Active learning for accelerated design of layered materials

• Published in: npj computational materials Vol. 4; no. 1; pp. 1 - 9
• Main Authors: Bassman Oftelie, Lindsay, Rajak, Pankaj, Kalia, Rajiv K., Nakano, Aiichiro, Sha, Fei, Sun, Jifeng, Singh, David J., Aykol, Muratahan, Huck, Patrick, Persson, Kristin, Vashishta, Priya
• Format: Journal Article
• Language: English
• Published: London Nature Publishing Group UK 10.12.2018; Nature Publishing Group
• Subjects: 639/301/1034/1037; 639/638/898; Active learning; Bayesian analysis; Chalcogenides; Characterization and Evaluation of Materials; Computational efficiency; Computational Intelligence; Computer applications; Computing costs; Conduction; Conduction bands; Data acquisition; Data collection; Dependence; Design optimization; Electronic structure; Energy gap; Gaussian process; Heterostructures; Layered materials; Learning algorithms; Machine learning; Material properties; MATERIALS SCIENCE; Mathematical and Computational Engineering; Mathematical and Computational Physics; Mathematical Modeling and Industrial Mathematics; Mathematical models; Optimization; Optoelectronic devices; Predictions; Regression analysis; Regression models; Source code; Statistical analysis; Theoretical; Thermoelectricity; Transition metal compounds; Transport properties; Valence band
• ISSN: 2057-3960; 2057-3960
• DOI: 10.1038/s41524-018-0129-0

Hetero-structures made from vertically stacked monolayers of transition metal dichalcogenides hold great potential for optoelectronic and thermoelectric devices. Discovery of the optimal layered material for specific applications necessitates the estimation of key material properties, such as electronic band structure and thermal transport coefficients. However, screening of material properties via brute force ab initio calculations of the entire material structure space exceeds the limits of current computing resources. Moreover, the functional dependence of material properties on the structures is often complicated, making simplistic statistical procedures for prediction difficult to employ without large amounts of data collection. Here, we present a Gaussian process regression model, which predicts material properties of an input hetero-structure, as well as an active learning model based on Bayesian optimization, which can efficiently discover the optimal hetero-structure using a minimal number of ab initio calculations. The electronic band gap, conduction/valence band dispersions, and thermoelectric performance are used as representative material properties for prediction and optimization. The Materials Project platform is used for electronic structure computation, while the BoltzTraP code is used to compute thermoelectric properties. Bayesian optimization is shown to significantly reduce the computational cost of discovering the optimal structure when compared with finding an optimal structure by building a regression model to predict material properties. The models can be used for predictions with respect to any material property and our software, including data preparation code based on the Python Materials Genomics (PyMatGen) library as well as python-based machine learning code, is available open source. Materials design: Bayesian optimization High accuracy predictions of materials properties can be obtained using Bayesian optimization (BO). A team led by Priya Vashishta at University of Southern California developed a Gaussian regression model capable of predicting the band gap value and thermoelectric properties of three-layered van der Waals heterostructures of transition metal dichalcogenides. A BO model further allowed identification of optimal heterostructures using a minimal number of ab initio calculations. BO models were computed to find either heterostructures with maximum band gap or heterostructures with a band gap value closest to 1.1 eV, relevant for optoelectronic and thermoelectric applications. BO was found to identify nearly optimal materials configurations with high probability, whilst significantly reducing the computational cost of discovering ideal structures using regression models.