Jump to Main Content
Assessment of the hindered transport model in predicting the rejection of trace organic compounds by nanofiltration
- Kong, Fan-xin, Yang, Hong-wei, Wang, Xiao-mao, Xie, Yuefeng F.
- Journal of membrane science 2016 v.498 pp. 57-66
- artificial membranes, chloramphenicol, ciprofloxacin, convection, drugs, glucose, haloacetic acids, hydrophobicity, mass transfer, models, molecular weight, nanofiltration, prediction, sodium chloride, solutes
- The DSPM&DE (Donnan steric pore model and dielectric exclusion) model was employed to predict the rejection of six haloacetic acids (HAAs) and six pharmaceuticals (PhACs), selected to have different molecular weight, hydrophobicity and charge, by two commercial nanofiltration (NF) membranes (HL and NF270). Increasing filtration pressures were applied to vary the rejection ratios. Glucose and NaCl were used as the probe solutes for the determination of the three adjustable parameters involved in the model. Results showed that the model could accurately predict the rejection of the HAAs by both NF membranes with general deviations less than 5%, but it generally over-predicted the rejection of the PhACs. According to the DSPM&DE model, diffusion was the predominant mass transport mechanism in the membrane for both the HAAs and the probe solutes. Experimental determination by conducting diffusion cell test however showed that diffusion only played a minor role in the overall mass transport normally with a contribution less than 10%. The disagreement of model calculation from experimental determination might be due to the improper quantification of the coefficients for solute partitioning between the water phase and the solid (membrane) phase and the hindrance factors for convection and diffusion by the DSPM&DE model. The high accuracy of the model in predicting the rejection of HAAs was owing to the high similarity in physicochemical properties of HAAs with the used probe solutes. If nizatidine were used as the probe solute, the rejection of ciprofloxacin and chloramphenicol by both HL and NF270 would be well predicted.