Photophysical, electrochemical, and photoelectrochemical properties of new azulene-based dye molecules

Article abstract of DOI:10.1039/b613703b

Four new azulene-based dye molecules, 3-(azulen-1-yl)-2-cyanoacrylic acid (Azu-1), 3-(7-isopropyl-1,4-dimethylazulen-3-yl)-2-cyanoacrylic acid (Guai-1), 5-(azulen-1-yl)-2-cyanopenta-2,4-dienoic acid (Azu-2), and 5-(7-isopropyl-1,4- dimethylazulen-3-yl)-2-cyanopenta-2,4-dienoic acid (Guai-2), were synthesized and their photoelectrochemical properties were studied in dye-sensitized solar cells (DSSCs). All of them exhibit, in the visible region, a strong absorption band coming from the S₀-S₂ transition and a very weak band coming from the S₀-S₁ transition, and the transition assignments are supported by theoretical calculations using time-dependent density functional theory (TD-DFT) at the B3LYP/6-31G* level. In sensitization of nanocrystalline TiO₂ electrodes, reducing their adsorption amount on the TiO₂ surface (by co-adsorption with deoxycholic acid) mitigates dye aggregation and improves their photoelectric conversion efficiency greatly. Also, extending the conjugated side chain (Azu-2 vs. Azu-1 or Guai-2 vs. Guai-1) not only shifts their photoelectric response to longer wavelengths and therefore enhances the short-circuit photocurrent, but also increases the open-circuit photovoltage significantly. Moreover, it was found that the electron injection efficiencies varied remarkably with excitation wavelength, suggesting direct electron injection from the S₂ state of these dye molecules. The Royal Society of Chemistry.

Full text of DOI:10.1039/b613703b

View Article Online / Journal Homepage / Table of Contents for this issue
PAPER
www.rsc.org/materials | Journal of Materials Chemistry
Photophysical, electrochemical, and photoelectrochemical properties of new
azulene-based dye molecules{
ab
Xue-Hua Zhang, Chao Li, Wei-Bo Wang, Xue-Xin Cheng,* Xue-Song Wang* and Bao-Wen Zhang*
a
a
a
a
a
Received 20th September 2006, Accepted 13th November 2006
First published as an Advance Article on the web 1st December 2006
DOI: 10.1039/b613703b
Four new azulene-based dye molecules, 3-(azulen-1-yl)-2-cyanoacrylic acid (Azu-1), 3-(7-
isopropyl-1,4-dimethylazulen-3-yl)-2-cyanoacrylic acid (Guai-1), 5-(azulen-1-yl)-2-cyanopenta-
2
,4-dienoic acid (Azu-2), and 5-(7-isopropyl-1,4-dimethylazulen-3-yl)-2-cyanopenta-2,4-dienoic
acid (Guai-2), were synthesized and their photoelectrochemical properties were studied in dye-
sensitized solar cells (DSSCs). All of them exhibit, in the visible region, a strong absorption band
0 2 0 1
coming from the S –S transition and a very weak band coming from the S –S transition, and the
transition assignments are supported by theoretical calculations using time-dependent density
functional theory (TD-DFT) at the B3LYP/6–31G* level. In sensitization of nanocrystalline TiO
2
electrodes, reducing their adsorption amount on the TiO
2
surface (by co-adsorption with
deoxycholic acid) mitigates dye aggregation and improves their photoelectric conversion
efficiency greatly. Also, extending the conjugated side chain (Azu-2 vs. Azu-1 or Guai-2 vs. Guai-1)
not only shifts their photoelectric response to longer wavelengths and therefore enhances the
short-circuit photocurrent, but also increases the open-circuit photovoltage significantly.
Moreover, it was found that the electron injection efficiencies varied remarkably with excitation
2
wavelength, suggesting direct electron injection from the S state of these dye molecules.
isomeric alternant aromatic molecule, azulene shows a charac-
teristic blue color, and its color can be varied remarkably by
simple substitution, depending on the substituent nature (e.g.
1
Introduction
As a novel renewable and clean solar-to-electricity conversion
system, dye-sensitized solar cells (DSSCs) based on meso-
porous TiO2 electrodes have attracted extensive interest
alkylated, fluorinated or formylated) and substitution posi-
25,26
tion.
Once it loses an electron, azulene will become a
1
following their discovery in 1991. While ruthenium poly-
resonance-stabilized azulenylium carbocation, a 6p-electron
27,28
pyridyl complexes, such as cis-di(thiocyanato)bis(4,49-dicar-
boxy-2,29-bipyridine)ruthenium(II) (N3) and tri(thiocyanato)
aromatic tropylium carbocation analog (Scheme 1).
It is
estimated that the azulenylium carbocation is about
21
(
4,49,40-tricarboxy-2,29:69,20-terpyridine)ruthenium(II) (Black
2
1
0 kcal mol
more stable than azulene itself, accounting
29,30
–6
dye),
still represent the most efficient sensitizers so far,
for the electron-donating feature of azulene.
The unique
metal-free organic chromophores are increasingly applied in
DSSCs by virtue of their high molar extinction coefficient, easy
preparation, low cost, as well as their rich structures. The rich
structures lead to diverse photophysical, photochemical, and
electrochemical properties, which stimulate the continuous
exploration of new types of organic dye molecules for
DSSC application. Over the past decades, many kinds of
electron-donating property of azulene has been fully utilized
to construct highly efficient nonlinear optical (NLO) mate-
31–33
34–37
rials,
conducting polymers,
charge-transfer com-
38,39
40
plexes,
bacteriorhodopsin pigments.
that 1-carboxyazulene can inject an electron in its excited state
into the conduction band of TiO within the fs time scale,
implying its great potential to sensitize TiO nanocrystalline
molecular
switches,
41–43
and
red-shifted
Recently, it was reported
2
7
,8
9
organic dyes, including cyanines, hemicyanines, triphenyl-
2
1
0,11
12,13
14–16
17–20
methanes,
perylenes,
coumarins,
porphyrins,
have been utilized in
DSSCs, and some of them achieved respectable photoelectric
44
electrodes.
