(645c) Modeling of Skinning Effect and Optimization of Dry-Casting Process

Dry casting process is one of the methods to produce a polymer membrane film by casting polymer solution on the continuously moving support film and supplying heated airflow to vaporize solvent form the solution. The polymer membrane film can be produced with merely overheated airflow, but there is probability of occurring skinning effect which deforms the morphology of the membrane film. The main cause of skinning effect is a low solvent concentration near the surface of the polymer solution. When solvent is vaporized with overheated airflow, solvent concentration near the surface is reduced more rapidly than that of the solution due to the sharp vaporization at the surface. If solvent concentration decreases, the diffusion coefficient of solvent in the solution also decreases, which limits the diffusion of the solvent near the surface. As a result, skinning effect occurs; thin solid layer is formed on the top of the solution and hinders solvent inside the solution from evaporating [5, 6]. Therefore, prediction and prevention of skinning effect is an important issue in operating dry-casting process.

The process conditions causing skinning effect can be analyzed through experiments and established empirically. However, it is costly and time consuming to collect sufficient data for analysis. The analysis of skinning effect will be efficient if causal conditions of skinning effect are predictable via model simulation. Therefore, it is aimed to find the process conditions that produce the polymer film that meets the desired value while minimizing the skinning effect. The process in the simulation includes 6 zones through which the polymer solution passes in order and temperatures of these zones are control variables to adjust properties of the final polymer film. The nonlinear model which consists of Flory-Huggins theory and Free volume theory is used for the simulation [1-4]. Flory-Huggins theory accounts for mixing effect of polymer solution different from that of general solutions. Free volume theory describes mass transfer equations of polymer solution and the interactions between solvent and polymer [7, 8]. The system for the simulation is polystyrene/toluene as polymer/solvent. Diffusion coefficient of solvent is traced so as to analyze skinning effect via the simulation since skinning effect cannot be observed directly in the model simulation. First, temperatures of 6 zones are randomly selected to figure out the tendency of diffusion coefficient of solvent to temperature distribution. Trajectories of diffusion coefficient of solvent are described through thickness direction of the polymer solution and temperature distribution where diffusion coefficient of solvent is drastically declined is grasped the trajectories. Furthermore, optimization is also performed based on the model equations to find the temperature distribution that minimizes the curvature of the trajectories and to be compared with the results of random simulations.

Temperature distribution that makes diffusion coefficient of solvent decrease slowly was found through temperature random sampling simulations. Solvent is removed by evaporation on the top of the solution and by diffusion at the bottom. Thus, concentration gradient need to be made by heated airflow so as to diffuse the solvent of the bottom upward. However, overheated airflow accelerates evaporation of the solvent on the surface, decreases diffusion coefficient rapidly and eventually induces skinning effect. Therefore, time for diffusion needs to be given with lowering temperature after concentration gradient is formed to prevent skinning effect. Temperature distribution via optimization also has a similar result, which is high temperature at early stage and relatively low temperature in a next zone. The strategy for prediction of skinning effect could be obtained by model simulation and optimization. Skinning effect in other dry-casting process with different properties of polymer solution is predictable by using the model simulation and the number of experiments to find out temperature conditions for preventing skinning effect is minimized.

References

[1] Altinkaya, S. A., & Ozbas, B. (2004). Modeling of asymmetric membrane formation by dry-casting method. Journal of Membrane Science, 230(1-2), 71-89.

[2] Arya, R. K. (2012). Drying Induced Phase Separation. Journal of Chemical Engineering, 27, 12-20.

[3] Bouyer, D., Werapun, W., Pochat-Bohatier, C., & Deratani, A. (2010). Morphological properties of membranes fabricated by VIPS process using PEI/NMP/water system: SEM analysis and mass transfer modelling. Journal of Membrane Science, 349(1-2), 97-112.

[4] Bouyer, D., & Pochat‐Bohatier, C. (2013). Validation of mass‐transfer model for VIPS process using in situ measurements performed by near‐infrared spectroscopy. AIChE Journal, 59(2), 671-686.

[5] Ciampi, E., & McDonald, P. J. (2003). Skin formation and water distribution in semicrystalline polymer layers cast from solution: a magnetic resonance imaging study. Macromolecules, 36(22), 8398-8405.

[6] Vinjamur, M., & Cairncross, R. A. (2001). A high airflow drying experimental set-up to study drying behavior of polymer solvent coatings. Drying Technology, 19(8), 1591-1612.

[7] Vrentas, J. S., & Vrentas, C. M. (1995). Determination of free-volume parameters for solvent self-diffusion in polymer-solvent systems. Macromolecules, 28(13), 4740-4741.

[8] Vrentas, J. S., & Vrentas, C. M. (2003). Evaluation of the free‐volume theory of diffusion. Journal of Polymer Science Part B: Polymer Physics, 41(5), 501-507.