Relationship between Atomic Scale Structure and Reactivity of Pt Catalysts: Hydrodeoxygenation of m-Cresol over Isolated Pt Cations and Clusters

Article abstract of DOI:10.1021/acscatal.9b04330

Atomically dispersed late transition-metal cations have attracted significant attention as next-generation heterogenous catalysts. However, relationships between the catalytic behavior of atomically dispersed metal cations and active sites on metal nanoparticles have been difficult to establish, in large part because of the difficulty in characterizing the local atomic structure of these metal species. Here, we use the hydrodeoxygenation (HDO) of m-cresol, a model bio-oil compound, to understand relationships between metal structure and reactivity down to the limit of atomically dispersed active sites. Through a combination of kinetic studies, spectroscopic characterization, and density functional theory calculations, we find that isolated Pt cations supported on TiO₂ are significantly less active than small Pt clusters for m-cresol HDO due to their lower activity for hydrogen dissociation and their weaker interaction with m-cresol. We demonstrate that m-cresol HDO reaction kinetics are particularly sensitive to the active Pt structure, suggesting that the catalytic reactivity can be a more reliable indicator of catalyst structure than commonly used characterization approaches. These findings provide insights into the ability of isolated Pt cations to catalyze elementary processes critical for hydrogenation catalysis.

Full text of DOI:10.1021/acscatal.9b04330

Subscriber access provided by Nottingham Trent University
Article
Relationship between atomic scale structure and reactivity of Pt catalysts:
Hydrodeoxygenation of m-cresol over isolated Pt cations and clusters
Joaquin Resasco, Feifei Yang, Tong Mou, Bin Wang, Phillip Christopher, and Daniel E. Resasco
ACS Catal., Just Accepted Manuscript • DOI: 10.1021/acscatal.9b04330 • Publication Date (Web): 03 Dec 2019
Downloaded from pubs.acs.org on December 4, 2019
Just Accepted
“Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted
online prior to technical editing, formatting for publication and author proofing. The American Chemical
Society provides “Just Accepted” as a service to the research community to expedite the dissemination
of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in
full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully
peer reviewed, but should not be considered the official version of record. They are citable by the
Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore,
the “Just Accepted” Web site may not include all articles that will be published in the journal. After
a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web
site and published as an ASAP article. Note that technical editing may introduce minor changes
to the manuscript text and/or graphics which could affect content, and all legal disclaimers and
ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or
consequences arising from the use of information contained in these “Just Accepted” manuscripts.
is published by the American Chemical Society. 1155 Sixteenth Street N.W.,
Washington, DC 20036
Published by American Chemical Society. Copyright © American Chemical Society.
However, no copyright claim is made to original U.S. Government works, or works
produced by employees of any Commonwealth realm Crown government in the course
of their duties.

Page 1 of 18
ACS Catalysis
1
2
3
4
5
6
7
8
9
Relationship between Atomic Scale Structure and Reactivity of Pt catalysts:
Hydrodeoxygenation of m-cresol over Isolated Pt Cations and Clusters
Authors: Joaquin Resasco1,†, Feifei Yang2,†, Tong Mou2, Bin Wang2, Phillip Christopher1,*,
Daniel E. Resasco2,*
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
Affiliations:
1. Department of Chemical Engineering, University of California, Santa Barbara, Santa Barbara,
CA 93106, United States
2. School of Chemical, Biological, and Materials Engineering and Center for Interfacial Reaction
Engineering, University of Oklahoma, Norman OK 73019, United States
† These authors contributed equally
*Corresponding authors: pchristopher@ucsb.edu, resasco@ou.edu
Abstract
Atomically dispersed late transition metal cations have attracted significant attention as next
generation heterogenous catalysts. However, relationships between the catalytic behavior of
atomically dispersed metal cations and active sites on metal nanoparticles have been difficult to
establish, due in large part to the difficulty in characterizing the local atomic structure of these
metal species. Here, we use the hydrodeoxygenation (HDO) of m-cresol, a model bio-oil
compound, to understand relationships between metal structure and reactivity down to the limit
of atomically dispersed active sites. Through a combination of kinetic studies, spectroscopic
characterization, and density functional theory calculations, we find that isolated Pt cations
supported on TiO2 are significantly less active than small Pt clusters for m-cresol HDO due to
their lower activity for hydrogen dissociation and their weaker interaction with m-cresol. We
demonstrate that m-cresol HDO reaction kinetics are particularly sensitive to the active Pt
structure, suggesting that catalytic reactivity can be a more reliable indicator of catalyst structure
than commonly used characterization approaches. These findings provide insights into the ability
of isolated Pt cations to catalyze elementary processes critical for hydrogenation catalysis.
Keywords: Atomically dispersed catalysts, single atom catalysis, m-cresol, hydrodeoxygenation, Pt
Introduction
Since the proposal that catalytic surfaces contain active sites with differing reactivity, there has
been ongoing interest in understanding the link between the structure of these sites and their
properties.1 For instance, many studies have explored the relationship between the size of
supported metal particles and their reactivity.2-6 Recently, there has been significant interest in
extending these relationships to the limit of atomically dispersed late-transition metal cations.7-9
However, it has proven experimentally difficult to devise synthetic methods for producing
exclusively atomically dispersed species with uniform structures and properties.10 Instead, many
1
ACS Paragon Plus Environment

ACS Catalysis
Page 2 of 18
1
2
3
4
5
6
7
8
9
synthetic methods produce a distribution of metal structures including atomically dispersed
species with varying coordination to the oxide support, and small metal particles.10-11
Furthermore, recent studies have demonstrated the difficulty in differentiating these atomic scale
structures using common characterization techniques.12-13 These challenges preclude the
understanding of the catalytic behavior of atomically dispersed metal cations and the
relationships between their properties and those of extended metal 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
While common characterization tools are often insensitive to subtle changes in molecular level
structure of these materials, reaction kinetics are not. Therefore, it is possible that catalytic
reactivity itself can be used as a probe of the structures present in a catalyst material. This is
most practical for reactions in which the kinetics are known to be strongly dependent on active
site structure. One such class of reactions is the hydrodeoxygenation (HDO) of biomass derived
molecules. These reactions are used to remove excess oxygen content from biomass for potential
use as fuels or fuel additives.14-15 Typically, these reactions require the ability of the catalyst to
facilitate hydrogenation and dehydrogenation steps, as carbon-oxygen bond cleavage steps. This
required multifunctionality could exist at metal-oxide catalyst interfaces, or over catalysts in
which one site can enable both chemistries.16-21
In previous studies, we explored the mechanistic routes for HDO of cresol and related molecules
on extended metal surfaces.22-26 We have shown that while oxophilic metals are able to catalyze
HDO directly on the metal surface, the reactivity of noble metals, such as Pt and Pd, are
promoted by the existence of exposed oxophilic cations on reducible oxide supports.22 On
catalysts such as Pt and Pd on TiO2 or ZrO2 supports, HDO occurs at the metal-support interface
with O in cresol binding to exposed Ti or Zr cations and the metal catalyzing H2 activation and
hydrogenation.27 This reaction is therefore of particular interest to understand the ability of
isolated metal cations for catalyzing complex reactions. As every metal atom in an atomically
dispersed metal catalyst resides at the metal-support interface, this reaction allows us to
understand whether isolated Pt cations (Ptiso) mimic the ability of perimeter sites on Pt
nanoparticles supported on TiO2, or whether the cationic nature of Ptiso and lack of metal
ensembles imparts fundamentally different reactivity.
Here, we evaluate the ability of Ptiso species to catalyze HDO of m-cresol, a molecule
representative of phenolic compounds in bio-oil and widely used as a model compound to
understand the reactivity of families of biomass-derived species. We find that Ptiso on TiO2 is a
poor catalyst for HDO of m-cresol compared to Pt clusters largely due to the unfavorable
energetics of H2 dissociation over Ptiso species. Further, the strong coordination of the Ptiso to
oxygen atoms from the support imparts stability to the isolated metal cation under reductive
environments but may reduce its propensity to bind cresol and activate carbon oxygen bonds. We
also synthesize materials containing a mixture of Ptiso and small Pt clusters to show that clusters
dominate their catalytic behavior. Therefore, reactions themselves can be used as a sensitive
probe of atomic scale active site structure. These findings demonstrate the limitations of isolated
Pt cations in catalyzing hydrogenation reactions, and the relationships between their reactivity
and their stabilizing coordination to the oxide support.
2
ACS Paragon Plus Environment

