Al2O3 Catalysts: Evidence

Feb 25, 2018 - The mechanism of CO oxidation over supported gold catalysts has long been debated, with two prevailing mechanisms dominating the discus...
1 downloads 11 Views 2MB Size
Subscriber access provided by MT ROYAL COLLEGE

Article

CO Oxidation Kinetics over Au/TiO2 and Au/Al2O3 Catalysts: Evidence for a Common Water-Assisted Mechanism Johnny Saavedra, Christopher J Pursell, and Bert D Chandler J. Am. Chem. Soc., Just Accepted Manuscript • DOI: 10.1021/jacs.7b12758 • Publication Date (Web): 25 Feb 2018 Downloaded from http://pubs.acs.org on February 25, 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.

Journal of the American Chemical Society 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 36 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

Journal of the American Chemical Society

CO Oxidation Kinetics over Au/TiO2 and Au/Al2O3 Catalysts: Evidence for a Common Water-Assisted Mechanism Johnny Saavedra, Christopher J. Pursell, and Bert D. Chandler* Department of Chemistry, Trinity University, San Antonio, TX 78212-7200 *

To whom correspondence should be addressed:

[email protected] (210) 999-7557 phone; (210) 999-7569 fax

1 ACS Paragon Plus Environment

Journal of the American Chemical Society 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 36

Abstract The mechanism of CO oxidation over supported gold catalysts has long been debated, with two prevailing mechanisms dominating the discussion: a water-assisted mechanism and a mechanism involving O-defect sites. In this study, we directly address this debate through a kinetic and mechanistic investigation of the role of water in CO oxidation over Au/TiO2 and Au/Al2O3 catalysts; the results clearly indicate a common water-assisted mechanism to be at work. Water adsorption isotherms were determined with infrared spectroscopy; the extracted equilibrium constant was essentially the same.

Added water decreases CO adsorption on

Au/TiO2, likely by blocking CO binding sites at the metal-support interface. Reaction kinetics (CO, O2, and H2O reaction orders) were essentially the same for both catalysts, as were measured O-H(D) kinetic isotope effects. These data indicate that the two catalysts operate by essentially the same mechanism under the conditions of these experiments (ambient temperature, significant amounts of water available).

A reaction mechanism incorporating the kinetic and

thermodynamic data, and accounting for different CO and O2 / COOH binding sites is proposed. The mechanism and kinetic data are treated with an active site (Michaelis-Menten) approach. This indicated water adsorption does not significantly affect reaction rate constants, only the number of active sites available at a given water pressure. Extracted water and O2 binding constants are similar on both catalysts and consistent with previous DFT calculations. Water adsorption constants are also similar to independently determined equilibrium constants measured by IR spectroscopy.

The likely roles of water, surface carbonates, and oxygen

vacancies at the metal-support interface are discussed. The result definitively show that, at least in the presence of added water, O-vacancies cannot play an important role in the room temperature catalysis; the water assisted mechanism is far more consistent with the preponderance of the kinetic data.

2 ACS Paragon Plus Environment

Page 3 of 36 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

Journal of the American Chemical Society

Introduction Since Haruta’s seminal discovery of low temperature CO oxidation activity over supported Au nanoparticles,1 gold catalysts have been investigated for a variety of organic oxidations,2-5 oxidative couplings,6-7 hydrogenations,8-9 and other non-aerobic processes.10 Even after 30 years, understanding the unexpectedly high CO oxidation activity of Au catalysts has been an exciting scientific challenge. The potential application of Au catalysts to hydrogen purification via the preferential oxidation of CO (PrOx) reaction,11-17 particularly for fuel cell vehicles,13, 17 has maintained current academic and technological interests. The defining characteristics of CO oxidation over Au catalysts include the large and well documented support effects.18-23 Support choice is critical to generate the highest CO oxidation activity, with “active” oxides (e.g. TiO2 and FeOx) typically associated with support reducibility. Non-reducible supports (e.g. alumina and silica) are typically considered to be less active.13 These effects go beyond the ability to stabilize appropriately small particles, and suggest that the support plays an important role in the catalysis. The early literature quickly demonstrated that three important factors impact CO oxidation over Au catalysts: (i) support identity, (ii) Au particle size, and (iii) Au-support interface structure.19,

24-25

It is now widely agreed that CO oxidation rates are optimized on

particles smaller than 10 nm in diameter; larger particles are basically inert.26 Herzing and coworkers have correlated CO oxidation activity with the presence of small bilayer Au clusters or rafts on some oxide supports,27 but others have found sub-nm clusters and single Au atoms to comparatively inactive.28 The role of cationic Au remains unclear as the presence of Au+ has not been consistently correlated with catalytic activity. For example, Guo et. al recently demonstrated catalytic rates to be much faster on reduced Au0 species present on nanoparticles, even when they coexisted with Au+ and Au3+ species.28 Gates and coworkers have also shown that Au+ is capable of catalyzing CO oxidation, albeit at substantially slower rates.29 Gruenert and co-workers have similarly shown activity by Au(III) at cryogenic temperatures; however, the Au(III) sites were unstable under reaction conditions and CO conversion at higher temperatures was attributed to sites containing exclusively Au0.30 Over the past 30 years, numerous reaction mechanisms have been proposed for CO oxidation over Au.31-35 Particularly noteworthy have been the attempts to explain the correlation 3 ACS Paragon Plus Environment

Journal of the American Chemical Society 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 4 of 36

between support reducibility and activity by a role for support oxygen vacancies in the reaction mechanism.34-37

Originally proposed by Bond, and modified by Haruta to incorporate their

more recent studies,34 the potential involvement of O-vacancies has arguably been the most widely utilized model for the past 15 years.

Indeed, Widmann and Behm have reported

convincing evidence indicating that O-vacancies are involved when the reaction is performed at higher temperatures (≥ 80 °C) under dry UHV conditions.35 Both the Haruta and Behm groups have also provided clear and convincing evidence that CO adsorption onto Au/TiO2 catalysts results in partial reduction of the TiO2 support;38-39 however, the potential role of this interesting chemistry in the catalysis remains unclear. In spite of this evidence, these mechanisms do not readily account for the widely reported promotional effect that water has on the catalysis.40-41 Based on kinetics and isotope effects studies over Au/TiO2, we recently proposed a new mechanism for CO oxidation over Au/TiO2 catalysts that incorporates a clear role for water in the catalysis (Scheme 1).42 This mechanism, which builds upon work from the Iglesia, Davis, Kung, and Haruta groups (among others),23, 4041, 43-47

focuses on the facile proton transfer chemistry associated with water48-49 rather than the

support properties. Although our mechanistic understanding was instrumental in dramatically improving catalytic performance in CO PrOx,11 it did not explain why the support identity has such a strong influence on catalytic activity.

4 ACS Paragon Plus Environment

Page 5 of 36 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

Journal of the American Chemical Society

Scheme 1. Proposed mechanism for CO oxidation over supported gold nanoparticles with the water co-catalyst adsorbed on the support at the metal-support interface. The charge in the blue oval indicates any formal charge distributed over the Au nanoparticle. (1) Adsorption of CO to the Au NP; (2) O2 adsorption and activation at the metal-support interface, mediated by a proton transfer from water adsorbed onto the support hydroxyl groups. This step involves a formal two electron oxidation of the Au NP and results in Au-OOH; (3) Reaction between Au-OOH and adsorbed CO yielding Au-COOH; (4) Rate limiting decomposition of Au-COOH releasing CO2. This step involves proton transfer back to water on the support and a formal two electron reduction of the Au nanoparticle; (5) Reaction between adsorbed CO and O generating a second equivalent of CO2. To better understand the factors that control CO oxidation catalysis over supported Au nanoparticles, we performed a detailed kinetic study of Au/TiO2 and Au/Al2O3 catalysts, allowing us to directly compare gold supported on a reducible (TiO2) to a non-reducible (Al2O3) support. These two catalysts are particularly important given the history of gold catalysts, and the often proposed role of oxygen vacancies and reduced supports in CO oxidation catalysis.34-37 To better understand the nature of the support effects in CO oxidation, we pursued this comparative 5 ACS Paragon Plus Environment

Journal of the American Chemical Society 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 6 of 36

study in the context of our proposed mechanism, asking a simple question:

What makes

Au/TiO2 more active for CO oxidation than Au/Al2O3? The answer to this question sheds further light onto a number of important areas in catalysis, including the role of water and proton transfer chemistry on small molecule activation, the role of the metal-support interface, and the potential role of oxygen vacancies. Results and Discussion Nearly 15 years ago, careful experiments from Haruta’s lab41,

47

first showed that the

presence of water increases the reaction rate; similar results have since been reported by several other groups.40, 45, 50-58 We investigated a pair of commercially available catalysts, both of which were prepared via deposition-precipitation. Figure 1 shows the effects of feed water content on CO oxidation activity for the Au/TiO2 and Au/Al2O3 catalysts. Both catalysts show similar trends, with the reaction rate reaching a maximum at 300-600 Pa water. The associated reaction order plots (Figure 1B) show both catalysts to have water reaction orders of about 0.3, which is generally consistent with what has been reported in the literature.40-42, 47, 50, 59 In Figure 1C, when the reaction order is reported with respect to the amount of surface water on the support, the reaction order increases to ~1.5 (see next section, water adsorption studies, for details). This suggests water may play multiple roles in the reaction mechanism. The Au/TiO2 catalyst is about 10 times more active than the Au/Al2O3 catalyst at the lowest water contents, which is consistent with the literature.23, 45 However, with sufficient water content to reach the maximum reaction rate for each catalyst, Au/TiO2 was only 2-3 times more active. This result is very similar to the results reported by Calla and Davis.45-46, 50 They also found that water has a larger promotional effect on the CO oxidation rate over Au/Al2O3; water also promoted CO oxidation over Au/ TiO2, but the rate enhancement was smaller.45

