The first nuclear clocks are ticking in Vienna and Beijing after a 50-year chase

A nuclear clock, one that counts time with the core of an atom instead of its electrons, has finally started ticking. In Vienna, a laser locked onto thorium nuclei sealed inside a crystal kept its own rhythm for more than a full day without anyone touching it, and a team in Beijing built a second clock of the same kind on its own.

That makes them the first working nuclear clocks, a device physicists have chased for decades. The nucleus is tiny and shielded by its own electron cloud, so stray fields, heat and vibration barely reach it. A clock built on it could one day keep time more steadily than the atomic clocks that run satellite navigation, the internet and the definition of the second.

The Vienna clock comes from Thorsten Schumm’s group at TU Wien and Ekkehard Peik’s group at PTB, Germany’s national metrology institute. The Beijing clock comes from a Tsinghua University team led by Shiqian Ding, working with 13 other Chinese institutions. Both papers appeared in the same issue of Nature, and both use the same rare isotope, thorium-229.

“The creation of a nuclear clock was something that physicists dreamt of for almost 50 years,” Schumm told Reuters. His own team has worked toward it since 2008.

How a crystal and a laser keep time

Every clock needs something that swings at a fixed rate. In a pendulum clock it is the pendulum. In an atomic clock it is the light an electron absorbs when it jumps between two energy levels, which in the best modern optical clocks oscillates far faster than any mechanical part could.

A nuclear clock moves that job into the nucleus. Protons and neutrons also have energy levels, but almost all of them sit far too high for any laser to reach. Thorium-229 is the exception: its nucleus has an excited state low enough to be driven by vacuum ultraviolet light at a wavelength of about 148 nanometres. Hitting it is still hard. The laser has to match the transition energy to within one millionth of an electron volt, according to PTB.

Schumm’s team grew calcium fluoride crystals doped with thorium-229, millimetre-sized pieces that hold vast numbers of nuclei at room temperature. “The basic idea is simple: you have a laser and you have thorium,” Schumm said. The nuclei absorb light only at one precise frequency. If the laser drifts, for instance because the room warms slightly, the crystal absorbs less, a detector registers the drop and a feedback loop steers the laser back. The nuclei become the reference, and the corrected laser frequency becomes the clock’s tick.

Two technical changes made this possible. PTB and the Max Born Institute in Berlin built a compact solid-state laser that emits continuously; the older system filled several square metres of optical tables and fired only in pulses. The teams also stopped counting the faint glow the nuclei give off after excitation, which took around ten minutes to read, far too slow for steering a laser. Instead they measure how much laser power the crystal swallows, a signal fast enough to close the loop.

Vienna against Beijing

The Vienna clock ran for more than 24 hours unattended while being compared with a conventional optical atomic clock. Its fractional instability was about 3 × 10⁻¹² over one second of averaging, and that scatter shrank toward 10⁻¹⁵ over a day. TU Wien translates that to an error of roughly one second in 30 million years, a figure measured over a day-long run, not observed over millions of years.

The Beijing clock is steadier over short times. Its instability over one second is about six times lower than Vienna’s, according to figures reported by Physics World. Ding’s team built its own continuous laser at 148.4 nanometres with more power, and it learned to grow small crystals from just 1.4 micrograms of thorium-229, because China has limited access to the isotope. Two independently grown crystals gave the same transition frequency, which also matched earlier measurements by JILA in Colorado.

Schumm sees the two results as complementary. The Vienna crystals carry a higher concentration of thorium and have better optical properties, he told Reuters, while the Beijing team has the stronger laser. “So already by putting these components together, we can build a significantly better clock,” he said.

“The two teams worked independently and reached operating thorium-229 nuclear clocks at the same time, using different experimental approaches,” Ding told Reuters, adding that this shows “the concept is robust.”

What these clocks do not do yet

Neither clock beats the atomic clocks it was measured against. “This is not yet at the level of the world’s best optical atomic clocks,” Schumm said. Peik put it more bluntly in the PTB announcement: the stability “does not yet set any world records.”

