An expanded conjugation photosensitizer with two different adsorbing groups for solar cells

Article abstract of DOI:10.1039/b302552g

Three novel hemicyanine derivatives bearing different electron donors, 2-[4-N, N-dimethylaminostyryl]-β-naphthothiazolium propylsulfonate (dye 1), 2-[4-N, N-diethylaminostyryl]-β-naphthothiazolium propylsulfonate (dye 2) and 2-[2-hydroxyl-4-N, N-diethylam

Full text of DOI:10.1039/b302552g

View Article Online / Journal Homepage / Table of Contents for this issue
An expanded conjugation photosensitizer with two different adsorbing
groups for solar cells
a
Qiao-Hong Yao, Lu Shan, Fu-You Li,* Dong-Dong Yin and Chun-Hui Huang*
b
a
b
a
a
State Key Laboratory of Rare Earth Materials Chemistry and Applications, Peking University,
Beijing 100871, P. R. China. E-mail: hch@chem.pku.edu.cn; leef@chem.pku.edu.cn;
Fax: +86-(10) 6275 1708; Tel: +86-(10) 6275 7156
b
Department of Chemistry, Beijing Normal University, Beijing 100875, P. R. China
Received (in New Haven, CT, USA) 6th March 2003, Accepted 15th April 2003
First published as an Advance Article on the web 7th July 2003
Three novel hemicyanine derivatives bearing different electron donors, 2-[4-N, N-dimethylaminostyryl]-
b-naphthothiazolium propylsulfonate (dye 1), 2-[4-N, N-diethylaminostyryl]-b-naphthothiazolium
propylsulfonate (dye 2) and 2-[2-hydroxyl-4-N, N-diethylaminostyryl]-b-naphthothiazolium propylsulfonate
(
dye 3), were designed and synthesized for use in dye-sensitized mesoporous TiO solar cells. Their visible light
2
absorption, luminescence emission, electrochemical properties, and photoelectrochemical properties were
systematically studied. Upon comparison, dye 3 was found to be the best sensitizer. Experimental data show
that introduction of a hydroxyl group can greatly enhance the electron injection efficiency, moreover,
this enhancement can greatly override the unfavourable factor originated from the decrease of the
adsorption amount of dye 3. Additionally bearing the hydroxyl group, the dye 3 sensitized solar cell gained
ꢀ
2
remarkably high photoelectric conversion yield of 6.3% under illumination of 80.0 mW cm white light from a
Xe lamp.
on mesoporous TiO
electricity conversion of over 2%.
2
electrode, and gave an overall light to
Introduction
1
9,20
Recently, we found that
films with hydrochloric acid, the
As an inexpensive alternative to the conventional silicon
photovoltaic cell, dye-sensitized solar cell (DSSC) is attracting
more and more interest because of its high photon to electron
efficiencies and long life cycle.
used in DSSC, the ruthenium polypyridyl derivatives, as a class
after a pretreating of TiO
2
ꢀ
conversion yield of a hemicyanine dye (containing the –RSO
group) sensitized TiO
3
2
solar cell can be greatly enhanced up to
1
–3
21
Among the photosensitizers
5.1%. The results encouraged us to design and synthesize new
hemicyanine dyes in a quest for more efficient photosensitizers.
In this paper, we describe how we designed and synthesized
three new hemicyanine derivatives (Scheme 1), 2-[4-N, N-
dimethylaminostyryl]-b-naphthothiazolium propylsulfonate
(dye 1), 2-[4-N, N-diethylaminostyryl]-b-naphthothiazolium
propylsulfonate (dye 2) and 2-[2-hydroxyl-4-N, N-diethylami-
nostyryl]-b-naphthothiazolium propylsulfonate (dye 3), and
systematically studied their visible light absorption, lumines-
cence emission, electrochemical properties, and photoelectro-
chemical properties in order to clarify the effect of different
electron donors on the photoelectric conversion properties of
the dyes. Dye 3 with two adsorption functional groups was
found to be the best when it served as a sensitizer and was
fabricated into a solar cell.
4
–6
of star dyes, have been studied extensively.
On the other
hand, pure organic dyes, compared with metal complexes,
have higher absorption coefficients, which could save the
amount of dyes to be used and reduce the thickness of the
semiconductor film. Moreover, organic dyes are easier to pre-
pare and cost less compared with those of metal Ru based
complexes. Recently, organic dye sensitized TiO
have made great progress, and the highest overall yield of solar
cells sensitized by pure organic dyes has exceeded 5%.
organic dyes are promising and expected to be a new type of
sensitizer for DSSCs.
In the last several years, our group has systematically stu-
died the photoelectric conversion properties of hemicyanine
dyes by the Langmuir–Blodgett technique.
results show that hemicyanine dyes with a strong donor
and acceptor linked by a p-conjugation bridge can produce
a higher quantum yield for photon-to-electron conversion.
