Partitioning and transport in random media
Lisa A. Fanti · Scholarly Commons (University of Pennsylvania) · 1989
The partitioning and transport of solutes within porous matrices is of importance in a variety of processes including gel permeation chromatography and membrane separations. The majority of the materials used in these devices possess extremely disordered microstructures, which presents a major challenge in the modeling of the above phenomena. The present thesis addresses partitioning into disordered porous media by examining several model random media, which can be described statistically in terms of only a few parameters. Particular attention is paid to fibrous, sintered-type and sponge-like structures. A statistical mechanical and thermodynamic approach is taken, in which the methods of equilibrium Monte Carlo simulation, density functional and integral equation theories are developed and implemented. The effects of pore geometry and solute concentration on the partition coefficient are investigated in depth. The issue of pore accessibility is also considered by introducing some percolation concepts. The ability of these materials to permit the macroscopic transport of solutes is assessed by computing a newly defined quantity: the filtration percolation threshold. Results of these studies show that the degree of partitioning is highly dependent upon the specific details of the pore geometry. There appears to be no universal parameter capable of characterizing the partitioning of solutes into disordered materials. Indications are that an accurate description of the pore space, in the form of a model random medium, is beneficial.