Application of Flash GC and linear and non-linear chemometric techniques as a rapid screening tool to discriminate monovarietal Grappa distillates

Grappa is a traditional Italian grape marc spirit protected by geographical indication. Monovarietal Grappa, from a single grape variety, exhibits distinctive compositional and sensory characteristics linked to the selected raw materials, including denominations such as Grappa di Barolo. Rapid screening methods are essential for product authentication. This study analyzed the volatile fingerprints of 45 Grappa samples from four grape varieties (Muscat, Prosecco, Barolo, Brunello) using Flash GC. Multivariate linear (PCA and PLS-DA) and non-linear (C-SVM) techniques were applied to different data representations including full chromatographic profiles, selected chromatographic regions and peak-area data to determine which representation offered the highest discriminant power. Comparable classification results were obtained using both PLS-DA and SVM models, reporting overall accuracy ranging from 77.8 to 84.1% in cross-validation. Tentative identification of discriminating compounds was also performed. The results suggest the potential of Flash GC and chemometrics as a rapid screening tool to discriminate monovarietal Grappa.