Pd/Nb2O5/SiO2 catalyst for the direct hydrodeoxygenation of biomass-related compounds to liquid alkanes under mild conditions

Article abstract of DOI:10.1002/cssc.201500053

A simple Pd-loaded Nb₂O₅/SiO₂ catalyst was prepared for the hydrodeoxygenation of biomass-related compounds to alkanes under mild conditions. Niobium oxide dispersed in silica (Nb₂O₅/SiO₂) as the support was prepared by the sol-gel method and characterized by various techniques, including N₂ adsorption, XRD, NH₃ temperature-programmed desorption (TPD), TEM, and energy-dispersive X-ray spectroscopy (EDAX) atomic mapping. The characterization results showed that the niobium oxide species were amorphous and well dispersed in silica. Compared to commercial Nb₂O₅, Nb₂O₅/SiO₂ has significantly more active niobium oxide species exposed on the surface. Under mild conditions (170°C, 2.5 MPa), Pd/10 %Nb₂O₅/SiO₂ was effective for the hydrodeoxygenation reactions of 4-(2-furyl)-3-buten-2-one (aldol adduct of furfural with acetone), palmitic acid, tristearin, and diphenyl ether (model compounds of microalgae oils, vegetable oils, and lignin), which gave high yields (>94 %) of alkanes with little C-C bond cleavage. More importantly, owing to the significant promotion effect of NbO_x species on C-O bond cleavage and the mild reaction conditions, the C-C cleavage was considerably restrained, and the catalyst showed an excellent activity and stability for the hydrodeoxygenation of palmitic acid with almost no decrease in hexadecane yield (94-95 %) in a 150 h time-on-stream test.

Full text of DOI:10.1002/cssc.201500053

DOI: 10.1002/cssc.201500053
Full Papers
Pd/Nb2O5/SiO2 Catalyst for the Direct Hydrodeoxygenation
of Biomass-Related Compounds to Liquid Alkanes under
Mild Conditions
Yi Shao, Qineng Xia, Xiaohui Liu, Guanzhong Lu,* and Yanqin Wang*[a]
A simple Pd-loaded Nb2O5/SiO2 catalyst was prepared for the
hydrodeoxygenation of biomass-related compounds to alkanes
under mild conditions. Niobium oxide dispersed in silica
(Nb2O5/SiO2) as the support was prepared by the sol–gel
method and characterized by various techniques, including N2
adsorption, XRD, NH3 temperature-programmed desorption
(TPD), TEM, and energy-dispersive X-ray spectroscopy (EDAX)
atomic mapping. The characterization results showed that the
niobium oxide species were amorphous and well dispersed in
silica. Compared to commercial Nb2O5, Nb2O5/SiO2 has signifi-
cantly more active niobium oxide species exposed on the sur-
face. Under mild conditions (1708C, 2.5 MPa), Pd/10%Nb2O5/
SiO2 was effective for the hydrodeoxygenation reactions of 4-
(2-furyl)-3-buten-2-one (aldol adduct of furfural with acetone),
palmitic acid, tristearin, and diphenyl ether (model compounds
of microalgae oils, vegetable oils, and lignin), which gave high
yields (>94%) of alkanes with little CÀC bond cleavage. More
importantly, owing to the significant promotion effect of NbOx
species on CÀO bond cleavage and the mild reaction condi-
tions, the CÀC cleavage was considerably restrained, and the
catalyst showed an excellent activity and stability for the hy-
drodeoxygenation of palmitic acid with almost no decrease in
hexadecane yield (94–95%) in a 150 h time-on-stream test.
Introduction
With the diminishment of fossil-fuel resources and the growth
of environmental concerns, the conversion of biomass into
liquid transportation fuels has attracted a lot of attention. As
comprehensively described by Dumesic et al.,[1] Lercher et al.,[2]
and Bitter et al.,[3] lignocellulose, inedible vegetable oils, and
microalgae oils are the three general feedstocks for the pro-
duction of “second-generation” biofuels. The conversion of
biomass to hydrocarbon fuels requires the removal of oxygen
to improve energy density.[1]
In the works of the groups of Dumesic[4] and Huber,[5] C8–C15
oxygenates were produced by the aldol condensation of ace-
tone with furfural (and/or 5-hydroxymethylfurfural) derived
from lignocellulose. The C8–C15 alkanes were obtained by sub-
sequent hydrogenation to remove the unsaturated C=C and
C=O bonds over Pd/Al2O3 (4–5.5 MPa H2) and hydrodeoxygena-
tion on a Pt-loaded solid acid catalyst (250–2658C, 5.2–6.0 MPa
H2). Very recently, Sutton et al.[6] developed a new process for
the hydrodeoxygenation of bioderived furans into alkanes
under mild conditions using a Pd catalyst under acidic condi-
tions. Firstly, the C=C bond outside furan ring was saturated,
the furan rings were then opened to form polyketones, and fi-
nally the polyketones were hydrodeoxygenated to alkanes.
