Rapid detection of the component contents in caryophylli flos by a handheld near infrared spectrometer based on digital light processing technology

A handheld and inexpensive near infrared spectrometer based on digital light processing technology was used to investigate the potential of fast quantitative detection of eugenol, beta-caryophyllene and eugenyl acetate in caryophylli flos. Gas chromatography was used to determine the reference values. The diffuse reflectance spectra of caryophylli flos powder were recorded and were pretreated by different methods, and then the partial least squares regression was applied to develop calibration models; furthermore, the competitive adaptive reweighted sampling was used for the wavelength selection to improve the performance of models. The results show that the best performance of the pretreatment methods is seen with the combination of the first derivative and standard normal variate, and the performance of calibration is improved by the competitive adaptive reweighted sampling. For the eugenol, the standard error of calibration and standard error of prediction are 0.46% and 0.60%, respectively, and the corresponding R-squares are 0.955 and 0.89; for beta-caryophyllene, the standard error of calibration and standard error of prediction are 0.11% and 0.14%, respectively, and the corresponding R-squares are 0.89 and 0.86; for eugenyl acetate, the standard error of calibration and standard error of prediction are 0.30% and 0.38%, respectively, and the corresponding R-squares are 0.89 and 0.80. The overall results of this work revealed the feasibility of the use of handheld near infrared spectrometers as a method for the quantitative on-site determination of eugenol, beta-caryophyllene and eugenyl acetate in caryophylli flos