Synergetic Catalysis of Nickel Oxides with Oxygen Vacancies and Nickel Phosphide for the Highly Efficient Hydrodeoxygenation of Phenolic Compounds

Article abstract of DOI:10.1002/cctc.201800010

The highly efficient hydrodeoxygenation of oxygenated chemicals at a low temperature is a critical issue for biomass crude oil upgrading to reduce equipment and operation expense. The unique electronic features of oxygen vacancies have been found to facilitate deoxygenation above 250 °C. In this work, a series of NiO_x/SiO₂ catalysts that contain oxygen vacancies was prepared by a sol–gel and controllable temperature-programmed reduction method. Results show that NiO_x/SiO₂ has a superior phenol hydrogenation activity in deoxygenation even at 180 °C to Ni₂P/SiO₂. The high hydrogenation activity derives from the flat adsorption of phenolics over low-valence Ni cations induced by connected oxygen vacancies. After the introduction of NiO_x into Ni₂P catalysts, composite NiO_x-Ni₂P/SiO₂ (0.43≤x<1) bifunctional catalysts demonstrate better phenol hydrodeoxygenation performances under mild reaction conditions in comparison with individual Ni₂P and NiO_x catalysts and most catalysts presented previously because the coexistence of oxygen vacancies from NiO_x and the acidity induced by Ni₂P results in the synergistic effect of the adsorption configuration and hydrodeoxygenation ability. Besides, this composite catalyst also has a high activity for the hydrodeoxygenation of different bio-derived cresol isomers. The reuse test results confirm that oxygen vacancies are relatively stable in comparison with the phosphide.

Full text of DOI:10.1002/cctc.201800010

Accepted Article
Title: Synergetic catalysis of NiOx with oxygen vacancies and Ni2P for
highly efficient hydrodeoxygenation of phenolic compounds
Authors: Yunhua Li, Yafei Zhao, Binghui Chen, and Wenju Wang
This manuscript has been accepted after peer review and appears as an
Accepted Article online prior to editing, proofing, and formal publication
of the final Version of Record (VoR). This work is currently citable by
using the Digital Object Identifier (DOI) given below. The VoR will be
published online in Early View as soon as possible and may be different
to this Accepted Article as a result of editing. Readers should obtain
the VoR from the journal website shown below when it is published
to ensure accuracy of information. The authors are responsible for the
content of this Accepted Article.
To be cited as: ChemCatChem 10.1002/cctc.201800010
Link to VoR: http://dx.doi.org/10.1002/cctc.201800010
A Journal of
www.chemcatchem.org

10.1002/cctc.201800010
ChemCatChem
FULL PAPER
Synergetic catalysis of NiOx with oxygen vacancies and Ni2P for
highly efficient hydrodeoxygenation of phenolic compounds
Yunhua Li,*, [a] Yafei Zhao,[a] Binghui Chen,[a], Wenju Wang†, [b]
Abstract: High efficient hydrodeoxygenation of oxygenated
chemicals at low temperature is a critical issue for biomass crude oil
upgrading to reduce equipment and operation expenses. Unique
electronic features of oxygen vacancies have been found to facilitate
occurrence of deoxygenation above 250°C. In this work, firstly a
series of NiOx/SiO2 catalysts containing some oxygen vacancies
were prepared by sol-gel and controllable temperature-programmed
reduction method. Results show that NiOx/SiO2 has superior phenol
hydrogenation activity to deoxygenation even at 180°C, which was
contrary to Ni2P/SiO2. High hydrogenation activity derives from flat
adsorption of phenolics over low valence Ni cations induced by
HDO of guaiacol proceeds efficiently on oxygen vacancies in
catalysts.[10] Prasomsri et al. pointed out that the C-O bands can
be further weakened when biomass-oxygenate adsorbed on the
oxygen vacancy sites of the MoO3 catalysts.[11] Moreover,
Alexander et al. also noted that RuO2 oxygen vacancies promote
C-O band scission.[12] In previous study, we has also found that
the oxygen vacancies on the RuO2/C catalysts favor the
cleavage of O-H in water.[13] Deoxygenation property of oxygen
vacancies is related to the generation of low valence cation,
which acts as Lewis acid and hydrogen dissociation site. Thus,
effects of acidity and hydrogenation capacity on reaction
connected oxygen vacancies. As introducing NiOx into Ni2P catalysts, performances have been considered simultaneously.
composite NiOx-Ni2P/SiO2 (0.43≤x<1) bifunctional catalysts
demonstrate excellent phenol HDO performances under mild
reaction conditions in comparison with individual Ni2P and NiOx
catalysts and most of results from literatures because the
coexistence of oxygen vacancies from NiOx and acid induced by
Ni2P results in synergistic effect of adsorption configuration and
hydrodeoxygenation ability, Besides, this composite catalyst also
has a high activity for hydrodeoxygenation of different bio-derived
cresol isomers. The reuse test result confirms that oxygen vacancies
are relatively stable in comparison with phosphide.