Corma[7] and Zhang[8] also developed new processes for alkane
formation. After the CÀC coupling reaction of 2-methylfuran
and lignocellulose-derived carbonyl compounds, the highly un-
saturated oxygenates were directly hydrodeoxygenated to
branched alkanes over carbon-supported noble-metal (Pt, Pd,
and Ir) catalysts at 350–3708C and 5–6 MPa H2. All of these
strategies are creative, but the hydrodeoxygenation process is
complicated (multiple steps) or is operated at high tempera-
ture and pressure.
Fatty acids and triglycerides from inedible oil crops such as
microalgae and jatropha are considered as promising renewa-
ble sources of biodiesel.[2,3,9] Currently, three methods can be
utilized for the deoxygenation of fatty acids and triglycerides
to alkanes. The first relies on supported Group 9 and 10 metal
catalysts, such as Ni/ZrO2, Pd/C, and Pt/C, for the decarbonyla-
tion and decarboxylation of carboxylic acids to alkanes at 260–
3308C.[10] However, the decarbonylation and decarboxylation
not only require high temperatures but also produce hydrocar-
bons with one less carbon atom. The second utilizes conven-
tional hydrodesulfurization catalysts (sulfide NiMo or CoMo),
which are more selective to hydrodeoxygenation at 300–
4508C;[11] however, the metal sulfide catalysts can suffer from
sulfur leaching, which leads to contaminated products and de-
activation. The third employs zeolite-supported metal catalysts,
such as Ni/HBEA (HBEA=protonated zeolite beta)[12] and Pt–
Re/H-ZSM-5,[13] which have high selectivity for the hydrodeoxy-
genation of fatty acids to alkanes at 250–3008C, but a consider-
able selectivity for decarbonylation/decarboxylation still exists
[a] Y. Shao, Q. Xia, X. Liu, Prof. G. Lu, Prof. Y. Wang
Key Laboratory for Advanced Materials
Research Institute of Industrial Catalysis
East China University of Science and Technology
No. 130 Meilong Road, Shanghai (PR China)
Fax: (+86)21-64253824
E-mail: gzhlu@ecust.edu.cn
wangyanqin@ecust.edu.cn
Supporting Information for this article is available on the WWW under
http://dx.doi.org/10.1002/cssc.201500053.
ChemSusChem 0000, 00, 0 – 0
1
ꢀ 0000 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
These are not the final page numbers! ÞÞ

Full Papers
(>20%) for triglyceride conversion owing to the high tempera-
ture and H2 pressure.
In our previous work,[14] we developed a direct and efficient
approach for the production of liquid alkanes from furfural-de-
rived aldol adducts under mild conditions (1708C, 2 MPa) over
a Pd/NbOPO4 catalyst. We found that NbOx species promote
the CÀO bond cleavage significantly, especially the bonds of
tetrahydrofuran rings, as was further confirmed by in situ dif-
fuse reflectance infrared Fourier transform spectroscopy
(DRIFTS) experiments and DFT calculations. Herein, we report
a simple Pd-loaded 10%Nb2O5/SiO2 catalyst for the quantita-
tive hydrodeoxygenation of 4-(2-furyl)-3-buten-2-one to n-
octane under mild conditions (1708C, 2.5 MPa). The
10%Nb2O5/SiO2 support was prepared readily by the sol–gel
method without any surfactant; therefore, it can be synthe-
sized on a large scale at a low cost. We also demonstrated that
the Pd-loaded 10%Nb2O5/SiO2 catalyst can convert palmitic
acid, tristearin, and diphenyl ether (Scheme 1) quantitatively
Figure 1. XRD patterns of various Nb2O5/SiO2 materials: a) 5%Nb2O5/SiO2,
b) 10%Nb2O5/SiO2, c) 20%Nb2O5/SiO2, d) 10%Nb2O5/SiO2-im, and e) commer-
cial Nb2O5.
crystalline Nb2O5 phase was detected in these Nb2O5/SiO2 sam-
ples, the amorphous niobium oxide species were well dis-
persed in the silica. For the 10%Nb2O5/SiO2-im sample, which
was prepared by impregnating SiO2 with niobium citrate solu-
tion, the clear observation of some small Nb2O5 peaks indicates
that the Nb2O5 on the SiO2 surface was crystalline.
The N2 adsorption–desorption isotherms and pore size distri-
butions of the Nb2O5/SiO2 materials are shown in Figure S1. All
of the samples with different Nb content exhibit typical type
IV isotherms (IUPAC definition), which are characteristic of mes-
oporous materials.[15] The appearance of H2-type hysteresis
loops at a relative pressure of 0.4–0.8 indicates the presence of
“ink-bottle-type” pores in these materials. All of the Nb2O5/SiO2
samples possess mesoporous characteristics with pore size dis-
tributions in the range 2–5 nm (Figure S1b). The BET surface
areas, pore volumes, and average pore diameters are summar-
ized in Table 1. The specific surface areas and pore volumes in-
Scheme 1. Biomass-related compounds used in this work.
into alkanes with less CÀC cleavage through hydrodeoxygena-
tion under the same mild conditions; these are model com-
pounds for microalgae oils, vegetable oils, and lignin, respec-
tively. In addition, the reaction pathway for palmitic acid hy-
drodeoxygenation and the stability of 4%Pd/10%Nb2O5/SiO2
catalyst were investigated.
Table 1. Texture properties of Nb2O5/SiO2 prepared by the sol–gel
method and other materials.
