Terahertz spectroscopy hits the market
Light moves through lenses, and radio moves through waveguides. But what of the waves in between—the wider-than-light but tighter-than-radio waves that dwell in the submillimeter realm? Water vapor in the atmosphere tends to absorb them. They can penetrate the human body no more than a few millimeters, so their medical scanning applications are similarly limited.
Yet though the territory be but little used by us, 98% of all detectable photons in the universe fall in this range, characterized by submillimeter wavelengths or terahertz (THz) frequencies. Astronomers love THz. Those frequencies emerge from parts of the early universe opaque to visible light.
Terahertz waves are also the best frequencies for gas phase rotational spectroscopy.
Gas phase rotational spectroscopy is what makes terahertz useful to the researchers at United Technology Corporation Aerospace Systems (UTAS) in Danbury, Connecticut. They create terahertz sensors that sniff out chemical threats in airports, shipping containers, and military terrain. The sensors pump a sample of air down to pressures much lower than Earth's atmosphere, typically 70 torr (0.09 atm). The device then shines terahertz frequencies through the sample and measures the rotational spectrum.
Electronic absorption spectroscopy—the kind most of us are familiar with—usually involves photons of visible light. When electrons in a molecular bond absorb a photon, they get excited and jump to a higher energy level.
Rotational spectroscopy, on the other hand, detects rotations of atomic nuclei or entire molecules. The rotations typically manifest themselves at submillimeter wavelengths or terahertz frequencies. Even slight differences in the mass of an atom's nucleus alter its rotation enough to shift its terahertz spectrum, so this type of spectrometry can distinguish even between two different isotopes of the same element. The weight might differ by a single neutron.
But to be that sensitive, the spectral analysis must be conducted at low pressure. At atmospheric pressures, the spectral lines broaden out to 5 or 10 GHz wide and blur together, making it hard to resolve one from another. That's why UTAS's device pumps the gas down to 70 torr (0.9 atm). At such low pressures the spectral lines are very sharp, easily seen with the device's 1-MHz resolution.
The rotation of molecules in the gas phase gives rise to terahertz spectral features that are broad at atmospheric pressure (top). The same features become narrower and easier to distinguish at 0.1 atm (bottom). CREDIT: Alex Majewski, UTAS
Resolution is important, particularly if you don't know exactly what you're looking for. Most molecules have very characteristic spectra at submillimeter frequencies, but if the lines all converge there's no way to tell what you're looking at. Laboratory instruments typically use cryogenic temperatures to reduce signal noise even more, but cryo-cooled detectors are pricey, complex, and hog energy.
"Better, faster, cheaper. That's what clients want," says Alex Majewski, a physicist in the group. "Better, faster, cheaper" is an old NASA mantra, and that should come as no surprise. The UTAS Danbury location was formerly part of the Perkin-Elmer optical group that built the Hubble Space Telescope's Optical Telescope Assembly. Perkin-Elmer subsequently sold the business and it passed through different hands before being bought by United Technologies Corporation in 2012.
Majewski joined the group in 2002, when it was owned by Goodrich. This was during the anthrax scare, and the federal government was reaching out to the aerospace and defense community for new methods of detection. Goodrich considered the problem, but decided terahertz wasn't a good fit for biological agents. It was, however, ideal for detecting chemical and radioactive threats. Gas phase rotational spectroscopy in the terahertz band had been practiced since the 1930s, but in 2002, no one had terahertz sources that combined high power with a wide range of frequencies.
Low-noise detectors that are independent of cryo-cooling weren't available, either. Instead, two technology choices existed. One was an electronic oscillator like the ones inside microwave ovens, which could multiply a single frequency many times to get up into the GHz range (THz frequencies range from 300 GHz to 3000 GHz) at high power, but narrow frequency. The other choice was a photonic technique, which used two lasers with slightly offset frequencies. By tuning the lasers, it was possible to get a wide range of terahertz frequencies.
Goodrich's team developed a novel photonic approach by driving a gallium arsenide coherent photomixer transceiver (developed by MIT Lincoln Labs) with a dual laser pump, and adding its own control electronics. The photomixer is nonlinear and radiates in many frequencies, so by tuning the laser pump they found they could get frequencies across the full terahertz spectrum, from microwaves up to the far IR. The technique is still low-power, but is a significant advance in frequency resolution and control, and can resolve frequencies to 1 MHz.
With a resolution of 1 MHz, scanning over the entire terahertz range, which is 3 million MHz wide, can take an awful lot of computing time. To speed up the analysis, the researchers repurposed an algorithm developed for another application. Together, the laser pump photomixer and the detection algorithm suppress noise and speed up signal extraction in spectrally cluttered environments, such as the ambient atmosphere of an airport.
When Physics Today asked what type of researcher UTAS looks to hire for its sensor work, Majewski said "terahertz engineers." Then he laughed. Terahertz engineers may not be as rare as quantum mechanics, but it's an uncommon specialty. More often, physicists, optical engineers, and electrical engineers come together to work on terahertz frequencies. Because they come with very different approaches, the collaboration is often a fruitful one. The team also employs mechanical and software engineers, as well as systems engineers. The systems engineers tend to be physicists who can consider the project holistically.
"I need to build something. What does it need to do? It needs to be x size, weigh y, use z current. That's often the kind of question we work on," said Majewski, a PhD physicist and systems engineer himself. And although UTAS's system avoids cryo-cooling when solving challenges on the job, Majewski and his coworkers do appreciate the cold in their spare time. UTAS Danbury fields its very own ice hockey team with no fewer than three PhD physicists. Majewski plays right wing.
Although the major application of terahertz frequencies reamins gas phase rotational spectroscopy, terahertz telecommunications applications also receive ample attention and funding. The appeal is clear: 3000 GHz of unclaimed frequencies beckon with their virtually unlimited bandwidth.
But atmospheric absorption and lack of affordable components make "THz telecom" a long shot, especially since we still have available frequency in wavelengths we can work with. UTAS hasn't pushed forward into the telecommunications realm yet. The company continues to focus its efforts on point sensor gas phase detection; improving the application that terahertz does best.