In previous study, we observed that Ni2P/SiO2, from the sol-gel
and temperature-programmed reduction (TPR) method,
presents the higher hydrodeoxygenation activity than the
traditional Ni2P supported catalysts due to highly dispersed
Ni2P.[14-16] Triggered by unique features of phosphide and
oxygen vacancies, in this work, we introduce oxygen vacancies
into Ni2P supported catalysts by modulating appropriate
phosphorous precursor and reduction temperature to obtain high
catalytic hydrodeoxygenation activity of catalysts. Study shows
under mild reaction conditions (180°C, 2 MPa), hydrogenation
capacity outweighs deoxygenation on NiOx containing oxygen
vacancies, whereas Ni2P is the opposite. By comparison with
individual nickel oxide and phosphide, NiOx-Ni2P/SiO2 containing
appropriate oxygen vacancies presents an excellent catalytic
activity. Adsorption-desorption of N2, Raman Spectroscopy, H2-
Introduction
Through the tireless efforts of researchers, biomass conversion
to liquid products has mainly involved gasification, hydrothermal,
transesterification and fast pyrolysis. Among them, fast pyrolysis
has been considered to be a most promising process because it
gives high liquid oil yield.[1] Since there exist high contents of
oxygenate compounds in the biomass oil produced by fast
pyrolysis,[2] prior to the use of the bio-oils, removal of oxygen
from the bio-oil through catalytic hydrodeoxygenation (HDO) has
been employed to upgrade the quality of the bio-oil.
Transition metal phosphide has generally been used for the
HDO process, especially the nickel phosphide[3-6] because of
their unique physical and chemical properties from phosphorus
inserting into nickel. Additionally, some vacancies in the surface
of catalysts have great promotion on the catalytic performance
of reactions due to distinct electronic characteristic. It has been
found that oxygen vacancies can facilitate deoxygenation of
oxygenates.[7-9] For instance, Foraita et al. reported that the
TPR,
X-ray
diffraction
(XRD),
temperature
programmed desorption of ammonia(NH3-TPD), inductively
coupled plasma atomic emission spectrometry (ICP-AES) ,
electron spin resonance (ESR), transmission electron
microscopy (TEM), Fourier-transform infrared spectroscopy of
pyridine adsorption (PY-FTIR), In-situ X-ray photoelectron
spectroscopy (XPS) characterization and endurance experiment
were conducted to unravel physiochemical features of catalysts.
HDO result of different phenolic oxygenates was also evaluated.
This study contributes to developing highly efficient catalysts
for hydrodeoxygenation at low temperature.
Results and Discussion
Characterization of catalysts
Temperature programmed reduction experiments were
conducted to determine appropriate reduction temperature to
form Ni2P or NiOx. Figure 1 shows TPR profiles of different
catalysts precursor. For the NiOx/SiO2 precursor, around 520°C
the bulk nickel oxide is reduced. By comparison, the peak
temperature of Ni2P/SiO2 is relatively high and that for starting
reduction is around 500°C . NiOx-Ni2P/SiO2 precursor presents
two prominent peaks corresponding to nickel oxide and nickel
oxy-phosphates, respectively.[17] Through TPR results, reduction
temperature in catalyst preparation process was programmed at
550°C to obtain NiOx containing oxygen vacancies or Ni2P.
[a]
[b]
Associate professor Yunhua Li
Department of Chemical and Biochemical Engineering, College
of Chemistry and Chemical Engineering
Xiamen University
Xiamen, 361005, PR China
E-mail: yunhuali@xmu.edu.cn
School of Energy and Power Engineering, Nanjing University of
Science and Technology, Nanjing 210094, China
E-mail: wangwenju1982@gmail.com
This article is protected by copyright. All rights reserved.

10.1002/cctc.201800010
ChemCatChem
FULL PAPER
To ensure the existence of oxygen vacancies, the effect of
reduction time was carefully investigated at 550°C . X-ray
diffraction patterns of the fresh NiOx/SiO2 catalysts reduced at
different times are shown in Figure 2a. XRD pattern for fresh
NiOx/SiO2 shows peaks at 44.5º and 51.8º (JCPDS: 04-0850),
attributed to Ni metal. The diffraction intensity for Ni gradually
increases with the increment of reduction time and the sample
reduced at 550°C for 3 h has more prominent diffraction peak for
Ni. The catalyst reduced at 550°C for 1 h has relatively low and
wide diffraction peaks, implying that small NiOx crystallites might
be highly dispersed on the silica support.
Based on the above results, Ni2P/SiO2, NiOx/SiO2 and NiOx-
Ni2P/SiO2 were reduced at 550°C for 1 h and the corresponding
XRD profiles are presented in Figure 2b. Diffraction peaks at
40.8°, 44.6°, and 47.3° (JDPDS: 03-0953) are ascribed to Ni2P
nanocrystal. Thus, we could control the proportion of NiOx and
Ni2P of catalysts through modulating the phosphorus loading
and calcination conditions.
Figure 1. H2-TPR profiles of NiOx/SiO2, Ni2P/SiO2 and NiOx-Ni2P/SiO2
precursor.
a
a
b
b
Figure 2. a. XRD profiles of NiOx/SiO2 reduced at different time, b. XRD
profiles of Ni2P/SiO2, NiOx/SiO2, NiOx-Ni2P/SiO2 reduced at 550°C for 1 h.
