Aqueous-phase hydrodeoxygenation of bio-derived phenols to cycloalkanes

Article abstract of DOI:10.1016/j.jcat.2011.02.001

The kinetics of the catalytic hydrodeoxygenation of phenol and substituted phenols has systematically been investigated on the dual-functional catalyst system Pd/C and H₃PO₄ in order to better understand the elementary steps of the overall reaction. The reaction proceeds via stepwise hydrogenation of the aromatic ring, transformation of the cyclic enol to the corresponding ketone, hydrogenation of the cycloalkanone to the cycloalkanol and its subsequent dehydration as well as the hydrogenation of the formed cycloalkene. The presence of dual catalytic functions is indispensible for the overall hydrodeoxygenation. The dehydration reaction is significantly slower than the hydrogenation reaction and the keto/enol transformation, requiring a significantly larger concentration of Bronsted acid sites compared to the available metal sites for hydrogenation.

Full text of DOI:10.1016/j.jcat.2011.02.001

Journal of Catalysis 280 (2011) 8–16
Contents lists available at ScienceDirect
Journal of Catalysis
journal homepage: www.elsevier.com/locate/jcat
Aqueous-phase hydrodeoxygenation of bio-derived phenols to cycloalkanes
1
⇑
Chen Zhao, Jiayue He, Angeliki A. Lemonidou , Xuebing Li, Johannes A. Lercher
Catalysis Research Center, Department of Chemistry, Technische Universität München, Garching 85747, Germany
a r t i c l e i n f o
a b s t r a c t
Article history:
The kinetics of the catalytic hydrodeoxygenation of phenol and substituted phenols has systematically
been investigated on the dual-functional catalyst system Pd/C and H3PO4 in order to better understand
the elementary steps of the overall reaction. The reaction proceeds via stepwise hydrogenation of the aro-
matic ring, transformation of the cyclic enol to the corresponding ketone, hydrogenation of the cycloal-
kanone to the cycloalkanol and its subsequent dehydration as well as the hydrogenation of the formed
cycloalkene. The presence of dual catalytic functions is indispensible for the overall hydrodeoxygenation.
The dehydration reaction is significantly slower than the hydrogenation reaction and the keto/enol trans-
formation, requiring a significantly larger concentration of Brønsted acid sites compared to the available
metal sites for hydrogenation.
Received 20 August 2010
Revised 1 February 2011
Accepted 4 February 2011
Keywords:
Bio-oil
Phenol derivatives
Aqueous-phase catalysis
Hydrodeoxygenation
Palladium
Ó 2011 Published by Elsevier Inc.
Dual-functional catalysis
1. Introduction
as the main component of bio-oil can act as suitable solvent for
the selective hydrogenation or oxidation of the bio-derived chem-
Bio-oil, produced by pyrolysis or liquefaction of lignocellulosic
biomass, is a promising potential raw material for liquid energy
carriers and in particular liquid transportation fuels. It cannot be
used directly, because it contains a high concentration of oxygen,
and the presence of unsaturated and phenolic moieties makes it
instable [1–3]. In contrast to the conventional crude oil, which is
a mixture of hydrocarbons containing 98 wt.% carbon and hydro-
gen, 1.8 wt.% sulfur and only 0.1 wt.% oxygen [1], bio-oil is an
acidic (pH = 2.5) aqueous solution containing up to 50 wt.% oxygen
[4,5]. Such high oxygen values cause a rapid catalyst deactivation
in hydrodeoxygenation using sulfide CoMo and NiMo catalysts
[6]. The highly active oxo-functionalized molecules also readily oli-
gomerize, requiring advanced catalytic technology to directly
hydrodeoxygenate the crude bio-oil under mild conditions.
Bio-oil contains ca. 30 wt.% of lignin-derived phenolic compo-
nents (phenols, guaiacols, and syringols) which are of high energy
density [2,3]. Hydrodeoxygenation (HDO) of such phenolic com-
pounds to cycloalkanes has been reported using heterogeneous
metal sulfide catalysts at moderate temperatures (573–873 K)
with high-pressure hydrogen (13–16 MPa) in fixed bed reactors
[1,7]. However, the incorporation of sulfur into products over these
catalysts stimulated us to develop catalysts, which are not based
on sulfides [1]. Combined with supported metal catalysts, water
icals [8–14]. Water also allows straightforward product separation,
if the targeted products are hydrocarbons without polar functional
groups.
Recently, we reported a new route for quantitative conversion
of an aqueous mixture of bio-derived phenolic monomers to cyclo-
alkanes and methanol using metal catalysts (Pd or Ni) in the pres-
ence of acids (H3PO4 or Nafion/SiO2) [15,16]. Here, we describe the
detailed kinetic behavior of phenol, anisole, catechol, and guaiacol
on the dual-functional catalysts of Pd/C and H3PO4 to form the tar-
geted cycloalkanes, including the pathway of the individual reac-
tion steps and the most important reaction steps of the
conversion of phenol. We have also explored the relationship be-
tween catalytic activity and reaction conditions, such as the reac-
tion temperature, pH of the aqueous solution, the nature of the
metal as well as the catalyst support.
2. Experimental section
2.1. Chemicals
All chemicals were obtained from commercial suppliers and
used as provided: phenol (Merck, 99.5% GC assay), anisole (Aldrich,
99.0% GC assay), guaiacol (Fluka, >98.0% GC assay), catechol (Al-
drich, crystalline, >99.0% GC assay), 4-n-propylphenol (SAFC,
>97.0% GC assay), 4-methylguaiacol (Aldrich, 99.0% GC assay), 4-
ethylguaiacol (Aldrich, >98.0% GC assay), 4-n-propylguaiacol
(SAFC, >99.0% GC assay), eugenol (Aldrich, 99.0% GC assay),
4-hydroxy-3-methoxyphenylacetone (Aldrich, 96.0% GC assay),
⇑
Corresponding author. Fax: +49 89 28913544.
E-mail address: johannes.lercher@ch.tum.de (J.A. Lercher).
Present address: Department of Chemical Engineering, Aristotle University of
1
Thessaloniki, Thessaloniki 54124, Greece.
0021-9517/$ - see front matter Ó 2011 Published by Elsevier Inc.
doi:10.1016/j.jcat.2011.02.001

C. Zhao et al. / Journal of Catalysis 280 (2011) 8–16
9
4-allyl-2,6-dimethoxy-phenol (Alfa Aesar, 98.0% GC assay), cyclo-
hexanol (Sigma–Aldrich, 99.0% GC assay), 2-methoxycyclohexanol
(Alfa Aesar, 99.0% GC assay), palladium (II) nitrate hydrate (Strem
Chemicals, 99.9%-Pd), hydrogen (Air Liquide, >99.999%).
Then, the autoclave was pressured with 5 MPa H2 (ambient tem-
perature). Reactions were conducted at 473 K for 0.5 h with a stir-
ring speed of 1000 rpm. After cooling to ambient temperature, the
organic products were extracted by ethyl acetate. The organic
phase and aqueous phase were both analyzed by a gas chromato-
graph (GC, Shimadzu 2010, flame ionization detector) with a HP-5
2.2. Commercial catalysts used
capillary column (30 m ꢂ 250
lm). A gas chromatograph–mass
Pd/C (Aldrich, loading amount: 5 wt.%), Ru/C (Aldrich, loading
amount: 5 wt.%), Pt/C (Aldrich, loading amount: 5 wt.%), Rh/C (Al-
drich, loading amount: 5 wt.%), Pd/C (Aldrich, loading amount:
1 wt.%).
spectrometer combination (GC–MS, Shimadzu QP 2010S) was used
to identify the organic compounds. Internal standards (i.e., 2-iso-
propylphenol for the organic phase and acetone for the aqueous
phase) were used to determine the liquid product concentration
and carbon balance. The calculations of conversion and selectivity
were based on carbon mole basis. Conversion = (the amount of aro-
matic ring change during reaction/total amount of aromatic
ring) ꢂ 100%. Selectivity = (C atoms in each product/total C atoms
in the products) ꢂ 100%. The carbon balance in the liquid
phase for all reported experiments was better than 95 3% in this
work.
