Tuning the HOMO energy levels of organic dyes for dye-sensitized solar cells based on Br-/Br3- electrolytes

Article abstract of DOI:10.1002/chem.201000460

A series of novel metal-free organic dyes TC301-TC310 with relatively high HOMO levels were synthesized and applied in dye-sensitized solar cells (DSCs) based on electrolytes that contain Br^-/Br₃^- and I^-/I₃^-. The effects of additive Li⁺ ions and the HOMO levels of the dyes have an important influence on properties of the dyes and performance of DSCs. The addition of Li⁺ ions in electrolytes can broaden the absorption spectra of the dyes on TiO₂ films and shift both the LUMO levels of the dyes and the conduction band of TiO₂, thus leading to the increase of J_sc and the decrease of V_oc. Upon using Br^-/Br₃^- instead of I^-/I₃^-, a large increase of V_oc is attributed to the enlarged energy difference between the redox potentials of electrolyte and the Fermi level of TiO₂, as well as the suppressed electron recombination. Incident photon to current efficiency (IPCE) action spectra, electrochemical impedance spectra, and nanosecond laser transient absorption reveal that both the electron collection yields and the dye regeneration yields (Φ_r) depend on the potential difference (the driving forces) between the oxidized dyes and the Br^-/Br ₃^- redox couple. For the dyes for which the HOMO levels are more positive than the redox potential of Br^-/Br₃ ^- sufficient driving forces lead to the longer effective electron-diffusion lengths and almost the same efficient dye regenerations, whereas for the dyes for which the HOMO levels are similar to the redox potential of Br^-/Br₃^-, insufficient driving forces lead to shorter effective electron-diffusion lengths and inefficient dye regenerations. A new middleman: The Br^-/Br₃^- redox mediator was proved to be a promising alternative to I^-/I ₃^- for application in dye-sensitized solar cells (see graphic). Studies show that the energy gap between the HOMO level of the dye and the potential of the Br^-/Br₃^- redox mediator influences their charge-transfer processes and solar-energy conversion efficiency.

Full text of DOI:10.1002/chem.201000460

FULL PAPER
DOI: 10.1002/chem.201000460
Tuning the HOMO Energy Levels of Organic Dyes for Dye-Sensitized Solar
Cells Based on Brꢀ/Br3 Electrolytes
ꢀ
Chao Teng,[a] Xichuan Yang,*[a] Shifeng Li,[a] Ming Cheng,[a] Anders Hagfeldt,[a, b]
Li-zhu Wu,[c] and Licheng Sun*[a, d]
ꢀ
Abstract: A series of novel metal-free
organic dyes TC301–TC310 with rela-
tively high HOMO levels were synthe-
sized and applied in dye-sensitized
solar cells (DSCs) based on electrolytes
Voc. Upon using Brꢀ/Br3 instead of Iꢀ/
yields and the dye regeneration yields
(Fr) depend on the potential difference
(the driving forces) between the oxi-
ꢀ
I3 , a large increase of Voc is attributed
to the enlarged energy difference be-
tween the redox potentials of electro-
lyte and the Fermi level of TiO2, as
well as the suppressed electron recom-
bination. Incident photon to current ef-
ficiency (IPCE) action spectra, electro-
chemical impedance spectra, and nano-
second laser transient absorption
reveal that both the electron collection
ꢀ
dized dyes and the Brꢀ/Br3 redox
couple. For the dyes for which the
HOMO levels are more positive than
that contain Brꢀ/Br3 and Iꢀ/I3 . The
effects of additive Li+ ions and the
HOMO levels of the dyes have an im-
portant influence on properties of the
dyes and performance of DSCs. The
addition of Li+ ions in electrolytes can
broaden the absorption spectra of the
dyes on TiO2 films and shift both the
LUMO levels of the dyes and the con-
duction band of TiO2, thus leading to
the increase of Jsc and the decrease of
ꢀ
ꢀ
ꢀ
the redox potential of Brꢀ/Br3 suffi-
cient driving forces lead to the longer
effective electron-diffusion lengths and
almost the same efficient dye regenera-
tions, whereas for the dyes for which
the HOMO levels are similar to the
redox potential of Brꢀ/Br3 , insufficient
ꢀ
Keywords:
bromides
charge transfer · dyes/pigments ·
energy conversion · sensitizers
·
driving forces lead to shorter effective
electron-diffusion lengths and ineffi-
cient dye regenerations.
Introduction
Dye-sensitized solar cells (DSCs), as a promising technology
for low-cost photovoltaics, have attracted much attention in
the past two decades owing to their high efficiency of ap-
proximately 12% and their high stability under prolonged
light and thermal dual stress.[1] A DSC employs a nanocrys-
talline porous semiconductor metal oxide such as TiO2 that
is sensitized by dye and infiltrated by liquid electrolyte that
contains a redox mediator. The dye is excited by solar-light
absorption and injects an electron into the conduction band
of TiO2. Thus the oxidized state of the dye is obtained. The
oxidized state of the dye receives an electron from a redox
mediator in the electrolyte and is regenerated. The common
choice of redox mediator for obtaining high solar-cell effi-
[a] C. Teng, Prof. X. Yang, S. Li, M. Cheng, Prof. A. Hagfeldt,
Prof. L. Sun
State Key Laboratory of Fine Chemicals
DUT-KTH Joint Education
and Research Center on Molecular Devices
Dalian University of Technology (DUT), Dalian (China)
Fax : (+86)411-837-02185
E-mail: yangxc@dlut.edu.cn
[b] Prof. A. Hagfeldt
School of Chemical Science and Engineering
Center of Molecular Devices, Physical Chemistry
Royal Institute of Technology (KTH), Stockholm (Sweden)
[c] Prof. L.-z. Wu
Key Laboratory of Photochemical Conversion
and Optoelectronic Materials
Technical Institute of Physics and Chemistry
Chinese Academy of Sciences, Beijing (China)
ꢀ
ciencies in liquid electrolyte is iodide/triiodide (Iꢀ/I3 ).[2]
The lowest unoccupied molecular orbital (LUMO) level of
the dye should be sufficiently more negative than the con-
duction band of TiO2 (Ecb =ꢀ0.5 V vs. normal hydrogen
electrode (NHE)) for efficient electron injection, and mean-
while the highest occupied molecular orbital (HOMO) level
of the dye should be sufficiently more positive than the po-
[d] Prof. L. Sun
Department of Chemistry, Organic Chemistry
Royal Institute of TechnologyACTHNUGRTNEUNG(KTH), Stockholm (Sweden)
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/chem.201000460.
Chem. Eur. J. 2010, 16, 13127 – 13138
ꢀ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
13127

