Preparation of semi-hydrogenation catalysts by embedding Pd in layered double hydroxides nanocages via sacrificial template of ZIF-67

Article abstract of DOI:10.1016/j.apcata.2020.117540

In this work a Pd/Fe-Co-Ni layered double hydroxide (LDH) composite was prepared and investigated in the semi-hydrogenation of 2-methyl-3-butyn-2-ol (MBY). Hollow nanocages of ternary Fe-Co-Ni LDH were constructed by stacking the LDH nanosheets edge-to-face which inherited the rhombic dodecahedral structure of ZIF-67 templates. Uniform palladium (Pd) nanocrystals with cubic shapes were imbedded in the LDH nanocages via a solvothermal method. The hydroxyl groups on LDH supports were critical to inhibit the association of excess hydrogen atoms on the catalyst, so less MBY molecules were converted to 2-methyl-2-butanol (MBA) during the hydrogenation catalysis. Under optimized reaction conditions the hydrogenation activity of Pd/Fe-Co-Ni LDH was significantly improved, and the conversion of MBY was higher than 99% with selectivity of 2-methyl-3-buten-2-ol (MBE) exceeding 99%. The catalyst was easily recovered and had little reduction in catalytic performance during the cycling reactions.

Full text of DOI:10.1016/j.apcata.2020.117540

Applied Catalysis A, General 597 (2020) 117540
Contents lists available at ScienceDirect
Applied Catalysis A, General
journal homepage: www.elsevier.com/locate/apcata
Preparation of semi-hydrogenation catalysts by embedding Pd in layered
double hydroxides nanocages via sacrificial template of ZIF-67
a
a
a
a
b
a
a
Peng Zhang , Ziyan Wang , Yan Zhang , Jian Wang , Wenqing Li , Lina Li , Peiping Zhang ,
a
a,
Cundi Wei , Shiding Miao *
Key Laboratory of Automobile Materials of Ministry of Education, Solid Waste Recycling Engineering Research Center of Jilin, School of Materials Science and
a
Engineering, Jilin University, Changchun, 130022, Jilin Prov., China
b
Key Laboratory of Mineral Resources Evaluation in Northeast Asia, Ministry of Land and Resources, Changchun 130061, Jilin Prov., China
A R T I C L E I N F O
A B S T R A C T
Keywords:
In this work a Pd/Fe-Co-Ni layered double hydroxide (LDH) composite was prepared and investigated in the
semi-hydrogenation of 2-methyl-3-butyn-2-ol (MBY). Hollow nanocages of ternary Fe-Co-Ni LDH were con-
structed by stacking the LDH nanosheets edge-to-face which inherited the rhombic dodecahedral structure of
ZIF-67 templates. Uniform palladium (Pd) nanocrystals with cubic shapes were imbedded in the LDH nanocages
via a solvothermal method. The hydroxyl groups on LDH supports were critical to inhibit the association of
excess hydrogen atoms on the catalyst, so less MBY molecules were converted to 2-methyl-2-butanol (MBA)
during the hydrogenation catalysis. Under optimized reaction conditions the hydrogenation activity of Pd/Fe-
Co-Ni LDH was significantly improved, and the conversion of MBY was higher than 99% with selectivity of 2-
methyl-3-buten-2-ol (MBE) exceeding 99%. The catalyst was easily recovered and had little reduction in cata-
lytic performance during the cycling reactions.
Palladium catalyst
Semi-hydrogenation
Layered double hydroxides
2
-Methyl-3-butyn-2-ol
Selectivity
1. Introduction
be an ideal choice for preventing NCs agglomeration. Mesoporous
supports with different levels of pores would save the dosage amount of
Catalytic hydrogenation of acetylenic alcohols to alkenols is a key
novel metals by providing enough surface areas. Sometimes there were
special interactions occurs between catalytic species and supports,
which would benefit the catalytic performance by the so-called syner-
getic effects [9]. For instance, the palladium supported on N-doped
carbons [10], metal-organic frameworks [11], inorganic materials
(ZnO) [12], organic materials (polymers, resins, organic frameworks)
[13], and metallic/non-metallic (Pd-Zn, Pd-Si, etc.) alloys have been
found as catalysts in hydrogenation of acetylenic alcohols [14–19].
Although homogeneous Pd catalysts were found to show excellent se-
lectivity in the semi-hydrogenation of MBY [20], there still remained
some shortages. The high costs and difficulty in separation were always
encountered, and in most cases the high catalytic activity of Pd resulted
in over-hydrogenation. To solve the problem of over-hydrogenation,
poisoning or isolation was used to reduce activity of Pd NCs. The facet
sites of Pd NCs were assumed to be active sites for semi-hydrogenation,
and sites from corners and edges were responsible for over-hydro-
genation [21]. Hydrogenation of acetylenic alcohol was demonstrated
not only to be a structure-sensitive reaction but also a size-independent
on Pd particles (6–13 nm). The particle size did not affect selectivity to
MBE, but did affect the formation of dimers and saturated by-products,
step for fine chemicals and intermediate products. One example is the
hydrogenation of 2-methyl-3-butyn-2-ol (MBY) to 2-methyl-3-buten-2-
ol (MBE), which is critical for synthesis of vitamins, fragrances and
agrochemicals, and has been conducted over lead (Pb)-poisoned pal-
ladium (Pd) supported on calcium carbonate (CaCO
3
) as current Lindlar
catalysts [1,2]. Lindlar catalysts (Pd-Pb/CaCO ) were widely used due
3
to its high selectivity toward semi-hydrogenation [3]. However, draw-
backs were found as toxicity of Pb, necessity of adding quinolone, and
instability of CaCO supports, which resulted in harmful and irrever-
3
sible chemical transformations [4,5]. Various strategies have been
proposed to replace the lead poisoned Lindlar catalysts and to regulate
catalytic activity as well for special use in semi-hydrogenation [6,7].
Selecting metallic palladium as catalysts for semi-hydrogenation was
preferred due to the stronger adsorption of alkynes compared to in-
teraction of alkene on the surface of Pd particles, which was caused by
the high electron density and restricted rotation of carbon-carbon triple
bonds [8]. However, the Pd(0) nanocrystals (NCs) would aggregate due
to the high surface energy caused by their nano-size nature. The het-
erogeneous catalyst which contains Pd NCs and porous supports would
⁎
Corresponding author.
E-mail addresses: miaosd@jlu.edu.cn, miaosd@iccas.ac.cn (S. Miao).
https://doi.org/10.1016/j.apcata.2020.117540
Received 12 December 2019; Received in revised form 15 March 2020; Accepted 21 March 2020
Available online 10 April 2020
0926-860X/ © 2020 Elsevier B.V. All rights reserved.