2.3. Preparation of supported catalysts Pd/Al2O3, Pd/SiO2, and Pd/ASA
We synthesized 2 wt.% of Pd-based catalysts supported on alu-
mina (Degussa), silica (Degussa), and amorphous-silica–alumina
(ASA, containing 20% of alumina) with incipient wetness impreg-
nation using palladium (II) nitrate hydrate as Pd precursor, then
air-calcined at 573 K for 4 h and hydrogen-reduced at 588 K for
4 h [17].
The ASA support was prepared from a gel formed at pH 7.5 by
mixing a solution of acetic acid and AlCl3ꢀ6H2O in distilled water
(pH = 1.5) with a solution containing sodium silicate in NH4OH
(pH = 12). The gel was then thoroughly washed with a diluted solu-
tion of ammonium acetate to eliminate Na+ cations, dried at 393 K
overnight at ambient atmosphere, and then calcined at 943 K for
2 h in flowing air (100 ml minꢁ1).
The composition of the gas phase was determined by GC (HP
6890) equipped with a plot Q capillary column (30 m ꢂ 250
lm)
with thermal conductivity detector (TCD). We found 99% hydrogen
and trace amounts of CO2 and methanol in the gas phase, so in this
work only the changes in the liquid phase were considered.
2.5.2. Kinetic study of phenol hydrodeoxygenation network
The kinetic study of the aqueous-phase hydrodeoxygenation of
phenol to cyclohexane was divided into four steps: (i) phenol
hydrogenation to cyclohexanone, (ii) cyclohexanone hydrogena-
tion to cyclohexanol, (iii) cyclohexanol dehydration to cyclohex-
ene, and (iv) cyclohexene hydrogenation to cyclohexane. These
reactions were carried out at 473 K in the batch autoclave reactor.
The reaction conditions were as follows: (i) phenol (12.5 g), Pd/C
(1 wt.%, 0.020 g), 5 MPa H2; (ii) cyclohexanone (60 g), Pd/C
(1 wt.%, 0.020 g), 5 MPa H2; (iii) cyclohexanol (140 g), H3PO4-H2O
(0.5 wt.%, 80 ml), 5 MPa N2; (iv) cyclohexene (50 g), Pd/C (1 wt.%,
0.020 g), 5 MPa H2.
After reaction, ethyl acetate was used to extract the organic
phase and the extract was analyzed by GC and GC–MS using the
internal standard 2-isopropylphenol. As the aqueous-phase HDO
reaction is a two-phase process, the real-time sampling during
reaction is very difficult. Thus, the conversion/selectivity data as
a function of reaction time were collected from separate batch
experiments with varying duration.
2.4. Catalyst characterization
2.4.1. Surface areas and pore diameters
The surface areas and pore diameters were determined from
nitrogen adsorption measurements carried out at 77 K on a PMI
automated BET sorptometer. The samples were first outgassed at
523 K for 20 h before measurements. The surface area, micropore
and mesopore distributions were calculated according to the BET
and BJH theories [18].
2.4.2. H2 chemisorption
The Pd catalysts were activated in vacuum at 588 K for 1 h and
then cooled to 313 K. A hydrogen adsorption isotherm (chemisorp-
tion together with physisorption) was subsequently measured at a
pressure range from 1 to 40 kPa. Then, the sample was outgassed at
the same temperature for 1 h to remove the physisorbed H2, fol-
lowed by measuring another adsorption isotherm (physisorption).
The concentration of chemisorbed hydrogen on the metal was
determined by extrapolating the isotherm to zero hydrogen pres-
sure and using the value to estimate the dispersions of the metal
on the supports. Besides H2 chemisorption, we also calculated
the Pd dispersion from a TEM micrograph (Fig. S1). The dispersion
of the 5 wt.% Pd/C calculated from the TEM measurement was 52%,
which is very similar to the 55% derived from the H2 chemisorp-
tion, showing that hydride formation and H2 uptake into the bulk
of Pd do not take place under the used conditions.
Turnover frequencies (TOFs) were determined as initial reaction
rates when the conversion was below 10%, defined as mole of con-
sumed reactant per mole of surface Pd atom (or per mole of H+) per
hour.
2.5.3. Kinetic study of anisole, catechol, and guaiacol
hydrodeoxygenation conversion
The kinetic study of anisole, catechol, or guaiacol was carried
out as follows. The reactant (0.016 mol) was mixed with 0.5 wt.%
H3PO4–H2O solution (80 ml) and 5 wt.% Pd/C (0.020 g), and then
such mixture was hydrodeoxygenated at 423 K and 5 MPa H2 with
a stirring speed of 1000 rpm. After reaction, ethyl acetate was used
to extract the organic phase, and the liquid products including
aqueous phase and organic phase were analyzed through GC and
GC–MS with internal standards (i.e., 2-isopropylphenol for the or-
ganic phase and acetone for the aqueous phase). As mentioned in
Section 2.5.2, the conversion/selectivity data versus reaction time
were collected from the corresponding separate batch experiments
with varying duration.
2.5. Measurements of catalytic reactions
2.5.1. Aqueous-phase HDO of bio-derived phenolic compounds
The HDO reactions were optimized by varying the pH of the
aqueous solution (neutral, acidic, basic), the metal (Pd, Pt, Ru,
Rh), the catalyst support (C, Al2O3, SiO2, ASA), as well as by varying
the reaction temperature (423, 473, 523 K). The reaction conditions
used are noted in the footnotes of the tables. In a typical experi-
ment, phenol or other lignin-derived monomers (0.0106 mol),
H3PO4-H2O solution (80 ml, 0.5 wt.%, pH = 2.1), and Pd/C (0.040 g,
5 wt.%) were added into an autoclave (Parr, Series 4843, 300 ml).
2.5.4. Aqueous-phase HDO of phenolic mixture and catalyst recycle
Resembling the HDO of bio-oil, reaction of a phenolic monomer
mixture (3.38 g) containing 2-methoxy-4-n-propylphenol (0.83 g,

10
C. Zhao et al. / Journal of Catalysis 280 (2011) 8–16
0.005 mol), 4-n-propylphenol (0.68 g, 0.005 mol), 4-hydroxy-3-
methoxyphenylacetone (0.90 g, 0.005 mol) and 4-allyl-2,6-dime-
thoxyphenol (0.97 g, 0.005 mol), Pd/C (0.080 g, 5 wt.%), and
H3PO4–H2O solution (80 ml, 0.5 wt.%) was carried out at 523 K
and 5 MPa H2 (ambient temperature) for 0.5 h, stirred at
1000 rpm. After reaction, the other treatments were kept the same
as mentioned in Section 2.5.1.
To test the catalyst recyclability, the first run for hydrodeoxy-
genation of the phenolic mixture was carried out at 523 K and
5 MPa H2 (ambient temperature) for 2 h, stirred at 1000 rpm. After
reaction, the Pd/C catalyst was separated by decantation, followed
by sequential washing with ethyl acetate and water, and drying at
393 K for 2 h. The catalyst was then reused in the next catalytic re-
cycle run without any further treatment.
