Charge transporting enhancement of NiO photocathodes for p-type dye-sensitized solar cells

Article abstract of DOI:10.1016/j.electacta.2012.01.081

A p-type NiO film was prepared by doctor-blading of Ni(OH)₂ paste onto FTO glass, followed by sintering at 450 °C for 30 min. The influence of nucleation condition of Ni(OH)₂ on the morphology of NiO films and consequently charge transporting behavior is investigated. A smooth and compact NiO film is obtained with smaller Ni(OH)₂ sol-gel particle size, which is confirmed by the surface topography examination. The hole transporting ability of NiO semiconductor is enhanced with such a compact film because of better interconnection between particles, as evidenced from electrochemical impedance analysis. The NiO film obtained is employed as the photocathode in p-type dye-sensitized solar cells (DSSCs) using arylamine-based dyes. The short-circuit photocurrent is two times improved to ca. 2.0 mA cm ^-2 compared to rough films.

Full text of DOI:10.1016/j.electacta.2012.01.081

Electrochimica Acta 66 (2012) 210–215
Contents lists available at SciVerse ScienceDirect
Electrochimica Acta
journal homepage: www.elsevier.com/locate/electacta
Charge transporting enhancement of NiO photocathodes for p-type
dye-sensitized solar cells
Chih-Yu Hsua,1, Wei-Ting Chena, Yung-Chung Chena, Hung-Yu Weib, Yung-Sheng Yena,
Kuan-Chieh Huangc, Kuo-Chuan Hob,c,1, Chih-Wei Chud, Jiann T. Lina,∗
a Institute of Chemistry, Academia Sinica, Taipei 11529, Taiwan
b Institute of Polymer Science and Engineering, National Taiwan University, Taipei 10617, Taiwan
c Department of Engineering, National Taiwan University, Taipei 10617, Taiwan
d Research Center for Applied Sciences, Academia Sinica, Taipei 11529, Taiwan
a r t i c l e i n f o
a b s t r a c t
Article history:
A p-type NiO film was prepared by doctor-blading of Ni(OH)2 paste onto FTO glass, followed by sintering
at 450 ◦C for 30 min. The influence of nucleation condition of Ni(OH)2 on the morphology of NiO films and
consequently charge transporting behavior is investigated. A smooth and compact NiO film is obtained
with smaller Ni(OH)2 sol–gel particle size, which is confirmed by the surface topography examination.
The hole transporting ability of NiO semiconductor is enhanced with such a compact film because of better
interconnection between particles, as evidenced from electrochemical impedance analysis. The NiO film
obtained is employed as the photocathode in p-type dye-sensitized solar cells (DSSCs) using arylamine-
based dyes. The short-circuit photocurrent is two times improved to ca. 2.0 mA cm−2 compared to rough
films.
Received 16 November 2011
Received in revised form 16 January 2012
Accepted 22 January 2012
Available online 1 February 2012
Keywords:
Dye-sensitized solar cell (DSSC)
Nickel oxide
P-type semiconductor
Electrochemical impedance spectroscopy
Organic sensitizer
© 2012 Elsevier Ltd. All rights reserved.
1. Introduction
for photoanodes. On the contrary, p-type semiconductor NiO [9] is
studied intensively for photocathodes in DSSCs. The working mech-
Since the seminal report on solar-energy-to-electricity conver-
sion by dye-sensitized solar cells (DSSCs) in 1991 [1], the sprung up
literatures in the past two decades have witnessed the promising
feature and even commercialization of DSSCs. DSSCs can con-
vert environmentally friendly solar energy at promising efficiency
exceeding 11% [2,3] with low manufacturing costs and are very
competitive with silicon-based solar cell [4–6]. A DSSC consists of
a counter electrode, electrolyte layer, and a photoelectrode with
anchoring dye molecules that harvest light. The dye-generated
charge, i.e., electron or hole, injects into the photoelectrode and
transports to the outer circuit for electric power supply. Therefore,
efficient conductivity, sufficient dye-adsorbing sites and matched
energy levels with dyes are the criteria for photoelectrodes in
DSSCs.
anisms of these two types of photoelectrodes are different: the
photoexcited dye injects an electron into the conduction band of
a n-type semiconductor from its LUMO (lowest unoccupied molec-
ular orbital), while a hole at the HOMO (highest occupied molecular
orbital) is injected into the valence band of a p-type semiconduc-
tor. Up to date, the gained efficiencies of tandem DSSCs compared
to single cells are still limited by the NiO electrodes despite the
predicted high efficiency of 43% [15]. To enhance the efficiency
of p-type DSSCs for practical application in tandem DSSCs [9,16],
improving the hole injection and hole mobility on NiO and prevent-
ing intensive charge recombination with redox couple are required.
