Separation of oligostyrene isomers in a complex mixture using two-dimensional heart-cutting reversed-phased liquid chromatography

The development of a two-dimensional liquid chromatographic system requires a process of assessment that can yield an optimum performing system with minimal experimental evaluation. Information Theory and a geometric approach to Factor Analysis are two techniques that when used in combination, provide important information on the expected two-dimensional performance. In the present study, we compare the predicted separation performance of two-dimensional systems that have been subjected to analysis by Information Theory and Factor Analysis to that of actual chromatographic separation performance. Our test separation comprised a mixture of 32 oligostyrene, structural isomers and stereoisomers. The optimal combination as determined by Information Theory and Factor Analysis consisted of a C18 column with a methanol mobile phase in the first dimension and a carbon clad zirconia column with an acetonitrile mobile phase in the second dimension. This system was also shown to be the most successful practical system when a heart-cutting approach was employed. The practical results were in total agreement with the results from Information Theory and Factor Analysis. The number of isomers resolved using this system was 27. A second system, namely one comprising of a C18 column and methanol mobile phase in the first dimension and a carbon clad zirconia column with a methanol mobile phase in the second dimension was also predicted to be a system with high separation potential. However, practical assessment of this system did not realise the theoretical predictions, largely due to the long separation times required in the second dimension. Furthermore, all combinations that employed a C-18 column with an acetonitrile mobile phase in the first dimension failed to realise the theoretical separation potential due to high solute crowding, low orthogonality and a disordered arrangement of bands along the first separation axis. This was also predicted by the theoretical assessment. (C) 2004 Elsevier B.V. All rights reserved.