Hydrogenation and cleavage of the C-O bonds in the lignin model compound phenethyl phenyl ether over a nickel-based catalyst

Article abstract of DOI:10.1016/S1872-2067(12)60535-X

Phenethyl phenyl ether (PPE) was selected as a typical lignin model compound and hydrogenated and cleaved over two readily accessible nickel-based catalysts, which could be easily separated from the product mixture. The results revealed that the reduction of the nickel catalyst with gaseous hydrogen produced a species capable of achieving higher activity towards C-O-C bond cleavage compared to the Ru/C and Pd/C catalysts. The selectivity of the C-O-C bond cleavage over the Ni/C catalyst was 85%, and higher than the corresponding values in the Ru/C (40%) and Pd/C (69%) systems. Using the carbothermal reduction method for the production of the Ni/C-C catalyst, the conversion and selectivity levels reached 99%, with 40% of the benzene rings in PPE being reserved. In comparison, no benzene ring containing products wer observed over the noble metal catalysts. This difference was attributed to the interaction between the carbon support and the nickel nanoparticles.

Full text of DOI:10.1016/S1872-2067(12)60535-X

Chinese Journal of Catalysis 34 (2013) 651–658
ava i l a b l e a t w w w. s c i e n c ed i r e c t . c o m
j o u r n a l h o m e p a g e : w w w. e l s e v i e r. c o m / l o c a t e /c h n j c
Communication
Hydrogenation and cleavage of the C–O bonds in the lignin model
compound phenethyl phenyl ether over a nickel‐based catalyst
SONG Qia,b, CAI Jiayinga,b, ZHANG Junjiea,b, YU Weiqianga, WANG Fenga,*, XU Jiea,#
a Dalian Institute of Chemical Physics, Chinese Academy of Sciences, Dalian 116023, Liaoning, China
b Graduate University of Chinese Academy of Sciences, Beijing 100049, China
A R T I C L E I N F O
A B S T R A C T
Article history:
Phenethyl phenyl ether (PPE) was selected as a typical lignin model compound and hydrogenated
and cleaved over two readily accessible nickel‐based catalysts, which could be easily separated from
the product mixture. The results revealed that the reduction of the nickel catalyst with gaseous
hydrogen produced a species capable of achieving higher activity towards C–O–C bond cleavage
compared to the Ru/C and Pd/C catalysts. The selectivity of the C–O–C bond cleavage over the Ni/C
catalyst was 85%, and higher than the corresponding values in the Ru/C (40%) and Pd/C (69%)
systems. Using the carbothermal reduction method for the production of the Ni/C‐C catalyst, the
conversion and selectivity levels reached 99%, with 40% of the benzene rings in PPE being re‐
served. In comparison, no benzene ring containing products wer observed over the noble metal
catalysts. This difference was attributed to the interaction between the carbon support and the
nickel nanoparticles.
Received 30 December 2012
Accepted 30 January 2013
Published 20 April 2013
Keywords:
Lignin model compound
Phenethyl phenyl ether
Hydrogenation and cleavage
Nickel catalyst
Aromatic compound
© 2013, Dalian Institute of Chemical Physics, Chinese Academy of Sciences.
Published by Elsevier B.V. All rights reserved.
Aromatic compounds are important bulk chemicals in the
petroleum oil industry, with the majority of these compounds
being produced from fossil fuels such as petroleum, coal and
natural gas. The production of aromatic compounds from fossil
fuels, however, represents a non‐sustainable and challenging
supply chain, because of the dwindling supplies of fossil fuels
and their nonrenewable properties. The development new of
methods for the production of aromatic compounds from re‐
newable biomass resources is therefore of critical importance.
Lignin is a three‐dimensional natural polymer, which serves
as a glue to adhere cellulose and hemicelluloses and provide
strength to plants. The basic unit of lignin is phenylpropane,
with these units being connected to each other via C–O–C
bonds (accounting for 2/3–3/4 of all linkages) and C–C bonds
(accounting for 1/4–1/3 of all linkages). Aromatic compounds
could be obtained from lignin if a suitable method was devel‐
oped to exploit this natural resource. For example, the devel‐
opment of techniques for the selective cleavage of the C–O–C
bonds without disrupting the benzene rings would provide
some useful aromatic compounds. The efficient use of lignin to
produce aromatic compounds therefore represents a poten‐
tially sustainable route to aromatic materials.
