Pore size analysis from retention of neutral solutes through nanofiltration membranes. The contribution of concentration–polarization

Abstract

Pore size distribution is one of the most important characteristics of a membrane. This can be obtained from the fitting of pore radius calculated fromretention versus flux measurements for a set of solute solutions. In thiswork a set of non-charged similar molecules is chosen as solutes to minimize other interactions apart of those related to size. The hydrodynamic model will be used to characterize the behavior of the membrane to uncharged solutes, assuming that membrane pores are straight and cylindrical. As is known, the phenomenon of concentration polarization must be taken into account because true retention is not experimentally accessible by concentration measurements. Frequently, the film layer model is applied for the dependence of concentration with experimental conditions; but the application of this model requires prior knowledge of the mass transfer coefficientwhich is evaluated by different dimensionless correlations (Sherwood correlation). Here we show a review of different alternatives in doing it and analyze their consequences when computing the pore size distribution. Experimental data were obtained from dead-end filtration experiments of a set of four ethylene glycol solutions with a nanofiltration membrane. Obtained results show the importance of the mass transfer model in the pore size value obtained.