Magnetically separable and reusable rGO/Fe3O4 nanocomposites for the selective liquid phase oxidation of cyclohexene to 1,2-cyclohexane diol

Article abstract of DOI:10.1039/c9ra04685b

A series of magnetically separable rGO/Fe₃O₄ nanocomposites with various amounts of graphene oxide were successfully prepared by a simple ultrasonication assisted precipitation combined with a solvothermal method and their catalytic activity was evaluated for the selective liquid phase oxidation of cyclohexene using hydrogen peroxide as a green oxidant. The prepared materials were characterized using XRD, FTIR, FESEM, TEM, HRTEM, BET/BJH, XPS and VSM analysis. The presence of well crystallized Fe₃O₄ as the active iron species was seen in the crystal studies of the nanocomposites. The electron microscopy analysis indicated the fine surface dispersion of spherical Fe₃O₄ nanoparticles on the thin surface layers of partially-reduced graphene oxide (rGO) nanosheets. The decoration of Fe₃O₄ nanospheres on thin rGO layers was clearly observable in all of the nanocomposites. The XPS analysis was performed to evaluate the chemical states of the elements present in the samples. The surface area of the nanocomposites was increased significantly by increasing the amount of GO and the pore structures were effectively tuned by the amount of rGO in the nanocomposites. The magnetic saturation values of the nanocomposites were found to be sufficient for their efficient magnetic separation. The catalytic activity results show that the cyclohexene conversion reached 75.3% with a highest 1,2-cyclohexane diol selectivity of 81% over 5% rGO incorporated nanocomposite using H₂O₂ as the oxidant and acetonitrile as the solvent at 70 °C for 6 h. The reaction conditions were further optimized by changing the variables and a possible reaction mechanism was proposed. The enhanced catalytic activity of the nanocomposites for cyclohexene oxidation could be attributed to the fast accomplishment of the Fe²⁺/Fe³⁺ redox cycle in the composites due the sacrificial role of rGO and its synergistic effect with Fe₃O₄, originating from the conjugated network of π-electrons in its surface structure. The rapid and easy separation of the magnetic nanocomposites from the reaction mixture using an external magnet makes the present catalysts highly efficient for the reaction. Moreover, the catalyst retained its activity for five repeated runs without any drastic drop in the reactant conversion and product selectivity.

Full text of DOI:10.1039/c9ra04685b

RSC Advances
View Article Online
View Journal | View Issue
PAPER
Magnetically separable and reusable rGO/Fe3O4
nanocomposites for the selective liquid phase
oxidation of cyclohexene to 1,2-cyclohexane diol†
Cite this: RSC Adv., 2019, 9, 32517
*
*
Qingming Jia, Yanan Dong, Zhongxiao Yue and Shaoyun Shan
Manoj Pudukudy,
A series of magnetically separable rGO/Fe3O4 nanocomposites with various amounts of graphene oxide
were successfully prepared by simple ultrasonication assisted precipitation combined with
a
a solvothermal method and their catalytic activity was evaluated for the selective liquid phase oxidation
of cyclohexene using hydrogen peroxide as a green oxidant. The prepared materials were characterized
using XRD, FTIR, FESEM, TEM, HRTEM, BET/BJH, XPS and VSM analysis. The presence of well crystallized
Fe3O4 as the active iron species was seen in the crystal studies of the nanocomposites. The electron
microscopy analysis indicated the fine surface dispersion of spherical Fe3O4 nanoparticles on the thin
surface layers of partially-reduced graphene oxide (rGO) nanosheets. The decoration of Fe3O4
nanospheres on thin rGO layers was clearly observable in all of the nanocomposites. The XPS analysis
was performed to evaluate the chemical states of the elements present in the samples. The surface area
of the nanocomposites was increased significantly by increasing the amount of GO and the pore
structures were effectively tuned by the amount of rGO in the nanocomposites. The magnetic saturation
values of the nanocomposites were found to be sufficient for their efficient magnetic separation. The
catalytic activity results show that the cyclohexene conversion reached 75.3% with a highest 1,2-
cyclohexane diol selectivity of 81% over 5% rGO incorporated nanocomposite using H2O2 as the oxidant
and acetonitrile as the solvent at 70 ꢀC for 6 h. The reaction conditions were further optimized by
changing the variables and a possible reaction mechanism was proposed. The enhanced catalytic activity
of the nanocomposites for cyclohexene oxidation could be attributed to the fast accomplishment of the
Fe2+/Fe3+ redox cycle in the composites due the sacrificial role of rGO and its synergistic effect with
Fe3O4, originating from the conjugated network of p-electrons in its surface structure. The rapid and
easy separation of the magnetic nanocomposites from the reaction mixture using an external magnet
makes the present catalysts highly efficient for the reaction. Moreover, the catalyst retained its activity for
five repeated runs without any drastic drop in the reactant conversion and product selectivity.
Received 22nd June 2019
Accepted 28th September 2019
DOI: 10.1039/c9ra04685b
rsc.li/rsc-advances
position. However, if the oxidation occurs at the allylic position,
then it leads to the formation of 2-cyclohexenol (enol) and 2-
cyclohexenone (enone).3 All of these chemicals have promising
1. Introduction
In recent years, surface catalysis has become a hot area of
research in synthetic chemistry because of its signicance in
organic transformation to produce high value added chemicals
for various industries. Cyclohexene is one of the most widely
used raw materials in chemical industries because of its low
cost, availability and chemical structure.1 Since there are two
potential oxidation sites existing in its structure, a mixture of
different oxidation products was formed in its oxidation.2
Cyclohexene oxide (epoxide) and 1,2-cyclohexane diol (diol) are
formed if the oxidation occurs at the double bond (C]C)
use in various elds such as organic synthesis, materials science
and pharmaceuticals.4 Therefore, the selective oxidation of
cyclohexene to achieve any of these products is highly signi-
cant. Among the different oxidation products, diols have
a signicant role in the synthesis of pharmaceutical interme-
diates and it can be used as a direct precursor for the synthesis
of adipic acid.5 In addition to this, it has a signicant role in the
synthesis of liquid crystals, resins and catechol. In the synthesis
of epoxy resins, it is used as a diluter. Moreover, it acts as
a chiral auxiliary in the asymmetric synthesis.6 Therefore, the
selective synthesis of diol is highly important in pharmaceutical
and chemical industries.
Faculty of Chemical Engineering, Kunming University of Science and Technology,
Kunming, 650500, Yunnan, People's Republic of China. E-mail: manojpudukudy@
gmail.com; jiaqm411@163.com
The common methods for diol synthesis are direct hydro-
lysis of epoxide and the dihydroxylation of cyclohexene. Dihy-
droxylation is the most widely accepted method and it uses the
† Electronic supplementary information (ESI) available. See DOI:
10.1039/c9ra04685b
This journal is © The Royal Society of Chemistry 2019
RSC Adv., 2019, 9, 32517–32534 | 32517

View Article Online
RSC Advances
Paper
tetroxides of Os and Ru to catalyze the process, which acts as completed within 3–5 hours, when the reaction performed with
a catalyst and oxidant in this process.7 Because of the high cost, [3,5-(CF3)2C6H3Se]2. Even though these catalysts presented high
toxicity and non-selective nature the process, the use of these efficiency for the reaction, their homogeneous nature limits
oxidants is limited and it is not possible for an industrial their recyclability. Hence the need of heterogeneous catalysts is
application. Furthermore, the reaction is very difficult to control highly substantial in this reaction.
using these oxidants. Therefore, it is highly desirable to develop
an ecofriendly catalytic path for cyclohexene conversion to diol selective oxidation of cyclohexene to diol and the results shows
using greener oxidants and more efficient low-cost catalysts. that a highest cyclohexene conversion of 92.63% with 64.22%
Huang et al.32 reported the direct use of polyaniline for the
In the past few years, different types of homogeneous and diol selectivity was noticed for the emeraldine base polyaniline
heterogeneous catalysts, oxidants and solvents have been used with the size of 10–20 nm. The activity of the polyaniline is due
for cyclohexene conversion.8 However, the conversion and to its small size and redox states. Zhou et al.33 reported the
selectivity of the products in cyclohexene oxidation was found to solvent free oxidation of cyclohexene to diol using acidic poly-
be completely depended on the type of catalyst, co-catalyst, aniline@halloysite nanotubes. Approximately 98% conversion
oxidant and solvents used in the reaction.9 The commonly with 99.5% diol selectivity was achieved over PANI@HA/1 M/
used catalysts for the oxidation of cyclohexene consists of 2.04-HCl catalyst at the optimized reaction conditions of
metals-free or metals incorporated organic catalysts such as 0.02 g catalyst, 2.5 mL H2O2 at 70 ꢀC for 24 h of reaction. As per
organometallic compounds and metal–organic frameworks, them, the enhanced activity of the catalysts could be ascribed to
bare metal oxides, metal oxide hybrids, supported metal oxides the reversible redox reaction of PANI in the composite for H2O2
and so on.10–12 Homogeneous catalysts were reported to be more decomposition. Boudjema et al.34 reported the montmorillonite
effective for the liquid phase oxidation reactions and highly clay supported 11-molybdovanado-phosphoric acid catalyst for
suitable to investigate the reaction mechanism. However, due to the H2O2 assisted catalytic oxidation of cyclohexene to diol. The
the difficulties associated with their isolation from reaction cyclohexene conversion and diol yield were reported to be
mixture and high susceptibility to drastic reaction conditions, 81.5% and 77% respectively for a 3 h of reaction at 70 ꢀC. In
their recyclability is restricted. These disadvantages of the other work, Guzik et al.35 reported cyclohexene oxidation over
homogeneous catalysts could be overcome by the use of FDU-1-supported Nb(Co) catalyst, using H2O2 as oxidant in
heterogeneous catalysts.13 The metal-based catalysts are usually acetonitrile. The results indicated that a maximum cyclohexene
more advantageous than organocatalysts in the liquid phase conversion and diol selectivity of 61% and 56% was observed
oxidation because of their high stability, ease of separation and over the catalyst respectively.
recyclability.14,15 Different types of oxidants were also used to
It is reported that the oxides and complexes of transition
catalyze the oxidation of cyclohexene.16 The main oxidants used metals especially Fe, Cu, V, Os, Mn, Ru, etc. were effectively used
are alkyl hydroperoxides, hydrogen peroxide, molecular oxygen for dihydroxylation reaction.7,38 Since iron is the cheap and
and air. Among these, air, oxygen and hydrogen peroxide are second most abundant metal, more efforts were given to the
considered as the green oxidants as their by-products are not development of iron-based catalysts.39 Magnetic Fe3O4 is the
harmful to environment.17 Moreover, among various alkyl main type of heterogeneous catalyst in this category due to its
hydroperoxides, tert-butyl hydroperoxide (TBHP) is the mostly magnetic recovery from the reaction mixture for recyclability.40
used oxidant for cyclohexene oxidation. It usually yields cyclo- However, due to the sintering and aggregation of particles, their
hexene oxide and allylic oxidation products based on the type of catalytic activity was limited for an efficient reaction. This can
catalysts.18,19 The use of molecular oxygen as an oxidant for the be avoided by anchoring the metal oxides on a support material.
reaction to produce allylic oxidation products was also reported The anchoring of metal oxides on a support has several
over a set of different metal-based catalysts.20–22 The H2O2 advantages like the improvement of their catalytic activity and
assisted cyclohexene oxidation to 1,2-diol in a catalytic route has stability together with ease of recovery and recyclability. Carbon
been also reported over a set of homogeneous and heteroge- nanomaterials with two dimensional sheets like structures such
neous catalysts.23–37
as graphene, graphene oxide and reduced graphene oxide are of
For example, Usui et al.26 reported a maximum diol yield of great importance in materials chemistry due to their ability to
98% in a solvent- and metal-free cyclohexene oxidation using disperse the metal oxides on their surface.41,42 Thus, the devel-
H2O2, catalyzed by immobilized sulfonic acids. In another work, opment of highly efficient and low-cost catalysts for cyclohexene
Rosatella et al.27 reported the application of Brønsted acids in conversion with high diol selectivity is worthy.
the catalytic conversion of cyclohexene to diol. Among the
In this work, a series of magnetically separable and reusable
various catalysts used the para-toluene sulfonic acid was found rGO/Fe3O4 nanocomposites were successfully prepared by an
to be showing the highest diol yield of 98%. Theodorou et al.28 ultrasonication assisted precipitation combining solvothermal
used 2,2,2-triuoroacetophenone as an organocatalyst for the method and effectively used for the selective oxidation of cyclo-
oxidation of cyclohexene to diol using H2O2 and nearly 96% hexene to 1,2-cyclohexane diol using H2O2 as a green oxidant in
yield was observed for the product. Yu et al.29,30 reported the use liquid phase for the rst time. It is reported that compared with
of organoselenium compounds for the selective oxidation of co-precipitation method, the solvothermal route provides a more
cyclohexene to diol using H2O2 as the oxidant. Approximately controlled growth and size of Fe3O4 nanoparticles for an efficient
96% yield was observed for 42 hours of reaction using (PhSe)2 as dispersion on the surface of graphene oxide.43 Additionally, it
catalyst at room temperature, whereas the reaction was reduces the surface groups of graphene oxide and reduced
32518 | RSC Adv., 2019, 9, 32517–32534
This journal is © The Royal Society of Chemistry 2019

