Electroreduction of Oxygen at a Pt/C-Modified Electrode with a Cobaltporphyrin Complex

Article abstract of DOI:10.1246/bcsj.77.401

Cobalt complexes that reversibly bind and release O₂ functioned as O₂ carriers, or O₂-enriching media, at the surface of a cathode to enhance the current for the reduction of O₂. The effect of the cobalt complex to accumulate O₂ was evidenced by the higher open-circuit potential, which reflected the equilibrated concentration of O ₂ at the electrode surface. The combination of the cobalt complex and the conventional Pt/C catalyst resulted in a significant increase in the steady-state current for the four-electron reduction of O₂, particularly at small overpotentials where typical fuel cells operate, based on the facilitated transport of O₂ from the atmosphere to the catalyst at the electrode surface.

Full text of DOI:10.1246/bcsj.77.401

ꢀ 2004 The Chemical Society of Japan
Bull. Chem. Soc. Jpn., 77, 401–405 (2004)
401
Electroreduction of Oxygen at a Pt/C-Modified Electrode
with a Cobaltporphyrin Complex
y
ꢀ
Baoqing Shentu, Kenichi Oyaizu, and Hiroyuki Nishide
Department of Applied Chemistry, Waseda University, Tokyo 169-8555
Received August 8, 2003; E-mail: nishide@waseda.jp
Cobalt complexes that reversibly bind and release O2 functioned as O2 carriers, or O2-enriching media, at the surface
of a cathode to enhance the current for the reduction of O2. The effect of the cobalt complex to accumulate O2 was evi-
denced by the higher open-circuit potential, which reflected the equilibrated concentration of O2 at the electrode surface.
The combination of the cobalt complex and the conventional Pt/C catalyst resulted in a significant increase in the steady-
state current for the four-electron reduction of O2, particularly at small overpotentials where typical fuel cells operate,
based on the facilitated transport of O2 from the atmosphere to the catalyst at the electrode surface.
The development of new materials designed to improve the
performance of polymer electrolyte fuel cells is a highly topical
research field.1–3 The basic elements of a typical fuel cell con-
sist of an electrolyte phase in contact with an anode and a cath-
ode. The fuel and oxidant gases flow past the backside of the
anode and cathode, respectively, and react electrochemically
in the region of the three-phase boundary established as the
gas/electrolyte/electrode interface. The nature of this interface
plays a critical role in the electrochemical performance of fuel
cells. Thus, the cell current and voltage depend on both the in-
terfacial electron transfer at the electrode and the transport of
O2 onto the electrode from the atmosphere.4 The inherent slow
kinetics of the electrochemical reactions of O2 have prompted
extensive studies to enhance the rate of electron transfer, which
have spawned a number of electrocatalysts.5–8 However, less
attention has been given to achieve electrode materials with en-
hanced O2 transport properties. Most of the previous studies
have been devoted to maintain a large interfacial contact area
using a porous structure (i.e., gas diffusion electrodes) to
achieve practical current densities. Indeed, recent issues for
the establishment of high-performance fuel cells concern cath-
ode materials to avoid ‘‘suffocation’’ of the fuel cell. For this
purpose, several specific properties would be required for the
cathode material, such as reversible O2 binding from the atmos-
phere, a rapid release of O2 to allow facile reduction at the elec-
trode, and durability under strongly acidic conditions.
trolyte solution to prevent an irreversible oxidation of the co-
balt(II) complex. Under those conditions, however, one must
apply such a large overpotential of ca. 1 V to reduce O2 that
the cell voltage becomes inevitably very low. Furthermore,
the liquid membrane of the organic solvent is disadvantageous
in handling use for practical applications. Here, we describe
that a simple tetraphenylporphyrinatocobalt(II)–benzylimida-
zole ([Co(tpp)(bim)]) complex, less susceptible to proton-in-
duced decomposition and the irreversible oxidation even in
strongly acidic media such as an aqueous HClO4 and a Nafion
membrane, can be employed as an effective O2-enricher. We
also report that a desirable system for fuel cell applications
can be realized when the O2-enriching layer is combined with
the conventional Pt/C catalyst, a platinum particle embedded in
a carbon particle. The [Co(tpp)(bim)] complex helps O2 reduc-
tion by complexing O2 from the electrolyte solution and there-
by increasing the local concentration of O2 at or near the active
catalytic surface. The main evidence for this is the higher open-
circuit potential and the larger reduction current of O2 at small
overpotentials in the presence of the cobalt complex. In this re-
port, we focus on the improved performance of the modified
electrode; a detailed elucidation of the mechanism of O2 en-
richment awaits future investigations.
