Simplistic determination of the membrane pore charge density in presence of mixture of electrolytes

Abstract

Membrane pore charge density (Xd) is an effective but non-measurable parameter for determining the performance of low cut-off ultrafiltration (LCUF) and nanofiltration (NF) membranes. Xd along with the physicochemical properties of the solution determines the membrane zeta potential. These properties include the solution pH, ionic strength as well as the ratio and nature of electrolytes in the solution. Hence, with the prior knowledge of the solution physicochemical properties and the zeta potential, determination of Xd is possible. In this study, the effect of binary mixture of electrolytes, i.e., NaCl/Na2SO4 on the membrane zeta potential was investigated for different solution pH (3–9) as well as the total salt concentration (2–10 g L−1). An empirical correlation between Xd and the other experimentally measurable parameters was developed. Salt rejection experiments were carried out with the above mentioned combinations of pH and total salt concentration along with different weight ratios of NaCl/Na2SO4 namely, 100:0, 90:10, 70:30, 50:50 and 0:100 for six LCUF membranes. For PAN PDA/PEI 12 membrane, at pH 7 and 2 g L−1 of total salt concentration, the rejection of divalent anion (SO42−) was increased from 60% to 75.3%, for monovalent cation (Na+), it was increased from 33% to 41.2% with the concentration of Na2SO4. However, the corresponding rejection of monovalent anion (Cl−) was decreased from 33% to 16.1%. For each of the experimental condition, Xd obtained from the Donnan Steric Pore Model (DSPM) solution was correlated with the zeta potential (ζ) and weight fraction of divalent salt (r) as: Xd=a+br+cζ. The effect of the solution pH and ionic strength is included in the membrane zeta potential. The correlation was successfully tested for the commercial membranes in predictive mode. The empirical correlation proposed in this paper acts as a useful tool to determine Xd for amine based LCUF membranes. This helps to predict the membrane performance for real life desalination treatment without performing a series of rejection experiments.