Chromatographic fingerprint of Semen Armeniacae Amarae based on high-performance liquid chromatogram and chemometric methods

Abstract

Thirteen real Semen Armeniacae Amarae (SAA) samples and 28 commercial SAA samples were obtained from four typical sources of China. The high-performance liquid chromatographic (HPLC) fingerprints of them were firstly recorded. Chromatographic fingerprint was estimated by adding all the UV spectra recorded at each wavelength in the range between 210 and 360 nm. The chromatographic fingerprint obtained in this way provided more information than the chromatograms recorded at single wavelengths. Principal component analysis (PCA) based on the chromatographic fingerprint obtained as referred in previous paragraphs was performed to visualize and classify the studied samples in principal spaces. Similarity analysis (SA) and absolute peak areas (APA) were applied to estimate the concentrations of their components. Hierarchical clustering analysis (HCA) was also carried out to validate the clustering results. Furthermore, amygdalin and another four components were checked out by loadings plots of PCA and could be used as potential chemical markers for discrimination among different sources of samples. The results indicate that herbal medicines from different geographical conditions lead to different constituent contents. The total chromatographic fingerprint adding UV spectra recorded at each wavelength between 210 and 360 nm was reliable for the chemical fingerprint analysis of herbal medicine and would provide a useful reference for quality control of herbal medicines.