Influence of TiO2 Bulk Defects on CO Adsorption ... - ACS Publications

Feb 7, 2017 - ABSTRACT: Electronic metal−support interactions (EMSIs) are demonstrated to severely affect the CO oxidation activity and the CO adsor...
0 downloads 11 Views 535KB Size
Subscriber access provided by University of Newcastle, Australia

Article 2

Influence of TiO bulk defects on the CO adsorption and CO oxidation on Au/ TiO – Electronic metal-support interactions (EMSI) in supported Au catalysts 2

Yuchen Wang, Daniel Widmann, and R. Jürgen Behm ACS Catal., Just Accepted Manuscript • DOI: 10.1021/acscatal.7b00251 • Publication Date (Web): 07 Feb 2017 Downloaded from http://pubs.acs.org on February 12, 2017

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 free 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 accessible to all readers and 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.

ACS Catalysis 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 25

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

ACS Catalysis

Influence of TiO2 bulk defects on the CO adsorption and CO oxidation on Au/TiO2 – electronic metal-support interactions (EMSI) in supported Au catalysts Yuchen Wang, Daniel Widmann and R. Juergen Behm* Institute of Surface Chemistry and Catalysis, Ulm University, D-89069 Ulm, Germany Abstract Electronic metal-support interactions (EMSI) are demonstrated to severely affect the CO oxidation activity and the CO adsorption properties of Au/TiO2 catalysts. Bulk oxygen vacancies, generated by a strongly reductive pre-treatment of Au/TiO2 at elevated temperature in 10% CO/N2, significantly lower the catalytic activity for CO oxidation at 80°C. With time on stream, the activity slowly increases until reaching the same steady-state value as obtained for a previously calcined and, hence, defect poor Au/TiO2 catalyst (activation period), where the time required for the activation period decreases with reaction temperature, but is independent from the oxygen partial pressure. Considering the similar Au particle size and Au loading, we conclude that the different activities originate from the presence of bulk oxygen vacancies generated during pre-treatment, which are slowly replenished during reaction. Insitu IR spectroscopy measurements reveal that the lower activity in the presence of bulk defects is coupled with and likely results from a strong modification of the CO adsorption strength on the reduced Au/TiO2 catalysts due to electronic metal-support interactions (EMSI). A possible mechanism explaining how these EMSI may be induced by the presence of bulk defects is discussed. Keywords: Au/TiO2, CO oxidation, CO adsorption, Oxygen vacancies, Bulk defects, Electronic Metal-Support Interactions (EMSI), Reaction mechanism, Kinetic measurements *Corresponding author:

Fax: Homepage: Email:

(++49) 731 502 5452 www.uni-ulm.de/iok [email protected]

ACS Paragon Plus Environment

ACS Catalysis

Page 2 of 25

2 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 Strong-metal support interactions (SMSI) are well known to severely affect the catalytic properties of catalysts supported on reducible oxides.1-3 For a variety of catalysts it was established that this is related with a change in the catalysts structure, by partial encapsulation of the catalytically active metal nanoparticles by a thin layer of reduced oxide.2;4-6 This lowers the accessible surface area of these NPs, and on the other hand creates a new surface phase (reduced oxide). More recently it was shown, however, that the oxide support may affect the catalyst activity also in a different way, by modifying the electronic properties of the active metal NPs.7 These electronic metal-support interactions (EMSI) were demonstrated to significantly improve, e.g., the ability of a Pt/CeO2 catalyst to dissociate the O-H bonds in water.7;8 In the present contribution we demonstrate that similar interactions can also strongly modify the activity of oxide supported Au catalysts, using the CO oxidation reaction on TiO2 supported catalysts as example. Specifically, we will demonstrate for bulk reduced Au/TiO2 catalysts that the presence of oxygen bulk vacancies in the TiO2 support strongly affects the CO oxidation activity. Au catalysts consisting of small Au nanoparticles supported on metal oxides, which have been introduced by Haruta et al.,9 are well known for their high activity in various oxidation and reduction reactions.10;11 Prime example and most intensively studied is the CO oxidation reaction, where these catalysts were shown to be highly active already at low temperatures, around room temperature and even below.11-13 Despite enormous efforts the physical origin of the high activity of these catalysts is still under debate (see reviews in

11;14;15

) It is, however,

well established that under typical reaction conditions, at room temperature and above, Au catalysts supported on reducible metal oxides (e.g., Au/TiO2, Au/CeO2, and Au/Fe2O3) are significantly more active than Au catalysts based on non-reducible oxides (e.g., Au/Al2O3 and Au/SiO2).14;16-18 This was rationalized by different dominant reaction pathways on the soACS Paragon Plus Environment

