Versatile heterogeneous dipicolinate complexes grafted into kaolinite: Catalytic oxidation of hydrocarbons and degradation of dyes

Article abstract of DOI:10.1016/j.cattod.2013.09.031

New heterogeneous catalysts were prepared by immobilization of Me(II)-dipicolinate complexes (Me = Co, Mn or Ni) on kaolinite (Ka). The precursor material was kaolinite grafted with dipicolinic acid (dpa) obtained via melting of the acid. The catalysts we

Full text of DOI:10.1016/j.cattod.2013.09.031

GModel
CATTOD-8593; No. of Pages11
ARTICLE IN PRESS
Catalysis Today xxx (2013) xxx–xxx
Contents lists available at ScienceDirect
Catalysis Today
journal homepage: www.elsevier.com/locate/cattod
Versatile heterogeneous dipicolinate complexes grafted into
kaolinite: Catalytic oxidation of hydrocarbons and degradation of dyes
Francielle R. Araújoa, Joel G. Baptistaa, Liziane Marc¸ ala, Katia J. Ciuffia, Eduardo J. Nassara,
Paulo S. Calefia, Miguel A. Vicenteb, Raquel Trujillanob, Vicente Rivesb, Antonio Gilc,
Sophia Korilic, Emerson H. de Fariaa,∗
a Sol-Gel Group, University of Franca, Av. Dr. Armando Salles Oliveira, Parque Universitário, 201, 14404-600, Franca, SP, Brazil
b GIR QUESCAT, Departamento de Química Inorgánica, Universidad de Salamanca, Plaza de la Merced, S/N, 37008 Salamanca, Spain
c Departamento de Química Aplicada, Universidad Pública de Navarra, 31006 Pamplona, Spain
a r t i c l e i n f o
a b s t r a c t
Article history:
New heterogeneous catalysts were prepared by immobilization of Me(II)-dipicolinate complexes
(Me = Co, Mn or Ni) on kaolinite (Ka). The precursor material was kaolinite grafted with dipicolinic acid
(dpa) obtained via melting of the acid. The catalysts were prepared by suspending the Ka-dpa precursors
(lamellar or exfoliated) in Me2+ solutions with a cation/ligand ratio of 1:3. The grafted complexes were
characterized by thermal analyses, X-ray diffraction, UV/Vis and infrared spectroscopies, and transmis-
sion electron microscopy. The catalysts were tested in three reactions: (i) epoxidation of cis-cyclooctene
to cis-cyclooctenoxide reaching 55% yield, with total selectivity to cyclooctenoxide; (ii) oxidation of
cyclohexane, reaching 22% yield, with total selectivity towards cyclohexanone; and (iii) Fenton-like decol-
orization of the dyes metanil yellow, methylene blue and green light, reaching 70–100% decolorization,
the maximum effectiveness being observed for methylene blue degradation. The high conversion in
the oxidation reactions and the high levels of decolorization of dyes confirmed the versatility of these
heterogeneous catalysts based on kaolinite.
Received 25 June 2013
Received in revised form
15 September 2013
Accepted 23 September 2013
Available online xxx
Keywords:
Kaolinite
Heterogeneous catalysis
Dipicolinate complexes
Oxidation reactions
Dyes degradation
© 2013 Elsevier B.V. All rights reserved.
1. Introduction
In recent years, much interest has been given to hybrid materials
based on clays. These materials are produced by several reactions
The need to develop effective methods and catalysts for
oxidation reactions is becoming ever more important to be com-
monly applied to synthesis of pharmaceuticals, agrochemicals, and
flavours or fragrances, among others. From an industrial point of
view, homogeneous catalysis methods remain unpractical, partic-
ularly due to the high cost of the catalysts and the difficulty of their
recovery and reuse [1]. Thus, from an economic, environmental
and technical point of view, heterogeneous supported catalysis is
preferred because of the handling, separation and recycling abili-
ties. The functionalization of clay minerals, such as kaolinite, with
organic units renders the possibility of heterogeneization of active
sites and promotes the reuse. Recently, our research group has
reported the functionalization of natural kaolinite with pyridine-
carboxylic acids, alkoxides, and ironporphyrins by the soft-guest
displacement method [2,3].
and can lead to intercalation compounds with various applications,
depending on the type of interaction between the inorganic matrix
and the intercalated species. Transition metal complexes show
high catalytic activity and selectivity in hydrocarbon oxidations in
homogeneous media, but they have high difficulty to be recovered
and reused. An alternative for applications requiring high tem-
perature during the process, and more catalytic cycles with high
conversion and selectivity, is based on the use of these complexes
immobilized on/into inorganic matrices. A “heterogeneization” is
made in this method, as the complexes are immobilized in inor-
ganic matrices such as kaolinite, able to modify the kinetics and
mechanism of the reactions. Hybrid materials have been explored
for this purpose and, among them, the kaolinite-based materials
stand out due to the high thermal stability of the hybrids obtained.
The search for efficient and environmentally friendly oxidation
systems that could be used in green processes for several oxidation
reactions has been recently intensified [4,5]. Transition metals are
used in a large variety of catalytic processes. Bergamaschi et al.
[6] used transition metals supported on zirconia in reactions of
steam reforming of ethanol to obtain hydrogen. Thomas et al. [7]
∗
Corresponding author.
E-mail address: eh.defaria@unifran.br (E.H. de Faria).
0920-5861/$ – see front matter © 2013 Elsevier B.V. All rights reserved.
http://dx.doi.org/10.1016/j.cattod.2013.09.031
Please cite this article in press as: F.R. Araújo, et al., Versatile heterogeneous dipicolinate complexes grafted into kaolinite: Catalytic oxidation of
hydrocarbons and degradation of dyes, Catal. Today (2013), http://dx.doi.org/10.1016/j.cattod.2013.09.031

GModel
CATTOD-8593; No. of Pages11
ARTICLE IN PRESS
2
F.R. Araújo et al. / Catalysis Today xxx (2013) xxx–xxx
reported the use of manganese and cobalt catalysts in the oxi-
dation of toluene by hydrogen peroxide. Sen et al. [8] and Gupta
et al. [9] demonstrated the effectiveness of nickel and manganese
heterogeneous catalysts in oxidation reactions. Previous studies
from our group [2,10] have demonstrated the efficient use of kaolin-
ite as a support for metalloporphyrin and iron(III)-picolinate and
dipicolinate complexes to obtain catalytic systems with good per-
formance and high selectivity in cis-cyclooctene, cyclohexane and
Baeyer-Villiger oxidation.
in the oxidation of cyclohexane to cyclohexanol and cyclohex-
anone, and in the degradation of three dyes (metanil yellow,
methylene blue and green light).
2. Experimental
2.1. Catalyst preparation
2.1.1. São Simão’ kaolinite purification
The transfer of oxygen atoms to carbon substrates, to under-
stand or to mimic enzymatic porphyrin systems, has been
previously reported [11], as well as the study of non-heme enzy-
matic model systems, such as metal complexes, to catalyze these
reactions. In particular, Schiff base complexes are potentially appli-
cable as catalysts in several oxidation reactions, including alkene
epoxidation and hydrocarbon oxidation [9,12].
Industrial processes normally involve oxidation reactions in
the presence of a metal complex as a catalyst, together with
acids and permanganate as oxidant, in an homogeneous medium.
However, these processes involve serious problems, namely, diffi-
culty to recover the catalyst from the reaction medium, corrosion,
and the need for expensive, hazardous, non-selective oxidants
[9,12]. One strategy to overcome these drawbacks, is to het-
erogeneize the metal complex by immobilizing it on a polymer
or an inorganic matrix and to accomplish the reaction using
cleaner oxidants like hydrogen peroxide or molecular oxygen,
which generates water as the only byproduct. However, many
efforts are still necessary to develop heterogeneous catalysts
that do not undergo deactivation during the catalytic process
[9,12,13].
The kaolinite used in this work came from the municipality of
São Simão in the State of São Paulo, Brazil, and was kindly supplied
by the mining company Darcy R. O. Silva & Cia. It belongs to the
ball-clay type, known for its fine granulometry and for being rich in
hexagonal kaolinite. Kaolinite was purified by dispersion in water,
followed by sedimentation, according to Stock’s law and using the
procedure previously described [4,21–23]. The structural formula
of purified kaolinite is Si2.0Al1.96Fe0.03Mg0.01K0.02Ti0.03O7.06 [3].
2.1.2. Kaolinite intercalation with dimethylsulfoxide
To obtain the precursor intercalated with dimethylsulfoxide
(Ka-DMSO), the method described by Detellier et al. was followed
[8,9]. A portion of 20 g of the purified kaolinite was suspended
in a mixture of 180 cm3 of DMSO and 20 cm3 of H2O, which was
maintained at 60 ◦C under agitation for 10 days. The material was
centrifuged at 2000 rpm, washed with ethanol and oven-dried at
60 ◦C. The resulting complex, designated as Ka-DMSO, has the stoi-
chiometry Ka(DMSO)0.45
.
