Hydrodeoxygenation of Phenol over Pd Catalysts. Effect of Support on Reaction Mechanism and Catalyst Deactivation

Article abstract of DOI:10.1021/acscatal.6b02022

This work investigates the effect of the type of support (SiO₂, Al₂O₃, TiO₂, ZrO₂, CeO₂, and CeZrO₂) on the performance of Pd-based catalysts for the hydrodeoxygenation of phenol at 573 K using a fixed-bed reactor. Product distribution is significantly affected by the type of support. Benzene was the major product over Pd/TiO₂ and Pd/ZrO₂; on the other hand, cyclohexanone was the main compound over Pd/SiO₂, Pd/Al₂O₃, Pd/CeO₂, and Pd/CeZrO₂. A reaction mechanism based on the tautomerization of phenol was proposed on the basis of DRIFTS experiments and catalytic tests with the intermediate products. The high selectivity to benzene over Pd/TiO₂ and Pd/ZrO₂ catalysts is likely due to the oxophilic sites of this support represented by incompletely coordinated Ti⁴⁺ and Zr⁴⁺ cations in close proximity to the periphery of metal particles. The greater interaction between oxygen in the keto-tautomer intermediate with oxophilic sites promotes the selective hydrogenation of C=O bond. Pd/SiO₂, Pd/Al₂O₃, Pd/TiO₂, and Pd/ZrO₂ catalysts significantly deactivated during TOS. However, Pd/CeO₂ and Pd/CeZrO₂ were more stable, and only slight losses in activity were observed. Carbon deposits were not detected by Raman spectroscopy after reaction. DRIFTS experiments under reaction conditions revealed a buildup of phenoxy and intermediate species during reaction. These species remained adsorbed on the Lewis acid sites, blocking those sites and inhibiting further reactant adsorption. The growth of Pd particle size and the reduction in acid site density during HDO of phenol were the primary routes of catalyst deactivation. The higher stability of Pd/CeO₂ and Pd/CeZrO₂ catalysts is likely due to the higher amount of oxygen vacancies of these supports. (Graph Presented).

Full text of DOI:10.1021/acscatal.6b02022

Subscriber access provided by UB + Fachbibliothek Chemie | (FU-Bibliothekssystem)
Article
Hydrodeoxygenation of phenol over Pd catalysts. Effect of
support on reaction mechanism and catalyst deactivation
Priscilla Magalhães de Souza, Raimundo C. Rabelo-Neto, Luiz E. P. Borges,
Gary Jacobs, Burtron H. Davis, Daniel E. Resasco, and Fabio B. Noronha
ACS Catal., Just Accepted Manuscript • DOI: 10.1021/acscatal.6b02022 • Publication Date (Web): 02 Feb 2017
Downloaded from http://pubs.acs.org on February 3, 2017
Just Accepted
“Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted
online prior to technical editing, formatting for publication and author proofing. The American Chemical
Society provides “Just Accepted” as a free service to the research community to expedite the
dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts
appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been
fully peer reviewed, but should not be considered the official version of record. They are accessible to all
readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered
to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published
in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just
Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor
changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers
and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors
or consequences arising from the use of information contained in these “Just Accepted” manuscripts.
ACS Catalysis is published by the American Chemical Society. 1155 Sixteenth Street
N.W., Washington, DC 20036
Published by American Chemical Society. Copyright © American Chemical Society.
However, no copyright claim is made to original U.S. Government works, or works
produced by employees of any Commonwealth realm Crown government in the course
of their duties.

Page 1 of 57
ACS Catalysis
1
2
3
4
5
6
7
8
9
Hydrodeoxygenation of phenol over Pd catalysts. Effect
of support on reaction mechanism and catalyst
deactivation
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
1,2
1
2
3
Priscilla M. de Souza , Raimundo C. RabeloꢀNeto , Luiz E.P. Borges , Gary Jacobs ,
3
4
1,2,
Burtron H. Davis , Daniel E. Resasco , Fabio B. Noronha *
1
Catalysis Division, National Institute of Technology, Av. Venezuela 82, 20081ꢀ312, Rio
de Janeiro, RJ 20081ꢀ312, Brazil.
2
Chemical Engineering Department, Military Institute of Engineering, Praça Gal. Tiburcio
8
0, 22290ꢀ270, Rio de Janeiro, RJ, Brazil.
3
Center for Applied Energy Research, University of Kentucky, 2540 Research Park Dr.,
Lexington, KY 40511, USA.
4
Center for Biomass Refining, School of Chemical, Biological, and Materials Engineering,
The University of Oklahoma, Norman, OK 73019, USA
*
Corresponding Author: email: fabio.bellot@int.gov.br
Submitted to ACS Catalysis
Revised
January 2017
1
ACS Paragon Plus Environment

ACS Catalysis
Page 2 of 57
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
Abstract
This work investigates the effect of the type of support (SiO , Al O , TiO , ZrO ,
2
2
3
2
2
CeO
of phenol at 573 K using a fixed bed reactor. Product distribution is significantly affected
by the type of support. Benzene was the major product over Pd/TiO and Pd/ZrO ; on the
other hand, cyclohexanone was a major product over Pd/SiO , Pd/Al , Pd/CeO and
2 2
and CeZrO ) on the performance of Pdꢀbased catalysts for the hydrodeoxygenation
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
2
2
O
3
2
Pd/CeZrO . A reaction mechanism based on the tautomerization of phenol was proposed
2
based on DRIFTS experiments and catalytic tests with the intermediate products. The high
2 2
selectivity to benzene over Pd/TiO and Pd/ZrO catalysts is likely due to the oxophilic
4
+
4+
sites of this support represented by incompletely coordinated Ti and Zr cations in close
proximity to the periphery of metal particles. The greater interaction between oxygen in the
ketoꢀtautomer intermediate with oxophilic sites promotes the selective hydrogenation of
C=O. Pd/SiO
2
, Pd/Al
2
O
3
, Pd/TiO
2
2
and Pd/ZrO catalysts significantly deactivated during
TOS. However, Pd/CeO
2
and Pd/CeZrO
2
were more stable and only slight losses in
activity were observed. Carbon deposits were not detected by Raman spectroscopy after
reaction. DRIFTS experiments under reaction conditions revealed a buildꢀup of phenoxy
and intermediate species during reaction. These species remained adsorbed on the Lewis
acid sites, blocking those sites and inhibiting further reactant adsorption. The growth of Pd
particle size and the reduction in acid site density during HDO of phenol were the primary
routes of catalyst deactivation. The higher stability of Pd/CeO
2 2
and Pd/CeZrO catalysts is
likely due to the higher amount of oxygen vacancies of these supports.
Keywords: Phenol; Bioꢀoil; Hydrodeoxygenation; Stability; Pd catalysts; Oxophilic sites.
2
ACS Paragon Plus Environment

