Selective hydrogenation of nitroarenes over MOF-derived Co@CN catalysts at mild conditions

Article abstract of DOI:10.1016/j.mcat.2019.04.008

N-doped porous carbons incorporating highly-dispersed non-noble metallic cobalt (Co) nanoparticle materials were synthesized by rapid pyrolysis of a zeolitic-type metal-organic framework (ZIF-9) and their structures, morphologies, topologies and relevant physical and chemical properties were fully measured by different characterization technologies. The resulting derived Co/CN materials were further evaluated as catalysts in nitrobenzene hydrogenation. It was found that Co/CN materials showed remarkbly catalytic activity and chemoselectivity for mild hydrogenation of nitrobenzene. Amongst the studied a series of Co/CN-x (pyrolyzed at x °C) materials, sample Co@CN-800 shows the most superior catalytic activity for hydrogenation of nitrobenzene. In particular, the catalytic conversion activity of Co@CN-800 is 100% with aniline being the sole product at 70 °C for 2 h, 27.5 times higher than that of cobalt powder (Co). It is believed that the large pore size, adsorption of nitroarenes substrate with high selectivity and strong interaction of Co nanoparticles with the doped N species can result in the high activity of Co@CN-800. This work therefore offers a cost-effective approach in developing highly efficient catalytic materials towards mild hydrogenation of nitrobenzene.

Full text of DOI:10.1016/j.mcat.2019.04.008

MolecularCatalysis472(2019)27–36
Contents lists available at ScienceDirect
Molecular Catalysis
journal homepage: www.elsevier.com/locate/mcat
Selective hydrogenation of nitroarenes over MOF-derived Co@CN catalysts
at mild conditions
Ao Hua, Xinhuan Lua,⁎, Dongming Caia, Haijun Pana, Run Jinga, Qinghua Xiaa,⁎, Dan Zhoua,
⁎⁎
Yongde Xiaa,b,
a Hubei Collaborative Innovation Center for Advanced Organic Chemical Materials & Ministry-of-Education Key Laboratory for the Synthesis and Application of Organic
Functional Molecules, Hubei University, Wuhan 430062, PR China
b College of Engineering, Mathematics and Physical Sciences, University of Exeter, Exeter EX4 4QF, UK
A R T I C L E I N F O
A B S T R A C T
Keywords:
N-doped porous carbons incorporating highly-dispersed non-noble metallic cobalt (Co) nanoparticle materials
were synthesized by rapid pyrolysis of a zeolitic-type metal-organic framework (ZIF-9) and their structures,
morphologies, topologies and relevant physical and chemical properties were fully measured by different
characterization technologies. The resulting derived Co/CN materials were further evaluated as catalysts in
nitrobenzene hydrogenation. It was found that Co/CN materials showed remarkbly catalytic activity and che-
moselectivity for mild hydrogenation of nitrobenzene. Amongst the studied a series of Co/CN-x (pyrolyzed at x
°C) materials, sample Co@CN-800 shows the most superior catalytic activity for hydrogenation of nitrobenzene.
In particular, the catalytic conversion activity of Co@CN-800 is 100% with aniline being the sole product at 70
°C for 2 h, 27.5 times higher than that of cobalt powder (Co). It is believed that the large pore size, adsorption of
nitroarenes substrate with high selectivity and strong interaction of Co nanoparticles with the doped N species
can result in the high activity of Co@CN-800. This work therefore offers a cost-effective approach in developing
highly efficient catalytic materials towards mild hydrogenation of nitrobenzene.
Mild hydrogenation
Heterogeneous catalyst
Co@CN
Nitroarenes
1. Introduction
precious metal catalysts are not chemoselection for dehalogenation and
hydrogenation of other unsaturated bonds (i.e. C]C, C]O, etc.), which
The reduction of nitroaromatic compounds to the substituted ani-
lines has a wide range of applications in the manufacturing of various
polymers, pharmaceuticals, agrochemicals, and fine chemicals [1–4].
Moreover, du to the ease of further derivation for this type of reactions,
they are also used in the synthesis of intermediates for novel high-value
organic compounds in laboratory [5]. Traditionally, aromatic amines
are formed via the hydrogenation of the corresponding nitroaromatic
compounds in the presence of a large number of reducing agents such as
tin (Sn), zinc (Zn), iron (Fe) or sodium hydrosulfite [6–9]. However,
plenty of acids or alkalis are needed to separate the required products
from the reaction mixtures, which not only result in the corrosion of the
reactor, but also generate a large amount of waste metal ions and un-
necessary by-products such as hydroxylamine [10–12].
inevitably lead to some undesirable by-products that are difficult to
separate [17]. On the other hand, the high cost of the noble metals also
constrains their widespread applications [18]. As a result, tremendous
efforts have been dedicated to the developing of non-noble metals (i.e.
Fe, Co, Ni, etc) [19,20], as cost-effective catalysts for hydrogenation
reactions. Although these non-noble metals are cheap and readily
available, they may subject to sintering, leaching and oxidation pro-
blemsin harsh terms [21,22]. As green chemistry becomes more and
more important in modern society, heterogeneous catalysts are playing
increased important role in catalysis research. Actually the discovery of
a large number of heterogeneous catalysts can enable one to choose the
best performed catalysts for a given heterogeneous catalytic reaction
system. Therefore, the exploration of advanced non-noble metal sup-
ported heterogeneous catalysts for the hydrogenation of nitrobenzene
under mild conditions with highly efficient and cost-effective remains
to be a great sought-after target.
It has been also widely reported that aromatic amine compounds are
commonly produced via reactions that depend heavily on noble metal-
based catalysts, such as Ru, Rh, Pd, Pt, Au [13–16]. Nevertheless, these
⁎ Corresponding authors.
⁎⁎ Corresponding author at: Hubei Collaborative Innovation Center for Advanced Organic Chemical Materials & Ministry-of-Education Key Laboratory for the
Synthesis and Application of Organic Functional Molecules, Hubei University, Wuhan 430062, PR China.
