Microscopic View of the Active Sites for Selective Dehydrogenation of Formic Acid on Cu(111)

Article abstract of DOI:10.1021/acscatal.5b01994

Formic acid is an important molecule, due to its potential for hydrogen storage and the role of formate in methanol synthesis. Formic acid can decompose on metals and oxides via dehydrogenation or dehydration, although dehydrogenation is preferred for most applications. These two pathways are linked via the water-gas shift reaction (WGSR), making them hard to separate, and debate over the mechanisms still exists. Cu catalysts are known to selectively decompose formic acid via dehydrogenation to produce CO₂ and H₂. Formic acid's interaction with Cu(110) has been extensively studied, but despite the (111) facet being predominant in many nanoparticles, Cu(111) has received little attention. Using temperature-programmed desorption/reaction (TPD/R) and scanning tunneling microscopy (STM), we have probed key steps in the decomposition of formic acid on Cu(111) at the atomic scale, observing intact adsorption and surface intermediates, as well as the surface after product desorption. Our model system allows us to investigate the reaction under conditions where WGSR is inactive. We find that Cu(111) decomposes formic acid 100% selectively through dehydrogenation. At 85 K, formic acid adsorbs molecularly on Cu(111), forming hydrogen-bonded chains in the β configuration. The acid loses a H atom by 160 K, producing the formate intermediate and surface-bound H atoms, both of which are visualized by STM. All molecules at surface step edges react to formate, but on the Cu(111) terraces desorption of formic acid competes with formate production, which limits formate production to 0.05 monolayer. H atoms formed by O-H bond cleavage recombine to form H₂ in a desorption rate limited process by 360 K. CO₂ and H₂ desorb from the surface in reaction rate limited processes at 400 and 450 K due to formate decomposition on terraces and steps, respectively.

Full text of DOI:10.1021/acscatal.5b01994

Subscriber access provided by CMU Libraries - http://library.cmich.edu
Article
A Microscopic View of the Active Sites for Selective
Dehydrogenation of Formic Acid on Cu(111)
Matthew D Marcinkowski, Colin J Murphy, Melissa L. Liriano,
Natalie A. Wasio, Felicia R. Lucci, and E. Charles H. Sykes
ACS Catal., Just Accepted Manuscript • DOI: 10.1021/acscatal.5b01994 • Publication Date (Web): 30 Oct 2015
Downloaded from http://pubs.acs.org on November 4, 2015
Just Accepted
“Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted
online prior to technical editing, formatting for publication and author proofing. The American Chemical
Society provides “Just Accepted” as a free service to the research community to expedite the
dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts
appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been
fully peer reviewed, but should not be considered the official version of record. They are accessible to all
readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered
to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published
in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just
Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor
changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers
and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors
or consequences arising from the use of information contained in these “Just Accepted” manuscripts.
ACS Catalysis is published by the American Chemical Society. 1155 Sixteenth Street
N.W., Washington, DC 20036
Published by American Chemical Society. Copyright © American Chemical Society.
However, no copyright claim is made to original U.S. Government works, or works
produced by employees of any Commonwealth realm Crown government in the course
of their duties.

Page 1 of 29
ACS Catalysis
1
2
3
4
5
6
7
8
9
A Microscopic View of the Active Sites for Selective
Dehydrogenation of Formic Acid on Cu(111)
Matthew D. Marcinkowski, Colin J. Murphy, Melissa L. Liriano,
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Natalie A. Wasio, Felicia R. Lucci and E. Charles H. Sykes*
Department of Chemistry, Tufts University, 62 Talbot Ave., Medford, Massachusetts 02155,
United States.
Charles.Sykes@tufts.edu
Abstract
Formic acid is an important molecule due to its potential for hydrogen storage and the role of
formate in methanol synthesis. Formic acid can decompose on metals and oxides via dehydrogenation
or dehydration, although dehydrogenation is preferred for most applications. These two pathways are
linked via the water-gas shift reaction (WGSR), making them hard to separate and debate over the
mechanisms still exists. Cu catalysts are known to selectively decompose formic acid via
dehydrogenation to produce CO2 and H2. Formic acid’s interaction with Cu(110) has been extensively
studied, but despite the (111) facet being predominant in many nanoparticles, Cu(111) has received
little attention. Using temperature programmed desorption/reaction (TPD/R) and scanning tunneling
microscopy (STM), we have probed key steps in the decomposition of formic acid on Cu(111) at the
atomic-scale, observing intact adsorption and surface intermediates, as well as the surface after
product desorption. Our model system allows us to investigate the reaction under conditions where
WGSR is inactive. We find that Cu(111) decomposes formic acid 100% selectively through
dehydrogenation. At 85 K, formic acid adsorbs molecularly on Cu(111), forming hydrogen bonded chains
in the β configuration. The acid loses a H atom by 160 K producing the formate intermediate and surface
bound H atoms, both of which are visualized by STM. All molecules at surface step edges react to
formate, but on the Cu(111) terraces desorption of formic acid competes with formate production
which limits formate production to 0.05 monolayers. H atoms formed by deprotonation recombine to
form H2 in a desorption rate limited process by 360 K. CO2 and H2 desorb from the surface in a reaction
rate limited processes at 400 and 450 K due to formate decomposition on terraces and steps
respectively.
ACS Paragon Plus Environment

