.png){kind=icon}
.png){kind=icon}
Return
Diffusion Explorer for the COF Space: Data-Driven Discovery of High-Performing COF Membranes for Gas Separations
.png){kind=icon}
OA
AI
G
S
DOI：10.1016/j.ccst.2025.100559.png)
.png){kind=icon}
.png){kind=icon}
Share.png){kind=icon}
Save
Abstract
En 中文
Covalent organic frameworks (COFs) have recently emerged as highly promising candidates for membrane-based gas separations, exhibiting superior performance relative to conventional membrane materials. Nevertheless, the rapidly expanding number of COFs renders the experimental evaluation of each material’s membrane performance infeasible. In this study, we investigated the COF space comprising approximately 70000 synthesized and hypothetical materials using high-throughput molecular dynamics (MD) simulations and machine learning (ML) for computing the diffusivities of CO2, CH4, H2, N2, and O2 gases. We generated an online toolbox by integrating our ML models to estimate gas diffusivities of any given COF material in seconds. Using the ML-predicted diffusivities, gas permeabilities and selectivities of COF membranes were assessed for seven industrially relevant separations; CO2/CH4, CO2/N2, H2/CO2, H2/N2, H2/CH4, O2/N2, N2/CH4. The performance of COF membranes was compared to traditional membrane materials, and the most promising COFs were identified and analyzed using molecular fingerprinting to reveal the critical structural and chemical features for accelerating the design of next-generation COF membranes.
Keywords:
covalent organic frameworks
machine learning
molecular simulation
membrane
separation
AI Summary
Key information extracted from the uploaded paper, including a brief overview, abstract, background, key highlights, visual analysis, and future outlook.
Journal
C
IF:
10.5
Papers:
492
Citations:
1.9K
Organization
No organization information available
Cited Papers