Two independent teams have built the first working clocks based on transitions inside atomic nuclei, reporting fractional frequency instabilities down to 5 x 10⁻¹³/√τ/s and publishing the results in Nature. Both devices tick via an ultraviolet transition in thorium-229 nuclei embedded in millimetre-sized calcium fluoride crystals — the culmination of more than 20 years of research into a concept first proposed in the late 1990s.

The achievement matters for metrology budgets and fundamental-physics portfolios alike. Today's optical atomic clocks, which underpin satellite navigation and telecommunications timing, also feed dark-matter searches and gravitational-wave detection. Nuclear clocks could, in principle, outperform them: nuclei are less disturbed by electromagnetic fields, and their UV transitions tick at higher frequencies, enabling finer time measurement.

Who built the clocks?

The first team is led by Ekkehard Peik at the Physikalisch-Technische Bundesanstalt (PTB) in Germany and Thorsten Schumm of TU Wien in Austria, with colleagues in Germany, Austria and Czechia. In 2003, Peik and his then-colleague Christian Tamm identified the thorium-229m transition as a promising clock basis. The group went on to demonstrate laser excitation of that nuclear transition and to embed thorium-229 in CaF₂ crystals, creating a solid-state device that — unlike optical clocks — needs no trapped atoms or ions.

The second team, led by Shiqian Ding at Tsinghua University, includes collaborators at 13 other institutions across China. Its prior milestones include the first continuous-wave narrow-linewidth UV laser at the critical 148.4 nm wavelength. Because China has limited access to thorium-229, the team also developed a crystal-growth approach that requires only 1.4 μg of the isotope — a supply-chain constraint turned into an engineering result.

How do the two devices compare?

Both teams drove the embedded thorium with narrow-linewidth continuous-wave UV lasers, using different nonlinear processes to generate the light. Both monitored excitation of the nuclear resonance via direct absorption spectroscopy through the crystal and locked the laser frequency onto the transition with a feedback loop, then compared the stabilized nuclear clocks continuously against frequency standards derived from atomic clocks.

Here the approaches diverged:

• The PTB-TU Wien team embedded more thorium-229 into its crystals.
• The Tsinghua team generated higher laser powers.
• Both features improve signal-to-noise, but the Tsinghua device achieved the lower fractional frequency instability: 5 x 10⁻¹³/√τ/s versus 3 x 10⁻¹²/√τ/s for the PTB-TU Wien collaboration.

The teams also used their devices differently. Peik's group applied its clock to fundamental physics, constraining ultralight scalar dark-matter couplings and fluctuations or drifts in the nuclear transition energy. The resulting limits on how dark matter couples to photons and the strong nuclear force are comparable to those from today's best atomic clocks. Ding's team focused on technical questions — high-power vacuum ultraviolet interrogation, clock performance, inter-crystal reproducibility and metrological consistency — and found that transition frequencies measured in two independently fabricated crystals agreed with each other and with earlier VUV frequency-comb measurements made at JILA in the US.

What limits performance now?

Measured performance still sits far from theoretical limits. Peik said: "We have now seen 30 kHz linewidth, but the natural linewidth should be way below 1 Hz, which would enable more stable clock operation." His team is now boosting laser power, improving detector signal-to-noise, and studying the structure of the thorium dopant to understand what sets the observed linewidth.

Ding identified better host crystals than CaF₂ as the next step — one that could narrow the transition and improve both stability and accuracy. He also wants a trapped-ion thorium clock that reuses quantum-control techniques from atomic and ion clocks: "It is technically much more difficult, but I think this route may ultimately provide the highest accuracy."

Victor Flambaum, a theorist at the University of New South Wales who was not involved in either project, called the devices the "first prototypes of nuclear clocks" and noted their promise for detecting new physics effects expected to be very small. "We have shown that in nuclear clocks these effects are strongly enhanced," he said.

Whether nuclear clocks displace optical atomic standards will depend on closing the linewidth gap and scaling thorium-229 supply — but with two independent prototypes now measured, characterized and cross-checked against JILA data, the field has its first experimental baseline to improve on.