A Comparison of Models to Describe the Maximal Retention of Organic Molecules in Nanofiltration

Abstract

Nanofiltration is used in a growing number of applications for the treatment of drinking water, wastewater, and process water. Trial-and-error is generally used to test the applicability of nanofiltration and to select the membranes. In particular for organic molecules, a model that describes retention as a function of molecular parameters and membrane characteristics has not yet been established. In this paper four models for maximal retention, represented by the reflection coefficient, were compared: the steric hindrance pore model, the model of Zeman and Wales, the log-normal model, and an adapted version of the log-normal model. The calculated results were compared to reflection coefficients determined experimentally for a broad range of relatively small organic molecules. Each of the models yielded acceptable results, although the steric hindrance pore model and the model of Zeman and Wales are based on a somewhat idealized view of membrane structure. The log-normal model calculates reflection coefficients from a distribution of pore sizes. The adapted log-normal model also includes hydrodynamic lag, caused by sterical hindrance in the membrane pores. It was found that this effect is very small. The log-normal model appeared to be most useful to predict reflection coefficients in practical applications.