While the

importance of water has long been appreciated in promoting catalytic activity, this comparison suggests that water may also play a role in the reported support effects for this reaction.18, 22, 60-61

6 ACS Paragon Plus Environment

Page 7 of 36 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

Journal of the American Chemical Society

Figure 1. Effects of feed water content on the CO oxidation reaction rate over Au/TiO2 and Au/Al2O3 catalysts shown through plots of (A) Rate vs. PH2O (B) reaction order plots, and (C) reaction order based on the amount of weakly adsorbed water (wH2O). Reaction conditions: 22 °C, 1% CO, 20% O2, variable H2O, balance N2. Total Flow = 100 mL/min. Prior to reaction, catalysts were dried for 2h at 120 °C under flowing N2, then adsorbed water was equilibrated for 1h using flowing N2 saturated with variable moisture contents. Water adsorption studies. We performed several studies to better understand the effects of adsorbed water on the catalyst. Specifically, we examined the water adsorption energy and the effects of water adsorption on the catalyst surface. Both weakly and strongly adsorbed water exist on the catalysts; the weakly adsorbed water (hereafter “wH2O”) is defined as water that can be removed with flowing N2 at 120 °C. The amount of weakly and strongly adsorbed water was determined with thermogravimetric analysis experiments (details in the SI) and compared with in-situ IR spectroscopy data, as we have previously reported.42 Prior to catalysis experiments, the catalysts were equilibrated with weakly adsorbed water for 1h. We previously showed that increases in reaction rate correspond to changes in the amount of wH2O;42 strongly bound water does not change during catalysis and there does not appear to be a correlation with strongly bound water or surface hydroxyl groups. We speculate that this strongly bound water is likely a combination of water bound to exposed Ti4+ sites and dissociated water in the form of surface hydroxyls. Surface hydroxyl groups are required to help anchor water, but do not appear to be directly involved in the catalysis when water is present.42

7 ACS Paragon Plus Environment

Journal of the American Chemical Society 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 8 of 36

Figure 2. Adsorption isotherms (A) and linear Langmuir plots (B) at 20 °C for weakly adsorbed water on Au/TiO2 and Au/Al2O3. The bending δHOH band (centered at 1640 cm-1) from the IR spectrum was collected using the dry pellet as background (20 mg catalyst, dried under flowing N2 at 120 °C until no changes were evident).

Using infrared spectroscopy to monitor changes in the δ(HOH) bending vibration centered at 1640 cm-1, we measured wH2O adsorption isotherms for both catalysts (Figure 2A). The larger water uptake on Au/Al2O3 is consistent with the higher catalyst surface area (230 m2/g for Au/Al2O3 vs. 45 m2/g for Au/TiO2). This experimental method ensures we only monitor changes in the amount of wH2O, and not any potential changes in the surface hydroxyl groups.62 In this case, however, the band at 1640 cm-1 is linearly correlated with changes in the broad IR absorbance from ~3600-3000 cm-1.42 This indicates that the adsorbed water is involved in hydrogen bonding with the support and/or other water molecules. The wH2O adsorption isotherms in Figure 2A, which were collected at sub-monolayer water coverages, are well 8 ACS Paragon Plus Environment

Page 9 of 36 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

Journal of the American Chemical Society

described by the Langmuir adsorption model, so we used linear Langmuir plots (Figure 2B) to extract an equilibrium constant for water adsorption. The extracted equilibrium constants for these measurements are very similar (0.029 Pa-1 and 0.024 Pa-1 for Au/TiO2 and Au/Al2O3, respectively).

Figure 3. Effects of added water on CO adsorption onto Au/TiO2 at 25 °C. (A) Au-CO peak area, measured via infrared spectroscopy, as a function of water pressure. (B) Au-CO peak area during drying with 1% CO/N2 after saturating with water at several PH2O. In both experiments, the CO peak area was normalized to the maximum CO peak area in the absence of added water.

Although water is a key co-catalyst, the ambient temperature reaction rate declines when water pressures exceed about 600 Pa.40, 42, 47 Since water promotes O2 binding/activation,42 we examined the effect of added water on CO adsorption using infrared spectroscopy. Figure 3 9 ACS Paragon Plus Environment

Journal of the American Chemical Society 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 10 of 36

shows the results of two studies carried out in a transmission IR flow cell. First, we flowed a 1% CO/N2 gas feed over a dried Au/TiO2 sample, allowing CO to equilibrate with the catalyst. The CO containing feed was then passed through a water saturator and the saturator temperature was adjusted to control the feed water content. The sample was equilibrated at each water pressure, and an infrared spectrum was recorded. The plot in Figure 3A clearly shows that the Au-CO peak, which was normalized to the Au-CO peak area with no water in the feed, drops as water was added to the system. At 720 Pa water, where the reaction rate begins to drop, only 35% of the CO adsorption sites have been lost. In the second set of experiments, the catalyst was saturated with a 1% CO + H2O/N2 gas feed. The water was then removed from the feed and the catalyst was dried at 20 °C in flowing 1% CO/N2. Infrared spectra were collected during drying and both the νCO peak (~ 2100 cm-1) and δHOH peak (~ 1640 cm-1) were monitored over time. Figure 3B shows there is an essentially linear relationship between the amount of CO adsorbed on the gold and the amount of water adsorbed on the support. In other words, full CO adsorption is recovered when the wH2O was removed from the support. The water adsorption experiments are consistent with previously reported volumetric measurements for Au/TiO2 and is consistent with water adsorbing on the support.11 There is little experimental precedent for water adsorption on Au at room temperature in the literature and DFT calculations show a strong preference for water binding to the support over Au.42, 63-64 Indeed, the van der Waals interactions between Au and water are so weak that Au is considered to be essentially hydrophobic,63-65 and water adsorption on Au surfaces is only observed at very low temperatures.66-70 Even in these observations, water tends to form 2-dimensional structures in order to maximize water-water hydrogen bonding interactions and minimize the water-gold interactions.67-70 In comparison water readily engages in hydrogen-bonding interactions with support hydroxyl groups.71 The KwH2O values (0.029 and 0.024 Pa-1 for Au/TiO2 and Au/Al2O3, respectively) are likely similar because they describe the energetics of water interacting with surfaces dominated by O-H groups, originating from the support hydroxyls and strongly bound water. It is also worth noting that, the reaction rate is highest when the support is covered with water: the maximum rate occurs when the amount of water adsorbed is roughly equivalent to 1-2 monolayers of H2O on the support (based on the simple Langmuir isotherms in Figure 2).11, 42 10 ACS Paragon Plus Environment

Page 11 of 36 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

Journal of the American Chemical Society

Most importantly, the small differences in water binding between the two catalysts are unlikely to account for the marked differences in catalytic activity.

Scheme 2. Effects of support water coverage on CO adsorption on Au. The blue region represents water adsorbed onto the support

The IR studies are consistent with a physical blocking of CO adsorption sites at or near the metal-support interface. Haruta’s group reached a similar conclusion based on their early work describing the promotional effects of water on these catalysts.41,

47

This effect can be

thought of as a wetting of the support surface, and is pictured in Scheme 2. As greater amounts of water are adsorbed onto the support, more CO adsorption sites are blocked, preventing CO from accessing the entire Au nanoparticle. At high enough water pressures, the water layer can become deep enough that the Au nanoparticles are “flooded” with water and can no longer adsorb CO. We have previously shown that as little as 2 Torr H2O blocks 50% of the CO adsorption sites on Au/TiO2.11 This results in a lower reaction rate at the highest water pressures. As we will discuss further, this indicates that two types of adsorption sites are important for the catalysis: CO adsorption sites on the top of the nanoparticles and Au sites with access to protons from the water adsorbed on the support. These latter sites are required for (fast) O2 activation and (rate determining) Au-COOH decomposition.42

11 ACS Paragon Plus Environment

Journal of the American Chemical Society 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 12 of 36

Figure 4. Infrared spectra of H2O and D2O exchange on (A) Au/TiO2 and (B) Au/Al2O3.

