Metal–Organic Framework Mediated Cobalt/Nitrogen-Doped Carbon Hybrids as Efficient and Chemoselective Catalysts for the Hydrogenation of Nitroarenes

Article abstract of DOI:10.1002/cctc.201700095

A Co@N-doped carbon (Co@NC) hybrid was synthesized by thermal decomposition of the metal–organic framework (MOF) ZIF-67 under N₂ atmosphere. These hybrid materials exhibit outstanding catalytic activity and chemoselectivity for the conversion of a wide range of substituted nitroarenes to their corresponding anilines under relatively mild reaction conditions. The high catalytic performance is attributed to the formation of cobalt nanoparticles and to the presence of atomically dispersed Co species in close interaction with nitrogen-doped graphene. Both active species are formed in situ during the pyrolytic transformation of ZIF-67. The catalysts could be reused in consecutive runs, exhibiting a slightly lower activity ascribed to blockage of the active sites by strongly adsorbed reaction species. These results open up a pathway for the design of noble-metal-free solid catalysts for industrial applications.

Full text of DOI:10.1002/cctc.201700095

Accepted Article
Title: Metal-Organic-Framework Mediated Cobalt/N-Doped Carbon
Hybrids as Efficient and Chemoselective Catalysts for the
Hydrogenation of Nitroarenes
Authors: Xiaohui Sun, Alma I Olivos Suarez, Lide Oar Arteta, Elena
Rozhko, Dmitrii Osadchii, Anastasiya Bavykina, Freek
Kapteijn, and Jorge Gascon
This manuscript has been accepted after peer review and appears as an
Accepted Article online prior to editing, proofing, and formal publication
of the final Version of Record (VoR). This work is currently citable by
using the Digital Object Identifier (DOI) given below. The VoR will be
published online in Early View as soon as possible and may be different
to this Accepted Article as a result of editing. Readers should obtain
the VoR from the journal website shown below when it is published
to ensure accuracy of information. The authors are responsible for the
content of this Accepted Article.
To be cited as: ChemCatChem 10.1002/cctc.201700095
Link to VoR: http://dx.doi.org/10.1002/cctc.201700095
A Journal of
www.chemcatchem.org

10.1002/cctc.201700095
ChemCatChem
FULL PAPER
Metal-Organic-Framework Mediated Cobalt/N-Doped Carbon
Hybrids as Efficient and Chemoselective Catalysts for the
Hydrogenation of Nitroarenes
Xiaohui Sun, Alma I. Olivos-Suarez, Lide Oar-Arteta, Elena Rozhko, Dmitrii Osadchii, Anastasiya
Bavykina, Freek Kapteijn, and Jorge Gascon*
Abstract: A Co@N-doped carbon (Co@NC) hybrid was synthesised
via thermal decomposition of the metal organic framework (MOF)
ZIF-67 under N2 atmosphere. These hybrid materials exhibit
outstanding catalytic activity and chemoselectivity for the conversion
of a wide range of substituted nitroarenes to their corresponding
anilines under relatively mild reaction conditions. The high catalytic
performance is attributed to the formation of cobalt nanoparticles
and to the presence of atomically dispersed Co species in close
interaction with nitrogen doped graphene. Both active species are
formed in situ during the pyrolytic transformation of ZIF-67. The
catalysts could be reused in consecutive runs, exhibiting a slightly
lower activity ascribed to blockage of the active sites by strongly
adsorbed reaction species. These results open up a pathway for the
design of noble-metal free solid catalysts for industrial applications.
development of cheaper first-row transition metal-based
hydrogenation catalysts (e.g. Fe, Co and Ni).[1a, 9] However, most
of these supported noble metal-free catalysts are not yet ideal
because of lack of stability, activity and in some cases
reusability. Major problems are aggregation, metal leaching
and/or metal surface modification under reaction conditions.[9-10]
Different strategies have been proposed to improve the
stability of metal nanoparticles, including coating with organic
molecules, polymers or oxides.[11] For instance, cobalt NPs were
reported to be protected against agglomeration and air oxidation
by capping with organic ligands. In general, these coatings
possess poor electron conductivity and can effectively block
charge transfer between catalytic site and reagents, thereby
negatively influencing the catalytic activity. In contrast, graphite-
like carbon coatings with unique chemical, electrical, and
functional properties are capable of protecting transition metal
(e.g. Co, Fe) NPs from agglomeration and re-oxidation by air,[12]
Introduction
while preserving high activities in
a variety of catalytic
reactions.[12-13] Recently, the group of Beller developed a novel
heterogeneous catalyst in which cobalt nanoparticles are
embedded in a N-doped carbon matrix via direct carbonization
of non-volatile Co-amine coordination complexes.[3a] This
cobalt/N-doped carbon hybrid shows outstanding activity in the
Anilines and, in particular their functionalised derivatives, are
key
organic
intermediates
for
manufacturing
dyes,
pharmaceuticals, pigments, and agrochemicals obtained via
reduction of their corresponding nitroarenes.[1] Conventional
Béchamp (using Fe/HCl) or sulfide reduction (with H2S or NaSH
as reducing agents) processes are widely applied to produce
functionalised anilines in industry.[2] However, these non-
catalytic processes generate large amounts of undesirable
waste, resulting in severe environmental problems. Because of
these reasons, the direct catalytic hydrogenation of nitroarenes
has been extensively studied over the last few years.[3]
hydrogenation
of
different
nitroarenes
along
with
chemoselectivity to their substituted anilines. Although cobalt is
claimed to play a crucial role in this system, the reasons behind
such high activity still remain elusive, largely because of the
complexity of the system. Cheng et al. suggested that those
exposed cobalt nanoparticles does not play a role in the
hydrogenation activity, and proposed that the real active sites
are highly dispersed Co-Nx species located at the carbon
matrix.[14] However, Liu et al. indicated no detectable
hydrogenation activity after acid leaching in a Co@C catalyst,
and concluded that the hydrogenation activity is to be attributed
to those accessible metallic cobalt NPs.[15] Similarly, Wang et al.
neither detected any hydrogenation activity after leaching Co
In this spirit, a large number of molecular catalysts have been
developed.[4] However, these homogeneous catalysts suffer
from difficulties in recycling and separation from products.
