molecules

Article
Supported Ni Catalyst for Liquid Phase
Hydrogenation of Adiponitrile to
6-Aminocapronitrile and Hexamethyenediamine
Chengqiang Wang 1 , Zekun Jia 1 , Bin Zhen 1,2 and Minghan Han 1, *
1 Department of Chemical Engineering, Tsinghua University, Beijing 100084, China;
thuwcq@163.com (C.W.); jiazekun1991@163.com (Z.J.); zhenbin@tjut.edu.cn (B.Z.)
2 College of Chemistry and Chemical Engineering, Tianjin University of Technology, Tianjin 300384, China
* Correspondence: hanmh@tsinghua.edu.cn; Tel.: +86-10-6278-1469

Received: 25 November 2017; Accepted: 30 December 2017; Published: 4 January 2018

Abstract: Supported Ni catalysts prepared under different conditions, for liquid phase hydrogenation of
adiponitrile (ADN) to 6-aminocapronitrile (ACN) and hexamethyenediamine (HMD), were investigated.
The highly reactive imine intermediate can form condensation byproducts with primary amine products
(ACN and HMD), which decreased the yield coefficient of primary amines. The catalysts support, condition
of catalyst preparation and dosage of additive were studied to improve the yield. A highly dispersed
Ni/SiO2 catalyst prepared by the direct reduction of Ni(NO3)2/SiO2 suppressed the condensation reactions
by promoting the hydrogenation of adsorbed imines, and it gave the improved hydrogenation activity of
0.63 mol·kgcat −1 ·min−1 and primary amine selectivity of 94% when NaOH was added into the reactor.

Keywords: supported Ni catalyst; adiponitrile hydrogenation; 6-aminocapronitrile; hexamethyenediamine

1. Introduction
The hydrogenation of ADN to ACN and HMD is an important industrial process [1–4], which is
mainly catalyzed by Raney Ni catalysts [5,6]. Although the skeleton structure of Raney Ni catalysts
favors a high catalytic activity, it causes low mechanical strength and bad regeneration capability.
In addition, a safety hazard also exists as the Raney Ni catalysts are flammable when exposed to air.
Due to these drawbacks, many researchers have been working on developing supported Ni catalysts
to replace Raney Ni. The use of a supported Ni catalyst can dramatically reduce the cost of the catalyst,
making the process easy to operate and improving the safety of the hydrogenation reaction. The main
challenge is that so far, the supported Ni catalysts are inferior in terms of catalytic activity and primary
amine selectivity [7,8].
The reaction scheme for the hydrogenation of ADN is given in Figure 1 [9–11]. ADN is hydrogenated
to ACN and then to HMD. During the sequential hydrogenation, a highly reactive imine intermediate
is formed, which has a strong tendency towards intermolecular condensation with the primary amine
products (ACN and HMD) and intramolecular cyclization to undesired condensation byproducts and
hexamethyleneimine (HMI). The formation of the primary amines has to compete with these side reactions.
From the chemistry, it can be seen that increasing the hydrogenation activity of the catalyst would be an
effective way to suppress the byproducts.
In this work, supported Ni catalysts were studied from several aspects, such as the type of
support, conditions of catalyst preparation and dosage of additive, in order to increase the yield of
primary amines. A highly dispersed Ni/SiO2 catalyst prepared by the direct reduction of a Ni(NO3 )2
precursor [12] was found able to promote the hydrogenation of the imine intermediate so that its
performance was close to that of Raney Ni.

Molecules 2018, 23, 92; doi:10.3390/molecules23010092 www.mdpi.com/journal/molecules

Molecules 2018, 23, 92 2 of 11
Molecules 2018, 23, 92 2 of 10

Figure 1.
Figure 1. Mechanism
Mechanism of
of ADN hydrogenation.
ADN hydrogenation.

In thisand
2. Results work, supported Ni catalysts were studied from several aspects, such as the type of
Discussion
support, conditions of catalyst preparation and dosage of additive, in order to increase the yield of
primary
2.1. amines.
Catalyst A highly dispersed Ni/SiO2 catalyst prepared by the direct reduction of a Ni(NO3)2
Support
precursor [12] was found able to promote the hydrogenation of the imine intermediate so that its
Since the catalyst support has an important influence on the catalytic performance, the effect of
performance was close to that of Raney Ni.
different supports was investigated. Many oxides can be used as the support, among which Al2 O3
and SiO2 are the most widely used. The characterization result of the specific surface area and pore
2. Results and Discussion
structure of the Al2 O3 , SiO2 and Ni/Al2 O3 , Ni/SiO2 used are shown in Table 1. The evaluation result
in
2.1.Table 2 showed
Catalyst Supportthe Ni/SiO2 was better than Ni/Al2 O3 in terms of the selectivity of primary amines
and condensation byproducts.
Since the catalyst support has an important influence on the catalytic performance, the effect of
different supports
Table was investigated.
1. Specific Many
surface area oxides
and pore can beofused
structure as the support,
the supports among which Al2O3
and the catalysts.
and SiO2 are the most widely used. The characterization result of the specific surface area and pore
Sample 2 g−1 3 ·g−1
structure of the Al2OSpecific
3, SiO2 Surface
and Ni/AlArea/m
2O3, Ni/SiO Pore Volume/cm
2 used are shown in Table 1.Average Pore Size/nm
The evaluation result in
Table 2Al showed
O
2 3 the Ni/SiO 2 was
242better than Ni/Al 2O 3 in terms
0.64 of the selectivity of primary
9.9 amines
Ni/Al O
and condensation
2 3 byproducts. 175 0.41 9.5
SiO2 245 0.94 17.8
Ni/SiO2 190 0.72 17.4
Table 1. Specific surface area and pore structure of the supports and the catalysts.