Sample
SBET
[m2 gÀ1
Pore volume
[cm3 gÀ1
]
Pore size[a]
[nm]
]
SiO2
786
927
794
490
411
6
0.65
0.75
0.68
0.58
0.35
–
3.5
3.5
3.5
4.5
3.8
–
5%Nb2O5/SiO2
10%Nb2O5/SiO2
20%Nb2O5/SiO2
10%Nb2O5/SiO2-im
commercial Nb2O5
Results and Discussion
Characterization
The X-ray powder diffraction patterns of various Nb2O5/SiO2
materials and reference commercial Nb2O5 are shown in
Figure 1. For commercial Nb2O5, all of the diffraction peaks
could be indexed to orthorhombic Nb2O5 (JCPDS card 30-
0873). The SiO2 sample prepared by the sol–gel method is
amorphous, as only one broad band was detected from 2q=
158 to 408. As no reflections were observed in the low-angle
range (18<2q<108), long-range ordering is absent in these
mesoporous Nb2O5/SiO2 materials. The appearance of only one
very broad band (2q=158–408) in the wide-angle 2q region
implies the amorphous characteristic of these materials. As no
[a] Average pore diameters calculated from the desorption branches by
the Barrett–Joyner–Halenda (BJH) model.
crease first and then decrease with increasing Nb2O5 content,
[16]
[17]
as was also observed for TiO2/SiO2 and WO3/SiO2 materials
prepared by the sol–gel method.
The TEM images of the 10%Nb2O5/SiO2, 20%Nb2O5/SiO2, and
Pd-loaded Nb2O5/SiO2 catalysts are shown in Figure 2. No clear
phase separation can be observed in the whole detected
&
ChemSusChem 0000, 00, 0 – 0
www.chemsuschem.org
2
ꢀ 0000 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ÝÝ These are not the final page numbers!

Full Papers
Hydrodeoxygenation of 4-(2-furyl)-3-buten-2-one
The results for the hydrodeoxygenation of 4-(2-furyl)-3-buten-
2-one over different Pd-based catalysts in a batch reactor are
summarized in Table 2. The Pd/10%Nb2O5/SiO2 catalyst
showed the highest catalytic activity, and the octane yield was
95.5% (Table 2, entry 3), which is as good as that of Pd/
NbOPO4;[14b] therefore, the Pd/10%Nb2O5/SiO2 has an excellent
ability to catalyze the hydrodeoxygenation of furan-based
compounds under mild reaction conditions. The other 3.6% al-
kanes were mainly n-heptane and isohexadecane. Without the
addition of niobium oxide species, Pd/SiO2 showed much
lower activity, and the octane yield was only 16% (Table 2,
entry 6). Through the impregnation of Nb2O5 in SiO2, the yield
of octane improved to 31.2% for the 4%Pd/10%Nb2O5/SiO2-im
catalyst (Table 2, entry 7). These results suggest that the high
hydrodeoxygenation activity for furan-based compounds over
the Pd/10%Nb2O5/SiO2 catalyst should be ascribed mainly to
the presence of niobium oxide species. This behavior agrees
well with the previous report[14b] that NbOx species could pro-
mote the CÀO bond cleavage effectively. However, there were
still 64.1% oxygenates unconverted over the 4%Pd/
10%Nb2O5/SiO2-im catalyst. This is because the Nb2O5 impreg-
nated in the SiO2 crystallized to larger particles and, therefore,
there were less niobium oxide species exposed on the surface
of the catalyst. In addition, the much weaker acidity of
10%Nb2O5/SiO2-im than that of 10%Nb2O5/SiO2 may be anoth-
er reason for the poor hydrodeoxygenation activity of 4%Pd/
10%Nb2O5/SiO2-im. The Pd/5%Nb2O5/SiO2 catalyst was slightly
less active than Pd/10%Nb2O5/SiO2, as its lower Nb content led
to a slightly slower reaction rate. For an extended reaction
time of 30 h, the octane yield increased to 90.1% (Table 2, en-
tries 1 and 2). For the Pd/20%Nb2O5/SiO2 catalyst with a higher
Nb content, 95.3% octane yield was obtained in a shorter time
(16 h), and no oxygenates remained (Table 2, entry 4). These re-
sults further confirmed that NbOx species could promote the
CÀO bond cleavage and assist the hydrodeoxygenation pro-
cess. The Pd/commercial Nb2O5 catalyst showed the lowest ac-
tivity with only 1.9% octane yield (Table 2, entry 5) owing to its
small specific surface area and crystallinity.
Figure 2. TEM images of a) 10%Nb2O5/SiO2, b) 20%Nb2O5/SiO2, c) 4%Pd/
5%Nb2O5/SiO2, d) 4%Pd/10%Nb2O5/SiO2, and e) 4%Pd/20%Nb2O5/SiO2;
f) histogram of the Pd particle size distribution (for 300 particles).
region (Figure 2a and b), which implies that the niobium oxide
species are dispersed uniformly in the Nb2O5/SiO2 materials. It
is also clear from the Nb atomic maps (Figure S3 f) that the Nb
atoms were uniformly dispersed in the silica. The histogram of
the Pd particle size distribution shows that most of the Pd par-
ticles are distributed in the range 2–9 nm (Figure 2 f). Accord-
ing to the statistical analysis, the average Pd particle sizes in
4%Pd/5%Nb2O5/SiO2, 4%Pd/10%Nb2O5/SiO2 and 4%Pd/
20%Nb2O5/SiO2 were 5.9, 5.1, and 4.5 nm, respectively, and de-
creased gradually as Nb2O5 content increased.