NiO was first applied on p-type DSSC in 1999 by Lindquist and
co-workers [17]. Nickel acetate was hydrolyzed to form Ni(OH)2
which was used as the precursor slurry to obtain NiO after high tem-
perature sintering. Afterwards, there were attempts to increase the
hole mobility or dye adsorbing sites of NiO for improving photocur-
rent. Sol–gel Ni(OH)2 was also prepared from nickel chloride for
better Ni(OH)2 particle quality [18–21]. Later, additives were added
during the sol–gel process, such as polymers [22,23] for increas-
ing surface area to porous nanostructure film or graphene [24]
for assisting conduction. Among them, addition of F108 triblock
copolymer turns out to be a popular route, and good photocur-
rent output in p-type DSSCs can be achieved [23,25–29]. On the
A tandem DSSC with two complementary photoelectrodes in
electric circuit is designed for improving light harvest using dyes
absorbing in complementary spectral region [7–9]. TiO2 [1–4] and
ZnO [10–14] are excellent candidates as n-type semiconductors
∗
Corresponding author. Tel.: +886 2 2789 8522; fax: +886 2 2783 1237.
E-mail address: jtlin@chem.sinica.edu.tw (J.T. Lin).
ISE Member.
1
0013-4686/$ – see front matter © 2012 Elsevier Ltd. All rights reserved.
doi:10.1016/j.electacta.2012.01.081

C.-Y. Hsu et al. / Electrochimica Acta 66 (2012) 210–215
211
Fig. 1. Molecular structures of dye 1 and 2.
other hand, strategy towards improvement of cell photovoltage
2.2. Device fabrication and measurement of DSSCs
by introduction of an additional layer to retard hole recombina-
tion with redox species was reported [30–32]. There are also few
reports [8,33] on commercial NiO nanopowders. Accordingly, the
photovoltaic performance of mesoporous NiO film depends on the
properties of NiO or Ni(OH)2 particles. Studies [34–36] on Ni(OH)2
have shown improved electrochemical properties with a smaller
crystalline size and more defects, which can be controlled by the
preparation condition.
In this paper we studied the effect of particle size of Ni(OH)2,
the precursor of NiO, on NiO by varying the preparation condi-
tions of Ni(OH)2 particles via sol–gel process. This was paid less
attention in the literature of p-type DSSCs. A substantial increase
in photocurrent of the p-type DSSC can be achieved by reducing the
particle size, as evidenced by the promising results using two novel
arylamine-based sensitizers we developed recently (Fig. 1) [37].
The corresponding electrochemical properties and performance of
p-type DSSCs will be also discussed.
The sintered NiO electrode was immersed in a 0.3 mM dye
solution in THF for 16 h, where the synthetic route of the dyes is
described elsewhere [37]. The counter electrodes were prepared
by coating the FTO plate with H2PtCl6 solution (2 mM in ethanol),
followed by heating at 400 ◦C for 15 min. The cell was assembled
with the two electrodes separated by the electrolyte composed of
1.0 M of lithium iodide (LiI), 0.1 M of iodine (I2), and 0.5 M of 4-
tert-butylpyridine in acetonitrile and an adhered polyester tape
(3 M) with thickness of 60 ␮m. The cell had an active area of
0.25 cm2. The cell was sealing with the Torr Seal® cement (Var-
ian, MA, USA). The DSSCs were energized under an Oriel Class A
solar simulator (Oriel 91195A, Newport Corp.) under simulated
light illumination. Light intensity at the measuring (cell) position
was controlled at 1.0 sun upon calibration from an Oriel refer-
ence solar cell (Oriel 91150, Newport Corp.). Photoelectrochemical
characteristics and electrochemical impedance spectra (EIS) of the
NiO films or DSSCs were recorded with a potentiostat/galvanostat
(CHI650B, CH Instruments, Inc., USA). The frequency range explored
in impedance measurements was 10 mHz to 100 kHz with sinu-
soidal signal (single sine) and an ac amplitude (ꢀEac) of 10 mV.
