Discovering surface reaction pathways using accelerated molecular dynamics and network analysis tools.

Area of Science:

• Computational Chemistry
• Surface Science
• Catalysis

Background:

• Mapping surface reaction pathways is crucial for understanding catalysis.
• Traditional methods often require extensive experimental data and human intervention.
• Accurate prediction of reaction mechanisms at the atomic level is challenging.

Purpose of the Study:

• To develop an automated computational method for mapping surface reaction pathways.
• To enable pathway discovery with minimal experimental data and human input.
• To apply the method to industrially relevant catalytic reactions.

Main Methods:

• Utilized bias potentials to enhance sampling of surface reactions within ab initio molecular dynamics (MD) simulations.
• Developed an automated procedure to extract elementary reactions from MD trajectories.
• Employed network analysis tools to map extracted reaction data onto pathways.

Main Results:

• Successfully mapped surface reaction pathways for methane steam reforming on Rh(111) and propane reforming on Pt(111) and Pt3Sn(111).
• Discovered novel, energetically favorable reaction pathways for both investigated reactions.
• Reproduced experimentally observed trends in surface reaction activity and selectivity, including material dependence.

Conclusions:

• The automated method provides an efficient and data-minimal approach for elucidating surface reaction mechanisms.
• The findings offer new insights into catalytic processes and can guide catalyst design.
• This computational strategy significantly advances the study of heterogeneous catalysis.