Ag/Au mixed sites promote oxidative coupling of methanol on the alloy surface

Article abstract of DOI:10.1002/chem.201304837

Nanoporous gold, a dilute alloy of Ag in Au, activates molecular oxygen and promotes the oxygen-assisted catalytic coupling of methanol. Because this trace amount of Ag inherent to nanoporous gold has been proposed as the source of oxygen activation, a thin film Ag/Au alloy surface was studied as a model system for probing the origin of this reactivity. Thin alloy layers of Ag _xAu_1-x, with 0.15≤x≤0.40, were examined for dioxygen activation and methanol self-coupling. These alloy surfaces recombine atomic oxygen at different temperatures depending on the alloy composition. Total conversion of methanol to selective oxidation products, that is, formaldehyde and methyl formate, was achieved at low initial oxygen coverage and at low temperature. Reaction channels for methyl formate formation occurred on both Au and Au/Ag mixed sites with a ratio, as was predicted from the local 2-dimensional composition. Activation and promotion: Au/Au mixed sites promotes selective oxidation of methanol to formaldehyde and methyl formate without producing any CO₂ (see scheme).

Full text of DOI:10.1002/chem.201304837

DOI: 10.1002/chem.201304837
Full Paper
&
Bimetallic Catalysis
Ag/Au Mixed Sites Promote Oxidative Coupling of Methanol on
the Alloy Surface
Bingjun Xu,[a, c] Cassandra G. F. Siler,[b] Robert J. Madix,[b] and Cynthia M. Friend*[a, b]
Abstract: Nanoporous gold, a dilute alloy of Ag in Au, acti-
vates molecular oxygen and promotes the oxygen-assisted
catalytic coupling of methanol. Because this trace amount of
Ag inherent to nanoporous gold has been proposed as the
source of oxygen activation, a thin film Ag/Au alloy surface
was studied as a model system for probing the origin of this
reactivity. Thin alloy layers of AgxAu1Àx, with 0.15ꢀxꢀ0.40,
were examined for dioxygen activation and methanol self-
coupling. These alloy surfaces recombine atomic oxygen at
different temperatures depending on the alloy composition.
Total conversion of methanol to selective oxidation products,
that is, formaldehyde and methyl formate, was achieved at
low initial oxygen coverage and at low temperature. Reac-
tion channels for methyl formate formation occurred on
both Au and Au/Ag mixed sites with a ratio, as was predict-
ed from the local 2-dimensional composition.
Introduction
Two striking features of the nanoporous Au catalyst are: 1) a
remarkable resistance to deactivation and 2) its activation of
O2 in the absence of a supporting oxide. The first characteristic
is likely the result of its ligament structure.[13] However, its fa-
cility for activating molecular oxygen cannot be explained by
the mechanisms proposed in the literature for supported gold
catalyst; namely, a size effect[15,16] or activation at the particle-
support interface.[17] However, nanoporous gold is typically pre-
pared by de-alloying Ag/Au alloy, leading to at least 1% of re-
sidual Ag. This small amount of Ag present in nanoporous Au
has been suggested to dissociate molecular oxygen, producing
adsorbed atomic oxygen, which can then spill over to Au and
initiate the catalytic cycle for selective oxidation. Indeed, acti-
vation of dioxygen on nanoporous Au has been demonstrated
in ultrahigh vacuum conditions.[18,19]
Selective oxidation is one of the key transformations in organic
chemistry, the main challenge being suppression of combus-
tion or secondary oxidation of desired products. Gold-based
catalysts have great potential for such processes, for example,
epoxidation of olefins,[1,2] oxidative coupling of alcohols,[3–6] al-
dehydes[7] and amines,[8,9] and oxidative carbonylation reac-
tions.[10]
Many of these reactions share a general mechanistic frame-
work. Alcohols, aldehydes, and amines adsorb and desorb re-
versibly on otherwise clean Au(111).[11,12] However, adsorbed
atomic oxygen selectively activates OÀH and NÀH bonds in al-
cohols and amines on Au to form the corresponding alkoxy or
amido groups.[11] Oxidative coupling then proceeds via a nucle-
ophilic attack of these species at electron-deficient centers, for
example, in aldehydes.[11,12] This mechanistic framework, estab-
lished by studies of reactions with supporting spectroscopic
identification of reaction intermediates on well-defined surfa-
ces under ultrahigh vacuum conditions, can be directly trans-
ferred to catalysis by nanoporous Au under ambient conditions
at modest temperatures.[13,14]
The presence of Ag in the nanoporous gold also has the po-
tential to alter bonding and reactivity of intermediates bound
to the surface. In alloys, the electronic and geometric structure
of the surface can be altered and the nature of bonding and
reaction sites changed substantially. These effects—often re-
ferred to as electronic and ligand effects—can change selectiv-
ity and activity in catalytic processes in a way that is not
simply additive.[20,21]
Herein, we report the effect of alloying the surface of
Au(111) with Ag on the mechanistic oxidative coupling of
methanol. We show that the Ag/Au alloy exhibits different re-
activity from either pure Au or Ag surface and attribute the
phenomenon to presence of mixed Ag/Au sites on the surface.
