Photophysical studies of dipolar organic dyes that feature a 1,3-cyclohexadiene conjugated linkage: The implication of a twisted intramolecular charge-transfer state on the efficiency of dye-sensitized solar cells

Article abstract of DOI:10.1002/chem.201001294

A detailed study of the synthesis and photophysical properties of a new series of dipolar organic photosensitizers that feature a 1,3-cyclohexadiene moiety integrated into the ?-conjugated structural backbone has been carried out. Dye-sensitized solar cel

Full text of DOI:10.1002/chem.201001294

FULL PAPER
DOI: 10.1002/chem.201001294
Photophysical Studies of Dipolar Organic Dyes That Feature a 1,3-
Cyclohexadiene Conjugated Linkage: The Implication of a Twisted
Intramolecular Charge-Transfer State on the Efficiency of Dye-Sensitized
Solar Cells
Kuan-Fu Chen,[a, b] Che-Wei Chang,[a] Ju-Ling Lin,[a] Ying-Chan Hsu,[a]
Ming-Chang P. Yeh,*[b] Chao-Ping Hsu,*[a] and Shih-Sheng Sun*[a]
Abstract: A detailed study of the syn-
thesis and photophysical properties of
a new series of dipolar organic photo-
sensitizers that feature a 1,3-cyclohexa-
diene moiety integrated into the p-con-
jugated structural backbone has been
carried out. Dye-sensitized solar cells
(DSSCs) based on these structurally
simple dyes have shown appreciable
photo-to-electrical energy conversion
efficiency, with the highest one up to
4.03%. Solvent-dependent fluorescence
studies along with the observation of
dual emission on dye 4b and single
emission on dyes 4a and 32 suggest
that dye 4b possesses a highly polar
emissive excited state located at
a
states to the lowest TICT state renders
the back-electron transfer process a
forbidden one and significantly retards
the charge recombination to boost the
photocurrent. The electrochemical im-
pedance under illumination and transi-
ent photovoltage decay studies showed
smaller charge resistance and longer
electron lifetime in 4b-based DSSC
compared to the DSSCs with reference
dyes 4a and 32, which further illus-
trates the positive influence of the
TICT state on the performance of
DSSCs.
lower-energy position than at the
normal emissive excited state. A de-
tailed photophysical investigation in
conjunction with computational studies
confirmed the twisted intramolecular
charge-transfer (TICT) state to be the
lowest emissive excited state for dye
4b in polar solvents. The relaxation
from higher-charge-injection excited
Keywords:
cyclohexadiene · density functional
calculations dyes/pigments
photochemistry
charge transfer
·
·
·
Introduction
(DSSCs) based on organic and coordination complexes have
received great attention since the seminal report by Grꢀtzel
and co-workers in 1991.[1] RuII-based complexes such as N3,
N719, black dye, and others have shown the best photocon-
version efficiency—up to approximately 11%—under air
mass (AM) 1.5 irradiation thus far.[2] Although the efficiency
of DSSCs based on organic dyes are generally lower than
the cells based on RuII dyes, the highest photoconversion ef-
ficiency of organic dye-sensitized solar cells has exceeded
9%.[3] The advantages of organic dyes over RuII-based
metal complexes in DSSCs include the typically high molar
absorption coefficients of organic dyes (e<20000mꢀ1 cmꢀ1),
a variety of specific functional groups of organic dyes avail-
able for tuning the absorption spectra coverage, the relative-
ly low costs of organic dye compared with those of rutheni-
um-based metal complexes, and essentially no limitation of
resources.[4]
The search for clean and alternative energy has become an
important and difficult issue in the 21st century. Among sev-
eral new energy technologies, solar energy seems to be one
important answer to this question. Dye-sensitized solar cells
[a] K.-F. Chen, C.-W. Chang, J.-L. Lin, Dr. Y.-C. Hsu, Dr. C.-P. Hsu,
Dr. S.-S. Sun
Institute of Chemistry, Academia Sinica (Taiwan, ROC)
Fax : (+886)2-2783-1237
E-mail: cherri@chem.sinica.edu.tw
sssun@chem.sinica.edu.tw
[b] K.-F. Chen, Prof. M.-C. P. Yeh
Department of Chemistry
National Taiwan Normal University (Taiwan, ROC)
Fax : (+886)2-2932-4249
E-mail: cheyeh@ntnu.edu.tw
The generation of photocurrents in a DSSC occurs
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/chem.201001294.
through the following processes.[5] First, the dye sensitizer
Chem. Eur. J. 2010, 16, 12873 – 12882
ꢁ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
12873

absorbs a photon (sunlight) to generate the photoexcited
state of the dye (S*). Next, the photoexcited dye (S*) injects
an electron into the conduction band (CB; ꢀ0.5 V versus
the normal hydrogen electrode (NHE)) of TiO2. The result-
ing oxidized dye is subsequently reduced back to its original
tion of a TICT state would generate a significant charge sep-
aration between the donor–acceptor pair and thus a large
dipole moment in the excited state. A notable characteristic
of the TICT state is its nearly electronically decoupled
donor–acceptor pair with forbidden radiative transition
from the TICT state to the ground state. We envision that
an organic dye with a strong electronically coupled dipolar
donor–acceptor bridged by a suitable p-conjugated spacer
with an accessible low-lying TICT state could be advanta-
geous for DSSCs. The relaxation from the Franck–Condon
state to the TICT state upon photoexcitation of the dipolar
dye followed by charge injection into the TiO2 conduction
band would enhance the photocurrent due to the forbidden
nature of charge recombination between the injected elec-
tron and the oxidized dye.
Previously, we have reported a series of structurally
simple dipolar organic dyes that feature a cyclohexadiene
moiety in the conjugation framework.[12] The DSSCs assem-
bled based on these dyes showed appreciable efficiency of
up to 4.0% relative to the 5.8% of N719 dye efficiency
under the same experimental conditions. The relatively
good performance of these dyes is partially ascribed to the
compact structures that increased the dye loading onto the
TiO2 surface and compensated for the narrow spectral cov-
erage of light harvesting. The comparable dye loading and
absorption spectral coverage between dyes 4a and 4b would
lead to the expectation of similar light-harvesting ability and
photocurrent density generated between these two dyes.
However, the incident photon-to-current efficiency (IPCE)
value at absorption maxima and photocurrent density of the
4a-based device are almost 20% less than the 4b-based
device. Herein, we report the photophysical studies, electro-
chemical impedance spectroscopy, and transient photovolt-
age measurements for dyes 4a, 4b, and model compound 32
to elucidate the possible TICT mechanism and its implica-
tion on the efficiency of DSSCs. The results have clearly
shown that the TICT state in dye 4b plays an important role
in the photocurrent process.
ꢀ
state by electron donation from the redox mediator (Iꢀ/I3 )
in the electrolyte solution. The injected electrons move
through the network of interconnected TiO2 nanoparticles
to arrive at the transparent conducting oxide (fluorine-
doped tin oxide, FTO) and then through the external circuit
to the counter-electrode (Pt-coated glass). The Iꢀ ion is re-
ꢀ
generated by the reduction of the triiodide ion (I3 ) at the
counterelectrode through the donation of electrons from the
external circuit. Then the circuit is completed. During this
electron-flow cycle, there are undesirable side processes: the
electrons injected into the CB of the TiO2 electrode may re-
ꢀ
combine either with oxidized dye, or with I3 at the TiO2
surface, thus resulting in lower photovoltaic performances of
the DSSCs.
