Organic redox couples and organic counter electrode for efficient organic dye-sensitized solar cells

Article abstract of DOI:10.1021/ja2030933

A series of organic thiolate/disulfide redox couples have been synthesized and have been studied systematically in dye-sensitized solar cells (DSCs) on the basis of an organic dye (TH305). Photophysical, photoelectrochemical, and photovoltaic measurements were performed in order to get insights into the effects of different redox couples on the performance of DSCs. The polymeric, organic poly(3,4-ethylenedioxythiophene) (PEDOT) material has also been introduced as counter electrode in this kind of noniodine-containing DSCs showing a promising conversion efficiency of 6.0% under AM 1.5G, 100 mW·cm^-2 light illumination. Detailed studies using electrochemical impedance spectroscopy and linear-sweep voltammetry reveal that the reduction of disulfide species is more efficient on the PEDOT counter electrode surface than on the commonly used platinized conducting glass electrode. Both pure and solvated ionic-liquid electrolytes based on a thiolate anion have been studied in the DSCs. The pure and solvated ionic-liquid-based electrolytes containing an organic redox couple render efficiencies of 3.4% and 1.2% under 10 mW·cm^-2 light illumination, respectively.

Full text of DOI:10.1021/ja2030933

ARTICLE
pubs.acs.org/JACS
Organic Redox Couples and Organic Counter Electrode for Efficient
Organic Dye-Sensitized Solar Cells
Haining Tian,† Ze Yu,‡ Anders Hagfeldt,§,^ Lars Kloo,*,‡,^ and Licheng Sun*,†,^
†Organic Chemistry, Center of Molecular Devices, Department of Chemistry, School of Chemical Science and Engineering,
Royal Institute of Technology (KTH), SE-10044 Stockholm, Sweden
‡Inorganic Chemistry, Center of Molecular Devices, Department of Chemistry, School of Chemical Science and Engineering,
Royal Institute of Technology (KTH), SE-10044 Stockholm, Sweden
§Department of Physical and Analytical Chemistry, Uppsala University, SE-75105 Uppsala, Sweden
^State Key Laboratory of Fine Chemicals, DUT-KTH Joint Education and Research Center for Molecular Devices, Dalian University of
Technology (DUT), 116012 Dalian, China
S Supporting Information
b
ABSTRACT: A series of organic thiolate/disulfide redox
couples have been synthesized and have been studied system-
atically in dye-sensitized solar cells (DSCs) on the basis of an
organic dye (TH305). Photophysical, photoelectrochemical,
and photovoltaic measurements were performed in order to
get insights into the effects of different redox couples on the
performance of DSCs. The polymeric, organic poly(3,4-
ethylenedioxythiophene) (PEDOT) material has also been
introduced as counter electrode in this kind of noniodine-
containing DSCs showing a promising conversion efficiency
of 6.0% under AM 1.5G, 100 mW cmꢀ2 light illumination. Detailed studies using electrochemical impedance spectroscopy and
3
linear-sweep voltammetry reveal that the reduction of disulfide species is more efficient on the PEDOT counter electrode surface
than on the commonly used platinized conducting glass electrode. Both pure and solvated ionic-liquid electrolytes based on a
thiolate anion have been studied in the DSCs. The pure and solvated ionic-liquid-based electrolytes containing an organic redox
couple render efficiencies of 3.4% and 1.2% under 10 mW cmꢀ2 light illumination, respectively.
3
’ INTRODUCTION
illumination conditions (100 mW cmꢀ2).3 For DSCs fabricated
3
with an organic dye (TH30535), our group has adopted a sulfide-
based redox couple derived from mercapto-5-methyl-1,3,4-thia-
diazole (McMT), which showed a 4.0% efficiency.26 On the other
hand, Li et al. applied tetramethylthiourea derivatives to N3 dye-
based DSCs and obtained 3.1% efficiency using a carbon counter
electrode.25 Liu et al. used the same organic redox couple in
organic D131 dye-based DSCs, which shows 3.9% efficiency on
the basis of a carbon CE under standard light illumination.32 Just
as for the photosensitizers, structural modification may finetune
the physical and electrochemical properties of the thiolate/
disulfide redox couples, which significantly can affect the photo-
voltaic properties of DSCs. Also, according to a previous study,
the fill factor (ff) of the DSCs based on this thiolate/disulfide
electrolyte is very low because of a large reduction resistance of
the disulfide on the standard platinized CE leading to a decrease in
the overall conversion efficiency of the DSCs.26 This implies that
platinum is not a suitable catalytic material for thiolate/disulfide
electrolyte reduction at the CE. Thus, a more effective counter
electrode material is required. Although some efforts have been
Dye-sensitized solar cells (DSCs), significantly improved by
O’Regan and Gr€atzel in 1991, represent a simple, efficient, and
economical photovoltaic device based on a new strategy for solar
energy-to-electricity conversion.1 Like other electrochemical de-
vices, the DSCs consist of three main components: a photoactive
working electrode (WE), a counter electrode (CE), and a redox-
active electrolyte to connect the electrodes. In recent years, many
attempts have been made to optimize this device from essentiallꢀy
every possible aspect.2 The replacement of the ubiquitous Iꢀ/I3
electrolyte redox couple would be a critical improvement because
of its disadvantages in terms of corrosion of metal-based current
collectors (especially for silver), evaporation losses, and visible
light absorption.3 Until now, three main kinds of iodine-free
liquid electrolytes, based on metal complexes,4ꢀ13 inorganic
materials,14ꢀ22 or organic redox couples,3,7,23ꢀ32 have been
developed for the DSCs. One of the organic redox couples, the
thiolate/disulfide system with multielectron-transfer capability,
has shown potential applications in lithium-ion batteries as well as
in DSC electrolytes.3,25,26,33,34 Recently, Wang et al. reported an
interesting thiolate/disulfide redox couple derived from 5-mercapto-
1-methyltetrazole, which showed an efficiency of 6.4% together
with a ruthenium-based dye (Z907Na) in DSCs under standard
Received: February 9, 2011
Published: May 18, 2011
r
2011 American Chemical Society
9413
dx.doi.org/10.1021/ja2030933 J. Am. Chem. Soc. 2011, 133, 9413–9422
|

Journal of the American Chemical Society
ARTICLE
Figure 1. Structures of the compounds used in the organic redox couples.
