Simulations to Predict Temperature and Plasticization Effects for Gas Separation in 3D Canal Ladder Polymers

Approximately 10-15% of the world’s energy consumption is used for chemical separations via traditional distillation and absorption methods. Gas separation research focuses on membrane-based technologies to reduce these energy requirements by up to 90%. Industrial membranes currently fall well-below the Robeson upper bound due to high selectivity’s required for industrial useful separations, yielding low permeabilities. Polymers of intrinsic microporosity (PIMs) have been used to create membranes that have pushed the limits of the Robeson upper bound due to their rigid and contorted structure that leads to poor packing—leading to high permeability but selectivity’s below industrial separation requirements. Previously, our group reported PIMs made of fused norbornyl benzocyclobutene repeat units made via catalytic arene-norbornene annulation (CANAL) with 3D contortions that exhibit strong permselectivity. Specifically, CANAL-Me-Me₂F has a selectivity two times higher and a permeability 100 times higher than cellulose acetate membranes for CO₂/CH₄ separations.

This work presents the first simulated results for system design, material properties, and sorption for CANAL polymers. Simulated CANAL-Me-Me₂F was validated against experimental with bulk and skeletal densities, wide angle x-ray scattering (WAXS), and small gas adsorption isotherms. Nonlocal density functional theory (NLDFT), a conventional but indirect method of obtaining pore size distribution (PSD), is seen to be unable to detect the angstrom-sized pores due to experimental limitations. To understand the crucial importance of size-sieving in gas separation membranes, geometric PSD calculated via these simulations is seen to be a far more physically consistent approach. Sorption of plasticizing gases was accurately simulated by integrating molecular dynamics into iterative grand canonical Monte Carlo (GCMC) simulations. Sorption at temperatures inaccessible to experimental determination, but relevant for industrial separations, was used to examine the unique physical aging behavior of CANALs. These methods allowed for generation of sorption coefficients without the necessity of the sorption-diffusion framework. Dual-mode sorption (DMS) models were used to predict multi-gas sorption selectivity. The physically consistent results suggest that these simulations are a vital companion to experimental results, providing previously inaccessible data, validating assumptions, and suggesting new experimental directions.