Scientists at the University of Rochester report algorithms that reveal how nanoscale catalysts turn propane into propylene, a vital feedstock for many everyday products. The work builds on a 2021 Science study that showed tandem nanoscale catalysts can combine multiple process steps into a single reaction, a strategy that can increase yield and reduce cost. The new paper appears in the Journal of the American Chemical Society.
The algorithms identify the atomic features that drive the complex chemistry, addressing reactions that are complicated by materials coexisting in different states, such as metallic and oxide phases. The researchers used computational screening to narrow the large number of possible atomic arrangements and to focus on the most relevant sites at the catalyst surface.
Siddharth Deshpande and PhD student Snehitha Srirangam report an unexpected result: the oxide material tends to grow selectively around defective metal sites. This selective growth proved important for catalyst stability, and the oxide remained associated with those defective sites and kept its functional role despite variations in composition. The team suggests the approach is general and could help explain other industrial reactions, for example methanol synthesis, and help companies seek more efficient, less empirical production methods.
Difficult words
- algorithm — Computer method to solve a problem.algorithms
- nanoscale — Extremely small size at nanometer scale.
- catalyst — Material that speeds up a chemical reaction.catalysts
- feedstock — Raw material used to make products.
- tandem — Two things working together in sequence.
- oxide — A compound that contains oxygen and metal.
- stability — Ability to stay the same under change.
- screening — Process of testing many options to choose.
Tip: hover, focus or tap highlighted words in the article to see quick definitions while you read or listen.
Discussion questions
- How might selective oxide growth around defective metal sites improve catalyst stability in industrial production? Give reasons or examples from the article.
- What are the possible benefits and limitations of using algorithms and computational screening to design catalysts?
- If companies move away from empirical methods, how could that change their development of chemical processes?