Aerobic oxidative coupling of alcohols and amines to imines over iron catalysts supported on mesoporous carbon

Article abstract of DOI:10.1016/S1872-2067(16)62506-8

Direct oxidative coupling of an alcohol and amine, with air or molecular oxygen as the oxygen source, is an environmentally friendly method for imine synthesis. We developed an Fe catalyst supported on mesoporous carbon (denoted by FeO_x/HCMK-3) for this reaction with excellent activity and recyclability. FeO_x/HCMK-3 was prepared by impregnating HNO₃-treated mesoporous carbon (CMK-3) with iron nitrate solution. The highly dispersed FeO_x species give FeO_x/HCMK-3 high reducibility and are responsible for the high catalytic performance. Imine synthesis over FeO_x/HCMK-3 follows a redox mechanism. The oxygen species in FeO_x/HCMK-3 participate in the reaction and are then regenerated by oxidation with molecular O₂. The reaction involves two consecutive steps: oxidative dehydrogenation of an alcohol to an aldehyde and coupling of the aldehyde with an amine to give an imine. Oxidative dehydrogenation of the alcohol is the rate-determining step in the reaction.

Full text of DOI:10.1016/S1872-2067(16)62506-8

Chinese Journal of Catalysis 37 (2016) 1451–1460
a v a i l a b l e a t w w w . s c i e n c e d i r e c t . c o m
j o u r n a l h o m e p a g e : w w w . e l s e v i e r . c o m / l o c a t e / c h n j c
Article
Aerobic oxidative coupling of alcohols and amines to imines over
iron catalysts supported on mesoporous carbon
Longlong Geng, Jinling Song, Bin Zheng, Shujie Wu, Wenxiang Zhang, Mingjun Jia, Gang Liu*
Key Laboratory of Surface and Interface Chemistry of Jilin Province, College of Chemistry, Jilin University, Changchun 130012, Jilin, China
A R T I C L E I N F O
A B S T R A C T
Article history:
Direct oxidative coupling of an alcohol and amine, with air or molecular oxygen as the oxygen
source, is an environmentally friendly method for imine synthesis. We developed an Fe catalyst
supported on mesoporous carbon (denoted by FeOx/HCMK‐3) for this reaction with excellent activ‐
ity and recyclability. FeOx/HCMK‐3 was prepared by impregnating HNO3‐treated mesoporous car‐
bon (CMK‐3) with iron nitrate solution. The highly dispersed FeOx species give FeOx/HCMK‐3 high
reducibility and are responsible for the high catalytic performance. Imine synthesis over
FeOx/HCMK‐3 follows a redox mechanism. The oxygen species in FeOx/HCMK‐3 participate in the
reaction and are then regenerated by oxidation with molecular O2. The reaction involves two con‐
secutive steps: oxidative dehydrogenation of an alcohol to an aldehyde and coupling of the aldehyde
with an amine to give an imine. Oxidative dehydrogenation of the alcohol is the rate‐determining
step in the reaction.
Received 30 May 2016
Accepted 12 July 2016
Published 5 September 2016
Keywords:
Iron catalyst
Imine synthesis
Oxidative coupling
Mesoporous carbon
Molecular oxygen
© 2016, Dalian Institute of Chemical Physics, Chinese Academy of Sciences.
Published by Elsevier B.V. All rights reserved.
1. Introduction
heterogeneous catalysts such as supported Au, Ru, Pd, and Pt
have been reported [17–24]. However, the high price of these
Imines are important nitrogen compounds; they contain a
reactive C=N moiety [1,2]. Imines are nitrogen sources and can
undergo various transformations. They are widely used in bio‐
logical, agricultural, and pharmaceutical synthetic processes
[3–6]. The condensation of aldehydes or ketones with amines is
the traditional method for imine synthesis, and is usually per‐
formed in the presence of an acidic catalyst [7–9]. Recently, the
direct oxidative coupling of an alcohol and amine to give an
imine has attracted increasing attention because of its envi‐
ronmentally friendly properties [10–16]. Air or molecular oxy‐
gen can be used as the oxygen source for the production of
various types of desirable imines from appropriate alcohols
and amines as starting reagents, with water as the only
by‐product. Effective noble‐metal‐based homogeneous and
noble metals and the need to use inorganic or organic bases as
additives in the reaction system limit their large‐scale applica‐
tion in imine synthesis.
Recently, a few metal oxide catalysts that can catalyze this
imine synthesis have been reported [25–27]; for example, CeO2
is an effective catalyst for imine formation from benzyl alcohol
and aniline at low temperature (60 °C), with a high yield of
96% after 24 h [25]. Manganese oxides supported on hydroxy‐
apatite (MnOx/HAP) also show high activity in the oxidative
coupling of alcohols and amines [26]. When benzyl alcohol and
aniline were used, a 92% yield of the imine was obtained after
reaction for 24 h at 80 °C. The catalytic activity, especially the
capacity to activate molecular oxygen, is highly dependent on
the redox properties of the metal oxides. They determine the
* Corresponding author. Tel: +86‐431‐85155390; Fax: +86‐431‐88499140; E‐mail: lgang@jlu.edu.cn
This work was supported by the National Natural Science Foundation of China (21473073, 21473074) and the "13th Five‐Year" Science and
Technology Research of the Education Department of Jilin Province (2016403).
DOI: 10.1016/S1872‐2067(16)62506‐8 | http://www.sciencedirect.com/science/journal/18722067 | Chin. J. Catal., Vol. 37, No. 9, September 2016

1452
Longlong Geng et al. / Chinese Journal of Catalysis 37 (2016) 1451–1460
catalytic behavior of the metal oxide in the oxidative dehydro‐
genation step. Tuning the redox properties of metal oxide cata‐
lysts should therefore be an efficient method for developing
high‐performance catalysts for imine synthesis via oxidative
coupling.
Among various metal oxides, iron oxides have the ad‐
vantages of ready availability, low cost, and low toxicity
[28–32]. However, iron oxides show low activities in the oxida‐
tive coupling of alcohols and amines to imines [25]. This could
be related to the intrinsic inert properties of a bulk material,
such as a low surface‐to‐bulk atomic ratio and low surface en‐
ergy. We recently reported that the redox properties of iron
oxides can be tuned by forming highly dispersed iron oxide
species on carbon supports [33,34]. The catalysts show high
redox activities in O2 activation. This property of these iron
oxide species might enable them to catalyze oxidative coupling
of alcohols and amines to imines.