Sample Specific
Table 2. Evaluation
Surface Area/m reaction
2g−1 resultVolume/cm
Pore of Ni/SiO2 and Ni/Al
3·g−1
2 O3 .
Average Pore Size/nm
Al2O3 242 0.64 9.9
Selectivity/%
Catalysts 175 ·kgcat −1 ·min−1
Ni/Al2O3 Catalytic Activity/mol 0.41 9.5
ACN HMD HMI Condensation Byproducts
SiO2 245 0.94 17.8
Ni/SiO2 0.25 50 4 9 37
Ni/SiO
Ni/Al2 O3
2 1900.29 43 0.724 8 17.445

Table 2. Evaluation reaction result of Ni/SiO2 and Ni/Al2O3.

Using the bifunctional mechanism [13], it is believed that in a nitrile hydrogenation system,
the hydrogenation reactions are catalyzed by metal sites while condensation reactions are catalyzed
Selectivity/%
Catalysts
by acid Catalytic
sites. This wouldActivity/mol·kg cat−1·min−1
indicate that the acidic properties of the support are crucial to the catalytic
ACN HMD HMI Condensation Byproducts
performance. A stronger acidity favors the conversion of nitrile but decreases the selectivity to primary
Ni/SiO2 0.25 50 4 9 37
amines. The NH3 -TPD data for Al2 O3 and SiO2 and their corresponding catalysts are presented in
Ni/Al2O3 0.29 43 4 8 45
Figure 2a. Al2 O3 has a large NH3 desorption peak in the temperature range from 100 to 500 ◦ C while
SiO2 does not. After the introduction of the Ni, Ni/Al2 O3 still showed an obvious NH3 desorption
Using the bifunctional mechanism [13], it is believed that in a nitrile hydrogenation system, the
peak in contrast to Ni/SiO2 as shown in Figure 2b. The acid sites can absorb imine intermediate and
hydrogenation reactions are catalyzed by metal sites while condensation reactions are catalyzed by
primary amine and cause them to react with each other so that condensation products are produced,
acid sites. This would indicate that the acidic properties of the support are crucial to the catalytic
which lead to the decreased selectivity of the target products.
performance. A stronger acidity favors the conversion of nitrile but decreases the selectivity to
primary amines. The NH3-TPD data for Al2O3 and SiO2 and their corresponding catalysts are
presented in Figure 2a. Al2O3 has a large NH3 desorption peak in the temperature range from 100 to
500 °C while SiO2 does not. After the introduction of the Ni, Ni/Al2O3 still showed an obvious NH3
Molecules 2018, 23, 92 3 of 10
desorption peak in contrast to Ni/SiO2 as shown in Figure 2b. The acid sites can absorb imine
intermediate andinprimary
presented Figure 2a. amine
Al2O3 hasand cause
a large NH3them to react
desorption with
peak in each otherrange
the temperature so that
from condensation
100 to
Molecules 2018, 23, 92 3 of 11
products500
are°Cproduced, which
while SiO2 does not.lead
Aftertothe
the decreasedofselectivity
introduction of2Othe
the Ni, Ni/Al target
3 still products.
showed an obvious NH3
desorption peak in contrast to Ni/SiO2 as shown in Figure 2b. The acid sites can absorb imine
intermediate and primary amine and cause them to react with each other so that condensation
products are produced, which lead to the decreased selectivity of the target products.

(a) (b)
(a) (b)
Figure 2. NH
Figure
Figure 2.
3-TPD
NH
2. NH -TPD
33-TPDofof
of(a)
(a)Supports
(a)
and
Supports and
Supports and (b)(b)
(b)
Catalysts.
Catalysts.
Catalysts.

Therefore, SiO2 isSiO

Therefore, the ismore
22 is the suitable
the more
more support
suitable
suitable supportfor
support for thehydrogenation
for the
the hydrogenation
hydrogenation of ADN
of ADN
of ADN toand
to ACN
to ACN ACN
and and HMD,
HMD,
HMD,
and SiO was
2 was used
used as
as the
the support in the following
and SiO2 was used as the support in the following study.
2 study.

2.2. Direct
2.2. Direct Reduction
Reduction Method
Method
2.2. Direct Reduction Method
The supported
The supportedNi
Nicatalyst
catalyst was
was prepared
prepared by wetness
by the the wetness incipient
incipient impregnation
impregnation method,
method, which
Thewhich
supported
consists
consists of three Ni catalyst
ofsteps:
three steps: was prepared
impregnation,
impregnation, by the
and wetness
calcination
calcination incipient
and reduction.
reduction. AfterAfter impregnation
calcination
calcination method,
the which
and reduction,
and reduction,
the Ni(NO
consists Ni(NO
of three )
steps:
3)2 loaded
3 2 loaded on the
impregnation,
on supports
the supports became became
calcination metal particles
and reduction.
metal particles with catalytic
After
with catalytic activity.
calcination
activity. The calcination
and reduction,
The calcination and the
and reduction
reduction are are steps
key key steps
and and
have have important
important influence
influence on on catalytic
the the catalytic performance.
performance.
Ni(NO3)2 loaded on the supports became metal particles with catalytic activity. The calcination and
Table 33 shows
Table shows meanmean particle
particle sizes
sizes ofof Ni
Ni estimated
estimated byby the
the two
two independent
independent methods,
methods, TEMTEM and
and
reduction areanalysis
XRD key steps [14]. and
Figure have
3 important
showed the TEM influence
images of on the
two catalytic
catalysts. The performance.
particle size of nickel on the
XRD analysis [14]. Figure 3 showed the TEM images of two catalysts. The particle size of nickel on
Table
the3Ni/SiO
NiDR shows
DR/SiO2 is2 mean
around
is around particle
7 nm,
7 nm,whilesizes
while that ofonNi
that onthe estimated
theNiNi
CRCR /SiO22by
/SiO is the two
is much
much independent
larger.
larger. XRD analysis methods,
XRD analysis results
results of TEM and
of the
the
XRD analysis [14].
catalysts
catalysts show
show Figure
the 3 showed
the similar
similar the TEM
information,
information, 6.4
6.4 and
and images
17.7 nm,ofrespectively.
17.7 nm, two catalysts. The particle size of nickel on
respectively.
the NiDR/SiO2 is around 7 nm, while that on the NiCR/SiO2 is much larger. XRD analysis results of the
Table3.3.NiNiparticle
Table particlesize
sizeofofNiNi
CR /SiO
/SiO22 and
and Ni
NiDR /SiOfrom
DR/SiO2 2
fromTEM
TEMandandXRD
XRDanalysis.
analysis.
catalysts show the similar information, 6.4CRand 17.7 nm, respectively.
Sample
Sample Ni Particle Size(a)
Ni Particle Size(a) 1 /nm Ni Ni
1/nm Particle Size(b)
Particle 2 /nm
Size(b)2/nm
Table 3. Ni Ni
particle
NiCR
CR/SiO2size
/SiO 2
of NiCR/SiO
19 219and NiDR/SiO2 from17.7
TEM
17.7 and XRD analysis.
NiDR
Ni DR /SiO
/SiO2 2 7 7 6.46.4
1Sample
Estimated
1 Estimated by TEM Ni
by TEM Particle
analysis.
analysis. Size(a)
2 Calculated
2 Calculated
1/nm
by XRD
by Niaccording
XRD analysis
analysis Particle Size(b)
to Scherrer
according to
2/nm
equation.
Scherrer equation.
NiCR/SiO2 19 17.7
NiDR/SiO2 7 6.4
1 Estimated by TEM analysis. 2 Calculated by XRD analysis according to Scherrer equation.

