A structural similarity based data-mining algorithm for modeling multi-reactant heterogeneous catalysts.

Area of Science:

• Computational Chemistry
• Materials Science
• Surface Science

Background:

• Density Functional Theory (DFT) is crucial for heterogeneous catalyst research.
• High computational expense and vast configuration spaces limit DFT's application in complex catalytic systems.
• Understanding multi-reactant catalysis on diverse surfaces requires efficient simulation methods.

Purpose of the Study:

• To develop and apply an innovative similarity algorithm for quantifying structural differences between atomic configurations.
• To reduce the number of computationally expensive DFT simulations needed for heterogeneous catalysis research.
• To enable efficient data mining of configurational phase-space for identifying stable atomic models.

Main Methods:

• Development of a novel similarity algorithm to quantify structural differences in atomic configurations.
• Application of the algorithm to identify structurally dissimilar configurations with minimal human intervention.
• Utilizing the algorithm to significantly reduce DFT simulations for multi-reactant catalysis studies.

Main Results:

• The similarity algorithm effectively identifies unique atomic configurations, drastically reducing the required DFT simulations.
• Demonstrated application in studying CO*-OH* co-adsorption on a stepped Pt surface, simulating only 2% of configurations.
• Showcased versatility by analyzing bidentate adsorption on a stepped Pt surface.

Conclusions:

• The developed similarity algorithm is a powerful tool for accelerating heterogeneous catalysis research.
• This approach significantly lowers computational barriers in studying complex multi-reactant systems.
• The method serves as a vital step towards understanding key multi-reactant heterogeneous catalytic chemistries.