Figure 3. a. Low temperature ESR spectra of NiOx/SiO2 reduced at different
conductions, b. Low temperature ESR spectra of samples reduced at 550 °C
for 1 h.
Table 1. Physiochemical properties of catalysts
Vacancies[c]
(µmol/g)
Lewis acid
amount[d]
(µmol/g)
55
Ni[a]
Ni:P
Ratio[a]
Average
size [b] (nm)
SBET
(m2·g -1)
Vpore
(cm3·g -1)
Dpore
(nm)
Sample
(wt %)
270
171
130
30
9.2
8.7
9.1
8.5
1.0:0.00
1.0:0.24
1.0:0.41
1.0:0.51
3.2
3.5
3.3
3.6
444.2
360.5
332.8
279.3
0.528
0.488
0.472
0.431
3.972
4.369
4.293
4.644
10%NiOx/SiO2 (10%NiOx/SiO2-550°C-1.0 h)
6.7% NiOx-3.3 % Ni2P/SiO2
3.1% NiOx-6.9 % Ni2P/SiO2
10% Ni2P/SiO2
176
242
303
[a] Results derive from ICP characterization. [b] Data are from the Scherrer equation (Dc = Kλ/βcos θ). [c] Results come from ESR.[d] Data are from NH3-TPD.
This article is protected by copyright. All rights reserved.

10.1002/cctc.201800010
ChemCatChem
FULL PAPER
Table 1 shows structure and composition characterization
results of Ni supported catalysts. Since catalyst precursor comes
from the same NiO/SiO2 precursor, the crystallite sizes of all the
catalysts are similar, about 3.5 nm, in agreement with TEM
result (Figure S1). The as-obtained samples also have the
similar Ni loading, about 9.0 wt.%. Moreover, the Ni/P ratio of
Ni2P/SiO2 catalyst is very close to the expected stoichiometry of
Ni2P, implying the formation of a large amount of Ni2P phase.
This result is consistent with XRD pattern. N2 adsorption-
desorption results present that in comparison with NiOx, the
composite catalysts suffer loss of surface area and pore volume
to some extent with the addition of P. We listed also the physic-
chemical characterization data from literatures in Tables S1.
Comparison of Tables 1 and Table S1 shows that under the
conditions of the similar Ni loading, the average size of Ni
particles (about 3.5 nm) in this work is a bit smaller and the
specific surface area (350m¬2•g-1) is higher than results from
literatures.. This mainly is from the addition of nickel precursor in
the sol-gel process, lower reduction temperature and short
reduction time.
a
b
Figure 5. a.NH3-TPD profiles and b. PY-FTIR spectra of Ni2P/SiO2, NiOx/SiO2
and NiOx-Ni2P/SiO2 reduced at 550°C for 1 h.
In order to clarify the acidity of Ni supported catalysts, NH3-TPD
and PY-FTIR characterization were also shown in Figure 5. All
the catalysts have a peak at 1445 cm-1 corresponding to weak
Lewis acidic sites. For Ni2P/SiO2, this peak is assigned to Lewis
acid Niδ+ and the amount of acid sites is 303µmol/g. The acidity
of composite catalysts derives from the low valence nickel ion
due to coexistence of NiOx and Ni2P phase. Additionally, among
these catalysts, NiOx/SiO2 with the higher content of oxygen
vacancies demonstrates the least acid amount 55µmol/g in turn.
In addition to the acidity of catalysts, XPS of the catalysts was
employed to understand the surface chemical components
(Figure 6). As illustrated in Figure 6a, the XPS peaks at 852.0
and 856.0 eV were assigned to Niδ+ (0≤δ<2) and Ni2+.[18] With
the prolonging of reduced time, the former is enhanced whereas
the latter peak gradually decreases for NiOx/SiO2, indicating that
there is the more metallic Ni generated with time. Although low
valence Ni species induced by oxygen vacancies are different
from Ni metal, it might partially form Ni0 by rearrangement under
the certain conditions. For the sample reduced at 550°C for 1 h,
Ni oxide and metal coexist and at this moment the catalyst
contains the more oxygen vacancies. In order to estimate the
value for x, we fitted the peak areas of Ni2+ and Niδ+ in the XPS
results to obtain individual contents and assumed that these
values come from NiO and Ni, respectively, to determine the
value of x. The value for x in NiOx-Ni2P/SiO2 and NiOx/SiO2 is in
the range of 0.43~1 (Table S2). For the individual NiOx/SiO2,
with the increasing in the reduction time from 0.5 h to 3 h, the
Figure 4. Raman spectra of Ni2P/SiO2, NiOx/SiO2 and NiOx-Ni2P/SiO2.
To further confirm the existence of oxygen vacancies, low
temperature ESR was performed (Figure 3), which is sensitive to
the presence of oxygen vacancies. The amount of oxygen
vacancies of fresh catalysts is also listed in Table 1. Figure 3a
shows that with the increasing in the reduction time, a signal
peak at g≈2.02, characteristic of surface oxygen vacancies,[7, 18,
19]
increases distinctly. NiOx/SiO2 reduced at 550°C for 1 h
presents the strongest characteristic peak and the amount of
oxygen vacanies is 270µmol/g. This indicates in the reduction
process of NiO/SiO2 to Ni/SiO2, oxygen vacancies are generated
gradually. When the reduction of the NiO phase is almost
completed, the oxygen vacancies disappear completely.
