Synthesis of an Ni2P catalyst supported on Na-MCM-41 with highly activity for dibenzothiophene HDS under mild conditions

Article abstract of DOI:10.1007/s11164-018-3423-z

A novel and simple method to synthesize supported Ni₂P/Na(x)-MCM-41 catalysts (where x is the mass fraction of Na-to-MCM-41 in terms of percentage) at a lower reduction temperature by incorporation of Na was described. The catalysts were characterized by H₂ temperature-programmed reduction (H₂-TPR), X-ray diffraction (XRD), N₂ adsorption–desorption, CO uptake, X-ray photoelectron spectroscopy (XPS), and transmission electron microscopy (TEM). The effect of Na on the structure of catalysts and catalytic properties for the dibenzothiophene (DBT) hydrodesulfurization (HDS) was investigated, which confirmed that a suitable amount of Na can promote highly dispersed Ni₂P particles. The Na preferentially interacts with phosphate to generate the sodium phosphate and therefore suppresses the formation of stronger P–O–P bonds, which enables the phosphide catalyst to be easily formed at a lower reduction temperature. Compared with conventional phosphate (973–1273?K), the reduction temperature of Ni₂P/Na(x)-MCM-41 catalyst was relatively low (773?K). The Ni₂P/Na(x)-MCM-41 catalyst with x = 1.0 showed the maximum DBT conversion of 91.6%, which is higher than that of Ni₂P/M41 without Na (80.3%).

Full text of DOI:10.1007/s11164-018-3423-z

Res Chem Intermed
https://doi.org/10.1007/s11164-018-3423-z
Synthesis of an Ni2P catalyst supported on Na-MCM-41
with highly activity for dibenzothiophene HDS
under mild conditions
Hua Song1
Fuyong Zhang1 Nan Jiang1 Maosen Chen1
•
•
•
•
Feng Li1 Zijin Yan1
•
Received: 9 October 2017 / Accepted: 2 April 2018
Ó Springer Science+Business Media B.V., part of Springer Nature 2018
Abstract A novel and simple method to synthesize supported Ni2P/Na(x)-MCM-41
catalysts (where x is the mass fraction of Na-to-MCM-41 in terms of percentage) at
a lower reduction temperature by incorporation of Na was described. The catalysts
were characterized by H2 temperature-programmed reduction (H2-TPR), X-ray
diffraction (XRD), N2 adsorption–desorption, CO uptake, X-ray photoelectron
spectroscopy (XPS), and transmission electron microscopy (TEM). The effect of Na
on the structure of catalysts and catalytic properties for the dibenzothiophene (DBT)
hydrodesulfurization (HDS) was investigated, which confirmed that a suit-
able amount of Na can promote highly dispersed Ni2P particles. The Na preferen-
tially interacts with phosphate to generate the sodium phosphate and therefore
suppresses the formation of stronger P–O–P bonds, which enables the phosphide
catalyst to be easily formed at a lower reduction temperature. Compared with
conventional phosphate (973–1273 K), the reduction temperature of Ni2P/Na(x)-
MCM-41 catalyst was relatively low (773 K). The Ni2P/Na(x)-MCM-41 catalyst
with x = 1.0 showed the maximum DBT conversion of 91.6%, which is higher than
that of Ni2P/M41 without Na (80.3%).
Keywords Ni2P Á Na-MCM-41 Á Sodium Á Hydrodesulfurization Á
Dibenzothiophene
Electronic supplementary material The online version of this article (https://doi.org/10.1007/s11164-
018-3423-z) contains supplementary material, which is available to authorized users.
&
Hua Song
songhua2004@sina.com
1
Provincial Key Laboratory of Oil and Gas Chemical Technology, College of Chemistry and
Chemical Engineering, Northeast Petroleum University, Daqing 163318, Heilongjing, China
123