Therefore, heterogeneous catalysts consisting of supported
metal nanoparticles (NPs) are more attractive. Supported Pt
catalysts have been proposed as alternative and are used in the
direct hydrogenation of nitroarenes.[5] However, despite very
efficient for the activation of nitro groups, Pt is also an excellent
catalyst for the reduction of carbonyl groups and double bonds,
which usually leads to unselective hydrogenations.[6] An
alternative to improve the chemoselectivity of supported Pt
catalysts is to introduce additives, such as PdO or H3PO2, albeit
in most cases at the expense of activity.[6-7] Lack of selective
hydrogenation is not the only associated problem. Side products
such as phenylhydroxylamine derivatives, one type of byproduct
formed with Pt-PbO or Pt-H3PO2 catalysts during the
hydrogenation process, can be explosive even at low
concentrations.[6a] Recently, Corma et. al investigated this
reaction using gold instead of platinum. This less active
hydrogenation metal surprisingly exhibited high activity and
with aqua regia on
a Co@NC catalyst using ZIF-67 as
template.[16] On the other hand, along the already mentioned
issues in the identification of the active sites in Co-based
catalysts, recycling is also confronted with great challenges.
Indeed, most of these cobalt-based catalysts deactivate in a
couple of runs. Chen et al. observed the gradual oxidation of
cobalt in Co@C during the hydrogenation process.[17] However,
Wei and Liu both claimed that oxidized cobalt can be in situ
reduced to its metallic phase by H2 under hydrogenation
conditions.[9a, 15]
In order to further contribute to this discussion, we prepared
N-doped Co@NC composites via one-step pyrolysis of ZIF-67 in
a N2 atmosphere along with an in-depth investigation on the
nature of the active sites and the reasons behind deactivation
during the hydrogenation of different nitroarenes. We
8]
chemoselectivity to the corresponding substituted anilines.[3b,
Along with the use of Au, recent efforts have led to the
This article is protected by copyright. All rights reserved.

10.1002/cctc.201700095
ChemCatChem
FULL PAPER
demonstrate that the pyrolysis temperature and heterogeneity of
metal species play a major role in the challenging hydrogenation
of nitroarenes over Co based catalysts.
decrease due to the lower amount of support after different
pyrolysis temperatures, the structure of the support does not
change, as inferred by the similar SBET and Vpore values obtained
per gram of support.
Figure 1a shows the thermogravimetric (TG) analyses under
air atmosphere of the Co@NC-T samples. All samples present
approximately the same oxidation profile. There are two major
mass losses directly attributed to two carbonaceous species
bearing different thermal stabilities.[22] Despite these similar
profiles, the final residual mass, attributed to Co3O4, increases
with pyrolysis temperature, related to the fact that the carbon
content decreases with pyrolysis temperature. According to AAS
Results and Discussion
Catalyst Characterization Results
ZIF-67 was synthesised by mixing a cobalt nitrate methanolic
solution and MeIm at room temperature. The relative intensity
and peak positions of the powder X-ray diffraction (PXRD)
pattern confirm the formation of pure crystalline ZIF-67 (Figure
S1a).[18] Thermogravimetric (TG) analysis of the synthesized
ZIF-67 under N2 atmosphere suggests that ZIF-67 crystals are
thermally stable up to 480 °C (Figure S1b). The sample weight
then decreases sharply along with temperature increase until
600 °C, which can be ascribed to the decomposition of the
organic linker (MeIm).[19] After that, the sample weight decline
slows down steadily up to 800 °C, implying that the majority of
organic fraction has decomposed below 600 °C.
The specific BET area (SBET) and pore volume (Vpore) of ZIF-
67 and Co@NC-T samples were determined by N2 adsorption-
desorption isotherms (Table S1 and Figure S2). ZIF-67 with a
high SBET of 1930 m2 g-1, displays a type I isotherm with a steep
N2 uptake at low relative pressures, typically associated with
microporosity.[20] After pyrolysis at 600 °C, the SBET of Co@NC-
600 drastically decreases to ~400 m2 g-1, due to the collapse of
the well-defined microporous structure of ZIF-67 during the
(atomic
adsorption
spectroscopy)
analysis,
the
Co
concentrations in Co@NC-T catalysts are 31.0 wt.% 32.9 wt.%
and 37.4 wt.% for Co@NC-600, -700, and -800, respectively
(Table S2). More importantly, the pyrolysis temperature also
impacts the initial oxidation temperature of the carbon matrix in
Co@NC-T samples. The higher the pyrolysis temperature used,
the higher the temperature for carbon oxidation, pointing out to a
higher graphitization degree of the carbon structure. This can be
also inferred from the Raman spectra in Figure 1b. All the
Raman spectra of Co@NC-T reveal the characteristic G and D
bands of carbon at 1580 cm-1 and 1350 cm-1, which are
correlated to graphitic sp2-carbon and disordered or defect
carbon, respectively.[23] When a higher pyrolysis temperature is
employed (i.e. 700 °C), a new band around 2700 cm-1 appears
and sharpens with further increasing pyrolysis temperature to
800 °C. This signal can be assigned to the 2D band of graphitic
carbon.[19a, 24]
carbonization process (note that this area is referred to the
The TEM images of Co@NC-T in Figure 2a-c indicate that
cobalt NPs are uniformly dispersed in the N-doped carbon
matrix. The average particle size of Co NPs in Co@NC-T is
dependent on the pyrolysis temperature (°C), increasing from 8
nm in ‘-600’ to 12 nm and 18 nm for ‘-700’ and ‘-800’,
respectively (Figure 2d-f). Clearly, higher pyrolysis temperatures
induce aggregation of Co to larger NPs.[19a] The high-resolution
(HR) TEM image of Co@NC-T (see insert in Figure 2a-c)
reveals a lattice fringe of 2.05 Å attributed to the (111) plane of
Co0, and confirms that the Co0 NP is embedded in N-doped
carbon matrix and encapsulated by a few graphitic-like carbon
layers, inferring a highly stable catalytic system.[12b] In view of
these results, it can be proposed that Co2+ in the structure of
ZIF-67 is reduced during the pyrolysis process to Co0, and that
the final size of the formed Co nanoparticles can be controlled
by varying the pyrolysis temperature. At the same time, these
NPs facilitate the formation of graphitic carbon layers
surrounding the nanocrystals during the pyrolysis and cooling
down period.[25]
21]
support, as explained above).[19a,
A hysteresis loop also
appears with gradual N2 uptake at a relative pressure from 0.45
to 1.0, caused by the capillary condensation of N2 in mesopores
with a wide size distribution. The microporous volume of
Co@NC-600 is considerably smaller (0.1 cm3 g-1) compared to
the parent ZIF-67 (0.68 cm3 g-1). In contrast, an increase in
mesoporous volume for the pyrolyzed sample from 0.03 cm3 g-1
to 0.14 cm3 g-1 is also observed. The pore structure change
proves that the original microporosity of ZIF-67 was transformed
into an hierarchically interconnected micro/mesoporous structure
during the pyrolysis process. Interestingly, further rising of the
pyrolysis temperature does not significantly influence the SBET
and Vpore of Co@NC. Although the overall SBET and Vpore do
The presence of Co0 and crystalline carbon is also observed
in the XRD patterns shown in Figure 3a. All Co@NC samples
exhibit a diffraction peak at 30.6o, indexed to the (002) planes of
graphitic carbon (JCPDS No. 75-1621)[26]. In addition, the two
peaks at 51.8o and 60.6o indicate the diffraction of the (111) and
(200) planes of metallic cobalt, respectively (JCPDS No. 15-
0806)[27]. The diffraction peaks of both graphite and Co0 become
narrower and sharper as the pyrolysis temperature increases,
pointing to a higher graphitization degree of the carbon matrix
Figure 1. (a) TGA profile of Co@NC-T under air flow (100 ml min-1 STP,
heating rate 5 K min-1), (b) Raman spectra of Co@NC-T.
