An atomistical elucidation of the 6FDA-durene-based mixed matrix membrane and ionic liquid mixed matrix membrane for CO2/CH4 separation

• Developed computational framework for 6FDA-durene/Si ILMMMs for CO 2 /CH 4 separation. • Framework showed high accuracy with MARD<0.1362 against experimental data. • ILs ion size and strength affected free volume and chain spacing in ILMMMs. • Solubility depended on free volume, and membrane structure limited IL effects on gas solubility. • ILs improved PISi CO 2 /CH 4 selectivity by 28–40 % at 11 wt.% loading. CO 2 removal is crucial in reducing greenhouse gas emissions and ensuring safer natural gas transmission. Pristine membranes suffer low selectivity-permeability trade-offs, limiting their application in industries. The mixed matrix membranes (MMM) offer an alternative, but their utilisation was challenged by low filler-polymer adhesion and agglomeration. Ionic liquid (ILs) mitigated the agglomeration effect and improved materials adhesion, but most research focused on experimental work, lacking an established and validated computational framework to study ionic liquid mixed matrix membranes (ILMMMs), resulting in limited atomistic insights into membrane physical and gas separation properties. This research developed a validated computational framework using Material Studio and the COMPASS force field to study the physical characteristics of CO 2 /CH 4 transport behaviour. Validation against experimental data showed a mean averaged relative deviation (MARD) below 0.1362. Silica reduced the total free volume (FFV) while increasing the d-spacing, indicating a larger pore size. ILs strengthened membrane adhesion, further reducing FFV (14.3 %–21.2 %) and d-spacing (11.1 %–16.8 %). Binding energy analysis confirmed strong integration of silica and ILs ([EMIM][Tf₂N], [BMIM][Tf₂N], [BMIM][PF₆]) with 6FDA-durene membranes by two to three-folds. Silica incorporation increased the CO 2 and CH 4 diffusivity by 40.7 % and 13.9 %, respectively, while the ILs generally reduced the gas diffusivity (approximately 20 %). Gas solubility is strongly correlated with FFV and surface area. While MMM portrayed no major performance change, the selectivity of the ILMMMs improved up to 39.42 %. This research established a validated computational framework for ILMMMs research, enabling better membrane design and optimisation for industrial CO 2 removal.