Preparation of Fe@Cu-BDC and its catalytic performance for benzene hydroxylation

Article abstract of DOI:10.1002/aoc.4662

Novel Fe@Cu-BDC catalysts were synthesized by adsorbing various contents of Fe ions on the surface and into the channels of Cu-BDC support through an in situ method under ultrasonication. The structure of the catalysts was characterized using X-ray diffraction, scanning electron microscopy, Fourier transform infrared spectroscopy, energy-dispersive X-ray spectrometry, X-ray photoelectron spectroscopy and thermogravimetric analysis. The catalytic activity of catalysts for benzene hydroxylation was evaluated. The results showed that the Fe@Cu-BDC catalysts had favorable catalytic activity for benzene hydroxylation. The yield of products can reach ca 37% and the selectivity of phenol can reach ca 62% over the 6% Fe@Cu-BDC catalyst at 60°C in a reaction time of 75 min. Furthermore, the contents of Fe and solvent played a key role in benzene hydroxylation. After the reaction, the catalyst could be easily separated from the reaction mixture by centrifugation and reused for five times without significant decrease in activity.

Full text of DOI:10.1002/aoc.4662

Received: 11 July 2018
DOI: 10.1002/aoc.4662
Revised: 2 September 2018
Accepted: 29 September 2018
F U L L P A P E R
Preparation of Fe@Cu‐BDC and its catalytic performance
for benzene hydroxylation
Boliang Liu1
Yaru Li1
| Songhai Wu1 | Yong Liu2 | Yanfeng Liu3 | Wei Cai1
| Qianyi Sun1 |
1 School of Chemical Engineering and
Technology, Tianjin University, Tianjin
300350, China
2 School of Chemistry and Chemical
Engineering, Tianjin University of
Technology, Tianjin 300384, China
3 College of Materials and Chemical
Engineering, Hainan University, Haikou
570228, China
Novel Fe@Cu‐BDC catalysts were synthesized by adsorbing various contents of
Fe ions on the surface and into the channels of Cu‐BDC support through an in
situ method under ultrasonication. The structure of the catalysts was character-
ized using X‐ray diffraction, scanning electron microscopy, Fourier transform
infrared spectroscopy, energy‐dispersive X‐ray spectrometry, X‐ray photoelec-
tron spectroscopy and thermogravimetric analysis. The catalytic activity of cat-
alysts for benzene hydroxylation was evaluated. The results showed that the
Fe@Cu‐BDC catalysts had favorable catalytic activity for benzene hydroxyl-
ation. The yield of products can reach ca 37% and the selectivity of phenol
can reach ca 62% over the 6% Fe@Cu‐BDC catalyst at 60°C in a reaction time
of 75 min. Furthermore, the contents of Fe and solvent played a key role in
benzene hydroxylation. After the reaction, the catalyst could be easily sepa-
rated from the reaction mixture by centrifugation and reused for five times
without significant decrease in activity.
Correspondence
Songhai Wu, School of Chemical
Engineering and Technology, Tianjin
University, Tianjin 300350, China.
Email: wusonghaitjuedu@163.com
Funding information
Program of Introducing Talents of Disci-
pline to Universities, Grant/Award Num-
ber: B06006; National Natural Science
Foundation of China, Grant/Award
Number: 41373114
KEYWORDS
benzene hydroxylation, catalyst, Fe@Cu‐BDCMOFs
1 | INTRODUCTION
most. Unfortunately, the peroxide intermediate is not
stable in the process. In order to solve these problems,
Phenol, as a significant raw material in the chemical
industry, is commercially produced mainly by the
cumene process, which involves three steps and suffers
from a need for high pressure and high temperature.
Some other disadvantages such as low one‐pass yield of
phenol (ca 5%), low atomic efficiency (61.8%)[1] and a
large quantity of acetone by‐product are also features of
this process. One of the most effective alternative
methods is the one‐step synthesis of phenol by direct
oxidation of benzene under mild conditions, which is a
challenge from the economic and environmental points
of view. And a number of systems have been reported
so far.[2] Several oxidants have been employed, such as
molecular oxygen,[3] dinitrogen oxide[4] and hydrogen
peroxide, among which hydrogen peroxide has been used
studies of the direct oxidation of benzene to phenol with
various metal‐based catalysts have been reported in the
literature.[5–11] For example, Fernandes et al.[5] discussed
the reaction over the compound [Fe(gma)(PBu3)].
Though the yield of phenol was not low, the reaction time
was a little long. Esmelindro et al.[12] prepared mononu-
clear [Cu(BTMEA)2Cl]Cl and applied it in the reaction,
but the yield and the selectivity of phenol were only
2.23 and 24.9%, respectively.
Metal–organic frameworks (MOFs) are kinds of
porous materials and formed by metal ions and organic
ligands. Due to their advantages of large surface area,
tunable pore size and easy functionalization and design,
they have attracted much attention in the field of hetero-
geneous catalysis.[13–15] Unfortunately, this aspect is still
Appl Organometal Chem. 2018;e4662.
wileyonlinelibrary.com/journal/aoc
© 2018 John Wiley & Sons, Ltd.