View Article Online
Paper
RSC Advances
graphene oxide (rGO) could be formed. The amount of GO was a solution of FeCl2$4H2O and FeCl3$6H2O in 1 : 2 mol ratio was
varied from 1 wt% to 10 wt% in the nanocomposites. To the best prepared in 200 ml of deionized water and it was mixed with the
of our knowledge, the use of rGO/Fe3O4 nanocomposites for the prepared exfoliated GO suspension and sonicated for 30
liquid phase oxidations is very rarely reported as many of the minutes. To precipitate the divalent and trivalent iron ions in
works were focused on the degradation of dyes by a Fenton-type the solution, aqueous ammonia was added slowly to the
reaction. Previously Balasubramanyan et al.39 reported the use of suspension by dropwise until the pH of the solution turned to
a series of GO/Fe3O4 nanocomposites for the liquid phase 11. The blackish suspension was then directly transferred into
oxidation of cyclohexene. Herein, a complete physico-chemical a 500 ml Teon lined stainless steel autoclave, sealed and kept
and magnetic characterization of the as-prepared rGO/Fe3O4 in an air driven oven at 120 ꢀC for 8 h. Aer solvothermal
nanocomposites were studied and discussed in detail. The effects treatment, the autoclave was allowed to cool naturally and the
of various reaction parameters on cyclohexene conversion and precipitates were collected by manual ltration, washed several
product selectivity were investigated and correlated with catalytic times with distilled water followed by ethanol and dried in
ꢀ
properties. A plausible reaction mechanism for the improved a vacuum oven at 80 C for 6 h to obtain the rGO/Fe3O4 nano-
activity/selectivity for oxidation is also reported.
composites. For comparison, bare Fe3O4 was also prepared by
the same method, without the addition of GO. The dispersed
GO suspension was also treated in similar solvothermal
condition for studying its catalytic properties and activities. All
of the samples were stored in air tight plastic containers for
catalytic studies. The samples were labelled as x% rGO/Fe3O4,
where x stands for the nominated amount of GO (wt%) in the
nanocomposites.
2. Experimental
2.1 Materials
All of the chemicals used in the present study were
purchased from Aladdin, China. These chemicals were of
analytical grade and were used as received without further
purication.
2.4 Materials characterization
2.2 Synthesis of graphite oxide
The as-prepared nanocomposites were characterized by using
powder X-ray diffraction (XRD), Fourier transform infrared
spectroscopy (FTIR), eld emission scanning electron
microscopy (FESEM), transmission electron microscopy
(TEM), high resolution transmission electron microscopy
(HRTEM), nitrogen physisorption, X-ray photoelectron spec-
troscopy (XPS) and vibrating sample magnetometer (VSM)
analyses. The XRD analysis was performed in a Bruker D8
Focus X-ray Diffractometer with Cu Ka radiation at a wave-
length of 0.1540 nm in the scanning range of 5 to 80ꢀ with
a scan rate of 0.025ꢀ. The Scherrer's equation was used to
calculate the crystallite size of the iron oxide from the full
width at half maximum value of the intense diffraction peak,
at the 2q value of 35.6ꢀ. A Bruker VERTEX 70 FTIR instrument
was used to record the FTIR spectra of the samples using KBr
pellet method. The spectra were recorded in the region of 400–
4000 cmꢁ1 in transmission mode, aer a background scan
with pelletized KBr. The FESEM images of the prepared
samples for the determination of their surface morphology
were performed in a ZEISS, MERLIN system at an accelerating
voltage of 3 kV. The samples were placed on a carbon tape,
coated with iridium and photographed. A TECNAI G2 TF30 S-
Twin TEM instrument was utilized for internal morphology
A previously reported modied Hummer's method was used
to synthesize the graphite oxide.44 In a typical procedure, 1 g
of graphite powder and 0.5 g of sodium nitrate were added
into 25 ml of conc. sulfuric acid taken in a 250 ml beaker. The
mixture was kept in an ice bath to keep the temperature
below 10 ꢀC with vigorous magnetic stirring. Next 3 g of
potassium permanganate was gradually added to the
suspension. The temperature of the mixture was maintained
below 15 ꢀC during the addition of KMnO4 for a period of 3 h.
Aer the complete addition of KMnO4, the ice bath was
removed and the mixture was kept in a water bath at 35 ꢀC
and again stirred for 2 hours. Then 150 ml of deionized water
was added to the mixture followed by the addition of 20 ml of
30% H2O2. This is to ensure the removal of residual
permanganate and manganese dioxide in the samples. The
suspension was then centrifuged to get a paste of graphite
oxide. It was then washed four times with 10% HCl and ve
times with deionized water to remove any impurities in it.
The as-syntꢀhesized graphite oxide was then dried in a vacuum
oven at 60 C for 24 h.
2.3 Synthesis of rGO/Fe3O4 nanocomposites
An ultra-sonication assisted precipitation combining sol- characterization of the nanocomposites, which was operated
vothermal method was used to prepare the rGO/Fe3O4 nano- at an accelerating voltage of 300 kV. The fast Fourier transform
composites. The amount of graphite oxide was varied during (FFT) electron diffraction patterns were also obtained together
synthesis, according to the nominated amount of graphene with the TEM micrographs. A small amount of the sample was
oxide (GO) in the nanocomposites. In the present study, dispersed in absolute ethanol by ultrasonication and a drop of
different compositions of GO were used, namely, 1 wt%, the solution was placed on a carbon coated copper grid, dried
2.5 wt%, 5 wt%, 7.5 wt% and 10 wt%. Firstly, the required under dim IR light and directly used for TEM imaging.
amount of graphite oxide was dispersed in 200 ml of ethanol–
The nitrogen physisorption characterization was carried out
ꢀ
water mixture (100 ml each) in an ultrasonic bath for 4 h to in a Micromeritics ASAP 2020 system at ꢁ196 C. The samples
obtain the suspension of exfoliated graphene oxide. Next, were degassed in vacuum at 100 ꢀC for 2 h before analysis. The
This journal is © The Royal Society of Chemistry 2019
RSC Adv., 2019, 9, 32517–32534 | 32519

