Molecular engineering of cyclopentadithiophene-containing organic dyes for dye-sensitized solar cell: Experimental results vs theoretical calculation

Article abstract of DOI:10.1016/j.dyepig.2013.08.016

Four donor-conjugated spacer-acceptor (D-π-A) organic dyes containing cyclopentadithiophene unit as a part of the conjugated spacer are prepared and studied to investigate the effects of the conjugated spacer and donor on the short-circuit current and conversion efficiency of the corresponding DSSCs. Theoretical calculations reveal that a shorter conjugated spacer or a stronger donor leads to more effective electron injection upon photo excitation. Therefore, D-π-A typed dye with longer conjugated spacer may have higher absorption coefficient and longer λ_max. Nevertheless, it may also have lower effective electron injection upon photo excitation. Calculation data also show that cyclopentadithiophene although is a good moiety for extending the conjugation length of the organic dyes, it also acts as an electron sink, decreasing the effective electron injection. The transition probability calculated with the electron density difference analyses combining with the experimental absorption and photovoltaic performance data provides a useful protocol for designing organic photosensitizers for DSSCs at the electronic and structural levels.

Full text of DOI:10.1016/j.dyepig.2013.08.016

Dyes and Pigments 99 (2013) 1091e1100
Contents lists available at ScienceDirect
Dyes and Pigments
journal homepage: www.elsevier.com/locate/dyepig
Molecular engineering of cyclopentadithiophene-containing organic
dyes for dye-sensitized solar cell: Experimental results vs theoretical
calculation
,
b
,
a
a
a
Chun-Guey Wu a , Wei-Tsung Shieh , Chin-Sheng Yang , Chun-Jui Tan ,
*
Chiao-Hsin Chang a, Su-Chien Chen a, Chia-Yin Wu c, Hui-Hsu Gavin Tsai a
,
**
a Department of Chemistry, National Central University, Jhong-Li 32001, Taiwan, ROC
b Research Center of New Generation Photovoltaics, National Central University, Jhong-Li 32001, Taiwan, ROC
c Photovoltaics Technology Center, Industrial Technology Research Institute, Hsin-Chu 310, Taiwan, ROC
a r t i c l e i n f o
a b s t r a c t
Article history:
Four donoreconjugated spacereacceptor (DepeA) organic dyes containing cyclopentadithiophene unit
Received 3 July 2013
Received in revised form
20 August 2013
Accepted 21 August 2013
Available online 30 August 2013
as a part of the conjugated spacer are prepared and studied to investigate the effects of the conjugated
spacer and donor on the short-circuit current and conversion efficiency of the corresponding DSSCs.
Theoretical calculations reveal that a shorter conjugated spacer or a stronger donor leads to more
effective electron injection upon photo excitation. Therefore, DepeA typed dye with longer conjugated
spacer may have higher absorption coefficient and longer lmax. Nevertheless, it may also have lower
effective electron injection upon photo excitation. Calculation data also show that cyclopentadithiophene
although is a good moiety for extending the conjugation length of the organic dyes, it also acts as an
electron sink, decreasing the effective electron injection. The transition probability calculated with the
electron density difference analyses combining with the experimental absorption and photovoltaic
performance data provides a useful protocol for designing organic photosensitizers for DSSCs at the
electronic and structural levels.
Keywords:
DSSC
Organic sensitizer
Cyclopentadithiophene
Electron injection
Donor
Conjugated spacer
Ó 2013 Elsevier Ltd. All rights reserved.
1. Introduction
component of DSSC devices, molecular engineering of the sensi-
tizers to achieve high photovoltaic performance is an extensive
Global warming and finite fossil fuels have made renewable
energy to be an active research topic. The Sun, providing a huge
amount of energy (3.78 ꢀ 1024 J per year) [1] to the earth, is one of
the best resources of the renewable energy. The most convenient
way to use solar energy is convert it into electricity via photovoltaic
devices. Dye-sensitized solar cells (DSSCs) have been regarded as
one of the most promising candidates for the photovoltaic appli-
cation by virtue of their low manufacturing cost and impressive
conversion efficiency [2]. DSSCs based on ruthenium sensitizers
currently achieve a remarkable conversion efficiency over 11% [3]
and the stability was also demonstrated [4]. Recently the effi-
ciency of DSSC based on zinc porphyrin dye and Co(II)/Co(III)
electrolyte reaches more than 12% [5]. Since the sensitizer is a vital
research topic [6e8].
In general the efficiency of DSSCs based on organic dyes is lower
than that of those sensitized with ruthenium complexes. Never-
theless, organic dyes have the advantages of high molar absorption
coefficient, variety of the structure, environmentally-friendly
property and essentially unlimited resources. Therefore, various
organic dyes for DSSCs, such as coumarin- [9], cyanine- [10], hemi-
cyanine- [11], triarylamine- [12], thiophene- [13], oligoene- [14],
porphyrin- [15], and perylene [16]-based molecules have been
developed, and up to 10% efficiency has been attained [17]. The
structure of the (De
typically consists of an electron donor, an anchoring acceptor, and a
-conjugated spacer for tuning the spectral coverage. Remarkable
peA) organic dyes employed in the DSSCs
p
progress has been made, and the structure-related photovoltaic
performance has also been reviewed [18]. The conjugated spacer in
the sensitizers is widely recognized as an important parameter for
the performance of the DepeA organic dye. It is valuable to be able
to predict their performance before the dye molecules are prepared
and their performance is explored. Thereforea greatnumberof work
* Corresponding author. Department of Chemistry, National Central University,
Jhong-Li 32001, Taiwan, ROC. Tel.: þ886 3 4227151x65903; fax: þ886 3 4227664.
** Corresponding author.
E-mail address: t610002@cc.ncu.edu.tw (C.-G. Wu).
0143-7208/$ e see front matter Ó 2013 Elsevier Ltd. All rights reserved.
http://dx.doi.org/10.1016/j.dyepig.2013.08.016

1092
C.-G. Wu et al. / Dyes and Pigments 99 (2013) 1091e1100
focuses on the theoretical calculations using density functional
theory (DFT) and time-dependent DFT (TD-DFT), complementing
experiments, to study the structure, electronic absorption spectra,
energy levels of the sensitizers both in solution [19e24] and on TiO2
clusters [25e29]. The results provide useful information to expedite
the development of appropriate dyes for DSSCs.
2.3.1. Synthesis of 3,6-di-tert-butyl-9-9H-carbazole-2-thiophen-
4,4-dimethyl-2H-cyclo-penta[2,1-b;3,4-b0]dithiophene-2-acrylic
acid (W-1)
(W-1) was prepared by refluxing the mixture of 3,6-di-tert-
butyl-9-9H-carbazole-2-thiophen-4,4-dimethyl-2H-cyclopenta
[2,1-b;3,4-b0]di-thiophene-2-car-boxaldehyde (1.0 g, 1.7 mmol, see
ESI), NH4OAc (0.20 g, 2.6 mmol), and CNCH2COOH (0.22 g,
2.6 mmol) in CH3COOH (50 mL) for 6 h under Ar. CHCl3 was used to
extract the product, after recrystallization, W-1 dye was obtained as
a red solid, mp ¼ 285 ꢂC (decomposed). IR (1679 cmꢁ1 (C]O);
2213 cmꢁ1 (C^N), see Figure S2, ESI). 1H NMR (Figure S3 (ESI),
300 MHz, dH/ppm in DMSO-d6): 1.45 (18H, s), 1.60 (6H, s), 7.48 (1H,
d, 3.8 Hz), 7.60 (5H, m), 7.73 (1H, s), 8.10 (1H, s), 8.41 (2H, s), 8.52
(1H, s). 13C NMR (Figure S4 (ESI), 300 MHz, dH/ppm in DMSO-d6):
24.41, 31.75, 34.56, 44.99, 109.53, 116.88, 117.28, 118.81, 123.17,
123.48, 124.14, 125.40, 132.35, 132.96, 134.17, 136.63, 137.53, 139.20,
141.85, 143.71, 146.58, 147.00, 160.63, 164.08, 165.95. MS: m/z 660
([M]ꢃ); HRAB-MS found: m/z 660.3426. Elemental analysis: calcd.
for C39H36N2O2S3.H2O: C, 69.02; H, 5.60; N, 4.13; S, 14.16%. Found: C,
68.48; H, 5.73; N, 4.17; S, 14.10%.
