Surface properties of Ni/MgO catalysts for the hydrogenation of lauronitrile

Article abstract of DOI:10.1007/s10562-010-0297-0

60%Ni/MgO (wt%) catalysts were prepared by the co-precipitation method and the influence of n-butanol treatment was investigated. The results showed that the treatment with n-butanol improved the dispersion and reducibility of supported nickel, resulted in an increase of H₂ uptake from 410 to 582 μmol/g, corresponding to an increase of active Ni surface area from 32 to 46 m²/g (increased by 42%). Accordingly, the catalytic activity for the hydrogenation of toluene to methyl cyclohexane was significantly increased. Microcalorimetric adsorption of H₂ and CO indicated that the treatment with n-butanol increased the amount of active metal sites on the surface, without the change of electron densities of supported nickel surface. Microcalorimetric adsorption of CO₂ and NH₃ revealed the strong surface basicity and weak surface acidity for the Ni/MgO catalysts, especially for the reduced ones. The initial heat for the adsorption of acetonitrile was measured to be about 130 kJ/mol on the Ni/MgO catalysts, indicating the strong interaction between acetonitrile and the supported nickel, which might be an important factor determining the activity of nickel for the hydrogenation of aliphatic nitriles. The surface basicity of the Ni/MgO catalysts might play a role in inhibiting the formation of secondary and tertiary amines and therefore improved the selectivity to primary amine during the hydrogenation of lauronitrile to laurylamine. In addition, the Ni/MgO-B catalyst prepared with n-butanol treatment seemed more active for the hydrogenation of lauronitrile.

Full text of DOI:10.1007/s10562-010-0297-0

Catal Lett (2010) 135:246–255
DOI 10.1007/s10562-010-0297-0
Surface Properties of Ni/MgO Catalysts for the Hydrogenation
of Lauronitrile
•
Hui Chen Mingwei Xue Jianyi Shen
•
Received: 6 December 2009 / Accepted: 1 February 2010 / Published online: 17 February 2010
Ó Springer Science+Business Media, LLC 2010
Abstract 60%Ni/MgO (wt%) catalysts were prepared by
the co-precipitation method and the influence of n-butanol
treatment was investigated. The results showed that the
treatment with n-butanol improved the dispersion and
reducibility of supported nickel, resulted in an increase of
H2 uptake from 410 to 582 lmol/g, corresponding to an
increase of active Ni surface area from 32 to 46 m2/g
(increased by 42%). Accordingly, the catalytic activity for
the hydrogenation of toluene to methyl cyclohexane was
significantly increased. Microcalorimetric adsorption of H2
and CO indicated that the treatment with n-butanol
increased the amount of active metal sites on the surface,
without the change of electron densities of supported nickel
surface. Microcalorimetric adsorption of CO2 and NH3
revealed the strong surface basicity and weak surface
acidity for the Ni/MgO catalysts, especially for the reduced
ones. The initial heat for the adsorption of acetonitrile was
measured to be about 130 kJ/mol on the Ni/MgO catalysts,
indicating the strong interaction between acetonitrile and
the supported nickel, which might be an important factor
determining the activity of nickel for the hydrogenation of
aliphatic nitriles. The surface basicity of the Ni/MgO cat-
alysts might play a role in inhibiting the formation of
secondary and tertiary amines and therefore improved the
selectivity to primary amine during the hydrogenation of
lauronitrile to laurylamine. In addition, the Ni/MgO-B
catalyst prepared with n-butanol treatment seemed more
active for the hydrogenation of lauronitrile.
Keywords Ni/MgO catalysts Á Drying with n-butanol Á
Microcalorimetric adsorption Á Surface basicity Á
Hydrogenation of toluene Á Hydrogenation of lauronitrile
1 Introduction
Aliphatic amines are widely used as solvents, surfactants,
intermediates for pharmaceuticals and resins, textile addi-
tives, disinfectants, rubber stabilizers and corrosion inhib-
itors [1–5]. Catalytic hydrogenation of nitriles is an
important process for the preparation of diverse amines
industrially. For example, hexamethylenediamine, stearyl-
amine, dodecylamine, benzylamine, and xylyenediamine
are usually prepared by the hydrogenation of corresponding
adiponitrile, stearonitrile, lauronitrile, benzonitrile, and
1,4-dicyanobenzene [1–10]. These reactions are usually
catalyzed in liquid phase at elevated temperatures and
pressures. Due to the high reactivity of reaction interme-
diates, condensation reactions may occur during the
hydrogenation process, resulting in a mixture of primary,
secondary and tertiary amines. Since the separation of these
reaction products is usually difficult, the control of selec-
tivity becomes an important issue for the hydrogenation of
a nitrile to an amine [8, 9].
The product distribution in the hydrogenation of a
nitrile to an amine depends on a few factors, such as the
catalysts used (metal, support and promoter), structure of
reactants and reaction conditions [1–4, 10–12]. Supported
or unsupported Co, Ni, Ru, Cu, Rh, Pd, and Pt have
been tested for the hydrogenation of nitriles to amines. It
was found that while Co, Ni, and Ru were selective to
primary amines, Cu and Rh were selective to secondary
amines, and Pt and Pd were quite selective to tertiary
amines [8, 9, 12]. Raney nickel is probably the most
H. Chen Á M. Xue Á J. Shen (&)
Lab of Mesoscopic Chemistry, School of Chemistry and
Chemical Engineering, Nanjing University, 210093 Nanjing,
China
e-mail: jyshen@nju.edu.cn
123

Surface Properties of Ni/MgO Catalysts
247
frequently used catalyst with high activity for the
industrial production of primary amines. However, the
use of Raney nickel may have some disadvantages. For
instance, the preparation of Raney nickel may consume a
great deal of energy and cause environmental pollutions,
since Ni–Al alloy must be prepared first and Al in the
alloy must be leached by a concentrated base. Moreover,
Raney nickel is pyrophoric, and it is usually difficult to
use in a fixed-bed reactor. Thus, it is desirable to
develop more active and selective catalysts for the
hydrogenation of nitriles to amines [3, 13–15].
agglomerations of small particles may occur in the
co-precipitates during drying and calcination procedures,
owing to the high surface tension and chemical bonding
among hydrated particles. Such coagulations may be
avoided in the supercritical fluid drying process, by
removing water at low surface tensions. However, the
supercritical fluid process is expensive, hazardous and
incompatible with many processing apparatuses [27].
