Direct Catalytic Hydroxylation of Benzene to Phenol Catalyzed by FeCl3 Supported on Exfoliated Graphitic Carbon Nitride

Article abstract of DOI:10.1007/s10562-019-03003-2

Abstract: Direct hydroxylation of benzene in the presence of H₂O₂ is a promising approach for the production of phenol. Most of transition-metal compound catalysts have high activity yet suffer from difficulty in catalyst recycling a

Full text of DOI:10.1007/s10562-019-03003-2

Catalysis Letters
https://doi.org/10.1007/s10562-019-03003-2
Direct Catalytic Hydroxylation of Benzene to Phenol Catalyzed by FeCl3
Supported on Exfoliated Graphitic Carbon Nitride
Zhou‑Hong Yu1 · Yu‑Lin Gan1 · Jie Xu1 · Bing Xue1
Received: 16 September 2019 / Accepted: 15 October 2019
© Springer Science+Business Media, LLC, part of Springer Nature 2019
Abstract
Direct hydroxylation of benzene in the presence of H2O2 is a promising approach for the production of phenol. Most of
transition-metal compound catalysts have high activity yet sufer from difculty in catalyst recycling and product purifca-
tion. In this work, graphitic carbon nitride (g-C3N4) material was exfoliated, and then applied as a catalyst support to load
FeCl3. The physical and chemical properties of the synthesized materials were characterized by N2 adsorption–desorption,
XRD, FT-IR, UV–Vis, XPS, and benzene-TPD. The results exhibited that the exfoliation improved the surface area and
pore volume of g-C3N4, and FeCl3 was stably anchored on the surface of eg-C3N4. As a heterogeneous catalyst, FeCl3/eg-
C3N4 showed good catalytic activity and selectivity in hydroxylation of benzene under H2O2 afording a maximum yield of
phenol as 22% at 60 °C.
Graphic Abstract
Keywords Graphitic carbon nitride (g-C3N4) · FeCl3 · Benzene · Hydroxylation · Phenol
1 Introduction
Electronic supplementary material The online version of this
article (https://doi.org/10.1007/s10562-019-03003-2) contains
supplementary material, which is available to authorized users.
As an important commodity chemical in contemporary
chemical industry, phenol is extensively applied for the man-
ufacture of synthesis of resin, dyestufs, pharmaceuticals,
pesticides, etc. [1, 2]. Currently, more than 90% of world-
wide output of phenol is produced based on the three-step
cumene process, which was indeed invented by Hock and
* Jie Xu
shine6832@163.com
* Bing Xue
xuebing_cn@163.com
Extended author information available on the last page of the article
Vol.:(0123456789)
1 3

Z.-H. Yu et al.
Lang dating back to the 1940s [3–5]. However, the tradi-
tional process sufers from several disadvantages including
high energy consumption, strong acidic condition, low one-
pass yield (∼ 5%) of phenol, and concomitant dependence
on the by-product of acetone [6–8]. Alternatively, one-step
direct hydroxylation of benzene, featuring an environmen-
tally benign procedure and relatively mild operation con-
ditions, has been recently regarded as the most promising
strategy for the synthesis of phenol [2, 9, 10].
addition to active sites, the catalytic activity is also basi-
cally related to the surface area of the catalyst support.
Chen et al. reported the Fe–g-C3N4-catalyzed oxidation of
benzene to phenol using H2O2 [31]. Although the reaction
was deliberately designed and driven by photocatalysis,
the fnal conversion of benzene obtained over Fe–g-C3N4
was ca. 4.8% under 60 °C at a reaction time of 4 h. In
fact, the g-C3N4 synthesized by direct thermal condensa-
tion method has very limited surface area and poor poros-
ity [27, 28, 35]. In this work, we synthesized exfoliated
g-C3N4 material and employed it as a catalyst support
to load FeCl3. The characterization results show that the
thermal exfoliation enhanced the surface area of g-C3N4,
and g-C3N4 could steadily anchor FeCl3 species. In the
direct hydroxylation of benzene to phenol in the presence
of H2O2, the synthesized FeCl3/eg-C3N4 demonstrated sta-
ble and good catalytic activity.
Among several oxidants including O2, air, H2O2, and N2O
developed for the direct hydroxylation of benzene, the use
of H2O2 under liquid phase has been widely studied owing
to the milder reaction conditions than its counterparts. It
should be stressed that, the target molecular (i.e. phenol) is
more reactive than benzene in the oxidation process [11, 12].
In this sense, it is still a challenge to develop a catalyst that
could provide efcient activity together with high selectivity,
therefore becomes a hot topic in both catalysis and organic
synthesis. Up to now, most catalysts proposed for the direct
hydroxylation process contain transition metal compounds
as catalytic active components, including Fe3+ [13–15], V5+
[3, 11, 12, 16, 17], Cu2+ [18, 19], vanadyl acetylacetonate
[VO(acac)2] [20, 21], and polyoxometalates [22, 23]. Despite
their high catalytic activity, such metal compounds are dis-
solved in the liquid phase of the hydroxylation reactions.
