Enhanced Hydroconversion of Lignin-Derived Oxygen-Containing Compounds Over Bulk Nickel Catalysts Though Nb2O5 Modification

Article abstract of DOI:10.1007/s10562-017-2085-6

A series of bimetallic Nb–Ni oxide catalysts with different Nb/Ni molar ratio have been prepared by chemical precipitation method. XRD, Raman and XPS results indicate that amorphous Nb₂O₅ species exist in the samples with a Nb/Ni ratio about 0.087. The as-synthesized bimetallic Nb–Ni oxides effectively promote the dispersion of NiO active components, as a result effectively inhibit the agglomeration of NiO particles. Ni_0.92Nb_0.08O sample with the largest surface area of 173?m²/g mainly consists of fold-like nanosheets and the amorphous Nb₂O₅ species are well-dispersed all over the bulk NiO. After the reduction in hydrogen, the Nb-promoted bulk nickel catalysts display better catalytic performance for hydrodeoxygenation of lignin-derived anisole to biofuels than bulk Ni catalyst. The selectivity to deoxygenated products with using Ni_0.92Nb_0.08 catalyst increases 2.5 fold to that with bulk Ni catalyst at 160 °C and 3?MPa H₂, as a result of the synergistic effect between amorphous Nb₂O₅ species and metal Ni active sites. In addition, with further increase in the reaction temperature to 200 °C, deoxygenation almost goes quantitatively. Graphical Abstract: High-specific-surface-area Nb–Ni oxides are prepared by using chemical precipitation, and display excellent HDO performance for lignin-derived compounds. Selectivity to deoxygenated products increases 2.5 folds over Ni_0.92Nb_0.08 than over bulk Ni catalyst. [Figure not available: see fulltext.].

Full text of DOI:10.1007/s10562-017-2085-6

Catal Lett
DOI 10.1007/s10562-017-2085-6
Enhanced Hydroconversion of Lignin-Derived Oxygen-
Containing Compounds Over Bulk Nickel Catalysts Though
Nb O Modification
2
5
1
1
2
2
Shaohua Jin · Weixiang Guan · Chi-Wing Tsang · Dickson Y. S. Yan ·
2
1
Cho-Yin Chan · Changhai Liang
Received: 9 February 2017 / Accepted: 15 May 2017
Springer Science+Business Media New York 2017
©
Abstract A series of bimetallic Nb–Ni oxide catalysts
with different Nb/Ni molar ratio have been prepared by
chemical precipitation method. XRD, Raman and XPS
results indicate that amorphous Nb O species exist in the
reaction temperature to 200°C, deoxygenation almost goes
quantitatively.
Graphical Abstract High-specific-surface-area Nb–Ni
oxides are prepared by using chemical precipitation, and
display excellent HDO performance for lignin-derived
compounds. Selectivity to deoxygenated products increases
2
5
samples with a Nb/Ni ratio about 0.087. The as-synthesized
bimetallic Nb–Ni oxides effectively promote the dispersion
of NiO active components, as a result effectively inhibit
the agglomeration of NiO particles. Ni Nb O sample
2
.5 folds over Ni Nb than over bulk Ni catalyst.
0.92 0.08
0
.92
0.08
2
with the largest surface area of 173 m /g mainly consists
of fold-like nanosheets and the amorphous Nb O species
2
5
are well-dispersed all over the bulk NiO. After the reduc-
tion in hydrogen, the Nb-promoted bulk nickel catalysts
display better catalytic performance for hydrodeoxygena-
tion of lignin-derived anisole to biofuels than bulk Ni cata-
lyst. The selectivity to deoxygenated products with using
Ni Nb catalyst increases 2.5 fold to that with bulk Ni
0
.92
0.08
catalyst at 160°C and 3 MPa H , as a result of the syner-
2
gistic effect between amorphous Nb O species and metal
2
5
Ni active sites. In addition, with further increase in the
Electronic supplementary material The online version of this
article (doi:10.1007/s10562-017-2085-6) contains supplementary
material, which is available to authorized users.
*
Chi-Wing Tsang
ctsang@vtc.edu.hk
Keywords Nb O · Bulk nickel catalysts, synergistic
2
5
*
Changhai Liang
effect · Hydroconversion · Anisole
changhai@dlut.edu.cn
1
2
Laboratory of Advanced Materials and Catalytic
Engineering, School of Chemical Engineering, Dalian
University of Technology, Dalian 116023, China
1
Introduction
Faculty of Science and Technology, Technological
and Higher Education Institute of Hong Kong, Hong Kong,
China
Exploiting the abundant and renewable carbon resources
in nature is one of the best options to reduce the society’s
Vol.:(0
1
123456789)
3

