The characteristics of a non-linear optimization technique for resolution of overlapping chromatographic peaks are examined. A modified Meiron-Marquardt method was used. The estimates of the parameters of overlapping peaks in simulated chromatograms were investigated to indicate the limitations of present mathematical methods and, hopefully, to improve their ultimate utility. Gaussian shapes as well as exponential-Gaussian convolutes were used to simulate the chromatographic peaks. Effects on the overall performance of varying heights, widths, and separation of two peaks were determined. Random additive noise and base line drift were also simulated. For illustrative purposes, the performance of the parameter estimation techniques was expressed in terms of relative errors in estimating the second (or smaller) peak's area, height and location. The results presented indicate the relative importance of noise, skewness, height and width ratios and peak separation on the maximum resolution achievable by numerical methods in an automated chromatographic system.
Bibliographical noteFunding Information:
*Based in part on a thesis submitted by I. Phyo to the Graduate School of the University of Minnesota in partial fulfillment of the degree of Doctor of Philosophy, December 1971. Supported in part by U. S. Public Health Service Research Grant No. RR-267 from the Biotechnology Resources Branch of the Research Resources Division of the National Institutes of Health.