Effects of p-substituents on electrochemical CO oxidation by Rh porphyrin-based catalysts

Article abstract of DOI:10.1039/b925413g

Electrochemical CO oxidation by several carbon-supported rhodium tetraphenylporphyrins with systematically varied meso-substituents was investigated. A quantitative analysis revealed that the p-substituents on the meso-phenyl groups significantly affected CO oxidation activity. The electrocatalytic reaction was characterized in detail based on the spectroscopic and X-ray structural results as well as electrochemical analyses. The difference in the activity among Rh pophyrins is discussed in terms of the properties of p-substituents along with a proposed reaction mechanism. Rhodium tetrakis(4-carboxyphenyl)porphyrin (Rh(TCPP)), which exhibited the highest activity among the porphyrins tested, oxidized CO at a high rate at much lower potentials (<0.1 V vs. a reversible hydrogen electrode, at 60°C) than the present PtRu catalysts. This means that CO is electrochemically oxidized by this catalyst when a slight overpotential is applied during the operation of a proton exchange membrane fuel cell. This catalyst exhibited little H₂ oxidation activity, in contrast to Pt-based catalysts.

Full text of DOI:10.1039/b925413g

View Article Online / Journal Homepage / Table of Contents for this issue
PAPER
www.rsc.org/pccp | Physical Chemistry Chemical Physics
Effects of p-substituents on electrochemical CO oxidation
by Rh porphyrin-based catalystsw
Shin-ichi Yamazaki,*a Yusuke Yamada,a Sahori Takeda,b Midori Goto,c
Tsutomu Ioroi,a Zyun Siromaa and Kazuaki Yasudaa
Received 2nd December 2009, Accepted 12th April 2010
DOI: 10.1039/b925413g
Electrochemical CO oxidation by several carbon-supported rhodium tetraphenylporphyrins
with systematically varied meso-substituents was investigated. A quantitative analysis
revealed that the p-substituents on the meso-phenyl groups significantly affected CO oxidation
activity. The electrocatalytic reaction was characterized in detail based on the spectroscopic
and X-ray structural results as well as electrochemical analyses. The difference in the activity
among Rh pophyrins is discussed in terms of the properties of p-substituents along with a
proposed reaction mechanism. Rhodium tetrakis(4-carboxyphenyl)porphyrin (Rh(TCPP)), which
exhibited the highest activity among the porphyrins tested, oxidized CO at a high rate at much
lower potentials (o0.1 V vs. a reversible hydrogen electrode, at 60 1C) than the present
PtRu catalysts. This means that CO is electrochemically oxidized by this catalyst when
a slight overpotential is applied during the operation of a proton exchange membrane
fuel cell. This catalyst exhibited little H2 oxidation activity, in contrast to Pt-based
catalysts.
catalysts is difficult during fuel cell operation. The high over-
1. Introduction
potential for CO oxidation prevents CO removal, and results in
CO poisoning of Pt catalysts.
Carbon monoxide (CO), which is strongly adsorbed on the Pt
anode catalyst in a proton exchange membrane fuel cell
(PEMFC), interferes with H2 oxidation on the catalyst.1,2 Even
in the presence of less than 100 ppm CO, H2 oxidation activity
dramatically decreases.3 Many CO conversion catalysts4,5 and
CO-tolerant anode catalysts6–12 have been developed to solve
this problem of CO poisoning. However, CO poisoning remains
a serious problem in the realization of a stationary PEMFC.
CO can be removed from the surface of a Pt catalyst by
electrochemical oxidation [eqn (1)].
If CO could be removed by electrochemical oxidation
during fuel cell operation, it may be possible to counteract
CO poisoning in PEMFC. We previously reported that
carbon-supported Rh octaethylporphyrin (Rh(OEP)/C) can
catalyze electrochemical CO oxidation at a high rate at low
overpotentials.13 We also demonstrated a direct CO-PEMFC
based on the high CO oxidation activity of Rh(OEP).14 This
PEMFC uses CO, which is a strong poison for Pt catalyst, as a
fuel and delivers high power (>95 mA cmꢀ2 at 0.47 V, 44 mW
cmꢀ2). The overpotential for CO oxidation with Rh(OEP)/C is
much lower than that with PtRu catalyst. However, the onset
potential for CO oxidation is still slightly higher than 0 V vs. a
reversible hydrogen electrode (RHE). Thus, the onset potential
needs to be decreased.
CO + H2O - CO2 + 2H+ + 2eꢀ
(1)
However, the overpotential for CO electro-oxidation on a
Pt-based catalyst is extremely high. Even PtRu catalysts,
which are usually used as CO-tolerant anode catalysts, cata-
lyze CO oxidation at much higher potentials (more than
ca. 0.3 V8) than the potential regions for PEMFC (at least
below 0.1 V). Electrochemical removal of CO by Pt-based
In some electrochemical reactions such as thiol oxidation
and disulfide reduction, the substituents have significant
effects on the electrocatalytic activity of metalloporphyrins
or metallophthalocyanins.15,16 Also, in the case of Rh porphyrins,
a change of substituents may improve the activity of CO oxidation.
To increase the activity by changing substituents, it is
important to understand the relationship between the substi-
tuents and activity. Such an understanding can be obtained
by clarifying the mechanism of the electrocatalytic reaction.
In previous studies, the effects of the substituents on thiol
oxidation and disulfide reduction could be clearly explained by
the mechanisms of these reactions.15,16 This information
should be useful for identifying good catalysts for the reaction.
A mechanistic analysis of electrocatalytic CO oxidation by Rh
porphyrins should also be helpful for the design of substituents
for improving the catalysts.
a Research Institute for Ubiquitous Energy Device, National Institute
of Advanced Industrial Science and Technology (AIST),
1-8-31 Midorigaoka, Ikeda, Osaka 563-8577, Japan.
E-mail: s-yamazaki@aist.go.jp
b Research Institute for Innovation in Sustainable Chemistry,
National Institute of Advanced Industrial Science and Technology
(AIST), 1-8-31 Midorigaoka, Ikeda, Osaka 563-8577, Japan
c Technical Service Center, National Institute of Advanced Industrial
Science and Technology (AIST), 1-1-1 Higashi, Tsukuba,
Ibaraki 305-8565, Japan
w Electronic supplementary information (ESI) available: Details of
experimental information, electrochemical data, UV-vis spectra, and
X-ray crystallography. CCDC reference numbers 729998 and 769814.