Page 3 of 18
ACS Catalysis
1
2
3
4
Results and discussion
5
6
Understanding the mechanism of HDO of m-cresol on Pt clusters
7
8
9
To begin to understand the relationship between atomic scale Pt structure and reactivity for HDO
of m-cresol, we first demonstrate the influence of reaction temperature and support composition
on m-cresol conversion over Pt clusters. Catalysts containing 1% weight (wt.) loading of Pt on
TiO2 and SiO2 supports comprised of small Pt clusters (dispersion of 29 % and 49 %
respectively, measured by CO pulse chemisorption) were synthesized by incipient wetness
impregnation and used for comparison to catalysts containing Ptiso species. While SiO2 was
chosen as an inert support where HDO is expected to occur solely over the metal surface,24 TiO2
was chosen as an active support where Ti cations at the metal-support interface can participate in
the reaction.28 Specific characteristics of these oxides are detailed in the methods section. The
gas phase conversion of m-cresol was studied in a fixed bed reactor at atmospheric pressure, and
care was taken to ensure all measurements were taken under strict kinetic control.29 At low
temperatures (<250 °C) over Pt clusters, ring hydrogenation occurs preferentially over
deoxygenation (Fig 1a). In previous studies we showed that at these low temperatures,
hydrogenation is thermodynamically favorable, while the barrier for cleavage of the C-O bond is
prohibitively high to produce significant amounts of the deoxygenation product toluene (Fig
1a).24, 26 At high enough temperatures (>300 °C), kinetic barriers for deoxygenation can be
overcome and toluene is produced with high selectivity. We will focus this study on the
influence of Pt structure on HDO reactivity at 350 °C. However, comparisons of reactivity for
ring hydrogenation at lower temperature are also discussed 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
Pt clusters (1 wt. %) on TiO2 showed higher reactivity towards deoxygenation to toluene than Pt
on SiO2 (Fig 1b). This result was confirmed on lower weight loading materials (0.1 wt. %) as
shown in Table S1. Previous studies have shown that this trend in activity is the result of active
sites at the perimeter of Pt particles and the oxophilic TiO2 support.17, 30 With this in mind, we
were interested in understanding whether a catalyst composed of all Ptiso species on TiO2, in
which every site interfaces with the support, would be equally active for this reaction, or whether
Ptiso species have different reactivity than perimeter sites on Pt clusters for this reaction.
3
ACS Paragon Plus Environment

ACS Catalysis
Page 4 of 18
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
Fig 1. Understanding HDO of m-cresol over Pt clusters. a) Yield to major products over 1 wt. %
Pt/TiO2 catalysts at low and high temperatures. Reaction conditions: W/F=0.05h; P=1atm H2;
TOS=20min. b) Comparison of toluene turnover frequency for Pt on different supports. Turnover
frequency values (+/- 1min-1) are based on number of surface sites counted by CO chemisorption.
Reaction conditions: T=350oC; P=1atm H2; TOS=20min; Conversion was maintained below 15% by
adjusting W/F.
Comparison of HDO activity of Ptiso and Pt clusters
To produce Ptiso catalysts, synthesis was conducted via strong electrostatic adsorption.31 We have
previously demonstrated that by using this synthetic technique and a support comprised of 5 nm
diameter anatase TiO2 nanocrystals at a weight loading of 0.025%, which correlates to nominally
one Pt atom for every two TiO2 particles, we can produce exclusively Ptiso species.10, 32-33 Prior to
analysis, the catalyst was calcined in air at 450 °C to remove remaining ligands from synthesis.34
We then characterized the structure and local environment of Ptiso species using CO probe
molecule Fourier transform infrared (FTIR) spectroscopy following in-situ pretreatment of the
catalyst in H2 for 1 hour at 250 °C. No significant difference was observed in IR characteristics
after pretreatment at 250 °C and 350 °C in H2. After a saturating dose of CO at 25 °C and
subsequent purging, a band centered at 2112 cm-1 with a full-width at half maximum of 11 cm-1
was observed (Fig 2a). During desorption of CO from the 0.025 % Pt/TiO2 catalyst, the vibrational
frequency did not change, indicating that Pt species were isolated and uniform in their local
environment (Fig S1). This IR signature is consistent with our previous work on Ptiso on TiO2.10,
32-33
Previously, through a combination of atomic resolution imaging, x-ray absorption
spectroscopy, and density functional theory (DFT) calculations, we conclusively demonstrated that
Ptiso species under these conditions exist as PtO2 structures bound to step or terrace sites on the
anatase TiO2 (101) surface.32-33 In contrast, after CO adsorption on the 1% Pt/TiO2, a broad band
centered at 2060 cm-1 was observed, consistent with adsorption of CO on metallic Pt clusters.32
Having produced a uniform and well-defined Ptiso/TiO2 catalyst, as well as a comparative sample
with only Pt clusters and a dispersion of 49% (which suggests a cluster size of ~2 nm), we then
compared their performance for HDO of m-cresol.
4
ACS Paragon Plus Environment