Figure 2. The general synthetic routes of the compounds used in the organic redox couples.
made to use a carbon-based electrode for the thiolate/disufide
redox couple, the ff is still not satisfactory.25,32 In this paper, we
report the successful development of a series of thiolate/disulfide
redox couples (see Figure 1) and further a systematic study of the
effect of redox couple structures on the photovoltaic properties of
DSCs on the basis of the TH305 organic dye. Furthermore, owing
to the poineering work on the poly(3,4-ethylenedioxythiophene)
(PEDOT) material for lithium-ion batteries andDSCs,30,33,34,36ꢀ46
we have also adopted the PEDOT material as CEs prepared by
electrochemical polymerization in organic dye-sensitized solar cells
on the basis of thiolate/disulfide electrolytes. Subsequently, we
have investigated the essential differences between platinized
conducting glass and PEDOT as substrates for disulfide reduction.
Meanwhile, pure imidazolium ionic-liquid electrolytes and solvated
ionic liquid based on thiolate anions have been prepared and
applied to DSCs and have been characterized by photophysical and
photovoltaic techniques.
obtained by neutralization of the corresponding mercaptan using
hydroxide compound, and the disulfide compounds were ob-
tained through oxidation of the corresponding mercaptan by
iodine in basic solution. Thiolates 1aꢀ and 1bꢀ are ionic liquids
at room temperature. The synthesis details of these redox
couples are presented in the Supporting Information. The
disulfide 2 is very sensitive to oxygen and can easily be further
oxidized to the corresponding sulfoxid. Therefore, the fresh
disulfide 2 has to be used in all measurements.
2. Different Redox Couples. The redox potential is one of the
most important parameters of redox couples in DSCs. To
determine the standard redox potentials of the organic redox
couples, cyclic voltammetric (CV) measurements were per-
formed using 0.2 M LiClO4 as supporting electrolyte at 20 ꢀC
at a scan rate of 50 mV sꢀ1. For the 1ꢀ/1, 2ꢀ/2, and 5ꢀ/5 redox
3
couples, the solution studied contained 20 mM of the reduced
component and 10 mM of the oxidized component in acetoni-
trile (MeCN). Because of the poor solubility of 3 and 4 in MeCN,
for the 3ꢀ/3 redox couple, the solution contained 20 mM of 3ꢀ
and was saturated with respect to 3 in MeCN. However, for the
4ꢀ/4 redox couple, the solution consisted of 20 mM of 4ꢀ and
was saturated with 4 in DMSO/DMF (6/4; v/v). The potential
’ RESULTS AND DISCUSSIONS
1. Synthesis. The synthetic routes of the organic redox
couples are depicted in Figure 2. All of the thiolate species were
9414
dx.doi.org/10.1021/ja2030933 |J. Am. Chem. Soc. 2011, 133, 9413–9422

Journal of the American Chemical Society
ARTICLE
Figure 3. UVꢀvis spectra of a thin layer of different electrolytes
confined between two conducting glass substrates sealed with 25 μm
Surlyn film.
versus the normal hydrogen electrode (NHE) was calibrated
using the reference Ferrocene/Ferrocenium (Fc/Fcþ) redox
couple by introducing a correction of 630 mV.48 From the CV
curves of these redox couples (Figure S1 of the Supporting
Information), the redox potentials of 1ꢀ/1, 2ꢀ/2, 3ꢀ/3, 4ꢀ/4,
and 5ꢀ/5 were determined to 0.15, 0.25, 0.37, 0.01, and 0.29 V
versus NHE, respectively. For the compounds with a similar
structure (1ꢀ/1 and 2ꢀ/2, 4ꢀ/4 and 5ꢀ/5, respectively), chan-
ging a S atom to O in the molecular framework makes the redox
potential more positive and, thus, the system more oxidizing.
Compared with 2ꢀ/2, 4ꢀ/4 containing extra phenyl groups also
gives a more positive redox potential probably because of the
stronger electron-withdrawing ability of the phenyl group as
compared to the methyl group in the 2ꢀ/2 redox couple.
Figure 4. The JꢀV curves (a) and IPCE spectra (b) of DSCs based on
different organic electrolytes.
The electrolytes A, B, C, D, and E were prepared from the 1ꢀ/
1, 2ꢀ/2, 3ꢀ/3, 4ꢀ/4, and 5ꢀ/5 redox couples in a mixed MeCN/
ethylene carbonate (EC) (6/4, v/v) solvent, respectively. The
electrolytes consisted of 0.2 M of the thiolate species, 0.2 M of
the disulfide species (1ꢀ/1 and 2ꢀ/2), or were saturated with
respect to the oxidized species (3 < 0.02 M, 4 < 0.05 M, and 5 <
0.2 M), 0.05 M LiClO4 and 0.5 M 4-t-butylpyridine (TBP).
Figure 3 shows the absorption sepctra of a thin layer of the
different organic redox couple based electrolytes compared to an
Table 1. Photovoltaic Properties of TH305-Based DSCs with
Different Organic Electrolytesa
electrolyte
Jsc [mA cmꢀ2
]
Voc [mV]
ff
ηb [%]
3
A
B
C
D
E
12.3
12.6
4.1
750
750
790
690
780
0.50
0.35
0.18
0.10
0.19
4.6
3.3
0.6
0.2
1.6
electrolyte containing Iꢀ/I3 redox couple (0.2 M TBAþIꢀ,
ꢀ
2.3
0.2 M I2, 0.05 M LiClO4, and 0.5 M TBP in MeCN/EC (6/4,
v/v)) confined between two conducting glass substrates sealed
with 25 μm Surlyn film. Thus, the absorption spectra can at least
be regarded as semiquantitative. All of the electrolytes containing
the organic redox couples essentially display no absorption in the
10.6
a The TiO2 film consists of 10 μm transparent layer and 2 μm
scattering layer. b Light intensity is AM 1.5G, 100 mW cmꢀ2
.
