Dye-sensitized solar cells based on novel donor-acceptor π-conjugated benzofuro[2,3-c]oxazolo[4,5-a]carbazole-type fluorescent dyes exhibiting solid-state fluorescence

Article abstract of DOI:10.1039/b706896d

Novel donor-acceptor π-conjugated benzofuro[2,3-c]oxazolo[4,5-a] carbazole-type fluorescent dyes (2a-2d) with substituents (R = H, butyl, benzyl and 5-nonyl) on the nitrogen of carbazole ring have been designed and synthesized as sensitizers in dye-sensit

Full text of DOI:10.1039/b706896d

View Article Online / Journal Homepage / Table of Contents for this issue
PAPER
www.rsc.org/njc | New Journal of Chemistry
Dye-sensitized solar cells based on novel donor–acceptor p-conjugated
benzofuro[2,3-c]oxazolo[4,5-a]carbazole-type fluorescent dyes exhibiting
solid-state fluorescence
Yousuke Ooyama, Akihiro Ishii, Yusuke Kagawa, Ichiro Imae and Yutaka Harima*
Received (in Durham, UK) 8th May 2007, Accepted 19th July 2007
First published as an Advance Article on the web 8th August 2007
DOI: 10.1039/b706896d
Novel donor–acceptor p-conjugated benzofuro[2,3-c]oxazolo[4,5-a]carbazole-type fluorescent dyes
(2a–2d) with substituents (R = H, butyl, benzyl and 5-nonyl) on the nitrogen of carbazole ring
have been designed and synthesized as sensitizers in dye-sensitized solar cells (DSSCs).
Absorption and fluorescence properties of 2a–2d are similar in solution. In the solid state,
however, the dyes 2c (R = benzyl) and 2d (R = 5-nonyl) with bulky substituents exhibit strong
solid-state fluorescence properties, which are attributed to the reduction of the p–p interaction
between the dyes. Photovoltaic parameters of DSSCs based on 2a–2d are measured with the
amount of dyes adsorbed on TiO2 film as a parameter. It is found that the 2d exhibits a 100%
efficiency for absorbed photon-to-current conversion, demonstrating that bulky substituents in 2d
can efficiently prevent intermolecular energy transfer between the dyes in molecular aggregation
states.
fluorescence quenching in molecular aggregation state. It is
Introduction
expected that the use of these dyes as sensitizers should enable
Dye-sensitized solar cells (DSSCs) fabricated using organic
elucidation of the influence of molecular aggregation on the
materials have attracted increasing interest because of their
photovoltaic performances of DSSCs.
high abilities to convert solar light to electricity at low cost.1–6
A large number of organic dyes have been developed and the
Results and discussion
relationship between the chemical structure and photovoltaic
performance of DSSCs based on the dyes is examined. In
particular, donor–acceptor p-conjugated dyes possessing
broad and intense spectral features are useful as sensitizers.2–6
Unfortunately, these dyes are liable to form p-stacked aggre-
gates between the dyes on metal oxides such as TiO2, which
leads to reduction in electron-injection yield from the dyes to
the conduction band of TiO2 owing to intermolecular energy
transfer.5,6 In order to prevent aggregation of dye molecules
on TiO2 surface, deoxycholic acid (DCA) has been used in a
number of research works.5,7 The key point in designing of
new effective donor–acceptor-type sensitizers for DSSCs is to
reduce possible intermolecular energy transfer between dyes
adsorbed on TiO2 surface. The influences of molecular aggre-
gation on the photovoltaic performances of DSSCs are not yet
fully understood because of difficulty in making a systematic
study of the regulation of molecular arrangement on TiO2
surface.
Synthesis of benzofuro[2,3-c]oxazolo[4,5-a]carbazole-type
fluorescent dyes (1a–1d) and (2a–2d)
The synthetic pathway of the dyes is shown in Scheme 1. We
used 7-(4-cyanophenyl)-3-dibutylaminobenzofuro[2,3-c]oxa-
zolo[4,5-a]carbazole 1a8 as a starting material. The reaction
of 1a with alkyl halide using sodium hydride gave 1b–1d in
74–82% yields. The compounds 2a–2d were synthesized in
68–90% yields by hydrolysis of 1a–1d.
