Exploring scaling relations and activity descriptors in methane pyrolysis on molten metal catalysts

초록

Scaling relations between binding energies and their connection to catalytic activity are central to descriptorbased catalyst design, but their applicability to dynamic molten metal catalysts remains largely unexplored. While such relations have been established for solid catalysts on well-defined crystalline planes, this approach is challenged in molten catalysts. Here, we use ab initio molecular dynamics (AIMD) combined with density functional theory (DFT) and a bootstrap statistical approach to investigate relationships between binding energies and catalytic activity in methane pyrolysis on molten pure and alloy metals. This approach samples a broad range of structural configurations and yields statistically meaningful binding energies for CHx and H, with the moving block bootstrap analysis enabling robust uncertainty quantification of the resulting scaling relations. Scaling relations are well preserved in pure molten p-block metals, where variations in binding strength are governed predominantly by local coordination environments. In contrast, scaling becomes weakened in p-block/ transition-metal binary alloys, where both structural and electronic contributions jointly determine the binding energetics. In addition, a linear relationship between the binding energy of the final dehydrogenated state and the experimental catalytic activity is observed in binary alloys, providing a reliable descriptor and consistent with the Br & oslash;nsted-Evans-Polanyi relation through the hypothetical one-step approximation of multistep dehydrogenation, whereas no such correlation is observed in pure metals. These findings demonstrate how statistical, configuration-resolved energetics can guide descriptor development for liquid catalysts.

키워드