PCA, PLS, and MCR-ALS applied to the analysis of miscibility and quantification of Cupuassu butter, Passion fruit oil and Sacha Inchi oil blended in Cocoa butter

• Cocoa butter is more miscible with vegetable fats than with oils. • Raman spectroscopy classifies oils and fats by their degree of unsaturation. • Raman spectroscopy and chemometrics classifies and quantifies vegetable oils and fats. Raman spectroscopy combined with chemometrics is widely used in food science due to its power of analysis, speed and environmental friendliness. In the search for new cocoa butter alternatives (CBA), the properties of other matrices such as butter and vegetable oils were studied. In the present study, Raman spectroscopy was used to analyze the chemical properties of cocoa butter (CB), cupuassu butter (CUB), passion fruit oil (PFO), sacha inchi oil (SIO) and its blends with the CB at five concentrations. These matrices have spectral differences based on the intensity of the unsaturated fatty acid peaks and the physical state at room temperature. The highest degree of unsaturation was established for PFO. PCA model found clusters between blend types and concentrations with an accuracy of 95.11–96.22 % and 94.37–98.26 %, respectively. PLS was able to predict the concentrations of CUB, PFO and SIO in blends with CB at RMSEP of 0.47 (R 2 P = 0.49), 0.25 (R 2 P = 0.85), and 0.48 (R 2 P = 0.62), respectively. MCR-ALS applied to the Raman chemical images of the blends established mathematical models with an accuracy of 99.06–99.73 %. The distribution maps analysis, relative standard deviation and histograms showed that CB and CUB at 3 % had the best miscibility and therefore can be used as CBA for the manufacture of chocolates.