Donor-acceptor dyes incorporating a stable dibenzosilole π-conjugated spacer for dye-sensitized solar cells

Article abstract of DOI:10.1039/c2jm30978e

Four novel organic dyes including three based on dibenzosilole (YS01-03) and one based on fluorene (YS04) were synthesized, and their photophysical properties and dye-sensitized solar cell (DSC) performances were characterized. The silicon-containing dibenzosilole-based dyes (YS01-03) were superior to the carbon analogue fluorene-based dye YS04 in incident-photon-to-current conversion efficiency (IPCE), and total solar-to-electric conversion efficiency (η), with YS03, which has the bulkiest and most branched electron donor group, achieving the highest η of 5.07% compared to 2.88% of YS04. To better understand how silicon influences the excited state oxidation potentials (S ^+/*) and absorption maxima (λ_max), the equilibrium molecular geometries of dyes YS01-04 were calculated using density functional theory (DFT) utilizing B3LYP energy functional and DGDZVP basis set. It was shown that the torsion angles (1 and 2) across the biphenyl linkages of dyes containing silicon (YS01-03) were less twisted than that of the silicon-free dye (YS04), which enhanced the π-π* overlap, and that translated into photocurrent enhancements in the silicon-containing dyes YS01-03. Moreover, the vertical electronic excitations and S^+/* of dyes YS01-04 were studied using different long-range corrected time-dependent DFT methods, including CAM-B3LYP, LC-BLYP, WB97XD, and LC-wPBE at the basis set level DGDZVP. Excellent agreement between the calculated, using CAM-B3LYP/DGDZVP, and experimental results was found. The Royal Society of Chemistry 2012.

Full text of DOI:10.1039/c2jm30978e

View Article Online / Journal Homepage / Table of Contents for this issue
Dynamic Article Links <
C
Journal of
Materials Chemistry
Cite this: J. Mater. Chem., 2012, 22, 10771
www.rsc.org/materials
PAPER
Donor–acceptor dyes incorporating a stable dibenzosilole p-conjugated
spacer for dye-sensitized solar cells†
a
a
b
c
Md. Akhtaruzzaman, Yohei Seya, Naoki Asao, Ashraful Islam, Eunsang Kwon,d Ahmed El-Shafei,e
*
*
c
Liyuan Han and Yoshinori Yamamoto
bf
*
Received 16th February 2012, Accepted 26th March 2012
DOI: 10.1039/c2jm30978e
Four novel organic dyes including three based on dibenzosilole (YS01–03) and one based on fluorene
(YS04) were synthesized, and their photophysical properties and dye-sensitized solar cell (DSC)
performances were characterized. The silicon-containing dibenzosilole-based dyes (YS01–03) were
superior to the carbon analogue fluorene-based dye YS04 in incident-photon-to-current conversion
efficiency (IPCE), and total solar-to-electric conversion efficiency (h), with YS03, which has the bulkiest
and most branched electron donor group, achieving the highest h of 5.07% compared to 2.88% of YS04.
To better understand how silicon influences the excited state oxidation potentials (S+/*) and absorption
maxima (lmax), the equilibrium molecular geometries of dyes YS01–04 were calculated using density
functional theory (DFT) utilizing B3LYP energy functional and DGDZVP basis set. It was shown that
the torsion angles (q1 and q2) across the biphenyl linkages of dyes containing silicon (YS01–03) were
less twisted than that of the silicon-free dye (YS04), which enhanced the p–p* overlap, and that
translated into photocurrent enhancements in the silicon-containing dyes YS01–03. Moreover, the
vertical electronic excitations and S+/* of dyes YS01–04 were studied using different long-range
corrected time-dependent DFT methods, including CAM-B3LYP, LC-BLYP, WB97XD, and
LC-wPBE at the basis set level DGDZVP. Excellent agreement between the calculated, using
CAM-B3LYP/DGDZVP, and experimental results was found.
low costs.1 Currently, the most popular design of organic sensi-
Introduction
tizers is based on D–p–A systems, in which the electron-donating
(D) and accepting (A) moieties are linked by p-conjugated
spacers such as polyene, oligophenylenevinylene, oligothiophene
and their derivatives.2 The overall DSC performances of these
organic sensitizers are lower than Ru-based dyes because they do
not absorb across a wide range of the solar radiation spectrum.3
Furthermore, they have inferior photostability and tend to
aggregate on the semiconductor TiO2 surface, which precludes
efficient electron injection into the semiconductor.4 It was
reported that the E0–0 energy gap of dyes for DSCs is one of the
key thermodynamic parameters that can be used effectively to
significantly increase the efficiency of DSCs up to 33% if and only
if the ground-state oxidation potential, S+/0, and excited-state
oxidation potential, S+/*, were maintained thermodynamically
favorable for replenishing the hole and electron injection,
respectively, while maintaining the E0–0 energy gap ꢀ1.4 eV.5
Therefore, proper design of organic sensitizers with small
E0–0 energy-gap that extends to the NIR region up to 880 nm
while maintaining the S+/0 and S+/* thermodynamically favorable
is highly desirable for furnishing panchromatic sensitizers with
efficient electron injection and dye regeneration.6 Hence, the
systematic molecular design of the D and A moieties along with
the p-conjugation linker is paramount for achieving the enviable
E0–0 energy-gap and the energy locations of the S+/0 and S+/* that
In recent years, the molecular design and realization of metal-
free organic sensitizers have attracted considerable attention in
the academia and industry for applications in dye-sensitized solar
cells (DSCs), owing to their high molar extinction coefficients,
flexible molecular structural modulations, facile synthesis, and
aDepartment of Chemistry, Graduate School of Science, Tohoku
University, Aramaki-Azaaoba 6-3, Aoba-ku, Sendai, 980-8578, Japan.
