Different Product Distributions and Mechanistic Aspects of the Hydrodeoxygenation of m-Cresol over Platinum and Ruthenium Catalysts

Article abstract of DOI:10.1021/acscatal.5b00765

Experimental measurements of the conversion of m-cresol over Pt and Ru/SiO₂ catalysts show very different product distributions, even when the reaction is conducted at similarly low conversions and the same operating conditions (300 °C, 1 atm). That is, although ring hydrogenation to 3-methylcyclohexanone is dominant over Pt, deoxygenation to toluene and C-C cleavage to C₁-C₅ hydrocarbons prevail over Ru. For understanding the differences in reaction mechanisms responsible for this contrasting behavior, the conversion of m-cresol over the Pt(111) and Ru(0001) surfaces has been analyzed using density functional theory (DFT) methods. The DFT results show that the direct dehydroxylation of m-cresol is unfavorable over the Pt(111) surface with an energy barrier of 242 kJ/mol. In turn, the calculations suggest that the reaction could proceed through a keto tautomer intermediate, which undergoes hydrogenation of the carbonyl group followed by dehydration to form toluene and water. At the same time, a low energy barrier for the ring hydrogenation path toward 3-methylcyclohexanone compared to the energy barrier for the deoxygenation path toward toluene over the Pt(111) surface is in agreement with the experimental observations, which show that 3-methylcyclohexanone is the dominant product over Pt/SiO₂ at low conversions. By contrast, the direct dehydroxylation of m-cresol becomes more favorable than the tautomerization route over the more oxophilic Ru(0001) surface. In this case, the deoxygenation path exhibits an energy barrier lower than that for the ring hydrogenation, which is also in agreement with experimental results that show higher selectivity to the deoxygenation product toluene. Finally, it is proposed that a partially unsaturated hydrocarbon surface species C₇H₇ is formed during the direct dehydroxylation of m-cresol over Ru(0001), becoming the crucial intermediate for the C-C bond breaking products C₁-C₅ hydrocarbons, which are observed experimentally over the Ru/SiO₂ catalyst.

Full text of DOI:10.1021/acscatal.5b00765

Subscriber access provided by UNIV OF NEBRASKA - LINCOLN
Article
Different Product Distributions and Mechanistic Aspects of the
Hydrodeoxygenation of m-Cresol over Platinum and Ruthenium Catalysts
Qiaohua Tan, Gonghua Wang, Lei Nie, Arne Dinse, Corneliu Buda, John Shabaker, and Daniel E. Resasco
ACS Catal., Just Accepted Manuscript • DOI: 10.1021/acscatal.5b00765 • Publication Date (Web): 14 Sep 2015
Downloaded from http://pubs.acs.org on September 20, 2015
Just Accepted
“Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted
online prior to technical editing, formatting for publication and author proofing. The American Chemical
Society provides “Just Accepted” as a free service to the research community to expedite the
dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts
appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been
fully peer reviewed, but should not be considered the official version of record. They are accessible to all
readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered
to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published
in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just
Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor
changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers
and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors
or consequences arising from the use of information contained in these “Just Accepted” manuscripts.
ACS Catalysis is published by the American Chemical Society. 1155 Sixteenth Street
N.W., Washington, DC 20036
Published by American Chemical Society. Copyright © American Chemical Society.
However, no copyright claim is made to original U.S. Government works, or works
produced by employees of any Commonwealth realm Crown government in the course
of their duties.

Page 1 of 59
ACS Catalysis
1
2
3
4
5
6
7
8
9
Different Product Distributions and Mechanistic Aspects of the
Hydrodeoxygenation of m-Cresol over Platinum and Ruthenium
Catalysts
1
1
1
2
2
2
Qiaohua Tan, Gonghua Wang, Lei Nie, Arne Dinse, Corneliu Buda, John Shabaker,
1
*
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
and Daniel E. Resasco
(
1) School of Chemical, Biological and Materials Engineering,
University of Oklahoma, Norman, OK 73019
(
2) Catalyst Discovery Laboratory, BP Products North America Inc., Naperville, IL 60563
Abstract
Experimental measurements of the conversion of mꢀcresol over Pt and Ru/SiO2
catalysts show very different product distributions, even when the reaction is conducted
at similarly low conversions and the same operating conditions (300°C, 1 atm). That is,
while ring hydrogenation to 3ꢀmethylcyclohexanone is dominant over Pt, deoxygenation
to toluene and CꢀC cleavage to C ꢀC hydrocarbons prevail over Ru. To understand the
differences in reaction mechanisms responsible for this contrasting behavior, the
conversion of mꢀcresol over the Pt(111) and Ru(0001) surfaces has been analyzed using
density functional theory (DFT) methods. The DFT results show that the direct
dehydroxylation of mꢀcresol is unfavorable over the Pt(111) surface, with an energy
barrier for the 242 kJ/mol. In turn, the calculations suggest that the reaction could
proceed through a keto tautomer intermediate, which undergoes hydrogenation of the
carbonyl group followed by dehydration to form toluene and water. At the same time, a
low energy barrier for the ring hydrogenation path towards 3ꢀmethylcyclohexanone
compared to the energy barrier for the deoxygenation path towards toluene over the
Pt(111) surface is in agreement with the experimental observations, which show that 3ꢀ
1
5
methylcyclohexanone is the dominant product over Pt/SiO at low conversions. By
2
contrast, the direct dehydroxylation of mꢀcresol becomes more favorable than the
tautomerization route over the more oxophilic Ru(0001) surface. In this case, the
deoxygenation path exhibits an energy barrier lower than that for the ring hydrogenation,
which is also in agreement with experimental results that show higher selectivity to the
deoxygenation product, toluene. Finally, it is proposed that a partially unsaturated
hydrocarbon surface species C H * is formed during the direct dehydroxylation of mꢀ
7
7
cresol over Ru(0001), becoming the crucial intermediate for the CꢀC bond breaking
products, C ꢀC hydrocarbons, which are observed experimentally over the Ru/SiO
2
1
5
catalyst.
Keywords:
Hydrodeoxygenation, mꢀcresol, hydrogenation, tautomerization, platinum, ruthenium,
density functional theory
__________________________
_
*
Corresponding author. Email address: resasco@ou.edu
1
ACS Paragon Plus Environment

