Insights into IL-modified silica/6FDA-DAM ionic liquid mixed matrix membranes for CO₂/CH₄ separation: Integrated...

Mixed Matrix Membranes (MMMs) are considered a promising alternative to conventional natural gas sweetening processes. However, their performance is often hindered by poor filler dispersion and weak polymer-filler compatibility, which result in non-selective gas transport. Ionic liquids (ILs) can improve dispersion and enhance interfacial interactions, thereby improving separation performance, but identifying an appropriate polymer-filler-IL combination at the experimental scale is challenging due to the vast structural variations of ILs. This study introduces a novel integrated computational approach that combines quantum mechanics with atomistic computations to investigate IL-supported silica (IL-Si)/6FDA-DAM ILMMMs, aiming to establish a predictive mechanistic framework for the material formulation in CO₂/CH₄ separation. COSMO-RS was employed to screen 150 ILs using activity coefficient (γ) predictions, from which [P8888][Tf₂N] was identified as the most CO₂-philic candidate. This IL was then supported on silica and incorporated into a 6FDA-DAM matrix. The influence of IL-Si loading (0, 8, and 16 wt%) on thermophysical and transport properties was systematically examined under pure and mixed-gas environments (30, 50, and 70% CO₂/CH₄) at 35 °C and 1 atm, enabling molecular-level insight into the effect of gas composition on separation performance. Results showed that the glass transition temperature (Tg) of the hybrid membrane increased from 409 °C (neat 6FDA-DAM) to 421 °C at 8 wt% IL-Si, indicating strong filler dispersion and interfacial adhesion. For separation performance, the optimized membrane at 16 wt% IL-Si achieved a CO₂ permeability of 1413.9 Barrer and a CO₂/CH₄ selectivity of 62.1 under 50 vol% CO₂/CH₄ conditions, corresponding to improvements of 114% in permeability and 193% in selectivity compared to neat 6FDA-DAM. The obtained results exhibited an error in the 0.7-9% range, confirming the reliability of the simulation framework. Overall, this study demonstrates the potential of simulation-driven design for tailoring IL-functionalized MMMs with enhanced stability, permeability, and selectivity for natural gas upgrading, while pointing toward future integration of molecular simulations with AI-based tools for accelerated discovery and optimization of next-generation membranes.