Water Poisons H2 Activation at the Au-TiO2 Interface by Slowing Proton and Electron Transfer between Au and Titania

Article abstract of DOI:10.1021/jacs.9b13729

Understanding the dynamic changes at the active site during catalysis is a fundamental challenge that promises to improve catalytic properties. While performing Arrhenius studies during H₂ oxidation over Au/TiO₂ catalysts, we found different apparent activation energies (E_app) depending on the feedwater pressure. This is partially attributed to changing numbers of metal-support interface (MSI) sites as water coverage changes with temperature. Constant water coverage studies showed two kinetic regimes: fast heterolytic H₂ activation directly at the MSI (E_app ～25 kJ/mol) and significantly slower heterolytic H₂ activation mediated by water (E_app ～45 kJ/mol). The two regimes had significantly different kinetics, suggesting a complicated mechanism of water poisoning. Density functional theory (DFT) showed water has minor effects on the reaction thermodynamics, primarily attributable to intrinsic differences in surface reactivity of different Au sites in the DFT model. The DFT model suggested significant surface restructuring of the TiO₂ support during heterolytic H₂ adsorption; evidence for this phenomenon was observed during in situ infrared spectroscopy experiments. A monolayer of water on the hydroxylated TiO₂ surface increased the H₂ dissociation activation barrier by ～0.2 eV, in good agreement the difference in experimentally measured values. DFT calculations suggested H₂ activation goes through a proton-coupled electron-transfer-like mechanism. During proton transfer to a basic support hydroxyl group, electron density is distributed through the gold nanorod and partially localized on the protonated support hydroxyl group. Water slows H₂ activation by slowing this H⁺ transfer, forcing negative charge buildup on the Au and increasing the transition state energy.

Full text of DOI:10.1021/jacs.9b13729

Subscriber access provided by Imperial College London | Library
Article
Water Poisons H2 Activation at the Au-TiO2 Interface by
Slowing Proton and Electron Transfer between Au and Titania
Sravan Kumar Kanchari Bavajigari, Todd Whittaker, Christine Peterson, Lars C. Grabow, and Bert D Chandler
J. Am. Chem. Soc., Just Accepted Manuscript • DOI: 10.1021/jacs.9b13729 • Publication Date (Web): 21 Feb 2020
Downloaded from pubs.acs.org on February 22, 2020
Just Accepted
“Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted
online prior to technical editing, formatting for publication and author proofing. The American Chemical
Society provides “Just Accepted” as a service to the research community to expedite the dissemination
of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in
full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully
peer reviewed, but should not be considered the official version of record. They are citable by the
Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore,
the “Just Accepted” Web site may not include all articles that will be published in the journal. After
a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web
site and published as an ASAP article. Note that technical editing may introduce minor changes
to the manuscript text and/or graphics which could affect content, and all legal disclaimers and
ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or
consequences arising from the use of information contained in these “Just Accepted” manuscripts.
is published by the American Chemical Society. 1155 Sixteenth Street N.W.,
Washington, DC 20036
Published by American Chemical Society. Copyright © American Chemical Society.
However, no copyright claim is made to original U.S. Government works, or works
produced by employees of any Commonwealth realm Crown government in the course
of their duties.

Page 1 of 30
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
Water Poisons H
2
Activation at the Au-TiO Interface by Slowing
2
Proton and Electron Transfer between Au and Titania
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
K. B. Sravan Kumar,A,B, † Todd Whittaker,A. † Christine Peterson,A
B
*,A
Lars C. Grabow, and Bert D. Chandler,
ADepartment of Chemistry, Trinity University, San Antonio, TX 78212-7200
BDepartment of Chemical and Biomolecular Engineering, University of Houston, Houston, TX
7
7204-4004
†These authors contributed equally to the work
*To whom correspondence should be addressed:
Bert.chandler@trinity.edu (210) 999-7557 phone; (210) 999-7569 fax
1
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 2 of 30
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
Abstract:
Understanding the dynamic changes at the active site during catalysis is a fundamental challenge
that promises to improve catalytic properties. While performing Arrhenius studies during H
oxidation over Au/TiO catalysts, we found different apparent activation energies (Eapp) depending
2
2
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
on the feed water pressure. This is partially attributed to changing numbers of metal-support
interface (MSI) sites as water coverage changes with temperature. Constant water coverage
studies showed two kinetic regimes: fast heterolytic H
2
activation directly at the MSI (Eapp ~25
kJ/mol) and significantly slower heterolytic H activation mediated by water (Eapp ~45 kJ/mol).
2
The two regimes had significantly different kinetics, suggesting a complicated mechanism of water
poisoning. Density functional theory (DFT) showed water has minor effects on the reaction
thermodynamics, primarily attributable to intrinsic differences in surface reactivity of different Au
sites in the DFT model. The DFT model suggested significant surface restructuring of the TiO
support during heterolytic H adsorption; evidence for this phenomenon was observed during in-
situ infrared spectroscopy experiments. A monolayer of water on the hydroxylated TiO surface
increased the H dissociation activation barrier by ~0.2 eV, in good agreement the difference in
experimentally measured values. DFT calculations suggested H activation goes through a proton-
2
2
2
2
2
coupled electron transfer-like mechanism. During proton transfer to a basic support hydroxyl
group, electron density is distributed through the gold nanorod and partially localized on the
+
protonated support hydroxyl group. Water slows H
2
activation by slowing this H transfer, forcing
negative charge build-up on the Au, and increasing the transition state energy.
2
ACS Paragon Plus Environment

Page 3 of 30
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
Introduction:
Supported metal catalysts, typically consisting of a metal nanoparticle immobilized on an
1
oxide support, are mainstays of the petrochemical and environmental remediation industries. The
interface between the metal nanoparticle and the oxide support is often considered the active site
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
hydrogenation,2-5 methanol synthesis,6
for many industrially important reactions, including CO
2
biomass upgrading,7 CO oxidation,8-10 the water-gas shift reaction,11-12 nitroarene
hydrogenation,13-14 and electrocatalytic energy conversion.15-16 In spite of the commercial
importance of these materials, developing an understanding of the nature of the catalytic active
site, and the factors that influence catalyst activity at the metal-support interface, remain significant
challenges.
The interplay between metal, support, and adsorbates is therefore an area of widespread
interest; our understanding of the complex chemistries involved continues to evolve.17-18 Charge
transfer between the two components may occur in either direction depending on the metal, the
reducibility of the support, the exposed metal facets, and the presence of defects.19-26 Reaction
1
0
9,
intermediates can bridge the two phases, and reactions can occur between species on each phase.
27
Reactive species can also be transferred from one phase to the other, as is the case of hydrogen
spillover.28-29 Finally, active sites at or near the interface can dynamically respond to the reaction
environment,30-31 experience poisoning or blocking by strongly adsorbing species, or deactivate
over extended periods of time.32-33
We are working to better understand the complexity associated with reactivity at the metal-
support interface (MSI) using the relatively simple, yet industrially important hydrogen oxidation
reaction over supported Au catalysts. More than 95% of industrial H is generated through
2
hydrocarbon steam reforming and water-gas shift reactions, which produces a ~1% CO impurity.
The preferential oxidation of CO (PrOx) is a potential means of removing this CO cheaply and
cleanly, but the reaction places large performance demands on the catalyst.34-37 Au nanoparticles
are exceptional CO oxidation catalysts8, 38-39 and have excellent selectivity in PrOx when an
O is present.40
optimal amount of physisorbed H
2
Water can have a significant effect on catalytic performance, particularly in selective
4
1
oxidation reactions, where the role of water varies. Ertl and coworkers concluded that water
preferentially adsorbs at coordinatively unsaturated (cus) Ru sites, poisoning CO oxidation over
RuO
2
catalysts.42 Conversely, Haruta et. al.,43 as well as others,8, 38, 44-45 demonstrated water
3
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 4 of 30
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
promotion of CO oxidation catalysis over supported Au at low pressures. Both the Behm and
Mullins groups have shown that water can change the reaction mechanism of CO oxidation on
Au/TiO
origins of this effect are not well understood, and appear to vary from system to system.
We recently showed water poisons H activation over Au, which takes place at the MSI via
dissociation. This finding was surprising given that most metals bind H through
2
.46-47 While it is clear that adsorbed water can profoundly affect catalytic systems, the
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
2
4
8
heterolytic H
2
2
dissociative chemisorption. Heterolytic activation of non-acidic bonds such as H-H and C-H bonds
has been claimed in some solid systems, but these systems typically require clever modifications
to yield strong Frustrated Lewis Pairs (FLP’s).49-58
The goals of this study are to elucidate the origins of the water poisoning effects and to
understand the driving force behind the heterolytic H
conclusion that water blocks the most reactive MSI sites and more deeply probe the details of
heterolytic H activation at the MSI. We find that the poisoning, and indeed the nature of the
active site, is far more dynamic than a simple “site-blocking” mechanism suggests. Additionally,
we show the Au/TiO catalysts have an intimate electronic interaction that allows for the rapid
2
activation. In particular, we examine the
2
2
exchange of both protons and electrons between the two components. This exchange is also tied
to structural changes on the support surface. Thus, water poisons the reaction not simply by
blocking the active site, but by interfering in the ability of the MSI to distribute these charges in
the H activation transition state.
2
Results and Discussion:
Constant PH O Arrhenius studies. We initially concluded the loss of activity with increased
2
water coverage resulted from losing the most active H activation sites at the metal-support
2
4
8
interface. Arrhenius studies carried out with 9 and 19 Torr water in the feed yielded two different
apparent activation barriers: 50 ± 3 kJ/mol and 76 ± 12 kJ/mol, respectively (see Supporting
Information). The feed water content should not affect a “true” activation barrier; further, these
values are considerably larger than the Eapp values Haruta and coworkers reported for H
2
– D
(36 kJ/mol). Given that both reactions occur at the MSI, and both are
activation limited, the measured apparent activation energies ought to be
2
59
equilibration over Au/TiO
2
considered to be H
2
reasonably similar. We also found the reaction kinetics varied substantially with the feed PH2O
4
ACS Paragon Plus Environment

