Promotional effect of Co and Ni on MoO3 catalysts for hydrogenolysis of dibenzofuran to biphenyl under atmospheric hydrogen pressure

Article abstract of DOI:10.1016/j.jcat.2020.01.035

Co(Ni)/MoO₃ catalysts were prepared, characterized and evaluated for the hydrogenolysis of bio-derived dibenzofuran (DBF), aiming at the understanding of the promoting effect of Co/Ni on MoO₃ for high-yield production of aromatic products. All Co(Ni)/MoO₃ catalysts selectively cleaved C-O bond, thus effectively transformed DBF to biphenyl (BP) at relatively moderate conditions. The experimental results from varying Mo species by adjusting the reduction temperature of MoO₃ together with XPS and in-situ XRD characteristic were evident that Mo⁵⁺ species was responsible as the major active specie for the reaction. Promotional effect between Co(Ni) and Mo in Co(Ni)/MoO₃ catalysts was observed, resulted from the presence of acidic Co(Ni)MoO₄ species and a large number of Mo⁵⁺ species both of which were created local to the Co(Ni)-O-Mo interface, as can be characterized by in-situ XRD, XPS, H₂-TPR, NH₃-TPD, in-situ FT-IR, Raman and TEM. The trend of the initial reaction rate follows: MoO₃ (0.18 μmol?g_cat^?1?s^?1) < Ni/MoO₃ (0.26 μmol?g_cat^?1?s^?1) < Co/MoO₃ (0.29 μmol?g_cat^?1?s^?1), corresponding to the decreasing activation barrier. And the best catalytic activity was observed for the 100% yield of BP over Co/MoO₃. A possible mechanism, including Co(Ni) facilitated reduction of Mo⁶⁺ to Mo⁵⁺ and Co(Ni) enhanced formation of acidic sites, is proposed to be responsible for the high activity in Co(Ni)/MoO₃ catalysts.

Full text of DOI:10.1016/j.jcat.2020.01.035

Journal of Catalysis 383 (2020) 311–321
Contents lists available at ScienceDirect
Journal of Catalysis
journal homepage: www.elsevier.com/locate/jcat
Promotional effect of Co and Ni on MoO3 catalysts for hydrogenolysis of
dibenzofuran to biphenyl under atmospheric hydrogen pressure
Jie Zhang a, Chuang Li a, Weixiang Guan a, Xiaozhen Chen a, Xiao Chen a, Chi-Wing Tsang b, Changhai Liang a,
a State Key Laboratory of Fine Chemicals, Laboratory of Advanced Materials and Catalytic Engineering, School of Chemical Engineering, Dalian University of Technology, Dalian
⇑
116023, China
b Faculty of Science and Technology, Technological and Higher Education Institute of Hong Kong, Hong Kong, China
a r t i c l e i n f o
a b s t r a c t
Article history:
Co(Ni)/MoO3 catalysts were prepared, characterized and evaluated for the hydrogenolysis of bio-derived
dibenzofuran (DBF), aiming at the understanding of the promoting effect of Co/Ni on MoO3 for high-yield
production of aromatic products. All Co(Ni)/MoO3 catalysts selectively cleaved C-O bond, thus effectively
transformed DBF to biphenyl (BP) at relatively moderate conditions. The experimental results from vary-
ing Mo species by adjusting the reduction temperature of MoO3 together with XPS and in-situ XRD char-
acteristic were evident that Mo5+ species was responsible as the major active specie for the reaction.
Promotional effect between Co(Ni) and Mo in Co(Ni)/MoO3 catalysts was observed, resulted from the
presence of acidic Co(Ni)MoO4 species and a large number of Mo5+ species both of which were created
local to the Co(Ni)-O-Mo interface, as can be characterized by in-situ XRD, XPS, H2-TPR, NH3-TPD, in-
situ FT-IR, Raman and TEM. The trend of the initial reaction rate follows: MoO3
Received 11 August 2019
Revised 3 January 2020
Accepted 31 January 2020
Keywords:
Hydrogenolysis
Dibenzofuran
Co(Ni)/MoO3
Biphenyl
(0.18
l l l
mol∙gcꢀa1t∙sꢀ1) < Ni/MoO3 (0.26 mol∙gcꢀa1t∙sꢀ1) < Co/MoO3 (0.29 mol∙gꢀca1t∙sꢀ1), corresponding to
Promotional effect
the decreasing activation barrier. And the best catalytic activity was observed for the 100% yield of BP
over Co/MoO3. A possible mechanism, including Co(Ni) facilitated reduction of Mo6+ to Mo5+ and Co
(Ni) enhanced formation of acidic sites, is proposed to be responsible for the high activity in Co(Ni)/
MoO3 catalysts.
Ó 2020 Elsevier Inc. All rights reserved.
1. Introduction
of biomass-derived model compounds has been extensively stud-
ied and the exploration of the reaction mechanisms has provided
The continuous depletion of fossil-fuel, the increasing demand
on transportation fuels and the ceaseless deterioration of the envi-
ronment during these last years arouse a growing need to develop
the use of the biomass as potential source for the production of
fuels and value-added chemicals [1–5]. However, the high oxygen
content of the bio-oil resulting from pyrolysis of biomass, leads to
some deleterious properties of the products that include high vis-
cosity, corrosiveness and thermal instability [6,7]. Thus
hydrogenolysis is one of the most prevalent and efficient strategies
to produce aromatic compounds, which consists of breaking the
C-O bond of oxygenated molecules under hydrogen pressure [8].
The process that bio-oil being directly converted into value-
added aromatic chemicals, seems to be an attractive route for
enhancing the utilization of biomass [3,9,10] with higher effi-
ciency. Thus, during the past several years, catalytic conversion
an insight into effective conversion of bio-oils.
Past efforts for hydrodeoxygenation catalyst development were
focused on conventional hydrotreating catalysts, particularly in the
petroleum refinery for several decades, such as CoMo- and NiMo-
based sulfides [11–16] and supported noble metal catalysts
[1,12,17–19]. The sulfided catalysts are active hydrodeoxygenation
catalysts, nevertheless, the presence of additional sulfur chemicals
for the purpose of avoiding catalyst deactivation results in the
inevitable sulfur contamination in the products. Besides, noble
metal catalysts could result in full ring saturation, resulting in
products with a lower octane number and an increased level of
H2 consumption [20,21]. Thus, more attention has been paid to
development of novel catalysts, such as transition metal oxides
systems. Recently, metal oxides are considered as more active cat-
alysts capable of C-O bond activation than commonly employed
transition metal catalysts [22]. Specifically, molybdenum oxide
shows efficient cleavage of C-O bonds on hydrogenation of molyb-
denum centers[10] under mild conditions, and thus exhibits high
selectivity to aromatics for the selective cleavage of C-O bonds in
⇑
Corresponding author.
E-mail address: changhai@dlut.edu.cn (C. Liang).
URL: http://amce.dlut.edu.cn (C. Liang).
https://doi.org/10.1016/j.jcat.2020.01.035
0021-9517/Ó 2020 Elsevier Inc. All rights reserved.

312
J. Zhang et al. / Journal of Catalysis 383 (2020) 311–321
exploring lignin derivatives upgrading [10,23–25]. Roman-Leshkov
et al. demonstrated that MoO3 effectively catalyzed the
hydrogenolysis of lignin-derived oxygenates to produce high yields
of aromatic hydrocarbons without ring saturation [26]. It was also
reported the selective conversion of guaiacol over MoO3 to produce
various alkylphenols in ethanol without the addition of gaseous
hydrogen, with 99% conversion and 94% selectivity of alkylphenols
[9]. Moreover, supported MoO3 catalysts also showed preferable
activity of C-O cleavage [4,10,27]. All these important contribu-
tions indicate the possibility for MoO3 to perform excellent ability
of C-O cleavage in moderate conditions. However, compared with
noble metal catalyst, MoO3 was relatively less active, and thus
the improvement of reactivity was crucial for the development of
Mo-based catalyst. Therefore, the rational design of catalysts,
which can selectively cleave the C-O bond and improve reaction
activity without aromatic saturation under mild conditions, are
attracting increasing attention. In previous reports, Co(Ni) as the
promoter could significantly enhance the activity of molybdenum
sulfide [13], and Co addition to supported molybdenum catalyst
enhanced the total hydrodeoxygenation selectivity by 45% in the
hydrodeoxygenation of anisole [28]. While, to the best of our
knowledge, no systematic studies have been performed to investi-
gate the promotion effect of Co and Ni on MoO3 catalyst and reac-
tion mechanism for hydrodeoxygenation reactions.
