In situhydrodeoxygenation of vanillin over Ni-Co-P/HAP with formic acid as a hydrogen source

Article abstract of DOI:10.1039/d1ra00979f

A new noble metal-free Ni-Co-P/HAP (hydroxyapatite) amorphous alloy catalyst was synthesized by an impregnation-chemical reduction method; the structure and properties of the catalysts were characterized by XRD, SEM, BET, XPS and DSC. Based on the model of the hydrodeoxygenation (HDO) of vanillin to 2-methoxy-4-methylphenol (MMP) with formic acid as a hydrogen source, the catalytic performance of the catalyst was studied. The results found that the Ni-Co-P/HAP catalyst exhibited excellent catalytic activity for thein situHDO reaction of vanillin compared with Ni-P and Ni-Co-P. The conversion of vanillin could be high to 97.86% with MMP selectivity of 93.97% under optimized reaction conditions. In addition, mechanism studies have shown that the side reaction of carbocation and vanillyl alcohol (HMP) condensation can be effectively reduced with increasing the hydrogenation rate, thereby the selectivity of MMP was effectively increased.

Full text of DOI:10.1039/d1ra00979f

RSC Advances
View Article Online
View Journal | View Issue
PAPER
In situ hydrodeoxygenation of vanillin over Ni–Co–
P/HAP with formic acid as a hydrogen source
Cite this: RSC Adv., 2021, 11, 10996
Mingxing Duan, Qingyan Cheng, * Mingming Wang and Yanji Wang
A new noble metal-free Ni–Co–P/HAP (hydroxyapatite) amorphous alloy catalyst was synthesized by an
impregnation-chemical reduction method; the structure and properties of the catalysts were
characterized by XRD, SEM, BET, XPS and DSC. Based on the model of the hydrodeoxygenation (HDO)
of vanillin to 2-methoxy-4-methylphenol (MMP) with formic acid as a hydrogen source, the catalytic
performance of the catalyst was studied. The results found that the Ni–Co–P/HAP catalyst exhibited
excellent catalytic activity for the in situ HDO reaction of vanillin compared with Ni–P and Ni–Co–P. The
conversion of vanillin could be high to 97.86% with MMP selectivity of 93.97% under optimized reaction
conditions. In addition, mechanism studies have shown that the side reaction of carbocation and vanillyl
alcohol (HMP) condensation can be effectively reduced with increasing the hydrogenation rate, thereby
the selectivity of MMP was effectively increased.
Received 5th February 2021
Accepted 3rd March 2021
DOI: 10.1039/d1ra00979f
rsc.li/rsc-advances
1
4
R. et al. studied the HDO reaction of vanillin and found that
the vanillin conversion and the selectivity of the product 2-
Introduction
With the rapid increase of the global population, the methoxy-4-methylphenol (MMP) were both above 90% with
consumption of petrochemical resources has become increas- N-doped carbon-supported Pd as the catalyst and formic acid
ingly serious, the energy demand has risen rapidly, and the as a hydrogen donor. Although noble metal catalysts have
supply of fossil fuels is insufficient, which makes the develop- a good catalytic effect on biomass oil HDO, the low reserves
1,2
ment of new energy sources imminent. The pyrolysis of and high cost of noble metals limit its large-scale industrial
lignocellulosic biomass is a prospective approach for producing production. Nickel-based amorphous alloy catalysts have
valuable liquid biomass oil. As an effective strategy, hydro- received widespread attention because of their low price,
deoxygenation (HDO) is of vital importance for improving the simple preparation process, and excellent catalytic perfor-
17
1
5,16
energy density of bio-oil by selectively removing oxygen- mance in the HDO reaction of vanillin.
Brandi et al.
3,4
containing groups. The traditional HDO technology requires synthesized a stable Ni catalyst and loaded it on nitrogen-
explosive fossil resource-derived hydrogen gas that is dangerous doped carbon (NDC). The catalytic performance of the cata-
and uncontrollable, oen leading to excessive hydrogenation lyst was evaluated in the aqueous hydrogenation reaction of
and C–C cleavage. Studies have found that as a new type of vanillin. Studies have found that the Ni/NDC catalyst exhibits
hydrogen storage material, formic acid has good stability, is high catalytic performance in the aqueous hydrogenation of
easy to store and transport, and has a large hydrogen capacity, vanillin with hydrogen as the hydrogen source, and the
and has been widely used as a hydrogen source in hydrogena- conversion of vanillin and the selectivity of product MMP are
5,6
7–10
tion reactions.
close to 100%.
In recent years, supported precious metal catalysts have
Although Ni-based amorphous alloy catalysts have exhibited
been used for the decomposition of formic acid and the excellent catalytic activity in the hydrogenation reactions, it still
hydrodeoxygenation of biomass oil.