Alternatively, a solvent replacing technique of drying
(usually using n-butanol) might be used to effectively
reduce the possibility of formation of chemical bonds
between particles and thus to prevent the formation of hard
agglomerates [28–30]. In particular, this technique has the
advantages of low cost and wide universality.
Supported nickel has been used for the hydrogenation
of nitriles to amines. Silica, silica-alumina, alumina,
zirconia, titania, magnesia, ceria, and carbon were
reported as supports for nickel catalysts [14–20]. Sup-
ports may have strong effect on the activity and selec-
tivity of supported nickel for the hydrogenation of
nitriles to amines, due to their different surface areas,
pore volumes and acid/base properties. For example, the
dispersion of nickel usually increases with surface areas
and the porosity of a support may affect the dispersion
of nickel, resistance to sintering and diffusion of reac-
tants and products [1]. The surface acidity of a support
may promote the formation of secondary and tertiary
amines. The addition of a basic material in the reaction
system may weaken the surface acidity and inhibit the
formation of secondary and tertiary amines. Industrially,
ammonia is usually added to improve the selectivity to
primary amines. The promoting effect of NH3 may come
from: (1) a thermodynamic influence on the reaction
between primary imine and amine, leading to secondary
imine and NH3, (2) a modification of the electronic
properties of the catalytic metal, and (3) a selective
poisoning of surface acid sites responsible for the cou-
pling reactions to by-products [1, 12, 21, 22]. The
decrease of surface acidity of the support usually favors
the formation of primary amines [16, 17, 21–23]. How-
ever, controversies exist in the literature as to the effect
of support on the hydrogenation of nitriles to amines.
Some authors believed that the nature of supports did not
have a significant effect on the selectivity of hydroge-
nation of nitriles [10, 24, 25]. Medina et al. [22] studied
the hydrogenation of acetonitrile over the Ni/MgAlO
catalysts prepared from the Ni/Mg/Al lamellar double
hydroxides (LDH). They found that the introduction of
Mg decreased the surface acidity and increased the
selectivity to the primary amine.
In this work, we prepared the Ni/MgO catalysts with
high loadings. N-butanol was used to replace water in
the samples during the drying process so that water may
be removed from the solution of n-butanol, rather than
removed directly from the samples. Varies techniques
such as X-ray diffraction (XRD), H2 temperature pro-
grammed reduction (TPR), transmission electron
microscopy (TEM), and H2/O2 adsorption were used to
characterize the reducibility, dispersion and particle sizes
of supported nickel. The surface acid/base properties
were determined by the microcalorimetric adsorption of
NH3 and CO2. The microcalorimetric adsorption of H2,
CO and acetonitrile were performed to determine the
properties of nickel surfaces in the catalysts. The cata-
lysts were evaluated by the liquid-phase hydrogenation
of lauronitrile to dodecylamine and the fixed-bed
hydrogenation of toluene to cyclohexane.
2 Experimental
2.1 Preparation of Catalysts
Ni/MgO catalysts with 60% Ni (weight percent) were
prepared by the co-precipitation method. Specifically, an
aqueous solution containing required amount of nickel and
magnesium nitrates was co-precipitated with a solution
containing a slightly excess of sodium carbonate. A green
precipitate was formed, filtered and washed thoroughly by
de-ionized water. Then, the filter cake was divided into two
parts. One was dried directly in an oven at 393 K over-
night, and denoted as Ni/MgO. The other one was added
into n-butanol, heated at 353 K for 12 h and then dried in
an oven at 393 K. The catalyst treated with n-butanol was
denoted as Ni/MgO-B.
Industrial catalysts are mostly high surface area solids,
on which active components are highly dispersed in the
form of small particles [26]. The dispersion of metals on
supports is affected by preparation method and treatment
conditions. Chemical co-precipitation is a widely used
method for preparing supported metal catalysts. However,
A support MgO-B and a catalyst Ni/SiO2-B (60% Ni)
were also prepared for comparison with the n-butanol
treatment procedure. The Raney Ni was an industrial cat-
alyst used by the Feixiang Chemicals Co., Ltd. in China.
123

248
H. Chen et al.
2.2 Characterization of Catalysts
Microcalorimetric adsorption of CO2, NH3, H2, CO and
acetonitrile was carried out on a Tian-Calvet heat-flux
apparatus. The microcalorimeter was connected to a gas
handling and volumetric adsorption system, equipped with
a Baratron capacitance manometer (MKS, USA) for
precision pressure measurement. The differential heat of
adsorption versus adsorbate coverage was obtained by
measuring the heats evolved when doses of a gas
(1–10 lmol) were admitted sequentially onto the catalyst
until the surface was saturated by the adsorbate. The
measurements were performed at 423 K for the adsorption
of CO2 and NH3, and at room temperature for the
adsorption of H2, CO and acetonitrile, respectively. The
samples were pretreated in H2 at 673 K for 2 h and evac-
uated at 673 K for 1 h before the measurements unless
stated otherwise.
X-ray diffraction (XRD) patterns were collected at ambient
atmosphere by a X’PERT diffractometer with a Cu Ka
˚
radiation (k = 1.5418 A) at a scan speed of 25°/min. The
applied voltage and current were 40 kV and 40 mA,
respectively.
