Improving ozone-based processes for persistent pesticide degradation

This study evaluated the efficiency of different ozonation-based systems for the removal of two recalcitrant herbicides, paraquat (PQ) and diquat (DQ). Solar photolysis and magnetite (Fe ₃ O ₄ ) as a catalyst did not promote degradation or mineralization, confirming the need for stronger oxidative conditions. Simple ozonation showed selectivity toward DQ (83%), while PQ reached 34%, with limited mineralization (26%). Increasing the ozone dose (1.8 to 3 gh ^− 1 ) improved mineralization by 8%, albeit with higher energy demand; 1.8 gh ^− 1 was selected as the best ozone dose to balance between efficiency and sustainability. Catalytic ozonation with magnetite (0.5–3 gL ^− 1 ) reduced efficiency. In technical grade DQ, mineralization was almost negligible at 1.8 gh ^− 1 , evidencing a higher oxidant demand. At pH 9 and higher O ₃ dose (3 gh ^− 1 ), > 85% degradation was achieved within 15 min, but with low mineralization (22%), indicating that the catalyst accelerates initial degradation without favoring mineralization. For PQ, catalytic ozonation did not outperform solar ozonation, even at high ozone doses, suggesting the formation of more stable byproducts. In contrast, solar photocatalytic ozonation with TiO ₂ (50 mgL ^− 1 ) improved degradation efficiency (> 85%). Toxicity assays with Allivibrio fischeri and Artemia franciscana confirmed that high achieved degradation rate adoes not ensure ecological safety, as photocatalytic ozonation generated the most toxic intermediates, while solar ozonation promoted detoxification. These findings highlight that advanced oxidation systems must be assessed not only for removal efficiency but also for their ecotoxicological outcomes, with, in this case, solar ozonation emerging as the most sustainable alternative for bipyridyl herbicides in water matrices. Graphical Abstract