Preparation and characterization of core-shell electrodes for application in gel electrolyte-based dye-sensitized solar cells

Article abstract of DOI:10.1016/j.electacta.2009.05.024

Core-shell electrodes based on TiO₂ covered with different oxides were prepared and characterized. These electrodes were applied in gel electrolyte-based dye-sensitized solar cells (DSSC). The TiO₂ electrodes were prepared from TiO₂ powder (P25 Degussa) and coated with thin layers of Al₂O₃, MgO, Nb₂O₅, and SrTiO₃ prepared by the sol-gel method. The core-shell electrodes were characterized by X-ray diffraction, scanning electron microscopy and atomic force microscopy measurements. J-V curves in the dark and under standard AM 1.5 conditions and photovoltage decay measurements under open-circuit conditions were carried out in order to evaluate the influence of the oxide layer on the charge recombination dynamics and on the device's performance. The results indicated an improvement in the conversion efficiency as a result of an increase in the open circuit voltage. The photovoltage decay curves under open-circuit conditions showed that the core-shell electrodes provide longer electron lifetime values compared to uncoated TiO₂ electrodes, corroborating with a minimization in the recombination losses at the nanoparticle surface/electrolyte interface. This is the first time that a study has been applied to DSSC based on gel polymer electrolyte. The optimum performance was achieved by solar cells based on TiO₂/MgO core-shell electrodes: fill factor of ～0.60, short-circuit current density J_sc of 12 mA cm^-2, open-circuit voltage V_oc of 0.78 V and overall energy conversion efficiency of ～5% (under illumination of 100 mW cm^-2).

Full text of DOI:10.1016/j.electacta.2009.05.024

Electrochimica Acta 55 (2010) 1468–1474
Contents lists available at ScienceDirect
Electrochimica Acta
journal homepage: www.elsevier.com/locate/electacta
Preparation and characterization of core-shell electrodes for application in gel
electrolyte-based dye-sensitized solar cells
César O. Avellaneda, Agnaldo D. Gonc¸ alves, João E. Benedetti, Ana F. Nogueira∗
Laboratório de Nanotecnologia e Energia Solar (LNES), Instituto de Química, Universidade Estadual de Campinas – UNICAMP, P.O. Box 6154, 13083-970 Campinas/SP, Brazil
a r t i c l e i n f o
a b s t r a c t
Article history:
Core-shell electrodes based on TiO2 covered with different oxides were prepared and characterized. These
electrodes were applied in gel electrolyte-based dye-sensitized solar cells (DSSC). The TiO2 electrodes
were prepared from TiO2 powder (P25 Degussa) and coated with thin layers of Al2O3, MgO, Nb2O5, and
SrTiO3 prepared by the sol–gel method. The core-shell electrodes were characterized by X-ray diffraction,
scanning electron microscopy and atomic force microscopy measurements. J–V curves in the dark and
under standard AM 1.5 conditions and photovoltage decay measurements under open-circuit conditions
were carried out in order to evaluate the influence of the oxide layer on the charge recombination dynam-
ics and on the device’s performance. The results indicated an improvement in the conversion efficiency
as a result of an increase in the open circuit voltage. The photovoltage decay curves under open-circuit
conditions showed that the core-shell electrodes provide longer electron lifetime values compared to
uncoated TiO2 electrodes, corroborating with a minimization in the recombination losses at the nanopar-
ticle surface/electrolyte interface. This is the first time that a study has been applied to DSSC based on gel
polymer electrolyte. The optimum performance was achieved by solar cells based on TiO2/MgO core-shell
electrodes: fill factor of ∼0.60, short-circuit current density Jsc of 12 mA cm−2, open-circuit voltage Voc of
0.78 V and overall energy conversion efficiency of ∼5% (under illumination of 100 mW cm−2).
© 2009 Elsevier Ltd. All rights reserved.