Table 2. Reaction results for the direct hydrodeoxygenation of 4-(2-furyl)-3-buten-2-one over different Pd-based catalysts in a batch reactor.[a]
Entry
Catalyst
t
[h]
Conversion
[%]
Yield [%]
MTHFA[c] BTHF[d]
octane
other alkanes
THFA[b]
MPTHF[e]
octanols
dioctyl ether
1
2
3
4
5
6
7
Pd/5%Nb2O5/SiO2
Pd/5%Nb2O5/SiO2
Pd/10%Nb2O5/SiO2
Pd/20%Nb2O5/SiO2
Pd/commercial Nb2O5
Pd/SiO2
24
30
24
16
24
24
24
100
100
100
100
100
100
100
84.1
90.1
95.5
95.3
1.9
6.0
6.3
3.6
4.7
0
0
0
0
0
49.0
0
0
0
0
0
41.5
0
0
0
0
0
0
0
0
0
0
9.9
3.6
0.9
0
0
5.4
4.3
0
0
0
4.6
19.3
17.8
1.7
24.1
18.5
0.3
27.4
23.5
16.0
31.2
1.1
1.4
Pd/10%Nb2O5/SiO2-im
0
0
[a] Reaction conditions: catalyst (0.2 g), 4-(2-furyl)-3-buten-2-one(0.2 g), cyclohexane(6.46 g),1708C, initial H2 pressure 2.5 MPa; Pd loading in all catalysts
was 4 wt%, and the catalysts were reduced in situ. [b] 4-(2-Tetrahydrofuryl)-2-butanol. [c] 5-Methyl-2-tetrahydrofuranpropanol. [d] 2-Butyltetrahydrofuran.
[e] 2-Methyl-5-propyltetrahydrofuran.
ChemSusChem 0000, 00, 0 – 0
www.chemsuschem.org
3
ꢀ 0000 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
These are not the final page numbers! ÞÞ

Full Papers
Hydrodeoxygenation of other compounds with different
oxygen-containing functional groups
acid was completely converted to alkanes with a yield of
86.1% for n-C16 alkane. These results suggest that the 10%Ni/
10%Nb2O5/SiO2 catalyst can be used for the highly selective
hydrodeoxygenation of fatty acids to fatty alcohols or alkanes,
owing to the promotion effect of 10%Nb2O5/SiO2 on CÀO
cleavage.
The 4%Pd/10%Nb2O5/SiO2 catalyst was also tested in the hy-
drodeoxygenations of palmitic acid, tristearin, and diphenyl
ether, which are model compounds for microalgae oils, vegeta-
ble oils, and lignin, respectively. These three model compounds
contained three different oxygen-containing functional groups,
namely, carboxy, ester, and ether groups. The results are
shown in Table 3. The reactions of palmitic acid and tristearin
The reaction pathways of the hydrodeoxygenation of palmit-
ic acid were investigated by products analysis at different reac-
tion times. The time course of the conversion of palmitic acid
catalyzed by 4%Pd/10%Nb2O5/SiO2 is shown in Figure 3. Hexa-
Table 3. Hydrodeoxygenation of diphenyl ether, palmitic acid, and tris-
tearin by 4%Pd/10%Nb2O5/SiO2.[a]
Entry
Substrate
Product
Conversion
[%]
Selectivity
[%]
1
2
3[b]
palmitic acid
tristearin
diphenyl ether
n-C16 alkane
n-C18 alkane
cyclohexane
100
100
100
95.3
94.2
98.2
[a] Reaction conditions: catalyst (0.2 g), substrate (0.2 g), cyclohexane
(6.46 g), 1708C, initial H2 pressure: 2.5 MPa, reaction time: 24 h; catalysts
were reduced in situ. [b] The solvent was dodecane (6.46 g) rather than
cyclohexane.
Figure 3. The time course of the hydrodeoxygenation of palmitic acid cata-
lyzed by 4%Pd/10%Nb2O5/SiO2. Reaction conditions: catalyst (0.2 g), palmitic
acid (0.2 g), cyclohexane (6.46 g), 1708C, initial H2 pressure 2.5 MPa.
under mild conditions (1708C, 2.5 MPa H2 pressure) resulted in
high yields (>94%) of alkanes with the same chain length as
that of the starting compound, and the 5–6% byproducts
were the alkanes with one carbon atom less than the main
product, which were produced by the terminal CÀC cleavage.
Therefore, it can be suggested that the 4%Pd/10%Nb2O5/SiO2
catalyst could catalyze efficiently the hydrodeoxygenation of
a fatty acid and a fatty acid ester with the CÀC cleavage reac-
tion suppressed. In addition, the hydrodeoxygenation
of diphenyl ether to cyclohexane with 98.2% yield in-
dicated that the 4%Pd/10%Nb2O5/SiO2 catalyst can
also be used in the hydrodeoxygenation of ether
groups.
decanol first accumulated to 63.2% at 3 h and then diminished
gradually as the reaction processed; this indicates that the pal-
mitic acid was initially hydrogenated rapidly to hexadecanol.
