Understanding materials plays a crucial role in our highly advanced society. Zeolitic imidazolate frameworks (ZIFs) are a very promising class of materials for applications such as catalysis, sensing, and gas separation. However, for industrial applications, it is important to have an in-depth understanding of these materials to achieve the highest productivity.
This dissertation investigates the transport properties of C2 species in nanostructured materials for separation. For this, we use computational methods to model the movement of the guest molecule and investigate the impact of the host material on the guest molecule motion down to the atomic level. A major breakthrough in modeling is the use of machine learning models, based on neural networks, which allow a significant speed-up while keeping high accuracy.
This research offers new insights into the impact of ZIF-8 on ethane and ethene diffusion, with the decomposition method demonstrating the entropic hindrance of ethane and ethene diffusion imposed by the host material. The newly developed method supports the design and development of more applicationspecific materials for separation purposes.