Photosensitization of thin SnO2 nanocrystalline semiconductor film electrodes with electron donor-acceptor metallodiporphyrin dyad

Article abstract of DOI:10.1016/j.chemphys.2004.11.027

A electron donor-acceptor porphyrin dyad (P_Zn-P) was synthesized by linking an electron acceptor porphyrin; 5,15-bis(4-carboxyphenyl)-10,20- bis(4-nitrophenyl) porphyrin (P) and an electron donor porphyrin; Zn(II) 5-(4-aminophenyl)-10,15,20-tris(4-methoxylphenyl) porphyrin (P_Zn) by amide bond. P_Zn-P dyad - thin SnO₂ nanocrystalline semiconductor film electrodes show higher spectral sensitized photocurrent quantum yield compared to the electrodes sensitized with either P_Zn or P monomers. Fluorescence analysis of dyad and the P moiety adsorbed on both, SnO₂ semiconductor and SiO₂ insulator, shows that the charge injection yields (Φ_inj) from the excited dyes to the SnO₂ are similar in both cases. Thus photocurrent enhancement is interpreted in terms of intramolecular electron transfer and preferential spatial orientation of the dyad on the SnO₂ surface that preclude back electron transfer.

Full text of DOI:10.1016/j.chemphys.2004.11.027

Chemical Physics 312 (2005) 97–109
www.elsevier.com/locate/chemphys
Photosensitization of thin SnO2 nanocrystalline semiconductor
film electrodes with electron donor–acceptor
metallodiporphyrin dyad
Miguel Gervaldo, Luis Otero, M. Elisa Milanesio, Edgardo N. Durantini,
*
Juana. J. Silber, Leonides E. Sereno
Departamento de Quı´mica, Universidad Nacional de Rı´o Cuarto, Agencia Postal 3, 5800 Rı´o Cuarto, Argentina
Received 19 July 2004; accepted 19 November 2004
Available online 13 December 2004
Abstract
A electron donor–acceptor porphyrin dyad (PZn–P) was synthesized by linking an electron acceptor porphyrin; 5,15-bis(4-car-
boxyphenyl)-10,20-bis(4-nitrophenyl) porphyrin (P) and an electron donor porphyrin; Zn(II) 5-(4-aminophenyl)-10,15,20-tris(4-
methoxylphenyl) porphyrin (PZn) by amide bond. PZn–P dyad – thin SnO2 nanocrystalline semiconductor film electrodes show
higher spectral sensitized photocurrent quantum yield compared to the electrodes sensitized with either PZn or P monomers. Fluo-
rescence analysis of dyad and the P moiety adsorbed on both, SnO2 semiconductor and SiO2 insulator, shows that the charge injec-
tion yields (Uinj) from the excited dyes to the SnO2 are similar in both cases. Thus photocurrent enhancement is interpreted in terms
of intramolecular electron transfer and preferential spatial orientation of the dyad on the SnO2 surface that preclude back electron
transfer.
ꢀ 2004 Elsevier B.V. All rights reserved.
1. Introduction
and efficient electron-transfer steps, which move elec-
trons and holes to opposite sides of the membrane
[16,17]. One approach to mimic the photosynthetic en-
ergy conversion system in nature is the use of complex
synthetic molecular arrays containing chromophores,
electron donors and electron acceptors linked by cova-
lent and non-covalent bonds [17–25]. The energy stored
in these systems as a photoinduced intramolecular
charge separation state can be converted in electric work
through an external circuit in a suitable photoelectro-
chemical cell. On the other hand, many studies on the
development of optoelectronic and energy conversion
devices incorporate several metallized and unmetallized
tetrapyrrolic compounds, as light receptors and charge
storage units [26–32].
The spectral sensitization of wide band gap metal
oxide semiconductors with organics dyes is of great
interest in the search of efficient energy conversion de-
vices [1–4]. A considerable number of organic structures
has been designed, synthesized and studied with this
purpose, and many researchers were inspired in this field
by the natural photosynthetic apparatus [5–15]. Photo-
initiated transmembrane electron transfer to yield
long-lived charge-separated states is the basic energy
conversion process of natural photosynthesis. In the
protein environment of photosynthetic reaction centers
the process is carried out by a series of short-range, fast,
It is interesting to remark the analogies between
the natural photosynthesis and the dye-sensitized nano-
crystalline semiconductor electrodes using electron
*
Corresponding author. Tel.: +54 358 4676541; fax: +54 358
4676233.
E-mail address: lsereno@exa.unrc.edu.ar (L.E. Sereno).
0301-0104/$ - see front matter ꢀ 2004 Elsevier B.V. All rights reserved.
doi:10.1016/j.chemphys.2004.11.027

98
M. Gervaldo et al. / Chemical Physics 312 (2005) 97–109
donor–acceptor molecular assemblies [33]. In a porous
film of nanometer-sized semiconductor particles, the
effective surface area can be greatly enhanced, producing
substantial light absorption [34–41] and charge separated
states formation. Light harvesting and charge separation
efficiencies in a range of porphyrin dyes adsorbed on
nanostructured semiconductor oxides were found com-
parable to those in natural photosynthesis, with a very
high (nearly 80%) incident-photon-to-current efficiency
(IPCE) [5,6]. With such high efficiencies, the photosensi-
tization of oxide semiconductors has become a practical
mean for solar energy conversion. Using Ru(II) bipyri-
dine complexes as photo-sensitizers, net power conver-
sion efficiencies up to 10% have been reported in diffuse
daylight on nanoporous TiO2 electrodes [1,2,42,43].
Imahori et al. [44–46] reported the production of pho-
tocurrent output with very high quantum yield by using
donor–acceptor linked molecules, composite molecular
nanoclusters [47] and quaternary self-organization of do-
nor–acceptor dye units [48]. Likewise, we recently syn-
thesized a phorphyrin-fullerene dyad molecule, which
showed to be effective in the spectral sensitization of wide
band gap SnO2 nanostructured semiconductor electrode,
producing anodic photocurrents under visible irradia-
tion [49]. Regardless of the fact that the porphyrin fluo-
rescence was strongly quenched by the C60 moiety in the
dyad, the photocurrent generation quantum yield was
around twice in the dyad when it is compared with the
same yield of the porphyrin moiety. A mechanism
involving the formation of an intramolecular photoin-
duced charge transfer state was proposed in order to
explain the observed photoeffect [49].
On the other hand, in a recent work, we used a por-
phyrin dyad molecule and its structural moieties in the
spectral sensitization of wide band gap SnO2 nanostruc-
tured semiconductor electrode [50,51]. SnO2 film modi-
fied with porphyrin dyes exhibited photoresponse in
the visible region, and nearly complete singlet–singlet
energy transfer from Zn-porphyrin to free base moiety
was observed in the dyad in the adsorbed state. Conse-
quently, a suitable ‘‘antenna’’ effect was obtained using
the dyad molecule. However, the photocurrent efficiency
in the dyad was lower than the one produced by the free
base moiety. This fact was explained considering that
the metallized porphyrin enhances the back electron
transfer process, producing low photocurrent yield
[52,53]. However, the lack of preferential orientation
of the dyad over the SnO2 surface (i.e., whether it is
the free base moiety or the Zn-porphyrin moiety that
gets attached to SnO2 surface) also play a role in deter-
mining the photocurrent quantum yield of the dyad
[51,54].
Thus, in order to improve the photocurrent quantum
yield, we designed and synthesized [55] a novel porphy-
rin-porphyrin supramolecular arrangement with partic-
ular structural and electronics properties. The new
asymmetric dyad PZn–P (Fig. 1) was designed in order
to increase the intramolecular electron transfer capacity;
one porphyrin moiety (PZn) bears electron-donating
methoxy groups and a zinc ion, which help to stabilize
H3C
O
NO2
N
N
N
NH
N
N
H
N
O
C
O
Zn
O
C
CH3
OH
C
O
H3C
HO
N
HN
O
NO2
CH3
P
P
Zn
H3C
O
Zn
O
NO2
N
N
N
N
H
N
NH
N
N
O
O
C
O
C
H3C
OH
HN
CH3
NO2
Dyad
P
-P
Zn
Fig. 1. Structures of the sensitizing dyes.

