The Walsworth Group

Atomic clock investigations at SAO


At present, the room temperature atomic hydrogen maser is the most stable time and frequency source available for intervals from seconds to days. Hydrogen masers are active oscillators that operate at 1420 MHz on the F = 1, mF = 0 to F = 0, mF = 0 hyperfine transition in the hydrogen electronic ground state. The frequency stability of state-of-the-art hydrogen masers, such as those designed and constructed here at the Smithsonian Astrophysical Observatory (SAO), is typically better than one part in 1015 for averaging intervals of 103­104 seconds. The hydrogen maser is an essential atomic clock for radio astronomy, time-keeping, and spacecraft navigation, and is used in precision tests of gravitation, relativity, and quantum mechanics [1]. For example, hydrogen masers developed at SAO are used in NASA's Deep Space Tracking Network, and in the U.S. Naval Observatory's ensemble of master clocks. A hydrogen maser built at SAO was flown on NASA's Gravity Probe A [2], providing the most accurate test to date of the gravitational red-shift (a consequence of the local position invariance of the Einstein Equivalence Principle). Hydrogen masers are also valuable instruments for experimental atomic physics [3], including work at low temperatures [4]. (See also the related SAO project, the Hydrogen Maser Clock.)

The Cryogenic Hydrogen Maser

We have developed a cryogenic hydrogen maser (CHM) that operates at 0.5 kelvin, and employs superfluid helium-coated walls to store the masing hydrogen atoms [5]. See Figs. 1 and 2. The CHM may provide frequency stability that is one to three orders of magnitude better than a room temperature hydrogen maser because of greatly reduced thermal noise and larger signal power. Such exceptional frequency stability will be required for spacecraft tracking in future deep-space missions, for space-based tests of relativity and gravitation, and for local (i.e., flywheel) oscillators used with "next-generation" absolute frequency standards such as the laser-cooled cesium fountain [6] and the linear ion trap [7]. These new devices, which are under development at NIST and other laboratories, are passive high-resolution frequency discriminators. Alone, they cannot function as superior atomic clocks; their effective operation depends critically on being integrated with an active local oscillator with excellent short term stability--such as that expected from the CHM.

Fig. 1. SAO cryogenic hydrogen maser (CHM) operated inside a 3He refrigerator.

Fig. 2. Schematic diagram of the SAO cryogenic hydrogen maser.

To date, we have measured the CHM's frequency stability (Allan deviation [8]) relative to a room temperature hydrogen maser to be approximately 1 x 10­13/t1/2, for averaging intervals, t, in the range of 10 sec < t < 300 sec. For intervals longer than a few minutes, the CHM's frequency stability is not yet as good as that of a room temperature hydrogen maser, because of large wall frequency shifts to the hydrogen hyperfine transition due to the sapphire substrate lying beneath the superfluid helium film. A typical set of data comparing CHM and room temperature hydrogen maser frequency stability is shown in Figure 3, along with the calculated limit to CHM performance set by thermal noise.

Fig. 3. Typical measured CHM and room temperature hydrogen maser frequency stabilities.

We are currently installing a thin-walled quartz bulb inside the sapphire resonant cavity to serve as the atomic hydrogen storage chamber, and thus to eliminate the large wall frequency shift that limits long-term CHM frequency stability. We are also installing a low temperature preamplifier to improve the CHM's short-term frequency stability to the thermal limit shown in Fig. 3.

In parallel we are using the CHM to study important effects in low temperature atomic physics, e.g., hyperfine-induced hydrogen-hydrogen spin-exchange collisions. These collisions depend sensitively upon: (i) details of the hydrogen-hydrogen interaction potential at long range that are otherwise experimentally inaccessible; (ii) non-adiabatic (i.e. non-Born-Oppenheimer) effects particular to cold atomic collisions; and (iii) inclusion of the intra-atomic hyperfine interaction (strength ~ 0.07 K) in addition to the electron exchange interaction (strength ~ 50,000 K), which fundamentally alters the rotational symmetry of the hydrogen-hydrogen collisional process. Furthermore, the calculated values for cold hydrogen-hydrogen collision cross-sections are sensitive to the theoretical techniques used in the calculations [9-11], making experimental comparisons particularly significant.

The Double-Bulb Rubidium Maser

We are developing the double-bulb rubidium maser (DBRM) as a new high-stability active oscillator [12,13]. The DBRM will operate at 6.835 GHz on the 87Rb hyperfine transition. The DBRM has the potential to provide frequency stability better than that of a room temperature hydrogen maser, but in a unit that is smaller, lighter, more robust, and less expensive to manufacture and maintain. See Fig. 4. The advantageous properties of the DBRM, relative to the hydrogen maser, result from its higher operating frequency and simpler design. Because the DBRM operates at 6.835 GHz, rather than at the hydrogen hyperfine frequency of 1.42 GHz, its resonant cavity has roughly 1% of the volume of a hydrogen maser cavity. Consequently, the surrounding mechanical components, particularly the magnetic shields, are significantly smaller. Furthermore, the DBRM uses a sealed, evacuated double-bulb cell to contain the rubidium atoms, thus dispensing with the hydrogen maser's vacuum system, depletable source of atoms, radio-frequency dissociator, and state-selecting magnets. All of these factors decrease the size, weight and complexity of the DBRM.

