Catalytic performance and deoxygenation path of methyl palmitate on Ni2P/SiO2 synthesized using the thermal decomposition of nickel hypophosphite

Article abstract of DOI:10.1039/c6ra02601j

In this paper, the catalytic performance and deoxygenation path of methyl palmitate on Ni₂P/SiO₂ catalysts were systematically studied in a continuous flow fixed-bed reactor. A series of Ni₂P/SiO₂ catalysts (with different molar ratios of P/Ni and Ni₂P loadings) were synthesized at 300°C using the thermal decomposition of nickel hypophosphite. The increased molar ratio of P/Ni generates phosphate-rich nickel phosphide catalysts and increasing conversion. Interestingly, Ni₂P/SiO₂ showed significantly higher conversion of methyl palmitate in comparison with Ni/SiO₂. Furthermore, an activation temperature higher than 500°C would significantly reduce the catalytic activity, as a result of the sintering of Ni₂P. The pressure in a range of 3.0 to 0.5 MPa almost has no effect on the deoxygenation of methyl palmitate, but significantly affects the reaction path and product distribution. Finally, a possible deoxygenation path over Ni₂P/SiO₂ was proposed based on a GC-MS investigation.

Full text of DOI:10.1039/c6ra02601j

RSC Advances
View Article Online
View Journal | View Issue
PAPER
Catalytic performance and deoxygenation path of
methyl palmitate on Ni P/SiO synthesized using
2
2
Cite this: RSC Adv., 2016, 6, 31308
the thermal decomposition of nickel
hypophosphite
a
a
ab
Qingxin Guan, Fei Han and Wei Li*
In this paper, the catalytic performance and deoxygenation path of methyl palmitate on Ni
2
P/SiO
catalysts (with
P loadings) were synthesized at 300 C using the thermal
decomposition of nickel hypophosphite. The increased molar ratio of P/Ni generates phosphate-rich
nickel phosphide catalysts and increasing conversion. Interestingly, Ni P/SiO showed significantly higher
. Furthermore, an activation temperature
P.
2
catalysts
were systematically studied in a continuous flow fixed-bed reactor. A series of Ni P/SiO
2
2
ꢀ
different molar ratios of P/Ni and Ni
2
2
2
conversion of methyl palmitate in comparison with Ni/SiO
2
Received 28th January 2016
Accepted 19th March 2016
ꢀ
higher than 500 C would significantly reduce the catalytic activity, as a result of the sintering of Ni
2
The pressure in a range of 3.0 to 0.5 MPa almost has no effect on the deoxygenation of methyl
palmitate, but significantly affects the reaction path and product distribution. Finally, a possible
DOI: 10.1039/c6ra02601j
www.rsc.org/advances
deoxygenation path over Ni
2
P/SiO
2
was proposed based on a GC-MS investigation.
Similar to hydrodesulfurization and hydrodenitrogenation,
1
. Introduction
10
HDO mainly involves the cleavage of C–O and C–C bond.
11,12
In recent times, biomass has attracted much attention as Several catalysts have been developed for HDO.
a sustainable resource with advantages over fossil fuels such as conventional transition metal sulde catalysts (such as CoMo/
renewability, lower expense, a neutral carbon cycle, and Al and NiMo/Al ) does not have sufficient activity with
However,
2
O
3
2 3
O
1
–5
protection of the environment. In the 1980s, rst-generation enough stability under HDO reaction conditions. On the other
biodiesel, its main component being a fatty acid methyl ester hand, sulfur content will be gradually reduced during the
13
(
FAME), was reported. However, there are too many problems reaction. Therefore, more and more studies focused on
when used in the engine. Second-generation biodiesel, being seeking for noble metal catalyst and other novel catalytic
composed of alkanes, can be obtained by hydrodeoxygenation materials to promote HDO efficiently. In this context, the
6
–10
(
HDO) and isomerization of FAME.
The second-generation transition metal phosphide catalyst showed excellent activity
14–18
biodiesel does not contain oxygen and sulfur, has a lower and selectivity.
2
Y. Yang and coworkers studied the Ni P/SBA-
density and viscosity, with a high cetane number and lower 15 as a HDO catalyst with enhanced selectivity for converting
cloud point. Hence, the second-generation biodiesel can also be methyl oleate into n-octadecane. They found that the Ni/SBA-15
termed as “renewable diesel” and resembles petrodiesel in displays a slightly higher activity than Ni
terms of hydrocarbon ingredient, and thus has been gaining higher pressure and lower temperature and space velocity
P/SBA-15. Overall,
2
14
great attention in recent years. The renewable diesel can be favors the formation of C18 hydrocarbon. J. A. Cecilia and
synthesized from a number of plant oils, such as jatropha oil, coworkers studied the HDO of dibenzofuran over silica-
palm oil, castor oil, and algae oil. These bio-oils usually contain supported nickel phosphide with different initial P/Ni ratios.
large amounts of oxygen and an upgrading process is necessary As a result, nickel phosphide based catalysts showed a high
17
to reduce oxygen content to improve heat value and stability. resistance to deactivation by coke formation. J. Moon and
Among several strategies, catalytic HDO has been considered as coworkers studied the active sites of Ni
P/SiO catalyst for HDO
2
2
9
a very promising route.
of guaiacol by using XAFS and DFT study. They revealed that the
active site of Ni P is composed of threefold hollow Ni and P sites
2
18
which lead to adsorption of H or OH groups.
