Activation of H2O2 and superoxide production using a novel cobalt complex based on a polyampholyte

Article abstract of DOI:10.1016/j.apcata.2013.08.002

A new catalyst based on Co(II) and a hydrogel with property of polyampholyte was characterized by equilibrium studies of Co(II) uptake, solid-state NMR and energy dispersive X-ray analysis. The matrix derived from methacrylic acid and 2-methylimidazole is easily synthesized in one-spot strategy, and combines coordination properties with chemical resistance. The catalytic activity of this material on H₂O₂ activation was studied by electron spin resonance, which confirmed the release of superoxide radical. A possible mechanism of interaction involves the simultaneous production of dioxygen, protons and water. The catalytic performance was assessed in the activation of H₂O₂ for the oxidation of two representative organic compounds of environmental concern. About 70% of methyl orange, a model azo dye, was removed from distilled water in 2 h by oxidation with H₂O₂ 60 mM and by adsorption on the catalyst. The amount of adsorbed dye was minimized in the presence of 0.1 M Na₂SO₄. The kinetics of the processes followed a pseudo-first-order empirical model, and the catalyst was recovered and re-used. Epinephrine was chosen as a pharmaceutical model susceptible to superoxide attack. About 80% of conversion to adrenochrome was reached in less than 6 min, following a pseudo-first-order kinetic model.

Full text of DOI:10.1016/j.apcata.2013.08.002

Applied Catalysis A: General 467 (2013) 342–354
Contents lists available at ScienceDirect
Applied Catalysis A: General
journal homepage: www.elsevier.com/locate/apcata
Activation of H2O2 and superoxide production using a novel cobalt
complex based on a polyampholyte
Lucía V. Lombardo Lupanoa, Juan Manuel Lázaro Martínezb,c, Lidia Leonor Piehld,
Emilio Rubin de Celisd, Viviana Campo Dall’ Ortoa,e,∗
a IQUIFIB-CONICET, Junín 956 (C1113AAD), Ciudad Autónoma de Buenos Aires, Argentina
b Departamento de Química Orgánica, Facultad de Farmacia y Bioquímica, Universidad de Buenos Aires, Junín 956 (C1113AAD), Ciudad Autónoma de
Buenos Aires, Argentina
c IFEG-CONICET, Medina Allende s/n (5000), Córdoba, Argentina
d Departamento de Fisicomatemática, Facultad de Farmacia y Bioquímica, Universidad de Buenos Aires, Junín 956 (C1113AAD), Ciudad Autónoma de
Buenos Aires, Argentina
e Departamento de Química Analítica y Fisicoquímica, Facultad de Farmacia y Bioquímica, Universidad de Buenos Aires, Junín 956 (C1113AAD), Ciudad
Autónoma de Buenos Aires, Argentina
a r t i c l e i n f o
a b s t r a c t
Article history:
A new catalyst based on Co(II) and a hydrogel with property of polyampholyte was characterized by
equilibrium studies of Co(II) uptake, solid-state NMR and energy dispersive X-ray analysis. The matrix
derived from methacrylic acid and 2-methylimidazole is easily synthesized in one-spot strategy, and
combines coordination properties with chemical resistance. The catalytic activity of this material on
H2O2 activation was studied by electron spin resonance, which confirmed the release of superoxide
radical. A possible mechanism of interaction involves the simultaneous production of dioxygen, protons
and water. The catalytic performance was assessed in the activation of H2O2 for the oxidation of two
representative organic compounds of environmental concern. About 70% of methyl orange, a model azo
dye, was removed from distilled water in 2 h by oxidation with H2O2 60 mM and by adsorption on the
catalyst. The amount of adsorbed dye was minimized in the presence of 0.1 M Na2SO4. The kinetics of
the processes followed a pseudo-first-order empirical model, and the catalyst was recovered and re-
used. Epinephrine was chosen as a pharmaceutical model susceptible to superoxide attack. About 80%
of conversion to adrenochrome was reached in less than 6 min, following a pseudo-first-order kinetic
model.
Received 7 July 2013
Accepted 1 August 2013
Available online xxx
Keywords:
Heterogeneous catalysis
Cobalt complexes
Polyampholytes
Solid-state NMR
Hydrogen peroxide
© 2013 Elsevier B.V. All rights reserved.
1. Introduction
The advanced oxidation processes using reactive radicals are
recognized as effective ways for the degradation of organic pollut-
ants dissolved in water.
Abbreviations: AICc, Akaike information criterion; Ce, ion concentration at
equilibrium; CP-MAS, cross-polarization and magic angle spinning; CW ESR, con-
tinuous wave electron spin resonance; DMPO, 5,5-dimethyl-1-pyrroline-N oxide;
D–R, Dubinin–Radushkevich; ε, Polanyi potential; E, mean free energy of sorption
(D–R); E◦, standard reduction potential; EDS, energy dispersive X-ray spec-
troscopy; KF, Freundlich sorption constant; MO, methyl orange; pI, isoelectric point;
Poly(EGDE-MAA-2MI), polyampholyte derived from ethylene glycol diglycidyl ether
Hydrogen peroxide (H2O2) is an attractive reagent in detox-
ification of water polluted with chemicals such as pesticides,
phenols, hydrocarbons, surfactants, and dyes. The low environmen-
tal impact of this peroxide is based on the fact that water is the
only by-product. However it is a rather slow oxidizing agent in the
absence of activators [1,2].
In this regard, many different transition metal complexes have
been introduced as catalysts due to their ability to activate oxi-
dants. The oxygen radicals from H2O2 are efficiently generated
only after proper chelation of the cation involved in the catalytic
mechanism.
methacrylic acid and 2-methylimidazole; qe, adsorption capacity in equilibrium; qm
,
maximum amount of adsorbate that can be adsorbed in micropores (DR); SEM, scan-
ning electron microscopy; R2, regression coefficient; SOD, superoxide dismutase;
TR-XRF, X-ray fluorescence with total reflection; ꢀ, zeta potential.
∗
Corresponding author at: Departamento de Química Analítica y Fisicoquímica,
Facultad de Farmacia
y Bioquímica, Universidad de Buenos Aires, Junín 956
(C1113AAD), Ciudad Autónoma de Buenos Aires, Argentina. Tel.: +54 1149648263;
fax: +54 1149648263.
E-mail addresses: vcdall@ffyb.uba.ar, vcdall@yahoo.com.ar (V. Campo Dall’ Orto).
0926-860X/$ – see front matter © 2013 Elsevier B.V. All rights reserved.
http://dx.doi.org/10.1016/j.apcata.2013.08.002