TPR was performed by using a quartz U-tube reactor
loaded with about 50 mg of a dried sample in 5.03% H2/N2
(V/V) at a flow rate of 40 mL/min. The hydrogen con-
sumption was monitored by a thermal conductivity detector
(TCD). The reducing gas was first passed through the
reference arm of the TCD before entering the reactor. The
exiting gas passed through a trap filled with soda lime to
remove water and possibly CO2 and then reached the
second arm of the TCD. The temperature was linearly
raised from 303 to 1,173 K with a rate of 10 K/min.
The surface areas and pore sizes were measured by the
Gemini V analyzer using N2 as an adsorbate at 77.3 K and
calculated by the BET and BJH (desorption isotherms)
methods, respectively. The samples were pretreated in N2
flow at 673 K for 4 h before the measurements.
2.3 Hydrogenation of Toluene
The gas phase catalytic hydrogenation of toluene was
carried out in a fixed-bed vertical down-flow glass micro-
reactor operating at 393 and 413 K under atmospheric
pressure. The catalyst powders (about 40 mg after reduc-
tion) was diluted by quartz sands (80 mesh) and supported
on a glass frit with a bed height of about 4 cm. The reaction
feedstock containing toluene with n-hexane as a diluent
(1:2 wt/wt) was fed into the reactor at a fixed rate by using
a 2ZB-1L10 dual-plunger infinitesimal quality metering
pump that was previously calibrated. H2 flow was moni-
tored by a mass flow controller in order to maintain the
molar ratio of H2/toluene = 5/1. Prior to hydrogenation,
the catalyst was reduced in situ in flowing H2 at 673 K for
2 h. The weight hourly space velocity (WHSV) of toluene
was varied from 20 to 80 h-1. The reaction effluent was
collected periodically with an ice-water trap and analyzed
by a GC 122 gas chromatograph equipped with a flame
ionization detector employing an FFAP capillary column.
Transmission electron microscopic measurement was
carried out on the reduced samples using a JEOL electron
microscope (JEM-2100) with an accelerating voltage
of 200 kV. The samples were passivated before the
measurements.
H2 and O2 adsorptions were performed on a home-made
volumetric adsorption system at room temperature and
673 K, respectively. The catalyst was reduced in H2 at
673 K for 2 h and evacuated at 673 K for 1 h before the
measurements. Doses of H2 were admitted sequentially and
coverage of H2 was measured until the equilibrium pres-
sure (PH2) reached about 50 kPa. The uptake of H2 was
determined by extrapolating the coverage of isotherm to
PH2 = 0. The surface area of metallic nickel was calcu-
lated by the uptake of H2 assuming the molar ratio of
2
˚
H/Nisurf = 1 and a surface area of 6.5 A for a Ni atom
[31]. After the adsorption of H2, the catalyst was evacuated
again at 673 K for 1 h. Doses of O2 were then admitted
sequentially and the adsorption isotherm of O2 was mea-
sured until the equilibrium pressure reached about 50 kPa.
The amount of metallic nickel atoms in the catalyst were
calculated according to the uptake of O2 and the molar ratio
of O/Ni = 1 [32]. The reduction degree was then calcu-
lated by the number of nickel atoms titrated by O2
adsorption divided by the number of nickel atoms present
in the sample. Finally, the dispersion of nickel (D) could be
calculated according to the following equation: D (%) =
uptake of H2/uptake of O2 9 100, while the average Ni
particle size d (nm) could be derived from the equation (d
(nm) = 101/D (%)) suggested by Smith et al. [33]
assuming the spherical shape of metal particles.
2.4 Hydrogenation of Lauronitrile
The liquid-phase hydrogenation of lauronitrile was carried
out in a 120 mL stainless high-pressure micro-autoclave
equipped with a magnetic stirrer. Prior to hydrogenation,
the catalyst was reduced (about 0.5 g after reduction) at
673 K for 2 h in H2 flow, cooled down and then transferred
(without exposure to air) into the autoclave containing 30 g
lauronitrile and a small amount of KOH inhibitor (0.05 g
KOH dissolved in 0.5 mL H2O) which was used to sup-
press the formation of secondary and tertiary by-products.
Then, the reactor was sealed and purged with H2 three
times to remove air. Temperature was then raised up and
H2 was charged to the specified pressure. The hydrogena-
tion of lauronitrile was performed at 393 K and 2.0 MPa
123

Surface Properties of Ni/MgO Catalysts
249
for 50 min with a vigorous stirring. After the reaction, the
autoclave was cooled down to about 360 K and the product
mixture was immediately filtered. The conversion of lau-
ronitrile and selectivity to primary amine were calculated
from the total amine value (TAV), secondary and tertiary
amine value (2/3AV) and tertiary amine value (3AV) of the
product mixture. The total amine value (TAV) is the
amount of KOH (mg) equivalent to the basicity in 1 g of
amine products. Correspondingly, the secondary and ter-
tiary amine value (2/3AV) is defined as the amount of
KOH (mg) equivalent to the total basicity of secondary and
tertiary amines in 1 g of amine products. TAV was deter-
mined by a regular potentiometric titration method and
2/3AV and 3AV were determined by using an indicator
(bromophenol blue). The concentrations of amines were
titrated by using the isopropanol solution of HCl (0.1 M).
A mask agent, salicylaldehyde, was added to react with
primary amine so that 2/3AV was determined. A mask
agent, acetic anhydride, was added to react with primary
and secondary amines so that 3AV was measured. TAV
was determined when no mask agent was added.
