It is the elephant in the room for dark-matter research: a claimed detection that is hard to believe, impossible to confirm and surprisingly difficult to explain away. Now, four instruments that will use the same type of detector as the collaboration behind the claim are in the works or poised to go online. Within three years, the experiments will be able to either confirm the existence of dark matter—or rule the claim out once and for all, say the physicists who work on them.

“This will get resolved,” says Frank Calaprice of Princeton University in New Jersey, who leads one of the efforts.

The original claim comes from the DAMA collaboration, whose detector sits in a laboratory deep under the Gran Sasso Massif, east of Rome. For more than a decade, it has reported overwhelming evidence for dark matter, an invisible substance thought to bind galaxies together through its gravitational attraction. The first of the new detectors to go online, in South Korea, is due to start taking data in a few weeks. The others will follow over the next few years in Spain, Australia and, again, Gran Sasso. All will use sodium iodide crystals to detect dark matter, which no full-scale experiment apart from DAMA’s has done previously.

Scientists have substantial evidence that dark matter exists and is at least five times as abundant as ordinary matter. But its nature remains a mystery. The leading hypothesis is that at least some of its mass is composed of weakly interacting massive particles (WIMPs), which on Earth should occasionally bump into an atomic nucleus.

DAMA’s sodium iodide crystals should produce a flash of light if this happens in the detector. And although natural radioactivity also produces such flashes, DAMA’s claim to have detected WIMPs, first made in 1998, rests on the fact that the number of flashes produced per day has varied with the seasons.

This, they say, is exactly what is expected if the signal is produced by WIMPs that rain down on Earth as the Solar System moves through the Milky Way’s dark-matter halo. In this scenario, the number of particles crossing Earth should peak when the planet’s orbital motion lines up with that of the Sun, in early June, and should hit a low when its motion works against the Sun’s, in early December.

There is one big problem. “If it’s really dark matter, many other experiments should have seen it already,” says Thomas Schwetz-Mangold, a theoretical physicist at the Karlsruhe Institute of Technology in Germany—and none has. But at the same time, all attempts to find weaknesses in the DAMA experiment, such as environmental effects that the researchers had not taken into account, have failed. “The modulation signal is there,” says Kaixuan Ni at the University of California, San Diego, who works on a dark-matter experiment called XENON1T. “But how to interpret that signal—whether it’s from dark matter or something else—is not clear.”

No other full-scale experiment has used sodium iodide in its detector, although the Korea Invisible Mass Search (KIMS), in South Korea, used caesium iodide. So the possibility remains that dark matter interacts with sodium in a different way to other elements. “Not until someone has turned on a detector made of the same material will you grow convinced that nothing is there,” says Juan Collar at the University of Chicago, Illinois, who has worked on several dark-matter experiments.

Many have found it challenging to grow sodium iodide crystals with the required purity. Contamination by potassium, which has a naturally occurring radioactive isotope, is a particular problem.

But now three dark-matter-hunting teams—KIMS; DM-Ice, run from Yale University in New Haven, Connecticut; and ANAIS, at the University of Zaragoza, Spain—have each obtained crystals with about twice the level of background radioactivity of DAMA’s. That is pure enough to test its results, they say.

The KIMS and DM-Ice teams have built a sodium iodide detector together at Yangyang Underground Laboratory, 160 kilometres east of Seoul. This instrument uses an ‘active veto’ sensor that will enable it to separate the dark-matter signal from the noise better than DAMA does, says Yeongduk Kim, the director of South Korea’s Center for Underground Physics in Daejeon, which manages KIMS.

ANAIS is building a similar detector in the Canfranc Underground Laboratory in the Spanish Pyrenees. Together, KIMS/DM-Ice and ANAIS will have about 200 kilograms of sodium iodide, and they will pool their data. That is comparable to DAMA’s 250 kilograms, enabling them to catch a similar number of WIMPs, they say. Even though the newer detectors will have higher levels of background noise, it should still be possible to either falsify or reproduce the very large DAMA signal, says Reina Maruyama of Yale, who leads DM-Ice.

But Calaprice argues that high purity is more important than mass. He and his collaborators have developed a technique to lower contamination, and in January announced that they were the first to obtain crystals purer than DAMA’s. He expects to reduce the background levels further, to one-tenth of DAMA’s.

The project, SABRE (Sodium-iodide with Active Background Rejection), will put one detector at Gran Sasso and the other at the Stawell Underground Physics Laboratory, which is being built in a gold mine in Victoria, Australia. SABRE will also use a sensor to pull out the dark-matter signal from noise, and will have a total mass of 50 kilograms.

SABRE should complete its research and development stage in about a year, and will build its detectors soon after that, says Calaprice. It will then make its technology available to other labs—something that DAMA did not do. And having twin detectors in both the Northern and Southern hemispheres could clarify whether environmental effects could have mimicked dark matter’s seasonality in DAMA’s results—if the signal is from WIMPs, then both detectors should see peaks at the same time.

DAMA will wait at least until 2017 to release its latest results, says spokesperson Rita Bernabei of the University of Rome Tor Vergata. She is not holding her breath about the upcoming sodium iodide detectors. “Our results have already been verified in countless cross-checks in 14 annual cycles, so we have no reason to get excited about what others may do,” she says. If other experiments do not see the annual modulation, she adds, her collaboration will conclude that they do not yet have sufficient sensitivity.

Could the teams prove DAMA right? “I was unwilling to believe the DAMA results or even take them seriously at first,” says Katherine Freese, a theoretical astroparticle physicist at the University of Michigan in Ann Arbor, who with her collaborators first proposed the seasonal modulation technique used by DAMA. But, as DAMA’s data have accumulated, and no other explanation for their signal has arisen, Freese is now excited by the possibility that dark matter may have been discovered after all. The fact that many have tried and failed to repeat DAMA’s experiment shows that it is not easy, says Elisabetta Barberio at the University of Melbourne, who leads the Australian arm of SABRE. “The more one looks into their experiment, the more one realizes that it is very well done.”

Editor's Note (4/18/16): The last paragraph has been amended to better reflect Katherine Freese’s views on the DAMA collaboration’s results.

This article is reproduced with permission and was first published on April 5, 2016.