The applied bias voltage was set at the open-circuit voltage of the
DSSC under illumination, between the Pt counter electrode and the
TiO2-dye working electrode. As for the NiO electrodes tested in an
acetonitrile solution of LiClO4 (0.1 M), 0.0 V (vs. Ag/Ag+) was set in
a typical three-electrode electrochemical system. The impedance
spectra were fitted to an equivalent circuit model interpreting mod-
ified electrodes [38] and DSSCs [39,40], and the model parameters
were obtained by ZView software.
2. Experimental
2.1. Preparation and characterization of NiO electrodes
The Ni(OH)2 paste was prepared according to the commonly
used sol–gel process in literatures [18–21] with some modifi-
cations. A precursor solution was prepared by adding 45 mL of
aqueous solution of NaOH (1.0 M) into a 100 mL of aqueous solution
of NiCl2 (0.2 M) with vigorous stirring. The addition of NaOH(aq.)
is controlled at two extreme rates, 50 and 0.025 mL s−1, namely
rapid and slow, respectively, for easy realization. The morphologies
of the resulting Ni(OH)2 particles were checked by a transmis-
sion electron microscopy (TEM, JOEL JEM-1230). This solution
was kept static overnight for precipitation. After removing the
clear solution part, the paste was then centrifuged at 6000 rpm
for 5 min and repeated for 3 times. Ethanol (30 wt.% vs. gel par-
ticles) was then added to the gel particles. Finally, ten drops
(ca. 50 ␮L per drop) of acetic acid were added into the paste to
finalize the preparation of precursor paste. The NiO films were
prepared by doctor-blading the paste onto fluorine-doped SnO2
(FTO) glass, and then sintered at 450 ◦C for 30 min. The film
thickness was measured by a profilometer (Dektak3, Veeco/Sloan
Instruments Inc., USA). The surface morphologies of the NiO films
were investigated using atomic force microscopy (AFM, Digi-
tal Instrument NS 3a controller equipped with a D3100 stage)
and scanning electron microscopy (SEM, Hitachi S-4700). X-ray
diffraction (XRD) and X-ray photoelectron spectroscopy (XPS) were
performed using a Philips X’Pert/MPD instrument (Cu K␣ radiation)
and a PHI 5000 VersaProbe (ULVAC-PHI, Chigasaki, Japan) system,
respectively.
3. Results and discussion
3.1. Ni(OH)2 particles
As mention in Section 2, a precursor solution was firstly pre-
pared by adding 45 mL of aqueous solution of NaOH (1.0 M) into
a 100 mL of aqueous solution of NiCl2 (0.2 M) with vigorous stir-
ring. The pH variation during nucleation of Ni(OH)2 resulted in
structural change [36], thus the addition rate of alkaline solution
was varied and checked. There was obvious difference in the TEM
images (Fig. 2a and b) of Ni(OH)2 sol–gel particles prepared from
different rates. A distribution of particle size from 40 to 200 nm was
found with a slow addition rate, while larger particles around 100 to
200 nm were observed in the rapid addition case. The particle size
distribution is shown in Fig. 2c. It is known for a sol–gel process of
metal oxides [41], condensation is facilitated in base environment
due to lower surface charge and less repulsion between particles
that leads to particle growth and aggregation. Thus, the particle size
of Ni(OH)2 is pH dependent and the slow addition of NaOH(aq.) may

212
C.-Y. Hsu et al. / Electrochimica Acta 66 (2012) 210–215
Fig. 2. TEM images of Ni(OH)2 sol–gel particles prepared with (a) rapid and (b) slow addition rates and (c) the corresponding particle size distribution.
allow smaller Ni(OH)2 particles. Besides, the nucleus can be clearly
seen at the center of each particles from TEM images since radius
growth of Ni(OH)2 is quick and thus loose [42,43]. The particles
may also aggregate from the loose structure as indicated in the ca.