Received 22 October 2008
Received in revised form 5 May 2009
Accepted 7 May 2009
Available online 14 May 2009
Keywords:
Dye-sensitized solar cells
Core-shell
Gel electrolyte
Recombination
1. Introduction
usually an organic solvent where the redox couple I3−/I− [2,3] is dis-
solved. Such DSSC configurations require a perfect sealant, in such
Dye-sensitized solar cells (DSSC) have currently been attract-
ing widespread academic and intense commercial investigation in
view of their potential low-cost and their use of environmentally
friendly materials endowed with good efficiencies, comparable to
the well-established silicon technology [1]. The energy conver-
sion working principle is based upon the injection of electrons
from the photoexcited state of a sensitizer into the conduction
band of an oxide semiconductor. A charge mediator, i.e., a suit-
able redox couple, may be added to the electrolyte to reduce
the oxidized dye. The mediator also needs to be renewed in the
counter-electrode, making the photoelectrochemical cell regener-
ative [2–3].
In general, these devices are sandwich-type photoelectrochem-
ical cells, in which the photoelectrode consists of a porous film
of dye-sensitized nanocrystalline TiO2 particles and the counter-
electrode is usually composed of a thin Pt film, both deposited on
the surface of a transparent electrode, e. g., a glass substrate coated
with a layer of fluorine-doped tin oxide (FTO). The electrolyte is
a way that, for long term stability, leakage and evaporation of the
solvent must be avoided, along with contamination from impurities
and water [4–7].
Several attempts have been made to substitute the liquid elec-
trolyte in the DSSC cells. In our laboratory, we have explored
polymer and gel electrolytes based on poly(ethylene oxide) deriva-
tives [6–8]. We also investigated the addition of inorganic clays
[9], crown ethers [10] and plasticizers [11,12] in these polymers in
order to improve the ionic conductivity and, as a consequence, the
device’s performance. However, the devices assembled with these
modified electrolytes all suffer from low open-circuit voltage (Voc
)
values. We believe that the recombination losses (photoinjected
electrons return to the dye cation and/or to the triiodide ions) are
accelerated when replacing a very basic polymer by such additives
which present lower basicity. The role of the basicity can be viewed
in two different manners: the ethylene oxide groups, present in the
polymer, interact strongly with Ti(IV) acid sites, acting as a pro-
tective layer against charge recombination or the smaller size of
such additives plus their coordinating ability drive the lithium ions
deeper inside the porous film, where the polymer cannot reach. The
increase in the concentration of lithium ions inside the porous film
corroborates for such potential drop.
∗
Corresponding author. Tel.: +55 19 3521 3029; fax: +55 19 3521 3023.
E-mail address: anaflavia@iqm.unicamp.br (A.F. Nogueira).
0013-4686/$ – see front matter © 2009 Elsevier Ltd. All rights reserved.
doi:10.1016/j.electacta.2009.05.024

C.O. Avellaneda et al. / Electrochimica Acta 55 (2010) 1468–1474
1469
High performance DSSCs require nanocrystalline TiO2 elec-
trodes with large surface areas, high crystallinity and without
cracks and a favorable electrical contact with the conducting glass
substrate so that high amounts of dye molecules can be adsorbed
and the electrons can be quickly transferred [13].
Instruments. The AFM images were obtained using a Veeco AFM
instrument.
2.2. Gel polymer electrolyte preparation and characterization
Poly(ethylene oxide-co-2-(2-methoxyethoxy)ethyl glycidyl
ether) P(EO-EM) was used as received from Daiso Co. (Osaka,
Japan). The ethylene oxide: 2-(2-methoxyethoxy)ethyl glycidyl
ether ratio in the copolymer was 78:22 and its molecular weight
was ca. 1 × 106 g mol−1, according to the supplier. The gel polymer
electrolytes were prepared by the dissolution of the copolymer,
LiI, I2, and ␥-butyrolactone (GBL) in 25 mL of acetone. The P(EO-
EM):GLB weight ratio was 0.3:0.7 (wt/wt). Electrolyte solutions
were kept under stirring for 1 week before use.
Ionic conductivity measurements were evaluated as a function
of salt concentration at constant temperature (25 ◦C). All samples
comprised a gel polymer electrolyte (film thickness ∼100 ␮m),
deposited by casting the electrolyte solution onto a Teflon disk
under a saturated atmosphere. Afterwards, the films were detached
from the Teflon disk by dipping into liquid nitrogen, and further
dried under vacuum for 144 h or more. Conductivity measure-
ments were carried out in an MBraun dry box (humidity < 10−4%,
under an argon atmosphere). The films were fixed between two
mirror-polished stainless steel disc-shaped electrodes (diame-
ter = 12 mm) using a Teflon® spacer 0.05 mm thick to prevent
short-circuits and also to maintain a fixed thickness. The con-
ductivity values were calculated from electrochemical impedance
spectroscopy (EIS) data, using an Eco-Chemie Autolab PGSTAT 12
with a computer-coupled FRA module over the frequency range
of 106 to 10 Hz with an ac amplitude of 10 mV under open-circuit
potential (Voc). The ionic conductivity was calculated from the bulk
electrolyte resistance value estimated from the complex impedance
diagram.
