Community and software applications on display at Fermilab quantum science workshop

From Sept. 12-14, more than 100 physicists, engineers and industry partners gathered at Fermilab for a workshop on quantum science and its applications to high-energy physics. Fermilab scientists organized the workshop, which included hands-on tutorials presented by Google.

“That’s the first time Google has given tutorials on their quantum software to the community in a format as big as this one,” said Panagiotis Spentzouris, head of quantum science at Fermilab.

It was just one example of the kind of boundary-crossing interactions that Spentzouris and his colleagues aimed to explore. Industry met with academia. Early-career researchers met with seasoned experts. And theorists and experimentalists discussed areas of common quantum-science interest they could tackle together.

Spentzouris said the workshop met both of its major goals: to foster relationships within the physics research community and to deliver software infrastructure that will allow researchers to access quantum computers and simulators.

“Our aim was to build community, identify algorithms that can be put on quantum hardware in the near future, and, through our partnership with Google, host hands-on tutorials so people can get an idea of how to run this stuff,” Spentzouris said. “That’s what we achieved.”

A wonderful problem

In a 1981 lecture that presaged the potential applications of quantum computers, Richard Feynman said, “Nature isn’t classical, dammit! And if you want to make a simulation of nature, you’d better make it quantum mechanical, and by golly it’s a wonderful problem, because it doesn’t look so easy.”

Developing quantum computers that can run quantum simulations of quantum physics phenomena is anything but easy. But the payoff for particle physics could be solutions to problems that are virtually impenetrable to classical computers.

Contemporary quantum computers typically have just a handful of qubits, or quantum bits, which correspond to classical computer bits. The most powerful quantum computers max out at around 50 qubits. Contrast that with the average laptop’s 40 billion bits.

But even a small number of qubits gives quantum computers extraordinary potential, because each added qubit doubles the computer’s memory and processing power.

Yet another challenge lies in the fact that the computers’ quantum logic gates — the decision-making building blocks of quantum computer circuits – are error-prone, or “noisy,” which limits the size of quantum circuits that scientists can currently construct.

“No one is sure how useful these noisy near-term quantum computers will be. We’ll experiment with them over the next few years to find out,” said John Preskill, a theoretical physicist at the California Institute of Technology, who presented the workshop’s colloquium.

“Eventually we’ll have truly scalable quantum computers, which will have interesting applications to simulation of quantum systems. These more powerful quantum computers will have many more qubits than today’s devices, and may still be many years away,” Preskill said. “To get there we’ll need quite a few advances, including better qubits and better ideas for controlling noise. It’s important for the basic research community to continue exploring hardware platforms that are different from the ones we’re using today.”

Beyond computing

The prospects for quantum computing are the main drivers behind quantum science, but the quantum science umbrella covers more than particle physics simulations.

Qubit-based technologies, such as quantum communication and sensors, were the focus of several workshop talks. For example, scientists are looking to exploit the phenomenon of quantum entanglement — something Einstein called “spooky action at a distance” — to send messages across fiber-optic networks. And they’re developing qubit-based sensors that can detect single photons of light, which could aid in the search for dark matter.

“Quantum is a big tent,” said Fermilab Deputy Director and Chief Research Officer Joe Lykken. “It’s for more than simulating complex physics problems. The fact that we can develop networks and detectors based on quantum principles is one of the most exciting aspects of this field.”

Connecting communities

Attendees included scientists whose work is focused on quantum physics, as well as researchers from condensed matter or nuclear physics backgrounds. Building a more interconnected research community was a key focus of the workshop.

“I think especially in the quantum world, you have people coming from very diverse backgrounds that don’t necessarily know each other but have a lot to teach each other,” Lykken said. “They are trying to solve similar problems to what we’re interested in in high-energy physics. It’s very much an example of a community that needs to know each other better and learn from each other. I think this workshop was an attempt to do that on a pretty large scale.”