Page 5 of 30
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
(
vida infra). Collectively, these data indicate the water-poisoning chemistry is more complex and
more interesting than simple “site blocking” of the most active H
Calculated elementary step energetics for H oxidation. We previously showed
homolytic activation of H on Au-Au sites is thermodynamically and kinetically less favorable
2
dissociation sites.
2
2
4
8
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
than is heterolytic activation at the MSI. The homolytic pathway is unlikely to be affected by the
water layer, beyond water’s ability to cover the more reactive surfaces. We therefore focus our
analysis on the (faster) heterolytic H
2
activation at Au MSI sites and Au metal-water interface
(MWI) sites.
Scheme 1. Schematic representation of the
Au/TiO
metal-support interface and metal-water
interface sites. The Au*(111)-HOcus and
2
computational model showing the
Au**(111)-HOcus sites are indicated on the left
side and right sides of the top portion of the
scheme, respectively. The metal-water interface
is described by the Au(111)-H
2
O and Au(211)-
H
2
O sites on the left side and right sides of the
bottom portion of the scheme. Further details on
the computational model can be found in the
Experimental section and the SI.
The various reaction sites in the computational model are described briefly here; full details
are available in the Supporting Information. A TiO (110) surface models the support and gold is
2
modeled as a nanorod with different facets. As Scheme 1 shows, the left side of the nanorod
truncates in the Au(111) surface, while the right side truncates in a Au(211) step edge; Miller
indices associated with individual surface atoms are also included. The TiO (110) surface was
2
fully hydroxylated with water, leading to two chemically distinct surface hydroxyl groups in close
5
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 6 of 30
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
proximity to the Au nanorod: (i) hydroxyl groups bound to coordinatively unsaturated Ti sites
(HOcus) and (ii) hydroxyl groups resulting from protonation of bridging O atoms (HObr). The
former are Brønsted bases; the latter have Brønsted acidity. Both Ti-OH sites alternate along the
length of the Au nanorod such that both hydroxyls are accessible to adsorbates on the Au.
Our model is designed to describe reactivity at the MSI in the presence and absence of
water, where metal, support, and Brønsted acid-base chemistry can all be explored. Each potential
reaction site is chemically distinct. Reactivity on the Au nanorod, away from the MSI is described
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
simply with the atoms involved (e.g. dissociative H chemisorption on Au(211)-Au(211) sites is
2
4
8
faster than on Au(111)-Au(111) sites). Both sides of the model have Au(111) atoms interacting
with the support, but the Au(111) structure is truncated and replaced by interactions with the
support O atoms. We define the combination of metal and support at this site as “Au*(111)-
HOcus”. The “*” indicates direct interaction with the support (and consequent lower Au
coordination); the “HOcus” indicates the close proximity of the HOcus group. The presence of the
step edge imposes a slightly different MSI coordination environment on the right side of the model.
We designate this site as “Au**(111)-HOcus” to distinguish the two similar MSI sites.
A monolayer of water adsorbed on the support (Scheme 1, bottom) physically blocks
adsorbates from interacting with the Au*(111)-HOcus and Au**(111)-HOcus sites. Under these
conditions, we describe the catalyst surface on the left side of the model as “Au(111)-H O”. This
2
indicates proximity between the available 9-coordinate Au(111) site and water adsorbed on the
support (not on Au). Similarly, the site defined by water adsorbed on the support and the Au(211)
step-edge is described as the “Au(211)-H
Our previous work indicated the fastest H
Au-H deprotonation followed by proton addition to various Au-O species (Au-OOH, Au-H
2
O” site.
2
oxidation pathway involves a combination of
2
O ,
2
4
8
Au-O, and Au-OH). We first address these O
2
activation and subsequent reaction steps, many of
which involve proton transfers that might be impacted by an adsorbed water layer. The energetics
of these elementary steps at the Au**(111)-HOcus sites (absence of water) and at the Au(211)-H
2
O
sites (presence of water) are presented in Table 1.
With one exception, the calculated barriers for proton transfer steps at MWI sites are within
.15 eV of the barriers at MSI sites. The only significant difference is the O activation barrier
0
2
near HObr sites, which is considerably higher in the presence of water. However, water mediates
+
proton transfer from acidic HObr to basic HOcus sites, and Au-OOH formation using H
2
Ocus sites
6
ACS Paragon Plus Environment

Page 7 of 30
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
is fast. Thus, the overall proton mediated O
2
activation pathway is largely unchanged by the
addition of water. Two proton transfer steps (Au-H deprotonation and Au-H
2
2
O protonation) have
modest barriers that affect the coverages of the adsorbed species, but neither step is significantly
impacted by the presence of water.
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Table 1 Elementary step energetics for H
with water. Elementary steps at the MSI were published previously.
Key H Oxidation Elementary Steps with and without H
2
oxidation at the MSI without water and at the MWI
48
2
2
O at the interface
E
a
ΔE
Entry
Reaction site/Reactant
Reaction
(
eV) (eV)
O
2
Activation
2
+ -
-
ES1
ES2
ES3
Au*(111)-HObr (MSI)
Au + O
Au + O
Au + O
2
+ HObr → Au - OOH + Obr
0.05 -0.49
0.48 -0.39
0.19 -0.52
2
+ -
-
Au(211)-H
2
O-HObr (MWI)
2
+ HObr → Au - OOH + Obr
+
+
2+ -
2
Au(211)-H O- H
2
Ocus (MWI)
2
+ H
2
O
cus → Au - OOH + HOcus
Au-OOH Protonation
+
2+ -
+
2+
ES4
ES5
Au*(111)- H
2
O
cus (MSI)
Au - OOH + H
2
O
cus → Au -H
2
O
+ HOcus 0.04 -0.27
2
+
2+ -
+
+
2+
Au(211)-H
2
O- H
2
O
cus (MWI) Au - OOH H
2
O
cus → Au -H
2
O
2
+ HOcus
0.19 -0.55
Au-H O
2 2
Protonation
2+
+
Au -H
2
O
2
+ Au + H
2
O
2
cus
→
+
ES6
Au*(111)- H
2
Ocus (MSI)
0.41 -1.81
0.37 -2.11
+
2+
Au -OH + Au -H O + HOcus
2
+
+
Au -H
2
O
2
+ Au + H
2
O
cus
→
+
ES7
Au(211)-H
2
O- H
2
Ocus (MWI)
+
2+
Au -OH + Au -H O + HOcus
2
Au-OH Protonation
+
+
+
2+
ES8
ES9
Au*(111)- H
Au(211)-H
2
O
cus (MSI)
cus (MWI)
Au -OH + H
2
O
cus → Au -H
2
O + HOcus
O + HOcus
0.00 -1.17
0.13 -0.70
+
+
+
2+
2
O- H
2
O
Au -OH + H
2
O
cus → Au -H
2
Au-O Protonation
ES10
+H
O
cus (MSI)
+
+
0.16 -1.04
0.02 -0.63
2
Au-O + H
Au-O + H
2
O
O
cus → Au -OH + HOcus
+
+
+
ES11
Au(211)-H
2
O- H
2
Ocus (MWI)
2
cus → Au -OH + HOcus
Au-H Deprotonation
-
2-
+
ES12
ES13
Au*(111)-HOcus (MSI)
Au(211)-H O-HOcus (MWI)
Au-H + HOcus → Au + H
Au-H + HOcus → Au + H
2
O
O
cus
0.39 -0.73
0.55 -0.71
-
2-
+
2
2
cus
Proton Exchange
-
+
ES14
ES15
HOcus & HObr (MSI)
HOcus & HObr (MWI)
HObr + HOcus → Obr + H
2
O
O
cus
0.49
0.44
0.08
0.00
-
+
HObr + HOcus → Obr + H
2
cus
Since the added water layer has a relatively small impact on the overall proton transfer and
activation kinetics, we conclude that the reaction likely proceeds through similar pathways in
O
2
the presence and absence of exogenous water. Additionally, nearly all proton transfer reactions
have barriers lower than 0.2 eV and can be considered quasi-equilibrated under typical reaction
conditions. Most importantly, the presence of the water layer has generally small effects on the
calculated thermodynamics and kinetics of these steps. While the reaction is conveniently
7
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 8 of 30
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
described as limited by hydrogen activation, this is an over simplification. Our previous kinetics
studies indicate that the reaction is H coverage limited rather than strictly limited by H
2
activation.48 However H
activation, has the largest kinetic barrier and is most significantly
2
impacted by the presence of water. The analysis below is not intended to describe the entire
reaction network, but focuses on the details of the key first step in the mechanism.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Table 2. DFT reaction energies (∆E) and activation energies (E
on Au/TiO
a
) for H activation on various sites
2
2
.
H
2
activation
mode
Reaction Site
∆E (eV)
E (eV) Reference
a
*
Ref 48
Au (111)-HOcus
Heterolytic
Heterolytic
Heterolytic
Heterolytic
Homolytic
Homolytic
-0.03
0.08
0.05
0.29
1.07
0.59
0.70
0.81
0.99
1.25
1.58
1.16
**
Au (111)-HOcus
this work
this work
this work
this work
Ref 48
Au(211)-H
Au(111)-H
2
O
O
2
Au(111)-Au(111)
Au(211)-Au(211)
-
+
Au(111)-H + H O
cus
0
.29 eV
2
-
+
Au(111)-H + H O
3
cus
0.20 eV
-
+
Au(211)-H + H O
3
cus
-
+
Au(211)-H + H O
2
cus
0.05 eV
Initial State
0.02 eV
-0.03 eV
*
-
+
Au (111)-H + H O
2
cus
-0.03 eV
*
Figure 1. Final states for H
Au(111)-H
side of the figure shows the calculated energies for moving the Au-H from the metal-support
2
adsorption at Au (111)-OHcus (red), Au(211)-H
2
O (left, green), and
2
O (left, blue). The horizontal lines show the calculated reaction energies. The right
*
-
-
-
interface site (Au (111)-H , red) to the Au(211)-H (green) and Au(111)-H (blue) sites.
8
ACS Paragon Plus Environment

