Understanding the Role of M/Pt(111) (M = Fe, Co, Ni, Cu) Bimetallic Surfaces for Selective Hydrodeoxygenation of Furfural

Article abstract of DOI:10.1021/acscatal.7b01682

Selectively cleaving the C=O bond of the aldehyde group in furfural is critical for converting this biomass-derived platform chemical to an important biofuel molecule, 2-methylfuran. This work combined density functional theory (DFT) calculations and temperature-programmed desorption (TPD) and high-resolution electron energy loss spectroscopy (HREELS) measurements to investigate the hydrodeoxygenation (HDO) activity of furfural on bimetallic surfaces prepared by modifying Pt(111) with 3d transition metals (Cu, Ni, Fe, and Co). The stronger binding energy of furfural and higher tilted degree of the furan ring on the Co-terminated bimetallic surface resulted in a higher activity for furfural HDO to produce 2-methylfuran in comparison to that on either Pt(111) or Pt-terminated PtCoPt(111). The 3d-terminated bimetallic surfaces with strongly oxophilic 3d metals (Co and Fe) showed higher 2-methylfuran yield in comparison to those surfaces modified with weakly oxophilic 3d metals (Cu and Ni). The effect of oxygen on the HDO selectivity was also investigated on oxygen-modified bimetallic surfaces, revealing that the presence of surface oxygen resulted in a decrease in 2-methylfuran yield. The combined theoretical and experimental results presented here should provide useful guidance for designing Pt-based bimetallic HDO catalysts.

Full text of DOI:10.1021/acscatal.7b01682

Subscriber access provided by GRIFFITH UNIVERSITY
Article
Understanding the Role of M/Pt(111) (M = Fe, Co, Ni, Cu)
Bimetallic Surfaces for Selective Hydrodeoxygenation of Furfural
Zhifeng Jiang, Weiming Wan, Zhexi Lin, Jimin Xie, and Jingguang G Chen
ACS Catal., Just Accepted Manuscript • DOI: 10.1021/acscatal.7b01682 • Publication Date (Web): 24 Jul 2017
Downloaded from http://pubs.acs.org on July 26, 2017
Just Accepted
“Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted
online prior to technical editing, formatting for publication and author proofing. The American Chemical
Society provides “Just Accepted” as a free service to the research community to expedite the
dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts
appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been
fully peer reviewed, but should not be considered the official version of record. They are accessible to all
readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered
to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published
in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just
Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor
changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers
and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors
or consequences arising from the use of information contained in these “Just Accepted” manuscripts.
ACS Catalysis is published by the American Chemical Society. 1155 Sixteenth Street
N.W., Washington, DC 20036
Published by American Chemical Society. Copyright © American Chemical Society.
However, no copyright claim is made to original U.S. Government works, or works
produced by employees of any Commonwealth realm Crown government in the course
of their duties.

Page 1 of 23
ACS Catalysis
1
2
3
4
5
6
7
8
Understanding the Role of M/Pt(111) (M = Fe, Co, Ni, Cu) Bimetallic
Surfaces for Selective Hydrodeoxygenation of Furfural
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
a,b
b
Zhifeng Jiang,¶ Weiming Wan,¶ Zhexi Lin,b Jimin Xie,a and Jingguang G. Chenb,c*
aInstitute for Energy Research, School of Chemistry and Chemical Engineering, Jiangsu University,
Zhenjiang, Jiangsu, 212013, P. R. China
bDepartment of Chemical Engineering, Columbia University, New York, NY, 10027, USA
cChemistry Department, Brookhaven National Laboratory, Upton, NY 11973, USA
Corresponding author: jgchen@columbia.edu (J. G. Chen)
¶ These authors contributed equally to this work.
ACS Paragon Plus Environment
1

ACS Catalysis
Page 2 of 23
1
2
Abstract
3
4
5
6
7
8
Selectively cleaving the C=O bond of the aldehyde group in furfural is critical for converting this
biomassꢀderived platform chemical to an important biofuel molecule, 2ꢀmethylfuran. This work
combined density functional theory (DFT) calculations, temperature programmed desorption (TPD)
and highꢀresolution electron energy loss spectroscopy (HREELS) measurements to investigate the
hydrodeoxygenation (HDO) activity of furfural on bimetallic surfaces prepared by modifying
Pt(111) with 3d transition metals (Cu, Ni, Fe and Co). The stronger binding energy of furfural and
higher tilted degree of the furan ring on the Coꢀterminated bimetallic surface resulted in a higher
activity for furfural HDO to produce 2ꢀmethylfuran than that on either Pt(111) or Ptꢀterminated
PtCoPt(111). The 3dꢀterminated bimetallic surfaces with strongly oxophilic 3d metals (Co and Fe)
showed higher 2ꢀmethylfuran yield than those surfaces modified with weakly oxophilic 3d metals
(Cu and Ni). The effect of oxygen on the HDO selectivity was also investigated on
oxygenꢀmodified bimetallic surfaces, revealing that the presence of surface oxygen resulted in a
decrease in 2ꢀmethylfuran yield. The combined theoretical and experimental results presented here
should provide useful guidance for designing Ptꢀbased bimetallic HDO catalysts.
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
Keywords: furfural, hydrodeoxygenation, 2ꢀmethylfuran, bimetallic surfaces, oxophilicity
ACS Paragon Plus Environment
2

