Improved Algorithm for Resolution of Overlapped Asymmetric Chromatographic Peaks

Abstract

A simple and improved procedure to resolve an overlapped asymmetric chromatogram into its component peaks is proposed. The overlapped asymmetric peak profile was assumed to be a convolution of its component peaks, which were characterized by an exponentially modified Gaussian, and further simplified by the use of its derivative chromatogram. A new technique is suggested for initial guessing of peak parameters. The simulation study showed that peak parameters were able to be recovered within 5% deviations when the reduced resolution (RR), the ratio of the resolution to the critical resolution, was larger than 1.0 and were sufficient for recovery process. For RR values from 1.0 to 0.6, however, the recovery was not so efficient with the initial guessed values alone and was achieved with a non-linear least-squares routine within 3% deviations in most instances utilizing these values as initial guesses of iterations. The lower limit of RR for this technique was found to be 0.6. The validity of this algorithm for recovered parameters was confirmed by comparison with experimental observations and its recovery ability was found to show no more than a 1.6% deviation from true values and a 2.2% standard deviation throughout the study.