E-mail addresses: xinhuan003@aliyun.com (X. Lu), xiaqh518@aliyun.com (Q. Xia), y.xia@exeter.ac.uk (Y. Xia).
https://doi.org/10.1016/j.mcat.2019.04.008
Received 22 January 2019; Received in revised form 25 March 2019; Accepted 7 April 2019
2468-8231/©2019ElsevierB.V.Allrightsreserved.

A. Hu, et al.
MolecularCatalysis472(2019)27–36
Metal-organic frameworks (MOFs), a class of porous crystalline
2.3. Preparation of Co@CN-x catalysts
materials connected by organic ligands and metal ions, are widely ap-
plied in catalysis [23–26], biomedical imaging and drug delivery
[27,28], gas storage and separation [29–31], solar energy harvesting
[32,33], sensing [34–36], and electron and ion conductivity [37,38].
MOF is manufactured based on the synergistic assembly of organic
linkers and metal ions. These structural features can aid the formation
of MOFs and MOF-derived materials with very small and highly dis-
persed nanoparticles, clusters, and even single atoms as catalytically
active sites on in-situ formed porous carbons. Hence, compared to other
classes of porous materials, MOF-based catalysts can offer higher ac-
tivities that are comparable to or even better than their homogeneous
counterparts [39–46]. It was recently reported that cobalt supported on
N-doped carbon materials prepared by the carbonization of MOFs ex-
hibited great activity and chemo-selectivity in hydrogenation of ni-
troaromatics [47–49]. For example, a high surface area of metal-or-
ganic framework (MOF), ZIF-67, is infused with red phosphorous,
followed by pyrolysis to promote the facile production of the phos-
phide-based catalysts (consisting of Co2P/CNx nanocubes) that exhibit
excellent catalytic performance in the selective hydrogenation of ni-
troarenes to anilines [50]. Furthermore, cobalt-based catalysts are
prepared by simple pyrolysis of ZIF-67, which were found to be highly
efficient in the chemoselective hydrogenation of nitroarenes [51].
Herein, we report the synthesis of functionalized Co-MOF (ZIF-9) by
one-step solvothermal crystallization, which is subject to further high-
temperature pyrolysis under N2 atmosphere to generate Co@CN ma-
terials. By applying rotating during the hydrothermal synthesis proce-
dure, the synthesis time for ZIF-9 is shortened (8 h), and the combi-
nation of metal and ligand is more stable during the synthesis of the
crystal compared with the conventional static hydrothermal synthesis
of MOF. Compared to other non-precious metal supported catalysts that
have been reported so far, these Co@CN materials exhibit excellent
hydrogenation activity and selectivity of nitroaromatic compounds
even under mild reaction conditions (70 °C, 2.0 MPa), due to the ex-
istence of Co-N active centers which conducive to adsorb the nitro
groups. In addition, these cobalt-supported catalysts can be easily se-
parated by magnetite for reuse, which is highly desirable in developing
a practical and economical aromatic amine synthesis process.
200 mg of ZIF-9 material synthesized under different conditions was
placed in a crucible and loaded in the center of a tube furnace with N2
atmosphere, with a ramp rate of 10 °C/min to the target heat treatment
temperature of 400, 500, 600, 700 and 800 °C and hold for 3 h at the
target temperature. The furnace was then left to cool down under N2 till
room temperature. The resulting materials were named as Co@CN-x (x
equal heat temperature) respectively.
2.4. Characterization of catalysts
The X-ray diffraction (XRD) patterns of materials were obtained
using
a X-ray powder diffractometer (D8A25, Bruker, Germany)
equipped with Cu Kα radiation (λ = 1.54184 Å). Fourier Transform
Infrared Spectrophotometer (FTIR, OPUS) was set to scan range from
400 cm−1 to 4000 cm−1. Raman spectroscopy was record using a
532 nm argon ion laser range from 100 cm−1 to 3200 cm−1 (Renishaw
Instruments, England). N2 adsorption measurement was using an au-
tomatic gas adsorption system at −196 °C (Micromeritics, TriStar II).
The morphologies of Co@CN-x were collected on a scanning electron
microscope (SEM, JEOL-JSM6700 F) at an accelerating voltage of 15 kV
and on a transmission electron microscope (TEM, JEOL-JEM2010 F) at
an accelerating voltage of 200 kV. Thermal gravimetric analysis (TGA)
were conducted under N2 atmosphere on a PerkinElmer thermal ana-
lyzer. X-ray photoelectron spectroscopy (XPS) characterization of
samples was obtained on a PerkinElmer PHI ESCA system. H2 tem-
perature-programmed reduction (H2-TPR) patterns of samples was
carried out on an automatic gas sorption analyzer (TP-5076) with N2
flow of 40 ml/min (10 vol% H2).
2.5. Evaluation of the catalytic performance of hydrogenation reactions
In a typical process, the catalytic hydrogenations of nitro-based
compounds were conducted in a 50 ml Teflon-lined under stainless steel
autoclave. Firstly, certain amount of catalyst, 0.5 mmol Nitrobenzene
and 15 mg catalyst were dispersed in 2.5 mL H2O, and then transferred
to an autoclave. The autoclave was first purged three times with N2 to
remove air inside, followed by further purging with pure H2 three times
to replace the N2. The reaction was maintained several hours under
vigorous stirring. The filtrate was analyzed by GC-FID (with 30 m ca-
pillary column Rtx@-5).
2. Experimental section
2.1. Chemical reagents
3. Results and discussion
Co(NO3)2·6H2O, Ni(NO3)2·6H2O, Cu(NO3)2·3H2O, Fe(NO3)3·9H2O,
HCOONa, Benzimidazole were purchased from Aladdin Chemicals Co.
Ltd and used as received.