Thus, one of the key issues to enhance the performance of
a DSSC is how efficient the system is able to shuffle the
hole generated from a photoexcited dye away from the TiO2
surface to prevent the undesirable charge-recombination
processes. The electron density of the HOMOs in thiocya-
nate-substituted RuII dyes is typically spread over the RuII
center and NCS ligands, which reduces the probability of
charge recombination.[6] A similar strategy of integration of
electron-rich moieties such as thiophene or furan in the an-
cillary ligand in RuII complexes or metal free dyes also par-
tially delocalizes the HOMO electron density distribution
away from TiO2 surface.[7] Organic dyes with a structural
framework that features a D–D–p–A arrangement have
been designed for creating charge-transfer cascades to en-
hance charge separation.[8] Incorporation of secondary elec-
tron-transfer cascades within the dye structure to increase
the separation between the oxidized dye and the TiO2 sur-
face is also capable of effectively retarding the charge-re-
combination kinetics and produces a long-lived charge-sepa-
rated state.[9] On the other hand, an excited-state conforma-
tional distortion after charge injection to render the charge
recombination a forbidden process would be an ideal alter-
native to slow down the charge-recombination kinetics be-
tween the injected electron and the oxidized dye.
Photoinduced intramolecular charge transfer (ICT) in a
strong electron donor–acceptor-substituted conjugated
system normally displays
a prominent solvatochromic
effect.[10] In a nonpolar solvent, the locally excited (LE)
state takes a planar conformation and the excited state is
stabilized with increasing solvent polarity. In some extreme
cases, the donor and acceptor fragments are twisted and the
excited molecule is further relaxed to a twisted intramolecu-
lar charge-transfer (TICT) state in polar solvents.[11] The
representative example with such low-lying TICT state in
Results and Discussion
Synthesis of the materials: The synthesis of 4a and 4b has
been reported previously.[12] Scheme 1 describes the synthe-
sis of dye 32. This molecule was designed to serve as a
model compound by fusing the benzene phenylamino
polar
solvents
is
N,N-dimethylaminobenzonitrile
ꢀ
(DMABN), in which twisting about the benzonitrilo–dime-
moiety to prevent the twisting of the C N bond in the excit-
ed states to assess the accessibility of the TICT state in 4a
and 4b. The key intermediate, 5-bromo-1-phenylindoline,
ꢀ
thylamino C N bond in the LE state has been proposed to
be responsible for the observed dual emission.[11] The forma-
12874
www.chemeurj.org
ꢁ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2010, 16, 12873 – 12882

Dipolar Organic Dyes
FULL PAPER
Scheme 1. Synthesis of dye 32.
ꢀ
was prepared by palladium-catalyzed Buchwald–Hartwig C
N coupling reaction of 5-bromoindoline with 1-iodobenzene
according to the published methods.[13] All three dyes with
1,3-cyclohexadiene in the struc-
Figure 1. Electronic absorption spectra of 4a, 4b, and 32 in acetonitrile.
tural framework exhibit fairly
[a]
Table 1. Absorption (labs) and fluorescence (lem
)
maxima, Stokes shifts (Dnst),[b] fluorescence quantum yields
good thermal stability and show
no decomposition up to 2008C
as evidenced by the thermogra-
vimetric analysis (TGA) data.
(Ff), lifetimes (t), and radiative (kr) and nonradiative (knr) rate constants for dyes 4a, 4b, and 32 in different
solvents.
Dye
Solvent
labs [nm]
lem [nm]
Dnst [cmꢀ1
]
Ff
t [ns]
kr [108 sꢀ1
]
knr [108 sꢀ1
]
4a
diethyl ether
toluene
THF
380
400
385
392
398
400
384
389
391
411
395
399
406
413
430
481
434
435
458
530
423
430
448
496
1685
787
0.11
0.14
0.20
0.08
0.11
0.12
0.14
0.03
0.07
0.14
0.10
0.08
1.5
1.5
1.6
2.5
1.8
1.7
1.7
0.72
0.93
1.21
0.32
0.60
0.66
0.80
0.08[e]
0.44
0.84
0.58
0.31
5.82
5.79
4.92
3.72
4.93
5.09
4.98
2.75[e]
5.97
4.98
5.20
3.57
2718
4720
2084
2011
4208
6839
1935
1075
2995
4902
Electronic absorption spectra:
The absorption spectra in ace-
tonitrile for the three dyes are
shown in Figure 1. The related
photophysical data are summar-
ized in Table 1. The general fea-
tures of the absorption spectra
consist of one broad band in
the visible region and a less in-
tense absorption band in the
near-UV region, which corre-
spond to the ICT transition and
the locally excited p–p* transi-
tion, respectively.[12] The ab-
MeCN
4b
32
diethyl ether
toluene
THF
MeCN
2.2,[c] 3.6[d]
1.6
1.7
1.7
2.6
diethyl ether
toluene
THF
MeCN
[a] Fluorescence data are from corrected spectra. [b] Dnst =nabsꢀnem. [c] Fitted lifetime monitored at a wave-
length of 450 nm. [d] Fitted lifetime monitored at a wavelength of 600 nm. [e] The values were calculated
based on the quantum yield of the whole emission profile and lifetime value of 3.6 ns.
sorption spectra for all three dyes vary little with a differ-
ence of only a few nanometers with respect to the solvent
polarity, thus suggesting a small difference in dipole mo-
ments between the ground state and Franck–Condon excited
state. The slightly more redshifted ICT band in dye 32 than
the ICT bands in dyes 4a and 4b is in accordance with the
more planar geometry in 32 due to the structural fixation of
rescence maximum upon increasing solvent polarity. The
emission maxima of bathochromic 4a shift from 406 nm in
diethyl ether to 481 nm in acetonitrile, whereas the emission
maxima of bathochromic 32 shift from 423 nm in diethyl
ether to 496 nm in acetonitrile. The emission is apparently
dominated by ICT for both 4a and 32. The related photo-
physical parameters including quantum yields and excited-
state lifetimes for these two dyes are also comparable.
ꢀ
the C N bond rotation.
The emission of 4b showed even more prominent batho-
chromic shift than 4a and 32 with increasing solvent polari-
ty, which indicates a larger dipole moment for the emissive
state of 4b than 4a and 32 (see Figure 2). It is consistent
with the order of the obtained excited-state dipole moments
(me) in which the me values follow an order of 4a<32<4b
(vide infra). However, unlike the cases of 4a and 32 in
which the excited-state decay in acetonitrile showed single
exponential process, the excited-state decay of 4b in aceto-
nitrile displayed a biexponential process with fitted lifetime
values of 2.2 and 3.6 ns for monitoring the emission profile
Fluorescence properties and solvatochromic effect: The
fluorescence quantum yields, excited-state lifetimes, and cal-
culated radiative and nonradiative rate constants for dyes
4a, 4b, and 32 in diethyl ether, toluene, THF, and acetoni-
trile at room temperature were collected in Table 1. In con-
trast to the minor solvent-dependent absorption spectral
features, the fluorescence spectra of 4a, 4b, and 32 vary sig-
nificantly with increasing solvent polarity, which is indicative
of a pronounced charge-transfer nature of the emissive
states. Both 4a and 32 exhibit a monotonic redshift of fluo-
Chem. Eur. J. 2010, 16, 12873 – 12882
ꢁ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
12875

S.-S. Sun, C.-P. Hsu, M.-C. P. Yeh et al.
moments of 4b from Lippert–Mataga analysis is larger than
the corresponding emissive excited-state dipole moments of
4a and 32. With a large emissive excited-state dipole
moment for 4b and comparable excited-state dipole mo-
ꢀ
ments for 4a and C N-bond-fused 32, it may be possible for
the emissive excited state of 4b in polar solvents to originate
from a TICT state.