P. Zhang, et al.
Applied Catalysis A, General 597 (2020) 117540
i.e., the larger particle size the lower ratio of dimers to MBA obtained
NCs. The hydrogen atoms were indicated to diffuse into cages and form
−OH species followed by reacting with MBY. The target product MBE
was obtained while restraining the yield of by-products.
[
22]. It is desirable to prepare Pd catalysts with appropriate size and
specific morphologies that have different types of active sites, e.g.,
planes, corners and edges to fulfill the semi-hydrogenation catalysis.
Usually a corner position in catalysts was designed to improve the
catalytic activity, and the carrier was used to prevent over-hydro-
genation. The Brønsted base, e.g., −OH was found to boost adsorption
and activation of acetylenic alcohols. Mechanisms were identified by
the high basicity of ligands which increased electron density of catalyst
center, functioned as electron donors, and thus the oxidative addition of
2. Experimental
2.1. Chemicals and materials
Chemicals used in this research were provided by Aladdin-E.Com
including Co(NO
‘mim’, AR., ≥99.0%), Fe(NO
≥98%), sodium tetrachloropalladate (Na
3
)
2
·6H
2
O (AR., ≥99.5%), 2-methylimidazole (short for
·9H O (AR), Ni(NO ·6H O (AR.,
PdCl AR., ≥98%), KCl (GR.,
H
2
was able to be accelerated. Examples were found as Brønsted bases
3
)
3
2
3
)
2
2
being beneficial to increase ‘electron’ density of Pd surfaces, which
decreased alkene adsorption, and resulted in a high yield of semi-hy-
drogenation [23]. However, the above conclusions were drawn from
homogeneous catalysts in the semi-hydrogenation, and no trials were
conducted by using of hard-templates with solid alkali as in forms of
supports. Layered double hydroxides (LDH), known as brucite Mg
2
4
99.8%), KBr (AR., 99.0%), ascorbic acid (short for ‘AA’, AR. > 99.0%),
polyvinylpyrrolidone (58,000, K29-32, short for ‘PVP’). 2-Methyl-3-
butyn-2-ol (short for ‘MBY’, purum, ≥99%), 2-methyl-3-buten-2-ol
(short for ‘MBE’, purum, ≥97%), and 2-methyl-2-butanol (short for
‘MBA’, purum, ≥98%). All chemicals were used without further pur-
ification.
2
−
(
OH)
2
-like with intercalated anions (CO
3
3
, NO -, OH-) and water
molecules confined in interlayer galleries, are a kind of basic materials
2
+
3+
z+ n−
z/n⋅mH
2+
3+
with formula [M1−x
M
x
(OH)
2
]
A
2
O. The M
and M
re-
2.2. Preparation of the Pd/Fe-Co-Ni LDH catalyst
present divalent and trivalent metal cations (Fe , Ni , Co2+, etc.),
and are located in host layers with ‘x’ ranging 0.2∼0.33 [24]. The LDH
materials have been demonstrated as heterogeneous catalysts in pho-
tocatalysis, electrocatalysis, and thermocatalysis [25]. Due to the OH-
groups, LDHs were thought as suitable catalysts for hydrogenation of
alkynes. The support induced self-poison/self-isolation, and will make
Pb poisoning unnecessary in Lindlar catalysts. The combination use of
Pd and LDHs have been observed to play an important role in tuning
electronic structure of supported metal NCs [26]. The alkalinity was
found to influence the mobility of Pd atoms, and thus prevent over-
sintering during thermal treatments [27]. It was known that the acti-
vation of hydrogen on Pd catalysts involved a hybrid dissociation that
3
+
2+
2.2.1. Synthesis of ZIF-67
The rhombic dodecahedron ZIF-67 were synthesized via a modified
3 2 2
method [32]. Typically, Co(NO ) ·6H O (1.0 mmol) and 2-methylimi-
dazole (4.0 mmol) were dissolved in 25 mL methanol, respectively. The
two solutions were homogenized via ultrasonication (∼10 min, RT).
The mixture was kept still for about 24 h to precipitate ZIF-67. The ZIF-
67 particles were washed several times with methanol, and dried under
vacuum (60 °C) prior for further use.
2.2.2. Synthesis of Pd NCs
To ensure selectivity of the catalyst, the size of Pd NCs was syn-
thesized at the size of ∼10 nm as possible. The histogram of size dis-
tribution of Pd NCs was shown in Fig. S1. The Pd NCs were synthesized
by mixing PVP (58,000 Mw, 105 mg), AA (60 mg), KBr (5 mg) and KCl
(185 mg) dissolved in deionized water (8 mL) under magnetic stirring
+
produced a negatively charged Pd-Hδ . The positively charged -H
atoms were assumed to bond with −OH groups. Although LDHs have
D structures, the flexible composition and confinement effect make
2
this type of materials serve as supports for noble metals, yet the limited
specific area poses hindrance from their wide applications [28].
Therefore more efforts in preparing porosity with 3D hollow nanos-
tructures have been developed to get higher specific areas [29]. The
cavity constructed by packing LDH platelets would provide shorter
diffusion paths for reactions, and thus affect the adsorption-desorption
of substrate molecules. Recently Chen et al. [30], utilized zeolitic imi-
dazolate framework-67 (ZIF-67) as templates to synthesize hollow LDH
polyhedra via a sacrificial template method. Superior pseudocapaci-
tance properties were obtained by benefiting from the larger specific
area and hierarchical/submicroscopic 3D structures.
2 4
(∼10 min, at RT). Another portion of Na PdCl (57 mg) was dissolved
in deionized water (3.3 mL). The two kinds of solution were mixed, and
were kept under magnetic stirring for 3 h at 80 °C. After reaction certain
amount of acetone (∼15 mL) was added to precipitate Pd NCs via
centrifugation (11,000 rpm, ∼20 min).
2.2.3. Syntheses of the catalyst Pd/Fe-Co-Ni LDH
To load Pd NCs within LDH, the as-prepared template (ZIF-67) was
transferred into a round bottomed flask consisting with 0.01 g Fe
3 3 2 3 2 2
(NO ) ·9H O, 0.08 g Ni(NO ) ·6H O and 25.0 mL ethanol. The Pd NCs
Owning to the special morphology of ZIF-67 as well as inspired by
Chen’s findings [30], we selected the micro-sized ZIF-67 particles as
templates to synthesize Fe-Co-Ni LDH hollow nanocages. The composite
Pd/Fe-Co-Ni LDH was prepared by embedding the pre-synthesized Pd
NCs. The LDH nanocages inherited the ZIF-67 framework perfectly, and
shells were composed of exfoliated LDH nanoflakes. The co-precipita-
dispersed in methanol (4.0 wt.%) were added to the above ethanol so-
lution. The mixture was treated via ultrasonication for 10 min, and was
then refluxed at 80 °C for ca. 1 h. After the solvothermal treatment
composites were collected by centrifugation, washed with anhydrous
ethanol, and dried at 80 °C overnight. The schematic diagram of Pd/Fe-
Co-Ni LDH nanocages were illustrated as Scheme 1. For comparison, the
catalytic hydrogenations were also presented that were catalyzed by the
same size of Pd NCs loaded on different supports, including ZIF-67, Co
LDH, Co-Ni LDH, Fe-Co-Ni LDH been treated at 500 °C, with co-pre-
cipitation of Fe-Co-Ni LDH, and several other inorganic oxide supported
3
+
3+
2+
tion of Fe , Co
and Ni
leads to formation of [(Ni,Co)1-x(Fe,Co)
x
−
(
OH)
2
]
z+(NO
3
)
z
+
∙mH
2
O, in which the positive charge z + was tunable
3
3+
2+
by dosages of Fe , Co and Ni . The composites were found to have
sufficient active sites for catalysis. The performance was investigated by
using Pd/Fe-Co-Ni LDH nanocages as catalysts in the semi-hydrogena-
tion of MBY. Under pressure of PH2 = 0.5 MPa at room temperature
2 3 2
catalysts of commercial activated carbon (AC), γ-Al O and SiO . Re-
sults of the hydrogenation were listed in Table 2.
(
RT=∼60 °C), the conversion of MBY reached almost ∼100%, and the
selectivity of MBE was higher than 99% when reactions were conducted
under solvent-free condition. This work validates that LDH can readily
function as nanocages by deliberately controlling synthetic reaction
kinetic balance between precipitation of shells and simultaneous
etching of sacrificial templates [31]. The face-centered cubic of Pd NCs
as catalytic species were successfully embedded in these LDH cages. The
2.3. Characterizations
X-ray diffraction (XRD) patterns were collected on a Regaku D/max-
5000 powder diffractometer equipped with Cu Kα (λ = 0.15418 nm)
radiation operating at 40 kV and 20 mA. Data were recorded within 2θ
of 3-80° at a speed of 0.05°/20 s per step. The morphology was ex-
amined by scanning electron microscope (SEM, Hitachi SU8020) and
2
H adsorption assisted by Brønsted-base could be activated on the Pd
2

