Study on alumina-supported cobalt-nickel oxide catalyst for synthesis of acetonitrile from ethanol

Article abstract of DOI:10.1007/s10562-010-0468-z

A new alumina-supported cobalt-nickel oxide catalyst for the synthesis of acetonitrile from ethanol and ammonia was prepared by coprecipitation-kneading method. The parameters influencing the reaction were studied thoroughly and an optimized process, which is running the reaction at 380 °C under atmospheric pressure while keeping the ammonia/alcohol molar ratio of 5 and GHSV of 1,163 h^-1, was obtained. Under the optimized conditions the catalyst reached its best performance when being on stream for 40 h, at which the yield of acetonitrile was 92.6%. Then the selectivity to acetonitrile decreased gradually but the yield of acetonitrile always remained higher than 81% within 720 h. The samples of the fresh and used catalyst were characterized by XRD, XPS, TEM, EDX and N₂ adsorption-desorption analysis. The results revealed that carbon deposition and formation of metal carbides from the active species in the catalytic runs led to the deterioration of the catalyst. Graphical Abstract: A new Co-Ni/γ-Al₂O₃ catalyst for the amination of ethanol to acetonitrile was studied in a continuous fixed-bed reactor and the reaction parameters were optimized. The mechanisms for the formation of acetonitrile and butyronitrile were discussed. The reasons leading to the deterioration of the catalyst were investigated through the characterization of the fresh, used and regenerated samples of the catalyst by means of XRD, XPS, TEM and N₂ adsorption-desorption techniques.

Full text of DOI:10.1007/s10562-010-0468-z

Catal Lett (2011) 141:168–177
DOI 10.1007/s10562-010-0468-z
Study on Alumina-Supported Cobalt–Nickel Oxide Catalyst
for Synthesis of Acetonitrile from Ethanol
Cheng Feng
Yanlong Wen
•
Yuecheng Zhang
Jiquan Zhao
•
Yining Zhang
•
•
Received: 2 July 2010 / Accepted: 8 October 2010 / Published online: 26 October 2010
Ó Springer Science+Business Media, LLC 2010
Abstract A new alumina-supported cobalt–nickel oxide
catalyst for the synthesis of acetonitrile from ethanol and
ammonia was prepared by coprecipitation-kneading
method. The parameters influencing the reaction were
studied thoroughly and an optimized process, which is
running the reaction at 380 °C under atmospheric pressure
while keeping the ammonia/alcohol molar ratio of 5 and
1 Introduction
Acetonitrile is an important fine chemical product which has
been widely used as the synthetic intermediate for pharma-
ceutical, agricultural, and functional material chemicals [1].
It also used as a general purpose solvent for many com-
pounds and movable phase in high-performance liquid
chromatographic analysis. Currently, it is mainly produced
as a by-product during the propylene ammoxidation process
to acrylonitrile [2, 3]. With the increasing demand, it is dif-
ficult to meet the need in the future to obtain acetonitrile only
from the production of acrylonitrile. Furthermore, the pro-
duction of acetonitrile as a by-product of acrylonitrile has the
disadvantages of concomitance of toxic hydrogen cyanide
and difficulty in separation of the crude product [4]. There-
fore, it is important to develop a green and advanced process
to produce acetonitrile. Many methods have been reported to
synthesize acetonitrile from other chemical substances up till
now. The raw materials include carbon monoxide [5–8],
methane or other hydrocarbon compounds [9–14], ethanol
[15–21] and acetic acid [22, 23]. Among all the raw materials
ethanol becomes the first choice because it can provide a
process which avoids the concomitance of high toxic
hydrogen cyanide and gives high quality acetonitrile. Two
approaches so-called amination-dehygrogenation [20, 21]
and ammoxidation of ethanol [15–20] can be employed to
the transfer of ethanol to acetonitrile. However, only the
former can meet the need of avoiding formation of high toxic
hydrogen cyanide. In 1960s, Kryukov et al. [24] first reported
the conversion of alcohols to acetonitriles by amination-
dehygrogenation process over a fused iron catalyst. Since
then, several catalysts have been reported to transfer ethanol
to acetonitrile [4, 20, 25]. But all the catalysts suffer from a
disadvantage of either a high ammonia/ethanol ratio
dependence [20] or low acetonitrile selectivity [4]. Besides,
-
1
GHSV of 1,163 h , was obtained. Under the optimized
conditions the catalyst reached its best performance when
being on stream for 40 h, at which the yield of acetonitrile
was 92.6%. Then the selectivity to acetonitrile decreased
gradually but the yield of acetonitrile always remained
higher than 81% within 720 h. The samples of the fresh and
used catalyst were characterized by XRD, XPS, TEM,
EDX and N adsorption–desorption analysis. The results
2
revealed that carbon deposition and formation of metal
carbides from the active species in the catalytic runs led to
the deterioration of the catalyst.
Keywords Cobalt–nickel oxide Á Catalyst Á Acetonitrile Á
Ethanol Á Deterioration Á Metal carbides
C. Feng Á Y. Zhang Á Y. Zhang Á Y. Wen Á J. Zhao (&)
School of Chemical Engineering and Technology, Hebei
University of Technology, Tianjin 300130,
People’s Republic of China
e-mail: zhaojq@hebut.edu.cn
C. Feng
College of Science, Agriculture University of Hebei,
Baoding 071001, People’s Republic of China
1
23

Study on Alumina-Supported Cobalt–Nickel Oxide Catalyst
169
no clear information about the lifetime and structures of the
catalysts was mentioned in any of the reports.
of the samples were recorded with a Rigaku D/max 2500
X-ray diffractometer using Cu Ka radiation (40 kV, 150
mA) in the range 2h = 10°–90°. X-ray photoelectron
spectroscopy (XPS) was performed with a PHI 1600
spectroscope using Mg Ka X-ray source for excitation.
