Development of an Electronic Tongue for the Analysis of Particulate Matter Samples Collected in the City of Rio de Janeiro

Abstract

This study presents a voltammetric electronic tongue for the discrimination of fine particulate matter (PM 2.5) collected at three distinct urban locations in Rio de Janeiro, Brazil. A Screen-Printed Electrode (SPE) was developed using a low-cost handmade ink (optimized at 20% w/w glass varnish, 40% w/w graphite, and 40% w/w solvent, with PET bottles used as the substrate). The electrode was characterized by electrochemical impedance spectroscopy (EIS), cyclic voltammetry (CV), scanning electron microscopy (SEM), and Raman spectroscopy. Following acid digestion, PM samples collected from the Rio de Janeiro Botanical Garden (RJBG), the Rebouças Tunnel (RT), and the Institute of Chemistry at UFRJ were analyzed by square wave voltammetry (SWV). Exploratory analysis using principal component analysis (PCA) and hierarchical cluster analysis (HCA) demonstrated that raw voltammetric data provided significantly superior sample separation compared to normalized data. Temporal evaluation revealed that seasonal factors and dry weather conditions influenced the grouping behavior, causing partial overlap between RJBG and RT samples. Finally, an artificial neural network (ANN) classification model constructed using raw data achieved a classification accuracy of 75% in the independent test set, classifying samples according to their sampling-site labels, although temporal and seasonal factors may also have contributed to the observed differences.