In general, a TICT state is likely to form with a strong
electron donor–acceptor pair linked with a short p-conjugat-
ed framework in a polar medium in which the electron den-
sity of the charge-separated species is nearly localized on
the donor–acceptor ends. Han et al. have reported a thio-
phene p-conjugated donor–acceptor system with emission
that originates from a low-lying TICT state in polar sol-
vents.[11n] Tang and co-workers also reported a series of di-
phenylamine boron dipyrromethene dipolar p-conjugated
molecules that show emission from TICT in highly polar sol-
vents.[11m] A series of aminostilbenes reported by Yang et al.
has been demonstrated to possess an accessible TICT state,
and the molecules are able to relax to the TICT state upon
Figure 2. Normalized fluorescence spectra of 4a (top), 4b (middel), and
32 (bottom) in diethyl ether (black curve), toluene (pink curve), THF
(blue curve), CH2Cl2 (green curve), and acetonitrile (red curve).
photoexcitation in polar solvents by twisting the correspond-
[11b,o]
ꢀ
ꢀ
at 450 and 600 nm, respectively. The same excitation-spec-
tral profiles observed when monitoring at these two wave-
lengths indicate that the two excited states under equilibri-
um originate from the same ground-state species.[14]
The solvatochromic properties of the three dyes were fur-
ther examined by Lippert–Mataga analysis (Figure S1 in the
Supporting Information).[15] The excited-state dipole
moment (me) can be extracted from the slope of the solvato-
chromic plot (mf) of the Stokes shift (nst) against the solvent
parameter Df according to Equations (1–3):
ing C N bond.
In our current studies, when the C N
bond is fused to restrict the rotation around the benzene di-
ꢀ
phenylamino C N bond as in the case of 32, essentially all
the photophysical parameters for 32 are similar to 4a in
which typical ICT properties were observed. The radiative
rate constant of 4b in acetonitrile is four times smaller than
the corresponding values for 4a and 32. It should be noted
here that the calculated radiative rate constant for 4b was
based on the quantum yield of a full emission profile in
which two emissive states are under equilibrium, and thus
the obtained kr value represents only an upper limit for the
radiative rate constant. These results suggest that a less-per-
mitted radiative transition from the emissive state to the
ground state exists in 4b when compared to the excited-
state deactivation pathways of 4a and 32. The emissive
states of 4b in acetonitrile, therefore, are tentatively as-
signed to a higher-energy ICT state with a lifetime of 2.2 ns
and a lower-energy TICT state with a lifetime of 3.6 ns.
Upon relaxation from the Franck–Condon state, polar ace-
tonitrile can effectively stabilize the ICT state, which is fol-
1
Dnst ¼ ꢀ½ð = pe0Þð2=hca3Þꢁ½meðmeꢀmgÞꢁDf þ constant
ð1Þ
ð2Þ
ð3Þ
4
Df ¼ ðeꢀ1Þ=ð2e þ 1Þꢀ0:5ðn2ꢀ1Þ=ð2n2 þ 1Þ
1
=
3
a ¼ ð3M=4NpdÞ
in which mg is the ground-state dipole moment, e is the sol-
vent dielectric constant, e0 is the vacuum permittivity, n is
the solvent refractive index, h is the Planck constant, c is the
velocity of light, and a is the Onsager radius of solute, which
can be derived from the Avogadro number N, molecular
weight M, and density d. The value of mg was calculated
using density functional theory with the B3LYP/6-31G*
basis set. The obtained dipole moments of the emissive ex-
cited state and related parameters are summarized in
Table 2. Indeed, the obtained emissive excited-state dipole
ꢀ
lowed by further geometric twisting of the C N bond to
relax to a TICT state.
The torsional barriers for the deactivation process from
an ICT state to a TICT state were evaluated by tempera-
ture-dependent fluorescence in acetonitrile and DMSO. As-
suming such torsional motion is the only activated singlet-
decay process and the intersystem crossing is negligible, the
activation energy, Ea, for the torsional motion can be esti-
mated from Equation (4):[16]
Table 2. Ground- and excited-state dipole moments for 4a, 4b, and 32.
[b]
Dye
a [ꢂ][a]
mf [cmꢀ1
]
mg [D][c]
me [D]
lnðFfꢀ1ꢀ1Þ ¼ lnðA0=krÞꢀEa=RT
ð4Þ
4a
4b
32
4.66
5.35
5.1
9409
16457
8465
5.97
3.48
4.04
13.18
17.77
12.79
in which Ff is the fluorescence quantum yield measured at
temperature T, and A0 is the Arrhenius constant. An Arrhe-
nius plot of lnCAHTUNGTRENNNUG
[a] Onsager radius calculated by Equation (3) with d=1.0 gcmꢀ3 for 4a,
4b, and 32. [b] Calculated on the basis of Equation (1). [c] Calculations
at the B3LYP/6-31G* level.
12876
www.chemeurj.org
ꢁ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2010, 16, 12873 – 12882

Dipolar Organic Dyes
FULL PAPER
Table 3. Calculated vertical electronic transitions for all compounds in
planar and twisted conformations.
the torsional motion. Figure 3 illustrates the Arrhenius plots
of 4b in acetonitrile and DMSO. The calculated torsional
barriers for 4b in acetonitrile and DMSO are 1.1 and
Dye Conformation State labs [nm]
f
Excitation[a]
0.7853 H!L
0.057
Hꢀ1!L
H!L+2
%
4a
planar
S1
S2
398
300
100
92
6
twisted
planar
twisted
planar
S1
S2
S1
S2
S1
S2
S1
S2
439
352
410
330
465
349
450
350
0.0003 H!L
0.7088 Hꢀ1!L
0.8833 H!L
0.0339 H!L+1
0.1300 H!L
0.7041 Hꢀ1!L
0.0548 H!L
0.6589 Hꢀ1!L
97
100
99
96
98
100
98
97
4b
32
[a] H and L represent HOMO and LUMO, respectively.
~
&
Figure 3. Arrhenius plots of 4b in acetonitrile ( ) and DMSO ( ).
2.4 kcalmolꢀ1, respectively. The small activation barriers for
the torsional motion of 4b in polar solvents indicate a facile
conversion from an ICT state to a relaxed TICT state. It is
interesting to note that there is a small difference in the tor-
sional barriers in acetonitrile and DMSO. This difference
can be attributed to the more viscous DMSO relative to ace-
tonitrile, which leads to a higher activation energy for the
conformational rearrangement in DMSO. Assuming kr is
temperature-independent in the experimental temperature
range, the pre-exponential A0 values in acetonitrile and
DMSO are 1.7ꢃ109 and 2.7ꢃ1010 sꢀ1, respectively, which are
smaller than the values for typical gas-phase unimolecular
reactions (1012–1014 sꢀ1), thereby indicating a strong solva-
tion effect in these polar solvents during the torsional
motion to yield a more organized activated complex than
the pretwisted conformation and result in a negative entropy
of activation.