For prolonged reaction times, the yield of n-C16 alkane in-
creased gradually as the hexadecanol decreased gradually. The
proposed pathways are shown in Scheme 2. According to the
A non-noble-metal Ni-loaded 10%Nb2O5/SiO2 cata-
lyst was also investigated for the hydrodeoxygena-
Scheme 2. Proposed reaction pathway for the hydrodeoxygenation of palmitic acid into
C16 alkane.
tion of palmitic acid to test the effect of 10%Nb2O5/
SiO2 on CÀO cleavage, and the results are shown in
Table 4. Notably, a high yield (97%) of hexadecanol
was obtained, and this compound is used widely in surfac-
tants, lubricants, plasticizers, emulsifiers, cosmetics, and bio-
fuels.[18] As the temperature increased to 2208C, the palmitic
product distribution versus reaction time, the hydrodeoxyge-
nation of palmitic acid to n-C16 alkane is a consecutive reac-
tion, and palmitic acid hydrogenation to hexadecanol is fol-
lowed by hexadecanol hydrogenolysis to n-C16 alkane; this is
consistent with the proposal by Lercher for the hydrodeoxyge-
nation of stearic acid to octadecane over Ni/HBEA catalyst.[12]
Table 4. Hydrodeoxygenation of palmitic acid over 10%Ni/10%Nb2O5/
SiO2.[a]
Stability test of 4%Pd/10%Nb2O5/SiO2
Entry
T
Conversion
Yield [%]
[8C] [%]
n-C16 alkane n-C15 alkane hexadecanol
The stability of the catalyst is important for its practical usage;
therefore, a cycle usage test of 4%Pd/10%Nb2O5/SiO2 for the
hydrodeoxygenation of palmitic acid was conducted, and the
results are shown in Figure 4. The nearly constant activity of
the 4%Pd/10%Nb2O5/SiO2 catalyst after reuse five times at
1
2
180 100
220 100
1.8
1.2
97.0
0
86.1
13.9
[a] Reaction conditions: catalyst (0.2 g), palmitic acid (0.2 g), cyclohexane
(6.46 g), initial H2 pressure: 2.5 MPa, reaction time: 24 h.
&
ChemSusChem 0000, 00, 0 – 0
www.chemsuschem.org
4
ꢀ 0000 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ÝÝ These are not the final page numbers!

Full Papers
Figure 6. Product distribution in the direct hydrodeoxygenation of a 3 wt%
palmitic acid in dodecane solution in a fixed-bed reactor over 4%Pd/
10%Nb2O5/SiO2 catalyst. Reaction conditions: 1708C, 2.5 MPa, 2.4 hÀ1 WHSV,
and 20 mLminÀ1 gas flow rate.
Figure 4. The cycle usage test and hydrothermal stability test of 4%Pd/
10%Nb2O5/SiO2 for the hydrodeoxygenation of palmitic acid. Reaction con-
ditions: catalyst (0.2 g), substrate (0.2 g), cyclohexane (6.46 g), 1708C, initial
H2 pressure: 2.5 MPa, reaction time: 5 h. The catalyst in the hydrothermal
stability test was 4%Pd/10%Nb2O5/SiO2 catalyst after hydrothermal treat-
ment at 2008C for 24 h.
the batch reactor; the yield of hexadecane reached 94–95%,
and the other 5–6% product was pentadecane (Figure 6).
Clearly, the catalyst showed high activity and good stability
with almost no decrease in hexadecane yield throughout the
150 h test. Owing to the mild reaction conditions, there was
no clear growth or aggregation of Pd particles after the 150 h
time-on-stream test on the basis of the TEM image and the his-
togram for the Pd particle size distribution shown in Figure 7.
There was only a little Pd leaching according to the inductively
coupled plasma atomic emission spectroscopy (ICP-AES) analy-
sis, and the Pd loadings of the fresh catalyst and used catalyst
1708C for 5 h in each cycle indicates the good stability of the
catalyst. In addition, the hydrothermal stability of the 4%Pd/
10%Nb2O5/SiO2 catalyst was investigated, because water is
ubiquitous in biomass conversion. The 4%Pd/10%Nb2O5/SiO2
catalyst was first treated hydrothermally at 2008C for 24 h and
then tested for the hydrodeoxygenation of palmitic acid at
1708C for 5 h. The reaction result (Figure 4) shows that the cat-
alytic performance of 4%Pd/10%Nb2O5/SiO2 remains almost
the same after the hydrothermal treatment. No crystalline
phases of niobium oxide species appeared in the XRD pattern
of the hydrothermally treated 4%Pd/10%Nb2O5/SiO2 catalyst
(Figure 5c). The addition of silica can inhibit the crystallization
Figure 5. XRD patterns of 4%Pd/10%Nb2O5/SiO2: a) fresh catalyst, b) used
catalyst after the 150 h time-on-stream test in a fixed-bed reactor, and c) cat-
alyst hydrothermally treated at 2008C for 24 h.
of niobia in niobia–silica catalysts to improve the hydrothermal
stability.[19] Therefore, the 4%Pd/10%Nb2O5/SiO2 catalyst has
a good hydrothermal stability.
