Machine Learning Assisted Imprinted Ag@PANI/CoFe₂O₄/C Heterojunction with Simultaneous Improvement of Selectivity and Activity for Antibiotic Photodegradation

Improvement of selectivity and activity of imprinted photocatalysis is a major challenge for antibiotic photodegradation due to the functional monomers of imprinting hindering the photogenerated carrier migration. Here, an organic imprinted Ag-polyaniline/CoFe₂O₄/Carbon photocatalyst (IM-Ag-PANI/CoFe₂O₄/C) was successfully prepared by photo-initiated polymerization which achieved selective photodegradation of tetracycline (TC). The heterojunction formed by the functional monomer Ag@PANI and CoFe₂O₄/C not only facilitates the separation of photo-excited carriers and the exposure of active sites but also contributes to the selective adsorption capability by imprinted cavity on Ag@PANI, thereby improving the photocatalytic activity and selectivity simultaneously. In addition, the Corncob conversion carbon matrix method and iron-based magnetic character enable a more environmentally friendly and recyclable capability of IM-Ag-PANI/CoFe₂O₄/C. After using machine learning models to train and predict experimental parameters by changing experimental parameters, the IM-Ag-PANI/CoFe₂O₄/C can photodegrade 82.23% of TC within 2 h, and it has a selective degradation ability compared to enrofloxacin hydrochloride (EH). This research provides a new idea for the construction of imprinted photocatalytic materials that can improve photocatalytic activity.