How fast do defects migrate in halide perovskites: insights from on-the-fly machine-learned force fields

Area of Science:

• Materials Science
• Solid-State Physics
• Computational Chemistry

Background:

• Defect migration significantly impacts halide perovskite stability.
• Experimental and conventional simulation methods face limitations in studying defect dynamics at the atomic scale.
• Accurate simulation of defect migration is essential for understanding and improving perovskite materials.

Purpose of the Study:

• To investigate the dynamical behavior of halide interstitials and vacancies in CsPbI3 and CsPbBr3.
• To compare the migration rates of interstitials versus vacancies.
• To elucidate the factors influencing defect migration speed in different halide perovskite compositions.

Main Methods:

• Utilizing machine-learned force fields trained with on-the-fly active learning.
• Employing accurate density functional theory (DFT) calculations for training data.
• Performing simulations to probe the differences in dynamical behavior of halide defects.

Main Results:

• Machine-learned force fields enable atomic-scale resolution of defect migration.
• Interstitials exhibit faster migration than vacancies due to shorter paths.
• Defect migration is more rapid in CsPbI3 compared to CsPbBr3.

Conclusions:

• The study successfully demonstrates the capability of machine-learned force fields for defect migration analysis.
• Less compact ion packing in CsPbI3 facilitates more frequent defect migration jumps.
• Findings provide insights into the stability and dynamics of halide perovskites, crucial for their technological applications.