2.1.3. Kaolinite grafting with dipicolinic acid
The hybrid organic-inorganic material was obtained by keeping
a mass of the precursor (Ka-DMSO) in the presence of the melted
dipicolinic acid for 48 h. The dipicolinic acid/kaolinite DMSO com-
plex molar ratio was 5:1. The temperature for grafting H2dpa acid
was 190 ◦C. The resulting materials were washed several times with
isopropanol or water and oven-dried at 80 ◦C. The first sample was
washed several times with isopropanol, the second washed with
water, and the third with isopropanol and suspended in an ultra-
sonic bath for 8 h to promote the exfoliation of the hybrid precursor;
finally, all these samples were oven-dried at 80 ◦C and designated
as Ka-dpa-IP, Ka-dpa-wt and Ka-dpa-exf, respectively.
About 15% of the total world production of dyes is lost dur-
ing the dyeing process and is released to the textile effluents,
and this release of these colored waste waters in the ecosystem
is a dramatic source of non-aesthetic pollution, eutrophication
and perturbations in the aquatic life [14]. As international envi-
ronmental standards are becoming more severe, new methods
for the removal of organic pollutants, such as industrial dyes,
have been very much studied by various groups. Among them,
physical methods (as adsorption) [15], biological methods (as
biodegradation) [16], and chemical methods (as chlorination or
ozonation) [17], are the most frequently used ones. In this context,
many studies have demonstrated the applicability of heteroge-
neous catalytic technologies in “Fenton-like” systems, advanced
oxidation processes that use hydrogen peroxide (H2O2) and metal-
lic species for the treatment of hazardous organic and inorganic
pollutants existing in aqueous media. The Fenton-type processes
are very promising since they are easy to operate and maintain,
they offer a cost effective source of hydroxyl radicals, and achieve
high conversions. The conventional production of HO• radicals by
the Fenton mechanism occurs by means of addition of H2O2 to
Fe2+ according to the reaction: Fe2+ + H2O2 → Fe3+ + HO• + HO− [18].
Recently, some other metals have been used to generate HO• rad-
icals [19,20].
2.1.4. Synthesis of the complexes immobilized into kaolinite
To obtain the metal-containing catalysts, Ka-dpa solids were
suspended in a 0.1 mol/dm3 Me2+ chloride solution (Me2+ = Co2+
,
Mn2+ and Ni2+), using the volume needed to reach a cation/ligand
ratio of 1:2, and the mixture was stirred at 80 ◦C for 3 h. The suspen-
sions were then centrifuged, and the resulting solids were washed
with ethanol five times. The catalysts obtained were designated as
Me(Ka-dpa)-IP, Me(Ka-dpa)-wt and Me(Ka-dpa)-exf, making ref-
erence to the active metal and the method of preparation of the
hybrid precursor.
On the other hand, dyes have been extensively used as model
compounds for the evaluation of new catalysts, among other fac-
tors by the simplicity for following their reactions, especially
decolorization. New research in the preparation of innovative
Fenton-like reactions has been motivated by two main issues,
namely, the improvement of the efficiency in the use of green oxi-
dants, and the immobilization of the active sites into inert supports
to facilitate the separation of the catalysts when the reaction has
finished.
2.2. Characterization techniques
The powder X-ray diffractograms of the solids were acquired
on a Siemens D-500 diffractometer operating at 40 kV and 30 mA
(1200 W), using filtered Cu K␣ radiation. All the analyses were
processed at a scan speed of 2◦ per minute.
Infrared absorption spectra were obtained on a Perkin-Elmer
1739 spectrophotometer with Fourier transform, using the KBr pel-
let technique.
This work reports the catalytic study of Me(II)-dipicolinate com-
plexes (Me = Co, Mn or Ni) immobilized into kaolinite. The grafted
complexes were used as heterogeneous environmental-friendly
catalysts in the epoxidation of cis-cyclooctene to cyclooctenoxide,
UV-Vis spectra were recorded in the 200–800 nm range on a
HP 8453 Diode Array Spectrophotometer. The spectra of the solid
samples were recorded in a quartz cell with 0.1 cm path length, in
ethanol suspension.
Please cite this article in press as: F.R. Araújo, et al., Versatile heterogeneous dipicolinate complexes grafted into kaolinite: Catalytic oxidation of
hydrocarbons and degradation of dyes, Catal. Today (2013), http://dx.doi.org/10.1016/j.cattod.2013.09.031

GModel
CATTOD-8593; No. of Pages11
ARTICLE IN PRESS
F.R. Araújo et al. / Catalysis Today xxx (2013) xxx–xxx
3
Thermal analyses were carried out in a TA Instruments SDT Q600
Simultaneous DTA-TGA thermal analyzer, in the temperature range
between 25 and 1100 ◦C, at a heating rate of 10 ◦C/min and air flow
of 100 cm3/min.
3. Results and discussion
3.1. Characterization of the catalysts
Transmission electron microscopy (TEM) was performed using
a Zeiss-902 microscope over samples ground and dispersed in
ethanol by using an ultrasonic apparatus; then, a drop of the sus-
pension was placed on a Cu grid and air dried before the study.
The first steps of the synthesis procedure (purification of kaolin-
ite, swelling with dimethylsulfoxide) are common for all samples,
and have been previously reported [4]. The samples will be
described in two series, in one of them the solids washed with water
or isopropanol after treatment with dipicolinic acid, and in a second
series the solids submitted to ultrasonic exfoliation.
2.3. Catalytic performance
3.1.1. Kaolinite-dipicolinate hybrid systems washed with water
or isopropanol
The X-ray diffractograms of the catalysts derived from kaolin-
ite are shown in the Supplementary Material (Fig. S1). The solids
2.3.1. Oxidation of cis-cyclooctene and cyclohexane
Catalytic oxidation reactions were carried out at 25 ◦C in a
2.0 cm3 glass reactor, sealed with a Teflon-coated silicone sep-
tum and equipped with a magnetic stirrer. When the oxidant was
hydrogen peroxide, the Me-(Ka-dpa) catalyst was suspended in
1 cm3 of the solvent mixture (1,2-dichloroethane/acetonitrile, 1:1,
v/v), and the substrate was added, resulting in a constant cata-
lyst/oxidant/substrate molar ratio of 1:300:100.
˚
grafted with dipicolinic acid displayed basal spacings of 12.00 A
˚
when washed with isopropanol and 11.75 A when washed with
water, characteristic values of kaolinite expanded with dpa [3].
The sample washed with water was much less ordered than that
washed with isopropanol and it showed a clear peak at about 7.20 A,
corresponding to non-expanded layers, effect that is almost negli-
gible in the sample washed with isopropanol.
Washing of grafted samples with isopropanol promoted the
rearrangement of kaolinite layers, while washing with water pro-
moted the disorder of kaolinite layers [4]. The hydrogen bonds
between the aluminol (Al–OH) and siloxane groups (Si–O) in
kaolinite were weakened in grafted samples, and further washing
with high polar solvents such as water promoted the structural dis-
order by repulsion of the aluminol grafted to the dpa hydrophobic
surface and attracted by the siloxane surface.
˚
When the oxidant was iodosylbenzene (PhIO), it was first
obtained through hydrolysis of iodosylbenzene diacetate [24],
and its purity was evaluated by iodometric titration [25]. PhIO
(0.023 mmol) was added to the reactor containing the cata-
lyst (10 mg) and the dichloroethane/acetonitrile solvent mixture
(1 cm3). Then, 1.15 mmol of the substrate (cis-cyclooctene or cyclo-
hexane) and 10−2 cm3 of di-n-butyl ether as internal standard were
added.
The evolution of the reactions was followed by analyzing the
products at fixed times of 2, 4, 24, or 48 h. The products were iden-
tified using a HP 6890 Series GC System gas chromatograph (with a
flame ionization detector) equipped with a HP-INNOWax-19091N-
133 (polyethylene glycol length 30 m, internal diameter 0.25 ␮m)
capillary column. The products were quantified using a calibration
curve obtained with a standard solution. When the oxidant was
hydrogen peroxide, the conversion was based on the substrate,
while when the oxidant was iodosylbenzene, the yield was based
on the oxidant.
At the end of the reactions, the catalysts were recovered by
centrifugation, dried for 3 h at 60 ◦C before being used again in a
further catalytic cycle. The supernatant liquids were maintained in
the reactor for 24 h, aftermost the possible products formed were
quantified by GC. This simple test gave evidences of the leaching of
active Me(II)-species from the solid to the liquid.
All solvents and reagents were purchased from Mallinckrodt,
Aldrich, or Acros Organics and were of commercial grade, unless
otherwise stated. Dichloromethane (DCM) was suspended on anhy-
The reaction ratio was quantified, and the percentage of layers
intercalated decreased from 90% in Ka-dpa-IP to 59% in Ka-dpa-wt.
