Spectroscopic and electrochemical studies of novel model compounds for cytochrome c oxidase

Article abstract of DOI:10.1016/j.ica.2010.03.026

Different metalated porphyrin compounds were studied as model complexes for cytochrome c oxidase. All models contain a tyrosine molecule and a copper binding site. Two of the compounds are bearing an axial pyridine ligand that could possibly coordinate wi

Full text of DOI:10.1016/j.ica.2010.03.026

Inorganica Chimica Acta 363 (2010) 2201–2208
Contents lists available at ScienceDirect
Inorganica Chimica Acta
journal homepage: www.elsevier.com/locate/ica
Spectroscopic and electrochemical studies of novel model compounds
for cytochrome c oxidase
Kalliopi Ladomenou a,b, Georgios Charalambidis a, Athanassios G. Coutsolelos a,
*
a Department of Chemistry, University of Crete, Laboratory of Bioinorganic Chemistry, Voutes Campus, P.O. Box 2208, 71003 Heraklion, Crete, Greece
b Department of Human Nutrition and Dietetics, Technological Educational Institute (TEI) of Crete, I. Kondylaki 46 Street, 723 00 Siteia, Crete, Greece
a r t i c l e i n f o
a b s t r a c t
Article history:
Different metalated porphyrin compounds were studied as model complexes for cytochrome c oxidase.
All models contain a tyrosine molecule and a copper binding site. Two of the compounds are bearing
an axial pyridine ligand that could possibly coordinate with Fe porphyrins. All complexes were studied
using NMR and UV–Vis spectroscopies and it was found that the coordination of the axial ligand is pos-
sible only in one of the porphyrins. Moreover, the synthesized catalysts were studied as promising
enzyme mimics using a rotating disc electrode in the presence of molecular oxygen.
Ó 2010 Elsevier B.V. All rights reserved.
Received 30 October 2009
Received in revised form 4 March 2010
Accepted 11 March 2010
Available online 18 March 2010
Keywords:
Enzyme models
Porphyrins
Oxidase
Dioxygen
Metalloenzymes
1. Introduction
of metalloenzymes. Such detailed information and systematic
variations are difficult to obtain from wild-type enzymes or their
The consumption of oxygen in aerobic cellular respiration is pri-
marily mediated by the multiunit enzyme cytochrome c oxidase
(CcO) [1,2]. This system spans the mitochondrial membrane and
catalyzes the reduction of dioxygen to water while concomitantly
building a transmembrane electrochemical gradient by proton
translocation. Thus, dioxygen serves the dual role as terminal elec-
tron acceptor for oxidative catabolism and as energy source for the
majority of ATP production. The three-dimensional structures of
CcOs from Paracoccus denitricans [3,4] and bovine heart [5–7] have
been determined by X-ray diffraction. The catalytic active site of
the enzyme consists of a heme (heme a3), a tricoordinated copper
ion (CuB) and a tyrosine residue (Y244) which is covalently con-
nected to one of the copper bound histidines (H240) (Fig. 1).
Extensive spectroscopic and crystallographic studies of CcO
have illuminated many characteristics that lead to its efficient
4eÀ reduction of dioxygen, although many controversial issues re-
main, especially concerning the role of the copper ion and the tyro-
sine residue. The development of biomimetic models to investigate
the structural–functional relationships of native metalloenzymes
has proven to be a successful strategy [8–11]. Simulation and var-
iation of synthetic models provide insight into the coordination
environments, spectroscopic properties, and catalytic mechanisms
mutants because of their restricted availability and difficulty in
mutagenesis. To date, many synthetic models have been reported
by Karlin and co-workers [12–15] and Collman et al. [16–19] with
models bearing either a Tyr244 mimic or not. However, in all of the
cases the tyrosine mimic is only a phenol ring and not the whole
tyrosine molecule. In the present study, we have developed a series
of models bearing a tyrosine molecule and an axial ligand that
could coordinate with iron and block the lower face of the model,
leaving the upper face free to bind an oxygen molecule.
2. Experimental
2.1. General
1H NMR spectra were recorded unless otherwise specified, as
deuteriochloroform solutions using the solvent peak as internal
standard on a Bruker AMX-500 MHz spectrometer. UV–Vis spectra
were recorded on a Shimadzu Multispec-1501 instrument. High-
resolution mass spectra were performed on a MS/MS ZABSpec
TOF spectrometer at the University of Rennes I (C.R.M.P.O.). Ele-
mental analyses were carried out using a Carlo-Erba EA 1110,
CHNS Eager 200 analyzer. Thin layer chromatography was
performed on silica gel 60 F254 plates. Chromatography refers to
flash chromatography and was carried out on SiO2 (silica gel 60,
SDS, 70–230 mesh ASTM). All dry solvents used were dried by
the appropriate technique. Organic extracts were dried over
* Corresponding author. Fax: +30 2810545001.
