Researchers at Lawrence Berkeley National Laboratory have developed a data-driven framework to identify the synthesis variables that most strongly influence the performance of chiral two-dimensional metal halide perovskites, materials under investigation for future spin-based optoelectronic technologies.
The work addresses a persistent reproducibility problem. Reported performance measurements for nominally identical chiral perovskite materials have varied by more than two orders of magnitude among laboratories, making it difficult for researchers to compare results and refine fabrication methods.
Published in Matter, the study examined how multiple processing variables affect the films’ chiroptical properties, or their interactions with circularly polarized light. These properties are important for proposed technologies such as light-emitting diodes and photodetectors that could use electron spin and polarized light to encode and transmit information.
Identifying the most influential synthesis variables
First author Raphael Moral prepared thin films from single-crystal precursor solutions and used X-ray techniques at Berkeley Lab’s Advanced Light Source to examine how the materials crystallized during fabrication.
Moral and co-first author Maher Alghalayini then used statistical tools, including correlation analysis and machine-learning methods supported by Berkeley Lab’s Center for Advanced Mathematics for Energy Research Applications, to identify and model the fabrication parameters that most strongly affected material performance.
The analysis identified solvent choice as the most influential variable. Films made with acetonitrile produced the strongest and most consistent chiroptical signals. Annealing temperature and film thickness also affected signal strength.
The team used X-ray diffraction experiments at the Advanced Light Source to validate the model’s predicted results.
Creating a roadmap for process optimization
The framework offers researchers a structured alternative to relying exclusively on trial-and-error experimentation. It connects fabrication conditions with measured material responses, allowing researchers to rank influential variables and test predicted processing methods through independent characterization.
“It is surprising that the same material can produce different chiroptical properties depending on the processing method,” Moral said.
The Berkeley Lab team plans to apply lessons from the study to future machine-learning-driven experiments involving other chiral molecules.
Implications for laboratory practice
For lab managers, the study provides an example of how detailed process documentation, statistical analysis, and independent validation can help research teams investigate experimental variability. It also suggests that integrating data science approaches earlier in experimental design may reduce time spent on iterative trial-and-error, improve reproducibility across different facilities, and support more efficient scaling of promising materials systems from exploratory research to applied development.
This article was created with the assistance of Generative AI and has undergone editorial review before publishing.
