Seven-Year Dark Matter Hunt Tests Whether the Universe's Missing Mass Has 'Dark Nuclei'

For decades, physicists hunting dark matter have assumed it is made of simple, indivisible particles. A seven-year experiment deep under an Italian mountain has now tested a very different idea: that dark matter might be built like atoms, with its own "dark nuclei" made of billions of smaller pieces.

The DarkSide collaboration reported the results in Physical Review D in a paper titled "First nuclear ultraheavy dark matter search in argon time projection chambers with the DarkSide-50 experiment." The detector found no sign of the composite particles. But the team says the search rules out a wide range of their possible properties and shows that today's detectors can test dark matter candidates that look nothing like the standard picture.

Dark matter makes up about 85% of the mass in the universe, yet it has never been directly detected. The two leading candidates, axions and weakly interacting massive particles, or WIMPs, differ in many ways but share one assumption: dark matter is elementary and cannot be broken into smaller parts, unlike an atomic nucleus, which splits into protons and neutrons. As detectors have come up empty, some physicists have looked elsewhere. "About a decade ago, my collaborators and I explored the possibility that dark matter might instead have its own nuclear physics," said Jocelyn Monroe of the University of Oxford. In that picture, "dark particles bind together in the early universe to form very large 'dark nuclei,' potentially containing billions or vastly more constituents."

From 2013 to 2020 the collaboration ran DarkSide-50, a cylindrical tank of liquid argon at the Italian National Institute for Nuclear Physics' Laboratori Nazionali del Gran Sasso, the world's largest underground research facility. Like other dark matter detectors, it watched for faint flashes of light that should appear when a passing dark matter particle bumps an argon nucleus and makes it recoil. A simple WIMP would cause a single bump. A big composite object is different. "The dark matter has a finite size and a corresponding form factor, and a single object can potentially scatter several times as it passes through the detector," Monroe said. The new analysis combed the full data set for exactly that multiple-hit signature.

None turned up. Still, the null result carries information. "An interesting feature of the result is that the sensitivity depends on the properties of the constituents — the 'dark nucleons' — showing that the internal structure of the dark matter can have observable consequences," Monroe said. In other words, if dark matter does have structure, detectors can in principle see it, much as scattering experiments a century ago revealed the structure inside atoms.

The next test is close. In 2027 the collaboration plans to switch on DarkSide-20k, which will hold roughly 1,000 times more liquid argon than its predecessor. "This will open up the long-term possibility of potentially observing 'dark chemistry' signatures," Monroe said. "More immediately, this work shows that existing experiments can test dark matter candidates that look very different from the conventional WIMP."