7
–21
2
solar cells
1
7,21
So
2
2–27
Experimental
Experimental section
ꢀ
Materials
Later, we introduced the –RSO
3
as attaching group into
hemicyanine dyes and used these dyes as photosensitizers in
DSSCs. The dyes produced good charge separation properties
Titanium(IV) tetraisopropoxide, propylene carbonate (PC),
2-methyl-b-nathphothiazole, 4-(dimethylamino)benzaldehyde,
Scheme 1 Molecular structures of the three dyes.
DOI: 10.1039/b302552g
New J. Chem., 2003, 27, 1277–1283
1277
This journal is # The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2003

View Article Online
1
62.66; H, 5.41; N, 5.57. H NMR (500 MHz, DMSO):
4
hyde and 1,3-propane sultone were purchased from Acros.
-(diethylamino)benzaldehyde,
4-(diethylamino)salicylalde-
d ¼ 1.18 (t, 6H, 2NCH CH , J ¼ 7.01 Hz), 2.44 (m, 2H,
2
3
ꢀ
ꢀ
N-Methyl-4-methylpyridinium iodide (PyI) was synthesized
colloidal ethanol solution
–CH
(q, 4H, 2NCH
2
SO
3
), 2.86 (t, 2H, –CH
2
CH
2
SO
3
, J ¼ 5.80 Hz), 3.46
2
2
according to the literature. TiO
8
+
2
2
, J ¼ 7.09 Hz), 5.27 (t, 2H, N CH , J ¼ 8.21
was prepared by hydrolysis of titanium(IV) tetraisopropoxide
in the acidified ethanol (pH around 3) at a temperature of
Hz), 6.21 (s, 1H, Ar–H), 6.44 (d, 1H, –CH=, J ¼ 9.27 Hz),
7.78–7.88 (m, 3H, Ar–H), 8.10 (d, 1H, Ar–H, J ¼ 9.09 Hz),
8.17–8.26 (m, 4H, Ar–H), 8.76 (d, 1H, –CH=, J ¼ 8.62 Hz),
ꢁ
29
C. All other solvents and chemicals used were produced
0
by Beijing Chemical Factory, China (reagent grade) and used
as received. Optically transparent conducting glass (CTO glass,
10.67 (s, 1H, –OH).
2
fluorine-doped SnO over layer, transmission > 70% in the
visible region, sheet resistance 20 O per square) was obtained
from the Institute of Nonferrous Metals of China.
Preparation of the nanocrystalline TiO
2
electrodes
Detailed procedures for preparing nanocrystalline TiO
2
films
The thickness of
film used in this work was ca. 5 mm. The TiO film elec-
2
0,32
have been described in the literature.
TiO
Synthesis of the photosensitizers
2
2
ꢀ
4
trode thus obtained was coated with dye in a 3 ꢂ 10 mol
Dye 1 was synthesized according to the following two-step
as illustrated in Scheme 2.
ꢀ3
dm chloroform solution for at least 12 h at room tempera-
3
0,31
procedure,
ture. After completion of the dye adsorption, the electrode
was withdrawn from the solution, washed with chloroform,
and dried under a stream of dry air.
Step one: synthesis of 2-methyl-b-naphthothiazolium pro-
pylsulfonate. 1.99 g (0.01 mol) 2-methyl-b-nathphothiazole
and 1.22 g (0.01 mol) 1,3-propane sultone were refluxed for
4
8 h in 20 mL benzene. When cooled to ambient temperature,
the reaction mixture was filtered and then washed with ben-
zene to yield 2.18 g (68%) of white solid.
Step two: synthesis of dye 1. 0.321 g (0.001 mol) 2-methyl-b-
naphthothiazolium propylsulfonate and 0.149 g (0.001 mol) 4-
Methods
1
H NMR spectra were recorded on a Bruker Avance 500
spectrometer, and elemental analyses were performed on a
Vario EL elemental analyzer. Film thickness was determined
with a DEKTAK 3 profilometer. Absorption measurements
were made with a Shimadzu mode 3100UV-VIS-NIR spectro-
photometer and fluorescence measurements with a Hitachi
F-4500 fluorescence spectrophotometer. Cyclic voltammetry
(CV) was carried out in electrochemical cells on a Model
600 voltammetric analyzer (CH Instruments, Cordova, TN).
A three-electrode cell was composed of a glassy carbon as
working electrode, a platinum wire as counter electrode,
(
dimethylamino)benzaldehyde were added into 30 mL absolute
ethanol and refluxed for 16 h, with piperidine as the catalyst.
After being cooled to ambient temperature, the mixture was
concentrated to about 10 mL and then 30 mL dimethyl ether
were added. Finally, the mixture was placed in a refrigerator
and kept overnight. The resulting precipitate was filtered and
recrystallized from ethanol to give a purple red product:
ꢁ
yield 78%, mp 262–263 C. Elemental analysis: calcd for
24 24 2 3 2
C H N O S (%), C, 63.69; H, 5.34; N, 6.19; found, C,
1
3.62; H, 4.97; N, 5.92. H NMR (500 MHz, DMSO):
6
and a Ag/AgCl as reference electrode. The supporting elec-
ꢀ3
3ꢀ
2
2
3
ꢀ
4
d ¼ 2.46 (m, 2H, –CH
2
SO ), 2.88 (t, 2H, –CH
CH
SO
,
,
trolyte was 0.1 mol dm LiClO
in acetonitrile. The scan
+
), 5.38 (t, 2H, N CH
ꢀ1
J ¼ 6.11 Hz), 3.12 (s, 6H, 2N–CH
3
2
rate was 0.1 V s . All potentials reported in this work refer
to the Ag/AgCl electrode.
