Qiang Li, Maoyuan Hu, Jianlong Li, Jinmi Zhang, Zhiqiang Jiang, Xiaoli Xiong
Rutin is widely employed as a therapeutic agent, but excessive intake may induce adverse reactions such as gastric discomfort, headache and dermatitis. However, reliably quantifying it in complex matrices remains a significant challenge. Herein, a high-performance electrochemical sensing platform is constructed based on machine learning and interfacial energy-barrier modulation of Ti 3 C 2 /CuCoSe 2 p-n heterojunction. The heterojunction offers a large specific surface area and abundant defect sites, facilitating enhanced adsorption and electrocatalytic oxidation of rutin. Meanwhile, the built-in energy barrier inhibits carrier recombination and accelerates charge transfer, thereby inducing a significant electrochemical sensing response. Density functional theory (DFT) calculations demonstrate that rutin undergoes an energetically favorable and reversible redox process at the Ti 3 C 2 /CuCoSe 2 interface, exhibiting high affinity and selectivity for rutin. When coupled with a machine learning-based CatBoost-Decision Tree (CatBoost-DT) model, the sensor achieves markedly improved precision and accuracy with a relative standard deviation of 1.23%, recovery rates ranging from 98.89% to 100.95%, and a detection limit of 0.02 μM, outperforming most traditional electrochemical sensing methods. This study provides an attractive strategy for the rapid, high-sensitive and accurate detection of rutin in complex real samples, and highlights the potential of combining heterojunction engineering with data-driven optimization for next-generation electrochemical sensing.