In-situ generated highly dispersed nickel nanoclusters confined in MgAl mixed metal oxide platelets for benzoic acid hydrogenation

Article abstract of DOI:10.1016/j.jcat.2019.03.012

A new and cost-effective NiMgAl mixed metal oxide (Ni₂Mg_0.5Al₁-MMO) catalyst derived from hierarchical flower-like Ni-Mg-Al layered double hydroxides (NiMgAl-LDHs) was fabricated by a hydrothermal-calcination-reduction method. This catalyst showed excellent catalytic performance in the selective hydrogenation of benzoic acid to cyclohexanecarboxylic acid. Notably, recycling experiments demonstrated that this catalyst could be used at least ten times without significant losses in activity and selectivity under harsh reaction conditions; thus, it presents similar behavior to most of noble metal catalysts. A series of characterizations were performed to investigate the relationship between the structure and the catalytic performance of this catalyst and elucidate the mechanism of its good stability. The results demonstrated that the Ni₂Mg_0.5Al₁-MMO catalyst exhibited highly dispersed nickel species due to the well-defined flower-like structure of NiMgAl-MMO platelets as well as the confined effect of Mg and Al oxide species.

Full text of DOI:10.1016/j.jcat.2019.03.012

Journal of Catalysis 372 (2019) 258–265
Contents lists available at ScienceDirect
Journal of Catalysis
journal homepage: www.elsevier.com/locate/jcat
In-situ generated highly dispersed nickel nanoclusters confined in MgAl
mixed metal oxide platelets for benzoic acid hydrogenation
⇑
⇑
Huiling Zhang, Jie Dong, Xianliang Qiao, Jingru Qin, Haofei Sun, Aiqing Wang, Libo Niu , Guoyi Bai
Key Laboratory of Chemical Biology of Hebei Province, College of Chemistry and Environmental Science, Hebei University, Baoding 071002, PR China
a r t i c l e i n f o
a b s t r a c t
Article history:
A new and cost-effective NiMgAl mixed metal oxide (Ni2Mg0.5Al1-MMO) catalyst derived from hierarchi-
cal flower-like Ni-Mg-Al layered double hydroxides (NiMgAl-LDHs) was fabricated by a hydrothermal-cal
cination-reduction method. This catalyst showed excellent catalytic performance in the selective hydro-
genation of benzoic acid to cyclohexanecarboxylic acid. Notably, recycling experiments demonstrated
that this catalyst could be used at least ten times without significant losses in activity and selectivity
under harsh reaction conditions; thus, it presents similar behavior to most of noble metal catalysts. A ser-
ies of characterizations were performed to investigate the relationship between the structure and the cat-
alytic performance of this catalyst and elucidate the mechanism of its good stability. The results
demonstrated that the Ni2Mg0.5Al1-MMO catalyst exhibited highly dispersed nickel species due to the
well-defined flower-like structure of NiMgAl-MMO platelets as well as the confined effect of Mg and
Al oxide species.
Received 29 November 2018
Revised 5 March 2019
Accepted 9 March 2019
Keywords:
Benzoic acid
Hydrogenation
Layered double hydroxides
Stability
Harsh reaction conditions
Ó 2019 Elsevier Inc. All rights reserved.
1. Introduction
activity and stability is of vital importance and remains a challenge
in BA hydrogenation.
Nowadays, the hydrogenation of aromatic substrates to obtain
saturated or partially saturated compounds is of great importance
in the field of chemical and pharmaceutical industry, due to its
readily obtained feedstocks, environmentally friendly and straight-
forward routes [1,2]. For instance, selective hydrogenation of
benzoic acid (BA) is of practical significance to produce high
value-added cyclohexanecarboxylic acid (CCA), which is an impor-
tant organic synthesis intermediates for the production of prazi-
quantel, ansatrienin and hexanolactam, etc [3,4]. However,
production of CCA by BA hydrogenation endures some disadvan-
tages in economic and technical fields. For example, aromatic ring
hydrogenation needs to overcome higher resonance energy during
synthesis, so hydrogenation of BA generally requires more active
catalysts and relatively harsh reaction conditions [5,6]. Most of
the active catalysts used in this transformation are noble metal cat-
alysts, such as Ir/r-Al2O3, Pd/AC, RuPd/CN, etc. [7–10]. In spite of
their advantages, the high price of noble catalysts has limited their
industrial applications. In addition, both noble metal and non-
noble metal catalysts are prone to agglomeration and sintering
under harsh reaction conditions [11,12]. Hence, how to develop
cost-effective metal catalysts with simultaneously enhanced
In recent years, nickel-based catalysts have attracted much
interest in hydrogenation reactions due to their unique isotropic
nature, high concentration of coordinative unsaturated sites, and
low cost [13–15]. However, due to the aggregation and sintering
of nanoparticles during the preparation and reaction process,
nickel-based catalysts, especially the unsupported catalysts, were
unstable. How to improve nickel-based catalysts to be as active
and stable as noble metal catalysts is a challenging issue in hetero-
geneous catalysis. Great efforts have been made to this issue, such
as deposing active Ni on certain supports, coating mesoporous
shell over active sites and in-situ generating active site from pre-
cursor material [16–18]. For example, Zhao et al. demonstrated
that the catalytic performance of Ni/TiO2 in aqueous phase
nitrobenzene hydrogenation could be significantly improved after
a carbon coating [18]. Nevertheless, the practical application of
nickel-based catalysts is still limited due to their poor thermal sta-
bility under harsh reaction conditions [19]. Recently, we have pre-
pared a Ni/mSiO2-AE catalyst by an ammonia evaporation method,
which can achieve better stability (4 cycles) in BA hydrogenation
[20]. However, its stability under harsh reaction condition was still
unsatisfactory, although it was better than the NiB amorphous
alloy catalysts we used in this transformation [21,22].