[a] Dr. B. Xu, Prof. C. M. Friend
Department of Chemistry and Chemical Biology
Harvard University, Cambridge, MA 02138 (USA)
Fax: (+1)617-496-8410
E-mail: cfriend@seas.harvard.edu
[b] C. G. F. Siler, Prof. R. J. Madix, Prof. C. M. Friend
School of Engineering and Applied Sciences
Harvard University, Cambridge, MA 02138 (USA)
Results
[c] Dr. B. Xu
Present address: Department of Chemical and Biological Engineering
University of Delaware, Newark, DE 19716 (USA)
Oxygen binding and desorption
The temperature-programmed desorption profile of molecular
oxygen from recombination of atomic oxygen on the Ag/Au
Supporting information for this article is available on the WWW under
http://dx.doi.org/10.1002/chem.201304837.
Chem. Eur. J. 2014, 20, 1 – 8
1
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
&
These are not the final page numbers! ÞÞ

Full Paper
thin film alloy surface is sensitive to the composition of the
surface and local bonding environment of the adsorbed
oxygen atoms. We focus on Ag0.4Au0.6 herein to exemplify the
general trend of the binding and desorption of oxygen on the
Ag/Au alloy surface. Distinctive O2 features from recombina-
tion of adsorbed O atoms are observed, at approximately 525
and 540 K from the Ag0.4Au0.6 alloy; the relative amount of O2
produced in these peaks depends on the initial oxygen cover-
age (Figure 1). The peak temperature for the 525 K peak (Fig-
540 K growing with the increase of Ag content in the alloy
(Figure S1 in the Supporting Information).
Adsorbed atomic oxygen oxidizes both Ag and Au on
Ag0.4Au0.6. The O (1 s) peak at 529.5 eV is evident, similar to
that observed on Au(111).[9,12] Its presence (qO =1 ML) causes
the binding energies of both Ag (3d) and Au (4f) peaks to shift
(Figure 2 and Figure S3 in the Supporting Information). A new
Figure 2. A) Au(4 f) and B) Ag(3d) X-ray photoelectron difference spectra of
atomic oxygen covered Ag0.4Au0.6 (qO =1 ML) at a) 200 K, b) 440 K, c) 500 K,
d) 520 K, and e) 540 K. The contributions of metallic Au and Ag to the spec-
tra have been subtracted for clarity. The spectra before subtraction are
given in Figure S3 in the Supporting Information.
shoulder of the Au (4f7/2) peak appears at 85.5 eV, correspond-
ing to oxidized Au.[22] The peak increases in area by approxi-
mately 70% after annealing the surface to 440 K and disap-
pears above 500 K with the desorption of the oxygen. A similar
shift of the Au(4f) binding energy was observed for O/Au(111)
(qO =1 ML). Adsorbed oxygen produces a new Ag (3d5/2) peak
at 367.2 eV, corresponding to oxidized Ag.[23] The integrated
area of the Ag (3d5/2) peak is largest at 440 K, and it does not
disappear until above 520 K, higher than the temperature at
which the feature for oxidized gold vanishes (Figure 2B).[24] The
higher stability of the surface oxide of Ag is evident. The intro-
duction of atomic oxygen on Ag/Au alloy thin film does not
change the ratio of the integrated area of Ag (3d5/2) and Au
(4f7/2) peaks, indicating that within the limit of our XPS meas-
urements no preferential segregation of either Ag or Au to the
surface occurs.