tential of redox mediator in electrolyte for efficient dye re-
generation. Efficient operation of DSC relies on both effi-
cient electron injection and efficient dye regeneration. So
zole derivatives (TC307–TC310) were used as electron
donors, and cyanoacrylic acid as an electron-accepting/an-
choring moiety. The performance of solar cells based on
ꢀ
far, the most efficient DSCs with record high efficiency of
these new organic dyes and the redox couple Brꢀ/Br3 in
ꢀ[3]
ꢀ
around 12% are based on Ru dye N719 and Iꢀ/I3
mediator.[4]
redox
comparison with Iꢀ/I3 has been investigated by electro-
chemistry, impedance spectra, and transient absorption spec-
troscopy, as well as time-dependent density functional
theory (TD-DFT) calculations.
Compared to the great efforts spent on the design and
synthesis of different dyes, less study has been devoted to
the redox mediators. Ferrocene/ferrocenium (Fc/Fc+) has
been tested as a redox mediator, but it yielded a low open-
circuit voltage (Voc) and a small short-circuit photocurrent
density (Jsc).[5] CoII complexes showed an impressive effi-
ciency of 8% at low light intensity but their performance
drops remarkably under full-sun illumination.[6] Stable or-
ganic radical 2,2,6,6-tetramethyl-1-piperidinyloxy (TEMPO)
has been shown to give a Voc of
Results and Discussion
Photophysical properties: Absorption spectra of solutions of
TC301–TC310 in CH3CN are displayed in Figure 1, and the
corresponding data are presented in Table 1. All the dyes
0.86 V (vs. NHE) and an effi-
ciency of 5.4% under simulated
AM
1.5
full
sunlight
(100 mWcmꢀ2) irradiation.[7]
The theoretical maximum of
Voc is a crucial parameter for
DSC and is determined by the
energy difference between the
potential of redox mediator in
electrolyte and the Fermi level
of TiO2. Therefore one possible
method to increase Voc is to
find a new redox mediator with
a more positive potential than
EACHTUNGTRENNUNG
(Iꢀ/I3 ).[8] Brꢀ/Br3 is a prom-
ꢀ
ꢀ
ising candidate because of its
much more positive redox po-
tential (up to 1.1 V vs. NHE)
ꢀ
than
E
N
0.4 V).[9] Recently, we published
the synthesis and characteriza-
tion of two novel carbazole
dyes (TC301 and TC306,
Scheme 1) for DSCs with high
Voc up to 1 V and good solar-
energy conversion efficiencies
based on electrolytes that con-
ꢀ [10]
tain Brꢀ/Br3 .
To fine-tune
the HOMO level of the dye
and to gain further insight into
the relationship between the
structure of the dye and the
performance of DSC based on
ꢀ
the Brꢀ/Br3 redox couple, here
we report the design, synthesis,
and characterization of a series
of novel metal-free organic
dyes (TC301–TC310, the de-
tailed structures are shown in
Scheme 1) in which triphenyla-
Scheme 1. Molecular structures of a series of metal-free organic dyes TC301–TC310 for application in DSCs
ꢀ
mine (TC301–TC306) or carba- with Brꢀ/Br3 as redox couple.
13128
www.chemeurj.org
ꢀ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2010, 16, 13127 – 13138

Organic Dyes
FULL PAPER
to LUMO transitions, are observed.[11] The values of molar
extinction coefficient (e) at lmax are in the range of 1.0ꢁ104–
3.7ꢁ104 mꢀ1 cmꢀ1. The absorption spectrum of TC301 shows
a lmax at 374 nm (e=12300mꢀ1 cmꢀ1; Figure 1a). Compared
to TC301, the absorption spectra of TC302 and TC303 are
slightly redshifted because the introduction of tert-butyl or
carbazolyl groups enhances the electron-donating abilities of
an electron donor. Enlarging the p conjugation of TC301 to
give TC305 and TC306 caused a further redshift to 416 nm
(e=27400mꢀ1 cmꢀ1) and 429 nm (e=21000mꢀ1 cmꢀ1), respec-
tively. Compared to TC307, the absorption spectrum of
TC308 changes only little (Figure 1b), thereby revealing that
the introduction of a bromo group into the electron-donor
part has almost no influence on the absorption spectrum of
the dye. The redshifted absorption spectra of TC309 and
TC310 relative to that of TC307 are also caused by the en-
hancement of the electron donor and the increase in p con-
jugation. We can see from Figure 2 that a blueshifting effect
of the absorption spectra of the dyes TC301, TC306, and
TC310 was observed with an increase in the polarity of the
solvent from CH2Cl2 to CH3CN.[12]
When the dyes were attached onto TiO2 films, the absorp-
tion spectra may shift to the blue or red region compared to
that in solutions because of the interaction between the dyes
and the semiconductor surface. In general, the larger shift of
the absorption spectra of the dyes on TiO2 films compared
with that in solutions could indicate a higher tendency to
form aggregation on TiO2 film.[13] The idea of suppressing
the dye aggregation on TiO2
Figure 1. Absorption spectra of solutions of the dyes in CH3CN (2ꢁ
10ꢀ5 m): a) TC301–TC306 with triphenylamine derivatives as electron
donors; b) TC307–TC310 with carbazole derivatives as electron donors.
film and anchoring the dye
monolayer onto the TiO2 sur-
Table 1. Absorption, emission, and electrochemical properties of TC301–TC310.
abs [a]
max
em [b]
max
[c]
Dye
l
e
l
Eox
lmax on
TiO2 [nm]
E0–0 abs/em[f]
EoxꢀE0–0
face has been considered to
give higher efficiency of a DSC.
We can see from Figures 1 and
2 and Table 1 that all the ab-
sorption spectra of TC301–
TC310 on TiO2 films only shift-
ed slightly in comparison to
those in CH3CN solutions. This
indicates that TC301–TC310
have less of a tendency to form
aggregation on TiO2 films and
can likewise achieve higher effi-
ciencies of DSCs.
[nm]
A
]
[nm]
[V vs. NHE]
1.59
[V]
[V vs. NHE]
384[d]
392[e]
391[d]
407[e]
387[d]
398[e]
400[d]
407[e]
411[d]
425[e]
405[d]
412[e]
383[d]
389[e]
407[d]
414[e]
400[d]
407[e]
2.82
2.68
2.71
2.58
2.83
2.70
2.63
2.51
2.44
2.34
2.63
2.48
2.66
2.54
2.54
2.43
2.50
2.40
ꢀ1.23
ꢀ1.09
ꢀ1.35
ꢀ1.12
ꢀ1.38
ꢀ1.25
ꢀ1.15
ꢀ1.03
ꢀ1.06
ꢀ0.96
ꢀ1.30
ꢀ1.15
ꢀ1.23
ꢀ1.11
ꢀ1.30
ꢀ1.19
ꢀ1.38
ꢀ1.28
TC301
TC302
TC303
TC305
TC306
TC307
TC308
TC309
TC310
374
12300
7800
527
381
551
1.46
378
416
429
381
379
418
421
10800
27400
21000
10700
37300
12800
22400
423
543
553
515
560
557
559
1.45
1.48
1.38
1.33
1.43
1.24
1.12
Electrochemical properties: To
fine-tune the HOMO levels of
the dyes, we changed the struc-
tures of the dyes by introduc-
tion of different electron donors
and linkers. We can see from
Table 1 that the first oxidation
potentials (Eox) that correspond
to the HOMO levels of dyes
changed negatively upon en-
[a] Absorption in CH3CN (2ꢁ10ꢀ5 m) at RT. [b] Emission spectrum in CH3CN (2ꢁ10ꢀ5 m) at RT. [c] The oxida-
tion potential of the dyes was measured in DMF with TBAPF6 (0.1m) as electrolyte (working electrode: glassy
carbon; reference electrode: Ag/Ag+; calibrated with Fc/Fc+ as an internal reference and converted to NHE
by addition of 630 mV;[16] counterelectrode: Pt). [d] Absorption of dye loaded on TiO2 film immersed in
CH3CN without electrolyte. [e] Absorption of dye loaded on TiO2 film immersed in CH3CN with 0.8m LiBr in
electrolyte. [f] E0–0 was determined by the onset wavelength of the corresponding absorption spectrum on TiO2
films.
with a strong absorption maximum (lmax) located in the
spectral range of 370–430 nm, which corresponds to HOMO
hancing the electron donor and increasing p conjugation.
ꢀ
The HOMO level of T310 is almost similar to E
N
Chem. Eur. J. 2010, 16, 13127 – 13138
ꢀ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
13129

