🔬 physics

Phase-continuous comparison of three all-optical time scales over 20 days

This paper demonstrates the creation and 20-day continuous operation of three parallel all-optical time scales using optical flywheels and an optical frequency standard, which achieve superior short-term stability and sub-100 picosecond total time difference compared to traditional maser-based systems, paving the way for the future of optical timekeeping.

Original authors: Dahyeon Lee, Kyungtae Kim, Zoey Hu, Ben Lewis, William Warfield, Kai Zhou, Alejandra Collopy, Jeffrey Sherman, Abijith Kowligy, Parth Patel, Jonathan Roslund, Arman Cingoz, Martin Boyd, Jun Ye

Published 2026-08-06

📖 4 min read☕ Coffee break read

CC BY 4.0

Original authors: Dahyeon Lee, Kyungtae Kim, Zoey Hu, Ben Lewis, William Warfield, Kai Zhou, Alejandra Collopy, Jeffrey Sherman, Abijith Kowligy, Parth Patel, Jonathan Roslund, Arman Cingoz, Martin Boyd, Jun Ye

Original paper licensed under CC BY 4.0 (https://creativecommons.org/licenses/by/4.0/). ✨ This is an AI-generated explanation of the paper below. It is not written or endorsed by the authors. For technical accuracy, refer to the original paper. Read full disclaimer

The Race for the Perfect Tick

Imagine time not as a smooth river, but as a series of tiny, rhythmic ticks. For centuries, humanity has relied on the swing of a pendulum or the vibration of a quartz crystal to count these ticks. But the most precise clocks we have today are like a super-accurate metronome that only works when someone is standing right next to it, watching it. These are "optical frequency standards," the current champions of timekeeping that use light to tick incredibly fast. However, they have a problem: they can't run 24/7 without a break. When they stop, we need a "flywheel"—a backup clock that keeps the rhythm going until the main clock wakes up.

Currently, that backup is usually a hydrogen maser, a type of radio-frequency clock. Think of the maser as a reliable, old-fashioned grandfather clock: it keeps ticking even when no one is looking, but it's a bit wobbly and loses its rhythm quickly. The optical clocks are like a high-tech digital watch that is perfect for a split second but drifts if left alone for too long. The big question in the world of physics is: Can we build a timekeeping system that uses the perfect rhythm of light for the backup, too? If we could, we wouldn't just be measuring time; we would be redefining the very second itself, making our clocks so precise they could detect the bending of space-time or the subtle shifts in the Earth's gravity. This is the frontier where scientists are trying to replace the wobbly grandfather clock with a laser-powered flywheel.

The 20-Day Laser Marathon

In this study, a team of scientists from JILA, NIST, and Vector Atomic decided to test a bold idea: what if we built a time scale using only light? They set up a race between three different "all-optical" time scales, each running on a different type of laser-based flywheel. Two of these flywheels were cryogenic silicon cavities—basically, super-cooled blocks of silicon that act like ultra-stable mirrors for laser light. The third was a commercial iodine optical clock, which uses iodine gas to lock a laser's frequency.

To keep these lasers on track, the team used a "Sr optical frequency standard" (a strontium atomic clock) as the referee. This referee would check the lasers' speed and gently nudge them back to the right pace whenever it was awake. The goal was to see if these laser-only systems could keep time together for a long stretch without drifting apart.

The results were a resounding success. The team ran this experiment for over 20 days. During this time, the three independent optical time scales stayed incredibly synchronized. When they compared the time kept by one laser-clock against another, the difference was less than 100 picoseconds (that's 0.0000000001 of a second) over the entire 20-day period. To put that in perspective, if these clocks were running for the age of the universe, they would still be off by less than a second.

The study found that these all-optical systems reached a level of stability (how much they wobble) that is better than 10⁻¹⁶ after just a few days of averaging. This is significantly faster than traditional systems using hydrogen masers, which would take weeks to reach the same level of precision. Even when the "referee" (the strontium clock) took a break—sometimes for about 6 hours—the optical flywheels kept such a steady beat that they only drifted by about 20 picoseconds. The team also verified that they could convert these ultra-fast optical ticks down to the radio frequencies (100 MHz) that our current electronics use, without losing any of that incredible precision.

The paper explicitly argues against the idea that we must rely on hydrogen masers as the long-term backbone of timekeeping. They show that masers introduce too much "noise" (jitter) and that optical flywheels are not only possible but superior for short-term stability. While the study confirms that these systems work and are stable, it notes that for them to become the global standard, we need more reliable optical clocks that can run with even higher "uptime" (less downtime) and better long-distance fiber networks to share the time signal. But for now, this experiment proves that a future where our time is kept entirely by light is not just a dream, but a reality we can measure today.