This article reports the photovoltaic performance of a series new
De
part of the conjugated spacer. It was known that dithiophene is
generally used as a linker for high-efficiency De eA photo-
peA organic dyes containing cyclopentadithiophene moiety as a
p
sensitizers. Compared to dithiophene, cyclopentadithiophene
(CPDT) unit has a more rigid conjugation structure, higher absorp-
tion coefficient [30] and better charge mobility [31] as well as facile
structure modification. Therefore CPDT has been generally used as a
monomer for the low band-gap polymer [32e35]. Furthermore,
cyclopentadithiophene moiety may act as a photon absorber where
the charge separation occurs, and migrates to the opposite di-
rections by the presence of the donor and acceptor units at each side,
increasing the separation of the excitons. Nevertheless, cyclo-
pentadithiophene is not generally employed in the high-efficient
organic dyes. Therefore, knowing the parameters which deter-
mine the photovoltaic performance of cyclopentadithiophene-
containing organic dyes is the focus of this article.
2.3.2. Synthesis of 3,6-di-tert-butyl-9-[5-(thiophen-2-yl)thiophen-
2-yl]-9H-carbazole-4,4-dimethyl-2H-cyclopenta[2,1-b;3,4-b0]
dithiophene-2-acrylic acid (W-2)
2. Experimental
The preparation procedures of W-2 dyes are the same as those
for synthesizing W-1 dye except 3,6-di-tert-butyl-9-[5-(thiophen-
2-yl)thiophen-2-yl]-9H-carbazole instead of 3,6-di-tert-butyl-9-
(thiophen-2-yl)-9H-carbazole (2.93 g, 6.8 mmol, see ESI) was used
to react with Me3SnCl (1.5 M, 3 mL). 3,6-di-tert-butyl-9-[5-(thio-
phen-2-yl)thiophen-2-yl]-9H-carbazole was prepared by Stille
coupling between 3,6-di-tert-butyl-9-[5(trimethylstannanyl)-thio-
phen-2-yl]-9H-carbazole and 2-bromothiohene. W-2 dye was iso-
lated as a dark green solid, mp ¼ 300 ꢂC (decomposed). IR
(1679 cmꢁ1 (C]O); 2213 cmꢁ1 (C^N), see Figure S2, ESI). 1H NMR
2.1. Materials
All reagents were obtained from the commercial resources and
used as received unless specified. Solvents were dried over sodium
or CaH2 before use. Unless otherwise specified, all reactions were
performed under an Ar atmosphere using the standard Schlenk
techniques. All chromatographic separations were carried out on
the silica gel (60M, 240e400 mesh) column. The structures of dyes
and their intermediates were identified with 1H NMR spectra. The
structure of the final products was further confirmed by IR, 13C NMR
spectra, FAB-MS and elemental analysis.
spectrum was displayed in Figure S5 (ESI). 1H NMR (300 MHz, dH
/
ppm in DMSO-d6): 1.51 (18H, s), 1.57 (6H, s), 7.50 (8H, m), 7.71 (1H,
s), 8.06 (1H, s), 8.39 (2H, s), 8.47 (1H, s). 13C NMR (Figure S6,
300 MHz, dH/ppm in DMSO-d6): 24.87, 32.21, 35.01, 45.50, 109.99,
117.27, 119.08, 119.18, 123.59, 123.86, 124.63, 125.59, 125.92, 126.03,
130.60, 133.54, 133.91, 135.62, 136.52, 137.69, 137.84, 139.72, 140.93,
144.02, 144.18, 161.05, 165.28, 165.55. MS: m/z 742 ([M]ꢃ); LRFAB-
MS found: m/z 742.2990. Elemental analysis: calcd. for
2.2. Physicochemical studies
1H and 13C NMR spectra were recorded with a Bruker 300 MHz
NMR spectrometer in CDCl3 or [D6]DMSO. FAB-MS spectra were
obtained using the JMS-700 HRMS. UV/Vis spectra of dyes in THF
C43H38N2O2S4.0.5H2O: C, 68.71; H, 5.19; N, 3.72; S, 17.04%. Found: C,
and adsorbed on TiO2 (4
mm thickness) films were measured using a
68.69; H, 5.04; N, 3.27; S, 17.05%.
Cary 300 Bio spectrometer. Photoluminescence spectra (in THF
solutions, Figure S1, ESI) were obtained using a Hitachi F-4500
spectrophotometer in the laboratory atmosphere at room temper-
ature. Cyclic voltammetric measurement was performed in a
single-compartment, three-electrode cell with an ITO glass work-
ing electrode and a Pt wire counter electrode. The reference elec-
trode was Ag/Agþ and the supporting electrolyte was 0.1 M LiClO4
in THF using an Autolab system (PGSTAT 30, Autolab, Eco-Chemie,
the Netherlands), the scan rate is 50 mV sꢁ1 and the ferrocene/
ferrocinium redox couple was used as an external calibration
standard. Elemental analysis was carried out with a Heraeus CHNe
OeS Rapid-F002 analysis system. The thickness of TiO2 film was
measured by a profile-meter (Dektak3, Veeco/Sloan Instruments
Inc., USA).
2.3.3. Synthesis of 3,6-di-tert-butyl-9-(3-octylthiophen-2-yl)-9H-
carbazole-4,4-dimethyl-H-cyclopenta[2,1-b;3,4-b0]di-thiophene-2-
acrylic acid (W-3)
The preparation procedures of W-3 dye is the same as those for
the synthesis of W-1 dye except 2-bromo-3-octylthiophene instead
of 2-bromothiophene was used to react with 3,6-ditert
butylcarbazole. W-3 dye was isolated as a red solid, mp ¼ 240 ꢂC
(decomposed). IR (1679 cmꢁ1 (C]O); 2213 cmꢁ1 (C^N), see
Figure S2, ESI).1H NMR spectrumwas displayed in Figure S7 (ESI).1H
NMR (300 MHz, dH/ppm in DMSO-d6): 0.81 (3H, t, 7.0 Hz),1.08 (10H,
s),1.48 (18H, s),1.51 (8H, s), 2.33 (2H, t, 7.2 Hz), 7.23 (1H, s), 7.26 (1H,
s), 7.59 (3H, m), 7.71 (1H, s), 8.09 (1H, s), 8.38 (2H, s), 851 (1H, s). 13
C
NMR (Figure S8 (ESI), 300 MHz, dH/ppm in DMSO-d6): 14.04, 22.54,
24.94, 27.61, 29.04, 29.08, 29.79, 31.70, 31.99, 34.75, 45.66, 92.65,
109.63, 116.27, 116.70, 117.57, 123.54, 123.89, 124.56, 130.84, 132.59,
133.09, 134.98, 136.43, 140.57, 142.17, 143.55, 144.28, 148.05, 149.53,
161.30, 166.51, 169.01. MS: m/z 772 ([M]ꢃ); LRFAB-MS found: m/z
772.4457. Elemental analysis: calcd. for C47H52N2O2S3.0.5H2O: C,
72.21; H, 6.91; N, 3.59; S,12.29%. Found: C, 72.62; H, 6.86; N, 3.55; S,
12.55%.