P. Zhang, et al.
Applied Catalysis A, General 597 (2020) 117540
Scheme 1. Schematic illustration of the fabrication of Pd/Fe-Co-Ni LDH nanocages and using as catalysts in hydrogenation of 2-methyl-3-butyn-2-ol (MBY).
transmission electron microscope (TEM, JEOL JEM-2100 F) with an
accelerating voltage of 200 kV. The thermal behavior was investigated
by using thermo gravimetric analysis (TG, Netzsch STA 449 F3 Jupiter)
products were evaluated by the GC factor of a certain component which
was obtained by known weight of standard substance to the internal
standard (n-hexane) according to GC analysis. By adding a certain
amount of n-hexane to a sample mixture with known weight, the GC
factor for a certain substance was obtained by the peak ratio of in-
tegrated area of internal standard to those tested components based on
the chromatographic profiles. By considering the GC factor, the per-
centage of each measured component can be achieved. The yield of
each component Yi (where ‘i’ represented MBY, MBE, MBA [34], and
others, e.g., dimmers [35]) was calculated from peak areas of the GC
signal factors. Selectivity to MBE, MBA, and other was defined as
Si = Yi × X × 100%, where X was a MBY conversion. As a measure of
activity, the turnover frequency (TOF) of MBY was calculated as con-
version of moles of MBY per molar Pd in one second. To test the sta-
bility of catalysts, the mixture was recovered (Fig. S2) and reused. The
diagram for hydrogenation of MBY was denoted as Scheme 2.
performed in N
were loaded into the thermal analyzer, and were heated from RT to
000 °C at 5 °C/min. To evaluate the specific surface area (SSA) and
pore-size distribution, samples were degassed at 60 °C for 12 h prior to
the N adsorption-desorption, and this measurement was performed on
a Micromeritics Tristar II 3020 M. The X-ray photoelectron spectrum
XPS) was recorded on a PHI-5300 ESCA spectrometer (Perkin Elmer)
2
atmosphere during heating steps. Samples of ∼10 mg
1
2
(
with its energy analyzer working in the pass energy mode at 30.0 eV,
and the Al Kα line was used as the excitation source. The binding en-
ergy of C1 s was shifted to 284.8 eV as the reference. The XPS profiles
were deconvoluted by using an X-peak program with a Gaussian-
Lorentzian mix function and Shirley background subtraction. Fourier
transform infrared spectroscopy (FT-IR) was collected on a Bruker IFS
6
6 v/s spectrometer (linked with an in-situ test apparatus) which was
−1
registered between 4000 and 400 cm
Temperature-programmed desorption of hydrogen (H
with 128 scans per spectrum.
-TPD) was mea-
3. Results and discussion
2
sured by using a Micromeritics ChemiSorb 2920 equipped with an on-
line mass spectrometry, and a thermal conductivity detector (TCD). The
Pd content was determined by inductively coupled plasma atomic
emission spectroscopy (ICP-AES, Thermo Scientific).
3
.1. Material characterizations
Fig. 1 presented XRD patterns of the pre-synthesized Pd NCs, tem-
plate of ZIF-67, cages of LDH (Fe-Co-Ni LDH), and the final Pd/Fe-Co-Ni
LDH composite after embedding with Pd NCs. The Pd NCs had three
XRD reflections observed at 2θ = 40°, 46° and 68° which could be in-
dexed to (111), (200), and (220) planes of a face-centered cubic (fcc) Pd
(0) (PDF 46-1043) [36]. The XRD pattern of ZIF-67 matched well with
the simulated zeolitic imidazolate framework-67 published elsewhere
[37], demonstrating the synthesized templates as phase-pure ZIF-67.
For the hydrothermal synthesized LDH, i.e., the sample of Fe-Co-Ni LDH
was prepared by sacrificing ZIF-67 frameworks. Three conspicuous
diffraction peaks at 2θ values of 11.7°, 23.5°, and 34.6° were assigned to
reflections of plan (003), (006) and (012) of the LDH phase [38]. The
basal spacing of LDH could be calculated from plane (003) reflection at
2θ = 11.7°. When the Pd NCs were added into Fe-Co-Ni LDH, the re-
lative intensity became weaker. This suggested the composite exhibited
a lower crystallinity, or we could attribute to the exfoliation of LDH
layers. There was no shift in the (003) reflection peak, indicating no
intercalation took place in this process. As determined by the ICP-AES,
the content of Pd in Pd/Fe-Co-Ni LDH was evaluated to be 3.8 wt.%,
which was in good consistence with the initial incorporation of Pd
2.4. Catalyst performance measurements
The catalytic performance on hydrogenation of MBY was evaluated
in a 50 mL stainless vessel equipped with a magnetic stirrer heating
jacket and a Parr high-pressure hydrogen supply system. In a typical
process 10.0 mg of catalyst was mixed with 5.0 mL of MBY, and were
introduced into the vessel. The reactor was flushed vacuum/H alter-
2
natively to remove air inside for four cycles, and heated up to the target
temperature within 20 min. During the reaction the reactor was carried
2
out at desired pressure of H and under intensive stirring ∼2000 rpm.
The starting reaction time was recorded at the point when the above
conditions were achieved. When the reaction was completed, the device
was cooled in an ice bath and depressurized. Samples were periodically
withdrawn from the reactor, and were analyzed by gas chromatography
(
GC, SP-6890) equipped with a FAAP column (30 m, 0.25 mm) and an
FID detector. The injector and flame ionization detector temperatures
were set at 200 °C and 250 °C, respectively. The oven temperature was
maintained at 50 °C for 5 min, and increased to 200 °C at a ramp rate of
2 4
(4.0 wt.%) from Na PdCl , suggesting the efficient use of noble metals.
2
°C/min. Due to the fact that only MBY, MBE and MBA were included
As shown by the field-emission SEM (FESEM) image (Fig. 2a), the
shape of ZIF-67 was of typical rhombic dodecahedron [39]. A magni-
fied image revealed smooth surfaces over the ZIF-67 crystallites
in our system, and the carbon balance was evaluated to be 100 ± 2%
according to the weight assessment [33]. Quantities of reactant/
3

P. Zhang, et al.
Applied Catalysis A, General 597 (2020) 117540
Scheme 2. The hydrogenation of 2-methyl-3-butyn-2-ol (MBY) to 2-methyl-3-buten-2-ol (MBE, semi-hydrogenation) and/or 2-methyl-2-butanol (MBA, over-hy-
drogenation). Other products such as isomers, dimmers were omitted due to the slight amount.
This transformation might be explained by the conversion from ZIF-67
3
+
2+
2+
shell to Fe-Co-Ni LDH nanoflakes in hydrolysis of Fe , Ni , Co ions
(
Scheme 1) during which H+ protons were generated to break bonds in
ZIF-67. Probably part of the Co ions in ZIF-67 were oxidized to Co
2+
3+
after embedding with Pd NCs [30]. The SEM image of final Pd/Fe-Co-Ni
LDH was shown in Fig. 2d denoting fine textures of nanocages. Cages of
Pd/Fe-Co-Ni LDH seemed to be encapsulated by flocculated substances,
and layered substances were assembled with edge-to-face stacking. The
prepared composite Pd/Fe-Co-Ni LDH inherited the morphology and
dimension of ZIF-67 as in forms of rhombic dodecahedron.
To get more structural and compositional information, the TEM
characterization linked with energy dispersive x-ray microanalysis
(
EDS) was carried out. Fig. 3a showed a typical TEM image of the as-
synthesized ZIF-67, which exhibits as polyhedral crystallites. The single
crystallites with smooth surface exhibited an average particle size of ca.
Fig. 1. XRD patterns of Pd NCs, ZIF-67, Fe-Co-Ni LDH and Pd/Fe-Co-Ni LDH.
600 nm. In sample of Fe-Co-Ni LDH the flocculated nanosheets were
found to assemble into cages. Facets of polyhedrons were observed to
construct porous stacking. The contrast between shells and hollow in-
terior clearly revealed the internal cavity. Insert in Fig. 3b was an en-
larged TEM image given to cages as LDH nanosheets with d-spacing of
(
Fig. 2b). After solvothermal treatment with nitrates of iron and nickel,
the SEM of Fe-Co-Ni LDH was given in Fig. 2c. Uniform cages were still
observed, but rhombic dodecahedrons remained less regularly shaped.
Fig. 2. SEM images of ZIF-67 (a, b), the Fe-Co-Ni LDH (c), and the final product of Pd/Fe-Co-Ni LDH with loaded Pd NCs (d).
4

P. Zhang, et al.
Applied Catalysis A, General 597 (2020) 117540
Fig. 3. TEM images of ZIF-67 (a), the as-prepared Fe-Co-Ni LDH nanocages (b), the Pd/Fe-Co-Ni LDH catalyst (c), and the HRTEM of Pd NCs (d). Insert in Fig. 3d is
the fast Fourier transform (FFT) pattern.
Fig. 4. STEM image and elemental mapping of an individual particle of the Pd/Fe-Co-Ni LDH hollow nanocages.
0
.26 nm attributed to plane (006) of LDH. The morphology of co-pre-
with the (111) lattice spacing of face-centered cubic (fcc) Pd [40].
According to EDS analysis the content of Pd in Pd/Fe-Co-Ni LDH was
evaluated to be 3.6 %, and this value was close to ICP-AES measure-
ments. The most striking feature was the uniform sizes of Pd NCs being
monodispersed on surfaces of LDH. A scanning TEM (STEM) image with
EDX mapping confirmed the uniform distribution of Pd, Fe, Co and Ni
throughout the hollow nanocage (Fig. 4). It was worth noting that the
cipitation LDH (Fig. S3) was layered structures with large lateral size of
disorder stacking and small thickness. Fig. 3c showed the TEM image of
Pd/Fe-Co-Ni LDH. Numerous particles were observed with sizes of
∼
10 nm, and particles were uniformly deposited on the LDH layers.
The high-resolution TEM (HRTEM) image (Fig. 3d) of Pd NCs showed
the interplanar spacing of 0.226 nm (inset in Fig. 3d), which agreed
5

