Ni-Co-Cu supported on pseudoboehmite-derived Al2O3: Highly efficient catalysts for the hydrogenation of organic functional groups

Article abstract of DOI:10.1016/j.apcata.2012.03.003

Ni-Co-Cu catalysts supported on pseudoboehmite-derived Al₂O ₃ were developed for efficient hydrogenation of organic functional groups: CN and NO₂ associated with benzene as well as CC in butynediol. Factors such as thermo-treatment of pseudoboehmite, Ni loading, Cu and Co content, and catalyst operating parameters were systematically studied. Among the catalysts, 20%Ni-3%Cu-5%Co/Al₂O₃ performed the best, showing 100% feed conversion and 84-99.8% selectivity of target products. Techniques such as N₂ sorption measurement, scanning electron microscope, hydrogen temperature-programmed reduction and desorption, X-ray photoelectron spectroscopy, and in situ X-ray diffraction were used to characterize the catalysts. The superior performance of the catalyst can be attributed to synergetic interaction among the Ni-Cu-Co constitutes as well as enhancement of Ni²⁺ reduction and Ni⁰ dispersion that are resulted due to the co-presence of Cu and Co.

Full text of DOI:10.1016/j.apcata.2012.03.003

Applied Catalysis A: General 425–426 (2012) 68–73
Contents lists available at SciVerse ScienceDirect
Applied Catalysis A: General
journal homepage: www.elsevier.com/locate/apcata
Ni–Co–Cu supported on pseudoboehmite-derived Al O : Highly efficient
2
3
catalysts for the hydrogenation of organic functional groups
a
a
a
a,∗
b,∗∗
Tianbao Shi , Hui Li , Lianghong Yao , Weijie Ji , Chak-Tong Au
a
Key Laboratory of Mesoscopic Chemistry, MOE, School of Chemistry and Chemical Engineering, Nanjing University, Nanjing 210093, China
Department of Chemistry, Hong Kong Baptist University, Hong Kong, China
b
a r t i c l e i n f o
a b s t r a c t
Article history:
Ni–Co–Cu catalysts supported on pseudoboehmite-derived Al O were developed for efficient hydro-
2
3
Received 18 August 2011
Received in revised form 26 February 2012
Accepted 1 March 2012
genation of organic functional groups: CN and NO2 associated with benzene as well as
C
C
in
butynediol. Factors such as thermo-treatment of pseudoboehmite, Ni loading, Cu and Co content, and cat-
alyst operating parameters were systematically studied. Among the catalysts, 20%Ni–3%Cu–5%Co/Al2O3
performed the best, showing 100% feed conversion and 84–99.8% selectivity of target products.
Techniques such as N2 sorption measurement, scanning electron microscope, hydrogen temperature-
programmed reduction and desorption, X-ray photoelectron spectroscopy, and in situ X-ray diffraction
were used to characterize the catalysts. The superior performance of the catalyst can be attributed to
Available online 9 March 2012
Keywords:
Hydrogenation
Nitro, nitrile and alkynyl groups
Dispersed Ni–Cu–Co catalyst
2
+
synergetic interaction among the Ni–Cu–Co constitutes as well as enhancement of Ni reduction and
0
Ni dispersion that are resulted due to the co-presence of Cu and Co.
©
2012 Elsevier B.V. All rights reserved.
1
. Introduction
In the cases of Pt–Ni/Al O3 and Co/Ni–Ti/diatomite catalysts,
2
the reactions were operated under rather severe conditions (tem-
◦
Meta-xylylenediamine (m-XDA) is widely used as an important
perature up to 180 C and pressure to 22 MPa). It is known that
intermediate of fine-chemicals and a curing agent of epoxy. It is
mainly produced by the hydrogenation of isophthalonitrile (IPN).
In the presence of catalysts, successive hydrogenation of IPN occurs
both Ni and Co are elements active for catalyzing nitriles hydro-
genation. For nitrile compounds, the adsorption strength is weak
on Pt and Pd, medium on Cu and Rh, and strong on Co and Ni.
The adding of certain amount of Cu to Ni can purposely modify
the electronic property of Ni; meanwhile the electronic property of
Ni–Co metals can be tuned through the control of Ni/Co ratio. The
change in electronic property of metal constitutes can in turn bring
about alteration in adsorption and reaction behavior of reactant.
Besides the hydrogenation of IPN to m-XDA, the hydrogenation of
nitrobenzene (NB) to aniline, butynediol (BD) to 2-butane-1,4-diol
(BDO), and 4-nitrophenol (4-NP) to 4-aminophenol are also impor-
tant processes in chemical industry. The corresponding products
are essential chemicals or fine chemical intermediates widely used
in the fields of medicine, pesticides, dye, and perfume; the demand
for them is getting higher and higher in recent years.
