After speaking to his adviser, Bob Svoboda at University of California, Davis, about the exciting science potential of DUNE, Gardiner recalled, “I became a neutrino person and never looked back.” He joined Fermilab in 2018, bringing a unique background in neutron simulations to the study of some of the most elusive particles in the universe.

This career track recently earned Gardiner a Department of Energy Early Career Award, which provides funding to explore the low-energy research potential of DUNE. While the experiment is primarily designed for high-energy beam physics, Gardiner is advocating for its use as an unusual kind of telescope. Rather than collecting rays of light to study the universe, DUNE will be sensitive to ghostly low-energy neutrinos coming from outer space, including those from the Sun, supernovae, black holes and possibly dark matter.

“Even though DUNE is designed for beam physics, you can go way lower in energy, and it still performs,” Gardiner said. He believes his specific background allows him to make a “unique contribution due to my career trajectory,” taking him from neutrons to neutrinos.

The technical heart of this research involves upgrading a computer simulation code called MARLEY, or Model of Argon Reaction Low Energy Yields. Because neutrinos interact so weakly with ordinary matter, they are not directly visible in detectors. Instead, physicists must look for extremely rare collisions between neutrinos and atomic nuclei. Like a subatomic version of a car crash investigation, the tracks left by particles coming out of each collision provide clues about what originally happened. To put these puzzle pieces back together, scientists rely on detailed simulations of the collision physics, which is where MARLEY becomes essential.  Results from these simulations will help researchers tell the difference between uninteresting noise, low-energy cosmic neutrinos and potential signals from undiscovered new particles. The ultimate goal of this work is to push the boundaries of physics. “My hope is that as a result of this project, we are thinking about new physics beyond the Standard Model,” Gardiner said.

He also emphasized science opportunities going far beyond the microscopic world. “Neutrinos reveal the inner workings of stars to us, whether it’s the core of our own Sun or the first moments of a supernova explosion,” Gardiner said. By refining the models used to interpret neutrino interactions, he is helping to open a wider view of the distant universe. “DUNE is a new window to the cosmos,” he said. “It gives us a new way of seeing into a hidden world where all kinds of crazy stuff might be going on.”

The DOE funding ensures that low-energy neutrino research may expand our understanding of the fundamental building blocks of the universe through the unparalleled sensitivity of the DUNE detectors.

But finding this elusive material is easier said than done. For one thing, nobody knows exactly what dark matter is made of, so the range of particle masses and their signal frequencies is incredibly broad. Also, dark matter interacts infrequently with ordinary matter and light. To observe it, scientists rely on extremely sensitive detectors to capture very weak signals produced by dark matter particles.

In a study published in Physical Review Letters, scientists at Fermi National Accelerator Laboratory, University of Chicago, Stanford University and New York University used a state-of-the-art detector to speed up the search for one theorized dark matter particle — the dark photon — with unprecedented precision. If it exists, the dark photon would be distantly related to the photon, a visible particle of light.

Their research is enabled by the U.S. Department of Energy’s Quantum Information Science Enabled Discovery program, which partners Fermilab and university scientists to advance quantum sensor development for future high-energy physics experiments.

“Fermilab’s longstanding expertise in designing and building ultrasensitive, low-noise electronics makes it the ideal place to further this technology for next-generation quantum science research like dark matter searches,” said Aaron Chou, a scientist at Fermilab who worked on the study.

The dark photon resides in a narrow frequency band, which means to see its signal a radio-like detector must be carefully tuned to its exact frequency. Scientists developed this detector to be capable of capturing weak signals from dark photons by placing an electrically-tunable instrument called a superconducting quantum interference device — or SQUID — inside a three-dimensional microwave cavity. The device’s superconductance means it has no resistance to energy and can therefore pick up even the faintest signals, such as those from a dark photon.

Key to the detector’s ability to speed up the search for a tiny signal in a broad range of frequencies is flux tuning, which uses electricity to tune the device instead of manually.  

“Rather than physically turning a dial to a specific frequency like with a radio, we apply electromagnetic flux to the SQUID, precisely controlling its ability to oppose changes in electricity flowing through it,” said Fang Zhao, a former Fermilab postdoctoral researcher who led the study.