Page 3 of 23
ACS Catalysis
1
2
1. Introduction
3
4
5
6
7
8
Biomass represents sustainable carbon sources for the production of transportation fuels and
chemicals.1 Furfural, one of the most promising platform molecules that is derived from
hemicellulose via the hydrolysis and dehydration of xylose, has received significant attention
because it offers the advantages of being widely available, renewable, and potentially
carbonꢀneutral.2 The selective removal of the oxygen atom in the carbonyl group is a major barrier
in processing furfural for fuel applications.3 An ideal hydrodeoxygenation (HDO) catalyst for
furfural should selectively cleave the carbonyl C=O bond while keeping the CꢀO/CꢀC/C=C bonds of
the furan ring intact. The desirable HDO product, 2ꢀmethylfuran (2ꢀMF), has shown promise as a
fuel additive in a practical road trial (high octane number, RON ~131, energy density and low water
solubility).4
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
Bimetallic catalysts often exhibit unique properties that differ from those of their parent metals and
offer the opportunity to design catalysts with improved activity and selectivity that exceed those
achievable with a monometallic catalyst.5 For example, Resasco et al.6 have reported that the
aromatic ring hydrogenation of anisole and guaiacol was suppressed when Pt was modified with Sn.
Similar result has also been reported by Vohs et al.7 that alloying Pt with Zn could improve the
selectivity for the conversion of furfural to the desired HDO product of 2ꢀMF, while the
decarbonylation products, such as furan and CO, were suppressed.
Although traditional precious metal catalysts, such as Pd8 and Pt9, exhibit relatively high activity,
the undesirable side reactions reduce the HDO selectivity to produce 2ꢀMF. To date, a variety of
bimetallic catalysts with different structures were reported for the HDO of furanic molecules.10
Gorte et al.10a studied the HDO of 5ꢀhydroxymethylfurfural over PtCo nanocrystals and found that
the Pt3Co2 (with Ptꢀrich core and a monolayer Co oxide shell) catalyst showed the highest
2,5ꢀdimethylfuran yield of 98%, which can be attributed to the oxide monolayer that interacts
weakly with the furan ring to prevent side reactions. Huang10b and coꢀworkers reported that the
CuPd alloys showed advantages over the monometallic catalyst in the hydrogenation of furfural to
ACS Paragon Plus Environment
3

ACS Catalysis
Page 4 of 23
1
2
3
4
5
6
7
8
9
2ꢀMF and 2ꢀmethyltetrahydrofuran (2ꢀMTHF). Yu et al.10c prepared a FeNi bimetallic structure and
showed that the ML FeNi(111) bimetallic surface had a higher 2ꢀMF yeild toward furfural HDO. In
addition, studies concerning the oxidation state of catalysts were widely investigated for the HDO
of furanic molecules.11 The work by Zhang et al.11a has shown that compared with Pd/SiO2, the
incorporation of FeOx species into the Pd/SiO2 significantly enhanced the HDO activity of furan
compounds, which was attributed to the generation of PdꢀFe alloy and partially reduced FeOx
species in the PdꢀFeOx/SiO2 system. Vlachos et al.11b have studied the reaction of
5ꢀhydroxymethylfurfural over RuꢀRu2O/C catalysts and shown that the selective production of 2,
5ꢀdimethylfuran could reach a 72% yield by generating a partially oxidized Ru catalyst. Hronec et
al.11c reported that the formation of partially oxidized Cu+ in the PdꢀCu/C system favored the
selective rearrangement of furfural to cyclopentanone rather than 2ꢀMF.
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
Considering the uncertainties and/or inconsistencies of the previous work, surface science studies
which can control the bimetallic structure (metalꢀtermination) and oxygen modification were
performed to investigate their effects on the HDO reaction of furfural. To this end, the primary
objective of the current paper is to understand how the modification of the Pt(111) surface by
oxophilic metals alters the HDO selectivity of furfural. The results presented here combine density
functional theory (DFT) calculations and surface science experiments using temperature
programmed desorption (TPD) and highꢀresolution electron energy loss spectroscopy (HREELS) to
study the effect of oxophilic 3d metal/Pt(111) bimetallic surfaces on the adsorption geometry and
selective HDO of furfural. The bimetallic surfaces were prepared by modifying Pt(111) with four
3d transition metals with weak (Cu and Ni) or strong (Co and Fe) oxophilicity. Both the subsurface
sandwich Ptꢀ3dꢀPt(111) structure and the overlayer 3dꢀPtꢀPt(111) sturcture were investigated.
Additionally, the effect of surface oxygen on the HDO of furfural was investigated on
oxygenꢀmodified bimetallic surfaces.
2. Experimental and computational methods
2.1. DFT calculations
ACS Paragon Plus Environment
4

Page 5 of 23
ACS Catalysis
1
2
3
4
5
6
7
8
9
All DFT calculations were performed with the Vienna Ab initio Simulation Package (VASP)
software.12 The core electrons were represented with the PAW formalism,13 while the valence
region was represented with the PBE exchangeꢀcorrelation functional.14,15 Total energies were
minimized selfꢀconsistently using a convergence threshold of 10–6 eV. The nuclear degrees of
freedom were optimized to a force convergence threshold of 0.05 eV Å–1. All slab calculations were
performed using a periodic 4 x 4 unit cell of four metal layers, with six equivalent layers of vacuum
between slabs. The two bottom layers of the Pt(111) slab were frozen, while the top two layers were
allowed to relax to reach the lowest energy configuration. The monolayer 3dꢀPtꢀPt(111) surfaces
were constructed by replacing the top layer of Pt atoms with 3d metal atoms. The subsurface
Ptꢀ3dꢀPt(111) structures were constructed by replacing the second layer of Pt atoms with 3d metal
atoms. The calculations were carried out implenting the spinꢀpolarization. The binding energy of
furfural on each surface was calculated by subtracting the energies of the bare slab and free
molecule from the total energy of the slab with adsorbed furfural.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
2.2. Surface science experiments
2.2.1. Preparation of bimetallic surfaces
Single crystal Pt(111) was purchased from Princeton Scientific Corporation. The Pt(111) crystal has
a purity of 99.999%, a diameter of 8 mm and a thickness of 1.5 mm. The preparation of 3d/Pt(111)
bimetallic structures was performed using identical procedures reported in a previous study.12 The
3dꢀPtꢀPt(111) surface structure was prepared by depositing 1 ML of 3d metal on Pt(111) at 300 K.
The Ptꢀ3dꢀPt(111) subsurface structure was obtained by depositing 1 ML of 3d metal at 600 K, with
the 3d metal diffusing beneath the top layer of Pt(111) at this temperature.12 The 3d metal coverage
was estimated using Auger electron spectroscopy (AES) based on the AES peak intensities of
Fe(646 eV), Co(777 eV), Ni(718 eV), Cu(922 eV) and Pt(922 eV), as described in a previous
study.12 The Pt(111) surface was cleaned by Ne+ sputtering at 300 K, followed with O2 treatment to
remove residual surface carbon, and then annealing at 1050 K. The O2 treatment was repeated until
negligible amount of C was detected by AES.
ACS Paragon Plus Environment
5

