Improved catalytic performance of acid-activated sepiolite supported nickel and potassium bimetallic catalysts for liquid phase hydrogenation of 1,6-hexanedinitrile

Article abstract of DOI:10.1016/j.molcata.2016.10.029

Different inorganic acids were used to activate sepiolite, and the acid-activated sepiolites supported nickel and potassium bimetallic catalysts were prepared. Nitrogen adsorption-desorption, hydrogen chemisorption, ammonia temperature programmed desorption (NH₃-TPD), temperature programmed reduction (TPR), powder X-ray diffraction (XRD), scanning electron microscopy (SEM), transmission electron microscopy (TEM), Fourier Transform Infrared Spectroscopy (FTIR) and energy dispersive X-ray (EDX) were used to characterize the catalysts. The catalytic performance of the acid-activated sepiolite supported K-Ni bimetallic catalysts were investigated in 1,6-hexanedinitrile (HDN) hydrogenation in liquid phase. It was revealed that the potassium could increase the alkalinity of the catalyst with the aim of inhibiting the formation of the 1-azacycloheptane (ACH). And the addition of potassium reduces the particle size of nickel and improves its dispersion. Compared with hydrochloric acid and sulfuric acid, nitric acid treatment increases more silanol groups (Si[sbnd]OH) on the sepiolite surface, which is helpful to nickel particles adsorption and dispersion. Nitric acid activated sepiolite supported nickel and potassium bimetallic catalysts (K-Ni/NASEP) present the best catalytic performance, the conversion of HDN comes up to 92.0% under moderate conditions of lower temperature and pressure, the selectivity to 6-aminocapronitrile (ACN) and 1,6-hexanediamine (HDA) is up to 95.2%.

Full text of DOI:10.1016/j.molcata.2016.10.029

Accepted Manuscript
Title: Improved catalytic performance of acid-activated
sepiolite supported nickel and potassium bimetallic catalysts
for liquid phase hydrogenation of 1,6-Hexanedinitrile
Author: Yang Lv Fang Hao Pingle Liu Shaofeng Xiong He’an
Luo
PII:
S1381-1169(16)30451-4
DOI:
Reference:
http://dx.doi.org/doi:10.1016/j.molcata.2016.10.029
MOLCAA 10091
To appear in:
Journal of Molecular Catalysis A: Chemical
Received date:
Revised date:
Accepted date:
9-7-2016
13-10-2016
24-10-2016
Please cite this article as: Yang Lv, Fang Hao, Pingle Liu, Shaofeng
Xiong, He’an Luo, Improved catalytic performance of acid-activated
sepiolite supported nickel and potassium bimetallic catalysts for liquid phase
hydrogenation of 1,6-Hexanedinitrile, Journal of Molecular Catalysis A: Chemical
http://dx.doi.org/10.1016/j.molcata.2016.10.029
This is a PDF file of an unedited manuscript that has been accepted for publication.
As a service to our customers we are providing this early version of the manuscript.
The manuscript will undergo copyediting, typesetting, and review of the resulting proof
before it is published in its final form. Please note that during the production process
errors may be discovered which could affect the content, and all legal disclaimers that
apply to the journal pertain.

Improved catalytic performance of acid-activated sepiolite
supported nickel and potassium bimetallic catalysts for liquid
phase hydrogenation of 1,6-Hexanedinitrile
Yang Lv, [a, b] Fang Hao, *[a] Pingle Liu, * [a, b] Shaofeng Xiong, [a] Hean Luo [a, b]
[
[
a]
b]
College of Chemical Engineering, Xiangtan University, Xiangtan 411105, China
National & Local United Engineering Research Centre for Chemical Process Simulation and
Intensification, Xiangtan University.
﹡
Corresponding author. Tel.: +86 73158298005; Fax: +86 73158298267.
E-mail address: liupingle@xtu.edu.cn, haofang.happy@163.com
1

Graphical abstract
Research Highlights

Acid-activated sepiolite supported nickel and potassium bimetallic catalysts for 1,6-hexanedinitrile
hydrogenation.

Nitric acid treatment increases silanol groups which efficiently improve nickel particles adsorption and
dispersion.

K-Ni/NASEP catalysts present high activity and selectivity to 6-aminocapronitrile and 1,6-hexanediamine.
K-Ni/NASEP catalysts have a good industrial application prospect.

2

Abstract Different inorganic acids were used to activate sepiolite, and the acid-activated sepiolites supported nickel and potassium
bimetallic catalysts were prepared. Nitrogen adsorption-desorption, hydrogen chemisorption, ammonia temperature programmed
3
desorption (NH -TPD), temperature programmed reduction (TPR), powder X-ray diffraction (XRD), scanning electron microscopy (SEM),
transmission electron microscopy (TEM), Fourier Transform Infrared Spectroscopy (FTIR) and energy dispersive X-ray (EDX) were used
to characterize the catalysts. The catalytic performance of the acid-activated sepiolite supported K-Ni bimetallic catalysts were investigated
in 1,6-hexanedinitrile (HDN) hydrogenation in liquid phase. It was revealed that the potassium could increase the alkalinity of the
catalyst with the aim of inhibiting the formation of the 1-azacycloheptane (ACH). And the addition of potassium reduces the particle size of
nickel and improves its dispersion. Compared with hydrochloric acid and sulfuric acid, nitric acid treatment increases more silanol groups
(Si-OH) on the sepiolite surface, which is helpful to nickel particles adsorption and dispersion. Nitric acid activated sepiolite supported
nickel and potassium bimetallic catalysts (K-Ni/NASEP) present the best catalytic performance, the conversion of HDN comes up to 92.0 %
under moderate conditions of lower temperature and pressure, the selectivity to 6-aminocapronitrile (ACN) and 1,6-hexanediamine (HDA)
is up to 95.2 %.
Keywords 1,6-hexanedinitrile; Hydrogenation; 6-aminocapronitrile; 1,6-hexanediamine; Acid-activated sepiolites supported
catalyst
3