M. Gervaldo et al. / Chemical Physics 312 (2005) 97–109
99
the radical cation; the second porphyrin moiety (P) fea-
tures electron-withdrawing nitro substituent, which sta-
bilize the porphyrin radical anion. It has been
demonstrated that peripheral substitution in para-
substituted tetraphenylporphyrin [H2(p-X)TPP] [56]
and metalloporphyrins [57–59] ([M(p-X)TPP] where
M = Mn2+, Fe2+, Co2+, Ni2+, Cu2+ and Zn2+) has a
noticeable effect in modulating their electron donor/
acceptor capabilities. Hammet linear correlations be-
tween electron donating or withdrawing ability of p-X
substituents showed that the stronger effects are pro-
duced by –OCH3 and –NO2 groups, the same that were
selected by us in the design of the PZn–P dyad. In PZn–P
dyad the moieties are linked by amide bond, which al-
lows electronic interaction between the p systems of
the two chromophores [60,61].
On the other hand P moiety bears a remnant carbox-
ylic acid group. This group was conveniently located in
the electron-withdrawing porphyrin moiety in order to
benefit the orientation of the PZn–P dyad with respect
to the basic semiconductor oxide electrode [62,63]. The
preferential orientation is able to improve charge injec-
tion and to keep the photooxidized porphyrin moiety
away from the photoinjected electrons. This fact could
diminish back electron transfer process [2,11,55].
Through the combination of intramolecular and interfa-
cial photoinduced charge injection processes is expected
to increase the photocurrent quantum yield of the dye
sensitized photovoltaic system [64,65].
the reaction of meso-(4-nitrophenyl) dipyrromethane
with 4-carboxymethylbanzaldehyde in a MacDonald-
type 2 + 2 condensation [66]. The hydrolysis in basic
medium of ester-P yielded 5,15-bis(4-carboxyphenyl)-
10,20-bis(4-nitrophenyl) porphyrin (P). 5-(4-Acetami-
dophenyl)-10,15,20-tris(4-methoxylphenyl) porphyrin
(amido-P) was formed from the condensation of
meso-(4-methoxyphenyl)dipyrromethane with a binary
mixture of 4-methoxybenzaldehyde and 4-acetamido-
benzaldehyde according to the procedure previously
described [67]. Basic hydrolysis of amido-P produced
5-(4-amidophenyl)-10,15,20-tris(4-methoxylphenyl) por-
phyrin (amino-P). Treatment of amido-P and amino-P
with Zn(II) acetate was used to obtain the correspond-
ing Zn(II) metallo complexes. To synthesize the target
dyad PZn–P, the carboxylic acid group of the porphyrin
P was transformed to acid chloride using thionyl
chloride in dry toluene and pyridine. The acid chloride
was immediately used for coupling with Zn(II) 5-(4-
aminophenyl)-10,15,20-tris(4-methoxylphenyl) porphy-
rin. The reaction mixture was allowed to react for 15 h
at room temperature. The product was purified by flash
chromatographic column. Spectroscopic data of the
porphyrin derivatives have been previously reported
[55].
2.2. Preparation of nanocrystalline films
The SnO2 nanostructured film thin electrodes were
prepared using a modification of the KamatÕs procedure
[38] by the method already described [12,49]. As base
contact optically transparent electrodes of indium tin
oxide (ITO, 100 X/square, Delta Technologies) were
used. ITO/SnO2 electrodes were prepared by spin coat-
ing using a P6204-A model Specialty Coating System.
The final films were annealed at 450 ꢁC for 1 h. The
SnO2 films were analyzed by X-ray diffractommetry
(Siemes D5000, Cu Ka radiation, XRD) in order to
2. Experimental
2.1. Synthesis of porphyrine dyes
The structures of the porphyrins used are shown in
Fig. 1. 5,15-Bis(4-carboxymethylphenyl)-10,20-bis(4-
nitrophenyl) porphyrin (ester-P) was synthesized from
Fig. 2. Top view (a) and three-dimensional plot (b) AFM of the nanocrystalline SnO2 film.

100
M. Gervaldo et al. / Chemical Physics 312 (2005) 97–109
assess the crystallinity of the sample. The peaks in the X-
ray diffraction pattern show lines which were in good
agreement with the crystalline SnO2 cassiterite standard
[68,69]. To further investigate the morphological charac-
teristics of the films, atomic force microscopy (Nano-
scope III, AFM) has been used, operating in the
taping mode. Fig. 2(a) presents 2 lm · 2 lm top view
of the films, meanwhile Fig. 2(b) shows a three-dimen-
sional surface plot. The films consist of interconnected
grain particles fused together to build up high moun-
tains and deep valleys. The average grain diameter is
about 15–20 nm and compares well the particle size of
the initial colloid solution. This shows that no significant
aggregation occurred during sintering of the film. The
fractal analysis shows that the effective surface area is
several times greater than the one of a flat non-fractal
surface. Such surface not only permits the adsorption
of a great number of dye molecules, but also creates a
rough environment where multiple light reflections can
occur, thus considerable increasing the amount of ad-
sorbed photons. The nanoparticulate film obtained from
colloidal suspension has a highly porous structure that
permits a better wetting of the film by the electrolyte,
which favors the interaction between the oxidized dye
and the sacrificial donor.
SiO2 nanostructured films were formed with the same
procedure used for SnO2 photoelectrodes, using a SiO2
colloidal suspension in water (Ludox SM 30, Aldrich).
ITO/SnO2 electrodes were modified with PZn–P dyad
and the PZn and P porphyrin moieties (Fig. 1) soaking
two hours ITO/SnO2 films in a saturated n-hexane/
dichloromethane (DCM Aldrich 99.8% grade) 80/20
solution of the corresponding dye. The electrodes were
then washed with the same solvent, dried in nitrogen
stream at room temperature and stored in vials. A cop-
per wire was connected to the electrodes surface with an
indium solder to achieve electric contact.
ferrocene was added, and the potential axis was cali-
brated against the formal potential for ferrocenium/fer-
rocene redox couple (E0 = 0.46 V vs SCE; 0.70 V vs
NHE). All the potentials were referred to the normal
hydrogen electrode (NHE) [51]. Redox potentials for
the first anodic (Eox) and cathodic (Ered) charge transfer
of each molecule were calculated using the expression
(Ep forward ꢀ Ep backward)/2. The Eox relative to the
conventional scale (NHE) were converted to the abso-
lute energy scale taken the energy of the electrons in
NHE as ꢀ4.5 eV [70].
The oxidation potentials of the dye-excited state ðEꢁoxÞ
were estimated by subtracting the excitation energy from
the redox potentials of the molecule in the ground state [70].
Absorption spectra were recorded on a Shimadzu
UV-2401PC spectrophotometer. Fluorescence spectra
were recorded on a Spex FluoroMax fluorimeter, pro-
vided of solid sample holder. The fluorescence quantum
yields (/f) were obtained in degassed toluene solutions
by comparison with that of tetra phenyl porphyrine
(TPP, /fl = 0.11) used as fluorescence standard. Relative
fluorescence intensities from photoelectrodes (Ifl) were
calculated from the area below the absorption corrected
emission spectrum in air. Light harvesting efficiency
(LHE) of the dye adsorbed over the photoelectrodes
were obtained from LHE ¼ 1–10ꢀA , where Ak is the
k
monochromatic absorbance [41].
2.4. Photoelectrochemical measurements
Photoelectrochemical experiments were conducted in
aqueous solutions (0.01 M) of hydroquinone (H2Q, Al-
drich, recrystallized from toluene), with phosphate buf-
fer (pH = 5.2) prepared from 0.05
M NaH2PO4
(Anedra > 99%) and NaOH (Baker). These solutions
were degassed by bubbling with Ar and maintained it
in the top of the cell by continuous stream. The mea-
surements were carried out under potential control in
an already described quartz photoelectrochemical cell
[49–51] equipped with Ag/AgCl as reference electrode,
and Pt foil as the auxiliary electrode, using a computer
controlled battery operated low noise potentiostat (Palm
Sens). Action spectra were obtained by illumination of
the photoelectrodes with monochromatic light obtained
from a 75 W high-pressure Xe lamp (Photon Technol-
ogy Instrument, PTI) and a computer controlled PTI
high intensity grating monochromator. The steady-state
photocurrents were obtained in front face configuration
(illuminated area: 1 cm2). The incident light intensities at
different wavelengths were measured with a Coherent
Laser-Mate Q radiometer (sensitivity 1 lW).
2.3. General
The voltammetric characterization of the redox
process for the porphyrine dyes were acquired with a
potentiostat-galvanostat Autolab (Electrochemical
Instruments) using a Pt disc working electrode of
0.204 cm2 and a Pt counter electrode in a conventional
three compartment Pyrex cell. A freshly prepared Ag/
AgCl quasireference electrode was used. This electrode
was prepared by plating a thin layer of AgCl onto an
Ag wire from a 1 · 10ꢀ3 M KCl solution. The working
electrode was cleaned between each experiment by pol-
ishing with 0.3 lm alumina paste, followed by solvent
rinses. Studies were carried out in N2-degased DCM
with 0.1 M tetrabutylammonium hexaflourophosphate
(TBAHFP, Aldrich) supporting electrolyte. The voltam-
mograms were corrected for IR drop by positive feed-
back technique. After each voltammetric experiment,
2.5. Laser-flash photolysis experiments
Laser-flash photolysis set-up has been described else-
where [49,71]. A Q-switched Nd:YAG laser (Spectron