Page 9 of 30
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
H
2
activation thermodynamics. While there may be small differences in the Au coverage
of various species, the primary difference between the dry and wet feeds is the water coverage on
TiO2, and therefore surface hydroxyl availability. Hydrogen adsorption reaction energies and
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
associated activation barriers are summarized in Table 2. The left side of Figure 1 shows the final
*
states and their relative energies for H
Au(211)-H
proton transferred to the support HOcus (Figure 1). This is essentially the same final state as for
activation directly at the MSI, and is a consequence of the HOcus being the most basic hydroxyl
group in the system.
The three Au-H adsorption sites are located at chemically distinct Au atoms. To understand
water’s influence on H adsorption, we first determined the impact of the Au surface atom
coordination environment by evaluating how Au-H stability affects H adsorption energy. Using
2
adsorption at the Au (111)-HOcus, Au(111)-H
2
O, and
2
O sites. When the water layer is present, the calculations show the final state has a
H
2
2
2
*
*
-
+
the final state of H
2
adsorption at the Au (111)-HOcus site (i.e. Au (111)H + H Ocus) as the
2
reference state, the hydride was moved to the other two calculated positions and the system energy
was calculated. The final states and calculated energies are shown on the right side of Figure 1.
The energetic differences associated with moving the Au-H closely mirror the overall energy
differences in H
in Au-H stability. We therefore conclude the added water layer, in spite of mediating the proton
transfer at the Au(111)-H O and Au(211)-H O sites, has almost no effect on the H adsorption
thermodynamics. The water layer must therefore have large effects on the kinetics and dynamics
of H adsorption, which we explore further below.
Experimental determination of H oxidation kinetic parameters at constant θH2O.
2
adsorption, indicating these differences are almost entirely due to the differences
2
2
2
2
2
Although the measured Eapp values extracted from constant PH2O data are determined in a manner
consistent with the literature, they miss a key component to this system: under constant water
pressure, the support water coverage changes significantly with temperature. Since water blocks
the fast H
2
activation sites at the MSI, changing water coverage effectively changes the number of
active sites on the catalyst. Over this relatively small temperature range, water coverage (θH2O
)
can change by more than 30%. This significantly impacts the number of available MSI sites. For
example, the nominal fraction of exposed hydroxyl groups triples when θH2O changes from 0.9
MLE (10% exposed hydroxyls) to 0.7 MLE (30% exposed hydroxyls). We note that this example
9
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 10 of 30
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
is illustrative; it is difficult to know the true number of available active hydroxyl groups at the MSI
under any given condition.
We therefore carried out experiments at constant water coverage to ensure the accuracy of
the experimental measurements. Judicious choice of reaction conditions, careful control of H
2
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
conversion, and long reaction equilibration times were required to ensure steady-state activity; this
allowed us to maintain consistent θH2O within reasonable experimental errors. The reaction
requirements limited our sub-monolayer measurements to between 0.8-0.9 MLE of water; under
these conditions, we believe H activation occurs at the MSI with surface Ti-OH groups.
2
Table 3. H
2
oxidation reaction orders at θH2O < 1 (reference 48) and θH2O > 1 (from this work).
θ
H O < 1a
θ
H O > 1b
Reactant/Species
2
2
O
H
2
0.20 ± 0.07
0.64 ± 0.05
-0.64 ± 0.02
-1.41 ± 0.06
0.16 ± 0.05
1.8 ± 0.2
-1.21 ± 0.07
2
H
2
O (w.r.t pressure)
O (w.r.t coverage)
H
2
-2.8 ± 0.2
a
3
Reference 46: P
H
2: 23-456 Torr., P
O
2: 7.6-76 Torr., PH2O: 5-18 Torr. Trxn = 60 °C. WHSV: 2.3 x 10 L/gcat/hr.
bThis Work: P
H
2: 243-304 Torr., P
O
2: 7.6-23 Torr., PH2O: 9-18 Torr. Trxn = 40 °C. WHSV: 86 L/gcat/hr.
Kinetic data collected at low (θH2O < 1 MLE) and high (θH2O > 1 MLE) θH2O are reported
in Table 3 and incorporated into observed rate laws (Equations 1 and 2). The O reaction orders
are within error of each other, indicating water has no significant effect on O activation. This is
consistent with our observations for CO oxidation,8, 38, 40 and with the elementary step analysis
reported in Table 1. The changes in H and H O reaction orders were surprisingly large. For θH2O
1 MLE, water inhibition is far more pronounced (i.e. larger negative reaction order) than for
MLE water coverages less than unity. The concomitant large increase in the H reaction order
indicates the loss in activity is associated with large changes in hydrogen activation kinetics.
2
2
2
2
>
2
0
.7 0.2
ꢅ
ꢅ
ꢆ
ꢇ
ꢀꢁꢂꢃ = ꢄ′
2
2
(1)
(2)
For θH2O < 1:
For θH2O > 1:
ꢁꢂꢃ
1.4
2
ꢈ
ꢆ
ꢇ
1.8 0.2
ꢅ
ꢅ
ꢆ
ꢇ
ꢀꢁꢂꢃ = ꢄ′
2
2
ꢁꢂꢃ
2.8
2
ꢈ
ꢆ
ꢇ
10
ACS Paragon Plus Environment

