Detection of ammonia based on a novel fluorescent artificial nose and pattern recognition

A simple and rapid fluorescent sensor array for identification and quantification of different concentrations (ppb level) of ammonia was proposed in this paper. Employing porphyrin, porphyrin derivative and chemically responsive dyes as the sensing elements, the developed sensor array of artificial nose showed a unique pattern of fluorescence changes upon its exposure to ammonia for just 4 min. And the eigenvalues from raw fluorescence spectra were analyzed by means of pattern recognition algorithm, including hierarchical cluster analysis (HCA), principal component analysis (PCA) and back-propagation neural network (BPNN). The results showed that HCA and PCA, which were used to assess the feasibility and effectiveness of discrimination of fluorescent sensor array, revealed a distinct separation between samples. BPNN were used for automatically classifying and predicting concentration of ammonia, and the accuracy was 97.55% while the relative standard deviation (RSD) was up to 3.67% for real samples. It indicates the fluorescent artificial nose system is a rapid and feasible sensing platform for the identification and quantitative analysis of gases, and also shows the possibilities in the field of environmental gas detection and environmental gas monitoring.