470 nm peak in Fig. 2c for the rapid addition case.
NiO particles owing to the loose structure of the Ni(OH)2 precur-
sor. The stacked Ni(OH)2 with loose periphery tends to aggregate
during the pyrolysis process at high temperature sintering. There-
fore, the grain particle size of NiO is larger for the rapid film (ca.
250 nm) than that for the slow film (ca. 100 nm), as evidenced from
the AFM images. The roughness of these two films is also differ-
ent because of different particle packing, and the root mean square
roughness (Rrms) obtained from the AFM is 89.9 and 22.3 nm from
the rapid and slow films, respectively. Thus, more smooth NiO film
is obtained from relatively more compact packed film of smaller
Ni(OH)2 particles formed by a slow addition rate during the sol–gel
preparation.
Fig. 5 displays XRD patterns of the rapid and slow NiO films
along with a bare FTO glass as a reference. The result indicates the
formation of crystalline cubic structured NiO covered on the FTO
conducting glass according to (1 1 1), (2 0 0) and (2 2 0) diffractions
[18]. In agreement with the morphologies observed by electron
microscopes (vide supra), the peaks assigned to FTO (gray squares
in Fig. 5) are reduced in the slow film due to the more compact
coverage of NiO. The average crystalline sizes of NiO estimated
by Scherrer equation are 19 and 16 nm for rapid and slow films,
respectively. The crystalline size is in the same range with those
studies prepared by sol–gel process in other studies [23,36]. The
different Ni(OH)2 particle size did not greatly affect the lattice size
after transforming to NiO polycrystals, i.e., the crystallinity prop-
erty is still identical. It implies that the different grain particle size
observed from AFM is due to the aggregation of Ni(OH)2 particles
instead of the structure change after transformation from Ni(OH)2
to NiO.
3.2. NiO films
The Ni(OH)2 paste was used to prepare NiO films on FTO glasses,
herein rapid and slow films represent rapid and slow addition of
NaOH solution during the formation of Ni(OH)2 paste. Both films
have an average film thickness of about 1.5 ␮m from the cross-
sectional measurement using a profilometer as shown in Fig. 3. The
film is smoother in the case of slow addition of alkaline solution as it
was formed from stacking of smaller Ni(OH)2 particles. Apparently
the surface morphology of NiO films is controlled by the particle
size of Ni(OH)2. The morphology difference can be recognized from
their appearance. Rough film with larger particle size leads to inter-
ference of the incident light, i.e., scattering effect. Accordingly, the
rapid film looks less pervious to light and non-uniform compared
to the slow film, as shown in the insets of Fig. 3. The micro-scale
morphology is verified by SEM and AFM as shown in Fig. 4. The
SEM image of rapid film reveals prominent aggregation between
XPS spectra in Fig. 6 reveal the chemical composition of the
NiO/FTO layers. The peak assigned as Sn 3d, contributed from FTO,
disappeared in the slow film owing to the same reason described
in the XRD part. The detection depth of XPS is about 20 angstrom
while a compact coverage of NiO is about 1.5 ␮m. According to ear-
lier report on NiO [44,45], the signal of Ni 2p3/2 is stemmed from
Ni2+ (854.3 eV) and Ni3+ (857.4 eV) since both oxidized states coex-
ist after oxidation of Ni metal. The peaks of our NiO films confirm
the coexistence feature and the content ratio of these two oxidized
states can be determined after Gaussian deconvolution (R2 = 0.926
and 0.920 for rapid and slow films, respectively) as shown from
the dotted lines in the inset of Fig. 6. The non-stoichiometric
Fig. 3. Profilometer results of NiO films on FTO glasses prepared with rapid and slow
addition paste. Insets show corresponding photograph.