Page 11 of 30
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
Arrhenius studies were carried out by adjusting the feed water pressure at each reaction
temperature to ensure constant θH2O at different reaction temperatures within reasonable
′
ꢁ
experimental errors. Equations 1 and 2 were used to determine ꢄ
values and normalize the
ꢂꢃ
small differences θH2O, which are presented alongside each data point in Figure 2A. Direct plots
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
′
ꢁ
of the ꢄ values are in the Supporting Information.
ꢂꢃ
Figure 2. (A) Arrhenius studies at constant
water coverage of 0.86 ± 5% MLE (blue) and
1
.26 ± 2% MLE (green). Each data point is
labeled with the estimated water coverage
under the conditions of the experiment. (B)
Hydrogen evolution energetics from proton-
hydride pairs; reaction energies were
referenced to the initial state of the proton-
hydride pair.
We also re-examined the 9 Torr constant pressure data, in which θH2O was always less than
unity, using Equation 1 (see Supporting Information). This yielded an Eapp value of 26 ± 3 kJ/mol,
in good agreement with the constant coverage data (31 ± 3 kJ/mol). We therefore consider the
value of 25-30 kJ/mol as our experimental measure of the apparent activation energy for H
2
2
oxidation at Au-MSI sites. This value is slightly lower than Haruta’s determination for H
5
9
activation (36 kJ/mol) from H
2
– D equilibration kinetics, but the two agree reasonably well. The
2
Arrhenius data for θH2O = 1.25 (45 ± 2 kJ/mol), is significantly larger. These Eapp values (25-30 ±
11
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 12 of 30
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
3
kJ/mol for θH2O < 1 and 45 ± 2 kJ/mol for θH2O > 1) clearly show the two different kinetic regimes
predicted in the DFT calculations. While the experimental values describe the overall reaction and
are therefore imperfect models for H activation, these values represent our best experimental Eapp
2
measurements for comparison with the computational model.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
The principle behind Arrhenius experiments is to measure the rate constant dependence on
temperature; however, care must be taken to correlate Arrhenius data with physically meaningful
activation barriers. As shown above, dramatically different Eapp values result from constant water
pressure and constant water coverage experiments. Standard Arrhenius studies using constant feed
composition involve the intrinsic assumption that coverages remain relatively constant over the
temperature range studied.
For many systems, this is a reasonable assumption; however, the concentration of active
sites can change when active sites are dynamically created, modified, or destroyed in rapid
response to changes in the reaction environment (e.g., temperature or partial pressure). For
interfacial reactivity, adsorbates on both the metal and support must be considered. The kinetically
important water is weakly adsorbed to the support; as these data show, coverage can change
significantly over a relatively small temperature range. Since water masks the support hydroxyl
groups that participate in (faster) direct proton transfer from H , changes to the water coverage
2
impact the number of available surface hydroxyl groups, and thus the reaction kinetics. In such
circumstances, coverage must be controlled to keep the catalyst surface in a constant functional
state (i.e. maintain a relatively constant number of available support hydroxyl groups) in order to
make physically meaningful conclusions from Arrhenius data.
Comparing DFT and Experiments. The absolute values of the experimental Eapp values,
which describe the temperature sensitivity of overall reaction rate constants, cannot be directly
compared with DFT barriers calculated for elementary reaction steps. Moreover, the experimental
measurements occur over a range of sites, most of which likely have lower coordination numbers
than the DFT model and are expected to be more reactive. Further, the experimental values are
apparent (not intrinsic) activation energies, and include the complexities of the reaction network.
48
However, the experimental data presented here and published previously collectively suggest the
E
app values are dominated by the H activation step. Qualitative comparisons with DFT
2
calculations focused on H
2
adsorption are therefore appropriate.
12
ACS Paragon Plus Environment

Page 13 of 30
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
Because each surface Au atom in the DFT model is chemically distinct, we cannot directly
compare reactions in the presence and absence of water without accounting for the differences in
surface reactivity. To do this, we examined H
2
evolution from the four MSI and MWI active sites,
evolution and H adsorption must pass
since the principle of microscopic reversibility dictates H
2
2
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
through the same transition state. Referencing all the initial states to the same energy also accounts
for the energy differences associated with the unique coordination environments of the different
Au sites, more clearly identifying the effect of water.
Figure 2B shows this comparison, referencing all of the adsorbed proton-hydride pairs to
the same initial energy. Visualized this way, the effect of water on the transition state energy is
clear: for this DFT model, the calculated activation barrier is ~0.2 eV higher in the presence of
the water layer. This corresponds to ~20 kJ/mol difference between proton transfer directly to a
support hydroxyl group and proton transfer mediated by water; the experimental Eapp values are
15-25 kJ/mol larger when θH2O > 1. This constitutes very good qualitative agreement between
theory and experiment, and imparts greater confidence that conclusions from the DFT calculations
have physically meaningful implications for the experimental system.
Charge Distribution After H
towards understanding the details of how water impacts the H
charge density distributions about the H molecule from final state calculations for H
2
Adsorption. Additional DFT calculations were directed
activation kinetics. Figure 3 shows
adsorption
2
2
2
on three different sites. The final states, which are quite similar in the presence and absence of
water, have essentially a full positive charge on the proton, yet only a small partial negative charge
on the (formal) hydride.
13
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 14 of 30
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
Final states
*
A
Au (111)-HO
B
Au(111)-H O-HO
C
Au(211)-H O-HO
2 cus
cus
2
cus
-0.06 |e-|
-
0.06 |e-|
0.93 |e-|
+0.95 |e-|
+
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
-0.09 |e-|
+0.94 |e-|
∆
E = -0.03 eV
∆E = 0.29 eV
∆E = 0.05 eV
Figure 3. Charge density difference plots for the final states of H
O and (C) the Au(211)-H
charge) shows electron accumulation, and blue shading (positive charge) shows electron depletion.
2
activation across (A) the
*
Au (111)-HOcus site, (B) the Au(111)-H
2
2
O site. Green shading (negative
To identify the distribution of the remaining negative charge, we examined Bader charge
differences (final state – initial state) for all atoms in the system (Figure 4). A negative value
indicates the corresponding atom/atoms develop additional negative charge, while a positive value
indicates positive charge is developed (or negative charge is removed). The positive charge resides
almost exclusively on the proton, which is always closely associated with the HOcus. Although H
2
adsorption adds two electrons to the system, there is surprisingly little negative charge on either
the Au nanorod or the hydride; rather, the compensating negative charge is highly delocalized,
distributed across the hydride, Au, Ti, and remaining support O atoms.
Final State
1.2
1
Total charge
(+) (-)
Au (111)-HOcus 0.94 -0.95
*
Figure 4. Bader charges of (A) final states (FS)
of H activation. Bader charges are referenced
0
0
0
.8
.6
.4
*
*
Au (111)-HOcus 1.04 -1.03
2
-
w.r.t. those of the initial state (IS). H : hydride;
H : proton; Au: all gold atoms; Ti: all titanium
Au(111)-H
Au(211)-H
2
O
O
0.93 -0.93
0.95 -0.95
+
2
0.2
0
atoms; HOcus: cus-hydroxyl abstracting the
proton; H
H-
Au Ti HOcus
other O
2
Ocus: water molecule abstracting the
H+
2 cus
H O
proton; other O: all other oxygen atoms.
-0.2
-0.4
0.6
-
14
ACS Paragon Plus Environment

Page 15 of 30
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
Figure 5 shows the Au(111) side of the model with and without water to further highlight
how H
increase in the Ti-O bond length; essentially, the combination of the proton transfer and negative
charge localization on the HOcus partially pulls the OH group out of the plane of the TiO surface.
2
adsorption affects the Ti-OHcus bond. Proton transfer to the HOcus induces a ~0.4 Å
2
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
This change in the final state geometry is consistent throughout the calculations regardless of the
presence or absence of the water layer, and did not depend significantly on the Au atom(s) involved
in stabilizing the hydride (see SI for further details) This is a substantial surface restructuring, and
is supported by experimental FTIR measurements.
*
A
Au (111)-HO
B
Au(111)-H O
2
cus
w/o H (IS)
w/ H (FS)
w/o H (IS)
w/ H (FS)
2
2
2
2
1
.84 Å
2.29 Å
1.85 Å
2.21 Å
Ti-Ocus bond elongation by 0.45 Å.
Ti-Ocus bond elongation by 0.36 Å.
Figure 5. Variation of bond length of Ti-Ocus upon H
2
activation across (A) the MSI site in (111)
facet and (B) the MWI site in (111) facet.
Infrared spectroscopy during H
was observable in FTIR spectra by exchanging support hydroxyl protons with deuterons.48
Experimental evidence for surface structural changes during CO adsorption on Au/TiO has also
2
adsorption. We previously showed H adsorption on Au
2
2
6
0
been reported. In that study, CO adsorption correlated with an overall loss of light transmission
through the sample, which was observed as a “broad band IR” (BB-IR) absorption and
6
1
quantitatively correlated with the IR absorption band associated with CO adsorbed on Au. The
effect was concluded to arise from decreased light transmittance associated with surface
roughening of the titania, which was hypothesized to be due to electron transfer from Au to the
support adsorption.60
15
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 16 of 30
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Figure 6: (A) FTIR spectra of H
2
adsorption on a D
2
O-exchanged Au/TiO
2
catalyst. Reaction
Conditions: T: 50°C, WHSV = 40 L/gc a- t/h. Temperature dependence of the integrated area under
1
-1
the spectral regions: (B) 1800-2100 cm (C) 3800-3950 cm . (D) Arrhenius fit of initial rates of
different spectral regions during H adsorption on Au/TiO
2
2
.
Experimental evidence for the structural changes induced by H
in infrared spectra collected after deuterated Au/TiO is exposed to H
contain various features including water bending vibrations (1500-1800 cm ), hydroxyl stretching
2
adsorption can be found
2
2
(Figure 6). These spectra
-
1
-1
vibrations (2800-3700 cm ) and the broad-band baseline shift. We monitored the BB-IR signal
-1
-1
by determining the area under the IR spectral curve from 1800 to 2100 cm , and 3800-3950 cm ,
as was previously reported for static CO adsorption measurements.60 Measuring the BB-IR signal
over time at various temperatures (50-70 °C) provided an independent measure of the H
2
activation
kinetic barrier.
16
ACS Paragon Plus Environment

