Covalent Organic Frameworks: Structure, Filling, Nucleation

Abstract

Covalent organic frameworks, although only studied for ten years, have found their way into hydrogen/ammonia storage, catalysis, carbon dioxide adsorption, molecular sensing, solar cells, and charge storage. This diversity in application comes from the diversity in composition of covalent organic frameworks (COFs), where many structural components with many different functionalities can be predictably assembled into a rigid, porous network. This thesis covers a wide gamut of topics concerning COFs: nucleation and growth, stacking structure of the two-dimensional variants, and scaffolding qualities in complementary semiconducting devices. Our analysis of stacking in the class of experimental and as-yet unsynthe˚ sized boronate-ester COFs unveils 1 to 2.6 A offsets present between layers that adds disorder to the layer structure, corrugates the inner surface, affects charge transport through the layers, and reduces accessible volume. We have produced correlations between components used and degree of offset present to enable the rapid screening of starting materials for minimizing the offset when necessary for guest molecule diffusion, or maximizing the offset when needing to selectively adsorb molecules. Our simulated offset structures match experimental powder x-ray diffraction patterns in nine compositionally diverse COFs. We also completed a study on the potential for COF-based organic solar cells. Organic solar cells are plagued by many architectural inefficiencies, one of which is the requirement for a potential bias to help free tightly-bound excitedelectrons. In organic solar cells this potential bias comes from the interface between two complementary semiconducting molecules. COFs alleviate this burden by providing a scaffold to hold fullerene molecules in place to maximize the interfacial area between COF and fullerene. We use kinetic Monte Carlo to simulate the filling of crystalline phthalocyanine-COFs and report the effects of kinetic trapping that limit full loading capacity. We also share the implications on electron transport through the structurally disordered fullerene domains. Finally, we establish the foundation for understanding nucleation and growth of COFs in solution. For COFs to be realized in many applications, the crystalline domains must be larger and more defect-free. Understanding how solvent, temperature, linkage chemistry, and the substrate affects nucleation and growth are steps towards achieving that goal. We used free energy techniques to map out the reaction mechanisms and activation energies of three fundamental reactions of prototypical COF-5 involving the common catalyzing agents water and methanol. Our crystallization studies also conclusively eliminate certain proposed mechanisms of growth while strengthening the case for template polymerization as a likely growth mechanism for COF crystals.