Design rules for ternary CO₂ hydrogenation catalysts via literature-sourced network construction and analysis

Area of Science:

• Catalysis science
• Materials informatics
• Chemical engineering

Background:

• Catalyst development for CO2 hydrogenation often relies on time-consuming trial-and-error methods.
• Exploring novel ternary catalysts is costly and inefficient.
• There is a need for advanced screening strategies to accelerate catalyst discovery.

Purpose of the Study:

• To develop a data-driven informatics approach for efficient screening of CO2 hydrogenation catalysts.
• To identify promising unreported ternary catalysts using network analysis of literature data.
• To validate the performance of newly identified catalysts experimentally.

Main Methods:

• Restructuring literature data into networks to analyze relationships between reaction conditions and CO2 conversion.
• Extracting knowledge from catalyst combination networks to identify potential ternary catalysts.
• Experimental validation and characterization of selected catalysts, including NiMnPr/Al2O3.

Main Results:

• A data-driven network approach successfully identified promising ternary catalysts.
• NiMnPr/Al2O3 and NiMnCe/Al2O3 demonstrated higher CO2 conversion compared to binary counterparts.
• Experimental validation confirmed the superior performance of the identified ternary catalysts.

Conclusions:

• Mapping multidimensional data into networks is a powerful strategy for catalyst development.
• This approach facilitates intuitive and highly efficient discovery of correlated variables for catalyst optimization.
• The study highlights a significant advancement in accelerating catalyst design and understanding for CO2 hydrogenation.