Where is the dark matter? Scientists who have hunted for decades for the stuff that comprises most of the cosmos’ mass are starting to worry that they are looking in the wrong places. After the latest null results came out this summer from the most sensitive search yet for the particles thought to make up dark matter, a limited theoretical range of masses and other characteristics remains viable for the particles. Now physicists have proposed two new methods to trawl this slim remaining territory, which has been out of reach to experiments so far.

Most of the universe’s mass is dark matter, around 80% of it. Although we cannot see or touch it, scientists know its gravity distorts images of distant objects and holds galaxies together. Since the 1980s, experiments buried deep in mountains and mines have waited patiently to see if a dark matter particle will pass through. Europe’s Large Hadron Collider (LHC) slams other particles together hoping to create some dark matter in the process. But so far, the elusive substance has failed to turn up at either the LHC or the Large Underground Xenon (LUX) experiment in South Dakota. Researchers are now facing growing hints that existing experiments may be targeting the wrong kinds of particles, and finding dark matter will require new techniques.

Dark matter searches to date have mostly searched for “weakly interacting massive particles” (WIMPs), theoretical particles that would weigh between 1 giga-electron volt (GeV) and 1 tera-electron volt (TeV), or between one and 1,000 times the mass of a proton. Many physicists have long viewed them as the most promising dark matter candidates, because theory implies that WIMPs should contribute about as much mass to the universe as the amount of dark matter astronomers have measured, but the particles have so far failed to appear. These experiments tend to search for rare instances of WIMPs impacting atoms in some detecting material; in the case of LUX, the material is liquid xenon, but others have used solid germanium or other substances.

“The WIMP paradigm is under siege” after so many failures to find them have limited the number of places they could still be hiding, says Kathryn Zurek of the Lawrence Berkeley National Laboratory in Berkeley, California. Zurek led two recent studies proposing new ways of searching for dark matter in the form of particles that would be lighter than WIMPs, such as so-called asymmetric dark matter. Such particles might interact with the normal particles we know of via some yet-undiscovered dark force. “The idea is that you can have this hidden sector where dark matter is really light … and can have individual particle interactions with [regular] particles,” Zurek says. “It’s not a paradigm people had been really thinking about until less than a decade ago.”

In place of traditional dark matter detector materials, Zurek’s team’s first method uses superconducting aluminum, a substance whose electrons are free to move without any resistance. Within the superconductor electrons bind themselves with partner electrons in so-called “Cooper pairs.” Energy from an incoming dark matter particle could break up one of these pairs and send vibrations through the superconductor, which hypersensitive heat detectors called transition edge sensors (TESs) would read out. The researchers published this method last January in Physical Review Letters.

The second method, published last month in Physical Review Letters, uses superfluid helium, a zero-viscosity liquid of ultra-cold helium atoms that can move around each other without any resistance. An incoming dark matter particle could interact with a helium nucleus, causing a chain reaction that sends a set of phonons, quantum sound waves, to TESs. Both methods require a much slighter knock from dark matter into the detecting material to generate a signal than existing experiments, and can therefore spot particles as light as 1 keV, a millionth the mass of a proton. Traditional experiments are sensitive only to particles as light as 10 MeV, ten thousand times heavier than a keV.

The current generation of dark matter experiments are getting upgrades; LUX is becoming the LUX-ZEPLIN (ZonEd Proportional scintillation in LIquid Noble gases) or LZ experiment, XENON100 in Italy is becoming XENON1T and the Super Cryogenic Dark Matter Search (SuperCDMS)  in Minnesota will move to a new Canadian site. But even the improved versions can only probe down to around 10 MeV at best. If they cannot find anything, scientists will likely look to proposals like Zurek’s to probe even lighter masses of potential particles. Yet such experiments will require research and development into what superconducting aluminum or superfluid helium detectors will actually look like, and where they should build such a detector. “These experiments will be technically challenging but not very expensive,” Zurek says.

“[This research] is the direction the field is moving partly because we haven't found the standard WIMPs,” says Dan Bauer, a scientist at the Fermi National Accelerator Laboratory in Illinois and the SuperCDMS spokesperson. Although scientists are still holding out hope that higher mass WIMPs will appear, “we realized that we’ve always been looking under particular lampposts. There's a lot of territory available for lighter mass dark matter particles.”

And as experimentalists build detectors that can spot lighter particles, theoretical physicists are likely to come up with more ideas for types of dark matter candidate particles that could be found there. “Theorists are very creative,” says Bob Jacobsen, a University of California, Berkeley, physicist who works on LUX and LZ. “If there's a [mass] region that hasn't been explored, the theorists will say can they do something that's mathematically consistent. If they publish it, it's our job to rule it out.”

Superconductor and superfluid detector proof-of-concepts will ultimately require physicists to divide their time between current searches and research and development, says Chris Tully, a Princeton University physicist. “You have to build these technologies in parallel with running experiments” he says. He hopes to begin seeing mockups in five to 10 years’ time, depending on funding (though Zurek herself thought it would be closer to 10 years). Whereas the experiment Tully works on, the Princeton Tritium Observatory for Light, Early-Universe, Massive-Neutrino Yield (PTOLEMY) , and  SuperCDMS  have TESs already developed, scientists must further test future detectors to ensure they can pick out dark matter particles from contaminating radiation that causes a similar signal in the detector. Current experiments are located deep underground or in mountains to shield against cosmic rays, high-energy particles from space that produce signals that can obscure dark matter. Superfluid or superconductor experiments would instead need shielding from stray electromagnetic waves, such as those from cell phones, Zurek says. Bauer says that SuperCDMS’ Sudbury Neutrino Observatory Laboratory is building such shielding now. Ultimately, “It's amazing how little we know,” says Jacobsen. “We're looking for the first clue in the crime scene. If you don't have the first clue, you don't know where to look.”