Direct amination of benzene to aniline with H2O2 and NH3·H2O over Cu/SiO2 catalyst

Article abstract of DOI:10.1039/c4cy00432a

The direct amination of benzene to aniline with H₂O₂ and NH₃·H₂O was studied using a series of Cu/SiO₂ catalysts with mesoporous structures under mild conditions, and an acceptable yield (5.4%) and selectivity (74.0%) to aniline was obtained. Ultraviolet-visible diffuse reflectance spectroscopy, Fourier transform infrared spectroscopy and X-ray photoelectron spectroscopy demonstrated that some of the Cu atoms were incorporated into the mesoporous silica network forming Cu-O-Si species, which enabled the selective activation of the N-H bond of NH ₃. The medium-strong acid sites were beneficial for the amination reaction. The concentration of the reactants was found to be another critical factor for aniline formation. The amination of the aromatic ring of substituted benzenes was also investigated under the optimized conditions. the Partner Organisations 2014.

Full text of DOI:10.1039/c4cy00432a

Catalysis
Science &
Technology
View Article Online
View Journal | View Issue
PAPER
Direct amination of benzene to aniline with H2O2
and NH3·H2O over Cu/SiO2 catalyst†
Cite this: Catal. Sci. Technol., 2014,
4, 3159
a
a
a
a
Tianhua Yu,ab Ruiguang Yang, Sheng Xia, Guiying Li and Changwei Hu
*
*
The direct amination of benzene to aniline with H2O2 and NH3·H2O was studied using a series of Cu/SiO2
catalysts with mesoporous structures under mild conditions, and an acceptable yield (5.4%) and selectivity
(74.0%) to aniline was obtained. Ultraviolet-visible diffuse reflectance spectroscopy, Fourier transform
infrared spectroscopy and X-ray photoelectron spectroscopy demonstrated that some of the Cu atoms
were incorporated into the mesoporous silica network forming Cu–O–Si species, which enabled the
selective activation of the N–H bond of NH3. The medium-strong acid sites were beneficial for the
amination reaction. The concentration of the reactants was found to be another critical factor for aniline
formation. The amination of the aromatic ring of substituted benzenes was also investigated under the
optimized conditions.
Received 7th April 2014,
Accepted 4th May 2014
DOI: 10.1039/c4cy00432a
www.rsc.org/catalysis
to aniline generally suffers from a relatively low conversion of
benzene and/or relatively low selectivity to aniline, so it has not
1. Introduction
Aniline is an important precursor or intermediate for the
production of agrochemicals, dyes and pharmaceuticals.1–3
Current commercial methods for the preparation of aniline
involve multiple reaction steps and have many disadvantages,
such as the use of corrosive materials, complicated appara-
tus, the production of waste, and related environmental
problems.4–9 A very attractive and challenging method for the
preparation of aniline directly from benzene was first reported
by Wibaut in 1917, whereby aniline could be obtained by
passing benzene and ammonia over a reduced iron/nickel
catalyst at 550–600 °C.10 Later, many researchers tried to shift
the reaction equilibrium towards aniline production by
removing the hydrogen formed.11–17 However, these methods
still suffer from the disadvantages of high energy consumption,
high cost of equipment and harsh reaction conditions. In
2013, Yoshida et al. synthesized aniline (0.066% yield) over a
platinum loaded titanium oxide photocatalyst with NH3·H2O
under optimized conditions with emphasis on the reaction
mechanism.18 Our previous studies indicated that H2O2 is a
desired oxidant for the direct synthesis of aniline from benzene
with ammonia, although the yield of aniline was still
not satisfactory.19–21 The direct, one-step amination of benzene
been adopted in industry.11,12 However, many researchers are
still deeply attracted by this fantastic research field.
Cu-based catalysts are well known as being efficient catalysts
for various reactions, such as, the hydrogenation process,22–25
and the activation of benzene halide and ammonia.3,26,27
Moreover, copper oxide can also be used as a reducible metal
oxide in the direct amination of benzene to aniline.16 Our pre-
vious work20 showed that titanium-silicalite-1 (TS-1) catalyst
displayed acceptable activity to the direct amination of ben-
zene to aniline. The loading of copper species on TS-1 remark-
ably enhanced the activity of the catalysts and increased the
yield of aniline to 1% under mild conditions. Recently, we
found that the use of a reactive distillation (RD) reactor could
further promote the yield of aniline to 12.4% over a hierar-
chical TS-1 (h-TS-1) supported copper catalyst.19
Although TS-1 and h-TS-1 are excellent supports for the
direct amination of benzene to aniline, there are some inher-
ent problems in their preparation, such as difficult duplication,
high cost and the requirement for complex technology, which
severely restrict their practical application. Finding a new
support as an alternative is a good way of overcoming the
above shortcomings. As a widely used support for many cata-
lysts, silicalite-1 (S-1) with its mesoporous structure possesses a
similar crystal and pore structure as well as a similar surface
area (SBET) to TS-1 and h-TS-1. However, no study regarding
the use of Cu/S-1 catalyst for the direct amination of benzene
to aniline has been reported even though the preparation of
S-1 is relatively easier. The aim of this study was develop a new
catalyst which could be produced simply, and to gain further
insight into the amination of benzene to aniline.
a Key Laboratory of Green Chemistry and Technology, Ministry of Education,
College of Chemistry, Sichuan University, 29 Wangjiang Road, Chengdu, Sichuan,
610064, China. E-mail: changweihu@scu.edu.cn, gchem@scu.edu.cn;
Fax: +86 28 85411105; Tel: +86 28 85411105
b School of Life Sciences, Guizhou Normal University, 116 Baoshan North Road,
Guiyang 550001, China
† Electronic supplementary information (ESI) available. See DOI: 10.1039/
c4cy00432a
This journal is © The Royal Society of Chemistry 2014
Catal. Sci. Technol., 2014, 4, 3159–3167 | 3159

View Article Online
Paper
Catalysis Science & Technology
X-ray photoelectron spectroscopy (XPS) was carried out using
an AXIS Ultra DLD (KRATOS) spectrometer with Al-Kα radiation.
The obtained binding energy (BE) was calibrated using the
C 1s peak at 284.8 eV as the reference, and the Shirley back-
ground was subtracted from all the spectra. Then peak fitting
was performed using an 80/20 Lorentz–Gauss function.
