On the search of rigorous thermo-kinetic model for wet phase inversion technique

Abstract

Polymeric membranes are usually fabricated by phase inversion process because the membrane morphology and its separation properties can be altered directly by the process conditions. The experimental study of this process and involved subsequent phenomenological events is almost impossible in many cases, mainly due to the fast rates of solvent/nonsolvent exchange and small thickness of cast polymeric film. Hence, many researchers have mathematically tried to develop predictive models. However, to establish a predictive model, an appropriate coupling of thermodynamic stability states of polymeric solution and the different possible involved kinetic mechanisms must be incorporated as investigated in the present work. Here, first, it has been shown that previous models had incorporated incorrect assumptions during model development and inconsistent form of governing equations for numerical calculations has been used which violates the conservative characteristics of mass conservation laws. The governing equations were obtained and used in simulations. In addition, interpretation of interface between the dope phase and coagulation phase was implemented using a kinetic approach instead of local pseudo thermodynamic phase equilibria. The compressible regular solution model was used for calculations of thermodynamic properties as well as construction of phase diagrams and boundaries. The equation of changes in thickness of cast film was established by using the fractional volume changes upon mixing allowing the assessment of cast film geometry effect on the process dynamics directly. Also, the mass average velocity equation was corrected and new expression was obtained. As case study, the cellulose acetate/acetone/water ternary system was used for evaluation of the developed model. The developed thermo-kinetic model was successful in prediction of membranes properties/morphology as compared to the available data. It has been shown that the practical precipitation path is different to those of reported in literature except for the regions near the interface as this skin layer reaches to its final stable condition in early moments of immersion process. The results are presented and discussed in details.