Luca Toscani de Col (top right) working on the nuclear clock in the Vienna lab. Credit: Matthias Heisler, TU Wien
Radioactive Crystal Clock Now Rivals Top Atomic Clocks in Dark Matter Hunt
In A Nutshell
- A thorium-229 nuclear clock now corrects itself using its own feedback loop, a first for this kind of device.
- In dark matter searches, it matched top atomic clocks for light-based interactions and probed strong-force interactions 10 to 1,000 times more sensitively.
- A day of data showed no dark matter signal, but it narrows how strongly the hypothetical particles could interact with ordinary matter.
- Results drifted from day to day, which the authors suspect comes from crystal flaws, and the planned upgrades remain projections.
A tiny radioactive crystal has become a clock sensitive enough to compete with some of the best atomic clocks on Earth in the search for dark matter, the invisible material that makes up most of the universe’s matter but has never been directly detected. The new thorium-229 nuclear clock, built by a team in Vienna, locks itself onto a faint flicker deep inside a thorium nucleus and corrects itself as it runs, a first for this kind of experiment. Results appear in the journal Nature.
Physicists have chased the idea of a nuclear clock since 2003, when Ekkehard Peik and Christian Tamm proposed that an oddly low-energy quirk inside the thorium nucleus could serve as a timekeeping signal. Because the flicker lives in the nucleus rather than the outer electrons that ordinary atomic clocks rely on, outside interference should disturb it far less, and the thorium can sit inside a solid crystal at room temperature.
What sets this work apart is the feedback loop, plus the payoff in the dark matter hunt. For certain possible interactions involving the strong nuclear force, the clock probed ten to a thousand times more sensitively than earlier atomic clock experiments. For interactions involving light, it matched the best atomic clocks.
Nuclear Clock Locks Onto Thorium Signal Without Outside Help
Top atomic clocks typically hold atoms or ions in vacuum chambers and cool them to extremely low temperatures. This clock skips that. Thorium atoms sit embedded in a clear calcium fluoride crystal only a few millimeters across. An ultraviolet laser shines through it. When the laser hits exactly the right frequency, the thorium nuclei absorb a sliver of the light, and a detector registers the dip.
Staying on that frequency is the hard part, since lasers drift. The system nudges the laser back and forth, checks which way it has strayed, and steers it back, much like a driver correcting course to stay in a lane. A correction can come as often as every 20 seconds, with each taking about a second to apply. Speed matters because a thorium nucleus holds its extra energy for roughly 10 minutes, and earlier methods that watched for a glow had to wait that long. The clock then ran for about a day without anyone stepping in, checked against a trusted atomic clock in Austria.
Nuclear Clock Finds No Dark Matter Signal but Narrows the Search
Some theories hold that dark matter consists of extremely light, undiscovered particles. Such particles would barely interact with ordinary matter, but they could make the strength of nature’s forces wobble slightly in a steady rhythm, which would shake the frequency of the thorium flicker too.
Thorium makes an unusually sensitive probe. Its nucleus has such low energy because two enormous forces, the electric repulsion among its protons and the strong nuclear force holding it together, nearly cancel out. A small change in either force throws off that balance, and theory predicts the thorium flicker reacts far more strongly than the electron-based signals in ordinary atomic clocks. That sensitivity lets the clock compete despite being less steady than the best atomic clocks.
Researchers combed through about 23 hours of data for rhythmic wobbles lasting 20 seconds to a full day, plus a slow drift. No wobble rose above the noise that computer simulations predicted from the equipment alone, and the drift was consistent with zero. That does not rule out the particles, but it caps how strongly they could interact with ordinary matter, assuming they make up most of the dark matter.
Day-to-Day Results Drift, and Crystal Flaws May Be to Blame
This is an early-stage device, and it shows. Within a single run, the clock behaved smoothly. From one day to the next, results shifted noticeably. The laser must be realigned before each run, so the beam crosses a slightly different part of the crystal every time.