Ni; MgO, NiO or NiO-MgO; Unidentified
Ni/MgO-B
Ni/MgO
Ni/MgO-B(R)
Ni/MgO(R)
20
40
60
80
2 theta (degree)
Fig. 1 XRD patterns of Ni/MgO and Ni/MgO-B after drying at
393 K and reduction (with R as suffix) at 673 K in H2 for 2 h. The
reduced samples were passivated in N2 flow containing 0.1% O2 at
room temperature for 48 h before the XRD measurements
3 Results and Discussion
Ni/MgO-B(R)
Ni/MgO(R)
3.1 Structural and Textural Properties of Catalysts
Figure 1 shows the wide angle X-ray diffraction patterns for
the Ni/MgO and Ni/MgO-B catalysts after drying at 393 K
and reduction in H2 at 673 K, respectively. No clear dif-
fraction peaks were observed for the two unreduced sam-
ples, suggesting that the catalysts were amorphous before
the reduction. After the reduction, two clear diffraction
peaks at 44.5 and 51.78 for metallic Ni were detected for the
two catalysts. The diffraction peaks at 37.0, 43.1 and 62.58
might belong to MgO, NiO or the solid solution formed
between NiO and MgO [34, 35]. These phases are difficult
to distinguish owing to their similar crystal structures [36,
37]. Since about 80% nickel were reduced in the catalysts
according to the results of O2 adsorption (data will be
presented shortly below), these diffraction peaks were not
likely attributable to NiO, but MgO-like phases (MgO and
NiO-MgO solid solution). Figure 1 showed that the two
XRD peaks around 43.1 and 44.5° for MgO-like phases and
metallic nickel, respectively, were weak, broadened and
overlapped. A Lorentz function was applied to deconvolute
the two XRD peaks (Fig. 2). The average particle sizes of
nickel in the Ni/MgO and Ni/MgO-B were then calculated
to be about 8.5 and 6.8 nm, respectively, according to the
peak widths and Scherrer equation.
40
42
44
46
48
50
2 theta (degree)
Fig. 2 Deconvolution of the two XRD peaks around 43.1° for MgO-
like phases and 44.5° for metallic nickel in the Ni/MgO and Ni/MgO-B
catalysts after the reduction in H2 at 673 K. The peak deconvolution
was performed using the Lorentz function
to increase the surface area of Ni/MgO-B after the calci-
nation, but it did increase the surface area from 61 m2/g for
Ni/MgO to 86 m2/g for Ni/MgO-B after the reduction. The
pore sizes (3.5–3.6 nm) and volumes (0.21–0.23 mL/g)
were similar for the two catalysts after the reduction.
3.2 Reducibility and Dispersion of Supported Nickel
Figure 3 shows the TPR profiles for the Ni/MgO and
Ni/MgO-B catalysts. Three reduction peaks were observed
for the Ni/MgO with peak temperatures at 666, 714 and
933 K, while only two reduction peaks occurred for the
Table 1 summarized the textural properties of the
Ni/MgO and Ni/MgO-B catalysts after the calcination and
reduction. It is seen that the use of n-butanol did not seem
123

250
H. Chen et al.
Table 1 Surface areas and pore structures of the Ni/MgO catalysts
Catalyst
Ni/MgO
Treatment
Surface area (m2/g)
Pore size (nm)
Pore volume (cm3/g)
Calcined
Reduced
Calcined
Reduced
167
61
4.5
3.6
3.6
3.5
0.40
0.21
0.23
0.23
Ni/MgO-B
133
86
Note: The catalysts were calcined at 673 K in N2 flow for 4 h, and reduced at 673 K in H2 flow for 2 h. The reduced catalysts were passivated in
N2 flow containing 0.1% O2 at room temperature for 48 h. The pore size was determined according to the desorption isotherms. The Ni/MgO was
dried at 393 K in oven directly, while the Ni/MgO-B was treated in butanol at 353 K for 12 h before they were further dried at 393 K in an oven
Ni/MgO-B at 618 and 659 K. Obviously, the treatment
with n-butanol favored the reduction of supported nickel
species since the temperatures of TPR peaks for Ni/MgO-B
were significantly lower than those for Ni/MgO.
the Ni/MgO-B and Ni/MgO, corresponding to the active
nickel areas of 46 and 32 m2/g, respectively. According to
the uptakes of O2 and H2, the dispersions of reduced nickel
were found to be 15 and 12% for the Ni/MgO-B and
Ni/MgO, respectively. The average particle sizes of
reduced Ni were then calculated to be 8.5 and 6.6 nm for
the Ni/MgO and Ni/MgO-B, respectively, in agreement
with the results of XRD line broadening.
The reduction degree usually increased with the reduc-
tion temperature. However, the higher reduction tempera-
tures might cause the decrease of dispersion of reduced
nickel on the surface, leading to decreased uptake of H2. In
this work, it was found that the reduction at 673 K pro-
duced the catalysts with the highest uptakes of H2. Figure 4
shows the H2 adsorption isotherms and uptakes at room
temperature for the Ni/MgO and Ni/MgO-B catalysts
reduced in H2 at 673 K. Table 2 summarizes the infor-
mation obtained by the adsorption of H2 and O2. According
to O2 uptakes, the degrees of reduction were 79 and 87%
for the Ni/MgO and Ni/MgO-B catalysts, respectively,
suggesting that the n-butanol treatment increased the
reducibility of nickel, consistent with the TPR results. The
H2 uptakes were measured to be 582 and 410 lmol/g on
Figure 5 displays TEM images for the Ni/MgO and
Ni/MgO-B reduced in H2 at 673 K. The images showed
that the size distribution of nickel particles was quite uni-
form. The average metallic Ni particle sizes in Ni/MgO and
Ni/MgO-B might be estimated to be about 7–8 nm. Table 3
summarized the particle sizes of supported nickel obtained
by different methods.
