Detection of cumin powder adulteration with allergenic nutshells using FT-IR and portable NIRS coupled with chemometrics

Cumin (Cuminum cyminum L.) powder is a flavoured and valuable spice, usually adulterated with allergenic nutshells that may affect consumer health. Identification of these adulterants is normally performed with expensive analytical equipment using time-consuming methods. In this work, we compared a portable Near Infrared spectrometer (NIRS) and Fourier Transform infrared (FT-IR) benchtop spectrometer to detect and quantify peanut, pecan and walnut shells in cumin powder. Principal Component Analysis (PCA) allowed a good class separation (pure cumin vs. adulterated cumin), with loadings peaks related to essential oils in cumin and cellulosic compounds in nutshells. Data Driven Soft Independent Class Analogy (DD-SIMCA) was implemented to authenticate pure cumin, with sensitivity (SEN) = 82.4% and specificity (SPS) = 85.4%, while classification using FT-IR data reached SEN = 94.1% and SPS = 91.7%. Partial Least Square Regression (PLSR) models reached residual prediction deviation (RPD) values higher than 3.5 for portable NIR and 8.9 for FT-IR benchtop. Although FT-IR spectroscopy provided better results than NIRS, a portable NIRS could be used as a screening method throughout the supply chain due to the practicality, easy implementation and low cost, complementing a laboratory analysis with benchtop FT-IR.