M. Gervaldo et al. / Chemical Physics 312 (2005) 97–109
101
20
15
10
5
SL400) was used as the excitation source operating at
532 nm (20 ns halfwidth) in order to excite the porphy-
rin compounds adsorbed over ITO/SnO2 electrodes
placed in the already described photoelectrochemical
cell. Has been reported [53] that the electron injection
kinetics into semiconductor are rather insensitive to
experimental conditions. In contrast, the recombination
kinetics between dye cation and photoinjected electrons
are very strongly dependent upon experimental condi-
tions, with variation in the recombination rate from
picoseconds to milliseconds depending upon the excita-
tion intensity, solvent/electrolyte composition and ap-
plied electrical bias [52,53,78]. To compare the
recombination kinetics between the different sensitizer
dyes, considerable care was taken to minimize such
experimental variation. Applied potential on ITO/
SnO2/dye electrodes was keep under control in a tree
electrode configuration, in an Ar – degassed phosphate
buffer solutions without sacrificial donor hydroquinone.
The laser beam was abated by using neutral filters in or-
der to avoid multiphotonic processes and/or photodeg-
radation of the samples. The sample electrodes were
matched in absorbance at 532 nm, the wavelength of
the exciting laser pulse, within 10%. The output of the
detector was coupled to a digital oscilloscope (Hew-
lett-Packard HP-54504A). About twenty shots were usu-
ally needed for averaging decay times, in order to get a
good signal to noise ratio. The averaged signals were
analyzed as multiple exponential decays by using a cal-
culation software (Microcal Origin 4.1, Northampton,
USA).
0
-5
-10
0.0 0.3 0.6 0.9 1.2 1.5 1.8
E / V vs NHE
Fig. 3. Cyclic voltammogram at a platinum electrode (A = 0.204 cm2)
of PZn 0.20 mM in DCM containing 0.1 M TBAHFP as supporting
electrolyte, sweep rate = 0.100 V/s.
withdrawing nitro substituents, consequently, the reduc-
tion process (Fig. 1) for P (Ered = ꢀ0.84 V) is easier than
the corresponding free base porphyrin without nitro
groups. Taking into account all these values, it can be
calculated that the PZn–P state in the dyad has an en-
ergy of 1.76 eV above its ground state.
Åþ Åꢀ
3.2. Absorption and fluorescence spectra of dyes
The photochemistry of the monomers and dyad mol-
ecule was analyzed by steady-state absorption and emis-
3. Results and discussion
sion spectroscopy. The absorption spectra of P, PZn
,
3.1. Solution electrochemistry
and PZn–P compounds are shown in Fig. 4, both in solu-
tion (toluene) and adsorbed over ITO/SnO2 nanostruc-
tured films. As can be observed, the spectrum of the
dyad (Fig. 4(c)) features the four Q-bands at kmax 518,
554, 594 and 652 nm, whereas the Soret band appears
at kmax = 424 nm. The absorption spectrum is identical,
within experimental error, with a linear combination of
the spectra of the component chromophores (Fig. 4(a)
and (b)) [67,72]. Thus, the absorption spectra are consis-
tent with only a weak interaction between the moieties,
and the two chromophores retain their individual iden-
tities in ground state [60]. On the other hand, porphyrin
adsorption onto the semiconductor produces violet-
green coloration of the film; the UV–Vis absorption
spectra of the dye compounds on ITO/SnO2/dye elec-
trode show the Q and Soret bands with electronic tran-
sitions similar to those observed for monomeric
absorption in solution. However, on the electrodes the
bands are broader and are shifted in comparison
with those in solution. This may be caused by the inter-
action of the porphyrins with the polar surface of the
The voltammetric characterization of the reduction
and oxidation process of the dyes was carried out in or-
der to know electron donor–acceptor capacity of each
moiety, and the energy of the possible intramolecular
charge separated state in PZn–P dyad. It has been re-
ported [51,72] that the reduction and oxidation poten-
tials of the dimeric porphyrins are similar to those of
the combined individual monomeric units, hence the en-
Åþ Åꢀ
ergy of the charge separated state in PZn–P has been
estimated from the first oxidation potential (Eox) of
PZn and first reduction potential (Ered) of ester-P. The
oxidation voltammogram of PZn is shown in Fig. 3.
The first charge transfer process occurs in this metallized
moiety at Eox = 0.92 V, that is, at 0.48 V less anodic po-
tential that for P (Eox = 1.40 V). This result is expected
due to the presence of the central metal and the electron
donor methoxy groups substituents in PZn. This is
agreement with previous results [56]. On the other hand,
the free base porphyrin possesses two strong electron