1
Introduction
Nylon-6 fibers and resins are manufactured from -caprolactam (CPL), an important petrochemical materials. The problem for the
production of -caprolactam mainly lies in its complicated process and a great deal of by-product of ammonium sulfate [1]. Recently,
researchers are developing new methods to obtain -caprolactam at lower cost and less environmental pollution. In comparison to the
current ketone-hydroxylamine processes, the route of butadiene hydrocyanation for -caprolactam production have such advantages, it
involves the synthesis of 1,6-hexanedinitrile (HDN), HDN hydrogenation to 6-aminocapronitrile (ACN) and ACN cyclization to

-caprolactam in water [2-4]. Among these steps, HDN hydrogenation to ACN is the critical step in this novel route, 1,6-hexanediamine
(HDA) and Schiff bases are the deep hydrogenation products [4]. Moreover, HDA is an significant industrial intermediate mainly applied to
produce nylon-66 [2, 5].
Researchers have tried HDN hydrogenation either in gas or liquid phase [6-12]. Nickel-based amorphous catalysts have shown higher
catalytic performance in various hydrogenation processes [13, 14]. However, the poor thermal stability limits its application in industry. For
instance, Hexing Li et al. [15, 16] have studied the effects of amorphous Ni-P and Ni-B for HDN hydrogenation in gas phase under higher
temperature and space velocity, and the selectivity to ACN is as low as 30 % and HDA is the main product. The HDN hydrogenation in gas
phase over Ni-MgO catalysts was recently reported by Serra et al. [6-8]. Although these catalysts exhibited better catalytic performance of
-
1
8
7 % HDN conversion and 83 % selectivity to ACN, the process was carried out at space velocity of 7162 h , 473 K and high consumption
of hydrogen. The characterization and activities of several potassium-doped nickel, alumina-supported nickel catalysts and iron-nickel
catalysts in HDN hydrogenation have been investigated by Medina et al. [17-20]. As to HDN hydrogenation in liquid phase catalyzed by
Raney Ni or Co in ammonia or alkali metal hydroxide solution at high pressure [21-26], Raney Ni or Co are very active, however, they are
pyrophoric and hard to divide from the process fluid. Moreover, a large quantity of ammonia is necessary to impede side reactions leading
to the conformation of secondary and tertiary amines. It has been reported that it is helpful to improve the selectivity to ACN by surface
modification of the catalyst [10, 11, 17, 19, 20, 27], and the additives and the supported metal type have great effects on the activity of the
nickel catalysts [11, 15-17, 19, 20, 27].
8 4 2 4 2
Si12O30(OH) (OH ) O and open porous framework
.nH
Sepiolite is a kind of fibrous clay material with the chemical formula of Mg
28-30]. Due to its peculiar structure and broad distribution in nature, sepiolite is widely used as catalyst support [28, 30]. Recently,
researchers pay much attention to active sepiolite by various surface modification methods in order to improve the adsorptivity of sepiolite
31]. During the acid activation process, the magnesium ions located in octahedral layers are dissolved and the tetrahedral silicate sheets are
[
[
retained and then the sepiolite changes into amorphous silica and leads to a great number of octahedral vacancies, which are beneficial to
integrate with various mental cations [32, 33]. Meanwhile, a large amount of silanol group form on the surface of sepiolite during the acidic
hydrolysis and these surface silanol groups (Si-OH) are the main active centers for adsorption [32, 34, 35]. Many researchers have been
studied on the modified sepiolite supported active metal catalysts in hydrogenation or oxidation reaction [36, 37]. However, HDN
4

hydrogenation over acid-modified sepiolite supported nickel and potassium bimetallic catalyst has not been reported.
In this study, different acid-activated sepiolites were used as catalyst support for loading nickel and potassium. The acid-activated
sepiolite supported catalysts were investigated in HDN hydrogenation under moderate conditions. It shows high activity and selectivity to
ACN and HDA.
2
Experimental
2
.1 Materials
2
Natural sepiolite was obtained from Henan Xinglei Sepiolite Limited Company, 95 % content of sepiolite (42.21 % SiO , 20.57 %
MgO, 18.83 % CaO, 0.37 % Al
2
O
3
, 0.21 % Fe
5.57 % Lost on ignition) was obtained by purification. 1,6-Hexanedinitrile (98.5 wt.%), 6-aminocapronitrile (98 wt.%) and
,6-hexanediamine (99 wt.%) were purchased from J&K scientific LTD. KNO , Ni(NO .6H O and ethanol were a n a l y t i c a l
(99.99 %) was obtained from Zhuzhou Diamond
2 3 2 2 2 5 2
O , 0.18 % K O, 0.08 % Na O, 0.05 % MnO, 0.02 % P O , 0.017 % TiO , 0.012 % FeO and
1
1
3
3
)
2
2
g r a d e a n d obtained from Sinopharm Chemical Reagent Corporation Limited. H
2
Gas Company.
2
.2 Catalyst preparation
The pre-prepared sepiolite powder (SEP) was washed by deionized water and acidified by nitric acid, hydrochloric acid or sulfuric acid
with the solid-to-liquid ratio of 1:15 at 348 K under mechanical agitation for 24 h. And the mixture was filtered, the precipitate was washed
with distilled water and dried at 393 K for 3 h. The sepiolite treated with nitric acid, hydrochloric acid or sulfuric acid was labeled as NASEP,
CASEP, SASEP. During the acid treatment, it can be seen from Scheme 1 that the Si-O-Mg-O-Si bond in sepiolite transform into two Si-OH
bonds and the Si-OH bonds condense to Si-O-Si bonds [31, 33].
The catalysts were synthesized by the incipient wetness method and through the following steps. Firstly, 5 g of acid-activated sepiolite
3 2 2 3
was impregnated in 25 ml of an aqueous solution including 49.50 mg of Ni(NO ) ·6H O and 0.65 mg of KNO and stirred for 12 h, followed
by drying at 393 K for 12 h, calcining at 623 K for 4 h in the fluid nitrogen stream, and reducing at 673 K for 4 h in the flowing hydrogen
stream. The prepared catalysts were labeled as Ni/CASEP, K-Ni/CASEP, Ni/NASEP, K-Ni/NASEP, Ni/SASEP and K-Ni/SASEP,
respectively. The amount of nickel is 20 wt.% and the amount of the potassium is 0.5 wt.%, respectively. Scheme 2 illustrates the process of
acid treatment and nickel adsorption.
2
.3 Catalyst characterization
The specific surface area, pore size distribution and pore volume were measured on Quantachrome NOVA-2200e automated gas sorption
system by the nitrogen adsorption-desorption. The specific surface area was calculated according to the Brunauer-Emmett-Teller (BET). The
Barrett- Joyner-Halenda (BJH) method was used to analyze the pore volume and pore size distributions.
Fourier Transform Infrared Spectroscopy (FTIR) was performed on Shimadzu IR Prestige-21 spectrometer with a KBr beam splitter.
5

