Electrochimica Acta 114 (2013) 318–324

Contents lists available at ScienceDirect

Electrochimica Acta
journal homepage: www.elsevier.com/locate/electacta

Thickness-dependent electron transport performance of mesoporous

TiO2 thin film for dye-sensitized solar cells
Wanyu Zhao, Hari Bala ∗ , Jingkuo Chen, Yujie Zhao, Guang Sun ∗ ,
Jianliang Cao, Zhanying Zhang
School of Materials Science and Engineering; Cultivating Base for Key Laboratory of Environment-Friendly Inorganic Materials in University of Henan
Province, Henan Polytechnic University, Jiaozuo 454000, People’s Republic of China

a r t i c l e i n f o a b s t r a c t

Article history: Anatase titania (TiO2 ) nanocrystals with an average size of 12 nm were synthesized by combining sol–gel
Received 8 June 2013 method and hydrothermal crystallization technology. Mesoporous TiO2 thin film electrodes (MTTFEs)
Received in revised form were prepared by variations in layer of screen-printed TiO2 nanocrystals pastes, and dye-sensitized solar
18 September 2013
cells (DSCs) based on different thickness of MTTFEs were fabricated with dye C106. Under the irradiation
Accepted 29 September 2013
of 100 mW cm−2 air mass 1.5 global (AM1.5G) sunlight, the short-circuit photocurrent density (Jsc ) and
Available online 19 October 2013
overall conversion efficiency () was improved when the thickness of MTTFEs increased from 2.1 ␮m to
9.8 ␮m, and obvious decreases of the Jsc and  were found when the film thickness was further increased
Keywords:
TiO2 nanocrystals
over 9.8 ␮m. The kinetic parameters of electron transport were investigated by the transient photoelec-
Thin film electrode trical and electrical impedance measurements, and then the influence of MTTFEs thickness on electron
Thickness transport performance in DSCs were further discussed. The results showed that the density of states
Dye-sensitized solar cells (DOS) and recombination rate (Kr ) increased with the MTTFEs thickness increasing. Furthermore, when
Electron transport the thickness of MTTFEs was in the range of 2.1–9.8 ␮m, the electron lifetime ( n ), the electron diffu-
sion coefficient (Dn ), and the electron diffusion length (Ln ) of devices increased, but these parameters
decreased gradually with the further increased thickness of MTTFEs, which influenced the electronic
transport performance of DSCs.
© 2013 Elsevier Ltd. All rights reserved.

1. Introduction coated glass, an electrolyte containing redox couple (usually I− /I3 −

couple), and a platinized counter electrode. The semiconductor film
In recent years, developments in solar energy have led differ- electrode is the key part of DSC. Nanosized TiO2 materials have
ent countries to take a more affirmative attitude toward the future become the preferred semiconductor electrode material because
prospects of solar energy. However, the traditional silicon solar of their excellent physical and chemical properties [8,9]. When the
cells cannot be widely applied because of their high material cost device is exposed to sunlight, the excited electrons in the dye are
and unresolved core technology issues. Currently, third-generation injected into the conduction band of nanocrystalline TiO2 , then, the
solar cells have been developed [1]. Dye-sensitized solar cells electrons are transport to the bottom of the TiO2 film and collected
(DSCs) based on nanocrystalline TiO2 have attracted a lot of sci- by the FTO glass. Meanwhile, the dye in the oxidized state strips one
entific and technological interest since their (Grätzel and O’Regan) electron from the electrolyte, the electrolyte then recovers its miss-
breakthrough in 1991 [2], the innumerable advancements made ing electrons from the counter electrode by diffusion [10]. During
on this unique device has now achieved the overall conversion these processes, electrons in the conduction band of semiconduc-
efficiency () of 12.3% [3]. tor may be recombined with the oxided dye sensitizers or electron
DSCs have attracted considerable attention due to their reason- acceptor species in the electrolyte solution. The electron collection
able conversion efficiency, low production cost, simple fabrication efficiency in DSCs is determined by the competition between trans-
process and environment friendliness compared to silicon solar port of electron in the porous semiconductor and recombination of
cells [4–7]. Typically, a DSC is constructed with a dye-anchored electrons with I3 − ions in electrolyte. Obviously, the most efficient
metal oxide semiconductor supported on an F-doped SnO2 (FTO) DSCs are based on a mesoporous TiO2 film that serves as both a sub-
strate for the dye and the electron conductor [11]. This provides not
only large surface area for the adsorption of dyes for light harvest-
∗ Corresponding authors. Tel.: +86 0391 3986952. ing, but also a network for electron transport to the FTO glass then
E-mail addresses: [email protected] (H. Bala), [email protected] (G. Sun). to an external circuit. Thus the different parameters such as particle