The current industrial processes for these products, however,
are either environment-unfriendly (for instance, the hydrogena-
tion of NB by Fe powder with acid [9]) or energy-demanding (high
temperature and pressure adopted in the Reppe process for BD
hydrogenation [10–13]). The motivation of this study is to develop
hydrogenation catalysts that are efficient yet less costly, and can be
applied to organic molecules with different functional groups. The
strategy of catalyst design is: (1) to develop a kind of supported Ni-
based catalyst using comparatively lower Ni loading; (2) to select
pseudoboehmite-derived Al2O3 as support and to tune its property
to produce single- (with one CH NH2 group) and di-amine (with
2
meta-CH NH groups in benzene ring); under severe conditions,
2
2
there is condensation of m-XDA product via NH2 groups to gen-
erate the condensed by-products. The reaction can be catalyzed
by a variety of catalysts such as Raney Ni [1], Raney Co [2], Fe-
or Cr-doped Raney Ni [3,4], Pt–Ni/Al O3 [5], Co–Ti/diatomite or
2
Ni–Ti/diatomite [6], Co–Rh/diatomite [7], and Rh/Al O [8]. In these
2
3
supported Ni or Co catalysts, high loading (up to 60 wt%) is usu-
ally adopted. The reported maximum m-XDA yields over Raney
Ni, modified Raney Ni, and dispersed Ni catalysts were approxi-
mately 85%, 95%, and 95%, respectively. The m-XDA yield, however,
was strongly dependent on reaction parameters such as IPN con-
centration, catalyst amount, reaction temperature and pressure.
Generally speaking, Co is more selective than Ni for nitrile for-
mation, but also more costly than Ni. Also, the skeletal-type Ni
catalyst is more expensive than the supported Ni counterpart and
deactivates in the reaction.
∗
Corresponding author. Tel.: +86 25 83686270.
Corresponding author. Tel.: +852 34117810.
E-mail addresses: jiwj@nju.edu.cn (W. Ji), pctau@hkbu.edu.hk (C.-T. Au).
∗
∗
0
926-860X/$ – see front matter © 2012 Elsevier B.V. All rights reserved.
doi:10.1016/j.apcata.2012.03.003

T. Shi et al. / Applied Catalysis A: General 425–426 (2012) 68–73
69
by controlled thermo-treatment; (3) selecting Co and Cu as addi-
tional components to form ternary Ni–Co–Cu catalysts; and (4) to
generate synergetic interaction among Ni–Co–Cu constitutes for
enhancement of catalytic activity. It’s the first time to report the
o
4
00 C
versatility of the optimized Ni–Co–Cu/Al O3 catalyst in highly effi-
2
cient hydrogenation of different organic functional groups (
NO , and N) under mild conditions.
C C ,
o
C
500 C
2
o
5
50 C
2
. Experimental
o
6
00 C
2.1. Catalyst preparation
o
6
50 C
The Al O support was obtained through the calcination of pseu-
2
3
◦
doboehmite precursor in air at 550 C for 4 h. The Ni, Co, and Cu
components were introduced onto the support separately or at the
same time by dipping/wetness impregnation methods. The catalyst
o
7
00 C
◦
precursors were dried in air at 100 C overnight and activated first
20
40
60
80
◦
◦
at 400 C in air for 4 h and then at 400 C in H flow (150 ml/min)
2θ/°
2
for 5 h.
Fig. 1. XRD patterns of pseudoboehmite precursor subject to thermo-treatment at
different temperatures.
2.2. Catalyst characterization
Both the catalyst precursors and the activated catalysts were
characterized by means of N2 sorption measurement, scanning
electron microscope (SEM), hydrogen temperature-programmed
stirring speed of 500 rpm for 0.5 h. For 4-NP hydrogenation, reac-
tant of 5 g was dissolved in anhydrous ethanol of 100 ml and then
catalyst of 3 g was added into the solution; the reaction was con-
◦
reduction and desorption (H -TPR/TPD), X-ray photoelectron spec-
ducted at 70, 90, and 110 C respectively and 1.0 MPa with a stirring
2
troscopy (XPS), and in situ X-ray diffraction (in situ XRD). The
specific surface area and pore volume of catalysts were measured
by N2 adsorption at 77 K on a Micrometrics ASAP 2020 instru-
speed of 500 rpm for 5 h. In the cases of IPN, NB and BD hydro-
genation, the recovered catalysts were washed with anhydrous
ethanol and reused multiple times under the same reaction condi-
tions. The reaction products were analyzed off-line by Agilent-6820
GC with FID and HP-1 capillary column (0.32 mm × 30 m) using
a temperature-programmed method: the oven temperature was
◦
ment. Prior to N2 sorption, the samples were treated at 350 C
under vacuum for 8 h. The specific surface areas were estimated
using the BET method, while the pore volumes were calculated by
the t-plot method. XRD measurement was conducted on a Philips
X’Pert diffractometer with Cu-Ka radiation (operated at 40 kV and
◦
◦
◦
◦
increased from 20 to 290 C (10 C/min), 80–220 C (10 C/min),
◦
◦
◦
◦
100–260 C (20 C/min), and 130–280 C (15 C/min) for analysis of
◦
◦
4
0 mA) in the 2ꢀ = 10 –80 range. Several samples were also mea-
m-XDA, aniline, BDO, and 4-aminophenol, respectively.
