Herein, experimental and theoretical approaches were used to design a new composite membrane for desalination by pervaporation. The theoretical approaches demonstrate the possibility to reach high mass transfer coefficients quite close to those obtained with conventional porous membranes if two conditions are verified: (i) a dense layer with a low thickness and (ii) a support with a high-water permeability. For this purpose, several membranes with a cellulose triacetate (CTA) polymer were prepared and compared with a hydrophobic membrane prepared in a previous study. The composite membranes were tested for several feed conditions, i.e., pure water, brine and saline water containing a surfactant. The results show that, whatever the tested feed, no wetting occurred during several hours of desalination tests. In addition, a steady flux was obtained together with a very high salt rejection (close to 100%) for the CTA membranes. Lastly, the CTA composite membrane was tested with real seawater without any pretreatment. It was shown that the salt rejection was still very high (close to 99.5%) and that no wetting could be detected for several hours. This investigation opens a new direction to prepare specific and sustainable membranes for desalination by pervaporation.
Keywords: CTA composite membranes; desalination; mass transfer prediction; seawater; thermo-pervaporation.