2
Energy dispersive X-ray (EDX) elementary mappings were obtained from an FEI Tecnai G F30 S-TWIN field emission transmission
electron microscope equipped with an EDX spectrometer operating at 300 kV.
Powder X-ray diffraction (XRD) analysis were recorded by using a D/max 2500 TC diffractometer equipped with a back monochromator
α
operating under the conditions of tube voltage of 40 kV, a tube current of 30 mA by using Cu K radiation (λ = 1.542Å). The samples were
-
1
scanned at a rate of 1° min over the range from 5° to 90°.
The reducibility of the calcined samples was analyzed by temperature programmed reduction (TPR) on a Quantachrome ChemBET-3000
equipped with a thermal conductivity detector (TCD) and a procedure temperature-controlled furnace at the range of 273 K-1273 K. Firstly,
the samples were preheated to 473 K for 2 h in argon-atmosphere, and then constantly heated from 473 K to 1173 K at 5 K min-1 with the
-
1
existence of 40 mL min 5 vol.% of H
Ammonia temperature programmed desorption (NH
reduction (TPR). Firstly, the samples were preheated to 523 K for 1 h in the fluid helium gas atmosphere and cooled to 353 K for 30 min in
2
/Ar flow.
3
-TPD) was performed on the same instrument as that of temperature programmed
-
1
the same helium-atmosphere before exposed to a 10 vol.% of NH
3
/He flow with a rate of 40 mL min . The easily adsorbed NH
3
was purged
by helium flow for 30 min until consecutive pulses base line became stable, and then the NH
3
-TPD were programmed from 353 K to 1123 K
-
1
at a rate of 10 K min in the fluid helium-atmosphere.
Hydrogen chemisorption was also carried out on Quantachrome ChemBET-3000. Firstly, the samples were re-reduced by heating in a
flow of hydrogen stream at 523 K for 12 h, and then the system was purged with flowing argon stream to remove the adsorbed hydrogen on
the nickel surface and cooled to 303 K in the flowing argon stream. The injection of hydrogen pluse size was 0.02 ml until consecutive
pulses base line became stable. The number of nickel atoms on the surface area was reckoned by assuming the stoichiometric of one
hydrogen molecule per two surface nickel atoms. The surface area of nickel was reckoned from the nickel atomic cross-sectional area of a
nickel atom of 6.49×10-20 m /Ni atom.
2
Scanning electron microscope (SEM) micrographs of the samples was observed by a JEOL JSM-6610LV scanning microscope. The
SEM composites were firstly dispersed in dilute ethanol on a copperplate, and then evaporated the ethanol at the room temperature.
Transmission electron microscopy (TEM) was performed on JEOL 2010 FEG microscope at 200 kV. The samples were dispersed in
ethanol by the process of sonicate, and then the suspension was dropped onto amorphous carbon films supported on a copper grid.
2
.4 Procedure for the catalytic test
The reaction of liquid phase 1,6-Hexanedinitrile (HDN) hydrogenation was actualized as follows: 1,6-Hexanedinitrile (5 g), ethanol (50
ml), catalysts (1 g) were added in a 100 mL Teflon-lined autoclave fitted with pressure controller, temperature detector and magnetic stirrer.
After being sealed, the reactor was excluded air by using the hydrogen, and then the autoclave was heated to setting temperature and pressure
under vigorous stirring. During the experiment, the pressure was maintained constantly at 2 MPa, the stirring rate was kept at 1000 rpm, and
6

