Regulatory-oriented discrimination and quantification of λ-cyhalothrin in virgin olive oil using near-infrared and Fourier transform infrared spectroscopy

Pesticide residue control is a critical component of regulatory surveillance in the olive oil market. In this preliminary study, a regulatory-oriented preliminary screening framework was developed for the discrimination and quantification of λ-cyhalothrin in virgin olive oil (VOO) using near-infrared (NIR) and Fourier transform infrared (FTIR) spectroscopy. To do this, 220 samples at 11 concentration levels (0.1-1.4 mg/kg oil) were prepared to cover the concentration range relevant to an EU compliance assessment example for λ-cyhalothrin in olive oils. Spectral characteristics were first analyzed to identify λ-cyhalothrin concentration-related features in both NIR and FTIR regions. Subsequently, multi-class discrimination models were constructed to categorize samples into three study-defined maximum residue level (MRL)-related screening groups, namely Below-MRL, Within-MRL, and Above-MRL, followed by regression modeling for quantitative determination. The NIR-XGBoost model achieved perfect classification performance (accuracy = 1.0) for both calibration and prediction sets. For quantitative prediction, the NIR-SVM model showed strong performance (R p 2 = 0.909, RMSEP = 0.113 mg/kg, RPD = 3.349), with stable prediction across clearly low- and high-risk ranges, although increased uncertainty was observed near the study-defined screening thresholds (0.37 and 1.11 mg/kg oil). In contrast, FTIR-based models exhibited limited generalization capability in both discrimination and quantification due to spectral overlap with dominant triglyceride absorptions. Overall, NIR spectroscopy combined with machine learning showed potential as a rapid preliminary screening tool for λ-cyhalothrin in VOO, particularly prior to confirmatory chromatographic analysis.