Two papers published in Nature on 7 October describe clocks that keep time by locking a laser to a transition inside the nucleus of thorium-229. They are not yet more accurate than the best atomic clocks, but they open a route to timekeeping — and tests of fundamental physics — that physicists have pursued for more than two decades.
Every modern timekeeping system, from the satellites that guide a phone’s map to the networks that stamp financial transactions, ultimately relies on atoms. Atomic clocks count the oscillations of light absorbed or emitted when electrons jump between energy levels; the caesium transition still defines the second. On Wednesday, 7 October, two research teams reported that they had built a different kind of clock, one that listens not to electrons but to the atomic nucleus itself.
The papers, published together in Nature, come from a collaboration led by Thorsten Schumm of TU Wien in Austria and Ekkehard Peik of Germany’s national metrology institute, PTB, and from a team led by Shiqian Ding at Tsinghua University in Beijing, working with collaborators at 13 other Chinese institutions. In an accompanying commentary, Akio Kawasaki of Japan’s National Metrology Institute described the work as the first implementations of a nuclear clock and a milestone in metrology.
Why thorium-229 is special
Nuclear energy levels are usually separated by enormous energies, requiring gamma rays far beyond the reach of any laser. Thorium-229 is the exception. Its nucleus has a low-lying excited state, an “isomer”, only about 8.4 electronvolts above the ground state — within reach of vacuum-ultraviolet light. Physicists Ekkehard Peik and Christian Tamm proposed a clock based on this transition in 2003, but its exact energy remained elusive for years. The breakthrough came in 2024, when Schumm’s team and others excited the nucleus with a laser for the first time, and a group at JILA in Colorado measured the frequency precisely using a vacuum-ultraviolet frequency comb.
The appeal is fundamental. The nucleus is tens of thousands of times smaller than the atom and shielded by its electron cloud, making it less sensitive to stray electric and magnetic fields that disturb atomic clocks. That suggests nuclear clocks could, in time, be both extremely stable and robust enough to be built into solid crystals rather than delicate traps of single ions or atoms.
How the two clocks work
Both teams embedded thorium-229 in millimetre-sized crystals of calcium fluoride, which is transparent to ultraviolet light. They shone narrow-linewidth, continuous-wave ultraviolet lasers through the crystal and watched how much light was absorbed. When the laser frequency drifted off the nuclear resonance, absorption fell; a feedback loop then steered the laser back. The nucleus, in effect, disciplines the laser — the defining step that turns a spectroscopy experiment into a clock. Each device was compared continuously against frequency standards derived from atomic clocks.
The approaches differed in emphasis. The PTB–TU Wien team packed more thorium-229 into its crystals and reported a fractional stability of 3 × 10⁻¹² divided by the square root of the averaging time in seconds. Interesting Engineering, reporting TU Wien’s account, said the Vienna clock ran unattended for more than 24 hours. The Tsinghua team built what it describes as the first continuous-wave narrow-linewidth laser at 148.4 nanometres, achieved a better stability figure of 5 × 10⁻¹³ per root second, and developed a method of growing usable crystals with just 1.4 micrograms of thorium-229 — an important feat given how scarce the isotope is. Its measured transition frequencies agreed across two independently made crystals and with JILA’s earlier measurements.
Popular reports have translated the performance into everyday terms, stating that the clocks would lose about one second in 30 million years. Such conversions depend on averaging time and should be treated as illustrative.
Not yet the best — and why that is acceptable
The researchers are clear that these are prototypes. The best optical atomic clocks, based on strontium or ytterbium atoms or single trapped ions, are currently more stable and accurate. One limitation is the width of the observed nuclear resonance: Peik told Physics World that his team has seen a linewidth of about 30 kHz, whereas the natural linewidth should be well below 1 Hz. Interactions between thorium nuclei and their crystal surroundings are thought to broaden the line, so understanding and engineering the host crystal is now a priority. Higher laser power and better detectors are also on the agenda; the Tsinghua group ultimately aims to build a trapped-ion thorium clock, which Ding believes may provide the highest accuracy.
What nuclear clocks could do
The prize is not merely a better second. Because the thorium transition depends on a delicate balance between the strong nuclear force and electromagnetism, it is expected to be unusually sensitive to any variation in the fundamental constants of nature. The PTB–TU Wien team has already used its clock to search for ultralight dark matter — hypothetical fields that would make such constants oscillate — and set limits comparable to those from leading atomic clocks. Victor Flambaum of the University of New South Wales, who was not involved, called the devices the “first prototypes of nuclear clocks” and said they could become the “most accurate instrument humans have ever built”.
Practical applications are further off but plausible: compact solid-state clocks for navigation where satellite signals are unavailable, and clocks so sensitive to gravity’s effect on time that they could map density variations underground — relevant to geodesy, mineral exploration and earthquake research. Schumm has said his group envisions an eventual chip-scale device and plans a second clock in Innsbruck within a year so that two nuclear clocks can be compared directly.
Relevance for India
India maintains its national time standard at CSIR-National Physical Laboratory in New Delhi, which is developing optical clocks and pursuing the government’s “One Nation, One Time” initiative to distribute Indian Standard Time with high precision. Indian research groups in atomic physics and quantum technologies, supported under the National Quantum Mission, are well placed to contribute to the theory and spectroscopy of thorium-229. Access to the isotope and to vacuum-ultraviolet laser technology will be the practical hurdles.
–Yadavalli VR Srinivas
Key facts
- Papers: Toscani De Col et al., “A thorium-229 optical nuclear clock with feedback loop” (doi:10.1038/s41586-026-11084-4); Huang et al., “A nuclear clock synchronized to ²²⁹Th” (doi:10.1038/s41586-026-11122-1); Nature, 7 Oct 2026
- Teams: PTB (Germany) and TU Wien (Austria), led by Ekkehard Peik and Thorsten Schumm; Tsinghua University, led by Shiqian Ding
- Method: thorium-229 in calcium fluoride crystals; laser locked to the ~8.4 eV nuclear transition (148.4 nm)
- Stability: 3 × 10⁻¹²/√τ (PTB–TU Wien); 5 × 10⁻¹³/√τ (Tsinghua); observed linewidth ~30 kHz vs natural linewidth expected well below 1 Hz
- Status: prototypes; not yet better than leading optical atomic clocks