However, the major problem confronting the economic via-
bility of this technology is still the lower energy conversion
efficiency when compared with silicon solar cells. This is primarily
due to charge recombination losses arising at the semiconduc-
tor/dye/electrolyte interface and low dye absorption towards the
infrared region [14]. Recombination with the dye cations and/or
the electrolyte species (triiodide) can drastically affect the Voc val-
ues [15]. The introduction of a very thin oxide layer, such as ZnO
[16], CaCO3 [17], Nb2O5 [18], SrTiO3 [19], and MgO [20], onto
the TiO2 nanoparticles, creating a core-shell structure, provides
means to minimize the charge recombination losses in DSSCs. The
improvement in the solar cell energy conversion efficiency by the
overcoating approach may be assigned to the following factors:
(i) the wide band gap coating delays the electron back trans-
fer to the electrolyte and minimizes charge recombination [21],
(ii) the coating layer also enhances the dye adsorption onto the
porous electrode and, as a consequence, the dye loading, increas-
ing the photocurrent [17]. In this work, a very thin film of the
following oxide materials (Nb2O5, SrTiO3, Al2O3 and MgO) were
deposited onto a TiO2 electrode prepared using TiO2 powder (P25
Degussa). The films were characterized by powder X-ray diffraction
(XRD), energy dispersive X-ray diffraction (EDS), scanning elec-
tronic microscopy (SEM) and atomic force microscopy (AFM) to
analyze the morphology of the films. The effects of these oxide lay-
ers on the performance of DSSCs assembled with a gel polymer
electrolyte were investigated.
2. Experimental
2.1. Working electrode preparation and characterization
2.3. Dye-sensitized solar cell assembly and characterization
The mesoporous layer of TiO2 was prepared by grinding with a
pestle 6 g of TiO2 powder (P25 Degusa) and 2.4 g of polyethylene
glycol (PEG 20000) (Merck, reagent grade) in 2 mL of distilled H2O
and 0.2 mL of acetylacetone (Aldrich, reagent grade) in a mortar for
40 min. Finally, 8.0 mL of distilled water and 0.1 mL of Triton X-100
(Aldrich, reagent grade) were slowly added under continuous mix-
ing for 10 min. Several drops of this suspension were spread onto
the electrode surface using a glass rod. The electrode was dried
in air and heated at 450 ◦C for 30 min. The TiO2 films presented a
thickness of approximately 7 ␮m measured with an Alpha-Step 200
profilometer (Tencor Instruments). The films were visually homo-
geneous without any cracks.
Al2O3, MgO, Nb2O5, and SrTiO3 precursor solutions were pre-
pared using a sol–gel method described in the literature [22].
Aluminum nitrate nonahydrate (Al(NO3)3·9 H2O, Aldrich, reagent
grade), magnesium nitrate hexahydrate (Mg(NO3)2·6H2O), Vetec,
98%), niobium ethoxide (Nb(OCH2CH3)5, Aldrich, 99%), and stron-
tium chloride hexahydrate (SrCl2·6H2O, Acros, 99%) were used
as precursors. Thin Al2O3, MgO, Nb2O5, and SrTiO3 layers were
deposited by dip-coating onto the TiO2 film with a speed of
12 cm min−1 under ambient conditions. Afterwards the samples
were again treated at 500 ◦C for 30 min in air.
DSSCs were assembled using FTO (Hartford glass, Rs ≤ 10 ꢁ m2)
and ITO (Delta Technologies, Rs ≤ 12 ꢁ m2) glass as conducting
substrates for the preparation of the photoelectrodes and the
counter-electrodes, respectively. The counter-electrodes were pre-
pared by sputtering thin Pt films (400 Å) onto the ITO substrate.
