ASTFS-based multivariate analysis for discrimination and simultaneous quantification of multiple adulterants in cow...

Ghee is a high-value dairy fat vulnerable to economically motivated adulteration. Despite its strong intrinsic fluorescence, fluorescence-based authentication remains underexplored. This study presents angular sweep total fluorescence spectroscopy (ASTFS) combined with multivariate analysis to discriminate and quantify common adulterants: vanaspati, beef tallow, and pork lard in cow ghee. After orthogonal signal correction and Pareto scaling, Partial Least Squares- Discriminant Analysis (PLS-DA) achieved complete discrimination between pure and adulterated samples, with 100% accuracy, sensitivity, and specificity. Beyond single-adulterant detection, robust PLS regression models were developed for dual-adulterant systems (vanaspati-beef tallow and vanaspati-pork lard). Calibration included home-made ghee and five commercial brands to capture matrix variability. The models showed high predictive performance (R² > 0.95), low RMSEC ≤1.37% (v/v) and RMSEP ≤1.49% (v/v). External validation confirmed strong generalizability (RMSEP <1.6% v/v). Overall, ASTFS with PLS-based modelling effectively captures matrix variability and quantifies multi-adulterant ghee mixtures.