the reaction temperature was at the range of 363 K to 373 K. when the reaction finished, the catalyst was removed through filtration, and then
the compositions of sample were determined by GC (Agilent 7820A) on a 30 m DB-1701 capillary column and the content of reactants and
products were analyzed by internal standard method.
3
Results and discussion
3
.1 Characterization of catalyst
It can be seen from Table 1 that the surface area of the acid-activated sepiolite (ASEP) is larger than the original sepiolite (SEP). Since
acid activation can unfold the closed cell in the sepiolite, get rid of some impurities and increase the specific surface area and the porosity
32, 34]. The surface areas of the sepiolite treated with different acid is in the following order: SASEP < NASEP < CASEP. During acid
[
treatment, it removes Mg2+ situated in its octahedral layer but leaves Si4+ coordinated in its tetrahedral layer and the originally microporous
structure transforms into mesoporous one when the sepiolite transforms to amorphous-silica (Si-O-Si) [31, 33]. Moreover, the
amorphous-silica (Si-O-Si) has the tendency to condense, so does the surface area decrease. It can be seen from Scheme 1 that
Si-O-Mg-O-Si bond in sepiolite transforms into two Si-OH bonds after acid treatment with hydrochloric acid or nitric acid, and the Si-OH
bond condense to Si-O-Si bond after acid treatment with sulfuric acid. Furthermore, The BET surface area decreases after the metals are
loaded on the support.
Figure 1-S shows the adsorption isotherms curves and BJH pore size distributions of the samples. The samples show type IV isotherms,
indicating nitrogen capillary condensation in mesoporous sepiolite structure. All the samples show type H4 hysteresis loops, it indicates the
slit-shaped pore structure. The area of the hysteresis loop is related with the number of the pores, the larger area indicates more pores. And
the pore volume result in Table 1 also confirmed this.
The morphology and microstructure of the samples are shown in Figure 1. It is worthy to note that the original fibrous morphology of
sepiolite is kept after acid treatment. Fig. 1(a) shows that sepiolite has a stick-like aggregation made up of lots of fibers. It can be seen from
Fig. 1(b), Fig. 1(c) and Fig. 1(d) that the majority of those fibers were cleaved after acidification. Fig. 1(e) and Fig. 1(f) indicates that
nickel and potassium particles cover on the surface of the support.
The FT-IR spectra of SEP, CASEP, NASEP and SASEP are shown in Figure 2. The bands at 3437 cm-1 correspond to stretching
vibrations of silanol groups (Si-OH) in the octahedral sheet [38]. The bands around 1021 cm-1 which attribute to Si-O-Si stretching still
exist in acid-activated sepiolite [39], it means that the basic structure of sepiolite is not destoryed. The Si-O-Si vibration band at 1100 cm-1
appear obvious shoulder peak in acid-activated sepiolite [40], indicating that the bond angle of Si-O-Si band in tetrahedral sheet is changed.
The reason may be that the Mg2+ in the lattice is replaced by H . The Mg-OH bending vibration at 3680 cm-1 disappears after acid
treatment of nitric acid or sulfuric acid. And the band around 1635 cm-1 is ascribed to the bending vibration of -OH bond of zeolitic water
+
7

[
41]. The bands at 3437 cm-1 and 1645 cm-1 are progressively broadened in NASEP, it suggests that considerable proportion of silanol
-
1
-1
groups form after nitric acid treatment. It can be seen that the band of Si-OH (3437 cm ) and -OH (1645 cm ) is visibly diminished in
SASEP, it indicates that the silanol groups (Si-OH) is progressively transformed into amorphous silica (Si-O-Si) after acid treatment of
sulfuric acid.
3 3
Figure 3 shows the NH -TPD plots of the supports. Two NH desorption peaks are detected in acid-activated sepiolite in the range of
5
23-673 K and 773-873 K, these are corresponding to weak acid sites and strong acid sites respectively. However, sepiolite only shows the
desorption peak of weak acid sites. Moreover, it can be seen that the weak acid peaks are significantly shifted toward higher temperature
when sepiolite is treated with sulfuric acid.
Energy dispersive X-ray (EDX) pattern of the samples are shown in Figure 2-S. And the signals for Ni and K are well attributed to the
active component, the signals for Mg, Si, Ca and O are from sepiolite (Fig. 2-S (a)). It can be seen from Fig. 2-S (b), (c) and (d) that the
intensities of the Mg peaks decreased significantly after acid treatment. Fig. 2-S (c) shows that the peak of Mg is weaked and the peak of
Ca is almost vanished after nitric acid treatment. Fig. 2-S (d) shows that the peaks of Ca and S with strong intensity are present in sepiolite
treated with sulfuric acid. Fig. 2-S (e), (f) and (g) show the EDX patterns of acid-activated sepiolite supported K-Ni catalysts and
semi-quantitative EDX analysis of the samples are shown in Table 2. Each datum reported is the average value determined from the three
locations on the catalyst surface. A large number of highly active silanol groups (Si-OH) on the surface is helpful to grasp the active
component and make them well dispersed. Fig. 2-S (h) shows the elementoverlay of K-Ni/NASEP, it indicates a homogeneous distribution
of the components.
The TPR profiles of the unreduced samples are shown in Figure 4. It can be seen from Fig. 4 that the sepiolite treated with different
acid presents different initial reduction temperature, and the TPR peaks are shifted toward lower temperature when the sepiolite was treated
with nitric acid. The order of the initial reduction temperature of the samples are NiO/SASEP > NiO/CASEP > NiO/NASEP. The number
of silanol groups (Si-OH) in NASEP increases much more than SASEP and CASEP, and the silanol groups (Si-OH) in the sepiolite are
helpful to nickel particles adsorption and dispersion. There is a reduction peak around 1110 K in NiO/SASEP which is attributed to the
4 2
reduction peak of CaSO [42]. Furthermore, the peak of hydrogen consumption of NiO/ASEP is broader than that of K O-NiO/ASEP, it
means that the reduction rate of NiO/ASEP decreases because of the excessive sintering of nickel oxide particles [27]. This may be the
reason that the nickel particle size on Ni/ASEP is larger than that of K-Ni/ASEP.
8