2-Methoxy-4- has problems with small specic surface area, easy agglomer-
methylphenol (MMP) is an important intermediate for ation, and poor thermal stability, etc. Studies have shown that
medicine and perfume, which can be obtained by the the degree of disorder of the Ni-based amorphous alloy can be
hydrodeoxygenation of vanillin. Meanwhile, the HDO reac- increased and the agglomeration of the catalyst can be reduced
tion of vanillin was oen used as a probe reaction to inves- by introducing the transition metal Co. Meanwhile, as a mate-
1
1–13
tigate the hydrogenation performance of the catalyst.
Nie rial with high reserves, low cost and high activity, phosphorus
can form a chelated structure with nickel in Ni–Co–P amor-
phous alloy, thereby improving the stability of the catalyst.
There is a large difference between the size of P atom and
School of Chemical Engineering, Hebei University of Technology, Key Laboratory of
Green Chemical Technology and High Efficient Energy Saving of Hebei Province, transition metal Ni, Co atom. The defect sites of the catalyst can
Tianjin Key Laboratory of Chemical Process Safety, Tianjin, 300130, China. E-mail:
be easily produced when P was incorporated into the material
chengqingyan@hebut.edu.cn
10996 | RSC Adv., 2021, 11, 10996–11003
© 2021 The Author(s). Published by the Royal Society of Chemistry

View Article Online
Paper
RSC Advances
ꢀ
structure framework, which is benecial to the formation of ethanol three times, and dried in vacuum at 50 C for 8 h to
amorphous alloy structure. In addition, the specic surface area obtain a Ni–P amorphous alloy catalyst. The preparation
and stability of the catalyst can be improved by supporting the process of Ni–Co–P amorphous alloy is similar to the above
amorphous alloy on the carrier, which enables the activity of the process. NiCl $6H O (2.5 mmol, 0.5942 g), NaH PO $H O
2
2
2
2
2
catalyst to be improved. Among many carriers, HAP was widely (20 mmol, 1.060 g), CH
3 2
COONa$3H O (1.25 mmol, 0.1701 g),
used as a catalyst carrier due to its mesoporous structure, large Co(NO $6H O (2 mmol, 0.5821 g) and deionized water (30
3
)
2
2
specic surface area, high porosity and good thermal stability. mL) were added to a 150 mL three-necked ask, the other steps
Moreover, although Ni-based amorphous alloy catalysts have are the same as the preparation of Ni–P amorphous alloy
good catalytic performance in HDO reactions using hydrogen as catalyst.
the hydrogen source, there are few reports on the use of formic
acid as the hydrogen source. Therefore, it is of great signicance Synthesis of supported Ni–Co–P/HAP amorphous alloy
for the purication of biomass oil and large-scale industrial
The Ni–Co–P/HAP amorphous alloy was prepared by
production through the development of cheap, efficient, and
impregnation-chemical reduction method. NiCl $6H O
2
2
stable Ni-based amorphous alloy catalysts to catalyze the
decomposition of formic acid to produce hydrogen and further
catalyze the HDO reaction of vanillin.
(
2.5 mmol, 0.5942 g), NaH PO $H O (20 mmol, 1.060 g), CH -
2 2 2 3
COONa$3H O (1.25 mmol, 0.1701 g), Co(NO ) $6H O (2 mmol,
2
3 2
2
0
.5821 g) and deionized water (30 mL) were added to the three-
necked ask, adjusted the pH of solution to 9.0 with 25 wt%
NH $H O, followed by HAP (0.3 g) was added to the above
solution. The reaction solution was magnetically stirred at room
In this work, the Ni–Co–P/HAP (hydroxyapatite) amor-
phous alloy catalyst was prepared by the impregnation-
chemical reduction method. The catalytic performance of
Ni–Co–P/HAP in the in situ HDO reaction of vanillin was
studied with formic acid as the hydrogen source, which
provides a new method for the rening of biomass oil.
Meanwhile, the effect of hydrogen protons on the selectivity
of MMP was studied, the mechanism of the vanillin HDO
reaction catalyzed by Ni–Co–P/HAP with formic acid as the
hydrogen source was proposed.
3
2
temperature for 12 h. Then, a few drops of KBH
4
solution
ꢁ1
(3 mol L ) were added slowly to the above solution to induce
ꢀ
the reaction to occur. The reaction was carried out at 30 C for
a certain period of time, and the reaction was stopped when no
bubbles are formed. The product was separated by centrifuga-
tion, washed with deionized water to neutrality, then washed
ꢀ
with anhydrous ethanol three times, and vacuum dried at 50 C
for 8 h to obtain Ni–Co–P/HAP amorphous alloy.
