Determination of Arabica and Robusta species in blends of roasted coffee by Mid Infrared spectroscopy in association with mixture design

This article presents a blend design methodology using a real case study of Arabica and Robusta coffee blends, which are the primary coffee species produced worldwide. Arabica coffee is known for its superior sensory quality compared to Robusta, resulting in significant price differences. However, undisclosed mixtures between these species after roasting and grinding are common in the coffee market and challenging to detect, especially when Robusta beans are added to coffees claimed to be 100 % Arabica. Thus, this method proposes a rapid and reliable solution to ensure the authenticity of coffee beans. The calibration curve is constructed using only three points and demonstrates excellent predictive efficiency. The objective of this study was to determine the proportion of coffee species in a blend. For this purpose, a quadratic model was fitted using Fourier transform mid-infrared measurements (FT-MIR). By employing mixture design, the number of samples was reduced. The adjusted model was y^=5.63±(0.03)X1+1.28±(0.03)X2−1.36±(0.13)X1X2. Additional measurements were conducted to verify the method's efficiency, and the predicted versus observed values exhibited a high correlation, with R² = 0.99 and Q² = 0.45. The results demonstrate a promising approach for quantifying blends of commercialized coffees. This methodology offers a rapid and cost-effective solution to authenticate coffee blends, enabling producers, distributors, and consumers to ensure the accuracy and quality of coffee products in the market. The proposed technique can contribute to the prevention of fraudulent practices and uphold the integrity of the coffee industry. Further research and application of this method can be extended to other food products where the authentication of blends is crucial for quality control and consumer satisfaction.