3
visible part of the spectrum in contrast to the Iꢀ/I3 based
ꢀ
TiO2, charge extraction and electron lifetime (see Figure 5)
measurements were performed. The quasi-Fermi level (EF,n) in
TiO2 was derived from the equation Voc = |EF,n ꢀ Eredox|, where
Voc is the open-circuit voltage of the DSC and Eredox is the redox
potential of the electrolyte. Electrolyte A gives the highest
photon-to-electricity conversion efficiency (η) of 4.6% with a
electrolyte.
For DSC fabrication, the organic dye TH305 was employed as
photosensitizer, and optimized platinized fluorine-doped tin
oxide (FTO) conducting glass (∼8.6 μg cmꢀ2) was used as
3
CE. The photocurrent densityꢀphotovoltage (JꢀV) curves
(Figure 4a) of DSCs based on different organic electrolytes were
short-current density (Jsc) of 12.3 mA cmꢀ2, an open-circuit
3
obtained under 100 mW cmꢀ2 simulated sunlight illumination,
voltage (Voc) of 750 mV, and a fill factor (ff) of 0.50. This is in
spite of the fact that electrolyte B, with the 2ꢀ/2 redox couple, is
characterized by a more positive redox potential, 0.25 V versus
NHE, than the 1ꢀ/1 redox couple, 0.15 V versus NHE. The
3
and the corresponding photovoltaic properties are collected in
Table 1. To further study the effects of different electrolytes on
the quasi-Fermi level of TiO2 (EF,n) and electron lifetime in
9415
dx.doi.org/10.1021/ja2030933 |J. Am. Chem. Soc. 2011, 133, 9413–9422

Journal of the American Chemical Society
ARTICLE
Figure 7. Light-intensity dependence of the current density under
short-circuit conditions for DSCs containing the different electrolytes.
show the same effect on the conduction band (CB) of the TiO2;
however, the DSCs containing electrolyte A render a much
longer electron lifetime than those containing electrolyte B at a
given Qoc value. This indicates that electron recombination losses
of injected photoelectrons to electrolyte B are much higher than
for electrolyte A. Notably, electrolyte B gives a much lower ff,
0.35, than electrolyte A, 0.50. To further investigate the differ-
ence between electrolyte A and electrolyte B in DSCs, electro-
chemical impedance spectroscopy (EIS) measurments were
performed under dark conditions and an applied bias voltage
of ꢀ0.7 V. The Nyquist plots of DSCs based on different
electrolytes are shown in Figure 6. From these results, we can
observe that DSCs containing the electrolyte A show consider-
ably lower charge-transfer resisitance at the CE (RCE), 320 Ω, as
compared to DSCs containing the electrolyte B with an RCE of
500 Ω. This may be the reason for the higher ff observed for the
electrolyte A. The results indicate that the component 1 more
easily reduced on the platinized CEs than the component 2.
Nevertheless, both RCE's are far too high to give competitive
conversion efficiencies, and the main reason can be attributed to
the fact that the Pt particles deposited on the FTO glass substrate
surface simply do not act as efficient catalysts for the reduction of
the disulfide electrolyte species in contrast to iodine/triiodide,
the iodine-based electrolyte systems. A general conclusion is that
other types of counter electrode materials adapted to the sulfur-
based organic redox systems must be identified. The light-
intensity dependence of the current density under short-circuit
conditions for DSCs with different electrolytes (Figure 7) again
identifies the electrolytes A and B as the best working ones
showing a linear relationship between DSC photocurrent and
light intensity. This implies that there is no serious mass-
transport problem in the electrolytes A and B. However, the
electrolytes C and D show a clear mass-transport problem that is
probably caused by their large molecules as well as by the low
concertration of oxidized state species in the electrolytes im-
posed by their low solubility.
Figure 5. Charge extraction (a) and lifetime (b) of DSCs based on
different electrolytes.
Figure 6. Nyquist plots obtained from DSCs containing the different
electrolytes under dark conditions and with an applied bias voltage of ꢀ0.7 V.
Although no such serious mass-transport problem was found
for electrolyte E, the lower ff lowered the overall efficiency of such
DSCs. The large RCE observed for the electrolytes CꢀE is the key
reason for the low ff of DSCs based on these electrolytes. The
electrolytes C and D shift the CB position of TiO2 to more
DSCs based on electrolyte B render about the same photovoltage
as those of electrolyte A. According to electron lifetime and
charge-extraction results, we can observe that both electrolytes
9416
dx.doi.org/10.1021/ja2030933 |J. Am. Chem. Soc. 2011, 133, 9413–9422

Journal of the American Chemical Society
ARTICLE
positive and negative values, respectively, probably because of
different effects on the surface chemistry of titania of these two
redox couples. From incident photon-to-current conversion
efficiency (IPCE) spectra (Figure 4b), the DSCs based on all
of these electrolytes show almost the same active range and
values revealing that these electrolytes give similar photocurrents
under very low light illumination. On the basis of these results,
further modification of this kind of redox couple should focus on
solving the mass-transport problem, the solubility of the oxidized
component of the redox couple, and the identification of a better
counter electrode material to obtain efficient charge transfer.
3. Counter Electrodes. Since the standard type of counter
electrode based on platinized FTO, which is optimized for an
iodine-based redox electrolyte, obviously is anything but opti-
mized for the sulfur-based redox electrolytes, advanced counter
electrode materials have been investigated. The main objective is
to introduce a CE material that offers a large effective surface for
charge transfer. Obvious candidates involve conducting carbon
materials and polymeric materials.25,30,32 In this section, the
performance of PEDOT as counter electrode material is pre-
sented and discussed. The PEDOT CEs were prepared by
electrochemical polymerization on an FTO conducting glass
substrate using the 3,4-ethylenedioxythiophene (EDOT) mono-
mer at a concentration of 20 mM in a 0.1 M LiClO4 solution of
MeCN via CV scans amounting to one or two cycles with the
Figure 8. JꢀV curves of electrolyte A based DSCs using platinized FTO
and PEDOT CEs.
sweep rate of 50 mV sꢀ1 over the potential range ꢀ0.6 to 1.2 V
3
versus Ag/Agþ. The UVꢀvis spectrum of the PEDOT CE (see
Figure S2 of the Supporting Information) displays two absorp-
tion bands at 400ꢀ700 nm and 700ꢀ900 nm. According to the
literature,34 the 400ꢀ700 nm absorption band is ascribed to the
reduced form of PEDOT and the other 700ꢀ900 nm band is
ascribed to the oxidized form of PEDOT. This means that the
PEDOT film prepared under our experimental conditions could
be a mixture of the reduced and oxidized forms. The electrolyte
used in the investigations of the new CE material is electrolyte A.
