(223ad) In silico Design of Nanostructured Porous Materials for Environmental Applications

Computer simulations and molecular modeling techniques have played an important role in elucidating and providing understanding of observed phenomena in a multitude of fields.  Particularly for porous materials, computer simulations have provided useful insights about the role of pore size, pore size distribution, chemical functionalities, and other physical properties in gas storage, separations, catalysis, and other applications.  New computational techniques are starting to emerge such as large-scale in silico generation of porous materials and high-throughput screening that hold particular promise to help identify promising materials for particular applications, discover structure-property relationships, and determine the potential performance limits of these materials.

            Our recent work has focused on the role of water in adsorption phenomena in porous materials.  In particular, it has focused on the capture of carbon dioxide under humid conditions.  Many materials excel in carbon dioxide adsorption under dry conditions, but they can decompose or their adsorption performance is severely affected by water due to the competition for adsorption sites within these materials.  Hence, research efforts have taken place to develop materials that are hydrophobic but still can adsorb carbon dioxide.  An aspect that has been somewhat overlooked is the role that defects can have in the adsorption of both water and carbon dioxide.  We have used grand canonical Monte Carlo (GCMC) simulations to assess the role of defects in water adsorption in a class of zirconium metal-organic frameworks (MOFs).  We show how a relatively small number of defects can shift the water adsorption behavior, going from hydrophobic to hydrophilic.  Additionally, we discovered that related metal-organic materials contain both hydrophobic and hydrophilic cavities.

            Another important potential application of porous materials is hydrogen storage.  MOFs hold particular promise in this area due to their large pore volumes and surface areas.  However, these materials falter near room temperature due to their low heats of adsorption.  To address this problem, we have investigated MOFs containing open metal sites to increase the heats of adsorption.  Previous work had used quantum mechanical calculations to quantify the interaction energies between hydrogen and the open metal sites, and the results were fit to a Morse potential.  We used this potential in GCMC simulations to assess the performance of these materials for hydrogen storage.  Mg alkoxide functionalization turned out to be a promising candidate.  Then, using recently developed methods for large-scale generation of MOFs, we generated over 18,000 structures containing various numbers of Mg alkoxides and a wide range of physical properties.  In a high-throughput manner, the hydrogen storage capability of these structures was assessed using GCMC simulations, and we identified promising candidates, obtained novel structure-property relationships, and quantified what could be the performance limits of these materials.  Currently, we are working on novel ways to generate new MOF structures with various topologies.  This could lead to new, undiscovered porous materials that can hold promise for various environmentally relevant applications.