BATAVIA, ILLINOIS—After 4 days Lazarus rose from the grave, but physicists here at Fermi National Accelerator Laboratory (Fermilab) are resurrecting a massive particle detector by lowering it into a tomblike pit and embalming it with a chilly fluid. In August 2018, workers eased two gleaming silver tanks bigger than shipping containers, the two halves of the detector, into a concrete-lined hole. Hauled from Europe 2 years ago, ICARUS—an outdated acronym for Imaging Cosmic And Rare Underground Signals—will soon start a second life seeking perhaps the strangest particles physicists have dreamed up, oddballs called sterile neutrinos.
On a bright afternoon in May, ICARUS lies in its insulated sarcophagus while technicians mill about on top of it, quietly connecting electronics to ports, as indifferent to observers as stonemasons caught up in their work. Once the detector is wired, workers will seal it under a concrete lid 3 meters thick. They will then fill ICARUS with frigid liquid argon, explains Angela Fava, a Fermilab physicist, looking on from a railing above. "We hope to start taking data by the end of the year."
The experiment gives ICARUS another shot at glory. It ran nearly a decade ago deep underground at Gran Sasso National Laboratory in L'Aquila, Italy, to study the properties of ordinary neutrinos—and achieved underwhelming results. "They were late and they had a much smaller detector than was expected, so they weren't competitive," says Giovanni De Lellis, a physicist at the University of Naples Federico II in Italy and spokesperson for the Oscillation Project with Emulsion-tracking Apparatus (OPERA), a rival neutrino experiment at the same lab that beat ICARUS to a significant observation.
Still, ICARUS succeeded in another way: It proved that a new kind of detector called the liquid argon time projection chamber—the brainchild of Italian Nobel laureate Carlo Rubbia—could capture the rare interactions of neutrinos with atomic nuclei with unprecedented precision. "To be honest, I didn't think it was going to work," says Chang Kee Jung, a neutrino physicist at the State University of New York in Stony Brook. "I'm kind of in awe." The approach worked so well that in 2012, U.S. physicists chose the relatively risky technology for their ultimate neutrino experiment, to be built next decade.
In the meantime, ICARUS is going back to work on the other side of the Atlantic Ocean. After a 2017 journey from Europe to Illinois, by sea, river, and road, it will resume the neutrino hunt here at Fermilab—this time looking for sterile neutrinos. If they exist, sterile neutrinos would be new additions to physicists' standard model of particles and forces. They would resolve nagging puzzles about their fellow neutrinos, but would be far harder to detect. Ordinary neutrinos interact—just barely—with other matter. But sterile neutrinos, in keeping with their name, would interact with nothing at all except other neutrinos, emerging and disappearing through a subtle identity swapping that can be detected only indirectly.
Since the late 1990s, experiments have hinted at their existence. Although far from convincing, the case for sterile neutrinos is a scab that even skeptical physicists have to pick. "We can't ignore it," Jung says. "We really need closure." Patrick Huber, a theorist at Virginia Polytechnic Institute and State University in Blacksburg, says the hunt makes even more sense given the lack of signs of new physics at large particle colliders. "If you're going to look for new physics, why not look in places where there appears to be something going on?" Huber says.
Neutrinos outnumber every other type of particle in the universe except photons. Every second trillions of them pass unnoticed through each of us. The particles come in three flavors—electron, muon, and tau—depending on how they are born. For example, nuclear interactions produce electron neutrinos, whereas muon neutrinos come from the decays of particles called muons, heavier cousins of electrons, which can emerge when cosmic rays strike the atmosphere.
Weirdly, neutrinos can change flavors. That revelation came 2 decades ago, when physicists discovered that the sun appeared to emit half as many electron neutrinos as predicted and that a fair fraction of the muon neutrinos raining down from the atmosphere seemed to disappear before they reached detectors. Flavor swapping, also called neutrino oscillations, explains those deficits. It also proves that neutrinos, long thought to be massless, actually have the slightest heft: Relativity requires a massless particle to travel at light speed, in which case time dilation stops the clock, preventing any flavor changes.
Physicists have roughed out the three-flavor theory by studying oscillations in neutrinos generated by nuclear reactors and fired from accelerators through Earth to detectors hundreds of kilometers away. The world's current big neutrino experiments are Fermilab's NOvA, which shoots muon neutrinos to a detector 810 kilometers away in Ash River, Minnesota, and Japan's T2K, which directs a similar beam from a laboratory in Tokai to Super-Kamiokande, a detector in a zinc mine 295 kilometers away. Such long "baselines" are necessary because the three neutrino flavors oscillate only over great distances, at rates that depend on their energies and the tiny differences in their masses.
