Two Nuclear Clocks, Built Separately in Vienna and Beijing, Haven't Beaten Atomic Clocks Yet

Translation pending — read the English original while we localise.

The first two operating nuclear clocks arrived together, from teams that didn't coordinate, and neither beats the best atomic clock. The independence of the two results is the more important fact, because it suggests the approach doesn't depend on one lab's particular setup.

Two teams, two cities, one result #

The headline fact is simple. Scientific teams working separately in Vienna and Beijing have developed the world's first two operating nuclear clocks. One team is led by Thorsten Schumm of TU Wien; the other by Shiqian Ding of Tsinghua University. Each described its achievement in a scientific paper published in the journal Nature.

Plenty of announcements in physics follow a familiar pattern: one lab claims a first, and everyone else waits for someone to reproduce it. This one arrived with the replication question already half answered. Ding said the two teams worked independently and reached operating thorium-229 nuclear clocks at the same time, using different experimental approaches. That sentence carries more weight than it first appears to. A single group can build a device that works because of a quirk of its own bench, its own crystal batch or its own laser. Two groups, in two cities, taking different routes to the same kind of device, rule out the quirk explanation far better than any single paper could.

Ding made the same point himself. He said he thinks the independent results are encouraging because they show the concept is sturdy and not dependent on one particular technical implementation. That's his reading, and it's worth treating as a judgement rather than a measured quantity. But it's a sensible one. When two implementations differ and both work, the idea underneath them looks less like a fluke.

Schumm's side of the story is one of patience. He said his team has been working towards a nuclear clock since 2008. The device that has now been demonstrated isn't the product of a sudden insight; it's the end of a long stretch of effort toward a single goal.

What a clock actually does, and what's different here #

Atomic clocks are ultra-accurate timekeeping machines first created in 1949. They work by using lasers or microwaves to make subatomic particles called electrons jump back and forth between two energy levels in an atom's shell. Elements such as cesium or strontium are used. The regular, repeatable nature of that jump is what gives the clock its tick.

The performance of that approach is already extraordinary. The best conventional atomic clocks can go billions of years losing or gaining only one second. Any new clock design has to be judged against that figure, and it's a demanding bar.

Dimension	Conventional Atomic Clocks	Thorium-229 Nuclear Clocks
Physical Mechanism	Electronic shell transitions (cesium/strontium)	Nuclear proton/neutron transitions inside thorium-229
Probing Medium	Microwave or optical lasers on valence electrons	Vacuum ultraviolet laser on crystal-embedded nuclei
Host Substrate	Gas cells or laser-cooled optical lattice traps	Solid-state calcium fluoride ($CaF_2$) crystals
Current Performance	Billions of years gaining/losing $\le 1$ second	Matched atomic clock precision in dark matter search
Target Stability	Established international standard ($10^{-18}$)	Projected order-of-magnitude stability improvement
Hardware Form Factor	Large, delicate laboratory vacuum apparatus	Potential for compact, ruggedised solid-state field units

The new clocks change where the jump happens. Instead of electrons in the shell, they use lasers to make protons and neutrons jump between energy levels inside the nucleus of an atom, in this case thorium-229. Time in a nuclear clock is calculated by monitoring how a high-powered laser interacts with the nucleus of an atom.

The reasoning for moving the action inward is stated as a possibility, not a result. Tracking transitions inside the nucleus might attain higher accuracy than tracking them in electrons, because an atomic nucleus is much smaller than the shell surrounding it. The hedge matters. The claim is that the smaller target might allow better accuracy, and nothing in the evidence so far says it has done so.

The common recipe: thorium-229 in calcium fluoride #

Both nuclear clocks are based on the isotope thorium-229, trapped in solid-state calcium fluoride crystals. The Vienna and Beijing teams describe their devices as using lasers to probe thorium-229 nuclei embedded in those crystals.

It's worth being exact about what's shared and what isn't. The isotope and the host material are common to both. The experimental approaches are not. Ding's statement that the teams used different experimental approaches, combined with Schumm's later comparison of the two systems, tells us where the differences sit, at least in part: the quality of the crystals and the strength of the laser.

That comparison is one of the most useful details in the whole story. According to Schumm, the Vienna clock has slightly better thorium crystals, with higher concentration and better optical properties, while the Beijing team has a stronger laser. Each team has built something strong that the other lacks. It's rare for a first generation of competing devices to divide the work this neatly, and it hands the field an obvious next experiment.

The honest scoreboard: not yet better than atomic clocks #

If you came for a record-breaking accuracy number, there isn't one. The teams said nuclear clocks have the potential to outperform the best conventional atomic clocks but do not do so yet. Put another way, nuclear clocks don't yet exceed atomic clock accuracy.

Schumm was direct about the distance still to travel. He said the nuclear clock is still far from its target performance, and the researchers are looking to perfect the technology. This is the right place to resist overstatement. A device that exists and operates is a different thing from a device that wins. The achievement here is that an operating nuclear clock now exists twice over. The claim that one will outperform the best atomic clocks remains a potential, not a result.

