Going beyond Conventional Catalytic Oxidation of Formaldehyde via Machine-Learning-Accelerated Design of Oxide-Silver Tandem Catalysts.
Designing high-performance catalysts for efficient HCHO oxidation under mild conditions is essential, yet remains challenging to achieve through conventional experimental screening. Here, we present a machine-learning-accelerated strategy for designing novel oxide-Ag tandem systems that overcome this limitation by coupling oxide-mediated HCHO activation with subsequent intermediate oxidation on Ag. Guided by an adsorption-energy-based activity descriptor, four theoretically predicted oxide-Ag systems (TiO2, Nb2O5, Ga2O3, SnO2) exhibit markedly enhanced tandem catalytic performance compared to conventional Ag catalysts. In particular, a representative TiO2/Ag-γ-Al2O3, prepared by simple physical mixing of commercial anatase and Ag-γ-Al2O3 catalyst, achieves an HCHO oxidation rate of 0.56 μmol gAg-1 s-1 at 55 °C, surpassing Ag-γ-Al2O3 alone by over 2 orders of magnitude in performance. Combining experimental and theoretical studies further reveals a cascade reaction mechanism, in which TiO2 catalyzes the HCHO-to-methyl formate transformation via a surface OH-mediated pathway, followed by efficient methyl formate oxidation to CO2 on Ag-γ-Al2O3. Particularly, the local oxygen environment over oxide surfaces is identified as a key factor governing HCHO adsorption and conversion. This work establishes a generalizable design principle for tandem catalysts and provides a data-driven framework for advancing low-temperature HCHO oxidation technologies.