Spectroscopic properties of 2a–2d in solution and in the solid
state
The visible absorption and fluorescence spectral data of 2a–2d
in 1,4-dioxane are summarized in Table 1 and the spectra of
2a–2d are depicted in Fig. 1. The absorption and fluorescence
spectra of the fluorescent dyes 2a–2d closely resemble each
other, showing that the effect of N-alkylation of the carbazole
ring on the photophysical properties of 2 is negligible. The
absorption maxima of the dyes appear at around 415 and
350 nm, and the corresponding fluorescence bands are ob-
served at around 515 nm. The fluorescence quantum yields (F)
of 2a–2d are close to 100%. In the solid state, however, the
fluorescent dye 2d exhibits a stronger fluorescence band than
the other compounds (Table 1). The fluorescence quantum
yields increase in the order of 2d (F = 0.11) 4 2c (F = 0.10)
4 2b (F = 0.04) c 2a (F E 0). It was difficult to determine
exactly a quantum yield of 2a because of a feeble fluorescence
intensity. These results demonstrate that the N-alkylation of
In the present study, we have designed and synthesized
novel donor–acceptor p-conjugated benzofuro[2,3-c]oxazolo-
[4,5-a]carbazole-type fluorescent dyes (2). The solid-state
fluorescence properties of these dyes are changed by introdu-
cing bulky substituents in the dye skeleton. The bulky groups
reduce the p–p interaction between the dyes and prevent the
Department of Applied Chemistry, Graduate School of Engineering,
Hiroshima University, Higashi-hiroshima, 739-8527, Japan. E-mail:
harima@mls.ias.hiroshima-u.ac.jp; Fax: (+81) 82-424-5494
ꢀc
This journal is the Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2007
2076 | New J. Chem., 2007, 31, 2076–2082

View Article Online
Scheme 1 Synthesis of fluorescent dyes 1a–1d and 2a–2d: Reagents and conditions: (a) NaH, alkyl halide, acetonitrile, RT, 5–24 h; (b) NaOH aq,
ethanol, reflux, 16–24 h.
Table 1 Optical properties of 2a–2d in 1,4-dioxane and in the crystalline state
In 1,4-dioxane
In the crystalline state
lambasx/nm (emax /dm3 molꢁ1 cmꢁ1
)
lflmax/nm
F
lemxax/nm
lflmax/nm
F
a
2a
2b
2c
2d
a
416 (25 700), 347 (25 400)
419 (26 300), 351 (29 600)
408 (23 400), 348 (23 400)
411 (28 700), 349 (30 900)
525
519
508
507
0.97
0.97
0.97
0.97
545
529
508
537
582
571
562
580
0.04
0.10
0.11
Too weak to be measured.
the carbazole ring in 2a effectively reduces the intermolecular
p–p interaction such as causing the fluorescence quenching in
the solid state.9–11
clear than the first one, showing a poor stability of the second
and third oxidized states of the dyes.
On the bases of the spectral analyses and CVs, we estimated
the HOMO and LUMO energy levels of the four dyes. The
HOMO energy levels for 2a–2d are evaluated to be 0.87, 0.88,
0.87 and 0.88 V with respect to NHE, respectively. The
LUMO energy levels of the dyes are estimated from the first
oxidation potential, and an intersection of absorption and
fluorescence spectra (480 nm (2.58 eV), 472 (2.63), 470 (2.64)
and 475 (2.61) for 2a–2d, respectively) corresponding to the
energy gap between HOMO and LUMO. They are ꢁ1.71,
ꢁ1.75, ꢁ1.77 and ꢁ1.73 V for 2a–2d, respectively. Evidently,
the LUMO energy levels for these dyes are higher than the
energy level of the TiO2 conduction band (ꢁ0.5 V), showing
that these dyes can inject efficiently electrons to the conduction
band of the TiO2 electrode. The results of absorption and
fluorescence spectra and CV for these dyes demonstrate that
all these dyes have similar HOMO and LUMO energy levels.
Electrochemical properties of 2a–2d and their HOMO and
LUMO energy levels
The electrochemical properties of 2a–2d were determined by
cyclic voltammetry (CV) in acetonitrile containing 0.1 M
Et4NClO4. As an example, the CV of 2d is shown in Fig. 2.
The CV curves of these compounds showed three redox waves
at similar potentials irrespective of the compounds (Table 2).
Three oxidation peaks of these compounds are determined to
be 0.31–0.32, 0.80–0.85 and 0.97–0.99 V vs. Ag/Ag+. The
corresponding reduction peaks appear at 0.25–0.26 V for the
first redox step and the peak separations are ca. 60 mV. The
first oxidation and reduction peaks are clearly observed,
suggesting that the first oxidized state of the dyes is stable.
On the other hand, the second and third redox waves are less
Semi-empirical MO calculations (AM1, INDO/S)
The photophysical and electrochemical properties of 2a–2d
were analyzed by using semi-empirical molecular orbital (MO)
calculations. The molecular structures were optimized by
using the MOPAC/AM1 method,12 and then the INDO/S
method13 was used for spectroscopic calculations. The calcu-
lated absorption wavelengths and the transition characters of
the first absorption bands are collected in Table 3. The
observed and calculated absorption spectra of the compounds
compare well with each other with respect to both the absorp-
tion wavelength and the absorption intensity, although the
calculated absorption wavelengths are slightly blue-shifted.
The deviation of the INDO/S calculations, giving transition
Fig. 1 (a) Absorption and (b) fluorescence spectra of 2a–2d in
1,4-dioxane.