Technical Summary: Phase-continuous comparison of three all-optical time scales over 20 days

Problem Statement
Despite the rapid advancement of optical frequency standards over the past two decades, which offer orders-of-magnitude improvements in stability and accuracy compared to radio-frequency (rf) standards, time scales such as Coordinated Universal Time (UTC) have not seen commensurate performance gains. Current time scales rely exclusively on rf flywheel oscillators, primarily hydrogen masers, to maintain continuous timekeeping. These masers exhibit frequency stability roughly a thousand times worse than optical standards. Consequently, when optical standards are used to steer maser-based time scales, the "Dick effect"—noise introduced by incompletely sampled local oscillator noise during steering gaps—imposes a severe performance limit. This necessitates weeks of averaging for a maser-based system to reach the stability an optical standard achieves in one second, leading to a significant underutilization of optical resources.

Methodology
To address this limitation, the authors realized three parallel, independent all-optical time scales. The system architecture consists of:

Optical Flywheels: Three distinct oscillators serving as the time base:
- Two cryogenic silicon cavities (labeled Si3, a 21 cm cavity, and Si6, a 6 cm cavity).
- One commercial molecular iodine optical clock (VA21).
Steering Mechanism: A high-uptime $^{87}$ Sr optical lattice frequency standard was used to measure and steer the frequencies of the three flywheels. The steering algorithm compensated for predictable linear drifts in the silicon cavities (using a 1-minute averaging interval for correction and 10-hour extrapolation for gaps) and utilized a Kalman filter for the iodine clock, similar to approaches used for maser steering.
Comparison and Conversion: The three time scales were compared phase-continuously against each other in both the optical and rf domains. Optical frequency combs were employed to downconvert the optical signals to 100 MHz rf signals, allowing for standard rf-domain analysis while preserving phase coherence.
Duration: The experiment ran for 27 days (MJD 61043–61070), with a specific focus on a 20-day period where the Sr standard operated with ~70% uptime.

Key Results

Stability Performance: The three all-optical time scales achieved a fractional frequency instability of less than $10^{-16}$ after just a few days of averaging. This is significantly faster than the weeks required for maser-based hybrid systems to reach similar stability levels. The measured instability did not exceed mid- $10^{-16}$ at any averaging time and reached low- $10^{-16}$ or below after a few days.
Time Difference Accumulation: During typical steering gaps of approximately 6 hours, the accumulated time difference between the scales was ~20 ps. Even when gaps exceeded 10 hours, the accumulated difference remained manageable. Over the entire measurement period, the peak-to-peak time difference was ~200 ps, with the final relative time difference between the scales remaining below 100 ps. This is substantially lower than the nanosecond-level differences typical of conventional time scales over similar durations.
Flywheel Characteristics:
- Si3: Demonstrated the best short-term stability (over a few hours), making it the optimal flywheel for gaps under 10 hours.
- Si6: Showed no discernible drift over the 20-day period but was terminated early due to a power outage affecting the cryostat.
- VA21: Exhibited a predictable drift of low- $10^{-15}$ per day and short-term stability of $3 \times 10^{-14}$ at 1 s.
Optical-to-RF Fidelity: The conversion of optical signals to 100 MHz rf signals via frequency combs was verified to be highly accurate. The additive timing error from the conversion process was measured to be less than 2 ps, and the frequency division process showed no significant residual slope, confirming the fidelity of the downconversion.

Significance and Claims
The paper claims that all-optical time scales are a feasible and superior alternative to current maser-based systems, particularly in the context of the anticipated redefinition of the SI second by an optical frequency. The authors argue that:

Performance Leap: By utilizing optical flywheels (specifically cryogenic silicon cavities and iodine clocks) with superior short-term stability, the Dick noise limit is drastically reduced, allowing time scales to reach $10^{-18}$ -level stability potential much faster than current hybrid systems.
Reliability: The experiment demonstrates that modern optical components, including cryogenic silicon cavities and commercial iodine clocks, possess the reliability and uptime necessary to function as time scale flywheels, even when steered by optical standards with less than 100% uptime.
Future Outlook: With the proliferation of long-distance optical fiber links and the commercialization of optical flywheels, the authors anticipate that all-optical time scales will become the future of timekeeping, offering a natural realization of the optical second that maintains compatibility with existing infrastructure while providing orders-of-magnitude performance improvements.

The study concludes that cryogenic silicon cavities and commercially available iodine clocks are ready for immediate adoption to improve time scales, bridging the gap between the theoretical performance of optical standards and the practical realization of robust, high-accuracy timekeeping.

The Race for the Perfect Tick

The 20-Day Laser Marathon

Technical Summary: Phase-continuous comparison of three all-optical time scales over 20 days

More like this