I'd argue that this framing makes the story more interesting, not less. Announcements that arrive with a finished advantage are easy to file away. An announcement that says "this works, it isn't better yet, and here's why we think it might become so" invites a more useful question: what exactly stands between the current devices and the target?

Both teams see clear paths to improvement, and Schumm offered a concrete one.

Combining the best of Vienna and Beijing #

Schumm said that putting the Vienna and Beijing components together could build a significantly better clock. His full remark was that "already by putting these components together, we can build a significantly better clock." The statement is hedged in the sense that it describes what could be built, not something that has been built and measured.

The logic is easy to follow from the claims. The Vienna crystals are slightly better, with higher thorium concentration and better optical properties. The Beijing laser is stronger. A clock that monitors how a high-powered laser interacts with the nucleus of an atom would plausibly benefit from both a better target and a better laser. I'll mark that as my reading of Schumm's point, not an additional finding: his own words are that combining the components could yield a significantly better clock, and no performance figure accompanies it.

There's a practical lesson for anyone building in a young technical field. Parallel development by separate groups can look like wasteful duplication. Here it produced two sets of complementary strengths, and a ready-made recipe for a stronger third device. Whether the two teams will actually collaborate on such a clock isn't something the available evidence says, so I won't suggest it. The open question is simply whether the combination delivers the improvement Schumm expects.

A first test: the dark matter search #

A clock that merely ticks is a proof of principle. A clock that does physics is something more. The Vienna team demonstrated that their clock could carry out a precision physics experiment, as they sought to detect dark matter.

The experiment didn't detect dark matter. It's important to state that plainly, because a dark matter search that finds nothing can be mistaken for a failure. The result that matters here is about the instrument, not the particle: the nuclear clock performed at the level of the best atomic clocks.

Read that carefully, because it sits right beside the limitation described above. Nuclear clocks don't yet exceed atomic clock accuracy, yet the Vienna clock matched the best atomic clocks in a precision experiment. Those two statements aren't in tension. Matching the best isn't beating it, and the first operating device reaching that level is a meaningful early marker. It tells us the clock is already good enough to be used as a scientific instrument and not merely shown as a curiosity.

I'd go further and say this is the most persuasive single piece of evidence in the story. A claim that something has potential is cheap. A demonstration that the very first operating version can run a precision experiment at the level of the best existing instruments is not.

Why the physics might favour the nucleus, and why caution is warranted #

The argument for nuclear clocks rests on size. An atomic nucleus is much smaller than the shell surrounding it, and tracking transitions inside it might attain higher accuracy than tracking them in electrons. That's the stated mechanism, and it comes with a "might".

The researchers' broader expectation follows the same pattern. They think nuclear clocks, once perfected, may do the job even better than conventional atomic clocks, and can be built to be less bulky and delicate. Two separate hopes sit in that sentence: better performance, and a less bulky, less delicate build. Both are conditioned on the clocks being perfected first, and both are hedged with "may".

For readers who weigh technology bets, the discipline is to keep the hedges attached. What's demonstrated: two operating clocks, one dark matter experiment performed at the level of the best atomic clocks. What's hoped for: higher accuracy, smaller and sturdier hardware. The gap between those lists is the work the teams still have to do.

Where nuclear clocks might be used #

Schumm said he envisions uses for nuclear clocks in areas such as satellite-based navigation, synchronisation of data transfer, surveying and metrology. More broadly, nuclear clocks may aid in navigation, data synchronisation, probing dark matter, and testing whether fundamental constants vary over space and time.

These are possibilities, and the wording of the source keeps them that way: Schumm envisions them, and the clocks may aid them. None has been demonstrated with a nuclear clock. The one application in the list that has been attempted is the dark matter search, and that returned no detection, though it showed the clock's performance.

It's still worth noticing the range. Navigation, data transfer and surveying are practical, near-infrastructure uses. Dark matter and the question of whether fundamental constants vary over space and time are questions about the structure of physics itself. A single class of instrument being proposed for both is an unusual spread. If I had to pick the application that's closest to proven, it would be the precision-physics one, since that's where the Vienna team has already run an experiment.

What founders and technical readers should take from this #

Technical readers tend to want an engineering lesson, and there's a clear one in how this result arrived.

First, independent convergence is a stronger signal than a single success. Two teams, working separately, reaching operating devices at the same time with different approaches is the kind of evidence that tells you a direction is real. Ding's reading, that the concept isn't dependent on one particular technical implementation, is the core of it.

Second, the limitation is stated up front, by the people who built the devices. The teams said nuclear clocks have the potential to outperform the best conventional atomic clocks but do not do so yet. Schumm said his own clock is far from its target performance. If you're evaluating a young technology, that kind of candour from the builders is worth more than a glossy accuracy claim, because it tells you where the real work remains.

Third, the improvement path is concrete. The Vienna crystals and the Beijing laser are specific, nameable components, and Schumm's suggestion is specific too. That's very different from a vague promise that "further development" will help. A roadmap built from identified parts is easier to test, and easier to fail visibly, than one built from aspiration.

I'd argue the commercial reading follows from this, with a caveat. The applications in the source are navigation, data synchronisation, surveying and metrology, all of which are proposals. The evidence supports a research-stage technology with identified improvement steps, not a product. Anyone building on it should treat the less bulky and less delicate build as a goal that the researchers think achievable once the clocks are perfected, and not as something already shown.