Page 3 of 25

ACS Catalysis

3 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

called active (reducible) and inert (non-reducible) catalysts under these reaction conditions. For Au catalysts based on reducible metal oxides we recently proposed a Au-assisted Marsvan Krevelen mechanism as dominant, rate determining reaction mechanism for reaction under dry conditions in the above temperature range.19;20 This reaction pathway includes the continuous formation and replenishment of surface oxygen vacancies on the metal oxide support, at the perimeter of the Au nanoparticles, upon interaction with CO or O2, respectively. In the meantime, this has been supported also by other experimental and theoretical studies.21-23 On Au catalysts based on non-reducible oxides, in contrast, this is not possible, and the CO oxidation reaction is proposed to proceed dominantly via a gold only mechanism,14;16;18 where CO and O2 adsorption and activation take place directly on the surface of the Au NPs.24;25 In the Au-assisted Mars-van Krevelen mechanism on active Au catalysts, the nature of the supporting oxide mainly appeared in its ability to support the generation and replenishment of surface oxygen vacancies at the perimeter of the supporting oxide, while bulk oxygen vacancies had found little attention so far. They are subject of the present study, where we investigated the influence of TiO2 bulk defects on the catalytic performance of Au/TiO2 catalysts in the CO oxidation by a combination of kinetic and in-situ IR spectroscopic measurements. Comparing the reaction characteristics of Au/TiO2 catalysts pre-treated in strongly oxidizing (O2/N2) or in strongly reducing (CO/N2) atmospheres at elevated temperatures (400°C), we want to elucidate effects of TiO2 bulk defects formed upon interaction with CO on the catalytic performance (activity and stability during CO oxidation) and on the CO adsorption behavior at a typical reaction temperature of 80°C. Based on these results we will demonstrate that both the metal-COad interactions and the activity for CO oxidation are strongly affected by electronic metal-support interactions (EMSI) in the reduced Au/TiO2 catalyst.

ACS Paragon Plus Environment

ACS Catalysis

Page 4 of 25

4 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

Experimental Section Preparation and pre-treatment of Au/TiO2 A commercially available Au/TiO2 catalyst with a nominal Au loading of 1 wt.% Au (AUROliteTM, prepared via deposition precipitation) was used in this study. Prior to all measurements the as received, already calcined Au/TiO2 catalyst was first i) dried in-situ in a flow of 20 Nml·min-1 N2 at 100°C for 16 h, followed by ii) a further pre-treatment at 400°C either in 10% O2/N2 (O400, 20 Nml·min-1) or 10% CO/N2 (CO400, 20 Nml·min-1) for 30 min before cooling to the reaction temperature of 80°C. Heating and cooling took place in a flow of pure N2. In-line water filters (Agilent, CP17971) were used during this pre-treatment as well as during the reaction (described below) in order to reduce the amount of residual water in all gases and, hence, to realize dry reaction conditions (< 0.2 ppm H2O). Au particle size The mean Au particle size and the Au particle size distribution in the previously calcined samples before and after reaction were determined from Transmission Electron Microscopy (TEM, JEOL 1400) images. For detailed information on the Au particle size (dVA: volumearea mean diameter, DAu: Au dispersion), at least 600 Au particles were evaluated for each sample. Assuming half-spherical Au NPs and 1.15·1015 surface Au atoms·cm-2, dVA and DAu values were calculated from the experimentally determined diameter (di) for a number of Au nanoparticles (ni) as follows:

 =

∑  

∑  

 = 6 ×

 ⁄ 



(equation 1)  ×∑ 

= 6 ×  ×∑  





 

Here vm equals the volume of a Au atom, and am the surface area of a Au atom.

ACS Paragon Plus Environment

(equation 2)

Page 5 of 25

ACS Catalysis

5 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

Kinetic measurements Catalytic activities of the Au/TiO2 catalyst for the continuous CO oxidation after different pre-treatments were determined in a conventional quartz glass micro-reactor at atmospheric pressure. All measurements were performed at a reaction gas flow of 60 Nml·min-1 containing 1% CO, 1% O2, balance N2, and at a reaction temperature of 80°C. Influent and effluent gases were analyzed by on-line gas chromatography (DANI, GC 86.10HT). Based on the relative conversion of CO to CO2 (XCO) under isothermal and differential reaction conditions Au mass normalized reaction rates (rAu) and turnover frequencies (TOFs) were calculated according to the following equations:  =

!  =

 ,  ,  , " ×  , #$% × &

'()  =

=

=

  ,  ,

  , #

* ×+ ,

(equation 3)

(equation 4) (equation 5)

Here nCO,in and nCO2,out are the molar flow rates of CO entering and CO2 leaving the reactor, mcat and mAu are the masses of catalyst /Au used in the experiment, wAu is the actual gold loading of 1.0 wt.-%, and MAu is the molar mass of Au. Since these calculations are only correct under differential reaction conditions, with conversions below ca. 20%, only 1 mg of the Au/TiO2 catalyst diluted 1:80 with inert Al2O3 were used (ca. 80 mg in total). In-situ infrared spectroscopy (DRIFTS) In-situ IR investigations were performed in a DRIFTS (Diffuse Reflectance Infrared Fourier Transform Spectroscopy) configuration with a Nicolet 6700 spectrometer from Thermo Fisher Scientific, equipped with a MCT narrow band detector and a commercial in-situ reaction cell unit from Harricks (HV-DR2). The pre-treatment as well as the reaction conditions during these measurements were identical to those in kinetic measurements in order to enable a direct

ACS Paragon Plus Environment

ACS Catalysis

6 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

comparison between the build-up of surface species during reaction (DRIFTS) and the catalytic activity (kinetic measurements). Prior to the experiments, background spectra were recorded on the freshly calcined catalysts at the reaction temperature of 80°C in a flow of pure N2 (60 Nml·min-1). From the reflectivity measured in pure N2 (background spectra) and that measured during the CO oxidation reaction the absorption of adsorbed surface species was calculated in terms of Kubelka-Munk units.26;27 These are assumed to be linearly related to the adsorbate coverage. Note that in all spectra presented contributions of the gas phase CO signal were already removed by subtracting a spectrum recorded on pure α-Al2O3 in COcontaining atmosphere at the same temperature from the spectral region of gas phase CO (2040-2240 cm-1). This way only adsorbed species are visible in the spectra illustrated.