H. Song et al.
Introduction
It is well known that the current commercial hydrodesulfurization (HDS) catalysts,
such as sulfides catalysts [1], are not up to the regulated levels [2]. Therefore,
development of new types of deep HDS catalysts has drawn great attention.
Transition metal phosphides have recently received extensive attention as a new
family of non-sulfided HDS catalysts because they exhibit higher HDS activity
[3, 4]. Among these, Ni2P showed the highest HDS activity [5, 6].
Some previous studies [7, 8] have shown that Ni2P catalysts were prepared by H2
temperature-programmed reduction (TPR) of oxidic precursors. The TPR method is
convenient and simple; however, it requires high reduction temperature for the
phosphate species mainly because the strong P–O bond in phosphate, which requires
much energy and a long production time. Therefore, it is necessary to explore novel
approaches featuring mild conditions for preparation of Ni2P catalyst. Recently,
there have been reports of some new approaches for preparing Ni2P catalysts.
Berhault et al. [9] found that unsupported nickel phosphides could be produced by
nickel thiophosphate (NiPS3) at lower temperatures. Lee [10] reported a liquid-
phase phosphidation method by which Ni2P nanostructures can be obtained by co-
reaction of nickel acetylacetonate (Ni(acac)2) with trioctylphosphine (TOP) in
coordinating solvent trioctylphosphine oxide (TOPO) at 573 K. Our previous study
[11, 12] used Ni(acac)2 and low-cost triphenylphosphine (TPP) in the presence of
the coordinating solvent tri-n-octylamine (TOA) to synthesize Ni2P catalyst at
603 K in an N2 atmosphere. It is well known that the reduction temperature could be
effectively decreased by using low-state phosphide instead of high-state phosphide.
Our group also found Ni2P catalysts can be successfully prepared by H2 reduction of
a mixture of NH4H2PO2 and NiCl2Á6H2O precursor at low temperature, and with
this preparation method, the catalysts do not need to be calcined, which saved
energy [13]. Brussell et al. [14] reported a novel preparation method of Ni2P
catalysts, in which the hypophosphite-based precursor (H3PO2) was used instead of
high-state phosphide phosphorus salts. One of its advantages is that it can be
conducted under mild conditions. But so far, most of the methods reported used
different phosphide precursors such as low-state phosphide to reduce the reduction
temperature of the Ni2P catalysts. Sawada [15] found that the reduction
temperatures of Rh2P supported on Al2O3 and SiO2 catalysts could be reduced by
adding metal sodium (Na) to the corresponding catalyst, showing that metal Na
plays an important role in reducing the reduction temperature of catalysts. Therefore
it appears reasonable to speculate that metal Na could affect the reduction
temperature of Ni2P catalyst precursor and promote the HDS activity of Ni2P
catalyst. However, to the best of our knowledge, up to now, there are no reported
works focused on the influence of Na on the reduction of Ni2P catalysts under mild
conditions and the performance of HDS. Therefore, it is interesting to study the
effect of Na on the reduction of Ni2P precursor under mild conditions and HDS
performance of Ni2P catalyst for refractory sulfur compounds such as dibenzoth-
iophene (DBT).
123

Synthesis of an Ni2P catalyst supported on Na-MCM-41 with…
In this paper, Ni2P/Na-MCM-41 catalysts were successfully prepared by the TPR
method with Na-doped Na-MCM-41 support at a low reduction temperature of
773 K. The effect of Na mass fraction on the structure of catalysts and catalytic
properties for DBT HDS were studied.
Experimental
Preparation of supports and catalysts
Siliceous MCM-41 was synthesized using tetraethyl orthosilicate (TEOS) as the
silica source and cetyltrimethylammonium bromide (CTAB) as the template,
following the procedure as described in the literature [16]. The Na-doped MCM-41
support was prepared by impregnation of the NaNO3 aqueous solution with MCM-
41 support, and then the water was evaporated to obtain a solid product. The
impregnated supports were dried at 383 K for 24 h and calcined at 673 K for 1 h
(with a heating rate of 2.5 K min-1). The supports obtained were named ‘M41’ for
MCM-41 and ‘Na(x)M41’ for Na-MCM-41, where x is the mass fraction of Na-to-
M41 in terms of percentage.
The supported Ni2P catalyst precursors were prepared by impregnating
diammonium hydrogen phosphate ((NH4)2HPO4) and nickel nitrate hexahydrate
(Ni(NO3)2Á6H2O) solution with the mesoporous supports. The precursors were
prepared with Ni loading of 9.73 wt%, and the initial Ni-to-P molar ratios was 1:2.
In a typical synthesis technique, (NH4)2HPO4 and Ni(NO3)2Á6H2O were dissolved in
deionized water at room temperature to form a uniform solution. The M41 and
Na(x)M41 were wet-impregnated with the above solution for 12 h. After the water
was evaporated, the resulting solid was dried at 393 K for 12 h and calcined at
773 K for 3 h to obtain the final oxidic precursor.
The catalyst precursors were reduced in a fixed-bed reactor by heating from room
temperature to 773 K at a rate of 2.0 K min-1 in a flow of H2 (200 mL min-1) and
held for 2 h. The obtained catalysts were allowed to cool naturally to room
temperature in a continuous H2 flow and then passivated in an O2/N2 mixture (0.5
vol.% O2) with a flow rate of 20 mL min-1 for 2 h. The corresponding catalysts are
denoted as Ni2P/M41 and Ni2P/Na(x)M41, respectively.
Characterization of catalysts
The reducibility of precursors was characterized by the H2 TPR (H2-TPR) using a
PC-1200 gas adsorption analyzer, in which 0.05 g of the precursors was loaded in a
quartz U tube reactor, and reduced in flow of a 10 vol.% H2/Ar (20 mL min-1) at a
heating rate of 5 K min-1. The TPR spectrum was determined using a thermal
conductivity detector (TCD) to monitor hydrogen consumption.
The small-angle XRD patterns were obtained with a RINT2000-X-ray diffrac-
tometer using Cu Ka radiation under the setting conditions of 35 kV, 50 mA, scan
range from 0.5° to 10° at a rate of 1/6 °C min-1. The large-angle XRD analysis was
carried out on a D/max-2200PC-X-ray diffractometer using Cu Ka radiation under
123