In this work, carbon‐supported FeOx (FeOx/HCMK‐3) was
used as a catalyst for oxidative coupling of alcohols and amines
to imines. A 98.8% yield of imine was obtained in the oxidative
coupling of benzyl alcohol and aniline after reaction for 6 h at
80 °C. The FeOx/HCMK‐3 catalyst was efficient in the synthesis
of a series of imine compounds. The correlations between the
activity and physicochemical properties of the catalyst were
investigated based on a series of characterization results. The
reaction mechanism of imine synthesis over FeOx/HCMK‐3 was
also explored.
bon–silica composite was obtained by pyrolysis at 900 °C for 6
h under an Ar flow and then washing in 10 wt% HF aqueous
solution to remove the silica template.
The carbon‐supported iron oxide catalyst was prepared us‐
ing a wet impregnation method. Typically, an Fe‐containing
solution was obtained by dissolving Fe(NO3)3·9H2O in a certain
amount of water. Before added to the iron nitrate solution, the
CMK‐3 support was treated in 4 mol/L HNO3 solution at 60 °C
for 6 h (denoted by HCMK‐3). After stirring for about 3 h at
room temperature, the mixture was heated at 80 °C under at‐
mospheric pressure to evaporate the water. The as‐synthesized
catalyst was thermally treated at 400 °C for 4 h under an Ar
flow. The iron oxide loading was 5 wt% (calculated based on
Fe2O3) and the resultant material was denoted by
FeOx/HCMK‐3. For comparison, a sample consisting of Fe3+ ions
supported on HCMK‐3 (denoted by Fe‐HCMK‐3) was also pre‐
pared by the same impregnation method and using
HNO3‐treated CMK‐3 as a support. The difference was that after
impregnation the sample was dried in an oven at 80 °C for 12 h
without further thermal treatment at 400 °C in an Ar flow for 4
h. The Fe species loading was the same as that on
FeOx/HCMK‐3. Bulk Fe2O3 was also prepared using a conven‐
tional precipitation method with Fe(NO)3·9H2O as the Fe
source and ammonia as the precipitating agent (pH 9). The
resultant material was calcined at 400 °C for 4 h in air.
2.3. Catalyst characterization
2. Experimental
Powder X‐ray diffraction (XRD) patterns were recorded
with a Rigaku X‐ray diffractometer using Cu Kα radiation (λ =
1.5418 Å). Transmission electron microscopy (TEM) images
and high‐angle annular dark field scanning TEM
(HAADF‐STEM) images were obtained using a FEI Tecnai F20
instrument with an accelerating voltage of 200 kV, equipped
with an energy‐dispersive X‐ray spectroscopy analyzer. X‐ray
photoelectron spectroscopy (XPS) was performed using a
Thermo ESCA LAB 250 system with an Mg Kα source (1254.6
eV). N2 adsorption‐desorption isotherms were recorded at
−196 °C using a Micromeritics ASAP 2010N analyzer. Raman
spectra were recorded using a Bruker RFS 100 Raman spec‐
trometer with an Ar laser (532 nm) as the excitation source.
Temperature‐programmed reduction (TPR) was performed
using a Tianjin Xianquan TP‐5079 adsorption analyzer. Before
examination, the catalysts (30 mg of FeOx/HCMK‐3, 10 mg of
Fe2O3) were treated in an Ar (99.99%) flow at 400 °C for 30
min. The zeta‐potential curve was plotted as a function of pH,
from 2 to 8 (Zeta PALS analyzer, Brookhaven Instruments
Corporation, USA). The sample was suspended in aqueous 0.01
mol/L NaCl solution. The pH was adjusted using 1.0 mol/L
NaOH and HCl solutions. The Fe content was estimated using
inductively coupled plasma atomic emission spectroscopy
(ICP‐AES; Perkin‐Elmer emission spectrometer). Fourier
transform infrared (FT‐IR) spectra were recorded using a
Thermo Nicolet 6700 FT‐IR spectrometer. The Boehm titration
method was used to determine the amounts of surface groups.
The carbon support (100 mg) was placed in 10 mL of a 0.05
mol/L solution containing NaOH, Na2CO3, and NaHCO3. The
2.1. Materials
All chemicals were analytical grade and used without fur‐
ther purification. Double‐distilled water was used in all exper‐
iments.
2.2. Catalyst preparation
SBA‐15 templates were prepared using a triblock copoly‐
mer, EO20‐PO70‐EO20 (P123; EO: ethylene oxide; PO: propylene
oxide), as the surfactant and tetraethyl orthosilicate (TEOS) as
the silica source, using the procedure reported by Zhao et al.
[35]. Typically, TEOS (4.68 mL) was added to 2 mol/L HCl (60
mL) containing P123 (2 g) at 35 °C. After stirring in a water
bath (35 °C) for 24 h, the mixture was transferred to Tef‐
lon‐lined stainless steel autoclaves (100 mL) and heated to 100
°C for 3 d. The resultant solid product was separated by filtra‐
tion, dried at 100 °C for 12 h, and calcined at 550 °C for 6 h.
CMK‐3 was prepared using SBA‐15 as a hard template and
sucrose as the carbon source [36]. Typically, SBA‐15 silica (1.0
g) was added to a solution containing sucrose (1.25 g), sulfuric
acid (0.14 g), and distilled water (5.0 g). The mixture was kept
in an oven for 6 h at 80 °C. The temperature was then increased
to 160 °C and maintained for 6 h. The heating procedure was
repeated after addition of the carbon precursor (0.8 g of su‐
crose, 0.09 g of H2SO4, and 5.0 g of H2O) to achieve complete
infiltration of the internal pores of the SBA‐15 silica. The car‐

Longlong Geng et al. / Chinese Journal of Catalysis 37 (2016) 1451–1460
1453
vials were sealed and shaken for 24 h, and the contents were
filtered. A known quantity of HCl was added to each filtrate.