Results and Discussion Catalyst characterization The Au loading of the catalyst was determined by optical emission spectroscopy. In accordance with the nominal loading it is 1.0 wt.%. The mean Au particle size and the Au particle size distribution after both pre-treatments and subsequent CO oxidation at 80 °C for 1000 min were determined by transmission electron microscopy. As shown in Fig. S1 it is about 2.5 nm after both pre-treatments and after subsequent reactions. Hence, differences in the activity after those very different pre-treatments originate only to a minor extent from Au particle size effects, which are well known to strongly affect the catalytic performance of supported Au catalysts.28-30 Note that a smaller Au particle size after reductive pre-treatment, which was reported in earlier studies,31;32 is not observed in our case, most probably because in the commercial STREM Au/TiO2 catalyst the Au NPs were already formed during a previous ex-situ calcination by the manufacturer.33

ACS Paragon Plus Environment

Page 6 of 25

Page 7 of 25

ACS Catalysis

7

Influence of TiO2 bulk defects on the CO oxidation activity To study the influence of TiO2 bulk defects on the catalytic behavior of Au/TiO2 for the CO oxidation at 80°C we first measured and compared the reaction rates after in-situ pretreatment of the Au/TiO2 catalysts in strongly oxidizing (O400: 10% O2/N2) and in strongly reducing (CO400: 10% CO/N2) atmospheres. The Au mass normalized reaction rates obtained under differential reaction conditions are shown in Figure 1. 8

a) 80°C O400

6

4 -1 Reaction rate / —10-3—mol—gAu —s-1

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

2

CO400

0

b) 120°C 15

O400 10

CO400

5

0 0

400

800

1200

Time / min

Figure 1:

Au mass normalized reaction rates recorded during CO oxidation (1% CO, 1% O2, balance N2) on Au/TiO2 after in-situ pre-treatment at 400°C in oxidative (O400: 10% O2/N2) or reductive (CO400: 10% CO/N2) atmosphere at reaction temperatures of (a) 80°C and (b) 120°C.

After calcination by O400 the activity is highest right in the beginning of the reaction, with an initial activity of 6.8·10-3 mol·gAu-1·s-1 (TOF: 3.7 s-1), followed by a rather strong deactivation by about 22% within 200 min on stream (78% of the initial activity). In the further course of ACS Paragon Plus Environment

ACS Catalysis

8 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 reaction it continues to slightly decrease, until reaching a steady-state situation after about 1200 min (5.1·10-3 mol·gAu-1·s-1, TOF: 2.8 s-1). This typical deactivation behavior has been convincingly explained to originate from the formation and accumulation of reaction inhibiting surface species, mainly surface carbonates, during the CO oxidation.18;33;34 Compared to previous studies on home-made Au/TiO2 catalysts,34 however, the surface carbonate formation is less pronounced on the commercial Au/TiO2 catalyst used in this study. Accordingly, we observed slightly higher reactions rates and a less pronounced deactivation. After pre-treatment in CO/N2 at 400°C (CO400), in contrast, the reaction behavior is completely different. Here the catalyst is least active in the initial period of the reaction, and its Au mass normalized reaction rate after 10 min is by a factor of about 10 lower compared to the initial activity after O400 (0.6·10-3 mol·gAu-1·s-1, TOF: 0.3 s-1, see Fig. 1). In addition to the initial rates, also the temporal evolution of the reaction rate is completely different after CO400. Instead of a deactivation, the reaction rate continues to increase with time on stream. After 1200 min it reaches 1.8·10-3 mol·gAu-1·s-1 (TOF: 1.0 s-1), which is still a factor of 3 lower compared to that after O400 at this point. But even after that time a steady-state situation is not yet reached, and the activity still increases with time on stream. Additional kinetic measurements revealed that a steady-state situation is reached only after about 6000 min. Interestingly, the reaction rates in steady-state are almost identical for both samples, indicating that any modifications of Au/TiO2 catalysts during reduction at elevated temperatures (CO400) are fully reversible upon interaction with the (net oxidizing) reaction gas mixture for extended reaction times (see Fig. S2). From recent studies in our group it is evident that after O400 pre-treatment no adsorbed molecular oxygen is present on the catalyst surface.19;20 The only oxygen species which are present on Au/TiO2 directly after O400 are TiO2 surface lattice oxygen and to some extent ACS Paragon Plus Environment

Page 8 of 25

Page 9 of 25

ACS Catalysis

9 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

also atomic oxygen species on the surface of the Au NPs. The latter ones, however, are immediately removed during the CO oxidation (by reaction with CO), and no more replenished under present reaction conditions. Considering the strongly reducing reaction atmosphere (10% CO at 400°C) during CO400 pre-treatment and the identical Au particle size after both pre-treatments, we propose that the different reaction characteristics are due to the presence or absence of oxygen vacancies in TiO2. In general, these vacancies may be present in the surface and/or in the bulk of TiO2. During reaction, however, surface oxygen vacancies are known to be quickly replenished by reaction with gas phase O2,19 and the surface of TiO2 in Au/TiO2 catalysts is close to fully oxidized under net oxidizing reaction conditions.21 Therefore we can rule out TiO2 surface oxygen vacancies as physical origin for the significantly lower CO oxidation activity after CO400 pre-treatment. Instead, we relate the lower CO oxidation activity and its increase with time on stream to the presence of oxygen vacancies in the bulk of the TiO2 support and their slow replenishment with time on stream, e.g. via oxygen bulk diffusion during reaction. The formation of surface and bulk oxygen vacancies in TiO2 during the pre-treatment in CO at elevated temperatures (CO400) fits also to previous results of temperature-programmed reduction measurements on Au/TiO2 by Idakiev et al., who observed a weak surface reduction peak (at around 100°C) and a strong bulk reduction peak (at around 400°C).35 A similar catalytic behavior after pre-treatment in reducing atmosphere has also been reported for a Au/Fe2O3 catalyst during the CO oxidation at room temperature by Gupta and coworkers.36 When pre-treated under H2 atmosphere, the CO oxidation activity as well as the CO adsorption capacity of the catalyst were strongly suppressed at the beginning of the reaction, due to a partial reduction of Fe2O3 to Fe3O4, and restored during time on stream, i.e., during a prolonged exposure to CO and O2 during which Fe3O4 is re-oxidized.36 These authors also demonstrated that the difference in activity after reductive pre-treatment

