Hydrogen peroxide ($H_2O_2$) is a green oxidant that produces only water as its reaction product and hence widely used in almost all industrial areas, particularly in fine chemical industry and environmental remediation. Beyond the conventional anthraquinone process, catalytic direct synthesis of $H_2O_2$ from $H_2$ and $O_2$ could be an attractive solution for production of $H_2O_2$ with low energy input and clean reactions. Despite the simplicity of the process design, the direct synthesis of $H_2O_2$ still faces a series of challenges such as the high flammability of $H_2/O_2$ mixtures and low $H_2O_2$ yields due to the thermodynamically favored competing formation of $H_2O$ and the subsequent disproportionation/hydrogenation of the produced $H_2O_2$ into $H_2O$. The work presented in this thesis addresses two different ways to develop the catalysts with having high selectivity and productivity toward $H_2O_2$.
In the first part, the results showed that a high density of oxygen functional groups on the carbon surface was essential for synthesizing highly dispersed bimetallic catalysts with effective AuPd alloying, which is a prerequisite for achieving high $H_2O_2$ selectivity. Regarding porous structure, a solely mesoporous carbon support was superior to microporous ones because $H_2O_2$ produced from AuPd catalyst in the micropores was more prone to subsequent disproportionation/hydrogenation into $H_2O$ due to retarded diffusion of the $H_2O_2$ out of the microporous structure. In the second part, selective dissociation of $H_2$ over $O_2$ was realized by depositing $H_2$ -selective carbon diffusion layers on the top of a Pt catalyst. Because $O_2$ cannot access to the carbon-coated Pt surface, $O_2$ hydrogenation occurs at the carbon surface via spilt-over hydrogen rather than at the Pt surface where O-O dissociation is likely. This leads to the great suppression of O-O dissociation, which allows highly selective synthesis of $H_2O_2$.