< Back to previous page

Project

Design of Fe-based Catalysts Using Machine Learning Potentials to Recycle CO2 to Jet Fuel

The energy system of the future will be based on renewable energy, yet carbon-based materials will remain important. Seasonal energy storage and long-distance transport will rely on e-fuels made from recycled CO2 and climate-neutral H2. While the conversion of CO to jet fuel is one of the largest catalytic processes (Fischer-Tropsch Synthesis, CO-FT), the direct conversion of CO2 to jet fuel (CO2-FT) suffers from poor selectivity and low carbon efficiencies. Direct conversion of CO2 to jet fuel requires a careful balance of sites for CO2 activation and chain growth. Promoted Fe-based catalysts have shown the most promise, but Co-based FT catalysts might also be modified for CO2-FT by introducing suitable promotors. At the LCT, a detailed model for the structure and the kinetics for Co-based catalyst has been developed using DFT calculations. The structural and phase complexity of promoted Fe-based catalysts has hampered a similar level of understanding. In this project, we will use simulations based on machine learning potentials (MLPs) to model the structural complexity of promoted Fe-based catalysts and identify potential active sites. The activity of those sites will then be modeled by DFT. Based on such structure-activity relationships, carbon-efficient Fe-based catalysts will be designed to be experimentally tested at the LCT.

Date:1 Nov 2022 →  Today
Keywords:Machine Learning Potentials, CO2, Catalysis
Disciplines:Chemical kinetics and thermodynamics, Theoretical and computational chemistry not elsewhere classified, Heterogeneous catalysis, Modelling, simulation and optimisation, Carbon capture engineering