The excess acid left in the solution was titrated with NaOH us‐
ing an automatic potentiometer titrator (LeiCi ZDT‐4A). The
amounts of acidic sites were calculated based on the assump‐
tion that NaOH neutralizes carboxylic, phenolic, and lactonic
groups, NaHCO3 neutralizes carboxylic groups, and Na2CO3
neutralizes carboxylic and lactonic groups.
2.4. Catalytic tests
The aerobic oxidative coupling of an alcohol and amine was
performed in a 50‐mL two‐necked flask at atmospheric pres‐
sure. A quantity of catalyst (0.3 g) was added to the reactor
containing alcohol (1.0 mmol), amine (2.0 mmol), and toluene
(10 mL). The mixture was in contact with air, and the reaction
temperature was kept at 80 °C. The reaction mixture was ex‐
tracted via a sampling pipe with a filtrator and then transferred
to a vial. The products were analyzed using a gas chromato‐
graph equipped with an HP‐5 column and a flame ionization
detector. The conversion and yield of the imine product were
calculated based on the alcohol to imine ratio. The products
were identified using standard compounds and gas chroma‐
tography‐mass spectrometry. The turnover frequencies (TOFs)
were calculated based on benzyl alcohol conversion in the ini‐
tial stage (60 min) and the amount of FeOx in the catalyst.
An oxygen‐free experiment was performed in a Schlenk
tube. First, the FeOx/HCMK‐3 catalyst was put in a Schlenk
tube, and degassed at 80 °C for 1.5 h under vacuum to remove
adsorbed O2. High‐purity N2 (99.99%) was injected into the
Schlenk tube until the pressure was 1 atm. The degassed tolu‐
ene, benzyl alcohol, and aniline were added to the Schlenk tube.
The system was further degassed and balanced with high puri‐
ty N2. The Schlenk tube was then put in an oil bath (80 °C) to
start the reaction. After each sample extraction, the system was
degassed and balanced with high purity N2. All these treat‐
ments were to ensure the removal of molecular O2 from the
system. In the recycling experiments, the solid catalyst was
separated by filtration and treated at 400 °C for 1 h before the
next cycle.
Fig. 1. Schematic illustration of synthetic route for FeOx/HCMK‐3 prep‐
aration.
abundant functional groups endow the HCMK‐3 support with a
hydrophilic surface. This facilitates introduction of iron nitrate
solution into the mesopores of the carbon support. The dried
solid was calcined at 400 °C for 4 h in an Ar flow to give the
FeOx/HCMK‐3 catalyst.
The XRD patterns show that FeOx/HCMK‐3 retains the or‐
dered arrangement of the parent HCMK‐3 support (Fig. 2a).
The three diffraction peaks in the range 2θ = 0.75°–3° can be
indexed to the (100), (110), and (200) reflections of the
HCMK‐3 hexagonal space group. Wide‐angle XRD patterns
show that both samples have two broad diffraction peaks, cen‐
tered at 24.5° and 44° (Fig. 2b), which are assigned to diffrac‐
tions from the (002) and (100) graphite planes of the carbon
supports [37]. No crystalline phase related to iron oxides was
observed, indicating that the iron oxide particles are small or
the degree of crystallization is low. The Raman spectra of
FeOx/HCMK‐3 and HCMK‐3 show two peaks, at 1358 and 1598
cm−1, which are assigned to the D and G bands of the carbon
support (Fig. 2c) [38,39]. No peak below 1000 cm−1 attributable
to crystalline iron oxides can be observed [40,41]; this also
indicates that the iron oxide particles in FeOx/HCMK‐3 are
small. The N2 adsorption‐desorption isotherms show that
FeOx/HCMK‐3 gives isotherms typical of an ordered
mesostructure (type IV) with a well‐defined step in the adsorp‐
tion curve near p/p0 = 0.5 (Fig. 2d). The surface area, pore
volume, and pore size of FeOx/HCMK‐3 are similar to those of
HCMK‐3, indicating that the iron oxides do not block the mes‐
opores of the CMK‐3 support.
3. Results and discussion
3.1. Catalyst structure and surface properties
Fig. 1 shows the preparation of the carbon‐supported FeOx
(5 wt%, calculated based on Fe2O3) catalyst. The catalyst sup‐
port was ordered mesoporous carbon CMK‐3 prepared via a
hard‐template method. The CMK‐3 was treated with 4 mol/L
HNO3 at 60 °C for 6 h before use as the catalyst support. The
resultant carbon was denoted by HCMK‐3. This was a key step
in preparing a catalyst with highly dispersed FeOx species. A
large amount of oxygen‐containing functional groups formed
on the HCMK‐3 surface. Table 1 shows that the amounts of
functional groups on HCMK‐3, including carboxylic, lactonic,
and phenolic groups, were all higher than those on CMK‐3. The
total amount of functional groups was about 1.47 mmol/g. The
TEM images confirm that FeOx/HCMK‐3 retains the ordered
structure of the CMK‐3 support (Fig. 3a). No Fe particles were
observed in the region examined, even at high resolution (Fig.
Table 1
Boehm titration results for CMK‐3 and HCMK‐3.
Content of functional groups (mmol/g)
Sample
Carboxylic
Lactonic
Phenolic
Acidic
CMK‐3
HCMK‐3
0.01
0.24
0.11
0.34
0.12
0.89
0.24
1.47

1454
Longlong Geng et al. / Chinese Journal of Catalysis 37 (2016) 1451–1460
(a)
(b)
HCMK-3
HCMK-3
FeOx/HCMK-3
FeOx/HCMK-3
1.0
1.5
2.0
2/(o )
2.5
3.0
3.5
10
20
30
40
2/(o )
50
60
70
800
(d)
(c)
600
400
200
0
HCMK-3
FeOx/HCMK-3
0
4
8
12
16
PPooreredDiaimameeteterr((nnmm))
500
1000
1500
2000
2500
3000
0.0
0.2
0.4
0.6
0.8
1.0
Raman shift (cm1)
Relative pressure (p/p0)
Fig. 2. Low‐angle (a) and wide‐angle (b) XRD patterns for HCMK‐3 and FeOx/HCMK‐3; (c) Raman spectra of HCMK‐3 and FeOx/HCMK‐3; (d) N2 ad‐
sorption‐desorption isotherms of FeOx/HCMK‐3 (inset shows pore size distribution).