The working electrodes at the temperature of 70 ◦C
were immersed in
a
0.5 × 10−3 mol L−1 ethanolic solution
of the sensitizer cis-dithiocyanate-N,Nꢀ-bis(4-carboxylate-4-
tetrabutylammoniumcarboxylate-2,2ꢀ-bipyridine)ruthenium(II),
also known as Ruthenium N719 (Solaronix), for 16 h at room
temperature. The electrodes were then rinsed with ethanol and
dried. The gel polymer electrolyte was deposited onto the porous
nanostructured electrodes by casting at a temperature of 60 ◦C to
remove the acetone from the electrolyte solution. The DSSCs were
finally assembled by pressing the Pt counter-electrode against
the sensitized porous electrode coated with the gel polymer
electrolyte. A thin PVC (polyvinyl chloride) film was used as the
spacer.
The DSSCs were characterized on an optical bench using an
Oriel Xe(Hg) 250 W lamp, lenses, water and AM1.5 filters (all from
Oriel). The light intensity was measured with a Newport optical
power meter, model 1830-C. Current–potential curves (J–V curves)
were measured by linear sweep voltammetry using the Eco Chimie-
Autolab PGSTAT 12 potentiostat. The J–V data were measured in the
dark and under standard AM 1.5 condition (1 sun, 100 mW cm−2).
The J–V curves were fitted using a two diode equation [23]. The
ideality factor m1 and series resistance Rs values were estimated
initially from the I–V curves in the dark. Then m2 and k (fitting con-
stant) were calculated from the I–V curve at 1 sun [23]. The voltage
decay measurements were carried out under open-circuit condi-
In order to identify the structural properties of the samples,
X-ray diffraction patterns were carried out in a Rigaku D/max-
EB diffractometer with Cu K␣ radiation over the 2ꢀ range of
10–50◦. The measurements were carried out by using powder
samples of TiO2 and TiO2 core-shell samples. The morphologi-
cal properties of the TiO2 and TiO2 core-shell electrode surface
were analyzed by SEM microscope coupled to an Energy Dispersive
X-ray detector using the SEM-EDS-J6360-LV from Thermo Noran

1470
C.O. Avellaneda et al. / Electrochimica Acta 55 (2010) 1468–1474
Fig. 1. Nyquist plots of the gel polymer electrolytes as a function of LiI content.
Fig. 2. X-ray diffraction patterns of TiO2 and the TiO2/core-shell electrodes.
tion by switching off the illumination and monitoring the Voc decay
in the dark for a given period of time [24,25].
The gel polymer electrolyte containing 20 wt % of LiI was then
chosen to continue this work due to its high ionic conductivity. Fre-
itas et al. reported that, for a similar polymer electrolyte system
above this salt concentration, the charge recombination is accel-
erated and no increase in the dye cation regeneration could be
observed [32].
3. Results and discussion
3.1. Electrolyte properties
The XRD patterns were carried out in order to evaluate any struc-
tural modification of the TiO2 electrode after coating with Al2O3,
Typical Nyquist plots are shown in Fig. 1. The Nyquist plots did
not exhibit a semicircle at the high frequency region (as shown in
Fig. 1, inset), as normally observed for polymer electrolyte samples
(salt and polymer only). The disappearance the semi-circle is due
to the high ionic conductivity of our samples. A straight line was
observed instead. The intersection of this straight line with the real
axis provides the resistance value usually employed to calculate the
ionic conductivity of the gel polymer electrolyte according to Eq. (1)
[26,27]
ꢂ = l/ARb
(1)
where ꢂ is the ionic conductivity, l is the film thickness, A is the
electrode area and Rb is the electrolyte resistance obtained from
the intercept of the Nyquist plot with the real axis.
Table 1 shows the ionic conductivity of the polymer electrolyte
as a function of LiI content. As may be observed, the ionic conduc-
tivity increases slightly up to 10 wt% of LiI. For a common polymer
electrolyte system (polymer and salt only), a decrease in the ionic
conductivity would be expected after a certain salt concentration
due to the formation of ionic pairs and crosslinking sites [28,29].