2.3. Synthesis of W-series dyes
The detailed synthesis and characterization of all intermediates
and the IR spectra as well as 1H and 13C NMR spectra (Figures S2e
S10) of these four metal free dyes can be found in the electronic
supporting information (ESI).

C.-G. Wu et al. / Dyes and Pigments 99 (2013) 1091e1100
1093
2.3.4. Synthesis of 2-cyano-{5-[4-(diphenylamino)phenyl]-2-
thienyl]-4,4-dimethyl-2H-cyclopenta[2,1-b;3,4-b0]di-thiophene-2-
acrylic acid (W-4)
dye adsorbed on TiO2 thin film indicating that the adsorbed dyes
were mainly in the deprotonated form. Therefore, we used depro-
tonated dye to model the dye-(TiO2)38 system [45]. The dye mol-
ecules were adsorbed on (TiO2)38 cluster via its two oxygen atoms
of the carboxylic acid group in a bidentate manner because the
bidentate adsorption mode is more stable than the monodentate
adsorption mode [46,47]. The optimized geometries of the dye-
(TiO2)38 from the CareParrinello [41] damped molecular dynamics
calculations were used for TD-DFT calculations to determine the
lowest excitation energies calculated at TD-CAM-B3LYP/6e
31G(d,p) level using Gaussian 09 program [37].
Under
Ar,
5-[4-(diphenylamino)phenyl]-2-thiophen-4,4-
dimethyl-2H-cyclopenta [2,1-b; 3,4-b0]di-thiophene-2-carbox-
aldehyde (0.7 g, 1.1 mmol), NH4OAc (0.14 g, 1.8 mmol), CNCH2COOH
(0.16 g, 1.9 mmol) and CH3COOH (50 mL) were mixed and refluxed
for 6 h. The product was extracted with CHCl3, after recrystalliza-
tion from dichloromethane/hexane, W-4 dye was isolated as a dark
green solid, mp ¼ 280 ꢂC (decomposed). IR (1679 cmꢁ1 (C]O);
2213 cmꢁ1 (C^N) see Figure S2, ESI). 1H NMR (Figure S9, 300 MHz,
dH/ppm in DMSO-d6): 1.58 (6H, s), 7.14 (8H, m), 7.40 (6H, m), 7.66
(3H, m), 8.06 (1H, s), 8.48 (1H, s). 13C NMR (Figure S10, 300 MHz, dH
/
3. Results and discussion
ppm in DMSO-d6): 24.41, 30.64, 44.95, 118.17, 122.78, 123.56,
124.00,124.42,125.64,126.36,126.87,129.64, 132.13,132.43,135.03,
136.48, 139.20, 142.34, 142.85, 146.72, 147.10, 160.49, 164.12, 166.09,
206.48. MS: m/z 626 ([M]ꢃ); LRFAB-MS found: m/z 626.2602.
Elemental analysis: calcd. for C37H26N2O2S3.H2O: C, 68.94; H, 4.35;
N, 4.34; S, 14.91%. Found: C, 68.91; H, 4.24; N, 4.00; S, 15.45%.
3.1. Synthesis and optical properties of W-series dyes
W-1eW-4 dyes were synthesized according to Schemes 1 and
2. Two types of donor: triphenyl amine and carbazole are used as
a donor moiety. Cyanoacetic acid is used as an anchor and
acceptor. The conjugated spacer is a combination of (oligo)
thiophene and cyclopentadithiophene moieties. A palladium
catalyzed Stille coupling reaction [48] is used to connect the
conjugated spacer, donor and acceptor. Cyclopentadithiophene
reacts with n-butyl lithium and DMF in THF to produce the
aldehyde group, which is used for the production of the
anchoring/donor group. The aldehyde then reacts with the cya-
noacetic acid by Knoevenagel reaction using ammonium acetate/
acetic acid as a catalyst to produce the final products. W-1 and
W-3 dyes are deep red color, whereas W-2 and W-4 dyes are
dark green powders (the detailed preparation and characteriza-
tion of all intermediates can be found in the Electronic
Supporting Information (ESI)). All dyes exhibit fairly good ther-
mal stability: TGA data (shown in Fig. 1) reveal that they all can
stable up to 200 ꢂC under nitrogen atmosphere.
All dyes are soluble in THF (tetrahydrofuran) with the absorp-
tion coefficient higher than 60,000 molꢁ1 dm3 cmꢁ1 at their lmax
(see Fig. 2a and the optical data are summarized in Table 1). The
general feature of the absorption spectra for W-series dyes (except
W-4) consists one broad band in the visible region and a weak
absorption band in the UV region which correspond to the pe
p*
transitions (or intramolecular charge transfer) for the molecule and
dialkylcyclopentadithiophene unit, respectively. The broad ab-
2.4. Fabrication and characterization of W-series dyes sensitized
DSSCs
The procedure for the fabrication and characterization of DSSCs
is the same as what we reported previously [36]. The characteristic
current densityevoltage (JeV) curves of the cells under the illu-
mination of AM 1.5G simulated sunlight (Yamashita Denso Corpo-
ration, YSS-50A,100 mW cmꢁ2) were obtained by applying external
potential bias to the cell and measuring the photocurrent output
with a Keithley model 2400 digital source meter (Keithley, USA).
Incident photon to electron conversion efficiency (IPCE) spectra
were recorded with a custom-designed instrument provided by
Enlitech Co., Taiwan, in which a 300 W Xenon lamp was focused
through a Gemini-180 double monochromator to produce the
excitation beam and Keithley 2400 Source Meter (Keithley) was
used to record the current.
2.5. Computational methods
The ground-state molecular geometries were optimized by the
Becke, three-parameter, LeeeYangeParr (B3LYP) functional and 6e
31G (d, p) basis set, as implemented in the Gaussian 09 program
[37]. The Conductor-like Polarizable Continuum Model (C-PCM)
[38] was used to account for THF solvation effect. For the conve-
nience of the calculations, the long alkyl chain was replaced by a
methyl group. Tetramethylammonium (TMA) was used to model
the counter ion. The time-dependent DFT (TD-DFT) calculations
were performed to calculate the UVeVis spectra. To account for the
charge transfer excitations, the coulomb-attenuating method
(CAM) [39] was applied (calculated at TD-CAM-B3LYP/6e31G(d,p)
level) to calculate the UVeVis spectra of the sensitizers based on
their optimized geometries calculated at B3LYP/6e31G(d,p) level.
Electron density difference map (EDDM) of the electronic transi-
tions was calculated using the Gaussum program (version 2.5.5)
[40].
sorption in the visible region for the intramolecular
transfer transition is a typical behavior for organic dyes with a De
eA architecture. It was found that lmax of the dyes depends on the
pe
p* charge
p
conjugated spacer and the strength of the donor. The lmax of the
dye with a triphenyl amine donor (W-4) are red-shifted compared
to that with a carbazole donor (W-1) and dye with a longer
conjugation spacer has a longer lmax. However, when a long side
chain was added in the thiophene ring (i.e. W-3), the lmax of the dye
was slightly blue-shifted by 5 nm (see Table 1). This is probably
because of the conformation of W-3 is not as planar as W-1, due to
the steric effect of the large octyl group.
The absorption spectra of the W-series dyes adsorbed on TiO2
thin films are displayed in Fig. 2b and the related data are also listed
in Table 1. Compared to the spectra of the corresponding dyes in
THF, the absorption spectra of the dyes adsorbed on TiO2 are
slightly broader, and at the same time, lmax blue-shifts by 30e
47 nm. The broadening of the absorption peak can be attributed to
the interaction between the carboxylate group and TiO2 surface or
the molecular conformation of the dyes adsorbed on TiO2 may not
be totally the same. The blue-shift in lmax suggests that there is an
H-aggregation of the dyes or the dye molecules are deprotonated
(carboxylate is a weaker acceptor compared to carboxylic acid)
when adsorbed on TiO2 [49].