View Article Online
RSC Advances
Paper
specic surface area and pore parameters of the samples were
measured by Brunauer–Emmett–Teller (BET) and Barrett–Joy-
ner–Halenda (BJH) methods respectively. The desorption
branch of nitrogen isotherms was used for the calculations. The
XPS characterization of the representative samples was studied
in an ESCALAB 250Xi, Thermo Fisher Scientic XPS system. The
advantage 5.967 surface chemical analyzer soware was used to
process and analyze the results. The binding energies of the
elements were internally calibrated using the binding energy of
C 1s (284.6 eV) as a reference. The narrow scan spectra were
curve tted using Gaussian–Lorentz function. The magnetic
measurements were conducted on a MPMS3 SQUID vibrating
sample magnetometer (Quantum Design) at room temperature.
The electron spin resonance (ESR) spectroscopic analysis was
carried out in a JEOL-JES FA200 ESR spectroscope, using 5,5-
dimethyl-1-pyrroline-N-oxide (DMPO) as a radical trapping
agent.
Fig. 1 XRD patterns of the prepared samples.
2.5 Liquid phase oxidation of cyclohexene
The catalytic performance of the prepared samples was inves-
tigated for the liquid phase oxidation of cyclohexene. The
reactions were carried out in a two necked round bottom ask
connected to a water condenser. In a typical procedure, the
weighed amount of catalyst (0.05 g) was dispersed in a mixture
of 2 mmol cyclohexene dissolved in 5 ml of acetonitrile solvent.
Next the RB ask was placed in an oil bath at 70 ꢀC and 10 mmol
of H2O2 was slowly added to the reaction mixture as an oxidant
(30% H2O2 in water, dropwise addition). The reaction was per-
formed for 5 h. A constant magnetic stirring was provided in the
whole period of reaction, for the proper mixing of the reactants,
solvent and catalyst. Aer reaction, the catalysts were separated
from the reaction mixture using an external magnet, and the
reaction products were treated with anhydrous sodium sulphate
to remove any residual water and were analyzed by a gas chro-
matograph (Agilent) connected to a ame ionization detector.
An Elite-1 series capillary column was used for the separation
and identication of the reaction products. The reactions were
triplicated to conrm the results. The conversion of cyclohexene
and selectivity of the formed products were calculated using the
following equations, from the calibrated results of the standard
samples. Cyclohexene conversion (%) ¼ (([CH]0 ꢁ [CH]t)/[CH]0)
ꢂ 100, where [CH]0 and [CH]t represents the initial and nal
concentration of cyclohexene in moles respectively. Products
selectivity (%) ¼ (moles of individual product/total moles of
products) ꢂ 100. The reaction parameters were further varied to
study their effect on the cyclohexene conversion and products
selectivity.
method. The interplanar distance (d) was measured to be
0.83 nm, which is in good agreement with the lamellar crys-
talline structure of graphite oxide.45 The other diffraction peak
at the 2q value of 20.3ꢀ might be initiated from the sample
holder used for the XRD analysis. Fig. 1 shows the XRD patterns
of the iron oxide and rGO/Fe3O4 nanocomposites. The diffrac-
tion peaks located at the 2q values of 18.4ꢀ, 30.2ꢀ, 35.6ꢀ, 43.3ꢀ,
53.6ꢀ, 57.3ꢀ, 62.9ꢀ and 74.3ꢀ were directly indexed to the
formation of cubic phase spinel crystalline structure of
˚
magnetite, Fe3O4, with the lattice constant of a ¼ 8.375 A
(JCPDS card number: 00-065-0731). All of peaks of Fe3O4 were
found to be completely retained in the prepared rGO/Fe3O4
nanocomposites. It indicates the successful anchoring of Fe3O4
on reduced graphene oxide matrix. However, no peaks related to
graphene oxide or rGO were observed in the prepared nano-
composites. This could be due to the excellent dispersion of
highly exfoliated reduced graphene oxide sheets or due to the
low amount of GO in the nanocomposites. No other diffraction
peaks were observed, which further points to the absence of
impurities in the nanocomposites. The high intensity of the
diffraction peaks in the nanocomposites suggests that the Fe3O4
particles are well crystallized. However, the incorporation of
rGO contributed a small reduction in the intensity and
a broadening of the diffraction peaks was detected in the
composites. This could be attributed to the size-controlled
growth of Fe3O4 nanoparticles on reduced graphene oxide.
The incremental incorporation of graphene oxide from 1 wt% to
10 wt% does not altered the crystallinity of Fe3O4 in nano-
composites. The crystallite size of Fe3O4 calculated using
Scherrer's equation was completely complemented with this
observation. The average crystallite size of Fe3O4 was calculated
to be 11 nm and 7 ꢃ 2 nm in the bare and rGO incorporated
Fe3O4 nanocomposites respectively. It is previously reported
that the graphene oxide is a good support for nanoparticles as it
provides necessary surface features through its functional
groups.46 Therefore, the growth of particles could be effectively
3. Results and discussion
3.1 Characterization of the prepared materials
The crystalline properties of the prepared samples were studied
using XRD and the diffraction patterns are shown in Fig. 1. A
sharp diffraction peak was observed at the 2q value of 10.5ꢀ in
Fig. S1.† This peak could be assigned to the (001) plane of the
graphite oxide that has been synthesized by modied hummer's
32520 | RSC Adv., 2019, 9, 32517–32534
This journal is © The Royal Society of Chemistry 2019

View Article Online
Paper
RSC Advances
controlled without any severe particle agglomeration. A similar of graphene oxide was resulted in the nanocomposites. It could
view is observed here in the case of the prepared be ascribed to the applied conditions of the solvothermal
nanocomposites.
treatment.
The morphological features of the prepared samples were
The surface chemistry and bonding characteristics of the
prepared materials were studied using FTIR spectroscopy and studied using FESEM analysis and the images are shown in
the spectra are shown in Fig. 2. The broad and intense trans- Fig. 3. The sheet like morphology of the graphite oxide is clearly
mission band found in the region of 495–775 cmꢁ1 and a less shown in Fig. 3(a), whether the surface layers of graphene oxide
intense transmission band at 445 cmꢁ1 were related to the are clearly visible. A number of wrinkles and ripples were also
stretching vibration of Fe–O bond.47 These two transmission seen on its puffy surface. A cluster of more or less spherical
bands conrm the presence of iron oxide, specically Fe3O4 in shaped Fe3O4 particles with a moderate degree of particle
the samples. This is highly consistent with the XRD results. The aggregation is shown in Fig. 3(b). The size of the Fe3O4 particles
other transmission bands centered at 1619 cmꢁ1 and 3423 cmꢁ1 was found to be uniform and it was measured to be in the nano
were attributed to the bending and stretching vibrations of the range of 10–12 nm. The introduction of graphene oxide allowed
O–H group, originated from the adsorbed water molecules. the homogeneous dispersion of Fe3O4 nanoparticles on the
These bands were found to be retained in the nanocomposites surface of reduced graphene oxide in the nanocomposites.
and their intensity increased with the increasing amount of
As shown in Fig. 3(c), the Fe3O4 nanoparticles were uniformly
graphene oxide in the composites. This could be attributed to distributed over reduced graphene oxide sheets in the 1% rGO/
the presence of more surface hydroxyl groups in the compos- Fe3O4 nanocomposite. However, due to its low amount, the
ites.48 Some other weakly resolved and less intense transmission individual layered structure of reduced graphene oxide was not
bands were also identied in the nanocomposites with seen. It is expected that the rGO surfaces were completely covered
increasing amount of GO. It could be due to presence of oxygen with Fe3O4 nanoparticles. When the amount of graphene oxide
containing functional groups in the composites arising from increased to 2.5 wt% and more, the layers of reduced graphene
partially-reduced graphene oxide. The transmission band oxide sheets were clearly visible in the composites as shown in
located at 1043 cmꢁ1 could corresponds to the stretching Fig. 3(d)–(f). The Fe3O4 nanoparticles were found to be nely
vibration of C–O originated from the presence of oxygen on the dispersed and anchored on the surface layers of reduced gra-
surface of graphene oxide formed by the oxidation of graphite.49 phene oxide. The rate of growth and aggregation of particles is
The transmission band observed at 1226 cmꢁ1 was related to found to be fully controlled by the incorporation of graphene
the symmetric stretching vibration of the epoxy and C–OH oxide in the nanocomposites. The surface edges of reduced gra-
groups in the nanocomposites. The transmission bands located phene oxide layers were also found to be fully decorated with
at 1627 cmꢁ1 and 1735 cmꢁ1 denote the OH bending and C]O spherical shaped iron oxide nanoparticles as seen in the SEM
stretching vibration of the COOH group.50 The aromatic C]C images of 2.5% to 7.5% rGO combined composites. The presence
bond and carbonyl bond also showing the same transmission of oxygenated surface functional groups in graphene oxide such
bands in these region.51 The low intensity of the functional as hydroxyl, epoxy and carboxyl groups could enable the homo-
groups fully validates the reduction of surface groups of GO geneous anchoring and decoration of iron oxide nanoparticles on
aer solvothermal treatment. However, only a partial reduction the layered surfaces of the reduced graphene oxide with a tight
bond between the individual components of the nano-
composites.52 Moreover, the two-dimensional layered nano-
structure of reduced graphene oxide was not affected by the
loading of Fe3O4 nanoparticles and its structure was preserved in
all of the nanocomposites. The ultra-dispersion of Fe3O4 nano-
particles on rGO surface restricts the particle growth and hence
large Fe3O4 particles and bigger aggregates of Fe3O4 clusters were
not seen in the prepared nanocomposites. The size of Fe3O4
nanoparticles was measured to be approximately 6–8 nm in all of
the nanocomposites, which is smaller than that observed in the
case of bare Fe3O4. Moreover, the aggregation of Fe3O4 nano-
particles was found to be decreased with increasing the amount
of graphene oxide in the nanocomposites. It could be due to the
availability of sufficient rGO surface for an effective dispersion of
Fe3O4 nanoparticles without any severe aggregation of particles.
The internal structure and degree of particle dispersion in the
prepared samples were evaluated by TEM analysis and the
resulting images are shown in Fig. 4. The transparent layered
surface structure of the rGO sheets and a cluster of aggregated
Fe3O4 nanoparticles prepared without graphene oxide are shown
in Fig. 4(a) and (b) respectively. The electron diffraction pattern
Fig. 2 FTIR spectra of the prepared samples.
shown in the inset of Fig. 4(b) indicates the polycrystalline nature
This journal is © The Royal Society of Chemistry 2019
RSC Adv., 2019, 9, 32517–32534 | 32521

View Article Online
RSC Advances
Paper
Fig. 3 FESEM images of the as-synthesized samples (a) GO, (b) Fe3O4, (c) 1% rGO/Fe3O4, (d) 2.5% rGO/Fe3O4, (e) 5% rGO/Fe3O4 and (f) 7.5% rGO/
Fe3O4 nanocomposites.
of the Fe3O4 nanoclusters. The presence of Fe3O4 nanoparticles displays clear lattice fringes of iron oxide. The lattice fringe
decorated on the surface of exfoliated reduced graphene oxide spacings were measured to be 0.295 nm and 0.25 nm, which
layers was displayed in the TEM image of 5% rGO/Fe3O4 nano- corresponded to the (220) and (311) reections of the Fe3O4
composite. The rGO layers and Fe3O4 nanoparticles were clearly respectively. Fig. 4(f) represents the HRTEM image and fast
marked in the TEM photograph of the representative nano- Fourier transform pattern of a single crystalline Fe3O4 particle
composite, as shown in Fig. 4(c). The size of more or less in the nanocomposite. The particle was found to have a size of
spherical shaped Fe3O4 nanoparticles distributed on rGO layers ꢄ8 nm and a lattice spacing of 0.253 nm, which is consistent
was measured to be varying from 5–8 nm, which is in good with its (311) plane.
agreement with the XRD and SEM observations.
The textural properties of the Fe3O4 and rGO/Fe3O4 nano-
The inset gure shown in the high magnied TEM image of composites were studied using BET/BJH analysis and the
the 5% rGO/Fe3O4 nanocomposite in Fig. 4(d) comprises well nitrogen physisorption isotherms are shown in Fig. 5. Accord-
developed concentric diffraction rings with sharp diffraction ing to IUPAC classication, the samples show isotherms that
spots. It indicates the polycrystalline nature of Fe3O4 in the belong to type IV, with a H1 hysteresis loop. It indicates the
composites. The high resolution TEM image showed in Fig. 4(e) presence of mesopores in the prepared samples.53
32522 | RSC Adv., 2019, 9, 32517–32534
This journal is © The Royal Society of Chemistry 2019