ꢀc
This journal is the Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2007
New J. Chem., 2007, 31, 2076–2082 | 2077

View Article Online
the dyes: 13.9 ꢂ 1016, 7.9 ꢂ 1016, 2.5 ꢂ 1016 and 2.3 ꢂ 1016
molecules cmꢁ2 for 2a–2d, respectively. Fig. 4 includes also the
absorption spectra of the dyes 2a–2d adsorbed on the TiO2
electrodes. Here, IPCE is defined by the following equation:
1240ðeV nmÞJscðlÞðmA cmꢁ2
Þ
IPCEð%Þ ¼
ꢂ 100
ð1Þ
lðnmÞIðlÞðmW cmꢁ2
Þ
where Jsc(l) is the short-circuit photocurrent density generated
by monochromatic light of a wavelength l and I(l) is an
intensity of the monochromatic light. The photocurrent–
voltage curves of Fig. 5 are obtained under AM 1.5 irradiation
(60 mW cmꢁ2). Photovoltaic performances of the DSSCs
deduced from Fig. 5 are summarized in Table 1. The solar
energy-to-electricity conversion yield (Z(%)) is expressed by
the following equation:
Fig. 2 Cyclic voltammogram of compound 2d in acetonitrile contain-
ing 0.1 M Et4NClO4 at a scan rate of 20 mV sꢁ1. The arrow denotes
the direction of the potential scan.
JscðmA cmꢁ2ÞVocðVÞff
Zð%Þ ¼
ꢂ 100
ð2Þ
I0ðmW cmꢁ2
Þ
where I0 is the intensity of an incident white light (60 mW
cmꢁ2), Voc is the open-circuit photovoltage, and ff represents
the fill factor.
energies greater than the experimental values, has been gen-
erally observed.14 The calculations show that the excitation
bands of the dyes at ca. 400 nm are mainly assigned to a
transition from HOMO to LUMO, where HOMOs are mostly
localized on the 3-dibutylaminobenzofuro[2,3-c]oxazolo-
[4,5-a]carbazole moiety for all dyes, and LUMOs are mostly
localized on the carboxylphenyl moiety. The changes in the
calculated electron density accompanying the first electron
excitation are shown in Fig. 3, which reveals a strong migra-
tion of intramolecular charge transfer from the 3-dibutyl-
aminobenzofuro[2,3-c]oxazolo[4,5-a]carbazole moiety to the
carboxylphenyl moiety for all dyes. Furthermore, the calcula-
tions show that the second excitation bands of the dyes are
mainly assigned to the transition from HOMO to LUMO+1,
where a moderate migration of charge from the 3-dibutylami-
nobenzofuro[2,3-c]oxazolocarbazole moiety to the carboxyl-
phenyl moiety is also observed. These calculation results
indicate that the effects of N-alkylation of the carbazole ring
on the photophysical and electrochemical properties of these
dyes are negligible, consistent with the experiment results of
the absorption and fluorescence spectra and the CVs.
Fig. 4 shows that the absorption peak wavelengths of
adsorbed dyes on TiO2 films are similar to those in 1,4-dioxane
and the maximum adsorption of the four sensitizers 2a–2d on
TiO2 films are quite similar. However, noteworthy is that
considerable differences in the onsets of absorption bands
were observed between the four dyes; the onsets of absorption
bands are red-shifted by 150 nm for 2a and 2b and 90 nm for
2c and 2d. Such a red-shift has been observed also in other
donor–acceptor p-conjugated dyes and attributed to the dye
aggregation on TiO2 electrode.4c,5c On the other hand, the
maximum IPCE values are very similar; 48% for 2a, 46% for
2b, 39% for 2c, and 45% for 2d, and the photocurrent–voltage
curves for four dyes also closely resemble each other. As
shown in Table 4, no significant differences in photovoltaic
parameters are seen among the dyes, although the maximum
amounts of dyes adsorbed on TiO2 film differ considerably.
The occupancy areas of 2a–2d calculated with the maximum
amounts of adsorbed dyes and a real surface area of 1470 cm2
for 1 cm2 geometric area of TiO2 were 106, 186, 588 and
639 A2, respectively, indicating that the dyes 2a and 2b are
more highly aggregated on TiO2 surface than 2c and 2d.
To see explicitly the influence of the dye aggregation on the
probability of charge injection from the dye to the conduction
band of TiO2, photovoltaic parameters were evaluated for
DSSCs with varying adsorption amounts of the dyes. Fig. 6
depicts the peak IPCE value (IPCEmax) plotted against the
light harvesting efficiency (LHE) at that wavelength for 2a and
Photovoltaic performances of DSSCs based on 2a–2d
DSSCs based on 2a–2d were prepared at different adsorption
amounts of the dyes and their photovoltaic performances were
evaluated for the respective DSSCs. The incident photons-to-
current conversion efficiency (IPCE) spectra and photo-
current–voltage curves depicted in Fig. 4 and 5, respectively,
are obtained for the DSSCs with the maximum adsorption of
Table 2 Electrochemical properties of 2a–2d
Compound
Epaa/V (vs. Ag/Ag+
)
Epcb/V (vs. Ag/Ag+
)
DEp/mV
E1/2/V
2a
2b
2c
2d
a
0.31, 0.82, 0.98
0.32, 0.85, 0.99
0.31, 0.80, 0.97
0.32, 0.82, 0.97
0.25, —, —
0.25, —, —
0.25, 0.74, —
0.26, 0.77, —
60,—, —
70,—, —
60, 60, —
60, 50, —
0.28,—, —
0.29,—, —
0.28, 0.77, —
0.29, 0.80, —
b
Epa and Epc are the anodic and cathodic peak potentials in acetonitrile (0.1 M Et4NClO4). Epa and Epc are the anodic and cathodic peak
potentials in acetonitrile (0.1 M Et4NClO4).
