Metal-organic frameworks (MOFs), an important class of nanoporous materials, have attracted considerable interest due to their widespread potential applications in diverse fields such as gas storage and separation, catalysis, photovoltaics, and drug delivery. Nevertheless, the synthesis of MOF materials is highly non-trivial, and often the rational synthesis of a target MOF structure is an elusive goal. Currently, targeted synthesis is currently hindered by the lack of comprehensive and predictive models of the fundamental processes occurring at the MOF-solution interface under synthetically-relevant conditions. Consequently, our research focuses on developing realistic models related to MOF synthesis, with a specific focus on MOF interfacial structure and stability, crystal growth, nucleation, and post-synthetic modification.
Currently, we are focused on simulating the nucleation and growth process(es) of zeolitic imidazole frameworks (ZIFs), an experimentally well-characterized class of MOFs in which tetrahedral metal ions (e.g. Zn, Co) are bridged by imidazolate (Im) linkers. Nucleation and growth processes in ZIFs are challenging for two reasons: firstly, because accurate force fields capable of modeling ZIF reactivity have not been developed, and secondly, because the high apparent activation barriers involved in nucleation and growth prevent us from using standard molecular simulation techniques. In order to address the first challenge of force field development, we are using genetic algorithms in order to develop reactive force fields that can simultaneously treat ZIF framework flexibility, differentiate between various ZIF topologies, and model complex metal-linker bond reactivity. As for the second challenge (that of simulating high activation barriers), we use a combination of sophisticated sampling techniques, namely kinetic Monte Carlo (KMC) and grand canonical Monte Carlo (GCMC), in order to examine the complex dynamics of ZIF nucleation and growth.
By utilizing these next-generation force fields in conjunction with novel sampling approaches, we will be able to simulate fundamental ZIF interfacial processes, thus enabling us to elucidate the mechanisms and provide predictive information for ZIF syntheses. We anticipate that these newly-developed models for MOF synthesis will be particularly useful for the synthetic community, and will ultimately aid in the development of rational strategies for the targeted engineering of a wide variety of MOF materials.