Quantitatively Investigating the Relationship between Rheology of Polymer Solutions and Their Resulting Membrane Morphology.

Current membrane-based separation research strives to design membranes with high separation efficiency and throughput while reducing energy consumption. To best meet this goal, optimizing membrane morphology with a focus on enhancing the membranes’ physical properties for end-use applications is critical. In most industrial membrane separations, an asymmetric morphology is preferred. Asymmetric membranes are characterized by a thin and dense skin layer supported by a thick porous layer. During fabrication of asymmetric membranes via nonsolvent-induced phase separation, large pores known as macrovoids may form within the porous support. Process conditions dictate if macrovoids are acceptable. High flux separations prefer macrovoids, since their presence decreases support layer mass transfer resistance. However, macrovoids compromise the mechanical integrity of the support and are thus undesirable for separations performed at high temperatures or pressures. A fundamental understanding of macrovoid formation and growth trends is necessary to tailor the structure and mechanical integrity of asymmetric membranes for specific applications.

This poster presents a quantitative analysis of the relationship between the rheology of binary pairings of polymers and polar aprotic solvents and their resulting membrane morphology. The rheology data are processed to understand concentration-dependent viscosity trends and how they vary by polymer/solvent pairing. Using scanning electron microscope (SEM) cross sections of resulting membranes, macrovoid densities are calculated as a proxy for macrovoid size, using image processing techniques written in MATLAB scripts.

While most studies in the open literature use microscopy as a qualitative descriptor for macrovoid size and show SEM images, we will present a quantitative analysis of the macrovoid size. We have quantitatively observed that (a) solutions with lower entanglement concentrations tend to yield less less macrovoids, (b) solvent-normalized viscosities seem to have a positive correlation with calculated and observed macrovoid sizes for polar aprotic solvents, and (c) negative correlation between the concentration of polymer solutions and their resultant membranes’ macrovoid density. These quantitative comparisons are paramount in broader efforts develop models that predict the presence and size of macrovoids in asymmetric membranes.