Consequently, the use of the homogeneous catalysts inevi-
tably are confronted with the difculties in product separa-
tion and catalyst recycling [13, 24]. Concerning the issue,
from the viewpoints of economy and environment, it is very
imperative to design an efcient heterogeneous catalyst for
the hydroxylation of benzene to phenol,
2 Experimental Section
2.1 Catalyst Preparation
2.1.1 Synthesis of g‑C3N4 and eg‑C3N4
5 g of dicyandiamide was placed into a crucible with a
cover, and heated from room temperature to 550 °C with
a ramping rate of 3 °C min−1, and then tempered at this
temperature for another 3 h. The synthesized yellow solid
was g-C3N4. The exfoliated g-C3N4 was prepared through
thermal treatment. 1 g of g-C3N4 was further heated at a
desired temperature for 2 h (with a ramping rate of 3 °C
min−1 from room temperature). The resultant powder was
labeled as eg-C3N4-T, where T indicated the heating tem-
perature. Unless other specifed, T is 500 °C.
Graphitic carbon nitride (g-C3N4) is a novel and appeal-
ing carbon-based polymeric material. Owing to the com-
bination of multiple physicochemical properties, the past
decade has witnessed the signifcant research advance of
g-C3N4 in diverse felds including photocatalysis [25, 26],
thermal catalysis [27–29], fuel cells [30], etc. g-C3N4 is
composed of π-conjugated heptazine tectonic heterocy-
cles bridged by nitrogen atoms. Antonietti et al. found that
g-C3N4 was able to activate benzene and other aromatic mol-
ecules through chemical interaction of g-C3N4 and benzene
via HOMO–LUMO hybridization [27, 31]. Recently, several
groups including ours have revealed that g-C3N4 materials
could catalyze Friedel–Crafts acylation and alkylation of
benzene [27, 28, 32]. However, in the case of direct hydroxy-
lation of benzene, because of the lack of Lewis-acid species
as mentioned above, the reaction is hardly to proceed with
the pure g-C3N4 as a catalyst.
2.1.2 Preparation of FeCl3/eg‑C3N4
0.54 g (2 mmol) of FeCl3·6H2O and 3–4 drops of con-
centrated HCl (a.q.) were initially dissolved in 40 mL
of ethanol. 1.0 g eg-C3N4-T was added into the solution
and stirred for 2 h. After that, the mixture was heated
under 50 °C until removal of ethanol, and the dark yel-
low solid was ground and dried. Afterwards, the powder
was transferred into a tubular furnace (N2 atmosphere,
10 mL min−1), and heated from room temperature to
300 °C for 2 h (with a ramping rate of 2.5 °C min−1 from
room temperature). After the temperature was cooled
down to ca. room temperature, the sample was collected
and designated as xFeCl3/eg-C3N4-T, where x indicated the
molar amount (mmol) of FeCl3·6H2O.
Besides its two dimensional stacking structure, g-C3N4
possesses abundant nitrogen-containing functionalities,
which thus enable it to anchor or embed various transi-
tion metal species, such as Au [33], Zn [25, 34], Fe [31],
and Cu [18], on the g-C3N4 support. On the other hand, in
1 3

Direct Catalytic Hydroxylation of Benzene to Phenol Catalyzed by FeCl3 Supported on…
calculated by means of an area-normalization method, and
the detailed calculation equations were as follows:
2.2 Material Characterization
The specifc surface areas and porous properties of the mate-
rials were analyzed by N2 adsorption–desorption isotherms
at − 196 °C using an ASAP 2020 (Micromeritics) instru-
ment. Before the analysis, the samples were pretreated in
vacuum at 150 °C for 6 h. The surface areas were calculated
by Brunauer–Emmet–Teller (BET) method.
A
phenol × fphenol + Abenzoquinone × fbenzoquinone
benzene + Aphenol × fphenol + Abenzoquinone × fbenzoquinone
phenol × fphenol
phenol × fphenol + Abenzoquinone × fbenzoquinone
Conv. =
A
A
Sel. =
A
X-ray difraction (XRD) patterns were recorded on a D/
max 2500 X-ray difractometer (Rigaku) employing a graph-
ite monochromator (40 kV, 40 mA) equipped with Ni-fl-
tered Cu-Kα radiation.
where A and f were the peak area and response factor,
respectively, for each component analyzed by GC.
Fourier transform infrared (FT-IR) spectra were tested in
a Tensor 27 (Bruker) spectrometer based on the transmission
mode with a resolution of 4 cm−1. Each spectrum was based
on 32 scans (4000–400 cm−1).
3 Results and Discussions
3.1 Materials Characterization
UV–Vis difuse refectance spectra (DRS) were carried
out on a UV-3600 spectrophotometer (Shimadzu) using
BaSO4 as a standard reference.
N2 adsorption–desorption isotherms of the materials are
depicted in Fig. 1A. The pristine g-C3N4 shows a type-IV
isotherm with an H3 hysteretic loop. For the g-C3N4 mate-
rials subjected to the thermal treatment (i.e. eg-C3N4), the
adsorption quantities become much higher, corresponding
to higher pore volumes. The hysteretic loops in the range of
at p/p0 = 0.8–0.99 indicate that both g-C3N4 and eg-C3N4
samples consists of typical mesoporous structures. As
refected in Fig. 1B, both g-C3N4 and eg-C3N4 samples pos-
sess relatively narrow pore size distributions (ca. 3.7 nm).
In the case of the FeCl3/eg-C3N4 materials, the isotherms
also suggest the existence of mesoporous structures like in
the eg-C3N4. However, as the loading amount of FeCl3 is
increased, the adsorption quantities of N2 acquired on FeCl3/
eg-C3N4 become gradually lower. This is probably due to the
collapse and/or blockage of mesopores under high loading
amounts of FeCl3.