0013-4686/$ – see front matter © 2013 Elsevier Ltd. All rights reserved.
http://dx.doi.org/10.1016/j.electacta.2013.09.165
 W. Zhao et al. / Electrochimica Acta 114 (2013) 318–324 319

size, thickness, the surface area of TiO2 film play important roles on K␣ target. The morphology and structure of as-prepared TiO2
the performance of DSCs. nanocrystals were investigated on a JEOM 2100 transmission
Up till now, various strategies have been developed to improve electron microscope (TEM). The samples for TEM analysis were pre-
the conversion efficiency of DSCs, such as changing the TiO2 crys- pared by dropping the final TiO2 colloid into ethanol solvent and
tal type, particles size, and treating TiO2 film with TiCl4 solution then spread it on copper grids coated with an amorphous carbon
[12–14], to improve the contact between the particles within film film.
and enhance the performance of electron transport; doped TiO2
with transition metal or rare earth element to change its energy 2.3. Preparation and characterization of MTTFEs
level structure [15,16]; coating with a layer of metal oxide and
rare earth element oxide on the TiO2 film surface to form core- MTTFEs were prepared by a screen-printing method on TiCl4
shell structure of the semiconductor or dielectric barrier layer to per-treated FTO glass. A layer of TiO2 paste was coated on the FTO
reduce the electron recombination, etc. [17,18]. However, the elec- glass, kept in a clean box for 5 min and then dried at 130 ◦ C for 6 min.
tron transport mechanism, the charge recombination mechanism This screen-printing procedure was repeated to get an appropriate
and influence factors in the nanocrystalline film region are not very thickness for the working electrode. Then the films were sintered
clear. The rapid electron transport in the mesoporous TiO2 film in air at 500 ◦ C for 30 min. After TiCl4 post-treatment, the MTTFEs
is important to ensure their efficient collection by the conducting were stored for fabricating devices. The thickness of TiO2 films were
substrate when in competition with the recombination processes observed by scanning electron microscope (SEM) (EVO18, Germany
[19]. Therefore, a series of micro processes that light induces elec- ZEISS) and XP-1 high resolution surface profiler (Ambios Technol-
tron transport in DSCs, such as charge separation, transport, and ogy, Inc., USA).
recombination, still need to be systemically understood. This will
help to optimize devices then design the DSC with better per- 2.4. Fabrication of dye-sensitized solar cells
formance and to pursuit the higher efficiency, resulting in laying
a foundation on the practical application of DSCs eventually. In After being heated at 500 ◦ C for 30 min and then cooled to 80 ◦ C,
this paper, TiO2 nanocrystals were synthesized by sol–gel method the electrodes were stained overnight by immersing MTTFEs into
and hydrothermal crystallization technique. Different thickness a dye solution containing C106 sensitizer [22] (300 ␮M) and cheno
MTTFEs-based DSCs were fabricated by using C106 as sensitizer. (300 ␮M) in a mixture of acetonitrile and tert-butanol (1/1, v/v). The
Electrical impedance and transient photoelectrical measurements liquid electrolyte was composed of 1.0 M DMII, 50 mM LiI, 30 mM
have been performed to study the kinetic issues of electron trans- I2 , 0.5 M TBP and 0.1 M GNCS in the solvent of acetonitrile. The FTO
port, and further research the influence of MTTFEs thickness on the glass was coated with the Pt catalyst using a drop of H2 PtCl6 in iso-
electron transport performance in DSCs. propyl alcohol solution, followed by heating at 450 ◦ C for 15 min.
The electrodes were separated by a Bynel (DuPont) hot-melt gas-
