A highly active non-precious metal catalyst based on Fe-N-C@CNTs for nitroarene reduction

Article abstract of DOI:10.1039/c6ra20666b

Research on transition metal-nitrogen-carbon (M-N-C) materials revealed their potential as catalysts in several important traditional reactions. However, the activity of M-N-C still needs to be further improved and the real active center of M-N-C catalysts is still under debate. In this work, an efficient Fe-N-C@CNTs for the hydrogenation of nitroarenes was prepared by pyrolysis of FeCl₃, phenanthroline and CNTs. Fe-N-C supported on CNTs is much more active than that supported on activated carbon, showing the promotion effect of CNTs. The characterization results suggest that the high activity of Fe-N-C is mainly attributed to the formation of ?-Fe₃N, which is the active site for the hydrogenation reaction. Nitrogen/carbon atoms contacted to the active centers could serve as bridges to transport the dissociated hydrogen atoms via spillover effect. The catalytic performance of Fe-N-C was also tested on fixed bed reactor under continuous flow condition for the first time and could smoothly catalyze the reaction for over 300 hours.

Full text of DOI:10.1039/c6ra20666b

View Article Online
View Journal
RSC Advances
This article can be cited before page numbers have been issued, to do this please use: Y. Qihua, J. Chen,
Y. Yao, J. Zhao, Y. Zhao, Y. Zheng and M. Li, RSC Adv., 2016, DOI: 10.1039/C6RA20666B.
This is an Accepted Manuscript, which has been through the
Royal Society of Chemistry peer review process and has been
accepted for publication.
Accepted Manuscripts are published online shortly after
acceptance, before technical editing, formatting and proof reading.
Using this free service, authors can make their results available
to the community, in citable form, before we publish the edited
article. This Accepted Manuscript will be replaced by the edited,
formatted and paginated article as soon as this is available.
You can find more information about Accepted Manuscripts in the
Information for Authors.
Please note that technical editing may introduce minor changes
to the text and/or graphics, which may alter content. The journal’s
standard Terms & Conditions and the Ethical guidelines still
apply. In no event shall the Royal Society of Chemistry be held
responsible for any errors or omissions in this Accepted Manuscript
or any consequences arising from the use of any information it
contains.
www.rsc.org/advances

Please do not adjust margins
RSC Advances
Page 1 of 8
View Article Online
DOI: 10.1039/C6RA20666B
RSC Advances
ARTICLE
A Highly Active Non-precious Metal Catalyst Based on Fe-N-
C@CNTs for Nitroarenes Reduction
Jian Chen,a,b, c Yi Yao,a,b,c Jiao Zhao,a Yaopeng Zhao,a Yuanyuan Zheng, Mingrun Li,a and Qihua
a
Received 00th January 20xx,
Accepted 00th January 20xx
Yang*a
DOI: 10.1039/x0xx00000x
Research on transition metal-nitrogen-carbon (M-N-C) materials revealed their potential as catalysts in several important
traditional reactions. However, the activity of M-N-C still needs to be further improved and the real active center of M-N-C
catalysts is still under debate. In this work, an efficient Fe-N-C@CNTs for the hydrogenation of nitroarenes was prepared
by pyrolysis of FeCl3, phenanthroline and CNTs. Fe-N-C supported on CNTs is much more active than that supported on
activated carbon, showing the promotion effect of CNTs. The characterization results suggest that the high activity of Fe-N-
C is mainly attributed to the formation of ε-Fe3N, which is the active site for the hydrogenation reaction. Nitrogen/carbon
atoms contacted to the active centers could serve as bridges to transport the dissociated hydrogen atoms via spillover
effect. The catalytic performance of Fe-N-C was also tested on fixed bed reactor under continuous flow condition for the
first time and could smoothly catalyze the reaction for over 300 hours.
www.rsc.org/
by the support material, for the support could impose
structural, electrical and other differences on the active sites.8
Carbon nanotubes (CNTs) are widely used as catalyst
Introduction
The Iron based heterogeneous catalyst is used in several
important chemical industrial processes such as Fischer-
Tropsch synthesis,1 ammonia synthesis2 and methane
aromatization3 due to its ability for activating a series of small
molecules like carbon monoxide, nitrogen,4 methane and etc.
In recent studies iron-nitrogen-carbon catalysts were applied
in hydrogenation reactions to transform nitroarenes into
anilines.5 Compared to precious metal based catalyst,6 the
hypothetical Fe-N-C catalysts show excellent chemoselectivity
to aromatic nitro-group above sensitive functional groups such
as aldehydes, ketones, olefins and etc., which made these
catalytic systems especially suitable for the production of
important intermediates of high economic value in
pharmaceuticals, agrochemicals, dyes, and fine chemicals.
