Abstract

The variable viscosity Poiseuille model (VVPM), developed recently to treat gas-liquid displacement (GLD) porometry data for track-etched polyethylene terephthalate (PET) membranes, is applied to polyethersulfone (PES) and polypropylene (PP) microfiltration membranes prepared by phase inversion. It is found that in general aspects the pore size distribution as estimated by the model, with the intrinsic assumption of isolated capillary pores, perfectly reproduces wet and dry fluxes during porometry measurements, but the estimated porosity appeared to be much higher than unity. To eliminate the discrepancy, two additional parameters, namely non-uniformity and tortuosity coefficient, have been introduced in the flux and porosity estimating equations. The apparent viscosity of the gas in porometry analysis is found to be several fold higher than that available in literature, and also the viscosity range appeared to be membrane type dependent. The pore size distribution curve shifts along the diameter axis depending on the gas flow direction applied in porometry data acquisition. However, because the membranes had narrow pore size disribution, the estimated parameters had been averaged within acceptable confidence range. With the inclusion of non-uniformity and tortuosity of the capillaries, the revised version of the VVPM made an advancement in porometry data analysis of membranes with both isolated and interconnected pore structure; it characterizes the assumed model capillaries in terms of Young-Laplace pore diameter, but also non-uniformity and tortuosity of pores. The procedure for data treatment has been illustrated in detail and the method could easily be adapted to gas-liquid displacement porometer software system for regular data treatment. Based on a set of readily available experimental data, including the typical output from standard GLD porometry, the here established extended version of the VVPM allows the step-by-step determination of all the parameters which characterize the pore structure of the membranes. Also the absolute pore number density distributions can be obtained, including estimates of additional pore characteristics such as non-uniformity and tortuosity.

Full Text
Published version (Free)

Talk to us

Join us for a 30 min session where you can share your feedback and ask us any queries you have

Schedule a call