Instrumental Broadening Correction in Size Exclusion Chromatography. Comparison of Several Deconvolution Techniques

Abstract

Several deconvolution techniques (1–7) and a novel method herein presented, are compared in relation to their ability for correcting size exclusion chromatograms for the undesirable effect of instrumental broadening. Such methods are evaluated on the same computer, and through a “synthetic” example of known solution. Methods based on the frequency domain are only applicable to uniform deconvolution problems with stationary statistics. However, in the herein presented heuristic method (based on the Wiener filter in the frequency domain), it is possible to relax this last restriction, and generate solutions that are equivalent to considering signals with time-varying statistics. The evaluated techniques are compared on the basis of quality of results, computational considerations and adjustment facility. Stochastic techniques provide the best (and nearly identical) numerical solutions. This is mainly due to their increased facility to introduce “a priori” information about the expected solution. As a counterpart, stochastic techniques are conceptually more complex, more difficult to implement, and normally inolve more elaborate adjustment procedures.