Page 3 of 57
ACS Catalysis
1
2
3
4
5
6
7
8
9
1.
Introduction
The hydrodeoxygenation (HDO) of bioꢀoil process is a fundamental step in the
1, 2, 3
production of biofuels and chemicals from lignocellulosic biomass.
In recent years,
HDO has been explored for a number of molecules that are representative of families of
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
4
compounds derived from biomass. One important molecule is phenol, which is not only
present in the bioꢀoil, but is also an intermediate produced during HDO of different
molecules.
According to the literature, HDO of phenol is a bifunctional reaction that requires a
5
,6
metallic site and an acid site.
The metal particles are responsible for hydrogenation/
dehydrogenation reactions, whereas alkylation and polymerization reactions are catalyzed
by the Lewis acid sites. Then, the design of the proper catalyst for HDO of phenol requires
an understanding of the reaction pathways.
Various reaction routes have been suggested for phenol HDO.
Direct
2
deoxygenation (DDO) involves hydrogenolysis of the relatively stable C(sp )−O bond,
whose scission is difficult due to ring stabilization. Thus, under conventional conditions,
4
the probability that this path occurs is low. However, the reaction may happen on
oxophilic metals such as Ru as well as supports like TiO , whose strong interaction with
2
the oxygen atom reduces the energy barrier for the direct cleavage of the CꢀO bond of the
7
,8
aromatic ring. Another proposal is the metalꢀcatalyzed successive ring hydrogenation
(
HYD), which reduces the energy for C−O bond scission. After hydrogenation, acidꢀ
9 ꢀ 12
catalyzed cyclohexanol dehydration takes place on the acid sites.
This pathway
proceeds on catalysts containing sufficient acidity. Recently, we proposed that phenol
conversion likely involves tautomerization, producing cyclohexadienone as an
1
3
intermediate. Due to lack of stability, this intermediate may be hydrogenated to produce
2
ꢀcyclohexenꢀ1ꢀone, which is converted by hydrogen addition to the ketone, and then the
3
ACS Paragon Plus Environment

ACS Catalysis
Page 4 of 57
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
alcohol. In contrast, hydrogen addition to the carbonyl functional group of the
cyclohexadienone intermediate assisted by the metal produces 2,4ꢀcyclohexadienol, which
dehydrates to benzene on the acid sites. Thus, in this case, the mechanism depends on the
support. For supports containing oxophilic sites such as zirconia (incompletely coordinated
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
+
3+
Zr or Zr cations), the carbonyl group is selectively hydrogenated on metal sites at the
metalꢀsupport interface, increasing the probability for benzene formation. It is clear that the
type of support may drive the reaction by different pathways.
14ꢀ
Different catalysts and operating conditions have been used for HDO of phenol.
2
5
An optimal catalyst for upgrading of bioꢀoil should exhibit high activity for
26
deoxygenation as well as stability under reaction conditions.
Catalyst deactivation is currently one of the main challenges facing catalyst
developers for the HDO reaction and it is mainly due to carbon deposition or sintering of
2
7
metal particles. Carbon formation occurs via polymerization and polycondensation
reactions on the acid sites of the support, producing polyaromatic species that may block
2
8
the active sites. Therefore, the density and the nature of acid sites of the support play an
important role on catalyst deactivation during HDO of bioꢀoil. To date, only few studies
have been dedicated to catalyst deactivation during HDO of guaiacol, anisol and mꢀ
2
9,30
29
cresol.
Gao et al. investigated the mechanism of deactivation of carbon supported
catalysts for HDO of guaiacol. Pd/C, Rh/C and Ru/C strongly deactivated during 300 min
of time on stream (TOS) whereas Pt/C remained quite stable during reaction. Metal
crystallite growth was ruled out using TEM images. Thermogravimetric analysis (TGA)
showed that the amount of carbon deposits was the same on both catalysts. However, the
type of carbon was different. Ru/C deactivation was attributed to a relatively higher
30
concentration of naphthalene. GonzalezꢀBorja and Resasco studied the performance of
Pt, Sn and PtSn catalysts for HDO of guaiacol and anisole. Deactivation was observed over
4
ACS Paragon Plus Environment

Page 5 of 57
ACS Catalysis
1
2
3
4
5
6
7
8
9
Inconel supported catalysts and depended on the molecule. Guaiacol caused a more severe
catalyst deactivation than anisole, consistent with pulse experiments that revealed stronger
adsorption of guaiacol than anisole. Therefore, guaiacol is retained adsorbed on the
surface, hindering the deoxygenation sites and causing catalyst deactivation. However, to
date, there have been no studies reported concerning the deactivation of catalysts during
HDO of phenol.
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
This work aims at investigating the mechanism of catalyst deactivation during
2 2 3 2 2
HDO of phenol using supported Pd on various oxide carriers (SiO , Al O , TiO , ZrO ,
CeO and CeZrO ). The investigation of reaction pathways in combination with diffuse
2
2
reflectance infrared Fourier transform spectroscopy (DRIFTS) measurements, Raman
spectroscopy and model reactions were carried out to determine the causes of catalyst
deactivation.
2.
2.1 Catalyst synthesis
2 2 2 2
SiO , Al , TiO , ZrO , CeO and CeZrO were used to prepare supported Pd
Experimental
2
2
O
3
catalysts. SiO (HiꢀSil 915), Al O (Puralox) and TiO (Aeroxide TiO P25) supports were
2
2
3
2
2
provided by PPG Industries, Sasol and Evonik, respectively. Al
K at a rate of 10 K/min for 5 h. ZrO , CeO and CeZrO
3 2
or coꢀprecipitation method. For ZrO , a 2.0 M solution of 35 wt.% ZrO(NO )
O
2 3
was calcined at 1273
2
2
2
were synthesized by precipitation
(Sigmaꢀ
2
Aldrich) was added to a solution containing 4.0 M NH OH (Vetec) at room temperature. It
4
was stirred vigorously for half an hour. The precipitate formed was collected by filtration
and rinsed with distilled H
2
O until the pH was neutral. The preparation of CeO
2
followed
the same procedure as previously described for ZrO
2
but using ammonium cerium (IV)
nitrate ((NH
5
4
)
2
Ce(NO
)
3 6
2
, SigmaꢀAldrich) as the precursor salt. CeZrO mixed oxide with a
ACS Paragon Plus Environment

ACS Catalysis
Page 6 of 57
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
Ce/Zr molar ratio = 1 was prepared by coꢀprecipitation using zirconyl nitrate and
ammonium cerium (IV) nitrate solutions. The precipitates were dried for 12 h at 383 K and
calcined under flowing air at 773 K for 6 h and the ZrO
obtained. The 2 wt. % Pd catalysts (Pd/SiO , Pd/Al
Pd/CeZrO
aqueous solution of palladium nitrate (Pd(NO
93 K for a period of 12 h followed by calcination at 673 K for 3h in 50 mL/min of
2
, CeO
2
and CeZrO
2
supports were
2
2
O
3
, Pd/TiO
2
, Pd/ZrO
2
, Pd/CeO and
2
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
) were prepared by incipient wetness impregnation of the supports using an
, Umicore). The powder was then dried at
3 2
)
3
flowing air.
2.2. Catalyst characterization
The chemical composition of each sample was determined by Xꢀray fluorescence
XRF) in wavelength dispersive mode (S8 Tiger Bruker Spectrometer, Rh tube, 4 kW).
(
The samples (300 mg) were analyzed by a semiꢀquantitative method (QUANTꢀ
EXPRES/Bruker). Surface areas were measured using a Micromeritics ASAP 2020
instrument by nitrogen adsorption at 77 K. Xꢀray diffraction (XRD) was carried out in a
Rigaku Miniflex instrument with Cu Kα radiation using the following parameters: 2θ =
10ꢀ80º; scan rate = 0.04 º/step; and scan time = 1 s/step. The Raman spectra were recorded
using a Horiba LabRam HRꢀUV800/JobinꢀYvon Spectrometer, equipped with HeꢀNe laser
(λ = 632 nm) with 10 mW of intensity, a CCD detector and an Olympus BX41 microscope
with objective lens of 100×. The acid properties of the catalysts was investigated by
temperatureꢀprogrammed desorption of ammonia (NH ꢀTPD). The samples (350 mg) were
3
reduced at 573 K for 1 h under a flow of H
30 min. After reduction, the sample was cooled to 423 K and the feed composition was
switched to a mixture containing 4% NH in He (30 mL/min) for 30 min. The physisorbed
2
of 60 mL/min and then purged in He flow for
3
ammonia was flushed out with flowing He for 1 h. Then, the catalyst was heated at 10
6
ACS Paragon Plus Environment