X-ray photoelectron spectroscopy (XPS) measurements
were performed using an ESCALAB 250 XI spectrometer
(Thermo Fisher) working in the constant energy mode with
Mg Kα radiation as the excitation source.
Benzene temperature-programmed desorption (TPD) was
carried out on a ChemBET-3000 analyzer (Quantachrome).
First, 100 mg of the sample was pretreated at 200 °C for
1 h in dry He fow (30 mL min−1), cooled down to 50 °C,
and then exposed to He-containing benzene for 0.5 h. After
purging the material with He for 0.5 h, the TPD data based
on TCD signals were collected from 50 to 350 °C (ramping
rate: 10 °C min−1).
Benzene adsorption tests were based on an method
reported previously [36]. A certain amount of benzene (6.7,
9.0 and 11.2 mmol) was dissolved in a solution containing
5 mL of acetonitrile and 1 mL of acetic acid (80 wt%, a.q.).
Next, 50 mg of dried materials was added into the solution,
and the mixture was stirred at 30 °C for 2 h.
The detailed surface areas and pore volumes of the mate-
rials are listed in Table 1. The surface areas as well as pore
volumes of eg-C3N4 are much larger than that obtained in
the pristine g-C3N4. As described above, g-C3N4 owns abun-
dant uncondensed nitrogen-containing functionalities. Such
thermally unstable species are liable to decompose by itself
under high temperatures, thereby creating a certain amount
of mesostructured pores in the g-C3N4. In this sense, the
thermal treatment (i.e. exfoliation) has been widely applied
to enhance the surface area of g-C3N4. It should be noted
that, adopting higher temperature might result in much lower
yield of eg-C3N4 as the C–N skeleton of g-C3N4 would be
destroyed. As for FeCl3/eg-C3N4, the surface areas and pore
volumes decrease with the increase of the loading of FeCl3.
XRD patterns of the materials are given in Fig. 2. The
pristine g-C3N4 shows an intensive difraction peak centered
at 2θ=27.4°, characteristic of the interlayer stacking struc-
ture of g-C3N4, i.e. (002) plane. The other weak difraction
peak emerging at 12.8° is associated with the intralayer
2.3 Catalytic Evaluation
The catalytic hydroxylation of benzene was carried out in a
thee-necked round-bottom fask equipped with a condenser.
1 mL of benzene (11.2 mmol), 5 mL of acetonitrile, and
1 mL of acetic acid (HOAc) were mixed well, followed by
addition of 50 mg of the catalyst. After the mixture was
heated up to 60 °C, 3 mL of H2O2 (a.q., 30 wt%) was added
into the fask through a peristaltic pump (The feeding time
lasted for ca. 30 min). After that, the reaction was performed
for 1–6 h. During the process, a small amount of reaction
mixture (< 0.15 mL) was collected periodically, and cen-
trifuged. The liquid phase was analyzed by a gas chroma-
tograph using a SP-7820 capillary column. The conversion
(Conv.) of benzene and selectivity (Sel.) to phenol were
1 3

Z.-H. Yu et al.
Fig. 1 N2 adsorption–des-
orption isotherms (A) of
A
B
g-C3N4 (a), eg-C3N4-400 (b),
eg-C3N4-500 (c), 0.5FeCl3/
eg-C3N4 (d), 1.0FeCl3/eg-C3N4
(e), 1.5FeCl3/eg-C3N4-500 (f),
2.0FeCl3/eg-C3N4-500 (g), and
2.5FeCl3/eg-C3N4-500 (h) and
the corresponding pore size
distributions (B)
h
h
g
f
g
f
e
e
d
20
D
c
V
d
c
b
b
a
a
0.0
0.2
0.4
0.6
0.8
1.0 2
6
10
14
18
22
26
30
p/p0
Pore size (nm)
motif of graphitic structure corresponding to (100) plane.
The eg-C3N4 samples exhibit the similar patterns to that
of g-C3N4. However, the overall peak intensity is weaken
after the thermal treatment. Furthermore, compared with
that of g-C3N4, the location of (002) peaks of eg-C3N4 sam-
ples shifts towards higher 2θ angles. This means that the
interlayer stacking structure of the original g-C3N4 under-
goes shrinkage during the thermal treatment. The reduction
of interplanar distance has also been found in the cases of
exfoliated [37, 38] and/or thermal treated g-C3N4 [26, 39]
materials reported previously.
Table 1 Textural parameters of various materials
Sample
SBET (m2 g−1
)
Pore volume
(cm3 g−1
)
g-C3N4
eg-C3N4-400
eg-C3N4-500
0.5FeCl3/eg-C3N4
1.0FeCl3/eg-C3N4
1.5FeCl3/eg-C3N4
2.0FeCl3/eg-C3N4
2.5FeCl3/eg-C3N4
14
17
25
22
15
10
17
7
0.04
0.08
0.13
0.17
0.07
0.04
0.07
0.02
For FeCl3/g-C3N4 materials, the difraction peak corre-
sponding to (002) and (100) planes are almost unchanged
compared with the eg-C3N4 support (Fig. 3), suggesting
that the incorporation of FeCl3 does not change the original
graphitic structure of eg-C3N4. In addition, no signifcant
difraction peak related to FeCl3 and Fe2O3 is detected in the
pattern of FeCl3/eg-C3N4, implying that Fe species might
disperse well on the surface of eg-C3N4. Nonetheless, with
the increase of the loading amount of FeCl3, the intensity
of (002) planes is weakened gradually and the (100) plane
seems to disappear under higher loading amount. The simi-
lar phenomena have been reported in previous papers con-
cerning g-C3N4 materials doped with Fe and Zn [25, 34].
