Infrared-Photoacoustic Spectroscopy and Multiproduct Multivariate Calibration to Estimate the Proportion of Coffee Defects in Roasted Samples

Infrared-photoacoustic spectroscopy (IR-PAS) and partial least squares (PLS) were tested as a rapid alternative to conventional methods evaluate the proportion of coffee defects in roasted ground coffees. Multiproduct multivariate calibration models obtained from spectra healthy beans Coffea canephora C. arabica (Arabica) blends composed defective Arabica different proportions. The blends, named selections, contained sour, black, broken, whole beans, skin, woods. Six built using samples. model was optimized through outlier evaluation, parameters merit such accuracy, sensitivity, limits detection quantification, inverse analytical linearity, adjustment computed. presented predictive capacity high sensitivity determining defects, all being predicted with suitable correlation coefficients (ranging 0.7176 0.8080) presenting adequate performance. displayed promising results, prediction developed for %defects can be safely used an reference method. Furthermore, new method is fast, efficient, in-line application quality control industrial processing.