L.V. Lombardo Lupano et al. / Applied Catalysis A: General 467 (2013) 342–354
343
The application of the transition metal cobalt in organic
2. Experimental
transformation has attracted increasing interest especially due
to its lower cost when compared with its higher homologues
[3].
2.1. Materials and reagents
Cobalt (II) appears to be a good oxygen transfer agent in oxidiz-
ing olefins. Reversible redox cycle between Co(II)–Co(III) oxidation
states causes oxygen atom transfer to the double bonds by cobalt
complexes [4–6].
2-Methylimidazole (2MI; 99 wt%), methacrylic acid (MAA;
99 wt%), ethylene glycol diglycidyl ether (EGDE; 50 wt% in ethylene
glycol dimethylether), epinephrine bitartrate and 5,5-dimethyl-1-
pyrroline-N oxide (DMPO) were purchased from Sigma–Aldrich.
Benzoyl peroxide was purchased from Fluka. Superoxide dismu-
tase (SOD) was purchased from Biosidus (Buenos Aires, Argentina).
Acetonitrile from Baxter was of HPLC grade.
CoSO4 and cobalt acetate were purchased from Mallincrodt
(USA). Na2SO4 and H2O2 (30 wt%) were acquired from Merck
(Darmstadt, Germany). Methyl Orange was from Santa Cruz
Biotechnology, Inc.
Besides, the combination of
a transition metal with a
hydrogel as ligand for heterogeneous catalysis is an inter-
esting approach which facilitates the recovery, regeneration
and recycling of the catalyst. In this way, the material
can be used in continuous processing, overcoming secondary
pollution.
In spite of the good performance of Co(II) catalysts, only a limited
number of complexes with polymeric non-soluble supports have
been presented.
Water was distilled with a FIGMAY glass apparatus (Córdoba,
Argentina). All other reagents were of analytical grade.
Cobalt complexes with modified crosslinked polyacrylamide
have been reported to be efficient heterogeneous catalysts for
selective oxidation of olefins and alkyl halides with H2O2 in aque-
ous media [7].
2.2. Instruments
Cobalt-phtallocyanines immobilized on activated carbon fibres
have been used as activators of H2O2 on the oxidative removal of
4-nitrophenol [8].
High-resolution 13C solid-state spectra for the polymers were
1
recorded using the ramp { H} → {13C} CP-MAS (cross-polarization
and magic angle spinning) sequence with proton decoupling during
acquisition. All the solid-state NMR experiments were performed
at room temperature in a Bruker Avance II-300 spectrometer
equipped with a 4-mm MAS probe. The operating frequency for
protons and carbons was 300.13 and 75.46 MHz, respectively.
Glycine was used as an external reference for the 13C spectra
and to set the Hartmann–Hahn matching condition in the cross-
polarization experiments. The recycling time was 4 s. Two different
contact times during CP were employed: 800 and 1500 ␮s. The
SPINAL64 sequence was used for heteronuclear decoupling dur-
ing acquisition with a proton decoupling field H1H satisfying
ω1H/2␲ = ꢁHH1H = 62 kHz [21]. The spinning rate for all the samples
was 10 kHz.
The surface of the polymers was characterized using scanning
electron microscopy (SEM) and the elemental composition of cobalt
and sulfur in different parts of the materials was performed by
energy dispersive X-ray spectroscopy (EDS). SEM imaging and EDS
were carried out with a scanning electron microscope Field Emis-
sion SEM (Zeiss Gemini DSM 982) operated at a 0.3 kV acceleration
voltage and an INCA Energy (Oxford Instrument), respectively.
The electron spin resonance (ESR) spectra were obtained at 20 ◦C
using an X-band ESR Spectrometer Bruker EMX plus (Bruker Instru-
ments, Inc., Berlin, Germany).
In this work we tested the synthetic polyampholyte Poly(EGDE-
MAA-2MI) as ligand for Co(II). The particular relevance of this
material is that it combines the benefits of synthetic polymers,
i.e. the high adsorption capacity, chemical stability, low costs
and simplicity of the synthetic procedure. This polyampholyte is
synthesized in a one-step procedure and bears carboxylic and 2-
methylimidazole residues, which endows it coordination capacity
[9–12].
Here, we studied the catalytic activity of this coordination com-
plex on H2O2 activation. The polyampholyte has already been
successfully applied on the development of a catalyst based on
Cu(II) for the activation of this oxidant [13]. Nevertheless, the Co(II)-
derived catalyst was expected to activate H2O2 by means of a
different mechanism.
In order to explore the applicability of the Co(II)-Poly(EGDE-
MAA-2MI)catalyst in the environmental and industrial fields, we
chose two different types of organic substrates: methyl orange (a
model compound of synthetic dyes), and epinephrine (a model
compound of pharmaceuticals).
Methyl orange (MO) is used as pH indicator in 0.1% aqueous
solution, as well as in dyeing and printing textiles. It is considered
a hazardous substance according to OSHA 29 CFR 1910.1200. It is
chemically stable and both, its persistence in water and soil and its
mobility are high [14].
Partial pressure of dioxygen was measured with an Orbisphere
A1100 – Oxygen Sensor.
The degradation products of MO by the action of highly reactive
oxygen-containing species are well described. The photo-Fenton
degradation of MO produces 4-dimethylamino aniline, suggesting
an N N bond cleavage of the azo linkage [15]. The hydroxylated and
non-hydroxylated species have been identified by Baiocchi et al.
[16] and He et al. [17].
UV–visible spectrophotometric measurements were made on a
Hewlett-Packard instrument, HP 8452A model with diode array.
Cobalt ions released from the complex to the solution were
analysed with a BRUKER X-ray fluorescence spectrometer with
geometry of total reflection (TR-XRF), PICOFOX S2 model.
Parshetti et al. studied the phytotoxicity and microbial toxic-
ity of the dye and some degradation products [18] and found that
MO is toxic and that the metabolites obtained, such as 4-amino
sulfonic acid and N,N-dimethyl p-phenyldiamine, are nontoxic for
experimental plants (Triticum aestivum and Phaseolus mungo) and
bacteria (Kocuria rosea, Pseudomonas aurugenosa and Azatobacter
vinelandii).
2.3. Cobalt uptake
Cobalt binding studies were performed with 0.0500 g of polyam-
pholyte and 4.0 mL of CoSO4 solution in a 2.6–21 mM concentration
range. The samples were centrifuged and filtered after 48 h of con-
tact time at 24 ◦C. Free Co(II) concentration at the equilibrium in
each supernatant solution could be determined spectrophotomet-
rically since Co2+ forms a coordination compound with SCN− in
presence of HCl and acetone, which absorbs radiation at 622 nm
[22].
Epinephrine is
a catecholamine, susceptible to peroxide
attack. It can be partially oxidized by superoxide to the
industrial product adrenochrome (catecholamine o-quinone),
a
precursor of haemostatic drugs such as carbazochrome
[19,20].

344
L.V. Lombardo Lupano et al. / Applied Catalysis A: General 467 (2013) 342–354
2.4. Model fitting by non-linear regression
The spectrometer settings were: centre field: 3515 G, sweep
width: 100 G, microwave power: 10 mW, microwave frequency:
9.85 GHz, conversion time: 2.56 ms, time constant: 2.56 ms, mod-
ulation frequency: 50 KHz, modulation amplitude: 0.11 G, gain:
2.00 × 104, resolution: 1024 points. All spectra were the accumula-
tion of 20 scans. DMPO adduct signal intensities were measured as
the total height of the first field peak in the respective ESR spectra.
The ESR spectra were simulated by means of an Electron Spin
Resonance Simulator using for the fitting the Levenberg–Marquardt
algorithm. Molar proportion yields and hyperfine interaction con-
stant values were calculated from the simulated spectra.
The isotherm and kinetic parameter sets were determined by
non-linear regression. The algorithm based on the Gauss–Newton
method was used. The error function employed to evaluate the fit
was the second order corrected Akaike information criterion (AICC)
[23]. The Akaike weights provide information about the strengths
of evidence supporting the two competing models. The evidence
ratio rising from the Akaike criterion represents the relative like-
lihood favouring the better of two competing models. As reported
by other authors, an evidence ratio greater than 20, would indicate
extremely strong evidence favouring the better model [24,25].
2.8. Methyl orange decolorization
2.5. Co(II)-Poly(EGDE-MAA-2MI) complex preparation
An amount of 0.1000 g of Co(II)-Poly(EGDE-MAA-2MI) was sus-
pended in 100 mL of 41 1 ␮M MO and 63 mM H2O2 solution.
The medium of reaction was alternatively distilled water or 0.1 M
Na2SO4 solution. The absorbance of the solution was monitored as a
function of time at 464 nm. For turbidity correction, the absorbance
was measured at 700 nm and subtracted to the absorbance at
464 nm.
The complex was prepared mixing 0.5000 g of polyampholyte
powder and 40 mL of 70 mM CoSO4 solution in distilled water at
24 ◦C for 48 h. The samples were then centrifuged, washed with
three 20-mL portions of distilled water to remove the Co(II) weakly
adsorbed on the particles, filtered, dried at 60 ◦C and milled in a
mortar. It must be mentioned that the last portion of water from
the wash step did not present detectable amounts of free Co(II) in
solution, when it was tested by the colorimetric method described
in Section 2.3.
The control of MO adsorption on the catalyst was made in
absence of H2O2.
The control of H2O2 activation in homogeneous phase was made
with a 0.35 mM CoSO4 solution, 41 ␮M MO, 63 mM H2O2 and 0.1 M
Na2SO4. This control was made to determine if any free and soluble
Co(II) species could interact with H2O2 and generate free radicals
in solution.
For the evaluation of the chemical stability of the catalyst on
reutilization, the experiment of MO decolorization was repeated
three times with the same recycled catalyst and new aliquots of
solution. Each cycle had an extension of 30 min, and was carried
out in 0.1 M Na2SO4 solution. After each catalytic cycle, the sus-
pension was centrifuged, the supernatant was discarded, the solid
was washed twice with 0.1 M Na2SO4 and with a portion of distilled
water, and then it was dried at 40 ◦C for a new use.
In order to explore the regeneration of the catalyst used for
the first catalytic cycle, the solid was recovered by centrifugation,
washed with three portions of distilled water and dried at 40 ◦C.
Then 0.1000 g of the used catalyst were reloaded with 8 mL of
70 mM CoSO4 solution in distilled water at 24 ◦C for 24 h. The mix-
ture was centrifuged, washed with three 20-mL portions of distilled
water to remove the Co(II) weakly adsorbed on the particles, fil-
tered, dried at 40 ◦C and milled in a mortar to be used in a new
catalytic cycle.
Another control dealt with the chemical stability of the MO
molecules adsorbed on the particles. An aliquot of 0.2000 g of Co(II)-
Poly(EGDE-MAA-2MI) was suspended in 200 mL of 41 1 ␮M MO
solution in 0.1 M Na2SO4, the suspension was centrifuged, the solid
was washed with three portions of distilled water and dried at 40 ◦C.
An amount of 0.1000 g of the MO-loaded Co(II)-Poly(EGDE-MAA-
2MI) was put in contact with 100 mL of a 63 mM H2O2 solution in
0.1 M Na2SO4, which had been previously activated with the cat-
alyst during 10 min. The experiment was repeated with another
0.1000-g aliquot of the MO-loaded Co(II)-Poly(EGDE-MAA-2MI)
and the 63 mM H2O2 solution in 0.1 M Na2SO4 without previous
activation. In both cases, we looked for a decrease of the colour
intensity of the particles.
2.6. Measurement of H2O2 concentration
This peroxide reacts with phenol and 4-aminoantipyrine in
presence of soybean peroxidase, giving a product that absorbs radi-
ation at 505 nm [26]. The absorbance of the mixture is proportional
to the concentration of H2O2 when this peroxide is the reagent in
defect. The reaction is presented in Scheme A.1 (Supplementary
content).
In this way, a solution of the 4-aminoantipyrine reagent was
prepared with 9 mg of 4-aminoantipyrine, 200 mg of phenol and
0.1 M phosphate buffer at pH 7.0 to a final volume of 10 mL.
An amount of 0.0500 g of Co(II)-Poly(EGDE-MAA-2MI) was sus-
pended in 50 mL of 42 mM H2O2 and 0.1 M Na2SO4 solution. The
concentration of H2O2 was monitored as a function of time from
20-␮L samples of the reaction, which were mixed with 60 ␮L of
300 U mL−1 of soybean peroxidase extract, 100 ␮L of the solution
of 4-aminoantipyrine reagent and 480 ␮L of 0.1 M phosphate buffer
at pH 7.0 [26].
The salt, Na2SO4, was added to most of the solutions in order to
minimize the adsorption of organic compounds on the surface of
the catalyst.
2.7. Measurement of free radicals by ESR
The initial H2O2 concentration was set close to 60 mM. Higher
concentration levels of H2O2 were expected to be less efficient in
oxidative processes due to possible deleterious effect on the poly-
meric catalyst [13].
An amount of 0.0500 g of Co(II)-Poly(EGDE-MAA-2MI) was sus-
pended in 50 mL of 63 mM H2O2 solution and the production of
free radicals was followed by ESR. The medium of reaction was
alternatively distilled water or 0.1 M Na2SO4 solution.
For a control experiment, we used a Co(II)-Poly(EGDE-MAA-
2MI) complex obtained from cobalt acetate instead of the original
Co(II)-Poly(EGDE-MAA-2MI) complex obtained from CoSO4.
The experiment was also performed in presence of SOD with an
activity of 200 U mL−1, added to the system before of H2O2.
Aliquots of 32 ␮L taken at different times of reaction were mixed
with 16 ␮L of 3 M DMPO (spin trap), and the continuous wave (CW)
ESR spectra of the DMPO spin adducts were recorded at 20 ◦C 3 min
after the end of incubation.
For the detection of cobalt intermediate species catalytically
active, 0.1000 g of Co(II)-Poly(EGDE-MAA-2MI) were put in contact
with 100 mL of 63 mM H2O2 and 0.1 M Na2SO4 solution during
10 min. The suspension was centrifuged, the supernatant was dis-
carded, the particles were washed with three portions of distilled
water, and they were not dried. The activated particles of the
catalyst were put in contact with 100 mL of 41 1 ␮M MO solution
in 0.1 M Na2SO4 and the MO concentration was monitored as a

