This is an interesting ...

This is an interesting question. Pore space models for membrane materials are by themselves generally too limited to be a reliable predictive tool for specific contaminant removal. Many other chemical factors contribute to organic contaminant rejection by membranes, including organic-membrane interactions (hydrophobic, estatic, VDW, pi-pi, etc.), organic shape, polarity, structure, flexibility, charge distributions, isomeric forms and more. Another consideration is the dynamics of the nano-pores themselves. How flexible are they? What is their distribution? Is the local pore chemistry different in some way from the bulk material? The dielectric properties of the pores may be quite different than the bulk material.