Previous reports have shown that MoO3 follow a reverse redox
reaction mechanism that would result in the removal of the oxy-
gen atom from the oxygenates upon the adsorption of the sub-
strates on the oxygen vacancy sites to form a Mo-O bond with
subsequent regeneration of the vacancy with H2 to produce water
[26]. The oxygen in the reactant then fills the oxygen vacancy,
leading to oxygen transfer to the catalyst and the formation of
the unsaturated product [25,29]. Considering the origin of the cat-
alytic activity of these materials, it is hypothesized that the oxygen
vacancy (i.e., an uncoordinated metal site on the metal oxide) is
responsible for hydrodeoxygenation over both pure and supported
MoO3 catalysts [25,30,31]. Shanks et al. reported that the MoO3
catalyst needed to be reduced to a more active form in order to
fully deoxygenate the pyrolysis vapors [25]. Therefore, the catalyst
requires a constant stream of H2 for continuous reduction and
vacant site formation [25]. Moreover, Roman-Leshkov and co-
workers used combined Raman, XRD, and XPS spectroscopy aug-
mented with the deactivation study, to reveal that the activity of
MoO3 could be predominantly promoted by uncoordinated Mo5+
species, whereas Mo4+ species should be related to the less activity
for hydrodeoxygenation reaction [2,21,25,31,32] corresponding to
the deactivation reason of MoC2 that was subsequently trans-
formed to MoO2 [24]. The above studies suggest the critical role
of intermediate state(s) of Mo species on its catalytic performance.
Moreover, extensive literatures have been reported on the acidic
sites that could also serve as active sites to adsorb the oxygen atom
of oxygen-containing compounds and then improve the activity of
catalysts [8,33].
mechanistic insights into the active sites of catalysts in the
hydrogenolysis of lignin-derived DBF is key to provide new
insights for catalyst design and potential applications for bio-oils
upgrading.
2. Experimental
2.1. Materials
Ammonium molybdate ((NH4)6Mo7O24ꢁ4H2O, ꢂ99.0%), cobalt
nitrate (Co(NO3)2ꢁ6H2O, ꢂ98.5%) and nickel nitrate (Ni(NO3)2ꢁ6H2-
O, ꢂ98.0%) were obtained from Sinopharm Chemical Reagent Co.
Methanol (CH3OH, 99.8%) and n-decane (C10H22, 98%) were pur-
chased from Tianjin Kermel Chemical Reagent Co. Dibenzofuran
(C12H8O, 98.0%) and n-dodecane (C12H26, 99.0%) were obtained
from Aladdin Chemical Reagent Co. All of the materials were ana-
lytical reagent grade and utilized without further purification.
2.2. Preparation of catalysts
MoO3 was prepared by calcining (NH4)6Mo7O24 under Ar/O2
(40/20 mL minꢀ1) flow at 500 °C for 4 h. 3 wt% Co(Ni)/MoO3 were
prepared by the wetness impregnation method with methanol
solution of Ni(NO3)2 or Co(NO3)2. Typically, a certain amount of
Ni(NO3)2ꢁ6H2O or Co(NO3)2ꢁ6H2O was dissolved in 60 mL methanol
at room temperature and then 2.00 g MoO3 was added under vig-
orous stirring for 8 h. After that, methanol was removed using a
rotary evaporator at 40 °C and the slightly damp sample was dried
overnight at 80 °C. Finally, the samples were calcined at 400 °C for
4 h under Ar/O2 (40/20 mL minꢀ1) flow for further characterization
and catalytic measurements. Co(Ni)MoO4 were prepared using a
precipitation method described in detail in the Supporting
Information.
MoO3 reduced at different temperature were denoted as MoO3-
n (n = 300, 400, 500 and 600 °C). In addition, MoO3, Ni/MoO3 and
Co/MoO3 represented catalysts reduced at 300 °C.
2.3. Characterization of catalysts
2.3.1. Nitrogen adsorption
N2 adsorption–desorption isotherms of the catalysts were mea-
sured at ꢀ196 °C using a Quantachrome Autosorb IQ instrument.
Prior to these measurements, the sample was loaded in a glass tube
and outgassed at 200 °C under vacuum for 8 h to remove any vola-
tile adsorbates from the surface. The resulting adsorption iso-
therms were used to calculate the specific surface area (S) by the
Brunauer-Emmett-Teller (BET) method. The average pore volume
(Vp) and the pore diameter (dp) was obtained using the Barrett-
Joyner-Halenda (BJH) method.
2.3.2. In-situ X-ray diffraction
In-situ X-ray diffraction (XRD) patterns were recorded on a
Rigaku Smartlab instrument, using a Cu Ka monochromatized radi-
ation source. Diffraction patterns were collected in the range of 2h
from 5° to 90° with a scan speed of 8° minꢀ1, operated at 40 kV and
100 mA. Prior to test, the samples were placed into the reaction
chamber and reduced in flowing H2 (40 mL minꢀ1) at correspond-
ing temperature for 2 h, and then the diffraction patterns were
collected.
In the present work, we select dibenzofuran (DBF) as the lignin
model compound [34–38] and focus on the influence of Co(Ni) over
MoO3 on the catalytic activity of catalysts during the hydrogenol-
ysis of DBF to biphenyl at atmospheric hydrogen pressure. First,
active sites on pure MoO3 for the reaction was determined by alter-
ing the reduction temperature of MoO3 to expose the diverse Mo
species, combined with the characterization techniques such as
XPS and in-situ XRD. Then, supported catalysts were synthesized
and then used to investigate the activity in order to identify the
structural-activity relationship of catalysts as evidenced by in-
situ XRD, in-situ FT-IR, Raman, XPS, H2-TPR and NH3-TPD. As a
result, an insight into a possible reaction mechanism was proposed
to illustrate the role of the surface Mo5+ and promotional Co(Ni) on
catalytic performance over the present catalysts. Developing
2.3.3. Temperature-programmed desorption
The acidic properties of catalysts were characterized by using
the temperature-programmed desorption of ammonia (NH3-TPD).
Experiments were performed on a CHEMBET-3000 chemisorption
instrument. Prior to the experiments, the calcined samples

J. Zhang et al. / Journal of Catalysis 383 (2020) 311–321
313
(0.10 g) were pre-reduced with 10% H2/He at 400 °C for 2 h at a
heating ramp rate of 10 °C minꢀ1. After pretreatment, the temper-
ature was cooled to 120 °C and the sample was saturated with a
10% NH3/He stream for 1 h. After removing physically absorbed
NH3 by purging with He at 120 °C for 1 h, the samples were heated
to 500 °C at a heating rate of 10 °C minꢀ1 in a He flow. The amount
of acidic sites on the catalyst was calculated from the desorption
amount of NH3, which was determined by measuring the areas
of the desorption profiles.
900 °C at a continuous heating ramp of 5 °C minꢀ1. The response
of the detector was recorded by gas chromatograph using the TCD.
2.3.10. Elemental analysis (ICP-AES)
The elemental composition of as-prepared catalysts was mea-
sured by means of inductively coupled plasma atomic emission
spectroscopy (ICP-AES) on a Perkin-Elmer Optima 2000 DV device.
