A universal graph deep learning interatomic potential for the periodic table

Chen, Chi; Ong, Shyue Ping

doi:10.1038/s43588-022-00349-3

Cited by 235 publications

(240 citation statements)

References 55 publications

Supporting

Mentioning

164

Contrasting

Unclassified

Order By: Relevance

“…In an independent but related effort, Chen and Ong predicted the stability of 31 million hypothetical structures using the M3gnet framework. 39 Of our 285 DFT-confirmed compositions, 102 appear in the matterverse.ai web platform, with 47 of these predicted to have a negative decomposition energy. This agreement between methods with separate training data and computational approaches lends further evidence to the potential stability of these materials.…”

Section: Discussionmentioning

confidence: 93%

“…Our method reveals 285 novel structures, many of which appear reasonable when compared with structures currently being explored for this application. In an independent but related effort, Chen and Ong predicted the stability of 31 million hypothetical structures using the M3gnet framework . Of our 285 DFT-confirmed compositions, 102 appear in the matterverse.ai web platform, with 47 of these predicted to have a negative decomposition energy.…”

Section: Discussionmentioning

confidence: 99%

See 1 more Smart Citation

Upper-Bound Energy Minimization to Search for Stable Functional Materials with Graph Neural Networks

Law

Pandey

Gorai

et al. 2022

JACS Au

View full text Add to dashboard Cite

The discovery of new materials in unexplored chemical spaces necessitates quick and accurate prediction of thermodynamic stability, often assessed using density functional theory (DFT), and efficient search strategies. Here, we develop a new approach to finding stable inorganic functional materials. We start by defining an upper bound to the fully relaxed energy obtained via DFT as the energy resulting from a constrained optimization over only cell volume. Because the fractional atomic coordinates for these calculations are known a priori, this upper bound energy can be quickly and accurately predicted with a scale-invariant graph neural network (GNN). We generate new structures via ionic substitution of known prototypes, and train our GNN on a new database of 128 000 DFT calculations comprising both fully relaxed and volume-only relaxed structures. By minimizing the predicted upper-bound energy, we discover new stable structures with over 99% accuracy (versus DFT). We demonstrate the method by finding promising new candidates for solid-state battery (SSB) electrolytes that not only possess the required stability, but also additional functional properties such as large electrochemical stability windows and high conduction ion fraction. We expect this proposed framework to be directly applicable to a wide range of design challenges in materials science.

show abstract

Section: Discussionmentioning

confidence: 93%

Section: Discussionmentioning

confidence: 99%

Upper-Bound Energy Minimization to Search for Stable Functional Materials with Graph Neural Networks

Law

Pandey

Gorai

et al. 2022

JACS Au

View full text Add to dashboard Cite

show abstract

“…We note that the IS2RE task can also be performed indirectly via initial structure to relaxed structure (IS2RS) followed by structure to energy and force (S2EF). , While direct IS2RE using D-CGCNN does not provide “pseudo” relaxed structures, the indirect approach does by optimizing initial structures using numerical optimization algorithms, such as Bayesian optimization (BO) or Broyden–Fletcher–Goldfarb–Shanno (BFGS) . We compared two approaches in terms of efficiency and accuracy.…”

Section: Discussionmentioning

confidence: 99%

Direction-Based Graph Representation to Accelerate Stable Catalyst Discovery

2022

View full text Add to dashboard Cite

To realize a renewable and sustainable energy cycle, there has been a lot of effort put into discovering catalysts with desired properties from a large chemical space. To achieve this goal, several screening strategies have been proposed, most of which require validation of thermodynamic stability and synthesizability of candidate materials via computationally intensive quantum chemistry or solid-state physics calculations. This problem can be overcome by reducing the number of calculations through machine learning methods, which predict target properties using unrelaxed crystal structures as inputs. However, numerical input representations of most of the previous models are based on either too specific (e.g., atomic coordinates) or too ambiguous (e.g., stoichiometry) information, practically inapplicable to energy prediction of unrelaxed initial structures. In this work, we develop a direction-based crystal graph convolutional neural network (D-CGCNN) with the highest accuracy toward formation energy predictions of the relaxed structures using the initial structures as inputs. By comparing with other approaches, we revealed correlations between crystal graph similarities and model performances, elucidating the origin of the improved accuracy of our model. We applied this model to the ongoing high-throughput virtual screening project, where the model discovered 1,725 stable materials from 15,318 unrelaxed structures by performing 3,966 structure optimizations (∼25%).

show abstract

“…Their benchmarking on the OC20 [6] dataset and lower accuracy requirements suggest that the approach could be generalizable across a wide-class of material systems and thus significantly expand the availability of structural descriptors. Similarly, Chen et al demonstrated that a variant of MEGNET could perform high fidelity relaxations of unseen materials with diverse chemistries and that leveraging the resulting structures could improve downstream ML predictions of energy when compared with unrelaxed inputs [165]. The strong performance of these approaches and their potential to significantly increase the scale and effectiveness of computational screening motivates high-value research questions concerning the scale of data sets required for training, the generalizabiltiy over material classes, and the applicability to prediction tasks beyond stability.…”

Section: Prediction From Unrelaxed Crystal Prototypesmentioning

confidence: 99%

Representations of Materials for Machine Learning

Damewood¹,

Karaguesian²,

Lunger³

et al. 2023

Preprint

View full text Add to dashboard Cite

High-throughput data generation methods and machine learning (ML) algorithms have given rise to a new era of computational materials science by learning relationships among composition, structure, and properties and by exploiting such relations for design. However, to build these connections, materials data must be translated into a numerical form, called a representation, that can be processed by a machine learning model. Datasets in materials science vary in format (ranging from images to spectra), size, and fidelity. Predictive models vary in scope and property of interests. Here, we review context-dependent strategies for constructing representations that enable the use of materials as inputs or outputs of machine learning models. Furthermore, we discuss how modern ML techniques can learn representations from data and transfer chemical and physical information between tasks. Finally, we outline high-impact questions that have not been fully resolved and thus, require further investigation.

show abstract

A universal graph deep learning interatomic potential for the periodic table

Cited by 235 publications

References 55 publications

Upper-Bound Energy Minimization to Search for Stable Functional Materials with Graph Neural Networks

Upper-Bound Energy Minimization to Search for Stable Functional Materials with Graph Neural Networks

Direction-Based Graph Representation to Accelerate Stable Catalyst Discovery

Representations of Materials for Machine Learning

Contact Info

Product

Resources

About