For ESI and crystallographic data in CIF or other electronic format
see DOI: 10.1039/b925413g
ꢁc
This journal is the Owner Societies 2010
8968 | Phys. Chem. Chem. Phys., 2010, 12, 8968–8976

View Article Online
In this study, we paid special attention to the effects of
substituents on the activity, and to the mechanism of the
electrocatalytic CO oxidation. We concentrated on the effects
of the p-substituents of the meso-phenyl group attached to the
porphyrin ring, since the p-substituent effect can be discussed
systematically by the Hammett approach. In this paper,
several Rh tetraphenylporphyrins (Scheme 1) with a wide
variety of p-substituents were examined with regard to their
CO oxidation activity. The effects of p-substituents on the CO
oxidation activity were examined quantitatively and systemati-
cally. The mechanism of CO oxidation was investigated by
electrochemical, structural, and spectroscopic approaches. We
found a drastic difference in CO oxidation activity among
carbon-supported Rh tetraphenylporphyrins, even though
their chemical structures are different only in the p-substituents.
The effects of p-substituents are also discussed from mecha-
nistic viewpoints. We found that a rhodium 5,10,15,20-tetra-
kis(4-carboxyphenyl)porphyrin (Rh(TCPP))-derived catalyst
exhibited the highest activity among the Rh porphyrin
catalysts tested. This complex can electrochemically oxidize
CO below 0.1 V (vs. RHE), which has not been previously
realized using Pt-based anode catalysts.
of the synthesis are described in the ESI.w The abbrevia-
tions for the complexes are as follows; [RhIII(TCPP)]+
(TCPP
= 5,10,15,20-tetrakis(4-carboxyphenyl)porphinate),
[RhIII(T(–COOCH3)PP)]+ (T(–COOCH3)PP = 5,10,15,20-tetra-
kis(4-methylcarboxyphenyl)porphinate), [RhIII(T(–OCH3)PP)]+
(T(–OCH3)PP = 5,10,15,20-tetrakis(4-methoxyphenyl)porphinate),
[RhIII(TPP)]+ (TPP
= 5,10,15,20-tetraphenylporphinate),
[RhIII(T(–CH3)PP)]+ (T(–CH3)PP = 5,10,15,20-tetra(4-methyl-
phenyl)porphinate), [RhIII(TFPP)]+ (TFPP = 5,10,15,20-tetra-
kis(4-fluorophenyl)porphinate), and [RhIII(TBPP)]+ (TBPP =
5,10,15,20-tetrakis(4-bromophenyl)porphinate).
2.3 Adsorption of Rh porphyrins on a carbon support by the
evaporation-to-dryness (impregnation) method
Carbon-supported [RhIII(porphyrin)]+ (Rh(porphyrin)/C)
was prepared by the evaporation-to-dryness (impregnation)
method as follows. Carbon Black (Vulcan XC 72R) (30 mg)
was added to 20 mL of appropriate solvent (see below)
containing a porphyrin in a sample vial. The suspension was
shaken vigorously, subjected to sonication for a few minutes,
and stirred for 30 min. The solvent was removed with a rotary
evaporator and the resulting powder was collected. The
amount of Rh porphyrins deposited was fixed at 30 mmol
gꢀ1
black. The solvents used in this preparation were
2. Experimental
carbon
ethanol for Rh(TCPP)/C and Rh(T(–CH3)PP)/C, and CH2Cl2
for the other catalysts.
2.1 General
Tetracarbonyldi-m-chlorodirhodium(I) (Rh2Cl2(CO)4) was
purchased from Wako Chemicals. All of the porphyrin ligands
used in this study are commercially available. The water used
in this study was purified using an Organo Puric-MX unless
otherwise stated. UV-vis spectra were measured using a
Shimadzu UV-1500 PC photodiode array spectrophotometer
equipped with a quartz cell (light path = 1 cm). Electrochemical
measurements were performed using an ALS electrochemical
analyzer (Model 627B) equipped with a rotation controller
(BAS RDE-2). ESI-MS spectroscopy was performed using an
Esquire3000plus ion trap mass spectrometer (Bruker Daltonics).
IR spectra were measured using a FT-IR spectrophoto-
meter (IFS-66v/S, Bruker). Product analysis was performed
by a gas chromatograph (Shimadzu, GC 2014) with a TCD
detector.
2.4 Adsorption of a Rh porphyrin on a carbon support (carbon
black, Vulcan XC 72R) by the equilibrium adsorption method
Rh(TCPP)/C was also prepared by the equilibrium adsorption
method as follows. Two conditions were tested for prepara-
tion. The two conditions differed by the ratio of amount of
carbon to the volume of solvent.
Condition A: Carbon Black (Vulcan XC 72R) (30 mg) was
added to 10 mL of ethanol containing porphyrin (0.46 mM) in
a sample vial. The suspension was homogenized as described
above. The suspension was stirred for 3 h and filtered through
filter paper (ADVANTEC, No.5C). The pellet was then
collected.
Condition B: Carbon Black (Vulcan XC 72R) (40 mg) was
added to 5 mL of ethanol containing porphyrin (0.46 mM) in a
sample vial. Other procedures were the same as in
Condition A.
2.2 Synthesis
Rh porphyrins were synthesized by refluxing Rh2Cl2(CO)4
and the corresponding ligands in appropriate solvents, and
purified by a conventional method. The product was analyzed
by elemental analysis and ESI-MS. The experimental details
The equilibrium amount of porphyrin adsorbed per unit
carbon mass was determined by measuring the filtrate con-
centration by UV-vis spectroscopy after the adsorption.
2.5 Electrochemical measurements
A modified rotating disk glassy carbon electrode (modified
RDE, described below) was used as a working electrode. A
Ag|AgCl|KCl(sat.) and a reversible hydrogen electrode (RHE)
were used as reference electrodes in the experiments at 25 and
60 1C, respectively. In the experiments at 60 1C, RHE was
placed at room temperature. All of the potentials in the results
are vs. RHE at the temperature at which the measurements
were performed (see ESI).w A platinum coil was used as a
Scheme 1 Structures of Rh porphyrins tested in this work.
ꢁc
This journal is the Owner Societies 2010
Phys. Chem. Chem. Phys., 2010, 12, 8968–8976 | 8969

View Article Online
counter electrode. The measurement was performed in a 0.1 M
H2SO4 solution. All voltammograms were measured at a scan
rate of 10 mV sꢀ1
.