This suggests that washing with small and strongly polar molecules
promotes the partial lixiviation of the intercalated ligands. The
intercalation ratio was calculated on the basis of the relative inten-
sities of the 0 0 1 peak, characteristic of stacking of the layers in
the c-direction, so the low value found for the sample washed with
water is justified by the strong affinity of the siloxane layer to water
molecules, that promotes the structural disorder (see Table S1).
The infrared spectra of Ka, Ka-DMSO, Ka-dpa-IP and Ka-dpa-wt
displayed bands at 3699 and 3618 cm−1, characteristic of inner-
and inner surface hydroxyls, respectively (Fig. S2 and Table S2).
The other bands for inner surface hydroxyls existing in kaolinite
and in the kaolinite-DMSO sample were weakened (at 3653 cm−1
)
or even were not observed (3668 cm−1), thereby confirming the
functionalization of the kaolinite matrix. A new band at 3600 cm−1
was assigned to the formation of hydrogen bonds between the
unreacted inner-surface hydroxyl groups of kaolinite with the car-
boxylate groups of the dpa molecules. It is also possible that a low
amount of intercalated dpa molecules (in sample Ka-dpa-IP) are
located in the interlayer space of kaolinite, since it might not have
been totally removed, despite the exhaustive washing process. This
band at 3600 cm−1 was formerly observed upon grafting and/or
intercalation of molecules such as polyols, ethylpyridinium, and
D-sorbitol into kaolinite [26–28]. The diffusion of these interca-
lated molecules depends on the interactions between the grafted
molecules and unreacted hydroxyl groups, which would contribute
˚
drous CaCl2 for 2.5 h, filtered, distilled over P2O5, and kept over 4 A
molecular sieves. cis-Cyclooctene, cyclohexane, cyclohexanol, and
cyclohexanone were purified on a basic alumina column imme-
diately before use. Hydrogen peroxide (solution in water) was
donated by Peróxidos do Brasil and was iodometrically titrated
before use.
2.3.2. Degradation of dyes in Fenton-like heterogeneous catalysis
The degradation of the dyes-metanil yellow (MY, anionic), green
light (GL, anionic) and methylene blue (MB, cationic) was tested in
heterogeneous “Fenton-like” systems, using 10 mg of each Me(Ka-
dpa) catalyst, 5 cm3 of 25 mg/dm3 dye solutions, and 0.1 cm3 of
hydrogen peroxide (50% v/v), under a constant catalyst:dye:H2O2
mole ratio of 1:300:2000. The kinetics of dyes degradation was
accomplished between 1 and 1740 min. The dye degradation and
final concentration of each solution was quantified by the calibra-
tion curve accomplished in solution by UV-Vis spectroscopy using
the characteristic band of each dye (MB: 665 nm, MY: 475 nm, GL:
625 nm).
to the formation of the new band at 3600 cm−1
.
Bands characteristic of antisymmetric and symmetric stretching
modes of the carboxylate group were observed at 1689, 1566 and
1478 cm−1. The shift of the bands corresponding to inner surface
hydroxyls and the development of new bands in regions related
to the pyridine-carboxylic groups gave evidence of the grafting of
dipicolinic acid molecules into the interlayer space of kaolinite.
At the same time, the absence of C H and S O vibrations in the
spectra showed the total substitution of DMSO molecules by the
Please cite this article in press as: F.R. Araújo, et al., Versatile heterogeneous dipicolinate complexes grafted into kaolinite: Catalytic oxidation of
hydrocarbons and degradation of dyes, Catal. Today (2013), http://dx.doi.org/10.1016/j.cattod.2013.09.031

GModel
CATTOD-8593; No. of Pages11
ARTICLE IN PRESS
4
F.R. Araújo et al. / Catalysis Today xxx (2013) xxx–xxx
wash the hybrid material. The second mass loss stage (235 ◦C,
20
105
100
95
90
85
80
75
70
65
60
TG
DTG
DTA
3.8%) was assigned to removal of residual DMSO and/or dipicolinic
acid intercalated into kaolinite, and the third stage (345 ◦C, 7.4%)
was assigned to the decomposition of dpa grafted in the inter-
layer space of kaolinite. The appearance of a single stage mass
loss in this region can be attributed to the removal of interca-
lated acid moieties during the washing process, so, the mass loss
corresponding to grafted dpa decreased from 9.5% (Ka-dpa-IP) to
7.4% (Ka-dpa-wt). The fourth mass loss stage (487 ◦C, 15.2%) was
attributed to the dehydroxylation of kaolinite and the decompo-
sition of residual carbon from the interlayer space of kaolinite.
Another effect noted is the variation in the temperature of kaolin-
ite dehydroxylation, which decreases to 473 ◦C in Ka-dpa-IP and
487 ◦C in Ka-dpa-wt, respectively (513 ◦C in pure kaolinite). This
decrease suggested a covalent bond between organic and inor-
ganic units, by reaction with the aluminol surface of kaolinite
[23,25,29–33].
15
10
5
exo
endo
0
-5
-10
100
200
300
400
500
600
700
800
900 1000
Temperature (ºC)
Fig. 1. TG, DTG and DTA curves of Ka-dpa-IP recorded in O2 atmosphere.
Based on the thermal and chemical analysis results, the ratio of
dpa molecules per unit cell of kaolinite was quantified, the resulting
formulae being Ka(dpa)0.684 for sample Ka-dpa-IP and Ka(dpa)0.430
for sample Ka-dpa-wt. The C/N ratios in the functionalized solids
were very close to that in the pure dpa molecule.
dipicolinic acid [26–28]. In the specific case of the Ka-dpa-wt sam-
ple, characteristic antisymmetric and symmetric stretching mode
bands from the carboxylate groups evidenced the functionalization
of the kaolinite. The appearance of two bands (1695 and 1354 cm−1
)
Fig. 2 shows that the basal reflection of intercalated kaolinite
disappeared after complexation with the metallic cations. This hap-
pened in all solids, together with an increase in the intensity of the
peak of non-expanded kaolinite, which suggests a different planar
arrangement of kaolinite layers and dpa molecules when complex-
ing the Me(II) ions. This rearrangement led to the formation of a
in the spectrum of the solid washed with water suggests that some
carboxylate groups, possibly free or weakly bound to the clay struc-
ture, were interacting with water or hydroxyl groups existing in
the interlayer space of kaolinite. At the same time, the character-
istic vibration of the inner surface aluminol group (938 cm−1) was
absent, this being a further evidence that kaolinite was indeed func-
˚
crystalline phase with peaks at 8.39, 6.28, 5.95 and 5.48 A, compat-
ible with pseudo-tubular halloysite [34]. The structural change of
kaolinite leading to the formation of nanotubes has been reported
by Matusik et al. [35–37]. Halloysite is a clay mineral with a chemi-
cal formula very similar to that of kaolinite, and naturally occurs in
a tubular form [37–39]. Formation of the tubular phase in kaolin-
ite functionalized with dipicolinate complexed with Me(II) cations
can be justified by the interaction between the Me(II)-dipicolinate
complex and the inorganic matrix, thereby reducing the interac-
tions between the layers and promoting their curling [40–43]. After
immobilization of Me2+ species into the Ka-dpa precursor another
fact that can contribute to the exfoliation process is the possible
leaching of intercalated moieties of H2dpa via a deintercalation
process promoting the decrease of the basal spacing. However, the
grafted species is very stable and remained grafted after insertion
of cationic species into the interlayer space and remained intact in
the exfoliated moities. The XRD patterns revealed that the char-
tionalized, via covalent Al
dipicolinic acid [26–28].
O C bonds between the clay and the
The thermal behavior of kaolinite and the kaolinite-DMSO pre-
cursor has been previously reported [8,16,21,23–27]. The TG, DTG
and DTA curves for grafted derivatives Ka-dpa-IP and Ka-dpa-wt
are shown in Fig. 1 and S3 and the data summarized in Table S3.
The TG curve for sample Ka-dpa-IP showed five mass loss stages;
the first one with an endothermic peak at 52 ◦C (1.9%) was assigned
to the elimination of the solvent used to wash the hybrid material.
The second stage, with the maximum mass loss at 219 ◦C (7.7%), was
assigned to removal of residual DMSO and/or acid intercalated into
the kaolinite interlayer space. The third and fourth stages, centered
at 343 ◦C (5.4%) and 373 ◦C (4.1%), were assigned to the decomposi-
tion of dpa molecules grafted into kaolinite. And the last mass loss
(473 ◦C, 16.6%) was assigned to dehydroxylation of kaolinite and
elimination of residual carbon from the interlayer space of kaolin-
ite. These results were similar to those reported by de Faria et al.
[10].
˚
acteristic peak of non-expanded kaolinite at 7.14 A, that remained
with a small intensity in Ka-DMSO and Ka-dpa precursors, becomes
broad after Me(II) incorporation, because of the presence of water
molecules coordinated to the Me(II) complexes located in the inter-
layer of kaolinite.