Figure 4. Frontier molecular orbitals for 4a, 4b, and 32 in planar and
twisted conformations.
the LUMO is localized on the acceptor moiety for the per-
pendicular conformation. Thus, a more prominent charge
separation exists in the perpendicular conformation than the
planar conformation.
Potential-energy curves as a function of the twisting angle
ꢀ
between the C N bond were calculated for the ground and
low-lying excited states of 4a and 4b with TD-DFT at the
B3LYP/6-31G* level. The results are depicted in Figure 5.
Unlike the case of 4a, in which the lowest-energy conforma-
tions in both S0 and S1 states are planar, the conformation of
4b reaches a low-energy plateau in the S1 state with twisted
angle between 70 and 908. Based on the potential-energy
curve of 4b, the S1 state at the twisted conformation is ap-
proximately 0.28 eV lower in energy than the planar confor-
mation. It should be noted here that the conformation of 4b
in the TICT state is not necessarily twisted at 908 along the
Computational approach: To further support the existence
of an accessible TICT state in 4b, detailed quantum calcula-
tions were carried out for 4a, 4b, and 32. The molecular
structures of ground and electronically excited states of all
three compounds in planar and twisted conformations were
optimized. The frontier molecular orbitals (MOs) and elec-
tronic transitions were calculated with DFT and time-depen-
dent (TD)-DFT at the B3LYP/6-31G* level; results are sum-
marized in Table 3 and Figure 4. The S1 state of all three
compounds corresponds to a HOMO!LUMO transition,
whereas the S2 state corresponds to either HOMOꢀ1!
LUMO or HOMO!LUMO+1. In Figure 5 it can clearly be
seen that the frontier MOs are more localized in twisted
conformations than in planar structures. The electron densi-
ty of the HOMO is localized on the donor moiety, whereas
ꢀ
C N bond, which is slightly different from the conventional
definition of the TICT state. The potential-energy curve of
the ground state has a minimum of 328 twisted geometry for
ꢀ
4b. The partial pretwisted C N bond of 4b in the ground
state ensures that there is less energy required to overcome
the rotation barrier to cross the potential surface from the
Chem. Eur. J. 2010, 16, 12873 – 12882
ꢁ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
12877

S.-S. Sun, C.-P. Hsu, M.-C. P. Yeh et al.
Figure 6. Normalized absorption spectra of dyes adsorbed on 1.5 mm TiO2
~
&
*
films ( =4a, =4b, and =32).
tion of the carboxylic acid.[7c] The spectral coverage of 4b is
apparently broader than the ones of 4a and 32, which im-
plies potentially better light harvesting for 4b. The amount
of dyes adsorbed on the TiO2 surface was estimated from
the difference in concentration of each solution before and
after TiO2 film immersion followed by rinsing with acetoni-
trile to ensure a minimum of the physically adsorbed dyes
left on the films. The amount of dye adsorbed onto the TiO2
film is extremely large, most likely due to the compact mo-
lecular size of dyes.[18] The high dye loading on the TiO2
film indicates that these structurally simple dyes with small
molecular sizes formed a well-organized monolayer on the
TiO2 surface, which is beneficial for light-harvesting efficien-
cy.[19]
Figure 5. Calculated potential-energy curves as a function of twisting
angle for different electronic states for 4a (top) and 4b (bottom) with
!
&
*
TD-DFT at the B3LYP/6-31G* level ( =S0, =S1, and =S2).
planar ICT state to the lowest-energy TICT state when com-
pared to the relaxation process in 4a.
The oxidation potentials EACHTNUGRTNEUNG
(S+/S) of the dyes were mea-
Properties of DSSCs assembled based on dyes 4a, 4b, and
32: The amount of dye loading and the electrochemical data
of DSSCs assembled based on dyes 4a, 4b, and 32 are col-
lected in Table 4. Figure 6 shows the absorption spectra of
the three dyes adsorbed on a transparent TiO2 film. Com-
sured by cyclic voltammetry (CV) and the results are col-
lected in Table 4. To achieve efficient electron injection
from the excited dye to the CB of the TiO2, the energy level
of the LUMO of the dye must be higher than the CB of the
TiO2 electrode. On the other hand, to achieve efficient re-
generation of the oxidized state by electron transfer from
ꢀ
I3 /Iꢀ redox couple in the electrolyte, the energy level of the
Table 4. Electrochemical data and the amount of dye loading.
ꢀ
HOMO of the dye must be lower than the I3 /Iꢀ redox po-
Dye E0–0
E
(S+/S)
E
(S+/S*)
Egap
Dye loading
tential. The oxidation potentials ranging from 0.69 to 1.04 V
versus NHE are more positive than the redox mediator Iꢀ/
I3ꢀ (ꢂ0.4 V vs. NHE),[20] which ensures the efficient thermo-
dynamic driving force (0.29–0.64 V) for dye regeneration.
[eV]
[V][a]
[V][a]
[V][b]
[10ꢀ6 molcmꢀ2
]
4a[c] 2.84
0.87
1.04
0.69
ꢀ1.97
ꢀ1.77
ꢀ2.12
1.47
1.27
1.62
9.59
2.79
1.90
4b[c] 2.81
32
2.81
The excited-state oxidation potential E
can be obtained from the difference between the ground-
state oxidation potential E
(S+/S) and zero–zero excitation
(S+/S*) of these dyes
ACHTUNGTRENNUNG
[a] Potentials are versus NHE. [b] Energy gap between the excited-state
oxidation potential and TiO2 conduction-band edge. [c] Data taken from
ref. [12].
AHCTUNGTRENNUNG
energy E0–0, which was estimated from the cross section of
the normalized absorption and emission spectra. As shown
in Table 4, the excited-state oxidation potentials of the three
dyes are more negative than the potential of the TiO2 con-
duction-band edge (ꢀ0.5 V vs. NHE), which provides suffi-
cient thermodynamic driving force (1.27–1.62 V) for elec-
tron injection.
pared to the corresponding spectra in acetonitrile solution,
the absorption spectral coverage of the dyes becomes broad-
er upon adsorption on TiO2, but the absorption maxima
shifted to higher energy when compared to that of in solu-
tion. Such a blueshift of the absorption spectra is attributed
to either the formation of H-type aggregate[17] or deprotona-
12878
www.chemeurj.org
ꢁ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2010, 16, 12873 – 12882

Dipolar Organic Dyes
FULL PAPER
Figure 7 illustrates the J–V curves and action spectrum of
IPCE plots for the DSSCs based on the dyes studied in this
work. The device performance of solar cells based on the
es, VOC, are in the order of 4b>32>4a. The reduced photo-
voltage observed in 4a can be attributed to the less-bulky
dimethylamino group in 4a that is incapable of effectively
ꢀ
blocking the infusion of the oxidized electrolyte I3 toward
the TiO2 surface and results in faster charge recombination
ꢀ
between injected electron and oxidized electrolyte I3 when
compared to the more bulky phenyindoline group in 32 and
diphenylamino group in 4b. The largest dark current gener-
ated from the DSSC based on 4a among the three con-
firmed the fastest charge recombination in this case. This
viewpoint is further supported by measuring the transient
photovoltage decay at the open circuit.[21] Figure 8 shows the
Figure 8. Traces of transient photovoltage decay of DSSCs based on dyes
~
*
*
with 0.5m LiI as the electrolyte in CH3CN ( =4a, =4b, and =32).
transient photovoltage decay plots with LiI as the electro-
lyte. The electron lifetime was determined by fitting a decay
of photovoltage transient with exp
G
and t is an average time constant before recombination. The
injected electron lifetimes obtained for 4a, 4b, and 32 are
2.93, 4.14, and 3.28 ms, respectively, which are in the order
of 4b>32>4a and in accordance with the observed order
of VOC shown in Table 5.