Figure 1. Temperature-programmed desorption of oxygen on Ag0.4Au0.6 sur-
face alloy at various initial oxygen coverages (a–e). Temperature-pro-
grammed desorption of oxygen on Au(111) (f, qO =1 ML) and Ag(111) (g,
qO =0.4 ML). Atomic oxygen was formed on the Ag/Au alloy by ozone expo-
sure at 200 K; on Ag(111), it was formed by NO2 exposure at 500 K.
ure 1a–e) is identical to that on Au(111) (Figure 1 f), suggesting
that the recombination and desorption occurs from sites simi-
lar to those on Au(111), referred to herein as Au-like sites. A de-
sorption peak at 540 K suggests recombination involving ad-
sorbed O atoms bound at sites with a local environment signif-
icantly different from either pure Au(111) or Ag(111). We refer
to those sites as mixed sites. No O2 desorption feature near
595 K was observed, suggesting desorption did not occur from
surface sites similar to those on Ag(111) (Ag-like sites). Similar
results were also observed on alloys with other compositions
(AgxAu1Àx, with 0.15ꢀxꢀ0.40).
Methanol oxidative coupling
The relative amounts of oxygen desorption from the two
channels depends on the initial oxygen coverage. At low
oxygen coverage (qO <0.8 ML), desorption from mixed sites
dominates (Figure 1). With increasing initial oxygen coverage
the Au-like sites become increasingly important, dominating
for qO ꢁ0.8 ML. Similar oxygen desorption profiles were ob-
served for Ag/Au thin film alloys with Ag molar fraction from
0–40%, with the oxygen desorption peak at approximately
Oxidative coupling of methanol to give methyl formate occurs
efficiently on the Ag/Au alloy after activation by atomic
oxygen (qO =0.1 ML) at 200 K (Figure 3). Methyl [D4]formate
was observed as the primary product of CD3OH.
[D2]Formaldehyde and CO2 were also detected; they are attrib-
uted to selective oxidation and combustion of CD3OH.[25,26] D2
was also detected at the same temperature range as formalde-
&
&
Chem. Eur. J. 2014, 20, 1 – 8
www.chemeurj.org
2
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ÝÝ These are not the final page numbers!

Full Paper
Figure 3. Oxidative coupling of [D3]methanol (CD3OH) on O/Ag0.4Au0.6
(qO =0.1 ML) to produce methyl [D4]formate is demonstrated by using tem-
perature-programmed reaction experiments. Atomic oxygen on Ag0.4Au0.6 is
created by ozone exposure at 200 K.
Figure 4. A) Product distribution of oxidative coupling of CD3OH on
O/Ag0.4Au0.6 (qO =0.1 ML) as a function of initial oxygen coverage. B) Abso-
lute amount of methyl [D4]formate formed and residual oxygen as a function
of initial oxygen coverage.
hyde and methyl formate; neither formaldehyde nor D2 were
observed for the oxidative coupling of methanol on oxygen-
activated Au(111).
At higher O coverages the methyl formate and formaldehyde
yields diminish, whereas CO2 increases. Coverages ꢀ0.05 ML
clearly favor partial oxidation.
The selectivity for the oxidative coupling on the alloy surfa-
ces depends on the initial oxygen coverage in a manner similar
to that observed for pure Au(111). As illustrated for Ag0.4Au0.6
(Figure 4A), low oxygen concentrations lead to high selectivity
for the coupling to form methyl formate, the selectivity reach-
ing 60% at an initial oxygen coverage of 0.05 ML.[3] Formalde-
hyde is also formed appreciably at this low O coverage: there
is no detectable amount of CO2.
The kinetics for methyl formate evolution reflects the distri-
bution of the sites available on the alloy film. When the reac-
tion is carried out on surfaces ranging from pure Au to pure
Ag (qO =0.1 ML), methyl formate is formed at three distinct
temperatures (Figure 5A). The gold- and grey-hatched areas in
Figure 5A are attributed to the contributions form self-cou-
At higher oxygen coverage (>0.1 ML), combustion of meth-
anol dominates (Figure 4A and Table 1) and excess oxygen re-
mains on the surface, based on the evolution of O2 above
500 K (Figure 4B). The yield of methyl formate initially rises,
reaching a maximum for an initial oxygen coverage of 0.1 ML.