J ¼ 7.14 Hz), 6.84 (d, 2H, J ¼ 8.28 Hz), 7.83 (t, 1H, CH=,
J ¼ 7.23 Hz), 7.90 (t, 1H, CH=, J ¼ 7.82 Hz), 8.01–8.11 (m,
The photovoltaic performance of the solar cells based on
hemicyanine dyes was carried out in a standard two-electrode
4
H, Ar–H), 8.23–8.33 (m, 3H, Ar–H), 8.78 (d, 1H, Ar–H,
3
,33
J ¼ 8.52 Hz).
system.
2
Dye-coated TiO film as working electrode was
Dyes 2 and 3 were synthesized according to the same proce-
dure as dye 1 except that 4-(dimethylamino)benzaldehyde was
replaced with 4-(diethylamino)benzaldehyde or 4-(diethyl-
amino)salicylaldehyde, respectively.
placed on top of an ITO glass as a counter electrode, on which
the 200 nm thick Pt was sputtered. The redox electrolyte was
introduced into the interelectrode space by capillary force.
The solar cells were illuminated in the front through the con-
duction glass substrate. The effective area is 0.188 cm and
the redox electrolyte solution was composed of 0.5 mol dm
LiI, 0.05 mol dm I2 , and 0.6 mol dm N-Methyl-4-methyl-
pyridinium iodide (PyI) in propylene carbonate (PC). A 500 W
xenon lamp (Ushio Electric, Tokyo, Japan) served as a white
light source in conjunction with an IRA-25S filter (Schott,
USA) and a GG400 cutoff filter (Toshiba, Japan). Monochro-
matic light in the range of 400 nm to 800 nm was produced
by setting an IRA-25S filter and a suitable band-pass filter
(Schott, USA) in the path of the light beam before the solar
cell. The incident light intensity was measured with a Light
Gauge radiometer/photometer (Coherent, USA).
ꢁ
2
Dye 2: mp 271–272 C. Elemental analysis: calcd for
ꢀ3
C
6
26
H
28
N
2
O
3
S
2
(%), C, 64.97; H, 5.89; N, 5.83; found, C,
4.61; H, 5.61; N, 5.61. H NMR (500 MHz, DMSO):
1
ꢀ3
ꢀ3
d ¼ 1.17 (t, 6H, 2NCH
2
CH
3
, J ¼ 6.60 Hz), 2.47 (m, 2H,
ꢀ
ꢀ
–
CH SO ), 2.88 (t, 2H, –CH CH SO , J ¼ 6.36 Hz), 3.51
2
3
2
2
3
+
(
q, 4H, 2NCH
2
, J ¼ 6.78 Hz), 5.36 (t, 2H, N CH
2
,
J ¼ 7.56 Hz), 6.81 (d, 2H, Ar–H, J ¼ 8.49 Hz), 7.82 (t, 1H,
CH=, J ¼ 7.41 Hz), 7.89 (t, 1H, Ar–H, J ¼ 7.33 Hz), 7.96–
8
.07 (m, 4H, Ar–H), 8.21–8.31 (m, 3H, Ar–H), 8.77 (d, 1H,
CH=, J ¼ 8.59 Hz).
Dye 3: mp 254–255 C. Elemental analysis: calcd for
(%), C, 62.88; H, 5.68; N, 5.64; found, C,
ꢁ
26 28 2 4 2
C H N O S
Scheme 2 The synthetic procedure of the three dyes.
1
278
New J. Chem., 2003, 27, 1277–1283

View Article Online
Table 1 Absorption data of the hemicyanine dyes in different solvents (5 ꢂ 10ꢀ mol dmꢀ3
6
)
Dye 1
Dye 2
Dye 3
ꢀ
1
ꢀ1
ꢀ1
ꢀ1
ꢀ1
ꢀ1
Absmax
e
(M cm
)
Absmax
e
(M cm
)
Absmax
e
(M cm
)
Dielectric constant
Methanol
Ethanol
532
536
538
547
566
531
530
532
532
13 125
5305
5556
8333
6153
4597
4431
11 639
7333
542
548
554
557
578
542
540
542
542
27 667
17 800
18 133
25 350
22 283
20 633
26 017
31 117
21 567
552
574
566
560
565
557
554
559
565
43 143
30 964
31 125
32 714
54 161
24 250
12 214
29 321
30 089
32.6
24.3
13.9
8.1
Pentanol
Decanol
Chloroform
Acetone
Acetonitrile
DMSO
4.8
20.7
37.5
46.5
109.5
DMF
Results and discussion
extinction coefficient enlarged in the sequence of dye 1, dye 2
and dye 3. We believe that this fact is originated from the
increase of the delocalization extent of the dyes.