We employed damped CareParrinello [41] molecular dynamics
to optimize the structure of dye-(TiO2)38 system using CPMD pro-
gram [42]. Damped CareParrinello [41] molecular dynamics was
performed in a vacuum, using the PBE functional [43] in conjunc-
tion with PBE-consistent ultra soft pseudopotentials [44] with a
cut-off of 25 Ry. The (TiO2)38 cluster has an Anatase (101) surface.
The dye-(TiO2)38 system was centered in a periodically repeated
orthorhombic box and the system was separated by a vacancy
ꢀ
w10.0 A wide. The calculated UVeVis spectra of the free deproto-
nated dyes in THF were similar to the experimental spectra of the

1094
C.-G. Wu et al. / Dyes and Pigments 99 (2013) 1091e1100
Scheme 1. Synthetic procedures for the preparation of W-1, W-2, and W-3 dyes.
3.2. Electro-optical properties and photovoltaic performance of the
W-series dye
voltammetry (see Figure S11, ESI) and the band gap energy, E0e0,
defined by the intersection of the absorption and emission spectra,
see Figure S1, ESI. As depicted in Table 2, based on W-1, adding one
more thiophene unit to the dye (i.e. W-2) does slightly drive up the
HOMO level, but keeps the LUMO level almost the same, therefore
The oxidation potential and energy level of the frontier orbitals
for the W-series dyes listed in Table 2 are determined by cyclic
Scheme 2. Synthetic procedures for the preparation of W-4 dye.

C.-G. Wu et al. / Dyes and Pigments 99 (2013) 1091e1100
1095
Table 1
100
80
60
40
20
0
The optical data of W-series dyes.
W1
W2
W3
W4
Dye
lmax in THF
(nm)
3
at lmax
lmax adsorbed
on TiO2 (nm)
(Mꢁ1 cmꢁ1
)
W-1
W-2
W-3
W-4
490
501
485
505
65,157
71,446
62,232
84,006
450
466
455
458
the E0e0 energy decreases. Adding an octyl group to the thiophene
ring (i.e. W-3) increases the oxidation potential and band-gap en-
ergy (E0e0) simultaneously, probably due to the steric effect of the
octyl chain making W-3 dye less planar (effectively a shorter
conjugation pathway). Nevertheless, LUMO for all dyes (ꢁ1.12 to
ꢁ1.32 V vs NHE) is higher than TiO2 conduction band edge (ꢁ0.5 V
vs NHE), which provides a sufficient thermodynamic driving force
(0.62e0.82 V) for the electron injection [50]. HOMO for the W-se-
ries dyes falls to 0.90e1.20 V (vs NHE) which is more positive than
the redox potential (0.4 V vs NHE) of the iodide/triiodide couple
used in DSSC, facilitating the dye regeneration. The frontier orbital
energy levels of the W-series dyes match the requirement for a dye
to be used in DSSCs.
200
400
600
800
o
Temperature ( C)
Fig. 1. TGA of the W-series dyes.
Preliminary photovoltaic performance tests are carried out and
the resulting JeV curves and IPCE plots are shown in Fig. 3. The
photovoltaic parameters are summarized in Table 3 and the
detailed DSSC device fabrication and photovoltaic performance
measurements can be found in the ESI. As listed in Table 3, DSSC
based on the W-4 dye has the highest efficiency of 5.0%, which is
75% of that for N719 (the structure was also shown in Scheme 2)
sensitized device fabricated using the same fabrication procedures.
The good photovoltaic performance of the W-4 dye may be
attributed to its high absorption coefficient and the absorption
profile matches better with the AM 1.5G Sunlight, as well as its
structure has a strong triphenyl amine donor, results in higher
short-circuit current (Jsc). Furthermore, the quantity of the dye
molecules adsorbed on the practical TiO2 electrode (displayed in
Table 3) is also an important parameter for the photovoltaic per-
formance of a DSSC. The values can be estimated by desorbing the
dye molecules from the practical TiO2 electrode using a basic so-
lution as reported [51]. It was found that, except for W-3, a similar
amount of molecules is adsorbed on TiO2 electrode. The fact that a
fewer W-3 molecules adsorbed on TiO2 is due to W-3 has large octyl
chains on the thiophene ring, which occupies a larger TiO2 surface.
The energy levels of the frontier orbitals and E0e0 energy for W-
2 are close to those for W-4 (see Table 2) but the conversion effi-
ciency of W-2 based DSSC is much lower. This may be attributed to
the difference in the donor group (triphenyl amine for W-4 dye vs
carbazole for W-2 dye) or the higher absorption intensity of W-4
self-assembled on TiO2 (see Fig. 2b). On the other hand, compared
to W-2, W-1 has the same donor (carbazole), a similar number of
molecules adsorbed on TiO2 and lower absorption coefficient.
Nevertheless, the conversion of a W-1 sensitized device is higher
than for a W-2 based device. The aggregation (see Fig. 2b) of W-2
dye when adsorbed on TiO2 may be the reason for its low efficiency.
Table 2
The electro-optical data of W-series dyes.
Dye E(Dþ/D)a (vs NHE) E0e0 (eV) HOMO (V vs NHE) LUMO (V vs NHE)
b
W-1 0.49
W-2 0.41
W-3 0.61
W-4 0.41
2.26
2.22
2.32
2.21
0.98
0.90
1.20
0.90
ꢁ1.28
ꢁ1.32
ꢁ1.12
ꢁ1.31
a
HOMO (V vs NHE) ¼ Eox ꢁ Efc/fcþ þ 0.63.
b
Fig. 2. UV/Vis absorption spectra of the W-series dyes (a) in THF, (b) adsorbed on TiO2.
E0e0 was defined by the intersection of the absorption and emission spectra.

1096
C.-G. Wu et al. / Dyes and Pigments 99 (2013) 1091e1100
20
16
12
8
100
(a)
W-1
W-1
W-2
W-3
W-4
(b)
W-2
80
W-3
60
W-4
N719
40
20
0
4
0
400
500
600
700
800
0.0
0.2
0.4
0.6
0.8
Wavelength (nm)
Voltage (V)
Fig. 3. (a) IeV curves, (b) IPCE plots of DSSCs sensitized with the W-series dyes.
The aggregation of W-2 is due to its long molecular axis and planar
conformation. We try to reduce the aggregation of W-2 dye by
adding more (30 mM) co-adsorbent (Chenodeoxycholic acid) in the
dye solution, but the efficiency of the corresponding device de-
creases due to the completing adsorption between the dye mole-
cule and co-adsorbent.
excitation. For convenience of the discussion, herein we abbreviate
these two structural parameters with referring to their two adja-
cent rings. For example, the dihedral angle between the phenyl ring
on the TPA unit and thiophene ring (T) is abbreviated as
and the length of the bond connecting these two adjacent rings is
abbreviated as d(TPAeT); N-arylcarbazole and cyclo-
s(TPAeT)
Interestingly, as mentioned in the previous paragraph that
higher conversion efficiency of W-4 sensitized DSSC (compared to
W-2 based device) is due to the higher absorption intensity
(combining the effect of the absorption coefficient and dye loading)
of W-4 dye adsorbed on TiO2. On the other hand, comparing W-1
and W-2 dyes we found that W-2 dye adsorbed on TiO2 has higher
absorption intensity but worse photovoltaic performance of the
corresponding device. This may be due to the fact that W-2 dye has
a higher HOMO energy level, therefore dye regeneration is more
difficult even the LUMO level of W-2 dye is higher than W-1 dye,
which may be benefit for electron injection from dye to TiO2. These
results reveal that it is hard to predict the photovoltaic performance
of a dye simply by one of its physicochemical properties or mo-
lecular architecture. Therefore, instead of using various spectro-
scopic methods to identify what are the major parameters for
pentadithiophene rings are abbreviated as NCAR and CPDT,
respectively. Regardless of the molecular state, the electron
donating groups, NCAR and TPA are non-planar with its neigh-
boring thiophene ring. For example, the
s
(NCAReT) dihedral angles
(TPAeT)
in W-1, W-2, and W-3 were in the range of 60ꢂe85ꢂ;
s
dihedral angles in W-4 are in the range of 17ꢂ w 18ꢂ. These
calculation results indicate that the terminal electron donating
groups, NCAR and TPA, are only partially conjugated with the
molecular backbone. For the dihedral angle between the conju-
gated spacer and acceptor such as s(TeT) and s(TeCPDT), the
dihedral angles indicated a certain degree of non-planarity but they
are less twisted than those of the donor-spacer. On the other hand,
the anchoring groups generally remain nearly planar with its
neighboring CPDT rings.
