AVS2014 Session SS+AS+EN-TuM: Synthesis, Structure and Characterization of Oxides

Cerium oxide is a principal component in many heterogeneous catalytic processes. One of its key characteristics is the ability to provide or remove oxygen in chemical reactions. Recent work has demonstrated how the reactivity and selectivity of various molecules are dramatically altered on different crystallographic faces of cerium oxide. The structure and composition of different faces determine the number of coordination vacancies (CV) surrounding surface atoms, the availability of adsorption sites, the spacing between adsorption sites and the ability to remove O from the surface. The Ce cation sites are less accessible and have fewer coordination vacancies (CV) on CeO₂(111) than on CeO₂(100). Even though the Ce is in the second layer on CeO₂(100), molecules can adsorb in the open bridge sites between two Ce cations. While there have been numerous studies of the adsorption and reaction of various molecules on CeO₂(111) only recently have comparable experiments been conducted on CeO₂(100).

To investigate the role of surface orientation on reactivity, CeO₂ films with different orientations were grown by two different methods. CeO₂(100) films were grown ex situ by pulsed laser deposition on Nb-doped SrTiO₃(100). CeO₂(111) films were grown in situ by thermal deposition of Ce metal onto Ru(0001) in an oxygen atmosphere. The chemical reactivity was characterized by the adsorption and decomposition of various molecules such as CO₂, H₂O, alcohols, aldehydes and organic acids. In general the CeO₂(100) surface was found to be more active, i.e. molecules adsorbed more readily and reacted to form new products, especially on a fully oxidized substrate. However the CeO₂(100) surface was less selective with a greater propensity to produce CO, CO₂ and water as products. Experiments are underway to determine if CeO₂(110), where the Ce adsorption sites are in the top layer and have 2 CV but the O has only 1 CV, will produce an active yet more selective catalyst.

It is possible to synthesize high surface area shape-selected nanoparticles (octahedra and cubes), i.e. powders that expose a single, well-defined surface. Experiments have shown similarities between the single crystal surfaces and shape-selected nanoparticles, e.g. CeO₂(111)/octahedra are less active than CeO₂(100)/cubes. However there have also been significant differences in selectivity and the types of products formed. Possible explanations for the differences on the single crystal surfaces vs. the nanoshapes will be considered.

Research sponsored by the US Department of Energy, Office of Science, Basic Energy Sciences, Chemical Sciences, Geosciences, and Biosciences Division.

The growth morphology and atomic geometry of ceria nanostructures on Cu(110) have been investigated by STM, LEED and XPS. Ceria grows epitaxially in a two-dimensional (2-D) hexagonal layer, which is associated with a CeO₂(111)-type trilayer structure forming a (3x11) coincidence lattice. An important experimental parameter is the oxygen pressure during growth: it influences the stoichiometry of the ceria overlayer as well as the Cu surface oxide phase, which coexists with the ceria for coverages below the full monolayer. For oxygen pressures in excess of 10^-7 mbar, stoichiometric CeO₂ and coexisting Cu-c(6x2) surface oxide are formed, whereas for lower oxygen pressures, in the 10^-8 mbar range, slightly substoichiometric ceria (CeO_~1.9) and a Cu-(2x1) surface oxide are observed. The ceria overlayer grows essentially 2-D, but displays a peculiar nanostripe pattern, with varying periodicities ranging from 4-8 nm and a corrugation amplitude of 0.2-0.3 nm. This nanostripe pattern is due to a topographic modulation of the overlayer caused by the frustration of overlayer-substrate bonding as a result of the epitaxial mismatch at the ceria-Cu interface. Detailed STM investigation reveals a distortion of the ceria lattice in the transition region between dark (low) and bright (high) stripes, which gives rise to periodic regions of anisotropic lattice strain - socalled "lattice strain defects". It is speculated that these lattice strain defects may support particular chemical reactivity.

Work supported by the ERC Advanced Grant "SEPON" and by the COST Action CM1104

Zinc oxide is a wurtzite-structured polar crystal with dramatic polarization direction-dependent surface chemistry. Meanwhile, chromium III oxide is a non-polar material catalytically active for a number of industrial chemical reactions, most notably dehydrogenation. Therefore, the Cr₂O₃/ZnO system has been chosen to demonstrate how the polarization direction of a substrate can be exploited tailor the surface properties of catalytically active non-polar thin films. Photoelectron spectroscopy and electron diffraction have been performed to determine the growth mode as well as the film quality. The growth is 2D; however, the films appear initially disordered on both positive and negatively poled substrates. On both substrates the order was observed to improve with thickness. Small band offsets between Cr₂O₃ and oppositely poled substrates were observed that were consistent with charge compensation at the Cr₂O₃/ZnO interface. No obvious change in the oxidation state of the chromium was seen and so it is assumed that the charge compensation at the interface only involves Zn and/or O. The offset between the Cr peaks on positively and negatively poled substrates was obvious at the initial growth stages but then decayed with film thickness, suggesting that the compensating charges at the interface may migrate to the film surface. Valence band spectra were analyzed to characterize the bonding at the interface. The surface chemical behavior of Cr₂O₃ on the two zinc oxide surfaces is being characterized to determine how significantly the polar interface impacts the surface properties of thin supported layers.

X-ray satellite lines come about in x-ray fluorescence spectroscopy as a result of shake-off events in the excitation process. The ratio of their intensities has been shown to be a sensitive function of their oxidation states. We are able to observe the K satellite lines in the x-ray spectra of oxides and glasses, excited by the beam of an electron microscope and detected by a high resolution x-ray microcalorimeter detector. The results show surprising departures from the expected states of some metal elements that we expect to be fully oxidized. The satellite ratio is a way of determining the chemical environment of insulators without charging or ultra-high vacuum.

The intrinsic point defects associated with oxygen vacancies and Ti³⁺ ions play a crucial role in the usage of titanium dioxide (TiO₂) in various technological applications including catalysis and photochemistry. It is well known that the interactions between H atoms and surface oxygen in TiO₂ lead to the formation of Ti³⁺ ions at elevated temperatures. However the Ti³⁺ ion formation and accumulation as a function of elevated temperatures in UHV conditions during hydrogen diffusion in TiO₂ is not well understood. In this study, we have used ion implantation method to incorporate hydrogen in single crystal TiO₂ (110) samples and investigated the behavior of point defects in both pure and hydrogen implanted TiO₂ as a function elevated temperatures using Rutherford backscattering spectrometry (RBS), nuclear reaction analysis (NRA), x-ray photoelectron spectroscopy (XPS) and ultra violet photoemission spectroscopy (UPS). TiO₂ single crystals were implanted with 40 keV hydrogen ions at room temperature with ion fluences of 1x10¹⁵, 1x10¹⁵ and 1x10¹⁷ atoms/cm². Samples were isochronally annealed in vacuum for 30 minutes at each temperature up to 1100K and hydrogen and Ti³⁺ defects were quantified. Hydrogen depth profile measurements obtained from 1x10¹⁷ atoms/cm² implanted sample reveal that hydrogen diffused towards the surface at lower temperatures and it slowly diffuses out from the samples at higher temperatures. XPS and UPS measurements from the hydrogen implanted samples show significantly higher Ti³⁺ defects in comparison to pure TiO₂ at these temperatures under UHV conditions. These defects reach a maximum around 880 K in which almost all hydrogen was removed from the sample. When the implanted sample further annealed to high temperatures, the amount of Ti³⁺ in hydrogen implanted samples started to decrease and reaches the values from the pure TiO₂ samples around 1100K.