Journal of Power Sources 350 (2017) 35e40

Contents lists available at ScienceDirect

Journal of Power Sources

journal homepage: www.elsevier.com/locate/jpowsour

Ferroelectric BaTiO3 dipole induced charge transfer enhancement in

dye-sensitized solar cells
Keyuan Feng a, b, Xiaoyan Liu b, *, Donghui Si a, b, Xiao Tang b, An Xing b, Minoru Osada c,
Peng Xiao a, **
a
College of Physics, Chongqing University, Chongqing, 400030, China
b
College of Metallurgy and Materials Engineering, Chongqing University of Science and Technology, Chongqing Key Laboratory of Nano/Micro Composites
and Devices, Chongqing, 401331, China
c
International Center for Materials Nanoarchitectonics, National Institute for Materials Science, Tsukuba, Ibaraki, 305-0044, Japan

h i g h l i g h t s g r a p h i c a l a b s t r a c t

 BaTiO3/TiO2 nanocomposite films

were fabricated as photoanodes of
DSCs.
 Ferroelectric dipole induced electric
field would reduce charge
recombination.
 The highest power conversion effi-
ciency was achieved in 1.0 wt%
BaTiO3 addition.

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

Article history: BaTiO3/TiO2 nanocomposite films with varied amount of BaTiO3 are fabricated and applied as photo-
Received 29 December 2016 anodes for dye-sensitized solar cells (DSCs) and demonstrated enhanced power conversion efficiency.
Received in revised form Ferroelectricity of BaTiO3 in the film after subjected to a annealing process up to 450  C is examined by
28 February 2017
Switching Spectroscopy Piezoresponse Force Microscopy (SSPFM). The highest performance is achieved
Accepted 13 March 2017
in 1.0 wt% BaTiO3 addition as a result of increased photocurrent density (Jsc) and fill factor (FF), regardless
of reduction of dye-loading. Electrochemical impedance spectroscopy (EIS) measurements at different
bias voltages (&Voc) in dark suggest that ferroelectric dipole induced electric field has positive effects on
Keywords:
Ferroelectric dipole induced electric field
enhancing electron mobility and suppressing charge recombination. Although more detailed experi-
Charge recombination ments are needed in designing of the nanocomposite films for compensating characteristics of
Electron mobility dye-loading and electron mobility, introduction of ferroelectric dipole induced electric field into the
Dye-sensitized solar cells photoanode would be a good strategy in achieving further improvement of power conversion efficiency
of DSCs through improved charge transfer properties.
© 2017 Elsevier B.V. All rights reserved.

1. Introduction

Dye-sensitized solar cells (DSCs) have received great attention

* Corresponding author.
due to their facile fabrication, potentially low-cost and competitive
** Corresponding author. power conversion efficiency [1e4]. A typically DSC consists of a
E-mail addresses: xyliu@cqust.edu.cn (X. Liu), xiaopeng@cqu.edu.cn (P. Xiao). transparent fluorine doped tin oxide (FTO) glass substrate as a

http://dx.doi.org/10.1016/j.jpowsour.2017.03.049
0378-7753/© 2017 Elsevier B.V. All rights reserved.
36 K. Feng et al. / Journal of Power Sources 350 (2017) 35e40