This article is protected by copyright. All rights reserved.

10.1002/cctc.201700095
ChemCatChem
FULL PAPER
nization process. The pyridinic-N to graphitic-N atomic ratio
(Table S3) decreases from 3.3 for ‘-600’ to 2.4 for ‘-700’ and 1.3
for ‘-800’, demonstrating that relatively more nitrogen atoms
incorporate into the graphitic carbon layers at higher pyrolysis
temperatures. According to elemental analysis, the N contents in
the three catalysts are 7.8 wt.% for Co@NC-600, 3.7 wt.% for
Co@NC-700, and 2.4 wt.% for Co@NC-800 (Table S2),
indicating
a
loss of nitrogen with increasing thermal
decomposition temperature.[3c, 28]
e
f
d
Catalytic Performance in Nitrobenzene Hydrogenation.
The catalytic activity of the Co@NC-T materials was assessed in
the hydrogenation of nitrobenzene (Table 1). In order to
clearly understand the catalytic activity of the Co@NC-T
materials, a series of control experiments were performed.
Neither blank runs (without catalyst) nor the parent ZIF-67
(Table 1, entries 1, 2) showed any conversion, indicating that the
active species should be formed during the pyrolysis process.
Furthermore, the comparable NC sample[30] synthesised using
ZIF-8(Zn) as precursor (Figure S1c) exhibited no activity (Table
1, entry 13), whereas the as-synthesized Co@NC-T catalysts
Figure 2. TEM images and corresponding size distribution of Co NPs in
Co@NC-T: (a, d) 600 °C, (b, e) 700 °C, (c, f) 800 °C.
and to the formation of larger cobalt NPs. Importantly, no cobalt
oxide and/or cobalt carbide reflections are found.
X-ray photoelectron spectroscopy (XPS) was further
employed for all Co@NC-T samples to analyse the bonding
state of nitrogen, as stated in Figure 3b. N 1s signals of
Co@NC-T were deconvoluted into three types of nitrogen
species with binding energies around 398.8 eV, 400.8 eV, and
exhibit good hydrogenation activity together with
a >99%
selectivity to aniline, suggesting that cobalt-containing species
are responsible for the catalytic action in this reaction. More
interestingly, the hydrogenation activity of these Co@NC-T
catalysts (Table 1, entries 3-5) correlates with the pyrolysis
temperature, with conversions after 3 h reaction ranging from 48%
to 69% as the pyrolysis temperature increases from 600 to
800 °C (note that the amount of Co used in all cases was
similar). Given that the SBET and Vpore of Co@NC do not vary
with pyrolysis temperature (Table S1), the different
hydrogenation activity for these Co@NC has to be attributed to:
(i) the different feature of the N-doped carbon matrix, such as
the graphitization degree, that might have an impact on
hydrogen dissociation;[31] (ii) the cobalt speciation or particle size.
The impact of reaction temperature on the catalytic activity for
Co@NC-800 further shows that the conversion of nitrobenzene
is below 15% for a reaction temperature below 70 °C (Table S4).
However, the activity sharply rises with the reaction temperature,
and reaches 99% conversion after 3 h at a reaction temperature
of 110 °C. Moreover, similar turnover numbers (TONs) are
obtained for experiments performed with different initial
concentrations of nitrobenzene on Co@NC-800 (Table S5),
indicating an apparent zeroth order reaction with respect to this
reactant, attributed to a strong adsorption of nitrobenzene on the
catalyst in these experiments.
402.3 eV, which can be attributed to pyridinic-N, graphitic-N
[28]
(pyrrolic and quaternary–N) and oxidized-N,23,
respectively.
The peak at a binding energy of 398.8 eV could also be partially
due to a contribution from nitrogen bound to the metal (Me-N),
due to the small difference between binding energies of N-Me
and pyridinic N.[29] Evidently, the N atoms in the pentagonal ring
of the original imidazole units were mostly converted into two
types of N moieties (pyridinic- and graphitic-N) during the carbo-
The Co@NC-T catalysts were further post-treated in a 0.5 M
HCl solution to remove those accessible cobalt NPs (hence, acid
leached samples are denoted as Co@NC-T (al)). AAS analysis
indicates that cobalt content in the acid leached samples is 27.1
wt.% for Co@NC-600 (al), decreasing to 4.1 wt.% for Co@NC-
900 (al), (Table S2). Apparently, a larger fraction of cobalt can
be leached out by acid when ZIF-67 is carbonized at a higher
temperature. At the same time, XRD analysis (Figure S3 and
Table S6) clearly demonstrates that these acid leached
Co@NC-T (al) samples still contain metallic cobalt nano-
Figure 3. (a) XRD patterns of Co@NC-T, N 1s XPS region of (b) Co@NC-600,
(c) Co@NC-700, (d) Co@NC-800.
This article is protected by copyright. All rights reserved.

10.1002/cctc.201700095
ChemCatChem
FULL PAPER
detectable graphitic-N atoms after removal of some cobalt NPs.
