Mines scientists lead novel measurement to advance proton decay searches

The LHC Physics Center at Fermi National Accelerator Laboratory hosted the 2026 CMS Data Analysis School in January, marking 15 years and 34 sessions of the program. Since the first school was held at Fermilab in January 2011, the program has provided intense, hands-on training for the next generation of physicists working on the CMS experiment at CERN.

The CMS experiment is one of the primary particle detectors at the Large Hadron Collider and was one of two detectors used to confirm the existence of the Higgs boson in 2012. The LHC Physics Center at Fermilab, commonly referred to as LPC, serves as a resource and physics analysis hub, primarily for the 700 U.S. physicists working in the international CMS collaboration.

The school’s rigorous curriculum covers physics, detector technology, computing, software and data analysis. This year’s session included 48 students — primarily graduate students with a small number of undergraduates — along with 48 facilitators and 11 lecturers. Facilitators included volunteers from the CMS experiment and the 2026 LPC Distinguished Researchers.

Innovations in artificial intelligence and analysis

The program combined traditional lectures with exercises in statistics, particle reconstruction, event simulation and machine learning. While machine learning has been a staple of the curriculum since 2019, this year featured a special demonstration of agentic AI used to simulate high-energy physics particle events.

During the latter half of the week, the focus shifted to an intensive group project. Teams of 10 students each completed all aspects of a full CMS analysis in just three days. The session culminated in a friendly competition where a panel of senior CMS researchers assessed each group’s technical understanding, teamwork and presentation skills.

Distinguished speakers and lecturers

Guest speakers from across the physics community presented lectures and offered insights throughout the week. Among the presenters were Fermilab Director Norbert Holtkamp, Fermilab Deputy Director for Science and Technology Bonnie Fleming and CMS collaboration Spokesperson Anadi Canepa. Additional experts in particle physics included Paddy Fox, Jeffrey Eldred and Scarlet Norberg of Fermilab; Matteo Cremonesi of Carnegie Mellon University; Tova Holmes of the University of Tennessee; Christoph Paus of MIT; Alejandro Gómez Espinoza of Carnegie Mellon; and Isobel Ojalvo of Princeton.

Fostering collaboration

The school also featured activities intended to foster collaboration, including group meals, a banquet featuring a trivia contest and evening social gatherings at a Fermilab recreational facility. Despite an intense week of early mornings and late nights, students maintained their enthusiasm, beginning with insightful questions during the opening lectures and concluding with “drum rolls” as the winners of the friendly competition were announced.

LPC Distinguished Researcher and school lecturer Christoph Paus of the Massachusetts Institute of Technology commented, “The event turned out to be a full success, with students learning lessons needed for graduate school, as well as making essential connections with their peers and some of the more senior members of the CMS collaboration. I highly recommend the event to any student who is just starting or has already been involved in CMS for a while.”

Interesting Engineering

Beneath Earth’s surface, shielded from the effects of most cosmic rays, is the Northwestern Experimental Underground Site, or NEXUS. Located about 350 feet underground at Fermi National Accelerator Laboratory, the research facility enables scientists to study the behavior of quantum devices in their quest to find evidence of dark matter.

It’s here that a multi-institutional team of scientists took measurements of correlated charge noise in a chip comprised of multiple superconducting qubits for the first time.

As reported in Nature Communications, their work will help inform the design of future quantum-based particle physics detectors, as well as develop noise reduction strategies to reduce qubit errors and decoherence.

“Understanding whether a charge burst could affect multiple qubits as the charge moves through the chip … is crucial to scientists who use quantum sensors …”

Superconducting qubits are a leading option for building quantum computers. However, they are sensitive to disturbances from their environment and can make errors. By understanding how electrical fluctuations called charge noise affect superconducting qubits, scientists can find ways to reduce these errors and improve quantum computers.

When an ionizing particle, like a cosmic ray or gamma ray passes through such a chip, it can create bursts of charge that can impact information stored in qubits. Scientists can directly measure these events because the qubits used in the study are incredibly sensitive to fluctuations in charge.

“Understanding whether a charge burst could affect multiple qubits as the charge moves through the chip — what researchers call correlated charge noise — is crucial to scientists who use quantum sensors to detect very faint signals that are possibly from dark matter, and to computer scientists, who are interested in correcting errors,” said Daniel Bowring, a scientist at Fermilab and organizer of this study.

This research is an extension of that conducted in 2019 by collaborators at the University of Wisconsin-Madison. In that study, scientists measured correlated charge noise on the Earth’s surface using the same chip comprised of four superconducting qubits. Among their findings were signals detected from both cosmic rays and gamma rays.

In the more recent study at Fermilab, scientists used the very same four-qubit chip and placed it in the underground NEXUS lab to block most cosmic rays. With a lead shield surrounding the dilution refrigerator that houses the qubit chip, they took measurements with the shield both open and closed to isolate and compare the effects of gamma radiation.

“Qubits are sensitive to different types of faint signals. If we want to use them as particle detectors, we need to be sure we can tell these signals apart from each other.”

“Qubits are sensitive to different types of faint signals. If we want to use them as particle detectors, we need to be sure we can tell these signals apart from each other,” said Bowring.

More specifically, they wanted to see what effect making things as quiet as possible would have on the rate of charge bursts on qubits and whether these effects were correlated in multiple qubits.

When comparing measurements, scientists expected to find a marked decrease in charge bursts with the shield closed. They did find a reduction, though overall less than expected. Interestingly, they found that, even with the shield closed, some correlated charge noise was present, indicating the presence of some other source of background interference.

