Highly efficient light harvesting ruthenium sensitizers for dye-sensitized solar cells featuring triphenylamine donor antennas

Article abstract of DOI:10.1039/c3ta15134d

Two new heteroleptic ruthenium complexes, coded as KW1 and KW2, featuring triphenylamine electron-donating antennas, have been synthesized and used in dye-sensitized solar cells (DSSCs). Compared to the reference dye Z907, these new dyes exhibit broad absorption and efficient light harvesting properties. Particularly the KW2 dye exhibits a low-energy metal-ligand charge transfer band centred at 554 nm with a high molar extinction coefficient of 2.43 × 10⁴ M^-1 cm^-1 arising from an extended conjugation in the donor antenna ligand. Photovoltaic devices using these sensitizers in conjunction with a volatile electrolyte show high photovoltaic conversion efficiencies of ～10.7% under full sunlight irradiation (AM 1.5G, 100 mW cm^-2). This efficiency is nearly 20% higher than that of the Z907-based reference device, which is attributed to a largely improved short circuit current. The distinct effects of different donor antennas incorporated in the ancillary ligands on the primary photovoltaic parameters in these devices are investigated with transient photoelectrical decays and impedance spectroscopy measurements. The devices utilizing these highly efficient light harvesting ruthenium sensitizers featuring triphenylamine donor antennas in combination with low volatility electrolytes exhibit good durability during the accelerated tests (60°C for 1000 h in a solar simulator, 100 mW cm ^-2). This journal is the Partner Organisations 2014.

Full text of DOI:10.1039/c3ta15134d

Journal of
Materials Chemistry A
View Article Online
View Journal | View Issue
PAPER
Highly efficient light harvesting ruthenium
sensitizers for dye-sensitized solar cells featuring
Cite this: J. Mater. Chem. A, 2014, 2,
triphenylamine donor antennas†
4945
Kun Cao,a Jianfeng Lu,a Jin Cui,a Yan Shen,a Wei Chen,a Getachew Alemu,a
a
a
b
a
Zhong Wang, Huailiang Yuan, Jie Xu, Mingkui Wang and Yibing Chengc
*
*
Two new heteroleptic ruthenium complexes, coded as KW1 and KW2, featuring triphenylamine electron-
donating antennas, have been synthesized and used in dye-sensitized solar cells (DSSCs). Compared to
the reference dye Z907, these new dyes exhibit broad absorption and efficient light harvesting
properties. Particularly the KW2 dye exhibits a low-energy metal–ligand charge transfer band centred at
554 nm with a high molar extinction coefficient of 2.43 ꢀ 104 Mꢁ1 cmꢁ1 arising from an extended
conjugation in the donor antenna ligand. Photovoltaic devices using these sensitizers in conjunction with
a volatile electrolyte show high photovoltaic conversion efficiencies of ꢂ10.7% under full sunlight
irradiation (AM 1.5G, 100 mW cmꢁ2). This efficiency is nearly 20% higher than that of the Z907-based
reference device, which is attributed to a largely improved short circuit current. The distinct effects of
different donor antennas incorporated in the ancillary ligands on the primary photovoltaic parameters in
these devices are investigated with transient photoelectrical decays and impedance spectroscopy
measurements. The devices utilizing these highly efficient light harvesting ruthenium sensitizers featuring
triphenylamine donor antennas in combination with low volatility electrolytes exhibit good durability
during the accelerated tests (60 ꢃC for 1000 h in a solar simulator, 100 mW cmꢁ2).
Received 11th December 2013
Accepted 14th January 2014
DOI: 10.1039/c3ta15134d
www.rsc.org/MaterialsA
A great effort has been made to optimize the molecular
structures of ruthenium complexes by changing ancillary
Introduction
ligands, typically bipyridines (bpy) or terpyridines (tpy), which
can be tuned by different substituents (alkyl, aryl, heterocycle,
etc.) to modulate their photochemical and photo-physical
properties, and thus, improve the device photovoltaic perfor-
mance.5 Recently a novel class of thiocyanate free Ru(II) sensi-
tizers with functionalized pyridyl azolate ancillary ligands has
been reported, achieving high efficiencies over 10%.6 It has
been well-accepted that enhancing the molar extinction coeffi-
cient of a sensitizer is an elegant strategy to improve the
photovoltaic performance.7 Incorporation of high electron
delocalization donor antenna substituents into the ancillary
ligands can substantially improve the light harvesting property
of such photo-sensitizers, therefore increasing the device over-
all power conversion efficiency (PCE).8 Ruthenium complexes
with ligands functionalized by carbazole-thiophenyl groups
were well-designed to enhance the light harvesting capacity,
demonstrating that the carbazole moiety in the dye could
enhance the performance of the device.9 A series of heteroleptic
ruthenium(II)-polypyridyl sensitizers were reported recently, in
which electron-donating NMe2 was changed to electron-with-
drawing CN incorporated in the benzene-substituted bipyridine
ligand.10 The DSSC devices using these sensitizers have ach-
ieved efficiencies close to 10–11%. It has been demonstrated
that by introduction of electron-donor groups, such as
Dye-sensitized solar cells (DSSCs) have received considerable
attention due to their relatively low production cost and other
attractive features, such as transparency and exibility.1
A
photo-sensitizer is one of the most important components of
this type of devices, which is responsible for the light harvesting
and electron injection. Various photo-sensitizers, including
metal complexes, phthalocyanines, porphyrins and metal-free
organic dyes, have been investigated and applied for DSSCs in
the past decades.2 Particularly, ruthenium complexes have
received intensive interest due to their favorable photo-elec-
trochemical properties and high stability.3 So far several DSSCs
based on ruthenium complexes have achieved photovoltaic
efficiencies of over 10% under standard measurement condi-
tions and the highest value is 11.5%.4
a
¨
Michael Gratzel Center for Mesoscopic Solar Cells, Wuhan National Laboratory for
Optoelectronics, Huazhong University of Science and Technology, 1037 Luoyu Road,
430074 Wuhan, P. R. China. E-mail: mingkui.wang@mail.hust.edu.cn
bSchool of Materials Science and Engineering, Wuhan Textile University, FangZhi
Road, 430073, Wuhan, P. R. China. E-mail: xujie0@mail.ustc.edu.cn
cDepartment of Materials Engineering, Monash University, Melbourne, Victoria, 3800,
Australia
† Electronic supplementary information (ESI) available: Synthesis routes of
ancillary ligands KW-Ll and KW-L2, their NMR and HRMS characterization, and
dark-current measurement. See DOI: 10.1039/c3ta15134d
This journal is © The Royal Society of Chemistry 2014
J. Mater. Chem. A, 2014, 2, 4945–4953 | 4945

View Article Online
Journal of Materials Chemistry A
Paper
3-phenylpropionic acid (PPA) were purchased from Aldrich.
THF was dried with sodium sand and dimethylformamide
(DMF) was distilled prior to use. 4,40-Dibromo-2,20-bipyridine,15
2-(4-diphenylaminopheny)-4,4,5,5-tetramethyl-1,3,2-dioxabor-
olane,16 and N,N-diphenyl-4-(5-(4,4,5,5-tetramethyl-1,3,2-dioxa-
borolan-2-yl)thiophen-2-yl) aniline17 were prepared according to
the literature procedures.3,14 All reactions were performed under
a dry argon atmosphere.