However, the TG curve for sample Ka-dpa-wt showed four mass
loss stages. The first stage, with an endothermic peak at 64 ◦C
(1.4%), was assigned to the removal of adsorbed water used to
Co(Ka-dpa)-IP
Mn(Ka-dpa)-IP
Ni(Ka-dpa)-IP
Ka-dpa-IP
Co(Ka-dpa)-wt
Mn(Ka-dpa)-wt
Ni(Ka-dpa)-wt
Ka-dpa-wt
10
20
30
5
10
15
2
20
25
30
2
Degrees)
Degrees)
Fig. 2. X-ray powder diffraction patterns of the precursor Ka-dpa-IP (left) and Ka-dpa-wt (right), and heterogeneous catalysts obtained after complexation with Me(II) cations
in aqueous media.
Please cite this article in press as: F.R. Araújo, et al., Versatile heterogeneous dipicolinate complexes grafted into kaolinite: Catalytic oxidation of
hydrocarbons and degradation of dyes, Catal. Today (2013), http://dx.doi.org/10.1016/j.cattod.2013.09.031

GModel
CATTOD-8593; No. of Pages11
ARTICLE IN PRESS
F.R. Araújo et al. / Catalysis Today xxx (2013) xxx–xxx
5
20
15
10
5
100
80
60
40
20
0
TG
100
DTG
DTA
95
exo
endo
90
0
85
-5
Co(Ka-dpa)-IP
-10
-15
-20
80
75
Mn(Ka-dpa)-IP
Ni(Ka-dpa)-IP
Ka-dpa
-20
4000
3500
3000
2500
2000
1500
1000
500
100
200
300
400
500
600
700
Wavenumber (cm-1)
Temperature (ºC)
Fig. 3. Infrared spectra of the Ka-dpa-IP precursor and heterogeneous catalysts
Me(Ka-dpa)-IP.
Fig. 4. TG, DTG and DTA curves of Ni(Ka-dpa)-IP carried out in O2 atmosphere.
Me-containing solids, the bands at 196, 220, and 265 nm were
associated to ligand to metal charge-transfer (LMCT) processes.
These bands could reasonably be attributed to the complexation
of Me(II) species with the dipicolinate units grafted on kaolinite
[46]. However, other authors have considered that these bands
were due to the d-d transfer to 2Eg state [47]. In our solids, this
LMCT band might be due to an electron transfer from the nitrogen
atoms or carboxylate groups of dpa anions grafted into kaolinite to
the empty Me(II) orbitals. Another evidence of formation of typical
Mn(II)-dpa complexes is observed in the visible region; the char-
acteristic broad band with maximum at 430 nm is, recorded after
complexation, confirming the complexation of Mn(II) ions into the
Ka-dpa hybrid matrix. Similar results were observed for the solids
containing Ni(II) and Co(II).
The infrared absorption spectra of the metal-containing cata-
lysts, and the spectra of the precursors, are shown in Figs. 3 and
S4, and the data summarized in Table S4. The vibrations of inner
and inner surface hydroxyls of kaolinite at 3622 and 3698 cm−1
,
respectively [3,29,30,32], are not shifted after complexation with
Ni2+, Mn2+ or Co2+, showing that complexation did not promote
the leaching of the ligands functionalized to aluminol groups of
kaolinite [2,10]. The characteristic bands of grafted dipicolinate,
due to the carboxylate group and the pyridine ring, were recorded
in the central region of the spectrum, appearing at 1695, 1674,
1597, 1572, 1485, 1387 and 1354 cm−1 in the precursors, and shif-
ting to 1651, 1599, 1572 and 1385 cm−1 after complexation with
Ni2+; to 1654, 1597, 1572, 1481 and 1381 cm−1 in the case of Mn2+
and to 1653, 1595, 1570, 1481 and 1385 cm−1 for Co2+, confirming
the coordination of the cations to the carboxylate groups and the
pyridine ring [3]. Álvaro et al. [44] have reported the incorpora-
tion of Fe(III) picolinate into zeolite Y and Na-mordenite, taking
the band at 1480 cm−1 as a spectroscopic evidence of the forma-
tion of the Fe(pa)3 complex. Gardolinski et al. [45] observed the
development of a weak band at 3545 cm−1 upon treating kaolinite
in aqueous solution, this band being attributed to the presence of
water molecules in the interlayer region of kaolinite, interacting
with the internal and external hydroxyls via hydrogen bonds. This
band was evident in the spectra of all our solids, both when washed
with water or with isopropanol.
The thermal behavior of some of the samples was investi-
gated under oxygen atmosphere; the curves from Ni(Ka-dpa)-IP
are given, as a representative example, in Fig. 4, while other curves
and a summary of the thermal effects are included in the Supple-
mentary Material (Fig. S5 and Table S5). The TG curves showed
three mass losses, the first stage, centered at 60–70 ◦C, was assigned
to removal of small amounts of solvent remaining in the solids.
The second mass loss, centered at 355–390 ◦C, was assigned to the
decomposition of the complex in the interlayer space of the clay,
with the subsequent decomposition and elimination of dipicolinate
ligands. The third stage, centered at 481–496 ◦C, corresponded to
kaolinite dehydroxylation, and the elimination of residual carbon
in the interlayer space of the kaolinite remaining from the previ-
ous process. Removal of residual carbon at these temperatures has
been previously shown in kaolinite hybrids by Faria et al. [10] and
Detellier and Tunney [22].
3.1.2. Kaolinite-dipicolinate hybrid systems exfoliated via
ultrasounds
Fig. 6 shows the powder XRD diffractograms of kaolinite func-
tionalized with dipicolinic acid after submission to exfoliation
treatments for 1–8 h in an ultrasound bath, compared to the original
hybrid material Ka-dpa-IP. A structural rearrangement of kaolin-
ite after the exfoliation treatment was observed, leading to the
formation of the crystalline phase above described, similar to hal-
loysite with a pseudo-tubular nature. This rearrangement was even
evident from short times and complete for the longest times con-
sidered.
Mn(Ka-dpa)
MnCl 2
dpa
Ka-dpa
Fig. 5 and Fig. S6 show the UV-Vis spectra of solids Me(Ka-dpa),
compared to the solution spectra of the Me(II) chlorides, the dpa
ligand, and the compound Ka-dpa. The UV-Vis spectra of Ka-dpa
displayed absorption bands at 204, 220, and 260 nm, characteris-
tic of the absorption of dipicolinate grafted to kaolinite. For the
200
250
300
350
400
450
500
550
600
Wavelength (nm)
Fig. 5. UV/Vis spectra of Mn2+ chloride, dipicolinic acid (H2dpa), Ka-dpa-IP, and
catalyst Mn(Ka-dpa)-IP.
Please cite this article in press as: F.R. Araújo, et al., Versatile heterogeneous dipicolinate complexes grafted into kaolinite: Catalytic oxidation of
hydrocarbons and degradation of dyes, Catal. Today (2013), http://dx.doi.org/10.1016/j.cattod.2013.09.031

GModel
CATTOD-8593; No. of Pages11
ARTICLE IN PRESS
6
F.R. Araújo et al. / Catalysis Today xxx (2013) xxx–xxx
20
15
10
5
TG
DTG
DTA
100
Ka-dpa-exf-8h
80
60
40
20
0
Ka-dpa-exf-7h
Ka-dpa-exf-6h
Ka-dpa-exf-5h
Ka-dpa-exf-4h
exo
endo
0
Ka-dpa-exf-3h
Ka-dpa-exf-2h
Ka-dpa-exf-1h
Ka-dpa-IP
-5
-10
800
100
200
300
400
500
600
700
Temperature (ºC)
10
20
30
Fig. 7. Thermal curves of Mn(Ka-dpa)-exf carried out in O2 atmosphere.
2
Degrees)
The second mass loss, exothermic, was centered close to 500 ◦C,
and assigned to the decomposition of the complexes and to the
dehydroxylation of the pseudo-halloysite structure (the endother-
mic character of this process is masked by the decomposition of the
ligands). As in the precursor, the decomposition of the clay occurs
at a higher temperature when it has the pseudo-tubular structure.
The nature of lamellar, exfoliated or pseudo-tubular kaolinite
was investigated by TEM (Fig. 8). The micrographs of the precursor
revealed the presence of hexagonal plates, confirming that grafting
did not change the morphology of the kaolinite particles. Such as
previously reported [41], insertion of large polar molecules into the
interlayer space could decrease the strong interaction between sil-
ica (Si O) and aluminol (Al OH) groups in kaolinite layers, thereby
promoting the exfoliation. After complexation with the divalent
cations, the structure changes, confirming the exfoliation observed
by powder XRD.
Fig. 6. X-ray powder diffraction patterns (detailed onset of the 2ꢀ = 5–30◦ region)
of the exfoliated solids compared to the precursor Ka-dpa-IP.
