AVS 50th International Symposium
    Homeland Security Topical Conference Wednesday Sessions
       Session HS+MM-WeA

Paper HS+MM-WeA8
A Novel Chemical Detector Using Cermet Sensors and Pattern Recognition Methods

Wednesday, November 5, 2003, 4:20 pm, Room 309

Session: Detection of Explosives and Other Chemicals for Homeland Security
Presenter: S.L. Rose-Pehrsson, Naval Research Laboratory
Authors: S.L. Rose-Pehrsson, Naval Research Laboratory
J. Ziegler, General Atomics
M.H. Hammond, Naval Research Laboratory
D. Gary, General Atomics
K. Caudy, General Atomics
Correspondent: Click to Email
Smart microsensor arrays are being developed by combining cermet electrochemical sensors, intelligent firmware and software to drive the sensors and analyze the data. The chemical microsensors offer a small size, light weight, low power and low cost alternative to conventional electrochemical sensors. The chemical microsensor architecture may be modified for detection selectivity of a variety of chemical species including chemical agents and combustible or corrosive gases. The microsensor arrays have potential application for monitoring hazardous chemicals in the part-per-million to part-per-billion range in a variety of internal and external environments. The arrays sense analytes using pattern recognition techniques to determine the presence of vapors of interest. General Atomics and the Naval Research Laboratory are developing this technology for the detection of chemical warfare agents and toxic industrial compounds (TICs). A test demonstrator has been developed with a three-sensor array, readout electronics, and system control software. The three-sensor array was exposed to 15 test vapors. The 15-analyte sources, including 2 blood agents, 10 TICs and 3 simulants were generated at 5 different concentrations in humid air. The cermet sensor array provided unique responses for the various analytes tested. Similar analyte types produced similar results. The sensitivity is sufficient to detect all the analytes at their respective exposure limits. Two different pattern recognition methods were developed to identify the analytes.