Near Infrared Spectroscopy Identification Method for Five Chemical Types of Cinnamomum camphora

Abstract

To more quickly and accurately identify the five chemical types of Cinnamomum camphora, a desktop and portable near-infrared spectrometer was used to collect the near-infrared spectra of the leaves of the five chemical types of C. camphora. Combined with chemometric methods, the near-infrared spectra of different chemical types of C. camphora were discriminated and analyzed. Using a desktop near-infrared spectrometer in the 1000-2000 nm wavelength range, we applied dg2 and SNV preprocessing techniques, combined with the SIMCA algorithm, to build a model. This model successfully identified the three chemical types- Linalool, Borneol, and Camphor-with 100% accuracy. Only two samples of Cineol and Iso-nerolidol were not correctly identified. In the wavelength range of 1600-2400 nm, the portable near-infrared spectrometer was pretreated with S-G smoothing and SNV pretreatment. The optimal model was established after spectral matching and the sample recognition rate of the model was as high as 99.73%. A sample of Cineol and Iso-nerolidol was incorrectly identified. In terms of the external verification test, the recognition rates of the two models reached 99.20 and 98% respectively, showing high recognition ability. This study provides a robust and efficient method for the rapid, field-based identification of five different chemotypes of C. camphora, which could significantly benefit practical applications in the field.