Related Experiment Video
Updated: Mar 25, 2026
11:38
Influence of Hybrid Perovskite Fabrication Methods on Film Formation, Electronic Structure, and Solar Cell Performance
Published on: February 27, 2017
19.1K
Machine Learning for Designing Perovskites and Perovskite-Inspired Solar Materials: Emerging Opportunities and
Yangfan Zhang1, Yiming Xia1, Ali Shakiba1
1School of Photovoltaic and Renewable Energy Engineering, University of New South Wales, Sydney, New South Wales, Australia.
Advanced Science (Weinheim, Baden-Wurttemberg, Germany)
|March 24, 2026
Summary
Machine learning (ML) accelerates the discovery of efficient, non-toxic solar materials like perovskites and perovskite-inspired materials (PIMs). This review details ML workflows for predicting material properties, aiding the development of next-generation solar energy technologies.
Area of Science:
- Materials Science
- Renewable Energy
- Computational Chemistry
Background:
- Perovskites and perovskite-inspired materials (PIMs) are crucial for efficient solar energy conversion.
- Halide perovskites offer excellent optoelectronic properties but face challenges with toxicity and stability.
- Traditional methods for material discovery are costly and slow, necessitating data-driven approaches.
Purpose of the Study:
- To review machine learning (ML)-driven strategies for predicting properties of perovskites and PIMs.
- To outline a comprehensive ML workflow for materials discovery.
- To assess the applicability of ML models across diverse PIMs.
Main Methods:
- Data collection and target identification for ML models.
- Feature engineering and selection for predictive accuracy.
- Application of supervised, unsupervised, and reinforcement learning frameworks.
- Evaluation of ML model transferability from halide perovskites to PIMs.
Main Results:
- ML effectively predicts key properties like bandgap, stability, and lattice constants.
- Established ML workflows can be adapted for PIMs discovery.
- Transferability of ML strategies to chemically diverse PIMs is a key focus.
Conclusions:
- ML integration is vital for rational design of next-generation solar materials.
- Data-driven approaches can accelerate the discovery of non-toxic, stable, and efficient solar absorbers.
- This review provides a roadmap for advancing ML in solar materials innovation.