P. Zhang, et al.
Applied Catalysis A, General 597 (2020) 117540
-
1
stretching in 2-methylimidazole were observed at 2670 cm and
−1
-1
1849 cm
[41]. Moreover, the band at 3132 cm was assigned to
aliphatic chains. However, the band intensity of imidazole ligands de-
creased sharply after addition of Fe and Ni precursors. The Ni-Co LDH
−1
spectrum displayed a new broad absorption peak at about 620 cm
which was attributed to the Ni-O tensile vibration. The absorption band
−1
of Ni-O-Co had a lower wavenumber within 600-1000 cm . This in-
dicated that part of the ZIF-67 templates were removed and converted
to Ni-Co LDH nanosheets. The Ni-O vibration mode of NO
3
- showed an
−
1
-1
intense band at 840 cm
mation of LDH. A broad band at 3425 cm corresponding to stretching
and 1385 cm− which decreased with for-
1
3
+
of hydroxyl group (−OH) was found to be dependent on ratios of M
/
2
+
3+
3+
(
Fe
M
+ M ) [42]. In order to increase the content of M , ions of
were added to the synthetic LDH cages. For spectra of Fe-Co-Ni
3+
Fig. 5. Typical TG curves of Ni-Co LDH, Ni-Co-Fe LDH and final catalyst of Pd/
Ni-Co-Fe LDH.
−
1
LDH, the IR bands below 900 cm were found as the Fe-O and Fe-O-Fe
vibrational modes [43]. Compared to Ni-Co LDH, modes of v(MeOH) in
−1
M3+/M2+ molar ratio and solvent composition played a crucial role in
Fe-Co-Ni LDH appeared at 527, 555 and 1036 cm [44]. The band of
−1
Fe-Co-Ni LDH at 1631 cm
was particularly observable, which was
the growth of Fe-Co-Ni LDH. The removal of internal templates during
hydrolysis reaction was essential. The template corrosion rate could be
caused by the H-bound −OH groups. The position of −OH absorption
−
1
3
+
band shifted towards lower wavenumber at 3444 cm
3
(vs.
accelerated by adding more Fe
(Fig. S4), but the growth was de-
−1
3+
425 cm ). This was due to the trivalent cations (Fe ) that had
creased by using ethanol as solvents. As a result, the ZIF-67 precursors
−
1
2
+
stronger polarity bonded with −OH. The band around 3636 cm was
were not eaten up completely, and the released Co ions produced the
Fe-Co-Ni LDH nanosheets as confirmed by EDS mapping.
3
+
indicated to free −OH groups, therefore, the addition of Fe
would
enhance alkalinity of LDH. For composite Pd/Fe-Co-Ni LDH, the band of
The TG curves were given in Fig. 5, and all samples had three stages
of weight loss. The slow trend of sample weight loss from RT to 228 °C
was attributed to physically adsorbed substances, e.g., interlayer water
and a small amount of free 2-methylimidazole. The next stage was re-
moval of hydroxyl groups of the laminate and interlayer of carbonate
ions. Probably the structure of laminate was collapsed to form multi-
metal composite oxides. Comparing the three curves in Fig. 5, the
weight loss temperature of Ni-Co LDH was terminated at 340 °C and Fe-
Co-Ni LDH at 270 °C, which suggested the sample Fe-Co-Ni LDH was
more difficult to remove physically adsorbed water. This might also be
related to decomposition of LDH nanosheets. It was found that with
−
1
−1
MeOH at 1036 cm
slightly shifted to 1018 cm
which can be a
+
proof of oxidative Pdδ species. The interlayer hydrogen-bonded
−
1
shifted to lower wavenumber at 2934 cm , and this could be some
impurity with addition of Pd NCs that resulted in disordered structure.
To verify effects of hydroxyl groups on different supports as well as the
catalytic performance, a comparison sample Pd/Al
pared by loading Pd NCs on γ-Al , was also investigated by FT-IR. As
can be seen from Fig. 6 (curve Pd/Al
assigned to the stretching vibration of AlO
2 3
O which was pre-
2 3
O
−1
2
O
3
), The band at 750 cm
in Al [45], and the OeH
bond was observed at ∼3400 cm . Although the position band of
was
4
2 3
O
−
1
3
+
2+
Al−OH in Pd/Al O3 was not shifted by comparison with Pd/Ni-Co-Fe
increment the molar ratio of M /M the content of removable water
decreased, which could be due to the introduction of iron, exhibiting as
increased d(003) reflection, and this was identified by Fig. 1. This
phenomenon was also indicated to increase charge density of the LDH
2
LDH, the band of Al−OH seems to be broadened in Pd/Ni-Co-Fe. This
means that the loading of Pd NCs in Pd/Ni-Co-Fe LDH could be more
complicated, and the binding environment varied from that of Pd/
3
+
Al O . It was known that metal catalysts supported on γ-Al O deliv-
sheets. The more presence of M also enhanced the hydrogen bonding
of water molecules, strengthened the interaction between water mole-
cules and interlayer carbonate ions. By comparing the weightlessness
curves of Fe-Co-Ni LDH and Pd/Fe-Co-Ni LDH, the amount of Pd loaded
on the nanosheets was evaluated to be 3.7 wt.%.
2
3
2 3
ered poor activity and selectivity unless certain promoters were added
if one wanted to get better performance [46]. Suitable promoters were
found as metal hydroxides, for example, the alkaline catalysts would
accelerate the reaction [46]. Explanations were given as that the −OH
groups assisted by alkalinity provided conductive pathways for active
catalyst sites, and this pathway was essential to the hydrogen activation
The FT-IR spectra of Pd NCs, ZIF-67, Ni-Co LDH, Fe-Co-Ni LDH and
Pd/Fe-Co-Ni LDH were illustrated in Fig. 6. The characteristic IR bands
of ZIF-67 was mainly attributed to ligand 2-methylimidazole. The band
[
47]. In other literatures this was also interpreted as the adsorption/
−1
desorption balance afforded by catalyst species with aid of supports,
and has been demonstrated to affect catalysis selectivity notably [48].
XPS measurements were carried out to investigate surface chemistry
of the prepared samples. The wide-range XPS spectrum in Fig. 7a
confirmed the presence of C, N, Fe, Co, Ni and Pd elements in sample of
Pd/Fe-Co-Ni LDH. All XPS spectra were referenced to the C 1s peak at
at 1580 cm
corresponded to stretching mode of the C]N bond of
ligands. The stretching and bending modes of imidazole ring appeared
−
1
at 600–1500 cm . The out-of-plane NeH···N bending and NeH
284.8 eV. Elements of nitrogen and carbon were observed due to the
presence of ligands PVP and/or ZIF templates. Fig. 7b–d displayed high-
resolution of XPS profiles of Fe 2p, Co 2p and Ni 2p. The measured
samples were Fe-Co-Ni LDH and Pd/Fe-Co-Ni LDH as shown in each of
the three figures. To illustrate the possible states of transition metals
(
Fe, Co, Ni, Pd) the high-resolution XPS profiles were deconvoluted
using a Gaussian-Lorentzian mix function. As can be seen (dashed lines
in Fig. 7b–d), the XPS bands of Fe 2p, Co 2p and Ni 2p in Pd/Fe-Co-Ni
LDH were positively shifted (to higher energy) at about 1.0 eV, 0.8 eV
and 1.0 eV by comparison with those in Fe-Co-Ni LDH. These shifts
suggested the occurrence of strong electronic interactions when Pd NCs
were introduced to the Fe-Co-Ni LDH nanocages. These unknown in-
teractions were demonstrated to influence the binding energy of Fe, Co
and Ni in LDH platelets. For example, the Fe 2p core-level spectra
Fig. 6. FT-IR spectroscopy of the synthesized samples.
6

P. Zhang, et al.
Applied Catalysis A, General 597 (2020) 117540
(a)
(
(
c)
(b)
d)
(
e)
Fig. 7. (a) The wide-range survey of XPS spectra, (b) High-resolution of XPS profiles and the corresponding deconvoluted curves of Fe 2p, (c) Co 2p and (d) Ni 2p fort
the Fe-Co-Ni LDH and Pd/Fe-Co-Ni LDH, (e) Pd 3d in sample Pd/Fe-Co-Ni LDH.
(
Fig. 7b) of Pd/Fe-Co-Ni LDH had two spin orbit peaks located at 724.3
energy separation between Ni 2p3/2 and Ni 2p1/2 being16.2 eV could be
assigned to Ni(OH) . The high-resolution XPS profile of Pd was shown
an 710.5 eV, which could be ascribed with Fe 2p3/2 and Fe 2p1/2 of
Fe . The relatively weak bands centered at 710.2 eV could be assigned
2
3
+
in Fig. 7e, displaying characteristic peaks at 335.5 eV and 340.8 eV
which were coincided with Pd 3d3/2 and Pd 3d5/2, respectively. The
element of Pd mainly existed in form of metallic Pd(0) with percentage
of 67.45%, in minority of oxidative Pd2+ of 18.06%, Pdδ of 8.79%
and satellites of 5.7%. There was a signal (336.0 eV) between Pd° and
2
+
to Fe . Another portion of Fe 2p band located at 723.2 eV can be at-
tributed to the satellites [49]. The presence of iron species ensured a
robust hollow nanocage and synergistic effect with the residual imi-
dazole ligands, providing more active sites to bind with Pd NCs, which
would improve catalytic performance as a result [50]. The high re-
solution XPS spectra of Co were displayed in Fig. 7c. As seen from the
fitted curve, two sets of bands representing Co 2p3/2 and Co 2p1/2 were
found at 781.5 eV↔796.7 eV and 782.4 eV↔797.6 eV, respectively.
These two types of bands were attributed to cobalt in form of +2 and
+
2+
Pd , which was found to have a shift of ∼0.8 eV toward higher value
vs. Pd° [12,53], suggesting electron transfer from Pd NCs to LDH sup-
ports, and the Pd NCs can be immobilized by the basic sites on Fe-Co-Ni
LDH. This interactions will lead to the formation of electron-deficient
+
metal atoms (Pdδ ), i.e., there is strong interactions between basic sites
+
3 oxidative states in the cobalt-doped LDH. Moreover, two extra sa-
of supports and Pd [54]. As is known, the metal loading on a support
tellite peaks appeared at 785.8 and 802.7 eV in the deconvoluted curve,
would construct a metal-support interface, which provides an effective
connection for charge transfer [55]. The reason that the state of Pdδ
2
+
+
which was indicative of high-spin states Co . However, this portion of
2
+
Co seems to vary due to the incorporation of Pd NCs, suggesting the
Pd NCs were able to influence chemical environments of LDH as in
terms of binding energy [51]. Similarly the strong peaks of Ni 2p were
also observed and ascribed to Ni 2p3/2 and 2p1/2 located at 855.4 and
existed in Pd/Fe-Co-Ni LDH could be attributed to the large amount of
−OH in Fe-Co-Ni LDH, and the electronegativity of −OH groups would
attract more electrons from surface of Pd NCs. This interaction was
confirmed by the variation of chemical environment of Fe, Co and Ni
+
872.9 eV by fitting the XPS spectrum of Ni in Fig. 7d. Two shakeup
when Pd NCs were incorporated. Therefore, part of the oxidative Pdδ
satellites were observed at about 861.3 and 880.0 eV [52]. The spin-
were indicated to appear in Pd/ Fe-Co-Ni LDH [56]. These interactions
7