1 of 10
https://doi.org/10.1002/aoc.4662

2 of 10
LIU ET AL.
in its infancy and also few strategies were performed.[16]
For example, MOFs could be directly used as heteroge-
neous catalysts (the active site was either the metal or
organic ligands in the structure of MOFs). Farzaneh and
Mortazavi[17] explored the alkylation of toluene with ben-
zyl bromide over Zn(BTC). Chung et al.[18] investigated
the Friedel–Crafts acylation of p‐xylene with benzyl
chloride over sulfonated UiO‐66 catalysts. Also, active
components can be loaded in porous MOFs. For instance,
Esken et al.[19] prepared a series of Au@MOF‐5 catalysts,
and applied them in liquid‐phase aerobic alcohol oxida-
tion. Gao et al.[20] investigated the Sonogashira coupling
reaction over a Pd/MOF‐5 catalyst. However, the solvent
effect of these MOFs catalysts has seldom been discussed
in the literature.
simultaneously, Fe2+ partially replaced Cu2+ in the crys-
tal framework to prepare Fe/Cu‐BTC bimetallic material.
In the work reported here, owing to the novelty of
loaded MOF catalysts, and their favorable catalytic activ-
ity for benzene hydroxylation, we aimed to adsorb Fe ions
on the surface and into the channels of Cu‐BDC support
and thus prepare a novel loaded MOF catalyst. By this
means, the efficiency of Fe ions in benzene hydroxylation
could be enhanced due to the prominent ability of
Cu‐BDC to adsorb benzene and desorb phenol. Based
on this notion, Fe@Cu‐BDC catalysts were synthesized
by adsorbing various contents of Fe ions on the surface
and into the channels of Cu‐BDC support using an in situ
method under ultrasonication. Then, the structure of the
catalysts was characterized using a series of measure-
ments. The performance of these catalysts in benzene
hydroxylation was evaluated.
Fenton reagent, which mainly consists of hydrogen
peroxide and free Fe ions, is an excellent catalyst for ben-
zene hydroxylation. Unfortunately, the yield of reaction is
not satisfactory because the catalytic activity of free Fe
ions is not high in the solution as expected. From another
aspect, on account of the fact that the activity of adsorbed
ions is much higher than that of free ions in solution
because the former are always in an unsaturated state
while the latter are always saturated, we aim to let Fe
ions adsorb on the surface and into the channels of a
MOF material to prepare a Fenton‐like reagent with
adsorbed Fe ions and apply it in benzene hydroxylation
to increase the yield of reaction. In order to achieve this
target, metal ions that have the same valence state as
and similar ionic radius to the Fe ion are often selected.
Among plenty of typical MOF materials that have the
above properties, Cu‐BDC is undoubtedly a notable one
because the metal ion of Cu‐BDC is the Cu ion that is
qualified to meet the demands mentioned above and,
most importantly, the organic ligand of Cu‐BDC is 1,4‐
benzenedicarboxylic acid (H2BDC) with two carboxyl
groups. Since the carboxyl group is hydrophilic, the
hydrophobicity of Cu‐BDC is stronger to some extent
than that of other classic MOF materials whose organic
ligands have three or more carboxyl groups. Meanwhile,
in benzene hydroxylation, the reactant benzene is hydro-
phobic but the product phenol is hydrophilic. As a result,
when the reaction is proceeding initially, the reactant
benzene is prone to adsorb on the surface of Cu‐
BDC.[21] After benzene has converted into phenol, phenol
is prone to desorb from the surface of the Cu‐BDC MOF
material. In this way, it creates a favorable condition to
increase the yield of reaction. Thus, we decided to use
Cu‐BDC rather than other classic MOFs to prepare a
novel Fe‐loaded material. For example, Nenoff and
co‐workers[22] used FeCl2⋅4H2O to partly replace Cu of
Cu‐BTC in N,N‐dimethylformamide (DMF) solution.
With maintaining of the integrity of the crystal structure
2 | EXPERIMENTAL
2.1 | Preparation of samples
Pure Cu‐BDC was synthesized according to a typical
method. In detail, Cu(OAc)2⋅H2O (0.659 g, 3.1 mmol)
and H2BDC (0.506 g, 3.1 mmol) were dissolved in DMF
(20 ml) under vigorous stirring. Then H2BDC solution
was added into the copper salt solution dropwise with
vigorous stirring at room temperature. After that, the
mixture reacted for 2 h under ultrasonication at 200 W
(KQ‐5200DB, Kunshan Ultrasonic Instruments Co. Ltd).
The obtained light‐blue Cu‐BDC powder was separated
by centrifugation and washed three times each
with DMF and methanol. Finally, the sample was put
in a freeze dryer under vacuum for 24 h. In order to
verify the purity of the as‐synthesized sample, it was
subjected to elemental analysis using an Elementar
Analysensysteme to determine the contents of C, H and
N. The elemental analysis values of Cu‐BDC were as fol-
lows: C, 40.86%; H, 1.94%; N, 1.19%. In the theoretical cal-
culation, the contents of elements were as follows: C,
42.11%; H, 1.75%; N, 0%. The actual measurement values
were basically in accordance with the theoretical values.
Various amounts of FeCl2⋅4H2O (0.099, 0.132, 0.198,
0.297, 0.396 mmol) and Cu(OAc)2⋅H2O (0.659 g,
3.1 mmol) were dissolved in DMF (20 ml). The other syn-
thesis steps were the same as described above. The
obtained blue‐green Fe@Cu‐BDC powder was separated
by centrifugation and washed three times each with
DMF and methanol. Finally, the sample was put in a
freeze dryer under vacuum for 24 h. Fe@Cu‐BDC cata-
lysts were prepared with various contents of FeCl2⋅4H2O:
3% Fe@Cu‐BDC, 4% Fe@Cu‐BDC, 6% Fe@Cu‐BDC, 9%

LIU ET AL.
