Machine-learning-based selective sampling procedure for identifying the low-energy region in a potential energy surface: A case study on proton conduction in oxides

Toyoura, Kazuaki; Hirano, Daisuke; Seko, Akira; Shiga, M.; Kuwabara, Akihide; Karasuyama, Masayuki; Shitara, Kazuki; Takeuchi, Ichiro

doi:10.1103/physrevb.93.054112

Cited by 63 publications

(44 citation statements)

References 50 publications

(61 reference statements)

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…Examples can be found in the literature (e.g., Refs. [1][2][3][4]). Other candidates are simply a binary digit representing the presence of each element in a compound ( Fig.…”

Section: Compound Descriptorsmentioning

confidence: 99%

Descriptors for Machine Learning of Materials Data

2018

Self Cite

View full text Add to dashboard Cite

Descriptors, which are representations of compounds, play an essential role in machine learning of materials data. Although many representations of elements and structures of compounds are known, these representations are difficult to use as descriptors in their unchanged forms. This chapter shows how compounds in a dataset can be represented as descriptors and applied to machine-learning models for materials datasets.Keywords Machine-learning interatomic potential ⋅ Lattice thermal conductivity ⋅ Recommender system ⋅ Gaussian process ⋅ Bayesian optimization IntroductionRecent developments of data-centric approaches should accelerate the progress in materials science dramatically. Thanks to the recent advances in computational power and techniques, the results from numerous density functional theory (DFT) calculations with predictive performances have been stored as databases. A combination of such databases and an efficient machine-learning approach should realize prediction and classification models of target physical properties. Consequently, machine-learning techniques are becoming ubiquitous. They are used to explore materials and structures from a huge number of candidates and to extract meaningful information and patterns from existing data.A key factor in controlling the performance of a machine-learning approach is how compounds are represented in a data set. Representations of compounds are called "descriptors" or "features". To perform machine-learning modeling, available descriptors must be determined according to the evaluation cost of the target property

show abstract

“…Examples can be found in the literature (e.g., Refs. [1][2][3][4]). Other candidates are simply a binary digit representing the presence of each element in a compound ( Fig.…”

Section: Compound Descriptorsmentioning

confidence: 99%

Descriptors for Machine Learning of Materials Data

2018

Self Cite

View full text Add to dashboard Cite

show abstract

“…Consequently, Fig. 2.7 Rank correlation between the actual and the preliminary PEs [12]. Open circles and crosses show the grid points in P α and N α , respectively.…”

Section: Low-fn Region Identificationmentioning

confidence: 99%

“…2.4b). [12]. Yellow surface in each plot is the isosurface corresponding to the PE threshold at α = 0.…”

Section: Low-fn Region Identificationmentioning

confidence: 99%

“…The goal is to identify all αN grid points in the low-PE region as efficiently as possible. [12]. (b) Grid points at which the force norm acting on a proton is less than 0.2 eV/Å For simplicity, α is assumed to be prespecified.…”

Section: Problem Statementmentioning

confidence: 99%

“…This chapter introduces a novel selective sampling procedure for PES mapping based on a machine learning technique [12]. This sampling procedure preferentially evaluates a partial PES in the region of interest characterizing ionic conduction.…”

mentioning

confidence: 99%

See 2 more Smart Citations

Potential Energy Surface Mapping of Charge Carriers in Ionic Conductors Based on a Gaussian Process Model

Toyoura

Takeuchi

2018

Nanoinformatics

Self Cite

View full text Add to dashboard Cite

The potential energy surface (PES) of a charge carrier in a host crystal is an important concept to fundamentally understand ionic conduction. Such PES evaluations, especially by density functional theory (DFT) calculations, generally require vast computational costs. This chapter introduces a novel selective sampling procedure to preferentially evaluate the partial PES characterizing ionic conduction. This procedure is based on a machine learning method called the Gaussian process (GP), which reduces computational costs for PES evaluations. During the sampling procedure, a statistical model of the PES is constructed and sequentially updated to identify the region of interest characterizing ionic conduction in configuration space. Its efficacy is demonstrated using a model case of proton conduction in a well-known proton-conducting oxide, barium zirconate (BaZrO 3 ) with the cubic perovskite structure. The proposed procedure efficiently evaluates the partial PES in the region of interest that characterizes proton conduction in the host crystal lattice of BaZrO 3 .

show abstract

Protonic Conduction in La₂NiO₄₊_δ and La_2‐_xA_xNiO₄₊_δ (A = Ca, Sr, Ba) Ruddlesden–Popper Type Oxides

Zhong

Toyoura

Jiang

et al. 2022

Advanced Energy Materials

Self Cite

View full text Add to dashboard Cite

A significant hydration and protonic conduction in La2NiO4+δ‐based Ruddlesden–Popper (RP) oxides will enable their use as positrode materials in proton ceramic electrochemical cells (PCECs). Unlike perovskite electrolytes and positrode materials, where protons reside and jump on regular oxide ions, RP oxides such as La2NiO4+δ may contain oxygen interstitials and offer the possibility that protons locate on and migrate by help of these, suggesting a mechanism by which positrodes may operate by triple conduction. Here, theoretical calculations that support proton migration between interstitial oxide ions in the rock‐salt layer via unshared oxide ions in the perovskite layer are reported. Furthermore, water contents of pristine La2NiO4+δ and La2‐xAxNiO4+δ (A = Ca, Sr, Ba) are reported, showing that hydration increases with the level and basicity of the dopant. Hebb‐Wagner blocking electrode measurements show significant partial protonic conductivity in all samples. The effects of doping on hydration and of pH2O on the p‐type conductivity suggest that protons reside primarily on structural oxide ion, but a role of interstitial oxide ions for dissolution and migration of protons cannot be ruled out in cases with oxygen excess. The results are instructive for further studies and optimization of RP oxides as potential positrode materials for PCECs.

show abstract

Machine-learning-based selective sampling procedure for identifying the low-energy region in a potential energy surface: A case study on proton conduction in oxides

Cited by 63 publications

References 50 publications

Descriptors for Machine Learning of Materials Data

Descriptors for Machine Learning of Materials Data

Potential Energy Surface Mapping of Charge Carriers in Ionic Conductors Based on a Gaussian Process Model

Protonic Conduction in La₂NiO₄₊_δ and La_2‐_xA_xNiO₄₊_δ (A = Ca, Sr, Ba) Ruddlesden–Popper Type Oxides

Contact Info

Product

Resources

About

Machine-learning-based selective sampling procedure for identifying the low-energy region in a potential energy surface: A case study on proton conduction in oxides

Cited by 63 publications

References 50 publications

Descriptors for Machine Learning of Materials Data

Descriptors for Machine Learning of Materials Data

Potential Energy Surface Mapping of Charge Carriers in Ionic Conductors Based on a Gaussian Process Model

Protonic Conduction in La2NiO4+δ and La2‐xAxNiO4+δ (A = Ca, Sr, Ba) Ruddlesden–Popper Type Oxides

Contact Info

Product

Resources

About

Protonic Conduction in La₂NiO₄₊_δ and La_2‐_xA_xNiO₄₊_δ (A = Ca, Sr, Ba) Ruddlesden–Popper Type Oxides