S. Jin et al.
reliance on limited fossil fuel reserves [1–5]. To do this,
an efficient catalytic process for biomass transformation
is needed. The bio-fuels produced from the transforma-
tion process have higher energy density than the raw
biomass, making them more suitable for industrial appli-
cations [6]. As the main constituent of lignocellulosic
biomass, lignin which occupies 15–30% by weight and
2 Experimental
2.1 Catalyst Preparation
Bulk Ni–Nb–O catalysts were prepared by chemical pre-
cipitation method using aqueous ammonium and sodium
hydroxide mixture. Niobium (V) oxalate hydrate was added
to an aqueous solution of nickel nitrate in an appropriate
proportion where Nb/(Ni+Nb) molar ratios equal to 0.03,
0.08, 0.15 and 0.20 respectively. This solution was main-
tained under stirring at 70°C for 1 h to ensure good mix-
ing of the starting compounds. The temperature was then
rapidly increased to 80°C and an aqueous ammonium
(0.5 M) and sodium hydroxide (0.1 M) mixture was added
dropwise to keep the mixture at pH 9.0. Then, the resulting
suspension was kept at 120°C overnight to ensure complete
precipitation of remaining nickel nitrate. Subsequently,
the obtained solids were washed repeatedly with deion-
4
0% by energy in the plants can be depolymerized to phe-
nolic and other oxygen-containing aromatic compounds
7, 8]. Due to the presence of the large oxygen contents,
[
crude biofuels have poor chemical stability and low heat-
ing value, thus limiting their direct application as alterna-
tives to fossil fuels [9–11]. Our goal is to devise a highly
efficient catalytic system for removal of oxygen of lignin-
derived compounds, whilst conservation of the represent-
ative aromatic rings, hydroxyl groups, keto groups, and
ether linkages etc.
Recently, a series of highly efficient Ni-based catalysts
for one-pot hydroupgrading lignin-derived model com-
pounds have been developed [5, 11–15]. Lignin-derived
model substrates were explored and a comprehensive
description of the reactivity of bulk Raney Ni toward the
hydrogenation and hydrogenolysis of these substrates were
described [5]. In particular, the C–O bonds of α-O-4, β-O-
+
ized water to remove residual Na . Finally, the solids were
dried at 110°C and calcined in synthetic air at 450°C for
5 h. The Ni Nb O catalysts with different atomic compo-
1
−x
x
sitions of x=0.03, 0.08, 0.15 and 0.20 and pure NiO have
been synthesized. Besides, the pure Nb O was obtained
2
5
from niobic acid (Nb O ·nH O, CBMM) in synthetic air at
2
5
2
4
and 4-O-5 etc. can be cleaved by direct hydrogenolysis
450°C for 5 h.
over Ni-based catalysts and the produced oxygen-contain-
ing monomers can further be upgraded to cyclic alcohols
by complete saturation of aromatic rings. Generally, the
efficient removal of oxygen cannot be easily achieved over
acid-free catalysts and without relatively high temperature
2.2 Catalyst Characterization
BET surface areas, pore volumes and pore size distribu-
tions of the catalysts (approximately 0.1 g sample) were
[
16–20]. Brønsted and Lewis acid sites are necessary to
determined by nitrogen physisorption at liquid N tempera-
2
efficiently remove the oxygen for converting phenolic and
oxygen-containing aromatics to hydrocarbons if a milder
reaction temperature is desirable [11, 21–25].
ture with an autosorb iQ automated gas sorption analyzer.
The phase structure of the samples was determined by
powder X-ray diffraction (XRD) analysis in a D/MAX-2400
diffractometer using a Cu Kα radiation (λ=1.5418 Å),
operated at 40 kV and 100 mA.
Niobium oxide and related materials, which consist both
Brønsted and Lewis acid sites, have been proven to be an
efficient and hydrothermally stable catalyst for biomass uti-
lization [26–30]. Among these niobium-containing materi-
als, Ni–Nb–O bulk oxide catalysts have been widely used
for ethane oxidative dehydrogenation and the high effi-
ciency of the catalysts was attributed to the surface inter-
action between the NiO active phase and an amorphous
Nb O phase [31–34]. Meanwhile, the introduction of nio-
Temperature-programmed reduction of hydrogen
(H -TPR) was performed in a stream of 10% H in Ar with
2
2
3
a flow rate of 50 cm /min. The samples were heated up to a
final temperature of 800°C at 10°C/min and H consump-
2
tion was monitored by a thermal conductivity detector.
The morphology and structural composition of the bulk
Ni–Nb–O catalyst was characterized and analyzed by tung-
sten filament scanning electron microscopy (SEM, Nova
NanoSEM 450 from FEI Co.), equipped with an energy-
dispersive X-ray (EDX) analyzer. The particle size and
dispersion of the samples were analyzed by transmission
electron microscopy (TEM). Powder samples were ultra-
sonicated in ethanol and dispersed on holey carbon films on
copper grids.
2
5
bium promotes the dispersion of NiO nanoparticles and the
obtained catalysts own higher surface area than bulk NiO.