(a) (b)
Figure3.3.TEM
TEMimages
imagesofof(a)(a)NiNiCR/SiO
/SiO2 and (b) NiDR/SiO2.
Figure CR 2 and (b) NiDR /SiO2 .

(a) (b)
Figure 3. TEM images of (a) NiCR/SiO2 and (b) NiDR/SiO2.
Molecules 2018, 23, 92 4 of 10

The H2 uptake was 76.2 µmol/g by NiDR/SiO2 while that by NiCR/SiO2 was 15.4 µmol/g,
Molecules 2018,
corresponding 23, 92
to the active nickel areas of 30.0 and 6.0 m2/g [15]. We can conclude that Ni 4 of
DR11/SiO2 had
a superior Ni dispersion of 4.50% while that of NiCR/SiO2 is 0.90%, which is in accordance with the
TEM and theThe XRD H2 analysis.
uptake was Ni(NO 3)2/SiO2by
76.2 µmol/g catalyst
NiDR /SiOprecursor would
2 while that by NibeCRtotally reduced
/SiO2 was at 700 °C from
15.4 µmol/g,
corresponding to the active nickel areas of 30.0 and 6.0 m 2 /g [15]. We can conclude that Ni
H2-TPR profiles. The H2-chemisorption data showed the H2 adsorption capacity of a catalyst DR /SiO2 partially

reducedhad a superior Ni dispersion of 4.50% while that of NiCR /SiO2 is 0.90%, which is in accordance with
at 450 °C. This means that H2-chemisorption data here cannot be used to calculate the mean
the TEM and the XRD analysis. Ni(NO3 )2 /SiO2 catalyst precursor would be totally reduced at 700 ◦ C
particle size
from accurately.
H2 -TPR profiles. The H2 -chemisorption data showed the H2 adsorption capacity of a catalyst
Figure 4 showed
partially reducedthe XRD
at 450 ◦ C.patterns
This means ofthat
the H
catalysts. The characteristic diffraction peaks of NiO in
2 -chemisorption data here cannot be used to calculate
NiC/SiO2the
aremean
obvious,
particlewhich indicated that the nickel mainly existed in the form of nickel oxide. After
size accurately.
the reduction Figure
in H42,showed
NiCR/SiO the2XRD patterns
showed of thesharp
a very catalysts. The characteristic
Ni peak while NiDRdiffraction peaks of
/SiO2 showed NiONiO and
both
in NiC /SiO2 are obvious, which indicated that the nickel mainly existed in the form of nickel oxide.
Ni diffraction peaks. Louis et al. [16] have studied supported Ni catalyst and found existence of nickel
After the reduction in H2 , NiCR /SiO2 showed a very sharp Ni peak while NiDR /SiO2 showed both
phyllosilicates, which were located at the surface of catalyst precursor after nickel nitrate
NiO and Ni diffraction peaks. Louis et al. [16] have studied supported Ni catalyst and found existence
impregnated onto
of nickel silica. The nickel
phyllosilicates, phyllosilicates
which were located at theact as anchoring
surface of catalyst sites for metallic
precursor Ninitrate
after nickel particles and
suppress the formation of large Ni particles. However, calcination before reduction leads to the
impregnated onto silica. The nickel phyllosilicates act as anchoring sites for metallic Ni particles
formationandofsuppress
larger the formation
metal of large
particles andNitoparticles. However, calcination
their aggregation. This isbefore reduction
probably dueleads
to the to partial
the formation of larger metal particles and to their aggregation. This is probably due to the partial
decomposition of nickel phyllosilicates during calcination, the breaking of the anchoring sites for
decomposition of nickel phyllosilicates during calcination, the breaking of the anchoring sites for
impregnated nickel, which favors nickel migration. Direct reduction of Ni(NO3)2/SiO2 was a moderate
impregnated nickel, which favors nickel migration. Direct reduction of Ni(NO3 )2 /SiO2 was a moderate
reduction process,
reduction which
process, sacrificed
which thethereduction
sacrificed degreebutbut
reduction degree gave
gave a superior
a superior Ni dispersion.
Ni dispersion.

Figure
Figure XRDpattern
4. 4.XRD pattern of
ofSiO2 , 2Ni
SiO C /SiO
, Ni 2 , Ni
C/SiO Ni/SiO
2, CR 2 and
CR/SiO NiDR
2 and Ni/SiO 2.
DR/SiO 2.

