Feasibility of using NIR spectroscopy to detect herbicide residues in intact olives

This study sought to assess the feasibility of using Near-Infrared Spectroscopy (NIRS) technology to detect levels of diuron herbicide in intact olives. The experimental set comprised 216 olive samples fortified with varying concentrations of diuron at values above and below the MRL (Maximum Residue Limit). A diuron-free sample set was used as control. A range of strategies was tested based on NIRS data for the construction of discriminant models using the Partial Least Squares-Discriminant Analysis (PLS-DA). The best model developed correctly classified 85.9% of samples used in the validation set by olives with a content of diuron above and below the MRL, applying a second order derivative with Multiplicative Scatter Correction (MSC). It can be concluded that the proposed reflectance NIR spectroscopy is appropriate for the quality control of herbicide diuron in intact olives.