Experimental
Materials.
5,10,15,20-Tetraphenylporphyrinatocobalt(II)
Cobalt complexes that reversibly bind and release O2 in re-
sponse to the O2 concentration are candidates for O2-enriching
materials.9 Recently, we reported that polymer–cobalt com-
plexes efficiently acted as O2-enriching materials, and that
the current for the reduction of O2 at a glassy carbon electrode
significantly increased compared with a control electrode with-
out the cobalt complex.10,11 The modified electrode composed
of a liquid membrane of picket-fence cobaltporphyrin in benzo-
nitrile adsorbed at the flat surface of the glassy carbon elec-
trode, which was placed in an aqueous buffered (pH = 10) elec-
([Co(tpp)]), benzylimidazole (bim), and Nafion 117 were pur-
chased from Aldrich Co. A platinum-embedded carbon particle
(Pt/C, ꢀ ¼ 40 nm) was obtained from Tanaka Kikinzoku Kogyo
Co. Perchloric acid (70%) was purchased from Kanto Chem. Co.
All solvents were purified by distillation in the usual manner prior
to use.
Preparation of the Modified Electrode.
Solutions of
[Co(tpp)] (1.3 mg, 0.0020 mmol) in THF (0.4 mL) and bim (1.2
mg, 0.0075 mmol) in THF (0.4 mL) were mixed (1/3.8 in mol/
mol) to complex the imidazolyl residue of bim with the fifth coor-
dination site of [Co(tpp)] (Chart 1). To the resulting solution was
dispersed the Pt/C catalyst (5.65 mg). The prepared dispersion
(5.0 mL) was carefully transferred to the surface of a glassy carbon
y Permanent address: Institute of Polymer Engineering, Zhejiang
Univeristy, Hangzhou 310027, China
Published on the web February 10, 2004; DOI 10.1246/bcsj.77.401

402 Bull. Chem. Soc. Jpn., 77, No. 2 (2004)
O2 Reduction at Pt/C–CoTPP
the rate-determining step for the overall electroreduction of O2.
The concentration of O2 (CO ;x, where x is the distance from
the electrode surface) is a function of an electrode potential
2
equilibrated with the O2/H2O redox couple (i.e., O2 + 4eꢁ
+
4Hþ = 2H2O). While a large overpotential is required to reduce
O2 at the bare surface of a glassy carbon electrode, the electrode
potential of the O2-reducing Pt/C surface is assumed to be
equilibrated with the O2/H2O couple. In this case, the electrode
potentials is approximated as
E ¼ E00 þ ðRT=4FÞ ln CO ;0
ð1Þ
2
Chart 1. Tetraphenylporphyrinatocobalt(II)–benzylimida-
zole ([Co(tpp)(bim)]) complex.
where
0
E0 ¼ E0 þ ðRT=FÞ ln CH ꢁ ðRT=2FÞ ln CH O
ð2Þ
þ
disk electrode (0.28 cm2), which was dried in air. Then, a solution
of Nafion (0.08 mg, 5.0 mL) was transferred to the electrode surface
to cover the composite layer of the catalyst and the cobalt complex.
The thickness of the Nafion film was ca. 1.7 mm, which was calcu-
lated based on the amount of Nafion, the surface area of the elec-
trode, and the density of Nafion (1.58 g cmꢁ3). After the solvent
was completely removed by evaporation, the resulting modified
electrode was immersed in an aqueous solution of HClO4 (0.1
M) to record the open-circuit potential and the current for the re-
duction of O2.
