POTENTIAL USE OF AN ELECTRONIC TONGUE FOR PROCESS MONITORING OF FLUE GAS CONDENSATE : WITH FOCUS ON ANTIMONY DETECTION

Sammanfattning

The increasing need for real-time monitoring of heavy metal emissions from waste incineration facilities has sparked interest in alternative measurement tools. This study investigates the feasibility of using a pulsed voltammetric electronic tongue to detect and distinguish differences in antimony concentrations in flue gas condensate from Vattenfall’s waste incineration facility in Uppsala. Three different pulse protocols were applied: Negative, clean and large amplitude pulse voltammetry, and the materials of the working electrodes were gold, platinum and rhodium. Multivariate data analysis using principal component analysis was used to interpret the complex and vast data produced by the voltammetric electronic tongue. The results indicated that the gold electrode under negative pulse conditions, no oxidative potentials, provided the most accurate differentiation between antimony levels due to its low hydrogen evolution activity and strong surface affinity for antimony. The effects of antimony fouling on the gold electrode significantly enhanced detection sensitivity, while the clean pulse protocol, designed to avoid fouling by incorporating oxidative steps or positive potential, showed weak discriminatory power. The large amplitude pulse protocol demonstrated the electronic tongue’s partial ability to distinguish between flue gas condensate samples of varying metal composition and pH, however the overlapping signals and sample measurements exceeding the detection threshold, limited interpretability. Besides the detection analysis, the accumulation of antimony on the gold electrode highlights the possibility of electrochemical heavy metal capture, opening new avenues for integrating sensing and capture in a single system. While the electronic tongue shows potential as an analytical instrument, the configuration used in this study is not yet suitable for in-situ process monitoring, and future work should focus on expanding voltammetric protocols, improving robustness, and advancing data modeling techniques for complex matrices such as flue gas condensate, for instance through the application of machine learning and artificial intelligence.