C.-Y. Hsu et al. / Electrochimica Acta 66 (2012) 210–215
213
Fig. 4. Representative images of the surface morphologies of the NiO films. (a, b) SEM images of (a) rapid and (b) slow films. (c, d) Corresponding AFM images scanned in
10 ␮m × 10 ␮m area.
composition generally known as Ni1−xO [46] is calculated accord-
ingly and x equals to 0.148 and 0.159 for rapid and slow films,
respectively. For a p-type metal oxide semiconductor, it is known
that the metal (cation) vacancy acts as hole hopping site. Thus, there
are more nickel vacancies in the slow film according to x values,
inferring a better hole transporting ability. Besides, the amounts
of other cations, specifically Na+ if any, was under detecting limit
of XPS after washing Ni(OH)2 with deionized H2O for 3 times (see
Section 2).
The charge transporting ability of the NiO electrodes can be
evaluated by the electrochemical impedance analysis at the high
frequency range interpreting the charge transfer control region
(Fig. 7). The rate constant of the charge transfer (k0) can be obtained
by the following equation: k0 = RT/n2F2RctAC [38,47], where A is the
area of the electrode, C is the bulk concentration of the ion and Rct
is the charge transfer resistance at the NiO layer obtained by fit-
ting the EIS data to the equivalent circuit as shown in the inset of
Fig. 7. The other symbols have their usual electrochemical mean-
ings. The fitted Rct values are 8274 and 2999 ꢁ for rapid and slow
cases, respectively. The values of k0 were consequently calculated
as 3.22 × 10−7 cm s−1 for the rapid film and 8.87 × 10−7cm s−1 for
the slow film, respectively. The charge (hole) transporting ability
for the slow NiO film is indeed superior to that for the rapid film.
We believe that the pronounced enhancement mainly results from
the compact packing which facilitates hole hopping through the
nickel vacancy.
Fig. 6. XPS spectra of rapid and slow NiO films along with a bare FTO glass as a
reference. The inset shows the Gaussian deconvolution results (dotted lines) of Ni
2p3/2 peaks.
Fig. 5. XRD patterns of rapid and slow NiO films along with a bare FTO glass as a
reference. Peaks of NiO are labeled as white circle while peaks of FTO, labeled as
gray squares, come from the substrate.

214
C.-Y. Hsu et al. / Electrochimica Acta 66 (2012) 210–215
Table 1
Summary of the photovoltaic performance results.
Dye
NiO photocathodes
JSC (mA cm−2
)
VOC (V)
FF
ꢂ (%)
Rct2 (ꢁ)
C
␮2
(mF)
ꢃR (ms)
1
Rapid
Slow
0.93
2.18
0.118
0.123
0.34
0.35
0.038
0.092
259
156
0.58
0.94
151
146
2
Rapid
Slow
0.84
2.05
0.117
0.131
0.32
0.32
0.032
0.087
219
129
0.60
0.94
132
121
transporting ability within the NiO reflects on the photocurrent.
The short-circuit photocurrent of DSSCs using slow films is about
two times of those using rapid films (0.94 to 2.18 mA cm−2 for dye
1 and 0.84–2.05 mA cm−2 for dye 2, respectively). It is in confor-
mity with the charge transporting ability (k0) analyses based on
EIS studies (vide supra). A slight increase in the photovoltage of ca.
5 to 10 mV for DSSCs using slow NiO film was observed. Probably
the compact NiO morphology reduced the contact between the hole
and the redox mediator that changes the chemical potential of NiO.
As a result, the energy conversion efficiencies enhanced from 0.038
to 0.092% for dye 1 and 0.032 to 0.087% for dye 2, respectively.
We further applied EIS technique to investigate the p-type DSSC
under 1 sunlight illumination as shown in Fig. 8b. EIS is a useful
technique for the analysis of charge and ionic transport processes
in an electrochemical device including p-type DSSC [28]. Three
semicircles are typically displayed for a DCCS structure, which are
assigned to electrochemical reaction at the Pt/electrolyte (high fre-
quency region), charge transfer at the semiconductor-dye electrode
(medium frequency region) and Warburg diffusion in the elec-
trolyte (low frequency region), as interpreted by Bisquert [48,49].