Experimental
Materials
Catalyst characterization
Powder X-ray diffraction (XRD) patterns were recorded on a D/
MAX-2500 X-ray diffractometer from Nippon Science, Inc.,
with Cu Ka rays at a scan rate of 8 min , tube voltage and
NiCl
2 2 3 2 2 2 2 2 4
$6H O, Co(NO ) $6H O, NaH PO $H O, KBH and 25 wt%
NH $H O were purchased from Sailboat Chemical Reagent
ꢀ
ꢁ1
3
2
Technology Co., Ltd. (Tianjin, China). Vanillin, vanillyl alcohol
current of 40 kV and 100 mA respectively, and a scan range of
(
(
HMP), 2-methoxy-4-methylphenol (MMP) and hydroxyapatite
Ca10(PO (OH) HAP) were purchased from Aladdin
ꢀ
2
0–70 . The surface morphology of the catalyst was character-
4
)
6
2
,
ized by FEI Nova SEM-450 scanning electron microscope (SEM)
from Hong Kong Co., Ltd. Specic surface area and porosity
adsorption (BET) were collected using an ASAP 2020 specic
surface area and porosity analyzer from Micromeritics
Biochemical Technology Co., Ltd. (Shanghai, China). CH
COONa$3H O, ethanol and isopropanol were purchased from
Reagent No. 1 Factory (Tianjin, China). N was purchased from
Sizhi Gas Co., Ltd. (99.999%, Tianjin, China). All chemicals are
analytical pure and used as received without further
purication.
3
-
2
2
Company, USA, using N
2
as adsorbate and He as carrier gas, the
ꢀ
samples were vacuum degassed at 90 C for 8 h. The X-ray
photoelectron spectroscopy (XPS) was carried out with ESCA-
LAB 250Xi from Thermo Fisher Scientic Co., Ltd., and Al Ka
(1486.6 eV) rays were used as the excitation source. The thermal
stability of the catalyst was investigated by SDT Q-600 differ-
ential scanning calorimeter (DSC) from TA Instruments, USA,
Synthesis of Ni–P and Ni–Co–P amorphous alloys
The Ni–P and Ni–Co–P amorphous alloys were prepared by the
chemical reduction method. NiCl $6H O (2.5 mmol, 0.5942 g),
NaH PO $H O (20 mmol, 1.060 g), CH COONa$3H O
2
2
ꢁ1
under N
a heating rate of 10 C min
2
atmosphere, with a ow rate of 100 mL min and
2
2
2
3
2
ꢀ
ꢁ1
.
(
1.25 mmol, 0.1701 g) and deionized water (30 mL) were added
to the three-necked ask, adjusted the pH of the solution to 9.0
ꢀ
Formic acid dehydrogenation
with 25 wt% NH
3
$H
2
O at 30 C. Under the protection of N
2
,
ꢁ1
a few drops of KBH
4
solution (3 mol L ) were added dropwise Formic acid dehydrogenation reactions were conducted in
to the solution to induce the reaction to occur, during which a 100 mL stainless steel autoclave with a magnetic stirrer to
a black precipitate was formed and accompanied by bubbles. evaluate the catalytic activity of the Ni–Co–P/HAP in the
ꢀ
The reaction was carried out at 30 C for a certain period of decomposition of formic acid. Formic acid (HCOOH, 0.6086 g),
time, and the reaction was stopped when no bubbles are catalyst (0.05 g) and isopropanol solvent (20 mL) were added to
formed. The product was separated by centrifugation, washed the autoclave, sealed the autoclave, and replaced the air in the
with deionized water to neutrality, washed with anhydrous autoclave with nitrogen (99.999%) three times. Heat to the
©
2021 The Author(s). Published by the Royal Society of Chemistry
RSC Adv., 2021, 11, 10996–11003 | 10997

View Article Online
RSC Advances
Paper
ꢀ
ꢁ1
required temperature at a speed of 10 C min , and react for
a period of time while stirring at a speed of 600 rpm. Aer the
reaction, the autoclave was quickly cooled to room temperature,
and the total volume of the H and CO products produced was
2
2
measured. A gas chromatograph (SP-3420A) equipped with
a thermal conductivity detector (TCD) and a hydrogen ame
ionization detector (FID) was used to determine the gas and
liquid products.
The HDO reaction of vanillin
Vanillin (2 mmol, 0.3043 g), formic acid (0.6086 g), catalyst (0.05
g) and 20 mL isopropanol were added to the autoclave. Seal the
autoclave and replace the air in the autoclave with nitrogen
(
99.999%) three times. Heat to the required temperature at
Fig. 2 SEM images of (a) Ni–P, (b) Ni–Co–P, (c) Ni–Co–P/HAP and (d)
HAP.
ꢀ
ꢁ1
a speed of 10 C min , and react for a period of time while
stirring at a speed of 600 rpm. Aer the reaction, it was cooled to
room temperature. The product was quantitatively analyzed
using high performance liquid chromatography (HPLC) from
Waters, USA.
intensity was obviously weakened, which indicates that Ni–Co–
19
P amorphous alloy has been loaded on HAP.
