Artificial neural network potential for gold clusters*

Ling-zhi, Cao; Wang, Pengju; Sai, Linwei; Fu, Jie; Duan, Xiangmei

doi:10.1088/1674-1056/abc15d

Cited by 6 publications

(7 citation statements)

References 58 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…We choose radial basis function of G 2 type and angle basis function of G 4 type [37]. The details about ANN settings can be found elsewhere [28].…”

Section: Artificial Neural Network Potentialmentioning

confidence: 99%

“…Thorn et al benchmarked an ANN potential against a Gupta potential and an embedded atom model in the search of stable Au n clusters (30 n 80) to accelerate the ground-state structures (GSSs) search by firstprinciples [27]. Cao et al also developed ANN potential for Au 11-100 based on DFT data and test its accuracy by calculating energy and force of structures from previous reports [28]. As dedicated by the authors, ANN trained for a wide range of size is highly demanded by the applications, but now, its accuracy is coarse especially in the deal with competitive local energyminimum isomers.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Artificial neural network potential for Au₂₀ clusters based on the first-principles

Ling-zhi¹,

Guo²,

Han³

et al. 2022

J. Phys.: Condens. Matter

Self Cite

View full text Add to dashboard Cite

The search of ground-state structures (GSS) of gold (Au) clusters is a formidable challenge due to the complexity of potential energy surface (PES). In this work, we have built a high-dimensional artificial neural network (ANN) potential to describe the PES of Au20 clusters. The ANN potential is trained through learning the GSS search process of Au20 by the combination of density functional theory (DFT) method and genetic algorithm (GA). The root mean square errors (RMSE) of energy and force are 7.72 meV/atom and 217.02 meV/Å, respectively. As a result, it can find the lowest-energy structure (LES) of Au20 clusters that is consistent with previous results. Furthermore, the scalability test shows that it can predict the energy of smaller size Au16-19 clusters with errors less than 22.85 meV/atom, and for larger size Au21-25 clusters, the errors are below 36.94 meV/atom. Extra attention should be paid to its accuracy for Au21-25 clusters. Applying the ANN to search the GSS of Au16-25, we discover two new structures of Au16 and Au21 that are not reported before and several candidate LESs of Au16-18. In summary, this work proves that an ANN potential trained for specific size clusters could reproduce the GSS search process by DFT and be applied in the GSS search of smaller size clusters nearby. Therefore, we claim that building ANN potential based on DFT data is one of the most promising ways to effectively accelerate the GSS pre-screening of clusters.

show abstract

“…We choose radial basis function of G 2 type and angle basis function of G 4 type [37]. The details about ANN settings can be found elsewhere [28].…”

Section: Artificial Neural Network Potentialmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Artificial neural network potential for Au₂₀ clusters based on the first-principles

Ling-zhi¹,

Guo²,

Han³

et al. 2022

J. Phys.: Condens. Matter

Self Cite

View full text Add to dashboard Cite

show abstract

“…Alternatively, the emerging machine learning or deep learning (DL) techniques have provided a solution for rapid prediction of the energies of isomers for a molecule or cluster with precision comparable to DFT. − For instance, using a geom-C60 database with four symmetric cage isomers and 29 unique C–C bonds, Aghajamali and Karton examined the performance of 12 carbon force fields and found that a machine-learning-based Gaussian approximation potential, namely, GAP-20, outperforms the empirical force fields. In addition to binding energies, Calvo et al created a large database of 753,184 infrared spectra of C n clusters ( n = 24, 33, 42, 52, 60) with different shapes (including fullerene-like cages, graphene-like flakes, pretzel-like and branched structures) using density functional-based tight-binding calculations and developed an interpolation scheme to reproduce the spectral features by encoding the structures using appropriate descriptors and selecting them through principal component analysis and Gaussian regression.…”

Section: Introductionmentioning

confidence: 99%

Predicting Binding Energies and Electronic Properties of Boron Nitride Fullerenes Using a Graph Convolutional Network