-carboxyazulene is poor, as the result of its weak absorptivity
for the S –S transition and high energy for the S –S
2
However, the light-harvesting efficiency of
2
1,22
23,24
squaraines,
and indoline dyes,
1
0
1
0
1
6,24
conversion efficiencies.
transition (below 400 nm). This shortcoming may be overcome
by insertion of methine units between the azulene chromo-
phore and the anchoring carboxyl group. It was found that the
Azulene is well known as a non-alternant, non-benzenoid
aromatic hydrocarbon. Compared with naphthalene, its
2
position of the S band is highly sensitive to the conjugated
a
side chain appended on C-1 of azulene, and undergoes a
Technical Institute of Physics and Chemistry, Chinese Academy of
Sciences, Beijing, 100080, P. R. China. E-mail: g203@mail.ipc.ac.cn;
Fax: +86-10-62554670; Tel: +86-10-82543592
Graduate School of Chinese Academy of Sciences, Beijing, 100039,
b
P. R. China
{
Electronic supplementary information (ESI) available: optimized
ground-state geometries and dihedral angles from the optimized
structures of the dyes. See DOI: 10.1039/b613703b
Scheme 1 Schematic representation of the radical cation of azulene.
This journal is ß The Royal Society of Chemistry 2007
6
42 | J. Mater. Chem., 2007, 17, 642–649

View Article Online
Scheme 2 Molecular structures of Azu-1, Azu-2, Guai-1, and Guai-2.
4
1,45
bathochromic shift with increasing conjugation length.
1-formylazulene (2a) and 1-formylguaiazulene (2b), respec-
tively. Then, the condensation of the corresponding aldehyde
with cyanoacetic acid in acetonitrile in the presence of a
catalytic amount of piperidine, followed by recrystallization
of the formed precipitate from ethanol, gave pure Azu-1 and
Guai-1 in yields of 73% and 69%, respectively (Scheme 3).
Moreover, the longer conjugated side chain will move the
azulenylium carbocation far away from the TiO surface, and
2
therefore may restrict the recombination of the injected
electron with the azulenylium carbocation. In this regard,
four new azulene derivatives, Azu-1, Azu-2, Guai-1, and Guai-2
1
(
Scheme 2), were designed and synthesized for DSSC applica-
Azu-1: H NMR (400 MHz, DMSO) d 7.71 (d, 1H, J =
tion, and their sensitizing abilities were discussed based on
their structures. To the best of our knowledge, this is the first
report on the photoelectrochemical behavior of azulene-based
sensitizers in DSSCs.
4.72 Hz), 7.75 (d, 1H, J = 10.54 Hz), 7.79 (s, 1H), 8.09 (t, 1H,
J = 9.9 Hz), 8.73 (d, 1H, J = 9.46 Hz), 8.81 (s, 1H), 8.86 (d, 1H,
J = 4.47 Hz), 9.01 (d, 1H, J = 9.81 Hz), 13.34 (s, 1H). EI-MS:
+
m/z 223 (M ). HR-MS: 223.0635 (Calc. Mass: 223.0633).
1
Guai-1: H NMR (300 MHz, DMSO) d 1.31 (d, 6H, J =
2
Experimental
6.84 Hz), 2.56 (s, 3H), 2.99 (s, 3H), 3.07 (m, 1H), 7.28 (d, 1H,
J = 10.77 Hz), 7.60 (d, 1H, J = 10.72 Hz), 8.21 (s, 1H), 8.38 (s,
+
H), 8.86 (s, 1H), 12.86 (br s, 1H). EI-MS: m/z 293 (M ). HR-
2
.1 General measurements, materials, and dye synthesis
1
1
MS: 293.1417 (Calc. Mass: 293.1416).
H NMR spectra were recorded on a Varian GEMINI 300 or
on a Bruker AM400 spectrometer. Mass spectra were taken
on a Shimadzu GCMS-QP2010 spectrometer and a GCT-MS
Micromass UK spectrometer. UV-vis spectra were run on
a Shimadzu UV-1601PC spectrophotometer. Fluorescence
spectra were measured on a Hitachi F-4500 fluorescence
spectrophotometer.
Syntheses of 5-(azulen-1-yl)-2-cyanopenta-2,4-dienoic acid
(Azu-2) and 5-(7-isopropyl-1,4-dimethylazulen-3-yl)-2-cyano-
penta-2,4-dienoic acid (Guai-2) were as follows. The Wittig
reaction of 1-formyl-azulene (2a) or 1-formyl-guaiazulene
(2b) with ((1,3-dioxolan-2-yl)methyl)-tri(n-butyl)phosphonium
4
8
bromide following the reported method gave 3-(azulen-1-
yl)acrylaldehyde (3a) or 3-(7-isopropyl-1,4-dimethylazulen-3-
yl)acrylaldehyde (3b), from which Azu-2 and Guai-2 were
prepared by condensation with cyanoacetic acid (Scheme 3) in
yields of 68% and 63%, respectively.
Azulene, guaiazulene, trimethyl orthoformate, boron tri-
fluoride–diethyl etherate, cyanoacetic acid, piperidine, TiCl4,
LiI, and I2 from Aldrich, tri-n-butylphosphine from TCI,
2
-bromomethyl-1,3-dioxolane and deoxycholic acid (DCA)
1
from Acrose, TiO2 paste (Nanoxide-T) and 1,2-dimethyl-3-
propylimidazolium iodide (DMPImI) from Solaronix,
Switzerland, were used as received. All solvents used were
Azu-2: H NMR (400 MHz, DMSO) d 7.10 (m, 1H), 7.34
(m, 2H), 7.51 (d, 1H, J = 4.28 Hz), 7.76 (t, 1H, J = 9.83 Hz),
7.82 (d, 1H, J = 11.86 Hz), 7.96 (d, 1H, J = 14.93 Hz), 8.34 (d,
1H, J = 4.26 Hz), 8.42 (d, 1H, J = 9.30 Hz), 8.74 (d, 1H, J =
4
6
analytical grade and purified following general procedures.