Transmission electron micrographs (TEM) were obtained
on a JEOL 100CX-II instrument equipped with an energy
dispersive X-ray (EDX) detector (Oxford Instruments) at
an accelerating voltage of 200 kV. Specimens for TEM
analysis were prepared by ultrasonic dispersion in n-buta-
nol where a drop of the resultant suspension was evapo-
rated on a lacey-carbon/Cu grid. The surface area, total
pore volume and pore size distribution of the samples were
measured at 77 K by nitrogen adsorption using a
Micromeritics ASAP 2020 Surface Area and Porosity
Analyzer. The crushing strength of the catalyst was mea-
sured on a ZQJ-II particle strength tester produced by
Dalian Intelligent Tester Plant. The applied force and idle
stroke velocity were automatically set on 5 N/s and 1 mm/s,
respectively.
Recently, we prepared an alumina-supported cobalt–
nickel oxide catalyst for the amination-dehygrogenation of
ethanol to acetonitrile by kneading method. The catalyst
showed high catalytic activity, good selectivity and long
lifetime [21]. However, the catalyst was fragile during the
course of regeneration for further investigation. For
improving the mechanical strength of the catalyst, we used
coprecipitation-kneading method instead of kneading
method to prepare a new alumina-supported cobalt–nickel
oxide catalyst. The catalyst also showed high activity, good
selectivity to acetonitrile and long lifetime as that prepared
by kneading method, but had good mechanical strength in
catalytic runs. Therefore, we characterized the catalyst
thoroughly to explore the deterioration reasons of the new
catalyst in catalytic runs. Herein, we present the results of
the study.
2
Experimental Section
The IR spectra of adsorbed pyridine were recorded using
a Thermo Nicliet Nicolet Nexus 470 spectrometer equip-
ped with a heatable and evacuatable IR cell with CaF2
windows, connected to a gas dosing-evacuating system.
The powdered samples were pressed into self-supporting
wafers with a diameter of 20 mm and a weight of 50 mg.
Prior to analysis, all samples were pretreated at 400 °C for
2
.1 Catalyst Preparation
All the chemicals were reagent grade. The alumina-sup-
ported cobalt–nickel oxide catalyst was prepared by
coprecipitation-kneading method as the following proce-
dures. 44.4 g of Co(NO ) Á6H O and 6.7 g of Ni(NO ) Á
5
1 h, under high vacuum conditions (5 9 10 Pa), followed
3
2
2
3 2
6
H O were dissolved in 302.6 g of distilled water, then into
2
by cooling to 200 °C. Then, pyridine was adsorbed at this
temperature for 15 min. The physisorbed pyridine was
removed by evacuating during 1 h at 200 °C, under high
this solution was dropped 20% (w/w) sodium carbonate
solution under vigorous stirring at 60 °C until the pH
reached 8. The reaction mixture was submitted to an aging
treatment at 80 °C for 0.5 h while maintained the pH of 8.
The resulting mixture was filtered and the solid separated
was washed with distilled water until the pH of the filtrate
reached 7 to remove the excess of sodium completely.
Afterwards, the resulting solid was dried and pulverized to
obtain a black powder finally. The black powder was mixed
with 34.7 g of c-Al O pretreated with about 10 mL of 1%
5
vacuum conditions (5 9 10 Pa). Then the infrared spectra
were recorded.
2.3 Catalytic Experiments
Catalytic tests were carried out in a continuous fixed-bed
reactor. 30.0 mL of the catalyst sample was loaded into a
reactor (i.d. = 12 mm; length = 1,100 mm), which created
a catalyst zone of 100 mm. The temperature in the catalyst
zone was kept constant and measured using a thermocouple
located in the center of the catalyst bed. The ethanol (95%)
was dosed into the reactor at a speed of 7.0 mL/h by a
syringe pump and the flux of ammonia was regulated by a
PID cascade controller (STP 220 mL/min). The liquid
products were separated from a gas–liquid separator and
analyzed by a gas chromatograph equipped with a 30 m of
PEG-20M capillary column and a thermal conductivity
detector (TCD) with H as carrier gas. The exhaust (H ,
2
3
diluted nitric acid. The mixture was kneaded for 3 h in a
kneader and the resulting kneaded material was then pro-
cessed in an extruder to obtain extrudates in a diameter of
2
mm and length of 2.5 mm. The extrudates were superfi-
cially dried at 120 °C for 12 h, and then calcined at 550 °C
for 4 h. The contents of Co and Ni were determined by
inductively coupled plasma (ICP) spectroscopy.
2
.2 Catalyst Characterization
2
2
For determining the metal content, the sample was first
NH , C H et al.) was analyzed by the same gas chro-
3 2 4
dissolved in concentrated HNO and HF, and then the
3
matograph with N as carrier gas. The components of the
2
metal content of the solution was analyzed by a T.J.A. ICP-
mixture were identified by a HP5971 MS equipped with a
30 m of PEG-20M capillary column.