Figure 3b further presents low temperature ESR spectra of
Ni2P/SiO2, NiOx-Ni2P/SiO2 and NiOx/SiO2. By comparison,
NiOx/SiO2 has the strongest characteristic peak. With the
increasing in P content, the characteristic peak of oxygen
vacancies decreases gradually and that of Ni2P/SiO2 is the
smallest. Raman peaks at around 500 cm-1 on catalysts with
oxygen vacancies (Figure 4) shift towards the lower
wavenumber with the reduction of P in the catalysts, further
confirming the presence of an induced effect by surface
defects.[20-23]
This article is protected by copyright. All rights reserved.

10.1002/cctc.201800010
ChemCatChem
FULL PAPER
value for x decreased from 0.88 to 0.17. The NiOx can be seen
as an intermediate state between Ni and NiO. As phosphorus
was introduced into catalysts, the ratio of Niδ+ and Ni2+ in 3.1%
NiOx-6.9% decreased accompanied by the decrement of oxygen
vacancies. In comparison with Ni2P/SiO2, Ni 2p XPS peak in
NiOx/SiO2 shifts towards higher binding energy, indicating that Ni
element in NiOx/SiO2 is the more electron deficient than Ni2P.
This indicates that in this case low valence Ni derives from both
Ni2P and induction of oxygen vacancies. XPS peaks of P 2p in
catalysts at 128.1 and 132.9 Ev are ascribed to reduced P
species in the Ni2P phase and phosphorus oxide, respectively, in
Figure 6c.[15] Phosphorus oxide is intensified in Ni2P/SiO2 the
more Ni2P was passivated with the increasing in the P loading
over catalysts.
Figure 7. CO-FTIR spectroscopy for different catalysts.
a
CO-FTIR characterization was conducted to further determine
the chemical properties of the catalyst samples in Figure 7. Two
main absorption bands around 2073 cm−1 and 1942 cm−1 are
observed for these catalysts, which are assigned to linear and
bridge CO stretching vibrations.[16] By comparison, the bridge
CO stretching vibration in NiOx-Ni2P/SiO2 catalyst shifts towards
a higher frequency as the Ni2P content drops. At the same time,
linear CO vibration gradually disappears. According to the
Blyholder model,[24] the adsorbed CO shifts toward the higher
frequencies is due to the decrement of electron transfer to the
CO π ∗ antibonding orbital. Electron deficiency induced by
oxygen vacancies results in lower capacity of electron donation
of NiOx than Ni2P. This follows electronegativity principle to
some extent. Electron deficiency of low valence Ni species
contributes to a flat adsorption of aromatic ring in phenol, which
has the relatively higher electron density.
b
c
Catalytic performances over phosphide catalysts
As we all known, oxygen vacancies facilitate adsorption and
deoxygenation of oxygenates at temperature higher than
250°C.[9, 22] Table S3 shows the conversion and selectivity for
phenol HDO over NiOx/SiO2 catalysts. At the lower temperature,
the presence of highly dispersed low valence Ni species favors
hydrogenation of phenolics adsorbed (entry 1, 2 and 3). Along
with the increase of reaction temperature (entry 2, 4 and 5),
phenol conversion increases gradually and when the reactions
were 200°C and 0.5 h, almost all the phenol has already been
converted. In the meanwhile, the selectivity to cyclohexanol
decreases and the selectivity to the cyclohexane increases.
Cyclohexanol is intermediate for phenol to cyclohexane and high
temperature and long reaction time (entry 2, 4, 5, 6 and 7)
benefit hydrodeoxygenation cascade reaction to proceed. It has
generally accepted that the existence of oxygen vacancies
favors deoxygenation. However, lower temperature contributes
to the adsorption of reactants. At lower temperature the total
hydrogenation of the aromatic ring probably involves a flat
adsorption of phenol through its aromatic ring on connected
9]
oxygen vacancies or neighboring vacancies over NiOx.[8, In
addition, in order to explore whether the catalyst NiOx/SiO2 is a
mixture of NiO/SiO2 and metallic Ni/SiO2, we compared catalytic
performance of 0.1 g 10% NiOx/SiO2 with a mixture of
0.05g10%NiO/SiO2 and 0.05g 10%Ni/SiO2 under the same
reaction conditions for phenol hydrodeoxygenation. As shown in
Figure 6. XPS spectra of Ni 2 p and P 2 p of the different catalysts, a. Ni 2 p of
NiOx/SiO2 reduced at different reduction time, b. Ni 2 p of NiOx- Ni2P/SiO2, c. P
2 p of NiO -Ni P/SiO
.
x
2
2
This article is protected by copyright. All rights reserved.

