Statistical Model Based HPLC Analytical Method Adjustment Strategy to Adapt to Different Sets of Analytes in Complicated Samples.

IntroductionOn account of the complicated compositions of the products like traditional Chinese medicines (TCMs) and functional foods, it is a common practice to determine different sets of analytes in the same product for different purposes. ObjectiveTo efficiently develop the corresponding HPLC methods, a statistical model based analytical method adjustment (SMB-AMA) strategy was proposed. MethodsIn this strategy, the HPLC data acquired with design of experiments methodology were efficiently utilised to build the retention models for all the analytes and interferences shown in the chromatograms with multivariate statistical modelling methods. According to the set of analytes under research, Monte-Carlo simulations were conducted based on these retention models to estimate the probability of achieving adequate separations between all the analytes and their interferences. Then the analytical parameters were mathematically optimised to the point giving a high value of this probability to compose a robust HPLC method. Radix Angelica Sinensis (RAS) and its TCM formula with Folium Epimedii (FE) were taken as the complicated samples for case studies. ResultsThe retention models for the compounds in RAS and FE were built independently with correlation coefficients all above 0.9799. The analytical parameters were tactfully adjusted to adapt to six cases of different sets of analytes and different sample matrices. In the validation experiments using the adjusted analytical parameters, satisfactory separations were acquired. ConclusionThe results demonstrated that the SMB-AMA strategy was able to develop HPLC methods rationally and rapidly in the adaption of different sets of analytes. Copyright (c) 2017 John Wiley & Sons, Ltd.