Surface-enhanced Raman Scattering Biosensor by Combining Rare Earth Elements: Integration with Machine Learning Tools for Diosgenin and Tectorigenin Detection and Classification
摘要
In this study, we developed a high-performance biosensor by combining rare earth elements, specifically europium (Eu) with cicada wings (C.w.), and Ag nanoparticles (NPs) for surface-enhanced Raman scattering (SERS). Using 4-aminothiophenol (4-ATP) as the probe molecule, the biosensor achieved an exceptionally low detection limit (LOD) of 10−10 M. Characteristic peaks for two pharmacodynamic substances were identified at concentrations down to 10−3 M. The experimental enhancement factor (EEF) was calculated to be 1.18 × 106. Among four machine learning classifiers, the decision tree (DT) demonstrated the highest accuracy in classifying diosgenin and tectorigenin, reaching 90.91%.