ꢀc
This journal is the Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2007
2078 | New J. Chem., 2007, 31, 2076–2082

View Article Online
Table 3 Calculated absorption spectra for 2a–2d
However, for dye 2d, the amounts of adsorbed dyes on TiO2
film was increased to 6.5 ꢂ 1016 molecules cmꢁ2, by using a
thicker TiO2 electrode (ca. 10 mm). As a consequence, the
IPCEmax reached 60%, and the Jsc and Z values were increased
to 2.17 mA cmꢁ2 and 1.02%, respectively. This result strongly
demonstrates that the dyes 2c and 2d with bulky substituents
can avoid effectively the molecular aggregation state even in
the absence of DCA.
Absorption (calc.)
Compound
lmax/nm
fa
CI componentb
2a
402
328
0.92
0.59
HOMO - LUMO (75%)
HOMO - LUMO + 1(43%)
HOMOꢁ1 - LUMO (13%)
2b
2c
407
331
0.89
0.61
HOMO - LUMO (76%)
HOMO - LUMO+1 (43%)
HOMOꢁ1 - LUMO (16%)
Conclusions
405
330
0.87
0.50
HOMO - LUMO (76%)
HOMO - LUMO+1 (34%)
HOMOꢁ1 - LUMO (14%)
Novel donor–acceptor p-conjugated benzofuro[2,3-c]oxazo-
lo[4,5-a]carbazole-type fluorescent dyes exhibiting solid-state
fluorescence have been designed and synthesized, and photo-
voltaic performance of DSSCs based on these dyes are in-
vestigated. A comparison of the IPCEmax vs. LHE plots for 2a
and 2d demonstrates that bulky substituent in 2d prevent the
dye molecules from forming strong intermolecular p–p inter-
actions in the solid state, resulting in higher photovoltaic
performances. We believe that fluorescent dyes exhibiting
solid-state fluorescence can be a class of potential sensitizers
for DSSCs.
2d
409
331
0.88
0.59
HOMO - LUMO (76%)
HOMO - LUMO+1 (42%)
HOMOꢁ1 - LUMO (17%)
a
b
Oscillator strength. The transition is shown by an arrow from one
orbital to another, followed by its percentage CI (configuration
interaction) component.
2d, where LHE is defined as 1 ꢁ T and T means transmittance
due to dyes deposited on the nanoporous TiO2 electrode. Both
of the plots appear to fit straight lines passing through the
origin, suggesting that the charge injection probability is in
proportion to the amount of absorbed photons for the two
sorts of dyes showing highly different aggregation behaviors.
A salient feature to be noted in Fig. 6 is the difference of the
slope values. It is reasonable to presume that less efficient
charge injection for 2a is ascribable to stronger intermolecular
p–p interactions in molecular aggregation state of 2a. Surpris-
ingly, the slope for 2d is close to unity, indicating that almost
all absorbed photons are converted to a current flow. This fact
emphasizes that the introduction of 5-nonyl group as bulky
substituent on the carbazole ring to the chromophores can
efficiently prevent intermolecular energy transfer between the
dyes in molecular aggregation states and lead to a higher
photon-to-current conversion efficiency.
Experimental
Melting points were measured with a Yanaco micro melting
point apparatus MP model. IR spectra were recorded on a
Perkin Elmer Spectrum One FT-IR spectrometer by the ATR
method. Absorption spectra were observed with a Shimadzu
UV-3150 spectrophotometer and fluorescence spectra were
measured with a Hitachi F-4500 spectrophotometer. The
fluorescence quantum yields were determined by a Hama-
matsu C9920-01 equipped with CCD by using a calibrated
integrating sphere system (lex = 325 nm). Cyclic voltammo-
grams were recorded in MeCN–Et4NClO4 (0.1 M) solution
with a three-electrode system consisting of Ag/Ag+ as refer-
ence electrode, Pt plate as working electrode, and Pt wire as
counter electrode, by using a Hokuto Denko HAB-151
potentiostat equipped with a functional generator. Elemental
analyses were made with a Perkin Elmer 2400 II CHN
analyzer. 1H NMR spectra were taken on a JNM-LA-400
(400 MHz) FT NMR spectrometer with tetramethylsilane
(TMS) as an internal standard. Column chromatography
In the presence of DCA, the Jsc values for 2a and 2b were
increased to 2.84 and 2.78 mA cmꢁ2, and the Z values reached
1.62 and 1.49%, respectively. This shows that the co-adsorp-
tion of the DCA prevents dye aggregation and increases the
electron injection yield. Dyes 2c and 2d, on the other hand, did
not adsorb on the TiO2 electrodes in the presence of DCA.
Fig. 3 Calculated electron density changes accompanying the first electronic excitation of 2a–2d. The black and white lobes signify decrease and
increase in electron density accompanying the electronic transition. Their areas indicate the magnitude of the electron density change.
ꢀc
This journal is the Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2007
New J. Chem., 2007, 31, 2076–2082 | 2079

View Article Online
Fig. 4 Absorption spectra of the dyes adsorbed on TiO2 film and the IPCE spectra for DSSCs based on the dyes; (a) 2a, (b) 2b, (c) 2c and (d) 2d.