Figure 4 shows the FT-IR spectra of the materials and
all the spectra exhibit similar bands. The broad transmit-
tance centered at ca. 3200 cm−1 are ascribed to the stretch-
ing modes of N–H of primary and secondary amino groups
[40]. The multiple bands appearing in the range of 1600 to
1200 cm−1 are associated with the aromatic nitrogen-con-
taining heterocycles [41]. The sharp bands of ca. 808 cm−1
Fig. 2 XRD patterns of g-C3N4 (a), eg-C3N4-400 (b), and eg-
C3N4-500 (c) materials. The right is the magnifed region from 26°
to 29°
1 3

Direct Catalytic Hydroxylation of Benzene to Phenol Catalyzed by FeCl3 Supported on…
Fig. 3 XRD patterns of eg-C3N4 (a), 0.5FeCl3/eg-C3N4 (b), 1.0FeCl3/
eg-C3N4 (c), 1.5FeCl3/eg-C3N4 (d), 2.0FeCl3/eg-C3N4 (e), and 2.5
FeCl3/eg-C3N4 (f) materials
Fig. 5 XPS spectra of eg-C3N4, 1.5FeCl3/eg-C3N4, 2.0FeCl3/eg-C3N4,
and FeCl3 materials
ratios of 1.5FeCl3/eg-C3N4 and 2.0FeCl3/eg-C3N4 are 0.91
and 0.97, respectively. The signifcant increase in terms of
carbon proportion points out the apparent loss of nitrogen
species of eg-C3N4 after the loading of FeCl3, which has
also been revealed in many metal-doped g-C3N4 materials
[9, 44, 45]. Wherein, the incorporation of transition metal
halides into g-C3N4 would facilitate further decomposi-
tion of nitrogen-containing species of g-C3N4. Besides the
three elements, two weak signals centered at ca. 713 eV and
199 eV, have been found in the spectra of 1.5FeCl3/eg-C3N4
and 2.0FeCl3/eg-C3N4, which are assigned as the Fe 2p and
Cl 2p orbits, respectively. This suggests that FeCl3 has been
smoothly supported on the surface of eg-C3N4.
In order to investigate the chemical environment of Fe
on the surface of FeCl3/eg-C3N4 materials, the detailed Fe
2p spectra in the range of 740–705 eV have been scanned,
as shown in Fig. 6. The spectra of 1.5FeCl3/eg-C3N4 and
2.0FeCl3/eg-C3N4 materials display two pronounced peaks
at 724.4 and 710.7 eV, corresponding to 2p1/2 and 2p3/2
orbits of Fe, respectively. Compared with the spectrum of
the pure FeCl3 compound, the two peaks show apparent
shift towards lower binding energies. That is, the positive
charging density of Fe3+ undergoes partial decline after the
supporting onto g-C3N4. Indeed, such a phenomenon has
been previously reported by several research groups includ-
ing ours [9, 25, 46, 47]. As explained by Wang et al. [25],
since g-C3N4 contains abundant nitrogen-containing species
(especially “nitrogen pots” bringing out adjacent tectonic
heptazine heterocycles, Scheme S1), the electrons of some
nitrogen functionalities interact with the 3d orbits of Fe. In
this sense, it is reasonable to infer that FeCl3 species on the
surface of eg-C3N4-500 is not merely physically adsorbed
but likely anchored by Fe–N coordinate bonds.
Fig. 4 FT-IR spectra of g-C3N4 (a), eg-C3N4 (b), eg-C3N4 (c),
0.5FeCl3/eg-C3N4 (d), 1.0FeCl3/eg-C3N4 (e), 1.5FeCl3/eg-C3N4 (f),
2.0FeCl3/eg-C3N4 (g), and 2.5FeCl3/eg-C3N4 (h) materials
are indexed as the breathing mode of triazine-based func-
tionalities, characteristic bands of g-C3N4 materials [41, 42].
Judging from the FT-IR spectra, it can be inferred that the
exfoliation and incorporation of FeCl3 have not changed the
basic chemical functionalities of g-C3N4.
The surface chemical composition of the materials were
further analyzed by XPS. The survey spectra (Fig. 5) evi-
dence the existence of carbon, nitrogen, and oxygen ele-
ments in the eg-C3N4, 1.5FeCl3/eg-C3N4, and 2.0FeCl3/
eg-C3N4 materials. Calculated from the peak areas, the
molar ratio of carbon to nitrogen (nC/nN) of eg-C3N4 was
ca. 0.745, slightly lower than the theoretical value of ideal
eg-C3N4 (nC/nN = 0.75) [43]. The corresponding nC/nN
1 3

Z.-H. Yu et al.
highest binding energy (401.0 eV) corresponds to the ter-
minal amino group (–NH2 and –NH–) located at the edges
of graphitic sheets of g-C3N4 [34]. The spectra of FeCl3/
eg-C3N4 resemble that of the eg-C3N4, whereas weak but
appreciable shoulders appear at ca. 400 eV in the cases of
the two FeCl3/eg-C3N4 spectra.