E-mail address: coutsole@chemistry.uoc.gr (A.G. Coutsolelos).
0020-1693/$ - see front matter Ó 2010 Elsevier B.V. All rights reserved.
doi:10.1016/j.ica.2010.03.026

2202
K. Ladomenou et al. / Inorganica Chimica Acta 363 (2010) 2201–2208
roform to give solid 1Fe (34 mg, 79%). Rf (CH2Cl2/MeOH 9:1): 0.45.
Anal. Calc. for C82H74N12O4Fe: C, 73.10; H, 5.54; N, 12.47. Found: C,
73.01; H, 5.44; N, 12.58%. UV–Vis absorption: kmax, nm (e ,
, cmÀ1
M
À1), 415 (8.5 Â 104), 506 (2.3 Â 103). HRMS (ES+) Calc. for
C82H74N12O4Fe [M]+ 1346.5305. Found: 1346.5321.
2.3. General procedure of Cu metalation reactions
To a solution of 1 equivalent of iron porphyrin in a mixture of
MeCN/EtOH (95:5) 1.1 equivalents of copper acetate (dissolved in
MeCN), was added and the mixture was stirred at 60 °C for
10 min. The completion of the reaction was confirmed using TLC
chromatography. Then, the solvent was removed under vacuum
and the resulting solid was dissolved in chloroform. The solid
was precipitated using pentane, filtered and washed with pentane
giving FeCu porphyrin as brown-purple solid.
Fig. 1. Structure of the bovine cytochrome c oxidase active center.
magnesium sulfate unless indicated otherwise. Evaporation of the
solvents was accomplished on a rotary evaporator. Synthesis and
characterization of all free base porphyrins has been reported else-
where [20].
2.3.1. Synthesis of porphyrin 1FeCu
Porphyrin 1Fe (10 mg, 7.9
lmol) and copper acetate [(0.8 ml,
8.7
l
mol), from a standard solution of 50 mg of Cu(CH3COOH)2 in
25 ml of MeCN] were dissolved in MeCN/EtOH (5 ml). Product
1FeCu was obtained quantitatively. Rf (CH2Cl2/MeOH 9:1): 0.70.
, cmÀ1, MÀ1), 418 (9.0 Â 104), 509
2.2. General procedure of Fe metalation reactions
UV–Vis absorption: kmax, nm (
e
(2.2 Â 103). HRMS (ES+) Calc. for C74H64N10O5FeCu [MÀCl]+
In a Schlenk flask and under an inert atmosphere, 1 equivalent
of porphyrin and 10 equivalents of iron dibromide (FeBr2) were
added. Then deoxygenated and anhydrous THF inserted in the flask
and the solution was stirred at 55 °C for 12 h. The color of the solu-
tion changed from purple to orange. The completion of the reaction
was confirmed by UV–Vis spectroscopy and TLC chromatography.
The solvent removed under vacuum and the residue dissolved in
chloroform and washed twice with 2 N HCl, then once with satu-
rated NaHCO3, dried, filtered, and concentrated under vacuum.
The resulting product was purified by column chromatography
and recrystallized in CH2Cl2 pentane, to obtain Fe porphyrin, as
brown-purple solid. Throughout the purification procedures excess
of NaCl was used in order to confirm that a Cl atom remains as the
axial Fe ligand.
1291.3707. Found: 1291.3715.
2.3.2. Synthesis of porphyrin 2FeCu
Porphyrin 2Fe (7 mg, 5.1
lmol) and copper acetate [(0.5 ml,
5.6
l
mol), from a standard solution of 50 mg of Cu(CH3COOH)2 in
25 ml of MeCN] were dissolved in MeCN/EtOH (4 ml). Product
2FeCu was obtained quantitatively. Rf (CH2Cl2/MeOH 9:1): 0.41.
UV–Vis absorption: kmax, nm (
e
, cmÀ1, MÀ1), 417 (8.2 Â 104), 507
(2.0 Â 103). HRMS (ES+) Calc. for C82H74N12O4FeCu [MÀCl]+
1409.4601. Found: 1409.4617.
2.3.3. Synthesis of porphyrin 3FeCu
Porphyrin 3Fe (20 mg, 14.5
lmol) and copper acetate [(1.5 ml,
15.9 mol), from a standard solution of 50 mg of Cu(CH3COOH)2
in 25 ml of MeCN] were dissolved in MeCN/EtOH (8 ml). Product
l
2.2.1. Synthesis of porphyrin 1Fe
Porphyrin 1 (30 mg, 0.026 mmol) and FeBr2 (56 mg, 0.26 mmol)
were dissolved in THF (5 ml). Column chromatography (1–5%
MeOH in chloroform). The product was eluted at 2% MeOH in chlo-
roform to give solid 1Fe (27 mg, 82%). Rf (CH2Cl2/MeOH 9:1): 0.71.