TEM analysis of acid leached samples reveals that some cobalt
NPs together with some hollow carbon shells (Figure 4a) are
present in these samples. HR-TEM (Figure S6a) along with H2
chemisorption results (see experimental section for detail) both
demonstrate that the remaining cobalt NPs are tightly
encapsulated by multiple graphitic carbon shells which render
them inaccessible, even to H2. Figure 4b shows a high-angle
annular dark-field (HAADF) image of the Co@NC-800 (al)
sample, in which cobalt nanoparticles can be easily
distinguished. At the same time, a cloud-like structure dispersed
along the carbon support is also noticeable. Although no cobalt
nanoparticles can be observed in these regions, cobalt (Figure
S6c) is still detected by energy-dispersive X-ray spectroscopy
(EDXS) analysis (area 2 in Figure 4b). It has been reported that
metal atoms (e.g. Co, Fe) released during high-temperature
pyrolysis are extremely reactive towards other heteroatoms such
as nitrogen to form atomically dispersed metal-Nx species
embedded in the carbon matrix.[34] These species, once formed,
can be very resistant towards acid leaching. To further
demonstrate this claim, we performed additional hydrogenation
experiments in the presence of thiocyanate ions (SCN-), a well-
known inhibitor for metal centres in homogeneous catalysts.[35]
After the addition of SCN-, the hydrogenation activity of Co@NC-
800 (al) decreases from 62% to 39% (Table 1, entry 12). Taking
into account that no accessible cobalt NPs are present in the
acid leached Co@NC-800 (al), it is proposed that the highly
dispersed Co-Nx sites do participate in the hydrogenation
process. The reduced activity can be attributed to a competitive
adsorption on Co-Nx sites between SCN- ions and -NO2 group in
nitrobenzene molecules, as it was reported that the thiocyanate
does not permanently bind to or alter the metal-centred site but
rather interacts competitively.[36] In summary, the superior
catalytic activity of Co@NC-800 can be attributed to the
presence of accessible cobalt nanoparticles and atomically
dispersed Co in close interaction with the nitrogen doped
graphene. Both active species are formed in situ during the
pyrolytic transformation of ZIF-67.
Table 1. Results of the catalytic hydrogenation of nitrobenzene over Co@NC-T
and Co@NC-T (al).[a]
Co
content
(wt. %)
Substrate to
cobalt molar
ratio
X
Entry
Catalyst
TON[c]
(%)[b]
1
2
3
4
5
6
7
8
9
10
11
12[d]
13
14
Blank
ZIF-67
-
-
37
37
37
37
24
35
31
43
60
85
85
-
< 1
< 1
48
57
69
99
59
79
43
38
62
39
< 1
36
-
-
26.0
31.0
32.9
37.4
37.4
32.9
37.4
27.1
19.4
13.6
13.6
-
Co@NC-600
Co@NC-700
Co@NC-800
Co@NC-800
Co@NC-700
Co@NC-800
Co@NC-600 (al)
Co@NC-700 (al)
Co@NC-800 (al)
Co@NC-800 (al)
NC-800 (al)
18
21
24
24
21
24
18
22
53
33
-
Co@NC-900 (al)
4.1
282
101
[a] Reaction conditions: 1 mmol nitrobenzene, 5 mL ethanol, 110 °C, 3 MPa
H2, 3 h, aniline selectivity in all cases > 99% (no byproducts were detected by
GC analysis, carbon balance nears 100% in all cases). [b] Determined by GC
(internal standard: n-hexadecane), X = mol nitrobenzene consumed divided by
total mol nitrobenzene. [c] TON = mol nitrobenzene consumed divided by total
mol cobalt. [d] 0.02 M NaSCN as catalyst inhibitor with SCN- to cobalt molar
ratio of 8.5.
particles with a similar average Co particle size. These Co
nanoparticles exhibit high resistance toward acid leaching,
owing to a well encapsulation by graphitic carbon layers on the
surface.[32] As it is well known, acid leaching under mild
conditions is only able to remove those Co nanoparticles that
are not fully encapsulated by the graphite shells.[3 3 ]
Hydrogenation experiments were further performed with these
leached samples using a similar total amount of catalyst (5.1 mg)
of Co@NC-T (al) and Co@NC-T (Table 1, entry 3, 7-11,14).
Obviously, the Co@NC-T (al) samples exhibit a lower
nitrobenzene conversion compared to their counterparts, albeit
with similar aniline selectivity, suggesting that those leached
cobalt nanoparticles did participate in the hydrogenation reaction.
Interestingly, the Co@NC-T (al) samples (T = 600, 700, 800 °C)
still retain a considerable activity and Co@NC-800 (al) exhibits
the highest nitrobenzene conversion (62%) among the three
Co@NC-T (al) samples, indicating that acid-leaching resistant
cobalt species are able to perform the hydrogenation as well.
Recently, Wang et al., reported a similar strategy to prepare
Co@NC catalyst by pyrolysis of ZIF-67, and claimed that no
hydrogenation activity was observed after acid leaching in aqua
regia.[16] Bearing in mind the superior oxidation capacity of aqua
regia, this is not surprising. Indeed, it has been reported by other
authors that N-doped graphene can be easily destroyed by acid
treatment in concentrated HNO3, one of the main components of
aqua regia.[14] In order to gain more insight into the nature of the
active sites in our Co@NC-800 (al), a series of characterization
measurements were performed. The Raman and C 1s XPS
spectra of Co@NC-800 (al) (Figure S4 and Figure S5a,b) are
similar to those of the non-acid leached sample, suggesting that
a possible carbon surface modification is not responsible for a
change in activity. Besides, a relatively higher graphitic-N
content in Co@NC-800 (al) (Figure S5d) than in Co@NC-800
(Figure 3d) is also observed. Taking into consideration that XPS
is a surface analysis technique, this is presumably due to more
A comparison (Table S8) between the TONs achieved by the
catalysts here reported and literature data reveals that the TON
value of Co@NC-900 (al) obtained for
a single run is
comparable (TON = 101) with the best catalysts reported to date
by Beller et al. (TON = 97)[3a], and larger than those for other Co
based catalysts. For repeated runs with Co@NC-800 and 900(al)
b
a
1
2
30 nm
50 nm
Figure 4. (a) TEM pictures of Co@NC-800 (al), (b) HAADF image of Co@NC-
800 (al).
This article is protected by copyright. All rights reserved.

10.1002/cctc.201700095
ChemCatChem
FULL PAPER
a total TON of 118 and 393 is reached (Table S8), respectively
(vide infra). Furthermore, when productivity is calculated per
gram of solid instead of per amount of Co all our catalysts stand
out. In view of the characterization performed and the fact that
the TON doubles already after leaching (Table 1), it is
reasonable to argue that the presence of inaccessible, i.e.
inactive cobalt NPs which are fully encapsulated by the graphite
layers, result in lower TON values than actually should apply.