3.1. Characterization of Co@CN materials
In order to obtain high quality of ZIF-9, the synthesis conditions of
precursor ZIF-9 including synthesis temperature, synthesis duration and
rotation speed were first optimized and the quality of ZIF-9 synthesized
under different conditions were evaluated by Powder XRD. As shown in
Fig. S1, the synthesis temperature was raised from 140 to 170 °C, and
the characteristic peak at 2θ = 9.1° of ZIF-9 gradually increases, until
reaching the strongest intensity at 170 °C. However, further increase in
the synthesis temperature to 180 °C results in weakened intensity of the
characteristic peak. Therefore, the aging of the reaction mixture at
170 °C results in the formation of ZIF-9 with the best quality. The
synthesis duration was also optimized and their XRD results are pre-
sented in Fig. S2. The characteristic diffraction peak of ZIF-9 can be
observed even when the aging duration is 2 h. Under optimal synthesis
temperature and rotation speed, the characteristic diffraction peak ZIF-
9 gradually enhances with the increase of crystallization time, and the
crystallinity of the ZIF-9 material is the highest when the reaction
reaches 10 h. The rotation speeds used for the synthesis of ZIF-9 was
also optimized (Fig. S3). The XRD patterns of the resulting materials
confirm the formation of ZIF-9, and the intensity of the main diffraction
2.2. Synthesis of ZIF-9
The preparation of ZIF-9 was based on rotational crystallization of
the solvent by sovlothermal synthesis. Typically, 3.104 g benzimidazole
and 1.789 g sodium formate was fully dissolved in 20 g DMF solution
under stirring at room temperature, then 6.5 mmol metal nitrate [Co
(NO3)2·6H2O, Ni(NO3)2·6H2O, Cu(NO3)2·3H2O or Fe(NO3)3·9H2O] was
added as a metal source to the solution. After that, the solution was
slowly added with another 50 g DMF and keep stirring another 0.5 h.
The mixture was then transferred to an autoclave and aged at tem-
perature of 130, 140, 150, 160, 170 and 180 °C, with crystallization
time of 2, 4, 6, 8 and 10 h under different rotational speed of the au-
toclave at 0, 56, 80, 120 and 150 rpm. Next, the autoclave was slowly
cooled to room temperature. Collected by filtration, the synthesized
purple precipitations were washed three times with deionized water,
followed by further washing with ethanol at 60 °C in a water bath
overnight. Finally, the resulting product was transferred to a vacuum
drying box at 80 °C for 24 h to obtain the ZIF-9 materials.
28

A. Hu, et al.
MolecularCatalysis472(2019)27–36
Fig. 1. Powder XRD patterns of ZIF-9 precursor (a) and Co@CN (b) materials prepared at various heating treatment temperatures.
peak at 9.1° 2θ gradually increases with increasing rotating speed up to
120 rpm. However, when the rotation speed is 150 rpm, the structure of
the resulting material starts to degrade and the crystallinity turns to
decrease.
further up to 700 °C, the amount of Co metal nanoparticles increases
and the edges of the material turn to rough and uneven, which is
consistent with the formation of more crystalline Co nanoparticles as
demonstrated in the XRD pattern of Co@CN-x at higher pyrolysis
temperatures in Fig. 1. When the pyrolysis temperature is 800 °C, the Co
nanoparticle content is the highest and the average particle size of Co
particles is 34.67 nm. Moreover, as demonstrated by the diagram of Co
particle size distribution in Fig. 2, the metal Co nanoparticles in the
sample Co@CN-800 are dominated with the size of smaller than 40 nm.
Obviously, as the pyrolysis temperature increases, the pyrolysis level of
the precursor increases gradually until it fully decomposes and transfers
to Co particles and carbon substrate under higher pyrolysis tempera-
ture, with more Co particles deposited on the surface of the formed
Co@CN materials.
The representative scanning electron microscopy (SEM) images of
Co-ZIF precursor and Co@CN-X materials are presented in Fig. 3. Ob-
viously, the crystals in the samples largely retain their original mor-
phology when the pyrolysis temperature is lower than 500 °C. As the
pyrolysis temperature rises gradually higher than 500 °C, defects start
to appear on the surface of resulting materials, and eventually the
surfaces of the samples turn to much rougher. The change of the mor-
phology of the samples is in agreement with the XRD results in Fig. 1
which clearly indicates that as the pyrolysis temperature increases to
higher than 500 °C, the ZIF-9 precursor turned into Co/Carbon com-
posites.
The powder XRD patterns of the catalytic materials obtained by heat
treatment of ZIF-9 precursor (160 °C, 120 rpm and 8 h) in N2 atmo-
sphere under different temperatures are presented in Fig. 1. When the
heat treatment temperature is up to 500 °C, the resulting materials
generally maintain the characteristic diffraction peak of ZIF-9 pre-
cursor, but the diffraction intensities are gradually decreased. However,
as the heat treatment temperature is higher than 500 °C, the obtained
samples ZIF show no characteristic diffraction peak of ZIF-9 precursor
any more, but accompanied with the appearance of other three peaks in
the XRD patterns at 2θ of 44.3°, 51.6° and 75.8° corresponding to the
(111), (200), and (220) diffractions of Co (JCPDS 15-0806). Moreover,
with increasing heat treatment temperature, the peak intensities for Co
phase in the XRD increased significantly, suggesting the formation of
well crystallized Co particles. In addition, an apparent diffraction peak
at 26° attributed to the reflection of (002) of graphite-type carbon is
also observed at higher heat treatment temperatures. Obviously, sample
Co@CN-800 obtained at higher treatment temperature of 800 °C ex-
hibited the highest diffraction peak intensities for both Co phase and
carbon phase, implying higher pyrolysis temperature results in well-
crystallized Co particles and higher level of graphitic carbon.
The FT-IR spectra of the as-synthesized ZIF-9 precursor and the
derived Co@CN-x catalysts are presented in Fig. S4. Obviously, when
the pyrolysis temperature is up to 500 °C, the FTIR spectra for the
Co@CN-x catalysts are generally the same as ZIF-9 precursor and
maintain the stretching vibration bands of the organic linkers in ZIF-9.
However, with increase the pyrolysis temperature up to 600 °C, the
FTIR spectra for the Co@CN-x catalysts showed that the stretching vi-
bration bands of the organic linkers in the ZIF precursor disappear, due
to the carbonization resulting in structure changes of the organic linkers
in ZIF precursor under higher pyrolysis temperature, which is con-
sistent with the results observed in their XRD patterns shown in Fig. 1.
