Mystery of dark-matter signal deepens with replication attempts

Physicists at a detector in Italy have long claimed to see the Universe’s missing mass — but copycat experiments don’t yet see the same.

Search for this author in:

Part of the ANAIS Experiment

A component of the ANAIS dark-matter detector in Spain.Credit: LSC/ANAIS Collaboration

For more than two decades, only one experiment in the world has consistently reported detecting a signal of dark matter — the Universe’s missing mass that physicists have long tried to identify.

Now, two experiments designed to replicate the results using the same detector technology have presented their first findings. A definite answer remains elusive: although initial data from one experiment seem to be compatible with the original results, the other detector’s findings point in the opposite direction. But scientists say that thanks to these experiments and others scheduled to come online soon, a final answer on the mysterious signal’s nature is now within reach.

“There is really no conclusion to be drawn at this point, other than mounting suspense,” says Juan Collar, a physicist at the University of Chicago in Illinois who has worked on several dark-matter experiments. “But the instruments seem to have sufficient sensitivity to give conclusive results soon,” Collar says.

Ordinary interactions

Observations of how galaxies rotate and of the cosmic microwave background — the ‘afterglow’ of the Big Bang — suggest that most of the matter in the Universe is invisible. This ‘dark’ matter would make its presence known almost exclusively through gravitational interactions with other objects, but a host of experiments have been attempting for decades to pick up signs of its other interactions with ordinary matter.

Since 1998, the underground DAMA detector at the Gran Sasso National Laboratory in central Italy, and its successor DAMA/LIBRA, have recorded a seasonal variation in data. The detector records flashes of light created as particles hit atomic nuclei in a highly purified sodium iodide crystal.

These flashes could be signs of dark matter or stray background radiation — but the experiment’s physicists say that the seasonal change arises because the Earth is moving through a halo of dark-matter particles that surrounds the Milky Way, resulting in a repeating pattern. In March 2018, the DAMA collaboration presented the first results from the detector after it was upgraded in 2010. The dark-matter signature still appeared to be there1.

Over the years, several experiments using various techniques have come up with results that apparently contradict DAMA’s. But COSINE-100 and ANAIS are the first projects to have come online that aim to test DAMA’s claims using the same materials, and each has been operating for more than a year.

ANAIS in the Canfranc Underground Laboratory in the Pyrenees, Spain, reported its first results on 11 March. On the basis of 18 months of data, the findings seem to disagree with DAMA’s2.

The ANAIS data show fluctuations, but they aren’t the same as DAMA’s yearly cycle, in which signals peak in early June and bottom out in early December.

Close but not quite

Another sodium iodide experiment called COSINE-100, under a mountain in South Korea, unveiled a similar analysis to ANAIS’s at conferences this month. The detector also sees a fluctuation in its data. However, “ours is a little bit closer to DAMA’s”, says Reina Maruyama, the co-leader of COSINE-100 and a physicist at Yale University in New Haven, Connecticut. (Both ANAIS and COSINE-100’s results are still preliminary, and have not yet been peer reviewed.)

The ANAIS analysis has “no impact” on the results of DAMA/LIBRA and its predecessors — the data have already confirmed over 20 independent annual cycles, says Rita Bernabei, a physicist at the University of Rome Tor Vergata who has long led the DAMA collaboration.

Even if one experiment were to find strong evidence against the DAMA/LIBRA claim, it will be worthwhile for several detectors to keep taking data for several more years, Collar says: “When one experiment is seeing something like that and the other one isn’t, you are left wondering if someone screwed up.”

Researchers agree that it will take years of data recording by multiple experiments to truly put the matter to rest. “With a few more years of data,” says David Spergel, a cosmologist who first predicted the seasonal oscillation in 1986 with two co-authors. “They should be able to make a definitive statement.”

ANAIS spokesperson Marisa Sarsa, who is also a physicist at the University of Zaragoza in Spain, says that whatever the eventual outcome, her experiment should help to explain what caused the seasonal signal at Gran Sasso. “I have a desire to understand DAMA/LIBRA,” she says, “not just to rule out the result.”


  1. 1.

    Bernabei, R. et al. Nucl. Phys. At. Energy 19, 307–325 (2018).

  2. 2.

    Amaré, J. et al. Preprint available at (2019).

  3. 3.

    Drukier, A. K., Freese, K. & Spergel, D. N. Phys. Rev. D 33, 3495–3508 (1986).

Download references

Nature Briefing

An essential round-up of science news, opinion and analysis, delivered to your inbox every weekday.