Measurements. Conventional electrochemical two-compart-
ment cells and commercially available instrumentations were em-
ployed. The glassy carbon (GC) disk was used as the working elec-
trode and polished before each experiment. The auxiliary elec-
trode, a coiled platinum wire, was separated from the working so-
lution by a fine-porosity frit. The reference electrode was a
saturated calomel electrode (SCE), which was placed in the main
cell compartment. The electrolyte solution was an aqueous solution
of HClO4 (0.1 M). The formal potential for the reduction of O2 to
H2O was Ef(O2/H2O) = 0.99 V vs this SCE. A Nikko Keisoku
DPGS-1 dual potentiogalvanostat and a Nikko Keisoku NFG-3
universal programmer were employed with a Graphtec WX
2400X-Y recorder to obtain voltammograms at room temperature.
2
Under strongly acidic aqueous conditions, E00 is assumed to be
constant. The bulk concentration of O2 (CO ;1) is reflected in
the rest potential (Er), or the open-circuit potential, and hence
the shift in rest potentials (ꢀEr) accompanied by a change in
the bulk O2 concentrations from C1 to C2 is given by
2
ꢀEr ¼ ðRT=4FÞ lnðC2=C1Þ
ð3Þ
Table 1 gives the rest potentials measured at a Pt/C-modified
electrode immersed in an aqueous solution of HClO4 saturated
with argon, air, or O2. It is known that O2 can permeate through
a Nafion membrane. Note that the potentials under argon might
have been influenced by the presence of unremovable O2 ad-
sorbed at the electrode surface (vide infra). On the other hand,
they evidently shifted positively with an increase in CO ;1
.
2
Without the cobalt complex, the potential shift of ꢀEr ¼ 0:01
V caused by the changes in electrolyte conditions from the
air- to the O2-saturated solutions is in almost agreement with
the calculated value of ꢀErcalc ¼ 0:01 V for C2=C1 ¼ 1=0:2
(atm/atm) (Eq. 3), although the potentials are supposed to be
lower than the equilibrated values because of the inherent slow
kinetics of the electroreduction of O2. However, the coinci-
dence of ꢀEr with the calculated value suggests that the rest po-
tentials are a good approximation of the equilibrium potentials
due to a lack of substantial current during the measurement. Re-
markably, Er shifted more positively in the presence of the co-
balt complex (Table 1), which could suggest an enhanced activ-
ity of the catalyst. We propose that the local concentration of
Results and Discussion
Tetraphenylporphyrinatocobalt(II) ([Co(tpp)]) has been
found to be a stable metal complex with O2-transporting prop-
erties.12,13 When [Co(tpp)] is complexed with bim as an axial
ligand, the O2-binding affinity of the [Co(tpp)(bim)] complex
is presumed to be still low but significantly enhanced with
p50 (partial pressure at which half of the cobalt binds O2) of
ca. 790 cmHg.12 Remarkably, the binding and releasing of O2
to and from the cobalt complex, determined by the laser flash
photolysis, are very rapid with rate constants of 1:3 ꢂ 109
O2 equilibrated at the electrode (CO ,0) could be even higher
than CO ;1 as a result of O2 enrichment by the [Co(tpp)(bim)]
complex. The enriched O2 roughly amounts to CO ;0=CO ;1
100 for a shift of ꢀEr ¼ 0:03 V in an O2 saturated solution
(Table 1). The remarkably high O2 concentration in the en-
riched layer suggests that it should be very thin, which is rem-
iniscent of the O2 hopping mechanism from the cobalt complex
2
2
ꢃ
2
2
M
ꢁ1 sꢁ1 and 1:7 ꢂ 108 sꢁ1, respectively.12 These results sug-
gest that the O2-binding and -releasing processes may not be
Table 1. Open-Circuit Potentials (Er) at Modified Electrodes ꢄImmersed in an Aqueous Solution
of HClO4 (0.1 M) Saturated with Argon, Air, or O2 at 20 C
Er under argon
(V vs SCE)
Er under air
(V vs SCE)
Er under O2
(V vs SCE)
Electrodea)
GC
Pt/C on GC
Pt/C + Co on GCb)
0.51
0.55
0.58
0.55
0.68
0.69
0.56
0.69
0.72
a) All electrodes were coated with Nafion (see Experimental Section). b) GC electrode modi-
fied with Pt/C and the [Co(tpp)(bim)] complex.