Page 17 of 30
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
Panayotov and Yates employed a similar approach using IR absorbance data to determine
6
2
the H
roughening, and general method for evaluating H
and Pursell using static H
adsorption experiments.61 Yates and Morris reported similar findings,
2
activation barrier on Au/TiO
2
in their system. We note that this description of surface
2
adsorption was described by Pollock, Peterson,
2
describing their IR spectra in terms of H atom spillover.62-64 A proton and electron transferred to
the support can be considered as equivalent to H atom spillover,65 so Yates and Morris’
interpretation is consistent with our findings. Similar changes to IR spectra have also been reported
during low temperature (90 K) photoexcitation of titania, and suggested to be caused by the
injection of electrons into the titania conduction band.66 Our observations are also broadly
consistent with this interpretation.
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
The extracted kinetic data are shown in panels B and C of Figure 6; Arrhenius plots are
shown in Panel D. The activation barrier measured with the BB-IR signal was 25 ± 2 kJ/mol for
-1
-1
the 1800 to 2100 cm region and 20 ± 1 kJ/mol for the 3800-3950 cm spectral region. These
values, which we interpret as estimates of Eapp for H adsorption when θH2O < 1 MLE, agree well
with the Eapp value determined in the reaction kinetics (i.e. packed bed reactor studies).
Considering that no O or additional H O/D O was present in the feed during the IR experiments,
the IR data support the conclusion that the Eapp value measured during H oxidation kinetics is
dominated by H adsorption. Thus, all of the Eapp measurements are internally consistent. They
therefore provide a strong experimental foundation for focusing the DFT calculations on the
dominant H adsorption reaction.
Electronic Changes Associated with H
of this large structural change prompted us to more carefully examine the structural and electronic
changes that result from H adsorption. While the Au-H species is a formal hydride for electron
2
2
2
2
2
2
2
2
Adsorption. The observation and characterization
2
counting purposes, there is relatively little charge on the H atom in the final state. This indicates
the Au-H bond is largely covalent with significant electron donation into the Au nanorod. This is
not surprising given that Au is slightly more electronegative than hydrogen. While many refer to
this species as a “hydrogen atom” we maintain the description of a formal hydride in order to
scrupulously maintain electron counting and distinguish between the proton and hydride.
We also compared the Bader charges on the water-free structures to the calculated reaction
energies. The small variance in the negative charges were normalized to a total charge of -1; this
normalization was ~5% of the total charge (details in the SI). In broad terms, the distribution of
17
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 18 of 30
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
negative charge is similar in all these structures: about ¼ of an electron remains distributed
between the hydride and the Au nanorod, about ½ an electron is localized on the HOcus, about ¼
of an electron is distributed across the rest of the support (Ti and O atoms, see SI Table S1 for
details). The only broad correlation between the final state energies and the distributed charges
was in the relative amounts of charge on the hydride and the Au nanorod. This is shown in Figure
7, which includes a plot showing the correlation between final state energy and the fraction of
negative charge on the Au nanorod.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Figure 7. Hydrogen adsorption final state energies
plotted against the Bader charge on the formal
hydride and the Au nanorod. The inset shows the
energy as a function of the fraction of the total Au
-
+
H charge on the Au nanorod.
Based on Figure 7, it is unclear if this correlation is due to the destabilization of Au as
excess negative charge builds on the nanorod, or if the hydride works to withdraw electron density
from the system. Given the similar electronegativities of Au and H, it may be that the system is
simply most stable when the negative charge is spread more or less evenly over these two parts of
the system. This is tied to the reactivity of the specific surface site: less reactive surface sites (i.e.
9
-coordinate Au(111) sites) are already electronically saturated. We note that this is also consistent
with the broader chemistry of Au. With a full d-band and half full s-band, Au is electronically
stable. This drives the propensity for heterolytic H-H activation, which avoids the formal
oxidation associated with dissociative chemisorption. Similarly, excess electron density over-
populates the s-band, so the system transfers the electron density to the support. In this case, it
largely localizes on the most electronegative atom in the system (HOcus) where it stabilizes the
generated proton. The reaction might therefore be considered as going through a proton coupled
electron transfer-type mechanism.
18
ACS Paragon Plus Environment

Page 19 of 30
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
As discussed above, the water layer causes negligible changes in the overall reaction
energy (< 0.1 eV, Figure 1, SI Table S1). However, water perturbs the final state charge
distribution, shifting ~0.1-0.2 electrons onto the Au at the expense of the support. The charge on
the proton, which resides on the HOcus, remains very close to +1 electron, so it is not immediately
obvious why water stabilizes greater charge on the nanorod. One possibility is that the water layer,
which extends along the length of the nanorod, provides a weak solvation effect that helps to
stabilize greater charge density on the Au surface.
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Transition states
*
A
Au (111)-HO
B
Au(111)-H O-HO
C
Au(211)-H O-HO
2 cus
cus
2
cus
-
0.21 |e-|
-
0.19 |e-|
0.39 |e-|
+
0.45 |e-|
+
-
0.18 |e-|
+0.30 |e-|
E = 0.70 eV
a
E = 1.25 eV
a
E = 0.99 eV
a
*
Figure 8. Charge density difference plots of transition states for H
HOcus, (B) Au(111)-H O, (C) the Au(211)-H O. Green shading (negative charge) shows electron
accumulation, and blue shading (positive charge) shows electron depletion.
2
activation across (A) Au (111)-
2
2
Semi-quantitative structural and electronic analysis of H
2
activation.
Calculated
transition states (Figure 8) show the influence of water on H activation kinetics / dynamics;
2
structural data can be found in Figure 9A and in the Supporting Information. The calculated
activation barrier correlates closely with the degree of charge build-up on the developing proton
(Figure 9A). The total charge on the Au and hydride also tracks with this value, as does the H-H
bond distance (see SI). Thus, the higher barrier is associated with greater charge separation and a
later transition state. Notably, the charge on the hydride is relatively similar in all four structures;
most of the additional charge associated with the later transition state is distributed across the Au
nanorod. Additionally, there is surprisingly little charge distributed to the HOcus or the support in
any of the transition states.
19
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 20 of 30
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Figure 9. Transition state charge distribution
and structural changes. (A) H activation
2
transition state (TS) Bader charges. Bader-
charges are referenced to the initial state (IS). H
+
:
hydride; H : proton; Au: all gold atoms; Ti: all
titanium atoms; HOcus: cus-hydroxyl abstracting
the proton; H cus: water molecule abstracting
the proton; other O: all other oxygen atoms. The
inset shows the calculated H activation energies
2
O
2
plotted against the partial charge of the
developing proton. (B) Ti-Ocus bond length and
the sum of gold and hydride charge in the H
2
activation transition state plotted against the
proton partial charge for various sites at the MSI
and MWI of Au/TiO .
2
The underlying phenomenon we would like to understand is why / how the water layer
shifts the reaction to a later, higher energy transition state. A clue to this question lies in the
structural changes associated with the HOcus group. Figure 9B shows the negative charge buildup
on the hydride and Au nanorod plotted against the charge on the developing proton. As suggested
above, these correlate, with greater negative charge developing to compensate the positive charge.
However, the Ti-O bond shows exactly the opposite trend. Based on the final state geometry, the
Ti-O bond eventually undergoes a ~0.4 Ǻ elongation; however, in the later transition states, the
Ti-O bond is slightly shorter than in the initial state. As Figure 10B shows, this strong correlation
suggests an important difference in the presence of water.
20
ACS Paragon Plus Environment