2.4. Catalytic activity
2.3.4. In-situ Fourier transform infrared spectra
Fourier transform Infrared (FT-IR) spectra was collected at room
temperature on a Thermo Fisher iN10 spectrometer with a resolu-
tion of 4 cmꢀ1 for 16 scans in the region 400–4000 cmꢀ1. The pel-
lets were prepared by mixing 10 mg of sample in 150 mg KBr and
then placed into an in-situ environmental cell. Afterward, the sam-
ples were heated from 30 °C to 300 °C with a ramp rate of 5 °-
C minꢀ1 and maintained for 2 h in 40 mL minꢀ1 H2 flow. After
the reactor was cooled to room temperature, H2 was switched to
He with a total flow rate of 40 mL minꢀ1 for 20 min and then the
spectra were recorded.
The hydrogenolysis of DBF was performed in a vapor-phase
down-flow fixed-bed reactor, as described in our previous works
[8,38]. The temperature inside the reactor was monitored by a K-
type thermocouple extended into the catalyst bed. The catalyst
(0.40 g) was diluted with 60–80 mesh quartz SiO2 (5 mL) to keep
the catalyst bed isothermal. Prior to catalytic tests, all the synthe-
sized catalysts were reduced in fixed-bed reactor at different tem-
perature (300 °C, 400 °C, 500 °C and 600 °C) and 1.0 MPa for 2 h
with a hydrogen flow rate of 40 mL minꢀ1. Then, the temperature
and pressure of the reactor were adjusted to reaction conditions
(0.1 MPa H2 pressure and 360 °C). The reactants consisted of
2.0 wt% DBF, 1.0 wt% n-dodecane (as an internal standard), and
97.0 wt% n-decane (as solvent), were fed into the reactor from
the top of the reactor by a pump along with H2 gas. The mixture
of reactant and H2 gas moved downward and reacted in the middle
of the reactor. Catalytic tests were conducted until reaching the
2.3.5. Raman spectroscopy
Raman spectra was recorded at room temperature on a Thermo
Scientific DXR Raman Microscope (Renishaw England) equipped
with a liquid-nitrogen cooled charge coupled device (CCD) detector
and a confocal microscope. The line at 532 nm Ar laser was used as
excitation source and the scanning range was 50–3500 nmꢀ1. The
laser was focused on the sample under a microscope with the
steady state. The Weighted Hourly Space Velocity (WHSV, hꢀ1
)
was calculated from the catalyst weight and the reactant flow as
defined as follows:
diameter of the analyzed spot being ~1 lm. The wavenumber val-
ues reported are accurate to within 2 cmꢀ1
.
2.3.6. X-ray photoelectron spectroscopy
WHSV ¼ ðgDBF=hÞ=gcatalyst
ð1Þ
X-ray photoelectron spectra (XPS) was measured using a photo-
electron spectrometer (ESCALAB250) with a monochromatic Al Ka
X-ray source (1486.6 eV, operating at 15 kV and 20 mA). The bind-
ing energy was preliminarily calibrated using the peak positions of
the adventitious carbon (C 1s, 284.6 eV). The composition of Mo
oxidation states was estimated by the deconvolution of Mo 3d dou-
blet. Relative element concentrations were determined by the inte-
gral areas of the core-level spectra. After background subtracted by
the Shirley method, the spectra were fitted into several peaks using
a convolution of Gaussian and Lorentzian functions.
The feed and the reaction products were analyzed off-line by an
Agilent gas chromatograph 7890A equipped with a flame ioniza-
tion detector (FID) and a 0.5
column. The products were determined by GC–MS with an Agilent
l
m ꢃ 0.32 mm ꢃ 30 m FFAP capillary
7890B with 5977A MSD and a 0.25
capillary column.
lm ꢃ 0.25 mm ꢃ 30 m HP-5
The conversion (X) of DBF and the selectivity (Si) of products
were defined as follows:
n0 ꢀ nDBF
X ¼
Si ¼
ꢃ 100%
ð2Þ
ð3Þ
n0
2.3.7. Scanning electron microscope
Scanning electron microscope (SEM) images of the samples
were obtained on a FEI Nova Nano SEM 450 electron microscope
with a voltage in the range of 0.5–30 kV. The results were dis-
cussed in supporting information (Figs. S1 and S2).
ni
ni
P
ꢃ 100%
where n0 and nDBF define the moles of DBF in the feed and product,
P
2.3.8. Transmission electron microscopy
respectively, ni denotes the mole of product i and ni are the total
moles of products.
The carbon balance of products was within 5% with no gas
generated.
Transmission electron microscopy (TEM) was performed on a
Philips CM 120 instrument operated at 120 keV. Samples were dis-
persed ultrasonically in ethanol and deposited on a copper grid and
dried at room temperature.
moles of carbon of all products
moles of carbon of reactant fed
l
mol∙gꢀca1t∙sꢀ1) was calculated at low conver-
Carbon balanceð%Þ ¼
ð4Þ
2.3.9. Temperature programmed reduction
Temperature programmed reduction (H2-TPR) measure was
carried out on AutoChem II 2920 Chemisorption Analyzer with a
thermal conductivity detector (TCD). 50 mg of fresh sample was
loaded in an isothermal zone of a quartz U-tube reactor and heated
from 30 °C to 300 °C with a ramp rate of 10 °C minꢀ1 and main-
tained for 2 h in 30 mL minꢀ1 Ar gas, which facilitates desorption
of the physically adsorbed water. After the reactor was cooled to
room temperature, Ar was switched to 10% H2 in Ar at a total flow
rate of 30 mL minꢀ1 and the temperature was linearly raised to
The reaction rate (
sion based on the formula.
FwDBF
M
X
wDBFX
The reaction rate ¼
¼
DBFW MDBFs
ð5Þ
where WDBF denotes the weight fraction of DBF in the reactant and
MDBF denotes the molecular mass of DBF.

314
J. Zhang et al. / Journal of Catalysis 383 (2020) 311–321
3. Results and discussion
3.1. Determination of the active sites on pure MoO3 catalyst.
Considering the complexity of valence states of Mo, MoO3 at
different reduction temperature which were exposed as varying
Mo species, were selected as the catalyst to better determine the
nature of the active sites (Fig. 1). On an equivalent Mo mass basis,
the conversions over MoO3 at different reduction temperature
exhibited a considerable discrepancy, in which MoO3-300 catalyst
reached the best activity. At the reduction temperature of 300 °C,
the conversion increased from 75% to 83% at 0.13 gDBF/gcatalyst∙h,
followed by a reduction with further increase in reduction temper-
ature of MoO3. However, the selectivity toward deoxygenated pro-
duct (BP) remained at 100% over all MoO3 catalysts even with
lower WHSV (not shown), which was inconsistent with our previ-
ous results over supported noble metals [38,39]. This indicated
that the hydrogenolysis of DBF does not produce the o-
phenylphenol (OPP) by just cleaving one C-O bond and thus Mo
species could facilitate the direct removal of oxygen atom in DBF.
To determine the nature of the active sites, in-situ XRD and XPS
were used. In-situ XRD patterns of MoO3 catalysts reduced at dif-
ferent temperature (300, 400, 500 and 600 °C) are shown in
Fig. 2. First of all, it can be seen that synthetic MoO3 presented
the characteristic peaks of the MoO3 crystalline phase (PDF#35-
0609), which suggested that pure MoO3 was obtained and the
structure was consistent with the orthorhombic molybdenum tri-
oxide. At the reduction temperature of 300 °C, no obvious disparity
was found in reduced MoO3 compared to unreduced MoO3. When
the temperature was increased to 400 °C, the new peaks at 26° and
37° appeared, which indicated that a little MoO2 has been formed
based on the comparison of the observed XRD peak intensities
from both MoO3 and MoO2 phases. The peak intensity of MoO3
decreased obviously as temperature was increased. Further
increasing the temperature to 500 °C, the major phase had been
changed to MoO2 (PDF#32-0671). When the temperature was
increased to 600 °C, almost all species were transformed into metal
Mo0.
Fig. 2. In-situ XRD patterns of MoO3 catalysts reduced at different temperature for
2 h.
tion of the XPS spectra, the contribution of the various Mo species
such as Mo6+, Mo5+, Mo4+ has been calculated (Table 1). Mo5+ spe-
cies was observed in a mild reduction of MoO3 at 300 °C (Fig. 3a).