B. Shentu et al.
Bull. Chem. Soc. Jpn., 77, No. 2 (2004)
403
lapped with the peak for the reduction of the oxide layer near
0.38 V, and was gradually diminished in repeated scans as a re-
sult of a decrease in the active surface area. The presence of a
reduction peak near 0.3 V in the first scan, even in the argon-
saturated HClO4 solution, indicates that the Pt/C catalyst is al-
so reducing O2 irreversibly adsorbed at the surface of the cata-
lyst by virtue of exposure to the atmosphere during electrode
preparation (Fig. 1(a)). However, the peak current increased
in the O2-saturated solution (Fig. 1(b)) and became even larger
in the presence of the [Co(tpp)(bim)] complex (Fig. 1(c)). The
potential for the cobalt-catalyzed reduction of O2 is by ca. 0.4 V
more negative than that for the Pt/C-catalyzed reduction, and
hence the Pt/C surface reduces O2 much faster than the cobalt
complex, which means that the cobalt complex does not act as
an O2-reducing catalyst. We can only assume from the en-
hanced reduction current in the presence of the cobalt complex
that it acts as an O2-enriching medium.10,11
Current–potential curves obtained at a rotating disk electrode
modified with the cobalt complex and the Pt/C catalyst at var-
ious rotation rates are shown in Fig. 2(a). The steady-state cur-
rent at rotating electrodes corresponds to the flux of electroac-
tive species transported from the bulk solution to the electrode
surface, and thus does not contain any contribution from the ad-
sorbed species. The limiting currents correspond to the amount
of O2 reduced at sufficient overpotentials, and are controlled by
the diffusion-limited transport of O2 from the bulk solution. In-
deed, the limiting currents significantly increased with an in-
crease in the rotation rate of the electrode. The limiting currents
were plotted as the Koutecky–Levich plots in Fig. 2(b).15 The
number of electrons transferred during the O2 reduction
(n ¼ 3:5) can be roughly evaluated from the slope of the plots
by comparison with the diffusion-limited currents calculated
for the two- and four-electron reduction of O2 (dashed lines
in Fig. 2(b)).16 Note that the electrode area used is not the area
of the catalyst surface, but the physical area of the electrode,
because the Koutecky–Levich equation refers to the rate of
O2 transport from the bulk solution to the electrode surface.
Parallel experiments using electrodes with different physical
areas gave almost an identical number of electrons transferred
for the O2 reduction, suggesting that almost all of the electrode
surface was successfully covered with the catalyst. The limiting
currents were almost identical to, or slightly larger than, those
obtained without the cobalt complex, indicating that at least the
cobalt complex does not poison the Pt/C catalyst. The relative-
ly small kinetic current in Fig. 2(b) also indicates that the cobalt
complex does not impede the catalysis of O2 reduction by the
Pt/C surface. On the other hand, the very small dependence
of the current on the electrode rotation rate at smaller overpo-
tentials is considered to arise from the O2 enrichment effect.
This effect is more clearly shown in the steady-state polariza-
tion curves in Fig. 3 (vide infra).
Fig. 1. (a) Cyclic voltammogram recorded for an aqueous
solution of HClO4 (0.1 M) using a glassy carbon disk elec-
trode (A ¼ 0:28 cm2) modified with the Pt/C catalyst un-
der argon. The active surface area of Pt/C determined from
the amount of charge for the reoxidation of adsorbed H2
was 5.85 cm2. Scan rate = 25 mV sꢁ1. (b) Repeat of (a) un-
der O2. (c) Repeat of (a) using the electrode modified with
the [Co(tpp)(bim)] complex and the Pt/C catalyst under
O2. The active surface area of Pt/C was 5.58 cm2.
to the Pt/C catalyst, although a fuller mechanistic study was
outside the scope of this study.