Page 21 of 30
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
It appears that H
2
activation proceeds through a proton-coupled electron transfer (PCET)-
like mechanism. The HOcus carries relatively little charge in any of the transition states, suggesting
that the positive charge is required to facilitate electron transfer to the support. Even though proton
transfers are relatively fast, the water layer imposes a transport barrier to moving the positive
charge, slowing that process. This, in turn, forces greater negative charge build-up on the Au
nanorod, which the final state calculations show is a net destabilizing influence on the system.
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Thus, water appears to slow H
which is a necessary pre-condition for fast electron transfer from Au to the support.
We initially concluded that water poisoning of H oxidation resulted from simple blocking
2
activation by slowing the proton transfer to the support OH groups,
2
of the fastest reaction sites at the MSI. While this is certainly true, the details of this poisoning are
more complicated and illuminating. In particular, the poisoning appears to occur by slowing the
dynamics of charge transfer between the Au and the support; this is a far more dynamic
representation of the active site than can be understood with a simple site blocking model. Further,
it indicates that both proton and electron transfer play important roles in H
has implications for similar systems.
2
activation; this likely
Hydrogen Spillover and Reducible Supports. Hydrogen activation over Au can
considered in the context of two limiting descriptions of this chemistry: support reduction and
hydrogen spillover. The process involves the transfer of both a proton and an electron to the
support. Since a proton and electron can be reasonably considered as equivalent to a hydrogen
6
5
atom, it can also be described as “hydrogen spillover”. Similarly, H activation over Au involves
2
the transfer of charge to the support and partial localization of the charge onto a surface hydroxyl.
While this might more aptly be described as a redistribution and localization of charge within the
system, it is still consistent with the concept of support reduction.
However, the transfer of negative charge to the support described by the DFT model
presents a remarkably different picture of “support reduction” than is commonly considered.
Typically, a “reducible support” such as TiO
oxygen vacancies produced, as in the classic Mars van Krevelen mechanism. Behm’s group in
particular has provided compelling evidence for the role of O vacancies in CO and H oxidation
Under the conditions of our experiments, where monolayer
2
implies that metal sites can be reduced and / or
2
9
, 67-69
over Au/TiO above ~100 °C.
2
equivalents of water are present at relatively low temperatures, O vacancies cannot be reasonably
invoked as participants in the reaction mechanism. Instead, support reducibility in this case is
21
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 22 of 30
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
governed by the ability / necessity of stabilizing protons and electrons on the support, not by the
propensity to form O vacancies.
It is also worth noting that, in the context of the formalisms associated with oxidation states
and electron counting, there is no “formal” reduction of the support during hydrogen activation.
Both formal oxidation states and organometallic electron counting schemes are valuable for
understanding these systems, but both are ultimately limiting descriptions of the chemistry and
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
should be considered in appropriate context. In this case, H activation is properly considered as
2
heterolytic, generating a proton on the support and a formal hydride on the Au. This process is
driven by the well-known disinclination of Au to undergo oxidation. However, this description,
while valuable and necessary to maintain charge neutrality in describing the process, is ultimately
limiting. The reality, at least as suggested by the DFT calculations, is that the negative charge of
the hydride is distributed throughout the system through the Au. In this sense, the Au functions as
a nanowire to distribute one of the electrons in the H-H bond throughout the system.
2
-
The oxidation states typically used to describe the support (e.g. Ti(IV) and O ) are
similarly limiting descriptions. While these descriptions are necessary to maintain charge
neutrality and provide categories for broad descriptive chemistries, the reality is more complex
with electrons shared between the formal oxide and formal Ti(IV). Numerous other models,
including of course band theory, provide other more detailed but similarly limiting descriptions of
these interactions. Thus, the support does not undergo a “formal” reduction during H activation
2
in the sense of generating a Ti(III) site or producing an O vacancy. Rather, the charge in the system
is partially redistributed and localized on a surface hydroxyl group where it can stabilize (or is
stabilized by) the proton.
This description or interpretation has the appeal of transferring / localizing charge onto the
most electronegative atom(s) in the system. Further, because it does not require formal reduction
of a support cation, it is likely to be a more general description of this chemistry. This is important
for supported Au catalysts, where reducible and non-reducible supports can show similar
chemistries. This is particularly true for CO and H
2
oxidation, which show nearly identical kinetics
over Au/TiO and Au/Al
2
2
O
3
, at least under conditions where water coverage is controlled.38, 48
Thus, beyond expanding our understanding of how “support reducibility” can be considered, this
study provides a clear example of how reducible and non-reducible supports can influence reaction
22
ACS Paragon Plus Environment

Page 23 of 30
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
rates through subtle changes in electronics, without necessitating the invocation of larger formal
electron transfers.70
Finally, we note the complex interplay between metal, support, protons, and electrons
evidenced in this chemistry provide new ways to think about reactions at the metal-support
interface. As noted by a reviewer of this manuscript, this study shows how adsorbates on the metal
and at the metal-support interface can modify the broader electronic structure of the catalyst.
Adsorbate-modified catalysts may similarly have lower energy reaction pathways available that
have not been previously considered. This may be particularly useful for similar reactions over
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
0
d metals where the metal-support interface is known to be important, such as alkyne partial
hydrogenation over Au71-77 or Ag,78 CO
synthesis catalysts.6
hydrogenation over Cu based catalysts,2-5 or methanol
2
Conclusion
This study highlights the complex and multiple ways that adsorbed water can impact
catalysis over supported nanoparticles. The strong product poisoning imposes the experimentally
challenging requirement that kinetic parameters be determined at constant water coverage in order
to maintain a consistent number of metal-support interface sites. The mechanism of water
poisoning, and the insight it provides into the reactivity of Au catalysts are particularly noteworthy.
Rate limiting hydrogen activation appears to go through a PCET-like mechanism, whereby a
proton from H is transferred to a support hydroxyl group. The associated negative charge is
2
distributed across the formal hydride, the Au, and the support, with the largest concentration of
charge on the hydroxyl group to which the proton was transferred. Water slows this process by
providing a transport barrier to moving the proton to the support hydroxyl group. This pushes the
reaction to a later transition state, forcing greater (and unfavorable) negative charge build-up on
the Au, thereby increasing the transition state energy by ~0.2 eV. A similar change in apparent
activation energy was determined experimentally through reaction kinetics and in-situ infrared
spectroscopic studies.
Acknowledgements
The authors gratefully acknowledge the National Science Foundation (Grants CHE-
1465148, CHE-1465184 and CBET-1803769) and the Research Corporation for Science
23
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 24 of 30
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
Advancement for supporting this work. The authors thank Prof. Siris Laursen and Evan Williams
for helpful discussions. BDC thanks Profs. Jeff Bode and Christophe Copéret as well as the
Laboratorium für Organische Chemie and the Laboratorium für Anorganische Chemie at ETH
Zürich for support during his academic leave.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Supplementary Information
Methods and materials; Development details of DFT model water layer; Effect of hydride
position on H
2
dissociation energetics; DFT examination of H activation across the MWI sites;
2
DFT charge analysis; DFT Bond length analysis; Additional kinetics data
References:
1.
Ponec, V.; Bond, G. C., Catalysis by Metals and Alloys. Elsevier: Amsterdam, 1995; Vol. 95.
2.
Kattel, S.; Yan, B.; Yang, Y.; Chen, J. G.; Liu, P., Optimizing Binding Energies of Key
Intermediates for CO2 Hydrogenation to Methanol over Oxide-Supported Copper. J. Am. Chem. Soc. 2016,
138 (38), 12440-12450.
3. Kattel, S.; Ramirez, P. J.; Chen, J. G.; Rodriguez, J. A.; Liu, P., Active sites for CO2 hydrogenation
to methanol on Cu/ZnO catalysts. Science (Washington, DC, U. S.) 2017, 355 (6331), 1296-1299.
4.
Larmier, K.; Liao, W.-C.; Tada, S.; Lam, E.; Verel, R.; Bansode, A.; Urakawa, A.; Comas-Vives,
A.; Coperet, C., CO2-to-Methanol Hydrogenation on Zirconia-Supported Copper Nanoparticles: Reaction
Intermediates and the Role of the Metal-Support Interface. Angew. Chem., Int. Ed. 2017, 56 (9), 2318-2323.
5.
Lam, E.; Larmier, K.; Noh, G.; Wolf, P.; Comas-Vives, A.; Coperet, C.; Corral-Perez, J. J.;
Urakawa, A.; Comas-Vives, A.; Urakawa, A., CO2 Hydrogenation on Cu/Al2 O3 : Role of the
Metal/Support Interface in Driving Activity and Selectivity of a Bifunctional Catalyst. Angew Chem Int Ed
Engl 2019, 58 (39), 13989-13996.
6.
Behrens, M.; Studt, F.; Kasatkin, I.; Kuehl, S.; Haevecker, M.; Abild-Pedersen, F.; Zander, S.;
Girgsdies, F.; Kurr, P.; Kniep, B.-L.; Tovar, M.; Fischer, R. W.; Norskov, J. K.; Schloegl, R., The Active
Site of Methanol Synthesis over Cu/ZnO/Al2O3 Industrial Catalysts. Science (Washington, DC, U. S.)
2012, 336 (6083), 893-897.
7.
Robinson, A. M.; Hensley, J. E.; Medlin, J. W., Bifunctional Catalysts for Upgrading of Biomass-
Derived Oxygenates: A Review. ACS Catal. 2016, 6 (8), 5026-5043.
8.
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), 1599-1602.
9.
Wang, Y.; Widmann, D.; Heenemann, M.; Diemant, T.; Biskupek, J.; Schloegl, R.; Behm, R. J.,
The role of electronic metal-support interactions and its temperature dependence: CO adsorption and CO
oxidation on Au/TiO2 catalysts in the presence of TiO2 bulk defects. J. Catal. 2017, 354, 46-60.
24
ACS Paragon Plus Environment