ACS Catalysis
Page 2 of 29
1
2
3
4
5
6
Keywords: Formic acid, dehydrogenation, Cu catalysts, hydrogen storage molecules,
temperature programmed reaction, scanning tunneling microscopy.
7
8
9
Introduction
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Hydrogen is prevalent in both current and future energy applications since it shows
great promise as an alternative to carbon based sources. Proton exchange membrane fuel cells
(PEMFCs) provide electricity by reacting hydrogen cleanly with oxygen to form H2O. While
hydrogen gas can be fed directly into PEMFCS, the storage and transportation of compressed
hydrogen gas is a potential safety hazard.1–4 In order for hydrogen to be a practical source of
energy, alternative strategies for hydrogen storage are necessary. Methanol has been used as a
hydrogen source in direct methanol fuel cells, but exhibits significant membrane crossover with
Nafion, the typical membrane used in PEMFCs.5–8 As a result, interest in formic acid as a
potential hydrogen storage molecule is increasing. Formic acid can reformed to release
hydrogen3,4,9, or the intact molecule can be fed into a direct formic acid fuel cell (DFAFC).4,5,10,11
While the hydrogen content of formic acid is less than that of methanol, it is less susceptible to
diffusion across the permeable Nafion membrane and thus higher concentrations of formic acid
can be used in a fuel cell.5,6,10 Formic acid can decompose on metal surfaces either through
dehydrogenation to produce CO2 and H2 or dehydration to produce CO and H2O. As hydrogen is
the product of interest for PEMFCs, high selectivity to dehydrogenation is preferred. In
addition, since fuel cells operate at relatively low temperatures, the CO formed via dehydration
can poison the catalyst.10–16 Typically, DFAFCs employ Pt or Pd as the decomposition catalyst,
ACS Paragon Plus Environment

Page 3 of 29
ACS Catalysis
1
2
3
4
5
6
and while these metals display high activity, they are expensive and can produce the undesired
dehydration products.4,5,10–16
7
8
9
Low Miller index Cu surfaces are known to selectively decompose formic acid through
dehydrogenation,17–41 producing the intermediate formate, which is also involved in important
reactions such as methanol synthesis on Cu.42–45 A formate pathway is known to promote high
selectivity to the dehydrogenation.15,46,47 Pt and Pd surfaces also primarily decompose formic
acid to formate,48,49 but unlike Cu, are known to produce some reactively formed CO.12,14,15,50–52
While the decomposition of formic acid to CO2 and H2 has been thoroughly examined on the
Cu(110) surface17–31, the interaction of formic acid on Cu(111) has received very little
attention36–41 , despite the fact that the (111) facet is dominant in many nanoparticles. There
are a few studies in the literature that focus on the formate intermediate on Cu(111), although
formate is often synthesized by the hydrogenation of CO2 rather than formic acid
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
decomposition, as this process is of interest for methanol synthesis.36,37,39 Intact formic acid and
formate have been previously examined on Cu(111) using scanning tunneling microscopy
(STM), x-ray photoelectron spectroscopy, and infrared reflection adsorption spectroscopy, but
to the best of our knowledge, no formic acid temperature programmed desorption/reaction
(TPD/R) work has been performed on Cu(111).37,38,40,41
In this study we examine the adsorption and reaction sites of formic acid on Cu(111)
over a range of temperatures. This allows us to elucidate the energetics of the elementary steps
occurring during the decomposition of formic acid. Using high-resolution STM and TPD/R, each
step of the reaction is observed including the intact adsorption of the formic acid, formation of
ACS Paragon Plus Environment

ACS Catalysis
Page 4 of 29
1
2
3
4
5
6
7
8
9
the formate intermediate, and evolution of reactively formed CO2 and H2 at 400 K on terraces
and 450 K on steps. We show that intact formic acid adsorbs on Cu(111) in the β-polymorphic
configuration at 80 K, and dehydrogenates to formate by 160 K. Our TPR results demonstrate
that defects in the form of step edges on the Cu surface are 100% saturated with formate while
reaction to formate on the Cu(111) terraces is limited to 0.05 monolayers (ML) by competition
with intact desorption of formic acid. While the surface step edges are more reactive than the
terraces, counter intuitively, CO2 formation at steps actually occurs at a higher temperature
than on terraces (450 K vs 400 K). This is because the initial dehydrogenation step to produce
formate is more exothermic on step edges than terraces. Since formate is bound more strongly
to step edges, the second dehydrogenation step to produce CO2 becomes more endothermic,
leading to a higher reaction barrier on step edges.53 Therefore, in catalysis one would expect a
tradeoff between the step edges being initially more reactive to intact HCOOH, but more
susceptible to being blocked and unavailable for further reaction due to the higher
decomposition temperature of the formate intermediates bound there. Importantly, no
restructuring of the Cu surface itself is observed.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Experimental Methods
Experiments were performed in two ultrahigh vacuum (UHV) instruments which have
both been described previously.54 STM experiments were performed using an Omicron
Nanotechnology LT-UHV STM. The base pressure in the chamber was <1 x 10-11 mbar. The
Cu(111) crystal was cleaned by Ar+ sputtering followed by an anneal to 1000 K. The STM stage
ACS Paragon Plus Environment

Page 5 of 29
ACS Catalysis
1
2
3
4
5
6
7
8
9
was held at 5 K during all the experiments described below. Images were taken using tunneling
currents between 30-200 pA and sample biases between 20-200 mV. Formic acid with a purity
of 98% was purchased from Sigma-Aldrich. Liquid formic acid was degassed through freeze-
pump thaw cycles before it was introduced into the chamber. Formic acid vapor was introduced
into the STM chamber through a collimated leak valve with a direct line of sight to the Cu(111)
sample. A preparation chamber attached to the STM was used for sample preparation. A highly
defective Cu surface was prepared by sputtering with Ar+ for 1 min at 1 keV without an anneal
step.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
The second UHV chamber incorporates a Hiden quadrupole mass spectrometer for
performing TPD/R experiments. The base pressure in this chamber is typically <1 x 10-10 mbar. A
clean Cu(111) sample was prepared via Ar+ sputtering and annealing to 750 K. Formic acid was
deposited on the Cu(111) crystal by backfilling the chamber with formic acid to a desired
pressure while the sample was held at 85 K. TPD/R experiments were run using a linear heating
rate of 1 K/s. Molecular formic acid was tracked by both the major contributor to its cracking
pattern m/z 29 and its parent ion m/z 46. Our results shown in this study are from the m/z 29
trace. CO2 and H2 desorption were detected by tracking their parent ions, m/z 44 and m/z 2
respectively. In addition, m/z 28 and m/z 18 were always monitored for detection of CO and
H2O, respectively. A highly defective Cu surface was prepared in this chamber by sputtering
with Ar+ for 10 min at 1.50 keV.
ACS Paragon Plus Environment