When a large overpotential is applied, the diffusion layer
will be thicker than the enriched layer of O2, and thus the con-
tribution from the enriched O2 to the overall reduction current
will be overshadowed by the larger diffusion-limited mass
transfer from the bulk solution. In this case, the lack of poison-
ing of the Pt/C surface by the cobalt complex can be assessed
from the magnitude of the diffusion-limited current. For this
purpose, the behavior of the modified electrode was examined
by cyclic and rotating-disk voltammetry. Figure 1 shows cyclic
voltammograms obtained in argon- and O2-saturated HClO4
solutions. The reduction peak near ꢁ0:2 V and the oxidation
peak near ꢁ0:1 V corresponded to the reduction of protons in
the Nafion membrane and the re-oxidation of hydrogen mole-
cules adsorbed at the platinum surface, respectively. The
charge consumed for the oxidation of the adsorbed hydrogen
(H2(ad)) was used to evaluate the electrochemically active sur-
face area of the Pt particle on the electrode, assuming a value
of 220 mC cmꢁ2 for the oxidation of H2(ad) on the Pt surface.14
Throughout voltammetric experiments, Pt/C-modified elec-
trodes with almost the same active surface area (5.6–5.9 cm2)
were employed to allow a comparison of the electrode perform-
ance. The current for the reduction of O2 near 0.3 V was over-
The polarization curves for the kinetically limited steady-
state currents at the rotating disk electrode for the reduction
of O2 at small overpotentials, normalized by the active surface
area of the Pt/C catalyst, are shown in Fig. 3. When mass trans-
port and electron transfer occur in series, the rates of both proc-
esses must be the same at steady state. Thus, for the case where
electron transfer at the electrode surface is the rate-limiting
process,

404 Bull. Chem. Soc. Jpn., 77, No. 2 (2004)
O2 Reduction at Pt/C–CoTPP
Fig. 3. Steady-state polarization curves for O2 reduction on
rotated modified electrodes recorded in an O2-saturated
aqueous solution of HClO4 (0.1 M). The electrode was a
glassy carbon disk modified both with the [Co(tpp)(bim)]
complex and the Pt/C catalyst with the active surface area
of 5.85 cm2 ( ), or the disk modified only with the Pt/C
catalyst with the active surface area of 5.58 cm2 ( ). Open
circles ( ) and squares ( ) represent plots of data obtained
from repeated experiments using the same Pt/C-modified
electrodes with and without the [Co(tpp)(bim)] complex,
respectively. Electrode rotation rate = 900 rpm.
Fig. 2. (a) Current–potential curves for the reduction of O2
obtained with the rotated glassy carbon electrode modified
with the [Co(tpp)(bim)] complex and the Pt/C catalyst.
The active surface area of Pt/C was 5.58 cm2. The electro-
lyte solution was an O2-saturated aqueous HClO4 (0.1 M).
Electrode rotation rate was 400, 900, and 1600 rpm. Scan
rate = 5 mV sꢁ1. (b) Koutecky–Levich plots of the plateau
currents in (a) versus (electrode rotation rates)ꢁ1=2. Dashed
lines represent the diffusion limited currents for the two-
(n ¼ 2) and four-electron (n ¼ 4) reduction of O2 calculat-
Scheme 1.
ed using D ¼ 2:9 ꢂ 10ꢁ5 cm2 sꢁ1
,
C ¼ 1:3 ꢂ 10ꢁ6
mol cmꢁ3, v (kinetic viscosity) = 0.01 cm2 sꢁ1, and A ¼
0:28 cm2.
is sufficiently stable under these electrolyte conditions. The
similarity of the linear region and the slope of the plots to those
obtained without the cobalt complex confirms that the cobalt
complex does not take part in the catalysis at small overpoten-
kETCO ,0 ¼ i=nFA
ð4Þ
2
where kET is a heterogeneous rate constant as a function of the
electrode potential. In this case, the effect of O2 enrichment by
the cobalt complex can be manifested in the magnitude of the
current. The presence of the cobalt complex led to a significant
increase in the current density at a given potential, as shown in
Fig. 3. The larger steady-state current means that the transport
of O2 to the Pt/C surface can be facilitated both by the very rap-
id binding and releasing of O2 at the cobalt center and by the
high carrier concentration at the electrode surface. Thus, the co-
balt complex provides another path for O2 transport (path 2 in
Scheme 1), in addition to diffusion from the bulk solution (path
1). The highly reproducible values of the current, even after
prolonged electrolysis, indicated that the cobalt–Pt/C system
tials, but contributes only to increase CO ;0
.