L.V. Lombardo Lupano et al. / Applied Catalysis A: General 467 (2013) 342–354
345
Scheme 1. Representative chemical structure of Poly(EGDE-MAA-2MI).
function of time by visible spectrophotometry. The activation of
the solid with H2O2 was repeated twice, and the catalyst was put
in contact with the same MO solution after each activation cycle
in order to monitor the concentration of the dye.
2.9. Epinephrine oxidation
An amount of 0.0400 g of Co(II)-Poly(EGDE-MAA-2MI) was sus-
pended in 20 mL of 2 mM epinephrine and 46 mM H2O2 solution
(higher concentrations of H2O2 were proved to be less effective).
The medium of reaction was 0.1 M Na2SO4 solution. The absorbance
of the solution was monitored as a function of time at 480 nm. For
turbidity correction, the absorbance was measured at 700 nm and
subtracted to the absorbance at 480 nm.
A control experiment of epinephrine oxidation in homogeneous
system was made with a solution of 2 mM epinephrine and 46 mM
H2O2 in 0.1 M Na2SO4.
Another control experiment was made with 0.0400 g of Co(II)-
Poly(EGDE-MAA-2MI) suspended in 20 mL of 2 mM epinephrine
and 0.1 M Na2SO4, in absence of H2O2. This experiment was
repeated with previous and continuous bubbling of argon in the
epinephrine solution, in order to displace the dissolved O2.
Fig. 1. Adsorption isotherm of Co(II) onto the polyampholyte at 24 ◦C. Experimental
data fitted to the D–R model.
3. Results and discussion
3.1. Synthesis and characterization of the polyampholyte
The best fit of experimental adsorption data was obtained with
the D–R (Fig. 1), Temkin and Freundlich isotherms (Supplemen-
tary content, Fig. A.1), according with the evidence ratio based on
the Akaike criterion [25]. The Langmuir model had no statistical
support due to its significantly larger residual sum of squares.
The Freundlich model can be described as:
The material was synthesized via reaction of the selected
monomers in the presence of benzoyl peroxide. Scheme 1 shows
the representative chemical structure for Poly(EGDE-MAA-2MI).
The mechanism of synthesis and the chemical structure of this
material have been studied by means of a variety of spectroscopic
methods, and carefully described in previous reports [9–12]. Here,
we present the characterization of the polyampholyte production
batch used for these experiments [25].
qe = KF(C1/n
)
(1)
e
The material is synthesized as a monolith, and then milled into
particles. The isoelectric point (pI) for this material determined
by the measurements of ꢀ (zeta potential) was 6.77. The apparent
diameter of the particles was 64 4 ␮m [27].
where qe is the adsorption capacity in equilibrium with the corre-
sponding Ce (mmol L−1), which is the concentration of metal ion.
This model assumes that the surface is heterogeneous in the
sense that the adsorption energy is distributed and the surface
topography is patchwise [28]. The sites with the same adsorption
energy are grouped together into one patch (the adsorption energy
here is the energy of interaction between adsorbate and adsorbent).
Each patch is independent of each other (i.e., there is no interaction
between patches) and the Langmuir equation is applicable for the
description of equilibrium of each patch.
These particles have 0.89 mmol g−1 of titrable carboxylate and
1.09 mmol g−1 of titrable 2MI residues, besides the hydroxyl groups
(Scheme 1). The pI calculated from the titrable functional groups
(6.75) coincided with the value obtained from the ꢀ measurements.
3.2. Cobalt adsorption equilibrium
The mechanism of cobalt uptake was explored analysing the
results of the adsorption isotherm for the polyampholyte at 24 ◦C.
Four isotherm models (Langmuir, Dubinin–Radushkevich (D–R),
Freundlich and Temkin) were used to fit the experimental adsorp-
tion data by non-linear regression.
KF can be defined as the sorption or distribution coefficient and
represents the amount of Co(II) sorbed onto the polymer surface
normalized by the equilibrium concentration and 1/n is a mea-
sure of surface heterogeneity, ranging between zero and one which
becomes more heterogeneous as its value gets closer to zero [29].