Figure 5. Nitrobenzene conversion upon reuse of Co@NC-800 (a). Reaction
conditions: 3 mmol nitrobenzene, substrate to cobalt molar ratio of 73, 5 mL
ethanol, 110 °C, 3 MPa H2, 3 h for 1st - 4th run and 12 h for the 5th run. (b) TGA
curve of Co@NC-800 (fresh) and Co@NC-800 (spent) under air flow (100 ml
min-1 STP, heating rate 5 K min-1).
Reusability and Catalyst Deactivation
Catalyst stability and reusability is a key factor to evaluate
heterogeneous catalysts. In this sense, the most active catalysts,
Co@NC-800 and 900(al), were tested in successive runs. The
yield of aniline decreases gradually after each run (Fig. 5a and
S10), but the aniline selectivity is well preserved in five
consecutive runs (> 99%). In spite of this activity decline, after 4
cycles Co@NC-800 has still half of its activity: extending the
reaction time to 12 h leads to an aniline yield as high as 65%.
Overall a TON of 118 is attained with this catalyst (Table S8).
Loss of the active component is one of causes of catalyst
deactivation and usually results from leaching in liquid phase
reactions. Wei et al. observed that 40 wt.% of cobalt was lost by
the CoOx@NCNT hybrid after 11 runs in nitrobenzene
hydrogenation, and attributed the deactivation to this loss of
cobalt.[9a] However, in our case the negligible amount of cobalt
found in the solution (< 3 ppm, below the ICP detection limit,
Table S7) after each run suggests that the graphite shells
prevent leaching of the cobalt NPs under reaction conditions,
implying that the observed deactivation cannot be attributed to a
loss of cobalt during the hydrogenation. At the same time, both
the XRD pattern and C 1s XPS spectrum of the spent Co@NC-
800 catalyst exhibit the same features as the fresh Co@NC-800
(Figure S7 and Figure S8), indicative of a well preserved
structure after recycling.
lity of all samples (Figure S10). The lower content of Co,and the
smaller fraction of encapsulated nanoparticles suggest that the
highly dispersed Co is associated with the activity in this catalyst
system.
Hydrogenation of Substituted Nitroarenes
To investigate the general applicability of Co@NC-800, various
substituted nitroarenes were explored under standard
conditions. A variety of industrially relevant anilines such as
chloro- and fluoroanilines were produced in high yields and
selectivities (Table 2, entries 1-4). The most-challenging
substrates bearing other easily reducible groups such as iodides,
nitrile, aldehyde, ketone and alkene (Table 2, entries 5-9) could
be converted easily and selectively into the corresponding
anilines. Even a sterically hindered nitroarene (Table 2, entry 10)
was hydrogenated to the substituted aniline with outstanding
yield. These results demonstrate again that this Co@NC-800
catalyst displays an intriguing activity and chemoselectivity in the
hydrogenation of substituted aromatic compounds in general.
TG analysis in air (Figure 5b) indicates a 50 °C lower initial
oxidation temperature of the carbon matrix in fresh than in spent
Co@NC-800. Bearing in mind that the carbon structure in the
fresh and spent Co@NC-800 is the same, the temperature
deviation is ascribed to the number of exposed cobalt sites,
considering that oxidized cobalt species can catalyse carbon
oxidation at low temperature.[37] For the spent Co@NC-800
catalyst, a more intense Co 2p XPS signal from CoOx is
detected compared to the fresh sample (Figure S9). Oxidized
cobalt can be in situ reduced to its metallic phase by H2 under
similar hydrogenation conditions;[9a, 15] therefore the increase of
the CoOx signal in our case might be attributed to the re-
oxidation of cobalt in air after the recycling. A ~3 wt.% lower
residual mass in the spent Co@NC-800 than in the fresh
catalyst is also observed. As no mass decrease due to possible
adsorbed water release is observed below 200 oC, and the delay
in the weight loss, the extra ~3 wt.% mass loss in Co@NC-800
(spent) is ascribed to the removal of some strongly adsorbed
reaction species (of unknown nature) on the catalyst. Indeed,
the presence of strongly adsorbed species would explain the
delay in oxidation of carbon support observed in the TGA
analysis. Co@NC-900 (al) exhibits the highest activity and stabi-
Conclusions
The one-step thermal decomposition of ZIF-67 under a N2
atmosphere is a straightforward approach to synthesize a
scalable, recyclable and active heterogeneous catalyst for the
hydrogenation of nitroarenes. The resulting cobalt containing
carbon composites exhibit an excellent catalytic performance in
the hydrogenation of a wide range of nitroarenes to their
substituted anilines under relatively mild conditions. By
controlling the pyrolysis temperature and with a subsequent acid
leaching treatment, it is possible to control the total metal
loading, resulting in high performance catalysts. The high activity
of the Co-N-C system is attributed to cobalt nanoparticles and
the presence of highly dispersed Co in close interaction with
nitrogen-doped graphene, both of which are formed in situ
during the pyrolytic transformation of ZIF-67. Furthermore, these
atomically dispersed Co-Nx species exhibit a much higher
turnover number (TON) than cobalt nanoparticles, probably due
to its higher cobalt atom utilization in the hydrogenation reaction.
Unfortunately, with this MOF-mediated synthesis method, it is
This article is protected by copyright. All rights reserved.

10.1002/cctc.201700095
ChemCatChem
FULL PAPER
2-Methylimidazole (MeIm, purity 99%), cobalt nitrate hexahydrate
(Co(NO3)2·6H2O, >99%), zinc nitrate hexahydrate (Zn(NO3)2·6H2O, 98%),
hydrochloric acid (37%), and methanol (>99.8%) were purchased from
Sigma-Aldrich Chemical Co. All chemicals were used without further
purification.
Table 1. Results of the chemoselective hydrogenation of substituted
nitroarenes over Co@NC-800 [a]
Entry
Substrate
Product
t (h)
Y (%)[b]
1
4
98
Synthesis of ZIF-67 and ZIF-8.
ZIF-67 precursors were synthesized according to a previous report with
some modification.[38] Typically, 2.933 g Co(NO3)2·6H2O and 6.489 g
MeIm were separately dissolved in 200 mL methanol. The clear linker
solution was rapidly poured into the metal solution under magnetic
stirring and kept at room temperature for 24 h. Afterwards, the bright
purple product was collected by filtration, washed 3 times with methanol,
and dried at 80 °C under vacuum. For the synthesis of ZIF-8, having the
same structure, all the steps are the same except substituting
Co(NO3)2·6H2O with Zn(NO3)2·6H2O.