To evaluate the morphologies and microstructures of the samples,
the TEM images are given in Fig. 2. As seen, when the pyrolysis tem-
perature is up to 500 °C, the crystal surface of the samples does not
change significantly, it’s likely due to the maintenance of the pre-
cursor’s structures. However, as the pyrolysis temperature gradually
rises to above 500 °C, some black particles start to appear on the surface
of the crystal, suggesting the formation of Co particles and carbon due
to the higher pyrolysis temperature of the ZIF precursor, which is
identical to our XRD results in Fig. 1. As the temperature increases
The N2 gas sorption analysis result of the representative sample
Co@CN-800 that derived from the highest pyrolysis temperature is
presented in Fig. 4. The adsorption and desorption curves of Co@CN-
800 exhibit a typical type IV isotherm with a hysteresis loop between
the adsorption and desorption branch, which in the relative pressure P/
P0 range of 0.4–0.9, indicating the presence of mesopores in the sample
with pore size around 3 nm. This sample has a specific surface area of
280.1 m2 g−1 as calculated by the Brunauer-Emmett-Teller (BET)
method, which is typical for Co/carbon based composites.
The representative materials Co@CN-600, Co@CN-700 and
Co@CN-800 were further measured by XPS (Fig. 5(a)). In Fig. 5(b), Co
2p spectrum of Co@CN-x shows an obvious shift of binding energy
when the pyrolysis temperature increases from 600 to 800 °C. It implies
that during the thermolysis process, the Co species in the precursor are
likely reduced to metallic Co particles. Actually, the Co° content in
Co@CN-600 and Co@CN-700 was less than 35%, while there was 48%
in Co@CN-800. Meantime, the N1s signal of these Co@CN samples can
be divided into three types of nitrogen with pyridine-N (398.6 eV),
pyrrole-N (399.6 eV) and graphitic-N(400.9 eV), respectively
29

A. Hu, et al.
MolecularCatalysis472(2019)27–36
Fig. 2. Representative TEM images of ZIF-9 precursor and Co@CN materials prepared at various temperatures. The inset in Co@CN-800 is the diagram of Co particle
size distribution of the sample.
(Fig. 5(c)). The proportion of graphitic N in these Co@CN samples in-
creases with the pyrolysis temperature of the sample, accompanied
with the decrease of the proportion of pyridinic N. Actually, the gra-
phitic N accounts for 66.3% of the total N content in sample Co@CN-
800 which has the highest Co nanoparticle content, indicating that
higher pyrolysis temperature is favorable for the formation of graphitic
nitrogen together with excessive amount of Co metal particles.
As shown in Fig. 5(d), the C1s binding energy region at approxi-
mately 284.5 eV, is distributed to CeCo bonding and CeC]N groups
[52], respectively. With the increase of the calcination temperature,
30

A. Hu, et al.
MolecularCatalysis472(2019)27–36
Fig. 3. Representative SEM images of ZIF-9 precursor and Co@CN materials prepared at various heating treatment temperatures.
this CeC]N signal is weakened significantly, meantime, an
sp3–bonded C peak is appeared at 288.8 eV. This means that a portion
of aromatic or pyridine-like N is transformed into graphitic N. It can
also be verified by the XRD characteristic diffraction peak of Co@CN-X.
As proved, when the reduction potential is higher than −0.27 V, the
metal ions (such as Co2+, Ni2+ and Cu2+) can be readily in-situ re-
duced to metal nanoparticles by the formed carbons during the pyr-
olysis process in an inert atmosphere [53].
plane vibrated structure of sp2 carbon atoms [54], while the D band is
evidenced the disorderly structure of graphite by the defect-caused
Raman characteristic [55]. Higher ID/IG ratio (intensity ratio of D band
to G band) indicates more defects in the carbon substrate of the Co@CN
samples. With temperature increasing from 400 to 800 °C, the ID/IG
ratio increases from 0.99 to 1.17, implying more defects are created at
higher temperature. Therefore, the pyrolysis at 800 °C or above gen-
erates the most defects in the resulting Co@CN composites.
As can be seen from the Raman spectra in Fig. 6, different from the
precursor ZIF-9, all the Co@CN samples derived from different heat
treatment temperatures exhibit clearly observable two Raman peaks at
the ∼1580 and ∼1350 cm−1 are attributed to the G band and the D
band, and the intensities of these two bands in these samples vary with
increasing pyrolysis temperature. The G band is attributed to the in-
The TGA profiles of ZIF-9 and Co@CN-x in nitrogen atmosphere are
presented in Fig. 7. As can be seen from Fig. 7, ZIF-9, Co@CN-400,
Co@CN-500 show similar thermal stabilities, maybe due to the fact that
these samples are basically ZIF-9 structures (according to XRD results of
Fig. 1) and their weight losses above 500 °C are mainly dominated with
the partial decomposition of the organic linkers, accompanied with
31

A. Hu, et al.
MolecularCatalysis472(2019)27–36
minor weight loss event from room temperature to 500 °C that is maybe
due to the loss of water molecules and/or the adsorbed moistures from
the porous structure of the resulting samples. Meanwhile, the other
three samples show similar thermal stability. In their TGA profiles, the
weight losses are contributed from the removal of the adsorbed
moisture at temperature below 500 °C and the burn-off of carbon in
Co@CN composites at temperature higher than 500 °C, respectively. As
Co@CN-800 has the highest level of graphitic carbon (based on XRD
results in Fig. 1), it exhibits the highest thermal stability among these
three composite samples. The weight loss and the relevant weight loss
temperatures of the studied samples were listed in Table S1.
