(631e) Systematic Identification of Active Site for Propene Combustion Using Pd/Pt Bimetallic Nanoparticle Catalysts and Theoretical Insights

Emission control catalysts have gained increasing interests due to the needs to reduce pollutants, such as hydrocarbons and CO. Currently, most catalysts use Pt and Pd to remove the hazardous species since these elements are known for their high efficiency in combusting hydrocarbons. However, there still are open questions regarding the catalytic behavior of Pd and Pt, since the conventional way to prepare catalysts always results in nonuniform metal morphology and site distribution. Thus, we report a systematic study using well-defined Pd/Pt nanoparticles (NPs) to provide a fundamental understanding of propene combustion, which is chosen to be the model molecule because of its large contribution to the diesel emissions.

We first demonstrated that Pd/Pt with equimolar ratio is the most active composition from light-off experiments. A surprising effect of water promoting the reaction was observed, with its improvement of activity proved to create more active surface sites on NPs. We then set to investigate the effect of size on the equimolar Pd/Pt sample by synthesizing Pd/Pt NPs ranging from 2.3 to 10.2 nm. A clear size-dependence trend in turnover frequency is observed, where the intrinsic activity is increased with increasing NP size. Post-catalysis TEM characterization and rate order measurements, corroborated by computational results, reveal that a specific ensemble of step sites that are active for a variety of representative elementary steps in propene combustion was formed on larger Pd/Pt NPs. We further demonstrated that the increase in the fraction of step-site ensemble is in line with the increase in turnover frequency, proving that the particular ensemble of step sites is the most active site for propene combustion. Taken all together, these insights help in the design of improved catalytic materials for propene combustion where the use of noble metals can be optimized.