ACS Paragon Plus Environment

ACS Catalysis

10 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

compared to that after oxidative pre-treatment depends on the temperature during pretreatment in H2, with less active catalysts obtained after reduction at higher temperature. They explained their results by a higher degree of reduction with increasing temperature, but without a clear assignment whether this relates to bulk or surface defects.36 Compared to our results, however, the re-activation with time on stream was much faster in their study, despite the lower reaction temperature in their case, taking only about 100 minutes until the highest activity was reached. Here it should be noted that reduction by hydrogen also results in the formation of surface hydroxyl groups and/or adsorbed water on the catalyst surface, which are well known to significantly affect the CO oxidation activity of supported Au catalysts as well.37-39 Accordingly, we propose that a reductive pre-treatment by CO (at elevated temperatures) is even more suitable to detect the influence of surface and bulk vacancies on the catalytic performance as compared to a reductive pre-treatment by H2. As it is demonstrated by IR spectra recorded at 80°C in pure N2 directly after the pre-treatment by CO400, there is no more CO adsorbed on the catalyst surface after this pretreatment (see also Fig. S3). To obtain more insight into the processes limiting the activity after CO400 pre-treatment we performed similar experiments at a slightly higher reaction temperature of 120°C (Fig. 1b) and at higher O2 content (Fig. 2). For reaction at 120°C, a similar activity of both samples, O400 and CO400 sample, is already reached after about 300 min. Obviously, the re-activation process after CO400 is accelerated at the higher reaction temperature, which could be related to a higher mobility of bulk defects (oxygen vacancies) in TiO2. Previous studies on TiO2 indeed provided clear evidence that diffusion of bulk defects in TiO2 is thermally activated at temperatures above ~125°C,40;41 confirming our conclusion that the lower activity after CO400 originates from TiO2 bulk defects.

ACS Paragon Plus Environment

Page 10 of 25

Page 11 of 25

ACS Catalysis

11

10

a) 20% O2

Reaction rate -1 / —10-3—mol—gAu —s-1

8

6

4

1% O2

2

0

b) 300

Relative activity / %

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

20% O2

1% O2

200

100

0

400

800

1200

Time / min

Figure 2:

(a) Au mass normalized reaction rates and (b) relative activities during CO oxidation at 80°C on a Au/TiO2 catalyst after reductive pre-treatment at 400°C (CO400, 10% CO/N2) with different O2 partial pressures (red/white circles: 1% CO, 1% O2, balance N2; blue/white triangles: 1% CO, 20% O2, balance N2).

Similar measurements performed on the CO400 pre-treated catalyst in O2-rich atmosphere, with 20% O2 in the reaction gas feed (Fig. 2), revealed much higher rates, as expected from the positive reaction order of O2 in the CO oxidation reported earlier.42 In the present case the rates differ by about a factor of 4.5 on the CO400 pretreated sample, while for a O400 pretreated catalyst, this was typically a factor of slightly more than 2.42 This means that the presence of bulk defects in the TiO2 support affects also the influence of the oxygen partial pressure. The time required for the activation process is, however, comparable to that determined for reaction with 1% O2 in the reaction gas. Obviously, the re-activation process, which is observed by the increase in the relative rate, is hardly affected by the O2 partial

ACS Paragon Plus Environment

ACS Catalysis

12 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

pressure. This confirms our assignment that the lower activity of the CO400 sample and the re-activation with time on stream indeed originates from bulk defects and their slow replenishment during reaction, while the re-oxidation of surface oxygen vacancies should be accelerated for an increasing O2 content in the gas phase. Influence of TiO2 bulk defects on the CO adsorption (in-situ IR spectroscopy) For gaining more insight into how the bulk oxygen vacancies may affect the CO oxidation activity of the reduced Au/TiO2 catalyst after CO400 pre-treatment, we followed the accumulation of adsorbed species during the reaction under identical reaction conditions by in-situ IR spectroscopy (DRIFTS) measurements. These measurements enable us to identify possible effects of the bulk defects on the adsorption properties of the Au/TiO2 catalyst. Series of spectra, which are shown in the Supporting Information (Fig. S4), reveal that the most significant differences after both pre-treatments, CO400 and O400, appear in the CO adsorption properties. Therefore, we here present sequences of representative spectra covering the range between 2250 cm-1 and 1950 cm-1 (see Fig. 3). Note that contributions from gas phase CO are already subtracted in these spectra. On the O400 pre-treated catalyst the exposure to the reaction gas mixture (1% CO, 1% O2, balance N2) at 80°C results in the formation of two absorption bands for adsorbed CO, a more pronounced band at 2119 cm-1 that can be assigned to COad on Au NPs,43;44 and a very weak band at 2181 cm-1 typical for CO adsorbed on coordinatively unsaturated Ti4+ sites on the support (see Fig. 3a).45-47 The intensity of the band at 2119 cm-1 is essentially constant for over 1000 min on stream (see also the temporal evolution of the band intensity in Fig. 3b), indicating a constant COad coverage during the reaction. For the COad coverage on the support material, the intensity is too low to quantitatively evaluate subtle changes in the intensity with time on stream (see Supporting Information, Fig. S5). Again, the situation is very different for the Au/TiO2 catalyst pre-treated in reducing atmosphere (CO400). While the adsorption of ACS Paragon Plus Environment