Page 5 of 18
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 2: Activity of isolated Pt cations and Pt clusters for HDO of m-cresol. a) CO probe molecule
infrared spectroscopy of 0.025% and 1% Pt on TiO2. The sharp band centered at 2112 cm-1 is
characteristic of isolated Pt cations. The broad band centered at 2060 cm-1 is characteristic of Pt clusters.
b) Comparison of TOF for toluene formation for 0.025% Pt and 1% Pt on TiO2. Turnover frequency (+/-
1 min-1) for 1% Pt is based on number of surface sites counted by CO chemisorption; a dispersion of
100% is assumed for 0.025% Pt. Reaction conditions T=350oC; P=1atm H2; TOS=20min. Conversion
was maintained below 15% by adjusting W/F. c) CO probe molecule infrared spectroscopy of 0.025% Pt
on TiO2 before and after reaction testing. Identical spectra indicate isolated Pt cations are stable under
reaction conditions.
As shown in Fig 2b, we find that the activity of Ptiso for deoxygenation of m-cresol to produce
toluene is ~40x lower on a per-site basis than it is over Pt clusters. Comparison of reactivity of
0.025% Pt/TiO2 with TiO2 supports not loaded with Pt demonstrates Ptiso does have catalytic
activity for HDO (Fig S2). To confirm that Ptiso catalysts were stable under HDO reaction
conditions, CO probe molecule IR was conducted before and after reactivity testing. After mild
oxidation to remove carbonaceous deposits and reduction (300 °C in O2 for 30 min, 250 °C in H2
for 30 min), a nearly identical IR signature was observed, indicating that Ptiso species are stable
and that we can accurately assess their reactivity for HDO (Fig 2c). The stability of these Ptiso
species throughout pretreatment and reaction testing under reducing conditions contrasts with
commonly reported behavior for atomically dispersed catalysts. We hypothesize that the unusual
stability stems from the low Pt loading, which allows us to probe Pt cations that only occupy their
most thermodynamically stable sites on the support.35
To further demonstrate the sensitivity of HDO reactivity to Pt structure, a set of catalysts were
synthesized using electrostatic adsorption with increasing weight loading to vary the ratio of Ptiso
species and Pt clusters. To ensure that samples included both Ptiso species and Pt clusters, a higher
Pt surface loading was used by choosing a lower surface area TiO2 support (BET surface area 35-
65 m2/g for the low surface area support, 290 m2/g for the high surface area support). This same
low surface area support was used above for the 1% Pt/TiO2 material. We have previously shown
that at a fixed weight loading, a higher surface loading results in increased formation of Pt
clusters.10 Examining the reactivity of lower surface area TiO2 samples containing Pt loadings of
0.025-1 wt.% we see that the turnover frequency (TOF) for forming toluene increases with
increasing Pt weight loading, until it saturates at ~0.3% Pt (Fig 3a). Further increases in weight
loading up to 1.0% Pt resulted in minimal changes in per site activity. Furthermore, the TOF at a
fixed weight loading of 0.025% Pt is ~5x higher in this case than for the sample containing
exclusively Ptiso.
5
ACS Paragon Plus Environment

ACS Catalysis
Page 6 of 18
1
2
3
4
5
6
To understand these trends in reactivity, we examined the structure of these different catalysts
through CO probe molecule IR. For the 0.025% Pt sample deposited on the lower surface area
support, we see a clear feature associated with Ptiso, as before, with the band centered at 2112 cm-
1
and a tight FWHM of 12 cm-1 indicating that these species coordinated to oxygen at uniform
7
positions on the support. This further confirms that the local coordination for Ptiso species are
similar and uniform on the two TiO2 supports with different surface areas. However, unlike the
sample synthesized at lower Pt surface loading, here a weak, broad band centered at 2060 cm-1
was also observed, indicating the presence of CO bound to metallic Pt clusters.32 As seen in Fig
3b, IR indicates the fraction of Pt existing in the form of clusters rather than Ptiso increases with
weight loading, as expected. To understand the relative contributions of these sites to measured
activity, the IR and reactivity data were used together to extract estimates of relative extinction
coefficients of CO bound to Ptiso and Pt clusters (Fig S3). This analysis allows conversion of IR
intensities into estimates of the fractions of Pt atoms exposed as Ptiso species or at the surface of
Pt clusters (see methods for more details). Using this methodology, we find the extinction
coefficient CO bound to Ptiso is ~5x larger than that of CO on metallic Pt clusters, consistent with
previous measurements on isolated Rh cations and Rh clusters.36 This implies that while Ptiso -
species have a small contribution to measured reactivity in samples containing a mixture of Pt
structures, their contribution to measured CO IR spectra is much larger. Using this procedure, we
predict that less than 10% of Pt existed in the form of Pt clusters for the 0.025% sample on the
lower surface area TiO2, however the TOF for toluene formation was ~8x larger than the sample
containing exclusively Ptiso. Therefore, while Ptiso species may have a larger contribution to signals
observed in IR, the catalytic activity measured for this sample is dominated by the presence of the
Pt clusters. Thus, reactivity for m-cresol HDO is a much more sensitive probe of the existence of
Pt clusters in a given catalyst than CO IR analysis. However, IR is quite effective at identifying
the presence of Ptiso species. This analysis suggests that the activity trends observed with weight
loading can be described by normalization of the activity to different fractions of active Pt cluster
sites and minimally active Ptiso sites. Rather than a higher weight loading catalyst possessing
intrinsically higher activity sites, it simply contains a smaller fraction of inactive sites that reduce
the calculated TOF.
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
To more conclusively substantiate this point, temperature dependent reactivity measurements were
taken to obtain apparent activation barriers. Comparing the samples synthesized at higher surface
loading, which contain both Pt clusters and Ptiso, very little change in barrier is observed with all
catalysts showing a barrier of ~85 kJ/mol. This supports the idea that the same type of site, an
interfacial site on a Pt cluster, is responsible for the observed activity of all of these Pt catalysts.
In stark contrast, the barrier over the 0.025% Pt catalyst which contained exclusively Ptiso species,
showed a significantly higher apparent activation energy of 189 kJ/mol. Again, we note that the
sample produced at a weight loading of 0.025% Pt on the lower surface area support, while
showing only a weak feature for Pt clusters in IR, shows a barrier consistent with the behavior of
Pt clusters, rather than Ptiso.
6
ACS Paragon Plus Environment

Page 7 of 18
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 3: Activity of catalysts containing both Ptiso and Pt clusters for HDO of m-cresol. a) Turnover
frequency values (+/- 1min-1) for deoxygenation of m-cresol to toluene at 350 °C for samples of different
weight loadings from 0.025-1.0 wt.%. Turnover frequency for higher Pt loadings is based on number of
surface sites counted by CO chemisorption; a dispersion of 100% is assumed for lower weight loadings
(<0.10% Pt). TOF of catalyst synthesized on higher surface area support containing exclusively Ptiso
shown for comparison. Reaction conditions T=350oC; P=1atm H2; TOS=20min. Conversion was
maintained below 15% by adjusting W/F. b) CO probe molecule infrared spectroscopy of the same
catalysts. Data normalized to maximum peak intensity and offset for clarity. Dotted lines added to make
weak features more easily visible. c) Arrhenius plots showing apparent activation barriers for the same
catalysts. Reaction conditions: T=275-350 °C for catalysts containing Pt clusters; T=330-370 °C for the
catalyst containing only Ptiso; P=1atm H2; TOS=20 min. Conversion was maintained below 15% by
adjusting W/F.
Understanding reactivity of Ptiso
Having observed that Ptiso species have low reactivity for HDO of m-cresol, we set out to
understand why this is the case, and what fundamental differences exist between Ptiso species and
the perimeter site of a Pt cluster. We hypothesized that two limiting factors in the HDO of cresol
could be responsible for the low Ptiso reactivity: the difficulty of dissociating hydrogen over Ptiso
species, or the weak interaction between Ptiso and cresol rendering it unable to effectively cleave
the C-O bond.
To test the first possibility, we used cyclohexene hydrogenation as a probe reaction (Fig 4a).37-38
Comparing the 0.025% Pt catalysts composed of only Ptiso with the catalysts containing both Ptiso
and Pt clusters, a significant difference in activity is observed. The catalyst composed of
exclusively Ptiso again shows significantly lower activity at a fixed temperature of 120 °C, and a
temperature difference of 40°C (120°C vs 80°C) is needed to obtain similar cyclohexene
conversion over both catalysts. This again suggests Ptiso species have low activity for cyclohexene
hydrogenation and the reactivity of the higher surface loading sample is dominated by Pt clusters.
This is further supported by cyclohexene hydrogenation reactivity measured over a 1% Pt on TiO2
sample, which quantitatively converts cyclohexene under the same conditions. This probe reaction
indicates that hydrogen dissociation and alkene hydrogenation is not facile over Ptiso. This is
consistent with recent comparative studies of alkene hydrogenation over isolated Pt and Pd cations
and small clusters.39-41
Further insight into the hydrogenation activity of Ptiso could be obtained by examining the trend in
m-cresol conversion rates at a lower temperature of 220 °C, where HDO no longer occurs (Fig
4b). Turnover rates to hydrogenation products (methylcyclohexanone, methylcyclohexanol, and
7
ACS Paragon Plus Environment