ket and sealed up by heating. The internal space was filled with
2. Experimental
the electrolyte using a vacuum-back-filling system. The electrolyte-
injecting hole on the counter electrode glass substrate, made with
2.1. Materials
a sandblasting drill, was sealed with a Bynel sheet and a thin glass
cover by heating.
C106 [NaRu(4,4 -bis(5-(hexylthio)thiophen-2-yl)-2,2 -bipyr-
idine)(4-carboxylic acid-4 -carboxylate-2,2 -bipyridine)(NCS)2 ]
2.5. Photovoltaic measurements
dye was obtained from research group of Prof. Wang. Ethyl
cellulose (EC), chloroplatinic acid (H2 PtCl6 ) and 3␣,7␣-dihyroxy-
Optical absorbance measurements were carried out on a TU-
5␤-cholic acid (Cheno) were purchased from Fluka. Guanidinium
1810PCS UV–vis spectrophotometer (Beijing Purkinje General Ins.
thiocyanate (GNCS), 1,3-dimethylimidazolium iodide (DMII), 4-
Co. Ltd., CHINA). A BOS-X-1000G solar simulator (Changchun
tert-butyl pyridine (TBP), LiI and I2 were purchased from Aldrich.
Bosheng, Com. Inc., CHINA) was used to give irradiance of
All other chemicals were purchased from Tianjin chemical reagent
100 mW cm−2 (the equivalent of one sun at AM1.5G, intensity
factory and used as received.
adjust by using PMA2100/PMA2144 Pyranometer (Solar Light Co.,
Inc., USA)) at the surface of testing cell. The current-voltage char-
2.2. Preparation and characterization of TiO2 nanocrystals and acteristics were investigated by applying external potential bias
pastes to the cell and measuring the dark current and photocurrent with
a Keithley model 2602A digital source meter (Keithley, USA). A
TiO2 nanocrystals were synthesized by a modified process testing device was covered with a metal mask with an aperture
[20,21]. In a typical procedure, a white precipitate was obtained via area of 0.16 cm2 . A transient photoelectrical decay technique was
the hydrolysis of acetic acid-modified butyl titanate in water. The applied to take a closer look at the charge recombination kinet-
prepared precipitate was then mixed with concentrated nitric acid ics. In addition, the devices were measured with electrochemical
and heated at 80 ◦ C for 75 min. The resultant colloid was transferred impedance spectrum in 1 mHz–100 kHz frequency range through
into a Teflon-lined autoclave and heated at 220 ◦ C for 18 h. After electrochemical workstation (IM6ex, Germany, Zahner).
being added 2.4 mL nitric acid, the cooled colloid was dispersed
ultrasonically, centrifuged to remove nitric acid, and finally washed 3. Results and discussion
with de-ionized water and ethanol three times to produce TiO2
nanocrystals. For the preparation of TiO2 pastes, two kinds of ethyl During the process of preparing MTTFEs for DSC, the TiO2
cellulose 4.5 g of EC (9–11 mPas) and 3.5 g of EC (45–55 mPas) pow- nanocrystals in fact underwent a two-step heat treatment at 500 ◦ C
ers and 64.9 g of terpineol were dissolved prior to usage in ethanol. for 30 min. Therefore, a similar heat treatment at 500 ◦ C for 1 h has
The ethanolic mixtures were firstly added to the as-prepared TiO2 been preformed on the as-prepared TiO2 . Fig. 1a shows the XRD
nanocrystlas with stirring and sonicating, and then the mixture pattern of the heat treated TiO2 . For comparison, the XRD pattern
solution was concentrated with a rotary-evaporator. of the original TiO2 sample without any heat treatment is showed in
The phase and purity were recorded by X-ray diffraction (XRD) Fig. 1b. In Fig. 1, all the diffraction peaks can be indexed as anatase
patterns via a Bruker D8 ADVANCE diffractometer with a Cu TiO2 phase (JCPDS file No.21-1272), and no characteristic peaks
320 W. Zhao et al. / Electrochimica Acta 114 (2013) 318–324

