Atomic fingerprints for high throughput screening of 2D monolayers

Nayamadi Mahmoodabadi A., Modarresi M., MOĞULKOÇ A.

Applied Physics Letters, vol.125, no.19, 2024 (SCI-Expanded, Scopus) identifier

  • Publication Type: Article / Article
  • Volume: 125 Issue: 19
  • Publication Date: 2024
  • Doi Number: 10.1063/5.0226435
  • Journal Name: Applied Physics Letters
  • Journal Indexes: Science Citation Index Expanded (SCI-EXPANDED), Scopus, Academic Search Premier, Applied Science & Technology Source, Chemical Abstracts Core, Compendex, Computer & Applied Sciences, INSPEC, DIALNET
  • Ankara University Affiliated: Yes

Abstract

In materials science, artificial intelligence is used to create machine learning models to predict material properties and discover different compounds. An atomic fingerprint is designed to reflect the structure and atomic features of 2D monolayer crystals. A neural network-based model is trained using fingerprints and the computational 2D materials database to predict 2D monolayers. The model predicted the formation energies of crystalline compounds, including some previously unexplored monolayers, which could be potential candidates for future technologies.