View Article Online
Paper
RSC Advances
Fig. 4 TEM images of the prepared samples (a) rGO, (b) Fe3O4 (c, d) 5% rGO/Fe3O4 nanocomposite and (e, f) high resolution TEM images of the
representative composite (FFT patterns were shown in the inset).
The hysteresis loops were observed to be widened by the nanocomposites. The specic surface area of Fe3O4 nano-
incorporation of a small amount of graphene oxide and the particles was measured to be 56.37 m2 gꢁ1 by the BET method.
trend was continued with its increasing amount in the However, the incorporation of 2.5% GO, increased the surface
This journal is © The Royal Society of Chemistry 2019
RSC Adv., 2019, 9, 32517–32534 | 32523

View Article Online
RSC Advances
Paper
area to 106.28 m2 gꢁ1, i.e. the surface area of Fe3O4 was almost
The magnetic properties of the prepared samples were
doubled. This could be due to the efficient dispersion of Fe3O4 studied using VSM analysis and the magnetization curves of the
nanoparticles over rGO without particle agglomeration. The bare Fe3O4 and rGO/Fe3O4 nanocomposites were tested in the
surface area of graphite oxide prepared in this work was magnetic eld of ꢃ20 kOe at room temperature. The resulting
measured to be 738.34 m2 gꢁ1. The incremental addition of M–H curves are shown in Fig. 6. The S-shaped magnetization
graphene oxide further increased the surface area of the hysteresis loops indicates the superparamagnetic nature of the
nanocomposites signicantly due to the synergistic effect of prepared samples with immeasurable remanence and coer-
Fe3O4 nanoparticles on the surface layers of rGO. The surface civity. The saturation magnetization value of Fe3O4 nano-
area was measured to be 138.15 m2 gꢁ1 and 239.67 m2 gꢁ1 for particles was measured to be 51.2 emu gꢁ1
, which is
the 5% and 10% GO incorporated nanocomposites respec- signicantly lower than the theoretical magnetization value of
tively. The high surface area of the nanocomposites further 92 emu gꢁ1 for the bulk Fe3O4. This could be due to the effects
conrms that the graphene oxide is a good support for the of surface contribution and non-stoichiometry in the prepared
effective immobilization of Fe3O4 nanoparticles on its surface. sample.57 The surface contribution could be mainly from the
Similar results were previously reported by other researchers.54 pinning of the surface spins at the particle interface. However,
The average pore size and total pore volume of the samples this value is very close to the most of the reported values of
were measured using BJH method and was found to be 4.8 nm, Fe3O4 nanoparticles.58 The saturation magnetization values of
16.6 nm 18.6 nm and 20.1 nm and 0.175 cm3 gꢁ1, 0.219 cm3 the rGꢁO1 incorporated nanocomposites were found to be 49.3
gꢁ1, 0.258 cm3 gꢁ1 and 0.328 cm3 gꢁ1 for the bare Fe3O4, 2.5%, emu g , 44.3 emu gꢁ1, 42.9 emu gꢁ1, 40.5 emu gꢁ1 and 38.02
5% and 10% rGO incorporated composites respectively. Like emu gꢁ1 respectively for the 1%, 2.5%, 5%, 7.5% and 10%
surface area, the pore parameters of Fe3O4 were also found to graphene oxide incorporated nanocomposites. It is clear that
be increased by the incorporation of graphene oxide and its the saturation magnetization value of Fe3O4 was slightly
amount in the nanocomposites. The BJH plots of the nano- decreased to 49.3 emu gꢁ1 with the incorporation of 1% rGO. It
composites indicated the presence of different types of pores was further decreased to 38.02 emu gꢁ1 for the 10% nano-
in them, whereas only one kind of pore with the size of 4.8 nm composite. The decrease in the saturation magnetization could
was noted for the bare Fe3O4 and which could be originated be ascribed to the low quantity of magnetic Fe3O4 and the
from the inter-aggregation of Fe3O4 nanoparticles. It further presence of non-magnetic graphene part in the nano-
indicates the enhanced role of GO in tailoring the pore composites. Furthermore, the strong anchoring of highly
structures of the nanocomposites.55 Moreover, the increased dispersed Fe3O4 nanoparticles on the surface rGO layers
pore volume of the nanocomposites could be due to the reduces the magnetization moment.59 It additionally conrms
formation of secondary pores as discussed before, which may the formation of composite structure. However, the low satu-
originate from the close stacking between the rGO and Fe3O4 ration magnetization values of the composites are still more
nanoparticles in the composites as reported.56 The large than enough to ensure its efficient magnetic separation. The
surface area and pore volume of the nano-composites makes superparamagnetic behavior and the non-signicant coercivity
more catalytic active sites for a catalytic reaction and it would and remanence make the easy and quick magnetic recovery of
be better for a facile mass diffusion and transmission, the as-prepared nanocomposites from the solution by applying
providing improved catalytic activity.
an external magnetic eld. The removal of magnetic eld again
Fig. 5 Nitrogen sorption isotherms of the prepared samples.
32524 | RSC Adv., 2019, 9, 32517–32534
Fig. 6 Magnetic hysteresis loops of the prepared samples.
This journal is © The Royal Society of Chemistry 2019

View Article Online
Paper
RSC Advances
allows its dispersion in solution. Fig. S2† shows the effective higher binding energies were assigned to the oxygenated
magnetic recovery of the nanocomposite from reaction mixture. surface groups in graphene oxide. The peak at 286.6 could be
The chemical composition and surface states of the elements related to the C–O bonding whereas the peak at 288.0 eV could
present in the prepared samples were investigated using XPS be assigned to the C]O bonds.62 The high intensity of the peak
analysis and the wide scan survey spectra of rGO, Fe3O4 and 5% at 284.6 eV (C–C/C]C) compared to the other two peaks at
rGO/Fe3O4 nanocomposite are shown in Fig. 7(a). As shown, the 286.6 eV and 288 eV with low intensities conrm the reduction
rGO contained C and O as the elements at the binding energies of graphene oxide by the solvothermal treatment. The same
of 284.6 eV and 531.1 eV respectively. The Fe3O4 and 5% rGO/ peaks were found to be retained in the C 1s spectrum of the
Fe3O4 nanocomposite contained C, O and Fe as the elements at nanocomposite. However, the peak positions were slightly
the same binding energy positions of 284.6 eV, 531.1 eV and shied to higher binding energies. The binding energies of the
710.5 eV respectively.60 The absence of other elemental peaks peaks related to C–C/C]C, C–O, C]O bondings were measured
conrms the high purity of the prepared samples. The peak for to be 284.95 eV, 286.72 eV and 288.66 eV respectively. The shi
C in the Fe3O4 sample could be originated from the carbon used of the binding energies of C 1s peaks in the rGO/Fe3O4 nano-
for XPS analysis, which cannot be discarded. The presence of composite is not surprising since similar observations were
rGO in the composite was correctly identied by comparing the previously reported.63 It could be ascribed to the interaction of
peak intensities of C 1s. As seen, the peak intensity of C 1s in the these groups with Fe3O4 nanoparticles anchored on the surface
composite was higher than that of the same in bare Fe3O4, of rGO.
conrming the formation of nanocomposites. The curve tted
Fig. 7(c) shows the deconvoluted narrow scan spectra of Fe
narrow scan XPS spectra of C 1s are shown Fig. 7(b). As shown in 2p. There are two main peaks were observed at the binding
the C 1s of rGO, the peak at the binding energy of 284.6 eV was energies of 710.8 eV and 724.1 eV in the Fe 2p spectra of bare
assigned to the sp3 and sp2 carbon species of the reduced gra- Fe3O4. These values were attributed to the Fe 2p3/2 and Fe 2p1/2
phene oxide such as C–C and C]C bonding.61 The peaks at spin orbit peaks of Fe3O4 respectively.64 The other deconvoluted
Fig. 7 XPS spectra of the prepared samples (a) survey scan spectra and curve fitted narrow scan spectra of (b) C 1s, (c) Fe 2p and (d) O 1s of the
rGO, Fe3O4 and 5% rGO/Fe3O4 nanocomposites.
This journal is © The Royal Society of Chemistry 2019
RSC Adv., 2019, 9, 32517–32534 | 32525

