Although caesium atomic clocks were developed as early as 1955, they have remained crucial for time measurement. Since 1967, the second has been defined as “the duration of 9 192 631 770 periods of the radiation corresponding to the transition between the two hyperfine levels of the ground state of the caesium 133 atom.”
In modern caesium clocks, the atoms fall like water droplets in a fountain through a microwave resonator. The frequency of this resonator is tuned to the atomic excitation frequency. However, the search for alternatives to caesium clocks began at an early point in time. Towards the end of the 1970s, it was Günter Werth at Mainz University who was already pursuing the approach of keeping ionized atoms confined in an ion cage (a so-called Paul trap) by means of electric fields for application as an atomic clock. In ultra-high vacuum, this cage offered the atoms much better protection against collision with background gas molecules and other disturbing ambient influences than a standard caesium clock. In addition, it was possible to observe the atoms for nearly any length of time. These two aspects were crucial for enabling much more accurate frequency measurement. For his work titled “The Ion Cage as a Frequency Standard”, Prof. Dr. Werth was awarded the 1985 Helmholtz Prize, which was endowed with 8,000 DM.
Werth used the ions of mercury and later of ytterbium and barium as “frequency generators”. As with caesium atoms, these ions have two hyperfine levels in their electronic ground state whose energy levels differ by several GHz. A dye laser was used to ascertain the extent of the occupation of these two hyperfine levels and to change their population density by means of optical pumping. From the laser excited state, the ions returned to the ground state while emitting light.
If the atoms were irradiated by a microwave field whose frequency was tuned to the transition between the two hyperfine levels, then the occupation of these two levels changed, causing a change in the intensity of the luminescence observed. The intensity of the light allowed a statement to be made on how well the microwave frequency agreed with the hyperfine splitting. Finally, the resonant microwave frequency was determined referenced to a caesium clock.
Based on the recorded resonance lines, it was possible to measure, for example, the hyperfine splitting of the 171Yb ions with an uncertainty of 3 × 10-12. At that time, Werth assumed that it was possible to achieve an uncertainty of 10‑15 by suppressing several error sources – which has since been experimentally realized by various groups worldwide. Today, frequency standards in the microwave range based on trapped mercury ions are being considered as potential clocks for the European navigation system called Galileo. It has become possible to cool down single trapped atoms to very low temperatures using laser light and to keep them at very low velocity in ion traps for months, which further reduces the uncertainty.
Thanks to the frequency comb technique – for which Theodor Hänsch and John Hall were awarded the Nobel Prize in 2005 – it is now possible to measure frequencies of “forbidden” optical transitions of trapped ions in the visible part of the spectrum. Optical atomic clocks, which are operated, for example, with ytterbium ions, exhibit relative instabilities of a few parts in 10‑18. While this competes well with standard microwave caesium clocks, it is of the same order of magnitude as clocks based on caesium atoms confined in optical lattices.