the high resolution surface profiler. Results of the surface profiler

test (Fig. 3d) were consistent with the observation of the film thick-
ness by SEM. In the following experiments, MTTFEs with one layer,
there layers, five layers and seven layers were selected to fabricate
DSCs, which are named device 1, device 3, device 5, and device 7,
respectively.
In the practical DSC, sensitizer functions, as the only light-
harvesting component, make contribution to charge generation.
The influence of the thickness of MTTFEs on dye adsorption was
investigated by measuring absorption spectra of the dye-sensitized
films to make a preliminary evaluation on their light-harvesting
capacity, in which the not-sensitized films with corresponding
thickness were used as reference. Fig. 4 shows the absorption spec-
tra of dye-sensitized MTTFEs with different thickness. The peaks
appeared near 550 nm correspond to the characteristic absorption
of C106 dye. The C106 sensitizer bestows an improved light absorp-
tion coefficient (molar extinction coefficient at 550 nm for C106 is
18.7 × 103 M−1 cm−1 ) [23] to the dye-sensitized TiO2 film, ensur-
Fig. 1. XRD patterns of as-prepared TiO2 nanocrystals: (a) before heat treatment,
ing better light-harvesting efficiency with the increasing of film
(b) after heat treatment at 500 ◦ Cfor 1 h.
thickness. From Fig. 4, it can be see that with the increasing of
film thickness, the intensity of the characteristic absorption band of
from other impurities were detected, indicating the formation of C106 dye increases obviously, and a red-shift was observed when
pure TiO2 phase in both samples. Moreover, it was found that the the thickness of MTTFEs was increased from 2.1 ␮m to 13.7 ␮m,
(1 0 1) peak for the heat treated TiO2 sample was obviously sharper indicating that the increasing film thickness can improve the light-
than that of the original TiO2 sample, demonstrating that the heat harvesting capacity of MTTFEs. It may be possible for the higher
treatment process improved the crystallinity of TiO2 . According to light harvesting efficiencies due to the larger internal surface area
Scherrer formula, the calculated average crystal sizes are 11.6 nm of films and the loading more amount of dye with increasing thick-
and 15.4 nm for original TiO2 and heat treated TiO2 , respectively. ness of MTTFEs, as well as the improved light absorbing and broad
TEM and HRTEM were employed to further investigate the scale of the solar spectrum induced by the enhancing light scatter-
morphology and structure of the heat treated TiO2 sample. From ing. However, the absorption curves show a decreasing rate of rise
the TEM image showed in Fig. 2a, it can be seen that the pre- as the thickness increases. The absorptivity ratio for C106 sensiti-
pared TiO2 nanocrystals have prismatic shapes, and their size range zers with the film thickness of 2.1 ␮m, 6.2 ␮m, 9.8 ␮m and 13.7 ␮m
is 6.4–17.6 nm, being consistent with the results given by XRD. is 1:2.65:3.54:3.96. Due to the increase of stacked layers and inner
The clear and regular lattice fringes in the corresponding HRTEM defects of films, the amount of dye loading is not strictly propor-
image (Fig. 2b) confirmed the highly crystalline nature of the TiO2 tional to the increasing thickness.
nanocrystals. The measured lattice fringe spacing is 0.35 nm, which The J–V characteristics for the DCSs fabricated by MTTFEs with
is found to correspond to (1 0 1) plane of anatase TiO2 . different thickness are presented in Fig. 5 for 100 mW cm−2 (AM
The thickness of TiO2 film was investigated by SEM. Fig. 3a–c 1.5G sunlight) illumination and in the dark, and the detailed pho-
shows the SEM images corresponding to the cross-section of meso- tovoltaic parameters are listed in Table 1. The overall conversion
porous TiO2 thin films with different thickness. Obviously, the efficiency () of the DSCs is given by  = (Jsc × Voc × FF)/Pin , where
increasing of printing layers results in the increase of film thick- Jsc , Voc , and FF are short-circuit current density, open-circuit volt-
ness. The thickness of the films with 1∼7 layers were measured age, and fill factor, respectively. When the film thickness of MTTFEs
to be about 2.1 ␮m, 4.2 ␮m, 6.2 ␮m, 7.9 ␮m, 9.8 ␮m, 11.8 ␮m, increased from 2.1 to 9.8 ␮m, an obvious increase of Jsc from 10.55
13.7 ␮m, respectively. The thickness of films was further tested by to 17.46 mA cm−2 was happened, resulting in a corresponding

Fig. 2. TEM (a) and HRTEM (b) images of TiO2 nanocrystals after heat treatment at 500 ◦ C for 1 h.
 W. Zhao et al. / Electrochimica Acta 114 (2013) 318–324 321

Fig. 3. SEM image of mesoporous TiO2 thin film electrodes with the 500 ◦ C heat treatment for 1 h: (a) one layer, (b) there layers, (c) five layers, and (d) profiles of MTTFEs
with different thickness.

