Monitoring wine fermentation using ATR-MIR spectroscopy and chemometric techniques.

Identifier:  TDX:3229

Handle: https://hdl.handle.net/20.500.11797/TDX3229

Authors:  Cavaglia Pietro, Julieta

Abstract:
Wine is one of the most appreciated high added-value products in the world and therefore, controlling wine production has always been a priority for most wineries. Implementing at-line analyses such as Process Analytical Technologies (PAT) guidelines, not only enables a control of the final wine but also gives the possibility to apply correcting measures throughout the process, thus avoiding a defective final product. In this doctoral thesis, we investigated the possibility of implementing different strategies to control and detect deviations during wine alcoholic fermentation using a fast and portable equipment: an Attenuated Total Reflectance Mid-Infrared (ATR-MIR) spectrometer which allows obtaining, in a few seconds, a large amount of information about the fermentation process, which was processed with different chemometric techniques. First, using the spectral data and Partial Least Square Regression, different chemical parameters were predicted during alcoholic fermentation. Secondly, we compared the spectra from both Normal Operation Conditions and deviated fermentations using Partial Least Squares Discriminant Analysis. ANOVA–simultaneous component analysis was applied to study the influence of several factors into the variance of the spectra. Multivariate Curve Resolution Alternating Least Squares was used to model both alcoholic and malolactic fermentations. Finally, a PAT methodology