ACS Catalysis
Page 8 of 18
1
2
3
4
5
6
7
8
9
methylcyclohexane) increased with Pt weight loading before saturating at the same loading that
CO-IR showed Pt existed exclusively as Pt clusters. The catalyst material that IR indicated was
composed of all Ptiso had no measurable reactivity under these conditions. This further
demonstrates that Ptiso sites had no measurable activity for m-cresol conversion, and that turnover
rates for all catalysts were dominated by the amount of Pt in the form of clusters. We previously
showed a similar relationship between the relative fractions of isolated and clustered Rh species
and the selectivity for CO2 reduction.36
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
Thermogravimetric analysis was performed of spent catalysts after HDO of m-cresol (T = 350 °C,
TOS = 3.5 h) to understand differences in coke formation and their relation to hydrogenation
activity. As shown in Fig S4, negligible coke formation was observed on the 1% Pt/TiO2 sample
containing a high density of Pt clusters. For the samples previously determined to contain a small
fraction of Pt clusters (0.025% - 0.1% Pt) increasing amounts of coke were observed with
decreasing Pt loading. At the lowest weight loading where Pt existed predominantly in the form
of Ptiso, similar amounts of coke were observed as a pure TiO2 support indicating that the metal
was unable to effectively clear coke from the oxide support. Similarly, significantly higher (~3x)
amounts of coke were found on the sample containing exclusively Ptiso species after reaction due
to the higher surface area of this TiO2 support (support surface area is 4-8x larger). These results
taken together suggest that Pt clusters that are able to effectively dissociate hydrogen, allowing for
spillover onto TiO2 and hydrogenation of carbonaceous species on the oxide support, cleaning the
surface from coking. Ptiso species that are unable to effectively dissociate hydrogen are not able to
clear the surface from coke.
DFT calculations were performed to examine the dissociative adsorption of H2 over Ptiso in order
to corroborate these experimental measurements. To model the Ptiso structure on TiO2, we used a
structure comprised of an adsorbed PtO2 moiety at an anatase TiO2 (101) terrace (Fig 5a). This
structure was previously found to match experimental measurements for Ptiso on TiO2.32 To
understand the effect of Pt local coordination, we also considered hydrogen dissociation on a Pt
atom coordinated to two lattice-oxygen at a bridging position on the TiO2 (101) surface (Fig 5b).
We find that the dissociative adsorption of H2 on the PtO2 structure is highly exothermic (-282
kJ/mol). However, the most favorable adsorption sites are support oxygens adjacent to the Pt atom
(forming a surface hydroxyl group). We expect that hydrogen atoms strongly bound to these
support oxygens are not active for hydrogenation. In contrast, hydrogen adsorption on Pt atoms
coordinated to two support oxygens is exothermic by only 20 kJ/mol using gas-phase H2 as the
reference (Fig 5a) indicating H2 dissociation is not facile. The most favorable adsorption sites of
two H atoms are found to be on the Pt atom and on the support oxygen adjacent to the Pt atom.
Over a Pt (111) surface, the adsorption process has a higher exothermicity (-101 kJ/mol). H atoms
dissociatively adsorb on the fcc sites of the Pt (111) surface (Fig 5c). Hydrogen dissociation
barriers over the Pt (111) surface are known to be negligible, as contiguous metal sites provide
favorable sites for dissociated hydrogen atoms to adsorb.42 Unlike a hydrogen atom bound to Pt,
the surface hydroxyls formed after H2 dissociation on Ptiso is unreactive for hydrogenating
unsaturated bonds. Interestingly, recent surface science studies for Pd atoms on Fe3O4 supports
have experimentally shown a H2 dissociation mechanism consistent with the picture our DFT
calculations provide for H2 dissociation over Ptiso.43 We note the contrast here between Ptiso on an
oxide support and isolated Pt atoms in a metal lattice studied for so called single atom alloys. The
ability of these single atom alloys to dissociate hydrogen onto nearby metal sites allows them to
be efficient catalysts for hydrogenation chemistry.42, 44 These experimental and computational
8
ACS Paragon Plus Environment

Page 9 of 18
ACS Catalysis
1
2
3
4
5
results together suggest that difficult dissociation of hydrogen is to a large extent responsible for
the low activity of Ptiso for HDO of cresol.
6
7
8
9
We also considered the possibility that HDO of m-cresol over Ptiso could be hindered by the
difficulty in performing C-O bond scission over these sites. In previous studies we have shown
that the oxygen binding energy of the metal is an important descriptor for the barrier for the
kinetically relevant C-O scission step in HDO.23 Therefore, weaker interaction between Ptiso and
cresol could render it inactive for catalyzing this reaction step. We have previously shown that the
binding energy of small adsorbates over stable Ptiso sites is significantly lower than it is over
extended Pt surfaces.10, 32 This can be intuitively described by a compromise between stability of
a metal ion garnered from strong coordination to oxygens from the support, and a propensity of
this ion to form additional bonds to adsorbates. Similar tradeoffs have been demonstrated for more
extended metal particles on well-defined oxide supports through calorimetry studies.45-48 We
expected therefore that weakened binding of cresol to Ptiso species would also result in an increased
barrier for C-O bond scission, reducing activity for HDO. We expect that the binding energy of
cresol through carbon should scale with the binding energy of CO, supporting weakened binding
relative to an extended metal surface.49 Furthermore, adsorption of cresol and phenolics is known
to be enhanced by π interactions between the aromatic ring and metal terraces, which would not
be possible over an isolated metal atom.50 To substantiate this further, DFT calculations of cresol
adsorption were done over the two Ptiso structures (Fig S5). Adsorption energies of -206 and -104
kJ/mol were found over the PtO2 structure and the Ptiso coordinating to two support oxygen ligands,
respectively. We find the adsorption of cresol over Pt (111) to be exothermic with an adsorption
energy of -229 kJ/mol. We note, however, that cresol adsorbed at the metal-support interface is
likely the more kinetically relevant species than cresol adsorbed to Pt terraces, and that the
presence of coadsorbates could influence binding energies. Thus, the relevant adsorption energy
over Pt clusters is likely intermediate between that of Ptiso and Pt (111).51 Taken together, these
results show that the lower reactivity of Ptiso for HDO of m-cresol is mainly related to the difficulty
in dissociating hydrogen over an isolated metal cation, while weak interactions between Ptiso and
cresol could potentially contribute to reactivity to a lesser extent.
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
9
ACS Paragon Plus Environment