Page 12 of 25

Page 13 of 25

ACS Catalysis

13 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

CO on the support surface, indicated by the weak absorption band at 2181 cm-1, is comparable to that on the O400 catalyst, the COad coverage on the Au NPs is very low at the beginning of the reaction. As we have already discussed above, background spectra recorded in N2 directly after pre-treatment and cooling down the catalyst to reaction temperature (80°C) revealed that there is essentially no adsorbed CO present when starting the reaction (Fig. S3). With time on stream, it continuously increases, together with a small blue-shift of the absorption band from 2116 cm-1 to 2119 cm-1. Note that the slight red-shift compared to the catalyst sample after O400 in the beginning of the reaction (2116 cm-1 and 2119 cm-1 for CO400 and O400, respectively) may result from CO adsorption on less positively charged Au species. This was also observed in previous IR studies during the CO adsorption on Au/TiO2 by Boccuzzi et al. and Sterrer et al., although the red-shift was much more pronounced in their case.48;49 The less pronounced shift of the COad absorption band in our study originates most likely from the fact that during the initial period of the CO oxidation TiO2 surface oxygen vacancies close to the Au NPs are immediately fully replenished, in contrast to the studies of Boccuzzi et al. and Sterrer et al., which were performed in the absence of O2 in the gas phase (CO adsorption). Even after 1000 min the intensity is much lower than on the O400 catalyst, but still increases with time on stream. Here we note that the observed blue-shift is not related to the increasing COad coverage, since for CO adsorption on Au nanoparticles, and also on extended Au surfaces this usually goes along with a red-shift of the absorption band.50-53

ACS Paragon Plus Environment

ACS Catalysis

Page 14 of 25

14

a)

1 min 1080 min CO titration

Au-CO 2119 cm-1

-3

5⋅10 KMU 4+

Ti -CO 2181 cm

-1

O400 2116 cm

-1

CO400 2200

2100

2000

Wavenumber / cm-1 10

b) Au-COad intensity / a.u.

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

8

O400 6

4

2

CO400

0 0

400

800

1200

Time / min

Figure 3:

Sequence of in-situ DRIFT spectra recorded during CO oxidation (1% CO, 1% O2, balance N2) on Au/TiO2 at 80°C directly after in-situ pre-treatment at 400°C in oxidative (O400: 10% O2/N2) or reductive (CO400: 10% CO/N2) atmosphere. For comparison we also show a spectrum recorded during CO adsorption (1% CO, balance N2, 10 min) on a CO400 catalyst. For all spectra the gas phase CO signal is already subtracted.

We instead propose that this blue-shift is due to changes in the electronic properties of the Au NPs, which go along with significant changes in the CO adsorption strength. After 1000 min of reaction, however, the position of the absorption band for both COad species is identical after both pre-treatments. Overall, there is a good correlation in the temporal evolution between the COad coverage on the Au NPs and the activity for CO oxidation after CO400 (see Figs. 1a and 3b). Obviously, the presence of TiO2 bulk defects has a strong influence also on

ACS Paragon Plus Environment

Page 15 of 25

ACS Catalysis

15 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 amount of adsorbed CO on the Au/TiO2 catalyst, resulting in significantly lower COad coverages on the Au NPs in the presence of such defects. With respect to similar effects on the adsorption of molecular oxygen (O2,ad) it should be noted that based on TPD measurements in the TAP reactor, the amount of adsorbed molecular oxygen is below the detection limit (in the range of 1% of a monolayer). Nevertheless, under reaction conditions there will still be minute amounts of adsorbed O2 on the surface, which we think act as precursor in the replenishment of surface lattice oxygen vacancies. Although we cannot rule out that that the presence of oxygen bulk vacancies may affect also the lifetime and hence concentration of O2,ad species, this would be an indirect effect, in contrast to COad, which is a direct reaction partner (with surface lattice oxygen). Accordingly, similar effects of the presence of bulk defects on the adsorption properties may occur also for adsorbed O2,ad, but i) their steady-state coverage is too low to be experimentally accessible under present reaction conditions and ii) any variation in the O2,ad coverage would affect the reaction only indirectly. A similar relation between the presence of TiO2 bulk defects and the amount of adsorbed CO was also observed in additional CO adsorption experiments, where a freshly CO400 pretreated catalyst was exposed to CO only (1% CO/N2) at 80°C. The corresponding spectrum recorded after CO adsorption for 10 min is also presented in Fig. 3 (blue line). In this case, in the absence of O2 in the gas phase, CO adsorption on the surface of the Au NPs is completely suppressed, and even after 1000 min exposure of the catalyst to CO/N2 there is still no CO adsorption detected. Obviously, the interaction of the catalyst with oxygen is a precondition for an increasing CO adsorption on the Au NPs on the CO400 pre-treated catalyst with time on stream, in good agreement with our above proposal that the reactivation of the CO400 pretreated Au/TiO2 catalyst is related to the re-oxidation of TiO2.