The infrared spectrum of the hybrid material exfoliated for 8 h
with an ultrasonic bath and the precursor material (Ka-dpa-IP)
is shown in Fig. S7. The exfoliated solid showed characteris-
tic vibrations at 3699 (␯OHinner surface) and 3622 cm−1 (␯OHinner
)
[3,29,30,32]. Characteristic vibrations of the carboxylate groups
appeared at 1647 and 1383 cm−1, corresponding to the antisym-
metric and symmetric stretching modes of these groups, indicating
that the hybrid material kept grafted after the exfoliation process,
what is confirmed by the absence of the characteristic vibration of
the aluminol group from the inner surface (938 cm−1). The decrease
in the intensity of the characteristic C–H bands confirmed the pres-
ence of less dpa molecules in comparison to the precursor material
Ka-dpa-IP.
Fig. S8 shows the thermal curves of the exfoliated hybrid
material (Ka-dpa-exf). The mass loss stages between 60–405 ◦C,
characteristic of the decomposition of functionalized dpa, disap-
peared. The increased thermal stability of the hybrid may have been
promoted by the folding of the structure of kaolinite, which led dpa
molecules to be encapsulated within the pseudo-tubular structure
[2,3,40]. The number of dpa molecules per unit cell of kaolinite
was calculated based on thermal analysis, leading to the formula
Ka(dpa)0.16, clearly showing the leaching of intercalated dpa units
promoted by ultrasonics. That is, the number of dpa molecules
is much lower, but these molecules are strongly grafted into the
kaolinite tubular structure.
Summarizing, all the catalysts have similar structures, and the
complexation media can favor the exfoliation of kaolinite. Actually,
the washing media control the leaching of organic moieties inter-
calated (non-grafted) in the kaolinite interlayer, and the amount of
complexes was slighted lower in the samples washed with water.
3.2. Catalytic activity of the solids
3.2.1. Oxidation of cis-cyclooctene and cyclohexane
To check the activity of the catalysts prepared herein, the oxi-
dation of the diagnostic substrate (Z)-cyclooctene by PhIO in the
presence of Ka-dpa and the catalysts Me(II)-(Ka-dpa) at room
temperature and pressure, was initially carried out (Table 1). (Z)-
cyclooctene was selected as a diagnostic substrate; indeed, the high
Complexation of Me(II) cations of the exfoliated precursor did
not promote structural changes in the pseudo-halloysite like struc-
ture, all the XRD peaks remained in the same positions and with
similar intensities for all metals (Fig. S9); the precursor Ka-dpa
Table 1
Yield of cyclooctenoxide (%) obtained by epoxidation of cis-cyclooctene with
iodosylbenzene at room temperature and pressure using the indicated catalysts.
˚
˚
showed a basal spacing of 7.81 A, but 7.70, 7.59 and 7.59 A after
the complexation with Co2+, Mn2+, and Ni2+, respectively. Grafting
followed by exfoliation into kaolinite-dipicolinate layers difficults
the formation of the pseudo-tubular phases.
Catalyst
Time (h)
2
4
24
48
After complexation with the divalent cations, the characteris-
tic infrared bands of intra and interlayer hydroxyls and aluminol
groups in kaolinite did not shift (Fig. S10), so the ligands continued
covalently bonded to aluminol groups in the surface of kaolinite
[10].
Thermogravimetric analysis of the catalysts based on Ka-dpa-
exf (the curve for Mn-solid is given as an example in Fig. 7; those for
the Ni- and Co-solids are included in Fig. S11) exhibited two mass
loss stages, the first, endothermic one centered at 45–55 ◦C, due to
removal of small amounts of solvent remaining from the synthesis.
Mn(Ka-dpa)-IP
Ni(Ka-dpa)-IP
Co(Ka-dpa)-IP
Mn(Ka-dpa)-wt
Ni(Ka-dpa)-wt
Co(Ka-dpa)-wt
Mn(Ka-dpa)-exf
Ni(Ka-dpa)-exf
Co(Ka-dpa)-exf
20
3
5
25
4
6
25
4
6
50
6
10
50
8
7
50
8
54
9
54
32
31
62
29
33
62
29
33
27
53
23
23
53
23
23
7
Catalytic reaction conditions: 10 mg of catalyst; 0.023 mmol of PhIO; 1.15 mmol of
cis-cyclooctene. 1 mL of dichloroethane/acetonitrile solvent mixture.
Please cite this article in press as: F.R. Araújo, et al., Versatile heterogeneous dipicolinate complexes grafted into kaolinite: Catalytic oxidation of
hydrocarbons and degradation of dyes, Catal. Today (2013), http://dx.doi.org/10.1016/j.cattod.2013.09.031

GModel
CATTOD-8593; No. of Pages11
ARTICLE IN PRESS
F.R. Araújo et al. / Catalysis Today xxx (2013) xxx–xxx
7
Fig. 8. TEM micrographs of the precursors (Ka-dpa)-IP (1) and Ka-dpa-wt (2), and catalysts Mn-(Ka-dpa)-IP (3) and Mn-(Ka-dpa)-wt (4).
stability of cis-cyclooctenoxide, its main (sometimes exclusive) oxi-
dation product, facilitates catalyst efficiency evaluation [48]. In
addition, epoxides are very useful intermediates in the chemical
industry; they are the starting chemicals to prepare a wide vari-
ety of products. PhIO was used as an oxidant because of its good
oxidant yields; it is relatively inert in the absence of metal com-
plexes, generates high valent oxo-metal species and PhI, and is a
polymeric solid that does not contain weak O–H bonds, thus avoid-
ing free-radical chain reactions normally initiated by oxidants such
as alkyl hydroperoxides (R–O–O–H) [49].
Control reactions were carried out in the absence of metal com-
plexes and in the absence of oxidant (separately). These control
reactions did not yield any product, not even after 48 h, although
parent kaolinite contained small amounts of Fe and Ti ions.
As observed, all the Me(II)-(Ka-dpa) materials catalyzed (Z)-
cyclooctene epoxidation by PhIO. Considering the preparation
methods, Me(II)-(Ka-dpa)-IP, Me(II)-(Ka-dpa)-wt, and Me(II)-(Ka-
dpa)-exf displayed a similar catalytic behavior, so these catalysts
will be discussed together as Me(II)-(Ka-dpa). The selectivity to
cyclooctenoxide was 100% in all cases. The epoxide yield increased
along time, the best yields were achieved at 48 h of reaction. It
was noted that the Mn-complex reached about 80% of their total
epoxide yield at 4 h of reaction, whereas the Co- and Ni-complexes
took 24 h to afford this same product yield. Furthermore, Mn(II)-
(Ka-dpa) afforded a higher epoxide yield than the other catalysts.
Catalytic yields decreased in the order Mn(II)-(Ka-dpa) > Co(II)-(Ka-
dpa) > Ni(II)-(Ka-dpa). Two reasons account for these observations:
the recognized high reactivity of the manganese cations, and the
fast product diffusion from the catalyst active sites into the reac-
tion solution [50]. It was also noted that clay exfoliation contributed
to release the product into the reaction medium; indeed, epoxide
yield increased linearly with time.
It has been previously reported that when Mn-Schiff base com-
plexes (salen or salophen) are immobilized onto a polystyrene
matrix in the presence of nitrogen ligands, they are highly efficient
oxidation catalysts when using single oxygen atom donors as oxi-
dants [9,46,51]. Mirkhani et al. [51] reported that nitrogen ligands
existing in the matrix act as co-catalysts, giving rise to 98% epoxide
yield during (Z)-cyclooctene oxidation by NaIO4 in CH3CN/H2O.
Although our systems afforded lower cis-cyclooctenoxide yield
as compared to Schiff base catalysts, it may be emphasized that
they offer some advantages over organic polymers, such as the
high thermal and chemical stability associated to the clay. In addi-
tion, the non-toxic Mn associated to the clay is very attractive for
being used in industrial processes within the principles of Green
Chemistry. Reaching 100% selectivity for cyclooctenoxide is also
remarkable.
The greatest issue when using transition metals as hetero-
geneized catalysts is to avoid catalyst leaching from the support.
The UV-Vis spectra of the supernatant solutions obtained after
the oxidation reactions showed that catalysts were not leached
from the support in any of the studied conditions. To prove that
catalysis was genuinely heterogeneous and to show the impor-
tance of adding the metal complexes to kaolinite in these reactions,
the solid catalysts were filtered off the reaction mixture, extra
oxidant was added to the resulting supernatant liquid, and the
oxidation reaction allowed proceeding under the same initial con-
ditions for further 48 h. After this period, cis-cyclooctenoxide was
not detected, indicating that the catalytic activity of the materi-
als had been really heterogeneous. However, the possibility that
Please cite this article in press as: F.R. Araújo, et al., Versatile heterogeneous dipicolinate complexes grafted into kaolinite: Catalytic oxidation of
hydrocarbons and degradation of dyes, Catal. Today (2013), http://dx.doi.org/10.1016/j.cattod.2013.09.031

GModel
CATTOD-8593; No. of Pages11
ARTICLE IN PRESS
8
F.R. Araújo et al. / Catalysis Today xxx (2013) xxx–xxx
Table 2
When the corresponding Mn-, Co- and Ni-chlorides were tested
as catalysts in an homogeneous medium, the release of bubbles was
observed after about five minutes of reaction, which indicated that
hydrogen peroxide disproportionated occurred even before oxy-
gen transfer from the oxidant to the substrate. Therefore, both the
dipicolinic complexes present on the clay and the clay itself opti-
mize the use of the peroxide oxidant, mimicking the site isolation
principle of biological enzymes.