ACS Catalysis
Page 10 of 18
1
2
3
4
5
6
7
8
9
Figure 4: Understanding activity of isolated Pt cations a) Activity of samples containing Ptiso and Pt
clusters for a probe hydrogenation reaction (cyclohexene hydrogenation). Reaction conditions: P=1 atm H2;
W/F=0.007h; TOS=3min. (b) Turnover frequency (+/- 0.1 min-1) for ring hydrogenation of m-cresol to
hydrogenation products (methylcyclohexanone, methylcyclohexanol, methylcyclohexane) at 250 °C for
samples of different weight loadings from 0.025-1.0 wt.%. Reaction conditions: T=220 C; P=1atm H2;
W/F=0.652h, TOS=20min. Note that the sample containing exclusively Ptiso had no measurable activity
under these conditions.
o
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: Adsorption of hydrogen on Pt structures. DFT optimized adsorbed structures with calculated
heats of adsorption for (a,b) dissociative adsorption of H2 over Ptiso and (c) Pt (111). Ptiso in (a) is modeled
by a PtO2 structure described previously.32 Ptiso in (b) is modeled by a Pt atom coordinating to two oxygen
ligands from the support on the anatase (101) surface. The balls refer to the atoms of H (white), O (red), C
(grey), Pt (blue) and Ti (green), respectively.
Conclusions
In this study, we sought to understand the influence of Pt structure (down to the limit of single Pt
cations) on reactivity for the upgrading of bio-oil derived molecules, specifically the HDO of m-
cresol. In doing so, we showed that the catalytic reactivity of different types of Pt species is
extremely sensitive to their structure. Combining these results with our previous work showing the
difficulty in accurately differentiating sub-nanometer Pt species from Ptiso species, we argue that
often catalytic reactivity is the most accurate probe of atomic scale structure.12, 36 We find that
Ptiso species show low HDO activity (~40x lower TOF to toluene than Pt clusters) due to their
inability to effectively dissociate hydrogen and hydrogenate unsaturated bonds. We further
synthesized materials in which Ptiso and Pt clusters coexist and show that while Ptiso species can
have a large contribution to spectroscopic measurements, the catalytic activity is dominated by the
clusters. This suggests that the sites most easily visible using characterization tools are not always
those that most significantly govern measured catalytic properties.
These reactivity studies give some general insights about the propensity of isolated metal cations
(particularly Pt) for catalyzing specific reaction steps. Results are beginning to emerge that show
isolated cations in their thermodynamically most stable site on oxides often coordinate strongly to
supports, but weakly to adsorbates in comparison to metal clusters. This suggests that their inherent
ability to break bonds may be less than clusters, but also may be highly tunable based on modifying
their adsorption site on the support or the support composition. Therefore, it can be concluded that
10
ACS Paragon Plus Environment

Page 11 of 18
ACS Catalysis
1
2
3
4
5
atomically dispersed catalysts, while not a panacea for catalysis, may provide well-defined and
tunable catalytic sites for some reaction chemistries.
6
7
8
Methods:
9
Materials
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 following materials were purchased from Sigma-Aldrich: Tetraammineplatinum nitrate (99.995%)
(TAPN) with SKU no. 482293, SiO2 (99.8%) with SKU no. 381276, reagent grade NH4OH (> 25%) with
SKU no. 17093, m-Cresol (99%) with SKU no. C85727, cyclohexene (≥ 99.0%) with SKU no. of 29240.
Cyclohexene was purified by distillation before use.
For TiO2 supported catalysts, two different materials were used as supports. To promote the formation of
Pt clusters, lower surface area TiO2 (99.5%) was purchased from Sigma-Aldrich with SKU no. 718467.
To promote formation of exclusively atomically dispersed Pt species, a higher surface area TiO2
composed of 5 nm anatase crystals was purchased from US Research Nanomaterials (Stock no. US3838).
The quoted surface areas of the two supports were 35-65 m2/g and 290 m2/g, respectively.
Catalyst preparation
1 wt.% Pt/TiO2 and 1 wt.% Pt/SiO2 catalysts were prepared by an incipient wetness impregnation
method. In a typical synthesis, 60.15 mg TAPN was dissolved in H2O (4 ml for TiO2, 10 ml for SiO2), and
dropwise added to 3 g pre-dried TiO2 or SiO2 support. After impregnation, the catalysts were dried at
room temperature overnight.
Lower weight loading catalysts (0.025-0.5 wt.%) were prepared by strong electrostatic adsorption. For
the synthesis of 0.025 wt.%Pt/TiO2, 10 mg TAPN was dissolved in 5 ml H2O, from which 1.5 ml was
taken and mixed with 8.5 ml NH4OH to prepare the precursor solution. 3 g TiO2 support powder was
dried in a vacuum oven at 120 °C for 12 h, and then dissolved to a mixture solution of 200 ml NH4OH
and 50 ml H2O. The TAPN precursor solution was injected by a syringe pump at a rate of 0.5 ml/h to the
support solution under constantly stirring. The final solution was heated to 70 °C until completely dried.
The synthesis of Pt/TiO2 catalysts with different Pt loading catalysts (0.05-0.5 wt.%) follows the same
procedure as above, but with different concentration of precursor solutions.
All dried catalysts were placed in a vacuum oven at 120 °C for 12 h, and further calcined at 450 °C for
4 h in flowing air, at a ramping rate of 2 °C/min. Before reactivity testing, the catalyst powders were
pressed, crushed, and sieved to 40-60 mesh.
Catalytic activity
Catalytic conversion of m-cresol or cyclohexene was carried out in a fixed-bed stainless-steel tube (6
mm o.d.) reactor at atmospheric pressure. The catalyst was pelletized between 40 and 60 mesh, mixed
with inert silica beans as a diluent (40-60 mesh, 100 mg), and packed between two layers of quartz wool
on top of a thermocouple placed inside the reactor tube. The catalyst sample was reduced in flowing H2 at
350 °C for 1 h after heating at a rate of 10 °C/min. After reduction, the reaction was conducted at 80-370
°C depending on the measurement of interest. The m-cresol or cyclohexene was fed using a syringe pump
(kd scientific) and vaporized (180 °C for m-cresol, 60 °C for cyclohexene) before entering the reactor.
The reactor gas lines were heated at 250 °C to avoid any condensation. The products were analyzed and
quantified online by a gas chromatograph (HP 5890), equipped with an HP-Innowax column (30 m × 0.5
mm ID × 0.25 μm) and a flame ionization detector. The effluent was trapped by methanol in an ice-water
11
ACS Paragon Plus Environment

