The capacity of spectrofluorometric fingerprints to discern changes of wine composition: applications in classifying wine additives and tracking red wine maturation and ageing

Abstract

Fluorescence spectroscopy combined with chemometrics has shown advantages in wine analysis due to being rapid, sensitive, and selective to fluorescent molecules. Especially due to the abundant phenolic compounds [1], the molecular fingerprints afforded by fluorescence spectroscopy can potentially be used to discern and track the change of wine composition, with two innovative investigations having been implemented.

Processes and additives can be used during winemaking for different purposes, with the overall goal of controlling composition and quality. Regulatory frameworks are imposed but analytical methods are needed to determine aspects such as permitted use of additives, abuse of illegal processes or adulteration, and false label claims [2]. Fluorescence spectroscopy combined with chemometrics is being tested for this purpose, with an investigation of additives and processes. Initial results reveal high classification accuracies of Chardonnay wine treated with nine additives (96.4%) at three levels (95.5%) based on spectrofluorometric datasets and machine learning modelling.

Maturation and bottle ageing are crucial dynamic stages associated with the quality and commercial value of red wine. The capacity of fluorescence spectroscopy to track wines through production [3] or changes in red wine compositions during fermentation [4] have been studied. As a further innovation, the molecular fingerprints afforded by fluorescence spectroscopy have been investigated for monitoring long-term ageing and tracking the dynamic change of red wine to enable the prediction of red wine evolution. To this end, a number of monovarietal red wines have been periodically analysed by fluorescence spectroscopy during maturation and various bottled wines are being regularly studied. Changes in colour, phenolic composition, and sensory properties are also being determined and used for machine learning modelling with the spectrofluorometric datasets to provide understanding that may guide decisions during winemaking and storage.

References

[1] Airado-Rodríguez, D., Durán-Merás, I., Galeano-Díaz, T., Wold, J. P. (2011). J. Food Compos. Anal., 24(2), 257-264.

[2] Ranaweera, R. K. R., Capone, D. L., Bastian, S. E. P., Cozzolino, D. and Jeffery, D. W. (2021). Molecules, 26(14), 4334.

[3] Ranaweera, R. K. R., Gilmore, A. M., Bastian, S. E. P., Capone, D. L., Jeffery, D.W. (2022). OENO One, 56(1), 189-196.

[4] dos Santos, I., Bosman, G., du Toit,W., Aleixandre-Tudo, J. L. (2023). Food Control, 147, 109616.