Nevertheless, the use of hydrogen reduced Ni–Nb–O cata-
lyst towards the hydroconversion of bio-fuels has rarely
been reported. Herein, we report the preparation, charac-
terization and catalytic performance of a Ni–Nb–O mixed
oxides systems with various Nb/Ni ratios. The aim of this
study is to explore the potential of these reduced Ni-Nb-O
catalysts for one-pot hydroconversion of phenolic and oxy-
gen-containing aromatic compounds (Fig. S1).
Raman spectra were obtained by an Invia Reflex Laser
Micro-Raman spectrometer (Renishaw, England) with a
+
532 nm line of Ar laser.
1
3

Enhanced Hydroconversion of Lignin-Derived Oxygen-Containing Compounds Over Bulk Nickel…
Surface compositions were investigated by X-ray
photoelectron spectroscopy (XPS) by employing an
ESCALAB250 (Thermo VG, USA) spectrometer with Al
K (1486.6 eV) radiation with a power of 150 W. All core-
α
level spectra were referenced as the C 1s neutral carbon
peak at 284.6 eV.
2
.3 Catalytic Hydroconversion of Lignin Model
Compounds
Hydroconversion tests were conducted in a 50 mL stain-
less steel autoclave equipped with a magnetic stirrer and
an electric temperature controller. Prior to reaction, the
obtained Ni–Nb–O catalysts were reduced by H at 400°C
2
for 2 h and then passivated in Ar overnight. 8 wt% sub-
strate dispersed into 20 mL n-decane with 2 wt% n-dode-
cane as an internal standard for quantitative GC analysis
and 0.1 g reduced catalysts were rapidly introduced into
the autoclave to prevent from contacting with air for long.
Subsequently, the autoclave was sealed, purged repeatedly
Fig. 1 XRD patterns of precipitate formed by mixing nickel nitrate
and niobium (V) oxalate hydrate in water at 70°C and after calcina-
tion at 700°C
with H to eliminate air, and kept an atmosphere pressure.
2
After heated to the desired temperature at 700 rpm stirring
speed, the autoclave was pressurized to 3.0 MPa and the
reaction zero time point was counted. At completion, the
reactor was cooled to room temperature and products were
analyzed by gas chromatography (Echrom A90, FID, OV-1
capillary column 30 m×0.32 mm×0.25 μm) and identi-
fied by an Agilent 6890N GC (HP-5 MS capillary column,
pattern of the formed precipitate and it is well matched
with that of hydrated nickel oxalate (JCPDS 25-0581).
Afterwards, the precipitate was calcined at 700°C in air
for 2 h and interestingly the peaks assignable to crystal-
line NiNb O phase at 2θ of 26.8°, 35.2° and 52.9° (JCPDS
2
6
76-2355) were observed.
2
+
5+
3
0 m×0.25 mm×0.25 μm) with 5973 MSD. Due to the
To completely precipitate residual Ni and Nb cati-
ons, a mixture of aqueous ammonium and sodium hydrox-
ide was subsequently added dropwise to the former sus-
pension at a pH of 9.0. The obtained solids were washed
repeatedly with deionized water to remove the residual
formation of low carbon compounds, there is carbon loss
but is less than 5%. The conversion of lignin model com-
pound is defined as the moles of C reacted divided by total
moles of C in the reactant, multiplied by 100%. The selec-
tivity is defined as the moles of C in the product of interest
divided by the total moles of C in all liquid products, mul-
tiplied by 100%.
+
Na . Finally, the solids were calcined in synthetic air at
450°C for 5 h. The corresponding XRD patterns are shown
in Fig. 2. All the Ni–Nb–O oxide samples show the charac-
teristic diffraction peaks of crystalline NiO bunsenite struc-
ture (JCPDS 89-7130). The main diffraction peaks of NiO
are located at 2θ of 37.1°, 43.2°, 62.5°, 74.8° and 78.7°,
which corresponds to the (111), (200), (220), (311) and
(222) reflections, respectively. As the Nb loading increases,
a broad background peak with very low intensity appears
at around 26°, which could be due to the emergence of
amorphous niobium oxides by the chemical precipitation
3
Results and Discussion
3
.1 Catalyst Characterization
Ni–Nb–O mixed oxide catalysts are usually prepared
by an evaporation method and the samples obtained by
5
+
this method generally own a relatively small surface area
processing of Nb and hydroxyl ion [30]. After calcina-
tion at 700°C, Ni Nb O sample shows the peaks attrib-
2
(
<85 m /g) [31, 34–37]. So our work used the precipita-
0
.8
0.2
tion method instead of the evaporation method for prepara-
tion of Ni–Nb–O mixed oxides and we only changed the
Nb/Ni ratio without varying any other parameters. First, a
certain amount of nickel nitrate was dissolved in water and
then the niobium (V) oxalate hydrate was added slowly. A
precipitate was gradually formed at room temperature and
the XRD results were analyzed. Figure 1 shows the XRD
uted to Ni–Nb–O mixed phase [32], indicating that amor-
phous but not crystalline Ni–Nb composite phase may
exist in the Ni–Nb–O samples after calcination at 450°C.