346
L.V. Lombardo Lupano et al. / Applied Catalysis A: General 467 (2013) 342–354
Table 1
Isotherm parameters for Co(II) adsorption on the polyampholyte. Regression coefficients (R2), values of AICC criterium and evidence ratio for Freundlich, Temkin,
Dubinin–Radushkevich (D–R) and Langmuir models.
Freundlich
Temkin
Dubinin–Radushkevich
qm (mmol g−1
0.521
Langmuir
n
5.87 S.D.: 0.30
0.2059S.D.: 0.0031
bT (kJ mol−1
)
61.1S.D.: 2.8
181 S.D.: 48
)
S.D.: 0.018
qm, (mmol g−1
KL
)
0.298S.D.: 0.013
0.332S.D.: 0.067
KF (mmol g−1
)
KT (mmol L−1
)
B (mol2 kJ−2
)
1.60 × 10−3S.D.:7 × 10−5
E (kJ mol−1
R2
)
17.68
0.9754
−137.4
1
S.D.: 0.39
R2
AICC
Evidence ratio
0.9666
−133.0
8.6
R2
AICC
Evidence ratio
0.9731
−134.6
4.0
R2
AICC
Evidence ratio
0.815
AICC
Evidence ratio
−107.3
3.2 × 106
The Temkin model assumes that the heat of adsorption (which is
a function of temperature) of all molecules in the layer will decrease
linearly rather than logarithmically with coverage [30].
3.3. SEM and EDS analysis
Co(II)-Poly(EGDE-MAA-2MI) was characterized using SEM in
order to evidence surface morphological changes when compared
with the polyampholyte. The SEM images are depicted in Fig. 3A
and B. The surface of Co(II)-Poly(EGDE-MAA-2MI) exhibited a sig-
nificant difference in homogeneity degree. The image also suggests
a small increase in surface area.
RT
qe
=
ln(KTCe)
(2)
bT
The Dubinin–Radushkevich isotherm is a semi-empirical equa-
tion which was originally developed for sub-critical vapours in
microporous solids, where the adsorption process follows a pore-
filling mechanism:
The identity of Co2+ counterion was manifested in EDS
(Fig. 3C–E). The sulfur in the sample was present when CoSO4 was
used to obtain the complex and absent in complexes derived from
2
ε
qe = qme−B
(3)
D
2−
cobalt acetate. The presence of SO4
evidences the existence of
fixed positive charges on the surface of the particles, coming from
the Co(II) and/or NH+ from the 2MI residues. SO42− could eventually
be exchanged by other anions, as discussed below.
where qm is the maximum amount of adsorbate that can be
adsorbed in micropores, and BD is a constant related to the energy.
The value estimated for qm was 30.7 mg of Co(II) per gram of
polyampholyte, significantly higher than that predicted by Lang-
muir model (17.6 mg g−1), summarized in Table 1. This loading
capacity for Co(II) was lower than the capacity for Cu(II), estimated
in 0.90 mmol or 57 mg g−1 [10].
Liquid-phase adsorption data can also be analysed by the equa-
tion below, where the amount adsorbed corresponding to any
adsorbate concentration is assumed to be a Gaussian function of
the Polanyi potential (ε),
3.4. Solid-state NMR analysis
Solid-state NMR spectroscopy was used to characterize the
Co(II)-complexes to obtain an approximation of the ligands
involved in the coordination sphere of the metal ion in the poly-
mer matrix, since the paramagnetic ions produce a strong source
of relaxation. In particular, we have previously observed that the
carbon signal of the carboxylic acid in the 13C CP-MAS is a sensitive
parameter to sense the union of the paramagnetic Cu2+ ion to this
ligand (RCO2H) [11].
ꢀ
ꢁ
1
Ce
ε = RT ln 1 +
.
(4)
The 13C CP-MAS spectra for the Poly(EGDE-MAA-2MI) poly-
mer and its cobaltous complexes with 8.7 and 30.7 mg of Co2+ per
gram of polymer material are shown in Fig. 4. The most signifi-
cant changes were observed with the highest amount of adsorbed
Co2+, where the signals of the carbons from the imidazole ring
(C7–8) and the carboxylic acid (C1) were completely depleted in
the 13C CP-MAS spectra. This behaviour indicates that the para-
magnetic centres of Co2+ strongly interacted with these ligands,
leading to the broadening and disappearance of the resonance sig-
nals at 30.7 mg of Co2+ per gram of polymer. For that reason, the
contact time with the cross-polarization, in the 13C CP-MAS experi-
ments, was reduced to 800 ␮s, in comparison with the 1500 ␮s used
in the other 13C CP-MAS spectra. Additionally, since the polymer
matrix did not present positive permanent domains (i.e. absence
R, T and Ce represent the gas constant (8.314 × 10−3 kJ mol−1 K−1),
absolute temperature (K), and ion concentration at equilibrium (g
ion per gram of solution), respectively [31].
The equation is generally applied to express the adsorption
mechanism with a Gaussian energy distribution onto a heteroge-
neous surface. The approach is usually applied to distinguish the
physical and chemical adsorption of metal ions by means of the
equation
1
E =
.
(5)
ꢂ
2BD
The parameter E is the mean free energy of sorption [32], whose
magnitude is a way to estimate the type of sorption process.
In the case E is lower than 8 kJ mol−1, weak physical forces, such
as van der Waals and hydrogen bonding, may affect the sorption
mechanism. If this value is between 8 and 16 kJ mol−1, the sorption
process can be explained by ion exchange. If this value is higher
than 16 kJ mol−1, the sorption process can be explained by other
chemical reactions such as coordination [33].
The E value and other parameters calculated for all the mod-
els tested are summarized in Table 1. The E value higher than
16 kJ mol−1 K−1 suggests the coordination of Co(II) with the OH,
COO− and/or 2MI residues of the polyampholyte. The deep vio-
let colour of the particles of Co(II)-poly(EGDE-MAA-2MI) is another
evidence of coordination (Fig. 2).
For further experiments, we prepared a Co(II)-Poly(EGDE-MAA-
2MI) complex with a load of 30.7 mg Co(II) per gram estimated from
the isotherm fit.
Fig. 2. Picture of Poly(EGDE-MAA-2MI) (left) and Co(II)-Poly(EGDE-MAA-2MI)
(right).

L.V. Lombardo Lupano et al. / Applied Catalysis A: General 467 (2013) 342–354
347
Fig. 3. SEM images for the surface of Poly(EGDE-MAA-2MI) (A) and Co(II)-Poly(EGDE-MAA-2MI) (B) at a magnification of 2K×. EDS spectra of Poly(EGDE-MAA-2MI) (C),
Co(II)-Poly(EGDE-MAA-2MI) obtained from CoSO4 (D) and Co(II)-Poly(EGDE-MAA-2MI) obtained from cobalt acetate (E).
of N1,N3-disubstitution of the imidazole moieties), as determined
from the potentiometric titration and zeta-potential measures, we
observed a complete reduction of the imidazole signals, making a
difference with a Cu(II)-complex where the polymer matrix con-
tains 45% of N1,N3-disubstitution of the imidazole units [11]. At
3.5. H2O2 activation
We have previously demonstrated that Poly(EGDE-MAA-2MI)
has no catalytic activity on H2O2 activation [13]. Now, the Co(II)-
Poly(EGDE-MAA-2MI) complex was put in contact with 42 mM
H2O2 in 0.1 M Na2SO4, and the concentration of H2O2 was mon-
itored as a function of time (Supplementary content, Fig. A.2). The
reaction followed a pseudo-first-order kinetics, in agreement with
the evidence ratio based on the Akaike criterion:
low levels of adsorbed cobaltous ion (8.7 mg Co2+ −1), the effect
g
of the ion was slightly observed in the intensity of the signals of C9
and C10. Although this change might be attributed to modifications
in the reticulation of the materials, the signals of the ligands were
not perturbed (C1–7–8). That made a difference with some Cu(II)-
complexes where the signal of C1 was specially affected with 8 mg
of copper per gram of polymer [11], having into account that the
Cu2+ ions have a higher reduction effect in the relaxation times
of 1H than Co2+ ions [34]. With these results, the cobaltous ions
might be coordinated with the oxygens of the carboxylic acid and
the nitrogens of the imidazole moieties at 30.7 mg of Co2+ per gram
of polymer.
[H2O2]t = [H2O2] + [H2O2]deg × e−k×t
(6)
∞
with 37.53 mM [H O ] (S.D.: 0.30), 4.63 mM [H O ]deg (S.D.: 0.45)
∞
2
2
2
2
and 0.076 min−1 k (S.D.: 0.018); R2 was 0.9642.
A fast decrease of H2O2 concentration was observed during the
first 20 min of reaction, followed by an asymptotic approach to
37.53 mM.
When the initial amount of H2O2 was increased to 63 mM, the
changes of concentration as a function of time were less significant
due to the large excess of this oxidant. Nevertheless, the activation
of the peroxide was evidenced by the set of experiments described
below.
Scheme 2 shows the possible chemicals structures of the Co(II)-
Poly(EGDE-MAA-2MI) complexes. The most probable geometries
are square-planar and octahedral, with the cobaltous ion coordi-
nated by different combinations of COO− and 2MI groups.
The appearance of a certain amount of bubbles suggested the
formation of dioxygen as a possible product of H2O2 activation.
The partial pressure of the gas in the system containing the catalyst
was measured before and after the addition of H2O2. The increase
by 66% in the O2 partial pressure confirmed the formation of molec-
ular oxygen during the first 5 min of the reaction (Supplementary
content, Fig. A.3). Besides, the pH decreased from 6.0 to 4.8 during
the reaction.
The evolution of gas bubbles from the surface of the particles
of Co(II)-Poly(EGDE-MAA-2MI) as a consequence of the addition of
H2O2 and only in the presence of immobilized Co(II), clearly evi-
dences that the activation of hydrogen peroxide does not involve a
homogeneous catalysis.
The production of free radicals that diffuse into the solution was
confirmed by spin trapping experiments with DMPO. Fig. 5A shows
the CW ESR spectrum produced by aliquots of supernatant from the
Fig. 4. 13C CP-MAS spectra for Poly(EGDE-MAA-2MI) (A) and for its Co(II)-complexes
containing 8.7 (B) and 30.7 mg of Co2+ per gram of polymer (C). The numbering
corresponds to that in Scheme 1.

