Closing the carbon cycle through nanoscale engineering

The increasing levels of carbon dioxide (CO2) in our atmosphere has let to environmental challenges, such as global warming and ocean acidification. To combat these issues, researchers are exploring the conversion of CO2 into value-added fuels and chemicals by means of a process called catalytic hydrogenation of CO2. Using renewable hydrogen for the catalytic hydrogenation of CO2 is a practical approach to not only reduce emissions but also transition towards a more sustainable carbon economy. However, selectively converting CO 2 into valuable fuels and chemicals, such as lower olefins and alcohols, poses significant challenges: the high stability of CO2 and various competing reactions hinder the production of desired products, and traditional catalysts face limitations in selectivity and often are prone to deactivation. Due to the vast number of possible materials, the rational design of efficient catalysts for such complex reactions is generally unfeasible without clear guidance from theoretical calculations. These calculations are based on the principles of quantum mechanics and require powerful supercomputers, and they can provide valuable insights into the reaction mechanism at the molecular level. Then, if quantum mechanical calculations are combined with kinetic modelling simulations, it is possible to predict the catalytic activity and selectivity of novel materials, and, most importantly, unveil structure-activity relationships. In the quest to find novel catalysts, the recently discovered family of two-dimensional (2D) materials named MXenes, have garnered increasing attention for nearly a decade by virtue of their versatile composition and structure, stability under conditions of interest in heterogeneous catalysis, and numerous appealing properties. This project envisions understanding how MXenes can be tailored to enable and optimise their catalytic activity for the selective conversion of CO2 to value-added fuels and chemicals, including methanol, dimethyl ether, alkanes, lower olefins, and higher alcohols. Achieving this high-level aim will rely on using state-of-the-art quantum mechanical calculations and kinetic modelling simulations. Our study will identify which are the key features determining the selectivity of a CO2 hydrogenation catalyst towards a given product. Thus, the outcomes of the proposed project will have broader impact within a wide range of practical catalysis applications. Moreover, the results obtained will be discussed with experimental researchers at ETH Zurich (Switzerland) throughout the duration of the project, who will prepare and test novel MXene catalysts based on our findings and provide valuable data and feedback. Overall, the knowledge gained from this project will aid in the rational design and fabrication of more active and selective catalysts for CO 2 conversion, making the process more efficient.

MXenes are a relatively new family of two-dimensional materials made of extremely thin layers of metals and carbon or nitrogen. Their composition and surface chemistry can be tuned in many ways, making them promising for applications such as clean energy technologies, sustainable fuel production and environmental protection. However, this flexibility also makes them difficult to understand: the same MXene can behave very differently depending on the atoms attached to its surface, how its layers are stacked, or the material used to support it. This project investigated these effects in molybdenum-based MXenes by combining experimental measurements with computer simulations. The first study examined what happens when the MXene is supported on silica, a common catalyst support. The results showed that silica restricts structural changes in the MXene and helps retain oxygen-containing groups on its surface. This changes the chemical reactions taking place and the products that are formed. In particular, the supported material produced more oxygen-containing compounds, including methanol and dimethyl ether, whereas the unsupported material mainly produced hydrocarbons. These results show that the support is not simply an inactive carrier but can directly influence catalytic performance. The second study focused on unusual molybdenum atoms with an oxidation state of 5+, which are frequently observed experimentally but whose atomic structure was previously unknown. By combining advanced experimental measurements with computer simulations of different structural models, we identified the most likely arrangement of these atoms. This provides a more realistic picture of the catalyst surface and improves the computer models used to predict catalytic behaviour. The third study investigated the space between stacked MXene layers, which determines whether molecules can reach the internal surfaces of the material. The simulations showed that completely clean layers collapse together, while small amounts of hydrogen keep the layered structure intact but leave almost no usable space between the sheets. Other surface atoms and trapped water maintain a more open structure. The project also produced a machine-learning model that makes it possible to simulate much larger systems over much longer times than was previously possible. Together, these results show how the support, surface chemistry and layered structure determine the behaviour of MXene catalysts. This knowledge will help guide the development of more efficient catalysts for producing fuels and chemicals from resources such as carbon dioxide and hydrogen.