In light of complex structure and composition of native lig‐
nin, methods for investigating C–O bond cleavage in lignin
model compounds and studying the associated cleavage rules
have been widely studied. In the majority of cases, research has
focused on cleaving the C–O bonds of lignin model compounds
using homogeneous catalysis. For example, Toste’s group [1]
*Corresponding author. Tel/Fax: +86‐411‐84379762; E‐mail: wangfeng@dicp.ac.cn
# Corresponding author. Tel/Fax: +86‐411‐84379245; E‐mail: xujie@dicp.ac.cn
This work was supported by the National Natural Science Foundation of China (21073184, 21273231, 21233008) and One Hundred Person Project of
the Chinese Academy of Sciences.
DOI: 10.1016/S1872‐2067(12)60535‐X | http://www.sciencedirect.com/science/journal/18722067 | Chin. J. Catal.,Vol. 34, No. 4, April 2013

SONG Qi et al. / Chinese Journal of Catalysis 34 (2013) 651–658
developed a series of vanadium–oxo complexes and applied
catalyst was reduced in a horizontal furnace in a N2 flow (30
ml/min) at 450 °C for 2 h. The Ru/C (5 wt%) and Pd/C (5 wt%)
catalysts used in this study were purchased from Aladdin Rea‐
gent Co. Ltd (Shanghai, China).
them to the non‐oxidative cleavage of the C–O bonds of dimeric
lignin model compounds to give aryl enones with 95% conver‐
sion. Hansen’s group [2–4] also studied a vanadium complex
catalyzed lignin model oxidation reaction, and reported that the
C–O and C–C bonds could be selectively cleaved by changing
the ligand involved in the catalysis [5]. The use of hydrogena‐
tion as another potential method for the conversion of lignin
has also been the subject of considerable attention. Hartwig’s
group [6] recently reported the use of a nickel carbene complex
as an effective catalyst for the hydrogenolysis of the aryl ether
bonds in several lignin model compounds in the presence of the
organic base NaOtBu, achieving 99% conversion. Bergman’s
group [7] reported the use of a ruthenium xantphos catalyst for
the cleavage of the C–O bond of a lignin model compound using
transfer hydrogenation with a 62%–98% conversion.
The phenethyl phenyl ether (PPE) reactant was synthesized
from phenol and 1‐bromo‐2‐phenylethane using the William‐
son ether synthesis method [12]. Briefly, powdered caustic
soda (18.0 g) was added to a solution of phenol (47.0 g) in
benzene (200 g), and the resulting mixture was heated to 95 °C.
1‐Bromo‐2‐phenylethane (46.0 g) was then added, and the
resulting mixture was stirred for 12 h at 95 °C. Upon comple‐
tion of the reaction, the mixture was cooled, and washed and
extracted with water and benzene, respectively. The solvent
was then removed in vacuo to give the crude PPE product,
which was purified by column chromatography. The purity of
the final PPE product was great than 99.6% (determined by
1
There are several advantages associated with the use of
homogeneous catalysts, including their high catalytic perfor‐
mance and controllable activities. The main limitations of using
homogenous catalysts, however, including difficulties associ‐
ated with their separation and recycling. In contrast, heteroge‐
neous catalysts can be easily separated from reaction mixtures
and reused. The choice of conversion method is also an im‐
portant factor in determining the efficiency of the lignin con‐
version. For example, the oxidation reactions of lignin can ex‐
hibit relative poor selectivity towards the oxidation of the C–O
and C–C bonds, resulting in the over oxidation of lignin and
damage to the benzene ring [8]. In contrast, the hydrogenation
reactions of lignin can cleave of C–O bonds selectively and re‐
serve benzene ring. These reactions therefore provide good
atomic economy with low levels of hydrogen consumption. In
our previous studies, we developed a method of the hydro‐
genation of lignosulfonate into aromatic compounds with 60%
conversion and 70%–95% selectivity [9]. These results
demonstrated that nickel not only has a high level of activity for
C–O bond cleavage but that it also possesses a propensity to‐
wards the preservation of the benzene ring. To further investi‐
gate the C–O bond hydrocracking and gaining C–O cleavage
rule, we prepared a series of carbon supported nickel catalysts
and studied the hydrocracking behavior of a lignin model
compound over the nickel catalysts.