Syntheses of 3-(azulen-1-yl)-2-cyanoacrylic acid (Azu-1) and
-(7-isopropyl-1,4-dimethylazulen-3-yl)-2-cyanoacrylic acid
Guai-1) were performed as follows. The formylation of
azulene (1a) and guaiazulene (1b) with trimethyl orthoformate
+
9.89 Hz), 12.75 (br s, 1H). EI-MS: m/z 249 (M ). HR-MS:
3
249.0792 (Calc. Mass: 249.0790).
1
Guai-2: H NMR (300 MHz, DMSO) d 1.29 (d, 6H, J =
(
6.60 Hz), 2.55 (s, 3H), 2.97 (s, 3H), 3.07 (m, 1H), 6.88 (t,
4
7
in the presence of boron trifluoride–diethyl etherate gave
1H, J = 12.9 Hz), 7.07 (d, 1H, J = 10.77 Hz), 7.44 (d, 1H,
Scheme 3 Synthesis routes to Azu-1, Azu-2, Guai-1, and Guai-2.
This journal is ß The Royal Society of Chemistry 2007
J. Mater. Chem., 2007, 17, 642–649 | 643

View Article Online
J = 10.65 Hz), 7.70 (d, 1H, J = 11.55 Hz), 8.01 (m, 3H) 12.98
+
br s, 1H). EI-MS: m/z 319 (M ). HR-MS: 319.1615 (Calc.
S370 single channel optometer equipped with a Model 262 flat
response sensor head (Graseby Optronics).
(
Mass: 319.1617).
2
.5 Molecular orbital calculation
2
.2 Preparation of the dye-sensitized nanocrystalline TiO
2
All calculations were made on the Gaussian 03 program
electrodes
4
9
package by using time-dependent density functional theory
(TD-DFT). B3LYP along with the 6-31G(d) basis set was
used. Before optimizing ground-state geometries by B3LYP/6–
31G*, the molecular structures were initially optimized in
semiempirical PM3 calculations as recommended in the
Gaussian 03 program package. All the geometries and
electronic properties were calculated by assuming the target
molecules to be isolated molecules in the gas phase.
The TiO
2
paste was coated on a conducting fluorine-doped tin
oxide (FTO) glass (transmission .85%, sheet resistivity of
2
0 V square ) by the doctor-blade method with Scotch tape
1
2
as the spacer. The thickness of the TiO film was controlled by
2
the number of Scotch tape layers. After air drying at room
temperature for 30 min, the film was heated to 450 uC at a rate
2
1
of 20 uC min and maintained at 450 uC for 30 min. Then the
sintered TiO film was treated with 0.05 M TiCl for 30 min at
0 uC, followed by heating at 450 uC for another 30 min. The
2
4
4
7
3
Results and discussion
electrode was then immersed in a solution of 0.5 mM dye in
DMSO (with or without 1 mM DCA) while it was still hot
3
.1 Photophysical properties
(about 80 uC) and soaked for at least 12 h at room tem-
perature. After that, the electrode was rinsed thoroughly with
Fig. 1 shows the electronic absorption spectra of Azu-1, Azu-2,
Guai-1, and Guai-2 both in DMSO solutions and adsorbed on
ethanol and dried under a dry air stream. The thus-obtained
nanocrystalline TiO
maxima, extinction coefficients at absorption maxima, and
adsorption amount on TiO electrodes are collected in Table 1.
2
electrodes. The corresponding absorption
TiO
Profilometer. The amount of the dye adsorbed on the TiO
electrode was estimated spectroscopically after the dye was
desorbed in a 1.0 mM solution of NaOH in methanol–H
50 : 50 volume ratio).
2
film had a thickness of 12 mm, determined by a Dektak 3
2
2
The absorption spectra of the four dye molecules in DMSO
2
O
exhibit similar profiles, i.e. a strong absorption band within
(
3
6
75–550 nm and a very weak absorption tail extending beyond
50 nm. The weak absorption band can be attributed to the
2
.3 Electrochemical measurements
S
0
–S
1
transition of these azulene-based dye molecules, while
–S
the strong absorption band can be ascribed to the S
0
2
Cyclic voltammetry measurements were carried out on a
Potentiostat/Galvanostat model 283A (EG&G Princeton
Applied Research) to determine the oxidation potentials of
the dyes in DMF solutions. The measurement details are as
follows: Pt working and counter electrodes, saturated calomel
electrode (SCE) as reference electrode, 0.1 M of tetrabutyl-
transition. Such assignments are supported by the theoretical
calculations (see below). Extension of the conjugated side
chain (Azu-2 vs. Azu-1 or Guai-2 vs. Guai-1) results in a
significant red shift of the absorption maxima, making them
more efficient in harvesting blue and green light. The alkyl
substitutions on the azulene ring (Guai-1 vs. Azu-1 or Guai-2
vs. Azu-2) also make a noticeable contribution to the spectrum
red shift, a typical characteristic of azulene chromophores that
the absorption is very sensitive to ring substitution. Besides red
shifting the absorption spectrum, the alkyl groups present in
Guai-1 and Guai-2 may play an important role in restricting
the dye aggregation. As shown in Fig. 2, Azu-1 undergoes a
remarkable red shift of the absorption maximum, broadening
of the absorption band, as well as variation of the absorption
line shape with increasing concentration, implying strong
intermolecular interactions. In sharp contrast, the concentra-
tion effect on the absorption spectrum of Guai-1is greatly
diminished. Similar behavior was also found for Azu-2 and
Guai-2. Dye aggregation often opens up new deactivation
pathways for the excited state, which may therefore interfere
with the desired process, such as electron injection at the
dye/semiconductor interface. The better sensitizing ability of
guaiazulene-based dye molecules in DSSCs may partly
originate from their restricted aggregation (see below).
ammonium perchlorate ((C
4
H
9
)
4
NClO
4
) as supporting elec-
2
1
trolyte, scan rate of 100 mV s , and temperature of 20 uC.
