Two-frequency separated oscillating fields technique for atomic and molecular beam spectroscopy

Abstract

HE RAMSEY TECHNIQUE [ 11 of separated oscillatT ing fields for atomic beam spectroscopy is widely used in atomic frequency standards, specifically in the cesium beam standard which forms the present basis for the definition of frequency and time interval. The technique offers the advantages of narrow linewidth, relative freedom from first-order Doppler effects, relaxation of certain constraints on field homogeneity in the drift region, and relative ease of implementation. A difference 6 in the phase of the interrogating RF signals as experienced by the atomic beam in the two Ramsey interaction regions leads to a displacement of the maximum transition probability from the true atomic resonance frequency by 6 / ( T ) where ( T ) represents the average flight time between the interaction regions. Care in fabrication and assembly of atomic beam apparatus may reduce but cannot ultimately eliminate this source of error. Beam reversal, a procedure which is only practical for laboratory devices, yields information on the value of 6, but the accuracy of this technique is limited by a similar effect, that of “distributed” phase error, which occurs as a result of a phase change across the transverse dimension ofthe interaction region. This latter effect is much less tractable in analytical treatment [2]. These two effects are presently the most serious source of uncertainty in the evaluation of primary frequency standards. In NBS 6, the IJ.S. primary cesium frequency standard, phase-shift effects limit the accuracy to 10-13. The long-term stability ( may also be limited by phase-shift effects. In commercial cesium standards phase-shift effects may be major contributors to inaccuracy and long-term drift. We are attacking the phase-shift problem by relaxing the constraint 6 = 0 and allowing the relative phase of the two interaction regions to advance (or recede) at a constant rate [3]. This will be implemented by driving the two spatially separated cavities each with a different frequency near the cesium atomic resonance. Fig. 1 depicts such an interroga, tion scheme. The transition probability of an atom travers-