Photophysical characteristics of the dyes in solution
The luminescence properties for the three dyes in several lin-
ear alcohol solutions are also measured and the results are
listed in Table 2. It appears that the fluorescence maxima red
shifted slightly and the relative fluorescence intensities
increased gradually for the three dyes with decreasing polarity
of the solvent, that is, the relative fluorescence intensities
increased when the solvents changed in the sequence methanol,
ethanol, pentanol, and decanol. For example, for dye 2 in
methanol solution the fluorescence maximum is at 604 nm
and the relative fluorescence intensity is 4.6, while in decanol
solution the former changed to 608 nm and the latter increased
by a factor of 7 compared with that in methanol solution. A
similar tendency was also observed for both dye 1 and dye 3.
Absorption properties for the three dyes in different solvents
are listed in Table 1, and, as an example, the absorption spec-
tra for the three dyes in methanol are illustrated in Fig. 1.
From Table 1, it is easy to find a symmetric solvatochromic
effect on the photophysical behavior of dye 1 and dye 2 in lin-
ear alcohol and chloroform solvents, that is, the absorption
peak of a dye in solution is normally blue shifted with increas-
ing the polarity of solvent. In solvents other than alcohols and
chloroform, the solvatochromic phenomenon became compli-
cated due to the tendency of the molecules to form aggre-
1
1,12
gates.
It is interesting to find that even in linear alcohol
solvents, the maximum absorption wavelength of dye 3 did
not show any solvatochromic phenomenon, indicating that
the dye 3 molecules, additionally bearing a hydroxyl group,
have stronger tendency to aggregate. However, an obvious ten-
dency can be concluded from the complicated data in Table 1
that in any kind of solvents, except for acetonitrile and
DMSO, the absorption maxima red shifted and the molar
ꢀ6
On the other hand, in any solvent investigated (5 ꢂ 10 mol
ꢀ
3
dm ), the change in the sequence of relative fluorescence
intensity is dye 1 < dye 2 < dye 3. For example, in ethanol
solution, the relative emission intensities are 3.2, 10.4, and
3
6.1 for dye 1, dye 2, and dye 3, respectively. The value for
dye 2 is three times as high as that for dye 1, suggesting that
the replacement of the N,N-dimethyl group by an N,N-diethyl
group favors fluorescence emission. Comparing dye 3 with dye
1
and dye 2, the relative fluorescence intensity for dye 3 in etha-
nol solution is 11.2 and 3.5 times larger than that for dye 1 and
dye 2, indicating that, besides the replacement of the N,N-
dimethyl group by an N,N-diethyl group, the introduction of
hydroxyl group also favors fluorescence generation.
When TiO nanoparticles are introduced into ethanol solu-
2
ꢀ
3
tion (10 g dm ), the relative fluorescence intensity of the
three dyes is drastically decreased. The relative emission
intensities decreased from 3.2, 10.4 and 36.1 to 1.4, 8.6 and
1
6.3 for dyes 1, 2, and 3, respectively. As an example, the
fluorescence spectra for dye 1 both in ethanol solution and
in TiO2 ethanol colloid solution are given in Fig. 2. Since
fluorescence emission is produced by the excited electrons
through deactivating from the singlet excited state to the
ground state, the decreased fluorescence emission elucidated
efficient electron injection from the excited state dye mole-
Fig. 1 Absorption spectra of the three dyes in methanol solution
(
5 ꢂ 10ꢀ mol dm ).
6
ꢀ3
2
cules to the TiO particles.
Table 2 Fluorescence properties for the three dyes in several linear alcohol solutions (5 ꢂ 10ꢀ6 mol dmꢀ3
)
a
Methanol
Ethanol
Pentanol
Decanol
2
TiO colloid
b
b
b
b
b
FLmax
FLintensity
FLmax
FLintensity
FLmax
FLintensity
FLmax
FLintensity
FLmax
FLintensity
dye 1
dye 2
dye 3
602
604
600
1.7
4.6
604
608
601
3.2
10.4
36.1
606
608
603
5.3
18.3
70.4
607
608
604
6.6
33.1
603
608
600
1.4
8.6
34.3
132.2
16.3
a
b
in ethanol (10 g/L). arbitrary unit.
New J. Chem., 2003, 27, 1277–1283
1279

View Article Online
Table 3 Parameters of the dye sensitized solar cells
a
b
sc
I
sc
V
oc
Z
(%) (molecules cm
N
I
(mA molecule
ꢀ2
ꢀ2
ꢀ1
(
mA cm
)
(mV) FF
)
)
17
17
16
ꢀ17
ꢀ17
ꢀ16
Dye 1 10.7
Dye 2 11.1
Dye 3 15.6
496
468
512
0.616 4.1 2.3 ꢂ 10
4.6 ꢂ 10
7.4 ꢂ 10
3.5 ꢂ 10
0.615 4.0 1.5 ꢂ 10
0.631 6.3 4.4 ꢂ 10
ꢀ
2
a
Illumination: 80.0 mW cm white light from a Xe lamp. N : adsorption amount
film; I bs c: short-circuit photocurrent generated
of dye per unit square area of TiO
per dye molecule.