The reducibility of the samples and the reciprocity between Co and
carrier are further characrerized by H2-TPR. The reduction capacity and
interaction intensity between Co and carriers are corrosponding the
different reduction temuerature locations. As shown in Fig. 8, the TPR
profile for ZIF-9 precursor exhibits two signal peaks centred at 605 and
736 °C, which can be assigned to the consective transformation of the
Co species from ZIF-9 to CoO and Co°. This is due to the fact that ZIF-9
needs to be first pyrolyzed and carbonized to form CoO particles and
carbons, then the formed CoO is reduced by H2 and the formed carbons
at a higher temperature to Co metallic particles. However, in their TPR
profiles, the H2 reduction temperatures of the pyrolyzed materials are
significantly lower than those of ZIF-9. For sample Co@CN-400 which
partially remains structures of its precursor ZIF-9 (as demonstrated in
Fig. 4. N2 adsorption/desorption isotherm and the corresponding pore size
distribution profile of sample Co@CN-800.
Fig. 5. XPS spectra of Co@CN-x (a), Co2p (b), N1s (c) and C1s (d).
32

A. Hu, et al.
MolecularCatalysis472(2019)27–36
Fig. 6. Raman spectra of ZIF-9 and the different Co@CN materials.
Fig. 8. H2-TPR of ZIF-9 and different Co@CN materials.
Table 1
Comparison of nitrobenzene hydrogenation over various catalysts.a
Entry
Sample
Conversion
(%)
Aniline selectivity (%)
1
2
ZIF-9
Co
CN
Co3O4
10%Co/C
Co@CN-800
Fe@CN-800
Ni@CN-800
Cu@CN-800
6.9
3.6
> 99
43.7
> 99
48.6
3b
4c
5d
6
19.7
12.3
15.2
100
21.3
46.6
4.6
33.6
> 99
> 99
> 99
> 99
7
8
9
a
Reaction conditions: Nitrobenzene (0.5 mmol), catalyst (15 mg), H2O
(2.5 mL), 2.0 MPa H2, 70 °C, 2 h.
b
Prepared by acid-washed Co@CN-800 with sulfuric acid (33%) overnight
and dried.
c
Prepared by calcination of Co@CN-800 in an air atmosphere at 300 °C for
2 h.
d
Prepared by impregnation method of 15%Co on active carbon (C).
3.2. Catalytic property of Co@CN-x catalysts
Fig. 7. TGA curves of ZIF-9 and Co@CN materials under nitrogen atmosphere.
In order to deeply understand the interactive effects of the various
components, a series of controlled test were performedin the ni-
trobenzene hydrogenation, and the results are shown in Table 1. In the
preliminary catalyst screening, the catalytic performance of N-doped
carbon supported with the metal of Fe, Co, Ni, and Cu and relevant
materials as catalysts in nitrobenzene hydrogenation at 2.0 MPa H2 and
70 °C was studied. The conversion of Fe@CN-800, Ni@CN-800 and
Cu@CN-800 is only 21.3, 46.6 and 4.6% respectively. In contrast, the
Co@CN-800 catalyst exhibits 100% conversion and almost 100% se-
lectivity of aniline (AN). Moreover, compared with other Co-based
catalysts reported in the literature (Table S2) [19,47,56,57], the
Co@CN catalysts in this study not only exhibits higher conversion and
selectivity for aniline, but also achieve better catalytic performance
under much lower reaction temperature, lower pressure H2 and green
H2O as reaction media, which is much advantage over the use of
XRD results in Fig. 1), three TPR reduction peaks appeared at 434, 526
and 578 °C, which can be ascribed to the reduction of the un-pyrolyzed
Co-MOF to CoO and Co°, as well as the reduction of the formed CoOx
from the pyrolysis of the precursor to Co°, respectively. For samples
Co@CN-600 and Co@CN-700 which lost their precursor structures and
are fully pyrolyzed to Co@CN at higher temperatures, three TPR re-
duction peaks appeared at 368, 473 and 578 °C for Co@CN-600, and at
355, 473 and 592 °C for Co@CN-700, which can be ascribed to the
reduction of CoOx particles to CoO to Co°. Nevertheless, the reduction
temperature of Co@CN-800(ca. 672 °C) is transferred higher, it shows
that the metal and the carrier have strong interaction with each other.
Comparatively, the peak of Co@CN-800 is relatively weak. Combined
with the TEM image (Fig.2), multitudes of the Co particals are highly-
dispersed on the surface of CN.
33

A. Hu, et al.
MolecularCatalysis472(2019)27–36
Fig. 10. The influence of the various solvents on the catalytic performance of
Co@CN in nitrobenzene hydrogenation (Reaction conditions: Nitrobenzene
(0.5 mmol), catalyst (Co@CN-800, 15 mg), solvent (2.5 mL), 2.0 MPa H2, 70 °C,
2 h).
Fig. 9. The effect of pyrolysis temperature on the catalytic performance of
Co@CN in nitrobenzene hydrogenation (Reaction conditions: Nitrobenzene
(0.5 mmol), catalyst (Co@CN-x, 15 mg), H2O (2.5 mL), 2.0 MPa H2, 70 °C, 2 h).
optimized the influence of different solvents on the hydrogenation re-
action performance. Among a range of solvents examined (Fig. 10), H2O
is the optimum reaction media for nitrobenzene hydrogenation, with
100% conversion of nitrobenzene as well as > 99% selectivity of ani-
line after 2 h. While polar solvent is used as the reaction media, such as
methanol, ethanol and isopropanol as organic solvents in nitrobenzene
hydrogenation, the catalytic reaction exhibits nitrobenzene conversion
of 65.3–88.9% with aniline selectivity of 89.7- > 99%. The use of weak
polar solvent, such as THF, propylene oxide, dioxane, toluene and cy-
clohexane gives a relatively low nitrobenzene conversion of 7.9–36.0%
with aniline selectivity of 60.7- > 99%. Obviously, the catalytic per-
formance of nitrobenzene hydrogenation reaction in polar solvents is
much greater than in non-polar solvents. The possible explanation
could be that Co@CN-800 catalyst with N doping usually possesses high
hydrophilic properties, which resulted the catalyst dispersion well in
water so as to improve the catalyst exposure to the reactant systems,
consequently increase its catalytic performance.