Sai,

Fu,

Zhao

2023

J. Chem. Inf. Model.

View full text Add to dashboard Cite

As isoelectronic counterparts of carbon fullerenes, medium-sized boron nitride clusters also prefer cage structures composed of even-sized polygons. As the cluster size increases, the number of cage isomers grows rapidly, and determining the ground state structure requires a tremendous amount of DFT calculations. Herein, we develop a graph convolutional network (GCN) that can describe the energy of a (BN) n cage by its topology connection. We define a vertex feature vector on a dual polyhedron by the permutation of the neighbor vertices' degree and aggregate the information on vertices by two graph convolutional layers to learn the local feature of the dual polyhedron. The GCN is trained on (BN) 28 and subsequently tested on (BN) 23 and (BN) 24 data sets, which satisfactorily reproduce the order of isomer energies from DFT calculations. We further employ the trained GCN to predict the ground state structures within the size range of n = 25−32, which agree well with DFT results. Using the same GCN framework, we also successfully trained the highest-occupied or lowest-unoccupied orbital energies of (BN) 28 isomers. The present graph convolutional network establishes a direct mapping between the topological connection and the energetic or electronic properties of a cage-like cluster or molecule.

show abstract

“…Specifically, it has been used to simulate the properties of silicon with accuracy comparable to DFT, and efficiency comparable to empirical potentials. For example, Cao et al [15,27] learned the PES of Au 20 clusters by training an artificial neural network (ANN) potential to DFT data, which can give out binding energy within DFT accuracy. Ouyang et al have found a new Au 58 cluster by the combination of neural network potential and basinhopping method [28].…”

Section: Introductionmentioning

confidence: 99%

Revisiting the stable structures of gold clusters: Au _n (n = 16–25) by artificial neural network potential

Guo

2023

J. Phys. D: Appl. Phys.

Self Cite

View full text Add to dashboard Cite

Identifying the stable structures of gold (Au) clusters is a huge challenge in cluster science. In this work, we have searched the ground-state structures of neutral Aun (n = 16 – 25) clusters by an artificial neural network (ANN) potential trained to density functional theory (DFT) data. Compared with the DFT data, the root mean square error of binding energy predicted by the ANN potential is about 8.66 meV/atom. Applying the ANN potential to search the ground-state structures by comprehensive genetic algorithm, we have found several new candidates of Au18, Au22, and Au23, which are not reported before. Au18 is hollow cage structure, Au22 and Au23 are flat cage structures. From the electronic analysis, we elucidate the stability mechanism of the newly found structures that are associated with the electronic shell closure of superatomic orbitals. Besides, we also clarified how to clean a database to train an efficient ANN potential in details. Overall, this work proves that applying machine learning in the description of atomic interactions can accelerate the searching of ground state structures of clusters and help to find new candidates of stable cluster structures.

show abstract

Artificial neural network potential for gold clusters*

Cited by 6 publications

References 58 publications

Artificial neural network potential for Au₂₀ clusters based on the first-principles

Artificial neural network potential for Au₂₀ clusters based on the first-principles

Predicting Binding Energies and Electronic Properties of Boron Nitride Fullerenes Using a Graph Convolutional Network

Revisiting the stable structures of gold clusters: Au _n (n = 16–25) by artificial neural network potential

Contact Info

Product

Resources

About

Artificial neural network potential for gold clusters*

Cited by 6 publications

References 58 publications

Artificial neural network potential for Au20 clusters based on the first-principles

Artificial neural network potential for Au20 clusters based on the first-principles

Predicting Binding Energies and Electronic Properties of Boron Nitride Fullerenes Using a Graph Convolutional Network

Revisiting the stable structures of gold clusters: Au n (n = 16–25) by artificial neural network potential

Contact Info

Product

Resources

About

Artificial neural network potential for Au₂₀ clusters based on the first-principles

Artificial neural network potential for Au₂₀ clusters based on the first-principles

Revisiting the stable structures of gold clusters: Au _n (n = 16–25) by artificial neural network potential