H. Song et al.
the setting conditions of 40 kV, 30 mA, scan range from 10 to 80° at a rate of
10° min-1
.
N2 adsorption–desorption was carried out on a NOVA2000e instrument at 77 K.
All the sample was pretreated at 473 K for 2 h until the vacuum pressure was 6 mm
Hg before measurement.
CO uptake measurements were performed by using a Setaram Tian-Calvet C-80
heat-flux microcalorimeter. Passivated catalyst (0.1 g) was loaded in the reactor and
reduced in a H2 flow (30 mL min-1) at 500 °C for 1 h. After pretreatment in a
continuous He flow and cooling the sample to room temperature, CO pulses were
injected into He, and the CO uptake was measured using a TCD until there was no
further CO adsorption after consecutive injections.
Transmission electron microscopy (TEM) examinations were performed using a
JEOL JEM-2100 instrument operated at 200 kV. The powder samples were
ultrasonically dispersed in ethanol and placed on a carbon-coated copper grid.
The XPS spectra were acquired using an ESCALAB MKII spectrometer, using
monochromatic Mg Ka radiation (E = 1253.6 eV) and equipped with a hemi-
spherical analyzer operating at fixed pass energy of 40 eV. The recorded
photoelectron binding energies were referenced against the C 1 s contamination
line at 284.8 eV.
Catalytic activities
The HDS of DBT over prepared catalysts was performed in a flowing high-pressure
fixed-bed reactor using a feed consisting of a decalin solution of DBT (1 wt%). The
conditions of the HDS reaction were 613 K, 3.0 MPa, WHSV = 2.5 h-1, and
hydrogen-to-oil ratio of 550 (V/V). Prior to reaction, 0.8 g of the catalysts were
pretreated in situ with flowing H2 (30 mL min-1) at 773 K for 2 h. The liquid
products of HDS reaction was collected and analyzed by flame ionization detector
(FID) gas chromatography (GC) with a GC-14C-60 column. Turnover frequency
(TOF) values of the samples containing nickel phosphide were calculated using
Eq. (1) [17]:
F X
TOF ¼
ð1Þ
W M
where F is the molar rate of DBT fed into the reactor (mol s-1), X is the conversion
of DBT (%), W is the weight of catalyst (g), and M is the mole of sites loaded which
is decided by the CO uptake.
Results and discussion
H2-TPR
The H2-TPR profiles of the Ni2P/M41 and Ni2P/Na(x)M41 catalyst precursors are
shown in Fig. 1. It can be observed that the reduction of Ni2P/M41 precursors
started around 910 K and the reduction peak was centered at 1028 K, which can be
123

Synthesis of an Ni2P catalyst supported on Na-MCM-41 with…
Fig. 1 H2-TPR profiles of the
Ni2P/M41 and Ni2P/Na(x)M41
catalyst precursors
ascribed to the reduction of highly thermodynamically stable P–O–P bonds and the
co-reductions of the nickel species in phosphate [18]. The reduction of Ni2P/
Na(x)M41 precursors started around 760 K and the reduction peaks were shifted to a
much lower region, indicating that the incorporation of Na into the Ni2P/M41 could
significantly decrease the reduction temperature. This is possibly because the Na
preferentially interacts with phosphate to generate the sodium phosphate and
therefore suppresses formation of stronger P–O–P bonds, which can produce the
phosphide catalyst at a lower reduction temperature [15]. A more detailed
discussion is shown in Sect. 3.8. With increasing the Na mass fraction, the
reduction temperature decreased to 823 K and then increased to 875 K, showing
that addition of too much Na increases the reduction temperature. This is possibly
because the excess Na causes formation of small amounts of the Ni5P4 phase
(Fig. 2, XRD analysis), which suppresses formation of Ni2P to some extent.
Fig. 2 XRD patterns of the
Ni2P/Na(x)M41 catalysts
123