pling on O/Au(111) and O/Ag(111), respectively. On Ag0.15Au0.85,
approximately 75% of the methyl formate evolves around
210 K and can be attributed to reaction on Au-like sites.[3,12]
The methyl formate formed near 260 K, which accounts for ap-
proximately 25% of the total methyl formate production, is at-
tributed to reaction at mixed Ag/Au, because it is evolved
about 90 K lower than on Ag(111). The percentage of coupling
due to mixed sites increases with the molar fraction of Ag in
the alloy (Figure 5B), and the contribution from Au sites de-
creases correspondingly. Only a very small amount of methyl
formate evolves from Ag-like sites, even at the highest Ag con-
centration in the alloy surface (Ag0.6Au0.4 (<1%)), and no de-
tectable amount was observed for alloys with lower Ag con-
tents. Ester formed from mixed sites is the dominant route on
Ag0.6Au0.4. The yield for methyl formate increases with the Ag
content, the amount of methyl formate produced on
O/Ag0.6Au0.4 being 50% higher than that on O/Au(111) (qO =
0.1 ML in both cases).
Table 1. Product distribution of selective oxidation of methanol on
O/Ag0.4Au0.6
.
qO [ML]
Selectivity[a] [%] (Æ10%)
Methyl formate
Formaldehyde
CO2
0.05
0.1
0.2
0.4
0.6
1.0
58
46
7
6
5
42
22
0
0
0
0
32
93
94
95
96
4
0
[a] Parent ions of methyl [D4]formate (64 amu), [D2]formaldehyde
(32 amu), and CO2 (44 amu) were used as signature masses in the quan-
tification process, following the procedure detailed in Ref. [10].
The oxygen-assisted coupling proceeds through the reaction
of the adsorbed methoxy intermediate with formaldehyde
Chem. Eur. J. 2014, 20, 1 – 8
www.chemeurj.org
3
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
&
These are not the final page numbers! ÞÞ

Full Paper
Figure 5. A) Self-coupling of methanol on atomic-oxygen-covered
(qO =0.1 ML) Au(111), Ag(111), and Ag/Au surface alloys with various Ag con-
tents. Cross-coupling between co-adsorbed methanol and formaldehyde on
atomic-oxygen-covered Ag(111) (qO =0.1 ML). B) Percentage of methyl
[D4]formate attributed to self- and cross-coupling, as well as the total yield
of methyl [D4]formate as a function of Ag content in the Ag/Au alloy.
Atomic oxygen on Ag(111) is introduced by NO2 exposure at 500 K. Atomic
oxygen on Au(111) and Ag/Au surface alloys was introduced by ozone expo-
sure at 200 K. All chemicals were introduced to the surface at 150 K.
Figure 6. CO2 production in the self-coupling of methanol on a) O/Ag (111)
(qO =0.1 ML), O/Ag0.4Au0.6 with initial oxygen coverages of b) 1, c) 0.6, d) 0.4,
e) 0.2, f) 0.1, and g) 0.05 ML, and h) O/Au(111) (qO =0.1 ML). The gold- and
grey-shaded areas represent contributions of CO2 formation from Au-like
and Ag-like sites, respectively.
when it is formed by b-H elimination from methoxy itself.[12] It
is thus possible that b-H elimination on Au-like sites, which
proceeds at lower temperature than on Ag(111), could releas-
ing formaldehyde to react with methoxy adsorbed on Ag-like
or mixed sites. To test this possibility, adsorbed methoxy and
formaldehyde were reacted on Ag(111) by co-adsorbing meth-
anol and formaldehyde on at 0.1 ML oxygen coverage (Fig-
ure 5A). The oxidative coupling reaction between co-adsorbed
methanol and formaldehyde occurred at approximately 260 K
on O/Ag(111), which corresponds to the temperature attribut-
ed to methyl formate formation at mixed sites on the alloy,
supporting this suggestion.
sensitivity for the distribution of Ag and Au within the top
layers of the alloy surface. According to a simple estimate as-
suming only the top two layers of metal atoms are involved in
alloying (procedure is shown in the Supporting Information),
only changes of Ag percentage that are greater than 25% can
be detected by XPS measurements. For example, a change of
Ag percentage from 100 and 0% (in the first and second layer,
respectively) to 75 and 25% due to oxygen adsorption will be
indistinguishable. Therefore, slight surface enrichment of Ag
into the top layer cannot be ruled out based on the current re-
sults.
CO2 also appears to be produced from different types of sur-
face sites. When the initial oxygen coverage on Ag0.4Au0.6 is in-
creased, its formation progresses from a pathway characteristic
of clean Ag(111) to that of clean Au(111), exhibiting formation
via mixed states at intermediate coverage (Figure 6). The low
temperature peak for CO2 at approximately 330 K, appearing
at initial oxygen coverages below 0.4 ML (gold shade, Fig-
ure 6d–f), lines up with that on Au(111), and is therefore attrib-
uted to combustion occurring on Au-like sites. The high-tem-
perature-desorption feature at approximately 405 K at initial
oxygen coverage >0.2 ML is similar to that on Ag(111) (grey
shade, Figure 6b–e) and is attributed to Ag-like site-mediated
combustion. Mixed sites are proposed to be responsible for
the CO2 at approximately 375 K, because the desorption tem-
perature is in between that on the pure Au and Ag surfaces
(unshaded area, Figure 6a–g).
