Cerium Oxide Nanostructures on Titania: Effect of ... - ACS Publications

Subscriber access provided by Kaohsiung Medical University

C: Surfaces, Interfaces, Porous Materials, and Catalysis

Cerium Oxide Nanostructures on Titania: Effect of the Structure and Stoichiometry on the Reactivity Toward Ethanol Oxidation Luca Artiglia, and Stefano Agnoli J. Phys. Chem. C, Just Accepted Manuscript • DOI: 10.1021/acs.jpcc.8b05807 • Publication Date (Web): 14 Aug 2018 Downloaded from http://pubs.acs.org on August 20, 2018

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 21 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

The Journal of Physical Chemistry

Cerium Oxide Nanostructures on Titania: Effect of the Structure and Stoichiometry on the Reactivity Toward Ethanol Oxidation Luca Artiglia*1,2, Stefano Agnoli*1 1

Dipartimento di Scienze Chimiche, Universita’ degli Studi di Padova, 35100 Padova, Italy.

2

Current address: Laboratory of Sustainable Chemistry and Catalysis and Laboratory of Environmental

Chemistry, Paul Scherrer Institute, 5232 Villigen, Switzerland. *

Corresponding authors email addresses: [email protected], [email protected].

ACS Paragon Plus Environment

1

The Journal of Physical Chemistry 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 2 of 21

Abstract

Nanocatalysis is a multisciplinary field in which catalytic reactions are explored, understood and improved by means of surface science methods. Model systems are a good playground to develop new interfaces where uncommon geometries and special active sites are stabilized. Cerium oxide deposited on titanium dioxide represent a good example of this. Thanks to a charge transfer from the titania support, Ce3+ sites grow at the interface. Thick cerium oxide layers show both the Ce3+ and Ce4+ oxidation states, whereas only Ce3+ is detected in the monolayer. This demonstrates that the charge transfer is limited to the interface. In the present study we grew and characterized two cerium oxide layers having different thicknesses, namely 2 and 6 monolayer equivalents, on rutile (110), and investigated their reactivity toward the partial oxidation of ethanol. Our results show that the selectivity and the reaction mechanism are affected by the thickness of the cerium oxide. In particular, a stoichiometric reaction involving surface oxygen sites takes place on the thinner reduced film whereas at higher ceria coverage the presence of the Ce3+/Ce4+ redox couple favor a catalytic reaction in which gas phase oxygen is activated on ceria and then reacts with adsorbed ethanol.


2

Page 3 of 21 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Introduction

Cerium oxide (CeO2) is well known and widely used in catalysis due to its ability to store and release oxygen.1,2,3 Either as a support for metal nanoparticles or doped with transition metal oxides, it has demonstrated to improve the activity, selectivity and thermal stability of catalysts, especially toward the oxidation of alcohols and hydrocarbons.4,5 Moreover, it has found application in several fields, such as the removal of soot from diesel engine exhaust,6,7 the removal of pollutants from wastewaters8 and in fuel cell technology.9,10 One of the main tasks in the field of nanocatalysis is to try to stabilize new active chemical species (e.g. coordinatively unsaturated sites)11 that can be created either forcing uncommon bond geometries or by interfacial hybridization of electronic states.12,13,14 A promising strategy, so far poorly explored, is to create mixed oxide materials where the overlap of the electronic states causes a charge transfer that stabilizes highly active sites at the interface. Model studies, where oxide/oxide interfaces are prepared in very controlled conditions and can be modeled by exploiting density functional theory calculations,15,16,17,18 are useful as a base to understand how to design the synthesis of new catalysts showing the same special properties.19 In this work we deposited selected amounts of CeOx on a reducible oxide single crystal support, TiO2rutile (110), which is the thermodinamically most stable surface termination of rutile and has been widely characterized both from the experimental and theoretical point of view.20 Theoretical calculations showed that Ce3+ sites are stabilized at the CeOx/TiO2 interface due to the energy decrease of the Ce 4f levels as a result of the mixing with the O 2p band of titania.21 In agreement with this work, Agnoli et al. demonstrated that thin CeOx nanostructures deposited by reactive molecular beam epitaxy on TiO2-rutile (110) are fully reduced.22 As the thickness of the overlayer increases, and thus CeOx does not experience any interaction with TiO2, stoichiometric ceria nanoislands start to grow. This paper shows that cerium oxide overlayers deposited on titania, whose structure and defectivity depend on their thickness, display a distinctive reactivity and selectivity with respect to ethanol oxidation.