In previous studies, different metal phosphides were
a
College of Chemistry, Key Laboratory of Advanced Energy Materials Chemistry
(
Ministry of Education), Nankai University, Tianjin, 300071, China. E-mail: weili@ synthesized using the thermal decomposition of metal hypo-
19–23
nankai.edu.cn; Fax: +86-22-23508662; Tel: +86-22-23508662
phosphite and metal phosphate precursors.
As shown in
b
Collaborative Innovation Center of Chemical Science and Engineering (Tianjin),
Scheme 1, Ni P/SiO was synthesized for converting methyl
2
2
Nankai University, Tianjin, 300071, China
31308 | RSC Adv., 2016, 6, 31308–31315
This journal is © The Royal Society of Chemistry 2016

View Article Online
Paper
RSC Advances
impregnation. Then, the wet solid was evaporated slowly to
ꢀ
dehydrate, and was dried at 80 C for 8 h to obtain the
precursor. At the beginning of the thermal decomposition, the
air in the reactor was removed by owing Ar, aer which the
ꢀ
Scheme 1 HDO reaction of methyl palmitate over Ni
2
P/SiO
2
.
precursor was treated at 300 C for 0.5 h in a static Ar atmo-
ꢀ
ꢁ1
sphere (heating rate: 5–20 C min ). The product was cooled to
ambient temperature under Ar and was washed several times
with deionized water to remove ion impurities, aer which the
wet material was dried at 120 C for 3 h. Nickel phosphide
catalysts with different mole ratio of P/Ni and different Ni
palmitate to pentadecane and hexadecane. Compared with
palmitic acid, methyl palmitate as a model compound for
triglyceride feedstock having a lower melting point (30 C), is
ꢀ
ꢀ
loading have similar synthesis procedure.
more suitable for laboratory studies. However, all of the pub-
lished papers on the HDO of methyl palmitate were used sup-
ported nickel catalysts and carried out in a batch stainless steel
The Ni/SiO was synthesized using the thermal decomposi-
tion of nickel nitrate precursor, the precursor being prepared
using incipient wetness impregnation. The wet precursors were
evaporated slowly to dehydrate, and were dried at 120 C for 5 h.
Subsequently, the precursor was calcined in a muffle furnace at
50 C for 3 h, aer which the sample was reduced at 350 C for
3 h in owing H atmosphere. Finally, the product was cooled to
ambient temperature under owing H and was passivated for
2
24–26
autoclave.
Therefore, it would be interesting to study the
ꢀ
Ni P/SiO for the HDO of methyl palmitate in a continuous ow
2
2

xed-bed reactor. In our recent studies, we have carried out
ꢀ
ꢀ
3
kinetic investigation of deoxygenation of methyl palmitate over
2
7
SiO
Ni P/SiO
2
2
-supported nickel phosphide catalysts. Herein, a series of
catalysts (with different molar ratio of P/Ni and Ni
2
2
2
P
2
ꢀ
1 h under owing 1% O /N .
2 2
loadings) were synthesized at 300 C using the thermal
decomposition of nickel hypophosphite. Catalytic performance
and deoxygenation path of methyl palmitate on Ni P/SiO
2
For comparison, the bulk Ni P was synthesized using the
temperature-programmed reduction (TPR) method. Initially,
nickel nitrate and diammonium hydrogen phosphate, with
a mole of P/Ni ¼ 0.5, were dissolved in deionized water and
2
2
catalysts were systematically studied in a continuous ow xed-
bed reactor. Different with Yang and co-workers published
ꢀ
1
4
stirred for 1 h. Then, the solution was evaporated at 120 C for
results, Ni
2
P/SiO
2
showed signicantly higher conversion in
. Interestingly, the activation temper-
ꢀ
3
h and calcined at 550 C for 2 h to obtain the oxidic precursor.
comparison with Ni/SiO
2
Subsequently, the precursor was loaded into the quartz tube
reactor. The temperature program of the TPR method was as
follows. The ow rate of H in the whole procedure was 200
2
atures and reaction pressures signicantly affect the conversion
and product distribution. Finally, a possible HDO mechanism
for the present HDO over Ni
MS investigation.
2 2
P/SiO was proposed based on GC-
ꢁ
1
ꢀ
ꢁ1
mL min . A heating rate of 5 C min was used from room
ꢀ
ꢀ
temperature to 400 C, and maintained at 400 C for 1 h. Then,
a heating rate of 1–2 C min was used from 400 to 600 C, and
maintained at 600 C for 3 h. Subsequently, the product was
ꢀ
ꢁ1
ꢀ
2. Experimental methods
ꢀ
cooled to ambient temperature under owing H , and was
passivated for 2 h under owing 1% O /N .