3 of 10
Fe@Cu‐BDC and 12% Fe@Cu‐BDC. These catalysts are
denoted as Fe@Cu‐BDC‐1, Fe@Cu‐BDC‐2, Fe@Cu‐
BDC‐3, Fe@Cu‐BDC‐4 and Fe@Cu‐BDC‐5, respectively.
The ultrasonication process is presented in Figure 1.
performed with a Pyris 1 TGA instrument (PerkinElmer
Corp., USA) from 30 to 500°C in nitrogen atmosphere at
a constant rate of 10°C min−1. A Shimadzu HPLC instru-
ment was used for analyzing products, which consisted of
a binary LC‐20 AD pump, an SIL‐20 AC auto sampler, a
CTO‐20A column oven, an SPD‐M20A PDA detector, a
CBM‐20A system controller and a Shim‐pack VP‐ODS
column (150 mm × 4.6 mm, 5 μm).
2.2 | Characterization
X‐ray diffraction (XRD) patterns of as‐synthesized sam-
ples were obtained using a Rigaku D/Max 2500 VB+/PC
diffractometer with a graphite monochromator, using
Cu Kα radiation operating at 200 mA and 40 kV. Elemen-
tal analysis was performed with an Elementar
Analysensysteme to determine the contents of C, H and
N. Scanning electron microscopy (SEM) was performed
with a Hitachi S‐4800 microscope. Energy‐dispersive
X‐ray spectrometry (EDS) was performed with a VEGA‐
Tescan to determine the contents of Fe and Cu. Fourier
transform infrared (FT‐IR) spectra were obtained with a
Bruker Tensor‐27 Fourier transform infrared spectrome-
ter in the range 400–4000 cm−1 (KBr discs). The chemical
composition of the samples was analyzed using an X‐ray
photoelectron spectrometry (XPS) instrument equipped
with an Al K X‐ray source operated at 1486.6 eV. The
anode was operated at 12 kV and 20 mA. A hemispherical
energy analyzer (Specs EA 10 Plus) operating in a vac-
uum better than 10−7 Pa was used to determine the
core‐level binding energies of photoelectrons emitted
from the surface. Binding energy values were calibrated
by fixing the C (1s) core level with a binding energy of
284.5 eV. Thermogravimetric analysis (TGA) was
2.3 | Evaluation of benzene hydroxylation
Benzene hydroxylation was carried out in a flask reactor
with a condenser and magnetic stirring. The process of
the reaction is exhibited in Figure 2. The effects of
various reaction times, catalysts and solvent dosage were
investigated. Briefly, certain amounts of benzene (1 ml),
hydrogen peroxide (3 ml), acetonitrile (4–10 ml) and cat-
alyst (150 mg) were added into a round‐bottom flask.
After the reaction, the supernatant was collected by cen-
trifugation and analyzed through HPLC. The mobile
phase consisted of a mixture of methanol and water
with a volume ratio of 35:65. The flow rate was 0.5
ml min−1. The column was maintained at 35°C and
detection was at a wavelength of 270 nm. The yield
and the selectivity of products were calculated by exter-
nal standard method. The solid catalyst separated by
centrifugation was washed with plenty of methanol after
reaction, and was put in a freeze dryer under vacuum
for 24 h.
FIGURE 1 Process of ultrasonication
FIGURE 2 Process of benzene hydroxylation

4 of 10
LIU ET AL.
could retain its structure during the synthesis procedure.
What is more, the intensity of Cu‐BDC diffraction peaks
gradually decreased with increasing Fe content. Thus,
when preparing loaded MOF materials through an in situ
method, the quantity of the loaded metal should be con-
trolled to avoid its excessive effect on the crystal structure
of the MOF.
3 | RESULTS AND DISCUSSION
3.1 | Characterization of samples
3.1.1 | XRD patterns of as‐synthesized
catalysts
XRD patterns of as‐synthesized samples are shown in
Figure 3. Diffraction peaks at 10.1°, 12.0°, 17.0°, 20.4°
and 24.7° of Cu‐BDC were consistent with the litera-
ture.[23] This validated the formation of Cu‐BDC during
the ultrasonic process. The results of different contents
of Fe encapsulated in Cu‐BDC are presented as patterns
(1) to (5). The diffraction peaks of Cu‐BDC were still kept
after encapsulation with Fe, illustrating that Cu‐BDC
3.1.2 | SEM of as‐synthesized catalysts
Figure 4 shows SEM images of as‐synthesized Cu‐BDC
and Fe@Cu‐BDC samples. It can be seen from
Figure 4a that the prepared Cu‐BDC sample had spher-
ical and cubic particles, and the size of them was about
500 nm. As can be seen from Figure 4b, the size of par-
ticles varies from dozens of nanometers to 100 nm or
more after encapsulation with Fe2+. As shown in
Figure 4c–f, it is apparent that the particles were almost
small granules of dozens of nanometers. In addition, the
surfaces of samples were not smooth. It seems like small
particles agglomerated together and formed larger
Fe@Cu‐BDC particles.