Anal. Calc. for C74H64N10O5FeCl: C, 70.28; H, 5.10; N, 11.08. Found:
3FeCu was obtained quantitatively. Rf (CH2Cl2/MeOH 9:1):
0.44. UV–Vis absorption: kmax, nm (
e
, cmÀ1, MÀ1), 415 (8.0 Â 104),
509 (2.1 Â 103). HRMS (ES+) Calc. for C82H74N12O4FeCu [M]+
1409.4601. Found: 1409.4612.
C, 70.09; H, 5.23; N, 11.13%. UV–Vis absorption: kmax, nm (e ,
, cmÀ1
M
À1), 418 (9.5 Â 104), 509 (2.4 Â 103). HRMS (ES+) Calc. for
C74H64N10O5Fe [MÀCl]+ 1228.4411. Found: 1228.4435.
2.4. Electrochemical studies
The diameter of the graphite disk of the RRDE (Pine Instru-
ments) was 6 mm, a bipotentiostat (Solea-Tacussel) was used to
control its potential and maintained the platinum ring potential
at 1.0 V so as to detect, through its oxidation, H2O2 produced by
the reduction of O2 at the graphite electrode. The voltammograms
were recorded on a SEFRAM T.2Y x–yy recorder. All the potentials
are referred to SCE (KCl saturated). The aqueous solution in contact
with the electrode was buffered at pH 6.86 (0.025 M KH2PO3,
0.025 M Na2HPO3) and saturated with O2 at 1 atm. The electrode
was abraded with wet SiC paper (grade 600), sonicated in deion-
ized water for 1 min, washed with water and acetone, and dried.
The adsorption of the catalysts on the graphite surface was done
by immersing the electrode for 5 min in a solution of the catalyst
in CH2Cl2. The electrode was then rinsed with CH2Cl2 and used
for the measurements.
2.2.2. Synthesis of porphyrin 2Fe
Porphyrin 2 (15 mg, 0.012 mmol) and FeBr2 (26 mg, 0.12 mmol)
were dissolved in THF (3 ml). Column chromatography (1–10%
MeOH in chloroform). The product was eluted at 5% MeOH in chlo-
roform to give solid 2Fe (13 mg, 78%). Rf (CH2Cl2/MeOH 9:1): 0.41.
Anal. Calc. for C82H74N12O4FeCl: C, 71.22; H, 5.39; N, 12.15. Found:
C, 71.39; H, 5.22; N, 12.04%. UV–Vis absorption: kmax, nm (e ,
, cmÀ1
M
À1), 418 (8.4 Â 104), 507 (2.1 Â 103). HRMS (ES+) Calc. for
C82H74N12O4Fe [MÀCl]+ 1346.5305. Found: 1346.5319.
2.2.3. Synthesis of porphyrin 3Fe
Porphyrin 3 (40 mg, 0.031 mmol) and FeBr2 (67 mg, 0.31 mmol)
were dissolved in THF (7 ml). Column chromatography (1–10%
MeOH in chloroform). The product was eluted at 5% MeOH in chlo-

K. Ladomenou et al. / Inorganica Chimica Acta 363 (2010) 2201–2208
2203
are shown in Scheme 1. For the metalation reaction we used a
10-fold excess of iron dibromide, therefore Fe could also coordi-
nate to tridentate ligand (Cu coordination site). However, this coor-
dination is weak and the iron is liberated from that site, due to
protonation of nitrogen atoms by the use of HCl extractions [25].
The incorporation of iron is confirmed by UV–Vis spectroscopy
[26] and TLC chromatography. The Soret band of the porphyrin ring
was shifted to shorter wavelengths, characteristic of iron binding
with the four pyrrolic nitrogens. Also, fewer Q bands and a de-
3. Results and discussion
3.1. Chemical structures of synthesized models
The synthesis of novel free base porphyrins 1–3 (Scheme 1),
was previously reported by our group [20]. The series of these
complexes are featuring a covalently attached tridentate ligand
site, where a Cu could be coordinated, a tyrosine molecule, and
two of them contain a proximal pyridine base. These new models
crease of extinction coefficient (e
) were observed. Moreover, 1H
are easily made using the a,b atropisomer of the same porphyrin.
NMR spectra of compounds 1Fe–3Fe were obtained. The electronic
configuration of FeIII is [Ar]3d5 and possesses unpaired electrons in
both high-spin and low-spin complexes. As a result, the proton sig-
nals in the 1H NMR spectrum were spread from 80 ppm down to
0 ppm, giving broad peaks that were difficult to assign. High-reso-
lution mass spectrometry (HRMS) and elemental analysis sup-
ported the identification of the derivatives.