The carbonecarbon bond lengths between two rigid rings
ꢀ
determining the photovoltaic performance of a DepeA organic dye
depicted in Fig. 4 were in the range of 1.440e1.463 A. These car-
(especially those containing cyclopentadithiophene unit), we used
the theoretical calculation to probe the electronic state of the dyes
adsorbed on TiO2 before and after photo excitation to understand
how to design high efficiency organic dyes.
bonecarbon bond lengths are shorter than that of a standard car-
ꢀ
bonecarbon single bond (1.54 A) and longer than a standard
ꢀ
carbonecarbon double bond (1.34 A). The characteristic carbone
carbon partial double bond indicates the existence of the
p-
conjugation between two rigid aromatic rings. Interestingly, the
bond lengths of the anchoring groups at the protonated and
deprotonated states are different. When the dye molecules are
deprotonated, the bonds of the anchoring groups generally elon-
gate. These results suggest that the deprotonated dyes are less
conjugated with the molecular framework than their correspond-
ing protonated states. The calculated frontier orbitals of the pro-
tonated W-series dyes in THF are displayed in Fig. 5, only HOMO
and LUMO orbitals are shown. The distribution of LUMO of all dyes
is similar, mainly locating on the CPDT and acceptor/anchor group.
The distribution of HOMO (although all are mainly on the conju-
gated spacer) is different: triphenyl amine donor contributes
significantly but little dispensation on carbazole donor. It is hard to
predict precisely the effective charge transfer of these W-series
dyes simply by the isodensity surface plots of the frontier orbitals.
More detailed calculation is needed.
3.3. Calculated ground state molecular geometry and electron
localization of the frontier orbitals of W-series dyes
Fig. 4 displays the selected structural parameters of W-1eW-4
molecules in their protonated and deprotonated states in THF.
These structural parameters are the bond lengths and dihedral
angles of the two adjacent rigid aromatic rings, which are more
sensitive to the chemical environments and will influence the de-
gree of
p-conjugation of the dye molecule. Furthermore, the degree
of -conjugation also affects the absorption spectra and coupling
p
strength between the donor and acceptor moieties, which conse-
quently affects the probability of the electron transfer upon photo
Table 3
The photovoltaic performance and dye loading of W-series dyes sensitized DSCs.
Dye
Voc (V)
Jsc (mA/cm2)
FF
h
(%)
Dye loading (mole/cm2)
3.4. Calculated UVeVis spectra of the W-series dyes in solution and
W-1
W-2
W-3
W-4
N719
0.66
0.62
0.66
0.66
0.66
10.52
9.74
7.21
11.91
17.37
0.69
0.71
0.71
0.69
0.60
4.78
4.33
3.38
5.39
6.7
0.95 ꢀ 10ꢁ8
0.94 ꢀ 10ꢁ8
0.40 ꢀ 10ꢁ8
0.88 ꢀ 10ꢁ8
1.81 ꢀ 10ꢁ8
adsorbed on TiO2
Table 4 lists the computational and experimental electronic
absorptions of the W-series dyes in THF. Here, the trends in the
experimental UVeVis spectra and their relationship to the

C.-G. Wu et al. / Dyes and Pigments 99 (2013) 1091e1100
1097
ꢂ
ꢀ
Fig. 4. Selected molecular geometries for the four studied dye molecules. The bond lengths (A) are labeled around their corresponding bonds. The dihedral angles ( ) are labeled
around their corresponding central bonds; their four consecutive atoms are highlighted in bold. The data for dye molecules are arranged, from top to bottom, for the protonated
state in THF and the deprotonated state in THF.
Table 4
shifted 11 nm. On the other hand, W-4 and W-2 dyes have the same
spacer and acceptor, but with different donor; compared to W-2
dye, lmax of W-4 dye red-shifted by 4 nm. This result indicates that
Experimental and computational electronic absorptions of W-series dyes in THF
solution.
Dyes Experiment Calculation
Calculation (deprotonated in THF)
TPA is more conjugated as a conjugated spacer than N-arylcarba-
zole, as evidenced by the smaller
(TPAeT) in W-4 dye (ca. 18ꢂ) than
(NCAReT) in W-1 dye (ca. 60ꢂ).
The degree of protonation of the dye molecules in solution is
(in THF)
(protonated in THF)
s
lmax
lmax
Deviation
from EXP.
lmax
Deviation
from EXP.
s
(nm)
(nm)
(nm)
important because it affects the adsorption of the dyes on TiO2
anode. Table 4 summarizes the calculated electronic absorptions for
the W-series dyes in THF at their protonated and deprotonated
states. The calculated electronic absorptions of the dye molecules at
their protonated states in THF agree well with the experimental
observations; the smallest absolute deviation from the experi-
mental value is 6 nm blue-shift for W-1 dye and the largest absolute
deviation was a 9 nm red-shift for W-2 and W-4 dyes. In contrast,
the calculated values of lmax for the W-series molecules in THF
based on their deprotonated state deviate widely from the exper-
imental data; all absolute deviations are greater than 20 nm, with
the largest absolute deviation up to 46 nm occurring for W-1 dye.
Based on the calculated UVeVis spectra, the W-series dyes may
adopt protonated states in THF solution. Interestingly, the calcu-
lated values of lmax for the W-series dyes in their deprotonated
states in THF agree well with the experimental lmax of dyes
adsorbed on TiO2, consistent with what is well-known that dye
adsorbed on TiO2 is in a deprotonated form.
W-1 490
484
510
478
514
ꢁ6
þ9
ꢁ7
þ9
444
472
444
469
ꢁ46
ꢁ29
ꢁ41
ꢁ36
W-2 501
W-3 485
W-4 505
molecular geometries are discussed first. Only one major absorp-
tion for all studied dyes are observed. Relative to W-1 dye, W-3 dye
features a less planar structure with its lmax blue-shifted by 5 nm;
W-2 dye features a longer p-conjugation length, with its lmax red-
3.5. Calculated UVeVis spectra of W-series dyes adsorbed on TiO2
The experimental electronic transitions of the dye molecules
adsorbed on TiO2 thin film are all blue-shifted by 30e47 nm with
respect to their corresponding spectra taken in THF solutions. This
is a general phenomenon for the sensitizers containing cyanoacetic
acid acceptors [52e55]. The calculated electronic transitions of the
dye molecules adsorbed on (TiO2)38 clusters are listed in Table 5.
We assume that the dye molecules are in their deprotonated states
and adsorbed on (TiO2)38 cluster in a bidentate manner. The
calculated lmax values for the dyes on (TiO2)38 clusters have abso-
lute deviations of 1e26 nm from their experimental values. The
Fig. 5. Isodensity surface plots of the frontier orbitals for the W-series dyes.

1098
C.-G. Wu et al. / Dyes and Pigments 99 (2013) 1091e1100
Table 5
Experimental and computational electronic absorptions of W-series dyes adsorbed on (TiO2)38 cluster.