collector electrode, a dye-sensitized nanocrystalline titanium di- (SSPFM). Photocurrent density-voltage (J-V) characteristics and
oxide (TiO2) film as a photoanode, a Pt-coated glass substrate as a electrochemical impedance spectroscopy (EIS) measurements
counter electrode, and electrolyte containing iodide (I )/tri- demonstrate that the device performance was strongly dependent
iodide(I3 ) redox couple filled between the photoanode and the on the amount of BaTiO3, and the best performance was achieved
counter electrode [5e9]. DSC is the only solar cell that separates with 1.0 wt% BaTiO3 addition resulting from both enhanced Jsc and
two functions of light harvesting and electron transport [10], where FF. EIS measurements at low electron concentrations suggest that
light is absorbed by a monolayer of dye adsorbed on the interfaces ferroelectric dipole induced electric field has positive effects on
of TiO2, and electrons are injected from the dye into the TiO2 upon enhancing electron mobility and suppressing charge recombina-
photo-excitation. Once the electrons are injected into the TiO2 tion in the DSC.
nanoparticles, they experience two types of processes: transport
and recombination. Transport is a forward movement of electrons 2. Experimental
to the collector electrode, whereas recombination is a back flow of
electrons to oxidized dyes and tri-iodide ions [11e13]. Competition 2.1. Preparation of mesoporous BaTiO3/TiO2 films
between transport and recombination can be denoted as a charge
collection efficiency (hc), indicating the degree of electrons reach- A TiO2 paste composed of 3.0 g P25 (TiO2, Degussa, P25) was
ing the collector electrode before getting recombined. Since there is ground using 1.5 g ethyl cellulose (AR, Aladdin) and 10.0 g terpineol
no significant electric field present in the TiO2 film [14], transport of (AR, Aladdin). Commercialized crystalline BaTiO3 nanoparticles
injected electrons to the collector electrode occurs mainly through (Sakai Chemical Industry) at the amounts of 0.0 wt%, 0.5 wt%, 1.0 wt
diffusion. %, 1.5 wt% and 2.5 wt% dispersed into 25.0 ml ethanol were added
In the DSC, the photoanode consisting of mesoporous TiO2 into the TiO2 paste (in the case of 0.0 wt% BaTiO3, 25.0 ml ethanol
nanoparticles provides a large surface area for the adsorption of dye was directly added into the TiO2 paste), stirred sufficiently to get
molecules and pathways for electron transport [15,16]. However, the final paste containing 4.0 ml ethanol. The doctor-blade tech-
the mesoporous structure of TiO2 nanoparticle film with high sur- nique was used to prepare BaTiO3/TiO2 films on an FTO glass sub-
face area has been also considered to escalate charge recombina- strate and the thickness was controlled by a tape. The films were
tion by reducing electron diffusion length and impeding charge first dried at 120  C for 30 min and then annealed at 450  C for
transport in view of the random and highly grained surface and 30 min to eliminate organic additives and to form a mesoporous
boundaries. Improving electron mobility and thus suppressing structure.
charge recombination in the TiO2 nanoparticle film has been
recognized to be one of the key strategies to enhance the power 2.2. Fabrication of the DSCs devices
conversion efficiency of practical DSCs.
Introducing local electric fields into DSCs has been considered as The as-prepared mesoporous BaTiO3/TiO2 films were sensitized
an effective way to enhance the electron mobility. Ferroelectric with 0.5 mM N719 dye at room temperature for 17 h. After rinsing
materials have unique surface properties arising from spontaneous with ethanol to remove excess dye adsorbed on the films and
polarization, which induces bound charges at the surface resulting drying by air flow, the sensitized film was assembled with a counter
in a permanent electric field [17e19]. The polarization-induced electrode made of Pt-coated silicon. The two electrodes were
properties of ferroelectrics have been attracting considerable in- separated by a 30-mm-thick spacer, and the gap between the two
terests in the research field of solar cells. Recently, studies on electrodes was filled with an electrolyte (Yingkou OPV Tech New
introduction of ferroelectric polymer into organic photovoltaic Energy Co., Ltd) composed of LiI, I2, and 4-tert-butylpyridine, 1-
devices have drawn attentions because of their increased perfor- propyl-3-methylimidazolium iodide, guanidine thiocyanate, and
mance. Yuan et al. [20] observed efficiency enhancement in poly- that the concentration of I was 0.07 M.
mer solar cells through utilizing a ferroelectric polyvinylidene
fluoride (PVDF) layer between the cathode electrode and the 2.3. Characterizations
photoactive layer to promote the exciton dissociation. Nalwa et al.
[21] and Z. Xiao et al. [22] increased power conversion efficiency of The crystallinity of BaTiO3 nanoparticles and BaTiO3/TiO2 films
a polymer solar cell by introducing a ferroelectric polymer as an before and after the annealing process was investigated by a X-ray
additive in the photoactive layer. They verified that performance diffraction (XRD, DX-2700 diffractometer) using Cu Ka radiation
improvement in the ferroelectric-incorporated devices was due to (35 kV-25 mA). Field-emission scanning electron microscopy
the increase in Jsc and FF resulting from enhanced exciton dissoci- (FESEM, 7800F, JEOL) was used to characterize the morphologies of
ation by the local electric field of ferroelectric dipoles. By applying the BaTiO3 nanoparticles and the BaTiO3/TiO2 films. Switching
dual phase SrTiO3:ZnO nanocomposite films as cathodic buffer Spectroscopy Piezoresponse force microscopy (SSPFM, Asylum
layers, Lan et al. [23] demonstrated an enhanced power conversion Research, Cypher™ S) was used to characterize ferroelectricity of
efficiency on inverted polymer solar cells. They concluded that local BaTiO3 nanoparticles by measuring phase-voltage hysteresis loop
ordering structure with aligned TiO6 octohedra form spontaneous and amplitude-voltage butterfly loop. The J-V characterization of
polarization, and induce a self-built electric field to prevent elec- the DSCs was performed by a Keithley 2400 multisource meter
tron recombination on the interface of the bulk heterojunction under a simulated AM 1.5 G illumination (Newport 92250A-1000)
active layer and cathodic buffer layer, resulting in high power with an intensity of 100 mW/cm2. The dye-loading amount was
conversion efficiency. measured via UVevis absorption spectra of solutions containing
In this work, ferroelectric BaTiO3 nanoparticles were used to dyes desorbed from the sensitized BaTiO3/TiO2 films using a 0.5 M
make mesoporous BaTiO3/TiO2 nanocomposite films as photo- NaOH aqueous solution. Charge transport properties including
anodes in DSCs. Fig. 1 illustrates the schematic of the photoanode charge transfer resistances (Rct) and transport resistance (Rt) were
consisting of mesoporous BaTiO3/TiO2 nanocomposite film. characterized by the electrochemical impedance spectra (EIS) un-
Ferroelectricity of the BaTiO3 nanoparticles incorporated in the der dark condition with a frequency ranging from 100 mHz to
film after subjected to the annealing process was examined by 105 Hz, keeping an ac amplitude of 10 mV by an electrochemical
phase-voltage hysteresis loop and amplitude-voltage butterfly loop workstation (PASTAT 4000) [24e29]. The photoluminescent (PL)
using Switching Spectroscopy Piezoresponse force microscopy spectra were obtained using photoluminescence mapping system
 K. Feng et al. / Journal of Power Sources 350 (2017) 35e40 37

