Predicting the Electric Conductivity and Potassium Leaching of Coffee by NIR Spectroscopy Technique

This study aimed to investigate the relationship between the experimental values of electrical conductivity and potassium leaching and near-infrared (NIR) spectra of green coffee beans. Analyses of electrical conductivity and potassium leaching were performed on two hundred fifty samples of green coffee beans. The near-infrared spectra of these samples were collected. The spectra were subjected to pretreatments of multiplicative scatter correction (MSC), standard normal variate (SNV), first derivative, and second derivative. The multivariate statistical method of partial least square (PLS) regression was used in the elaboration of calibration models, which were later applied to forecast external samples. The results obtained demonstrate that good calibration models were developed. The best model obtained for estimating electrical conductivity showed a correlation of 0.97, while for potassium leaching, the maximum correlation between experimental and predicted values was 0.88. These results demonstrate good prospects for predicting the coffee bean cell membranes integrity through spectroscopic techniques.