Degradation of organic contaminants by non-thermal plasma: Unraveling the pH-dependent mechanism based on experiment, density functional theory analysis and machine learning.

This paper studies the dependence of different pollutant degradation by dielectric barrier discharge plasma (DBD) technology on pH through experiments at different pH levels, density functional theory (DFT), and machine learning. When pH = 3.1, methyl orange (MO) and ibuprofen (IBP) showed a high degradation rate, but the degradation effect was poor under alkaline conditions. Bisphenol A (BPA), sulfamethoxazole (SMX), ciprofloxacin (CIP) and tetracycline (TC) showed excellent removal efficiency at pH = 10.6. The effects of reactive oxygen species (ROSs) on pollutant degradation were investigated using trapping agents and electron spin resonance (ESR) technology. A pivotal finding revealed the pH-dependent transformation of dominant ROSs during plasma degradation: superoxide radicals (O2-) primarily drove MO degradation under acidic conditions, whereas hydroxyl radicals (·OH) played the dominant role in SMX degradation under alkaline conditions. DFT calculations further identified the sulfonamide bond and benzene ring in SMX as preferential attack sites for ROSs, with their reactivity being significantly modulated by pH variations. Furthermore, ecological toxicity assessment demonstrated that the intermediates derived from SMX degradation under alkaline conditions exhibited significantly lower toxicity compared to those formed under acidic conditions, a conclusion consistently validated through both ECOSAR predictions and seed germination experiments. Finally, by innovatively integrating interpretable machine learning with feature engineering, this study successfully identified key molecular descriptors-such as pKa and the Highest Occupied Molecular Orbital (HOMO) energy level-that substantially influence degradation efficiency under varying pH conditions. These findings provide crucial theoretical insights and practical guidance for the targeted optimization of pollutant degradation processes.