The structures of KW1 and KW2 were veried by NMR, MS,
UV-vis spectroscopy and electrochemical measurement. 1H
NMR spectra were recorded on a Bruker AV400 400 MHz with
tetramethylsilane as an internal standard. High resolution mass
spectra (HRMS) were measured with a Bruker micro TOF mass
spectrometer. Electronic absorption spectra were observed with
a PE950 spectrophotometer and uorescent emission spectra
were obtained with a Jasco FP-6500 spectrophotometer. Square-
wave voltammograms of various dyes were measured on a
CHI660C electrochemical workstation in a three-electrode
single-compartment cell with a glassy carbon working elec-
trode, a platinum wire counter electrode and an Ag/AgCl refer-
ence electrode. All potentials were internally referenced to the
ferrocene/ferrocenium (Fc/Fc+) couple.
Scheme 1 Molecular structures of the ruthenium complex sensitizers
KW1, KW2, and Z907.
triphenylamine (TPA), into the ancillary moieties, a positive
charge or ‘hole’ can be shied onto this moiety, potentially
increasing the distance between the dye–cation moiety and the
TiO2 surface, and thus retarding the interfacial charge recom-
bination dynamics.11 Based on these strategies, several novel
ruthenium dyes have been designed and synthesized, whose
ancillary bipyridine ligands are modied by a covalent attach-
ment of triarylamine-based electron-donating groups.12 DSSC
devices utilizing these dyes present relatively high PCEs in the
range of 7–10%.11 These sensitizer dyes exhibit not only efficient
light harvesting capability, i.e., an extended absorption range
and high molar absorption coefficient, but also an effective
intramolecular spatial charge separation of the excited state of
dye sensitizers, which are critically important for highly effi-
cient DSSCs.
Herein we communicate an investigation of the inuence of
two electron-donating antennas upon the photo-physical and
photo-electrochemical properties of ruthenium complexes and,
especially, their photovoltaic performance in the DSSCs. Two
new ruthenium sensitizers, coded as KW1 and KW2 (see
Scheme 1), were designed and synthesized for this aim. For the
KW1 dye molecule, the TPA is directly attached to the bpy unit;
however, for the KW2 dye a thiophene used as the conjugated
spacer is incorporated between the TPA and the bpy unit. The
thiophene conjugated spacer has been demonstrated to have a
profound effect on the photo-physical properties of [Ru(tpy)2]2+
complexes because of its polarizability and low resonance
energy.13 We investigated their inuence on DSSC device
performance. Especially, the interfacial charge transfer
processes were investigated with transient photovoltage/
photocurrent decay measurements together with electrical
impedance analysis. Furthermore, for comparing the inuences
of the electron-donating groups on spectral and photovoltaic
properties, we correlated these novel sensitizers with a model
dye Z907,14 in which the aromatic amine electron-donating
groups are substituted by long alkyl chains.
Synthesis of 4,40-bis(N,N-diphenyl-4-aminophenyl)-2,20-
bipyridine (ligand-1)
A mixture of compound 2-(4-(diphenylamino)-4,4,5,5-tetra-
methyl-1,3,2-dioxaborolane) (0.408 g, 1.1 mmol), 4,40-dibromo-
2,20-bipyridine (0.157 g, 0.5 mmol), and Pd(PPh3)4 (45 mg,
0.04 mmol) was dissolved in toluene–EtOH (2 : 1, v/v, 30 mL)
and stirred for 10 min. K2CO3 (1.1 g, 8 mmol) in 10 mL H2O was
subsequently added and the mixture was reuxed under argon
for 40 h. Aer the mixture was cooled to room temperature, the
organic layer was separated and the product was extracted with
CHCl3 (3 ꢀ 30 mL). The combined organic layers were dried
with anhydrous Na2SO4 and the solvent was evaporated under
reduced pressure. Purication of the crude product by column
chromatography (CH2Cl2–hexane: 1 : 5) afforded ligand-1 as a
yellow solid (0.170 g, yield: 53%). 1H NMR (400 MHz, CDCl3, d):
8.70 (s, 2H), 8.69 (d, J ¼ 5.2 Hz, 2H), 7.68 (d, J ¼ 7.8 Hz, 4H), 7.53
(d, 2H), 7.31–7.27 (t, J ¼ 7.5 Hz, 8H), 7.17–7.14 (m, 12H), 7.06–
7.09 (t, J ¼ 7.3 Hz, 4H). MS (ESI) m/z calcd for (C46H34N4):
642.79. Found: 643.4.
Synthesis of 4,40-bis(5-(N,N-diphenyl-4-aminophenyl)-
thiophen-2-yl)-2,20-bipyridine (ligand-2)
N,N-Diphenyl-4-(5-(4,4,5,5-tetramethyl-1,3,2-dioxaborolan-2-yl)
thiophen-2-yl) aniline (0.498 g, 1.1 mmol), 4,40-dibromo-2,20-
bipyridine (0.157 g, 0.5 mmol), K2CO3 (0.552 g, 4 mmol), and
Pd(PPh3)4 (45 mg, 0.04 mmol) were dissolved in THF–H2O (9 : 1,
v/v, 50 mL). The solution was degassed and heated at reux for
24 h upon which a half equivalent of Pd(PPh3)4 (22 mg,
0.019 mmol) was added to the reaction mixture. The reaction
Experimental section
Materials and characterization
All solvents and reagents were of puriss grade quality and used was heated at reux for another 16 h and then the product was
as received except those specied. Dichloro(p-cymene) ruth- extracted with CHCl3 (3 ꢀ 30 mL). The combined organic layers
enium(II) dimer and 4,40-dicarboxy-2,20-bipyridine were were dried with anhydrous Na2SO4 and aer the rotary evapo-
purchased from Alfa Aesar. Ammonium thiocyanate and ration of solvents, the resulting solid was puried by column
4946 | J. Mater. Chem. A, 2014, 2, 4945–4953
This journal is © The Royal Society of Chemistry 2014

View Article Online
Paper
Journal of Materials Chemistry A
chromatography on silica gel using 1 : 10 CH3OH–CHCl3 as an
eluent to afford the target compound (0.194 g, 48% yield) as a
Device fabrication
A screen-printed single or double layer lm of interconnected TiO2
particles was used as a mesoporous negative electrode. The 20 nm-
sized TiO2 particles were rst printed on the uorine-doped
conducting glass (FTO) substrates and then coated with a 5 mm
thick second layer of 400 nm-sized light-scattering anatase
particles (WER2-O, Dyesol) if needed. The lm thickness was
measured using a Prole-system (DEKTAK, VECCO, Bruker). The
detailed preparation procedures of TiO2 lms have been described
1
yellow solid. H NMR (400 MHz, CDCl3, d): 8.71 (d, J ¼ 5.3 Hz,
2H), 8.69 (s, 2H), 7.65 (d, J ¼ 3.8 Hz, 2H), 7.55 (m, 6H), 7.33–7.29
(m, 10H), 7.17–7.07 (m, 16H). MS (ESI) m/z calcd for
(C54H38N4S2): 807.04. Found: 807.6.