H2O/D2O Exchange and Kinetic Isotope Effects. We previously determined the OH(D) kinetic isotope effect for CO oxidation over Au/TiO2 by exchanging the surface water and hydroxyl groups with D2O (Figure 4).42 The exchange occurs quickly on Au/TiO2, and was complete in 30 minutes. This process was similar on Au/Al2O3; however, this catalyst required longer times (2+ hours) or higher temperatures (120 °C) to complete the exchange. This is at

12 ACS Paragon Plus Environment

Page 13 of 36 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

Journal of the American Chemical Society

least partially due to the higher surface area of the alumina support, and therefore greater total amount of adsorbed water that must be exchanged. Infrared spectra of the two catalysts after the exchange (Figure 4) showed essentially complete exchange of the OH groups for OD after these treatments, with the exception of a small amount of remaining HOD (~3450 cm-1). After completing the H/D exchange, we measured the O-H(D) kinetic isotope effect over Au/Al2O3 (kH/kD = 1.8 ± 0.2; Figure 5A); this is the same value as we previously reported for Au/TiO2 (kH/kD = 1.8 ± 0.1).42 This is the KIE under low water conditions. When 700 Pa H2O/D2O was added to the feed, the reaction rate increased and the KIE value dropped to 1.5 (Figure 5B). For Au/TiO2, the KIE under 700 Pa H2O/D2O was 1.4.42 When the H2O/D2O was removed, the reaction rate decreased, and the KIE increased, approaching the original value. This is consistent with the behavior we observed for Au/TiO2.42 Additionally, the KIE values under ~ 700 Pa H2O are similar to the low KIE values determined by other research groups under high water pressures (kH/kD ≈ 1 to 1.2).40, 43

Figure 5. CO oxidation H(D) kinetic isotope effects experiments for Au/Al2O3. (A) Average KIE for 7 experiments with H2O and D2O exchanged catalysts. (B) Effects of adding and removing 700 Pa H2O/D2O to Au/Al2O3 on the reaction rate. All tests were performed at 20 °C with a 1% CO and 20 % O2 feed (balance N2). Reaction kinetics. Tables 1 and 2 compare the kinetics and thermodynamics parameters for both catalysts at low conversions (< 15% conversion); the reaction parameters for the two catalysts are strikingly similar. The water and CO reaction orders vary within reasonable experimental errors for different amounts of water in the feed. Both catalysts show small fractional oxygen reaction orders; the Au/Al2O3 oxygen order is slightly larger, suggesting small 13 ACS Paragon Plus Environment

Journal of the American Chemical Society 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 14 of 36

differences in O2 binding. Similarly, the reaction order based on the amount of surface water is slightly larger for Au/Al2O3.

Table 1. Kinetics parameters for CO oxidation over Au/TiO2 and Au/Al2O3 at 20 °C. Parameter

Au/TiO2

Au/Al2O3

CO rxn order1

< 0.1

< 0.1

O2 rxn order2

0.1-0.3

0.3-0.4

H2O(g) rxn order3

0.30

0.28

wH2O rxn order3

1.3

1.8

KIE (exchanged)4

1.8 ± 0.1

1.8 ± 0.2

KIE (700 Pa H2O)

1.4

1.5

1

Reaction conditions: Reaction conditions: 3 Reaction conditions: 4 Reaction conditions: 2

20% O2, 0.6-1.1% CO, 0.1 and 60 Pa H2O. 1% CO, 50-300 Pa H2O, 5-25% O2. 1% CO, 20% O2, 50-300 Pa H2O. 1% CO, 20% O2, surface exchanged H2O/D2O.

Table 2. Reaction parameters for CO oxidation over Au/TiO2 and Au/Al2O3 at 20 °C. Parameter

Au/TiO2

Au/Al2O3

KwH2O (ads, atm-1) 1

2890 ± 110

2360 ± 80

KwH2O (kinetics, atm-1) 2

2700 ± 500

2190 ± 150

0.030 ± 0.009

0.066 ± 0.011

34 ± 10

15 ± 3

KR (atm)2 KO2 (atm-1)2 1 2

Equilibrium constant value determined from FTIR water adsorption data with PH2O > 100 Pa. Determined from kinetics measurements; reaction conditions: 1% CO, 50-300 Pa H2O, 5-25% O2.

The reaction kinetics (water binding, CO, and O2 reaction orders; KIE) are essentially the same for Au/Al2O3 and Au/TiO2, so the two catalysts almost certainly operate via the same reaction mechanism under our experimental conditions. This has important implications for other mechanisms suggested in the literature, primarily mechanisms that invoke oxygen defect 14 ACS Paragon Plus Environment

Page 15 of 36 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

Journal of the American Chemical Society

sites and/or a reduction of the titania support. In essence, the reaction kinetics indicate that, at least in the presence of several hundred Pa water, this reducibility (which leads to O-vacancies on the surface) does not play an important role in CO oxidation catalysis. The logic behind this conclusion is as follows: alumina cannot be reduced under these reaction conditions; therefore, oxygen vacancies cannot be invoked in the reaction mechanism for Au/Al2O3. Since the kinetics are essentially identical for Au/Al2O3 and Au/TiO2, the two catalysts likely operate via the same reaction mechanism. Therefore, oxygen vacancies and the reducibility of titania should not be invoked to explain the catalysis by Au/TiO2 under these reaction conditions. The strong positive effect of water on the catalysis is also inconsistent with O-vacancies playing an important role. Indeed, the maximum reaction rate coincides with an amount of adsorbed water roughly equivalent to 1-2 monolayers of H2O on the support. Under these conditions, (i) highly reactive O-vacancies should quickly react to reduce protons to H2 and (ii) any remaining O-vacancies will be blocked with adsorbed water and therefore cannot directly participate in the catalysis. There is similarly little evidence for oxygen activation on alumina, which would be necessary for a Mars-Van Krevelen type mechanism. Therefore, it is unlikely that O2 activation on the support is an important pathway for Au/TiO2, at least under reaction conditions where the catalysis is fastest (i.e. with an optimal amount of water on the support). We note that other mechanisms are certainly possible under other reaction conditions. Indeed, the mechanism proposed by Neurock and Yates, which invokes a Ti-O-O-Au intermediate under low temperature UHV conditions, goes through fundamentally very similar O2 activation chemistry as the mechanism that we have proposed.

72-73

There may also be

reaction conditions where the reducibility of titania plays a more important role in catalytic activity. In particular, a number of elegant experiments by Widmann and Behm have provided convincing evidence that such mechanisms are likely available at higher temperatures where support surfaces have much lower surface water coverages.12, 18, 35, 74 Similarly, we cannot comment directly on mechanisms that might operate on other supports beyond suggesting that the water assisted mechanism should be available to most CO oxidation catalysts, provided that the support can supply water to the gold nanoparticles. Indeed, Davis and coworkers have shown that Au/C, which is completely inactive for gas phase CO oxidation, becomes quite active when immersed in 1 M NaOH.2, 53 It would therefore be prudent to investigate the possible mechanistic role(s) of water in the catalysis on other supports. Water 15 ACS Paragon Plus Environment

Journal of the American Chemical Society 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 16 of 36

is extremely difficult to exclude from oxide supports, even when using UHP gases. For example, we have observed that passing UHP gases through a -80 °C dry ice/isopropanol trap removed sufficient residual water to reduce the reaction rate over Au/TiO2 by about 30%. Once the role of water is properly evaluated, the reducibility of the support may indeed be important for some metal oxides, under certain conditions, or for other reactions. However, based on the reaction kinetics of Au/Al2O3 and Au/TiO2, and contrary to many reports in the literature, the reducibility of titania does not appear to play a significant role in the fast room temperature CO oxidation. Rate Law and Active Site Model. Our previous work42, 75-78 utilized an “active site” model for the catalysis, employing many of the core ideas of Michaelis-Menten kinetics. This kinetic treatment can be a useful tool for evaluating and comparing catalysts. It has been particularly useful in identifying changes to catalytic activity that result from changes in the number of active sites42,

75-77

versus changes to the inherent reactivity of the active site.78 However, the

characterization mechanism that we originally developed77 did not include a clear role for water, and was therefore incomplete. In order to properly compare the two catalysts, we derived a complete rate law based on the mechanism shown in Scheme 1. In this mechanism, the rate determining step is the decomposition of Au-COOH, which involves a proton transfer from – COOH to water as CO2 is released by the catalyst. The full derivation can be found in the SI; we highlight only the key issues and assumptions here. The overall reaction can be broken down into the following elementary steps:



(1) 2   + ↔  − 



(2) ∗ +  +  ↔ ∗ ( )



(3) ∗ ( ) +  ↔ ∗ −  + (   )



(4) ∗ −  +  − ↔  −  + ∗ − !"

(5)  −  + ( ) →  +  + ($ % ) &' !)

(6) ∗ − +  − →  +  +  *+,- ./01 01+

3

(7) (  ) + ($ % ) 45 2 ( ) *+,- ./01 01+

(8) 789:;?7@: 2 +  → 2 

The reaction kinetics and proposed mechanism indicate that a far more nuanced view of the active site must be employed for this reaction. In essence, there are three types of important 16 ACS Paragon Plus Environment

Page 17 of 36 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

Journal of the American Chemical Society

sites: (i) Au sites away from the interface where CO binding occurs (Step 1), (ii) support sites at the metal-support interface (MSI) where water binding occurs, (iii) Au sites at or near the MSI where O2 binding can occur (Steps 2 and 3). We note that Step 2 is included to express that an MSI site must have a water molecule adsorbed on the support at the MSI in order to provide a proton to the incoming O2. This water could come from the gas phase or from the support on non-MSI support sites.

Assuming that water adsorption on the support at the MSI is not

drastically different than water adsorption on the remainder of the support (this is very likely), this equilibrium is essentially the same at both types of support sites. This also means that the fraction of Au* sites available for O2 adsorption should be roughly equivalent to the water coverage on the bulk of the support, and we can consider water transport from the support to be equivalent to direct adsorption from the gas phase. There is a critical interplay between the amount of adsorbed water and the fraction of the Au perimeter sites that might be active for O2 binding and Au-COOH decomposition. To deal with this complexity, we defined two types of Au sites. First, the Au sites at or near the MSI with access to water adsorbed on the support are defined as Au* sites. The access to water provides these sites with the capability of binding O2 as Au*-OOH (Step 3). Secondly, we define the CO binding sites on Au as Au’ sites.