Figure 7.
&
*
J–V curves (top) and IPCE plots (bottom) of the dyes ( =4a, =4b,
~
and =32).
As shown in Tables 4 and 5, the amount of dye adsorbed
onto the TiO2 surface qualitatively correlated with the
values of JSC, with the lowest JSC generated from the DSSC
with the least amount of adsorbed dye. Beyond the better
light-harvesting ability due to the higher surface-dye load-
ing, it would be informative to correlate the photocurrent
density to the accessibility of the TICT state of these dyes
since the forbidden nature of radiative transition from a
TICT state to the ground state is expected to slow down the
charge recombination between the injected electron and
photo-oxidized dye. To realize this argument, the electro-
chemical ac impedance of the cells was measured under the
illumination conditions.[22] Figure 9 shows the electrochemi-
cal impedance Nyquist plots for DSSCs based on dyes 4a,
4b, and 32 measured under open-circuit conditions and
under illumination of 100 mWcmꢀ2. The semicircle located
in the midfrequency region is assigned to the charge transfer
dyes is summarized in Table 5. The performance of the devi-
ces based on dyes 4b reached up to 70% of the standard
N719 dye. The photocurrent densities, JSC, of the DSSCs are
in the order of 4b>4a>32, whereas the open circuit voltag-
Table 5. Photovoltaic performance parameters of the dyes.[a]
Dye
VOC [mV]
JSC [mAcmꢀ2
]
ff
h [%][b]
4a[c]
4b[c]
32
577
660
641
695
7.89
9.40
5.44
11.9
0.59
0.65
0.67
0.71
2.69ꢃ0.20
4.03ꢃ0.01
2.34ꢃ0.08
5.87ꢃ0.02
N719[c]
[a] Experiments were performed using TiO2 photoelectrodes with ap-
proximately 16 mm thickness and 0.25 cm2 working area on the FTO
(15 W per sq.) substrates with electrolyte composed of 0.05m I2, 0.5m LiI,
and 0.5m tert-butylpyridine in acetonitrile solution under AM 1.5 illumi-
nation. [b] Average of three measurements. [c] Data taken from ref. [12].
Chem. Eur. J. 2010, 16, 12873 – 12882
ꢁ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
12879

S.-S. Sun, C.-P. Hsu, M.-C. P. Yeh et al.
Experimental Section
Materials and general procedures: The chemical 3-ethoxycyclohex-2-
enone was synthesized according to the published method.[12] All other
chemical reagents were commercially available and were used without
further purification unless otherwise noted. Reactions were monitored by
TLC using aluminum plates precoated with a 0.25 mm layer of silica gel
that contained a fluorescent indicator.
Absorption spectra were obtained using a Perkin–Elmer Lambda 900
UV/Vis-NIR spectrophotometer or a Varian Cary 300 UV/Vis spectro-
photometer. Emission spectra were recorded in an air-equilibrated solu-
tion using a Fluorolog III photoluminescence spectrometer.[25] Cyclic vol-
tammetry experiments were performed using a CHI electrochemical ana-
lyzer. All measurements were carried out at room temperature with a
conventional three-electrode configuration that consisted of a platinum
working electrode, a platinum wire auxiliary electrode, and a nonaqueous
Ag/AgNO3 reference electrode. The potentials were quoted against the
ferrocene internal standard. The solvent contained 1.0 mm of the dye and
0.1m tetrabutylammonium hexafluorophosphate as supporting electrolyte
in all experiments.
Figure 9. Electrochemical impedance Nyquist plots of DSSCs based on
dyes measured under one sun illumination with open-circuit conditions
The photoelectrochemical characterizations on the solar cells were car-
ried out using a modified light source, a 450 W Xe lamp (Oriel, 6266)
equipped with a water-based IR filter and an AM 1.5 filter (Oriel,
81075). Light intensity, attenuated by neutral density filter (Optosigma,
078-0360) at the measuring (cell) position, was estimated to be around
100 mWcmꢀ2 according to the reading from a radiant power meter
(Oriel, 70260) connected to a thermopile probe (Oriel, 70263). Photocur-
rent/voltage curves of the DSSCs were recorded through the potentio-
stat/galvanostat (PGSTAT 30, Autolab, Eco-Chemie, Netherland).
~
&
*
(
=4a, =4b, and =32).
at the TiO2/dye/electrolyte interface.[23] The radius of the in-
termediate frequency semicircle of 4b is much smaller than
those of 4a and 32, thus indicating the much improved elec-
tron generation and transport in 4b among the three dyes.[24]
This result also corresponds well to that of the short-circuit
current density shown in Table 5 in which the JSC of 4b is
the highest among the three dyes.
Electrochemical impedance spectra were recorded for DSSCs under the
same conditions at VOC potential at room temperature. The frequencies
explored ranged from 10 mHz to 65 kHz. The ac amplitude was set at
10 mV. The photovoltage transients of assembled devices were recorded
using a digital oscilloscope (Tektronix, TDS 3012b). Pulsed laser excita-
tion was applied by a Q-Switched Nd:YAG laser (Quanrel, brilliant B)
with 10 Hz repetition rate at 532 nm and a 7 ns pulse width at half-
height. The beam size was slightly larger than 0.25 cm2 to cover the area
of the device with incident energy of 3 mJcmꢀ2. The recombination life-
time of injected electrons with oxidized dyes was measured by transient
photovoltages at open circuit with the presence of LiI electrolyte (0.5m).
The average electron lifetime can be estimated approximately by fitting a
Conclusion
In summary, a series of organic dyes that feature a 1,3-cyclo-
hexadiene conjugated moiety integrated into the p-conjugat-
ed framework has been synthesized and studied. DSSCs
based on these dyes have shown appreciable photo-to-elec-
trical energy conversion efficiency with the highest one up
to 4.03%. DSSCs based on the dye 4b, which contain a di-
phenylamine moiety as the donor group, have better effi-
ciency than DSSCs based on dyes 4a and 32. Dye 32 with a
decay of the open circuit voltage transient with exp
time and t is an average time constant before recombination.
