Electro-oxidation of Polystyrene Nanoplastics Using a Titanium-based Anode Modified with Ag–Sn/Pb–PbO₂: ANN-PSO Modeling

Abstract Polystyrene nanoplastics (PS NPs) are persistent water contaminants that conventional treatment processes struggle to remove. In this study, a multilayer Ti/TiO₂–Ag–Sn/Pb–PbO₂ (TASP) anode was synthesized and evaluated for the electro-oxidation of PS NPs. The TASP electrode exhibited a high oxygen-evolution potential of 1.97 V vs. Ag/AgCl, reduced charge-transfer resistance, and an extended accelerated service life of 188 s under harsh test conditions, indicating greater relative stability than the reference electrodes. The effects of pH, initial PS-NP concentration, and current density on removal efficiency were investigated using a maximin Latin hypercube design (LHS) augmented with a two-level factorial, and both RSM and ANN-PSO models were developed for process prediction. The ANN-PSO model showed superior predictive performance ( R² = 0.953) compared with the RSM model ( R² = 0.919), confirming its superior ability to capture the nonlinear behavior of the system. Coupling ANN-PSO with a genetic algorithm identified the optimum operating conditions to be pH 6.46, initial concentration 20 ppm, and current density 11.44 mA cm⁻², with a predicted removal efficiency of 90.64%. Experimental validation under the optimized conditions gave 91.43% removal within 90 min, confirming the reliability of the proposed optimization framework. Kinetic analysis showed that PS-NP degradation followed pseudo-first-order behavior, with the TASP electrode exhibiting the highest apparent rate constant, approximately 3.2 times greater than that of the unmodified Ti/TiO 2 /Pb-PbO 2 electrode. These results demonstrate that combining an Ag–Sn-modified PbO₂-based anode and a data-driven modeling and optimization strategy provides an effective and interpretable approach for PS-NP electro-oxidation.