Abstract
This study applied differential scanning calorimetry (DSC) coupled with chemometric analysis, specifically, Principal Component Analysis (PCA) and Partial Least Square Regression (PLSR), for the detection of fat adulteration in butter. Adulteration was simulated by adding varying concentrations (2–30%, w/w) of palm stearin and coconut oil to butter. Thermograms acquired from DSC were subjected to chemometric analysis to detect alterations in the butter melting pattern. The results showed that DSC is a highly sensitive technique for detecting even small changes in the butter melting pattern. Discriminant analysis performed using K-Nearest Neighbors (kNN) on 11 distinct butter samples, adulterated with palm and coconut oils at concentration of 10, 20 and 30% (w/w), achieved an accuracy rate higher than 92.1 % in differentiating authentic from adulterated samples. Hierarchical cluster analysis (HCA) enabled the discrimination of the type of adulterant—palm stearin versus coconut oil—at concentrations exceeding 5% (w/w). Compared to traditional methods, DSC coupled with chemometric analysis presents a simple yet effective tool for screening adulterated butter samples, thereby offering potential applications in quality control within the food industry.