It was reported that high Nb content leads to the formation
of Ni–Nb–O mixed phase (such as NiNb O , Ni Nb O ,
2
6
4
2
9
Ni Nb O , etc.) and the formation of Ni–Nb–O mixed
3
2
8
phase could be detrimental to the catalytic efficiency [34].
1
3

S. Jin et al.
the pure NiO are presented in Fig. 3a–e. Pure nickel oxide
consists of well-defined sheet-like microcrystallites in the
nano size range with small void spaces contribute to the
surface area and pore volume. Doping NiO with niobium
resulted in significant changes in the crystalline morphol-
ogy. For the Ni Nb O and Ni Nb O samples, a
0
.97
0.03
0.92
0.08
sponge-like and fold-like appearance exist on the sheet-
like NiO surface and in the small void spaces [34]. The
sponge-like species can be identified as a Ni–Nb solid
solution formed by the incorporation of a small amount
of Nb in the NiO lattice structure due to the similar ionic
2
+
5+
radius of Ni and Nb of 0.69 and 0.64 Å [31, 38]. For
the Ni Nb O sample, the sheet-like appearance is
0
.85
0.15
hardly observed totally but rather sponge-like and block-
like appearance. Further increasing the Nb content led to
the formation of round crystallites with fewer small void
spaces (Ni Nb O sample).
Fig. 2 XRD patterns of Ni–Nb–O mixed oxide catalysts after calci-
0
.8
0.2
nation at 450°C in air for 5 h
In order to confirm the distribution of niobium oxide in
bulk NiO and to determine the surface Nb/Ni ratio, X-ray
maps of Ni and Nb element for Ni Nb O sample
The surface morphology of the synthesized materials
was examined by SEM. Micrographs of the Ni–Nb–O
mixed oxides of all compositions, together with that of
0
.92
0.08
Fig. 4 X-ray maps of Ni, Nb and EDX results of Ni0.92Nb0.08O sam-
Fig. 3 SEM images of NiO and Ni–Nb–O mixed oxides
ple
1
3

Enhanced Hydroconversion of Lignin-Derived Oxygen-Containing Compounds Over Bulk Nickel…
after calcination at 450 °C are recorded, as shown in
Fig. 4. It can be seen clearly that niobium oxide are well-
dispersed all over the bulk NiO by the chemical precipita-
tion method. In addition, the EDX result suggests that the
surface Nb/Ni ratio of the as-prepared Ni Nb O sam-
0
.92
0.08
ple is 0.11/0.92, which is slightly higher than the bulk
theoretical value of 0.08/0.92, indicating that the surface
is niobium-enriched.
To accommodate a sufficient number of active sites for
catalytic reaction, suitable porosity is required especially
for the bulk catalysts. Surface areas, pore size and pore
volume of the as-prepared Ni–Nb–O samples are listed in
Table 1. The bulk NiO obtained from chemical precipita-
2
tion method owns a surface area of 136 m /g. The incor-
poration of Nb into the precipitation synthesis with a Nb/
Ni ratio blow 0.087 contributes to a rapid growth of the
surface area of the resulting Ni–Nb oxide materials, which
can be attributed to the use of niobium oxalate as the pre-
cursor of Nb. The generated nickel oxalate hydrated from
XRD result in Fig. 1. A more porous structure will be
formed by the decomposition (during calcination) of nickel
oxalate hydrated with organic nature [34]. In addition, the
formation of the amorphous Nb O species with low Nb
2
5
content is beneficial to the dispersion of NiO particles [31,
3
4, 35]. With further increasing in Nb/Ni ratio to 0.25, a
significant reduction in surface area is recorded, due to the
formation of block-like large crystallites (see SEM results).
In addition, the nickel crystallite size of the Ni–Nb–O
mixed oxides calculated by Scherrer formula is also listed
in Table 1. We can see that the nickel crystallite size has a
direct relation to the corresponding surface area.
Fig. 5 XPS spectra of the Ni 2p3/2 region and Nb 3d for NiO,
Ni Nb and Ni Nb samples
0
.92
0.08
0.80
0.20
XPS results of the selected samples as shown in Fig. 5
are used to investigate the surface chemical valence state
and Nb/Ni ratio. The Nb/Ni ratio of Ni Nb O and
structures (such as NiNb O , Ni Nb O and Ni Nb O ,
2 6 4 2 9 3 2 8
etc.). Besides, the binding energy values obtained for nio-
5
+
bium species well match with that of Nb [32, 40], sug-
gesting that niobium is fully oxidized into its maximum
valence of +5.
