A novel route to the preparation of carbon supported nickel phosphide catalysts by a microwave heating process

Article abstract of DOI:10.1007/s10562-010-0280-9

A simple and efficient approach based on microwave heating process was developed to prepare carbon supported nickel phosphide. In this approach, red phosphorus was used as a P source and carbon acted as both the support and the microwave absorbent. The red phosphorus was homogeneously mixed with Ni-impregnated carbon by milling, and then subjected to microwave heating. After several minutes by microwave heating in Ar or H₂ atmosphere, the nickel phosphide, Ni₂P, was produced on the carbon support, while the temperature of the sample bed was only 473 K or even lower during the reaction. It was also found that the preparation atmosphere had significant effects on the phosphide formation. Compared to the preparation in Ar, the nickel phosphides prepared in H₂ were more readily formed and more highly dispersed on the carbon support due to PH₃ formation during the reduction process. The as-prepared nickel phosphide catalysts exhibited much higher activities in selective hydrogenation of 1,3-butadiene compared to that prepared by the conventional heating method, which was attributed to the high dispersion of Ni₂P prepared by the microwave heating method.

Full text of DOI:10.1007/s10562-010-0280-9

Catal Lett (2010) 135:305–311
DOI 10.1007/s10562-010-0280-9
A Novel Route to the Preparation of Carbon Supported Nickel
Phosphide Catalysts by a Microwave Heating Process
Lining Ding
Tao Zhang
•
Mingyuan Zheng
•
Aiqin Wang
•
Received: 15 December 2009 / Accepted: 18 January 2010 / Published online: 5 February 2010
Ó Springer Science+Business Media, LLC 2010
Abstract A simple and efficient approach based on
microwave heating process was developed to prepare car-
bon supported nickel phosphide. In this approach, red
phosphorus was used as a P source and carbon acted as
both the support and the microwave absorbent. The red
phosphorus was homogeneously mixed with Ni-impreg-
nated carbon by milling, and then subjected to microwave
heating. After several minutes by microwave heating in Ar
or H atmosphere, the nickel phosphide, Ni P, was pro-
1 Introduction
Over the past decade, many research communities have
shown great interest in transition metal phosphides due
to their promising catalytic performances in a variety of
hydrotreating reactions [1–11]. For instance, nickel phos-
phide (Ni P) has been found to be highly active for hydro-
2
genation [2, 3], hydrodenitrogenation (HDN) [4, 11], and
hydrodesulfurization (HDS) [4–10]. Stimulated by these
potential applications, innovations in synthesis methods
have been greatly encouraged, such as reduction of phos-
phates [1–11], phosphine method, [12, 13] solvothermal
synthesis using elemental P as the phosphorus source
[14–16], thermal decomposition of hypophosphites [17],
and thermal decomposition of metal-phosphine complexes
[18, 19]. For the purpose of catalysis, the catalyst must have
a large surface area so as to create active sites as many as
possible. In this respect, the temperature programmed
reduction (TPR) of phosphates which was first developed by
Oyama et al. [11], seems to be the most effective method.
2
2
duced on the carbon support, while the temperature of the
sample bed was only 473 K or even lower during the
reaction. It was also found that the preparation atmosphere
had significant effects on the phosphide formation. Com-
pared to the preparation in Ar, the nickel phosphides pre-
pared in H were more readily formed and more highly
2
dispersed on the carbon support due to PH formation
3
during the reduction process. The as-prepared nickel
phosphide catalysts exhibited much higher activities in
selective hydrogenation of 1,3-butadiene compared to that
prepared by the conventional heating method, which was
attributed to the high dispersion of Ni P prepared by the
2
However, the formation process of Ni P from the phos-
2
microwave heating method.
phates by TPR method is neither thermodynamically nor
kinetically favorable, and thereby requires a high tempera-
ture and a slow heating rate [20]. Typically, the TPR process
Keywords Carbon ꢀ Nickel phosphide ꢀ Red phosphorus ꢀ
Microwave ꢀ High dispersion ꢀ Hydrogenation ꢀ
for preparing a Ni P catalyst takes 8 h or even longer.
2
1
, 3-butadiene
Clearly, such a low preparation efficiency needs to be
improved greatly.