Figure 4. Comparison of the frequency stability of the hydrogen maser and the traditional passive rubidium clock with the expected stability of the double-bulb rubidium maser. The right-hand axis shows the corresponding precision of range rate measurement.

Because it would provide high frequency stability in a compact package, the DBRM has wide potential usage in a variety of terrestrial, airborne and space-based applications.

Navigation and communication applications of the DBRM include serving as a portable high-stability reference oscillator for secure, high-speed spread-spectrum communications and for bistatic radar, which has the potential for imaging targets without revealing the location of the radar receiver. The DBRM would be attractive as an improved clock for Global Positioning System (GPS) satellites, and its use on geosynchronous spacecraft would increase the precision of locating such spacecraft when used in conjunction with an on-board GPS receiver. Onboard aircraft, the DBRM could improve the reliability of the Federal Aviation Administration's enhanced GPS system of aircraft navigation. In this application a compact, stable clock would reduce the number of simultaneously tracked GPS satellites required to detect and eliminate erroneous GPS data. Furthermore, the DBRM could enable the development of high performance non-GPS guidance systems, allowing real-time measurements of range and range rate with great precision.

In research, metrology and time-keeping, the DBRM would be ideal as the flywheel oscillator for narrow-linewidth atomic fountain and trapped ion standards. Such devices, which are under development at several laboratories and may be used in space, hold promise of a substantially improved level of long-term frequency stability; however, their operation depends critically upon the use of an active local oscillator with short term stability on the order of that expected from the DBRM. The DBRM would be useful in astronomy and space science as the local oscillator for earth-based Very Long Baseline Interferometry (VLBI) and deep-space tracking, and would be ideal as the local oscillator for space-based VLBI. Other space applications for a small, stable DBRM include high-precision global time transfer, which would improve the ability of standards laboratories and other users to compare their clocks, and space-based experiments in relativity and gravitation.

Fig. 5. Photograph of the prototype double-bulb rubidium maser.

Fig. 6. Schematic diagram of the double-bulb rubidium maser.

We have developed a prototype DBRM using equipment adapted from our hydrogen masers (see Figs. 5 and 6). The DBRM has two connected quartz bulbs, one for hyperfine state selection via diode laser optical pumping and the other for maser radiation in a resonant microwave cavity. This separation of the optical pumping and maser functions is designed to eliminate the light shift that afflicts current rubidium clocks. Furthermore, the DBRM avoids any buffer gas pressure shift by coating the inner walls of the double-bulb cell with tetracontane, a material that minimizes rubidium-wall interactions. As a result of the DBRM's double-bulb wall-coated design, the dominant systematic effects that perturb the frequency of the traditional passive rubidium clock--light shifts and pressure shifts--should be avoided. To date, we have state-selected 87Rb atoms via optical pumping with diode lasers, detected transport of the state-selected atoms between the two quartz bulbs, and observed their 6.835 GHz hyperfine signals (see Fig. 7).

Fig. 7. Observed 87Rb hyperfine signals from prototype double-bulb rubidium maser.<


[1] J. Vanier and C. Audoin, The quantum physics of atomic frequency standards (Adam Hilger, Bristol, 1989).
[2] R.F.C. Vessot, et al., Phys. Rev. Lett. 45, 2081 (1980).
[3] R.L. Walsworth, I.F. Silvera, E.M. Mattison, and R.F.C. Vessot, Phys. Rev. A 46, 2495 (1992).
[4] M.E. Hayden, M.D. Hurlimann, and W.N. Hardy, Phys. Rev. A 53, 1589 (1996).
[5] R.L. Walsworth, I.F. Silvera, H.P. Godfried, C.C. Agosta, R.F.C. Vessot, and E.M. Mattison, Phys. Rev. A 34, 2550 (1986).
[6] A. Clairon, S. Ghezali, G. Santarelli, S.N. Lea, M. Bahoura, E. Simon, S. Weyers, and K. Szymaniec, Symposium on Frequency Standards and Metrology, edited by J. Bergquist (World Scientific, Singapore, 1996), 49.
[7] R.L. Tjoelker, J.D. Prestage, and L. Maleki, Symposium on Frequency Standards and Metrology, edited by J. Bergquist (World Scientific, Singapore, 1996), 33.
[8] D. Allan, Proc. IEEE 54, 105 (1966).
[9] B.J. Verhaar, J.M.V.A. Koelman, H.T.C. Stoof, O.J. Luiten, and S.B. Crampton, Phys. Rev. A 35, 3825 (1987).
[10] J.M.V.A. Koelman, S.B. Crampton, H.T.C. Stoof, O.J. Luiten, and B.J. Verhaar, Phys. Rev. A 38, 3535 (1988).
[11] C.J. Williams and P.S. Julienne, Phys. Rev. A 47, 1524 (1993).
[12] W.M. Golding, Proceedings of the 1994 IEEE International Symposium on Frequency Control (IEEE, Piscataway, NJ, 1994), 724.
[13] R.L. Walsworth, E.M. Mattison, R.E. Stoner, K.Y.N. Tan, R.F.C. Vessot, and I.A. Yu, Symposium on Frequency Standards and Metrology, edited by J. Bergquist (World Scientific, Singapore, 1996), 187.