50 % (on average over 36 and 45 % in green and roasted coffee extracts, respectively). Significant differences between the AC values determined for CB and PA samples were noticed only for the MCA and DPPH methods which reflect the different molecular mechanisms underlying each of the assays. Additionally, the statistical methods, including principal component analysis, applied to results of antioxidant capacity obtained with different analytical techniques confirmed their feasibility to distinguish between coffee brews with different degrees of roasting, regardless of coffee origin. Abstract The antioxidant capacity (AC) of boiled-type coffee brews (CB) and phenolic acids (PA) isolated from them, obtained from the caffeinated and decaffeinated beans of different geographical origins and species and with different roasting degrees, was examined. The AC of PA and CB samples was tested in five antioxidant assays: a total antioxidants reducing capacity assay using a Folin-Ciocalteu reagent (FCR), a ferric ion reducing antioxidant power (FRAP) assay, a DPPH · radical-scavenging activity (DPPH) assay, a metal chelating activity (MCA) assay and a total radical trapping antioxidant parameter (TRAP) assay. In most samples, the total amount of phenolic acids, determined by HPLC, decreased with the increasing degree of roasting the coffee beans, leading to reduced AC. All used methods showed that CB exhibits higher AC compared with the PA samples. Phenolic acids isolated from CB samples have the main contribution (on average over 95 and 84 % in green and roasted coffee extracts, respectively) in AC of the CB samples in FCR, FRAP and TRAP assays, whereas in DPPH and MCA tests, the phenolic acid contribution in AC of CB samples was below
Keywords