As described
As described above,above,
the thetwotwocatalysts
catalysts prepared
prepared bybythethe
different methods
different had different
methods had different
microstructures, which probably caused their catalytic performance for hydrogenation of ADN to be
microstructures, which probably caused their catalytic performance for hydrogenation of ADN to be
different. Table 4 shows the evaluation result of the catalysts. NiDR /SiO2 exhibited a better hydrogenation
different.activity
Table 4 shows the evaluation result of the catalysts. NiDR/SiO2 exhibited a better
of 0.56 mol·kgcat −1 ·min−1 compared to NiCR /SiO2 with 0.25 mol·kgcat −1 ·min−1 . The selectivity
hydrogenation activity
to the primary ofproducts
amine 0.56 mol·kg cat−1·min
over Ni
−1 compared to NiCR/SiO2 with 0.25 mol·kgcat−1·min−1. The
DR /SiO2 was higher than that over NiCR /SiO2 , which, respectively,
selectivity
wereto79%
theand
primary amine products
54%. Meanwhile, over
the selectivity NiDR/SiO2 was
to condensation higherwere
byproducts than that over
inhibited NiDR
over Ni /SiO2 ,2, which,
CR/SiO
which decreased from 37% to 3%. In the sequential reaction, shown in
respectively, were 79% and 54%. Meanwhile, the selectivity to condensation byproducts wereFigure 1, the highly reactive
inhibitedimine
overintermediate
NiDR/SiO2can further react along three pathways: further hydrogenation, intramolecular
, which decreased from 37% to 3%. In the sequential reaction, shown in
condensation to HMI, and intermolecular to condensation byproducts. The three pathways are parallel,
Figure 1,that
theis, highly reactive imine intermediate can further react along three pathways: further
the formation of primary amine competes with the side reactions. NiDR /SiO2 exhibited a
hydrogenation, intramolecular
superior hydrogenation condensation
activity, which promoted thetohydrogenation
HMI, and intermolecular
and improved to tocondensation
the selectivity ACN
byproducts. The three pathways are parallel, that is, the formation of primary amine competes with
and HMD.
the side reactions. NiDR/SiO2 exhibited a superior hydrogenation activity, which promoted the
hydrogenation and improved the selectivity to ACN and HMD.

Table 4. Evaluation reaction result of NiCR/SiO2 and NiDR/SiO2.

Selectivity/%
Catalysts Catalytic Activity/Mol·kgcat−1·min−1
 Molecules 2018, 23, 92 5 of 11

Table 4. Evaluation reaction result of NiCR /SiO2 and NiDR /SiO2 .

Selectivity/%
Catalysts Catalytic Activity/Mol·kgcat −1 ·min−1
ACN HMD HMI Condensation Byproducts
MoleculesNi
2018, 23, 92
/SiO 0.25 50 4 9 37 5 of 10
CR 2
NiDR /SiO2 0.50 66 13 18 3
The condensation byproducts were secondary and tertiary amines with a high molecular weight
which wereThe not eluted and
condensation cannot bewere
byproducts detected in the
secondary GC
and analysis.
tertiary Thewith
amines product
a highmixture
molecular was analyzed
weight
by awhich
masswere
spectrometer to confirm the existence of condensation byproducts. The result
not eluted and cannot be detected in the GC analysis. The product mixture was analyzed is shown in
Figure
by a5.mass
The spectrometer
mass to charge ratio of
to confirm theprotonated
existence ofHMI, ACN and
condensation HMD are,
byproducts. respectively,
The result is shownm/zin= 100,
113, 117. The peaks of m/z = 200~300 indicated the formation of C12 and C18 oligomers. These100,
Figure 5. The mass to charge ratio of protonated HMI, ACN and HMD are, respectively, m/z = showed
113, that
clearly 117. The
the peaks of m/z = 200~300
condensation indicated
byproducts the Ni
from formation
DR/SiO2ofwas
C12 and C18 oligomers.
decreased, whichThese showed
verified the GC
clearly that
evaluation result.the condensation byproducts from Ni DR /SiO 2 was decreased, which verified the GC
evaluation result.

(a) (b)
Figure 5. Mass
Figure 5. Massspectra
spectraofof
(a)(a)
NiNi CR/SiO2 and (b) NiDR/SiO2.
CR /SiO2 and (b) NiDR /SiO2 .

NiDRNi
/SiO/SiO
2 had a higher H2 adsorption capacity from H2 uptake result, which would indicate that
DR 2 had a higher H2 adsorption capacity from H2 uptake result, which would indicate
amines
that amines andintermediates
and imine are more
imine intermediates abundantly
are more abundantlyadsorbed on on
adsorbed NiDR
Ni/SiO 2. The C atoms of imines
DR /SiO2 . The C atoms of
wereimines
subjected to nucleophilic
were subjected attack by
to nucleophilic N atoms
attack of the
by N atoms of amines [17],
the amines leading
[17], leadingtotothethe formation
formation of a
secondary imine which
of a secondary can be
imine which canfurther hydrogenated
be further hydrogenated totoHMI.
HMI. NiNiDR
DR/SiO
/SiO2 2 was
was more
more selective for
selective
for intramolecular
intramolecular side reaction.
side reaction. As seen
As seen in Table
in Table 4, the
4, the selectivity
selectivity to to
HMIHMI overNi
over Ni CR
CR /SiO
/SiO 2 was
2 was 9%9%while
while that over
that over NiDR/SiO2 DRNi /SiO
was 18%.
2 was 18%. From this point, further developing of the supported
From this point, further developing of the supported Ni catalyst Ni catalyst
prepared by the direct reduction
prepared by the direct reduction method method isis still
still necessary.
necessary.The Theconditions
conditionsof catalyst preparation
of catalyst and and
preparation
reaction conditions have to be optimized to inhibit the formation of HMI and improve the selectivity
reaction conditions have to be optimized to inhibit the formation of HMI and improve the selectivity
to primary amines.
to primary amines.
2.3. Reduction Teamperature
2.3. Reduction Teamperature
The microstructure of supported Ni catalyst prepared by direct reduction method is affected by
The microstructure
reduction temperatureof supported The
significantly. Ni catalyst prepared
TEM images by direct
of catalysts reduction
reduced method
at different is affected by
temperature
are shown
reduction in Figure 6.
temperature The mean particle
significantly. sizesimages
The TEM of Ni became larger reduced
of catalysts with the increase of reduction
at different temperature
temperature obviously.
are shown in Figure 6. The mean particle sizes of Ni became larger with the increase of reduction
temperature obviously.
 to primary amines.

