A compact diketopyrrolopyrrole dye as efficient sensitizer in titanium dioxide dye-sensitized solar cells

Article abstract of DOI:10.1016/j.jphotochem.2011.09.023

Two novel TiO₂ sensitizers, based on the highly stable diketopyrrolopyrrole (DPP) skeleton, have been synthesized for application in the field of dye sensitized solar cells. The obtained dyes, DPP1 and DPP2 bear respectively a cyanoacrylic acid and a rhodanine acid anchoring groups, thus tuning the extent of the electronic communication with the semi-conducting oxide. The two chromophores were characterized by solution phase spectroscopy and electrochemistry. DFT calculations gave deeper insight into the electronic structure of both dyes, through the disclosure of their frontier orbitals. Photovoltaic performances unravelled the undisputable advantage of DPP1 over DPP2, owing to the combination of a favourable dipolar moment interaction with TiO₂, and more intimate orbital blending between the chemisorbed dye and the conduction band. Chenodeoxycholic acid proved to be useful in limiting the formation of dye aggregates, improving to a great extent the performances of DPP1 based DSSCs, reaching in our conditions a 4.47% yield and 57% IPCE at 500 nm.

Full text of DOI:10.1016/j.jphotochem.2011.09.023

Journal of Photochemistry and Photobiology A: Chemistry 226 (2011) 9–15
Contents lists available at SciVerse ScienceDirect
Journal of Photochemistry and Photobiology A:
Chemistry
journal homepage: www.elsevier.com/locate/jphotochem
A compact diketopyrrolopyrrole dye as efficient sensitizer in titanium dioxide
dye-sensitized solar cells
Julien Warnan, Ludovic Favereau, Yann Pellegrin, Errol Blart, Denis Jacquemin∗, Fabrice Odobel∗
Université de Nantes, CNRS, Chimie et Interdisciplinarité: Synthèse, Analyse, Modélisation (CEISAM), UMR CNRS n◦ 6230, 2, rue de la Houssinière – BP 92208 – 44322 Nantes Cedex
3, France
a r t i c l e i n f o
a b s t r a c t
Article history:
Two novel TiO2 sensitizers, based on the highly stable diketopyrrolopyrrole (DPP) skeleton, have been
synthesized for application in the field of dye sensitized solar cells. The obtained dyes, DPP1 and DPP2
bear respectively a cyanoacrylic acid and a rhodanine acid anchoring groups, thus tuning the extent of the
electronic communication with the semi-conducting oxide. The two chromophores were characterized
by solution phase spectroscopy and electrochemistry. DFT calculations gave deeper insight into the elec-
tronic structure of both dyes, through the disclosure of their frontier orbitals. Photovoltaic performances
unravelled the undisputable advantage of DPP1 over DPP2, owing to the combination of a favourable
dipolar moment interaction with TiO2, and more intimate orbital blending between the chemisorbed
dye and the conduction band. Chenodeoxycholic acid proved to be useful in limiting the formation of
dye aggregates, improving to a great extent the performances of DPP1 based DSSCs, reaching in our
conditions a 4.47% yield and 57% IPCE at 500 nm.
Received 20 August 2011
Accepted 24 September 2011
Available online 14 October 2011
Keywords:
Dye-sensitized solar cells
Organic sensitizers
TiO2
TD-DFT calculations
Diketopyrrolopyrrole
© 2011 Elsevier B.V. All rights reserved.
1. Introduction
To provide a good solubility of the dyes, the nitrogen of lac-
tam units were alkylated by branched fatty alkyl chains. Moreover,
Discovering efficient and robust organic sensitizers is an impor-
tant goal in the field of dye-sensitized solar cell (DSSC) and has
therefore stimulated countless investigations for more than a
decade [1]. Originally, ruthenium polypyridine complexes were the
dominating sensitizers used in this area [2] but since 2000 many
classes of organic dyes were systematically investigated and some
of them achieved impressive photo-conversion efficiencies around
10% and even higher [3–5]. Diketopyrrolopyrrole (DPP) is a well-
known chromogen with particularly high photostability [6] and
which was quite thoroughly used as electron donor unit in organic
solar cells [7]. However, there are only three reports on the use of
this class of dyes in DSSC [8,9]. For its outstanding photo-stability,
high absorbance in the visible region, high synthetic accessibility,
we felt that DDP derivatives could constitute efficient sensitizers
for DSSC and deserved further investigations. In this study, we
prepared two new compact dyes as we strive for obtaining a signif-
icant sensitizer efficiency with a minimal size. The two compounds
essentially differ by their anchoring group: cyanoacrylic acid for
DDP1 and rhodanine-3-acetic acid for DPP2 (Chart 1).