E-mail: akhtar@m.tohoku.ac.jp; Fax: +81(0)22-217-6165; Tel: +81(0)
22-217-6177
bWPI-Advanced Institute for Materials Research (WPI-AIMR), Tohoku
University, Katahira 2-1-1, Aobaku, Sendai 980-8577, Japan. E-mail:
yoshi@m.tohoku.ac.jp; Fax: +81-22-217-5129; Tel: +81-22-217-5130
cPhotovoltaic Materials Unit, National Institute for Materials Science, 1-2-
1 Sengen, Tsukuba, Ibaraki 305-0047, Japan. E-mail: ISLAM.Ashraful@
nims.go.jp; Fax: +81(0)29 859 2301; Tel: +81(0)29 859 2129
dResearch and Analytical Center for Giant Molecules, Graduate School of
Science, Tohoku University, 6-3 Aramaki-AzaAoba, Japan 980-8578
ePolymer and Color Chemistry Program, North Carolina State University,
1000 Main Campus Dr. Raleigh, NC 27695, USA
fChemistry Department, Faculty of Science, and Center of Excellence for
Advanced Materials Research, King Abdulaziz University, P.O. Box
80203, Jeddah, Saudi Arabia
† Electronic supplementary information (ESI) available: Experimental
procedures for synthesis, ionization potential and all fundamental data
of DSCs. See DOI: 10.1039/c2jm30978e
This journal is ª The Royal Society of Chemistry 2012
J. Mater. Chem., 2012, 22, 10771–10778 | 10771

View Article Online
lead to sensitizers that can harvest energetic photons across
a wide range of the solar spectrum, and the desirable molecular
geometry that precludes dye aggregations.
spectra were obtained on a BRUKER APEXIII spectrometer.
Column chromatography was carried out employing Silica Gel
60N (spherical, neutral, 40–100 mm, KANTO Chemical Co.).
Analytical thin-layer chromatography (TLC) was performed on
0.2 mm precoated plate Kieselgel 60 F254 (Merck). All other
reagents and solvents commercially available were used without
further purification unless otherwise noted.
Silicon-based dyes have been considered to have very prom-
ising chemistry for DSCs due to their photostability and thermal
stability, and the ability to effectively tune their HOMO–LUMO
energy levels.7 Polymers and copolymers containing a silole unit
were recently investigated for organic photovoltaics (OPV),
organic field-effect transistors (OFETs) and sensors because of
their low-lying LUMO arising from the strong s*–p* conjuga-
tion between the s* orbital of the exocyclic two Si–C bonds and
the p*orbital of the butadiene moiety.8 However, the silole
chemistry has not been investigated extensively for the molecular
design of organic sensitizers for DSC applications. From the
open literature, very few examples have been found based on
dithienosilole heterocyclic aromatic motifs with an electron
donor at one end of the molecule and an electron acceptor at the
other end for DSCs application.9 In the present paper, novel
D–p–A sensitizers (Fig. 1) based on silole chemistry were
successfully designed and synthesized, containing indolized
analogs as the electron donor, and cyanoacrylic acid as the
electron acceptor moiety, and both are connected with a diben-
zosilole p-conjugated unit to furnish dyes YS01–03, to study and
better understand the photophysical and photoelectrochemical
properties of silole-containing dyes for DSCs. YS01 had the
simplest structure, with methyl substituted indoline as the elec-
tron donor. Two different indoline derivatives, methoxy in YS02
and di-tert-butyl in YS03, were also introduced to further
increase the electron-donating strength of the donor moiety and
to generate the bulkiest group for precluding dye aggregations,
respectively. To study the influence of non-silicon containing
sensitizers on photovoltaic performance in DSCs, dye YS04,
a fluorene-based one was synthesized. The photophysical prop-
erties, equilibrium molecular geometry calculations and the
performance of DSCs based on these organic dyes are reported.
Synthesis of compounds YS01–04
To a mixture of intermediates 8a–c, 11 (0.14 mmol), 2-cyano-
acetic acid (0.448 mmol) and ammonium acetate (0.08 mmol) in
acetonitrile (5 mL) was added 10 mL of glacial acetic acid, and
the mixture was refluxed for 3 h under Ar. After evaporation of
the solvent, the crude solid was dissolved into CH2Cl2 and
washed with water. The organic layer was dried over Na2SO4.