The biggest gap is the sharpness of the resonance. “We have now seen 30 kHz linewidth, but the natural linewidth should be way below 1 Hz,” Peik told Physics World. A blurred resonance means a blurred tick. The crystal itself is part of the puzzle: thorium sits in several different sites in the calcium fluoride lattice, and the teams are still working out how that shapes the resonance. Ding has said that calcium fluoride host crystals limit performance and that better host materials are needed.

Both machines are also laboratory prototypes. They depend on reference clocks for comparison, on custom ultraviolet lasers and on an isotope that is hard to obtain. A rugged device that a ship, a satellite or a data centre could carry is still a long engineering path away.

Why physicists want a better clock

Clocks already decide more of daily life than most people notice. Satellite navigation works out position from tiny differences in signal timing. Data networks are synchronised against atomic time. Surveyors and geophysicists use the fact that clocks tick at different rates at different heights in Earth’s gravity.

A nuclear clock adds two things. First, a solid-state design, with no trapped atoms or ions held in a vacuum chamber, could eventually make precise timekeeping smaller and tougher. Second, the thorium transition depends on the forces inside the nucleus, so it should react far more strongly than atomic transitions if a constant of nature drifts or if some form of dark matter pulls on ordinary matter. The Vienna team already ran that test with its prototype. It found no dark matter signal, but its limits on dark matter coupling to light and to the strong force are already comparable to those set by the best atomic clocks.

Victor Flambaum of the University of New South Wales, who was not involved, called the devices the “first prototypes of nuclear clocks.” Schumm’s verdict was wider: “It gives access to a whole new physics universe.”

Common questions about the nuclear clock

What is a nuclear clock?

A nuclear clock keeps time using a transition inside an atomic nucleus instead of a jump between electron energy levels. A laser is tuned to the exact frequency the nucleus absorbs, and that frequency becomes the clock’s tick. Thorium-229 is the only known nucleus with a transition low enough in energy to reach with a laser.

Is a nuclear clock radioactive or dangerous?

It has nothing to do with fission or fusion and produces no nuclear energy. Thorium-229 is a radioactive isotope, but the clocks use tiny amounts of it locked inside a crystal, and the laser only nudges the nucleus into a slightly higher energy state.

Is a nuclear clock more accurate than an atomic clock?

Not yet. The Vienna and Beijing prototypes are less stable than the best optical atomic clocks. The expectation is that the nucleus, being far smaller and better shielded than the electron shell, will eventually allow better performance once lasers and crystals improve.

What could nuclear clocks be used for?

Researchers point to satellite navigation, network synchronisation, surveying and fundamental physics. Because the thorium transition is highly sensitive to the forces inside the nucleus, these clocks are also tools for searching for dark matter and for changes in the constants of nature.

Stronger lasers, better crystals, single ions

The thorium nucleus was first excited with a laser in April 2024, by Schumm’s and Peik’s teams, decades after Peik proposed the nuclear clock principle in 2003. That autumn the setup was linked to a conventional optical atomic clock. The self-stabilising Vienna clock is described in a Nature paper published on 7 October 2026, alongside the Tsinghua paper in the same issue and a companion PTB and TU Wien study of continuous-laser absorption by the thorium nucleus, published in Nature in September.

The next steps are concrete. Peik’s group plans to raise laser power, improve the detector’s signal-to-noise ratio and map how thorium sits in the crystal to find what blurs the resonance. Ding’s team is searching for host crystals better than calcium fluoride and is interested in a clock built on single trapped thorium ions, which Ding called “technically much more difficult” but possibly the route to the highest accuracy.

References: Toscani De Col et al., “A thorium-229 optical nuclear clock with feedback loop,” Nature, 2026. DOI: 10.1038/s41586-026-11084-4; Huang et al., “A nuclear clock synchronized to 229Th,” Nature, 2026. DOI: 10.1038/s41586-026-11122-1

Tags: physics, atomic clock, nuclear clock, PTB, thorium-229, Tsinghua University