these substituents could also wrap the dye core and limit their
aggregation upon binding on the TiO2 surface. They can also prevent
the I3 from approaching the surface of TiO2 and inhibit the inter-
ception of the injected electron by the electrolyte [5,10]. In spite
of its simplicity, we demonstrate that DPP1 clearly represents a
valuable unit to prepare efficient diketopyrrolopyrrole-based sen-
sitizers, since it gives the best photo-conversion efficiency reported
for this class of dye [9].
−
2. Synthesis
Preparation of the DDP core follows a previously described
method consisting in a pseudo-Stobbe condensation of 1-bromo-
4-cyanobenzene and diethylsuccinate in presence of a strong base
[11,12]. A double alkylation was subsequently achieved by depro-
tonation of the lactam units by potassium tertbutanolate which
was reacted with the branched alkyl 1-bromo-2-ethylhexane in
N-methylpyrrolidone to furnish compound 1 in a modest yield
of 26% [13]. Then, a Suzuki cross-coupling reaction with the 4-
formylthiophenyl boronic acid in presence of sodium carbonate
and tetrakis(triphenylphosphine)palladium as catalytic system led
to the aldehyde 2 with a yield of 34% (Scheme 1) [9,11,14]. The
two other main products were the starting material 1 and the
bis-coupled product but they were easily separated by column
chromatography due to their very different polarities. Finally, two
∗
Corresponding authors. Tel.: +33 02 51 12 54 29; fax: +33 02 51 12 54 02.
E-mail addresses: denis.jacquemin@univ-nantes.fr (D. Jacquemin),
Fabrice.Odobel@univ-nantes.fr (F. Odobel).
1010-6030/$ – see front matter © 2011 Elsevier B.V. All rights reserved.
doi:10.1016/j.jphotochem.2011.09.023

10
J. Warnan et al. / Journal of Photochemistry and Photobiology A: Chemistry 226 (2011) 9–15
Scheme 1. Synthesis of DPP1 and DPP2.
transitions correspond to a partial charge transfer (CT), in which
the electron density shifts from the DPP core towards the
anchoring moiety. On TiO2, the spectral coverage spans from
380 nm to 600 nm, with a shallower valley around 450 nm and
slightly larger absorption window for DPP2 (Fig. 2). The presence
of chenodeoxycholic acid (CDCA) during the chemisorption step
induces a small but noticeable blue-shift of the absorption spec-
tra (Fig. 2). This supports the fact that CDCA certainly prevents the
formation of aggregates of these dyes on TiO2, though aggregation
was originally very limited, thanks to, the alkyl chains on the lactam
units. DPP1 and DPP2 are both strongly fluorescent molecules and
they display a strong emission band around 650 nm, which gives a
singlet excited state lying at about 2.25 eV (Fig. 1 and Table 1).
4. Electrochemistry
The redox potentials of the dyes were recorded by cyclic voltam-
metry to assess the thermodynamic feasibilities of the electron
transfer processes. Both dyes exhibit a reversible oxidation at
around 1.2 V vs. SCE with marginal influence of the anchoring group
which is consistent with the topology of the HOMO orbital. Indeed,
the latter is rather localized on the DPP unit with almost no con-
tribution of the anchoring group (Fig. 3). Using the well-accepted
values of −0.5 V vs. NHE [15] (−0.74 V vs. SCE) for the conduction
band of TiO2 and of 0.40 V vs. NHE [16] (0.16 V vs. SCE) for the redox
couple I2•−/I−, the electron injection Gibbs free energy (ꢀGinj) and
the dye regeneration Gibbs free energy (ꢀGreg) have been calcu-
Chart 1. Structure of the dyes investigated in this study.
Knoevenagel reactions of compound 2 with cyanoacrilic acid or
with rhodanine-3-acetic acid afforded respectively the final targets
DPP1 or DPP2 with high yield (>95%).
