Microplastic-regulated MOFs-derived Co/C nanocomposites for enhanced peroxymonosulfate activation in tetracycline degradation: catalytic performance, machine learning prediction, and mechanistic insights

This work proposes an innovative method for recycling microplastics, specifically polystyrene microplastics, to fabricate metal-organic framework (MOF)-derived catalysts. A MOF-derived Co/C nanocomposite (PS@ZIFC) was prepared by carbonization of ZIF-67, regulated by the incorporation of microplastics. PS@ZIFC shows a core-shell structure and large surface area (223.75 m 2 /g), and contains carbon nanotubes (CNTs) and cobalt nanoparticles. The catalytic performance of PS@ZIFC in peroxymonosulfate (PMS) activation for tetracycline degradation was thoroughly evaluated. The PS@ZIFC/PMS system achieves high effectiveness in a short treatment time (30 min) with a high rate constant of 0.1083 min -1 . The system is effective in a wide pH range of 3.0 to 9.0 and has good recycling performance. Notably, the degradation capacity is 1129.2 mg tetracycline per gram of PS@ZIFC, surpassing that of previously reported catalysts. Machine learning prediction reveals that the extreme gradient boosting model is highly trustworthy for predicting tetracycline degradation compared to the other two models (random forest model and support vector regression model). Regarding the catalytic mechanism, both radical pathways (driven by hydroxyl radicals (·OH) and sulfate radicals (SO 4 · - )) and non-radical pathways (singlet oxygen and electron transfer) are involved in tetracycline degradation. The mechanism of tetracycline degradation, deduced from the identified degradation intermediates, was confirmed by liquid chromatography-mass spectrometry (LC-MS). The toxicity evaluation revealed a significant decline in the toxicity of degradation intermediates. • MOF-derived Co/C nanocomposite catalyst regulated by addition of microplastic • Improvement of MOF stability for routine applications • PS@ZIFC/PMS system achieved high degradation capacity of 1129.2 mg/g • Machine learning predicts tetracycline degradation by extreme gradient boosting model • Toxicity evaluation reveals significant decline in the toxicity of degradation intermediates