View Article Online
RSC Advances
Paper
peaks at the binding energies of 714.14 eV, 718.79 eV and
728.16 eV could correspond to their satellite peaks. These values
were highly consistent with the standard binding energies of
Fe3O4.65 As observed in the case of C 1s, the positions of Fe 2p
peaks in the composite were also shied to higher binding
energies. The binding energies of Fe 2p3/2 and Fe 2p1/2 were
found to be 711.17 eV and 724.66 eV, due to their strong
interaction with rGO through Fe–O–C bonds. Their satellite
peaks also shadowed the same shi to higher binding energies.
The curve tted O 1s spectra of all of the samples were shown in
Fig. 7(d). Only one peak at the binding energy of 532.1 eV was
observed in the O 1s of reduced graphene oxide. This peak could
be assigned to the presence of partially reduced oxygen con-
taining groups such as C–O, O–C]O and O–H groups in it.66
The deconvoluted O 1s spectra of Fe3O4 showed two peaks at the
binding energies of 530 eV and 531.1 eV. The peak at 530 eV
corresponded to the lattice oxygen of Fe3O4, whereas the peak at
531.1 eV was related to the surface adsorbed hydroxyl groups.
The O 1s spectra of the composite also showed two deconvo-
luted peaks at the binding energies of 530.3 eV and 531.98 eV.
The peak due to lattice oxygen was found to be quite shied to
higher binding energies compared to that of bare Fe3O4. This
could be attributed to their interaction with reduced graphene
oxide via Fe–O–C bondings.67 However, the broad peak at
531.98 eV could be related to the presence of oxygen containing
groups such as C–O, O–C]O and O–H groups in the compos-
ites, le aer the partial reduction of GO in the solvothermal
reaction.
3.2 Catalytic activity of the rGO/Fe3O4 nanocomposites
The catalytic activity of the as-prepared nanocomposites was
investigated for the liquid phase oxidation of cyclohexene using
H2O2 as a green oxidant. The main products obtained in the
present study were identied to be 1,2-cyclohexane diol, cyclo-
hexene oxide, 2-cyclohexenol and 2-cyclohexenone as discussed
in the introduction. However, high selectivity was observed for
the diol product over rGO/Fe3O4 nanocomposites using H2O2.
Some preliminary experiments were rst conducted to study the
effect of catalytic reaction.
Fig. 8 Catalytic performance of the as-synthesized samples for the
liquid phase oxidation of cyclohexene (a) preliminary experiments (b)
reactions over rGO/Fe3O4 nanocomposites with various amounts of
GO (cyclohexene: 2 mmol, H2O2: 10 mmol, acetonitrile: 5 ml, catalyst
dosage: 0.05 g, temperature: 70 ꢀC, reaction time: 5 hours, atmo-
spheric pressure).
As seen in Fig. 8(a), a maximum cyclohexene conversion of
5.8% with 83% epoxide selectivity was reached in absence of
catalysts for 5 h of reaction. When bare Fe3O4 and rGO was used
as a catalyst, the conversion reached to 36.9% and 42.2% with
a diol selectivity of 35% and 48% respectively. The low activity of
the bare Fe3O4 could be attributed to the aggregation Fe3O4
nanoparticles. The activity of the rGO could be due to its
enhanced self-sacricial donor–accepter character due to the
presence of sp2 and sp3 carbon atoms together with the pres-
ence of unreduced surface acidic functional groups present in
it.67,68 However, cyclohexene oxide was also formed in high
proportions with the selectivity of 42% and 23% respectively. In
both cases the formation of allylic oxidation products such as
cyclohexenol and cyclohexenone were observed, which dropped
the overall selectivity of the products. In the blank run no allylic
oxidation products were obtained.
be clearly seen that the introduction of 1% rGO onto Fe3O4
increased the cyclohexene conversion to 53.3% with a diol
selectivity of 63%. It indicates the enhanced role of rGO/Fe3O4
nanocomposites for the selective liquid phase oxidation of
cyclohexene using H2O2 as oxidant. With increasing the
amount of rGO, the conversion and diol selectivity increased
signicantly until an optimum value of rGO and thereaer
a drastic drop was noticed in the catalytic activity. Using 2.5%
rGO/Fe3O4, the conversion increased to 68.8% with 69%
selectivity for the diol product. However, a maximum cyclo-
hexene conversion of 73.2% with a highest diol selectivity of
81% was obtained over 5% rGO/Fe3O4 nanocomposite, which
is the highest value among the prepared samples for 5 h of
reaction. The further increase of graphene oxide in the
composite resulted in a low conversion of cyclohexene and
Fig. 8(b) shows the effect of different amount of rGO in the
rGO/Fe3O4 nanocomposites for cyclohexene oxidation. It can
32526 | RSC Adv., 2019, 9, 32517–32534
This journal is © The Royal Society of Chemistry 2019

View Article Online
Paper
RSC Advances
a low selectivity for diol formation. Over 7.5% and 10% rGO such as acetonitrile, dichloromethane and chloroform were
incorporated composites, the cyclohexene conversion and diol selected for cyclohexene oxidation. As shown in Fig. 9(a), the
selectivity were calculated to be 61.3% and 48.5% and 66% and maximum cyclohexene conversion and diol selectivity was
53% respectively.
observed when the reaction carried out in acetonitrile solvent.71
From these results it is seen that the amount of rGO is Only 21% and 12% cyclohexene conversion with diol selectiv-
a crucial factor for the activity of the nanocomposite for cyclo- ities of 12% and 11% was observed when dichloromethane and
hexene oxidation. Even though a low amount of rGO (1%) chloroform was used as the solvents. Cyclohexene oxide was the
makes the Fe3O4 more effective, it is not enough to achieve main product with these solvents. The high conversion of
a good conversion and selectivity of the product. This could be cyclohexene in acetonitrile solvent could be due to its high
probably connected to the dispersion rate of Fe3O4 nano- dipole moment.72 Since it is polar aprotic solvent, it is expected
particles on the surface of rGO along with their surface area. to increase the substrate concentration to favour the reaction
The low amount of GO might not be sufficient for the dispersion kinetics and hence cyclohexene conversion could be increased.
of Fe3O4 nanoparticles effectively on the surface of rGO without The amount of acetonitrile was further increased to 10 ml and
particle aggregation. For an effective catalytic reaction, the 15 ml to investigate its effect on cyclohexene conversion and
active metal dispersion is highly important, and it is expected products selectivity. It can be seen that no appreciable change
that the 5% rGO provides a better dispersion for Fe3O4 nano- was noticed in the cyclohexene conversion for 10 ml solvent
particles compared to 1% and 2.5% rGO incorporated nano- whereas the conversion decreased to 32% when the amount of
composites. This makes the 5% rGO/Fe3O4 nanocomposite acetonitrile increased to 15 ml. However, the selectivity towards
more effective for the cyclohexene oxidation reaction. However, diol remained almost same in all amounts of acetonitrile. The
the decreased performance of the nanocomposites with high low cyclohexene conversion in high volume of solvent could be
rGO content such as 7.5% and 10% could be due to the blocking due to the less contact time of the substrate with active catalyst
of catalytic active sites of Fe3O4 by excess rGO layers together sites because of the reduced substrate concentration in the
with the decreased amount of Fe3O4 in the nanocomposites. reaction mixture. Therefore 5 ml of acetonitrile was selected for
Thus, it is clear that an optimum amount of rGO is compulsory the remaining studies.
to get a good conversion with high product selectivity.
Reaction temperature is the other important factor which
The high catalytic performance of the rGO/Fe3O4 nano- determines the efficiency of a liquid phase redox reaction. At
composites for the cyclohexene oxidation could be related to the room temperature, nearly 8.25% cyclohexene conversion with
synergistic effects of Fe3O4 nanoparticles with rGO in the high epoxide selectivity (ꢄ95%) was observed. Therefore, the
composites. A catalytic oxidation reaction is usually accompa- effect of reaction temperature in the range of 50 ꢀC to 80 ꢀC was
nied by a redox reaction.69 In the present case it is expected that further studied for cyclohexene oxidation and the results are
the Fe3O4 undergoes a redox cycle such as Fe2+ to Fe3+ (oxida- shown in Fig. 9(b). The cyclohexene conversion was found to be
tion) and Fe3+ to Fe2+ (reduction) during the reaction and the signicantly increased from 23.5% to 73.2% with increasing the
transfer of electron could be facilitated by the presence of rGO.70 reaction temperatures from 50ꢀ to 70 ꢀC. The selectivity to diol
The rGO enhances the transfer of electrons compared to GO due remained more or less same (80 ꢃ 1%) in all of these temper-
to the presence of a well-developed conjugated network of p atures. Furthermore, no allylic oxidation products were
electrons in it, originated by the partial reduction of surface observed at 50 ꢀC and 60 ꢀC whereas cyclohexene oxide was
groups in GO. Therefore, the active catalyst sites were generated detected as other side product with a selectivity of 20 ꢃ 1% at
rapidly in each redox cycle for further catalytic reaction. A these temperatures. The formation of allylic oxidation products
plausible reaction mechanism and the active species involved in was found to be starting from 70 ꢀC onwards. When the reaction
reaction was studied separately at the end of the article.
temperature increased to 80 ꢀC, the cyclohexene conversion was
In order to study the effect of reaction parameters on cyclo- slightly increased to 76.6% with a low diol selectivity of 58%.
hexene conversion and products selectivity, the parameters The oxidant H2O2 was added very carefully, without its
were varied individually by keeping the other parameters decomposition. This is highly desired at this temperature. The
constant in the cyclohexene oxidation. The 5% rGO/Fe3O4 selectivity of cyclohexene oxide, enol and enone was calculated
ꢀ
nanocomposite was used for the optimization studies since it to be 8%, 13% and 11% respectively at 80 C. Thus, it is seen
gave maximum cyclohexene conversion and diol selectivity. At that the high reaction temperature increases the selectivity
rst, the effect of solvent was studied since it plays a signicant towards the allylic oxidation products rather than epoxidation
role in the reaction and product selectivity. No conversion was product. However, our result is slightly deviated from the results
observed in the solvent free liquid phase oxidation of cyclo- of Yang et al.,73 possibly due to the difference in the catalysts.
hexene. When water was used as solvent, nearly 10.7% cyclo- They reported that at low temperatures, the reaction proceeds
hexene conversion with 47.3% epoxide and 51.6% diol was through the allylic oxidation pathway resulting in the selective
formed without any allylic oxidation products. The low effi- formation of a,b unsaturated alcohol and ketone whereas at
ciency of cyclohexene oxidation in water could be attributed to high temperatures of 70–80 ꢀC, the reaction resulted in the
the insolubility of cyclohexene in water and also due to its protic formation of more epoxide products such as epoxide and diol
nature. It is known that the aprotic solvents increase the over TiO2 modied V2O5/MoO3 catalysts. However, the forma-
substrate concentration around the active catalytic sites more tion of allylic oxidation products could be due to the radical
effectively than protic solvents. Therefore, three aprotic solvents polymerization/chain mechanism as reported by Zhong et al.
This journal is © The Royal Society of Chemistry 2019
RSC Adv., 2019, 9, 32517–32534 | 32527