2.6 Modification of RDE with Vulcan XC 72R-supported Rh
porphyrins
RDE was modified as described in the literature.17 Five
milligrams of the Vulcan XC 72R-supported Rh porphyrins
was dispersed in 0.5 mL of a mixed solvent (water–ethanol =
1 : 1) containing 5 mL Nafion (Aldrich, 5%). A portion
(2 mL) of this suspension was dropped onto a rotating glassy
carbon (GC) disk electrode (BAS) and the solvent was allowed
to evaporate at room temperature. The catalyst (0.02 mg)
was then mounted on a glassy carbon working electrode
(0.0707 cm2).
2.7 Structure determination by X-ray crystallography
Crude [RhIII(TCPP)(CO)(Cl)] and [RhIII(T(–OCH3)PP)(C2H5OH)-
(Cl)] complexes were recrystallized to obtain fairly good single
crystals suitable for X-ray analysis from a mixed solvent
(2-propanol and hexane) and ethanol, respectively. X-Ray
diffraction data were collected by a Bruker SMART AXS
CCD area-detector diffractometer using Mo Ka radiation
(l = 0.71073 A) at a temperature of 183 K. All crystallo-
graphic calculations were conducted using SHELXTL18 (Bruker
AXS Inc.). Other details are given in the ESI.w
Fig. 1 ORTEP drawings of (A)[RhIII(TCPP)(C3H7OH)(Cl)](H2O)-
(C3H7OH)2 and (B) [RhIII(T(–OCH3)PP)(C2H5OH)(Cl)] with 25%
probability ellipsoids. Hydrogen atoms and uncoordinated solvents
are omitted for clarity.
by recrystallization from a mixed solvent system (2-propanol :
hexane = 2 : 9) and from ethanol, respectively. Their crystal
structures were determined by single-crystal X-ray diffraction
analysis, and are shown in Fig. 1. In the crystal structure of
[RhIII(TCPP)(C3H7OH)(Cl)], C3H7OH, which is used for crystalli-
zation, coordinates on the Rh atom. This indicates that the
axial ligands of a Rh porphyrin are labile. In both structures,
phenyl groups at the meso positions are situated at various angles
(94, 61, 127, and 821 in [RhIII(TCPP)(C3H7OH)(Cl)] and 66, 64, 71
and 731 in [RhIII(T(–OCH3)PP)(C2H5OH)(Cl)], see ESIw) vs. the
porphyrin ring. The overlap between the orbitals of the
porphyrin ring and meso-phenyl groups is reduced in this
configuration. Hence the electronic effect of the p-substituents
on meso-phenyl groups would be slight. In the structure of
[RhIII(TCPP)(C3H7OH)(Cl)], a molecule is connected to the
neighboring molecules via hydrogen bonds between carboxy
groups. Some solvent molecules participate in these hydrogen
bonds. The O6 and O7 atoms form direct hydrogen bonds
without participation of the solvent. In the structure of
[RhIII(T(–OCH3)PP)(C2H5OH)(Cl)], the O4 atom participates
in the formation of a hydrogen bond with the O1 atom in
ethanol.
2.8 Product analysis
Electrolysis for analysis of the products of CO oxidation by
Rh(TCPP)/C was performed as follows; a 0.1 M H2SO4
solution was placed in a cell. This cell was composed of two
compartments separated by a glass filter. The Rh(TCPP)/
C-modified working electrode (A = 4.5 cm2) and the reference
electrode were immersed in one compartment. The counter
electrode was immersed in the other compartment. The com-
partment with the working and reference electrodes was
purged with neat CO gas, and the compartment with the
counter electrode was purged with argon gas. After the two
compartments were purged, the cell was sealed with a silicone
plug to prevent gas exchange during electrolysis. The potential
of the working electrode was held at 0.31 V (vs. RHE) at
25 1C. The charge was recorded during electrolysis. After the
electrolysis was finished, the gas phase above the solution was
sampled and analyzed by gas chromatography.
3. Results
3.2 Electrochemical CO oxidation activity of
carbon-supported Rh porphyrins prepared by the
evaporation-to-dryness (impregnation) method
3.1 Characterization of Rh porphyrin complexes
Elemental analysis and ESI-MS spectroscopy clearly indicated
the generation of trivalent Rh porphyrin complexes. UV-vis
spectra of the Rh porphyrins are shown in Fig. S1.w All Rh
porphyrins exhibited similar spectra regardless of the electronic
properties of the p-substituents. The p-substituents have little
effect on the electronic properties of the porphyrin ring in Rh
porphyrins.
Since Rh porphyrins were used as an electrocatalyst, they were
adsorbed on Carbon Black, which has good conductivity.
Adsorption was performed by the evaporation-to-dryness
method. In this method, the amount of Rh porphyrins on
the carbon support can be controlled. The amounts were fixed
at 30 mmol gꢀ1. We measured the electrochemical CO oxida-
tion activity by carbon-supported Rh porphyrins prepared by
this method.
Single crystals of [RhIII(TCPP)(C3H7OH)(Cl)] and
[RhIII(T(–OCH3)PP)(C2H5OH)(Cl)] complexes were obtained
ꢁc
This journal is the Owner Societies 2010
8970 | Phys. Chem. Chem. Phys., 2010, 12, 8968–8976

View Article Online
Fig. 2 shows voltammograms of several carbon-supported
Rh porphyrins under CO atmosphere with electrode rotation
at 6400 rpm in 0.1 M H2SO4. Rh porphyrins exhibited clear
CO oxidation activity. Carbon-supported inorganic salts of
Rh such as Rh2Cl2(CO)4 showed little CO oxidation activity
compared to Rh porphyrins (data not shown), while some
researchers have reported CO oxidation activity with a
similar water-soluble Rh complex ([Rh(CO)2Br2]ꢀ).19 CO
electro-oxidation by Rh black required large overpotentials
(data not shown), as in the case of Pt catalyst. These data
suggest that the porphyrin structure is very important for a
high rate of CO oxidation in low potential regions.