sured under H atmosphere in the range of room temperature (RT)
2
◦
to 500 C. The SEM images of samples were taken on a Hitachi S-
3
3
3
. Results and discussion
3
400N SEM instrument. H -TPR was performed with the sample
2
kept in a H2 (5%)–N2 (balanced) flow (40 ml/min) and heated at
a rate of 10 C/min from RT to 550 C. H -TPD was conducted by
using a thermo-conductive detector (TCD). Catalyst of 0.1 g was first
reduced in a H2 (5%)–N2 (balanced) flow (40 ml/min) at 500 C for
.1. Characterization
◦
◦
2
.1.1. XRD
◦
The structure and property of alumina support was verified
2
h. Then the sample was cooled to RT. After that, H2 adsorption
before the loading of active component(s). The XRD results of alu-
mina calcined at different temperatures are shown in Fig. 1. The
sample calcined at 400 C essentially retains the structure of pseu-
doboehmite. Dehydration of pseudoboehmite occurs at 450 C, and
the structure of ␥-Al O3 starts to form at 500 C. At 500 C, the
was conducted at RT for 1 h. Finally, H -TPD was carried out in an
Ar flow (30 ml/min) with the sample being heated to 500 C at a rate
of 10 C/min. XPS measurement was performed on a VG ESCALAB
MK II instrument using Mg K␣ radiation. The binding energy (BE)
was calibrated against the C1s signal (284.6 eV) of contaminant car-
bon. Elemental surface composition was estimated on the basis of
peak areas which were normalized using the Wagner factors.
2
◦
◦
◦
◦
◦
◦
2
␥
-Al O structure is rather amorphous. With further rise of calcina-
2 3
tion temperature, the ␥-Al O peak intensity increases (especially
2
3
◦
◦
those in the range of 2ꢀ = 10 –20 ), suggesting an increase in sam-
ple crystallinity as well as a change in sample morphology. In order
to remain sufficiently high in surface area while keeping the ␥-
Al O structure, thermo-treatment of pseudoboehmite was kept at
2.3. Catalyst evaluation
2
3
◦
550 C. In situ XRD investigation on the as-obtained Al O sample
2 3
The performance evaluation of catalysts for different reac-
tions was conducted in a continuously stirred tank reactor (CSTR).
(not shown) suggests that the support is stable under H2 atmo-
sphere in the 30–500 C range.
◦
For IPN hydrogenation, the following conditions were generally
◦
adopted: 70 C, 6.0 MPa H , 450 rpm, 1.5 h. A mixture of toluene and
Fig.
cursors
2
shows the XRD patterns of catalyst pre-
20%Ni(–O)/Al O , 20%Ni–3%Cu(–O)/Al O ,
2
methanol (Vtoluene/Vmethanol = 4:1) was used as the IPN solvent. The
IPN concentration in solution was 5–12.5 wt%. For NB hydrogena-
tion, NB of 10 ml was dissolved in anhydrous ethanol of 100 ml and
then catalyst of 3 g was added into the solution; the reaction was
2
3
2
3
20%Ni–3%Cu–5%Co(–O)/Al O , and 20%Ni–5%Cu–5%Co(–O)/Al O
2
3
2
3
(before H reduction) as well as that of ␥-A1 O . It was pointed out
2
2
3
by Jiang et al. that loading of active component(s) might induce
◦
operated at 100 C and 1.0 MPa with a stirring speed of 500 rpm for
structural rearrangement of ␥-A1 O , resulting in a change of
2
3
2
.5 h. For BD hydrogenation, BD of 5 g was dissolved in anhydrous
support morphology [14]. The characteristic diffraction peaks of
NiO phase can be readily identified (2ꢀ = 43 , 63 ). The intensity
of NiO peaks declines with Cu and/or Co introduction, suggesting
◦
◦
ethanol of 100 ml and then catalyst of 3 g was added into the solu-
tion; the reaction was performed at 40 or 60 C and 1.7 MPa with a
◦

7
0
T. Shi et al. / Applied Catalysis A: General 425–426 (2012) 68–73
Δ(200)
•
γ
-Al O3
Δ
NiO
∇
(220)
2
Δ
∇
CoO
∇
•
(440)
∇
•
(400)
4
3
∇
•
•
•
•
•
2
1
•
•
•
(
311)•
•(
222)
Al O
2
3
1
0
20
30
40
50
60
70
80
2θ
(°)
Fig. 2. XRD patterns of support and catalyst precursors before H2-reduction: (1)
0%Ni(–O)/Al2O3, (2) 20%Ni–3%Cu(–O)/Al2O3, (3) 20%Ni–3%Cu–5%Co(–O)/Al2O3,
2
and (4) 20%Ni–5%Cu–5%Co(–O)/Al2O3.
a possible enhanced dispersion of NiO component caused by the
addition of Cu and/or Co [15–17]. Different from the Cu component,
Co species can be identified over the Co-containing samples, and
this could be due to the fact that the dispersion of Co species is
relatively low.