G
Compound 29: nBuLi (2.5m in n-hexane, 8.8 mL, 22 mmol) was added to
a 250 mL flask charged with 5-bromo-1-phenylindoline (5.5 g, 20 mmol)
and THF (100 mL) at ꢀ788C and stirred for 2 h. 3-Ethoxycyclohex-2-
enone (2.5 g, 20 mmol) in THF (45 mL) was transferred to this mixture
and the solution was stirred at ꢀ788C for 2 h. The reaction was quenched
with aqueous NH4Cl, extracted with ethyl acetate, dried with MgSO4,
and evaporated the solvent to afford the crude product. Column chroma-
tography eluted with ethyl acetate/n-hexane (1:3) afforded compound 29
in 69% yield as a pale yellow solid. 1H NMR (400 MHz, CDCl3, 258C):
ꢀ
C N bond, which fused to prevent the structural twisting
between the phenylamino and benzene ring, served as a
model compound for 4a and 4b to evaluate the related pho-
tophysical properties. Comprehensive photophysical studies
and quantum calculations have suggested the accessibility of
the TICT state for 4b but not 4a in polar solvents. The vir-
tually forbidden nature of the radiative relaxation from the
TICT state to the ground state renders DSSCs based on 4b
potent to resist the charge recombination between the pho-
toinjected electron and the oxidized 4b. The good correla-
tion between the proposed TICT model and photocurrent
generated from the DSSCs in our current work implies an
useful guideline for designing highly efficient organic dyes
for future progress of DSSCs. Our work on efficient dyes for
DSSCs along this line is currently in progress.
d=7.40–7.23 (m, 4H), 7.23 (d, 3J
8.5 Hz, 1H), 7.025 (t, 3J
(H,H)=7.7 Hz, 1H), 6.39 (s, 1H), 4.03 (t,
3J(H,H)=8.5 Hz, 2H), 2.74 (t, 3J(H,H)=6.1 Hz, 2H), 2.45 (t, 3J
(H,H)=
(H,H)=7.7 Hz, 2H), 7.08 (d, 3J
ACHUTGTNRNEUNG ACHTUNGTRENNUNG(H,H)=
AHCTUNGTRENNUNG
A
R
ACHTUNGTRENNUNG
7.3 Hz, 2H), 2.15–2.08 ppm (m, 2H); 13C NMR (100 MHz, CDCl3, 258C):
d=199.9, 159.4, 149.2, 143.1, 131.8, 129.3, 128.5, 126.3, 122.9, 122.10,
122.08, 118.3, 107.6, 52.4, 37.2, 27.8, 22.8 ppm; HRMS (EI): m/z: calcd
for C23H23NO2: 289.1467 [M+]; found: 289.1467.
Compound 31: Compound 29 (1.04 g, 3.58 mmol) was dissolved in THF
(45 mL) and the solution was cooled to ꢀ788C. The resulting mixture
was transferred to a 100 mL flask that contained lithium diisopropyl-
amide (LDA) prepared from diisopropylamine (4.65 mmol) and n-butyl-
lithium (1.6m solution in hexane, 3.0 mL, 4.8 mmol) at ꢀ788C for 45 min.
Subsequently, ethyl chloroformate (0.6 mL, 5.37 mmol) was added to the
reaction flask and stirred at ꢀ788C for another 45 min. The reaction was
quenched with aqueous NH4Cl, extracted with ethyl acetate, dried over
12880
www.chemeurj.org
ꢁ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2010, 16, 12873 – 12882

Dipolar Organic Dyes
FULL PAPER
MgSO4, and evaporated the solvent to afford the crude product. Flash
column chromatography eluted with ethyl acetate/n-hexane (1:3) afford-
Angew. Chem. 2006, 118, 5954–5957; Angew. Chem. Int. Ed. 2006,
45, 5822–5825; f) F. Gao, Y. Wang, D. Shi, J. Zhang, M. Wang, X.
Jing, R. Humphry-Baker, P. Wang, S. M. Zakeeruddin, M. Grꢀtzel,
J. Am. Chem. Soc. 2008, 130, 10720–10728; g) B.-S. Chen, K. Chen,
Y.-H. Hong, W.-H. Liu, T.-H. Li, C.-H. Lai, P.-T. Chou, Y. Chi, G.-
H. Lee, Chem. Commun. 2009, 5844–5846.
ed ethyl 2-oxo-4-(1-phenylindolin-5-yl)cyclohex-3-enecarboxylate as
a
pale yellow solid. HRMS (EI): m/z: calcd: 361.1684 [M+]; found:
361.1678.
NaBH4 (0.29 g, 7.8 mmol) was added to a 100 mL flask charged with the
ethyl 2-oxo-4-(1-phenylindolin-5-yl)cyclohex-3-enecarboxylate crude
product (0.56 g, 1.6 mmol) and MeOH (50 mL) in an ice bath and the re-
action progress was monitored by TLC. The reaction was quenched with
aqueous NH4Cl. After evaporating the volatile portion, the residue was
extracted with ethyl acetate, dried over MgSO4, and evaporated the sol-
vent to afford the crude product. The crude product was redissolved in
CH2Cl2 (50 mL) followed by addition of methanesulfonyl chloride
(0.2 mL, 2.1 mmol) and NEt3 (0.25 mL, 1.9 mmol) in the ice bath. The
mixture was stirred at room temperature for 2.5 h. The reaction was
quenched with aqueous NH4Cl, extracted with CH2Cl2, dried over
MgSO4, and evaporated the solvent to afford the brownish residue.
[3] a) S. Ito, S. M. Zakeeruddin, R. Humphry-Baker, P. Liska, R. Char-
vet, P. Comte, M. K. Nazeeruddin, P. Pꢆchy, M. Takata, H. Miura, S.
Uchida, M. Grꢀtzel, Adv. Mater. 2006, 18, 1202–1205; b) T. Horiu-
chi, H. Miura, K. Sumioka, S. Uchida, J. Am. Chem. Soc. 2004, 126,
12218–12219; c) S. Hwang, J. H. Lee, C. Park, H. Lee, C. Kim, C.
Park, M.-H. Lee, W. Lee, J. Park, K. Kim, N.-G. Park, C. Kim,
Chem. Commun. 2007, 4887–4889; d) S. Ito, H. Miura, S. Uchida,
M. Takata, K. Sumioka, P. Liska, P. Comte, P. Pꢆchy, M. Grꢀtzel,
Chem. Commun. 2008, 5194–5196; e) G. Zhang, H. Bala, Y. Cheng,
D. Shi, X. Lv, Q. Yu, P. Wang, Chem. Commun. 2009, 2198–2200.
[4] a) R. Neil, Angew. Chem. 2006, 118, 2398–2405; Angew. Chem. Int.
Ed. 2006, 45, 2338–2345; b) A. Mishra, M. K. R. Fischer, P. Bꢀuerle,
Angew. Chem. 2009, 121, 2510–2536; Angew. Chem. Int. Ed. 2009,
48, 2474–2499.
CH2Cl2 (50 mL) and 1,8-diazabicycloACTHNUGRTNEUNG[5.4.0]undec-7-ene (DBU; 0.3 mL,
2.1 mmol) were added to this residue in an ice bath and the mixture was
stirred at room temperature for 8 h before being quenched with aqueous
NH4Cl. The residue was extracted with CH2Cl2, dried over MgSO4, and
evaporated the solvent to afford the crude product. Column chromatog-
raphy eluted with ethyl acetate/n-hexane (1:30) afforded compound 31 in
[5] a) A. Hagfeldt, M. Grꢀtzel, Chem. Rev. 1995, 95, 49–68; b) M. Grꢀt-
zel, Nature 2001, 414, 338–344.
[6] a) A. Hagfeldt, M. Grꢀtzel, Acc. Chem. Res. 2000, 33, 269–277; b) F.
Matar, T. H. Ghaddar, K. Walley, T. DosSantos, J. R. Durrant, B.
OꢄRegan, J. Mater. Chem. 2008, 18, 4246–4253; c) K. Chen, Y.-H.