H. Song et al.
XRD
Figure S1 illustrates the small-angle XRD patterns of the M41 and Na(1.0)M41
supports. The M41 indicated some peaks at 2h = 1.2–6° corresponding to 100, 110,
200, and 210 planes of the typical long hexagonal structure of M41. The
Na(1.0)M41 exhibited similar peaks. However, the peaks were slightly shifted to
higher angle values and intensities became lower, which indicates that a certain
amount of Na atoms could cause some deformation or distortion of the original M41
tetrahedral coordination structure. Figure S2 shows the small-angle XRD patterns of
the Ni2P/M41 and Ni2P/Na(1.0)M41 catalysts. Upon loading Ni2P, the diffraction
peaks were shifted to higher angle values and the intensities became lower as
compared to original supports, showing that loading Ni2P causes some lattice
distortion or lattice collapse.
The XRD patterns of Ni2P/Na(x)M41 catalysts with different Na mass fractions
are shown in Fig. 2. As can be seen from Fig. 2, all samples exhibit a broad feature
owing to the amorphous nature of mesoporous Na(x)M41 at 2h & 22°. Further-
more, the Ni2P/Na(x)M41 samples also showed peaks of Ni2P at 2h = 40.6, 44.5,
47.1 and 54.1° (PDF: 03-0953). This indicates that Ni2P is the main active phase for
all samples. The Ni2P/M41 showed relatively weak peaks of Ni2P, while the peaks
of Ni2P phase for the Ni2P/Na(x)M41 samples become more intense. In addition,
with increasing the Na mass fraction, the peaks intensities increased, which reveals
the Na can promote formation of Ni2P. This showed that incorporation of Na could
promote formation of more and higher-crystallinity Ni2P phase. Moreover, the Ni2P/
Na(2.0)M41 exhibits some weak diffraction peaks at 30.3, 31.5, 45.1, 49.7° and
52.9° (PDF: 18-0883) owing to small amounts of the Ni5P4 phase, which indicates
4-
that excess Na suppresses the reduction of PO3 into Ni2P. The Dc of the Ni2P
crystallites estimated by Scherrer’s equation [19, 20] is listed in column 5 of
Table 1. The crystallite size (Dc) of Ni2P/Na(0.5)M41 was 22.5 nm. With
increasing the Na mass fraction, the Ni2P particles increased, which shows that
the crystallinity of Ni2P phase increased. In other words, this result also shows that
the incorporation of Na can decrease the reduction temperature of Ni2P. Sawada
[20] similarly showed addition of Na can decrease the reduction temperature of
Rh2P and promote the formation of the active Rh2P phase.
N2 adsorption–desorption
Table 1 illustrates the textural properties of Na(x)M41 support and Ni2P/Na(x)M41
samples. The surface area and pore volume of the M41 support were 910.6 m2 g-1
and 0.96 cm3 g-1, respectively. Upon incorporating Na into M41, a decrease in
specific surface area (SBET) of Na(x)M41 was observed, accompanied with
decreases in pore volume (Vp) and pore diameter (d). In addition, with increase
of Na mass fraction, the SBET, Vp and d became much smaller, showing that the Na
was incorporated into the pores of the sieves. In parallel, a remarkable drop in the
SBET and Vp of the Ni2P/Na(x)M41 catalysts was observed after loading Ni2P, when
compared with the corresponding supports. This indicates that the pores of supports
were blocked by Ni2P. Moreover, the surface area of Ni2P/Na(x)M41 indicated a
123

Synthesis of an Ni2P catalyst supported on Na-MCM-41 with…
123

H. Song et al.
dramatic decreased upon increasing the Na mass fraction. The SBET of Ni2P/
Na(1.0)M41 was almost two-fold when compared with that found for Ni2P/
Na(2.0)M41, showing the growth of the Ni2P particle sizes (Table 1, Dc obtained by
XRD), which resulted in the decrease of SBET
.
The N2 adsorption–desorption isotherms of the supports and catalysts are
presented in Fig. 3. As can be seen from Fig. 3, all the supports and catalysts
exhibited a type IV isotherm and a standard H4 type hysteresis loop according to the
IUPAC classification, which reveals that some mesopores are present for all
samples.
CO uptake
The CO adsorption was used to estimate the surface density of exposed Ni sites on
the catalysts [21]. The CO uptakes at room temperature of the Ni2P/M41 and Ni2P/
Na(x)M41 catalysts are listed in column 6 of Table 1. Actually, CO molecules may
also be adsorbed on the surface of P sites. However, because of the relatively small
amount, it can be ignored in the following discussion [22]. As shown in Table 1, the
CO uptake of the Ni2P/M41 catalyst was 28 lmol g-1. Upon incorporation of Na,
the CO uptakes of the Ni2P/Na(x)M41 catalysts were increased. In addition, with
increasing the Na mass fraction, the CO uptakes of the Ni2P/Na(x)M41 catalysts
increased initially and decreased after a maximal value was achieved. These results
showed that an appropriate amount of Na can promote more exposed nickel atoms
on the surface. The amount of exposed Ni atoms over the catalyst is mainly affected
by two factors, on one hand, by the amount of the active Ni2P particles and their
dispersion, and on the other hand, by the P and Na coverage on the surface of
catalyst. With increasing the Na mass fraction (\ 1.0), more Ni2P particles were
formed (XRD analysis) and a better dispersion of the active Ni2P phase (TEM
analysis, will discuss later) can be achieved. Compared with Na, the P was
presented in a relatively large proportion. Therefore, at lower Na content (\ 1.0),
the P coverage is the key factor. With addition and increasing the Na mass fraction
(\ 1.0), the surface P-to-Ni ratio decreased significantly. For all these reason, the
CO uptake increased with increasing the Na mass fraction. However, when Na mass
Fig. 3 The nitrogen adsorption/desorption isotherms for M41, Na(x)M41, and Ni2P/Na(x)M41 catalysts
123

