A version of this story appeared in Science, Vol 376, Issue 6592.
Kent Irwin has a vision: He aims to build a glorified radio that will reveal the nature of dark matter, the invisible stuff that makes up 85% of all matter. For decades, physicists have struggled to figure out what the stuff is, stalking one hypothetical particle after another, only to come up empty. However, if dark matter consists of certain nearly massless particles, then in the right setting it might generate faint, unquenchable radio waves. Irwin, a quantum physicist at Stanford University, plans to tune in to that signal in an experiment called Dark Matter Radio (DM Radio).
No ordinary radio will do. To make the experiment practical, Irwin’s team plans to transform it into a quantum sensor—one that exploits the strange rules of quantum mechanics. Quantum sensors are a hot topic, having received $1.275 billion in funding in the 2018 U.S. National Quantum Initiative. Some scientists are employing them as microscopes and gravimeters. But because of the devices’ unparalleled sensitivity, Irwin says, “dark matter is a killer app for quantum sensing.”
DM Radio is just one of many new efforts to use quantum sensors to hunt the stuff. Some approaches detect the granularity of the subatomic realm, in which matter and energy come in tiny packets called quanta. Others exploit the trade-offs implicit in the famous Heisenberg uncertainty principle. Still others borrow technologies being developed for quantum computing. Physicists don’t agree on the definition of a quantum sensor, and none of the concepts is entirely new. “I would argue that quantum sensing has been happening in one form or another for a century,” says Peter Abbamonte, a condensed matter physicist and leader of the Center on Quantum Sensing and Quantum Materials at the University of Illinois, Urbana-Champaign (UIUC).
Still, Yonatan Kahn, a theoretical physicist at UIUC, says quantum sensors open the way to testing new ideas for what dark matter might be. “You shouldn’t just go blindly looking” for dark matter, Kahn says. “But even if your model is made of bubblegum and paperclips, if it satisfies all cosmological constraints, it’s fair game.” Quantum sensing is essential for testing many of those models, Irwin says. “It can make it possible to do an experiment in 3 years that would otherwise take thousands of years.”
Astrophysical evidence for dark matter has accreted for decades. For example, the stars in spiral galaxies appear to whirl so fast that their own gravity shouldn’t keep them from flying into space. The observation implies that the stars circulate within a vast cloud of dark matter that provides the additional gravity needed to rein them in. Physicists assume it consists of swarms of some as-yet-unknown fundamental particle.
In the 1980s, theorists hypothesized what soon became the leading contender: weakly interacting massive particles (WIMPs). Emerging in the hot soup of particles after the big bang, WIMPs would interact with ordinary matter only through gravity and the weak nuclear force, which produces a kind of radioactive decay. Like the particles that convey the weak force, the W and Z bosons, WIMPs would weigh roughly 100 times as much as a proton. And just enough WIMPs would naturally linger—a few thousand per cubic meter near Earth—to account for dark matter.
Occasionally a WIMP should crash into an atomic nucleus and blast it out of its atom. So, to spot WIMPs, experimenters need only look for recoiling nuclei in detectors built deep underground to protect them from extraneous radiation. But no signs of WIMPs have appeared, even as detectors have grown bigger and more sensitive. Fifteen years ago, WIMP detectors weighed kilograms; now, the biggest contain several tons of frigid liquid xenon.
The second most popular candidate—and one DM Radio targets—is the axion. Far lighter than WIMPs, axions are predicted by a theory that explains a certain symmetry of the strong nuclear force, which binds quarks into trios to make protons and neutrons. Axions would also emerge in the early universe, and theorists originally estimated they could account for dark matter if the axion has a mass between one-quadrillionth and 100-quadrillionths of a proton.
In a strong magnetic field, an axion should sometimes turn into a radio photon whose frequency depends on the axion’s mass. To amplify the faint signal, physicists place in the field an ultracold cylindrical metal cavity designed to resonate with radio waves just as an organ pipe rings with sound. The Axion Dark Matter Experiment (ADMX) at the University of Washington, Seattle, scans the low end of the mass range, and an experiment called the Haloscope at Yale Sensitive to Axion CDM (HAYSTAC) at Yale University probes the high end. But no axions have shown up yet.
