Geographical classification of Chinese Cabernet Sauvignon wines by data fusion of ultraviolet-visible and synchronous fluorescence spectroscopies: the combined use of multiple wavelength differences

Background and Aims: We report the chemometric classification of commercial Chinese Cabernet Sauvignon dry red wines according to the winegrowing region based on the coupling of synchronous fluorescence ( SF) and ultraviolet ( UV)-visible ( VIS) spectroscopies. Methods and Results: Total SF spectra at a wavelength difference (Delta lambda) of 10-150 nm with 10-nm intervals, UV spectra in the 240 to 380 nm range and VIS spectra in the 380- to 700-nm range were recorded. A simple classification model was constructed by principal component analysis and linear discriminant analysis and was evaluated by external validation. The combined use of SF and UV-VIS spectroscopies offered clearer classification than the individual techniques. The Delta lambda in SF spectroscopy was revealed to play an important role in the classification. The combination of SF spectroscopy at Delta lambda 30, 60, 90 and 120nm and UV-VIS spectroscopy achieved the highest prediction rate ( 79.2%). Conclusions: The fusion of data from UV-VIS and SF spectroscopies at multiple Delta lambda with certain pitched intervals, which has retained more complementary and useful information, can mostly improve the classification of wines. Significance of the Study: The joint use of the two complementary spectroscopies provides an alternative to discriminate wines from different geographical origins.