With further reduction of MoO3 at 400 °C, in addition to the peak of
Mo5+ surface species, peaks corresponding to Mo4+ also appeared
(Fig. 3b). Previous DFT calculations suggested that a portion of
the surface oxygen of MoO3 can be removed to generate oxygen
vacancy sites [26] at 400 °C, even though the partial reduction to
form uncoordinated metal sites was undetectable by XRD.
XRD did not specifically distinguish the change of the surface
Mo species of MoO3 at the reduction temperature of 300 °C and
400 °C. The chemical state of the Mo species in catalyst surface
played an important role in the oxygen-removal reaction. There-
fore, XPS spectra was used to investigate the nature of the surface
Mo species [12]. Prior to any deconvolution of the XPS data, it is
already evident that all catalyst samples contain a significant frac-
tion of Mo6+ species in the form of MoO3 (Fig. 3). After deconvolu-
The specific surface area of calcined MoO3 (17 m2 gꢀ1) is very
low as shown in Table 2, and then thus the change of surface area
of MoO3 at different reduction temperature should not result in
such a considerable difference in activity, thereby confirming that
the hydrogenolysis activity arose from the Mo species, as charac-
terized by XRD and XPS. In comparison with pure MoO3, as the pro-
portion of Mo5+ slightly increased, there existed an increasing
trend in catalytic activity. When reduction temperature was
increased, the presence of Mo4+ species reduced the catalytic per-
formance. Further increasing the temperature up to 500 °C led to
the formation of the major phase MoO2 and thus a sharp decrease
in the activity occurred, suggesting that MoO2 is an inactive phase
for the reaction which was in agreement with previous reports
[2,24,31]. Additionally, as reduction further proceeded, the gener-
ated Mo0 showed lower activity than MoO2. Hence, the creation
of strong Mo active sites for the removal of oxygen were highly
dependent on the reduction temperature by H2 pre-treatment
and thus influenced the activity of the catalyst. These observed
results suggested that uncoordinated Mo sites had a significant
impact on the conversion of dibenzofuran and the extent of the
effect follows: Mo5+ > Mo6+ ꢄ Mo4+ > Mo0.
The results demonstrated that surface oxygen vacancies derived
from Mo5+ play a central role in the HDO of dibenzofuran. In the
next step, we selected 300 °C as the reduction temperature of
the catalyst and focused on understanding the role of Co(Ni) in
Fig. 1. Hydrogenolysis conversion of DBF as a function of WHSV (gDBF/gcatalyst∙h)
using MoO3 catalysts reduced at 300, 400, 500 and 600 °C for 2 h.

J. Zhang et al. / Journal of Catalysis 383 (2020) 311–321
315
Fig. 3. XPS spectra of the Mo (3d) energy region in MoO3 reduced at 300 °C (a) and 400 °C (b). The ratios displayed correspond to the proportion of oxidation states of Mo6+
,
Mo5+ and Mo4+
.
Table 1
The proportion of different Mo species from the XPS.
Sample
The proportion of different Mo species (%)
Mo6+
Mo5+
Mo4+
MoO3-300
MoO3-400
82
79
42
58
18
15
56
23
0
6
2
19
a
Co/MoO3
b
Ni/MoO3
a,b
Catalysts reduced at 300 °C for 2 h.
enhancing the hydrogenolysis rate of MoO3 during the hydrogenol-
ysis reaction at atmospheric H2 pressure.
3.2. Interaction of Co(Ni) and MoO3
Fig. 4. H2-TPR for the calcined MoO3 and Co(Ni)/MoO3 catalysts.
The Co(Ni)/MoO3 catalysts were extensively characterized to
explore the interaction between Co(Ni) and MoO3. Temperature-
programmed reduction with hydrogen (H2-TPR) was used to assess
the interaction between the metal and the support (Fig. 4). The TPR
profile of pure MoO3 showed two major peaks at 639 and 699 °C
corresponding to the reduction of MoO3 to MoO2, and one minor
reduction peak at 756 °C corresponding to conversion of MoO3 to
Mo4O11[2,40]. The further reduction of MoO2 to Mo occurred at
about 1000 °C (not shown) [40]. These reduction process were con-
sistent with previous study using in-situ XAFS and XRD [41]. For Co
(Ni)/MoO3 catalysts, the reduction peaks in the range of 400–
699 °C were observed, which can be associated with the reduction
process of octahedrally coordinated Mo6+ presented in MoO3 and
Co(Ni)MoO4 to Mo4+ [21,28,42,43], in which the peak at 464 °C
for the Ni/MoO3 belonged to the reduction of nickel oxide
[44,45]. The peak at 355 °C for the Co/MoO3 was attributed to
the reduction of Co3O4 to CoO [28,46,47]. And the peak of further
reduction of CoO to Co may have overlapped with that correspond-
ing to the reduction of MoO3 at 450–500 °C [47]. Higher tempera-
ture (~800 °C) was attributed to further reduction of MoO2 species
formed in the first reduction process at 644 °C [42]. Compared to
MoO3 sample spectra, the H2-TPR of Co(Ni)/MoO3 catalysts proved
that the addition of Co(Ni) appeared to increase the reducibility of
the MoO3 species. The main peak at 699 °C on MoO3 corresponding
to the reduction of MoO3 to MoO2 shifted to lower temperatures
(633 and 644 °C) with the respective addition of Ni and Co. Besides,
Ni had a stronger promotion on the reducibility of the MoO3 spe-
cies than Co. These observations were consistent with previous
report, showing that the reduction of metal oxides was facilitated
by the presence of the other adjacent metal considering the
reducibility of the catalysts [48]. It should be noted that the con-
centration of hydrogen in H2-TPR is far less than that in prior
pre-treatment reaction and thus H2-TPR was just used to compare
the reducibility of catalysts.
In-situ XRD was used to further assess the effect of Co(Ni) on the
reducibility of MoO3 in Fig. 5. In contrast to the XRD results with
the standard PDF card of MoO3, it was obvious that all the catalysts
showed a typical MoO3 pattern. XRD did not identify the presence
of oxidized Co(Ni) on the calcined catalysts resulting from the cor-
responding impregnation of MoO3 with the cobalt or nickel salt, an
observation that could be explained by the fact that only a small
Table 2
The catalytic performance of MoO3 and Co(Ni)/MoO3 catalysts at WHSV of 0.13 gDBF/gcatalyst∙h.
Samples
MoO3
Co/MoO3 11
Ni/MoO3 11
Sa (m2 gꢀ1
)
Initial reaction rate(
l
mol∙gꢀca1t∙sꢀ1
)
Conversion (%) BP selectivity (%) The amount of acidityb (mmol gꢀ1
)
Mo5+c (%) Ea (kJ molꢀ1
)
17
0.18
0.29
0.26
83
97
91
100
100
86
0.004
0.048
0.056
18
56
23
152
131
140
a
Determined using calcined samples.
Determined using reduced samples at 300 °C for 2 h.
b,c

316
J. Zhang et al. / Journal of Catalysis 383 (2020) 311–321
Fig. 5. In-situ XRD patterns of MoO3 and Co(Ni)/MoO3 catalysts at (a) oxidation states and (b) reduction states (reduction at 300 °C for 2 h).
amount of modifier was added. However, the new peaks of Co/
MoO3 (2h = 25.2°) and Ni/MoO3 (2h = 14.3, 28.5°) were detected
at oxide states compared with MoO3, suggesting cobalt molybdate
and nickel molybdate were formed [28,49]. And the cobalt/nickel
molybdate still existed under the present reduction conditions.
Furthermore, the peak intensity of MoO3 decreased obviously on
MoO3 modified by Co(Ni), and the peaks of MoO2 appeared, which
demonstrated that the reduction of MoO3 is facilitated by the pres-
ence of the Co(Ni) [48] and this was consistent with H2-TPR results.
Moreover, Ni-modified MoO3 was more prompt to be reduced than
that modified with Co due to the fact that the peak intensity of
MoO2 in Ni/MoO3 was much stronger than Co/MoO3 (Fig. 5b), as
also evidenced by H2-TPR.