Despite of its high selectivity, the activity of Fe-N-C catalyst
was relatively low and usually long reaction time was needed
to achieve complete conversion even at a substrate/catalyst
ratio of around 20. Therefore, to improve the catalytic activity
is now the main issue for fulfilling the potential practical
applications of Fe-N-C catalysts.
support.9,10 Its nano-scale diameter, unique electronic
properties and abundant surface groups could provide
improved mass transfer properties and catalytic performance
for active center.11,12 P. de Jong and co-workers found that
iron nanocrystals supported on CNTs show excellent catalytic
performance in F-T synthesis.13 Recently, C-Co3O4@N-doped
CNTs are in-situ generated from D-glucosamine hydrochloride,
melamine and Co(NO3)2•6H2O, which was found to be more
active than Fe-N-C/AC regarding the hydrogenation of
nitroarenes.14
The M-N-C catalysts are very complicated and their active
sites are still not very clear.15 In the pioneering work of
Matthias et al. they declare that the unique activity of nitrogen
doped carbon supported iron catalyst is derived from some
particular FeNx centers.5 Zhang and co-workers indicate that
Co single atoms bonded with N within graphitic sheets are the
active sites of Co−N−C catalyst in C−C bond-forming
reactions.16 Wang and co-worker attribute the high catalytic
performance of C-Co3O4@N-doped CNTs to the synergistic
effects of the unique structure of grapheme layers-coated Co0
and the electronic activation of doped nitrogen.14
Understanding the catalytic active sites is very important for
the catalyst design. Herein, we report a very efficient Fe-N-
C@CNTs catalyst for chemoselective hydrogenation of
nitroarenes. Our studies suggest that ε-Fe3N is the most active
site for the hydrogenation of nitrobenzene.
Previously, activated carbon or metal oxides are usually
used as support for the preparation of Fe-N-C catalysts.5,7 It is
well known that activity of catalysts could be influenced largely
a.State Key Laboratory of Catalysis, Dalian Institute of Chemical Physics, Chinese
Academy of Sciences, Dalian 116023, China Address here.
b.University of Chinese Academy of Sciences, Beijing 100039, China.
c. The authors contribute equally to this work.
* E-mail: yangqh@dicp.ac.cn
Electronic Supplementary Information (ESI) available: [details of any
supplementary information available should be included here]. See
DOI: 10.1039/x0xx00000x
Experimental section
Reagents.
This journal is © The Royal Society of Chemistry 2016
RSC Advances, 2016, 00, 1-3 | 1
Please do not adjust margins

Please do not adjust margins
RSC Advances
Page 2 of 8
View Article Online
DOI: 10.1039/C6RA20666B
ARTICLE
Journal Name
1,10-Phenanthroline and nitrobenzene was purchased from was placed in a stainless steel autoclave. The reactor was
J&K Inc. All other nitro compounds were purchased from Inno- purged and sealed with 5 MPa H2 and kept at 110 oC under
Chem Inc. Carbon nanotubes with a length of 0.5-2 µm and an stirring for a certain time. After the reaction, the autoclave
OD (Outer Diameter) of 10-20 nm was purchased from was cooled to room temperature and the product was
timesnano Inc. Solvents such as absolute ethanol and analyzed by Agilent 6890 gas chromatography equipped with a
tetrahydrofuran were purchased from SINOPHARM Co. Ltd. All flame ionization detector and a HP-5 capillary column.
materials and reagents were used as received without further Stability test by fixed-bed reactor.
purification.
A pre-degassed solution of nitrobenzene (6.9 ml) in 400 ml
Characterization.
H2O/THF (1/1 v/v) was fed continuously with a syringe pump
over a catalyst bed of 350 mg of Fe-N-C@CNTs-1.5 (0.3 mmol)
under 2.0 Mpa H2 atmosphere at 140 oC. Simultaneously, H2
was also continuously introduced with a flow rate of 7 ml/min.
The outlet of the reactor was sampled at different time
intervals. The crude mixtures were directly analyzed by GC.
The transmission electron microscopy (TEM) was undertaken
using a Hitachi HT-7700 at an acceleration voltage of 100 kV.
The samples were placed onto an ultrathin carbon film
supported on a copper grid. High resolution transmission
electron microscopy (HRTEM) images were recorded on a FEI
Tecnai F30 microscope with a point resolution of 0.20 nm
operated at 300 kV. The powder X-ray diffraction data were
collected on a Rigaku D/Max2500PC diffractometer with Cu Kα
radiation (λ=1.5418 Å) over the 2θ range of 5o - 80o with a scan
speed of 5o/min at room temperature. The 57Fe Mössbauer
spectra of the prepared samples were recorded on a Topologic
500A spectrometer and a proportional counter at room
temperature. 57Co(Rh) moving in a constant acceleration mode
was used as radioactive source. Ambient temperature high-
pressure hydrogen isotherms were recorded on a HPVA-2000
system.
Results and discussion
Synthesis and characterization of Fe-N-C@CNTs and Fe-N-C/CNTs-
1.5.
Fe-N-C@CNTs was prepared using phenanthroline, FeCl3 and
CNTs as precursors (Scheme 1). For comparison, Fe-N-C/CNTs-
1.5 was also prepared. The only difference in the preparation
method is that ultrasonic treatment of phenanthroline, FeCl3
and CNTs was performed for Fe-N-C@CNTs and Fe-N-C/CNTs-
1.5 was prepared without the ultrasonic treatment. After
solvent evaporation and successive pyrolysis at 800 oC under
N2 atmosphere, Fe-N-C@CNTs and Fe-N-C/CNTs-1.5 were
obtained.
CNTs treatment.