A significant difference was observed in the activities of the
Rh porphyrin catalysts. Some of the Rh porphyrins tested
exhibited strong CO oxidation activity even in low potential
regions (o0.25 V vs. RHE). Especially with Rh(TCPP)/C,
Faradaic anodic current starts to flow at ca. 0.1 V vs. RHE. In
contrast, Rh(TPP), Rh(TFPP), Rh(TBPP), and Rh(T(–CH3)PP)
exhibit significantly lower activity. The only difference in their
structures is the difference between the p-substituents on the
meso-phenyl group. This difference in the activity is somewhat
surprising if we consider that the meso-phenyl groups in tetra-
phenylporphyrin ligands are not conjugated with a porphyrin
ring (Fig. 1) and the substituents on the meso-phenyl group
would have a less significant effect on the electronic property of
the ring.
Fig.
2
Linear sweep voltammograms of (a) Rh(TCPP)/C,
(b) Rh(T(–COOCH3)PP)/C, (c) RhT(–OCH3)PP/C, (d) Rh(T(–CH3)PP)/C,
(e) Rh(TPP)/C, (f) Rh(TBPP)/C, and (g) Rh(TFPP)/C prepared
by the equilibrium adsorption method. The measurements were
performed under a CO atmosphere with electrode rotation at 6400 rpm
(scan rate = 10 mV sꢀ1) in 0.1 M H2SO4 at 25 1C.
The activity is also not related to steric hindrance by
p-substituents. Thus, the drastic difference in activity in
Fig. 2 cannot be explained by the difference in the properties
of porphyrin alone. This strongly suggests that the configura-
tion of porphyrin molecules on the surface of the carbon
support might play a role in this difference in activity.
The dependence of the activity on the electrode rotation rate
was also examined with all of the catalysts tested. The current
below 0.25 V did not depend on the rotation rate. Therefore, the
currents below 0.25 V should be free of the effect of diffusion in
bulk solution and reflect the activity of Rh porphyrin catalysts.
CO oxidation in these potential regions (o0.25 V vs. RHE) is
important for fuel cell applications. We adopted the current at
0.25 V as an index of the catalytic activity since in the very low
potential regions (o0.25 V) some catalysts hardly exhibit activity
and quantitative comparison is difficult. The results are shown
in Table 1. The CO oxidation activity at 0.25 V increases in
3.3 Electrochemical CO oxidation activity of carbon-
supported Rh porphyrins prepared by the equilibrium adsorption
method
The possible importance of the configuration of the molecules
on the surface encouraged us to examine the effect of the
preparation method on the activity of Rh porphyrin-based
catalysts. Therefore, we tested the equilibrium adsorption
method for preparing carbon-supported Rh porphyrins for
comparison to the results with the evaporation-to-dryness
method. Two different conditions (A and B, see section 2.4.)
were used to prepare the catalyst using the equilibrium adsorp-
tion method.
the order: Rh(TCPP)/C
Rh(T(–OCH3)PP)/C Rh(T(–CH3)PP)/C
Rh(TFPP)/C, Rh(TBPP)/C (Table 1).
>
Rh(T(–COOCH3)PP)/C
>
>
>
Rh(TPP)/C,
In general, the evaporation-to-dryness method can be used
to control the amount of complexes on Carbon Black. However,
a disadvantage is that even molecules that interact less with
carbon might also be adsorbed on the catalyst surface, so as to
cause the undesirable aggregation and multilayer adsorption
of the porphyrins. In contrast, in the equilibrium adsorption
method, molecules that interact less with carbon would be
washed away, and only molecules that strongly interact with
There appears to be no clear relationship between the
activities of the catalysts and the Hammett constants of the
p-substituents on the meso-phenyl groups (Table 1). For
example, although –COOH, –COOCH3, –Br, and –F are all
electron-withdrawing groups, their activities are totally different.
This indicates that the electronic properties of p-substituents
are hardly correlated with the rate of CO electro-oxidation.
Table 1 Hammett constant of p-substituent and CO oxidation activity of Rh porphyrins
Hammett constant of
p-substituent (sp)
CO oxidation current
at 0.25 V/mA
Porphyrins
Rh(TCPP)
0.44
0.44
0.22
0.06
0
46
20
Rh(T(–COOCH3)PP)
Rh(TBPP)
Rh(TFPP)
Rh(TPP)
Rh(T(–CH3)PP)
Rh(T(–OCH3)PP)
2.4
0.4
1.4
3.4
6.7
ꢀ0.14
ꢀ0.28
ꢁc
This journal is the Owner Societies 2010
Phys. Chem. Chem. Phys., 2010, 12, 8968–8976 | 8971

View Article Online
method. We were able to discuss the meso-substituent
effect and the superiority of the equilibrium adsorption
method using this quantitative evaluation method, and dis-
covered a new CO oxidation catalyst that has high activity
per Rh atom, while the study by van Baar et al. paid little
attention to the effects of meso-substituents and the adsorption
method.
3.4 Product analysis of the electrochemical CO oxidation
activity of carbon-supported Rh porphyrins prepared by the
equilibrium adsorption method
To investigate the reaction product and the overall number of
electrons transferred, the dependence of current on the rota-
tion rate was examined (Fig. S2). In potential regions above
0.3 V, the current slightly increased with the rotation rate;
diffusion in the bulk affects the current. The Koutecky-Levich
plot at 0.45 V is shown in Fig. S2. The number of electrons
transferred in CO oxidation is calculated to be 2.16 from the
slope of this plot (see ESIw). This value is close to 2, and thus
stoichiometry is consistent with reaction (1).
Fig. 3 Linear sweep voltammograms of (A) Rh(TCPP)/C prepared
by the equilibrium adsorption method (condition A), (B) Rh(TCPP)/C
prepared by the equilibrium adsorption method (condition B), and
(C) Rh(TCPP)/C prepared by the evaporation-to-dryness method
(taken from Fig. 2). The measurements were performed under a
CO atmosphere with electrode rotation at 6400 rpm (scan rate =
10 mV sꢀ1) in 0.1 M H2SO4 at 25 1C.
carbon would be adsorbed on the catalyst surface. Aggrega-
tion and multilayer adsorption can be minimized with this
method. The amount of porphyrins (Table 1) adsorbed on
Carbon Black was evaluated by measuring the difference
between concentrations before and after adsorption.
This stoichiometry was also confirmed by a product analysis.
The generation of a significant amount of CO2 from the electro-
chemical oxidation of CO was verified using gas chromato-
graphy. The amount of CO2 generated (14 mmol) is nearly
equivalent to half the amount of electrons transferred (38 mmol).
This also supports the stoichiometry of eqn (1).