2 2
2
SiO
2
was purchased from Tianjin Chemist Scientic Ltd.,
2
ꢁ1
China. It has a specic surface area of 206 m g , a pore volume
3
ꢁ1
of 0.8 cm g . Other reagents were analytical pure grade and
purchased from Tianjin Guangfu Fine Chemical Research
Institute, China.
2.2. Characterization
The characterization was performed for passivated sample,
which was exposed to the air and without any reduction treat-
ment. Powder X-ray diffraction (XRD) was performed on
2.1. Synthesis of bulk and supported nickel phosphides
The preparation of bulk Ni
2
P is illustrated as follows. Nickel a Bruker D8 focus diffractometer, with Cu Ka radiation at 40 kV
PO ) were and 40 mA. The transmission electron microscope (TEM)
chloride (NiCl
2
) and sodium hypophosphite (NaH
2
2
2
dissolved in deionized water under stirring with a mole ratio of images were acquired using a Philips Tecnai G F-20 eld
P/Ni ¼ 1.5/1. Aer stirring for 1 h, the solution was evaporated emission gun TEM. Scanning electron microscope (SEM)
ꢀ
slowly to dehydrate, and was dried at 80 C for 8 h to obtain the images were acquired using a JEOL JSM-7500F SEM equipped
precursor. At the beginning of the thermal decomposition, the with an energy dispersive spectrometer (EDS). The composi-
air in the reactor was removed by owing Ar, aer which the tions of the sample were measured by inductively coupled
ꢀ
precursor was treated at 300 C for 0.5 h in a static Ar atmo- plasma-atomic emission spectrometry (ICP-AES). The Bru-
ꢀ
ꢁ1
sphere (heating rate: 5–20 C min ). The product was cooled to nauer–Emmett–Teller (BET) specic surface areas were ob-
ambient temperature under Ar and was washed several times tained using nitrogen adsorption/desorption measurements at
with deionized water to remove ion impurities, aer which the 77 K with a BELSORP-Mini instrument. Samples were degassed
ꢀ
ꢀ
wet material was dried at 120 C for 3 h.
at 300 C for 3 h before measurement. The surface area was
The preparation of Ni P/SiO is illustrated as follows. NiCl
2
calculated using a multi-point BET model. X-ray photoelectron
2
2
and NaH PO were dissolved in deionized water under stirring spectroscopy (XPS) was carried out using a Kratos Axis Ultra
2
2
with a mole ratio of P/Ni ¼ 1.5/1. Aer stirring for 1 h, the DLD spectrometer employing a monochromated Al-Ka X-ray
solution was impregnated onto the SiO
2
using incipient wetness source (hn ¼ 1486.6 eV), hybrid (magnetic/electrostatic) optics
This journal is © The Royal Society of Chemistry 2016
RSC Adv., 2016, 6, 31308–31315 | 31309

View Article Online
RSC Advances
Paper
and a multi-channel plate and delay line detector (DLD). All XPS
spectra were recorded using an aperture slot of 300 ꢂ 700
microns, survey spectra were recorded with a pass energy of 80
eV, and high-resolution spectra with a pass energy of 40 eV. In
order to subtract the surface charge effect, the C 1s peak was

xed at a binding energy of 284.6 eV. The CO chemisorption was
performed with Micromeritics Chemisorb 2750 gas-adsorption
equipment. The sample was loaded into a quartz reactor and
ꢀ
pretreated in 10% H
2
/Ar at 450 C for 3 h. Aer cooling in He,
ꢁ
1
pulses of 10% CO/He in a He carrier (25 mL (NTP) min ) were
injected at 30 C through a loop tube.
ꢀ
2.3. Catalytic activity test
The HDO catalytic activities were evaluated using 50 wt%
methyl palmitate in decalin. The HDO reaction was carried out
in a xed-bed microreactor. The catalyst was pelleted, crushed,
and sieved with 20–40 mesh; 2.0 g of the catalyst was diluted
with SiO
2
to a volume of 10.0 mL in the reactor. Prior to the
reaction, catalysts were pretreated in situ with owing H (100
2
ꢁ
1
mL min ) for 3 h. The testing conditions for the HDO reaction
ꢁ
1
were 3 MPa, weight hourly space velocity (WHSV) ¼ 3 h , and
/oil ¼ 1000. Liquid products were collected every hour aer
a stabilization period of 6 h. Both feed and products were
H
2
Fig. 1 The XRD patterns of bulk nickel phosphides prepared using the
thermal decomposition of a nickel hypophosphite precursor with
analyzed with an Agilent 7890A/5975C GC-MS equipped with different mole ratios of P/Ni.
a ame ionization detector and an HP-5 column. The same
catalyst was used to investigate the effects of different temper-
atures. The reaction temperature is being investigated in given in Fig. 2. The typical diffraction peaks of NaCl were
descending order, and each temperature is maintained 3 h. The detected in unwashed sample, and which were disappeared in
methyl palmitate conversion and turnover frequency (TOF) were washed sample. As shown in the SEM images, Ni P nano-
2
used to evaluate the catalytic activity. The TOF was calculated particles (about 200 nm) are embedded in sodium chloride in
using eqn (1):
unwashed sample (Fig. 2a). This embedded coexistence can
effectively prevent sintering of particles. Particle size of Ni2P
F
A0
X
A
TOF ¼
(1)
obtained using hypophosphite method is smaller than that of
W COuptake
using TPR method (500–1000 nm) in Fig. 2d and e.
