Silicon Clock Challenges Atomic Timekeeping Norms

For decades, atomic clocks were the most stable means of measuring time. They measure time by oscillating in time with the resonant frequency of atoms, a method so precise that it is the basis for defining a second.

Today, a new challenger has emerged in the field of timekeeping. Researchers recently developed a tiny MEMS-based clock that uses silicon doping to achieve record stability. After running for eight hours, the clock deviated by only 102 nanoseconds, approaching the standard for atomic clocks while requiring less physical space and less energy to operate. This has been a challenge in the past due to the chaos that even slight temperature variations can introduce into timing.

The group presented its new clock at the 71st annual meeting of the IEEE International Electron Devices last week.

Save space and energy

The MEMS clock is built from a few tightly connected parts, all integrated onto a chip smaller than the side of a sugar cube. At its center, a silicon plate topped with a piezoelectric film vibrates at its natural frequencies, while a nearby electronic circuit measures these vibrations. A small, built-in heater gently keeps the entire structure at an optimal temperature. Because the resonator, electronics, and heater are all close together, they can work as a coordinated system: the resonator creates the timing signal, the electronics monitor and adjust it, and the heater prevents temperature variations from causing drift.

This clock is unique in several ways, says Roozbeh Tabrizian, project advisor and MEMS engineer at the University of Michigan. On the one hand, the resonator is “extremely stable despite variations in the environment,” he says. “You could actually raise the temperature from minus 40°C to 85°C and you see virtually no change in frequency.”

The resonator is so stable because the silicon from which it is made has been doped with phosphorus, Tabrizian explains. When a material is doped, impurities are added to it, usually to change its conductive properties. Here, however, the group used doping specifically to stabilize mechanical properties. “We control the mechanics very strictly so that the elasticity of the material does not change depending on temperature variations,” he explains.

Some other materials, like the commonly used timing crystal quartz, can also be doped for added robustness. But “we cannot miniaturize [quartz] “Semiconductor manufacturing benefits from size miniaturization,” says Tabrizian. “Semiconductor manufacturing benefits from size miniaturization,” so it’s an obvious choice for next-generation clocks.

Doping also allows the electronics to actively eliminate any small frequency drift over long periods of time. This attribute is “the most distinctive aspect of the physics of our devices compared to previous MEMS clocks,” says Tabrizian. By making silicon conductive, doping allows electronics to subtly adjust the mechanical drive force of the device, which counteracts slow changes in frequency.

This system is also unique in its integration of autonomous temperature sensing and adjustment, says Banafsheh Jabbari, a University of Michigan graduate student who led the project. “This clock resonator operates in two modes [or, resonant frequencies]basically. The main clock mode is very stable and we use it as [time] reference. The other is the temperature sensor. The latter acts as an internal thermometer, helping the electronics automatically detect temperature changes and adjust both the heating and the main timing mode itself. This built-in self-correction helps the clock maintain a stable time even when the environment changes.

These features make it the first MEMS clock to run for eight hours and deviate by only 102 billionths of a second. Linearly extended out to a week of operation, this equates to just over two microseconds of drift. This is worse than high-end laboratory atomic clocks by a few orders of magnitude, but it rivals the stability of miniaturized atomic clocks.

Additionally, the MEMS clock has a significant space and power saving advantage over its atomic competitors. The more isolated they are from their environment and the more energy they consume, the more precisely atomic clocks can probe the oscillations of atoms, Tabrizian explains. They are therefore generally the size of a wardrobe and consume a lot of energy. Even chip-scale atomic clocks are 10 to 100 times larger than the MEMS clock, he says. And, “most importantly,” this new clock requires 1/10th to 1/20th the power of mini-atomic clocks.

Timing for next generation technologies

Jabbari’s work grew out of a DARPA project that aimed to create a clock that could run for a week and deviate by just one µs. So there is still a lot to do. One of the challenges the team faces is the behavior of doped silicon over longer operating periods, such as a week. “You see some diffusion and changes in the material,” Tabrizian says, but only time will tell how well the silicon will hold up.

It is important for both researchers to continue their efforts because of the broad applications they envision for a small, energy-efficient MEMS clock. “Essentially, every modern technology we have needs some sort of synchronization,” Jabbari says, and she thinks the clock could fill the time synchronization gaps that currently exist.

For situations where the technology has robust access to GPS satellites, there are no problems to solve, she says. But in more extreme scenarios, like space exploration and underwater missions, navigation technology is forced to rely on internal timekeeping, which must be extremely bulky and power-intensive to be accurate. A MEMS clock could be a small, less power-hungry replacement.

There are also more everyday applications, says Tabrizian. In the future, when more information needs to be transmitted more quickly to every phone (or whatever device we’ll use in 50 years), precise timing will become crucial for transmitting data packets. “And, of course, you can’t put a big atomic clock in your phone. You can’t use that much power,” he says, so a MEMS clock might be the answer.

Even with promising applications, the road ahead for this project could be difficult due to existing competition. SiTime, a company already producing MEMS clocks, has already integrated its chips into Apple and Nvidia devices.

But Tabrizian is confident in his team’s abilities. “Companies like SiTime put a lot of emphasis on system design,” increasing the complexity of systems, he says. “Our solution, on the other hand, is entirely physics-based and looks at the very complex and very fundamental physics of a semiconductor. We are trying to get around the need for a complex system by making the resonator 100 times more precise than the SiTime resonator.”