Fabrication optimization of acrylonitrile butadiene styrene (ABS)/polyvinylpyrrolidone (PVP) nanofiltration membrane using response surface methodology

This paper introduces response surface methodology (RSM) as an efficient approach for modeling and optimization nanofiltration (NF) membrane preparation via phase inversion. RSM statistical design (full factorial) was applied to develop the predictive regression models for optimization. This design provides a model with small prediction error, and permits a judgment of the model adequacy. NF acrylonitrile butadiene styrene (ABS)/polyvinylpyrrolidone (PVP) blend membranes were fabricated using phase inversion induced by immersion precipitation method. Dimethylacetamide and water were used as solvent and coagulant respectively. Polymer concentration, evaporation (EVP) time and additive concentration were the preparation factors which their main and interaction effects on membrane morphology and performance were investigated. Membrane morphology was characterized by scanning electron microscopy (SEM). Cross flow permeation experiments were applied for performance tests. According to the analysis of variance (ANOVA) all three independent parameters were statistically significant and the final model was accurate. Response surfaces and contours were plotted for representation of the regression equations and interpretation. Finally. the RSM optimizer supplied three optimized conditions for each main ABS concentration level. These optimum conditions were proven to have high performance, where deviations between predicted and actual responses fall within 4%. (C) 2011 Elsevier B.V. All rights reserved.