2.3. Reduction Teamperature

The microstructure of supported Ni catalyst prepared by direct reduction method is affected by
reduction temperature significantly. The TEM images of catalysts reduced at different temperature
Moleculesare
2018, 23, 92in Figure 6. The mean particle sizes of Ni became larger with the increase of reduction 6 of 11
shown
temperature obviously.

Molecules 2018, 23, 92 6 of 10

(a) (b)
Molecules 2018, 23, 92 6 of 10

(c) (d)
Figure 6. TEM images of the catalysts reduced at different temperatures: (a) 350 °C; (b) 400 °C; (c) 450
Figure 6. TEM images of the (c)catalysts reduced at different temperatures: (a)(d) 350 ◦ C; (b) 400 ◦ C; (c) 450 ◦ C;
°C; (d)
◦ C.
500 °C.
(d) 500Figure 6. TEM images of the catalysts reduced at different temperatures: (a) 350 °C; (b) 400 °C; (c) 450
°C;
The(d)H500 °C. profiles of Ni(NO3)2/SiO2 were given in Figure 7. The TPR profiles of the
2-TPR

impregnated
The catalysts,
H2 -TPR profiles ofperformed
Ni(NO3 )2without
/SiO2 were any previous
given incalcination,
Figure 7. The exhibit
TPR three peaksofofthe
profiles hydrogen
impregnated
The H2-TPR
consumption at 300,profiles
400, and ofabout
Ni(NO550 3)2/SiO2 were given in Figure 7. The TPR profiles of the
°C, attributed to the decomposition of nickel nitrate into NiO,
catalysts, performed without any previous calcination, exhibit three peaks of hydrogen consumption at
impregnated
the reductioncatalysts,
of NiO performed
and without
the reduction any previous
of nickel calcination,
phyllosilicates. Afterexhibit
TPR up three peaks
to 700 °C, of
thehydrogen
nickel is
300, 400, and about 550 ◦ C, attributed to the decomposition of nickel nitrate into NiO, the reduction of NiO
consumption
totally at 300,
reduced. 400,complete
Before and aboutreduction,
550 °C, attributed to the
the nickel decomposition
phyllosilicates areofprobably
nickel nitrate intoatNiO,
located the
and thethereduction
reduction of
of nickel
NiO andphyllosilicates.
the reduction ofAfter
nickelTPR up to 700 ◦After
phyllosilicates. C, theTPRnickel
up tois700
totally
°C, reduced.
the nickel is Before
interface between silica and the remaining nickel, for which they act as anchoring sites. During TPR,
completeafterreduction,
totally reduced.
nitrate the nickel
Before
decomposition, phyllosilicates
complete
the NiOreduction, arethe
particles probably located at theare
nickelonphyllosilicates
located interface
the phyllosilicates probably between
are reduced intosilica
located at
Ni0theatand the
interface
remaining nickel,between
for silica
which and
they the
act remaining
as anchoring nickel, for
sites. which
During they
TPR,act as anchoring
after nitrate
400 °C without any migration. Between 400 and 700 °C, the increase in metal particle size is due not sites. During
decomposition, TPR,the NiO
after
particles
only nitrate
located
to NiO decomposition,
onmigration
the phyllosilicatesthe NiO
induced byare particles
reduced
thermal located
into
effect butNi on
0
also the
atto400phyllosilicates
◦
the C withoutofany
reduction are reduced
themigration. into Ni 0 at
Between
phyllosilicates by 400
400
◦ C,°C without of any migration. Between 400 and 700 °C, the increase in metal particle size is due not
and 700the weakening
the increasethe inanchoring strength.
metal particle size is due not only to NiO migration induced by thermal effect
only Theto NiO migration
H2-TPR profilesinduced by thermal effect but also to the reduction of the phyllosilicates by
but alsothetoweakening
the reduction of the
the of
ofanchoring Ni(NO 3)2/SiO2 are in accordance with the TEM analysis of the four
phyllosilicates by the weakening of the anchoring strength.
catalysts reduced at 350, 400, 450strength.
and 500 °C, respectively.
The HThe 2 -TPR profiles
H2-TPR profilesof Ni(NO
of Ni(NO 3 )23)/SiO are in accordance with the TEM analysis of the four
2/SiO22 are in accordance with the TEM analysis of the four

catalysts reduced at 350, 400,400,

450450
and 500 ◦
catalysts reduced at 350, and 500 C,
°C,respectively.
respectively.

Figure 7. H2-TPR of catalyst precursors.