improvement of  from 5.88% to 8.73%. Compared with device 1, Voc decreased gradually with the increasing of TiO2 film thickness,
the Jsc and  of device 5 were increased by 65.5% and 48.5%, respec- and the FF was trivial change. The surface states and trap states
tively. But the Jsc and  decreased to 16.21 mA cm−2 and 7.85%, of MTTFEs increased with the thickness of film increasing, many
respectively, when the MTTFEs thickness was further increased to TiO2 surface states can cut down the Fermi level [24], which
13.7 ␮m. Therefore, this could be mainly ascribed that when the made the Voc of devices decrease. With the increasing of MTTFEs
total surface area of TiO2 was increased, more dye molecules could thickness, the Jsc increases, the Voc and fill factor FF reduces, while
be adsorbed with the thickness of MTTFEs increasing, resulting the increasing rate of Jsc is greater than the decreasing rate of Voc
in the light harvesting increasing, the generation of more photo- and FF. So, when the thickness of MTTFEs is less than five layers,
electrons, and further increased the Jsc of devices. However, the the increasing thickness of MTTFEs can improve the . On the

Table 1
The photovoltaic performance parameters of the devices.

Printing layers Thickness of Short-circuit current Open-circuit voltage Fill factor (FF) Overall conversion
MTTFEs (d/␮m) density (Jsc /mA cm−2 ) (Voc /mV) efficiency (/%)

1 2.1 10.55 787.5 0.707 5.88

2 4.2 14.37 752.4 0.689 7.45
3 6.2 16.15 736.3 0.685 8.14
4 7.9 16.88 721.2 0.683 8.31
5 9.8 17.46 717.1 0.697 8.73
6 11.8 16.87 693.3 0.692 8.10
7 13.7 16.21 683.1 0.685 7.85
322 W. Zhao et al. / Electrochimica Acta 114 (2013) 318–324

Fig. 4. Absorption spectra of the dye-sensitized MTTFEs with different thickness by

using not-sensitized MTTFEs with the corresponding thickness as reference.

contrary, excessive film thickness of electrodes, such as more than

five layers, MTTFEs thickness with 11.8 ␮m and 13.7 ␮m, makes
decreased the photovoltaic performance.
Generally, the increasing thickness of MTTFEs can enhance the
light harvesting of the photo-anode and improve the photovoltaic
performance of devices, but which does not agree with above
results of J-V test. This indicates that the Jsc and  of devices do
not increase with the continuously increasing of film thickness. The
charge transport and recombination were further investigated by
transient photoelectrical measurements. As presented in Fig. 6a,
the chemical capacitances (C ) of MTTFEs with different thick-
ness increase exponentially along with the increase of Voc , which
was generated by applying a gradually enhanced light intensity. In
each case, a strong shift in the capacitance curve is apparent for
the device 7, its C is maximum at a certain Voc . The C is pro-
portional to the differential density of states (DOS) at Femi level
Fig. 6. Plots of (a) chemical capacitance, (b) density of states versus open-circuit
in the TiO2 particles. Fig. 6b shows the DOS calculated for these voltage of DSCs with different thickness MTTFEs.
films using DOS = 6.24 × 1018 × C /[d(1 − p)] [25], where C is the
chemical capacitance/cm2 , d is the thickness of the TiO2 film in cm,
With the Voc increasing, TiO2 /electrolyte interface recombination
and p is the porosity. The DOS also increased as index form with
rate constant increase as index form. The Kr of device 1 is mini-
the changing of Voc . Under the certain light, C is closely related
mum, Kr grows up with the MTTFEs thickness increasing, and the
with DOS under the conduction band, the chemical capacitance of
Kr of device 7 is maximum. For instance, when the Voc is close to the
devices are so large that the TiO2 surface states and trap states
device work Voc = 680 mV, the Kr of devices increased from 21.6 s−1
are increased. Fig. 7 shows the relationship of recombination rate
to 43.5 s−1 with the thickness of MTTFEs increased from 6.2 ␮m to
constant (Kr ) and Voc of the DSCs with different thickness MTTFEs.
13.7 ␮m, the increased rate is about 2 times. It is thus clear that