0
.92
0.08
Ni Nb O samples by XPS is 0.11/0.92 and 0.34/0.80
0
.80
0.20
respectively, indicating that the surface is niobium-enriched
[
32]. The main Ni2p3/2 peaks are located at around 853.8
Figure 6 shows the Raman spectra of the Ni–Nb–O
mixed oxides and reference materials NiO and Nb O
2
+
3+
and 855.8 eV, which correspond to Ni and Ni , respec-
2
5
2
+
3+
tively [39]. An evolution of the Ni and Ni near-sur-
face populations indicates a gradual consumption by nio-
bium, probably associated with the formation of Nb–Ni–O
collected at ambient conditions. All the Ni-rich samples
−
1
display a strong band at around 500 cm due to the NiO
phase. The reference amorphous Nb O oxide displays
2
5
Table 1 Physicochemical
2
3
Catalysts
Nb/Ni ratio
SBET (m /g)
d (nm)
Vtotal (cm /g)
Crystallite
performances of NiO,
size (nm)a
Nb O ·nH O and Ni–Nb–O
2
5
2
mixed oxides
NiO
0
136
158
173
139
110
122
18.0
16.5
9.6
12.5
12.0
6.7
0.61
0.65
0.41
0.43
0.33
0.20
9.3
8.0
5.4
11.8
14.5
–
Ni0.97Nb0.03
Ni0.92Nb0.08
Ni0.85Nb0.15
Ni0.80Nb0.20
O
O
O
O
0.031
0.087
0.176
0.250
–
Nb O ·nH O
2
5
2
aDetermined considering the NiO (200) peak higher intensity
1
3

S. Jin et al.
Fig. 7 Temperature-programmed reduction profiles of the Ni–Nb–O
Fig. 6 Raman spectra of the Ni–Nb–O mixed oxides and reference
catalysts and reference material NiO
materials NiO and Nb O
2
5
the bands near 248, 659 and a broad shoulder band in the
−1
8
00–1000 cm . These bands can be attributed to the bend-
hence the oxygen can be removed much easier at lower
temperatures near 340 °C than in pure NiO, owing to the
decrease of nickel electronic density in the presence of
niobium [34]. As a result, the low-temperature reduction
peak becomes increasingly apparent with niobium addi-
tion. Besides, Nb O has been reported to be irreducible
ing modes of the Nb–O–Nb linkages, slightly and highly
distorted octahedral NbO species, which can give weak
6
Brönsted acidity and strong Lewis acidity, respectively [30,
3
4]. For Ni Nb O sample after calcination at 700°C,
0.80 0.20
−
1
two additional bands detected near 784 and 866 cm can
be assigned to vibrations of bridging Nb–O–Ni bonds [32].
2
5
under the conditions used in the TPR measurements [32].
−1
The Raman band near 784 cm can also be detected at
higher Nb content (Ni Nb O and Ni Nb O) after
TEM images of Ni Nb O and Ni Nb
O
0.20
0
.92
0.08
0.80
samples obtained after calcination at 450 °C in air and
0
.85
0.15
0.80
0.20
calcination at 450°C, indicating the existence of Ni–Nb–O
phase. Nb may exist as amorphous niobium oxide for
Ni Nb O and Ni Nb O samples without any
reduced Ni Nb O sample in H with the correspond-
0
.92
0.08
2
ing high-resolution TEM are displayed in Fig. 8. As
expected, Ni Nb O sample owns a fold-like sheet
0
.97
0.03
0.92
0.08
0.92
0.08
−
1
Raman bands observed in the range of 780–870 cm .
Temperature-programmed reduction curves of Ni-
rich materials are shown in Fig. 7. The TPR profile of
bulk nickel oxide exhibits a reduction peak at around
structure and Ni Nb O sample exhibits a block-like
0.80 0.20
structure consisting of small units with the spherical
shape in accordance with SEM results (Fig. 8a, b). After
the reduction in hydrogen, the fold-like sheet structure
disappears. Instead, large amounts of well-defined metal-
lic nickel nanoparticles (8–10 nm) with amorphous struc-
ture likely due to the noncrystalline niobium oxides for
Ni Nb O sample are observed. The HRTEM image
2
+
0
4
30 °C, which can be attributed to the Ni →Ni reduc-
tion step. The promotion of Nb caused the appearance
of two reduction peaks. The high-temperature reduc-
tion peak near 520 °C can be caused by the formation of
smaller crystalline nickel particles from XRD and SEM
results. Due to the formation of block-like large crys-
tallites with an increase in niobium content, the high-
temperature reduction peak gets smaller. On the other
hand, Nb incorporation weakens the Ni–O–Ni bonds, and
0
.92
0.08
in Fig. 8d presents the lattice spacing of the metallic
nickel as 1.99 Å, calculated on the basis of two-dimen-
sional fast Fourier transform (FFT), which is in accord-
0
ance with the lattice spacing value of Ni (111) (JCPDS
88-2326).
1
3

Enhanced Hydroconversion of Lignin-Derived Oxygen-Containing Compounds Over Bulk Nickel…
Fig. 8 TEM and HRTEM
images of a calcined
Ni Nb O, b calcined
0.92
0.08
Ni Nb O and c, d reduced
0.80
0.20
Ni Nb O samples
0.92
0.08
Table 2 Hydroconversion of anisole over bulk Nb–Ni catalysts
and 3 MPa H . The HDO activities using bulk Nb–Ni cata-
2
lysts with different Nb/Ni molar ratio are summarized in
Table 2.
