It took Odysseus twenty years to get home from Troy. It took about as long for the nuclear clock to travel from an idea to a working machine.
The idea of a nuclear clock goes back to 2003, when FQxI member Ekkehard Peik and Christian Tamm proposed one for the very first time. Atomic clocks (the most precise clocks we have at the moment) and nuclear clocks work on the same principle: a laser is tuned until its frequency matches the exact energy a particle needs to jump between two of its states, and because the particle absorbs only that one energy, the jump makes a fixed reference to keep time against. In an atomic clock, the particle that jumps is an electron. However, twenty-odd years ago, Peik and Tamm asked whether the nucleus could take its place, reasoning that a nucleus, shielded from the jostling outside world beneath its cloud of electrons, would hold even stiller. The problem is that most nuclei jump only when struck by a violent burst of gamma rays, nothing like the lower-energy, tunable light of a laser. One nucleus in the whole periodic table is the exception, its jump small enough for a laser to reach: thorium-229.
This June, two independent teams reported that they had finally built a nuclear clock. FQxI’s Ekkehard Peik, who heads the time and frequency department at Germany’s national metrology institute, the PTB, in Braunschweig, co-led the European team behind one of the papers, with Thorsten Schumm of TU Wien. A second group, led by Shiqian Ding at Tsinghua University in Beijing, described a clock of its own days later. Both are yet to be peer reviewed. “It is just beautiful,” Schumm told Live Science’s Kenna Hughes-Castleberry, “how a very ‘wild’ idea such as manipulating an atomic nucleus with a laser has turned into reality.”
Each clock is a millimetre-sized crystal of calcium fluoride grown with thorium-229 inside and probed by an ultraviolet laser. Most of the hard work was already behind them by 2024, when physicists first kicked the thorium nucleus into its higher-energy state and then pinned down the exact frequency of that jump. FQxI’s own Year in Physics Review for 2024 counted this among the biggest stories of the year, with quantum physicist Ian Durham hailing the arrival of the “first nuclear clock.” But as Elizabeth Gibney, a senior physics reporter at Nature, explained, the one ingredient still missing was a way to lock the laser onto that natural timekeeper and stop its tick from drifting. That is the self-correction both teams have now supplied. A feedback loop watches how much laser light the thorium absorbs, and corrects the laser the instant it drifts off the nucleus’s frequency.
As Gibney reported, both clocks ran reliably for a full day, drifting by only about a second every three million years. That is still well short of the best atomic clocks, which gain or lose a second just once every 40 billion years. Seems like a stark difference in precision for the claim that nuclear clocks could become the most precise ever made, right? Sure, but those atomic clocks have had decades of refinement; these are weeks old.
So, they do not win at telling time yet. But they are already doing physics. Schumm’s team pointed their nuclear clock at ultralight dark matter, hunting for the faint wobble it would leave in the nucleus’s energy levels. They found none, but their limits already rival the best atomic clocks, and on dark matter’s coupling to the strong nuclear force they push past anything measured before. Because the nucleus answers to the strong force as well as to electromagnetism, comparing a nuclear clock with an atomic one is a uniquely sharp way to test whether the fundamental constants really hold still, and to hunt for the new particles and “fifth forces” predicted by theories beyond the Standard Model. “This is an outstanding result,” the theorist Victor Flambaum, who was not involved in the research, told Science News’s senior physics writer Emily Conover. “The race for building super-accurate nuclear clocks just started.”
To actually beat the atomic clocks, Peik told Gibney, thorium will need to be freed from its crystal and isolated, “an important route that remains to be explored.” He is neither surprised nor troubled that the first part took twenty years. Peik, who was honoured last year with the Lower Saxony Science Award for laying the groundwork, has waited this long with good cheer. “I have always been optimistic about the success of this project,” he told Conover. “I am certain that this momentum will continue.”
FQxI has long been interested in what a clock fundamentally is. For the thermodynamic price of keeping ever-better time, see “The Entropic Price of Building the Perfect Clock,” a Q&A with Natalia Ares, and the podcast “The Thermodynamic Cost of Timekeeping” with Marcus Huber.
Explore more:
“A thorium-229 optical nuclear clock with feedback loop,” by L. Toscani De Col et al., including Ekkehard Peik: arXiv
“A nuclear clock based on ²²⁹Th,” by Beichen Huang et al.: arXiv
“Clocks made from an atomic nucleus just ticked on for the first time,” by Emily Conover: Science News
“The first ticking ‘nuclear clocks’ are here, what can they do?” by Elizabeth Gibney: Nature
“The world’s first nuclear clock just ticked on,” by Kenna Hughes-Castleberry: Live Science
“The First Nuclear Clock Will Test if Fundamental Constants Change,” by Joseph Howlett: Quanta Magazine
“Testing Times for Nature’s Constants,” by Sophie Hebden: QSpace
A crystal of calcium fluoride that is infused with thorium atoms (shown) is at the heart of a new nuclear clock.
Luca Toscani De Col/TU Wien