Hong, Y. Chi, W.-H. Liu, B.-S. Chen, P.-T. Chou, J. Mater. Chem.
2009, 19, 5329–5335; d) J.-F. Yin, D. Bhattacharya, Y.-C. Hsu, C.-C.
Tsai, K.-L. Lu, H.-C. Lin, J.-G. Chen, K.-C. Ho, J. Mater. Chem.
2009, 19, 7036–7042.
1
39% yield as a dark yellow solid. H NMR (500 MHz, CDCl3, 258C): d=
7.35–7.32 (m, 3H), 7.25–7.20 (m, 4H), 7.16 (d, 3J
ACHTUNGTRENNUNG
7.08 (d, 3J
3J(H,H)=6.0 Hz, 1H), 4.22 (q, 3J
8.4 Hz, 2H), 3.13 (t, 3J
(t, 3J
ACHTUNGTRENNUNG ACHTUNGTRENNUNG
(H,H)=8.4 Hz, 1H), 6.98 (t, 3J
G
A
ACHTUNGTRENNUNG
ACHTUNGTRENNUNG
ACHTUNGTRENNUNG
147.3, 143.6, 143.5, 134.6, 131.7, 130.4, 129.2, 125.0, 122.2, 121.5, 117.9,
117.1, 107.7, 60.2, 52.3, 27.9, 26.1, 22.0, 14.4 ppm; HRMS (EI): m/z: calcd
for C23H23NO2: 345.1729 [M+]; found: 345.1738.
[7] a) A. Abbotto, N. Manfredi, C. Marinzi, F. De Angelis, E. Mosconi,
J.-H. Yum, Z. Xianxi, M. K. Nazeeruddin, M. Grꢀtzel, Energy Envi-
ron. Sci. 2009, 2, 1094–1101; b) K. R. J. Thomas, Y.-C. Hsu, J. T. Lin,
K.-M. Lee, K.-C. Ho, C.-H. Lai, Y.-M. Cheng, P.-T. Chou, Chem.
Mater. 2008, 20, 1830–1840; c) Y.-S. Yen, Y.-C. Hsu, J. T. Lin, C.-W.
Chang, C.-P. Hsu, D.-J. Yin, J. Phys. Chem. C 2008, 112, 12557–
12567; d) J. T. Lin, P.-C. Chen, Y.-S. Yen, Y.-C. Hsu, H.-H. Chou,
M.-C. P. Yeh, Org. Lett. 2009, 11, 97–100; e) S. Kim, J. K. Lee, S. O.
Kang, J. Ko, J.-H. Yum, S. Fantacci, F. De Angelis, D. D. Censo,
M. K. Nazeeruddin, M. Grꢀtzel, J. Am. Chem. Soc. 2006, 128,
16701–16707; f) A. Abbotto, C. Barolo, L. Bellotto, F. de Angelis,
M. Grꢀtzel, N. Manfredi, C. Marinzi, S. Fantacci, J.-H. Yum, M. K.
Nazeeruddin, Chem. Commun. 2008, 5318–5320.
[8] a) Z. Ning, H. Tian, Chem. Commun. 2009, 5483–5495 and referen-
ces therein; b) H. Tian, X. Yang, J. Pan, R. Chen, M. Liu, Q. Zhang,
A. Hagfeldt, L. Sun, Adv. Funct. Mater. 2008, 18, 3461–3468; c) G.
Li, Y.-F. Zhou, X.-B. Cao, P. Bao, K.-J. Jiang, Y. Lin, L.-M. Yang,
Chem. Commun. 2009, 2201–2203.
[9] a) N. Hirata, J.-J. Lagref, E. J. Palomares, J. R. Durrant, M. K. Na-
zeeruddin, M. Grꢀtzel, D. D. Censo, Chem. Eur. J. 2004, 10, 595–
602; b) S. A. Haque, S. Handa, K. Peter, E. Palomares, M. Thelak-
kat, J. R. Durrant, Angew. Chem. 2005, 117, 5886–5890; Angew.
Chem. Int. Ed. 2005, 44, 5740–5744.
Compound 32: A 100 mL flask was charged with compound 31 (100 mg,
0.29 mmol), KOH (50 mg, 0.89 mmol), and MeOH/THF (50 mL, 5:1 in v/
v) and the mixture was heated at reflux for 12 h. The volatile portion was
removed under reduced pressure. The residue was neutralized to pH 7
with 1m HCl. The aqueous solution was extracted with CH2Cl2 and dried
over MgSO4. Subsequent column chromatography eluted with ethyl ace-
tate/n-hexane (1:1) afforded dye 32 in 53% yield as a pale orange solid.
1H NMR (500 MHz, CDCl3, 258C): d=7.48 (s, 1H), 7.39–7.340 (m, 3H),
7.8 (d, 3J
3J(H,H)=5.0 Hz, 1H), 6.99 (t, 3J
5.0 Hz, 1H), 4.03 (t, 3J(H,H)=10.0 Hz, 2H), 3.17 (t, 3J
2H), 2.70 (t, 3J(H,H)=10.0 Hz, 2H), 2.55 ppm (t, 3J
ACHTUNGTRENNUNG
G
ACHTUNGTRENNUNG
E
E
ACHTUNGTRENNUNG
G
ACHTUNGTRENNUNG
U
2H); 13C NMR (125 MHz, CDCl3, 258C): d=168.3, 148.2, 144.6, 144.3,
135.4, 133.0, 130.4, 125.9, 125.2, 122.1, 118.6, 117.6, 111.4, 108.4, 52.9,
31.6, 26.5, 22.8 ppm; HRMS (EI): m/z: calcd for C21H19NO2: 317.1416
[M+]; found: 317.1420.
Acknowledgements
[10] a) Z. R. Grabowski, K. Rotkiewicz, W. Rettig, Chem. Rev. 2003, 103,
3899–4031 and references therein; b) R. Chen, G. Zhao, X. Yang,
X. Jiang, J. Liu, H. Tian, Y. Gao, X. Liu, K. Han, M. Sun, L. Sun, J.
Mol. Struct. 2008, 876, 102–109.
We thank the National Science Council in Taiwan, Academia Sinica, and
the National Taiwan Normal University for support of this research.
[11] a) E. Lippert, W. Rettig, V. Bonacic-Koutecky, F. Heisel, J. A.
Miehe, Adv. Chem. Phys. 1987, 68, 1–174; b) J.-S. Yang, K.-L. Liau,
C.-Y. Li, M.-Y. Chen, J. Am. Chem. Soc. 2007, 129, 13183–13192;
c) S. S. Palayangoda, X. C. Cai, R. M. Adhikari, D. C. Neckers, Org.
Lett. 2008, 10, 281–284; d) M. S. Yuan, Z. Q. Liu, Q. Fang, J. Org.
Chem. 2007, 72, 7915–7922; e) S. Cogan, S. Zilberg, Y. Haas, J. Am.
Chem. Soc. 2006, 128, 3335–3345; f) E. Lippert, W. Lꢅder, F. Moll,
W. Nꢀgele, H. Boos, H. Prigge, I. Seibold-Blankenstein, Angew.
Chem. 1961, 73, 695–706; g) L. Serrano-Andrꢆs, M. Merchꢇn, B. O.
Roos, R. Lindh, J. Am. Chem. Soc. 1995, 117, 3189–3204; h) W.
Rettig, Angew. Chem. 1986, 98, 969–986; Angew. Chem. Int. Ed.