The hydroxyl group of all compounds can be protected/deprotec-
ted in order to investigate the role of this tyrosine mimic. The tri-
dentate ligand with a tyrosine molecule is attached on the
porphyrin ring via an urea link, because the affinity of dioxygen
might be stronger compared to an amide link [21]. Moreover, in
complexes 2 and 3 an axial pyridine ligand is covalently attached
to the porphyrin in order to block the lower face of the porphyrin
ring. Additionally, it has been found that models bearing covalently
attached Fe axial ligands, appear to enhance further their stability
[16]. Thus, two different substituted pyridine molecules were used
in order to investigate which one will better coordinate to the iron.
Preliminary structural investigation was attempted by comparing
the 1H NMR chemical shifts of the free ligands and final molecules.
The spectra comparison showed that the tridentate ligand is not
suspended over the porphyrin plane and that the 3-pyridine base
in model 3 is oriented closer under the porphyrin ring compared
to 2-pyridine of model 2 [20].
3.3. Copper metalation reactions
In the next stage, metalation reactions with Cu were performed.
The insertion of Cu is done at the iron porphyrins, due to harsh
conditions used for Fe metalation. Otherwise copper metal would
be replaced by iron. For the introduction of copper, Cu+2 was used
[24]. To a solution of porphyrin derivatives in MeCN/ethanol, equi-
molar quantity of [Cu(CO2CH3)2] in MeCN was added. The mixture
was stirred at 60 °C for 10 min and the product was obtained pure
after precipitation using pentane. The metalated Cu–Fe products
(Scheme 2) could not be submitted to further purification due to
weak coordination of Cu ion. The compounds could not be fully
characterized by NMR spectroscopy, because Fe that is present is
paramagnetic. HRMS aided in the identification of the formed
complexes and the results were in agreement with the target
compounds.
3.2. Iron metalation reactions
Subsequently, metalation reactions of compounds 1–3 were
preformed, using iron dibromide (FeBr2), under anaerobic condi-
tions, following the same experimental conditions for all models
[22,23]. In a glove box iron dibromide was added to a solution of
porphyrin in dry THF and the mixture was stirred at 60 °C for
12 h. Insertion of iron induce a color change of the solution from
purple to orange, characteristic for ferrous porphyrins. After com-
pletion of the reaction, we exposed the solution to the atmosphere
and the Fe was immediately oxidized from the +2 to the more sta-
ble +3 oxidation state. During this oxidation iron can be complexed
with atmospheric oxygen and various oxygenated species can be
formed. The separation of the above species is difficult, thus extrac-
tions with dilute solution of HCl were carried out [24]. With this
procedure a chlorine atom was coordinated with iron and the di-
mer units were splitted. Consequently, we obtained exclusively
monomeric species, where iron was pentacoordinated and had an
axial chlorine ligand. The structures of iron porphyrins 1Fe–3Fe
3.4. 1H NMR studies
As referred above, even though NMR spectroscopy can not be
used to fully characterize the metalated complexes, it is a useful
technique to determine both the oxidation state and the spin of
iron [27]. All NMR spectra of models 1Fe–3Fe that were obtained
showed that the oxidation state of iron is +3 and is high-spin pen-
tacoordinated, since in all cases a signal due to pyrrolic protons is
present at around 80 ppm. Following, we investigated the possible
coordination of the proximal base as a fifth ligand for the iron
atoms of porphyrins 2Fe and 3Fe, when Fe is at +2 oxidation state.
It is well known in the literature that iron porphyrins in which iron
Scheme 1. Free-based and metalated models 1–3.

2204
K. Ladomenou et al. / Inorganica Chimica Acta 363 (2010) 2201–2208
Scheme 2. Metalated complexes 1FeCu–3FeCu (the Cu counter anions are not represented).
is hexacoordinated appear to have a diamagnetic NMR spectrum
and the pyrrolic protons appear at about 9 ppm [28]. Initially the
oxidized porphyrin 3Fe was diluted in air-free CDCl3 and NMR
spectra were obtained (Fig. 2). A peak at 79.31 ppm is present,
due to pyrrolic protons, which is characteristic of high-spin Fe+3
porphyrins. Subsequently, excess of carbon monoxide and a solu-
tion of Na2S2O4 in air-free D2O were introduced in the NMR tube.
After stirring the solution the organic phase changed color from
brown to red due to the reduction of iron in the +2 oxidation state.
Then the NMR spectrum obtained corresponds to a diamagnetic
complex, since there are no peaks above 10 ppm. The chemical
shift of the pyrrolic protons has shifted at 8.61 ppm. The complex
that is formed is low-spin, with total spin S = 0 and the Fe is six
coordinated [29]. The above results show that in complex 3Fe pyr-
idine is coordinated to Fe+2 and CO is the sixth ligand that induces
an extra stability to the complex. Additionally, the possibility that
the sixth axial ligand is another CO molecule is not likely [30].
