Chemometric Classification of Morphologically Similar Umbelliferae Medicinal Herbs by DART‐TOF‐MS Fingerprint

INTRODUCTION: It needs many years of special training to gain expertise on the organoleptic classification of botanical raw materials and, even for those experts, discrimination among Umbelliferae medicinal herbs remains an intricate challenge due to their morphological similarity. OBJECTIVE: To develop a new chemometric classification method using a direct analysis in real time–time of flight–mass spectrometry (DART‐TOF‐MS) fingerprinting for Umbelliferae medicinal herbs and to provide a platform for its application to the discrimination of other herbal medicines. METHODOLOGY: Angelica tenuissima, Angelica gigas, Angelica dahurica and Cnidium officinale were chosen for this study and ten samples of each species were purchased from various Korean markets. DART‐TOF‐MS was employed on powdered raw materials to obtain a chemical fingerprint of each sample and the orthogonal partial‐least squares method in discriminant analysis (OPLS‐DA) was used for multivariate analysis. RESULTS: All samples of collected species were successfully discriminated from each other according to their characteristic DART‐TOF‐MS fingerprint. Decursin (or decursinol angelate) and byakangelicol were identified as marker molecules for Angelica gigas and A. dahurica, respectively. Using the OPLS method for discriminant analysis, Angelica tenuissima and Cnidium officinale were clearly separated into two groups. Angelica tenuissima was characterised by the presence of ligustilide and unidentified molecular ions of m/z 239 and 283, while senkyunolide A together with signals with m/z 387 and 389 were the marker compounds for Cnidium officinale. CONCLUSION: Elaborating with chemoinformatics, DART‐TOF‐MS fingerprinting with chemoinformatic tools results in a powerful method for the classification of morphologically similar Umbelliferae medicinal herbs and quality control of medicinal herbal products, including the extracts of these crude drugs.