10.1002/cctc.201800010
ChemCatChem
FULL PAPER
the Table S3, the NiOx/SiO2 catalyst containing oxygen
vacancies has much better catalytic activity than the mixture of
NiO/SiO2 and Ni/SiO2. Therefore, the catalyst isn’t simply a
mixture of NiO and metallic Ni. The TOF of NiOx/SiO2 is higher
than Ni/SiO2, indicating that the oxygen vacancies in NiOx/SiO2
are beneficial for the catalytic activity for hydrogenation.
Figure 8. A Scheme displaying the different catalytic steps and the type of
involved active sites
Table 2 compares catalytic results of phenol HDO over Ni2P
/SiO2, NiOx-Ni2P/SiO2, NiOx/SiO2 and literature data. Under the
silimar reaction conditions, in spite of the lowest conversion of
phenol, only 53.5%, the higher selectivity of cyclohexane was
56.6% over NiP2/SiO2 catalyst in comparison with other catalysts.
Lewis acid site contributes to the deoxygenation of oxygenated
chemicals. Since Ni2P/SiO2 (303µmol/g) contains by far the
more Lewis acid sites than NiOx/SiO2 (55µmol/g), the
deoxygenated product cyclohexane selectivity (56.6%) on
Ni2P/SiO2 is rather higher than that of NiOx/SiO2 (8.2%). In
contrast, NiOx/SiO2 catalyst has 100.0% phenol conversion and
90.8% hydrogenated product cyclohexanol selectivity. The
content of oxygen vacaneis in the NiOx/SiO2 (270µmol/g) much
higher than Ni2P/SiO2 (30µmol/g) contributes to more flat
adsorptions of phenol on low valence Ni speicies induced by
oxygen vacancies to hydrgenate. At this moment, the low
selectivity of product cyclohexane was due to the less of Lewis
acidic sites in the catalyst. Thus, the high intermediate product
cyclohexanol is obtained. For the 3.1% NiOx-6.9% Ni2P/SiO2
catalysts, the conversion of phenol was 97.3% and selectivity of
cyclohexane was 73.7%. The yield of cyclohexane, 71.7%, over
3.1% NiOx-6.9% Ni2P/SiO2 was much higher than other three
catalysts and those reported in literatures,[25-29] which confirmed
there are distinct synergistic effects of oxygen vacancies in NiOx
for hydrogenation and Lewis acid sites in Ni2P for deoxygenation.
(Figure 8). From TOF results (Table 2), we observe that
consumed phenol converts into a large amount of cyclohexanol
on connected oxygen vacancies of NiOx, which is benefit for flat
adsorption of phenol to hydrogenate. With the presence of
phosphorus, the acidity of 3.1% NiOx-6.9% Ni2P/SiO2 is
strengthened. In this respect, cyclohexane is generated a lot
since the coexistence of Ni2P and NiOx leads to the more acid
sites and oxygen vacancies.
For comparison, 10% Ni(H2PO2)2/SiO2 was characterized
(Figure S2) and evaluated. Ni(H2PO2)2/SiO2 demonstrates the
XRD pattern consistent with pure Ni(H2PO2)2 powder, which is
different from that of NiOx-Ni2P/SiO2. Based on the similar Ni
loading, catalytic activity of Ni(H2PO2)2/SiO2 is by far the lower
than NiOx-Ni2P/SiO2. This indicates that NiOx-Ni2P/SiO2 contains
individual nickel oxide and phsophide characteristic phase rather
than hypophosphite.
Table 3 shows the reuse results over 3.1% NiOx-6.9% Ni2P/SiO2
for HDO reaction. In three cycles, phenol conversion almost
keeps constant accompanied by the slight change in the content
of oxygen vacancies (Table S4). At the same time, cyclohexane
selectivity decreases from 73.7% to 52.6% due to P loss.
HDO of other bio-derived cresol isomers in biomass crude oil
over 3.1% NiOx-6.9% Ni2P/SiO2 catalyst was evaluated in Table
4. Since solubility of partial phenolics in dodecane is limited,
used solvent was changed into N-hexane. As shown in Table 4,
3.1% NiOx-6.9% Ni2P/SiO2 presents the similar catalytic
performances for HDO of a series of phenolic compounds.
Table 2. Conversion and selectivity of phenol HDO[a] over NiOx/SiO2, Ni2P/SiO2, NiOx-Ni2P/SiO2 and literature data
Selectivity (%)
TOF[c](10-3 s-1)
Time
(h)
Temp
(°C)
P
Con
(%)
Yield of
ANE
Catalyst
NOL[b]
37.6
ANE[b]
ONE[b]
BZ[b]
1.3
NOL
ANE
1.03
(MPa)
0.68
0.82
2.34
10% Ni2P/SiO2
3
3
3
180
180
180
2
2
2
53.5
97.3
56.6
3.2
30.3
71.7
29.4
3.1% NiOx-6.9 % Ni2P/SiO2
6.7% NiOx-3.3% Ni2P/SiO2
26.5
70.5
73.7
29.4
0.5
0.0
0.3
0.2
2.28
0.98
100.0
2.85
10% NiOx/SiO2
3
3
3
3
2
3
7
3
180
180
180
180
200
180
220
180
2
2
2
2
4
5
4
4
100.0
3.2
90.8
45.2
31
8.2
26.8
31
0.3
19.3
4
0.4
8.7
33
4
8.2
0.9
0.26
10% Ni(H2PO2)2/ SiO2
0.04
0.03
[25]
Ni2P/SiO2
48
14.9
23
-
-
-
-
-
-
-
-
-
-
-
-
[25]
1%Pd-10%Ni2P/SiO2
92
71
25
0
Ni/CZn-Gl –SO3H[26]
Ru/substrate[27]
Ni–Mo–B-3[28]
100
100
94.9
78
50
25
5
-
12.5
11
89
11
-
-
86.2
3
9.1
52
0.8
33
3.8
8
8.6
Ni/HZSM-5[29]
40.6
[a] Reaction conditions: Phenol (0.25 g), dodecane (20 g), Cat (0.1 g), 2 MPa H2, 3 h, Stirring at 700 rpm. [b] NOL, ANE, ONE and BZ represent cyclohexanol,
cyclohexane, cyclohexanone, and benzene respectively. [c] TOF was calculated on the actual Ni content of the catalysts.
