Comparing high-accuracy clocks to open up new opportunities in fundamental physics and technological applications
Femtosecond combs and optical fibre links allow clocks operating at different frequencies and in different locations to be compared at accuracies commensurate with their performance. We participate in pan-European clock comparisons designed to check the international consistency of the new generation of optical clocks and to determine their operating frequencies with the lowest possible uncertainty relative to current primary frequency standards. This work is paving the way for a redefinition of the second.
The ability to compare optical clocks with high accuracy is of interest well beyond the metrology community. Experiments performed at NPL, and in collaboration with European partners, have provided new tests of fundamental physical theories, such as Lorentz invariance and the time invariance of fundamental constants. In another collaborative European project we demonstrated that it is possible to measure gravity potential differences by comparing well-separated optical clocks linked by optical fibre.
The enhanced stability and accuracy achievable with frequency transfer over optical fibre networks has the potential to benefit other applications such as radio astronomy and precision spectroscopy.
Optical frequency combs
NPL's femtosecond optical frequency combs are used to measure the frequencies of our optical atomic clocks relative to our caesium fountain primary frequency standards and to compare optical standards operating at different frequencies. For both types of measurement we have demonstrated that the uncertainty introduced by the frequency comb is small compared to the uncertainties of the standards themselves. As well as checking the performance of our optical clocks, these measurements have been used to set stringent limits on any possible variation of fundamental physical constants.
We also use our combs to transfer stability from one frequency to another. We operate a comb-based universal synthesiser that transfers the stability of our best optical oscillator (a 1064 nm laser with sub-Hz linewidth) to the frequencies we need to operate our optical atomic clocks. The same system transfers the stability of the 1064 nm laser to a 1542 nm laser used to compare our optical clocks with others across Europe via optical fibre networks. It also produces a microwave output signal with fractional frequency stability superior to that of the best commercially available microwave oscillators. This is used for our caesium fountain primary frequency standard.
One of our frequency combs is designed to be readily transportable, and is equipped with a GPS-disciplined frequency reference so that it can be operated away from the NPL site. Although this limits the frequency measurement accuracy that can be achieved by several orders of magnitude compared to what is possible using our primary standards, the performance of the GPS-referenced comb is adequate for many applications. The transportable comb can also be used for direct optical frequency comparisons, and played a key role in a recent collaborative project where we demonstrated that optical atomic clocks can be used to measure gravity potential differences.
Optical frequency transfer
State-of-the-art optical frequency standards already exceed the best caesium standards in terms of stability and accuracy. But in most cases, the assessment is made locally, against frequency standards within the same laboratory. This is not enough – we need to be sure that clocks tick at the same rate across the globe, and this is an important step towards the redefinition of the SI second.
Using the optical fibre network that carries the internet, optical frequency transfer enables comparison of optical frequency standards over thousands of kilometres, without degrading their performance. NPL operates a 760 km long fibre link to Paris, which is part of a growing European network of fibre links for optical frequency transfer. This network supports an ongoing programme of intercomparisons between optical as well and caesium frequency standards at NPL, Observatoire de Paris (France) and Physikalisch-Technische Bundesanstalt (Germany).
Optical frequency transfer can also be used to disseminate an ultrastable frequency reference signal to places where no adequate frequency standard is available locally, such as universities or ground stations for spacecraft. Areas of research that might benefit from the enhanced stability and accuracy available via fibre include radio astronomy and cutting-edge physics research.
Optical frequency standards in conjunction with fibre links are also opening up entirely new ways of observing our planet. The gravitational redshift causes clock rates to slightly change according to local gravity potential. In comparisons of frequency standards, this needs to be taken into account. Therefore, knowledge of the precise altitude of each clock is a key factor in every comparison. The precise locations of NPL's optical frequency standards, along with those at LNE-SYRTE (France), PTB (Germany) and INRiM (Italy) have been surveyed to 2 – 3 cm accuracy. Conversely, comparing two distant, accurate optical frequency standards allows the difference in altitude to be determined – a technique known as 'chronometric levelling'. This technique has been demonstrated in a proof-of-principle experiment performed by NPL in collaboration with PTB and INRiM. With future optical clock technology, it will become possible to resolve height differences to the nearest centimetre, across thousands of kilometres.
Fibre links for optical frequency transfer act like giant interferometers. Changes of the optical path length of only a fraction of the wavelength of light, typically 1.5 micrometers, are readily detected in the form of phase shifts. For the purpose of frequency transfer, this is a potential error that needs to be corrected or accounted for. However, scientists at NPL and INRiM have come up with a way to exploit the same extraordinary sensitivity to detect minute ground movements, such as those caused by earthquakes .
The stability of optical atomic clocks is critically dependent on the short term stability of the lasers, enabling the short-term noise in the detection of the atomic absorption signal to be averaged out.
Even the best commercially available free-running lasers have spectral linewidths that are far too broad for high resolution spectroscopy of optical clock transitions. For this reason, the frequency of a clock laser must be pre-stabilised to a narrow resonance of a highly engineered optical cavity. The optical cavity consists of two highly-reflective mirrors bonded to a spacer made from ultra-low-expansivity glass, and resonates only at very specific optical frequencies. The frequency of the laser is actively steered so that it always remains in resonance with the cavity, and with a properly designed feedback system it thus takes on the frequency characteristics of the cavity.
In NPL's state-of-the-art ultrastable laser systems, careful attention is paid to temperature control and vibration isolation of the cavity. Our cavities are specially designed to be insensitive to vibrations, with the result that the frequency stability becomes limited by thermal fluctuations in the mirror coatings, which cause minute changes in the cavity length. In the quest for continual improvements in optical clock stability we are therefore developing new cavities with lower thermal noise limits, for example through optimal choice of the materials used for the cavity spacer, mirror substrates and coatings.