102
M. Gervaldo et al. / Chemical Physics 312 (2005) 97–109
(a)
(c)
(b)
1.0
0.8
0.6
0.4
0.2
0.0
1.0
0.8
0.6
0.4
0.2
0.0
1.0
0.8
0.6
0.4
0.2
0.0
300
400
500
600
700
300
400
500
600
700
300
400
500
600
700
Wavelength / nm
Wavelength / nm
Wavelength / nm
Fig. 4. Absorption spectra of porphyrins: in toluene solution (–––) and in adsorbed state over ITO/SnO2/electrodes (------). (a) P; (b) PZn and
(c) PZn–P dyad. Absorbance has been normalized to one at Soret band for comparison.
electrode, as well as the possible formation of porphyrin
aggregates [51].
frequencies of the longest wavelength absorption max-
ima and the shortest wavelength emission maxima.
The obtained values for PZn and P first excited singlet
states, 2.08 and 1.91 eV, respectively, show that there
is a significant energy gap between both moieties in
the dyad, thus, singlet–singlet energy transfer from the
PZn to P is an energetically favorable process. The re-
sults show that the PZn moiety serves as a reasonable
efficient antenna for P by transferring singlet excitation
energy to it. However, it should be remarked that emis-
sions from both porphyrin moieties are quenched in
PZn–P dyad, relative to either monomer (fluorescence
from the PZn and P are diminished with a quenching
efficiency (gq) of 0.62 and 0.35, respectively). Thus, the
energy transfer is not complete, and the first excited sin-
glet states of both porphyrins are quenched by another
On the contrary, in the emission spectra this situation
changes. Fig. 5(a) shows the absorption corrected fluo-
rescence spectra of the monomers units and the dyad
in toluene solution. The PZn has maxima at 604 and
652 nm, whereas P has two maxima at 653 and 719
nm. In this case, the emission spectra of the PZn–P dyad
resemble that of the unmetallized porphyrin, with low
emission from the PZn moiety, even though that both
porphyrin species absorb at the excitation wavelength
(550 nm). These results suggest singlet–singlet energy
transfer in the PZn–P dyad from PZn moiety to the free
base P [50,51,55,73]. Analyzing the possible photophysic
mechanisms, the energies of the porphyrins first excited
singlet states can be calculated from the average of the
(b)
(a)
600
650
700
750
800
600
650
700
750
800
Wavelenght / nm
Wavelength / nm
Fig. 5. (a) Corrected fluorescence emission spectrum in toluene solution of: P (-------); PZn (•••••) and PZn–P (
) dyad, kex = 550 nm.
) and SnO2 (x2) (-------), kex = 420 nm. All the emission spectra were
(b) Fluorescence emission spectrum of PZn–P dyad adsorbed on SiO2 (
corrected by the absorbance and instrumental response.

M. Gervaldo et al. / Chemical Physics 312 (2005) 97–109
103
competitive process that could involve electron transfer
an external circuit could collect the injected electrons,
producing observable photocurrent, although, the effi-
ciency of charge harvesting is adversely influenced by
charge recombination. On the other hand, when the
dye is adsorbed on ITO/SiO2 (band gap = 6.9 eV) [74],
the oxidation potential of the dye-excited state ðEꢁoxÞ lies
at much lower energy that the conduction band edge of
the insulator. Consequently, electron transfer is not pos-
sible and the fluorescence is not affected by this addi-
tional mechanism [9]. On the contrary, for a dye
adsorbed on ITO/SnO2 there is a driving force for pho-
toinduced electron transfer froꢁm the dye to the semicon-
ductor, defined as DE ¼ ꢀeðEox ꢀ EFBÞ [75], where EFB
is the potential of the semiconductor flat band
(ꢂꢀ0.02 V vs NHE) [38] and e is the electronic charge.
Since electron transfer is in competition with radiative
and non-radiactive process, one can express the charge
injection yield (Uinj) from the excited dye to the semicon-
ductor as [9,76]:
from the PZÅnꢀ to the P moiety, producing the ion pair
Åþ
state, PZn–P Åþ[55,60,61]. Since the electrochemical data
Åꢀ
shows that PZn–P has a energy of 1.76 eV, this state
lies at about 0.32 eV below PZn first excited singlet state
ðP1Zn–PÞ, making the intramolecular electron transfer
quenching energetically feasible.
On the other hand, when the PZn–P dyad is in ad-
sorbed state (both, over SnO2 and SiO2 nanocrystalline
films) the emission spectrum of an ITO/Oxide/PZn–P
electrode also resembles that of the P porphyrin, with
no significant emission from the metallized moiety
(Fig. 5(b)). This result suggests that singlet–singlet en-
ergy transfer from the PZn to P also occurs when the
dye is adsorbed. However, the fluorescence intensity of
the dyad adsorbed over SnO2 film is much smaller than
the same over the insulator SiO2 (Fig. 5(b)). In other
words, an additional quenching pathway is effective
when the dyad is adsorbed on the semiconductor sur-
face. This quenching can be interpreted in terms of elec-
tron transfer from the excited dye to the semiconductor
[39]. In the present case, the energies of the excited dyad
(in both possible singlet states, P1Zn–P or PZn–P1) are
found to lie at higher values than the semiconductor
conduction band edge [38] (vide infra). With all the de-
scribed information an energy level diagram of the dif-
ferent states of PZn–P dyad and the possible
mechanism for the heterogeneous electron transfer in
an ITO/SnO2/PZn–P electrode can be drawn as shown
in Fig. 6.
ket
Uinj
¼
;
ð1Þ
ket þ kr þ knr
where ket is the heterogeneous electron transfer rate con-
stant, kr is the radiative deactivation rate constant and
knr is the non-radiative deactivation rate constant. The
electron transfer efficiency can be correlated with fluo-
rescence quenching if the fluorescence intensity in ab-
sence of charge transfer (Ifl(0)) is written as:
kr
Iflð0Þ ¼ A ꢃ I
;
ð2Þ
exc kr þ knr
After production of the excited state by photon
absorption, several deactivation pathways are possible:
radiative relaxation, thermal (non-radiative) relaxation,
or interfacial charge injection. When the latter occurs
where A is an instrumental factor and Iexc is the excita-
tion intensity. On the other hand, fluorescence intensity
in the presence of electron transfer (Ifl(e)) can be written
as:
kr
0
IflðeÞ ¼ A ꢃ Iexc
:
ð3Þ
P1Zn-P
ket þ kr þ knr
PZn-P1
e
P+Zn-P-
The reorganization of the Eqs. (2) and (3) gives Uinj in
terms of normalized fluorescence intensities as follow:
ET
hν
-4
CB
hν
1
f
Iflð0Þ ꢀ IflðeÞ
hν
2
Uinj
¼
:
ð4Þ
Iflð0Þ
The relative fluorescence intensity (Ifl) obtained from
the area below the corrected emission spectrum of pho-
toelectrodes in air (Fig. 5(b)) allows to calculate the
charge injection yield from dyes into the semiconductor
nanoparticles (Eq. (4)). Uinj in this conditions was found
to be 0.89 and 0.88 for PZn–P dyad and P porphyrin,
respectively. Thus, the electron injection efficiency is
similar for both compounds.
PZn-P / P+Zn-P
-6
-8
VB
VB
3.3. Photoelectrochemistry
SnO2
SiO2
The ITO/SnO2/dye electrodes are photoelectrochem-
ically active in the range between 300 and 700 nm. The
Fig. 6. Energy level diagram in absolute scale for the spectral
sensitization of ITO/Oxide/PZn–P electrode. ET = energy transfer.