The samples were also characterized by NH3 temperature
programmed desorption (NH3-TPD). A multichannel HP R-20QIC
mass spectrometer was also used to analyze the released gases
(NH3-TPD-MS). 0.20 g of sample in a stainless steel U-tube was
flushed with N2 at 300 °C for 1 h. After cooling to 60 °C, the
gas was switched to pure NH3 with a flow rate of 30 ml min−1
for 25 minutes. Weakly adsorbed NH3 was removed by pure He
sweeping at 60 °C until the baseline of Thermal Conductivity
Detector and MS signals stabilized. Under an atmosphere of
purified He with a flow rate of 30 ml min−1, NH3-TPD was
conducted from 60 to 800 °C at a ramping rate of 10 °C min−1.
The possible molecules (N2O, NO, NO2, and NH3) from
desorption or reactive desorption of NH3 were monitored
with MS signals of m/z = 44, 30, 46 and 17.
2. Experimental
2.1. Catalyst preparation
S-1 was synthesized basically according to the procedures
reported in the literature.28 The details are given in the ESI† (S1).
For the synthesis of mesoporous silica (Meso-1),29 2.4 g of
cetyltrimethylammonium bromide and 120.0 g of H2O were
mixed and stirred until the solution became clear. 8.0 ml of
28% ammonia solution was added, followed by the addition
of 9.4 g of tetraethylorthosilicate. After stirring for 24 h,
the precipitate was filtered, dried at 100 °C overnight, and
calcined at 550 °C for 6 h in air. As a reference support in the
previous work,20 commercial porous silica from Qingdao
Xinchanglai Silicone Limited Company was also used and
denoted as Meso-2 in the present work.
Cu/SiO2 catalyst was prepared by an impregnation method.
First, an aqueous solution of Cu(NO3)2 was dropped onto the
SiO2 support in a 50 ml beaker, and the slurry was left for 24 h
at room temperature. The excess H2O was removed by heating
in a water bath at 80 °C, and the resultant solid was dried
overnight under vacuum at 110 °C, and then calcined at 550 °C
for 6 h in air. Finally, the catalysts were obtained and denoted
as x wt% support (x the controlled Cu content).
The chemical analysis of the samples was performed
by using the inductively coupled plasma (ICP) technique to
measure the Cu loading (Thermo E.IRIS).
2.3. Measurements of catalytic activity
2.2. Catalyst characterization
Direct amination of benzene was performed in a 50 ml single
necked flask equipped with a reflux condenser and a stirrer
in a sealed system. Typically, 0.5 g of 2.5 wt% Cu/S-1 and
2.8 mmol benzene were loaded into the flask. The reaction
was carried out with stirring when the temperature of the
reactor reached 60 °C. A mixture consisting of 40.0 ml of
H2O, 133.0 mmol NH3·H2O and 18.1 mmol H2O2 was pre-
pared and kept at 0 °C. The mixture was fed into the reaction
system at a velocity of 0.2 ml min−1 from the top of the con-
denser by means of a peristaltic pump. After the addition of
the mixture was completed, the reaction was carried out for
another 0.5 h under the same reaction conditions. Then, the
reaction system was cooled down to room temperature.
X-ray diffraction (XRD) measurements of the samples were
performed with an LTD DX-1000 CSC diffractometer using
Cu-Kα radiation (λ = 0.15406 Å) with a scanning angle (2θ)
from 5 to 75°. The tube voltage was 40 kV and the current
was 30 mA.
N2 adsorption–desorption isotherms were measured by
static N2 physisorption at −196 °C with a Micromeritics
Tristar 3020 SBET and pore analyser. Prior to N2 adsorption–
desorption, the samples were degassed at 100 °C for 1 h and
then evacuated at 300 °C for 4 h to remove the physically
adsorbed impurities. SBET was calculated using the Brunauer–
Emmett–Teller (BET) method. The total pore volume (Vp) was
derived from the adsorbed N2 volume at a relative pressure of
approximately 0.99, and the Barrett–Joyner–Halenda method
was used to calculate the pore size distributions according to
the desorption branches of the isotherms.
Ultraviolet-visible diffuse reflectance (DR UV-vis) spectra were
obtained in the range from 200 to 800 nm using a TU-1901
spectrometer equipped with a reflectance attachment. BaSO4
was used as the reference material.
Fourier transform infrared (FT-IR) spectra were collected at
2 cm−1 resolution using a Nicolet Nexus 670 FT-IR spectrometer
equipped with a mercury cadmium telluride detector. The
powder samples were mixed with KBr (1 wt%) and pressed
into translucent disks at room temperature.
The analysis of the products and related calculation methods
were the same as those detailed in our previous report (S2†).19
3. Results and discussion
3.1. Catalyst characterisation
3.1.1 XRD. Fig. 1A shows the XRD patterns of the three
bare supports and fresh samples containing 2.5 wt% copper.
For Meso-1 and Meso-2, no obvious diffraction peaks could
be observed, indicating that Meso-1 and Meso-2 were present
in an amorphous state.30 The diffraction peaks of S-1 were
coincident with the mobil five (MFI) zeolitic topology with a
monoclinic symmetry (the signals at 2θ were 7–9°, 23–25°
and 45°).31 In addition, no diffraction peaks from cupric
oxide or any related cupreous substances could be detected
in the three samples containing 2.5 wt% copper, suggesting
that the copper species were highly dispersed on the three
supports.
Transmission electron microscopy (TEM) images were
obtained with an FEI Company Tecnai G220 Stwin instrument
operated at 200 kV. The samples were ultrasonically dispersed
in ethanol at room temperature for 6 h. The as-obtained solu-
tions were dropped onto copper grids for TEM.
3160 | Catal. Sci. Technol., 2014, 4, 3159–3167
This journal is © The Royal Society of Chemistry 2014

View Article Online
Catalysis Science & Technology
Paper
showed that the structure of the samples remained the same
after use.
3.1.2 N2 adsorption–desorption isotherm. N2 adsorption–
desorption isotherms and pore size distributions of the samples
are shown in Fig. S2–S4.† According to the International Union
of Pure and Applied Chemistry classification,33 all the isotherms
exhibited characteristic type IV curves with capillary condensation
that signified the presence of mesopores. SBET and pore structure
parameters of all the samples are summarized in Table 1. The
average pore diameters of all the samples resided in the range
of 25.9–81.0 Å, confirming that all the samples possessed a
mesoporous structure. Among the three bare supports, SBET
followed the order: Meso-1 > Meso-2 > S-1.
