A Fundamentally New Way of Realizing Pressure . . . Without Mercury

Glass-enclosed cavity for prototype pressure device.

In 1643, Evangelista Torricelli, in the course of solving a Tuscan plumbing problem, made the first barometer with a meter-long tube of mercury. Ever since then, civilization has used the toxic liquid metal to measure pressure. Today, the United States’ primary pressure standard -- NIST’s Ultrasonic Interferometer Manometer -- stands three meters tall, contains about 230 kg of mercury, and has an uncertainty of about 2.5 parts per million (ppm). But it is basically the culmination of numerous incremental improvements on the original 17th-century design.

The long-standing quicksilver hegemony, however, is about to end, says Jay Hendricks of PML’s Sensor Science Division. “We expect to make a fundamental change, a paradigm leap forward in the way we realize units of pressure, replacing mercury manometers with the world’s first portable, quantum-based primary barometric pressure standard,” Hendricks says.

Hendricks, Thermodynamic Metrology Group Leader Gregory Strouse, Doug Olson, and Jack Stone from PML’s Semiconductor & Dimensional Metrology Division make up the team pursuing the ambitious project, which recently won funding of approximately $1 million per year for up to five years in NIST’s competitive Innovations in Measurement Science program.

The plan is to develop ultra-precise optical interferometer cavity devices for pressure standards that will be accurate to about 1.4 ppm and will at the same time provide a new method of realizing and disseminating temperature and length. “We’re looking to revolutionize the way we measure all three basic quantities,” Strouse says.

Such devices –initially developed at the size of a small mailing tube – would find immediate applications in commercial aviation, semiconductor manufacturing, and military activities, and they would also eliminate persistent difficulties in pressure metrology.

“We feel we can’t get any further improvements in accuracy with mercury,” Hendricks says. “And in practical terms, there are real problems with using mercury systems to disseminate pressure to industry. Most companies don’t want to own a big manometer that uses large quantities of a dangerous neurotoxin that is being rapidly phased out worldwide.” Such systems can be expensive to operate and costly to acquire. For example, a widely used commercial manometer costs in the range of five hundred thousand dollars and uses 10 kg of mercury.

The pressure and temperature of a gas are directly related to its density. Gas density is, in turn, directly related to the refractivity of the gas. In the instrument the PML team plans to develop, the refractivity of ultra-pure helium gas will be determined by locking lasers to the resonances of a variable-length, multiple Fabry-Perot (FP) cavity. A FP cavity is formed by careful alignment of two highly reflective mirrors at each end of an enclosure. A laser is locked to a FP cavity by tuning its frequency such that an integer number of half-wavelengths fit between the mirrors.

In a simplified scheme, the procedure would be as follows (see image):

Laser light at 633 nm will be locked to the resonance of two FP cavities; one of the cavities will be at vacuum (n=1, where n is refractive index), the other will be filled with helium gas at a known temperature. The mirror substrate forming one end of the cavities will be translated by a physical length of ΔL (about 100 mm with a coarse translation stage). By relocking the lasers to the translated cavity resonances and measuring the change in laser frequencies, the value of ΔL will be determined to picometer accuracy. Since the physical length change ΔL should be identical for both cavities, the refractivity of the helium gas is derived by n-1 = n*ΔL_gas - ΔL_vac. As mentioned before, from refractivity comes gas density, and from gas density at a known temperature comes pressure.

The design allows for displacement of one reflector to accommodate different resonance lengths for the waves in vacuum and in helium.

The plan depends on several critical innovations. One is improving knowledge of the refractive index of helium by an order of magnitude and taking the first observational data to accompany the values generated by quantum theory. The team is collaborating with a physical chemist to provide the theoretical calculations, which will then have to be confirmed by the team’s observations to an uncertainty less than 0.5 ppm.

“Of course, companies don’t want to be messing around with getting super-high-purity helium. They want to use easy to obtain, fairly high purity nitrogen,” Hendricks says. “So to create a practical standard we can deploy in the field, we need to transfer those very accurate theory measurements for helium index of refraction over to nitrogen. We can do that with the variable length cavity.”

The disseminated standard would consist of a fixed-length interferometer with both vacuum and gas cavities that would allow users to correct the index of refraction in real-time, enabling length measurements in air at a level of 2 ppb uncertainty. That degree of precision is unavailable today, but needed nonetheless. “On a factory floor, the density of air is changing all the time,” Hendricks says. “There are variations in temperature and humidity. But our planned device would allow real-time corrections at the scale of tens of nanometers or better.”

The same device will provide exquisitely sensitive measurements of temperature and length. Because of the fundamental relationships between refractivity of a gas (refractive index-1) and gas density which is determined by pressure and temperature-- if any two are known, the other can be determined. “So when we have an accurate measurement of pressure, we can determine temperature to within about 1 ppm,” Strouse says. “And eventually we can finally liberate realization of the kelvin from the triple-point of water.” Similarly, “we can use measurement of laser frequencies to make an ultraprecise determination of length.”Now this is where it gets fun.

The first experimental interferometers will be about 30 cm long. “But we think that eventually it could shrink down to the size of a deck of cards,” Hendricks says. “There’s no theoretical limit to how small we can make it. And the devices could probably be made for something in the range of $150,000 – about one-third the cost of high-quality mercury manometers.”

The device’s combination of unprecedented accuracy and extreme portability, the team believes, would make it useful for commercial aviation. “When airspace around an airport is filling up,” Hendricks says, “you want to stack planes as close together as you can, consistent with safety, to get them ready to land. One way to do that is to reduce the vertical separation from the current standard of around 2000 feet to maybe 1000 feet. That’s not really possible with current instrumentation and calibration standards. This device, however, could easily have the resolution and accuracy for high-precision vertical flight measurements.”

Representatives of Boeing and Lockheed Martin have endorsed the PML project, as have officials from the U.S. Army and U.S. Navy Primary Standards Laboratories. Defense applications include improved positional data for aircraft flying low and following contours of the surface landscape.

The high degree of portability and its use of gas thermodynamic properties for traceability has another advantage: calibrations can occur at the factory floor. “Companies will no longer have to send their pressure standards to NIST for calibration, saving them time and money. And at NIST we’ll have more time to do the fun stuff in our labs,” says Olson.

Finally, with better length metrology semiconductor manufacturing can increase feature density. “The international industry consensus plan, the International Technology Roadmap for Semiconductors, calls for uncertainty in length measurements below 3 parts in 10⁸. With the new technology, we can deliver a capability that is more than a factor of 10 better than this and enable it to be done at conditions other than vacuum,” says Stone.

“And of course, Strouse emphasizes, “it would hasten the elimination of mercury from pressure instruments employed around the world by national metrology institutes, standards labs, and industrial facilities.”