The main products detected by GC and GC–MS are
methoxycyclohexane, cyclohexanol and cyclohexane, sug-
gesting that the HDO of anisole under the current condi-
tions follows the HYD route [24]. Anisole can almost be
fully hydrogenated to methoxycyclohexane over the bulk
Ni catalyst after 2 h. However, the selectivity of oxygen-
free cyclohexane obtained by Ni-catalyzed hydrogenoly-
sis is only 12.9%, probably due to the poor acidity of bulk
Ni [41]. On the other hand, the conversion of anisole over
the niobium-doped Ni catalysts decrease slightly caused
by the reduced content of nickel for hydrogenation, which
is directly related to the surface area of the correspond-
ing samples. Nevertheless, the deoxygenation activity of
Nb–Ni catalysts obviously increases, due to the presence of
amorphous Nb O species having the acid sites for deoxy-
Catalyst
bulk Ni
Ni0.97Nb0.03
Ni0.92Nb0.08
Ni0.85Nb0.15
Conversion/%
Selectivity/%
1
2
3
99.3
93.9
95.3
88.7
84.2
–
12.9
22.1
31.8
18.9
17.9
–
2.1
8.3
10.1
2.3
5.8
–
85.0
69.6
58.1
78.8
76.3
–
Ni0.80Nb0.20
Nb O ·H O
2
5
2
Reaction conditions: 0.1 g catalyst, 8 wt% anisole, 20 mL C ,
1
0
1
60°C, 3 MPa, 2 h, 700 rpm
2
5
3
.2 Catalytic Hydroconversion of Lignin Model
Compounds
genation reaction. Ni Nb catalyst exhibits the best
0
.92
0.08
deoxygenation activity and the selectivity of oxygen-free
cyclohexane increases 2.5 folds compared with that over
In consideration of the methoxyl being the most abundant
group in lignin structure, we conducted the hydrodeoxy-
genation (HDO) test for anisole in an autoclave at 160°C
bulk Ni catalyst at 160°C and 3 MPa H . When the Nb/
2
Ni molar ratio exceeds 0.087, the deoxygenation activity
decreases owing to the formation of the inactive NiNb O
2
6
1
3

S. Jin et al.
Fig. 9 Hydroconversion of
phenol (a) and anisole (b) over
Ni Nb catalyst. Condi-
0.92
0.08
tions: 0.1 g Ni Nb catalyst,
0
.92
0.08
8
3
wt% anisole, 20 mL C ,
10
MPa, 700 rpm
used to investigate the effect of temperature on the product
distribution and effectiveness of the overall process. The
corresponding results are shown in Fig. 10. Anisole was
almost completely converted at 160°C within 2 h due to the
fast saturation of aromatic-rings, however, the selectivity
for oxygen-free cyclohexane is only 30%. The deoxygena-
tion activity of the Ni Nb catalyst obviously increases
0.08
0
.92
with increasing reaction temperature and the selectivity of
deoxygenated product at 180°C rises 2.5 folds of that at
1
2
60°C. With further increasing the reaction temperature to
00°C, the degree of deoxygenation reaches 100%. In addi-
tion, a small amount of cyclohexanol (10% selectivity) was
formed at 160°C and then fast dehydrated to cyclohexane
with increasing temperature to 170°C. Accordingly, higher
reaction temperature favors HDO activity for Nb-doped
bulk Ni catalysts and the only shorter reaction time is
required for the conversion of lignin-derived phenolic and
oxygen-containing biofuels to oxygen-free alkanes.
Fig. 10 Effect of temperature on HDO of anisole over Ni Nb
0.08
0
.92
catalyst. Conditions: 0.1 g Ni Nb catalyst, 8 wt% anisole, 20 mL
0
.92
0.08
C , 3 MPa, 2 h, 700 rpm
10
Hydrodeoxygenations of several typical lignin-derived
monomers and dimers over Ni Nb catalyst for the
0.08
0
.92
species. Besides, the blank niobic acid as a reference was
also used for HDO of anisole and however, it was inactive.
production of oxygen-free alkanes were then conducted
and the corresponding results are listed in Table 3. β-O-4
dimer was synthesized according to published literature
[42] and the characterization results are shown in Fig. S2.
