This work describes the use of an array of potentiometric sensors and an artificial neural network response model to determine perchlorate and sulfide ions in polluted waters, by what is known as an electronic tongue. Sensors used have been all-solid-state PVC membrane selective electrodes, where their ionophores were different metal-phtalocyanine complexes with specific and anion generic responses. The study case illustrates the potential use of electronic tongues in the quantification of mixtures when interfering effects need to be counterbalanced: relative errors in determination of individual ions can be decreased typically from 25% to less than 5%, if compared to the use of a single proposed ion-selective electrode.
Keywords: artificial neural network; electronic tongue; ion-selective electrode; perchlorate; phtalocyanine ionophores; sulfide.