Synthesis of KW1 (Ru[(dcbpy)(ligand-1)(NCS)2])
Dichloro(p-cymene)ruthenium(II) dimer (0.11 g, 0.18 mmol) and
ligand-1 (0.231 g, 0.36 mmol) were dissolved in DMF (50 mL)
and the solution was degassed. The reaction mixture was stirred
elsewhere.4 The TiO2 lm was rst sintered at 500 C for 30 min
and then cooled to about 80 C in air. Then the TiO2 lm elec-
ꢃ
ꢃ
ꢃ
at 75 C for 4 h under Ar in the dark. Subsequently, 4,40-dicar-
trodes were stained by immersing them into a dye solution con-
taining a KW1 or Z907 sensitizer (300 mM) in a mixture of
acetonitrile and tert-butyl alcohol (volume ratio: 1/1) or a dye
solution containing KW2 (300 mM) with 10% DMSO in a acetoni-
trile and tert-butyl alcohol (volume ratio: 1 : 1) mixture overnight.
Aer washing with acetonitrile and drying by air ow, the sensi-
tized titania electrode was assembled with a thermally platinized
FTO electrode. The electrodes were separated by a 45 mm thick hot
melt ring (Surlyn, DuPont) and sealed by heating. The internal
space was lled with a liquid electrolyte with use of a vacuum back
lling system. In this study, a conventional liquid triiodide/iodide
electrolyte (coded Z960) was used for a high efficiency device,
containing 0.1 M LiI, 0.05 M I2, 0.6 M PMII, and 0.5 M 4-tert-
butylpyridine in a mixture of valeronitrile and acetonitrile (volume
ratio, 15 : 85). A low volatility electrolyte was used for stability
testing, containing 1.0 M DMII, 0.15 M I2, 0.5 M N-butylbenzimi-
dazole, and 0.1 M GNCS in methoxypropionitrile (MPN).
boxylic acid-2,20-bipyridine (0.088 g, 0.36 mmol) was added and
ꢃ
the reaction mixture was heated up to 145 C for 4 h. Next, an
excess of ammonium thiocyanate (0.82 g, 10.80 mmol) was
added to the resulting dark solution and it was continued for
additional 4 h at the same temperature. Then the reaction
mixture was cooled to room temperature overnight. The solvent
was removed on a roto-evaporator until little DMF remained
and water (100 mL) was added to obtain a suspended solution.
The resulting solid was collected on a sintered glass crucible by
suction ltration, further washed with distilled water and
diethyl ether, and dried under a vacuum. The crude product was
dissolved in basic methanol (NaOH) and puried on a Sephadex
LH-20 column with methanol as the eluent. The main band was
collected and the pH value of the collected solution was lowered
to ca. 3 by adding dilute HNO3. The collected precipitate was
washed with water and dried under a vacuum. Aer purica-
tion, 0.167 g (yield: 42%) of KW1 was obtained. 1H NMR
(400 MHz, DMSO-d6 d) 9.52 (d, J ¼ 5.8 Hz, 1H), 9.21 (d, J ¼ 6.1
Hz, 1H), 9.17 (s, 1H), 9.14 (s, 1H), 8.99 (s, 2H), 8.35 (d, J ¼ 5.6 Hz,
1H), 8.29 (d, J ¼ 5.9 Hz, 1H), 8.12 (d, J ¼ 8.6 Hz, 2H), 7.95 (d, J ¼
5.9 Hz, 1H), 7.83 (d, J ¼ 5.2 Hz, 2H), 7.68 (d, J ¼ 6.2 Hz, 1H),
7.68–7.36 (m, 10H), 7.19–7.10 (m, 14H), 7.00–6.98 (d, J ¼ 8.4 Hz,
2H). HRMS m/z calcd for (C60H42N8O4RuS2) [M ꢁ NCS]+,
1046.2062; found: 1046.2069.
DSSC device characterization
A 450 W xenon light source solar simulator (Oriel, model 9119)
with an AM 1.5G lter (Oriel, model 91192) was used to give an
irradiance of 100 mW cmꢁ2 (1 sunlight intensity) at the surface
of solar cells. The current–voltage characteristics of the cell were
obtained by applying an external potential bias to the sample
and measuring the generated photocurrent with a Keithley
model 2400 digital source meter (Keithley, USA). A similar data
acquisition system was used to control the incident photon-to-
current conversion efficiency (IPCE) measurement. A white light
bias (1% sunlight intensity) was used to bring the total light
intensity on the device under test closer to operating condi-
tions. The devices with the photoanode area of 0.16 cm2 were
tested using a metal mask with an aperture area of 0.09 cm2.
Synthesis of KW2 (Ru[(dcbpy)(ligand-2)(NCS)2])
Subsequent to the one-pot synthesis described above for KW1,
except for starting from ligand-2 (0.291 g, 0.36 mmol), the
solvent was removed on a roto-evaporator until little DMF
remained and water (100 mL) was added to obtain a suspended
solution. The insoluble solid was collected and washed
intensively with water, further washed with diethyl ether and
dried under a vacuum. The black solid was dissolved in DMF
and precipitated aer addition of diethyl ether. This procedure
was repeated twice and the product was obtained as a black
Electrochemical impedance spectroscopy and transient
photovoltage decay measurements
solid. Aer purication, 0.155 g (yield: 34%) of KW2 was Impedance measurements of DSSCs were carried out with an
obtained. 1H NMR (400 MHz, DMSO-d6 d): 9.49 (d, J ¼ 5.6 Hz, Autolab Frequency Analyzer set-up, which consists of an Auto-
1H), 9.18 (d, J ¼ 6.0 Hz, 1H), 9.13 (s, 1H), 9.11 (s, 1H), 8.98 (s, lab PGSTAT 30 (Eco Chemie B. V., Utrecht, The Netherlands)
1H), 8.96 (s, 1H), 8.34 (d, J ¼ 5.8 Hz, 1H), 8.30 (d, J ¼ 2.1 Hz, producing
a small-amplitude harmonic voltage, and a
1H), 8.21 (d, J ¼ 5.3 Hz, 1H), 8.10 (d, J ¼ 3.2 Hz, 1H), 7.99–7.96 frequency response analyzer module. The determination of the
(m, 2H), 7.75–7. 71 (m, 3H), 7.65–7.59 (m, 4H), 7.42–7.32 (m, interfacial recombination rate constant was carried out by
9H), 7.15–7.05 (m, 14H), 7.00–6.98 (d, J ¼ 8.8 Hz, 2H). HRMS m/ performing transient photovoltage decay measurements and
z calcd for (C68H46N8O4RuS4) [M ꢁ NCS]+, 1210.1817; found: charge-extraction experiments. Details can be found in our
1210.1760.