The same experimental procedure was followed in the absence
of CO. The NMR spectrum obtained was exactly the same with the
initial spectrum of the oxidized 3Fe showing a peak at 79.31 ppm,
therefore this complex is reoxidized to iron +3 quite fast. In gen-
eral, pentacoordinated complexes of reduced iron porphyrins are
oxygen sensitive and there are only a few reports in the literature
that such compounds have been isolated and characterized at
room temperature.
Moreover, the above experimental procedure was followed for
complex 2Fe. The initial NMR spectrum of the compound showed
pyrrolic proton signals at 7502 ppm, characteristic of oxidized
Fig. 2. (a) 1H NMR spectrum of oxidized complex 3Fe, (b) 1H NMR spectrum of reduced model 3Fe in the presence of CO. Solvent used CDCl3, at 299.2 K.

K. Ladomenou et al. / Inorganica Chimica Acta 363 (2010) 2201–2208
2205
high-spin iron porphyrins. After the addition of CO and reduction
of the porphyrin it was not observed any difference at the NMR
spectrum, indicating that pyridine remains unbound. Conse-
quently, the above findings showed that the pyridine axial ligand
coordinates with iron in the case of 3Fe porphyrin, but not in the
case of 2Fe. These results are in accordance with the previous
structural investigation that was done using 1H NMR spectroscopy
for compounds 2 and 3 [23].
the same time is oxidized. This procedure lasts about 7.0 min and
the Soret band of the product appears at 421 nm and the Q band at
510 nm. It is known that pentacoordinated iron porphyrins that
contain bulky groups and a base as an axial ligand, form an inter-
mediate Fe+2–[O2] upon reaction with oxygen [31]. Oxygen binding
is reversible since bubbling argon into the solution gave back the
reduced complex. The derivative remains stable for about 1 h,
where after that time the Soret band is shifted to smaller wave-
lengths. Finally at the end of 2 h the Soret band appeared at
418 nm and the Q band at 508 nm, similar absorption bands with
the initial 3Fe (Section 2). In conclusion, model 3Fe is capable of
binding oxygen and forms a quite stable derivative. Therefore, it
can be used to further studies in order to understand the mecha-
nism of oxygen binding.
3.5. UV–Vis absorption studies
UV–Vis spectroscopy was used in order to study iron metalated
models. Our first goal was to investigate the possible, axial pyri-
dine ligand binding to iron atom. Therefore, we reduce the iron
porphyrins and then compare the UV–Vis spectra of both reduced
and oxidized forms of the models. The reduction of iron was
achieved with Na2S2O4 in anaerobic conditions using deoxygen-
ated solvents. Porphyrin 3Fe was first dissolved in deoxygenated
chloroform and the UV–Vis spectrum was obtained, showing a Sor-
et band at 415 nm and a Q band at 506 nm. Then, in the cuvette
saturated aqueous solution of Na2S2O4 was added and the mixture
was stirred. UV–Vis spectrum was obtained, and both the Soret and
Q bands were shifted to 434 nm and 533 nm respectively (Fig. 3).
This red shift was due to the reduction of iron and the coordination
of the axial pyridine ligand with the iron. Continuing, the above
procedure was repeated using model 2Fe and the UV–Vis spectrum
after the addition of the reducing agent showed no shift of the Sor-
et band. This indicates that in the case of 2Fe the axial pyridine
does not coordinate with Fe. Moreover, in order to examine if
one of the pyridines present at the copper binding site coordinates
with Fe, we followed the above procedure using 1Fe. The results
showed that the Soret band did not shift, therefore there is no coor-
dination of pyridine to Fe.
3.6. Electrochemical studies using rotating ring-disk electrode
Following, these porphyrin derivatives were assessed as prom-
ising enzyme mimics electrochemically using a rotating ring-disk
electrode (RRDE) in the presence of molecular oxygen. Our aim
was to compare the catalytic activity of the newly synthesized
compounds. The goal of this investigation was firstly to understand
the role of copper in the catalytic activity, and secondly to observe
if the presence of an axial ligand provides an extra stability to the
complex. On the one hand, the main advantage of this technique is
its ability to detect H2O2 by a ring current at the anode (made of a
platinum ring). It is important to note that the H2O2 detected at the
anode was produced by the reduction of dioxygen at the cathode
(composed of a graphite disk) and had migrated significantly. On
the other hand, a disadvantage of this technique consists in the
irreversible adsorption of the catalyst (the synthetic model) on
the surface of the electrode. Indeed, there is no direct way to ob-
serve how the catalyst evolves during the catalytic process. In or-
der to carry out the electrochemical experiments, the molecules
were adsorbed on the disk electrode, which was a section of a high
ordered graphite rod, with the planes of graphite being perpendic-
ular to the surface of the electrode (edge-plane graphite electrode,
EPGE). The potential of this modified graphite electrode was
scanned from high to lower values so as to record the voltammo-
gram for the reduction of O2. The graphite electrode was
surrounded by a platinum ring electrode, the potential of which
Our next objective was to study the oxygen binding to iron
models. Compound 3Fe was used to study O2 binding, since is
the only model of the three that has an axial iron ligand. This is
in accordance with the catalytic active site of the enzyme of cyto-
chrome c oxidase. Following, the reduced complex 3Fe was ex-
posed to the atmosphere and successive UV–Vis spectra were
obtained. A shift to smaller wavelengths is observed, due to oxygen
binding. Iron binds molecular oxygen from the atmosphere and at
Fig. 3. Successive spectra of 3Fe during its reduction. At the oxidized form Soret band 415 nm and at reduced form 434 nm.