3. Electronic absorption and emission spectra
The electronic absorption spectra of the two new dyes were
recorded in dichloromethane solution and on TiO2. The spec-
tra are respectively illustrated in Figs. 1 and 2 and the data
are collected in Table 1. Diketopyrrolopyrroles usually exhibit
intense ␲–␲* transitions in the visible. In DPP1 and DPP2
the absorption bands around 400 and 500 nm are dominated
by a HOMO–LUMO and (HOMO−1) − LUMO/HOMO − (LUMO+1)
contributions, respectively (see below QM calculations). These
lated (Table 1). Clearly, the electron injection driving force (ꢀGinj
)
is relatively weak leaving the possibility of a sluggish reaction for
both dyes while the regeneration reaction of the oxidized sensi-
tizer (S+ + 2I− → S + I2•) is highly favourable (ꢀGreg < −1 eV) in both
cases (Table 1).
1,4
1
0,8
0,6
0,4
0,2
0
DPP1 without CDCA
DPP2 without CDCA
DPP2 with CDCA
DPP1 with CDCA
1,2
1
0,8
0,6
0,4
0,2
0
380
430
480
530
580
630
680
300
400
500
600
700
800
Wavelength (nm)
Wavelength/nm
Fig. 1. Normalized absorption (straight line) spectra and emission (dashed line) of
DPP1 (red) and DPP2 (black) recorded in dichloromethane. (For interpretation of
the references to color in this figure legend, the reader is referred to the web version
of this article.)
Fig. 2. Normalized absorption spectra of DPP1 (red) and DPP2 (black) recorded TiO2
electrodes with chenodeoxycholic acid (straight line) and without (dashed line). (For
interpretation of the references to color in this figure legend, the reader is referred
to the web version of this article.)

J. Warnan et al. / Journal of Photochemistry and Photobiology A: Chemistry 226 (2011) 9–15
11
Table 1
Absorption and emission characteristics along with redox potential and injection Gibbs free energies of DPP1 and DPP2.
Dye
EOx(S+/S) V vs. SCE
ꢁabs/ε(nm/M−1cm−1
)
ꢁem (nm)
aE00(S*) (eV)
bEOx(S+/S*) V vs. SCE
cꢀGinj (eV)
dꢀGreg (eV)
DPP1
DPP2
1.20
1.18
403 (2.1 × 104); 493 (1.4 × 104)
416 (1.3 × 104); 503 (1.2 × 104)
653
641
2.24
2.25
−1.04
−1.07
−0.30
−0.33
−1.04
1.02
SCE, saturated calomel electrode.
a
Calculated with the wavelength at the intersection (ꢁinter) normalized absorption and emission spectra with the equation E00 = 1240/ꢁinter
.
b
c
Calculated according to the equation: EOx(S+/S*) = EOx(S+/S) − E00
Calculated according to the equation: ꢀGinj = 0.74 + EOx(S+/S*).
Calculated according to the equation: ꢀGreg = 0.16 − EOx(S+/S).
.
d
Fig. 3. Graphical representation of the optimized structure with representation of CT of both absorption bands (see text) and of the frontier molecular orbitals of DPP1 (left)
and DPP2 (right). Branched alkyl chains were replaced by methyl groups for faster calculations.

12
J. Warnan et al. / Journal of Photochemistry and Photobiology A: Chemistry 226 (2011) 9–15
Table 2
5. Quantum chemical calculations
Photovoltaic performances for DPP1 and DPP2 cell in response to 100 mW/cm2
illumination.