In our case, the Warburg diffusion is not obvious and negligible
since we used very thin spacer (60 ␮m) and low viscous sol-
vent (0.37 cP for acetonitrile). The equivalent circuit (shown in
the inset of Fig. 8b) was employed to fit the impedance spectra
of the DSSCs. Meanwhile, the charge recombination lifetime (ꢃR)
Fig. 7. EIS analysis of NiO electrodes. The inset shows the equivalent circuit describ-
ing thin-film electrodes.
3.3. DSSC performances
The corresponding two NiO films were then used as photo-
cathodes sensitized by two novel organic dyes for a p-type DSSC
and the photovoltaic J–V characteristic and EIS were analyzed. The
results are shown in Fig. 8 with summarizing data in Table 1. A con-
spicuous difference on the photocurrent in Fig. 8a was observed
for both dyes adsorbed on rapid and slow NiO films. The charge
can be obtained by ꢃR = Rct2 × C
where Rct2 and C
represent
␮2
␮2
the charge transfer resistance and the interfacial capacitance at the
NiO/dye/electrolyte interface, respectively [50]. The parameters are
listed in Table 1. It was found the Rct2 values for the DSSCs using
slow films (dye 1: 156 ꢁ; dye 2: 129 ꢁ) were indeed smaller than
those using rapid films (dye 1: 259 ꢁ; dye 2: 219 ꢁ) due to the
enhanced charge transfer ability. While the interfacial capacitance
for the slow films of ca. 0.94 mF is higher than that for the rapid
films of ca. 0.60 mF. Likely the smaller grain particle size of the slow
NiO film provide more interfaces between NiO particles, and con-
sequently increment of the double layer capacitance. Besides, the
ꢃR value for the slow film is only slightly smaller than that of the
rapid film. The photocurrent improvement is more likely resulted
from the enhanced charge transfer ability. Accordingly, synthetic
procedure of Ni(OH)2, the precursor paste of NiO plays an impor-
tant role for the quality of the NiO photocathode and therefore the
performance of the DSSCs.
4. Conclusions
The growth of Ni(OH)2 particles depends on the addition rate
during sol–gel preparation: slow addition of NaOH(aq.) into NiCl2(aq.)
leads to smaller particle sizes due to pH environment or local con-
centration of reactants during nucleation. The NiO films prepared
from doctorblading Ni(OH)2 sol–gel paste followed by a sintering
process exhibit smooth and compact morphology for slow addi-
tion case. The charge transfer rate constant (k0) of slow NiO film
is 8.87 × 10−7 cm s−1 which is about two times higher than that
for the rapid addition case of 3.22 × 10−7 cm s−1. The enhancement
in the charge transporting ability is attributed to relative compact
packing of NiO films. When the NiO films were used as the pho-
tocathode of p-type DSSCs and tested with two different organic
Fig. 8. (a) Photovoltaic J–V characteristics of p-type DSSCs and (b) the corresponding
EIS analysis results with the equivalent circuit describing for DSSCs (inset).

C.-Y. Hsu et al. / Electrochimica Acta 66 (2012) 210–215
215
dyes, the short-circuit photocurrents of the slow NiO films were
about two times improvement compared to the fast films due to
smaller charge transfer resistance of the slow NiO films. Efforts on
improving the hole transporting behavior of NiO should be made
for enhancing the power output performance for p-type DSSCs.
[19] A. Morandeira, G. Boschloo, A. Hagfeldt, L. Hammarström, J. Phys. Chem. B 109
(2005) 19403.
[20] P. Qin, H. Zhu, T. Edvinsson, G. Boschloo, A. Hagfeldt, L. Sun, J. Am. Chem. Soc.
130 (2008) 8570.
[21] P. Qin, M. Linder, T. Brinck, G. Boschloo, A. Hagfeldt, L. Sun, Adv. Mater. 21 (2009)
2993.
[22] A. Nakasa, H. Usami, S. Sumikura, S. Hasegawa, T. Koyama, E. Suzuki, Chem.
Lett. 34 (2005) 500.
Acknowledgements
[23] S. Sumikura, S. Mori, S. Shimizu, H. Usami, E. Suzuki, J. Photochem. Photobiol.
A: Chem. 199 (2008) 1.