Firstly, 8 g of commercial CNTs (with diameter of 10-20 nm)
o
were refluxed in 400 ml of concentrated nitric acid at 120 C
for 11 h. After cooling down to room temperature, the mixture
was diluted and washed with copious amount of water to
neutral. After that, the collected CNTs were dispersed in 400
ml of ethanol by ultrasound for 4 h. Treated CNTs was
obtained after filtering and drying.
Synthesis of Fe-N-C@CNTs.
FeCl3 (0.0145 g) and phenanthroline (0.0531 g) were refluxed
in ethanol (50 ml) for 12 h to form solution A. CNTs (1 g) were
dispersed in solution A with the aid of ultrasonic treatment for
4 h. The solid precursor was obtained by rotary evaporation of
the solvent under vacuum. The precursor was pyrolyzed in
Scheme 1 Schematic illustration for the synthesis of Fe-N-C@CNTs.
The SEM images of Fe-N-C@CNTs-1.5 and Fe-N-C/CNTs-1.5
clearly show the tangles of CNTs with diameter of 10-20 nm
(Fig. S1). The small metal particles with size less than 10 nm
scatters in the CNTs and also small amount of particle with size
about 30-50 nm and above 100 nm could be found in the SEM
image of Fe-N-C@CNTs-1.5. In addition to small metal
nanoparticles (less than 10 nm), larger metal particles (300-
500 nm) also exist in the SEM image of Fe-N-C/CNTs-1.5. The
TEM images confirm the above SEM results that both small
nanoparticles and large metal particles co-exist in both
samples and Fe-N-C/CNTs-1.5 has more amount of large-sized
metal particles than Fe-N-C@CNTs-1.5 does (Fig. 1). For Fe-N-
C@CNTs-1.5 and Fe-N-C/CNTs-1.5, most of metal particles are
dispersed outside of the CNTs, although some metal particles
still exist inside CNTs. According to the elemental mapping (Fig.
S2), the N and Fe element are uniformly distributed in the
entire skeletal framework, which suggests that Fe atom may
o
o
nitrogen at 800 C for 2 h with heating rate of 5 C /min and
the gas flow of 20 ml/min. Black powder of Fe-N-C@CNTs-0.5
was obtained after cooling to room temperature. For other Fe-
N-C@CNTs, the synthetic process was similar, expect that FeCl3
(0.043 g) and phenanthroline (0.1593 g) were used for the
synthesis of Fe-N-C@CNTs-1.5 and FeCl3 (0.0871 g) and
phenanthroline (0.3186 g) were used for the synthesis of Fe-N-
C@CNTs-3. Fe@CNT-1.5 was synthesized with the same
conditions to Fe-N-C@CNTs-1.5 except that only FeCl3 was
used. Fe-N-C/CNT-1.5 was prepared in a similar method to Fe-
N-C@CNT-1.5 but no ultrasonic treatment was performed
after mixing solution A with CNTs.
Hydrogenation of nitrobenzene.
In a typical experiment (S/C=22), 80 mg of Fe-N-C@CNTs-1.5,
46 μl of nitrobenzene, 2 ml of H2O and 2 ml of THF were placed
in a 5 ml ampoule. After ultrasonic treatment, the ampoule
2 | RSC Advances, 2016, 00, 1-3
This journal is © The Royal Society of Chemistry 20xx
Please do not adjust margins

Please do not adjust margins
RSC Advances
Page 3 of 8
View Article Online
DOI: 10.1039/C6RA20666B
Journal Name
ARTICLE
have some interaction with
N
atom. The above 2A). After comparing with the standard pattern of different
characterization results suggest that the ultrasonic treatment iron nitrides, carbides and oxides, it is still difficult to assign
could facilitate the uniform distribution of precursors on CNTs this new phase. Fe-N-C@CNTs and Fe-N-C/CNTs-1.5 were
for the formation of small metal NPs. The TEM images of Fe-N- prepared with the same amount of iron and phenanthroline.
C@CNTs-0.5 and Fe-N-C@CNTs-3 shown in Fig. S3 are similar Even with higher amount of iron and phenanthroline, the
to that of Fe-N-C@CNTs-1.5. However, the particle size outside above mentioned diffraction peaks still could not be found in
the CNTs steadily grows to over several hundred nanometers the XRD patterns of Fe-N-C@CNTs. The reason is still not clear
with the increment of iron loading. With Fe-N-C@CNTs-1.5
at present stage, but in consideration of the only difference
during the preparation of two kinds of catalysts (ultrasonic
treatment), it is probably related with the higher dispersion
degree of iron precursor, for long period ultrasonic treatment
would result in changes to functional groups on the surface of
highly oxidized CNTs,17 which may influence its interaction
with iron precursors during the pyrolysis.