P. Zhang, et al.
Applied Catalysis A, General 597 (2020) 117540
would surely enhance the dispersion of Pd and resistance to aggregation
during hydrogenation, thus improving the hydrogen dissociative ad-
sorption in semi-hydrogenation. Okhlopkova et al. [35], found that Pd
NCs modified with PVP would cause XPS band shift to higher energies.
Fig. S6 showed the high-resolution of XPS profiles of N 1s in sample Pd/
Fe-Co-Ni LDH. Even after complete washing, the adsorbed PVP still
remained on Pd NCs. In this experiment the average size of synthesized
Pd NCs was ∼10 nm. The remaining PVP after washing would provide
modification of polymer molecules (PVP) on Pd NCs surface. Although
partial modification by PVP deactivated the Pd catalysts, the certain
extent of deactivation would favor the semi-hydrogenation, i.e., impede
the over-hydrogenation. The presence of PVP prevented aggregation of
Pd NCs during the reaction as well [57]. Therefore, the catalyst per-
formance in our Pd/Fe-Co-Ni LDH was observed almost no dependence
on the initial PVP/Pd ratios when the material (Pd/Fe-Co-Ni LDH) was
used after washing. We indicated the change in electronic properties of
Pd affected the adsorption of alkene, thereby accelerated desorption of
alkene and improved the selectivity as a result [58]. The Pd element in
Pd/Fe-Co-Ni LDH after reusing for 8 times was also checked (Fig. S5).
Table 1
Surface area of ZIF-67, Ni-Co LDH, Fe-Co-Ni LDH and Pd/Fe-Co-Ni LDH.
2
Sample
Specific surface area (m /g)
Average pore size (nm)
ZIF-67
Ni-Co LDH
1593.5
83.2
2.4
6.6
Fe-Co-Ni LDH
Pd/Fe-Co-Ni LDH
139.8
158.2
11.5
14.2
Table 2
a
Hydrogenation of MBY with different samples prepared in this work. .
Entry Catalyst
Time/h Conversion Selectivity TOFb (s―1
)
1
2
3
4
5
6
7
8
ZIF-67
Fe-Co-Ni LDH
Pd NCs
3
3
1
1
1
1
1
1
3.2
＜1
–
–
–
–
89.2
78.3
88.1
92.5
＞99.0
93.6
54.2
49.7
68.5
80.9
＞99.0
79.6
27.1
23.7
26.6
28.0
30.2
28.3
Pd/ZIF-67
Pd/Co-LDH
Pd/Ni-Co LDH
Pd/Fe-Co-Ni LDH
Pd/Fe-Co-Ni LDH
+
2+
The observed less amount of Pdδ and increment of Pd were due to
5
00 °C
being oxidized during the cycling reactions. The XPS of the chlorine
c
9
1
11
Pd/Fe-Co-Ni LDH
Pd/C
1
3
3
5
92.9
＞99.0
85.1
53.8
75.4
38.7
40.6
28.1
10.1
8.6
−
−
(
Cl ) was shown in Fig. S7. The Cl signal was quite weak, whic−h
0
suggested that multiple washings after synthesis would remove Cl
ions completely.
Pd/Al
Pd/SiO
2 3
O
1
2
2
80.2
4.9
2
The N -adsorption/desorption measurements were performed to
better understand structures concerning surface area and pore sizes.
The Fe-Co-Ni LDH, Ni-Co LDH and ZIF-67 were used as reference
samples. The template of ZIF-67 showed N absorption occurs under
2
relatively low pressure, indicating that ZIF-67 was a microporous ma-
a
Reaction condition: catalyst 10.0 mg with 2.0% wt. Pd, MBY 20.0 mL, in-
, T = 60 °C; Entries 9–11 were results catalyzed by commercial
itial 0.5 MPa H
2
2 3 2
catalysts Pd/C, Pd/Al O and Pd/SiO which purchased from Leader Catalyst,
Shanghai.
b
TOF (turnover frequency) was calculated as conversion of moles of MBY
per molar Pd in 1 s.
terial with a Type I behavior. The specific surface area (SSA) of ZIF-67
2
−1
was evaluated to be 1593.5 m g
via the Brunauer-Emmett-Teller
(
BET) method [59]. After solvothermal treatment, the surface area of
calculated by Barrett-Joyner-Halenda means in Fig. 8b. The catalyst of
Pd/Fe-Co-Ni LDH showed a relatively wider distribution. The pore size
distribution was found to range from 4.2 nm to 14.2 nm, and the mul-
timodal pore-size distribution indicated pores were formed by stacking
of LDH. The size of the Pd NCs is about 10 nm, while the average pore
size of Fe-Co-Ni LDH support is about 11.5 nm. In this way it can be
concluded that some Pd nanoparticles entered the pore channels, and
this encapsulation would prevent the loss of nanoparticles during cat-
alysis. The BET specific surface areas and pore size distribution of
samples were listed in Table 1.
The catalytic performance of Pd/Fe-Co-Ni LDH for semi-hydro-
genation of liquid MBY was investigated. All reactions were carried out
at 60 °C, 0.5 MPa and 2000 rpm without using initiator. As a type of
terminal alkynes MBY is easily over-hydrogenated to alkanes because of
higher reactivity of alkynes. The catalytic performances of Pd/Fe-Co-Ni
LDH and other samples (ZIF-67, Fe-Co-Ni LDH, Pd NCs, et al.) prepared
was listed in Table 2. Images of gas chromatography of Pd/Fe-Co-Ni
LDH reduce MBY to MBE were shown Fig. S8. The performance
2
−1
material was decreased to 83∼139 m g , suggesting the collapse of
ZIF-67 framework and transformation into LDH platelets. The SSA of
samples Fe-Co-Ni LDH and Ni-Co LDH was 139.8 m g
8
our LDH-based material was found to be a relative high level by com-
parison with other’s LDH materials in literatures [53]. Furthermore, the
sample Pd/Fe-Co-Ni LDH with loading of Pd NCs had a SAA of
2
-1
and
2
−1
3.2 m g , respectively. Due to the nanocage structure, the SAA of
2
−1
1
58.2 m g
which was much larger than those bare LDH nanocages.
This means the nano-sized Pd in LDH cages contributed to a higher
surface area. In Fig. 8a the hysteresis loop of Pd/Fe-Co-Ni LDH was
larger than others’. The hysteresis loop were observed for Pd/Fe-Co-Ni
LDH, Fe-Co-Ni LDH and Ni-Co LDH in range of P/P = 0.4–0.8, sug-
0
gesting the hollow cages belonged to an adsorption-desorption behavior
of typical IV. The uniform mesopores might come from ZIF-67 tem-
plates. The hysteresis loop at relatively high pressures (P/P > 0.9)
0
indicated the presence of macropores in Pd/Fe-Co-Ni LDH, which might
be formed by packing of LDH sheets. The pore size distribution was
Fig. 8. (a) N
2
-adsorption/desorption isotherms and (b) Barrett-Joyner-Halenda of the prepared ZIF-67, Ni-Co LDH, Fe-Co-Ni LDH and Pd/Fe-Co-Ni LDH samples.
8

