Synthesis of Co-N-C immobilized on carbon nanotubes for ethylbenzene oxidation

Article abstract of DOI:10.1016/j.molcata.2016.09.011

The catalysts, namely noble-metal-free Co-N-C immobilized carbon nanotubes (CNTs), are synthesized via heating nitrogen-rich cobalt tetraphenyl porphyrin (CoTPP) supported on CNTs in N₂ atmosphere. The obtained catalysts have been characterized by BET, Raman, XRD, TEM, HRTEM and XPS. It is found that the synergistic effect between carrier and active sites plays an important role in the catalytic performance of Co-N-C/CNTs for ethylbenzene oxidation. When the mass ratio of CoTPP to CNTs is 0.15, the catalyst exhibits the highest catalytic performance for ethylbenzene oxidation (i.e. 19.9% for ethylbenzene conversion, 72.9% selectivity of acetophenone). It can be attributed to the well-dispersed Co-N-C species and the enhanced interaction chances between substrate and active sites with the introduction of CNTs.

Full text of DOI:10.1016/j.molcata.2016.09.011

Journal of Molecular Catalysis A: Chemical 424 (2016) 276–282
Contents lists available at ScienceDirect
Journal of Molecular Catalysis A: Chemical
journal homepage: www.elsevier.com/locate/molcata
Editor’s choice paper
Synthesis of Co-N-C immobilized on carbon nanotubes for
ethylbenzene oxidation
Yuyuan Qiu, Congqiang Yang, Jia Huo, Zhigang Liu∗
State Key Laboratory of Chemo/Biosensing and Chemometrics, School of Chemistry and Chemical Engineering, Hunan University, Changsha, Hunan 410082,
China
a r t i c l e i n f o
a b s t r a c t
Article history:
Received 25 July 2016
Received in revised form 2 September 2016
Accepted 7 September 2016
Available online 8 September 2016
The catalysts, namely noble-metal-free Co-N-C immobilized carbon nanotubes (CNTs), are synthesized
via heating nitrogen-rich cobalt tetraphenyl porphyrin (CoTPP) supported on CNTs in N2 atmosphere. The
obtained catalysts have been characterized by BET, Raman, XRD, TEM, HRTEM and XPS. It is found that the
synergistic effect between carrier and active sites plays an important role in the catalytic performance
of Co-N-C/CNTs for ethylbenzene oxidation. When the mass ratio of CoTPP to CNTs is 0.15, the catalyst
exhibits the highest catalytic performance for ethylbenzene oxidation (i.e. 19.9% for ethylbenzene con-
version, 72.9% selectivity of acetophenone). It can be attributed to the well-dispersed Co-N-C species and
Keywords:
Cobalt catalysts
Co-N-C
Ethylbenzene oxidation
Cobalt porphyrin
Carbon nanotubes
the enhanced interaction chances between substrate and active sites with the introduction of CNTs.
© 2016 Elsevier B.V. All rights reserved.
1. Introduction
as well as electrocatalytic ORR [13–15]. In addition, the biomass
derived nitrogen-containing carbon in styrene epoxidation [16],
Recently, carbon nanomaterials have attracted great atten-
tion in many fields, such as energy conversion, energy storage,
drug delivery, catalysis, data storage, optoelectronic devices, etc
[1–5]. The applications are attributed to the distinctive proper-
ties of nano-carbon material including the admirable conducting
and semiconducting performance, quite large specific surface
area, good chemical stability, biocompatibility as well as splen-
did photoelectric response [6,7]. In the field of catalysis, graphene,
graphene nanoribbons, nano-diamond, nanotubes, nanofibers and
other types of porous carbon are used as the supports or as the
catalysts of metal-free oxygen reduction reaction (ORR) [8,9]. In
most cases, the original carbon material has limited or even neg-
ligible ORR catalytic activity, the introduction of heteroatom into
carbon seems to be a desirable alternative to solve the problem. The
improved catalytic performance of heteroatom-doped carbon cat-
alysts has been attributed to the charge redistribution adjacent to
the dopants, which lower the activation energy, boost the stability
of the transition state and weaken the O O bond [10,11]. In gen-
eral, nitrogen-doped carbon nanomaterials are proved to be high
efficient catalysts in water splitting [12], water oxidation, selective
oxidation of alcohols, heterogeneous hydrogenation and oxidation
MOFs derived porous carbons in selective aerobic oxidations [17]
and heteroatoms doped single-wall carbon nanohorns in nitroben-
zene reduction [18] are realized. Morever, nitrogen-doped [19],
sulfur-doped [20], phosphorus-doped [21] and boron-doped [22]
carbon nanotubes are synthesized successfully and expressed out-
standing ORR activity, which also shows considerable catalytic
performance in selective oxidation of cycloalkanes and the degra-
dation of organic compounds [23,24]. It is reported that the
structural defects, surface modification and functionalization play
significant roles in the carbon-catalyzed reactions [25,26]. For
instance, carbonyl/quinone groups and surface acid properties have
been proposed to be the active sites for the oxidative dehydrogena-
tion of hydrocarbons and oxidation of alcohols [26,27].