The DSC devices showed an unexpected efficiency of 6.0% with a
Jsc value of 12.5 mA cmꢀ2, a Voc value of 745 mV, and an ff
3
of 0.65 under 100 mW cmꢀ2 simulated sunlight illumination.
3
This result is comparable to the reported efficiency of 6.4%
obtained using ruthenium-dye-based DSCs with another thiolate/
disulfide-based electrolyte.3
To optimize the PEDOT CEs, one and two CV cycles were
introduced to prepare PEDOT CEs. Using a different number of
CVs most likely influences both film thickness and morphology.
The RCE of DSCs based on PEDOT films prepared by two cycles
is much lower than that of DSCs based on PEDOT films
prepared by one cycle (see Figure S3 of the Supporting In-
formation) leading to a higher ff, 0.65. However, more CV cycles
made the PEDOT film too brittle and too weakly attached to the
FTO surface. As a result, the films became easy to peel off from
the FTO surface, and the interface contact must be rather bad.
Consequently, PEDOT films from two CV cycles were used in
the following experiments. Noticeably, when a PEDOT film is
employed as CE in DSCs, an additional semicircle (2ꢀ3 Ω) in
the high-frequency region is observed. This effect is likely to
originate from the charge-transfer resistance (RCE0) at the FTO/
PEDOT interface.
Figure 9. Nyquist (a) and Bode (b) plots of DSCs containing electro-
lyte A based on platinized FTO and PEDOT CEs with ꢀ0.7 V bias
voltage.
bias voltage under dark conditions (Figure 9a). Two semicircles
are clearly observed in the Nyquist plots of DSCs based on
platinized FTO of which the two effects originate from the CE
and the dye-sensitized WE, respectively. By fitting the Nyquist
plots to an electrochemical model,49 the electrolyte reduction
resistance on CE and the electron recombination resistance from
WE of DSCs based on platinized FTO CEs are 320 Ω and 500 Ω,
respectively. The WE resistance of DSCs based on PEDOT CEs
The corresponding JꢀV curves of DSCs based on platinized
FTO and PEDOT CEs are shown in Figure 8. To better
understand the reasons for the difference in performance be-
tween the PEDOT and platinized FTO CEs in DSCs based on
the 1ꢀ/1 redox couple, EIS of DSCs was performed using ꢀ0.7 V
9417
dx.doi.org/10.1021/ja2030933 |J. Am. Chem. Soc. 2011, 133, 9413–9422

Journal of the American Chemical Society
ARTICLE
Figure 10. Thin-layer cells based on the CE//electrolyte//CE system
with 0 V bias voltage.
Figure 11. LSV of PEDOT (a), platinized FTO (b), and bare FTO (c)
in 2 mM 1, 0.1 M LiClO4 in MeCN solution, and PEDOT (d) in 0.1 M
LiClO4 in MeCN solution.
is ca. 210 Ω. The electron lifetime (τe), giving an estimate of the
electron recombination rate between the electrolyte and the
TiO2 film, displays a difference between DSCs with a PEDOT
CE and a platinized FTO. The τe extracted from the angular
frequency (ωrec) in the Bode phase plots (Figure 9b), using the
relation τe = 1/ωrec, gives a lifetime of 93 ms for DSCs based on
the PEDOT CEs. This is significantly lower than that of the
platinized FTO CEs, 250 ms. This may explain the slightly lower
Voc value obtained using PEDOT CEs. As shown by eqs 1ꢀ3, the
injected electron will be engaged in three different recombination
processes with the oxidized redox couple species as well as with
the radical species formed from the WE or CE. Therefore, the
effective reduction of oxidized species on the CE could increase
the concentration of radical species in the electrolytes (see eq 5)
accelerating the recombination eq 3. This may explain the
unexpected indirect effect of the CE material on recombination
losses at the WE.
Recombination processes:
RꢀSꢀSꢀR þ WEðeꢀÞ f RꢀSꢀ þ RꢀS
RꢀSꢀSꢀR þ WEð2eꢀÞ f 2RꢀSꢀ
ð1Þ
ð2Þ
ð3Þ
3
RꢀS þ WEðeꢀÞ f RꢀSꢀ
3
Reduction processes:
RꢀSꢀSꢀR þ CEð2eꢀÞ f 2RꢀSꢀ
ð4Þ
ð5Þ
RꢀSꢀSꢀR þ CEðeꢀÞ f RꢀSꢀ þ RꢀS
3
Noticeably, the semicircle originating from the PEDOT CE
significantly overlaps with the WE semicircle in PEDOT-based
DSCs. Therefore, it is difficult to accurately fit the Nyquist plot
data to extract a fully reliable CE resistance. Therefore, EIS
investigations of the thin-layer CE//electrolyte A//CE systems
were employed under ꢀ0.7 V (Figures S4 and S5 of the
Supporting Information) and 0 V (Figure 10) bias voltage in
order to extract the CE resistance. As shown in Figure 10, the
Nyquist plots should consist of the resistance originating from
the CE and from the electrolyte, denoted Warburg (frequency-
independent) impedance.3 The latter can be clearly observed at
ꢀ0.7 V bias voltage and is very low, 5 Ω. It means that the
diffusion of the redox couple components in the DSC electrolytes
Figure 12. Five cycles of LSV of platinized FTO (a) and PEDOT (b) in
2 mM 1, 0.1 M LiClO4 in MeCN solution.
