Non-targeted detection of butter adulteration using pointwise UHPLC-ELSD and UHPLC-UV fingerprints with chemometrics

A non-targeted chemometric method was devised to detect possible butter adulteration without prior knowledge of the adulterant and marker compounds. Nine common edible oils including vegetable oils, animal fats and margarines were selected as potential adulterants to build a unified classification model. The samples were analyzed using the high-performance liquid chromatography hyphenated with an evaporative light scattering detector (UHPLC-ELSD) and an ultraviolet detector (UHPLC-UV), with the pointwise chromatograms instead of individual peaks for modelling. Both models achieved over 95% correct classification in external validation at the adulteration levels as low as 5% (w/w). The root mean squared errors of prediction (RMSEP) of the regression model were 0.9865 and 1.9080 for UHPLC-ELSD and UHPLC-UV, respectively. Non-targeted chemometrics analyses based on pointwise chromatographic profiles could be valuable for detecting adulterated butter.