Metal macrocyclic compounds are readily available carbon
and nitrogen-rich precursors to synthesize metal-coordinating
nitrogen-doped carbon materials (M-N-C) via a simple pyrolysis
procedure. In addition, the M-N-C compounds are good substi-
tute for the noble metal catalyst for its low-price, earth-abundant,
environment-friendly and better leaching resistance in electro-
chemical ORR or organic redox [28–31]. Metalloporphyrin is a
decent alternative precursor for the M-N-C owing to its nitrogen-
rich macrocycle and the intrinsic metal-nitrogen bond which acts
as the pivotal factor in the enhancement of catalytic performance
in oxide reduction [31,32]. The importance of carbon support in the
promotion of catalyst efficiency has been widely known. The intro-
∗
Corresponding author.
E-mail address: liuzhigang@hnu.edu.cn (Z. Liu).
http://dx.doi.org/10.1016/j.molcata.2016.09.011
1381-1169/© 2016 Elsevier B.V. All rights reserved.

Y. Qiu et al. / Journal of Molecular Catalysis A: Chemical 424 (2016) 276–282
277
duction of carrier can not only increase the contact surface between
the active sites and reactant, but provide a steady atmosphere for
the active species which could delay the aggregation, slow the mor-
phology change and keep its original properties [33,34].
the mixture in batches and kept for 1 h. After refrigerated overnight,
the resulting product was filtered and washed with deionized water
for several times. The end product was obtained and denoted as
CoTPP after drying.
The transformation of less expensive alkanes to high value-
added chemicals remains a significant task in the current industrial
and fine-chemical processes. For instance, the selectivity oxida-
tion of alcohols to their corresponding aldehydes or ketones, the
oxidative dehydrogenation of cyclohexane to cyclohexene, the aer-
obic oxidation of ethylbenzene and the like are still on the way for
further optimization [35]. Acetophenone is an important interme-
diate in medicine, resin, flavouring agent, perfumes and esters, and
also used as a solvent for cellulose esters and plasticizers in plas-
tics [36–38]. Traditionally, the synthesis of acetophenone utilizes
Friedel-Crafts acylation of benzene using acyl halides or acid anhy-
drides in the existence of Lewis acids, or utilizes the oxidation of
alkylarenes with inorganic metal salt such as permanganate and
dichromate [39,40]. There is no doubt that the traditional prepa-
ration process will cause a mass of wastes and pollution including
toxic effluent, offscum, etc.
Herein, considering the low catalytic activity of original CNTs,
the combination of the advantages of CNTs and metal macrocyclic
compound is adopted. Besides, the complex and tedious covalently
bonded procedure is successfully replaced by a facile impregnation
method in which the loading amount of M-N-C is easily con-
trollable [41]. After that, a high-performance catalyst is obtained
by pyrolyzing cobalt tetraphenyl porphyrin coated multi-walled
carbon nanotubes in N2 atmosphere and the catalytic activity of
Co-N-C is greatly improved (Scheme 1).
2.1.2. Synthesis of Co-N-C/CNTs
Commercial carboxylated multi-walled carbon nanotubes
(CNTs) were dried in a vacuum oven overnight without any other
purification. Accordingly, 0.01 g CoTPP (0.0149 mmol) was dis-
solved with 2 mL dichloromethane, then 0.1 g CNTs was added to
the solution and sonicated for 30 min to disperse them. Following
that, the paste was placed in vacuum oven overnight. The obtained
product was transferred to a quartz boat and heated to 500 ◦C with
heating rate 10 ◦C per minute in nitrogen atmosphere and kept at
500 ◦C for 3 h. The final products were named as Co-N-C/CNTsx in
which “x” represented the mass ratio of CoTPP to CNTs.
2.2. Characterization of catalysts
The surface area and pore volume of the samples were char-
acterized by nitrogen adsorption/desorption analysis instrument
at liquid nitrogen temperature on Novav1000e apparatus from
Quantachrome Instrument. The Brunauer-Emmett-Teller (BET),
Barret-Joyner-Halenda (BJH) and Dubinin-Astakhov (DA) equation
were applied to calculate the surface area, pore size distri-
bution and total/meso/micropore volume. Raman spectra were
obtained in a Labram–010 micro Raman spectrometer with an
excitation wavelength at 632 nm with 3 cm−1 spectral resolution.
Transmission electron microscopy (TEM) and high resolution trans-
mission electron microscopy (HRTEM) images were collected on
JEM-3010 ultra-high resolution transmission electron microscopy.
