Rapid Analysis of Multiple Components in Tobacco Using the Antaris II FT-NIR Analyzer

FT-NIR was evaluated as a method to analyze 16 components (analytes) in tobacco. For each component, spectra of 578 to 785 samples were utilized respectively as calibration samples to develop the calibration model. For all calibrations, the raw spectra were pretreated as first derivative, and Partial Least Square (PLS) was used as the calibration algorithm. About 50 randomly selected validation samples were used to test the prediction accuracy of each calibration model. The root mean square error of prediction (RMSEP) for nicotine, total sugar, reductive sugar, total nitrogen, potassium, chlorine, total volatile acids, total volatile bases, sulfate, starch, cellulose, polyphenols, total petroleum ether extracts, petroleum ether extracts neutral, ash, water soluble ash bases are 0.170, 1.17, 0.92, 0.0882, 0.186, 0.0529, 0.00530, 0.0205, 0.159, 0.564, 0.00855, 2.78, 0.00420, 0.00361, 0.945, 0.226, respectively. The results indicate that FT-NIR is a viable alternative to traditional wet chemical methods to rapidly analyze multiple components in tobacco. In turn, the utilization of the AntarisTM II FT-NIR analyzer is shown to be a cost-effective QC tool for the tobacco and cigarette industries.