X. Yang, L. Sun et al.
Figure 2. Absorption spectra of solutions of the selected dyes in CH3CN
(2ꢁ10ꢀ5 m) compared to those in CH2Cl2 (2ꢁ10ꢀ5 m).
whereas the HOMO levels of TC301–TC309 are all more
ꢀ
positive than E
N
the dyes could help us scrutinize the relationship between
the structure of the dye and the performance of DSC based
ꢀ
on electrolytes that contain Brꢀ/Br3 .
The LUMO level of the dye was calculated by EoxꢀE0–0
,
in which E0–0 was determined by the onset wavelength of
the corresponding absorption spectrum on TiO2 films
(Figure 3). It is evident that the values of EoxꢀE0–0 are more
Figure 4. Absorption spectra of the selected dyes attached on TiO2 films
immersed in CH3CN with 0.8m LiBr and without electrolyte. a) TC301
on TiO2 films; b) TC306 on TiO2 films; c) TC308 on TiO2 films; and
d) TC310 on TiO2 films.
Figure 3. Absorption spectra of the dyes on TiO2 films: a) TC301–TC306
with triphenylamine as electron donors; b) TC307–TC310 with carbazole
as electron donors.
negative than Ecb, thus all the dyes can complete the process
of electron injection into the conduction band (cb) of TiO2
13130
www.chemeurj.org
ꢀ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2010, 16, 13127 – 13138

Organic Dyes
FULL PAPER
Table 2. Frontier molecular orbitals of the HOMO and LUMO calculated with DFT at the B3LYP/6-31+G(d)
level of the selected dyes.
change, thus the LUMO levels
of TC301–TC310 shifted posi-
tively about 100 mV in the pres-
ence of 0.8m LiBr.
Dye
HOMO
LUMO
DFT study: To gain insight into
TC301
the
geometrical
electronic
structures of the dyes, we per-
formed DFT calculations on
the dyes using the Gaussian 03
program package. In particular,
we used B3LYP as exchange-
correlation functional and 6-
31+G (d) as the basis set.[17] At
the ground state (HOMO) of
the dyes, electrons are homoge-
neously distributed on electron-
donor part. At the excited state
TC306
TC308
TC310
(LUMO),
intramolecular
charge-transfer occurs, thereby
resulting in the electron move-
ment to the electron-acceptor
part (cyanoacrylic acid). The
frontier molecular orbital of the
dyes reveals that HOMO–
LUMO excitation moves the
electron-density
distribution
from the electron-donor moiety
to the electron-acceptor moiety
by means of different p conju-
to form the oxidized dyes. Noticeably, the relatively large
energy gaps between the LUMO levels of the dyes and the
conduction band of TiO2 allow for the addition of tributyl
phosphate (TBP) into the electrolyte, which can shift the
conduction band of TiO2 negatively about 0.3 V and conse-
quently improve Voc and total solar-energy-to-electricity
conversion efficiency.[14]
gation. Thus the change in electron distribution induced by
photoexcitation results in an efficient charge separation
(Table 2).
Photocurrent–photovoltage characteristics: To compare the
performance of DSCs based on electrolytes that contain
ꢀ
ꢀ
Brꢀ/Br3 and those which contain Iꢀ/I3 and the effects of
the addition of Li+ ions, three kinds of devices (devices A,
B, and C) were constructed. Devices A are DSCs based on
Li+-ion effects on photoelectrochemical properties: With a
view towards practical development of high-performance
DSCs, the effects of additives to electrolyte have been ex-
tensively investigated. It is well known that the adsorption/
intercalation of potential determining cations such as Li+
ions into the porous TiO2 electrode can broaden the absorp-
tion of the dye on TiO2.[15] Figure 4 shows the absorption
spectra of the selected dyes on TiO2 films immersed in
CH3CN with 0.8m LiBr as additive and without electrolyte.
The various solvents used here were exchanged between
measurements while the samples were maintained in a fixed
position in the measurement cell. Thus, the changes of ab-
sorbance observed in different solvents most likely corre-
sponded to changes in the absorption coefficient of the dyes
on the TiO2 surface. We can see from Figure 4 and Table 1
that the absorption spectra of TC301–TC310 on TiO2 films
immersed in CH3CN solutions with 0.8m LiBr are all red-
shifted about 10 nm relative to that without electrolyte.
Redshift of the absorption on TiO2 films led to the E0–0
ꢀ
Brꢀ/Br3 and contain 0.9m 1,2-dimethyl-3-butylimidazolium
bromide (DMBIBr), 0.08m Br2, and 0.5m TBP electrolyte in
ꢀ
dry CH3CN. Devices B are DSCs also based on Brꢀ/Br3
and contain 0.9m DMBIBr, 0.8m LiBr, 0.08m Br2, and 0.5m
TBP electrolyte in dry CH3CN. As reference, devices C are
ꢀ
DSCs based on Iꢀ/I3 and contain 0.6m 1, 2-dimethyl-3-pro-
pylimidazolium iodide (DMPII), 0.06m LiI, 0.04m I2, and
0.4m TBP electrolyte in dry CH3CN. Figure 5 shows the
photocurrent–photovoltage characteristics for TC301–TC310
and for devices A, B, and C under simulated AM 1.5 irradia-
tion (100 mWcmꢀ2); the detailed photovoltaic parameters
ꢀ
are listed in Table 3. A DSC that consisted of Brꢀ/Br3 and
TC301 accomplished a very high open-circuit voltage (up to
1.156 V) with a good solar-energy-conversion efficiency of
3.68% (Figure 5a). For the TC306-based DSC, a high open-
circuit voltage (up to 0.939 V) with a considerably good
solar-energy-conversion efficiency of 5.22% was also ach-
ꢀ
ieved with an electrolyte that contained Brꢀ/Br3 (Fig-
Chem. Eur. J. 2010, 16, 13127 – 13138
ꢀ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
13131

