Chi’s Talk on Constructing Accurate Quantitative Structure-Property Relationships via Materials Graph Networks

Chi Chen gave a talk at nanoHub’s Hands-on Data Science and Machine Learning Training Series on how to develop MatErials Graph Network (MEGNet) models for predicting various materials properties from crystal structure. He also demonstrates how the MEGNet framework can be adapted to work with multi-fidelity data sources to improve predictions on high-value small datasets (e.g., experimental data). Extensive examples are shown using Jupyter notebooks. The video is available on the Materials Virtual Lab Youtube Channel.

The megnet package used extensively in these tutorials can be found on Github.

Yunxing’s Workshop Talk on Machine Learning Interatomic Potential Development with MAML

Yunxing gave a talk at NanoHUB’s Hands-on Data Science and Machine Learning Training Series today on how to conveniently develop machine learning interatomic potentials (ML-IAPs) using the Materials Machine Learning (maml) library. ML-IAPs describe the potential energy surface using local environment descriptors and has been demonstrated to be able to achieve near-DFT accuracy with linear scaling with respect to the number of atoms. The recording of this talk is now available on the Materials Virtual Lab’s Youtube channel.

To find out more about the maml package, check out our Github repository. You can also read Yunxing’s excellent paper benchmarking the performance and cost of various ML-IAPs to learn more.

Machine learning properties of ordered and disordered materials

Our paper on “Learning properties of ordered and disordered materials from multi-fidelity data” has just been published in the inaugural issue of Nature Computational Science! In this work, we address two major impediments to ML for materials science. The first impediment is that valuable accurate data is much more expensive to obtain than less accurate data. Using multi-fidelity materials graph networks (MEGNet), we show that we can use the lower quality data to improve underlying structural representations in models, and in the process significantly improve predictions on smaller, more valuable data (e.g., experimental measurements). The second impediment is that making predictions on disordered materials, which is the vast majority of known materials, is much more difficult than on ordered materials. We show that the elemental representations (embeddings) learned by our MEGNet models can be used to directly model disordered materials.

The article is available here. For an independent perspective on the findings, check out the Nature News & Views article.

All data and code are available from http://crystals.ai and the Github repository.

Design Principles for Cation-Mixed Sodium Solid Electrolytes

Zhuoying’s paper on “Design Principles for Cation-Mixed Sodium Solid Electrolytes” is the first publication from the Materials Virtual Lab in 2021! All-solid-state sodium-ion batteries are highly promising for next-generation grid energy storage with improved safety. Published in Advanced Energy Materials, this work develops design rules for the highly promising family of cation-mixed Na superionic conductors Na3PnS4-Na4TtS4. We show that cation mixing results in the “worst of both worlds” in terms of electrochemical stability, but can potentially lead to improved ionic conductivity and moisture stability. In particular, the recently reported Na11Sn2PnS12 superionic conductors are shown to be stable and Na11Sn2AsS12 is identified as a hitherto unexplored stable sodium superionic conductor with higher Na + conductivity and better moisture stability than those already reported experimentally.

Check out the work here.

High-entropy Na-ion cathode

Our joint publication with Prof Hailong Chen’s group on “Multiprincipal Component P2-Na0.6(Ti0.2Mn0.2Co0.2Ni0.2Ru0.2)O2 as a High-Rate Cathode for Sodium-Ion Batteries” has just been published in JACS Au. In this work, we extended the “high-entropy” concept in metal alloys and ceramics to layered oxide cathode materials, which stablizes the crystal structure and enhances diffusion. Chi Chen from our group showed using AIMD calculations and NEB calculations that the high-entropy concept leads to a percolating network of low barrier pathways for fast, macroscopic Na diffusion, resulting in the observed high rate performance.

Check out the work here.