2
Conclusion
The [Co(tpp)(bim)] complex that reversibly binds and re-
leases O2 functions as an O2-enriching medium for cathodes
to enhance the current for the reduction of O2. A combination
of the conventional Pt/C catalyst and the O2 enricher resulted
in a significant positive shift in the open-circuit potential, which
reflected the larger O2 concentration equilibrated at the elec-
trode. An increase in the current for the reduction of O2 was es-
tablished, particularly at small overpotentials where typical fuel
cells operate, based on the O2 enrichment effect of the cobalt
complex without poisoning of the Pt/C catalyst.

B. Shentu et al.
Bull. Chem. Soc. Jpn., 77, No. 2 (2004)
405
8
D. R. Lawson, L. D. Whiteley, C. R. Martin, M. N.
Szentirmay, and J. I. Song, J. Electrochem. Soc., 135, 2247 (1988).
H. Nishide and E. Tsuchida, ‘‘Transport Phenomena and
Separation of Small Molecules,’’ in ‘‘Macromolecules–Matal
This work was partially supported by a Grant-in-Aid for
Scientific Research (No. 13031072) and for 21COE Research
‘‘Practical Nano-Chemistry’’ from MEXT. B. S. expresses
her thanks to the scholarship from MEXT.
9
Complexes,’’ ed by F. Ciardelli, E. Tsuchida, and D. Wohrle,
¨
Springer-Verlag, Berlin (1996), Chap. 4.2, pp. 175–189.
10 B. Shentu, K. Oyaizu, and H. Nishide, Chem. Lett., 2002,
712.
11 K. Oyaizu, H. Nakano, B. Shentu, and H. Nishide, J. Mater.
Chem., 12, 3162 (2002).
12 H. Shinohara, T. Arai, and H. Nishide, Macromol. Symp.,
186, 135 (2000).
13 J. Won, Y. S. Kang, and H. Nishide, ‘‘Binding of Small
Molecules,’’ in ‘‘Metal Complexes and Metals in Macromole-
cules,’’ ed by D. Wohrle and A. D. Pomogailo, Wiley-VCH,
¨
Weinheim (2003), Chap. 9, pp. 363–404.
14 E. Antolini, R. R. Passos, and E. A. Ticianelli, Electrochim.
Acta, 48, 263 (2002).
15 A. J. Bard and L. R. Faulkner, ‘‘Electrochemical Methods,’’
2nd ed, John Wiley, New York (2001).
16 E. Yeager, J. Mol. Catal., 38, 5 (1986).
References
1 M. Watanabe, Y. Zhu, and H. Uchida, J. Phys. Chem. B,
104, 1762 (2000).
2 A. Noda, M. A. B. H. Susan, K. Kudo, S. Mitsushima, K.
Hayamizu, and M. Watanabe, J. Phys. Chem. B, 107, 4024 (2003).
3 K. Miyatake, H. Zhou, H. Uchida, and M. Watanabe, Chem.
Commun., 2003, 368.
4 K. Kinoshita, ‘‘Electrochemical Oxygen Technologies,’’
John Wiley, New York (1992).
5
J. Maruyama, M. Inaba, T. Morita, and Z. Ogumi, J.
Electroanal. Chem., 504, 208 (2001).
J. Maruyama, M. Inaba, and Z. Ogumi, Electrochim. Acta,
45, 415 (1999).
6
7 J. B. Floriano, E. A. Ticianelli, and E. R. Gonzalez, J.
Electroanal. Chem., 367, 157 (1994).