104
M. Gervaldo et al. / Chemical Physics 312 (2005) 97–109
generated photocurrent is anodic, and negative open cir-
cuit photopotentials are observed upon excitation, indi-
cating that the electrons flow from the solution to ITO
base contact through the illuminated electrode. The
photocurrent is reproducible under several (hundreds)
repeated on–off illumination cycles. Contrarily, ITO/
SiO2/dye electrodes do not have photoresponse in the
visible range, in agreement with the expected lack of
charge injection into the insulator.
The action spectrums of ITO/SnO2/dye electrodes
(measured under 0.05 V positive bias vs. Ag/AgCl) clo-
sely match the absorption spectrum of the adsorbed or-
ganic molecules in all cases, as is shown in Fig. 7(a)–(c).
The IPCE were evaluated from Eq. (5) [34,38,77].
Table 1 resume the results of IPCE, LHE and the
product Uinj gc for the three dyes at the Soret band. Uinj
gc allows to made a comparative analysis of photoelec-
tric generation efficiency of the different dyes under sim-
ilar experimental conditions.
As it can be observed P gives lower photocurrent effi-
ciency than the metallized porphyrin, contrarily to the
results that we obtained using structurally related mole-
cules [51]. However, it must be noted that both porphy-
rins bear substituent groups that could affect the
photoelectrochemical performance. In the case of the
free base porphyrin P possess electron-withdrawing ni-
tro substituents, which could be involved in a less effec-
tive heterogeneous electron injection into the
semiconductor and/or faster back electron transfer. This
deleterious process showed to be the main problem in
achieving high IPCE values, producing low efficiency
of photon to current efficiency [52,53]. Moreover, ana-
IPCE ð%Þ ¼ 100ðisc1240Þ=ðIinckÞ;
ð5Þ
where isc is the short circuit photocurrent (A cmꢀ2), Iinc
is the incident light intensity (W cmꢀ2), and k is the exci-
tation wavelength (nm). The close correspondence be-
tween the photocurrent action spectrum and the
absorption spectrum of the electrodes confirms that
photosensitization has successfully extended the photo-
current response of the electrodes into the visible region
in all cases.
Table 1
Photoelectrochemical properties of dyes
*,b
Dye
SEa Eox
IPCE %c LHE % Uinj · gc kb ·
106 sꢀ1d
P
1.91 (ꢀ0.51)*
2.08 ꢀ1.16
–
3
43
14
38
0.07
0.14
0.26
4.15 (79)
0.11 (21)
2.7 (77)
0.09 (23)
2.37 (73)
0.10 (29)
Maximum IPCE of ꢂ3%, 2% and 10% were obtained
at the Soret band of ITO/SnO2/P, ITO/SnO2/PZn and
ITO/SnO2/PZn–P, respectively. There are several rea-
sons for this difference and may be linked to the factors
that determine IPCE, namely, the LHE of the dye, the
charge injection yield (Uinj) from the excited dye to the
semiconductor, and the charge collection efficiency, gc,
of the system. The IPCE in terms of these parameters
can be written [34,41] as:
PZn
PZn–P
a
2
10
Energy of the first excited singlet state (eV).
Oxidation potential of excited dyes (V vs NHE).
Maximum IPCE at Soret bands under bias voltage = 0.05 V.
Back electron transfer kinetic fitted by bi-exponential function, in
b
c
d
parenthesis relative amplitudes.
*
IPCE % ¼ LHE ð%ÞUinjgc:
ð6Þ
Taken from oxidation potential of ester-P.
(a)
(b)
(c)
10
2.0
1.5
1.0
0.5
0.0
3
8
6
4
2
0
2
1
0
300
400
500
600
700
300
400
500
600
700
300
400
500
600
700
Wavelength / nm
Wavelength / nm
Wavelength / nm
Fig. 7. Absorption spectra (
) and photocurrent action spectra, IPCE (-O-) of ITO/SnO2/dye electrodes modified with (a) P; (b) PZn; and
(c) PZn–P dyad.

M. Gervaldo et al. / Chemical Physics 312 (2005) 97–109
105
0.30
lyzing the Uinj we found it to be large (0.88) and compa-
rable to the same of PZn–P dyad, which produces higher
photocurrent quantum yield. Thus, low charge collec-
tion efficiency gc of the ITO/SnO2/P system could be
the limiting factor through back electron transfer (i.e.
the charge recombination between the photoinjected
electron and the porphyrin cation). It has been showed
by Tachibana and coworkers [53] that the electron injec-
tion kinetics from porphyrin and ruthenium complex
sensitizers into TiO2 conduction band are indistinguish-
able, meanwhile the recombination kinetics between the
dye cations and electrons injected into the oxide film are
strongly dependent upon experimental conditions, with
variations from picoseconds to milliseconds in the
recombination rates. It should be noted that PZn–P
dyad bears the remnant carboxylic acid group on the
P moiety, which is located in the electron-withdrawing
porphyrin moiety in order to benefit the orientation of
the PZn–P dyad on the semiconductor tin oxide elec-
trode, and making the free base porphyrin in direct con-
tact with SnO2 substrate in a PZn–P–SnO2 arrangement.
The dependence of the photocurrent efficiency on the
applied potential has been analyzed by other authors
[52,78,79] and showed that recombination rates between
photoinjected electrons and a dye cation increase as
more negative potentials are applied to the semiconduc-
tor. Lower cell efficiencies at negative bias correlate with
this increase in the charge recombination rate. These
studies suggested that the increase in recombination rate
constant at more negative potentials implies the level of
the steady-state trap occupancy in the nanostructured
semiconductor. In order to avoid this factor in the com-
parative analysis of the efficiency of the different dyes
used in this work, the measurements were carried out
under potential control. The effect of the applied poten-
tial on the observed photocurrent generation is shown in
Fig. 8 for different ITO/SnO2/dye electrodes. In this fig-
ure the value of the photocurrent quantum yield ex-
pressed as the Uinj gc value of the systems at the Soret
bands are plotted vs. the applied potential. Uinj gc in-
creases with the anodic bias in all cases, mainly due to
the increase in the collection efficiency of the injected
electrons.
0.25
0.20
0.15
0.10
0.05
0.00
-300
-200
-100
0
100
200
Applied Potential (V)
Fig. 8. Dependence of the product of the charge injection yield (Uinj
)
and charge collection efficiency gc at the maximum of the Soret bands
of ITO/SnO2/PZn–P (d), ITO/SnO2/PZn (m) and ITO/SnO2/P (j)
electrodes on applied bias.
1.5
1.2
0.9
0.6
0.3
0.0
500
550
600
650
700
Wavelength / nm
Fig. 9. Photocurrent (-------) and absorption (
PZn–P dyad in Q band wavelength region, normalized to one at 520 nm.
) spectra of
Fig. 8 also shows that the photocurrent quantum
yield is higher in the dyad molecule than for the sepa-
rated moieties in all applied potential range. Moreover,
the analysis of the photocurrent action spectra of the
PZn–P dyad in the 500–700 nm region shows that both
porphyrin moieties contribute with similar quantum
yield to the observed photoelectric effect. Fig. 9 shows
the photocurrent and absorption spectra normalized to
one at 520 nm, where only free base moiety absorbs
radiation. The closely match between both spectra dem-
onstrate that there is not difference in the contribution
from the two moieties to the global photocurrent ob-
served. Thus, the excitation of either moiety results in
electron injection into SnO2 conduction band, with sim-
ilar electron harvesting capacity.
3.4. Laser-flash photolysis
The changes in the absorbance measured at 430 nm
(dyeÕs Soret band) following excitation after 3.0
mJ cmꢀ2 laser pulse at 532 nm for ITO/SnO2/dye elec-
trodes are shown in Fig. 10. Electron injection results
in fast bleaching followed by regeneration of the ground
state absorbance of the porphyrin compounds. As it has
been reported for several dyes, including porphyrins