To test that explanation, researchers measured four different spots. The thorium signal moved from spot to spot by as much as 13 times what measurement noise alone would explain, while returning to one spot brought back the same frequency. The authors suspect small strains from uneven crystal structure, a hypothesis not yet proven. More uniform crystals, or a beam confined to one reproducible spot, could help.
Stronger Lasers and New Crystals Could Close the Gap With Atomic Clocks
A stronger ultraviolet laser, crystals with thorium built into the material, and a sharper nuclear signal could bring a future thorium clock to the stability of today’s best atomic clocks, according to the paper’s projections. Sensitivity to slight shifts in the basic numbers that set nature’s rules could climb roughly ten-thousand-fold over this version.
A clock built around a crystal small enough for a lab bench is holding its own, in this search, against instruments that took decades to refine. Whether it catches a real signal or only keeps shrinking the places one could hide, it has earned a spot in the hunt.
Paper Notes
Limitations
Within a single continuous run, the clock reached strong stability, but results between runs on different days were limited to roughly 5 parts in 10 trillion, much larger than the instability within a single run. The team linked this partly to realigning the laser before each run, which sends the beam through a slightly different region of the crystal. Tests at four crystal positions showed frequency shifts of up to 1.7 kilohertz, versus 0.13 kilohertz expected from noise, though within the range reported in earlier studies. The authors conjecture that local strain or structural unevenness in the crystal causes the shifts. They note that this reproducibility limit would become a constraint if the dark matter search ran for months or years instead of about 23 hours.
Dark matter limits from the study also carry assumptions. They apply to ultralight scalar particles, assuming those particles make up most of the dark matter, and they depend on theoretical estimates of how sensitive the thorium nucleus is to changes in fundamental forces. The paper states that precise values of these sensitivity factors cannot at present be calculated, though they are predicted to exceed those of the most sensitive atomic clock transitions by several orders of magnitude. The projected upgrades have not been demonstrated.
Funding and Disclosures
Part of the work was funded by the European Research Council under the European Union’s Horizon 2020 research and innovation programme and by the Austrian Science Fund. Support also came from the Österreichische Nationalstiftung für Forschung, Technologie und Entwicklung (AQUnet project), the Deutsche Forschungsgemeinschaft, and the Max-Planck-RIKEN-PTB-Center for Time, Constants and Fundamental Symmetries. The project 23FUN03 HIOC received funding from the European Partnership on Metrology, co-financed by the European Union’s Horizon Europe Research and Innovation Program and by participating states. The Vienna team acknowledged funding from the Defense Advanced Research Projects Agency, and TU Wien provided open access funding. The authors declare no competing interests.
Publication Details
“A thorium-229 optical nuclear clock with feedback loop” was published online October 7, 2026, in Nature. Authors are L. Toscani De Col, T. Riebner, I. Morawetz, F. Schneider, N. Sempelmann, J. Schlachet-Lépinay, F. Schaden, M. Bartokos, G. A. Kazakov, K. Beeks, B. Gerstenecker, M. Pimon, S. Lahs, A. Hellerschmied, T. Lercher, H. Denker, J. Premper, A. Niessner, M. Matus, M. Čížek, O. Číp, V. Lal, G. Zitzer, V. Petrov, J. Tiedau, M. V. Okhapkin, E. Peik, and T. Schumm. Institutions include the Vienna Center for Quantum Science and Technology at TU Wien, the Bundesamt für Eich- und Vermessungswesen in Vienna, the Wolfgang Pauli Institut, the Institut für Erdmessung at Leibniz Universität Hannover, the Institute of Scientific Instruments in Brno, the Physikalisch-Technische Bundesanstalt in Braunschweig, and the Max-Born-Institute for Nonlinear Optics and Ultrafast Spectroscopy in Berlin. Corresponding authors are Ekkehard Peik and Thorsten Schumm. DOI: https://doi.org/10.1038/s41586-026-11084-4







