Arsenic is a naturally occurring element that can cause vascular disease, non-malignant skin alterations, and internal malignancies cancers and is thus considered to be a serious hazard to human health. It is commonly found in mine tailings, industrial wastewater, and even natural river systems. Due to its toxicity, the EPA reduced the maximum contaminant limit for arsenic to 10 ìg/L in 2006. Current techniques for the removal of arsenic from drinking water and high concentration brines include ion exchange, adsorption and reverse osmosis. Previous works have shown that negatively charged ultrafiltration membranes, such as regenerated cellulose, can separate arsenic oxyanions from water as long as no other ions are present, and that the addition of cationic surfactants improves separation. This is accomplished by arsenic oxyanions binding to the micelles, which can then be retained using relatively coarse ultrafiltration membranes. The main benefit of this method is that ultrafiltration can be operated at relatively low pressures and that the surfactant in the permeate can potentially be recycled by processes such as foam fractionation. The micelle-driven separation process is known as micellar-enhanced ultrafiltration.

This research focuses on the quantification of ionic strength effects on micellar ultrafiltration of arsenate ions, the optimization of the membrane pore size, and the quantification of the concentration polarization effect apparent at the membrane surface. The surfactant used was cetyl pyridinium chloride, and the membranes used were regenerated cellulose membranes with pore sizes ranging from 1 kDa to 100 kDa. Results show that the addition of surfactant below the critical micelle concentration (CMC) lowers separation by the membrane as the surfactant monomers screens the surface charges of the membrane, making the pores more permeable to arsenate ions. The addition of surfactant above the CMC has to overcome this negative effect before actual separation enhancement can be observed. The addition of surfactant does not improve separation if the membrane pore size is larger than the surfactant micelle. The average molecular weight of a CPC micelle is 32 kDa. Results show that CPC concentrations that are higher than the CMC lead to arsenic rejections of 85% and higher for membranes with pore sizes less than or equal to 30 kDa. The number of arsenate ions bound to each micelle varies between 28 and 0.5 as the concentration of surfactant is varied in the range from 1.0 mM to 15 mM. Due to pumping costs and pressure requirements, the ideal membrane size would be somewhere between 10 kDa and 30 kDa. Concentration polarization at the membrane surface was found to increase arsenic rejection due to a presieving effect; however, the surfactant rejection decreased due to a build up of surfactant micelles and monomers at the surface, creating a sizeable surfactant concentration gradient across the membrane. Increase in the ionic strength of the solutions by means of additional ionic species led to a decrease in arsenic rejection due to ion competition and charge screening, but also to a decrease in surfactant rejection due to the lowering of the surfactant CMC. The results support the use of micellar-enhanced ultrafiltration for the removal of arsenate ions within a specific range of ionic strengths.