After removal of the solvent under reduced pressure, the residue
was purified by silica-gel column chromatography to give
compounds YS01–04.
Dye YS01. Yield 73%; 1H NMR (400 MHz, DMSO-d6, d) 8.42
(s, 1H), 8.20 (d, J ¼ 8.4 Hz, 2H), 8.14 (d, J ¼ 1.6 Hz, 1H), 8.09–
7.93 (m, 5H), 7.88 (dd, J ¼ 8.0 Hz, J ¼ 1.6 Hz, 1H), 7.69 (dd, J ¼
8.4 Hz, J ¼ 1.6 Hz, 1H), 7.53 (brs, 1H), 7.41 (dd, J ¼ 8.4 Hz, J ¼
1.6 Hz, 1H), 7.25 (d, J ¼ 8.8 Hz, 4H), 7.21 (d, J ¼ 8.8 Hz, 2H),
6.96 (d, J ¼ 8.4 Hz, 1H), 4.94–4.85 (m, 1H), 3.94–3.86 (m, 1H),
2.32 (s, 3H), 2.18–1.61 (m, 6H), 1.45–1.14 (m, 12H), 1.12–0.97
(m, 4H), 0.85–0.69 (m, 6H); 13C NMR (100 MHz, DMSO-d6, d)
163.2, 153.5, 148.1, 146.6, 145.0, 144.3, 139.7, 139.5, 138.22,
138.20, 136.6, 135.4, 131.5, 131.3, 130.26, 130.22, 130.17, 129.96,
128.9, 127.6, 126.9, 125.6, 122.7, 121.7, 121.3, 119.0, 116.2, 107.2,
102.9, 68.2, 44.7, 34.79, 34.77, 33.2, 24.0, 23.1, 21.5, 20.4,13.8,
11.6. HRMS (ESI positive): [M + H]+ calcd for C50H53N2O2Si,
741.3871; found, 741.3867%.
Dye YS02. Yield 85%; 1H NMR (400 MHz, CDCl3, d) 8.33 (s,
1H), 8.12 (d, J ¼ 8.4 Hz, 2H), 7.95–7.77 (m, 6H), 7.71 (d, J ¼ 7.6
Hz, 1H), 7.63 (d, J ¼ 7.6 Hz, 1H), 7.41 (s, 1H), 7.32 (d, J ¼ 8.0
Hz, 1H), 7.25 (d, J ¼ 8.8 Hz, 2H), 6.93 (d, J ¼ 8.8 Hz, 2H), 6.81
(d, J ¼ 8.0 Hz, 1H), 4.83–4.74 (m, 1H), 3.95–3.86 (m, 1H), 3.83 (s,
3H), 2.15–1.56 (m, 6H), 1.49–1.19 (m, 12H), 1.05–0.95 (m, 4H),
0.87–0.76 (m, 6H); 13C NMR (100 MHz, DMSO, d) 163.2, 154.5,
153.2, 148.1, 147.7, 144.9, 144.2, 139.6, 138.20, 138.17, 136.6,
135.4, 135.0, 131.5, 131.2, 130.3, 130.2, 129.5, 128.9, 127.5, 126.9,
125.6, 122.7, 121.9, 121.7, 121.3, 116.4, 114.5, 106.4, 103.4, 68.8,
55.2, 44.7, 34.9, 34.8, 33.1, 24.0, 23.1, 21.5, 13.8, 11.6. ESI-MS:
m/z calcd for [M (C50H52N2O3Si) ꢁ H]: 755.37; found, 755.1%.
Experimental section
General information for materials synthesis
1H NMR and 13C NMR spectra were recorded on JEOL JMTC-
270/54/SS (JASTEC, 400 MHz) and BRUKER (600 MHz)
1
spectrometers. H NMR spectra are reported as follows: chem-
ical shift in ppm (d), integration, multiplicities (s ¼ singlet, d ¼
doublet and m ¼ multiplet), and coupling constants (Hz). 13C
NMR spectra are reported in ppm (d). High-resolution mass
Dye YS03. Yield 81%; 1H NMR (400 MHz, DMSO-d6, d) 8.09
(brs, 1H), 8.07–7.93 (m, 6H), 7.89 (d, J ¼ 8.4 Hz, 2H), 7.81 (d,
J ¼ 8.0 Hz, 1H), 7.69 (d, J ¼ 8.0 Hz, 1H), 7.52 (brs, 1H), 7.46 (d,
J ¼ 8.4 Hz, 1H), 7.18 (brs, 1H), 6.93 (d, J ¼ 8.4 Hz, 1H), 4.88–
4.82 (m, 1H), 3.92–3.85 (m, 1H), 3.69 (s, 3H), 2.20–1.15 (m, 18H),
1.45 (s, 18H), 1.10–1.00 (m, 4H), 7.83–0.71 (m, 6H). FAB-MS
(EI positive): m/z calcd for C58H68N2O3Si, 868.5; found, 868.5%.