The activities of Rh(TCPP)/C prepared with the equilibrium
adsorption method are shown in Fig. 3 and Table 2 in
comparison to the activity of the catalyst prepared by the
evaporation-to-dryness method. The results in Fig. 3 clearly
indicate that the activity is increased with the equilibrium
adsorption method. The activity per Rh atom was also
increased with the equilibrium adsorption method (Table 2).
Especially with catalyst (B), even though the amount adsorbed
is less than with the evaporation-to-dryness method, higher
activity was observed. Such a drastic difference in activity
between the two preparation methods supports the hypothesis
that the configuration of porphyrin molecules on the surface of
Carbon Black is important for their activity. The onset potential
for CO oxidation with the catalyst (A) is much lower than those
for PtRu catalysts and for Rh(OEP)/C reported previously.13
The catalysts prepared with this method (condition A) were used
for further experiments.
3.5 Electrochemical H2 oxidation activity of Rh(TCPP)/C
To discuss the application of Rh porphyrin-based catalysts to
fuel cell technology, we also examined their H2 oxidation
activity. Fig. 4 shows voltammograms of Rh(TCPP)/C prepared
by the equilibrium adsorption method (condition A) in atmo-
spheres of (a) argon and (b) H2. The change from argon to H2
only slightly increased the anodic current. Electrode rotation
did not increase the current (line c), indicating that the H2
oxidation current is completely determined by the kinetic
activity of the catalyst. This indicates that Rh(TCPP)/C has
slight H2 oxidation activity, but this activity is very low
compared to its CO oxidation activity (Fig. 3). If we subtract
the anodic current of line (a) from that of line (b), we obtain
the intrinsic H2 oxidation current. The current at 0.25 V is
ca. 1/110 of that observed in CO oxidation. The saturated
concentration of H2 (0.78 mM) is comparable to that of CO
(1 mM) based on their solubility. Considering the difference in
solubility, the activity of Rh(TCPP)/C toward CO is 85 times
as high that toward H2. Our previous study also suggested that
the Rh(OEP) catalyst was selective in a PEMFC.14 van Veen
et al. stated that Rh porphyrin catalyst has high selectivity
In their pioneering work, van Baar et al. described the
electrochemical CO oxidation activity of carbon-supported
Rh tetraphenylporphyrins, as measured by a floating electrode
technique.20,21 However, the activity per Rh atom is low,
and their evaluation methods are unsuitable for quantitative
electrochemical analysis. In this study, we measured CO
oxidation activity using a quantitative rotating electrode
Table 2 CO oxidation activity of the Rh(TCPP)/C catalysts prepared with several methods
CO oxidation current at
0.25 V per Rh porphyrin
1 mmol/mA mmolꢀ1
Adsorption amount
(mmol per 1 g Vulcan XC 72R)
CO oxidation current
at 0.25 V/mA
Preparation methods
Equilibrium adsorption method
Condition A
Condition B
49
22
30
111
95
46
113
216
77
Evaporation-to-dryness method
ꢁc
This journal is the Owner Societies 2010
8972 | Phys. Chem. Chem. Phys., 2010, 12, 8968–8976

View Article Online
Fig. 4 Voltammograms (scan rate = 10 mV sꢀ1) of Rh(TCPP)/C
prepared by the equilibrium adsorption method (condition A) under
(a) an argon atmosphere, (b) a H2 atmosphere, and (c) a H2 atmo-
sphere with electrode rotation at 6400 rpm. Other details of the
measurements are shown in the captions for Fig. 2 and 3.
using CO–H2 and CO–N2 mixtures.22 They also demonstrated
H2 oxidation by heat-treated noble-metal porphyrins.22
However, there has been no previous demonstration of the
quantitative relation between H2 and CO oxidation activities
of Rh porphyrins. The present experiments directly and
quantitatively demonstrate the selectivity of Rh(TCPP)/C
toward CO. The high selectivity is a distinguishing property
of porphyrin-based catalysts compared to bulk noble-metal
catalysts.
Fig. 5 The CO oxidation activity at 60 1C of Rh(TCPP)/C prepared
by the equilibrium adsorption method (condition A). (A) Voltammo-
grams of Rh(TCPP)/C under (a) an argon atmosphere, (b) a CO
atmosphere without electrode rotation, and a CO atmosphere with
electrode rotation at (c) 400, (d) 900, (e) 1600, (f) 2500, (g) 3600,
(h) 4900, and (i) 6400 rpm (scan rate = 10 mV sꢀ1). (B) A logarithmic
plot of potential vs. current of CO oxidation by Rh(TCPP)/C.
3.6 Electrochemical CO oxidation activity of Rh(TCPP)/C
at 60 1C.
The CO oxidation activity of Rh(TCPP)/C at a higher tempera-
ture (60 1C) was also examined because actual fuel cell
catalysts are used at a somewhat higher temperature. The
experiment at 60 1C is also important for the analysis of the
dependence of activity on temperature. The results are shown
in Fig. 5A. The increase in temperature reduced the over-
potentials for CO oxidation. The current started to flow below
0.1 V, and the current at 0.15 V exceeded 70 mA, while the
current at 0.15 V at 25 1C was only 8 mA (Fig. 3A).
The results of a Tafel analysis of Rh(TCPP)/C from 0.08 V
to 0.12 V are shown in Fig. 5B. In these potential regions, the
current does not depend on the rotation rate (Fig. 5A),
indicating that the rate is a completely kinetic process. Hence,
the current at 6400 rpm is adopted as a kinetic current. In this
figure, the logarithms of the kinetic currents are plotted
against the potentials. The slope of this plot (Tafel slope) is
62 mV/decade. Other Rh porphyrin-based catalysts
(RhT(–COOCH3)PP/C, RhT(–OCH3)PP/C, and Rh(TPP)/C)
gave similar Tafel slopes (69, 74, and 67 mV/decade, respec-
tively) (Fig. S3). This indicates that the same rate-determining
step is involved in the CO oxidation by the Rh porphyrins
used.
introduction of CO gas to the CH2Cl2 solution of the Rh
porphyrins caused a clear red shift and hyperchromic effects
(line b). The solution after the introduction of CO was exposed
to air, however, the spectra did not change (line c). These
results indicate that CO can easily coordinate on Rh porphyrins,
and the product complex is somewhat stable toward oxygen in
CH2Cl2. Similar absorption behaviors were observed for other
Rh porphyrins. The spectra of other Rh porphyrins after CO
coordination are shown in Fig. 6B. The p-substituents on the
meso-phenyl groups have little effect on either the CO coordi-
nation process or the electronic properties of the CO complex
generated.
