Quantification and modelling of organic micropollutant removal by reverse osmosis (RO) drinking water treatment

Reverse osmosis (RO) is the most promising membrane technology in organic micropollutants (MPs) removal of drinking water treatment. For 78 MPs, passage and removal were evaluated with an ESPA3 RO membrane and the robustness of RO against MPs was studied. The MPs were classified according to their charge and hydrophobicity. The results showed that the size of neutral compounds was negatively correlated with their passage. This correlation was weaker for neutral hydrophobic MPs than neutral hydrophilic MPs. The lowest passage (0.2%-4%) was displayed by anionic MPs because of electrostatic repulsion between the negatively charged solute and negatively charged membrane surface. Cationic MPs showed a higher passage (around 0.4%-40%) due to electrostatic sorption and Donnan exclusion. The relationship between physical-chemical properties of MPs and their passage was evaluated by the one-way analysis of variance (ANOVA). We performed a qualitative analysis of variables using Principal Component Analysis (PCA) in order to examine the physical-chemical properties of compounds that affect the membrane removal of MPs. After analysis with Multiple Linear Regression (MLR), we concluded that properties such as molecular width, equivalent molecular width, pKa and solubility can be considered as significant descriptors for prediction of the membrane removal. The influence of feed water temperature on MPs passage was also assessed. The results revealed that a rise of water temperature from 5 to 19 degrees C, increases the average passage of MPs by 6.5%.