2206
K. Ladomenou et al. / Inorganica Chimica Acta 363 (2010) 2201–2208
was set at fixed value (1.0 V) such that H2O2 eventually produced
at the disk was oxidized to O2. Catalytic studies of compound 1
are shown in Fig. 4. First, compound 1FeCu was studied (Fig. 4,
curve a), and from comparison with the 2eÀ reduction wave on a
bare EPGE, the number of electrons exchanged per O2 molecule
was 3. H2O2 is detected through its oxidation, as soon as O2 is re-
duced from the current increase at the platinum electrode encom-
passing the graphite disk. This means that the 2eÀ and 4eÀ
reduction mechanisms occur simultaneously. From the study of
complex 1Fe without Cu (Fig. 4, curve b), a reduced catalytic activ-
ity was observed since there was a decrease of electrons in the
reduction reaction from 3 to 2.4. Moreover, in the case of 1Fe more
H2O2 was detected. Therefore, the presence of Cu ion favors the 4eÀ
reduction of oxygen to water. Possibly in model 1FeCu copper is
present in a right electronic environment that allows the interac-
tion with oxygen bound on Fe porphyrin. Due to this interaction
O–O bond is splitted easier favoring 4eÀ mechanism and the
amount of H2O2 is less. These results are in accordance with the
studies of Karlin et al. in which a similar model to 1Fe was studied
in the presence or absence of a copper ion. In the case of the two
models under similar experimental conditions, they observed a
2eÀ and 4eÀ reduction of O2 which is in accordance with our find-
ings [32]. They also observed that the complex that contained cop-
per appeared to have a slightly increased catalytic activity and at
the same time there was a decrease at the H2O2 that was released.
Models 1Fe and 1FeCu were unstable and decomposed after suc-
cessive scans. After carrying out a second scan with the same elec-
trode reduction of the catalytic activity was observed, that
approached the one of the bare graphite electrode. The decomposi-
tion of the catalyst resulted the poisoning of the platinum elec-
trode from the catalyst molecules. As
a result the current
intensity can be used in order to arise only qualitative results.
The most precise method of determining the number of elec-
trons that participate in the reduction reaction of oxygen is the
study of the current intensity of graphite electrode. For that reason
at the voltammograms of models 2 and 3 (Fig. 5), is presented only
the current of the electrode where the molecule is absorbed. Com-
pounds 2 and 3 show similar catalytic activity (Fig. 5). In the case of
molecules that have metal ions (Fig. 5, curves a and d), oxygen
reduction takes place at negative potential values (À0.1 V), where
the number of electrons required for the reduction of oxygen is
ꢀ2.5. The presence of copper did not affect the catalytic activity
of these models (Fig. 5, curves b and e), therefore Cu is not involved
in the catalytic pathway. Boitrel and co-workers also suggested
that the presence of copper in the distal coordination site is not
essential for the 4eÀ reduction of O2 to H2O [33]. Moreover, Coll-
man and co-workers observed that copper does not affect the
kinetics of the catalysis or the stability of the catalysts [34].
Catalysts 2FeCu and 3FeCu reduce oxygen via two mechanisms
(2eÀ and 4eÀ), however their catalytic activity is lower compared
to catalyst 1FeCu. The reduction take place at lower potential val-
ues (lower than the ones that O2 is reduced from the bare graphite
electrode) and the number of electrons that are exchanged during
the reaction is decreased. Since models 2 and 3 contain a pyridine
group compared to 1, it seems that this axial ligand is possibly
responsible for the decrease of the catalytic activity. Nevertheless,
these results should be explained carefully, because different sub-
stitution of the models can affect the adsorption of the catalyst on
graphite electrode. The reduced adsorption could result the de-
crease of the catalytic center at the surface of the electrode, fact
that would lead in the reduction of the current observed from
graphite electrode. The concentration of the catalyst at the surface
Fig. 4. RRDE voltammograms (a) 1FeCu, (b) 1Fe, and (c) reduction of O2 from bare
graphite electrode.