9
000(N?M) type ICP-AES instrument. The XRD patterns
123

1
70
C. Feng et al.
2
.4 Regeneration Procedure
100
Regeneration of the used catalyst sample for further cata-
lytic evaluation was carried out by continuously blowing
air (STP 250 mL/min) into the fixed-bed at 470 °C for 5 h
on-line.
80
conversion of ethanol
selectivity of acetonitrile
selectivity of butyronitrile
selectivity of ethylene
selectivity of pyridine base
selectivity of higher boiling
components
6
0
2
1
0
0
0
3
Results and Discussion
3
.1 Catalytic Activity Test
The mechanical strength of the catalyst was evaluated by
measuring the crushing strength of the catalyst. The new
catalyst had a crushing strength of 20 N/mm higher than
that (13 N/mm) of the previously reported catalyst prepared
by kneading method [21]. The result indicated that the
catalyst had good mechanical strength as we expected. The
increase of the mechanical strength of the catalyst can be
attributed to the difference of the precursors between the
two catalysts. For the catalyst prepared by coprecipitation-
kneading method, the cobalt and nickel nitrates were
transferred to their basic cobalt carbonates in the copre-
cipitation process, therefore, the actual precursor was basic
cobalt and nickel carbonates. The decomposition behaviors
of nitrates and basic carbonates are different in calcination,
which led to the difference in matrix between the two
catalysts.
3
40
360
380
400
420
440
Reaction temperature (centigrade)
Fig. 1 The effect of reaction temperature on the performance of the
catalyst. Reaction conditions: NH /C OH molar ratio = 5,
1
GHSV = 1,163 h , pressure = 1 atm. The data were obtained at
3
2 5
H
-
the time when the catalyst was on stream for 40 h
conversion of ethanol increased with the increasing of
temperature from 340 to 380 °C, and reached 100% at
380 °C, and then maintained constant at the temperatures
higher than 380 °C. However, the selectivity of acetonitrile
reached its maximum of 92.6% at 380 °C and then
decreased significantly with the temperature increasing. As
shown in Fig. 1, the selectivity of high boiling point
compounds increased sharply from 0.02% at 380 °C to
20.9% at 420 °C. Other components had no significant
change with the reaction temperature increase. Therefore,
the decrease of the selectivity of acetonitrile can be mainly
attributed to the formation of high boiling point compounds
at high temperature.
The catalytic performance of the catalyst on the ami-
nation-dehygrogenation of ethanol to acetonitrile was tes-
ted after the strength measurement. First the catalytic test
was run under the optimized conditions used in the previ-
ous catalyst that are running the reaction at 420 °C under
the ammonia/ethanol ratio of 5 [21]. However, the yield of
acetonitrile was not good as we expected. Only about 70%
of yield of acetonitrile was received, which was much
lower than that catalyzed by the previous one. For probing
into the reasons causing the low yield of acetonitrile, we
analyzed the reaction mixture preciously by means of GC-
MS combined with distillation. It was found that the
mixture was composed of pyridines, ethylene, butyronitrile
and high boiling point compounds besides acetonitrile. The
results indicated that the reaction parameters could have
strong effects on the selectivity of acetonitrile. Therefore,
the effects of reaction parameters on the performance of the
catalyst were studied thoroughly. Because initial experi-
mental results had showed that the catalyst reached its best
performance when it was on stream for 40 h, all the opti-
mized parameters were determined at this moment in the
subsequent experiments.
Besides acetonitrile and high boiling point compounds,
the condensed components also included pyridine bases
and butyronitrile. It is a general phenomenon that pyridine
bases are concomitant in the synthesis of acetonitrile from
ethanol [26] or acetaldehyde [27]. However, it is interest-
ing that butyronitrile was detected with a selectivity as high
as 7.6% at 340 °C in the reaction mixture. Moreover, little
ethylamine was observed in the mixture, which was dif-
ferent from hydroamination of ethanol on Co/SiO [28] and
2
other catalysts [29]. The facts indicated that not the for-
mation of ethylamine but the dehydrogenation of ethanol to
acetaldehyde was the first step in the synthesis of aceto-
nitrile from ethanol and ammonia catalyzed by the catalyst
here. Once the generation of acetaldehyde, it acted as the
starting material to give acetonitrile, pyridine bases and
butyronitrile through a series of reactions. A proposed
mechanism for the formations of the compounds is shown
in Scheme 1. The acetonitrile was generated as the path-
way described in our previous report [21] and the forma-
tion of pyridines was similar to the mechanism given by
others [30–32]. The formation of butyronitrile is in need of
The influence of reaction temperature on the catalytic
performance of the catalyst was first investigated and the
results are shown in Fig. 1. It can be seen that the
1
23

Study on Alumina-Supported Cobalt–Nickel Oxide Catalyst
171
describing in detail. The butyronitrile was formed as fol-
lows. First, two molecules of acetaldehyde from the
dehydrogenation of ethanol underwent aldol condensation
on the acid sites of the catalyst to give crotonaldehyde.
After the crotonaldehyde molecules were formed, there
will be two possible pathways (A and B) as shown in
Scheme 1 to afford butyronitrile. In path A, crotonalde-
hyde condensed with ammonia to generate the intermediate
imine which underwent dehydrogenation to give crotonit-
rile and hydrogen intermediately. The hydrogen generated
from the dehydrogenation of the imine in situ hydrogenated
the carbon–carbon double bond in the crotonitrile molecule
generating butyronitrile. In path B, hydrogenation first took
place on crotonaldehyde to give butyroaldehyde, then the
butyroaldehyde condensed with ammonia followed by
dehydrogenation to give butyronitrile.
weakened the acidity of the catalyst, which brought down
the condensation of acetaldehyde with ammonia to imine
being a key step in the formation of acetonitrile as shown in
Scheme 1.
