Visible light-driven copper vanadate/biochar nanocomposite for heterogeneous photocatalysis degradation of tetracycline: Performance, mechanism, and application of machine learning

Environ Res. 2024 Dec 31:267:120747. doi: 10.1016/j.envres.2024.120747. Online ahead of print.

Abstract

Water pollution caused by antibiotics is considered a major and growing issue. To address this challenge, high-performance copper vanadate-based biochar (CuVO/BC) nanocomposite photocatalysts were prepared to develop an efficient visible light-driven photocatalytic system for the remediation of tetracycline (TC) contaminated water. The effects of photocatalyst mass, solution pH, pollutant concentration, and common anions on the TC degradation were investigated in detail. Analytical techniques indicated that the CuVO exhibited a nanobelt-like structure with a uniform distribution on the wrinkled biochar surface. The XRD spectrum confirmed that the as-prepared nanomaterial was composed of Cu3V2O7(OH)2·2H2O. Meanwhile, XPS analysis revealed that copper was present in two forms: monovalent and divalent, while vanadium remained pentavalent. The CuVO/BC exhibited excellent stability and high visible light photocatalytic activity towards TC degradation over a wide pH range. The presence of SO42-, H2PO4-, CO32-, and citric acid inhibited the degradation process due to the consuming of photogenerated h+ and •OH, while Cl- enhanced the efficiency of photocatalytic reactions due to generating chlorine oxidizing species. The CuVO/BC showed lower electron-hole recombination rate, more effective separation of photogenerated carriers, lower charge transfer resistance, and higher visible light absorption capacity comparing to pure CuVO by the addition of BC, thus improving the overall photocatalytic performance. In terms of oxidation mechanism, the EPR test and quenching experiment revealed that the contribution of the active species to the degradation of TC followed the order h+ > 1O2 > •OH > •O2-. Through the application of machine learning models to analyse the influencing factors of photocatalytic processes, it was discovered that the GBDT model exhibited optimal reliability for the photocatalytic system, and the simulation results were in agreement with the experimental findings.

Keywords: Biochar; Copper vanadate; Machine learning models; Tetracycline; Visible-light photocatalysis.