ACS Catalysis
Page 12 of 18
1
2
3
4
5
6
bath and its components were identified by a gas chromatography-mass spectrometer (Shimadzu
QP2010s). Fresh catalyst was used for each reactivity measurement at each space time (W/F, gcatgreactant
1h). The conversion and yield are reported in molcarbon%.
-
7
8
Chemisorption Measurements
9
CO pulse chemisorption dispersion measurements were performed using a Micromeritics Autochem II
2920 gas analyzer. In a typical measurement, approximately 500 mg of calcined catalyst was loaded into a
U-shaped sample tube and reduced at 623 K for 1 h in 10% H2/Ar (Airgas). The catalyst was then flushed
with He (99.999%, Airgas) while ramping to 798 K, and held at that temperature for 30 min. After the
sample cooled to ambient conditions, pulse chemisorption measurements were performed with 10%
CO/He (Airgas), while monitoring the effluent with a thermal conductivity detector. The total flow rate for
each experiment was 50 sccm. Dispersion was calculated assuming hemispherical clusters and a Pt:CO
ratio of 1:1. The volume of active gas adsorbed on the surface of each catalyst per gram of material was
used in calculation of turnover frequencies (TOFs). Dispersions of 81, 64, and 49% were measured for
0.3, 0.5, and 1.0 wt. % Pt/TiO2 respectively. A dispersion of 29% was measured for Pt/SiO2. Dispersions
of 100% were assumed for lower weight loadings (0.025-0.1%) due to the observation of significant
amounts of Ptiso species and the weak adsorption of CO at room temperature over Ptiso sites on TiO2.
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
Thermogravimetric Analysis of spent catalysts
Coke formation on spent catalysts were analyzed by Thermo Gravimetric Analysis (TGA). To compare
coke formation, the catalysts were subjected to HDO reactivity testing at identical reaction conditions
(Catalyst mass: 100 mg, T = 350 °C, m-Cresol = 0.1 ml/h, TOS = 3.5 h, H2 = 58 ml/min). After reaction,
the catalyst was cooled to room temperature in flowing H2.
TGA analysis was carried out using a TGA Q500 instrument in flowing air. Approximately 30 mg of
spent catalyst was loaded into the instrument. The temperature was initially rapidly increased to 150 °C
and kept isothermal for 15 min. After this period the temperature was increased from 150 °C to 850 °C at
a heating rate of 10 °C/min.
Infrared spectroscopy
FTIR experiments were carried out in a diffuse reflectance reaction chamber (Harrick Scientific)
equipped with ZnSe windows, mounted inside a Praying Mantis diffuse reflectance adapter (Harrick
Scientific), and coupled to a Thermo Scientific Nicolet iS10 FTIR spectrometer with a liquid-nitrogen-
cooled HgCdTe (MCT-A) detector. During measurements, the FTIR and Praying Mantis diffuse
reflection accessory were purged with dry N2. In a typical experiment, the reactor was loaded with ~80
mg of pure Al2O3 support, followed by 25 mg of calcined catalyst packed on top of the inert support. The
catalyst was pretreated in-situ in 50 sccm of a 10% H2/Ar or 100% O2 at various pretreatment
temperatures for 1 hour prior to analysis. All feed gases were passed through a desiccant and liquid
nitrogen cold trap to remove any humidity prior to the reaction chamber. Upon completion of the
pretreatment, the catalyst was cooled to room temperature. All FTIR spectra were collected in difference
mode. A background spectrum was then collected before introducing a flow of 10% CO in Ar for 10
minutes. CO was subsequently purged out by Ar flow.
Site fraction quantification
To estimate the relative fractions of Ptiso and Pt clusters in samples containing a distribution of both
species, infrared measurements were correlated to experimental reactivity data. For each material, the area
of the infrared feature associated with CO adsorbed to Ptiso or Pt clusters was integrated numerically. The
12
ACS Paragon Plus Environment

Page 13 of 18
ACS Catalysis
1
2
3
4
5
site fraction can be obtained by weighting of the integrated area with the extinction coefficient for the
site:
6
7
8
퐴푖/(휀푖 × 푁푖)
푋푖 =
푛
∑
푖 = 1[퐴푖/(휀푖 × 푁푖)]
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
where 푖 is the fraction of total adsorbed molecules bound to each type of site; 푖 is the absorption peak
푋
퐴
휀
푁
푖
area associated with each type of site; 푖 is the relative extinction coefficient for each type of site; and
is number of metal atoms coordinated to the adsorbate, for n different types of sites. In this case we
consider two types of sites (Ptiso and Pt clusters) and the CO:Pt stoichiometry is taken as 1:1. The
expected turnover frequency can then be obtained from a linear combination of the TOF of a material
containing exclusively Ptiso and one containing exclusively Pt clusters:
푇푂퐹푐푎푙푐 = 푋푖푠표푇푂퐹푖푠표 + 푋푐푙푢푠푡푒푟푇푂퐹푐푙푢푠푡푒푟
As the extinction coefficient of Ptiso is not known, this methodology was employed to obtain relative
extinction coefficients by allowing them to vary such that the error between calculated and experimental
turnover frequencies were minimized. This procedure was applied to reactivity data obtained at both 220
푇푂퐹
푖푠표 values were obtained from data on
°C and 350 °C to obtain consistent extinction coefficients.
푇푂퐹
푐푙푢푠푡푒푟 values were obtained from 0.3%
0.025% Pt samples on high surface area TiO2 supports while
Pt samples.
Density Functional Theory Calculations
All first-principles calculations were performed with the Vienna Ab initio Simulation Package
(VASP52) with the projector-augmented wave method.53 The electronic exchange-correlation energy was
described using the Perdew-Burke-Ernzerhof (PBE) functional generalized gradient approximation.54
Considering the self-interaction problem in the standard DFT formulation, the DFT + U
method was used by introducing a Hubbard U correction with a value of 3
eV acting on 3d electrons of Ti.55 The van der Waals interaction was included through
Grimme’s DFT-D3 semi-empirical method.56 A kinetic cutoff energy of 400 eV was used in all the
calculations. The energy and force convergence criteria were set to 10 eV and 0.02 eV -1, respectively.
A vacuum region of at least 15 Å was placed to eliminate interactions between the top and bottom of the
slab model.
To model the Ptiso structure on TiO2, we used a structural model comprised of an adsorbed Pt atom
coordinated to two lattice-oxygen at a bridging position on the TiO2 (101) surface. We included five TiO2
(101) atomic layers with all the atoms being fully relaxed. In comparison, to model the Pt clusters, Pt
(111) was chosen to represent the energetically most stable facet. The close-packed Pt (111) slab was
modeled with a (4 × 4) supercell composed of four atomic layers. The bottom two layers were fixed to
reflect the bulk, while the top two layers and molecules were allowed to relax. We used the Gamma point
and a (3 × 3 × 1) k-mesh for calculations of the Ptiso structure and the Pt(111) surface, respectively. The
heats of dissociative adsorption of H2 and adsorption of m-cresol were calculated according to the
following equations:
-5
Å
∆Ediss(H2) = E(2H*/Pt-slab) – [E(Pt-slab) + E(H2)]
∆E(m-cresol*) = E(m-cresol*/Pt-slab) – [E(Pt-slab) + E(m-cresol)]
13
ACS Paragon Plus Environment