IR measurements of adsorbed pyridine were applied to
determine the acidity of the fresh catalyst and the results
-
1
are shown in Fig. 3. Two peaks at 1,545 and 1,450 cm
are appeared in the figure. In view of no species with
carbon–oxygen bond absorbing in the same region is
present in the fresh catalyst due to the calcination in the
catalyst preparation processes, the peaks can be attributed
to the characteristic absorptions of pyridine molecules
adsorbed on the Brønsted and Lewis acid centers of the
catalyst, respectively [33–35]. Therefore, it can be con-
cluded that high concentration of ammonia led to more
ammonia molecules adsorbed on the acid centers of the
catalyst and then slowed down the reaction rate in the
catalytic runs.
The influence of ammonia/alcohol molar ratio on the
catalytic reaction was also investigated and the results are
presented in Fig. 2. It was found that when the ammonia/
alcohol molar ratio increased from 2 to 5, meanwhile the
temperature was kept at 380 °C and the pressure was
For a given reaction that is run in a fixed-bed reactor,
gas hourly space velocity (GHSV) is another very
important parameter that one must consider. Therefore,
the catalytic performance of the alumina-supported
cobalt–nickel oxide catalyst was evaluated under different
GHSV. The results are shown in Fig. 4. It was found that
1
atm, the conversion of ethanol was always maintained
around 100%, but the selectivity of acetonitrile increased
from 76.5 to 92.6%. Figure 2 clearly shows that high
content of ammonia could prevent the formation of the
high boiling point compounds, which led to the increase of
the selectivity of acetonitrile. Of course it is not to say that
we can increase the ammonia/alcohol molar ratio unlimit-
edly for increasing the selectivity of acetonitrile. When the
ammonia/alcohol molar ratio was higher than 5, both the
conversion of ethanol and selectivity of acetonitrile had
tendencies of decrease. The reason may be that the excess
ammonia diluted the concentration of ethanol and also
-
1
when the GHSV was increased from 379 to 1,163 h
,
the conversion of ethanol was kept at 100% but the
selectivity of acetonitrile increased from 75.9% to its
maximum of 92.6%, and then both of them decreased in
some degree with the increase of GHSV. It was also
observed that low GHSV was in favor of the formation of
high boiling point compounds due to long detention time
of materials on the catalyst leading to oligomerization or
coking.
Scheme 1 The possible
mechanisms for the synthesis of
H2
C
NH2
H C
3
H O
2
H C CH2
2
H2
acetonitrile and the formation of
butyronitrile
OH
C
H
H2
NH3
H O
2
H2
C H OH
CH CHO
NH2
H C
3
C
H
NH
H C C N
3
2
5
3
H C
3
acid
CH3CHO
CH CHCH CHO
3
2
OH
NH3
H O
2
H C
3
C
C
CHO
B
H
H
A
H O
2
CH CH CH CHO
H3C
C
H
C C N
3
2
2
H2
H
NH3
CH CH CH CN
3
2
2
H O
2
123

1
72
C. Feng et al.
1
00
100
80
8
0
conversion of ethanol
selectivity of acetonitrile
selectivity of butyronitrile
selectivity of ethylene
selectivity of pyridine base
selectivity of higher boiling
components
conversion of ethanol
selectivity of acetonitrile
selectivity of butyronitrile
selectivity of ethylene
selectivity of pyridine base
selectivity of higher boiling
components
6
2
0
0
6
1
0
0
10
0
0
400
600
800 1000 1200 1400 1600 1800
2
3
4
5
6
-1
Ammonia/alcohol (molar ratio)
GHSV (h )
Fig. 2 The effect of molar ratio of ammonia to alcohol on catalytic
performance of the catalyst. Reaction conditions: T = 380 °C,
GHSV = 1,163 h , pressure = 1 atm. The data were obtained at
Fig. 4 The effect of GHSV on catalytic performance of the catalyst.
/C OH molar ratio = 5,
pressure = 1 atm. The data were obtained at the time when the
Reaction conditions: T = 380 °C, NH
3
2 5
H
-
1
the time when the catalyst was on stream for 40 h
catalyst was on stream for 40 h
1
1
2
0
8
6
4
2
0
conditions the yield of acetonitrile reached 92.6% at the
moment that the catalyst was on stream for 40 h. The above
results indicate that the new catalyst can afford an identical
yield of acetonitrile at a lower reaction temperature com-
pared with the previous one [21].
-
1
3.2 Lifetime Test
1
545 cm
Initial study showed that the new catalyst prepared by
coprecipitation-kneading method is an effective one for
the amination-dehygrogenation of ethanol to acetonitrile.