Page 25 of 30
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
1
0.
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.
11. Carrasquillo-Flores, R.; Ro, I.; Kumbhalkar, M. D.; Burt, S.; Carrero, C. A.; Alba-Rubio, A. C.;
Miller, J. T.; Hermans, I.; Huber, G. W.; Dumesic, J. A., Reverse Water-Gas Shift on Interfacial Sites
Formed by Deposition of Oxidized Molybdenum Moieties onto Gold Nanoparticles. J. Am. Chem. Soc.
2015, 137 (32), 10317-10325.
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
12. Zhao, Z.-J.; Li, Z.; Cui, Y.; Zhu, H.; Schneider, W. F.; Delgass, W. N.; Ribeiro, F.; Greeley, J.,
Importance of metal-oxide interfaces in heterogeneous catalysis: A combined DFT, microkinetic, and
experimental study of water-gas shift on Au/MgO. J. Catal. 2017, 345, 157-169.
13.
Wang, L.; Zhang, J.; Wang, H.; Shao, Y.; Liu, X.; Wang, Y.-Q.; Lewis, J. P.; Xiao, F.-S., Activity
and Selectivity in Nitroarene Hydrogenation over Au Nanoparticles on the Edge/Corner of Anatase. ACS
Catal. 2016, 6 (7), 4110-4116.
14.
Corma, A.; Serna, P., Chemoselective Hydrogenation of Nitro Compounds with Supported Gold
Catalysts. Science 2006, 313 (5785), 332-334.
15.
Zhu, Y. P.; Guo, C.; Zheng, Y.; Qiao, S.-Z., Surface and Interface Engineering of Noble-Metal-
Free Electrocatalysts for Efficient Energy Conversion Processes. Acc. Chem. Res. 2017, 50 (4), 915-923.
16.
Song, F.; Li, W.; Yang, J.; Han, G.; Yan, T.; Liu, X.; Rao, Y.; Liao, P.; Cao, Z.; Sun, Y., Interfacial
Sites between Cobalt Nitride and Cobalt Act as Bifunctional Catalysts for Hydrogen Electrochemistry. ACS
Energy Lett. 2019, 4 (7), 1594-1601.
17.
Ro, I.; Resasco, J.; Christopher, P., Approaches for Understanding and Controlling Interfacial
Effects in Oxide-Supported Metal Catalysts. ACS Catalysis 2018, 8 (8), 7368-7387.
18.
Mehta, P.; Greeley, J.; Delgass, W. N.; Schneider, W. F., Adsorption Energy Correlations at the
Metal-Support Boundary. ACS Catal. 2017, 7 (7), 4707-4715.
19.
Schauermann, S.; Nilius, N.; Shaikhutdinov, S.; Freund, H.-J., Nanoparticles for Heterogeneous
Catalysis: New Mechanistic Insights. Accounts of Chemical Research 2013, 46 (8), 1673-1681.
20.
Campbell, C. T., Catalyst-support interactions: Electronic perturbations. Nat Chem 2012, 4 (8),
597-598.
21.
Liu, Z.; Grinter, D. C.; Lustemberg, P. G.; Nguyen-Phan, T.-D.; Zhou, Y.; Luo, S.; Waluyo, I.;
Crumlin, E. J.; Stacchiola, D. J.; Zhou, J.; Carrasco, J.; Busnengo, H. F.; Ganduglia-Pirovano, M. V.;
Senanayake, S. D.; Rodriguez, J. A., Dry Reforming of Methane on a Highly-Active Ni-CeO2 Catalyst:
Effects of Metal-Support Interactions on C-H Bond Breaking. Angew. Chem., Int. Ed. 2016, 55 (26), 7455-
7459.
22.
Bruix, A.; Rodriguez, J. A.; Ramirez, P. J.; Senanayake, S. D.; Evans, J.; Park, J. B.; Stacchiola,
D.; Liu, P.; Hrbek, J.; Illas, F., A New Type of Strong Metal-Support Interaction and the Production of H2
through the Transformation of Water on Pt/CeO2(111) and Pt/CeOx/TiO2(110) Catalysts. J. Am. Chem.
Soc. 2012, 134 (21), 8968-8974.
25
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 26 of 30
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
2
3.
Lykhach, Y.; Kozlov, S. M.; Skala, T.; Tovt, A.; Stetsovych, V.; Tsud, N.; Dvorak, F.; Johanek,
V.; Neitzel, A.; Myslivecek, J.; Fabris, S.; Matolin, V.; Neyman, K. M.; Libuda, J., Counting electrons on
supported nanoparticles. Nat. Mater. 2016, 15 (3), 284-288.
24.
Divins, N. J.; Angurell, I.; Escudero, C.; Perez-Dieste, V.; Llorca, J., Influence of the support on
surface rearrangements of bimetallic nanoparticles in real catalysts. Science (Washington, DC, U. S.) 2014,
3
46 (6209), 620-623.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
2
5. Huang, L.; Han, B.; Xi, Y.; Forrey, R. C.; Cheng, H., Influence of Charge on the Reactivity of
Supported Heterogeneous Transition Metal Catalysts. ACS Catal. 2015, 5 (8), 4592-4597.
2
6.
Pacchioni, G., Electronic interactions and charge transfers of metal atoms and clusters on oxide
surfaces. Physical Chemistry Chemical Physics 2013, 15 (6), 1737-1757.
27.
Hartadi, Y.; Behm, R.; Widmann, D., Competition of CO and H2 for Active Oxygen Species during
the Preferential CO Oxidation (PROX) on Au/TiO2 Catalysts. Catalysts 2016, 6 (2), 21.
28.
Karim, W.; Spreafico, C.; Kleibert, A.; Gobrecht, J.; VandeVondele, J.; Ekinci, Y.; van Bokhoven,
J. A., Catalyst support effects on hydrogen spillover. Nature 2017, 541, 68.
29.
Briggs, N. M.; Barrett, L.; Wegener, E. C.; Herrera, L. V.; Gomez, L. A.; Miller, J. T.; Crossley, S.
P., Identification of active sites on supported metal catalysts with carbon nanotube hydrogen highways.
Nat. Commun. 2018, 9 (1), 1-7.
30.
Liu, J.; Hibbitts, D.; Iglesia, E., Dense CO Adlayers as Enablers of CO Hydrogenation Turnovers
on Ru Surfaces. Journal of the American Chemical Society 2017, 139 (34), 11789-11802.
31.
Hibbitts, D.; Dybeck, E.; Lawlor, T.; Neurock, M.; Iglesia, E., Preferential activation of CO near
hydrocarbon chains during Fischer–Tropsch synthesis on Ru. Journal of Catalysis 2016, 337, 91-101.
32.
Wang, Y.-G.; Cantu, D. C.; Lee, M.-S.; Li, J.; Glezakou, V.-A.; Rousseau, R., CO Oxidation on
Au/TiO2: Condition-Dependent Active Sites and Mechanistic Pathways. Journal of the American Chemical
Society 2016, 138 (33), 10467-10476.
33.
Polierer, S.; Jelic, J.; Pitter, S.; Studt, F., On the Reactivity of the Cu/ZrO2 System for the
Hydrogenation of CO2 to Methanol: A Density Functional Theory Study. The Journal of Physical
Chemistry C 2019, 123 (44), 26904-26911.
34.
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.
35.
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.
36.
Nilekar, A. U.; Alayoglu, S.; Eichhorn, B.; Mavrikakis, M., Preferential CO Oxidation in
Hydrogen: Reactivity of Core-Shell Nanoparticles. J. Am. Chem. Soc. 2010, 132 (21), 7418-7428.
37.
Rossignol, C.; Arrii, S.; Morfin, F.; Piccolo, L.; Caps, V.; Rousset, J.-L., Selective oxidation of CO
over model gold-based catalysts in the presence of H2. J. Catal. 2005, 230 (2), 476-483.
26
ACS Paragon Plus Environment

