Use of SNIF-NMR and IRMS in combination with chemometric methods for the determination of chaptalisation and geographical origin of wines (the example of Slovenian wines)

Citation
Ij. Kosir et al., Use of SNIF-NMR and IRMS in combination with chemometric methods for the determination of chaptalisation and geographical origin of wines (the example of Slovenian wines), ANALYT CHIM, 429(2), 2001, pp. 195-206
Citations number
21
Categorie Soggetti
Spectroscopy /Instrumentation/Analytical Sciences
Journal title
ANALYTICA CHIMICA ACTA
ISSN journal
00032670 → ACNP
Volume
429
Issue
2
Year of publication
2001
Pages
195 - 206
Database
ISI
SICI code
0003-2670(20010223)429:2<195:UOSAII>2.0.ZU;2-T
Abstract
Isotopic ratios determined by SNIF-NMR and IRMS methods were used in combin ation with chemometric methods PCA, KANN and cluster analysis for the detec tion of wine chaptalisation and differentiation of wines according to the g eographical origin. Selected wines from the three wine-growing regions of S lovenia were chosen for the analysis since Slovenia, small by surface, but pedologically and climatically diverse, offers interesting possibilities fo r studying the influence of this diversity upon the isotopic ratios in mole cules of natural products. We found that the separation of wines of the Coa stal region from the wines of the continental regions Sava and Drava is ver y good. On the other hand, the separation between wines from the Sava and t he Drava region is not satisfactory. This is not surprising since the clima tic differences between these two regions are small when compared with the differences between the continental and Coastal regions. Regarding the chap talisation, the considerable improvement by combining both SNIF-NMR and IRM S is obtained. PCA and KANN give comparable separation of wines according t o the geographical origin, but PCA is less time consuming. In the case of e nrichment of wines from the Drava region by beet sugar better separation be tween natural and enriched wines is obtained using KANN method. In all case s cluster analysis shows poorer efficiency than PCA and KANN. (C) 2001 Else vier Science B.V. All rights reserved.