Dye YS04. Yield 83%; 1H NMR (400 MHz, DMSO-d6, d) 8.42
(s, 1H), 8.21 (d, J ¼ 8.0 Hz, 2H), 8.05 (d, J ¼ 8.4 Hz, 2H), 7.97–
7.92 (m, 2H), 7.89 (d, J ¼ 7.6 Hz, 1H), 7.83 (d, J ¼ 8.0 Hz, 1H),
Fig. 1 Chemical structures of dyes YS01–04 where q1 and q2 refer to the
torsion angles across the biphenyl linkages.
10772 | J. Mater. Chem., 2012, 22, 10771–10778
This journal is ª The Royal Society of Chemistry 2012

View Article Online
7.72 (s, 1H), 7.63 (d, J ¼ 8.0 Hz, 1H), 7.57 (s, 1H), 7.45 (d, J ¼
8.8 Hz, 1H), 7.26 (d, J ¼ 8.0 Hz, 2H), 7.22 (d, J ¼ 8.0 Hz, 2H),
6.98 (d, J ¼ 8.4 Hz, 1H), 4.94–4.88 (m, 1H), 3.97–3.89 (m, 1H),
2.33 (s, 3H), 2.20–1.36 (m, 10H), 1.13–0.99 (m, 8H), 0.73–0.55
(m, 10H); 13C NMR (100 MHz, DMSO-d6, d) 163.3, 153.5, 151.3,
146.6, 144.6, 141.1, 139.8, 137.9, 136.8, 135.4, 131.3, 130.4, 130.3,
130.1, 129.6, 129.0, 127.1, 126.9, 125.9, 125.8, 124.6, 122.9, 121.1,
120.5, 120.1, 119.8, 119.1, 116.3, 107.3, 103.0, 68.2, 54.9, 44.7,
34.8, 33.3, 31.51, 31.5, 24.0, 23.2, 21.7, 20.4, 13.8. HRMS (ESI
positive): [M + H]+ calcd for C51H53N2O2, 725.4107; found,
725.4084%.
DFT energy functional methods including CAM-B3LYP,19 LC-
BLYP,20 WB97XD,21 or LC-wPBE22 with the basis set
DGDZVP, and the electronic absorption spectra were extracted.
The MM3 calculations were performed on a desktop computer
equipped with four processors at 2.6 GHz and 8 GB of RAM.
The TD-DFT calculations were performed in DMF using the
polarizable conductor calculation model (PCM). The DFT
calculations were performed on the East Carolina University’s
Super Computer Jasta using 8 processors and 4 GB of RAM.
The solvent effect was accounted for using the polarizable
conductor calculation model (PCM), implemented in Gaussian
09. All calculations were performed using Gaussian 09.
Molecular modelling
Fabrication of dye sensitized solar cells
The ground state equilibrium molecular geometry of dye YS01
was calculated by generating an energy search map in molecular
mechanics using augmented MM3 parameters implemented in
Scigress Workspace Version 7.7.0.47. To generate the energy
search map, a conformation search was performed using MM3
parameters by varying the torsion angles across the two
peripheral biphenyl linkages (q1 and q2) from ꢁ180ꢂ to +180ꢂ
using 120 steps of 3ꢂ each, which furnished a potential energy
map of a total of 14 641 conformers.
A double-layer TiO2 photoelectrode (10 + 5) mm in thickness
with a 10 mm thick nanoporous layer and a 5 mm thick scattering
layer (area: 0.25 cm2) was prepared by screen printing on a con-
ducting glass substrate. A dye solution of 3 ꢃ 10ꢁ4 M concen-
tration in acetonitrile–tert-butyl alcohol (1/1, v/v) was used to
uptake the dye onto the TiO2 film. Deoxycholic acid (DCA)
(20 mM) as a co-adsorbent was added into the dye solution to
prevent aggregation of the dye molecules. The TiO2 films were
immersed into the dye solution and then kept at 25 ꢂC for 30 h.
Photovoltaic measurements were performed in a sandwich type
solar cell in conjunction with an electrolyte consisting of a solu-
tion of 0.6 M dimethylpropyl-imidazolium iodide (DMPII), 0.05
M I2, 0.1 M LiI and 0.5 M tert-butylpyridine (TBP) in acetoni-
trile (AN). The dye-deposited TiO2 film and a platinum-coated
Ground state equilibrium molecular geometries of YS01–04
were calculated using the hybrid DFT energy functional
B3LYP15,16 and the basis set 3-21G* or DGDZVP (density Gauss
double-z with polarization functions).17,18 Subsequently, a single
point energy calculation was performed on the optimized
geometry using different long-range corrected time-dependent
Scheme 1 Synthesis routes of dyes YS01–03.
Scheme 2 Synthesis routes of dye YS04.