Subsequently, its lifetime and regeneration were investi-
gated and the results are shown in Fig. 5. It can be seen
that the conversion of ethanol was 100% initially and
maintained with the reaction time extension. However, the
selectivity of acetonitrile decreased gradually and dropped
to 81.3% from its maximum of 92.6% after the catalyst
had been run for 720 h. The catalyst was regenerated on-
line similar to the procedures reported by others [36–38]
and the reaction was run once again. The curve repre-
senting the changing of the selectivity of acetonitrile with
the running time of the regenerated catalyst almost
superposed with that of the fresh one, which indicated
that the catalyst was regenerated on-line successfully. We
can speculate that the main reason leading to the decrease
of selectivity of acetonitrile is the deposition of carbon
from coking which prevents the contact of reactants with
the active centers of the catalyst. This was confirmed by
the TEM analysis and will be discussed in the charac-
terization section. All results indicated that the new alu-
mina-supported cobalt–nickel oxide catalyst performed
excellently and will be suitable for application in large
scale.
-
1
1
450 cm
1
700 1650 1600 1550 1500 1450 1400 1350 1300
-1
Wavenumber (cm )
Fig. 3 IR spectrum of the pyridine absorbed on the fresh catalyst
The pressure effect was also investigated within the
range of atmospheric pressure to 0.3 MPa in the reaction. It
was noted that the selectivity of acetonitrile decreased
monotonically with increasing the pressure. In addition, the
amination-dehygrogenation of ethanol to acetonitrile is an
entropy increasing process, so increasing pressure is
unfavorable for the formation of acetonitrile. Therefore,
the reasonable pressure in the reaction was atmospheric
pressure.
From the study of parameter effects on the amination-
dehygrogenation of ethanol to acetonitrile, we obtained an
optimized process for synthesizing acetonitrile over the
alumina-supported cobalt–nickel oxide catalyst, which is
running the reaction at 380 °C under atmospheric pressure
while keeping the ammonia/alcohol molar ratio of 5 and
-
1
GHSV of 1,163 h , respectively. At the optimized
1
23

Study on Alumina-Supported Cobalt–Nickel Oxide Catalyst
173
1
1
05
00
#
+
a
b
# Co O
3
4
+
Al O
2
3
%
&
HOPG
Co
0
#
#
9
9
8
8
7
7
5
0
5
0
5
0
#
#
* CoO
#
#
+
+
+
#
#
d
&
#
+
#
+
%
%
#
&
#
&
c
#
+
&
*
#
#
#
+
+
#
+
&
*
#
b
Conversion of ethanol
Selectivity of acetonitrile
#
#
#
+
#
#
+
#
#
a
0
200
400
600
800
1000
1200
1400
Reaction time (h)
2
0
40
60
80
Two-Theta (degree)
Fig. 5 The ethanol conversion and the acetonitrile selectivity as
functions of reaction time. (a) The process of reaction on the fresh
catalyst, (b) the process of reaction on the regenerated catalyst.
Fig. 6 XRD patterns of the catalyst. (a) Fresh; (b) run for 40 h;
c) run for 720 h; (d) Regenerated
(
Reaction conditions: T = 380 °C, NH
GHSV = 1,163 h , pressure = 1 atm
3 2 5
/C H OH molar ratio = 5,
-1
two new species indicated that some of the Co O was
3
4
0
The new alumina-supported cobalt–nickel oxides cata-
lyst showed good performance over the previous one in the
aspect that the reaction can be run at lower temperature.
However, compared to the previous one [21] the new cat-
alyst must be on stream for about 40 h then could reach its
optimum state (see Fig. 5 and Ref. [21]). It is hard to say
why the two catalysts exhibited so obvious difference in
their catalytic properties. The only possible reason is that
cobalt nitrate and nickel nitrate were transferred to their
corresponding basic carbonates during the coprecipitation,
which might cause slight differences between the two
catalysts in matrix structure and crystalline phase of cata-
lytic active species during calcination processes due to
different decomposition behaviors of the nitrates and
carbonates.
transformed to CoO and Co after the catalyst was on
stream for 40 h. According to lifetime test (Fig. 5), the
highest yield of acetonitrile was found after the catalyst
was on stream for 40 h. Therefore, it may be concluded that
0
CoO and Co are active species of the catalyst too and are
more active than Co O . We can speculate that Co O as
3 4
3
4
the only active species catalyzed imine dehydrogenation to
acetonitrile and hydrogen in the beginning. Then Co O
3
4
0
was transformed to Co by the H generated in situ grad-
2
0
ually. Once the generation of Co , it accelerated the imine
dehydrogenation leading to the increase of the selectivity
of acetonitrile. High temperature could largely shorten the
0
transformation time of Co O to Co , therefore, the
3
4
induction period was not observed in the previous com-
munication [21].
Comparing the pattern of the sample run for 720 h with
that of the one run for 40 h, the peaks representing the CoO
crystalline phase were intensively weakened and the ones
3
.3 Catalyst Characterization
0
For elucidating the active species and deterioration mech-
anism of the catalyst in the catalytic run, the catalyst was
first characterized by XRD. Figure 6 shows the diffraction
patterns of the samples of the fresh, run for 40 h, run for
representing the Co crystalline phase strengthened, dem-
0
onstrating that most of CoO was transformed to Co .
Meanwhile, an additional peak attributed to the highly
oriented pyrolytic graphite (HOPG) [36, 41] appeared,
indicating the growth of HOPG around the catalyst similar
to the carbon deposition on the active sites in the ethyl-
benzene dehydrogenation [36] and CO /CH reforming
7
20 h and regenerated catalyst. The patterns of the fresh
catalyst revealed that the fresh catalyst had a Co O crys-
3
4
talline phase same as that of the previous one reported by
us recently [21]. The sharpness of the XRD peaks indicated
the higher crystallinity of Co O in the fresh catalyst. After
2
4
reaction processes [42], which led to the decrease of
selectivity to acetonitrile. After regeneration, the deposed
3
4
0
carbon was removed and the Co was oxidized to Co O
the catalyst was run for 40 h, the peaks of the Co O
3
4
3
4
crystalline phase in the XRD patterns were intensively
weakened and some of which were not detectable. Mean-
while, several new peaks attributed to the CoO [39] and
again, therefore, the regenerated catalyst displayed the
same behavior as the fresh one (Fig. 5).