3b and c). STEM and energy‐dispersive X‐ray mapping were
performed to investigate iron oxide dispersion (Fig. 3d and e).
The results show that Fe species are nearly homogeneously
dispersed on the HCMK‐3 surface, but the particle size still
cannot be ascertained based on these measurements. A sample
with a low FeOx loading (0.25 wt%) was prepared to clarify the
FeOx dispersed state on the HCMK‐3 surface. FeOx particles still
cannot be observed in the HRTEM image (not shown), but
some light spots (about 2–3 nm) can be distinguished in the
HAADF‐STEM image (Fig. 3f). These results suggest that FeOx
species are present as two‐dimensional or even sub‐nano clus‐
ters. These clusters could be the constituent units of the highly
dispersed FeOx on the carbon support. The HRTEM and
HAADF‐STEM images together confirm that the FeOx species
are highly dispersed on the HCMK‐3 support surface.
XPS shows that there are no obvious differences between
the Fe 2p binding energies of FeOx/HCMK‐3 and bulk Fe2O3
(Fig. 4a). The Fe 2p1/2 and Fe 2p3/2 peaks appear at 724.2 and
710.9 eV, respectively. This indicates that most of the Fe spe‐
cies are present in the iron oxides in the +3 valence state
[42,43]. Clear differences can be observed in the O 1s spectra.
The binding energies in the deconvoluted O 1s XPS spectrum
are in according with those reported in the literature [44,45].
The O 1s spectrum of bulk Fe2O3 can be deconvoluted into two
peaks (Fig. 4b). Peak I, in the range 529.5–530.5 eV, is assigned
to O2− ions in the iron oxide lattice. Peak II, centered at about
532 eV, is related to oxygen in surface OH groups or oxygen
vacancies in Fe2O3. The peak II/peak I area ratio is 0.21 for bulk
Fe2O3. The FeOx/HCMK‐3 O 1s spectrum can be deconvoluted
into three peaks. The new peak (peak III) at 534 eV is attribut‐
ed to oxygen functional groups remaining on the catalyst sur‐
face. The peak II/peak I area ratio is 3.54 for FeOx/HCMK‐3,
which is much higher than that for bulk Fe2O3. This is caused by
the small size of the iron oxide particles and interactions be‐
tween iron oxides and the carbon support.
Diffuse reflectance infrared Fourier transform (DRIFT)
spectra show that some oxygen‐containing groups remain on
the FeOx/HCMK‐3 surface (Fig. 5a). Further FT‐IR spectra were
obtained to clarify whether the Fe species were located in iron
oxides or present as Fe3+ ions coordinated with these functional
groups. A sample consisting of Fe3+ ions supported on HCMK‐3
(denoted by Fe‐HCMK‐3) was prepared via the same impreg‐
nation method as was used to prepare FeOx/HCMK‐3, with
HNO3‐treated CMK‐3 as a support. The Fe species loading was
the same as that for FeOx/HCMK‐3. The difference is that after
impregnation, the sample was dried in an oven at 80 °C for 12 h
to obtain Fe‐HCMK‐3, whereas further thermal treatment at
400 °C in an Ar flow for 4 h is needed to obtain FeOx/HCMK‐3.
The peak at 469 cm−1 in the FT‐IR spectrum of FeOx/HCMK‐3
(Fig. 5b) can be attributed to the stretching vibration of Fe–O,
similar to that of Fe2O3. However, this signal is different in the
Fe‐HCMK‐3 spectrum. A combination of the FT‐IR and XPS re‐
sults confirms that Fe is mainly located in iron oxide species
(i.e., FeOx). The formation of FeOx species in FeOx/HCMK‐3 is a
result of the high treatment temperature. The oxygen‐contain‐
ing functional groups in HCMK‐3 can interact with these
two‐dimensional FeOx species.
The H2‐TPR profiles (Fig. 6) show that the iron oxide species
in FeOx/HCMK‐3 are more reducible than those in bulk Fe2O3.
There are three reduction peaks present in the profile of bulk

Longlong Geng et al. / Chinese Journal of Catalysis 37 (2016) 1451–1460
1455
(a)
(a)
(b)
Fe 2p
FeOx/HCMK-3
Fe2O3
100 nm
(d)
740
735
730
725
720
715
710
705
(c)
Binding energy (eV)
(b)
Peak II
O 1s
Peak III
Peak I
FeOx/HCMK-3
C
(f)
(e)
Fe2O3
538
536
534
532
530
528
526
200 nm
200 nm
Binding energy (eV)
Fe
O
Fig. 4. Fe 2p (a) and O 1s (b) XPS spectra of FeOx/HCMK‐3 and bulk
Fe2O3.
20 nm
TPR were 2.13 and 18.10 mmol/g, respectively. The FeOx con‐
tent in FeOx/HCMK‐3 was 5 wt%, therefore the stoichiometric
consumption of H2 by FeOx should be 0.93 mmol/g, which is
lower than the actual amount of H2 consumed (2.13 mmol/g). A
combination of the H2‐TPR, XPS, and DRIFT results shows that
additional H2 is consumed (1.2 mmol/g) because of the pres‐
ence of oxygen‐containing functional groups in FeOx/HCMK‐3.
The quantity of reduced functional groups in FeOx/HCMK‐3
was determined based on H2‐TPR of HCMK‐3. The results show
that H2 consumption by HCMK‐3 was about 1.08 mmol/g,
which is consistent with H2 consumption by the carbon support
in FeOx/HCMK‐3 (1.2 mmol/g).
200 nm
200 nm
Fig. 3. TEM (a), HRTEM (b, c), HAADF‐STEM (d), and energy‐dispersive
X‐ray mapping images (e) of FeOx/HCMK‐3; (f) HAADF‐STEM image of
0.25wt% FeOx/HCMK‐3 catalyst.
