The hyperfine structure of the ground state of the 133Cs atom serves as the definition of one metric second.

The previous decades have seen impressive progress in the development of very precise atomic frequency standards (eg. atomic clocks). The fractional change in frequency of certain narrow microwave and optical transitions can be measured to 10-15, with the prospect of measurements to a precision of 10-18 on the horizon. This presents great opportunities for applied and fundamental science, in particular in searching for any variation of fundamental constants.

Many modern theories that unify gravity with other interactions or describe a non-stationary universe allow fundamental constants to vary. Analysis of quasar absorption spectra performed at our school suggests that the fine structure constant, a, might have been smaller in the past: 10 billion years ago it was smaller by a tiny (10-5) fraction of its current value. If we assume that a is changing at a constant rate over the cosmological time scale we would expect it to still be changing at the relative rate of 10-15 per year. A variation such as this could be observed with atomic clocks. Different atomic clocks depend differently on a so by comparing their frequencies over a long time interval (a year or more) any change in a can be studied.

Note that these measurements can only reveal the change in the ratio of the atomic frequencies. Atomic calculations are needed to link this change to any variation of a. These calculations are performed by our group. The table shows recent constraints on possible present day a variation obtained from using our results in conjunction with the comparison of some extremely stable atomic frequencies to the cesium primary frequency standard. The cesium primary frequency standard is the frequency of the microwave transition between the components of the 133Cs hyperfine structure. The sensitivity of atomic clocks to the variation of a is now approaching that of quasar absorption spectra. It is worth mentioning that the possible measurement of the variation of a using atomic clocks was mentioned in the description of the work awarded this year’s Noble prize.

Another way to improve the sensitivity of atomic measurements is by finding atomic frequencies that change much faster that a if a varies. A unique example of this can be found in the dysprosium atom. The dysprosium atom possesses two very close levels of opposite parity and different electron configurations. The small transition frequency between these levels changes 108 times faster than a, if a varies. An experiment utilizing this sensitivity is currently underway at Berkley. The first very promising results have already been obtained and the prospects for further improvement in accuracy are very good.

Elizabeth Angstmann, Vladimir Dzuba
and Victor Flambaum