The removal of three emerging pollutants: carbamazepine, ketoprofen, and bisphenol A, has been studied using the nanofiltration flat sheet membrane NF99HF. The removal efficiencies of the membrane have been evaluated by two system characteristic parameters: permeate flux and rejection coefficient. The influence of two operating variables has been analysed: operating pressure and feed concentration. Before and after the tests with emerging pollutants, the membrane has been characterized by determining its water permeability coefficient and its magnesium chloride rejection coefficient to find out if the removal of emerging pollutants causes membrane fouling. The results show that operating pressure has significant separation effects, obtaining the highest efficiencies at a pressure of 20 bar for pollutant concentrations between 5 and 25 mg/L. Moreover, rejection of ketoprofen was found to be dependent on electrostatic repulsion, while rejection of bisphenol A was significantly affected by adsorption onto the membrane. Finally, the experimental data have been fitted to the solution diffusion model and to the simplified model of Spiegler-Kedem-Katchalsky to predict the behaviour of the nanofiltration membrane in the removal of the tested pollutants. Good agreement between the experimental and predicted carbamazepine and bisphenol A data has been obtained with each model, respectively.
Keywords: Spiegler-Kedem-Katchalsky model; bisphenol A; carbamazepine; emerging pollutants; ketoprofen; nanofiltration membranes; solution-diffusion model.