Quality Evaluation of Chocolate by using Near Infrared Spectroscopy

Abstract

Near infrared (NIR) diffuse reflectance spectra were measured on melted and solid chocolate in order to propose a calibration model which predicts their degree of temper. Twenty-three chocolate samples which had different degrees of temper were prepared artificially, and the degree of temper investigated was in the ranges of 1.4-7.6. Partial least squares (PLS) regression was applied to the original NIR spectra data set. The correlation coefficient was calculated to be 0.85 and 0.97 for the melted chocolate and solid chocolate, respectively, and the standard error of prediction (SEP) was found to be 1.11 and 0.57, respectively. The regression coefficients plot for the calibration models showd that bands due to the overtone and combination modes of the CH₂ and CH₃ groups, and C=O bonds, were related to the prediction of the degree of temper. Consequently, it was concluded that NIR spectroscopy had a possibility for evaluating the degree of temper of chocolate.