Synthesis of an Ni2P catalyst supported on Na-MCM-41 with…
fraction increased from 1.0 to a relatively higher value of 2.0, the surface P-to-Ni
ratio increased instead of decreased (Table 2, XPS analysis). Therefore, the
coverage of P and Na became more impactful, occupying some of the Ni sites and
leading to a lower CO uptake of Ni2P/Na(2.0)M41. In addition, the growth in Ni2P
particle sizes and decrease in surface area (See Table 1), which lead to poor
dispersion, were other reasons for the lower CO uptake of Ni2P/Na(2.0)M41.
XPS
XPS patterns were used to further identify the influence of Na mass fraction on the
catalysts and its surface composition. Figure 4 illustrates the XPS spectra of the
Ni2P/M41 and Ni2P/Na(x)M41 samples in the Ni(2p) and P(2p) regions, and
Table 2 shows the binding energies and the P-to-Ni atomic ratios for all samples. As
shown in Fig. 4, all spectra were decomposed, taking into account the spin-orbital
splitting of the Ni 2p3/2 and Ni 2p1/2 lines (about 17 eV) and the presence of satellite
peaks at about 5 eV higher than the binding energy of the parent signal [23]. It is
well known that Ni 2p3/2 core-level spectrum consists of three components. The
bands centered at 852.4–852.6 eV can be ascribed to Nid? in the Ni2P phase. The
second at 856.5–856.9 eV corresponds to the Ni2? species interacting with
phosphate species as a consequence of superficial passivation. This is accompanied
with the broad satellite at approximately 5.0 eV higher than that of the Ni2? species
and this shake-up peak is assigned to divalent species [24]. Other broad peaks on the
high-binding-energy side can be assigned to the Ni 2p1/2 signal from oxidized Ni
species [25]. P 2p binding energy involves the peaks at 128.9–129.3 eV can be
assigned to Pd- species on the metal phosphides [23] and the peak at
134.6–134.8 eV is due to phosphate (P5?) arising from superficial oxidation of
nickel phosphide particles [26].
As can be seen from Fig. 4a and Table 2, the XPS spectra for the as-prepared
Ni2P/M41 exhibited Ni 2p peaks at 852.6 and 856.9 eV, which can be attributed to
the Nid? band forming a Ni2P phase and Ni2? species, respectively. No obvious
change can be seen for the binding energy of Ni species between the Ni2P/
Na(x)M41 and Ni2P/M41. However, the peaks of Nid? and Ni2? species shift to a
Table 2 Spectral parameters obtained by XPS analysis for the catalysts
Sample
Binding energy (eV)
Superficial atomic ratio
Ni 2p3/2
P 2p3/2
Ni2?
Satellite
Nid?
P5?
Pd-
Na-to-Ni
P-to-Ni
Na-to-P
Ni2P/M41
856.9
856.7
856.6
856.5
856.8
861.7
861.5
861.7
861.8
862.2
852.6
852.5
852.4
852.4
852.6
134.8
134.6
134.8
134.6
134.8
128.9
129.3
129.3
129.3
129.2
–
3.14
2.12
1.96
1.84
1.95
–
Ni2P/Na(0.5)M41
Ni2P/Na(0.7)M41
Ni2P/Na(1.0)M41
Ni2P/Na(2.0)M41
0.20
0.21
0.23
0.40
0.09
0.11
0.13
0.21
123

H. Song et al.
Fig. 4 XPS spectra of the Ni2P/Na(x)M41 catalysts
slightly lower value with the incorporation of Na. This result shows that there is
interaction between the Ni2P particles and Na(x)M41 support. In addition, it is
worth noting that the intensity of the Nid? species peaks located around 852.4 eV
initially increased, reaching a maximum value at x = 1.0, and then decreased with
further increasing Na mass fraction in the catalyst. This indicates more Ni2P
particles may be exposed on the surface of Ni2P/Na(x)M41 and the incorporation of
Na can promote the formation of Ni2P phase. This result were also confirmed by
XRD. As can be seen from Fig. 4b, for Ni2P/M41, the peaks centered at 134.8 eV
for PO43- was observed, together with a low intensity peak at 128.9 eV that can be
attributed to the Pd- band forming Ni2P. However, the binding energies of Pd- over
the Ni2P/Na(x)M41 catalysts are slightly higher than the value reported for Ni2P/
M41 (128.9 eV) without Na addition. This shows the incorporation of Na can lead
to a small charge transfer from Ni to P in the catalyst [27]. This special effect favors
the reduction of PO34- into Ni2P. Therefore, an appropriate amount of Na (x = 1.0)
incorporated can promote the formation of active Ni2P phase.
Table 2 summarizes the binding energy and the surface Na-to-Ni, P-to-Ni, and
Na-to-P atomic ratios for the Ni2P/M41 and Ni2P/Na(x)M41 catalysts. The initial
P-to-Ni ratio corresponding to the precursor materials was 2:1. However, the surface
P-to-Ni ratio of Ni2P/M41 was 3.14, which is higher than the initial P-to-Ni ratio.
This is due to the enrichment of phosphorous on the surface of the catalysts. Note
that upon addition of Na, the surface P-to-Ni ratio of Ni2P/M41 decreased
remarkably. With addition and increasing the Na mass fraction (\ 1.0), the P-to-Ni
ratio decreased. This shows that the incorporation of Na could suppress the
enrichment of P on the surface, which could lead to the increase in the exposed
nickel sites on the surface. The CO uptake analysis confirmed this. The Ni2P/
Na(2.0)M41 showed a slightly higher P-to-Ni ratio than that of Ni2P/Na(1.0)M41,
which possibly because the formation of high P-to-Ni ratio of Ni5P4 phase (Fig. 2,
XRD analysis). The surface Na-to-P ratio of Ni2P/Na(x)M41 decreased with
increasing the Na mass fraction. This is understandable since the surface P
decreases with Na mass fraction increase.
123