It has been established that microwave irradiation is a
low energy cost and high efficiency heating mode. Com-
pared with conventional heating mode, the microwave
heating often results in higher yields and shorter reaction
time in material synthesis [21–25]. Especially, it presents
unique advantages in synthesis of carbon-based materials
because carbon itself is a good microwave absorbent. For
example, interconnected cable-like Ag/C [26], spherical
L. Ding ꢀ M. Zheng ꢀ A. Wang ꢀ T. Zhang (&)
State Key Laboratory of Catalysis, Dalian Institute of Chemical
Physics, CAS, 116023 Dalian, People’s Republic of China
e-mail: taozhang@dicp.ac.cn
L. Ding
Graduate School of the Chinese Academy of Sciences,
1
00049 Beijing, China
123

3
06
L. Ding et al.
Se/C nanocomposites [27], superparamagnetic porous
Fe O /C nanocomposites [28] and nearly monodispersed
2.2 Catalyst Characterization
3
4
nanostructured tungsten carbide particles on carbon nano-
tubes [29], have been successfully prepared by microwave
irradiation. However, there has been no any report so far, to
the best of our knowledge, on the synthesis of carbon
supported transition metal phosphides by microwave irra-
diation in spite of their excellent catalytic performances in
HDS [30].
The XRD patterns of the catalysts were recorded on a
PW3040/60 X’ Pert PRO (PANalytical) diffractometer
equipped with a Cu Ka radiation source (k = 0.15432 nm),
operating at 40 kV and 40 mA. The morphologies of
samples were observed with transmission electron
microscopy (TEM) using a JEOL JEM-2000EX micro-
scope operated at 120 kV. The gaseous products during the
synthesis were analyzed with an in situ mass spectrometer
(Omini-star, GSD-300). The chemisorption of CO was
conducted on a Micromeritics autochem II 2910 automated
catalyst characterization system. Usually, 0.1 g of a sample
Herein we describe a new and facile method for syn-
thesis of carbon-supported Ni P catalysts by microwave
2
heating. In this synthesis, activated carbon (AC) acts both
as the support to disperse Ni P nanoparticles and as the
2
microwave absorbed heating medium, while red phospho-
rus is used as the P source because it can readily react with
metal halide at a relatively low temperature to form metal
phosphide [12–14]. With the aid of microwave irradiation,
the whole reaction process took only less than 7 min, sig-
nificantly more efficient than the conventional heating.
was loaded in a quartz reactor and pretreated in H at
2
773 K for 2 h. After the sample was cooled in He stream,
pulses of 5% CO/He were dosed in and the metal disper-
sions were auto-calculated accordingly. Both the metal and
P contents were determined by inductively coupled plasma
(ICP) on an IRIS Intrepid II XSP instrument (Thermo
Electron Corporation).
More important, the Ni P/AC catalysts prepared by this
2
method exhibited superior performance in the selective
hydrogenation of 1,3-butadiene in contrast to that prepared
via the conventional heating method.
2.3 Catalytic Activity Tests
The catalytic performances of the Ni P/AC samples were
2
evaluated for selective hydrogenation of 1, 3-butadiene.
The reaction was conducted in a fixed-bed continuous-flow
reactor at ambient pressure. Prior to the reaction, 0.05 g of
2
Experiments
2
.1 Catalyst Preparation
the catalyst was reduced in H
2
at a flow rate of
0 mL min at 673 K for 2 h and cooled to room tem-
perature in H . Then, the reacting gas containing 1 vol%
1,3-butadiene, 50% H and balance N was allowed to pass
-
1
2
To prepare Ni P/AC precursor, 3.0 g of the AC powders
2
2
2
Norit, SBET = 709 m /g) were impregnated with an
aqueous solution of 1.677 g of nickel chloride hydrate
(
2
2
-
1
through the reactor at a flow rate of 30 mL min . The
reaction products were on line analyzed by gas chroma-
tography (Agilent 6890N) equipped with a flame ionization
detector and a DSP capillary column.
(
NiCl ꢀ6H O), and then dried at 298 K overnight and at
2
2
3
93 K for another 3 h in air. 0.2 g of this precursor was
then milled with red phosphorus at a desired Ni/P ratio in
an agate mortar to give the Ni P/AC precursor.
2
The as-prepared precursor was loaded in a quartz tube
reactor which was aligned vertically at the center of the
microwave cavity so that the region was seated in the
3 Results and Discussion
microwave field at maximum intensity [31]. Ar or H flow
2
3.1 Characterization of the Catalysts
Before being subjected to the microwave irradiation, the
was allowed to pass through the reactor at a rate of
-
50 mL min . The temperature of the reaction bed was
1
1
measured using an IR imager. The reaction was initiated by
microwave irradiation at the working power of 50 W, and
stopped until no white fume released from the reactor.