It was also found that the loading of copper on the bare
supports obviously decreased the SBET and Vp. For the sam-
ples containing 2.5 wt% copper, the decrease of SBET and
Vp were 44% and 21% for Cu/Meso-1, 23% and 14% for
Cu/Meso-2, and only 8% and 5% for Cu/S-1, respectively. It
was likely that the porous space inside the support was occu-
pied and the smaller pore mouth was blocked by the copper
oxide introduced. As shown in Fig. S2,† the pore size distribu-
tions became broader after Cu loading.34 For the fresh Cu/S-1,
the SBET and Vp gradually decreased with the increase of
Cu loading from 0.5 to 14.0 wt%. It seemed that more porous
space inside the support was occupied and the larger pore
mouth was blocked with the increasing amount of copper
oxide introduced.
For the used samples, the loss of SBET was more drastic
for Cu/Meso-1 and Cu/Meso-2 than that for Cu/S-1; especially,
on the used 2.5 wt% Cu/Meso-1, SBET decreased by 68%. The
decrease in SBET may be attributed to the partial collapse of
the mesoporous network, caused by the vigorous stirring con-
ditions and the liquid-phase nature of the reaction.35 Because
the structure of S-1 possessed strong stability, the SBET loss of
Cu/S-1 was low after use.
Fig. 1 XRD patterns of the samples: (A) the bare supports and fresh
2.5 wt% Cu loading samples. (B) The fresh catalysts with different Cu
loadingssthe on S-1 samples.
XRD patterns of the samples with different Cu loading on
S-1 are shown in Fig. 1B. When the Cu loading was greater
than 5.5 wt%, two diffraction peaks at 2θ of 36.5° and 38.8°
characteristic of CuO were observed, and their intensity
increased with an increase of the Cu loading, implying that
copper oxide aggregated into large particles.32
3.1.3 DR UV-vis and FT-IR spectroscopy. In the literature,36,37
the 600–800, 440 and 230 nm DR UV-vis bands were assigned
to the d–d transition of Cu(II) ions in the octahedral ligand
field generated by the oxygen ions, the charge-transfer (CT) of
For the used samples containing 2.5 wt% copper
(Fig. S1†), the peak positions were not changed, which
Table 1 The textural properties and chemical compositions of the catalysts
SBET (m2 g−1
)
VP (cm3 g−1
)
ΦP (Å)
Cuc (wt%)
Cud (wt%)
b
Sample
S-1
Meso-1
420
889
478
400
387
375
368
339
341
500
281
390
324
0.43
0.58
0.96
0.42
0.41
0.41
0.38
0.35
0.47
0.46
0.45
0.83
0.91
41.3
25.9
81.0
41.8
57.0
44.5
41.6
41.6
55.7
36.8
63.8
84.9
88.2
—
—
—
—
—
—
Meso-2
0.5Cu/S-1a
0.49
2.48
5.47
9.95
13.92
2.28
2.47
1.73
2.48
1.87
0.12
0.33
0.52
0.80
0.88
0.30
1.98
1.39
5.77
0.60
2.5Cu/S-1
5.5Cu/S-1
10.0Cu/S-1
14.0Cu/S-1
2.5Cu/S-1(used)
2.5Cu/Meso-1
2.5Cu/Meso-1(used)
2.5Cu/Meso-2
2.5Cu/Meso-2(used)
a
b
0.5 wt% Cu/S-1 was abbreviated to 0.5Cu/S-1 and other samples were also abbreviated analogously. Adsorption average pore width.
Total Cu loading determined by ICP. d Surface Cu loading determined by XPS.
c
This journal is © The Royal Society of Chemistry 2014
Catal. Sci. Technol., 2014, 4, 3159–3167 | 3161

View Article Online
Paper
Catalysis Science & Technology
O
bridge → Cu transition in the Cu2O2 moiety and CT of O → Cu
from 1.0/0.8/3.0 to 1.0/3.0/6.0 with an increase of the Cu load-
ing to 14.0 wt%. The results revealed that the doped copper
atoms were incorporated mainly into the mesoporous silica
oxide network at a low Cu loading, and generated dispersed
CuO species after the Cu loading reached 2.5 wt%, and then
formed Cu–O–Cu species with a higher Cu loading. The
amount of Cu–O–Cu species increased obviously with an
increase of the Cu loading and that of Cu–O–Si species
decreased dramatically with an increase in the amount of
Cu–O–Cu species.
Fig. 3 shows the FT-IR spectra of the samples. For the
three bare supports (Fig. 3A), both the bands at 3430–3450
and 1630–1640 cm−1 are attributed to the stretching vibration
and the bending vibration of O–H bond in physisorbed H2O,
respectively, indicating the hydrophilic character of SiO2.38
The band at 470 cm−1 is assigned to the Si–O–Si bending
vibration, while the bands at 800–816 and 1090–1115 cm−1
are caused by Si–O–Si symmetric stretching vibration and the
Si–O–Si asymmetry stretching vibration, respectively.38,39 For
the S-1, a band near 550 cm−1 is observed which is due to the
stretching vibration of structural double five-member rings.40
Thus the appearance of this band is evidence that the S-1
support possesses an atomic ordering of the MFI structure.
These results compare well with the XRD measurements.
transitions of isolated Cu(II) ions in coordination with lattice
oxygen, respectively. Fig. 2A shows the spectra of the three bare
supports and fresh samples containing 2.5 wt% copper. For
the three bare supports, no peaks were observed over the exam-
ined wavelength range of 200–800 nm. After copper species
was introduced (Fig. 2A and S5†), the spectra were character-
ized by a d–d band in the region of 600–800 nm and an intense
CT band at 230 nm. For the 2.5 wt% Cu/Meso-2, an obvious
peak at about 440 nm was observed, and S230/S440/S600–800
(the ratio of the peak area at 230 nm to the peak area at 440 nm
to the peak area at 600–800 nm) was 1.0/0.9/0.9. However, no
band at about 440 nm was detected over 2.5 wt% Cu/Meso-1
and Cu/S-1, and S230/S600–800 was 1.0/0.4 and 1.0/1.2, respectively.