The product of the CO coordination of a Rh porphyrin was
collected by removal of the solvent. The powder was subjected
to IR analysis. The IR spectra of [RhIII(T(–OCH3)PP)(Cl)] are
shown in Fig. 7. After the reaction with CO, absorbance bands
in wavenumber regions around 2100 cmꢀ1 emerged, while
absorbance was not observed before CO coordination. This
absorbance can be reasonably ascribed to CO coordinated on
the Rh atom in the Rh porphyrins. The wavenumber regions
are significantly higher than those of other metalloporphyrin-CO
complexes (for example, for Fe porphyrins, o2010 cmꢀ1).23,24
These regions are also higher than that of CO adsorbed on a Pt
metal catalyst (o2050 cmꢀ1),25,26 and close to that of free CO
(2143 cmꢀ1). Similar IR spectra of CO on Rh(TPP) have been
reported in the literature.23,27
3.7 Analysis of CO coordination on Rh porphyrin
To investigate the mechanism of CO electro-oxidation,
the coordination of CO on Rh porphyrin was examined by
UV-vis spectroscopy and IR spectroscopy. Fig. 6A shows
UV-vis spectra of [RhIII(T(–OCH3)PP)(Cl)] in CH2Cl2. The
ꢁc
This journal is the Owner Societies 2010
Phys. Chem. Chem. Phys., 2010, 12, 8968–8976 | 8973

View Article Online
Scheme 2 A proposed reaction mechanism for the electrocatalytic
CO oxidation by a Rh porphyrin.
undergo nucleophilic attack more easily than other CO com-
plexes. These experimental results lead us to a plausible
reaction mechanism (Scheme 2). In this scheme, CO first
coordinates on the Rh atom in a Rh porphyrin. Next, the
coordinated CO undergoes nucleophilic attack by a water
molecule to generate a Rh porphyrin–COOH complex via a
transient intermedate ([Rh porphyrin-C(O)OH2]+). Finally,
electron transfer from the Rh porphyrin–COOH complex to
the Carbon Black surface occurs concomitant with proton
release to generate CO2 and free Rh porphyrin. Similar
reaction mechanisms have been reported for electrocatalytic
CO oxidation.20,21,28
According to the mechanism in Scheme 2, the reaction rate
can be shown as eqn (2) if we consider the equilibrium of CO
binding (K1) and H2O addition to the CO complex (K2), and
assume that the electron transfer (k3) is the rate-determining
step. The kinetic current (Ik) can be shown as follows:
ꢀ
ꢁ
K2
K2þ1
K1
K2þ1
½COꢂ
Fig. 6 UV spectra of [RhIII(T(–OCH3)PP)(Cl)] in CH2Cl2 (a) before,
(b) after CO bubbling, and (c) after exposure to air following the
CO bubbling. (B) UV spectra of CO complexes of Rh porphyrins:
(a) [RhIII(T(–OCH3)PP)(Cl)], (b) [RhIII(T(–COOCH3)PP)(Cl)],
(c) [RhIII(TPP)(Cl)], and (d) [RhIII(TFPP)(Cl)]. The concentration of
Rh porphyrins was fixed at 3 mM. Neat CO was bubbled for 3 min.
Exposure to air was performed for 10 min.
ꢀ
The value k3 can be expressed using the Butler-Volmer
ꢁ
Ik ¼ nFAuG k3
ð2Þ
þ ½COꢂ
equation.
ꢀ
ꢁ
K2
K2þ1
K1
K2þ1
ꢂ
ꢃ
½COꢂ
nFð1 ꢀ aÞ
ðE ꢀ Erꢃds
Þ
ꢀ
ꢁ
Ik ¼ nFAuG
k0 exp
RT
þ ½COꢂ
ð3Þ
where n indicates the number of electrons trnasferred.The
parameter u is the ratio of complex used to the total amount
(the ratio of effective molecules). F, A, G, a, and Erds are Faraday
constant, surface area, amount of complex on electrode, transfer
coefficient, and the redox potential of the rate-determining
process, respectively.
The value of n was determined to be 2 in section 3.4.
Assuming that a is 0.5, the Tafel slope is theoretically
66 mV/decade at 60 1C by replacing n with 2. Actually, the
obtained Tafel slope was within 62–74 mV/decade. Eqn (3) can
be changed to eqn (4) by replacing n with 2.
ꢀ
ꢁ
K2
K2þ1
K1
K2þ1
ꢂ
ꢃ
Fig. 7 FT-IR spectra of [RhIII(T(–OCH3)PP)(Cl)] (a) after and
½COꢂ
2Fð1 ꢀ aÞ
ðE ꢀ Erꢃds
Þ
ꢀ
ꢁ
Ik ¼ 2FAuG
k0 exp
(b) before CO addition. (c) FT-IR spectra of T(–OCH3)PPH2. Inset:
an expanded figure from 1800 cmꢀ1 to 2200 cmꢀ1
RT
þ ½COꢂ
.
ð4Þ
4. Discussion
A UV-vis analysis clearly indicates that CO easily coordinates
on the Rh atom in Rh porphyrins (the first process in
Scheme 2). The CO complexes of Rh porphyrins were charac-
terized using IR. The higher wavenumber regions in the IR
spectra indicate that electron-donation from Rh metal to CO
is weak; the electron density on CO would be lower than those
in CO-complexes of other metalloporphyrins such as Fe
4.1 Plausible reaction mechanism
The easy CO coordination on Rh atom observed in UV-vis
spectra (Fig. 6) suggests the participation of Rh porphyrin–
CO complex in the catalytic reaction. The higher wavenumber
region in IR spectra (Fig. 7) indicates that the CO complex can
ꢁc
This journal is the Owner Societies 2010
8974 | Phys. Chem. Chem. Phys., 2010, 12, 8968–8976

View Article Online
porphyrins. The significance of the reduced electron density on
CO with regard to electrochemical CO oxidation by Co
porphyrin was documented by Shi et al.28 The electron
deficiency in CO on Rh porphyrin assists the nucleophilic
attack by water (the second process in Scheme 2).
of porphyrin molecules on the carbon surface and the u values
in eqn (2)–(4).