ACS Catalysis
Page 14 of 18
1
2
3
4
Supporting Information
5
6
7
8
CO IR spectroscopy of Ptiso during CO desorption, comparison of activity for Pt/TiO2 catalysts and pure
TiO2 supports, extinction coefficient analysis, thermogravimetric analysis, DFT calculations for m-cresol
adsorption on Pt clusters and Ptiso.
9
Acknowledgements
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
PC acknowledges funding from the National Science Foundation (NSF) CAREER grant number
CBET-1554112 and UCSB. DER and BW acknowledge support from the US Department of
Energy, Office of Science BES grant number DE-SC0018284. FY is thankful for the support of
the China Scholarship Council in the funding the exchange program to study at the University of
Oklahoma.
References
1.
Taylor, H. S.; Armstrong, E. F., A theory of the catalytic surface. Proc. Royal Soc. Lon. A, 1925,
108, 105-111.
2.
3.
Boudart, M., Catalysis by Supported Metals. Adv. Catal., 1969, 20, 153-166.
Boudart, M.; Rumpf, F., The catalytic oxidation of CO and structure insensitivity. React. Kinet.
Catal. Lett. 1987, 35, 95-105.
4.
5.
Iglesia, E.; Soled, S. L.; Fiato, R. A., Fischer-Tropsch synthesis on cobalt and ruthenium. Metal
dispersion and support effects on reaction rate and selectivity. J. Catal. 1992, 137, 212-224.
Jacobsen, C. J. H.; Dahl, S.; Hansen, P. L.; Törnqvist, E.; Jensen, L.; Topsøe, H.; Prip, D. V.;
Møenshaug, P. B.; Chorkendorff, I., Structure sensitivity of supported ruthenium catalysts for
ammonia synthesis. J. Molec. Catal. A 2000, 163, 19-26.
6.
Rioux, R. M.; Song, H.; Hoefelmeyer, J. D.; Yang, P.; Somorjai, G. A., High-Surface-Area
Catalyst Design:ꢀ Synthesis, Characterization, and Reaction Studies of Platinum Nanoparticles in
Mesoporous SBA-15 Silica. J. Phys. Chem. B 2005, 109, 2192-2202.
7.
Liu, L.; Corma, A., Metal Catalysts for Heterogeneous Catalysis: From Single Atoms to
Nanoclusters and Nanoparticles. Chem. Rev. 2018, 118, 4981-5079.
8.
Flytzani-Stephanopoulos, M.; Gates, B. C., Atomically Dispersed Supported Metal Catalysts.
Annu. Rev. Chem. Biomol. Eng. 2012, 3, 545-574.
9.
Wang, A.; Li, J.; Zhang, T., Heterogeneous single-atom catalysis. Nat. Rev. Chem. 2018, 2, 65-
81.
10.
DeRita, L.; Dai, S.; Lopez-Zepeda, K.; Pham, N.; Graham, G. W.; Pan, X.; Christopher, P.,
Catalyst Architecture for Stable Single Atom Dispersion Enables Site-Specific Spectroscopic and
Reactivity Measurements of CO Adsorbed to Pt Atoms, Oxidized Pt Clusters, and Metallic Pt
Clusters on TiO2. J. Am. Chem. Soc. 2017, 139, 14150-14165.
11.
Hoffman, A. S.; Fang, C.-Y.; Gates, B. C., Homogeneity of Surface Sites in Supported Single-
Site Metal Catalysts: Assessment with Band Widths of Metal Carbonyl Infrared Spectra. The J.
Phys. Chem. Lett. 2016, 7, 3854-3860.
12.
13.
Resasco, J., Uniformity is key in defining structure-function relationships for atomically
dispersed catalysts: the case of Pt/CeO2. J. Am. Chem. Soc. (Under Revision) 2019.
Resasco, J.; Dai, S.; Graham, G. W.; Pan, X.; Christopher, P. Combining In-Situ Transmission
Electron Microscopy and Infrared Spectroscopy for Understanding Dynamic and Atomic-Scale
Features of Supported Metal Catalysts. J. Phys. Chem. C 2018, 122, 25143-25157
Huber, G. W.; Iborra, S.; Corma, A., Synthesis of Transportation Fuels from Biomass:ꢀ
Chemistry, Catalysts, and Engineering. Chem. Rev. 2006, 106, 4044-4098.
13.
14
ACS Paragon Plus Environment

Page 15 of 18
ACS Catalysis
1
2
3
4
5
6
7
8
9
14.
15.
16.
Pham, T. N.; Shi, D.; Resasco, D. E., Evaluating strategies for catalytic upgrading of pyrolysis oil
in liquid phase. Appl. Catal. B 2014, 145, 10-23.
Robinson, A. M.; Hensley, J. E.; Medlin, J. W., Bifunctional Catalysts for Upgrading of Biomass-
Derived Oxygenates: A Review. ACS Catal. 2016, 6, 5026-5043.
Omotoso, T. O.; Baek, B.; Grabow, L. C.; Crossley, S. P., Experimental and First-Principles
Evidence for Interfacial Activity of Ru/TiO2 for the Direct Conversion of m-Cresol to Toluene.
ChemCatChem 2017, 9, 2612-2612.
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.
18.
19.
Wang, C.; Mironenko, A. V.; Raizada, A.; Chen, T.; Mao, X.; Padmanabhan, A.; Vlachos, D. G.;
Gorte, R. J.; Vohs, J. M., Mechanistic Study of the Direct Hydrodeoxygenation of m-Cresol over
WOx-Decorated Pt/C Catalysts. ACS Catal. 2018, 8, 7749-7759.
Ren, H.; Yu, W.; Salciccioli, M.; Chen, Y.; Huang, Y.; Xiong, K.; Vlachos, D. G.; Chen, J. G.,
Selective Hydrodeoxygenation of Biomass-Derived Oxygenates to Unsaturated Hydrocarbons
using Molybdenum Carbide Catalysts. ChemSusChem 2013, 6, 798-801.
Shetty, M.; Murugappan, K.; Prasomsri, T.; Green, W. H.; Román-Leshkov, Y., Reactivity and
stability investigation of supported molybdenum oxide catalysts for the hydrodeoxygenation
(HDO) of m-cresol. J. Catal. 2015, 331, 86-97.
Lee, W.-S.; Wang, Z.; Wu, R. J.; Bhan, A., Selective vapor-phase hydrodeoxygenation of anisole
to benzene on molybdenum carbide catalysts. J. Catal. 2014, 319, 44-53.
de Souza, P. M.; Rabelo-Neto, R. C.; Borges, L. E.; Jacobs, G.; Davis, B. H.; Sooknoi, T.;
Resasco, D. E.; Noronha, F. B., Role of keto intermediates in the hydrodeoxygenation of phenol
over Pd on oxophilic supports. ACS Catal. 2015, 5, 1318-1329.
Duong, N.; Tan, Q.; Resasco, D. E., Controlling phenolic hydrodeoxygenation by tailoring metal–
O bond strength via specific catalyst metal type and particle size selection. C. R. Chim. 2018, 21,
155-163.
Nie, L.; Resasco, D. E., Kinetics and mechanism of m-cresol hydrodeoxygenation on a Pt/SiO2
catalyst. J. Catal. 2014, 317, 22-29.
Tan, Q.; Wang, G.; Long, A.; Dinse, A.; Buda, C.; Shabaker, J.; Resasco, D. E., Mechanistic
analysis of the role of metal oxophilicity in the hydrodeoxygenation of anisole. J. Catal. 2017,
347, 102-115.
Tan, Q.; Wang, G.; Nie, L.; Dinse, A.; Buda, C.; Shabaker, J.; Resasco, D. E., Different product
distributions and mechanistic aspects of the hydrodeoxygenation of m-cresol over platinum and
ruthenium catalysts. ACS Catal. 2015, 5, 6271-6283.
Nelson, R. C.; Baek, B.; Ruiz, P.; Goundie, B.; Brooks, A.; Wheeler, M. C.; Frederick, B. G.;
Grabow, L. C.; Austin, R. N., Experimental and Theoretical Insights into the Hydrogen-Efficient
Direct Hydrodeoxygenation Mechanism of Phenol over Ru/TiO2. ACS Catal. 2015, 5, 6509-6523.
de Souza, P. M.; Nie, L.; Borges, L. E.; Noronha, F. B.; Resasco, D. E., Role of oxophilic
supports in the selective hydrodeoxygenation of m-cresol on Pd catalysts. Catal. Lett. 2014, 144,
2005-2011.
20.
21.
22.
23.
24.
25.
26.
27.
28.
29.
Madon, R. J.; Boudart, M., Experimental criterion for the absence of artifacts in the measurement
of rates of heterogeneous catalytic reactions. Ind. Eng. Chem. Fund. 1982, 21, 438-447.
de Souza, P. M.; Rabelo-Neto, R. C.; Borges, L. E.; Jacobs, G.; Davis, B. H.; Resasco, D. E.;
Noronha, F. B., Hydrodeoxygenation of phenol over Pd catalysts. Effect of support on reaction
mechanism and catalyst deactivation. ACS Catal. 2017, 7, 2058-2073.
Brunelle, J., Preparation of catalysts by metallic complex adsorption on mineral oxides. Pure
Appl. Chem. 1978, 50, 1211-1229.
Thang, H. V.; Pacchioni, G.; DeRita, L.; Christopher, P., Nature of stable single atom Pt catalysts
dispersed on anatase TiO2. J. Catal. 2018, 367, 104-114.
30.
31.
32.
DeRita, L.; Resasco, J.; Dai, S.; Boubnov, A.; Thang, H. V.; Hoffman, A. S.; Ro, I.; Graham, G.
W.; Bare, S. R.; Pacchioni, G.; Pan, X.; Christopher, P., Structural evolution of atomically
dispersed Pt catalysts dictates reactivity. Nat. Mater. 2019, 18, 746-751.
15
ACS Paragon Plus Environment