The structural parameters of both dyes have been optimized
using Density Functional Theory (see section 10 for details), and
have been found to be nearly planar, e.g. for DPP1 the twist angles
between the diketopyrrolopyrrole moiety and the side phenyl rings
are ca. 30◦, whereas the phenyl-thiophene dihedral angle is 21◦. In
order to get insight into the electronic structure, Time-Dependent
DFT (TD-DFT) calculations were performed on both dyes. For
DPP1 (DPP2), we compute two strongly dipole-allowed transi-
tions in the visible, at 492 and 384 nm (494 and 411 nm), which
fits experiments (Fig. 1), including the significant bathochromic
shift of the second band when going from DPP1 to DPP2. The
TD-DFT fluorescence wavelengths from the lowest excited-states
are 648 and 650 nm for DPP1 and DPP2 and also nicely match
with the experimental values. The excited-state geometry is sig-
nificantly more planar (e.g. phenyl-thiophene dihedral angle of
0.2◦ for DPP1) than the ground state, which partly explains the
significant Stokes shift. For both dyes, the ca. 490 nm absorption
band is dominated by an HOMO–LUMO transition but also includes
a smaller HOMO–LUMO+1 component, whereas the ca. 400 nm
absorption includes almost equal shares of HOMO−1–LUMO and
HOMO–LUMO+1 contributions. These frontier molecular orbitals
are shown in Fig. 3. We observe that the methylene carboxylic acid
moiety of rhodanine-3-acetic acid anchoring group in DPP2 ori-
ents perpendicularly to the rest of the molecule. On the contrary,
DPP1 is a relatively planar dye, which ensures a high degree of ␲-
conjugation throughout the molecule. In both dyes, the electron
density on HOMO is located on the diketopyrrolopyrrole core and
particularly on the double bonds and the oxygen of the lactam
units. The LUMO and LUMO+1 are distributed on the anchor-
ing group and spreads up to the carboxylic acid group for DPP1,
whereas in DPP2 there is no contribution on the carboxylic acid
group in DPP2 there is no contribution on the carboxylic acid
group. Another way to qualitatively account for the importance
of electron transfer is to estimate the CT vector, as recently sug-
gested by Ciofini and co-workers [17]. For both absorption bands,
the CT distance is found to be significantly smaller for DPP2 (ca.
Dye
CDCA
No
t-BuPy
Voc (mV)
Jsc (mA/cm2)
ff (%)
Á(%)
DPP1
No
Yes
No
Yes
No
Yes
No
605
615
625
635
475
525
485
525
7.78
2.17
9.71
3.09
1.72
0.37
1.30
0.44
70.7
73.8
73.7
79.2
76.6
68.5
70.1
71.8
3.33
0.99
4.47
1.55
0.63
0.13
0.44
0.16
Yes
No
DPP2
Yes
Yes
sensitizer on the TiO2, because the surface occupation is shared
with two different molecules instead of one. Second, the proton
release during the chemisorption step acidifies the TiO2 surface
and induces a downward bending of the conduction band [19].
This latter effect is certainly important here as both dyes dis-
play rather low injection driving force (Table 1). Therefore, upon
band bending the exergonicity of the charge injection is increased
which enhances the photocurrent density. This beneficial effect
is only observed with DPP1, because DPP2 exhibits a poor elec-
tronic coupling, therefore the slight increase of the injection driving
force is counter-balanced by the lower dye loading. The 4-tert-
butylpyridine also plays two roles. On the one hand, it binds to
the TiO2 surface and causes an upward conduction band bend-
ing; on the other hand it prevents close contact between the
TiO2 surface and triiodide, subsequently reducing current losses
by recombination. These two effects essentially raise the Voc. Not
surprisingly, the presence of 4-tert-butylpyridine increases the Voc
,
but importantly decreases the overall photovoltaic performances
of both DPP1 and DPP2 by dramatically diminishing the Jsc, which
is the direct consequence of the already low injection Gibbs free
energy, a feature that is exacerbated by the upward bending of
the conduction band. The very large difference of power con-
version efficiencies of DPP1, anchored with cyanoacrylic group,
compared to that of DPP2 linked to TiO2 via rhodanine-3-acetic
acid is quite impressive (Table 2 and Fig. 4). This is the direct
consequence of the lower electronic coupling with DPP2 due
to the presence of a methylene unit in rhodanine-3-acetic acid,
which interrupts the ␲-conjugation between the dye and TiO2 and
slows down charge injection. Recently, it was also proposed that
cyanoacrylic acid injects electrons more deeply into the bulk of
TiO2, while with rhodanine-3-acetic acid the electron resides in
the surface at a shorter distance than from the oxidized dye sup-
porting thus a higher probability of charge back recombination
[20]. Another interesting difference is the higher Voc measured
with DPP1 (Voc = 625 mV) than with DPP2 (Voc = 485 mV). It prob-
ably results from the higher dipolar moment of DPP1 which
˚
˚
3.5 A) than for DPP1 (ca. 4.7 A), as illustrated at the bottom of
Fig. 3. As a result, DPP1 certainly features a much higher electronic
communication between DPP1 and with the conduction band of
TiO2 combined to a stronger CT character, hence a higher elec-
tronic coupling and a faster electron injection reaction with DPP1
than with DPP2. Another striking difference between DPP1 and
DPP2 is their ground-state dipole moments, which are estimated
to 10.4 D and 4.9 D, respectively. It is therefore likely that DPP1
induces a larger bending of the conduction band than DPP2, a fac-
tor, which was previously attributed to have an impact on the Voc
[18].