Interestingly, the activity of Rh(TCPP)/C per Rh atom
increased with the equilibrium adsorption method. This suggests
that the configuration of Rh porphyrins on Carbon Black
significantly affects the activity. A porphyrin molecule is
adsorbed on Carbon Black due to various kinds of inter-
actions such as p–p stacking interactions, electrostatic inter-
actions, van der Waals interactions, and hydrogen bond
formation. These interactions cause a favourable configura-
tion (monolayer adsorption), and increase the u value in
eqn (4). On the other hand, a porphyrin molecule can interact
with neighboring porphyrin molecules via p–p stacking. The
latter interaction results in an unfavourable configuration
(aggregation or multilayer adsorption), and reduces the use
of Rh porphyrins. This decreases the u value (eqn (4)).
Rh(TCPP), RhT(–COOCH3)PP, and RhT(–OCH3)PP
exhibited higher activity. These Rh porphyrins have p-substituents
that include O atoms. There are many O-derived func-
tional groups (–OH, –O–, COOH, etc.) on the surface of
Carbon Black.31–34 These functional groups can interact with
the substituents of Rh(TCPP), RhT(–COOCH3)PP, and
RhT(–OCH3)PP via a hydrogen bond. This interaction with
carbon promotes the favourable configuration of porphyrin
molecules, and is associated with high activity. A carboxy
group can make hydrogen bonds most easily with O-containing
functional groups. This can explain the highest activity of
Rh(TCPP)/C. The other porphyrins (Rh(TPP), Rh(T(–CH3)PP,
Rh(TFPP), and Rh(TBPP)) exhibited much lower activity
than the three O-containing porphyrins. These porphyrins have
no p-substituents that can interact with the functional groups on
the surface of Carbon Black, and tend to aggregate via inter-
molecular p–p stacking. This possible aggregation of porphyrin
molecules may be responsible for their lower activity.
The Tafel slope of 62–74 mV/decade suggests that the last
process in Scheme 2 is an apparent 2-electron-transfer process
(see ESI).w It follows that the Rh–COOH complex would
undergo heterolytic cleavage, as shown by the arrows in
Scheme 2. Similar values for the Tafel slope for some Rh
porphyrins indicate that there is no difference in the properties
of the rate-determining step.
UV-vis analysis of CO coordination on Rh porphyrins
strongly suggested that there is little difference in K1 among
the porphyrins tested. The similar spectra of CO complexes of
Rh porphyrins (Fig. 6) suggests similar reactivty of the CO
complexes, and implies that there is little difference in K2 and
k3 among the porphyrins. The drastic difference in the CO
oxidation activity can be ascribed to the parameter u.
4.2 Explanation of the difference in the activity between Rh
porphyrins
A significant difference in CO oxidation activities among Rh
porphyrins was observed in Fig. 2. Zagal et al. extensively
studied the metalloporphyrins and metallophthalocyanins
with various substituents with regard to the activities of thiol
electro-oxidation and disulfide electro-reduction.15,16 They
found a close correlation between the electronic properties of
the substituents and the activities.
The substituents cause a change in the redox potentials of
the porphyrins/phthalocyanines, and eventually affect the
electrocatalytic activity. The electronic effect of the substi-
tuents on the redox potential were clearly explained in terms of
Hammett constants.29,30 Hence, in the reactions in previous
works, the electrocatalytic activity is directly related to the
Hammett constants of the substituents.
4.3 Reason for the low activity toward electrochemical H2
oxidation
In this work, the activity of CO oxidation was not related to
the Hammett constants of the p-substituents. This situation is in
marked contrast to the results in the previous works mentioned
above. In this work, the substituents are not directly on the
porphyrin ring but rather on meso-phenyl groups. The X-ray
crystal structure in Fig. 1 clearly shows that the porphyrin ring is
scarcely conjugated with meso-phenyl groups. The electronic
properties of p-substituents have little influence on the electronic
state of the Rh atom in the Rh porphyrins tested in this work.
This is consistent with the findings that the UV-vis spectra of the
Rh porphyrins were almost the same (Fig. S1).w This could
explain why the electronic properties of substituents hardly
affected the activity of CO oxidation activity.
Many other CO oxidation catalysts (e.g. Pt-based catalysts)
have strong activity in H2 electro-oxidation. However, Rh
porphyrin catalysts exhibited little activity toward H2 oxida-
tion. The H2 oxidation activity of Pt catalysts has been
well documented. In the proposed mechanism, the oxidative
cleavage of H2 on the Pt surface generates two Pt–H bonds.
These Pt–H bonds are oxidized electrochemically. The key
point of this catalytic cycle is that two adjacent Pt atoms are
involved. Rh atoms in the Rh porphyrin catalyst are separated
from each other, unlike in bulk metal catalysts, and this
prevents the easy cleavage and coordination of H2. This may
cause the low H2 oxidation activity. The UV-vis spectra
of some Rh porphyrins in CH2Cl2 did not change after
H2 bubbling, in contrast to CO bubbling. This indicates that
H2 does not coordinate on the Rh atom, and supports the
hypothesis discussed above.
The X-ray structure also indicates that a p-substitent is too
far from the Rh atom to have any steric effect. The activity is
also not related to the bulkiness of the p-substituents; a bulky
COOCH3 group enhances activity, while a bulky Br does not.
Thus, the difference in activity cannot be explained by
electronic and steric properties of a single molecule of porhyrin.
Porphyrin–Carbon Black interactions and porphyrin–
porphyrin interactions should be taken into consideration to
explain this difference. These interactions affect the configuration
Conclusions
We investigated CO oxidation by several carbon-supported
Rh tetraphenylporphyrins with systematically varied meso-
substituents. We found that Rh(TCPP)/C exhibits the highest
ꢁc
This journal is the Owner Societies 2010
Phys. Chem. Chem. Phys., 2010, 12, 8968–8976 | 8975

View Article Online
activity among the porphyrins tested. This catalyst can
catalyze CO oxidation at low overpotentials (less than 0.1 V vs.
RHE, at 60 1C). This result means that CO is electrochemically
oxidized by this catalyst when a slight overpotential is applied
under PEMFC operation. This situation is in marked contrast
to that with Pt-based catalysts, which require a large over-
potential for CO oxidation. The catalyst shows high selectivity
toward CO compared to H2. The lack of H2 oxidation activity
indicates that this catalyst should be combined with a Pt
catalyst, which can strongly oxidize H2. The optimal combina-
tion of these two catalysts should be examined further for use
in fuel cell applications.