The SEM of Ni–P, Ni–Co–P, Ni–Co–P/HAP and HAP are
shown in Fig. 2. It can be seen from Fig. 2 that Ni–P and Ni–Co–
P amorphous alloys were spherical particles with partial defects
on the surface. The grain size of Ni–P amorphous alloy is not
Results and discussion
Catalysts characterization
The XRD spectra of Ni–P, Ni–Co–P, Ni–Co–P/HAP and HAP are uniform, in the range of 100 to 200 nm, accompanied by
shown in Fig. 1. As can be seen from Fig. 1, the broad dispersion a certain degree of agglomeration. By comparing with Ni–P, it
diffraction peaks of Ni–P and Ni–Co–P both appeared around at can be seen that the particles of the Ni–Co–P and Ni–Co–P/HAP
ꢀ
2
q ¼ 45 , which are the characteristic diffraction peaks of the amorphous alloy modied by Co were relatively loose, which
amorphous alloy formed by metal Ni and metalloid P, indi- indicates that the agglomeration of amorphous alloy particles
16,18
cating that the samples present the amorphous structure.
was effectively suppressed when the auxiliary Co was added
Compared with Ni–P amorphous alloys, the characteristic during the preparation of catalyst. Comparing Fig. 2(c) and (d),
diffraction peaks of Ni–Co–P amorphous alloys are more it can be seen that the Ni–Co–P amorphous alloy forms spher-
ꢀ
dispersed around 2q ¼ 45 , which indicates that the introduc- ical particles with a diameter of about 80 nm, which are
tion of Co is benecial to the formation of amorphous struc- dispersed on the HAP.
tures. At the same time, compared with Ni–Co–P, the diffraction
peak of Ni–Co–P/HAP near 2q ¼ 45 was wider, and the peak areas of Ni–P and Ni–Co–P are basically the same, which is 8.28
The BET results in Table 1 show that the specic surface
ꢀ
2
ꢁ1
and 7.97 m g , respectively. The pore diameter of the amor-
phous alloys Ni–P and Ni–Co–P is 10.64 and 10.45 nm, respec-
3
ꢁ1
tively, the pore volume is 0.0282 and 0.0277 cm
g
respectively, which is also basically the same. Furthermore, HAP
has a large specic surface area, pore size and pore volume,
2
ꢁ1
3
ꢁ1
which is 65.94 m g , 22.18 nm and 0.4036 cm g , respec-
tively, so HAP can be used as a good carrier for Ni–Co–P. When
Ni–Co–P was loaded on HAP, the specic surface area of Ni–Co–
2
ꢁ1
P/HAP was reduced to 51.27 m g , and the pore size and pore
volume were reduced to 20.68 nm and 0.3372 cm
3
ꢁ1
g ,
Table 1 BET characterization results of Ni–P, Ni–Co–P, Ni–Co–P/
HAP and HAP Samples
2
ꢁ1
3
ꢁ1
Catalyst
S
BET/m g
Pore size/nm
Pore volume/cm g
Ni–P
Ni–Co–P
Ni–Co–P/HAP
8.28
7.97
51.27
65.94
10.64
10.45
20.68
22.18
0.0282
0.0277
0.3372
0.4036
Fig. 1 XRD patterns of (a) Ni–P, (b) Ni–Co–P, (c) Ni–Co–P/HAP and HAP
d) HAP.
(
10998 | RSC Adv., 2021, 11, 10996–11003
© 2021 The Author(s). Published by the Royal Society of Chemistry

View Article Online
Paper
RSC Advances
0
addition, the peak height of P in Ni–Co–P/HAP was sharply
0
decreased compared with the p of Ni–P and Ni–Co–P, which is
3
+
because there is a strong and broad peak attributable to P in
HAP. According to related reports, the standard binding
0
0
energies (BEs) of Ni are 853.0 and 869.9 eV, and the BEs of Co
2
1,27
0
are 777.8 eV and 796.7 ev.
It can be seen that the BEs of Ni ,
0
Co of the catalysts were decreased, compared with the stan-
dard BEs, which indicates that the electrons have been trans-
ferred, that is, electrons are transferred from P to Ni, which
result in the Ni of catalyst in an electron-rich state. According
to relevant literature reports that the electron-rich Ni exhibits
an excellent ability in adsorbing hydrogen and can better
activate hydrogen, which is benecial to the hydrogenation
2
8
reaction.
Fig. 3 XPS spectra of Ni 2p, Co 2p and P 2p of Ni–P, Ni–Co–P and Ni–
Co–P/HAP samples (a) Ni 2p (b) Co 2p and (c) P 2p.
The DSC curves of Ni–P, Ni–Co–P and Ni–Co–P/HAP are
shown in Fig. 4. It can be seen from Fig. 4 that Ni–P and Ni–
Co–P amorphous alloys appeared exothermic peaks near
ꢀ ꢀ
3
65 C and 402 C, respectively, which are typical crystalliza-
respectively, which shows that Ni–Co–P was successfully loaded
on HAP.
