Rapid identification of pearl powder from Hyriopsis cumingii by Tri-step infrared spectroscopy combined with computer vision technology

Abstract

Pearl powder, an important raw material in cosmetics and Chinese patent medicines, is commonly uneven in quality and frequently adulterated with low-cost shell powder in the market. The aim of this study is to establish an adequate approach based on Tri-step infrared spectroscopy with enhancing resolution combined with chemometrics for qualitative identification of pearl powder originated from three different quality grades of pearls and quantitative prediction of the proportions of shell powder adulterated in pearl powder. Additionally, computer vision technology (E-eyes) can investigate the color difference among different pearl powders and make it traceable to the pearl quality trait–visual color categories. Though the different grades of pearl powder or adulterated pearl powder have almost identical IR spectra, SD-IR peak intensity at about 861cm−1 (v2 band) exhibited regular enhancement with the increasing quality grade of pearls, while the 1082cm−1 (v1 band), 712cm−1 and 699cm−1 (v4 band) were just the reverse. Contrastly, only the peak intensity at 862cm−1 was enhanced regularly with the increasing concentration of shell powder. Thus, the bands in the ranges of (1550–1350cm−1, 730–680cm−1) and (830–880cm−1, 690–725cm−1) could be exclusive ranges to discriminate three distinct pearl powders and identify adulteration, respectively. For massive sample analysis, a qualitative classification model and a quantitative prediction model based on IR spectra was established successfully by principal component analysis (PCA) and partial least squares (PLS), respectively. The developed method demonstrated great potential for pearl powder quality control and authenticity identification in a direct, holistic manner.