ACS PRF | ACS
All e-Annual Reports
41278-AC5
The Structure of Reactivity of Palladium Oxide Surfaces
Palladium has been identified as the best catalyst for the total oxidation of hydrocarbons. Although ambient pressure studies have suggested that bulk PdO is the most active phase for this reaction, surface science studies have shown that oxygen inhibits the dissociative adsorption of hydrocarbons on Pd, the rate-limiting step in the reaction. Our work has been focused on resolving this conundrum, and to develop a general understanding of the conditions under which metal oxides are more reactive than metals. We have taken two approaches: 1) characterizing the reactivity of PdO clusters formed on Pd(100) single crystals by exposure to atomic oxygen; and 2) epitaxial growth of well-ordered PdO thin films. In this past year we completed characterization of the PdO clusters and worked on structural and chemical characterization of epitaxial PdO thin films grown in Co3O4(110) grown on top of magnesium aluminate (110).
Last year, we showed that oxidation of Pd(100) by atomic oxygen leads to the formation of large, poorly-ordered, three-dimensional PdO clusters on the surface. Temperature programmed desorption (TPD) and propene titration of the surface oxygen revealed a propene sticking coefficient near unity, 20 times higher than on the well-ordered surface oxide formed by exposure to O2. It was further shown that propene oxidation on the PdO clusters follows a direction reaction mechanism in which adsorbed molecules react to form CO2 and H2O simultaneously, in contrast to adsorbed oxygen phases on metallic Pd where propene dissociates and then the fragments are oxidized. Because the direct mechanism has a lower activation energy, the PdO clusters are more reactive than Pd metal. Meanwhile the high sticking coefficient on the bulk oxide clusters makes them more reactive than the surface oxide. This year we showed that there is little difference between how strongly bound oxygen is to surface and bulk oxides, and, using ion scattering spectroscopy, that the density of exposed Pd sites on the two surfaces are very similar. Thus, neither differences in chemical bonding of oxygen at the surface nor differences in the density of potentially reactive metal sites on the surface can explain the observed higher reactivity of bulk PdO clusters. Rather, it is suggested that structural differences between the well-ordered surface oxide and the poorly ordered PdO clusters are responsible for the differences in reactivity.
To test the idea that the surface oxide is less reactive than the bulk PdO clusters because hydrocarbon oxidation on PdO is structure sensitive, we have been working on growing and characterizing well-ordered epitaxial PdO. We have tried several different substrates and growth sequences in our attempt to grow epitaxial PdO. As noted in prior reports, we have had the greatest success using Co3O4(110) films grown on top of magnesium aluminate (110) as the substrate. During this past year, we found using reflection high energy electron diffraction (RHEED) that PdO grows smoothly and epitaxially on this substrate up to a thickness of roughly 7 nm, then roughens to more of a 3D hillock structure. X-ray photoelectron spectroscopy (XPS) indicated that as the films thicken they become slightly reduced, which may account for the roughening. X-ray diffraction of the films revealed two peaks that could be associated with the PdO film: one that corresponds quite well to PdO(110) and the other at small angles that does not correspond to any known phases containing Pd, Co and O. After annealing in flowing air at 900 K, the low angle peak disappeared while the PdO(110) peak intensified; reducing the sample caused the low angle peak to reappear. Thus, the low angle peak was attributed to the formation of reduced Pd-O phases with a long range periodicity; and it was concluded that the PdO grows with a [110]-orientation. In addition to XPS, we have also characterized the valence band structure of the epitaxial PdO with ultraviolet photoelectron spectroscopy (UPS) and the results showed an interesting transition in the structure as the film surface roughened. We have begun to characterize the reactivity of the ordered PdO surface using CO TPD and isothermal titration experiments. In accord with the expectation that structural order plays a large role in determining the reactivity of PdO, we have found that the ordered PdO is almost completely inert towards CO. We will be continuing this work by looking at the reaction of propene with epitaxial PdO.
It should also be noted that we have observed interesting structural transitions in the Co3O4(110) films we have grown as substrates. In the [110] direction the Co3O4 structure is composed of alternating Co2O2 and CoO2 planes. After growth, the surface exposes CoO2 planes but annealing in flowing air produces a Co2O2 termination, the opposite of what would be expected. Although this is not central to the original project, we have pursued this interesting finding and will be publishing a paper on it.
