Eco-friendly AOP for Amoxicillin removal: hybrid optimization, by-products identification via QUECHERS-HRMS technique, and toxicity assessment

Abstract This study investigates NaClO/FeSO 4 -based Fenton-like oxidation for Amoxicillin (AMX) degradation in aqueous systems. Process optimization using Response Surface Methodology coupled with Central Composite Design (RSM-CCD) identified optimal conditions of [NaClO] = 800 µM, [AMX] = 35 mg L −1 , [FeSO 4 ] = 5 mg L −1 , and pH = 3, yielding 91.8% removal. A hybrid RSM, Artificial Neural Networks (ANN), and Genetic Algorithm (GA) modeling approach further enhanced performance, predicting 99% efficiency (R 2 = 0.991) and achieving 97% removal under refined conditions (AMX concentration of 34.85 mg L −1 , [NaClO] = 508.4 µM, [FeSO 4 ] = 7.83 mg L −1 , and a pH of 3.14). Green-synthesized IONPs nanoparticles supplemented the homogeneous system for by-products formation and toxicity assessment experiments, achieving 99.9% AMX removal within 60 min ( k = 0.114 min −1 ) when combined with hydroxylamine. Trace and ultra-trace analysis using QuEChERS extraction coupled with Direct Infusion-High Resolution Mass Spectrometry (DI-HRMS) enabled quantification of the parent compound and seven transformation by-products. Ecotoxicity testing with Chlorella vulgaris demonstrated transient inhibition followed by substantial detoxification.