ACS Catalysis
Page 6 of 23
1
2
2.2.2. TPD measurements
3
4
5
6
7
8
Furfural and 2ꢀMF, purchased from Sigma Aldrich with purity of 99%, were injected into glass
cylinders and purified using freezeꢀpumpꢀthrow cycles. All gas samples, hydrogen, neon, and
carbon monoxide, were of research purity and used without further purification. The reagents were
dosed into the UHV chamber with a stainless steel dosing tube and the purity was verified by mass
spectrometry. The TPD experiments were performed in a UHV chamber with a base pressure of 4 ×
10ꢀ10 Torr, equipped with AES, a mass spectrometer, a sputter gun and metal sources, as described
previously.16 The metal source was constructed with a high purity metal wire (Alfa Aesar, 99.99%)
wrapped around a tungsten filament and was mounted within a tantalum enclosure. The Pt(111)
crystal was placed at the center of the UHV chamber by directly spotꢀwelding to two tantalum posts,
allowing resistive heating and cooling with liquid nitrogen.10c For the TPD experiments 4 L (1 L =
1x10ꢀ6 Torr.s) furfural and 5 L H2 were first dosed into UHV chamber. The Pt(111) or
metalꢀmodified Pt(111) surface was then heated from 110 K to 800 K with a linear rate of 3 K/s and
the products desorbing from the surface were detected using a mass spectrometer.
2.2.3. HREELS measurements
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
The HREELS experiments on the Pt(111), PtCoPt(111) and CoPtPt(111) surfaces were performed
in a separate UHV chamber with a base pressure of 1×10ꢀ10 Torr. Each molecule was dosed to
asꢀprepared surfaces at 100 K. During the HREELS measurements, the surfaces were heated to
higher temperatures at 3 K/s and cooled down to 100 K before each HREEL spectrum was recorded.
The prime beam energy was 6 eV and the angles of the incidence and reflection beam were 60o with
respect to the surface normal. The count rate of the elastic peak was typically 2 × 105 counts/s and
the resolution was 50 cmꢀ1 full width at half maximum.
3. Results and discussion
3.1. Reaction of furfural on Co/Pt(111) surfaces
3.1.1. DFT calculations of furfural on Co/Pt(111) surfaces
ACS Paragon Plus Environment
6

Page 7 of 23
ACS Catalysis
1
2
3
4
5
6
7
8
DFT calculations of furfural were performed on Pt(111), PtCoPt(111) and CoPtPt(111) surfaces.
The top and side views of the optimized configurations of furfural on the three surfaces are shown
in Figure 1(A). The numerical assignment of the carbon and oxygen atoms in furfural is shown in
9
Figure 1(B). The correlation between the dꢀband center and the binding energy of furfural on each
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
surface is presented in Figure 1(C). The binding energy values of furfural are predicted to increase
as the surface dꢀband center moves closer to the Fermi level, in agreement with trends observed for
other adsorbates in previous reports.12, 17 The Coꢀterminated bimetallic surface shows the highest
binding energy of furfural among the three surfaces.
ACS Paragon Plus Environment
7

ACS Catalysis
Page 8 of 23
1
2
3
4
5
6
7
8
9
Figure 1. (A) Top and side views of DFTꢀoptimized structures of furfural on Pt(111), PtCoPt(111) and
CoPtPt(111) surfaces (Pt: fuchsia, Co: kelly, C: gray, O: red, H: white); (B) the assignment of the carbon and
oxygen atoms in furfural; (C) the correlation between the dꢀband center and the binding energy of furfural on each
surface; (DꢀF) DFT calculated binding energies of reaction intermediates of furfural on the Co/Pt(111) surfaces (F
= 2ꢀfuranyl)
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Table 1 Comparison of bond lengths, bond length ratio and binding energies of furfural adsorbed on the
corresponding surfaces
Pt(111)
3.444
2.860
0.830
ꢀ1.564
PtCoPt(111)
3.093
CoPtPt(111)
1.903
O1ꢀM (Å)
O2ꢀM (Å)
3.302
1.068
ꢀ0.996
2.448
1.286
ꢀ2.926
d
(O2ꢀM)/
d
(O1ꢀM)
Binding energy (eV)
Table 1 summarizes binding energies (eV) of adsorbed furfural and the tilted degree of the furan
ring on the three surfaces. The binding energies of furfural show a trend of CoPtPt(111) > Pt(111) >
PtCoPt(111). The bond distances of O1ꢀM and O2ꢀM, and the d(O2ꢀM)/d(O1ꢀM) ratio are compared
to investigate the extent of the furan ring tilting away from the surface. The distances of O1ꢀM and
O2ꢀM on Coꢀterminated CoPtPt(111) are smaller than those on Pt(111) and PtCoPt(111), indicating
that furfural has stronger interaction with the Coꢀterminated bimetallic surface than the other two,
consistent with the higher furfural binding energy on CoPtPt(111). The d(O2ꢀM)/d(O1ꢀM) ratio of
furfural shows a trend of CoPtPt(111) > PtCoPt(111) > Pt(111), suggesting that the addition of 1
ML Co on Pt(111) makes the furan ring of furfural tilted further away from the surface.
In order to predict the reaction pathways of furfural on Co/Pt(111) surfaces, the reaction energies of
five possible reaction intermediates were calculated. As shown in Figure 1DꢀF, possible
intermediates from furfural include FꢀCHOH (orange line, F = 2ꢀfuranyl) or FꢀCH2O (black line)
through hydrogenation, FꢀCH (red line) by C=O bond cleavage, furanyl (purple line) through CꢀC
bond scisson and FCO (blue line) by dehydrogenation. The Bronsted–Evans–Polanyi principle
suggests that a more negative reaction energy should in general lead to a lower activation barrier.18
On the Pt(111) surface (Figure 1D), the dehydrogenation of furfural is perferred as the first step and
all the following steps in the blue curve are exothermic, suggesting that FꢀCO would break the CꢀC
bond to form CO and furanyl, which would be further hydrogenated to form furan. Figure 1E shows
that on PtCoPt(111) furfural can be converted to FꢀCHOH in the first step through hydrogenation.
ACS Paragon Plus Environment
8

