Machine Vision-Enabled Octahedral Network Reconstruction and Structural Analysis of Perovskite Quantum Dots

Area of Science:

• Materials Science
• Nanotechnology
• Crystallography

Background:

• Metal-halide perovskites feature corner-sharing PbX6 octahedra crucial for optoelectronic properties.
• Analyzing octahedral configurations in perovskite quantum dots (QDs) is challenging due to data limitations.

Purpose of the Study:

• To develop a machine vision approach for high-fidelity analysis of PbX6 octahedral networks in perovskite QDs.
• To establish atomic-scale structure-property relationships in perovskite nanomaterials.

Main Methods:

• Integrated self-supervised denoising (S2SRED) for low-dose scanning transmission electron microscopy (STEM) data.
• Automated atomic species classification and PbX6 octahedral network reconstruction.
• Precise lattice parameter extraction and analysis of octahedral tilting.

Main Results:

• Observed reduced PbX6 octahedral tilting in CsPbI3 QDs, forming isotropic core-shell structures.
• Identified inhomogeneous, anisotropic PbX6 tilting in mixed-halide CsPbI3-xBrix QDs, linked to dopant segregation.
• Corroborated findings with photoluminescence measurements, showing impaired phase stability in mixed-halide QDs.

Conclusions:

• The machine vision method provides robust analysis of perovskite QD structures at the atomic scale.
• Octahedral geometry and lattice parameters are critical for understanding phase stability and optoelectronic performance.
• Established a standardized approach for linking atomic structure to properties in perovskite nanomaterials.