ACS Catalysis
Page 2 of 59
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
1
. Introduction
Biomass pyrolysis is a promising method being widely explored to displace
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
petroleumꢀderived liquid fuels. Unfortunately, the liquid product generated via biomass
pyrolysis (bioꢀoil) contains highly oxygenated species and could not be used as a
transportation fuel without catalytic upgrading to reduce its oxygen content. Among the
recalcitrant oxygenated species present in bioꢀoil, phenolic compounds formed from the
thermal breakdown of lignin are major components. Significant research efforts have
been directed to the development of catalysts for hydrodeoxygenation (HDO) of phenolic
1ꢀ7
compounds such as phenol, cresol, and guaiacol. Traditional hydrotreating catalysts
7
ꢀ9
composed of metal sulfides such as NiMoS or CoMoS have been tested. However, they
exhibit low HDO activity and require coꢀfeeding of sulfur to prevent catalyst degradation.
Supported metal catalysts that do not suffer from these limitations, including noble
metals such as Pt, Pd, Rh and Ru as well as nonꢀnoble metals such as Ni and Fe, have
2ꢀ4,10ꢀ13
become more attractive.
Several studies have shown that the composition of the
metal catalyst can influence the activity and selectivity of HDO reactions of phenolic
1
0
compounds. For instance, Mortensen et al. investigated the phenol HDO on different
carbonꢀsupported catalysts and found that the order of activity for phenol HDO is Ru/C >
Pd/C > Pt/C, which follows the trend of metal oxophilicity. Our recent study of mꢀcresol
HDO shows that not only the activity, but also the product distribution greatly vary over
3
SiO ꢀsupported Ni, Fe, and NiꢀFe bimetallic catalysts. That is, over Ni, the
2
hydrogenation product 3ꢀmethylcyclohexanone is dominant, while the HDO product,
toluene, appears in small amounts, even at high conversions. By contrast, over Fe and
2
ACS Paragon Plus Environment

Page 3 of 59
ACS Catalysis
1
2
3
4
5
6
7
8
9
NiꢀFe bimetallic catalysts, the dominant product is toluene while the hydrogenation
products are practically negligible in the entire range of conversions. Significant effects
of the support on activity and selectivity have also been reported. For example, reducible
oxide supports such as ZrO and TiO can enhance the selectivity to toluene in the HDO
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
of mꢀcresol over Pt and Ru catalysts compared to those supported on inert supports such
2
,10
as SiO or carbon.
2
Even though a number of studies have reported differences in activity and HDO
selectivity with varying metals and supports, the precise reaction mechanisms on the
different catalysts and the exact influence of each metal and/or support on the reaction are
still a matter of debate. In general, two types of mechanisms have been proposed for the
5
,14,15
HDO of phenolic compounds: (1) hydrogenation/deoxygenation (HYD) route
and
8,16,17
(2) direct deoxygenation (DDO) route via CꢀO hydrogenolysis,
as shown in Figure
1
(left). The HYD route typically starts with hydrogenation of the aromatic ring of
phenolic compounds to their corresponding saturated alcohols followed by the oxygen
removal via dehydration. The HYD route has been proposed to occur over biꢀfunctional
catalysts such as Ni/HZSMꢀ5 with the metal site providing hydrogenation activity to
saturate the aromatic ring and the acid site catalyzing the dehydration path. The DDO
route was proposed as the HDO pathway for the phenolic compounds on CoMo sulfide
1
0
catalysts at high temperature and further extended to metal catalysts.
Our group has recently investigated the HDO of mꢀcresol at relatively mild
temperatures (300ºC) on SiO ꢀsupported Pt catalyst, which yield 3ꢀmethylcyclohexanone,
2
2
,10
3
ꢀmethycyclohexanol and toluene as the major products.
Importantly, when the ringꢀ
hydrogenated intermediate 3ꢀmethylcyclohexanol was fed over this catalyst at the same
3
ACS Paragon Plus Environment

ACS Catalysis
Page 4 of 59
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
reaction conditions, no toluene was observed. That is, neither the DDO mechanism nor
the HYD mechanism can explain the behavior of this type of catalyst. The former
mechanism is believed to require a high energy barrier which will be investigated in the
following sections. Moreover, the absence of toluene production when feeding 3ꢀ
methylcyclohexanol contradicts the latter mechanism, which considers toluene as derived
from the dehydration of 3ꢀmethycyclohexanol. When 3ꢀmethylcyclohexanol was reacted
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
over the SiO support alone, the conversion was extremely low, reflecting the low acidity
2
2
of SiO . Thus, an alternative mechanism that starts with the tautomerization of mꢀcresol
2
and forms an unstable ketone intermediate (3ꢀmethylꢀ3,5ꢀcyclohexadienone) was
2
,3
proposed, as shown in Figure 1 (right) . This unstable ketone intermediate can be
hydrogenated at either the ring C=C bond or the carbonyl C=O bond. If the C=C
2
3
hydrogenation is dominant, which may occur over a Pt or Ni surface, the 3ꢀ
methylcyclohexanone and 3ꢀmethylcyclohexanol are the main products. However, if the
3
C=O hydrogenation is more favorable, as over Fe or NiꢀFe catalysts, the main product is
toluene, which is formed by the hydrogenation of the C=O group followed by
dehydration. This new tautomerization mechanism was further extended to the HDO of
phenolic compounds over other catalysts to explain the influence of the oxophilicity of
metals and reducible supports such as TiO as the more oxophilic metal and the reducible
2
,
1
2,18
oxide supports may facilitate the ꢀC=O bond hydrogenation and CꢀO activation.
4
ACS Paragon Plus Environment

Page 5 of 59
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 1. Different mechanisms proposed for the HDO reaction of phenolic compounds with mꢀcresol as
14ꢀ17
2,3
an example. (Left) DDO and HYD mechanisms
; (right) mechanism involving a keto tautomer
.
While this tautomerization mechanism can explain experimental results quite well,
in this contribution, we have further examined this possible pathway on two different
metals (Pt and Ru), using firstꢀprinciple theoretical methods. To this end, density
functional theory (DFT) has enabled us to understand the structure of the catalyst and the
chemical reactions at the atomic level, and thus has given a more inꢀdepth understanding
of how the specific atomic and electronic structure of the catalyst influence the catalytic
1
9,20
performance. In recent work, Lu et al.
have used DFT calculations to explore the
2
1
mechanisms of HDO of guaiacol over Pt(111) and Ru(0001) surfaces. Chiu et al. and
13
Lee et al.
have also investigated the plausible paths of the HDO of guaiacol over
Ru(0001) surface and Pt(111) surface separately. However, since catechol or phenol is the
main product for the HDO of guaiacol over Pt(111) and Ru(0001) surfaces, their
investigation mainly focused on the removal of CH Oꢀ and suggested that
3
dehydrogenation at the methyl group occurs before the CꢀO bond breaking. Hensley et
5
ACS Paragon Plus Environment