The Au (4f7/2) peak at 85.5 eV and the Ag (3d5/2) peak at
367.2 eV observed upon adsorption of atomic oxygen (qO =
1 ML) on Ag0.4Au0.6 at 200 K indicate the presence of oxidized
Au and Ag species.[22,23] The amount of oxidized Ag is approxi-
mately four times as much as that of Au at 200 K, suggesting
that the surface oxide of Ag is more stable than Au (Table S1
in the Supporting Information). Also, because the oxygen cov-
erage is near saturation, the higher amount of oxidized Ag
suggests some preferential accumulation of Ag into the top-
most layer, assuming that the oxygen cannot diffuse into the
surface at 200 K. The fractions of oxidized Ag and Au grow
from 29 and 7% at 200 K to 45 and 13%, respectively, after an-
nealing the surface to 440 K, indicating further restructuring of
the surface with annealing in the presence of oxygen, though
annealing may allow atomic oxygen to diffuse into the surface
or to form agglomerates of surface oxide.
Discussion
Based on the XPS measurements, no detectable enrichment of
Ag occurs in the in Ag/Au thin film alloy upon the adsorption
of atomic oxygen. However, XPS measurements have limited
Previous experimental and computational studies showed
that atomic oxygen and surface intermediates, such as meth-
oxy, prefer to adsorb on threefold hollow sites,[27] suggesting
&
&
Chem. Eur. J. 2014, 20, 1 – 8
www.chemeurj.org
4
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ÝÝ These are not the final page numbers!

Full Paper
the local environment of the threefold sites could play a critical
role in mediating surface reactions. The distribution of the
three types of surface sites can be semi-quantitatively predict-
ed by a three-atom model, in which Au-like, mixed and Ag-like
sites are defined as having 3, 1–2, and 0 neighboring Au
atoms, respectively. The fraction of surface sites attributed to
Au-like, mixed and Ag-like can be computed statistically as-
suming a random distribution of surface Ag and Au atoms
(Figure 7A, open symbols). The percentage of mixed sites esti-
increases when the Ag molar fraction rises in the alloy, demon-
strating that the stability of methoxy on the mixed sites can
be tuned over a wide range.
In addition, the higher temperature of methyl formate for-
mation from the mixed sites indicates that the mobility of
methoxy adsorbed on the mixed is limited at the temperatures
over which this reaction occurs, consistent with it being more
strongly bound at the mixed sites. Otherwise, methoxy bound
to the mixed sites would migrate to Au-like sites and form
methyl formate at the same temperature as Au-like
sites. In contrast, the weak interaction between form-
aldehyde and both Au[28] and Ag[26] suggests formal-
dehyde formed on Au-like sites migrates to nearby
mixed sites to react with methoxy. It is not until the
formaldehyde from Au-like sites is depleted that the
slower b-H elimination on mixed sites commences.
Additionally, mixed sites are able to directly acti-
vate the b-CÀH bonds in CH3O(a), leading to higher
yield of methyl formate. The fact that D2 is formed
along with [D2]formaldehyde and methyl [D4]formate
(Figure 3) when CD3OH reacts on O/Ag0.4Au0.6, is clear
evidence for the surface-assisted b-CÀD bond activa-
tion. Because hydrogen evolution is never observed
in these reactions on Au(111)[12] but is on
Figure 7. A) Percentages of Au-like, mixed and Ag-like sites (open orange, green, and
blue symbols) as a function of Ag content in the Ag/Au alloy predicted from the three-
atom model. The percentage of mixed sites was obtained by deconvoluting the oxygen-
desorption profiles in Figure S1 (in the Supporting Information; black solid symbols).
B) Comparison of the percentage of methyl formate formed at 210 and 260 K (solid red
and green symbols) with the distribution of surface sites.
Ag(110),[25,26] surface-assisted CÀH bond activation ap-
parently occurs on mixed sites. The increase in yield
for methyl formate with the Ag content in the alloy
could be attributed in part to the extra formaldehyde
formed via surface-assisted b-H elimination of meth-
mated by deconvoluting the oxygen desorption profiles from
the alloy surfaces at constant initial oxygen coverage is also
shown in Figure 7A (qO =1 ML; see also Figure S1 in the Sup-
porting Information), and the agreement is excellent.
oxy. The ability of the Ag-modified surface to facilitate direct b-
H elimination of methoxy hemiacetal to methyl formate and
H2 is also likely to be another pathway leading to the in-
creased yield of methyl formate.