106
M. Gervaldo et al. / Chemical Physics 312 (2005) 97–109
0.018
0.016
0.014
0.012
0.010
0.008
0.006
0.004
0.002
0.000
-0.002
electron percolation through the external circuit and
recombination with the oxidized species, seems to be
responsible of the observed different photocurrent
quantum yield. Similar relationship between gc and
pseudo first order back electron transfer kinetic con-
stant has been reported for a series of ruthenium
(II) polypyridyl complexes [85].
However, it should be taking in account that in the
kinetic analysis (Table 1) we assumed pseudo first order
kinetic, fitted by a bi-exponential function. But it is
known that the back electron transfer process is a very
complex phenomenon which involves electrons occupy-
ing conduction band, superficial traps and bulk traps
states [78,83], and consequently a quantitative analysis
of the relation between kb and Uinj gc would be highly
speculative. Moreover, has been reported that a propor-
tion of charge recombination occurs on time scales less
than the time resolution of our laser flash photolysis
set-up [78]. On the other hand, as it has already been
pointed out by Boschloo et al. [86], in practice rather
high laser pulse intensities are needed in order to obtain
sufficient signal in laser flash photolysis experiments on
dye sensitized semiconductor electrodes, making a direct
comparison with conditions that exit under normal illu-
mination slight difficult.
A
B
0.0
1.0
2.0
time / µs
3.0
4.0
5.0
Fig. 10. Kinetic of dye neutral state recovery on SnO2 followed at 430
nm after laser pulse for (a) ITO/SnO2/P and (b) ITO/SnO2/PZn–P
electrodes, kexc = 532 nm, applied bias 0.05 V.
In order to explain our results, a possible energy and
electron transfer mechanism in the ITO/SnO2/PZn–P
electrodes is given in the Scheme 1, according to the
photophysical properties of PZn–P dyad and the ob-
served photoelectric effects.
According this Scheme, both porphyrins moieties in
the dyad can be excited by light absorption (steps a
and b), but the lack of fluorescence from the metallized
unit (both in solution and in adsorbed state) indicates
that energy transfer from P1Zn state to P moiety occurs
with high yield [51,54] (step c).
[53,80], the primary step of electron injection across the
sensitizer/semiconductor interface is extremely fast in
the case of TiO2 semiconductor (<100 fs for tetracarbox-
yphenyl porphyrins [53]). In addition, Benko¨ et al. [81]
demonstrated that the time constant and overall quan-
tum yield of electron injection from RuN3 dye are sim-
ilar for to both TiO2 and SnO2 semiconductors, the
author concluded that electron injection process is al-
most independent of the semiconductor used.
The recovery processes, in absence of sacrificial do-
nor hydroquinone, observed in Fig. 10 are associated
with the reduction of the oxidized species formed after
excitation and electron injection. The regeneration pro-
cess, which occurs by recombination between photoin-
jected electron in the semiconductor and the oxidized
dye, present a comparatively fast step followed by an-
other slower one, showing a complex kinetics [82,83].
Such behavior is often observed in similar heteroge-
neous systems, and is attributed to multiple electrons
trapping in surface defects, which affect the recovery
process [84,85].
hν
PZn-P---SnO2
PZn-P---SnO2
PZn-P1 ---SnO2
PZn1-P ---SnO2
(a)
(b)
hν
k1
PZn1-P---SnO2
PZn-P1---SnO2
PZn-P1 ---SnO2
(c)
(d)
PZn-P+• ---SnO2 (e)
As can been observed in Table 1 (assuming pseudo
first order kinetic constant denoted as kb) back elec-
tron transfer is faster in the case of ITO/SnO2/P elec-
trode than ITO/SnO2/PZn–P. As was already
mentioned, in despite of the fact that the electron
injection efficiency is comparable for both compounds,
P dye is less efficient in the generation of photoelectric
effects than the dyad molecule. Thus, the charge col-
lection efficiency of the systems, which is related to
k
et
PZn-P+• ---SnO2 (e)
PZn+•-P ---SnO2 (e)
PZn-P ---SnO2
(e)
(f)
kb
PZn+•-P ---SnO2 (e)
PZn+•-P ---SnO2 (e) + H2Q
PZn-P ---SnO2 (e) + HQ + H+ (g)
Scheme 1.

M. Gervaldo et al. / Chemical Physics 312 (2005) 97–109
107
The heterogeneous electron transfer process able to
generate photoelectric effect (step d) produces the for-
mation of the radical cation PZn–P+Å, but the intramo-
lecular electron transfer from the PZn to the oxidized
free base (step e) is an exergonic process, as it is inferred
from the 0.48 V difference in the oxidation potentials. As
a consequence, regardless the electron injection path-
tion mechanism is operative in extending the photocur-
rent response of ITO/SnO2/dye electrode to the visible
region. Also it demonstrates that the excitation of either
moiety results in electron injection into SnO2 conduction
band, with similar electron harvesting capacity. It is pro-
posed that inhibition of back electron transfer is favored
by the slower charge recombination kinetic in the
PþZnÅ–P–SnO2(e) state with respect to the case where the
oxidized cation is in direct contact with the semiconduc-
tor. Moreover, fluorescence analysis of PZn–P dyad and
the P moiety adsorbed on both SnO2 and SiO2 shows
that the charge-injection yields (Uinj) from the excited
dye to the tin oxide semiconductor in both cases are sim-
ilar. Thus photocurrent enhancement is interpreted in
terms of intramolecular electron transfer and preferen-
tial spatial orientation of the dyad on the SnO2 surface
that preclude back electron transfer. The architecture
of dyad is a promising organic material for spectral sen-
sitization of semiconductor solar cells.
þÅ
way, it is proposed that the P –P cation is formed (step
Zn
e). Nonetheless, in a previous work we found that the
presence of the oxidized PZn enhanced the back electron
transfer process in the case of a similar metallodiporph-
yrin [51]. This effect produces lower photocurrent yields
in that dyad molecule in comparison with the corre-
sponding monomers.
On the other hand, regardless of the fact that the
compounds with higher back electron transfer rate con-
stant have a corresponding lower photocurrent quan-
tum yield (see Table 1), in agreement with the model
that we suggested, Pzn and Pzn–P showed not so different
recombination rates in the time scale of our laser flash
photolysis experiment. As it was already mentioned,
our kinetic analysis is an oversimplification of a complex
one, and it is possible that more sophisticated experi-
ments are needed (for example ultrafast transient
absorption measurements) in order to obtain a more
certain relationship between kb and Uinj gc.
Acknowledgements
We are grateful to Consejo Nacional de Investigaci-
´
´
ones Cientıficas y Tecnicas (CONICET-Argentina);
´
´
´
Agencia Nacional de Promocion Cientıfica y Tecnolog-
Nevertheless, in the present case, the photocurrent
quantum yield is higher in the dyad molecule, when it
is compared with the free base moiety. Since the dyad
PZn–P was designed with capacity for preferential orien-
tation when it is absorbed onto the semiconductor oxide
(denoted in the Scheme 1 as PZn–P–SnO2), the recombi-
nation process PZþnÅ–P–SnO2ðeÞ ! PZn–P–SnO2 is shown
to be less effective that when the oxidized sensitizer is in
direct contact whit the semiconductor, due to the expo-
nential decay of the electron transfer rate constant with
the distance. Consequently it is highly probable that the
slower charge recombination kinetic in PZn–P in com-
parison with P is the responsible of the photocurrent
yield observed in the dyad sensitized photoelectrodes.
´
´
ica (ANPCYT-Argentina), Secretarıa de Ciencia y Tec-
´
nica de la Universidad Nacional de Rıo Cuarto
(SECYT-UNRC) and Fundacion Antorchas for finan-
´
cial support. L.O., E.N.D. and J.J.S. are scientific mem-
bers of CONICET. M.E.M. and M.G. thank CONICET
for a research fellowship. Thanks are also due to C.M.
Previtalli for his help in Laser Flash Photolysis
experiments.
References
[1] B. OÕRegan, M. Gra¨tzel, Nature 353 (1991) 737.
[2] A. Kay, M. Gra¨tzel, Chem. Mater. 14 (2002) 2930.
[3] R.D. McConnell, Renew. Sustainable Energy Rev. 6 (2002) 271.
[4] T. Stergiopoulos, I.M. Arabatzis, H. Cachet, P. Falaras, J.
Photochem. Photobiol. A: Chem. 155 (2003) 163.
[5] A. Kay, M. Gra¨tzel, J. Phys. Chem. 97 (1993) 6272.
[6] A. Kay, R. Humphry-Baker, M. Gra¨tzel, J. Phys. Chem. 98
(1994) 952.
[7] S. Hotchandani, P.V. Kamat, Chem. Phys. Lett. 191 (1992) 320.
[8] I. Bedja, S. Hotchandani, R. Carpentier, R. Fessenden, P.V.
Kamat, J. Appl. Phys. 75 (1994) 5444.
[9] I. Bedja, P.V. Kamat, S. Hotchandani, J. Appl. Phys. 80 (1996)
4637.
[10] H. Deng, H. Mao, Z. Lu, J. Li, H. Xu, J. Photochem. Photobiol.
A: Chem. 110 (1997) 47.
[11] G.K. Boschloo, A. Goossens, J. Phys. Chem. 100 (1996) 19489.
[12] L. Otero, H. Osora, W. Li, M.A. Fox, J. Porphyrins Phthalocy-
anines 2 (1998) 123.
[13] H.H. Deng, Z.H. Lu, Supramol. Sci. 5 (1998) 669.
[14] J. Wienke, T.J. Schaafsma, A. Goossens, J. Phys. Chem. B 103
(1999) 2702.
4. Conclusions
A new porphyrin-porphyrin non-symmetric dyad has
been used as light receptor in the spectral sensitization of
nanocrystalline SnO2 semiconductor electrode. The
dyad contains structural moieties with both different sin-
glet state energy and redox properties, and it is able to
produce an electron transport – molecular antenna ef-
fect. The heterogeneous charge injection from PZn–P is
favored in the dyad due to the preferential orientation
through the free carboxylic acid group of the P moiety
able to bind to the semiconductor surface. The close
match between the photocurrent action spectrum and
the absorption spectrum shows that the photosensitiza-