We further optimized the effect of different synthesis conditions of
ZIF-9 precursor on the catalytic performance of derived Co@CN ma-
terials in nitrobenzene hydrogenation reaction. The catalytic activity of
Co@CN-800 derived from ZIF-9 synthesized at different rotation speeds
is shown in Fig. 11(a). There is slightly increase in the aniline selectivity
with increasing rotation speed. However, the conversion increases
significantly with increasing rotation speed up to 120 rpm, at which the
reaction reaches 100% of nitrobenzene conversion and > 99% of ani-
line selectivity. Further increase the rotation speed to 150 rpm results in
a downward trend of the catalytic activity with nitrobenzene conver-
sion of 68.4%. This downturn could be related to the deterioration of
the crystallinity when the rotation speed is 150 rpm, as shown in Fig.
S3.
relative higher reaction temperature (up to 120 °C), higher H2 pressure
and use organic solvents like ethanol or THF as reaction media in the
same catalytic reaction.
Various parameters that can influence the catalytic performance of
nitrobenzene hydrogenation were systematically optimized. The effect
of pyrolysis temperature of the catalysts on the catalytic activity of
hydrogenation reaction and selectivity of aniline was first investigated.
As presented in Fig. 9, the catalytic hydrogenation activity of ni-
trobenzene significantly increases with increasing pyrolysis tempera-
ture of the catalysts. Co@CN-400 shows almost non-reactivity, but
Co@CN-700 gives 98.2% conversion and Co@CN-800 exhibits > 99%
conversion and aniline selectivity. These results indicate that the pyr-
olysis temperature have a notable effect on the activity of the Co@CN-x
catalyst, and 800 °C results in catalytic activity of 100% and aniline
selectivity > 99%. This can be explained as follows: At lower pyrolysis
temperature (400 °C), the sample still preserves the structure of pre-
cursor ZIF-9 without signification change (as shown the XRD in Fig. 1),
less carbon and Co are formed in the pyrolyzed material, so it shows
almost no catalytic activity. When the pyrolysis temperature was gra-
dually increased, the content of Co° in the resulting catalyst is gradually
increased, so does the catalytic activity. At medium pyrolysis tem-
perature of 500 and 600 °C, as demonstrated by the XRD (Fig. 1), XPS
(Fig. 5) and Raman spectra (Fig. 6), the resulting Co@CN samples
contain Co nanoparticles and low level of graphitic carbon with some
defections according to the Raman spectra; therefore these samples
show medium activities. At high pyrolysis temperature (700 and
800 °C), the resulting samples are Co-based nanoparticles (Fig. 1) and
highly graphitic carbon with more defections (Fig. 6), consequently
these catalysts display high activity. Obviously, with the pyrolysis
temperature increases from 400 to 800 °C, the ID/IG value of carbon in
the prepared samples increases from 0.99 to 1.17, implying more de-
fects are created at higher pyrolysis temperature. In addition, combined
XRD and the XPS results, it is clear that Co@CN-x sample obtained at
higher pyrolysis temperature has higher graphitic-N accounted and
more metallic Co nanoparticles formed, indicating that excessive metal
Co particles is favorable for the formation of graphitic nitrogen at high
temperature. Therefore, the synergistic effect of N-doping and graphitic
carbon and Co lead to the high hydrogenation activity and selectivity.
Using sample (Co@CN-800) which shows the highest activity and
selectivity in nitrobenzene hydrogenation as a model catalyst, we also
The effect of synthesis (crystallization) temperature of precursor
ZIF-9 on the hydrogenation performance of nitrobenzene over the
Co@CN-x catalysts is presented in Fig. 11(b). The hydrogenation ac-
tivity of nitrobenzene increases up to 160 °C, and then decreases with
increasing synthesis temperature of ZIF-9. In particular, the conversion
of nitrobenzene increases from 18.1% at 130 °C to 100% at 160 °C.
Further increase in the synthesis temperature of precursor ZIF-9 to 170
and 180 °C, the catalytic activity however decreases to 91.1 and 57.7%
respectively. As shown in Fig. 11 (c), the synthesis duration of precursor
ZIF-9 can also effect on the hydrogenation activity of nitrobenzene over
the Co@CN-x catalysts. The hydrogenation activity of nitrobenzene
34

A. Hu, et al.
MolecularCatalysis472(2019)27–36
Fig. 11. The effect of (a) rotation speed, (b) synthesis temperature and (c) synthesis time of the ZIF-9 precursor on the catalytic performance of derived Co@CN
materials in nitrobenzene hydrogenation (Reaction conditions: Nitrobenzene (0.5 mmol), catalyst (15 mg), H2O (2.5 mL), 2.0 MPa H2, 70 °C, 2 h).
gradually increases from 74.4% conversion and 93.1% selectivity at 2 h
synthesis duration to 100% conversion and > 99% selectivity at 8 h
synthesis duration. Further increase the synthesis duration to 10 h,
however, leads to the nitrobenzene hydrogenation in the conversion
and the selectivity decrease to 67.2% and 88.3%, respectively. This is
probably because the crystal size of ZIF gradually increases as the
crystallization time increases, which may adversely affect the catalytic
performance of the derived Co@CN catalysts.
The recycling experiments of Co@CN-800 catalyst for the hydro-
genation of nitrobenzene are conducted. As observed from Fig. S5, the
decrease in catalyst performance is much milder with successive cycles.
Obviously, Co@CN-800 has still above 85.2% after the fifth run. ICP-
OES analysis of H2O spent solutions detect < 3 ppm metal, indicating
that leaching was not the cause of the decreased activity. When the
catalyst is used for the 6th times, the reaction time is extended to 3 h,
and the conversion of nitrobenzene can still reach 100%. This indicates
that the catalyst is not deactivated during recycling. It may be that the
organic compounds partially covers the surface of the catalyst during
the reaction, which partially affects the activity of the catalyst. At the
same time, the XRD pattern of the spent Co@CN-800 catalyst exhibit
the same features as the fresh Co@CN-800 (Fig. S6), indicative of a well
preserved structure after recycling.
is generally higher than 80% (entry 2–6). On the contrary, if the sub-
stituent on the aromatic ring of nitrobenzene is an electron-accepting
group, such as halogen-based -F, -Cl, -Br or -I group (entry 7–12), these
single halogen substituted nitrobenzenes show catalytic conversion of
nitrobenzene lower than 66%. However, the catalytic conversion of
multi-halogen substituted nitrobenzene like 2,3,5-trifluoronitrobenzene
(entry 13) over Co@CN-800 catalytic is only 48.8%, much lower than
the single halogen substituted nitrobenzene. In addition, the catalytic
hydrogenation conversion of electron-accepting group eCHO sub-
stituted nitrobenzene with the same conditions is around 10–53% de-
pending on the locations of the eCHO group on the aromatic ring, only
around one tenth to half of that of non-substituted nitrobenzene.
