Some Accuracy Limiting Effects in an Atomic Beam Frequency Standard

The accurate resonance frequency of the transition (F, M ) = (4, 0) ++(3 , 0) in the ground state of cesium-l33 is expressed In the form of an operational equation f o r an atomic beam spectrometer. Emphasized are the terms in this equation which correct for the beam direction dependence and radiation field dependence of measured resonance frequencies: F where i and j refer to opposite beam directions through the apparatus, P i s the microwave power exciting the transition, and S . and S. a r e \he rates of linear frequency shift. The results of a detailed theoretical analysis are give,n which specify the contributions to these terms by various apparatus and fundamental shiftinducing effects. 1 3 This approach to accuracy specification is applied to the United States frequency standard, a National Bureau of Standards atomic beam machine designated NBS 111, through a set of experiments using an atomic hydrogen maser as a highly stable reference frequency source. The corrections determined are -3.2 x 10for beam direction dependence, -2. 2 x 1 0-l2 for power dependence, and t 0.4 x for second-order Doppler shift. The uncertainties in these corrections and contributions from other sources give a present 10 estimate of accuracy capabili ty off 1. 1 x 10-l' for NBS 111. This figure should be reducible by one order of magnitude through efforts to eliminate systematic errors in the 12 measurements of L, andv res r e s i j Up to the present t ime the lowest inaccuracy f igure that has been 1 quoted for any frequency standard is f 1. 1 x .l 0 a s a single standard deviation ( l o ) estimate for a National Bureau of Standards cesium beam machine designated NBS III. Attempts to improve on this value are resisted by the difficulty in identifying and making corrections for resonance frequency shifts induced by a variety of effects.. -12