ACS Catalysis
Page 6 of 59
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
22
al. have also examined five different mechanisms for the HDO of phenol over Fe(110)
and Pd(111) surfaces by DFT calculations. They showed that the elementary reaction step
for each mechanism is endothermic on the Pd(111) surface while exothermic on Fe(110).
They also suggested that the direct cleavage of the CꢀO bond is likely to be the most
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
2
3
energetically and kinetically favorable on Fe(110) surface. Li et al. investigated the
hydrogenation of phenol over Pt(111) and Pd(111) surfaces by DFT calculations and
suggested that the phenol dissociation is crucial for the different product selectivity over
these two metal surfaces, although they only examined the first hydrogenation step at the
phenyl ring.
In this work, we have extended our efforts combining experimental HDO of mꢀ
cresol over SiO supported Pt and Ru catalysts with DFT calculations to compare the
2
fundamental mechanisms of this reaction, including the hydrodeoxygenation and the
hydrogenation paths over these two metal surfaces of different oxophilicity. We will
show how the nature of the metal catalyst influences the activity and selectivity of the
HDO reaction.
2
. Methods
2.1 Catalyst preparation and characterization
The 1 wt.% Pt/SiO and 9.4 wt.% Ru/SiO catalysts used in this study were prepared
2
2
2
by incipient wetness impregnation of the SiO support (HiSil 210, PPG, S
=135 m /g)
BET
2
with an aqueous solution of Pt(IV) chloride hydrate and Ru(III) chloride hydrate,
respectively. After impregnation, the catalysts were dried at 120 ºC for 12 h and calcined
at 400 ºC for 4 h. Before carrying out the reactions, the catalysts were pelletized, crushed,
and sieved to 250ꢀ420 ꢁm size range (Mesh no. 40ꢀ60). TEM images for the 1% Pt/SiO2
6
ACS Paragon Plus Environment

Page 7 of 59
ACS Catalysis
1
2
3
4
5
6
7
8
9
catalyst shown in previous work indicate an average Pt particle size of about about 3.5
2
nm. A TEM image for the 9.4% Ru/SiO catalyst is shown in Figure 2. The average
2
particle size for this catalyst was estimated to be 5 nm, as summarized in Table 1. The
dispersion of the Pt and Ru particles were estimated accordingly from the observed
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
2
4
particle sizes to be about 0.3 and 0.25, respectively.
Figure 2. TEM of 9.4 wt.% Ru/SiO after calcination at 400 ºC in air, reduction at 400 ºC by H and
2
2
passivation at room temperature.
Table 1. Characteristics of the Pt/SiO and Ru/SiO catalysts
2
2
Pt/SiO2
Ru/SiO2
Metal loading
1 wt. %
9.4 wt. %
dp
3.5 nm
0.3
5.2 nm
0.25
Metal Dispersion
2.2 Reaction conditions
The HDO reactions of mꢀcresol over the Pt/SiO and Ru/SiO catalysts were carried
2
2
out in a fixedꢀbed quartz tube reactor (6 mm outer diameter) under H flow at
2
atmospheric pressure. The flow reaction system is equipped with a mass flow controller
and a syringe pump for the continuous injection of mꢀcresol. The pelletized catalysts
7
ACS Paragon Plus Environment

ACS Catalysis
Page 8 of 59
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
were packed in the reactor between two layers of glass beads and quartz wool to improve
the temperature uniformity. The catalysts were preꢀreduced in flow of H (60 ml/min) for
2
1
h at 400 ºC. After reduction, the reaction was conducted at 300 ºC under the same H2
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
flow rate. A 0.5 ml/h flow of liquid mꢀcresol was fed continuously from the syringe pump
and vaporized into the H stream. All lines were heated to 250 ºC to avoid product
2
condensation. The reaction products were analyzed and quantified online by gas
chromatography using an HPꢀ5 column (30 m, 0.25 ꢁm) and a flame ionization detector
(FID). The selectivity for each product was calculated as:
Selectivity (%)
=
moles of carbon in the target product / moles of carbon in all the products
where the moles of carbon in the product are calculated as the number of carbon atoms in
the product molecule multiplied by the moles of product.
2
.3 Density Functional Theory Calculations
The calculations were carried out using periodic planeꢀwave gradientꢀcorrected
density functional theory methods implemented in the Vienna ab initio Simulation
2
5ꢀ28
29,30
Package (VASP).
The optPBEꢀvdW
exchange correlation functional was used to
provide the nonꢀlocal gradientꢀcorrections to exchange and correlation energies with the
correction of van der Walls (vdW) interactions. Projector augmented wave potentials
31
(PAW) within a cutoff energy of 400 eV were used for wave functions.
The closeꢀpacked 4x4 Pt(111) and Ru(0001) surfaces, with four metal layers and 15
Å of vacuum separating slabs in the zꢀdirection were used to model the Pt and Ru
catalysts. The top two metal layers were allowed to relax in the calculations, whereas the
bottom two layers were held fixed to their bulk position. The electronic energies were
8
ACS Paragon Plus Environment

Page 9 of 59
ACS Catalysis
1
2
3
4
5
6
7
8
9
ꢀ6
converged to within 10 eV. The geometric structures for all of reactants, intermediates
and products were optimized until the forces were converged to within 0.01 eV/ Å. A
3
2
2
×2×1 kꢀpoint mesh was used to sample the first Brillouin zone. The energies of the
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
adsorbates in vacuum were calculated spinꢀpolarized using an 18×18×18 Å unit cell with
the Γ only kꢀpoint mesh.
3
3
Transition state searches were performed using the dimer method with the initial
guesses for the transition state structure and the reaction trajectory obtained through the
3
4
nudged elastic band (NEB) method. The located transition states were further verified
by vibrational frequency analysis.
The Gibbs free energies for the adsorption of closeꢀshell species at 300 ºC and 1 atm
were estimated by including the entropy, which in turn was calculated according to
35
standard procedures. That is, the vibrational entropy was calculated from vibrational
frequencies derived from DFT frequency calculations. The translational and rotational
entropy for adsorbed species were assumed to be zero. For gas phase species, the
35
translational and rotational entropies were calculated by statistical thermodynamics.
3
. Results and Discussions
3.1 Product distribution over Pt/SiO and Ru/SiO catalysts
2
2
The product selectivity observed during the conversion of mꢀcresol over Pt/SiO and
2
Ru/SiO catalysts at 300 ºC is summarized in Table 2. Over Pt/SiO , the mꢀcresol
2
2
conversion yielded more 3ꢀmethylcyclohexanone (60.2%) than toluene (33.3%). This
result is consistent with our previous work that showed faster ring hydrogenation than
2
hydrodeoxygenation over pure Pt catalysts. In contrast, the Ru/SiO catalyst favored
2
HDO over ring hydrogenation, with a selectivity to toluene (38.5%) much higher than
9
ACS Paragon Plus Environment

