Hyperfine Splitting of Rubidium-87

Abstract

RUBIDIUM has been used in the development of gas cell frequency standards, and from measurements of the variation of frequency with the pressure of the buffer gas used in the cell it has been deduced¹ that the frequency corresponding to zero gas pressure is: where the second of time is that giving 9 192 631 770 c./s. for the cæsium hyperfine structure splitting at zero field². It has been found, however, that when cæsium is used in a gas cell the extrapolated frequency is 30 c./s. lower than the value given above obtained by an atomic beam experiment³. There is therefore considerable interest in the atomic beam value for rubidium. For this purpose the oven of the long beam chamber⁴ at the National Physical Laboratory was filled with a mixture of rubidium-87 and cæsium-133. A strong beam of atoms was obtained at an oven temperature of 120° C., and the application of the appropriate radio frequency produced, alternatively, resonances of the rubidium or cæsium atoms. The amplitude of the cæsium resonance was 50 per cent greater than that of rubidium, the higher vapour pressure of cæsium presumably compensating for the fact that a greater proportion of the rubidium atoms take part in the resonance. In both cases the signal to noise ratio was adequate for the precise measurement of the frequencies corresponding to the resonance peaks. The rubidium resonance, which was 71 c./s. wide, was excited in cavities independently tuned and separated by a distance of 254 cm. It was slightly asymmetrical, but the asymmetry was measured and a correction applied. The value on the atomic time-scale corrected for the asymmetry and magnetic field is: It is proposed to continue measurements of the two atomic transition frequencies.