Efficient Screening of Metal Promoters of Pt Catalysts for C-H Bond Activation in Propane Dehydrogenation from a Combined First-Principles Calculations and Machine-Learning Study

Platinum-based materials are the most widely used catalystsinpropane direct dehydrogenation, which could achieve a balanced activitybetween both propane conversion and propene formation. One of thecore issues of Pt catalysts is how to efficiently activate the strongC-H bond. It has been suggested that adding second metal promoterscould greatly solve this problem. In the current work, first-principlescalculations combined with machine learning are performed in orderto obtain the most promising metal promoters and identify key descriptorsfor control performance. The combination of three different modesof adding metal promoters and two ratios between promoters and platinumsufficiently describes the system under investigation. The activityof propane activation and the formation of propene are reflected bythe increase or decrease of the adsorption energy and C-H bondactivation of propane and propene after the addition of promoters.The data of adsorption energy and kinetic barriers from first-principlescalculations are streamed into five machine-learning methods includinggradient boosting regressor (GBR), K neighbors regressor (KNR), randomforest regressor (RFR), and AdaBoost regressor (ABR) together withthe sure independence screening and sparsifying operator (SISSO).The metrics (RMSE and R (2)) from differentmethods indicated that GBR and SISSO have the most optimal performance.Furthermore, it is found that some descriptors derived from the intrinsicproperties of metal promoters can determine their properties. In theend, Pt3Mo is identified as the most active catalyst. Thepresent work not only provides a solid foundation for optimizing Ptcatalysts but also provides a clear roadmap to screen metal alloycatalysts.

Automated summary

Platinum-based materials are widely used in propane direct dehydrogenation as catalysts for achieving balanced activity. This study combines first-principles calculations and machine learning to identify promising metal promoters and key descriptors for controlling performance.