For further analysis, the N 1s spectra of the three materi-
als were deconvoluted and the results are summarized in
Table 2. The percentages of Na and Nb atoms of the eg-C3N4
support are 74.7% and 12.5% respectively. Upon incorporat-
ing FeCl3, the distribution of nitrogen species of eg-C3N4
undergoes obvious variation. Namely, the percentages of
Na species decline whereas the Nb and Nc species increase
remarkably from eg-C3N4 to FeCl3/eg-C3N4. The variation
of the nitrogen distribution is probably derived from partial
destruction of heptazine heterocycles (Na) of eg-C3N4 after
the loading of FeCl3, which has been observed in the cases
of transition-metal-doped g-C3N4 materials [9, 34, 51].
Fig. 6 Fe 2p spectra of 1.5FeCl3/eg-C3N4 (a), 2.0FeCl3/eg-C3N4 (b),
and FeCl3 (c) materials
The corresponding fne N 1s spectra of eg-C3N4 and
FeCl3/eg-C3N4 materials are depicted in Fig. 7. The nitro-
gen spectrum of eg-C3N4 can be deconvoluted into three
independent peaks. The primary signal with the lowest
binding energy (398.6 eV) is indexed as the sp2-hybridized
nitrogen atoms (Na, Scheme S1) in the form of C=N–C
functionalities in hepatize rings [43, 48]. The peak located
at 399.7 eV is assigned as sp2-hybridized nitrogen atom
(Nb) bonded with three other atoms (basically N–C3 and
C–N(–C)–H groups) [49, 50]. The last peak with the
Table 2 Distributions of various nitrogen species of eg-C3N4,
1.5FeCl3/eg-C3N4, and 2.0FeCl3/eg-C3N4 materials
Samples
Percentage of N component (mol%)
Nc
Nb
Na
eg-C3N4
1.5FeCl3/eg-C3N4
2.0FeCl3/eg-C3N4
12.8
15.7
13.8
12.5
18.9
19.1
74.7
65.4
67.1
A
B
C
Fig. 7 N 1s spectra of eg-C3N4 (A), 1.5FeCl3/eg-C3N4 (B), and 2.0FeCl3/eg-C3N4 (C) materials
1 3

Direct Catalytic Hydroxylation of Benzene to Phenol Catalyzed by FeCl3 Supported on…
Table 3 Catalytic performances of various catalysts in the hydroxyla-
tion of benzene to phenol
The metal–support interaction of FeCl3 and eg-C3N4 was
analyzed by UV–Vis DRS. As shown in Fig. 8, the spectra of
g-C3N4 and eg-C3N4 exhibit major optical absorption with
a cutting edge at ca. 457 nm, corresponding to a band gap
of ca. 2.7 eV. The absorption is derived from the band gap
between HOMO and LUMO in the polymeric melon units of
g-C3N4 [31, 34]. Compared with that of eg-C3N4, the spec-
tra of FeCl3/eg-C3N4 demonstrate noticeable red shifts, and
with the increase of the loading of FeCl3, the optical absorp-
tion is moved monotonously towards longer wavelengths.
The change indicates the variation of electronic structure of
eg-C3N4, which is attributed to the d–p repulsion between
Fe 3d and N 2p orbits of FeCl3 and eg-C3N4, respectively
[47]. Combining the UV–Vis and XPS spectra, it can be
inferred that the eg-C3N4 support has smoothly anchored
FeCl3. That is, FeCl3 is not merely physically supported on
the surface of eg-C3N4.
Entry
Catalyst
Con. (%)
Sel. (%)
Yield (%)
1
2
3
4
5
6
7
8
9
–
–
–
–
–
>99
>99
>99
>99
>99
>99
–
g-C3N4
<1
<1
12
13
17
18
22
22
<1
<1
12
13
17
18
22
22
eg-C3aN4
FeCl3
0.5FeCl3/eg-C3N4
1.0FeCl3/eg-C3N4
1.5FeCl3/eg-C3N4
2.0FeCl3/eg-C3N4
2.5FeCl3/eg-C3N4
11.2 mmol of benzene,
3 mL of aqueous H2O2 solution
(30wt %),1 mL of HOAc and 5 mL of acetonitrile as solvent,
Wcatal =50 mg, T=60 °C, and t=3 h
a50 mg of FeCl3
3.2 Catalyst Activity
catalyst recovery. In the cases of FeCl3/eg-C3N4, the catalytic
conversions of benzene have been remarkably improved. The
catalytic activity of 0.5FeCl3/eg-C3N4, notwithstanding its
low loading amount of FeCl3, is higher than that obtained in
the pure FeCl3. Meanwhile, the selectivity to phenol is up to
99%. The catalytic results suggest that the combination of
g-C3N4 and FeCl3 efectively enhances the hydroxylation of
benzene. Moreover, with the increase of the loading amount
of FeCl3, the catalytic conversion of benzene is gradually
increased. Among the catalysts evaluated, 2.0FeCl3/eg-C3N4
shows the highest catalytic yield of phenol up to 22%.