ACS Catalysis
Page 10 of 59
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
that to 3ꢀmethylcyclohexanone (7.4%). That is, at comparable conversion levels (5ꢀ7 %),
the HDO/HYD ratio is only 0.5 on Pt but greater than 5 on Ru. At the same time,
significant amounts of CH (43.4%) and C ꢀC hydrocarbons (10.7%) were formed on Ru
4
2
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
via CꢀC bond hydrogenolysis. However, these products were essentially absent over
Pt/SiO under the same reaction conditions and comparable conversion levels. Turnover
2
frequencies (TOF) were calculated from the observed conversion, W/F, and estimated
ꢀ1
metal dispersions for Pt/SiO and Ru/SiO catalysts. Pt/SiO (TOF=19 min ) appears as
2
2
2
ꢀ1
significantly more active than Ru/SiO (TOF= 0.1 min ).
2
To further understand how the C ꢀC hydrocarbons were formed, we used the two
1
6
HYD products, 3ꢀmethylcyclohexanone and 3ꢀmethylcyclohexanol, as feed in two
separate sets of experiments as shown in Figure 3. It is found that CH , a product of the
4
CꢀC hydrogenolysis reaction over Ru/SiO is only observed in very small amounts when
2
the feed was 3ꢀmethylcyclohexanone or 3ꢀmethylcyclohexanol. It is important to note
that CH is clearly a secondary product of both 3ꢀmethylcyclohexanone and 3ꢀ
4
methylcyclohexanol as the rates of CH formation are zero at low conversions (i.e. as
4
W/F approaches zero). By contrast, CH is a primary product when mꢀcresol is the feed,
4
as demonstrated by the high selectivity to CH at low mꢀcresol conversion (Table 2).
4
Interestingly, mꢀcresol is a primary product when feeding the ketone 3ꢀ
methylcyclohexanone, but is a secondary product when feeding the alcohol 3ꢀ
methylcyclohexanol. This suggests that mꢀcresol is directly formed from the ketone 3ꢀ
methylcyclohexanone rather than the alcohol 3ꢀmethylcyclohexanol. However, in both
cases, the CꢀC cracking product CH is directly formed from mꢀcresol rather than from its
4
hydrogenation products 3ꢀmethylcyclohexanone or 3ꢀmethylcyclohexanol over Ru/SiO2
10
ACS Paragon Plus Environment

Page 11 of 59
ACS Catalysis
1
2
3
4
5
6
7
8
9
catalyst. This is an important result that illustrates the radically different reaction paths
that occur over Pt and Ru catalysts. It is clear that, over Pt catalyst, the hydrogenation
reaction path is more favorable, while over the more oxophilic Ru catalyst, the
hydrodeoxygenation and CꢀC breaking paths are more favorable. This observation
suggests that an intermediate may exist over Ru that is ready for CꢀC breaking
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
immediately following the CꢀO cleavage, but this does not occur over Pt.
Table 2. Product distributions for the HDO of mꢀcresol over Pt/SiO and Ru/SiO catalysts
2
2
at 300 ºC and atmospheric pressure.
Pt/SiO2
Ru/SiO2
W/F (gcat.hr/mol)
4.0
35.3
Conversion (%)
7.2
19
4.5
0.1
-1
TOF (min )
Selectivity (%)
3
3
-methylcyclohexanone
-methylcyclohexanol
60.2
6.5
7.4
ꢀ
Toluene
33.3
38.5
C -C hydrocarbon
CH4
ꢀ
ꢀ
10.7
43.4
2
5
Figure 3. Conversion and yields of products of feeds of 3ꢀmethylcyclohexanone (left) and 3ꢀ
o
methylcyclohexanol (right) over Ru/SiO as a function of W/F at 300 C with H at the pressure of 1 atm.
2
2
11
ACS Paragon Plus Environment

ACS Catalysis
Page 12 of 59
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
In the following sections, we describe the density functional theory (DFT)
calculations that we have performed to examine the mꢀcresol hydrogenation and
hydrodeoxygenation pathways over Pt and Ru metal surfaces and explain the
experimental observations.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
3
3
.2 DFT exploration of pathways for hydrogenation and hydrodeoxygenation
.2.1 Adsorption of m-cresol on the Pt(111) and Ru(0001) surfaces
Different adsorption configurations of mꢀcresol on the Pt(111) and Ru(0001)
surfaces were analyzed by DFT. As illustrated in Figure 4, the energyꢀoptimized
configuration shows that mꢀcresol adsorbs flat on ‘bridge 30’ sites on both Pt(111) and
Ru(0001) (i.e. through the aromatic ring carbons). This configuration is similar to
previously reported favorable adsorption modes of benzene and phenol on these, as well
3
6ꢀ39
as other, metal surfaces.
The adsorption energy of mꢀcresol on Pt(111) is calculated to
be ꢀ176 kJ/mol with the vdW corrections as listed in Table 3, which is similar to the
37
reported binding energy of phenol on the Pt(111) surface (ꢀ172 kJ/mol). Similarly, the
adsorption energy of mꢀcresol on a Ru(0001) surface is calculated to be ꢀ239 kJ/mol,
which is similar to the reported adsorption energy of phenol on Ru(0001) surface (ꢀ242
19
kJ/mol). The stronger adsorption of mꢀcresol on Ru(0001) compared to Pt(111) is
probably due to a stronger carbonꢀmetal interaction on the Ru(0001) surface. A similar
trend has been reported for benzene, which binds more strongly on Ru(0001) than on
4
0
Pt(111).
12
ACS Paragon Plus Environment

Page 13 of 59
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 4. DFT optimized structures of mꢀcresol and its two keto tautomers in gas phase, and on the Pt(111)
and Ru(0001) surfaces. The energies shown in the top and bottom are the reaction energies for the
tautomerization of mꢀcresol to each of its keto tautomers (a and b, respectively).
Table 3. Adsorption energies (kJ/mol) of reactant, products, and closeꢀshell intermediates in the
hydrodeoxygenation process of mꢀcresol on Pt(111) and Ru(0001) surfaces with vdW corrections. KetCO is
the keto tautomer of mꢀcresol, and KetHOH is the intermediate formed by hydrogenation of KetCO
tautomer at the carbonyl group. Free adsorption energies ꢂG were calculated at 300 ºC and 1 atm.
Pt(111)
Ru(0001)
ꢂE
ꢀ176
ꢀ208
ꢀ116
ꢀ170
ꢀ38
ꢂG
ꢀ50
ꢀ83
ꢀ5
ꢀ44
ꢀ
ꢂE
ꢀ239
ꢀ301
ꢀ165
ꢀ225
ꢀ48
ꢂG
ꢀ130
ꢀ185
ꢀ63
ꢀ121
ꢀ
mꢀcresol
KetCO
ꢀmethylcyclohexanone
3
Toluene
H
H O
2
ꢀ37
ꢀ
ꢀ47
ꢀ
KetHOH
ꢀ232
ꢀ85
ꢀ242
ꢀ101
13
ACS Paragon Plus Environment

