Perovskite-based ternary nanocomposite for advanced ascorbic acid sensing assisted by machine learning
Oumayma Bouhajeb, Ramzi Hammami, Adel Megriche, Soner Cakar, Mahmut Ozacar, Emna Bouhajeb
article (Uluslararası, Hakemli) | Yazar(lar): Oumayma Bouhajeb, Ramzi Hammami, Adel Megriche, Soner Cakar, Mahmut Ozacar, Emna Bouhajeb | Yıl: 2026 | Dergi/Kurum: FOOD BIOSCIENCE | İndeks: WOS.SCI | DOI: 10.1016/j.fbio.2026.109178 | Anahtar Kelimeler: Ascorbic acid, Perovskites, Electrochemical sensors, Machine learnin
This study reports the rational design and sonochemical synthesis of a novel ternary LaMO3 (M = Mn, Fe, Co)/ MAPbCl3/g-C3N4 nanocomposites for the electrochemical sensing of ascorbic acid (AA). Comprehensive analyses via XRD, FTIR, FE-SEM, EDS, TEM, HRTEM, PL, UV-DRS, EIS, and Mott-Schottky techniques confirmed the successful formation of well-defined heterostructures with strong interfacial coupling and synergistic functionality. Glassy carbon electrodes (GCEs) modified with the nanocomposites enabled ultrasensitive electrochemical detection of AA in 0.1 M phosphate buffer (pH 7.4). Among the prepared electrodes, LaCoO3/MAPbCl3/g-C3N4 exhibited the lowest band gap energy (1.63 eV), narrowest PL intensity, lower RCT, lower limit of detection (0.023 mu M) and the highest sensitivity, outperforming the Fe- and Mn-based ternary analogues. Integration of the optimized LaCoO3/MAPbCl3/g-C3N4 sensor with an XGBoost-based machine learning model enabled accurate quantification of AA in commercial vitamin C tablets. Furthermore, the electrode exhibited excellent repeatability and long-term cyclic stability, emphasizing its potential as a robust, cost-efficient candidate for nextgeneration electrochemical sensing applications.