Table 3. Equilibrium constants for the proposed mechanism. BC =

[ − ] [ ]GHI

B = BJKI

[ ∗ ( ) ] = [∗ ][ ]GK I

[ ∗ − ][(   ) ] B$ = [ ∗ ( ) ]GI B$

[∗ − ] B$ = = [∗ ( ) ]GI [(   ) ]

[ − ][∗ − ] BL = [ ∗ − ][ − ]

17 ACS Paragon Plus Environment

Journal of the American Chemical Society 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

BL =

Page 18 of 36

[ − ] BL = [ ∗ − ][ − ] [ ∗ − ]

The reaction mechanism then consists of the following equilibrated (or quasi-equilibrated) substrate binding steps: (1) CO to the Au’ sites described by K1; (2) wH2O binding to the support described by K2, which is equivalent to KwH2O measured by IR spectroscopy; and (3) O2 binding and activation at a MSI site yielding Au*-OOH, described by K3. Equilibrium constants are provided in Table 3. The O2 binding and activation reaction results in the generation of OHon the support; the concentration of this species is likely constant during the reaction, so it is included in the modified equilibrium constant K3’. The equilibrium reaction between Au*-OOH and Au’-CO (Step 4) is also considered to be fast; this is justified based on the high reactivity of Au*-OOH predicted by DFT calculations.42 As in the case of O2 binding and activation, this step results in a species that likely has a constant concentration during the reaction (Au*-O), so it is described by the modified equilibrium constant K4’. A site balance equation can then be written for each type of Au binding site as follows: (9)

[ M ] = [ ] + [ − ]

(10)

[∗M ] = [ ∗ ] + [∗ ( ) ] + [∗ − ] + [∗ − ]

And the experimental rate given by Step 5, the rate determining step, can be written as: (11)

&/1+ = NO [ − ][( ) ]

Using the above expressions, the following rate law can be derived: (12)

&/1+ = NO BL [( ) ] P

Q RS

C% Q RS

[∗M ]T U

 RVS

C%  RVS

[ M ]W

This expression is essentially the Langmuir-Hinshelwood expression where the rate depends on adsorbed water, activated oxygen coverage on the Au* sites, and adsorbed carbon monoxide coverage on the Au’ sites. Under our experimental conditions BC GHI ≫ 1,79-81 therefore the last

term reduces to [ M ] (this term represents the total number of CO binding sites in close proximity to the O2 binding sites). The overall rate then becomes: (13)

&/1+ = NO BL [( ) ][ M ] P

Q RS

C% Q RS

[∗M ]T

18 ACS Paragon Plus Environment

Page 19 of 36 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

Journal of the American Chemical Society

Inverting this rate expression in order to create a double-reciprocal expression yields: (14)

C

YZ[\

C

= Q !

e

C

Q ∗ Q  "  [(K I)]^_ ]`abc d`abc d RS

f + !

C

Q ∗ Q "  [(K I)]^_ ]`abc d`abc d

The terms ghZi and BY can be defined as (15)

ghZi = NO BL [( ) ][ M ][∗M ] and

BY = 1/B$

Yielding the simplified expression: (16)

C

YZ[\

= l

k

mno

C

eR f + l S

C

mno

Figure 6. (A) Double reciprocal plots for the O2 dependence data over Au/Al2O3 with varying amounts of water in the feed and (B) extracted KR and νmax values plotted against the wH2O coverage on the support.

19 ACS Paragon Plus Environment

Journal of the American Chemical Society 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 20 of 36

Note that the KR value is essentially the inverse of the O2 binding constant (KO2), which should be independent of the water coverage. This is consistent with our original interpretation of KR based on the incomplete reaction mechanism.77 Double reciprocal plots for the Au/Al2O3 catalyst are shown in Figure 6A and in the SI; data for the Au/TiO2 catalyst have been previously reported.42 Figure 6B also plots the extracted KR and νmax values as a function of the normalized surface coverage of wH2O on the support. These plots are in excellent agreement with the rate law derived above. For each catalyst, the KR values are constant within a very reasonable margin of error and the νmax values vary linearly with water coverage. This suggests that added water has little effect on the electronics of either O2 binding or Au-COOH decomposition. This is consistent with the proposed mechanism, and indicates water’s primary role as a proton donor / acceptor. Table 2 shows the extracted KR values for Au/TiO2 and Au/Al2O3 differ only slightly, indicating that O2 binding as Au*-OOH is similar on the two catalysts. The above kinetic treatment allows us to extract the net O2 binding and activation equilibrium constants (K3’) from the KR determinations (Equations 13 & 14). The extracted values (34 atm-1 and 15 atm-1 for Au/TiO2 and Au/Al2O3, respectively) are reasonable in the context of the literature on Au catalysts: if there is one consistent conclusion throughout the literature on Au chemistry and catalysis, it is that O2 binding is exceptionally weak. However, the K3’ values are somewhat greater than unity; this is consistent with our DFT calculations, which indicate that O2 binding as Au*-OOH is thermodynamically favorable.42 The key to controlling O2 binding, then, appears to lie in controlling the amount of water on the catalyst surface in order to maximize the number of O2 binding sites. As an additional test of the proposed mechanism, we examined the kinetic water binding data by considering the reaction rate as a function of feed water content at constant O2 pressures. The rate law in Eq. 11 indicates that the reaction rate is proportional to the amount of wH2O at the MSI; the νmax values also show a dependence on water coverage. Using the water binding equilibrium expression (K2 = KwH2O, Table 2) and a site balance for the MSI sites on the support (Eq. 16), we expanded the rate law to include a Langmuir-Hinshelwood term for water adsorption on the support: (17)

[ ] M = [ ] + [∗ ( ) ] 20 ACS Paragon Plus Environment

Page 21 of 36 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

Journal of the American Chemical Society

(18)

3p S Rp S

&/1+ = NO BL PC%

3p S Rp S

Q R

RVS

 S [∗M ]T U  [ ] M T PC% Q R C% 

 RVS

S

[ M ]W

Under constant O2 and CO pressures, the O2 and CO Langmuir-Hinshelwood terms become constant, allowing the rate law to be simplified82 to (19)

&/1+ = NO BL [ ∗M ][ M ] P

3p S Rp S

C% 3p S Rp S

[ ] M T

A double reciprocal plot of this equation yields: (20)

C

YZ[\

= !

C

Q ∗ Q "  `abc d`abc d[bqq]^_ ]c 3p S

eR

C

p S

f + !

C

Q ∗ Q "  `abc d`abc d[bqq]^_ ]c

This allows for the extraction of KwH2O from the kinetic data. Double reciprocal plots for this data are in the SI. The extracted KwH2O values, which can be found in Table 2, are in excellent agreement with the binding constants measured independently with IR spectroscopy.

This

provides very strong support for the proposed mechanism and the kinetic treatment that we have employed. Origins of differences in catalytic activity. The KO2 values in Table 2 suggest that O2 binding may be slightly stronger on the Au/TiO2 catalyst, but this difference cannot account for the difference in catalytic activity. Indeed, all the kinetic and thermodynamic parameters are essentially the same, with the exception of the overall rate. Based on the similarities in all of the values determined from the kinetic and thermodynamic data, we conclude that any electronic differences between catalysts arising from a combination of particle size differences and different support interactions are not large enough to cause the observed differences in activity. The only substantial difference between the two catalysts is the sensitivity to the amount of wH2O. This is best seen in the νmax vs. water plots in Figure 6B, where the slope of the line for the Au/Al2O3 catalyst is far steeper than for Au/TiO2. Given that the electronics of the catalysts have proven to be only subtly different, we are therefore left to conclude that the difference in activity is primarily due to differences in the number of active sites available during the reaction. However, the nanoparticles on Au/Al2O3 (~2 nm, see SI) are smaller than on

21 ACS Paragon Plus Environment

Journal of the American Chemical Society 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 22 of 36

Au/TiO2 (~3 nm, see SI); the Au/Al2O3 catalyst therefore has a greater number of gold perimeter sites at the MSI, and would be expected to have higher activity, not lower. Exhausting these possibilities, we now consider the potential role of surface carbonates. We have previously shown the loss of catalytic activity was closely correlated to the appearance of carbonates on the support of Au/TiO2 catalysts.75 Other groups have also reported significant carbonate poisoning.16,

52, 54, 83-88

We therefore examined potential differences in carbonate

production and carbonate stability between Au/TiO2 and Au/Al2O3. Figure 7 shows IR spectra collected from both catalysts during an experiment designed to evaluate carbonate stability. In this experiment, we first deposited carbonates (IR bands in the 1700-1200 cm-1 region are attributed to carbonates)75 onto the Au/TiO2 catalyst by drying the catalyst and performing CO oxidation for 10 minutes. Similar to the careful work by Behm’s group, who convincingly showed the role of carbonate poisoning,52,

85-88

we found that these carbonates were readily

removed by flowing 600 Pa of water over the catalyst for 30 minutes. The last spectrum (C) in the top panel of Figure 7 shows the cleaned catalyst after an additional 30 minutes of flowing N2 to remove the surface water.