Layered double hydroxides (LDHs) are a typical class of highly
ordered two-dimensional anionic clays, which exhibit the lamellar
form based on M2+(OH)6 octahedral unites sharing edges to form M
⇑
Corresponding authors.
E-mail addresses: libo_niu@126.com (L. Niu), baiguoyi@hotmail.com (G. Bai).
https://doi.org/10.1016/j.jcat.2019.03.012
0021-9517/Ó 2019 Elsevier Inc. All rights reserved.

H. Zhang et al. / Journal of Catalysis 372 (2019) 258–265
259
(OH)2 brucite-like layers. In these materials, trivalent cations could
replace some of divalent ions, and exchangeable anions are situ-
ated in the interlayer space to balance the charge [23–26]. Due
to the unique characteristic, such as surface adsorption capacity,
ion-exchange ability, surface tunable basicity and the memory
effect, LDHs have attracted great attention in the field of
adsorption, catalysis, drug delivery and supramolecular chemistry
[27–30]. After calcination at specified temperature, LDHs could be
converted to mixed metal oxides (MMO) with large surface area,
high metal dispersion, tunable acid/basic properties and good ther-
mal stability. Based on these properties, MMO can be widely used
as promising heterogeneous catalysts or supports [30,31]. This
inspired us to fabricate a nickel-based MMO catalyst derived from
NiMgAl-LDHs for the selective hydrogenation of BA. The resulting
nickel-based catalyst may possess following advantages compared
to the traditional supported catalysts: (i) Ni nanparticles can be in-
situ formed in MMO matrix, the confined microenvironment can
disperse and anchor metal particles uniformly; (ii) thermal and
mechanical stability of the MMO structure can improve the stabil-
ity of the catalyst during reaction; (iii) tunable preparation method
promotes the metal-support interaction, which could facilitate the
stability of the catalyst.
As continuation of our previous work [20–22], NiMgAl-MMO
catalysts were fabricated from hierarchical flower-like NiMgAl-
LDHs precursor and applied in the selective hydrogenation of BA
to CCA under harsh reaction conditions. These in-situ generated
catalysts exhibited better activity compared to the catalyst pre-
pared by an impregnation method. Calcination process was found
to play a significant role on the stability of NiMgAl-MMO catalysts.
A Ni2Mg0.5Al1-MMO catalyst calcined under 500 °C for 8 h showed
the highest activity and stability with negligible loss of its activity
during 10 cycles. A series of characterizations were employed to
investigate the structure-activity relationship of the prepared cat-
alysts. To the best of our knowledge, there has been no report on
nickel-based mixed metal oxides with hierarchical flower-like
structure, which showed excellent stability in BA hydrogenation
under harsh reaction conditions.
MgAl-MMO support was added to an Ni(NO3)2Á6H2O aqueous
solution under ultrasound for 30 min, using the same Ni content
as that in NiMgAl-MMO. The resulting paste was dried at 100 °C
for 12 h and then calcined under a certain condition to obtain a
supported metal oxide. Furthermore, Ni/Al2O3, Ni/SiO2 and Ni/
MgO were prepared by the same method as described above using
different supports.
2.3. Catalyst characterization
The actual Ni and Mg content in the catalysts were evaluated
by ICP-OES (Varian Vista-MPX) on an Agilent 7700 Â spectrome-
ter. The textural properties of the samples were measured on a
Micromeritics Tristar II 3020 surface area and pore analyzer. All
the samples were dried at 200 °C for 4 h in a vacuum oven before
its measurement. The specific surface areas and average pore
diameters were determined using the Brunauer-Emmett-Teller
(BET) and the Barett–Joyner-Halenda (BJH) methods, respectively.
Pore size distributions were calculated according to the desorp-
tion regimes of the isotherms. H2-chemisorption was performed
on a Micromeritics Autochem II 2920 instrument. Before the test,
all the samples were reduced by a 99.99% H2 flow (50 mLÁminÀ1
)
at 450 °C for 3 h. Then chemisorption was performed by pulsing
of a mixture of 10% H2/Ar (50 mLÁminÀ1). The nickel surface areas
and metal dispersion were calculated by setting the stoichiomet-
ric ratio Hadsorbed/Nisurface = 1 and the density of active sites on the
surface to 1.54 Â 1019 atomsÁm2. X-ray diffraction (XRD) patterns
were obtained on a Bruker D8 Advance x-ray diffractometer using
Cu Ka radiation at 40 kV and 40 mA with a step size of 0.02° over
a 5–80° range at a scanning rate of 0.01 s per step. Scanning elec-
tron micrographs (SEM) were obtained on a JEOL JSM-7500
instrument. Transmission electron microscopy (TEM), scanning
transmission electron microscopy (STEM), and elemental mapping
of the catalysts were carried out on a FEI Tecnai G2 F20 S-TWIN
microscope that operates at an accelerating voltage of 200 kV. The
samples were prepared by dispersing the catalyst powder in etha-
nol under ultrasound and then supported on a carbon film of cop-
per grid. X-ray photoelectron spectroscopy (XPS) analysis was
2. Experimental
carried out on a PHI 1600 spectrometer using Mg Ka X-ray source
for excitation. The binding energy was calibrated to the C1s signal
(284.8 eV) as reference. Temperature-programmed reduction of
hydrogen (H2-TPR) test was also carried out on the Micromeritics
Autochem II 2920 instrument. The sample (100 mg) was pre-
treated at 200 °C for 1 h in Ar flow (50 mLÁminÀ1) to remove
the adsorbed water and impurities. After cooling to 50 °C, the cat-
alyst was heated to 900 °C at a rate of 10 oCÁminÀ1 under a 10%
H2-Ar (50 mLÁminÀ1) gas mixture; the TCD signal continuously
recorded.