[24] H. Yang, G.H. Guai, G. Guo, Q. Song, S.P. Jiang, Y. Wang, W. Zhang, C.M. Li, J. Phys.
Chem. C 115 (2011) 12209.
[25] L. Li, E.A. Gibson, P. Qin, G. Boschloo, M. Gorlov, A. Hagfeldt, L. Sun, Adv. Mater.
22 (2010) 1759.
[26] P. Qin, J. Wiberg, E.A. Gibson, M. Linder, L. Li, T. Brinck, A. Hagfeldt, B. Albinsson,
L. Sun, J. Phys. Chem. C 114 (2010) 4738.
[27] E.A. Gibson, A.L. Smeigh, L. Le Pleux, L. Hammarström, F. Odobel, G. Boschloo,
A. Hagfeldt, J. Phys. Chem. C 115 (2011) 9772.
[28] Z. Ji, G. Natu, Z. Huang, Y. Wu, Energy Environ. Sci. 4 (2011) 2818.
[29] L. Le Pleux, A.L. Smeigh, E.A. Gibson, Y. Pellegrin, E. Blart, G. Boschloo, A.
Hagfeldt, L. Hammarström, F. Odobel, Energy Environ. Sci. 4 (2011) 2075.
[30] S. Uehara, S. Sumikura, E. Suzuki, S. Mori, Energy Environ. Sci. 3 (2010) 641.
[31] X.L. Zhang, F. Huang, A. Nattestad, K. Wang, D. Fu, A. Mishra, P. Bäuerle, U. Bach,
Y.B. Cheng, Chem. Commun. 47 (2011) 4808.
[32] Z. Bian, T. Tachikawa, S.C. Cui, M. Fujitsuka, T. Majima, Chem. Sci. 3 (2012) 370.
[33] A. Nattestad, M. Ferguson, R. Kerr, Y.B. Cheng, U. Bach, Nanotechnology 19
(2008) 295304.
[34] F.S. Cai, G.Y. Zhang, J. Chen, X.L. Gou, H.K. Liu, S.X. Dou, Angew. Chem. Int. Ed.
43 (2004) 4212.
[35] R.S. Jayashree, P.V. Kamath, G.N. Subbanna, J. Electrochem. Soc. 147 (2000)
2029.
[36] Q.S. Song, Z.Y. Tang, H.T. Guo, S.L.I. Chan, J. Power Sources 112 (2002) 428.
[37] Y.S. Yen, W.T. Chen, C.Y. Hsu, H.H. Chou, J.T. Lin, M.C.P. Yeh, Org. Lett. 13 (2011)
4930.
[38] F. Sundfors, J. Bobacka, A. Ivaska, A. Lewenstam, Electrochim. Acta 47 (2002)
2245.
[39] L.Y. Han, N. Koide, Y. Chiba, T. Mitate, Appl. Phys. Lett. 84 (2004) 2433.
[40] J.Y. Han, N. Koide, Y. Chiba, A. Islam, T. Mitate, Comptes Rendus Chim. 9 (2006)
645.
[41] C.J. Brinker, G.W. Scherer, Sol–Gel Science: The Physics: Chemistry of Sol–Gel
Processing, Academic Press, Boston, 1990.
[42] H. Orikasa, J. Karoji, K. Matsui, T. Kyotani, Dalton Trans. (2007) 3757.
[43] M.X. Peng, X.Q. Shen, J. Cent. South Univ. Technol. 14 (2007) 310.
[44] C.D. Wagner, W.M. Riggs, L.E. Davis, J.F. Moulder, G.E. Muilengerg, Handbook
of X-ray Photoelectron Spectroscopy, Perkin-Elmer, Minnesota, 1979, p. 80.
[45] I. Hotovy´, J. Huran, J. Janík, A.P. Kobzev, Vacuum 51 (1998) 157.
[46] K. Nakaoka, J. Ueyama, K. Ogura, J. Electroanal. Chem. 571 (2004) 93.
[47] A.J. Bard, L.R. Faulkner, Electrochemical Methods: Fundamentals and Applica-
tions, 2nd ed., Wiley, New York, 2001.