GC), and its structure was confirmed by H NMR analysis and
mass spectroscopy. The PPE hydrocracking reaction was car‐
ried out in 75 ml autoclave under 2 MPa of hydrogen pressure.
The reaction products were qualified by gas chromatog‐
raphy‐MS (GC‐MS) (Agilent 6890 N GC/5973 MS, SE‐54 col‐
umn) and quantified by GC (Agilent 7890, FID detector, HP‐5
column, 30 m × 0.25 mm × 0.32 mm)
In our recent report, a nickel precursor supported on acti‐
vated carbon was completely reduced to the nickel nanoparti‐
cles (d = 20 nm) at 400 °C [11]. However, the exact nickel re‐
duction temperature of the carbothermal reduction method is
not clear. As a result, we attempted to reduce the Ni/C precur‐
sor using the carbothermal method in N2 at temperatures of
300, 400, and 450 °C. The Ni/C products obtained in each case
were characterized by X‐ray diffraction (XRD) analysis (see Fig.
1). The curves B and C in Figure 1 represent the Ni/C treated at
400 and 300 °C, respectively. The peaks positioned at 2θ = 37°,
43°, and 63° were assigned to NiO (PDF 47‐1049), suggesting
that nickel nitrate decomposed to NiO, which was reduced no
further by the carbon when the reduction temperature was
below 400 °C. When the Ni/C precursor was reduced at a tem‐
perature of 450 °C, however, the XRD pattern of the Ni catalyst
was different to those for the samples treated at 400 °C (curve
B) and 300 °C (curve C). The XRD pattern of the sample heated
Nickel supported on activated carbon catalysts (Ni/C) were
prepared by the incipient‐wetness impregnation method and
further reduced by hydrogen (denoted as Ni/C‐H) or by the
carbothermal reduction (denoted as Ni/C‐C) method [10,11].
The nickel metal loading in these catalysts was set as 10 wt %.
The catalyst precursors were typically prepared as follows:
nickel nitrate (II) hexahydrate (Ni(NO3)2·6H2O, 2.76 g) was
added to water (8.0 ml) in a beaker. The solution was stirred
for 30 min to give an impregnation stock solution. Activated
carbon (5.0 g, water absorption capacity 1.8 ml/g) was added
to this solution and the beaker was then covered with a culture
dish to keep the sample wet for 24 h. The sample was then
dried at 120 °C overnight before being milled (80 mesh). The
Ni/C‐H catalyst was reduced in a horizontal furnace in a H2
flow (30 ml/min). The reduction temperature reached 400 °C
in 60 min and was maintained at 400 °C for 2 h. The Ni/C‐C
Ni(111)
(1)
(2)
Ni(200)
Ni(220)
NiO(200)
NiO(111)
NiO(200)
(3)
20
30
40
50
60
70
80
2/(o )
Fig. 1. XRD patterns of the Ni/C catalyst reduced using the carbother‐
mal method in N2 at 450 °C (1), 400 °C (2), and 300 °C (3).

SONG Qi et al. / Chinese Journal of Catalysis 34 (2013) 651–658
40
(a)
(b)
d = 13 nm
30
20
10
0
0
4
8
12
16
20
24
28
50nm
Diameter (nm)
Fig. 2. TEM image (a) and Ni particle size distribution (b) of the Ni/C catalyst carbothermally reduced at 450 °C.
at 450 °C contained new diffraction peaks at 2θ = 44°, 52°, and
contacts and consequently formed a surface with close contacts
between the nickel and the carbon support, as illustrated in Fig.
3. The occurrence of such close surface contact plays an im‐
portant role in enhancing nickel‐carbon interactions, and can
potentially result in enhanced catalytic activity.
76°, which were assigned to metal Ni (PDF 3‐65‐2865). This
result suggested that the nickel nitrate had been reduced by the
carbon support at 450 °C.
