On the importance of catalyst-adsorbate 3D interactions for relaxed energy predictions - Inria - Institut national de recherche en sciences et technologies du numérique Access content directly
Conference Papers Year : 2023

On the importance of catalyst-adsorbate 3D interactions for relaxed energy predictions

Abstract

The use of machine learning for material property prediction and discovery has traditionally centered on graph neural networks that incorporate the geometric configuration of all atoms. However, in practice not all this information may be readily available, e.g. when evaluating the potentially unknown binding of adsorbates to catalyst. In this paper, we investigate whether it is possible to predict a system's relaxed energy in the OC20 dataset while ignoring the relative position of the adsorbate with respect to the electro-catalyst. We consider SchNet, DimeNet++ and FAENet as base architectures and measure the impact of four modifications on model performance: removing edges in the input graph, pooling independent representations, not sharing the backbone weights and using an attention mechanism to propagate non-geometric relative information. We find that while removing binding site information impairs accuracy as expected, modified models are able to predict relaxed energies with remarkably decent MAE. Our work suggests future research directions in accelerated materials discovery where information on reactant configurations can be reduced or altogether omitted.
Fichier principal
Vignette du fichier
76_on_the_importance_of_catalyst_.pdf (331.55 Ko) Télécharger le fichier
Origin : Files produced by the author(s)

Dates and versions

hal-04407906 , version 1 (21-01-2024)

Licence

Attribution

Identifiers

  • HAL Id : hal-04407906 , version 1

Cite

Alvaro Carbonero, Alexandre Duval, Victor Schmidt, Santiago Miret, Alex Hernandez-Garcia, et al.. On the importance of catalyst-adsorbate 3D interactions for relaxed energy predictions. AI4MAt 2023 - workshop on AI for Accelerated Design of the 37th Conference on Neural Information Processing Systems (NeurIPS), Dec 2023, New Orleans (Louisiana), United States. ⟨hal-04407906⟩
10 View
7 Download

Share

Gmail Facebook X LinkedIn More