The aim of this study was to develop a new and sensitive ambient mass spectrometric method in combination with multivariate data analysis to classify and quantify contaminations of the traditional herb Equiseti herba through the similar looking E. palustre. Due to its morphological similarity and the often-overlapping habitats, mix-ups of the two plants are common. This is concerning since E. palustre contains substances that are potentially toxic to humans, such as the piperidine alkaloids palustrine and palustridiene. Next to macroscopic and microscopic examination, the identification of Equiseti herba in the European Pharmacopoeia is carried out by means of thin layer chromatography, which must detect up to 5% of foreign compounds. The results of the chemometric models were built from the mass spectrometric data. Furthermore, they were compared with a near infrared spectroscopy method combined with multivariate data analysis, since this type of spectroscopy has already been established in quality control. As a result, the linear discriminant model obtained from the mass spectrometric data could classify an independent set of artificial contaminations of E. arvense through E. palustre with high correctness as low as 1% dry weight contamination grade. In addition, quantification through a partial least square model showed adequate results for a location of the grade of contamination. In conclusion, the new ambient mass spectrometric method presents itself superior to the respective near infrared method and the chemometric models built from its data as well as the official thin layer chromatography of the European Pharmacopoeia.
Read full abstract