The introduction of data science in catalysis is about to fundamentally change how catalysts are designed. The idea of catalysts informatics is to unveil the trends and patterns in catalysts data using data science techniques such as machine learning and data mining. Here, the concept of catalysts informatics along with two case studies of synthesis of catalysts in oxidative coupling of methane reaction is presented. Machine learning and deep learning are implemented to design catalysts for the oxidative coupling of methane (OCM) reaction. In particular, OCM dataset is prepared by high throughput experiments where greater than 12,000 OCM data is generated, which is approximately 5 times more than previously collected OCM literature data. Combined with machine learning and deep learning, several unreported catalysts are designed where those catalysts are found to have high performance in experimental confirmation. In addition, low temperature active OCM catalysts are designed via data mining. In particular, data mining identifies the common physical rules among the low temperature active OCM catalysts. Based on the discovered rules, unreported low temperature active OCM is designed and confirmed in experiment. Thus, catalysts informatics can potentially change how catalysts are explored and designed.

Design and synthesis of catalysts in oxidative coupling of methane via catalysts informatics

ACS Spring Meeting 2021

The official archive and repository for all ACS Meetings & Expositions abstracts and select posters or presentations.