3.2.1. Effect of catalyst support
In neutral water under conditions of 473 K, 5 MPa H2 and 0.5 h,
Pd/C yielded 98% cyclohexanol (Table 2). This is line with the
expectation that high hydrogen pressure, high temperature, and
strong adsorption of phenol on carbon lead to cyclohexanol being
the dominant product [25]. On the other hand, Pd/Al2O3, Pd/SiO2,
and Pd/ASA led to about 70% cyclohexanol and 30% cyclohexanone
at 90% conversion under the identical conditions. Liu et al. found
that Lewis acids such as AlCl3, ZnCl2, InCl3, and SnCl2 stabilized
cyclohexanone intermediates in the presence of Pd/C [26], achiev-
ing a 100% cyclohexanone yield in dichloromethane solvent at low
temperature (303 K) and hydrogen pressure (1 MPa). The present
results qualitatively agree with the above suggestion, as over Lewis
supports of Pd/Al2O3, Pd/SiO2, and Pd/ASA catalysts, the cyclohex-
anone intermediate is stabilized, thus its yield reached up to 30%.
2.5.5. Aqueous-phase dehydration of cyclohexanol
3.2.2. Effect of the pH value of aqueous solution
A typical experiment for the dehydration of cyclohexanol
(1.06 g, 0.0106 mol) was carried out in the Parr reactor with a
H3PO4-H2O solution (80 ml, 0.5 wt.%) at 373, 423, 453, and 473 K
and 5 MPa H2 (ambient temperature) for 0.5 h, stirred at
1000 rpm. After cooling, the organic layer was extracted with ethyl
acetate and the internal standard 2-isopropylphenol was added to
the organic phase to calculate the product concentrations. The li-
quid organic products were analyzed through GC and GC–MS.
In neutral water, Pd/C yielded 98% cyclohexanol at 473 K. When
a basic solution (80 ml of 0.05 wt.% NaOH solution, pH = 12) was
added, the performance (conversion and selectivity) of the Pd/C
catalyst was nearly unchanged by the presence of OH– ions, leading
to a yield of 97% cyclohexanol (see Table 2). However, when acetic
acid was introduced, i.e., the main organic acid in crude bio-oil (up
to 32 wt.%) [27], the product distribution was changed (Table 2).
When using Pd/C with acetic acid (5 wt.%, pH = 2.6), 75% cyclohex-
ane, 20% cyclohexanol, and 2% cyclohexanone were obtained. Small
amounts of esters and ethers, i.e., 1% cyclohexyl acetate and 2%
dicyclohexyl ether, were also catalytically produced with acetic
acid. Applying a stronger mineral acid, such as phosphoric acid
(0.5 wt.%, pH = 2.1), led to 85% cyclohexane selectivity with 99%
phenol conversion, indicating that the acidity of the aqueous solu-
tion is the essential factor for cyclohexane formation.
3. Results
3.1. Catalyst characterization
The main physicochemical properties of the Pd catalysts are
compiled in Table 1. The surface area of 1 wt.% Pd/C (934 m2 gꢁ1
)
By comparison, Pd/Al2O3, Pd/SiO2, and Pd/ASA catalysts pro-
duced a 7:3 ratio of cyclohexanol to cyclohexanone at ca. 80% con-
version in neutral water. When NaOH was added to the aqueous
solution, the catalytic activity was drastically decreased by 10%,
32%, and 24%, respectively. This is attributed to the observed par-
or 5 wt.% Pd/C (845 m2 gꢁ1) was much higher than that of 2 wt.%
Pd/SiO2 (174 m2 gꢁ1), Pd/Al2O3 (92 m2 gꢁ1), and Pd/ASA
(269 m2 gꢁ1). The pore diameter in Pd/C was 3.6 nm, while those
in the other three materials Pd/SiO2, Pd/Al2O3 and Pd/ASA were
14–18 nm as compiled in Table 1. In agreement with the higher
surface area of the carbon support, the dispersion of Pd in 5 wt.%
Pd/C calculated from H2 chemisorption data (55%) was much high-
er than that of Pd supported on Al2O3 (12%), SiO2 (24%), and ASA
(40%).
Table 2
Aqueous-phase hydrodeoxygenation of phenol.a
Catalyst T (K) Solvent
Conv. (%) Selectivity (C%)
Cyclo Cyclo
Hexane Hexanol Hexanone
Cyclo
3.2. Hydrodeoxygenation of phenol
Pd/C
Pd/Al2O3 473
Pd/SiO2
Pd/ASA
Pd/C
473
H2O
H2O
H2O
H2O
100
82
87
96
100
72
0
0
0
0
0
0
0
0
75
20
20
21
0
98
72
62
73
97
69
53
64
20
20
13
24
74
94
10
1
2
28
38
27
3
31
47
36
2
56
64
53
25
4
Phenol was selected as the simple phenolic model compound to
explore the fundamental chemistry in the aqueous-phase hydro-
deoxygenation. In industry, hydrogenation of phenol (using palla-
dium and platinum-based catalysts) is mainly used to produce
cyclohexanone, which is the key material for manufacturing nylon
6 and polyamide resins [18–25]. Cyclohexanol and cyclohexanone
are the main competing hydrogenated products from phenol
hydrogenation, while cyclohexane is formed on supported palla-
dium catalysts at a low yield (less than 20%) [24].
473
473
473
NaOH–H2O
NaOH–H2O
NaOH–H2O
NaOH–H2O
CH3COOH–H2O 100
CH3COOH–H2O
CH3COOH–H2O
Pd/Al2O3 473
Pd/SiO2
Pd/ASA
Pd/C
473
473
473
55
72
Pd/Al2O3 473
88
95
Pd/SiO2
Pd/ASA
Pd/C
473
473
423
423
473
523
473
473
473
CH3COOH–H2O 100
H3PO4–H2O
H3PO4–H2O
H3PO4–H2O
H3PO4–H2O
H3PO4–H2O
H3PO4–H2O
H3PO4–H2O
100
100
99
100
100
100
100
Pd/Cb
Pd/C
1
85
98
86
88
92
4
0
0
1
Table 1
Pd/C
Pt/C
Ru/C
Rh/C
Textural properties of the Pd-supported catalysts.
13
10
7
Catalyst
BET surface area
Pore diameter
(nm)
Dispersion
(%)
Metal loading
(wt.%)
0
(m2 gꢁ1
)
a
Reaction conditions: phenol (1.0 g, 0.0106 mol), H2O (80 ml), H3PO4–H2O
Pd/C
Pd/C
Pd/Al2O3
Pd/SiO2
Pd/ASA
934
845
92
174
269
3.7
3.6
16
18
14
51
55
12
24
40
1
5
2
2
2
(0.5 wt.%, 80 ml), CH3COOH–H2O (5 wt.%, 80 ml), NaOH–H2O (0.05 wt.%, 80 ml), Pd/
C (5 wt.%, 0.040 g), Rh/C (5 wt.%, 0.040 g), Ru/C (5 wt.%, 0.040 g), Pt/C (5 wt.%,
0.070 g), Pd/Al2O3 (2 wt.%, 0.10 g), Pd/SiO2 (2 wt.%, 0.10 g), Pd/ASA (2 wt.%, 0.10 g),
5 MPa H2, 0.5 h, stirred at 1000 rpm.
b
Reaction time: 1 h.