XPS was carried out to identify the chemical states of Mo spe-
cies (Fig. 6). The surface fraction of Mo5+ predominantly increased
along with the addition of Co(Ni) compared with MoO3 catalyst,
indicating the addition of Co (Ni) could accelerate the reduction
process from Mo6+ to Mo5+. However, Ni also promotes the reduc-
tion of Mo5+ to Mo4+ as well as Mo6+ to Mo5+ due to the stronger
promotion to MoO3 reduction as proved by XRD and H2-TPR. It
seems that Co prevented over-reduction to lower oxidation states
compared to Ni species.
by Shanks and coworkers, in which the fresh MoO3 was found to
contain some acidity, whereas by reducing the catalyst at 350 °C
for one hour, the acidity greatly diminished as compared with
the fresh catalyst [25]. It can be concluded that the addition of
Co(Ni) could significantly increase the acidity of catalyst, and the
explanation for this may be related to the formation of Co(Ni)
MoO4 considering the lesser acidity of metal Co(Ni), as evidenced
by NH3-TPD of CoMoO4 (Fig. S4).
To explore the acidic sites, Raman spectroscopy was used to
determinate the formation of Co(Ni)MoO4 (Fig. 7b). The band at
998 cmꢀ1 can be assigned to the terminal (Mo-O) stretching vibra-
tions [21,42,43,49,50]. The band at 820 cmꢀ1 can be attributed to
the antisymmetric (Mo-O-Mo) stretching modes [21,42]. The band
centered at 673 cmꢀ1 is due to an asymmetric stretching of the
Mo-O-Mo [42]. The band at 292 cmꢀ1 belongs to a doublet com-
prised of wagging modes of the terminal oxygen connected to
Mo atoms [42]. The Raman spectra of Co/MoO3 sample not only
possessed the characteristic peaks of MoO3, but also presented
new bands at about 692 and 938 cmꢀ1, which stood for the main
bands of CoMoO4 [48,51]. In addition, the Raman spectra of Ni/
MoO3 sample exhibited additional band of NiMoO4 at about
963 cmꢀ1 [49]. Raman spectroscopic data of the catalysts showed
a superficial interaction between Co(Ni) and Mo in catalysts cal-
cined at 400 °C, probably with the formation of Co(Ni)-O-Mo
phases presented in the surface.
The preceding characterization had been shown that the Co(Ni)
can promote the reduction of MoO3. Further, temperature pro-
grammed desorption with ammonia (NH3-TPD) was performed to
determine the total amounts of acid centers as well as their acid
strength of the three catalysts (Fig. 7a). The corresponding values
Apart from the characterization of Raman, the Co(Ni)MoO4 spe-
cies was also proved by the in-situ FT-IR (Fig. 7c). Three types of
connections between Mo and O atoms were presented [21,49,52].
The broad and complex band centered at about 625 cmꢀ1 is due
(l
mol gꢀ1) were listed in Table 2. The reduced MoO3 was found
to have few acidity which was consistent with the results reported
Fig. 6. X-ray photoelectron spectra (XPS) of the Mo (3d) energy region in Co/MoO3 (a) and Ni/MoO3 (b) catalysts reduced at 300 °C for 2 h. The ratios displayed correspond to
the proportion of oxidation states of Mo6+, Mo5+ and Mo4+
.

J. Zhang et al. / Journal of Catalysis 383 (2020) 311–321
317
The combined characterization of NH3-TPD, Raman and in-situ
FT-IR confirmed that Co(Ni)MoO4 were the acidic sites of catalysts.
To further visually demonstrate the presence of acidic site, TEM
was used to probe the Co/MoO3 catalyst. The obtained Co/MoO3
was nanosheet (Fig. 8a). HRTEM further showed existence of
CoMoO4 with the interplanar distance of 0.33 nm, which corre-
sponded to the (0 0 2) plane (Fig. 8b).
On the basis of the above results, the addition of Co(Ni) pro-
motes the reduction of MoO3 and the generation of acidic Co(Ni)
MoO4 species, thus increases the active sites of catalysts.
3.3. The role of Co(Ni) on the hydrogenolysis of dibenzofuran
The catalysts were tested for the hydrogenolysis activity and
the results were shown in Fig. 9. In general, strong promotional
effect was observed, and Co/MoO3 was observed to have a stronger
promotional effect than Ni/MoO3, which was consistent with the
previous report[53]. The initial reaction rate was increased from
0.18
0.26
l
l
mol∙gꢀca1t∙sꢀ1 (MoO3) to 0.29
l
mol∙gꢀca1t∙sꢀ1 (Co/MoO3) and
mol∙gꢀca1t∙sꢀ1 (Ni/MoO3). The promotional effects observed
can be explained by an increase of the number of active sites. BP
was the major product with the selectivity of 95% over Ni/MoO3
and 100% over MoO3 and Co/MoO3. As the WHSV decreased, the
BP selectivity was gradually decreased over Ni/MoO3 accompanied
by an increased selectivity to cyclohexylbenzene (Fig. S5), whereas
the BP selectivity over Co/MoO3 still maintained at 100%, due to the
higher hydrogenation activity of Ni than Co [16,54]. The apparent
activation energy (Ea) of catalysts was calculated (Fig. S6) and
the results were shown in Table 2. The addition of Co(Ni) could
lead to a decrease in the activation barrier of hydrogenolysis of
DBF and further result in an increase in the reaction activity. These
results showed that MoO3 promoted by Co(Ni) could catalyze the
oxygen-removal reactions with high efficiency.
3.4. Determination of the active sites on Co(Ni)/MoO3
To determine the active sites for the Co(Ni)/MoO3 catalysts, the
studies on reaction rates as a function of acidic sites, active Mo5+
species, could provide an in-depth understanding of active centers
and their individual contribution toward the reaction. The activity
of Co(Ni)/MoO3 catalysts reduced at different reduction tempera-
ture was explored in order to expose the varying quantity of acidic
sites and Mo5+ species (Table S2, Figs. S7–S9). In this study, the
increase of reaction rates and acidity was not a linear relationship
over Co(Ni)/MoO3 catalysts (Figs. S10a and S11a), illustrating that
acidic sites serve as the active site but not an exclusive factor to
govern catalytic performance. Furthermore, it should be men-
tioned that the Mo species in Co(Ni)/MoO3 also strongly affect
the reaction rates of catalysts (Figs. S10b and S11b). These results
suggest that the hydrogenolysis activity of DBF is not only associ-
ated with the acidic sites, but also related to the Mo species of
catalysts.
Fig. 7. (a) NH3-TPD, (b) Raman spectra and (c) in-situ FT-IR spectra of MoO3 and Co
(Ni)/MoO3 catalysts reduced at 300 °C for 2 h.
to the stretching vibration of the OA atoms each linked to three Mo
atoms. The two absorption patterns at 820 cmꢀ1 and 894 cmꢀ1
could be attributed to the vibration of the OB atoms in Mo-O-Mo
entities. A band at 992 cmꢀ1 was attributed to the vibration of
the terminal oxygen in Mo-O. The bands corresponding to OA
stretching modes shifted to lower wavenumber with the addition
of Co(Ni), as evidenced by the shifting of the band at 625 cmꢀ1
for MoO3 to 594 cmꢀ1 for Co/MoO3 and Ni/MoO3. Furthermore,
the band at 894 cmꢀ1 corresponding to OB stretching modes also
shifted to lower wavenumber. The position of terminal Mo-O and
the band at 820 cmꢀ1, however, was not significantly affected. Note
that there was a strong absorption band at 945 cmꢀ1 over Co/
MoO3, assigned to CoMoO4 species, which was typical of the octa-
hedral configuration of Mo [52]. These results elucidated that Co
(Ni) could interact with Mo-OA and Mo-OB and then the new spe-
cies Co(Ni)MoO4 were formed, which was also evidenced by XRD
(Fig. 4) and Raman spectroscopy (Fig. 7b).
