Morphological Characterization Versus Membrane Performance

customers who are interested in using membranes in their processes have serious problems on choosing the right membrane. When they examine the manufacturer s catalog, they often find two parameters which are supposed to describe the membrane performance permeability and cut-off. Permeability corresponds to the normalized water flow, pressure and membrane area (usually litres per square meter per bar per hour LMHB) and the cut-off (often in kilo-Daltons or Daltons) which, by definition, corresponds to the size of the particles that are 90% retained by the membrane. 

The permeability of the test fluid that has to be filtered is very difficult to predict, since it depends on other factors. On one hand, there are fouling and polarization problems, which can reduce dramatically the flow rate. These depend on the type of sample, the concentration, etc. On the other hand but related, the design of the membrane module also plays an important role. A correct design of the module can prevent stagnant flows inside the module (which reduces the effective membrane area) and can significantly increase the shear stress on the membrane, which can prevent the mentioned phenomena. 

More important is the cut-off value of the membrane given by the manufacturer. The fact that this value is widely used impUes that it is almost an intrinsic property of the membrane, but reality shows a completely different view. The cut-off not only depends highly on the type of solute that has to be filtered, but also on process parameters like pressure or temperature. Considering also that membrane 

An experimental way to measure the molecular weight cut-off (MWCO) of a membrane is to perform filtration experiments with a fluid containing known amounts of model substances of different molecular weight and for which the molecule radii are known. It is desirable to use substances with a narrow size dispersion to avoid overlapping. It is interesting to repeat experiments at different transmembrane pressures in order to know the correlation with the cut-off results. From the analysis of the permeate sample as well as the retentate one and after cdiecJdng the mass balance, a scatter plot between the size of each component and its retention can be developed and a curve (often exponential) can be obtained. From the curve, the MWCO can be calculated (corresponding size at a retention of 90%). 

From the experimental results obtained, the mean membrane pore diameter can be determined by using correlations found in the literature. Among them, the Ferry-Faxen equation [40] correlates the membrane retention R) with the radius 

---