The XPS spectra of Ni 2p, Co 2p and P 2p of Ni–P, Ni–Co–P
and Ni–Co–P/HAP amorphous alloys are shown in Fig. 3. It can
be seen from Fig. 3(a) that the Ni 2p of the three amorphous
alloys all appeared six peaks near 852, 855, 861, 869, 873 and
tion exothermic peaks of amorphous alloys. Compared with
Ni–P amorphous alloy, the crystallization temperature of Ni–
Co–P amorphous alloy was signicantly higher, which indi-
cates that the introduction of Co can improve the thermal
stability of the amorphous structure. Comparing the DSC
curves of Ni–Co–P and Ni–Co–P/HAP, it can be seen that Ni–
Co–P/HAP has higher thermal stability, which is also the key to
the better catalytic performance of the Ni–Co–P/HAP in the
HDO reaction of vanillin. Meanwhile, it can be seen from the
SEM characterization (Fig. 2(c)) that the Ni–Co–P amorphous
alloy supported on HAP is benecial to the dispersion of the
Ni–Co–P amorphous alloy catalyst, which can effectively
prevent the agglomeration or crystallization of amorphous
alloys. This is also the main reason why the stability of Ni–Co–
P/HAP amorphous alloy was signicantly improved compared
8
79 eV, among which the peaks near 852, 869 eV were attrib-
0
2+
uted to Ni , and the remaining peaks are attributed to Ni .
The oxidation state of Ni may originate from the NiO formed
during the preparation of the catalyst,
2
0–22
and the presence of
2
3,24
NiO can catalyze the decomposition of formic acid.
The
rapid decomposition of formic acid is benecial to the
hydrodeoxygenation of vanillin. The XPS of the three catalysts
0
Co 2p is shown in Fig. 3(b). It can be seen that the peaks of Co
in Co 2p appear about 777 and 796 eV. The XPS spectrum of P
2
p is shown in Fig. 3(c), in terms of Ni–P, Ni–Co–P and Ni–Co–
2
9
with Ni–P and Ni–Co–P.
0
P/HAP, the peaks of P appear around 129 and 130 eV, while
the peaks near 132 eV were attributed to P , which is derived
from H
3
+
ꢁ
25,26
Decomposition of formic acid
2
PO
3
generated during the preparation process.
In
Under the reaction conditions of 20 mL isopropanol, 0.6086 g
ꢀ
formic acid, 0.05 g catalyst, and reaction temperature of 190 C,
Fig. 5 Catalytic decomposition of formic acid with Ni–Co–P/HAP
Fig. 4 DSC curves of (a) Ni–P, (b) Ni–Co–P and (c) Ni–Co–P/HAP.
catalyst.
©
2021 The Author(s). Published by the Royal Society of Chemistry
RSC Adv., 2021, 11, 10996–11003 | 10999

View Article Online
RSC Advances
Paper
a
Table 2 The HDO reaction of vanillin catalyzed by different catalysts
acid can release hydrogen through the dehydrogenation reac-
tion (HCOOH / H + CO ), and the process is usually accom-
panied by an undesirable dehydration pathway (HCOOH /
2
2
Catalyst
Xvanillin (%)
SMMP (%)
SHMP (%)
H O + CO). However, CO and acetone were not detected by gas
None
HAP
Ni–P
Ni–Co–P
Ni–Co–P/HAP
<1
<1
76.88
82.96
84.23
None
None
50.41
61.72
65.36
None
None
16.22
7.32
2
chromatography, which conrms that formic acid was the main
hydrogen source of the reaction and no side reactions occurred
in the dehydrogenation reaction of formic acid, while iso-
propanol was only used as a solvent in the hydrogenation
6.42
a
19,31
Reaction conditions: 2 mmol vanillin, catalyst 0.05 g, solvent 20 mL reaction of vanillin.
In addition, the maximum total volume
ꢀ
isopropanol, m(formic acid) : m(vanillin) ¼ 2 : 1, 190 C, 3 h.
of H and CO produced by the decomposition of formic acid is
2
2
6
9
35 mL, and the conversion of formic acid is calculated at
8.16%.
the catalytic performance of Ni–Co–P/HAP amorphous alloy for
the decomposition of formic acid was investigated. The result is
shown in Fig. 5, in the absence of a catalyst, the total volume of
Evaluation of the activity of the HDO reaction of vanillin
The HDO reaction of vanillin was chosen as a model reaction
to investigate the catalytic performance of catalysts. The effect
of the catalyst on the HDO reaction of vanillin is shown in
Table 2. It can be seen from Table 2 that the HDO reaction of
vanillin cannot proceed when no catalyst was added or only
carrier HAP was added. The activity of the catalyst was
signicantly improved when using Ni–P amorphous alloy as
the catalyst, and the conversion of vanillin reached 76.88%,
accompanied by 50.41% MMP and 16.22% vanillin alcohol
2 2
H and CO (625 mL) remained unchanged aer 100 minutes of
formic acid decomposition reaction. The rate of formic acid
dehydrogenation was increased signicantly aer introducing
the catalyst into the reaction system. The dehydrogenation of
formic acid was completed when the reaction time reaches 70
minutes, the reason is that the decomposition of formic acid
can be catalyzed effectively by Ni–Co–P/HAP amorphous alloy
18,30
catalyst, which is consistent with the previous reports.
In
addition, acetone are usually produced as by-products in the
reaction system when isopropanol as a hydrogen source. Formic
(
HMP) selectivity. Compared with Ni–P, the selectivity of MMP
was increased to 61.72% when using Ni–Co–P amorphous
Fig. 6 Influence of (a) formic acid amount (b) reaction temperature (c) reaction time and (d) catalyst amount on vanillin HDO reaction.