C. Zhao et al. / Journal of Catalysis 280 (2011) 8–16
11
Table 3
tial dissolution of amorphous inorganic oxides in the aqueous basic
Aqueous-phase dehydration of cyclohexanol.a
solution at 473 K, and thus, the well-dispersed Pd particles aggre-
gate, leading to less surface Pd atoms for the hydrogenation reac-
tion [11]. In basic environment, the selectivity ratio of
cyclohexanol to cyclohexanone remained approximately at 7:3.
Additionally, ca. 20% of cyclohexane, 20% of cyclohexanol, and
55% of cyclohexanone were formed on Pd/Al2O3, Pd/SiO2, and Pd/
ASA with acetic acid. Severe dissolution of the amorphous silicon
and aluminum oxides occurred after reaction with Brønsted liquid
acids in water at 473 K. Approximately 20% cyclohexane was cata-
lytically produced by acetic acid introduction. From the point of
stability in non-neutral aqueous solutions, robust carbon supports
are more practical for hydrodeoxygenation.
T (K) Catalyst Solvent
Conv. (%) Selectivity (C%)
Cyclohexene Cyclohexanone
473
473
473
373
393
423
433
453
473
473c
–
Pd/C
Pd/C
–
–
–
–
–
–
–
H2O
H2O
0
0
0
1.2
0
100
0.1
0
H3PO4–H2O 98
H3PO4–H2O
99.9b
0
0
H3PO4–H2O 0.1
H3PO4–H2O 1.2
H3PO4–H2O 2.3
H3PO4–H2O 8.6
H3PO4–H2O 95
H3PO4–H2O 93
62
69
83
95
99.9
100
38
31
17
5.0
0.1
0
a
Reaction conditions: cyclohexanol (1.06 g, 0.0106 mol), Pd/C (5 wt.%, 0.04 g),
3.2.3. Effect of the metal sites
H2O (80 ml), H3PO4–H2O (0.5 wt.%, 80 ml), 5 MPa H2 (ambient temperature), 0.5 h,
stirred at 1000 rpm.
Besides Pd/C, other noble metal catalysts such as Rh/C, Ru/C,
and Pt/C were explored. The results showed that these four cata-
lysts exhibited similar activities (85% cyclohexane yields) in phenol
conversion with H3PO4 at 473 K. The materials also performed
quite similarly in neutral and basic water solution, yielding ca.
100% cyclohexanol (Table S1). Because of the slight difference in
activities, Pd/C was selected as the typical catalyst to perform the
following experiments. For this reason, these three metals (Rh,
Ru, Pt) will not be discussed in detail in the discussion section.
b
Cyclohexane selectivity.
Repeat of the experiment carried out with H3PO4 at 473 K.
c
the mineral acid H3PO4 at 373 K, cyclohexanol was stable with zero
conversion in water solvent (Table 3). From 393 to 453 K, the dehy-
dration conversion of cyclohexanol over H3PO4 increased from 0.1%
to 8.6%. The apparent activation energy, calculated from the data
presented in Table 3 for the aqueous-phase cyclohexanol dehydra-
tion with H3PO4, was 120 kJ molꢁ1. At 473 K, cyclohexanol was
quantitatively dehydrated to cyclohexene within half an hour in
the presence of H3PO4 (Table 3). In previous studies, cyclohexanol
was dehydrated to cyclohexene with 80% yield over Lewis acid alu-
mina at 603–623 K [32], or it was dehydrated by Brønsted liquid
acids, such as sulfuric acid and phosphoric acid, and by solid acids
in the liquid phase at 403–473 K in the absence of solvents [33,34].
In the aqueous phase, water as solvent inhibits cyclohexanol dehy-
dration with respect to the chemical equilibrium, because it is also
a reaction product. Thus, the reaction temperature has to be in-
creased to 473 K in water compared to 403 K with sulfuric acid
without a solvent. Catalyzed by the Pd/C and H3PO4 dual-func-
tional components, phenol was rapidly and quantitatively hydro-
deoxygenated to cyclohexane (Table 3). In summary, cyclohexane
formation from phenol requires dual-functional catalysis, i.e., the
presence of hydronium ions for dehydration at appropriate tem-
peratures, and a metallic function able to hydrogenate in the pres-
ence of water.
3.2.4. Effect of temperature
With the Pd/C and H3PO4 combination, different temperatures
(423, 473, 523 K) were applied for phenol conversion. Cyclohexane
was not formed at 423 K after 0.5 and 1 h, cyclohexanol and cyclo-
hexanone being the main products. The different reaction routes
for phenol conversion are determined by the hydronium ion con-
centration of the aqueous solution, but the reaction temperature
needs to be sufficiently high to catalyze the more demanding dehy-
dration to cyclohexene at appreciable rates. Thus, increasing the
reaction temperature to 473 and 523 K led to 85% and 98% cyclo-
hexane yields over Pd/C and H3PO4 catalysts, respectively (Table
1). Note that the yields of cyclohexane achieved are far higher than
those achieved with the most active CoMo-sulfided catalysts, pro-
ducing approximately 34% benzene and 4% cyclohexane from phe-
nol at 673 K in a fixed bed reactor at comparable residence times
[28].
3.2.5. Hydrodeoxygenation reaction pathway for phenol conversion
The mechanism for gas-phase phenol conversion to cyclohexane
with heterogeneous sulfide catalysts has been proposed in Refs.
[7,29]: (i) direct hydrogenolysis to benzene (
followed by benzene hydrogenation (
(ii)hydrogenationtocyclohexanol(
by direct cyclohexanol hydrogenolysis (
ever, in the present aqueous-phase process with Pd/C, phenol and
cyclohexanol were unequivocally shown not to follow the classical
hydrogenolysis route. Thus, the overall hydrodeoxygenation route
does not follow the hydrogenation/hydrogenolysis sequential
mechanism.
In our previous work [15], it was shown that phenol was hydro-
genated to cyclohexanone over Pd/C in the first step. It was
concluded to have actually been formed by fast isomerization of
cyclohexenol (cyclic keto/enol transformation) [30,31]. Then,
cyclohexanone was gradually hydrogenated to cyclohexanol with
the hydrogenation catalysts. Pd/C does not catalyze the (structure
sensitive) hydrogenolysis of phenol, and thus, Pd as well as Pt-, Ru-,
and Rh-based catalysts favor phenol hydrogenation to cyclohexa-
nol in neutral water (Table S1).
The overall reaction pathway for the aqueous-phase hydro-
deoxygenation of phenol to cyclohexane proceeds via an initial
metal-catalyzed aromatic ring hydrogenation and naphthenic alco-
hol dehydration and then metal-catalyzed cycloalkene hydrogena-
tion (Fig. 1) [15].
H
D
H
= ꢁ62 kJ molꢁ1),
H
D
H
= ꢁ206 kJ molꢁ1), and
H
D
H
= ꢁ190 kJ molꢁ1), followed
H
D
H
= ꢁ78 kJ molꢁ1). How-
3.3. Hydrodeoxygenation of phenolic monomers
After optimizing the reaction conditions on such aqueous sys-
tem (Table S2), phenol and substituted phenols including catechol
and 4-n-propylphenol were quantitatively converted to cyclohex-
ane at 523 K and 5 MPa H2 in half an hour over Pd/C and H3PO4 cat-
alysts (Table 4). Five guaiacols, 4-methylguaiacol, 4-ethylguaiacol,
4-n-propylguaiacol, 4-allylguaiacol, and 4-acetonylguaiacol, were
fully converted to cycloalkanes, methanol, and intermediates.