ACS Catalysis
Page 6 of 29
1
2
3
4
Results and Discussion
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 1. Reaction of Formic Acid on Cu(111). TPR spectra of CO2 and H2 from decomposition of
on Cu(111) after depositing varying amounts of formic acid. H2 spectra are offset on the y-axis
for clarity. The S and T peaks in CO2 and H2 spectra result from formate decomposition on steps
and terraces, respectively. The D peak in the H2 spectra originates from Ha recombination from
the dehydrogenation of formic acid to formate. Spectra highlighted in red correspond to the
red highlighted peaks in Figure 2 where no intact molecular formic acid is observed desorbing.
CO2 and H2, the sole reaction products from the decomposition of formic acid adsorbed
at 85 K on Cu(111) are shown in Figure 1. Importantly, no high temperature reactively formed
CO or H2O are observed in our TPR spectra, indicating that only dehydrogenation occurs. CO2
and CO can be interconverted by the water-gas shift reaction (WGSR)47,55,56, but we do not
observe any reaction between CO and H2O on a Cu(111) surface allowing us to unambiguously
rule out this effect. Although some signal for m/z 28 is observed, it coincides with m/z 44 and is
consistent with the expected fragmentation pattern for CO2 (Figure S1). Formic acid
ACS Paragon Plus Environment

Page 7 of 29
ACS Catalysis
1
2
3
4
5
6
7
8
9
decomposition on Cu(111) therefore proceeds solely through the dehydrogenation pathway.
This is consistent with TPD results on the more open Cu(110) facet17–19,21,25 and theoretical
predictions for Cu(111).15,53 CO2 has two distinct desorption states exhibiting first-order order
kinetics at 400 K and 450 K. This is in contrast with the single desorption state seen between
450-473 K in previous studies on clean Cu(110) single crystal surfaces. 17–19,21,25 At low
exposures below 0.1 Langmuir (1 L = 1x10-6 torr∙ s), only the peak at 450 K is present. This peak
shifts slightly from 459 K at low coverage to 451 K at saturation, which may be indicative of
repulsive interactions in the densely packed fomate intermediates at step edges. At higher
exposures the peak at 400 K begins to grow, but saturates at a 5 L exposure. Since CO2 does not
typically have stable binding sites on Cu single crystals, even at low temperature, these CO2
desorption peaks are due to reaction rate limited processes.18
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
For H2 desorption, three distinct peaks are observed in our TPD spectra. The lowest
desorption temperature is observed at 360 K at low coverage and the symmetric peak’s
desorption temperature shifts to lower temperature as coverage increases, consistent with
second-order order kinetics. This temperature range is consistent with H2 desorption from
Cu(111)57 and we will show with our high-resolution STM results that the evolution of H2 at
~360 K results from a desorption rate limited process. The two higher temperature H2 peaks
coincide with the reactively formed CO2 at 400 K and 450 K, exhibiting first-order kinetics. This
H2 desorbs at a higher temperature than expected from Cu(111)57 (310–370 K) and therefore
desorbs in a reaction rate limited process. Up to exposures of 0.1 L only the H2 desorption at
360 K and 450 K are observed. Above these exposures the H2 desorption feature at 400 K grows
in the same manner as described for CO2. We assign the desorption rate limited H2 peak at 360
ACS Paragon Plus Environment

ACS Catalysis
Page 8 of 29
1
2
3
4
5
6
7
8
9
K to recombination of adsorbed H atoms resulting from the cleavage of the O-H bond in formic
acid that yields surface bound formate.15,36–41,53 This peak is labeled D for dehydrogenation of
formic acid in Figure 1. We assign the two reaction rate limited desorption states at 400 K and
450 K to formate decomposition on Cu(111) terraces and steps (labeled in Figure 1 as T and S),
respectively. Experimental justification for these assignments will be presented below.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 2. Formic Acid Desorption from Cu(111). (A) Molecular desorption of formic acid from a
Cu(111) surface as measured by TPD after varying exposures of formic acid. Dehydrogenation
products from these experiments are shown in Figure 1. Spectra highlighted in red show low
formic acid coverage, conditions at which no intact formic acid desorbs. Inset TPD traces are a
magnification of the 0.05 L–0.5 L doses. (B) Wide-scale STM image showing ~1 ML of intact
formic acid on terraces and steps annealed to 80 K. Scale bar = 10 nm. (C) STM image showing
formic acid forms well-ordered domains of hydrogen bonded chains in the β conformation. The
system was annealed to 120 K to equilibrate the intact formic acid molecules before STM
imaging occurred. Scale bar = 3 nm.
ACS Paragon Plus Environment

