High dynamic separation performance of metal-organic frameworks for D₂/H₂: Independent or competitive adsorption?

Most traditional D-2/H-2 separation techniques are energy-intensive with low efficiency. Metal-organic frameworks (MOFs) provide a promising solution for D-2/H-2 separation due to their excellent chemical and structural characteristics. Here, machine learning-assisted high-throughput computational screening was employed to identify the high-performance MOFs for the dynamic D-2/H-2 separation. Extensive data analysis reveals that there were two adsorption behaviors in the optimal MOFs, independent adsorption and competitive adsorption, and the independent adsorption was favorable for the preferential adsorption of D-2. To quantify these two adsorption behaviors, we introduced and defined overlap degree (OD) and independence degree (ID), and developed a software for the rapid assessment of OD/ID. After batch simulation of the breakthrough curves of 2000 optimal MOFs, similar to 80% MOFs exhibited independent occupancy, confirming its contribution to good dynamic separation capabilities. This work provides a new idea for designing MOFs with independent adsorption behavior to improve the dynamic separation performance of hydrogen isotopes.