Regression model for stabilization energies associated with anion ordering in Perovskite materials
Masanori Kaneko a, Mikiya Fujii a, Takashi Hisatomi b, Koichi Yamashita a, Kazunari Domen a b
a Department of Chemical System Engineering, School of Engineering, The University of Tokyo, 7-3-1 Hongo Bunkyo-ku, Tokyo 113-8656, Japan
b Center for Energy & Environmental Science, Interdisciplinary Cluster for Cutting Edge Research, Shinshu University, 4-17-1 Wakasato, Nagano-shi, Nagano 380-8553, Japan
Proceedings of International Conference on Advances in Organic and Hybrid Electronic Materials (AOHM19)
Dubrovnik, Croatia, 2019 March 17th - 20th
Organizers: Alejandro Briseno, Thuc-Quyen Nguyen and Natalie Stingelin
Poster, Masanori Kaneko, 062
Publication date: 8th January 2019

Recently, various perovskite materials have been studied due to characteristic electrical properties. The structure, bandgap, and stability of the mixed anion perovskite material can be controlled by changing the anion ratio. Anion ordering in the perovskite affects electrical properties, but it is difficult to synthesize specific anion ordering by experiment, understanding anion ordering is still challenging. It is also difficult to comprehensively calculate anion ordering by standard DFT calculation from the viewpoint of calculation cost. Therefore, this work shows that anion ordering of large supercell can be rapidly predicted by using machine learning, using BaNbO2N perovskite (a promising photocatalyst capable of oxidizing water under irradiation up to 740 nm[1]) as an example. A machine learning model for predicting the relationship between the structures and energies obtained by first-principles calculation with some anion orderings was created by using machine learning. By using this model, it became possible to calculate the total energy of arbitrary anion ordering. By combining this model with the Metropolis Monte Carlo method, we were able to search stable anion orderings of large supercells, which is difficult with the first-principles calculation. This study demonstrates a means to predict the properties of functional materials based on the most realistic element ordering with a reasonable computational cost. This method can also be applied to other mixed anion perovskites. MAPbI3 (MA=CH3NH3+), which have been well studied as a photovoltaic device, have been tried to control chemical stability and bandgap by replacing iodine with bromine. Therefore, we also applied the method to MAPb(I1-xBrx)3 and compared with BaNbO2N.

This work was financially supported by Grants-in-Aid for Scientific Research (A) (no. 16H02417) and for Young Scientists (A) (no. 15H05494) from the Japan Society for the Promotion of Science (JSPS). The DFT calculations in this study were performed using the facilities at the Supercomputer Center in the Institute for Solid State Physics (ISSP) at the University of Tokyo and at the Research Center for Computational Science in the Institute for Molecular Science (IMS), Okazaki, Japan. This research was partly supported by MEXT as "Priority Issue on Post-K computer” (Development of new fundamental technologies for high-efficiency energy creation, conversion/storage and use).

© Fundació Scito
We use our own and third party cookies for analysing and measuring usage of our website to improve our services. If you continue browsing, we consider accepting its use. You can check our Cookies Policy in which you will also find how to configure your web browser for the use of cookies. More info