The hydrodeoxygenation of palmitic acid was also investi-
gated in a continuous-flow system with a fixed-bed reactor at
1708C, 2.5 MPa, 2.4 hÀ1 weighted hourly space velocity (WHSV),
and a gas flow rate of 20 mLminÀ1. The results obtained from
the fixed-bed reactor were almost consistent with those from
Figure 7. a) TEM image of used 4%Pd/10%Nb2O5/SiO2 catalyst after the
150 h time-on-stream test in a fixed-bed reactor; b) histogram of Pd particle
size distribution of fresh 4%Pd/10%Nb2O5/SiO2 (black) and used 4%Pd/
10%Nb2O5/SiO2 (red) after the 150 h time-on-stream test (for 300 particles).
ChemSusChem 0000, 00, 0 – 0
www.chemsuschem.org
5
ꢀ 0000 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
These are not the final page numbers! ÞÞ

Full Papers
CA/H2O molar ratio of 1.0:0.31:11. After aging for 2 h, the mixture
was heated at 508C in the open air to remove water and all other
volatiles to afford the as-synthesized precursor. Finally, the dried
solid was calcined at 6008C for 3 h (heating rate 10 KminÀ1), and
the 10 wt%Nb2O5/SiO2 sample was obtained. The 5 wt%Nb2O5/SiO2
and 20wt%Nb2O5/SiO2 catalysts were prepared by the same
method with the amount of niobium citrate changed to 5.34 and
25.35 g, respectively, and the volume of TEOS and the TEOS/CA/
H2O molar ratio of 1:0.31:11 were unchanged. As a reference
sample, pure mesoporous silica (SiO2) was prepared under the
same conditions in the absence of niobium citrate. For comparison,
a 10 wt%Nb2O5/SiO2-im sample was also prepared by the incipient
wetness impregnation method. After the impregnation of SiO2
(prepared as above) with the calculated amount of an aqueous so-
lution of niobium citrate, the sample was dried at 508C for 12 h
and then calcined at 5008C for 3 h.
were 4.0 and 3.8 wt%, respectively. In addition, the XRD pat-
terns (Figure 5) showed that no crystalline phases of niobium
oxide species appeared, and the niobium oxide species re-
mained amorphous. The above evidence indicates that the
4%Pd/10%Nb2O5/SiO2 demonstrated excellent activity and sta-
bility for the hydrodeoxygenation reaction.
Conclusions
We have reported a simple and effective Pd/10%Nb2O5/SiO2
catalyst for hydrodeoxygenation reactions under mild condi-
tions. The Nb2O5/SiO2 support material prepared by the sol–gel
method possesses mesoporous characteristics with pore size
distributions in the range 2–5 nm, and amorphous niobium
oxide species were well dispersed in the silica. The Pd/
10%Nb2O5/SiO2 catalyst showed excellent activity for the hy-
drodeoxygenation of 4-(2-furyl)-3-buten-2-one (furan ring-
based compound) with a 95.5% yield of octane at 1708C and
a H2 pressure of 2.5 MPa. Palmitic acid and tristearin can also
be converted under the same conditions and gave high yields
(>94%) of alkanes with the same chain length as those of the
starting compounds. The catalyst can last for either 150 h
time-on-stream test or five reuse cycles without deactivation in
the hydrodeoxygenation of palmitic acid. The hydrodeoxyge-
nation of diphenyl ether to cyclohexane with 98.2% yield indi-
cates that the 4%Pd/10%Nb2O5/SiO2 catalyst can also be used
for the hydrodeoxygenation of ether groups. In conclusion, the
Pd/10%Nb2O5/SiO2 catalyst can catalyze efficiently the hydro-
deoxygenation of furan-ring-based compounds, fatty acids,
fatty acid esters, and ethers. Therefore, Pd/10%Nb2O5/SiO2 is
a versatile catalyst for the hydrodeoxygenation of biomass-re-
lated compounds to alkanes under mild conditions.
The Pd- and Ni-based catalysts were prepared by the incipient wet-
ness impregnation method with appropriate amounts of aqueous
solutions of Pd(NO3)2·xH2O and Ni(NO3)2·6H2O, respectively. The ob-
tained sample was dried at 508C for 12 h and then calcined at
5008C for 3 h (heating rate 1 KminÀ1). The Pd loading was 4 wt%,
and the Ni loading was 10 wt%. The Ni-based catalysts were re-
duced previously in a flowing 10% H2/Ar mixture at 5008C for 3 h.
Characterization
The powder XRD patterns were recorded with a Rigaku D/max-
2550VB/PC diffractometer by using CuKa (l=0.15406 nm) radia-
tion.
The N2 adsorption–desorption isotherms were measured at 77 K by
using a NOVA 4200e analyzer (Quantachrome Co. Ltd). Before the
measurements, the samples were outgassed at 2008C for 12 h
under vacuum to remove moisture and volatile impurities.
The TEM images were recorded with a FEI Tecnai F20s-TWIN instru-
ment, and the electron-beam accelerating voltage was 200 kV.