Fig. 1 TEM images of (A) Fe-N-C@CNTs-1.5 and (B) Fe-N-C/CNTs-1.5, (C) HR-TEM
images of the typical metal particles with Fe-N-C@CNTs-1.5 as model: (a-b) iron
nanoparticles wrapped with amorphous carbon shells inside CNTs, (c-d) iron
nanoparticles wrapped with amorphous carbon shells outside CNTs, (e-f) iron crystals
tightly encapsulated by highly graphitized carbon layers.
as model, metal nanoparticles could be classified as metal
nanoparticles wrapped with amorphous carbon shells and iron
crystals tightly encapsulated by highly graphitized carbon
layers (Fig. 1C). These species were formed by pyrolysis of
phenanthroline, FeCl3 and carbon (or Al2O3).5 By measuring the
d-space of iron containing crystals we could acquire some
information on species inside different kinds of carbon layers.
For particles wrapped with a layer of poorly graphitized carbon
shell, lattice fringes attributed to (201) or (111) facet of Fe3C (b
and d in Fig. 1C) were observed while for particles coated with
a layer of highly graphitized carbon lattice fringes attributed to
(110) facet of α-Fe were identified (Fig. 1Cf).
Fig. 2 (A) X-ray powder diffraction patterns of (a) Fe-N-C@CNTs-0.5, (b) Fe-N-C@CNTs-
1.5, (c) Fe-N-C/CNTs-1.5, (d) Fe-N-C@CNTs-1.5 after storing in air for a month, (e) Fe-N-
C@CNTs-1.5 after the 10th recycle and (f) Fe-N-C@CNTs-3.0 (▼Fe0, ■unknown
species, ◆FeN0.076, ★Fe3O4). Dashed lines indicate the characteristic peaks of
graphite. Mössbauer spectra (298K) of (B) Fe-N-C@CNTs-1.5, (C) Fe-N-C/CNTs-1.5, and
(D) Fe-N-C@CNTs-3.0. The parameters of doublet and sextets were listed in Table S1.
Mössbauer spectra were employed to characterize iron
species of Fe-N-C@CNTs and Fe-N-C/CNTs-1.5 because
Mössbauer spectra may give more details on the state of iron
species (Fig. 2). Due to low Fe content, no signals could be
found for Fe-N-C@CNTs-0.5. For Fe-N-C@CNTs-1.5 and Fe-N-
C@CNTs-3.0, the signals could be well fitted into one doublet
and three sets of sextets. The doublet with the isomer shift (IS)
of~0.2 mm s-1 and the quadrupole splitting (QS) of~1.0 mm s-1
might be attributed to Fe(II) or Fe(III) species of some poorly
crystallized iron oxide.18,19 Sextet 1 and Sextet 2 could be
attributed to α-iron and χ-Fe2C5/Fe3C respectively;18,20,21 Sextet
3 may be ε-Fe3N in which Fe was coordinated with 3 nitrogen
atoms.22 It is interesting to mention than Fe-N-C@CNTs affords
all three sets of sextets and Fe-N-C/CNTs-1.5 only shows one
doublet and two of sextets (Sextet 1 and Sextet 2). This
indicates that Fe-N-C/CNTs-1.5 with poorly crystallized iron
oxide, α-iron and χ-Fe2C5/Fe3C has no ε-Fe3N species. The
The crystal phase of iron nanoparticles was further testified
using X-ray powder diffraction technique (Fig. 2A). The
diffraction peaks attributed to graphitic carbon, metallic iron
(α-Fe, JCPDS-PDF: 65-4899) and metal nitrides (FeN0.076)
could be clearly observed in the XRD patterns of Fe-N-C@CNTs
and Fe-N-C/CNTs. As Fe loading increases from 0.5 to 1.5, the
intensity of α-Fe species increases. This is probably due to the
formation of iron species with larger particle size at high Fe
loading based on TEM characterizations. It is noteworthy to
mention that there was a set of diffraction patterns appeared
in the XRD pattern of Fe-N-C/CNTs-1.5 (indicated by
■ in Fig.
This journal is © The Royal Society of Chemistry 20xx
RSC Advances, 2016, 00, 1-3 | 3
Please do not adjust margins

Please do not adjust margins
RSC Advances
Page 4 of 8
View Article Online
DOI: 10.1039/C6RA20666B
ARTICLE
Journal Name
unknown species in the XRD pattern of Fe-N-C/CNTs-1.5 may under the same reaction conditions, such as the same S/C ratio
be related with χ-Fe2C5/Fe3C based on the results of and the same solvent ect. Though Fe-N-C/CNTs-1.5 is less
Mössbauer.
active than Fe-N-C@CNTs-1.5, it is still more active than Fe-N-
C/AC, suggesting the promotion effect of CNTs.
Catalytic performance of Fe-N-C@CNTs in the hydrogenation of
nitroarenes.
With Fe-N-C@CNTs-1.5 as model catalyst, the influence of
the solvent on the catalytic performance was tested (Table 1).
Fe-N-C@CNTs-1.5 could efficiently catalyze the reaction to
afford 99% conversion in 6 h under solvent free condition,
which is quite important for energy saving and economic
production of chemicals. Under similar reaction conditions,
much lower conversion was achieved using pure H2O and pure
THF as solvent. The TOF of Fe-N-C@CNTs-1.5 is 46.8 h-1 in
TOF/H2O and 33.3 h-1 under neat conditions. The above results
suggest that the mixture of TOF and H2O is the best solvent for
Fe-N-C@CNTs.