In recent years physicists have begun to consider other possibilities. Maybe axions are either more or less massive than previously estimated. Instead of one type of particle, dark matter might even consist of a hidden “dark sector” of multiple new particles that would interact through gravity but not the three other forces, electromagnetism and the weak and strong nuclear forces. Rather, they would have their own forces, says Kathryn Zurek, a theorist at the California Institute of Technology. So, just as photons convey the electromagnetic force, dark photons might convey a dark electromagnetic force. Dark and ordinary electromagnetism might intertwine so that rarely, a dark photon could morph into an ordinary one.
To spot such quarry, dark matter hunters have turned to quantum sensors—a shift partly inspired by another hot field: quantum computing. A quantum computer flips quantum bits, or qubits, that can be set to 0, 1, or, thanks to the odd rules of quantum mechanics, 0 and 1 at the same time. That may seem irrelevant to hunting dark matter, but such qubits must be carefully controlled and shielded from external interference, exactly what dark matter hunters already do with their detectors, says Aaron Chou, a physicist at Fermi National Accelerator Laboratory (Fermilab) who works on ADMX. “We have to keep these devices very, very well isolated from the environment so that when we see the very, very rare event, we’re more confident that it might be due to the dark matter.”
The interest in quantum sensors also reflects the tinkerer culture of dark matter hunters, says Reina Maruyama, a nuclear and particle physicist at Yale and co-leader of HAYSTAC. The field has long attracted people interested in developing new detectors and in quick, small-scale experiments, she says. “This kind of footloose approach has always been possible in the dark matter field.”
For some novel searches, the simplest definition of a quantum sensor may do: It’s any device capable of detecting a single quantum particle, such as a photon or an energetic electron. “I call a quantum sensor something that can detect single quanta in whatever form that takes,” Zurek says. That’s what is needed for hunting particles slightly lighter than WIMPs and plumbing the dark sector, she says.
Such runty particles wouldn’t produce detectable nuclear recoils. A wispy dark sector particle could interact with ordinary matter by emitting a dark photon that morphs into an ordinary photon. But that low-energy photon would barely nudge a nucleus.
In the right semiconductor, however, the same photon could excite an electron and enable it to flow through the material. Kahn and Abbamonte are working on an extremely sensitive photodiode, a device that produces an electrical signal when it absorbs light. Were such a device shielded from light and other forms of radiation and cooled to near absolute zero to reduce noise, a dark matter signal would stand out as a steady pitter-pat of tiny electrical pulses.
The trick is to find a semiconductor sensitive to very low-energy photons, Kahn says. The industrial standard, silicon, releases an electron when it absorbs a photon with an energy of at least 1.1 electron volts (eV). To detect dark sector particles with masses as low as 1/100,000th that of a proton, the material would need to unleash an electron when pinged by a photon of just 0.03 eV. So Kahn, Abbamonte, and colleagues at Los Alamos National Laboratory are exploring “narrow bandgap” semiconductors such as a compound of europium, indium, and antimony.
Even lighter dark-sector particles would create photons with too little energy to liberate an electron in the most sensitive semiconductor. To hunt for them, Zurek and Matt Pyle, a detector physicist at the University of California, Berkeley, are developing a detector that would sense the infinitesimal quantized vibrations set off when a dark photon creates an ordinary photon that pings a nucleus. It would “only rattle that nucleus and produce a bunch of vibrations,” Pyle says. “So the detectors must be fundamentally different.”
Their detector consists of a single crystal of material composed of two types of ions with opposite charges, such as gallium arsenide. The feeble photon spawned by a dark photon would nudge the different ions in opposite directions, setting off quantized vibrations called optical phonons. To detect these vibrations, Zurek and Pyle dot the crystal with small patches of tungsten and chill it to temperatures near absolute zero, where tungsten becomes a superconductor that carries electricity without resistance. Any phonons would slightly warm the tungsten, reducing its superconductivity and leading to a noticeable spike in its resistance.