P. Zhang, et al.
Applied Catalysis A, General 597 (2020) 117540
commercial Pd/C, Pd/Al
2
O
3
, and Pd/SiO
2
were also given in Table 2.
and concentration of MBE, MBA were plotted versus time (Fig. 9a). The
concentration of MBA tended to have an increasing trend at the early
stage, and the maximum was found at about ∼50 min which was fol-
lowed by an almost baseline of MBA concentration (Fig. 9a). The
maximum yields of MBE exceeded 99% at a conversion over 99% when
the reaction time was 60 min. The over-hydrogenation of MBY to MBA
was not observed before 200 min. Moreover, dimers of MBY were not
observed throughout the hydrogenation. This indicated that the −OH
groups of LDH in the acetylenic alcohol/enol system actually binds to
Pd sites upon adsorption, thereby preventing polymerization [71]. The
unique selectivity of Pd/Fe-Co-Ni LDH catalyst could be further ex-
tended to selective hydrogenation of C^C and C]C, another experi-
ment using a 1:1 mixture of MBY and MBE was carried out (Fig. 9b).
The result showed that the only product was MBE with 99% yield, in-
dicating that over-hydrogenation of MBE never occurred at the first
stage of 200 min.
As to the recyclability, recycling runs of catalysis were carried out
on Pd/Fe-Co-Ni LDH. The Pd/Fe-Co-Ni LDH displayed an extremely
high stability as no deactivation was observed even after up to 8 cycling
tests (Fig. 10). The conversion of MBY was found to be always higher
than 99.0%, and the selectivity of MBY to MBE remained constant. To
detect the loss of Pd amount in the recycling tests, samples were col-
lected and were analyzed by the ICP-AES. The Pd loss percentage of Pd/
Fe-Co-Ni LDH was < 0.5% after eight circles of reuse.
To verify roles of supports, the Fe-Co-Ni LDH was used for hydro-
genation reaction under the same condition, but no noticeable con-
version of MBY was observed (Entry 2). This confirmed the Fe-Co-Ni
LDH was inactive for hydrogenation catalysis without Pd NCs. For
catalysis using pure ZIF-67, the amount of MBY was only observed, and
almost no products were formed. The hydrogenation performance of Pd
NCs to MBY was excellent, i.e., the conversion rate is 89.2%, but the
selectivity is only 54.2%. This indicated that MBY on Pd catalyst was
easily to be over-hydrogenated. The Pd/ZIF-67 of conversion in MBY is
78.3%, and the selectivity was 49.7%. The catalyst of Pd/Co-LDH
showed a conversion of 88.1%, and the selectivity was 68.5%, sug-
3
+
gesting the doping of Co was beneficial to catalysis. More addition of
2
+
Ni would give synergistic effect to provide enough active sites [60],
and the conversion was improved from 88.1% to 92.5%. The Pd/Fe-Co-
Ni LDH had excellent performance for semi-hydrogenation of MBY to
MBE by comparison with Pd/Ni-Co LDH. This suggested that Fe3+ had
modulating effect on Ni- and Co-based LDH. To verify the effect of
−
OH on selective hydrogenation, the LDH supports were fully calcined
at 500 °C to remove −OH from supports, catalysts were post-prepared
by loading with Pd NCs. The conversion of this sample reached 93.6%,
but the selectivity to MBE was only 79.6%. This demonstrates that
groups of −OH in catalyst supports were important to the selectivity.
The commercial Pd/C showed high activity for conversion as shown
Table 2, but the selectivity was low. The selectivity of catalysts Pd/
To validate the selective adsorption of MBY and MBE on catalyst
supports, we used FT-IR spectrum to detect species adsorbed on the Pd/
Fe-Co-Ni LDH. Pure MBY and MBE were flowed into a quartz tube at the
reaction temperature (100 °C) for 60 min, and then purged with Ar flow
for about 15 min to remove the gaseous and physically adsorbed MBY
or MBE. As expected, bonds of C^C and C]C were detected. Fig. 10a
showed characteristic bands of MBY, MBE and MBA. The absorption
2 3 2
Al O and Pd/SiO were even lower, suggesting the LDH played an
essential role in hydrogenation of MBY. To confirm effects of LDH on
selective hydrogenation, the turnover frequency (TOF) was derived.
Obviously the introduction of LDH nanocages had higher activity and
selectivity in the selective hydrogenation of MBY. Although the TOF
value of Pd monometallic catalyst is close to those of several types of
LDH supported catalysts, the selectivity is low. Compared to other
−1
peak at 2250 cm
was a typical telescopic vibration of C^C [72].
-
1
different supports, Pd/SiO
times as that of Pd/Fe-Co-Ni LDH. The TOF of Pd/Al
2
shows the lowest TOF which was only 0.16
Other band at 1626 cm was assigned to the C]C [73]. It should be
noted that the band of C]C was observed in Fig. 10b, which indicated
the MBE group appeared on the catalyst within 30 min since the start of
reaction. This might be the conversion of MBY to MBE or the adsorption
of MBE on catalysts. As the reaction time was prolonged to 60 min, the
competition adsorption of MBY and MBE occurred. However, the C^C
remained an stabilized state on the catalyst. Bands at 1496 and 1450
−1
2 3
O is 8.6 s , and
-
1
the selectivity is even lower. It is interesting that TOF of 10.1 s and
high selectivity values were observed for Pd/C. Among the LDH support
catalysts, the Pd/Fe-Co-Ni LDH has the highest TOF and selectivity.
Transition metals of Fe, Co and Ni have synergistic effects with Pd NCs
in hydrogenation. Moreover, the selectivity of Pd/Fe-Co-Ni LDH with
hollow nano-cages for hydrogenation is much higher than LDH pre-
pared with traditional co-precipitation under similar TOF. To evaluate
the level of our prepared catalysts literatures concerning on hydro-
genation of MBY were checked throughout recent publications, and the
literature data were summarized in Table 3. The performance of Pd/Fe-
−
1
cm assigning to ν(CeC) of the MBA were not observed. Therefore, it
could be concluded that the adsorption of C^C on Pd/Fe-Co-Ni LDH
was stronger than C]C, which favors the detachment of C]C bond in
MBE instead of the C^C in forms of MBY, resulting in hydrogenation of
MBE until MBY was completely converted.
Co-Ni was comparable to those of Lindlar catalysts (Pd-Pb/CaCO
3
)
The mechanism can be drawn as that hydrogenation of alkyne firstly
occurred on surface of Pd nanocubic particles, and the support acted as
a co-catalyst to reduce electron density of the d-band of Pd, and thus
decreased the adsorption energy of alkenes. The LDH support inhibited
the formation of intermediate specie, e.g., β-palladium hydrides (β-
PdH) with positively charged H-atoms, which was prone to over-
[
61], but our prepared Pd/Fe-Co-Ni LDH had advantages by conducting
semi-hydrogenation solvent-free, and no heavy metals were used in
view of avoiding potential dangers.
For the kinetics study, the Pd/Fe-Co-Ni LDH was further monitored
during the hydrogenation reaction of MBY, and data of MBY conversion
Table 3
Comparison of our prepared composite in catalysis of the MBY hydrogenation with previous reported catalysts in literatures.
Catalyst
Reaction condition
MBY Conversion
MBE Selectivity
Ref.
Pd NCs (13 nm)
333 K, 0.28 MPa, Solvent free
323 K, 0.1 MPa, in hexane
333 K, 0.7 MPa,
313 K, 0.5 MPa, in methanol
308 K, 0.5 MPa, in water
363 K, 0 MPa, toluene
303 K 0.1 MPa, toluene
343 K, 0.5 MPa
335K in capillary in hexane
313 K, 0.3 MPa, ethanol
313 K, 0.3 MPa, ethanol
333 K, 0.5 MPa,
∼97%
–
90%
86.5%
∼95%
95.1%
∼100%
> 99%
∼95%
95%
96%
94-96%
88%
74.5%
95%
97.9%
100%
97%
[62]
[33]
[63]
[64]
[65]
[66]
[67]
[68]
[69]
[70]
[61]
this work
Pd-Bi/SiO
Pd/ZnO Nanowires
PdZn/TiO
PdZn/CN@ZnO
Pd/Fe /γ- Fe
Pd S/C
Pd/ZnO
Pd–Bi/TiO
.2 wt.%Pd/MN270
Pd-Pb/CaCO
Pd/Fe-Co-Ni LDH
2
2
3
O
4
2
O
3
3
3 4
N
2
93 ± 1.5%
98.7%
98%
0
3
∼100%
∼100%
99%
9

P. Zhang, et al.
Applied Catalysis A, General 597 (2020) 117540
(a)
(b)
Fig. 9. (a) Percentage of MBY and products versus reaction time during hydrogenation over Pd/Fe-Co-Ni LDH catalyst. (b) The experiment using a 1:1 mixture of
MBY and MBE for verifying selectivity of Pd/Fe-Co-Ni LDH catalyst.
Fig. 10. Reuse of the Pd/Fe-Co-Ni LDH and reactions were conducted under
conditions: T = 60 °C, PH2 = 0.5 MPa, and reaction time = 2.0 h.
Fig. 12. H -TPD profiles of the catalysts Pd/Fe-Co-Ni LDH and Pd/Co-Fe LDH.
2
species. The higher temperature peak (150–500 °C) was attributed to
dissociative chemisorbed hydrogen. In case of the Pd/Fe-Co-Ni-LDH, all
peaks were observed at lower temperatures compared to the Pd/Co-Ni
LDH. The reason lies in the introduction of Fe3+ which accelerate the
hydrogenation to alkanes [74]. Due to the positively charged H-atoms
the intermediate species were tended to bond with chemical groups
surrounding the support. The Fe-Co-Ni LDH was assumed to have −OH
groups as confirmed by the IR characterization, and these groups would
attract a positively charged H (Fig. 11). Therefore, the LDH inhibited
the formation of β-PdH, and the semi-hydrogenation was achieved
electron transfer ability [76]. Both curves showed at least four H
desorption peak regions. The desorption peaks between 120–350 °C
corresponded to desorption of H from Pd sites onto the surface of LDH
which is due to the strong interaction between Pd-LDH and H at the
interface sites. The peak at higher temperature was attributed to the H
on layers of LDH and/or spillover H [77]. Compared to Pd NCs sup-
2
2
[
75].
It was well known that the H
portant step during hydrogenation. In this research the H
periment was carried out to investigate the H -adsorption behavior on
2
2
activation/dissociation was an im-
-TPD ex-
2
2
2
2
ported on Fe-Co-Ni LDH, the peak area at 394 °C for Pd/Co-Ni LDH was
much larger than that of Pd/Fe-Co-Ni LDH (371 °C), i.e., the weightage
of this peak in all H-desorption peaks was much higher in Pd/Fe-Co-Ni
LDH based catalysts. The addition of Fe to LDH was verified to make the
catalyst more prone to hydrogen dissociation and alleviate the over-
hydrogenation. The H -TPD profiles were presented in Fig. 12. As can
2
2
LDH, thereby indicating the activation/dissociation of H was more
be seen, both TPD curves of Pd/Fe-Co-Ni LDH and Pd/Co-Ni LDH were
observed to have two kinds of desorption peaks. The lower temperature
peak (50–150 °C) represented desorption of weakly adsorbed hydrogen-
prone to occur on Pd/Fe-Co-Ni nanocages, thus contributing to im-
provement of activity and selectivity.
To investigate the effect of different components on catalytic
Fig. 11. (a) The FT-IR spectra of the alkynol, alkene and alkyne; (b) The adsorption and desorption of alkene and alkyne for Pd/Fe-Co-Ni LDH at 30 min and 60 min.
10