Fig. 1. (a) Schematic diagram of the photoanode consisting of BaTiO3/TiO2 nanocomposite film. (b) Charge transfer enhancement nearby BaTiO3 dipoles.

(LabRAM HR Evolution) at an excitation wavelength of 325 nm, 1.0 wt% BaTiO3 (Fig. 2(d)) indicate that the addition of BaTiO3 has no
ranging from 350 nm to 800 nm. obviously influence on the morphology of the films. Thickness at
about 12 mm was also controlled for all the films as shown in Fig. S1.
3. Results and discussion Mesoporous structure of the film with an addition of 1.0 wt% BaTiO3
was confirmed by the low angle XRD spectrum as shown in Fig. S2.
Commercialized crystalline BaTiO3 nanoparticles used in this Crystallinity of the films made with the pure P25 and with an
work are pure and highly crystallized, which is confirmed by a XRD addition of 1.0 wt% BaTiO3 was also confirmed by comparing XRD
pattern as shown in Fig. 2(a). The splitting of (202) and (220) peaks spectra of the both films before and after the annealing process, as
indicates that the BaTiO3 nanoparticles possess tetragonal structure shown in Fig. S3.
suggesting that they are ferroelectrics in nature [30]. The SSPFM (Fig. S4) was employed to examine ferroelectricity of the
morphology of the BaTiO3 nanoparticles and BaTiO3/TiO2 films after BaTiO3 nanoparticles before and after the annealing process.
subjected to the annealing process were investigated by the FESEM. Fig. 3(a) shows phase-voltage hysteresis and amplitude-voltage
Fig. 2(b) shows that the BaTiO3 nanoparticles have a diameter butterfly loops of BaTiO3 nanoparticles before the annealing pro-
ranging from 40 nm to 120 nm. Surface topographic images of the cess, while Fig. 3(b) displays the loops of the BaTiO3 nanoparticles
films made with the pure P25 (Fig. 2(c)) and with an addition of incorporated into the film which has been subjected to the

