Abstract
The development of CO(2) hydrogenation catalysts has largely depended on researchers's trial and error efforts. However, exploring unreported ternary catalysts requires substantial time and cost, highlighting the need for more efficient screening strategies. This study adopts a data-driven informatics approach where literature data is restructured into networks to reveal systematic relationships between reaction conditions and CO(2) conversion. Knowledge extracted from the catalyst combination network further enables the identification of promising ternary catalysts. NiMnPr/Al(2)O(3) and NiMnCe/Al(2)O(3) for CO(2) hydrogenation are rapidly identified and experimentally validated, exhibiting higher CO(2) conversion than their corresponding binary catalysts. Moreover, detailed characterization is carried out for NiMnPr/Al(2)O(3). These findings and analysis demonstrate that mapping multidimensional data into networks provides a powerful strategy for uncovering correlated variables and facilitating intuitive, highly efficient catalyst development and understanding.