Servicenavigation

Detailseite

Zurück

Projekt Druckansicht

Untersuchungen zur Oberflächendiffusion und effektiven Porendiffusion in der Umkehrphasenchromatographie mittels molekulardynamischer Simulation

Fachliche Zuordnung Technische Thermodynamik
Chemische und Thermische Verfahrenstechnik
Förderung Förderung von 2018 bis 2022
Projektkennung Deutsche Forschungsgemeinschaft (DFG) - Projektnummer 397067997
 
Erstellungsjahr 2022

Zusammenfassung der Projektergebnisse

We presented an extensive set of MD simulations in an RPLC mesopore model with W‒ACN mobile phases. The model consisted of a planar silica slab, whose surfaces were modified with alkyl chains and trimethylsilyl groups for endcapping. The slab was placed in the center of a simulation box and faced identical solvent reservoirs at each side. Due to periodic boundary conditions, the system is equivalent to a slit pore. To relate the diffusive and retentive behavior observed for simple solutes to the actual conditions found in chromatographic practice, we used four aromatic hydrocarbons (ethylbenzene, benzene, acetophenone, benzyl alcohol) that cover the polarity range (apolar to moderately polar) of typical RPLC analytes. During system equilibration, ACN molecules accumulate around the ends of the alkyl chains and form an ACN-rich region (the ACN ditch) between bonded phase and bulk phase. Analyte density profiles obtained from simulations with an endcapped, C18-modified surface allowed to differentiate between the two retention mechanisms: (i) partitioning into the bonded-phase chains and (ii) adsorption at the terminal parts of the alkyl chains. Our simulations recovered the dependence of retention on analyte polarity and size observed in RPLC practice: The retention increases with decreasing polarity and, at comparable polarity, with increasing size of a compound. Furthermore, we confirmed the presence of an ACN ditch in the interface region for mobile phases containing up to 80 vol% ACN. The ACN excess in the interfacial region (compared to the bulk region) decreased with increasing ACN content of the mobile phase. The spatially-dependent diffusion coefficient profiles of the analytes showed an increase in lateral mobility in the ACN ditch for mobile phases of up to 20/80 (v/v) W/ACN. Like retention, the lateral mobility gain decreased with the ACN content of the mobile phase for ethylbenzene, benzene, and acetophenone, whereas the lateral mobility gain of benzyl alcohol (the most polar and least retained compound of the analytes) remained approximately constant between 80/20 and 40/60 (v/v) W/ACN. Additionally, we showed that the lateral mobility of analyte molecules in the ACN ditch can profit from contact with the flexible ends of the bonded-phase chains. This lubricating effect was only observed with W-rich mobile phases, when analyte molecules have many contacts with bonded-phase groups. To analyze the influence of alkyl chain length and ligand density on surface diffusion, we compared an endcapped, C18-modified surface with (i) an endcapped, C8-modified surface at identical ligand density and (ii) a high-density, C8-modified surface. At identical ligand density, the decreased flexibility of C8 chains compared to C18 chains is disadvantageous for surface diffusion of retained analytes, but high ligand density can overcome this disadvantage. In W-rich mobile phases, the high-density, C8-modifed surface provides a higher lateral mobility gain from surface diffusion than the endcapped, C18-modified surface. The results of this study suggest the existence of an optimal chain length and ligand density to produce a maximum lateral mobility gain from surface diffusion. The data obtained from simulations with the endcapped, C18-modified surface were then used in ground-breaking multiscale simulation studies of diffusion and hydrodynamic dispersion in porous media. The goal was to calculate the effective mesopore and bed diffusivities as well as dispersion coefficients for hierarchical, macro‒mesoporous, silica-based materials employed as fixed beds for chromatographic separations (and likewise in heterogeneous catalysis) using physical reconstructions of mesopore space and macropore space available in our group. The spatially-dependent analyte density and mobility profiles obtained from the MD simulations characterized the interfacial dynamics under explicit consideration of surface chemistry, mobilephase and analyte properties. This information was incorporated into Brownian dynamics simulations in the reconstructed mesopore space morphology to derive the effective diffusion coefficient in the mesopore space. Mass transfer between the pore space hierarchies was simulated through an effective homogeneous medium approach for the mesoporous domain in the reconstructed macropore space, arriving at the macroscopic bed diffusivity. Implementation of fluid flow in the macropore space using the lattice-Boltzmann method allowed us to simulate longitudinal and transverse dispersion coefficients. The calculated diffusivities and dispersivities have immediate value as input parameters to transport models of separation (and also catalytic processes), expanding the basis for predictive column modeling. We addressed a fundamental stationary-phase property largely outside the chromatographer's control at the end of this DFG project, namely pore geometry and diameter of the silica support structure. Simulations with cylindrical pore models representing an average and a small pore (9 nm and 6 nm, respectively) for a W-rich mobile phase and an apolar analyte showed that an increase in pore curvature (from planar to 9 nm to 6 nm) has large effects on the bonded-phase, solvent, and analyte density distributions with consequences for the surface diffusivity and the effective pore diffusivity. The presented data revealed higher surface diffusivity with increasing pore curvature. Additionally, we found that a diameter of 6 nm represents the limit for the existence of a bulk liquid region inside the pore, which implies that mass transport relies on surface diffusion alone, whereas mass transport in larger pores combines surface and pore diffusion. Further simulations (currently ongoing) comprising a 12 nm pore, additional analytes, and an ACN-rich mobile phase will illuminate the influence of pore diameter, analyte properties, and ACN content of the mobile phase on surface diffusion.

Projektbezogene Publikationen (Auswahl)

 
 

Zusatzinformationen

© 2025 DFG Kontakt / Impressum / Datenschutz

Textvergrößerung und Kontrastanpassung