LZ records one unexplained flash a mile under South Dakota and stops short of a discovery

A detector filled with about seven tonnes of liquid xenon, sitting nearly a mile under the old Homestake gold mine in Lead, South Dakota, recorded a single recoil that the LUX-ZEPLIN collaboration cannot easily blame on ordinary backgrounds. Samuel Eriksen of the University of Bristol presented the event on 1 September at the TeV Particle Astrophysics conference in Tendo, Japan. A paper has gone to Physical Review Letters. A preprint is on the experiment's site.

The flash deposited 248 keV. That is a harder kick than the low-energy nuclear recoils LZ was built to catch first. If a weakly interacting massive particle caused it, the particle would weigh at least 200 GeV, and more likely around 1,000 GeV, or several hundred times the mass of a proton. The analysis used 220 days of data taken from March 2023 to April 2024. The candidate showed up in June 2023, when Earth is moving into the galaxy's dark-matter wind, the season when a real WIMP signal should be slightly stronger.

The statistical weight is 2.6 sigma. There is about a half-percent chance the event is a known background. Particle physics treats 5 sigma as a discovery. Eriksen was plain: it is one event, and the team is not claiming dark matter. Rick Gaitskell of Brown University, the LZ spokesperson, said the collaboration does not want to get ahead of itself. It has seen something it wants other physicists to inspect.

How the detector is supposed to work

LZ is managed by Lawrence Berkeley National Laboratory and run at the Sanford Underground Research Facility. Two hundred and fifty scientists from 39 institutions work on it. A dual-phase xenon time projection chamber sits inside veto layers of extra xenon and a water tank. Rock overhead cuts cosmic rays. When a particle hits a xenon nucleus, photomultiplier tubes record a flash. The shape of that flash is how the team tells a nuclear recoil from an electron recoil.

Decades of WIMP searches have produced null results tight enough that many theorists moved on to other candidates. Katherine Freese of the University of Texas at Austin, whose work helped set the WIMP agenda, said she has waited 40 years for a positive result and is excited by this one. Wick Haxton of the University of California, Berkeley, who is not on the paper, called it the most interesting thing in the field in recent times and noted that the event also hints at a mechanism, not only a count.

Knut Morå of the University of Zurich, who is on the collaboration, said the instrument was built so the needle would jump on an event like this. The jump is real. The label on the needle is not yet agreed.

What has to happen next

LZ already holds the world's largest dark-matter exposure and is still taking data. If the 248 keV region collects more events, the significance will rise or the candidate will drown in background. A single June event can be a WIMP, a rare neutron, or a flaw in the background model. Only more live time sorts those three.

For now the honest sentence is the one Gaitskell used. They have not seen dark matter. They have seen one recoil in the place they were watching, at an energy that is awkward for the simplest WIMP charts, under a mile of South Dakota rock.