2.1. Materials
Unless otherwise noted, all chemicals were of analytical reagent
grade and were purchased from Baoding Huaxin Reagent and
Apparatus Co., Ltd. Ultrapure Ar (99.999%) and 10 vol% H2/Ar were
purchased from Baoding Zhuoda Gas Co., Ltd.
2.2. Preparation of nickel-based catalysts
In a typical run for the synthesis of flower-like NiMgAl-LDHs
precursor (Fig. 1), Ni(NO3)2Á6H2O, Mg(NO3)2Á6H2O, Al(NO3)3Á
9H2O and urea with a molar ratio 2:0.5:1:16 were dissolved
in a mixed solvent of ethylene glycol and water with a volume
ratio 9:1. The above mixture was stirred for 30 min and then
transferred into a 100 mL Teflon-lined autoclave, sealed and
heated at 160 °C for 6 h. The obtained powder was filtered,
washed with deionized water and ethanol several times and
dried in an oven at 100 °C for 12 h to afford the NiMgAl-LDHs
precursors. Finally, these precursors were calcined in static air
at 500 °C for certain time to obtain NiMgAl-MMO-x (x denotes
as calcination time).
2.4. Hydrogenation reaction
Before the activity test, all the catalysts were reduced under a
99.99% H2 flow (50 mLÁminÀ1) at 450 °C for 3 h in a microreactor
and then used for the reaction. The hydrogenation reactions were
carried out in a 100 mL stainless autoclave reactor equipped with
a mechanical stirrer and an electric heating system. In a typical
procedure, 0.3 g catalyst was dispersed in 60 mL cyclohexane and
then 1.0 g BA was added into the solution. The reactor was then
sealed, purged with H2 for three times, and then pressurized to
5.0 MPa. The reaction was conducted at 150 °C with a stirring
speed of 400 rpm for 3 h. The reaction mixture was analyzed by
Agilent 7820 gas chromatography (GC) and the product structures
were identified by gas chromatography-mass spectrometry (GC-
MS) using an Agilent 5975C spectrometer.
For comparison, Ni/MgAl-IMP was prepared by an incipient
impregnation method. Firstly, MgAl-MMO was prepared using
the same method as NiMgAl-MMO-x without the addition of Ni
(NO3)2Á6H2O and used as support for Ni/MgAl-IMP. Then, the

260
H. Zhang et al. / Journal of Catalysis 372 (2019) 258–265
Fig. 1. Schematic illustrating the preparation of NiMgAl-MMO.
order to further improve its stability, Ni2Mg0.5Al1-MMO was pre-
3. Results and discussion
pared at 500 °C with different calcination time and then tested in
BA hydrogenation. As expected, the cycle times of
Ni2Mg0.5Al1-MMO-x increased with the increase of calcination
time and the sample calcined at 500 °C for 8 h showed the best sta-
bility (Fig. S1). The conversion of BA is still 91.20% in the tenth
cycle (Fig. 2a), comparable to the fresh catalyst. However, increas-
ing the calcination time to 10 h caused the decrease of stability. We
speculated the agglomeration of MMO microplates and coverage of
active sites may occur when the calcination time was over 8 h,
which diminished the active sites in the catalyst. These results
indicate that the stability of Ni2Mg0.5Al1-MMO could be signifi-
cantly improved by adjusting the calcination conditions. To the
best of our knowledge, this is the best stability result in BA hydro-
genation over nickel-based catalysts that has been reported in the
open literature. Its performance is comparable to most of the noble
metal catalysts (Table S3) and the outstanding nickel-based cata-
lysts in other heterogeneous catalytic reactions [32–34].