previous report.18
This journal is © The Royal Society of Chemistry 2014
J. Mater. Chem. A, 2014, 2, 4945–4953 | 4947

View Article Online
Journal of Materials Chemistry A
Paper
p–p* charge transition of 4,40-dicarboxylic acid-2,20-bipyridine
(dcbpy) and the ancillary bipyridine ligand. Another absorption
band in the UV region was observed at 397 nm for the KW1 dye
and 434 nm for the KW2 dye. It can be attributed to the p–p*
transition of the ancillary bipyridine ligand and higher-energy
metal-to-ligand charge transfer (MLCT) transition. The elec-
tronic absorption spectra of the KW1 and KW2 dyes presented
the characteristic MLCT absorption bands in the visible region,
similar to other heteroleptic polypyridyl ruthenium(II) com-
plexes.3c,19 In DMF, the lower energy MLCT bands centred at
538 nm for the KW1 dye and 554 nm for the KW2 dye were
observed, which were 18 nm and 34 nm red-shied compared to
that of the reference Z907 dye, respectively. The molar extinc-
tion coefficients (3) of the low-energy MLCT absorption bands
for KW1 and KW2 were calculated to be 18.4 ꢀ 103 Mꢁ1 cmꢁ1
and 24.3 ꢀ 103 Mꢁ1 cmꢁ1, respectively. These values are
signicantly higher than those of the Z907 model sensitizer
(12.0 ꢀ 103 Mꢁ1 cmꢁ1) and the CYC-B6S sensitizer (16.1 ꢀ 103
Mꢁ1 cmꢁ1).9a The latter has alkyl-substituted carbazole moieties
in the thiophene-substituted bipyridine ligand.9a The molar
extinction coefficient enhancement and bathochromic shi
come from the extension p-conjugation in the ancillary ligand
by incorporating electron-rich moieties.20 Clearly, the DSSCs
can benet from dyes with a high molar extinction coefficient.
Considering the same surface area being occupied by sensi-
tizers with different light harvesting efficiencies, those with
enhanced molar extinction coefficients allow a high light har-
vesting yield, and eventually, a large reduction in lm thickness.
As shown in Fig. 1b, the polypyridyl ruthenium(II) dyes con-
taining TPA-donor antennas increased the visible light har-
vesting capacity due to a larger absorption cross-section. This
would fairly augment the device overall PCE because of reduced
transport losses in the nanoporous environment for a thinner
lm.12 The molar absorption coefficients for the MLCT bands of
KW1 and KW2 are much higher than those of analogous
ruthenium complexes containing 5,50-disubstituted-2,20-bipyr-
idine ancillary ligands, 5,50-bis(N,N-diphenyl-4-amino phenyl)-
2,20-bipyridine (6.4 ꢀ 103 Mꢁ1 cmꢁ1) and 5,50-bis(5-(N,N-
diphenyl-4-aminophenyl)-thiophen-2-yl)-2,20-bipyridine (3.2 ꢀ
103 Mꢁ1 cmꢁ1).21 This could be the main reason for the
performance enhancement in devices with KW1 and KW2 dyes.
The sensitizers displayed similar spectral features when
adsorbed onto 2.3 mm transparent TiO2 lms, except for a slight
blue shi of the absorption maxima due to interaction of
anchoring groups with the surface.
Device stability test
The test was performed under illumination with visible light (1
sun; 100 mW cmꢁ2) at 60 ꢃC, and the photovoltaic performance
during the test was recorded. The physicochemical measure-
ments were conducted at room temperature aer the cell was
equilibrated and cooled down to room temperature.
Results and discussion
New ruthenium sensitizers coded as KW1 and KW2, which
incorporate TPA– and TPA–thiophene-substituted bipyridines
as the ancillary ligands, were synthesized using a stepwise
synthesis protocol. The detailed synthesis procedures are
illustrated in the ESI (Scheme S1†). The desired disubstituted
bipyridine ancillary ligands were obtained from the corre-
sponding boronic acid with dibromopyridine by means of a
Suzuki coupling reaction. The typical one-pot synthetic proce-
dure developed for heteroleptic polypyridyl ruthenium
complexes was employed for the preparation of the new sensi-
tizers at a good yield.14a The detailed synthesis procedure for the
new dyes and their identication with proton NMR (Fig. S1 and
S2†), HRMS (Fig. S3 and S4†), electronic absorption and emis-
sion spectroscopy (Fig. 1) are described in the Experimental
section, which are found to be in good agreement with their
structures.
The photo-physical and electrochemical properties of the
KW1, KW2, and Z907 dyes were investigated. The results are
summarized in Table 1. Three absorption bands were observed
in the 300–800 nm range both for the KW1 and KW2 dyes
measured in a DMF solution (Fig. 1a, le ordinate). The intense
absorption band around 300 nm is attributed to the intra-ligand
Fig. 1 (a) Electronic absorption spectra (left coordinate) of the KW1,
KW2, and Z907 sensitizers dissolved in DMF and the emission spectra
(right coordinate) of KW1 and KW2. (b) Electronic absorption spectra of
the KW1, KW2, and Z907 anchored on a 2.3 mm-thick mesoporous
titania film.
Table 1 Absorption wavelength, fluorescence maxima and redox potentials of various ruthenium complexes (KW1, KW2, and Z907) in DMF
labs (3 ꢀ 103 Mꢁ1 cmꢁ1) [nm]
Emissiona [nm]
Eoxb [V] vs. NHE
E0–0 [eV]
E0–0*d [eV]
a
c
Dye
KW1
KW2
Z907
310 (56.2), 399 (41.7), 538 (18.4)
307 (60.1), 392 (56.4), 554 (24.3)
298 (50.1), 372 (11.9), 520 (12.0)
750
785
769e
0.84
0.90
0.74e
1.86
1.78
—
ꢁ1.02
ꢁ0.88
—
a
b
Absorption and emission data were measured in DMF. Recorded in DMF/0.1 M TBAP/N2GC working and Pt counter electrodes, Ag/AgCl
reference electrode, scan rate ¼ 50 mV sꢁ1
.
The band-gap, E0–0, was estimated from the intersection wavelengths of the normalized UV-vis
c
d
e
absorption and the uorescence spectra. Excited state oxidation potential was calculated according to the equation Eox ꢁ E0–0
.
Z907 was
synthesized according to ref. 14a.