3.1.3 | FT‐IR spectra of as‐synthesized
catalysts
Figure 5 presents the FT‐IR spectra of pure Cu‐BDC, the
Fe@Cu‐BDC catalysts and H2BDC. As can be seen, the
strong vibration peak at 1685 cm−1 in spectrum (6) could
be ascribed to the ─COOH stretching vibration of
H2BDC.[23,24] For spectrum (0), the strong vibration at
1605 cm−1 could be attributed to the connection of
deprotonated ─COOH with Cu ion.[25,26] In other words,
FIGURE 3 XRD patterns of as‐synthesized samples: (0) Cu‐BDC,
(1) Fe@Cu‐BDC‐1, (2) Fe@Cu‐BDC‐2, (3) Fe@Cu‐BDC‐3, (4)
Fe@Cu‐BDC‐4 and (5) Fe@Cu‐BDC‐5
FIGURE 4 SEM images of as‐synthesized samples: (a) Cu‐BDC, (b) Fe@Cu‐BDC‐1, (c) Fe@Cu‐BDC‐2, (d) Fe@Cu‐BDC‐3, (e) Fe@Cu‐
BDC‐4 and (f) Fe@Cu‐BDC‐5

LIU ET AL.
5 of 10
TABLE 1 Molar ratio of Fe and Cu of as‐synthesized samples
Value of theoretical
calculation
Value of actual
measurement
Sample
Fe@Cu‐BDC‐1
Fe@Cu‐BDC‐2
Fe@Cu‐BDC‐3
Fe@Cu‐BDC‐4
Fe@Cu‐BDC‐5
3:100
4:100
6:100
9:100
12:100
3.0:100
3.7:100
5.6:100
6.9:100
18.9:100
for Fe@Cu‐BDC‐2, Fe@Cu‐BDC‐3 and Fe@Cu‐BDC‐4,
values of actual measurement were lower than those of
the theoretical calculation. This may be attributed to a
small amount of Fe not being encapsulated successfully
into the channels of the Cu‐BDC support. For Fe@Cu‐
BDC‐5, it was noticeable that the value of actual mea-
surement was higher than that of theoretical calculation.
This may be ascribed to a small quantity of Fe occupying
the sites which were originally supposed to be occupied
by Cu under the condition of high ratio. As a conse-
quence, the actual ratio was a little higher than the
theoretical one. Therefore, when preparing loaded MOF
materials by an in situ method, the amount of the loaded
metal ought to be controlled to avoid its excessive influ-
ence on the metal active site of the MOF.
FIGURE 5 FT‐IR spectra of as‐synthesized samples: (0) as‐
synthesized Cu‐BDC, (1) Fe@Cu‐BDC‐1, (2) Fe@Cu‐BDC‐2, (3)
Fe@Cu‐BDC‐3, (4) Fe@Cu‐BDC‐4, (5) Fe@Cu‐BDC‐5 and (6)
H2BDC
this strong band was caused by the asymmetric stretching
vibration of ─COOH. It was notable that the intensity of
the band at 1605 cm−1 gradually decreased with increas-
ing Fe content. This is ascribed to the Fe ion weakening
the connection of deprotonated ─COOH with Cu ion.
The band at 1382 cm−1 was mainly owing to the symmet-
ric stretching vibration of ─COOH.[25,27] In comparison
with spectrum (6), these vibration peaks were shifted to
lower wavenumber. For the Fe@Cu‐BDC samples, the
strong asymmetric and symmetric vibrations were also
evident at 1605 and 1382 cm−1, respectively. This was
consistent with the pure Cu‐BDC. This indicated the
formation of Cu‐BDC structure in the process of synthe-
sizing Fe@Cu‐BDC and also verified the XRD results of
as‐synthesized samples. However, the above asymmetric
and symmetric stretching vibration bands were all moved
to lower wavenumber compared with spectrum (6).
Moreover, the vibration band at 1501 cm−1 was pres-
ent in all spectra. This could attributed to the C═C
stretching vibration of p‐disubstituted aromatic ring,
which meant the vibration of 1,4‐dicarboxylate
substituted on benzene.[28] Furthermore, several bands
were observed in the range 568–1190 cm−1 which could
be attributed to the out‐of‐plane vibration of BDC.
3.1.5 | XPS of as‐synthesized catalysts
XPS was performed to further determine the surface
chemical composition of the Fe@Cu‐BDC‐3 catalyst and
the valence states of Fe species present therein. Figure 6
a illustrates the XPS survey spectrum of the fresh
Fe@Cu‐BDC‐3 catalyst that consists of C, O, Cu and Fe
elements, with sharp photoelectron peaks appearing at
binding energies of 284.5 (C 1s), 533 (O 1s) and
934.5 eV (Cu 2p), and a weak photoelectron peak at
around 720 eV (Fe 2p). In Figure 6b, the Fe 2p spectrum,
two characteristic peaks at binding energies of 716.2 and
724.2 eV were assigned to Fe 2p3/2 and Fe 2p1/2, respec-
tively, which are typical for the standard XPS spectrum
of metallic Fe in O2/Fe/Cu.[29,30] Also, Figure 6c exhibits
the XPS survey spectrum of the used Fe@Cu‐BDC‐3 cata-
lyst that is composed of C, O, Cu and Fe elements, with
sharp photoelectron peaks appearing at binding energies
of 284.5 (C 1s) and 533 eV (O 1s) and two weak photo-
electron peaks at around 934.5 (Cu 2p) and 716 eV (Fe
2p). It is also observed that the intensity of Fe 2p peak
for the used Fe@Cu‐BDC‐3 catalyst was lower than that
for the fresh Fe@Cu‐BDC‐3 catalyst, which is attributed
to the loss during the experimental process. What is
more, the XPS spectrum (Figure 6d) of the Fe 2p region
3.1.4 | EDS of as‐synthesized catalysts
The molar ratios of Fe and Cu of Fe@Cu‐BDC‐1,
Fe@Cu‐BDC‐2, Fe@Cu‐BDC‐3, Fe@Cu‐BDC‐4 and
Fe@Cu‐BDC‐5 are presented in Table 1. The value of
actual measurement was consistent with the theoretical
calculation for Fe@Cu‐BDC‐1. This indicated that Fe
really existed in Fe@Cu‐BDC‐1 as expected. However,

6 of 10
LIU ET AL.