348
L.V. Lombardo Lupano et al. / Applied Catalysis A: General 467 (2013) 342–354
Scheme 2. Most probable geometries of the coordination complex Co(II)-poly(EGDE-MAA-2MI).
Table 2
Co(II)-Poly(EGDE-MAA-2MI)/63 mM H2O2 heterogeneous system,
in the presence of DMPO spin trap measured beyond 10 min after
the addition of H2O2 in 0.1 M Na2SO4.
The identity of the free radicals had to be confirmed by
additional experiments. According with the profile of the signal,
Hyperfine interaction constant values for the DMPO/OOH and the DMPO/OH
adducts. The data were obtained by simulation and fits of the experimental spectra.
Radical
aN (G)
aH␤ (G)
aHꢁ (G)
DMPO/OOH
DMPO/OH
13.8
14.7
11.0
13.8
Not resolved
–
•
−
the probable adducts were: DMPO/OSO3H from SO4
[35] and
•
•
DMPO/OOH from O2 [36]. The first option, SO4 −, was consid-
ered the least probable due to the expected higher E◦ of this free
radical, when compared with the E◦ of OH• (+2.8 V vs HNE).
Then, the experiment was repeated in distilled water instead of
0.1 M Na2SO4, in order to evaluate the reactivity of SO42− ions. The
ESR spectrum did not change significantly (Supplementary content,
Fig. A.4B) when compared with Fig. 5A. It must be pointed out that
the kinetics of the free radical indicated a maximal production at
10 min of reaction (Supplementary content, Fig. A.3), reaching a
constant concentratio•n level beyond the 20 min.
−
The molar fractions of DMPO/OOH and DMPO/OH adducts were
0.88 and 0.12 respectively.
Fig. 6 shows simulated DMPO/OOH and DMPO/OH adducts sep-
arated from each other. The simulated spectrum of Figs. 5B and 6C
was obtained by the addition of the above mentioned adducts.
It should be noted that the Hꢁ hyperfine interaction of the
DMPO/OOH adduct was not resolved. Table 2 shows the hyper-
fine interaction constant values for the DMPO/OOH and DMPO/OH
adducts.
The presence of the DMPO/OOH adduct was also confirmed by
the addition of SOD with an activity of 200 U mL−1. In this case
we expected the dismutation of the superoxide radical and the
absence of an ESR signal. Instead, we found a spectrum with the
characteristic shape of hydroxyl radical adduct.
Fig. 7A shows the CW ESR spectrum obtained when the experi-
ment was performed with the Co(II)-poly(EGDE-MAA-2MI)/63 mM
H2O2 system in the presence of DMPO and SOD, in distilled water.
The simulation (Fig. 7B) and fits of the experimental spectrum
allowed establishing the presence of DMPO/OH. The same result
−
2−
species could still arise from the SO4
However, the SO4
counterions of the positive sites on Co(II)-Poly(EGDE-MAA-2MI)
particles. So, the experiment was repeated in distilled water
with the Co(II)-poly(EGDE-MAA-2MI) catalyst obtained from cobalt
acetate salt instead of CoSO4 (i.e. the absence of sulfur had been con-
firmed by EDS, Fig. 3E). Again, the same profile in the ESR spectrum
was obtained eve•n in the absence of SO42−, thus discarding the
−
production of SO4 radicals (Supplementary content, Fig. A.5B).
•
When considering the second option, O2 −, we found that the
simulation and fits of the experimental spectrum from Fig. 5A
allowed establishing the presence of two species: DMPO/OOH and
•
DMPO/OH adducts from O2 and OH• , respectively (Fig. 5B) [37].
−
Fig. 5. Experimental (A) and simulated (B) CW ESR in band X spectra from spin trapping experiments with DMPO. Heterogeneous system: Co(II)-Poly(EGDE-MAA-2MI)/63 mM
H2O2 in 0.1 M Na2SO4. The simulation and fits of the experimental spectrum allowed establishing the presence of two species: a DMPO/OOH and a DMPO/OH adducts. The
molar fractions of DMPO/OOH and DMPO/OH adducts were 0.88 and 0.12 respectively.

L.V. Lombardo Lupano et al. / Applied Catalysis A: General 467 (2013) 342–354
349
A
B
C
Fig. 6. Simulated DMPO/OOH adduct (A), simulated DMPO/OH adduct (B) and simulated spectrum (C) obtained by the addition of (A) and (B). It should be noted that the Hꢁ
hyperfine interaction of DMPO/OOH adduct was not resolved.
was obtained when the experiment was performed in the presence
of SOD in 0.1 M Na2SO4 (Supplementary content, Fig. A.6).
•
HO2• ꢀ O2 − + H+
Summarizing, the identity of• the ESR signal was confirmed:
−
the DMPO/OOH adduct from O2
was generated by the Co(II)-
H
+ + OH− ꢀ H2O
poly(EGDE-MAA-2MI)/63 mM H2O2 system, and was absent in the
presence of SOD due to a fast dismutation of the radical.
Based on this experimental evidence, we propose two mecha-
nisms of H2O2 activation catalysed by immobilized Co(II).
The stage of OH• degradation must be the slowest in the presence
of the enzyme SOD, which would explain the observation of the
DMPO/OH signal.
Here we found a significant difference in the mechanism
exhibited by a related catalyst, Cu(II)-Poly(EGDE-MAA-2MI). This
copper-based material led to the release of OH• species on the acti-
vation of H2O2 [13]. Besides, the cation suffered a reduction to Cu(I)
in the catalytic process [38].
The most probable mechanism of activation involves the pro-
•
−
duction of O2, O2 and H+ according with [36]:
•
−
L-Co(II) + 2H2O2 ꢀ L-Co(III) + 2H2O + O2
2L-Co(II) + H2O2 ꢀ 2L-Co(III) + O2 + 2H+
Other speculative mechanism of H2O2 activation compatible
with the production of O2 and radicals would be the oxidation of
Co(II)-Poly(EGDE-MAA-2MI) by H2O2 as follows [39]:
where L-Co(II) represents Co(II)-Poly(EGDE-MAA-2MI).
In order to explain the occurrence of the signal attributed to
the DMPO/OH adduct as a minor product of the reaction and as the
main product after the addition of SOD, the first step should be split
as follows:
L-Co(II) + 2H2O2 ꢀ HO2 + OH• + [(H2O)L-Co(III)]+
•
•
HO2• ꢀ O2 − + H+
L-Co(II) + H2O2 ꢀ L-Co(III) + OH− + OH•
OH• + H2O2 ꢀ HO2• + H2O
2[(H2O)L-Co(III)]+ + H2O2 ꢀ 2L-Co(II) + H2O + O2 + 2H+
Fig. 7. Experimental (A) and simulated (B) CW ESR in band X spectra from spin trapping experiment with DMPO. Heterogeneous system: Co(II)-Poly(EGDE-MAA-2MI)/63 mM
H2O2 with 200 U mL−1 SOD, in distilled water. The simulation and fits of the experimental spectrum allowed establishing the presence of DMPO/OH. It should be noted that
the DMPO/OOH adduct is not present.

