Simulation of the synthesis of metal-organic framework materials
Cessford, Naomi Faye
MetadataShow full item record
The objective of this work was to develop a molecular simulation method with the capacity to represent the synthesis of metal-organic framework (MOF) structures to the extent of being able to accurately predict the MOF structures that form under specified reaction conditions. MOFs are a class of porous, crystalline solids composed of metal-ion vertices coordinated by organic linker molecules. MOFs are created in a self-assembly process in which the building blocks (reactants) retain their integrity. Under different experimental synthesis conditions, a particular combination of building blocks can react to form differing MOF structures. The structure of MOFs confers a large degree of tunability, allowing almost limitless potential for the materials to be designed with the capacity to fulfil the requirements of a specific application. Consequently, MOFs have shown promise for a variety of applications including gas storage, separation and catalysis. Thus, the ability to accurately predict the MOF formed by specifying reaction parameters such as temperature, pH and the concentrations of reactants has great potential because, upon identification of a promising hypothetical structure for a particular application, the synthesis conditions ascertained via the simulation method could be used as the basis for the determination of an experimental synthesis procedure. In addition, a simulation method with the capacity to predict MOF structures affords the opportunity to gain a fundamental understanding of the influence of the experimental synthesis conditions on the structures formed, so as to enable progress towards the rational design of MOFs. In this work, the experimental synthesis of MOFs via self-assembly is modelled using a kinetic Monte Carlo approach. Ideally, simulation of the self-assembly of the building blocks would be modelled atomistically with all atoms in the reactant and solvent molecules represented explicitly. However, due to the prohibitive computational requirements of such a simulation, in this work a “potential-of-mean-force” (PMF) approach was used to represent the solvent implicitly by encompassing the solvent-mediated behaviour in the interactions between building blocks, thereby reducing the computational cost. The PMF approach to the implicit representation of the solvent involved the utilisation of effective pairwise interactions between the constituents of the reactant species. Following extensive testing to ensure that the explicit-solvent behaviour of the reactants could be replicated using the PMF method, this approach allowed computationally efficient implicit-solvent simulations of the synthesis of MOF materials to be performed. Thorough assessment of a method developed to simulate the synthesis of MOFs required investigation of a system which, under different reaction conditions, forms differing structures. In this respect, the cobalt succinates represent an unparalleled test because under different reaction temperatures, reactant concentrations, pH and reaction time, seven different phases have been identified. Furthermore, the parameters within which the different phases form have been clearly delineated experimentally. The method developed has been employed, under the appropriate reaction conditions, to simulate the synthesis of two of the seven identified phases of the cobalt succinates. Whilst still subject to computational limitations, the MOF-synthesis simulation method yields structures characteristic of those expected experimentally under corresponding reaction conditions.