This article is protected by copyright. All rights reserved.

10.1002/cctc.201800010
ChemCatChem
FULL PAPER
Table 3. Catalytic performance[a] over catalysts 3.1% NiOx-6.9% Ni2P/SiO2
Ni(NO3)2·6H2O, 2.1 g urea and 110 g mL H2O was kept in a
beaker at room temperature, and diluted HNO3 solution was
used to adjust its pH value to 2.5. A solution of tetraethyl
orthosilicate and ethanol was drop wise added to the above
mixture under stirring for 4 h. Then it was heated to 363 K for 12
h to lead to the formation of gel. After calcination in a muffle
furnace, NiO/SiO2 was obtained. (2) The conventional
temperature programmed reduction method was to synthesize
NiOx/SiO2 catalysts. NiO/SiO2 sample was reduced under certain
conditions in H2 atmosphere and was passivated for 1 h in the
flow of 0.5% O2/N2 at room temperature. Thus NiOx/SiO2
catalysts obtained were denoted as NiOx/SiO2-X-Y. X and Y
indicate the reduction temperature and time.
for HDO reaction
Selectivity (%)
Con
(%)
Times
NOL
26.5
35.9
45.3
ANE
73.7
63.1
52.6
ONE
BZ
0.3
0.4
0.8
1 st (Fresh Catalyst)
97.3
100.0
96.0
0.5
0.6
1.6
2 nd
3 rd
[a] Reaction conditions: Phenol (0.25 g), dodecane (20 g), Cat (0.1 g), 180°C,
3 h, 2 MPa H2, Stirring at 700 rpm.
Table 4. Conversion and selectivity of phenol HDO[a] over 3.1% NiOx-6.9%
Ni2P/SiO2
10% Ni2P/SiO2 catalyst was prepared through incipient wetness
impregnation method and TPR method according to our
previous work.[16] NiO/SiO2 sample was impregnated with an
aqueous solution of NH4H2PO4. After drying at 383 K for 12 h,
the precursor (nominal Ni/P=1.0) was prepared. The precursor
was reduced at 550°C for 1 h in H2 atmosphere and was
passivated for 1 h in the flow of 0.5% O2/N2 at room
temperature. Thus the Ni2P/SiO2 catalyst was synthesized.
NiOx-Ni2P/SiO2 catalysts preparation process was similar to
Ni2P/SiO2. NiOx/SiO2 was impregnated with different
concentration of aqueous solution of NH4H2PO4. After drying at
383 K for 12 h, the different mole ratio of Ni/P of catalyst
precursors (i.e., Ni/P=2, Ni/P=4) was prepared. These
precursors were reduced at 823 K for certain reduce time in H2
atmosphere and were passivated for 1 h in the flow of 0.5%
O2/N2 at room temperature. Thus, silica-supported NiOx-
Ni2P/SiO2 catalysts were prepared, and were denoted as x%
NiOx-y% Ni2P /SiO2, in which x% and y% indicates the NiOx and
Ni2P mass fraction in the catalysts based on the same Ni loading,
10 wt.%. The catalyst evaluation tests have been repeated more
than twice. The experimental errors in conversion and selectivity
were within 5%.
For comparison, 10% Ni(H2PO2)2/SiO2 catalysts was also
prepared. The synthesis of this catalyst is similar to Ni2P/SiO2. In
a typical process, Firstly, SiO2 was prepared via a sol-gel
process. SiO2 was then impregnated with aqueous solution of
Ni(H2PO2)2•6H2O under stirring for 4 h at 300 K and dried at 373
K. The as-prepared sample was denoted as Ni(H2PO2)2/SiO2.
Considering that Ni(H2PO2)2 is easy to decomposed under high
temperature, that the as-obtained catalyst wasn’t calcined before
evaluation test of the catalyst.
Selectivity (%)
ANE[c]
Con.