This journal is ª The Royal Society of Chemistry 2012
J. Mater. Chem., 2012, 22, 10771–10778 | 10773

View Article Online
conducting glass were separated by a Surlyn spacer (40 mm thick)
and sealed by heating the polymer frame. Photocurrent density–
voltage (I–V) of sealed solar cells was measured under AM 1.5G
simulated solar light at a light intensity of 100 mW cmꢁ2 with
a metal mask of 0.25 cm2. The photovoltaic parameters, i.e. short
circuit current (Jsc), open circuit voltage (Voc), fill factor (FF),
and power conversion efficiency (h) were estimated from I–V
characteristics under illumination.
transfer from D to A in silicon-based dyes than that of the fluo-
rene-based dye (YS04). Moreover, the anodic shift increased the
negative free energy of electron injection, which explained why
YS01–03 furnished more photocurrent than YS04. The onset of
the optical energy gap (E0–0) of YS01–04 dyes was in the range of
2.43–2.53 eV (Table 1).13 The excited-state oxidation potential,
S+/*, of sensitizers YS01–04 was in the range of ꢁ3.15 to ꢁ3.18 eV
(Table 1) and lay above the conduction band edge (ꢁ4.2 eV)1c of
the nanocrystalline TiO2 ensuring that the excited states of these
dyes are thermodynamically driven for efficient electron injection
into the conduction band edge of TiO2.
Results and discussions
A key factor in predicting accurately the electronic and ther-
modynamic properties of any compound is to obtain first the
correct molecular geometry. Hence, to gain more insights and
The synthesis routes of compounds YS01–04 are shown in
Schemes 1 and 2. The intermediate compounds 2a, 2b, 5 and 10
were synthesized according to the published methods.10,11 Other
intermediate compounds (2–8, 10, 11) were obtained in high
yields following standard reaction conditions as described in the
ESI†. The target compounds YS01–04 were prepared via
condensation of their corresponding aldehyde derivatives and
cyanoacetic acid in the presence of ammonium acetate in high
yields as described in the Experimental section.
The photophysical properties of dyes YS01–04 are summa-
rized in Table 1. Their wavelengths of absorption maxima (lmax
)
were in the range of 385–391 nm, shown in Fig. 2a. The low-lying
absorption bands, due to the p–p* transition, for YS01, YS02
YS03, and YS04 exhibited lmax at 386, 390, 391, and 385 nm,
respectively. When absorbed on a transparent thin TiO2 film, all
dyes exhibited broad absorption spectra similar to those in
solution, due to the interaction between the carboxylate group
and TiO2 (Fig. 2b). Upon adsorption on the TiO2 film, the tail of
the absorption spectrum of all dyes extended up to 550 nm, which
is highly desirable for harvesting more of the solar spectrum and
translates into greater photocurrent.
The ionization potential (IP) of YS01–04 bound to the nano-
crystalline TiO2 film was measured using the photoemission yield
spectrometer (Riken Keiki, AC-3E). The IP values of YS01–04
were found in the range of ꢁ5.59 to ꢁ5.68 eV (Fig. S2, ESI†). The
IP values of YS01–04, which correspond to S+/0, were sufficiently
ꢁ
lower than the redox potential of Iꢁ/I3 (ꢁ5.20 eV),12 which
ensured favorable thermodynamic ground states for efficient
regeneration of the dye through the reaction of the oxidized dye
with iodide. The IP value of the silyl containing sensitizers
(YS01–03) was shifted anodically compared to the fluorene-based
dye (YS04) (Table 1). This anodic shift is because silicon is more
electropositive than carbon, exhibiting more electron donation
relative to carbon (b-silicon effect), which furnished more charge
Fig. 2 Absorption spectra of YS01–04 (a) in DMF and (b) on a trans-
parent TiO2 film.
Table 1 Photophysical properties and DSCs performance parameters of dyes YS01–04
Dye
lmax/nma (3 ꢃ 104/Mꢁ1 cmꢁ1
)
lmax/nmb on TiO2
IPc (eV)
E0–0 (eV)d
S+/* (eV)e
Jsc (mA cmꢁ2
)
Voc (V)
FF
h (%)
YS01
YS02
YS03
YS04
386(4.5)
390(4.5)
391(4.8)
385(5.4)
382
379
392
396
ꢁ5.66
ꢁ5.60
ꢁ5.59
ꢁ5.68
2.48
2.46
2.43
2.53
ꢁ3.18
ꢁ3.15
ꢁ3.16
ꢁ3.15
8.23
7.69
9.14
5.10
0.767
0.798
0.778
0.772
0.735
0.740
0.712
0.732
4.64
4.55
5.06
2.88
a
Absorption measured in DMF at room temperature. b Absorption measured on a transparent 4 mm TiO2 film. c Ionization potential (IP) of absorbed
dyes on the nanocrystalline TiO2 film was determined by using the photoemission yield spectrometer (Riken Keiki, AC-3E). d E0–0 was estimated from
the absorption onset of the dye loaded on a TiO2 film. e The excited-state oxidation potential (S+/*) ¼ IP ꢁ E0–0. Measurements were performed under
AM 1.5 irradiation on the DSC devices with 0.25 cm2 active surface area defined by a metal mask. Jsc, short circuit current; Voc, open circuit voltage; FF,
fill factor; h, conversion efficiency.