Page 7 of 57
ACS Catalysis
1
2
3
4
5
6
7
8
9
K/min under He to 773 K. The density of acid sites of the catalysts before and after HDO
3
1
of phenol was measured by the cyclohexanol dehydration reaction. This probe reaction
was performed using a fixedꢀbed reactor made of quartz at 1 atm and 543 K. The fresh
catalyst was previously reduced with hydrogen (60 mL/min) at 573 K for a period of 1 h.
The reactant mixture was obtained by flowing He (30 mL/min) through a saturator
containing cyclohexanol, which was maintained at 336 K. A GCMS (Agilent
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
7890A/5975C) with capillary column (HPꢀInnowax) and flame ionization detector was
used to analyze the products of the reaction. The dehydration rate was calculated by the
sum of cyclohexene, cyclohexane and benzene yields. The nature of acid sites was
determined by DRIFTS of adsorbed pyridine using a Nicolet Nexus 870 instrument with a
DTGSꢀTEC detector and a Thermo SpectraꢀTech reaction chamber with ZnSe windows.
To reduce the sample, H was flowed at 573 K for a period of 1 h and the temperature was
2
decreased to 373 K while He was flowed to record a spectrum for use as a background.
Pyridine adsorption involved bubbling He (30 mL/min) through a saturator for 30 min,
ꢀ1
followed by a He purge. The scan resolution was 4 cm and 512 scans were taken. For
DRIFTS of adsorbed CO, each sample was subjected to the same H reduction treatment.
2
After cooling the catalyst to 473 K, a 1.3 % CO in He mixture was flowed for 25 min.
Spectra were recorded in flowing CO/He to measure formate band intensities by reaction
with defectꢀassociated OH groups with CO. For DRIFTS of adsorbed cyclohexanone, the
2
samples underwent the same H reduction treatment previously described. Then, the
catalyst was cooled to 323 K and He was flowed through a bubbler containing
cyclohexanone at 60 ml/min and spectra were recorded at 323, 373, 423, and 473K. Peak
fitting was performed on the band corresponding to adsorbed cyclohexanone by using
Gaussian peaks and the nonlinear generalized reduced gradient (GRG) algorithm. The
dehydrogenation of cyclohexane was used as an insensitive structure reaction to determine
7
ACS Paragon Plus Environment

ACS Catalysis
Page 8 of 57
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 2
the metal dispersion of supported Pd catalysts. For Pd/CeO and Pd/CeZrO catalysts,
more traditional techniques such as H or CO chemisorption are not recommended for
2
33,34
these catalysts due to the possibility of adsorption of both gases on ceria.
In addition,
the determination of Pd particle size by transmission electron microscopy (TEM) is
challenging due to the low contrast between the Pd and CeO and CeZrO supports.
Therefore, Pd dispersion of Pd/SiO , Pd/Al and Pd/ZrO catalysts of this work was
determined by chemisorption of carbon monoxide. Prior to chemisorption, the catalysts (50
mg) were activated with H (60 ml/min) at 573 K for 1h, and then the temperature was
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
2
2
O
3
2
2
decreased to room temperature. The catalysts were purged in helium for 30 min. Pulses of
a mixture containing 5%CO in He were injected until the saturation of the surface. The
gasses were analyzed by quadrupole mass spectrometry using an MKS Cirrus 200
instrument. Dispersion of Pd was calculated considering that the Pd:CO ratio was unity.
Then, a direct relationship between the cyclohexane dehydrogenation rate and the
dispersion of palladium was established and used to determine the Pd dispersion of the
other catalysts of this study. In order to validate this methodology, the Pd dispersion of
Pd/ZrO
2
catalyst was also obtained by high resolution transmission electron microscopy
35
(
HRꢀTEM) and scanning transmission electron microscopy (STEM) analyses. The results
obtained by both techniques were quite close, revealing that this procedure can be used for
the determination of metal dispersion. This procedure enables the measurement of the Pd
dispersion of fresh and used catalysts (after HDO of phenol for 24 h). A fixedꢀbed reactor
was used to carry out the dehydrogenation of cyclohexane reaction over fresh and used
catalysts. The fresh samples (10 mg) were first reduced at 573 K for 1 h and then cooled to
the cyclohexane dehydrogenation reaction temperature (543 K). The reaction mixture
ꢀ1
(WHSV = 170 h ) was fed into the reactor after bubbling H
2
through a saturator containing
cyclohexane maintained at 285 K (H
2 6
/C H12 = 13.2). For the used catalysts, after HDO of
8
ACS Paragon Plus Environment

Page 9 of 57
ACS Catalysis
1
2
3
4
5
6
7
8
9
phenol for 24 h, the saturator containing phenol was bypassed and hydrogen was allowed
to flow through the catalyst for 30 min at 573 K, which removed hydrocarbons that
remained adsorbed on the surface. The reactor was cooled to 543 K under hydrogen. Then,
2
a cyclohexane/H mixture was passed through the reactor. The effluent was analyzed by
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
GCMS (system previously described). To probe the reaction mechanism, experiments were
conducted in the inꢀsitu DRIFTS cell previously described for acidity measurements using
conditions representative of HDO. The sample (40 mg) was reduced under H (60 ml/min)
2
at 573 K and the temperature was decreased to 323 K to record a background spectrum.
Hydrogen was passed through a saturator containing phenol at 60 ml/min and spectra were
ꢀ1
recorded at 373, 473, 573, 673 and 773 K. Scan resolution was 4 cm and 1024 scans were
recorded at each point. The DRIFTS cell has a very small system volume, and the
residence time inside the chamber is only 3.4 s, such that steady state is achieved very
rapidly. To study catalyst deactivation, spectra were taken for 6h at steady state under
2
reaction conditions at 573 K using a H flow of 60 ml/min passing through the phenol
saturator as previously described.
2.3. Catalytic activity
A fixed bed reactor made of quartz was used for the gas phase conversion of the
various oxygenate compounds (i.e., phenol, cyclohexanone, or cyclohexanol). The reactor
was operated at 1 atm and 573 K. Preliminary tests were carried out following the same
criteria proposed by Madon and Boudart (1982) to ensure that external and internal mass
3
6
transfer limitations were eliminated. The powder catalysts (53ꢀ106 µm range) were
diluted with inert material (mSiC/mcatalyst = 3/1) to avoid localized overheating. Before
2
reaction, the catalyst was activated in H (60 mL/min) for 1 h at 573 K. Hydrogen was
9
ACS Paragon Plus Environment

