Computational Screening and Analysis of Functional Porous Materials

Abstract

Functional porous materials are a class of materials that have found use in many industrial applications. In particular, extended framework materials, such as metal–organic frameworks (MOFs) and porous aromatic frameworks (PAFs), which are the subject of this thesis, show significant promise for applications including gas storage and separations, catalysis, drug delivery, microelectronics and sensing. This broad scope of applications stems from the immense chemical diversity afforded by their modular bottom-up synthesis and design. Additionally, the rational choice of building blocks allows for the precise control of the properties of the pore networks of crystalline extended porous materials. However, the process of finding optimal porous materials for emerging applications is slow due to arduous trial-and-error experimental approaches. The application of computational methods to analyze porous materials allows for the development of design principles, which can guide experimental endeavors. Furthermore, high-throughput screening can be used to expand on experimental findings by efficiently exploring chemical space for the best candidates for a given application. This thesis reports several studies in which novel computational protocols are developed and applied to more rapidly screen porous functional materials for applications. A coarse-grained molecular dynamics model was developed to investigate the formation mechanism of PAFs and the role of structural and dynamics factors in determining their highly porous, amorphous networks. PAF formation, which is kinetically controlled, was found to robustly lead to a high degree of defects and porosity, and that relatively weak dispersion interactions are responsible for inducing porosity-reducing interpenetration. The simulations suggest that bulky reaction intermediates or building blocks with diminished dispersion interactions can be used to eliminate interpenetration and increase material porosity. Highly-ordered MOF thin films with macroscale in-plane and out-of-plane alignment have many potential applications, but only a handful of examples have been reported to date. Therefore, a high-throughput screening process was developed to suggest new MOFs that are likely to undergo aligned growth. The screening process was parameterized from a set of experimental observations of the aligned growth of copperbased MOFs from copper(II) hydroxide (Cu(OH)2) and allows for the screening of thousands of MOF structures in a few days. Importantly, the number of known MOFs that are likely to grow aligned from Cu(OH)2 was expanded and some design principles were uncovered. In particular, it was found that the substrate imparts a directing effect on the MOFs able to grow aligned, but does not limit the possible pore network properties of aligned MOFs. The biomimetic mineralization of MOFs around biomacromolecules was investigated in two joint experimental and computational studies. Biomimetic mineralization is a general and facile method for encapsulating biological entities to, for example, enhance their stability in harsh conditions. Systematic experimental studies of the encapsulation of proteins and carbohydrates by zeolitic imidazolate framework-8 (ZIF- 8) found that the electrostatic properties of the biomacromolecule govern biomimetic mineralization and showed that chemical functionalization can be used to control this process. Computational modelling verified the role of the negative charge on biomacromolecules in inducing ZIF growth as a result of enhancement of the surrounding zinc ion concentration. Furthermore, calculations of the surface electrostatic potential and pI of a protein were shown to accurately and efficiently predict whether a biomacromolecule seeds MOF growth. Finally, a high-throughput screening process was developed to explore enzymatic reaction space to discover candidate reactions for MOF-encapsulated enzymes. This screening process uses the molecular size of the components of a reaction to predict whether the reaction can occur inside MOFs. The number of possible enzymatic reactions that have been carried out inside ZIF-8 is very small, and many of those reactions were found to have components that are likely too big to diffuse through ZIF-8. Therefore, the screening process was applied to suggest reactions that can investigate the relationship between the size of reaction components and enzymatic activity inside ZIF-8. In this process, a reaction of significant commercial value was identified that should occur in a MOF-encapsulated enzyme.