The amounts of adsorbed dyes on TiO2 film are 13.9 ꢂ 1016, 7.9 ꢂ 1016, 2.5 ꢂ 1016 and 2.3 ꢂ 1016 molecules cmꢁ2 for 2a, 2b, 2c and 2d, respectively.
The 7 mm thick TiO2 electrodes were used.
was performed on silica gel (KANTO CHEMICAL, 60 N,
spherical, neutral).
extract was dried over MgSO4, filtered, and concentrated. The
residue was chromatographed on silica gel (CH2Cl2 as eluent)
to give 1b (0.91 g, yield 82%); mp 230–231 1C; IR (KBr): n~ =
Synthesis of 9-butyl-7-(4-cyanophenyl)-3-
2163 cmꢁ1 1H NMR (CDCl3, TMS) d 1.00–1.03 (9H, m),
;
dibutylaminobenzofuro[2,3-c]oxazolo[4,5-a]carbazole (1b)
1.42–1.68 (10H, m), 2.00 (2H, m), 3.42 (4H, t), 4.92 (2H, t),
6.87 (1H, dd, J = 8.70 and 2.20 Hz), 6.92 (1H, m, J = 2.20
Hz), 7.37–7.41 (1H, m), 7.53–7.57 (1H, m), 7.77 (2H, d, J =
8.21 Hz), 8.10 (1H, d, J = 8.22 Hz), 8.33 (2H, d, J = 8.21 Hz),
8.42 (1H, d, J = 8.70 Hz), 8.62 (1H, d, J = 7.73 Hz);
elemental analysis: calc. (%) for C38H38N4O2: C 78.32, H
6.57, N 9.61; found: C 78.22, H 6.44, N 9.61.
A solution of 1a (1.0 g, 1.90 mmol) in dry acetonitrile was
treated with sodium hydride (60%, 0.23 g, 5.70 mmol) and
stirred for 1 h at room temperature. Iodobutane (1.75 g,
9.49 mmol) was added in dropwise manner over 30 min and
the solution was stirred at room temperature for 5 h. After
concentrating under reduced pressure, the resulting residue
was dissolved in CH2Cl2, and washed with water. The organic
Synthesis of 9-benzyl-7-(4-cyanophenyl)-3-
dibutylaminobenzofuro[2,3-c]oxazolo[4,5-a]carbazole (1c)
A solution of 1a (1.0 g, 1.90 mmol) in dry acetonitrile was
treated with sodium hydride (60%, 0.23 g, 5.70 mmol) and
stirred for 1 h at room temperature. Benzyl bromide (1.62 g,
9.49 mmol) was added in dropwise manner over 30 min and
the solution was stirred at room temperature for 10 h. After
concentrating under reduced pressure, the resulting residue
was dissolved in CH2Cl2, and washed with water. The organic
Table 4 Photovoltaic parameters of DSSCs based on 2a–2d
Molecules
cmꢁ2 a/1016
Dye
Jsc/mA cmꢁ2
Voc/mV
ff
Z (%)
2a
2b
2c
2d
2.00
1.82
1.67
1.73
504
464
452
536
0.54
0.52
0.60
0.58
0.89
0.72
0.74
0.88
13.9
7.9
2.5
2.3
Fig. 5 Photocurrent–voltage curves of the same DSSCs as those used
a
Adsorption amount per unit area of TiO2 film.
in Fig. 4.
ꢀc
This journal is the Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2007
2080 | New J. Chem., 2007, 31, 2076–2082

View Article Online
(CH2Cl2–ethyl acetate = 2 : 5 as eluent) to give 2a (0.91 g,
yield 88%); mp 296–297 1C (decomposition); IR (KBr): n~ =
1
3240, 1692 cmꢁ1; H NMR (DMSO-d6, TMS) d 0.97 (6H, t),
1.37–1.43 (4H, m), 1.58–1.63 (4H, m), 3.38 (4H, t) overlap
peak of dissolved water in DMSO-d6, 6.94–6.96 (1H, m),
7.03–7.05 (1H, m), 7.33–7.38 (1H, m), 7.46–7.50 (1H, m),
7.66 (1H, d, J = 8.08 Hz), 8.20 (2H, d, J = 8.32 Hz), 8.38
(2H, d, J = 8.32 Hz), 8.46 (1H, d, J = 8.32 Hz), 8.60 (1H, d,
J = 7.84 Hz), 12.50 (1H, s, NH); elemental analysis: calc. (%)
for C34H31N3O4: C 74.84, H 5.73, N 7.70; found: C 74.72, H
5.63, N 7.67.
9-Butyl-7-(4-carboxyphenyl)-3-dibutylaminobenzofuro[2,3-c]
oxazolo[4,5-a]carbazole (2b)
Fig. 6 Plots of IPCEmax against LHE for 2a (&) and 2d (n).
