Accurate prediction of the retention behaviour of polydisperse macromolecules based on a minimum number of experiments

Comparison of retention models for polymers

The suitability of three different retention models to predict the retention times of poly(ethylene glycol)s (PEGs) in gradient and isocratic chromatography was investigated. The models investigated were the linear (LSSM) and the quadratic solvent strength model (QSSM). In addition, a model describing the retention behaviour of polymers was extended to account for gradient elution (PM). It was found that all models are suited to properly predict gradient retention volumes provided the extraction of the analyte specific parameters is performed from gradient experiments as well. The LSSM and QSSM on principle cannot describe retention behaviour under critical or SEC conditions. Since the PM is designed to cover all three modes of polymer chromatography, it is therefore superior to the other models. However, the determination of the analyte specific parameters, which are needed to calibrate the retention behaviour, strongly depend on the suitable selection of initial experiments. A useful strategy for a purposeful selection of these calibration experiments is proposed.

Molar mass distributions by gradient liquid chromatography: predicting and tailoring selectivity

Interactive liquid chromatography (iLC) for polymer analysis is usually applied to the characterisation of distributions other than molar mass. In particular, its use for the determination of chemical-composition, functionality-type and tacticity distributions has been demonstrated. The application of iLC for the determination of molar mass distributions (MMDs), however, has not yet been fully explored. An expanded version of the reversed-phase liquid chromatography model has been developed to describe and predict how the retention behaviour of polydisperse polystyrene samples changes with molar mass. The relationship between molar mass and the parameters of the model has been investigated in some detail and non-linear correlations were found. From the model and the relationships between the model parameters and molar mass, calibration curves (retention time versus molar mass) were constructed to predict changes in chromatographic selectivity across a given molar mass range. These calibration curves were compared to experimentally obtained curves and, in most cases, excellent agreement was found. The dramatic enhancement in selectivity that can be obtained with iLC in comparison to size-exclusion chromatography (SEC) was illustrated by measuring matrix-assisted laser desorption ionisation (MALDI) MS spectra of fractions collected during a gradient-LC separation. In the low-molar mass range, essentially monodisperse fractions were obtained. Calibration curves, predicted by the model and validated experimentally using narrow-dispersity standards and MALDI-MS spectra of fractions, were used to determine the molar mass distribution of some narrowly distributed polystyrene samples. Molar mass distributions for such standards were found to be somewhat lower than the values reported by the manufacturers. The results also deviated from those obtained by MALDI-MS.