Network modeling of the pore structure of porous media for determining the diffusion and convective flow of a solute in monoliths and columns packed with chromatographic particles under unretained and adsorbing conditions -- comparison between theory and experiments

A versatile pore network model (cubic lattice network) is constructed describing the void structure in monoliths and in porous chromatographic particles packed in a column with respect to pore size distribution, pore connectivity, and pore spatial distribution. Constitutive equations for mass transfer are developed for a single pore and are employed in the network model to determine a priori values of the intraparticle convective velocity and pore diffusivity for a given pore structure for different solutes of interest. The network is employed to calculate intraparticle convective velocities and pore diffusivities for adsorbate molecules under retained (adsorption) and unretained conditions where the combined effects of steric hindrance at the entrance to the pores, frictional resistance within the pores, molecular size of the adsorbate and ligand (active site), and fractional saturation of active sites are considered. By obtaining the values of the pore diffusivity of a solute calculated in an a priori fashion from pore network model analysis for a given pore structure, a model was constructed and used to evaluate the dynamic behavior, scale-up, and design of monoliths and packed beds employing adsorbent particles. Breakthrough curves and dynamic adsorptive capacities were calculated based on different pore size distributions and pore connectivities for chromatographic systems operating under retained and unretained conditions. Simulations of nitrogen adsorption/desorption, mercury intrusion, size exclusion chromatography and internal surface area accessibility were performed in order to determine the properties of different pore structures. Also, pore structure parameters (pore size distribution, pore connectivity, and pore spatial distribution) were determined by the comparison of experimental nitrogen adsorption and mercury intrusion data with pore network model simulations; the agreement between theory and experiment was found to be good --Abstract, page iv