2
.4 Photovoltaic measurements
The sandwich-type solar cell consisted of a dye-sensitized TiO
2
film as working electrode, a Pt-coated FTO conducting glass
as counter electrode and a redox electrolyte solution which was
composed of 0.1 M LiI, 0.05 M I
2
, and 0.6 M DMPImI in
film was illumi-
methoxyacetonitrile. The dye-sensitized TiO
2
nated through the conducting glass support and the effective
2
irradiated area was 0.2 cm . The current density–voltage (J–V)
curves were obtained from a computer controlled Keithley
2
400 source meter under illumination by AM 1.5 Global
2
2
simulated light (75 mW cm ) from an Oriel 91192 Solar
Simulator.
In the measurements of the action spectra of the incident
monochromatic photon-to-current conversion efficiencies
(
IPCEs), the solar cell was irradiated by monochromatic
light in the range of 400–800 nm at 10 nm intervals. The
monochromatic light was obtained from a 500 W Xe lamp in
combination with a SpectraPro-150 monochromator (Acton
Research Co.). The generated short-circuit photocurrents were
recorded on a model 283 A Potentiostat/Galvanostst. The
incident monochromatic light intensity was measured with a
Upon adsorption onto the TiO nanocrystalline electrodes,
2
the absorption spectra of the four dye molecules were all
extended greatly to longer wavelengths (Fig. 1b), probably
resulting from the interactions between dye molecules
and TiO2 and/or among the dye molecules. Such spectrum
broadening allows the dye molecules to harvest visible light
6
44 | J. Mater. Chem., 2007, 17, 642–649
This journal is ß The Royal Society of Chemistry 2007

View Article Online
2
6
2
Fig. 1 Absorption spectra of Azu-1, Azu-2, Guai-1, and Guai-2 in DMSO solutions (a, 1.5 6 10 M) and on nanocrystalline TiO films
(b, obtained by soaking the film (12 mm) in a DMSO solution containing 0.5 mM dye and 1.0 mM DCA).
Table 1 Photophysical and electrochemical properties as well as the
adsorption amount on nanocrystalline TiO films (12 mm) for Azu-1,
2
Azu-2, Guai-1, and Guai-2
molecules were adsorbed alone or adsorbed simultaneously
with DCA (deoxycholic acid). It seems that there is a good
correlation between the dye adsorption amount and dye
molecular size, i.e. the bigger the dye molecules, the smaller
the adsorption amount. Furthermore, the alkyl groups
substituted on the azulene ring reduce the adsorption amount
more effectively, and consequently weaken the dye aggregation
more efficiently, which may account for the higher photo-
electric conversion efficiencies of Guai-1 and Guai-2 with
respect to Azu-1 and Azu-2, respectively (see Table 1).
Azu-1
Azu-2
Guai-1
Guai-2
l
max (DMSO)/nm
402.0
7.8
0.95
20.95
2.7
429.5
6.9
0.85
21.05
2.6
423.5
5.3
0.91
20.99
2.0
453.0
5.8
0.81
21.09
1.9
4
21
e (DMSO)/10 M cm
21 a
b
E
ox/V (vs. SCE)
ox*/V (vs. SCE)
c
E
N/10 mol cm
N/10 mol cm
27
22 d
22 e
27
4.1
3.7
2.4
2.1
a
b
Molar extinction coefficient at lmax
.
Ground state oxidation peak
potential in DMF. Excited state oxidation potential calculated by
subtracting the 0–0 transition energy from ox (see the text).
c
One special property of azulene is its fluorescence emission
E
from the S
rather than the S
Rule results from the unusually large S
the presence of a chemical channel for rapid relaxation of the
state. This violation of Kasha’s
2
1
d
e
Adsorption amount under co-adsorption with DCA. Adsorption
amount without DCA co-adsorption.
50
1
–S energy gap and
2
5
1
1
S state. However, no fluorescence emission was observed for
more efficiently, particular for Azu-2 and Guai-2 based TiO2
electrodes.
Azu-1, Azu-2, Guai-1, and Guai-2. The absence of fluorescence
may be the result of the red shift of the S band, which reduces
2
The dye adsorption amount decreased in the order Azu-1 .
Azu-2 . Guai-1 . Guai-2, no matter whether the dye
the S –S gap and accordingly enhances internal conversion
1
2
from S
2
1
to the nonemissive S state.
2
6
26
Fig. 2 Normalized absorption spectra of DMSO solutions of Azu-1 (a) and Guai-1 (b) at varied concentrations: a, 1.5 6 10 ; b, 2.5 6 10
c, 5.0 6 10 ; d, 1.0 6 10 ; e, 2.0 6 10 M.
;
2
6
25
25
This journal is ß The Royal Society of Chemistry 2007
J. Mater. Chem., 2007, 17, 642–649 | 645

View Article Online
Fig. 3 Electron distribution in the HOMOs and LUMOs of the optimized ground states of Azu-1, Azu-2, Guai-1, and Guai-2.
3
.2 Electrochemical properties
the methyl group at C-4 of the azulene ring (for Guai-1) makes
the vinyl group have some torsion vs. the azulene plane.
Fig. 3 shows the electron distribution of the HOMOs and
LUMOs for these dye molecules. In the cases of Azu-1, Azu-2
or Guai-1, the HOMO electron density is mainly located in the
five-membered ring of azulene and the conjugated side chain.