2
3ꢀ
tion with other anchoring group of –RSO , the contact area
between the dye 3 molecule and the TiO nanoparticle maybe
2
increased, which subsequently leads to a decreased adsorption
of dye 3 per unit surface area of TiO film. The decrease in
ꢀ
6
ꢀ3
Fig. 2 Fluorescence spectra of 5 ꢂ 10 mol dm dye 1 in ethanol
2
ꢀ3
solution (solid line) and in 10 g dm TiO
dashed line).
2
ethanol colloid solution
adsorption would have a negative impact on the photoelectric
conversion properties of dye 3. Alternatively, the enlarged
contact area is maybe caused by two adsorption groups exist-
ing in the molecule, which possibly favors electron injection
from the excited state dye molecules to the conduction band
(
Photophysical characteristics of the dyes on TiO electrode
2
of the TiO electrode. This point of view is well supported by
2
the data discussion later.
Absorption spectra of the three dyes on TiO films are shown
2
in Fig. 3. The absorption peaks for the three dyes on TiO
2
films are all extremely broadened, compared with their corre-
sponding absorption peaks in methanol solutions (Fig. 1),
suggesting that the dye molecules may aggregate in plane-
to-plane stacking to form H-aggregate or in a head-to-tail
Energy levels of HOMO and LUMO
To thermodynamically judge the possibility of electron trans-
fer from the excited dye molecules to the conduction band
of TiO , cyclic voltammetry was performed to determine the
2
3
4
arrangement to form J-aggregate, when more and more
dye molecules adsorb on TiO film. From Fig. 3 it is easily
2
redox potentials for the three hemicyanine dyes. Redox pot-
entials of 0.83, 0.82 and 0.80 V vs. Ag/AgCl, calculated by
averaging the related oxidation and reduction potentials, are
roughly regarded as HOMO energy levels for dyes 1, 2 and
seen that the absorption peaks of dye 1 and dye 2 are almost
of the same breadth while dye 3 has a slightly narrower-band
spectrum. To determine the exact amount of the dyes
adsorbed on TiO
methanol to desorb the dye completely. The yield solution
25 mL) was used to measure its absorbance, which was
2
films, dye-coated films were soaked in
3. The absorption maximum in the visible region is at 563,
65 and 529 nm, respectively, for dyes 1, 2 and 3, which corre-
5
(
sponds to the energy gap between the HOMO and LUMO
energy levels for the three dyes. So the HOMO and LUMO
energy levels for dye 1 are evaluated to be ꢀ5.55 eV and
divided by the molar extinction coefficient and the thickness
of liquid layer, to give the number of the dye molecules
adsorbed on TiO
per square centimeter TiO2 film for the three dyes is listed
2
films. The number of adsorbed molecules
ꢀ
3.35 eV on the absolute scale, with reference to the redox
1
7
potential 0.83 V (vs. Ag/AgCl) and the band gap 2.20 eV
563 nm). Similarly, the HOMO and LUMO energy levels
in Table 3. The number for dye 1 is 2.3 ꢂ 10 . When the
N,N-dimethylanilino group is replaced by the N,N-diethylani-
lino group, the amount of dye 2 adsorbed decreases by about
(
on the absolute scale are assumed to be ꢀ5.54 eV and ꢀ3.34
eV for dye 2, ꢀ5.52 eV and ꢀ3.17 eV for dye 3, respectively.
Obviously, the excited-state energy levels for the three dyes
4
increased steric effect. The number of adsorbed molecules
0%. This decrease in adsorption is possibly caused by the
1
6
are all higher than the energy level of TiO conduction band
2
for dye 3 is 4.4 ꢂ 10 , which is about 19% of that for dye
3
edge (ꢀ4.40 eV), showing that the electron injection should
1
and 29% of that for dye 2. It is believed that besides the
be possible thermodynamically. No emission signal was
observed for the three dyes on TiO2 films, suggesting that
the injection of the excited electron from the excited dye to
steric effect there might be other factors that cause dye 3 to
have such low adsorption. Comparing the structure of dye 2,
dye 3 has a hydroxyl group on the electron donor part.
This hydroxyl group can bond with the oxygen atom on
the TiO particles is efficient.
2
the TiO film through H-bonding. Considering the coopera-
2
Photoelectrochemical properties of the dye-sensitized
TiO electrode
2
Photocurrent action spectra for the dye-coated TiO electrodes
2
are shown in Fig. 4. The monochromatic incident photo-to-
electron conversion efficiency (IPCE), defined as the number
of electrons generated by light in the outer circuit divided by
the number of incident photons, was obtained by the following
equation:
240IscðmA cmꢀ
2
Þ
1
IPCEð%Þ ¼
lðnmÞPinðW mꢀ
2
Þ
where the constant 1240 is derived from units conversion, Isc is
the short-circuit photocurrent generated by monochromatic
light, and l is the wavelength of incident monochromatic light,
whose intensity is Pin . The losses of the light reflection and
absorption by the conducting glass were not corrected. Com-
paring the action spectra with the absorption spectra on
2
Fig. 3 Absorption spectra of the three dyes on TiO electrode.