Synthesis of an Ni2P catalyst supported on Na-MCM-41 with…
TEM
TEM images of the Ni2P/M41, Ni2P/Na(0.7)M41, and Ni2P/Na(1.0)M41 catalysts
are shown in Fig. 5. It can be seen from Fig. 5a that the average Ni2P particle size of
the Ni2P/M41 catalyst is approximately 35 nm. Unlike the typical stacked
morphologies of Mo and W sulfides, Ni2P is not layered and forms spherical
particles that can be well dispersed on supports [28]. The average Ni2P particle sizes
of Ni2P/Na(0.7)M41 and Ni2P/Na(1.0)M41 were 25.0 nm and 26.0 mm, respec-
tively, which are smaller than that of Ni2P/M41. And the average Ni2P particle sizes
correspond with the results of XRD analysis (See Dc in Table 1). The Ni2P particle
sizes of Ni2P/Na(0.7)M41 (Fig. 5b) and Ni2P/Na(1.0)M41 (Fig. 5c) were more
uniform and an improved Ni2P dispersion can be seen in comparison with Ni2P/
M41. With increasing the Na mass fraction from 0.7 to 1.0, the particle sizes
changed little. Combined with XRD analysis, the increase of the Ni2P peak
intensities with increasing the Na mass fraction, one can conclude that addition of
Na promotes formation of more Ni2P particles on the surface.
Fig. 5 TEM spectra of a Ni2P/M41, b Ni2P/Na(0.7)M41, and c Ni2P/Na(1.0)M41
123

H. Song et al.
HDS activity
According to the literature [29], there are two parallel reaction pathways in the DBT
HDS reaction, which are direct desulfurization (DDS) and hydrogenation (HYD).
DDS is responsible for the formation of biphenyl (BP) and a small amount of the
formed BP further transforms into cyclohexylbenzene (CHB), while HYD is mainly
responsible for the formation of CHB. Because the transformation of BP into CHB
is negligible in the HDS reaction, the selectivity of DDS is almost equal to the BP
selectivity, and HYD selectivity is equal to that of the CHB selectivity.
The Ni2P/M41 and Ni2P/Na(x)M41 catalysts are evaluated by the HDS of DBT,
and the results are shown in Fig. 6. As can be seen from Fig. 6, the DBT
conversions for all samples increased with time on stream initially, during which
more active intermediate phase gradually formed, and then tended to stabilize,
showing the intermediate phase was completely formed. The active intermediate
phase is regarded as a superficial phosphosulfide with a stoichiometry represented
by NiPxSy, which is more active than Ni2P phase [21]. The DBT conversion of Ni2P/
M41 reached the value of about 80.3% after 8 h. It is noted that the DBT conversion
of Ni2P/Na(x)M41 initially increased and then decreased with increasing the Na
mass fraction. The Ni2P/Na(1.0)M41 catalyst showed the maximum HDS activity of
91.6% after 8 h, which was 14.0% higher than that of Ni2P/M41 without Na. This
result is contrary to the trend of the P-to-Ni ratio of Ni2P/Na(x)M41 (Table 2),
which has a maximum value at x = 1.0. The enhanced HDS activity may be
attributed to the smaller size (TEM) and better dispersion of the active Ni2P phase
(CO uptake), as well as the lower coverage of phosphorus on their surfaces (XPS
analysis).
Temperature, 613 K; pressure, 3.0 MPa; H2-to-oil ratio, 550 (V/V); WHSV,
.
2.5 h-1
The DBT conversions obtained at T = 553 K, p = 3.0 MPa, WHSV = 2.5 h-1
,
and H2-to-oil ratio = 550 (V/V), and CO uptake were used to calculate the HDS
TOF for the samples and the results are presented in Table 1. As can be seen from
Table 1, the HDS TOF of Ni2P/M41 is low (3.2 9 10-3 s-1). The TOF of the Ni2P/
Na(x)M41 significantly increased, reaches maximum at x = 1.0 (the TOF was
Fig. 6 The conversion of DBT
over the Ni2P/Na(x)M41
catalysts
123