Page 9 of 29
ACS Catalysis
1
2
3
4
5
6
7
8
9
The TPD spectra in Figure 1 and 2 originate from the same experiments and Figure 2A
shows that intact formic acid desorbs at low temperature from Cu(111). At low exposures, no
intact formic acid desorbs and only dehydrogenation products are observed, suggesting all
formic acid has reacted. These low coverage spectra are highlighted in red in Figures 1 and 2.
The inset of Figure 2A gives an expanded view of the five lowest coverages in the main figure
(0.05–0.5 L). At the three lowest exposures (0.05–0.1 L) only the 360 K H2 peak and 450 K CO2
and H2 peaks are present. At higher exposure, there are two desorption states for intact formic
acid at 155 and 170 K both exhibiting zero order desorption kinetics. The peak at 170 K
saturates before the appearance of the 155 K peak. The 155 K state continues to grow for
larger exposures and does not saturate, consistent with desorption from multilayers. The 170 K
state is therefore assigned to desorption from the monolayer of formic acid. Support for these
assignments include Hayden et al. who reported that formic acid multilayers desorb from
Cu(110) between 147–159 K.21 No intact formic acid was observed desorbing from sub-
monolayer coverages on Cu(110).17–19,21,25 A previous study using a combination of XPS, IR and
STM found that on Cu(111) multilayers desorbed between 150–175 K and monolayer
desorption occurred between 175–200 K.41 Interestingly, in our experiments, desorption of
intact formic acid in the TPD spectra ~170 K coincides with the appearance of the
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
dehydrogenation products at 400 K. This result will be discussed further in the context of our
STM results which allow us to assign the active sites for formic acid decomposition on Cu(111).
We also examined the intact formic acid using STM (Figure 2B and C). Figure 2B shows a
wide-scale STM image resulting from ~1 ML of formic acid annealed to 80 K. At this coverage
all the Cu step edges are saturated and there are only a small amount of vacant areas on the
ACS Paragon Plus Environment

ACS Catalysis
Page 10 of 29
1
2
3
4
5
6
7
8
9
terraces. Formic acid adsorbed on terraces appears as bright protrusions and forms well-
ordered domains of molecular chains (Figure 2C). Figure 2C was imaged after a 120 K anneal,
but the surface is essentially unchanged between 80–120 K.41 We believe that these are the
first direct images of hydrogen bonded formic acid chains which have previously been discussed
in relation to IRRAS41 measurements on Cu(111) and are also consistent with the structure of
3D crystalline formic acid.58–64
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 3. Formic Acid Chain Structure on Cu(111). High-resolution STM image of the formic acid
chains after annealing to 80 K. The repeating distance along the chain is 6.7 ± 0.2 Å as
highlighted by the red line. A to-scale representation of formic acid that accounts for the
expected bond lengths and angles in the β conformation is overlaid on the STM image for
comparison. Scale bar = 0.5 nm.
Solid phase formic acid was previously proposed to be polymorphic 60,61 based on far-
infrared spectra which showed examples of peak splitting.59 The β configuration of bulk formic
acid has been shown to be the most stable63,64, but previous results on Cu(111) suggested that
the α configuration is present on the Cu(111) surface.41 These conclusions were based on
ACS Paragon Plus Environment

Page 11 of 29
ACS Catalysis
1
2
3
4
5
6
7
8
9
earlier IR assignments of crystalline formic acid.18–20 Our high-resolution STM results provide
direct, real space images of the structure which is most consistent with the β configuration.
Figure 3 shows a high-resolution STM image of the intact chains of hydrogen bonded formic
acid. We took measurements along the chain length as shown by the red line in Figure 3 and
found an average value of 6.7 ± 0.3 Å. This measurement is in close agreement with the
distance of 6.38 Å from carbon 1 to carbon 3 in β polymorphic formic acid as calculated by a
study using an ab initio crystal orbital method.63 A to-scale representation based on the formic
acid chain dimensions in the β configuration calculated by this study63 is overlaid on our STM
image in Figure 3. The α polymorph of formic acid has an expected carbon 1 to carbon 3
distance of 4.5 Å, well below our measured result. A recent theoretical study used a
computational electron-correlation method and found that the α configuration is unlikely and
the peak splitting observed in the crystallographic IR study results from in-phase and out-of-
phase vibrations in the β configuration.64 This agrees with the explanation initially given by the
original IR study of crystalline formic acid.59
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
ACS Paragon Plus Environment

ACS Catalysis
Page 12 of 29
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 4. Formate and H Atom Intermediates on Cu(111). (A) STM image showing formate
(bright protrusions) and Ha (dark depressions) on the Cu(111) terrace after annealing the sample
to 160 K. Scale bar = 5 nm. (B) Formate saturates step edges. This image was obtained after a
220 K anneal. Scale bar = 2 nm. (C) High-resolution image of a formate cluster on the Cu
terrace. Scale bar = 0.5 nm. (D) High-resolution image of hydrogen adatoms on the Cu terrace.
Scale bar = 1 nm.
ACS Paragon Plus Environment

Page 13 of 29
ACS Catalysis
1
2
3
4
5
6
7
8
9
STM allows us to track the formation of intermediates on the surface that are
inaccessible to TPR. We find that formic acid remains intact on the surface until 160 K, above
which, significant changes are observed by STM. Figure 4 shows STM images that result from
annealing the formic acid covered surface in Figure 2B between 160 and 220 K. As seen in
Figure 4A, the formic acid chains have disappeared from the surface and have been replaced by
clusters of formate (bright protrusions in Figure 4C) on the terraces and steps as well as
adsorbed hydrogen atoms (Ha; dark depressions in Figure 4D). We assign formate as the
intermediate in formic acid decomposition on our Cu(111) surface based on previous results on
Cu(110)18–27,29–31, Cu(100)32–35, Cu(111)36–38,40,41 as well as recent theoretical studies which
predict formate can be produced with a lower barrier than carboxyl on Cu(111).15,53 Formate
was identified as the surface intermediate in early TPD studies on Cu(110) using HCOOD18 and
DCOOH.19 More specifically on Cu(111), formate was identified by comparing the O 1s XPS
spectra of intact formic multilayers of formic acid adsorbed at 130 K to a monolayer of formic
acid annealed to 200 K.40 The monolayer peak was 1.2 eV narrower than the multilayer peak
consistent with the presence of two equivalent oxygen atoms, which is evidence of a formate
rather than a carboxyl pathway. IRRAS further supports the formation of formate on Cu(111), as
the OH and C=O peaks dominating the spectra of multilayers of formic acid disappear and are
replaced with peaks representing OCO stretches upon annealing to 175 K.41 In terms of the
dark depressions seen in the STM images, we have previously studied the quantum tunneling
mediated diffusion of Ha with STM on Pd/Cu(111) single atom alloys and these studies guide our
assignment of the depressions as Ha that result from cleavage of the O-H bond in formic acid to
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
produce formate.54,65–67
.
ACS Paragon Plus Environment