After the reaction, the sample was cooled down to ambient
temperature and then passivated in 0.5% O /N
red phosphorus was homogeneously mixed with NiCl -
2
impregnated AC support by milling. XRD pattern of the
precursor with Ni/P ratio of 1/3 shows diffraction lines of
well-crystallized NiCl
ꢀ2H
O phase (Fig. 1). The absence
2
2
of red phosphorus in the XRD pattern is due to its amor-
phous nature. This result indicates no reaction occurs
2
2
-
1
(
20 mL min ) for 3 h.
For comparison, a conventional heating method was also
between NiCl and red phosphorus at this stage. Upon
2
employed to prepare Ni P/AC catalyst. The precursor was
2
being microwave irradiation in the Ar atmosphere, how-
ever, a fast chemical reaction took place accompanied with
release of a significant amount of white fume. During the
reaction, some light sparkles occurred randomly
heated in a flow reactor system as described above, but
-
1
with an electric oven at a rate of 2 K min and held at the
set temperature for 90 min.
1
23

A Novel Route to the Preparation of Carbon Supported Nickel Phosphide Catalysts
307
♦
Ο
♦
♦
•
♦
•
(a)
♦
•
•
♦
♦
∗
♦
Ο
•
•
•
•
•
(b)
•
∗
∇
∗
∇
(c)
10
20
30
40
50
60
70
80
2
Theta/Degree
10
20
30
40
50
60
70
80
2
Theta/Degree
Fig. 1 XRD pattern of the precursor with Ni/P molar ratio of 1/3.
filled circle): NiCl O; (open circle): activated carbon
ꢀ2H
(
2
2
Fig. 2 XRD patterns of Ni
Ni/P ratios by microwave heating in Ar. a Ni/P = 1/2, b Ni/P = 1/3,
c Ni/P = 1/5. (filled diamond): Ni P; (asterisks): Ni ; (filled circle):
2
P/AC catalysts prepared with different
throughout the catalyst bed due to electronic discharge
between the sharp points in the carbon support [32], but no
obvious plasma was observed in the heating zone. Mean-
while, the temperature of the catalyst bed was rapidly
increased up to 433 K within the beginning 2 min and then
kept basically at this temperature in the following steady
stage. The white fume produced from the heated catalyst
bed was condensed at the outlet of the quartz reactor,
forming some light yellow powders which are identified as
white phosphorus because they will ignite upon exposed to
air. After a period of less than 7 min, the white fume van-
ished and the reaction was stopped accordingly. Figure 2
presents the XRD patterns of the resultant samples from
precursors with different Ni/P ratios. At the molar ratio of
Ni to P being 1/2 (Fig. 2a), strong diffraction peaks cor-
2
5 4
P
NiCl
2
; (inverse triangle): NiP
2
; (open circle): activated carbon
♦
♦
Ο
♦
♦
(
e)
♦
♦
♦♦
∗
∗
∗
∗
∗∗
(d)
∗∗
∇
∗∗ ∗
∗
∗
∗
∇
(c)
(b)
•
•
•
•
•
(a)
10
20
30
40
50
60
70
80
2
Theta/Degree
responding to Ni P are observed, in addition to a small
2
Fig. 3 XRD patterns of Ni
temperatures by conventional TPR method in Ar with Ni/P = 1/3:
a 623 K, b 723 K, c 773 K, d 823 K, e 923 K. (filled diamond): Ni P;
; (inverse triangle): NiP
2
P/AC catalysts prepared at different
peak at 2h = 15.1° which is identified as NiCl ꢀ2H O. This
2
2
2
result indicates that most NiCl reacted with red phos-
2
(
(
asterisks): Ni
5
P
4
; (filled circle): NiCl
2
2
;
phorus under microwave irradiation, forming well-crys-
tallized Ni P phase. Nevertheless, a small fraction of NiCl
open circle): activated carbon
2
2
was still intact. The reason for this is that the sublimation
of red phosphorus took place simultaneously with the
phosphorus-rich nickel phosphides. It is also observed that
the intensities of XRD peaks decrease with increasing the P
content, suggesting that the excess P may lead to a decrease
of the crystallinity of nickel phosphides. Evidently, Ni/P =
1/3 is the optimum ratio for the synthesis of phase-pure
reaction with NiCl , as indicated by the release of white
2
fume and formation of yellow powders at the outlet of the
reactor. The sublimation of red phosphorus is inevitable,
which leads to insufficient amount of P that can be used for
Ni P. It should be pointed out that in the three investigated
2
reaction with NiCl . To promote the total conversion of
2
samples, the formation of Ni P seems to be inevitable
5
4
NiCl to nickel phosphide, the amount of P must further be
2
although its amount is very tiny.
increased.