Figure 5 presents the XRD curves of the samples. The diffraction peak of 2θ angle at 44.0, 52.0, 76.0 are ascribed to nickel and 37.0,
3.0, 62.0 are ascribed to nickel oxide. Table 3 depicts the crystalline phases and crystallite sizes of the samples. The peaks of nickel in
4
Ni/ASEP are more spiculate than those in K-Ni/ASEP, it means that K-Ni/ASEP possess smaller nickel particles. It indicates that doping of
potassium is benificial to the dispersion of nickel particles, and this is also testified by the following hydrogen chemisorption results in
Table 4. Some researchers have found that doping of the potassium can effectively inhibits the excessive sintering of nickel during the
reduction process [17, 18, 20, 27]. Compared with other catalysts, K-Ni/NASEP has broader diffraction peaks of nickel, and K-Ni/NASEP
presents smaller nickel crystallite size. The XRD patterns show that nickel oxide and nickel crystalline phase exist in Ni/CASEP, and the
peaks of nickel oxide in K-Ni/CASEP are more spiculate than those in Ni/CASEP, it indicates that nickel oxide is harder to be reduced with
the introduction of potassium.
2
Table 4 shows the data of H chemisorption. It indicates that the addition of potassium remarkably increase the amount of hydrogen
-
1
chemisorption. Compared with the results of other catalysts, K-Ni/NASEP presents the largest hydrogen uptake quantity of 50.1 μL g , the
2
-1
best nickel dispersion of 6.4 % and the largest metallic surface of 8.3 m g .
Figure 6 shows the TEM images of the samples. The dark dots represent the nickel nanoparticles on the tube-like acid-activated
sepiolite surfaces. As shown in Fig. 6 (c), (f), the nickel nanoparticles in K-Ni/SASEP seriously agglomerate. But nanoparticles of nickel in
K-Ni/NASEP (Fig. 6 (b), (e)) and K-Ni/CASEP (Fig. 6 (a), (d)) are well-distributed. The size of the nickel particles are about 40 nm in
K-Ni/SASEP and 20 nm in K-Ni/CASEP, However, they are approximate 10 nm in K-Ni/NASEP, it indicates that silanol groups (Si-OH)
increase through acidic hydrolysis and efficiently decrease the nickel nanoparticle size and improve its distribution.
3
.2 Catalytic performance
HDN hydrogenation was carried out by using ethanol as solvent over the Ni-based/ASEP catalysts. It can be seen from Scheme 3 that the
hydrogenation products consist of 6-aminocapronitrile (ACN), 1,6-hexanediamine (HDA), 1-azacycloheptane (ACH) and C12 compounds
7, 9, 10, 15]. Crystalline Ni-K/α-Al catalyst presented good activity in HDN hydrogenation [27]. Yet, the nickel amount is as high as 30
[
2 3
O
wt.% and the reaction temperature is as high as 423 K. The results of HDN hydrogenation catalyzed by different sepiolite supported catalysts
are shown in Table 5. It shows that CASEP and NASEP supported catalysts present higher selectivity to ACN and HDA, while SASEP
supported catalysts show higher selectivity to ACN and ACH. Generally, the conversion of HDN increases while the ACN selectivity
decreases evidently with the raise of reaction temperature. Overall, sepiolite supported catalysts doped with potassium possess higher
selectivity to ACN and HDA. This means that potassium doped catalysts can effectively inhibit the transimination reactions leading to ACH
and C12 by-products. Lanthanum, potassium, chromium, titanium or other dopant agents have been shown great influences on the catalytic
9

performance of nickel supported catalysts in liquid or gas phase hydrogenation of HDN [5, 11, 17, 19, 20, 27]. Thorpe-Ziegler intramolecular
condensation of HDN to 2-aminomethylcyclopentylamine (C-C bond formation) is inhibited due to the existence of potassium [10]. It has
been reported that the selectivity to aminonitrile increases when dinitriles hydrogenation over Raney Ni was carried out in the reaction
medium with the addition of inorganic base [2, 21, 22]. Some researchers explained that the effect of inorganic base addition by the
stabilization of an intramolecular interaction of ACN to an heterocyclic form containing imino form of the CN group, and such cyclic
conformation would hinder the aminonitrile re-adsorption, which can further hydrogenate to azacycloheptane-2-amine [10, 22, 44]. The
azacycloheptane-2-amine could be transformed to 1-azacycloheptane (ACH) by the hydrogenation and deamination (Scheme 3). And the
alkalinity on catalyst surface increases due to the addition of potassium [11, 17, 19, 20, 27], which may impede 1-amino-6-imino-hexane
intramolecular condensation to ACH [7, 10, 15]. Didier et al. prepared the Ni-based catalysts elaborated from lamellar double hydroxide
precursors and applied in selective half-hydrogenation of HDN, the HDN conversion is about 70 % and the ACN selectivity is 66 % under
2
.5 Mpa H
2
pressure and 353 K reaction temperature [10]. Compared with the Ni-based acid-activated sepiolite catalyst, the preparation
O and inorganic base (KOH) during the HDN hydrogenation,
method of the catalyst is complicated. Moreover, it needs a certain amount of H
2
which can result in serious equipment corrosion and large amount of wastewater. From the characterization results (Fig. 5, Table 4, Fig. 6),
we can see that the nickel
dispersion is improved with the addition of potassium, and nickel crystallite size decreases as well as the metallic surface area increases. This
may be the reason that potassium doped catalysts show higher activity at the same reaction conditions. Table 5 shows that K-Ni/CASEP
gives 78.0 % HDN conversion, while Ni/CASEP gives 70.4 % HDN conversion at 373 K. Similarly, K-Ni/SASEP gives 64.8 % HDN
conversion at 373 K and the HDN conversion is only 60.8 % over Ni/SASEP at the same reaction conditions. Moreover, the HDN conversion
is 92.0 % over K-Ni/NASEP and 81.4 % over Ni/NASEP at 373 K respectively. Nitric acid treated sepiolite supported nickel and potassium
bimetallic catalysts present the best catalytic performance, the HDN conversion comes up to 92.0 % under moderate reaction conditions of
3
73 K and 2.0 MPa, the ACN and HDA selectivity is up to 95.2 %.
It has been reported that two kinds of adsorption sites exist on the internal and external surface of sepiolite, one is the silanol groups
(
Si-OH) and the other is water molecules integrated with Mg2+ [34, 35, 43]. And some researchers have found that silanol groups (Si-OH) on
the surface of sepiolite possess synergistic effect with the active component and it is in favor of its high dispersion [37]. The results of FT-IR
spectra indicates that the density of silanol groups (Si-OH) increase highly after nitric acid treatment, and the silanol groups on the sepiolite
surface might aid in nickel adsorption and dispersion, this may be the reason that K-Ni/NASEP and Ni/NASEP catalysts possess higher
catalytic performance. And it can be seen from the spectra of FT-IR that the silanol groups (Si-OH) is progressively transformed into
amorphous silica (Si-O-Si) after sulfuric acid treatment. However, amorphous silica (Si-O-Si) is not beneficial to nickel adsorption and
dispersion. The results of TEM and XRD patterns also show that SASEP supported catalysts present larger nickel particles and agglomerate
10