Engl. 1986, 25, 971–988; i) C. Jamorski Jçdicke, H. P. Lçthi, J. Am.
Chem. Soc. 2003, 125, 252–264; j) D. Rappoport, F. Furche, J. Am.
[1] B. OꢄRegan, M. Grꢀtzel, Nature 1991, 353, 737–740.
[2] a) M. K. Nazeeruddin, A. Kay, I. Rodicio, R. Humphry-Baker, E.
Mꢅller, P. Liska, N. Vlachopoulos, M. Grꢀtzel, J. Am. Chem. Soc.
1993, 115, 6382–6390; b) M. K. Nazeeruddin, P. Pꢆchy, T. Renouard,
S. M. Zakeeruddin, R. Humphry-Baker, P. Comte, P. Liska, L.
Cevey, E. Costa, V. Shklover, L. Spiccia, G. B. Deacon, C. A.
Bignozzi, M. Grꢀtzel, J. Am. Chem. Soc. 2001, 123, 1613–1624;
c) M. Grꢀtzel, J. Photochem. Photobiol. A 2004, 164, 3–14; d) C.-Y.
Chen, J.-G. Chen, S.-J. Wu, J.-Y. Li, C.-G. Wu, K.-C. Ho, Angew.
Chem. 2008, 120, 7452–7455; Angew. Chem. Int. Ed. 2008, 47, 7342–
7345; e) C.-Y. Chen, S.-J. Wu, C.-G. Wu, J.-G. Chen, K.-C. Ho,
Chem. Eur. J. 2010, 16, 12873 – 12882
ꢁ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
12881

S.-S. Sun, C.-P. Hsu, M.-C. P. Yeh et al.
Chem. Soc. 2004, 126, 1277–1284; k) I. Gꢈmez, M. Reguero, M.
Boggio-Pasqua, M. A. Robb, J. Am. Chem. Soc. 2005, 127, 7119–
7129; l) U. Resch-Genger, Y. Q. Li, J. L. Bricks, V. Kharlanov, W.
Rettig, J. Phys. Chem. A 2006, 110, 10956–10971; m) R. Hu, E.
Lager, A. Aguilar-Aguilar, J. Liu, J. W. Y. Lam, H. H. Y. Sung, I. D.
Williams, Y. Zhong, K. S. Wong, E. PeÇa-Cabrera, B. Z. Tang, J.
Phys. Chem. C 2009, 113, 15845–15853; n) G.-J. Zhao, R.-K. Chen,
M.-T. Sun, J.-Y. Liu, G.-Y. Li, Y.-L. Gao, K.-L. Han, X.-C. Yang, L.
Sun, Chem. Eur. J. 2008, 14, 6935–6947; o) J.-S. Yang, C.-K. Lin,
A. M. Lahoti, C.-K. Tseng, Y.-H. Liu, G.-H. Lee, S.-M. Peng, J.
Phys. Chem. A 2009, 113, 4868–4877.
Cui, M. Takahashi, H. Sekiguchi, A. Mori, T. Kubo, A. Furube, K.
Hara, Chem. Mater. 2008, 20, 3993–4003; d) Y. Ooyama, A. Ishii, Y.
Kagawa, I. Imae, Y. Harima, New J. Chem. 2007, 31, 2076–2082.
[19] J. R. Mann, M. K. Gannon, T. C. Fitzgibbons, M. R. Detty, D. F.
Watson, J. Phys. Chem. C 2008, 112, 13057–13061.
[20] N. Koumura, Z.-S. Wang, S. Mori, M. Miyashita, E. Suzuki, K. Hara,
J. Am. Chem. Soc. 2006, 128, 14256–14257.
[21] N. Kopidakis, K. D. Benkstein, J. van de Lagemaat, A. J. Frank, J.
Phys. Chem. B 2003, 107, 11307–11315.
[22] a) J. van de Lagemaat, N.-G. Park, A. J. Frank, J. Phys. Chem. B
2000, 104, 2044–2052; b) Q. Wang, J.-E. Moser, M. Grꢀtzel, J. Phys.
Chem. B 2005, 109, 14945–14953; c) D. Kuang, S. Ito, B. Wenger, C.
Klein, J.-E. Moser, R. Humphry-Baker, S. M. Zakeeruddin, M. Grꢀt-
zel, J. Am. Chem. Soc. 2006, 128, 4146–4154.
[23] a) C. Longo, A. F. Nogueira, M.-A. de Paoli, H. Cachet, J. Phys.
Chem. B 2002, 106, 5925–5930; b) R. Kern, R. Sastrawan, J. Ferber,
R. Stangl, J. Luther, Electrochim. Acta 2002, 47, 4213–4225.
[24] a) H. Choi, S. Kim, S. O. Kang, J. Ko, M.-S. Kang, J. N. Clifford, A.
Forneli, E. Palomares, M. K. Nazeeruddin, M. Grꢀtzel, Angew.
Chem. 2008, 120, 8383–8387; Angew. Chem. Int. Ed. 2008, 47, 8259–
8263; b) J.-J. Kim, H. Choi, J.-W. Lee, M.-S. Kang, K. Song, S. O.
Kang, J. Ko, J. Mater. Chem. 2008, 18, 5223–5229.
[12] K.-F. Chen, Y.-C. Hsu, Q.-Y. Wu, M.-C. P. Yeh, S.-S. Sun, Org. Lett.
2009, 11, 377–380.
[13] J.-S. Yang, K.-L. Liau, C.-M. Wang, C.-Y. Hwang, J. Am. Chem. Soc.
2004, 126, 12325–12335.
[14] T.-P. Lin, C.-Y. Chen, Y-S. Wen, S.-S. Sun, Inorg. Chem. 2007, 46,
9201–9212.
[15] W. Baumann, H. Bischof, J.-C. Frhling, C. Brittinger, W. Rettig, K.
Rotkiewicz, J. Photochem. Photobiol. A 1992, 64, 49–72.
[16] a) J. Saltiel, A. Marinari, D. W.-L. Chang, J. C. Mitchener, E. D.
Megarity, J. Am. Chem. Soc. 1979, 101, 2982–2996; b) J.-S. Yang, S.-
Y. Chiou, K.-L. Liau, J. Am. Chem. Soc. 2002, 124, 2518–2527.
[17] Z.-S Wang, K. Hara, Y. Dan-oh, C. Kasada, A. Shinpo, S. Suga, H.
Arakawa, H. Sugihara, J. Phys. Chem. B 2005, 109, 3907–3914.
[18] a) W. Xu, B. Peng, J. Chen, M. Liang, F. Cai, J. Phys. Chem. C 2008,
112, 874–880; b) Z. Ning, Q. Zhang, W. Wu, H. Pei, B. Liu, H. Tian,
J. Org. Chem. 2008, 73, 3791–3797; c) Z.-S. Wang, N. Koumura, Y.
[25] a) C.-Y. Wu, M.-S. Chen, C.-A. Lin, S.-C. Lin, S.-S. Sun, Chem. Eur.
J. 2006, 12, 2263–2269; b) C.-L. Chen, T.-P. Lin, Y.-S. Chen, S.-S.
Sun, Eur. J. Org. Chem. 2007, 3999–4010.
Received: May 13, 2010
Published online: September 30, 2010
12882
www.chemeurj.org
ꢁ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2010, 16, 12873 – 12882