.png){kind=icon}
.png){kind=icon}
Return
From atoms to systems: machine learning in the clean energy materials revolution
.png){kind=icon}
OA
AI
M
C
E
DOI：10.1007/s40243-026-00374-6.png)
.png){kind=icon}
.png){kind=icon}
Share.png){kind=icon}
Save
Abstract
En 中文
Machine learning (ML) has become a pervasive tool in clean-energy materials research, accelerating virtual screening, inverse design, catalyst discovery, process optimisation, and autonomous experimentation across a range of domains. This review offers a systematic and deliberately critical assessment of what that acceleration has and has not yet delivered. Across twelve application domains spanning CO$$_2$$ reduction, hydrogen evolution, nitrogen reduction, ML interatomic potentials, generative inverse design, battery lifetime prediction, sustainable manufacturing, critical-material recycling, and operando characterisation, we introduce the Materials–ML Maturity Matrix (M4) to rate each domain on four dimensions: data reliability, model robustness, experimental and industrial readiness, and generalisation stability. No domain currently achieves high ratings on all four dimensions simultaneously. Four findings emerge from this analysis. First, ML performance outside training distributions is consistently and substantially lower than benchmark results indicate. Second, experimental validation rates for ML-generated candidates remain below 5–10% in inverse-design workflows, and adsorption-energy predictions in catalysis routinely diverge from experiment by more than 0.3 eV. Third, cross-facility reproducibility of autonomous laboratory results has not been established at scale. Fourth, no ML-designed material has yet outperformed the state of the art across the combined metrics of activity, selectivity, stability, and cost in any application area reviewed. Battery lifetime prediction and manufacturing optimisation represent genuinely mature and industrially relevant applications. NRR catalysis and fully autonomous discovery remain at an early, largely conceptual stage. The most binding constraints across domains are infrastructural—data quality, benchmark integrity, and uncertainty quantification—rather than architectural. A five-year priority roadmap grounded in the M4 analysis identifies the interventions most likely to convert ML’s demonstrated acceleration into genuine transformation of clean-energy materials development.
Keywords:
Machine learning
Clean energy materials
AI-driven discovery
Sustainable manufacturing
CO2 reduction catalysts
Green hydrogen
Operando characterisation
Materials informatics
AI Summary
Key information extracted from the uploaded paper, including a brief overview, abstract, background, key highlights, visual analysis, and future outlook.
Journal
IF:
5.5
Papers:
223
Citations:
955
Organization
Cited Papers