10774 | J. Mater. Chem., 2012, 22, 10771–10778
This journal is ª The Royal Society of Chemistry 2012

View Article Online
Table 2 Calculated absorption spectra for YS01 using different starting
optimized geometries and long range corrected DFT methods
lmax (nm)
Long-range corrected energy functional/basis set
Geometry CAM-B3LYP/ LC-BLYP/ WB97XD/ LC-wPBE/
optimization DGDZVP DGDZVP DGDZVP DGDZVP Expt.
B3LYP/
3-21G*
B3LYP/
DGDZVP
371
376
328
332
355
359
329
333
386
Fig. 3 Potential energy map generated by varying two torsion angles.
The black ball at the bottom of the map represents the global energy
minimum conformer.
Table 2 shows a comparison between the calculated and
experimental lmax for YS01. These data clearly demonstrate that
the geometry optimization of YS01 using the energy functional
B3LYP and the basis set DGDZVP, coupled with TD-DFT
CAM-B3LYP energy functional and the basis set DGDZVP
furnished the closest lmax (376 nm) to the experimental result
(386 nm). These findings are vital because they clearly demon-
strate the high sensitivity of the lmax to the twist across the
biphenyl linkage. Hence, a novel generation of these dyes that
absorbs at much longer wavelength can be achieved if the twist
across the biphenyl is prevented.
better understanding into the electronic and thermodynamic
properties of dyes YS01–04, the equilibrium molecular geometry
for each dye was calculated first followed by an energy calcula-
tion, and the S+/*, S+/0, vertical electronic excitation values, and
the delocalization of the HOMO and LUMO were extracted for
each dye.
To identify an accurate and reliable molecular modelling
method for the prediction of the equilibrium molecular geom-
etry, different geometry optimization methods were performed
first on dye YS01. First, a conformation search was performed
on dye YS01 in molecular mechanics using augmented MM3
parameters, which generated an energy map of 14 641
conformers, shown in Fig. 3. Fig. 4 shows the lowest confor-
mational energy structure, the global energy minimum
conformer, in which the two terminal torsion angles q1 and q2
were twisted by 36.05ꢂ (q1) and 36.26ꢂ (q2), respectively.
To determine the accuracy of the global energy minimum
conformer calculated using MM3 parameters, the ground state
equilibrium molecular geometry of dye YS01 was calculated
using high level DFT calculations utilizing the hybrid energy
functional B3LYP at two different basis sets, 3-21G* and
DGDZVP.14 In the case of the basis set 3-21G*, the two torsion
angles (q1 and q2) of the equilibrium molecular geometry were
twisted by 38.67ꢂ and 40.67ꢂ while in the case of the basis set
DGDZVP the twist was slightly less showing 34.42ꢂ and 36.11ꢂ,
respectively, which was comparable to those obtained using
MM3 calculations.
Based on the data presented in Table 2, it was deemed that the
aforementioned DFT and TD-DFT methods will be the methods
of choice for the prediction of the optimized geometries and lmax
,
respectively, for YS02–04. Table 3 shows the calculated versus
experimental lmax and S+/* for dyes YS01–04. It also shows the
Table 3 A comparison between the experimental and calculated S+/* and
lmax of YS01–04
Torsion
angles (ꢂ)
S+/* (eV)
Calcd
lmax (nm)
Dye
q1
q2
Expt.
Calcd
Expt.
YS01
YS02
YS03
YS04
34.42
34.51
35.29
35.05
36.11
37.57
36.12
39.07
ꢁ3.361
ꢁ3.228
ꢁ3.229
ꢁ3.263
ꢁ3.18
ꢁ3.15
ꢁ3.16
ꢁ3.15
376
377
377
376
386
390
391
385
Fig. 4 Global energy conformer extracted from the potential energy
map shown in Fig. 3, calculated using augmented MM3 parameters. The
terminal torsion angles are twisted by 36.05ꢂ (q1) and 36.26ꢂ (q2), from
left to right.
Fig. 5 Calculated absorption spectra of YS01–YS04, using the energy
functional CAM-B3LYP and the basis set DGDZVP.
This journal is ª The Royal Society of Chemistry 2012
J. Mater. Chem., 2012, 22, 10771–10778 | 10775

View Article Online
calculated torsion angles (q1 and q2). The calculated lmax and S+/*
are in excellent agreement with experimental results, which
demonstrates the high level of accuracy of the DFT/TD-DFT
methods used for this class of dyes. Moreover, YS04 exhibited
the largest twist across the biphenyl linkage for q2 (39.09ꢂ), which
reduced the p–p overlap across the quarterphenyl, and furnished
the weakest charge transfer from the D to A in the series.