Fig. 2. (a) XRD pattern of the commercialized crystalline BaTiO3 nanoparticles, SEM images of commercialized crystalline BaTiO3 nanoparticles (b), nanocomposite film with pure
P25 (c) and that with an addition of 1.0 wt% BaTiO3 after subjected to the annealing process (d).
38 K. Feng et al. / Journal of Power Sources 350 (2017) 35e40

Fig. 3. Representative phase-voltage hysteresis and amplitude-voltage butterfly loops of BaTiO3 nanoparticles before the annealing (a) and that of BaTiO3 nanoparticles in the
nanocomposite film after the annealing (b).

annealing process. The phase-voltage hysteresis loops show the when the amount of BaTiO3 is more than 1.0 wt%, the power con-
polarization switchable behavior, verifying the ferroelectric prop- version efficiency starts to reduce, and it is caused mainly by the
erties of the BaTiO3 nanoparticles before and after the annealing decrease in Jsc and FF, which is probably attributed to low dye-
process. The difference in amplitude value between the two sam- loading of BaTiO3 and random alignment of BaTiO3 dipoles, as it
ples might be caused by size variation of the BaTiO3 nanoparticles, will be discussed further later in the paper. The change in the open
and varied alignment of BaTiO3 nanoparticles against the substrate circuit voltage (Voc) is very small with the addition of BaTiO3. No
when the BaTiO3 nanoparticles were incorporated into the film. significant difference in open circuit voltage might be due to
BaTiO3/TiO2 films with varied amounts of BaTiO3 are employed neglectable change in work function of BaTiO3/TiO2 with small
as photoanodes in DSCs. J-V curves of the DSCs are shown in amount of BaTiO3. The maximum power conversion efficiency was
Fig. 4(a) and the corresponding photovoltaic parameters are sum- achieved in a DSC with the photoanode film consisting of 1.0 wt%
marized in Table 1. BaTiO3.
In addition, dye-loading amounts were measured by UVevis With a small amount of BaTiO3 in the photoanode film, a local
absorption spectra of solutions containing dyes desorbed from the electric field induced by the ferroelectric dipole might enhance the
sensitized BaTiO3/TiO2 films as shown in Fig. 4(b), and that was also electron mobility in the film, and reduce the charge recombination
listed in Table 1. From Table 1, Compare with the pure TiO2, with a at the interfaces of TiO2/dye/electrolyte, which shows improved
small amount of BaTiO3 (<1.0 wt%) addition, the Jsc is found to in- device performance by Jsc and FF. However, while the BaTiO3
crease from 9.23 to 9.55 mA cm 2, and the FF also increases slightly amount is more than 1.0 wt%, the dye-loading amount drops
from 0.67 to 0.71, although the dye-loading amount decreases greatly and thus decreases electron concentration in the film,
gradually from 74.0 to 69.0 nmol/cm2. As a result, the overall power resulting in a small Jsc. Further, as the BaTiO3 amount exceeds
conversion efficiency increases from 4.86% to 5.39%. However, 1.0 wt%, the dipole-induced local electric field may conversely

Fig. 4. (a) The J-V curves of the DSCs based on BaTiO3/TiO2 films with varied amounts of BaTiO3. (b) UVevis absorption spectra of solutions containing dyes desorbed from the
sensitized BaTiO3/TiO2 films.
 K. Feng et al. / Journal of Power Sources 350 (2017) 35e40 39

Table 1 Table 2
Photovoltaic parameters of the DSCs with varied amounts of BaTiO3 in photoanodes. Charge transport properties of BaTiO3/TiO2 films with varied amounts of BaTiO3.