Castaman et al. [52] used a binuclear carboxylated bridged
manganese complex immobilized on silica for epoxidation of
cyclooctene by iodosylbenzene and hydrogen peroxide in homo-
geneous and heterogeneous media. These authors reported that
Mn-complex catalysts were very efficient and selective in the epox-
idation of cyclooctene by iodosylbeneze (29–74% yield of epoxide),
but lower yields (1–14%) of epoxide were obtained with hydrogen
peroxide. These low yields with hydrogen peroxide were attributed
to the competitive parallel reaction of homolytic cleavage of hydro-
gen peroxide decomposition promoted by the catalysts.
Addition of the radical scavenger hydroquinone to the reaction
medium did not alter the epoxide yields, dismissing the radical
mechanism in our systems. Hence, only the high valent oxo-metal
species were originated in the reaction medium, via oxidative
heterolytic cleavage of hydrogen peroxide. This intermediate per-
formed (Z)-cyclooctene oxidation and is the same active species
reported for metalloporphyrins used as catalysts in epoxidation
reactions with hydrogen peroxide as oxidant [53].
Berkessel et al. [54] used the Mn-salen complex to catalyze
epoxidation reactions oxidized by hydrogen peroxide in the pres-
ence of imidazole, as co-catalyst. These authors reported that the
peroxide underwent heterolytic cleavage, to give the metal-oxo
active species. Apparently, heterolytic cleavage occurred without a
co-catalyst in our system. Stamatis et al. [55] obtained high catalytic
efficiency, slightly higher than ours, when using symmetrical acety-
lacetone Mn-based Schiff bases immobilized on a silica surface (by
grafting and sol–gel) as catalyst for (Z)-cyclooctene epoxidation by
hydrogen peroxide (55 vs. 70%), but their catalysts were effective
and selective towards the epoxide only in the presence of ammo-
nium acetate. Thus, our system is advantageous, because it does
not require any salt or co-catalyst. Moreover, our catalysts can be
reused thrice without efficiency loss (Fig. 9), while the catalysts
prepared by the aforementioned authors were destroyed in the first
reuse.
To confirm the efficacy and selectivity of the catalysts prepared
herein, as well as the active oxidant species involved in these sys-
tems, the performance of the Me(II)-(Ka-dpa) materials as catalysts
for cyclohexane oxidation was tested. Cyclohexane was chosen
because it is relatively inert, has a great industrial importance and
its selective oxidation into a mixture of cyclohexanol and cyclohex-
anone (called KA oil) is essential for the synthesis of polymers such
as nylon. However, the industrial process in this case is poorly selec-
tive and carried out under drastic conditions. Therefore, the search
for catalysts efficient at mild temperatures is a great challenge for
both researchers and industries [56].
Table 3 shows the results obtained in the cyclohexane oxidation
by PhIO catalyzed by Me(II)-(Ka-dpa) catalysts. All the Me(II)-(Ka-
dpa) materials catalyzed the oxidation of the cyclohexane saturated
C-H bonds, leading to 100% selectivity towards cyclohexanone. Nei-
ther cyclohexanol nor other by-products were detected.
These results highlight various aspects. First, 100% selectivity
towards cyclohexanone for all the catalysts is a significant out-
come. In all series the conversion rose with time; 48 h of reaction
provided the best product yields. Control reactions were carried
out in the absence of metal complexes and in the absence of oxi-
dant (separately), and did not yield any products, not even at 48 h.
The same reaction carried out under inert atmosphere showed a
decreasing in the performance, this reduction of the yields give
Yield of cyclooctenoxide (%) obtained by epoxidation of cis-cyclooctene with hydro-
gen peroxide at room temperature and pressure using the indicated catalysts.
Catalyst
Time (h)
2
4
24
48
Mn(Ka-dpa)-IP
Ni(Ka-dpa)-IP
Co(Ka-dpa)-IP
Mn(Ka-dpa)-wt
Ni(Ka-dpa)-wt
Co(Ka-dpa)-wt
Mn(Ka-dpa)-exf
Ni(Ka-dpa)-exf
Co(Ka-dpa)-exf
6
3
5
5
2
4
5
4
5
12
8
10
10
7
28
18
17
25
15
16
25
20
23
55
35
43
45
30
41
58
30
39
9
10
11
9
Catalytic reaction conditions: catalyst/H2O2/cis-cyclooctene = 1:300:100; 1 cm3 of
dichloroethane/acetonitrile solvent mixture.
homogenous catalytically inactive species leached into solution
cannot be ruled out.
It may be noted that the (Z)-cyclooctene epoxidation was not
carried out under inert atmosphere, so oxygen and alkenes could
have generated free radicals in the presence of the metal catalyst.
To find out whether the oxidizing species originated from the reac-
tion between the Me(II)-(Ka-dpa) materials and PhIO was different
from the radicalar species normally detected in systems that do not
involve oxo-metal species, (Z)-cyclooctene epoxidation was stud-
ied in the presence of hydroquinone. This radical scavenger did
not diminish the product yields, so the involvement of the radical
species as the epoxidizing agent can be dismissed. This fact could
support the sole involvement of the active high-valent metal-oxo
species in these reactions.
Catalyst reuse confirmed the high efficiency and stability of the
Me(II)-(Ka-dpa) systems. The solid catalysts were separated from
the reaction mixture after each experiment by simple filtration and
dried before being used in a subsequent run. All the catalysts could
be reused in three consecutive runs without activity loss.
It is a priority to devote attention to applying heterogeneous
catalysts in combination with mild oxidants such as hydrogen per-
oxide [47]. This oxidant has a number of advantages: water is
the only byproduct of the oxidation reactions, hydrogen perox-
ide shows a higher active oxygen content than other commercially
available oxidants, and most of the other oxidants result from
hydrogen peroxide derivatization. However, using this oxidant
poses a major challenge to avoid its self-destructive dispropor-
tionation in the presence of water. Thus, to prove the efficiency of
the Me(II)-(Ka-dpa) catalysts, (Z)-cyclooctene oxidation was car-
ried out using hydrogen peroxide as an oxidant, and the results
are included in Table 2. Again, the materials prepared by the
three methods, Me(II)-(Ka-dpa)IP, Me(II)-(Ka-dpa)wt, and Me(II)-
(Ka-dpa)exf, exhibited a similar behavior, so they will be discussed
in general terms by referring to them as Me(II)-(Ka-dpa).
Table 2 evidences that the epoxide yield rose with time; 48 h
of reaction provided the best product yields (although longer reac-
tion times were not tested). Mn(II)-(Ka-dpa) furnished the largest
alkene conversion among the catalysts here tested: Mn(II)-(Ka-
dpa) > Ni(II)-(Ka-dpa) > Co(II)-(Ka-dpa). Although the manganese
compound showed again the maximum efficiency, in contrast with
the use of PhIO as oxidant, when hydrogen peroxide was the
oxidant, the nickel complex was more efficient than the cobalt
one.
No leaching of the catalysts from the matrix was detected.
Therefore, catalysis was truly heterogeneous even under drastic
oxidation conditions. Blank tests also confirmed that the epoxide
only originated in the presence of Me(II)-(Ka-dpa). The high (Z)-
cyclooctene conversion into the epoxide indicated that the active
species during oxidation was the high valent oxo-metal species.
Please cite this article in press as: F.R. Araújo, et al., Versatile heterogeneous dipicolinate complexes grafted into kaolinite: Catalytic oxidation of
hydrocarbons and degradation of dyes, Catal. Today (2013), http://dx.doi.org/10.1016/j.cattod.2013.09.031

GModel
CATTOD-8593; No. of Pages11
ARTICLE IN PRESS
F.R. Araújo et al. / Catalysis Today xxx (2013) xxx–xxx
9
Fig. 9. Catalysts reuse of the indicated catalysts for three cycles in cis-cyclooctene oxidation reaction with hydrogen peroxide at room temperature and pressure.
good evidences of the possible radicalar mechanisms promoted by
O2 from air.
Many authors have reported that by using metal complex as cat-
alysts and hydrogen peroxide as oxidant, oxidation reactions can
be improved by the addition of imidazole or sodium bicarbonate,
since they act as co-catalysts and hinder hydrogen peroxide dispro-
portionation by the metal complex (catalase-type activity) or the
support [9,46,52,57]. To confirm that the homolytic mechanism did
not occur in our systems, the reactions were carried out in the pres-
ence of imidazole or NaHCO3 in the presence and absence of water.