Natalie Klco, a graduate student at the University of Washington who presented a talk on quantum computing, also noted the importance of bringing experimentalists and theorists together.

“It’s really nice to have the ability to converse with them to find out what’s hard on the hardware, what’s easy, what can we take advantage of to make some really clever experiments,” she said.

And, of course, the workshop provided a prime opportunity for scientists to talk with members of industry.

In the workshop tutorial, Google software developers Craig Gidney and Kevin Sung introduced scientists to Cirq, an open-source framework for developing quantum algorithms, and OpenFermion, a library that can compile and analyze them. Gidney presented on Cirq, which is designed to help scientists ascertain whether intermediate-scale quantum computers can handle complex computational problems. In the tutorial, Sung demonstrated the software’s ability to solve the Schrödinger equation for the Fermi-Hubbard model, a classic quantum mechanics problem.

“I’m coming from a bit more of a traditional physics perspective, so the Google software tutorial, for example, helped a lot,” said Andy Li, a postdoctoral research associate. “It will be useful for my work in digital quantum simulation.”

All kinds of opportunities

Lykken highlighted the importance of making early-career scientists aware of the fact that quantum information is a field that is part of the future of physics research.

“It’s not just a fad; it’s not going away,” he said. “And if younger researchers want to spend some fraction or all of their research time delving into the many aspects of quantum information sciences, then that’s a good thing for them to do. It’s something that’s opening up. The opportunities are boundless.”

 

Fermilab scientists are harnessing quantum technology in the search for dark matter.

For decades, physicists have been searching for the elusive stuff, which doesn’t emit light but appears to make up the vast majority of matter in the universe. Several theoretical particles have been proposed as dark matter candidates, including weakly interacting massive particles (WIMPs) and axions.

Fermilab’s Aaron Chou is leading a multi-institutional consortium to apply the techniques of quantum metrology to the problem of detecting axion dark matter. The project, which brings together scientists at Fermilab, the National Institute of Standards and Technology, the University of Chicago, University of Colorado and Yale University, was recently awarded $2.1 million over two years through the Department of Energy’s Quantum Information Science-Enabled Discovery (QuantISED) program, which seeks to advance science through quantum-based technologies.

If the scientists succeed, the discovery could solve several cosmological mysteries at once.

“It’d be the first time that anybody had found any direct evidence of the existence of dark matter,” said Fermilab’s Daniel Bowring, whose work on this effort is supported by a DOE Office of Science Early Career Research Award. “Right now, we’re inferring the existence of dark matter from the behavior of astrophysical bodies. There’s very good evidence for the existence of dark matter based on those observations, but nobody’s found a particle yet.”

The axion search

Finding an axion would also resolve a discrepancy in particle physics called the strong CP problem. Particles and antiparticles are “symmetrical” to one another: They exhibit mirror-image behavior in terms of electrical charge and other properties.

The strong force – one of the four fundamental forces of nature – obeys CP symmetry. But there’s no reason, at least in the Standard Model of physics, why it should. The axion was first proposed to explain why it does.

Finding an axion is a delicate endeavor, even compared to other searches for dark matter. An axion’s mass is vanishingly low — somewhere between a millionth and a thousandth of an electronvolt. By comparison, the mass of a WIMP is expected to be between a trillion and quadrillion times more massive — in the range of a billion electronvolts — which means they’re heavy enough that they could occasionally produce a signal by bumping into the nuclei of other atoms. To look for WIMPs, scientists fill detectors with liquid xenon (for example, in the LUX-ZEPLIN dark matter experiment at Sanford Underground Research Facility in South Dakota) or germanium crystals (in the SuperCDMS Soudan experiment in Minnesota) and look for indications of such a collision.

“You can’t do that with axions because they’re so light,” Bowring said. “So the way that we look for axions is fundamentally different from the way we look for more massive particles.”