350
L.V. Lombardo Lupano et al. / Applied Catalysis A: General 467 (2013) 342–354
Table 3
Parameters estimated for the changes in MO concentration due to adsorption on
Co(II)-Poly(EGDE-MAA-2MI), from the empirical model presented in Eq. (7). The
adsorption was tested in two different media (0.1 M Na2SO4 and distilled water),
using 0.1000 g of the Co(II)-Poly(EGDE-MAA-2MI) complex in 100 mL of 41 1 ␮M
MO solution.
MO adsorption in 0.1 M Na2SO4
MO adsorption in water
2.51
S.D.: 0.10
0.246
S.D: 0.022
37.19
S.D.: 0.05
22.59
S.D.: 0.38
0.1011
S.D.: 0.0035
20.80
S.D.: 0.36
0.0007
[MO]ads (␮M)
kads (min−1
[MO]ads 1 (␮M)
)
kads 1 (min−1
[MO]ads 2 (␮M)
)
[MO] (␮M)
∞
kads 2 (min−1
)
S.D.: 0.0002
0.9982
R2
0.9545
R2
Scheme 3. Chemical structures of MO, epinephrine and adrenochrome.
was slowly uptaken by the matrix. The empirical kinetic model
involves two terms describing the decrease in MO (Table 3).
The dissolved Na2SO4 clearly inhibited the adsorption of MO on
the particles, since only 6.3% of the initial amount was withdrawn
from the solution. The empirical model for this medium includes
the amount of non-adsorbed dye (Table 3).
As a second approach, we explored the effect of the catalytic
activation of H2O2 on the azo dye concentration (Fig. 8). The initial
amount of H2O2 was set in 63 mM to optimize the efficiency in
terms of degradation.
Without the polyampholyte, the homogeneous 0.35 mM
Co(II)/63 mM H2O2 system oxidized less than 2% of the dye in 1 h.
With these results we concluded that the amount of free radicals
generated by free and soluble Co(II) species from the interaction
with H2O2, was negligible.
The concentration of MO in the Co(II)-Poly(EGDE-MAA-
2MI)/H2O2 system exhibited an exponential decay according to the
empirical kinetic expression (8) in 0.1 M Na2SO4 and (9) distilled
water. The parameters estimated are presented in Table 4
This second possible mechanism predicts a molar fraction equal
to 0.5 for each radical specie (HO2• and OH• ), which is not consistent
•
with our experimental evidence of a O2 − preponderance (Fig. 5B).
Agboola et al. presented spectroscopic evidence of the formation
of cobalt intermediate species with catalytic activity ([(HOO)L-
Co(III)]), when Co(II)-based homogeneous catalysts were oxidized
by H2O2 [39,40]. In our heterogeneous system, both the [(HOO)L-
Co(III)] intermediate and the radicals could eventually react with
oxidizable compounds.
It must be taken into account that the surface of our catalyst
is heterogeneous in terms of chemical composition. The differ-
ent probable combinations of ligands on the coordination sphere
of Co(II) in this heterogeneous catalyst would explain some het-
erogeneity in redox potentials of the active sites and thus the
coexistence of both proposed mechanisms.
3.6. Decolorization of methyl orange
×t
×t
[MO]t = [MO]ads × e−k
+ [MO]deg × e−k
(8)
(9)
ads
deg
The oxidative activity of the free radicals from activated H2O2
was tested using the azo dye methyl orange (MO) as a model
compound (Scheme 3). This dye was selected because of its low ten-
dency to adsorb on the polyampholyte and its known susceptibility
towards the advanced oxidation processes.
×t
×t
[MO]t = [MO]ads × e−k
+ [MO]deg × e−k
+ [MO]
ads
deg
∞
The experiments for removal of MO were carried out in distilled
water and in 0.1 M Na2SO4 solution. Na2SO4 was used to minimize
the dye adsorption on the catalyst, since this negatively-charged
organic compound is electrostatically attracted by the positive sites
on the polyampholyte surface. Phosphate, chloride and nitrate were
discarded since each of them affected the efficiency of the overall
process.
As a first approach, we studied the adsorption of MO on the
catalyst.
The kinetic profiles of MO removal in the presence of Co(II)-
poly(EGDE-MAA-2MI) and in the absence of H2O2 were compared
for the two media (Supplementary content, Fig. A.7). The experi-
mental data were fitted to the following empirical model:
×t
[MO]t = [MO]∞ + [MO]n × e−k
(7)
n
where [MO] is the amount of non-adsorbed dye, [MO]n is the
∞
amount of dye removed by adsorption, and kn is the pseudo-
first-order kinetic constant for the exponential decay of the dye
concentration from the nth term of the equation. The parameters
estimated for the different data sets are presented in Table 3.
In the absence of H2O2, MO was adsorbed on Co(II)-Poly(EGDE-
MAA-2MI) from an initial dye concentration of 41 1 ␮M, to an
extent that depended on the composition of the medium. In dis-
tilled water, 52% of MO was removed in less than 1 h, and the rest
Fig. 8. MO concentration profile in the heterogeneous system, which consisted on
0.1000 g of the Co(II)-Poly(EGDE-MAA-2MI) complex in 100 mL of 41 1 ␮M MO and
63 mM H2O2 solution. The reaction media were: 0.1 M Na2SO4 (ꢁ) and distilled water
(ꢂ). The catalyst was reloaded with Co(II) after the first cycle, and the experiment
was repeated in 0.1 M Na2SO4 (♦). The homogeneous system was also tested with
41 ␮M MO, 0.35 mM Co(II), 63 mM H2O2 and 0.1 M Na2SO4 (*).

L.V. Lombardo Lupano et al. / Applied Catalysis A: General 467 (2013) 342–354
351
Table 4
Parameters estimated for the changes in MO concentration due to adsorption and degradation by the heterogeneous system, from the empirical models presented in Eqs. (8)
and (9). The reaction was tested in two different media (0.1 M Na2SO4 and distilled water), using 0.1000 g of the Co(II)-Poly(EGDE-MAA-2MI) complex in 100 mL of 41 1 ␮M
MO and 63 mM H2O2 solution. The catalyst was tested on its first cycle of use, and on the second cycle with previous reloading of Co(II).
Condition
MO decoloration in water
First catalytic cycle
MO decoloration in 0.1 M Na2SO4
First catalytic cycle
MO decoloration in 0.1 M Na2SO4
Second catalytic cycle; Co(II) reloaded
9.09
1.96
5.5
[MO]ads (␮M)
S.D.: 0.56
0.343
S.D.: 0.047
21.96
S.D.: 0.51
0.0174
S.D.: 0.0014
9.33
S.D.: 0.14
0.558
S.D: 0.079
29.09
S.D.: 0.20
0.0092
S.D.: 0.0010
9.50
S.D.: 3.9
0.036
S.D.: 0.018
19.3
S.D.: 2.9
0.0075
S.D.: 0.0024
13.7
kads (min−1
[MO]deg (␮M)
kdeg (min−1
)
)
[MO] (␮M)
∞
S.D.: 0.63
0.9990
S.D.: 0.22
0.9991
S.D.: 1.4
0.9862
R2
where [MO] represents the free MO that decomposed with a very
was complete after 12 h. Evidently, the immobilized dye was also
susceptible to the oxidative degradation, and the adsorbed MO
would affect in some extent the kinetics of H2O2 activation.
The effect of the amount of catalyst on the decolorization rate
was also studied. When we changed the solution volume (L) to mass
of catalyst (g) ratio from 1 to 2.5 in distilled water, there was an
increase of 33.4% in [MO]deg and a decrease of 85.6% in kdeg, com-
pared with the corresponding parameters for a ratio equal to 1
(Supplementary content, Fig. A.8 and Table A.1). Besides, a lower
H2O2 initial concentration significantly affected the [MO]ads (lead-
ing to higher amounts) and the kdeg, as expected (Supplementary
content, Table A.1).
∞
low kinetic constant since a complete removal was observed after
24 h of reaction; [MO]ads is the total amount of MO per litre of solu-
tion that would be removed by adsorption; and [MO]deg is the total
amount of MO per litre of solution that would be removed by degra-
dation. The pseudo-first-order kinetic constants (k) were assigned
to each process.
The overall process of dye removal from the solution has a com-
plex description involving surface adsorption on the catalyst and
oxidative degradation. The degradation of MO is expected to be
slower than the adsorption process which only involves mass trans-
port and non-covalent bond formation.
Instead, the degradation process would involve mass transport
of H2O2 to the catalyst and free radical formation on the catalytic
surface in parallel with the eventual appearance of catalytically
active cobalt intermediate species. The oxidative effect of the free
radicals would require their mass transport to the bulk solution
and the chemical oxidation reaction of the dye. Besides, the MO
molecules should be transported to the particles surface to be oxi-
dized by the cobalt intermediate species. In agreement with this,
we assigned the term with the highest k to the adsorption process
(Table 4 and Fig. 8).
Regarding the degradation products, we observed that the
UV–visible spectra of the treated extracts of MO exhibited absorp-
tion bands at 400 nm even when the solution was already
decolorized (Supplementary content, Fig. A.9). This was consis-
tent with the presence of aromatic products, indicating that the
mineralization was not achieved [15–17].
The next experiments were conducted to put in evidence the
main mechanism of MO degradation, the role of the solid matrix
and the potential participation of cobalt intermediate species in
the MO oxidative reaction.
From the tabulated values of the empirical parameters it can
be observed that only 4.8% of MO was adsorbed on the catalyst in
0.1 M Na2SO4 (Table 4) and that 71.7% was degraded by means of
radicals. Besides, the amount of adsorbed MO in 0.1 M Na2SO4 was
close to the estimated in the absence of H2O2 (Tables 3 and 4).
On the other hand, only 54.4% of MO was degraded in distilled
water, but the kinetic constant was 1.89 times higher than the kdeg
estimated in the saline medium (Table 4).
Fig. 8 shows that near 70% of MO was removed from distilled
water in 2 h. Instead, 70% of decolorization was achieved after
260 min in 0.1 M Na2SO4. Here, we considered both adsorption and
degradation ways of decolorization.
It must be mentioned that in these experiments, the adsorbed
MO was observed on the surface of the particles even when the
solutions were almost decolorized. Then, we performed control
experiments to evaluate the chemical stability of the adsorbed dye.
In the first control experiment, the MO was adsorbed on
particles of the catalyst. In parallel, a H2O2 solution was put in
contact with the solid particles of Co(II)-Poly(EGDE-MAA-2MI)
during 10 min in order to activate the H2O2 and to get the maximal
concentration of free radicals. Then, the MO-loaded Co(II)-
poly(EGDE-MAA-2MI) was put in contact with activated H2O2.
Here we observed a partial decolorization of the particles during
the first 60 min of contact, and the complete absence of adsorbed
MO after 8 h. When the control experiment was repeated without
the previous step of H2O2 activation, the MO colour was preserved
on the particles during the first 60 min but the decolorization
First, the H2O2 solution was activated during 10 min with the
solid particles of Co(II)-Poly(EGDE-MAA-2MI). Then, the suspen-
sion was filtered, the supernatant was mixed with concentrated
MO solution, and the MO concentration was monitored as a func-
tion of time. The concentration of the dye exhibited a lag phase
during the first 10 min, followed by a pseudo-first order expo-
nential decay with a kdeg equal to 0.0089 0.0002 min−1 (Fig. 9).
This parameter coincided with the kinetic constant estimated for
the degradation process (0.0092 0.0010 min−1) when the whole
Co(II)-Poly(EGDE-MAA-2MI)/63 mM H2O2 heterogeneous system
was used (Table 4). In this supernatant solution, the degrada-
tion of MO can be attributed exclusively to the oxidative action
of the free radicals from the previously activated H2O2. The con-
trol experiment performed with MO and H2O2 in absence of the
catalyst, showed no changes in the absorbance of the solution
(Fig. 9).
Besides, we combined the independent results obtained from
the MO adsorption on the catalyst and the results from the degrada-
tion of MO by action of the free radicals from the supernatant. The
simulated profile arisen from this combination clearly coincided
with the MO decolorization performed with the whole Co(II)-
Poly(EGDE-MAA-2MI)/63 mM H2O2 heterogeneous system (Fig. 9).
In a second approach, we intended to evidence the formation of
cobalt intermediate species and to study their role on MO degrada-
tion. This time, the solid particles of Co(II)-Poly(EGDE-MAA-2MI)
were activated with H2O2 and put in contact with a MO solution;
the resulting profile of dye concentration was consistent with the