ACS Catalysis
Page 14 of 59
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
3
.2.2 Direct deoxygenation of m-cresol over the Pt(111) surface
As summarized above, the direct deoxygenation mechanism (DDO) for the HDO of
phenolic compounds including mꢀcresol has been proposed in several experimental
studies. Some theoretical calculations have also been carried out on HDO of phenol over
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
2
2
22
41
13
Fe (110) and Pd(111) surfaces, as well as guaiacol over CoMoS, Pt(111) and
1
9,21
Ru(0001)
surfaces. Here, we further examine the direct deoxygenation of mꢀcresol
over the Pt(111) surface and compare it to other possible pathways. As shown in Figure 5,
mꢀcresol adsorbs flat on the Pt(111) surface through the aromatic ring. It must be noted
that the hydroxyl group ꢀOH does not interact with the metal surface as indicated by the
rather long PtꢀO distance (3.11 Å). On the other hand, the C ꢀOH bond distance in mꢀ
α
cresol is about 1.36 Å, which is shorter than a typical CꢀO bond in aliphatic alcohols
(~1.45 Å) and indicates a stronger CꢀO bond due to the conjugation with the aromatic
ring. However, the CꢀO bond gradually elongates along the reaction coordinate for the
direct dehydroxylation over Pt(111). As shown in Figure 5, in the transition state, the CꢀO
bond increases to 2.46 Å from the original 1.36 Å in the reactant. At the same time, the
hydroxyl group is bound to the metal surface at the atop site, which stabilizes the
transition state. After the CꢀO bond is cleaved to form the product structure, the hydroxyl
group changes from the atop site to the bridge site on the Pt(111) surface as the bridge
site binding mode is the most stable for the hydroxyl group on the Pt(111) surface. At the
same time, the aromatic ring tilts up about 45° from the flat adsorbed configuration. This
tilted configuration is about 23 kJ/mol more stable than the flat configuration, as shown
in Figure 5.
However, it is very important to note that the calculated activation energy for this
14
ACS Paragon Plus Environment

Page 15 of 59
ACS Catalysis
1
2
3
4
5
6
7
8
9
direct deoxygenation path over Pt(111) is exceedingly high (242 kJ/mol) and the reaction
energy is highly endothermic (116 kJ/mol). These energies confirm our previous
assumption that the breaking of the aromatic CꢀO bond is very difficult and may not
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
2,3
occur at all on Pt at the temperature (300 ºC) of the experimental HDO studies.
Figure 5. DFT optimized structures for reactant, transition state, and product (tilted) in the direct
dehydroxylation of mꢀcresol over the Pt(111) surface.
3
.2.3 Tautomerization of m-cresol to its keto tautomer
Based on our previous kinetic analyses of the HDO of mꢀcresol over SiO ꢀsupported
2
3
NiꢀFe and Pt catalysts, we propose a mechanism that involves the tautomerization of the
phenolic compound to their corresponding keto tautomer. This tautomer could undergo
ring hydrogenation (forming a hydrogenated ketone) or hydrogenation of the carbonyl
group followed by dehydration (forming the deoxygenated aromatic product). This
mechanism explains the experimental kinetic results. Here, we have used DFT
calculations to examine the reaction pathways involved in this mechanism.
It is wellꢀknown that phenol exhibits ketoꢀenol tautomerism with its unstable keto
tautomer (cyclohexadienone), even though only a tiny fraction may exist as the ketone
4
2
form. A similar ketoꢀenol tautomerism should exist for mꢀcresol. The structures of the
two keto tautomers of mꢀcresol (KetCO-a and KetCO-b) are shown in Figure 4 above.
The reaction energy for the ketoꢀenol tautomerization of mꢀcresol in the gas phase to
15
ACS Paragon Plus Environment

ACS Catalysis
Page 16 of 59
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
KetCO-a and KetCO-b is calculated to be 73 kJ/mol and 79 kJ/mol respectively, which
suggests the phenol form of mꢀcresol is much more stable than its keto tautomers due to
its aromaticity. Therefore, the keto tautomer can only be in a very small amount in the gas
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
42
phase, similar to the case of phenol. The calculations also suggest that KetCO-b with
the ꢀCH ꢀ at the ortho position of methyl group is more stable than KetCO-a, in which the
2
ꢀ
CH ꢀ is at the para position of methyl group. As shown in Figure 4, KetCO-b is not only
2
more stable than KetCO-a in the gas phase, but also on Pt(111) and Ru(0001) surfaces.
Thus, in this paper, we have only examined the reaction paths involving the more stable
tautomer KetCO-b of mꢀcresol. In the following sections this keto tautomer is referred as
KetCO, for simplicity.
On the Pt(111) surface, this keto tautomer (KetCO) may adsorb through the two
C=C bonds and form four carbonꢀmetal bonds with the metal surface, as shown in Figure
4
. However, the carbonyl group does not bind to the metal surface, but instead tilts away
from the surface with a carbonyl CꢀPt distance of 2.86 Å and a carbonyl OꢀPt distance of
.67 Å. This adsorption configuration is similar to the reported adsorption mode of
3
4
3
40
41
butadiene , furan, and furfural on Pd(111) and Pt(111) surfaces, all molecules
containing two C=C bonds. The adsorption energy of KetCO on the Pt(111) surface is
calculated to be ꢀ208 kJ/mol, which is stronger than the adsorption energy of mꢀcresol on
the Pt(111) surface, as shown in Table 3. The reaction energy for the tautomerization of
mꢀcresol over Pt(111) surface is calculated to be 41 kJ/mol. That is, compared to the gas
phase, the keto tautomer KetCO is more stable on the Pt(111) surface, with a much less
endothermic tautomerization reaction.
Similarly, the keto tautomer KetCO adsorbs on Ru(0001) surface, also through the
16
ACS Paragon Plus Environment

Page 17 of 59
ACS Catalysis
1
2
3
4
5
6
7
8
9
two C=C double bonds. However, different from the Pt(111) surface, the ꢀC=O binds to
the metal surface in the ‘diꢀδ’ mode, as shown in Figure 4, due to the stronger
oxophilicity of Ru. The adsorption energy of KetCO on the Ru(0001) surface is
calculated to be ꢀ301 kJ/mol, which is also stronger than mꢀcresol on the Ru(0001)
surface, as shown in Table 3. Thus, the reaction energy for the tautomerization of mꢀ
cresol is calculated to be 11 kJ/mol, which is much less endothermic compared to the
Pt(111) surface and in the gas phase. This result suggests that the more oxophilic the
metal is, the more strongly the keto tautomer is stabilized on the metal surface.
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
We now discuss in detail the energy barrier for the tautomerization of mꢀcresol
2
2
over Pt(111) and Ru(0001) surfaces (Figure 6). Hensley et al. studied the
tautomerization of phenol over Fe(110) surface, and found a very high energy barrier
(196 kJ/mol). However, it is important to note that, in their study, the tautomerization
proceeded via direct H transfer from the hydroxyl group onto the adjacent carbon.
2
3
Recently, Li et al. suggested that the tautomerization on metal surfaces may proceed via
metalꢀassisted OꢀH dissociation, followed by metalꢀassisted CꢀH combination. We have
analyzed the metalꢀassisted tautomerization of mꢀcresol over Pt(111) and Ru(0001)
surfaces and the results are shown in Figure 6.
It can be observed that, over both Pt(111) and Ru(0001) surfaces, the OꢀH
dissociation of mꢀcresol is easy with energy barriers of 39 kJ/mol over the Pt(111) surface
and 45 kJ/mol over Ru(0001) to form an mꢀcresoxy intermediate and H*. This OꢀH
dissociation step is endothermic over Pt(111) with a reaction energy of 22 kJ/mol, while it
is very exothermic over the Ru(0001) surface with a reaction energy of ꢀ89 kJ/mol. After
the OꢀH bond is dissociated, the H* can then be combined with the βꢀC of the mꢀcresoxy
17
ACS Paragon Plus Environment