Dyes
W-1
Experimental (on TiO2)
Calculated [adsorbed on (TiO2)38]
lmax (nm)
Deviation from that in THF
lmax (nm)
Deviation from EXP.
Oscillator strength
2.157
Transitiona
450
ꢁ40
459
þ9
HOMO / LUMO þ 5 (66%),
HOMO / LUMO þ 6 (19%),
HOMO / LUMO þ 4 (5%)
HOMO / LUMO þ 2 (89%)
HOMO / LUMO þ 5 (77%),
HOMO / LUMO þ 4 (8%),
HOMO / LUMO þ 6 (7%)
HOMO / LUMO þ 4 (50%),
HOMO / LUMO þ 6 (29%),
HOMO / LUMO þ 5 (7%)
W-2
W-3
466
455
ꢁ35
ꢁ30
486
456
þ20
þ1
2.591
2.081
W-4
458
ꢁ47
484
þ26
2.528
a
Transitions with contributions greater than 5% are listed.
electronic transitions are mainly arisen from HOMO located at dye
molecules to higher unoccupied orbitals located at (TiO2)38 clusters
and/or dye molecules.
the electron density of W-1 is directly transferred to (TiO2)38 cluster
after photo excitation (EDDM shown Fig. 6a) indicating that the
electron injection of W-1 follows both the direct and indirect
photosensitization processes. The direct photosensitization pro-
cesses arise from the optical electron transfer, which the excited
state of the sensitizer can be considered as an oxidized state and
TiO2 can be described as a reduced state [56]. The indirect photo-
sensitization (based on photo induced electron transfer) is a
multistep process involving consecutive electron transfer to the
appropriate band of a semiconductor (e.g. TiO2). Therefore, the
direct photosensitization should give rise to a more efficient elec-
tron injection although the back electron transfer process might
also be attainable [57]. On the other hand, only 4% of the electron
density of W-2 was directly transferred to the (TiO2)38 cluster after
photo excitation (EDDM shown Fig. 6b) indicating that the electron
injection of W-2 mainly followed the indirect photosensitization.
The indirect photosensitization is generally considered less effi-
cient for the electron transfer to the conduction band of (TiO2)38
cluster but the back electron transfer is retarded by an energy
barrier. It is known that the Jsc is proportional to the efficiency of
the electron injection from the excited sensitizers to TiO2 anode
[58]. Therefore, our calculations suggest that W-1 based device
should have a larger Jsc compared to W-2 sensitized device,
consistent with the experimental data. Moreover, W-2 dye which
has higher LUMO energy level should have a high electron injection
driving force compared to W-1 dye. We calculate a model molecule
(W-5 dye, structure displayed in Figure S12 of ESI) with a molecular
framework similar to W-1 but without the thiophene spacer. The
result shows that 75% of electron density on W-5 is transferred to
(TiO2)38 cluster directly (detailed data not shown). Therefore the
lower efficiency of the electron transfers (from dye to (TiO2)38) for
W-2 dye compared to W-1 dye may be due to W-2 dye has a longer
electron transfer path (an additional thiophene ring).
3.6. Calculated electronic transition upon photo excitation for the
dye adsorbed on TiO2
Table 6 lists the characteristics of the electron densities of the
W-series dyes adsorbed on (TiO2)38 clusters before and after photo
excitation. Before photo excitation, the electron density of each dye
molecule mainly resided on CPDT unit (50e65%) and had no pop-
ulation on (TiO2)38 cluster. After excitation, for W-1, W-3, and W-4,
(TiO2)38 cluster received significant amount (31e58%) of electron
density from the dye molecule. In contrast, after excitation, (TiO2)38
cluster received a only very minor amount (4%) of electron from the
W-2 dye molecule. These results indicate that the electron injection
of W-2 dye to (TiO2)38 is less efficient than other dyes studied in this
article.
Data listed in Tables 1e3 reveal that W-1 and W-2 dyes possess
some interesting relationships: (a). W-1 and W-2 dyes have similar
amounts of dye loading on TiO2 anode; (b). W-2 dye owns an
approximately 40% larger optical-absorbance at its lmax and has
higher absorbance at all wavelengths compared to W-1 dye; and
(c). W-1 and W-2 have similar open-circuit voltages (Voc). These
three relationships provide us a system suitable to study whether
the distinct electronic transitions affects the short-circuit current
densities (Jsc) and thereafter the conversion efficiencies (h) of the
corresponding devices or not. From the absorption profiles, we may
expect W-2 sensitized DSSC should have larger Jsc than W-1 based
device. However, experimental data show that Jsc of W-2 based
device (9.74 mA/cm2) is ca. 8% lower than that (Jsc ¼ 10.52) of W-1
sensitized device. However, the calculation results show that 44% of
Among all dyes studied, W-4 sensitized device has the highest
h
and largest Jsc. These results may be ascribed to the highest dye
Table 6
Characteristics of electron densities of deprotonated organic W-series dyes adsorbed
on (TiO2)38 cluster before and after transition.
Dye Transition Percent contribution (%)
Donor Thiophenea Thiophene CPDT Anchor (TiO2)38
W-1 Before
After
6
0
17
8
e
e
61
32
16
16
0
44
W-2 Before
After
2
0
9
9
20
18
56
47
13
22
0
4
W-3 Before
After
1
0
16
10
e
e
65
39
18
20
0
31
Fig. 6. Electron density difference maps (EDDMs) of the transitions for (a) W-1 dye
and (b) W-2 dye on TiO2. Cyan indicates an increase in electron density; light blue
indicates a decrease. (For interpretation of the references to color in this figure legend,
the reader is referred to the web version of this article.)
W-4 Before
After
17
2
21
6
e
e
50
22
12
12
0
58
a
This thiophene is connected with the donor.

C.-G. Wu et al. / Dyes and Pigments 99 (2013) 1091e1100
1099
[3] Nazeeruddin MK, De Angelis F, Fantacci S, Selloni A, Viscardi G, Liska P, et al.
Combined experimental and DFTeTDDFT computational study of
photoelectrochemical cell ruthenium sensitizers. J Am Chem Soc 2005;127:
16835e47.
loading and largest optical absorbance of W-4 dye. On the other
hand, our calculations show that W-4 has more than 50% of its
electron density transferred to (TiO2)38 cluster after photo excita-
tion, which mainly followed the direct photosensitization. As dis-
cussed above, the direct photosensitization should result in a
[4] Wang P, Zakeeruddin SM, Moser JE, Nazeeruddin MK, Sekiguchi T, Grätzel M.
A
stable quasi-solid-state dye-sensitized solar cell with an amphiphilic
ruthenium sensitizer and polymer gel electrolyte. Nat Mater 2003;2:402e7.
[5] Yella A, Lee HW, Tsao HN, Yi C, Chandiran AK, Nazeeruddin M, et al.
Porphyrin-sensitized solar cells with cobalt (II/III)-based redox electrolyte
exceed 12 percent efficiency. Science 2011;334:629e34.
[6] (a) Wang P, Zakeeruddin SM, Moser JE, Humphry-Baker R, Comte P, Aranyos V,
et al. Stable new sensitizer with improved light harvesting for nanocrystalline
dye-sensitized solar cells. Adv Mater 2004;16:1806e11;
higher Jsc and therefore h, consistent with the experimental data.
Furthermore, the molecular framework of W-4 is analogous to that
of W-1, but with a TPA (instead of carbazole) unit as a donor. The
higher charge transfer efficiency for W-4 dye compared to W-1 dye
(58% vs 44%) suggests TPA is a better electron donor moiety
compared to carbazole.
(b) Chen CY, Wu SJ, Wu CG, Chen JG, Ho KC. A new ruthenium complex with
super-high light harvesting capacity for dye-sensitized solar cells. Angew
Chem Int Ed 2006;45:5822e5.