The catalytic performance of Fe-N-C@CNTs was tested in the
hydrogenation of nitrobenzene and compared with Fe-N-
C/CNTs-1.5 (Table 1). All samples tested could efficiently
catalyze the hydrogenation of nitrobenzene with aniline as the
only product. Fe-N-C@CNTs-0.5 and Fe-N-C@CNTs-1.5 could
afford about 98% conversion in 0.5 h with H2O/THF (1/1 v/v) as
solvent. Under similar reaction conditions, the reaction time
needs to prolong to 1 h for Fe-N-C@CNTs-3.0 to achieve >99%
conversion. Due to the fast reaction rate of Fe-N-C@CNTs, it is
hard to get precise kinetic curves to measure the turn over
frequency. Thus, S/C was increased from 22 to 300 and TOF
was calculated with conversion less than 30%. As Fe loading
increases, the TOF increases and reaches the maximum at Fe
loading of 1.5 wt%. Fe-N-C@CNTs-1.5 with TOF of 46.8 h-1
shows the highest activity among all Fe-N-C@CNTs samples.
Fe-N-C@CNTs-1.5 could also afford >99% conversion in 6 h,
indicating that Fe-N-C@CNTs-1.5 is active enough to catalyze
the reaction even with low catalyst amount. It takes 10 h for
Fe-N-C/CNTs-1.5 to achieve full conversion. Due to its low
activity, the TOF of Fe-N-C/CNTs-1.5 was measured at S/C of
22. Fe-N-C@CNTs-1.5 is much more active than Fe-N-C/CNTs-
1.5 (TOF: 46.8 versus 5.1 h-1). As far as we know, Fe-N-
C@CNTs-1.5 is the highest active Fe-N-C based catalyst ever
reported for nitrobenzene hydrogenation.
The active sites of Fe-N-C catalysts are still not fully
understood though previous reports suggest that iron particles
surrounded with C and N species are the active sites. Recently,
Zhang and co-workers indicates that Co single atoms bonded
with N within graphitic sheets are the active sites in the
Co−N−C catalyst for C−C bond-forming reactions.16 In a just
published report by Su et al., researchers attributed the real
active center of Co-N-C catalyst to chelate complexes with
cobalt atoms bonded to 2 to 3 nitrogen atoms in the graphene
lattice which probably like the pyridinic vacancy.23
Fe-N-C/CNTs-1.5 without ε-Fe3N (Sextet 3) is less active
than Fe-N-C@CNTs-x (x=1.5 and 3) possessing ε-Fe3N,
suggesting that ε-Fe3N should be one kind of active sites. Fe-N-
C/CNTs-1.5 shows catalytic activity for the hydrogenation of
nitrobenzene though the activity is not very high. This
indicates that there should be another kind of active sites in
addition to ε-Fe3N. The results of XRD and Mössbauer show
that Fe-N-C/CNTs-1.5 has α-Fe, FeN0.076, iron oxide, and χ-
Fe2C5/Fe3C. A control sample, Fe@CNTs-1.5 was prepared
under similar conditions to Fe-N-C@CNTs-1.5 but without the
addition of phenanthroline. This sample shows only trace
amount conversion of nitrobenzene, but the TEM and XRD
result shows the presence of α-Fe in this sample (Fig. S3C and
Fig. S4). Thus, another kind of active site may possibly involve
FeN0.076 and χ-Fe2C5/Fe3C. Comparing the activity of Fe-N-
C@CNTs and Fe-N-C/CNTs-1.5, it can be stated that ε-Fe3N
should be much more active than other kind of active sites
(FeN0.076 and χ-Fe2C5/Fe3C). Fe-N-C@CNTs-1.5 and Fe-N-
C@CNTs-3.0 have similar amount of ε-Fe3N based on
Mössbauer characterization, the higher catalytic activity of
former sample is possibly due to its smaller particle size. This is
reasonable due to the fact that smaller particles could increase
the surface area of active sites.
Table 1. The catalytic performance of Fe-N-C@CNTs and Fe-N-C/CNTs in the
hydrogenation of nitrobenzene[a].
TOF
(h-1)
Cat.
S/C
Time (h)
Conv. (%)
98.5
Fe-N-C@CNTs-0.5
Fe-N-C@CNTs-1.5
Fe-N-C@CNTs-3.0
Fe-N-C/CNTs-1.5
Fe-N-C/AC[b]
22
22
0.5
0.5
1
-
-
>
>
99
99
22
-
22
10
15
6
99
99
5.1
22
Fe-N-C@CNTs-0.5
Fe-N-C@CNTs-1.5
Fe-N-C@CNTs-3.0
Fe-N-C@CNTs-1.5[c]
Fe-N-C@CNTs-1.5[d]
Fe-N-C@CNTs-1.5[e]
300
300
300
300
300
300
>
>
99
99
33.5
46.8
17
6
6
37
6
>
99
33.3
-
6
69.5
43.7
6
-
[a] Reaction conditions: 0.5 mmol nitrobenzene, S/C of 22, H2O/THF (4 ml, 1/1
v/v), 110 oC, 5 MPa H2. TOF was calculated based on kinetic curves with
conversion less than 30%. [b] data from ref 5, reaction temperature is 120 oC. [c]
No solvent was used. [d] H2O was used as solvent. [e] THF was used as solvent.