3.1.2. N2 sorption
The surface area decreases with increasing NiO loading, but only
declines insignificantly with the addition of Cu and/or Co (Table 1).
Based on the data shown in Table 1, one can deduce that NiO is
largely deposited in the pores of Al O3 since notable decrease in
2
pore volume and diameter as well as in specific surface area can
be observed, while the introduction of Cu and/or Co components
has little effect on Al O3 texture (little change in pore volume and
diameter), though the supported CoOx entities are relatively less
dispersed than the CuO species according to the XRD results (Fig. 2).
2
Fig. 3. SEM images of H2-reduced catalysts: (a) 20%Ni/Al2O3, (b)
0%Ni–3%Cu–5%Co/Al2O3.
2
the latter are notably smaller than those in the former, confirming
that the Cu–Co components promote the dispersion of metallic Ni.
3
.1.3. SEM
Based
on
the
SEM
images
20%Ni–3%Cu(–O)/Al O ,
of
catalyst
precur-
and
sors
2
20%Ni(–O)/Al O ,
0%Ni–3%Cu–5%Co(–O)/Al O3 (not shown), one can deduce
2
3
2
3
3.1.4. XPS
2
The Ni2p binding energy (BE) and surface composition of rep-
that the supported components are highly dispersed. After Cu
introduction, there is improvement in NiO dispersion but little
change in morphology of NiO particles. In the case of CoOx addi-
tion, however, beside enhancement of NiO dispersion, there is
change in morphology. To enhance NiO dispersion through the
introduction of Cu and Co components is attractive because with
higher dispersion of active metals, more active sites are available.
Under the conditions for hydrogenation reaction, the metal oxide
species could be transformed to metallic entities. To study the
resentative catalyst precursors were analyzed by means of XPS.
Due to technical limitation, we were unable to perform in situ H2
2+
reduction of samples. With Cu presence, the Ni2p3/2 BE of Ni
decreases from 857.1 to 856.4 eV (Table S1). The BE of Ni2+ fur-
ther decreases to 856.2 eV in the co-presence of Cu and Co. The
results suggest that the addition of Cu and Co into Ni-based catalysts
results in a net transfer of electrons towards Ni and a decrease in
Ni2p3/2 BE. The surface Ni concentration for the single-, binary-, and
ternary-component catalyst is 9.6, 14.6, and 17.1 at%, respectively
dispersion of metallic species, the reduced 20%Ni(–O)/Al O3 and
2
tigation (Fig. 3). One can observe that the supported particles in
2
(Table S1), indicating that the dispersion of Ni species is enhanced
0%Ni–3%Cu–5%Co(–O)/Al O catalysts were subject to SEM inves-
2 3
due to the addition of Cu and/or Co. This is in good agreement with
the SEM observations. Doping Cu and Co to Ni may generate a syner-
gistic interaction amidst the Ni, Cu, and Co components, resulting in
enhanced electron density of Ni as well as improved Ni dispersion.
The enhanced Ni exposure should favor the adsorption and acti-
vation of reactants, including hydrogen, consequently promoting
hydrogenation reaction.
Table 1
Specific surface area, pore volume, and pore diameter of support and supported
catalysts.
Sample
Surface area
Pore volume
(cm
Pore size
(nm)
2
−1
3
−1
g )
(
m
g
)
3
.1.5. H -TPR
2
2
2
0%Ni(–O)/Al2O3
0%Ni–3%Cu(–O)/Al2O3
0%Ni–3%Cu–5%Co(–O)/Al2O3
138.2
133.8
133.4
190.8
0.27
0.26
0.26
0.40
5.2
5.3
5.1
6.0
2
As shown in Fig. 4, the TPR profiles of Al O , 20%Ni(–O)/Al O ,
0%Ni–3%Cu(–O)/Al O , and 20%Ni–3%Cu–5%Co(–O)/Al O indi-
2
3
2
3
2
2 3 2 3
Al2O3 support
cate that with the addition of Cu and/or Co, there is a change

T. Shi et al. / Applied Catalysis A: General 425–426 (2012) 68–73
71
♦
a
Oct. Ni2+
◊
NiO, ♦ Ni
♦
Tetra. Ni2+
o
5
50 C
o
5
00 C
o
4
4
50 C
o
00 C
d
o
3
50 C
o
3
2
00 C
o
50 C
c
o
2
00 C
b
o
1
50 C
a
◊
◊
RT
^
2
00
400
Temp. (
600
800
°C)
2
0
30
40
50
60
70
80
2θ/°
Fig. 4. H2-TPR profiles of (a) Al2O3, (b) 20%Ni(–O)/Al2O3, (c) 20%Ni–3%Cu(–O)/Al2O3,
and (d) 20%Ni–3%Cu–5%Co(–O)/Al2O3.
