Recognition of Chemical Warfare Agents Based on Nanosensor Array and Probabilistic Neural Network

Nanosensor is a new technology of highly integrating between nanotechnology and chemical sensors. For the advantage of good stability, high sensitivity, strong anti-interference and so on, it has good application foreground in the latest filed of gaseous sample detection. In this work, a novel analytical system based on nanosensor array and probabilistic neural network (PNN) was developed to detect chemical warfare agents such as sarin and mustard gas. The array consisted of four quartz crystal microbalance with a fundamental frequency at 10MHz. To improve the selectivity and sensitivity, four nano-membrane materials such as nano-zeolite modified with copper ion (CuZSM-5), modified carbon nano-tubes (CNT), hydrogen bond acidic fluorinated polymethyldrosiloxane (mTFPS) and polyepichloro-hydrin (PECH) were synthesized and selected as the Sensitive membrane material of chemical warfare agents. Then, a nanosensor array with these nano-film materials was developed. Combined with pattern recognition methods, a qualitative and quantitative identifying mode has been set up. The signals obtained from the array were analyzed with PNN to identify the toxic gases. The success rate of identification was 96.15%. The study integrate a variety of modern high-tech, which has novel methods and a high level of technology, an independent innovation research has been made for China´s anti-chemical warfare detector technology, opening up a new generation of chemical detectors and equipment developed in new field.