The 9th Symposium on Frequency Standards and Metrology

The second symposium took place in 1976 in Copper Mountain, the USA organized by Helmut Hellwig.

Click here for the proceedings

The second meeting, called 2nd Frequency Standards and Metrology Symposium was held at the ski resort, Copper Mountain, Colorado, USA, from July 5 to 8, 1976. The organizer and chairman was H. Hellwig, National Bureau of Standards (now renamed the National Institute for Standards and Technology, NIST). The number of participants was 94. The Proceedings of the meeting (soft covers) were assembled and published by NBS (now NIST).

(Extract from David Wineland 2016 J. Phys.: Conf. Ser. 723 012001)

Cesium standard inaccuracy had steadily improved (to around 10^-13) and there were several reports on advances in hydrogen masers. One of the most interesting of these was a preliminary report by Bob Vessot on the results of a sub-orbital rocket flight (“Gravity Probe A”) that carried a maser [3]. During the approximately 2 hour, 10,000 km high flight, the frequency of this maser was compared to one on the earth’s surface. Because of the different time dependences of the relativistic frequency shifts caused by changes in the gravitational potential and time dilation, both effects could be confirmed with an uncertainty less than 10-4. This was a pioneering experiment showing the viability of carrying precise atomic clocks into space, a topic of ever increasing interest up to this day. Moreover, from a basic physics standpoint, it is still the most accurate direct measurement of the gravitational potential red shift. As a side note, the transponder technique used to determine and correct for the first-order Doppler shift between masers is now commonly used to stabilize the phase between separate sites in optical clock and frequency standard comparisons [4].

Optical frequency standards continued to improve. An interesting effect in saturated-absorption spectroscopy was the observation [5] of the splitting of the feature due to photon recoil [6] where the position of the two lines is governed by the condition where the transverse velocity of either the upper or lower state in the transition is at zero velocity. For many years, Christian Bordé, who has attended these symposia, has been a key person in explaining how the effects of recoil must be accounted for in atomic clocks, at all ranges of frequency. Recoil effects are of course an integral part of the atom interferometer experiments whose dramatic advances are reported at the current Symposium. To increase resolution in saturated absorption experiments, some laboratories started pushing towards laser beams with larger waists and heavier absorbers such as SF6 and OsO4, for increased interaction times. In the early 70’s, the dye laser was becoming an extremely useful tool in AMO physics; for optical frequency standards, it’s wide tunability meant that atomic or molecular reference transitions need not be in near coincidence with an existing fixed frequency laser line and one could now focus on choosing a reference that had particularly low systematic offsets. Ted Hänsch described his early very impressive experiments on the 1S – 2S two-photon Doppler-free transition in hydrogen, using a pulsed dye laser at 486 nm (peak power ~ 50 kW) doubled to 243 nm in a lithium-formate crystal.

Frequency chains were featured by a number of laboratories. Here, through a sequence of lasers of increasing frequency referenced to each other with harmonic mixers, phase coherence could be maintained. At the highest frequencies, metal-on-metal diodes were used but these lost efficiency as one approached the visible range of the spectrum. Therefore, a difficult step was to go from frequencies corresponding to wavelengths of a few micrometers to the visible range. As possible solutions to this problem, Veniamin Chebotayev discussed 4-wave mixing experiments in Ne to sum three infrared laser beams to the visible (λ ≅ 0.65 µm) and also the use of nonlinear crystals (e.g. LiNbO3). In any case, with the Cs to 3.39 µm frequency chain, it was now possible to connect the 86Kr wave length standard to Cesium via a comparison of the 86Kr wavelength to the 3.39 µm wavelength in a shared reference cavity [7]. This led to a measurement of the speed of light to about 3 parts in 109 prior to the second Symposium [8] (see also [9]).

Even at this relatively early stage in the Symposium series, the benefits of low temperature reference cavities were demonstrated. Sam Stein presented results of a microwave parametric oscillator whose signal component was stabilized to a Nb reference cavity that had a Q of around 1010, yielding an instability σy(100 s) ≅ 6 × 10-16. Of course cold, cavity-stabilized microwave and optical oscillators continue to be intensely studied today.

Cliff Will and Robert Pound gave very interesting lectures proposing various fundamental investigations. These included tests of the equivalence principle for clocks whose frequencies depended in different ways on the fundamental forces, a laboratory test of frame dragging due to a nearby rotating object, space-born detection of gravitational waves, and searches for a time rate of change in the ratios of the fundamental constants. These topics endure today, benefitting from ever-increasing sensitivity. Prof. Pound pointed out the high Q’s available in Mössbauer spectroscopy and gave an example of the 92 keV transition in 67Zn whose Q could be measured using one 67Zn sample as the source and another 67Zn sample as the absorber, a direct comparison of two essentially identical samples. Here, the relative frequencies of the two samples could be tuned via the first-order Doppler shift by moving one of the samples relative to the other. The linewidth measured in this way was 0.8 µm/s corresponding to a Q of around 3 ×1014. As a personal aside, when I was a graduate student at Harvard working on masers in Norman Ramsey’s group, Prof. Pound would tease us in a light-hearted way about our meager Q’s of only around 109. I had always hoped we might compete using atoms, and finally, in 2011 our lab at NIST was able to compare the 1S0 – 3P0 transitions of two 27Al+ ions in a similar way to Mössbauer spectroscopy and demonstrate a Q of 3.4 × 1016 [10].

For atomic ions at this symposium, Guenter Werth reported on laser optical pumping in Ba+ ions; however, more interesting to me was that between this and the previous symposium, he and Fouad Major had measured the 40.5 GHz hyperfine transition in 199Hg+, where optical pumping and state detection could be accomplished with fluorescence from a 202Hg+ resonance lamp [11].This work was extended by Len Cutler and colleagues at HP and John Prestage, Lute Maleki, and colleagues at JPL, where it has undergone dramatic improvements, continuing today. Hans Dehmelt described his “shelved electron amplifier” detection idea[12] where state-dependent fluorescence on an allowed transition can be used to discriminate between two levels of a clock transition with essentially 100% efficiency. In this case, noise in the measurement of transitions reduces to the fundamental quantum “projection” noise associated with the fluctuations in the measured state for a superposition of two clock states, the so- called “standard quantum limit” for uncorrelated atoms. Impressively, with the current precise control of technical noise in the clock experiments, the quantum noise can still dominate on even larger numbers of atoms (N~106) where the quantum noise-to-signal ratio scales as N . Dehmelt also described early laser-cooling proposals [13, 14].