Usefulness of Fluorescence Excitation−Emission Matrices in Combination with PARAFAC, as Fingerprints of Red Wines

Abstract

The possibility of using front-face fluorescence spectroscopy to characterize red wines was investigated, and a tentative identification of their main fluorescent components was attempted. Fifty-seven red wine samples from different origins were included in the present study. Their fluorescence excitation-emission matrices (EEMs) were registered directly on 3-mL aliquots of untreated samples. The assayed excitation and emission ranges were 245-340 and 300-500 nm, respectively. The set of 57 EEMs was analyzed by means of parallel factor analysis (PARAFAC). Thus, the spectral excitation and emission profiles of possible "pure" fluorescence components (PARAFAC loadings) and the relative contribution of each component to the individual EEMs (PARAFAC scores) were obtained. The red wine system contained four main fluorescence components, and the excitation and emission loadings had maxima at the wavelength pairs 260/380, 275/323, 330/410, and 280/364 nm, respectively. A tentative identification of fluorophores was done by matching PARAFAC score values with HPLC measurements on the same 57 samples, as well as fluorescence measurements on pure compounds typically present in red wine. It was found that the third component was highly correlated with concentrations of catechin and epicatechin. When the PARAFAC score values were plotted against each other, they did to some extent discriminate the wines according to origin (country) and grape variety.

Modeling excitation–emission fluorescence matrices with pattern recognition algorithms for classification of Argentine white wines according grape variety

“…For example, Dufour et al, employed both excitation and emission spectra for French and German wine classification (Dufour, Letort, Laguet, Lebecque, & Serra, 2006), and Sádecká et al reported the classification of brandies and wine distillates using total luminescence and synchronous fluorescence spectra (Sádecká, Tóthová, & Májek, 2009), but only first-order data and algorithms were used in an unsupervised manner in both cases. On the other hand, Ariado-Rodrigues and collaborators employed EEM and PARAFAC for fingerprinting of red wines (Airado-Rodríguez, Galeano-Díaz, Durán-Merás, & Wold, 2009) and quality control in the wine industry (Airado-Rodríguez, Durán-Merás, Galeano-Díaz, & Wold, 2011). Interestingly, and different from the present report, an unsupervised approch was used in both reports.…”

Synchronous front-face fluorescence spectroscopy as a promising tool for the rapid determination of spoilage bacteria on chicken breast fillet

“…Any data set that can be modeled adequately with PARAFAC can be thus also be modeled by Tucker-3 or two-way PCA. PARAFAC has already been used in many food systems including meat (Moller, Parolari, Gabba, Christensen, & Skibsted, 2003), fish oil (Pedersen, Munck, & Engelsen, 2002), milk (Boubellouta & Dufour, 2008), cheese (Christensen, Povlsen, & Sorensen, 2003), sugar (Bro, 1999), beer (Sikorska et al, 2008), red wine (Airado-Rodriguez, Galeano-Diaz, Duran-Meras, & Wold, 2009) and oils (Dupuy et al, 2005;Guimet, Ferré, & Boqué, 2005a;Guimet, Ferré, Boqué, & Rius, 2004;Guimet, Ferre, Boque, Vidal, & Garcia, 2005b;Sikorska et al, 2004).…”

Calibration, standardization, and quantitative analysis of multidimensional fluorescence (MDF) measurements on complex mixtures (IUPAC Technical Report)

“…This approach has been implemented successfully for the quantitative analysis of various complex biogenic samples from, e.g. wine [50], milk [40,51], food [5,52,53], and cell culture media [4,35,54]. In many of these applications, there are generally multiple fluorophores present, often at high concentration.…”