Classification of Chardonnay grapes according to geographical indication and quality grade using attenuated total reflectance mid-infrared spectroscopy

Abstract

Rapid analytical methods based on infrared spectroscopy in combination with chemometrics have found wide application in the food and beverage industry. These methods have the potential to qualitatively analyse and classify or authenticate samples including grapes and wines, or be used as a tool for objective decision-making while grapes are still ripening, ultimately offering better control over the winemaking process. Thus, an initial investigation examined the use of attenuated total reflectance (ATR) mid-infrared (MIR) spectroscopy to discriminate Chardonnay grape samples from different geographical origins and industry-allocated quality grades with minimal sample preparation. Classification of samples according to region of origin using partial least squares discriminant analysis (PLS-DA) of the fingerprint region of the MIR spectra (1500–800 cm⁻¹) had an overall success rate of 83 and 81% for the 2014 and 2016 vintages, respectively. It was also possible to classify sample quality successfully using this same approach. Correct classification of Chardonnay grapes according to quality grade was of the order of 83% in 2014 and 79% in 2016. The ability to predict juice titratable acidity and total soluble solids was also shown. We have demonstrated the potential use of ATR-MIR as a rapid tool to classify samples according to geographical origins and quality grades, which has implications for authenticity determination and for optimising the streaming of fruit to the most appropriate winemaking processes.