ACS Catalysis
Page 10 of 57
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
passed through a bubbler containing the reactant (phenol, cyclohexanol, cyclohexanone),
which was kept at the temperature needed to achieve H /reactant molar ratio of 60.
2
Two different series of experiments were performed to determine the initial phenol
conversion and to study catalyst deactivation. In the first series, each catalyst was tested at
various W/F values by changing the amount of the catalyst (2.5 to 120 mg). W/F is catalyst
mass (g) divided by mass flow rate (g/h) of the organic feed. The initial phenol conversion
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
was taken after 5 min of TOS in order to avoid catalyst deactivation. In the second series,
ꢀ1
6
0 mg of catalyst and a total flow rate of 60 mL.min were used for all catalysts in order to
observe catalyst deactivation within 24 h TOS. To avoid condensing the gas, the tubing
was heat traced to 523 K. The GCMS used for analysis was previously described. Product
yield and selectivity were defined as:
mol of product produced
yield(%) =
×100
(1)
(2)
mol of phenol fed
mol of product produced
mol of phenol consumed
Selectivity (%) =
×100
10
ACS Paragon Plus Environment

Page 11 of 57
ACS Catalysis
1
2
3
4
5
6
7
8
9
3.1- Results and discussion
.1 Catalyst characterization
As shown in Table 1, Pd content varied from 1.83 to 2.46 wt.% and was close to the
3
2 2 3
expected value (2 wt.%). Pd/SiO and Pd/Al O samples exhibited the highest specific
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
surface areas (172 and 134 m /g, respectively), while the smallest ones were observed for
2
Pd/CeO
2
and Pd/TiO
2
(51 and 54 m /g, respectively).
Table 1. Pd loading, BET surface area, reaction rate of dehydrogenation of cyclohexane
DHO), Pd dispersion, reaction rate of dehydration of cyclohexanol (DHA) of fresh
catalysts and total amount of ammonia desorbed.
(
Rate of DHA
(mmol.g cat.min )
Ammonia
desorbed
(ꢀmol/gcat
[
a]
wt %
Pd
BET
(m /g)
Rate of DHO
D
ꢀ1
ꢀ1
Catalyst
Pd/SiO
Pd/Al
Pd/TiO2
Pd/ZrO
Pd/CeO
2
ꢀ1
ꢀ1
(mmol.g Pd.min )
(%)
)
2
1.92
2.28
1.83
2.24
2.44
2.46
172
134
54
0.22
15.9
1.4
83.0
31.4
12.6
1.9
8
2
O
3
0.25
0.91
0.57
0.62
0.55
18.4
63.3
40.4
43.6
38.7
339
225
114
27
2
84
2
51
Pd/CeZrO
2
82
1.2
56
[a]
Pd dispersion of the fresh catalyst.
The diffractograms of Pdꢀbased catalysts are provided in Figure S1 (Supporting
Information). With the exception of the Pd/SiO catalyst, the lines characteristic of PdO
were not observed, which could be attributed to their overlapping by the diffraction lines of
the supports or the small PdO particle size. The Pd/SiO catalyst also exhibited a broad
peak at 22.7° corresponding to amorphous SiO . The diffractogram of Pd/Al presented
diffraction lines corresponding to various phases of Al O : θꢀAl O (JCPDS 35ꢀ0121), δꢀ
2
2
2
2 3
O
2
3
2
3
Al
2
O
3
2 3 2
(JCPDS 46ꢀ1131), γꢀAl O (JCPDS 10ꢀ0425). The diffractogram of Pd/TiO
11
ACS Paragon Plus Environment

ACS Catalysis
Page 12 of 57
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
revealed lines characteristic of both anatase (JCPDS 21ꢀ1272) and rutile (JCPDS 21ꢀ1276)
titania phases. For Pd/ZrO catalyst, the diffractogram has the diffraction lines of two
2
zirconia phases, tetragonal ZrO
Pd/CeO and Pd/CeZrO catalysts showed the diffraction pattern of the cubic fluorite
structure (JCPDS 34ꢀ0394) typical of ceria. However, for Pd/CeZrO catalyst, the lines in
2
(JCPDS 17ꢀ0923) and monoclinic (JCPDS 37ꢀ1484).
2
2
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 2
the diffractogram are shifted to higher 2θ positions, suggesting that a CeO /ZrO solid
37
solution was formed.
The Pd dispersion calculated by the cyclohexane dehydrogenation reaction depended
on the support (Table 1). The Pd/TiO catalyst had the highest Pd dispersion whereas the
lowest values were observed for Pd/SiO and Pd/Al catalysts. Pd/ZrO , Pd/CeO and
Pd/CeZrO exhibited approximately the same dispersion.
2
2
2
O
3
2
2
2
The nature of acid sites was studied by DRIFTS of adsorbed pyridine (Figure 1). For
ꢀ1
Pd/Al
2 3
O , the band at 1593 cm is assigned to the vibrations of pyridine coordinated to
3+
ꢀ1
octahedral Al sites, whereas the band at 1615 cm corresponds to pyridine adsorption on
3
+
ꢀ1
tetrahedral Al cations. The bands at 1500 and 1449 cm can be attributed to physically
38
adsorbed pyridine.
12
ACS Paragon Plus Environment

Page 13 of 57
ACS Catalysis
1
2
3
4
5
6
7
8
9
1449
1615
1593
1500
(a)
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
1
602
1
583
(
b)
1442
1574
1
483
(c)
(
d)
e)
1591
1438
1
560
1478
(
1600
1550
1500
1450
1400
-1
Wavenumber (cm )
Figure 1. DRIFTS spectra of adsorbed pyridine at 373 K over (a) Pd/Al
2
3 2
O ; (b) Pd/TiO ;
(
2 2 2
c) Pd/ZrO ; (d) Pd/CeZrO ; (e) Pd/CeO .
2
The infrared spectrum after pyridine adsorption on Pd/TiO catalyst showed bands at
ꢀ1
1
446, 1475, 1487, 1583, and 1602 cm . These bands can be attributed to physically
ꢀ1
adsorbed pyridine (1446, 1475, 1487, 1583 cm ), and pyridine adsorbed on unsaturated
4
+
ꢀ1
Ti cations (1602 cm ). The DRIFTS spectra of adsorbed pyridine on Pd/ZrO
2 2
, Pd/CeO
and Pd/CeZrO were quite similar to that of the Pd/TiO catalyst. The main difference was
2
2
that the bands were shifted to lower wavenumber in the presence of Ce.
The DRIFTS spectra for adsorption of pyridine are dominated by Lewis acid sites.
Slight features that might indicate a minor contribution from Bronsted acid sites (range of
ꢀ1
1500 – 1550 cm ) might be present on Pd/ZrO
2 2
and Pd/TiO , but the features were within
13
ACS Paragon Plus Environment