ACS Catalysis
Page 16 of 18
1
2
3
4
5
6
7
8
9
33.
34.
Oudenhuijzen, M.; Kooyman, P.; Tappel, B.; van Bokhoven, J. A.; Koningsberger, D.,
Understanding the influence of the pretreatment procedure on platinum particle size and particle-
size distribution for SiO2 impregnated with [Pt2+(NH3)4](NO3 )2: A combination of HRTEM,
−
mass spectrometry, and Quick EXAFS. J. Catal. 2002, 205, 135-146.
Hoffman, A. S.; Debefve, L. M.; Zhang, S.; Perez-Aguilar, J. E.; Conley, E. T.; Justl, K. R.;
Arslan, I.; Dixon, D. A.; Gates, B. C., Beating Heterogeneity of Single-Site Catalysts: MgO-
Supported Iridium Complexes. ACS Catal. 2018, 8, 3489-3498.
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
35.
36.
37.
Matsubu, J. C.; Yang, V. N.; Christopher, P., Isolated Metal Active Site Concentration and
Stability Control Catalytic CO2 Reduction Selectivity. J. Am. Chem. Soc. 2015, 137, 3076-3084.
Gonzo, E.; Boudart, M., Catalytic hydrogenation of cyclohexene: 3. Gas-phase and liquid-phase
reaction on supported palladium. J. Catal. 1978, 52, 462-471.
Davis, S.; Somorjai, G., Correlation of cyclohexene reactions on platinum crystal surfaces over a
ten-order-of-magnitude pressure range: Variations of structure sensitivity, rates, and reaction
probabilities. J. Catal. 1980, 65, 78-83.
Rossell, M. D.; Caparrós, F. J.; Angurell, I.; Muller, G.; Llorca, J.; Seco, M.; Rossell, O.,
Magnetite-supported palladium single-atoms do not catalyse the hydrogenation of alkenes but
small clusters do. Catal. Sci. Tech. 2016, 6, 4081-4085.
Liu, L.; Meira, D. M.; Arenal, R.; Concepcion, P.; Puga, A. V.; Corma, A., Determination of the
Evolution of Heterogeneous Single Metal Atoms and Nanoclusters under Reaction Conditions:
Which Are the Working Catalytic Sites? ACS Catal. 2019, 9, 10626-10639.
Kuo, C.-T.; Lu, Y.; Kovarik, L.; Engelhard, M.; Karim, A. M., Structure Sensitivity of Acetylene
Semi-Hydrogenation on Pt Single Atoms and Subnanometer Clusters. ACS Catalysis 2019, 9,
11030-11041.
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., Isolated metal atom geometries as a strategy
for selective heterogeneous hydrogenations. Science 2012, 335, 1209-1212.
Doudin, N.; Yuk, S. F.; Marcinkowski, M. D.; Nguyen, M.-T.; Liu, J.-C.; Wang, Y.; Novotny, Z.;
Kay, B. D.; Li, J.; Glezakou, V.-A.; Parkinson, G. S.; Rousseau, R.; Dohnalek, Z., Understanding
Heterolytic H2 Cleavage and Water-Assisted Hydrogen Spillover on Fe3O4(001)-Supported
Single Palladium Atoms. ACS Catal. 2019, 9, 7876-7887.
Lucci, F. R.; Liu, J.; Marcinkowski, M. D.; Yang, M.; Allard, L. F.; Flytzani-Stephanopoulos, M.;
Sykes, E. C. H., Selective hydrogenation of 1,3-butadiene on platinum–copper alloys at the
single-atom limit. Nat. Commun. 2015, 6, 8550.
Hemmingson, S. L.; Campbell, C. T., Trends in Adhesion Energies of Metal Nanoparticles on
Oxide Surfaces: Understanding Support Effects in Catalysis and Nanotechnology. ACS Nano
2017, 11, 1196-1203.
Campbell, C. T.; Parker, S. C.; Starr, D. E., The Effect of Size-Dependent Nanoparticle
Energetics on Catalyst Sintering. Science 2002, 298, 811-814.
38.
39.
40.
41.
42.
43.
45.
46.
47.
48.
49.
Campbell, C. T.; Mao, Z., Chemical potential of metal atoms in supported nanoparticles:
dependence upon particle size and support. ACS Catal. 2017, 7, 8460-8466.
Campbell, C. T., The Energetics of Supported Metal Nanoparticles: Relationships to Sintering
Rates and Catalytic Activity. Acc. Chem. Res. 2013, 46, 1712-1719.
Wang, S.; Vorotnikov, V.; Sutton, J. E.; Vlachos, D. G., Brønsted–Evans–Polanyi and Transition
State Scaling Relations of Furan Derivatives on Pd(111) and Their Relation to Those of Small
Molecules. ACS Catal. 2014, 4, 604-612.
Carey, S. J.; Zhao, W.; Mao, Z.; Campbell, C. T., Energetics of Adsorbed Phenol on Ni(111) and
Pt(111) by Calorimetry. J. Phys. Chem. C 2019, 123, 7627-7632.
Mehta, P.; Greeley, J.; Delgass, W. N.; Schneider, W. F., Adsorption Energy Correlations at the
Metal–Support Boundary. ACS Catal. 2017, 7, 4707-4715.
50.
51.
52.
Kresse, G.; Furthmüller, J., Efficiency of ab-initio total energy calculations for metals and
semiconductors using a plane-wave basis set. Comput. Mater. Sci. 1996, 6, 15-50.
16
ACS Paragon Plus Environment

Page 17 of 18
ACS Catalysis
1
2
3
4
5
6
7
8
9
53.
54.
Blochl, P. E., Projector Augmented-Wave Method. Phys. Rev. B. 1994, 50, 17953-17979.
Perdew, J. P.; Burke, K.; Ernzerhof, M., Generalized Gradient Approximation Made Simple.
Phys. Rev. Lett. 1996, 77, 3865-3868.
Chen, H.-Y. T.; Tosoni, S.; Pacchioni, G., Adsorption of ruthenium atoms and clusters on anatase
TiO2 and tetragonal ZrO2 (101) surfaces: a comparative DFT study. J. Phys. Chem. C 2014, 119
(20), 10856-10868.
Grimme, S.; Antony, J.; Ehrlich, S.; Krieg, H., A consistent and accurate ab initio parametrization
of density functional dispersion correction (DFT-D) for the 94 elements H-Pu. The Journal of
chemical physics 2010, 132 (15), 154104.
55.
56.
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
ACS Paragon Plus Environment

ACS Catalysis
Page 18 of 18
1
2
3
4
5
6
7
8
9
Table of contents figure
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
18
ACS Paragon Plus Environment