BaTiO3 (wt%) Jsc (mA cm 2

) Voc (V) FF s (%) dye-loading (nmol cm 2
) BaTiO3 wt% Rct (U) tn (ms)
0.0 9.23 0.786 0.67 4.86 74.0 0.0 66.2 14.0
0.5 9.53 0.786 0.69 5.24 72.8 0.5 82.9 16.5
1.0 9.55 0.796 0.71 5.39 69.0 1.0 101.9 16.7
1.5 9.38 0.787 0.69 5.09 62.7 1.5 89.2 16.0
2.5 8.97 0.777 0.68 4.74 61.8 2.5 66.9 15.4

impede electron transport due to random alignment of dipoles generally involves bulk resistance of semiconductor materials and
(Fig. 1), and thus resulting in a low FF. Electrochemical impedance interfacial resistances, whereas the shunt resistance (Rsh) repre-
spectroscopy (EIS) measurements were performed to investigate sents the electron transfer resistance at the TiO2/electrolyte inter-
the effects of ferroelectric dipoles induced electric field on charge face donated as interfacial charge recombination or Rct [4].
transport properties of DSCs. The impedance results of the DSCs Increasing the Rsh or Rct will lead to higher FF, thus resulting in
with varied amount of BaTiO3 under the open voltage in dark greater efficiency. As it can be seen from Table 2, the DSCs with
condition are shown in Fig. 5 and the corresponding electro- 1.0 wt% BaTiO3 addition in the photoanode has the largest Rct, and
chemical parameters are summarized in Table 2. All the EIS spectra thus highest FF, which is in agreement with J-V characteristics as
were fitted by using Z-View software with an equivalent circuit shown in Fig. 4(a).
based on the general transmission line mode that is inserted in To further demonstrate the effect of ferroelectric dipole induced
Fig. 5(a). In the circuit, R1 represents Faraday resistance of the electric field on charge transport properties of DSCs, varied bias
redox reaction of I /I3 at the Pt/electrolyte interface at the high- from 0.76 V to 0.70 V were applied for EIS measurements. Under
frequency (kHz), Rt is the electron transport resistance denoting the dark condition, electrons are injected through an external cir-
how effective the electrons in the conduction band of TiO2 cuit into a photoanode film from a FTO substrate and the photo-
network [31], and Rct represents the charge transfer resistance at anode film gets charged by prepagation of the injected electrons,
the interfaces of TiO2/dye/electrolyte and within the TiO2 films therefore different bias voltage will result in different electron
[32,33]. The Rct is shown as the second and large semicircle in the concentration in the film [38]. Fig. 6 shows bias voltage dependence
Nyquist plot (Fig. 5(a)), and it is the most decisive factor for the of charge transfer resistance (Rct) and electron lifetime (tn) in the
performance of DSCs [34]. In the dark under a forward bias, elec- pure P25 photoanode and the photoanode with 1.0 wt% BaTiO3
trons are injected in the conduction band of the TiO2 nanoparticles addition. As shown in Fig. 6(a), Rct increased with reduction of the
and then transport through the mesoscopic TiO2 network, react bias voltage for both devices due to a lower electron concentration
with I3 in the electrolyte (recombination process), and reduce the under a smaller bias. Compared with the pure P25 photoanode,
I3 to I , followed by the oxidation of I to I3 at the interface be- charge transfer resistance (Rct) in the photoanode with 1.0 wt%
tween the electrolyte and counter electrode. BaTiO3 addition has larger increment indicating that the ferro-
A larger Rct implies less charge recombination at the interfaces electric dipole induced electric field has even more obvious effects
of TiO2/dye/electrolyte [35, 36]. From Table 2, the DSCs with the on suppressing charge recombination at lower electron concen-
photoanode film consisting of 1.0 wt% BaTiO3 has a largest value of tration. As shown in Fig. 6(b), the same trend was also found in bias
Rct. This result suggests that the electrons in the photoanode voltage dependence of electron lifetime (tn). The larger increment
transport more efficiently with a decrease in the recombination of tn in the photoanode with 1.0 wt% BaTiO3 addition further sug-
rate, which is in agreement with the frequency response shown in gests more significant effects on electron mobility and charge
Fig. 5(b), revealing a longer lifetime (tn) for the device with the recombination at the lower electron concentration [39]. In addition,
photoanode film consisting of 1.0 wt% BaTiO3. Longer lifetime can photoluminescence spectroscopy (PL) shows that the PL intensity
be attributed to higher electron mobility induced by ferroelectric was found to be decreased in the photoanode with addition of
dipole induced electric field in the photoanode film. 1.0 wt% BaTiO3 comparing with the pure P25 photoanode (Fig. S5).
Fill factor reflects electrical and electrochemical losses occurring This result again suggests that recombination of photo-induced
during operation of the DSCs, which can be contributed by series electrons and holes can be suppressed in the BaTiO3/TiO2 nano-
and shunt resistance in the device [37]. The series resistance (Rs) composite films.

