Authentication of oats and discrimination from their gluten-containing adulterants using NIR diffuse reflectance spectroscopy and multivariate analysis

Abstract

Oat (Avena sativa L.) is recognized for its nutritional value and gluten-free status, however, ensuring its authenticity and purity is crucial, given the potential for contamination with gluten-containing grains. In this study, near-infrared spectroscopy coupled with chemometrics were employed to authenticate oat flour (from different forms of oat; oat groats, steel-cut and rolled oats) and distinguish it from common gluten-containing adulterants including wheat, farro, triticale, barley, rye, and ryegrass. Both unsupervised and supervised chemometric methodologies, encompassing PCA, SIMCA, and OPLS-DA, were applied. Both SIMCA and OPLS-DA models displayed 100% sensitivity, enabling reliable identification of oat flour and detection of potential adulteration with these gluten-containing grains with specificity of 97.78% and 100%, respectively. In the SIMCA model for oat groats and adulterated mixtures, samples with 1% and 2% adulteration were incorrectly classified as oat groats, yet successful discrimination from deliberately-adulterated mixtures was accomplished through the OPLS-DA model with 100% specificity. Moreover, PLS regression analysis was employed to precisely quantify the levels of adulterants in oat groats flour. The models exhibited reliable performance, as reflected in Root Mean Square Error of Calibration (RMSEC) values. Validation of these models using test samples provided further confirmation of their efficacy in detecting sample adulteration and ensuring product authenticity, all while maintaining a high sample throughput rate.