P. Zhang, et al.
Applied Catalysis A, General 597 (2020) 117540
Acknowledgements
We acknowledged financial supports by National Natural Science
Foundation of China (51874145), Open Foundation of State Key
Laboratory of Mineral Processing (BGRIMM-KJSKL-2019-07), Project
guide of Jilin Science and technology development plan in 2020, China
Ocean Mineral Resources R&D Association (COMRA) Special
Foundation (DY135-46, DY135-R2-1-01), and the Province/Jilin
University co-construction project
SXGJSF2017-3).
– funds for new materials –
(
Appendix A. Supplementary data
Fig. 13. CV curves of Pd, Fe-Co-Ni LDH and Pd/Fe-Co-Ni LDH in pH = 14 of
KOH, 40 mV/S.
Supplementary material related to this article can be found, in the
online version, at doi:https://doi.org/10.1016/j.apcata.2020.117540.
activity in Pd/Fe-Co-Ni LDH, the electrocatalytic tests of Pd, Fe-Co-Ni
LDH and Pd/Fe-Co-Ni LDH were conducted in 1 M KOH by cyclic vol-
tammogram (CV, Fig. 13). The CV of samples were carried out in po-
tential range of -0.14∼1.14 V vs. Ag/AgCl (saturated with KCl). The
activity was evaluated by electrochemically active surface area (ECSA),
i.e., the hydrogen adsorption-desorption. It was found that the ECSA of
Pd/Fe-Co-Ni LDH was higher than Pd NCs or Fe-Co-Ni LDH, which
suggests the Pd/Fe-Co-Ni LDH has the highest hydrogenation activity
References
[
1] R. Chinchilla, C. Najera, Chem. Rev. 114 (2014) 1783–1826.
[
2] M. Eggersdorfer, D. Laudert, U. Letinois, T. McClymont, J. Medlock, T. Netscher,
W. Bonrath, Angew. Chem. Int. Ed. 51 (2012) 12960–12990.
[
3] G. Vile, N. Almora-Barrios, S. Mitchell, N. Lopez, J. Perez-Ramirez, Chem. Eur. J. 20
(2014) 5926–5937.
[
4] M. Garcia-Mota, J. Gomez-Diaz, G. Novell-Leruth, C. Vargas-Fuentes, L. Bellarosa,
B. Bridier, J. Perez-Ramirez, N. Lopez, Theor. Chem. Acc. 128 (2011) 663–673.
[
78]. By comparison with free Pd NCs or Pd/Fe-Co-Ni LDH, the Fe-Co-
[5] N. Semagina, M. Grasemann, N. Xanthopoulos, A. Renken, L. Kiwi-Minsker, J. Catal.
51 (2007) 213–222.
2
Ni LDH was found to have an obvious oxidation peak within 0.8∼0.7 V
and reduction potential of 0.76∼0.54 V (see the rectangular area in
Fig. 13), which benefits from the reversible charge transfer to Pd in Fe-
Co-Ni LDH.
[
[
6] K.M. Engle, T.S. Mei, M. Wasa, J.Q. Yu, Acc. Chem. Res. 45 (2012) 788–802.
7] S. Zhou, L. Shang, Y. Zhao, R. Shi, G.I.N. Waterhouse, Y.C. Huang, L. Zheng,
T. Zhang, Adv. Mater. 31 (2019).
[
[
8] A. Molnar, A. Sarkany, M. Varga, J. Mol. Catal. A Chem. 173 (2001) 185–221.
9] M. Tian, X. Cui, C. Dong, Z. Dong, Appl. Surf. Sci. 390 (2016) 100–106.
[
[
[
10] X.X. Li, Y. Pan, H. Yi, J.C. Hu, D.L. Yang, F.Z. Lv, W.D. Li, J.P. Zhou, X.J. Wu,
A.W. Lei, L.N. Zhang, ACS Catal. 9 (2019) 4632–4641.
11] J. Yang, F.J. Zhang, H.Y. Lu, X. Hong, H.L. Jiang, Y. Wu, Y.D. Li, Angew. Chem. Int.
Ed. 54 (2015) 10889–10893.
4
. Conclusions
12] N. Iwasa, S. Masuda, N. Ogawa, N. Takezawa, Appl. Catal. A Gen. 125 (1995)
In summary, a new type of catalysts based on layered double hy-
145–157.
[
[
13] S.A. Jagtap, B.M. Bhanage, Mol. Catal. 460 (2018) 1–6.
14] Y. Wan, H.Y. Wang, Q.F. Zhao, M. Klingstedt, O. Terasaki, D.Y. Zhao, J. Am. Chem.
Soc. 131 (2009) 4541–4550.
droxides nanocages were developed, and was successfully employed in
the semi-hydrogenation of 2-methyl-3-butyn-2-ol. This catalysts have
its superiority in none use of heavy metals, e.g., the doped Pb, which
was always found in the previous Lindlar catalysts. The prepared Pd/
Fe-Co-Ni LDH delivers brilliant performance on semi-hydrogenation of
[15] F. Tao, M.E. Grass, Y.W. Zhang, D.R. Butcher, J.R. Renzas, Z. Liu, J.Y. Chung,
B.S. Mun, M. Salmeron, G.A. Somorjai, Science 322 (2008) 932–934.
[
[
16] Z.L. Li, J.H. Liu, C.G. Xia, F.W. Li, ACS Catal. 3 (2013) 2440–2448.
17] M. Armbruster, K. Kovnir, M. Behrens, D. Teschner, Y. Grin, R. Schlogl, Chem. Soc.
132 (2010) 14745–14747.
2-methyl-3-butyn-2-ol to 2-methyl-3-buten-2-ol. Under pressure of
P
H2 = 0.5 MPa at ∼25 °C), the conversion of 2-methyl-3-butyn-2-ol
[18] W.X. Niu, Y.J. Gao, W.Q. Zhang, N. Yan, X.M. Lu, Angew. Chem. Int. Ed. 54 (2015)
8
271–8274.
reached ∼100%, and the selectivity of 2-methyl-3-buten-2-ol was
higher than 99% when reactions were conducted under solventless
condition. Even at prolonged reaction time, the hydrogenation product
still remained at alkene. This work validates that LDH can function as
supports by deliberately control of precipitation of shell materials and
simultaneous etching of sacrificial templates (ZIF-67). The transition
metal doped LDH, Fe-Co-Ni LDH, was demonstrated to play a vital role
in suppressing the over-hydrogenation of alkynes to alkanes. The cat-
alyst has advantage in the selective semi-hydrogenation of various al-
kynes by the synergy effect between Pd and Fe-Co-Ni LDH.
[
19] F. Studt, F. Abild-Pedersen, T. Bligaard, R.Z. Sorensen, C.H. Christensen,
J.K. Norskov, Angew. Chem. Int. Ed. 47 (2008) 9299–9302.
[20] P. Hauwert, G. Maestri, J.W. Sprengers, M. Catellani, C.J. Elsevier, Angew. Chem.
Int. Ed. 47 (2008) 3223–3226.
[
21] A. Prestianni, M. Crespo-Quesada, R. Cortese, F. Ferrante, L. Kiwi-Minsker, D. Duca,
J. Phys. Chem. C 118 (2014) 3119–3128.
[22] J.P. Boitiaux, J. Cosyns, S. Vasudevan, Appl. Catal. 6 (1983) 41–51.
[
[
[
23] R. Noyori, T. Ohkuma, Angew. Chem. Int. Ed. 40 (2001) 40–73.
24] Q. Wang, D. O’Hare, Chem. Rev. 112 (2012) 4124–4155.
25] G. Fan, F. Li, D.G. Evans, X. Duan, Chem. Soc. Rev. 43 (2014) 7040–7066.
[26] M. Xu, M. Wei, Adv. Funct. Mater. 28 (2018) 20.
[
27] Y. Tan, X.Y. Liu, L.L. Zhang, A.Q. Wang, L. Li, X.L. Pan, S. Miao, M. Haruta,
H.S. Wei, H. Wang, F.J. Wang, X.D. Wang, T. Zhang, Angew. Chem. Int. Ed. 56
(2017) 2709–2713.
[
[
28] V. Holler, K. Radevik, I. Yuranov, L. Kiwi-Minsker, A. Renken, Appl. Catal. B:
Environ. 32 (2001) 143–150.
29] W. Kong, J. Li, Y. Chen, Y. Ren, Y. Guo, S. Niu, Y. Yang, Appl. Surf. Sci. 437 (2018)
CRediT authorship contribution statement
1
61–168.
Peng Zhang: Investigation, Data curation. Ziyan Wang: Validation.
Yan Zhang: Software. Jian Wang: Visualization. Wenqing Li: Formal
analysis. Lina Li: Supervision. Peiping Zhang: Resources. Cundi Wei:
[30] Z. Jiang, Z. Li, Z. Qin, H. Sun, X. Jiao, D. Chen, Nanoscale 5 (2013) 11770–11775.
[31] A. Berkessel, T.J.S. Schubert, T.N. Müller, J. Am. Chem. Soc. 124 (2002)
8
693–8698.
32] B. Seoane, J.M. Zamaro, C. Tellez, J. Coronas, CrystEngComm 14 (2012)
103–3107.
[
Project
Methodology, Writing
Supervision.
administration.
Shiding
review
Miao:
Conceptualization,
3
-
&
editing, Funding acquisition,
[33] N. Cherkasov, A. Ibhadon, A.J. McCue, J.A. Anderson, S.K. Johnston, Appl. Catal. A
Gen. 497 (2015) 22–30.
34] M. Sudhakar, G. Naresh, G. Rambabu, C. Anjaneyulu, A.H. Padmasri, M.L. Kantam,
A. Venugopal, Catal. Commun. 74 (2016) 91–94.
35] L.B. Okhlopkova, S.V. Cherepanova, I.P. Prosvirin, M.A. Kerzhentsev,
Z.R. Ismagilov, Appl. Catal. A Gen. 549 (2018) 245–253.
[
[
[
[
Declaration of Competing Interest
The authors declared no interests in this this work.
36] Y. Sun, B. Huang, N. Xu, Y. Li, M. Luo, C. Li, Y. Qin, L. Wang, S. Guo, Sci. Bull. 64
(2019) 54–62.
37] R. Banerjee, A. Phan, B. Wang, C. Knobler, H. Furukawa, M. O’Keeffe, O.M. Yaghi,
11