ACS Catalysis
Page 14 of 29
1
2
3
4
5
6
7
8
9
Comparison of Figure 4A with Figures 2B and C shows that heating the sample results in
significant desorption of formic acid from the Cu(111) terraces which competes with
dehydrogenation to formate. On Cu(110), dehydrogenation to formate is complete and intact
formic acid is only observed at multilayer coverages, whereas we observe only ~5% conversion
to formate on the Cu(111) terraces. 17–19,21,25 This can be explained by the Cu atoms in the
Cu(110) surface being more reactive due to their under-coordination with a coordination
number of 7 compared to 9 for Cu(111).68 On average our STM results reveal that formate
coverage is ~0.05 ML after annealing a monolayer of formic acid to 160 K indicating that on
terraces, desorption of intact formic acid is the dominant pathway.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Comparing Figure 4B with Figure 2B shows that the Cu step edges, unlike the terraces,
remained saturated with adsorbates between 85–220 K. Our TPD results showed that at low
coverage (red traces in Figures 1 and 2), no intact formic acid desorbs and the dehydrogenation
products desorb in a single peak at 450 K. This suggests that the CO2 and H2 peaks at 450 K are
due to dehydrogenation occurring at the step edges. We assign the CO2 and H2 peaks at 400 K
to dehydrogenation occurring on terraces as these peaks are only observed in TPR when the
formic acid coverage is high enough to occupy all the step sites and some terrace sites (Figure
1) and STM confirms the presence of formate intermediates on the terraces (Figure 4A).
ACS Paragon Plus Environment

Page 15 of 29
ACS Catalysis
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 5. Structure Sensitivity of Formic Acid Dehydrogenation. (A) Comparison of TPR spectra
of formic acid and decomposition products from a sputtered Cu surface (red) vs. a flat Cu(111)
surface (black). The S and T peaks in CO2 and H2 spectra result from formate decomposition on
steps and terraces, respectively. The D peak in the H2 spectra originates from Ha recombination
from the dehydrogenation of formic acid to formate. The intensity of the CO2 and H2 peaks at
450 K are greatly increased on the sputtered surface compared to the Cu(111) surface. (B) STM
Image of clean Cu(111) compared to (C) sputtered Cu. Scale bars = 5 nm.
In order to confirm these proposed reaction sites, we performed an experiment where
we sputtered the Cu(111) surface with Ar+ without a subsequent anneal. Compared to an
annealed Cu(111) surface (Figure 5B) this treatment results in a surface with a much higher
step density than Cu(111) (Figure 5C). Figure 5A shows a comparison of two TPR experiments:
one resulting from 1 L of formic acid on a Cu(111) surface (black) and the other resulting from 1
L of formic acid on the sputtered Cu surface (red). It can be seen from these two TPD spectra
that the intensity of the CO2 and H2 desorption peaks at 450 K have increased significantly for
the sputtered Cu surface, confirming that they originate from formic acid decomposition at step
edges. In addition, only trace amounts of intact formic acid are observed to desorb from this
ACS Paragon Plus Environment

ACS Catalysis
Page 16 of 29
1
2
3
4
5
6
7
8
9
sputtered Cu surface compared to the flat Cu(111) surface meaning that the highly stepped Cu
surface resulting from sputtering is more reactive than flat Cu(111). One other interesting
observation, is the lower temperature hydrogen peak has shifted to 300 K on the sputtered Cu
surface compared to 350 K on clean Cu(111). This effect is most likely due to the increase in
formate concentration on the sputtered Cu surface. Previous TPD results examining formic acid
decomposition on Cu(110) also observed a 50 K shift in this hydrogen peak and proposed it was
due to repulsive interactions with the surface formate.18,19 It stands to reason that we do not
observe this effect on clean Cu(111) terraces as the formate concentration there is typically
very low (0.05 ML).
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Formate decomposes on Cu(111) steps at 450 K, a temperature consistent with previous
TPR results on Cu(110).17–19,21,25 We therefore propose the active site on defective Cu(111), i.e.
steps edges, is similar to that on Cu(110) terraces. Formate is known to preferentially bind to
the top of rows of Cu(110) in a bidentate configuration.22,28,29 Atoms in Cu(111) step edges and
in the top layer of Cu(110) terraces both have a coordination number of 7 and therefore both
decompose formate at a temperature of 450 K. Comparatively, atoms in Cu(111) terraces have
a coordination of 9, so it is not unreasonable that we see two distinct desorption states on this
surface at 400 and 450 K for terraces and steps respectively. Normally, it would be expected
that the reaction would take place at lower temperature on the more under-coordinated and
hence more reactive site; the step edge. However, we see the opposite trend in our TPD results
shown in Figures 1 and 5. This can be explained by the fact that the decomposition of formic
acid to CO2 and H2 takes place in a two-step process. According to a recent theoretical
investigation53, on under-coordinated surfaces the first reaction step to dehydrogenate formic
ACS Paragon Plus Environment

Page 17 of 29
ACS Catalysis
1
2
3
4
5
6
7
8
9
acid to formate is more exothermic than on close packed surfaces. However, this results in a
more endothermic second reaction (formate decomposition to CO2 and H2) for the under-
coordinated surface. This theoretical study specifically applied this principle to different metal
surfaces, but stated that the principle can most likely be applied to under-coordinated sites on
a surface. We show evidence that this is true when comparing two types of active sites on
Cu(111), namely step edges and terraces.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 6. Reactivity of Formic Acid on Cu(111). Graph comparing desorption of intact formic
acid (black) and reactively formed CO2 (red) as a function of total formic acid surface coverage.
We also examined the extent of formic acid dehydrogenation over a wide range of formic
acid coverages (0.1–9.0 ML). Our TPD results are summarized in Figure 6, which shows the
ACS Paragon Plus Environment