4948 | J. Mater. Chem. A, 2014, 2, 4945–4953
This journal is © The Royal Society of Chemistry 2014

View Article Online
Paper
Journal of Materials Chemistry A
The excitation of the low-energy MLCT transitions of the Z907 sensitizer. These cells employed a volatile acetonitrile-
KW1 or KW2 sensitizer in a DMF solvent produces an emission based electrolyte, containing 1.0 M 1,3-dimethylimidazolium
maximum located at 750 or 785 nm (as shown in Fig. 1a, right iodide (DMII), 50 mM LiI, 30 mM I2, 0.5 M tert-butylpyridine,
ordinate). Based on the absorption and emission spectra, the and 0.1 M guanidinium thiocyanate (GNCS) in the mixed
excitation transition energies (E0–0) of the KW1 and KW2 dyes solvent of acetonitrile and valeronitrile (v/v, 85/15).14c Typical
were evaluated to be 1.86 and 1.78 eV, respectively. Cyclic vol- values of the photovoltaic parameters open circuit voltage (VOC),
tammetry measurements were performed in a DMF solution short circuit photocurrent density (JSC) and ll factor (FF) of all
using 0.1 M tetrabutylammonium hexauorophosphate as the three devices measured under 1 sun irradiation conditions (the
supporting electrolyte. The redox potentials [40 (S+/S)] of the KW1 dye for device A, the KW2 dye for device B and the Z907 dye
KW1 and KW2 sensitizers were estimated to be 0.84 and 0.90 V for device C) are summarized in Table 2. Encouragingly, device
(vs. NHE, see Table 1). These are 0.34 and 0.4 V higher than that A (KW1 with TPA-substituted bipyridines) exhibited a VOC of
of the iodide electron-donor (ꢂ0.5 V vs. NHE), providing suffi- 0.764 V, a JSC of 8.89 mA cmꢁ2, and a FF of 0.73, giving an overall
cient driving force for dye regeneration. Furthermore, their PCE of about 5% when a 2.3 mm-thick TiO2 lm was used. Even
excited state formal redox potentials [40 (S+/S*)] were calculated with such a thin transparent titania lm (2.3 mm), device B (the
to be ꢁ1.02 and ꢁ0.88 V (vs. NHE). The negative offset of [40 (S+/ KW2 dye with TPA–thiophene-substituted bipyridines) exhibi-
S*)] relative to the conduction band edge of TiO2 ensures the ted a signicant increase in the short circuit current density (JSC
thermodynamic driving force for charge generation. The excited ¼ 11.24 mA cmꢁ2), yielding an overall PCE of 6.3% under illu-
state potentials of the two dyes obtained in DMF are more mination with standard AM 1.5G simulated sunlight (100 mW
negative than the conduction band of TiO2, beneting electron cmꢁ2). Under the same conditions, the JSC of device C (with the
injection from these excited states to TiO2.
Z907 dye) was about 7.44 mA cmꢁ2 and the PCE of device C
Frontier molecular orbitals for KW1 and KW2 were calcu- (ꢂ4%) was less than two-thirds of device B. These results clearly
lated via the density functional theory (DFT) method in DMF at revealed the advantages of the ruthenium dyes with high molar
the cam-B3LYP/lanl2dz level with the Gaussian 09 package.22 extinction coefficients for DSSC application.8d,e This observation
The HOMO and LUMO proles of these two dyes are displayed most probably stems from the superior molar extinction coef-
in Fig. 2. The HOMO of Z907 has an essentially ruthenium t2g cient and the red shi in the absorption spectrum of the KW2
character with sizable contribution coming from the NCS dye relative to the Z907 dye, which can be further veried from
groups.23 However, the HOMO of KW1 is delocalized partially the corresponding IPCE spectra as exhibited in Fig. S5.†
over the NCS groups and TPA unit. In contrast, the HOMO of
ꢁ
The competitive light absorption of triiodide from the Iꢁ/I3
KW2 is located mainly on the TPA and thiophene moieties, redox couple could decrease the IPCE spectrum of a DSSC
resulting in an increased separation of the HOMO from the TiO2 device at around 400 nm.3d In this work, there are strong
surface compared to Z907 and KW1. Apparently, the thiophene absorption bands for the donor antenna dyes with high molar
moiety can signicantly tune the HOMO due to the enhanced extinction coefficients of about 44 ꢀ 103 Mꢁ1 cmꢁ1 and 57 ꢀ 103
electronic delocalization.
Mꢁ1 cmꢁ1 at 397 nm for KW1 and 434 nm for KW2, which are
Some preliminary photovoltaic experiments were conducted signicantly higher than those of Z907 dye (3 ¼ 10.9 ꢀ 103 Mꢁ1
ꢁ
to evaluate the performance of the new dyes employing trans- cmꢁ1) and I3 (3 ¼ 25 ꢀ 103 Mꢁ1 cmꢁ1) at this wavelength.3d
A
parent and double layer titania lms with different thicknesses. large molar extinction coefficient in this regiꢁon will offset the
Initially, the transparent TiO2 thin lms (with the thickness of competitive visible light absorption due to I3 and recover the
2.3 mm) were used to fabricate DSSCs, taking the advantage of dip in the IPCE spectra around 400 nm. Actually, we found an
the high optical cross-section of the new dyes compared to the obvious peak in this region in the IPCE spectra for KW1 and
KW2-sensitized devices utilizing a thin lm (with a thickness of
2.3 mm, Fig. S5†). It is noteworthy that devices A and B have an
intensely increased IPCE response in the range of 400–450 nm.
Table 2 Detailed photovoltaic parameters of DSSC devices with the
KW1, KW2, and Z907 sensitizers of varying film thicknesses in the
presence and absence of PPA as the co-adsorbent under full sunlight
intensity
TiO2
JSC
Device
thickness [mm] VOC [V] [mA cmꢁ2
]
FF
PCE [%]
A (KW1)
B (KW2)
C (Z907)
D (KW1)
E (KW2)
F (Z907)
2.3
2.3
2.3
8.5 + 5
8.5 + 5
8.5 + 5
0.764
0.755
0.758
0.740
0.725
0.730
0.767
0.747
8.89
11.24
7.44
18.47
18.90
16.72
18.63
19.40
0.73 4.98
0.74 6.30
0.71 4.05
0.74 10.08
0.74 10.07
0.72 8.88
0.74 10.64
0.74 10.73
G (KW1 + PPA) 8.5 + 5
H (KW2 + PPA) 8.5 + 5
Fig. 2 The HOMO and LUMO for the KW1 and KW2 dyes from the DFT
calculations (isodensity contour ¼ 0.02 a.u.).
This journal is © The Royal Society of Chemistry 2014
J. Mater. Chem. A, 2014, 2, 4945–4953 | 4949

View Article Online
Journal of Materials Chemistry A
Paper
The TiO2 lm thickness plays a vital role in the DSSC device broad plateau of ꢂ90% in the region around 600 nm. We can
performance. To further explore the potential of these new dyes, conclude that the substitution of donor antennas for long alkyl
solar cells were fabricated by using a double layer TiO2 lm chains improves the absorbed photon-to-current conversion
(8.5 mm ꢀ 20 nm-sized TiO2 transparent lm with 5 mm ꢀ efficiency and thus, its capacity to develop photon-generated
400 nm-sized TiO2 scattering layer) in conjunction with a vola- charge.
tile electrolyte (see the Experimental section) and with or
The photovoltage of a DSSC is intrinsically determined by
without a co-adsorbent (3-phenylpropionic acid, PPA). Fig. 3a the potential difference between the quasi-Fermi level of TiO2
presents the photocurrent density–voltage curve (J–V) charac- and the redox potential of the electrolyte.25 The VOC can be
teristics of the devices based on these new donor antenna affected by a shi of the TiO2 conduction band edge26 as well
sensitizers and Z907 dye, using the above-mentioned volatile as the degree of electron recombination reaction.27 Unraveling
electrolyte under illumination with air mass (AM) 1.5 sunlight of the details of the electron recombination dynamics between
(100 mW cmꢁ2). The photovoltaic parameters of various devices the photo-injected electrons at the TiO2 and the oxidized
are tabulated in Table 2. The device D (with the KW1 dye) had a electrolyte in various devices was undertaken by employing
JSC of 18.47 mA cmꢁ2, a VOC of 0.74 V, and a FF of 0.74, yielding transient photovoltage measurements and charge-extraction
an overall conversion efficiency of 10.08%. For the cell (device E) measurements. Fig. 4a illustrates the plot of the extracted
with the KW2 dye the corresponding device parameters (JSC
,
charge density versus open circuit voltage for devices D, E, and
VOC, FF, and PCE) were 18.90 mA cmꢁ2, 0.725 V, 0.74, and F. The charge density was chosen as the abscissa in order to
10.07%, respectively. Here we observed that for the device E compare the VOC values obtained from the three photoanodes
with the KW2 dye the improvement in the short circuit current at the same electron concentration in the dye-sensitized TiO2
was accompanied by a lower open circuit voltage compared to mesoporous lms. The charge density for various dyes
KW1-sensitized device D. A reduced interfacial charge recom- increased exponentially along with the increase of VOC, which
bination in device E was also evident from the dark-current data was generated by applying a gradually enhanced light intensity.
