Application of High Performance Liquid Chromatography for the Profiling of Complex Chemical Mixtures with the Aid of Chemometrics.
Yongnian Ni,Liangsheng Zhang,Jane Churchill,Serge Kokot
DOI: https://doi.org/10.1016/j.talanta.2007.02.003
IF: 6.1
2007-01-01
Talanta
Abstract:In this paper, chemometrics methods were applied to resolve the high performance liquid chromatography (HPLC) fingerprints of complex, many-component substances to compare samples from a batch from a given manufacturer, or from those of different producers. As an example of such complex substances, we used a common Chinese traditional medicine, Huoxiang Zhengqi Tincture (HZT) for this research. Twenty-one samples, each representing a separate HZT production batch from one of three manufacturers were analyzed by HPLC with the aid of a diode array detector (DAD). An Agilent Zorbax Eclipse XDB-C18 column with an Agilent Zorbax high pressure reliance cartridge guard-column were used. The mobile phase consisted of water (A) and methanol (B) with a gradient program of 25-65% (v/v, B) during 0-30min, 65-55% (v/v, B) during 30-35min and 55-100% (v/v, B) during 35-60min (flow rate, 1.0mlmin(-1); injection volume, 20mul; and column temperature-ambient). The detection wavelength was adjusted for maximum sensitivity at different time periods. A peak area matrix with 21objectsx14HPLC variables was obtained by sampling each chromatogram at 14 common retention times. Similarities were then calculated to discriminate the batch-to-batch samples and also, a more informative multi-criteria decision making methodology (MCDM), PROMETHEE and GAIA, was applied to obtain more information from the chromatograms in order to rank and compare the complex HZT profiles. The results showed that with the MCDM analysis, it was possible to match and discriminate correctly the batch samples from the three different manufacturers. Fourier transform infrared (FT-IR) spectra taken from samples from several batches were compared by the common similarity method with the HPLC results. It was found that the FT-IR spectra did not discriminate the samples from the different batches.