A toy model, not measured data. The 133 K and 385 K positions follow the page; the curve shapes, the 1,000-fold drop, the cooling-warming gap and the size of the field shift are drawn to show the logic of each check.
Case two: lutetium hydride
In March 2023 Nature published a paper from Ranga Dias's group at the University of Rochester claiming that lutetium hydride with a little nitrogen in it superconducts at 294 K, room temperature, under about 1 billion pascals, around 10,000 atmospheres. That pressure is high, but far below what other hydrides need. Xue Ming, Hai-Hu Wen and colleagues in Nanjing made the compound, confirmed that its crystal structure matched, and measured it from 0.4 to 40.1 billion pascals. It behaved as an ordinary metal down to 2 K, and magnetic measurements at 1 and 2.1 billion pascals found no sign of the field being pushed out.
In November 2023 Nature retracted the paper at the request of eight of its eleven authors, who said it did not accurately describe where the samples came from, what was measured, or how the data were processed. The journal's own investigation found separate concerns about the resistance data credible. An investigation for the university concluded that Dias had fabricated and falsified data. An earlier Nature paper from the same group, on a compound of carbon, sulfur and hydrogen, had been retracted in 2022.
What would count
A claim becomes a finding when the recipe is published, other labs make the same compound and show that it is the same, the resistance is measured finely enough to tell zero from merely small, the field is pushed out, the transition moves the right way in a field, and all of that holds in more than one place. Neither 2023 claim got past the second lab.
A finding is still not a material in the useful sense. A superconductor in a cable or a magnet has to carry large currents while sitting in a strong field, and has to be made by the tonne, not the microgram. Hydrides that need a million atmospheres are a long way from that, and Hg-1223's 151 K lives in specks that came out of a diamond press. In 2026 a group of sixteen researchers set out a plan for the search. They argue that no law of physics rules out a superconductor at room temperature, and that a central problem is that most compounds a computer predicts to be good superconductors cannot actually be made.
In one breath
A superconductor must show resistance that is truly zero and a field pushed out, and another lab must see the same thing in the same compound. The established records are 151 K at ordinary pressure and about 250 K at more than a million atmospheres; a 2025 claim of 298 K at 2.6 million atmospheres has not yet been seen by a second lab. LK-99's drop was an impurity changing its crystal structure; the lutetium hydride paper was retracted, and an investigation for the University of Rochester found that the group's leader, Ranga Dias, had fabricated and falsified data. A room-temperature claim is not yet a material because, so far, none has survived the second lab.
Where this comes from
- Conventional superconductivity at 203 K at high pressures linked only, not reproduced
arxiv.org/abs/1506.08190 - Superconductivity at 250 K in lanthanum hydride under high pressures linked only, not reproduced
arxiv.org/abs/1812.01561 - Hydride superconductivity: here to stay reuse permitted with attribution
arxiv.org/abs/2411.10522 - Superconductor Pb_10-xCu_x(PO_4)_6O showing levitation at room temperature and atmospheric pressure and mechanism reuse permitted with attribution
arxiv.org/abs/2307.12037 - First order transition in Pb_10-xCu_x(PO_4)_6O (0.9<x<1.1) containing Cu_2S linked only, not reproduced
arxiv.org/abs/2308.04353 - Single crystal synthesis, structure, and magnetism of Pb_10-xCu_x(PO_4)_6O reuse permitted with attribution
arxiv.org/abs/2308.06256 - Absence of near-ambient superconductivity in LuH_2pmxN_y linked only, not reproduced
arxiv.org/abs/2303.08759 - Room-Temperature Superconductivity at 298 K in Ternary La-Sc-H System at High-pressure Conditions linked only, not reproduced
arxiv.org/abs/2510.01273 - Stability Analysis of Superconductivity in P6/mmm-LaSc_2H_24 and its Experimental Reproducibility from La-Sc Alloys linked only, not reproduced
arxiv.org/abs/2605.29985 - Ambient-pressure 151-K superconductivity in HgBa2Ca2Cu3O8+δ via pressure quench linked only, not reproduced
www.pnas.org/doi/10.1073/pnas.2536178123 - First-order transition in LK-99 containing Cu2S linked only, not reproduced
doi.org/10.1016/j.matt.2023.11.001 - LK-99 isn't a superconductor: how science sleuths solved the mystery linked only, not reproduced
www.nature.com/articles/d41586-023-02585-7 - Retraction Note: Evidence of near-ambient superconductivity in a N-doped lutetium hydride linked only, not reproduced
www.nature.com/articles/s41586-023-06774-2 - Exclusive: official investigation reveals how superconductivity physicist faked blockbuster results linked only, not reproduced
www.nature.com/articles/d41586-024-00976-y - The path to room-temperature superconductivity: A programmatic approach linked only, not reproduced
www.pnas.org/doi/10.1073/pnas.2520324123