Page 9 of 23
ACS Catalysis
1
2
3
4
5
6
7
8
9
However, it is unfavorable to break the CꢀO bond of FꢀCHOH to form FꢀCH, suggesting that the
PtCoPt(111) surface should have a low HDO activity. Since furfural has a low binding energy on
PtCoPt(111), the molecular desorption of furfural would likely occur before undergoing subsequent
surface reactions. On the CoPtPt(111) surface (Figure 1F), furfural prefers to be hydrogenated to
form FꢀCH2O in the first step. It is also favorable to break the CꢀO bond of FꢀCH2O to form FꢀCH2
and surface oxygen, with FꢀCH2 being further hydrogenated to form 2ꢀMF.
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
In summary, DFT calculations provide two parameters to predict the surface activity: the
binding energy of furfural and the tilted degree of the furan ring. The surface with high furfural
binding energy and a tilted furan ring, CoPtPt(111), prefers to produce 2ꢀMF. The surface with high
furfural binding energy and a flat furan ring, Pt(111), prefers to produce furan. Finally, the surface
with low binding energy, PtCoPt(111), should have a low activity for furfural conversion.
3.1.2. TPD experiments of furfural on Co/Pt(111) surfaces
TPD experiments were performed to investigate the products from Pt(111), PtCoPt(111) and
CoPtPt(111) surfaces after exposure to 5 L H2, corresponding to half of the saturated H coverage,
and 4 L furfural, corresponding to a saturation of chemisorbed furfural. Some of the possible
reaction pathways of furfural are listed as follows:
Reaction (1) represents the desired HDO reaction as it produces 2ꢀMF, reaction (2) is the undesired
decarbonylation pathway involving the C1ꢀC2 bond scission to produce furan and CO, and reaction
(3) includes both HDO and the hydrogenation of the furan ring to produce 2ꢀMTHF and Oad. The
stoichiometric coefficients in the above equations, a, b and c, represent the amount of chemisorbed
furfural (molecule per metal atom) to produce 2ꢀMF, furan and 2ꢀMTHF, respectively. The yields
ACS Paragon Plus Environment
9

ACS Catalysis
Page 10 of 23
1
2
3
4
5
of the aboveꢀmentioned gasꢀphase products are estimated from the TPD peak areas of the cracking
patterns of 2ꢀMF (m/e = 82), furan (m/e = 68) and 2ꢀMTHF (m/e = 71), as shown in Figure 2.
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 2. TPD spectra of (A) 2ꢀMF (m/e = 82), (B) furan (m/e = 68) and (C) 2ꢀMTHF (m/e = 71) with an
exposure of 4 L furfural on hydrogen preꢀdosed Pt(111), PtCoPt(111) and CoPtPt(111) surfaces.
The molecular desorption of furfural from the asꢀprepared surfaces occurs at about 200 K (spectra
not shown). As can be observed from Figure 2A, the desorption peak of 2ꢀMF is observed at 310 K
from CoPtPt(111), while only trace amount of 2ꢀMF is observed from PtCoPt(111) and 2ꢀMF is not
detected from the Pt(111)4d surface. Figure 2B shows the production of furan from the
decarbonylation pathway. The Pt(111) surface produces the largest amount of furan among the three
surfaces, which is consistent with the conclusion from the Kyriakou group.4d Figure 2C indicates
that 2ꢀMTHF is not produced from the three surfaces under UHV conditions. In addition to the
furanꢀring containing products, ethylene was also detected as a reaction product as shown in Figure
S1 and Table S1 in the Supporting Information. To quantify the amount of furfural undergoing each
of the aboveꢀmentioned reaction pathways, the yields of 2ꢀMF, furan and 2ꢀMTHF were calculated
from an experimentally determined sensitivity factor relative to CO by comparing the mass
spectrometer sensitivity factors at equal concentrations of 2ꢀMF, furan,2ꢀMTHF and CO. According
to the reported literature,19a the saturation coverage of CO on Pt(111) is 0.68 ML (θCsOat = 0.68) and
this value is used for the quantification. Detailed procedures for quantifying the TPD products were
ACS Paragon Plus Environment
10

Page 11 of 23
ACS Catalysis
1
2
3
4
5
described in a previous work.19b Table 2 summarizes the activity and selectivity of furfural on
Pt(111), PtCoPt(111) and CoPtPt(111) surfaces.
6
7
8
9
The 2ꢀMF yield follows the trend of CoPtPt(111) > PtCoPt(111) ≈ Pt(111) and the furan yield
follows the trend of Pt(111) > PtCoPt(111) > CoPtPt(111). The TPD results are consistent with the
DFT predictions (Figure 1) that CoPtPt(111) is active toward 2ꢀMF formation, Pt(111) is active to
form furan, and PtCoPt(111) has the lowest activity for furfural decomposition.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Table 2 Quantification of furfural on hydrogen preꢀdosed Co/Pt(111) surfaces from TPD measurements
Activity (monolayer per metal atom)
Selectivity of furanꢀring
containing products (%)
2ꢀMF
0
Furan
0.036
2ꢀMTHF
0
Total
0.036
2ꢀMF
Furan
Pt(111)
0
100
PtCoPt(111)
CoPtPt(111)
0.001
0.012
0.018
0.008
0
0
0.019
0.020
5.3
94.7
40.0
60.0
3.1.3. HREELS measurements of furfural on Co/Pt(111) surfaces
The HREELS experiments were performed to determine the adsorption configuration of
furfural and to identify surface reaction intermediates on Pt(111), PtCoPt(111) and CoPtPt(111).
The vibrational features of furfural observed at 100 K are summarized in Table 3, along with
Raman frequencies of liquid furfural. Figure 3 shows the HREELS spectra of 4 L furfural on
hydrogen preꢀdosed Pt(111), PtCoPt(111) and CoPtPt(111) surfaces at 100 K and the vibrational
frequencies were generally similar with those of the Raman spectrum of liquid furfural, suggesting
that furfural adsorbed molecularly on the three surfaces at 100 K. After the surfaces were heated to
200 K, the intensities of all vibrational modes decreased. The relative intensiy of the ν(CH)ring at
3080 cmꢀ1 decreased compared to that of the ν(CH)aliphatic at 2838 cmꢀ1, indicative of the interaction
between the skeletal furan structure and metal surfaces. Meanwhile, the intensity of the ν(C=O)
mode at 1643 cmꢀ1 decreased and shifted to a lower frequency, resulting from the interaction
between the carbonyl group and the metal surfaces to form the η2(C,O)ꢀfurfural intermediate, as
predicted from the DFT results.
ACS Paragon Plus Environment
11