The distribution of the three types of surface sites calculated
from the three-atom model also correlates well with the
amount of methyl formate evolved in the three reaction routes
(Figure 7B). This match indicates the assumption that local
composition of the threefold sites largely determines their re-
activity is a good approximation. Furthermore, only one Ag
atom in three atoms surrounding a threefold site appears to
be sufficient to convert it from Au-like to a mixed site.
Conclusion
Atomic-oxygen-covered Ag/Au thin film alloys selectively medi-
ate oxidative coupling of methanol to methyl formate at low
temperature. Surface sites on an Ag/Au alloy can be classified
into Au-like, Ag-like, and mixed sites, with Au- and Ag-like sites
exhibiting reactivity characteristic of Au(111) and Ag(111), re-
spectively. The kinetics of recombination of adsorbed atomic
oxygen and of the oxidative coupling of methanol from mixed
active sites of Ag/Au reflect the modification of the reactivity
of metallic gold due to stabilization of the atomic oxygen and
methoxy, respectively, by local coordination to Ag. The relative
surface concentrations of the Au-like and mixed sites calculat-
ed from a simple statistical model correlate semi-quantitatively
with the conversion to methyl formate.
The higher evolution temperature of methyl formate from
mixed sites compared to Au-like sites could be due to the en-
hanced stability of methoxy thereon. The rate-determining
step for the oxidative coupling of methanol on O/Au(111) is b-
H elimination of methoxy to form formaldehyde, consistent
with the fact that reacting formaldehyde with pre-adsorbed
methoxy leads to the formation of methyl formate at signifi-
cantly lower temperatures than methanol self-coupling on
both O/Au(111)[7] and Ag(111) (Figure 5A). Therefore, in the ab-
sence of an external source of formaldehyde, the stability of
methoxy with respect to b-H elimination determines the for-
mation temperature of formaldehyde, and in turn, methyl for-
mate. It is likely that the stability of methoxy on mixed sites
lies between that on the pure Au and Ag surfaces, leading to
an intermediate temperature at which formaldehyde, and
hence methyl formate, is formed. This temperature gradually
Experimental Section
All experiments were performed in an ultrahigh-vacuum (UHV)
chamber with a base pressure below 2ꢁ10À10 Torr. The preparation
of the clean Au(111) surface has been described elsewhere.[29] Thin
films of Ag/Au alloys were prepared by physical vapor deposition
(Omicron EFM 3 metal evaporator) of Ag atoms onto Au(111) at
Chem. Eur. J. 2014, 20, 1 – 8
www.chemeurj.org
5
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
&
These are not the final page numbers! ÞÞ

Full Paper
200 K, followed by annealing to 650 K for 10 min to allow the
mixing of Ag and Au atoms.
2:3 and 3:4, respectively, due to the spin–orbit coupling effect. The
Ag (3d), Au (4f), and O (1 s) spectra were accumulated with 10, 10,
and 100 scans, respectively, to enhance the signal-to-noise ratio.
The composition of the alloy film was calculated from the mea-
sured XPS intensities.
The composition of the surface alloy was monitored by calculating
the relative ratio of the integrated area of Ag (3d5/2) and Au (4f7/2
)
peaks (corrected for the atomic sensitivity factors of Ag and Au) in
the X-ray photoelectron spectrum taken immediately after prepara-
tion. The surface compositions obtained by this method are aver-
age values for the first 4–6 atomic layers, estimated from the elec-
tron-escape depths for Ag and Au (10–20 ꢂ) estimated based on
the kinetic energies of the Au(4f) and Ag(3d) photoelectrons creat-
ed with the MgKa radiation used in our experiments. We refer to
the Ag/Au surface alloy as AgxAu1Àx, with x being the molar frac-
tion of Ag in the alloy thin film.
Acknowledgements
We gratefully acknowledge the support of the U.S. Department
of Energy, Basic Energy Sciences, under Grant No. FG02–84-
ER13289 (C.M.F.) and the National Science Foundation, Division
of Chemistry, Analytical and Surface Science (R.J.M.) CHE-
0952790. B.X. also acknowledges the Harvard University Center
for the Environment for support through the Graduate Consor-
tium in Energy and the Environment.