Fig. 5. J–V characteristics of DSCs are measured in the dark and under an irradiance Fig. 7. Plots of charge recombination rate constant versus open circuit voltage of
of 100 mW cm−2 AM1.5G sunlight (curves with hollow mark). DSCs with different thickness MTTFEs.
 W. Zhao et al. / Electrochimica Acta 114 (2013) 318–324 323

Fig. 8. Plots of (a) electronic lifetime, (b) electronic diffusion coefficient and (c) electronic diffusion length versus applied bias voltage of DSCs with different thickness MTTFEs.

the thickness of mesoporous TiO2 films has significant influence on The Ln reflects the electron transfer path before recombination, the
the recombination rate constant. Furthermore, the charge recom- greater the Ln is, the higher electronic collection efficiency is, and it
bination probability between the electrode surface and electrolyte is an important factor for DSCs performance. The Ln in DSC has been
grew with the increasing of film thickness, suggesting that the elec- estimated by measuring Dn and  n [28]. When the MTTFEs is too
tron transport efficiency would decrease when the film is too thick thick, the electronic diffusion is more difficult, a large number of
although the light harvesting enhances. The electrons captured by photogenerated electrons are recombinated without transferring
traps increased because the excited photoelectrons can not effec- to the collecting electrode, making the Ln and electronic collection
tively transfer to the external circuit and form the photocurrent in efficiency decrease, the  also reduces.
the internal film. Thus, the Jsc and  decreased, this is consistent
with the above analysis of J–V characteristics. 4. Conclusions
The electrochemical impedance spectra were also measured in
the frequency range of 1 mHz–100 kHz. A Z-view date processing In summary, TiO2 nanocrystals with an average particle size
software was applied to calculate the kinetic parameters of elec- of 12 nm were synthesized by combining technologies of sol–gel
tronic transport. Fig. 8 shows the relationship between electronic method and hydrothermal crystallization. MTTFEs with different
lifetime ( n ), electronic diffusion coefficient (Dn ) and electronic dif- thickness were prepared by screen-printed method. DSCs based on
fusion length (Ln ) of DSCs fabricated by differently thick MTTFEs MTTFEs with different thickness were fabricated with dye C106.
versus applied bias voltage (Vb ). It can be seen that under the same The Jsc and  of devices significantly improved through increasing
bias voltage, the values of  n and Dn increase with MTTFEs thickness the thickness of MTTFEs in the range of 2.1–9.8 ␮m. However, when
increasing from 1 to 5 layers (9.8 ␮m of thickness), but decrease the MTTFEs thickness was more than 9.8 ␮m, the values of Jsc and
when the thickness is more than five layers. The  n refers to the time  both decreased under the irradiation of 100 mW cm−2 air mass
interval that photogenerated electrons start from injection into 1.5 global (AM1.5G) sunlight. Further investigation on the kinetic
the TiO2 conduction band to transfer into the oxidation-reduction parameters of electron transport performance in DSCs was carried
system [26]. Due to the main composite reaction of the internal out using the transient photoelectrical and electrical impedance
device is the reaction between TiO2 conduction band electronic measurements. It was found that the DOS and Kr increased with
and the redox couple of electrolyte and the recombination rate is the increasing of MTTFEs thickness. Furthermore, when the thick-
related to the electron transition frequency among trap sites. Dn ness of MTTFEs were from 2.1 ␮m to 9.8 ␮m, the  n , Dn and Ln of
depends on the photogenerated electrons concentration gradient, devices increased, but these parameters decreased gradually with
and the probability that electrons are captured by trap states, etc., the increasing thickness of MTTFEs, which influenced the elec-
which is the result of comprehensive competition, it reflects the tronic transport performance of DSCs. In general, any factor that
difficulty level of electronic transmission at concentration gradi- has relationship with the electronic transport process, including the
ent [27]. When the thickness of MTTFEs is increased, more dye electronic generation, transmission and recombination, will affect
molecules can be adsorbed, resulting in the production of more the performance of DSCs. So, these results should be references
photogenerated electrons in the MTTFEs, subsequently, the elec- to the design of high performance DSCs and the optimization of
trons concentration gradient inside the film is increased. And there devices.
are a lot of trap states and surface states, electrons constantly expe-
rience trap capturing and taking off in transport process, MTTFEs Acknowledgments
of deep trap sites will be filled by injection electrons, shallow trap
states release the captured electrons quickly, electrons transition This work was supported by the National Natural Science
frequency increases among trap states. So the time that electrons Foundation of China (60877028, 51172065), China Postdoctoral
stay within trap states increases, which delay the recombination Science Foundation funded project (2012M521394), Program for
rate between electrons and I3 − , the  n and Dn increase. However, Innovative Research Team in the University of Henan Province
when the MTTFEs is too thick, trap states largely increase inside (2012IRTSTHN007), Foundation of Henan Scientific and Technol-
the film, the electronic transport paths lengthen, recombination ogy key project (132102210251, 112102310425, 112102310029),
rate increases rapidly, and the probability of trap capturing and the Education Department Natural Science Foundation of Henan
taking off electrons is increased, making the electronic diffusion Province (13A430315, 2010B420012, 2011B150009). We are grate-
more difficult, the reflected result is  n and Dn decrease gradually. ful to Professor Peng Wang (Changchun Institute of Applied
324 W. Zhao et al. / Electrochimica Acta 114 (2013) 318–324