2
6
98
3
4
6
6
97
97
Preparation of Co@NC, Co@NC (al) and NC. The Co@NC catalysts
were prepared independently but in an almost similar manner as recently
reported by Wang and Li[16] Typically, carbonization of 1 g ZIF-67 was
performed in a ceramic crucible inside a quartz tubular reactor (approx. L
= 1.0 m x ID = 5.0 cm) horizontally situated in a ceramic fiber oven
(Carbolite, Sheffield). The reactor was flushed with nitrogen (150 ml min-1)
at 30 °C for 0.5 h, followed by further carbonization at different
temperature for 8 h under N2 (150 ml min-1) at a ramp of 2 °C min-1.
Before exposure to ambient conditions, these catalysts were passivated
for 2 h at room temperature in a stream of 5 vol% O2 in N2 and denoted
herein as Co@NC-T (T = 600, 700, 800, 900 °C), where T refers as the
pyrolysis temperature. 0.5 g of the ground Co@NC-T samples were
further immersed in 200 mL 0.5 M hydrochloric acid solution for 24 h at
ambient temperature to dissolve the exposed cobalt nanoparticles. The
leaching process was repeated 3 additional times, followed by washing
with deionized water until the pH reached ~7. After drying at 50 °C for 24
h under vacuum, these samples are defined as Co@NC-T (al) (T = 600,
700, 800 °C). For the preparation of NC sample, 1 g ZIF-8 was pyrolyzed
at 800 °C followed by a complete acid leaching step with 0.5 M HCl
solution, using the same procedures as Co@NC-800 (al) sample.
Characterization.
5
6
97
6
4
6
96
96
7c
8
6
98
9c
1
47d
99
10
12
The concentrations of Co and N in the samples were measured by
atomic adsorption spectroscopy (AAS) (AAnalyst 200, Perkin Elmer, USA)
and elemental analysis (Vario EL, Elementar, Germany), respectively.
The Brunauer-Emmett-Teller (BET) areas and porous structure of the
samples were determined using a Micromeritics Tristar 3020 apparatus
after degassing under vacuum overnight at 150 °C in Micromeritics
Vacprep 061 apparatus.[39]
[a] Reaction conditions: 1 mmol nitroarene, substrate to cobalt molar ratio of
24, 5 mL ethanol, 110 °C, 3 MPa H2, > 99% conversion observed in all
cases. [b] Determined by GC (internal standard: n-hexadecane). [c]
substrate to cobalt molar ratio of 12. [d] 1 h reaction with 53% conversion.
Competing reaction is the double bond hydrogenation[15]
impossible to investigate the differences in N-doped carbon
structure and cobalt particle size individually on the
hydrogenation performance, and a more in-depth work on these
two factors is under investigation. Furthermore, the mild
deactivation observed in the Co-N-C catalysts can be
attributedto the partial blockage of active sites by strongly
adsorbed reaction species.
The presented synthetic approach opens up an avenue to
develop and improve powerful non-noble metal catalysts for
industrial applications.
Thermogravimetric (TG) analysis was carried out using a Mettler Toledo
TGA/SDTA851e instrument. The experiments were performed from room
temperature to 800 °C with
continuous flow of N2 or air (100 ml min-1).
a ramping rate of 5
°C min−1 under
Raman spectra were obtained with a commercial Renishaw inVia reflex
confocal microscope using a 532 nm laser. Measurements were carried
out in samples without any pre-treatment and exposed to normal
conditions of air and humidity at room temperature. Spectra analysis was
carried with WIRE4.1 software after subtracting the baseline.
X-ray diffraction (XRD) patterns were recorded in Bragg-Brentano
geometry in Bruker D8 Advance X-ray diffractometer equipped with a
Vantec position sensitive detector and graphite monochromator.
Measurements were performed at room temperature with monochromatic
Co Kα radiation (λ = 0.179026 nm) in the 2θ range (10°-90°).
Experimental Section
Materials.
This article is protected by copyright. All rights reserved.

10.1002/cctc.201700095
ChemCatChem
FULL PAPER
[1]
aD. Cantillo, M. Baghbanzadeh, C. O. Kappe, Angewandte
Chemie International Edition 2012, 51, 10190-10193; bR. S.
Downing, P. J. Kunkeler, H. van Bekkum, Catalysis Today
1997, 37, 121-136; cN. Ono, The Nitro Group in Organic
Synthesis, Wiley-VCH, New York, 2001.
Transmission electron microscopy (TEM) and high resolution TEM (HR-
TEM) images were obtained by using a Talos F200X microscope (FEI,
Hillsboro, OR, USA) at 200 kV.
XPS measurements were performed on
a K-alpha Thermo Fisher
[2]
[3]
S. Nishimura, Handbook of Heterogeneous Catalytic Hydro-
genation for Organic Synthesis, Wiley-Interscience, New York,
2001.
aF. A. Westerhaus, R. V. Jagadeesh, G. Wienhöfer, M.-M.
Pohl, J. Radnik, A.-E. Surkus, J. Rabeah, K. Junge, H. Junge,
M. Nielsen, A. Brückner, M. Beller, Nat Chem 2013, 5, 537-
543; bA. Corma, P. Serna, Science 2006, 313, 332-334; cR. V.
Jagadeesh, A.-E. Surkus, H. Junge, M.-M. Pohl, J. Radnik, J.
Rabeah, H. Huan, V. Schünemann, A. Brückner, M. Beller,
Science 2013, 342, 1073-1076.
aF. L'Eplattenier, P. Matthys, F. Calderazzo, Inorganic
Chemistry 1970, 9, 342-345; bA. Corma, C. González-
Arellano, M. Iglesias, F. Sánchez, Applied Catalysis A:
General 2009, 356, 99-102; cH.-U. Blaser, H. Steiner, M.
Studer, ChemCatChem 2009, 1, 210-221; dK. Junge, K.
Schroder, M. Beller, Chemical Communications 2011, 47,
4849-4859.
aG. Vilé, N. Almora-Barrios, N. López, J. Pérez-Ramírez, ACS
Catalysis 2015, 5, 3767-3778; bA. Yarulin, C. Berguerand, I.
Yuranov, F. Cárdenas-Lizana, I. Prokopyeva, L. Kiwi-Minsker,
Journal of Catalysis 2015, 321, 7-12; cM. Carrus, M.
Fantauzzi, F. Riboni, M. Makosch, A. Rossi, E. Selli, J. A. van
Bokhoven, Applied Catalysis A: General 2016, 519, 130-138.
aA. Corma, P. Serna, P. Concepción, J. J. Calvino, Journal of
the American Chemical Society 2008, 130, 8748-8753; bG. C.
