Recommended by

Indexed by

Adsorbate chemical environment-based machine learning framework for heterogeneous catalysis

Pushkar Ghanekar1*, Siddharth Deshpande1,2*, Jeffrey Greeley1*

1 Davidson School of Chemical Engineering, Purdue University, West Lafayette, Indiana 47907, USA

2 Department of Chemical Engineering, University of Delaware, Newark, DE

* Corresponding authors emails: pghaneka@purdue.edu, sdeshpan@udel.edu, jgreeley@purdue.edu
DOI10.24435/materialscloud:td-hf [version v1]

Publication date: Apr 11, 2022

How to cite this record

Pushkar Ghanekar, Siddharth Deshpande, Jeffrey Greeley, Adsorbate chemical environment-based machine learning framework for heterogeneous catalysis, Materials Cloud Archive 2022.50 (2022), doi: 10.24435/materialscloud:td-hf.


Heterogeneous catalytic reactions are influenced by a subtle interplay of atomic-scale factors, ranging from the catalysts’ local morphology to the presence of high adsorbate coverages. Describing such phenomena via computational models requires generation and analysis of a large space of surface atomic configurations. To address this challenge, we present the Adsorbate Chemical Environment-based Graph Convolution Neural Network (ACE-GCN), a screening workflow that can account for atomistic configurations comprising diverse adsorbates, binding locations, coordination environments, and substrate morphologies. Using this workflow, we develop catalyst surface models for two illustrative systems: (i) NO adsorbed on a Pt3Sn(111) alloy surface, of interest for nitrate electroreduction processes, where high adsorbate coverages combine with the low symmetry of the alloy substrate to produce a large configurational space, and (ii) OH* adsorbed on a stepped Pt(221) facet, of relevance to the Oxygen Reduction Reaction, wherein the presence of irregular crystal surfaces, high adsorbate coverages, and directionally-dependent adsorbate-adsorbate interactions result in the configurational complexity. In both cases, the ACE-GCN model, having trained on a fraction (~10%) of the total DFT-relaxed configurations, successfully ranks the relative stabilities of unrelaxed atomic configurations sampled from a large configurational space. This approach is expected to accelerate development of rigorous descriptions of catalyst surfaces under in-situ conditions.

Materials Cloud sections using this data

No Explore or Discover sections associated with this archive record.


File name Size Description
5.6 KiB File containing description of the data included in the folders and how to access it.
84.4 KiB Raw data to recreate analysis of Fig 3
28.2 KiB Raw data to recreate analysis of Fig 4
2.0 KiB File to generate Fig 5
11.5 KiB processing script to generate graph objects
1.4 GiB Raw files (graph objects and atom trajectories) for Pt-OH analysis
378.6 MiB Raw files (graph objects and atom trajectories) for Pt3Sn-NO analysis


Files and data are licensed under the terms of the following license: MIT License.
Metadata, except for email addresses, are licensed under the Creative Commons Attribution Share-Alike 4.0 International license.


density-functional theory machine learning electronic structure

Version history:

2022.50 (version v1) [This version] Apr 11, 2022 DOI10.24435/materialscloud:td-hf