A blank test of hydroxylation of benzene without any cat-
alyst was performed and the result shows no product was
obtained (Table 3). Also, benzene is hardly converted in
the presence of g-C3N4 or eg-C3N4. After addition of FeCl3,
the conversion of benzene could reach up to 12%. Nonethe-
less, it is worth noting that FeCl3 is fully dissolved in liquid
phase of the reaction. Namely, the hydroxylation of benzene
catalyzed FeCl3 is homogeneous catalytic reaction, undoubt-
edly having the shortcoming in either product purifcation or
Fig. 8 UV–Vis spectra of
g-C3N4 (a), eg-C3N4-400 (b),
eg-C3N4-500 (c), 0.5FeCl3/eg-
C3N4 (d), 1.0FeCl3/eg-C3N4 (e),
1.5FeCl3/eg-C3N4 (f), 2.0FeCl3/
eg-C3N4 (g), and 2.0FeCl3/eg-
C3N4 (h) samples
1 3

Z.-H. Yu et al.
For further comparison, FeCl3 catalysts supported on var-
ious g-C3N4 materials have been evaluated under the same
reaction conditions and the results are listed in Table 4. It is
obvious that the catalyst supports have an important impact
on the fnal catalytic activity of the FeCl3 catalysts. More
specifcally, using the bare g-C3N4 as a support, the FeCl3
catalyst (2.0FeCl3/g-C3N4) exhibits a moderate benzene
conversion (12%). By comparison, using exfoliated g-C3N4
instead, the catalytic activity is substantially improved. One
possible reason is the textural properties of the support. The
eg-C3N4 materials own larger surface areas and higher pore
volumes than g-C3N4, therefore could expose large acces-
sible surface to disperse FeCl3. The benefcial efect of sur-
face area of g-C3N4 catalysts on the corresponding catalytic
activity has also been previously reported in many catalytic
cases [27, 52].
corresponding profle of the pure g-C3N4 material (Fig. 9A)
reveals a broad desorption peak in the range of 75 to 300 °C,
which is associated with its physical and chemical adsorp-
tion of benzene. Like g-C3N4, eg-C3N4-400 shows a similar
desorption peak in the TPD profle. By comparison, eg-
C3N4-500 exhibits a remarkable peak with higher intensity
than g-C3N4 and eg-C3N4-400, suggesting that the mate-
rial owns stronger adsorption ability than its counterparts.
Apparently, the TPD results correlate well the catalytic
activity described in Table 4.
Along with the benzene-TPD tests, we also performed
liquid-phase adsorption experiments to simulate the catalytic
reaction, and the resultant adsorption-equilibrium data were
given in Fig. 9B. For the three supports, The equilibrium
concentrations of benzene are all below the corresponding
initial concentrations, verifying that all the materials are
able to adsorb benzene in liquid phase. Moreover, the order
of adsorption capability of benzene is eg-C3N4-500 > eg-
C3N4-400>g-C3N4, which can reasonably explain the high-
est catalytic activity of FeCl3/eg-C3N4-500 in the direct
hydroxylation of benzene other two catalysts.
In both FeCl3/g-C3N4 and FeCl3/eg-C3N4 materials, FeCl3
is the catalytic active component for the catalytic hydroxy-
lation of benzene in this work. In order to further elucidate
the efect of the catalyst support in the benzene-involved
reaction, benzene-TPD experiments were carried out. The
To investigate the efects of reaction conditions on the
fnal catalytic performances, several experiments have been
conducted using 2.0FeCl3/eg-C3N4-500 as a representative
catalyst. As illustrated in Fig. 10, as the reaction temperature
is elevated from 50 to 60 °C, the conversion of benzene is
increased from 10% to 22%. On a further increase of the
temperature, the selectivity to phenol decreases. Moreover,
as the temperature is above 65 °C, the conversion also drops.
The declines of conversion and selectivity probably result
from the self-decomposition of H2O2 and over-oxidation of
phenol [8, 16] under higher temperature, respectively.
Table 4 Catalytic performances of FeCl3 catalysts supported on vari-
ous g-C3N4 materials
Catalyst
Con. (%)
Sel. (%)
Yield (%)
2.0FeCl3/g-C3N4
2.0FeCl3/eg-C3N4-400
2.0FeCl3/eg-C3N4-500
12
18
22
>99
>99
>99
12
18
22
11.2 mmol of benzene,
3 mL of aqueous H2O2 solution
(30 wt%),1 mL of HOAc and 5 mL of acetonitrile as solvent,
Wcatal =50 mg, T=60 °C, and t=3 h
Fig. 9 Benzene TPD profles
(A) and equilibrium concentra-
tions of benzene under various
initial concentrations (B) of
g-C3N4 (a), eg-C3N4-400 (b),
and eg-C3N4-500 (c)
A
B
1 3

Direct Catalytic Hydroxylation of Benzene to Phenol Catalyzed by FeCl3 Supported on…
catalytic productivity. This might due to the severe decom-
position of H2O2 induced by more amount of FeCl3.
Reproducibility is an important issue to examine an het-
erogeneous catalyst. A series of reactions have been submit-
ted using 2.0FeCl3/eg-C3N4 as a catalyst. After each test,
the catalyst was immediately fltrated, rinsed by ethanol,
and used for the subsequent catalytic test. Figure 12 shows
that, during the fve consecutive catalytic runs, the conver-
sion of benzene is very stable and meanwhile the selectivity
obtained in each run is also above 99%. The stable reproduc-
ibility of FeCl3/eg-C3N4 further evidences the steady inter-
action of FeCl3 on the eg-C3N4 support.
