Determining Covalent Organic Framework Structures Using Electron Crystallography and Computational Intelligence

The structural characterization of new materials often poses immense challenges, especially when obtaining single-crystal structures is difficult, which is a common difficulty with covalent organic frameworks (COFs). Despite this, understanding the atomic structure is crucial as it provides insights into the arrangement and connectivity of organic building blocks, offering the opportunity to establish the correlation of structure-function relationships and unravel material properties. In this study, we present an approach for determining the structures of COFs, an integration of electron crystallography and computational intelligence (COF+). By applying established chemistry knowledge and employing particle swarm optimization (PSO) for trial structure generation, we overcome existing limitations, thus paving the way for advancements in COF structural determination. We have successfully implemented this technique on four representative COFs, each with unique characteristics. These examples underline the accuracy and efficacy of our approach in addressing the challenges tied to COF structural determination. Furthermore, our approach has revealed new structure candidates with different topologies or interpenetrations that are chemically feasible. This discovery demonstrates the capability of our algorithm in constructing trial COF structures without being influenced by topological factors. Our new approach to COF structure determination represents a significant advancement in the field and opens new avenues for exploring the properties and applications of COF materials.