Samuel de Lint

Ph.D. thesis

Summary

The last two decades, considerable effort has been invested into the development of membranes that combine the good separation properties of Reverse Osmosis materials with the lower pressures required for Ultrafiltration membranes. This has resulted in the development of nanofiltration (NF) membranes. In a wet state, NF materials have a charged surface, and by means of electrostatic effects they can partly retain charged species that are about one order of magnitude smaller than their pore size. By controlling the electrostatic effects, the effective pore size of NF membranes can, to some degree, be regulated and therefore the separation of charged species can be performed without having to resort to smaller pore sizes and, consequently, higher pressures.

The number of applications in which NF is used is growing rapidly. Currently the reduction of hardness (removal of divalent ions like Ca²⁺, Mg²⁺, SO₄^2-) and dissolved organics from water is the most important application. Another example is the cleaning of water streams from metalworking plants by the removal of heavy metals (Ni, Fe, Cu, Zn). In the food industry, NF is used for the recovery of organic acids from fermentation broths, and the desalting of whey.

In this thesis, the separation properties of inorganic NF membranes are discussed for dilute aqueous electrolyte solutions. The separation mechanisms of NF membranes are not known in detail and therefore the aim was to improve the understanding of transport through NF membranes by the development of a model that is able to predict their separation characteristics. In the final chapter of the thesis the retention predictions of such a model are compared to the experimental retention behaviour for an inorganic membrane in NaCl and CaCl₂ as well as a mixture of these two salts. The promising conclusion is that the model predictions are in good agreement with the experimental retention data. Hence, the primary objective of the research project has been achieved; it is indeed possible to make a predictive multi-component mass transport description for NF separation of electrolyte solutions.

Our approach differs from already existing NF models primarily in the determination of the charging behaviour of the NF membrane. A method is used that determines the individual charging (adsorption) properties of ions and that eliminates the effects of electrostatic double layer overlap. The ion-material specific adsorption parameters are derived, combining electrophoretic mobility measurements with a site-binding model for the surface adsorption chemistry. The adsorption parameters for single ions can be used directly in a transport model to calculate the retention in a multi-component electrolyte solution. Because dilute suspensions of membrane particles are used in the mobility experiments, particle double layer overlap is eliminated.