1
280
New J. Chem., 2003, 27, 1277–1283

View Article Online
where Vopt and Iopt are voltage and current for maximum
power output respectively, Voc and Isc are open-circuit photo-
voltage and short-circuit photocurrent, respectively. The over-
all yield (Z) is expressed by the following equation:
Z ¼ ðVocIscFFÞ=Pin
where Pin is the power of incident white light. Here, the power
ꢀ
2
of incident white light is 80.0 mW cm . For dye 1 modified
ꢀ
2
solar cell, a short-circuit photocurrent of 10.7 mA cm
,
open-circuit photovoltage of 496 mV, and a fill factor of
0
4
.616 were obtained, corresponding to an overall yield of
.1%. According to our analysis from the action spectrum,
dye 2 generated a similar value of 4.0% (with I of 11.1 mA
sc
ꢀ
2
cm , Voc of 468 mV, and 0.615 for fill factor). Dye 3 gener-
ꢀ
2
ated the highest short-circuit photocurrent of 15.6 mA cm
2
Fig. 4 Action spectra of dye sensitized TiO electrodes. Effective area
ꢀ2
due to its higher IPCE values in a broad range. With a higher
open-circuit photovoltage of 512 mV and 0.631 for fill factor,
dye 3 gained a significantly high overall yield of 6.3%.
for illumination is 0.188 cm ; the IPCE values are not corrected for
the loss of light intensity due to absorption by the conducting glass
substrate.
2
TiO film (cf. Fig. 3), all three dyes have spectral selectivity for
Stability of dye 3 sensitized solar cell
different regions of visible light, that is, they convert more
efficiently the light in the shorter wavelength region. Despite
the spectral selectivity, the action spectra and the absorption
spectra of the three dyes overlapped well, indicating that the
photoelectrochemical conversion is achieved through photo-
sensitization, namely that dye molecules absorbed photons
and generated excited electrons and the excited electrons sub-
In the same way as for the measurement of I–V curves, the sta-
bility of dye 3 on the TiO electrode was tested in an unsealed
2
solar cell under illumination of white light from a Xe lamp.
Fig. 6A gives the time course of the light to electricity con-
version efficiency. It was seen that the cell showed a stable
efficiency up to 170 h. After that, the efficiency decreased sud-
denly because of an accident of the circuit. When the circuit
was fixed, the efficiency recovered to the same level as before
the accident, and then the efficiency decreased gradually after
sequently transferred to the conduction band of the TiO
trode.
to photocurrent in the region from 400 nm to 700 nm. The
2
elec-
All the three dyes can efficiently convert visible light
1
2,13
2
00 h. From Fig. 6B, C and D, we can see that, during the time
maximum IPCE values for dyes 1 and 2 are both about
4
course and except for the accident, FF increased gradually and
the Voc kept almost the same, while the Isc decreased slowly.
Before 200 h, the efficiency remained constant because the
decrease of the Isc was compensated by the increase of FF.
With the time over 200 h, FF increased very slowly while the
0%, while that for dye 3 is 57%. Even though the spectrum
of dye 1 is a little broader than that of dye 2, the action spectra
feature of the two dyes differ little, suggesting dye that 1 and
2
dye 2 sensitized TiO electrodes would generate the same over-
all yield efficiency. The IPCE values of dye 3, in the whole
photosensitization range from 400 nm to 700 nm, are about
Isc decreased quickly. Since the two changes can not be com-
pensated by each other, this consequently led to the decrease
1
.3 times higher than that of dyes 1 and 2, therefore, dye 3
has higher electron injection efficiency than dyes 1 and 2. From
the combination of its high IPCE and the broad feature of its
action spectrum, it is concluded that dye 3 should be a superior
sensitizer to either dye 1 or dye 2.
The photoelectrochemical properties of dye sensitized TiO
2
electrode are given in Table 3, while the photocurrent–voltage
curves are shown in Fig. 5. Fill factor (FF) is given by the fol-
lowing equation:
FF ¼ ðVoptIoptÞ=ðVocIscÞ
Fig. 5 Current–voltage curves of dye sensitized TiO
mination: 80.0 mW cm white light from a Xe lamp. Effective area:
2
solar cells. Illu-
ꢀ2
Fig. 6 Stability test of the dye 3 sensitized TiO solar cell under illu-
mination of 70.0 mW cm white light from a Xe lamp. Other condi-
tions as in Fig. 5.
2
ꢀ2
ꢀ3
ꢀ3
ꢀ2
0.188 cm . Electrolyte: 0.5 mol dm LiI, 0.05 mol dm
mol dm PyI in PC.