Synthesis of an Ni2P catalyst supported on Na-MCM-41 with…
3.6 9 10-3 s-1), and then decreased upon increasing the Na mass fraction. This
indicates that more active Ni2P phase was generated by addition of Na.
The effect of Na element on the HDS catalytic selectivity was also investigated in
our study. Table 1 shows the selectivities of DBT over Ni2P/M41 and Ni2P/
Na(x)M41 to BP and CHB. For all the samples, BP is formed in greater proportions,
indicating that DBT primarily removed by the DDS pathway over all the catalysts
[30]. Moreover, the BP selectivity of Ni2P/M41 was 85.2%, and the BP selectivity
of Ni2P/Na(x)M41 catalyst decreased slightly with increasing the Na mass fraction.
This result indicates that addition of Na has not dramatically changed for the HDS
route [31].
Synthetic mechanism
It is well known that the reduction of metal nickel salt together with ammonium
phosphate requires high temperature (973–1273 K) even with hydrogen [32].
Sawada et al. [15] studied the effect of Na addition on rhodium phosphide (Rh2P)
formation over MFI zeolite, SiO2, and Al2O3; they found that the (NH4)2HPO4
transforms into a polyphosphate such as diphosphate, tripolyphosphate, and
tetrapolyphosphate over the MFI zeolite, and SiO2 during the calcination process.
The formation of AlPO4 upon calcination is inevitable, but Na addition could
weaken the interaction between the phosphate and Al. The reduction of polyphos-
phate is more difficult in comparison with the phosphate and the Na addition inhibits
formation of polyphosphates [16]. This indicates that the amount of polymerized P–
O–P bonds, which have stronger bonding than P–O–H in the precursor of phosphide
catalyst, is decreased by Na addition [15]. The Na preferentially interacts with
phosphate to suppress formation of stronger P–O–P bonds, which produced the
phosphide catalyst at a lower reduction temperature. Both of the TPR profiles and
XRD measurement showed that reducibility of phosphates in the Rh-P catalyst was
remarkably enhanced by the addition of Na in the MFI support [15]. They explain
the decrease in reduction temperature by addition of Na as Na interacts with
phosphate to form sodium phosphate, such as NaH2PO4, in the impregnation step of
catalyst preparation. Therefore, Na addition inhibits the formation of polyphosphate
and decreases the reduction temperature. On the basis of the above discussion, a
possible synthetic mechanism of the Ni2P/M41 and Ni2P/Na-M41 catalysts may be
proposed to explain the effect of Na addition on the reduction temperature
(Scheme 1). After calcination, Na species on Na-M41 existed in the form of Na
which interacted with M41 and some NaNO3. Therefore, the Na preferentially
interacts with phosphate to generate the sodium phosphate during the impregnation
step, and the formation of polyphosphate is inhibited. Hence, the addition of Na into
the catalyst dramatically decreases the reduction temperature of Ni2P/Na-M41
catalyst. For the Ni2P/Na-M41 catalyst, the Ni2P phase was generated at the
reduction temperature of 773 K, which is at least 200 K lower than the reduction
temperature of conventional phosphate (973–1273 K). However, without adding
Na, the (NH4)2HPO4 transforms into polyphosphate during the calcination process
and the resistance of polyphosphate reduction becomes larger. The strong P–O–P
123

H. Song et al.
Scheme 1 Proposed synthetic mechanism of the Ni2P/Na-M41 catalysts
bands are difficult to reduce at 773 K; therefore, the amorphous nickel pyrophos-
phate was obtained at this temperature (XRD analysis).
Conclusions
In this paper, Ni2P/Na(x)M41 catalysts with different Na mass fractions and a higher
HDS property have been successfully prepared at a lower reduction temperature
(773 K), and the effect of Na mass fraction in the Ni2P/Na(x)M41 on the HDS
activities was investigated. Both XRD analysis and TEM photos indicate the
incorporation of Na is beneficial for dispersion of Ni2P particles. During the
calcination process, the phosphate transforms into a polyphosphate such as
diphosphate, tripolyphosphate, and tetrapolyphosphate. Therefore, a high reduction
temperature is indispensable, mainly due to the polymerized P–O–P bonds in the
precursor of phosphide catalyst. The addition of Na into the catalyst can
dramatically decrease the reduction temperature of the catalyst (about 200 K).
This is possibly because the Na preferentially interacts with phosphate to generate
the sodium phosphate and, therefore, suppresses the formation of stronger P–O–P
bonds. As a result, the amount of polymerized P–O–P bonds in the precursor of
phosphide catalyst, which have stronger bonding than P–O–H, is decreased by Na
addition. In addition, the incorporation of Na suppresses the enrichment of P on the
surface of the catalyst and leads to more exposed Ni sites, which also improves the
123