for these devices (Fig. S6†), which addressed the voltage This change in the VOC value can be explained qualitatively by a
dependence of the recombination between the electrons in the shi in the conduction band edge (Ecb) with respect to the
TiO2 and the electrolyte. The reduction in the dark-current in quasi-redox potential of the electrolytes. We observed that the
device D relative to device E is consistent with the recombina- conduction band edge of the TiO2 underwent an upward
tion dynamics as discussed below. In the presence of PPA as a displacement in devices compared to that of Z907. These
co-adsorbent (1 : 1 molar ratio in the dye staining solution), energy shis principally inuence the devices' open circuit
both the JSC and VOC were increased for devices G and H (see voltage.28 At an identical photo-induced charge density (1 ꢀ
Table 2), and conversion efficiency values of 10.64% and 10.73% 1018 cmꢁ3), the photovoltage of device D (sensitized with the
were achieved, respectively. This result can be attributed to the KW1 dye) was approximately 36 mV higher than that of F
fact that PPA was co-graed with the dye molecules onto the (sensitized with the Z907 dye). It indicated that addition of TPA
surface of TiO2 nanocrystals. The mixed dye–PPA monolayer into the ancillary ligand shied the TiO2 conduction band edge
appears to be more robust to impair the escape of photo- upward by about 36 mV, neglecting the change in the Nernst
injected electrons into electrolytes.24
potential of the electrolyte. Fig. 4b shows the recombination
The IPCE spectra of devices D (with the KW1 dye), E (with lifetime for various devices against the charge density. At a
the KW2 dye) and F (with the Z907 dye) are illustrated in given charge density, the device F had the longest lifetime. This
Fig. 3b. The broad bands covered almost the entire visible result indicated that the charge recombination was faster for
spectrum from 450 nm to 700 nm with the maxima of 90%, devices D and E than for device F. Though the incorporation of
91% and 86% for devices D, E, and F, respectively. These results
indicated that the increased JSC values for device D and device E
stemmed from the higher IPCE response, which attained a
Fig. 4 Transient photovoltage decay measurements of the device D
(with the KW1 dye), device E (with the KW2 dye), and device F (with the
Z907 dye): (a) the relationship between open circuit voltage and the
Fig. 3 (a) J–V characterization for DSSC device D (with the KW1 dye), photo-induced charge density and (b) the relationship between
device E (with the KW2 dye), device F (with the Z907 dye), device G recombination lifetime and the photo-induced charge density; and (c)
(with the KW1 and PPA), and device H (with the KW2 and PPA) under electron transport resistance Rt of the TiO2 film obtained from
AM 1.5G full sunlight intensity. (b) The photocurrent action spectra of impedance measurement in the dark at 20 ꢃC as a function of applied
DSSC devices D, E, and F, respectively.
forward bias voltage for the devices.
4950 | J. Mater. Chem. A, 2014, 2, 4945–4953
This journal is © The Royal Society of Chemistry 2014

View Article Online
Paper
Journal of Materials Chemistry A
TPA– and TPA–thiophene-substituted bipyridines as the ancil-
lary ligands increased the interfacial recombination as indi-
cated in devices D and E, the effect was compensated by an
increase in the conduction band edge of the TiO2 of these
devices. These results indicated that different donor antennas
incorporated in the ruthenium sensitizers had an effect on the
conduction band edge of TiO2 as well as the recombination
process at the same time.
Electronic impedance measurements (EIS) in the dark were
further performed to understand the differences in photovoltaic
behavior of devices with various dye sensitizers. The impedance
spectra for the DSSC devices (D, E, and F) revealed comparable
characteristic information about the various kinetic processes.
As presented in Fig. 4c, the logarithm of the Rt shows parallel
behavior for the three devices, which implies that the shi of
the resistance for the steady state electron transport in these
devices is caused by a change in position of the conduction
band edge (Ecb).14c Presuming that electron mobility is the same
among various devices, the TiO2 conduction band edge energy
level (Ecb) relative to the quasi-equilibrium potential for the
redox couple can be determined by linear extrapolation of
the experimental points in Fig. 4c to their intersection with the
abscissa. The Rt data from device D (sensitized with KW1) were
shied upward (ꢂ38 mV) from those of device F (Z907 used as a
sensitizer). Since the conduction band edge of the photoanode
(dye coated TiO2 lm) is independent of the redox potential of
electrolytes, the shi of the potentials in devices D and F can be
attributed to a band edge modication14c induced by changing
the molecular dipoles.29 Clearly, the trend of the electron
transport resistance of the various devices is well in agreement
with that of the above measured transient photovoltage decay
values.
As discussed in the absorption spectroscopy, impedance and
transient photovoltage investigation, these new donor antenna
sensitizers with different TPA-donor moieties have a benecial
inuence on the conduction band edge energy levels of TiO2
and light harvesting capability for the improvement of photo-
voltage and photocurrent. However, the electron recombination
increases along with the large dimension of donor antenna
moieties compared to that bearing long alkyl chains. Thus, a
high photovoltage would be expected in these devices with
donor antenna sensitizers if an efficient modication on the
Fig. 5 Evolution of the photovoltaic parameters (JSC, VOC, FF and PCE)
for the device based on an (8.5 + 5) mm TiO2 double layer film sensi-
tized with KW2 in the presence of PPA (1 : 1 molar ratio) as a co-
adsorbent and low volatility of a liquid electrolyte (with an MPN
solvent) during the visible light soaking (100 mW cmꢁ2) at 60 ꢃC.
photovoltaic parameters, i.e., the VOC, FF, JSC and PCE of the
fresh low volatility electrolyte based device with the KW2 dye
were 0.69 V, 0.74, 18.4 mA cmꢁ2, and 9.29%, respectively. Over
the 1000 hour light soaking test period, the photovoltaic
parameters JSC, VOC, FF, and PCE of the cell varied only slightly
from the initial values. During the aging process, the value of
the efficiency retained 95% of its initial value aer 1000 h light
soaking, a small drop of 46 mV in VOC being compensated by
an increase of 0.6 mA cmꢁ2 in the JSC values. Such an
impressively stable performance implies the robustness of the
KW2 dye itself as well as the dye-sensitized heterojunction
interface.