About 80% yields of cycloalkanes, 8–10% methanol, and 12–18%
intermediates (cycloalkanol or cycloalkanone) were produced from
guaiacols. Promoted by acid catalysis, cycloalkanes undergo carbon
skeletal isomerization at the extent of 6–10%. The most complex
lignin-derived compound syringol (4-allyl-2,6-dimethoxyphenol)
was also tested, achieving a yield of 67% cycloalkanes and 12%
methanol. These results suggest that the new catalytic approach
The intermediate product, cyclohexanol, was not hydrogeno-
lyzed to cyclohexane over Pd/C at 473 K in water either, but was
dehydrogenated to 100% cyclohexanone at 1.2% conversion. With

12
C. Zhao et al. / Journal of Catalysis 280 (2011) 8–16
OH
OH
O
H+
Pd/C, H2
TOF1
Pd/C, H2
TOF2
Pd/C, H2
TOF4
TOF3
Fig. 1. Reaction pathway for aqueous-phase phenol of hydrodeoxygenation on dual-functional catalysts of Pd/C and H3PO4 at 473 K.
Table 4
Aqueous-phase hydrodeoxygenation of bio-derived phenolic monomers with Pd/C and H3PO4 catalysts at 523 K.a
Reactant
Conv. (%)
Selectivity (C%)
CH3OH
Cycloalcohol/ketone
100
100
100
98
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
1.3
3.6
12
HO
96
–
HO
HO
87
HO
100
100
–
–
78b
–
–
–
–
–
–
–
10
10
10
H3CO
HO
–
80c
9.2
H3CO
HO
100
99
–
–
–
–
–
–
–
–
–
–
–
–
66
65
71
58
3.0
3.5
3.1
3.1
6.7
5.0
5.2
5.6
7.2
7.4
8.2
12
17
18
12
21
H3CO
HO
H3CO
HO
100
92
H3CO
O
HO
OCH3
HO
H3CO
a
Reaction conditions: phenol (1.0 g, 0.0106 mol) or other phenolic monomers (0.0106 mol), Pd/C (5 wt.%, 0.040 g), H3PO4-H2O (0.5 wt.%, 80 ml), 5 MPa H2 (ambient
temperature), 0.5 h, stirred at 1000 rpm.
b
Includes isomerization products: methylcyclopentane (selectivity: 1.8%) and 1,2-dimethylcyclopentane (selectivity: 4.6%).
Includes isomerization products: ethylcyclopentane (selectivity: 2.2%) and 2-methyl-1-ethylcyclopentane (selectivity: 7.6%).
c
can be effectively applied in the hydrodeoxygenation of the diverse
substituted phenolic monomers in bio-oil upgrading.
copy (ICP-AES) analysis of the organic phase and the aqueous solu-
tion after each cycle did not show even traces of leached Pd
(detection limit: 1 g/mL). Nearly quantitative conversions and
very high selectivities (ca. 80% cycloalkanes and 8% methanol)
were also obtained in the two subsequent recycle experiments (Ta-
ble 5). The stable catalyst recycle performance indicates that the Pd
active sites have not changed (no changes in the Pd dispersion or
Pd leaching).
The hydrodeoxygenation of a phenolic mixture was tested over
Pd/C and H3PO4. Phenolic monomers (4-n-propylphenol, 2-meth-
oxy-4-n-propylphenol, 4-hydroxy-3-methoxyphenylacetone, and
4-allyl-2,6-dimethoxyphenol) were mixed in equal molar amounts
to simulate the bio-oil mixture (Table 5). As expected, the substi-
tute bio-oil was effectively hydrodeoxygenated at conditions
523 K, 5 MPa H2, and 0.5 h, maintaining a high rate of the overall
reaction. The conversions of the components ranged from 82% to
100%, with selectivities of 70% in cycloalkanes and 8.0% in
methanol.
l
4. Discussion
Moreover, to test for catalyst recyclability, a batch of Pd/C cata-
lyst was repeatedly used on the substitute bio-oil mixture at 523 K
and 5 MPa H2 for 2 h (Table 5). The first run reached 100% conver-
sion and selectivities of 89% cycloalkanes and 9.0% methanol after
2 h testing. Inductively coupled plasma-atomic emission spectros-
4.1. Kinetics of the phenol hydrodeoxygenation network
The elementary reactions for the aqueous-phase phenol hydro-
deoxygenation network are divided into four processes (Fig. 1), (i)
phenol hydrogenation to cyclohexanone on Pd/C, (ii) cyclohexa-

C. Zhao et al. / Journal of Catalysis 280 (2011) 8–16
13
Table 5
Hydrodeoxygenation of phenolic mixture and catalyst recycle.
Conv. (%)
Selectivity (C%)
CH3OH
Cycloalcohol/ketone
HDO of phenolic mixturea
4-n-Propylphenol: 100
64
3.0
3.2
8.0
20
2-Methoxy-4-n-propylphenol: 92
4-Hydroxy-3-methoxyphenylacetone: 100
4-Allyl-2,6-dimethoxyphenol: 82
Catalyst recycle (the reaction was carried out at 523 K for 2 h)
Run 1
Run 2
Run 3
4-n-Propylphenol: 100
80
73
67
3.6
2.5
1.7
5.0
3.1
1.7
9.0
8.4
7.5
1.7
12
21
2-Methoxy-4-n-propylphenol: 100
4-Hydroxy-3-methoxyphenylacetone: 100
4-Allyl-2,6-dimethoxyphenol: 95
4-n-Propylphenol: 100
2-Methoxy-4-n-propylphenol: 98
4-Hydroxy-3-methoxyphenylacetone: 100
4-Allyl-2,6-dimethoxyphenol: 87
4-n-Propylphenol: 100
2-Methoxy-4-n-propylphenol: 88
4-Hydroxy-3-methoxyphenylacetone: 95
4-Allyl-2,6-dimethoxyphenol: 79
a
Reaction conditions: the phenolic monomer mixture (3.4 g) contains 2-methoxy-4-n-propylphenol (0.83 g, 0.005 mol), 4-n-propylphenol (0.68 g, 0.005 mol), 4-hydroxy-
3-methoxyphenylacetone (0.90 g, 0.005 mol), and 4-allyl-2,6-dimethoxyphenol (0.97 g, 0.005 mol), Pd/C (0.080 g, 5 wt.%), H3PO4-H2O solution (80 ml, 0.5 wt.%), 523 K, 5 MPa
H2 (ambient temperature), 0.5 h, stirred at 1000 rpm.
was 1:2 ꢂ 104 molcyclohexanone molꢁ1
hꢁ1 (Fig. 2b), a higher value
than that of the phenol hydrogenation.
Table 6
surf:Pd
Turnover frequencies for aqueous-phase phenol hydrodeoxygenation network.
TOF1 for
TOF2 for
TOF3 for
TOF4 for
phenol
cyclohexanone
hydrogenation
cyclohexanol
dehydration
cyclohexene
hydrogenation
hydrogenation
(hꢁ1
)
(hꢁ1
)
(hꢁ1
)
(hꢁ1
)
10.0
6.2 ꢂ 103
1.2 ꢂ 104
15
>9 ꢂ 106
Cyclohexanone hydrogenation to cyclohexanol
8.0
6.0
4.0
2.0
0.0
none hydrogenation to cyclohexanol on Pd/C, (iii) cyclohexanol
dehydration to cyclohexene with H3PO4, and (iv) cyclohexene
hydrogenation to cyclohexane on Pd/C.