3.4.1. Role of the Mo5+ species on Co(Ni)/MoO3
The experimental results from varying Mo species by adjusting
the reduction temperature of MoO3 have already confirmed that
the presence of these different Mo species demonstrated a signif-
icant effect on the catalytic performance and Mo5+ species and
this could be responsible for the improved activity, as determined
by the relationship between the reaction results and the charac-
terization results (XRD and XPS). With the addition of Co(Ni) to
MoO3, the reaction rate distinctly increased with the increase in
Mo5+ species (Table 2), indicating that Mo5+ species of Co(Ni)/
MoO3 played an important role for the reaction and Co(Ni) facil-
itated the reduction process from Mo6+ to Mo5+. Further, the
experimental results from exposing different amounts of Mo5+

318
J. Zhang et al. / Journal of Catalysis 383 (2020) 311–321
Fig. 8. TEM images (a and b) of Co/MoO3 reduced at 300 °C for 2 h.
3.4.2. Role of the acidic sites on Co(Ni)/MoO3
The addition of Co(Ni) increased the acidity of the catalysts in
comparison to MoO3 with no acidity (Fig. 7a), along with the for-
mation of Co(Ni)MoO4 species, as characterized by the Raman
(Fig. 7b) and FT-IR (Fig. 7c) spectroscopies. The NH3-TPD profiles
of synthetic CoMoO4 further confirmed that the acidic sites were
derived from the formation of CoMoO4 species (Fig. S4), and previ-
ous report [58] also verified the acidity of Co(Ni)MoO4. Moreover,
the Co(Ni) species of Co(Ni)MoO4 were more difficult to be reduced
than those of Co(Ni)/MoO3 (Fig. S3), which was in agreement with
previous report[58], thereby it may indicate the possibility of the
existence of Co(Ni)MoO4 in our pre-treatment condition, which
are in accord with the preceding results.
The synthetic Co(Ni)MoO4 samples were used as the catalysts in
an attempt to investigate the influence of acidic sites on the reac-
tion activity. The Co(Ni)MoO4 showed a similar conversion as the
MoO3 (~25%), both of which were lower than the conversion of
Co(Ni)/MoO3 (~40%) (Table S3). It indicated that the high activities
of Co(Ni)/MoO3 could result from the synergistic effect between
Mo species and the acidic sites rather than a single effect. In the
Co(Ni)MoO4, the Mo atoms were more difficult to be reduced than
the Co atoms, as evidenced by TPR (Fig. S3), hence unsaturated Mo
was not the main contributor to the activity of the corresponding
catalysts. In order to exclude the effect of Mo species, the reaction
rate per unit mole of Mo was calculated (Table S3). The reaction
Fig. 9. BP selectivity and conversion as a function of WHSV (gDBF/gcatalyst∙h) during
the hydrogenolysis of DBF over MoO3 and Co(Ni)/MoO3 reduced at 300 °C for 2 h.
species by changing the reduction temperature of Co(Ni)/MoO3
indicated that the Mo5+ species is the one active site for the reac-
tion (Fig. S10b).
Regarding the reducibility of the catalysts, the existence of a
synergetic effect between Co(Ni) and Mo should be noted since
the reduction of metal oxides was facilitated by the presence of
the other adjacent metal. XRD and XPS showed that the addition
of Co(Ni), even in very small amount, could increase the
reducibility of MoO3 species and thus increase the amount of
Mo5+. The prevalence of Mo5+ species, indicative of the creation
of active defects, arose from the reduction of MoO3 to MoO3ꢀx
[27]. Previous report proposed that the promoter (Ni and Co)
can donate electrons to molybdenum and lead to a weakening
of the metal–sulfur bond of MoS2, resulting in an increase in
the number of vacancy sites [16]. Tran et al. also reported that
cobalt not only facilitated the reduction and dispersion of nickel
oxide but also enhanced the HDO activity of bimetallic Ni-Co cat-
alyst [54]. Similarly, the Pd could facilitate reduction of Fe oxide
based on H2-TPR studies and DFT calculations [55]. Indeed, inter-
action of Co(Ni) and Mo also promotes the reduction of molybde-
num oxides [56]. It was shown that an oxygen atom between two
molybdenum atoms is less strongly bonded than an oxygen atom
bonded between a nickel (or cobalt) and a molybdenum atom and
can be more easily removed (Fig. S3), thus creating a vacancy,
which was different from the promotional effect of Co(Ni) on
MoS2, albeit both enhance reaction rate by creating a larger num-
ber of vacancy [57].
rates of CoMoO4 and NiMoO4 (1.17 and 1.07
respectively) were far more than MoO3 (0.66
l
l
mol∙molꢀM1o∙sꢀ1 102,
mol∙molꢀM1o∙sꢀ1 102),
which showed that acidic sites would be of great importance in
the reaction and the activity was significantly increased regardless
of the activity of Mo species. The experimental results from changing
the amounts of acidic sites by increasing the reduction temperature
of Co(Ni)/MoO3 further indicated that the acidic sites are the active
sites for the reaction (Figs. S10a and S11a). Subsequently, it can be
concluded that the high activities of Co(Ni)/MoO3 catalysts were
the results of the synergistic effect between Mo species and acidic
Co(Ni)MoO4 species.
3.5. Reaction mechanism
First of all, the reaction mechanism of pure MoO3 catalyzing the
cleavage of C-O bond should be proposed. The reduced MoO3 has
no acidity, as evidenced by NH3-TPD (Fig. 7a), which elucidated
that the cleavage of C-O bond was not going through acid-base cat-
alytic mechanism. On the other hand, as described in Section 3.1,
the results demonstrated that unsaturated Mo5+ species were the
major active species for MoO3. Therefore, surface oxygen vacancy

J. Zhang et al. / Journal of Catalysis 383 (2020) 311–321
319
derived from unsaturated Mo5+ played a pivotal role in the HDO of
dibenzofuran. It was well accepted that oxygen-removal reaction
involves an oxygen vacancy mechanism over MoO3 catalysts
[4,25,26]. This reaction mechanism shown in Scheme 1 had the fol-
lowing elementary reaction steps (1), (2) and (3): (1) in the initial
induction period, MoO3 was reduced by H2 as a reductant to form
adequate number of Mo5+ sites for reactions; (2) oxygen atoms of
dibenzofuran molecules interacted with Mo5+ sites to form Mo-O
bond, followed by cleavage of the C-O bond; (3) sequential desorp-
tion of the formative BP occurred. Eventually, H2 reduced the
already-oxidized catalyst surface to expose the active sites again
[26].
Next, the reaction mechanism of Co(Ni)/MoO3 was discussed. In
Co(Ni)/MoO3 catalysts, the metallic Co(Ni) species were observed,
as evidenced by the XPS (Fig. S12) and TPR (Fig. 4), which may
be associated with the activation of hydrogen and enrichment of
the surface hydrogen populations. The presence of Co2+/Ni2+ in
XPS also confirmed the formation of Co(Ni)MoO4 species and it
can be difficult to be reduced in current conditions (Fig. S12).
Metallic Co(Ni) species were capable for H2 activation and
increased availability of reactive hydrogen surface to aid in the
removal of oxygen [59]. However, Bartholomew stated that the
kinetics and energetics of hydrogen adsorption on cobalt and
nickel were significantly changed by metal-support interactions,
particularly in catalysts of low metal content or involving highly
reducible supports [60]. Prins proposed that hydrogen spillover
occurred mainly in the reducible oxides because the defect sites
can stabilize atomic hydrogen [61], hence the reducible MoO3
may provide the possibility for hydrogen spillover. The activated
hydrogen species on Co(Ni) can spill over and migrate onto the
MoO3 surface. The reduction of Mo6+ to Mo5+ species promoted
by the addition of Co(Ni) may be enhanced by means of hydrogen
spillover that can be improved by the increased acidity [8,62],
which corresponds well with the increase of acidity of catalysts.
However, H species by spillover do not travel much further than
the immediate interface between the metal particle and the
metal-oxide support [61]. Therefore, the oxygen atoms at the
boundary region of metal particle and oxide support, can be more
easily removed by an adsorbed H species, leaving behind an oxy-
gen vacancy [63]. It also indicated that the metal-support inter-
faces provide a chance for the hydrogen spillover and may be the
area where most of the H atoms react with DBF to generate BP
on the interface.
