Au octahedrons monolayer film SERS substrate coupled with a hybrid metaheuristic algorithm-optimized ELM model: An analytical strategy for rapid and label-free detection of zearalenone in corn oil

Abstract

This study developed a rapid, label-free analytical strategy for quantifying zearalenone (ZEN) in corn oil. A highly sensitive Au octahedrons (Ohs) monolayer film was synthesized as the surface-enhanced Raman spectroscopy (SERS) substrate. A hybrid metaheuristic algorithm that combines the particle swarm optimization (PSO) algorithm and the grey wolf optimizer (GWO) algorithms, was used to optimize an extreme learning machine (ELM) model (i.e., the PSOGWO-ELM model). The PSOGWO-ELM model analyzed the collected SERS spectra to determine ZEN contents in corn oil. The results demonstrated that the analytical strategy possessed excellent performance: the root mean squared error of the prediction set (RMSEP) = 0.2297 μg/mL, the coefficient of determination of the prediction set ($R_P^2$) = 0.9907, and the ratio of performance to deviation of the prediction set (RPD_P) = 10.3695. The proposed analytical approach shows considerable promise for the rapid, label-free, and accurate detection of trace levels of ZEN in corn oil.