ACS Paragon Plus Environment

ACS Catalysis

16 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 lower amount of CO adsorption can be caused by different effects. Most simply, it may arise from a lower CO adsorption strength on the Au NPs, caused e.g. by an electronic modification of the Au NPs. Considering the often observed SMSI effects for PGM/TiO2 (model) catalysts after exposure to reductive atmospheres at elevated temperatures,2;4-6 as described in the introduction, one may alternatively argue that the lower amount of COad on the CO400 catalyst originates from an encapsulation of the Au NPs by a layer of reduced TiOx species and, accordingly, a lower accessible Au surface area. Additional CO adsorption experiments at lower temperature (1% CO/N2, -20°C) revealed, however, that under these conditions CO can adsorb on the Au NPs also directly after the CO400 treatment, and that the amount of COad is of comparable order of magnitude to that obtained after the oxidizing O400 pre-treatment. Moreover, high-resolution TEM images recorded after CO400 pre-treatment (Fig. S1a) do not show any evidence for an encapsulation of the Au NPs by reduced TiOx. These additional results clearly support that the lower COad coverage observed on Au/TiO2 during reaction at 80°C after CO400 originates from a lower CO adsorption strength on the Au NPs due to EMSI effects in the presence of bulk defects rather than from a lower accessible Au surface area. From the fact that the time scale as well as the (relative) changes in both, activity and COad coverage are very similar, we propose that the lower activity of the CO400 pre-treated catalysts mainly originates from the lower CO adsorption strength and, hence, the lower amount of adsorbed COad on these catalysts during reaction. Finally it should be noted that the formation of carbonate species, which is widely accepted as the origin of deactivation of Au/TiO2 catalyst during CO oxidation reaction,13;33;34;54;55 was also observed in the present DRITFS measurements. In fact, the type of carbonate species and their formation rates are rather similar, independent of the pre-treatment procedure (see Fig. S4). Accordingly we suggest that the change of the catalytic activity of the CO400 sample ACS Paragon Plus Environment

Page 16 of 25

Page 17 of 25

ACS Catalysis

17 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

originates from two effects: replenishment of bulk-oxygen-vacancies (increasing activity) and build-up of surface carbonate species (decreasing activity), while for the O400 sample only the latter process is active. Besides the Au/Fe2O3 example discussed above,36 electronic metal support interactions (EMSI) have also been claimed recently for Au/Ce0.62Zr0.38O2 catalysts. This was based on the lower amount of CO adsorption and a slight shift of the Au(4f) signal by 0.2 eV after reductive pre-treatment compared to after oxidative pre-treatment.56 Similar to the present study there was no evidence for encapsulation of the Au NPs. On the other hand, Liu et al. concluded an encapsulation of the Au NPs on a Au/ZnO catalyst, driven by strong metal support interactions, after oxidative catalyst pre-treatment at 300°C, based mainly on high resolution TEM measurements and on a lower amount of CO adsorption.57 Interestingly, these modifications were observed after oxidative rather than after reductive pre-treatment, which is generally considered as prerequisite for the classical SMSI effect (encapsulation), while in all other cases, including the present study on Au/TiO2 catalysts, modifications were observed after reductive pre-treatment. Finally we would like to note that similar findings of a lower CO oxidation activity after exposure to CO were reported also for a Au/CeO2 catalyst (reaction temperature 120°C).58 In that case, however, the reduction of the CeO2 support was limited to the surface, resulting in the formation of surface oxygen vacancies close to the Au NPs. Accordingly, the activity was restored on a rather short time scale, in the range of a few seconds, during reaction in oxidative atmosphere. Considering other reactions, similar type effects were also reported for ammonia synthesis over a Ru catalyst supported on Zr-modified CeO2.59 In consequence, we believe that such kind of electronic modifications caused by interactions of the metal NPs with bulk oxygen vacancies (electronic metal-support interactions (EMSI)) are not limited to Au/TiO2 catalysts, but are a much more general phenomenon for noble ACS Paragon Plus Environment

ACS Catalysis

18 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

metal NP catalysts including Au catalysts supported on reducible oxide supports. These electronic modifications may be accompanied by structural modifications, e.g., by encapsulation of the metal NPs, and the latter effect may even be dominant for the catalytic activity, but this is not necessarily the case.8 In that sense, the often used equalization of SMSI effects and structural modifications due to encapsulation is often misleading and should be replaced by more precise considerations. A last point to be discussed is the mechanism governing this interaction. According to Di Valentina et al., in TiO2 (anatase) electrons from oxygen bulk vacancies transfer to neighboring, undercoordinated Ti4+ ions, leading to Ti3+ ions, with the additional electron in a highlying state below the bottom of the conduction band.60 These electrons may either tunnel or transfer via thermal excitation to a Au nanoparticle at the surface, leading to a negatively charged Au particle. Similar transport concepts were proposed to apply also for rutile, which is assumed to present qualitatively similar electron-transport properties.60 Obviously, depending on the reaction or on the nature of the catalyst, such kind of charging may result in different modifications of the catalyst activity, either lowering or increasing it, and this may occur not only for Langmuir-Hinshelwood type reactions but also for the Au-assisted Marsvan Krevelen mechanism proposed for CO oxidation on Au/TiO2 catalysts and comparable systems. At present we can neither quantify the maximum distance of the oxygen vacancies from the Au nanoparticles for such kind of charge transfer nor the amount of charge accumulated on the Au nanoparticles. Hence, more work is required to identify further details of the interactions between bulk oxygen vacancies and electronic nature of the Au NPs.