I
2
and 0.6
ꢀ3
New J. Chem., 2003, 27, 1277–1283
1281

View Article Online
in the efficiency. The results show that the decrease of the cell
efficiency was directly due to the decrease of I , and the main
Conclusion
sc
reason for the Isc decrease was electrolyte loss through eva-
poration since the dark purple color of the dye on the electrode
did not change during the stability test. Alternatively, we also
Three new hemicyanine dyes, bearing the same electron
acceptor but different electron donors, were synthesized and
used as photosensitizer on the nanocrystalline TiO electrode.
2
Through investigation the relationship between the photoelec-
trochemical properties and structures of the three hemicyanine
dyes, it is found that dye 3 is the best sensitizer, which has a
conversion yield of 6.3%, with a short-circuit photocurrent
checked the stability of dye 3 on TiO
the change of the absorption properties. The dye-loaded TiO
film was covered with the redox electrolyte solution (50 mL
2
electrode by measuring
2
ꢀ2
cm ) to form an investigating electrode (abbr. TiO
after being dried in air. For comparison, the dye 3 loaded
TiO electrode (abbr. TiO /D) was also prepared and served
2
/D/E)
ꢀ
2
of 15.6 mA cm , an open-circuit voltage of 512 mV and a fill
ꢀ
2
factor of 0.631 under illumination of 80.0 mW cm white light
from a Xe lamp. In addition, in an unsealed solar cell, dye 3
can generate a constant conversion yield for more than 200
h. It is found that the main reason for the gradual decrease
of the efficiency after 200 h is the evaporation of the solvent
in the redox electrolyte solution. Our study indicates that
hemicyanine dyes could perform excellent spectral sensitiza-
tion in DSSCs, and that more efficient hemicyanine dyes as
sensitizers could be designed after researching the relationship
between molecular structure and functional characteristic.
2
2
as a reference electrode. Fig. 7 gives the time course of the
maximum absorbance of the two electrodes. It was found that
the absorbance of the TiO
during the first 12 hours and then decreased gradually over a
prolonged time, while that of the TiO /D/E electrode showed
almost constant during the time course (250 h). It can be con-
cluded that the thin film of dry electrolyte can prohibit the TiO
2
/D electrode decreased quickly
2
2
film from degradation of the hemicyanine dye. This supports
our belief that the main reason for the decreasing performance
of the dye 3 sensitized solar cell was due to the evaporation of
the solvent in the redox electrolyte solution.
Acknowledgements
This work was financially supported by the State Key Program
of Fundamental Research (G1998061308), and NHTRDP of
P. R. China (2002 AA 324030 and 2002 AA 324080).
Molecular structure and functional characteristic
It is well known that the N, N-diethylanilino group is a stron-
ger electron donor than the N, N-dimethylanilino group.
However the experimental data show that dye 1, bearing the
N,N-dimethylanilino group as electron donor, has a slightly
higher overall photoelectric conversion yield than dye 2. Since
the short-circuit photocurrent of a dye sensitized solar cell is
proportional to the amount of adsorbed dye molecules under
certain conditions, so combining the value of short-circuit
photocurrent and the number of adsorbed dye molecules, we
can clarify the contribution of one dye molecule to the
generation of the short-circuit photocurrent (ref. Table 3).
References
1
2
3
A. Hagfeld and M. Gr a¨ tzel, Chem. Rev., 1995, 95, 49.
B. O’Regan and M. Gr a¨ tzel, Nature, 1991, 353, 737.
M. K. Nazeeruddin, A. Kay, I. Rodicio, R. Humphry-Baker, E.
Muller, P. Liska, N. Vlachopoulos and M. Gr a¨ tzel, J. Am. Chem.
Soc., 1993, 115, 6382.
4
5
M. K. Nazeeruddin, P. P e´ chy and M. Gr a¨ tzel, Chem. Commun.,
1
997, 1705.
M. K. Nazeeruddin, P. P e´ chy, T. Renouard, S. M. Zakeeruddin,
R. Humphry-Baker, P. Comte, P. Liska, L. Cevery, E. Costa,
V. Shllover, L. Spiccia, G. B. Deacon, C. A. Bignozzi and
M. Gr a¨ tzel, J. Am. Chem. Soc., 2001, 123, 1613.
1
7
Referring the number of adsorbed molecules 2.3 ꢂ 10 mole-
ꢀ
2
cules cm and the short-circuit photocurrent 10.7 mA cm ,
ꢀ2
the contribution per dye 1 molecule to the short-circuit photo-
ꢀ17
b
6
T. Renouard, R. A. Fallahpour, M. K. Nazeeruddin, R.
Humphry, S. I. Gorelsky, A. B. P. Lever and M. Gr a¨ tzel, Inorg.
Chem., 2002, 41, 367.
current (I ) is 4.6 ꢂ 10
mA. Similarly, those for dyes 2 and
mA and 3.5 ꢂ 10
sc
ꢀ17
ꢀ16
mA, respectively (Table
). Data show that the contribution of one dye 2 molecule to
3
are 7.4 ꢂ 10
3
7
8
9
C. Nasr, D. Liu, S. Hotchandani and P. V. Kamat, J. Phys. Chem.
B, 1996, 100, 11 054.