Figure 7. IR spectra of Au/TiO2 (top) and Au/Al2O3 (bottom) after (A) CO oxidation catalysis. The broad bands in the 1700-1200 cm-1 range of the spectra are associated with adsorbed carbonates. (B) H2O treatment (600 Pa in N2) to remove surface carbonates, and (C) flushing with N2 to remove surface water (note the different temperatures required to clean the two catalysts)

22 ACS Paragon Plus Environment

Page 23 of 36 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

Journal of the American Chemical Society

We then performed the same experiment with Au/Al2O3. Note that the absorbance scale bar for the Au/Al2O3 spectra is an order of magnitude larger than for Au/TiO2. This comparison is somewhat qualitative, since the surface area of the Au/Al2O3 catalyst is about 5x larger than that of the Au/TiO2 catalyst. Although the comparison is not perfect, the result suggests that carbonates are more efficiently delivered to the support on the Au/Al2O3 catalyst. We were unable to completely remove the carbonates on Au/Al2O3 by flowing water at room temperature over the catalyst for up to an hour. We then determined that a temperature of 60 °C was required to completely remove the carbonates with 600 Pa water (time was kept constant in order to directly compare to the Au/TiO2 experiment). Based on these experiments, it appears that the carbonates produced during the reaction over Au/Al2O3 are more strongly adsorbed onto the support, which results in a greater surface coverage of carbonate once the steady state is reached. While we cannot directly observe the chemistry at the Au perimeter sites, it seems reasonable to assume that these differences in carbonate stability are similar at the MSI. We performed an additional experiment to probe the mechanism of carbonate poisoning. Using a single sample of the Au/TiO2 catalyst, we used infrared spectroscopy to monitor how carbonate production affects the catalyst’s water adsorption capacity (Figure 8). First, the catalyst was heated under 20% H2/N2 at 250 °C for 1 hour and cooled to room temperature. The catalyst was then placed under CO oxidation conditions (1% CO, 20% O2, balance N2) with no added water.

Our previous work has shown that this treatment produces large amounts of

carbonates on the surface and largely poisons the catalytic activity.75

23 ACS Paragon Plus Environment

Journal of the American Chemical Society

750

Fresh

δHOH area (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

Page 24 of 36

600

Regenerated

450

After CO oxidation

300

Treatment Fresh After CO Ox. Regenerated

150

Amax 936 632 854

Figure 8. Adsorption isotherms for wH2O on Au/TiO2. The fresh catalyst was treated under H2/N2 at 250 °C for 1 hour, cooled to room temperature, and treated with 1% CO + 20% O2 (balance N2) for 20 min to generate surface carbonates (see SI); the regenerated catalyst was treated with 600 Pa H2O/N2 for 1 hour at 120 °C. Maximum area (Amax) values in the table were extracted from linear Langmuir plots of the adsorption data.

0 0

50

100

150

200

250

PH2O (Pa)

Figure 8 shows that surface carbonates produced during catalysis clearly reduce the wH2O adsorption capacity of the catalyst. This shows that one of the mechanisms by which carbonates poison the catalytic activity is by lowering the water coverage on the support and reducing the number of wH2O binding sites. Finally, although we cannot directly probe the MSI, carbonates are produced by the reaction on the Au NP. It is therefore likely that site blocking at the MSI (where carbonates are produced)52, 85-88 is more severe than on the support surface as a whole. We explored the possibility of regenerating the catalyst with a water treatment, as Behm’s group has suggested.16, 88 Treating the carbonate covered Au/TiO2 catalyst with 600 Pa H2O/N2 at 120 °C for 1 hour, and cooling under the H2O/N2 returned the wH2O adsorption capacity to essentially its original value. These experiments, along with the catalysis data suggest that there is an intimate interplay between adsorbed water and carbonates in determining the reaction rate and catalytic activity. This interplay can essentially be seen as a competition between water and carbonates for surface sites on the support. Carbonates bind more strongly than water, but nominally high pressures of water also can be used to remove the carbonates. For supports where carbonates bind more strongly, such as alumina, higher temperatures are required to remove the carbonates. Considering this competition, it is likely that operating these catalysts with relatively high water coverage may improve both catalyst activity and stability by preventing or slowing carbonate deposition.11

24 ACS Paragon Plus Environment

Page 25 of 36 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

Journal of the American Chemical Society

Thus, based on all the thermodynamic and kinetic evidence, we conclude that the stronger binding of carbonates to the alumina support is the primary cause of the lower activity for Au/alumina.

While it is likely that there are small electronic differences between the

catalysts, at least for this reaction, those differences appear to be less important than the role of water and the deleterious role of carbonates. The support plays a pivotal role in CO oxidation catalysis, and the choice of support clearly impacts the catalytic activity; however, at least at ambient temperatures when water is present, it appears that the support’s largest role is in supplying water to the catalytic active site and mitigating carbonate poisoning.

Conclusions Reaction kinetics (CO, O2, H2O reaction orders, O-H(D) KIE) and water binding measurements were essentially the same for Au/TiO2 and Au/Al2O3 catalysts. This kinetic and mechanistic investigation shows that the two catalysts operate by essentially the same reaction mechanism under these conditions (ambient temperature with water present in the reaction feed). Since the kinetics are nearly identical despite the large differences in support reproducibility, and the catalysis is fastest when the support is completely covered with water, it is unlikely that Ovacancies associated with the reducibility of titania play a significant role in the catalysis under these conditions. This conclusion is contrary to much of what appears in the literature.

Employing an active site or Michaelis-Menten treatment of the kinetic data provided substantial insight into the role of water in the catalysis. In particular, the amount of water on the catalysts was shown to affect the number of available reaction sites, but had no significant effect on the intrinsic rate constants, which also appear to be similar on both catalysts. We therefore proposed a mechanism for the reaction that deals with three distinct but important sites for the catalysis: (i) Au-CO binding sites away from the MSI, (ii) water binding sites on the support at the MSI, and (iii) Au sites near the MSI where O2 activation and Au-COOH decomposition can

25 ACS Paragon Plus Environment

Journal of the American Chemical Society 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 26 of 36

occur. The latter of these sites must have access to water on the support in order for the proton transfer chemistry to occur. The primary differences between the two catalysts appear to be the number of active sites and the sensitivity to carbonate poisoning. In particular, carbonates bind more strongly to Au/Al2O3, which in turn, is more sensitive to carbonate poisoning. The mechanism of carbonate poisoning is not entirely clear, although it appears to be associated with either blocking the MSI sites or sequestering water from those sites. This is reflected in the higher sensitivity to the amount of water in the feed for Au/Al2O3. This mechanistic understanding provides clear guidance for improving Au based CO oxidation catalysts. Namely, improved catalysts should seek to (i) maximize the Au-support interface with small particles, (ii) maximize water adsorption at or near the MSI, and (iii) minimize carbonate adsorption on the support and decrease the stability of surface carbonates.

Materials and Methods Materials. The catalysts used in this study were commercial AUROlite™ samples purchased from STREM Chemicals (nominal 1% Au/TiO2 and Au/Al2O3). These catalysts were prepared by deposition-precipitation and pretreated by the manufacturer to ensure that particles were of appropriate size (2.9 ± 0.9 nm for Au/TiO2, and 2.2 ± 0.7 nm for Au/Al2O3; details in SI) to be active for CO oxidation. BET surface areas were determined to be 45 m2/g for Au/TiO2 and 230 m2/g for Au/Al2O3. The catalyst was crushed and stored in a dark refrigerator. Powdered Silicon Carbide (400 mesh) was purchased from Aldrich. Gases (N2, H2, O2, and 5% CO/He) were 5.0 grade supplied by Praxair and used with no additional purification. Deuterium oxide (99.9%) was purchased from Cambridge Isotope Laboratories. Water was purified to a resistivity of 18.6 Ω with a Barnstead Nanopure system. CO Oxidation Catalysis. The CO oxidation reactor consisted of a home-built laboratory scale single pass plug-flow micro-reactor. The reaction zone consisted of finely ground fresh catalyst (4 mg Au/TiO2 or 12 mg Au/Al2O3) diluted in 750 mg of silicon carbide. Gas flows were controlled with 4 electronic low pressure mass flow controllers (Porter Instruments). The composition of the feed and reactor effluent (CO and CO2) was determined using a Siemens Ultramat 23 IR gas analyzer.

