This article illustrates that miniaturization promises to bring chemical analysis out of the lab and into the field. Current research, based on varying approaches, is aimed at ultimately developing small, portable chemical analysis systems that are fast, accurate, and field-friendly. Some applications have been commercialized, and many more are still under development. One approach, taken by Sandia National Laboratories in Livermore, CA, is based on detecting elemental signatures—telltale traces of elements that fluoresce when the material is hit by an X-ray or gamma ray. An integrated, field-portable µChemLab will take a chromatographic approach to performing chemical analysis on gases and liquids. The enabling technology of Sandia’s approach is based on a new type of solid-state detector designed with a semiconductor crystal alloy of cadmium, zinc, and telluride, which can operate at room temperature—a key advantage over conventional silicon- and germanium-based devices. A different approach, based on chromatography, is the focus of another Sandia project to develop a field-portable chemical analysis device.
Chemical analysis is an extremely complex endeavor, requiring specialized facilities, expensive instruments, and highly trained individuals to perform analyses and interpret data. But now new developments in integrated microsystems—tiny intelligent devices that sense, act, and communicate—are poised to simplify the process.
Current research, based on varying approaches, is aimed at ultimately developing small, portable chemical analysis systems that are fast, accurate, and field-friendly. Some applications have been commercialized, and many more are still under development. In the not-too-distant future, these efforts could lead to a range of reliable, easy-to-use chemical sensors for nondestructive testing in the field.
One approach, taken by Sandia National Laboratories in Livermore, Calif., is based on detecting elemental signatures—telltale traces of elements that fluoresce when the material is hit by an X-ray or gamma ray. Ralph James, a scientist who is working on the project at Sandia, said: “By being able to measure the individual signatures, we can learn a lot about the composition of practically all objects. And we use that for a variety of applications.” Field instruments excite the material with a tiny gamma-ray source that is emitted from a lightweight probe weighing about 2 pounds. The targeted material fluoresces, releasing energies that are read by the detector in the probe, thereby identifying the element, James said.
Although the technique of irradiating elements has been used before, until now it has depended on detectors based on lithium-drifted silicon or high-purity germanium. Both of these materials must be cryogenically cooled, requiring large, bulky, and heavy cooling apparatus—making sensors impractical in many field applications. Silicon and germanium, which are excellent semiconductors, generate thermal “noise,” or current, at room temperature, degrading the resolution of the device, James explained. This requires that the detectors be cooled to be useful. Silicon- and germanium-based instruments, together with cooling equipment, weigh from about 35 to a few hundred pounds, he said.
The enabling technology of Sandia’s approach is based on a new type of solid-state detector designed with a semiconductor crystal alloy of cadmium, zinc, and telluride, which can operate at room temperature—a key advantage over conventional silicon- and germanium-based devices. Thermal noise that flows through the device is reduced by increasing the zinc component of the new chip. The new detector has a magnitude of thermal noise many levels lower than those of silicon and germanium at room temperature, eliminating the need for cryogenic cooling while offering roughly comparable performance, James said. The sensor can be housed in a unit that weighs about 10 pounds, can be strapped to the shoulder, and last all day, he said.
An integrated, field-portable pChemLab will take a chromatographic approach to performing chemical analysis on gases and liquids.
Typical size of the cadmium, zinc, and telluride chip is about one centimeter square and a millimeter thick. When struck by an incident X-ray, electrons and holes are produced in the crystal. By applying a voltage bias to the electrodes, the electrons and holes produce a small current pulse, which is unique for each X-ray of the element that is being detected. Each element has a specific “signature” in its fluorescence, which is given by the energy of the X-ray. These small current pulses are detected, recorded, and measured, allowing identification of the elements by their X-ray “fingerprints.” Currently, the detectors can identify all elements having atomic numbers between 11 and 92, James said. However, with the exception of hydrogen, these devices can, in principle, read virtually all the elements on the periodic table as well as the compounds associated with them.
James sees three big potential uses for this sensor: materials separation and identification, detection of toxins in the environment, and land mine detection.
One successful application of the sensors, commercialized since about 1996, is the detection of lead in paint. “We can hold these up to a wall that has paint on it. It will immediately read out everything that is in the paint sample—not only what is on the surface, but lead-containing paint from older layers that are much deeper,” said James. It also can be used by fabricators, scrap dealers, and recyclers to identify unknown materials, he added.
Because of its ability to probe through a few centimeters of soil and sand and detect many different elements, the device also can be effective for screening of suspect objects, such as land mines, James noted. “We can see particular materials found in antipersonnel land mines that are inconsistent with naturally occurring substances.” Another plus for land mine detection is that the device is minimally intrusive, and the samples and surrounding soil can be left undisturbed, he added.
One remaining challenge of these sensors is to increase their sensitivity, which is currently limited to detecting elements in part-per-million concentrations. “There seems to be no end to demand in terms of how sensitive we’d like to measure all sorts of elements and chemicals in the environment,” said James. As the device is moved away from the source, fewer X-rays come into contact with the detector, so more time is needed to characterize the elemental composition of unknown materials, he said. Work is being done to refine the crystal growth to produce larger, single pieces of the material, which will help increase the sensitivity of the detector. Room-temperature solid-state detectors produced from other semiconductors, such as mercuric oxide, are also under development and may present advantages for identification of some materials.
Tiny ‘Chem Lab’
A different approach, based on chromatography, is the focus of another Sandia project to develop a field-portable chemical analysis device. The device, known as μChem-Lab, will use chromatographic techniques to separate and analyze different components present in complex chemical mixtures.