3.3 Microcalorimetric Adsorption of H2 and CO
The results of microcalorimetric adsorption of H2 and CO
on the Ni/MgO and Ni/MgO-B catalysts at room temper-
ature were shown in Fig. 6. The initial heats were similar
for the adsorption of H2 on the Ni/MgO and Ni/MgO-B
catalysts (about 80 kJ/mol), in consistent with the results
659
618
800
582
Ni/MgO-B
600
410
666
400
714
Ni/MgO
933
200
Ni/MgO-B
Ni/MgO
0
0
10
20
30
40
50
H2 pressure (kPa)
500
600
700
800
900
1000
T(K)
Fig. 4 Adsorption isotherms of H2 at room temperature for the
Ni/MgO and Ni/MgO-B catalysts after reduction at 673 K in H2 for
2 h, followed by evacuation at 673 K for 1 h
Fig. 3 TPR profiles for the Ni/MgO and Ni/MgO-B catalysts dried at
393 K
123

Surface Properties of Ni/MgO Catalysts
251
Table 2 Adsorption of H2 and O2 for the Ni/MgO catalysts
Sample
H2 ads. (lmol/g)
O2 ads. (lmol/g)
Surf. area of Ni (m2/g)
Diam. (nm)
Disp. (%)
Red. (%)
Ni/MgO
410
582
3,460
3,827
32
46
8.5
6.6
12
15
79
87
Ni/MgO-B
Note: All the catalysts were reduced in H2 at 673 K for 2 h and evacuated at 673 K for 1 h before the measurements. Adsorptions of H2 and O2
were performed at room temperature and 673 K, respectively. Reduction degree (Red.) was calculated according to the uptake of O2. Average
diameter (Diam.) and dispersion (Disp.) of supported nickel particles were calculated according to the uptake of H2 and the amount of nickel
reduced
150
H2/Ni/MgO-B
H2/Ni/MgO
120
CO/Ni/MgO-B
CO/Ni/MgO
90
60
30
0
0
100
200
300
400
500
600
Coverage (µmol/g)
Fig. 6 Differential heat versus adsorbate coverage for the adsorption
of H2 (filled square and open square) and CO (filled circle and open
circle) on the Ni/MgO and Ni/MgO-B catalysts at 298 K, respec-
tively. Samples were reduced at 673 K in H2 for 2 h and then
evacuated at 673 K for 1 h
similar to those reported in literature [22, 38, 41]. The heats
of adsorption remained nearly constant up to the CO cov-
erage of about 270 lmol/g, beyond which the differential
heats declined quickly with the further increase of CO
coverage.
CO is usually adsorbed on metallic nickel in linear or
bridged forms [41, 42]. Thus, the CO/H2 ratio of uptakes
should be between 2 (for linear adsorption of CO) and 1
(for bridged adsorption of CO). The CO/H2 ratio of uptakes
was smaller than 1 in this study, suggesting some extent of
hydrogen spillover occurred during the adsorption of H2
[43–45].
Fig. 5 TEM images of the reduced Ni/MgO (a) and Ni/MgO-B (b)
catalysts. The reductions were performed in H2 at 673 K for 2 h,
followed by passivation in N2 flow containing 0.1% O2 at room
temperature for 48 h
Table 3 Particle sizes of nickel in the Ni/MgO catalysts calculated
according to different methods
Since the initial heats of CO adsorption on the Ni/MgO
and Ni/MgO-B were almost the same, the surface electron
density of nickel particles in the two samples might be
similar.
Sample
XRD (nm) Uptakes of H2 and O2 (nm) TEM (nm)
Ni/MgO
8.5
8.5
6.6
8
7
Ni/MgO-B 6.8
3.4 Surface acidic and Basic Properties
reported in literature [22, 38–40]. However, the uptake of
H2 on the Ni/MgO-B (540 lmol/g) was significantly higher
than that on the Ni/MgO (425 lmol/g).
Microcalorimetric adsorptions of NH3 and CO2 are usually
used to characterize the surface acid–base properties of
catalysts. Figure 7 presents the results of microcalorimetric
adsorption of CO2 at 423 K on the supports and supported
The initial heats of CO adsorption on the Ni/MgO and
Ni/MgO-B catalysts were also similar (about 125 kJ/mol),
123

252
H. Chen et al.
nickel catalysts. The initial heat of CO2 adsorption on
Ni/MgO (128 kJ/mol) was slightly higher than on Ni/MgO-
B (112 kJ/mol), indicating that the n-butanol treatment
decreased the surface basicity of the catalyst. Before the
reduction, Ni/MgO exhibited lower basicity than MgO.
However, after the reduction, the surface basicity was
increased due to the reduction of NiO in the NiO-MgO
solid solution that led to the more exposure of clean MgO.
The reduced Ni/MgO catalyst exhibited strong surface
basicity, with the initial heat of about 153 kJ/mol and
saturation coverage of about 540 lmol/g for the CO2
adsorption. Similarly, before the reduction, the surface
basicity was lower on Ni/MgO-B than on MgO-B, while it
was higher on Ni/MgO-B than on MgO-B after the
reduction.
Ni/MgO and Ni/MgO-B were 37 and 75 kJ/mol, respec-
tively, showing the weak surface acidity of the two cata-
lysts. It was reported [46] that NiO also possessed the weak
surface acidity with an initial heat lower than 40 kJ/mol for
the adsorption of ammonia. After the reduction, the heats
of NH3 adsorption were increased to 97 and 130 kJ/mol for
the Ni/MgO and Ni/MgO-B, respectively, but they
decreased rapidly with the increase of NH3 coverage. We
suspected that the increased initial heats might be attributed
to the dissociative adsorption of NH3 on the metallic nickel
surfaces [47–49]. Thus, the microcalorimetric results for
NH3 adsorption indicated the weak surface acidity for the
Ni/MgO and Ni/MgO-B catalysts before and after the
reduction.
Figure 7 shows that the surface basicity was lower on
MgO-B than on MgO. Thus, the treatment with n-butanol
decreased the surface basicity of MgO. It is currently dif-
ficult to explain this phenomenon. Figure 7 shows that the
reduced Ni/MgO exhibited even higher surface basicity
than MgO. Since the samples were precipitated by sodium
carbonate, carbonate species were present in the samples
before the reduction and evacuation at 673 K, which were
the conditions used before the microcalorimetric adsorp-
tion of CO2. The surface basicity of NiO-MgO solid
solution should be weaker than that of MgO. Therefore,
carbonate species in NiO-MgO must be easier to remove
than in MgO, probably resulting in the cleaner MgO sur-
face in the Ni/MgO after reduction than in the MgO
sample.