26 ACS Paragon Plus Environment

Page 27 of 36 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

Journal of the American Chemical Society

After loading into a glass U-tube, the diluted catalyst was stabilized for 4h with variable contents of moisture in the gas (0.1 to 700 Pa). CO oxidation reaction rates were measured in a 60 min experiment immediately following the pretreatment. The feed (1% CO, 20% O2, balance N2, 180 mL/min; WHSV = 2.2 x 103 L/h/gcat) was held constant and the reaction temperature was maintained at 20 ºC using a water bath. Moisture in the gaseous feeds used in these experiments was controlled using a coil submerged into a dry ice trap (-78.5 °C), which removes nascent water to an approximate value of 0.1 Pa. Moisture in the gas was controlled by subsequently passing the gas through a water column submerged in a dry ice / isopropanol bath set at a constant, but adjustable temperature (-78.5 to 20 °C). Prior to CO oxidation measurements, the catalyst was equilibrated with flowing water vapor (1, 5, 20, 50, or 125 Pa) for 4 hours. O2 reaction orders were determined with 10, 15, 18, 20, 24 % O2 (by volume). CO reaction orders were performed using 0.56, 0.80, 1.0, 1.2, and 1.4 % CO (by volume) at 0 and 60 Pa of water vapor. The O2 content was held constant at 20%. All catalytic rates are normalized to the total amount of Au in the catalyst; no additional adjustments to account for the number of active sites were made. Infrared spectroscopy A previously described home built flow IR cell was used for the in situ FTIR experiments.75, 89-92 The cell consisted of a stainless steel chamber wrapped by a heating mantle (up to 400⁰C) with two IR transparent KBr windows. Gases were mixed in an external stainless steel manifold using low pressure rotameters. The gas cell was fed with 100 mL/min of gas mixtures and heated at a constant rate of 5⁰ C/min. The catalyst (~25 mg) was finely ground in an agate mortar, pressed into a 13 mm circular pellet using a stainless steel die and a manual hydraulic press (5 metric tons of pressure for 2 min). The pellet was mounted into the cell and placed in the sample compartment of a Nicolet FTIR spectrometer where a thermocouple adjacent to the pellet monitored the temperature. Prior to any treatment, a background spectrum of the catalyst pellet in the sample cell was collected. The catalyst pretreatment was carried out in situ using 100 mL/min of the gas mixture at different temperatures (150, 250 and 350 ºC; WHSV= 2.00 x 102 L·h-1·gcat-1). After treatment, the sample was purged for 1 hr and cooled to room temperature under N2 flow. Spectra were collected at 20 ºC and the temperature was kept constant using a coil with recirculating water.

27 ACS Paragon Plus Environment

Journal of the American Chemical Society 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 28 of 36

Immediately after treatment, the sample was cooled and a reference spectrum was collected once the temperature was equilibrated at 20 ºC.

To measure water adsorption

isotherms, the moisture in the gas was controlled by passing the feed through a water column submerged in a dry ice / isopropanol bath set at a constant, but adjustable temperature (-78.5 to 20 °C). The catalyst pellet was then used for the in situ CO oxidation reaction. The CO oxidation reaction consisted of 4 steps: (a) 10 min flowing with a 1% CO mixture, (b) 10 min with our standard CO oxidation mixture (1% CO, 20% O2), (c) 10 min flowing with a 1% CO, and (d) 10 min purging with N2. Spectra were collected every 2.5 min.

Acknowledgments: The authors gratefully acknowledge the U.S. National Science Foundation (Grant numbers CBET-1160217, CHE-1012395) for financial support of this work. Supporting Information: Methods and materials, H/D Isotope Exchange and KIE measurements, Additional Kinetics and Spectroscopy Data, Full Rate Law Derivation.

References 1. Bamwenda, G. R.; Tsubota, S.; Nakamura, T.; Haruta, M., The influence of the preparation methods on the catalytic activity of platinum and gold supported on TiO2 for CO oxidation. Catalysis Letters 1997, 44 (1,2), 83-87. 2. Ide, M. S.; Davis, R. J., The Important Role of Hydroxyl on Oxidation Catalysis by Gold Nanoparticles. Acc. Chem. Res. 2014, 47 (3), 825-833. 3. DellaPina, C.; Falletta, E.; Rossi, M., Update on selective oxidation using gold. Chem. Soc. Rev. 2012, 41 (1), 350-369. 4. Corma, A.; Garcia, H., Supported gold nanoparticles as catalysts for organic reactions. Chemical Society Reviews 2008, 37 (9), 2096-2126. 5. Turner, M.; Golovko, V. B.; Vaughan, O. P. H.; Abdulkin, P.; Berenguer-Murcia, A.; Tikhov, M. S.; Johnson, B. F. G.; Lambert, R. M., Selective oxidation with dioxygen by gold nanoparticle catalysts derived from 55-atom clusters. Nature 2008, 454 (7207), 981. 6. Xu, B.; Madix, R. J.; Friend, C. M., Predicting Gold-Mediated Catalytic OxidativeCoupling Reactions from Single Crystal Studies. Accounts of Chemical Research 2014, 47 (3), 761-772.

28 ACS Paragon Plus Environment

Page 29 of 36 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

Journal of the American Chemical Society

7. Grirrane, A.; Corma, A.; Garcia, H., Gold-Catalyzed Synthesis of Aromatic Azo Compounds from Anilines and Nitroaromatics. Science 2008, 322 (5908), 1661-1664. 8. Corma, A.; Serna, P., Chemoselective Hydrogenation of Nitro Compounds with Supported Gold Catalysts. Science 2006, 313 (5785), 332-334. 9. Ward, T.; Delannoy, L.; Hahn, R.; Kendell, S.; Pursell, C. J.; Louis, C.; Chandler, B. D., Effects of Pd on Catalysis by Au: CO Adsorption, CO Oxidation, and Cyclohexene Hydrogenation by Supported Au and Pd-Au Catalysts. ACS Catal. 2013, 3 (11), 2644-2653. 10. Stratakis, M.; Garcia, H., Catalysis by Supported Gold Nanoparticles: Beyond Aerobic Oxidative Processes. Chem. Rev. 2012, 112 (8), 4469-4506. 11. Saavedra, J.; Whittaker, T.; Chen, Z.; Pursell, C. J.; Rioux, R. M.; Chandler, B. D., Controlling activity and selectivity using water in the Au-catalysed preferential oxidation of CO in H2. Nat Chem 2016, 8 (6), 584-589. 12. Widmann, D.; Hocking, E.; Behm, R. J., On the origin of the selectivity in the preferential CO oxidation on Au/TiO2 - Nature of the active oxygen species for H2 oxidation. J. Catal. 2014, 317, 272-276. 13. Lakshmanan, P.; Park, J.; Park, E., Recent Advances in Preferential Oxidation of CO in H2 Over Gold Catalysts. Catal. Surv. Asia 2014, 18 (2-3), 75-88. 14. Maeda, N.; Meemken, F.; Baiker, A., Insight into the Mechanism of the Preferential Oxidation of Carbon Monoxide by Using Isotope-Modulated Excitation IR Spectroscopy. ChemCatChem 2013, 5 (8), 2199-2202. 15. Kandoi, S.; Gokhale, A. A.; Grabow, L. C.; Dumesic, J. A.; Mavrikakis, M., Why Au and Cu are more selective than Pt for preferential oxidation of CO at low temperature. Catal. Lett. 2004, 93 (1-2), 93-100. 16. Schubert, M. M.; Venugopal, A.; Kahlich, M. J.; Plzak, V.; Behm, R. J., Influence of H2O and CO2 on the selective CO oxidation in H2-rich gases over Au/α-Fe2O3. J. Catal. 2004, 222 (1), 32-40. 17. Liu, K.; Wang, A.; Zhang, T., Recent Advances in Preferential Oxidation of CO Reaction over Platinum Group Metal Catalysts. ACS Catal. 2012, 2 (6), 1165-1178. 18. Widmann, D.; Liu, Y.; Schueth, F.; Behm, R. J., Support effects in the Au-catalyzed CO oxidation - Correlation between activity, oxygen storage capacity, and support reducibility. J. Catal. 2010, 276 (2), 292-305. 19. Haruta, M.; Tsubota, S.; Kobayashi, T.; Kageyama, H.; Genet, M. J.; Delmon, B., Lowtemperature oxidation of carbon monoxide over gold supported on titanium dioxide, α-ferric oxide, and cobalt tetraoxide. J. Catal. 1993, 144 (1), 175-92.

29 ACS Paragon Plus Environment

Journal of the American Chemical Society 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 30 of 36