The LUMO electron density, however, is mainly confined in
the seven-membered ring of azulene. This phenomenon is very
similar to the calculated result for azulene itself, in which there
is a substantial charge transfer from the five-membered ring to
The oxidation potentials of the four dye molecules were
measured with cyclic voltammetry (Table 1). It is clear that
both alkylation on the azulene ring and the extension of the
conjugated side chain shift the oxidation potential cathodi-
cally. The excited state oxidation potential (Eox*) of a dye
molecule can be estimated by subtracting the 0–0 transition
energies (E0-0) from the ground state oxidation potential (Eox).
1
The low extinction coefficient for the S band limits the
accurate estimation of E0-0, so a E0-0 of 1.9 eV (corresponding
2
6
the seven-membered ring upon HOMO–LUMO transition.
to a transition at 650 nm) was used for the four dyes to
For Guai-2, the electron density within the five-membered ring
moves both to the conjugated side chain and the seven-
membered ring upon HOMO–LUMO transition.
1
calculate the oxidation potentials of their S states (Table 1).
Obviously, all the S1 state potentials are far more negative
than the band edge energy of the nanocrystalline TiO2
electrode (20.8 V vs. SCE), suggesting that the electron
injection from the S1 state to the TiO2 conduction band is
Table 2 presents the calculated absorption maxima in the
visible region, their oscillator strengths, and their compositions
for the four dye molecules. Similar to the experimental data
energetically permitted. Due to the higher E0-0 for the S
state, a larger driving force and higher efficiency may be
2
(
Fig. 1 and Table 1), the calculation also produces a strong
absorbing band at higher energy and a very weak band at
lower energy. Moreover, the calculated absorption maxima
for the second excited states are in good agreement with the
measured ones, supporting the band assignments in Fig. 1(a).
envisioned for electron injection from the S
discussion below).
2
state (see the
3
.3 Quantum-chemical calculation
To gain insight into the electronic structures of these new
azulene-based dyes at the molecular level, we performed TD-
DFT (time-dependent density functional theory) calculations
3
.4 Photovoltaic performance
A series of DSSCs were prepared and compared to get insight
into the relations between the sensitizing behavior of the four
dye molecules and their structures. Fig. 4 shows the incident
monochromatic photo-to-electron conversion efficiencies
(IPCE) as a function of the wavelength of the incident light,
which is defined as the number of electrons generated by light
in the external circuit divided by the number of incident
4
9
at the B3LYP/6-31G* level with Gaussian 03, revision B.03.
The optimized geometries and atom labeling are shown in ESI
Fig. 1S,{ while the main dihedral angles, by which the
optimized geometries of the ground state can be defined, are
tabulated in Table 1S.{ Largely, the four dye molecules adopt
almost planar conformations in their energy minima, though
Table 2 Calculated energies (lmax), oscillator strengths (f), and compositions of the transitions in the visible region for Azu-1, Azu-2, Guai-1, and
Guai-2
Compound
Azu-1
Azu-2
Guai-1
Guai-2
a
Excited state
l
max
f
Composition
1
2
1
2
1
2
1
2
535
392
587
422
562
423
611
448
0.003
0.414
0.005
1.031
0.006
0.466
0.007
1.081
20.110x(57A60) + 0.681x(58A59)
0.261x(57A59) + 0.576x(58A60)
20.105x(64A66) + 0.568x(65A66) + 0.370x(65A67)
20.104x(64A66) 2 0.328x(65A66) + 0.481x(65A67)
0.632x(78A79) + 0.252x(78A80)
20.174x(77A79) 2 0.177x(77A80) 2 0.217x(64A66) + 0.544x(78A80)
0.204x(85A86) + 0.676x(85A87)
0.178x(84A87) + 0.557x(85A86) 2 0.166x(85A87)
Orbital 58 is the HOMO and 59 the LUMO of Azu-1, orbital 65 is the HOMO and 66 the LUMO for Azu-2, orbital 78 is the HOMO and
9 the LUMO for Guai-1, orbital 85 is the HOMO and 86 is the LUMO for Guai-2.
7
6
46 | J. Mater. Chem., 2007, 17, 642–649
This journal is ß The Royal Society of Chemistry 2007

View Article Online
Fig. 5 Photocurrent–photovoltage curves of the DSSCs based on
Azu-1, Azu-2, Guai-1, and Guai-2.
Fig. 4 Photocurrent action spectra of the DSSCs based on Azu-1,
Azu-2, Guai-1, and Guai-2.
2
photons and is calculated using the following equation, where
also found in Azu-1, Azu-2 and Guai-1 sensitized solar cells.
The higher electron injection efficiency from the S
result from the larger driving force and/or longer S
2
state may
lifetime. It
Jsc is the short-circuit photocurrent density, l is the excitation
2
wavelength, and Pin is the incident light intensity.
will be useful to take measurements on the lifetimes of the S1
and S states, the electron injection rate constants, as well as
ꢀ
ꢁ
ꢁ
{
2
2
1
240Jsc mA cm
2
IPCE ð%Þ~
ꢀ
{
the recombination rates for these azulene-based dye molecules,
though the features of our time-resolved spectrum techniques
(absorption and emission at ns or longer time regimes) limit us
presently.
lðnmÞPin W m
The four dye molecules exhibit photoelectric activities over
their whole absorption bands (Fig. 1(b)), indicative of the
participation of the S
the electron injection process. It is well known that the S
of azulene and its derivatives has an ultrafast deactivation
1
state (and/or even the dye aggregate) in
Once the four dye molecules are compared to each other
(Fig. 4), it immediately appears that Azu-2 and Guai-2 possess
better sensitizing capabilities than Azu-1 and Guai-1, respec-
tively, mainly due to their red shifted absorption (Fig. 1).
Fig. 5 presents the photocurrent–photovoltage (J–V) curves
of the corresponding solar cells under AM1.5G irradiation
1
state
5
2
pathway, which in turn indicates that the electron injection
rates for the four dyes should be fast enough to compete
effectively with their intrinsic decay from the S
1
state.
