C.J.D. Fell

Heat and mass transfer in membrane distillation

Abstract Equations for heat and mass transfer in membrane distillation (MD) have been developed and tested experimentally. The concept of temperature polarisation is introduced and shown to be important in the interpretation of experimental results. Vapour transport through the membranes tested is reasonably described by combined Knudsen and molecular diffusion. The significance of temperature polarisation in the design and operation of large-scale MD modules is discussed, and hollow fibre and tubular systems shown to be potentially the most effective.

A review of fouling and fouling control in ultrafiltration

Abstract This paper discusses the properties of ultrafiltration (UF) membranes which make them susceptible to fouling. Various types of flux decline are described from early usage to long-term effects. For protein UF it is shown that flux decline occurs due to protein deposition, and that this depends on membrane and solute type, solution environment and operating conditions. Attempts to model UF fouling are reviewed and selected examples of fouling control are also described.

The effect of ph and ionic environment on the ultrafiltration of protein solutions with retentive membranes

Abstract This paper discusses the ultrafiltration (UF) of protein (BSA) solutions with retentive membranes over a range of pH values (2—10) and salt concentrations. UF flux transients, following a step input of salt, produced flux drops at pH 2 and 10, but flux increase at pH 5. For long-term runs with fixed feed the flux declined at all pHs and salt contents. Initial and long-term fluxes varied with pH. A flux minimum occurred at pH 5 in the absence of salts, but in the presence of 0.2 M NaCl the monotonically with pH. The final UF flux was found to correlate with the amount of adsorbed protein, particularly in the absence of salts. Maximum adsorption occurred at the isoionic/isoelectric points (≊pH 5), and it was greater in the presence of salts. The results are explained in terms of the effect of different ionic environments on the permeability of the deposited protein. Conformational changes and charge properties of the BSA appear to be the dominant factors determining the flux.

Factors affecting flux in membrane distillation

Abstract This paper summarises a model of the Direct Contact Membrane Distillation (MD) process. The factors affecting flux are shown to be film heat transfer, membrane ‘thermal’ permeability, the partial pressure of air within the pores and the presence of solute in the feed. Solute effects mainly influence vapour pressure and film transfer coefficients. Deaeration significantly increases gas phase permeability, but the effect on flux is modest due to increased temperature polarisation. MD of salt and sucrose solutions at concentrations up to 25 and 30% respectively give fluxes of 60 to 70% of that for a pure water feed. The flux reduction for salt is largely due to vapour pressure reduction, and for sucrose is due to increased viscosity.

The relationship between membrane surface pore characteristics and flux for ultrafiltration membranes

Abstract Surface porosities of Amicon XM100A and XM300 membranes have been measured by electron microscopy and found to be less than 1 per cent. From the measured pore size distributions it is deduced that 50 per cent of the solvent flow is through 20 to 25 per cent of the pores. The conventional model for concentration polarisation in ultrafiltration (UF), which assumes a homogeneously permeable membrane surface, has been modified to account for regions of differing permeability. An effective free area correction factor (≤ 1.0) has been introduced to allow for the effect of membrane surface properties on gel-polarised UF flux. Ultrafiltration experiments with protein solutions and membranes with a range of water fluxes confirm that gel-polarised UF flux is dependent on membrane permeability and surface properties. Effective free area correction factors vary from about 0.4 to 1.0 with values Support for the effective free area concept in UF is provided by an analogy between a gel-polarised UF membrane and a composite reverse osmosis membrane. In both cases the magnitude of the upper ‘controlling’ resistance may be influenced by the pore size and spacing of the lower supporting structure.

Flux decline in protein ultrafiltration

Abstract This paper examines the link between flux decline and protein which becomes deposited (bound) onto the membrane in protein ultrafiltration. For the conditions studied deposition kinetics were relatively slow, with the rate dependent on feed concentration but the “plateau” (steady-state) amount insensitive to this parameter. The amount of deposition was dependent on system hydrodynamics, membrane type and solution environment. Specific resistances of the deposited layer and the labile boundary layer were measured by analysis of unstirred and stirred permeation rates. A semi-empirical relationship, including the deposition kinetics and the deposited layer resistance, gives reasonable prediction of the observed flux decline.

Gas and vapour transport through microporous membranes. I. Knudsen-Poiseuille transition

The transport of gases at subatmospheric pressure through typical microporous membranes falls in the transition between Knudsen and Poiseuille flow. For a range of gases and membranes it is found that flux can be correlated by the simple expression, J=aPbΔP, where b is a measure of the extent of viscous (Poiseuille) flow. The results show how b increases with gas molecular weight and membrane pore size. The use of the relationship for water vapour transport in membrane distillation (MD) is discussed.

Fouling mechanisms of membranes during protein ultrafiltration

Abstract Fouling mechanisms of various ultrafiltration membranes were studied by high resolution filed emission scanning electron microscopic (FESEM) examination of deposits formed during ultrafiltration of albumin protein. Fouling appears to be a surface phenomenon even for membranes that pass substantial amounts of protein. There was no FESEM observable protein within the pores. Two different types of foulant deposit were observed on the membrane surface; fouling by multilayer (cake) coating and fouling by aggregates of proteins. While aggregating type fouling was observed for membranes with high initial UF fluxes, those with lower initial UF flux showed cake formation. Aggregation may be initiated by rapid supersaturation of proteins above the pores due to high convective flows.

Optimal channel spacer design for ultrafiltration

Abstract Ultrafiltration experiments have been performed with a Dextran T500 solution in a thin channel fitted with various types and orientations of spacer. Mass transfer can be described by a laminar flow type of correlation, whereas channel pressure loss is best described by turbulent flow effects. In both cases, and particularly for pressure loss, the performance varies significantly with the characteristic angle of flow diversion induced by the spacer. An analysis of processing costs, including equipment and operating (energy) components, shows that optimal spacers can be identified. For the range of conditions considered the minimum cost is found for a spacer with a characteristic angle of ca. 80°. The minimum varies with feed concentration and flow-rate.

Ultrafiltration of protein solutions through partially permeable membranes — the effect of adsorption and solution environment☆

Abstract In the ultrafiltration of proteins through partially-permeable membranes both flux and rejection are time-dependent. Flux shows an initially rapid and then a more gradual decline which is attributable to loss of porosity by internal adsorption followed by surface adsorption. Rejection rises steadily as adsorption reduces the effective sieving coefficient of the membrane. The influence of pH and ions on long-term flux is explained by conformational changes and charge effects in the solute deposited on the membrane. Subtle differences in rejection as a function of solution environment can also be explained by these effects.