Solute adhesion —membrane fouling

Membranes are very finely porous structures and like all such porous structures used in an industrial context are susceptible to fouling caused by adhesion of components of the materials being processed. This fouling can be minimised or avoided if suitable polymers are used in membrane manufacture. However, the selection of membrane polymers suited to particular separations has until now been a matter of experience (and failure) rather than science. However, an AFM used with the colloid probe technique [23] can provide a rapid means of assessing the adhesion of solutes to membrane materials and is hence a powerful tool for the membrane technologist. 

The difference in force between B and C is a direct measurement of the adhesive interaction, in this case 1.98 mN/m. For the modified membrane, curve (b) the adhesive interaction is just 0.38 mN/m, so the use of a mixture 

The development of ab initio methods for the prediction of the performance of membrane separation processes has made substantial developments. Sophisticated methods now exist for such prediction, and these have been experimentally verified in the laboratory. The present challenges are two-fold. Firstly, to continue the fundamental development to more complex separations. Secondly, to apply the verified methods in the design of full-scale industrial processes. The existence of good predictive methods is expected to further expand the application of membrane processes. 

All research requires a source of inspiration. The original inspiration for WRB s research lies in the intellectual stimulation provided by WJA as tutor and doctoral research supervisor. This work described in this chapter was funded by the UK BBSRC, the UK EPSRC and the EC. 

New models for the prediction of molecular diffusion coefficients are described, and compared to previously established ones. These are based on solute molecular size, solvent viscosity solvent molecular size, and temperature. The data set of diffusion coefficients used was primarily the one developed by Wilke and Chang and upon which their commonly used diffusion model is based (A.I.Ch.E. Journal, 1 (1955), 264). 

---

During reverse osmosis and ultrafiltration membrane concentration, polarization and fouling are the phenomena responsible for limiting the permeate flux during a cyclic operation (i.e., permeation followed by cleaning). That is, membrane lifetimes and permeate (i.e., pure water) fluxes are primarily affected by the phenomena of concentration polarization (i.e., solute build up) and fouling (e.g., microbial adhesion, gel layer formation, and solute adhesion) at the membrane surface [11]. 

Bowen et al. [40] used sulfonated poly(ether ether ketone) (SPEEK) as an additive in the polysulfone (PSf)/SPEEK/N-methyl-2-pyrrolidone (NMP) system. Membrane characterization was carried out using filtration studies and AEM. Membranes prepared from solutions in the region of polyelectrolyte behavior [41] showed more pronounced and systematic improvement of membrane permeability and salt rejection. A small decrease in pore size and surface roughness was also followed by an increase in SPEEK content. Compared with a - 28.5 mN m adhesion force of a 4 pm silica particle for a SPEEK free PSf membrane, a SPEEK modified membrane showed greatly reduced adhesion of - 0.75 mNm h This, together with the surface smoothening effect, leads to the reduction of membrane fouling when the surface is modified by the addition of SPEEK. 

A potential solution to prevent fouling is to develop fouling release membranes that do not resist the adhesion of foulants, but have an active layer with a low surface energy so that adhered foulants can be readily washed away by hydrodynamic mixing in the membrane module [21] provided the water flux and salt rejection of the resulting membrane are not compromised. 

Such layers are usually made of highly cross-linked material and show good adhesion to the substrate. Examples of such membrane surface treatment are plasma polymerization of allyl alcohol and allyl amine (Gancarz et al. 2002, 2003). It was shown that the membranes modified with allyl amine do not foul as intensively during UF of the BSA solutions compared with the unmodified membranes. Similar behavior was also shown for membranes modified by the deposition of plasma-polymerized n-butylamine however, in this case, the modified layer deposited on the manbrane surface was not as enriched in amines as the polymer formed from aUyl amine (Gancarz et al. 2002). 

The data show that the protein-modified probe has significantly lower adhesion to the development membrane (XP117) than to the existing commercial membrane (ES404) and hence the modified membrane has significantly lower fouling potential in the processing of such protein solutions. 

---