Within 5 years, the researchers hope to improve their detector’s sensitivity by a factor of 10 so that they can sense a single phonon and hunt dark-sector particles weighing one-millionth as much as a proton. To provide the dark matter, such particles would have to be so numerous that a detector weighing just a few kilograms should be able to spot them or rule them out. And because so few experiments have probed this mass range, even little prototype detectors unshielded from background radiation can yield interesting data, Pyle says. “We run just in our lab aboveground, and we can get world-leading results.”
Some physicists argue that true quantum sensors should do something more subtle. The Heisenberg uncertainty principle states that if you simultaneously measure the position and momentum of an electron, the product of the uncertainties in those measurements must exceed a “standard quantum limit.” That means no measurement can yield a perfectly precise result, no matter how it’s done. However, the principle also implies you can swap greater uncertainty in one measurement for greater precision in the other. To some physicists, a quantum sensor is one that exploits that trade-off.
It can make it possible to do an experiment in 3 years that would otherwise take thousands of years.
- Kent Irwin
- Stanford University
Physicists are using such schemes to enhance axion searches. To make up dark matter, those lightweight particles would be so numerous that en masse they’d act like a wave, just as sunlight acts more like a light wave than a hail of photons. So with their metal cavities, ADMX and HAYSTAC researchers are searching for the conversion of an invisible axion wave into a detectable radio wave.
Like any wave, the radio wave will have an amplitude that reveals how strong it is and a phase that marks its exact synchronization relative to whatever ultraprecise clock you might choose. Conventional radio circuits measure both and run into a limit set by the uncertainty principle. But axion hunters care only about the signal’s amplitude—is a wave there or not?—and quantum mechanics lets them measure it with greater precision in exchange for more uncertainty in the phase.
HAYSTAC experimenters exploit that trade-off to tamp down noise in their experiment. The vacuum—the backdrop for the measurement—can itself be considered a wave. Although that vacuum wave has on average zero amplitude, its amplitude is still uncertain and fluctuates to create noise. In HAYSTAC a special amplifier reduces the vacuum’s amplitude fluctuations while allowing those in the irrelevant phase to grow bigger, causing any axion signal to stand out more readily. Last year, HAYSTAC researchers reported in Nature that they had searched for and ruled out axions in a narrow range around 19-quadrillionths of a proton mass. By squeezing the noise, they increased the speed of the search by a factor of 2, Maruyama says, and validated the principle.
Such “squeezing” has been demonstrated for decades in laboratory experiments with lasers and optics. Now, Irwin says, “These techniques for beating the standard quantum limit [have] been used to actually do something better, as opposed to do something in a demonstration.” In the DM Radio experiment, he hopes to use a related technique to probe for even lighter axions as well as dark photons.
Instead of a resonating cavity, DM Radio consists of a radio circuit containing a charge-storing capacitor and a current-storing inductor—a carefully designed coil of wire—both placed in a magnetic field. Axions could convert to radio waves within the inductor coil to create a resonating signal in the circuit at a certain frequency. Researchers can also look for dark photons by reconfiguring the coil and turning off the magnetic field.
To read out the signal, Irwin’s scheme plays on another implication of quantum mechanics, that by measuring a system’s state you may change it. The researchers couple their resonating circuit to a second, higher frequency circuit, so that, much as in AM radio, any dark matter signal would make the amplitude of the higher frequency carrier wave warble. The stronger the coupling, the bigger the warbling, and the more prominent the signal. But stronger coupling also injects noise that could stymie efforts to measure dark matter with greater precision.
