Estudio del estado de madurez y la aptitud enológica en uva mediante análisis de imagen hiperespectral

Résumé

Wine quality mainly depends on the characteristics of harvested grapes. Traditionally, variety control, grape maturity, phenolic composition or phenolic compounds extractability are determined by physical and chemical analyses which are usually time consuming. During the past several decades, spectroscopic techniques have grown significantly as technology improved. This growth has allowed developing fast, non-destructive and green chemistry methods for the screening of different parameters of value for agriculture. Among them, hyperspectral imaging methods provide special and spectral information of the sample. In this work, near infrared hyperspectral imaging has been applied for the spectral characterisation of grapes. Afterwards, this spectral data has been used for the development of different methods for the evaluation of grape maturity and oenological capability. The ability of near infrared hyperspectral imaging for discriminating among different grape varieties has been checked. This technique shows similar results than traditional techniques, such as chromatographic analysis, for this purpose. Furthermore, hyperspectral imaging has been applied to grapes in order to develop different calibration models for the screening of several parameters linked to wine sector (i.e., sugar concentration, titratable acidity and pH of grape must, and total phenolic and anthocyanic compounds in grape skin). Moreover, characteristic vector analysis has been applied in order to reduce the amount of spectral data which is necessary for the calibration model development. In addition, calibration models have been developed for the screening ofextractable contentof anthocyanins, flavanols and total phenolic compounds ingrape skins. These models allow knowing how easily phenolic compounds are released from grape seed and skin to wine. In order to provide more information in this respect, Fourier transform infrared and Raman spectroscopy have been applied to grape seed and skin non–extracted material (solid residue obtained after exhaustive phenolic extraction). The main spectral features have been linked to phenolic extractability and it has been confirmed that grape seed and skin phenolic extractability is closely linked to the cell wall composition. Finally, the effect that copigments coming from white grape seeds has on the anthocyanin extraction from red grape skins has been evaluated. Grape seed contains copigments that can improve or stabilize wine colour, however, they might hamper the extraction equilibrium of anthocyanins compounds. In this work it has been proven that the addition of copigments coming from white grape seeds has not reduce the amount of extracted anthocyanins from red grape skin.