Though the active sites for Fe-N-C based catalysts are very
complicated, our results indicated that ε-Fe3N should be the
most active site for the hydrogenation of nitrobenzene. Also,
smaller particle size benefits higher catalytic activity. Fe-N-
C@CNTs-x and Fe-N-C/CNTs-1.5 all show higher activity than
Fe-N-C/AC, suggesting the unique structure of CNTs brought
about profound influences on the active species. Since the
surface area of CNTs is much smaller than AC, the extension of
It takes 10 h and 15 h respectively for Fe-N-C/CNTs-1.5 and
Fe-N-C/AC to achieve full conversion, suggesting that Fe-N-
C@CNTs-1.5 is much more active than Fe-N-C/AC reported in
5
the literature. Due to the fact that no TOF or the kinetic
curves of Fe-N-C/AC was reported in the literature, it is
reasonable to compare the activity of Fe-N-C/CNTs-1.5 and Fe-
N-C/AC with reaction times needed to achieve full conversion
4 | RSC Advances, 2016, 00, 1-3
This journal is © The Royal Society of Chemistry 20xx
Please do not adjust margins

Please do not adjust margins
RSC Advances
Page 5 of 8
View Article Online
DOI: 10.1039/C6RA20666B
Journal Name
ARTICLE
dispersion of active sites should not be a reason for the dehelogenation within 0.5 to 1 h. Amazingly, the substrates
difference in activity. The reason is not clear at present stage. bearing other easily reducible functional group (aldehydes, CN,
Another indispensable part of the catalyst was nitrogen pyridine) could be also converted to corresponding substituted
atom. Langmuir–Hinshelwood mechanism of the nitroarenes amines without affecting the substitution groups. It is
reduction involves surface adsorption of reducing agent and noteworthy to mention that it takes no more than 3 h to
substrate on the catalyst.24,25 Chen and his co-workers confirm achieve 96% to >99% conversion for all substrates listed in
that N doping is suitable to introduce active sites for the
adsorption of substrates on catalyst, which is critical to the
following catalytic process.26 In Matthias’ report they revealed
that carbon/alumina supported iron catalyst could only yield
trace amount of product under the same reaction conditions,
indicating that N is an essential element for the reaction.5
Although from XPS results we got some information on the
distribution of N species (Fig. S5 and S6), the contribution of
interaction between iron and nitrogen towards the reaction
still remained largely unknown. In a recent report, Gianvito
Vilé claimed that nitrogen atoms around metal center may
facilitate heterolytic dissociation of hydrogen molecule,27
according to the result we applied high pressure hydrogen
adsorption measurement on different materials to elucidate
the function of nitrogen atoms in the catalyst (Fig. 3). The
measurements were performed on a high pressure adsorption
Fig. 3. High-pressure hydrogen isotherms at 298 K for pure CNTs and CNTs supported
apparatus and the equilibrium time for each point was 20 min.
iron based catalysts with or without nitrogen/carbon bridges.
Considering the possibility of chemical adsorption, the data
acquired was a dynamic result rather than equilibrium. Pure
CNTs could only adsorb trace amount of hydrogen, which is
within the range of physical adsorption. The quantity of
hydrogen gas adsorbed on Fe-N-C/CNTs-1.5 and Fe@CNTs-1.5
was even lower, indicating that the addition of iron and
nitrogen via simple impregnation merely increased the weight
of the material, while single component of iron could not fulfill
the activation of hydrogen molecule either. The hydrogen
adsorption amount for Fe-N-C@CNTs-1.5 increased
significantly to ~0.6% at the pressure of 6 MPa, a value which
is at the same level for carbon bridged Pd on AC.28,29 Similar
high hydrogen take up was reproducible for Fe-N-C@CNTs-3.0.
The difference in hydrogen adsorption between Fe-N-C@CNTs-
1.5 and Fe@CNTs-1.5 should originate from the
nitrogen/carbon coating layer derived from pyrolysis of
phenanthroline. Fe-N-C@CNTs-1.5 and Fe-N-C/CNTs-1.5
synthesized with FeCl3 and phenanthroline both have
nitrogen/carbon coating layer. The higher H2 adsorption
properties of the former sample may be attributed to the
existence of ε-Fe3N. Hydrogen molecules could be dissociated
on the surface of iron containing active centers and then
diffuse on to the surface of CNTs support.29,30 By comparing
the catalytic results, we can conclude that the ability for
activating and storing hydrogen under high pressure
undoubtedly contributes to the high activity of Fe-N-C@CNTs
catalysts.
Table 2, showing the high catalytic activity of Fe-N-C@CNTs-
1.5. For substrate bearing ethylene group, the product
selectivity is 87%, slightly lower than Co-N-C or Fe-N-C
catalysts reported in literature. This is probably due to the high
catalytic activity of Fe-N-C@CNTs-1.5.
The stability of Fe-N-C@CNTs-1.5 was tested using batch
reactor and fixed-bed reactor (Fig. 4). The recycle results based
on batch reactor shows up to 99% conversion could be
obtained for the nine cycles though the reaction time is
extended from 1 h (for 1-5 cycles) to 2 h (for 6-10 cycles). For
the tenth cycles, the conversion decreases from 99% to 95%.