Fig. 5. (a) Nyquist plot and (b) Bode plot of the DSCs based on BaTiO3/TiO2 films with varied amount of BaTiO3 under the open voltage in dark. Color circles represent experimental
data.
40 K. Feng et al. / Journal of Power Sources 350 (2017) 35e40

Fig. 6. (a) Charge transfer resistance (Rct) and (b) electron lifetime (tn) measured at different bias voltages (&Voc) in dark.

4. Conclusions [8] G. Calogero, G.D. Marco, S. Caramori, S. Cazzanti, R. Argazzi, C.A. Bignozzi,
Energy Environ. Sci. 2 (2009) 1162e1172.
[9] H. Wang, X. Liu, Z. Wang, H. Li, D. Li, A. Qingbo Meng, L. Chen, J. Phys. Chem. C
BaTiO3/TiO2 nanocomposite films have been fabricated and 110 (2006) 5970e5974.
demonstrated as photoanodes in dye-sensitized solar cells for [10] W.T. Sun, Y. Yu, H.Y. Pan, X.F. Gao, A. Qing Chen, L.M. Peng, J. Am. Chem. Soc.
improved power conversion efficiency. The device performance 130 (2008) 1124e1125.
[11] B.A. Gregg, Coord. Chem. Rev. 248 (2004) 1215e1224.
was found to be strongly dependent on the amount of BaTiO3 in the [12] Y.Y. Wang, J.H. Cai, L. Sun, J.J. Gong, C.J. Lin, Chin. J. Inorg. Chem. 28 (2012)
photoanode. With a small amount of BaTiO3 (<1.0 wt%), ferro- 647e656.
electric BaTiO3 dipole induced electric field would increase electron [13] P.R. Barnes, K. Miettunen, X. Li, A.Y. Anderson, T. Bessho, M. Gratzel,
B.C. O'Regan, Adv. Mater 25 (2013) 1881e1922.
mobility and reduce charge recombination. However, further in- [14] M. Wang, P. Chen, R. Humphrybaker, S.M. Zakeeruddin, M. Gr€ atzel, Chem.
crease in the amount of BaTiO3 led to lower photocurrent density Phys. Chem. 10 (2009) 290e299.
and smaller fill factor. Our present study concluded that the pho- [15] Q. Zhang, G. Cao, Nano Today 6 (2011) 91e109.
[16] Q. Zhang, C.S. Dandeneau, X. Zhou, G. Cao, Adv. Mater. 21 (2009) 4087e4108.
toanode with 1.0 wt% BaTiO3 addition provided the best device [17] A. Pecchia, D. Gentilini, D. Rossi, d.M.M. Auf, C.A. Di, Nano Lett. 16 (2015)
performance, and the power conversion efficiency increased by 11% 988e992.
comparing with the pure TiO2 photoanode. The highest power [18] P. Raksa, S. Pinitsoontorn, S. Maensiri, Ferroelectrics 492 (2016) 150e158.
[19] F. Wu, X. Li, Y. Tong, T. Zhang, J. Power Sources 342 (2017) 704e708.
conversion efficiency achieved in 1.0 wt% BaTiO3 addition suggests [20] Y. Yuan, T.J. Reece, P. Sharma, S. Poddar, S. Ducharme, A. Gruverman, Y. Yang,
that designing of the nanocomposite films for compensating J. Huang, Nat. Mater 10 (2011) 296e302.
characteristics of electron concentration, electron mobility and [21] K.S. Nalwa, J.A. Carr, R.C. Mahadevapuram, H.K. Kodali, S. Bose, Y. Chen,
J.W. Petrich, B. Ganapathysubramanian, S. Chaudhary, Energy Environ. Sci. 5
charge transport is important for high performance DSCs.