X. Yang, L. Sun et al.
ure 5c). These results suggest
ꢀ
that Brꢀ/Br3 could be a prom-
ꢀ
ising alternative to Iꢀ/I3 for
DSCs with dyes for which the
HOMO levels are more posi-
ꢀ
tive than E
(Brꢀ/Br3 ).
ꢀ
When replacing Iꢀ/I3 with
Brꢀ/Br3
in TC301–TC310-
based DSCs, both devices A
and B yielded a significant in-
crease in Voc than devices C.
This was attributed to the en-
larged energy difference be-
tween the redox potential of
the electrolyte and the Fermi
level of TiO2. The Voc of devi-
ces B are slightly lower than
those of devices A due to the
positive shift of the conduction
band of TiO2 in the presence of
Li+ ions.[8b,18] And the higher Jsc
in devices B compared with de-
vices A was attributed to the
broader absorption of the dyes
on TiO2 films in the presence
of Li+ ions. We can see from
Figure 5 and Table 3 that upon
ꢀ
ꢀ
ꢀ
using Brꢀ/Br3 instead of Iꢀ/I3 ,
DSCs sensitized by TC301–
TC308 produced much higher
Voc and led to much higher h;
the DSC sensitized by TC309
produced much higher Voc and
almost the same h, but the DSC
sensitized by TC310 produced
much lower h along with much
higher Voc because of the large
decrease in Jsc. One of the main
reasons for such differences in
solar-cell performance may be
the different energy gaps be-
Figure 5. Photocurrent density versus voltage curves for DSC sensitized by the selected dyes with different
electrolytes: device A based on 0.9m DMBIBr, 0.08m Br2, and 0.5m TBP electrolyte in dry CH3CN; device B
based on 0.9m DMBIBr, 0.8m LiBr, 0.08m Br2, and 0.5m TBP electrolyte in dry CH3CN; and device C based
on 0.6m DMPII, 0.06m LiI, 0.04m I2, and 0.4m TBP electrolyte in dry CH3CN. a) DSC sensitized by TC301,
b) DSC sensitized by TC305, c) DSC sensitized by TC306, d) DSC sensitized by TC308, e) DSC sensitized by
TC309, and f) DSC sensitized by TC310.
tween the HOMO levels of the
ꢀ
dyes and
E
N
lead to the different driving
forces for dye regeneration.
vices A and B produced larger IPCE values compared to de-
vices C in this spectral region, except for DSC sensitized by
Incident photon-to-current con-
version efficiency (IPCE) action spectra: For DSCs sensi-
tized by TC301–TC310, the IPCE plateaus are all between
360 and 500 nm (Figure 6), and IPCE maxima are above
TC310. And we find that IPCE maxima produced by elec-
ꢀ
trolytes that contain Brꢀ/Br3 are very similar to those pro-
ꢀ
duced by those that contain Iꢀ/I3 , also except for DSC sen-
ꢀ
ꢀ
70% upon using electrolytes that contain Brꢀ/Br3 or Iꢀ/I3 .
Considering the light absorption and scattering loss by the
conducting glass, the maxima efficiencies for absorbed
sitized by TC310. Assuming the electron injection is inde-
pendent of the redox mediator,[19] the similar IPCE maxima
of DSC sensitized by TC301–TC309 may be due to the simi-
ꢀ
photon-to-collected-electron
conversion
efficiencies
lar dye-regeneration efficiencies upon using Brꢀ/Br3 and Iꢀ/
ꢀ
(APCEs) are very high, thus suggesting very high charge-
collection yields. Because of the lower absorption of bro-
mine relative to that of iodine below 400 nm (Figure 7), de-
I3 . For DSC sensitized by TC310, the much smaller IPCE
ꢀ
maxima of DSC based on Brꢀ/Br3 may be due to much less
ꢀ
dye-regeneration efficiency than that based on Iꢀ/I3 .
13132
www.chemeurj.org
ꢀ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2010, 16, 13127 – 13138

Organic Dyes
FULL PAPER
tron transfer from Pt countere-
lectrode to the oxidized species
in the electrolyte, that is, the re-
duction of the oxidized species
to the reduced species. The
large semicircle in the Nyquist
plots, which corresponds to the
midfrequency peaks in the
Bode phase plots, represents
the charge recombination that
arises from electrons in TiO2
ꢀ
film recombining with I3 ion in
the electrolyte.
From Figure 8 we find that
the high-frequency peaks in the
Bode phase plots also shifted to
lower frequency upon using
ꢀ
ꢀ
Brꢀ/Br3 instead of Iꢀ/I3 . Corre-
spondingly, in the Nyquist plots,
the small semicircle showed a
ꢀ
larger resistance for Brꢀ/Br3
than for Iꢀ/I3 . This means that
ꢀ
the reduction of bromine is
slower than that of iodine at
the Pt counterelectrode.
By fitting the EIS curves, we
can obtain some important pa-
rameters for the devices. One is
the electron lifetime (t), which
could be extracted from the an-
gular frequency (wmin) at the
midfrequency peak in the Bode
phase plot by using t=1/wmin
and expresses for electron re-
combination in TiO2 films.[20]
Rct and Rt represent charge-
transfer resistance at the dye/
TiO2/electrolyte interface relat-
ed to electron recombination
and the electron-transport re-
sistance in TiO2 film, respec-
tively. Another important pa-
rameter is the effective elec-
tron-diffusion length (Ln) in
TiO2 films, which could be ex-
Figure 6. IPCE action spectra for dye-sensitized solar cells sensitized by the dyes TC301–309 with different
electrolytes: device A based on 0.9m DMBIBr, 0.08m Br2, and 0.5m TBP electrolyte in dry CH3CN; device B
based on 0.9m DMBIBr, 0.8m LiBr, 0.08m Br2, and 0.5m TBP electrolyte in dry CH3CN; and device C based
on 0.6m DMPII, 0.06m LiI, 0.04m I2, and 0.4m TBP electrolyte in dry CH3CN. a) DSC sensitized by TC301,
b) DSC sensitized by TC305, c) DSC sensitized by TC306, d) DSC sensitized by TC308, e) DSC sensitized by
TC309, and f) DSC sensitized by TC310.
1
2
Electrochemical impedance spectroscopy: We can see from
tracted by using Ln =L
A
ꢀ
Figure 5 and Table 3 that upon using Brꢀ/Br3 instead of Iꢀ/
(L is the thickness of TiO2 film);[21] it reflects the competi-
tion between charge collection and recombination.
ꢀ
I3 , distinctive solar-energy-conversion efficiencies were ob-
tained by DSCs sensitized by TC308 and TC310. To gain
The electron lifetimes (t) and the charge-transfer resistan-
ces (Rct) of devices A and B are much larger than those of
devices C. The significant increase in t and Rct means that
the electron recombination that arises from electrons in
further insight into the interfacial charge-transfer processes
ꢀ
in DSCs based on electrolytes that contain Iꢀ/I3 or Brꢀ/
ꢀ
Br3 , we chose TC308 and TC310 as subjects for electro-
ꢀ
chemical impedance spectra (EIS) analysis. The Nyquist
plots and Bode phase plots are shown in Figure 8, and the
detailed parameters are shown in Table 4. The small semicir-
cle in the Nyquist plots, which corresponds to the high-fre-
quency peaks in the Bode phase plots, represents the elec-
TiO2 film with I3 ion in electrolyte is suppressed upon re-
ꢀ
ꢀ
placing Iꢀ/I3 with Brꢀ/Br3 . Thus we can conclude that sig-
ꢀ
nificant increase in Voc by using Brꢀ/Br3 -based electrolytes
ꢀ
instead of those based on Iꢀ/I3 were attributed to not only
the enlarged energy difference between the redox potential
Chem. Eur. J. 2010, 16, 13127 – 13138
ꢀ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
13133