where F is the molar rate of reactant fed into the reactor
A0
ꢁ
Fig. 3 shows the XRD patterns of 10 wt% Ni P/SiO catalysts
2
2
1
(
mmol s ), W is the catalyst weight (g), CO
is the uptake of
uptake
prepared using the thermal decomposition of a nickel hypo-
phosphite precursor with various mole ratios of P/Ni. No typical
ꢁ
1
chemisorbed CO (mmol g ), and X is the reactant conversion (%).
A
2
diffraction peaks of Ni P were detected, which indicates that the
crystal of nickel phosphide is smaller than the detection limit of
XRD and has a good dispersion on the SiO2 support. Low
calcination temperature can signicantly reduce the sintering
3
. Results and discussion
3.1. Synthesis of bulk and supported nickel phosphides
In our previous published paper, Ni
2
P was synthesized from of crystal grain. Fig. 4 shows the XRD patterns of Ni
2 2
P/SiO
nickel hypophosphite in the mole ratio of P/Ni range of 1.5 to catalysts prepared using the thermal decomposition of a nickel
23
1
.75. To further study the inuence of the mole ratio of P/Ni on hypophosphite precursor at a mole ratio of P/Ni ¼ 1.5 with
Ni P crystal forms, a series of Ni P were prepared using the different loadings. No typical diffraction peaks of Ni P were
thermal decomposition of a nickel hypophosphite precursor detected when the loading was less than 15 wt%. The crystal size
with different mole ratios of P/Ni. As shown in Fig. 1, we can of Ni P gradually becomes large with increasing loading. As
clearly see that Ni
P can be synthesized in the mole ratio of P/Ni shown in Fig. 4, weak diffraction peaks of Ni P were detected in
range of 1.5 to 1.75. The weak diffraction peaks of Ni
were 15 wt% Ni P/SiO , and diffraction peaks of Ni P increase as the
detected when the mole ratio of P/Ni is 2.0, which indicates that loading increases from 15 to 25 wt%. The result shows that the
P is a slight excess for the synthesis of bulk Ni
P. Not surpris- increased loading could cause the sintering of nickel phosphide
x
x
2
2
2
2
5
P
4
2
2
2
2
ingly, XRD detected more diffraction peaks of Ni
mole ratio of P/Ni is 2.5.
5
P
4
when the crystal grain. In order to observe the distribution of nickel
phosphide particles, we performed TEM characterization.
To study the effect of preparation method (hypophosphite
method and TPR method) on bulk Ni
P particles, SEM images thermal decomposition of a nickel hypophosphite precursor at
and XRD patterns of the samples before and aer washing were a mole ratio of P/Ni ¼ 1.5 with different loadings. The TEM
2 2
Fig. 5 shows the TEM images of Ni P/SiO prepared using the
2
31310 | RSC Adv., 2016, 6, 31308–31315
This journal is © The Royal Society of Chemistry 2016

View Article Online
Paper
RSC Advances
crystallographic planes of Ni
2
P, in good agreement with the
calculated values (PDF# 65-1989). It shows the formation and
good dispersion of nanocrystalline Ni P on the SiO support.
2
2
Fig. 6 shows the SEM and EDS mapping images of different
catalysts (red dot is nickel atom and blue dot is phosphorus).
Fig. 6a and b show that the Ni/SiO
2
has better metal dispersion
than the Ni P/SiO with the same Ni loading. Fig. 6c gives the
2
2
mole ratio of P/Ni ¼ 0.71 measured by EDS, which is slightly
higher than the ICP-AES result, P/Ni ¼ 0.62. Considering that
EDS is a semiquantitative analysis method, and a small amount
of phosphate ion is not washed away, the mole ratio of P/Ni ¼
0.71 is a reasonable value. Fig. 6b and d show that Ni and P
Fig. 2 The SEM images and XRD patterns of bulk nickel phosphides
prepared using the thermal decomposition of a nickel hypophosphite
precursor with a mole ratio of P/Ni ¼ 1.5; (a) unwashed sample, (b)
washed sample, (c) XRD patterns of unwashed and washed samples,
(
2
d) and (e) bulk Ni P synthesized using TPR method.
2 2
Fig. 4 The XRD patterns of Ni P/SiO catalysts prepared using the
thermal decomposition of a nickel hypophosphite precursor at a mole
ratio of P/Ni ¼ 1.5 with different loadings.