11000 | RSC Adv., 2021, 11, 10996–11003
© 2021 The Author(s). Published by the Royal Society of Chemistry

View Article Online
Paper
RSC Advances
alloy as the catalyst, and the selectivity of HMP was decreased
to 7.32%, which indicates that the introduction of Co can
improve the activity and selectivity of the catalyst, and
promote the hydrogenation of HMP to MMP. Meanwhile, the
activity of the catalyst was slightly increased, the selectivity of
HMP was further reduced, while the selectivity of MMP was
increased to 65.36% when using Ni–Co–P/HAP amorphous
alloy as the catalyst. This is because the Ni–Co–P amorphous
alloy was uniformly dispersed, the agglomeration of the cata-
lyst was reduced and the stability of the catalyst was improved
when Ni–Co–P was loaded on the HAP, thereby the activity of
the catalyst was increased.
The effect of the amount of formic acid, reaction temper-
ature, reaction time and the amount of catalyst on the HDO
reaction of vanillin were investigated, the results were shown
in Fig. 6. Here, the effect of the amount of formic acid on the
vanillin HDO reaction was shown in Fig. 6(a), under the
reaction conditions of 2 mmol vanillin, 0.05 g catalyst, 20 mL
Fig. 7 Mechanism proposed for the HDO reaction of vanillin using
formic acid as the hydrogen source.
ꢀ
isopropanol, 190 C and 3 h. It can be seen from Fig. 6(a) that
in the initial stage of the reaction, due to the small amount of
formic acid in the system, the decomposition of formic acid
cannot provide sufficient hydrogen, and HMP cannot be
hydrogenated in time to generate MMP, resulting in the side
was almost completely converted at 300 min, the conversion
rate of vanillin was 96.79%, the selectivity of the target product
MMP was 81.47%, and they remained unchanged as the reac-
tion time was further increased. Fig. 6(d) shows the effect of the
amount of Ni–Co–P/HAP amorphous alloy catalyst on the
vanillin HDO reaction under the reaction conditions of 2 mmol
3
2,33
reaction of aldol condensation,
and the selectivity of the
product MMP was 49.55%. As the amount of formic acid
increases, the hydrogen source in the reaction system
increased, the selectivity of MMP was increased signicantly.
The highest selectivity of product MMP was 65.36% when
m(formic acid) : m(vanillin) ¼ 2 : 1. When the amount of
formic acid was further increased, the selectivity of MMP was
decreased. This is because hydrogen protons will combine
with formic acid to generate carbocations when there is too
much formic acid in the reaction system, which intensies
vanillin, 20 mL isopropanol, m(formic acid) : m(vanillin) ¼ 2 : 1,
ꢀ
2
00 C and 300 min. It can be seen from Fig. 6(d) that the
conversion of vanillin was increased slightly from 93.28% to
7.86%, while the selectivity of MMP was increased signicantly
9
from 60.89% to 93.97% when the increase of the amount of Ni–
Co–P/HAP catalyst from 0.04 g to 0.06 g. This is because the Ni–
Co–P/HAP amorphous alloy plays a dual role in the HDO reac-
tion of vanillin: on the one hand, the NiO component in Ni–Co–
P/HAP has the function of catalyzing the decomposition of
3
3
the occurrence of HMP coupling side reactions. Fig. 6(b)
presents the effect of reaction temperature on the HDO
reactions of vanillin under the reaction conditions of 2 mmol
vanillin, 0.05 g catalyst, 20 mL isopropanol, m(formic acid)-
formic acid to obtain H ; on the other hand, Ni–Co–P/HAP can
2
promote the HDO reaction of vanillin. Therefore, the active
component of the catalyst is less than enough to catalyze the
decomposition of formic acid and the vanillin HDO reaction
when the amount of the catalyst is too small, the formic acid
cannot provide a sufficient hydrogen source for the vanillin
HDO, which leads to low selectivity of the MMP.
:
m(vanillin) ¼ 2 : 1 and 3 h. As shown in Fig. 6(b), the best
ꢀ
results were achieved under 200 C, which showed a 94.48%
conversion of vanillin and 72.22% selectivity of MMP. When
the temperature further increased, although the conversion
of vanillin slightly increased, the selectivity of the product
MMP was signicantly decreased. The reason is that the HDO
reaction of vanillin is an exothermic reaction, but formic acid
Mechanism study
decomposition to H
2
is an endothermic reaction, the rise of Based on experimental studies and literature reports, a plau-
temperature is benecial to the decomposition of formic acid sible reaction mechanism for the HDO reaction of vanillin with
35–37
to produce hydrogen, but the too high temperature will cause formic acid as a hydrogen source has been proposed (Fig. 7).
the coupling reaction of MMP which the selectivity of MMP First, formic acid can generate hydrogen protons, the carbonyl
3
4
was decreased.