studied here is sufficiently high. By fitting the Nyquist plots
obtained at 0 V bias voltage to an electrochemical model, it can be
deduced that PEDOT as CE materials displays an average
charge-transfer resistance on the electrolyte/CE interface
(RCE) of 45 Ω. This is considerably lower than that of DSCs
with the platinized FTO CE giving an RCE of 355 Ω, which is
almost 8 times higher. This implies that the reduction of 1 is
9418
dx.doi.org/10.1021/ja2030933 |J. Am. Chem. Soc. 2011, 133, 9413–9422

Journal of the American Chemical Society
ARTICLE
Table 2. Photovoltaic Properties of DSCs Containing the
Electrolytes IL1 and IL2 under Different Intensity Light
Illuminationa
light intensityb
[sun]
Jsc
Voc
ff
η
electrolyte
IL1
[mA cmꢀ2
]
[mV]
[%]
[%]
3
0.1
1
0.82
2.76
0.72
0.96
652
714
640
708
0.65
0.37
0.17
0.19
3.4
0.7
0.8
0.1
IL2
0.1
1
a The thickness of the TiO2 film is 6 μm transparent layer. b One sun is
equal to 100 mW cmꢀ2
.
3
much less disfavored on the PEDOT surface than on the
platinized FTO surface.
To further elucidate the essential differences between the
PEDOT and the platinized FTO with respect to reduction of 1,
linear sweep voltammetry (LSV) was performed using a 0.1 M
LiClO4 in MeCN solution at a sweep rate of 50 mV sꢀ1 over the
3
potential range from 0 to ꢀ0.8 V and with 2 mM 1 employing the
PEDOT and platinized FTO as working electrodes. As refer-
ences, also LSV for bare FTO (Figure 11c) and PEDOT
(Figure 11d) in an analogous electrolyte with and without 1
added was included, respectively.
As shown in Figure 11a, the PEDOT electrode renders an
obvious reduction peak at ꢀ0.52 V, which is ascribed to the
reduction peak of 1. The platinized FTO working electrode
shows a reduction peak of 1 at ꢀ0.61 V (Figure 11b). Futher-
more, the PEDOT electrode shows higher reduction current
density of 0.74 mA cmꢀ2 than platinized FTO, 0.58 mA cmꢀ2
.
3
3
Figure 13. The JꢀV curves (0.1 sun) and IPCE spectra of DSCs
In addition, as shown in Figure 12a, and in contrast to the
PEDOT electrode, the potential of 1 reduction using the
platinized FTO electrode is not constant upon cycling and shifts
to more negative potentials ending at ꢀ0.72 V. Such an effect
could be explained by the adsorption of the reduced species onto
the Pt nanoparticle surfaces, which may inhibit the reduction of 1.
Therefore, in continuously operating DSCs, an adsorption
process on the CE could cause significant degradation of
performance with time. However, this phenomenon was not
observed for the PEDOT electrode for which the reduction peak
of 1 was constant at ꢀ0.52 V upon voltage cycling (Figure 12b).
The less effective reduction ability (low reduction current density)
and a more negative reduction potential of 1 on the platinized FTO
surface may decrease the reduction rate of 1 and may thus constitute
the main reason for the lower ff observed in the DSCs.
containing the electrolytes IL1 and IL2.
4. Solvent-Free and Solvated Ionic Liquid Electrolytes.
Solvent-free iodide ionic liquid (IL) electrolytes have been
employed successfully in DSCs showing the potential of com-
mercial application because of the high efficiency and nonvolatile
properties.50ꢀ56 Fortunately, the thiolates 1aꢀ and 1bꢀ are pure
ILs at room temperature, which provide us with a chance to study
the solvent-free thiolate IL electrolytes in DSCs for the first time.
The electrolytes IL1 and IL2 were prepared through dissolving
0.2 M 1 in 1aꢀ and 1bꢀ, respectively.
Table 2 shows the best photovoltaic properties of DSCs based
on IL1 and IL2 under different light intensity illumination, 1 and
0.1 sun, respectively. The JꢀV curves (0.1 sun) and the IPCE
spectra of DSCs based on the two different electrolytes are
shown in Figure 13. As we can learn from these data, efficiencies
of DSCs based on these IL electrolytes benefit from lower light
intensities. This result could probably imply that the mass
Figure 14. Nyquist plots of PEDOT//IL electrolyte//PEDOT systems
with ꢀ0.2 V bias voltage.
transport of the redox couple in the electrolytes plays an
important role with respect to the efficiencies of the DSC devices.
Under 0.1 sun illumination, the DSCs based on IL1 show the best
efficiency of 3.4% indicating a potential for indoor applications.
As shown in Figure 13, in comparison to the electrolyte IL1, the
DSCs based on the electrolyte IL2 show slightly lower photo-
current as well as IPCE values but a similar photovoltage and a
lower fill factor resulting in the much lower efficiency observed to
0.8% under 0.1 sun illumination. The disappointing ff values are
9419
dx.doi.org/10.1021/ja2030933 |J. Am. Chem. Soc. 2011, 133, 9413–9422

Journal of the American Chemical Society
ARTICLE
semicircle can be observed. Thus, it is difficult to fit the semicircle
to get accuratedata. The estimated Rd of the IL2-based model isca.
700 Ω, which isalso higher than that obtained from the IL1 model.
The result is in coherence with the higher viscosity observed for
the electrolyte IL2. All the observed higher resistances of IL2-
based system result in a lower ff as well as in a lower photocurrent.
To decrease the viscosity of the IL electrolytes, we also
introduced the concept of solvated ionic liquids to the experi-
ments. When the thiolate 1ꢀ is mixed with 1,2-dimethylimida-
zole under heating, a solvated ionic liquid (SIL) forms. In
addition, another SIL also can be formulated when phenol, as a
hydrogen-bond donor, is mixed with 1ꢀ. For the traditional
iodine-based electrolyte, similar, so-called eutectic solvents (ESs)
have been used as a promising strategy to prepare a conductive,
less flammable, and less volatile solvent with a low melting point
(m.p.) and a high boiling point (b.p.)50,57 and good DSC
performance. Although the viscosity of the SILs is significantly
decreased, the best efficiency is only 1.2% obtained for the
Figure 15. Light-intensity dependence of the current density under
short-circuit conditions for DSCs containing the IL electrolytes.
phenol component-based DSCs with a Jsc of 0.7 mA cmꢀ2
,
likely to be caused by the counter electrode or diffusion
resistances in the electrolyte. To test this hypothesis, EIS of
the thin-layer systems (PEDOT//IL electrolyte//PEDOT)
based on the electrolytes IL1 and IL2 was performed with a
series of bias voltage (Figure S6 of the Supporting Information).