Miniaturization will allow integrated sensing technologies to be placed on a chip, or series of chips, that can be housed in a handheld, portable, battery-operated unit, according to Duane Lindner, a senior manager at Sandia’s Livermore, Calif., laboratory. Speaking about the project at the Sandia-sponsored Microsystems Exposition held last September in Santa Clara, Calif., Lindner said that miniaturization is bringing chemical analysis into a highly integrated device that relatively unsophisticated users can exploit in a routine, user-friendly way.
The μChemLab project has demanding design challenges: part-per-billion sensitivity, fast response time, bare minimum false positive rate, portability, and low cost. To accomplish that, Sandia is looking at integrating a range of capabilities, in areas such as microoptics, microlasers, and separation technology.
Chromatography is a well-established suite of lab techniques to analyze the composition of gases and liquids. In chromatography, the mixture of chemicals—gas or liquid—is introduced to a “column,” or narrow tube filled with packing material. As the sample flows through the column, various elements of the mixture interact with the packing material in different ways. Some chemicals are “stickier” than others are, and the material impedes their flow. Various chemicals in the mixture are separated as the mixture advances through the column. A detector placed at the end of the stream monitors the separated chemicals as they come off the column.
A key challenge here is “figuring out on a microscale what is the correct analog of what you do in a laboratory,” according to Deon Anex, a staff scientist at Sandia/California. For example, conventional laboratory-scale chromatography is accomplished with a metal tube filled with packing material, hooked up with compression fittings and a pump. On a microscale, however, filled metal tubes and compression fittings are not an option. “Instead of tubes, we are using etched channels in a substrate, and seal those with a cover plate,” said Anex. Electric fields, rather than mechanical pumps, will be used to move liquids in microchannels.
Polymer materials with controlled porosity and high surface area are critical in separations and fluid control. New microporous polymers are being developed as packing materials, noted Anex. These can be UV-initiated, so that a lithography approach can be used to locate these polymers in the channels on the chip, he said.
Anex points to two advantages of chromatographic techniques. One is that chemicals in a mixture are separated and analyzed individually, minimizing interference of one component with another. Also, the device will be capable of running different kinds of separations in parallel. “You are now developing a signature, in terms of the arrival times on different channels for each compound. That allows you to look at more complex samples and have a higher confidence in the answer you are getting.”
The device will draw on microsensor detection technologies to be used with chemical separations. These include miniaturized surface wave acoustics, laser-induced fluorescence analysis, and electrochemical detection systems.
The design of liquid phase separations incorporated work being done at Sandia National Laboratories in Albuquerque, N.M., for optical communications, said Anex. "They were building small diode lasers and small optics for communications applications,” he said. “When we were developing our liquid phase separations capabilities, we recognized that these lasers and optics could be used to build small fluorescence-based detectors.”
The sensors incorporate two major optical elements developed at Sandia in Albuquerque. One is a vertical cavity surface emitting diode laser, or VCSEL. This is a small laser, on the order of 50 microns, which is very efficient, reducing the excess heat that would heat up the chip. The other component is micromachined optics, known as a Fresnel lens, which diffracts light.
The ChemLab project combines the VCSEL diode laser with the micromachined optics to steer the light through the chip, illuminating a section of the channel near its end. The optics collect the fluorescent light from the various bands that pass by the detector, and send that light through a small photo-diode. “In the space of a few millimeters are the light source, the optics, and the collection, mounted onto the chip. And adjacent to that is a small photo-diode to collect the light,” said Anex.
Sandia expects to have a field prototype early this year. Although the primary, and original, mission for this technology is national defense, such as analyzing explosive mixtures, Lindner also sees commercial applications. Among them: monitoring food for spoilage, or health monitoring.
“Although our first targets are explosives, we are trying to build a general platform, with modular separation channels,” said Anex. These modules will fit into a handheld unit with common connections, user interfaces, and communications for different applications. To switch to another application, say from explosives detection to biological weapons detection, one separation module could be switched for another, he said.
A new generation of photo-gravimetric-based microelectromechanical sensors for chemical detection are being developed at Oak Ridge National Laboratory in Oak Ridge, Tenn. The lab has developed sensors based on the microcantilever bending approach, according to Panos Datskos, the research scientist involved with the project. The cantilever, basically a tiny diving board that is roughly 100 to 150 microns long, 20 to 50 microns wide, and 0.2 to 0.5 micron thick, bends as molecules adsorb on its surface. “By looking at the bending, people can tell if something has adsorbed on it,” said Datskos.
The problems of most chemical sensors are related to issues of sensitivity and selectivity, said Datskos. Traditionally, selectivity is achieved by coating the cantilever with a chemical layer that has an affinity for a particular molecule or class of molecules, said Datskos. However, many coatings lack specificity—that is, they are sensitive to more than one type of molecule.
Oak Ridge has refined its microcantilever sensors with chemical coatings that change their chemical responsiveness when irradiated with a tiny diode laser. “With our technology, you can change the stress state of the coating and/or the substrate,” thereby increasing the selectivity and sensitivity of the sensor, Datskos said.
The MEMS device developed by Oak Ridge is based on a microcantilever that consists of a substrate and coating. The coating, substrate, or both can be irradiated to cause expansion or contraction. The ground state molecule has different reaction properties than the excited state molecule, Datskos explained. “You can change the selectivity of the device using selected wavelengths of photons. The sensitivity can be changed as well.”
Datskos is incorporating this concept into a small, portable device with a dynamic range that can go from neat, basically 100 percent, concentration to parts per trillion, by adjusting the sensitivity and selectivity. “We are able to change the selectivity and sensitivity in real time, without altering the coating,” he said. He envisions a sensor with five to 10 cantilevers that can be tailored to identify different materials.
Datskos sees various applications for such sensors: detecting fires, dangerous solvents, or chemical weapons.