phenomenon. It can be noted that SASEP supported catalysts possess relatively lower conversion of HDN and lower selectivity to ACN and
HDA. The reason may be that nickel nanoparticles agglomerate seriously on the surface of the sulfuric acid treated sepiolite. Another reason
4
may be that the formed CaSO in SASEP may go against nickel particles adsorption and dispersion. And a certain amount of strong acid sites
on Ni/SASEP or K-Ni/SASEP help to generate ACH and C12 by-products, so does the selectivity to ACN and HDA decrease.
3
.3 The recycle of the catalysts
The stability of K-Ni/NASEP has been tested and the recycling results are shown in Table 6. When the reaction finished, K-Ni/NASEP
was separated through filtration, washed with ethanol repeatedly and then dried at room temperature. The K-Ni/NASEP was re-activated in
hydrogen fluid at 673 K for 4 h. The results in Table 6 indicate that the HDN conversion maintains at 87.4 % and selectivity to ACN and
HDA keeps at 85.7 %. However, the catalytic activity decreases slightly in third runs. Through the apparatus of inductively coupled plasma
atomic emission spectroscopy (ICP-AES) analysis, the nickel content of the third recycled K-Ni/NASEP indicates that the leaching of the
nickel can be neglected. Figure 7 shows XRD patterns and TEM image of the third recycled K-Ni/NASEP. It can be seen that the
diffraction peaks of nickel in the third recycled K-Ni/NASEP is more spiculate than the fresh K-Ni/NASEP, which indicates that larger
nickel particles exist in third recycled K-Ni/NASEP, and the TEM image also proves this. The reason may be excessive sintering of the
nickel particles during the process of regeneration.
4
Conclusions
The sepiolite was treated with different inorganic acids and the acid-activated sepiolites supported nickel and potassium bimetallic
catalysts were obtained through incipient wetness method. The prepared catalysts were characterized by various methods such as nitrogen
adsorption-desorption, hydrogen chemisorption, ammonia temperature programmed desorption (NH -TPD), temperature programmed
3
reduction (TPR), powder X-ray diffraction (XRD), transmission electron microscopy (TEM), scanning electron microscopy (SEM), Fourier
Transform Infrared Spectroscopy (FTIR) and energy dispersive X-ray (EDX). The acid-activated sepiolite supported nickel and potassium
bimetallic catalysts were investigated in liquid phase HDN hydrogenation. The products consist of ACN, HDA, ACH and C12. It was
revealed that the potassium could increase the alkalinity of the catalyst with the aim of inhibiting the formation of the 1-azacycloheptane
(ACH). And the addition of potassium reduces the particle size of nickel and improves its dispersion. Compared with the hydrochloric acid
and sulfuric acid, nitric acid treatment increases more silanol groups (Si-OH) on the sepiolite surface, which is in favor of nickel particles
adsorption and dispersion. The results show that nitric acid activated sepiolite supported nickel and potassium bimetallic catalysts
(
K-Ni/NASEP) present the best catalytic performance, the HDN conversion is as high as 92.0 % under moderate reaction conditions of 373
K and 2.0 MPa, the selectivity to ACN and HMDA is up to 95.2 %.
11

Acknowledgments
This work was supported by NSFC (21276218, 21306160), Natural Science Foundation of Hunan Province (2016JJ2123) and Project of
Hunan Provincial Science and Technology Department (2015GK1060)
12

