Article pubs.acs.org/JPCC
Water Dissociation on CeO2(100) and CeO2(111) Thin Films David R. Mullins,*,† Peter M. Albrecht,† Tsung-Liang Chen,† Florencia C. Calaza,† Michael D. Biegalski,‡ Hans M. Christen,‡ and Steven H. Overbury† †
Chemical Sciences Division and ‡Center for Nanophase Materials Sciences, Oak Ridge National Laboratory, Oak Ridge, Tennessee 37831-6201, United States ABSTRACT: This study reports and compares the adsorption and dissociation of water on oxidized and reduced CeO2(100) and CeO2(111) thin films. Water adsorbs dissociatively on both surfaces. On fully oxidized CeO2(100) the resulting surface hydroxyls are relatively stable and recombine and desorb as water over a range from 200 to 600 K. The hydroxyls are much less stable on oxidized CeO2(111), recombining and desorbing between 200 and 300 K. Water produces 30% more hydroxyls on reduced CeO1.7(100) than on oxidized CeO2(100). The hydroxyl concentration increases by 160% on reduced CeO1.7(111) compared to oxidized CeO2(111). On reduced CeO1.7(100) most of the hydroxyls still recombine and desorb as water between 200 and 750 K. Most of the hydroxyls on reduced CeO1.7(111) react to produce H2 at 560 K, leaving O on the surface. A relatively small amount of H2 is produced from reduced CeO1.7(100) between 450 and 730 K. The differences in the adsorption and reaction of water on CeOX(100) and CeOX(111) are attributed to different adsorption sites on the two surfaces. The adsorption site on CeO2(100) is a bridging site between two Ce cations. This adsorption site does not change when the ceria is reduced. The adsorption site on CeO2(111) is atop a single Ce cation, and the proton is transferred to a surface O in a site between three Ce cations. When the CeOX(111) is reduced, vacancy sites are produced which allows the water to adsorb and dissociate on the 3-fold Ce cation sites.
1. INTRODUCTION Cerium oxide is an important component in many catalytic processes.1,2 Many of these catalytic reactions involve water either directly as in the water−gas shift (WGS) reaction, as a product as in the dehydration of organic oxygenates, or as a spectator during reactions in an aqueous environment or in the exhaust stream of an automotive three-way catalyst. An understanding of water’s interaction with a ceria surface also provides insight for other molecules containing an −OH functionality such as alcohols and organic acids. One of the principal issues with regards to the interaction of water with a surface is whether the molecule adsorbs molecularly or whether it dissociates into (OH)− and H+. The adsorption of water on surfaces has been extensively reviewed by Thiel and Madey3 and by Henderson.4 In general, water appears to have a weaker interaction, and is less likely to dissociate, on metal surfaces than on oxide surfaces. The strength of adsorption and the likelihood for dissociation on oxide surfaces is related to the presence, accessibility, and coordination of metallic cations on the surface. Therefore, the interaction of water with a particular oxide is expected to be strongly dependent on the structure of the oxide surface. CeO2(111) and CeO2(100) produce very different environments for the Ce4+ cations. CeO2(111) is a Tasker Type 2 oxide5 with a repeating A-B-A-A-B-A layer structure, where A is the oxygen anion and B is the metal cation. The elemental concentration in each layer is the same which maintains charge neutrality and results in no surface dipole. CeO2(100) is a Tasker Type 3 oxide with an A-B-A-B layer structure. This © 2012 American Chemical Society
surface is unstable due to a surface dipole. This dipole can be eliminated, and thereby the surface can be stabilized, by moving half of the O from one face and placing it on the opposing face.6 Surface structures for CeO2(111) and CeO2(100) are shown in Figure 1. On CeO2(111) the O and Ce each have a single coordination vacancy. In addition, the presence of the O in the top layer and the spacing between Ce cations make it unlikely that an adsorbate could bind to more than one Ce cation site, e.g., bridge-bonded. The surface termination for CeO2(100) is less obvious, but the structure shown in Figure 1 has been the one most frequently proposed based on both experimental observations7,8 and theoretical calculations.6,9 On this surface the O and the Ce each have two coordination vacancies. In addition, an adsorbate could bridge-bond between two adjacent Ce cations in a position identical to the O in the top layer. There have been several studies of water on ceria surfaces. On fully oxidized CeO2(111) experimental observations have generally led to the conclusion that water adsorbs molecularly and desorbs below room temperature.10,11 Different theoretical studies of water adsorption on CeO2(111) have come to contradictory conclusions. Watkins et al.12 concluded that water will dissociate on oxidized CeO2(111), whereas Fronzi et al.13 and Marrocchelli and Yildiz14 concluded it would not. Recently, Molinari et al.9 calculated that dissociation is favored Received: June 29, 2012 Revised: August 15, 2012 Published: August 16, 2012 19419
dx.doi.org/10.1021/jp306444h | J. Phys. Chem. C 2012, 116, 19419−19428
The Journal of Physical Chemistry C
Article
Henderson et al. observed no H2 desorption from reduced CeO2−X(111) and Ce 3d XPS indicated further reduction of the Ce in response to water adsorption.10 Alternatively, in our previous work we observed a significant amount of H2 desorption, a weak indication of increased oxidation in the Ce 3d and Ce 4d XPS following a single TPD cycle, and significant reoxidation if the reduced surface was exposed to water above the H2 desorption temperature.11 Our more recent results support our earlier observations, and we will consider possible explanations as to why these results differ so dramatically from those of Henderson et al. Experimental observations of water adsorption have not been previously reported on reduced CeO2−X(100). The calculations of Molinari et al.9 indicate that dissociation will be more favorable on the reduced surface compared to the oxidized surface; however, the effect is not as significant as in the case of CeO2−X(111) vs CeO2(111). Our results are consistent with this prediction in that the OH concentration increases on CeO2−X(100) but by a comparatively modest amount relative to CeO2−X(111).
2. EXPERIMENTAL SECTION CeO2(100) films were grown at the Center for Nanophase Materials Sciences at Oak Ridge National Laboratory (ORNL) by pulsed laser deposition on 0.05% Nb-doped SrTiO3(100) (Nb-STO). Similarly grown films were used in previous studies of CeO2(100).20,21 The Nb-dopant made the STO substrate conductive. The films were grown using a CeO2 ceramic target while the Nb-STO substrate was at 550 °C in a background pressure of 50 mTorr O2 and with a laser fluence of 1.4 J/cm2. The film growth was monitored using reflection high-energy electron diffraction (RHEED), and the films were grown to a thickness of 19 ± 1 nm. X-ray diffraction shows that the films are epitaxial with the CeO2 L00 peaks oriented out of plane with a lattice constant of 5.433 ± 0.002 Å (Figure 2a). The CeO2 lattice is 39% bigger than the STO lattice. However, good epitaxy along [100] can be achieved if the CeO2(100) lattice is rotated by 45° relative to that of the Nb-STO(100) substrate (inset, Figure 2b). This epitaxial relationship is demonstrated in the XRD azimuthal scans (Figure 2b) where the positions of the CeO2 ⟨101⟩ and the SrTiO3 ⟨101⟩ differ by 45°. Auger electron spectroscopy and XPS indicated that the CeO2(100) surface had S, C, K, and Cl impurities. The S and C were removed by exposure to O2 (1 × 10−6 Torr, 800 K, 5−10 min). The K and Cl were removed by gentle sputtering (