ACS Catalysis
Page 14 of 57
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
the experimental limits of detection. Therefore, only the Lewis acid sites were considered
to explain the results on HDO of phenol.
The titration of acid sites of the catalysts was measured by two different methods:
temperatureꢀprogrammed desorption of NH
reaction. Table 1 reports the amount of desorbed ammonia calculated from the NH
profiles. Pd/Al catalyst exhibited the highest number of acid sites (339 ꢀmol/gcat
whereas Pd/SiO , Pd/CeO and Pd/CeZrO catalysts showed the lowest values.
The dehydration of cyclohexanol reaction was also used to measure the number of acid
3
(NH
3
ꢀTPD) and the cyclohexanol dehydration
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
ꢀTPD
2
O
3
)
2
2
2
3
131
sites of the supports. Martin and Duprez
proposed to investigate the acidꢀbase
properties of oxides using the dehydration of cyclohexanol reaction. They reported that the
dehydration of cyclohexanol to cyclohexene would likely be catalyzed by the acid sites. In
our work, the reaction rate of dehydration of cyclohexanol for fresh catalysts is listed in
Table 1. Pd/Al
higher density of acid sites. Pd/ZrO
Pd/CeO and Pd/CeZrO catalysts. The following order was observed for the density of
acid sites: Pd/Al > Pd/TiO > Pd/ZrO >> Pd/SiO ≈ Pd/CeO ≈ Pd/CeZrO . This result
is in agreement with the NH ꢀTPD experiment. In fact, both techniques revealed that the
density of acid sites is low and approximately the same for Pd/SiO , Pd/CeO and
Pd/CeZrO catalysts.
2
O
3
exhibited the highest rate of dehydration of cyclohexanol, suggesting a
2
showed a reaction rate 10ꢀfold higher than Pd/SiO
2
,
2
2
2
O
3
2
2
2
2
2
3
2
2
2
3.2 HDO of phenol over Pd-based catalysts
The HDO of phenol was performed over all supports. The supports displayed
insignificant activity in the W/F range studied. Thus, activity can be ascribed to the metal
14
ACS Paragon Plus Environment

Page 15 of 57
ACS Catalysis
1
2
3
4
5
6
7
8
9
surface and/or the metalꢀsupport junction. The phenol conversion and product yield versus
W/F for HDO of phenol at 573 K over all catalysts are shown in Figure S2.
Among the catalysts, Pd/TiO and Pd/ZrO
2
2 2
were the most active, whereas Pd/SiO ,
Pd/Al
2
O
3
, Pd/CeO , and Pd/CeZrO exhibited approximately the same conversion at the
2
2
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
conditions of the study. Regarding product distributions, benzene, cyclohexanone (ONE),
cyclohexanol (OL) and C12 hydrocarbons such as biphenyl and cyclohexylbenzene were
the main products observed. Cyclohexane and C12 Oꢀcontaining hydrocarbons (2ꢀ
phenylphenol, 2ꢀcyclohexylcyclohexanꢀ1ꢀone and 2ꢀcyclohexylphenol) were also detected
in trace amounts at high conversion on some catalysts. However, product distribution
significantly depended on the support.
Cyclohexanone was mainly formed at low conversion for Pd/Al
2 3 2
O , Pd/SiO ,
Pd/CeO and Pd/CeZrO catalysts. Higher formation of benzene only occurred at high
2
2
conversions. Minor production of cyclohexanol was also observed for the Pd/SiO
Pd/CeO and Pd/CeZrO catalysts. On the other hand, benzene was the dominant product
over the entire W/F range for Pd/TiO and Pd/ZrO . In this case, only small amounts of
2
,
2
2
2
2
cyclohexanone and C12 hydrocarbons were observed. C12 hydrocarbons were formed on all
catalysts, except Pd/SiO . The highest yield of C hydrocarbons was obtained on Pd/Al O
3
2
12
2
2
and Pd/TiO catalysts. C12 hydrocarbons were likely the product of alkylation through the
28,39
reaction of cyclohexanone with phenolic / aromatic rings over Lewis acid sites.
These
4
0
C
12 hydrocarbons are quite interesting as a fuel due to their high cetane number.
The reaction rate of HDO of phenol was calculated and the selectivity of all
catalysts was compared at low phenol conversion (around 10%), as shown in Table 2.
15
ACS Paragon Plus Environment

ACS Catalysis
Page 16 of 57
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
Table 2. Reaction rate of HDO of phenol and product distribution at low conversion.
Catalyst Conversion
(%)
Rate of HDO
Selectivity(%)
ꢀ1
ꢀ1
(mmol.min .gcat
)
BZ
8.1
ONE
88.9
84.1
33.2
68.7
85.8
84.4
OL
C
12
Pd/SiO
2
7.0
7.5
0.09
0.13
0.72
0.59
0.14
0.09
3.0
2.0
0.0
1.3
9.4
11.0
0.0
2.8
0.0
0.0
0.0
0.0
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
Pd/Al O
11.4
66.8
29.9
4.9
2
3
Pd/TiO
2
7.0
Pd/ZrO
2
12.6
9.0
Pd/CeO
2
Pd/CeZrO
2
12.4
4.6
BZ: benzene; ONE: cyclohexanone; OL: cyclohexanol; C : bicyclic compounds
.
12
Pd/TiO
catalysts showed basically the same values. However, for Pd/TiO
1.2ꢀfold higher than that of Pd/ZrO . The reaction rates displayed the following trend:
2
and Pd/ZrO
2
exhibited the highest reaction rates, whereas the other
2
this value was about
2
Pd/TiO
2
> Pd/ZrO
The dependency of the product selectivity on the support is remarkable. Pd/TiO
and Pd/ZrO exhibit a higher selectivity to benzene, whereas the other catalysts showed a
higher selectivity to cyclohexanone. For instance, Pd/ZrO catalyst showed selectivity to
benzene that was 6ꢀfold higher than that of either Pd/CeO or Pd/CeZrO catalysts. Thus,
2
> Pd/Al
2
O
3
≈ Pd/CeO
2
≈ Pd/SiO
2
≈ Pd/CeZrO
2
.
2
2
2
2
2
the type of support significantly impacts not only the activity for deoxygenation but also
the distribution of products.
The HDO reaction of cyclohexanol and cyclohexanone (potential intermediate
products) was also conducted on supports and catalysts to investigate the effect of support
on the reaction pathways. In addition, DRIFTS spectra were also obtained under reaction at
different temperatures.
16
ACS Paragon Plus Environment