Fig. 5 shows the calculated absorption spectra of YS01–YS04,
using time-dependent DFT long-range corrected energy func-
tional CAM-B3LYP and the basis set DGDZVP. The compar-
ison between calculated (Fig. 5) and experimental absorption
spectra (Fig. 2a) showed excellent agreement between the
calculated and experimental lmax. Moreover, the shapes of
calculated absorption peaks are also identical to the experimental
spectra. Fig. 6 shows the HOMO (left) and LUMO (right) of
dyes YS01–04. It can be seen clearly that the HOMO is delo-
calized mainly on the electron donor regime where the aux-
ochrome (indole analog) resides, which facilitates the reduction
of the oxidized dye through the reaction with Iꢁ, making the dye
regeneration more efficient. On the other hand, the LUMO was
mainly delocalized on the p* cyanuric unit and has sizable
contributions from the carboxylic groups, facilitating electron
injection from the photoexcited sensitizer into the TiO2 semi-
conductor. This clearly confirms that this structural motif is an
excellent design for efficient charge transfer in DSCs. Moreover,
it can be seen from Fig. 6 that the torsion angle q2 across the
Fig. 6 The calculated HOMO and LUMO isodensity of YS01–YS04 showing the regions of delocalization.
10776 | J. Mater. Chem., 2012, 22, 10771–10778
This journal is ª The Royal Society of Chemistry 2012

View Article Online
biphenyl linkage for YS04 is the largest, which diminished the p–
p* overlap, and this is why its photocurrent and total conversion
efficiency were the lowest compared to dyes YS01–03.
potential to exhibit significant increase in the extinction coeffi-
cient and simultaneously harvest more energetic photons across
a wider range of the solar radiation spectrum.
Fig. 7 shows the monochromatic IPCE and photocurrent–
voltage (I–V) characteristics for DSCs based on YS01–04
demonstrating excellent sensitization of nanocrystalline TiO2
from 350 to 650 nm with a maximum quantum efficiency of 85%
in the plateau region. Dye YS03 showed the highest solar to
power conversion efficiency (h) (5.06%) among the four dyes with
the short-circuit current (Jsc), the open circuit voltage (Voc) and
the fillfactor (FF) of 9.14 mA cmꢁ2, 0.778 V and 0.712, respec-
tively. Under the same conditions the overall power conversion
efficiencies (h) of dyes YS01 and YS02 were 4.64 and 4.55%,
respectively. Silyl-containing dyes (YS01–03) showed higher
power conversion efficiency compared to the fluorene-based dye
YS04 (2.88%), as shown in Table 1 due to the strong electron
donation of silicon relative to carbon and the less twist across the
biphenyl linkages of q1 and q2, and the anodic shift of S+/* in dyes
YS01–03 compared to dye YS04. The overall power conversion
efficiency of these dyes was in the range 2.88–5.06% owing to the
lack of absorption of the energetic photons from the blue, red
and NIR regions. Therefore, work is underway in our laboratory
to develop a novel generation of more efficient dyes that have the
Conclusions
Four novel D–p–A sensitizers, YS01–04, were designed,
synthesized, and characterized for DSCs. It was demonstrated
that dyes with a p-conjugated-silyl unit linker (YS01–03) fur-
nished better charge separation and more photocurrent, resulting
in higher total conversion efficiency than the silicon-free dye
(YS04). Among dyes YS01–YS03, the photovoltaic properties of
YS03 showed the greatest light harvesting ability, suppressed dye
aggregation owing to the highly branched donor group, which
led to the highest overall conversion efficiency of 5.06%. More-
over, DFT calculations showed that the torsion angles across the
biphenyl linkage in dyes containing silicon (YS01–YS03) are less
than in the silicon-free dye (YS04), which contributed to better
charge separation, and enhancements in the total efficiency of
dyes YS01–03 compared to YS04. Results from this work
strongly indicate that the application of this class of organic dyes
as panchromatic sensitizers for DSCs is promising, and opens
new avenues for the design of more efficient organic dyes based
on the silyl moiety for DSCs. With the aid of molecular
modeling, work in our laboratory is underway for the synthesis
of novel silicon-based sensitizers that can harvest solar energy
across a wider range of the solar spectrum up to 900 nm.
Acknowledgements
This work was partly supported by Core Research for Evolu-
tional Science and Technology (CREST) of the Japan Science
and Technology Agency.
References
€
1 (a) B. O’Regan and M. Gratzel, Nature, 1991, 353, 737; (b) A. Mishra,
M. K. R. Fischer and P. Bauerle, Angew. Chem., Int. Ed., 2009, 48,
€
€
2474; (c) A. Hagfeld and M. Gratzel, Acc. Chem. Res., 2000, 33,
269; (d) A. Hagfeld and M. Gratzel, Chem. Rev., 1995, 95, 49; (e)
€
L. Han, A. Islam, H. Chen, C. Malapaka, B. Chiranjeevi, S. Zhang,
X. Yang and M. Yanagida, Energy Environ. Sci., 2012, 5, 6057.
2 (a) N. Robertson, Angew. Chem., Int. Ed., 2006, 45, 2338; (b)
Y. Ooyama and Y. Harima, Eur. J. Org. Chem., 2009, 2903.
ꢀ
3 (a) 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. Gratzel, J. Am. Chem. Soc., 2001, 123, 1613; (b) Y. Chiba,
A. Islam, Y. Watanabe, R. Komiya, N. Koide and L. Han, Jpn. J.
Appl. Phys., 2006, 45, L638.
4 D. Liu, R. W. Fessenden, G. L. Hug and P. V. Kamat, J. Phys. Chem.
B, 1997, 101, 2583.