Bond, Chemical Society Reviews 1991, 20, 441-475.
A. M. Tafesh, J. Weiguny, Chemical Reviews 1996, 96, 2035-
2052.
aA. Grirrane, A. Corma, H. García, Science 2008, 322, 1661-
1664; bM. Boronat, P. Concepción, A. Corma, S. González, F.
Illas, P. Serna, Journal of the American Chemical Society
2007, 129, 16230-16237.
aZ. Wei, J. Wang, S. Mao, D. Su, H. Jin, Y. Wang, F. Xu, H. Li,
Y. Wang, ACS Catalysis 2015, 5, 4783-4789; bO. Beswick, I.
Yuranov, D. T. L. Alexander, L. Kiwi-Minsker, Catalysis Today
2015, 249, 45-51; cR. K. Rai, A. Mahata, S. Mukhopadhyay, S.
Gupta, P.-Z. Li, K. T. Nguyen, Y. Zhao, B. Pathak, S. K. Singh,
Inorganic Chemistry 2014, 53, 2904-2909.
aL. Zhao, J. Chen, J. Zhang, Journal of Molecular Catalysis A:
Chemical 2006, 246, 140-145; bW. Lin, H. Cheng, J. Ming, Y.
Yu, F. Zhao, Journal of Catalysis 2012, 291, 149-154.
aJ. Lee, O. K. Farha, J. Roberts, K. A. Scheidt, S. T. Nguyen,
J. T. Hupp, Chemical Society Reviews 2009, 38, 1450-1459;
bV. K. Mittal, S. Bera, T. Saravanan, S. Sumathi, R. Krishnan,
S. Rangarajan, S. Velmurugan, S. V. Narasimhan, Thin Solid
Films 2009, 517, 1672-1676; cB. V. Appa Rao, M. Yakub Iqbal,
B. Sreedhar, Corrosion Science 2009, 51, 1441-1452; dJ. Lee,
D. H. K. Jackson, T. Li, R. E. Winans, J. A. Dumesic, T. F.
Kuech, G. W. Huber, Energy & Environmental Science 2014,
7, 1657-1660; eY.-Z. Chen, C. Wang, Z.-Y. Wu, Y. Xiong, Q.
Xu, S.-H. Yu, H.-L. Jiang, Advanced Materials 2015, 27, 5010-
5016.
aD. Deng, L. Yu, X. Chen, G. Wang, L. Jin, X. Pan, J. Deng, G.
Sun, X. Bao, Angewandte Chemie International Edition 2013,
52, 371-375; bJ. Deng, P. Ren, D. Deng, L. Yu, F. Yang, X.
Bao, Energy & Environmental Science 2014, 7, 1919-1923; cX.
Zheng, J. Deng, N. Wang, D. Deng, W.-H. Zhang, X. Bao, C.
Li, Angewandte Chemie International Edition 2014, 53, 7023-
7027.
Scientific spectrometer using monochromatic Al Ka radiation at ambient
temperature and chamber pressure of about 10-8 mbar. The spot size
was 400 µm. A flood gun was always used for charge compensation. All
the spectra measured were corrected by setting the reference binding
energy of carbon (1s) at 284.8 eV. The spectra were analyzed and
processed using Thermo Avantage v5.903 software (Thermo Fisher
Scientific). The peaks were fitted using Lorentzian–Gaussian product
function. Smart background (derived from the Shirley background) was
used over the peak width. The binding energy reported is within ± 0.2 eV.
H2 chemisorption was carried out with a Micromeritics ASAP 2020 C
instrument using the method reported by Bartholomew and co-workers.[40]
Typically, the Co@NC-800 (al) sample was first reduced for 3 h in H2 gas
flow at 350 °C, followed by evacuating at that temperature for 30 min.
Then, the temperature was cooled down to 150 °C, at which temperature
the H2 adsorption isotherms were measured. The total H2 uptake at zero
pressure was obtained by extrapolating the linear part of the isotherm.
The amount of chemisorbed hydrogen is negligible on Co@NC-800 (al)
sample.
[4]
[5]
[6]
General Procedure for Hydrogenation of Nitroarenes.
[7]
[8]
The hydrogenation reactions were performed in
a Parr 5000 Multi
Reactor Stirrer System.[41] The reaction vessels have a volume of 45
cm3 with stirring and an internal temperature controller. In a typical
experiment, hydrogenation reactions are carried out in a batch mode, for
which autoclaves are filled with nitrobenzene (1 mmol), an amount of
catalyst (Co@NC-T or Co@NC-T (al)) corresponding to a substrate to
cobalt molar ratio of 37 (unless otherwise stated), internal standard (n-
hexadecane, 0.34 mmol), and 5 ml ethanol as solvent. Before starting the
reaction, the autoclaves are purged 3 times with He to remove air, and
pressurized to 3 MPa H2, followed by heating to 110 °C under stirring at
500 rpm. Further increasing stirring speed did not improve the
hydrogenation activity. After a fixed reaction time, the autoclaves are
cooled down to room temperature and the hydrogen pressure is carefully
released. The resulting reaction mixture is filtered and liquids are
analyzed by GC (Agilent Technologies, GC 6890N). For recycling studies,
6.5 mg Co@NC-800 was added into the reactant mixtures (3 mmol
nitrobenzene, substrate to cobalt molar ratio of 73, 0.34 mmol n-
hexadecane, and 5 ml ethanol) under the same conditions as mentioned
above except using the recovered catalyst. The catalyst was recovered
by filtration, washed three times with ethanol, dried under vacuum at
50 °C for 2 h and then used for the next run without any reactivation or
purification.
[9]
[10]
[11]
[12]
[13]
Acknowledgements
aL. Zhang, A. Wang, W. Wang, Y. Huang, X. Liu, S. Miao, J.
Liu, T. Zhang, ACS Catalysis 2015, 5, 6563-6572; bY. Hou, Z.
Wen, S. Cui, S. Ci, S. Mao, J. Chen, Advanced Functional
Materials 2015, 25, 872-882.
T. Cheng, H. Yu, F. Peng, H. Wang, B. Zhang, D. Su,
Catalysis Science & Technology 2016, 6, 1007-1015.
L. Liu, P. Concepción, A. Corma, Journal of Catalysis 2016,
340, 1-9.
Xiaohui Sun acknowledges support by the China Scholarship
Council (CSC). The authors greatly thank Alla Dikhtiarenko for
the contribution to the design of the graphical abstract for this
paper.
