Optical frequency combs in atomic clock comparisons across three papers

Short answer: An optical frequency comb links the very different frequencies of optical atomic clocks so that two clocks can be compared directly. Three papers report comparisons with fractional instabilities and agreements at the 10^-17 to 10^-19 level, including a 2020 free-space link and a 2019 preprint on two strontium clocks. These are laboratory results from the authors, and one is an arXiv preprint.

Why compare clocks through a comb?

The 2017 Applied Physics Express abstract notes that all-polarization-maintaining, single-port erbium fibre combs offer long-term robust operation and high stability. The authors built two such combs and evaluated the transfer noise for linking optical clocks. The comb spectrum is uniformly broadened over 135 to 285 THz with a high signal-to-noise ratio. That allows optical frequency measurement of subharmonics of strontium, ytterbium and mercury optical lattice clocks, with fractional frequency-noise power spectral density of (1 to 2) x 10^-17 per root hertz at 1 Hz. By applying a synchronous clock comparison, they say the comb enables clock ratio measurements with 10^-17 instability at 1 second, one order of magnitude smaller than the best instability of the frequency ratio of optical lattice clocks that they cite.

What did the 2019 preprint do?

An arXiv preprint, which may not have been peer reviewed, says optical atomic clocks need local oscillators with exceptional optical coherence because their clock transitions have ultra-narrow linewidths. The authors use a local oscillator built on a cryogenic silicon reference cavity, along with a state-of-the-art frequency comb for coherence transfer, with two strontium optical lattice clocks. An anti-synchronous comparison gives a fractional instability for both clocks that the authors assess to be above 4.8 x 10^-17 over the square root of the averaging time in seconds. Synchronous interrogation reveals a quantum-projection-noise-dominated instability of 3.5(2) x 10^-17 over the square root of averaging time, which gives a precision of 5.8(3) x 10^-19 after a single hour of averaging. The authors say the ability to measure shifts below 10^-18 in such short times will affect applications in quantum sensing and fundamental physics.

What did the free-space comparison show?

A 2020 paper, published in Physical Review Research according to the arXiv record, uses frequency-comb-based optical two-way time-frequency transfer to measure the optical frequency ratio of ytterbium and strontium clocks separated by a 1.5 km open-air link. The free-space measurement is compared with a simultaneous measurement over a noise-cancelled fibre link. Despite non-stationary, picosecond-level time-of-flight variations in the free-space link, ratio measurements from the two links, averaged over 30.5 hours across six days, agree to 6 x 10^-19. The authors conclude that the method can support free-space atomic clock comparisons below the 10^-18 level.

How to read this

The three papers measure different things: comb transfer noise, clock stability, and agreement between two links. The numbers cannot be lined up as one scale. All values are the authors' own, and this post is based on abstracts and visible introductions. Full texts were not checked. It makes no claim about products or field use.

FAQ

What does a frequency comb do in a clock comparison?

It transfers coherence and frequency information between clocks or lasers, per the three abstracts.

What is the best precision reported here?

5.8(3) x 10^-19 after one hour of averaging, in the 2019 preprint.

Can clocks be compared over open air?

A 1.5 km free-space link agreed with a fibre link to 6 x 10^-19 in the 2020 paper.

Is the 2019 work peer reviewed?

The arXiv version is a preprint, and its publication status was not checked.