Computational Discovery of Metal-Organic Frameworks for Carbon Capture and Natural Gas Storage.

Abstract

Metal-organic frameworks (MOFs) have recently emerged as promising materials for the capture of carbon dioxide (CO2) and the storage of alternative fuels such as methane (CH4). Amongst the many possible MOFs, metal-substituted compounds based on Ni-DOBDC and HKUST-1 have demonstrated the highest capacities for CO2 and CH4. Here we explore the possibility for additional performance tuning within these compounds by computationally screening several metal-substituted variants with respect to their CO2 adsorption enthalpies and CH4 capacities. These compounds are denoted M-DOBDC and M-HKUST-1, where M refers to the composition of the coordinatively unsaturated metal site (CUS), which is varied amongst 18 possible substitutions. Calculations are performed using a variety of techniques, ranging from dispersion-corrected Density Functional Theory to classical Monte Carlo, with the latter simulations employing customized interatomic potentials tuned by first-principles calculations. Regarding CO2 capture, we find that substitutions involving alkaline earth metals and early transition metals yield relatively strong affinities for CO2. Several compositions having adsorption enthalpies within the targeted thermodynamic window were identified. The electronic structure of the MOF/CO2 interaction was characterized and used to rationalize trends in CO2 affinity. In particular, the partial charge on the CUS is found to correlate with the adsorption enthalpy, suggesting that this property may be used as a simple descriptor for carbon captures efficiency. Regarding methane storage, adsorption isotherms were predicted using Grand Canonical Monte Carlo simulations across the M-DOBDC and M-HKUST-1 series, and for hundreds of other MOFs mined from the Cambridge Structure Database. A distinguishing feature of this work is the development of tuned interatomic potentials that properly capture the interaction between CH4 and CUS. The potential developed here reproduces the experimental isotherm for the benchmark Cu-HKUST-1 system very well; consequently, this approach was extended to predict CH4 uptake across the M-HKUST-1 series. Our calculations suggest that Ca-HKUST-1 and Fe-HKUST-1 should exceed the performance of Cu-HKUST-1, which currently holds the record for highest measured methane storage capacity. These compositions are suggested as promising targets for experimental testing. Screening metal-substituted variants of additional MOFs containing the same Cu-paddlewheel unit present in HKUST-1 is suggested as an extension of this work.