ACS Catalysis
Page 18 of 29
1
2
3
4
5
6
7
8
9
partial pressure of desorbing HCOOH and CO2 as a function of initial coverage of formic acid.
CO2 production saturates at ~1 ML, meaning that multilayers (Figure S2) are not involved in the
reaction.17–19,21,25,41 The production of CO2 in our system is therefore limited by the extent to
which dehydrogenation can compete with desorption. The inset shows a zoom in of the lower
formic acid coverages up to 1.5 ML. As discussed before, at low coverage where only surface
step edges are populated, desorption of intact formic acid is not observed. As the surface
coverage increases and terraces start to become populated, the reaction cannot run to
completion and intact formic acid begins to desorb. We calculated our total CO2 production
from both steps and terraces at saturation based on the step edge density on our Cu(111)
crystal (Figure S3). We find that the maximum CO2 production is 0.04 ML according to our TPR
results, which corresponds well with our calculated formate coverage in STM of 0.05 ML.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
ACS Paragon Plus Environment

Page 19 of 29
ACS Catalysis
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 7. Reaction Pathway for Formic Acid Dehydrogenation on Cu(111). Summary of the key
reaction steps illustrated with STM images and schematics of the corresponding species.
Initially, between 80–160 K intact formic acid forms hydrogen bonded chains on the surface.
Scale bar = 3 nm. At 160 K formic acid deprotonates to yield formate. Scale bar = 5 nm.
Between 400–500 K formate decomposes to yield H2 and CO2, which desorb instantaneously in
a reaction rate limited process leaving a clean Cu surface. Scale bar = 5 nm.
Figure 7 summarizes the reaction steps for formic acid dehydrogenation on Cu(111) as
measured with TPR and STM. Formic acid exists as intact hydrogen bonded chains in the β
configuration up to 160 K (Figure 7 left panel), at which temperature it either desorbs intact as
seen in the TPD monolayer peak in Figure 2 or decomposes to form formate and Ha on the
surface (Figure 7 middle panel). The Cu steps are much more reactive than terraces and no
desorption of intact HCOOH is observed from step edges. Previous work on Cu(111) has
predominately observed formate adsorbing in a bidentate configuration.36–41 One recent study
on Cu(111) reported monodentate formate at lower temperatures, which switched to
ACS Paragon Plus Environment

ACS Catalysis
Page 20 of 29
1
2
3
4
5
6
7
8
9
bidentate formate by 300 K.41 This was predicted by theory on Cu(111)56 and agrees with
results reported on Cu(110).26 We observed no discernible difference of the formate in our STM
images after annealing between 160–350 K. The formate in Figure 7 resulted after annealing to
160 K so it is shown in the monodentate configuration in our schematic in accordance with the
results on Cu(111).41 Based on the IR results from this previous study, it was proposed the
monodentate formate interacts with itself through strong intermolecular interactions. This is
consistent with the formate clustering we observe on Cu(111); the STM image of formate
species shown in Figure 4C suggests that the chain like ordering of the intermediate is mediated
by hydrogen bonding between monodentate species. At 360 K the hydrogen formed from
cleavage of the formic acid O-H bond recombines and desorbs resulting in the TPD peak labeled
D in Figure 1. At 400 K formate on the Cu terraces decomposes to yield CO2 and H2 resulting in
the TPD peak labeled T in Figure 1. Finally, at 450 K formate on the step edges decomposes
resulting in the TPD peak labeled S in Figure 1. This leaves a clean Cu(111) surface, as seen in
the right panel in Figure 7. We confirmed the cleanliness of the Cu surface by running multiple
TPD experiments without cleaning the crystal between runs, which showed no change in
surface reactivity. Furthermore, despite the relatively strong binding of the formate
intermediates we did not observe any signs of Cu surface reconstruction at any point during the
reaction.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Conclusions
Using a combination of TPR and high-resolution STM we have tracked the selective
dehydrogenation of formic acid on Cu(111) from reactants, to intermediates, and products. This
ACS Paragon Plus Environment

Page 21 of 29
ACS Catalysis
1
2
3
4
5
6
7
8
9
reaction has been well studied by catalysis groups, but there is a distinct lack of fundamental
data on the reactivity of the (111) facet of Cu, which is a major termination of most
nanoparticles. Formic acid remains intact below 160 K and forms a β polymorphic structure.
Upon annealing above 160 K, formate and adsorbed hydrogen atoms are formed and visualized
with STM. Annealing above 400 K decomposes the formate further to CO2 and H2. Our results
indicate that the Cu(111) surface is significantly less reactive than the well-studied Cu(110)
facet. 18,19,21,23,25,30,31 However, roughening the Cu(111) surface lead to a significant increase in
reactivity. The lack of Cu surface reconstruction during the reaction indicates that the DFT
results on flat Cu slabs are relevant to the reaction and our results agree well with those
predictions. For example, of all the metal surfaces examined in a recent theoretical study,
Cu(111) was predicted to have one of the lowest turn over frequencies to CO2, but was also
predicted to have extremely low selectivity to CO which is important in terms of catalyst
poisioning.15 Interestingly, this study predicted that Cu3Pt alloys have the greatest potential as
catalysts for formic acid decomposition in terms of both rate and selectivity to CO2 and H2. Our
future work is aimed at investigating the reaction of formic acid on PtCu single atom alloy
surfaces.54
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
ACS Paragon Plus Environment