Cyclohexanone was found to be the dominant product (>95%
selectivity) (see Fig. 2a) for phenol hydrogenation (DH°(473K) =
0
20
40
60
80
100
120
ꢁ130 kJ molꢁ1) at 473 K on Pd/C at low conversions (lower than
Time (min)
10%). The TOF of phenol hydrogenation on Pd/C was
6:2 ꢂ 103 molphenol molꢁ1
hꢁ1 (see Table 6). For the further
surf:Pd
Fig. 2b. Aqueous-phase hydrogenation of cyclohexanone to cyclohexanol as
function of time. Reaction conditions: cyclohexanone (60 g), Pd/C (1 wt.%, 0.020 g),
473 K, 5 MPa H2, stirred at 1000 rpm.
a
cyclohexanone hydrogenation to cyclohexanol (
D
H°(473K) = ꢁ74
kJ molꢁ1), the TOF of the ketone hydrogenation on Pd/C at 473 K
10.0
10.0
Phenol hydrogenation to cyclohexanone
Cyclohexanol dehydration
8.0
6.0
4.0
2.0
0.0
8.0
6.0
4.0
2.0
0.0
0
20
40
60
80
100
120
0
20
40
60
80
100
120
Time (min)
Time (min)
Fig. 2a. Aqueous-phase hydrogenation of phenol to cyclohexanone as a function of
time. Reaction conditions: phenol (12.5 g), Pd/C (1 wt.%, 0.020 g), 473 K, 5 MPa H2,
stirred at 1000 rpm.
Fig. 2c. Aqueous-phase dehydration of cyclohexanol to cyclohexene as a function of
time. Reaction conditions: cyclohexanol (140 g), H3PO4-H2O (0.5 wt.%, 80 ml),
473 K, 5 MPa N2, stirred at 1000 rpm.

14
C. Zhao et al. / Journal of Catalysis 280 (2011) 8–16
The TOF rate of the acid-catalyzed alcohol dehydration at 473 K
hydrolyzed to phenol at 473 K in water with 3 wt.% acetic acid
[35]. The increasing cyclohexanol selectivity was accompanied by
a decreasing cyclohexanone selectivity (see Fig. 3a), which sug-
gests that after the rapid hydrolysis of anisole and the partial
hydrogenation of phenol, the hydrogenation of cyclohexanone
was somewhat slower. Note that the high selectivity to cyclohexa-
none indicates a higher adsorption constant of the aromatic rings
compared to the saturated cyclic compounds. The other indepen-
dent primary product was the directly hydrogenated methoxy-
cyclohexane (t = 0, selectivity ꢃ 10%). As the selectivity to meth-
oxy-cyclohexane and methanol was approximately constant at
30% and 10% from 10 to 190 min, respectively, the ratio of the aro-
matic ring hydrogenation to the ether hydrolysis is concluded to be
constant. This indicates that the hydrogenation of the aromatic
ring and methoxy hydrolysis take place as parallel reaction routes.
Moreover, the ratio of the selectivities of cyclohexanone and cyclo-
hexanol to methanol was 6:1, which coincides with the carbon mo-
lar ratio of aromatic ring to methoxy group in anisole. The TOF for
initial anisole hydrogenation on Pd/C at 423 K exceeded
(D
H°(473K) = 26 kJꢀmolꢁ1) was as low as 15 molcyclohexanol molHꢁþ1 hꢁ1
(Fig. 2c, Table 6). This indicates that the aqueous-phase dehydra-
tion is the key step in the overall hydrodeoxygenation reaction,
which is also in line with its high apparent activation energy.
The rate of cyclohexene hydrogenation was very high, so that
the conversion reached 87% at 473 K on Pd/C after 5 min (for the
reaction conditions see Section 2.5.2), thus the TOF value for cyclo-
hexene hydrogenation (
D
H°(473K) = ꢁ119 kJ molꢁ1) on Pd/C ex-
ceeded 9 ꢂ 106 molcyclohexene molꢁ1
hꢁ1 (Table 6). The extremely
surf:Pd
high intrinsic consumption rate of cyclohexene and the phase sep-
aration also shift the equilibrium of the dehydration reaction,
allowing nearly quantitative cyclohexane production.
4.2. Kinetics of anisole, catechol, and guaiacol hydrodeoxygenation
To probe the stepwise aqueous-phase conversion of various
substituted groups at the aromatic ring, i.e., a methoxy group,
two adjacent hydroxyl groups, and adjacent hydroxyl and methoxy
groups on Pd/C and H3PO4, anisole, catechol, and guaiacol were
selected.
Anisole (methyl phenyl ether) was reacted on Pd/C at 423 K and
5 MPa H2 with H3PO4 and the conversion and selectivity of prod-
ucts are displayed in Fig. 3a as a function of reaction time. The pri-
mary product (t = 0, selectivity = 80%) was cyclohexanone
originating from phenol hydrogenation. This implies that under
the selected conditions, hydrolysis of anisole to phenol is the dom-
inating reaction. The TOF for anisole hydrolysis on H3PO4 at 423 K
was 2:8 molanisole molꢁHþ1 hꢁ1. This hydrolysis rate is much lower
than the phenol hydrogenation rate (6120 hꢁ1), thus the phenol
intermediate is not detected and cyclohexanone appears to be
the primary product. Ratton also showed that anisole was 100%
O
OCH3
OH
(a)
H+/H2O
2H2
Pd
+
CH3OH
+ CH3OH
hydrogenation
hydrolysis
ketone/enol isomerization
H2/Pd hydrogenation
OCH3
hydrogenation
H2/Pd
OH
CH3OH
+
OH
O
OH
(b)
OH
OH
H2
2H2
OH
Pd
Pd
100
80
60
40
20
0
80
60
40
20
0
(a)
hydrogenation
hydrogenation
ketone/enol isomerization
dehydration H+
O
dehydration
H+
Conversion
cyclohexanol
methoxycyclohexane
methanol
OH
cyclohexanone
H2/Pd
hydrogenation
ketone/enol isomerization
OH
0
50
100
Time (min)
150
200
O
H2
Pd
(b) 100
80
60
40
20
0
80
hydrogenation
Conversion
60
40
20
0
2-hydroxycyclohexanol
cyclohexanol
OH
O
OH
(c)
2-hydroxycyclohexanone
cyclohexanone
OCH3
OCH3
OCH3
2H2
Pd
H2
Pd
hydrogenation
ketone/enol isomerization
H+, hydrolysis
hydrogenation
0
50
100
Time (min)
150
200
H+, hydrolysis
100
100
90
80
70
60
50
40
30
20
10
0
OH
O
(c)
OH
OH
Conversion
H2
Pd
80
60
40
20
0
CH3OH
CH3OH
methoxycyclohexane
2-hydroxycyclohexanone
methanol
cyclohexanol
cyclohexanone
H+, dehydration
H+, dehydration
ketone/enol isomerization
Pd, H2, hydrogenation
2-methoxycyclohexanone
2-methoxycyclohexanol
O
OH
H2
Pd
0
50
100
Time (min)
150
200
CH3OH
CH3OH
hydrogenation
Fig. 3. Product distribution after hydrogenation of phenolic monomers (a) anisole,
(b) catechol, and (c) guaiacol. Reaction conditions: reactant (0.0106 mol), Pd/C
(5 wt.%, 0.020 g), H3PO4-H2O (0.5 wt.%, 80 ml), 423 K, 5 MPa H2 (ambient temper-
ature), stirred at 1000 rpm.
Fig. 4. Proposed reaction pathway for aqueous-phase conversion of bio-derived
phenolic monomers (a) anisole, (b) catechol, and (c) guaiacol to cyclohexanol with
Pd/C catalyst and H3PO4 component.