X-ray diffraction (XRD) patterns were conducted on X-ray pow-
der diffractometer XRD-6100 (Shimadzu). The X-ray photoelectron
spectroscopy (XPS) was performed on a RBD upgraded PHI-5000C
ESCA system (Perkin Elmer) with Mg K␣ radiation (hꢀ = 1253.6 eV)
or Al K␣ radiation (hꢀ = 1486.6 eV). Binding energies were cali-
brated with the reference of C 1s peak at 284.5 ev.
2. Experimental
2.1. Preparation of catalysts
2.1.1. Synthesis of cobalt (II) tetraphenyl porphyrin (CoTPP)
Cobalt tetraphenyl porphyrin was synthesized as the litera-
ture reported previously [42]. During a typical synthesis process,
0.1 mol benzaldehyde was added into a three-neck flask contain-
ing 250 mL propanoic acid. Then the mixture was heated to the
target temperature (130 ◦C) slowly and 0.1 mol freshly distilled
pyrrole was added dropwise. After stirring for another 1 h, the
obtained product was refrigerated overnight, then filtered and puri-
fied. 1.6 mmol pure trtraphenyl porphyrin was transferred into
100 mL N,N-dimethylformamide (DMF) solution, heating to reflux
subsequently. 0.01 mol cobalt chloride hexahydrate was added to
2.3. Catalytic performance
The selective oxidation of ethylbenzene was carried out at a
50 mL Teflon-lined stainless steel autoclave. Typically, 10 mL of
ethylbenzene (81.7 mmol) and 30 mg of catalyst was added to the
reactor and then sealed with the O2 pressure raised to 0.8 MPa. Sub-
Scheme 1. Schematic diagram showing the preparation procedure of Co-N-C decorated CNTs by thermal treatment using CoTPP and commercial available CNTs.

278
Y. Qiu et al. / Journal of Molecular Catalysis A: Chemical 424 (2016) 276–282
respectively, while the total pore volume and micropore volume of
Co-N-C0.3 is 0.095 cm3/g and 0.074 cm3/g. It is clearly showing that
the decrease of specific surface area, total pore volume as well as
the micropore volume are associated with the amount of Co-N-C
in Co-N-Cx/CNTs catalysts. The results can be explained that the
intrinsic pores and the aggregation pores are partially occupied by
Co-N-C (Table 1).
3.2. TEM
The morphology of prepared catalysts is investigated using TEM
and HRTEM. The TEM images show that both Co0.15/CNTs and Co-
N-C0.15/CNTs retain the hollow tubular structure similar to the raw
CNTs, indicating that the loading of cobalt and cobalt porphyrin
cause no damage to the morphology of raw CNTs during the ther-
mal treatment. Although Co is undoubtedly detected in XPS analysis
(see Fig. 5.), no uniformly distributed and clearly visible large-scale
aggregation of Co-N-C species can be found in the TEM images.
This indicates that cobalt is highly dispersed or poorly crystalline
[44,45], which is consistent with the X-ray diffraction (XRD) peaks
(see Fig. 3.). The uniform distribution of the active sites results
from the cooperative effect of the multiple ␲-stacking interactions
between the porphyrins and the nanotube as well as the dispersion
effect produced by ultrasonic vibrations, which is favorable for the
formation of smaller nanoparticles [46].
Fig. 1. Nitrogen adsorption/desorption isotherms of the prepared catalysts, includ-
ing CNTs, Co-N-Cx/CNTs (X = 0.05, 0.1, 0.15, 0.2, 0.3; Co-N-Cx/CNTs are abbreviated
as CoNCx/CNTs in all of the figures).
sequently, the reaction system was heated to 120 ◦C and kept for
5 h.
The final product was analyzed by gas chromatography
(ShimadzuGC-2014 equipped with a capillary column (RTX-5, 30 m,
ꢁ0.25 mm)) with internal standard method using bromobenzene
and 1,4-dichlorobenzene as reference. The catalysts were recov-
ered by simple centrifugation and washed with ethanol for several
times, then dried in air at 80 ◦C.