FIGURE 6 XPS spectra of as‐synthesized Fe@Cu‐BDC‐3 sample: (a) survey spectrum (fresh), (b) Fe 2p spectrum (fresh), (c) survey
spectrum (used) and (d) Fe 2p spectrum (used)
shows two peaks at 710.7 (Fe 2p3/2) and 724 eV (Fe 2p1/2).
This is characteristic of Fe3+ in Fe2O3, which is in good
agreement with earlier reports.[31,32] And the measure-
ment result indicated that the valence states of iron
species transformed from metallic to ferric iron after
the reaction.
of Fe is not beneficial for enhancing the thermal stability
of Fe@Cu‐BDC materials. Thus, when preparing
loaded MOF materials through an in situ method, the
quantity of the loaded metal should be controlled to
avoid its excessive impact on the thermal stability of
the materials.
3.1.6 | TGA of as‐synthesized catalysts
TGA curves of as‐synthesized samples are presented in
Figure 7. All the samples have a similar changing trend
from 30 to 290°C. For Cu‐BDC, its weight loss rate was
the least of all the samples, whereas the other samples
had almost the same trends. The skeleton structure of
Cu‐BDC began to degrade from 290°C and the final
product was CuO,[33] and there was Fe2O3[34] in the final
product of the Fe@Cu‐BDC samples in addition to CuO.
As can be seen, the weight loss of Fe@Cu‐BDC‐1,
Fe@Cu‐BDC‐2, Fe@Cu‐BDC‐3, Fe@Cu‐BDC‐4 and
Fe@Cu‐BDC‐5 samples rose with increasing content of
loaded Fe from 290 to 500°C. Moreover, it is obvious that
the temperature at which samples reached the minimum
mass point gradually decreased with increasing content
of loaded Fe in samples. This indicated that the addition
FIGURE 7 TGA curves for as‐synthesized samples: (0) as‐
synthesized Cu‐BDC, (1) Fe@Cu‐BDC‐1, (2) Fe@Cu‐BDC‐2, (3)
Fe@Cu‐BDC‐3, (4) Fe@Cu‐BDC‐4 and (5) Fe@Cu‐BDC‐5

LIU ET AL.
7 of 10
TABLE 2 Effect of reaction time on yield using Fe@Cu‐BDC‐3
catalyst
3.2 | Evaluation of catalytic performance
On the basis of the above characterization and analysis,
Fe@Cu‐BDC and pure Cu‐BDC catalysts were applied
in benzene hydroxylation, and, in particular, various
reaction times, catalysts, solvent dosages and the
recycling of the catalysts were investigated. The reaction
equation is shown in Scheme 1.
Reaction
time
(min)
Selectivity (%)
Yield
(%)
Entry
Phenol
Benzoquinone
1
2
3
4
5
6
7
30
45
28.74
23.11
37.38
43.27
11.64
26.21
28.42
71.26
76.89
62.62
56.73
88.36
73.79
71.58
9.96
10.41
17.30
37.56
7.78
60
75
3.2.1 | Effect of reaction time
90
120
150
5.39
Table 2 presents the catalytic activity of the Fe@Cu‐BDC‐
3 catalyst for benzene hydroxylation at 60°C. The content
of catalyst was fixed at 150 mg. Benzene reacted with
hydrogen peroxide to give two products in liquid phase,
namely phenol and benzoquinone.[35] Luo et al.[36]
explored the impact of TS‐1 catalysts on direct hydroxyl-
ation of benzene to phenol and found that a longer
reaction time resulted in an increased formation of
over‐oxidation products.[37] Furthermore, Capocasa
et al.[38] discussed the effect of iminopyridine Fe(II)
complex on the direct hydroxylation of benzene, and they
found that during the first 90 min, the yield undoubtedly
rose. Then, the yield decreased subsequently. As is evi-
dent from Table 2, the yield increased gradually up to
75 min. This may be attributed to that the effective utili-
zation ratio of hydrogen peroxide increased with increas-
ing reaction time. After that, an extension of reaction
time caused a certain amount of products to be polymer-
ized, which reduced the yield of the reaction, and this
corresponded to the above results. So the optimum reac-
tion time was 75 min.
4.62
Conditions: 150 mg of catalyst; benzene, 1 ml; H2O2, 3 ml; acetonitrile, 4 ml;
60°C.
increase in the molar ratio of catalyst to benzene from
0.019 to 0.027. From Table 3, when the content of loaded
Fe was 0%, the yield was very low compared with the
other catalysts, which indicated that pure Cu‐BDC cata-
lyst did not have a significant influence on the yield of
the benzene hydroxylation reaction. Therefore, the key
to benzene hydroxylation is the presence of Fe[10] and
the yield rose with increasing content of Fe as expected.