2 2
Fig. 3 The XRD patterns of 10 wt% Ni P/SiO catalysts prepared using
the thermal decomposition of a nickel hypophosphite precursor with
different mole ratios of P/Ni.
images show fairly uniform nanoparticles with a size of 2–4 nm,
indicating Ni P has good dispersion on the SiO support. The
2
2
weak diffraction peaks of Ni
2
2 2
P for 15, 20, and 25 wt% Ni P/SiO
catalysts may be caused by a few larger sintered Ni
shown in Fig. 5b, a magnied high resolution TEM image of 10
wt% Ni P/SiO , yields d-spacing values of 0.221 nm, for the (111)
2
2 2
P crystals. As Fig. 5 The TEM images of Ni P/SiO catalysts prepared using the
thermal decomposition of a nickel hypophosphite precursor at a mole
ratio of P/Ni ¼ 1.5 with different loadings, (a) 5 wt%, (b) 10 wt%, (c) 15
wt%, and (d) 20 wt%.
2
2
This journal is © The Royal Society of Chemistry 2016
RSC Adv., 2016, 6, 31308–31315 | 31311

View Article Online
RSC Advances
Paper
Fig. 6 The SEM and EDS mapping images of different catalysts (red
dot refer to nickel atom, and blue dot refer to phosphorus atom), (a) 10
wt% Ni/SiO
P/Ni ¼ 1.5, and (d) 25 wt% Ni
2
, (b) 10 wt% Ni
2
P/SiO
2
, P/Ni ¼ 1.5, (c) 10 wt% Ni
2 2
P/SiO ,
2
P/SiO
2
, P/Ni ¼ 1.5.
Fig. 7 The XPS spectra of 10 wt% Ni
2
P/SiO
2
prepared using the
thermal decomposition of a nickel hypophosphite precursor with
a mole ratio of P/Ni ¼ 1.5.
dispersion increased signicantly with the increase in Ni
loading.
To observe of the valence of Ni and P elements on 10 wt%
Ni
2
P/SiO
2
surface, we carried out XPS analysis. Fig. 7 shows the
P/SiO prepared using the thermal
different pretreatment temperatures. The CO uptakes of cata-
XPS spectra of 10 wt% Ni
decomposition of a nickel hypophosphite precursor with a mole
2
2
ꢀ
lysts aer pretreated at 330, 400, 500, and 600 C is 6.1, 6.2, 3.9,
ꢁ
1
3
.5 mmol g , respectively. These results indicate that the high
ratio of P/Ni ¼ 1.5. According to the most of the published XPS
pretreatment temperature might cause sintering of the Ni
2
P
17,19
results,
the Ni 2p spectrum of Ni P mainly involves two
2
nanoparticles. Therefore, HDO activity decreased when the
d+
peaks, the rst one is assigned to Ni of Ni
2
P and centered at
ꢀ
activation temperature was increased to 500, up to 600 C.
8
52.9 eV, and the second one is at 856.5 eV, corresponding to
To study the detailed reaction pathway of methyl palmitate
HDO, different WHSVs, reaction temperatures, reaction pressures,
and loadings were investigated in detail. Fig. 9 shows the catalytic
2
+
Ni ions interacting possibly with phosphate ions as a conse-
quence of a supercial passivation. For P 2p, the binding energy
at 129.9 eV is attributed to P of Ni
dꢁ
2
P and 134.0 eV is attributed
to surface nickel phosphate species. As a result of the XPS
characterization sample is 10 wt% Ni P/SiO , therefore, the
2
2
intensity of the peaks of Ni 2p and P 2p are relatively low.
3.2. Catalytic activity
The HDO catalytic activities were evaluated using 50 wt%
methyl palmitate in decalin.
In order to investigate the inuence of hydrogen activation,
different catalyst activation temperatures were rst studied.
Fig. 8 shows the catalytic activities of 10 wt% Ni/SiO
wt% Ni P/SiO catalysts aer different pretreatment tempera-
tures (testing conditions: 3 MPa, 290 C, WHSV ¼ 3 h , H
1000). The result indicates that Ni P/SiO showed signi-
cantly higher activity than Ni/SiO aer different pretreatment
2
and 10
2
2
ꢀ
ꢁ1
2
/oil
¼
2
2
2
temperatures. In particular, the catalytic activity of Ni P/SiO
aer being pretreated at 330 or 400 C was more than twice that
2
2
ꢀ
Fig. 8 The catalytic activities of 10 wt% Ni/SiO
2 2 2
and 10 wt% Ni P/SiO
of Ni/SiO
2
. To nd out why activation temperature affect activity
(
P/Ni ¼ 1.5) catalysts after different pretreatment temperatures (testing
ꢀ
ꢁ1
signicantly, we measured the CO uptakes for catalysts aer conditions: 3 MPa, 290 C, WHSV ¼ 3 h , H
2
/oil ¼ 1000).