X. Yang, L. Sun et al.
Table 3. Photovoltaic performance of DSC sensitized by TC301–TC310
in different electrolytes.
Dye
Device[a]
Jsc [mAcmꢀ2
]
Voc [V]
ff [%]
h [%]
A
B
C
A
B
C
A
B
C
A
B
C
A
B
C
A
B
C
A
B
C
A
B
C
A
B
C
4.00
4.93
4.42
4.71
5.27
4.64
3.88
4.27
4.13
5.83
6.31
6.35
7.12
7.36
8.76
3.73
4.02
4.90
5.08
5.19
4.95
4.41
5.23
6.39
6.12
6.06
9.39
1.156
1.041
0.696
1.091
1.009
0.721
1.077
0.969
0.692
1.033
0.959
0.665
0.939
0.915
0.621
0.946
0.881
0.620
1.045
0.976
0.673
1.015
0.960
0.710
0.988
0.945
0.706
0.796
0.713
0.767
0.788
0.749
0.778
0.784
0.756
0.775
0.766
0.751
0.764
0.781
0.752
0.754
0.750
0.774
0.744
0.765
0.760
0.746
0.755
0.723
0.764
0.728
0.735
0.791
3.68
3.66
2.36
4.05
3.98
2.60
3.28
3.13
2.21
4.61
4.54
3.23
5.22
5.07
4.10
2.65
2.74
2.26
4.06
3.85
2.48
3.38
3.63
3.47
4.40
4.21
5.24
TC301
TC302
TC303
TC305
TC306
TC307
TC308
TC309
TC310
[a] Device A is based on 0.9m DMBIBr, 0.08m Br2, and 0.5m TBP elec-
trolyte in dry CH3CN. Device B is based on 0.9m DMBIBr, 0.8m LiBr,
0.08m Br2, and 0.5m TBP electrolyte in dry CH3CN. Device C is based on
0.6m DMPII, 0.06m LiI, 0.04m I2, and 0.4m TBP electrolyte in dry
CH3CN.
Figure 7. Absorption spectra of the different electrolytes: electrolyte 1
ꢀ
for devices A based on Brꢀ/Br3 containing 0.9m DMBIBr, 0.08m Br2
and 0.5m TBP in dry CH3CN; electrolyte 2 for devices B also based on
ꢀ
Brꢀ/Br3 containing 0.9m DMBIBr, 0.8m LiBr, 0.08m Br2, and 0.5m TBP
ꢀ
in dry CH3CN; and electrolyte 3 for devices C based on Iꢀ/I3 containing
0.6m DMPII, 0.06m LiI, 0.04m I2, and 0.4m TBP in dry CH3CN.
Figure 8. Electrochemical impedance spectra, scanned from 10ꢀ1 to
5
~
&
*
10 Hz at RT, for devices A ( ), B ( ), and C ( ), respectively. a, b) Ny-
quist plots and Bode phase plots for TC308-based DSC; c, d) Nyquist
plots and Bode phase plots for TC310-based DSC. The cells were mea-
sured at ꢀ0.7 V in the dark. The alternating current (AC) amplitude was
set at 10 mV. Devices A are based on 0.9m DMBIBr, 0.08m Br2, and
0.5m TBP electrolyte in dry CH3CN; devices B are based on 0.9m
DMBIBr, 0.8m LiBr, 0.08m Br2, and 0.5m TBP electrolyte in dry
CH3CN; and devices C are based on 0.6m DMPII, 0.06m LiI, 0.04m I2,
and 0.4m TBP electrolyte in dry CH3CN.
of electrolyte and the Fermi level of TiO2, but also the sup-
pressed electron recombination.
For a DSC, the collection yield of injected electrons de-
pends on the effective electron-diffusion length (Ln). The
longer Ln results in better collection yield of injected elec-
ꢀ
ꢀ
trons. Upon replacement of Iꢀ/I3 with Brꢀ/Br3 , the Ln of
13134
www.chemeurj.org
ꢀ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2010, 16, 13127 – 13138

