Saw Gas Sensor Array Pattern Recognition Technology

By the end of the 20th century, the terrorism incidents have aroused the attention of various countries. The development of gas sensors able to detect chemical warfare agents at trace concentrations seems be more imperative. Surface Acoustic Wave (SAW) sensors presents several advantages, such as, high sensitivity, small size and low cost, and so on, but the lack of sufficient selectivity of the applied coatings prevented the breakthrough of the mass-sensitive chemical sensors in many applications. However, since several "semiselective" sensors have been combined into arrays and coupled with subsequent pattern recognition methods, even selective analyses have been successful. Consequently, a marked increase about this sensor technology has been observed in recent years, especially for the detection of organic gases and the dissertation was based on this background.In the dissertation, the gas sensing properties of different sensitive films have been investigated, and the optimization sensor array from some sensors coated with different polymer layers were operated using pattern recognition technology, in order to reduce the output data of the multi-sensor array and improve the selective performance. The main research results were as follow:1. Surface acoustic wave gas sensor array test system has been improved. The test system was constituted by a sensor array, a gas chamber, an air distributing equipment, several mas flow controllers, a frequency counter, an RF switching, a personal computer and so on. Opening house of reference SAW device would lower the noise level of sensor output signal and there by enhance the precision of the response of sensors. SAW sensors coated with poly {methyl [3-(2-hydroxyl,6-bistrifluoromethyl) phenyl] propylsiloxane} (DKAP) were fabricated, and the sensitive characteristics have been investigated in detail. It was shown that DKAP-SAW sensor was sensitive to dimethyl methyl phosphonate (DMMP) and the frequency shifts of the sensor, acting as the response, were proportion to the concentrations of DMMP in the low range.2. Three polymers and a pattern recognition approach were chosen for the SAW sensor array. Based on a data set collected from 10 candidate sensitive materials for 12 analytes, the selection was accomplished by principal component analysis (PCA) and system cluster analysis methods. It was indicated that changing of sensors with similar film thickness have less influence on the selection of the polymers and the detection of the sensor array, but the effect of change of analyte concentration on the selection should be considered. Then, the chosen polymers were deposited on three SAW devices combined into the array, and PCA was adapted to classify three gases, DMMP, DCP (1,5-Dichloropentane) and ethanol. The results showed that DMMP, DCP and ethanol could be clearly distinguished by the SAW sensor array, indicating high selectivity of the sensor array.