31312 | RSC Adv., 2016, 6, 31308–31315
This journal is © The Royal Society of Chemistry 2016

View Article Online
Paper
RSC Advances
activities of 10 wt% Ni
2
P/SiO
2
(P/Ni ¼ 1.5) at different WHSVs
ꢀ
(testing conditions: 3 MPa, 290 C, H /oil ¼ 1000). Not surpris-
2
ingly, HDO conversion and selectivity of pentadecane (C15) and
hexadecane (C16) were gradually decreased as the WHSV
increased. Meanwhile, the amount of intermediates (such as
pentadecene, palmitic acid, hexadecanal, and hexadecanol) grad-
ually increased. In order to better observe the changes of inter-
mediates, subsequent tests were carried out at high WHSV. Fig. 10
shows the catalytic activities of 10 wt% Ni
different reaction temperatures (testing conditions: 3 MPa, WHSV
2
P/SiO
2
(P/Ni ¼ 1.5) at
ꢁ
1
¼
9 h , H /oil ¼ 1000). HDO conversion and selectivity of C15
2
and C16 gradually increased as the reaction temperature
increased. Meanwhile, the amount of intermediates (such as
pentadecene, palmitic acid, hexadecanal, and hexadecanol) grad-
ually decreased.
Fig. 11 shows the catalytic activities of 10 wt% Ni
2
P/SiO
2
Fig. 11 The catalytic activities of 10 wt% Ni
2
P/SiO
2
(P/Ni ¼ 1.5) at
ꢀ
ꢁ1
different reaction pressures (testing conditions: 290 C, WHSV ¼ 9 h
,
(
2
P/Ni ¼ 1.5) at different reaction pressures (testing conditions:
ꢀ
ꢁ1
2
H /oil ¼ 1000).
90 C, WHSV ¼ 9 h , H /oil ¼ 1000). We can clearly see that
2
the pressures in a range of 3.0 to 0.5 MPa almost has no effect
on the conversion of methyl palmitate. However, selectivity of
C15, C16, palmitic acid, hexadecanal, and hexadecanol gradu-
ally decreased as the reaction pressure decreased, and the
content of pentadecene and hexadecene gradually increased.
The decreasing amount of C16 indicates that the cleavage of
C–O bond is unfavorable at lower pressure. The increased
contents of pentadecene and hexadecene show that the olen
hydrogenation reaction at low pressures is unfavorable.
Table 1 shows the catalytic activities of Ni/SiO
2
and Ni
2
P/SiO
2
ꢀ
at different loadings (testing conditions: 3 MPa, 290 C, WHSV
ꢁ
1
¼
3 h , H /oil ¼ 1000), which indicates that the HDO conver-
2
sion gradually increased as the Ni P loading increased. The
2
result also indicates that Ni P/SiO showed signicantly higher
2
2
2
activity than the Ni/SiO catalyst. The TOF was calculated using
eqn (1). Irreversible CO uptake measurements were used to
titrate the surface metal atoms and to provide an estimate of the
Fig. 9 The catalytic activities of 10 wt% Ni
different WHSVs (testing conditions: 3 MPa, 290 C, H
2
P/SiO
2
(P/Ni ¼ 1.5) at
/oil ¼ 1000).
ꢀ
2
active sites on the catalysts. We can see that 10 wt% Ni
2
P/SiO
2
ꢁ
1
and 10 wt% Ni/SiO have CO uptakes are 6.1, and 8.0 mmol g
2
,
ꢁ
1
and TOF are 0.10, and 0.04 s , respectively. The CO uptakes
and TOF values indicate that Ni P/SiO having lower metal
2
2
dispersion and higher catalytic activity than Ni/SiO2.
In our previous studies, increasing mole ratios of P/Ni of the
precursor increases the phosphorus content of the product
(Fig. 1), which might affect the catalytic activity. Fig. 12 shows
the catalytic activities of 10 wt% Ni P/SiO prepared using the
2
2
thermal decomposition of a metal hypophosphite precursor
with different mole ratios of P/Ni (testing conditions: 3 MPa,
ꢀ
ꢁ1
2
90 C, WHSV ¼ 3 h , H /oil ¼ 1000). The result indicates that
2
conversion and selectivity of C16 were gradually increased with
the increasing mole ratios of P/Ni in the precursor. We specu-
late that phosphorus-rich nickel phosphide could be more
favorable for HDO reaction.
Finally, a possible HDO reaction path for HDO of methyl
2 2
palmitate on Ni P/SiO was proposed based on GC-MS investi-
gation. As shown in Fig. 13, the reaction pathway of the HDO of
methyl palmitate mainly includes 10 steps. Methyl palmitate
Fig. 10 The catalytic activities of 10 wt% Ni
2
P/SiO
2
(P/Ni ¼ 1.5) at
different reaction temperatures (testing conditions: 3 MPa, WHSV ¼
ꢁ1
9
h
, H
2
/oil ¼ 1000).
This journal is © The Royal Society of Chemistry 2016
RSC Adv., 2016, 6, 31308–31315 | 31313

View Article Online
RSC Advances
Paper
ꢀ
ꢁ1
Table 1 The physical properties and HDO activities of different loading catalysts (testing conditions: 3 MPa, 290 C, WHSV ¼ 3 h , H
2
/oil ¼
1
000)
Catalysts
Ni/SiO 10 wt%
2
ꢁ1
ꢁ1
ꢁ1
S (m g
)
CO uptake (mmol g
)
Conversion (%)
TOF (s
)
Carbon balance (%)
2
179
189
175
156
132
116
8.0
3.8
6.1
7.8
8.4
7.5
20
28
39
80
74
74
0.04
0.11
0.10
0.15
0.15
0.15
98
97
97
97
97
97
Ni
2
Ni
2
Ni
2
Ni
2
Ni
2
P/SiO
P/SiO
P/SiO
P/SiO
P/SiO
2
2
2
2
2
5 wt%
10 wt%
15 wt%
20 wt%
25 wt%
2
9,30
and also could generate hexadecanal by dehydration (step 4).
