Volatile fingerprinting, sensory characterization, and consumer acceptance of pure and blended arabica coffee leaf teas

Coffee leaves contain several bioactive compounds and have been traditionally consumed as a medicinal infusion in the East for centuries. Coffee production generates large amounts of leaves as by-products, which are often wasted in most producing countries because of the low acceptability in the West. Nevertheless, processing and blending coffee leaves may increase aroma and flavor complexity. This study evaluated the volatile and sensory profiles and consumer acceptance of coffee leaf teas compared to two among the most consumed teas (black and maté teas) in Rio de Janeiro. Infusions were made with one experimental and one commercial coffee leaf tea (CLT), two black teas (BT), and one toasted maté tea (TMT) for volatile (GC-MS/MS) and sensory profiles. As an attempt to improve coffee leaf tea acceptance, CLT were also blended (50%) with BT or TMT. Acceptance, Check All That Apply (CATA), and Projective Mapping sensory tests were performed with untrained assessors aged 18-49 (n = 100). Volatile data were standardized by centering and normalization. Sensory data were treated by ANOVA/Fisher test, PCA, and AHCMFA, considering differences at p < 0.05. Ninety-two volatile compounds distributed in 12 classes were identified in different samples. CLT, BT, and TMT infusions shared 19 compounds, including 9 potential impact compounds for aroma and flavor: α-ionone, β-ionone, hexanal, nonanal, decanal, benzaldehyde, trans-linalool oxide, linalool, and dihydroactinidiolide. The most cited flavor attributes for CLT infusions were herbs/green leaf, woody and refreshing. For TMT and BT, herbs/green leaf, woody, burnt, and fermented were the most cited. These attributes agreed with the volatile profiles. CLT shared 22 compounds with TMT and 28 with BT. Considering pure infusions, TMT presented the highest mean acceptance scores (6.7), followed by Com. and Exp. CLT (6.1 and 5.8, on a 9-point-hedonic scale, respectively). Blending with TMT increased mean acceptance of Exp. CLT (6.4), while blending with BT, downgraded the mean acceptance of Com. CLT (5.3). In Projective Mapping, CLT was considered to have a higher sensory resemblance with TMT than BT. If produced adequately, CLT was shown to have good market potential to support sustainable coffee production and promote health.