Moreover, the stability of Ni/Mg0.5Al1-IMP, which was prepared
under the same condition and composition, was also tested in this
transformation. As shown in Fig. 2b, Ni/Mg0.5Al1-IMP showed not
only lower initial activity compared to Ni2Mg0.5Al1-MMO-8, but
also faster deactivation rate. The activity of Ni/Mg0.5Al1-IMP
decreased rapidly during recycling and the conversion of BA was
reduced to 42.50% only after 5 cycles. Notably, as shown in
Table S4, the Ni loading of Ni2Mg0.5Al1-MMO-8 only dropped from
29.6% to 26.0% after 10 cycles (0.36% per run) whereas the Ni load-
ing of Ni/Mg0.5Al1-IMP dropped sharply from 26.6% to 14.1% within
5 cycles (2.5% per run). The loss of Ni species may be the main rea-
son for the poor stability of Ni/Mg0.5Al1-IMP, which probably due
to the different interaction strength between nickel and MgAl-
MMO supports arising from different preparation methods. The
above results demonstrated that the in-situ generated
Ni2Mg0.5Al1-MMO, especially Ni2Mg0.5Al1-MMO-8, showed
3.1. Catalyst selection
Firstly, the catalytic performances of the prepared nickel-based
catalysts were tested in the hydrogenation of BA. It was found that
the conversion of BA was 56.26%, 81.46% and 68.38% for Ni/Al2O3,
Ni/SiO2 and Ni/MgO, respectively (Table S1). The reaction results of
these catalysts using traditional supports were lower than our
expectation. Considering the benefit of the MMO catalysts gener-
ated from LDHs, a series of NiMgAl-MMO catalysts were then pre-
pared and evaluated. As shown in Table1, the molar ratios of Ni and
Mg in these catalysts are similar to their theoretical values. As
expected, all NiMgAl-MMO catalysts exhibited higher activity than
the ordinary supported catalysts. It was found that the
Ni2Mg0.5Al1-MMO showed the best hydrogenation activity among
the three in-situ generated NiMgAl-MMO catalysts, with a conver-
sion of BA and yield of CCA of approximately 98.39% and 96.90%,
respectively. Ni2Mg0.5Al1-MMO also has the highest turnover
frequency (TOF) value in all the in-situ generated catalysts consid-
ered in this study. In addition, Ni/Mg0.5Al1-IMP, prepared by an
impregnation method, showed lower activity in BA hydrogenation,
compared to all in-situ generated catalysts.
Considering the effect of calcination conditions on the structure
and properties of LDHs derived MMO materials, different calcina-
tion temperature (400 °C, 500 °C and 600 °C) with the same calci-
nation time (3 h) was evaluated in the preparation process of
Ni2Mg0.5Al1-MMO and the obtained catalysts were applied in BA
hydrogenation (Table S2). Notably, although the original conver-
sions of BA were almost unchanged over these three catalysts,
there are obvious differences in their stability. Ni2Mg0.5Al1-MMO
obtained at 500 °C can be used for 4 cycles without distinct loss
of activity. In contrast, activity of the catalysts calcined at 400 °C
and 600 °C was decreased obviously during their second run. In
Table 1
Catalytic performance of different nickel-based catalysts.
Catalysta
Molar ratiob
(Ni:Mg)
Ni contentb
(wt%)
BA conversion (%)
CCA selectivity (%)
CCA yield (%)
TOFc
(hÀ1
)
Ni1.5Mg0.5Al1-MMO
Ni2Mg1Al1-MMO
Ni2Mg0.5Al1-MMO
Ni/Mg0.5Al1-IMP
0.63:0.23
0.83:0.40
0.88:0.23
0.73:0.18
22.6
29.2
30.1
31.1
95.31
96.48
98.39
78.55
99.01
98.64
98.49
97.53
94.37
95.17
96.90
76.60
23.27
21.93
25.60
19.62
Reaction conditions:1.0 g BA, 0.3 g catalyst, 60 mL cyclohexane, 150 °C, 5.0 MPa H2, 400 rpm, 3 h.
a
All catalysts were calcined under 500 °C for 3 h before reduction.
Based on ICP results.
The TOF values were calculated based on the conversion of BA in 1 h, according to the Ref. [2].
b
c

H. Zhang et al. / Journal of Catalysis 372 (2019) 258–265
261
Fig. 2. Stability of the catalysts: (a) Ni2Mg0.5Al1-MMO-8 and (b) Ni/Mg0.5Al1-IMP.
excellent activity and stability in BA hydrogenation. Thus, it was
selected as the catalyst of choice for further studies.
diameter around 20 nm, further demonstrating the formation of
mesopores in these materials. The hierarchical architecture in
these catalysts can promote the mass transfer process in catalytic
reactions. In addition, Ni/Mg0.5Al1-IMP presents a mixture hystere-
sis loop of H2 and H3 type with an abrupt closure at a relative pres-
sure of 0.6, indicating the presence of slit-shaped and ink-bottle
pore in this sample. Moreover, Ni/Mg0.5Al1-IMP also exhibits a nar-
row pore size distribution around 10 nm (Fig. 3d). Such textural
structures would not facilitate mass transportation and diffusion
during reaction compared to the in-situ generated samples.
XRD patterns of Ni2Mg0.5Al1-LDHs, calcined Ni2Mg0.5Al1-MMO-8,
reduced Ni2Mg0.5Al1-MMO-8 and reduced Ni/Mg0.5Al1-MMO-IMP
are shown in Fig. 4. As can be seen, Ni2Mg0.5Al1-LDHs exhibits
the characteristic diffraction peaks at about 11.30°, 23.7°, 34.6°
and 60.5° correspond to (0 0 3), (0 0 6), (0 0 9) and (1 1 0) reflec-
tions [36], respectively. The occurrence of these characteristic
reflections confirms the successful synthesis of LDHs. Meanwhile,
after the Ni2Mg0.5Al1-LDHs was calcined at 500 °C for 8 h, the char-
acteristic peaks of LDHs cannot be detected and NiO is observed in
the obtained Ni2Mg0.5Al1-MMO-8, as indicated by its characteristic
diffraction peaks at 2h value of 37.2°, 43.3°, 62.9° and 75.4° [37].