S. Ferrere, A. Zaban and B. A. Gregg, J. Phys. Chem. B, 1997,
the short-circuit photocurrent is 60% larger than that of dye
, showing that replacing of the methyl group by an ethyl
1
1
01, 4490.
group is favorable for photoelectric generation. This fact
agrees with the general rule. What is more interesting is the
finding that the short-circuit photocurrent generated by one
dye 3 molecule is 4.7 times larger than that of one dye 2, reveal-
ing that the introduction of the additional hydroxyl group to
the donor is favorable for the electron injection efficiency.
N. J. Cherepy, G. P. Smestad, M. Gr a¨ tzel and J. Z. Zhang,
J. Phys. Chem. B, 1997, 101, 9342.
1
1
0
1
T. N. Rao and L. Bahadur, J. Electrochem. Soc., 1997, 144, 179.
E. Stathatos, P. Lianos and A. Laschewsky, Langmuir, 1997,
13, 259.
12 E. Stathatos and P. Lianos, Chem. Mater., 2001, 13, 3888.
1
1
3
4
A. Kay and M. Gratzel, J. Phys. Chem., 1993, 97, 6272.
A. C. Khazraji, S. Hotchandani, S. Das and P. V. Kamat, J. Phys.
Chem. B, 1999, 103, 4693.
1
5
6
K. Sayama, M. Sugino, H. Sugihihara, Y. Abe and H. Arakawa,
Chem. Lett., 1998, 753.
K. Sayama, K. Hara, N. Mori, M. Satsuki, S. Suga, S.
Tsukagoshi, Y. Abe, H. Sugihara and H. Arakawa, Chem. Com-
mun., 2000, 1173.
1
17
18
19
20
21
K. Hara, K. Sayama, Y. Ohga, A. Shinpo, S. Suga and H.
Arakawa, Chem. Commun., 2001, 569.
K. Sayama, S. Tsukagoshi, K. Hara, Y. Ohga, A. Shinpou, Y.
Abe, S. Suga and H. Arakawa, J. Phys. Chem. B, 2002, 106, 1363.
Z.-S. Wang, F.-Y. Li and C.-H. Huang, Chem. Commun., 2000,
2
063.
Z.-S. Wang, F.-Y. Li, C.-H. Huang, L. Wang, M. Wei, L.-P. Jin
and N.-Q. Li, J. Phys. Chem. B, 2000, 104, 9676.
Z.-S. Wang, F.-Y. Li and C.-H. Huang, J. Phys. Chem. B, 2001,
1
05, 9210.
2
2
2
3
W.-S. Xia, C.-H. Huang and D.-J. Zhou, Langmuir, 1997, 13, 80.
T.-R. Cheng, C.-H. Huang and L.-B. Gan, J. Mater. Chem., 1997,
7, 631.
2
Fig. 7 Stability test of dye 3 on TiO electrode.
1
282
New J. Chem., 2003, 27, 1277–1283

View Article Online
24
25
26
F.-Y. Li, J. Zheng, C.-H. Huang, L.-P. Jin, J.-Y. Zhuang, J.-Q.
Guo and X.-S. Zhao, J. Phys. Chem. B, 2000, 104, 5090.
D.-G. Wu, C.-H. Huang, L.-B. Gan, W. Zhang and J. Zheng,
J. Phys. Chem. B, 1999, 103, 4377.
D.-G. Wu, C.-H. Huang, Y.-Y. Huang, L.-B. Gan, A.-Q. Yu,
L.-M. Ying and X.-S. Zhao, J. Phys. Chem. B, 1999, 103,
29 Y.-X. Weng, Y. Li, Y. Liu, L. Wang, G.-Z. Yang and J.-Q. Sheng,
Chem. Phys. Lett., 2002, 355, 294.
30 N. Rabjohn, R. T. Arnold and N. Leonard, J. Org. Synth. Coll.,
1963, 4, 331.
31 A. Hassner, D. Birnbaum and L. M. Loew, J. Org. Chem., 1984,
49, 2546.
7
130.
32 C. J. Barb e´ , F. Arendse, P. Comte, M. Jirousek, F. Lenzmann,
V. Shklover and M. Gr a¨ tzel, J. Am. Ceram. Soc., 1997, 80,
3157.
33 G. P. Smestad and M. Gr a¨ tzel, J. Chem. Educ., 1998, 75, 752.
34 A. Mishra, R. K. Behera, P. K. Behera, B. K. Mishra and G. B.
Behera, Chem. Rev., 2000, 100, 1973.
2
7
8
C.-H. Huang, F.-Y. Li and Y.-Y. Huang, Ultrathin Films for
Optics and Electronics, Peking University Press, Beijing, 2001,
and some references therein.
A.-D. Lang, J. Zhai, C.-H. Huang, L.-B. Gan, Y.-L. Zhao, D.-J.
Zhou and Z.-D. Chen, J. Phys. Chem. B, 1998, 102, 1424.
2
New J. Chem., 2003, 27, 1277–1283
1283