108
M. Gervaldo et al. / Chemical Physics 312 (2005) 97–109
[15] M. Linke-Schaetzel, A.D. Bhise, H. Gliemann, T. Koch, T.
Schimmel, T.S. Balaban, Thin Solid Films 451 (2004) 16.
[16] G. Feher, J.P. Allen, M.I. Okamura, D.C. Rees, Nature 339
(1989) 111.
[49] F. Fungo, L. Otero, C.D. Borsarelli, E.N. Durantini, J.J. Silber,
L. Sereno, J. Phys. Chem. B 106 (2002) 4070.
[50] F. Fungo, L. Otero, L. Sereno, J.J. Silber, E.N. Durantini, J.
Mater. Chem. 10 (2000) 645.
[17] D. Gust, T. Moore, in: D. Volman, G. Hammond, D. Neckers
(Eds.), Advances in Phototchemistry, vol. 16, 1991, p. 1.
[18] D. Kuciauskas, P.A. Liddell, S. Lin, T.E. Johnson, S.J. Weghorn,
J.S. Lindsey, A.L. Moore, T.A. Moore, D. Gust, J. Am. Chem.
Soc. 121 (1999) 8604.
[51] F. Fungo, L. Otero, E.N. Durantini, J.J. Silber, L. Sereno, J.
Phys. Chem. B 104 (2000) 7644.
[52] S.A. Haque, Y. Tachibana, D.R. Klug, J.R. Durrant, J. Phys.
Chem. B 102 (1998) 1745.
[53] Y. Tachibana, S.A. Haque, I.P. Mercer, J.R. Durrant, D.R.
Klug, J. Phys. Chem. B 104 (2000) 1198.
[19] M.R. Wasielewski, Chem. Rev. 92 (1992) 435.
´
[20] S.L. Springs, D. Gosztola, M.R. Wasielewski, V. Kral, A.
[54] R.B.M. Koehorst, G.K. Boschloo, T.J. Savenije, A. Goossens,
T.J. Schaafsma, J. Phys. Chem. B 104 (2000) 2371.
[55] M.A. Milanesio, M. Gervaldo, L. Otero, L. Sereno, J.J. Silber,
E.N. Durantini, J. Porphyrins Phthalocyanines 7 (2003) 42.
[56] K.M. Kadish, M.M. Morrison, J. Am. Chem. Soc. 98 (1976)
3326.
[57] K.M. Kadish, M.M. Morrison, Inorg. Chem. 15 (1976) 980.
[58] K.M. Kadish, M.M. Morrison, Bioinorg. Chem. 7 (1977) 107.
[59] F.A. Walker, D. Beroiz, K.M. Kadish, J. Am. Chem. Soc. 98
(1976) 3484.
Andrievsky, J.L. Sessler, J. Am. Chem. Soc. 121 (1999)
2281.
[21] H. Imahori, D.M. Guldi, K. Tamaki, Y. Yoshida, C. Luo, Y.
Sakata, S. Fukuzumi, J. Am. Chem. Soc. 123 (2001) 6617.
[22] H. Imahori, K. Tamaki, Y. Araki, T. Hasobe, O. Ito, A.
Shimomura, S. Kundu, T. Okada, Y. Sakata, S. Fukuzumi, J.
Phys. Chem. A 106 (2002) 2803.
[23] H. Imahori, Y. Mori, Y. Matano, J. Photochem. Photobiol. C:
Photochem. Rev. 4 (2003) 51.
[24] A.F. Noriega, L.F.O. Frutado, A.L.B. Formiga, M. Nakamura,
K. Araki, H.E. Toma, Inorgan. Chem. 43 (2004) 396.
[25] J.N. Clifford, E. Palomares, M.K. Nazeeruddin, R. Thampi, M.
Gratzel, J.R. Durrant, J. Am. Chem. Soc. 126 (2004) 5670.
[26] H. Deng, Z. Lu, Y. Shen, H. Mao, H. Xu, Chem. Phys. 231
(1998) 95.
[27] K. Takahashi, I. Nakajima, K. Imoto, T. Yamaguchia, T.
Komura, K. Murata, Sol. Energy Mater. Sol. Cells 76 (2003) 115.
[28] J. Chen, A.K. Burrell, W.M. Campbell, D.L. Officer, C.O. Too,
G.G. Wallace, Electrochim. Acta 49 (2004) 329.
[60] D. Gust, T.A. Moore, A.L. Moore, L. Leggett, S. Lin, J.M.
DeGraziano, R.M. Hermant, D. Nicodem, P. Craig, G.R. Seely,
R.A. Nieman, J. Phys. Chem. 97 (1993) 7926.
[61] D. Gust, T.A. Moore, A.L. Moore, H.K. Kang, J.M. DeGrazi-
ano, P.A. Liddell, G.R. Seele, J. Phys. Chem. 97 (1993) 13637.
[62] T. Ma, K. Inoue, K. Yao, H. Noma, T. Shuji, E. Abe, J. Yu, X.
Wang, B. Zhang, J. Electroanal. Chem. 537 (2002) 31.
[63] T. Ma, K. Inoue, H. Noma, K. Yao, E. Abe, J. Photochem.
Photobiol. A: Chem. 152 (2002) 207.
[64] P. Bonhoˆte, J.-E. Moser, R. Humphry-Baker, N. Vlachopoulos,
S.M. Zakeeruddin, L. Walder, M. Gra¨tzel, J. Am. Chem. Soc.
121 (1999) 1324.
[29] K. Miyairi, E. Itoh, Y. Hashimoto, Thin Solid Films 438 (2003)
147.
[30] J.H. Yu, F. Li, X. Wang, Y. Huang, B. Zhang, C. Huang, Y.
Cao, Opt. Mater. 21 (2003) 467.
[65] J.N. Clifford, E. Palomares, Md.K. Nazeeruddin, M. Gratzel, J.
Nelson, X. Li, N.J. Long, J.R. Durrant, J. Am. Chem. Soc. 126
(2004) 5225.
´
[31] D. Wrobel, J. Lukasiewicz, J. Goc, A. Waszkowiak, R. Ion, J.
Mol. Struct. 555 (2000) 407.