For comparison, conventional TPR method was also
employed to prepare nickel phosphide catalyst in Ar
atmosphere. The molar ratio of Ni to P was fixed at 1/3. As
shown in Fig. 3a, even being treated at 623 K for 90 min,
As expected, at the molar ratio of Ni/P = 1/3 (Fig. 2b),
the peak corresponding to NiCl ꢀ2H O disappears and
2
2
diffraction peaks of Ni P are very strong, indicating the Ni
2
species is completely transformed into nickel phosphides at
this Ni/P ratio. With a further increase of the P amount to
the sample still presented NiCl diffraction peaks and no
2
any nickel phosphides were detected by XRD. Further
increasing the temperature to 723 K (Fig. 3b) led to the
formation of mixed phases of Ni P and Ni P , and the
Ni/P = 1/5 (Fig. 2c), the intensities of the Ni P have some
2
decrease while two new peaks at 32.7° and 55.7° corre-
2
5 4
sponding to NiP phase appear, suggesting too much excess
unreacted NiCl could still be detected at this temperature.
2
2
phosphorus in the precursor favors the formation of
Until up to 773 K, the diffraction peaks of NiCl is no
2
123

3
08
L. Ding et al.
longer observed in the XRD patterns, indicating the Ni
species in the precursor was completely transformed into
Ni P, as well as a small amount of Ni P and NiP . When
Ο
2
5
4
2
♦
the precursor was treated at the temperature as high as
23 K, phase-pure Ni P was obtained probably due to
♦
♦
♦
♦
♦
(a)
♦
♦♦
9
∗
♦
2
♦
♦
decomposition of Ni P and NiP into Ni P.
5
4
2
2
∇
∇
(b)
♦♦
∗
∇
♦
In comparison with the above conventional TPR
method, the microwave irradiation manifests remarkable
advantages. The reaction conditions are rather mild. The
nickel phosphide can be formed at 433 K, which is around
♦♦
(c)
♦
♦∇
∗
10
20
30
40
50
60
70
80
3
00 K lower than that needed in the conventional heating
2
Theta/Degree
method. The low temperature can effectively prevent the
Ni P particles from sintering. Moreover, the reaction pro-
2
Fig. 4 XRD patterns of Ni
ratios by microwave heating in H
c Ni/P = 1/5. (filled diamond): Ni
triangle): NiP ; (open circle): activated carbon
2
P/C catalysts prepared with different Ni/P
. a Ni/P = 1/2, b Ni/P = 1/3,
P; (asterisks): Ni ; (inverse
2
cess under microwave irradiation is highly efficient since it
takes only several minutes to finish the reaction in contrast
to more than 5 h in the TPR method. According to litera-
ture [21], the excellent microwave absorbing characteris-
tics of carbon would promote the formation of some local
2
5 4
P
2
consistent with the results obtained in Ar atmosphere that
much excess phosphorous in the precursor favored the
formation of phosphorus-rich phosphides. For comparison,
conventional TPR method in H2 atmosphere was also
employed to prepare nickel phosphide catalyst. Phase-pure
‘‘hot spots’’ under the microwave irradiation. The reaction
takes place around the hot spots whereas the average
temperature of the whole catalyst bed is much lower. On
the other hand, microwave heating provides a uniform
reaction environment, which would facilitate the mass
transfer and then speed up the synthesis process. These
unique features of microwave irradiation can not be
achieved by conventional heating.
Ni P phase could form only at or above 923 K (the XRD
2
patterns are not shown here).