Fe2O3. The narrow peak (centered at 430 °C) can be ascribed to
the reduction of Fe2O3 to Fe3O4. The two overlapping peaks in
the range 450–800 °C can be ascribed to further reduction to
FeO/Fe species [46,47]. For FeOx/HCMK‐3, only one broad
peak is present in the range 200–600 °C. The start and center of
this peak are clearly at lower temperatures than those for bulk
Fe2O3, confirming the high reducibility of FeOx/HCMK‐3. The
amounts of H2 consumed by FeOx/HCMK‐3 and Fe2O3 during
These results show that oxygen functional groups play im‐
(a)
(b)
Fe2O3
HCMK-3
FeOx/HCMK-3
Fe-HCMK-3
FeOx/HCMK-3
CMK-3
3500
3000
2500
2000
1500
1000
1400
1200
1000
Wavenumber (cm1)
Fig. 5. (a) DRIFT spectra of CMK‐3, HCMK‐3, and FeOx/HCMK‐3; (b) FT‐IR spectra of FeOx/HCMK‐3, Fe‐HCMK‐3, and Fe2O3.
800
600
400
Wave number (cm1)

1456
Longlong Geng et al. / Chinese Journal of Catalysis 37 (2016) 1451–1460
100
80
60
40
20
0
FeOx/HCMK-3
Fe2O3
FeOx/HCMK-3
Fe2O3
HCMK-3
0
2
4
6
8
100
200
300
400
500
600
700
800
Reaction time (h)
Temperature (C)
Fig. 8. Time dependences of imine formation from aerobic oxidative
coupling of benzyl alcohol and aniline over FeOx/HCMK‐3, bulk Fe2O3,
and HCMK‐3. Reaction conditions: benzyl alcohol 1 mmol, aniline 2
mmol, catalyst 0.3 g, toluene 10 mL, 80 °C, air 1 bar.
Fig. 6. H2‐TPR profiles of FeOx/HCMK‐3 and bulk Fe2O3.
portant roles in the formation of highly dispersed iron oxide
species. They render the HCMK‐3 support surface negatively
charged. Zeta‐potential measurements show that HCMK‐3 is
negatively charged at pH 1.0–8.0 (Fig. 7). Untreated CMK‐3 is
positively charged at pH less than 5.1. FeOx/HCMK‐3 was pre‐
pared by impregnation with Fe(NO3)3 aqueous solution. The pH
of the Fe3+‐containing solution was 1.65 (Fig. 7). Under these
preparation conditions, there is an electrostatic force between
the positively charged metal ions and negatively charged car‐
bon surface (Fig. 1). This could be the main reason for the ho‐
mogeneous dispersion of Fe ions on the carbon surface. In the
subsequent thermal treatment, these surface oxygen functional
groups provide strongly active sites for anchoring FeOx species,
and suppress iron oxide aggregation.
of imine after reaction for 6 h was 98.8%. The TOF for
FeOx/HCMK‐3 in the initial stage was 2.9 h−1. The catalytic per‐
formance of FeOx/HCMK‐3 at 60 °C was also investigated for
comparison with those of other reported catalytic systems. The
TOF was 0.6 h−1. Comparisons with results reported in the lit‐
erature [19,25,26] show that this TOF is higher than those for
non‐noble‐metal oxide catalysts (e.g., CeO2 and MnOx/HAP), but
a little lower than those of supported Au catalysts under similar
reaction conditions. The reaction was also performed using
equimolar amounts of the alcohol and amine. A 78.4% yield of
the imine was obtained after reaction for 8 h. All these results
indicate that FeOx/HCMK‐3 is an effective catalyst for imine
formation.
Multiple aerobic oxidation cycles were performed to exam‐
ine the recoverability of FeOx/HCMK‐3 (Fig. 9). The catalyst
retained high activity after five cycles, with only 3% and 5%
decreases in the third and fifth cycles compared with the first
one. Loss of Fe from the catalyst to the reaction mixture was
determined using ICP‐AES. Trace amounts of Fe (0.02 ppm,
0.02‰ of total Fe) were detected, indicating that Fe active sites
3.2. Catalytic performance
Fig. 8 shows the catalytic activity of FeOx/HCMK‐3 in the
oxidative coupling of benzyl alcohol and aniline to an imine at
80 °C. The imine was identified using gas chromatog‐
raphy‐mass spectrometry. For comparison, the reaction was
also performed using bulk Fe2O3 and HCMK‐3. The activity of
bulk Fe2O3 was low. Only a 13.2% yield of imine was obtained
after reaction for 8 h at 80 °C. The HCMK‐3 support was almost
inactive under the test conditions. For FeOx/HCMK‐3, the yield
100
80
60
40
20
0
40
Fe3+ solution
CMK-3
30
pH = 1.65
HCMK-3
20
10
0
-10
-20
-30
-40
1
2
3
4
5
Cycle times
1
2
3
4
5
6
7
8
Fig. 9. Reusability of FeOx/HCMK‐3 in imine synthesis by aerobic oxida‐
tive coupling of benzyl alcohol and aniline. Reaction conditions: benzyl
alcohol 1 mmol, aniline 2 mmol, catalyst 0.3 g, toluene 10 mL, 80 °C, air
1 bar, 3 h. Recovered catalysts were treated under Ar at 400 °C for 1 h.
pH value
Fig. 7. Zeta‐potential profiles for CMK‐3 and HCMK‐3; pH is the value
for impregnation process, i.e., iron nitrate solution containing HCMK‐3.

Longlong Geng et al. / Chinese Journal of Catalysis 37 (2016) 1451–1460
1457
were not lost. The XRD patterns show that there was no aggre‐
Table 3
gation of iron oxide species in the used FeOx/HCMK‐3. All these
results suggest that iron oxide active sites are stable in the
FeOx/HCMK‐3 catalyst. The decrease in the imine yield was
mainly caused by loss of catalyst in the recycling process. The
catalyst was recycled by filtration and then thermally treated at
400 °C for 1 h under Ar to remove adsorbed reagents. Catalyst
loss (about 7%) is inevitable during this process.
Aerobic oxidative coupling of various alcohols and anilines over
FeOx/HCMK‐3 catalyst.