Compared with fresh catalyst, no obvious change in the
particles size was found for Fe-N-C@CNTs-1.5 after ten cycles
according to the TEM characterization (Fig. 1A and Fig. S7). The
X-ray diffraction pattern of catalyst after recycling is also
shown in Fig. 2A. Only Fe3O4 could be identified in the
material. Though the catalyst could still convert nitrobenzene
into aniline, iron oxide should not be an active site due to the
insensitivity of XRD for detecting low content species. We also
prepared an aged catalyst by exposing Fe-N-C@CNTs-1.5 to
the open air for a month. There was no obvious change in the
XRD result. However, this catalyst only affords 20% conversion
o
in 0.5 h with S/C of 22 with THF/H2O as solvent at 110 C. The
result shows the insensitivity of XRD for detecting the active
species for Fe-N-C catalysts. The results of fixed-bed reactor
show that 100% conversion could be obtained with flow rate
of 0.06 ml/min for 120 h. The conversion drops to 70% with
the flow rate increasing to 0.09 ml/min. After that the
continuous decreasing of conversion from 70% to 42% could
be observed. The above results suggest that the active site of
Fe-N-C is not stable during the catalytic process and the
stability of this kind of catalyst should be further improved.
With Fe-N-C@CNTs-1.5 as model catalyst, the general
substrate scope of Fe-N-C@CNTs catalysts was tested in the
hydrogenation of nitroarenes containing some sensitive
functional groups such as aldehydes, C=C, CN and so on (Table
2). In addition of nitrobenzene, hydrogenation of halogen-
substituted nitroarenes proceeds smoothly to produce
corresponding
haloaromatic
amines
without
any
This journal is © The Royal Society of Chemistry 20xx
RSC Advances, 2016, 00, 1-3 | 5
Please do not adjust margins

Please do not adjust margins
RSC Advances
Page 6 of 8
View Article Online
DOI: 10.1039/C6RA20666B
ARTICLE
Journal Name
there are at least two kinds of active species for Fe-N-C
catalysts and ε-Fe3N is the most active site for catalyzing the
nitroarene hydrogenation. Nitrogen/carbon atoms contacted
to the active centers could serve as bridges to transport the
dissociated hydrogen atoms and further enhance the activity.
The sonication treatment is one of the key factors for the
formation of ε-Fe3N species. Fe-N-C@CNTs catalysts also show
excellent durability which could run under continuous flow
condition over 100 hours.
Table 2. Hydrogenation of different types of nitroarenes catalyzed by Fe-N-C@CNTs-
1.5[a]
.
Substrate
Product
Time (h)
0.5
Sel. (%)
Conv. (%)
>99
>99
0.5
1
>99
>99
>99
>99
>99
>98
>99
>99
>99
>99
>99
87
Acknowledgements
This work was financially supported by NSFC (21325313,
21321002) and the strategitic Priority Research Program of the
Chineses Academy of Sciences (Grant No. XDB17020200).
I
NH2
0.5
3
References
1
2
3
H. M. Torres Galvis, J. H. Bitter, C. B. Khare, M. Ruitenbeek,
A. I. Dugulan and K. P. de Jong, Science, 2012, 335, 835.
N. Svend Erik, In Innovations in Industrial and Engineering
Chemistry, American Chemical Society: 2008, Vol. 1000, p 15.
X. Guo, G. Fang, G. Li, H. Ma, H. Fan, L. Yu, C. Ma, X. Wu, D.
Deng, M. Wei, D. Tan, R. Si, S. Zhang, J. Li, L. Sun, Z. Tang, X.
O2N
NO2
CH3
[b]
1.5
1.5
NH2
NH2
Pan and X. Bao, Science, 2014, 344, 616.
D. R. Strongin, J. Carrazza, S. R.Bare and G. A. Somorjai, J.
Catal., 1987, 103, 213.
R. V. Jagadeesh, A.-E.Surkus, H. Junge, M.-M. Pohl, J. Radnik,
J. Rabeah, H. Huan, V. Schünemann, A. Brückner and M.
Beller, Science, 2013, 342, 1073.
4
5
[b]
3
3
>99
>96
>99
>99
CN
[b]
N
N
6
7
H.-U. Blaser, H. Steiner and M. Studer, ChemCatChem
, 210.
H. Xiao, Z.-G. Shao, G. Zhang, Y. Gao, W. Lu and B. Yi, Carbon
2013, 57, 443.
Q. Fu and T. Wagner, Surf. Sci. Rep., 2007, 62, 431.
, 2009,
NH2
NH2
1
[a] Reaction conditions: 0.5 mmol substrate, 4.5 mol% Fe, 2 ml H2O+2 ml THF,
110 oC, 5 Mpa H2, [b] 90 oC.
,
8
9
X. Tan, W. Deng, M. Liu, Q. Zhang and Y. Wang, Chem.
Commun., 2009, 46, 7179.
10 Y. Yao, H. Xiao, P. Wang, P. Su, Z. Shao and Q. Yang, J. Mater.
Chem. A 2014, , 11768.
,
11 W. Chen, X. Pan and X. Bao, J. Am. Chem. Soc., 2007, 129
2
,
7421.
