Rubidium fountain atomic frequency standard.

We are developing a rubidium fountain standard at NPL, the stability of which will exceed the stability of our caesium fountains. The reason for this is that the cold collisional frequency shift for ⁸⁷Rb atoms is significantly smaller than that for caesium atoms, meaning that it is possible to operate with a higher number of atoms to yield higher signal-to-noise ratio.

The fractional fluctuations of the measured rubidium clock transition frequency caused by shot-to-shot fluctuations in collisional shift (due to variations in atomic density), are expected to be below 1 part in 10¹⁶. To take maximum advantage of the small collisional shift it is therefore important to suppress all other major sources of frequency fluctuations. As a local oscillator we will use a cryogenic sapphire oscillator with a fractional frequency stability of about 5 parts in 10¹⁵ for averaging times between 1 and 10 s. The quantum projection noise of the system is optimised by using an intense magneto-optical source of cold atoms to load the fountain.

Measurements of the rubidium hyperfine transition at 6.8 GHz are carried out by comparison with the NPL caesium primary frequency standard. To date only one group (at SYRTE in Paris) has reported precise measurements of this frequency, which is one of the transitions that may by used as secondary representations of the second.