These results indicated that the Cu–O–Si species was formed in
all the samples containing copper. The Cu–O–Si species was
generated over the Meso-1 support. However, the formation of
Cu–O–Cu species was easier on Cu/Meso-2.
For the fresh Cu/S-1 (Fig. 2B and S5†), as the Cu loading
increased from 0.5 to 4.0 wt%, S230/S600–800 initially increased,
reached a maximum of 1.0/1.2 at 2.5 wt%, and then decreased
to 1.0/3.0. When the Cu loading was greater than 5.5 wt%, the
band at about 440 nm appeared, and S230/S440/S600–800 varied
Fig. 2 DR UV-vis spectra of the samples: (A) a, 2.5 wt% Cu/Meso-2;
b, 2.5 wt% Cu/Meso-1; c, 2.5 wt% Cu/S-1; d, Meso-2; e, Meso-1; f, S-1.
(B) a, 14.0 wt% Cu/S-1; b, 10.0 wt% Cu/S-1; c, 7.0 wt% Cu/S-1; d, 5.5 wt%
Cu/S-1; e, 4.0 wt% Cu/S-1; f, 2.5 wt% Cu/S-1; g, 1.0 wt% Cu/S-1;
h, 0.5 wt% Cu/S-1.
Fig. 3 FT-IR spectra of the samples: (A) a, Meso-1; b, 2.5 wt% Cu/Meso-1;
c, Meso-2; d, 2.5 wt% Cu/Meso-2; e, S-1; f, 2.5 wt% Cu/S-1. (B) a, 14.0 wt%
S-1; b, 10.0 wt% Cu/S-1; c, 7.0 wt% Cu/S-1; d, 5.5 wt% Cu/S-1; e, 4.0 wt%
Cu/S-1; f, 2.5 wt% Cu/S-1; g, 1.0 wt% Cu/S-1; h, 0.5 wt% Cu/S-1.
3162 | Catal. Sci. Technol., 2014, 4, 3159–3167
This journal is © The Royal Society of Chemistry 2014

View Article Online
Catalysis Science & Technology
Paper
Compared to the three bare supports, these bands were still
preserved and were actually unchanged after copper was intro-
duced. For the Cu/S-1 (Fig. 3B), a band centred at 960 cm−1
,
ascribed to the formation of Cu–O–Si hetero-linkage, testifies the
successful incorporation of copper into the silica framework,41
which is in accordance with the DR UV-vis measurements.
However, no absorbance at 960 cm−1 was clearly detected
from the spectra of Cu/Meso-1 and Cu/Meso-2, which is
probably due to the overlap of the band from the Si–O–Si
asymmetry stretching vibration.38
3.1.4 TEM. The TEM images presented in Fig. S6† show
that 2.5 wt% Cu/S-1 possesses well crystalline monoclinic
particles with an average diameter of approximately 100–150 nm
(Fig. S6a and b†), indicating that the prepared S-1 has an MFI
crystal structure.42 However, 2.5 wt% Cu/Meso-1 and Cu/Meso-2
have an irregular mesostructure with sponge like mesoporous
morphologies (Fig. S6d, e, g and h†),43 and the pore size of
Meso-2 is larger than that of Meso-1. There was no detectable
agglomeration of copper species condensed on the supports,
suggesting that the metallic particles were dispersed uniformly.
The high dispersion could be associated with the relatively
larger SBET of the supports that provided a favourable dispersing
effect for the copper species.
For the used samples (Fig. S6c, f and i†), the crystalline
characteristics of Cu/S-1 were unchanged, but the structures
of Cu/Meso-1 and Cu/Meso-2 were severely destroyed in the
reaction. The results conformed to the BET measurements
and showed that Cu/S-1 was the most stable in surface and
pore structure among the three samples.
Fig. 4 XPS profiles of the samples: (A) the fresh 2.5 wt% Cu loading
samples. (B) The fresh catalysts with different Cu loadings on S-1
samples.
3.1.5 XPS. The XPS profiles (Cu 2p) of the samples are
reported in Fig. 4. The intense and broad photoelectron peak
at about 934.0 eV (Cu 2p3/2) observed for the three Cu-based
samples were in agreement with literature44 concerning copper
oxide systems supported on different oxidic phases.
therefore justifying major Cu species dispersed on the sur-
face of Meso-2. The cS,Cu/cT,Cu of 2.5 wt% Cu/S-1 was the
smallest among the three samples, indicating that Cu species
were thoroughly dispersed in the mesoporous walls.35 From
the analysis of the mass concentration, both stability and
SBET of the support greatly affected the distribution of the Cu
species. It could be presumed that the uniformity and stabil-
ity of S-1 and the high SBET of Meso-1 could provide a rela-
tively large distribution area for copper species. As a result,
this led to a larger degree of bonding between copper species
and hydroxyl-terminated silicone, promising the generation
of more Cu–O–Si species.
The XPS profiles (Cu 2p) of Cu/S-1 with different Cu
loadings are shown in Fig. 4B. Two Cu 2p3/2 peaks (about
933.0 and 934.5 eV), attributed to dispersed CuO species
and Cu–O–Si or Cu–O–Cu species, were observed.47–49 The
cS,Cu/cT,Cu decreased from 0.25 to 0.06 with an increase of the Cu
loading from 0.5 to 14.0 wt%. This meant that a large amount
of Cu species dispersed more easily within the pores than on
the surface in S-1. The appearance and increase of Cu–O–Cu
species caused a dramatic decrease of the Cu–O–Si species.
For the used samples containing 2.5 wt% copper, the losses
of surface Cu content for S-1, Meso-1 and Meso-2 were 9%,
30% and 90%, respectively, indicating that Cu species on S-1
were the most stable among the three catalysts.