Experimental Section
Catalytic reactions and product analysis
Materials
The batch reactions for the direct hydrodeoxygenation of 4-(2-
furyl)-3-buten-2-one were performed in a 50 mL stainless-steel au-
toclave. Typically, 4-(2-furyl)-3-buten-2-one (0.2 g) and the catalyst
(0.2 g) were mixed with cyclohexane (6.46 g) in the autoclave. The
reactor was then sealed, purged three times with nitrogen, and
charged to 2.5 MPa H2. The reactor was then heated to 1708C
under magnetic stirring at 600 rpm for 24 h. After the completion
of the reaction, the reactor was quenched in a water bath to room
temperature. The liquid solution was separated from the solid cata-
lyst by centrifugation and analyzed by GC–MS (Agilent 7890A-
5975C) with an HP-5 column. Tridecane was used as the internal
standard for the quantification of the liquid products. The batch
hydrodeoxygenation reactions of palmitic acid, tristearin, and di-
phenyl ether were conducted in the same way.
Pd(NO3)2·xH2O solution was purchased from Heraeus Materials
Technology Shanghai Co., Ltd, palmitic acid and tristearin were
purchased from TCI, diphenyl ether was purchased from Aladdin,
and commercial Nb2O5 was purchased from Sinopharm Chemical
Reagent Co., Ltd. 4-(2-Furyl)-3-buten-2-one was prepared by the
aldol condensation of furfural with acetone according to the litera-
ture procedure.[14b] All other chemicals were purchased from Sino-
pharm Chemical Reagent Co., Ltd. and used without purification.
Catalyst preparation
The Nb precursor (niobium citrate) was prepared according to the
literature procedure,[20] and a niobium citrate solution with a Nb
concentration of 0.4 mollÀ1 was prepared for use.
The direct hydrodeoxygenation of palmitic acid was also tested in
a fixed-bed reactor system. A feed composed of a 3 wt% solution
of palmitic acid in dodecane and a H2 co-feed were used for the
continuous-flow reaction test. The pelletized catalyst (1.0 g, 40–
60 mesh) was loaded into the stainless-steel tubular reactor (inner
diameter 6 mm, length 55 cm). Crushed quartz granules were
placed into both ends of the catalyst to maintain the bed height
and reduce the dead volume. After the loading, the reaction tem-
perature and pressure were adjusted to the desired value, and the
Nb2O5/SiO2 with different Nb2O5 content was prepared by a sol–gel
method in the presence of citric acid, which is an effective method
for the synthesis of thermally stable mesoporous aluminophos-
phate materials or mesoporous silica-based materials.[17,21] The de-
tailed process is as follows: A certain amount of citric acid (CA)
and distilled water were added to niobium citrate solution
(0.4 molLÀ1
20 mL) was added under stirring to afford a mixture with a TEOS/
, 11.26 g), and then tetraethyl orthosilicate (TEOS,
&
ChemSusChem 0000, 00, 0 – 0
www.chemsuschem.org
6
ꢀ 0000 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ÝÝ These are not the final page numbers!

Full Papers
Li, G. Li, W. Wang, A. Wang, X. Wang, Y. Cong, T. Zhang, ChemSusChem
2013, 6, 1149–1152.
[9] a) S. Lestari, P. Mꢂki-Arvela, J. Beltramini, G. Lu, D. Y. Murzin, Chem-
SusChem 2009, 2, 1109–1119; b) J. C. Serrano-Ruiz, E. V. Ramos-Fernꢃn-
dez, A. Sepffllveda-Escribano, Energy Environ. Sci. 2012, 5, 5638–5652.
[10] a) L. Li, E. Coppola, J. Rine, J. L. Miller, D. Walker, Energy Fuels 2010, 24,
feed flow was injected using an HPLC pump at a H2 flow of
10 mLminÀ1. The WHSV was determined by dividing the mass flow
rate of the liquid by the mass of the catalyst used. The liquid
phase was separated from the gas phase and collected with a gas–
liquid separator. The liquid phase analysis was performed with an
Agilent 7890A GC–MS instrument equipped with an HP-5 column.
Tridecane was used as the internal standard for the quantification
of the liquid products.
ˇ
1305–1315; b) M. Snåre, I. Kubickovꢃ, P. Mꢂki-Arvela, D. Chichova, K.
ˇ
Erꢂnen, D. Y. Murzin, Fuel 2008, 87, 933–945; c) M. Snåre, I. Kubickova,
P. Mꢂki-Arvela, K. Erꢂnen, D. Y. Murzin, Ind. Eng. Chem. Res. 2006, 45,
5708–5715; d) B. Peng, X. Yuan, C. Zhao, J. A. Lercher, J. Am. Chem. Soc.
2012, 134, 9400–9405.
Acknowledgements
[11] a) G. W. Huber, P. O’Connor, A. Corma, Appl. Catal. A 2007, 329, 120–
129; b) Y. Liu, R. Sotelo-Boyꢃs, K. Murata, T. Minowa, K. Sakanishi, Energy
Fuels 2011, 25, 4675–4685; c) R. Kumar, B. S. Rana, R. Tiwari, D. Verma,
R. Kumar, R. K. Joshi, M. O. Garg, A. K. Sinha, Green Chem. 2010, 12,
2232–2239; d) R. Sotelo-Boyꢃs, Y. Liu, T. Minowa, Ind. Eng. Chem. Res.
