Column selection for liquid chromatographic estimation of the kw′ hydrophobicity parameter

Abstract

Newer reversed-phase column technologies that incorporate polar groups either by an endcapping procedure or by embedding them into the stationary phase ligand have been receiving much attention in the literature for their robustness when highly aqueous conditions are used. We investigated their ability to accurately determine the chromatographic hydrophobicity value log k w′. The non-linear deviations of retention data as mobile phase conditions approach zero percent modifier are a large source of error when extrapolating to log k w′ values using the linear solvent strength model. Here, we compare a conventional reversed-phase stationary phase with others that have incorporated either polar embedded or polar endcapped phases, along with a hybrid-based particle derivatized with a polar embedded ligand. Our results show that polar endcapped phases perform very similarly to the conventional phase and do not show any improved ability for determining log k w′, but polar embedded phases have reduced curvature in the data, and therefore result in less error in extrapolation. We also investigated the solubility parameter model and the [E T(30)] model for their extrapolation efficiency, and have concluded that the [E T(30)] model shows the least error when extrapolating the data.