Furthermore, three obvious diffraction peaks at 2h of about 44.5°,
51.8° and 76.4°, which is characteristic fcc nickel phase (JCPDS
87-0712) [38,39], were obtained in the reduced sample, together
with the disappearance of the NiO peaks, demonstrating the full
reduction of NiO in the reduced Ni2Mg0.5Al1-MMO-8. In addition,
the reduced Ni/Mg0.5Al1-IMP sample also showed the characteris-
tic peaks of Ni, but the diffraction peaks are much stronger com-
pared to the reduced Ni2Mg0.5Al1-MMO-8 sample, indicating the
relatively larger size of Ni particles in this sample. Moreover, the
characteristic peaks of NiAl2O4 were also detected in Ni2Mg0.5Al1-
MMO-8 and Ni/Mg0.5Al1-IMP, demonstrating the formation of spi-
nel structure in these samples [40,41].
3.2. Catalyst characterization
The physicochemical properties of different nickel-based MMO
catalysts calcined with different times are summarized in Table 2.
As can be seen, Ni2Mg0.5Al1-MMO-8 has the largest specific surface
area, pore size, H2-chemisorption value, metal dispersion and
metallic surface area among all the in-situ generated
Ni2Mg0.5Al1-MMO catalysts. The outstanding physicochemical
properties of Ni2Mg0.5Al1-MMO-8 may be attributed to overflow
of CO2 released from decomposition of intercalated CO23À with
longer calcination time, resulting in more effective exposure of
active Ni sites. The reason for the obvious decrease of these prop-
erties in Ni2Mg0.5Al1-MMO-10 can be attributed to the collapse of
the flower-like morphology due to the prolonged calcination
time. This change caused agglomeration of nanoplatelets, which
results in a decrease of the surface area and coverage of the active
sites, thus affecting its stability performance. In addition,
Ni/Mg0.5Al1-IMP has a lower surface area, H2-chemisorption and
metal dispersion when compared to the in-situ Ni2Mg0.5Al1-MMO-x
catalysts. These combined effects produce a low hydrogenation activ-
ity, which is in good agreement with the hydrogenation results.
The nitrogen adsorption-desorption isotherms and pore size
distributions
of
Ni2Mg0.5Al1-MMO-4,
Ni2Mg0.5Al1-MMO-6,
Ni2Mg0.5Al1-MMO-8 and Ni/Mg0.5Al1-IMP are shown in Fig. 3. It
was found that all the samples exhibit mixed isotherms of type II
(without plateau at high P/P0) and type IV (low adsorption at low
P/P0) isotherms with a large hysteresis loops in the relative pres-
sure region of 0.40–1.0, which indicates the capillary condensation
takes place in mesopores [35]. H3 hysteresis loops were observed
in the in-situ generated samples, which suggested the presence of
the slit-shaped pore formed by aggregated plate-like particles.
The pore size distribution derived from the desorption branch indi-
cates that all Ni2Mg0.5Al1-MMO-x samples have average pore
Morphology of Ni2Mg0.5Al1-MMO-8 and Ni/Mg0.5Al1-IMP was
characterized by SEM and TEM. As shown in Fig. 5a, the as-
synthesized Ni2Mg0.5Al1-LDH presents a flower-like morphology,
Table 2
Physicochemical properties of different catalysts.
Surface area (m2 ÁgÀ1
)
Pore size (nm)
H2-chemisorption (cm3ÁgÀ1
)
Metal surface area (m2Ág
)
Metal dispersion (%)
À1
Catalyst
cat
Ni2Mg0.5Al1-MMO-4
Ni2Mg0.5Al1-MMO-6
Ni2Mg0.5Al1-MMO-8
Ni2Mg0.5Al1-MMO-10
Ni/Mg0.5Al1O-IMP
191
204
211
161
140
16.31
15.65
16.85
7.50
0.32
2.09
3.59
2.42
0.23
3.73
7.29
12.52
8.44
0.81
0.56
3.69
6.35
4.28
0.41
6.01

262
H. Zhang et al. / Journal of Catalysis 372 (2019) 258–265
Fig. 3. N2 adsorption-desorption isotherm and pore size distributions (inset) of (a) Ni2Mg0.5Al1-MMO-4, (b) Ni2Mg0.5Al1-MMO-6, (c) Ni2Mg0.5Al1-MMO-8 and (d) Ni/Mg0.5Al1-
IMP.
number of approximately spherical Ni nanoparticles are well dis-
persed on the support, which is in good agreement with the ele-
mental mapping observation of relatively homogeneous
distributions of Ni, Mg and Al atoms in this sample (Fig. S2). The
HRTEM image (Fig. 5f) of Ni2Mg0.5Al1-MMO-8 further demon-
strates that the Ni particles are immobilized on the support with
a high dispersion and narrow particle size distribution (ꢀ1 nm).