[7] Wang P, Klein C, Humphry-Baker R, Zakeeruddin SM, Grätzel M. A high molar
extinction coefficient sensitizer for stable dye-sensitized solar cells. J Am
Chem Soc 2005;127:808e9.
[8] (a) Snaith HJ, Zakeeruddin SM, Schmidt-Mende L, Klein C, Grätzel M. Ion-
coordinating sensitizer in solid-state hybrid solar cells. Angew Chem Int Ed
2005;44:6413e7;
4. Conclusion
A series of metal free donoreconjugated spacereacceptor (De
peA) organic dyes containing cyclopentadithiophene unit in the
spacer are synthesized and their photovoltaic performance is tested.
These designed photosensitizers provide a suitable system to
investigate the effects of the conjugated spacer and donor on the
short-circuit current (Jsc) of the corresponding DSSCs. The measured
Jsc values are well rational with the quantity of the direct electron
injection derived from the theoretical calculations. Our study shows
that extending the length of the conjugated spacer may red-shift the
lmax and increase the absorption coefficient of the dye molecule,
however, theoretical calculations are also efficiently rationalized
with the charge transfer from dye to (TiO2)38 cluster decreases upon
increasing the length of the conjugated spacer. The electron injec-
tion of W-1 and W-4 dyes toTiO2 follows both the direct and indirect
mechanisms; on the other hand, W-2 mainly follows an indirect
electron injection mechanism and thus has a smaller Jsc value. The
distribution of the frontier orbitals of the W-series dyes is also very
sensitive to the donor moiety; the calculation data show that TPA
group is a better donor than carbazole to drive the electron density
of the sensitizers to TiO2 directly. Therefore W-4 dye (with TPA as a
donor) sensitized DSSC has the highest efficiency (5.0%) among the
dyes studied in this article. The efficiency is w75% of the N719 based
(b) Haque SA, Handa S, Peter K, Palomares E, Thelakkat M, Durrant JR.
Supermolecular control of charge transfer in dye-sensitized nanocrystalline
TiO2 films: towards a quantitative structureefunction relationship. Angew
Chem Int Ed 2005;44:5740e4.
[9] Hara K, Dan-oh Y, Kasada C, Ohga Y, Shinpo A, Suga S, et al. Effect of additives
on the photovoltaic performance of coumarin-dye-sensitized nanocrystalline
TiO2 solar cells. Langmuir 2004;20:4205e10.
[10] Ehret A, Stuhl L, Spitler MT. Spectral sensitization of TiO2 nanocrystalline
electrodes with aggregated cyanine dyes. J Phys Chem B 2001;105:9960e5.
[11] Wang ZS, Li FY, Huang CH. Highly efficient sensitization of nanocrystalline
TiO2 films with styryl benzothiazolium propylsulfonate. Chem Commun 2000:
2063e4.
[12] Satoh N, Nakashima T, Yamamoto K. Metal-assembling dendrimers with a
triarylamine core and their application to a dye-sensitized solar cell. J Am
Chem Soc 2005;127:13030e8.
[13] Thomas KRJ, Hsu YC, Lin JT, Lee KM, Ho KC, Lai CH, et al. 2,3-Disubstituted
thiophene based organic dyes for solar cells. Chem Mater 2008;20:1830e40.
[14] Kitamura T, Ikeda M, Shigaki K, Inoue T, Anderson NA, Ai X, et al. Phenyl-
conjugated oligoene sensitizers for TiO2 solar cells. Chem Mater 2004;16:
1806e12.
[15] Campbell WM, Jolley KW, Wagner P, Wagner K, Walsh PJ, Gordon KC, et al.
Highly efficient porphyrin sensitizers for dye-sensitized solar cells. J Phys
Chem C 2007;111:11760e2.
[16] Shibano Y, Umeyama T, Matano Y, Imahori H. Electron-donating perylene
tetracarboxylic acids for dye-sensitized solar cells. Org Lett 2007;9:1971e4.
[17] Zeng W, Cao Y, Bai Y, Wang Y, Shi Y, Zhang M, et al. Efficient dye-sensitized
solar cells with an organic photosensitizer featuring orderly conjugated
ethylenedioxythiophene and dithienosilole blocks. Chem Mater 2010;22:
1915e25.
[18] Mishra A, Fischer MKR, Bauerle P. Metal-free organic dyes for dye-sensitized
solar cells: from structure: property relationships to design rules. Angew
Chem Int Ed 2009;48:2474e99.
[19] Pastore M, Mosconi E, De Angelis F, Gratzel M. A computational investigation
of organic dyes for dye-sensitized solar cells: benchmark, strategies, and open
issues. J Phys Chem C 2010;114:7205e12.
device (
h
¼ 6.7%) fabricated in the similar conditions. The calculated
electron density difference maps (EDDMs) before and after photo
excitation provide a better understanding of the photovoltaic per-
formance of cyclopentadithiophene-containing sensitizers at the
electronic and structural levels.
Acknowledgments
[20] Yen YS, Hsu YC, Lin JT, Chang CW, Hsu CP, Yin DJ. Pyrrole-based organic dyes
for dye-sensitized solar cells. J Phys Chem C 2008;112:12557e67.
[21] Jacquemin D, Perpete EA, Scuseria GE, Ciofini I, Adamo C. TD-DFT performance
for the visible absorption spectra of organic dyes: conventional versus long-
range hybrids. J Chem Theor Comput 2008;4:123e35.
[22] Edvinsson T, Li C, Pschirer N, Schoneboom J, Eickemeyer F, Sens R, et al.
Intramolecular charge-transfer tuning of perylenes: spectroscopic features
and performance in dye-sensitized solar cells. J Phys Chem C 2007;111:
15137e40.
[23] Chao WS, Liao KH, Chen CT, Huang WK, Lan CM, Diau EW. Spirally configured
cis-stilbene/fluorene hybrids as bipolar, organic sensitizers for solar cell ap-
plications. Chem Commun 2012;48:4884e6.
[24] Ma RM, Guo P, Yang LL, Guo LS, Zhang XX, Nazeeruddin MK, et al.
A theoretical interpretation and screening of porphyrin sensitizer candidates
with anticipated good photo-to-electric conversion performances for dye-
sensitized solar cells. J Phys Chem A 2010;114:1973e9.
The financial support provided by the National Science Council
and ITRI, Taiwan, ROC is gratefully acknowledged. We thank the
National Center for High-Performance Computing and the V’ger
computer cluster (Contribution Number: NCU-CCG100-0012) at the
National Center University of Taiwan for allowing access to the
computer time and facilities.
Appendix A. Supplementary material
Details of the experiment, including the synthesis and charac-
terization of all intermediates, IR, NMR spectra of the W-series
dyes, as well as the fabrication and photovoltaic performance
measurement of DSSCs can be found, in the online version, at
http://dx.doi.org/10.1016/j.dyepig.2013.08.016
[25] Yakhanthip T, Jungsuttiwong S, Namuangruk S, Kungwan N, Promarak V,
Sudyoadsuk T, et al. Theoretical investigation of novel carbazole-fluorene
based DepeA conjugated organic dyes as dye-sensitizer in dye-sensitized
solar cells (DSCs). J Comput Chem 2011;32:1568e76.
[26] Pastore M, Angelis FD. Aggregation of organic dyes on TiO2 in dye-sensitized
solar cells models: an ab initio investigation. ACS Nano 2010;4:556e62.
[27] De Angelis F, Fantacci S, Sgamellotti A. An integrated computational tool for
the study of the optical properties of nanoscale devices: application to solar
cells and molecular wires. Theor Chem Acc 2007;117:1093e104.
References
[1] Markvart T. Solar electricity. 2nd ed. Chichester: Wiley; 2000.
