Modeling combinatorial complexity in hydrocarbon catalysis
1st promotor and 2nd supervisor: Prof. Emiel Hensen
1st supervisor and co-promotor: Assistant Prof. Ivo Filot
Affiliation: Eindhoven University of Technology
Research theme: Storing electricity from renewable sources in chemical bonds
To develop next-generation catalytic materials for hydrocarbon processing, it is crucial that we develop a fundamental understanding of the underlying reaction mechanisms involved. Despite significant advances in computational power and modeling algorithms, there remain many challenges in hydrocarbon catalysis which are fundamentally difficult to model. The underlying principle of modeling is reducing the complexity of a system enabling the researcher to focus on the essentials, yet this complexity reduction can be severely hampered if any observable of interest can only be successfully calculated by averaging over many possible configurations with equal likeliness.
Two notable cases are lateral interaction in Fischer-Tropsch synthesis and the entropy sampling in high-temperature hydrocarbon conversion in zeolites. These systems both suffer from the principle of combinatorial complexity; the former due to the large number of adlayer configurations possible and the latter due to the large number of relevant configurational states within the confinement of the zeolite pore. Within this project, we aim to develop a new modeling procedure based on artificial neural networks that effectively targets the combinatorial complexity inherent in these systems.
- Computational Catalysis
- Artificial Neural Networks