Resumo:
The quality of the coffee is determined by the sensorial analysis, using subjective methods and it is known that this quality is related to the chemical constituents of the grains. The present study used infrared spectroscopy to correlate the chemical composition of the grain with the final quality of the product and to assist in the current evaluation method, generating more reliability in the process. This study applied the spontaneous and induced fermentation with starter cultures (yeast) to evaluate the sensory potential from the impacts generated by the types of processes adopted. The experiments were conducted in a randomized block design with five replicates, in a 3x4 factorial scheme, with three fermentation times: 24, 48, and 72 hours and four wet processes: Washed, Yeast fermentation, Fully washed without yeast, and Fully washed with yeast. Infrared spectra were taken in the medium region (FTIR-ATR) of the 34 samples generated by the treatments. Given that the method of determining the quality of coffee is subjective, based on scores given by cuppers, the FTIR-ATR spectra were used in the construction of chemometric models to predict sensory ratings given to the attribute ‘acidity’. The mean error of prediction of the model constructed for the sensory score was 0.11. Sensory results indicated a higher score for the acidity attribute, in dry fermentation with yeast at the time of 48 hours, suggesting a new strategy for the production of quality robusta coffee.
