On June 16, 2023, 39 seconds after 3:22 P.M. and one mile beneath the Black Hills of South Dakota, there was a sudden flash of light.

It happened in a dark, sealed tank of extremely cold liquid in a cavern where miners once chiseled gold from the walls. Caught by the tank’s many cameras, the flash also sent an electric charge through hair-thin wires submerged in the liquid.

The tank is the centerpiece of the LUX-ZEPLIN (LZ) experiment, one of several ongoing attempts by physicists to snare and study dark matter, a shadowy something they suspect is everywhere—yet somehow seems nowhere to be found. They’ve sought it for the past 40 years by placing detectors in some of Earth’s most silent and secluded spaces, places where dark matter’s whispers might register. But every time so far, they’ve come up empty, with every burst of light or surge of electricity having a more ordinary explanation.

On supporting science journalism

If you’re enjoying this article, consider supporting our award-winning journalism by subscribing. By purchasing a subscription you are helping to ensure the future of impactful stories about the discoveries and ideas shaping our world today.

This flash from three years ago is different. The charge and light together seem to indicate a violent collision between an unknown particle and the nucleus of one of the detector’s atoms. The event bears almost no resemblance to any of the “backgrounds”—mundane sources of flashes that mimic dark matter—which the LZ collaboration’s hundreds of scientists have spent their careers battling.

Yet the collision is also out of sync with the kind of dark matter signal they’ve been searching for. That is, it’s a total mystery. But after four decades of nothing, a new mystery is welcome.

“I have been waiting for a positive result for, oh, my god, 40 years,” says Katherine Freese, a theoretical physicist at the University of Texas at Austin, whose work helped inspire the search. “So you can bet I’m very, very excited by the LZ results.”

The LZ collaboration announced the finding in a presentation today at the TeVPA conference in Japan. An accompanying paper has been submitted to Physical Review Letters. According to the project’s best estimates, the flash’s chances of being a statistical fluke are about half a percent (which is still far above the 0.00003 percent threshold used by particle physics experiments to hail a discovery).

The universe we see when we look out into space is almost impossible to countenance without dark matter. The stuff our telescopes can track is swirling too fast, forming too quickly into clumps that are too dense, to easily explain. The easiest explanation, in fact, is that, compared with the matter we can see, there must be five times more stuff that’s invisible and inert, still carrying gravity’s weight but emitting no light. This dark matter is ubiquitous and plentiful, but we have no idea what it’s made of.

Detectors like LZ are predicated on a guess: that dark matter is made of weakly interacting massive particles (WIMPs). These are hypothetical particles that almost always slip through ordinary matter like ghosts. But very rarely, they should interact with it, colliding at random with an atom. LZ and its competitors are racing to pack as many atoms as possible into the quietest places they can find to catch these collisions.

Still, traces of background radioactivity sometimes break the subterranean silence, with that activity piercing through LZ’s mile of protective bedrock or even seeping out of the materials lining the detector’s walls. But physicists can discern many of these events from their best guess at dark matter’s signals and try to reduce the incidents as much as they can. That a single unexplained event can carry so much statistical significance shows how incredibly sensitive the technology has become. “We made our meter so that the needle would jump when it saw an event like this,” says Knut Morå, a physicist at the University of Zurich and a member of the LZ collaboration. “Everything is working as intended.”

The collision from June 2023 meets most of the team’s WIMPy expectations—except that it’s far too explosive. For LZ to find such a high-energy WIMP event, most standard theories say that it should have already seen dozens or hundreds of gentler ones—which it has not. So if what LZ saw is a WIMP, this particle must be exceptionally odd—interacting in a slightly different way with regular matter or somehow having been boosted to a higher speed.

“Given that we don’t know what the dark matter is, it doesn’t strike me as strange that it might not be exactly the thing we expect,” says Hugh Lippincott, an associate professor at the University of California, Santa Barbara, and a member of the LZ collaboration.

Whether or not this signal will prove to be genuine is presently impossible to say. What’s already certain is that it will spark a frenzy of papers seeking to explain how tweaks to various theoretical flavors of WIMPs could account for LZ’s mysterious event. “They are potentially seeing other types of interactions which are perfectly reasonable,” Freese says.

The collaboration itself also advises caution. “We shouldn’t be comfortable claiming specific things about dark matter or about new physics based on one event,” Morå says. Additionally, the team’s scientists analyzed the data without “blinding” themselves to it, an approach to avoid bias that’s considered the gold standard for particle physics. (They attempted an alternative approach called “salting” but stated that it was unsuccessful.)

Fortunately, more answers should arrive in the near future, based on a larger tranche of data that has been properly blinded. “This was 200 days of data. We actually have over 700 days blinded,” says Richard Gaitskell, a professor at Brown University and co-spokesperson of LZ. “So that puts us in a strong position to study this.”

Those data will either reveal this one event to be an anomaly or uncover more like it. In the latter case, two competing experiments, PandaX-4T and XENONnT, will help determine whether LZ has really seen WIMPs or some as-yet-unknown background.

For now, the LZ collaboration is trying its best to manage expectations. “You get these flashes of, like, ‘Wait, could this really be it?’” Morå says. “But it would be damaging to the analysis to be in that mindset, and I don’t think many people are.” Lippincott agrees, albeit with a nod toward the extraordinary. “It is far from any of the known backgrounds,” he says, “so it’s hard not to wonder.”

The mystery collision comes at a perilous time for the dark matter race. “It’s a good moment for nature to give us this gift,” says Juan Collar, a physicist at the University of Chicago, who is not part of the LZ collaboration. Detectors are becoming so large and so sensitive that they will soon be swamped with other ghostly particles called neutrinos that are generated in the sun and in Earth’s atmosphere. At that point, the race would have to cease—or at least to shift away from WIMPs to instead seek some other elusive quarry.

“Here’s your last window of opportunity to see a WIMP—if it’s not there, you have to close up shop at some point in the next 10 years,” Collar says. “This could be the beginning of something spectacular.”