In contrast, Ni/Mg0.5Al1-IMP did not present the flower-like mor-
phology and the average diameter size of Ni particle was 9 nm with
nonuniform Ni dispersion (Fig. S3). It is believed that the uniformly
distributed cations in the brucite-like layer of LDHs precursor is
benefit for the Ni dispersion. The obtained MgO and amorphous
Al2O3 phases act as supports and dispersing agents between Ni
nanoclusters to expose more Ni active sites. The hierarchical struc-
ture and uniform Ni dispersion in Ni2Mg0.5Al1-MMO-8 is benefit
for the diffusion and hydrogenation of reactant, explaining its
highest activity. The lattice fringes can be clearly observed with a
crystal plane spacing of 0.201 nm and 0.202 nm in Ni2Mg0.5Al1-
MMO-8 and Ni/Mg0.5Al1-IMP, which is consistent with the 111
Fig. 4. XRD patterns of (a) Ni2Mg0.5Al1-LDHs, (b) calcined Ni2Mg0.5Al1-MMO-8, (c)
reduced Ni/Mg0.5Al1O-IMP and (d) reduced Ni2Mg0.5Al1-MMO-8.
crystal plane in cubic (fcc) Ni (JCPDS 87–0712, inset of Figs. 5f
and S3b) [15]. In addition, the flower-like structure was not dis-
tinct in Ni2Mg0.5Al1-MMO-10, indicating that longer calcination
time may lead to the collapse of the flower-like morphology
(Fig. S4); the latter caused the decrease of surface area and the cov-
erage of the active sites (Table 2).
In order to explore the surface chemical states of Ni and inter-
action between Ni and support, the reduced nickel-based catalysts
were subjected to XPS analysis (Fig. 6). As can be seen, the Ni2p3/2
spectra present two states of nickel. Generally, the binding ener-
gies of Ni2p3/2 of metallic Ni and NiO are around 852 and 855 eV,
which is composed of numerous frizzy nanoplatelets intercrossing
with each other. In addition, the flower-like structure was kept in
Ni2Mg0.5Al1-MMO-8 after the LDH precursor was calcined at 500 °C
for 8 h (Fig. 5b), which is also confirmed by the TEM image (Fig. 5c).
A closer examination on the detailed morphology of Ni2Mg0.5Al1-
MMO-8 indicates that the hierarchical structure is composed of a
large number of microplates connecting with each other to form
a flower-like micro/nanostructure (Fig. 5d). The SEM and TEM
results revealed that the MMO inherits the original flower-like
morphology of the LDHs precursor. As shown in Fig. 5e, a large

H. Zhang et al. / Journal of Catalysis 372 (2019) 258–265
263
Fig. 5. SEM images of Ni2Mg0.5Al1-LDH (a) and Ni2Mg0.5Al1-MMO-8 (b); TEM (c, d, e) and HRTEM (f) images of Ni2Mg0.5Al1-MMO-8.
respectively [42,43]. Based on this phenomenon, the peaks at about
852.26–852.94 eV can be assigned to the metallic Ni [44], and the
peaks at 855.34–855.84 eV as well as the broad satellites centered
at 861.31–861.60 eV demonstrate the presence of NiO in these
samples [45]. The formation of NiO in the tested samples may be
caused by the oxidation of metallic Ni during the sample transfer
and settlement before the XPS analysis [42,46]. Notably, it could
be concluded that the binding energies of metallic Ni and NiO in
Ni2Mg0.5Al1-MMO-x samples are higher than those in Ni/Mg0.5Al1-
IMP, indicating a stronger interaction between nickel and supports
in the in-situ generated samples [46,47]. In addition, Ni2Mg0.5Al1-
MMO-8 has the highest binding energies of metallic Ni
(852.94 eV) and NiO (855.84 eV) among Ni2Mg0.5Al1-MMO-x sam-
ples, indicating the 8 h calcination time is suitable and benefit to
generate a stronger interaction between nickel and supports,
which might be the main reason for its better stability under harsh
reaction conditions compared to other catalysts.
TPR profiles of Ni/Mg0.5Al1-IMP and Ni2Mg0.5Al1-MMO-x are
shown in Fig. 7. According to previous studies on nickel-based
Fig. 6. Ni 2p3/2 XPS spectra of the reduced catalysts: (a) Ni/Mg0.5Al1-IMP, (b)
Ni2Mg0.5Al1-MMO-4, (c) Ni2Mg0.5Al1-MMO-8 and (d) Ni2Mg0.5Al1-MMO-10.

264
H. Zhang et al. / Journal of Catalysis 372 (2019) 258–265
different intension of metal-support interaction, leading to signifi-
cantly different stability behaviors of the Ni2Mg0.5Al1-MMO-x cat-
alysts. Notably, Ni2Mg0.5Al1-MMO-8 can be reused for ten cycles
without obvious loss of its initial activity, which is comparable to
most of noble metal catalysts. The simple preparation route, low
cost, excellent catalytic activity and stability make Ni2Mg0.5Al1-
MMO-8 a suitable and inexpensive candidate for large scale hydro-
genations under harsh reaction conditions.
Acknowledgments
The authors thank Professor Gang Ma for his kind help in
preparing this manuscript. Financial supports from the National
Natural Science Foundation of China (21676068), Hundred Out-
standing Innovative Personnel Support Plan of Hebei Universities
(SLRC2017020), 333 Talent Project of Hebei Province
(A2016005006), Natural Science Foundation of Hebei Province
(B2016201167), and Science and Technology Research Project of
Hebei Higher Education Institutions (QN2015037) are gratefully
acknowledged.
Fig. 7. H2-TPR profiles of (a) Ni/Mg0.5Al1-IMP, (b) Ni2Mg0.5Al1-MMO-4,
(c) Ni2Mg0.5Al1-MMO-6, (d) Ni2Mg0.5Al1-MMO-8 and (e) Ni2Mg0.5Al1-MMO-10.
catalysts [20,48], the reduction peaks at different temperatures
were attributed to the reduction of different nickel species. As
can be seen, Ni/Mg0.5Al1-IMP has three main reduction peaks at
351, 642, and 778 °C. The peak at 351 °C is ascribed to the reduc-
tion of the bulk or free crystallites of NiO, which interacts weakly
with the support [49,50]. The high reduction peaks at 642 and
778 °C are attributed to the reduction of NiO that interacts rela-
tively strong with the mixed metal oxide support and the spinel
Appendix A. Supplementary material
Supplementary data to this article can be found online at
https://doi.org/10.1016/j.jcat.2019.03.012.