♦
•
b
∇
•
♦
NiO,
Ni
∇ CoO,
in NiO reduction profile. As suggested by Ranade, there are
tetrahedrally and octahedrally coordinated Ni2 in supported
Ni catalysts [18], the former are less reducible than the latter
and with reduction peak positioned at higher temperature. For
+
♦
RT
o
5
00 C
2
0%Ni–3%Cu(–O)/Al O , Cu–Ni interaction enhances NiO disper-
2 3
o
4
50 C
sion and reduction. Note that Co addition not only significantly
reduces the reduction temperature of NiO but also increases the
reducibility of NiO. In other words, NiO reduction is notably
enhanced by Cu and Co presence. Over 20%Ni(–O)/Al O , the
o
4
00 C
o
3
50 C
2
3
o
high-temperature reduction peak (due to the presence of tetra-
300 C
hedrally coordinated Ni2 ) dominates, and this peak diminishes
+
o
250 C
o
over 20%Ni–3%Cu(–O)/Al O3 and becomes even smaller over
200 C
2
o
2
0%Ni–3%Cu–5%Co(–O)/Al O . The TPR results suggest that due
150 C
2
3
o
to the interaction among the Ni–Co–Cu components, the amount
•
30 C
of tetrahedrally coordinated Ni2+ over Al O3 is reduced; in other
^
2
words, the amount of NiAl O4 species is decreased while that of
2
2
0
30
40
50
60
70
80
NiO increased [19,20], which enhanced the reducibility of the sup-
ported NiO specimen and hence the exposure of metallic Ni in the
ternary-metal catalyst.
2θ/°
Fig. 6. In situ XRD patterns of catalyst precursors (a) 20%Ni(–O)/Al2O3 and (b)
0%Ni–3%Cu–5%Co(–O)/Al2O3 at elevated temperatures under H2 atmosphere.
2
3
.1.6. H -TPD
2
H -TPD
was
carried
out
over
20%Ni(–O)/Al O ,
2 3
2
to compare the exposure of metal on surfaces [21]. The introduc-
tion of Cu brings about a shift of desorption peak toward higher
temperature (i.e. stronger M–H interaction) and an increase in
peak area (i.e. larger metal exposure). Addition of both Cu and Co
causes a notable shift of desorption peak to lower temperature
2
0%Ni–3%Cu(–O)/Al O , and 20%Ni–3%Cu–5%Co(–O)/Al O (Fig. 5)
2
3
2
3
b
(i.e. weaker M–H interaction) and a further increase in peak area
even higher metal exposure). The observations suggest that a
(
higher and more uniform dispersion of surface Ni is achieved in
the Ni–Cu–Co/Al O3 catalyst [22]. The data are also consistent
2
with the results of SEM, XPS, and H -TPR. There is greater amount
2
a
c
of active H species exists on the ternary Ni–Cu–Co/Al O3 catalyst.
2
Comparing the H -TPR and H -TPD profiles, one can see that the
2
2
H2 desorption temperature is generally lower than the H2 reduc-
tion temperature, which excludes the possibility of certain oxide
species being reduced by the desorbed H2 [23], and ascertains that
desorption peaks are solely originated from adsorbed H species on
metal surfaces.
3
.1.7. In situ XRD
In situ XRD technique was applied to study the structural
100
200
300
400
500
C)
600
700
800
Temp. (
°
variation of Ni(–O)/Al O and Ni–Cu–Co(–O)/Al O under H atmo-
2
3
2
3
2
sphere at elevated temperatures (Fig. 6). From Fig. 6a, it is observed
that the single NiO phase starts to diminish due to reduction at
Fig. 5. H2-TPD profiles of (a) 20%Ni(–O)/Al2O3, (b) 20%Ni–3%Cu(–O)/Al2O3, and (c)
0%Ni–3%Cu–5%Co(–O)/Al2O3.
2

7
2
T. Shi et al. / Applied Catalysis A: General 425–426 (2012) 68–73
Table 2
a 120
Optimization of Ni loading for IPN hydrogenation.
