Desalination of brackish water from Tunisian Sahel using composite polymethylhydrosiloxane–cellulose acetate membranes

Composite cellulose acetate (CA) based membranes having a hydrophobic top layer made of polymethylhydrosiloxane (PMHS) were prepared with the aim of desalination of brackish waters from the Sahel region of Tunisia having a total dissolved salt content (TDS) of about 4100 g.m–3. The porous substructure made of CA was obtained by a phase separation process from acetone/formamide (2/1 wt/wt) casting solutions. Two polymer concentrations (20 and 22 wt. %) were used with a heat post-treatment temperature going from 60 to 90°C in order to produce membranes with pore sizes responding to nanofiltration needs. Overcoating the CA substructure with a PMHS thin layer gave rise to the composite membranes. Scanning electron microscopy analysis showed the asymmetric and composite feature of the CA–PMHS membrane morphology. Contact angle measurements clearly pointed out a marked hydrophobic character of the composite membranes due to the PMHS surface modification. When compared to the corresponding unmodified CA membranes, the salt rejection determined by conductivity in the range of 78– 93% was strongly enhanced whereas the pure water permeability of the composite membranes was reduced by a factor of 15–50% depending on the starting pore size. It results in the same salt retention — an improvement factor of 50–100% for the permeation rate when using the composite membranes. The ion rejection in decreasing order was SO42– > Cl– > NO3– and Mg2+ ≥ Ca2+ > Na+ for anions and cations, respectively.

A series of CA membranes of different porosity in the nanofiltration range were prepared by varying the formulation of the dope solution and the annealing conditions. Desalination of well brackish water from the Monastir area in Tunisia was performed with these membranes and compared with the same materials modified by a thin layer of PMHS onto the surface. The data obtained showed that the surface modification enables the CA supports to increase the salt rejection. Although a decline of the water flux was Table 6 Permeate chemical composition and rejection factor for the desalting experiments using MS20-80 NO3– Ion –3 Permeate (eq.m ) Rejection (%) Cl- HCO3– SO42– ΣΑ Na+ K+ Mg2+ Ca2+ ΣC 4.3 5.56 5.6 0.01 0.08 0.25 5.94 0.26 observed, a compromise can be reached for the membranes having more opened pores. In this case, the desalting performance of the composite membranes was higher than that of the unmodified CA supports. As the charge density of the studied materials was low, the separation mechanism was assumed to take place mainly through steric, dielectric and hydration interaction.