Researchers develop machine learning models to predict antibiotic degradation using TiO2/ZnO nanocomposites

Researchers developed machine learning models to predict the photocatalytic degradation of the antibiotic metronidazole using a synthesized TiO2/ZnO nanocomposite. Under optimal conditions, the nanocomposite degraded 94.92% of the antibiotic. Among the evaluated models, support vector regression achieved the highest predictive accuracy. Feature importance analysis showed that reaction time and pH were the most influential parameters.

TiO2/ZnO nanocomposite photocatalyst synthesis

• ▪Researchers synthesized a TiO2/ZnO nanocomposite as a photocatalyst using the sol-gel method
• ▪The synthesized TiO2/ZnO nanocomposite was characterized using EDX, TEM, FTIR, SEM, and XRD

Antibiotic pollution bacterial resistance

• ▪Inappropriate disposal of antibiotics in aquatic and soil environments makes bacteria resistant, posing a potential threat to humans and other organisms
• ▪Photocatalysis is considered an attractive option for degrading antibiotics because it is a simple, inexpensive, and eco-friendly process

Metronidazole photocatalytic degradation optimization

• ▪Under optimized conditions, the synthesized TiO2/ZnO nanocomposite is capable of degrading metronidazole by 94.92%
• ▪The effect of pH, irradiation time, metronidazole concentration, and catalyst dose on the photodegradation of metronidazole was optimized using response surface methodology based on central composite design

Machine learning prediction models

• ▪Researchers employed machine learning and deep learning models, guided by response surface methodology, to predict the photocatalytic degradation efficiency of metronidazole
• ▪The evaluated models for predicting metronidazole degradation efficiency included support vector regression, artificial neural network, fully connected neural network, and random forest
• ▪Among the evaluated models, support vector regression demonstrated the highest predictive accuracy, achieving an R-value of 0.9495 and an R2 of 0.8502

Reaction time pH influence

• ▪Feature importance analysis revealed that reaction time and pH were the most influential parameters in predicting metronidazole degradation efficiency
• ▪Feature importance analysis showed that catalyst dose had minimal impact on the prediction of metronidazole degradation efficiency compared to reaction time, pH, and metronidazole concentration