No improvement of yields or conversion was detected, confirming
once more that in our systems the reaction underwent by the het-
erolytic cleavage of the hydrogen peroxide without involvement of
the radical mechanism.
The versatile nature of the Me(II)-(Ka-dpa) catalysts in the
oxidation of various substrates and oxidants tested in this work
provides strong evidence that the active high valent oxo-metal
species are responsible for oxygen atom transfer from the oxi-
dant to the studied substrates, as in the case of cytochrome P450
enzymes [11].
As observed in the case of (Z)-cyclooctene oxidation, and
reported by several authors for Schiff base complex encapsulated
in zeolites [9], the Mn-catalysts furnished the largest cyclohexane
conversion among the metal complexes investigated here, while
the activity of Co- and Ni-catalysts changes from one series to
another. The catalysts were not leached from the matrix, therefore
catalysis was truly heterogeneous even under the most drastic oxi-
dation conditions used. Blank tests also confirmed that the reaction
only occurred in the presence of Me(II)-(Ka-dpa). As in the case of
the studies involving (Z)-cyclooctene as a substrate, the high selec-
tivity towards the ketone indicated that the active species during
the oxidation was the high valent oxo-metal species.
By using the Mn(III)salophen complex immobilized onto
polystyrene-bound imidazole as catalyst and NaIO4 as oxidant,
Mirkhani et al. [51] obtained ketone and alcohol as products, with
greater selectivity for the ketone. Our material showed an advan-
tage over this catalyst, as only the ketone was produced. Adding
the radical scavenger hydroquinone to the reaction medium did
not alter the ketone yields, dismissing the radical mechanism.
Hydrogen peroxide was also used to oxidize cyclohexane
(Table 4). Despite lower conversion results, all the solids selectively
catalyzed the oxidation of cyclohexane by hydrogen peroxide, and
the ketone was the sole product. Mn(II)-(Ka-dpa) furnished the
largest cyclohexane conversion among the metal complexes inves-
tigated here, Mn(II)-(Ka-dpa) > Co(II)-(Ka-dpa) > Ni(II)-(Ka-dpa). All
the tests confirmed that the catalysis was truly heterogeneous and
that the reaction proceeded only in the presence of the catalysts
and without a radical mechanism.
3.2.2. Degradation of dyes
Degradation of the dyes was evaluated in terms of decoloriza-
tion. Actually, total mineralization of complex molecules, such as
the cationic and anionic dyes here considered, leads to the con-
version of organic carbon into CO2 and of nitrogen and sulfur
heteroatoms into inorganic ions, requiring many oxidation steps
and involving a large number of intermediate species. The first step
of this process is usually the cleavage of the chromophoric bond,
leading to the decolorization of the contaminated waters.
Results of the degradation of the three dyes considered in the
present work in Fenton-like systems catalyzed by Me(II)-Ka-dpa
Table 3
Table 4
Yield of cyclohexanone (%) obtained by oxidation of cyclohexane with iodosylben-
zene at room temperature and pressure using the indicated catalysts.
Yield of cyclohexanone (%) obtained by oxidation of cyclohexane with hydrogen
peroxide at room temperature and pressure using the indicated catalysts.
Catalyst
Time (h)
2
Catalyst
Time (h)
2
4
24
48
4
24
48
Mn(Ka-dpa)-IP
Co(Ka-dpa)-IP
Ni(Ka-dpa)-IP
Mn(Ka-dpa)-wt
Co(Ka-dpa)-wt
Ni(Ka-dpa)-wt
Mn(Ka-dpa)-exf
Co(Ka-dpa)-exf
Ni(Ka-dpa)-exf
13
0
0
13
0
0
25
2
0
15
0
0
16
0
0
50
4
3
21
12
13
22
0
14
53
21
11
22
14
15
22
14
16
62
25
17
Mn(Ka-dpa)-IP
Ni(Ka-dpa)-IP
Co(Ka-dpa)-IP
Mn(Ka-dpa)-wt
Ni(Ka-dpa)-wt
Co(Ka-dpa)-wt
Mn(Ka-dpa)-exf
Ni(Ka-dpa)-exf
Co(Ka-dpa)-exf
0
0
0
0
0
0
0
0
0
1
0
0
4
1
2
5
2
3
7
4
4
10
8
8
12
5
6
12
8
7
15
10
9
18
10
13
Catalytic reaction conditions: 10 mg of catalyst; 0.023 mmol of PhIO; 1.15 mmol of
cyclohexane. 1 cm3 of dichloroethane/acetonitrile solvent mixture.
Catalytic reaction conditions: catalyst/H2O2/cyclohexane = 1:300:100; 1 cm3 of
dichloroethane/acetonitrile solvent mixture.
Please cite this article in press as: F.R. Araújo, et al., Versatile heterogeneous dipicolinate complexes grafted into kaolinite: Catalytic oxidation of
hydrocarbons and degradation of dyes, Catal. Today (2013), http://dx.doi.org/10.1016/j.cattod.2013.09.031

GModel
CATTOD-8593; No. of Pages11
ARTICLE IN PRESS
10
F.R. Araújo et al. / Catalysis Today xxx (2013) xxx–xxx
Fig. 10. Degradation of the dyes under Fenton-like conditions using the indicated catalysts.
catalysts is shown in Fig. 10 (similar results were obtained for all
series, the only series washed with isopropanol is given in Fig. 10).
Degradation of each of the dyes is different, as shown by the shape
of the curves. In the case of MB, degradation is easy, it was higher
than 80% after only 5 min of reaction, and reached 99% after 30 min.
The behavior is analogous for all the catalysts. In the case of GL,
although 40% of degradation was reached after 5 min of reaction,
this degradation remained practically constant, with a very small
increase, up to 4 h of reaction, and a time of reaction as long as 29 h
was needed to reach 90% of degradation. In this case, and for long
reaction times, the Ni-catalyst showed the best efficiency. In the
case of MY, the degradation reached 85% after 2 h of reaction but
Please cite this article in press as: F.R. Araújo, et al., Versatile heterogeneous dipicolinate complexes grafted into kaolinite: Catalytic oxidation of
hydrocarbons and degradation of dyes, Catal. Today (2013), http://dx.doi.org/10.1016/j.cattod.2013.09.031

GModel
CATTOD-8593; No. of Pages11
ARTICLE IN PRESS
F.R. Araújo et al. / Catalysis Today xxx (2013) xxx–xxx
11
decreased on extending the reaction time. This may be due to the
complex structure of the dye, whose degradation may give rise to
organic intermediates absorbing in close positions, a fact which can
tentatively explain the non-stability with time of the degradation of
MY. As previously discussed by Sleiman et al. [58], the initial steps
of the reaction generally involve the hydroxylation and cleavage
of –NH– bonds, resulting in the formation of benzene, aniline and
phenol.
[3] E.H. de Faria, O.J. Lima, K.J. Ciuffi, E.J. Nassar, M.A. Vicente, R. Trujillano, P.S.
Calefi, J. Coll. Interf. Sci. 335 (2009) 210.
[4] E.H. de Faria, G.P. Ricci, L. Marc¸ al, E.J. Nassar, M.A. Vicente, R. Trujillano, A. Gil,
S.A. Korili, K.J. Ciuffi, P.S. Calefi, Catal. Today 187 (2012) 135.
[5] A. Goti, F. Cardona, in: P. Tundo, V. Esposito (Eds.), Green Chemical Reactions:
NATO Science for Peace and Security Series C, Environmental Security, Springer,
2008, p. 135.
[6] V.S. Bergamaschi, IPEN-USP, Brazil (2005) (Ph.D. Thesis).
[7] J.M. Thomas, R. Raja, Catal. Today 117 (2006) 31.
[8] R. Sen, S. Bhunia, D. Mal, S. Koner, Y. Miyashita, K. Okamoto, Langmuir 25 (2009)
13672.
It is noticeable that the highest degradation efficiency was
observed for MB, the only cationic species among the dyes studied.
Although kaolinite has no cation exchange capacity, it seems that
this is the dye that most efficiently interacts with the clay layers,
favoring its further degradation. In summary, high decolorization
(70–100%) was observed for the treated dyes at relatively short
reaction times, and by means of heterogeneous processes. It is also
remarkable that the reaction was completely heterogeneous, as no
leaching of the catalysts was detected, which favors the reuse of
the catalysts.
[9] K.C. Gupta, A.K. Sutar, Coord. Chem. Rev. 252 (2008) 1420.
[10] E.H. de Faria, Universidade de Franca, Brazil (2011) (Ph.D. Thesis).
[11] M. Costas, Coord. Chem. Rev. 255 (2011) 2912.
[12] K.C. Gupta, A.K. Sutara, C.-C. Lin, Coord. Chem. Rev. 253 (2009) 1926.