Again, a quantum trade-off comes to the rescue. The researchers modify their carrier wave by injecting a tiny warble at the frequency they hope to probe. Just by random chance, that input warble and any dark matter signal will likely be somewhat out of sync, or phase. But the dark matter wave can be thought of as the sum of two components: one that’s exactly in sync with the added signal and one that’s exactly out of sync with it—much as any direction on a map is a combination of north-south and east-west. The experiment is designed to measure the in-sync component with greater precision while injecting all the disturbance into the out-of-sync component, making the measurement more sensitive and accelerating the rate at which the experiment can scan different frequencies.
Irwin and colleagues have already run a small prototype of the experiment. They are now building a larger version, and ultimately they plan one with a coil that has a volume of 1 cubic meter. Implementing the quantum sensing is essential, Irwin says, as without it, scanning the entire frequency range would take thousands of years.
Some dark matter hunters are explicitly borrowing hardware from quantum computing. For example, Fermilab’s Chou and colleagues have used a superconducting qubit—the same kind Google and IBM use in their quantum computers—to perform a proof-of-principle search for dark photons in a very narrow energy range. Like a smaller version of ADMX or HAYSTAC, their experiment centers on a resonating cavity, this one drilled into the edge of an aluminum plate. There a dark photon could convert into radio waves, although at a higher frequency than in ADMX or HAYSTAC. Ordinarily, experimenters would bleed the radio waves out through a hole in the cavity and measure them with a low-noise amplifier. However, the tiny cavity would generate a signal so faint it would drown in noise from the amplifier itself.
The qubit sidesteps that problem. Like any other qubit, the tiny superconducting circuit can act like a clock, cycling between different combinations of 0 and 1 at a rate that depends on the difference in energy between the circuit’s 0 and 1 states. That difference in turn depends on whether there are any radio photons in the cavity. Even one is enough to speed up the clock, Chou says. “We’re going to stick this artificial atomic clock in the cavity and see if it still keeps good time.”
The measurement probes only the amplitude of the radio waves and not their phase, obtaining greater precision in the former in exchange for greater uncertainty in the latter, the team reported last year in Physical Review Letters. It might speed up dark photon searches by as much as a factor of 1300, Chou says, and it could be extended to search for axions, if researchers could apply a magnetic field to the cavity while shielding the sensitive qubit.
One group has invented a scheme to search for WIMPs using another candidate qubit: a so-called nitrogen vacancy (NV) center within a diamond crystal. In an NV, a nitrogen atom replaces a carbon atom in the crystal lattice and creates an adjacent, empty site that collects a pair of electrons that can serve as qubit. A WIMP passing through a diamond can bump carbon atoms out of the way, leaving a trail of NVs roughly 100 nanometers long, says Ronald Walsworth, an experimental physicist at the University of Maryland, College Park. The NVs will absorb and emit light of specific wavelengths, so the track can be spotted clearly with fluorescence microscopy.
That scheme has little to do with quantum computing, but it would address a looming problem for WIMP searches. If current liquid xenon detectors get much bigger, they should start to see well-known particles called neutrinos, which stream from the Sun. To tell a WIMP from a neutrino, physicists would need to know where a particle came from, as WIMPs should come from the plane of the Galaxy rather than the Sun. A liquid xenon detector can’t determine the direction of a particle that caused a signal. A detector made of diamonds could.
Walsworth envisions a detector formed of millions of millimeter-size synthetic diamonds. A diamond would flash when pierced by a neutrino or WIMP, and an automated system would remove it and scan it for an NV track, using the time of the flash to determine the track’s orientation relative to the Sun and the Galaxy, the team explained last year in Quantum Science and Technology. Walsworth hopes to build a prototype detector in a few years. “I absolutely do not want to claim that our idea would work or that it’s better than other approaches,” he says. “But I think it’s promising enough to go forward.”
Physicists have proposed many other ideas for using quantum sensors to search for dark matter, and the influx of money should help transform them into new technologies, Zurek says. “Things can move faster when you’re funded,” she says. As tool builders, dark matter hunters embrace that push. “They have a great hammer, so they started looking for nails,” Walsworth says. Perhaps they’ll bang out a discovery of cosmic proportions.