(2012) 7042e7049.
[22] Z. Xiao, Q. Dong, P. Sharma, Y. Yuan, B. Mao, W. Tian, A. Gruverman, J. Huang,
Acknowledgment Adv. Energy Mater 3 (2013) 1581e1588.
[23] J.L. Lan, Z. Liang, Y.H. Yang, F.S. Ohuchi, S.A. Jenekhe, G. Cao, Nano Energy 4
(2014) 140e149.
This work was supported by the National Natural Science [24] T. Georgakopoulos, A. Apostolopoulou, N. Todorova, K. Pomoni, C. Trapalis,
Foundation of China (51472037), cooperative project of Chongqing E. Stathatos, J. Alloys Compd. 692 (2017) 313e321.
[25] Q. Wang, J.E. Moser, M. Gra €tzel, J. Phys. Chem. B 109 (2005) 14945e14953.
Key Laboratory of Nano/Micro Composite Materials and Devices
[26] L. Zhao, C. Zhong, Y. Wang, S. Wang, B. Dong, L. Wan, J. Power Sources 292
(CQKL-1501). (2015) 49e57.
[27] J. Chen, T. Peng, K. Fan, J. Xia, J. Mater. Chem. 21 (2011) 16448e16452.
Appendix A. Supplementary data [28] X. Gao, D.G. n, J. Huo, J. mChen, C. Yuan, Nanoscale 5 (2013) 10438e10446.
[29] S.A. Mozaffari, M. Ranjbar, E. Kouhestanian, H.S. Amoli, M.H. Armanmehr,
Mater. Sci. semicon. Process 40 (2015) 285e292.
Supplementary data related to this article can be found at http:// [30] P. Ganguly, A. Kumar, S. Tripathi, D. Haranath, A.M. Biradar, Appl. Phys. Lett.
dx.doi.org/10.1016/j.jpowsour.2017.03.049. 102 (2013) 222902e222904.
[31] K. Park, Q. Zhang, J. Xi, G. Cao, Thin Solid Films 588 (2015) 19e25.
[32] C.Y. Liu, K.C. Huang, P.H. Chung, C.C. Wang, C.Y. Chen, R. Vittal, C.G. Wu,
References W.Y. Chiu, K.C. Ho, J. Power Sources 217 (2012) 152e157.
[33] E. Antolini, L. Giorgi, A. Pozio, E. Passalacqua, J. Power Sources 77 (1999)
[1] B. O’regan, M. Grfitzeli, Nature 353 (1991) 737e740. 136e142.
[2] M. Gra €tzel, J. Photoch. Photobio. A 164 (2004) 3e14. [34] K. Wan, F. Wu, Y. Dou, L. Fang, C. Mao, J. Alloys Compd. 680 (2016) 373e380.
[3] A. Polman, M. Knight, E.C. Garnett, B. Ehrler, W.C. Sinke, Science 352 (2016). [35] K.M. Lee, C.W. Hu, H.W. Chen, K.C. Ho, Sol. Energy Mater. So. C 92 (2008)
[4] M. Gra €tzel, Nature 414 (2001) 338e344. 1628e1633.
[5] Ana Morandeira, Je ro
^me Fortage, Tomas Edvinsson, L.L.P. Errol Blart, [36] T. Georgakopoulos, A. Apostolopoulou, N. Todorova, K. Pomoni, C. Trapalis,
Gerrit Boschloo, Anders Hagfeldt, Leif Hammarstro €m, Fabrice Odobel, J. Phys. E. Stathatos, J. Alloys Compd. 692 (2016) 313e321.
Chem. C 112 (2008) 1721e1728. [37] B. Qi, J. Wang, Phys. Chem. Chem. Phys. 15 (2013) 8972e8982.
[6] C.Y. Lin, Y.C. Wang, S.J. Hsu, C.F. Lo, W.G. Diau, J. Phys. Chem. C 114 (2010) [38] Q. Wang, J.E. Moser, M. Gra €tzel, J. Phys. Chem. B 109 (2005) 14945e14953.
176e184. [39] E. Palomares, J.N. Clifford, S.A. Haque, T. Lutz, J.R. Durrant, Chem. Inf. 14 (2002)
[7] J. Fan, S. Liu, J. Yu, J. Mater. Chem. 22 (2012) 17027e17036. 1464e1465.