Methane adsorption and dissociation and oxygen adsorption and reaction with CO on Pd nanoparticles on MgO(1 0 0) and on Pd(1 1 1)

We present measurements of the desorption kinetics and dissociative sticking probability of methane on the surfaces of Pd(1 1 1) and Pd nanoparticles supported on MgO(1 0 0). A molecular beam system was used to directly probe the fraction of methane molecules that dissociate at the Pd surfaces as a function of the molecular beam energy and incident angle. Measurements on the Pd(1 1 1) surface confirm “normal energy scaling” for the methane dissociative sticking probability, consistent with an activation barrier normal to the surface, although there may be additional barriers in other degrees of freedom, but with little corrugation parallel to the surface. Sticking measurements on supported Pd particles (∼3 nm wide) with the methane beam directed normal to the MgO(1 0 0) surface results in a large fraction of the methane/Pd collisions occurring on regions of the particles where the beam direction is far from the local particle surface normal, resulting in lower sticking probability. We attempt to decouple this effect from the measured sticking probabilities in order to compare the intrinsic reactivity of the Pd particles with Pd(1 1 1). We find that the sticking probability on ∼3 nm Pd particle surfaces is at most twice as large as on Pd(1 1 1). This result depends on our assumption that these annealed Pd particles have the known equilibrium particle shape (truncated half octahedron). We also discuss the need for detailed structural knowledge of the particles and careful geometric analysis when probing direct collisional activation barrier crossing using molecular beams. Temperature programmed desorption studies of physisorbed (not dissociated) methane reveal that the Pd particles bind methane more strongly than Pd(1 1 1). Oxygen adsorbs on the Pd nanoparticles via a mobile, molecular O2 precursor state which is transiently adsorbed on the MgO(1 0 0) surface. An induction period is observed on Pd nanoparticles for the titration of adsorbed O by CO gas to make CO2 which is not observed on Pd(1 1 1). This is attributed to inhibition by adsorbed O, whose saturation coverage on the Pd particles is 41% greater than on Pd(1 1 1).