The gaseous products during the preparation were on
line analyzed with mass spectroscopy to probe the reaction
mechanism. As shown in Fig. 5a, the signals with
Since the nickel phosphide is usually used as a hydro-
treating catalyst, preparing it in a reducing atmosphere
would most likely produce a catalyst with highly reduced
surface and would be greatly beneficial to the catalytic
reactions. Subsequently, we investigated the synthesis of
nickel phoshpides under microwave irradiation in H2
atmosphere. Similar to the phenomena observed during the
preparation in the Ar atmosphere, some sparkles were
generated randomly in the catalyst bed upon the microwave
irradiation. The white fume of phosphorous was formed
m/z = 31 (P), 18 (H O), and 37 (HCl) were detected with
2
the reaction time in the Ar atmosphere. In the first 2 min,
only H O was detected, which originates from the crys-
2
talline water associated with nickel chloride. After 2 min
of reaction, both HCl and P began to be detected. The
reaction proceeded only about 2 min since no any other
compound could be detected except H O. It is well known
2
that white phosphorous is more reactive than the red one,
and we did observe the formation of white phosphorus
(white fume) during the reaction. Therefore, we propose
and then carried away with H stream. No evident plasma
2
that upon microwave irradiation, some local hot spots are
produced on the carbon support due to the excellent
microwave absorbing properties of the carbon. On or
around these hot spots, red phosphorus transformed into
more reactive white phosphorus, which then reacted with
was produced during the reaction. A noticeable difference
from the case in Ar atmosphere is the bed temperature. The
temperature of the catalyst bed was 473 K, which was
4
0 K higher than that in the case of Ar atmosphere. The
difference in the catalyst bed temperature may be caused
NiCl according to the following pathway:
2
by the different physical properties of H and Ar.
2
Figure 4 presents the XRD patterns of samples prepared
3NiCl þ 2P ¼ 3Ni þ 2PCl
ð1Þ
ð2Þ
ð3Þ
2
3
with different Ni/P molar ratios in H . In contrast to the
2
2Ni þ P ¼ Ni2P
case in Ar where the molar ratio of Ni/P must be smaller
than 1/2 for the complete transformation of NiCl to nickel
2
PCl3 þ 3H2O ¼ H3PO3 þ 3HCl
phosphide, the total transformation of NiCl occurred even
2
In this reaction pathway, the white phosphorus reduces
2
?
0
first Ni to Ni , and then reacts with the Ni to produce
0
at Ni/P = 1/2 in H atmosphere. Clearly, the presence of
2
H facilitates the formation of nickel phosphides. Further
2
nickel phosphides. At the same time, the PCl formed in
3
increasing the amount of phosphorus to the molar ratio of
the first step reacts with H O to form gaseous HCl. It
2
Ni/P = 1/3 and 1/5 led to the appearance of NiP2 in
should be pointed out that in the whole reaction process no
any carbon-containing gases, such as CH , CO, or CO
2
addition to the predominant phase of Ni P. This is
2
4
1
23

A Novel Route to the Preparation of Carbon Supported Nickel Phosphide Catalysts
309
(
a)
(b)
M=18
M=34
M=37
M=18
M=37
M=31
0
100
200
300
400
500
600
0
100
200
300
400
500
Reaction Time (s)
Reaction Time (s)
Fig. 5 Profiles of compounds with m/z of 18 (H
b in the H atmosphere
2
O), 31(P), 34 (PH
3
), and 37(HCl) in the outlet gas during the synthesis a in the Ar atmosphere;
2
were detected, indicating that the carbon support is inert at
this temperature (the maximum bed temperature was only
precursor was lost via volatilization during the preparation,
and the loss of phosphorus was even more pronounced in
the conventional TPR process.
4
33 K in the case of microwave). This is quite different
from our previously reported Ni P where the support car-
2
On the other hand, the Ni/P molar ratios in the four
samples are all apparently lower than 2, the stochiometric
bon acted also as the reducing agent for the phosphide
formation at 1,073 K [33].
ratio of Ni to P in the Ni P product, suggesting that some
2
When the Ni P/AC catalyst was prepared in the H
2
surplus phosphorus remains on the carbon support. More-
2
atmoshpere, besides the mass of 18 (H O) and 37 (HCl),
2
over, the Ni P/AC (microw-H ) has the P content one-third
2
2
the mass of 34 (PH ) was also detected (Fig. 5b). However,
3
less than that of the Ni P/AC (microw-Ar). This might be
2
no PH was formed when the red phosphorus was solely
3
related to the phosphine formation during the preparation
in the hydrogen atmosphere, which can readily remove the
loaded on the carbon support without nickel chloride. This
result strongly suggests that the formation of PH3 is
facilitated by the presence of Ni. Based on this, the possible
reaction pathway for the formation of nickel phosphides in
P from the catalyst precursor. Ni P/AC (oven-Ar) and
2
Ni P/AC (oven-H ), on the other hand, have similarly low
2 2
P content to that of the Ni P/AC (microw-H ) probably due
2
2
the H atmosphere can be depicted as following:
2
to the long heating process allowing more phosphorus
being removed.
NiCl2 þ H2 ¼ Ni þ 2HCl
ð4Þ
ð5Þ
ð6Þ
The CO adsorption measurements reveal the remarkable
difference in the number of active sites over these catalysts.