Synthesis of an Ni2P catalyst supported on Na-MCM-41 with…
CO uptake. By this means, the reduction temperature of Ni2P/Na(x)-MCM-41
catalyst was 773 K, which is decreased by about 200 K as compared to that of the
conventional phosphate (973–1273 K). The Ni2P/Na(x)-MCM-41 catalyst with
x = 1.0 showed the maximum DBT HDS activity of 91.6%, which was 14.0%
higher than that of Ni2P/M41-773 without Na (80.3%).
Acknowledgements The authors acknowledge the financial supports from the National Natural Science
Foundation of China (21276048).
References
1. M. Egorova, R. Prins, J. Catal. 255, 417 (2004)
2. Q.Y. Li, P.Y. Wu, L. Lan, H. Liu, S.F. Ji, Catal. Today 216, 38 (2013)
3. H.Y. Zhao, S.T. Oyama, H.J. Freund, R. Włodarczyk, M. Sierka, Appl. Catal. B 164, 204 (2015)
4. X. Li, Z.C. Sun, A.J. Wang, X.N. Yang, Y. Wang, Appl. Catal. A 417, 19 (2012)
5. S. Tian, X. Li, A. Wang, R. Prins et al., Angew. Chem. Int. Edit. 55, 4030 (2016)
´
´
6. C.P. Jimenez-Gomez, J.A. Cecilia, R. Moreno-Tost et al., ChemCatChem 9, 2881 (2017)
7. H. Song, F.Y. Zhang, H.L. Song, X.W. Xu, F. Li, Catal. Commun. 69, 59 (2015)
8. Y.Y. Shu, S.T. Oyama, Carbon 43, 1517 (2005)
´
9. H. Loboue, C. Guillot-Deudon, A.F. Popa, A. Lafond, B. Rebours, C. Pichon, T. Cseri, G. Berhault,
C. Geantet, Catal. Today 130, 63 (2008)
10. K.S. Cho, H.R. Seo, Y.K. Lee, Catal. Commun. 12, 470 (2011)
11. H. Song, M. Dai, H.L. Song, X. Wan, X.W. Xu, Z.S. Jin, J. Mol. Catal. A- Chem. 385, 149 (2014)
12. H. Song, M. Dai, H.L. Song, X. Wan, X.W. Xu, C.Y. Zhang, H.Y. Wang, Catal. Commun. 43, 151
(2014)
13. H. Song, M. Dai, H.L. Song, X.X.W. Xu, Appl. Catal. A 462–463, 247 (2013)
14. A.I. d’Aquino, S.J. Danforth, T.R. Clinkingbeard, B. Ilic, L. Pullan, M.A. Reynolds, B.D. Murray,
M.E. Bussell, J. Catal. 335, 204 (2016)
15. A. Sawada, Y. Kanda, M. Sugioka, M. Uemichi, Catal. Commun. 56, 60 (2014)
´
´
16. H.I. Melendez-Ortiz, L.A. Garcıa-Cerda, Y. Olivares-Maldonado, G. Castruita, J.A. Mercado-Silva,
Y.A. Perera-Mercado, Ceram. Int. 38, 6353 (2012)
17. V. Teixeira da Silva, L.A. Sousa, R.M. Amorimb, L. Andrini, S.J.A. Figueroa, F.G. Requejo, F.C.
Vicentini, J. Catal. 279, 88 (2011)
18. Z.Y. Pan, R.J. Wang, Z.Y. Nie, J.X. Chen, J. Energy Chem. 25, 418 (2016)
19. X. Wang, P. Clark, S.T. Oyama, J. Catal. 208, 321 (2002)
20. S.T. Oyama, X. Wang, Y.K. Lee, W.J. Chun, J. Catal. 221, 263 (2004)
21. S.T. Oyama, X. Wang, Y.K. Lee, K. Bando, F.G. Requejo, J. Catal. 210, 207 (2002)
22. K.A. Layman, M.E. Bussell, J. Phys. Chem. B 108, 10930 (2004)
´
ˇ
23. I.K. Tamas, V. Zdenek, G.P. Dilip, R. Ryong, S.K. Hei, J.M.H. Emiel, J. Catal. 253, 119 (2008)
24. J.N. Kuhn, N. Lakshminarayanan, U.S. Ozkan, J. Mol. Catal. A 282, 9 (2008)
25. L. Song, S. Zhang, Q. Wei, Catal. Commun. 12, 1157 (2011)
26. H. Song, M. Dai, Y.T. Guo, Y.J. Zhang, Fuel Proces. Technol. 96, 228 (2012)
27. P. Liu, J.A. Rodriguez, J. Am. Chem. Soc. 127, 14871 (2005)
28. H. Song, J. Wang, Z.D. Wang, H.L. Song, F. Li, Z.S. Jin, J. Catal. 311, 257 (2014)
29. S.T. Oyama, T. Gott, K. Asakura, S. Takakusagic, K. Miyazakib, Y. Koiked, K.K. Bando, J. Catal.
268, 209 (2009)
30. H. Song, X.W. Xu, M. Dai, H.L. Song, Chem. J. Chin. Univ. 34, 2609 (2013)
31. S.T. Oyama, Y.K. Lee, J. Catal. 258, 393 (2008)
32. R. Prins, M.E. Bussell, Catal. Lett. 142, 1413 (2012)
123