Besides the peaks in the patterns discussed above, two
peaks at 45.7° and 66.4° corresponding to those of c-Al O
0
Co [40] were observed. The appearance of the peaks of the
2
3
123

1
74
C. Feng et al.
[
40] were also found. However, no phase associated with
of the used catalyst also displayed a much lower C 1s
binding energy at 281.4 eV. The smaller binding energy of
C 1s in the used catalyst indicated that the carbon had
combined with Co or Ni in the catalytic runs [45–48].
Therefore, it may be concluded that the formation of metal
carbides in the catalytic runs is another reason for the
decrease of selectivity to acetonitrile of the catalyst. The
binding energy of Ni in the catalyst was detected near
854.7 eV, indicating that nickel was present as NiO [49] in
the fresh and regenerated catalysts. However, the binding
energy of Ni in the used catalyst was about 851.5 eV,
the nickel was identified, which indicated that the content
of the Ni or the crystalline related to Ni in the catalyst was
too small to be detected. The regenerated catalyst displayed
diffraction patterns identical to those of the fresh one,
0
indicating the transformation of Co to Co O and removal
3
4
of carbon by combustion during the regeneration process,
therefore, showed catalytic performance same as that of the
fresh one.
The XPS technique is much more sensitive than XRD
for the analysis of surface oxides. Therefore, the surface
compositions of the fresh, used and regenerated samples
were determined by XPS. Figure 7 shows the analysis
results. Both the fresh and regenerated samples showed that
the binding energy of Co 2p was 779.77 eV, indicating that
the Co was present as Co O [43]. For the used sample,
0
indicating that nickel was present as Ni . The metal content
of the fresh catalyst was analyzed by ICP-AES mentioned
above. The results showed the mass content of Co and Ni
was 20.75 and 3.12%, respectively.
The morphological features of the fresh, used and
regenerated samples of the alumina-supported cobalt–
nickel oxide catalyst can be assessed from the TEM
images as shown in Fig. 8. In all the samples the dark
portions attributed to Co O particles were seen in the
3
4
after a curve-fitting procedure was applied, the binding
energy of 777.33 eV was observed clearly, which indicated
the presence of metal cobalt [44]. This result is in agree-
ment with the XRD analysis above. Figure 7 also shows
the presence of the binding energy of C 1s. All the samples
had a C 1s binding energy around 284 eV derived from the
carbon contamination in the analysis. However, the sample
3
4
TEM images and dispersed on the surface of the c-Al O
2
3
with diameters of 20–30 nm. Besides, in the used catalyst
partial Co O particles were surrounded by the carbon as
4
3
Fig. 7 XPS patterns of the
catalysts. (a) Fresh; (b) used;
Co 2p
C 1s
3.5
3.0
2.5
Energy: 281.401
Energy: 779.767
(
c) regenerated
2.0
Energy: 284.050
Energy: 777.570
c
Energy: 780.531
Energy: 779.766
2.0
1.5
c
Energy: 284.350
b
a
1.5
Energy: 284.500
b
a
1.0
0.5
0.0
810
800
790
780
770
760
292 290 288 286 284 282 280 278 276
Binding energy (eV)
Binding energy (eV)
4
×
10
Co2p
Ni 2p
1
.95
Pos. Sep. %Area
77.33 0.00 55.68
780.31 2.98 44.32
b
Energy: 854.800
7
1
.9
c
1.6
1.85
Energy: 851.456
b
1.8
Energy: 854.750
1.75
1.4
a
1.7
1.65
784
782
780
778
776
774
890
880
870
860
850
840
Binding Energy (eV)
Binding energy (eV)
1
23

Study on Alumina-Supported Cobalt–Nickel Oxide Catalyst
175
Fig. 8 TEM micrographs of
a
b
a fresh, b, c used, d regenerated
samples and e EDX spectra on
points 1–3
1
2
3
c
d
eO Al
Co
C
Co
Cu
Cu
Ni
Ni
3
Co
O
Co
Al
C
2
1
C
Co
O
Co
7
NiCu
8
Al
0
1
2
3
4
5
6
9
10
Energy (KeV)
shown in the top left corner of Fig. 8b, and carbon-coated
cobalt nanocapsules were observed in Fig. 8c. These were
supported by the EDX analysis (Fig. 8e); element C was
mainly detected at the bright point 1 and Co and O were
mainly present at the dark point 2 as well as Al and O
were mainly found at point 3 by EDS analysis(Fig. 8e).
The carbon deposition phenomena were also observed in
the literatures [43, 50, 51]. TEM analysis directly
confirmed that deterioration of the alumina-supported
cobalt–nickel oxide catalyst was mainly caused by the
growth of carbon around cobalt oxide concluded from the
XRD analysis. In addition, the Co O particle size of the
3
4
regenerated sample (30 nm) was larger than those of the
fresh and used catalyst samples (20 nm), indicating that
cobalt oxide was partly sintered during the regeneration
process.
