Hydrophilic interaction chromatography (HILIC) is a liquid phase separation technique that has received considerable attention in recent years. Its widespread use is partly explained by modern pharmaceuticals, which are highly designed molecules targeting specific polar interactions, and by many naturally occurring compounds with biological activity, such as, e.g., nutraceuticals and marine toxins, which cannot be separated on alternative separation chemistries. What characterizes HILIC is the use of water-miscible solvents (notably acetonitrile; ACN) with increasing concentrations of water to separate polar compounds on polar stationary phases. This setup is of opposite polarity to the more conventional “reversed phase” liquid chromatographic scheme, where separations take place on hydrophobic phases using the organic member of a miscible water/solvent as strong eluent component. Flipping the polarity has created a whole array of possibilities in designing new stationary phases, and because the diversity in polar groups is potentially endless, the palette of available HILIC separation phases is therefore growing rapidly. Little is, however, known on the actual mechanisms that govern these separations. Over the last decade we have developed new HILIC phases and also used spectroscopic and other experimental methods to elucidate the processes that cause retention in HILIC, where most peers draw (sometimes far-fetched) conclusion merely from retention data. This is where we intend to shed light by correlating our more fundamental experimental investigations with modeling of the surface interaction processes.
The relative retention of solutes in liquid chromatography, and hence their separation, depends partly on attractive surface interactions and partly on “solvophobic expulsion” processes. Explaining HILIC systems to develop deterministic ways to predict retention has proven far more difficult than in systems based on low polarity stationary phases.
The experimental techniques we have employed are mainly based on nuclear magnetic resonance (NMR), and in these we observe interactions that cannot be explained by simplistic models based on how molecules behave in bulk phase. A typical HILIC eluent contains 5-30 % water in ACN with 5-100 mM ammonium formate or acetate as buffer. The common conception is that the retention in HILIC is caused by partitioning of solutes into a water-enriched layer on the surface; however our experiments show that these processes are far more complicated. Mixtures of ACN and water are first of all known to be “microinhomogeneous” already in bulk phase, i.e., the individual water and acetonitrile molecules have a tendency to segregate into transient clusters – a process that is augmented by the addition of electrolytes. When these highly dynamic structured solutions are contacted with polar surfaces, a whole ensemble of gradients are formed into which polar and amphiphilic molecules are partitioned. In recent experimental work we have seen interactions between stationary phases and solutes such as toluene, which has been taken for granted would not be even close to the surface and is hence used as void volume marker. We are therefore in urgent need to model these surface interactions molecular dynamics computation in order to support our experimental data.