Applicability of spectrofluorometry and voltammetry in combination with machine learning approaches for authentication of DOCa Rioja Tempranillo wines
The main objective of the work was to develop a simple, robust and selective analytical tool that allows predicting the authenticity of Tempranillo wines from DOCa Rioja. The techniques of voltammetry and absorbance-transmission and fluorescence excitation emission matrix (A-TEEM) spectroscopy have been applied in combination with machine learning (ML) algorithms to classify red wines from DOCa Rioja according to region (Alavesa, Alta or Oriental) and category (young, crianza or reserva).
Linear sweep voltammetry (using disposable carbon paste electrodes) and A-TEEM signals of 132 Tempranillo red wines were acquired. Data were analysed following non-supervised statistical strategies such as principal component analysis (PCA) to reduce the number of variables, and two-way ANOVA (origin and category) and supervised modeling strategies derived from machine learning algorithms.
The voltammogram in the region of 691-771 mV provided clear classification of the three ageing categories and Rioja Oriental and Rioja Alavesa/Alta could be separated, but Alavesa could not be differentiated from Alta based on voltammetric signals. Results showed that A-TEEM was more efficient in classifying subareas and ageing categories of Tempranillo Rioja wines, with an ML approach using extreme gradient boosting discriminant analysis (XGBDA) providing 100% correct class assignment for subregion and wine category. A-TEEM coupled with ML algorithms is presented as a powerful and rapid approach to classify Tempranillo Rioja wines according to their origin and style of ageing.
Acknowledgements: This project was funded by the Corporation DOCa Rioja and the José Castillejo program through the Ministerio de Universidades: Programa Estatal de Promoción del Talento y su Empleabilidad en I+D+i, Subprograma Estatal de Movilidad, del Plan Estatal de Investigación Científica y Técnica y de Innovación 2017-2020 (CAS21/00221).
Issue: ICGWS 2023
1Instituto de Ciencias de la Vid y del Vino (Universidad de La Rioja-Consejo Superior de Investigaciones Científicas-Gobierno de La Rioja). Departamento de Enología, Logroño, La Rioja, Spain
2School of Agriculture, Food and Wine, The University of Adelaide, PMB 1, Glen Osmond, South Australia 5064, Australia
Contact the author*
designated origin, A-TEEM, extreme gradient boosting, classification, red wine, statistical modeling