Figureresult
Figure 8 presents the evaluation
Figure 7. 7.HH2 -TPR
2-TPR offour
of the
of catalyst precursors.
catalysts.
catalyst The catalytic activity first decreased
precursors.
and then increased with the increase of reduction temperature, which can be explained by the
Figure
reduction 8 presents
degree and Nithedispersion.
evaluationWith
resulttheofincrease
the fourofcatalysts.
reductionThe catalytic activity
temperature, first decreased
the reduction degree
and
of the catalysts became higher while the Ni dispersion became lower because the by
then increased with the increase of reduction temperature, which can be explained the
nickel
reduction degree and Ni dispersion. With the increase of reduction temperature,
phyllosilicates were gradually decomposed. Ni/SiO2 reduced at 350 °C had a high hydrogenationthe reduction degree
of the because
activity catalystsofbecame
its high higher while When
Ni dispersion. the Nithedispersion became lower
reduction temperature wasbecause
increasedthe nickel
to 400 °C
phyllosilicates were gradually decomposed. Ni/SiO 2 reduced at 350 °C had a high hydrogenation
and 450 °C, the migration and aggregation of Ni particles led to a lower metallic Ni dispersion, which
 Molecules 2018, 23, 92 7 of 11

Figure 8 presents the evaluation result of the four catalysts. The catalytic activity first decreased
and then increased with the increase of reduction temperature, which can be explained by the reduction
degree and Ni dispersion. With the increase of reduction temperature, the reduction degree of the
catalysts became higher while the Ni dispersion became lower because the nickel phyllosilicates were
gradually decomposed. Ni/SiO2 reduced at 350 ◦ C had a high hydrogenation activity because of its
Molecules 2018,
high 23, 92
Ni dispersion. When the reduction temperature was increased to 400 ◦ C and 450 ◦ C, the migration 7 of 10
and aggregation of Ni particles led to a lower metallic Ni dispersion, which caused the catalytic activity
became even lower,
decrease. the catalytic
Although activity
the Ni dispersion of became
Ni/SiO2 reduced 500 ◦ C became
higheratbecause of its even
higher reduction
lower, degree. In
the catalytic
activity became higher because of its higher reduction degree. In conclusion, Ni
conclusion, Ni dispersion was the key factor to affect the catalytic activity for the reduction dispersion was the key
factor to affect the catalytic activity for the reduction temperature range from 350 ◦ C to 450 ◦ C while the
temperature range from 350 °C to 450 °C while the reduction degree was the key factor in the
reduction degree was the key factor in the temperature range from 450 ◦ C to 500 ◦ C, which explained
temperature range from
the variation 450 °C
of catalytic to 500 °C, which explained the variation of catalytic activity.
activity.

FigureFigure 8. Evaluation
8. Evaluation result
result ofofcatalysts
catalysts reduced
reduced at at
different temperatures.
different temperatures.

◦
450 considered
C was considered the optimum temperature for
450 °C was the optimum temperature forthe reduction
the when
reduction both catalytic
when activity activity
both catalytic
and selectivity to HMI were taken into account.
and selectivity to HMI were taken into account.
2.4. Additive
2.4. Additive
Most of the side reactions are deamination reactions as shown in Figure 1. Thermodynamically,
Mostadding
of theanside
alkaline substance
reactions areasdeamination
additive to the liquid reaction
reactions system
as shown inwould
Figure benefit the catalytic
1. Thermodynamically,
activity and selectivity to primary amines [5,18]. In ADN hydrogenation, H2 and nitrile adsorbed on
adding an alkaline substance as additive to the liquid reaction system would benefit the catalytic
metallic Ni first and then they reacted with each other to imine intermediates and amines. Imines and
activity and selectivity to primary amines [5,18]. In ADN hydrogenation, H2 and nitrile adsorbed on
amines desorbed from the catalyst with difficulty [19], which would not only suppress the following
metallic hydrogenation
Ni first and then theybut
of nitriles, reacted with each
also promote other to imine
the condensation intermediates
side reactions. and amines.
The addition Imines and
of an alkaline
amines desorbed from the
substance increases thecatalyst with difficulty
electron density of metallic[19],
Ni onwhich would
the catalyst not which
surface, only suppress
would benefitthethe
following
desorption of imine and amine [20].
hydrogenation of nitriles, but also promote the condensation side reactions. The addition of an
NaOH was chosen as an additive to improve the NiDR /SiO2 catalyst. The evaluation result
alkaline substance increases the electron density of metallic Ni on the catalyst surface, which would
is presented in Figure 9 and Table 5. The mass ratio NaOH/Ni was used as the independent
benefit the desorption of imine and amine [20].
variable. With NaOH/Ni increasing, the catalytic activity rose from 0.39 mol·kgcat −1 ·min−1 to
NaOH was·kg
0.63 mol chosen
−1
cat ·min
as−1an
andadditive to improve
the selectivity the NiDR
to condensation /SiO2 catalyst.
byproducts decreasedThe evaluation
from 12% to 1%result is
presented in Figure
while 9 and
that to HMI Table 5.
decreased The21%
from mass ratio
to 5%. WhenNaOH/Ni was used
the NaOH/Ni as the
was raised to independent
0.5, the primaryvariable.
With NaOH/Ni increasing, the catalytic activity rose from 0.39 mol·kgcat−1·min−1 to 0.63
mol·kgcat−1·min−1 and the selectivity to condensation byproducts decreased from 12% to 1% while that
to HMI decreased from 21% to 5%. When the NaOH/Ni was raised to 0.5, the primary selectivity was
as high as 94%. Clearly, the effect of NaOH on the catalytic activity and product selectivity was
 Molecules 2018, 23, 92 8 of 11

selectivity was as high as 94%. Clearly, the effect of NaOH on the catalytic activity and product
selectivity was positive.

Table 5. Evaluation reaction result of the catalysts with different NaOH/Ni mass ratio.