Joule & Energy and AI Symposium

Prof Ong gave a talk at the Joule & Energy and AI Joint Online Symposium on our recent work on “Multi-fidelity Graph Networks forMaterials Property Predictions”. Predicting the properties of a material from the arrangement of its atoms is a fundamental goal in materials science. In recent years, machine learning (ML) on ab initio calculations has emerged as a new paradigm to provide rapid predictions of materials properties across vast chemical spaces. However, the performances of ML models are determined by the quantity and quality of data, which tend to be inversely correlated with each other. In this talk, we show that multi-fidelity materials graph networks can transcend this trade-off to achieve accurate predictions of the experimental band gaps of ordered and disordered materials to within 0.3-0.5 eV. Further, such models can be readily extended to predict the band gaps of disordered crystals to excellent agreement with experiments, addressing a major gap in the computational prediction of materials properties.

You can check out the two works discussed in the presentation at:
(1) Chen, C.; Zuo, Y.; Ye, W.; Li, X.; Ong, S. P. Multi-Fidelity Graph Networks for Machine Learning the Experimental Properties of Ordered and Disordered Materials. arXiv:2005.04338 [cond-mat] 2020. (Accepted in Principle for Nature Computational Materials)
(2) Chen, C.; Ye, W.; Zuo, Y.; Zheng, C.; Ong, S. P. Graph Networks as a Universal Machine Learning Framework for Molecules and Crystals. Chem. Mater. 2019, 31 (9), 3564–3572. https://doi.org/10.1021/acs.chemmater.9b01294.

The MEGNet code is available at our Github repo at https://github.com/materialsvirtuallab/megnet.

ASTAR Webinar on The AI Revolution in Materials Science

Prof Ong gave a webinar talk on the AI Revolution in Materials Science for the Singapore Agency of Science Technology and Research (A*STAR). In this talk, he discussed the big challenges in materials science where AI can make a huge impact towards addressing as well as outstanding challenges and opportunities to bringing forth the AI revolution to the materials domain.

Disordered Rock Salt Anode for Fast-charging Lithium-ion Batteries

Li3V2O5

Our joint work with Ping Liu’s group on “A Disordered Rock Salt Anode for Fast-charging Lithium-ion Batteries” has been published in Nature. In this work, we report that disordered rock salt (DRS) Li3+xV2O5 as a fast-charging anode that can reversibly cycle two lithium ions for thousands of cycles. Because it operates at an average voltage of about 0.6 volts versus a Li/Li+, Li3+xV2O5 is less likely compared to graphite to plate lithium metal, alleviating a major safety concern, while still being 71% more energy dense than lithium titanate.

Zhuoying from the Materials Virtual Lab studied the new anode using DFT calculations. We propose a new redistributive lithium intercalation mechanism that suppresses the intercalation voltage and lowers energy barriers for diffusion. This low-potential, high-rate intercalation reaction can be used to identify other metal oxide anodes for fast-charging, long-life lithium-ion batteries.

Li3V2O5

Check out our work at here.

Press:

Design Principles for Aqueous Na-Ion Battery Cathodes

Xingyu’s second paper on “Design Principles for Aqueous Na-Ion Battery Cathodes” has just been published in Chemistry of Materials! In this work, we develop design rules for aqueous sodium-ion battery cathodes through a comprehensive DFT study of known cathode materials. We identified five promising aqueous sodium-ion battery cathode materials – NASICON-Na3Fe2(PO4)3, Na2FePO4F, Na3FeCO3PO4, alluadite-Na2Fe3(PO4)3, and Na3MnCO3PO4 – with high voltage, good capacity, high stability in aqueous environments, and facile Na-ion migration.

Check out this work at this link!

Predicting Thermal Quenching in Phosphors

TQ

Mahdi’s paper on “Predicting Thermal Quenching in Inorganic Phosphors” has just been published in Chemistry of Materials! Phosphors are used in energy efficient light emitting diodes (LEDs). A key performance metric of a phosphor is its thermal quenching (TQ), which is the percentage loss of emission at elevated temperatures during operation. In this work, we develop a new approach to predicting TQ in phosphors using ab initio molecular dynamics (AIMD) simulations and the band gap of the phosphor host. We demonstrate for the first time that TQ under the crossover mechanism is related to the local environment stability of the activator. Further, by accounting for the effect of the crystal field on the thermal ionization barrier, we show that a unified model can predict the experimental TQ in 29 known phosphors to within a root-mean-square error of ∼3.1−7.6%.

Check out the work at this link.