123

1
76
C. Feng et al.
Table 1 summarizes textural properties of the alumina-
0.06
0.04
Pore Diameter: 8.3176
Pore Diameter: 8.3956
supported cobalt–nickel oxide catalyst. Compared to the
fresh catalyst, the used catalyst possessed a relatively large
Pore Diameter: 7.2842
2
3
surface area (197.79 m /g), a small pore volume (0.50 cm /
g) and pore diameter (8.93 nm). The drop in the pore
volume could be the result that a certain amount of carbon
was deposited in the pores of the catalyst in the catalytic
runs. Table 1 also demonstrates that regenerated catalyst
had a lower surface area and pore volume than those of the
fresh one, which can be explained by the fact that the
cobalt oxide was partly sintered during the regeneration
process, as detected by TEM investigation.
0
0
.02
.00
a
b
c
There was a strange phenomenon that the surface area of
the used catalyst was larger than that of the fresh one and
could be recovered after regeneration. Based on the XRD,
XPS and TEM results, it could be concluded that the
increase of the surface area was brought about by the
deposition of activated carbon with large surface area on
the framework of the catalyst [51, 52]. In order to
strengthen this inference, the porous structure of the cata-
lyst was further analyzed. The adsorption–desorption iso-
therms are show in Fig. 9. All the samples displayed type
IV isotherms. In addition, the hysteretic loop of the used
0
10
100
0
10
100
0
10
100
Pore Diameter (nm)
Fig. 10 Pore size distribution curves as a function of pore diameter.
a) Fresh; (b) used; (c) regenerated
(
sample occurred with no bulge while the fresh and regen-
erated ones occurred with it, indicating that structural
properties of the used catalyst were similar to the type H4
hysteretic loop, which was a indicative of highly micro-
porosity with slit-like pores as the characteristic of acti-
vated carbon [53]. A pore size distribution of the catalyst
was determined using the BJH desorption method too
Table 1 Textural properties of the alumina-supported cobalt–nickel
oxide catalysts determined by N adsorption–desorption analysis
2
(
Fig. 10). The major differences between the fresh and
used samples were the decrease of the maximum distri-
bution of pore diameter and the increase of micropores in
the used catalyst, which indicated that many micropores as
the feature of activated carbon [54, 55], were formed
during the catalytic runs. These results confirmed further
that porous carbon deposition took place during the reac-
tion process.
a
2
-1
b
3
V (cm g
-1
c
S
BET (m g
)
)
p
d (nm)
Fresh
176.63
197.79
164.67
0.52
0.50
0.49
10.30
8.93
Used
Regenerated
10.67
a
BET surface area
b
c
BJH cumulative desorption pore volume
Mean pore diameter = 4 V/SBET
4
Conclusions
5
5
4
4
3
3
2
2
1
1
50
00
50
00
50
00
50
00
50
00
The alumina-supported cobalt–nickel oxide catalyst for
catalytic amination of alcohol to acetonitrile prepared by
coprecipitation-kneading method showed high mechanical
strength compared to the one prepared by kneading method
reported previously. The parameters which effect on the
performances of the catalyst were studied thoroughly and
an optimized process for synthesizing acetonitrile from
ethanol and ammonia over the catalyst was obtained. Under
the optimized conditions the catalyst can last for 720 h to
afford acetonitrile with a yield higher than 81%. The cat-
alytic performance can be recovered on-line by flowing of
air into the catalyst bed at 470 °C. The reasons leading to
the deterioration of the catalyst were studied through the
characterization of the catalysts by means of XRD, XPS,
TEM, EDX and N2 adsorption–desorption techniques.
Characterization results revealed that the deterioration of
desorption
adsorption
c
b
a
50
0.0
0.2
0.4
0.6
0.8
1.0
Relative Pressure (P/P0)
Fig. 9 Adsorption–desorption isotherms of the catalyst. (a) Fresh;
b) used; (c) regenerated
(
1
23

Study on Alumina-Supported Cobalt–Nickel Oxide Catalyst
177
the catalyst was attributed to the carbon deposition on the
catalyst and the formation of metal carbides from the active
species in the catalytic run.
24. Kryukov YB, Bashkirov AN, Zakirov NS, Novak FL (1966) Dokl
Akad Nauk SSSR 170:852
2
2
5. Pan WX (1992) CN Patent 1,062,303, Tsinghua University
6. Kulkarni SJ, Ramachandra Rao R, Subrahmanyam M, Rama Rao
AV (1994) Appl Catal A 113:1
Acknowledgments We thank the financial support by the National
Natural Science Foundation of China (Grant no. 20976034) and the
Natural Science Foundation for Young Scientists of Hebei Province,
China (Grant no. B2009000009) and Guide Project for the Devel-
opment of Science and Technology of Hebei Province, China (Grant
no. 072156136).
