Comparison of benchtop near infrared and micro near infrared spectrometer for quality control of dried ginger and its different degrees of processed products

Abstract Dried ginger (DG) is the dried rhizome of Zingiber offcinale Rosc. (Fam. Zingiberaceae), Ginger charcoal (GC) is the processed products of DG, due to the difficulty in controlling the processing degree, unqualified processed products (light carbonized ginger, LCG; heavy carbonized ginger, HCG) often appear, which can both affect clinical efficacy. To ensure the quality and safety of DG and its processed products, this paper mainly focused on the feasibility of rapidly determining the quality of DG and its processed products by benchtop-NIR spectrometer and Micro-NIR spectrometer. A total of 100 samples were scanned by benchtop-NIR spectrometer and Micro-NIR spectrometer respectively. The collected NIRS data were qualitatively and quantitatively analyzed by different models. The partial least squares discriminant models established based on the benchtop-NIR and Micro-NIR spectroscopy both have a discrimination rate of 100% for the prediction set. Three characteristic wavelength selection methods (CARS, VCPA-IRIV and VCPA-GA) were used to improve the performance of partial least squares regression model. The results showed that the benchtop-NIR model with the wavelength selected by the CARS algorithm performed best, and the Micro-NIR model performance of the characteristic wavelength selected by the VCPA-IRIV algorithm had a better prediction ability. The consistency evaluation of the prediction models showed that the two instruments had a very good coherence except for the zingerone model. Therefore, A micro-NIR spectrometer combined with an appropriate wavelength selection method can precisely distinguish DG, LCG, MCG and HCG from each other and accurately determine the five gingerol compounds in DG and its processed products.