Screening for lead-free inorganic double perovskites with suitable band gaps and high stability using combined machine learning and DFT calculation

• Published in: Applied surface science 2021-12, Vol.568, p.150916, Article 150916
• Main Authors: Gao, Zhengyang, Zhang, Hanwen, Mao, Guangyang, Ren, Jianuo, Chen, Ziheng, Wu, Chongchong, Gates, Ian D., Yang, Weijie, Ding, Xunlei, Yao, Jianxi
• Format: Article
• Language: English
• Subjects: Band gap; DFT calculation; Lead-free inorganic double perovskites; Machine learning; Material screening
• ISSN: 0169-4332; 1873-5584
• DOI: 10.1016/j.apsusc.2021.150916

[Display omitted] •The eXtreme Gradient Boosting Regression (XGBR) algorithm was first applied to build a robust and predictive machine learning (ML) model for perovskite materials.•The method of combining machine learning and DFT calculation used in this article can greatly save computing resources and time, and accelerate the discovery of renewable energy materials.•Two lead-free perovskite structures screened out by the combination of machine learning and DFT calculation have reasonable band gap values, high environmental stability and good optical absorption properties. To accelerate the application of perovskite materials in photovoltaic solar cells, developing novel lead-free perovskite materials with suitable band gaps and high stability is vital. However, laborious experiment and density functional theory (DFT) calculation are time-consuming and incapable to screen promising perovskites rapidly and efficiently. Here, we proposed a novel search strategy combining machine learning and DFT calculation to screen 5,796 inorganic double perovskites. The eXtreme Gradient Boosting Regression (XGBR) algorithm was first applied to build a robust and predictive machine learning (ML) model for perovskite materials. XGBR algorithm yielded a lower mean square error (MSE) than both Artificial Neural Network (ANN) algorithm and Support Vector Regression (SVR) algorithm. From the ML model, two novel lead-free inorganic double perovskites: Na2MgMnI6, K2NaInI6, were obtained, suitable direct bandgaps of 1.46 eV for K2NaInI6 and 1.89 eV for Na2MgMnI6, which are similar to the organic–inorganic perovskite (MAPI3) CH3NH3PbI3 (Eg = 1.6 eV), high thermal stability and good optical properties were also confirmed by DFT calculation. The method of combining ML and DFT exhibits high accuracy and significantly speeds up the discovery of promising perovskite materials.