31
Similar to Mavrikakis et al., we give an adsorption scheme of
palmitic acid on the Ni P surface (red rectangle). Hexadecanal
2
could directly generate pentadecene by take-off of the carbon
32
monoxide (step 5), which can be hydrogenated to produce
pentadecane (step 8). As shown in Fig. 11, pentadecene gradu-
ally increased and C15 decreased as the reaction pressure
decreased. The number of molecules in step 5 is increased, and
decreasing pressure more favorable promote the positive reac-
tion. Hence, the content of pentadecene is increased as the
reaction pressure decreased. Similarly, the number of mole-
cules in step 3 is also increased, but the content of C15 is
decreased, which shows that the C15 generated not only from
step 3 but also from step 8. Meanwhile, hexadecanal could
produce hexadecene from dehydration (step 6) or produce
2 2
Fig. 12 The catalytic activities of 10 wt% Ni P/SiO prepared using the hexadecanol by hydrogenation (step 7), which are parallel
thermal decomposition of a metal hypophosphite precursor with
33
reactions. The number of molecules in step 6 is not changed,
ꢀ
different mole ratios of P/Ni (testing conditions: 3 MPa, 290 C, WHSV
so the content of hexadecene almost is not changed as the
reaction pressure decreased. The number of molecules in step 7
is decreased, and decreasing pressure unfavorable promote the
positive reaction. Hence, the content of hexadecanol is
decreased as the reaction pressure decreased. C16 could be
ꢁ
1
¼
3 h , H
2
/oil ¼ 1000).
34
generated from hexadecene (step 9) and hexadecanol (step 10).
The number of molecules is decreased in the two reaction
pathway (step 6 + 9 and step 7 + 10), not surprisingly, content of
C16 is also decreased as the reaction pressure decreased. The
above results indicate that the pressures in a range of 3.0 to 0.5
MPa signicantly affect the intermediate products attribute
number of molecules in the different reaction paths are
different.
4. Conclusions
In this paper, a series of SiO supported Ni P catalysts were
2
2
synthesized using the thermal decomposition of nickel hypo-
phosphite precursors. The XRD and TEM results indicate that
P/ the crystal of nickel phosphide is smaller than the detection
2
Fig. 13 Reaction pathway of the HDO of methyl palmitate on Ni
SiO catalysts.
2
limit of XRD and has a good dispersion on the SiO
increased molar ratio of P/Ni (>1.75) gave phosphate-rich nickel
phosphide. The Ni P/SiO catalysts were characterized using
2
support. The
2
2
could produce palmitic acid via hydrogenation with methane XRD, TEM, SEM, EDS, ICP-AES, BET, and carbon monoxide
removal (step 1). At the same time, methyl palmitate could chemisorption and catalytic performance was evaluated for
produce hexadecanal through methanol abstraction (step 2). A HDO of methyl palmitate in a continuous ow xed-bed reactor.
2
8
similar result has been reported by Chen et al. Palmitic acid Notably, Ni P/SiO showed signicantly higher conversion of
2
2
26
could directly generate pentadecane by decarbonation (step 3),
methyl palmitate in comparison with Ni/SiO . The selectivity of
2
31314 | RSC Adv., 2016, 6, 31308–31315
This journal is © The Royal Society of Chemistry 2016

View Article Online
Paper
RSC Advances
C15 is much higher than that of C16 at different reaction 13 D. A. Ruddy, J. A. Schaidle, J. R. Ferrell, J. Wang, L. Moens
conditions. Furthermore, the catalytic activity of Ni P/SiO aer and J. E. Hensley, Green Chem., 2014, 16, 454–490.
2
2
ꢀ
being pretreated at 330 or 400 C was more than twice that of Ni/ 14 Y. Yang, C. Ochoa-Hern ´a ndez, V. A. De La Pe n˜ a O'Shea,
ꢀ
SiO . Activation temperature at higher than 500 C would
J. M. Coronado and D. P. Serrano, ACS Catal., 2012, 2, 592–
2
signicantly reduce the catalytic activity, as a result of the sin-
598.
tering of Ni
MPa almost does not affect conversion of methyl palmitate,
signicantly affect the intermediate products attribute number 16 Y. Yang, J. Chen and H. Shi, Energy Fuels, 2013, 27, 3400–
of molecules in the different reaction paths are different. 3409.
Finally, a possible HDO reaction path for HDO of methyl 17 J. A. Cecilia, A. Infantes-Molina, E. Rodr ´ı guez-Castell ´o n,
2
P. However, the pressures in a range of 3.0 to 0.5 15 Y. X. Yang, C. Ochoa-Hernandez, P. Pizarro, V. A. D. O'Shea,
J. M. Coronado and D. P. Serrano, Fuel, 2015, 144, 60–70.
palmitate on Ni P/SiO was proposed based on GC-MS
investigation.