“These comparisons will lead to new designs where we purposely engineer the amount of response to the environment,” said study coauthor Enectali Figueroa-Feliciano, a professor of physics and astronomy at Northwestern University. “Having that control will lead to quantum devices optimized to minimize their environmental response for use in quantum computing applications. It will also allow us to maximize it for quantum sensing applications.”

The study was conducted with support from the Quantum Science Center. Collaborating institutions include: Fermilab, Illinois Institute of Technology, Northwestern University, SLAC National Accelerator Laboratory, Stanford University, Tufts University, University of Wisconsin-Madison, University of Florida in Gainesville, Université Grenoble, University of Toronto, Université Paris-Saclay, and Wellesley College.

An accelerator technology test facility with a catchy name has achieved a major milestone at the U.S. Department of Energy’s Fermi National Accelerator Laboratory by successfully accelerating its first proton beams. The achievement marks an important step forward that will support research aimed at pushing the boundaries of stable, high-intensity beams in future particle accelerator programs.

“This marks a major advancement for our R&D program.”

“This marks a major advancement for our R&D program,” said Jonathan Jarvis, director of Fermilab’s Accelerator Research Division. “This new proton-beam capability lets us address the challenges we’ll face as we increase the beam power in Fermilab’s accelerators.”

The dedicated research and development accelerator complex is called FAST/IOTA — shorthand for Fermilab’s Accelerator Science and Technology facility and its Integrable Optics Test Accelerator ring. Particle accelerators empower scientists to study the universe at its smallest scales, and FAST/IOTA helps ensure the United States will continue to hold a preeminent position in high-energy particle physics.

FAST has two linear accelerators that are connected to the IOTA storage ring, one for electrons and now one for protons. The IOTA ring is equipped with unique magnets and other advanced technologies that enable researchers to study new concepts for cutting-edge accelerator systems and high-intensity beam physics.

Compared to Fermilab’s main accelerator complex, a large, well-established workhorse delivering high-energy protons for physics experiments, FAST/IOTA is designed to evaluate technologies that could improve the efficiency, reliability and performance of future particle accelerators.

“FAST/IOTA is a dedicated R&D accelerator complex that provides us the freedom to explore high-risk, high-reward ideas,” explained Jarvis. “These ideas could dramatically improve the way we design, build and operate particle accelerators.”

In accelerator physics, intensity refers to the number of particles packed into a beam and accelerated through the machine. With protons now circling the ring at around 7% of the speed of light, researchers can further investigate methods for mitigating beam instabilities, develop advanced control systems and discover ways to apply artificial intelligence to accelerator operations.

Fermilab is in the process of upgrading its high-power, proton-accelerator complex through the PIP-II project. This upgrade will enable higher-intensity beams for the lab’s neutrino science program, beginning with the Deep Underground Neutrino Experiment. FAST/IOTA will enhance this effort by providing advanced tools and understanding for navigating the challenges of such high-intensity operations.

By figuring out how to increase the number of particles in an accelerator beam and reduce beam losses, researchers can steer more particles toward a target. When more particles hit the target, they generate more secondary particles such as neutrinos. FAST/IOTA’s flexible, modular design makes it especially suited for developing new methods to increase and maintain beam intensity.

“Researchers can conduct tests and experiments without tying up the operations of a larger production facility,” said Trey Thompson, an engineering physicist on the FAST/IOTA team. “We can shut down, swap out part of the ring and install new experiments without the usual constraints.”

Before this upgrade, FAST/IOTA primarily worked with electron beams. For example, in 2022, Fermilab researchers published a landmark paper in Nature, reporting the first experimental demonstration of optical stochastic cooling — a breakthrough technique that cools a particle beam using its own emitted light. The team also conducted studies on single electrons, using the facility’s unique capabilities to track individual particles for hours at a time.

“Electrons were a perfect starting point and complement to our proton work,” said Jarvis. “They’re easy to work with, and they help us explore the advanced concepts and technologies that we are developing for our core proton program.”

With this upgrade, FAST/IOTA is also helping to pioneer the use of artificial intelligence in accelerator design and operations.

“About half of our program now focuses on enhancing accelerators with AI,” Jarvis said. “We’re building high-fidelity simulation environments that capture the complexity of the real machine and are fully integrated with its control systems. These virtual accelerators allow us to train AI systems to optimize performance and discover new configurations. The key is that we can then test those simulations in a real machine, at scale.”

AI is not only a research focus but also a practical tool in deploying the new proton systems.

“We used AI tools to help optimize the output of the proton source,” said Fermilab engineering physicist Chip Edstrom. “We’ve applied algorithms that treat the system as a black box — meaning we don’t need to understand every internal detail — and instead optimize parameters based on performance.”

This approach reflects Fermilab’s broader core value of ingenuity, which also makes FAST/IOTA an ideal environment for accelerator workforce development. Students and postdoctoral researchers gain “full-stack” experience, from design and construction to operation and data analysis for particle accelerators. At any given time, the program hosts multiple graduate students and postdoctoral researchers, exposing them to cutting-edge R&D spanning physics, engineering and AI.

“There aren’t many places where you can get hands-on experience with a facility of this scale,” Jarvis said. “It’s an incredible opportunity to train the next generation of accelerator scientists and technologists.”

And the impact of FAST/IOTA will extend well beyond Fermilab. “This is the realization of a vision that began over a decade ago,” Jarvis added. “We’ve built a uniquely capable facility that will help guide the operation and design of next-generation accelerators — from Fermilab’s own PIP-II project to machines around the world.”