Fig. 4 shows that copper existed in the oxidation state of
Cu2+ on all the samples, evidenced by the Cu 2p3/2 peak at
932.8–935.6 eV and the 2p → 3d satellite at 942.0–944.0 eV
characteristic of Cu2+ with an electron configuration of d9.45
As shown in Fig. 4A, the XPS profiles of the three samples
containing 2.5 wt% copper presented two contributions to
the Cu 2p3/2 peaks, that is, 933.0 and 934.5 eV for Cu/S-1;
933.3 and 935.6 eV for Cu/Meso-1; 933.1 and 934.5 eV for
Cu/Meso-2, respectively, indicating that two kinds of Cu(II)
species with different microenvironments existed on these
samples.46 Typically, dispersed CuO species were observed at
about 933.0 eV, and the shift to a higher BE was indicative of a
CT from the metal ion towards the support oxide, suggesting
the formation of the Cu–O–Si species.45
As shown in Table 1, the mass concentration ratios of the
surface Cu content (cS,Cu) obtained from XPS to the total
Cu content (cT,Cu) obtained from ICP, were calculated to be
0.13 for Cu/S-1, 0.80 for Cu/Meso-1 and 2.32 for Cu/Meso-2,
when the samples contained 2.5 wt% copper. The values
lower than 1 meant that cS,Cu < cT,Cu, suggesting that major
Cu species dispersed within the pores in S-1 and Meso-1. On
the contrary, values larger than 1 indicated cS,Cu > cT,Cu
,
This journal is © The Royal Society of Chemistry 2014
Catal. Sci. Technol., 2014, 4, 3159–3167 | 3163

View Article Online
Paper
Catalysis Science & Technology
3.1.6 NH3-TPD-MS. The NH3-TPD-MS measurements were
Interestingly, the desorption temperature of NH3 was
almost the same as that of NO gas. However, three N2O peaks
were detected at 450, 580 and 650 °C. The profile of Cu/Meso-2
possessed the largest peak area at 450 °C and the smallest one
at 650 °C. On the Cu/Meso-1, the area of the three peaks was
almost the same. In contrast, the profile of Cu/S-1 exhibited
the smallest peak area at 450 °C and the largest one at 650 °C.
These results indicated that the desorptive reaction of NH3
from the catalysts originated from the different degree of
activation of the N–H bond of NH3 and/or different interaction
of NOx formed with the catalysts.
carried out to evaluate the interaction between ammonia and
the catalysts containing 2.5 wt% copper, and the obtained
profiles are shown in Fig. 5. On the three catalysts, the main
evolved gases were NH3, NO and N2O with trace NO2. This
could be explained by the fact that part of the chemisorbed
NH3 was oxidized to NO, N2O and NO2 during the desorption
process by the catalysts. Three NH3 desorption peaks at 150,
270 and 400–800 °C were assigned to NH3 desorption on weak,
medium-strong and strong acid sites, respectively, indicating
that there were at least three kinds of interaction between the
catalysts and NH3.
The total amount of desorbed NH3 on 2.5 wt% Cu/Meso-2,
2.5 wt% Cu/Meso-1 and 2.5 wt% Cu/S-1 was 1.1, 0.9 and
1.8 mmol, respectively. The ratio of the three peaks corresponded
to the reactive amounts of the acid sites. The Sa/Sb/Sc (peak
area ratio) of NH3 which desorbed from the three catalysts
at 150, 270 and 400–800 °C were 1.0/0.6/1.2, 1.0/1.2/3.5 and
1.0/0.5/0.1, respectively. The proportion of weak and medium-
strong acid sites on Cu/S-1 was the largest, while the pro-
portion of strong acid sites was the lowest. In contrast,
the proportion of medium-strong acid sites on Cu/Meso-2
was the lowest, while the proportion of strong acid sites was
the largest.
3.2. Catalytic amination of benzene
The results for the amination of benzene to aniline over dif-
ferent catalysts are shown in Table 2. No aniline was detected
over the three bare supports. Aniline could be generated over
copper oxide and copper doped porous silica. When 2.5 wt%
Cu/S-1 was used as the catalyst, the yield of aniline (4.4%)
and the conversion of benzene (6.2%) was obtained with a
selectivity of 81.5% (entry 5) at 60 °C. The main by-products
detected were phenol (0.8%) and dihydroxybenzenes (DHB)
(0.2%) with trace nitrobenzene. The activities of 2.5 wt%
Cu/Meso-1 and Cu/Meso-2 were poor, and the yield of aniline
was significantly low (1.7%, 0.5%) with low selectivity (48.6%,
50.0%). The 2.5 wt% Cu/Meso-1 was more effective for the
production of phenol. The lowest yield of aniline (0.1%) was
obtained over CuO.
Fig. 6 illustrates the influence of Cu loading on the reac-
tion. With the increase of Cu loading, the yield of aniline,
conversion of benzene and selectivity to aniline went up ini-
tially, reaching maxima of 4.4%, 6.2% and 81.5% respectively
at 2.5 wt% Cu/S-1 and then decreased. However, the yield of
phenol initially decreased and then increased. The yield of
DHB steadily increased from 0.1 to 0.9% with an increase of
Cu loading from 0.5 to 14.0 wt%. It was clear that excessive
or inadequate copper species were more advantageous to
hydroxylation of benzene. These results are in agreement
with our previous reports.19,20 DR UV-vis (Fig. 2 and S5†) and
XPS (Fig. 4) indicated the existence of Cu–O–Si species,
Fig. 5 NH3-TPD-MS analysis of the samples: a 2.5 wt% Cu/Meso-2;
b 2.5 wt% Cu/Meso-1; c 2.5 wt% Cu/S-1.
Table 2 The activity over different catalystsa
Yield (mol%)
Conversion of
benzene (%)
Selectivity to
aniline (%)
Entry
Catalyst
Aniline
Phenol
DHB
1
2
3
4
5
6
7
8b
S-1
Meso-1
Meso-2
CuO
2.5 wt% Cu/S-1
2.5 wt% Cu/Meso-1
2.5 wt% Cu/Meso-2
2.5 wt% Cu/S-1
0.0
0.0
0.0
0.1
4.4
1.7
0.5
5.4
1.4
1.2
1.2
0.3
0.8
1.5
0.3
1.4
0.1
2.1
0.2
0.0
0.2
0.3
0.2
0.5
2.3(1.4)c
4.1(3.3)
2.2(1.4)
1.2(0.4)
6.2(5.4)
4.3(3.5)
1.9(1.0)
8.1(7.3)
0.0
0.0
0.0
25.0
81.5
48.6
50.0
74.0
a
Reaction condition: 1.0 g catalyst, 40.0 ml H2O, 2.8 mmol C6H6, 133.0 mmol NH3·H2O and 22.6 mmol H2O2, reaction temperature 60 °C,
b
reaction time 8 h. Optimal reaction condition: 0.5 g catalyst, 40.0 ml H2O, 2.8 mmol C6H6, 133.0 mmol NH3·H2O and 18.1 mmol H2O2,
reaction temperature 60 °C, reaction time 4 h. Value calculated by moles of all products/moles of initial benzene × 100, the difference
c
between the two value indicated some losses of benzene in the process.