When the content of loaded Fe was 6%, the yield reached
the maximum value, probably resulting from the pres-
ence of more catalytic active sites. But when we went
on to further increase the content of loaded Fe, the yield
was observed to gradually decrease. This may be attrib-
uted to a great deal of loaded Fe excessively occupying
the active sites of the MOF material so that benzene
cannot approach them easily. As a result, the yield of
reaction decreased. Therefore, the optimum content of
loaded Fe was 6%. In this work, the conclusion was
similar to that of previous reports.
3.2.2 | Effect of different catalysts
Table 3 presents the catalytic activity of the various cata-
lysts for benzene hydroxylation at 60°C. The content of
catalysts was set at 150 mg. Jourshabani et al.[39] explored
an Fe‐supported SBA‐16 type catalyst used in the direct
synthesis of phenol from benzene. They discussed the
impact of molar ratio of catalyst to benzene on the yield
of phenol in the range 0.011 to 0.027. And they found
the yield of phenol was increased when the molar ratio
of catalyst to benzene changed from 0.011 to 0.019. How-
ever, the yield of phenol was decreased with a further
TABLE 3 Effect of catalyst on yield
Content
of
Selectivity (%)
loaded
Fe (%)
Yield
(%)
Entry
Phenol
Benzoquinone
1
2
3
4
5
6
0
3
31.85
24.93
29.98
43.27
25.39
22.00
68.15
75.07
70.02
56.73
74.61
78.00
2.83
4.32
4
6.86
6
37.56
15.41
9.18
9
12
SCHEME 1 Benzene hydroxylation reaction over Fe@Cu‐BDC
and pure Cu‐BDC catalysts
Conditions: 150 mg of catalyst; benzene, 1 ml; H2O2, 3 ml; acetonitrile, 4 ml;
reaction time, 75 min; 60°C.

8 of 10
LIU ET AL.
ascribed to the fact that with the volume ratio continuing
to increase, a small quantity of benzene could easily pro-
duce phenol totally under the condition of sufficient
hydrogen peroxide. And after benzene had converted into
phenol entirely, phenol went on to react to produce ben-
zoquinone with a large amount of hydrogen peroxide
existing in the solution. As a consequence, the selectivity
of phenol began to decrease. Thus, the optimum volume
ratio of benzene and acetonitrile was 1:7. From what
has been discussed above, we can briefly arrive at a con-
clusion that the volume ratio of benzene and acetonitrile
had an optimum value which had a crucial influence on
the selectivity of phenol.
3.2.3 | Effect of solvent dosage
In order to solve the problem of deep oxidation,
researchers have investigated the influence of solvent,
catalyst and reaction process. They found that an appro-
priate solvent was conducive for improving the selectivity
of phenol by letting the reactant and the product dissolve
in different phases, respectively. At the same time,
the reactant benzene is hydrophobic while the oxidant
hydrogen peroxide is hydrophilic. So how to let the two
insoluble phases approach more easily is the key to the
reaction. Through comparison, researchers found
that among solvents that included water, acetonitrile,
propanonitrile, isobutanol, acetone, dioxane, DMF,
octane and so on, acetonitrile had the optimum promot-
ing effect.[40] So we decided to use acetonitrile as the
solvent, which can play a role that is equivalent to a kind
of phase transfer reagent. Also, we found that the volume
ratio of benzene and acetonitrile had a vital impact on the
selectivity of phenol during the process.
Table 4 presents the catalytic activity of the Fe@Cu‐
BDC‐3 catalyst for benzene hydroxylation at 60°C. The
content of catalyst was fixed at 150 mg. As is evident from
the table, the selectivity of phenol increased gradually
with increasing ratio of benzene and acetonitrile. This
may be attributed to that, at the beginning, the concentra-
tion of benzene became relatively low with increasing
volume ratio. Therefore, when benzene produced phenol
after the first step, it did not have enough power of
pushing phenol to continue reacting to produce benzo-
quinone, which caused the selectivity of phenol to rise.
And the selectivity of phenol reached a peak when the
volume ratio was 1:7. After that, the selectivity of phenol
declined gradually with increasing ratio. This may be
3.2.4 | Effect of recovery and recycle times
Generally, the separation and recycling of as‐synthesized
catalysts is an important factor in industry. In other
words, the lifespan and the contents of catalysts would
directly affect the yield of reaction. During the reaction
procedure, the solid catalysts could be dispersed in the
liquid system due to the high speed of stirring. Then after
the reaction, the solid catalysts could be separated by cen-
trifugation. Subsequently, fresh methanol was used to
wash the spent solid catalysts, and the recovered catalysts
could be used for the next run.