[2] O’Regan B, Grätzel M. A low-cost, high-efficiency solar cell based on dye-
sensitized colloidal TiO2 films. Nature 1991;353:737e40.
[28] De Angelis F, Fantacci S, Selloni A, Nazeeruddin MK, Gratzel M. Time-
dependent density functional theory investigations on the excited states of

1100
C.-G. Wu et al. / Dyes and Pigments 99 (2013) 1091e1100
Ru(II)-dye-sensitized TiO2 nanoparticles: the role of sensitizer protonation.
J Am Chem Soc 2007;129:14156e7.
[45] Planells M, Pelleja L, Clifford JN, Pastore M, De Angelis F, Lopez N, et al. Energy
levels, charge injection, charge recombination and dye regeneration dynamics
for donoreacceptor pi-conjugated organic dyes in mesoscopic TiO2 sensitized
solar cells. Energy Environ Sci 2011;4:1820e9.
[46] Vittadini A, Selloni A, Rotzinger FP, Grätzel M. Formic acid adsorption on dry
and hydrated TiO2 anatase (101) surfaces by DFT calculations. J Phys Chem B
2000;104:1300e6.
[29] De Angelis F, Tilocca A, Selloni A. Time-dependent DFT study of [Fe(CN)6] 4-
sensitization of TiO2 nanoparticles. J Am Chem Soc 2004;126:15024e5.
[30] Zhang M, Tsao HN, Pisula W, Yang C, Mishra AK, Mullen K. Field-effect tran-
sistors based on a benzothiadiazole-cyclopentadithiophene copolymer. J Am
Chem Soc 2007;129:3472e3.
[31] Tsao HN, Yi C, Moehl T, Yum JH, Zakeeruddin SM, Nazeeruddin MK, et al.
Cyclopentadithiophene bridged donoreacceptor dyes achieve high power
conversion efficiencies in dye-sensitized solar cells based on the tris-cobalt
bipyridine redox couple. ChemSusChem 2011;4:591e4.
[47] Nazeeruddin MK, Humphry-Baker R, Liska P, Gratzel M. Investigation of
sensitizer adsorption and the influence of protons on current and voltage of a
dye-sensitized nanocrystalline TiO2 solar cell. J Phys Chem B 2003;107:8981e
7.
[32] Bijleveld JC, Shahid M, Gilot J, Wienk MM, Janssen RAJ. Copolymers of
cyclopentadithiophene and electron-deficient aromatic units designed for
photovoltaic applications. Adv Funct Mater 2009;19:3262e70.
[48] Stille JK. The palladium-catalyzed cross-coupling reactions of organotin re-
agents with organic electrophiles [New Synthetic methods (58)]. Angew
Chem Int Ed 1986;25:508e24.
[33] Tsao HN, Cho DM, Park I, Hansen MR, Mavrinskiy A, Yoon Do Y, et al. Ultrahigh
[49] (a) Peyratout C, Daehne L. Aggregation of thiacyanine derivatives on poly-
electrolytes. Phys Chem Chem Phys 2002;4:3032;
mobility in polymer field-effect transistors by design.
2011;133:2605e12.
J Am Chem Soc
(b) Wang ZS, Hara K, Dan-oh Y, Kasada C, Shinpo A, Suga S, et al. Photophysical
and (photo)electrochemical properties of a coumarin dye. J Phys Chem B
2005;109:3907e14.
[34] Li Z, Tsang SW, Du XM, Scoles L, Robertson G, Zhang YG, et al. Alternating
copolymers of cyclopenta[2,1-b;3,4-b0]dithiophene and thieno[3,4-c] pyrrole-
4,6-dione for high-performance polymer solar cells. Adv Funct Mater
2011;21:3331e6.
[50] Hagfeldt A, Grätzel M. Light-induced redox reactions in nanocrystalline sys-
tems. Chem Rev 1995;95:49e68.
[35] Zhang Y, Zou JY, Yip HL, Sun Y, Davies JA, Chen KS, et al. Conjugated polymers
based on C, Si and N-bridged dithiophene and thienopyrroledione units:
synthesis, field-effect transistors and bulk heterojunction polymer solar cells.
J Mater Chem 2011;21:3895e902.
[36] Li JY, Chen CY, Ho WC, Chen SH, Wu CG. Unsymmetrical squaraines incor-
porating quinoline for near infrared responsive dye-sensitized solar cells. Org
Lett 2012;14:5420e3.
[51] Chen CY, Chen JG, Wu SJ, Li JY, Wu CG, Ho KC. Multi-functionalized ruthenium
based super-sensitizers for high efficient dye-sensitized solar cells. Angew
Chem Int Ed 2008;47:7342e5.
[52] Edvinsson T, Li C, Pschirer N, Schöneboom J, Eickemeyer F, Sens R, et al.
Intramolecular charge-transfer tuning of perylenes: spectroscopic features
and performance in dye-sensitized solar cells. J Phys Chem C 2007;111:
15137e40.
[37] Frisch MJT, Schlegel HB, Scuseria GE, Robb MA, Cheeseman JR, Scalmani G,
et al. Gaussian, Inc., Wallingford, CT; 2009.
[38] Cossi M, Rega N, Scalmani G, Barone V. Energies, structures, and electronic
properties of molecules in solution with the C-PCM solvation model. J Comput
Chem 2003;24:669e81.
[39] Yanai T, Tew DP, Handy NC. A new hybrid exchange-correlation functional
using the Coulomb-attenuating method (CAM-B3LYP). Chem Phys Lett
2004;393:51e7.
[40] O’Boyle NM, Tenderholt AL, Langner KM. A library for package-independent
computational chemistry algorithms. J Comput Chem 2008;29:839e45.
[41] Car R, Parrinello M. Unified approach for molecular dynamics and density-
functional theory. Phys Rev Lett 1985;55:2471e4.
[53] Hagberg DP, Marinado T, Karlsson KM, Nonomura K, Qin P, Boschloo G, et al.
Tuning the HOMO and LUMO energy levels of organic chromophores for dye
sensitized solar cells. J Org Chem 2007;72:9550e6.
[54] Tian H, Yang X, Chen R, Pan Y, Li L, Hagfeldt A, et al. Phenothiazine de-
rivatives for efficient organic dye-sensitized solar cells. Chem Commun
2007:3741e3.
[55] Wu CG, Chung MF, Tsai HHG, Tan CJ, Chen SC, Chang CH, et al. Fluorene
containing organic photosensitizers for dye sensitized solar cell. Chem-
PlusChem 2012;77:832e43.
[56] Persson P, Bergström R, Lunell S. Quantum chemical study of photoinjection
processes in dye-sensitized TiO2 nanoparticles. J Phys Chem B 2000;104:
10348e51.
[42] Hutter J, Alavi A, Deutsch T, Bernasconi M, Goedecker S, Marx DT, et al. MPI fur
Festkorperforschung and IBM Research Laboratory, Stuttgart and Zurich;
1995e2000.
[57] Macyk W, Szacilowski K, Stochel G, Buchalska M, Kuncewicz J, Labuz P.
Titanium(IV) complexes as direct TiO2 photosensitizers. Coord Chem Rev
2010;254:2687e701.
[43] Perdew JP, Burke K, Ernzerhof M. Generalized gradient approximation made
simple. Phys Rev Lett 1996;77:3865e8.
[58] Barnes PRF, Anderson AY, Koops SE, Durrant JR, O’Regan BC. Electron injection
efficiency and diffusion length in dye-sensitized solar cells derived from
[44] Vanderbilt D. Soft self-consistent pseudopotentials in a generalized eigen-
value formalism. Phys Rev B 1990;41:7892e5.
incident photon conversion efficiency measurements.
2009;113:1126e36.
J Phys Chem C