Optical clocks with 1E-18 uncertainty

Short Name: OC18, Project Number: 15SIB03
Image showing a Large Modern Hourglass
Large Modern Hourglass

High-accuracy time measurement for optical atomic clocks

Accurate global timekeeping is based on the use of a worldwide standard definition of the second, the SI base unit of time. The second is currently realised using caesium fountain atomic clocks, which use the precise transitions between the energy levels in caesium to measure the frequency of microwave laser light. Optical atomic clocks work through the same principle but use optical light instead, measuring time with accuracy improved by an order of magnitude.

However, previous demonstrations of optical clock performance have been based on estimates. For optical clocks to become a viable candidate to realise the SI second, these uncertainties needed to be reduced to 1x10-18 with experimental verification. To achieve this level, the frequency output of optical clocks must be averaged over a period of time, previously days or weeks. However, practical applications, such as Earth observation and investigations of fundamental science, required this averaging to take place over just a few hours. This in turn required instabilities in the lasers used in optical clocks to be at or below 1x10-16 after 1s.


This project built on the achievements of EMRP projects Ion Clock and  ITOC to improve laser stability in optical clocks. The project demonstrated state-of-the-art laser stabilisation, achieving levels of 1x10-16 after 1s in room-temperature glass vacuum cavities and as low as 4x10-17 after 1s in cryogenically-cooled silicon cavities. To take advantage of this improvement, scattering effects in the optical lattice were studied and new atomic traps were designed to extend the time over which the target atoms can be probed to 1s. These probe sequences were further optimised through a theoretical study to further reduce frequency instability and uncertainty. Frequency shifts caused by background effects such as black body radiation and collisions with background gases in the vacuum chamber were also characterised and controlled, allowing the target uncertainty of 1x10-18 to be achieved. Two independent optical clocks, using ytterbium ions, have been confirmed by the project to be operating at uncertainties of 1x10-18 through a direct extended comparison.


The work of this project has increased the accuracy of optical clocks, supporting their use for the proposed redefinition of the second and for work ‘in the field’”. This will find application in a broad range of scientific and commercial arenas, from aiding in the search for dark matter to improving the service from global navigation satellite systems.


EMPIR project CC4C builds on this work.


Project website
Atomic clocks for geodesy

Reports on Progress in Physics

Autobalanced Ramsey Spectroscopy

Physical Review Letters

Other Participants
Gottfried Wilhelm Leibniz Universität Hannover (Germany)
Kobenhavns Universitet (Denmark)
Uniwersytet Mikolaja Kopernika w Toruniu (Poland)