References
structure of NiAl2O4, respectively [51]. In contrast, all the Ni2Mg0.5
-
[1] A. Mukherjee, D. Srimani, S. Chakraborty, Y. Ben-David, D. Milstein, J. Am
Chem. Soc. 137 (2015) 8888–8891.
Al1-MMO-x samples do not have obvious reduction peak of NiAl2O4,
indicating fewer spinel structure formed in Ni2Mg0.5Al1-MMO-x
catalysts. Ni2Mg0.5Al1-MMO-4 exhibits one small reduction peak
at 360 °C together with a large reduction peak at 628 °C, which
indicates the existence of both weak and strong interactions
between Ni and supports in this sample. Notably, the strength of
lower reduction peak weakened with the increase of the calcina-
tion time and could not be detected in Ni2Mg0.5Al1-MMO-8, indi-
cating only strong interaction between Ni species and the MMO
support existed in Ni2Mg0.5Al1-MMO-8 [52]. The reduction peak
of Ni2Mg0.5Al1-MMO-10 is centered at 625 °C, indicating that pro-
longing the calcination time caused relatively weak interaction
between NiO and supports compared to the Ni2Mg0.5Al1-MMO-8
sample, which is consistent with the XPS results.
[2] X. Qiao, L. Niu, H. Zhang, X. Wen, Y. Cao, G. Bai, Appl. Catal. B: Environ. 218
(2017) 721–730.
[3] H. Cao, H. Liu, A. Dömling, Chem-Eur. J. 16 (2010) 12296–12298.
[4] X. Zhang, B. Zong, M. Qiao, AIChE. J. 55 (2009) 2382–2388.
[5] J.A. Anderson, F.M. McKenna, A. Linares-Solano, R.P. Wells, Catal. Lett. 119
(2007) 16–20.
[6] E. Grootendorst, R. Pestman, R. Koster, V. Ponec, J. Catal. 148 (1994) 261–269.
[7] M. Tang, S. Mao, X. Liu, C. Cen, M. Liu, Y. Wang, Green Chem. 19 (2017) 1766–
1774.
[8] J. Anderson, A. Athawale, F. Imrie, F.M. McKenna, A. McCue, D. Molyneux, K.
Power, M. Shand, R. Wells, J. Catal. 270 (2010) 9–15.
[9] X. Xu, M. Tang, M. Li, H. Li, Y. Wang, ACS Catal. 4 (2014) 3132–3135.
[10] M. Tang, S. Mao, M. Li, Z. Wei, F. Xu, H. Li, Y. Wang, ACS Catal. 5 (2015) 3100–
3107.
[11] W. Feng, H. Dong, L. Niu, X. Wen, L. Huo, G. Bai, J. Mater. Chem. A 3 (2015)
19807–19814.
[12] G. Bai, X. Wen, Z. Zhao, F. Li, H. Dong, M. Qiu, Ind. Eng. Chem. Res. 52 (2013)
2266–2272.
The characterization results presented above suggest that the
calcination time can significantly affect the structures of the LDHs
in-situ derived nickel-based catalysts, which will influence the
interaction between the active nickel species and the MMO sup-
port, and result in different cycle performance. Compared with
the traditional impregnation method, the in-situ generated method
could endow the Ni2Mg0.5Al1-MMO-x catalysts not only highly-
dispersed active nickel species but also strong interaction between
Ni and support, resulting in its high activity and excellent stability
under harsh reaction conditions.
[13] J. Xiong, H. Shen, J. Mao, X. Qin, P. Xiao, X. Wang, Q. Wu, Z. Hu, J. Mater. Chem.
22 (2012) (1932) 11927–11931.
[14] X. Li, D. Li, H. Tian, L. Zeng, Z. Zhao, J. Gong, Appl. Catal. B: Environ. 202 (2017)
683–694.
[15] Y. Cao, L. Niu, X. Wen, W. Feng, L. Huo, G. Bai, J. Catal. 339 (2016) 9–13.
[16] A.G. Boudjahem, S. Monteverdi, M. Mercy, M.M. Bettahar, J. Catal. 221 (2004)
325–334.
[17] Q. Fu, W. Deng, H. Saltsburg, M. Flytzani-Stephanopoulos, Appl. Catal. B:
Environ. 56 (2005) 57–68.
[18] W. Lin, H. Cheng, J. Ming, Y. Yu, F. Zhao, J. Catal. 291 (2012) 149–154.
[19] W. Huang, J.R. McCormick, R.F. Lobo, J.G. Chen, J. Catal. 246 (2007) 40–51.
[20] H. Zhang, X. Gao, Y. Ma, X. Han, L. Niu, G. Bai, Catal. Sci. Technol. 7 (2017)
5993–5999.
[21] X. Wen, Y. Cao, X. Qiao, L. Niu, L. Huo, G. Bai, Catal. Sci. Technol. 5 (2015) 3281.
[22] G. Bai, X. Wen, Z. Zhao, F. Li, H.X. Dong, M.D. Qiu, Ind. Eng. Chem. Res. 53
(2013) 2266–2272.