4. Conclusion
Co@CN catalysts have been synthesized by a facile carbonization
approach using precursor ZIF-9 as both the carbon and nitrogen
sources. The mixture exhibits excellent catalytic activity for the ni-
troarenes hydrogenation to relevant anilines under mild conditions.
Some of the most challenging that hydrogenated the nitroarenes with
high reduction to the relevant anilines yields over 80% and aniline
selectivity up to 100%. The excellent catalytic performance is ascribed
to the intimate interaction among the Co particles and the N species,
supporting a higher reducibility to Co. Therefore, this work clearly
demonstrates the interaction of different parameters with catalytic
properties, in which the porous C and N from ZIF-9 precursor play
crucial roles that can boost the formation of Co° in the homodispersion
of Co particles and especially the doped N. Moreover, the Co@CN-800
catalyst as a non-noble metal material has both economical and se-
parability. Consequently, this research offers a cost-effective approach
in the nitrobenzene reduction to aniline.
3.3. Hydrogenation of substituted nitroarenes
To illustrate the versatility of the prepared Co@CN materials in this
work, selective hydrogenation of a series of substituted nitrobenzenes
was further demonstrated, as presented in Table 2. While the sub-
stituent on the aromatic ring of nitrobenzene is an electron-donating
group, such as methyl, hydroxyl or amino group, under the same cat-
alytic reaction conditions, these substituted nitrobenzenes may exhibit
fast hydrogenation rate and their catalytic conversion in hydrogenation
35

A. Hu, et al.
MolecularCatalysis472(2019)27–36
Table 2
[5] Z. Rappoport, The Chemistry of Anilines, John Wiley & Sons, 2007, p. 169.
[6] J. Butera, J. Bagli, WO patent (1991) 91-09023.
[7] A. Burawoy, J. Critchley, Tetrahedron 5 (1959) 340–351.
[8] M. Suchy, J. Wenger, P. Winternitz, M. Zeller, WO Patent (1991) 91-02278.
[9] R.F. Kovar, F.E. Arnold, U.S. Patent (1976) 3975444.
[10] S.M. Kelly, B.H. Lipshutz, Org. Lett. 16 (2014) 98–101.
[11] K. Junge, B. Wendt, N. Shaikh, M. Beller, Chem. Commun. (Camb.) 46 (2010)
1769–1771.
Chemoselective hydrogenation of various substituted nitroarenes.a
Entry
1
Substrate–product
Conversion
100
Selectivity
> 99
2
92.2
97.3
[12] S.K. Mohapatra, S.U. Sonavane, R.V. Jayaram, P. Selvam, Appl. Catal. B 46 (2003)
155–163.
[13] S. Zhang, C.R. Chang, Z.Q. Huang, J. Li, Z. Wu, Y. Ma, J. Am. Chem. Soc. 138 (2016)
2629–2637.
[14] X.L. Yan, P. Duan, F.W. Zhang, H. Li, H.X. Zhang, M. Zhao, Carbon 143 (2019)
378–384.
[15] S.F. Cai, H.H. Duan, H.P. Rong, D.S. Wang, L.S. Li, W. He, ACS Catal. 3 (2013)
608–612.
[16] F.W. Zhang, C. Zhao, S. Chen, H. Li, H.Q. Yang, X.M. Zhang, J. Catal. 348 (2017)
212–222.
3
4
5
6
84.8
98.0
80.7
81.2
> 99
> 99
> 99
96.0
[17] W. She, T.Q.J. Qi, M.X. Cui, P.F. Yan, S.W. Ng, W.Z. Li, ACS Appl. Mater. Interfaces
10 (2018) 14698–14707.
[18] R.F. Nie, J.H. Wang, L.N. Wang, Y. Qin, P. Chen, Z.Y. Hou, Carbon 50 (2012)
586–596.
7
56.2
96.3
[19] X.H. Lu, Y. Shena, J. He, R. Jing, P.P. Tao, A. Hu, R.F. Nie, D. Zhou, Q.H. Xia, Mol.
Catal. 444 (2018) 53–61.
[20] T. Schwob, R. Kempe, Angew. Chem. Int. Ed. 55 (2016) 15175–15179.
[21] P. Munnik, P.E. de Jongh, K.P. de Jong, J. Am. Chem. Soc. 136 (2014) 7333–7340.
[22] L. Li, D.L. King, J. Liu, Q. Huo, K. Zhu, C. Wang, Chem. Mater. 21 (2009)
5358–5364.
[23] T. Sawano, P.F. Ji, A.R. McIsaac, Z.K. Lin, C.W. Abney, W.B. Lin, Chem. Sci. 6
(2015) 7163–7168.
[24] K.K. Tanabe, S.M. Cohen, Chem. Soc. Rev. 40 (2011) 498–519.
[25] K. Manna, T. Zhang, F.X. Greene, W. Lin, J. Am. Chem. Soc. 137 (2015) 2665–2673.
[26] Y.K. Hwang, D.Y. Hong, J.S. Chang, S.H. Jhung, Y.K. Seo, J. Kim, Angew. Chem. Int.
Ed. 47 (2008) 4144–4148.
8
65.0
62.6
62.7
64.9
65.7
81.6
> 99
95.7
97.1
97.1
9
10
11
12
[27] P. Horcajada, C. Serre, M. Vallet-Regí, M. Sebban, F. Taulelle, G. Férey, Angew.
Chem. 118 (2006) 6120–6124.
[28] P. Horcajada, R. Gref, T. Baati, P.K. Allan, G. Maurin, P. Couvreur, Chem. Rev. 112
(2012) 1232–1268.
