Particle size dependent adsorption and reaction kinetics on reduced and partially oxidized Pd nanoparticles

Abstract

Combining scanning tunneling microscopy (STM), IR reflection absorption spectroscopy (IRAS) and molecular beam (MB) techniques, we have investigated particle size effects on a Pd/Fe(3)O(4) model catalyst. We focus on the particle size dependence of (i) CO adsorption, (ii) oxygen adsorption and (iii) Pd nanoparticle oxidation/reduction. The model system, which is based on Pd nanoparticles supported on an ordered Fe(3)O(4) film on Pt(111), is characterized in detail with respect to particle morphology, nucleation, growth and coalescence behavior of the Pd particles. Morphological changes upon stabilization by thermal treatment in oxygen atmosphere are also considered. The size of the Pd particles can be varied roughly between 1 and 100 nm. The growth and morphology of the Pd particles on the Fe(3)O(4)/Pt(111) film were characterized by STM and IRAS of adsorbed CO as a probe molecule. It was found that very small Pd particles on Fe(3)O(4) show a strongly modified adsorption behavior, characterized by atypically weak CO adsorption and a characteristic CO stretching frequency around 2130 cm(-1). This modification is attributed to a strong interaction with the support. Additionally, the kinetics of CO adsorption was studied by sticking coefficient experiments as a function of particle size. For small particles it is shown that the CO adsorption rate is significantly enhanced by the capture zone effect. The absolute size of the capture zone was quantified on the basis of the STM and sticking coefficient data. Finally, oxygen adsorption was studied by means of MB CO titration experiments. Pure chemisorption of oxygen is observed at 400 K, whereas at 500 K partial oxidation of the particles occurs. The oxidation behavior reveals strong kinetic hindrances to oxidation for larger particles, whereas facile oxidation and reduction are observed for smaller particles. For the latter, estimates point to the formation of oxide layers which, on average, are thicker than the surface oxides on corresponding single crystal surfaces.

Particle-size dependent heats of adsorption of CO on supported Pd nanoparticles as measured with a single-crystal microcalorimeter

“…16 The surface of the Pd particles investigated in this work exposes mainly ͑111͒ terraces alongside with a smaller fraction of ͑100͒ facets and low-coordinated defect sites such as edges and corners. 10 According to theoretical calculations, such irregular sites exhibit higher CO adsorption energies than Pd͑111͒. 17 Experimentally, there is a general agreement that the CO adsorption energy on the ͑100͒ plane is by about 10 to 15 kJ mol −1 higher than on the ͑111͒ plane.…”

Selective Hydrogenation of Acrolein Over Pd Model Catalysts: Temperature and Particle‐Size Effects

“…[9] Briefly, a well-ordered ~10 nm thick Fe3O4 film was grown on a Pt(111) substrate (see references [10] for details) followed by Pd deposition onto the Fe3O4 film at 120 K by physical vapor deposition of Pd (Goodfellow, >99.9%) using a commercial evaporator (Focus EFM 3). After depositing Pd, the sample was annealed at 600 K and the Pd nanoparticles were stabilized by repeated cycles of oxidation and reduction at 500 K. [11] The size of the Pd nanoparticles was controlled by the nominal thickness of the Pd film deposited onto the Fe3O4 substrate at 120 K (see reference [9] for details). The Pd(111) crystal was cleaned by repeated cycles of Ar + sputtering at room temperature, annealing at 1000 K, and oxidation in 1x10 -6 mbar O2 at 750 K. The cleanliness of the Pd/Fe3O4 and Pd(111) samples was verified prior to every experiment by IRAS of adsorbed CO.…”

Properties of oxide thin films and their adsorption behavior studied by scanning tunneling microscopy and conductance spectroscopy

“…In most cases, the role of point defects in adsorption events was indirectly derived from the growth of metal deposits on flat oxide terraces in large distance to step edges and other apparent line defects. 48,251,429 The assumption of a defect-mediated nucleation mechanism was thereby rationalized by the higher binding potential of oxygen vacancies compared to regular surface sites, as calculated for many bulk oxides with DFT. 312,316,318 The actual atomic nature of the underlying defect and the dominant interaction mechanism remained therefore unclear in most of these studies.…”