C. Zhao et al. / Journal of Catalysis 280 (2011) 8–16
15
1100 molanisole molꢁ1
hꢁ1. The results also imply that the alkyl C-
dehydration followed by hydrogenation and the other via hydroge-
nation followed by dehydration. This dual pathway may explain
why cyclohexanone behaves as a primary product. Cyclohexanone
is in turn hydrogenated to cyclohexanol (Fig. 4b). Note that we can-
not rule out, however, that in a first step, only one double bond of
catechol is hydrogenated and that this diol undergoes acid-cata-
lyzed elimination of water to phenol, which subsequently is then
partially hydrogenated to cyclohexanone [38].
surf:Pd
O bond of an aryl alkyl ether (anisole) is much easier to hydrolyze
than that of a dialkyl ether (methoxy-cyclohexane), which is in line
with the findings of the hydrolysis of benzyl phenyl ether [36] and
diphenyl ether [37]. Hence, with a single methoxy group at the aro-
matic ring, anisole can be catalytically hydrolyzed and directly
hydrogenated at the same time on the dual-functional catalyst sys-
tem Pd/C and H3PO4 (Fig. 4a).
Catechol having two adjacent hydroxyl groups on the aromatic
ring is an important constituent from lignin depolymerization and
also an intermediate in the guaiacol or syringol conversion. When
Pd/C and a mineral acid were jointly used as catalysts for catechol
hydrodeoxygenation, it was rapidly hydrogenated at the aromatic
ring leading to 2-hydroxycyclohexanone (t = 0, selectivity = 80%)
(Fig. 3b). The TOF for catechol hydrogenation on Pd/C at 423 K ex-
Hydrogenation of the most representative phenolic monomer,
guaiacol, with adjacent methoxy and hydroxy functional groups
at the aromatic ring led to a variety of products including cycloal-
kanone, cycloalkanol, aromatics, and methanol at 423 K (Fig. S2).
The primary product was the intermediately hydrogenated 2-
methoxycyclohexanone (t = 0, selectivity = 100%) (Fig. 3c). Also, in
this case, the fastest step is the metal-catalyzed hydrogenation of
the aromatic ring and not the acid-catalyzed hydrolysis of the
methoxy group. The initial TOF for guaiacol hydrogenation on Pd/
ceeded 3500 molcatechol molꢁ1
hꢁ1. Subsequently, 2-hydroxycy-
surf:Pd
clohexanone was rapidly converted to cyclohexanone
(s = 0,
selectivity = 20%) via two parallel routes of sequential alcohol
dehydration/alkene hydrogenation (minor pathway) and ketone
hydrogenation/alcohol dehydration (major pathway) (Fig. 4b). It
was shown in a separate experiment that 1,2-cyclohexanediol
was dehydrated to cyclohexanone at 92% yield at 523 K with Pd/
C and H3PO4 in 10 min (Table S3). Note that the dehydration rate
of 1,2-cyclohexanediol was significantly lower at 423 K. It was
shown from Fig. 3b that cyclohexanone was then slowly hydroge-
nated to cyclohexanol, eventually reaching nearly 30% selectivity.
Thus, the catechol conversion proceeds in the first step via hydro-
genation, as initial metal-catalyzed partial aromatic hydrogenation
to form 2-hydroxycyclohexanone, and then, it was subsequently
converted to cyclohexanone via two parallel pathways: one via
C at 423 K exceeded 1500 molguaiacol molꢁ1
hꢁ1. In comparison
surf:Pd
with anisole, which has only a single methoxy group and which
is easily hydrolyzed at 423 K in water, the lower rate of the meth-
oxy group hydrolysis of guaiacol may be caused by the stabiliza-
tion of the carbocation transition state through electron donation
by the adjacent hydroxyl groups, preventing its further hydrolysis
to methanol. Note that Kanetake et al. reported that under more
drastic conditions, i.e., in supercritical water at 653–673 K, guaia-
col can be hydrolyzed and pyrolyzed to catechol, methanol, phenol,
and char [39].
The hydrogenated intermediate 2-methoxycyclohexanone is
converted to cyclohexanone and methanol via two routes
(Fig. 4c). One is based on sequential methoxy hydrolysis followed
OH
O
OH
R
OCH3
R
OCH3
R
OCH3
H2
Pd
2H2
Pd
R1
O
R1
hydrogenation
R1
hydrogenation
ketone/enol isomerization
OH
O
R
OH
R
H+
R
-H2O/H+
CH3OH
or
CH3OH
dehydration
R1
hydrolysis
R1
R1
ketone/enol isomerization
OH
R1
OH
OH
CH3OH
R
R
OH
H2
or
or
CH3OH
Pd
R1
hydrogenation
R1
R1
demethoxygenation
-H2O/H+
H2
or
CH3OH
CH3OH
Pd
R1
R1
R1
hydrogenation
dehydration
R=H, OCH3
R1=alkyl
Fig. 5. Proposed overall reaction pathway for hydrodeoxygenation of bio-derived phenolic monomers to cycloalkanes with Pd/C catalyst and H3PO4 component [15].

16
C. Zhao et al. / Journal of Catalysis 280 (2011) 8–16
by alcohol dehydration and alkene hydrogenation and the other
route is based on sequential ketone hydrogenation followed by
hydrolysis, diol dehydration, and ketone/enol isomerization. As
shown (Fig. 3c), 2-methoxycyclohexanone was mainly hydroge-
nated to 2-methoxycyclohexanol, eventually achieving an 80%
yield. Note that at 523 K, 2-methoxycyclohexanol was converted
to cyclohexanone and methanol with 85% yield on Pd/C and
H3PO4 catalysts in 10 min (Table S3). The primary reaction path-
way for methoxycyclohexanol conversion is speculated to be the
acid-catalyzed hydrolysis to methanol and 1,2-cyclohexanediol,
followed by the acid-catalyzed dehydration of the latter product
and the keto/enol isomerization to cyclohexanone. The pathway
to the rapid initial formation of cyclohexanone is unclear to us at
present. Independent of the pathway by which it is formed, cyclo-
hexanone is in turn gradually hydrogenated on Pd/C to cyclohexa-
nol at 423 K (Fig. 4c).
hydrolysis, dehydration, and isomerization. The balance between
these functions determines the observed catalytic chemistry.
The turnover frequencies of the acid-catalyzed dehydration reac-
tions are at least two orders of magnitude lower than the rates of
metal-catalyzed hydrogenation. This causes the acid-catalyzed
steps to determine the overall hydrodeoxygenation reaction, and
high concentrations of hydronium ions are required to efficiently
catalyze it. A balanced dual-functional catalyst allows to convert
phenol-based bio-oils to the mixture of cyclic alcohols or alkanes
depending on the chosen reaction conditions.
Acknowledgment
This work was supported by the Technische Universität Mün-
chen in the framework of the European Graduate School for Sus-
tainable Energy. Valuable discussions in the framework of the
network of excellence IDECAT are also gratefully acknowledged.