Somewhat like an electronic pendulum, this flux essentially changes how quickly or slowly the device moves. The microwave cavity is coupled to the SQUID, so changes in the SQUID correspondingly changes the speed of the cavity, allowing it to “listen to” different frequencies.

“Without the ability to electrically tune its frequency, you would have to build billions of detectors to capture the signal,” said Ziqian Li, a former University of Chicago graduate student who also worked on the study. “In contrast, we can build a few flux-tunable detectors and place them at various frequencies, enabling capture of possible signals much faster than before.”

Conventional tunable detectors require mechanically changing the shape of a cavity by physically exerting force or adding mechanical parts inside connected circuits. This poses a challenge because qubit-based detectors require ultracold temperatures to function properly, and extreme cold can cause these parts to seize and break. In addition, mechanical parts emit a lot of heat, which creates noise in the cavity, obscuring signals and decreasing the ability to read and understand the quantum information stored inside the detector.

But use of flux tuning not only enables rapid frequency scanning, it also generates very little heat. This overcomes a major challenge for dark matter searches — preserving coherence. Quantum coherence, says Zhao, is what makes these sensors so precise.

“It’s a fundamental requirement for quantum devices to be protected from anything like heat or noise that might obscure such fragile signals and preserve them long enough for us to detect them.”  

The scientists scanned a relatively large frequency range of 22-megahertz over three days. During this time, they were able to speed up the scanning rate by at least a factor of 20 over mechanical tuning methods. While their search did not turn up any dark photons, they were able to build on previous studies at multiple institutions and narrow the frequency range where dark matter can exist.

“What we’re really trying to do is to build a detector that is more sensitive than anybody else has ever made before; we did that,” said Chou. “We also showed that the detector was compatible with the qubit-based signal readout that we use for dark matter searches and that everything was integrated and everything just worked. It laid the foundation for larger dark matter searches.”

The current detector is very simple, with one cavity and one tunable device — the SQUID. However, work is underway to scale up this technology. Researchers could combine 10, 50 or even more cavities, each covering a different frequency range, with a single tunable element and simultaneously scan a 50 times wider range.

“While there is more work to do to improve scaling, we know now we can use the same detection technique to allow us to detect a large range of the dark photon within a few days, and then the full coverage search of the dark photon is within our reach,” said Li.

Ćiprijanović’s project aims to address a problem that she encountered almost a decade ago when she first started dabbling with AI in her research: the domain shift problem. The problem occurs when a machine learning model is trained on one dataset but is tested or deployed on data that comes from somewhere else; since the model only learned patterns from the training data, its performance often drops when the data changes.

The domain shift problem is persistent in high-energy physics research. Scientists often use simulations to train their AI models, but computational constraints, approximations and unknown physics create unavoidable differences between simulations and real data. This means simulation-trained AI models may perform poorly when they are applied to experimental data.

“Trying to solve this would help the high-energy physics community in general,” said Ćiprijanović.

So Ćiprijanović wants to create a universal AI analysis framework to bridge the gap between simulated and real data. “The main deliverable of this project is going to be a software package that is general and broad and easy to use for all communities,” she said. “You can just plug in your own dataset, easily choose the type of AI model that you want to train and a downstream task that you want to solve.”

While she plans to start with cosmological data — since that is her scientific home — Ćiprijanović also intends to test her project on collider and neutrino physics.

“I really do want to make a software framework that will be used across different high-energy physics frontiers,” she said.

The framework will have a modular structure and a number of data, model and task options to enable broad scientific use and solutions to any high-energy physics domain-shift problem.

Fermilab is uniquely positioned to support this project, said Ćiprijanović. The lab’s science and computing capabilities will enable the creation of a universal AI analysis framework that will work with a vast range of high-energy physics research needs and coding standards.

“We will need to find people to test out the code and give us inputs from cosmic and from neutrinos and from the collider world,” said Ćiprijanović. “Luckily, at Fermilab, we have experts from all these frontiers! Fermilab is the place to do this.”

Since 2010, the highly competitive DOE Office of Science Early Career Research Program has distributed funding annually to support outstanding early career scientists at universities, national laboratories and Office of Science user facilities.