[29] B. Li, H.M. Wen, Y.J. Cui, W. Zhou, G.D. Qian, B.L. Chen, Adv. Mater. 28 (2016)
8819–8860.
[30] N. Rosi, J. Eckert, M. Eddaoudi, D. Vodak, J. Kim, O.M. Yaghi, 300 (2003)
1127–1129.
13
14
48.8
24.2
99.7
81.2
[31] S.C. Xiang, Z.J. Zhang, C.G. Zhao, K.L. Hong, X.B. Zhao, D.R. Ding, Nat. Commun. 2
(2011) 204.
[32] W.B. Lin, J.R. Long, Inorg. Chem. 55 (2016) 7189–7191.
[33] C.A. Kent, D.M. Liu, L.Q. Ma, J.M. Papanikolas, T.J. Meyer, W.B. Lin, J. Am. Chem.
Soc. 133 (2011) 12940–12943.
[34] L.E. Kreno, K. Leong, O.K. Farha, M. Allendorf, R.P. Van Duyne, J.T. Hupp, Chem.
Rev. 112 (2012) 1105–1125.
15
16
10.1
52.4
65.2
43.6
[35] M.D. Allendorf, R.J.T. Houk, L. Andruszkiewicz, A.A. Talin, J. Pikarsky,
A. Choudhury, J. Am. Chem. Soc. 130 (2008) 14404–14405.
[36] A. Lan, K. Li, H. Wu, L. Kong, N. Nijem, D.H. Olson, Inorg. Chem. 48 (2009)
7165–7173.
[37] L. Sun, M.G. Campbell, M. Dincă, Angew. Chem. Int. Ed. 55 (2016) 3566–3579.
[38] C.G. Silva, A. Corma, H. García, J. Mater. Chem. 20 (2010) 3141–3156.
[39] J. Yang, F.J. Zhang, H.Y. Lu, X. Hong, H.J. Jiang, Y.E. Wu, Y.D. Li, Angew. Chem.
Int. Ed. 54 (2015) 10889–10893.
[40] C. Wang, X. Liu, N.K. Demir, J.P. Chen, K. Li, Chem. Soc. Rev. 45 (2016)
5107–5134.
a
Reaction conditions: substrate (0.5 mmol), catalyst (Co@CN-800, 15 mg),
H2O (2.5 mL), 2.0 MPa H2, 70 °C, 2 h.
Acknowledgments
[41] G. Zhang, J. Zhang, P. Su, Z. Xu, W. Li, C. Shen, Q. Meng, Chem. Commun. (Camb.)
53 (2017) 8340–8343.
[42] C.Y. Zhang, B. Wang, W.B. Li, S.Q. Huang, L. Kong, Z.C. Li, L. Li, Nat. Commun. 8
The authors gratefully acknowledge the financial support of the
project from the National Natural Science Foundation of China
(21503074, 21673069, 21571055), the Hubei Province Students
Innovation and Entrepreneurship Training Foundation of China
(201710512045), and the Hubei Province Outstanding Youth
Foundation (2016CFA040).
(2017) 1138.
[43] W.H. Li, A.F. Zhang, X. Jiang, C. Chen, Z.M. Liu, C.S. Song, X.W. Guo, ACS Sustain.
Chem. Eng. 5 (2017) 7824–7831.
[44] J. Zhao, Z. Tang, F. Dong, J. Zhang, Mol. Catal. 463 (2019) 77–86.
[45] Y. Guo, X. Chen, X. Zhang, S. Pu, X. Zhang, C. Yang, D. Li, Mol. Catal. 459
(2018) 1–7.
[46] C. Guo, Y. Zhang, X. Nan, C. Feng, Y. Guo, J. Wang, Mol. Catal. 440 (2017)
168–174.
[47] X. Sun, A.I. Olivos-Suarez, L. Oar-Arteta, E. Rozhko, D. Osadchii, A. Bavykina,
ChemCatChem 9 (2017) 1854–1862.
Appendix A. Supplementary data
[48] X. Wang, W. Zhong, Y. Li, Catal. Sci. Technol. 5 (2015) 1014–1020.
[49] X. Ma, Y.X. Zhou, H. Liu, Y. Li, H.L. Jiang, Chem. Commun. 52 (2016) 7719–7722.
[50] S.L. Yang, L. Peng, E. Oveisi, S. Bulut, D.T. Sun, M. Asgari, O. Trukhina, W.L. Queen,
Chem. Eur. J. 24 (2018) 4234–4238.
Supplementary material related to this article can be found, in the
online version, at doi:https://doi.org/10.1016/j.mcat.2019.04.008.
[51] X. Wang, Y.W. Li, J. Mol. Catal. A: Chem. 420 (2016) 56–65.
[52] G. Deniau, L. Azoulay, P. Jégou, G. Le Chevallier, S. Palacin, Surf. Sci. 600 (2006)
675–684.
References
[53] J. Long, K. Shen, Y. Li, ACS Catal. 7 (2017) 275–284.
[54] V. Chandra, J. Park, Y. Chun, J.W. Lee, I.C. Hwang, K.S. Kim, ACS Nano 4 (2010)
3979–3986.
[55] R. Nie, H. Jiang, X. Lu, D. Zhou, Q. Xia, Catal. Sci. Technol. 6 (2016) 1913–1920.
[56] M. Tamura, K. Kon, A. Satsuma, K. Shimizu, ACS Catal. 2 (2012) 1904–1909.
[57] B. Chen, F. Li, Z. Huang, G. Yuan, ChemCatChem 8 (2016) 1132–1138.
[1] A. Corma, P. Serna, Science 313 (2006) 332–334.
[2] H.U.B. Dr, H. Steiner, M.S. Dr, ChemCatChem 1 (2010) 210–221.
[3] H. Wei, X. Liu, A. Wang, L. Zhang, B. Qiao, X. Yang, Nature Commun. 5 (2014)
5634.
[4] G. Vilé, N. Almora-Barrios, N. López, J. Pérez-Ramírez, ACS Catal. 5 (2015)
3767–3778.
36