Page 17 of 57
ACS Catalysis
1
2
3
4
5
6
7
8
9
3
.3. Cyclohexanol and cyclohexanone conversion over Pd-based catalysts
Table 3 reports the yields of products for phenol, cyclohexanol or cyclohexanone
HDO at the same conversion over Pd catalysts. When cyclohexanone was used as the feed,
phenol was the dominant product formed, although minor quantities of cyclohexanol and
benzene were also produced over all catalysts. This result indicates that benzene is a
secondary product from HDO of phenol. For Pd/Al O , Pd/TiO and Pd/ZrO , significant
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
2
2
formation of C12 hydrocarbons was also observed. Nimmanwudipong et al. also observed a
28
2 3
high yield of C12 hydrocarbons for the reaction of cyclohexanone over Pt/Al O catalysts.
This result was attributed to the Lewis acid sites of this support and indicates that C12
hydrocarbons are preferentially formed via conversion of cyclohexanone during HDO of
phenol. The conversion of cyclohexanol yielded predominantly phenol and cyclohexanone
over Pd/SiO
cyclohexane and cyclohexene was observed for Pd/Al
benzene was only produced over alumina supported catalyst.
2
, Pd/ZrO
2
, Pd/CeO
2
and Pd/CeZrO
2
catalysts. Significant formation of
2
O
3
and Pd/TiO catalysts, whereas
2
This result suggests that the cyclohexanol dehydration to cyclohexene may occur
when Al O and TiO are used as supports. To study the role of the support on the
2
3
2
dehydration of cyclohexanol, this reaction was performed over the pure supports at the
same reaction conditions used for the HDO of phenol over the supported Pd catalysts.
2 2 2 2
SiO , ZrO , CeO and CeZrO supports showed low conversion of cyclohexanol to
cyclohexene, which suggests that the activity of the supports for this reaction is not
significant. On the other hand, Al O and TiO supports were very active for cyclohexanol
2
3
2
dehydration. However, benzene cannot be formed via the dehydration of cyclohexanol on
Pd/TiO . In contrast, the formation of benzene by cyclohexanol dehydration cannot be
2
ruled out over Pd/Al
17
2 3
O .
ACS Paragon Plus Environment

ACS Catalysis
Page 18 of 57
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
.4. DRIFTS of HDO of phenol over Pd-based catalysts
The DRIFT spectra of experiments performed using the mixture of phenol and
hydrogen over Pd/SiO
temperature range studied, indicating that the adsorption of phenol or intermediates on the
catalyst is very low. For Pd/Al catalyst, the DRIFT spectra obtained under reaction
2
did not reveal the presence of significant bands over the entire
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
O
3
conditions at different temperatures are shown in Figure 2. The vibrational modes of the
41ꢀ46
various species formed from phenol adsorption were obtained from the open literature.
At 373 K, phenol is dissociatively adsorbed over metal oxide cations with bands
attributed to modes of vibration of unidentate or bidentate phenoxy species (1289, 1495,
ꢀ1
1
595, 3034, and 3065 cm ). These bands are related to the following vibrational mode
ꢀ1
ꢀ1
assignments: ν(CO) stretching (1289 cm ); ν(C=Cring) (1495 and 1595 cm ); and ν(C–H)
ꢀ1
stretching (3034 and 3065 cm ).
13
As discussed previously from experiments with probe molecules, the bands at
ꢀ1
2
942, 2857 and 1687 cm likely correspond to 2ꢀcyclohexenꢀ1ꢀone species. In this case,
ꢀ1
ꢀ1
the band typical of the ν(C=O) is shifted to 1687 cm (from 1715 cm ), which is caused
by conjugation of the carbonyl with the unsaturated bond of the ring.
18
ACS Paragon Plus Environment

Page 19 of 57
ACS Catalysis
1
2
3
4
5
6
7
8
9
2 2 3 2 2 2
Table 3. Performance comparison of Pd/SiO , Pd/Al O , Pd/TiO , Pd/ZrO , Pd/CeO and
Pd/CeZrO
2
catalysts under different feeds.
W/F Feed
Catalyst
Yield %
X
(h)
ANE+ENE BZ ONE OL Phenol C12 (%)
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
0.053 Phenol
0.024 ONE
ꢀ
ꢀ
2.9 16.8 1.6
ꢀ
ꢀ
ꢀ
ꢀ
ꢀ
21.3
28.4
23.8
3.0
Pd/SiO
2
0.3
2.9
ꢀ
ꢀ
7.6
ꢀ
0.9 27.2
0
.098
0.160
.160 Phenol
0.092 ONE
.006
0.160
.053 Phenol
ꢀ
ꢀ
ꢀ
13.3
ꢀ
OL
SiO2
3.0
0
0.1
1.2
6.6 17.5
2.6
13.1 2.1
ꢀ
ꢀ
ꢀ
ꢀ
ꢀ
9.5
ꢀ
4.0 28.2
8.4 21.6
Pd/Al
2
O
3
ꢀ
0
11.8
100
ꢀ
ꢀ
27.1
100
OL
Al
2
O
3
ꢀ
ꢀ
ꢀ
0
0.1
3.0
13.3 5.3
ꢀ
1.1
ꢀ
ꢀ
4.0 22.7
7.0 21.4
Pd/TiO
2
0.046 ONE
1.3
ꢀ
5.8
ꢀ
9.0
0.3
ꢀ
0.012
OL
14.2
94.7
ꢀ
ꢀ
ꢀ
ꢀ
20.6
94.7
TiO2
0.160
ꢀ
0
.026 Phenol
0.023 ONE
.024
0.160
.026 Phenol
0.005 ONE
.006
0.160
.053 Phenol
0.011 ONE
.012
0.160
ꢀ
8.7 10.1 0.4
ꢀ
0.4 19.6
9.0 22.1
Pd/ZrO
2
ꢀ
1.2
4.0
ꢀ
ꢀ
9.3
ꢀ
2.2
9.7
5.6
ꢀ
0
0.4
3.8
ꢀ
ꢀ
ꢀ
ꢀ
19.3
3.8
OL
ZrO
2
0
ꢀ
1.0 17.1 3.1
0.2
0.4 16.3
ꢀ
ꢀ
ꢀ
ꢀ
ꢀ
21.2
19.2
26.8
3.3
Pd/CeO
CeO2
2
ꢀ
ꢀ
4.7 14.3
0
0.3
3.3
ꢀ
ꢀ
9.9
ꢀ
OL
ꢀ
ꢀ
0
ꢀ
1.7 21.1 6.0
0.3
1.1 11.8
ꢀ
ꢀ
ꢀ
ꢀ
ꢀ
28.9
25.4
23.6
4.0
Pd/CeZrO
2
0.1
0.8
4.0
ꢀ
3.9 21.1
0
ꢀ
ꢀ
9.8
ꢀ
OL
CeZrO
2
ꢀ
ꢀ
ANE + ENE: cyclohexane + cyclohexene; BZ: benzene; ONE: cyclohexanone; OL:
cyclohexanol; C12: bicyclic hydrocarbon
19
ACS Paragon Plus Environment

ACS Catalysis
Page 20 of 57
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
595 1495
1
289
1
687
2
942
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
3065
a)
(
2
857
1
459
(
b)
(
c)
(
d)
(
e)
3500
3000
2500
2000
1500
1000
-1
Wavenumber (cm )
Figure 2. DRIFTS spectra under HDO of phenol over Pd/Al
73 K, (b) 473 K, (c) 573 K, (d) 673 K, (e) 773 K.
2 3 2
O (H /phenol = 60) at (a)
3
Increasing the temperature to 473 K, band intensities characteristic of phenoxy
species and 2ꢀcyclohexenꢀ1ꢀone species diminished. In addition, the appearance of a new
ꢀ
1
ꢀ1
band at 1459 cm and the increase in the intensity of the bands at 2937 and 2863 cm
suggests that phenoxy species reacted, producing cyclohexanol species. Further increasing
temperature to 773 K resulted in a decrease in band intensities of phenoxy and
cyclohexanol species, which is indicative of facile desorption from the surface.
2
For Pd/TiO , the IR spectrum at 373 K exhibited bands positioned at 1283, 1491,
ꢀ
1
1591 and 3064 cm that are attributed to the phenoxy vibrational modes (Figure 3).
20
ACS Paragon Plus Environment