3.3. XRD
3. Results and discussion
Fig. 3 displays the XRD patterns of pristine CNTs, Co-N-C and
Co-N-C supported on CNTs. The crystalline structures of the cata-
lysts can be revealed and the chemical valence state information
can be presented further more. The strong diffraction peak at
26.0◦ appeared on the three samples can be assigned to the (002)
planes of graphitic carbon, which illustrates that the graphite struc-
ture of CNTs is retained in the presence of CoTPP. Besides, the
graphite carbon of Co-N-C is formed in the pyrolysis process. Gen-
erally, the weak peaks at 42.8◦, 44.4◦ and 53.5◦ can classified to
the graphitic carbon, which are similar to the commercial CNTs
[47]. The Co-N-C sample shows three characteristic peaks at 44.2◦,
51.5◦, 75.9◦ indexed to (111) (200) and (220) crystalline planes
of the face-centered cubic structure of metallic cobalt (PDF#15-
0806), indicating the existence of Co0. However, no similar typical
diffraction peaks of metallic cobalt or cobalt oxide was detected in
Co-N-C0.15/CNTs compared with the XRD spectra of pure Co-N-C
and CNTs, which demonstrates that the cobalt species are highly
dispersed on the CNTs since the weak signal could not be caught
by the XRD analysis owing to its limited sensitivity. As shown in
Fig. 2., there are no observable metal particles according to the TEM
images. In addition, a weak peak at 41.6◦ was noticed, the Co2N
(002) compound possesses the same diffraction peak in the posi-
3.1. BET
To investigate the specific surface area, pore size distri-
bution and the pore volume of the catalysts, N2 adsorption-
desorption measurement is conducted. The N2 adsorption-
desorption isotherms of Co-N-C/CNTs samples, presented in Fig. 1.,
belong to the type IV adsorption-desorption isotherm with a dis-
tinct hysteresis loops at high relative pressure nearly 0.9 (P/P0),
clarifying the existence of mesopores [43]. The obscure hystere-
sis loops at the medium relative pressure (P/P0) can be ascribed
to the weak capillary condensation in mesopores. The isotherms
of raw CNTs and the Co-N-Cx coated CNTs are similar in the over-
all trend, the break points for example, which mean no obvious
difference of the morphology of CNTs after modification. N2 adsorp-
tion amount at P/P0 decreased clearly with the increasing amount
of Co-N-C compounds, suggesting that part of the micropores and
mesopores are blocked because of the addition of Co-N-C. The spe-
cific surface area decreased from 112.9 m2/g to 100.6 m2/g along
with CoTPP increasing from zero to thirty percent (the mass ratio
of CoTPP to CNTs). Likewise, pristine CNTs possess the largest total
pore volume and micropore volume of 0.145 cm3/g and 0.104 cm3/g
Table 1
BET surface area of CNTs and Co-N-Cx/CNTs.
b
c
d
Catalyst
BET surface areaa [m2/g]
VTotal [cm3/g]
VMicro [cm3/g]
VMeso [cm3/g]
Co-N-C
–
–
–
–
CNTs
112.9
110.4
110.5
108.0
104.7
100.6
0.5377
0.4766
0.4716
0.3704
0.4088
0.3244
0.145
0.137
0.131
0.119
0.115
0.095
0.3927
0.3396
0.3406
0.2514
0.2938
0.2294
Co-N-C0.05/CNTs
Co-N-C0.1/CNTs
Co-N-C0.15/CNTs
Co-N-C0.2/CNTs
Co-N-C0.3/CNTs
a
The surface area was caculated by the BET method.
Total pore volume was estimated at a relative pressure P/P0 = 0.98.
The micropore volume was determined from the Dubinin-Astakhov (DA) equation.
The mesopore volume was obtained from the subtraction of micropore volume from total pore volume.
b
c
d

Y. Qiu et al. / Journal of Molecular Catalysis A: Chemical 424 (2016) 276–282
279
Fig. 2. TEM images of (a) Co/CNTs, (b) Co-N-C0.15/CNTs, (c) Co-N-C0.15/CNTs-R6.
Table 2
The ratio analysis of the peaks in XPS spectra of Co-N-C, Co-N-C0.15/CNTs, Co-N-
C
0.3/CNTs.
Samples
C(at%) O(at%) Co(at%) N(at%) Total N%
N1(%) N2(%) N3(%)
Co-N-C
87.93 6.37
0.60
0.17
0.26
5.10
1.71
1.48
0.662 0.245 0.093
0.442 0.358 0.199
0.435 0.374 0.191
Co-N-C0.15/CNTs 95.92 2.20
Co-N-C0.3/CNTs 95.21 3.06
higher ID/IG ratio of CoTPP/CNTs to pristine CNTs shows the success-
ful coating of CNTs with cobalt porphyrin. Furthermore, the ID/IG
ratio decreased after calcination, indicating the carbonization of
cobalt porphyrin and the formation of Co-N-C structure. However,
the ID/IG ratio of Co-N-C/CNTs increased to a higher degree after six
cycles. This is in agreement with the catalytic activity loss, which
may be caused by the corrosion of carbon matrix due to its par-
tial peeling by a large amount of the evolved oxygen in the oxide
reduction reactions [48].
Fig. 3. XRD patterns of Co-N-C, CNTs and Co-N-C0.15/CNTs.
