Synergistic degradation of the aqueous antibiotic norfloxacin by nonthermal plasma combined with defective titanium dioxide exposed {0 0 1} facets.

• Catalytic activity of defective TiO 2 -exposed {0 0 1} facets was higher than TiO 2 P25. • Synergistic factor significantly increased with increasing TiO 2-x dosage. • The contribution of 1O 2 was the most significant in NOR degradation. • NOR degradation achieved by ring opening, defluorination and mineralization. • The GA-ANN model had high accuracy in predicting the NOR degradation efficiency. The wide band gap of titanium dioxide (TiO 2) photocatalyst leads to its narrow optical response, which seriously affects the application of TiO 2. In order to enhance its catalytic efficiency, defective TiO 2 -exposed {0 0 1} facets (TiO 2-x) was prepared and applied to pollutants degradation by combing with the dielectric barrier discharge nonthermal plasma (DBD-NTP), so as to use the ultraviolet light and other active species produced by DBD-NTP. Taking aqueous norfloxacin (NOR) as a model pollutant, the degradation performance of NOR by DBD-NTP and TiO 2-x was evaluated, the synergistic catalytic mechanism and degradation pathways of NOR based on heterogeneous catalysis were proposed. An artificial neural network model based on genetic algorithm optimization (GA-ANN) was constructed to predict the NOR degradation efficiency. The characteristics of TiO 2-x were characterized, and the effects of influencing factors on the NOR degradation efficiency were investigated. The results showed a significant synergistic catalytic effect in NOR degradation by DBD-NTP combined with TiO 2-x. The synergistic catalytic effect substantially increased with increasing TiO 2-x dosage, and the synergistic factor increased by 3.3 times. The reactive agents 1O 2 , OH, O 2 −, h+ and e− played an essential role in NOR degradation, with the contribution of 1O 2 being the most significant. TiO 2-x had good catalytic stability. The degradation pathways of NOR mainly included breakage of the piperazine ring and quinoline ring, decarboxylation, benzene ring defluorination and mineralization. The established GA-ANN model showed high accuracy (mean square error = 0.0016, correlation coefficient = 0.9956) and could be used to predict the degradation efficiency of aqueous NOR. [ABSTRACT FROM AUTHOR]