Beyond Tolerance Factor: Using Deep Learning for Prediction Formability of ABX3 Perovskite Structures. Issue 5 (2nd April 2021)

• Record Type: Journal Article
• Authors: Fedorovskiy, Alexander E.; Queloz, Valentin I. E.; Nazeeruddin, Mohammad Khaja
• Is Part Of: Advanced theory and simulations. Volume 4:Issue 5(2021)
• Page Start: n/a
• Page End: n/a
• Publication Date: 2021-04-02
• Subjects: crystal structures -- deep learning -- formability prediction -- perovskites -- tolerance factor; Science -- Simulation methods -- Periodicals; Science -- Methodology -- Periodicals; Engineering -- Simulation methods -- Periodicals; Engineering -- Methodology -- Periodicals; 507.21
• DOI: 10.1002/adts.202100021
• Languages: English
• ISSNs: 2513-0390

Abstract

Abstract: Deep learning (DL) is a modern powerful instrument for multiple purposes, including classification. In this study, this technique is applied to the task of perovskites formability. A commonly known perovskite dataset is used to try to make an instrument superior to the 'classic' geometric approach. The authors found that the resulting models allow the finding of inaccuracies in the data and can successfully forecast perovskite formability with an accuracy of over 98% for the best case. Abstract : The study presents an improved way of predicting the formability of ABX 3 oxides and halide perovskites based on a deep learning approach. The impact of different input features are studied. The models displayed high accuracy and the ability to clarify the structure initial dataset's compounds. Therefore, the results could be an effective instrument for future compound screening.