What question did this study set out to answer?

To develop a hybrid computational workflow combining density functional theory and data-driven machine learning models to identify stable 1/1 Mackay-type quasicrystal approximants in scandium-rich intermetallic systems.

June 7, 2022Crystal Growth & Design

Hybrid DFT/Data-Driven Approach for Searching for New Quasicrystal Approximants in Sc-X (X = Rh, Pd, Ir, Pt) Systems

View Full Paper

Ask AI

Bookmark

Authors

RERoman A. EreminPresidential Executive Office IHInnokentiy S. HumonenDevelopment Agency of Serbia PZPavel N. ZolotarevUniversity of Milano-Bicocca

Discussion

Member takes

Implication

Computational study demonstrates machine learning-accelerated identification of over twenty quasicrystal approximants in scandium intermetallics, indicating high screening efficiency.

Key Points

To develop a hybrid computational workflow combining density functional theory and data-driven machine learning models to identify stable 1/1 Mackay-type quasicrystal approximants in scandium-rich intermetallic systems.
Generated simplified configuration spaces containing up to 2,107 structures each for Sc-Rh, Sc-Pd, Sc-Ir, and Sc-Pt systems and computed their thermodynamic properties using density functional theory (DFT).
Trained Random Forest and graph neural network (GNN) regression models on the DFT datasets to screen extended configuration spaces of 3,942 structures per system across multiple target energy metrics.
Predicted and DFT-validated more than 20 potentially new stable quasicrystal approximants in the Sc-Pt and Sc-Pd intermetallic systems.
Demonstrated that graph neural network predictive recall and performance vary based on the specific target energy metric, including relaxed energy, formation energy, and energy above the convex hull.

Cite This Study

Eremin et al. (2022) studied this question.

synapsesocial.com/papers/6a1ff91b3224f8dacd0db4c5 https://doi.org/10.1021/acs.cgd.2c00463

View Full Paper

Ask AI

Bookmark

Implication

Computational study demonstrates machine learning-accelerated identification of over twenty quasicrystal approximants in scandium intermetallics, indicating high screening efficiency.

Hybrid DFT/Data-Driven Approach for Searching for New Quasicrystal Approximants in Sc-X (X = Rh, Pd, Ir, Pt) Systems

Discussion

Member takes

Implication

Key Points

Cite This Study

Hybrid DFT/Data-Driven Approach for Searching for New Quasicrystal Approximants in Sc-X (X = Rh, Pd, Ir, Pt) Systems

Discussion

Member takes

Implication

Key Points

Cite This Study

Also Consider

Also Consider