Experimental The TiO2-rutile(110) single crystal (Mateck) was cleaned by several Ar+ sputtering cycles, followed by annealing in ultra high vacuum (UHV) at 970 K. After sputtering and annealing, low energy electron diffraction (LEED) revealed a sharp (1x1) surface reconstruction. CeOx films were deposited at 670 K ACS Paragon Plus Environment

3


Page 4 of 21

by electron beam evaporation of the metal in an oxygen background (1.0 x 10-6 mbar), followed by annealing at 970 K and cool down at the same oxygen pressure until 450 K. In the following work, we will show two CeOx coverages, i.e. about 2 and 6 monolayer equivalents (MLE). The MLE was determined based on angle resolved X ray photoelectron spectroscopy (XPS), acquiring the Ce (3d) and Ti (2p) core level peaks at different polar angles (with respect to the sample surface).23 Temperature programmed reaction (TPR) spectra were acquired in a multi-technique chamber equipped with a manipulator having 4 degrees of freedom, a 4 grid rear view LEED, an X ray photoelectron spectrometer (XPS - double anode soft x ray source and hemispherical single channeltron analyzer) and a quadrupole mass spectrometer. The TiO2 sample was fixed, by means of thermally conductive (dielectric) glue, to a Ta frame and connected to the manipulator end through 0.2 mm thick Ta wires. A K-type thermocouple was glued to the sample rear, so that the temperature of the TiO2 crystal could be monitored in real time. A precise control of the temperature was achieved through a PID controller coupled to a power supply. Such a configuration of the sample holder minimized its contribution to the desorption spectra. The quadrupole mass spectrometer is covered by a quartz shield to avoid any contribution to the desorption spectra arising from the chamber walls. During TPR, the circular aperture (6 mm diameter) of the quartz shield of the quadrupole was positioned at about 3.0 mm distance with respect to the sample surface. A typical TPR experiment consisted in: i) cool down of the sample with liquid nitrogen (ca. 120-130 K). ii) dosing of the reactant/s through a variable leak valve. The amount of gas was quantified in Langmuirs (L), where 1.0 L corresponds to 1.0 second dosing at the pressure of 10-6 Torr. iii) heating of the sample following a linear desorption ramp (β=heating rate=2 K/s). We dosed deuterium labelled ethanol (C2H5OD) to simplify the analysis of the reaction products. The alcohol (> 99.5 atom% D) was obtained from Sigma Aldrich. It was purified by several freeze-pumpthaw cycles before use. Although the C2H5OD was pure and the gas line was baked, we could not avoid the accidental co-dosing of water during the experiments. Based on previous literature report,24 the isotopic impurity of C2H5OD is in the order of 15%. Ethanol was monitored at mass 45 and 46 (the sum of mass 45 and 46, normalized to the relative intensity in the cracking pattern, was used to evaluate the selectivity shown in Table 1 and 2), water at mass 18 (H2O) and 20 (D2O), acetaldehyde at mass 29 and ethylene at mass 27. The m/z signals were corrected considering the cracking pattern of ethanol obtained by our quadrupole mass spectrometer as a function of its signal intensity.17 XPS of the Ce 3d core levels was acquired using Al Kα X-ray photon energy (hv=1486.7 eV). Resonant photoemission (ResPes) was acquired at the Elettra Synchrotron (Trieste), Materials Science beamline, ACS Paragon Plus Environment

4

Page 5 of 21 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


using linearly polarized light.25 The valence band spectra were acquired across the Ce 4d photoabsorption threshold.

