AUTHENTICATION OF GEOGRAPHICAL ORIGIN IN COFFEE USING UV-VIS SPECTROPHOTOMETRY COMBINED WITH DD-SIMCA MODELLING: A STUDY WITH ROBUSTA AMAZÔNICO

Robusta Amazônico coffee is grown in a sustainable agroforestry system in the state of Rondônia, located in the Brazilian Amazon region, in a terroir that gives the beans unique characteristics, making it a product with an Indicated Geographical Indication. This valuable coffee runs the risk of being fraudulently misrepresented as to its origin, requiring the implementation of authentication tools to safeguard the product's identity. The aim of this study was therefore to develop a method using ultraviolet-visible spectrophotometry (UV-VIS) combined with data-driven and independent modelling of class analogies (DD-SIMCA) for authenticating beans of Robosta Amazônico coffees. For this purpose, samples of Robusta Amazônico coffees (n=114), canephora coffees from other regions (n=108) and arabica coffees (n=12) were used. The UV-VIS spectra of the diluted extracts, obtained according to the process indicated by Souto (2017), were obtained on a quartz plate in a multimode microplate reader (BMG Labtech, model Fluostar Omega, Germany), with a resolution of 1.0 nm in the 200-1000 nm region. Spectral data were preprocessed with Multiplicative Scatter Correction. A DD-SIMCA was applied and using Kennard-Stone algorithm, the Robusta Amazônico was split into two sets (86 for calibration and 28 for external validation). The other coffees samples were added into the external validation set, and the quality of the model was assessed by sensitivity and specificity values. The calibration procedure showed no outliers or anomalous values (α=0.01, β=1). The model developed had a sensitivity value of 100% and a specificity of 95%, indicating that there were only 6 false-positives for the target class. In this sense, the combination of UV-VIS, a technique available in several laboratories, associated to DD-SIMCA proved effective in authenticating the coffees studied. Therefore, the proposed methodology could be useful for applications in quality control procedures and certification of origin of coffee beans with a geographical indication.