Structure motif–centric learning framework for inorganic crystalline systems

Banjade, Huta; Hauri, Sandro; Zhang, Shanshan; Ricci, Francesco; Gong, Weiyi; Hautier, Geoffroy; Vučetić, Slobodan; Yan, Qimin

doi:10.1126/sciadv.abf1754

Cited by 20 publications

(18 citation statements)

References 63 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Solely relying on these pre-calculated atomic composition features implies that the model ignores any 3D-structural environment information within the materials and thus misses the key attributes used in reference physics based simulations [15,16]. However, we anticipate that combining both chemical composition along with some 3D-structural information inside of a more complex deep learning method may lead to a significantly more accurate voltage prediction; as it has been shown for other crystals' properties predictions [17][18][19].…”

Section: Introductionmentioning

confidence: 97%

Accurate Prediction of Voltage of Battery Electrode Materials Using Attention-Based Graph Neural Networks

Louis

Siriwardane

Joshi

et al. 2022

ACS Appl. Mater. Interfaces

View full text Add to dashboard Cite

Performing first principle calculations to discover electrodes' properties in the large chemical space is a challenging task. While machine learning (ML) has been applied to effectively accelerate those discoveries, most of the applied methods ignore the materials' spatial information and only use pre-defined features: based only on chemical compositions. We propose two attention-based graph convolutional neural network techniques to learn the average voltage of electrodes. Our proposed method, which combines both atomic composition and atomic coordinates in 3D-space, improves the accuracy in voltage prediction by 17% when compared to composition based ML models. The first model directly learns the chemical reaction of electrodes and metal-ions to predict their average voltage, whereas the second model combines electrodes' ML predicted formation energy (E form ) to compute their average voltage. Our models demonstrates improved accuracy in transferability from our subset of learned metal-ions to other metal-ions.

show abstract

Section: Introductionmentioning

confidence: 97%

Accurate Prediction of Voltage of Battery Electrode Materials Using Attention-Based Graph Neural Networks

Louis

Siriwardane

Joshi

et al. 2022

ACS Appl. Mater. Interfaces

View full text Add to dashboard Cite

show abstract

“…13,14 Most of the ML models developed in computational materials science to date have been focused on individual scalar quantities. Common properties are those directly computed ab initio, such as the formation energy 11,15,16 , the shear-and bulk-moduli 11,15,17,18 , the band gap energy 11,16,18,19 , and the Fermi energy 11 . Additional targets include properties calculated from the ab initio output, which are typically performance metrics for a target application such as Seebeck coefficient 20,21 .…”

Section: Introductionmentioning

confidence: 99%

Density of States Prediction for Materials Discovery via Contrastive Learning from Probabilistic Embeddings

Kong,

Ricci,

Guevarra

et al. 2021

Preprint

Self Cite

View full text Add to dashboard Cite

Machine learning for materials discovery has largely focused on predicting an individual scalar rather than multiple related properties, where spectral properties are an important example. Fundamental spectral properties include the phonon density of states (phDOS) and the electronic density of states (eDOS), which individually or collectively are the origins of a breadth of materials observables and functions. Building upon the success of graph attention networks for encoding crystalline materials, we introduce a probabilistic embedding generator specifically tailored to the prediction of spectral properties. Coupled with supervised contrastive learning, our materials-to-spectrum (Mat2Spec) model outperforms state-of-the-art methods for predicting ab initio phDOS and eDOS for crystalline materials. We demonstrate Mat2Spec's ability to identify eDOS gaps below the Fermi energy, validating predictions with ab initio calculations and thereby discovering candidate thermoelectrics and transparent conductors. Mat2Spec is an exemplar framework for predicting spectral properties of materials via strategically incorporated machine learning techniques.

show abstract

“…Here we take the approach of physics informed machine learning, which incorporates physical principles into machine learning (ML) models [5][6][7][8]30]. We find that the polyhedral formed by cations and nearby anions can serve as an important optimization target to the machine learning framework of crystal materials [31]. Therefore, we added the coordination number of the cation as an additional optimization objective to the previous contact map-based CSP algorithm, the CMCrystal.…”

Section: Introductionmentioning

confidence: 99%

Crystal structure prediction using age-fitness multi-objective genetic algorithm and coordination number constraints

Yang,

Siriwardane,

2021

Preprint

View full text Add to dashboard Cite

Crystal structure prediction (CSP) has emerged as one of the most important approaches for discovering new materials. CSP algorithms based on evolutionary algorithms and particle swarm optimization have discovered a great number of new materials. However, these algorithms based on ab initio calculation of free energy are inefficient. Moreover, they have severe limitations in terms of scalability. We recently proposed a promising crystal structure prediction method based on atomic contact maps, using global optimization algorithms to search for the Wyckoff positions by maximizing the match between the contact map of the predicted structure and the contact map of the true crystal structure. However, our previous contact map based CSP algorithms have two major limitations: (1) the loss of search capability due to getting trapped in local optima; (2) it only uses the connection of atoms in the unit cell to predict the crystal structure, ignoring the chemical environment outside the unit cell, which may lead to unreasonable coordination environments. Herein we propose a novel multi-objective genetic algorithms for contact map-based crystal structure prediction by optimizing three objectives, including contact map match accuracy, the individual age, and the coordination number match. Furthermore, we assign the age values to all the individuals of the GA and try to minimize the age aiming to avoid the premature convergence problem. Our experimental results show that compared to our previous CMCrystal algorithm, our multi-objective crystal structure prediction algorithm (CMCrystalMOO) can reconstruct the crystal structure with higher quality and alleviate the problem of premature convergence.

show abstract

Structure motif–centric learning framework for inorganic crystalline systems

Cited by 20 publications

References 63 publications

Accurate Prediction of Voltage of Battery Electrode Materials Using Attention-Based Graph Neural Networks

Accurate Prediction of Voltage of Battery Electrode Materials Using Attention-Based Graph Neural Networks

Density of States Prediction for Materials Discovery via Contrastive Learning from Probabilistic Embeddings

Crystal structure prediction using age-fitness multi-objective genetic algorithm and coordination number constraints

Contact Info

Product

Resources

About