The addition of Co(Ni) could increase the number of Mo5+ sites
and promote the formation of Co(Ni)MoO4 species, both of which
were the active sites for the reaction. It should be noted that the
oxophilic Mo species could increase the interaction of the surface
with oxygen-containing functional groups [64]. Resasco et al. also
demonstrated that the metal oxophilicity makes the direct cleav-
age of the C-O bond in the surface easier [65]. Additionally, it
was known that the acidic sites of catalyst could accelerate the
adsorption of oxygen-containing compounds [8,66]. During the
deoxygenation of benzyl alcohol, the Mo species could cause the
aromatic ring to repel from the surface, resulting in alteration of
the adsorption orientation as well as a reduction of the adsorption
strength, as evidenced by the DFT calculation and experiment
studies [67]. These indicated that the DBF molecules were favor-
able to be adsorbed on the catalyst surface by the interaction
between the oxygen atom of DBF and two types of active sites in
non-coplanar manner, which may be one reason why smaller
hydrogenated product was observed. As previously indicated, the
HDO of DBF over Co(Ni)/MoO3 occurred both through the oxygen
vacancies and the acidic sites. Moreover, the HDO reaction may
also involve the enhanced hydrogen spillover by acidic sites of
Co(Ni)/MoO3 catalysts and the hydrogen by spillover could pro-
mote the cleavage of C-O bond adsorbed on Mo5+ species and
acidic sites. The supposed schematic of the C-O hydrogenolysis
mechanism of DBF over Co/MoO3 catalyst was illustrated with
Scheme 2. The dissociated hydrogen on the metallic Co(Ni) in
metal-support interface was provided by the spillover for the
unsaturated Mo5+ sites and acidic sites where the DBF molecules
can adsorb and activate. The reasons for the much higher activity
of Co(Ni)/MoO3 than the MoO3 catalyst included the increased
hydrogen species promoted by Co(Ni), the higher rate of reduction
Scheme 1. Supposed schematic of the C-O hydrogenolysis mechanism of DBF over Co/MoO3 catalyst.

320
J. Zhang et al. / Journal of Catalysis 383 (2020) 311–321
Scheme 2. Supposed schematic of the C-O hydrogenolysis mechanism of DBF over Co/MoO3 catalyst.
from Mo6+ to Mo5+ species assisted by Co(Ni) and finally, the effec-
tive formation of acidic sites. The results thus may act as a guid-
ance for future designing of highly active and selective
deoxygenation catalyst for conversion of lignin.
[3] J.N. Chheda, G.W. Huber, J.A. Dumesic, Angew. Chem. Int. Ed. 46 (2007) 7164–
7183.
[4] V.O.O. Gonçalves, C. Ciotonea, S. Arrii-Clacens, N. Guignard, C. Roudaut, J.
Rousseau, J.-M. Clacens, S. Royer, F. Richard, Appl. Catal. B 214 (2017) 57–66.
[5] A.M. Robinson, J.E. Hensley, J.W. Medlin, ACS Catal. 6 (2016) 5026–5043.
[6] D.P. Serrano, J.A. Melero, G. Morales, J. Iglesias, P. Pizarro, Catal. Rev. 60 (2017)
1–70.
4. Conclusion
[7] A.N.K. Lup, F. Abnisa, W.M.A.W. Daud, M.K. Aroua, Appl. Catal. A 541 (2017)
87–106.
[8] J. Zhang, C. Li, X. Chen, Y. Chen, L. Zhang, B. Zhang, C. Liang, J. Catal. 371 (2019)
346–356.
[9] K. Cui, L. Yang, Z. Ma, F. Yan, K. Wu, Y. Sang, H. Chen, Y. Li, Appl. Catal. B 219
(2017) 592–602.
[10] L.P. Xiao, S. Wang, H. Li, Z. Li, Z.-J. Shi, L. Xiao, R.C. Sun, Y. Fang, G. Song, ACS
Catal. 7 (2017) 7535–7542.
[11] A.L. Jongerius, R. Jastrzebski, P.C.A. Bruijnincx, B.M. Weckhuysen, J. Catal. 285
(2012) 315–323.
[12] G.D.L. Puente, J.J.P.A. Gil, P. Grange, Langmuir 15 (1999) 5800–5806.
[13] V.N. Bui, D. Laurenti, P. Afanasiev, C. Geantet, Appl. Catal. B 101 (2011) 239–245.
[14] A.E. Coumans, E.J.M. Hensen, Appl. Catal. B 201 (2017) 290–301.
[15] A. Popov, E. Kondratieva, L. Mariey, J.M. Goupil, J. El Fallah, J.-P. Gilson, A.
Travert, F. Maugé, J. Catal. 297 (2013) 176–186.
[16] Y. Romero, F. Richard, S. Brunet, Appl. Catal. B 98 (2010) 213–223.
[17] M. Grilc, B. Likozar, Chem. Eng. J. 330 (2017) 383–397.
[18] Z. Strassberger, A.H. Alberts, M.J. Louwerse, S. Tanase, G. Rothenberg, Green
Chem. 15 (2013) 768.
[19] F. Yang, D. Liu, H. Wang, X. Liu, J. Han, Q. Ge, X. Zhu, J. Catal. 349 (2017) 84–97.
[20] G.W. Huber, S. Iborra, A. Corma, Chem. Rev. 106 (2006) 4044–4098.
[21] M. Shetty, K. Murugappan, W.H. Green, Y. Román-Leshkov, A.C.S. Sustain,
Chem. Eng. 5 (2017) 5293–5301.
[22] K.A. Goulas, A.V. Mironenko, G.R. Jenness, T. Mazal, D.G. Vlachos, Nat. Catal. 2
(2019) 269–276.
[23] S. Boullosa-Eiras, R. Lødeng, H. Bergem, M. Stöcker, L. Hannevold, E.A. Blekkan,
Catal. Today 223 (2014) 44–53.
[24] P.M. Mortensen, H.W.P. de Carvalho, J.-D. Grunwaldt, P.A. Jensen, A.D. Jensen, J.
Catal. 328 (2015) 208–215.
[25] M.W. Nolte, J. Zhang, B.H. Shanks, Green Chem. 18 (2016) 134–138.
[26] T. Prasomsri, T. Nimmanwudipong, Y. Román-Leshkov, Energ, Environ. Sci. 6
(2013) 1732.
[27] C. Ranga, R. Lødeng, V.I. Alexiadis, T. Rajkhowa, H. Bjørkan, S. Chytil, I.H.
Svenum, J. Walmsley, C. Detavernier, H. Poelman, P. Van Der Voort, J.W.
Thybaut, Chem. Eng. J. 335 (2018) 120–132.
The results from varying the nature of Mo species by adjusting
the reduction temperature of MoO3 and the characterization of
these species by XPS and in-situ XRD strongly supported that Mo
species had a significant impact on the conversion of dibenzofuran
and the extent of the effect follows: Mo5+ > Mo6+ ꢄ Mo4+ > Mo0. A
strong promotional effect in activity between Co(Ni) and Mo in Co
(Ni)/MoO3 catalysts was observed, which was resulted from speci-
fic Co(Ni)-Mo interaction in which Co(Ni) facilitated the reduction
from Mo6+ to Mo5+ species and the formation of acidic Co(Ni)MoO4
species, as evidenced by in-situ XRD, XPS, H2-TPR, NH3-TPD,
Raman, in-situ FT-IR and TEM. Ni/MoO3 showed a lower activity
than Co/MoO3, which was related to the further reduction of
Mo5+ to Mo4+ species, and the observation of aromatic hydrogena-
tion ring on Ni/MoO3, both of which may be due to the higher abil-
ity to dissociate hydrogen molecules on Ni. Except for Ni/MoO3, all
catalysts were capable of exhibiting remarkable selectivity for C-O
bond cleavage without hydrogenation of aromatic rings, giving
striking biphenyl selectivity of 100% at relatively moderate condi-
tions. The best catalytic activity was realized over Co/MoO3 with a
100% yield of BP, which might be related to the interaction of Co
and Mo species as reflected by the decreased activation energy. A
possible mechanism, including Co(Ni) facilitated reduction of
Mo6+ to Mo5+ and Co(Ni) enhanced formation of acidic sites, was
proposed to be responsible for the high activity and BP selectivity
in Co(Ni)/MoO3 catalysts.