Such effects make the segmental motion of the polymeric chains
difficult and, as a consequence, lowers the ionic mobility. However,
above 10 wt% of LiI we observe a pronounced increase in the ionic
conductivity, reaching 1.9 × 10−3 S cm−1 for the sample with 20 wt%
of salt. This value is close to the values observed in a system based
on a salt dissolved in an organic solvent. Such high conductivity can
be explained by considering the formation of polyiodide species. It
is well known that polyiodides can be formed at high concentra-
tions of I2 in the presence of I−, improving the conductivity by the
introduction of a new electronic conduction pathway [30,31].
Table 1
Ionic conductivity values (S cm−1) as a function of the LiI content in the
gel polymer electrolyte. The amount of iodine was adjusted to be 10:1
(LiI:I2).
LiI concentration (wt%)
Conductivity (S cm−1
)
2
5
10
15
20
2.1 × 10−4
5.2 × 10−4
5.7 × 10−4
1.1 × 10−3
1.9 × 10−3
Fig. 3. SEM images of TiO2 (a) and TiO2/MgO (b) electrodes.

C.O. Avellaneda et al. / Electrochimica Acta 55 (2010) 1468–1474
1471
Fig. 4. EDS measurements of the TiO2 (a) and (b) TiO2–Al2O3, (c) TiO2–MgO, (d) TiO2–Nb2O5 and (e) TiO2–SrTiO3 core-shell films.
SrTiO3, Nb2O5 and MgO shells. The XRD patterns (Fig. 2) indicate
that none of the coatings affected the TiO2 composition as there
is no apparent shift in the existing peaks assigned to rutile and
anatase phases. We estimated that TiO2 P25 (Degussa) consists of
20% rutile and 80% anatase, consistent with the published crys-
tal data for the commercial powder (Degussa P 25). We believe
that the thickness of the coating layers consist of a few nanome-
ters, this would explain why no peaks from these layers were
detected.
Scanning electron microscopy (SEM) images, shown in Fig. 3a
and b, reveal that the surface of TiO2 and of the TiO2/MgO films
possesses a sponge-like structure, with extended roughness and
porosity, being more pronounced for the TiO2/MgO electrode. The
overlayers were not clearly observed by high-resolution transmis-
sion electron microscopy (HRTEM, data not shown), probably due
to their very thin thickness and/or equipment resolution. However,
the EDS measurements (Fig. 4) confirm the presence of Ti, Sr, Al and
Mg in the corresponding core-shell films.
of the TiO2 nanoparticles. In general, the appearance of the films
resembles a porous network with extended surface area, suitable
for heterogeneous energy conversion processes.
Atomic force microscopy (AFM) has been demonstrated to be
a very versatile and powerful tool for surface imaging at the sub-
micrometer level and to reveal the surface characteristics of films.
Fig. 5a and b depicts a typical three-dimensional representation
(1 ␮m × 1 ␮m surface plot) of the TiO2 and TiO2/MgO films. Both
films present a rough surface texture and consist of interconnected
grain particles fused together to build up high mountains and deep
valleys. The Rms (Rms = the standard deviation of the Z values, Z
being the total range analyzed) values are 17 and 26 nm for TiO2
and TiO2/MgO films, respectively. The AFM images are in agree-
ment with the SEM images, showing a slight increase in the surface
roughness after coating.
3.2. DSSC electrical characterization
The mean diameter of the nanocrystallites is controlled by the
original Degussa P25 semiconductor material. Results show that
the modification with the coating does not induce any aggregation
The current-voltage characteristics of the DSSCs based on TiO2
and TiO2 core-shell electrodes measured at a light intensity of
100 mW cm−2 are shown in Fig. 6. The inset shows the curves

1472
C.O. Avellaneda et al. / Electrochimica Acta 55 (2010) 1468–1474
Table 2
J–V features from the DSSC based on TiO2 and TiO2/core-shell electrodes under 1
sun illumination (100 mW cm−2), along with some parameters from J–V curve fitting
using a two diode model.
Electrode
Voc (V) Jsc (mA/cm2) FF
Efficiency (%) m1
m2
Rs (ꢁ)
TiO2
0.73
0.78
0.75
0.74
0.76
10.9
11.1
11.1
10.9
11.3
0.55
0.57
0.57
0.56
0.53
4.43
4.94
4.78
4.55
4.57
4.9
4.1
4.2
4.5
4.4
5.4
5.6
5.1
5.6
6.6
66
35
34
42
81
TiO2/MgO
TiO2/Nb2O5
TiO2/SrTiO3
TiO2/Al2O3
electrode with a different oxide endowed with higher conduction
band energy [34].