From the results, it was seen that the more numbers of
hydroxyl and methoxyl the less degree of deoxygenation
Then, we choose the Ni Nb catalyst having the
0.08
0
.92
optimal Nb/Ni molar ratio for HDO of phenol and anisole
under mild conditions and the products distribution with
the reaction time are shown in Fig. 9. Both phenol and ani-
sole can be hydrogenated to fully saturated cyclohexanol
and methoxycyclohexane respectively within 2 h and while
the subsequent deoxygenation process to form cyclohexane
needs a longer reaction time. This phenomenon indicates
that the Ni-catalyzed hydrogenation for aromatic-rings
saturation is the fastest step, followed by the dehydroxyla-
tion or demethoxylation process [21]. Compared with phe-
nol, the conversion of anisole through hydrodeoxygenation
needs higher reaction temperature of 160°C and longer
reaction time as 12 h, suggesting that anisole is a less active
oxygen-containing substrate than phenol [7].
is for lignin-derived monomers over Ni Nb catalyst
0.08
0
.92
and the sequence of difficulty is: vanillin>guaiacol>ani-
sole>phenol. This phenomenon should be attributed to
the steric hindrance effect of reactant molecules and com-
petitive adsorption on the surface of Ni Nb cata-
0.08
0
.92
lyst [43–45]. For the HDO of vanillin, more toluene are
obtained. This is due to the rather higher steric hindrance
caused by three oxygen-containing functional groups and
competitive adsorption between them and aromatic ring
units, inhibiting the complete hydrogenation of the aro-
matic rings on metal active sites. Furthermore, all three
lignin-derived dimers can be completely converted to liquid
alkanes at 200°C within 2 h, suggesting that Ni Nb
Ni Nb catalyst demonstrates the best deoxygena-
0.08
0
.92
tion activity in the HDO process of anisole, thus it was
0
.92
0.08
1
3

Enhanced Hydroconversion of Lignin-Derived Oxygen-Containing Compounds Over Bulk Nickel…
Table 3 HDO of typical lignin-derived monomers and dimers
Substrate
Conditions
Conversion%
>99
Selectivity %
OH
O
o
2
2
2
2
20 C, 2h
1
00
OH
O
o
40 C, 6h
>99
>99
>99
>99
8
00
8
1
19
O
O
o
00 C, 2h
1
4
4
O
o
00 C, 2h
52
54
HO
o
β-O-4
200 C, 2h
6
Reaction condition: 0.1g Ni0.92Nb0.08 catalyst, 8 wt% substrate, 20 mL C10, 3 MPa, 700 rpm.
catalyst exhibited both excellent C–O cleavage ability and
deoxygenation activity.
(2016YFB0600305), National Natural Science Foundation of China
(
Nos. 21573031 and 21373038), Program for Excellent Talents in
Dalian City (2016RD09) and Technological and Higher Education
Institute of Hong Kong (THEi SG1617105).
4
Conclusion
References
High specific surface area bulk Nb–Ni oxide catalysts
with different Nb/Ni molar ratio have been prepared using
chemical precipitation method. The amorphous Nb O spe-
1
.
.
Corma A, Iborra S, Velty A (2007) Chem Rev 107(6):2411–2502
Huber GW, Corma (2007) Angew Chem Int Ed
6(38):7184–7201
(2006) Chem Rev
2
5
2
A
cies effectively promote the dispersion of NiO active com-
ponents, consequently inhibiting the agglomeration of NiO
particles. On the contrary, high Nb/Ni molar ratio above
4
3. Huber GW, Iborra S, Corma
06(9):4044–4098
Zakzeski J, Bruijnincx PCA, Jongerius AL, Weckhuysen BM
A
1
4
.
0
.087 resulted in the formation of inactive NiNb O spe-
2 6
(
2010) Chem Rev 110(6):3552–3599
cies. Ni Nb O sample with the largest surface area
0
.92
0.08
5. Wang X, Rinaldi R (2012) Energy Environ Sci 5(8):8244
2
of 173 m /g mainly consists of fold-like nanosheets and
6. Gutierrez A, Kaila RK, Honkela ML, Slioor R, Krause AOI
(
2009) Catal Today 147(3–4):239–246
Ni Nb O sample is composed of block-like crystal-
0
.80
0.20
7
8
.
.
Deutsch KL, Shanks BH (2012) Appl Catal A 447–448:144–150
Zhang XH, Zhang Q, Long, JX, Xu Y, Wang TJ, Ma LL, Li YP
lites. The Nb-promoted bulk Ni catalysts after reduction in
hydrogen atmosphere displayed higher HDO performance
for lignin-derived oxygen-containing aromatic compounds
than bulk Ni catalyst. The selectivity to deoxygenated prod-
(
2014) Bioresources 9(2):3347–3360
9. Garcia-Perez M, Chaala A, Pakdel H, Kretschmer D, Roy C
2007) Biomass Bioenerg 31(4):222–242
0. Jin S, Xiao Z, Li C, Chen X, Wang L, Xing J, Li W, Liang C
(
1
uct cyclohexane over Ni Nb catalyst increases 2.5
0
.92
0.08
(
2014) Catal Today 234:125–132
folds of that over bulk Ni catalyst at 160°C and 3 MPa H ,
2
11. Zhao C, Kou Y, Lemonidou AA, Li X, Lercher JA (2010) Chem
Commun 46(3):412–414
which is mainly attributed to the catalytic synergistic effect
1
2. He J, Zhao C, Lercher JA (2012) J Am Chem Soc
between amorphous Nb O species and metal Ni active
2
5
1
34(51):20768–20775
sites. With further increasing in the reaction temperature
to 200°C, deoxygenation almost goes quantitatively. Fur-
thermore, lignin-derived dimers can also be completely
converted to liquid alkanes at 200°C within 2 h, which is
attributed to the excellent C–O cleavage ability and deoxy-
1
1
3. Sergeev AG, Hartwig JF (2011) Science 332(6028):439–443
4. Zhang XH, Zhang Q, Wang TJ, Ma LL, Yu YX, Chen LG (2013)
Bioresour Technol 134:73–80
1
1
5. Zhang XH, Wang TJ, Ma LL, Zhang Q, Jiang T (2013) Biore-
sour Technol 127:306–311
6. Bykova MV, Ermakov DY, Kaichev VV, Bulavchenko OA,
Saraev AA, Lebedev MY, Yakovlev VА (2012) Appl Catal B
genation activity of Ni Nb catalyst.