352
L.V. Lombardo Lupano et al. / Applied Catalysis A: General 467 (2013) 342–354
MO and H2O2, in the absence of the catalyst. On the other hand, the
amount of Co(II) leached was not dependent of the composition of
the medium.
For the second and third catalytic cycle, we detected 19.4 ␮g
(6.32%) and 0.92 ␮g (3.00%), respectively. This means that the 30%
of the adsorbed Co(II) interacts weakly with the polyampholyte,
and the other fraction is mostly retained by the matrix in succes-
sive cycles of use. Here, the H2O2 acting as ligand of Co(II) in the
activation reaction, would compete with the weaker ligand groups
of the polymer (such as COO−) for the coordination of the cation,
leading to some extent of cobalt release from the heterogeneous
catalyst. This assumption is supported on the fact that the poly-
electrolyte named poly(EGDE-MAA) has very low capacity for Cu(II)
and Co(II), requiring high concentration levels of the cation to form
the respective coordination complexes [10,11].
Cobalt salts are frequently used in catalysis of organic chemical
reactions, as well as in painting pigments, B12 vitamin manufac-
turing, and as foam stabilizer in beer, among other applications.
Regarding the environmental concern, cobalt ions have both ben-
eficial and harmful effects on health [41–43]. In order to overcome
the cobalt leak, we recommend the use of the non-complexed
polyampholyte (Poly(EGDE-MAA-2MI) as an adsorbent for the dis-
solved ions released from the catalyst after the treatment with
H2O2. The loading capacity of this resin for Co(II) ions was already
estimated in 30.7 mg g−1 (Table 1).
Fig. 9. MO concentration profile in 0.1 M Na2SO4 under different experimen-
tal conditions: with 63 mM H2O2 (*); with Co(II)-Poly(EGDE-MAA-2MI) (ꢃ); with
Co(II)-Poly(EGDE-MAA-2MI)/63 mM H2O2 (ꢁ); with the supernatant of Co(II)-
Poly(EGDE-MAA-2MI)/63 mM H2O2 (contact time with catalyst: 10 min) (ꢁ); the
sum of the individual effects of MO adsorption and degradation by supernatant (+).
In order to explore the regeneration of the catalyst used for the
first catalytic cycle, it was recovered, reloaded with Co(II) and tested
again on the activation of H2O2 for decolorization of MO solutions
(Fig. 8). Here we found that the kdeg did not change significantly
when compared with the first catalytic cycle, indicating that the
efficiency on H2O2 activation was recovered with Co(II) reloading
(Table 4). On the other hand, the kads decreased a 93.5% meanwhile
[MO]ads did not suffer a significant variation, indicating that some
reversible surface changes resulted from the overall treatment. The
MO molecules, which bear negative electric charge from the sul-
fonate group, would interact with the free 2MI residues at the time
that these groups become protonated due to the release of H+ from
H2O2 activation.
From these result we also concluded that the catalyst needs to
be reloaded with Co(II) ions between cycles in order to get the
most efficient performance. At this moment, our working group
is focussing the research towards the synthesis of stronger ligands
(alternative to 2MI) to be incorporated to the polyampholyte, in
order to overcome the cation leak.
decolorization by adsorption on the solid (Supplementary content,
Fig. A.10).
Then, the activation of the solid with H2O2 was repeated twice,
and the activated catalyst was put in contact with the same MO
solution after each activation cycle, resulting in a small decrease of
MO concentration when the equilibrium was reached. In this case,
the MO adsorbed on the first contact would have been degraded
during the second activation cycle, regenerating some binding
sites (Supplementary content, Fig. A.10). If any cobalt intermedi-
ate species were formed during the activation of the catalyst, there
was not enough evidence of their direct catalytic activity on MO
oxidative degradation.
Based on this experimental evidence, we conclude that the main
mechanism of MO degradation would involve the oxidative action
of the free radicals present in the bulk solution. The direct contribu-
tion of a hypothetical cobalt intermediate specie from the particles
on the dye oxidation would be minimal. The MO adsorbed on the
catalyst particles would not be degraded by the catalytic sites of the
•
−
surface; instead, it would suffer the oxidative attack of the O2
from the solution, decolorization with a slower kinetics than the
solubilized MO.
3.8. Oxidation of epinephrine
The catalytic activation of H2O2 was also assessed with
epinephrine as a model compound of pharmaceuticals. Epinephrine
is a catecholamine which can be oxidized by superoxide to
the strongly absorbing product adrenochrome (catecholamine o-
quinone; Scheme 3) [19]. Its presence is detected in solution by a
pink colour, which then turns brown upon polymerization.
3.7. Catalyst reuse and stability
The Co(II)-Poly(EGDE-MAA-2MI) complex was recovered and
re-used in three cycles of reaction in order to confirm that this
material is not a reagent and to demonstrate the catalytic activity
of the solid.
Both adsorption and degradation of MO took place during the
first cycle of 30 min in 0.1 M Na2SO4, and only degradation was
expected in the following cycles. In the third cycle, only 27% of MO
was degraded and the kdeg was the 82% of the value obtained in the
first cycle (Supplementary content, Fig. A.11).
Fig. 10 presents the results of the epinephrine conversion in the
presence of H2O2 and Co(II)-Poly(EGDE-MAA-2MI). About 80% of
conversion was reached in less than 6 min, with a pseudo-first-
order kinetic constant of 0.397 0.025 min−1
.
In the absence of catalyst, the oxidation of epinephrine by H2O2
was practically negligible.
Based on these results, we considered the cobalt leaking as a
possible cause of the loss of efficiency in successive cycles.
In our system, we found a release of 88.4 ␮g of Co(II) in 10.0 mL
of reaction at equilibrium after the first catalytic cycle, which rep-
resents 28.8% of the total Co(II) in the system. It must be pointed
out that this exhausted solution recovered from this first catalytic
cycle did not exhibit catalytic activity upon the addition of more
A control experiment performed with epinephrine and Co(II)-
Poly(EGDE-MAA-2MI) in the absence of H2O2 exhibited some
extent of catecholamine oxidation, being less efficient than the oxi-
dation by H2O2. Then, the control experiment was repeated, this
time with previous and continuous bubbling of argon to displace
the dissolved O2 and to saturate the atmosphere with the inert
gas. Here, the amount of adrenochrome released was negligible

