Molecular Dynamics Study of the Relation between Analyte Retention and Surface Diffusion in Reversed-Phase Liquid Chromatography

Abstract

In reversed-phase liquid chromatography (RPLC), analyte molecules retained on the hydrophobic stationary phase can undergo fast surface diffusion within an acetonitrile (ACN)-rich border layer between the stationary phase and the water (W)–ACN mobile phase. We perform molecular dynamics simulations in an RPLC mesopore model employing an endcapped C18 phase to determine retention and diffusive mobility data for four analytes at solvent ratios between 80/20 and 10/90 (v/v) W/ACN. Simulated retention data are validated by experimental retention factors acquired over the full range of studied W/ACN ratios. Our data show that for a given analyte, the lateral mobility gain from surface diffusion increases with the retention factor because both decrease with the elution strength (ACN content) of the mobile phase. A general correlation between analyte retention and surface diffusion is, however, ruled out, as analyte properties influence retention and surface diffusion differently. Complementary simulations of bulk diffusion in W–ACN mixtures show that the lateral mobility of analyte molecules in the ACN ditch can be higher than expected from the local solvent ratio. This occurs only for W-rich mobile phases, when analyte molecules have numerous contacts with bonded-phase groups, and suggests a bonded-phase contribution to surface diffusion through lubrication of retained analytes.

Connecting theory and simulation with experiment for the study of diffusion in nanoporous solids

“…While we primarily focus on applications in heterogeneous catalysis and adsorption based separations, the fundamental concepts of molecular diffusion have broader application to membrane science [12], chromatography [13], gas exploration [14], environmental remediation [15], etc. Greater cross-community collaborations can help provide new insights into diffusion processes that are valuable across disciplines.…”

PoreMS: a software tool for generating silica pore models with user-defined surface functionalisation and pore dimensions

“…As of version 0.2.0, only the structure of β-cristobalite SiO 2 has been implemented. Simulations in slit-pore models have proved that the faces of β-cristobalite SiO 2 , particularly its (111) face, support the envisioned ligand density and residual hydroxylation of functionalised and bare silica materials used in chromatographic columns [10,[12][13][14]24,25,34,40,41]. Moreover, MD simulations have shown that non-functionalised, cylindrical pores carved from β-cristobalite SiO 2 replicate the relevant properties of the bare silica surface in liquid chromatography well [9,11].…”

“…The specific length required depends on the system studied (the properties of the functionalised silica as well as of the solvent mixture). The solvent reservoir length chosen for this example was informed by prior MD simulation studies using a slit-pore model of C 18 functionalised silica equilibrated with water-acetonitrile mixtures [24,25]. The chosen length should suffice for modelling bonded phases of similar or shorter chain length than C 18 and common aqueousorganic mobile phases.…”

“…Molecular simulations, particularly atomistic molecular dynamics (MD) simulations, have already been used to elucidate the mass transport characteristics of different liquid chromatography modes [10][11][12][13][14][15][16] and recently to model catalytic [17][18][19] and drug delivery systems [20]. This suggests that molecular simulations are increasing in importance as an essential building block in multiscale modelling of heterogeneous catalysis [21] and the selection and development of chromatographic columns [22][23][24][25]. The molecular insight gained by such studies was used, for example, to obtain a unified description of diffusion inside mesoporous and microporous structures [26], to rationalise the effect of surface chemistry on olefin metathesis in confined geometries [17], and to provide fundamental insight into reactive ionic liquid films for supported ionic liquid phase (SILP) catalysis relevant for electrochemical systems or the water-gas shift reaction [27][28][29].…”

The Influence of Column Temperature on the Extrapolated Values of the Retention Factor in Reversed-Phase Liquid Chromatography for Water as Mobile Phase

“…According to the literature, a fair approximation of a mean value of the phase ratio φ over the used temperature interval can be considered as 0.25. 22,23 With this value the formula of calculating the standard entropy variation becomes: ΔS 0 = 19.15 x (a' + 0.602). The values resulted from these calculation are given in Table 3.…”