Conclusions
interface of TiO2–dye was adopted. Actually, when a co-adsor- In summary, we have designed and synthesized two new het-
bent (PPA) was added, an augment in photovoltage was obvi- eroleptic ruthenium sensitizers with a feature of incorporating
ously observed, showing high PCEs close to 11% (see Table 2). high electron delocalization TPA-donor antennas into the
The KW2-sensitized device H provides a JSC of 19.40 mA cmꢁ2, a ancillary ligands. These new dyes exhibit remarkable light
VOC of 0.744 V and a FF of 0.74, yielding an overall PCE of harvesting capacities and the corresponding DSSC cells achieve
10.73%.
impressive power conversion efficiencies close to 11% under
Low volatility electrolyte (detailed compositions given in AM 1.5G simulated sunlight. The different spacer between
the Experimental section) based devices were used to evaluate bipyridyl and TPA-donor clearly showed an inuence on the
the preliminary stability of the new sensitizers under photo-physical and electrochemical properties of this class of
moderate thermal stress and visible light soaking. The devices ruthenium complexes as well as their interface behaviors on the
with the KW1 and KW2 sensitizers showed similar behaviour TiO2, which were scrutinized by impedance and transient
during the stability testing. Fig. 5 presents the representative photovoltage decay measurements. Based on these studies and
variation in the photovoltaic parameters of KW2-sensitized ndings, the photo-activity of ruthenium sensitizers containing
cells when subjected to the acceleration test in a solar simu- donor antennas can be successfully improved by careful
lator at full sunlight (100 mW cmꢁ2) and 60 ꢃC. The molecular engineering.
This journal is © The Royal Society of Chemistry 2014
J. Mater. Chem. A, 2014, 2, 4945–4953 | 4951

View Article Online
Journal of Materials Chemistry A
Paper
K. Liao, G. Yang, C. Li, J. Kim and J. Ko, J. Phys. Chem. C,
2011, 115, 7747.
Acknowledgements
10 W. Han, J. Han, H. Kim, M. J. Choi, Y. Kang, C. Pac and
S. O. Kang, Inorg. Chem., 2011, 50, 3271.
11 (a) S. Haque, S. Handa, K. Peter, E. Palomares, M. Thelakkat
and J. Durrant, Angew. Chem., Int. Ed., 2005, 44, 5740; (b)
N. Hirata, J. Lagref, E. Palomares, J. Durrant,
Financial support from the 973 Program of China
(2014CB643506, 2013CB922104, and 2011CBA00703), NSFC
(201173091, 21103578, and 21161160445), the Fundamental
Research Funds for the Central Universities (HUST:
CXY12Q022, and 2013YQ051), and the CME with the Program of
New Century Excellent Talents in University (NCET-10-0416), is
gratefully acknowledged. The authors thank the Analytical
and Testing Centre at the HUST for performing characterization
of various samples. Y.S. and M.W. thank the Carl von
Ossietzky University of Oldenburg for the support as visiting
scientist.
¨
M. Nazeeruddin, M. Gratzel and D. Censo, Chem. – Eur. J.,
2004, 10, 595.
12 (a) C. Karthikeyan, H. Wietasch and M. Thelakkat, Adv.
Mater., 2007, 19, 1091; (b) H. Snaith, C. Karthikeyan,
A. Petrozza, J. Teuscher, J. Moser, M. Nazeeruddin,
¨
M. Thelakkat and M. Gratzel, J. Phys. Chem. C, 2008, 112,
7562; (c) J. Yum, S. Moon, C. Karthikeyan, H. Wietasch,
M. Thelakkat, S. Zakeeruddin, M. Nazeeruddin and
Notes and references
¨
M. Gratzel, Nano Energy, 2012, 1, 6.
1 G. Vougioukalakis, A. Philippopoulos, T. Stergiopoulos and 13 K. Robson, B. Koivisto, T. Gordon, T. Baumgartner and
P. Falaras, Coord. Chem. Rev., 2011, 255, 2602.
C. Berlinguette, Inorg. Chem., 2010, 49, 5335.
2 (a) J. Yina, M. Velayudham, D. Bhattacharya, H. Lin and 14 (a) M. Nazeeruddin, S. Zakeeruddin, J. Lagref, P. Liska,
¨
K. Lu, Coord. Chem. Rev., 2012, 256, 3008; (b) J. Clifford,
M. Planells and E. Palomares, J. Mater. Chem., 2012, 22,
24195; (c) L. Li and E. W. Diau, Chem. Soc. Rev., 2013, 42,
291; (d) M. Liang and J. Chen, Chem. Soc. Rev., 2013, 42, 3453.
P. Comte, C. Barolo, G. Viscardi, K. Schenk and M. Gratzel,
Coord. Chem. Rev., 2004, 248, 1317; (b) P. Wang,
S. Zakeeruddin, J. Moser, M. Nazeeruddin, T. Sekiguchi
and M. Gratzel, Nat. Mater., 2003, 2, 402; (c) M. Wang,
¨
¨
3 (a) O. Kohle, M. Gratzel, A. Meyer and T. Meyer, Adv. Mater.,
P. Chen, R. Humphry-Baker, S. Zakeeruddin and
¨
M. Gratzel, ChemPhysChem, 2009, 10, 290.
1997, 9, 904; (b) F. Gao, Y. Wang, D. Shi, J. Zhang, M. Wang,
X. Jing, R. Humphry-Baker, P. Wang, S. Zakeeruddin and 15 H. Staats, F. Eggers, O. Haß, F. Fahrenkrug, J. Matthey,
¨
¨ ¨
U. Luning and A. Lutzen, Eur. J. Org. Chem., 2009, 2009, 4777.
M. Gratzel, J. Am. Chem. Soc., 2008, 130, 10720; (c) Y. Cao,
Y. Bai, Q. Yu, Y. Cheng, S. Liu, D. Shi, F. Gao and P. Wang, 16 B. Liang, L. Wang, Y. Xu, H. Shi and Y. Cao, Adv. Funct.
J. Phys. Chem. C, 2009, 113, 6290; (d) L. Han, A. Islam,
Mater., 2007, 17, 3580.
H. Chen, C. Malapaka, B. Chiranjeevi, S. Zhang, X. Yang 17 S. Li, X. Yang, M. Cheng, J. Zhao, Y. Wang and L. Sun,
and M. Yanagida, Energy Environ. Sci., 2012, 5, 6057. Tetrahedron, 2012, 53, 3425.
4 C. Chen, M. Wang, J. Li, N. Pootrakulchote, L. Alibabaei, 18 J. Lu, X. Xu, Z. Li, K. Cao, J. Cui, Y. Zhang, Y. Shen, Y. Li,
¨
C. Ngocle, J. Decoppet, J. Tsai, C. Gratzel, C. Wu,
J. Zhu, S. Dai, W. Chen, Y. Cheng and M. Wang, Chem. –
Asian J., 2013, 8, 956.
¨
S. Zakeeruddin and M. Gratzel, ACS Nano, 2009, 3, 3103.
5 A. Hagfeldt, G. Boschloo, L. Sun, L. Kloo and H. Pettersson, 19 (a) S. Zakeeruddin, M. Nazeeruddin, R. Humphry-Baker,
Chem. Rev., 2010, 110, 6595.