2
P þ 3H2 ¼ 2PH3
The CO uptake on the Ni P/AC (microw-Ar) is
2
4
Ni þ 2PH3 ¼ 2Ni2P þ 3H2
-
1
9
.0 lmol g , approximately three times larger than that
on the Ni P/AC (oven-Ar) catalyst. More intriguing, the
CO uptake on the Ni P/AC (microw-H ) attains up to
24.8 lmol g , which doubles that on the Ni P/AC (oven-
) catalyst. Evidently, the microwave irradiation provides
the catalysts with higher dispersion due to the low bed
temperature and uniform heating mode. The H atmosphere
results in a much higher dispersion than the Ar thanks
probably to the formation of PH . It was previously
reported that silica-supported Ni P catalyst with high dis-
persion could be obtained by reducing Ni/SiO in PH and
[12, 13].
Figure 6 presents the TEM images of the four samples.
It can be clearly seen that the particles on the Ni P/AC
According to this proposed mechanism, NiCl was first
2
2
reduced by hydrogen to zero-valence nickel, and then the
2
2
-
1
metallic nickel catalyzed the reaction betwen H and P to
2
2
form PH . As reported in literature [12, 13], phosphine was
3
H
2
a good P source for the nickel phosphide preparation.
The nickel phosphide catalysts prepared under four
typical conditions, i.e., Ni P/AC (microw-H ), Ni P/AC
2
2
2
2
(
microw-Ar), Ni P/AC (oven-H ), and Ni P/AC (oven-Ar),
2
3
2
2
were characterized with various techniques. The Ni/P
molar ratios in the precursors are all 1/3. As shown in
Table 1, the actual Ni/P ratio in the catalysts strongly
depends on the preparation method. ICP analyses show that
the samples prepared under microwave irradiation have the
2
2
3
H
2
2
Ni/P molar ratio of 1:0.96 (in H ) and 1:1.32 (in Ar), while
2
prepared by conventional heating are not uniform (Fig. 6a,
d). Especially in Ar (Fig. 6a), both spherical and one-
dimensional nanorods are observed, and particle sizes vary
in the range of 20–200 nm. In contrast, the samples
the samples prepared under conventional heating have the
Ni/P molar ratio of 1: 0.88 (in H ) and 1:1.13 (in Ar). In
2
each case, more than half amount of phosphorus in the
123

3
10
L. Ding et al.
Table 1 Characterization results of various catalysts
a
c
Particle sizes (nm)
-1
CO uptake (lmol g )
Catalyst
Ni/P molar ratio
Ni
Ni
Ni
Ni
a
2
2
2
2
P/AC (microw-H
2
)
1:0.96
1:1.32
1:1.13
1:0.88
10–20
20–40
20–200
10–80
24.8
9.0
P/AC (microw-Ar)
b
P/AC (oven-Ar)
2.9
b
P/AC (oven-H
2
)
12.0
Molar ratio of Ni to P in the precursors is always fixed at 1/3
b
c
Ni
2 2 2
P/AC (oven-Ar) and Ni P/AC (oven-H ) were prepared at 923 K by the conventional heating method
Particles sizes were estimated by the TEM images
to 1-butene and 2-butene were 31–38% and 60–65%,
respectively, totally up to more than 96% towards the
formation of butenes. Compared to the predominant
occurence of 1,2-addition giving rise to 1-butene as the
major product over transition metals such as Ni, Co, Pt, and
Pd, [35] a distinctive feature of the Ni P lies in that the
2
selectivity to 2-butene is much higher than that to 1-butene,
indicating 1,4-addition governs the hydrogenation process
over Ni P catalysts. This is consistent with the previous
2
report [35, 36] that the presence of electronegative element
such as S or P could modify the electronic property of
transition metals and thereby reduce the ratio of 1-butene to
2-butene in the hydrogenation products of 1,3-butadiene. It
is also noted that the ratios of trans-/cis-2-butene vary in
the range of 2–3, which is in agreement with that reported
on transition metals including Ni, suggesting the isomeri-
zation from trans-butadiene to cis-butadiene takes place
facilely on either Ni or Ni P [35]. On the other hand, the
2
amount of P imposed a great effect on the catalytic per-
formance of Ni P/AC catalysts. For example, the Ni P
2
2
(1/2)/AC catalyst prepared with microwave heating in Ar
converted only 7.9% of 1, 3-butadiene at 573 K. The poor
activity of this catalyst is due to the incomplete transfor-
Fig. 6 TEM images of a Ni P/AC (oven-Ar), b Ni
2
2
P/AC (microw-
mation of NiCl which may cover the active sites of the
2
Ar), c Ni P/AC (microw-H ), and d Ni
2
2
P/AC (oven-H )
2 2
Ni P. In contrast, the Ni P (1/3)/AC catalyst exhibited
2
2
pretty good activity for the hydrogenation of 1,3-butadiene.