Chemistry, Chinese Academy of Sciences) supported this work for [14] L.M. Peter, K.G.U. Wijayantha, Intensity dependence of the electron diffusion
Dye C106. length in dye-sensitised nanocrystalline TiO2 photovoltaic cells, Electrochem-
istry Communications 1 (1999) 576.
[15] S. Bingham, W.A. Daoud, Recent advances in making nano-sized TiO2 visible-
References light active through rare-earth metal doping, Journal of Materials Chemistry
21 (2011) 2041.
[1] H. Chang, H.T. Su, W.A. Chen, K.D. Huang, S.H. Chien, S.L. Chen, C.C. Chen, [16] Y. Zhang, L.L. Wang, B.K. Liu, J.L. Zhai, H.M. Fan, D.J. Wang, Y.H. Lin, T.F. Xie,
Fabrication of mulitilayer TiO2 thin films for dye-sensitized solar cells with Synthesis of Zn-doped TiO2 microspheres with enhanced photovoltaic perfor-
high conversion efficiency by eletrophoresis deposition, Solar Energy 84 (2010) mance and application for dye-sensitized solar cells, Electrochimica Acta 56
130. (2011) 6517.
[2] B. O’Regan, M. Grätzel, A low-cost, high efficiency solar cell based on dyesen- [17] H. Choi, W.T. Chen, P.V. Kamat, Know thy nano neighbor. Plasmonic versus
sitized colloidal TiO2 films, Nature 353 (1991) 737. electron charging effects of metal nanoparticles in dye-sensitized solar cells,
[3] A. Yella, H.W. Lee, H.N. Tsao, C. Yi, A.K. Chandiran, M.K. Nazeeruddin, E.W.G. ACS Nano 5 (2012) 4418.
Diau, C.Y. Yeh, S.M. Zakeeruddin, M. Grätzel, Porphyrin-sensitized solar cells [18] J. Qi, X. Dang, P.T. Hammond, A.M. Belcher, Highly efficient plasmon-enhanced
with cobalt (II/III)-based redox electrolyte exceed 12 percent efficiency, Science dye-sensitized solar cells through metal@ oxide core-shell nanostructure, ACS
334 (2011) 629. Nano 5 (2011) 7108.
[4] A. Hegfleldt, M. Grätzel, Molecular photovoltaics, Accounts of Chemical [19] B. Tan, Y.Y. Wu, Dye-sensitized solar cells based on anatase TiO2 nanoparti-
Research 33 (2000) 269. cles/nanowire composites, Journal of Physical Chemistry B 110 (2006) 15932.
[5] M. Grätzel, Conversion of sunlight to electric power by nanocrystalline dye- [20] S. Ito, T.N. Murakami, P. Comte, P. Liska, C. Grätzel, M.K. Nazeeruddin, M. Grätzel,
sensitized solar cells, Journal of Photochemistry and Photobiology A 164 (2004) Fabrication of thin film dye sensitized solar cells with solar to electric power
3. conversion efficiency over 10%, Thin Solid Films 516 (2008) 4613.
[6] M. Grätzel, Recent advances in sensitized mesoscopic solar cells, Accounts of [21] P. Wang, S.M. Zakeeruddin, P. Comte, R. Charvet, R. Humphry-Baker, M.
Chemical Research 42 (2009) 1788. Grätzel, Enhance the performance of dye-sensitized solar cells by co-grafting
[7] H. Bala, L. Shi, L. Jiang, J.Y. Guo, G.Y. Yuan, L.B. Wang, Z.R. Liu, Preparation amphiphilic sensitizer and hexadecylmalonic acid on TiO2 nanocrystals, Jour-
of lamina-shape TiO2 nanoarray electrode and its electron transport in dye- nal of Physical Chemistry B 107 (2003) 14336.
