Peak deconvolution to correctly assess the band broadening of chromatographic columns

The present study provides experimental evidence for the fact that the peak deconvolution method can be applied to accurately measure the column-only dispersion of the current generation of high speed and high efficiency columns. Unlike the conventional variance difference method, it furthermore preserves any prevailing asymmetry of the column-only peak. This has been demonstrated by testing the same column on three different system configurations, with different extra-column volumes, and showing that, after deconvolution, the resulting column-only peaks coincide very well and produce very similar column-only plate height values (typical relative standard deviation comprising all runs on three different system configurations is 2-2.5%). Extensively studying a large set of theoretically produced peaks (with exactly known variance and asymmetry), it could be shown that the main criterion for the validity of the deconvolution method is that the variance of the system-only peak is minimum 1.5 times smaller than the variance of the column+system peak. The need to add a radial mixer unit to accurately assess the system-only contributions has been demonstrated as well. To illustrate its use and merits, the deconvolution method has been used to establish so-called multiple van Deemter curves, wherein plate height curves relating to different peak width definitions are shown in the same plot. These plots can give new insights in the intrinsic asymmetry of the column-only dispersion.