L.V. Lombardo Lupano et al. / Applied Catalysis A: General 467 (2013) 342–354
353
effective concentration of free radicals can also be minimized with
our method.
In this same context, the method can be useful to obtain partially
oxidized industrial products like adrenochrome from epinephrine
(a precursor of haemostatic drugs such as carbazochrome), which is
usually synthesized with Ag2O as oxidant [44,45]. Here, Ag+ could
be replaced by an oxidant of lower environmental impact.
As a summary, the applicability of this cobalt-based polymeric
catalyst combined with H2O2 is acceptable in terms of both perfor-
mance and low environmental impact.
Acknowledgements
The authors thank the financial support from University of
Buenos Aires (UBACyT 10-12/237 and 11-14/915), CONICET (CON-
ICET 2010-2012/PIP 076), ANPCyT (BID 1728/PICT 01778 and BID
Bicentenario/PICT 1957). L.V. Lombardo Lupano thanks the CON-
ICET for her doctoral fellowship and Dr. J.M. Lázaro Martínez thanks
the CONICET for his post-doctoral fellowship. The authors thank
Dr. Victor Busto (Department of Microbiology, Immunology and
Biotechnology, Faculty of Pharmacy and Biochemistry, University of
Buenos Aires) for the measurements of partial pressure of dioxygen.
Fig. 10. Percent of epinephrine conversion to adrenochrome in 0.1 M Na2SO4
solution: in the presence of Co(II)-Poly(EGDE-MAA-2MI) (ꢃ), in the presence of
Co(II)-Poly(EGDE-MAA-2MI) with argon bubbling (×), in the presence of Co(II)-
Poly(EGDE-MAA-2MI) and 46 mM H2O2 (ꢁ), and in the presence of 46 mM H2O2
(ꢄ).
Appendix A. Supplementary data
Supplementary data associated with this article can be found,
in the online version, at http://dx.doi.org/10.1016/j.apcata.2013.
08.002.
(Fig. 10), indicating that the oxidation of epinephrine could also
be achieved with atmospheric O2, when it was activated in some
way by Co(II)-Poly(EGDE-MAA-2MI).
References
4. Conclusions
[1] D.E. Richardson, H.R. Yao, K.M. Frank, D. Bennett, J. Am. Chem. Soc. 122 (2000)
1729–1739.
[2] C.W. Jones (Ed.), Applications of Hydrogen Peroxide and Derivatives, RSC, Lon-
don, 1999.
[3] W. Hess, J. Treutwein, G. Hilt, Synthesis (2008) 3537–3562.
[4] E.K. Beloglazkina, A.G. Majouga, R.B. Romashkina, N.V. Zyk, Tetrahedron Lett.
47 (2006) 2957–2959.
A chemically stable coordination complex was prepared with
Co(II) and the polyampholyte Poly(EGDE-MAA-2MI). The interac-
tion between the polymer and Co(II) was well characterized by
solid-state NMR spectroscopy.
The catalytic activity of this material on H2O2 activation could be
deeply studied by ESR. We could propose a mechanism of reaction
which involves the simultaneous production of superoxide radical,
dioxygen, protons and water. This mechanism was different from
the H2O2 activation by the Cu(II)-Poly(EGDE-MAA-2MI) complex
described in a previous article.
[5] D. Gao, Q. Gao, Catal. Commun. 8 (2007) 681–685.
[6] Y. Zhang, Z. Li, W. Sun, C. Xia, Catal. Commun. 10 (2008) 237–242.
[7] B. Tamami, S. Ghasemi, Appl. Catal. A 393 (2011) 242–250.
[8] W. Lu, W. Chen, N. Li, M. Xu, Y. Yao, Appl. Catal. B 87 (2009) 146–151.
[9] M.F. Leal Denis, R.R. Carballo, A.J. Spiaggi, P.C. Dabas, V. Campo Dall’ Orto, J.M.
Lázaro Martínez, G.Y. Buldain, React. Funct. Polym. 68 (2008) 169–181.
[10] J.M. Lázaro Martínez, M.F. Leal Denis, V. Campo Dall’ Orto, G.Y. Buldain, Eur.
Polym. J. 44 (2008) 392–407.
[11] J.M. Lázaro Martínez, A.K. Chattah, G.A. Monti, M.F. Leal Denis, G.Y. Buldain, V.
Campo Dall’ Orto, Polymer 49 (2008) 5482–5489.
[12] J.M. Lázaro Martínez, A.K. Chattah, R.M. Torres Sánchez, G.Y. Buldain, V. Campo
Dall’ Orto, Polymer 53 (2012) 1288–1297.
[13] J.M. Lázaro Martínez, M.F. Leal Denis, L.L. Piehl, E. Rubín de Celis, G.Y. Buldain,
V. Campo Dall’ Orto, Appl. Catal. B 82 (2008) 273–283.
[14] Methyl Orange Material Safety Data Sheet, Santa Cruz Biotechnology, Inc., 2007.
[15] F. Chen, Y. Xie, J. He, J. Zhao, J. Photochem. Photobiol. A: Chem. 138 (2001)
139–146.
[16] C. Baiocchi, M.C. Brussino, E. Pramauro, A. Bianco Prevot, L. Palmisano, G. Marcí,
Int. J. Mass Spectrom. 214 (2002) 247–256.
[17] Y. He, F. Grieser, M. Ashokkumar, Ultrason. Sonochem. 18 (2011) 974–980.
[18] G.K. Parshetti, A.A. Telke, D.C. Kalyani, S.P. Govindwar, J. Hazard. Mater. 176
(2010) 503–509.
[19] R. Alhasan, D. Njus, Anal. Biochem. 381 (2008) 142–147.
[20] Z. Tao, G. Wang, J. Goodisman, T. Asefa, Langmuir 25 (17) (2009) 10183–10188.
[21] B.M. Fung, A.K. Khitrin, K. Ermolaev, J. Magn. Reson. 142 (2000) 97–101.
[22] A.M. El-Brashy, M. El-Sayed Metwally, F.A. El-Sepai, J. Chin. Chem. Soc. 52 (2005)
77–84.
[23] M. Hadi, M.R. Samarghandi, G. McKay, Chem. Eng. J. 160 (2010) 408–416.
[24] K. Burnham, D. Anderson, K. Huyvaert, Behav. Ecol. Sociobiol. 65 (2011)
23–35.
[25] G.J. Copello, L.E. Diaz, V. Campo Dall’ Orto, J. Hazard. Mater. 217–218 (2012)
374–381.
[26] M.B. Ettinger, C.C. Ruchhoft, R.J. Lishka, Anal. Chem. 23 (12) (1951) 1783–1788.
[27] L.R. Denaday, M.V. Miranda, R.M. Torres Sánchez, J.M. Lázaro Martínez,
L.V. Lombardo Lupano, V. Campo Dall’ Orto, Biochem. Eng. J. 58–59 (2011)
57–68.
This complex combined effectiveness in H2O2 activation, ease
of recovery and aptitude for regeneration.
The high loading capacity of the polyampholyte for ionic species
may lead to the adsorption of solutes, which would affect the cat-
alytic efficiency in the oxidative treatment of complex samples. This
drawback can be conveniently overcome by the addition (if not yet
present) of a certain amount of inorganic salts such as sodium or
potassium sulfate.
The superoxide radicals are considered less reactive than
hydroxyl species, leading to an oxidative treatment of mild
strength.
In the treatment of MO, the mineralization of the organic com-
pound was not achieved, according with the spectral evidence.
We must recall that MO solutions have been found to be toxic for
some model plants and bacteria, which would prevent the choice of
biological methods as the starting way of decolorization of waste-
water. Then, our “green and mild” oxidative treatment could be
useful to initiate the degradation (until exhaustion of H2O2), which
could eventually be completed by microorganisms that use small
organic compounds as carbon and/or nitrogen source.
Superoxide radicals can also be useful to accelerate some degra-
dation processes in which the substrate has a low reduction
potential. Besides, the amount of oxidant necessary to produce an
[28] D.D. Do, Adsorption Analysis: Equilibria and Kinetics, vol. 2, Imperial College
Press, London, 1998.

354
L.V. Lombardo Lupano et al. / Applied Catalysis A: General 467 (2013) 342–354
[29] F. Haghseresht, G.Q. Lu, Energy Fuel 12 (1998) 1100–1107.
[30] C. Aharoni, D.L. Sparks, in: D.L. Sparks, D.L. Suarez (Eds.), Rates of Soil
Chemical Processes, Soil Science Society of America, Madison, WI, 1991,
pp. 1–18.
[38] J.M. Lázaro Martínez, E. Rodríguez-Castellón, R.M. Torres Sánchez, L.R. Denaday,
G.Y. Buldain, V. Campo Dall’ Orto, J. Mol. Catal. A: Chem. 339 (2011) 43–51.
[39] C. Shen, S. Song, L. Zang, X. Kang, Y. Wen, W. Liu, L. Fu, J. Hazard. Mater. 177
(2010) 560–566.
[31] C.-T. Hsieh, H. Teng, J. Chem. Technol. Biotechnol. 75 (2000) 1066–1072.
[32] J.P. Hobson, J. Phys. Chem. 73 (1969) 2720–2727.
[40] B. Agboola, K.I. Ozoemena, T. Nyokong, J. Mol. Catal. A: Chem. 227 (2005)
209–216.
[33] I.H. Gübbük, R. Güp, M. Ersöz, J. Colloid Interface Sci. 320 (2008) 376–382.
[34] I. Bertini, P. Turano, A.J. Vila, Chem. Rev. 93 (1993) 2833–2932.
[35] P.L. Zamora, F.A. Villamena, J. Phys. Chem. A 116 (2012) 7210–7218.
[36] S.-X. Liang, L.-X. Zhao, B.-T. Zhang, J.-M. Lin, J. Phys. Chem. A 112 (2008)
618–623.
[41] S. Rengaraj, S.-H. Moon, Water Res. 36 (2002) 1783–1793.
[42] M.F. Abdel-Sabour, J. Environ. Sci. 15 (3) (2003) 388–395.
[43] D.S. Pines, D.A. Reckhow, Environ. Sci. Technol. 36 (2002) 4046–4051.
[44] MacCarthy, Chim. Ind. Paris 55 (1946) 435.
[45] M. Verstraete, Haemostatic Drugs, Springer, Netherlands, The Hague, 1977, pp.
59–66.
[37] G.R. Buettner, Free Radical Biol. Med. 3 (1987) 259–303.