Chemical differentiation of coffee species is essential for quality control, authenticity verification, and the prevention of mislabeling in the coffee industry. This is particularly relevant in Panama, a recognized producer of specialty coffees such as Geisha, where studies on chemical characterization remain limited. This study investigated the phenolic composition, antioxidant activity, and caffeine content of roasted coffee as potential chemical markers for differentiating Coffea arabica and Coffea canephora. Ultrasound-assisted hydroethanolic extraction was optimized using the response surface methodology, with temperature, ethanol concentration, and extraction time as the experimental variables. The optimized extraction parameters were established at 75 °C, 44% ethanol, and 25 min, resulting in highrecovery of the total phenolic content, total flavonoid content, and antioxidant activity (ABTS and DPPH assays). Under these conditions, C. canephora samples obtained higher levels of phenolic compounds, antioxidant activity, and caffeine content than C. arabica. Pearson’s correlation analysis showed significant associations among phenolic compounds, caffeine content, and antioxidant activity. PCA explained 90.5% of the total variance and clearly discriminated between the two species, mainly based on differences in caffeine content and antioxidant activity. HCA confirmed this classification and revealed subgroups within C. arabica, particularly among the Geisha varieties. PLS-DA achieved complete separation between species, with zero classification error under cross-validation. These results indicated that combined chemical and multivariate approaches can be used to differentiate and assess the authenticity of coffee, particularly in underexplored production regions such as Panama.
Monrroy et al. (Tue,) studied this question.