4.3. Overall hydrodeoxygenation reaction pathway for conversion of
phenolic monomers to cycloalkanes
Appendix A. Supplementary data
Combining the results from phenolic molecules to the cycloalk-
anol (Figs. 1–4) and cycloalkanol to cycloalkane through dehydra-
tion (Table 3), the overall reaction pathway for converting phenolic
monomers to cycloalkanes is summarized in Fig. 5 [15]. Under
appropriate conditions, phenolic monomers such as methoxy-
substituted phenols are first hydrogenated on the aromatic ring
on the metal site to form 2-methoxycyclohexanone and subse-
quently hydrogenated at 2-methoxycyclohexanone to 2-methoxy-
cyclohexanol. The reaction sequence continues with the acid-
catalyzed hydrolysis of the cyclohexanol methyl ether to form
the cyclohexan-1,2-diol, which is in turn dehydrated to cyclohexa-
none. Subsequent hydrogenation of cyclohexanone leads to cyclo-
hexanol. Acid-catalyzed cyclohexanol dehydration and metal-
catalyzed cyclohexene hydrogenation finally lead to the targeted
cyclohexane. This highly integrated stepwise aqueous-phase
hydrodeoxygenation pathway of phenolic monomers is based on
dual-functional catalysis, i.e., coupling metal-catalyzed hydrogena-
tion and acid-catalyzed hydrolysis and dehydration. It differs dras-
tically from the C–O bond hydrogenolysis pathway proposed for
sulfided catalysts [25].
Supplementary data associated with this article can be found, in
the online version, at doi:10.1016/j.jcat.2011.02.001.
References
[1] A. Corma, S. Iborra, A. Velty, Chem. Rev. 107 (2007) 2411.
[2] S. Czernik, A.V. Bridgwater, Energy Fuels 18 (2004) 590.
[3] D. Mohan, C.U. Pittman, P.H. Steele, Energy Fuels 20 (2006) 848.
[4] K. Sipilä, E. Kuoppala, L. Fagernäs, A. Oasmaa, Biomass Bioenergy 14 (1998)
103.
[5] M. Garcia-Perez, A. Chaala, H. Pakdel, D. Kretschmer, C. Roy, Biomass Bioenergy
31 (2007) 222.
[6] T.-R. Viljava, R.S. Komulainen, A.O.I. Krause, Catal. Today 60 (2001) 83.
[7] M.J. Girgis, B.C. Gates, Ind. Eng. Chem. Res. 30 (1991) 2021.
[8] R.D. Cortright, R.R. Davda, J.A. Dumesic, Nature 418 (2002) 964.
[9] G.W. Huber, J.W. Shabaker, J.A. Dumesic, Science 300 (2003) 2075.
[10] N. Yan, C. Zhao, C. Luo, P.J. Dyson, H. Liu, Y. Kou, J. Am. Chem. Soc. 128 (2006)
8714.
[11] C. Zhao, W.J. Gan, X.B. Fan, Z.P. Cai, P.J. Dyson, Y. Kou, J. Catal. 254 (2008) 244.
[12] T. Wang, C.X. Xiao, L. Yan, L. Xu, J. Luo, H. Shou, Y. Kou, H. Liu, Chem. Commun.
(2007) 4375.
[13] N. Yan, C. Zhao, P.J. Dyson, C. Wang, L. Liu, Y. Kou, ChemSusChem 1 (2008) 626.
[14] C.X. Xiao, Z.P. Cai, T. Wang, Y. Kou, N. Yan, Angew. Chem. Int. Ed. 47 (2008) 746.
[15] C. Zhao, Y. Kou, A.A. Lemonidou, X.B. Li, J.A. Lercher, Angew. Chem. Int. Ed. 48
(2009) 3987.
[16] C. Zhao, Y. Kou, A.A. Lemonidou, X.B. Li, J.A. Lercher, Chem. Commun. 46 (2010)
412.
4.4. Acid-catalyzed isomerization of cycloalkanes
[17] M.F. Williams, B. Fonfé, C. Sievers, A. Abraham, J.A. van Bokhoven, A. Jentys,
J.A.R. van Veen, J.A. Lercher, J. Catal. 251 (2007) 485.
Catalyzed by acids, cycloalkanes undergo carbon skeletal isom-
erization to the extent of 6–10%. For example, n-propylcyclohexane
was isomerized to n-butylcyclopentane via ring contraction and
proton transfer and to 2-methylpropylcyclopentane via 1,3-carbon
backbone transformation.
[18] E.P. Barrett, L.G. Joyner, P.P. Halenda, J. Am. Chem. Soc. 73 (1951) 373.
[19] Y.Z. Chen, C.W. Liaw, L.I. Lee, Appl. Catal. A 177 (1999) 1.
[20] N. Mahata, V. Vishwanathan, J. Catal. 196 (2000) 262.
[21] S. Scire, S. Minico, C. Crisafulli, Appl. Catal. A 235 (2002) 21.
[22] E. Díaz, A.F. Mohedano, L. Calvo, M.A. Gilarranz, J.A. Casas, J.J. Rodríguez, Chem.
Eng. J. 131 (2007) 65.
[23] G. Yuan, M.A. Keane, Ind. Eng. Chem. Res. 46 (2007) 705.
[24] S. Velu, M.P. Kapoor, S. Inagaki, S. Suzuki, Appl. Catal. A 245 (2003) 317.
[25] G. Neri, A.M. Visco, A. Donato, C. Milone, M. Malentacchi, G. Gubitosa, Appl.
Catal. A 110 (1994) 49.
[26] H. Liu, T. Jiang, B. Han, S. Liang, Y. Zhou, Science 326 (2009) 1250.
[27] K. Takanabe, K. Aika, K. Seshan, L. Lefferts, J. Catal. 227 (2004) 101.
[28] D.C. Elliott, Energy Fuels 21 (2007) 792.
5. Conclusions
The reaction pathways of (substituted) phenols such as phenol,
anisole, catechol, and guaiacol are described together with the
principal kinetic pathways to outline the elementary catalytic
steps in dual-functional hydrodeoxygenation. Phenol is concluded
to be converted to cyclohexane by sequential phenol hydrogena-
tion, cyclohexenol transformation, cyclohexanone hydrogenation,
cyclohexanol dehydration, and cyclohexene hydrogenation. The
more complex guaiacol is first hydrogenated to 2-methoxycyclo-
hexanol, followed by the hydrolysis of the methoxy group to
cyclohexane-1,2-diol, and then dehydration to cyclohexenol/
cyclohexanone. The subsequent series of steps transform cyclo-
hexanone via hydrodeoxygenation to cyclohexane as discussed
above.
[29] E. Furimsky, Appl. Catal. A 199 (2000) 147.
[30] R.A. Correa, P.E. Lindner, D.M. Lemal, J. Am. Chem. Soc. 116 (1994) 10795.
[31] P.E. Lindner, R.A. Correa, J. Gino, D.M. Lemal, J. Am. Chem. Soc. 118 (1996)
2556.
[32] C.E. Parish, US Patent 3 275 698, 1966, to Signal Oil and Gas Company.
[33] F. Eschard, US Patent 2 957 034, 1960, to Institut Francais du Petrole.
[34] H. Aizawa, A, Kuroda, M. Minaga, K. Ohnishi, S. Matsuhisa, US Patent 3 974
232, 1976, to Toray Industries, Inc.
[35] S. Ratton, US 4 473 713, 1984, to Rhone-Poulenc Specialites Chimiques.
[36] V.M. Roberts, S. Fendt, A.A. Lemonidou, X. Li, J.A. Lercher, Appl. Catal. B 95
(2010) 71.
[37] V.M. Roberts, R.T. Knapp, X. Li, J.A. Lercher, ChemCatChem 2 (2010) 1407.
[38] J.S. Kudavalli, D.R. Boyd, D. Coyne, J.R. Keeffe, D.A. Lawlor, A.C. MacCormac,
R.A.M. O’Ferrall, S.N. Rao, N.D. Sharma, Org. Lett. 12 (2010) 5550.
[39] W.T. Kanetake, M. Sasaki, M. Goto, Chem. Eng. Technol. 30 (2007) 1113.
The dual-functional catalysis for hydrodeoxygenation requires
a metal function for hydrogenation and an acid function for