R2
H2O
FeOx/HCMK-3 catalyst
80 oC Air
+
Yield (%)
98.0
R2-NH2
R1
R1
+
OH
Entry
Product
Time (h)
6
Ph
N
1
2
3
Experiments were performed to identify the active sites in
FeOx/HCMK‐3 (Table 2). When Fe(NO3)3 was used as the cata‐
lyst, the yield of imine after reaction for 8 h was 1.8%. Even in
the co‐presence of Fe(NO3)3 and HCMK‐3, the imine yield was
only 5.0%. These results differ from those reported by Zhang et
al. [48], mainly because different reaction conditions were
used. Under their reaction conditions, Fe(NO3)3 shows moder‐
ate activity in imine synthesis and the activity can be greatly
enhanced by addition of 2,2,6,6‐tetramethyl‐1‐piperidinyloxy
[48]. We also investigated the catalytic performance of
Fe‐HCMK‐3. An imine yield of 24.6% was obtained over
Fe‐HCMK‐3 after reaction for 8 h; this is much lower than the
yield achieved using FeOx/HCMK‐3. Loss of Fe into the reaction
mixture was detected using ICP‐AES; the amount was about
33.65 ppm, which is 3.2% of the total amount of Fe loaded on
Fe‐HCMK‐3. All these results suggest that Fe3+ ions or coordi‐
nated Fe species were not active in the imine synthesis reac‐
tion. Supported FeOx species were the effective and stable ac‐
tive sites. The highly dispersed FeOx species on the HCMK‐3
surface could affect the redox properties of the FeOx/HCMK‐3
catalyst. The reducibility of these iron oxide species could be
responsible for the high activity in this reaction.
Ph
N
8
8
93.2
90.7
Ph
Ph
N
N
O
4
8
84.2
Ph
N
5
6
7
8
10
10
80.7
75.1
64.2
O3N
Ph
Ph
N
HO
Cl
N
N
Ph
8 *
9 *
10
12
12
8
55.9
36.4
85.8
Ph
N
The general applicability of the FeOx/HCMK‐3 catalyst was
also investigated. Eleven imines were obtained by oxidative
coupling of amines with various alcohols. Table 3 (entries 1–7)
shows that the FeOx/HCMK‐3 catalyst is active in reactions
using benzyl alcohol derivatives with either electron‐rich or
electron‐poor substituents, especially 4‐methylbenzyl alcohol,
4‐methoxybenzyl alcohol, and 4‐isopropylbenzyl alcohol (the
imine yields were 93.2%, 90.7%, and 84.2%, respectively, after
reaction for 8 h). FeOx/HCMK‐3 also catalyzed the reaction of
phenethyl alcohol and allylic alcohols (e.g., trans‐2‐hexenol)
with aniline to give the corresponding imines when the reac‐
tion temperature and time were increased (Table 3, entries 8
and 9). Table 3 shows that FeOx/HCMK‐3 is active in the aero‐
bic oxidative coupling of aliphatic amines (e.g., cyclohexylamine
Ph
N
11
8
80.6
Reaction conditions: alcohol 1 mmol, amine 2 mmol, catalyst 0.3 g,
toluene 10 mL, 80 °C, air 1 bar. *Temperature: 100 °C.
and n‐butylamine) with benzyl alcohol (Table 3, entries 10 and
11). These results clearly show that FeOx/HCMK‐3 is an effi‐
cient catalyst in oxidative coupling reactions.
Three additional experiments were performed to clarify the
reaction mechanism over FeOx/HCMK‐3; the results are shown
in Fig. 10. Fig. 10a shows that the oxygen species in
FeOx/HCMK‐3 are active in the aerobic oxidative reaction. Even
without the presence of O2, an 8.7% yield of imine was ob‐
tained after reaction for 8 h at 80 °C. This experiment was per‐
formed in a Schlenk tube (see the experimental section for de‐
tails). The FeOx/HCMK‐3 catalyst was first degassed at 80 °C for
1.5 h under vacuum to remove adsorbed O2. After adding the
reactants and solvent, the system was further degassed and
then balanced with high‐purity N2. All these treatments were
performed to ensure the removal of molecular O2 from the sys‐
tem. This result shows that the oxygen species in FeOx/HCMK‐3
participate in the reaction. The oxidative coupling of alcohols
and amines on FeOx/HCMK‐3 occur via a redox mechanism
[25,26]. During imine synthesis over FeOx/HCMK‐3, benzalde‐
hyde was detected in the initial stage and the amount slowly
decreased with time (Fig. 10b). This suggests that benzalde‐
Table 2
Imine synthesis by oxidative coupling of benzyl alcohol and aniline over
various catalysts.
Entry
1
2
3*
4*
5
6
7
Catalyst
FeOx/HCMK‐3
None
Time (h)
Yield (%)
98.6
0.9
1.8
5.0
24.6
13.2
5.1
6
8
8
8
8
8
8
Fe(NO3)3
Fe(NO3)3 + HCMK‐3
Fe‐HCMK‐3
Fe2O3
HCMK‐3
Reaction conditions: benzyl alcohol 1 mmol, aniline 2 mmol, catalyst 0.3
g, toluene 10 mL, 80 °C, air 1 bar. *The amount of Fe(NO3)3 was calcu‐
lated based on the Fe content in FeOx/HCMK‐3.

1458
Longlong Geng et al. / Chinese Journal of Catalysis 37 (2016) 1451–1460
100
100
80
60
40
20
0
100
80
60
40
20
0
(c)
(a)
(b)
FeOx/HCMK-3
Air atmosphere
80
60
40
20
0
Imine yield
Benzaldehyde yield
None
1.5
N2 atmosphere
0
2
4
6
8
0
1
2
3
4
5
6
0.0
0.5
1.0
2.0
Reaction time (h)
Reaction time (h)
Reaction time (h)
Fig. 10. (a) Effects of catalytic conditions on performance of FeOx/HCMK‐3 in imine formation by aerobic oxidative coupling of benzyl alcohol and
aniline. Reaction conditions: benzyl alcohol 1 mmol, aniline 2 mmol, catalyst 0.3 g, toluene 10 mL, 80 °C, air or N2 1 bar. (b) Time dependence of imine
formation by oxidative coupling of benzyl alcohol and aniline over FeOx/HCMK‐3. Reaction conditions: benzyl alcohol 1 mmol, aniline 2 mmol, catalyst
0.3 g, toluene 10 mL, 80 °C, air 1 bar. (c) Time dependences of imine formation by coupling of benzaldehyde and aniline with and without
FeOx/HCMK‐3 catalyst. Reaction conditions: benzaldehyde 1 mmol, aniline 2 mmol, catalyst 0.3 g, toluene 10 mL, 80 °C, air 1 bar.
hyde might be an intermediate and that the reaction proceeds
in two consecutive steps: (1) oxidative dehydrogenation of the
alcohol to benzaldehyde and (2) imine formation by reaction of
benzaldehyde with aniline.
tion of benzyl alcohol by oxygen species occurs at the redox
sites of FeOx/HCMK‐3. Iron is reduced from valence +3 to +2.