A TEM image of the Ni/C‐C material reduced at 450 °C is
shown in Fig. 2. The average size of the nickel nanoparticles
within this material was found to be 12 nm (calculated from
120 particles), which was smaller than that of the Ni/C‐H sam‐
ple (20 nm) [11]. In the carbothermal reduction method, the
activated carbon not only served as a support but also acted as
a reducing agent upon which nickel nitrate was reduced to
nickel metal at 450 °C (see Fig. 1 (1) and (2)). During the car‐
bothermal process, the nickel nitrate was initially decomposed
to NiO. Further increases in the temperature led to the reduc‐
tion of the NiO to nickel particles by the carbon support at the
nickel‐carbon interface, with some of the carbon being simul‐
taneously oxidized to COx [13,14]. During the reduction pro‐
cess, these COx species were effectively swept away by the flow
of N2 through the system. The consumption of carbon species
during Ni reduction resulted in the formation of a concave hole
in the carbon surface (Fig. 3) which was supported by the re‐
sults of BET analysis. The BET result showed that a mesopore
with a 7–15 nm pore size was present following the completion
of the carbothermal reduction process. The presence of the
concave hole allowed the Ni nanoparticles to burrow into the
surface and become partially embedded in the carbon matrix.
The nickel nanoparticles did not grow into larger particles be‐
cause of the restriction effect imposed by the concave hole. As a
result, the diameter of the nickel particles in Ni/C‐C material (d
= 13 nm) were smaller than those in the Ni/C‐H material (d =
20 nm). This reduction method created many Ni‐carbon point
We initially conducted hydrogenation and cleavage reac‐
tions of the PPE over different catalysts, including Pd/C, Ru/C
and Ni/C‐H, and compared the hydrogenation and cleavage
abilities of the different catalysts towards the C–O bonds in
PPE. The reactions were carried out in a 75 ml autoclave. Brief‐
ly, a portion of the catalyst (25 mg) was added to a solution of
the PPE (0.167 mmol) in methanol (10 ml), and the resulting
mixture was stirred at 150 °C under 2 MPa of H2 pressure for 2
h. The reaction was then quenched and analyzed by GC‐MS and
GC. The results are shown in Scheme 1. The material balance of
PPE in all of these runs was found to be 96%. No small alkane,
alcohol or ether compounds were detected in the reaction gas
according to analysis by MS. The Pd/C, Ru/C and Ni/C‐H cata‐
lysts exhibited high levels of activity towards the conversion of
PPE, providing conversions in excess of 99%. The three differ‐
ent catalysts, however, showed different selectivities towards
the C–O bond cleavage.
For the Ru/C catalyst, all of the products were found to be
saturated cyclohexane derivates, such as cyclohexanol (19%),
ethylcyclohexane (21%), and the fully hydrogenated PPE
product (60%). The C–O cleavage selectivity of PPE at full con‐
version over the Ru/C catalyst was only 40%. Based on the
product distribution, it was concluded that the major reaction
involved in this reaction was the hydrogenation of the benzene
rings rather than C–O bond cleavage. This result indicated that
the hydrogenation of the benzene rings was preferred to the
cleavage of the C–O bond over the Ru/C catalyst. PPE was also
converted to cyclohexane derivates over the Pd/C catalyst but
with a higher selectivity towards C–O bond cleavage (69%).
Ethylcyclohexane (32%) was observed as one of the major
products formed from the hydrogenation reaction of PPE over
Pd/C. The occurrence of this product was attributed to palla‐
dium’s hydrodeoxygenation ability [15]. For the Ni/C‐H cata‐
lyst, PPE was converted to cyclohexanol (50%) and ethylcyclo‐
hexane (35%) with >99% conversion and 85% C–O cleavage
selectivity. The selectivity result obtained for the Ni/C‐H cata‐
lyst was higher than those of the Ru/C and Pd/C catalysts, sug‐
Fig. 3. Representation of the Ni/C catalyst reduced by H2 (A) and acti‐
vated carbon (B), respectively.

SONG Qi et al. / Chinese Journal of Catalysis 34 (2013) 651–658
Scheme 1. Hydrogenation and cleavage of phenethyl phenyl ether over Ru/C, Pd/C and Ni/C‐H catalysts.