[66] G.R. Geier III, J.S. Lindsey, J. Chem. Soc. Perkins Trans. 2
(2001) 677.
´
[32] D. Wrobel, Comptes Rendus Chimie 6 (2003) 417.
[33] J.E. Moser, P. Bonhoˆte, L. Walder, M. Gra¨tzel, Chimia 51 (1997)
28.
[34] M.K. Nazeeruddin, A. Kay, I. Rodicio, R. Humphry-Baker, E.
Muller, P. Liska, N. Vlachopoulus, M. Gra¨tzel, J. Am. Chem.
Soc. 115 (1993) 6382.
[35] G. Smestad, Sol. Energy Mater. Sol. Cells 32 (1994) 273.
[36] A. Hagfeld, M. Gra¨tzel, Chem. Rev. 95 (1995) 49.
[37] P.V. Kamat, Prog. React. Kinetics 19 (1994) 277.
[38] I. Bedja, S. Hotchandani, P.V. Kamat, J. Phys. Chem. 98 (1994)
4133.
[39] P.V. Kamat, CHEMTECH 25 (1995) 22.
[40] P.V. Kamat, I. Bedja, S. Hotchandani, L. Patterson, J. Phys.
Chem. 100 (1996) 4900.
[67] E.N. Durantini, J.J. Silber, Synth. Commun. 29 (1999) 3353.
[68] T. Stergiopoulus, I.M. Arabatzis, H. Cachet, P. Falaras, J.
Photochem. Photobiol. A: Chem. 155 (2003) 163.
[69] N.-G. Park, M.G. Kang, K.S. Ryu, K.M. Kim, S.H. Chang, J.
Photochem. Photobiol. A: Chem. 161 (2004) 105.
[70] H. Gerischer, F. Willig, Top. Curr. Chem. 61 (1976) 31.
[71] C.D. Borsarelli, J.J. Cosa, C.M. Previtali, Photochem. Photobiol.
68 (1998) 438.
[72] K.M. Kadish, N. Guo, E. Van Caemelbeckie, R. Paolesse, D.
Monti, P. Tagliatesta, J. Porphyrins Phthalocyanines 2 (1998)
439.
[73] D. Gust, T. Moore, A. Moore, F. Gao, D. Luttrull, J. DeGra-
zaino, X. Ma, L. Makings, S.-J. Lee, T. Trier, E. Bittersmann, G.
´
[41] I. Bedja, P.V. Kamat, X. Hua, A.G. Lappin, S. Hotchandani,
Lagmuir 13 (1997) 2398.
[42] H. Sugihara, L.P. Singh, K. Sayama, H. Arakawa, M.K.
Nazeeruddin, M. Gra¨tzel, Chem. Lett. 10 (1998) 1005.
[43] S.D. Burnside, K. Brooks, A.J. McEvoy, M. Gra¨tzel, Chimia 52
(1998) 557.
Seely, S. Woodward, R. Bensansson, M. Rougee, F. De Schryver,
M. Van der Auweraer, J. Am. Chem. Soc. 113 (1991) 3638.
[74] T. Kajiwara, K. Hashimoto, T. Kawal, T. Sakata, J. Phys. Chem.
86 (1982) 4516.
[75] T. Sakata, K. Hashimoto, M. Hiramoto, J. Phys. Chem. 94 (1990)
3040.
[44] H. Imahori, H. Yamada, Y. Nishimura, I. Yamazaki, Y. Sakata,
J. Phys. Chem. B 104 (2000) 2099.
[76] K. Hashimoto, M. Hiramoto, T. Sakata, J. Phys.Chem. 92 (1988)
4272.
[45] H. Imahori, H. Norieda, H. Yamada, Y. Nishimura, I. Yamazaki,
Y. Sakata, S. Fukuzumi, J. Am. Chem. Soc. 123 (2001) 100.
[46] D. Hirayama, K. Takimiya, Y. Aso, T. Otsubo, T. Hasobe, H.
Yamada, H. Imahori, S. Fukuzumi, Y. Sakata, J. Am. Chem.
Soc. 124 (2002) 532.
[77] P. Vlachopoulos, P. Liska, J. Augustynski, M. Gra¨tzel, J. Am.
Chem. Soc. 110 (1988) 1216.
[78] S.A. Haque, Y. Tachibana, R.L. Willis, J.E. Moser, M.
Gratzel, D.R. Klug, J.R. Durrant, J. Phys. Chem.
(2000) 538.
B 104
[47] T. Hasobe, H. Imahori, S. Fukuzumi, P.V. Kamat, J. Phys.
Chem. B 107 (2003) 12105.
[79] D. Kuciauskas, M.S. Freund, H.B. Gray, J.R. Winkler, N.S.
Lewis, J. Phys. Chem. B 105 (2001) 392.
[48] T. Hasobe, H. Imahori, P.V. Kamat, S. Fukuzumi, J. Am. Chem.
Soc. 125 (2003) 14962.
[80] Y. Tachibana, Md.K. Nazeeruddin, M. Gratzel, D.R. Klug, J.R.
Durrant, Chem. Phys. 285 (2002) 127.

M. Gervaldo et al. / Chemical Physics 312 (2005) 97–109
109
[81] G. Benko¨, P. Myllyperkio¨, J. Pan, A.P. Yartsev, V. Sundstro¨m, J.
Am. Chem. Soc. 125 (2003) 1118.
[82] J. Bisquert, A. Zaban, P. Salvador, J. Phys. Chem. B 106 (2002)
8774.
[83] J. Bisquert, A. Zaban, M. Greenshtein, I. Mora-Sero, I.J. Am.
Chem. Soc. 126 (2004) 13550.
[84] C. Graziani Garcia, C.J. Kleverlaan, C.A. Bignozzi,
N.Y.M. Iha, J. Photochem. Photobiol. A: Chem. 147
(2002) 143.
[85] C. Graziani Garcia, A.K. Nakano, C.J. Kleverlaan, N.Y.M. Iha,
J. Photochem. Photobiol. A: Chem. 151 (2002) 165.
[86] G. Boschloo, A. Hagfeldt, Chem. Phys. Lett. 370 (2003) 381.