We appreciate kind assistance from Mr. Chun-Chieh Wang at
Department of Chemical Engineering, National Taiwan University
during this work. We would like to acknowledge the Institute of
Chemistry, Academia Sinica (AC), the Instrumental Center of Insti-
tute of Chemistry (AC), and National Science Council, Taiwan for
financial support.
References
[1] B. O’Regan, M. Grätzel, Nature 353 (1991) 737.
[2] F. Gao, Y. Wang, D. Shi, J. Zhang, M. Wang, X. Jing, R. Humphry-Baker, P. Wang,
S.M. Zakeeruddin, M. Grätzel, J. Am. Chem. Soc. 130 (2008) 10720.
[3] C.Y. Chen, M. Wang, J.Y. Li, N. Pootrakulchote, L. Albabaei, C.-h Ngoc-le, J.D.
Decoppet, J.H. Tsai, C. Grätzel, C.G. Wu, S.M. Zakeeruddin, M. Grätzel, ACS Nano
3 (2009) 3103.
[4] J.M. Kroon, N.J. Bakker, H.J.P. Smit, P. Liska, K.R. Thampi, P. Wang, S.M. Zakeerud-
din, M. Grätzel, A. Hinsch, S. Hore, U. Wurfel, R. Sastrawan, J.R. Durrant, E.
Palomares, H. Pettersson, T. Gruszecki, J. Walter, K. Skupien, G.E. Tulloch, Prog.
Photovoltaics 15 (2007) 1.
[5] M. Grätzel, Nature 414 (2001) 338.
[6] M. Grätzel, Prog. Photovoltaics 8 (2000) 171.
[7] E.A. Gibson, A.L. Smeigh, L. Le Pleux, J. Fortage, G. Boschloo, E. Blart, Y. Pelle-
grin, F. Odobel, A. Hagfeldt, L. Hammarström, Angew. Chem. Int. Ed. 48 (2009)
4402.
[8] A. Nattestad, A.J. Mozer, M.K.R. Fischer, Y.B. Cheng, A. Mishra, P. Bäuerle, U.
Bach, Nat. Mater. 9 (2010) 31.
[9] F. Odobel, L. Le Pleux, Y. Pellegrin, E. Blart, Acc. Chem. Res. 43 (2010) 1063.
[10] C. Xu, J. Wu, U.V. Desai, D.J. Gao, Am. Chem. Soc. 133 (2011) 8122.
[11] Q.R. Zhang, T.R. Chou, B. Russo, S.A. Jenekhe, G.Z. Cao, Angew. Chem. Int. Ed. 47
(2008) 2402.
[12] M. Law, L.E. Greene, J.C. Johnson, R. Saykally, P.D. Yang, Nat. Mater. 4 (2005)
455.
[13] H.M. Cheng, W.H. Chiu, C.H. Lee, S.Y. Tsai, W.F. Hsieh, J. Phys. Chem. C 112 (2008)
16359.
[14] Y.Z. Zheng, X. Tao, L.X. Wang, H. Xu, Q. Hou, W.L. Zhou, J.F. Chen, Chem. Mater.
22 (2010) 928.
[15] M.A. Green, Third Generation Photovoltaics: Advanced Solar Energy Conver-
sion, Springer-Verlag, Berlin, Heidelberg, 2003.
[16] A. Hagfeldt, G. Boschloo, L. Sun, L. Kloo, H. Pettersson, Chem. Rev. 110 (2010)
6595.
[48] J. Bisquert, J. Phys. Chem. B 106 (2002) 325.
[49] F. Fabregat-Santiago, J. Bisquert, G. Garcia-Belmonte, G. Boschloo, A. Hagfeldt,
Sol. Energy Mater. Sol. Cells 87 (2005) 117.
[50] K.M. Lee, Y.C. Hsu, M. Ikegami, T. Miyasaka, K.R.J. Thomas, J.T. Lin, K.C. Ho, J.
Power Sources 196 (2011) 2416.
[17] J. He, H. Lindström, A. Hagfeldt, S.E. Lindquist, J. Phys. Chem. B 103 (1999) 8940.
[18] G. Boschloo, A. Hagfeldt, J. Phys. Chem. B 105 (2001) 3039.