Scheme 2. Hydrogenation and cleavage of phenethyl phenyl ether over Ni/C‐C catalyst.
gesting that Ni had a pronounced ability to cleave C–O bonds,
which was consistent with similar reports from the literature
[16,17].
higher level of hydrogen activity, which was in direct contra‐
diction of the general rule provided above. The fact that the
capacity of the Ni catalysts to facilitate benzene hydrogenation
decreased with a reduction in the size of the nickel nanoparti‐
cles suggested that the benzene hydrogenation reaction was
governed by factor other than the size effect.
During the reduction of the Ni/C‐H catalyst, nickel nitrite
was reduced by hydrogen at 400 °C. The activated carbon
served as an intact support and was not involved in the reduc‐
tion process. Thus, the nickel‐carbon interaction in this case
was weak (see Fig. 3). In the carbothermal reduction method,
however, the nickel nanoparticles became semi‐embedded in
the carbon support because of the consumption of carbon at
the nickel‐carbon interface. This structure led to an increase in
Although the Ni/C‐H reduced by H2 provided a good per‐
formance towards C–O bond cleavage, the PPE molecule was
completely hydrogenation to the corresponding cyclohexyl
products. To retain the benzene rings in the products, we pre‐
pared the Ni/C‐C catalyst which was reduced by the car‐
bothermal method, and applied it to the PPE reaction. As
showed in Scheme 2, the hydrocracking of the PPE model mol‐
ecule occurred with >99% conversion and >99% C–O bond
selectivity over the Ni/C‐C catalyst, demonstrating that the
Ni/C‐C catalyst showed an excellent performance towards C–O
bond cleavage. Furthermore, aromatic compounds accounted
for 44% of the final products, indicating that that the Ni/C‐C
catalyst favored the preservation of the benzene rings in the
PPE.
Compared with traditional homogeneous catalysts, the
Ni/C‐C catalyst could be easily separated from the product
mixture using a magnetic field (Fig. 4).
It is well known that the size of the nanoparticles in a cata‐
lyst has a significant effect on the reaction activity. Generally,
the smaller the nanoparticles in a catalyst, the higher the activ‐
ity of the catalyst [18,19]. For the nickel‐catalyzed benzene
hydrogenation reaction, the Ni/C‐C catalyst (d = 13 nm) hy‐
drogenated 56% of the benzene rings in the PPE (retained 44%
of the benzene rings in the product), whereas the Ni/C‐H cata‐
lyst (d = 20 nm) hydrogenated >99% of the benzene rings.
These results suggested that the larger nickel size provided a
Fig. 4. Photograph of the Ni/C‐C catalyst separated by a magnet fol‐
lowing the reaction.

SONG Qi et al. / Chinese Journal of Catalysis 34 (2013) 651–658
Graphical Abstract
Chin. J. Catal., 2013, 34: 651–658 doi: 10.1016/S1872‐2067(12)60535‐X
Hydrogenation and cleavage of the C–O bonds in the lignin model
compound phenethyl phenyl ether over a nickel‐based catalyst
SONG Qi, CAI Jiaying, ZHANG Junjie, YU Weiqiang, WANG Feng*, XU Jie*
Dalian Institute of Chemical Physics, Chinese Academy of Sciences;
Graduate University of Chinese Academy of Sciences
The C–O bond in the lignin model compound, phenethyl phenyl ether,
was successfully hydrogenated and cleaved over a nickel‐based catalyst,
which had been prepared using the carbothermal reduction method.
[2] Sedai B, Diaz‐Urrutia C, Baker R T, Wu R L, Silks L A, Hanson S K.