3164 | Catal. Sci. Technol., 2014, 4, 3159–3167
This journal is © The Royal Society of Chemistry 2014

View Article Online
Catalysis Science & Technology
Paper
better than 2.5 wt% Cu/Meso-1. The result indicated that the
presence of Cu–O–Si species on catalysts is not the most
important factor affecting the catalytic performance.
Previous work reported that medium-strong acid sites on
catalysts could be important for amination.20 Compared to
the other two catalysts, 2.5 wt% Cu/S-1 possessed greater
amounts of medium-strong acid sites which provided enough
sites for the effective chemical adsorption of NH3, and more
Cu–O–Si species which provided enough active sites for the
activation of the N–H bond of NH3. When 2.5 wt% Cu/Meso-2
was employed as the catalyst, it could not provide enough
sites for the effective chemical adsorption of NH3 nor active
sites for the activation of N–H bond of NH3. Though 2.5 wt%
Cu/Meso-1 could provide more active sites for NH3 activation,
there were not enough sites for the effective chemical adsorp-
tion of NH3. The outstanding catalytic amination performance
was obtained due to the existence of greater amounts of both
medium-strong acid sites and Cu–O–Si species on 2.5 wt%
Cu/S-1. The results indicated that Cu–O–Si species in Cu/SiO2
played almost the same role as Cu–O–Ti species over Cu/TS-1
in amination. On the other hand, the adsorption of NH3 on
medium-strong acid sites also played an important role in the
total reaction.
In the present work, the formation of aniline, as well as
phenol and DHB over all the catalysts, showed that the C–H
bond of benzene could be activated in the presence of H2O2
under mild conditions. Although the conversion of benzene
(1.2%) (Table 2, entry 4) could be obtained over CuO, the
catalytic activity of CuO was inferior to that of the copper
doped catalyst. S-1 turned out to be the best support for
the activation of the C–H bond of benzene (Table 2, entry 5).
Remarkably, with an increase of the Cu loading, the conver-
sion of benzene went up initially, reached a maximum of
6.2% on 2.5 wt% Cu/S-1 (Fig. 6) and then decreased. The
results were in accordance with our previous work,19,20 that
is, the C–H bond of benzene could be effectively activated
over Cu/S-1 under the present reaction conditions. The for-
mation of aniline was dependent on both the selective activa-
tion of the N–H bond of NH3 and the C–H bond of benzene
on Cu/SiO2.
The effect of other reaction conditions on the amination
of benzene to aniline have been investigated (S3†). Under the
optimized conditions, the highest yield of aniline (5.4%) and
the highest selectivity of aniline (74.0%) was obtained over
Cu/S-1 (Table 2, entry 8). The results showed that there was a
competition between amination and hydroxylation of ben-
zene in the reaction system.19,50 It was revealed that a diluted
instantaneous concentration of both ammonia and H2O2 was
favourable to amination.
To further investigate the catalytic activity of the Cu/S-1 in
the direct amination of benzene, 2.5 wt% Cu/S-1 was used
under the optimized conditions in kettle-type20 and RD reac-
tors19 (S4†). The results showed that 2.5 wt% Cu/S-1 possessed
comparable catalytic activity to that of 2.5 wt% Cu/h-TS-1,
confirming that Cu/S-1 could play almost the same role as
Cu/h-TS-1 in the direct amination of benzene. However, the
Fig. 6 The effect of Cu loading on the yield of products, the conversion
of benzene and the selectivity to aniline. Reaction conditions: 1.0 g
of catalyst, 40.0 ml of H2O, 2.8 mmol C6H6, 133.0 mmol NH3·H2O and
22.6 mmol H2O2, reaction temperature 60 °C, reaction time 8 h.
dispersed copper species and Cu–O–Cu species on the
catalysts with different Cu loading on mesoporous silica.
Over the S-1 support (Fig. 2B and S5†), the relative amount
of Cu–O–Si species initially went up, reached a maximum on
2.5 wt% Cu/S-1, and then decreased gradually with an
increase of Cu loading. When the Cu loading was greater than
5.5 wt% (Fig. S5G–J†), the Cu–O–Cu species was generated
and its relative amount increased obviously with the increase
of Cu loading, accompanied by a decrease of Cu–O–Si species.
Remarkably, with an increase in the Cu loading, the catalytic
activity increased obviously and then decreased gradually.
The best catalytic amination performance could be obtained
on 2.5 wt% Cu/S-1.
Our previous work20 indicated that Cu–O–Ti could form by
doping a certain amount of Cu on the TS-1 zeolite. Cu–O–Ti
species could weaken the acidity of the initial strong acid sites
and promote the chemical adsorption of NH3 on Cu-TS-1,
thus enabling the selective activation of the N–H bond of NH3
and the oxyamination of benzene. Similar to our previous
NH3-TPD experiments over Cu-TS-1, four kinds of gaseous
molecules (NO2, N2O, NO and NH3) desorbed from the cata-
lyst were also observed (Fig. 5). The results demonstrated that
NH3 could be adsorbed and that the N–H bond of NH3 could
be selectively activated on the copper doped catalysts. There-
fore, it is suggested that Cu–O–Si plays a similar role to
Cu–O–Ti in direct amination. That is, Cu–O–Si species could
also promote the chemical adsorption of NH3 on Cu/SiO2,
therefore enabling the selective activation of the N–H bond of
NH3 and the oxyamination of benzene.
On the catalysts containing 2.5 wt% copper, no large particles
of CuO were detected by XRD (Fig. 1A) and TEM (Fig. S6†).
XPS (Fig. 4A) and ICP showed a high dispersion of Cu species
on S-1, which enabled the bonding of Cu to the hydroxyl-
terminated silicone, forming Cu–O–Si species on the frame-
work of the catalysts. With the highest SBET, Meso-1 could
improve the dispersion of Cu species, resulting in the largest
relative amount of Cu–O–Si species on 2.5 wt% Cu/Meso-1, as
indicated by DR UV-vis. However, 2.5 wt% Cu/S-1 performed
This journal is © The Royal Society of Chemistry 2014
Catal. Sci. Technol., 2014, 4, 3159–3167 | 3165

View Article Online
Paper
Catalysis Science & Technology
catalytic activity of Cu–O–Ti would be better than that of
Cu–O–Si.