ACS Paragon Plus Environment

Page 18 of 25

Page 19 of 25

ACS Catalysis

19 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

Summary / Conclusions In summary we have demonstrated by combined kinetic and in-situ IR spectroscopic measurements that electronic metal-support interactions (EMSI) may considerably affect the CO oxidation activity of Au/TiO2 catalysts and, most likely, of Au catalysts supported on reducible oxides in general. Electronic modifications of the support, by the formation of bulk oxygen vacancies in the TiO2 support upon pre-treating a commercial Au/TiO2 catalyst in a strongly reducing atmosphere (10% CO/N2, 400°C), were found to result in much less active catalysts than obtained in the absence of such bulk defects. During the reaction, however, activity increased continuously until reaching that of catalysts pre-treated in an oxidative atmosphere. We attribute these variations in the CO oxidation activity to electronic metalsupport interactions (EMSI), supported also by the observations that reaction at higher temperatures (faster replenishment of bulk oxygen vacancies due to more facile vacancy diffusion) results in a faster recovery of the activity, while using a higher O2 content in the reaction gas has little effect on this. A pronounced modification of the electronic properties of the Au NPs is indicated by the simultaneous modifications of the CO adsorption strength, more specifically a lower CO adsorption strength with increasing number of oxygen bulk vacancies. CO oxidation activity and CO adsorption strength are strictly correlated with each other, and the latter is proposed as a main reason for the modified activity. Considering also observations in previous reports, we propose that such kind of EMSI effects are of general validity for Au catalysts supported on reducible oxides.

ACS Paragon Plus Environment

ACS Catalysis

20 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

AUTHOR INFORMATION Corresponding Author * Author to whom correspondence should be addressed, e-mail: juergen.behm@uni-ulm Author Contributions The authors declare no competing financial interest. Supporting Information Additional figures including TEM images, kinetic data and infrared data (DRIFTS) in support of the manuscript. This material is available free of charge via the Internet at http://pubs.acs.org. ACKNOWLEDGMENT We gratefully acknowledge Dr. J. Biskupek (Ulm University) for excellent high resolution TEM images and Prof. G. Pacchioni (University of Milano) for stimulating discussions.

ACS Paragon Plus Environment

Page 20 of 25

Page 21 of 25

ACS Catalysis

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

References (1)

Tauster, S. J.; Fung, S. C.; Garten, R. L. J. Am. Chem. Soc. 1978, 100, 170-175.

(2)

Haller, G. L.; Resaco, D. E. Adv. Catal. 1989, 36, 173-234.

(3)

Bernal, S.; Calvino, J. J.; Cauqui, M. A.; Gatica, J. M.; Larese, C.; Pérez Omil, J. A.; Pintado, J. M. Catal. Today 1999, 50, 175-206.

(4)

Tauster, S. J. Acc. Chem. Res. 1987, 20, 389-394.

(5)

Diebold, U.; Li, M.; Dulub, O.; Hebenstreit, E. L. D.; Hebenstreit, W. Surf. Rev. Lett. 2000, 7, 613-617.

(6)

Bowker, M.; Stone, P.; Morrall, P.; Smith, R.; Bennett, R.; Perkins, N.; Kvon, R.; Pang, C.; Fourre, E.; Hall, M. J. Catal. 2005, 234, 172-181.

(7)

Bruix, A.; Rodriguez, J. A.; Ramírez, P. J.; Senanayake, S. D.; Evans, J.; Park, J. B.; Stacchiola, D.; Liu, P.; Hrbek, J.; Illas, F. J. Am. Chem. Soc. 2012, 134, 8968-8974.

(8)

Campbell, C. T. Nat. Chem. 2012, 4, 597-598.

(9)

Haruta, M.; Kobayashi, T.; Sano, H.; Yamada, N. Chem. Lett. 1987, 16, 405-408.

(10) Hutchings, G. J. Catal. Today 2005, 100, 55-61. (11) Bond, G. C.; Louis, C.; Thompson, D. T. In Catalysis by Gold; Imperial Press: London, 2007; Vol. 6;.p 1-366. (12) Haruta, M.; Tsubota, S.; Kobayashi, T.; Kageyama, H.; Genet, M. J.; Delmon, B. J. Catal. 1993, 144, 175-192. (13) Bollinger, M. A.; Vannice, M. A. Appl. Catal. B 1996, 8, 417-443. (14) Liu, X. Y.; Wang, A.; Zhang, T.; Mou, C. Y. Nano Today 2013, 8, 403-416. (15) Panayotov, D. A.; Morris, J. R. Surf. Sci. Rep. 2016, 71, 77-271. (16) Schubert, M. M.; Hackenberg, S.; van Veen, A. C.; Muhler, M.; Plzak, V.; Behm, R. J. J. Catal. 2001, 197, 113-122. (17) Arrii, S.; Morfin, F.; Renouprez, A. J.; Rousset, J. L. J. Am. Chem. Soc. 2004, 126, 1199-1205. (18) Widmann, D.; Liu, Y.; Schüth, F.; Behm, R. J. J. Catal. 2010, 276, 292-305. (19) Widmann, D.; Behm, R. J. Angew. Chem. Int. Ed. 2011, 50, 10241-10245. ACS Paragon Plus Environment