Page 27 of 30
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
3
8.
Saavedra, J.; Pursell, C. J.; Chandler, B. D., CO Oxidation Kinetics over Au/TiO2 and Au/Al2O3
Catalysts: Evidence for a Common Water-Assisted Mechanism. Journal of the American Chemical Society
2
018, 140 (10), 3712-3723.
3
9. 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.
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
4
0.
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.
4
1. Tran, H.-V.; Doan, H. A.; Chandler, B. D.; Grabow, L. C., Water-assisted oxygen activation during
selective oxidation reactions. Current Opinion in Chemical Engineering 2016, 13, 100-108.
4
2. Paulus, U. A.; Wang, Y.; Kim, S. H.; Geng, P.; Wintterlin, J.; Jacobi, K.; Ertl, G., Inhibition of CO
oxidation on RuO2(110) by adsorbed H2O molecules. The Journal of Chemical Physics 2004, 121 (22),
11301-11308.
43.
Daté, M.; Haruta, M., Moisture Effect on CO Oxidation over Au/TiO2 Catalyst. Journal of
Catalysis 2001, 201 (2), 221-224.
44.
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.
4
5.
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.
46.
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. Accounts of Chemical Research 2014, 47 (3),
7
4
40-749.
7. Yan, T.; Gong, J.; Flaherty, D. W.; Mullins, C. B., The Effect of Adsorbed Water in CO Oxidation
on Au/TiO2(110). The Journal of Physical Chemistry C 2011, 115 (5), 2057-2065.
48. Whittaker, T.; Kumar, K. B. S.; Peterson, C.; Pollock, M. N.; Grabow, L. C.; Chandler, B. D., H2
Oxidation over Supported Au Nanoparticle Catalysts: Evidence for Heterolytic H2 Activation at the Metal–
Support Interface. Journal of the American Chemical Society 2018, 140 (48), 16469-16487.
4
9.
Werner, K.; Weng, X.; Calaza, F.; Sterrer, M.; Kropp, T.; Paier, J.; Sauer, J.; Wilde, M.; Fukutani,
K.; Shaikhutdinov, S.; Freund, H.-J., Toward an Understanding of Selective Alkyne Hydrogenation on
Ceria: On the Impact of O Vacancies on H2 Interaction with CeO2(111). Journal of the American Chemical
Society 2017, 139 (48), 17608-17616.
50.
Dahal, A.; Mu, R.; Lyubinetsky, I.; Dohnálek, Z., Hydrogen adsorption and reaction on RuO2(110).
Surface Science 2018, 677, 264-270.
51.
Hu, G.; Wu, Z.; Jiang, D.-e., First Principles Insight into H2 Activation and Hydride Species on
TiO2 Surfaces. The Journal of Physical Chemistry C 2018, 122 (35), 20323-20328.
5
2.
Huang, Z.-Q.; Liu, L.-P.; Qi, S.; Zhang, S.; Qu, Y.; Chang, C.-R., Understanding All-Solid
Frustrated-Lewis-Pair Sites on CeO2 from Theoretical Perspectives. ACS Catalysis 2018, 8 (1), 546-554.
27
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 28 of 30
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
5
3.
Liu, W.; Chen, Y.; Qi, H.; Zhang, L.; Yan, W.; Liu, X.; Yang, X.; Miao, S.; Wang, W.; Liu, C.;
Wang, A.; Li, J.; Zhang, T., A Durable Nickel Single-Atom Catalyst for Hydrogenation Reactions and
Cellulose Valorization under Harsh Conditions. Angewandte Chemie 2018, 130 (24), 7189-7193.
54.
Riley, C.; Zhou, S.; Kunwar, D.; De La Riva, A.; Peterson, E.; Payne, R.; Gao, L.; Lin, S.; Guo,
H.; Datye, A., Design of Effective Catalysts for Selective Alkyne Hydrogenation by Doping of Ceria with
a Single-Atom Promotor. Journal of the American Chemical Society 2018, 140 (40), 12964-12973.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
55.
Shen, Y.; Yin, K.; An, C.; Xiao, Z., Design of a difunctional Zn-Ti LDHs supported PdAu catalyst
for selective hydrogenation of phenylacetylene. Applied Surface Science 2018, 456, 1-6.
56.
Phadke, N. M.; Mansoor, E.; Bondil, M.; Head-Gordon, M.; Bell, A. T., Mechanism and Kinetics
of Propane Dehydrogenation and Cracking over Ga/H-MFI Prepared via Vapor-Phase Exchange of H-MFI
with GaCl3. Journal of the American Chemical Society 2019, 141 (4), 1614-1627.
57.
Vityuk, A.; Khivantsev, K.; Aleksandrov, H. A.; Vayssilov, G. N.; Alexeev, O. S.; Amiridis, M.
D., Room-Temperature Ethene Hydrogenation Activity of Transition-Metal-Free HY Zeolites. ACS
Catalysis 2019, 9 (2), 839-847.
58.
Wu, D.; Hernández, W. Y.; Zhang, S.; Vovk, E. I.; Zhou, X.; Yang, Y.; Khodakov, A. Y.;
Ordomsky, V. V., In Situ Generation of Brønsted Acidity in the Pd-I Bifunctional Catalysts for Selective
Reductive Etherification of Carbonyl Compounds under Mild Conditions. ACS Catalysis 2019, 2940-2948.
59.
Fujitani, T.; Nakamura, I.; Akita, T.; Okumura, M.; Haruta, M., Hydrogen Dissociation by Gold
Clusters. Angewandte Chemie International Edition 2009, 48 (50), 9515-9518.
60.
Powell, C. D.; Daigh, A. W.; Pollock, M. N.; Chandler, B. D.; Pursell, C. J., CO Adsorption on
Au/TiO2 Catalysts: Observations, Quantification, and Explanation of a Broad-Band Infrared Signal. J.
Phys. Chem. C 2017, 121 (44), 24541-24547.
61.
Pollock, M. N.; Peterson, C.; Pursell, C.; Chandler, B. D. In Gold nanoparticle catalysts: Chemical
properties and catalytic behavior, American Chemical Society: 2017; pp INOR-436.
62.
Panayotov, D. A.; Yates, J. T., Spectroscopic Detection of Hydrogen Atom Spillover from Au
Nanoparticles Supported on TiO2:ꢀ Use of Conduction Band Electrons. The Journal of Physical Chemistry
C 2007, 111 (7), 2959-2964.
63.
Powell, C. D.; Daigh, A. W.; Pollock, M. N.; Chandler, B. D.; Pursell, C. J., CO Adsorption on
Au/TiO2 Catalysts: Observations, Quantification, and Explanation of a Broad-Band Infrared Signal. The
Journal of Physical Chemistry C 2017, 121 (44), 24541-24547.
64.
Panayotov, D. A.; Burrows, S. P.; Yates, J. T.; Morris, J. R., Mechanistic Studies of Hydrogen
Dissociation and Spillover on Au/TiO2: IR Spectroscopy of Coadsorbed CO and H-Donated Electrons. The
Journal of Physical Chemistry C 2011, 115 (45), 22400-22408.
65.
Prins, R., Hydrogen Spillover. Facts and Fiction. Chemical Reviews 2012, 112 (5), 2714-2738.
66.
Berger, T.; Sterrer, M.; Diwald, O.; Knoezinger, E.; Panayotov, D.; Thompson, T. L.; Yates, J. T.,
Jr., Light-Induced Charge Separation in Anatase TiO2 Particles. J. Phys. Chem. B 2005, 109 (13), 6061-
068.
6
28
ACS Paragon Plus Environment

Page 29 of 30
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
6
7.
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.
68.
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-
2
6
76.
9.
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
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.
70. Kumar, G.; Tibbitts, L.; Newell, J.; Panthi, B.; Mukhopadhyay, A.; Rioux, R. M.; Pursell, C. J.;
Janik, M.; Chandler, B. D., Evaluating differences in the active-site electronics of supported Au
nanoparticle catalysts using Hammett and DFT studies. Nature Chemistry 2018, 10, 268.
71.
Segura, Y.; Lopez, N.; Perez-Ramirez, J., Origin of the superior hydrogenation selectivity of gold
nanoparticles in alkyne + alkene mixtures: Triple- versus double-bond activation. J. Catal. 2007, 247 (2),
383-386.
72.
Nikolaev, S. A.; Smirnov, V. V., Synergistic and size effects in selective hydrogenation of alkynes
on gold nanocomposites. Catal. Today 2009, 147 (Suppl.), S336-S341.
73.
Bruno, J. E.; Sravan Kumar, K. B.; Dwarica, N. S.; Hüther, A.; Chen, Z.; Guzman IV, C. S.; Hand,
E. R.; Moore, W. C.; Rioux, R. M.; Grabow, L. C.; Chandler, B. D., On the Limited Role of Electronic
Support Effects in Selective Alkyne Hydrogenation: A Kinetic Study of Au/MOx Catalysts Prepared from
Oleylamine-Capped Colloidal Nanoparticles. ChemCatChem 2019, 11 (6), 1650-1664.
74.
Jenkins, A. H.; Musgrave, C. B.; Medlin, J. W., Enhancing Au/TiO2 Catalyst Thermostability and
Coking Resistance with Alkyl Phosphonic-Acid Self-Assembled Monolayers. ACS Applied Materials &
Interfaces 2019, 11 (44), 41289-41296.
75.
Luneau, M.; Shirman, T.; Foucher, A. C.; Duanmu, K.; Verbart, D. M. A.; Sautet, P.; Stach, E. A.;
Aizenberg, J.; Madix, R. J.; Friend, C. M., Achieving High Selectivity for Alkyne Hydrogenation at High
Conversions with Compositionally Optimized PdAu Nanoparticle Catalysts in Raspberry Colloid-
Templated SiO2. ACS Catalysis 2020, 10 (1), 441-450.
76.
Hugon, A.; Delannoy, L.; Louis, C., Influence of the reactant concentration in selective
hydrogenation of 1,3-butadiene over supported gold catalysts under alkene rich conditions: a consideration
of reaction mechanism. Gold Bull. (London, U. K.) 2009, 42 (Copyright (C) 2011 American Chemical
Society (ACS). All Rights Reserved.), 310-320.
77.
Hugon, A.; Delannoy, L.; Krafft, J.-M.; Louis, C., Selective Hydrogenation of 1,3-Butadiene in the
Presence of an Excess of Alkenes over Supported Bimetallic Gold-Palladium Catalysts. J. Phys. Chem. C
2010, 114 (24), 10823-10835.
78.
Oakton, E.; Vilé, G.; S. Levine, D.; Zocher, E.; Baudouin, D.; Pérez-Ramírez, J.; Copéret, C., Silver
nanoparticles supported on passivated silica: preparation and catalytic performance in alkyne semi-
hydrogenation. Dalton Transactions 2014, 43 (40), 15138-15142.
29
ACS Paragon Plus Environment