View Article Online
RSC Advances
Paper
Fig. 9 Effect of various reaction parameters: (a) solvents and its amount [cyclohexene: 2 mmol, H2O2: 10 mmol, solvent: 5 ml, catalyst: 5% rGO/
Fe3O4 nanocomposite, catalyst dosage: 0.05 g, temperature: 70 ꢀC, reaction time: 5 hours, atmospheric pressure], (b) effect of reaction
temperature [cyclohexene: 2 mmol, H2O2: 10 mmol, acetonitrile: 5 ml, catalyst: 5% rGO/Fe3O4 nanocomposite, catalyst dosage: 0.05 g, reaction
time: 5 hours, atmospheric pressure], (c) effect of cyclohexene/H2O2 molar ratio [cyclohexene: 2 mmol, acetonitrile: 5 ml, catalyst: 5% rGO/
Fe3O4 nanocomposite, catalyst dosage: 0.05 g, temperature: 70 ꢀC, reaction time: 5 hours, atmospheric pressure], (d) effect of catalyst dosage
[cyclohexene: 2 mmol, H2O2: 10 mmol, acetonitrile: 5 ml, catalyst: 5% rGO/Fe3O4 nanocomposite, temperature: 70 ꢀC, reaction time: 5 hours,
atmospheric pressure].
and Dou et al.74,75 Furthermore, the reaction was not conducted 20 mmol whereas the other reaction parameters were kept
at 90 ꢀC, since 82 ꢀC is the reux temperature of acetonitrile and constant. As shown in Fig. 9(c), the cyclohexene conversion
the further increase of reaction temperature might lead to the drastically increased from 6% to 73.2% with an increase in the
rapid decomposition of H2O2 without taking part in the reac- molar ratio of cyclohexene to H2O2 from 1 : 1 to 1 : 5. However,
tion. The high conversion of cyclohexene with increasing reac- when the ratio increased to 1 : 10, only a small increment in
tion temperatures could be due to the increased reaction rate cyclohexene conversion was noticed, which is not so signicant
according to Arrhenius theory of temperature dependent reac- and promising compared to the results of low substrate to
tion rates. By considering the highest diol selectivity, 70 ꢀC was oxidant ratios. The cyclohexene conversion was measured to be
selected for further optimization studies, without giving priority 6%, 22.9%, 73.2% and 76.9% respectively for the 1 : 1, 1 : 2, 1 : 5
ꢀ
to the maximum conversion of cyclohexene at 80 C.
and 1 : 10 cyclohexene to H2O2 molar ratios. The high conver-
The effect of substrate to oxidant molar ratio was studied by sion of cyclohexene with increasing H2O2 concentration could
changing the amount of H2O2 in the liquid phase cyclohexene be due to the availability of more oxidant species for the
oxidation. The amount of H2O2 was varied from 2 mmol to substrate molecules at the active catalyst interfaces. However,
32528 | RSC Adv., 2019, 9, 32517–32534
This journal is © The Royal Society of Chemistry 2019

View Article Online
Paper
RSC Advances
the selectivities of the reaction products were found to be
completely depending on the oxidant concentration. At low
molar ratios of cyclohexene to H2O2, i.e. for 1 : 1 and 1 : 2 molar
ratios, the main product was cyclohexene oxide, with highest
selectivities of 96% and 69% respectively. The remaining con-
tained only diol and no allylic oxidation products were detected.
The effect of long duration of reaction time was further inves-
tigated in cyclohexene oxidation with a low substrate to H2O2
ratio of 1 : 2 under same experimental conditions. It indicated
that there is no increase in cyclohexene conversion for a period
of 10 h of reaction without any signicant change in the product
selectivities. Moreover, no allylic oxidation products were
formed even aer a long duration of reaction.
When the oxidant ratio increased to 1 : 5, surprisingly, the
conversion reached to 73.2% with a maximum diol selectivity of
81%. Moreover, in addition to epoxide, allylic oxidation prod-
ucts also formed in small quantities. The further increase of
oxidant concentration (1 : 10) has an adverse effect on the diol Fig. 10 (a) Effect of reaction time on the cyclohexene conversion and
products selectivity (cyclohexene: 2 mmol, H2O2: 10 mmol, acetoni-
trile: 5 ml, catalyst: 5% rGO/Fe3O4 nanocomposite, catalyst dosage:
0.05 g, temperature: 70 ꢀC, atmospheric pressure).
selectivity. The diol selectivity was dropped to 62% and the
concentration of allylic oxidation products increased signi-
cantly. The selectivity of cyclohexene oxide, cyclohexenol and
cyclohexenone was calculated to be 8%, 21% and 19% respec-
tively. This indicates that the excessive use of H2O2 does not
favour the selective oxidation of cyclohexene into diol whereas
the selective formation of allylic oxidation products was initi-
ated. Therefore, 1 : 5 molar ratio of cyclohexene to H2O2 was
selected as the optimum ratio for the selective liquid phase
oxidation of cyclohexene to diol.
decreased gradually with time. At the same time a gradual
increase in the diol selectivity was observed along with the
formation of allylic oxidation products. A highest diol selectivity
of 81% was obtained at 5th hour of the reaction and it remained
unchanged for further prolonged duration of reaction. The
increased diol selectivity with reaction time could be due to the
hydrolysis of the formed epoxide.
Fig. S3† shows the effect of different oxidants in the cyclo-
hexene conversion and selectivity of the products. As seen, the
H2O2 assisted reaction achieved a maximum cyclohexene
conversion of 73.2% with 81% diol selectivity. However, the use
of tert-butyl hydroperoxide (TBHP) as an oxidant in the reaction
increased the substrate conversion to 82% with a highest
epoxide selectivity of 77%. The diol selectivity was found to be
only 13% in this case. It indicates the role of different oxidants
in the selective distribution of the products.
A number of mechanisms were previously reported for
cyclohexene oxidation depending on the product distribution,
type of oxidants and catalysts used for the reaction. Based on
the selective formation of diol, a plausible reaction mechanism
was proposed. The oxidation of cyclohexene follows different
pathways to achieve the products as reported. In the rst
pathway, the direct epoxidation of cyclohexene occurs and
yields cyclohexene oxide. It further undergoes hydrolysis to
produce 1,2-cyclohexene diol. In the second pathway, the
oxidation of cyclohexene occurs at allylic position and results in
the formation of 2-cyclohexenol. It further oxidizes to 2-cyclo-
hexenone. Since 1,2-cyclohexane diol was the major product
formed in all of the reactions along with the formation of direct
epoxidation and allylic oxidation products, a different mecha-
nism is expected together with the rst mechanistic pathway
(hydrolysis of epoxide), which is not the same, as reported in
many of the articles. Therefore, at rst, a radical scavenger study
The amount of catalyst was varied from 25 mg to 75 mg to
study its effect on cyclohexene conversion and products selec-
tivity. As seen in Fig. 9(d), the cyclohexene conversion increased
with increasing the catalyst weight as expected, due to the
availability of more active catalytic sites in the reaction mixture.
At a low catalyst weight of 25 mg, only epoxidation products
were observed with 46% diol and 54% epoxide selectivity with
a cyclohexene conversion of 28%. The further increment of
catalyst weight to 50 mg increased the cyclohexene conversion
to 73.2% with 81% diol selectivity. A small increase in cyclo-
hexene conversion (75.1%) was observed when the reaction
conducted with 75 mg catalyst. However, the selectivity of diol
decreased to 65% with the increased formation of allylic
oxidation products. Therefore, 50 mg of catalyst was found to be
optimum for getting high cyclohexene conversion with
maximum diol selectivity in the present experimental
conditions.
The effect of reaction time on cyclohexene conversion and
product selectivities was studied in the liquid phase oxidation
of cyclohexene and the results are shown in Fig. 10. The cyclo-
hexene conversion was found to be gradually increased with
reaction time. As shown in the gure, the cyclohexene conver-
sion increased from 34% to 73.2% as the time increased from 2
to 5 hours. Further increase of reaction time by one hour only
increased the conversion to 75.3%. Aer 6 hours of reaction, no
signicant change was noticed in the cyclohexene oxidation,
because the calculated cyclohexene conversion for 7th hour
remained same as that measured in the 6th hours of reaction.
Moreover, it is found that the selectivity of cyclohexene epoxide
This journal is © The Royal Society of Chemistry 2019
RSC Adv., 2019, 9, 32517–32534 | 32529

