Optimization Of Reaction Performance And Prediction Of Photocatalytic Treatment Of Antibiotic Wastewater Based On Artificial Neural Network

Antibiotics are widely used to prevent and treat bacterial infections and their residues are detected in various aqueous environments.Antibiotics and residues lead to serious environmental pollution,threaten human health and deteriorate biological ecology.Therefore,it is urgently needed to develop technology for effective and thorough treatment of antibiotic wastewater.Photocatalytic degradation of organic pollutants has the advantages of environmental friendliness,safety,and thoroughness.Photocatalytic technology was combined with neural networks to address environmental pollution caused by the widespread use of antibioticsin this work.To enhance the photocatalytic activity,TS-1/C₃N₄,Bi/Bi₂O₃/Bi₂WO₆ composite materials were prepared based on carbon nitride（g-C₃N₄）and bismuth tungstate（Bi₂WO₆）and used to photocatalytic degradation of the typical antibiotics（ofloxacin and sulfapyridine）in wastewater.Experimental study and theoretical modelling of the photocatalytic degradation process were studied in this work.The photocatalytic mechanism and degradation route were proposed.The corresponding artificial neural network（ANN）models were developed based on the experimental data.Genetic algorithm（GA）was used to optimize operational parameters.Experiment and model study on the synergistic effect of environmental factors（p H,wastewater composition,etc.）on photocatalytic treatment of antibiotic wastewater under optimal experimental conditions.This study can provide theoretical basis and technical support for the screening of catalysts and the application of photocatalytic treatment of raw antibiotic wastewater.The main content are shown as follows:（1）Titanium silicon molecular sieve（TS-1）loaded on carbon nitride（C₃N₄）（TS-1/C₃N₄composite）was synthesized and used for the photocatalytic treatment of ofloxacin（OFX）wastewater.The influence of operating parameters and synergistic effect of wastewater components on removal efficiency（RE）was studied.The reaction mechanism and degradation route of photocatalytic degradation of OFX was proposed.Based on the experimental data under different operating parameters,the optimal operating parameters（TS-1 loading at 58.60%,catalyst dose at 1.55 g/L,luminous power density（LPD）at 49.38 m W/cm²）was obtained by combining ANN model with GA method.The RE（82.92%）was measured under the optimal operating parameters and compared with the calculated value.The absolute relative deviation（ARD%）between the experimental value and the predicted value was 2.01%and indicated the ANN model has a good predictive ability.Based on the experimental data of the synergistic influence of wastewater components,an ANN model was established for photocatalytic treatment of OFX wastewater.Validation experiments about the synergistic effect of wastewater components were carried out.The ARD%of the synergistic effects of cations,metal ions and anions are 6.88%,1.04%and 1.77%,respectively.The above results show that the model has good predictive ability for the synergistic effect of wastewater components.（2）Bi₂WO₆ and Bi/Bi₂O₃/Bi₂WO₆ composite photocatalysts were synthesized.Firstly,the experimental and model study of Bi₂WO₆ photocatalytic treatment of OFX was carried out.The OFX wastewater RE was measured under different operating parameters.An ANN model was proposed based on the experimental data.GA was used to optimize operational parameters.The RE（98.16%）was measured under the optimal experimental conditions（catalyst dose at 0.96 g/L,p H at 7,LPD at 41.38 m W/cm²）.The ARD%between the experimental and predicted RE was1.4%,indicating that the neural network model had good predictive ability.Secondly,Bi/Bi₂O₃/Bi₂WO₆ composites were synthesized to improve the photocatalytic removal performance of sulfapyridine（SP）wastewater by Bi₂WO₆ catalyst.The effects of different operating conditions on the photocatalytic treatment of SP wastewater were measured,and the corresponding electrochemical properties of the composites were analyzed.The optimal operational parameters（catalyst concentration at 1.32 g/L,p H at 6.35,LPD at 48.68 m W/cm²）were obtained by combining ANN model and GA method.The ARD%between measured and predicted value was 1.34%,indicating that the ANN model has a good predictive ability.Besides,to further investigate the extrapolated ability of the ANN model to predict the synergistic effects of wastewater components,photocatalytic degradation experiments were performed on a more complex OFX solution（synergistic effects of metal ions,total nitrogen and total phosphorus）using Bi₂WO₆ as a photocatalyst.The corresponding ANN model was established based on the experimental data.Additional validation experiments were conducted out under the presence of wastewater constituents（included metal ions,total nitrogen,and total phosphorus）.The ARD%about the synergistic effect of 10 wastewater components（metal ions,total nitrogen and total phosphorus）were 6.28%,5.59%,1.81%,0.40%,7.31%and 7.67%,respectively.The above results show that the model has good extrapolated prediction ability for predicting the synergistic effect of wastewater components.