In addition, the possible reaction pathway of amination
for aniline formation was investigated (S5†). The experimen-
tal results showed that hydroxylamine was the possible inter-
mediate, which agreed well with the previous report.19
Among these substrates, the electron donating property of
the substituents followed the order: –OH > –OCH3 > –CH3 >
–Cl > –NO2. With an increase of the ability of electron
donation, the yield of the amination products increased from
0.2% to 7.4% for –OCH3. However, no amination products
were detected when phenol was used as the substrate. It
might be that the amination products were unable to desorb
from the catalyst surface and further inhibited their forma-
tion.18 Because the substituents of toluene, paraxylene and
mesitylene were involved in the reaction, the selectivity of
their amination products was low (22.2–48.3%). With an
increase of the number of substituents, the yield of products
on the aromatic ring obviously decreased due to the steric
hindrance of the substituents. It is worth noting that
ortho-isomers were preferentially produced from substituted
benzenes (Table S2†). These results confirmed that the reac-
tion possessed possibly electrophilic characters.
3.3 The stability and recyclability of the catalyst
The stability of Cu/S-1 (Table S1†) was measured as well. The
selectivity to aniline and catalytic activity decreased after the
first two runs. In the next three recycles, the catalytic perfor-
mance was almost unchanged, suggesting the good recycla-
bility of the catalyst.
The used Cu/S-1 was further investigated by XRD (Fig. S1†),
BET (Table 1 and Fig. S3†) and TEM (Fig. S6†). The results
showed that the used Cu/S-1 still possessed a mesoporous
and MFI crystal structure after being used. The ICP results
presented in Table S1† indicate that the Cu loading of the
catalyst decreased slightly from 2.48 to 2.13% after being used
in the first two runs and then remained almost unchanged with
further runs. These results demonstrated the good stability
of the Cu/S-1.
The literature51 and our previous5 work reported that the
active species were the electrophilic ammonia radical cation
+
(·NH3 ) when hydroxylamine was used as the aminating agent.
We found that hydroxylamine could be generated in the pres-
ent reaction system. The results suggested that ·NH3+ was the
possible active species and hydroxylamine was the reaction
intermediate in this reaction. More studies on the reaction
mechanism will be conducted in future work.
3.4. Catalytic amination of substituted benzenes
Under the optimized conditions, the amination was extended 4. Conclusions
to substituted benzenes. The results are listed in Table 3.
Copper doped S-1 was effective in the direct aromatic ring
Amination products could be detected for all of the substituted
benzenes except phenol. When anisole was used as the
substrate, the highest yields of amination (7.4%) and hydroxyl-
ation (4.3%) products on the aromatic ring were obtained
with a selectivity of 63.2% to arylamine. For chlorobenzene, the
high yield (4.8%) and selectivity (96.0%) to arylamine were
obtained. Although the selectivity reached a maximum of
100.0%, the yield of the amination products of nitrobenzene
was the lowest (0.2%). The results showed that the Cu/S-1 cata-
lyst could effectively activate the C–H bond on the aromatic
ring of substituted benzenes and the N–H bond of ammonia
in the presence of H2O2.
amination of benzene and substituted benzenes with NH3·H2O
and H2O2. The MFI structure of S-1 with hydroxyl-terminated
silicone was found to facilitate the formation of Cu–O–Si species.
This Cu–O–Si species was identified to be one of the important
factors responsible for the generation of aniline. Besides, the
medium-strong acid sites were also determined as a vital factor
for the amination process. Because 2.5 wt% Cu/S-1 possessed
more amounts of Cu–O–Si species and medium-strong acid
sites, it exhibited the best catalytic performance in the reaction.
Over 2.5 wt% Cu/S-1, the amination process could provide
5.4% aniline yield in a close system at 60 °C, while a yield of
10.8% was obtained using an RD reactor. The concentration of
the reactants was another crucial factor for the reaction, and a
diluted instantaneous concentration of both ammonia and
H2O2 was found to be favourable for amination.
Table 3 The results of catalytic amination of substituted benzenesa
Products yield on the
Selectivity
aromatic ring (mol%)
to aminated
Acknowledgements
Substrate
Amination Hydroxylation Other products (%)
Financial support from the NSFC (no. 21172157, no. 21372167
and no. 201321061) of China, and technical support from the
Analytical and Testing Centre of Sichuan University are cordially
acknowledged. The suggestions by the reviewers were highly
appreciated.
Nitrobenzene
Chlorobenzene 4.8
Benzene
Toluene
Anisole
Phenol
Paraxylene
Mesitylene
0.2
n.d
0.2
1.9
3.1
4.3
3.5
1.5
0.3
—
—
—
1.5
—
n.d
1.6
1.1
100.0
96.0
74.0
48.3
63.2
—
5.4
4.3
7.4
n.d
1.2
0.4
27.9
22.2
Notes and references
a
Reaction condition: 0.5 g of catalyst, 40.0 ml of H2O, 2.8 mmol
substituted benzenes, 133.0 mmol NH3·H2O and 18.1 mmol H2O2,
reaction time 4 h, reaction temperature 60 °C.
1 Q. Shen and J. F. Hartwig, J. Am. Chem. Soc., 2006, 128,
10028.
3166 | Catal. Sci. Technol., 2014, 4, 3159–3167
This journal is © The Royal Society of Chemistry 2014

View Article Online
Catalysis Science & Technology
Paper
2 D. S. Surry and S. L. Buchwald, J. Am. Chem. Soc., 2007, 129,
10354.
29 H. Wang, L. Liang, W. Zhang and J. Han, J. Porous Mater.,
2012, 19, 889.
3 H. Xu and C. Wolf, Chem. Commun., 2009, 3035.
4 H. Yuzawa and H. Yoshida, Chem. Commun., 2010, 46, 8854.
5 L. Zhu, B. Guo, D. Tang, X. Hu, G. Li and C. Hu, J. Catal.,
2007, 245, 446.
6 A. Hagemeyer, R. Borade, P. Desrosiers, S. Guan, D. M. Lowe,
D. M. Poojary, H. Turner, H. Weinberg, X. Zhou, R. Armbrust,
G. Fengler and U. Notheis, Appl. Catal., A, 2002, 227, 43.
7 S. Singha and K. M. Parida, Catal. Sci. Technol., 2011, 1, 1496.
8 K. M. Parida, D. Rath and S. S. Dash, J. Mol. Catal. A: Chem.,
2010, 318, 85.
