Application of benchtop total-reflection X-ray fluorescence spectrometry and chemometrics in classification of origin and type of Croatian wines

• Published in: Food Chemistry: X Vol. 13; p. 100209
• Main Authors: Vitali Čepo, D., Karoglan, M., Borgese, L., Depero, L.E., Marguí, E., Jablan, J.
• Format: Journal Article
• Language: English
• Published: Netherlands Elsevier Ltd 30.03.2022; Elsevier
• Subjects: Elemental composition; Multivariate characterization; TXRF; Wine; Wine; Elemental composition; Multivariate characterization; TXRF
• ISSN: 2590-1575; 2590-1575
• DOI: 10.1016/j.fochx.2022.100209

•A low-power benchtop TXRF system was used for analysis of 70 wine samples.•The metal content of K, Ca, Fe, Cu, Zn, Mn, Sr, Rb, Ba, Pb, Ni, Cr and V was estimated by chemometric methods.•Mn, K, Ni, Sr, Rb and Ba were the main variables used to differentiate by wine type and origin.•LDA showed good detection and prediction abilities with selected elements.•Classification of origin and type of Croatian wines by chemometric tools. The contents of selected metals (K, Ca, Fe, Cu, Zn, Mn, Sr, Rb, Ba, Pb, Ni, Cr and V) in 70 wine samples from Continental and Adriatic part of Croatia and different types of wine (red and white) were determined by TXRF. The aim of this study was to compare the elemental composition of wines from two different regions and to determine the discriminant ability of each variable and to indicate which variables discriminate between the four categories considered. Principal component analysis and cluster analysis showed that K, Mn, Ba and Ni can be considered as the most important characteristics to distinguish between Continental red and white wines, Rb, Ni and Ba for Continental red and Adriatic red wines while Sr is the only metal that completely distinguishes the samples of each category. Finally, linear discriminant analysis showed good recognition (100%) and prediction abilities (96.43%) using these selected elements.