Enhancing wine traceability through data fusion strategies combining ¹H-NMR and IR analysis

Abstract

Wine authentication is a major challenge in the context of an expanding global market and increasing fraudulent practices. Protecting designations of origin is essential to sustain the wine sector and maintain consumer confidence. Among the available analytical tools in the wine industry, infrared (IR) spectroscopy is widespread, while ¹H-NMR spectroscopy has demonstrated its effectiveness to assess wine authenticity. To enhance the discriminative capacity of analytical classification, a data fusion strategy combining ¹H-NMR and IR spectroscopic data was implemented. This approach leverages the complementary strengths of both techniques. IR spectroscopy provides information related to hydrogen bonding and asymmetric polar groups, whereas ¹H-NMR spectroscopy offers highly resolved insights into specific proton environments.

The complementary spectral features obtained from these methods enable the construction of fused datasets containing richer and more comprehensive chemical information. By exploiting this complementarity, data fusion enhances classification robustness and reliability. This strategy has been applied to discriminate 44 wines according to grape variety, geographical origin, and vintage. The results demonstrate the effectiveness of combining data from multiple analytical techniques to enhance wine traceability and authenticity assessment.