When an axion encounters a strong magnetic field, it should — at least in theory — produce a single microwave-frequency photon, a particle of light. By detecting that photon, scientists should be able to confirm the existence of axions. The Axion Dark Matter eXperiment (ADMX) at the University of Washington and the HAYSTAC experiment at Yale are attempting to do just that.

Those experiments use a strong superconducting magnet to convert axions into photons in a microwave cavity. The cavity can be tuned to different resonant frequencies to boost the interaction between the photon field and the axions. A microwave receiver then detects the signal of photons resulting from the interaction. The signal is fed through an amplifier, and scientists look for that amplified signal.

“But there is a fundamental quantum limit to how good an amplifier can be,” Bowring said.

Photons are ubiquitous, which introduces a high degree of noise that must be filtered from the signal detected in the microwave cavity. And at higher resonant frequencies, the signal-to-noise ratio gets progressively worse.

Both Bowring and Chou are exploring how to use technology developed for quantum computing and information processing to get around this problem. Instead of amplifying the signal and sorting it from the noise, they aim to develop new kinds of axion detectors that will count photons very precisely — with qubits.

The qubit advantage

In a quantum computer, information is stored in qubits, or quantum bits. A qubit can be constructed from a single subatomic particle, like an electron or a photon, or from engineered metamaterials such as superconducting artificial atoms. The computer’s design takes advantage of the particles’ two-state quantum systems, such as an electron’s spin (up or down) or a photon’s polarization (vertical or horizontal). And unlike classical computer bits, which have one of only two states (one or zero), qubits can also exist in a quantum superposition, a kind of addition of the particle’s two quantum states. This feature has myriad potential applications in quantum computing that physicists are just starting to explore.

In the search for axions, Bowring and Chou are using qubits. For a traditional antenna-based detector to notice a photon produced by an axion, it must absorb the photon, destroying it in the process. A qubit, on the other hand, can interact with the photon many times without annihilating it. Because of this, the qubit-based detector will give the scientists a much higher chance of spotting dark matter.

“The reason we want to use quantum technology is that the quantum computing community has already had to develop these devices that can manipulate a single microwave photon,” Chou said. “We’re kind of doing the same thing, except a single photon of information that’s stored inside this container is not something that somebody put in there as part of the computation. It’s something that the dark matter put in there.”

Light reflection

Using a qubit to detect an axion-produced photon brings its own set of challenges to the project. In many quantum computers, qubits are stored in cavities made of superconducting materials. The superconductor has highly reflective walls that effectively trap a photon long enough to perform computations with it. But you can’t use a superconductor around high-powered magnets like the ones used in Bowring and Chou’s experiments.

“The superconductor is just ruined by magnets,” Chou said. Currently, they’re using copper as an ersatz reflector.

“But the problem is, at these frequencies the copper will store a single photon for only 10,000 bounces instead of, say, a billion bounces off the mirrors,” he said. “So we don’t get to keep these photons around for quite as long before they get absorbed.”

And that means that they don’t stick around long enough to be picked up as a signal. So the researchers are developing another, better photon container.

“We’re trying to make a cavity out of very low-loss crystals,” Chou said.

Think of a windowpane. As light hits it, some photons will bounce off it, and others will pass through. Place another piece of glass behind the first. Some of the photons that passed through the first will bounce off the second, and others will pass through both pieces of glass. Add a third layer of glass, and a fourth, and so on.

“Even though each individual layer is not that reflective by itself, the sum of the reflections from all the layers gives you a pretty good reflection in the end,” Chou said. “We want to make a material that traps light for a long time.”

Bowring sees the use of quantum computing technology in the search for dark matter as an opportunity to reach across the boundaries that often keep different disciplines apart.

“You might ask why Fermilab would want to get involved in quantum technology if it’s a particle physics laboratory,” he said. “The answer is, at least in part, that quantum technology lets us do particle physics better. It makes sense to lower those barriers.”