Deep learning object detection in materials science: Current state and future directions

Computational Materials Science, Journal Year: 2022, Volume and Issue: 211, P. 111527 - 111527

Published: May 24, 2022

Language: Английский

---

Designing high-TC superconductors with BCS-inspired screening, density functional theory, and deep-learning

npj Computational Materials, Journal Year: 2022, Volume and Issue: 8(1)

Published: Nov. 22, 2022

Abstract We develop a multi-step workflow for the discovery of conventional superconductors, starting with Bardeen–Cooper–Schrieffer inspired pre-screening 1736 materials high Debye temperature and electronic density states. Next, we perform electron-phonon coupling calculations 1058 them to establish large systematic database BCS superconducting properties. Using McMillan-Allen-Dynes formula, identify 105 dynamically stable transition temperatures, T C ≥ 5 K. Additionally, analyze trends in our dataset individual including MoN, VC, VTe, KB 6 , Ru 3 NbC, V Pt, ScN, LaN 2 RuO TaC. demonstrate that deep-learning(DL) models can predict superconductor properties faster than direct first-principles computations. Notably, find by predicting Eliashberg function as an intermediate quantity, improve model performance versus DL prediction . apply trained on crystallographic open pre-screen candidates further DFT calculations.

Language: Английский

---

Graph neural network predictions of metal organic framework CO2 adsorption properties

Computational Materials Science, Journal Year: 2022, Volume and Issue: 210, P. 111388 - 111388

Published: April 14, 2022

Language: Английский

---

Band gap predictions of double perovskite oxides using machine learning

Communications Materials, Journal Year: 2023, Volume and Issue: 4(1)

Published: June 10, 2023

Abstract The compositional and structural variety inherent to oxide perovskites spawn wide-ranging applications. In perovskites, the band gap E g , a key material parameter for these applications, can be optimally controlled by varying composition. Here, we implement hierarchical screening process in which two cross-validated predictive machine learning models classification regression, trained using exhaustive datasets that span 68 elements of periodic table, are applied sequentially. model separates wide materials, with ≥ 0.5 eV, from materials have zero or relatively small gaps, namely < second regression quantitatively predicts value compounds. study down-selects 13,589 cubic perovskite compositions predicted experimentally formable, thermodynamically stable, gap. Of these, subset 310 compounds, stable formable confidence greater than 90%, identified further investigation. Our methodically analyzed via performance metrics inter-dependence features gain physical insight into prediction problem. Design maps identify variation substitution different also presented.

Language: Английский

---

Conditional diffusion-based microstructure reconstruction

Materials Today Communications, Journal Year: 2023, Volume and Issue: 35, P. 105608 - 105608

Published: Feb. 13, 2023

Language: Английский

---