3.5 Microcalorimetric Adsorption of Acetonitrile
Figure 9 shows the microcalorimetric results for CH3CN
adsorption on MgO and MgO supported nickel catalysts.
The adsorption of acetonitrile is complex. It adsorbs on
metallic Ni surfaces through the electron donation from the
nitrogen atom to the d orbital of a nickel atom [50, 51]. It
also adsorbs on acid sites since it is a base or an electron
donor owing to the lone electron pair on nitrogen atom. It
even adsorbs on basic sites since the hydrogen atoms in the
methyl group adjacent to the –CN group might exhibit
some characters of protons [52–55].
Figure 9 shows that the initial heat of acetonitrile
adsorption on the MgO-B support was about 70 kJ/mol,
along with the saturation coverage of about 650 lmol/g.
The initial heats of acetonitrile adsorption on the unreduced
Ni/MgO and Ni/MgO-B were about 90–95 kJ/mol, higher
than that on the MgO-B. This might be due to that the
Microcalorimetric results for NH3 adsorption at 423 K
on the Ni/MgO and Ni/MgO-B catalysts are displayed in
Fig. 8. The initial heats of NH3 adsorption on unreduced
160
150
Ni/MgO
Ni/MgO(R)
MgO
Ni/MgO
Ni/MgO(R)
Ni/MgO-B
120
120
Ni/MgO-B
Ni/MgO-B(R)
Ni/MgO-B(R)
MgO-B
90
80
40
0
60
30
0
0
150
300
450
600
750
0
100
200
300
400
500
CO2 coverage (µmol/g)
NH3 coverage (µmol/g)
Fig. 7 Differential heat versus adsorbate coverage for the adsorption
of CO2 at 423 K on the samples. MgO, MgO-B, Ni/MgO-B and
Ni/MgO were evacuated at 673 K while Ni/MgO-B (R) and Ni/MgO
(R) were reduced in H2 and then evacuated at 673 K
Fig. 8 Differential heat versus adsorbate coverage for the adsorption
of NH3 at 423 K on the Ni/MgO and Ni/MgO-B catalysts. Ni/MgO-B
and Ni/MgO were evacuated at 673 K while Ni/MgO-B (R) and
Ni/MgO (R) were reduced in H2 and then evacuated at 673 K
123

Surface Properties of Ni/MgO Catalysts
253
150
40
30
20
10
0
Ni/MgO at 393 K
Ni/MgO-B at 393 K
Ni/MgO at 413 K
Ni/MgO-B at 413 K
Ni/MgO
Ni/MgO(R)
Ni/MgO-B
Ni/MgO-B(R)
MgO-B
120
90
60
30
0
200
400
600
0
800
20
40
60
80
CH3CN coverage (µmol/g)
-1
WHSV (h )
Fig. 9 Differential heat versus adsorbate coverage for the adsorption
of acetonitrile at 308 K on the Ni/MgO and Ni/MgO-B catalysts.
MgO-B, Ni/MgO-B and Ni/MgO were evacuated at 673 K while
Ni/MgO-B (R) and Ni/MgO (R) were reduced in H2 and then
evacuated at 673 K
Fig. 10 Hydrogenation of toluene over the Ni/MgO and Ni/MgO-B
catalysts at 393 and 413 K with different weight hourly space velocity
(WHSV). The catalysts were reduced in flowing H2 at 673 K for 2 h
surface acidity was higher for the NiO-MgO solid solution
than for the MgO. After the reduction, the initial adsorption
heats increased to about 130 kJ/mol for the two catalysts,
significantly higher than those for the unreduced samples.
This clearly indicates the adsorption of acetonitrile on the
surface of metallic nickel, which might an important factor
for the activity of nickel for the hydrogenation of nitriles.
than Ni/MgAlO although the uptake of H2 was much higher
on Ni/MgAlO than on Ni/Al2O3. We ascribed this to the
inhibition effect of surface basicity on the hydrogenation of
toluene. Figure 7 shows that Ni/MgO displayed stronger
surface basicity than Ni/MgO-B. Thus, the more active sites
and less basic surface in Ni/MgO-B might account for its
significantly higher activity than Ni/MgO for the hydroge-
nation of toluene.
3.6 Catalytic Hydrogenation of Toluene
3.7 Catalytic Hydrogenation of Lauronitrile
The catalytic hydrogenation of toluene to methyl cyclo-
hexane (MCH) was performed in a fixed-bed micro-reac-
tor, which was used as a probe reaction to evaluate the
activity of the Ni/MgO and Ni/MgO-B catalysts. At the
reaction temperatures (393 and 413 K) employed in this
work, MCH was the unique product detected in the
hydrogenation reaction. Fig 10 shows the relationship
between the conversion of toluene and the weight hourly
space velocity (WHSV). Apparently, the conversion of
toluene was significantly higher on the Ni/MgO-B than on
the Ni/MgO at the same reaction temperature, consistent
with the results of H2 uptakes. Thus, the n-butanol treat-
ment procedure improved the dispersion and reducibility of
MgO supported nickel and therefore enhanced the catalytic
activity for the hydrogenation of toluene.
The liquid-phase hydrogenation of lauronitrile to dode-
cylamine was carried out in a batch micro-reactor. Table 4
summarized the results for the reaction over different cat-
alysts. The higher TAV implies the higher conversion of
lauronitrile. If the primary amine is the only product with
100% conversion of lauronitrile, TAV should be equal to
303 mg KOH/g. However, TAV would be decreased when
secondary and tertiary amines are formed. Thus, the higher
2/3AV means the lower content of primary amine in the
product mixture, i.e., the lower selectivity for the formation
of desired primary amine [5].