IPN conversion (%)
Effect of Ni loading (%)
100
m-XDA selectivity (%)
Supported component: x%Ni
Ni loading (%)
m-XDA yield (%)
10
75.4
15
80.1
20
85.0
25
78.8
8
6
4
2
0
0
0
0
0
Supported components: x%Ni–3%Cu–5%Co
Ni loading (%)
m-XDA yield (%)
10
85.2
15
89.1
20
95.0
25
86.8
◦
Reaction conditions: 70 C, 6.0 MPa, stirring speed of 450 rpm, and IPN concentration
of 5.0%.
Single amine selectivity (%)
◦
◦
3
50 C and vanishes completely at 500 C. With addition of Cu and
Co, the NiO phase starts to diminish at a notably lower temperature
◦ ◦
300 C) and vanishes completely at 450 C. Thus the results of in
(
7
8
9
10
11
12
13
situ XRD study clearly demonstrate the promotion effect of Cu and
Co on NiO reduction, implying a direct interaction among the Ni, Cu,
and Co constitutes. From Fig. 6b, it is observed that the supported
CoOx is more easily reduced than NiO and its reduction almost com-
IPN Concentration (%)
120
b
◦
pletes at 250 C. The easy reduction of CoOx may be due to strong
IPN conversion (%)
interaction between NiO and CoOx entities, and such interaction
enforces the promoting effects of Co on Ni. Therefore, the in situ
XRD observations clearly demonstrate the reduction behavior of
NiO in the presence of Cu and Co components, and that interaction
among the three metals is favorable.
100
m-XDA selectivity (%)
8
6
4
0
0
0
3.2. Catalytic performance
Prior to performance evaluation, the catalyst was activated in
◦
a H2 flow (150 ml/min) at 400 C for 5 h. The effect of calcination
20
0
◦
Single amine selectivity (%)
temperatures (450, 500, 550, and 600 C) of the support precursor
pseudoboehmite) on catalyst activity was first investigated,
and the optimal temperature for support thermo-treatment
was verified to be 550 C. The effect of Ni loading was studied
(
◦
4
5
6
7
8
9
10
11
12
13
in the 10–25 wt% range (Table 2) and the optimal value was
found to be 20 wt%. The catalyst constitution was systematically
IPN Concentration (%)
tuned: Ni/Al O ; Ni–Cu/Al O ; Ni–Co/Al O ; and Ni–Cu–Co/Al O .
2
3
2
3
2
3
2
3
Fig. 8. Dependence of IPN conversion and m-XDA selectivity on initial IPN concen-
The reaction performances are depicted in Supplementary Fig.
S1, Figs. 7 and 8 to make clearer comparisons among single-
metal (Ni/Al O ), binary-metal (Ni–Co/Al O and Ni–Cu/Al O ),
tration over (a) 20%Ni–5%Cu–3%Co/Al O , and (b) 20%Ni–3%Cu–5%Co/Al O .
2
3
2
3
2
3
2
3
2
3
and ternary-metal catalysts (Ni–Co–Cu/␥-Al O ). It is clearly
2
3
observed in Fig. S1, Supplementary information that introduction
of the secondary metal component (Cu or Co) to Ni notably
enhanced m-XDA selectivity. The amount of 3–4%Cu and 4–5%Co
showed better promotion effect on Ni. As revealed in Fig. 7,
the performances of the ternary-metal catalysts with Cu and
Co amount separately varied show the following sequence:
1
1
10
00
IPN conversion
2
0%Ni–3%Cu–5%Co/Al O > 20%Ni–5%Cu–3%Co/Al O > 20%Ni–3%
2 3 2 3
Cu–3%Co/Al O > 20%Ni–5%Cu–5%Co/Al O . The results indicate
2
3
2
3
9
8
7
6
5
0
0
0
0
0
m-XDA selectivity
that the activity of a ternary-metal catalyst is rather sensitive to
the variation in both Cu and Co amount, and this further suggest
an intensive and direct interaction among the Ni–Co–Cu ternary
metals. The Ni–Co–Cu catalyst comprising relatively low Cu (3%)
and high Co (5%) content outperformed (IPN conversion of 100%
and m-XDA selectivity of 98.2%) its counterparts in the catalytic
hydrogenation of IPN to m-XDA.
Fig. 8 shows the catalytic activities of 20%Ni–5%Cu–3%Co/Al O ,
2
3
and 20%Ni–3%Cu–5%Co/Al O3 of different Cu and Co contents.