20. Schubert, M. M.; Hackenberg, S.; van Veen, A. C.; Muhler, M.; Plzak, V.; Behm, R. J., CO Oxidation over Supported Gold Catalysts-"Inert" and "Active" Support Materials and Their Role for the Oxygen Supply during Reaction. J. Catal. 2001, 197 (1), 113-122. 21. Lopez, N.; Janssens, T. V. W.; Clausen, B. S.; Xu, Y.; Mavrikakis, M.; Bligaard, T.; Norskov, J. K., On the origin of the catalytic activity of gold nanoparticles for low-temperature CO oxidation. Journal of Catalysis 2004, 223 (1), 232-235. 22. Comotti, M.; Li, W.-C.; Spliethoff, B.; Schueth, F., Support Effect in High Activity Gold Catalysts for CO Oxidation. Journal of the American Chemical Society 2006, 128 (3), 917-924. 23. Kung, M. C.; Davis, R. J.; Kung, H. H., Understanding Au-Catalyzed Low-Temperature CO Oxidation. J. Phys. Chem. C 2007, 111 (32), 11767-11775. 24. Haruta, M.; Kobayashi, T.; Sano, H.; Yamada, N., Novel gold catalysts for the oxidation of carbon monoxide at a temperature far below 0°C. Chem. Lett. 1987, (2), 405-8. 25. Haruta, M.; Yamada, N.; Kobayashi, T.; Iijima, S., Gold catalysts prepared by coprecipitation for low-temperature oxidation of hydrogen and of carbon monoxide. J. Catal. 1989, 115 (2), 301-9. 26. Hutchings, G. J., Nanocrystalline gold and gold-palladium alloy oxidation catalysts: a personal reflection on the nature of the active sites. Dalton Transactions 2008, (41), 5523-5536. 27. Herzing, A. A.; Kiely, C. J.; Carley, A. F.; Landon, P.; Hutchings, G. J., Identification of Active Gold Nanoclusters on Iron Oxide Supports for CO Oxidation. Science 2008, 321, 13311335. 28. Guo, L.-W.; Du, P.-P.; Fu, X.-P.; Ma, C.; Zeng, J.; Si, R.; Huang, Y.-Y.; Jia, C.-J.; Zhang, Y.-W.; Yan, C.-H., Contributions of distinct gold species to catalytic reactivity for carbon monoxide oxidation. Nature Communications 2016, 7, 13481. 29. Aguilar-Guerrero, V.; Gates, B. C., Kinetics of CO oxidation catalyzed by highly dispersed CeO2-supported gold. Journal of Catalysis 2008, 260 (2), 351-357. 30. Gruenert, W.; Grossmann, D.; Noei, H.; Pohl, M.-M.; Sinev, I.; De Toni, A.; Wang, Y.; Muhler, M., Low-Temperature Oxidation of Carbon Monoxide with Gold(III) Ions Supported on Titanium Oxide. Angew. Chem., Int. Ed. 2014, 53, 3245-3249. 31. Bollinger, M. A.; Vannice, M. A., A kinetic and DRIFTS study of low-temperature carbon monoxide oxidation over Au-TiO2 catalysts. Appl. Catal., B 1996, 8 (Copyright (C) 2012 American Chemical Society (ACS). All Rights Reserved.), 417-443. 32. Chen, M. S.; Goodman, D. W., The Structure of Catalytically Active Gold on Titania. Science 2004, 306 (5694), 252-255. 33. Bond, G. C.; Louis, C.; Thompson, D. T., Catalysis by Gold Imperial College Press: London, 2006; Vol. 6, p 366. 30 ACS Paragon Plus Environment

Page 31 of 36 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

Journal of the American Chemical Society

34. Haruta, M., Spiers Memorial Lecture Role of perimeter interfaces in catalysis by gold nanoparticles. Faraday Discuss. 2011, 152 (Gold), 11-32. 35. Widmann, D.; Behm, R. J., Activation of Molecular Oxygen and the Nature of the Active Oxygen Species for CO Oxidation on Oxide Supported Au Catalysts. Acc. Chem. Res. 2014, 47 (3), 740-749. 36. Takei, T.; Akita, T.; Nakamura, I.; Fujitani, T.; Okumura, M.; Okazaki, K.; Huang, J.; Ishida, T.; Haruta, M., Heterogeneous catalysis by gold. Adv. Catal. 2012, 55, 1-126. 37. Romero-Sarria, F.; Plata, J. J.; Laguna, O. H.; Marquez, A. M.; Centeno, M. A.; Sanz, J. F.; Odriozola, J. A., Surface oxygen vacancies in gold based catalysts for CO oxidation. RSC Adv. 2014, 4 (25), 13145-13152. 38. Okumura, M.; Coronado, J. M.; Soria, J.; Haruta, M.; Conesa, J. C., EPR Study of CO and O2 Interaction with Supported Au Catalysts. J. Catal. 2001, 203 (1), 168-174. 39. Widmann, D.; Krautsieder, A.; Walter, P.; Brueckner, A.; Behm, R. J., How Temperature Affects the Mechanism of CO Oxidation on Au/TiO2: A Combined EPR and TAP Reactor Study of the Reactive Removal of TiO2 Surface Lattice Oxygen in Au/TiO2 by CO. ACS Catal. 2016, 6 (8), 5005-5011. 40. Ojeda, M.; Zhan, B.-Z.; Iglesia, E., Mechanistic interpretation of CO oxidation turnover rates on supported Au clusters. J. Catal. 2012, 285 (1), 92-102. 41. Date, M.; Okumura, M.; Tsubota, S.; Haruta, M., Vital role of moisture in the catalytic activity of supported gold nanoparticles. Angew. Chem., Int. Ed. 2004, 43 (16), 2129-2132. 42. Saavedra, J.; Doan, H. A.; Pursell, C. J.; Grabow, L. C.; Chandler, B. D., The critical role of water at the gold-titania interface in catalytic CO oxidation. Science 2014, 345 (6204), 15991602. 43. Costello, C. K.; Yang, J. H.; Law, H. Y.; Wang, Y.; Lin, J. N.; Marks, L. D.; Kung, M. C.; Kung, H. H., On the potential role of hydroxyl groups in CO oxidation over Au/Al2O3. Appl. Catal., A 2003, 243 (1), 15-24. 44. Costello, C. K.; Kung, M. C.; Oh, H. S.; Wang, Y.; Kung, H. H., Nature of the active site for CO oxidation on highly active Au/γ-Al2O3. Appl. Catal., A 2002, 232 (1-2), 159-168. 45. Calla, J. T.; Davis, R. J., Oxygen-exchange reactions during CO oxidation over titaniaand alumina-supported Au nanoparticles. J. Catal. 2006, 241 (2), 407-416. 46. Calla, J. T.; Davis, R. J., Investigation of Alumina-Supported Au Catalyst for CO Oxidation by Isotopic Transient Analysis and X-ray Absorption Spectroscopy. Journal of Physical Chemistry B 2005, 109 (6), 2307-2314. 47. Date, M.; Haruta, M., Moisture Effect on CO Oxidation over Au/TiO2 Catalyst. J. Catal. 2001, 201 (2), 221-224. 31 ACS Paragon Plus Environment

Journal of the American Chemical Society 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 32 of 36

48. Hibbitts, D.; Iglesia, E., Prevalence of Bimolecular Routes in the Activation of Diatomic Molecules with Strong Chemical Bonds (O2, NO, CO, N2) on Catalytic Surfaces. Acc. Chem. Res. 2015, 48, 1254-1262. 49. Hibbitts, D. D.; Loveless, B. T.; Neurock, M.; Iglesia, E., Mechanistic Role of Water on the Rate and Selectivity of Fischer-Tropsch Synthesis on Ruthenium Catalysts. Angew. Chem., Int. Ed. 2013, 52 (47), 12273-12278. 50. Calla, J. T.; Davis, R. J., Influence of Dihydrogen and Water Vapor on the Kinetics of CO Oxidation over Au/Al2O3. Industrial & Engineering Chemistry Research 2005, 44 (14), 5403-5410. 51. Debeila, M. A.; Wells, R. P. K.; Anderson, J. A., Influence of water and pretreatment conditions on CO oxidation over Au/TiO2-In2O3 catalysts. J. Catal. 2006, 239 (Copyright (C) 2012 American Chemical Society (ACS). All Rights Reserved.), 162-172. 52. Diemant, T.; Bansmann, J.; Behm, R. J., CO oxidation on planar Au/TiO2 model catalysts: Deactivation and the influence of water. Vacuum 2009, 84 (Copyright (C) 2012 American Chemical Society (ACS). All Rights Reserved.), 193-196. 53. Ketchie, W. C.; Murayama, M.; Davis, R. J., Promotional effect of hydroxyl on the aqueous phase oxidation of carbon monoxide and glycerol over supported Au catalysts. Top. Catal. 2007, 44 (1-2), 307-317. 54. Daniells, S. T.; Makkee, M.; Moulijn, J. A., The effect of high-temperature pre-treatment and water on the low temperature CO oxidation with Au/Fe2O3 catalysts. Catal. Lett. 2005, 100 (1-2), 39-47. 55. Bongiorno, A.; Landman, U., Water-Enhanced Catalysis of CO Oxidation on Free and Supported Gold Nanoclusters. Phys. Rev. Lett. 2005, 95 (10), 106102/1-106102/4. 56. Mullen, G. M.; Gong, J.; Yan, T.; Pan, M.; Mullins, C. B., The Effects of Adsorbed Water on Gold Catalysis and Surface Chemistry. Top. Catal. 2013, 56 (15-17), 1499-1511. 57. Yan, T.; Gong, J.; Flaherty, D. W.; Mullins, C. B., The Effect of Adsorbed Water in CO Oxidation on Au/TiO2(110). J. Phys. Chem. C 2011, 115 (5), 2057-2065. 58. Ojifinni, R. A.; Froemming, N. S.; Gong, J.; Pan, M.; Kim, T. S.; White, J. M.; Henkelman, G.; Mullins, C. B., Water-Enhanced Low-Temperature CO Oxidation and Isotope Effects on Atomic Oxygen-Covered Au(111). J. Am. Chem. Soc. 2008, 130 (21), 6801-6812. 59. Fujitani, T.; Nakamura, I.; Haruta, M., Role of Water in CO Oxidation on Gold Catalysts. Catal. Lett. 2014, 144, 1475-1486. 60. Schubert, M. M.; Plzak, V.; Garche, J.; Behm, R. J., Activity, selectivity, and long-term stability of different metal oxide supported gold catalysts for the preferential CO oxidation in H2-rich gas. Catal. Lett. 2001, 76 (3-4), 143-150.