Atomic oxygen was introduced to Au(111) and Ag/Au surface alloy
by ozone exposure at 200 K, because O2 does not dissociate on
the alloy surfaces within the range of composition studied in this
work under the conditions of our experiments (up to 1200 L of O2
dose within the temperature range of 200–400 K). There is no
measurable dissociation of O2 on either pure Ag(111) or Au(111)
under the conditions of our experiments, consistent with the ab-
sence of O2 activation on the alloy films.
Keywords: alloys · gold · mixed sites · oxidative coupling ·
silver
[1] A. K. Sinha, S. Seelan, S. Tsubota, M. Haruta, Top. Catal. 2004, 29, 95–
102.
The oxygen-atom coverage was calibrated by comparing the
amount of O2 evolution in temperature-programmed reaction to
that formed for a saturation coverage of oxygen atoms on clean
Au(111), which is 1.1 ML.[30] Variations in oxygen coverage of
Æ15% are expected due to the day-to-day variation in O3 concen-
tration of the ozone source. Atomic oxygen was deposited on
Ag(111) by exposing the surface to NO2 at 480 K.[31] We refer to the
atomic-oxygen-covered Au(111), Ag(111), and Ag/Au alloy surfaces
as O/Au(111), O/Ag(111), and O/AgxAu1Àx, respectively.
[2] T. Hayashi, K. Tanaka, M. Haruta, J. Catal. 1998, 178, 566–575.
[3] B. Xu, X. Liu, J. Haubrich, R. J. Madix, C. M. Friend, Angew. Chem. 2009,
121, 4270–4273; Angew. Chem. Int. Ed. 2009, 48, 4206–4209.
[4] X. Liu, B. Xu, J. Haubrich, R. J. Madix, C. M. Friend, J. Am. Chem. Soc.
2009, 131, 5757–5759.
[5] H. Miyamura, T. Yasukawa, S. Kobayashi, Green Chem. 2010, 12, 776–
778.
[6] A. Abad, C. Almela, A. Corma, H. Garcia, Chem. Commun. 2006, 3178–
3180.
Temperature-programmed reaction experiments were conducted
according to well-established protocols, which are described in
detail elsewhere.[29] Regardless of the temperature at which reac-
tants were dosed, the surface was cooled to 120 K before collect-
ing reaction data. The heating rate for all temperature-pro-
grammed experiments was nearly constant at 5 KsÀ1 up to the
maximum surface temperature of 600 K, chosen to minimize
changes in the alloy composition at the surface with heating. No
change in the Ag/Au ratio was detected after temperature-pro-
grammed reactions.
[7] B. Xu, X. Liu, J. Haubrich, C. M. Friend, Nat. Chem. 2010, 2, 61–65.
[8] R. J. Angelici, J. Organomet. Chem. 2008, 693, 847–856.
[9] B. Xu, L. Zhou, R. J. Madix, C. M. Friend, Angew. Chem. 2010, 122, 404–
408; Angew. Chem. Int. Ed. 2010, 49, 394–398.
[10] B. Xu, R. J. Madix, C. M. Friend, J. Am. Chem. Soc. 2011, 133, 20738–
20783.
[11] B. Xu, C. M. Friend, R. J. Madix, Faraday Discuss. 2011, 152, 241–252.
[12] B. Xu, C. M. Friend, Faraday Discuss. 2011, 152, 307–320.
[13] A. Wittstock, V. Zielasek, J. Biener, C. M. Friend, M. Baumer, Science
2010, 327, 319–322.
[14] K. Kosuda, A. Wittstock, C. M. Friend, Angew. Chem. 2012, 124, 1730–
1733; Angew. Chem. Int. Ed. 2012, 51, 1698–1701.
[15] M. Turner, V. B. Golovko, O. P. H. Vaughan, P. Abdulkin, A. Berenguer-
Murcia, M. S. Tikhov, B. F. G. Johnson, R. M. Lambert, Nature 2008, 454,
981–983.
[16] C. Lemire, R. Meyer, Angew. Chem. 2004, 116, 121–124; Angew. Chem.
Int. Ed. 2004, 43, 118–121.
[17] M. S. Chen, D. W. Goodman, Science 2004, 306, 252–255.
[18] A. Wittstock, B. Neumann, A. Schaefer, K. Dumbuya, C. Kubel, M. M.
Biener, V. Zielasek, H. P. Steinruck, J. M. Gottfried, J. Biener, A. Hamza, M.
