Abstract:Inorganic metal halide perovskites, such as CsPbI3 , have recently drawn extensive attention due to their excellent optical properties and high photoelectric efficiencies. However, the structural instability originating from inherent ionic defects leads to a sharp drop in the photoelectric efficiency, which significantly limits their applications in solar cells. The instability induced by ionic defects remains unresolved due to its complicated reaction process. Herein, to explore the effects of ionic defects on stability, we develop a deep learning potential for a CsPbI3 ternary system based upon density functional theory (DFT) calculated data for large-scale molecular dynamics (MD) simulations. By exploring 2.4 million configurations, of which 7,730 structures are used for the training set, the deep learning potential shows an accuracy approaching DFT-level. Furthermore, MD simulations with a 5,000-atom system and a one nanosecond timeframe are performed to explore the effects of bulk and surface defects on the stability of CsPbI3 . This deep learning potential based MD simulation provides solid evidence together with the derived radial distribution functions, simulated diffraction of X-rays, instability temperature, molecular trajectory, and coordination number for revealing the instability mechanism of CsPbI3 . Among bulk defects, Cs defects have the most significant influence on the stability of CsPbI3 with a defect tolerance concentration of 0.32 %, followed by Pb and I defects. With regards to surface defects, Cs defects have the largest impact on the stability of CsPbI3 when the defect concentration is less than 15 %, whereas Pb defects act play a dominant role for defect concentrations exceeding 20 %. Most importantly, this machine-learning-based MD simulation strategy provides a new avenue to explore the ionic defect effects on the stability of perovskite-like materials, laying a theoretical foundation for the design of stable perovskite materials.

Cover Feature: Exploring the Effects of Ionic Defects on the Stability of CsPbI 3 with a Deep Learning Potential (ChemPhysChem 7/2022)

Exploring the Effects of Ionic Defects on the Stability of CsPbI3 with a Deep Learning Potential

Significant Impact of Defect Fluctuation on Charge Dynamics in CsPbI3: A Study Combining Machine Learning with Quantum Dynamics

Machine Learning Mechanical Properties of Defect-Engineered Hexagonal Boron Nitride

Physics-Based Feature Makes Machine Learning Cognizing Crystal Properties Simple

Cover Feature: Molecular‐Level Insight into the Chlorofluorocarbons Adsorption by Defective Covalent Organic Polymers (ChemPhysChem 20/2024)

Electron Density Learning of Z-Bonds in Ionic Liquids and Its Application

Machine-Learning-based Investigation of Atomic Packing Effects: Chemical Pressures at the Extremes of Intermetallic Complexity

Defect formation in CsSnI$_3$ from Density Functional Theory and Machine Learning

Cover Feature: Controlling Chirogenic Effects in Porphyrin Based Supramolecular Systems: Theoretical Analysis Versus Experimental Observations (ChemPhysChem 11/2024)

Data-driven prediction of complex crystal structures of dense lithium

Machine Learning-Empowered Study of Metastable γ-CsPbI3 under Pressure and Strain

Improved Physics-Based Structural Descriptors of Perovskite Materials Enable Higher Accuracy of Machine Learning

Crystalline Material Discovery in the Era of Artificial Intelligence

Enhancing crystal structure prediction by combining computational and experimental data via graph networks

Accelerating materials discovery using integrated deep machine learning approaches

Prediction of Energetic Material Properties from Electronic Structure Using 3D Convolutional Neural Networks

Exploring Challenging Properties of Liquid Metallic Systems through Machine Learning: Liquid La and Li4Pb Systems

Discovering Unusual Structures from Exception Using Big Data and Machine Learning Techniques.

Machine learning potential assisted exploration of complex defect potential energy surfaces

Accelerating the calculation of electron-phonon coupling strength with machine learning