Machine-Learning-Accelerated Catalytic Activity Predictions of Transition Metal Phthalocyanine Dual-Metal-Site Catalysts for CO2 Reduction
Ускоренное машинным обучением прогнозирование каталитической активности двухметалльных каталитических центров на основе фталоцианинов переходных металлов для восстановления CO₂
2021-06-25
SCID: 54.1/sgtveuef
Discuss with AI
CO2 reduction reactiondensity functional theorydual-metal-site catalystsgradient boosting regressiontransition metal phthalocyanines
Figures from the paper
Abstract (AI)
The highly active and selective carbon dioxide reduction reaction (CO 2 RR) can generate valuable products such as fuels and chemicals and reduce the emission of greenhouse gases. Single-atom catalysts (SACs) and dual-metal-sites catalysts (DMSCs) with high activity and selectivity are superior electrocatalysts for the CO 2 RR as they have higher active site utilization and lower cost than traditional noble metals. Herein, we explore a rational and creative density-functional-theory-based, machine-learning-accelerated (DFT-ML) method to investigate the CO 2 RR catalytic activity of hundreds of transition metal phthalocyanine (Pc) DMSCs. The gradient boosting regression (GBR) algorithm is verified to be the most desirable ML model and is used to construct catalytic activity prediction, with a root-mean-square error of only 0.08 eV. The results of ML prediction demonstrate Ag-MoPc as a promising CO 2 RR electrocatalyst with the limiting potential of only −0.33 V. The DFT-ML hybrid scheme accelerates the efficiency 6.87 times, while the prediction error is only 0.02 V, and it sheds light on the path to accelerate the rational design of efficient catalysts for energy conversion and conservation.
Key Findings
1
A DFT-based, machine-learning-accelerated workflow was developed to evaluate CO2 reduction activity across hundreds of transition-metal phthalocyanine dual-metal-site catalysts.
2
Ag-MoPc was predicted as a promising CO2 reduction electrocatalyst, with a limiting potential of only −0.33 V.
3
Gradient boosting regression was identified as the best-performing machine-learning model, achieving a root-mean-square prediction error of 0.08 eV.
4
The approach enables rational, accelerated design and screening of dual-metal-site catalysts for energy-conversion applications.
5
The hybrid DFT–machine-learning scheme accelerated catalyst-screening efficiency by 6.87 times while maintaining a prediction error of only 0.02 V.
Research Object
Transition metal phthalocyanine dual-metal-site catalysts for CO2 reduction
Research Subject
CO2 reduction catalytic activity, selectivity, and limiting potential across the catalyst family, including Ag-MoPc
Publication Details
Publication Date
2021-06-25
Journal
Publisher
ISSN
Cited by
172
Access Type
Author Information
Download PDF
Subscribe to digest