"We were building the PA-IR spectrograph for materials-science applications, but it turned out to be about 1,000 times more sensitive than we expected, so we did a few calculations and realized it could be used to detect airborne chemical and biological agents," said John Rabolt, chairman of the University of Delaware's Materials Science and Engineering Department. Rabolt and Mei-Wei Tsao, a research professor, have patented the spectroscopy.

The PA-IR spectrograph can detect trace amounts of chemical and biological weapons before they become a biohazard, enabling alerted personnel to leave the area or get into protective clothing. The device can detect chemical agents in vapor, liquid or solid phases and is being fitted with a telescopic collection system for detection at a safe distance.

"The chemical agents we can detect just by looking in the air. . . . They have very specific signatures, and the PA-IR spectrograph can see them in very low concentrations — that is the key," said Rabolt.

To detect chemical agents, the device needs nothing but a sample to be tested. But to detect biological agents an analyte for that specific biological agent must be mixed with the sample before the test. The analyte's antibody qualities makes it bind chemically only with the specific biological agent for which it was designed to test. Then the PA-IR spectrograph can be tuned to quickly ascertain whether the analyte is present untainted, or whether it has become bound with a target biological agent.

Real-time monitor

The PA-IR spectrograph also has broad industrial applications on the factory floor, enabling it to monitor in real-time the thickness of films as well as the precise chemical composition of coatings and liquids as they are being produced. Manufacturing plants could cut their wastes by 25 percent, Rabolt said, by installing PA-IR spectrographs on their lines to perform real-time monitoring.

"We originally developed PA-IR spectroscopy for real-time monitoring of processes on the factory floor, for instance monitoring the growth of oxide on silicon wafers in real-time, or monitoring the thickness of films as they are manufactured. About 25 percent of production films end up in the landfill because they are of the wrong thickness, or the wrong orientation, or not stretched enough. You just can't monitor the process fast enough. so companies that produce films typically only measure them once a week, but we could set up a real-time monitoring system that sets off an alarm whenever the film goes out of spec," said Rabolt.

IR spectroscopy dates back to the 1960s when it was discovered that IR light selectively absorbs certain wavelengths in a unique pattern called a "fingerprint" or "signature." By using a prism or diffraction grating, it was discovered, the absorption signature could be quantified for each wavelength in the IR spectrum, thereby uniquely identifying the substance under test. Since then various laboratory setups have been cataloging the IR absorption signatures for practically all known substances. These IR signature catalogs are commercially available today, as are standard software suites to automatically make a comparison with an unknown sample's signature in about 30 seconds.

"You put a broad-spectrum source of light through the sample, and certain components of the light will be attenuated, so when we put the light through a dispersing element, like a prism or a grating, it creates a signature," said Rabolt.

Before the PA-IR spectrograph, the most advanced instrument was the Fourier Transform IR (FT-IR) spectrograph, which requires an air bearing, so it has to be hooked up to high-pressure air. In addition, the optics for the FT-IR spectrograph are very fragile, since it employs a single-element detector and a mirror that rides on a rail with only 1-micron clearance. The mirror reflects the various components of the signature onto the single-element sensors by moving back and forth to scan the entire absorption spectrum.

"The Fourier Transform IR spectrograph is usually confined to the laboratory, because just bouncing it around in the back of a truck could damage it," said Rabolt.

In contrast, the PA-IR spectrograph uses a chip with an array of hundreds of IR sensors, enabling it to photograph the entire absorption signature for a substance in an instant. So instead of filling a room with FT-IR equipment, the PA-IR spectrograph has no moving parts and can be packed into a rugged, shoebox-sized unit.

"The PA-IR spectrograph is basically just an infrared camera, which was developed by the military, but was only declassified about five years ago," said Rabolt. "We just add a light source and a prism."

The PA-IR spectrograph prototype built by Rabolt uses an IR sensor with 256 by 320 pixels, enabling it to capture a thermal signature in about 10 milliseconds today, and in as little as 10 microseconds in future devices, Rabolt said.

"Right now we have a fairly broadband instrument, but you can also get planar-array IR cameras that only measure very narrow regions. For instance," he said, "if you wanted to look for chemical agents only, then you could build it much more compactly. . . . Future devices might fit in a 4-inch cube that you could dispense over a battlefield to monitor the air, and with wireless technology all the information could be sent back to a laptop computer miles away."

The major expense of the PA-IR spectrograph is the IR camera itself, costing $30,000 or more today. The camera's sensor chip must be thermoelectrically cooled, accounting for most of its expense. That cost will come down by an order of magnitude in the next few years, Rabolt predicted, once manufacturing is ramped up for his "cubes" and similar anti-terrorist devices.

"There are only a few manufacturers of the IR cameras at this stage, but that will open up. With the current state of affairs after Sept. 11, there will be much more of a demand for these in many other applications besides just our PA-IR spectrograph, so the price will come down very quickly. That makes the idea of a couple-of-thousand-dollar cube that you can spread around a battlefield much more realistic from the visionary standpoint," said Rabolt.

Three models planned

Three new kinds of PA-IR spectrographs are on the drawing board for Rabolt's group: a small, inexpensive cube; a specialized version that measures only the part of the spectrum that detects chemical agents such as sarin nerve gas; and a generalized unit that measures the entire IR spectrum. The current $30,000 IR camera only measures about half the IR spectrum while the camera needed for the generalized version measures the entire spectrum but costs $250,000.

"We say miniaturize, specialize and generalize are the three things we are going after next. We are already working on the miniature cube version, so that will be ready in about six months. Then the specialized version is about one year out. But the generalized version will have to wait until a scientific agency provides us with the funds to purchase the expensive broad-spectrum IR camera," said Rabolt.

An audio recording of reporter R. Colin Johnson's full interview with John Rabolt can be found online at AmpCast.com/RColinJohnson.