No tertiary amine was produced for the hydrogenation
of lauronitrile over all the catalysts studied in this work
since 3AV was found to be zero. In addition, 2/3AV was
low for the hydrogenation of lauronitrile over the Ni/MgO
and Ni/MgO-B, suggesting that the primary amine was the
main product (99.9%) over the two catalysts. However, the
conversion of lauronitrile was higher on the Ni/MgO-B
(94%) than on the Ni/MgO (87%), indicating the higher
activity of Ni/MgO-B than Ni/MgO for the hydrogenation
of lauronitrile, consistent with the results of H2 uptakes.
It should be mentioned that the difference of Ni dispersion
between the two catalysts was about 25%, but Ni/MgO-B
was 2–5 times more active than Ni/MgO, as can be seen from
Fig. 10. In a separate work, we compared the activity of
hydrogenation of toluene over Ni/MgO, Ni/MgAlO and
Ni/Al2O3 catalysts with different surface acidity and basi-
city. It was found that Ni/Al2O3 was significantly more active
123

254
H. Chen et al.
PA sel. (%)
Table 4 Hydrogenation of lauronitrile
Catalyst
TAV (mg KOH/g)
2/3AV (mg KOH/g)
3AV (mg KOH/g)
Conv. (%)
Ni/MgO
265
284
274
254
0.2
0.1
1.1
29
0
0
0
0
87
94
91
93
99.9
99.9
99.6
88.6
Ni/MgO-B
Raney Ni
Ni/SiO2-B
Note: TAV is the total amine value, 2/3AV is the sum of secondary amine and tertiary amine values and 3AV is the amine value of tertiary amine
only. PA denotes primary amine. The reaction was carried out in a batch reactor at 393 K and 2.0 MPa for 50 min with the initial feed ratio of
lauronitrile/catalyst/KOH/H2O = 600/10/1/10 (wt/wt). The supported nickel catalysts were reduced in H2 at 673 K
Although the microcalorimetric adsorption of CO2
showed the stronger surface basicity on Ni/MgO than on
Ni/MgO-B, both catalysts exhibited strong surface basicity
and thus high selectivity to primary amine for the hydro-
genation of lauronitrile. The microcalorimetric adsorption
showed the similar initial heats for the adsorption of CO on
Ni/MgO than on Ni/MgO-B, indicating that the n-butanol
treatment did not change the electron density of surface
nickel. In addition, the initial heats for the adsorption of
acetonitrile on Ni/MgO and Ni/MgO-B were also similar,
indicating the similar interaction of acetonitrile with nickel
surfaces in Ni/MgO and Ni/MgO-B. Thus, the higher
activity of Ni/MgO-B for the hydrogenation of lauronitrile
might be attributed to the more active sites presented on
Ni/MgO-B.
4 Conclusions
Following conclusions may be drawn from the current
work:
(1) During the preparation of Ni/MgO catalysts by the
co-precipitation method, a treatment with n-butanol before
the oven-drying significantly improved the dispersion and
reducibility of the supported nickel, resulting in a signifi-
cant increase in the active nickel surface areas. With this
technique, the average particle size of supported nickel
could be smaller than 7 nm even with the high loading
(60 wt%) of nickel on MgO. The Ni/MgO-B catalyst thus
prepared exhibited high catalytic activity for the hydroge-
nation of toluene to methyl cyclohexane.
(2) The Ni/MgO catalysts possessed weak surface
acidity, but quite strong surface basicity, especially for the
catalysts after the reduction. The treatment with n-butanol
before the oven-drying decreased the surface basicity. The
Ni/MgO catalyst without the n-butanol treatment exhibited
strong surface basicity after the reduction.
As compared to the industrial Raney nickel, the Ni/
MgO-B exhibited a slightly higher activity for the hydro-
genation of lauronitrile. In addition, both the Ni/MgO and
Ni/MgO-B exhibited a slightly higher selectivity towards
primary amine than the Raney nickel. Thus, the Ni/MgO-B
might be used instead of Raney nickel for the hydrogena-
tion nitriles to corresponding primary amines. The use of
Ni/MgO-B instead of Raney nickel for the hydrogenation
of nitriles may have a few advantages. Firstly, the prepa-
ration of Ni/MgO-B seems easier than Raney nickel since it
does not involve the processes (the formation of Ni–Al
alloy at high temperatures and the dissolution of Al from
the alloy) that may cause environmental problems. Sec-
ondly, the Ni/MgO-B may be recycled more easily than
Raney nickel since the used Ni/MgO-B can be easily dis-
solved in nitric acid and then precipitated.
(3) Microcalorimetric adsorption of H2 and CO showed
that the n-butanol treatment did not affect the electron
density of surface nickel, since the initial heats for the
adsorption of CO on the Ni/MgO and Ni/MgO-B catalysts
were similar. The initial heat for the adsorption of aceto-
nitrile was found to be about 130 kJ/mol on the supported
metallic nickel surfaces, while it was lower than 95 kJ/mol
on MgO and unreduced Ni/MgO catalysts, indicating the
strong interaction between acetonitrile and the supported
nickel surfaces.
(4) The surface basicity of the Ni/MgO catalysts favored
the selectivity to the primary amine in the hydrogenation of
lauronitrile to lauryl amine, and the formation of undesir-
able secondary and tertiary amines catalyzed by acid sites
was greatly inhibited. The Ni/MgO-B catalyst prepared
with the n-butanol treatment procedure exhibited the
excellent activity and selectivity to the primary amine for
the hydrogenation of lauronitrile to lauryl amine.
A 60%Ni/SiO2 catalyst with high H2 uptake (850 lmol/g)
was also prepared for comparison. Under the same reaction
conditions, it exhibited 93% conversion of lauronitrile with
88.6% selectivity to the primary amine. Although the same
amount of KOH inhibitor was added for the reaction
(to suppress the secondary and tertiary by-products), sub-
stantial amount of secondary amine (11.4%) was produced
over the Ni/SiO2, indicating the importance of surface
basicity of Ni/MgO-B for the hydrogenation of lauronitrile to
the primary amine.