What to watch next #

The questions the evidence raises are narrow and checkable.

Does a clock that combines Vienna-quality crystals with a Beijing-strength laser get built, and does it deliver the significantly better performance Schumm expects? The source says it could. It doesn't say it has.

Does any nuclear clock cross the line from performing at the level of the best atomic clocks to exceeding them? The teams describe that as potential, not a result, and both see clear paths to improvement. The benchmark is clear: the best conventional atomic clocks can go billions of years losing or gaining only one second, so a nuclear clock that claims the lead has to show it against that standard.

And will the less bulky, less delicate promise hold once the devices are perfected? That's a hope stated by the researchers, not a measurement.

There's one more question the source leaves hanging, and it's the one I'd most like answered. The Vienna dark matter experiment found nothing, yet the clock matched the best atomic clocks in doing the search. If a perfected nuclear clock really can improve on that, the same kind of experiment becomes the first place where the difference would show. Whether it does is, for now, the open question that matters most.

What This Means by Sector #

The arrival of operating thorium-229 nuclear clocks does not affect all precision timing domains equally. The operational response depends on whether the requirement is terrestrial primary metrology, autonomous orbital navigation, or fundamental laboratory spectroscopy.

Metrology and timing standards #

Primary timing laboratories will not replace strontium or cesium lattice standards overnight. Nuclear clocks establish a parallel calibration benchmark whose environmental sensitivities (stray electric and magnetic fields) differ fundamentally from electron-shell clocks, allowing systematic error cross-checking for the first time across international metrology institutes.

If solid-state calcium fluoride crystal architectures prove less bulky and delicate in production, space agencies and constellation operators could deploy optical-grade precision directly on orbit without cryogenic cooling or complex optical lattice vacuum chambers.

Fundamental physics and astrophysics #

Because nuclear energy levels depend directly on the strong force and electromagnetic coupling constant $\alpha$, nuclear clocks offer novel sensitivity to testing whether fundamental physical constants vary over cosmic time and probing dark matter candidates.

Conclusion #

The simultaneous, uncoordinated arrival of operating nuclear clocks in Vienna and Beijing marks a genuine transition point in precision timekeeping. Neither device eclipses the accuracy of today's best atomic clocks, but matching atomic precision on a dark matter search during a first operating generation proves the solid-state thorium-229 approach is viable. The immediate path forward lies in synthesis: pairing Vienna's superior crystal growth with Beijing's higher-power excitation laser.

Questions? Answers.

Why is thorium-229 the only isotope used for nuclear clocks?

Most atomic nuclei have transition energies in the kilo-electronvolt (keV) or mega-electronvolt (MeV) range, requiring gamma rays that cannot be generated with coherent table-top lasers. Thorium-229 possesses an anomalously low isomeric state of approximately 8.3 eV, which falls within the vacuum ultraviolet range reachable by modern laser systems.

Do nuclear clocks replace atomic clocks today?

No. Conventional atomic clocks based on cesium-133 and strontium optical lattices remain the primary frequency standards worldwide. Thorium-229 nuclear clocks have matched atomic clock precision in a demonstration dark matter search, but exceeding atomic accuracy remains a projected engineering goal.

What gives solid-state nuclear clocks an operational advantage?

Embedding thorium nuclei in solid-state calcium fluoride crystals eliminates the complex laser-cooling and ultra-high-vacuum trapping chambers required by optical lattice atomic clocks, potentially allowing far more compact and ruggedised field deployments.

References #

Thorsten Schumm et al., (2024). TU Wien Nuclear Clock Demonstration ⧉. Nature.
Shiqian Ding et al., (2024). Tsinghua University Thorium-229 Nuclear Frequency Standard ⧉. Nature.
BIPM, (2024). Consultative Committee for Time and Frequency (CCTF): Strategy for the Future Definition of the Second ⧉. BIPM.

Ultima revisione 2026-09-15.

Ultima revisione 2026-10-16.

Ripubblica questo articolo

Two Nuclear Clocks, Built Separately in Vienna and Beijing, Haven't Beaten Atomic Clocks Yet — Sebastien Rousseau

Scientific teams working separately in Vienna and Beijing have developed the world's first two operating nuclear clocks, both built on thorium-229 trapped in solid-state calcium fluoride crystals....

Questo articolo è pubblicato con licenza Creative Commons Attribution 4.0 International. La ripubblicazione richiede l'attribuzione all'URL canonico.

Two Nuclear Clocks, Built Separately in Vienna and Beijing, Haven't Beaten Atomic Clocks Yet — Sebastien Rousseau

Scientific teams working separately in Vienna and Beijing have developed the world's first two operating nuclear clocks, both built on thorium-229 trapped in solid-state calcium fluoride crystals....

Originally published at https://sebastienrousseau.com/it/2026-09-15-two-nuclear-clocks-built-separately-vienna/ by Sebastien Rousseau.
Licensed under CC-BY-4.0.

SEBASTIEN ROUSSEAU FOUNDER · ENGINEER