A team at TU Wien has built the first self-stabilizing nuclear clock in collaboration with Germany's Physikalisch-Technische Bundesanstalt (PTB). The prototype ran stably for more than 24 hours without intervention, reaching a precision of about 10^-15 — roughly one second error in 30 million years. That falls short of the world's best optical atomic clocks but represents a remarkable first prototype, according to Thorsten Schumm.
How the clock self-regulates its laser frequency
Conventional atomic clocks use atoms and their electron energy states to stabilize a laser frequency. Schumm's team uses atomic nuclei instead, embedding thorium-229 nuclei in a crystal lattice at room temperature. A calcium fluoride crystal sits at the core, illuminated by laser light. While laser oscillations can in principle measure time, the frequency drifts due to temperature fluctuations and other disturbances. That is why high-precision clocks need a stabilization mechanism. The thorium nuclei absorb laser light only at precise frequencies. When the frequency deviates slightly, absorption drops measurably, triggering an automatic correction. Atomic nuclei are over 10,000 times smaller than atoms and respond far less to external disturbances. The technique relies on continuous absorption spectroscopy at the nuclear transition at 148 nanometers. For validation, the team continuously compared a subharmonic of that radiation with a single Yb+ ion clock at the Austrian Federal Office of Metrology (BEV).
Two decades from proposal to working prototype
PTB physicist Ekkehard Peik proposed the thorium nuclear clock concept in 2003. It took about 20 years to find the exact laser frequency needed. Thorium-229 has two exceptionally close energy states, allowing a single laser to change the nucleus state, unlike other nuclear isotopes with densely packed levels. In April 2024, Schumm's team achieved the first direct detection of this nuclear transition. By autumn 2024, they coupled the apparatus to a conventional optical atomic clock. The initial laser system required several square meters of optical tables and worked only in pulsed mode. In December 2025, Peik's PTB team and the Max Born Institute developed a compact, continuously emitting solid-state laser with very high frequency resolution. This new laser enabled measurement of the power absorbed by the nuclei without delay, allowing the team to lock the laser frequency to the nuclear resonance both short- and long-term.
Vienna and Beijing nearly simultaneous
The same day TU Wien announced its results in Nature, a team at Tsinghua University in Beijing published results on a self-stabilizing nuclear clock also based on thorium-229. According to Chemistry & Engineering News, these represent the first functioning nuclear clocks ever built — a concept physicists had described theoretically since 2003. Chinese media declared their clock superior, but experts note the two systems differ in approach and were not directly compared. The Beijing team measured 6 times better short-term stability but used their own control loop as reference and significantly higher laser power, while Vienna validated stability against a conventional atomic clock. Stability differs from accuracy: stability measures tick uniformity, accuracy measures closeness to true time. Neither team aimed for accuracy in this round. The gap to the world's best optical atomic clocks remains large. C&EN reports they deviate by one second in hundreds of billions of years.
Technical hurdles remain before nuclear clocks match optical standards
TU Wien plans major improvements: stronger lasers and better thorium crystals. Peik identifies clear technical limits: lasers and thorium crystals must improve further for the nuclear clock to become truly competitive. Early physics applications are already underway. The team used the new measurement capability to probe thorium nucleus positions in the crystal lattice and to test for possible couplings to certain forms of dark matter. During a roughly 23-hour measurement, they found no dark matter coupling at any of the frequencies tested. For now, Schumm views Vienna's nuclear clock as a first prototype but one with remarkable promise.