2
2
Another point worthy of concern is that whether the
(75 mW cm ), from which the short-circuit photocurrent
electron injection can directly occurs in the S
dye molecules, though no fluorescence from S
2
state of these
was observed
density (J ), the open-circuit photovoltage (V ), the fill factor
sc oc
2
(FF), and the overall photoelectric conversion efficiency (g)
can be obtained (Table 3). The relations between these photo-
voltaic parameters are as follows, where Pin is the intensity of
the incident light and Pmax the power output maximum of the
solar cells.
for them. With this in mind, we examined the electron injection
efficiencies (W ) of the dye molecules at different wavelengths.
in
The IPCE at a wavelength is the production of three
efficiencies, i.e. IPCE = LHE(l) 6 Win 6 W , in which
c
LHE(l) is the light harvesting efficiency at a wavelength of l
JscVocFF
Pin
Pmax
and W
back contact of the photoanode. W
electron transport within the TiO network, (2) the recombina-
tion of the injected electrons with the electrolyte (I
c
is the collection efficiency of the injected electron at the
g~
,
FF~
JscVoc
c
is associated with (1) the
2
2
3
), and (3)
Table 3 Photovoltaic performances of the DSSCs based on Azu-1,
Azu-2, Guai-1, and Guai-2
the recombination of the injected electrons with the oxidized
dyes, and therefore should be a term independent of the
excitation wavelength. Thus, IPCE/LHE(l) can be used to
compare Win at different wavelengths, while LHE(l) can be
calculated based on the absorption spectrum of the dye-
sensitized electrode (Fig. 1(b)). Taking Guai-2 as an example,
its IPCE/LHE(l) at 400 nm, 450 nm, 500 nm, 550 nm, and
a,b
Azu-1
Azu-2
Guai-1
Guai-2
22
J
V
sc/mA cm
oc/mV
6.30 (4.91)
349 (336)
57.9 (56.8)
1.7 (1.25)
8.48 (6.87)
433 (425)
57.6 (56.9)
2.8 (2.22)
5.23 (4.75)
353 (349)
57.9 (57.2)
1.4 (1.26)
7.88 (7.41)
468 (462)
58.8 (58.1)
2.9 (2.65)
22
FF (%)
g (%)
a
Irradiation light, AM 1.5 Global simulated light (75 mW cm
)
6
00 nm are respectively 0.57, 0.51, 0.34, 0.27, and 0.25. Such
wavelength dependence suggests that electron injection might
take place (at least partly) in the S state and with a higher
efficiency than in the S state (and/or dye aggregate). The
from Oriel 91192 Solar Simulator; photoelectrode, 12 mm thick TiO
2
2
film with DCA co-adsorption; irradiated area, 0.2 cm ; electrolyte,
2
0.1 M LiI, 0.05 M I , 0.6 M DMPImI in methoxyacetonitrile.
b
Values in parentheses are corresponding parameters obtained in
the DSSCs without DCA co-adsorption.
2
1
decrease of IPCE/LHE(l) with increasing wavelength was
This journal is ß The Royal Society of Chemistry 2007
J. Mater. Chem., 2007, 17, 642–649 | 647

View Article Online
In Table 3, the photovoltaic parameters both with and
without the co-adsorption of DCA are included for com-
parison. DCA was often utilized, by co-adsorption with dye
5
3
stuff, to improve the photovoltaic properties. The working
mechanisms of DCA have been well accepted to be (1)
2
blocking the access of I
3
2
to the TiO surface where not
occupied by dye molecules and therefore restricting recombi-
nation of the injected electrons, and (2) reducing the dye
5
4
adsorption amount and therefore refraining dye aggregation.
In the presence of DCA, the g was increased by 36% (Azu-1),
6% (Azu-2), 11% (Guai-1), and 9.4% (Guai-2), respectively.
2
The improved photoelectric conversion efficiencies mainly
stem from the enhanced photocurrent as evidenced by the fact
that Voc was increased by only 3.9% (Azu-1), 1.9% (Azu-2),
1
.1% (Guai-1), and 1.3% (Guai-2), FF was increased by only
.9% (Azu-1), 1.2% (Azu-2), 1.2% (Guai-1), and 1.2% (Guai-2);
1
Fig. 6 Dark currents of the DSSCs based on Azu-1, Azu-2, Guai-1,
and Guai-2.
in contrast, Jsc was increased by 28% (Azu-1), 23% (Azu-2),
1
0% (Guai-1), and 6.3% (Guai-2). Interestingly, the enhanced
Jsc was obtained in the case that less dye molecules were bound
on their chemical structures and on their mutual interactions
on the electrode (Table 1), implying the dye aggregation
occurred at higher loading impaired the electron injection
process. This is further proved by the fact that DCA promoted
the photovoltaic properties of Azu-1 (vs. Guai-1) and Azu-2
2
on the TiO surface. Reducing their adsorption amount on
TiO2 electrodes (by co-adsorption with deoxycholic acid)
mitigates dye aggregation and improves their photoelectric
conversion efficiency greatly. Extending the conjugated side
chain shows two-fold advantages: (1) increasing the short-
circuit photocurrent as the result of the red shifted absorption,
and (2) increasing the open-circuit photovoltage under a
mechanism ambiguous as yet but deserving further investiga-
tion. While the study of the important kinetic processes (such
as electron injection and recombination, lifetimes of the first
and second excited states) will aid the better understanding of
the photoelectrochemical properties of the four dye molecules,
the chemical modifications (such as the extension of the
conjugated side chain and introduction of more bulky groups
at proper positions) may lead to more efficient azulene-based
dyes for DSSC application.
(vs. Guai-2) more efficiently, because Azu-1 and Azu-2 are
more prone to aggregate than Guai-1 and Guai-2 as mentioned
above (Fig. 2).
