NASA NTRS2005
A proposed highly accurate clock or oscillator would be based on the concept of an opto-electronic oscillator (OEO) stabilized to an atomic transition. Opto-electronic oscillators, which have been described in a number of prior NASA Tech Briefs articles, generate signals at frequencies in the gigahertz range characterized by high spectral purity but not by longterm stability or accuracy. On the other hand, the signals generated by previously developed atomic clocks are characterized by long-term stability and accuracy but not by spectral purity. The proposed atomic clock would provide high spectral purity plus long-term stability and accuracy a combination of characteristics needed to realize advanced developments in communications and navigation. In addition, it should be possible to miniaturize the proposed atomic clock. When a laser beam is modulated by a microwave signal and applied to a photodetector, the electrical output of the photodetector includes a component at the microwave frequency. In atomic clocks of a type known as Raman clocks or coherent-population-trapping (CPT) clocks, microwave outputs are obtained from laser beams modulated, in each case, to create two sidebands that differ in frequency by the amount of a hyperfine transition in the ground state of atoms of an element in vapor form in a cell. The combination of these sidebands produces a transparency in the population of a higher electronic level that can be reached from either of the two ground-state hyperfine levels by absorption of a photon. The beam is transmitted through the vapor to a photodetector. The components of light scattered or transmitted by the atoms in the two hyperfine levels mix in the photodetector and thereby give rise to a signal at the hyperfine- transition frequency. The proposed atomic clock would include an OEO and a rubidium- or cesium- vapor cell operating in the CPT/Raman regime (see figure). In the OEO portion of this atomic clock, as in a typical prior OEO, a laser beam would pass through an electro-optical modulator, the modulated beam would be fed into a fiber-optic delay line, and the delayed beam would be fed to a photodetector. The electrical output of the photodetector would be detected, amplified, filtered, and fed back to the microwave input port of the modulator. The laser would be chosen to have the same wavelength as that of the pertinent ground-state/higher-state transition of the atoms in the vapor. The modulator/ filter combination would be designed to operate at the microwave frequency of the hyperfine transition. Part of the laser beam would be tapped from the fiberoptic loop of the OEO and introduced into the vapor cell. After passing through the cell, this portion of the beam would be detected differentially with a tapped portion of the fiber-optically-delayed beam. The electrical output of the photodetector would be amplified and filtered in a loop that would control a DC bias applied to the modulator. In this manner, the long-term stability and accuracy of the atomic transition would be transferred to the OEO.