However, comparing Ni P (1/3)/AC and Ni P (1/5)/AC
prepared by microwave irradiation have uniform and small
particle sizes, 10–20 nm for Ni P/AC (microw-H ), and
2
2
catalysts, one can see that the excess P in the catalyst
lowered the activity but meanwhile increased the selec-
tivity to 2-butene and the trans-/cis-2-butene ratio in the
2
2
2
0–40 nm for Ni P/AC (microw-Ar). The TEM results are
2
well consistent with the CO chemisorption data.
product. Furthermore, the activities of the Ni P/AC cata-
2
3
.2 Catalytic Performances in Selective Hydrogenation
of 1, 3-Butadiene Reaction
lysts prepared at the fixed Ni/P ratio of 1/3 depend strongly
on the preparation methods. For example, at a reaction
temperature of 433 K, the conversion of 1, 3-butadiene
reached 93.6% over the Ni P/AC (microw-H ) catalyst,
The catalytic performances of the Ni P/AC samples were
2
2
2
evaluated for selective hydrogenation of 1, 3-butadiene,
which is a subject of industrial importance for highly pure
butenes production for the polymer industry [34]. The
while it was only 65.3% over the Ni P/AC (oven-H ). Only
2 2
at an elevated reaction temperature (e.g., 463 K) was the
conversion of 1, 3-butadiene over the Ni P/AC (oven-H )
2
2
results are shown in Table 2. The four Ni P/AC catalysts
2
more than 90%. Similarly, the Ni P/AC (microw-Ar)
2
with nominal Ni/P molar ratio of 1/3 exhibited a high
selectivity towards partial hydrogenation, the selectivities
behaved more actively than the Ni P/AC (oven-Ar), but
2
less than the Ni P/AC (microw-H ). At the same Ni/P ratio
2
2
1
23

A Novel Route to the Preparation of Carbon Supported Nickel Phosphide Catalysts
311
2
Table 2 Results of selective hydrogenation of 1,3-butadiene over different Ni P/AC catalysts
a
Catalyst
Reaction temp. (K) Conv. (%) Sel. of butene (%) Selectivity (%)
n-Butane 1-Butene trans-2-Butene cis-2-Butene
Ni
Ni
Ni
Ni
Ni
Ni
Ni
Ni
Ni
a
2
2
2
2
2
2
2
2
2
P (1/3)/AC (microw-H
2
)
433
433
473
573
573
433
463
433
523
93.6
51.1
91.1
7.9
97.2
97.7
97.5
2.8
2.3
2.5
0.0
0.0
2.6
3.7
4.1
2.3
34.0
33.8
37.6
3.1
46.8
48.1
41.5
3.0
16.4
15.8
18.4
1.8
P(1/3)/AC (microw-Ar)
P(1/3)/AC (microw-Ar)
P(1/2)/AC (microw-Ar)
P(1/5)/AC (microw-Ar)
100
100
97.4
48.9
65.3
92.6
36.5
94.6
13.9
34.6
36.8
31.3
35.3
25.6
45.8
40.1
49.6
45.6
9.4
P(1/3)/AC (oven-H
P(1/3)/AC (oven-H
2
)
)
17.0
19.4
15.0
16.8
2
96.3
95.9
97.7
P(1/3)/AC (oven-Ar)
P(1/3)/AC (oven-Ar)
2
The figures in parenthesis after the Ni P indicate the molar ratios of Ni to P in the preparation precursors
of 1/3 (in the precursor), the activities follow the order of
5. Burns AW, Gaudette AF, Bussell ME (2008) J Catal 260:262
6
7
8
. Wang X, Clark P, Oyama ST (2002) J Catal 208:321
. Sawhill SJ, Phillips DC, Bussell ME (2003) J Catal 215:208
. Sawhill SJ, Layman KA, Van Wyk DR, Engelhard MH, Wang C,
Bussell ME (2005) J Catal 231:300
Ni P/AC (microw-H ) [ Ni P/AC (oven-H ) [ Ni P/AC
2
2
2
2
2
(
microw-Ar) [ Ni P/AC (oven-Ar), totally in line with the
2
dispersion of the active sites as measured by CO adsorp-
tion. Therefore, we can conclude that microwave irradia-
9. Wang A, Ruan L, Teng Y, Li X, Lu M, Ren J, Wang Y, Hu Y
2005) J Catal 229:314
0. Kor a´ nyi TI, V ´ı t Z, Poduval DG, Ryoo R, Kim HS, Hensen EJM
2008) J Catal 253:119
11. Li W, Dhandapani B, Oyama ST (1998) Chem Lett 3:207
(
tion in the H atmosphere is a facile and effective way to
2
1
prepare carbon supported phosphide catalysts with high
dispersion and excellent catalytic performance.