The length of the conjugated side chain is of importance as
well in the photovoltaic performances of these dye molecules
(
Fig. 5). While the enhanced Jsc for Azu-2 or Guai-2 can be
mainly attributed to the red shifted absorption (Fig. 1) and
therefore the improved IPCE beyond 550 nm (Fig. 4), the
reason underlying the increased Voc by the extension of
conjugated side chain is not so clear. In DSSCs, Voc can be
5
5
expressed as follows:
0
1
ꢂ
ꢃ
kT
e
Iinj
@
A
Voc~
ln
P
n
ki½Aꢀi
i
Acknowledgements
where n is the concentration of electrons in TiO
2
, and the
This work was financially supported by the Ministry of Science
and Technology of China (G2000028204) and NNSFC.
summation is for all electron transfer rate constants, k
2
i
,
to acceptors, [A] . Both I
and oxidized dye can serve as
i
3
acceptors. The diminished dark currents (Fig. 6) indicate that
2
References
the recombination between the electrons in TiO and I was
2
3
1
2
B. O’Regan and M. Gr a¨ tzel, Nature, 1991, 353, 737.
M. K. Nazeeruddin, A. Kay, I. Podicio, R. Humphry-Baker,
E. Muller, P. Liska, N. Vlachopoulous and M. Gr a¨ tzel, J. Am.
Chem. Soc., 1993, 115, 6382.
retarded to some extent in the cases of Azu-2 and Guai-2. The
side chain extension might vary the arrangement and orienta-
tion manner of the dye molecules, and thus influence the access
2
of I
3
onto the TiO
2
surface. The increased Voc might also
3 M. K. Nazeeruddin, S. M. Zakeeruddin, R. Humphry-Baker,
M. Jirousek, P. Liska, N. Vlachopoulos, V. Shklover, Christian-
H. Fischer and M. Gr a¨ tzel, Inorg. Chem., 1999, 38, 6298.
come from the restricted recombination of the injected
electrons with the oxidized dye molecules, since the extended
side chain may make the formed azulenylium carbocation
4
M. K. Nazeeruddin, P. Pechy, 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 and
M. Gr a¨ tzel, J. Am. Chem. Soc., 2001, 123, 1613.
P. Wang, S. M. Zakeeruddin, J. E. Moser, M. K. Nazeeruddin,
T. Sekiguchi and M. Gr a¨ tzel, Nat. Mater., 2003, 2, 402.
M. K. Nazeeruddin, F. D. Angelis, S. Fantacci, A. Selloni,
G. Viscardi, P. Liska, S. Ito, B. Takeru and M. Gr a¨ tzel, J. Am.
Chem. Soc., 2005, 127, 16835.
2
further away from the TiO surface. If this assumption is true,
improved photovoltaic performance will be expected in the
analog with longer conjugated side chain.
5
6
4
Conclusions
7
8
K. Sayama, K. Hara, Y. Ohga, A. Shinpou, S. Suga and
H. Arakawa, New J. Chem., 2001, 25, 200.
A. Ehret, L. Stuhl and M. T. Spitler, J. Phys. Chem. B, 2001, 105,
9960.
The four new azulene-based dye molecules were, for the first
time, utilized to sensitize nanocrystalline TiO electrodes. The
photovoltaic performances are found to be dependent strongly
2
6
48 | J. Mater. Chem., 2007, 17, 642–649
This journal is ß The Royal Society of Chemistry 2007

View Article Online
9
Y. S. Chen, C. Li, Z. H. Zeng, W. B. Wang, X. S. Wang and
B. W. Zhang, J. Mater. Chem., 2005, 15, 1654.
34 F. X. Redl, O. Kothe, K. Rockl, W. Bauer and J. Daub, Macromol.
Chem. Phys., 2000, 201, 2091.
1
1
1
0 G. Ramakrishna, H. N. Ghosh, A. K. Singh, D. K. Palit and
J. P. Mittal, J. Phys. Chem. B, 2001, 105, 12786.
1 P. M. Jayaweera, A. R. Kumarasinghe and K. Tennakone,
J. Photochem. Photobiol., A, 1999, 126, 111.
2 S. Ferrere, A. Zaban and B. A. Gregg, J. Phys. Chem. B, 1997, 101,
35 M. Porsch, G. Sigl-Seifert and J. Daub, Adv. Mater., 1997, 9, 635.
36 J. Bargon, S. Mohmand and R. Waltman, J. Mol. Cryst. Liq.
Cryst., 1983, 93, 279.
37 G. Tourillon and F. Garnier, J. Electroanal. Chem., 1982, 135, 173.
38 S. Schmitt, M. Baumgarten, J. Simon and K. Hafner, Angew.
Chem., Int. Ed., 1998, 37, 1077.
4490.
1
1
3 S. Ferrere and B. A. Gregg, New J. Chem., 2002, 26, 1155.
4 K. Hara, K. Sayama, H. Arakawa, Y. Ohga, A. Shinpo and
S. Suga, Chem. Commun., 2001, 569.
39 J. E. Frey, A. M. Andrews, S. D. Combs, S. P. Edens, J. J. Puckett,
R. E. Seagle and L. A. Torreano, J. Org. Chem., 1992, 57, 6460.
40 T. Mrozek, H. Gorner and J. Daub, Chem.–Eur. J., 2001, 7, 1028.
41 A. E. Asato, X. Y. Li, D. Mead, G. M. L. Patterson and
R. S. H. Liu, J. Am. Chem. Soc., 1990, 112, 7398.
15 K. Hara, Y. Tachibana, Y. Ohga, A. Shinpo, S. Suga, K. Sayama,
H. Sugihara and H. Arakawa, Sol. Energy Mater. Sol. Cell, 2003,
7
7, 89.
42 R. S. H. Liu, C. W. Liu, X. Y. Li and A. E. Asato, Photochem.
Photobiol., 1991, 54, 625.
43 R. S. Muthyala, M. Alam and R. S. H. Liu, Tetrahedron Lett.,
1998, 39, 5.