[28] C. Ranga, V.I. Alexiadis, J. Lauwaert, R. Lødeng, J.W. Thybaut, Appl. Catal. A 571
(2019) 61–70.
[29] T.M. Sankaranarayanan, A. Berenguer, C. Ochoa-Hernández, I. Moreno, P. Jana,
J.M. Coronado, D.P. Serrano, P. Pizarro, Catal. Today 243 (2015) 163–172.
[30] T. Prasomsri, M. Shetty, K. Murugappan, Y. Román-Leshkov, Energ. Environ. Sci.
7 (2014) 2660–2669.
Declaration of Competing Interest
The authors declare that they have no known competing finan-
cial interests or personal relationships that could have appeared
to influence the work reported in this paper.
[31] K. Murugappan, E.M. Anderson, D. Teschner, T.E. Jones, K. Skorupska, Y.
Román-Leshkov, Nat. Catal. 1 (2018) 960–967.
[32] C. Bouchy, C. Pham-Huu, B. Heinrich, C. Chaumont, M.J. Ledoux, J. Catal. 190
(2000) 92–103.
Acknowledgements
[33] F. Hoxha, B. Schimmoeller, Z. Cakl, A. Urakawa, T. Mallat, S.E. Pratsinis, A.
Baiker, J. Catal. 271 (2010) 115–124.
[34] J.A. Cecilia, A. Infantes-Molina, E. Rodríguez-Castellón, A. Jiménez-López, S.T.
Oyama, Appl. Catal. B 136–137 (2013) 140–149.
[35] Y. Wang, Y. Fang, T. He, H. Hu, J. Wu, Catal. Commun. 12 (2011) 1201–1205.
[36] H.W. Lee, B.R. Jun, H. Kim, D.H. Kim, J.-K. Jeon, S.H. Park, C.H. Ko, T.W. Kim, Y.K.
Park, Energy 81 (2015) 33–40.
This work was financially supported by the National Key
Research & Development Program of China (2016YFB0600305).
The authors thank Instrumental Analysis & Research Center for
their analysis assistance.
[37] D. Ballesteros-Plata, A. Infantes-Molina, M. Rodríguez-Cuadrado, E. Rodríguez-
Aguado, P. Braos-García, E. Rodríguez-Castellón, Appl. Catal. A 547 (2017) 86–
95.
Appendix A. Supplementary material
[38] J. Zhang, C. Li, X. Chen, W. Guan, C. Liang, Catal. Today 319 (2019) 155–163.
[39] J. Zhang, L. Wang, C. Li, S. Jin, C. Liang, Org. Process Res. Dev. 22 (2018) 67–76.
[40] K.R.R. Thallada Bhaskar, Chinthala Praveen Kumar, Mamidanna R.V.S. Murthy,
Komandur V.R. Chary, Appl. Catal. A 211 (2001) 189–201.
[41] R.E.J.T. Ressler, J. Wienold, M.M. Gunter, O. Timpe, J. Phys. Chem. B 104 (2000)
6360–6370.
Supplementary data to this article can be found online at
https://doi.org/10.1016/j.jcat.2020.01.035.
References
[42] Y. Cui, N. Liu, Y. Xia, J. Lv, S. Zheng, N. Xue, L. Peng, X. Guo, W. Ding, J. Mol.
Catal. A 394 (2014) 1–9.
[43] T. Hahn, U. Bentrup, M. Armbrüster, E.V. Kondratenko, D. Linke, ChemCatChem
6 (2014) 1664–1672.
[1] J. He, C. Zhao, D. Mei, J.A. Lercher, J. Catal. 309 (2014) 280–290.
[2] M. Shetty, K. Murugappan, T. Prasomsri, W.H. Green, Y. Román-Leshkov, J.
Catal. 331 (2015) 86–97.

J. Zhang et al. / Journal of Catalysis 383 (2020) 311–321
321
[44] S. Jin, Z. Xiao, C. Li, X. Chen, L. Wang, J. Xing, W. Li, C. Liang, Catal. Today 234
(2014) 125–132.
[56] R.G. Kukushkin, O.A. Bulavchenko, V.V. Kaichev, V.A. Yakovlev, Appl. Catal. B
163 (2015) 531–538.
[45] E. Heracleous, A. Lee, K. Wilson, A. Lemonidou, J. Catal. 231 (2005) 159–171.
[46] H. Song, L. Zhang, U.S. Ozkan, Green Chem. 9 (2007) 686–694.
[47] R. Nava, B. Pawelec, P. Castaño, M.C. Álvarez-Galván, C.V. Loricera, J.L.G. Fierro,
Appl. Catal. B 92 (2009) 154–167.
[48] C.P. Cheng, G.L. Schrader, J. Catal. 60 (1979) 276–294.
[49] L.M. Madeira, M.F. Portela, C. Mazzocchia, Catal. Rev. 46 (2004) 53–110.
[50] J. Jiao, J. Fu, Y. Wei, Z. Zhao, A. Duan, C. Xu, J. Li, H. Song, P. Zheng, X. Wang, Y.
Yang, Y. Liu, J. Catal. 356 (2017) 269–282.
[51] X. Xia, W. Lei, Q. Hao, W. Wang, X. Wang, Electrochim. Acta 99 (2013) 253–
261.
[52] C.A.C. Mazzocchia, C. Diagne, E. Tempesti, J.M. Herrmann, G. Thomas, Catal.
Lett. 10 (1991) 181–192.
[57] M. Egorova, R. Prins, J. Catal. 225 (2004) 417–427.
[58] N.F.Y.-S. Yoon, W. Ueda, Y. Moro-oka, K.-W. Lee, Catal. Today 24 (1995) 327–
333.
[59] M. Pozzo, D. Alfè, Int. J. Hydrogen Energ. 34 (2009) 1922–1930.
[60] C.H. Bartholomew, Catal. Lett. 7 (1990) 27–52.
[61] R. Prins, Chem. Rev. 112 (2012) 2714–2738.
[62] Y. Liu, J. Zhao, Y. He, J. Feng, T. Wu, D. Li, J. Catal. 348 (2017) 135–145.
[63] P.S. Antonio Ruiz Puigdollers, Sergio Tosoni, Gianfranco Pacchioni, ACS Catal. 7
(2017) 6493–6513.
[64] A. Robinson, G.A. Ferguson, J.R. Gallagher, S. Cheah, G.T. Beckham, J.A. Schaidle,
J.E. Hensley, J.W. Medlin, ACS Catal. 6 (2016) 4356–4368.
[65] Q. Tan, G. Wang, A. Long, A. Dinse, C. Buda, J. Shabaker, D.E. Resasco, J. Catal.
347 (2017) 102–115.
ˇ
[53] T. Klicpera, M. Zdrazil, J. Catal. 206 (2002) 314–320.
[54] N.T.T. Tran, Y. Uemura, S. Chowdhury, A. Ramli, Appl. Catal. A 512 (2016) 93–
100.
[55] Y. Hong, H. Zhang, J. Sun, K.M. Ayman, A.J.R. Hensley, M. Gu, M.H. Engelhard, J.
S. McEwen, Y. Wang, ACS Catal. 4 (2014) 3335–3345.
[66] A. Popov, E. Kondratieva, J.M. Goupil, L. Mariey, P. Bazin, J.-P. Gilson, A. Travert,
F. Mauge, J. Phys. Chem. C 114 (2010) 15661–15670.
[67] A.M. Robinson, L. Mark, M.J. Rasmussen, J.E. Hensley, J.W. Medlin, J. Phys.
Chem. C 120 (2016) 26824–26833.