View Article Online
RSC Advances
Paper
was performed using butylated hydroxytoluene as a radical conversion could be achieved. It is reported that the hydroxyl
scavenger. The results had shown that the addition of 5 mmol of radicals are more reactive than the hydroperoxyl radicals since
butylated hydroxytoluene to the reaction mixture drastically their reaction rate constants are very low compared to hydroxyl
decreased the cyclohexene conversion to 9% with a lowest diol radicals.78 Therefore, it is expected that the hydroxyl radicals
selectivity of 2.1%. Cyclohexene epoxide and allylic oxidation play a major role in the oxidation of cyclohexene. A schematic
products such as cyclohexenol and cyclohexenone were formed representation of plausible mechanism is shown in Fig. 11.
with the selectivities of 2.8%, 41.1% and 54% respectively. The
The electron spin resonance (ESR) spectroscopic method was
decreased cyclohexene conversion conrms that the epoxida- applied to conrm the formation of hydroxyl radicals. For this
tion proceeds through a radical mechanism. Previously Zubir study, 5% rGO/Fe3O4 composite was selected as a representative
et al.76 reported the formation of radical species by H2O2 sample and the spectrum was recorded using 5,5-dimethyl-1-
decomposition over GO/Fe3O4 composite. According to them, pyrroline-N-oxide (DMPO) as a radical trapping agent in the
a synergistic interfacial effect of Fe3O4 and graphene oxide was presence of H2O2. The resulting spectrum at two different time
observed in the composite for the facile formation of radical intervals of 5 and 10 minutes is shown in Fig. 12. It showed the
species through a Fenton like mechanism via electron transfer. presence of four peaks in the form of DMPO-cOH signals with
Initially, the oxidant H2O2 gets adsorbed on the solid liquid a peak intensity ratio of 1 : 2:2 : 1, which conrms the forma-
interface of the nanocomposite and the active catalytic sites of tion of hydroxyl radicals.
Fe3O4 (Fe2+ and Fe3+), anchored on the surface of reduced gra-
The practical application of a heterogeneous catalyst mainly
phene oxide decomposes the adsorbed H2O2 into hydroxyl depends on its reusability for successive run in the reaction.
(cOH) and hydroperoxyl (cOOH) radicals respectively. At this Therefore, four reusability studies were conducted aer sepa-
stage, Fe2+ and Fe3+ ions inter convert each other by their redox rating the catalyst from reaction mixture using an external
reaction as per eqn (1) and (2). In addition to Fe3O4, the rGO magnet. The magnetically separated catalysts were washed with
ꢀ
also takes part in this process. This is mainly due to the intrinsic deionized water followed by ethanol and dried at 75 C before
donor–acceptor surface characteristics of carbon nano- its use for the next run. The same experimental conditions were
materials.77 The adsorbed oxidant H2O2 further get activated on used for the reusability studies and the results are shown in
the surface of rGO and decomposes to hydroxyl and hydro- Fig. 13. It can be seen that the cyclohexene conversion and diol
peroxyl radicals according to eqn (3) and (4). Herein, the sp2 selectivity decrease slightly in each run with a simultaneous
(C]C) and sp3 (C–C) carbon atoms present in the rGO act as the increase in the epoxide selectivity. The cyclohexene conversion
reduced and oxidized carbon active sites respectively. There- was found to be 61.2% for the fourth reusability study with
fore, more radical species were formed in the reaction mixture a diol selectivity of 69%. It indicates that there is a gradual
according to eqn (1)–(4).75
activity loss in the continuous reaction. It could be attributed to
the leaching of metal species in the nanocomposite during its
continuous use. It was further conrmed by metal leaching test
measured by ICP-OES analysis. A part of the reaction mixture
obtained in the rst and fourth reusability run, aer the sepa-
ration of catalyst was analyzed by ICP-OES to check the possi-
bility of any iron content in the solution. The results show that
there is a small amount of iron loss due to metal leaching was
observed during the reusability studies. It was measured to be
less than 2% (1.8 ꢃ 0.2%) in both cases. This metal leaching
could be responsible for the decreased performance of the
catalyst for the continuous reaction.79 The reduction in the
reaction efficiency for continual run is not so much signicant
while concerning the run number. As the catalyst is active for
the continuous reaction, even though there is a slow decrease in
its performance; it is worth to mention their stability.
The stability of the nanocomposites for continuous reaction
could be due to the maintenance of its structural features.
Therefore, XRD and FTIR analysis were used to characterize the
reused catalysts for their crystalline and structural properties.
As shown in Fig. 14(a), there is no signicant crystalline change
was observed. The peaks of Fe3O4 were found to be retained in
the reused catalysts with a slight reduction in the intensity of
peak at 35.6ꢀ. This could be attributed to the metal leaching
process. Similar observation was also noted in the FTIR results.
Fig. 14(b) shows the FTIR spectra of the reused catalysts and it
indicate that the transmission bands related to Fe3O4 in the
nanocomposite was not changed in the spent catalysts,
Fe2+ + H2O2 / Fe3+ + cOH + OHꢁ
(1)
(2)
(3)
(4)
Fe3+ + H2O2 / Fe2+ + cOOH + H+
sp2 rGOC]C + H2O2 / sp3 rGOC–C + cOH + OHꢁ
sp3 rGOC–C + H2O2 / sp2 rGOC]C + cOOH + H+
Since reduced graphene oxide contains
a conjugated
network of p electrons, there is a good path for the rapid
transfer of electrons between rGO and active iron sites. The
presence of Fe–O–C bondings in the rGO/Fe3O4 nanocomposite
further facilitates the transfer of electrons in the composite with
an excellent reduction potential of graphene oxide.76 Thus, the
redox cycle of the active centers in Fe3O4 could be enhanced,
due to the synergistic effects of rGO and Fe3O4 nanoparticles. Or
more clearly it can be said that the redox cycle of Fe2+/Fe3+
system could be achieved quickly due of the presence of
conjugated network of p electrons in rGO. In addition to this,
the sacricial role of reduced graphene oxide for transferring
electron to Fe3O4 by the oxidation of its sp2 carbon domains is
signicant. This further explains the effectiveness of rGO/Fe3O4
nanocomposites for cyclohexene oxidation compared to their
individual components. The formed radicals especially hydroxyl
radicals further react with cyclohexene on the catalysts surface
and the selective formation of diol with high cyclohexene
32530 | RSC Adv., 2019, 9, 32517–32534
This journal is © The Royal Society of Chemistry 2019

View Article Online
Paper
RSC Advances
Fig. 11 Schematic representation of the plausible mechanism for cyclohexene oxidation.
obtained aer the rst and fourth reusability run of reaction. originated from the adsorbed CO2, probably on the surface of
Moreover, some addition transmission bands were observed. graphene oxide. The changes in the surface texture of the
These peaks could be originated from the oxidation of aromatic reduced graphene oxide as a result of the continuous run could
C]C bonds present in the rGO of nanocomposite in the H2O2 slow down the electron transfer through the disturbed conju-
environment.39,76 The transmission bands observed at gated network of p electrons in rGO to accomplish a redox cycle
1122 cmꢁ1 and 1187 cmꢁ1 could validate this assumption since of the active catalyst sites of Fe3O4. This could be another
these bands were related to oxygenated groups in graphene possible reason for the decreased activity of the composite for
oxide. The intense transmission band at 2363 cmꢁ1 could be continuous cyclohexene oxidation.
Fig. 12 DMPO-cOH profiles for 5% rGO/Fe3O4 nanocomposite Fig. 13 Reusability of the 5% rGO/Fe3O4 nanocomposite for cyclo-
measured by ESR analysis.
hexene oxidation.
This journal is © The Royal Society of Chemistry 2019
RSC Adv., 2019, 9, 32517–32534 | 32531

View Article Online
RSC Advances
Paper
the magnetic saturation values of Fe3O4 decreased by the
incorporation of rGO and its incremental amount in the
nanocomposites. However, the values were more than enough
for its separation from the solution with the use of an external
magnet. The increased of amount of rGO in the nano-
composites further enhanced the rate of particle dispersion and
decreased the aggregation rate of Fe3O4 nanoparticles. The
nitrogen physisorption analysis showed the enhancement of
textural properties size as specic surface area and pore
parameters aer the formation of nanocomposites. Among the
nanocomposites, 5% rGO/Fe3O4 composite showed highest
activity and selectivity for cyclohexene oxidation. The low and
high amounts of graphene oxide in the nanocomposite does not
positively impacted the conversion and product selectivity.
The study on the effect of reaction parameters proven that
a maximum cyclohexene conversion of 75.3% with a highest
diol selectivity of 81% was achieved with a substrate to oxidant
ꢀ
ratio of 1 : 5 using 0.05 g of 5% rGO/Fe3O4 catalyst at 70 C in
5 ml of acetonitrile solvent for 6 h of reaction. The enhanced
activity of the catalysts could be attributed to the synergistic
effects of Fe3O4 nanoparticles that were effectively dispersed on
the surface of rGO layers in the nanocomposites. The scavenger
studies on reaction indicated the active role of hydroxyl radicals
in the oxidation reaction, which could be formed by the
decomposition of H2O2 by a Fenton type mechanism in Fe2+
/
Fe3+ system via electron transfer in the nanocomposites.
Furthermore, due to the presence of a conjugated network of p
electrons in rGO, the Fe2+/Fe3+ redox cycle could be fastened in
the nanocomposites for the generation of more hydroxyl radi-
cals for oxidation. The additional sacricial role of rGO in the
electron transfer process is more effective for the fast accom-
plishment of Fe2+/Fe3+ redox cycle. Moreover, the composites
were successfully reused for four times in cyclohexene oxidation
without any drastic loss in the activity.
Conflicts of interest
Fig. 14 (a) XRD and (b) FTIR spectra of the reused samples.
There are no conicts of interests to declare.
4. Conclusions
Acknowledgements
An ultrasonication assisted precipitation combining sol-
vothermal method was reported to prepare a set of rGO/Fe3O4
nanocomposites with various amounts of graphene oxide and
were successfully used for the selective liquid phase oxidation of
cyclohexene to 1,2-cyclohexane diol using H2O2 as a green
oxidant. The prepared nanocomposites were characterized for
their crystalline, structural, textural and magnetic properties.
The active phase of iron was found to exist in the form of
magnetite in the nanocomposites. The eld emission scanning
and transmission electron microscopic analysis conrmed the
ne surface dispersion of more or less spherical shaped Fe3O4
nanoparticles on the surface layers of rGO nanosheets. The
Fe3O4 nanoparticles were found to be strongly anchored on the
surface of rGO layers through Fe–O–C bonding, as indicated by
the shi of binding energies of the representative elements in
the XPS characterization. The magnetic studies indicated that
This work was nancially supported by China Postdoctoral
Science Foundation (Grant No. 2019M653845XB), Postdoctoral
Research Funding of Kunming University of Science and
Technology (Grant No. 10988880) and National Natural Science
Foundation of China (Grant Nos. 21566014 and No. 21766016).
References
1 H. Cao, B. Zhu, Y. Yang, L. Xu, L. Yu and Q. Xu, Chin. J. Catal.,
2018, 39, 899.
2 B. G. Rao, P. Sudarsanam, P. R. G. Nallappareddy,
M. Y. Reddy, T. V. Rao and B. M. Reddy, Catal. Commun.,
2017, 101, 57.
3 S. El-Korso, I. Khaldi, S. Bedrane, A. C. Braham, F. T. Starzyk
and R. Bachir, J. Mol. Catal. A: Chem., 2014, 394, 89.
4 Z. L. Cheng and W. Sun, Chin. Chem. Lett., 2016, 27, 81.
32532 | RSC Adv., 2019, 9, 32517–32534
This journal is © The Royal Society of Chemistry 2019

