Within the market of fine cocoa (Theobroma cacao L.), quality and prices are mainly determined by geographical origin, making traceability indispensable. However, a solid scientific method for this purpose has not been developed yet.
Material and Methods
Various cocoa samples from 20 countries have been profiled using, for the first time for cacao, a combination of multielement isotope-ratio mass spectrometry (IRMS) and nuclear magnetic resonance (NMR).
Dimensionality reduction of NMR data resulted in six sets of signals, which could be assigned to substances well-suited for discrimination purposes. Principal component analysis (PCA) as well as cross-validated partial least squares discriminant analysis (PLS-DA) of combined isotope data (δ13C, δ15N, δ18O, δ2H, %C, %N, %O, %H) and 1H-NMR fingerprints achieved improved separation with increased classification rates compared to classification with data of the isolated methods. Loading plots revealed complementary properties of the two analytical techniques, as IRMS contributed primarily to discrimination between countries, while NMR significantly contributed to the separation of regions and varieties.
Combination of datasets from different analytical methods each providing a larger number of parameters, is proposed as a general tool to enhance both, accuracy and precision, in testing of foodstuffs for authenticity.