DOI: https://doi.org/10.29363/nanoge.interect.2022.009
Publication date: 11th October 2022
Using renewable energy to convert carbon dioxide to fuels and chemical is a goal pursued worldwide, but neither electro- nor photocatalytic processes have reached a maturity that would invite their application on the industrial scale. Performance, stability, and selectivity issues occur in either process option to different extents. Although progress is reported in the scientific literature almost daily, fundamental insight is still lacking that would allow a knowledge-driven improvement of photo- and electrocatalysts.
As will be shown in the current contribution, systematic studies in photocatalysis are severely hindered by the non-standardized and often irreproducible reaction conditions. Therefore, standardization approaches were undertaken that use, as a prerequisite, high-purity reaction conditions free of any (carbon-)impurity. Using, in addition, reaction engineering tools known from classical catalysis allows to determine the interdependence of yields and selectivity on the reaction conditions. This is not only a first step towards an understanding of the reaction fundamentals, but also builds a bridge towards machine learning approaches for (photo)catalyst improvement.
It will be shown that despite its absorption in the UV range, titania is still the best performing semiconductor for photocatalytic carbon dioxide reduction. However, oxidation and reduction reaction are inherently coupled because the pathway from carbon dioxide to methane on the titania surface also contains oxidative elementary steps involving photogenerated holes. On bare titania, methane and oxygen thus cannot be formed at the same time. Other approaches are needed in which new catalytic active sites allow a continuous reduction sequence. Furthermore, it will be shown how reduction and oxidation reaction can be separated onto different semiconductors in Z scheme and heterojunction systems, which also allows the development of switchable photocatalyst composites.