3.5. XPS
X-ray photoelectron spectroscopy (XPS) measurements are uti-
lized to determine the elemental compositions and valence states
of the Co-N-Cx/CNTs hybrids. As depicted in Fig. 5., the elements
of cobalt, oxygen, nitrogen and carbon are detected and listed in
Table 2 (see Fig. S3. for survey XPS results). Clearly, the cobalt
species possess the highest content (0.60%) in pure Co-N-C com-
pounds compared to Co-N-C0.15/CNTs (0.17%) and Co-N-C0.30/CNTs
(0.26%), which is in expectation. Besides, it is remarkable that the
elemental percentage of N in Co-N-C0.15/CNTs (1.71%) and Co-N-
C0.3/CNTs (1.48%) is lower than that in Co-N-C (5.10%), indicating
that the Co-N-C species is well separated. In order to obtain more
detailed information, such as distribution of different valence states
and chemical combination of the single element, the high reso-
lution XPS is carried out. The high-resolution N 1s spectrum of
Co-N-C and Co-N-Cx/CNTs can be fitted into three peaks with bind-
ing energy at 398.6ev, 400.1ev and 401.2 ev, which can correspond
to pyridinic, pyrrolic and graphitic N, respectively [49,50]. The rel-
ative content of the three types of nitrogen is given in Table 2, N1
refers to pyridinic N, N2, N3 correspond to pyrrolic and graphitic N,
respectively. The peaks of Co 2p1/2 and Co 2p3/2 locate at ∼796 ev
and ∼780 ev clearly presenting in the XPS spectrum. After decon-
volution, it can be noticed that the peak of Co0 locates at 778.5 ev
and that of Co3+ locates at 779.8 ev, the higher binding energy at
781.3 ev corresponds to Co2+, according to Zhu et al. [51]. The broad
peak at 786.1 ev and 804.2 ev can be attributed to the shake-up
satellite peaks [52].
Fig. 4. Raman spectra of raw CNTs, CoTPP0.15/CNTs, Co-N-C0.15/CNTs and Co-N-
C0.15/CNTs-R6 (abbreviated as R6, the catalyst was reused for 6 times).
tion (PDF#06-0647), which could be attributed to the formation of
Co-N bond in the pyrolysis.
3.4. Raman
Raman spectroscopy is widely used to measure CNTs. The
Raman spectra of the samples are listed in Fig. 4. The Raman spec-
trum of pristine CNTs clearly shows the peaks at 1322 cm−1 and
1571 cm−1, which corresponds to the D and G bands, respectively.
The band at 1600 cm−1, termed D’ band, appears as an inconspic-
uous peak on the shoulder of G band. The D band is related to the
disorderly structure of carbon nanotubes, which includes amor-
phous carbon in the CNTs, and the defects in graphene lattice.
Instead, the G band corresponds to the ideal sp2 hybridized car-
bon atom, showing the integrity of graphene layer, while the D’
band can be attributed to the irregular d002 spacing as the literature
reported [43]. The intensity ratio of D-band to G-band, termed ID/IG,
is used to estimate the defects of carbon-based nanostructures. The
3.6. Catalytic performance of Co-N-C/CNTs for ethylbenzene
oxidation
The selective oxidation of ethylbenzene to acetophenone is an
important reaction for its pivotal position in industry, and there-

280
Y. Qiu et al. / Journal of Molecular Catalysis A: Chemical 424 (2016) 276–282
Fig. 5. XPS spectra of the samples. (a) (c) (e) N1s spectra of Co-N-C0.3/CNTs, Co-N-C0.15/CNTs and Co-N-C. (b) (d) (e) Co2p spectra of Co-N-C0.3/CNTs, Co-N-C0.15/CNTs and
Co-N-C, respectively.
fore the transformation of ethylbenzene to acetophenone is used to
evaluate the catalytic performance of the as prepared catalysts. The
catalytic reaction is conducted at 120 ◦C and 0.8 MPa in the presence
or absence of catalyst with the molecular oxygen as final oxidant.
The main products of ethylbenzene oxidation in this reaction sys-
tem are acetophenone, benzaldehyde and 1-phenethyl alcohol.
Catalytic performances of various catalysts, including Co-N-
C, CNTs and Co-N-Cx/CNTs are summarized in Table 3. A blank
experiment indicated that the noncatalytic autoxidation of ethyl-
benzene gave 8.3% conversion (Table 3, entry 1). The conversion is
improved to 10.9% in the addition of original CNTs under the same
conditions, close to Co-N-C, which owes to its intrinsic catalytic
ability. As previously reported, the carbon skeleton and the oxygen-
containing groups, such as carbonyl and quinone, are proposed as
the active sites [26,27,53]. Although pure Co-N-C has higher metal
and nitrogen content, the fairly low surface area might impede the
access to reactants. As expected, the introducing of CNTs signifi-
cant enhances the conversion of ethylbenzene (Table 3, entry 7).
The conversion of ethylbenzene reaches to 19.9%, almost twice as
the pure Co-N-C.
the Co-N-C content was further investigated. The catalytic perfor-
mance increases with the amount of CoTPP increasing from 0.01 to
0.15 (Table 3, entries 4 to 7). Undoubtedly, the mass ratio of 0.15
is a turning point, after which the performance goes down with
the increasing of CoTPP. Co-N-C0.15/CNT exhibits the highest con-
version of ethylbenzene, namely 19.9%. Below the turning point,
the increasing performance originates from the increasing of active
centers. While the density of the active centers exceed the appro-
priate saturation point, the excessive active centers may aggregate
and form other less active compounds which should be responsible
for the decrease of catalytic performance.