2
Over the former, 100% IPN conversion can be maintained in IPN
concentration range of 7.5–12.5%, and 95–97% m-XDA selectiv-
ity can be obtained in IPN concentration of 7.5–11 wt%. Note
that there is significant increase of by-products (single- as well
as poly-amine) when the IPN concentration is higher than 11%,
resulting in decline of m-XDA selectivity to around 80%. Over
1
2
3
4
Catalyst
Fig. 7. IPN conversion and m-XDA selectivity over the Ni–Co–Cu/Al2O3 catalysts
with different Co and Cu contents. Catalyst 1: 20%Ni–3%Cu–5%Co/Al2O3; Cata-
lyst 2: 20%Ni–3%Cu–3%Co/Al2O3; Catalyst 3: 20%Ni–5%Cu–3%Co/Al2O3; Catalyst
◦
20%Ni–3%Cu–5%Co/Al O3 (less Cu but more Co), IPN conversion of
4
: 20%Ni–5%Cu–5%Co/Al2O3. Reaction conditions: 70 C, 6.0 MPa, stirring speed of
2
4
50 rpm, and IPN concentration of 7.5%.
100% can also be maintained at IPN concentration of 5.0–12.5%,

T. Shi et al. / Applied Catalysis A: General 425–426 (2012) 68–73
73
Table 3
In the cases of IPN, NB, and BD hydrogenation, the recovered
catalysts were reused up to 3 times under the same conditions.
There is no substantial decline in activity (overall ≤4%) observed
over the repeatedly used catalyst.
Reaction performances (in %) for hydrogenation of NB, BDO, and 4-NP in CSTR.
Catalyst
20%Ni/Al2O3
20%Ni–3%Cu/Al2O3
20%Ni–3%Cu–5%Co/Al2O3
NB hydrogenationa
Conv. (%)
Sel. (%)
100
95.8
100
93.8
100
99.7
4. Summary
BD hydrogenationb
Conv. (%)
Sel. (%)
100
98.1
100
98.9
100/100#
99.1/99.8
In the present study, ternary Ni–Cu–Co catalysts supported on
#
pseudoboehmite-derived Al O3 were developed for the hydro-
2
4
-NP hydrogenationc
genation of organic substrates containing CN, NO , and
2
C C
ꢀ
ꢀ
Conv. (%)
Sel. (%)
/
/
/
/
100 /100 /100*
groups. Factors such as thermo-treatment of pseudoboehmite, Ni
loading, content of Cu–Co components, and operating parameters
are found to have an effect on reaction performance, and have been
optimized through proper tuning. The reaction data proved that the
ternary Ni–Cu–Co catalysts are efficient and versatile for the target
reactions. Complete conversion and high selectivity (95–99%) to
desired product can be accomplished for IPN, NB, and BD hydro-
genation under mild conditions. The characterization results of
ꢀ
ꢀ
84 /79 /<27*
a
◦
The reaction was operated at 100 C and 1.0 MPa with a stirring speed of 500 rpm
for 2.5 h.
b
#
◦
The reaction was performed at 40 ( ) or 60 C and 1.7 MPa with a stirring speed
of 500 rpm for 0.5 h.
c
ꢀ
ꢀ
◦
The reaction was conducted at 70 ( ), 90 ( ), and 110 C (*) respectively and
.0 MPa with a stirring speed of 500 rpm for 5 h.
1
SEM, H -TPR/TPD, XPS, and in situ XRD revealed that with the
2
and m-XDA selectivity is ca. 98% within the 5.0–7.5% IPN con-
centration range. With increase in IPN concentration from 7.5 to
2+
introduction of Cu and Co, Ni reduction and electron density of
Ni as well as the dispersion of metallic Ni can be enhanced. As a
result, there is Ni–Cu–Co synergism that accounts for the superior
performance of the catalyst.
1
2.5%, m-XDA selectivity decreases continuously to ca. 80%. It is
apparent that despite catalyst constitutional elements and Ni load-
ing are the same, a small variation in the amount of Cu and Co
can notably affect product distribution: m-XDA selectivity (98.2%)
Acknowledgement
is the highest over 20%Ni–3%Cu–5%Co/Al O3 at IPN concentra-
2
tion of 7.5% while m-XDA selectivity remains high (95.5%) over
This work was supported by Jiangsu Science & Technology Sup-
porting Program (Industry Section) (BE2008006).
2
0%Ni–5%Cu–3%Co/Al O3 at IPN concentration of 11%. In other
2
words, if a concentrated IPN solution (11%) is used for reaction,
the latter can have a higher overall productivity (∼40% higher).
From Table 3, one can see that the Ni-based catalysts are also
active and selective for the hydrogenation of organic compounds
with different functional groups. For NB hydrogenation, the binary
Ni–Cu/Al O catalyst shows slightly lower selectivity (93.8%), prob-
Appendix A. Supplementary data
Supplementary data associated with this article can be found, in
the online version, at doi:10.1016/j.apcata.2012.03.003.
2
3
ably due to the strong M–H interaction as revealed in H -TPR study;
2
References
while the ternary Ni–Co–Cu/Al O3 catalyst shows an outstanding
2
selectivity (99.7%), thanks to the favorable Ni surface properties
[
1] P. Gallezot, in: I.T. Horvath (Editor-in-Chief), Encyclopedia of Catalysis, vol. 4,
Wiley-Interscience, New Jersey, 2003, p. 49.
