Loughborough University
Kevin Cliff Thesis - EC.pdf (1.75 MB)

A study of membrane swelling and transport mechanisms in solvent-resistant nanofiltration

Download (1.75 MB)
Version 2 2020-01-15, 10:23
Version 1 2011-11-22, 14:09
posted on 2020-01-15, 10:23 authored by Kevin Terry Cliff
Recently a large amount of interest has developed around separating out impurities of small size; pertinent examples are found within fuel and solvent processing. For such applications a leading candidate process is nanofiltration. This thesis focuses on SRNF (solvent resistant nanofiltration) composite membranes consisting of a dense polymer active layer bonded to a stronger, but ultimately more porous, support layer. The composite membranes that have been produced during the course of this work consist of a PDMS (polymdimethylsiloxane) active layer bonded to a commercially available support layer of PAN (polyacrylonitrile). To create the membrane a monomer was spread over the support layer and then polymerised to form the matrix which was responsible for separation. Commercially, either heat or radiation is often applied to cause polymerisation, however the membranes in the current work have been formed by the used of a homogeneous catalyst. This thesis investigates the transport and separation dynamics of the produced membranes for a series of fuel simulants composed of organometallics and poly-nuclear aromatic solutes dissolved in aromatic and alkane solvents. Membrane composition and the extent of polymer swelling were found to be the two key factors which had the greatest influence on solvent flux and solute rejection. By increasing catalyst concentration it was found that the dual effects of increased rejection and reduced flux occurred, with the converse also being true. The effective pore size of the membrane could also be controlled by varying the catalyst amount during manufacture as this directly affected the limit of crosslinking which formed. Polymer swelling was the most pronounced using solvents with a solubility parameter close to that of the polymer. The membrane transport mechanism was most accurately forecast by the solution diffusion model for flux predictions and the convection diffusion model for rejection predictions, however all the models tried were in close agreement. This was postulated to be due to the swelled polymer matrix which allows for both convective and diffusive transport to occur.



  • Aeronautical, Automotive, Chemical and Materials Engineering


  • Chemical Engineering


Loughborough University

Rights holder

© Kevin Terry Cliff

Publication date



A Doctoral Thesis. Submitted in partial fulfillment of the requirements for the award of Doctor of Philosophy of Loughborough University.

EThOS Persistent ID



  • en


Steve Tarleton

Qualification name

  • PhD

Qualification level

  • Doctoral

Usage metrics

    Chemical Engineering Theses


    Ref. manager