Fig. 5. RRDE voltammograms (a) 2FeCu, (b) 2Fe, (d) 3FeCu, and (e) 3Fe. Curves (c) and (f) reduction of O2 from bare graphite electrode.

K. Ladomenou et al. / Inorganica Chimica Acta 363 (2010) 2201–2208
2207
Fig. 6. Voltammograms obtained after successive scans for 2FeCu (left) and 3FeCu (right).
of the electrode can not be determined due to decreased stability of
these models. The decomposition of the catalysts is quite fast, as a
result the concentration of the active molecules on the electrode
surface to vary during a voltammogram. In previous studies it
has been reported by Collman et al. that models that the axial li-
gand of Fe is covalently linked to porphyrin ring appear further sta-
bility [35]. Therefore, comparison of experimental data of
compounds 1–3, can give as an insight on the possible effect of
the axial ligand at the stability of the catalysts.
complexes were studied for their ability to reduce dioxygen to
water on graphite electrodes in contact with an aqueous solution.
All models reduce oxygen via two mechanisms (2eÀ and 4eÀ), and
have low stability. The lack of stability was more profound in mod-
els bearing axial ligands 2FeCu and 3FeCu. Finally, the presence of
copper in the case of 1FeCu increases the catalytic activity of this
model favouring 4eÀ mechanism of oxygen to water, compared
to 1Fe.
After successive scans models 2FeCu and 3FeCu (Fig. 6) lost
their catalytic activity. Compound 1FeCu showed similar behavior
which decomposes after successive scanning. Subsequently, even
though the models have structural differences, they appear to have
similar stability, so the axial ligand does not seem to reinforce their
stability. The actual mechanism of the decomposition of the cata-
lysts has not been determined. First it was proposed that H2O2
was responsible for the decomposition of the catalyst, nowadays
it has been suggested that the destruction of the catalyst includes
the formation of hydroxyl radical (ÅOH). These radicals are more
reactive as far as thermodynamic and kinetics is concerned com-
pared to H2O2, and they are formed during the homolytic cleavage
of O–O. Consequently, the decomposition of the synthetic models
is attributed to the formation of hydroxyl radicals. The radicals
are very reactive thus they do not reach platinum ring, and instead
they decompose the catalyst. Therefore, the study of the stability of
an electrocatalyst during O2 reduction can comprise a quantitative
measurement of its selectivity towards the heterolytic cleavage in
relevance to the homolytic cleavage of O–O bond. Moreover, based
on the above findings some conclusions can be made about the cat-
alytic activity of the models bearing Cu ions. The reduced selectiv-
ity of the models that do not contain Cu is not because they can not
reduce oxygen to water, but due to their fast decomposition. Ab-
sence of Cu the homolytic cleavage of O–O is favored resulting a
better catalytic activity and selectivity towards 4eÀ reduction of
oxygen to water.
Acknowledgments
We acknowledge for the financial support the European Union
project Pythagoras I, and Herakleitos. Finally, the seventh frame-
work program (FP7), under Grand Agreement No. 229927, project
BIOSOLENUTI.
References
[1] S. Ferguson-Miller, G.T. Babcock, Chem. Rev. 96 (1996) 2889.
[2] T.L. Poulos, H.Y. Li, C.S. Raman, Curr. Opin. Chem. Biol. 3 (1999) 131.
[3] S. Iwata, C. Ostermeier, B. Ludwig, H. Michel, Nature 376 (1995) 660.
[4] C. Ostermeier, A. Harrenga, U. Ermlera, H. Michel, Proc. Natl. Acad. Sci. USA 94
(1997) 10547.
[5] T. Tsukihara, H. Aoyama, E. Yamashita, T. Tomizaki, H. Yamaguchi, K.
Shinzawa-Itoh, R. Nakashima, R. Yaono, S. Yoshikawa, Science 269 (1995)
1069.
[6] T. Tsukihara, H. Aoyama, E. Yamashita, T. Tomizaki, H. Yamaguchi, K.
Shinzawa-Itoh, R. Nakashima, R. Yaono, S. Yoshikawa, Science 272 (1996)
1136.
[7] S. Yoshikawa, K. Shinzawa-Itoh, R. Nakashima, R. Yaono, E. Yamashita, N.
Inoue, M. Yao, M.J. Fei, C.P. Libeu, T. Mizushima, H. Yamaguchi, T. Tomizaki, T.
Tsukihara, Science 280 (1998) 1723.
[8] R.H. Holm, E.I. Solomon, Chem. Rev. 104 (2004) 347.
[9] E.I. Solomon, R.K. Szilagyi, S.D. George, L. Basumallick, Chem. Rev. 104 (2004)
419.