These experiments have given researchers a handle on the differences in the three neutrinos' masses. But the masses themselves remain immeasurably small—no more than one-millionth the mass of an electron. That situation gives theorists heartburn, says Stephen Parke, a theorist at Fermilab, because physicists have few good ways to explain why the neutrino's mass is so exquisitely close to, but not exactly, zero.
Enter sterile neutrinos. Nearly massless neutrinos muddle a mathematical symmetry in the standard model, Parke explains. But theorists can patch up the symmetry and explain the tiny masses if they assume each type of neutrino is born as a quantum-mechanical mixture: a large helping of an ordinary neutrino and a dash of a far heavier sterile neutrino. The mixing would set up a "seesaw mechanism" that further drives the ordinary neutrino's mass down and the sterile neutrino's mass up.
The first experimental hint of sterile neutrinos came from the Liquid Scintillator Neutrino Detector (LSND), which ran from 1993 until 1998 at Los Alamos National Laboratory in New Mexico. Researchers shot protons into a target to generate a beam of pure muon neutrinos. But they spotted dozens of electron neutrinos in their detector 30 meters away, far more than they were expecting for such a short distance. The result suggested muon neutrinos were quickly morphing into heavier sterile neutrinos and then into electron neutrinos, in a kind of flavor-changing shortcut. The sterile neutrino's higher mass would make the oscillation happen more quickly.
Researchers with Fermilab's Mini-Booster Neutrino Experiment (MiniBooNE), which ran from 2002 until this year, found similar results by firing higher energy muon neutrinos to a bigger detector 500 meters away. Other experiments have found different, complementary hints. Numerous measurements suggest nuclear reactors produce fewer electron neutrinos than expected, perhaps because the neutrinos are oscillating to sterile form and disappearing. Some astrophysical observations hint that sterile neutrinos floating in space could be the undetected dark matter particles that are thought to make up most of the universe's mass.
Caveats abound. The sterile neutrinos that could explain the LSND and MiniBooNE results would be far, far lighter than those required by the seesaw mechanism—and too light to explain cosmic dark matter. Reactors might produce fewer neutrinos than expected because of poorly understood radioactive decays in the hot reactor fuel, not because of neutrino oscillations. "Getting all these observations to agree is very, very difficult," Huber says. But that tension could hint at an even bigger discovery, Parke says. "If it's not sterile neutrinos but some other new physics, that's even more exciting," he says. One way or another, Parke says, the LSND and MiniBooNE results need to be explained.
While hints of sterile neutrinos were accumulating, Rubbia and colleagues were developing ICARUS. Rubbia dreamed up the idea in 1977, 7 years before he shared the Nobel Prize in Physics for discovering particles of a very different kind: the massive W and Z particles responsible for the weak nuclear force that governs certain nuclear decays. In breaks from his prize-winning work, Rubbia envisioned a 5000-ton detector that would watch for an event expected to be theoretically possible, although vanishingly rare: the decay of the proton. The same detector could also study neutrinos from the atmosphere or fired from an accelerator.
A neutrino detector typically spots an electron neutrino when a nucleus absorbs it and spits out an energetic electron; for a muon neutrino the signal is a muon. Because the events are so rare, Rubbia needed many tons of detector material. That requirement typically requires a trade-off in precision.
Super-Kamiokande, for example, is a tank filled with 50,000 tons of ultrapure water and lined with 13,000 light detectors. When a neutrino spawns an electron or muon, that speeding charged particle radiates a shock wave of light that is picked up by the tank's detectors as a ring. The sharpness or fuzziness of the ring reveals the type of particle and the flavor of the original neutrino. But Super-Kamiokande can't track the electrons or muons precisely or spot other lower energy particles created in the interactions, limiting its ability to estimate the original neutrino's energy, a key variable for studying neutrino oscillations in detail.
Rubbia envisioned a detector both large and precise: a large tank of liquid argon with a fine grid of wires on one side and high-voltage electrodes on the other. The high energy electron or muon spawned by a neutrino in a flash of light would ionize the argon along its path, leaving a trail of electrons. A strong electric field would push these electrons to the grid. By registering which wires in the grid were hit by electrons—and when—the detector could trace the original electron or muon in 3D. The dense liquid provides lots of target nuclei, and its chemical properties enable electrons to drift long distances.
A liquid argon detector can track an electron or muon with millimeter precision and distinguish between the particles easily, says Janet Conrad, a neutrino physicist at the Massachusetts Institute of Technology in Cambridge. "The tracks are very different—kind of like rabbit tracks and deer tracks," she says. Such precision makes it easier to rule out spurious events. The detector can also track all the other charged particles blasted from an interaction, making it possible to better estimate the original neutrino's energy, which determines how quickly it should oscillate.