Organic Dyes
FULL PAPER
Table 4. The detailed parameters extracted from electrochemical impe-
dance spectra of DSC based on TC308 and TC310 and different electro-
lytes.
Dye
Device[a]
t
Rct
Rt
[ohmcmꢀ2
Ln
[c]
[d]
[ms][b]
G
]
G
]
[mm][e]
TC308
A
B
C
A
B
C
78
73
57
38
35
16
34
37
25
25
29
23
28
29
33
12
13
7
18
18
12
23
24
29
TC310
[a] Devices A are based on 0.9m DMBIBr, 0.08m Br2, and 0.5m TBP
electrolyte in dry CH3CN; devices B are based on 0.9m DMBIBr, 0.8m
LiBr, 0.08m Br2, and 0.5m TBP electrolyte in dry CH3CN; and devices C
are based on 0.6m DMPII, 0.06m LiI, 0.04m I2, and 0.4m TBP electrolyte
in dry CH3CN. [b] The electron lifetime (t) is expressed for electron re-
combination in TiO2 films and can be extracted from the angular fre-
&
Figure 9. Time-resolved transient-absorption spectra of TC308 ( ) and
~
TC310 ( ) on TiO2 films excited at 355 nm.
quency (wmin) at the midfrequency peak in the Bode phase plot by using
[20]
t=1/wmin
.
[c] Rct represents the charge-transfer resistance at the dye/
(TC308+) and TC310 (TC310+), respectively. The decay of
the transient-absorption signal of TC308+ and TC310+ were
recorded at 450 and 600 nm, respectively (Figure 10), and
the related data are collected in Table 5.
TiO2/electrolyte interface related to electron recombination. [d] Rt repre-
sents the electron-transport resistance in TiO2 film. [e] The effective elec-
tron-diffusion length (Ln) in TiO2 films reflects the competition between
1
2
charge collection and recombination: Ln =L
A
thickness of TiO2 film).[21]
The kinetics of TC308+ transient-absorbance decay exhib-
1
ited a typical half-lifetime of t = =92.3 ms. In the presence of
2
electrolytes, the decays of TC308+ were all significantly ac-
devices A and B became longer than that of devices C for
the DSC sensitized by TC308, thus a better collection yield
of injected electrons was obtained. However, for the DSC
sensitized by TC310, the Ln became shorter, which accounts
for the reduced collection yield of injected electrons ob-
ꢀ
ꢀ
tained upon replacement of Iꢀ/I3 with Brꢀ/Br3 . This is one
of main reasons why distinctive efficiencies were had for
DSCs sensitized by TC308 and TC310 upon replacement of
ꢀ
Brꢀ/Br3ꢀ with Iꢀ/I3 .
Dye regeneration: The solar-energy-conversion efficiency of
a DSC is strongly dependent on the kinetic competition be-
tween electron backtransfer from the conduction band of
TiO2 to the oxidized state of the dye (S+) and the intercep-
tion of S+ by the redox mediator in electrolyte, and is di-
rectly proportional to the dye-regeneration yield Fr, as
shown in Equation (1):
Fr ¼ kr=ðkrþkbÞ
ð1Þ
in which kr and kb are rate constants of dye regeneration
that occurs in the presence of redox mediator and electron
backtransfer that takes place between S+ and the conduc-
tion-band electron, respectively.[22]
The different energy gaps between the HOMO levels of
the dyes and the redox potential of electrolytes can lead to
different driving forces, and thus different Fr. To scrutinize
the different Fr that occurs in the presence of different elec-
trolytes, nanosecond laser transient-absorbance measure-
ments of TiO2 films sensitized by TC308 and TC310 were
performed. The time evolutions of the transient absorbance
are shown in Figure 9. The signals over wavelength ranges
that extend from 380 to 500 nm and from 430 to 700 nm are
assigned to the absorbance of the oxidized form of TC308
Figure 10. Transient-absorbance decay kinetics of the oxidized form of
a) TC308 and b) TC310 adsorbed on a mesoporous TiO2 film in different
electrolytes. Electrolyte 2 is based on the Brꢀ/Br3ꢀ couple, and electrolyte
ꢀ
3 is based on the Iꢀ/I3 couple. The dyes were adsorbed from solutions in
CH2Cl2. Absorbance changes were measured at a probe wavelength of
450 nm for TC308 and 600 nm for TC310 and employed 355 nm laser ex-
citation. The continuous lines drawn on top of the experimental data are
single-exponential fit curves.
Chem. Eur. J. 2010, 16, 13127 – 13138
ꢀ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
13135

X. Yang, L. Sun et al.
Table 5. The lifetimes and the rate constants extracted from transient-ab-
sorbance measurements of TC308 and TC310 absorbed TiO2 films in dif-
ferent electrolytes.
suppressed electron recombination. Electrochemical impe-
dance spectra studies reveal that the effective electron-diffu-
sion lengths (Ln) become longer upon using Brꢀ/Br3ꢀ instead
k [ꢁ105 sꢀ1
]
ꢀ
1
of Iꢀ/I3 for the dyes for which the HOMO levels are more
Dye
Electrolyte
t = [ms]
2
ꢀ
positive than the redox potential of Brꢀ/Br3 , but become
TC308
CH3CN only
electrolyte 2[a]
electrolyte 3[b]
CH3CN only
electrolyte 2[a]
electrolyte 3[b]
92.3
6.4
6.8
39.8
23.4
6.7
0.11
1.56
1.48
0.25
0.43
1.50
shorter for the dyes for which the HOMO levels are similar
ꢀ
to the redox potential of Brꢀ/Br3 . IPCE action spectra and
TC310
nanosecond laser transient-absorption study reveal that the
dye-regeneration yields (Fr) depend on the potential differ-
ence between the HOMO levels of the dyes and the redox
ꢀ
[a] Electrolyte 2 is based on a Brꢀ/Br3 couple. [b] Electrolyte 3 is based
on an Iꢀ/I3ꢀ couple.
ꢀ
couple of Brꢀ/Br3 .
ꢀ
1
celerated with t = =6.4 ms for electrolyte that contained Br /
Experimental Section
2
ꢀ
ꢀ
ꢀ
1
Br3 and t = =6.8 ms for I /I3 , respectively. Remarkably, the
2
Analytical methods and measurements: Absorption and emission spectra
were recorded in a quartz cell with 1 cm path length using Agilent HP
8453 (USA) and Photon Technology International 700 (USA) spectro-
dye-regeneration kinetics can be described by approximate
single-exponential decays with rate constant kr =1.56ꢁ
ꢀ
105 sꢀ1 for electrolyte that contains Brꢀ/Br3 and kr =1.48ꢁ
photometers, respectively. 1H NMR spectra were measured using
a
ꢀ
105 sꢀ1 for Iꢀ/I3 , respectively. The data for the electron
Varian INOVA 400 MHz (USA) with the chemical shifts measured
against TMS. MS data were obtained using GCT CA156 (UK), HP1100
LC/MSD (USA), and LC/Q-TOF MS (UK) instruments. Electrochemical
redox potentials were obtained by cyclic voltammetry (CV) using a
three-electrode cell and an electrochemistry workstation (BAS100B,
USA). The working electrode was a glass carbon disc electrode; the aux-
iliary electrode was a Pt wire, and Ag/Ag+ was used as reference elec-
trode. Tetrabutylammonium hexaflourophosphate (TBAPF6, 0.1m) was
used as supporting electrolyte in DMF. Ferrocene was added to each
sample solution at the end of the experiments and the ferrocenium/ferro-
cene (Fc/Fc+) redox couple was used as an internal potential reference.
The potentials versus NHE were calibrated by the addition of 630 mV to
the potentials versus Fc/Fc+.[11] Electrochemical impedance spectroscopy
(EIS) for DSC in the dark with bias of ꢀ0.7 V was measured using an im-
pedance/gain-phase analyzer (PARSTAT 2273, USA). The spectra were
scanned in a frequency range of 10ꢀ1–105 Hz at RT. The alternating cur-
rent (AC) amplitude was set at 10 mV. Transient-absorption measure-
ments were carried out using a nanosecond-laser flash photolysis setup
(LP920, Edinburgh Instrument Ltd., UK). Excitation pulses at 355 nm
(1.1 mJ, 7 ns full width at half-maximum) were obtained using a Quanta-
Ray master optical parametric oscillator (MOPO) pumped by a Quanta-
Ray 230 Nd:YAG laser (355 nm). The probe light was provided by a
75WXe arc lamp and was collinear with the excitation beam. After pass-
ing through the sample, the probe light was spectrally filtered with two
monochromators and finally detected using a Hamamatsu R928 photo-
multiplier tube. Individual trace kinetics were analyzed using the decon-
volution software Spectra Solve.[23]
backtransfer kinetics could also be fitted approximately with
a single-exponential decay, a rate constant of kb =1.1ꢁ
104 sꢀ1, thus leading to Fr of 93.4% for electrolyte that con-
tained Brꢀ/Br3 and 93.1% for Iꢀ/I3 , respectively. This
ꢀ
ꢀ
result reveals that TC308 can be regenerated efficiently by
ꢀ
ꢀ
both Brꢀ/Br3 and Iꢀ/I3 . On the other hand, the Fr of
TC310 are determined to be 63.2% for electrolyte that con-
ꢀ
ꢀ
tains Brꢀ/Br3 and 85.7% for Iꢀ/I3 , respectively. This means
ꢀ
that TC310 can only be regenerated efficiently by Iꢀ/I3 and
ꢀ
not by Brꢀ/Br3 . The HOMO level of TC308 is 1.43 V,
ꢀ
which is much more positive than both E
N
(Iꢀ/I3 ), thus the driving forces of dye regeneration are suffi-
ꢀ
ꢀ
ꢀ
cient with both Brꢀ/Br3 and Iꢀ/I3 . However, the HOMO
ꢀ
level of TC310 is more positive than E
N
ꢀ
E
(Brꢀ/Br3 ), thus the driving force of the reduction of
ꢀ
TC310+ is only sufficient with Iꢀ/I3 and insufficient with
ꢀ
Brꢀ/Br3 .
Conclusion
We have synthesized a series of novel metal-free organic
dyes TC301–TC310 with relatively high HOMO levels for
application in DSCs based on electrolytes that contain Brꢀ/
DSC fabrication: A layer of approximately 2 mm TiO2 (13 nm paste,
DHS-TPP3, Heptachroma, China) was coated on the F-doped tin oxide
conducting glass (TEC15, 15 W per square, Pilkington, USA) by screen
printing and then dried for 6 min at 1258C. This procedure was repeated
6 times (around 12 mm) and finally coated by a layer (around 4 mm) of
TiO2 paste (DHS-SLP1, Heptachroma, China) as scattering layer. The
double-layer TiO2 electrodes (area: 6ꢁ6 mm) were gradually heated
under an air flow at 3258C for 5 min, at 3758C for 5 min, at 4508C for
15 min, and at 5008C for 15 min. The sintered film was further treated
with 40 mm TiCl4 aqueous solution at 708C for 30 min, then washed with
ethanol and water, and annealed at 5008C for 30 min. After the film was
cooled to 408C, it was immersed into a 2ꢁ10ꢀ4 m dye bath in CH2Cl2 so-
lution and maintained in the dark for 8 h. The electrode was then rinsed
with CH2Cl2 and dried. The hermetically sealed cells were fabricated by
assembling the dye-loaded film as the working electrode and Pt-coated
ꢀ
ꢀ
Br3 and Iꢀ/I3 . Our purposes for the dye design are sum-
marized by two major points: 1) to study the Li+-ion effects
on the photoelectrochemical properties of the dyes and the
performance of the DSC, and 2) to scrutinize the relation-
ship between the HOMO levels of the dyes and the perfor-
mance of the DSC based on electrolytes that contain Brꢀ/
ꢀ
ꢀ
Br3 and Iꢀ/I3 . The addition of Li+ ions in electrolytes can
broaden the absorption spectra of the dyes on TiO2 films
and shift the LUMO levels of the dyes and the conduction
band of TiO2 towards positive positions, thereby leading to
the increase in Jsc and the decrease in Voc. Upon using Brꢀ/
conducting glass as the counterelectrode separated with
a hot-melt
Surlyn 1702 film (25 mm, Dupont). The electrolyte was introduced into
the cell by means of vacuum backfilling from a hole in the back of the
counterelectrode. Finally, the hole was also sealed using Surlyn 1702 film
and cover glass.
ꢀ
ꢀ
Br3 instead of Iꢀ/I3 , a large increase of Voc is attributed to
not only the enlarged energy gap between the redox poten-
tial of electrolyte and the Fermi level of TiO2, but also the
13136
www.chemeurj.org
ꢀ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2010, 16, 13127 – 13138