The produced benzaldehyde reacts with aniline over
FeOx/HCMK‐3 to afford the corresponding imine.
FeOx/HCMK‐3 is regenerated by oxidation of the reduced
FeOx/HCMK‐3 with O2.
We recently reported that FeOx/HCMK‐3 catalyzes the oxi‐
dation of benzyl alcohol to benzaldehyde with air as the oxygen
source [33]. Benzyl alcohol conversion of 72% was observed
after reaction for 8 h. In the imine synthesis, benzyl alcohol
conversion improved to 98.8% after reaction for 6 h. The sec‐
ond step in imine formation accelerates the conversion of ben‐
zyl alcohol. Fig. 10c shows that FeOx/HCMK‐3 efficiently cata‐
lyzes imine synthesis from benzaldehyde and aniline. The reac‐
tion rate is 372.8 mmol h−1 g−1 (calculated based on iron oxide
sites), which is 4.5 times higher than that for imine formation
from benzyl alcohol and aniline (82.0 mmol h−1 g−1). This con‐
firms that oxidative dehydrogenation of alcohol is the
rate‐determining step in the reaction. Based on the above re‐
sults and related literature reports [49,50], a plausible reaction
mechanism for imine formation from an alcohol and aniline on
FeOx/HCMK‐3 is proposed (Scheme 1). Oxidative dehydrogena‐
4. Conclusions
FeOx/HCMK‐3 with highly dispersed iron oxide species
shows high activity and recyclability in aerobic oxidative cou‐
pling of alcohols and amines to imines. The excellent catalytic
performance of FeOx/HCMK‐3 is ascribed to its high reducibil‐
ity. The lattice oxygens in FeOx/HCMK‐3 can participate in
imine synthesis, and the reaction follows a redox mechanism
over FeOx/HCMK‐3. Two consecutive steps are involved in the
reaction: oxidative dehydrogenation of an alcohol to an alde‐
hyde and coupling of the aldehyde with an amine to an imine.
Oxidative dehydrogenation of alcohol is the rate‐determining
step in the reaction.
NH2
O
N
+
H2O
+ H2O
O
O
Fe
Fe
Fe
O
O
O
III
Fe
II
Fe
III
Fe
III
Fe
Fe
Fe
Carbon support
Oxygen vacancy
O2
Scheme 1. Proposed reaction mechanism for imine formation over FeOx/HCMK‐3.

Longlong Geng et al. / Chinese Journal of Catalysis 37 (2016) 1451–1460
1459
Huang, X. Y. Liu, W. G. Liu, J. Y. Son, H. Oji, T. Zhang, Green Chem.,
2013, 15, 2680–2684.
References
[1] Z. B. Han, Z. Wang, X. M. Zhang, K. L. Ding, Angew. Chem. Int. Ed.,
2009, 48, 5345–5349.
[25] M. Tamura, K. Tomishige, Angew. Chem. Int. Ed., 2015, 54,
864–867.
[2] A. J. A. Watson, J. M. J. Williams, Science, 2010, 329, 635–636.
[3] J. Gawronski, N. Wascinska, J. Gajewy, Chem. Rev., 2008, 108,
5227–5252.
[4] S. I. Murahashi, Angew. Chem. Int. Ed., 1995, 34, 2443–2465.
[5] J. F. Soulé, H. Miyamura, S. Kobayashi, Asian J. Org. Chem., 2012, 1,
319–321.
[6] R. D. Patil, S. Adimurthy, Asian J. Org. Chem., 2013, 2, 726–744.
[7] V. Kumar, S. Sharma, U. Sharma, B. Singh, N. Kumar, Green Chem.,
2012,14, 3410–3414.
[8] L. Ravishankar, S. A. Patwe, N. Gosarani, A. Roy, Synth. Commun.,
2010, 40, 3177–3180.
[26] B. Chen, J. Li, W. Dai, L. Y. Wang, S. Gao, Green Chem., 2014, 16,
3328–3334.
[27] L. Y. Wang, B. Chen, L. H. Ren, H. Y. Zhang, Y. Lü, S. Gao, Chin. J.
Catal., 2015, 36, 19–23.
[28] A. H. Lu, J. J. Nitz, M. Comotti, C. Weidenthaler, K. Schlichte, C. W.
Lehmann, O. Terasaki, F. Schüth, J. Am. Chem. Soc., 2010, 132,
14152–14162.
[29] H. M. Torres Galvis, J. H. Bitter, C. B. Khare, M. Ruitenbeek, A. I.
Dugulan, K. P. de Jong, Science, 2012, 355, 835–838.
[30] Z. X. Wu, W. Li, P. A. Webley, D. Y. Zhao, Adv. Mater, 2012, 24,
485–491.
[9] J. S. M. Samec, A. H. Éll, J. E. Bäckvall, Chem. Eur. J, 2001, 11,
2327–2334.
[10] B. Chen, L. Y. Wang, S. Gao, ACS Catal., 2015, 5, 5851–5876.
[11] Q. Kang, Y. G. Zhang, Green Chem., 2012, 14, 1016–1019.
[12] S. Sithambaram, R. Kumar, Y. C. Son, S. L. Suib, J. Catal., 2008, 253,
269–277.
[31] A. R. Rajenahally, 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.
[32] S. A. Cotton, Annu. Rep. Prog. Chem., Sect. A, 2005, 101, 208–230.
[33] L. L. Geng, X. Y. Zhang, W. X. Zhang, M. J. Jia, G. Liu, Chem. Commun.,
2014, 50, 2965–2967.