In order to further explore the properties of the DSSC, the J–V
curves were fitted using a two diode model of the form
ꢀ
ꢀ
ꢁ
ꢁ
ꢀ
ꢀ
ꢁ
ꢁ
qVj
qVj
I = IL − I0 exp
− 1 − kIL exp
− 1
(2)
m1kBT
m2kBT
where IL is the light intensity dependent short-circuit current, kB
is Boltzmann’s constant, T is temperature, and I0, k, m1 and m2 are
fitting constants [23]. m1 is related to the charge recombination
and m2 expresses, along with the final term in Eq. (2), the light
dependent recombination current.
Fig. 5. AFM images of TiO2 (a) and TiO2/MgO (b) electrodes.
The bias drop across the internal junction, Vj, can be related to
the externally applied bias, V, through
under dark conditions. The significant changes associated with
the core-shell electrodes arise mainly on the open-circuit volt-
age (Voc) values. The highest performance was achieved by the
DSSC based on the TiO2/MgO core-shell electrodes yielding a fill
factor of ∼60%, Jsc of 12 mA cm−2, a Voc of 0.78 V, and overall
energy conversion efficiency (ꢃ) of ∼5%. This is a remarkable effi-
ciency considering the type of electrolyte employed. Since the
core-shell electrodes provide an inherent energy barrier at the
electrode–electrolyte interface, the electron recombination losses
are remarkably reduced and, consequently, the Voc values increase
[33]. Such increase in Voc may be clearly observed from the dark
current characteristics (see inset in Fig. 6), which show that the
core-shell films reduce the charge recombination. The higher Voc
valuesoftheDSSCsbasedonthecore-shellelectrodesarisefromthe
conduction band edge shift of the TiO2 brought about by coating the
Vj = V + IRs
(3)
where Rs is the series resistance of the system.
The first two terms on the right of Eq. (2) represent a voltage
independent photocurrent, IL, and a light independent recombina-
tion or “dark” current, and together compose the usual nonideal one
diode current voltage characteristic of a photovoltaic cell. The final
term in Eq. (2) is a light dependent recombination current, and it is
required to describe adequately the observed behavior for gel poly-
mer electrolyte cells [23]. The choice of this term is empirical but
follows from the observations that the additional recombination
current increases with light intensity, and hence with short-circuit
photocurrent, and that the net recombination current shows an
exponential dependence, as would be expected from the increas-
ing occupation of electron acceptor states as the Fermi level is
raised.
A remarkable decrease in series resistance (Rs, Table 2) is
observed for the DSSC based on core-shell electrodes with the
exception of the Al2O3 coated film. Such lower Rs values for DSSC
based on core-shell electrodes, compared to bare TiO2, might be
rationalized in terms of the better electrical contact among particles
arising from the core-shell structure. Such improved contact results
in better electron transport properties as will be confirmed in the
next section. The lower FF of the solar cells based on TiO2/Al2O3
provides some evidence for the higher Rs, along with a more pro-
nounced dependence of the photocurrent with the light intensity,
representedherebythe highest m2 value. Such discrepancy inthe Rs
value may be due to the fact that the Al2O3 overlayer has the largest
band gap among the oxides studied herein and/or its thickness that
might be different from the other electrodes.
The diode ideality factor m1 also displays a lower value when
comparing DSSC based on TiO2 and core-shell electrodes. Such
lower m1 values provide evidence for the role of the oxide layer
in reducing recombination losses [23].
Fig. 7 illustrates the voltage decay curves after switching off the
illumination for the DSSC based on TiO2 with or without the oxide
layers. The DSSC using core-shell structures presented a smooth Voc
decay compared to the DSSC employing pristine TiO2 electrodes.
This provides some evidence that the thin insulator layer used
Fig. 6. J–V curves of DSSCs based on TiO2 and on TiO2/core-shell electrodes and gel
polymer electrolyte under 1 sun (100 mW cm−2 AM1.5) and under dark condition
(inset).

C.O. Avellaneda et al. / Electrochimica Acta 55 (2010) 1468–1474
1473
same voltage range under open-circuit conditions. The differences
in the decay curves for all samples might be due to differences in
the thickness of these coatings.