[13] B.S. Lane, K. Burgess, Chem. Rev. 103 (2003) 2457.
[14] A. Houas, H. Lachheb, M. Ksibi, E. Elaloui, C. Guillard, J.-M. Herrmann, Appl.
Catal. B 31 (2001) 145.
[15] M. Toor, B. Jin, Chem. Eng. J. 187 (2012) 79.
[16] B. Joshi, K. Kabariya, S. Nakrani, A. Khan, F.M. Parabia, H.V. Doshi, M.C. Thakur,
Am. J. Environ. Protect. 1 (2013) 41.
[17] A.C. Gomes, J.C. Nunes, R.M.S. Simões, J. Hazard. Mater. 178 (2010) 57.
[18] S. Papic, D. Vujevic, N. Koprivanac, D. Sinko, J. Hazard. Mater. 164 (2009)
1137.
[19] Y. Tu, S. Tian, L. Kong, Y. Xiong, Chem. Eng. J. 185–186 (2012) 44.
[20] E.R. Bandala, M.A. Pelaez, M.J. Salgado, L. Torres, J. Hazard. Mater. 151 (2008)
578.
4. Conclusions
[21] R.L. Frost, J. Kristof, E. Horvath, J.T. Kloprogge, Spectrochim. Acta Part A 56
(2000) 1191.
Kaolinites covalently grafted with dipicolinate anions proved
to be suitable supports for the effective immobilization of various
metal cations by forming complexes. Complexation of Ni2+, Co2+
and Mn2+ cations into the functionalized starting materials was
efficient in all cases, leading to the formation of active heteroge-
neous catalysts. These catalysts were advantageous for oxidation
reactions since they allowed for the use of mild conditions (room
temperature and pressure, use of non-polluting oxidants) and pro-
moted high activity and product selectivity. The catalysts were
very active for cis-cyclooctene oxidation (about 55% yield, with
total selectivity towards the epoxide), the oxidation of cyclohex-
ane (about 22% yield, with total selectivity towards cyclohexanone),
and the decolorization of dye solutions (70–100%), in this case being
rather sensitive to the molecular structure of the dye. The main
advantage of the kaolinite grafted complexes is their easy separa-
tion from the reaction mixture by simple filtration of the solid, thus
enabling catalyst reuse.
[22] C. Detellier, J.J. Tunney, Chem. Mater. 5 (1993) 747.
[23] S. Letaief, C. Detellier, Chem. Comm. 1 (2007) 2613.
[24] J.G. Sharefkin, H. Saltzmann, Org. Synth. 43 (1963) 62.
[25] H.J. Lucas, E.R. Kennedy, M.W. Forno, Org. Synth. 43 (1963) 483.
[26] J. Murakami, T. Itagaki, K. Kuroda, Solid State Ionics 172 (2004) 279.
[27] K.B. Brandt, T.A. Elbokl, C. Detellier, J. Mater. Chem. 13 (2003) 2566.
[28] S. Letaïef, B. Casal, P. Aranda, M.A. Martín-Luengo, E. Ruiz-Hitzky, Appl. Clay Sci.
22 (2003) 263.
[29] L.R. Avila, E.H. de Faria, K.J. Ciuffi, E.J. Nassar, P.S. Calefi, M.A. Vicente, R. Trujil-
lano, J. Coll. Interf. Sci. 341 (2010) 186.
[30] E.H. de Faria, K.J. Ciuffi, E.J. Nassar, M.A. Vicente, R. Trujillano, P.S. Calefi, Appl.
Clay Sci. 48 (2010) 516.
[31] R.L. Frost, W.N. Martens, J. Kristof, E. Horvath, J. Phys. Chem. 106 (2002) 4162.
[32] S. Letaief, C. Detellier, Can. J. Chem. 86 (2008) 1.
[33] J.J. Tunney, C. Detellier, Can. J. Chem. 75 (2003) 1666.
[34] M.C. MacEwan, J.L. Amoros, Anales Edafol. Fisiol. Veg. 9 (1950) 363.
[35] J. Matusik, A. Gawel, E. Bielanska, K. Bahranowski, Clay Clay Miner. 59 (2011)
116.
[36] J. Matusik, E. Stodolak, K. Bahranowski, Appl. Clay Sci. 51 (2011) 102.
[37] J. Matusik, A. Gawel, E. Bielanska, W. Osuch, K. Bahranowski, Clay Clay Miner.
57 (2009) 452.
[38] N.G. Veerabadran, R.R. Price, Y.M. Lvov, Nano Brief Rep. Rev. 2 (2007) 115.
[39] F. Chavarria, D.R. Paul, Polymer 45 (2004) 8501.
[40] S. Nakagaki, G.S. Machado, M. Halma, A.A.S. Marangon, K.A.D.F. Castro, N. Mat-
toso, F. Wypych, J. Catal. 242 (2006) 110.
Acknowledgments
This work has been carried out in the frame of
a
[41] J.E.F.C. Gardolinski, G. Lagaly, Clay Miner. 40 (2005) 547.
[42] S. Nakagaki, F. Wypych, J. Coll. Interf. Sci. 315 (2007) 142.
[43] K. Weissermel, H.-J. Arpe, (C.R. Lindley, Trans.), Industrial Organic Chemistry,
vol. 237, third ed., VCH Publishers, Weinheim, 1997.
[44] M. Álvaro, B. Ferrer, H. Garcia, A. Sanjuán, Tetrahedron 55 (1999) 11895.
[45] J.E. Gardolinski, F. Wypych, H.P.M. Filho, Quim. Nova 26 (2003) 30.
[46] M. Tabatabaee, B.-M. Kukovec, V. Razavimahmoudabadi, Z. Natur. 66b (2011)
813.
[47] J. Limburg, R.H. Crabtree, G.W. Brudvig, Inorg. Chem. 297 (2000) 301.
[48] R.G. Pearson, J. Am. Chem. Soc. 85 (1963) 3533.
[49] F. Gozzo, J. Mol. Catal. A 171 (2001) 1.
[50] M. Salavati-Niasari, Z. Salimi, M. Bazarganipour, F. Davar, Inorg. Chim. Acta 362
(2009) 3715.
Spain–Brazil Interuniversity Cooperation Grant financed by
MEC (PHB2011–0164–PC) and CAPES (267/12), and a Cooperation
Grant from Universidad de Salamanca and FAPESP (2013/50216–0).
Spanish authors thank additional financial support from Junta de
Castilla y León (SA009A11–2). The Brazilian group thanks support
from Brazilian Research funding agencies FAPESP (2011/17660–8
and 2012/07410–7) and CNPq, and Peróxidos do Brasil (Solvay) for
kindly supplying the 70 wt.% aqueous hydrogen peroxide solution.
Appendix A. Supplementary data
[51] V. Mirkhani, M. Moghadamb, S. Tangestaninejada, B. Bahramian, Appl. Catal. A
311 (2006) 43.
[52] S.T. Castaman, S. Nakagaki, R.R. Ribeiro, K.J. Ciuffi, S.M. Drechsel, J. Mol. Catal.
A 300 (2009) 89.
[53] T.G. Traylor, C. Kim, W.-P. Fann, C.L. Perrin, Tetrahedron 54 (1998) 7977.
[54] A. Berkessel, M. Frauenkron, T. Schwenkreis, A. Steinmetz, G. Baum, D. Fenske,
J. Mol. Catal. A 113 (1996) 321.
Supplementary material related to this article can be
found, in the online version, at http://dx.doi.org/10.1016/
j.cattod.2013.09.031.
[55] A. Stamatis, D. Giasafaki, K.C. Christoforidis, Y. Deligiannakis, M. Louloudi, J.
Mol. Catal. A 319 (2010) 58.
References
[56] U. Schuchardt, D. Cardoso, R. Sercheli, R. Pereira, R.S. da Cruz, M.C. Guerreiro,
D. Mandelli, E.V. Spinacé, E.L. Pires, J. Mol. Catal. A 319 (2010) 58.
[57] M. Ghorbanloo, H.H. Monfared, C. Janiak, J. Mol. Catal. A 345 (2011) 12.
[58] M. Sleiman, D. Vildozo, C. Ferronato, J.-M. Chovelon, Appl. Catal. B 77 (2007) 1.
[1] C.L. Hill, Angew. Chem. Int. Ed. 43 (2004) 402.
[2] N. Bizaia, E.H. de Faria, G.P. Ricci, P.S. Calefi, E.J. Nassar, K.A.D.F. Castro, S. Naka-
gaki, K.J. Ciuffi, R. Trujillano, M.A. Vicente, A. Gil, S.A. Korili, Appl. Mater. Interf.
11 (2009) 2667.
Please cite this article in press as: F.R. Araújo, et al., Versatile heterogeneous dipicolinate complexes grafted into kaolinite: Catalytic oxidation of
hydrocarbons and degradation of dyes, Catal. Today (2013), http://dx.doi.org/10.1016/j.cattod.2013.09.031