ACS Catalysis
Page 12 of 23
1
2
3
4
5
6
7
8
A comparison of the HREELS spectra at 200 K provides information about the adsorption
configuration of furfural on the three surfaces. Vibrational peaks at 1440 cmꢀ1 and 1643 cmꢀ1
represent ν(C=C) in the furan ring and ν(C=O) in the carbonyl group, respectively. The ratio of
9
these two peaks, I(1440)/I(1643) (defined as
regarding the relative orientation of the furan ring in furfural with respect to the carbonyl group. As
compared in Table 4, on CoPtPt(111) showed the highest value, indicating that the furan ring in
α) could be used to provide a qualitative indicator
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
α
adsorbed furfural is tilted to the largest extent on CoPtPt(111) compared with that of Pt(111) and
PtCoPt(111), in agreement with the DFT prediction.
Table 3. Vibrational assignment of furfural
Mode
Frequency(cmꢀ1)
Ramana
Pt(111)b
PtCoPt (111)c
CoPtPt(111)d
τ (ring)
ω (CH)
δ (ring)
ν (CO)
ꢀꢀ
ꢀꢀ
ꢀꢀ
930
602
765
ꢀꢀ
602
765
ꢀꢀ
605
760
ꢀꢀ
ꢀꢀ
ꢀꢀ
ꢀꢀ
χ (CH)
950
ꢀꢀ
ꢀꢀ
ꢀꢀ
ν (CO)
δb (OꢀCꢀH)
ρ (CH)
δb (OꢀCꢀH)
ν (CC)
1025
1157
ꢀꢀ
1370
1023
1131
1223
ꢀꢀ
1016
1141
1202
ꢀꢀ
1018
1133
1216
ꢀꢀ
1395
ꢀꢀ
ꢀꢀ
ꢀꢀ
ν (C=C)
ν (C=O)
ν (CH)
1466, 1555, 1570
1445
1650
2832
3087
1440
1654
2827
3080
1437
1643
2837
3090
1684
ꢀꢀ
3153
ν (CH)
τ ꢀ torsion, ω ꢀ wagging, δ ꢀ deformation, ν ꢀ symmetric stretching, χ ꢀ scissoring, δb ꢀ bending, ρ ꢀ rocking
a Ref. [10c]
b 4 L furfural on Pt(111) at 100 K
c 4 L furfural on PtCoPt(111) at 100 K
d 4 L furfural on CoPtPt(111) at 100 K
ACS Paragon Plus Environment
12

Page 13 of 23
ACS Catalysis
1
2
3
4
5
6
765
602
3090
2838
1643
7
8
9
1018
1440
CoPtPt(111)
200 K
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
CoPtPt(111)
PtCoPt(111)
100 K
200 K
PtCoPt(111)
Pt(111)
100 K
200 K
100 K
Pt(111)
0
1000
2000
3000
-1
Wavelength (cm
)
Figure 3. HREELS of 4 L furfural on H2 preꢀdosed Pt(111), PtCoPt(111) and CoPtPt(111) surfaces at 100 Kand 200 K.
Table 4. Intensity ratio of I(1440)/I(1643) in HREEL spectra of furfural adsorption on H2 preꢀdosed Pt(111),
PtCoPt(111) and CoPtPt(111) surfaces
Surface
Pt(111)
PtCoPt(111)
CoPtPt(111)
Temperature (K)
I(1440)/I(1643)
200
200
200
0.9
1.2
2.8
The onꢀ and offꢀspecular HREELS measurements were further carried out to determine the different
adsorption configurations of furfural on the three surfaces. Based on the surface dipole selection
rule, virbrational modes with component of dipole moment perpendicular to the surface would have
a strong signal in the onꢀspecular spectra. Thus, the virbrational modes that are more tilted away
from the surface should show higher intensity in the onꢀspecular spectra and relatively low intensity
in the offꢀspecular spectra. The onꢀ and offꢀspecular HREELS spectra of 4 L furfural at 200 K on
hydrogen preꢀdosed Pt(111), PtCoPt(111) and CoPtPt(111) surfaces are presented in Figure 4. The
vibrational peaks at 765 cmꢀ1 and 1643 cmꢀ1 represent the ω(CꢀH) mode of the furan ring and the
ν(C=O) mode of the carbonyl group, respectively. The ratio of these two modes, (I(765)/I(1643)), is
ACS Paragon Plus Environment
13

ACS Catalysis
Page 14 of 23
1
2
3
4
5
6
7
8
9
defined as β and this value is used to compare the relative tilted degree of the furan ring and
carbonyl group. As shown in Table 5, the β values of the onꢀspecular spectra for Pt(111),
PtCoPt(111) and CoPtPt(111) follow a trend of 10.7 > 8.4 > 2.6, suggesting that the furan ring in
furfural adsorbed more parallel to the Pt(111) and PtCoPt(111) surfaces than to CoPtPt(111). In the
offꢀspecular spectra, the β values on Pt(111) and PtCoPt(111) surfaces decreased from 10.7 to 1.8
and 8.4 to 5.2. In contrast, on the CoPtPt(111) surface, β remained similar in the onꢀ and
offꢀspecular measurements from 2.6 to 2.9. The differences in the β values further suggest that the
furan ring in furfural is tilted in a larger angle than the carbonyl group on the CoPtPt(111) surface.
The HREELS analysis provides evidence to support the DFT prediction that the furan ring of
furfural is more tilted on CoPtPt(111) than on Pt(111). The titlted configuration of furfural on the
CoPtPt(111) surface allows a stronger interaction with the carbonyl group than with the furan ring,
leading to the selective HDO through the carbonyl group to produce 2ꢀMF, consistent with the DFT
predictions and the TPD experimental results.
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. On and offꢀspecular HREELS of 4 L furfural on H2 preꢀdosed Pt(111), PtCoPt(111) and CoPtPt(111) surfaces
at 200 K.
ACS Paragon Plus Environment
14

