On 16 June 2023, at 3:22 in the afternoon, a flash of light appeared in a tank of liquid xenon a mile under the Black Hills of South Dakota. On 1 September 2026, the collaboration that runs that tank told the world what it had seen: one particle interaction it cannot explain with any known background. It is the most compelling hint of dark matter the experiment has produced so far.
Run the event yourself
A WIMP enters the detector. Click the tank to fire one in, watch it scatter off a xenon nucleus, and see where the resulting flash sits against the discovery threshold.
The quietest room on Earth
LUX-ZEPLIN, LZ for short, is the most sensitive detector of its kind. Ten tonnes of ultrapure liquid xenon, watched by 494 photomultiplier tubes, buried under a mile of rock at the Sanford Underground Research Facility in South Dakota. It is built to catch WIMPs, weakly interacting massive particles, the leading candidate for the dark matter that makes up most of the mass in the universe. SLAC, which reported this result, helped build it: the purification systems that keep the xenon clean and the electrodes that collect the signals.
The number
The event is a single flash from the 220 live days of data the team analyzed, collected between March 2023 and April 2024. Its significance is 2.6 sigma, which means roughly a 0.5 percent chance it comes from known backgrounds. Physics reserves the word "discovery" for 5 sigma. The collaboration is explicit about where this stands. "We are not claiming to have seen dark matter," said Rick Gaitskell, the LZ spokesperson. "But we have seen something interesting."
Here is the part that makes it interesting rather than just a number. The event does not fit. If it is a WIMP, it would have to weigh at least 200 GeV, more than 200 times a proton, and interact with ordinary matter in a way the simplest models do not allow. LZ should have seen dozens of gentler hits first. It has not. The signal is simultaneously the best candidate the field has ever had and the wrong shape for the theory that motivated the search.
The process is the product
The collaboration is handling this the way a field keeps its credibility: in the open. The result was presented at the TeVPA conference in Japan, the paper is on arXiv and submitted to Physical Review Letters. Two competing experiments, XENONnT and PandaX-4T, will look at their own data. And LZ has more than 700 days of blinded data still unexamined, data collected without the analysts knowing what a dark matter signal would look like in it. Either the event grows into something, or it fades into the background it could not be identified as. Both outcomes are on the table.
None of this would carry weight without the record around it. Forty years of searches, zero robust detections. DAMA/LIBRA's annual modulation remains unconfirmed by any other experiment. The field is also running out of room: detectors are now so sensitive that solar neutrinos, the so-called neutrino fog, are about to set a hard floor under every WIMP search. This is, as one physicist put it, the last window.
Editorial
The restraint is the story
The sentence I keep returning to is not the one about dark matter. It is what the spokesperson said first: we are not claiming to have seen dark matter. After forty years of nothing, with the best candidate the field has produced sitting in their data, the official position is restraint. That restraint is the product. It is what lets the eventual claim, if it comes, be believed.
And the detail that stays with me: the event does not fit. Nature did not hand them the clean version of their prediction. If this flash survives the blinded data, the theory bends, not the experiment. The most interesting number in physics right now is not 5 sigma. It is 0.5 percent, and nobody is pretending it is more.