0
.92
0.08
1
13–114:296–307
Acknowledgements We gratefully acknowledge the finan-
cial support provided by National Key Research & Development
Program of the Ministry of Science and Technology of China
17. Lee CR, Yoon JS, Suh Y-W, Choi J-W, Ha J-M, Suh DJ, Park
Y-K (2012) Catal Commun 17:54–58
1
3

S. Jin et al.
1
1
2
2
8. Wang L, Li C, Jin S, Li W, Liang C (2014) Catal Lett
32. Rojas E, Delgado JJ, Guerrero-Pérez MO, Bañares MA (2013)
Catal Sci Technol 3(12):3173–3182
1
44(5):809–816
9. Ohta H, Kobayashi H, Hara K, Fukuoka A (2011) Chem Com-
mun 47(44):12209–12211
33. Rubio-Marcos F, Rojas E, López-Medina R, Guerrero-Pérez
MO, Bañares MA, Fernandez JF (2011) ChemCatChem
3(10):1637–1645
0. Sun J, Karim AM, Zhang H, Kovarik L, Li XS, Hensley AJ,
McEwen J-S, Wang YJ (2013) Catal 306:47–57
34. Heracleous E, Lemonidou AJ (2006) Catal 237(1):162–174
35. Heracleous E, Delimitis A, Nalbandian L, Lemonidou AA
(2007) Appl Catal A 325(2):220–226
1. Zhao C, Kou Y, Lemonidou AA, Li X, Lercher JA (2009) Angew
Chem Int Ed 48(22):3987–3990
2
2
2. Zhao C, Song W, Lercher JA (2012) ACS Catal 2(12):2714–2723
3. Yan N, Yuan Y, Dykeman R, Kou Y, Dyson PJ (2010) Angew
Chem Int Ed 49(32):5549–5553
36. Heracleous E, Lemonidou AJ (2006) Catal 237(1):175–189
37. Skoufa Z, Heracleous E, Lemonidou AA (2012) Catal Today
192(1):169–176
2
2
4. Wang H, Male J, Wang Y (2013) ACS Catal 3(5):1047–1070
5. Jin S, Chen X, Li C, Tsang C-W, Lafaye G, Liang C (2016)
ChemistrySelect 1(15):4949–4956
38. Lee S-H, Cheong HM, Park N-G, Tracy CE, Mascarenhas A,
Benson DK, Deb SK (2001) Solid State Ion 140(1):135–139
39. Salagre P, Fierro J, Medina F, Sueiras JJ (1996) Mol Catal A
106(1):125–134
2
2
2
2
6. Pham HN, Pagan-Torres YJ, Serrano-Ruiz JC, Wang D, Dumesic
JA, Datye AK (2011) Appl Catal A 397(1–2):153–162
7. Sumiya S, Oumi Y, Sadakane M, Sano T (2009) Appl Catal A
40. López-Medina R, Fierro J, Guerrero-Pérez MO, Bañares MA
(2010) Appl Catal A 375(1):55–62
3
65(2):261–267
41. Alonso DM, Wettstein SG, Dumesic JA (2012) Chem Soc Rev
41(24):8075–8098
8. West RM, Liu ZY, Peter M, Dumesic JA (2008) ChemSusChem
1
(5):417–424
42. Buendia J, Mottweiler J, Bolm
17(49):13877–13882
C (2011) Chem Eur J
9. Buitrago-Sierra R, Serrano-Ruiz JC, Rodríguez-Reinoso F,
Sepúlveda-Escribano A, Dumesic JA (2012) Green Chem
43. Philippe M, Richard F, Hudebine D, Brunet S (2010) Appl Catal
A 383(1–2):14–23
1
4(12):3318–3324
3
3
0. Jin S, Xiao Z, Chen X, Wang L, Guo J, Zhang M, Liang C
44. Sepúlveda C, Escalona N, García R, Laurenti D, Vrinat M (2012)
Catal Today 195(1):101–105
(
2015) Ind Eng Chem Res 54(8):2302–2310
1. Savova B, Loridant S, Filkova D, Millet JMM (2010) Appl Catal
A 390(1–2):148–157
45. Sun Y, Prins RJ (2009) Catal 267(2):193–201
1
3