A novel chemical detector using colorimetric sensor array and pattern recognition methods for the concentration analysis of NH3

Abstract

With a colorimetric sensor array comprising chemoresponsive dyes, a simple, rapid, and cost-effective integrated system for differentiating low-concentration gases was described. The system could be used to identify gases by detecting the color change information of the chemoresponsive dyes based on porphyrins before and after reaction with the target gas; the colorimetric sensor array images were collected by a charge coupled device and processed with image analysis to get the color changes of the dyes in the array. Temperature, humidity, and flux of the chamber could be detected and displayed on the personal computer screen. A low-concentration [30-210 ppb (parts per 10(9))] NH(3) was detected by the system. This prototype successfully differentiated four concentration levels of NH(3) in less than 1 min. Pattern recognition methods, such as the backpropagation neural network and the radial basis function neural network, validated the effect of the developed sensor system both with 100% classification with feature vectors at single time point as inputs.