P. Pechy, P. Quagliotto, C. Barolo, G. Viscard and
¨
6 (a) K. Wu, W. Ku, J. N. Clifford, E. Palomares, S. Ho, Y. Chi,
M. Gratzel, Langmuir, 2002, 18, 952; (b) P. Wang,
¨
S. Liu, P. Chou, M. K. Nazeeruddin and M. Gratzel, Energy
S. Zakeeruddin, J. Moser, R. Humphry-Baker, P. Comte,
¨
Environ. Sci., 2013, 6, 859; (b) S. Wang, K. Wu, E. Ghadiri,
V. Aranyos, A. Hagfeldt, M. Nazeeruddin and M. Gratzel,
M. G. Lobello, S. Ho, Y. Chi, J. Moser, F. D. Angelis,
Adv. Mater., 2004, 16, 1806.
¨
M. Gratzel and M. K. Nazeeruddin, Chem. Sci., 2013, 4, 423. 20 Q. Yu, S. Liu, M. Zhang, N. Cai, Y. Wang and P. Wang, J. Phys.
7 P. Wang, S. M. Zakeeruddin, J. E. Moser, R. Humphry-Baker,
Chem. C, 2009, 113, 14559.
P. Comte, V. Aranyos, A. Hagfeldt, M. K. Nazeeruddin and 21 F. Dai, W. Wu, Q. Wang, H. Tian and W. Wong, Dalton
¨
M. Gratzel, Adv. Mater., 2004, 16, 1806.
Trans., 2011, 40, 2314.
8 (a) C. Chen, N. Pootrakulchote, S. Wu, M. Wang, J. Li, J. Tsai, 22 M. Frisch, G. Trucks, H. Schlegel, G. Scuseria, M. Robb,
¨
C. Wu, S. M. Zakeeruddin and M. Gratzel, J. Phys. Chem. C,
J. Cheeseman, G. Scalmani, V. Barone, B. Mennucci,
G. Petersson, H. Nakatsuji, M. Caricato, X. Li,
H. Hratchian, A. Izmaylov, J. Bloino, G. Zheng,
J. Sonnenberg, M. Hada, M. Ehara, K. Toyota, R. Fukuda,
J. Hasegawa, M. Ishida, T. Nakajima, Y. Honda, O. Kitao,
H. Nakai, T. Vreven, J. Montgomery, J. Peralta, F. Ogliaro,
M. Bearpark, J. Heyd, E. Brothers, K. Kudin, V. Staroverov,
R. Kobayashi, J. Normand, K. Raghavachari, A. Rendell,
J. Burant, S. Iyengar, J. Tomasi, M. Cossi, N. Rega,
J. Millam, M. Klene, J. Knox, J. Cross, V. Bakken,
C. Adamo, J. Jaramillo, R. Gomperts, R. Stratmann,
O. Yazyev, A. Austin, R. Cammi, C. Pomelli, J. Ochterski,
2009, 113, 20752; (b) J. Yum, I. Jung, C. Baik, J. Ko,
M. Nazeeruddin and M. Gratzel, Energy Environ. Sci., 2009,
¨
2, 100; (c) K. Willinger, K. Fischer, R. Kisselev and
M. Thelakkat, J. Mater. Chem., 2009, 19, 5364; (d) M. Wang,
S. Moon, D. Zhou, F. Formal, N. Cevey-Ha, R. Humphry-
Baker, C. Gratzel, P. Wang, S. Zakeeruddin and M. Gratzel,
Adv. Funct. Mater., 2010, 20, 1821; (e) A. Mishra,
N. Pootrakulchote, M. Wang, S. Moon, S. Zakeeruddin,
M. Gratzel and P. Bauerle, Adv. Funct. Mater., 2011, 21, 963.
9 (a) C. Chen, J. Chen, S. Wu, J. Li, C. Wu and K. Ho, Angew.
Chem., Int. Ed., 2008, 47, 7342; (b) S. Fan, C. Kim, B. Fang,
¨
¨
¨
¨
4952 | J. Mater. Chem. A, 2014, 2, 4945–4953
This journal is © The Royal Society of Chemistry 2014

View Article Online
Paper
R. Martin, K. Morokuma, V. Zakrzewski, G. Voth, P. Salvador,
Journal of Materials Chemistry A
Chem. Soc., 2008, 130, 13558; (b) B. O'Regan, K. Walley,
M. Juozapavicius, A. Anderson, F. Mata, T. Ghaddar,
S. Zakeeruddin, C. Klein and J. Durrant, J. Am. Chem. Soc.,
2009, 131, 3541.
J. Dannenberg, S. Dapprich, A. Daniels, O. Farkas,
J. Foresman, J. Ortiz, J. Cioslowski and D. Fox, Gaussian
09, Revision A.01, Gaussian Inc., Wallingford CT, 2009.
23 S. Handa, H. Wietasch, M. Thelakkat, J. Durrant and 28 (a) E. Barea, J. Ortiz, F. Paya, F. Fernandez-Lazaro,
S. Haque, Chem. Commun., 2007, 1725.
24 (a) P. Wang, S. Zakeeruddin, R. Humphry-Baker, J. Moser
F. Fabregat-Santiago, A. Sastre-Santos and J. Bisquert,
Energy Environ. Sci., 2010, 3, 1985; (b) A. Allegrucci,
N. Lewcenko, A. Mozer, L. Dennany, P. Wagner, D. Officer,
K. Sunahara, S. Mori and L. Spiccia, Energy Environ. Sci.,
2009, 2, 1069.
¨
and M. Gratzel, Adv. Mater., 2003, 15, 2101; (b) P. Wang,
¨
S. M. Zakeeruddin, R. Humphry-Baker and M. Gratzel,
Chem. Mater., 2004, 16, 2694.
¨
25 D. Hagberg, J. Yum, H. Lee, F. Angelis, T. Marinado, K. Karlsson, 29 (a) M. Wang, C. Gratzel, S. Moon, R. Humphry-Baker,
¨
¨
R. Humphry-Baker, L. Sun, A. Hagfeldt, M. Gratzel and
N. Rossier-Iten, S. M. Zakeeruddin and M. Gratzel, Adv.
M. Nazeeruddin, J. Am. Chem. Soc., 2008, 130, 6259.
Funct. Mater., 2009, 19, 2163; (b) P. Chen, J. Yum,
26 Z. Zhang, S. Zakeeruddin, B. O'Regan, R. Humphry-Baker
F. Angelis, E. Mosconi, S. Fantacci, S. Moon, R. Baker,
¨
¨
J. Ko, M. Nazeeruddin and M. Gratzel, Nano Lett., 2009, 9,
and M. Gratzel, J. Phys. Chem. B, 2005, 109, 21818.
´
¨
¨
27 (a) A. Reynal, A. Forneli, E. Martinez-Ferrero, A. Sanchez-
2487; (c) J. Kruger, U. Bach and M. Gratzel, Adv. Mater.,
´
Dıaz, A. Vidal-Ferran, B. O'Regan and E. Palomares, J. Am.
2000, 12, 447.
This journal is © The Royal Society of Chemistry 2014
J. Mater. Chem. A, 2014, 2, 4945–4953 | 4953