Acknowledgments This work was supported by the NSFC
(20673055) and Jiangsu Province of China (BE2009145).
123

Surface Properties of Ni/MgO Catalysts
255
29. Shan H, Zhang Z (1997) J Euro Ceram Soc 17:713
30. Gao GM, Zou HF, Gan SC, Liu ZJ, An BC, Xu JJ, Li GH (2009)
Powder Tech 191:47
31. Narayanan S, Screekanth G (1989) J Chem Soc, Faraday Trans I
85:3785
32. Pannell RB, Chung KS, Bartholomew CH (1977) J Catal 46:340
33. Smith JS, Thrower PA, Vannice MA (1981) J Catal 68:270
34. Chen YG, Tomishige K, Yokoyama K, Fujimoto K (1999) J Catal
184:479
35. Furusawa T, Tsutsumi A (2005) Appl Catal A 278:207
36. Yoshida T, Tanaka T, Yoshida H, Funabiki T, Yoshida S (1996) J
Phys Chem 100:2302
References
1. Gomez S, Peters JA, Maschmeyer T (2002) Adv Synth Catal
344:1037
2. De Bellefon C, Fouilloux P (1994) Catal Rev-Sci Eng 36:459
3. Branco JB, Ballivet-Tkatchenko D, de Matos AP (2007) J Phys
Chem C 111:15084
4. McMillan ST, Agrawal PK (1988) Ind Eng Chem Res 27:243
¨
5. Ostgard DJ, Olindo R, Berweiler M, Roder S, Tacke T (2007)
Catal Today 121:106
6. Enthaler S, Junge K, Addis D, Erre G, Beller M (2008) Chem Sus
Chem 1:1006
7. Kukula P, Studer M, Blaser H (2004) Adv Synth Catal 346:1487
8. Barrault J, Pouilloux Y (1997) Catal Today 37:137
37. Serra M, Salagre P, Cesteros Y, Medina F, Sueiras JE (2000)
Solid State Ionics 134:229
38. Spiewak BE, Shen J, Dumesic JA (1994) J Phys Chem 99:17640
39. Weatherbee GD, Bartholomew CH (1984) J Catal 87:55
40. Prinsloo JJ, Gravelle PC (1980) J Chem Soc Faraday Trans I
76:221
¨
9. Chojecki A, Veprek-Heijman M, Mu¨ller TE, Scharringer P,
Veprek S, Lercher JA (2007) J Catal 245:237
10. Volf J, Pasek J (1986) Stud Surf Sci Catal 27:105
11. Baumeister P, Studer M, Roessler F (1997) Handb Heterog Catal
5:2186
12. Greenfield H (1967) Ind Eng Chem, Prod Res Dev 6:142
13. Xie S, Li H, Li H, Deng J-F (1999) Appl Catal A 189:45
14. Cordier G, Popa JM (1999) US Patent 6,005,145
15. Ansmann A, Benisch C (2003) US Patent 2003/0120115
16. Borninkhof F, Geus JM, Verhaak MJFM (1993) EP Patent
0566197
´
´
41. Cerro-Alarcon M, Bachiller-Baeza B, Guerrero-Ruiz A, Rodrı-
guez-Ramos I (2006) J Mol Catal A 258:221
42. Poncelet G, Centeno MA, Molina R (2005) Appl Catal A 288:232
¨
43. D’Ercole A, Giamello E, Pisani C, Ojamae L (1999) J Phys Chem
B 103:3872
44. Ito T, Kuramoto M, Yoshioka M, Tokuda T (1983) J Phys Chem
87:4411
45. Claus P, Berndt H, Mohr C, Radnik J, Shin E-J, Keane MA
(2000) J Catal 192:88
17. Reesink BH, van Gasteren N (2003) US Patent 6,524,994
18. Merger F, Priester CU, Witzel T, Koppenhoefer G, Schuster L
(1992) US Patent 5,166,443
46. Xue M, Ge J, Zhang H, Shen J (2007) Appl Catal A 330:117
47. Verhaak MJFM, van Dillen AJ, Geus JW (1993) Appl Catal A
105:251
19. Deckers G, Frohning D (1992) EP Patent 0547505
20. Oudejans JC (1989) EP Patent 0340848
´ ´
48. Borgna A, Frety R, Primet M, Guenin M (1991) Appl Catal 76:
21. Verhaak MJFM, van Dillen AJ, Geus JW (1994) Catal Lett 26:37
22. Medina Cabello F, Tichit D, Coq B, Vaccari A, Dung NT (1997)
J Catal 167:142
˘ ˘
23. Gluhoi AC, Marginean P, Stanescu U (2005) Appl Catal A
294:208
24. Huang Y, Sachtler WMH (1999) Appl Catal A 182:365
25. Rode CV, Arai M, Shirai M, Nishiyama Y (1997) Appl Catal A
148:405
233
49. Chattopadhyay A, Yang H, Whitten JL (1990) J Phys Chem 94:
6379
50. Bigot B, Delbecq F, Peuch VH (1995) Langmuir 11:3828
51. Yang H, Caves TC, Whitten JL (1995) J Chem Phys 103:8756
52. Lavalley JC (1996) Catal Today 27:377
53. Platero EE, Mentruit MP, Morterra C (1999) Langmuir 15:5079
54. Koubowetz F, Latzel J, Noller H (1980) J Colloid Interface Sci
74:322
26. Bell AT (2003) Science 299:1688
27. Prakash SS, Brinker CJ, Hurd AJ, Rao SM (1995) Nature 374:439
28. Hu ZS, Dong JX, Chen GX (1998) J Colloid Interface Sci
208:367
55. Prinetto F, Manzoli M, Ghiotti G, MdJM Ortiz, Tichit D, Coq B
(2004) J Catal 222:238
123