Organic Dyes
FULL PAPER
Scheme 2. Synthetic routes of the dyes TC301–TC310. a) compound 1a, tBuCl, RT, 24 h; b) 4-bromobenzaldehyde, PdACTHNUGTRNEUNG(OAc)2, tBu3P, toluene, reflux,
24 h; c) KI, KIO3, CH3COOH, 808C, 4 h; d) cyanoacetic acid, piperidine, CH3CN, reflux, 3 h; e) 1) NaBH4, CH2Cl2, C2H5OH, RT, 2 h; 2) PPh3·HBr,
CHCl3, reflux, 2 h; 3) terephthalaldehyde or thiophene-2,5-dicarbaldehyde, [18]crown-6, K2CO3, DMF, RT, 2 h; 4) I2, THF, reflux, 8 h; f) cyanoacetic
acid, piperidine, CH3CN, reflux, 3 h; g) Br2, CHCl3, ice–water, 3 h; h) KI, KIO3, CH3COOH, 808C, 4 h; p) compound 1a, PdACHTUNRTGENUNG(OAc)2, tBu3P, toluene,
reflux, 24 h; i) POCl3, DMF, ClCH2CH2Cl, reflux, 6 h; j) cyanoacetic acid, piperidine, CH3CN, reflux, 3 h; k) 1) NaBH4, CH2Cl2, C2H5OH, RT, 2 h;
2) PPh3·HBr, CHCl3, reflux, 2 h; 3) terephthalaldehyde, [18]crown-6, K2CO3, DMF, RT, 2 h; 4) I2, THF, reflux, 8 h; and l) cyanoacetic acid, piperidine,
CH3CN, reflux, 3 h.
Photovoltaic properties measurements: The irradiation source for the
photocurrent–voltage (J–V) measurement was an AM 1.5 solar simulator
(16S-002, SolarLight Co. Ltd., USA). The incident light intensity was
100 mWcmꢀ2 calibrated with a standard Si solar cell. The tested solar
cells were masked to a working area of 0.159 cm2. The current–voltage
curves were obtained by the linear-sweep voltammetry (LSV) method
using an electrochemical workstation (LK9805, Lanlike Co. Ltd., China).
The measurement of the incident photon-to-current conversion efficiency
(IPCE) was performed using a Hypermonolight (SM-25, Jasco Co. Ltd.,
Japan).
Preparation of the sample for laser-flash measurements: The transparent
TiO2 film (approximately 3 mm thick) was prepared from a commercial
TiO2 paste (DHS-TPP3, Heptachroma, China) by
a doctor-blading
method and sintering at 5008C for 30 min in air. After the film was
cooled to RT, it was immersed in a 2ꢁ10ꢀ4 m dye solution in CH2Cl2 and
kept in darkness for 5 min. The TiO2 films with anchoring dyes were cut
into the same size (2.5 cmꢁ1.3 cm) for laser-flash measurements.
Synthesis: The synthetic routes of the dyes TC301–TC310 are shown in
Scheme 2, and the detailed synthetic procedures are described in Sup-
porting Information.
Chem. Eur. J. 2010, 16, 13127 – 13138
ꢀ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
13137