32 ACS Paragon Plus Environment

Page 33 of 36 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

Journal of the American Chemical Society

61. Ivanova, S.; Pitchon, V.; Petit, C.; Caps, V., Support Effects in the Gold-Catalyzed Preferential Oxidation of CO. ChemCatChem 2010, 2 (Copyright (C) 2011 American Chemical Society (ACS). All Rights Reserved.), 556-563. 62. We note that there are essentially no changes to the IR peak area associated with the TiOH groups with the addition or removal of weakly adsorbed water (see reference XX for details). The primary changes are assocated changes to hydrogen-bonding interactions between the TiOH groups and the weakly adsorbed water. 63. Hussain, A.; Gracia, J.; Nieuwenhuys, B. E.; Niemantsverdriet, J. W., Explicit Roles of Au and TiO2 in a Bifunctional Au/TiO2 Catalyst for the Water-Gas Shift Reaction: A DFT Study. ChemCatChem 2013, 5 (8), 2479-2488. 64. Sun, K.; Kohyama, M.; Tanaka, S.; Takeda, S., Reaction Mechanism of the LowTemperature Water-Gas Shift Reaction on Au/TiO2 Catalysts. J. Phys. Chem. C 2017, 121 (22), 12178-12187. 65. Carrasco, J.; Klimeš, J.; Michaelides, A., The role of van der Waals forces in water adsorption on metals. The Journal of Chemical Physics 2013, 138 (2), 024708. 66. Pan, M.; Pozun, Z. D.; Yu, W.-Y.; Henkelman, G.; Mullins, C. B., Structure Revealing H/D Exchange with Co-Adsorbed Hydrogen and Water on Gold. J. Phys. Chem. Lett. 2012, 3 (14), 1894-1899. 67. Corem, G.; Kole, P. R.; Zhu, J.; Kravchuk, T.; Manson, J. R.; Alexandrowicz, G., Ordered H2O Structures on a Weakly Interacting Surface: A Helium Diffraction Study of H2O/Au(111). J. Phys. Chem. C 2013, 117 (45), 23657-23663. 68. Velasco-Velez, J.-J.; Wu, C. H.; Pascal, T. A.; Wan, L. F.; Guo, J.; Prendergast, D.; Salmeron, M., The structure of interfacial water on gold electrodes studied by x-ray absorption spectroscopy. Science (Washington, DC, U. S.) 2014, 346 (6211), 831-834. 69. Stacchiola, D.; Park, J. B.; Liu, P.; Ma, S.; Yang, F.; Starr, D. E.; Muller, E.; Sutter, P.; Hrbek, J., Water Nucleation on Gold: Existence of a Unique Double Bilayer. J. Phys. Chem. C 2009, 113 (34), 15102-15105. 70. Ikemiya, N.; Gewirth, A. A., Initial Stages of Water Adsorption on Au Surfaces. J. Am. Chem. Soc. 1997, 119 (41), 9919-9920. 71. Merte, L. R.; Bechstein, R.; Peng, G.; Rieboldt, F.; Farberow, C. A.; Zeuthen, H.; Knudsen, J.; Lægsgaard, E.; Wendt, S.; Mavrikakis, M.; Besenbacher, F., Water clustering on nanostructured iron oxide films. 2014, 5, 4193. 72. Green, I. X.; Tang, W.; Neurock, M.; Yates, J. T., Jr., Spectroscopic Observation of Dual Catalytic Sites During Oxidation of CO on a Au/TiO2 Catalyst. Science 2011, 333, 736-739. 73. Green, I. X.; Tang, W.; Neurock, M.; Yates, J. T., Jr., Insights into Catalytic Oxidation at the Au/TiO2 Dual Perimeter Sites. Acc. Chem. Res. 2014, 47 (3), 805-815. 33 ACS Paragon Plus Environment

Journal of the American Chemical Society 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 34 of 36

74. Tost, A.; Widmann, D.; Behm, R. J., Stable active oxygen on mesoporous Au/TiO2 supported catalysts and its correlation with the CO oxidation activity. J. Catal. 2009, 266 (Copyright (C) 2011 American Chemical Society (ACS). All Rights Reserved.), 299-307. 75. Saavedra, J.; Powell, C.; Panthi, B.; Pursell, C. J.; Chandler, B. D., CO oxidation over Au/TiO2 catalyst: Pretreatment effects, catalyst deactivation, and carbonates production. J. Catal. 2013, 307, 37-47. 76. Chandler, B. D.; Kendell, S.; Doan, H.; Korkosz, R.; Grabow, L. C.; Pursell, C. J., NaBr Poisoning of Au/TiO2 Catalysts: Effects on Kinetics, Poisoning Mechanism, and Estimation of the Number of Catalytic Active Sites. ACS Catalysis 2012, 2 (4), 684-694. 77. Long, C. G.; Gilbertson, J. D.; Vijayaraghavan, G.; Stevenson, K. J.; Pursell, C. J.; Chandler, B. D., Kinetic Evaluation of Highly Active Supported Gold Catalysts Prepared from Monolayer Protected Clusters: An Experimental Michaelis-Menten Approach for Determining the Oxygen Binding Constant During CO Oxidation Catalysis J. Am. Chem. Soc. 2008, 130, 10103-10115. 78. Chandler, B. D.; Long, C. G.; Gilbertson, J. D.; Vijayaraghavan, G.; Stevenson, K. J.; Pursell, C. J., Improving Oxygen Activation Over Supported Au Catalysts Through the Controlled Preparation of Bimetallic Ni-Au Nanoparticles. The Journal of Physical Chemistry C 2010, 114, 1149811508. 79. Pursell, C. J.; Chandler, B. D.; Manzoli, M.; Boccuzzi, F., CO Adsorption on Supported Gold Nanoparticle Catalysts: Application of the Temkin Model. The Journal of Physical Chemistry C 2012, 116 (20), 11117-11125. 80. Pursell, C. J.; Hartshorn, H.; Ward, T.; Chandler, B. D.; Boccuzzi, F., Application of the Temkin Model to the Adsorption of CO on Gold. J. Phys. Chem. C 2011, 115 (48), 2388023892. 81. Hartshorn, H.; Pursell, C. J.; Chandler, B. D., Adsorption of CO on Supported Gold Nanoparticle Catalysts: A Comparative Study. J. Phys. Chem. C 2009, 113, 10718-10725. 82. Under our experimental conditions, CO and O2 terms (~0.96 and ~0.8, respectively) are fairly close to one, and are constant for our determinations of KwH2O. We also point out that, even if the fractions were not close to one, they would represent a common constant in the two terms of Equation 20. Since KwH2O is determined from the ratio of the these two terms, that constant factors out and does not affect the determined value for KwH2O. 83. Ntho, T. A.; Anderson, J. A.; Scurrell, M. S., CO oxidation over titanate nanotube supported Au: Deactivation due to bicarbonate. J. Catal. 2009, 261 (1), 94-100. 84. Azar, M.; Caps, V.; Morfin, F.; Rousset, J.-L.; Piednoir, A.; Bertolini, J.-C.; Piccolo, L., Insights into activation, deactivation and hydrogen-induced promotion of a Au/TiO2 reference catalyst in CO oxidation. J. Catal. 2006, 239 (2), 307-312.

34 ACS Paragon Plus Environment

Page 35 of 36 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

Journal of the American Chemical Society

85. Denkwitz, Y.; Schumacher, B.; Kucerova, G.; Behm, R. J., Activity, stability, and deactivation behavior of supported Au/TiO2 catalysts in the CO oxidation and preferential CO oxidation reaction at elevated temperatures. J. Catal. 2009, 267 (1), 78-88. 86. Denkwitz, Y.; Makosch, M.; Geserick, J.; Hoermann, U.; Selve, S.; Kaiser, U.; Huesing, N.; Behm, R. J., Influence of the crystalline phase and surface area of the TiO2 support on the CO oxidation activity of mesoporous Au/TiO2 catalysts. Appl. Catal., B 2009, 91 (1-2), 470-480. 87. Denkwitz, Y.; Zhao, Z.; Hoermann, U.; Kaiser, U.; Plzak, V.; Behm, R. J., Stability and deactivation of unconditioned Au/TiO2 catalysts during CO oxidation in a near-stoichiometric and O2-rich reaction atmosphere. J. Catal. 2007, 251 (Copyright (C) 2012 American Chemical Society (ACS). All Rights Reserved.), 363-373. 88. Schumacher, B.; Denkwitz, Y.; Plzak, V.; Kinne, M.; Behm, R. J., Kinetics, mechanism, and the influence of H2 on the CO oxidation reaction on a Au/TiO2 catalyst. J. Catal. 2004, 224 (2), 449-462. 89. Korkosz, R. J.; Gilbertson, J. D.; Prasifka, K. S.; Chandler, B. D., Dendrimer Templates for Supported Au Catalysts. Catalysis Today 2007, 122, 370-377. 90. Singh, A.; Chandler, B. D., Mild Thermolysis Conditions for the Activation of Dendrimer Encapsulated Pt Nanoparticles. Langmuir 2005, 21, 10776-10782. 91. Lang, H.; Maldonado, S.; Stevenson, K. J.; Chandler, B. D., Synthesis and Characterization of Dendrimer Templated Supported Bimetallic Pt-Au Nanoparticles. Journal of the American Chemical Society 2004, 126, 12949-12956. 92. Lang, H.; May, R. A.; Iversen, B. L.; Chandler, B. D., Dendrimer Encapsulated Nanoparticle Precursors to Supported Platinum Catalysts. Journal of the American Chemical Society 2003, 125, 14832-14836.

TOC Graphic

35 ACS Paragon Plus Environment

Journal of the American Chemical Society 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 36 of 36

36 ACS Paragon Plus Environment