Page 15 of 23
ACS Catalysis
1
2
3
Table 5. Intensity ratio of I(765)/I(1643) in HREELS spectra of furfural adsorption on H2 preꢀdosed Pt(111),
4
5
6
7
8
9
PtCoPt(111) and CoPtPt(111) surfaces
Surface
Pt(111)
Type
I(765)/I(1643)
Onꢀspecular
Offꢀspecular
Onꢀspecular
Offꢀspecular
Onꢀspecular
Offꢀspecular
10.7
1.8
8.4
5.2
2.6
2.9
PtCoPt(111)
CoPtPt(111)
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
3.2. Trend in furfural reaction on M/Pt(111) (M = Cu, Ni, Fe and Co) surfaces
3.2.1. DFT calculations of furfural and oxygen on M/Pt(111) surfaces
In the previous section, two parameters, furfural binding energy and tilted degree of the furan ring,
were used to explain the HDO activity over Co/Pt(111) surfaces. This section will attempt to use
these two parameters to predict the performance of other 3d/Pt(111) surfaces. Since the ML
CoPtPt(111) structure shows better HDO activity and selectivity than PtCoPt(111), only the
Mꢀterminated bimetallic surfaces (M = Cu, Ni, Fe and Co) will be compared here. As summarized
in Table 6, the NiPtPt(111), FePtPt(111) and CoPtPt(111) surfaces show similar binding energies of
furfural, which are much larger than that of CuPtPt(111). This could be attributed to the different
absorbed configuration of furfural on these four surfaces. Furfural absorbs on NiPtPt(111),
FePtPt(111) and CoPtPt(111) in the η2(CꢀO) configuration in which both the C and O of carbonyl
group are bonded to the surface, whereas the adsorption on CuPtPt(111) adopts a top
η1(O)ꢀaldehyde20 binding mode. The ratio of d(O2ꢀM)/d(O1ꢀM) follows the order of CuPtPt(111) >
CoPtPt(111) > FePtPt(111) > NiPtPt(111) (Table 6). Although the furan ring in furfural is tilted in
the largest angle to CuPtPt(111) as compared to the other three surfaces, the HDO activity of
CuPtPt(111) may not be the highest because of the weak binding energy of furfural on the surface.
In addition, as indicated by the oxygen binding energies (Table 6), CoPtPt(111) and FePtPt(111)
surfaces are stronger oxophilic surfaces than NiPtPt(111) and CuPtPt(111). Correspondingly, the
CoPtPt(111) and FePtPt(111) surfaces have the smallest d(O1ꢀM) value, suggesting that these two
oxophilic surfaces have stronger interaction with the carbonyl group of furfural, which should
facilitate the cleavage of the C=O bond and lead to HDO activity. In terms of the degree of the tilted
ACS Paragon Plus Environment
15

ACS Catalysis
Page 16 of 23
1
2
3
ring, the d(O2ꢀM)/d(O1ꢀM) value is slightly lower on FePtPt(111) than the CoPtPt(111) surface,
4
5
6
suggesting that the furan ring is more tilted on CoPtPt(111).
7
8
9
Table 6 Comparison of bond lengths, bond length ratio and binding energies of oxygen and furfural adsorbed on
the corresponding surfaces.
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
CuPtPt(111)
NiPtPt(111)
FePtPt(111) CoPtPt(111)
OBE* (eV)
O1ꢀM (Å)
O2ꢀM (Å)
d(O2ꢀM)/d(O1ꢀM)
Binding energy (eV)
*OBE = Oxygen binding energy
ꢀ6.30
2.049
3.077
1.502
ꢀ1.665
ꢀ7.62
1.930
2.105
1.091
ꢀ3.256
ꢀ8.59
1.901
2.176
1.144
ꢀ3.496
ꢀ7.85
1.903
2.448
1.286
ꢀ2.926
3.2.2 TPD experiments of furfural on M/Pt(111) bimetallic surfaces
As shown in Figure 5, TPD experiments on the four bimetallic surfaces were further carried out to
verify the DFT prediction. The adsorbed molecular furfural desorbed at about 200 K from the four
bimetallic surfaces (spectra not shown). The CoPtPt(111), FePtPt(111) and NiPtPt(111) surfaces
show 2ꢀMF activity while CuPtPt(111) does not show HDO activity (Figure 5A), which could be
attributed to the weak furfural binding energy. Furan desorption occurs between 378 and 393 K
from all the bimetallic surfaces (Figure 5B). No 2ꢀMTHF desorption peak was observed from all the
surfaces (Figure 5C). Among three surfaces with HDO activity, the 2ꢀMF yield shows a trend of
CoPtPt(111) > FePtPt(111) > NiPtPt(111) (Table 7), which is consistent with the trend of the tilted
degree predicted by DFT calculations. In contrast, the furan yield shows the opposite trend of
NiPtPt(111) > FePtPt(111) > CoPtPt(111) (Table 7), which is consistent with the argument that a
less tilted furan ring would facilitate the C1ꢀC2 bond cleavage to produce furan. Likewise, the
reforming product of ethylene is also generated from the CoPtPt(111), FePtPt(111) and NiPtPt(111)
surfaces. From the DFT calculations and TPD results, it is evident that high furfural binding energy
is necessary to achieve HDO activity and the tilted degree of the furan ring affects the selectivity.
Adsorbed furfural with a more tilted furan ring prefers to form 2ꢀMF, while furfural with a less
tilted furan ring leads to decarbonylation to form furan and CO. Therefore, high furfural binding
energy and tilted furan ring are both necessary for selective HDO. In addition, both CoPtPt(111)
and FePtPt(111) show high yield of 2ꢀMF and they are also strongly oxophilic surfaces. Though the
ACS Paragon Plus Environment
16

Page 17 of 23
ACS Catalysis
1
2
3
4
5
6
7
8
oxygen binding energy alone is not sufficient to predict the HDO selectivity, in general oxophilic
surfaces should show higher HDO activity since it binds more strongly with oxygen of the carbonyl
group, facilitating the cleavage of the C=O bond.
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. TPD spectra of (A) 2ꢀMF (m/e = 82), (B) furan (m/e = 68) (C) 2ꢀMTHF (m/e = 71) with an exposure of
4 L furfural on hydrogen preꢀdosed CuPtPt(111), NiPtPt(111), FePtPt(111) and CoPtPt(111) surfaces.
Table 7 Quantification of furfural on hydrogen preꢀdosed MꢀPtꢀPt(111) surfaces from TPD measurements
Activity (monolayer per metal atom)
Selectivity of furanꢀring
containing product (%)
2ꢀMF
Furan
2ꢀMTHF
Total
2ꢀMF
Furan
CuPtPt(111)
NiPtPt(111)
0
0.008
0.024
0
0
0.008
0.028
0
100
0.004
14.3
85.7
FePtPt(111)
CoPtPt(111)
0.010
0.012
0.015
0.008
0
0
0.025
0.020
40.0
60.0
60.0
40.0
3.2.3. Effect of oxygen-modification on M/Pt(111) bimetallic surfaces
In many reactor studies, HDO catalysts were reported to be partially oxidized during the HDO
reactions. For example, Vlachos et al.21 reported Ru/RuOx/C to be active for furfural HDO. Zhang
et al.11a reported Pd/FeOx as an active catalyst for HDO of furan compounds. Gorte et al.10a reported
that the formation of a CoOx overlayer on Pt was responsible for the high activity and selectivity
toward furfural HDO. It remains unclear whether the oxophilicity of the metal or the presence of
ACS Paragon Plus Environment
17