4 Conclusion
Fig. 10 Infuence of reaction temperature on the catalytic perfor-
mance of 2.0FeCl3/eg-C3N4. Reaction conditions: Vbenzene =1.0 mL,
VH2O2 =3.0 mL, V CH3CN =5.0 mL, V HOAc =1.0 mL, Wcatal.=50 mg,
and t=3 h
In summary, FeCl3/eg-C3N4 catalysts were prepared by an
impregnation method. The thermal exfoliation improves the
surface areas and pore volumes of g-C3N4. The characteriza-
tion spectra demonstrate that FeCl3 was steadily anchored
on the g-C3N4 support via interaction between Fe3+ and
nitrogen atoms of g-C3N4. In the hydroxylation of benzene
to phenol in the presence of H2O2, FeCl3/eg-C3N4 materi-
als exhibit good catalytic activity, which was related to the
reaction conditions. The maximum yield of phenol is ca.
22%. Furthermore, FeCl3/eg-C3N4 shows superior activity to
FeCl3/g-C3N4, which is attributed to the diferent adsorption
and activation of benzene on the catalyst supports.
Figure 11A plots the catalytic results of hydroxylation
of benzene during 5 h. It can be found that the conversion
increases initially while stays at a constant level after 3 h.
However, the selectivity to phenol collected across the reac-
tion process is above 99%. In addition to the reaction tem-
perature and time, the catalytic performance is also sensitive
to the amount of catalyst. As described in Fig. 11B, in the
presence of only 25 mg of 2.0FeCl3/eg-C3N4, the phenol
yield reaches ca. 11%. Further adding the amount of the cat-
alyst could produce much higher yield of phenol. Whereas,
the use of excessive amount of the catalyst depresses the
Acknowledgements This work was supported by National Natu-
ral Science Foundation of China (Grant No. 21878027), PetroChina
Fig. 11 Efects of reaction time
(A) and amount of catalyst (B)
on the catalytic performance
of 2.0FeCl3/eg-C3N4. Reaction
conditions: Vbenzene =1.0 mL,
VH2O2 =3.0 mL, V
A
B
CH3CN =5.0 mL, V
HOAc =1.0 mL, T=60 °C,
Wcatal.=50 mg (A), and t=3 h
(B)
1 3

Z.-H. Yu et al.
13. Alsabagh AM, Yehia FZ, Eshaq G, Elmetwally AE (2017) ACS
Sustain Chem Eng 5:4811–4819
14. Jourshabani M, Badiei A, Shariatinia Z, Lashgari N, Ziarani GM
(2016) Ind Eng Chem Res 55:3900–3908
15. Zhang P, Gong Y, Li H, Chen Z, Wang Y (2013) RSC Adv
3:5121–5126
16. Hu L, Wang C, Ye L, Wu Y, Yue B, Chen X, He H (2015) Appl
Catal A 504:440–447
17. Gao Y, Gao Y, Shi L, Sun Q (2017) Catal Lett 147:2799–2806
18. Muniandy L, Adam F, Mohamed AR, Iqbal A, Rahman NRA
(2017) Appl Sur Sci 398:43–55
19. Sanjini NS, Velmathi S (2016) J Porous Mater 23:1527–1535
20. Borah P, Ma X, Nguyen KT, Zhao Y (2012) Angew Chem Int Ed
51:7756–7761
21. Xu J, Jiang Q, Shang J-K, Wang Y, Li Y-X (2015) RSC Adv
5:92526–92533
22. Long Z, Zhou Y, Chen G, Ge W, Wang J (2014) Sci Rep 4:3651
23. Wang H, Fang L, Yang Y, Zhang L, Wang Y (2016) Catal Sci
Technol 6:8005–8015
Fig. 12 Catalytic performances of 2.0FeCl3/eg-C3N4-500 dur-
ing fve consecutive runs. Reaction conditions: Vbenzene =1.0 mL,
VH2O2 =3.0 mL, V CH3CN =5.0 mL, V HOAc =1.0 mL, Wcatal =50 mg,
T=60 °C, and t=3 h
24. Jiang T, Wang W, Han B (2013) New J Chem 37:1654–1664
25. Wang X, Chen X, Thomas A, Fu X, Antonietti M (2009) Adv
Mater 21:1609–1612
26. Li Y, Xu H, Ouyang S, Lu D, Wang X, Wang D, Ye J (2016) J
Mater Chem A 4:2943–2950
27. Goettmann F, Fischer A, Antonietti M, Thomas A (2006) Angew
Chem Int Ed 45:4467–4471
Innovation Foundation (Grant No. 2018D-5007-0508), Advanced
Catalysis and Green Manufacturing Collaborative Innovation Center
(Grant No. ACGM2016-06-28), Natural Science Foundation of the
Jiangsu Higher Education Institutions (Grant Nos. 18KJA150001 and
19KJA430003), and the Top-notch Academic Programs Project of
Jiangsu Higher Education Institutions. J. Xu also thanks Dr. Jun-Jing
Ding of the Shiyanjia Lab (www.shiyanjia.com) for his kind help in
XPS characterization.