Recent studies have shown that a certain amount of iron could
improve significantly the catalytic property of carbon-based cat-
alyst [54]. In particular, Co (Fe)-N-C catalysts, which have got
intensive studies for the electrochemical ORR [29,30,55], were
recently reported to be effective for various organic redox trans-
formations [56,57]. Nevertheless, the nature of active sites has
not been clearly identified yet. It’s still ambiguous what the role
of metal/metal oxide participates in the catalysis, as it could be
encapsulated by carbon shell. Trying to figure out the problem,
Co/CNTs and NC/CNTs are synthesized in the absence of por-
phyrin or cobalt respectively. The amount of cobalt or porphyrin
As mentioned above, Co-N-Cx/CNTs display high activity in the
oxidation of ethylbenzene. In order to achieve an optimal catalyst,

Y. Qiu et al. / Journal of Molecular Catalysis A: Chemical 424 (2016) 276–282
281
Table 3
Catalytic performance of CNTs, Co-N-C, Co-N-Cx/CNTs, Co/CNTs and NC/CNTs in ethylbenzene oxidationa.
Entry
Catalyst
Conversion (%)
Selectivity (%)
Acetophenone
Phenethyl alcohol
Benzaldehyde
1
2
3
4
5
6
7
8
Blank
8.3
72.6
73.9
73.4
74.4
74.7
72.9
72.9
73.3
71.5
69.1
72.6
74.7
74.1
73.5
18.6
21.3
19.0
19.0
20.5
22.5
23.4
22.9
24.7
26.4
19.7
17.2
19.8
19.0
8.8
4.9
7.6
6.6
4.8
4.6
3.7
3.8
3.8
4.5
7.7
8.1
6.1
7.4
Co-N-C
11.3
10.9
12.5
16.2
17.8
19.9
17.5
15.2
14.5
11.3
8.9
CNTsR1
Co-N-C0.01/CNTs
Co-N-C0.05/CNTs
Co-N-C0.1/CNTs
Co-N-C0.15/CNTs
Co-N-C0.2/CNTs
Co-N-C0.3/CNTs
Co/CNTs
9
10
11
12
13
14
NC/CNTs
CNTs R2
Co-N-C0.15/CNTsR2
Co-N-C0.15/CNTsR3
15.0
12.5
a
Reaction conditions: ethylbenzene 10 mL, catalyst 30 mg, O2 pressure 0.8 MPa, reaction time 5 h, temperature 120 ◦C.
in Co/CNTs and NC/CNTs is equal to Co-N-C0.15/CNTs before pyrol-
ysis. Unsurprisingly, neither cobalt nor NC (compounds pyrolyzed
from porphyrin) makes evident contribution to the catalytic per-
formance in ethylbenzene transformation. As given in entry 10 and
11, the conversion of ethylbenzene is 14.5% for Co/CNTs and 11.3%
for NC/CNTs, which are close to Co-N-C0.05/CNTs and pristine CNTs,
respectively.
The efficient mass transfer among the CNT bundles is a signif-
icant factor in the reaction, which is dominated by the dispersion
degree of CNT bundles [58]. Although CNTs with large surface area
and excellent charge transport properties are ideal catalyst carrier,
much of the CNTs surface area and the associated catalyst sites are
blocked or lost because of restacking via the strong ␲-␲ interaction.
Conversion is 10.9% for the fresh CNTs and 8.9% for the recovered
CNTs (Table 3, entries 3,12), indicating that its catalytic perfor-
mance decreases in reuse, which results from the poor dispersion
in organic solvents, leading to the unfavorable aggregation in the
experiment process. When the catalyst recovered from the reac-
tion by filtration and dried in vacuum, the stable aggregation state
is kept, which is not easy to regain the catalytic performance com-
pared with the fresh CNTs. Not only that, the unstable combination
between the active sites and the carrier leads to the leaching of
the metal species in liquid reaction as well as the post-process in
recycling, which should be responsible for the loss of activity [59].
Key Laboratory of Mineralogy and Metallogeny in Chinese Academy
of Sciences (No. KLMM20150103).
Appendix A. Supplementary data
Supplementary data associated with this article can be found, in
the online version, at http://dx.doi.org/10.1016/j.molcata.2016.09.
011.
References
[1] M.F.L. De Volder, S.H. Tawfick, R.H. Baughman, A. John Hart, Science 339
(2013) 535–539.
[2] C. Liu, F. Li, L. Ma, H. Cheng, Adv. Mater. 22 (2010) 28–62.
[3] D. Jariwala, V.K. Sangwan, L.J. Lauhon, T.J. Marks, M.C. Hersam, Chem. Soc.
Rev. 42 (2013) 2824–2860.
[4] J.B. Cui, R. Sordan, M. Burghard, K. Kern, Appl. Phys. Lett. 81 (2002) 3260–3262.