References
[
[
(
[
1] H. Kajikuri, M. Kitamura, Y. Higashio, US Patent 5,304,643 (1994) to Sumitomo Chemical Company.
2] R. Fischer, P. Bassler, H. Luyken, F. Ohlbach, J.P. Melder, M. Merger, A. Ansmann, A. Rehfinger, G. Voit, US Patent 6,359,178B1
2002) to BASF Aktiengesellschaft.
3] P. Bassler, H. Luyken, G. Achhammer, T. Witzel, E. Fuchs, R. Fischer, W. Schnurr, US Patent 5,717,090 (1998) to BASF
Aktiengesellschaft.
[
[
[
[
[
[
[
[
[
[
[
[
[
[
[
[
[
[
[
[
[
[
[
[
4] R. Sheldon, Green Chem. 2 (2000) G1-G4.
5] V. Balladur, P. Fouilloux, C. Bellefon, Appl. Catal. A 133 (1995) 367-376.
6] M. Serra, J. Catal. 209 (2002) 202-209.
7] M. Serra, P. Salagre, Y. Cesteros, F. Medina, J.E. Sueiras, Appl. Catal. A 272 (2004) 353-362.
8] M. Serra, P. Salagre, Y. Cesteros, F. Medina, J.E. Sueiras, Solid State Ionics 134 (2000) 229-239.
9] S. Alini, A. Bottino, G. Capannelli, A. Comite, S. Paganelli, Appl. Catal. A 292 (2005) 105-112.
10] D. Tichit, R. Durand, A. Rolland, B. Coq, J. Lopez, P. Marion, J. Catal. 211 (2002) 511-520.
11] L. Zhao, C.Y. Wang, J.X. Chen, J.Y. Zhang, Chin. J. Chem. 18 (2007) 685-688.
12] Z. Jia, B. Zhen, M. Han, C. Wang, Catal. Commun. 73 (2016) 80-83.
13] H. Li, H. Li, W.L. Dai, J.F. Deng, Appl. Catal. A 207 (2001) 151-157.
14] G. Bai, Z. Zhao, L. Niu, H. Dong, M. Qiu, F. Li, Q. Chen, G. Chen, Catal. Commun. 23 (2012) 34-38.
15] X. Yu, H. Li, J.F. Deng, Appl. Catal. A 199 (2000) 191-198.
16] H. Li, Y. Xu, H. Li, J.F. Deng, Appl. Catal. A 216 (2001) 51-58.
17] F. Medina, P. Salagre, J.E. Sueiras, J.L.G. Fierro, Solid State Ionics 59 (1993) 205-210.
18] F. Medina, P. Salagre, J.E. Sueiras, J.L.G. Fierro, J. Chem. Soc. Faraday T. 89 (1993) 3507-3512.
19] F. Medina, P. Salagre, J.E. Sueiras, J.L.G. Fierro, J. Mol. Catal. 81 (1993) 387-395.
20] F. Medina, P. Salagre, J.E. Sueiras, J.L.G. Fierro, J. Chem. Soc. Faraday T. 89 (1993) 3981-3986.
21] M.C. Cotting, L. Gilbert, US Patent 5,981,790 (1999) to Rhone-Poulenc Fiber and Resin Intermediates.
22] S.B. Ziemecki, US Patent 5,151,543 (1992) to E.I. Du Pont de Nemours and Company.
23] M. Harper, WO Patent 00,27526A1 (2000) to E.I. Du Pont de Nemours and Company.
24] R.P. Beatty, R.A. Paciello, US Patent 5,554,778 (1996) to E.I. Du Pont de Nemours and Company.
25] R.P. Beatty, R.A. Paciello, US Patent 5,559,262 (1996) to E.I. Du Pont de Nemours and Company.
26] R.P. Beatty, R.A. Paciello, US Patent 5,599,962 (1997) to E.I. Du Pont de Nemours and Company.
27] H.G. Liao, S.H. Liu, F. Hao, P.L. Liu, K.Y. You, D.D. Liu, H.a. Luo, React. Kinet. Mech. Cat. 109 (2013) 475-488.
13

[
[
[
[
[
28] J.A. Anderson, S.E. Falconer, M. Galán-Fereres, Spectrochim. Acta, Part A 53 (1997) 2627-2639.
29] K. Núñez, R. Gallego, J.M. Pastor, J.C. Merino, Appl. Clay Sci. 101 (2014) 73-81.
30] S. Damyanova, L. Daza, J.L.G. Fierro, J. Catal. 159 (1996) 150-161.
31] E. Sabah, M. Turan, M.S. Çelik, Water Res. 36 (2002) 3957-3964.
32] J.L. Valentín, M.A. López-Manchado, P. Posadas, A. Rodríguez, A. Marcos-Fernández, L. Ibarra, J. Colloid Interface Sci. 298 (2006)
7
94-804.
[
[
[
[
[
[
[
[
[
[
[
[
33] F. Wang, J.S. Liang, Q.G. Tang, J.P. Meng, Z.Z. Wu, G.S. Li, Trans. Nonferrous Met. Soc. China 16 (2006) s406-s410.
34] E. Sabah, M. Majdan, J. Food Eng. 91 (2009) 423-427.
35] E. Sabah, J. Colloid Interface Sci. 310 (2007) 1-7.
36] A. Esteban-Cubillo, R. Pina-Zapardiel, J.S. Moya, M.F. Barba, C. Pecharromán, J. Eur. Ceram. Soc. 28 (2008) 1763-1768.
37] Y. Ma, G. Zhang, Chem. Eng. J. 288 (2016) 70-78.
38] Q. Zhu, Y. Zhang, F. Lv, P.K. Chu, Z. Ye, F. Zhou, J. Hazard. Mater. 217-218 (2012) 11-18.
39] S. Shoval, K.H. Michaelian, M. Boudeulle, G. Panczer, I. Lapides, S. Yariv, J. Therm. Anal. Calorim. 69 (2002) 205-225.
40] N. Zotov, H. Keppler, Chem. Geol. 184 (2002) 71-82.
41] M. Zhang, L. Shi, S. Yuan, Y. Zhao, J. Fang, J. Colloid Interface Sci. 330 (2009) 113-118.
42] B.S. Kim, H.Y. Sohn, Ind. Eng. Chem. Res. 41 (2002) 3092-3096.
43] S. Liu, M. Chen, L. Chu, Z. Yang, C. Zhu, J. Wang, M. Chen, Int. J. Hydrogen Energy 38 (2013) 3948-3955.
44] S. B. Ziemecki, Heterogeneous Catalysis and Fine Chemical III, in: M. Guisnet, J. Barbier, J. Barrault, C. Bouchoule, D. Duprez, G.
Perot, and C. Montassier (Eds.), Elsevier, Amsterdam, 1993, p. 283.
14

Figure 1 SEM images of SEP (a), CASEP (b), NASEP (c), SASEP (d) and K-Ni/NASEP (e, f)
15