5 M. Akhtaruzzaman, A. Islam, F. Yang, N. Asao, E. Kwon,
S. P. Singh, L. Han and Y. Yamamoto, Chem. Commun., 2011, 47,
12400.
6 A. Islam, H. Sugihara and H. Arakawa, J. Photochem. Photobiol., A,
2003, 158, 131.
7 (a) S. Furukawa, J. Kobayashi and T. Kawashima, Dalton Trans.,
2010, 39, 9329; (b) C. W. Keyworth, K. L. Chan, J. G. Labram,
T. D. Anthopoulos and S. Watkins, J. Mater. Chem., 2011, 21,
11800; (c) G. Lu, H. Usta, C. Risko, L. Wang, A. Facchetti,
M. A. Ratner and T. J. Marks, J. Am. Chem. Soc., 2008, 130, 7670.
8 (a) H. Usta, G. Lu, A. Facchetti and T. J. Marks, J. Am. Chem. Soc.,
2006, 128, 9034; (b) Y. Yabusaki, N. Ohshima, H. Kondo,
T. Kusamoto, Y. Yamanoi and H. Nishihara, Chem.–Eur. J., 2010,
16, 5581.
Fig. 7 (a) Photocurrent action spectra (IPCE) of nanocrystalline TiO2
film sensitized by YS01–04. (b) Photocurrent–voltage (I–V) characteris-
tics of dyes YS01–04. The redox electrolyte solution was a mixture of
0.6 M DMPII, 0.05 M I2, 0.1 M LiI and 0.5 M TBP in acetonitrile.
This journal is ª The Royal Society of Chemistry 2012
J. Mater. Chem., 2012, 22, 10771–10778 | 10777

View Article Online
9 (a) S. Ko, H. Choi, M.-S. Kang, H. Hwang, H. Ji, J. Kim, J. Ko and
Y. Kang, J. Mater. Chem., 2010, 20, 2391; (b) L.-Y. Lin, C.-H. Tsai,
K.-T. Wong, T.-W. Huang, L. Hsieh, S.-H. Liu, H.-W. Lin,
C.-C. Wu, S.-H. Chou, S.-H. Chen and A.-I. Tsai, J. Org. Chem.,
2010, 75, 4778; (c) W. Zeng, Y. Cao, Y. Bai, Y. Wang, Y. Shi,
M. Zhang, F. Wang,C. PanandP. Wang,Chem.Mater., 2010, 22, 1915.
10 M. Matsui, A. Ito, M. Kotani, Y. Kubota, K. Funabiki, J. Jin,
T. Yoshida, H. Minoura and H. Miura, Dyes Pigm., 2009, 80, 233.
11 Y. Nakao, J. Chen, M. Tanaka and T. Hiyama, J. Am. Chem. Soc.,
2007, 129, 11694.
C.-C. Wu, S.-H. Chou, S.-H. Chen and A.-I. Tsai, J. Org. Chem.,
2010, 75, 4778; (c) W. Zeng, Y. Cao, Y. Bai, Y. Wang, Y. Shi,
M. Zhang, F. Wang, C. Pan and P. Wang, Chem. Mater., 2010, 22,
1915.
15 A. D. Becke, Phys. Rev. A: At., Mol., Opt. Phys., 1988, 38, 3098.
16 C. T. Lee, W. T. Yang and R. G. Parr, Phys. Rev. B, 1988, 37, 785.
17 N. Godbout, D. R. Salahub, J. Andzelm and E. Wimmer, Can. J.
Chem., 1992, 70, 560.
18 C. Sosa, J. Andzelm, B. C. Elkin, E. Wimmer, K. D. Dobbs and
D. A. Dixon, J. Phys. Chem., 1992, 96, 6630.
12 G. Oskam, B. V. Bergeron, G. J. Meyer and P. C. Searson, J. Phys.
Chem. B, 2001, 105, 6867.
19 T. Yanai, D. Tew and N. Handy, Chem. Phys. Lett., 2004, 393, 51.
20 H. Iikura, T. Tsuneda, T. Yanai and K. Hirao, J. Chem. Phys., 2001,
115, 3540.
21 J.-D. Chai and M. Head-Gordon, Phys. Chem. Chem. Phys., 2008, 10,
6615.
13 E0–0 value was estimated from the 5% intensity level of the absorption
spectrum of dyes YS01–04 adsorbed onto a transparent TiO2 film.
14 (a) S. Ko, H. Choi, M.-S. Kang, H. Hwang, H. Ji, J. Kim, J. Ko and
Y. Kang, J. Mater. Chem., 2010, 20, 2391; (b) L.-Y. Lin, C.-H. Tsai,
K.-T. Wong, T.-W. Huang, L. Hsieh, S.-H. Liu, H.-W. Lin,
22 Y. Tawada, T. Tsuneda and S. Yanagisawa, J. Chem. Phys., 2004,
120, 8425.
10778 | J. Mater. Chem., 2012, 22, 10771–10778
This journal is ª The Royal Society of Chemistry 2012