group can be activated by hydrogen protons to form carboca-
In addition, under the reaction conditions of 2 mmol tions (Fig. 7, reaction a), and the carbocations undergoes
vanillin, 0.05 g catalyst, 20 mL isopropanol, m(formic acid)- condensation side reactions with HMP, which is not conducive
ꢀ
:
m(vanillin) ¼ 2 : 1 and 200 C, the effect of the reaction time to the further hydrogenation of HMP to produce MMP, which is
on the vanillin HDO reaction was investigated. It can be seen also the main reason why the selectivity of the product MMP
from Fig. 6(c) that in the initial reaction stage, the reaction was decreased a when there is too much formic acid in the
proceeded slowly and the reaction was incomplete. As the reaction system. Secondly, formate decomposes on the surface
ꢁ
reaction time increases, the conversion of vanillin and the of NiO to form CO
2
and [ONi–H] intermediates (Fig. 7, reac-
ꢁ
+
selectivity of the product MMP was increased gradually. Vanillin tions b and c), [ONi–H] intermediates combine with H ions to
©
2021 The Author(s). Published by the Royal Society of Chemistry
RSC Adv., 2021, 11, 10996–11003 | 11001

View Article Online
RSC Advances
form H and NiO (Fig. 7, reaction d), at this time, the NiO
Paper
2
8 D. A. Bulushev, S. Beloshapkin, P. E. Plyusnin, Y. V. Shubin,
V. I. Bukhtiyarov, S. V. Korenev and J. R. Ross, J. Catal., 2013,
299, 171–180.
9 M. Grasemann and G. Laurenczy, Energy Environ. Sci., 2012,
5, 8171–8181.
38
catalytic cycle is completed, returning to the initial state. The
H generated by the decomposition of formic acid was activated
2
0
by Ni and hydrogenated with vanillin to form HMP (Fig. 7,
reaction e), and HMP was further hydrogenated to form MMP
39,40
(Fig. 7, reaction f).
The experimental results showed that the 10 C. Hu, J. K. Pulleri, S. W. Ting and K. Y. Chan, Int. J. Hydrogen
condensation reaction could be greatly reduced (Fig. 7, reaction
g) when the hydrogenation reaction proceeded rapidly (Fig. 7, 11 M. O. Bengoechea, A. Hertzberg, N. Mileti ´c , P. L. Arias and
reaction e and f). Under the premise that hydrogenation and T. Barth, J. Anal. Appl. Pyrolysis, 2015, 113, 713–722.
condensation are competing reactions, improving the hydro- 12 C. Pu, J. Zhang, G. Chang, Y. Xiao, X. Ma, J. Wu and X. Yang,
Energy, 2014, 39, 381–390.
genation efficiency is an effective way to reduce condensation
side reactions.
Carbon, 2020, 159, 451–460.
13 L. Wang, B. Zhang, X. Meng, D. S. Su and F. S. Xiao,
ChemSusChem, 2014, 7, 1537–1541.
41,42
1
4 R. Nie, X. Peng, H. Zhang, X. Yu, X. Lu, D. Zhou and Q. Xia,
Catal. Sci. Technol., 2017, 7, 627–634.
Conclusions
Herein, the Ni–Co–P/HAP amorphous alloy catalyst was 15 A. Jia, X. Yao, L. Feng, Z. Ma, F. Li and Y. Wang, Eur. J. Inorg.
synthesized by impregnation-chemical reduction method, and Chem., 2020, 13, 1184–1191.
it exhibits good catalytic activity for in situ HDO reaction of 16 W. Wang, Y. Yang, H. Luo and W. Liu, React. Kinet., Mech.
vanillin with formic acid as the hydrogen source. The conver- Catal., 2010, 101, 105–115.
sion of vanillin could be high to 97.86% with MMP selectivity of 17 F. Brandi, M. B ¨a umel, V. Molinari, I. Shekova, I. Lauermann,
3.97%, under optimized reaction conditions of 2 mmol T. Heil and M. Al-Naji, Green Chem., 2020, 22, 2755–2766.
9
vanillin, 0.06 g catalyst, 20 mL isopropanol solvent, m(formic 18 M. H. Qiao, S. H. Xie, W. L. Dai and J. F. Deng, Catal. Lett.,
ꢀ
acid) : m(vanillin) ¼ 2 : 1, 200 C and 300 min. Ni–Co–P/HAP, as
2001, 71, 187–192.
a dual-function catalyst, can promote the decomposition of 19 G. Xie, W. Sun and W. Li, Catal. Commun., 2008, 10, 333–335.
formic acid and catalyze the in situ HDO reaction of vanillin, 20 M. H. Lin, B. Zhao and Y. W. Chen, Ind. Eng. Chem. Res.,
which provides a new catalytic way for the hydrogenation of
biomass using formic acid as a hydrogen source. Mechanism 21 J. Legrand, A. Taleb, S. Gota, M. J. Guittet and C. Petit,
studies have shown that the rapid hydrogenation of carboca- Langmuir, 2002, 18, 4131–4137.
