Chlorophenols identification in water using an electronic nose and ANNs (artificial neural networks) classification.

Electronic artificial noses are being developed as systems for the automated detection and classification of odours, vapors and gases. In the food industry, such devices are used as aids for quality control or process-monitoring tools. An electronic nose (EN) is generally composed of a chemical sensing system and a pattern recognition system (e.g. artificial neural network). An EN based on a non-specific conducting polymer array was used to monitor chlorophenols in water samples. Operational parameters for the EN were optimized by a Plackett-Burman factorial design. The experimental parameters studied were: sample volume, platen temperature, sample equilibration time, loop fill time, sample pressurization time and injection time. Optimal experimental conditions were applied to chlorophenols determination and differentiation in ultrapure water samples spiked with the EPA listed chlorophenols. Data analysis was carried out using principal component analysis (PCA) and artificial neural networks (ANNs) to predict the chlorophenols presence in water samples. The obtained results showed that it was possible to differentiate the five chlorophenol groups: monochlorophenol, dichlorophenol, trichlorophenol, tetrachlorophenol and pentachlorophenol. Differentiation of chlorophenol groups was based on Mahalanobis distance between the formed clusters. This Mahalanobis distance is designated by the Quality Factor, a value >2 for this quality factor means a good differentiation between the clusters.