Brought to you by Wiley
Login | Register
Ezine News Education Links
Webinars Podcasts Books & Journals Jobs Products Conferences Classifieds
Predictable wine Predictable wine
[October 1, 2007]
AddThis Social Bookmark Button

A simple tool for winemakers and researchers based on UV-vis spectroscopy allows them to predict levels of phenolic compounds, including anthocyanins, tannins, polymeric pigments, and iron-reactive phenols, in real-time and could lead to improved wine production methods as well as opening up new lines of research in enology. All the winery or researcher needs is access to a spectrometer interfaced to Microsoft's Excel spreadsheet software and access to the model available on-line.

Phenolic compounds play a critical role in the sensory properties of wine, including colour, how they feel in the mouth, and overall perceived quality, as well as playing a part in how a wine "ages". As such, a method for monitoring the extraction of these compounds during fermentation could allow wine makers to make better process decisions in crafting wines with phenolic profiles appropriate for the desired wine style. Currently, colorimetric and chromatographic techniques are used by wineries to quantify colour and tannin extraction, but these are time consuming in terms of sample preparation and analysis and expensive to contract out to a service laboratory.

Previous work in the development of predictive methods to quantitate phenolic compounds in wine focused on vis-NIR spectroscopy, however, there were limitations in these models with regard to tannin quantitation. Kirsten Skogerson, Roger Boulton, and colleagues in the Department of Viticulture & Enology at the University of California Davis have turned to UV-vis spectroscopy to develop a predictive model of phenolic content in wine. The model is based on values measured using the Harbertson-Adams assay, a comprehensive, robust assay developed in the Viticulture & Enology department at UCD.

The team developed the predictive function for each phenolic compound class using multivariate methods, and then used the models to estimate the corresponding phenolic concentrations for 200 independent wine samples. The researchers obtained successful correlations and Skogerson adds that, "The method has the potential for the rapid determination of these colour and phenol components during the fermentation of red wines."

The model was most adept at predicting anthocyanin, total phenolic, and tannin levels, and although non-tannin phenols were also well predicted, the concentrations of polymeric pigments was not quite tenable. However, the team reports that same approach is just as applicable to single-cultivar wines, Shiraz and Cabernet Sauvignon were tested, wine juices and musts as it is to finished wines.
One large winery in Washington State is already using our technique, and two large wineries in Australia trialled it in the 2007 harvest," Skogerson told SpectroscopyNOW, "I think we're only about two more years away from widespread adoption."

Related links:

Article by David Bradley

Click here for ezine index Click here for news index Click here for education index

Roger Boulton

Boulton, testing times for wine

Wine corks (Photo by David Bradley)
The phenolic content of wine determined by UV-vis before any corks are pulled