[5] Z. Liu, J.T. Robinson, S.M. Tabakman, K. Yang, H. Dai, Mater. Today 14 (2011)
316–323.
[6] Z. Li, Z. Liu, H. Sun, C. Gao, Chem. Rev. 115 (2015) 7046–7117.
[7] E. Cruz-Silva, F. Lopez-Urias, E. Munoz-Sandoval, B.G. Sumpter, H. Terrones,
J.C. Charlier, V.M, M. Terrones, Nanoscale 3 (2011) 1008–1013.
[8] M.E.M. Buan, N. Muthuswamy, J.C. Walmsley, D. Chen, M. Rønning, Carbon
101 (2016) 191–202.
[9] J. Wu, Z. Yang, X. Li, Q. Sun, C. Jin, P. Strasser, R. Yang, J. Mater. Chem. A 1
(2013) 9889–9896.
[10] J. Zhang, L. Dai, ACS Catal. 5 (2015) 7244–7253.
[11] K. Gong, F. Du, Z. Xia, M. Durstock, L. Dai, Science 323 (2009) 760–764.
[12] Y. Zhao, R. Nakamura, K. Kamiya, S. Nakanishi, K. Hashimoto, Nat. Commun. 4
(2014) 2390–2396.
[13] X. Li, Y. Fang, S. Zhao, J. Wu, F. Li, M. Tian, X. Long, J. Jin, J. Ma, J. Mater. Chem. A
4 (2016) 13133–13141.
4. Conclusions
[14] J. Long, X. Xie, J. Xu, Q. Gu, L. Chen, X. Wang, ACS Catal. 2 (2012) 622–631.
[15] M. Li, F. Xu, H. Li, Y. Wang, Catal. Sci. Technol. 6 (2016) 3670–3693.
[16] Y. Gao, X. Chen, J. Zhang, N. Yan, ChemPlusChem 80 (2015) 1556–1564.
[17] X. Wang, Y. Li, J. Mater. Chem. A 4 (2016) 5247–5257.
[18] X. Wu, L. Cui, P. Tang, Z. Hu, D. Ma, Z. Shi, Chem. Commun. 52 (2016)
5391–5393.
[19] H. Yu, F. Peng, J. Tan, X. Hu, H. Wang, J. Yang, W. Zheng, Angew. Chem. 123
(2011) 4064–4068.
[20] W. Li, D. Yang, H. Chen, Y. Gao, H. Li, Electrochimica. Acta 165 (2015) 191–197.
[21] Q. Li, F. Yuan, C. Yan, J. Zhu, J. Sun, Y. Wang, J. Ren, X. She, RSC Adv. 6 (2016)
33205–33211.
In summary, Co-N-Cx/CNTs displays remarkable catalytic
performance for selectivity oxidation of ethylbenzene. The intro-
duction of CNTs facilitates the mass transfer of O2 and substrate
between active sites, which promotes the catalytic process and
enhances the catalytic performance. When the loading ratio of
CoTPP to CNTs is 0.15, the catalyst Co-N-C0.15/CNTs shows the
highest catalytic activity than others for the selective oxidation
of ethylbenzene, which demonstrates the suitable distributions of
active sites is needed as well as the highly dispersion degree. Based
on a series of control experiments and extensive characterizations,
the CoNx compounds have been proved to be the high performance
active species among others.
[22] L. Yang, S. Jiang, Y. Zhao, L. Zhu, S. Chen, X. Wang, Q. Wu, J. Ma, Y. Ma, Z. Hu,
Angew. Chem. 123 (2011) 7270–7273.
[23] Y. Cao, H. Yu, F. Peng, H. Wang, ACS Catal. 4 (2014) 1617–1625.
[24] D.F.M. Santos, O.S.G.P. Soares, A.M.T. Silva, J.L. Figueiredo, M. Fernando, R.
Pereira, Appl. Catal. B: Environ. 199 (2016) 361–371.
[25] X. Wan, C. Zhou, J. Chen, W. Deng, Q. Zhang, Y. Yang, Y. Wang, ACS Catal. 4
(2014) 2175–2185.
[26] G. Wen, J. Diao, S. Wu, W. Yang, R. Schlogl, D. Su, ACS Catal. 5 (2015)
3600–3608.
[27] J. Luo, H. Yu, H. Wang, H. Wang, F. Peng, Chem. Eng. J. 240 (2014) 434–442.
[28] R. Zhou, M. Jaroniec, S. Qiao, ChemCatChem 7 (2015) 3808–3817.
[29] A. Shaabani, S. Keshipour, M. Hamidzad, S. Shaabani, J. Mol. Catal. A: Chem.
395 (2014) 494–499.
Acknowledgements
This work was supported by the National Natural Science Foun-
dation of China (No. 21103045, 1210040, and 1103312), the State
Key Laboratory of Heavy Oil in China (No. SKCHOP201504) and the
[30] P. Xu, W. Chen, Q. Wang, T. Zhu, M. Wu, J. Qiao, Z. Chen, J. Zhang, RSC Adv. 5
(2015) 6195–6206.