At a ꢀ0.2 V bias voltage (see Figure 14), the electrolyte
reduction characteristics on the CE and the electrolyte diffusion
resistance can be clearly observed in the Nyquist plots.
3
Voc of 498 mV, and ff of 0.32 under 0.1 sun illumination.
According to the detailed investigation of SIL-based electrolytes
(shown in the Supporting Information), we note that the
photovoltaic properties of DSCs based on SILs strongly depend
on the components used. The positive shift of the TiO2 CB of
phenol offers a better injection efficiency giving a higher overall
conversion efficiency in spite of the lower open-circuit voltage
caused by such a shift.
Consequently, the PEDOT//IL electrolyte//PEDOT systems
based on the electrolytes IL1 and IL2 were investigated and were
compared under the same applied bias voltage ꢀ0.2 V. All of the
resistances extracted for the system, including the series resistance
(Rs) from the conduction through the conducting glass at the WE
and CE, the counter electrode resistances consisting of FTO/
PEDOT (RCE0) and PEDOT/electrolyte (RCE) resistances, and
the electrolyte diffusion resistance (Rd) can be clearly observed
and separated in frequency space. By fitting the Nyquist plots to a
suitable electrochemical model, the corresponding resistance data
can be obtained. The Rs and RCE0, 48 Ω and 70 Ω, obtained from
the model based on the electrolyte IL2 is much higher than the
values obtained from the systems based on the electrolyte IL1, 13
Ω and 4 Ω. Especially, the RCE value, 2600 Ω, obtained from the
model based on the electrolyte IL2 is much higher than that of the
electrolyte IL1, 510 Ω. This means that the 1 component in DSCs
containing the electrolyte IL2 is far less easily reduced by the
PEDOT CE than cells based on the organic solvent-based
electrolyte A (vide supra). However, the concentration of 1 in
the electrolytes IL1 and IL2 is the same, and the PEDOT CEs are
also the same. Consequently, the difference in RCE between the
electrolytes IL1 and IL2 is probably caused by the viscosity of
the ionic liquids, which can affect the mass transport of 1 in the
electrolytes as well as within the solvated PEDOT film. Viscosity
measurements for these two ionic liquids were carried out, and the
results show that the viscosity of 1bꢀ is ca. 1300 cP, which is
almost 12 times higher than that of 1aꢀ, ca. 100 cP. Therefore, the
larger counter electrode resistance could presumably be caused by
the higher viscosity of the ionic-liquid solvents. Also, the light-
intensity dependence of the current density under short-circuit
conditions for DSCs (Figure 15) based on the IL electrolytes
shows that cells containing the electrolyte IL2 give worse restric-
tions in mass transport than the electrolyte IL1 even under low-
light intensity. Turning the attention to the Rd, the model based on
the electrolyte IL1 renders a value of 360 Ω. Unfortunately, for the
IL2-based model, only a portion of the electrolyte diffusion
’ CONCLUSIONS
In summary, a series of thiolate/disulfide redox couples with
weak absorption in the visible light region have been synthesized
and applied in organic dye (TH305) sensitized solar cells. The
1ꢀ/1 redox couple with platinized FTO CE provides the best
efficiency (η) of 4.6% with a short-current density (Jsc) of 12.3
mA cmꢀ2, an open-circuit voltage (Voc) of 750 mV, and a fill
3
factor (ff) of 0.50 under AM 1.5G, 100 mW cmꢀ2 simulated sun
3
light illumination (1 sun). The large charge-transfer resistance at
the electrolyte/platinized FTO interface limits the ff value of
DSCs based on this electrolyte. An alternative CE material, the
organic poly(3,4-ethylenedioxythiophene) (PEDOT) film, de-
posited on a conducting glass substrate by electrochemical
polymerization and used as CE in DSCs has been shown to
outperform the standard platinized FTO-type of CE for a
thiolate/disulfide-type of redox couple. This experiment presents
a DSC model based on all-organic dye, redox couple (1ꢀ/1), and
CE, and the most important result is a promising efficiency of
6.0% with a ff of 0.65 achieved under standard simulated 1 sun
illumination. As compared to a platinized FTO CE, the PEDOT-
based film shows considerably lower charge-transfer resistance at
the electrolyte/PEDOT interface and also offers a lower poten-
tial of reduction of 1. In combination, this gives a higher ff value
resulting in a higher conversion efficiency in such DSCs.
Furthermore, the concepts of pure and solvated ionic liquids
were introduced to this kind of thiolate/disulfide redox-couple-
based DSCs for the first time. The best efficiency of 3.4% was
achieved by DSCs based on the pure ionic liquid electrolyte with
1a-/1 redox couple under 10 mW cmꢀ2 (0.1 sun) illumination.
3
Future work will focus on strategies to decrease the electrolyte
mass-transport limitations and to improve the chemical stability
of this kind of electrolyte by structral modification as well as by
optimization of electrolyte composition.
9420
dx.doi.org/10.1021/ja2030933 |J. Am. Chem. Soc. 2011, 133, 9413–9422

Journal of the American Chemical Society
ARTICLE
’ ASSOCIATED CONTENT
(12) Daeneke, T.; Kwon, T.-H.; Holmes, A. B.; Duffy, N. W.; Bach,
U.; Spiccia, L. Nat. Chem 2011, 3, 213–217.
(13) Bai, Y.; Yu, Q.; Cai, N.; Wang, Y.; Zhang, M.; Wang, P. Chem.
Commun. 2011, 47, 4376–4378.
(14) Wang, Z.-S.; Sayama, K.; Sugihara, H. J. Phys. Chem. B 2005,
109, 22449–22455.