Page 21 of 57
ACS Catalysis
1
2
3
4
5
6
7
8
9
2
937
1491
1
454
1283
1
591
2
859
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
3064
1
666
(a)
(b)
(c)
(d)
(
e)
3
500
3000
2500
2000
1500
1000
-1
Wavenumber (cm )
Figure 3. DRIFTS spectra under HDO of phenol over Pd/TiO (H /phenol = 60) at (a) 373
2
2
K, (b) 473 K, (c) 573 K, (d) 673 K, (e) 773 K.
ꢀ1
The bands at 1454, 1666, 2859 and 2937 cm indicate the presence of both
cyclohexanol and 2ꢀcyclohexenꢀ1ꢀone species. Raising temperature to 473 K caused the
intensities of all bands to diminish, and they were no longer observed at 573 K.
DRIFTS spectra of the phenol/H
2
mixture over Pd/ZrO
catalyst.
2
catalyst (Figure 4) showed
some different features from those of the Pd/Al
2
O
3
21
ACS Paragon Plus Environment

ACS Catalysis
Page 22 of 57
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
559
1
494 & 1457
417
342
291
1
592
1
1
1
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
1556
3025 2931
064
3
2
852
(
a)
(
b)
1
285
(
c)
(
d)
(
e)
3
500
3000
2500
2000
1500
100
-
1
Wavenumber (cm )
2 2
Figure 4. DRIFTS spectra under HDO of phenol over Pd/ZrO (H /phenol = 60) at (a) 373
K, (b) 473 K, (c) 573 K, (d) 673 K, (e) 773 K.
Bands characteristic of the vibrational modes of phenoxy species were detected
ꢀ1
(
3064, 3025, 1592, 1494, 1291 cm ), as well as bands positioned at 2931, 2852, 1697
ꢀ1
(
shoulder), 1457 and 1049 cm , demonstrating that adsorbed phenol reacted to produce
ꢀ
cyclohexenꢀ1ꢀone and cyclohexanol species. Additional bands at 1559, 1417 and 1342 cm
1
were also observed, and may be assigned to ν(OCO) of phenol adsorbed on the support,
4
7
analogous to carboxylates formed by ethanol adsorption. In this case, the ν(OCO)
vibration produced by the bonding of carbon from the aromatic ring to oxygen from the
phenoxy molecule as well as oxygen from the support. Increasing temperature to 573 K
22
ACS Paragon Plus Environment

Page 23 of 57
ACS Catalysis
1
2
3
4
5
6
7
8
9
resulted in increases in band intensities associated with phenol adsorbed at defect sites.
Further increasing in temperature to the 673 ꢀ 773 K range significantly decreased the band
intensities.
The DRIFTS spectra obtained under HDO of phenol over Pd/CeO
2
.were shown 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
Fig. 5. At 373 K, the reaction of phenol in the presence of hydrogen leads to bands
ꢀ1
positioned at 2934, 2856, 1423, and 1055 cm ascribed to cyclohexanol. However, intense
ꢀ1
bands located at 1591 and 1491 cm correlate well with the ν(C=C) aromatic ring modes
ꢀ1
of phenoxy species, with 1262 cm being consistent with the ν(CO) monodentate mode of
ꢀ1
phenoxy species as well. The band at 1687 cm is due to the presence of 2ꢀcyclohexenꢀ1ꢀ
2
one. However, as we previously observed with Pd/ZrO , there is an intense band at close to
ꢀ1
1
560 cm that is too low to be consistent with an aromatic ring mode. Rather, it correlates
very well with an asymmetric ν(OCO) stretching mode of a vibration that resembles that of
ꢀ1
carboxylate species (e.g., similar to adsorbed formate or acetate). The band at ~1330 cm
may be the symmetric ν(OCO) stretching mode. Considering the tautomerization of
phenol, the result provides further confirmation for the coordination of the carbonyl
species of 2ꢀcyclohexadienone to the defect site, which results in carboxylateꢀlike modes
of vibration as the C of the carbonyl is now bonded to both O which came from the
carbonyl functional group, as well as oxygen adjacent to the defect site.
23
ACS Paragon Plus Environment

ACS Catalysis
Page 24 of 57
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
561
491
1
591
1
1
1
262
423
1
557 1055
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
2934
(
(
a) 3054 2856
1687
1429
1
330
252
1
1
777
b)
(
c)
1266
1479
1
586
(
d)
(
e)
3
500
3000
2500
2000
1500
1000
-1
Wavenumber (cm )
Figure 5. DRIFTS spectra under HDO of phenol over Pd/CeO
2
(H
2
/phenol = 60) at (a) 373
K, (b) 473 K, (c) 573 K, (d) 673 K, (e) 773 K.
When temperature was raised to 473 and 573 K, band intensities attributed to
coordination of phenol to defect sites increase, while those of cyclohexanol decrease. In
ꢀ1
addition, there is the development of a broad band between 1930 and 1750 cm . Further
increases in temperature diminished all band intensities. Figure 6 displays the DRIFTS
spectra of experiments carried out with the mixture of phenol and hydrogen using
Pd/CeZrO . Qualitatively, the trend is nearly identical to that of Pd/CeO .
2
2
24
ACS Paragon Plus Environment

Page 25 of 57
ACS Catalysis
1
2
3
4
5
6
7
8
9
1
562
1
597
1494
1
262
3026 2935
1
426
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
1169
2
853
2122
1687
3066
a)
(
1
345
1070
(b)
1585
(
c)
1
478
(
d)
(e)
3
500
3000
2500
2000
1500
1000
-1
Wavenumber (cm )
Figure 6. DRIFTS spectra under HDO of phenol over Pd/CeZrO
2
(H
2
/phenol = 60) at (a)
373 K, (b) 473 K, (c) 573 K, (d) 673 K, (e) 773 K.
3.5- Effect of support on the reaction mechanism for HDO of phenol
Our DRIFT spectra revealed that phenol adsorbs dissociatively over metal oxide
cations of the support. In addition, DRIFT spectra also revealed the formation of 2,4ꢀ
cyclohexadienꢀ1ꢀone, which is the keto tautomer of phenol. This result indicates that after
adsorption, phenol is tautomerized to 2,4ꢀcyclohexadienꢀ1ꢀone, which may follow two
different routes of reaction, and depends on support type. Ring hydrogenation of this ketoꢀ
tautomer intermediate led to the production of 2ꢀcyclohexenꢀ1ꢀone that further
hydrogenates to cyclohexanone. The catalytic tests with different feeds showed that
25
ACS Paragon Plus Environment