4. Conclusions
DSSC based on TiO2 electrodes coated with various thin oxide
layers such as Al2O3, Nb2O5, MgO and SrTiO3 were assembled and
the effects of the oxide layer on the Voc, Jsc and the resulting solar
cell energy conversion efficiencies were investigated. No significant
changes were observed for the short-circuit current density, but
the Voc values changed notably: from 0.73 V (for TiO2 electrodes)
to 0.78 V for TiO2/MgO core-shell electrodes. The latter electrodes
provided DSSC endowed with overall energy conversion efficien-
cies as high as 5% under 1 sun. This is a remarkable efficiency
considering the type of the electrolyte employed. The use of such
insulating coatings was effective in minimizing the charge recom-
bination and thus increasing the photovoltage in DSSC where the
liquid electrolyte is replaced by a gel polymer electrolyte.
Fig. 7. Voltage decay curves measured after switching off the illumination of DSSCs
based on TiO2 and TiO2/core-shell electrodes under open-circuit condition.
Acknowledgments
herein effectively acts in decreasing the charge recombination rate
at the oxide/electrolyte interface.
The behavior of the electron lifetime (ꢄe) of the DSSC under
open-circuit condition (Fig. 8) can be estimated using Eq. (4) and
the data presented in Fig. 6 according to [24,25].
The authors thank FAPESP (08/51001-9 and 06/58998-3), CNPq,
and Renami for financial support, Prof. Fernando A. Sigoli for the use
of the dip-coating equipment, Daiso (Osaka, Japan) for the donation
of the polymer samples and Prof. Carol H. Collins for English review.
ꢂ
ꢃ−1
kT dVoc
ꢄe = −
(4)
References
e
dt
The exponential-like increase in the electron lifetime as the volt-
age decreases observed for DSSC based on TiO2 under open-circuit
conditions is in excellent agreement to the results reported else-
where [25]. In this work, however, the voltage decays show an
almost linear increase in Voc under open-circuit conditions for all
DSSC based on TiO2 nanoparticles covered with a thin insulator
layer (core-shell structure), suggesting probably a different recom-
bination mechanism.
[1] B. O’Regan, M. Gratzel, Nature 353 (1991) 737.
[2] M.K. Nazeeruddin, A. Kay, I. Rodicio, R. Humphry-Baker, E. Muller, P. Liska, N.
Vlachopoulos, M. Gratzel, J. Am. Chem. Soc. 115 (1993) 6382.
[3] A. Hagfeldt, M. Gratzel, Chem. Rev. 95 (1995) 49.
[4] U. Bach, D. Lupo, P. Conte, J.E. Moser, F. Weissortel, J. Salbeck, H. Spreitzer, M.
Gratzel, Nature 395 (1998) 583.
[5] M. Matsumoto, H. Miyazaki, K. Matsuhiro, Y. Kumashiro, Y. Takaoka, Solid State
Ionics 89 (1996) 263.
[6] A.F. Nogueira, M.-A. De Paoli, Sol. Energy Mater. Sol. Cells 61 (2000) 135.
[7] A.F. Nogueira, J.R. Durrant, M.-A. De Paoli, Adv. Mater. 13 (2001) 826.
[8] A.F. Nogueira, C. Longo, M.-A. De Paoli, Coord. Chem. Rev. 248 (2004) 1455.
[9] B.I. Ito, J.N. Freitas, M.-A. De Paoli, A.F. Nogueira, J. Braz. Chem. Soc. 19 (2008)
688.
[10] J.E. Benedetti, M.-A. De Paoli, A.F. Nogueira, Chem. Commun. 9 (2008) 1121.
[11] V.C. Nogueira, C. Longo, A.F. Nogueira, M.A. Soto-Oviedo, M.-A. De Paoli, J. Pho-
tochem. Photobiol. A: Chem. 181 (2006) 226.
Moreover, the electron lifetime under open-circuit conditions is
longer in the DSSC employing core-shell electrodes. This is exactly
the expected trend, since the thin insulator layer is effectively acting
in decreasing the charge recombination rate in the oxide semicon-
ductor/electrolyte interface. Among the oxides dealt with herein,
SrTiO3 and Nb2O5 provided longer electron lifetime values in the
[12] I.C. Flores, J.N. Freitas, C. Longo, M.-A. De Paoli, H. Winnischofer, A.F. Nogueira,
Photochem. Photobiol. A: Chem. 189 (2007) 153.