A solution of 1b (0.10 g, 0.17 mmol) in ethanol (500 ml) was
added in dropwise aqueous NaOH (0.089 g, 1.72 mmol, 50
mL) with stirring under reflux. After further stirring for 24 h
under reflux, the solution was acidified to pH 4 with 2 M HCl,
and concentrated under reduced pressure. The residue was
dissolved in CH2Cl2, and washed with water. The organic
extract was dried over MgSO4, filtered, and concentrated. The
residue was chromatographed on silica gel (CH2Cl2–ethyl
acetate = 5 : 1 as eluent) to give 2b (0.093 g, yield 90%);
mp 301–302 1C; IR (KBr): n~ = 1693 cmꢁ1; 1H NMR (DMSO-
d6, TMS) d 0.97 (3H, t), 1.04 (6H, t), 1.23–1.26 (2H, m),
1.37–1.41 (4H, m), 1.59–1.63 (4H, m), 1.92–1.97 (2H, m), 4.39
(4H, t), 5.01 (2H, t), 6.94–7.06 (2H, m), 7.37–7.41 (1H, m),
7.53–7.57 (1H, m), 7.83 (1H, d, J = 8.28 Hz), 8.17–8.20 (2H,
m), 8.36–8.38 (2H, m), 8.47–8.49 (1H, m), 8.64 (1H, d, J =
8.28 Hz); elemental analysis: calc. (%) for C38H39N3O4: C
75.85, H 6.53, N 6.98; found: C 75.76, H 6.70, N 6.90.
extract was dried over MgSO4, filtered, and concentrated. The
residue was chromatographed on silica gel (CH2Cl2 as eluent)
to give 1c (0.93 g, yield 79%); mp 253–254 1C; IR (KBr): n~ =
2229 cmꢁ1 1H NMR (acetone-d6, TMS) d 1.04 (6H, t),
;
1.47–1.53 (4H, m), 1.72–1.76 (4H, m), 3.51–3.56 (4H, m),
6.37 (2H, s), 7.03–7.07 (3H, m), 7.20–7.28 (3H, m), 7.39–7.54
(3H, m), 7.78 (1H, d, J = 7.60 Hz), 8.06 (2H, d, J = 8.80 Hz),
8.52–8.57 (3H, m), 8.76 (1H, d, J = 7.60 Hz); elemental
analysis: calc. (%) for C41H36N4O2: C 79.84, H 5.88, N 9.08;
found: C 79.66, H 5.85, N 8.86.
Synthesis of 9-(5-nonyl)-7-(4-cyanophenyl)-3-dibutylaminoben-
zofuro[2,3-c]oxazolo[4,5-a]carbazole (1d)
A solution of 1a (1.0 g, 1.90 mmol) in dry acetonitrile was
treated with sodium hydride (60%, 0.23 g, 5.70 mmol) and
stirred for 1 h at room temperature. 5-Bromononane (1.62 g,
9.49 mmol) was added in dropwise manner over 30 min and
the solution was stirred at room temperature for 24 h. After
concentrating under reduced pressure, the resulting residue
was dissolved in CH2Cl2, and washed with water. The organic
extract was dried over MgSO4, filtered, and concentrated. The
residue was chromatographed on silica gel (CH2Cl2 as eluent)
to give 1d (0.92 g, yield 74%); mp 156–158 1C; IR (KBr): n~ =
9-Benzyl-7-(4-carboxyphenyl)-3-dibutylaminobenzofuro[2,3-c]
oxazolo[4,5-a]carbazole (2c)
A solution of 1c (0.10 g, 0.16 mmol) in ethanol (500 ml) was
added in dropwise aqueous NaOH (0.091 g, 1.16 mmol, 50
mL) with stirring under reflux. After further stirring for 24 h
under reflux, the solution was acidified to pH 4 with 2 M HCl,
and concentrated under reduced pressure. The residue was
dissolved in CH2Cl2, and washed with water. The organic
extract was dried over MgSO4, filtered, and concentrated. The
residue was chromatographed on silica gel (CH2Cl2 as eluent)
to give 2c (0.079 g, yield 77%); mp 250–251 1C; IR (KBr): n~ =
2227 cmꢁ1 1H NMR (acetone-d6, TMS) d 0.72 (6H, t),
;
0.85–0.90 (4H, m), 1.03 (6H, t), 1.26–1.51 (12H, m),
1.68–1.74 (4H, m), 3.50 (4H, t), 4.95–4.98 (1H, m), 6.96–7.00
(2H, m), 7.39–7.52 (2H, m), 7.95–8.05 (3H, m), 8.50–8.52 (3H,
m), 8.74 (1H, d, J = 7.56 Hz); elemental analysis: calc. (%) for
C43H48N4O2: C 79.11, H 7.41, N 8.58; found: C 78.81, H 7.64,
N 8.28.
1647 cmꢁ1 1H NMR (acetone-d6, TMS) d 1.04 (6H, t),
;
1.47–1.53 (4H, m), 1.72–1.76 (4H, m), 3.51–3.55 (4H, m),
6.00 (2H, s), 7.01–7.07 (3H, m), 7.20–7.28 (3H, m), 7.40–7.48
(3H, m), 7.78 (1H, d, J = 7.60 Hz), 8.19 (2H, d, J = 8.40 Hz),
8.43 (2H, d, J = 8.40 Hz), 8.53 (1H, d, J = 8.80 Hz), 8.71 (1H,
d, J = 7.60 Hz); elemental analysis: calc. (%) for C41H37N3O4:
C 77.46, H 5.87, N 6.61; found: C 77.35, H 6.02, N 6.78.
