Study Of Carbon Nanotube(CNT)-based Sensor Array For VOCs Detection

Volatile organic compounds(VOCs)are common pollutants in the environment and cause great harm to human health.Therefore,it is of great significance to develop VOC gas sensors with high response,low cost and low power consumption to detect or monitor the VOC gas in the environment.Carbon nanotubes are considered to be excellent gas-sensing materials due to their unique hollow structure and high specific surface area.This paper mainly explored the effects of various modification methods on the morphology and VOC gas sensing properties of carbon nanotubes,including changes in gas sensitivity and selectivity.Two different pattern recognition algorithms,namely principal component analysis and artificial neural networks,were used to analyze the ability of the sensor array composed of different carbon nanotube sensors to distinguish different VOC gases.Firstly,the effects of carboxylation,amination and polyaniline modification on the carbon nanotubes(CNTs)were investigated.The difference between the dielectrophoresis method and the drop casting method was also studied.The results showed that the carboxylation and amino modification can reduce the entanglement and agglomeration of carbon nanotubes.Polyaniline can be synthesized and wrapped round carbon nanotubes using oxidative polymerization method.Dielectrophoresis had better positioning of carbon nanotubes than the drop casting method.Polymer-carbon nanotube composites can be prepared by dissolving polymer particles in CNT dispersion solvent before dielectrophoresis.Secondly,the gas sensing properties of different modified carbon nanotubes on VOC gases were studied.The results showed that different modification could improve the response of carbon nanotubes to VOC gases and the selectivity of different VOCs also changed.Finally,6 carbon nanotube sensors with different modifications were combined to form a sensor array.The ability of the array to distinguish VOC gas was studied by using two pattern recognition algorithms:principal component analysis and artificial neural network.The principal component analysis showed that the array could distinguish the seven kinds of VOC gases studied.The accuracy of predicting the gas species was 92.3%after the artificial neural network was trained,and the total accuracy was 88%when the prediction of concentration was made at the same time.