Hard Modeling Methods for the Curve Resolution of Data from Liquid Chromatography with a Diode Array Detector and On-Flow Liquid Chromatography with Nuclear Magnetic Resonance Spectroscopy

摘要

Hard modeling methods have been performed on data from high-performance liquid chromatography with a diode array detector (LC-DAD)and on-flow liquid chromatography with ~1H nuclear magnetic spectroscopy (LC-NMR).Four methods have been used to optimize parameters to model concentration profiles,three of which belong to classical optimization methods (the simplex method of Nelder-Mead,sequential quadratic programming approach,and Levenberg-Marquardt method),and the fourth is the application of genetic algorithms using real-value encoding.Only classical methods worked well for LC-DAD data,while all of the methods produced good results when LC-NMR data were divided into small spectral windows of peak clusters and parameters were optimized over each window.