Multiplatform characterization and multi-block data analysis of Robusta Amazônico coffee husk cultivars using 1H NMR and NIR spectroscopies and relationship with their antioxidant potential

Coffee husk is the residue obtained from the coffee processing, representing 50–55% of the bean. The husk obtained from specialty coffees can be processed and dried to obtain flavored infusions. In Brazil, the husks from a specialty coffee named Robusta Amazônico produced in the Amazon region have been transformed into a pleasant beverage. However, the attributes that differentiate husks of different cultivars, as well as between Robusta Amazônico and Arabica are unknown. Therefore, this work characterized four husks of Robusta Amazônico red cultivar (102, 08, VR1, and 25), one yellow Robusta Amazônico husk cultivar, and one Arabica using near-infrared (NIR) and nuclear magnetic resonance (NMR) spectroscopies and differentiated among them using a multi-block chemometric analysis, taking into account their antioxidant potential. Trigonelline, formate, chlorogenic acids, sucrose, glucose, malate, acetate, quinic acid, -quinide, and lactate were metabolites often identified in the samples by NMR spectroscopy. However, 2D NMR spectroscopy should be performed to identify other relevant compounds. Such metabolites were correlated to the different NIR spectral vibrations. VR1 (64 mg of Tr. eq/mL) was significantly the most powerful sample with respect to the antioxidant potential, followed by 25 (104 mg of Tr. eq/mL), 08 (113 mg of Tr. eq/mL), Arabica (118 mg of Tr. eq/mL), yellow Robusta (136 mg of Tr. eq/mL), and 102 (154 mg of Tr. eq/mL). Multiplatform chemometric analysis of NMR, NIR and antioxidant potential data showed that Robusta Amazônico and Arabica husks are distinguishable perfectly. Exclusive cultivars of Robusta Amazônico husk also are distinguishable from each other. It is shown that the species and cultivar of coffee are determinant in the metabolomic composition of the husks and consequently in their antioxidant potential.