A sampling device smaller than the tip of a fingernail promises big results for detecting and analyzing trace chemicals. Developed by the Department of Energy's Sandia National Laboratory in Albuquerque, N.M., this tool is a super-miniaturized version of a traditional preconcentrator used to collect sample gases for analysis. The active area of the MEMS device is only 2 by 2 mm (Fig. 1). Already part of Sandia's initiative to build a handheld "chemistry laboratory," it potentially can be integrated with other microchemical detectors, including a mass spectrometer or an ion-mobility spectrometer.
The tiny size will allow chemical testing using small handheld instruments, eliminating the need to send samples to a large laboratory. This would be beneficial, for example, to soldiers in battle who must know immediately what chemical they're encountering. Obviously, there's no laboratory handy and no time to wait for an analysis. The preconcentrator is part of Sandia's µChemLAB, a miniature, handheld dual-channel gas-phase analysis instrument consisting of three cascaded microfabricated components (Fig. 2).
"Because it can work with different types of microanalytical systems, this \[preconcentrator\] is receiving a lot of attention," says researcher Ron Manginell, who has been working on the device for the past three years. "It's small, uses minute amounts of power, is extremely portable, and is inexpensive to produce, all making it very interesting to both industry and the military."
A traditional preconcentrator is usually large. It consists of a cigarette-sized stainless-steel tube, typically 100 mm long by 6 mm in diameter, and packed with adsorbent resins between two glass-wool plugs. A pump forces the sample gas through the tube, where it's adsorbed into the material. Next, the steel tube goes into a benchtop thermal desorber and is heated to 200°C. The gas escapes from the tube for analysis by a detector, such as a benchtop gas chromatograph system, that determines the chemical's nature.
According to Manginell, this traditional system is bulky, slow, and must be done in a laboratory setting, which isn't at all practical for field testing. Project leader Greg Frye-Mason says the microfabricated planar preconcentrator is a revolution in front-end sampling devices. Using standard IC microfabrication technology that allows 200 units to be built on a single 4-in. silicon wafer, it has a silicon base topped by a 0.5-µm layer of silicon nitride. The silicon-nitride membrane, formed by etching the silicon away, holds a patterned platinum heater called a microhotplate. A thin layer of a microporous adsorbent material is placed on the front surface of the heater.
To test the device, the adsorbent region is covered by a Pyrex lid with an inlet and outlet port located on the top. Gold pads surround the device and help connect the platinum heater to the macroscopic world electrically.
The miniature preconcentrator operates much like its larger relative. First, a small pump pulls air containing a chemical over the adsorbent material. Then, current flows through the platinum, heating up the microhotplate to 200°C. The high temperature causes the chemical to be released from the adsorbent material so it can be analyzed by a microdetector system.
"All this happens in the blink of an eye," Frye-Mason says. "It takes 6 ms and 100 mW of power to reach 200°C. That's 1000 times faster than using the conventional method." What makes this possible is the fact that the device is so small it doesn't take much current or time to heat up. Because of its tiny size and planar design, this micro preconcentrator is ideal for chip-based microanalytical systems such as Sandia's chem-lab-on-a-chip concept.
The adsorbent material most frequently used in testing the device is a sol gel created by Sandia researcher Jeff Brinker. This gel can be "tuned" to collect certain types of molecules and not others. Scientists also have tested other adsorbent materials with the microfabricated planar concentrator.
Manginell has been involved in all aspects of the preconcentrator's development from the beginning. "The initial development was fast," he recalls. "It took us six months to come up with a prototype. Since then we've been refining and modeling it."