Figure 8 presents the effect of recycling time after
Fe@Cu‐BDC‐3 catalyst was separated by centrifugation
from the liquid reaction system, washed and reactivated
in vacuum. As can be seen, the yield did not show an
obvious decline after recycling for five times after the
second run, which indicated this catalyst could be
TABLE 4 Effect of volume ratio of benzene and acetonitrile on
phenol selectivity using Fe@Cu‐BDC‐3 catalyst
Volume
ratio of
Selectivity (%)
benzene
and
acetonitrile
Yield
(%)
Entry
Phenol
Benzoquinone
1
2
3
4
5
6
7
1:4
1:5
1:6
1:7
1:8
1:9
1:10
37.38
40.32
49.43
62.00
44.33
32.77
26.80
62.62
59.68
50.57
38.00
55.67
67.23
73.20
17.30
6.27
2.60
9.14
5.09
1.86
0.96
FIGURE 8 Yield and selectivity with different recycling times
over Fe@Cu‐BDC‐3 catalyst. Conditions: 150 mg of catalyst;
benzene, 1 ml; H2O2, 3 ml; acetonitrile, 4 ml; reaction time, 75 min;
60°C
Conditions: 150 mg of catalyst; benzene, 1 ml; H2O2, 3 ml; reaction time,
60 min; 60°C.

LIU ET AL.
9 of 10
recognized as a real heterogeneous catalyst. What is
more, it is notable that although the yield decreased by
nearly half after the first run, the yield for the further
recycling times was still high compared with other
catalysts under the same condition, which can be seen
in Table 3. Simultaneously, the selectivity of phenol
exhibited different levels of decline from the second
run and reached the minimum value in the last run.
Correspondingly, the selectivity of benzoquinone was
the highest. Owing to the heterogeneity of the catalyst
and its easy separation by centrifugation from liquid,
the catalyst showed no significant loss during the reac-
tion. This meant the catalytic activity and selectivity of
the reaction could almost be maintained during the fifth
run. It can also be seen from Figure 8 that the as‐
synthesized catalyst could be recycled without a signifi-
cant loss after each run except the second run. This
revealed that the catalyst could be stable in the reaction
system and could be reused.
the process of benzene hydroxylation, when phenol
reached a certain concentration in the reaction system,
benzene hydroxylation would be inhibited. Because phe-
nol is more active than benzene, once the concentration
of phenol and hydroxyl radical are relatively high in the
reaction system, deep oxidation would be intensified. As
a result, phenol continued to react to produce benzoqui-
none. To sum up, steps (4)–(6) make up the catalytic cycle
process of side reaction.[41]
4 | CONCLUSIONS
Novel Fe@Cu‐BDC was prepared by adsorbing Fe ions on
the surface and into the channels of Cu‐BDC support
through an in situ method under ultrasonication. Then
a series of Fe@Cu‐BDC catalysts were synthesized by
varying the contents of Fe. Compared with the XRD
results of reported literature, the structure of Cu‐BDC in
the as‐synthesized catalysts could keep stable during the
synthesis procedure. Spherical and cubic particles were
formed in the process of preparing Cu‐BDC and
Fe@Cu‐BDC‐1. However, as‐synthesized catalysts did
not show an explicit morphology with the content of Fe
continuing to increase, which can be mainly due to the
agglomeration of catalyst particles. In addition, XPS anal-
ysis confirmed that the valence states of Fe species
changed from metallic to ferric iron after the reaction.
Moreover, the as‐synthesized catalysts exhibited an
excellent thermal stability but the temperature at which
samples reached the minimum mass point gradually
decreased with increasing content of loaded Fe. Further-
more, benzene hydroxylation was used to evaluate the
catalytic activity of the as‐synthesized catalysts. The
results showed that the yield of reaction could reach
37.56%, and the selectivity of phenol was about 62% for
a reaction time of 75 min at 60°C. Also, the as‐
synthesized catalysts exhibited excellent recycling rate
and reusability. The recovery experiment proved that
the yield of reaction could be kept at least 15.47% after
the fifth run. This novel Fe@Cu‐BDC catalyst may offer
a new idea for modification of MOFs and benzene
hydroxylation in the future.
3.3 | Catalytic mechanism
The whole catalytic process is presented in Scheme 2.
First, hydrogen peroxide oxidized ferrous iron to ferric
iron, and meanwhile hydrogen peroxide produced
highly active hydroxyl radical. Then the hydroxyl radical
added to benzene ring to produce cyclohexadiene
intermediate. Finally, ferric iron oxidized it to produce
phenol and, at the same time, ferric iron was reduced to
ferrous iron and continued the catalytic process. All in
all, steps (1)–(3) make up the catalytic cycle process of
the main reaction.
However, it is noticeable that cyclohexadiene inter-
mediate may be subject to other reaction routes. With
ACKNOWLEDGMENTS
We are grateful for financial support of the National
Natural Science Foundation of China (no. 41373114)
and we also acknowledge the Program of Introducing
Talents of Discipline to Universities (no. B06006).
SCHEME 2 Catalytic mechanism

10 of 10
LIU ET AL.
[20] S. Gao, N. Zhao, M. Shu, S. Che, Appl. Catal. A 2010, 388, 196.
ORCID
[21] T. Sugimoto, M. Nishiya1, M. Kobayashi, Colloid Polym. Sci.
2017, 295, 2405.
Boliang Liu https://orcid.org/0000-0001-7513-559X
Wei Cai https://orcid.org/0000-0002-3573-3496
[22] D. F. Sava Gallis, M. V. Parkes, J. A. Greathouse, X. Zhang, T.
M. Nenoff, Chem. Mater. 2015, 27, 2018.
REFERENCES
[23] T. T. Nguyen, N. T. S. Phan, Catal. Lett. 2014, 144, 1877.
[24] L. T. L. Ngyen, T. T. Nguyen, K. D. Nguen, N. T. S. Phan, Appl.
Catal. A 2012, 425, 44.
[1] Y.‐J. Lyu, T. Qi, H.‐Q. Yang, C.‐W. Hu, Catal. Sci. Technol.
