Computational Insights into Gas Separation of CO ₂ /CH ₄ /N ₂ Molecules in 6FDA-Based Polyimides

Polyimides (PIs) are pivotal materials for gas separation, owing to their exceptional chemical stability, tunable free volume, and selective permeability. On the basis of the solution-diffusion mechanism, herein we computationally investigated the dissolution, diffusion, and permeation performances of key CO 2, CH 4, and N 2 molecules in six 6FDA-based PI matrices (6FDA: 4,4′-hexafluoroisopropylidene-diphthalic anhydride) under pressures ranging from 1 to 20 atm by employing molecular dynamic (MD) simulations and Grand Canonical Monte Carlo (GCMC) method. We demonstrated that gas sorption (particularly CO 2 ) induces obvious swelling, significantly enhancing solubility coefficients and increasing the fractional accessible volume (FAV), which quantifies a probe-accessible void space in PIs and more accurately reflects the free volume available for gas transport. A strong correlation of gas permeation properties was observed with FAV. The methyl-abundant PIs, like 6FDA-Durene (Durene: 2,3,5,6-tetramethyl-1,4-phenylenediamine) and 6FDA-DAM (DAM: 2,4,6-trimethyl- m -phenylenediamine), exhibit higher solubility and diffusion performance primarily due to larger free volume and loose chain packing. Notably, both exponential and linear relations between solubility/diffusion coefficients and FAV can be established across the six PIs. As a result, the permeability coefficients show equivalent exponential and quadratic dependences on FAV, confirming the nonlinear relations and the govering role of free volume in gas transport by synergistically modulating both sorption and diffusion dynamic performances. Beyond validating the established exponential correlation between permeability and FAV, this work also proposes a quadratic permeability prediction model using FAV as the key descriptor. The simulated results pave the way for the rational design of PI-based gas separation membranes at the molecular level.