28. Xu J, Wu H-T, Wang X, Xue B, Li Y-X, Cao Y (2013) Phys Chem
Chem Phys 15:4510–4517
29. Yang Z-Z, He L-N, Dou X-Y, Chanfreau S (2010) Tetrahedron
Lett 51:2931–2934
30. Zheng Y, Liu J, Liang J, Jaroniec M, Qiao SZ (2012) Energy
Environ Sci 5:6717–6731
31. Chen X, Zhang J, Fu X, Antonietti M, Wang X (2009) J Am Chem
Soc 131:11658–11659
32. Goettmann F, Fischer A, Antonietti M, Thomas A (2006) Chem
Commun:4530-4532
Compliance with Ethical Standards
33. Datta KKR, Reddy BVS, Ariga K, Vinu A (2010) Angew Chem
Int Ed 49:5961–5965
Conflicts of interest There are no conficts of interest to declare for
each contributing author.
34. Xu J, Long K-Z, Wang Y, Xue B, Li Y-X (2015) Appl Catal A
496:1–8
35. Wang Y, Jiang Q, Shang J-K, Xu J, Li Y-X (2016) Acta Phys
-Chim Sin 32:1913–1928
References
36. Hu L, Wang C, Yue B, Chen X, He H (2016) RSC Adv
6:87656–87664
37. Xu J, Chen Y, Hong Y, Zheng H, Ma D, Xue B, Li Y-X (2018)
Appl Catal A 549:31–39
1. Wang X, Tan X, Meng B, Zhang X, Liang Q, Pan H, Liu S (2013)
Catal Sci Technol 3:2380–2391
38. Niu P, Zhang L, Liu G, Cheng HM (2012) Adv Funct Mater
22:4763–4770
2. Yang J-H, Sun G, Gao Y, Zhao H, Tang P, Tan J, Lu A-H, Ma D
(2013) Energy Environ Sci 6:793–798
39. Xu H, Yan J, She X, Xu L, Xia J, Xu Y, Song Y, Huang L, Li H
(2014) Nanoscale 6:1406–1415
3. Borah P, Datta A, Nguyen KT, Zhao Y (2015) Green Chem
18:397–401
40. Talapaneni SN, Anandan S, Mane GP, Anand C, Dhawale DS,
Varghese S, Mano A, Mori T, Vinu A (2012) J Mater Chem
22:9831–9840
4. Ye X, Cui Y, Wang X (2014) Chemsuschem 7:738–742
5. Carneiro L, Silva AR (2016) Catal Sci Technol 6:8166–8176
6. Han JW, Jung J, Lee Y, Nam W, Fukuzumi S (2017) Chem Sci
8:7119–7125
41. Bojdys MJ, Müller J-O, Antonietti M, Thomas A (2008) Chem
-Eur J 14:8177–8182
7. Yamada M, Karlin KD, Fukuzumi S (2016) Chem Sci
7:2856–2863
42. Wang X, Blechert S, Antonietti M (2012) ACS Catal 2:1596–1606
43. Thomas A, Fischer A, Goettmann F, Antonietti M, Müller J-O,
Schlögl R, Carlsson JM (2008) J Mater Chem 18:4893–4908
44. Yue B, Li Q, Iwai H, Kako T, Ye J (2011) Sci Technol Adv Mater
12:034401
8. Ding G, Wang W, Jiang T, Han B, Fan H, Yang G (2013) Chem-
CatChem 5:192–200
9. Xue B, Chen Y, Hong Y, Ma D-Y, Xu J, Li Y-X (2018) Chin J
Catal 39:1263–1271
45. Chen Z, Zhang J, Zheng S, Ding J, Sun J, Dong M, Abbas M,
Chen Y, Jiang Z, Chen J (2018) Mol Catal 444:90–99
46. Xu J, Gan Y-L, Hu P, Zheng H, Xue B (2018) Catal Sci Technol
8:5582–5593
10. Zhou Y, Ma Z, Tang J, Yan N, Du Y, Xi S, Wang K, Zhang W,
Wen H, Wang J (2018) Nat Commun 9:2931
11. Guo B, Zhu L, Hu X, Zhang Q, Tong D, Li G, Hu C (2011) Catal
Sci Technol 1:1060–1067
47. Tonda S, Kumar S, Kandula S, Shanker V (2014) J Mater Chem
2:6772–6780
12. Xu J, Jiang Q, Chen T, Wu F, Li Y-X (2015) Catal Sci Technol
5:1504–1513
1 3

Direct Catalytic Hydroxylation of Benzene to Phenol Catalyzed by FeCl3 Supported on…
48. Su Q, Sun J, Wang J, Yang Z, Cheng W, Zhang S (2014) Catal Sci
Technol 4:1556–1562
52. Yang S, Gong Y, Zhang J, Zhan L, Ma L, Fang Z, Vajtai R, Wang
X, Ajayan PM (2013) Adv Mater 25:2452–2456
49. Huang Z, Li F, Chen B, Lu T, Yuan Y, Yuan G (2013) Appl Catal
B 136:269–277
Publisher’s Note Springer Nature remains neutral with regard to
jurisdictional claims in published maps and institutional afliations.
50. Lan D-H, Wang H-T, Chen L, Au C-T, Yin S-F (2016) Carbon
100:81–89
51. Zhu J, Wei Y, Chen W, Zhao Z, Thomas A (2010) Chem Commun
46:6965–6967
Afliations
Zhou‑Hong Yu1 · Yu‑Lin Gan1 · Jie Xu1 · Bing Xue1
1
Jiangsu Key Laboratory of Advanced Catalytic Materials
and Technology, School of Petrochemical Engineering,
Changzhou University, Gehu Middle Road 21,
Changzhou 213164, Jiangsu, People’s Republic of China
1 3