A. Jim ´e nez-L ´o pez and S. T. Oyama, Appl. Catal., B, 2013,
136–137, 140–149.
2
2
1
1
8 J. S. Moon, E. G. Kim and Y. K. Lee, J. Catal., 2014, 311, 144–
152.
Acknowledgements
9 Q. Guan, X. Cheng, R. Li and W. Li, J. Catal., 2013, 299, 1–9.
This work was nancially supported by the NSFC (21376123, 20 Q. Guan and W. Li, Catal. Sci. Technol., 2012, 2, 2356–2360.
U1403293), MOE (IRT-13R30 and 113016A), NSFT 21 Q. Guan, C. Sun, R. Li and W. Li, Catal. Commun., 2011, 14,
(15JCQNJC05500), and the Research Fund for 111 Project
114–117.
(B12015).
22 Q. Guan and W. Li, J. Catal., 2010, 271, 413–415.
2
2
2
2
2
2
2
3
3
3
3
3
3 Q. Guan, W. Li, M. Zhang and K. Tao, J. Catal., 2009, 263, 1–
3.
References
4 I. V. Deliy, E. N. Vlasova, A. L. Nuzhdin, E. Y. Gerasimov and
G. A. Bukhtiyarova, RSC Adv., 2014, 4, 2242–2250.
5 H. L. Zuo, Q. Y. Liu, T. J. Wang, N. Shi, J. G. Liu and L. L. Ma,
J. Fuel Chem. Technol., 2012, 40, 1067–1073.
6 H. Zuo, Q. Liu, T. Wang, L. Ma, Q. Zhang and Q. Zhang,
Energy Fuels, 2012, 26, 3747–3755.
1
2
3
4
X. Zhang, T. Wang, L. Ma, Q. Zhang, X. Huang and Y. Yu,
Appl. Energy, 2013, 112, 533–538.
N. Sharma, H. Ojha, A. Bharadwaj, D. P. Pathak and
R. K. Sharma, RSC Adv., 2015, 5, 53381–53403.
K. L. Luska, P. Migowski and W. Leitner, Green Chem., 2015,
17, 3195–3206.
7 F. Han, Q. Guan and W. Li, RSC Adv., 2015, 5, 107533–
C. Li, X. Zhao, A. Wang, G. W. Huber and T. Zhang, Chem.
Rev., 2015, 115, 11559–11624.
1
07539.
8 J. Chen, H. Shi, L. Li and K. Li, Appl. Catal., B, 2014, 144, 870–
84.
5
6
N. Yan and X. Chen, Nature, 2015, 524, 155–157.
S. C. De Vries, G. W. J. van de Ven and M. K. van Ittersum,
Eur. J. Agron., 2014, 52, 166–179.
8
9 B. X. Peng, X. G. Yuan, C. Zhao and J. A. Lercher, J. Am. Chem.
Soc., 2012, 134, 9400–9405.
0 S. Brillouet, E. Baltag, S. Brunet and F. Richard, Appl. Catal.,
B, 2014, 148–149, 201–211.
7
M. Mohammadi, G. D. Najafpour, H. Younesi, P. Lahijani,
M. H. Uzir and A. R. Mohamed, Renewable Sustainable
Energy Rev., 2011, 15, 4255–4273.
1 M. Mavrikakis and M. A. Barteau, J. Mol. Catal. A: Chem.,
8
9
S. N. Naik, V. V. Goud, P. K. Rout and A. K. Dalai, Renewable
Sustainable Energy Rev., 2010, 14, 578–597.
Z. Zhu, H. Tan, J. Wang, S. Yu and K. Zhou, Green Chem.,
1998, 131, 135–147.
2 O. I. ¸S enol, E. M. Ryymin, T. R. Viljava and A. O. I. Krause, J.
Mol. Catal. A: Chem., 2007, 268, 1–8.
3 D. Li, P. Bui, H. Y. Zhao, S. T. Oyama, T. Dou and Z. H. Shen,
J. Catal., 2012, 290, 1–12.
4 R. W. Gosselink, S. A. W. Hollak, S.-W. Chang, J. Haveren,
K. P. Jong, J. H. Bitter and D. S. van Es, ChemSusChem,
2
014, 16, 2636–2643.
0 B. Van de Beld, E. Holle and J. Florijn, Appl. Energy, 2013,
02, 190–197.
1
1
1
1 P. M. Mortensen, J. D. Grunwaldt, P. A. Jensen,
K. G. Knudsen and A. D. Jensen, Appl. Catal., A, 2011, 407,
2013, 6, 1576–1594.
1
–19.
12 A. H. Zacher, M. V. Olarte, D. M. Santosa, D. C. Elliott and
S. B. Jones, Green Chem., 2014, 16, 491–515.
This journal is © The Royal Society of Chemistry 2016
RSC Adv., 2016, 6, 31308–31315 | 31315