ACS Catal, 2011, 1: 794
[3] Hanson S K, Baker R T, Gordon J C, Scott B L, Thorn D L. Inorg
Chem, 2010, 49: 5611
[4] Zhang G Q, Scott B L, Wu R L, Silks L A, Hanson S K. Inorg Chem,
2012, 51: 7354
[5] Hanson S K, Wu R, Silks L A P. Angew Chem, Int Ed, 2012, 51: 3410
[6] Sergeev A G, Hartwig J F. Science, 2011, 332: 439
[7] Nichols J M, Bishop L M, Bergman R G, Ellman J A. J Am Chem Soc,
2010, 132: 12554
[8] Gaspar A R, Gamelas J A F, Evtuguin D V, Neto C P. Green Chem,
2007, 9: 717
[9] Song Q, Wang F, Xu J. Chem Commun, 2012, 48: 7019
[10] Yu W Q, Zhao J, Ma H, Miao H, Song Q, Xu J. Appl Catal A, 2010, 383:
73
the level of nickel‐carbon interaction. Furthermore, the surface
structure of the carbon was a graphene in which the  electrons
were delocalized. Based on the close interaction between the
carbon and nickel atoms, the  electrons in Pz orbitals of the
carbon could become hybridized with the d electrons of the
nickel, leading to an increase in the electron mobility [20–22],
which may result in the nickel bearing a negative charge.
It is generally believed that reaction activity of benzene hy‐
drogenation is proportional to the amount of positive charges
on surfaces of metal nanoparticles [23–25]. For supported cat‐
alysts, this positive charge is predominantly formed as a con‐
sequence of the strong interactions between the metal and the
support. In the Ni/C‐C catalyst, the Pz  electrons of the carbon
surface could have become hybridized with the nickel d orbit‐
als and subsequently donated to nickel, resulting in negatively
charged nickel nanoparticles, which could lead to a decrease in
the benzene hydrogenation activity. The benzene ring would
therefore be preserved in the products, as observed, accounting
for the 44% of aromatic compounds in the final products fol‐
lowing the hydrocracking of PPE over the Ni/C‐C catalyst.
In summary, the lignin model compound phenethyl phenyl
ether was effectively hydrocracked over a nickel‐based catalyst.
The Ni/C‐C catalyst was produced by reduction using the car‐
bothermal method, and exhibited high levels of activity and
C–O bond selectivity. Compared with the noble catalysts, the
Ni/C‐C catalyst showed better benzene ring selectivity, with the
difference being attributed to the strong nickel‐carbon interac‐
tions in the catalyst. These results provide a better under‐
standing of the C–O bond hydrocracking rule and will be useful
for the design of new high performance catalysts.
[11] Yu W Q, Xu J, Ma H, Chen C, Zhao J, Miao H, Song Q. Catal Commun,
2010, 11: 493
[12] Klein M T, Virk P S. Ind Eng Chem Fundamen, 1983, 22: 35
[13] L’vov B V. Thermochim Acta, 2000, 360: 109
[14] Yao J Y, Li L X, Song H H, Liu C Y, Chen X H. Carbon, 2009, 47: 436
[15] Zhao C, Lercher J A. Angew Chem, Int Ed, 2012, 51: 5935
[16] Cao L Y, Jia Z G, Ji S F, Hu J Y. J Nat Gas Chem, 2011 20: 377
[17] Lu M, Lü P M, Yuan Zh H, Li H W, Xu J L. Chin J Catal, 2011, 32:
1017
[18] Carter J L, Cusumano J A, Sinfelt J H. J Phys Chem, 1966, 70: 2257
[19] Raberg L B, Jensen M B, Olsbye U, Daniel C, Haag S, Mirodatos C,
Sjastad A O. J Catal, 2007, 249: 250
[20] Khomyakov P A, Giovannetti G, Rusu P C, Brocks G, van den Brink
J, Kelly P J. Phys Rev B, 2009, 79: 195425
[21] Stokbro K, Engelund M, Blom A. Phys Rev B, 2012, 85: 165442
[22] Barraza‐Lopez S, Vanević M, Kindermann M, Chou M Y. Phys Rev
Lett, 2010, 104: 076807
[23] Feast S, Lercher J A. Stud Surf Sci Catal, 1996, 102: 363
[24] Simon L J, Kooyman P J, van Ommen J G, Lercher J A. Appl Catal A,
2003, 252: 283
References
[25] Bogomolova O, Ostrovskii N, Smolikov M, Belyi A, Duplyakin V.
React Kinet Catal Lett, 1989, 40: 131
[1] Son S, Toste F D. Angew Chem, Int Ed, 2010, 49: 3791