ACS Catalysis
Page 26 of 57
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
cyclohexanone is rapidly converted to cyclohexanol, or it may alkylate the phenolic ring
producing bicyclic compounds. This is the main reaction pathway for Pd/SiO , Pd/Al O ,
2
2
3
2 2 2 3
Pd/CeO and Pd/CeZrO catalysts. In the case of Pd/Al O catalyst, the dehydration of
cyclohexanol to benzene may also take place due to the high density of acid sites of this
support, as revealed by cyclohexanol dehydration test and temperatureꢀprogrammed
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
desorption of NH
deoxygenated products and the density of acid sites. The cyclohexanol dehydration tests
and NH ꢀTPD experiments showed that Pd/Al catalyst exhibited the highest density of
3 3
(NH ꢀTPD). However, there is no correlation between the selectivity to
3
2 3
O
2
acid sites whereas the highest selectivity to benzene was observed with Pd/ZrO and
Error! Bookmark not defined.
Pd/TiO
2
catalysts. Mortensen et al.
also proposed that the density of
acid sites of the support does not affect the activity for liquid phase phenol HDO using
supported nickel.
Therefore, the formation of benzene should proceed by another route on Pd/ZrO
2
and Pd/TiO , which do not have sufficient acidity. In this case, hydrogenating the CO
2
functional group of the ketoꢀtautomer intermediate results in the formation of 2,4ꢀ
cyclohexadienol and further conversion to benzene. This reaction pathway is favored over
Pd/ZrO and Pd/TiO catalysts.
2
2
There are some studies in the literature that suggested the oxygen vacancies play a
4
8
key role in the HDO of molecules like guaiacol. Schimming et al. investigated the HDO
of guaiacol over ceriaꢀzirconia catalysts. They reported a linear correlation between the
concentration of oxygen vacancies and the conversion of guaiacol at 573 K, indicating that
the HDO of guaiacol occurs on paired oxygen vacancies. In order to unveil the role of the
oxygen vacancies for the HDO of phenol, the IR of CO was used to measure the amount of
oxygen vacancies and the reaction was also carried out after reduction of Pd/CeZrO
catalyst at 773 K.
2
26
ACS Paragon Plus Environment

Page 27 of 57
ACS Catalysis
1
2
3
4
5
6
7
8
9
The IR bands after reduction of Pdꢀbased catalysts at 573 K and CO adsorption at
73 K are shown in Figure 7.
4
Pd/Al O
2
3
Pd/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
Pd/CeO2
Pd/CeZrO2
Pd/ZrO2
2
400 2200 2000 1800 1600 1400 1200 1000
-1
Wavenumber (cm )
Figure 7. DRIFTS spectra of adsorbed CO at 473 K over supported Pd catalysts.
ꢀ1
The bands at 2062 and 1960 cm are assigned to linear and bridge CO adsorption
ꢀ1
on metallic Pd particles. The bands between 1700 – 1200 cm are due to the different
ꢀ1
vibrational modes of monodentate and bidentate formate species: 1570 and 1361 cm
ꢀ1
ꢀ1
(ν(OCO) symmetric and asymmetric); 1381 cm (δ(CH)). The shoulder at 1531 cm and
ꢀ1
the bands at 1623, 1420 and 1336 cm could be due to the formation of hydrogen
carbonate and carbonate species. Formates are formed by the reaction between CO and
defectꢀassociated hydroxyl groups. Therefore, the ratio between the intensities of the
formate bands and the carbonyl bands on the metallic particles provides information about
27
ACS Paragon Plus Environment

ACS Catalysis
Page 28 of 57
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
the density of oxygen vacancies on the Pdꢀbased catalysts. This ratio is highest for the
Pd/CeZrO catalyst and decreases in the following order: Pd/CeZrO > Pd/CeO > Pd/ZrO
2
2
2
2
>
> Pd/Al
2
O
3 2
> Pd/TiO . This result is in agreement with the literature, since the addition of
4
9
zirconia to ceria structure improves the formation of oxygen vacancies. However,
Pd/ZrO catalyst exhibited much higher activity to deoxygenation and selectivity to
benzene than Pd/CeZrO and Pd/CeO catalysts. This result reveals that there was no clear
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
2
correlation between the concentration of oxygen vacancies and the deoxygenation activity
of phenol over these Pdꢀbased supported catalysts.
Table 4 shows a comparison between the selectivities obtained for HDO of phenol
2
over Pd/CeZrO catalyst after reduction at 573 and 773 K.
Table 4. Distribution of products for phenol HDO over Pd/CeZrO activated at 573 and
2
7
73K.
Reduction
temperature (K)
73
773
Conversion
(%)
W/F
(h)
Selectivity(%)
ONE OL
63.1
63.7
BZ
15.3
12.0
C12
0.2
0.2
5
33.0
35.3
0.16
0.80
21.1
23.7
2
Increasing the reduction temperature of Pd/CeZrO catalyst did not impact the
distribution of products for phenol HDO. However, the reduction at high temperature
increases the formation of oxygen vacancies, which is simultaneously followed by the
4+
3+
conversion of Ce to Ce to compensate for oxygen removal. Therefore, the higher
amount of oxygen vacancies formed after reduction at 773 K did not affect the selectivity
to deoxygenated products that remained unchanged in comparison to the catalyst reduced
4
+
3+
at 573 K. Furthermore, the oxophilicity of Ce and Ce cations should be similar and low,
28
ACS Paragon Plus Environment

Page 29 of 57
ACS Catalysis
1
2
3
4
5
6
7
8
9
since the selectivity to deoxygenated products was the same, regardless of the reduction
temperature.
Recently, it has been proposed in the literature that oxophilic sites (represented by
unreduced metals and metal oxide cations from the supports) promote the deoxygenation
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
7
,50,51
pathway.
The strong bond between the oxygen from the phenol molecule and the
oxophilic site might favor the hydrogenation of the carbonyl function or even the direct
5
2
cleavage of the CꢀO bond. Newman et al. also investigated the effect of the support on
liquid phase phenol HDO using Ru/SiO , Ru/Al , Ru/TiO , and Ru/C catalysts. The
highest selectivity to benzene was shown by Ru/TiO . This was attributed to the strong
2
2
O
3
2
2
3
+
interaction between the Ti sites (created during the reduction) with the oxygen from the
phenol molecule, which weakens the aromatic CꢀO bond and promotes the cleavage of this
8
bond. Griffin et al. also observed a higher selectivity to toluene over Pt/TiO than Pt/C for
2
gas phase mꢀcresol HDO at 623 K. They performed DFT calculations to study the most
energetically favorable reaction pathway over Pt(111) and anatase TiO
methylcyclohexanone and 3ꢀmethylcyclohexanol were mainly formed by the
hydrogenation of the ring over Pt(111) surface. For TiO (101) surface, tautomerization
2
(101) surfaces. 3ꢀ
2
and direct deoxygenation to toluene were also energetically favorable routes. In this case,
the reaction takes place at oxygen vacancy defects. This could explain the higher
selectivity to toluene observed for Pt/TiO
In the present work, the high selectivity to benzene over Pd/TiO
catalysts is likely due to the oxophilic sites of these supports represented by incompletely
2
catalyst.
2
and Pd/ZrO
2
4+
3+
4+
3+
coordinated Ti / Ti and Zr /Zr cations near the periphery of metal particles. Stronger
interactions between oxygen in phenol with metal oxide cations favors the selective
hydrogenation of the CO function of the tautomer intermediate, which is assisted by the
29
ACS Paragon Plus Environment