(15) Teng, C.; Yang, X.; Yuan, C.; Li, C.; Chen, R.; Tian, H.; Li, S.;
Hagfeldt, A.; Sun, L. Org. Lett. 2009, 11, 5542–5545.
(16) Rani, S.; Suri, P.; Mehra, R. M. Prog. Photovolt: Res. Appl. 2010,
19, 180–186.
S
Supporting Information. Experimental section; electro-
b
chemical data for different redox couples; the electrochemical
curves and data of different redox couples; UVꢀvis spectrum and
presumed structure of the prepared PEDOT material; Nyquist
plots of DSCs containing the electrolyte A based on different
PEDOT CEs; the EIS of thin-layer cells Pt CE//electrolyte A//
Pt CE and PEDOT CE//electrolyte A//PEDOT CE with ꢀ0.7
V bias voltage; the EIS of thin-layer cells PEDOT//IL1//
PEDOT and PEDOT//IL1//PEDOT at different applied bias
voltages; the collection of photovoltaic properties of DSCs based
on different dyes, counter electrodes and electrolytes; the DSC
photovoltaic properties of different concentrations of 1ꢀ using
PEDOT as counter electrode; JꢀV curves of N719-based DSCs
using different electrolytes and counter electrodes; the photo-
voltaic curves and properties of DSCs based on thiolate ionic
liquid electrolyte IL2 with different thickness TiO2 films under
(17) Teng, C.; Yang, X.; Li, S.; Cheng, M.; Hagfeldt, A.; Wu, L.; Sun,
L. Chem.—Eur. J. 2010, 16, 13127–13138.
(18) Oskam, G.; Bergeron, B. V.; Meyer, G. J.; Searson, P. C. J. Phys.
Chem. B 2001, 105, 6867–6873.
(19) Wang, P.; Zakeeruddin, S. M.; Moser, J.-E.; Humphry-Baker,
R.; Gr€atzel, M. J. Am. Chem. Soc. 2004, 126, 7164–7165.
(20) Bergeron, B. V.; Marton, A.; Oskam, G.; Meyer, G. J. J. Phys.
Chem. B 2005, 109, 937–943.
(21) Ning, Z.; Tian, H.; Qin, H.; Zhang, Q.; Ågren, H.; Sun, L.; Fu, Y.
J. Phys. Chem. C 2010, 114, 15184–15189.
(22) Li, L.; Yang, X.; Zhao, J.; Gao, J.; Hagfeldt, A.; Sun, L. J. Mater.
Chem. 2011, 21, 5573–5575.
(23) Gregg, B. A.; Pichot, F.; Ferrere, S.; Fields, C. L. J. Phys. Chem. B
2001, 105, 1422–1429.
(24) Snaith, H. J.; Zakeeruddin, S. M.; Wang, Q.; Pechy, P.; Gr€atzel,
M. Nano Lett. 2006, 6, 2000–2003.
(25) Li, D.; Li, H.; Luo, Y.; Li, K.; Meng, Q.; Armand, M.; Chen, L.
Adv. Funct. Mater. 2010, 20, 3358–3365.
(26) Tian, H.; Jiang, X.; Yu, Z.; Kloo, L.; Hagfeldt, A.; Sun, L. Angew.
Chem., Int. Ed. 2010, 49, 7328–7331.
10 mW cmꢀ2 sunlight illumination; light-intensity dependence
3
of the current density under short-circuit conditions for DSCs
with SIL electrolytes; charge extraction and electron lifetime of
DSCs based on the electrolytes IL1 and IL2; detailed investiga-
tions of solvated ionic liquid electrolytes. This material is
available free of charge via the Internet at http://pubs.acs.org.
’ AUTHOR INFORMATION
Corresponding Author
larsa@kth.se; lichengs@kth.se
(27) Zaban, A.; Micic, O. I.; Gregg, B. A.; Nozik, A. J. Langmuir 1998,
14, 3153–3156.
(28) Pichot, F.; Gregg, B. A. J. Phys. Chem. B 2000, 104, 6–10.
(29) Zhang, Z.; Chen, P.; Murakami, T. N.; Zakeeruddin, S. M.;
Gr€atzel, M. Adv. Funct. Mater. 2008, 18, 341–346.
(30) U.S. Patent 20080041438, 2007.
(31) WO Patent 2007109907, 2008.
(32) Liu, Y.; Jennings, J. R.; Parameswaran, M.; Wang, Q. Energy
Environ. Sci. 201010.1039/c1030ee00519c.
(33) Kiya, Y.; Henderson, J. C.; Hutchison, G. R.; Abru~na, H. D.
J. Mater. Chem. 2007, 17, 4366–4376.
(34) Kiya, Y.; Hatozaki, O.; Oyama, N.; Abru~na, H. D. J. Phys. Chem.
C 2007, 111, 13129–13136.
’ ACKNOWLEDGMENT
This work was supported by the Swedish Research Council,
the Swedish Energy Agency, and the Knut & Alice Wallenberg
Foundation. Z. Y. also would like to acknowledge the China
Scholarship Council (CSC) for financial support. We would like
to thank Kazuteru Nonomura (Uppsala University) for helpful
discussions and Dr. Xiao Jiang (KTH) for supplying the
TiO2 paste.
(35) Tian, H.; Yang, X.; Cong, J.; Chen, R.; Liu, J.; Hao, Y.; Hagfeldt,
A.; Sun, L. Chem. Commun. 2009, 6288–6290.
’ REFERENCES
(36) Hong, W.; Xu, Y.; Lu, G.; Li, C.; Shi, G. Electrochem. Commun.
2008, 10, 1555–1558.
(37) Ahmad, S.; Yum, J.-H.; Xianxi, Z.; Gr€atzel, M.; Butt, H.-J.;
Nazeeruddin, M. K. J. Mater. Chem. 2010, 20, 1654–1658.
(38) Pringle, J. M.; Armel, V.; MacFarlane, D. R. Chem. Commun.
2010, 46, 5367–5369.
(1) O’Regan, B.; Gr€atzel, M. Nature 1991, 353, 737–740.
(2) Hagfeldt, A.; Boschloo, G.; Sun, L.; Kloo, L.; Pettersson, H.
Chem. Rev. 2010, 110, 6595–6663.
