ULTRAVIOLET-VISIBLE (UV-VIS) SPECTROSCOPY AND CLUSTER ANALYSIS AS A RAPID TOOL FOR CLASSIFICATION OF MEDICINAL PLANTS

The ultraviolet-visible (UV-Vis) spectroscopy coupled with cluster analysis (CA) was evaluated for the classification of some medicinal plants of different geographical growing area. To have a deeper view, the experiment was carried out on herbs belonging to different families. The UV-Vis spectra of hydroalcoholic extracts were acquired in the range of 200-800 nm. The hierarchical clustering analysis (HCA) was applied to the data matrix provided by unprocessed, normalized and standardized spectra respectively. Different types of distance measuring of (dis)similarity between the samples as well as different kinds of linkage or amalgamation rule were taken into account. The best results for the classification of the selected medicinal plants were obtained using Ward's method as the amalgamation rule combined with 1-Pearson r clustering distance measurement. The obtained results reveal the ability of HCA with Ward and 1-Pearson r algorithm to identify plant species even when the raw material has different provenience areas and different pedoclimatic growing conditions. In addition, this methodology revealed a direct link between herbs from different families.