2009, 48, 7037–7043.
tions and HMP can effectively reduce the occurrence of 22 L. Wang, W. Liu and X. Wang, Tribol. Lett., 2009, 37, 381–387.
condensation side reactions, which is the key to further 23 W. Zhao, A. D. Doyle, S. E. Morgan, M. Bajdich, J. K. Nørskov
improving the selectivity of the target product MMP.
and C. T. Campbell, J. Phys. Chem. C, 2017, 121, 28001–28006.
4 T. M. Huynh, U. Armbruster, M. M. Pohl, M. Schneider,
J. Radnik, D. L. Hoang and A. Martin, ChemCatChem, 2014,
2
Conflicts of interest
6
, 1940–1951.
25 Z. Zhu, J. Ma, L. Xu, L. Xu, H. Li and H. Li, ACS Catal., 2012,
, 2119–2125.
6 J. H. Shen and Y. W. Chen, J. Mol. Catal. A: Chem., 2007, 273,
65–276.
The authors declare no conicts of interest.
2
2
Acknowledgements
2
This work was nancially supported by the Natural Science 27 Z. Li, H. Li, L. Wang, T. Liu, T. Zhang, G. Wang and G. Xie,
Foundation of Hebei Province (No. B2018202293).
Int. J. Hydrogen Energy, 2014, 39, 14935–14941.
8 W. Y. Wang, Y. Q. Yang, J. G. Bao and H. A. Luo, Catal.
Commun., 2009, 11, 100–105.
9 F. Li, J. Liang, K. Wang, B. Cao, W. Zhu and H. Song, Can. J.
Chem. Eng., 2017, 95, 2012–2017.
2
2
References
1
M. Saidi, F. Samimi, D. Karimipourfard, T. Nimmanwudipong,
B. C. Gates and M. R. Rahimpour, Energy Environ. Sci., 2014, 7, 30 W. Wang, Y. Yang, H. Luo, H. Peng, B. He and W. Liu, Catal.
03–129. Commun., 2011, 12, 1275–1279.
X. Li, G. Chen, C. Liu, W. Ma, B. Yan and J. Zhang, Renewable 31 J. Reyes, E. Arriola, J. A. Hern ´a ndez and G. A. Fuentes, Appl.
Sustainable Energy Rev., 2017, 71, 296–308. Catal., A, 2015, 497, 211–215.
P. R. Patwardhan, R. C. Brown and B. H. Shanks, 32 L. Zhang, Z. Xin, Z. Liu, G. Wei, Z. Li and Y. Ou, Renewable
ChemSusChem, 2011, 4, 1629–1636.
Energy, 2020, 147, 695–704.
Y. Y. Wang, L. L. Ling and H. Jiang, Green Chem., 2016, 18, 33 R. Fan, C. Chen, M. Han, W. Gong, H. Zhang, Y. Zhang and
1
2
3
4
4
032–4041.
G. Wang, Small, 2018, 14, 1801953.
34 M. J. Hidajat, A. Riaz and J. Kim, Chem. Eng. J., 2018, 348,
799–810.
5
6
H. Jiang, J. Chem. Res., 2019, 37, 761–763.
F. Huang, W. Li, Q. Lu and X. Zhu, Chem. Eng. Technol., 2010,
3
3, 2082–2088.
35 K. Mori, T. Taga and H. Yamashita, ACS Catal., 2017, 7,
3147–3151.
7
J. Du, J. Zhang, Y. Sun, W. Jia, Z. Si, H. Gao and L. Lin, J.
Catal., 2018, 368, 69–78.
11002 | RSC Adv., 2021, 11, 10996–11003
© 2021 The Author(s). Published by the Royal Society of Chemistry

View Article Online
Paper
RSC Advances
3
3
3
3
6 K. Tedsree, T. Li, S. Jones, C. W. A. Chan, K. M. K. Yu, P. A. Bagot 40 N. C. Ellebracht and C. W. Jones, Carbohydr. Polym., 2020,
and S. C. E. Tsang, Nat. Nanotechnol., 2011, 6, 302–307. 233, 115825.
7 M. Al-Naji, M. Popova, Z. Chen, N. Wilde and R. Gl ¨a ser, ACS 41 S. Zhou, F. Dai, C. Dang, M. Wang, D. Liu, F. Lu and H. Qi,
Sustainable Chem. Eng., 2019, 8, 393–402. Green Chem., 2019, 21, 4732–4747.
8 K. Mori, M. Dojo and H. Yamashita, ACS Catal., 2013, 3, 42 V. R. Bakuru, D. Davis and S. B. Kalidindi, Dalton Trans.,
114–1119. 2019, 48, 8573–8577.
1
ˇ
ˇ
´
9 O. Kikhtyanin, L. Capek, Z. Tisler, R. Velvarska, A. Panasewicz,
P. Dibl ´ı kov ´a and D. Kubi ˇc ka, Front. Chem., 2018, 6, 176.
©
2021 The Author(s). Published by the Royal Society of Chemistry
RSC Adv., 2021, 11, 10996–11003 | 11003