(
1
1
2. Yang S, Prins R (2005) Chem Commun 33:4178
3. Yang S, Liang C, Prins R (2006) J Catal 237:118
4
Conclusion
14. Xie Y, Su H, Qian X, Liu X, Qian Y (2000) J Solid State Chem
49:88
5. Aitken JA, Ganza-Hazen V, Brock SL (2005) J Solid State Chem
78:970
6. Barry BM, Gillan EG (2008) Chem Mater 20:2618
1
1
1
A novel method based on microwave heating has been
developed for synthesis of activated carbon supported
1
nickel phosphide from precursors containing NiCl and red
2
17. Xing G, Li W, Zheng M, Tao K (2009) J Catal 263:1
1
8. Perera SC, Tsoi G, Wenger LE, Brock SL (2003) J Am Chem Soc
25:13960
9. Hu X, Yu J (2008) Chem Mater 20:6743
phosphorus. In contrast to the high temperature and long
reaction time needed in the conventional TPR method, our
present microwave approach has the advantages of short
1
1
2
0. Wang A, Qin M, Guan J, Wang L, Guo H, Li X, Wang Y, Prins
R, Hu Y (2008) Angew Chem Int Ed 47:6052
1. Galema SA (1997) Chem Soc Rev 26:233
2. Zhu Y, Wang W, Qi R, Hu X (2004) Angew Chem Int Ed
4
reaction time and low bed temperature. The Ni P phase was
2
2
2
formed even in several minutes at a low temperature of
4
33–473 K. The atmosphere also imposed an important
3:1410
23. Panda AB, Glaspell G, El-Shall MS (2006) J Am Chem Soc
128:2790
effect on the dispersion of Ni P particles. H leads to higher
2
2
dispersion of active sites than Ar due to the formation of
2
2
4. Celer EB, Jaroniec M (2006) J Am Chem Soc 128:14408
5. Xu L, Ding Y, Chen C, Zhao L, Rimkus C, Joesten R, Suib SL
PH in the former case, and then results in a higher activity
3
for 1,3-butadiene hydrogenation. This method can also be
extended to synthesis of other transition metal phosphides.
(
2008) Chem Mater 20:308
2
2
6. Yu J, Hu X, Li Q, Zhang L (2005) Chem Commun 21:2704
7. Yu J, Hu X, Li Q, Zhang Z, Xu Y (2005) Chem Eur J 12:548
Acknowledgments Supports from the Natural Science Foundation
of China (NSFC Nos. 20573108, 20773124, 20773122) and from the
National Basic Research Program of China (2009CB226102) are
gratefully acknowledged.
28. Hu X, Yu J (2006) Chem Asian J 1:605
29. Liang C, Ding L, Wang A, Ma Z, Qiu J, Zhang T (2009) Ind Eng
Chem Res 48:3244
30. Shu Y, Oyama ST (2005) Carbon 43:1517
3
3
1. Tang J, Zhang T, Liang D, Sun X, Lin L (2000) Chem Lett 8:916
2. Wang X, Wang A, Wang X, Zhang T (2007) Energy Fuels 21:867
References
33. Wang H, Shu Y, Wang A, Wang J, Zheng M, Wang X, Zhang T
2008) Carbon 46:2076
(
3
3
4. Boitiaux JP, Cosyns J, Vasudevan S (1983) Appl Catal 6:41
5. Moyes R, Well P, Grant J, Salman NY (2002) Appl Catal A Gen
229:251
6. Nozaki F, Adachi R (1975) J Catal 40:166
1
2
3
4
. Oyama ST (2003) J Catal 216:343
. Wang H, Shu Y, Zheng M, Zhang T (2008) Catal Lett 124:219
. Nozaki F, Tokumi M (1983) J Catal 79:207
. Oyama ST, Wang X, Lee YK, Bando K, Requejo FG (2002) J
Catal 210:207
3
123