7-(4-Carboxyphenyl)-3-dibutylaminobenzofuro[2,3-c]oxazolo[4,5-a]
carbazole (2a)
A solution of 1a (1.0 g, 1.90 mmol) in ethanol (500 ml) was
added in dropwise aqueous NaOH (0.76 g, 19 mmol, 50 mL)
with stirring under reflux. After further stirring for 16 h
under reflux, the solution was acidified to pH 4 with 2 N
HCl, and concentrated under reduced pressure. The
residue was dissolved in CH2Cl2, and washed with water.
The organic extract was dried over MgSO4, filtered, and
concentrated. The residue was chromatographed on silica gel
9-(5-Nonyl)-7-(4-carboxyphenyl)-3-dibutylaminobenzofuro[2,3-c]
oxazolo[4,5-a]carbazole (2d)
A solution of 1c (0.15 g, 0.23 mmol) in ethanol (500 ml) was
added in dropwise aqueous NaOH (0.13 g, 2.30 mmol, 50 mL)
with stirring under reflux. After further stirring for 24 h under
reflux, the solution was acidified to pH 4 with 2 M HCl, and
concentrated under reduced pressure. The residue was
ꢀc
This journal is the Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2007
New J. Chem., 2007, 31, 2076–2082 | 2081

View Article Online
dissolved in CH2Cl2, and washed with water. The organic
extract was dried over MgSO4, filtered, and concentrated. The
residue was chromatographed on silica gel (CH2Cl2 as eluent)
to give 2d (0.105 g, yield 68%); mp 118–118 1C; IR (KBr): n~ =
in a mixed solvent of THF–DMF–NaOH aq 1 M (5 : 4 : 1),
which was used to determine the amount of dye molecules
adsorbed onto the film by measuring the absorbance. The
quantification of each dye was made based on the lmax and the
molar extinction coefficient of each dye in the above solution.
1
1690 cmꢁ1; H NMR (acetone-d6, TMS) d 0.71 (6H, t), 1.03
(6H, t), 1.18–1.88 (18H, m), 3.50 (4H, t), 5.10–5.15 (1H, m),
7.02 (1H, d, J = 2.20 and 8.80 Hz), 7.06 (1H, d, J = 2.20 Hz),
7.41–7.45 (1H, m), 7.49–7.53 (1H, m), 7.96 (1H, d, J = 8.28
Hz), 8.23 (2H, d, J = 8.56 Hz), 8.49 (2H, d, J = 8.56 Hz), 8.59
(1H, d, J = 8.80 Hz), 8.79 (1H, d, J = 8.28 Hz); elemental
analysis: calc. (%) for C43H49N3O4: C 76.87, H 7.35, N 6.25;
found: C 76.72, H 7.38, N 6.15.
Acknowledgements
We are grateful to Dr N. Koga (Hiroshima University) for the
measurement of absorption spectra of the dyes adsorbed on
TiO2 film. This work was supported by a Grant-in-Aid for
Scientific Research (B) (19350094) and a Grant-in-Aid for
Young Scientist (B) (18750174) from the Ministry of Educa-
tion, Science, Sports and Culture of Japan and by Electric
Technology Research Foundation of Chugoku.
Computational methods
The semi-empirical calculations were carried out with the
WinMOPAC Ver. 3 package (Fujitsu, Chiba, Japan). Geome-
try calculations in the ground state were made using the AM1
method.12 All geometries were completely optimized (keyword
PRECISE) by the eigenvector following routine (keyword
EF). Experimental absorption spectra of the four compounds
were compared with their absorption data by the semi-empiri-
cal method INDO/S (intermediate neglect of differential over-
lap/spectroscopic).13 All INDO/S calculations were performed
using single excitation full SCF/CI (self-consistent field/con-
figuration interaction), which includes the configuration with
one electron excited from any occupied orbital to any
unoccupied orbital, where 225 configurations were considered
[keyword CI (15 15)].
References
1 B. O’Regan and M. Gratzel, Nature, 1991, 353, 737.
¨
2 (a) Q.-H. Yao, F.-S. Meng, F.-Y. Li, H. Tian and C.-H. Haung, J.
Mater. Chem., 2003, 13, 1048; (b) Y.-S. Chen, C. Li, Z.-H. Zeng,
W.-B. Wang, X.-S. Wang and B.-W. Zhang, J. Mater. Chem.,
2005, 15, 246; (c) K. R. J. Thomas, J. T. Lin, Y.-C. Hsu and K.-C.
Ho, Chem. Commun., 2005, 4098; (d) D. P. Hagberg, T. Edvinsson,
T. Marinado, G. Boschloo, A. Hagfeld and L. Sun, Chem.
Commun., 2006, 2245.
3 T. Horiuchi, H. Miura, K. Sumioka and S. Uchida, J. Am. Chem.
Soc., 2004, 126, 12218.
4 (a) K. Hara, K. Sayama, Y. Ohga, A. Shinpo, S. Suga and H.