[13] J. Jiu, F. Wang, M. Sakamoto, J. Takao, M. Adachi, Sol. Energy Mater. Sol. Cells 87
(2005) 77.
[14] S.Y. Huang, G. Schlichthorl, A.J. Nozik, M. Gratzel, A.J. Frank, J. Phys. Chem. B 101
(1997) 2576.
[15] B.A. Gregg, P. Pichot, S. Ferrere, C.L. Fields, J. Phys. Chem. B 105 (2001) 1422.
[16] Z.S. Wang, C.H. Huang, Y.Y. Huang, Y.J. Hou, P.H. Xie, B.W. Zhang, H.M. Chen,
Chem. Mater. 13 (2001) 678.
[17] S. Lee, J.Y. Kim, K.S. Hong, H.S. Jung, H. Shin, Sol. Energy Mater. Sol. Cells 90
(2006) 2405.
[18] A. Zaban, S.G. Chen, S. Chappel, B.A. Gregg, Chem. Commun. 22 (2000)
22131.
[19] Y. Diamant, S.G. Chen, O. Melamed, A. Zaban, J. Phys. Chem. B 107 (2003)
1977.
[20] H.S. Wang, J.K. Lee, M. Natasi, S.W. Lee, J.Y. Kim, J.S. Park, K.S. Hong, H. Shin,
Langmuir 21 (2005) 10332.
[21] E. Palomares, J.N. Clifford, S.A. Haque, T. Lutz, J.R. Durrant, J. Am. Chem. Soc. 125
(2003) 473.
[22] C.J. Brinker, G.W. Scherer, Sol–Gel Science, The Physics and Chemistry of Sol Gel
Processing, Academic Press, Inc., San Diego, CA, USA, 1990.
[23] A.F. Nogueira, M.-A. De Paoli, I. Montanari, R. Monkhose, J. Nelson, J.R. Durrant,
J. Phys. Chem. B 105 (2001) 7517.
[24] A. Zaban, M. Greenshtein, J. Bisquert, Chemphyschem 4 (2003) 859.
[25] M. Quintana, T. Edvinsson, A. Hagfeldt, G. Boschloo, J. Phys. Chem. C 111 (2007)
1035.
[26] S. Rajendran, M. Sivakumar, R. Subadevi, Solid State Ionics 167 (2004) 335.
[27] S. Rajendran, M. Sivakumar, R. Subadevi, Mater. Lett. 58 (2004) 641.
[28] A.F. Nogueira, M.A.S. Spinacé, W.A. Gazotti, E.M. Girotto, M.A. De Paoli, Solid
State Ionics 140 (2001) 327.
Fig. 8. Electron lifetime (ꢄe) behavior under open-circuit condition for DSSCs based
[29] A.F. Nogueira, M.A. De Paoli, Sol. Energy Mater. Sol. Cells 61 (2000) 135.
on TiO2 and TiO2/core-shell electrodes.

1474
C.O. Avellaneda et al. / Electrochimica Acta 55 (2010) 1468–1474
[30] W. Kubo, K. Murakoshi, K. Kitamura, Y. Yoshida, M. Haruki, H. Hanabusa, H.
Shirai, Y. Wada, S. Yanagida, J. Phys. Chem. B 105 (2001) 12809.
[31] N. Papageorgiou, T. Athanassov, M. Armand, P. Bonhote, H. Pettersson, A. Azam,
M. Gratzel, J. Electrochem. Soc. 143 (1996) 3099.
[33] M. Law, L.E. Greene, A. Radenovic, T. Kuykendal, J. Liphardt, P. Yang, J. Phys.
Chem. B 110 (2006) 22652.
[34] Y. Diamant, S. Chappel, S.G. Chen, O. Melamed, A. Zaban, Coord. Chem. Rev. 248
(2004) 1271.
[32] J.N. Freitas, A.D. Gonc¸ alves, M.-A. De Paoli, J.R. Durrant, A.F. Nogueira, Elec-
trochim. Acta 53 (2008) 7166.