ACS Catalysis
Page 18 of 23
1
2
3
4
5
6
7
8
9
surface oxygen contributes to the HDO activity. Thus, furfural TPD experiments were performed on
clean and oxygenꢀmodified MꢀPtꢀPt(111) (M = Cu, Ni, Fe and Co) surfaces to determine the effect
of surface oxygen on furfural HDO. As shown in Figure 6(A), the 2ꢀMF product was not detected
from CuPtPt(111), 0.1O/CuPtPt(111) and 0.25O/CuPtPt(111) (0.1 and 0.25 represent the atomic
ratios of O/Pt), suggesting that partially oxidizing Cu on CuPtPt(111) surface did not have obvious
influence on the 2ꢀMF yield. As shown in Figure 6(BꢀD) for NiPtPt(111), FePtPt(111), CoPtPt(111),
the 2ꢀMF desorption peak area decreased after the surfaces were modified by oxygen. As the
oxygen amount increased, the desorption peak areas of 2ꢀMF decreased obviously on all three
surfaces, indicating that the presence of oxygen on these bimetallic surface is not in favor of
producing 2ꢀMF. Similarly, the yields of other generated products such as furan and ethylene also
decreased after the surfaces were modified by oxygen (Figure S2). Based on the TPD results, two
general conclusions can be reached: (1) the production of 2ꢀMF is attributed to the overlayer metal
surfaces rather than metal oxideꢀPtꢀPt(111) surfaces; (2) the addition of oxygen likely blocks the
activate sites of the bimetallic surfaces, resulting in a significant decrease in the yield of 2ꢀMF.
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
18

Page 19 of 23
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 6. TPD spectra of 2ꢀMF (m/e = 82) with an exposure of 4 L furfural on hydrogen preꢀdosed
oxygenꢀmodified MꢀPtꢀPt(111) surfaces (M = Cu, Ni, Fe and Co).
4. Conclusions
In summary, combined DFT calculations and surface science experiments were performed to study
the HDO reaction on 3d/Pt(111) bimetallic surfaces using furfural as a probe molecule for
biomassꢀderived oxygenates. The main conclusions are as follows:
(1) For the reaction of furfural on Co/Pt(111) surfaces, the Coꢀterminated CoPtPt(111) surface
shows higher HDO activity for furfural to produce 2ꢀMF than the Ptꢀterminated PtCoPt(111)
surface. The addition of ML Co atoms onto Pt(111) increases the furfural binding energy and
ACS Paragon Plus Environment
19

ACS Catalysis
Page 20 of 23
1
2
3
4
5
6
7
8
9
leads to the furan ring in furfural tilted away from the surface, as predicted by DFT calculations
and confirmed by the onꢀ and offꢀ specular HREELS results. Reaction network DFT calculation
also suggests that the addition of Co atoms can alter the reaction pathway of furfural to form
2ꢀMF.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
(2) For 3dꢀterminated bimetallic surfaces, the 2ꢀMF yield follows the order of CoPtPt(111) >
FePtPt(111) > NiPtPt(111) > CuPtPt(111). The high furfural binding energy is necessary for a
surface to show HDO activity and the tilted furan ring can further improve the selectivity.
CoPtPt(111) has similar furfural binding energy with FePtPt(111) and NiPtPt(111) while
showing the largest tilted degree of furan ring, which facilitates the selective HDO of furfural to
produce 2ꢀMF. In contrast, 2ꢀMF is not observed from CuPtPt(111), which can be attributed to
the weak interaction with furfural.
(3) Based on the comparison of the HDO activity before and after modifying bimetallic surfaces
with oxygen, the production of 2ꢀMF is related to the bimetallic surfaces rather than metal
oxideꢀPtꢀPt(111) surfaces. The addition of surface significantly decreases the yield of 2ꢀMF.
Acknowledgement
This article was based on work supported as part of the Catalysis Center for Energy Innovation
(CCEI), an Energy Frontier Research Center (EFRC) funded by the U.S. Department of Energy,
Office of Basic Energy Sciences under Award Number DEꢀSC0001004. The DFT calculations were
performed using computational resources at the Center for Functional Nanomaterials, a user facility
at BNL, supported by US Department of Energy under Contract No. DEꢀAC02ꢀ05CH11231.
Supporting Information
Supporting Information Available: additional TPD results for ethylene and furan desorption on the
M/Pt(111) and oxygen modified M/Pt(111) surfaces.
This material is available free of charge via the Internet at http://pubs.acs.org.
ACS Paragon Plus Environment
20