ACS Catalysis
Page 22 of 29
1
2
3
4
Supporting Information
5
6
7
8
9
Contains additional information about analysis of the CO2 cracking pattern, multilayers of
formic acid on Cu(111), and analysis of the step density of the Cu(111) crystal. This material is
available free of charge via the internet at http://pubs.acs.org.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Corresponding Author
*E-mail: Charles.Sykes@tufts.edu
Notes
The authors declare no competing financial interest.
Acknowledgements
The work was funded by the Division of Chemical Sciences, Office of Basic Energy Sciences,
CPIMS Program, U.S. Department of Energy under Grant No. FG02-10ER16170. M. D. M. thanks
Tufts University Department of Chemistry for an Illumina fellowship. M. L. L. thanks the NSF for
a graduate fellowship.
ACS Paragon Plus Environment

Page 23 of 29
ACS Catalysis
1
2
3
4
References
5
6
7
8
9
(1)
(2)
(3)
(4)
(5)
Schlapbach, L.; Zuttel, A. Nature 2001, 414, 353–358.
Graetz, J. Chem. Soc. Rev. 2009, 38, 73–82.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Joó, F. ChemSusChem 2008, 1, 805–808.
Johnson, T. C.; Morris, D. J.; Wills, M. Chem. Soc. Rev. 2010, 39, 81–88.
Jeong, K. J.; Miesse, C. M.; Choi, J. H.; Lee, J.; Han, J.; Yoon, S. P.; Nam, S. W.; Lim, T. H.;
Lee, T. G. J. Power Sources 2007, 168, 119–125.
(6)
(7)
(8)
(9)
Verbrugge, M. W. J. Electrochem. Soc. 1989, 136, 417–423.
McGrath, K. M.; Surya Prakash, G. K.; Olah, G. A. J. Ind. Eng. Chem. 2004, 10, 1063–1080.
Yu, X.; Pickup, P. G. J. Power Sources 2008, 182, 124–132.
Mellmann, D.; Barsch, E.; Bauer, M.; Grabow, K.; Boddien, A.; Kammer, A.; Sponholz, P.;
Bentrup, U.; Jackstell, R.; Junge, H.; Laurenczy, G.; Ludwig, R.; Beller, M. Chem. Eur. J.
2014, 20, 13589–13602.
(10) Rice, C.; Ha, S.; Masel, R. I.; Waszczuk, P.; Wieckowski, A.; Barnard, T. J. Power Sources
2002, 111, 83–89.
(11) Rice, C.; Ha, S.; Masel, R. I.; Wieckowski, A. J. Power Sources 2003, 115, 229–235.
ACS Paragon Plus Environment

ACS Catalysis
Page 24 of 29
1
2
3
4
(12) Li, Q.; He, R.; Gao, J.-A.; Jensen, J. O.; Bjerrum, N. J. J. Electrochem. Soc. 2003, 150,
5
6
A1599–A1605.
7
8
9
(13) Ha, S.; Larsen, R.; Masel, R. I. J. Power Sources 2005, 144, 28–34.
(14) Yu, X.; Pickup, P. G. Electrochem. Commun. 2009, 11, 2012–2014.
(15) Yoo, J. S.; Abild-Pedersen, F.; Nørskov, J. K.; Studt, F. ACS Catal. 2014, 4, 1226–1233.
(16) Baschuk, J. J.; Li, X. Int. J. Energy Res. 2001, 25, 695–713.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
(17) Wachs, I. E.; Madix, R. J. Surf. Sci. 1979, 84, 375–386.
(18) Ying, D. H. S.; Madix, R. J. J. Catal. 1980, 61, 48–56.
(19) Bowker, M.; Madix, R. J. Surf. Sci. 1981, 102, 542–565.
(20) Hayden, B. E.; Prince, K.; Woodruff, D. P.; Bradshaw, A. M. Vacuum 1983, 33, 876–877.
(21) Hayden, B. E.; Prince, K.; Woodruff, D. P.; Bradshaw, A. M. Surf. Sci. 1983, 133, 589–604.
(22) Crapper, M. D.; Riley, C. E.; Woodruff, D. P. Surf. Sci. 1986, 171, 1–12.
(23) Henn, F. C.; Rodriguez, J. A.; Campbell, C. T. Surf. Sci. 1990, 236, 282–312.
(24) Leibsle, F. M.; Haq, S.; Frederick, B. G.; Bowker, M.; Richardson, N. V. Surf. Sci. 1995, 343,
L1175–L1181.
(25) Bowker, M.; Rowbotham, E.; Leibsle, F. M.; Haq, S. Surf. Sci. 1996, 349, 97–110.
ACS Paragon Plus Environment

Page 25 of 29
ACS Catalysis
1
2
3
4
5
6
(26) Bowker, M.; Haq, S.; Holroyd, R.; Parlett, P. M.; Poulston, S.; Richardson, N. J. Chem. Soc.,
Faraday Trans. 1996, 92, 4683–4686.
7
8
9
(27) Haq, S.; Leibsle, F. M. Surf. Sci. 1997, 375, 81–90.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
(28) Poulston, S.; Bennett, R. A.; Jones, A. H.; Bowker, M. Phys. Rev. B 1997, 55, 888–891.
(29) Bowker, M.; Poulston, S.; Bennett, R. A.; Stone, P.; Jones, A. H.; Haq, S.; Hollins, P. J. Mol.
Catal. A Chem. 1998, 131, 185–197.
(30) Stone, P.; Poulston, S.; Bennett, R. A.; Price, N. J.; Bowker, M. Surf. Sci. 1998, 418, 71–83.
(31) Youngs, T. G. A.; Haq, S.; Bowker, M. Surf. Sci. 2008, 602, 1775–1782.
(32) Crapper, M. D.; Riley, C. E.; Woodruff, D. P. Surf. Sci. 1987, 184, 121–136.
(33) Woodruff, D. P.; McConville, C. F.; Kilcoyne, A. L. D. Surf. Sci. 1988, 201, 228–244.
(34) Krastev, E. T.; Kuhl, D. E.; Tobin, R. G. Surf. Sci. 1997, 387, L1051–L1056.
(35) Sambi, M.; Granozzi, G.; Casarin, M.; Rizzi, G. A.; Vittadini, A.; Caputi, L. S.; Chiarello, G.
1994, 6028, 309–322.
(36) Nakamura, J.; Kushida, Y.; Choi, Y.; Uchijima, T.; Fujitani, T. J. Vac. Sci. Technol. A 1997,
15, 1568–1571.
(37) Fujitani, T.; Choi, Y.; Sano, M.; Kushida, Y.; Nakamura, J. J. Phys. Chem B 2000, 104, 1235–
1240.
ACS Paragon Plus Environment

