We expect that future generations of atomic clocks will be based on optical transitions in laser-cooled ions and neutral atoms. The ultimate goal of this program is to measure the frequencies of optical frequency standards and to be able to compare them directly to the cesium primary atomic clock, NIST- 7. We will also develop the capability to synthesize and measure arbitrary optical frequencies in the visible and near visible spectral regions for metrological purposes (calibration of lasers for length measurements and communication systems etc). Toward this goal we are exploring new technologies that allow us to measure frequencies in the visible with a precision limited by the performance of the primary standard.

Our primary objective is to make connection between the cesium clock, NIST-7, (at 9.2 GHz) and the outstanding Hg+ optical reference (at 563 nm, 532 THz) developed by J. Bergquist et al. in the NIST Ion Storage Group. We also want to connect both of these to the Ca optical reference (657nm, 456THz) developed by C. Oates, as well as the stabilized Nd-Yag laser (1064nm, 281 THz) developed by J.L. Hall and collaborators at JILA. For example, the present scheme that we are developing to measure Hg+ and Ca relative to the cesium standard, and relative to each other, is diagramed below.

Here the connection between the various laser lines is by nonlinear mixing in periodically poled lithium niobate (PPLN). Thus, "DFG PPLN" indicates difference frequency generation (DFG) in PPLN, which can be used to put the 750nm laser at the midpoint of the frequency interval between 563nm and 1126nm. Similar mixings constrain the other lines.

To make the final connection to the cesium frequency standard we will follow the recent work of T. Udem, Th. Hansch and collaborators in Germany where they have demonstrated that mode-locked femtosecond lasers can be used to measure frequency intervals greater than 20 THz. We plan to use their method to span the interval between the 750 and 788nm lines shown above.

For the obvious reasons of simplicity, compactness, convenience, efficiency and reliability we are focusing most of our effort on semiconductor laser sources and domain-engineered nonlinear optical crystals. Work on diode lasers includes broadly tunable highly coherent lasers, semiconductor amplifiers, and advanced monolithic laser designs.

Leo Hollberg, NIST, hollberg@boulder.nist.gov.