44 C. Pagba, G. Zordan, E. Galoppini, E. L. Piatnitski, S. Hore,
K. Deshayes and P. Piotrowiak, J. Am. Chem. Soc., 2004, 126,
9888.
45 J. R. Bell, R. S. Muthyala, R. W. Larsen, M. Alam and R. S. H. Liu,
J. Phys. Chem. A, 1998, 102, 5481.
1
6 K. Hara, M. Kurashige, Y. Dan-oh, C. Kasada, A. Shinpo,
S. Suga, K. Sayama and H. Arakawa, New J. Chem., 2003, 27, 783.
7 F. Odobel, E. Blart, M. Lagree, M. Villieras, H. Boujtita,
N. E. Murr, S. Caramori and C. A. Bignozzi, J. Mater. Chem.,
1
2
003, 13, 502.
1
1
8 Y. Tachibana, S. A. Haque, I. P. Mercer, J. R. Durrant and
D. R. Klug, J. Phys. Chem. B, 2000, 104, 1198.
9 D. F. Watson, A. Marton, A. M. Stux and G. J. Meyer, J. Phys.
Chem. B, 2003, 107, 10971.
0 C. C. Wamser, H. S. Kim and J. K. Lee, Opt. Mater., 2002, 21, 221.
1 C. Li, W. B. Wang, X. S. Wang, B. W. Zhang and Y. Cao, Chem.
Lett., 2005, 34, 554.
2 Y. S. Chen, Z. H. Zeng, C. Li, W. B. Wang, X. S. Wang and
B. W. Zhang, New J. Chem., 2005, 29, 773.
3 T. Horiuchi, H. Miura and S. Uchida, Chem. Commun., 2003,
46 W. L. F. Armarego and C. L. L. Chai, Purification of Laboratory
Chemicals, Elsevier Science, 5th ed., London, 2003.
47 W. Treibs, Tetrahedron Lett., 1967, 8, 4707.
2
2
48 C. W. Spangler and R. K. Mccoy, Synth. Commun., 1998, 18, 51.
49 N. M. Klene, X. Li, J. E. Knox, H. P. Hratchian, J. B. Cross,
C. Adamo, J. Jaramillo, R. Gomperts, R. E. Stratmann, O. Yazyev,
A. J. Austin, R. Cammi, C. Pomelli, J. W. Ochterski, P. Y. Ayala,
K. Morokuma, G. A. Voth, P. Salvador, J. J. Dannenberg,
V. G. Zakrzewski, S. Dapprich, A. D. Daniels, M. C. Strain,
O. Farkas, D. K. Malick, A. D. Rabuck, K. Raghavachari,
J. B. Foresman, J. V. Ortiz, Q. Cui, A. G. Baboul, S. Clifford,
J. Cioslowski, B. B. Stefanov, G. Liu, A. Liashenko, P. Piskorz,
I. Komaromi, R. L. Martin, D. J. Fox, T. Keith, M. A. Al-Laham,
C. Y. Peng, A. Nanayakkara, M. Challacombe, P. M. W. Gill,
B. Johnson, W. Chen, M. W. Wong, C. Gonzalez and J. A. Pople,
GAUSSIAN 03 (Revision B.03), Gaussian, Inc., Pittsburgh PA,
2003.
2
2
2
2
2
3
036.
4 T. Horiuchi, H. Miura, K. Sumioka and S. Uchida, J. Am. Chem.
Soc., 2004, 126, 12218.
5 R. S. H. Liu, R. S. Muthyala, X. S. Wang, A. E. Asato, P. Wang
and C. Ye, Org. Lett., 2000, 2, 269.
6 S. V. Shevyakov, H. Li, R. Muthyala, A. E. Asato, J. C. Croney,
D. M. Jameson and R. S. H. Liu, J. Phys. Chem. A, 2003, 107,
3295.
2
2
2
7 J. N. Woodford, C. H. Wang, A. E. Asato and R. S. H. Liu,
J. Chem. Phys., 1999, 111, 4621.
8 R. Muthyala, D. Watanabe, A. E. Asato and R. S. H. Liu,
Photochem. Photobiol., 2001, 74, 837.
9 A. E. Asato, R. S. H. Liu, V. P. Rao and Y. M. Cai, Tetrahedron
Lett., 1996, 37, 419.
0 W. C. Herndon, J. Phys. Chem., 1981, 85, 3040.
1 C. Lambert, G. N o¨ ll, M. Zabel, F. Hampel, E. Schm a¨ lzlin,
C. Br a¨ uchle and K. Meerholz, Chem.–Eur. J., 2003, 9, 4232.
2 P. Wang, P. Zhu, C. Ye, A. E. Asato and R. S. H. Liu, J. Phys.
Chem. A, 1999, 103, 7076.
50 M. Kasha, Discuss. Faraday Soc., 1950, 9, 14.
51 M. J. Bearpark, F. Bernardi, S. Clifford, M. Olivucci, M. A. Robb,
B. R. Smith and T. Vreven, J. Am. Chem. Soc., 1996, 118, 169.
52 B. D. Wagner, M. Szymanski and R. P. Steer, J. Chem. Phys.,
1993, 98, 301.
53 K. Sayama, S. Tsukagoshi, T. Mori, K. Hara, Y. Ohga,
A. Shinpou, Y. Abe, S. Suga and H. Arakawa, Sol. Energy
Mater. Sol. Cells, 2003, 80, 47.
54 A. Kay and M. Gr a¨ tzel, J. Phys. Chem., 1993, 97, 6272.
55 C. C. Klark, G. J. Meyer, Q. Wei and E. Galoppini, J. Phys. Chem.
B, 2006, 110, 11044.
3
3
3
3
3 L. Cristian, I. Sasaki, P. G. Lacroix, B. Donnadieu, I. Asselberghs,
K. Clays and A. C. Razus, Chem. Mater., 2004, 16, 3543.
This journal is ß The Royal Society of Chemistry 2007
J. Mater. Chem., 2007, 17, 642–649 | 649