ACS Catalysis
Page 18 of 59
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
intermediate to form the ketone tautomer KetCO. The energy barrier for this CꢀH
combination step is calculated to be 77 kJ/mol over the Pt(111) surface and 80 kJ/mol
over Ru(0001). As shown in Figure 6, the overall barrier for this metalꢀassisted
tautomerization process is about 99 kJ/mol over Pt(111) surface and about 45 kJ/mol over
Ru(0001) surface, which are much lower than the 196 kJ/mol barrier calculated by
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
2
2
Hensley et al. for the direct tautomerization path over Fe(110). The lower overall barrier
for this tautomerization process over Ru(0001) than over Pt(111) also suggests that an
oxophilic metal not only stabilizes the keto tautomer more efficiently, but also lowers the
energy barrier for the tautomerization reaction.
Figure 6. DFT optimized structures for reactant, transition state and product in the metal assisted
tautomerization reactions of mꢀcresol over Pt(111) surface (top) and Ru(0001) surface (bottom). The
energies listed are in reference to the reactant.
3
.2.4 Hydrogenation and hydrodeoxygenation of the keto tautomer over Pt(111)
Four sequential hydrogenation steps are required to saturate the two C=C bonds of
18
ACS Paragon Plus Environment

Page 19 of 59
ACS Catalysis
1
2
3
4
5
6
7
8
9
keto tautomer KetCO and form the hydrogenated product 3ꢀmethylcyclohexanone. The
different sequences of hydrogenation of the four different carbons lead to 24 different
pathways from KetCO to 3ꢀmethylcyclohexanone as shown in Figure S1. After
examination of the different pathways, it was found that the pathway for hydrogenating
KetCO in the order C3ꢀC4ꢀC5ꢀC6 is the most favorable as shown in Figure 7. This result
is consistent with the reported hydrogenation of furan on Pd(111), in which the most
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
44
favorable hydrogenation pathway of the four carbons follows the order C1ꢀC2ꢀC3ꢀC4.
The structures for the intermediates in the stepꢀwise hydrogenation of KetCO (KetHCO,
KetH CO and KetH CO), as well as the final product 3ꢀmethylcyclohexanone on Pt(111),
2
3
are shown in Figure 7. We can see that in the structures of intermediates KetHCO and
KetH CO that are formed by addition of one H and two H, respectively, the carbonyl
2
group C=O does not bind to the metal surface, as indicated by the long OꢀPt distances
(3.64 ꢀ 4.07 Å) and the CꢀPt distance of ~2.90 Å, similar to that in KetCO. By contrast, in
the subsequent KetH CO and 3ꢀmethylcyclohexanone species formed by hydrogenation
3
of three carbons and four carbons, respectively, the carbonyl group C=O binds to the
metal surface through the carbonyl O due to the saturation of the carbons in the ring that
releases the steric constraint. The transition state structures for each hydrogenation step
are shown in Figure 7. All hydrogenation steps proceed through the metal insertion
mechanism. In each case, the adsorbed H inserts into the Cꢀmetal bond forming an HꢀCꢀ
metal triangular transition state. The calculated energy barriers and reaction energies for
each hydrogenation step are included in Figure 7. The activation barriers for the four
hydrogenation steps are similar at 93~107 kJ/mol. These four hydrogenation steps are
endothermic with the reaction energy of 22 kJ/mol, 11 kJ/mol, 3 kJ/mol and 15 kJ/mol
19
ACS Paragon Plus Environment

ACS Catalysis
Page 20 of 59
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
respectively, which is consistent with the reported energy trends for hydrogenation of
4
4,45
furan and benzene over Pt(111) and Pd(111) surfaces.
After 3ꢀmethylcyclohexanone is formed, it desorbs with a desorption energy of
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
116 kJ/mol (Table 3). It must be noted that the vdW interaction contributes significantly
to this energy compared to desorption energies calculated without vdW corrections, such
23
as that for cyclohexanone (26 kJ/mol) on Pt. It is noted that even though the calculated
desorption energy for 3ꢀmethylcyclohexanone is high (116 kJ/mol), the desorption
process may be entropy driven. When entropy is considered, the free energy of 3ꢀ
methylcyclohexanone adsorption (at 300 ºC and 1 atm) becomes only ꢀ5 kJ/mol as shown
in Table 3.
The overall energy diagrams for the hydrogenation of the KetCO intermediate is
summarized in Figure 8. Starting at the adsorbed keto tautomer KetCO and adsorbed
hydrogen (KetCO*+4H*) the overall energetics is uphill along the reaction coordinate,
with the last hydrogenation step transition state at the highest energy. The resulting
energy barrier for the hydrogenation of the KetCO intermediate is 143 kJ/mol. The
overall energy barrier, including the tautomerization step, from the adsorbed mꢀcresol and
adsorbed hydrogen (Cre*+4H*) to the final hydrogenation product (3ꢀ
methylcyclohexanone) is calculated to be 181 kJ/mol. Of course, even though the
tautomerization reaction and each hydrogenation step over a Pt(111) surface are
endothermic, the heat of adsorption of the reactants makes the overall reaction from gas
phase mꢀcresol and H is exothermic, with a reaction energy of ꢀ121 kJ/mol, as expected.
2
20
ACS Paragon Plus Environment

Page 21 of 59
ACS Catalysis
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 7. DFT optimized structures for the intermediates, transition state, and product in the stepꢀwise
hydrogenation of KetCO at the ring carbons over the Pt(111) surface.
Figure 8. Energy diagrams for the hydrogenation of mꢀcresol to form 3ꢀmethylcyclohexanone (red) and the
hydrodeoxygenation of mꢀcresol to form toluene (blue) through the keto tautomer KetCO over the Pt(111)
surface.
In addition to the hydrogenation of the keto tautomer KetCO at the carbon ring that
21
ACS Paragon Plus Environment

