Accurate time is critical in modern digital communications circuits to avoid interference and provide networking timing. GPS is good enough for many of these systems, but it doesn't work indoors and is prone to errors due to interference and variation in propagation. Atomic clocks can provide extreme accuracy, but they are huge.
A group at MIT and Draper Laboratory has come up with a new approach to atomic timekeeping that comes close to the accuracy of the atomic clocks in GPS satellites but could potentially be built in a device no larger than a Rubik's cube. While very small chip-sized atomic clocks (CSACs) are commercially available, they drift over time and are less accurate than the fountain clocks used as a time standard.
The group outlined its approach in the journal Physical Review. Co-author Krish Kotru, a graduate student in MIT's Department of Aeronautics and Astronautics explained the problem with today's fountain clocks: “You could put one in a pickup truck or a trailer and drive it around with you, but I’m guessing it won’t deal very well with the bumps on the road. We have a path toward making a compact, robust clock that’s better than CSACs by a couple of orders of magnitude, and more stable over longer periods of time.”
The most accurate atomic clocks use the resonant frequency of cesium as a reference: a second is defined as 9,192,631,770 oscillations of a cesium atom between two energy levels. Fountain clocks measure this frequency by creating small clouds of slow-moving cesium atoms a few feet high, like a pulsed fountain, and measure their oscillations as they pass up, and then down, through a microwave beam.
The MIT and Draper Laboratory group probe the atom's oscillations using laser beams, which are easier to control spatially and require less space. One problem with using laser beams is that they produce an electric field which can shift an atom's resonant frequency, an effect called “AC Stark shift.” Kotru said, “That's really bad, because we're trusting the atomic reference. If that’s somehow perturbed, I don’t know if my low-quality wristwatch is wrong, or if the atoms are actually wrong.”
Kotru's team minimizes this problem by applying laser pulses of changing intensity and frequency, similar to the technique used in nuclear magnetic resonance spectroscopy to probe features in individual molecules. He explains, “For our approach, we turn on the laser pulse and modulate its intensity, gradually turning it on and then off, and we take the frequency of the laser and sweep it over a narrow range. Just by doing those two things, you become a lot less sensitive to these systematic effects like the Stark shift.”
The group found their technique suppressed the AC Stark shift by a factor of 100 compared with a conventional laser-based system. The team's apparatus measures time in intervals of 10 milliseconds rather than the single second used by fountain clocks and their meter-high cesium cloud. This approach is less accurate, but much more compact. Kotru said, “That’s fine, because we’re not trying to make the world’s standard — we’re trying to make something that would fit in, say, a Rubik’s cube, and be stable over a day or a week.” He said the stability and accuracy of the system should be comparable to that of microwave-based atomic clocks on today's GPS satellites, which are bulky and expensive.
Not only is the group's system compact, but it can handle shaking. The group tested its response to physical forces, which is important if the device is to be used in a backpack or in a vehicle. While they didn't physically shake their experimental system, the group “created a displacement between the atoms and the laser beam,” moving the laser beam from side to side as it probed the cloud of atoms. Even with the simulated shaking, the system was able to measure the atoms' resonant frequency with a high degree of sensitivity.
The group is working on reducing the size of other components in the system, including the vacuum chamber and electronics. Kotru commented, “Additional miniaturization could ultimately result in a handheld device with stability orders of magnitude better than compact atomic clocks available today. Such a device would satisfy requirements for more technologically intensive applications, like the synchronization of telecommunications networks.”