Results and Discussion For convenience we will hereafter name the 2 and 6 MLE CeOx coverages as 2ML and 6ML, respectively. Previous scanning tunneling microscopy (STM) showed that, when the cerium oxide coverage is low (between 1 and 2 MLE), a superstructure having a rectangular unit cell and extended line defects grows epitaxially on the TiO2.22 None of the low index surfaces of CeO2 or Ce2O3 seems to match this structure. The 2ML structure could be due either to highly strained ceria surfaces or to an unknown mixed oxide comprising both cerium and titanium atoms epitaxially stabilized by the interaction with the substrate.22 The 6ML sample did not show any LEED pattern, neither coming from the TiO2 support (the 1x1 reconstruction of TiO2-rutile (110) is attenuated by the CeOx overlayer) nor from a new superstructure. The photoemission spectra of the Ce 3d core levels are reported in Figure 1 for the two coverages. The 2ML (green spectrum) shows the typical lineshape of fully reduced cerium oxide, i.e. two 3d3/2-3d5/2 spin-orbit splitting doublets, whose 3d5/2 peaks are found at 881.6 eV (labelled v0) and 885.6 eV binding energy (BE), and represent the different 4f configurations in the final state of Ce3+.26,27,28 The spectrum of the 6ML-CeOx/TiO2(110) (black) shifts toward higher binding energy as compared to the former. A new peak at 883.1 eV, labelled as v and associated with Ce4+, can be seen in the 3d5/2 region. The v0 component is still present as a low BE shoulder. A new satellite appears at 917.0 eV (labelled as uiii in Figure 1), which is a distinctive benchmark of Ce4+.27 Therefore, the photoemission data demonstrates that the 2ML thick film corresponds to an almost fully reduced cerium oxide, whereas the 6ML thick sample contains both the 3+ and the 4+ oxidation states. These results, in agreement with the study of Agnoli et al.,22 suggest that CeOx grown at the interface with the TiO2 substrate is stabilized in a reduced state. Such an effect has been demonstrated both theoreticaly (in the case of vanadia/ceria)29 and experimentally (for both vanadia/ceria and ceria/titania)1,15,30,31 The interfacial hybridization leads to an oxide ceria charge transfer and thus to a local reduction of cerium oxide.29 As the coverage is increased, and multilayer structures grow on the support, the surface hybridization effect is lost and stoichiometric ceria forms. At variance with these results, the deposition of cerium on metal substrates (such as Cu(111)) in similar oxygen reactive backgrounds, leads immediately to the growth of stoichiometric CeO2.32,33,34