30 C. Wen, A. Yin, Y. Cui, X. Yang, W.-L. Dai and K. Fan, Appl.
Catal., A, 2013, 458, 82.
31 Q. Hu, B. J. Dou, H. Tian, J. J. Li, P. Li and Z. P. Hao,
Microporous Mesoporous Mater., 2010, 129, 30.
32 R. Wang, X. Guo, X. Wang, J. Hao, G. Li and J. Xiu, Appl.
Catal., A, 2004, 261, 7.
33 K. Sing, D. Everett, R. Haul, L. Moscou, R. Pierotti, J. Rouquerol
and T. Siemieniewska, Pure Appl. Chem., 1985, 57, 603.
34 A. Patel, P. Shukla, T. Rufford, S. Wang, J. Chen, V. Rudolph
and Z. Zhu, Appl. Catal., A, 2011, 409, 55.
9 K. M. Parida, S. S. Dash and S. Singha, Appl. Catal., A, 2008,
351, 59.
35 E. S. Vasiliadou and A. A. Lemonidou, Appl. Catal., A, 2011,
396, 177.
10 J. P. Wibaut, Ber. Dtsch. Chem. Ges., 1917, 50, 541.
11 N. Hoffmann and M. Muhler, Catal. Lett., 2005, 103, 155.
12 N. Hoffmann, E. Loffler, N. A. Breuer and M. Muhler,
ChemSusChem, 2008, 1, 393.
36 K. V. Chary, G. V. Sagar, C. S. Srikanth and V. V. Rao, J. Phys.
Chem. B, 2007, 111, 543.
37 P. J. Smeets, M. H. Groothaert and R. A. Schoonheydt, Catal.
Today, 2005, 110, 303.
13 J. T. Anders, US Pat., 2009/0292144, 2009.
14 P. Desrosiers, S. Guan, A. Hagemeyer, D. M. Lowe, C. Lugmair,
D. M. Poojary, H. Turner, H. Weinberg, X. Zhou, R. Armbrust,
G. Fengler and U. Notheis, Catal. Today, 2003, 81, 319.
15 F. V. Laar, US Pat., 2009/0203941, 2009.
38 Z. Sun, C. Bai, S. Zheng, X. Yang and R. L. Frost, Appl.
Catal., A, 2013, 458, 103.
39 P. M. C. Zapata, M. L. Parentis, E. E. Gonzo and N. A. Bonini,
Appl. Catal., A, 2013, 457, 26.
40 M. V. Konduru and S. S. Chuang, J. Catal., 1999, 187, 436.
41 J. Andas, F. Adam and I. A. Rahman, Appl. Surf. Sci., 2013,
284, 503.
16 J. U. Becker and W. F. Hölderich, Catal. Lett., 1998, 54, 125.
17 S. A. Axon, WO Pat., 99/10 311, 1999.
18 H. Yuzawa, J. Kumagai and H. Yoshida, J. Phys. Chem. C,
2013, 117, 11047.
19 T. Yu, Q. Zhang, S. Xia, G. Li and C. Hu, Catal. Sci. Technol.,
2014, 4, 639.
42 D. V. Vu, M. Miyamoto, N. Nishiyama, S. Ichikawa, Y. Egashira
and K. Ueyama, Microporous Mesoporous Mater., 2008, 115, 106.
43 S. Zhao, H. Yue, Y. Zhao, B. Wang, Y. Geng, J. Lv, S. Wang,
J. Gong and X. Ma, J. Catal., 2013, 297, 142.
20 B. Guo, Q. Zhang, G. Li, J. Yao and C. Hu, Green Chem.,
2012, 14, 1880.
21 C. Hu, L. Zhu and Y. Xia, Ind. Eng. Chem. Res., 2007, 46, 3443.
22 F. Zaccheria, R. Psaro and N. Ravasio, Green Chem., 2009,
11, 462.
23 A. Cooper, B. Bachiller-Baeza, J. A. Anderson, I. Rodriguez-Ramos
and A. Guerrero-Ruiz, Catal. Sci. Technol., 2014, 4, 1446.
24 L. Xie, F. Liu, K. Liu, X. Shi and H. He, Catal. Sci. Technol.,
2014, 4, 1104.
44 L.-F. Chen, P.-J. Guo, M.-H. Qiao, S.-R. Yan, H.-X. Li,
W. Shen, H.-L. Xu and K.-N. Fan, J. Catal., 2008, 257, 172.
45 J. Gong, H. Yue, Y. Zhao, S. Zhao, L. Zhao, J. Lv, S. Wang
and X. Ma, J. Am. Chem. Soc., 2012, 134, 13922.
46 A. Gervasini, M. Manzoli, G. Martra, A. Ponti, N. Ravasio,
L. Sordelli and F. Zaccheria, J. Phys. Chem. B, 2006, 110, 7851.
47 Z. He, H. Lin, P. He and Y. Yuan, J. Catal., 2011, 277, 54.
48 S. Zhu, X. Gao, Y. Zhu, Y. Zhu, H. Zheng and Y. Li, J. Catal.,
2013, 303, 70.
25 J. Lin, X. Zhao, Y. Cui, H. Zhang and D. Liao, Chem. Commun.,
2012, 48, 1177.
26 N. Xia and M. Taillefer, Angew. Chem., Int. Ed., 2009, 48, 337.
27 J. Kim and S. Chang, Chem. Commun., 2008, 3052.
28 E. M. Flanigen, J. Bennett, R. Grose, J. Cohen, R. Patton and
R. Kirchner, Nature, 1978, 271, 512.
49 S. Natesakhawat, J. W. Lekse, J. P. Baltrus, P. R. Ohodnicki,
B. H. Howard, X. Deng and C. Matranga, ACS Catal., 2012, 2,
1667–1676.
50 F. Xiao, J. Catal., 2001, 199, 273.
51 N. I. Kuznetsova, L. I. Kuznetsova, L. G. Detusheva, G. P. Pez
and H. Cheng, J. Mol. Catal. A: Chem., 2000, 161, 1.
This journal is © The Royal Society of Chemistry 2014
Catal. Sci. Technol., 2014, 4, 3159–3167 | 3167