4. Conclusions
[23] J. Feng, Y. He, Y. Liu, Y. Du, D. Li, Chem. Soc. Rev. 44 (2015) 5291–5319.
[24] F.M. Andrew, R.L. Andrew, P.M. Gordon, H.O. Dermot, Chem. Mater. 11 (1999)
1194–1200.
A series of NiMgAl-MMO catalysts was successfully fabricated
by using
a hydrothermal-calcination-reduction method. The
[25] Y. You, G.F. Vance, H. Zhao, Appl. Clay Sci. 20 (2001) 13–25.
[26] S. Mandal, S. Mayadevi, Appl. Clay Sci. 40 (2008) 54–62.
[27] C.J. Roelofs, J.D. Lensveld, A.J. Dillen, J. Catal. 203 (2001) 184–191.
[28] S. Nishimura, A. Takagaki, K. Ebitani, Green Chem. 15 (2013) 2026–2042.
[29] X. Guo, F. Zhang, E.G. David, X. Duan, Chem. Commun. 46 (2010) 5197–5210.
[30] R.G. Williams, O’hare Dermot, J. Mater. Chem. 37 (2006) 3065–3074.
[31] T. Mitsudome, A. Noujima, T. Mizugaki, Green Chem. 11 (2009) 741–896.
[32] J. Lin, C. Ma, Q. Wang, Y. Xu, G. Ma, J. Wang, H. Wang, C. Dong, C. Zhang, M.
Ding, Appl. Catal. B: Environ. 243 (2019) 262–272.
resulting catalysts showed good catalytic performance in the selec-
tive hydrogenation of benzoic acid under harsh reaction condi-
tions. A Ni2Mg0.5Al1-MMO-8 catalyst inherits the original flower-
like morphology of the LDHs precursor with larger surface area,
pore size and higher stability compared to the catalyst that pre-
pared by the impregnation method with same metal composition.
It was demonstrated that different calcination times result in
[33] L. Lin, D. Tang, Y. Song, B. Jiang, Q. Zhang, Energy 149 (2018) 937–943.

H. Zhang et al. / Journal of Catalysis 372 (2019) 258–265
265
[34] Y. Liu, Y. Gu, Y. Hou, Y. Yang, S. Deng, Z. Wei, Chem. Eng. J. 275 (2015) 271–280.
[35] M. Shao, F. Ning, J. Zhao, M. Wei, E.G. David, J. Am. Chem. Soc. 134 (2012)
1071–1077.
[44] J. Wang, G. Fan, H. Wang, F. Li, Ind. Eng. Chem. Res. 50 (2011) 13717–13726.
[45] M.M. Telkar, J.M. Nadgeri, C.V. Rode, R.V. Chaudhari, Appl. Catal. A: Gen. 295
(2005) 23–30.
[36] K.J. You, C.T. Chang, B.J. Liaw, C.T. Huang, Y.Z. Chen, Appl. Catal. A: Gen. 361
(2009) 65–71.
[37] Z. Liu, Z. Li, F. Wang, J. Liu, J. Ji, J. Wang, W. Wang, S. Qin, L. Zhang, Mater. Lett.
65 (2011) 3396–3398.
[38] Z. Li, L. Mo, Y. Kathiraser, S. Kawi, ACS Catal. 4 (2014) 1526–1536.
[39] H. Ma, L. Zeng, H. Tian, D. Li, X. Wang, X. Li, J. Gong, Appl. Catal. B: Environ. 181
(2016) 321–331.
[40] L. Zhang, F. Li, Appl. Clay Sci. 50 (2010) 64–72.
[41] D. Atong, S. Thassanaprichayanont, V. Sricharoenchaikul, Int. J. Mater. Prod.
Technol. 44 (2012) 201–215.
[46] W. Lin, H. Cheng, L. He, Y. Yu, F. Zhao, J. Catal. 303 (2013) 110–116.
[47] H. Ren, Q. Hao, W. Wang, Y. Song, J. Cheng, Z. Liu, Z. Liu, J. Lu, Z. Hao, Int. J.
Hydrogen Energy. 39 (2014) 11592–11605.
[48] D. Liu, X.Y. Quek, W.N.E. Cheo, R. Lau, A. Borgna, Y. Yang, J. Catal. 266 (2009)
380–390.
[49] K.V. Chary, P.V.R. Rao, V. Vishwanathan, Catal. Commun. 7 (2006) 974–978.
[50] J. Liu, C. Li, F. Wang, S. He, H. Chen, Y. Zhao, M. Wei, D.G. Evans, X. Duan, Catal.
Sci. Technol. 3 (2013) 2627–2633.
[51] X. Yan, Y. Liu, B. Zhao, Z. Wang, Y. Wang, C. Liu, Int. J. Hydrogen Energy. 38
(2013) 2283–2291.
[42] H. Ren, Y. Song, Q. Hao, Z. Liu, W. Wang, J. Chen, J. Jiang, Z. Liu, Z. Hao, J. Lu, Ind.
Eng. Chem. Res. 53 (2014) 19077–19086.
[52] Z. Liu, J. Zhou, K. Cao, W. Yang, H. Gao, Y. Wang, H. Li, Appl. Catal. B: Environ.
125 (2012) 324–330.
[43] J. Ashok, S. Kawi, ACS Catal. 4 (2014) 289–301.