5


Page 6 of 21

Resonant photoemission of the valence band across the Ce 4d-4f photo-absorption region is an efficient method to reveal the 4f occupancy of cerium oxide (presence of Ce3+ sites) with very high sensitivity.35,36 Figure 2 shows the valence band of the 2ML and 6ML samples on the left and right side, respectively, acquired at increasing excitation photon energies in the 114-126 eV photon energy range. Figure 2 a and b were acquired in ultra high vacuum (UHV), Figure 2 c and d were acquired after dosing 3000 L of oxygen on the samples at room temperature and Figure 2 e and f were acquired after dosing 60 L of ethanol. The spectra collected off-resonance (hv=114 eV) show a broad valence band that develops between 3 and 8 eV: in the case of 2ML, the contribution of the TiO2 substrate states, i.e. O 2p and Ti 3d,20,37 is still predominant whereas the 6ML spectrum shows a more complex lineshape, due to the ceria derived states (6s, 5d).33 In both cases, when the photon energy increases, a peak centered at about 1.5 eV is detected, which is attributed to the 4f electrons of Ce3+. Its maximum intensity is found exciting with 121.0 eV energy photons. Also the region between 2.0 and 8.0 eV shows some modifications, which are well evident in the 6ML sample. In particular, a peak at about 4.0 eV, associated with the f0 electronic configuration of Ce4+ ions, appears above 119.0 eV and reaches its maximum intensity at 124.0 eV. In good agreement with our results, other ResPes studies of cerium oxide demonstrate that the Ce3+ valence states resonate strongly at 121.0 eV whereas Ce4+ around 124.0 eV.33,34,36,38 By subtracting the VB data acquired off-resonance, i.e. at 114.0 eV, to those acquired in resonance conditions, i.e. at 121.0 eV for Ce3+ and 124.0 eV for Ce4+, it is possible to outline selectively the signals coming from Ce3+ (at 1.5 eV BE) and of Ce4+ (at 4.0 eV BE). Figure 3 shows an example of the subtraction procedure operated on the VB spectra of clean 2ML and 6ML (dashed lines correspond to the difference spectra). As mentioned above, the peak at 1.5 eV (DCe3+) is visible on both samples and its relative intensity is peaked at 121.0 eV. On the contrary, the feature at 4.0 eV (DCe4+) is barely visible in the 2ML sample, whereas becomes extremely intense for the 6ML film. This is in good agreement with XPS results and proves that the 2ML sample is almost completely reduced, because ceria grows at the interface with the TiO2 support and a supportoverlayer charge transfer stabilizes the Ce3+ species. On the contrary, stoichiometric ceria is detected on the 6ML sample, because the external layers of thicker islands do not grow in contact with the support. Based on the results of Figure 3a and b, we estimated the relative variation of the oxidation state of cerium by calculating the resonant enhancement ratio (RER), which is defined as the ratio of the resonant peaks DCe3+/DCe4+ at different photon energies.35 The RERs (hv=121.0 and 124.0 eV) calculated for clean samples are reported in Figure 3c. As expected, the values calculated at 124.0 eV are lower than those


6

Page 7 of 21 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


at 121.0 eV, because Ce4+ is found on both samples. This proves that a small amount of CeO2 grows also at low coverage, probably because the thickness of the CeOx islands is not uniform. To test the reactivity of the samples toward the oxidation of ethanol we performed both ResPes and TPR experiments either after dosing a mixture of oxygen and alcohol or the alcohol only. Depending on the intrinsic nature of the sample, deuterated ethanol (C2H5OD) can be converted to different products: the oxidative dehydrogenation reaction path yields acetaldehyde (C2H4O), whereas ethylene (C2H4) is formed via dehydration. ResPes can give indications about the oxidation state of the sample, whereas TPR, through the analysis of carbon containing species, can provide a full overview about the reaction path and selectivity. From the analysis of the VB evolution across the Ce 4d photoabsorption region, we could evaluate the RER after dosing O2 and C2H5OD (Figure 3c). In the case of 6ML, both the RERs calculated at 121.0 and 124.0 eV show the same behavior, i.e. a decrease after oxygen exposure followed by an increase to the value corresponding to clean CeOx/TiO2 after dosing ethanol. The 2ML behaves differently, indeed the RERs are almost constant after the exposure to oxygen (a small decrease is observed for RER calculated at 121.0 eV) then they increase after dosing ethanol. Because the RER is calculated as DCe3+/DCe4+ at a specific photon energy value, such an increase could be due to the reaction between ethanol and active sites found at the surface of multilayer islands (O(s)), which can be considered a source of reactive oxygen species. We can speculate a stoichiometric reaction between the oxide and the alcohol probably following a Mars van Krevelen mechanism. The behavior of 6ML is extremely reversible: after the adsorption of oxygen Ce3+ is partially oxidized to Ce4+, suggesting that gas phase oxygen reacts with surface Ce3+ sites oxidizing them. The initial stoichiometry is restored dosing ethanol indicating that a catalytic reaction takes place. TPR data are shown in Figure 4 and in Figure 5 for the 2ML and 6ML sample, respectively. We want to remark that we acquired also m/z signals corresponding to other possible products, i.e. carbon monoxide (CO), carbon dioxide (CO2) acetic acid (C2H4O2), diethyl ether (C4H10O) formaldehyde (CH2O), ethane (C2H6) and methane (CH4), but none of them could be detected. We show the spectra acquired after adsorption on 2ML and 6ML at low temperature (T

Cerium Oxide Nanostructures on Titania: Effect of ... - ACS Publications

Recommend Documents