(%)
reactant
NOL[b]
29.5
Others
5.4
Phenol
96.8
100
65.1
O-cresol
25.2
28.4
35.4
68.2
66.7
58.5
6.3
4.9
6.1
M-cresol
P-cresol
100
98.4
[a] Reaction conditions: Reactant (0.25 g), N-hexane (20 g), Cat (0.1 g),
180°C, 2 MPa H2, 3 h, Stirring at 700 rpm. [b] NOL represents o-
methylcyclohexanol and m-methylcyclohexanol, p-methylcyclohexanol. [c]
ANE represents methyl cyclohexane.
Conclusions
In this work, oxygen vacancies in metal oxide with and without
phosphide were investigated for phenolic compounds. Results
show that both NiOx/SiO2 and Ni2P/SiO2 have hydrogenation and
deoxygenation activity simultaneously. However, under relatively
mild reaction conditions, hydrogenation capacity outweighs
deoxygenation on NiOx because of flat adsorption of phenol over
low valence Ni species induced by oxygen vacancies, whereas
Ni2P is the opposite. When introducing oxygen vacancies into
Ni2P, NiOx-Ni2P/SiO2 presents an excellent activity due to the
synergistic effect of connected oxygen vacancies from NiOx and
acid induced by Ni2P in comparison with individual nickel oxide
and phosphide and most of results from literatures. Similar to
phenol, HDO results of different cresol isomers are also superior.
Stability experiment of NiOx-Ni2P/SiO2 indicates that the
deactivation of oxygen vacancies is slight in comparison with
phosphide. This work demonstrates an efficient and promising
catalyst for HDO in industrial application and provides helpful
information about developing catalysts with highly efficient
hydrodeoxygenation activity.
Catalyst characterizations
H2-TPR analyses were conducted out in a tubular quartz reactor
equipped with a thermal conductivity detector (TCD). For a
typical test, the reactor was loaded with 50 mg catalyst
precursor. The reduction was performed with a 10 vol.%H2/Ar2
flow (50 ml/min) at a rate of 5 K/min.
X-ray diffraction ( XRD ) measurement was carried out on a
Rigaku UItima IV with Cu Kα radiation ( λ = 0.15418 nm ) at 35
kV, 15 mA, a scan range from 10° to 90° and a scan rate of 5°
min-1. The crystallite size of sample was estimated by using the
Scherrer equation. Transmission electron microscopy (TEM)
images were obtained on JEM-2100.
Experimental Section
Element contents in the catalysts were detected by Inductively
Coupled Plasma Atomic Emission Spectrometry (ICP-AES),
which was carried out using ICP-MS-4500. Electron Spin
Resonance (ESR) was performed on Bruker EMX-10/12 to
Catalyst synthesis
10% NiOx/SiO2 catalysts synthesis was based on a two-step
procedure: (1) NiO/SiO2 was prepared by using a sol-gel
process. In a typical process, a homogeneous solution of 2.31 g
This article is protected by copyright. All rights reserved.

10.1002/cctc.201800010
ChemCatChem
FULL PAPER
[13] Y. Li, Q. Zhang, N. Zhang, L. Zhu, J. Zheng, B. Chen, Int. J. Hydrogen
Energ. 2013, 38, 13360-13367.
detect oxygen vacancies of the catalysts. The ESR peak areas
in the vicinity of g≈2.02 for catalysts were calibrated by standard
samples, in which the amount of oxygen vacancies is
determined.
[14] K. Yan, Y. Li, X. Zhang, X. Yang, N. Zhang, J. Zheng, B. Chen, K. J.
Smith, Int. J. Hydrogen Energ. 2015, 40, 16137-16146.
[15] Y. Li, J. Fu, B. Chen, RSC Adv. 2017, 7, 15272-15277.
[16] Y. Li, X. Zhang, H. Zhang, B. Chen, K. J. Smith, J. Taiwan Inst. Chem. E.
2017.
[17] A. Berenguer, T. M. Sankaranarayanan, G. Gómez, I. Moreno, J. M.
Coronado, P. Pizarro, D. P. Serrano, Green Chem. 2016, 18, 1938-1951.
[18] Y. Wang, W. Wu, Y. Rao, Z. Li, N. Tsubaki, M. Wu, J. Mater. Chem. A
2017, 5, 6170-6177.
[19] D. Liu, Y. Lv, M. Zhang, Y. Liu, Y. Zhu, R. Zong, Y. Zhu, J. Mater. Chem.
A 2014, 2, 15377-15388.
[20] X. Liu, L. Xu, B. Zhang, J. Solid State Chem. 2014, 212, 13-22.
[21] S. H. Lee, H. M. Cheong, Solid State Ionics 2001, 140, 135-139.
[22] J. Yang, S. Wang, X. Tang, Y. Wang, Y. Li, J. Supercrit. Fluid. 2018, 131,
1-10.
[23] H. Sun, X. Xu, Z. Yan, X. Chen, F. Cheng, P. S. Weiss, J. Chen, Chem.
Mater. 2017, 29, 8539-8547.
[24] H. Aiza, S. Tsuneyuki, Surf. Sci. lett. 1998, 399, 364-367.
[25] Y. Li, X. Yang, L. Zhu, H. Zhang, B. Chen, RSC Adv. 2015, 5, 80388-
80396.