Figure 2 FT-TR spectra of SEP (a), CASEP (b), NASEP (c), SASEP (d)
16

3
Figure 3 NH -TPD profiles of SEP (a), CSEP (b), NASEP (c), SASEP (d)
17

2 2 2 2
Figure 4 H -TPR profiles of K O-NiO/CASEP (a), NiO/CASEP (b), K O-NiO/NASEP (c), NiO/NASEP (d), K O-NiO/SASEP (e), NiO/SASEP (f)
18

2 2
Figure 5 XRD patterns of the samples: Ni/CASEP (a), K-Ni/CASEP (b), Ni/NASEP (c), K-Ni/NASEP (d), Ni/SASEP (e), K-Ni/SASEP (f), Ni/ SiO (g), K-Ni/SiO (h)
19

Figure 6 TEM images of K-Ni/CASEP (a, d), K-Ni/NASEP (b, e), K-Ni/SASEP (c, f)
20

Figure 7 XRD spectra of the third recycled K-Ni/NASEP (a), the fresh K-Ni/NASEP (b) and TEM image of the third recycled K-Ni/NASEP
21

Scheme 1 The process of ion exchange in sepiolite during acid treatment
22

Scheme 2 Process of acid treatment and nickel adsorption
23

Scheme 3 The possible products obtained by catalytic hydrogenation of 1,6-hexanedinitrile
24

Table 1 Textural property of different catalysts
Catalysts
BET surface area
Average pore diameter
Pore volume
2
-1
3 -1
(
m g )
(nm)
6.0
(cm g )
0.03
0.12
0.15
0.09
0.01
0.08
0.09
0.06
0.06
0.06
0.06
0.02
0.02
SEP
28.3
CASEP
NASEP
SASEP
124.7
94.8
52.5
1.8
5.9
5.8
7.4
SiO
2
8.1
Ni/CASEP
40.2
34.5
26.2
15.5
22.2
21.1
5.9
11.4
8.8
K-Ni/CASEP
Ni/NASEP
11.0
14.0
10.5
11.0
10.7
10.2
K-Ni/NASEP
Ni/SASEP
K-Ni/SASEP
Ni/SiO
2
K-Ni/SiO
2
6.4
25

Table 2 Semi-quantiative EDX analysis of the samples
Weight (%)
SEP
42.4
15.3
25.8
16.5
--
Element
CASEP
50.9
9.2
NASEP
52.5
3.4
SASEP
43.6
--
K-Ni/CASEP
K-Ni/NASEP
K-Ni/SASEP
34.3
O
39.9
8.1
37.2
2.2
38.7
--
Mg
Si
Ca
K
--
33.6
6.3
44.1
--
27.8
16.0
--
30.2
4.8
28.9
14.9
--
--
0.5
0.6
21.3
--
0.3
Ni
S
--
--
--
--
16.5
--
11.3
--
--
--
12.6
10.3
26

Table 3 XRD identification of different samples
a
a
Samples
Crystalline phases
Crystallite sizes of NiO (nm)
Crystallite sizes of Ni (nm)
Ni/ SiO
2
Ni
--
57.1
38.3
28.3
18.3
22.1
11.9
37.8
32.2
K-Ni/SiO
2
NiO, Ni
NiO, Ni
NiO, Ni
Ni
7.6
4.5
12.8
--
Ni/CASEP
K-Ni/CASEP
Ni/NASEP
K-Ni/NASEP
Ni/SASEP
Ni
--
Ni
--
K-Ni/SASEP
Ni
--
a
Crystallite size estimated from the Scherrer equation.
27

2
Table 4 H chemisorption data of Ni-based series catalysts
-
1
2
-1
Catalysts
H
2
uptake quantity (μL g )
Metallic surface areas (m g )
Dispersion (%)
Ni/CASEP
18.7
31.1
36.3
50.1
12.3
21.1
4.2
5.6
7.1
8.3
2.4
3.5
3.1
4.1
4.8
6.4
1.2
1.8
K-Ni/CASEP
Ni/NASEP
K-Ni/NASEP
Ni/SASEP
K-Ni/SASEP
28

Table 5 Catalytic performance of Ni-based catalysts for the hydrogenation of HDN under the different reaction temperature
Selectivity (%)
Catalyst
HDN conversion (%)
Temperature (K)
b
ACN
58.5
60.2
65.8
59.5
53.2
72.3
65.2
51.3
64.2
60.1
50.6
69.5
61.8
37.1
54.2
46.8
48.9
56.8
50.4
54.3
HDA
7.1
Others
34.4
28.9
22.0
25.0
24.6
15.6
15.6
17.3
20.3
19.3
13.8
6.1
ACN and HDA
65.6
Ni/ SiO
2
55.4
60.5
373
373
363
373
383
363
373
383
363
373
383
363
373
383
363
373
383
363
373
383
K-Ni/SiO
2
10.9
12.2
15.5
22.2
12.1
19.2
31.4
15.5
20.6
35.6
24.4
33.4
50.6
1.3
71.1
6
4.7
78.0
a
Ni/CASEP
70.4
75.0
8
7
9.4
2.4
75.4
84.4
a
K-Ni/CASEP
78.0
84.4
9
7
2.6
2.6
82.7
79.7
a
Ni/NASEP
81.4
80.7
9
8
5.6
5.4
86.2
93.9
a
K-Ni/NASEP
92.0
4.8
95.2
9
5
9.3
1.4
12.3
44.5
52.8
48.4
41.8
46.7
39.3
87.7
55.5
a
Ni/SASEP
60.8
0.4
47.2
7
6
3.9
0.4
2.7
51.6
1.4
58.2
a
K-Ni/SASEP
64.8
6.6
2.9
53.3
7
6.4
60.7
a
Reaction conditions: Stirring rate, 1,000 rpm; PH2, 2.0 MPa; Reaction time, 360 min; Ethanol, 50 mL; HDN, 5 g; Catalyst, 1 g.
b
Others: 1-azacycloheptane (ACH), mainly.
29