282
Y. Qiu et al. / Journal of Molecular Catalysis A: Chemical 424 (2016) 276–282
[31] A. Morozan, S. Campidelli, A. Filoramo, B. Jousselme, S. Palacin, Carbon 49
(2011) 4839–4847.
[32] G. Lu, Y. Zhu, K. Xu, Y. Jin, Z. Jason Ren, Z. Liu, W. Zhang, Nanoscale 7 (2015)
18271–18277.
[46] I. Hijazi, T. Bourgeteau, R. Cornut, A. Morozan, A. Filoramo, J. Leroy, V.
Derycke, B. Jousselme, S. Campidelli, J. Am. Chem. Soc. 136 (2014) 6348–6354.
[47] R.S. Ribeiro, A.M.T. Silva, J.L. Figueiredo, J.L. Faria, H.T. Gomes, Appl. Catal. B:
Environ. 140–141 (2013) 356–362.
[33] Y. Chen, S. Zhao, Z. Liu, Phys. Chem. Chem. Phys. 17 (2015) 14012–14020.
[34] X. Ma, X. He, Appl. Mater. Today 4 (2016) 1–8.
[35] T. Gensch, M.N. Hopkinson, F. Glorius, J. Wencel-Delord, Chem. Soc. Rev. 45
(2016) 2900–2936.
[48] T. Ma, S. Dai, M. Jaroniec, S. Qiao, Angew. Chem. Int. Ed. 53 (2014) 7281–7285.
[49] X. Duan, H. Sun, Y. Wang, J. Kang, S. Wang, ACS Catal. 5 (2015) 553–559.
[50] T. Sun, L. Xu, S. Li, W. Chai, Y. Huang, Y. Yan, J. Chen, Appl. Catal. B: Environ.
193 (2016) 1–8.
[36] B. Gao, S. Meng, X. Yang, Org. Process Res. Dev. 19 (2015) 1374–1382.
[37] D. Habibi, A.R. Faraji, M. Arshadi, S. Heydari, A. Gil, Appl. Catal. A: Gen. 466
(2013) 282–292.
[51] J. Zhu, K. Kailasam, A. Fischer, A. Thomas, ACS Catal. 1 (2011) 342–347.
[52] M.A. Langell, M.D. Anderson, G.A. Carson, L. Peng, S. Smith, Condens. Matter
59 (1999) 4791–4798.
[38] M. Arshadi, M. Ghiaci, Appl. Catal. A: Gen. 399 (2011) 75–86.
[39] M. Arshadi, M. Ghiacia, A. Rahmanian, H. Ghaziaskar, A. Gil, Appl. Catal. B:
Environ. 119–120 (2012) 81–90.
[40] M. Arshadi, M. Ghiaci, A.A. Ensafi, H. Karimi-Maleh, S.L. Suib, J. Mol. Catal. A:
Chem. 338 (2011) 71–83.
[53] Y. Gao, P. Tang, H. Zhou, W. Zhang, H. Yang, N. Yan, G. Hu, D. Mei, J. Wang, D.
Ma, Angew. Chem. Int. Ed. 55 (2016) 3124–3128.
[54] S. Wang, Z. Cui, M. Cao, Electrochimica. Acta 210 (2016) 328–336.
[55] X. Cui, Y. Li, S. Bachmann, M. Scalone, A.E. Surkus, K. Junge, C. Topf, M. Beller, J.
Am. Chem. Soc. 137 (2015) 10652–10658.
[41] Y. Yao, H. Xiao, P. Wang, P. Su, Z. Shao, Q. Yang, J. Mater. Chem. A 2 (2014)
11768–11775.
[56] H. Yu, F. Peng, H. Wang, B. Zhang, D. Su, Catal. Sci. Technol. 6 (2016)
1007–1015.
[42] L. Fu, Y. Chen, S. Zhao, Z. Liu, R. Zhu, RSC Adv. 6 (2016) 19482–19491.
[43] D. Wang, G. Yang, Q. Ma, M. Wu, Y. Tan, Y. Yoneyama, N. Tsubaki, ACS Catal. 2
(2012) 1958–1966.
[44] A. Rezaeifard, M. Jafarpour, Catal. Sci. Technol. 4 (2014) 1960–1969.
[45] J. Zhu, K. Kailasam, A. Fischer, A. Thomas, ACS Catal. 1 (2011) 342–347.
[57] D. Habibi, A.R. Faraji, M. Arshadi, J.L.G. Fierro, J. Mol. Catal. A:Chem 372 (2013)
90–99.
[58] J. Shui, M. Wang, F. Du, L. Dai, Sci. Adv. 1 (2015) e1400129.
[59] Z. Wei, J. Wang, S. Mao, D. Su, H. Jin, Y. Wang, F. Xu, H. Li, Y. Wang, ACS Catal. 5
(2015) 4783–4789.