Voltammetric electronic tongue for the discrimination of antibiotic mixtures in tap water

The identification and quantification of antibiotics in different types of water is of growing interest due to the resistance their presence can induce in many bacterial species. Although the electrochemical analysis of electroactive antibiotics via voltammetric techniques is rapid, simple and cost-effective, their simultaneous detection is often challenging due to their overlapping oxidation peaks. To address this issue, we have developed a voltammetric electronic tongue for the discrimination and quantification of antibiotic mixtures in tap water. For that purpose, three different screen-printed carbon electrodes were modified with "green"-synthesized metallic nanoparticles (NPs) of copper and silver, and commercial titanium dioxide NPs. The electrochemical behavior of the electrodes against the detection of six antibiotics was first investigated by cyclic voltammetry, proving the catalytic effect of the selected NPs. Next, principal component analysis and artificial neural networks (ANNs) were used to achieve the accurate discrimination and quantification of three selected antibiotics, cefepime, daptomycin, and vancomycin, at the mg/L level with a normalized root mean square error (NRMSE) of 0.055 for the test subset. Finally, after validation of the ANN model, the analysis of tap water samples spiked with the three selected antibiotics was carried out to test its applicability for water quality monitoring, showing a good agreement with the expected values (average recovery of 95 %, and NRMSE of 0.034).