2018, 8, 176.
[25] D. Saha, S. Deng, J. Colloid Interface Sci. 2010, 348, 615.
[2] S. Goto, Y.‐J. Seo, Y. Mukai, T. Tagawa, J. Mol. Catal. A 1997,
120, 149.
[26] O. M. Yaghi, H. Li, T. L. Groy, J. Am. Chem. Soc. 1996, 118,
9096.
[3] M. Tani, T. Sakamato, S. Mita, S. Sakaguchi, Y. Ishii, Angew.
Chem. Int. Ed. 2005, 44, 2586.
[27] L. Xu, E. Y. Choi, Y. U. Kwon, Inorg. Chem. Commun. 2008, 11,
1190.
[4] V. N. Parnon, G. I. Panov, A. Uriarte, A. S. Noskov, Catal.
Today 2005, 100, 115.
[28] E. Biemmi, T. Bein, N. Stock, Solid State Sci. 2006, 8, 363.
[5] R. R. Fernandes, M. V. Kirillova, J. A. L. da Silva, J. da Silva, A.
[29] H. den Daas, M. Passacandanto, L. Lozzi, S. Santucci, P.
J. L. Pombeiro, Appl. Catal. A 2009, 353, 107.
Picozzi, Surf. Sci. 1994, 317, 295.
[6] H. Abbo, S. Titinchi, Appl. Catal. A 2009, 356, 167.
[30] L. Lozzia, M. Passacantandoa, P. Picozzia, S. Santuccia, H. den
Daas, Surf. Sci. 1995, 331, 703.
[7] D. Bianchi, L. Balducci, R. Bortolo, R. D'Aloisio, M. Ricci, G.
Spano, R. Tassinari, C. Tonini, R. Ungarelli, Adv. Synth. Catal.
2007, 349, 979.
[31] H. J. Mathieu, D. Landolt, Corros. Sci. 1986, 26, 547.
[32] B. J. Tan, K. J. Klabunde, P. M. A. Sherwood, Chem. Mater.
1990, 2, 186.
[8] B. Liptá ková, M. Báhidsky, M. Hronec, Appl. Catal. A 2004,
263, 33.
[33] S. S. ‐Y. Chui, S. M. ‐F. Lo, J. P. H. Charmant, A. G. Orpen, I. D.
Williams, Science 1999, 283, 1148.
[9] S. Niwa, M. Eswaramoorthy, J. Nair, A. Raj, N. Itoh, H. Shoji,
T. Namba, F. Mizukami, Science 2002, 295, 105.
[34] S. Shahid, K. Nijmeijer, J. Membr. Sci. 2014, 459, 33.
[35] S. Bhattacharjee, Indian J. Chem. 2018, 57A, 778.
[36] Y. Luo, J. Xiong, C. Pang, G. Li, C. Hu, Catalysts 2018, 8, 49.
[10] D. Bianchi, R. Bortolo, R. Tassinari, M. Ricci, R. Vignola,
Angew. Chem. Int. Ed. 2000, 39, 4321.
[11] B. Guo, L. Zhu, X. Hu, Q. Zhang, D. Tong, G. Li, C. Hu, Catal.
Sci. Technol. 2011, 1, 1060.
[37] X. Wang, T. Zhang, B. Li, Q. Yang, S. Jiang, Appl. Organometal.
Chem. 2014, 28, 666.
[12] M. C. Esmelindro, E. G. Oestreicher, M. Caovilla, J. A. Lessa,
C. Fernandes, C. Dariva, S. M. Egues, A. J. Bortoluzzi, O. A.
C. Antunes, J. Braz. Chem. Soc. 2006, 17, 1551.
[38] G. Capocasa, G. Olivo, A. Barbieri, O. Lanzalunga, S. Di
Stefano, Catal. Sci. Technol. 2017, 7, 5677.
[39] M. Jourshabani, A. Badiei, Z. Shariatinia, N. Lashgari, G. M.
[13] A. Corma, H. Garcia, F. X. L. Xamena, Chem. Rev. 2010, 110, 4606.
[14] L. Ma, C. Abney, W. Lin, Chem. Soc. Rev. 2009, 38, 1248.
Ziarani, Ind. Eng. Chem. Res. 2016, 55, 3900.
[40] C. Walling, R. A. Johnson, J. Am. Chem. Soc. 1975, 97, 363.
[41] C. Walling, Acc. Chem. Res. 1975, 8, 125.
[15] S. Neogi, M. K. Sharma, P. K. Bharadwaj, J. Mol. Catal. A 2009,
299, 1.
[16] M. Ranocchiari, J. A. van Bokhoven, Phys. Chem. Chem. Phys.
2011, 13, 6388.
How to cite this article: Liu B, Wu S, Liu Y,
et al. Preparation of Fe@Cu‐BDC and its catalytic
performance for benzene hydroxylation. Appl
Organometal Chem. 2018;e4662. https://doi.org/
10.1002/aoc.4662
[17] F. Farzaneh, S. Sadat Mortazavi, React. Kinet. Mech. Catal.
2017, 120, 333.
[18] Y. M. Chung, H. Y. Kim, W. S. Ahn, Catal. Lett. 2014, 144, 817.
[19] D. Esken, X. Zhang, O. I. Lebedev, F. Schroder, R. A. Fischer,
J. Mater. Chem. 2009, 19, 1314.