ACS Catalysis
Page 26 of 29
1
2
3
4
(38) Sotiropoulos, A.; Milligan, P. K.; Cowie, B. C. C.; Kadodwala, M. Surf. Sci. 2000, 444, 52–
5
6
60.
7
8
9
(39) Nakano, H.; Nakamura, I.; Fujitani, T.; Nakamura, J. J. Phys. Chem. B 2001, 105, 1355–
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
1365.
(40) Wuhn, M.; Weckesser, J.; Woll, C. Langmuir 2001, 17, 7605–7612.
(41) Baber, A. E.; Mudiyanselage, K.; Senanayake, S. D.; Beatriz-Vidal, A.; Luck, K. A.; Sykes, E.
C. H.; Liu, P.; Rodriguez, J. A.; Stacchiola, D. J. Phys. Chem 2013, 15, 12291–12298.
(42) Askgaard, T. S.; Norskov, J. K.; Ovesen, C. V.; Stoltze, P. J. Catal. 1995, 156, 229–242.
(43) Grabow, L. C.; Mavrikakis, M. ACS Catal. 2011, 1, 365–384.
(44) Yoshihara, J.; Campbell, C. T. Surf. Sci. 1998, 407, 256–267.
(45) Ludviksson, A.; Zhang, R.; Campbell, C. T.; Griffiths, K. Surf. Sci. 1994, 313, 64–82.
(46) Singh, S.; Li, S.; Carrasquillo-Flores, R.; Alba-Rubio, A. C.; Dumesic, J. A. AIChE J. 2014, 60,
1303–1319.
(47) Grabow, L. C.; Gokhale, A. A.; Evans, S. T.; Dumesic, J. A.; Mavrikakis, M. J. Phys. Chem. C
2008, 112, 4608–4617.
(48) Avery, N. R. Appl. Surf. Sci. 1982, 11-12, 774–783.
(49) Columbia, M. R.; Crabtree, A. M.; Thiel, P. A. J. Am. Chem. Soc. 1992, 114, 1231–1237.
ACS Paragon Plus Environment

Page 27 of 29
ACS Catalysis
1
2
3
4
(50) Sun, S. G.; Clavilier, J.; Bewick, A. J. Electroanal. Chem. 1988, 240, 147–159.
5
6
7
8
(51) Aslam, N. M.; Masdar, M. S.; Kamarudin, S. K.; Daud, W. R. W. APCBEE Procedia 2012, 3,
9
33–39.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
(52) Yu, W.-Y.; Mullen, G. M.; Flaherty, D. W.; Mullins, C. B. J. Am. Chem. Soc. 2014, 136,
11070–11078.
(53) Herron, J. A.; Scaranto, J.; Ferrin, P.; Li, S.; Mavrikakis, M. ACS Catal. 2014, 4, 4434–4445.
(54) Kyriakou, G.; Boucher, M. B.; Jewell, A. D.; Lewis, E. A.; Lawton, T. J.; Baber, A. E.; Tierney,
H. L.; Flytzani-Stephanopoulos, M.; Sykes, E. C. H. Science. 2012, 335, 1209–1212.
(55) Jakdetchai, O.; Nakajima, T. J. Mol. Struct. 2002, 619, 51–58.
(56) Gokhale, A. A.; Dumesic, J. A.; Mavrikakis, M. J. Am. Chem. Soc. 2008, 130, 1402–1414.
(57) Anger, G.; Winkler, A.; Rendulic, K. D. Surf. Sci. 1989, 220, 1–17.
(58) Holtzberg, F.; Post, B.; Fankuchen, I. Acta Cryst. 1952, 6, 127–130.
(59) Millikan, R. C.; Pitzer, K. S. J. Am. Chem. Soc. 1958, 80, 3515–3521.
(60) Mikawa, Y.; Jakobsen, R. J.; Brasch, J. W. J. Chem. Phys. 1966, 45, 4750.
(61) Mikawa, Y.; Brasch, J. W.; Jakobsen, R. J. J. Mol. Spectrosc. 1967, 24, 314–329.
(62) Nahringbauer, I. Acta Cryst. 1978, B34, 315–318.
ACS Paragon Plus Environment

ACS Catalysis
Page 28 of 29
1
2
3
4
(63) Karpfen, A. Chem. Phys. 1984, 88, 415–423.
5
6
7
8
(64) Hirata, S. J. Chem. Phys. 2008, 129, 204104.
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
(65) Tierney, H. L.; Baber, A. E.; Kitchin, J. R.; Sykes, E. C. H. Phys. Rev. Lett. 2009, 103, 246102.
(66) Marcinkowski, M. D.; Jewell, A. D.; Stamatakis, M.; Boucher, M. B.; Lewis, E. A.; Murphy,
C. J.; Kyriakou, G.; Sykes, E. C. H. Nat. Mater. 2013, 12, 523–528.
(67) Jewell, A. D.; Peng, G.; Mattera, M. F. G.; Lewis, E. A.; Murphy, C. J.; Kyriakou, G.;
Mavrikakis, M.; Sykes, E. C. H. ACS Nano 2012, 6, 10115–10121.
(68) Michel, A. C.; Lianos, L.; Rousset, J. L.; Deliche, P.; Prakash, N. S.; Massardier, J.; Jugnet,
Y.; Bertolini, J. C. Surf. Sci. 1998, 416, 288–294.
ACS Paragon Plus Environment

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