Selectivity/%
NaOH/Ni Catalytic Activity/mol·kgcat −1 ·min−1
ACN HMD HMI Condensation Byproducts
0 0.39 55 12 21 12
0.1 0.50 66 13 18 3
0.3 0.56 70 21 7 2
Molecules 2018, 23,
0.592 0.63 70 24 5 1 8 of 10

FigureFigure 9. Evaluation
9. Evaluation result
result of of catalysts with
catalysts with the
theuse
useof of
different NaOH/Ni
different ratios. ratios.
NaOH/Ni

The addition
The addition of NaOHof NaOH improved
improved thehydrogenation
the hydrogenation activity andand
activity improved the target
improved the products
target products
selectivity significantly. After the introduction of NaOH, the performance of NiDR /SiO2 , reduced at
selectivity significantly. After the introduction of NaOH, the performance of NiDR/SiO2, reduced at
450 ◦ C, is as excellent as Raney Ni, which is presented in Table 6.
450 °C, is as excellent as Raney Ni, which is presented in Table 6.
Table 6. Evaluation reaction result of the NiDR /SiO2 and Raney Ni.
Table 6. Evaluation reaction result of the NiDR/SiO2 and Raney Ni.
Selectivity/%
Catalysts Catalytic Activity/mol·kgcat −1 ·min−1
ACN HMD Selectivity/%
HMI Condensation Byproducts
Catalysts Catalytic Activity/mol·kgcat−1·min−1
Raney Ni 0.61 ACN
78 HMD
17 HMI
1 Condensation
4 Byproducts
NiDR /SiO2 0.63 70 24 5 1
Raney Ni 0.61 78 17 1 4
NiDR/SiO2 0.63 70 24 5 1
3. Materials and Methods

3. Materials and Methods

3.1. Catalyst Preparation
The supported Ni catalysts were prepared by incipient wetness impregnation, which included
3.1. Catalyst
threePreparation
steps: impregnation, calcination and reduction. Ni/SiO2 catalyst prepared by the direct reduction
(DR) of Ni(NO ) /SiO in H was denoted as Ni /SiO . For comparison, another catalyst prepared by
The supported 3Ni2 catalysts
2 2
were preparedDRby incipient
2
wetness impregnation, which included
calcination and reduction (CR) was denoted as NiCR /SiO2 . They were prepared as follows. The loading
three steps:
amountimpregnation,
of nickel was 20 calcination
wt.%. and reduction. Ni/SiO2 catalyst prepared by the direct
reduction (DR) of Ni(NO3)2/SiO2 in H2 was denoted as NiDR/SiO2. For comparison, another catalyst
prepared by calcination and reduction (CR) was denoted as NiCR/SiO2. They were prepared as
follows. The loading amount of nickel was 20 wt.%.
NiDR/SiO2: First, SiO2 (Alfa Aesar, Haverhill, MA, USA) was soaked in an aqueous solution of
Molecules 2018, 23, 92 9 of 11

NiDR /SiO2 : First, SiO2 (Alfa Aesar, Haverhill, MA, USA) was soaked in an aqueous solution of
Ni(NO3 )2 (98%, Alfa Aesar) at room temperature. Then, the mixture was treated with ultrasonic waves
for 2 h and dried at 80 ◦ C for 12 h. Finally, the catalyst precursor was reduced in H2 flow at 450 ◦ C for 8 h.
NiCR /SiO2 : The impregnated and dried Ni(NO3 )2 /SiO2 was calcined at 450 ◦ C for 4 h and
then reduced in H2 flow. The other steps were the same as those for NiDR /SiO2 . Before reduction,
the calcined (C) catalyst was denoted as NiC /SiO2 .

3.2. Catalyst Characterization

The specific surface area and pore structure of support and catalyst were determined by N2 adsorption
with a Quadrasorb SI instrument. The crystalline phase of the catalysts was characterized by a Bruker
Advance D8 X-ray diffractometer with a Cu Kα radiation source. The H2 uptake of the catalysts was
determined by H2 chemisorption on a Quantachrome ChemBET Pulsar TPR/TPD instrument. The active
Ni area was calculated assuming H/Nisurface = 1 and a surface area of 6.5 × 10−20 m2 per Ni atom.
A JEM-2010 transmission electron microscope was employed to examine the Ni particle size of catalysts.
Some product samples were analyzed by mass spectrometry (MS model: Q Exactive) to identify the relative
amount of condensation byproducts.

3.3. Catalytic Reaction

The hydrogenation of ADN was carried out in a stainless-steel autoclave equipped with a
temperature control system and magnetic stirrer. Typically, 5 g ADN (98%, Alfa Aesar, Haverhill, MA,
USA), 80 mL of methanol (>99.5%, Beijing Chemical Works, Beijing, China), 5 g pre-reduced catalyst
and 0.1 g NaOH (>96.0%, Beijing Chemical Works) were added into the reactor. The hydrogenation
conditions were: 80 ◦ C, 3 MPa, 500 rpm. The products were sampled online and analyzed by a gas
chromatograph (GC 7890F, Techcomp Instrument Company, Kwai Chung, Hong Kong) equipped
with a flame ionization detector and a KB-624 capillary column (30 m × 0.32 mm × 1.8 µm, Kromat).
The condensation byproducts were high molecular weight secondary and tertiary amines which were
not eluted in the GC. Phenylamine was used as the internal standard to determine the content of ADN,
ACN, HMD and HMI. The catalytic activity, the conversion of ADN and the selectivity to ACN and
HMD were calculated as:
moles of converted ADN
Catalytic Activity = (1)
Time × mCat

moles of converted ADN

ADN Conversion = × 100% (2)
moles of ADN feedstock
moles of ACN
ACN Selectivity = × 100% (3)
moles of converted ADN
moles of HMD
HMD Selectivity = × 100% (4)
moles of converted ADN

4. Conclusions
Supported Ni catalysts for the liquid phase hydrogenation of ADN were investigated. The following
conclusions are based on the results and discussion:

1. The support affects the catalytic performance of supported Ni catalysts by providing original
acid sites, which promote the condensation side reactions and decrease the selectivity to primary
amines. SiO2 is chosen as the support in this work providing less original acid sites.
2. The catalyst prepared by the direct reaction method, without a prior calcination step, gave smaller
Ni particles, this is, a higher Ni dispersion and H2 uptake capacity, which improved the catalytic
activity and inhibited the intermolecular condensation side reaction.
Molecules 2018, 23, 92 10 of 11

3. The reduction temperature had an appreciable effect on the catalytic performance of NiDR /SiO2 .
450 ◦ C was regarded as a preferable temperature to suppress the intramolecular condensation
reaction to HMI.
4. The addition of NaOH into the reaction system improved the hydrogenation activity and
significantly improved the target products selectivity. When the NaOH/Ni mass ratio was
increased to 0.5, the selectivity to primary amines attained 94%.

