Nutshell detection in cinnamon powder using a portable NIR spectrometer and hierarchical modelling

Cruz-Tirado J.P.; Fossati F.; Haughey S.A.; Elliott C.T.; Siche R.; Barbin D. · Journal of Food Composition and Analysis (2025) · DOI: 10.1016/j.jfca.2025.107226

Resumen

Cinnamon is a valuable spice susceptible to adulteration with allergenic nutshells, which can compromise food safety. This study aimed to detect peanut, pecan, and walnut shell adulteration in cinnamon powder using a low-cost, portable near-infrared (NIR) spectrometer and chemometrics. Two cinnamon varieties, Cinnamomum verum and Cinnamomum cassia, were tested with adulteration levels ranging from 1 % to 50 %. Principal Component Analysis (PCA) effectively distinguished pure C. verum and C. cassia, though adulterated samples clustered together. Data-Driven Soft Independent Modelling Class Analogy (DD-SIMCA) achieved high specificity for detecting adulterated C. cassia (SPE = 100 %). However, Partial Least Square Discriminant Analysis (PLS-DA) had lower sensitivity in distinguishing adulteration types, particularly in C. verum mixed with pecan shell (SEN = 50 %) and C. cassia with walnut shell (SEN = 66.7 %). A hierarchical model improved true positive rates (TPR) for C. verum adulterated with pecan shell (TPR = 80 %) and C. cassia with walnut shell (TPR = 83.3 %). These findings demonstrate the potential of portable NIR spectrometry and chemometrics for rapid, non-destructive detection of adulterated cinnamon, supporting food quality and food safety control in the spice industry.

Instituciones: Universidad Nacional de Trujillo