5 N. Bion, F. Epron, M. Moreno, F. Marino and D. Duprez, Top.
Catal., 2008, 51, 76–88.
6 H. Uchida, K. Izumi, K. Aoki and M. Watanabe, Phys. Chem.
Chem. Phys., 2009, 11, 1771–1779.
7 H. A. Gasteiger, N. M. Markovic and P. N. Ross, J. Phys.
Chem., 1995, 99, 8290–8301.
8 H. A. Gasteiger, N. M. Markovic and P. N. Ross, J. Phys. Chem.,
1995, 99, 16757–16767.
9 M. Arenz, V. Stamenkovic, B. B. Blizanac, K. J. Mayrhofer,
N. M. Markovic and P. N. Ross, J. Catal., 2005, 232, 402–410.
10 S. Ball, A. Hodgkinson, G. Hoogers, S. Maniguet, D. Thompsett
and B. Wong, Electrochem. Solid-State Lett., 2002, 5, A31–A34.
11 T. Ioroi, K. Yasuda, Z. Siroma, N. Fujiwara and Y. Miyazaki,
J. Electrochem. Soc., 2003, 150, A1225–A1230.
12 T. Matsui, K. Fujiwara, T. Okanishi, R. Kikuchi, T. Takeguchi
and K. Eguchi, J. Power Sources, 2006, 155, 152–156.
13 S. Yamazaki, Y. Yamada and K. Yasuda, Inorg. Chem., 2005, 44,
6512–6514.
14 S. Yamazaki, T. Ioroi, Y. Yamada, K. Yasuda and T. Kobayashi,
Angew. Chem., Int. Ed., 2006, 45, 3120–3122.
15 F. Bedioui, S. Griveau, T. Nyokong, A. J. Appleby, C. A. Caro,
M. Gulppi, G. Ochoa and J. H. Zagal, Phys. Chem. Chem. Phys.,
2007, 9, 3383–3396.
We also found that the p-substituents on the meso-phenyl
groups dramatically influenced CO oxidation activity. The
effects of the substituents on the meso-phenyl group were
discussed in terms of a proposed reaction mechanism of CO
activation. No clear relationship was observed between the
activity and the electronic or steric properties of p-substituents.
The CO coordination process and the CO complex of Rh
porphyrin were not significantly different among the Rh
porphyrins tested. These results strongly suggest that the differ-
ences in activity can be attributed not to the intrinsic properties
of a single porphyrin molecule, but rather to porphyrin–
carbon interactions. The substituents that are associated with
high CO oxidation activity contain an O atom, and this can
promote the formation of hydrogen bonds with functional
groups on the carbon surface. The increased interaction due
to this bond formation may be responsible for the high activity.
16 J. H. Zagal, M. A. Gulppi and G. Cardenas-Jiron, Polyhedron,
´ ´
2000, 19, 2255–2260.
17 S. Yamazaki, Y. Yamada, N. Fujiwara, T. Ioroi, Z. Siroma,
H. Senoh and K. Yasuda, J. Electroanal. Chem., 2007, 602,
96–102.
18 G. M. Sheldrick, SHELXTL. version 6.12, Bruker AXS, Madison,
Wisconsin, USA, 2000.
19 J. Wu and C. P. Kubiak, J. Am. Chem. Soc., 1983, 105,
7456–7457.
20 J. F. van Baar, J. A. R. van Veen and N. de Wit, Electrochim. Acta,
1982, 27, 57–59.
21 J. F. van Baar, J. A. R. van Veen, J. M. van der Eijk, Th. J. Peters
and N. de Wit, Electrochim. Acta, 1982, 27, 1315–1319.
22 J. A. R. van Veen, J. F. van Baar and K. J. Kroese, J. Chem. Soc.,
Faraday Trans. 1, 1981, 77, 2827–2843.
23 E. B. Fleischer, R. Thorp and D. Venerable, J. Chem. Soc. D, 1969,
475a.
Acknowledgements
The authors would like to express sincere gratitude to
Ms M. Ichimura for her kind support and technical assistance.
They would also like to thank Dr M. Yao for fruitful discussions.
They are grateful to the researchers from the Technical Service
Center (AIST) for elemental analysis. This study was supported
by the New Energy and Industrial Technology Development
Organization (NEDO) of Japan.
24 M. Matsu-ura, F. Tani and Y. Naruta, J. Am. Chem. Soc., 2002,
124, 1941–1950.
25 T. Sato, K. Kunimatsu, H. Uchida and M. Watanabe, Electrochim.
Acta, 2007, 53, 1265–1278.
´
26 F. Maillard, A. Bonnefont, M. Chatenet, L. Guetaz, B. Doisneau-
Cottignies, H. Roussel and U. Stimming, Electrochim. Acta, 2007,
53, 811–822.
27 I. A. Cohen and B. C. Chow, Inorg. Chem., 1974, 13, 488–489.
28 C. Shi and F. C. Anson, Inorg. Chem., 2001, 40, 5829–5833.
29 A. B. P. Lever, J. Porphyrins Phthalocyanines, 1999, 3, 488–499.
30 A. B. P. Lever, Inorg. Chim. Acta, 1993, 203, 171–174.
31 S. Yamazaki, Z. Siroma, T. Ioroi, K. Yasuda and K. Tanimoto,
J. Electroanal. Chem., 2008, 616, 64–70.
References
1 R. A. Lemons, J. Power Sources, 1990, 29, 251–264.
2 H. Igarashi, T. Fujino and M. Watanabe, J. Electroanal. Chem.,
1995, 391, 119.
32 E. Fuente, J. A. Menendez, M. A. Dıez, D. Suarez and
´ ´ ´
M. A. Montes-Moran, J. Phys. Chem. B, 2003, 107, 6350–6359.
´
3 M. Gotz and H. Wendt, Electrochim. Acta, 1998, 43, 3637–3644.
4 Y. Lei, N. W. Cant and D. L. Trimm, J. Catal., 2006, 239,
227–236.
¨
33 H. P. Boehm, Carbon, 2002, 40, 145–149.
34 H. P. Boehm, E. Diehl, W. Heck and R. Sappok, Angew. Chem.,
Int. Ed. Engl., 1964, 3, 669–678.
ꢁc
This journal is the Owner Societies 2010
8976 | Phys. Chem. Chem. Phys., 2010, 12, 8968–8976