ACS Catalysis
Page 22 of 59
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
leads to 3ꢀmethylcyclohexanone, the keto tautomer KetCO can also be hydrogenated at
the carbonyl group C=O. In this case, subsequent dehydration would form the HDO
product toluene. Therefore, we have also examined this path with DFT calculations.
First, we must note that the hydrogenation of the carbonyl group can proceed by two
alternative pathways: 1) through a hydroxylꢀalkyl intermediate by first H addition to the
O of the carbonyl group; or 2) through an alkoxide intermediate by first H addition to the
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
4
6
C of the carbonyl group. However, the formation of the alkoxide would require the four
ring C that bind to the metal surface to all tilt away, resulting in a high reaction energy of
1
77 kJ/mol. Therefore, the more favorable pathway for the hydrogenation of KetCO at
the carbonyl group is through the hydroxylꢀalkyl intermediate, followed by the second
hydrogenation at the carbonyl C. In the first hydrogenation of the carbonyl O of KetCO,
the tilted away carbonyl group gradually gets closer to the metal surface. The distance
between the carbonyl C decreases from 2.88 Å in the reactant to 2.56 Å in the transition
state, as shown in Figure 9A. At the same time, the H atom adsorbed on the metal surface
shifts from the fcc hollow site to the atop site and binds to the carbonyl O of KetCO. This
reaction has a low energy barrier (26 kJ/mol) and an exothermic reaction energy (ꢀ16
kJ/mol). The subsequent hydrogenation of the hydroxylꢀalkyl intermediate at the C
proceeds through the metal insertion mechanism, as shown in Figure 9B, similar to the
hydrogenation of KetCO at the ring carbons described above. The energy barrier is
calculated to be 88 kJ/mol, which is higher than the first hydrogenation step at the
carbonyl oxygen. This reaction is also endothermic with a reaction energy of 29 kJ/mol.
It can be noted that in the intermediate (KetCHOH) that is formed by the hydrogenation
of the carbonyl group of the keto tautomer KetCO, the hydroxyl group also tilts away
22
ACS Paragon Plus Environment

Page 23 of 59
ACS Catalysis
1
2
3
4
5
6
7
8
9
from the metal surface rather than binding to the metal surface due to the constraints
caused by the adsorption of the ring carbons, as shown in Figure 9B.
After the intermediate KetCHOH is formed by hydrogenation of the carbonyl group
of the keto tautomer KetCO, it can then undergo dehydration to remove the hydroxyl
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
group and H at the βꢀcarbon as H O to form toluene, as shown in Figure 9C. The CꢀOH
2
and βꢀC–H bonds gradually elongate, with the CꢀOH bond changing from 1.45 Å to 1.89
Å and the CꢀH bond changing from 1.11 Å to 1.36 Å. Simultaneously, the distance
between the hydroxyl group and the H at the βꢀcarbon gradually decreases from 2.25 Å in
the reactant to 1.31 Å in the transition state. After the CꢀOH bond and the βꢀC–H bond
are completely broken, a H O molecule is formed, leaving toluene adsorbed at the metal
2
surface, as shown in Figure 9C. Toluene desorbs with a desorption energy of 170 kJ/mol
(Table 3). This reaction is exothermic with a reaction energy of ꢀ69 kJ/mol because two
stable molecules (water and toluene) are formed as products. However, the calculated
energy barrier is rather high (152 kJ/mol). If KetCHOH desorbed and reꢀadsorbed with
the OH group facing the metal surface, the dehydration of KetCHOH would have a lower
energy barrier. However, the desorption energy of KetCHOH is calculated to be 232
kJ/mol, as listed in Table 3, which is even higher than the energy barrier for the direct
dehydration path of KetCHOH shown above (152 kJ/mol).
Therefore, the dehydration step rather than the hydrogenation steps is likely to be
rateꢀlimiting in the HDO process over Pt(111) surface (Figure 8). However, in any case,
the overall energy barrier relative to the adsorbed mꢀcresol and adsorbed hydrogen (Cre*
+
4H*) is about 203 kJ/mol, which is still about 39 kJ/mol lower than the energy barrier
required for the direct deoxygenation pathway (242 kJ/mol). This tautomerizationꢀ
23
ACS Paragon Plus Environment

ACS Catalysis
Page 24 of 59
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
hydrogenationꢀdehydration path is more favorable than the direct deoxygenation path for
the HDO of mꢀcresol over the Pt(111) surface, which supports our previous assumptions.
The energy diagram for the hydrogenation of mꢀcresol towards 3ꢀ
methylcyclohexanone can be compared in Figure 8 with the energy diagrams for the mꢀ
cresol HDO towards toluene. It is shown that the energy barrier for the deoxygenation of
mꢀcresol towards toluene is about 22 kJ/mol higher than that for hydrogenation of mꢀ
cresol towards 3ꢀmethylcyclohexanone, which is in agreement with the experimental
observations of higher selectivity to 3ꢀmethylcyclohexanone rather than toluene over
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Pt/SiO catalysts.
2
Figure 9. DFT optimized structures for the reactant, transition state and product over a Pt(111) surface in
the reaction of (A) hydrogenation of KetCO at the carbonyl oxygen; (B) continuing hydrogenation of
KetCO at the carbonyl carbon; (C) dehydration of the KetCHOH intermediate formed from hydrogenation
of KetCO.
3
.2.5 Hydrogenation and hydrodeoxygenation of keto tautomer over Ru(0001)
While the mechanism involving the keto tautomer KetCO over Pt(111) surface is in
agreement with experimental observations, the same may not be necessarily true over
24
ACS Paragon Plus Environment

Page 25 of 59
ACS Catalysis
1
2
3
4
5
6
7
8
9
other metals. Therefore, we examined on a Ru(0001) surface the same pathways for
hydrogenation of the keto tautomer as those studied on Pt(111). Similar to the case of Pt,
the most favorable path for hydrogenation of KetCO at the ring carbons on Ru(0001)
follows the sequence of C3ꢀC4ꢀC5ꢀC6. However, as shown in Figure 10, an important
differences between the two metals are observed in the structures for the intermediates
and transition states involved in the sequential hydrogenation of KetCO (KetHCO,
KetH CO and KetH CO), as well as that for the final product 3ꢀmethylcyclohexanone. In
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
3
contrast to Pt(111), over Ru(0001) the carbonyl group ꢀC=O of all the three intermediates
and final product binds to the surface in a ‘diꢀσ’ mode due to the stronger oxophilicity of
Ru. Each hydrogenation step also proceeds via the metal insertion mechanism, with the
formation of a HꢀCꢀmetal triangular transition state. The activation energies for these four
sequential hydrogenation steps are calculated to be 87, 81, 75, and 90 kJ/mol,
respectively, which are similar to those over Pt(111). These hydrogenation reactions are
all endothermic with positive reaction energies, consistent with those over the Pt(111)
surface.
From the energy diagram in Figure 11, we can see that along the reaction
coordinate of the hydrogenation of keto tautomer KetCO at the ring carbons over
Ru(0001), the energetics is also uphill, with the last hydrogenation step transition state
having the highest energy, similar to the case over Pt(111). Thus, in reference to the state
with the lowest energy, (i.e. adsorbed mꢀcresol and the adsorbed hydrogen (Cre*+4H*)),
the overall energy barrier for the hydrogenation of mꢀcresol is calculated to be 224
kJ/mol, which is higher than that obtained over the Pt(111) surface (181 kJ/mol). This is
consistent with previous calculations for the benzene hydrogenation over Pt(111) and
25
ACS Paragon Plus Environment