[10] P.V. Rao, R.H. Holm, Chem. Rev. 104 (2004) 527.
[11] E.Y. Tshuva, S.J. Lippard, Chem. Rev. 104 (2004) 987.
[12] E. Kim, E.E. Chufan, K. Kamaraj, K.D. Karlin, Chem. Rev. 104 (2004) 1077.
[13] H.C. Liang, M. Dahan, K.D. Karlin, Curr. Opin. Chem. Biol. 3 (1999) 168.
[14] R.A. Ghiladi, T.D. Ju, D.H. Lee, P. Moenne-Loccoz, S. Kaderli, Y.M. Neuhold, A.D.
Zuberbuhler, A.S. Woods, R.J. Cotter, K.D. Karlin, J. Am. Chem. Soc. 121 (1999)
9885.
[15] E. Kim, M.E. Helton, I.M. Wasser, K.D. Karlin, S. Lu, H. Huang, P.M. Loccoz, C.D.
Incarvito, A.L. Rheingold, M. Honecker, S. Kaderli, A.D. Zuberbuhler, Proc. Natl.
Acad. Sci. USA 100 (2003) 3623.
[16] J.P. Collman, R. Boulatov, C.J. Sunderland, L. Fu, Chem. Rev. 104 (2004) 561.
[17] J.P. Collman, L. Fu, P.C. Herrmann, X. Zhang, Science 275 (1997) 949.
[18] J.P. Collman, L. Fu, P. Herrmann, Z. Wang, M. Rapta, M. Bröring, R.
Schwenninger, B. Boitrel, Angew. Chem., Int. Ed. 37 (1998) 3397.
[19] J.P. Collman, N.K. Devaraj, R.A. Decréau, Y. Yang, Y.L. Yan, W. Ebina, T.A.
Eberspacher, C.E.D. Chidsey, Science 315 (2007) 1565.
4. Conclusions
In conclusion, metalated porphyrins 1FeCu–3FeCu were effec-
tively synthesised, containing a covalently attached tridentate li-
gand site, a tyrosine molecule and a proximal base. UV–Vis and
NMR spectroscopies were used in order to investigate the coordi-
nation of the axial pyridine ligand with iron. Furthermore, the

2208
K. Ladomenou et al. / Inorganica Chimica Acta 363 (2010) 2201–2208
[20] K. Ladomenou, G. Charalambidis, A.G. Coutsolelos, Tetrahedron 63 (2007)
[28] J.P. Collman, J.I. Brauman, K.M. Doxsee, T.R. Halbert, E. Bunnenberg, R.E. Linder,
G.N. La Mar, J. Del Gaudio, G. Lang, K. Spartalian, J. Am. Chem. Soc. 102 (1980)
4182.
2882.
[21] G.E. Wuenschell, C. Tetreau, D. Lavalette, C.A. Reed, J. Am. Chem. Soc. 114
(1992) 3346.
[29] R.M. Kretzer, R.A. Ghiladi, E.L. Lebeau, H.-C. Liang, K.D. Karlin, Inorg. Chem. 42
(2003) 3016.
[22] C. Ruzié, P. Even, D. Ricard, T. Roisnel, B. Boitrel, Inorg. Chem. 45 (2006)
1338.
[30] B.B. Wayland, L.F. Mehne, J. Swartz, J. Am. Chem. Soc. 100 (1978) 2379.
[31] R.A. Ghiladi, K.D. Karlin, Inorg. Chem. 41 (2002) 2400.
[32] H. Shin, D.-H. Lee, C. Kang, K.D. Karlin, Electrochem. Acta 48 (2003) 4077.
[33] D. Ricard, M. L’Her, P. Richard, B. Boitrel, Chem. Eur. J. 7 (2001) 3291.
[34] R. Boulatov, J.P. Collman, I.M. Shiryaeva, C.J. Sunderland, J. Am. Chem. Soc. 124
(2002) 11923.
[35] J.P. Collman, R. Boulatov, C.J. Sunderland, in: K.M. Kadish, K.M. Smith, R.
Guilard (Eds.), The Porphyrin Handbook, vol. 11, Academic Press, Boston, 2003,
p. 1.
[23] G. Charalambidis, K. Ladomenou, B. Boitrel, A.G. Coutsolelos, Eur. J. Org. Chem.
(2009) 1263.
[24] B. Andrioletti, D. Ricard, B. Boitrel, New J. Chem. 23 (1999) 1143.
[25] D. Ricard, A. Didier, M. L’Her, B. Boitrel, ChemBioChem 2 (2001) 144.
[26] M. Nakamura, T. Ikeue, S. Neya, N. Funasaki, N. Nakamura, Inorg. Chem. 35
(1996) 3731.
[27] F.A. Walker, in: The Porphyrin Handbook, Academic Press, 2000, p. 81 (Chapter
36).