Page 21 of 23
ACS Catalysis
1
2
Reference
3
4
5
6
7
8
(1) (a) Mariscal, R.; MairelesꢀTorres, P.; Ojeda, M.; Sádaba, I.; Granados, M. L. Energ. Environ.
Sci. 2016, 9, 1144ꢀ1189. (b) Huber, G. W.; Iborra, S.; Corma, A. Chem. Rev. 2006, 106,
4044ꢀ4098. (c) Corma, A.; Iborra, S.; Velty, A. Chem. Rev. 2007, 107, 2411ꢀ2502.
(2) (a) Xiong, K.; Yu, W.; Vlachos, D. G.; Chen, J. G. ChemCatChem 2015, 7, 1402ꢀ1421. (b)
Xiong, K.; Lee, W. S.; Bhan, A.; Chen, J. G. ChemSusChem 2014, 7, 2146ꢀ2149.
(3) (a) Wang, H.; Male, J.; Wang, Y. ACS Catal. 2013, 3, 1047ꢀ1070. (b) Vlachos, D. G.; Chen, J.
G.; Gorte, R. J.; Huber, G. W.; Tsapatsis, M. Catal. Lett. 2010, 140, 77ꢀ84.
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
(4) (a) Lee, W.ꢀS.; Wang, Z.; Zheng, W.; Vlachos, D. G.; Bhan, A. Catal. Sci. Technol. 2014, 4,
2340ꢀ2352. (b) Lange, J. P.; van der Heide, E.; van Buijtenen, J.; Price, R. ChemSusChem
2012, 5, 150ꢀ166. (c) RománꢀLeshkov, Y.; Barrett, C. J.; Liu, Z. Y.; Dumesic, J. A. Nature
2007, 447, 982ꢀ985. (d) M. J. Taylor, L. Jiang, J. Reichert, A. C. Papageorgiou, S. K.
Beaumont, K. Wilson, A. F. Lee, J. V Barth, G. Kyriakou. J. Phys. Chem. C 2017, 121,
8490ꢀ8497. (e) J. Luo, M. Monai, H. Yun, L. ArroyoꢀRamírez, C. Wang, C. B. Murray, P.
Fornasiero, R. J. Gorte. Catal. Lett. 2016, 146, 711ꢀ717.
(5) (a) Pallassana, V.; Neurock, M.; Hansen, L. B.; Nørskov, J. K. J. Chem. Phys. 2000, 112,
5435ꢀ5439. (b) Greeley, J.; Mavrikakis, M. Nat. Mater. 2004, 3, 810ꢀ815. (c) Kitchin, J. R.;
Nørskov, J. K.; Barteau, M. A.; Chen, J. Phys. Rev. Lett. 2004, 93, 156801ꢀ156804.
(6) GonzálezꢀBorja, M. Á.; Resasco, D. E. Energy Fuels 2011, 25, 4155ꢀ4162.
(7) Shi, D.; Vohs, J. M. ACS Catal.2015, 5, 2177ꢀ2183.
(8) (a) Sitthisa, S.; Resasco, D. E. Catal. Lett. 2011, 141, 784ꢀ791. (b) Zhao, C.; Kou, Y.;
Lemonidou, A. A.; Li, X.; Lercher, J. A. Angew. Chem. Int. Ed. 2009, 121, 4047ꢀ4050. (c)
Kim, S. M.; Lee, M. E.; Choi, J.ꢀW.; Suh, D. J.; Suh, Y.ꢀW. Catal. Commun. 2011, 16,
108ꢀ113.
ACS Paragon Plus Environment
21

ACS Catalysis
Page 22 of 23
1
2
3
4
5
6
7
8
(9) (a) Gutierrez, A.; Kaila, R.; Honkela, M.; Slioor, R.; Krause, A. Catal. Today 2009, 147,
239ꢀ246. (b) Runnebaum, R. C.; Nimmanwudipong, T.; Limbo, R. R.; Block, D. E.; Gates, B.
C. Catal. Lett. 2012, 142, 7ꢀ15.
9
(10) (a) Luo, J.; Yun, H.; Mironenko, A.; Goulas, K.; Lee, J.; Monai, M.; Wang, C.; Vorotnikov,
V.; Murray, C.; Vlachos, D.; Fornasiero, P.; Gorte. R. ACS Catal. 2016, 6, 4095ꢀ4104. (b)
Chang, X.; Liu, A.; Cai, B; Luo, J.; Pan, H.; Huang, Y. ChemSusChem 2016, 9, 3330ꢀ3337. (c)
Yu, W.; Xiong, K.; Ji, N.; Porosoff, M. D.; Chen, J. G. J. Catal. 2014, 317, 253ꢀ262.
(11) (a) Yang, J.; Li, S.; Zhang, L.; Liu, X.; Wang, J.; Pan, X.; Li, N.; Wang, A.; Cong, Y.; Wang,
X.; Zhang, T. Appl. Catal. B: Environ. 2017, 201, 266ꢀ277. (b) Jae, J.; Zheng, W.; Karim, A.;
Guo, W.; Lobo, R.; Vlachos, D. ChemCatChem 2014, 6, 848ꢀ856. (c) Hroneca, M.;
Fulajtárováa, K.; Vávrab, I.; Sotáka,T.; Dobročkab, E.; Mičušíkc, M. Appl. Catal. B: Environ.
2016, 181, 210ꢀ219.
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
(12) Humbert, M. P.; Chen, J. G. J. Catal. 2008, 257, 297ꢀ306.
(13) (a) Kresse, G.; Furthmuller, J. Phys. Rev. B 1996, 54, 11169ꢀ11186. (b) Kresse, G.; Hafner, J.
Phys. Rev. B 1993, 47, 558ꢀ561. (c) Kresse, G.; Hafner, J. Phys. Rev. B 1994, 49,
14251ꢀ14269.
(14) (a) Blöchl, P. E. Phys. Rev. B 1994, 50, 17953ꢀ17979. (b) Spear, K. E.; Frenklach, M. Pure
Appl. Chem. 1994, 66, 1773ꢀ1782.
(15) Perdew, J. P.; Burke, K.; Ernzerhof, M. Phys. Rev. Lett. 1996, 77, 3865ꢀ3868.
(16) Salciccioli, M.; Yu, W.; Barteau, M. A.; Chen, J. G.; Vlachos, D. G. J. Am. Chem. Soc. 2011,
133, 7996ꢀ8004.
(17) Yu, W.; Barteau, M. A.; Chen, J. G. J. Am. Chem. Soc. 2011, 133, 20528ꢀ20535.
(18) Evans, M. G.; Polanyi, M. Trans. Faraday Soc. 1936, 32, 1333ꢀ1360.
(19) (a) G. Ertl, M. Neumann, and K. M. Streit, Surf. Sci. 1977, 64, 393ꢀ410. (b) Myint, M.; Yan,
Y.; Chen, J. G. J. Phys. Chem. C 2014, 118, 11340ꢀ11349.
(20) Sitthisa, S.; Sooknoi, T.; Ma, Y.; Balbuena, P. B.; Resasco, D. E. J. Catal. 2011, 277, 1ꢀ13.
ACS Paragon Plus Environment
22

Page 23 of 23
ACS Catalysis
1
2
3
4
5
(21) Gilkey, M.; Panagiotopoulou, P.; Mironenko, A.; Jennes, G.; Vlachos, D.; Xu, B. ACS Catal.
2015, 5, 3988ꢀ3994.
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
TOC
ACS Paragon Plus Environment
23