Combination of linear solvent strength model and quantitative structure–retention relationships as a comprehensive procedure of approximate prediction of retention in gradient liquid chromatography

• Published in: Journal of Chromatography A 2002-07, Vol.962 (1), p.41-55
• Main Authors: Ba̧czek, Tomasz, Kaliszan, Roman
• Format: Article
• Language: English
• Subjects: Analytical chemistry; Chemistry; Chromatographic methods and physical methods associated with chromatography; Chromatography, High Pressure Liquid - methods; Exact sciences and technology; Other chromatographic methods; Quantitative Structure-Activity Relationship; Regression Analysis; Solvents - chemistry
• ISSN: 0021-9673
• DOI: 10.1016/S0021-9673(02)00557-5

Quantitative structure–retention relationships (QSRR) combined with the linear solvent strength (LSS) model are demonstrated to provide approximate predictions of gradient reversed-phase high-performance liquid chromatography (HPLC) retention time for any structurally defined analyte on a once characterized column. The approach requires at first the determination of retention times for a predesigned model series of 15 analytes in two gradient runs. Then by employing the LSS theory a given HPLC system of interest is quantitatively characterized. Structure of the model analytes is next described quantitatively by means of three structural descriptors from standard molecular modeling: total dipole moment, electron excess charge of the most negatively charged atom and water-accessible molecular surface area. With those data the general QSRR equations are derived which describe gradient retention times of the model analytes in the specific column/eluent system. Having now the structural descriptors for any analyte to be chromatographed in such a characterized HPLC system, one employs respective general QSRR equations to calculate its expected gradient retention time at given gradient conditions by means of appropriate LSS equations. Additionally, the chromatographic parameters log k w and S can be calculated and retention coefficients corresponding to chosen isocratic conditions evaluated. The approach provides retention predictions which can be treated as a first approximation of actual data. Predictions are not yet precise enough for practical separation purposes but can be of use in rational modification of analytical conditions aimed at optimization of separations.