(102e) Computational Design of Metal-Organic Frameworks for Gas Storage

Fuels such as natural gas (NG) and hydrogen produce less CO₂emissions than gasoline.  They are therefore attractive to power next-generation vehicles.  However, a challenge that must be addressed first is the storage of these gases in compact volumes such that the vehicle tank is of a reasonable size but allows for a practical driving range.  Vehicles powered by natural gas currently store the gas compressed to 250 bar (CNG) in cumbersome tanks.  An attractive alternative is to use an adsorbent material that enables densification of the fuel at safer, cheaper, and lower pressures. 

Metal-organic frameworks (MOF) are porous, crystalline materials resulting from the assembly of inorganic “nodes” and organic “linkers” into a three-dimensional network.  The essentially limitless possible combinations of inorganic and organic building blocks make MOFs highly tunable and promising for gas storage.  Nonetheless, a major concern in the applications of MOFs is the stability of this class of materials.  Recent synthesis of remarkably stable MOFs, however, suggests that the identity of the inorganic nodes may be the key to stable structures.  In this contribution, we first use a high-throughput computational screening approach, along with data analysis techniques, to investigate thousands of hypothetical MOF structures and explore the performance limits of nanoporous materials based on their methane (the major component of natural gas) “deliverable capacity” at various operation scenarios.  Then, we continue our computational work using a hypothesis-driven approach where we focus on building and investigating the adsorption characteristics of hypothetical MOFs based on stable Zr₆O₄(OH)₄(CO₂)_n inorganic nodes.

After computational characterization and simulation of methane adsorption in over 200 hypothetical zirconium MOFs, some relationships became apparent, which constitute important guidelines for material design.  One example is the link between methane deliverable capacity and the presence and location of triple bonds in the structure of the screened MOFs.  Following these emerging guidelines, we constructed and identified in silico a material with a predicted deliverable methane capacity of 197 cc(STP)/cc between 65 and 5.8 bar. The identified hypothetical zirconium MOF, based on 1,4-benzenedipropynoic acid linker, was synthesized and named NU-800.  Experimentally, it was possible to activate around 90% of the porosity of NU-800, which allowed us to confirm the promising methane adsorption properties of this material.  The measured deliverable capacity between 65 and 5.8 bar was 167 cc(STP)/cc, with the discrepancies between simulation and experiment probably due to the somewhat incomplete activation of the material.  Our screening also allowed us to identify zirconium MOFs with large gravimetric surface areas, which is a key MOF property to achieve high gravimetric hydrogen adsorption.  Some of these materials were synthesized, for instance MOF NU-1100, and their promising hydrogen adsorption properties were confirmed by experimental measurements.   All of the synthesized structures were confirmed to be stable over a dozen adsorption/desorption cycles.