Author Contributions: M.H., Z.J. and C.W. conceived and designed the experiments; Z.J. and C.W. performed
the experiments; Z.J. and C.W. analyzed the data; B.Z. and C.W. wrote the paper.
Conflicts of Interest: The authors declare no conflict of interest.

References
1. Dahlhoff, G.; Niederer, J.P.M.; Hoelderich, W.F. Epsilon-Caprolactam: New by-product free synthesis routes.
Catal. Rev. Sci. Eng. 2001, 43, 381–441. [CrossRef]
2. Tolman, C.A.; Mckinney, R.J.; Seidel, W.C.; Druliner, J.D.; Stevens, W.R. Homogeneous Nickel-Catalyzed
Olefin Hydrocyanation. Adv. Catal. 1985, 33, 1–46. [CrossRef]
3. Brunelle, J.; Seigneurin, A.; Sever, L. Method for Cyclizing Hydrolysis of an Aminonitrile Compound into
Lactam. U.S. Patent 6479658B1, 12 November 2002.
4. Leconte, P. Process for Preparing Lactams. U.S. Patent 20120095212A1, 19 April 2012.
5. Ziemecki, S.B. Selective Low Pressure Hydrogenation of a Dinitrile to an Aminonitrile. U.S. Patent 5151543A,
29 September 1992.
6. Boschat, V.; Brunelle, J.P.; Darrier, B.; Chevaljer, B.; Bobet, J.L. Hemihydrogenation Method for Dinitriles.
U.S. Patent 6521779B1, 18 February 2003.
7. Ziemecki, S.B.; Ionkin, A.S. Aminonitrile Production. U.S. Patent 6080884A, 27 June 2000.
8. Liao, H.; Liu, S.; Hao, F.; Liu, P.; You, K.; Liu, D.; Luo, H. Liquid phase hydrogenation of adiponitrile to
6-aminocapronitrile and hexamethylenediamine over potassium doped Ni/alpha-Al2 O3 catalyst. React. Kinet.
Mech. Catal. 2013, 109, 475–488. [CrossRef]
9. Chen, B.; Dingerdissen, U.; Krauter, J.G.E.; Rotgerink, H.G.J.L.; Mobus, K.; Ostgard, D.J.; Panster, P.;
Riermeier, T.H.; Seebald, S.; Tacke, T. New developments in hydrogenation catalysis particularly in synthesis
of fine and intermediate chemicals. Appl. Catal. A Gen. 2005, 280, 17–46. [CrossRef]
10. Tichit, D.; Durand, R.; Rolland, A.; Coq, B.; Lopez, J.; Marion, P. Selective half-hydrogenation of adiponitrile
to aminocapronitrile on Ni-based catalysts elaborated from lamellar double hydroxide precursors. J. Catal.
2002, 211, 511–520. [CrossRef]
11. Hoffer, B.W.; Moulijn, J.A. Hydrogenation of dinitriles on Raney-type Ni catalysts: Kinetic and mechanistic
aspects. Appl. Catal. A Gen. 2009, 352, 193–201. [CrossRef]
12. Jia, Z.; Zhen, B.; Han, M.; Wang, C. Liquid phase hydrogenation of adiponitrile over directly reduced
Ni/SiO2 catalyst. Catal. Commun. 2016, 73, 80–83. [CrossRef]
13. Verhaak, M.J.F.M. The selective hydrogenation of acetonitrile on supported nickel-catalysts. Catal. Lett.
1994, 26, 37–53. [CrossRef]
14. Nakagawa, Y.; Nakazawa, H.; Watanabe, H.; Tomishige, K. Total Hydrogenation of Furfural over a Silica-
Supported Nickel Catalyst Prepared by the Reduction of a Nickel Nitrate Precursor. ChemCatChem 2012, 4,
1791–1797. [CrossRef]
15. Li, H.X.; Xu, Y.P.; Li, H.; Deng, J.F. Gas-phase hydrogenation of adiponitrile with high selectivity to primary
amine over supported Ni-B amorphous catalysts. Appl. Catal. A Gen. 2001, 216, 51–58. [CrossRef]
16. Louis, C.; Cheng, Z.; Che, M. Charaterization of Ni/SiO2 catalysts during impregnation and future thermal
activation treatment leading to metal particles. J. Phys. Chem. 1993, 97, 5703–5712. [CrossRef]
17. Braun, J.V.; Blessing, G.; Zobel, F. Katalytische hydrierungen unter druck bei gegenwart von nickelsalzen, VI:
Nitrile. Berichte der Deutschen Chemischen Gesellschaft (A B Ser.) 1923, 56, 1988–2001. [CrossRef]
18. Sanchez, K.M. Preparation of 6-Aminocapronitrile. U.S. Patent 5296628A, 22 March 1994.
Molecules 2018, 23, 92 11 of 11

19. Huang, Y.Y.; Adeeva, V.; Sachtler, W.M.H. Stability of supported transition metal catalysts in the
hydrogenation of nitriles. Appl. Catal. A Gen. 1991, 182, 365–378. [CrossRef]
20. Balladur, V.; Fouilloux, P.; deBellefon, C. Monometallic Ni, Co and Ru, and bimetallic NiCr, NiTi and CoFe
Ziegler-Sloan-Lapporte catalysts for the hydrogenation of adiponitrile into hexamethylenediamine: Effect of
water and dopants. Appl. Catal. A Gen. 1995, 133, 367–376. [CrossRef]

Sample Availability: Samples of the compounds are available from the authors.

© 2018 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access
article distributed under the terms and conditions of the Creative Commons Attribution
(CC BY) license (http://creativecommons.org/licenses/by/4.0/).