[14]
[15]
[16]
X. Wang, Y. Li, Journal of Molecular Catalysis A: Chemical
2016, 420, 56-65.
Keywords: nitroarene hydrogenation, chemoselectivity, cobalt,
N-doped carbon, deactivation.
This article is protected by copyright. All rights reserved.

10.1002/cctc.201700095
ChemCatChem
FULL PAPER
[17]
[18]
[19]
B. Chen, F. Li, Z. Huang, G. Yuan, ChemCatChem 2016, 8,
1132-1138.
Müllen, Journal of the American Chemical Society 2013, 135,
16002-16005; cG. Wu, C. M. Johnston, N. H. Mack, K.
Artyushkova, M. Ferrandon, M. Nelson, J. S. Lezama-
Pacheco, S. D. Conradson, K. L. More, D. J. Myers, P.
Zelenay, Journal of Materials Chemistry 2011, 21, 11392-
11405.
L. Zhang, Z. Su, F. Jiang, L. Yang, J. Qian, Y. Zhou, W. Li, M.
Hong, Nanoscale 2014, 6, 6590-6602.
R. W. Coughlin, Product R&D 1969, 8, 12-23.
R. Banerjee, A. Phan, B. Wang, C. Knobler, H. Furukawa, M.
O'Keeffe, O. M. Yaghi, Science 2008, 319, 939-943.
aN. L. Torad, M. Hu, S. Ishihara, H. Sukegawa, A. A. Belik, M.
Imura, K. Ariga, Y. Sakka, Y. Yamauchi, Small 2014, 10,
2096-2107; bS. J. Yang, T. Kim, J. H. Im, Y. S. Kim, K. Lee, H.
Jung, C. R. Park, Chemistry of Materials 2012, 24, 464-470.
J. Tang, R. R. Salunkhe, J. Liu, N. L. Torad, M. Imura, S.
Furukawa, Y. Yamauchi, Journal of the American Chemical
Society 2015, 137, 1572-1580.
N. L. Torad, R. R. Salunkhe, Y. Li, H. Hamoudi, M. Imura, Y.
Sakka, C.-C. Hu, Y. Yamauchi, Chemistry – A European
Journal 2014, 20, 7895-7900.
S. Gadipelli, T. Zhao, S. A. Shevlin, Z. Guo, Energy &
Environmental Science 2016.
[30]
[20]
[21]
[31]
[32]
J. Deng, P. Ren, D. Deng, X. Bao, Angewandte Chemie
International Edition 2015, 54, 2100-2104.
[33]
aX. Zou, X. Huang, A. Goswami, R. Silva, B. R. Sathe, E.
Mikmeková, T. Asefa, Angewandte Chemie 2014, 126, 4461-
4465; bL.-B. Lv, T.-N. Ye, L.-H. Gong, K.-X. Wang, J. Su, X.-H.
Li, J.-S. Chen, Chemistry of Materials 2015, 27, 544-549.
W.-J. Jiang, L. Gu, L. Li, Y. Zhang, X. Zhang, L.-J. Zhang, J.-
Q. Wang, J.-S. Hu, Z. Wei, L.-J. Wan, Journal of the American
Chemical Society 2016, 138, 3570-3578.
H.-W. Liang, S. Bruller, R. Dong, J. Zhang, X. Feng, K. Mullen,
Nat Commun 2015, 6.
M. S. Thorum, J. M. Hankett, A. A. Gewirth, The Journal of
Physical Chemistry Letters 2011, 2, 295-298.
J. van Doorn, M. A. Vuurman, P. J. J. Tromp, D. J. Stufkens, J.
A. Moulijn, Fuel Processing Technology 1990, 24, 407-413.
Z. Jiang, Z. Li, Z. Qin, H. Sun, X. Jiao, D. Chen, Nanoscale
2013, 5, 11770-11775.
M. F. De Lange, T. J. H. Vlugt, J. Gascon, F. Kapteijn,
Microporous and Mesoporous Materials 2014, 200, 199-215.
aR. C. Reuel, C. H. Bartholomew, Journal of Catalysis 1984,
85, 63-77; bJ. M. Zowtiak, C. H. Bartholomew, Journal of
Catalysis 1983, 83, 107-120.
[22]
[23]
[24]
[34]
B. Y. Xia, Y. Yan, N. Li, H. B. Wu, X. W. Lou, X. Wang, Nature
Energy 2016, 1, 15006.
D. Yang, A. Velamakanni, G. Bozoklu, S. Park, M. Stoller, R.
D. Piner, S. Stankovich, I. Jung, D. A. Field, C. A. Ventrice Jr,
R. S. Ruoff, Carbon 2009, 47, 145-152.
aJ. Tang, T. Wang, X. Pan, X. Sun, X. Fan, Y. Guo, H. Xue, J.
He, The Journal of Physical Chemistry C 2013, 117, 16896-
16906; bN. A. M. Barakat, B. Kim, S. J. Park, Y. Jo, M.-H.
Jung, H. Y. Kim, Journal of Materials Chemistry 2009, 19,
7371-7378.
Z. Zhuang, S. A. Giles, J. Zheng, G. R. Jenness, S.
Caratzoulas, D. G. Vlachos, Y. Yan, Nature Communications
2016, 7, 10141.
W. Zhou, J. Zhou, Y. Zhou, J. Lu, K. Zhou, L. Yang, Z. Tang, L.
Li, S. Chen, Chemistry of Materials 2015, 27, 2026-2032.
J. R. Pels, F. Kapteijn, J. A. Moulijn, Q. Zhu, K. M. Thomas,
Carbon 1995, 33, 1641-1653.
[35]
[36]
[37]
[38]
[39]
[40]
[25]
[26]
[27]
[28]
[29]
[41]
S. Sartipi, M. J. Valero Romero, E. Rozhko, Z. Que, H. A. Stil,
J. de With, F. Kapteijn, J. Gascon, ChemCatChem 2015, 7,
3243-3247.
aJ. Casanovas, J. M. Ricart, J. Rubio, F. Illas, J. M. Jiménez-
Mateos, Journal of the American Chemical Society 1996, 118,
8071-8076; bH.-W. Liang, W. Wei, Z.-S. Wu, X. Feng, K.
This article is protected by copyright. All rights reserved.

10.1002/cctc.201700095
ChemCatChem
FULL PAPER
Entry for the Table of Contents (Please choose one layout)
Layout 1:
FULL PAPER
Text for Table of Contents
Author(s), Corresponding Author(s)*
Page No. – Page No.
Title
This article is protected by copyright. All rights reserved.