ACS Catalysis

22 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

(20) Widmann, D.; Behm, R. J. Acc. Chem. Res. 2014, 47, 740-749. (21) Maeda, Y.; Iizuka, Y.; Kohyama, M. J. Am. Chem. Soc. 2013, 135, 906-909. (22) Duan, Z.; Henkelman, G. ACS Catal. 2015, 5, 1589-1595. (23) Saqlain, M. A.; Hussain, A.; Siddiq, M.; Ferreira, A. R.; Leitao, A. A. Phys. Chem. Chem. Phys. 2015, 17, 25403-25410. (24) Remediakis, I. N.; Lopez, N.; Nørskov, J. K. Appl. Catal. A 2005, 291, 13-20. (25) Lopez, N.; Janssens, T. V. W.; Clausen, B. S.; Xu, Y.; Mavrikakis, M.; Bligaard, T.; Nørskov, J. K. J. Catal. 2004, 223, 232-235. (26) Hamadeh, I. M.; Griffiths, P. R. Appl. Spectrosc. 1987, 41, 682-688. (27) Armaroli, T.; Bécue, T.; Gautier, S. Oil & Gas Science and Technology - Rev. IFP 2004, 59, 215-237. (28) Haruta, M. Catal. Today 1997, 36, 153-166. (29) Kotobuki, M.; Leppelt, R.; Hansgen, D. A.; Widmann, D.; Behm, R. J. J. Catal. 2009, 264, 67-76. (30) Laoufi, I.; Saint-Lager, M. C.; Lazzari, R.; Jupille, J.; Robach, O.; Garaudée, S.; Cabailh, G.; Dolle, P.; Cruguel, H.; Bailly, A. J. Phys. Chem. C 2011, 115, 4673-4679. (31) Tsubota, S.; Cunningham, D. A. H.; Bando, Y.; Haruta, M. Stud. Surf. Sci. Catal. 1995, 91, 227-235. (32) Schumacher, B.; Plzak, V.; Cai, J.; Behm, R. J. Catal. Lett. 2005, 101, 215-224. (33) Saavedra, J.; Powell, C.; Panthi, B.; Pursell, C. J.; Chandler, B. D. J. Catal. 2013, 307, 37-47. (34) Denkwitz, Y.; Schumacher, B.; Kucerova, G.; Behm, R. J. J. Catal. 2009, 267, 78-88. (35) Idakiev, V.; Tabakova, T.; Yuan, Z.-Y.; Su, B.-L. Appl. Catal. A 2004, 270, 135-141. (36) Gupta, N. M.; Tripathi, A. K. J. Catal. 1999, 187, 343-347. (37) Daté, M.; Haruta, M. J. Catal. 2001, 201, 221-224. (38) Ide, M. S.; Davis, R. J. Acc. Chem. Res. 2014, 47, 825-833. (39) Saavedra, J.; Doan, H. A.; Pursell, C. J.; Grabow, L. C.; Chandler, B. D. Science 2014, 345, 1599-1602. ACS Paragon Plus Environment

Page 22 of 25

Page 23 of 25

ACS Catalysis

23 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

(40) Henderson, M. A. Surf. Sci. 1995, 343, L1156-L1160. (41) Henderson, M. A. Surf. Sci. 1999, 419, 174-187. (42) Schumacher, B.; Denkwitz, Y.; Plzak, V.; Kinne, M.; Behm, R. J. J. Catal. 2004, 224, 449-462. (43) Boccuzzi, F.; Chiorino, A. J. Phys. Chem. B 2000, 104, 5414-5416. (44) Mihaylov, M.; Knözinger, H.; Hadjiivanov, K.; Gates, B. C. Chem. Ing. Tech. 2007, 79, 795-806. (45) Martra, G. Appl. Catal. A 2000, 200, 275-285. (46) Boccuzzi, F.; Chiorino, A.; Manzoli, M. Surf. Sci. 2000, 454-456, 942-946. (47) Green, I. X.; Tang, W.; Neurock, M.; Yates, J. T. Science 2011, 333, 736-739. (48) Boccuzzi, F.; Chiorino, A.; Manzoli, M.; Andreeva, D.; Tabakova, T. J. Catal. 1999, 188, 176-185. (49) Sterrer, M.; Yulikov, M.; Fischbach, E.; Heyde, M.; Rust, H. P.; Pacchioni, G.; Risse, T.; Freund, H. J. Angew. Chem. Int. Ed. 2006, 45, 2630-2632. (50) Ruggiero, C.; Hollins, P. J. Chem. Soc. Faraday Trans. 1 1996, 92, 4829-4834. (51) Boccuzzi, F.; Chiorino, A.; Tsubota, S.; Haruta, M. J. Phys. Chem. 1996, 100, 36253631. (52) Schumacher, B.; Denkwitz, Y.; Plzak, V.; Kinne, M.; Behm, R. J. J. Catal. 2004, 224, (53) Lemire, C.; Meyer, R.; Shaikhutdinov, S.; Freund, H. J. Surf. Sci. 2004, 552, 27-34. (54) Konova, P.; Naydenov, A.; Venkov, C.; Mehandjiev, D.; Andreeva, D.; Tabakova, T. J. Mol. Catal. A Chem. 2004, 213, 235-240. (55) Raphulu, M.; McPherson, J.; Lingen, E.; Anderson, J. A.; Scurrell, M. S. Gold Bull 2010, 43, 334-344. (56) Cíes, J. M.; del Río, E.; López-Haro, M.; Delgado, J. J.; Blanco, G.; Collins, S.; Calvino, J. J.; Bernal, S. Angew. Chem. Int. Ed. 2010, 49, 9744-9748. (57) Liu, X.; Liu, M. H.; Luo, Y. C.; Mou, C. Y.; Lin, S. D.; Cheng, H.; Chen, J. M.; Lee, J. F.; Lin, T. S. J. Am. Chem. Soc. 2012, 134, 10251-10258. (58) Widmann, D.; Leppelt, R.; Behm, R. J. J. Catal. 2007, 251, 437-442.

ACS Paragon Plus Environment

ACS Catalysis

24 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

(59) Ma, Z.; Xiong, X.; Song, C.; Hu, B.; Zhang, W. RSC Adv. 2016, 6, 51106-51110. (60) Di Valentin, C.; Pacchioni, G.; Selloni, A. J. Phys. Chem. C 2009, 113, 20543-20552.

ACS Paragon Plus Environment

Page 24 of 25

Page 25 of 25

ACS Catalysis

25 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

Table of Contents

ACS Paragon Plus Environment