sensitized solar cells, Acta Physica Sinica 60 (2011) 088101. [22] H. Bala, L. Jiang, W.Y. Fu, G.Y. Yuan, X.D. Wang, Z.R. Liu, Growth of seaweed-
[8] J.Z. Chen, B. Li, J.F. Zheng, S.P. Jia, J.H. Zhao, H.W. Jing, Z.P. Zhu, Role of one- like TiO2 nanoarrays for dye-sensitized solar cells, Applied Physics Letters 97
dimensional ribbonlike nanostructures in dye-sensitized TiO2 -based solar cells, (2010) 153108.
Journal of Physical Chemistry C 115 (2011) 7104. [23] Y.M. Cao, Y. Bai, Q.J. Yu, Y.M. Cheng, S. Liu, D. Shi, F.F. Gao, P. Wang, Dye-
[9] L.Y. Liang, S.Y. Dai, X.Q. Fang, L.H. Hu, Research on the electron transport and sensitized solar cells with a high absorptivity ruthenium sensitizer featuring a
back-reaction kinetics in TiO2 films applied in dye-sensitized solar cells, Acta 2-(hexylthio)thiophene conjugated bipyridine, Journal of Physical Chemistry B
Physica Sinica 3 (2008) 1956. 113 (2009) 6290.
[10] M. Dubey, M. Shrestha, Y.H. Zhong, D. Galipeau, H.S. He, TiO2 nanotube mem- [24] Q.J. Yu, Y.H. Wang, Z.H. Yi, N.N. Zu, Y. Zhang, M. Zhang, P. Wang, High-efficiency
branes on transparent conducting glass for high efficiency dye-sensitized solar dye-sensitized solar cells: the influence of lithium ions on exciton dissociation,
cells, Nanotechnology 22 (2011) 285201. charge recombination, and surface states, ACS Nano 4 (2010) 6032.
[11] B.C. O’Regan, J.R. Durrant, P.M. Sommeling, N.J. Bakker, Influence of the TiCl4 [25] P.M. Sommeling, B.C. O’Regan, R.R. Haswell, H.J.P. Smit, N.J. Bakker, J.J.T.
treatment on nanocrystalline TiO2 films in dye-sensitized solar cells. 2. Charge Smit, J.M. Kroon, J.A.M. Roosmalen, Influence of the TiCl4 post-treatment on
density, band edge shifts, and quantification of recombination losses at short nanocrystalline TiO2 films in dye-sensitized solar cells, Journal of Physical
circuit, Journal of Physical Chemistry C 111 (2007) 14001. Chemistry B 110 (2006) 19191.
[12] A.M. Bakhshayesh, M.R. Mohammadi, D.J. Fray, Controlling electron transport [26] A. Zaban, M. Greenshtein, J. Bisquert, Determination of the electron lifetime in
rate and recombination process of TiO2 dye-sensitized solar cells by design of nanocrystalline dye solar cells by open-circuit voltage decay measurements,
double-layer films with different arrangement modes, Electrochimica Acta 78 Journal of Chemical Physics and Physical Chemistry 4 (2003) 859.
(2012) 384. [27] S. Nakade, Y. Saito, W. Kubo, T. Kitamure, Y. Wada, S. Yanagida, Influence of TiO2
[13] C.Y. Huang, Y.C. Hsu, J.G. Chen, V. Suryanarayanan, K.M. Lee, K.C. Ho, The effects nanoparticle size on electron diffusion and recombination in dye-sensitized
of hydrothermal temperature and thickness of TiO2 film on the performance TiO2 solar cells, Journal of Physical Chemistry B 107 (2003) 8607.
of a dye-sensitized solar cells, Solar Energy Materials and Solar Cells 90 (2006) [28] J. Bisquert, Theory of the impedance of electron diffusion and recombination
2391. in a thin layer, Journal of Physical Chemistry B 106 (2002) 325.