X. Yang, L. Sun et al.
[14] K. Hara, T. Sato, R. Katoh, A. Furube, T. Yoshihara, M. Murai, M.
Kurashige, S. Ito, A. Shinpo, S. Suga, H. Arakawa, Adv. Funct.
Mater. 2005, 15, 246.
[15] A. Furube, R. Katoh, K. Hara, T. Sato, S. Murata, H. Arakawa, M.
Tachiya, J. Phys. Chem. B 2005, 109, 16406.
Acknowledgements
We gratefully acknowledge the financial support of this work from the
following sources: China Natural Science Foundation (grant no.
20633020), the National Basic Research Program of China (grant no.
2009CB220009), the Ministry of Science and Technology (MOST) (grant
no. 2001CCA02500), the Ministry of Education (MOE), the Program for
Changjiang Scholars and Innovation Research Team in university
(PCSIRT), the Swedish Energy Agency, the Swedish Research Council,
and the K&A Wallenberg Foundation. The authors are grateful to Profes-
sor Can Li, Dr. Jingying Shi, and Min Zhong at the Dalian Institute of
Chemical Physics, CAS, China, for EIS measurement and helpful discus-
sions.
[16] D. P. Hagberg, T. Edvinsson, T. Marinado, G. Boschloo, A. Hagfeldt,
L. Sun, Chem. Commun. 2006, 2245.
[17] a) Gaussian 03, Revision C.02, M. J. Frisch, G. W. Trucks, H. B.
Schlegel, G. E. Scuseria, M. A. Robb, J. R. Cheeseman, J. A. Mont-
gomery, Jr., T. Vreven, K. N. Kudin, J. C. Burant, J. M. Millam, S. S.
Iyengar, J. Tomasi, V. Barone, B. Mennucci, M. Cossi, G. Scalmani,
N. Rega, G. A. Petersson, H. Nakatsuji, M. Hada, M. Ehara, K.
Toyota, R. Fukuda, J. Hasegawa, M. Ishida, T. Nakajima, Y. Honda,
O. Kitao, H. Nakai, M. Klene, X. Li, J. E. Knox, H. P. Hratchian,
J. B. Cross, V. Bakken, C. Adamo, J. Jaramillo, R. Gomperts, R. E.
Stratmann, O. Yazyev, A. J. Austin, R. Cammi, C. Pomelli, J. W.
Ochterski, P. Y. Ayala, K. Morokuma, G. A. Voth, P. Salvador, J. J.
Dannenberg, V. G. Zakrzewski, S. Dapprich, A. D. Daniels, M. C.
Strain, O. Farkas, D. K. Malick, A. D. Rabuck, K. Raghavachari,
J. B. Foresman, J. V. Ortiz, Q. Cui, A. G. Baboul, S. Clifford, J. Cio-
slowski, B. B. Stefanov, G. Liu, A. Liashenko, P. Piskorz, I. Komaro-
mi, R. L. Martin, D. J. Fox, T. Keith, M. A. Al-Laham, C. Y. Peng,
A. Nanayakkara, M. Challacombe, P. M. W. Gill, B. Johnson, W.
Chen, M. W. Wong, C. Gonzalez, J. A. Pople, Gaussian, Inc.: Wall-
ingford, CT, 2004; b) A. D. Becke, J. Chem. Phys. 1993, 98, 5648;
c) R. Ditchfield, W. J. Herhe, J. A. Pople, J. Chem. Phys. 1971, 54,
724.
[18] G. Redmond, D. Fitzmaurice, J. Phys. Chem. 1993, 97, 1426.
[19] G. Oskam, B. V. Bergeron, G. J. Meyer, P. C. Searson, J. Phys. Chem.
B 2001, 105, 6867.
[20] R. Kern, R. Sastrawan, J. Ferber, R. Stangl, J. Luther, Electrochim.
Acta 2002, 47, 4213.
[21] F. Fabregat-Santiago, J. Bisquert, L. Cevey, P. Chen, M. Wang, S. M.
Zakeeruddin, M. Grꢂtzel, J. Am. Chem. Soc. 2009, 131, 558.
[22] D. Kuang, S. Ito, B. Wenger, C. Klein, J. Moser, R. Humphry-Baker,
S. M. Zakeeruddin, M. Grꢂtzel, J. Am. Chem. Soc. 2006, 128, 4146.
[23] Spectra Solve 2.01, Lastek Pty. Ltd., 1997.
[1] a) B. O’Regan, M. Grꢂtzel, Nature 1991, 353, 737; b) A. Hagfeldt,
M. Grꢂtzel, Acc. Chem. Res. 2000, 33, 269.
[2] a) M. Grꢂtzel, Pure Appl. Chem. 2001, 73, 459; b) C. Nasr, S. Hotch-
andani, P. V. Kamat, J. Phys. Chem. B 1998, 102, 4944.
[3] A. Hagfeldt, M. Grꢂtzel, Chem. Rev. 1995, 95, 49.
[4] M. Grꢂtzel, J. Photochem. Photobiol. A 2004, 164, 3.
[5] a) B. A. Gregg, F. Pichot, S. Ferrere, C. R. Fields, J. Phys. Chem. B
2001, 105, 1422; b) B. A. Gregg, Coord. Chem. Rev. 2004, 248, 1215.
[6] a) H. Nusbaumer, J.-E. Moser, S. M. Zakeeruddin, M. K. Nazeerud-
din, M. Grꢂtzel, J. Phys. Chem. B 2001, 105, 10461; b) H. Nusbaum-
er, S. M. Zakeeruddin, J.-E. Moser, M. Grꢂtzel, Chem. Eur. J. 2003,
9, 3756; c) S. A. Sapp, C. M. Elliott, C. Contado, S. Caramori, C. A.
Bignozzi, J. Am. Chem. Soc. 2002, 124, 11215.
[7] Z. Zhang, P. Chen, T. N. Murakami, S. M. Zakeeruddin, M. Grꢂtzel,
Adv. Funct. Mater. 2008, 18, 341.
[8] a) S. Ito, P. Chen, P. Comte, M. K. Nazeeruddin, P. Liska, P. Pꢃchy,
M. Grꢂtzel, Prog. Photovoltaics 2007, 15, 603; b) G. Redmond, D.
Fitzmaurice, J. Phys. Chem. 1993, 97, 1426.
[9] Z. Wang, K. Sayama, H. Sugihara, J. Phys. Chem. B 2005, 109,
22449.
[10] C. Teng, X. Yang, C. Yuan, C. Li, R. Chen, H. Tian, S. Li, A. Hag-
feldt, L. Sun, Org. Lett. 2009, 11, 5542.
[11] P. V. Kamat, Chem. Rev. 1993, 93, 267.
[12] C. Reichardt, Chem. Rev. 1994, 94, 2319.
[13] S. Roquet, A. Cravino, P. Leriche, O. AlꢃvÞque, P. Frꢄre, J. Roncali,
J. Am. Chem. Soc. 2006, 128, 3459.
Received: February 22, 2010
Published online: October 4, 2010
13138
www.chemeurj.org
ꢀ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2010, 16, 13127 – 13138