27. Kumar RJ, Joshi PN, Chapekar GM (2005) WO Patent
2,005,000,816A, Council of Scientific and Industrial Research,
Jubilant Organosys Ltd
28. Rausch AK, van Steen E, Roessner F (2008) J Catal 253:111
29. Neylon MK, Bej SK, Bennett CA, Thompson LT (2002) Appl
Catal A 232:13
3
3
0. Higashio YS, Shoji T (2004) Appl Catal A 260:251
1. Ramachandra Rao R, Srinivas N, Kulkarni SJ, Subrahmanyam M,
Raghavan KV (1997) Appl Catal A 161:L37
References
3
3
2. Jin F, Cui YG, Li YD (2008) Appl Catal A 350:71
3. Volckmar CE, Bron M, Bentrup U, Martin A, Claus P (2009) J
Catal 261:1
4. Barthos R, Solymosi F (2005) J Catal 235:60
5. Modrogan E, Valkenberg MH, Hoelderich WF (2009) J Catal
1
2
3
. Iida S, Togo H (2007) Tetrahedron 63:8274
. Grasselli RK, Callahan JL (1969) J Catal 14:93
. Andersson A, Andersson SLT, Centi G, Grasselli RK, Sanati M,
Trifiro F (1993) Stud Surf Sci Catal 75:691
. Potaraju R, Vishnu JM (1986) EP Patent 0,206,632, Indian
Explosives Limited, ICI
3
3
2
61:177
4
5
6
3
6. Freire RM, de Sousa FF, Pinheiro AL, Longhinotti E, Mendes
Filho J, Oliveira AC, de Tarso C, Freire P, Ayala AP, Oliveira
AC (2009) Appl Catal A 359:165
7. Martin A, Narayana Kalevaru V, L u¨ cke B, Br u¨ ckner A (2008)
Appl Catal A 335:196
. Oliv e´ G, Oliv e´ S (1979) US Patent 4,179,462, Monsanto Com-
pany, St. Louis, MO
3
3
3
4
4
. Hummel AA, Badani MV, Hummel KE, Delgass WN (1993) J
Catal 139:392
8. Wang HF, Chen LG, Bai GY, Luan DZ, Li Y, Yan XL, Zhang
YC, Xing JD (2007) Catal Commun 8:145
9. Seehra MS, Dutta P, Manivannan A (2005) Div Fuel Chem
0:638
0. Tavasoli A, Malek Abbaslou RM, Dalai AK (2008) Appl Catal A
46:58
1. Zhang DS, Shi LY, Fang JH, Li XK, Dai K (2005) Mater Lett
9:4044
7
8
. Eshelman LM, Delgass WN (1994) Catal Today 21:229
. Badani MV, Eshelman LM, Delgass WN (1993) Stud Surf Sci
Catal 75:1223
5
9
. Sandner WJ, Fierce WL (1961) CA Patent 615,929, Pure Oil
Company
3
1
1
1
1
0. Fierce WL, Lake C, Sandner WJ (1957) US Patent 2,802,020,
Pure Oil Company
1. Krebaum LJ (1964) US Patent 3,129,241, Monsanto Chemical
Company, St. Louis, MO
2. Fierce WL, Sandner WJ (1959) CA Patent 586,008, Pure Oil
Company
3. Fierce WL, Lake C, Sandner WJ (1962) US Patent 3,056,826,
Pure Oil Company
5
4
4
2. Ruckenstein E, Wang HY (2002) J Catal 205:289
3. Wang ZH, Choi CJ, Kim BK, Kim JC, Zhang ZD (2003) Carbon
4
1:1751
4
4
4
4
4
4. Bechara R, Balloy D, Vanhove D (2001) Appl Catal A 207:343
5. Waldrop JR, Grant RW (1990) Appl Phys Lett 56:557
6. Kuivila CS, Butt JB, Stair PC (1988) Appl Surf Sci 32:99
7. Sinharoy S, Levenson LL (1978) Thin Solid Films 53:31
8. Ramqvist L, Hamrin K, Johansson G, Fahlman A, Nordling C
1
1
4. Morikawa K, Shiba T (1961) GB Patent 870,462A, State of Japan
5. Callahan JL, Stamm JA (1963) US Patent 3,106,574, Standard
Oil Company
(
1969) J Phys Chem Solids 30:1835
1
1
6. Stamm JA, Callahan JL (1965) CA Patent 718,939A, Standard
Oil Company
7. Zhang HT, Zhang YN, Chen J (2005) CN Patent 1,226,281C,
Zhang HT
4
5
9. Li CP, Proctor A, Hercules DM (1984) Appl Spectrosc 38:880
0. Wang ZH, Zhang ZD, Choi CJ, Kim BK (2003) J Alloys Compd
3
61:289
1. Nieto-M a´ rquez A, Valverde JL, Keane MA (2009) Appl Catal A
52:159
2. Alfred GC, Yardley P, Partha SG, Princeton NJ (1991) US Patent
,037,791, HRI, Inc
3. Hu ZH, Guo HM, Srinivasan MP, Yaming N (2003) Sep Purif
Technol 31:47
4. Hu ZH, Srinivasan MP (2001) Microporous Mesoporous Mater
5
5
5
5
5
1
1
8. Reddy BM, Manchar B (1993) J Chem Soc Chem Commun 3:234
9. Kulkarni SJ, Ramachandra Rao R, Subrahmanyam M, Rama Rao
AV (1994) J Chem Soc Chem Commun 3:273
0. Roger JC, Lawrence JS (1981) J Org Chem 46:754
1. Zhang YN, Zhang YC, Feng C, Qiu CJ, Wen YL, Zhao JQ (2009)
Catal Commun 10:1454
3
5
2
2
2
2
2. Kurina LN, Golovko AK, Galanov SI, Sidorova OI (2003) RU
Patent 2,214,396C2, Inst Khim Nefit So Ran
3. Hou XM, Yan L, Zhang X, Li YW, Li CB (2009) J Jilin Norm
Univ (Nat Sci Ed) 30:108
43:267
5. Inagaki M, Radovic LR (2002) Carbon 40:2279
123