(
Table S1) resulting from the synergistic interaction among the
ternary Ni, Co, and Cu constitutes.
Wang and Li [24] found that the amorphous Ni–B alloy sup-
ported on SiO2 was active and selective for NB hydrogenation
[2] J. Buttle, A. Walter, US Patent 4,186,146 (1980).
[
[
[
3] Z.H. Lu, Z.Q. Zhu, J. Eastern China Univ. Sci. Tech. 25 (1999) 257–260.
4] K.M. Kenichiro, JP 79 41,840 (1979).
5] R.L. Augustine, Heterogeneous Catalysis for the Synthetic Chemist, Marcel
Dekker, Inc., New York, 1996, pp. 495–496.
(
94.6% selectivity at 83% conversion). They made comparison over
Ni-based catalysts (including Ni–B/SiO2 and superfine Ni–B), and
interpreted performance difference in terms of electronic and geo-
metric effect. As for BD hydrogenation, similar results are obtained
over our Ni-based catalysts. Note that complete conversion of BD to
[6] P.R. Kurek, US Patent 4,482,741 (1984).
[
[
[
7] A. Howard, US Patent 4,247,478 (1981).
8] T. Makoto, EP 538,865 (1993).
9] S.Y. Li, X.Q. Zhou, Chin. Ind. Catal. 14 (2006) 7–10.
[10] Chem. Week 164 (16) (2002) 42.
[11] Chemical Market Reporter 269 (23) (2006) 17.
12] W. Richard, P. Spencer, US2950441 (1976).
13] V.H. Eugene, R.D. Waldo, GB2104794A (1982).
[14] X.Y. Jiang, R.X. Zhou, P. Pan, B. Zhu, X.X. Yuan, X.M. Zheng, Appl. Catal. A 150
1997) 131–141.
15] S. Stefan, S. Tapio, P.L. Lars, K. Outi, Appl. Catal. A 144 (1996) 177–194.
16] D.G. Mustard, C.H. Bartholomew, J. Catal. 67 (1981) 186–206.
[17] J.T. Richardson, T.S. Cale, J. Catal. 102 (1986) 419–432.
18] V.V. Ranade, Chem. Eng. Sci. 47 (1992) 1857–1869.
19] J.B. Peri, J. Phys. Chem. 69 (1965) 211–219.
20] J.L. Falconer, J.A. Schwarz, Catal. Rev. Sci. Eng. 25 (1983) 141–227.
[21] N.W. Hurst, S.J. Gentrey, A. Jones, B.D. McNicol, Catal. Rev. Sci. Eng. 24 (1982)
233–309.
22] C.W. Hu, Y. Chen, A.M. Tian, J. Mol. Catal. (China) 10 (1996) 345–350.
23] X.W. Xie, Y. Li, Z.Q. Liu, H. Masatake, Nature 458 (2009) 746–749.
[24] M.H. Wang, H.X. Li, Chin. J. Catal. 22 (2001) 287–290.
[25] G.S. Yu, Z.N. Wu, L.W. Xu, Petrochem. Tech. 18 (1989) 148–153.
26] M.M. Telkar, C.V. Rode, V.H. Rane, R. Jaganathan, R.V. Chaudhari, Appl. Catal. A
16 (2001) 13–22.
27] H.T. Wang, Y.H. Zhou, Z.H. Yang, J.L. Liu, Q.J. Sun, X.H. Lu, F. Xin, Chin. J. Catal.
30 (2009) 414–420.
◦
BDO can be accomplished at reaction temperature as low as 40 C.
[
[
Yu et al. [25] reported 95% selectivity of BDO over modified Renay
Ni in BD hydrogenation at high reaction temperature and moderate
pressure. Telkar et al. found that ca. 65% selectivity was obtained
over a Pd/C catalyst operated at 7 MPa pressure [26]. In the case
of 4-NP hydrogenation, although the ternary Ni–Cu–Co catalyst is
very active for the reaction (100% conversion), non-selective hydro-
(
[
[
[
[
[
◦
genation of benzene ring becomes significant especially at 110 C,
and this is likely to be due to the OH group associated with the
◦
benzene ring. Decrease in temperature form 110 to 70 C can con-
[
[
siderably improve the selectivity to desired product (∼85%). Wang
et al. [27] studied 4-NP hydrogenation over a Ni/mesoporous TiO2
catalyst and found that activity was the highest at 10% of Ni load-
ing. The catalyst is superior to Raney Ni and only deactivates slowly
after repeated use. Nevertheless, there is room for further improve-
ment of catalyst performance in 4-NP hydrogenation by tuning
parameters such as catalyst constitution and reaction medium.
[
2
[