P. Zhang, et al.
Applied Catalysis A, General 597 (2020) 117540
Science 319 (2008) 939–943.
38] Y. Liu, N. Wang, J.H. Pan, F. Steinbach, J.R. Caro, J. Am. Chem. Soc. 136 (2014)
4353–14356.
J.S. Wu, W.N. Zhang, F.W. Huo, ACS Appl. Mater. Interfaces 9 (2017)
15394–15398.
[
[
1
[60] H. Pourfarzad, M. Shabani-Nooshabadi, M.R. Ganjali, H. Kashani, Electrochim. Acta
317 (2019) 83–92.
39] J. Tang, R.R. Salunkhe, J. Liu, N.L. Torad, M. Imura, S. Furukawa, Y. Yamauchi, J.
Am. Chem. Soc. 137 (2015) 1572–1580.
40] J. Silvestre-Albero, G. Rupprechter, H.J. Freund, Chem. Commun. (2006) 80–81.
41] Y. Wang, C.Y. Wang, Y.J. Wang, H.K. Liu, Z.G. Huang, J. Mater. Chem. A 4 (2016)
[61] B. Tripathi, L. Paniwnyk, N. Cherkasov, A.O. Ibhadon, T. Lana-Villarreal, R. Gomez,
Ultrason. Sonochem. 26 (2015) 445–451.
[
[
[62] N. Semagina, A. Renken, D. Laub, L. Kiwi-Minsker, J. Catal. 246 (2007) 308–314.
[63] L.N. Protasova, E.V. Rebrov, K.L. Choy, S.Y. Pung, V. Engels, M. Cabaj,
A.E.H. Wheatley, J.C. Schouten, Catal. Sci. Technol. 1 (2011) 768–777.
[64] L.B. Okhlopkova, S.V. Cherepanova, I.P. Prosvirin, M.A. Kerzhentsev,
Z.R. Ismagilov, Appl. Catal. A Gen. 549 (2018) 245–253.
[65] L.F. Shen, S.J. Mao, J.Q. Li, M.M. Li, P. Chen, H.R. Li, Z.R. Chen, Y. Wang, J. Catal.
350 (2017) 13–20.
5
428–5435.
[
[
42] F. Cavani, F. Trifirò, A. Vaccari, Catal. Today 11 (1991) 173–301.
43] J.M. Hernandez-Moreno, M.A. Ulibarri, J.L. Rendon, C.J. Serna, Phys. Chem. Miner.
1
2 (1985) 34–38.
[
[
[
[
[
[
[
44] S. Abello, F. Medina, D. Tichit, J. Perez-Ramirez, J.C. Groen, J.E. Sueiras, P. Salagre,
Y. Cesteros, Chem-Eur. J. 11 (2005) 728–739.
45] D.M. Marinkovic, J.M. Avramovic, M.V. Stankovic, O.S. Stamenkovic,
D.M. Jovanovic, V.B. Veljkovic, Energy Convers. Manage. 144 (2017) 399–413.
46] Y. Chen, X. Kong, S. Mao, Z. Wang, Y. Gong, Y. Wang, Chin. J. Catal. 40 (2019)
[66] R. Easterday, C. Leonard, O. Sanchez-Felix, Y. Losovyj, M. Pink, B.D. Stein,
D.G. Morgan, N.A. Lyubimova, L.Z. Nikoshvili, E.M. Sulman, W.E. Mahmoud,
A.A. Al-Ghamdi, L.M. Bronstein, ACS Appl. Mater. Interfaces 6 (2014)
21652–21660.
1
516–1524.
47] S. Koukiou, M. Konsolakis, R.M. Lambert, I.V. Yentekakis, Appl. Catal. B: Environ.
6 (2007) 101–106.
[67] D. Albani, M. Shahrokhi, Z.P. Chen, S. Mitchell, R. Hauert, N. Lopez, J. Perez-
7
Ramirez, Nat. Commun. 9 (2018) 11.
[68] M. Crespo-Quesada, M. Grasemann, N. Semagina, A. Renken, L. Kiwi-Minsker,
Catal. Today 147 (2009) 247–254.
[69] N. Cherkasov, A.O. Ibhadon, E.V. Rebrov, Lab Chip 15 (2015) 1952–1960.
[70] L. Nikoshvili, E. Shimanskaya, A. Bykov, I. Yuranov, L. Kiwi-Minsker, E. Sulman,
Catal. Today 241 (2015) 179–188.
[71] S. Mao, B. Zhao, Z. Wang, Y. Gong, G. Lü, X. Ma, L. Yu, Y. Wang, Green Chem. 21
(2019) 4143–4151.
[72] A.A. Titov, V.A. Larionov, A.F. Smol’yakov, M.I. Godovikova, E.M. Titova,
V.I. Maleev, E.S. Shubina, Chem. Commun. 55 (2019) 290–293.
[73] S. Saikia, P. Gogoi, A.K. Dutta, P. Sarma, R. Borah, J. Mol. Catal. A: Chem. 416
(2016) 63–72.
48] P. Zhai, C. Xu, R. Gao, X. Liu, M. Li, W. Li, X. Fu, C. Jia, J. Xie, M. Zhao, X. Wang,
W. Li, Q. Zhang, X.D. Wen, D. Ma, Angew. Chem. Int. Ed. 55 (2016) 9902–9907.
49] X. Zhu, T. Jin, C.C. Tian, C.B. Lu, X.M. Liu, M. Zeng, X.D. Zhuang, S.Z. Yang, L. He,
H.L. Liu, S. Dai, Adv. Mater. 29 (2017) 6.
50] G. Jia, W. Zhang, G.Z. Fan, Z.S. Li, D.G. Fu, W.C. Hao, C.W. Yuan, Z.G. Zou, Angew.
Chem. Int. Ed. 56 (2017) 13781–13785.
[51] X.W. Teng, D. Black, N.J. Watkins, Y.L. Gao, H. Yang, Nano Lett. 3 (2003) 261–264.
[52] Q. Wang, X. Wang, C. Shi, Ind. Eng. Chem. Res. 57 (2018) 12478–12484.
[53] S. Zhao, K. Li, S. Jiang, J. Li, Appl. Catal. B: Environ. 181 (2016) 236–248.
[54] Z. Xu, D.V. Cuong, Y. Liu, W. Baaziz, B. Li, N.D. Lam, O. Ersen, P.H. Cuong, Appl.
Catal. B: Environ. 244 (2019) 128–139.
[55] F. Su, L. Lv, F.Y. Lee, T. Liu, A.I. Cooper, X.S. Zhao, J. Am. Chem. Soc. 129 (2007)
[74] D. Deng, Y. Yang, Y. Gong, Y. Li, X. Xu, Y. Wang, Green Chem. 15 (2013)
1
4213–14223.
2525–2531.
[
56] R.G. Rao, R. Blume, T.W. Hansen, E. Fuentes, K. Dreyer, S. Moldovan, O. Ersen,
[75] K. Radkowski, B. Sundararaju, A. Fuerstner, Angew. Chem. Int. Ed. 52 (2013)
355–360.
D.D. Hibbitts, Y.J. Chabal, R. Schloegl, J.P. Tessonnier, Nat. Commun. 8 (2017)
3
40.
[76] Y.K. Jin, G.H. Sun, F. Xiong, L.B. Ding, W.X. Huang, Chem.-Eur. J. 21 (2015)
[57] A. Yarulin, I. Yuranov, F. Cardenas-Lizana, P. Abdulkin, L. Kiwi-Minsker, J. Phys.
4252–4256.
Chem. C 117 (2013) 13424–13434.
58] H. Hirai, N. Yakura, Polym. Adv. Technol. 12 (2001) 724–733.
59] W.L. Ji, Z.L. Xu, P.F. Liu, S.Y. Zhang, W.Q. Zhou, H.F. Li, T. Zhang, L.J. Li, X.H. Lu,
[77] Y. Cesteros, P. Salagre, F. Medina, J.E. Sueiras, Appl. Catal. B: Environ. 25 (2000)
213–227.
[78] E. Rafiee, M. Kahrizi, F. Farjami, P. Dorraji, Electrochim. Acta 253 (2017) 1–10.
[
[
12