Executing Rubbia's vision was an immense challenge. The argon would have to be chilled to –186°C and contaminants reduced to levels of parts in a trillion. Through the 1990s, physicists at CERN, the European particle physics laboratory near Geneva, Switzerland, and Italy's National Institute for Nuclear Physics in Pavia developed ever bigger prototypes, trucking the final 760-ton detector—smaller than Rubbia wanted—into Gran Sasso in 2008. "It was in an underground tunnel, and it was a huge detector that needs a huge cryogenic system," says Ornella Palamara, a Fermilab physicist who worked on ICARUS at Gran Sasso. "All of this takes time."
Once in place, ICARUS monitored a stream of muon neutrinos fired from CERN for hints of an oscillation that had not yet been observed—from muon to tau neutrinos. It lost out to OPERA, an unusual detector that used bricks of photographic film to spot the decays of the heavy, short-lived tau particles spawned by the tau neutrinos. ICARUS's biggest scientific achievement may have been shooting down OPERA's claim in September 2011 that neutrinos coming from CERN traveled faster than light, a spurious signal eventually traced to a loose electrical cable. "ICARUS was the right detector in the wrong place," says Heidi Schellman, a neutrino physicist at Oregon State University in Corvallis.
Nevertheless, ICARUS altered the course of neutrino research. That's because it ran just as U.S. physicists were mulling the next great neutrino experiment: the Deep Underground Neutrino Experiment (DUNE), which will lurk 1500 meters down in an abandoned gold mine in Lead, South Dakota, and snare neutrinos shot from Fermilab 1300 kilometers away, a distance physicists say is ideal for measuring neutrino oscillations. Whereas current experiments focus on particular parameters of the three-flavor model, DUNE will scrutinize so many neutrinos with such great precision that it will measure all the parameters at once to more rigorously test the model. It will also look for an asymmetry between neutrinos and their antiparticles, antineutrinos, that could help explain how the early universe generated more matter than antimatter.
Early plans for DUNE envisioned a water detector as much as four times the size of Super-Kamiokande. But then physicists started to see images from ICARUS, says Sam Zeller, a physicist at Fermilab. "You'd look at them and say, ‘This is a game changer.’" In the end they opted for the riskier liquid argon technology: a multibillion-dollar detector containing 40,000 tons of the stuff, to be built by 2026.
Rubbia wasn't done with ICARUS, however. In 2013 he suggested bringing the detector to Fermilab, where researchers were hoping to build a similar detector to hunt sterile neutrinos. "So, of course, the Fermilab scientific committee said, ‘Sit down and work together to make a plan,’" Palamara says.
Although ICARUS was too small to compete with OPERA, it's just the right size to confirm or rule out the tantalizing LSND and MiniBooNE results, Schellman says. "It's a brilliant reuse of the detector. ICARUS is finally where it should be."
As in previous efforts, Fermilab physicists will blast high energy protons into a target to produce an intense beam of muon neutrinos. The neutrinos will shoot through a string of three detectors under Fermilab's prairie campus. At 110 meters from the neutrino source will sit the Short-Baseline Neutrino Detector, a 112-ton liquid-argon detector now under construction. At 470 meters from the source sits MicroBooNE, a 170-ton liquid argon detector that has been running since 2015. ICARUS itself, the biggest of the three detectors, will be 600 meters away.
That three-detector setup should address the biggest weakness in the LSND and MiniBooNE experiments. Both employed a single detector at a fixed distance from the neutrino source and essentially counted electron neutrinos in a muon neutrino beam, attributing their unexpected appearance to oscillations from muon to sterile to electron type. They could not rule out the possibility that extraneous particles like photons were producing spurious events that researchers mistook for electron neutrinos. With the three detectors, physicists can see whether the number of electron neutrinos increases with distance from the source, as it should for a bona fide neutrino oscillation.
To nail the case for a short-distance oscillation, physicists must also conduct a second, complementary measurement, Parke says. If a few muon neutrinos oscillate to sterile neutrinos and then to electron neutrinos, then many more must oscillate to the sterile form and disappear altogether. "If you don't see muon neutrinos disappearing at the same time, it's a dagger in the heart of the sterile neutrino," Parke says.
It's not quite as simple as counting the muon neutrinos in each detector, however. Higher energy neutrinos oscillate more slowly than lower energy ones, so physicists have to measure the muon neutrinos' energy spectrum in each detector to see whether its shape changes with distance the correct way. That's where the liquid argon detector's ability to tally up neutrino energy comes into play.
Experimenters have different hunches about whether the international $100 million program will end up finding its quarry. "The data prefers the model that has the sterile neutrino—by a lot," Conrad says. But Zeller, a skeptic, cautions that the hints from LSND and MiniBooNE could have conventional explanations. "What's the saying? If you hear hoofsteps behind you it's probably a horse and not a zebra," she says.
Although she's also skeptical, Schellman hesitates to write off the sterile neutrino because time and again neutrinos have proved stranger than expected. "I thought neutrino oscillations were a crock" until they were discovered, she Says. "I'm withholding judgment."