View Article Online
Paper
RSC Advances
5 S. O. Lee, R. Raja, K. D. M. Harris, J. M. Thomas, 33 T. Zhou, Y. Zhao, W. Han, H. Xie, C. Li and M. Yuan, J. Mater.
B. F. G. Johnson and G. Sankar, Angew. Chem., Int. Ed.,
2003, 42, 1520.
6 K. C. Bhowmick and N. N. Joshi, Tetrahedron, 2006, 17, 1901.
Chem. A, 2017, 5, 18230.
34 S. Boudjema, E. Vispe, A. C. Braham, J. A. Mayoral, R. Bachir
and J. M. Fraile, RSC Adv., 2015, 5, 6853.
7 B. Plietker, M. Niggemann and A. Pollrich, Org. Biomol. 35 A. F. Guzik, A. Wawrzynczak and I. Nowak, Microporous
Chem., 2004, 2, 1116. Mesoporous Mater., 2015, 202, 80.
8 M. S. Hamdy, Microporous Mesoporous Mater., 2016, 220, 81. 36 H. Huang, H. Zhang, Z. Ma, Y. Liu, H. Ming, H. Li and
9 A. Bejar, S. B. Chaabene, M. Jaber, J. F. Lambert and Z. Kang, Nanoscale, 2012, 4, 4964.
L. Bergaoui, Microporous Mesoporous Mater., 2014, 196, 158. 37 W. Lei, Z. Wu, J. Ma, P. Yu and Y. Luo, Res. Chem. Intermed.,
10 Y. Fu, D. Sun, M. Qin, R. Huang and Z. Li, RSC Adv., 2012, 2,
3309.
11 X. Sang, J. Zhang, T. Wu, B. Zhang, X. Ma, L. Peng, B. Han,
X. Kang, C. Liu and G. Yang, RSC Adv., 2015, 5, 67168.
12 W. N. Oloo, Y. Feng, S. Iyer, S. Parmelee, G. Xue and L. Que,
New J. Chem., 2013, 37, 3411.
2016, 42, 6893.
38 B. Plietker, M. Niggemann and A. Pollrich, Org. Biomol.
Chem., 2004, 2, 1116.
39 S. Balasubramanyan, S. Arayathody, S. Sugunan and
B. N. Narayanan, Mater. Chem. Phys., 2018, 211, 23–33.
40 M. Kohantorabi and M. R. Gholami, Mater. Chem. Phys.,
2018, 213, 472.
´
13 P. Cancino, V. P. Garcıa, P. Aguirre and E. Spodine, Catal. Sci.
Technol., 2014, 4, 2599.
41 R. Hu and J. Zhou, Appl. Surf. Sci., 2018, 457, 789.
14 F. P. da Silva, R. V. Gonçalves and L. M. Rossi, J. Mol. Catal. A: 42 R. Rajagopal and K. S. Ryu, J. Ind. Eng. Chem., 2018, 60, 441.
Chem., 2017, 426, 534.
15 C. Li, Y. Zhao, T. Zhu, Y. Li, J. Ruan and G. Li, RSC Adv., 2018,
8, 14870.
43 Z. Cao, X. Wen, P. Chen, F. Yang, X. Ou, S. Wang and
H. Zhong, Colloids Surf., A, 2018, 549, 94–104.
44 W. S. Hummers and R. E. Offeman, J. Am. Chem. Soc., 1958,
80, 1339.
16 A. S. Novikov, M. L. Kuznetsov, B. G. M. Rocha,
A. J. L. Pombeiro and G. B. Shulpin, Catal. Sci. Technol., 45 K. Ouyang, C. Zhu, Y. Zhao, L. Wang, S. Xie and Q. Wang,
2016, 6, 1343. Appl. Surf. Sci., 2015, 355, 562–569.
17 Z. Y. Cai, M. Q. Zhu, H. Dai, Y. Liu, J. X. Mao, X. Z. Chen and 46 X. Lv, D. Tian, Y. Peng, J. Li and G. Jiang, Appl. Surf. Sci.,
C. H. He, Adv. Chem. Eng. Sci., 2011, 1, 15. 2019, 466, 937.
18 A. Dali, I. R. Hammoumraoui, A. C. Braham and R. Bachir, 47 X. Wang, Z. Zhao, J. Qu, Z. Wang and J. Qiu, J. Phys. Chem.
RSC Adv., 2015, 5, 29167. Solids, 2010, 71, 673.
19 D. Habibi, A. R. Faraji, M. Arshadi and J. L. G. Fierro, J. Mol. 48 J. Wu, Z. Ye, W. Liu, Z. Liu and J. Chen, Ceram. Int., 2017, 43,
Catal. A: Chem., 2013, 372, 90. 13146.
20 Q. Luo, M. Ji, S. E. Park, C. Hao and Y. Li, RSC Adv., 2016, 6, 49 G. Xie, P. Xi, H. Liu, F. Chen, L. Huang, Y. Shi, F. Hou,
33048. Z. Zeng, C. Shao and J. Wang, J. Mater. Chem., 2011, 22, 1033.
21 S. El-Korso, S. Bedrane, A. C. Braham and R. Bachir, RSC 50 F. Zhang, Y. Song, S. Song, R. Zhang and W. Hou, ACS Appl.
Adv., 2015, 5, 63382. Mater. Interfaces, 2015, 7, 7251.
22 D. Yang, T. Jiang, T. Wu, P. Zhang, H. Han and B. Han, Catal. 51 D. Chen, G. Wang, S. He, J. Liu, L. Guo and M. Cao, J. Mater.
Sci. Technol., 2016, 6, 193. Chem. A, 2013, 1, 5996.
23 C. Antonetti, A. M. R. Galletti, P. Accorinti, S. Alini, P. Babini, 52 K. Wu, C. Jing, J. Zhang, T. Liu, S. Yang and W. Wang, Appl.
K. Raabova, et al., Appl. Catal., A, 2013, 466, 21. Surf. Sci., 2019, 466, 746.
24 X. Y. Shi, P. M. Wang, K. Y. Liu, X. F. Dong, X. Y. Han and 53 M. Thommes, K. Kaneko, A. V. Neimark, J. P. Olivier,
J. F. Wei, Appl. Organomet. Chem., 2014, 28, 760.
F. R. Reinoso, J. Rouquerol and K. S. W. Sing, Pure Appl.
Chem., 2015, 87(9–10), 1051.
¨
¨
25 D. Dalmizrak, H. Goksu and M. S. Gultekin, RSC Adv., 2015,
5, 20751.
54 M. Rastgar, A. Shakeri, A. Bozorg, H. Salehi and
V. Saadattalab, Appl. Surf. Sci., 2018, 441, 923.
55 L. Yu, J. Chen, Z. Liang, W. Xu, L. Chen and D. Ye, Sep. Purif.
Technol., 2016, 171, 80.
56 F. Liu, Z. Wu, D. Wang, J. Yu, X. Jiang and X. Chen, Colloids
Surf., A, 2016, 490, 207.
57 I. S. Lyubutin, A. O. Baskakov, S. S. Starchikov, K. Y. Shih,
C. R. Lin, et al., Mater. Chem. Phys., 2018, 219, 411.
58 J. P. Fan, B. Zheng, Y. Qin, D. Yang, D. D. Liao, X. K. Xu,
X. H. Zhang and J. H. Zhu, Appl. Surf. Sci., 2016, 364, 332.
59 S. R. Kumar, J. G. Kim, C. Viswanathan, W. B. Kim,
R. K. Selvan and N. Ponpandian, Mater. Res. Bull., 2018, 97,
272.
26 Y. Usui, K. Sato and M. Tanaka, Angew. Chem., Int. Ed., 2003,
42, 5623.
27 A. A. Rosatella and C. A. M. Afonso, Adv. Synth. Catal., 2011,
353, 2920.
28 A. Theodorou, I. Triandallidi and C. G. Kokotos, Eur. J. Org.
Chem., 2017, 2017, 1502.
29 L. Yu, J. Wang, T. Chen, K. H. Ding and Y. Pan, Chin. J. Inorg.
Chem., 2013, 33, 1096.
30 L. Yu, J. Wang, T. Chen, Y. G. Wang and Q. Xu, Appl.
Organomet. Chem., 2014, 28, 652.
31 Y. G. Wang, L. Yu, B. C. Zhu and L. Yu, J. Mater. Chem. A,
2016, 4, 10828.
32 H. Huang, Y. Liu, S. T. Lee and Z. Kang, J. Mater. Chem., 2012, 60 Y. Li, J. Chu, J. Qi and X. Li, Appl. Surf. Sci., 2011, 257, 6059.
22, 337.
61 Y. Yu, L. Yan, J. Cheng and C. Jing, Chem. Eng. J., 2017, 325,
647.
This journal is © The Royal Society of Chemistry 2019
RSC Adv., 2019, 9, 32517–32534 | 32533

View Article Online
RSC Advances
Paper
62 X. Zhao, Y. Jia and Z. H. Liu, J. Colloid Interface Sci., 2019, 71 S. Ganji, P. Bukya, V. Vakati, K. S. R. Rao and D. R. Burri,
536, 463. Catal. Sci. Technol., 2013, 3, 409.
63 B. Yu, Y. Bai, Z. Ming, H. Yang, L. Chen, X. Hu, S. Feng and 72 S. K. Roy, D. Dutta and A. K. Talukdar, Mater. Res. Bull., 2018,
S. T. Yang, Mater. Chem. Phys., 2017, 198, 283. 103, 38.
64 Y. Liu, H. Huang, D. Gan, L. Guo, M. Liu, J. Chen, F. Deng, 73 G. Yang, M. D. Huff, H. Du, Z. Zhang and Y. Lei, Catal.
N. Zhou, X. Zhang and Y. Wei, Ceram. Int., 2018, 44, 18571. Commun., 2017, 99, 43.
65 D. Zhao, X. Gao, C. Wu, R. Xie, S. Feng and C. Chen, Appl. 74 W. Zhong, M. Liu, J. Dai, J. Yang, L. Mao and D. Yin, Appl.
Surf. Sci., 2016, 384, 1. Catal., B, 2018, 225, 180.
66 H. Yang, F. Li, C. Shan, D. Han, Q. Zhang, L. Niu and 75 J. Dou and F. Tao, Appl. Catal., A, 2017, 529, 134.
A. Ivaska, J. Mater. Chem., 2009, 19, 4632.
67 N. A. Zubir, C. Yacou, J. Motuzas, X. Zhang and J. C. D. da
Costa, Sci. Rep., 2014, 4, 4594.
68 A. A. Rosatella, C. A. M. Afonso and L. C. Branco, J. Chem.
Educ., 2011, 88, 1002.
76 N. A. Zubir, C. Yacou, J. Motuzas, X. Zhang, X. S. Zhao and
J. C. D. da Costa, Chem. Commun., 2015, 51, 9291.
77 C. M. Domınguez, P. Ocon, A. Quintanilla, J. A. Casas and
J. J. Rodriguez, Appl. Catal., B, 2013, 140–141, 663.
´
78 M. P. Moya, M. Graells, G. Castells, J. Amigo, E. Ortega,
´
69 Z. Hua, W. Ma, X. Bai, R. Feng, L. Yu, X. Zhang and Z. Dai,
Environ. Sci. Pollut. Res., 2014, 21, 7737.
G. Buhigas, L. M. Perez and H. D. Mansilla, Water Res.,
2010, 44, 2533.
70 A. Arshad, J. Iqbal, I. Ahmad and M. Israr, Ceram. Int., 2018, 79 H. Zang, C. Miao, J. Shang, Y. Liu and J. Liu, RSC Adv., 2018,
44, 2643.
8, 16193.
32534 | RSC Adv., 2019, 9, 32517–32534
This journal is © The Royal Society of Chemistry 2019