Exploration and classification of chromatographic fingerprints as additional tool for identification and quality control of several Artemisia species.

The World Health Organization accepts chromatographic fingerprints as a tool for identification and quality control of herbal medicines. This is the first study in which the distinction, identification and quality control of four different Artemisia species, i.e. Artemisia vulgaris, A. absinthium, A. annua and A. capillaris samples, is performed based on the evaluation of entire chromatographic fingerprint profiles developed with identical experimental conditions. High-Performance Liquid Chromatography (HPLC) with Diode Array Detection (DAD) was used to develop the fingerprints. Application of factorial designs leads to methanol/water (80:20 (v/v)) as the best extraction solvent for the pulverised plant material and to a shaking bath for 30 min as extraction method. Further, so-called screening, optimisation and fine-tuning phases were performed during fingerprint development. Most information about the different Artemisia species, i.e. the highest number of separated peaks in the fingerprint, was acquired on four coupled Chromolith columns (100 mm × 4.6 mm I.D.). Trifluoroacetic acid 0.05% (v/v) was used as mobile-phase additive in a stepwise linear methanol/water gradient, i.e. 5, 34, 41, 72 and 95% (v/v) methanol at 0, 9, 30, 44 and 51 min, where the last mobile phase composition was kept isocratic till 60 min. One detection wavelength was selected to perform data analysis. The lowest similarity between the fingerprints of the four species was present at 214 nm. The HPLC/DAD method was applied on 199 herbal samples of the four Artemisia species, resulting in 357 fingerprints. The within- and between-day variation of the entire method, as well as the quality control fingerprints obtained during routine analysis, were found acceptable. The distinction of these Artemisia species was evaluated based on the entire chromatographic profiles, developed by a shared method, and visualised in score plots by means of the Principal Component Analysis (PCA) exploratory data-analysis technique. Samples of different quality could be indicated on the score plots. No multi-component analysis was required to reach the goal. Furthermore, differences related to the origin of some of the not-certified samples were shown. The importance of the specific herbal part used for its identification was also presented. In addition, no differences were observed among fingerprints of lyophilised or conditioned-air dried samples. Finally, a classification technique, Soft Independent Modelling by Class Analogy (SIMCA), was successfully evaluated as identification technique for unknown samples. Six additional Artemisia species (29 herbal samples) were identified as not belonging to any of the four modelled classes. The developed chromatographic fingerprints and the evaluation of the entire profiles provide an added value to the distinction, identification and quality control of the simultaneously investigated Artemisia species.