Potential use of nanofiltration membranes in treatment of industrial wastewater from Ni-P electroless plating

Abstract

Nanofiltration (NF) membranes are a relatively new class of membranes which have properties in between those of ultrafiltration (UF) membranes and reverse-osmosis (RO) ones. Their separation mechanisms involve both steric (sieving) effects and electrical (Donnan) effects. The significance of this membrane, besides having small pores is the membrane's surface charges, which allows charged solutes that are smaller than the membrane pores to be rejected along with bigger neutral solutes and salts. Furthermore it is capable of rejecting multivalent ions effectively, and on the other hand lets the monovalent ions pass through. These are the characteristics that make the nanofiltration membrane a potential process to reject heavy metal ions, which in general are multivalent ions. In this study, the potential use of a nanofiltration membrane for rejection of heavy metal ions such as Ni 2+ from wastewater from Ni-P electroless plating industry was studied. A composite NF membrane designated as HL membrane was used. The characteristics of the membranes were analyzed using single salt rejection studies of NaCl, MgCl 2, Na 2SO 4, NiSO 4. The wastewater from the Ni-P electroless plating industry was found to contain various cations such as nickel, zinc and sodium. In order to ascertain the effect of ion rejection in a multicomponent system, the rejection behaviour of Ni 2+ ion in mixed solutions of Na 2SO 4 and NiSO 4 was tested and the rejections were then compared to those obtained using real solutions. The results show that in general the rejection of Ni is high regardless of the presence of other ions. The experimental data were also analyzed using two predictive models, namely Donnan steric pore model (DSPM) and ion transport model. The models are based on the extended Nerrnst-Planck equation but solved using different approaches and parameters. The results from the models were compared to find which model gives better agreement with the experimental findings.