Updating chromatographic predictions by accounting ageing for single and tandem columns

Abstract

The most efficient optimization methodologies in liquid chromatography are based on the modeling and prediction of the chromatographic behavior for each compound in the sample. However, when the column suffers some ageing after the modeling process, predictions may differ significantly from the actual separation. Repeating the modeling is especially troublesome when several columns are involved, as is the case of coupled columns. We propose a shortcut to correct the time and peak profiles in these situations, after evaluating the effects of ageing. The original models are corrected by introducing parameters accounting for column ageing, obtained using the data of a small subset of compounds from those used to model the brand-new column. The ageing parameters are fitted from the discrepancies between the data predicted with the original retention models for the brand-new column and the experimental data measured for the aged column. The approach was developed and tested to predict the chromatographic behavior of 15 sulfonamides, analyzed with individual and tandem columns, using isocratic and gradient elution. Chromatograms more in line with the aged column performance were predicted. The agreement between predictions and experimental data in the aged columns was excellent.