Paired Diamond Defects Give Quantum Sensors a Window Into Hidden Magnetic Fluctuations

Nitrogen molecules moving at more than 30,000 feet per second strike the surface of a lab-grown diamond, and the diamond does not give. The molecule breaks apart. Its two nitrogen atoms, no longer bonded, shoot off in separate directions into the crystal. That collision is the first step in a magnetic sensor built by researchers at Princeton, described in a Nov. 27 paper in Nature.

The team, led by Nathalie de Leon, an associate professor of electrical and computer engineering and the paper's senior author, controls how much energy each molecule carries. That sets how deep the atoms go. They pass a few dozen layers of carbon atoms and stop about 20 nanometers below the surface, roughly 10 nanometers from each other. The diamonds are about the size of a large flake of sea salt and far purer than natural stones. Each defect the researchers engineer in them is tiny, one missing atom in a lattice of billions. Such defects interact strongly with magnetic fields, which is why they work as sensors.

Until now, these sensors have usually been treated as single points in space. De Leon's group placed two defects extremely close together, close enough to interact quantum-mechanically and become entangled. Their electrons then act in lock step, so a measurement of one gives a perfectly correlated measurement of the other. Because the two defects are still distinct points, like two eyes, they can triangulate signatures in noisy magnetic fluctuations and home in on the source. The team reports roughly 40 times greater sensitivity than previous techniques. De Leon says the method exposes magnetic fluctuations that stay buried in the statistical data of conventional approaches. "You have this totally new kind of playground," she said.

The idea began in 2020, when Jared Rovny joined her lab as one of the first Princeton Quantum Initiative postdoctoral fellows. The COVID-19 pandemic had limited lab access, so he and de Leon turned to theory. They asked whether diamond defects, called nitrogen vacancy centers, could detect correlations in the magnetic noise that hums behind condensed matter physics. De Leon called it "one of these weird, COVID, theory projects." At the time, sensing such correlations was not a topic of scientific conversation, and they started out of curiosity. Only after formalizing the idea, she said, did they see how powerful it was.

Rovny's background in nuclear magnetic resonance, where interacting particles and their correlations are central, pushed it further. He is now a physicist at the quantum computing startup Logiqal. Working first with Shimon Kolkowitz, an atomic physicist then at the University of Wisconsin-Madison and now at the University of California-Berkeley, the team studied correlations between two defects that were not entangled. That work produced a 2022 paper in Science, but it was technically onerous and too complex for most experiments.

Entanglement changed that. Rovny realized that entangling the defects meant the presence or absence of a correlation would put its fingerprint on the system. That let the team skip the most cumbersome steps and get the benefit of two sensors for about the cost of one. "Now all I have to do is a single measurement," de Leon said, "a single normal measurement."

Philip Kim, an experimental physicist at Harvard who was not part of the study, called it a new way of operating the sensor, one that probes something not possible before. Other techniques for getting at this information have been confined to carefully constructed arrays of atoms, he said, not real materials. The scale matters too. Between the atomic scale and the wavelength of visible light, scientists want to measure quantities such as how far an electron travels before bouncing off another particle, and how magnetic vortices evolve in superconductors under special conditions. Kim called it the length scale of interest.

Kim is now working with de Leon, using complementary techniques in his lab. He studies superconductors cooled to their critical temperatures with liquid nitrogen, and graphene, a material with promising uses that has been hard to engineer at scale. Those materials are the next test for the sensor.