[26] S. Kasakov, C. Zhao, E. Barath, Z. A. Chase, J. L. Fulton, D. M. Camaioni,
A. Vjunov, H. Shi, J. A. Lercher, Chem-Eur. J. 2015, 21, 1567-1577.
[27] J. Chen, J. Huang, L. Chen, L. Ma, T. Wang, U. I. Zakai, ChemCatChem.
2013, 5, 1598-1605.
The acidity of catalysts was detected by temperature
programmed desorption of ammonia (NH3-TPD) by the
semiautomatic Micromeritics TPD/TPR 2900 apparatus.
X-ray photoelectron spectroscopy (XPS) was conducted by a
Quantum 2000 Scanning ESCA Microprobe system which was
installed with a monochromatic Al Kα X-rays under ultrahigh
vacuum.
Pyridine-FTIR was collected on the Thermo Nicolet NEXUS
spectrometer. Self-supporting pellets made of the catalysts were
placed in the IR cell and were first reduced in H2/N2 mixture (10
vol.% of H2 ) flow at 723 K for 30 min. After cooling to room
temperature, the samples were saturated with pyridine vapor
and then evacuated at 373 K for 30 min. Spectra were recorded
at this temperature in the 4000-600 cm-1 range.
[28] W. Wang, Y. Yang, J. Bao, H. Luo, Catal. Commun. 2009, 11, 100-105.
[29] X. Zhang, T. Wang, L. Ma, Q. Zhang, T. Jiang, Bioresour Technol. 2013,
127, 306-31.
Hydrodeoxygenation evaluation
The hydrodeoxygenation evaluation reactions were carried out
in an autoclave, the liquid products were analyzed with a gas
chromatograph equipped with a FID detector. Then we utilize
the similar approach to evaluating other phenolic, such as p-
cresol, and o-cresol. For the sake of simplicity, cyclohexanol,
cyclohexane, cyclohexanone, and benzene were denoted as
NOL, ANE, ONE and BZ, respectively. Additionly, the turnover
frequency (TOF) over each supported catalyst was calculated on
the basis of the phenol conversion to NOL and ANE,
respectively, and the actual Ni content of the catalysts.
Acknowledgements
This work was supported by National Natural Science
Foundation of China (Grant No. 21476188, 21106118). We
would like to appreciate Sijing Chen and Sajid Ali’s contribution
to the preparation of the manuscript.
Conflicts of interest
There are no conflicts to declare.
Keywords: Oxygen Vacancies
•
Hydrodeoxygenation
•
Oxygenate Compounds • Nickel phosphide
[1] A. V. Bridgwater, Biomass Bioenerg. 2012, 38, 68-94.
[2] M. Garcia-Perez, A. Chaala, H. Pakdel, D. Kretschmer, C. Roy, Biomass
Bioenerg. 2007, 31, 222-242.
[3] S. T. Oyama, J. Catal. 2003, 216, 343-352.
[4] K. Li, R. Wang, J. Chen, Energ. Fuel. 2011, 25, 854-863.
[5] L. F. Feitosa, G. Berhault, D. Laurenti, T. E. Davies, V. T. D. Silva, J. Catal.
2016, 340, 154-165.
[6] J. Chen, Y. Chen, Q. Yang, K. Li, C. Yao, Catal. Commun. 2010, 11, 571-
575.
[7] T. N. Phan, C. H. Ko, Catal. Today 2018, 303, 219-226.
[8] V. O. O. Gonçalves, C. Ciotonea, S. A. Clacens, N. Guignard, C. Roudaut,
J. Rousseau, J. M. Clacens, S. Royer, F. Richard, Appl. Catal. B-Environ.
2017, 214, 57-66.
[9] S. M. Schimming, O. D. LaMont, M. Konig, A. K. Rogers, A. D. D'Amico, M.
M. Yung, C. Sievers, ChemSusChem. 2015, 8, 2073-2083.
[10] S. Foraita, J. L. Fulton, Z. A. Chase, A. Vjunov, P. Xu, E. Barath, D. M.
Camaioni, C. Zhao, J. A. Lercher, Chem-Eur. J. 2015, 21, 2423-2434.
[11] T. Prasomsri, T. Nimmanwudipong, Y. Román-Leshkov, Energ. Environ.
Sci. 2013, 6, 1732-1738.
[12] A. V. Mironenko, D. G. Vlachos, J. Am. Chem. Soi. 2016, 138, 8104-8113.
This article is protected by copyright. All rights reserved.

10.1002/cctc.201800010
ChemCatChem
FULL PAPER
Entry for the Table of Contents
FULL PAPER
Synergetic catalysis of NiOx with oxygen vacancies and Ni2P for
hydrodeoxygenation of phenolic compounds
Yunhua Li,*, [a] Yafei Zhao, [a] Binghui Chen, [a], Wenju Wang†, [b]
[a]
Associate professor Yunhua Li
Department of Chemical and Biochemical Engineering, College of
Chemistry and Chemical Engineering
Xiamen University
Xiamen, 361005, PR China
E-mail: yunhuali@xmu.edu.cn
[b]
School of Energy and Power Engineering, Nanjing University of
Science and Technology, Nanjing 210094, China
E-mail: wangwenju1982@gmail.com
This article is protected by copyright. All rights reserved.