Multi-level computational screening of anion-pillared metal-organic frameworks for propane and propene separation

Although the search of proper adsorbents for efficient propane/propene (PA/PE) separation has attracted great interest, the actual process-level performance of adsorbents has been rarely considered as a filtering criterion. In this work, a multi-level computational screening of a database containing 936 anion-pillared metal–organic frameworks (AP MOF) is performed to identify promising adsorbents for practical PA/PE separation in pressure/vacuum swing adsorption (PVSA) processes. At the crystal level, the geometry properties are computed. The adsorption isotherms at the phase level are predicted via molecular simulations with the newly calibrated force field parameters. Based on the results from these two levels, 216 candidates are discarded. Then, referring to a rigorous 4-step PVSA model, a simplified batch adsorber model is proposed and used to efficiently identify qualified AP MOF candidates that can meet the separation requirements on PE purity (≥99.5 %) and a certain recovery. The screening indicates that 15 AP MOFs can be used to separate 15 %/85 % PA/PE feed gas with 80 % PE recovery. Only 2 out of them are feasible for the separation of 50 %/50 % feed with 75 % PE recovery. Finally, one of the best candidates is sent for rigorous PVSA process optimization. It shows that the screened adsorbent consumes less energy in the separation of 15 %/85 % PA/PE mixture compared with the benchmark 4A zeolite process. Clearly, this work is anticipated to provide an effective approach for the rapid screening of adsorbents for PA/PE separation in PVSA processes.