12 X. Pan, Z. Fan, W. Chen, Y. Ding, H. Luo and X. Bao, Nat.
Mater., 2007, , 507.
6
13 M. Casavola, J. Hermannsdörfer, N. de Jonge, A. I. Dugulan
and K. P. de Jong, Adv. Funct. Mater., 2015, 25, 5309.
14 Z. Wei, J. Wang, S. Mao, D. Su, H. Jin, Y. Wang, F. Xu, H. Li
and Y. Wang, ACS Catal., 2015, 5, 4783.
15 F. A. Westerhaus, R. V. Jagadeesh, G. Wienhöfer, M.-M. Pohl,
J. Radnik, A.-E. Surkus, J. Rabeah, K. Junge, H. Junge, M.
Fig. 4 (A) Recycling of Fe-N-C@CNTs-1.5 in batch reactor, reaction condition: 0.5
mmol nitrobenzene, 4.5 mol% Fe, H2O+THF (1/1 v/v), 4 ml, 110 °C, 5 MPa H2, 1 h
for 1-5th cycle, 2 h for 6-10th. (B) The stability of Fe-N-C@CNTs-1.5 tested in
fixed-bed reactor, reaction condition: 6.9 ml nitrobenzene, H2O+THF (1/1 v/v),
400 ml, 350 mg Fe-N-C@CNTs-1.5 (0.3 mmol), 2 Mpa H2 with a flow rate of 7
ml/min, 140 oC.
Nielsen, A. Brückner and M. Beller, Nat. Chem., 2013,
16 L. Zhang, A. Wang, W. Wang, Y. Huang, X. Liu, S. Miao, J. Liu
and T. Zhang, ACS Catal., 2015, , 6563.
17 R. Zeineldin, M. Al-Haik and L. G. Hudson, Nano Lett., 2009,
, 751.
5, 537.
5
9
18 S. Zhang, B. Liu and S. Chen, Phys. Chem. Chem. Phys., 2013,
15, 18482.
19 U. I. Kramm, I. Herrmann-Geppert, S. Fiechter, G. Zehl, I.
Zizak, I. Dorbandt, D. Schmeißer and P. Bogdanoff, J. Mater.
Conclusions
Chem. A
20 Z. Yang, S. Guo, X. Pan and J. Wang, X. Bao, Energ. Environ.
Sci., 2011, , 4500.
21 K. Cheng, M. Virginie, V. V. Ordomsky, C. Cordier, P. A.
Chernavskii, M. I. Ivantsov, S. Paul, Y. Wang and A. Y.
Khodakov, J. Catal., 2015, 328, 139.
, 2014, 2, 2663.
In this report a series of CNTs supported Fe-N-C catalysts were
prepared and they show enhanced activity towards
4
hydrogenation of nitroarenes while keeping
a
high
chemoselectivity. Fe-N-C@CNTs possessing ε-Fe3N is much
more active than Fe-N-C/CNTs without ε-Fe3N. We believe that
6 | RSC Advances, 2016, 00, 1-3
This journal is © The Royal Society of Chemistry 20xx
Please do not adjust margins

Please do not adjust margins
RSC Advances
Page 7 of 8
View Article Online
DOI: 10.1039/C6RA20666B
Journal Name
ARTICLE
22 S. Kurian and N. S.Gajbhiye, Chem. Phys. Lett., 2010, 493
299.
,
23 T. Cheng, H. Yu, F. Peng, H. Wang, B. Zhang and D. Su, Catal.
Sci. Technol., 2016, , 1007.
6
24 T. Aditya, A. Pal and T. Pal, Chem. Commun., 2015, 51, 9410.
25 X. Kong, Z. Sun, M. Chen, C. Chen and Q. Chen, Energ.
Environ. Sci., 2013, 6, 3260.
26 H. Hu, X. Wang, D. Miao, Y. Wang, C. Lai, Y. Guo, W. Wang, J.
Xin and H. Hu, Chem. Commun., 2015, 51, 16699.
27 G. Vilé, D. Albani, M. Nachtegaal, Z. Chen, D. Dontsova, M.
Antonietti, N. López and J. Pérez-Ramírez, Angew. Chem. Int.
Ed., 2015, 54, 11265.
28 R. T. Yang and Y. Wang, J. Am. Chem. Soc., 2009, 131, 4224.
29 L. Wang and Yang R. T., Catal. Rev., 2010, 52, 411.
30 A. J. Lachawiec, G. Qi and R. T. Yang, Langmuir, 2005, 21,
11418.
This journal is © The Royal Society of Chemistry 20xx
RSC Advances, 2016, 00, 1-3 | 7
Please do not adjust margins

RSC Advances
Page 8 of 8
View Article Online
DOI: 10.1039/C6RA20666B
Table of Contents Graphic and Synopsis Here
An efficient FeꢀNꢀC@CNTs for the hydrogenation of nitroarenes was prepared by pyrolysis of
FeCl3, phenanthroline and CNTs. It was found that εꢀFe3N is the most active site of FeꢀNꢀC and
nitrogen/carbon atoms contacted to the active centers could serve as bridges to transport the
dissociated hydrogen atoms via spillover effect.
1