Authentication and Quality Assessment of Cocoa via FTIR and Chemometrics: Geographical Origin, Alkalization, and Quantitative Adulteration Detection

This is the authors' abstract. We don't add key points for this paper.

The adulteration, authentication, and quality of cocoa‐based products are very important issues for both consumers and producers in terms of health, economic concerns, and legal trade. In this study, a method combining Fourier transform infrared (FTIR) spectroscopy with chemometrics was developed for classification of origin‐based (West Africa, East Africa, North America, and South America) and alkalized/natural‐based cocoa samples and also for quantification of origin‐based cocoa liquor adulterations. For classification studies, recorded FTIR spectra of samples were evaluated with PCA and PLS‐DA methods. On the other hand, for quantification of adulteration, Peru‐ and Ghana‐origin liquor samples were mixed (2.5%–80%) and analyzed. Physicochemical analyses (moisture, pH, ash, and color) were also performed to interpret the modeling results. According to origin‐based classification, specificity and sensitivity values were 1.0 for calibration, cross‐validation, and prediction for PLS‐DA. Moreover, for the alkalized/natural separation, specificity and sensitivity values were 1.0 for calibration and prediction, respectively. In the PLS model for quantification of adulteration, calibration, cross‐validation, and prediction correlation coefficients ( R 2 , , and ) were found to be 0.98, 0.93, and 0.96, respectively. Furthermore, the root mean square error of prediction value was 0.06%. This is an exemplary study in the literature for the rapid and nondestructive determination of origin‐based adulteration.