2011, 50, 2791–2799.
[12] B. Peng, Y. Yao, C. Zhao, J. A. Lercher, Angew. Chem. 2012, 124, 2114–
2117.
[13] K. Murata, Y. Liu, M. Inaba, I. Takahara, Energy Fuels 2010, 24, 2404–
2409.
This project was supported financially by the National Natural
Science Foundation of China (No. 21273071), the Science and
Technology Commission of Shanghai Municipality (13520711400,
13JC1401902, and 10dz2220500), and the Fundamental Research
Funds for the Central Universities.
[14] a) W. Xu, Q. Xia, Y. Zhang, Y. Guo, Y. Wang, G. Lu, ChemSusChem 2011, 4,
1758–1761; b) Q. N. Xia, Q. Cuan, X. H. Liu, X. Q. Gong, G. Z. Lu, Y. Q.
Wang, Angew. Chem. Int. Ed. 2014, 53, 9755–9760; Angew. Chem. 2014,
126, 9913–9918.
[15] K. S. Sing, Pure Appl. Chem. 1985, 57, 603–619.
[16] A. Sinha, S. Seelan, M. Okumura, T. Akita, S. Tsubota, M. Haruta, J. Phys.
Chem. B 2005, 109, 3956–3965.
[17] G. Liu, X. Wang, X. Wang, H. Han, C. Li, J. Catal. 2012, 293, 61–66.
[18] A. Corma, S. Iborra, A. Velty, Chem. Rev. 2007, 107, 2411–2502.
[19] H. N. Pham, Y. J. Pagan-Torres, J. C. Serrano-Ruiz, D. Wang, J. A. Dumesic,
A. K. Datye, Appl. Catal. A 2011, 397, 153–162.
[20] A. Sarkar, P. Pramanik, Microporous Mesoporous Mater. 2009, 117, 580–
585.
[21] a) G. Liu, M. Jia, Z. Zhou, W. Zhang, T. Wu, D. Jiang, Chem. Commun.
2004, 1660–1661; b) G. Liu, Y. Liu, G. Yang, S. Li, Y. Zu, W. Zhang, M. Jia,
J. Phys. Chem. C 2009, 113, 9345–9351.
Keywords: alkanes · biomass · heterogeneous catalysis ·
hydrodeoxygenation · niobium
[1] D. M. Alonso, J. Q. Bond, J. A. Dumesic, Green Chem. 2010, 12, 1493–
1513.
[2] C. Zhao, T. Brꢁck, J. A. Lercher, Green Chem. 2013, 15, 1720–1739.
[3] R. W. Gosselink, S. A. Hollak, S. W. Chang, J. van Haveren, K. P. de Jong,
J. H. Bitter, D. S. van Es, ChemSusChem 2013, 6, 1576–1594.
[4] a) G. W. Huber, J. N. Chheda, C. J. Barrett, J. A. Dumesic, Science 2005,
308, 1446–1450; b) R. M. West, Z. Y. Liu, M. Peter, J. A. Dumesic, Chem-
SusChem 2008, 1, 417–424.
[5] R. Xing, A. V. Subrahmanyam, H. Olcay, W. Qi, G. P. van Walsum, H.
Pendse, G. W. Huber, Green Chem. 2010, 12, 1933–1946.
[6] A. D. Sutton, F. D. Waldie, R. Wu, M. Schlaf, L. A. Silks, J. C. Gordon, Nat.
Chem. 2013, 5, 428–432.
[7] A. Corma, O. de La Torre, M. Renz, Energy Environ. Sci. 2012, 5, 6328–
6344.
Received: January 12, 2015
Revised: February 12, 2015
Published online on && &&, 0000
[8] a) G. Li, N. Li, S. Li, A. Wang, Y. Cong, X. Wang, T. Zhang, Chem.
Commun. 2013, 49, 5727–5729; b) G. Li, N. Li, J. Yang, L. Li, A. Wang, X.
Wang, Y. Cong, T. Zhang, Green Chem. 2014, 16, 594–599; c) J. Yang, N.
ChemSusChem 0000, 00, 0 – 0
www.chemsuschem.org
7
ꢀ 0000 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
These are not the final page numbers! ÞÞ

FULL PAPERS
Y. Shao, Q. Xia, X. Liu, G. Lu,* Y. Wang*
&& – &&
Pd/Nb2O5/SiO2 Catalyst for the Direct
Hydrodeoxygenation of Biomass-
Related Compounds to Liquid Alkanes
under Mild Conditions
Mild mannered: Under mild conditions
(1708C, 2.5 MPa), Pd/10%Nb2O5/SiO2 is
effective for the hydrodeoxygenation re-
actions of 4-(2-furyl)-3-buten-2-one (the
aldol adduct of furfural with acetone),
palmitic acid, tristearin, and diphenyl
ether (model compounds of vegetable
oils, microalgae oils, and lignin) and
gives high yields (>94%) of alkanes
with little CÀC bond cleavage.
&
ChemSusChem 0000, 00, 0 – 0
www.chemsuschem.org
8
ꢀ 0000 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ÝÝ These are not the final page numbers!