Theoretical description of three major retention mechanisms in reversed-phase HPLC on the example of naphthalene, uracil and phenol molecules

Abstract

Nowadays in the high performance liquid chromatography (HPLC) method there is a number of
theoretical models allowing to predict chromatographic characteristics. For the last decade scientists have
developed new models and have improved earlier existing approaches for this analytical method. However in
all these models there is a one common aspect — a high degree of dependence on experimental data and a
small possibility of parameters spreading on a wide range of systems.
It is convenient to describe the interaction of molecules at the level of quantum mechanical concepts
of the chemical bond. The major approach is that the intermolecular interaction energy is defined by the dispersion, induction and orientation forces, and hydrogen bonding, which define character of all types of interaction between atoms in molecules and molecular complexes.
Intermolecular interaction energy correlates exponentially with distribution coefficient where the
difference between the special way calculated adsorption energy of an analyte and a modifier has been taken, and even a small inaccuracy in energy definition considerably affects its value. The retention factor is connected with distribution coefficients over the value of column porosity.
In this article the new approach to the description of chromatographic retention of three types of
analytes molecules for the HPLC isocratic elution mode is considered. In the adsorption effects examination
it is possible to determine three main cases: non-polar molecule retention, retention of solvates hindering the adsorption and heteropolar retention of molecules. Non-empirical calculations have been carried out on the basis of various interactions contribution to the chemical bond with the usage of expressions from the theory of the generalized charges. For the molecules differing on hydrophilicity/hydrophobicity chromatographic characteristics have been finally defined. High degree of correlation between theoretical and experimental data has been found. It has been shown that the developed approach is relevant for application in different systems.