(3) Wang, M.; Chamberland, N.; Breau, L.; Moser, J.-E.; Humphry-
Baker, R.; Marsan, B.; Zakeeruddin, S. M.; Gr€atzel, M. Nat. Chem. 2010,
2, 385–389.
(39) Lee, K. S.; Lee, H. K.; Wang, D. H.; Park, N.-G.; Lee, J. Y.; Park,
O. O.; Park, J. H. Chem. Commun. 2010, 46, 4505–4507.
(40) Saito, Y.; Kitamura, T.; Wada, Y.; Yanagida, S. Chem. Lett.
2002, 1060–1061.
(41) Yanagida, S.; Yu, Y.; Manseki, K. Acc. Chem. Res. 2009,
42, 1827–1838.
(4) Nusbaumer, H.; Moser, J.-E.; Zakeeruddin, S. M.; Nazeeruddin,
M. K.; Gr€atzel, M. J. Phys. Chem. B 2001, 105, 10461–10464.
(5) Sapp, S. A.; Elliott, C. M.; Contado, C.; Caramori, S.; Bignozzi,
C. A. J. Am. Chem. Soc. 2002, 124, 11215–11222.
(6) Nusbaumer, H.; Zakeeruddin, S. M.; Moser, J.-E.; Gr€atzel, M.
Chem.—Eur. J. 2003, 9, 3756–3763.
(42) Manseki, K.; Jarernboon, W.; Youhai, Y.; Jiang, K.-J.; Suzuki, K.;
Masaki, N.; Kim, Y.; Xia, J.; Yanagida, S. Chem. Commun. 2011,
47, 3120–3122.
(43) Senadeera, R.; Fukuri, N.; Saito, Y.; Kitamura, T.; Wada, Y.;
Yanagida, S. Chem. Commun. 2005, 2259–2261.
(44) Xia, J.; Masaki, N.; Lira-Cantu, M.; Kim, Y.; Jiang, K.; Yanagida,
S. J. Am. Chem. Soc. 2008, 130, 1258–1263.
(7) Cazzanti, S.; Caramori, S.; Argazzi, R.; Elliott, C. M.; Bignozzi,
C. A. J. Am. Chem. Soc. 2006, 128, 9996–9997.
(8) Gibson, E. A.; Smeigh, A. L.; Le Pleux, L.; Fortage, J.; Boschloo,
G.; Blart, E.; Pellegrin, Y.; Odobel, F.; Hagfeldt, A.; Hammarstr€om, L.
Angew. Chem., Int. Ed. 2009, 48, 4402–4405.
(9) Caramori, S.; Husson, J.; Beley, M.; Bignozzi, C. A.; Argazzi, R.;
Gros, P. C. Chem.—Eur. J. 2010, 16, 2611–2618.
(45) Saito, Y.; Azechi, T.; Kitamura, T.; Hasegawa, Y.; Wada, Y.;
Yanagida, S. Coord. Chem. Rev. 2004, 248, 1469–1478.
(46) Kim, Y.; Sung, Y.-E.; Xia, J.-B.; Lira-Cantu, M.; Masaki, N.;
Yanagida, S. J. Photochem. Photobiol., A: Chem. 2008, 193, 77–80.
(10) Feldt, S. M.; Gibson, E. A.; Gabrielsson, E.; Sun, L.; Boschloo,
G.; Hagfeldt, A. J. Am. Chem. Soc. 2010, 132, 16714–16724.
(11) Wang, H.; Nicholson, P. G.; Peter, L.; Zakeeruddin, S. M.;
Gr€atzel, M. J. Phys. Chem. C 2010, 114, 14300–14306.
9421
dx.doi.org/10.1021/ja2030933 |J. Am. Chem. Soc. 2011, 133, 9413–9422

Journal of the American Chemical Society
ARTICLE
(47) Jansen, M.; Rabe, H.; Strehless, A.; Dieler, S.; Debus, F.;
Dannhardt, G.; Akabas, M. H.; Lueddens, H. J. Med. Chem. 2008,
51, 4430–4448.
(48) Hagberg, D. P.; Edvinsson, T.; Marinado, T.; Boschloo, G.;
Hagfeldt, A.; Sun, L. Chem. Commun. 2006, 2245–2247.
(49) Fabregat-Santiago, F.; Bisquert, J.; Garcia-Belmonte, G.;
Boschloo, G.; Hagfeldt, A. Sol. Energy Mater. Sol. Cells 2005, 87, 117–
131.
(50) Bai, Y.; Cao, Y.; Zhang, J.; Wang, M.; Li, R.; Wang, P.;
Zakeeruddin, S. M.; Gr€atzel, M. Nat. Mater. 2008, 7, 626–630.
(51) Papageorgiou, N.; Athanassov, Y.; Armand, M.; Bonhote, P.;
Pettersson, H.; Azam, A.; Gr€atzel, M. J. Electrochem. Soc. 1996,
143, 3099–3108.
(52) Zakeeruddin, S. M.; Gr€atzel, M. Adv. Funct. Mater. 2009,
19, 2187–2202.
(53) Yu, Z.; Gorlov, M.; Nissfolk, J.; Boschloo, G.; Kloo, L. J. Phys.
Chem. C 2010, 114, 10612–10620.
(54) Wang, P.; Zakeeruddin, S. M.; Comte, P.; Exnar, I.; Gr€atzel, M.
J. Am. Chem. Soc. 2003, 125, 1166–1167.
(55) Kubo, W.; Kitamura, T.; Hanabusa, K.; Wada, Y.; Yanagida, S.
Chem. Commun. 2002, 374–375.
(56) Kuang, D.; Wang, P.; Ito, S.; Zakeeruddin, S. M.; Gr€atzel, M.
J. Am. Chem. Soc. 2006, 128, 7732–7733.
(57) Jhong, H.-R.; Wong, D. S.-H.; Wan, C.-C.; Wang, Y.-Y.; Wei,
T.-C. Electrochem. Commun. 2009, 11, 209–211.
9422
dx.doi.org/10.1021/ja2030933 |J. Am. Chem. Soc. 2011, 133, 9413–9422