[13] R. Cano, D. J. Ramón, M. Yus, J. Org. Chem., 2011, 76, 5547–5557.
[14] Z. X. Zhang, Y. H. Wang, M. Wang, J. M. Lü, L. H. Li, Z. Zhang, M. R. Li,
J. Y. Jiang, F. Wang, Chin. J. Catal., 2015, 36, 1623–1630.
[15] H. W. Tian, X. C. Yu, Q. Li, J. X. Wang, Q. Xu, Adv. Synth. Catal., 2012,
354, 2671–2677.
[16] B. Gnanaprakasam, J. Zhang, D. Milstein, Angew. Chem., 2010, 122,
1510–1513.
[17] J. F. Soulé, H. Miyamura, S. Kobayashi, J. Am. Chem. Soc., 2011, 133,
18550–18553.
[18] L. Tang, H. Y. Sun, Y. F. Li, Z. G. Zha, Z. Y. Wang, Green Chem., 2012,
14, 3423–3428.
[19] H. Sun, F. Z. Su, J. Ni, Y. Cao, H. Y. He, K. N. Fan, Angew. Chem. Int.
Ed., 2009, 48, 4390–4393.
[20] W. He, L. D. Wang, C. Y. Sun, K. K. Wu, S. B. He, J. P. Chen, P. Wu, Z.
K. Yu, Chem. Eur. J., 2011, 17, 13308–13317.
[21] D. Gnanamgari, E. L. O. Sauer, N. D. Schley, C. Butler, C. D. Incarvito,
R. H. Crabtree, Organometallics, 2009, 28, 321–325.
[22] A. Corma, T. Ródenas, M. J. Sabater, Chem. Eur. J., 2010, 16,
254–260.
[34] L. L. Geng, M. Zhang, W. X. Zhang, M. J. Jia, W. F. Yan, G. Liu, Catal.
Sci. Technol., 2015, 5, 3097–3102.
[35] D. Y. Zhao, J. L. Feng, Q. S. Huo, N. Melosh, G. H. Fredrickson, B. F.
Chmelka, G. D. Stucky, Science, 1998, 279, 548–552.
[36] S. Jun, S. H. Joo, R. Ryoo, M. Kruk, M. Jaroniec, Z. Liu, T. Ohsuna, O.
Terasaki, J. Am. Chem. Soc., 2000, 122, 10712–10713.
[37] G. Liu, Y. Liu, Z. L. Wang, X. Z. Liao, S. J. Wu, W. X. Zhang, M. J. Jia,
Microporous Mesoporous Mater., 2008, 116, 439–444.
[38] L. L. Geng, S. J. Wu, Y. C. Zou, M. J. Jia, W. X. Zhang, W. F. Yan, G. Liu,
J. Colloid Interface Sci., 2014, 421, 71–77.
[39] M. Zhang, W. C. Zhu, G. Liu, X. Y. Zhang, Y. H. Zu, W. X. Zhang, W. F.
Yan, M. J. Jia, Chin. J. Catal., 2012, 33, 465–472.
[40] X. Zhang, Y. A. Niu, X. D. Meng, Y. Li, J. D. Zhao, CrystEngComm,
2013, 15, 8166–8172.
[41] J. L. Tu, M. Y. Ding, Q. Zhang, Y. L. Zhang, C. G. Wang, T. J. Wang, L. L.
Ma, X. J. Li, ChemCatChem, 2015, 7, 2323–2327.
[42] J. Yang, H. W. Zhang, M. H. Yu, I. Emmanuelawati, J. Zou, Z. G. Yuan,
C. Z. Yu, Adv. Funct. Mater., 2014, 24, 1354–1363.
[43] M. Monti, B. Santos, A. Mascaraque, O. R. Figuera, M. A. Niño, T. O.
Menteş, A. Locatelli, K. F. McCarty. J. F. Marco, J. de la Figuera, J.
Phys. Chem. C, 2012, 116, 11539 –11547.
[23] S. Kegnaes, J. Mielby, U. V. Mentzel, C. H. Christensen, A. Riisager,
Green Chem., 2010, 12, 1437–1441.
[24] L. L. Zhang, W. T. Wang, A. Q. Wang, Y. T. Cui, X. F. Yang, Y. Q.
[44] M. D. Sánchez, P. Chen, T. Reinecke, M. Muhler, W. Xia,
Graphical Abstract
Chin. J. Catal., 2016, 37: 1451–1460 doi: 10.1016/S1872‐2067(16)62506‐8
Aerobic oxidative coupling of alcohols and amines to imines
over iron catalysts supported on mesoporous carbon
Longlong Geng,Jinling Song, Bin Zheng,Shujie Wu,Wenxiang
Zhang, Mingjun Jia, Gang Liu *
Jilin University
Mesoporous‐carbon‐supported iron oxide is a highly efficient and
recyclable catalyst for imine synthesis by oxidative coupling of
alcohols and amines with air as the oxygen source.

1460
Longlong Geng et al. / Chinese Journal of Catalysis 37 (2016) 1451–1460
ChemCatChem, 2012, 4, 1997–2004.
J. Catal., 2013, 299, 90–100.
[45] Q. Q. Zhuo, J. Gao, M. F. Peng, L. D. Bai, J. J. Deng, Y. J. Xia, Y. Y. Ma, J.
Zhong, X. H. Sun, Carbon, 2013, 52, 559–564.
[48] E. Zhang, H. W. Tian, S. D. Xu, X. C. Yu, Q. D. Xu, Org. Lett., 2013, 15,
2704–2707.
[46] X. L. Mou, B. S. Zhang, Y. Li, L. D. Yao, X. J. Wei, D. S. Su, W. J. Shen,
Angew. Chem. Int. Ed., 2012, 51, 2989–2993.
[47] L. Li, A. Q. Wang, B. T. Qiao, J. Lin, Y. Q. Huang, X. D. Wang, T. Zhang,
[49] P. N. Amaniampong, Q. T. Trinh, B. Wang, A. Borgna, Y. Yang, S. H.
Mushrif, Angew. Chem. Int. Ed., 2015, 54, 8928–8933.
[50] U. Menon, V. V. Galvita, G. B. Marin, J. Catal., 2011, 283, 1–9.