Arakawa, Chem. Commun., 2001, 569; (b) K. Hara, M. Kurashige,
S. Ito, A. Shinpo, S. Suga, K. Sayama and H. Arakawa, Chem.
Commun., 2003, 252; (c) K. Hara, M. Kurashige, Y. Dan-oh, C.
Kasada, A. Shinpo, S. Suga, K. Sayama and H. Arakawa, New J.
Chem., 2003, 27, 783.
5 (a) Z.-S. Wang, F.-Y. Li, C.-H. Hang, L. Wang, M. Wei, L.-P. Jin
and N.-Q. Li, J. Phys. Chem. B, 2000, 104, 9676; (b) A. Ehret, L.
Stuhl and M. T. Spitler, J. Phys. Chem. B, 2001, 105, 9960; (c) K.
Hara, T. Sato, R. Katho, A. Furube, Y. Ohga, A. Shinpo, S. Suga,
K. Sayama, H. Sugihara and H. Arakawa, J. Phys. Chem. B, 2003,
107, 597.
Preparation of the dye-sensitized solar cells based on dyes 2a–2d
The TiO2 electrodes used for dye-sensitized solar cells were
prepared as follows. Into a powder (1.30 g) of TiO2 (P-25, d =
30–40 nm) in mortar was added water (1.9 mL) in six portions
with rapid stirring. Then, three drops of 12 M nitric acid and
poly(ethylene glycol) (80 mg) were successively added, and the
mixture was well kneaded to a smooth paste. It was then
applied on a fluorine-doped-tin-oxide (FTO) substrate and
sintered for 30 min at 500 1C. The 7 mm thick TiO2 electrode
(0.5 ꢂ 0.5 cm2 in photoactive area) was immersed into a
0.3 mM acetonitrile solution (with or without 20 mM DCA)
of the dye for a number of hours enough to adsorb the
photosensitizer. The DSSCs were fabricated by using the
TiO2 electrode thus prepared, Pt-coated glass as a counter
electrode, and a solution of 0.05 M iodine, 0.1 M lithium
iodide and 0.6 M 1,2-dimethyl-3-n-propylimidazolium iodide
in acetonitrile as electrolyte. The photocurrent–voltage char-
acteristics were measured using a potentiostat under a simu-
lated solar light (AM 1.5, 60 mW cmꢁ2). IPCE spectra were
measured under monochromatic irradiation with a tungsten-
halogen lamp and a monochromator. BET surface areas of the
TiO2 particles were 73.5 m2 gꢁ1. Consequently, the total area
for 1 cm2 geometric area of the film was ca. 1470 cm2 for 7 mm
thick TiO2 films. Absorption spectra of the dye-adsorbed TiO2
films were measured in the diffuse-reflection mode by a
JASCO UV-VIS spectrophotometer with a calibrated integrat-
ing sphere system ISV-469. The dye-coated film was immersed
6 S.-L. Li, K.-J. Jiang, K.-F. Shao and L.-M. Yang, Chem. Com-
mun., 2006, 2792.
7 A. Kay and M. Gratzel, J. Phys. Chem., 1993, 97, 6272.
¨
8 (a) Y. Ooyama and Y. Harima, Chem. Lett., 2006, 902; (b) Y.
Ooyama, Y. Kagawa and Y. Harima, Eur. J. Org. Chem., 2007,
3613.
9 (a) K. Yoshida, Y. Ooyama, H. Miyazaki and S. Watanabe, J.
Chem. Soc., Perkin Trans. 2, 2002, 700; (b) Y. Ooyama, T.
Nakamura and K. Yoshida, New J. Chem., 2005, 29, 447; (c) Y.
Ooyama, T. Okamoto, T. Yamaguchi, T. Suzuki, A. Hayashi and
K. Yoshida, Chem.–Eur. J., 2006, 7827.
10 (a) K. Yoshida, Y. Ooyama, S. Tanikawa and S. Watanabe, J.
Chem. Soc., Perkin Trans. 2, 2002, 708; (b) Y. Ooyama and K.
Yoshida, New J. Chem., 2005, 29, 1204.
11 H. Langhals, T. Potrawa, H. Noth and G. Linti, Angew. Chem.,
Int. Ed. Engl., 1989, 28, 478.
12 (a) J. E. Ridley and M. C. Zerner, Theor. Chim.Acta, 1973, 32, 111;
(b) J. E. Ridley and M. C. Zerner, Theor. Chim. Acta, 1976, 42,
223; (c) A. D. Bacon and M. C. Zerner, Theor. Chim. Acta, 1979,
53, 21.
13 M. J. S. Dewar, E. G. Zoebisch, E. F. Healy and J. J. Stewart,
J. Am. Chem. Soc., 1985, 107, 389.
14 (a) M. Adachi, Y. Murata and S. Nakamura, J. Org. Chem., 1993,
58, 5238; (b) W. M. F. Fabian, S. Schuppler and O. S. Wolfbesis,
J. Chem. Soc., Perkin Trans. 2, 1996, 853.
¨
ꢀc
This journal is the Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2007
2082 | New J. Chem., 2007, 31, 2076–2082