ACS Catalysis
Page 26 of 59
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
Ru(0001) surfaces showing that the overall barrier over Ru(0001) is higher than over
4
0,47
Pt(111).
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Figure 10. DFT optimized structures for the intermediates, transition state, and product in the stepꢀwise
hydrogenation of KetCO at the ring carbons over Ru(0001).
Figure 11. Energy diagrams for the hydrogenation of mꢀcresol to form 3ꢀmethylcyclohexanone (red) and
the hydrodeoxygenation of mꢀcresol to form toluene (blue) through the keto tautomer KetCO over
Ru(0001) surface.
To complete the HDO reaction as done above for Pt, the pathways for the
26
ACS Paragon Plus Environment

Page 27 of 59
ACS Catalysis
1
2
3
4
5
6
7
8
9
hydrogenation of the keto tautomer KetCO at the carbonyl group followed by
dehydration to form toluene were also examined over Ru(0001). The corresponding
structures are summarized in Figure 12.
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
Similar to the case of Pt(111), the favorable path for hydrogenation of the
carbonyl group of KetCO on Ru(0001) is through the hydroxylꢀalkyl intermediate formed
by the first hydrogenation at the carbonyl O. The transition state for the first
hydrogenation at the carbonyl O moves from the bridge site to the atop site and, at the
same time, the adsorbed H inserts into the OꢀRu bond, as shown in Figure 12A. This
hydrogenation step is very endothermic with a reaction energy of 60 kJ/mol, more
endothermic than over Pt(111) (ꢀ16 kJ/mol). The energy barrier (136 kJ/mol) is also much
higher than on Pt(111) (26 kJ/mol). Clearly, the higher reaction energy and energy barrier
for the hydrogenation of the O in the carbonyl is due to the stronger interaction of the
carbonyl O with Ru(0001) than with Pt(111). This trend is consistent with the reported
4
8
hydrogenation of other ketones and aldehydes over Pt(111) and Ru(0001) surfaces.
Next, the structures of the reactant, transition state, and product for the continuing
hydrogenation of the hydroxylꢀalkyl intermediate (KetCOH*) are shown in Figure 12B.
In the transition state, the adsorbed H inserts into the CꢀRu bond and forms a HꢀCꢀRu
triangle, similar to the transition state structures of the hydrogenation of KetCO at the
ring carbons shown above for Pt. This reaction is also very endothermic, with a reaction
energy of 78 kJ/mol. The energy barrier is calculated to be 106 kJ/mol. Similar to Pt(111),
the intermediate (KetCHOH*) formed by hydrogenation of the carbonyl group adsorbs
on Ru(0001) with the –OH tilting away rather than binding to the metal surface.
After this intermediate (KetCHOH*) is formed it can dehydrate leading to toluene
27
ACS Paragon Plus Environment

ACS Catalysis
Page 28 of 59
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
on the metal surface and water in the gas phase, as shown in Figure 12C. The transition
state structure is similar to that on Pt(111), as shown in Figure 9C. This reaction over
Ru(0001) is even more exothermic (ꢀ145 kJ/mol) than over Pt(111) (ꢀ69 kJmol), due to
the stronger adsorption of the product toluene on a Ru(0001) surface (ꢀ225 kJ/mol), as
listed in Table 3. The energy barrier is calculated to be 120 kJ/mol, which is also about 30
kJ/mol lower than on Pt(111).
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Figure 12. DFT optimized structures for the reactant, transition state and product state over the Ru(0001)
surface in the reaction of (A) hydrogenation of KetCO at the carbonyl oxygen; (B) continuing
hydrogenation of KetCO at the carbonyl carbon; (C) dehydration of KetCOH intermediate formed from
hydrogenation of KetCO.
As shown in Figure 11, similar to Pt(111), all the energies go uphill, with the
transition state for the dehydration step at the highest energy. Thus, the overall energy
barrier referred to the adsorbed mꢀcresol and adsorbed hydrogen (Cre*+4H*) is about
2
83 kJ/mol, about 80 kJ/mol higher than that over Pt(111). Even though the energy
barrier for the dehydration step of KetCHOH is lower over Ru(0001), the more
endothemic nature of the hydrogenation step of the keto tautomer KetCO leads to a
28
ACS Paragon Plus Environment

Page 29 of 59
ACS Catalysis
1
2
3
4
5
6
7
8
9
higher overall energy barrier for the hydrodeoxygenation path to toluene on Ru(0001)
than over Pt(111).
Comparing the paths of the hydrogenation of mꢀcresol to 3ꢀmethylcyclohexanone
and the deoxygenation to toluene over the Ru(0001) surface through the keto tautomer
KetCO, one finds a higher overall energy barrier for the deoxygenation path (by 59
kJ/mol) than for the hydrogenation path. This result is in contradiction with the
experimental results that show a higher selectivity to toluene rather than 3ꢀ
methylcyclohexanone. Therefore, we will examine other possible paths over Ru(0001),
such as the direct dehydroxylation, which was unfavorable for Pt(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
3.2.6 Direct deoxygenation of m-cresol over Ru(0001)
21
19
Recently, Chiu et al.
and Lu et al.
have independently investigated the
deoxygenation of guaiacol over Ru(0001) using DFT calculations. They both have found
that the energy barrier for the direct CꢀO bond breaking of catecholate C H (OH)O is
6
4
22
rather low, about 100 kJ/mol. Hensley et al. also investigated the mechanism of the
deoxygenation of phenol on Fe(110) and found that a plausible path for the
deoxygenation of phenol on the oxophilic Fe(110) surface is the direct dehydroxylation.
They did not consider the hydrodeoxygenation path via the keto tautomer discussed here.
Therefore, we have analyzed the direct deoxygenation path of mꢀcresol over Ru(0001)
and compared the results with those of the keto tautomer path.
The structures for the reactant, transition state, and product for the direct
dehydroxylation of mꢀcresol over the Ru(0001) surface are shown in Figure 13A. Similar
to Pt(111), the dehydroxylation of mꢀcresol over Ru(0001) proceeds via the gradual
29
ACS Paragon Plus Environment