Bꢃumer, J. Phys. Chem. C 2009, 113, 5593–5600.
[19] K. J. Stowers, R. J. Madix, C. M. Friend, J. Catal. 2013, 308, 131–141.
[20] C. T. Campbell, Annu. Rev. Phys. Chem. 1990, 41, 775–837.
[21] J. A. Rodriguez, Surf. Sci. Rep. 1996, 24, 223–287.
[22] A. Krozer, M. Rodahl, J. Vac. Sci. Technol. A 1997, 15, 1704–1709.
[23] G. Hoflund, Z. Hazos, G. Salaita, Phys. Rev. B 2000, 62, 11126–11133.
[24] A. Bzowski, M. Kuhn, T. Sham, J. Rodriguez, J. Hrbek, Phys. Rev. B 1999,
59, 13379–13393.
The reaction products were identified by quantitative mass spec-
trometry (Hiden HAL/3F) by using fragmentation patterns obtained
from authentic samples, which were found to be in good agree-
ment to NIST reference data (see the Supporting Information).[32]
The detailed protocol for quantitative analysis is described in detail
elsewhere.[33]
Methanol and formaldehyde were introduced to surfaces at 150 K
in all experiments. A directed doser with an enhancement factor of
approximately 50 was used. Exposures, calculated from back-
ground pressure and corrected for the enhancement factor, are
given in units of Langmuir [L]; one Langmuir corresponds to the
equivalent of a dose of 1ꢁ10À6 torrsec. These exposures were not
corrected for relative ionization efficiencies in the ion gauge.
Unless otherwise noted, 6 L was the typical dose for organic mole-
cules.
[25] I. E. Wachs, R. J. Madix, Surf. Sci. 1978, 76, 531–558.
[26] I. E. Wachs, J. Catal. 1978, 53, 208–227.
[27] T. A. Baker, C. M. Friend, E. Kaxiras, J. Phys. Chem. C 2009, 113, 3232–
3238.
[28] B. Xu, J. Haubrich, T. A. Baker, E. Kaxiras, C. M. Friend, J. Phys. Chem. C
2011, 115, 3703–3708.
[29] B. K. Min, A. R. Alemozafar, D. Pinnaduwage, X. Deng, C. M. Friend, J.
Phys. Chem. B 2006, 110, 19833–19838.
X-ray photoelectron spectra were acquired with an analyzer pass-
ing energy of 17.9 eV and a multiplier voltage of 3 kV by using
MgKa X-rays (1253.6 eV, 300 W) as the excitation source. The bind-
ing-energy (BE) calibration was referenced to the Au (4f7/2) peak at
83.9 eV. The XPS peaks were fitted after the Shirley background
was subtracted. The relative ratios of the integrated area of Ag
(3d3/2) versus Ag(3d5/2), and Au(4f5/2) versus Au(4f7/2) was fixed to
&
&
Chem. Eur. J. 2014, 20, 1 – 8
www.chemeurj.org
6
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ÝÝ These are not the final page numbers!

Full Paper
[30] N. Saliba, D. H. Parker, B. Koel, Surf. Sci. 1998, 410, 270–282.
[31] S. Bare, K. Griffiths, W. Lennard, H. Tang, Surf. Sci. 1995, 342, 185–198.
[32] S. E. Stein, Linstrom, P. J. , W. G. Mallard, NIST Chemistry WebBook, NIST
Standard Reference Database Number 69 (Eds.: P. J. Linstrom, W. G. Mal-
lard), National Institute of Standards and Technology, Gaithersburg,
p.20899.
[33] B. Xu, R. J. Madix, C. M. Friend, J. Am. Chem. Soc. 2010, 132, 16571–
16580.
Received: December 10, 2013
Published online on && &&, 0000
Chem. Eur. J. 2014, 20, 1 – 8
www.chemeurj.org
7
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
&
These are not the final page numbers! ÞÞ

Full Paper
FULL PAPER
Activation and promotion: Au/Au
mixed sites promotes selective oxida-
tion of methanol to formaldehyde and
methyl formate without producing any
CO2 (see scheme).
&
Bimetallic Catalysis
B. Xu, C. G. F. Siler, R. J. Madix,
C. M. Friend*
&& – &&
Ag/Au Mixed Sites Promote Oxidative
Coupling of Methanol on the Alloy
Surface
&
&
Chem. Eur. J. 2014, 20, 1 – 8
www.chemeurj.org
8
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ÝÝ These are not the final page numbers!