60%
6. Photovoltaic measurements
DPP1
50%
40%
30%
20%
10%
0%
DPP2
Both dyes DPP1 and DPP2 were tested in DSSC and the data
are collated in Table 2 and incident photon-to-current conversion
efficiency (IPCE) spectra are depicted in Fig. 4. In these experi-
ments, we used the classical iodide/triiodide electrolyte composed
of 0.6 M 1,2-dimethyl-3-butylimidazolium iodide, 0.1 M LiI and
0.05 M I2 in acetonitrile and we also tested the influence on the
photovoltaic performances of 4-tert-butylpyridine additive (0.5 M)
and of chenodeoxycholic acid (CDCA) as co-adsorbate. First, we
observe that CDCA significantly improves the photocurrent den-
sity of DPP1, which passes from 7.78 to 9.71 mA/cm2 whereas it
has a detrimental impact on DPP2, whose photocurrent dropped
upon CDCA co-adsorption. The effects of CDCA are two folds. First,
it reduces dye aggregation but at the cost of a lower loading of the
350
400
450
500
550
600
650
700
Wavelength (nm)
Fig. 4. Photoaction spectra of DPP1 and DPP2 co-adsorbed with CDCA and in
absence of tert-butylpyridine in the electrolyte.

J. Warnan et al. / Journal of Photochemistry and Photobiology A: Chemistry 226 (2011) 9–15
13
bends more significantly the TiO2 conduction band than DPP2 and
consequently raises the Voc. Anyhow to the best of our knowledge,
DPP1 is the most efficient sensitizer based on the diketopy-
rrolopyrrole chromogen. Indeed, in our conditions, it exhibits a
Jsc = 9.71 mA/cm2, a Voc = 625 mV and a ff = 73.7% corresponding to
a ␩ of 4.47% with an IPCE curve demonstrating a continuous elec-
tricity production from 400 to −600 nm (Fig. 4).
at 45 ◦C for 0.5 h, then a solution of 4-formylthiophenylboronic
acid (3.4 × 10−4 mol) in 4 mL of THF was added. The
temperature was increased to 80 ◦C and maintained for 16 h.
Once back at room temperature, water was poured and the crude
extracted with dichloromethane. After two aqueous washings, the
organic phase was dried on MgSO4, filtered and concentrated to
give a red solid. The product was then purified on silicagel col-
umn chromatography with dichloromethane as eluent. A yellow
ring corresponding to the starting material was collected first,
followed by an orange fraction corresponding to pure compound
2 (1.0 × 10−4 mol, 34%). 1H NMR (300 MHz, CDCl3): ıH = 9.91 (1H,
s); 7.84 (2H, d, 3J = 8.6); 7.76 (3H, m); 7.76 (2H, d, 3J = 8.5); 7.62
(4H, bs); 7.48 (1H, d, 3J = 4.0); 3.75 (2H, d, 3J = 7.8), 3.71 (2H, d,
3J = 7.6), 1.45 (2H, m), 1.08 (16H, m), 0.74 (12H, m). 13C NMR
(75 MHz, CDCl3): ıc = 182.9, 162.6, 152.7, 147.9, 143.4, 137.4,
135.4, 132.2, 130.2, 129.6, 129.2, 127.3, 126.7, 125.7, 125.2, 110.3,
110.2, 45.2, 45.1, 38.7, 30.4, 28.4, 23.9, 23.0, 14.1, 10.6. MALDI-
TOF: m/z: Calcd for: 701.2407 [MH]+, Found: 701.2419 [MH]+,
ꢀ = 1.7 ppm.
7. Conclusions
In this work, we prepared and characterized two new com-
pact diketopyrrolopyrrole dyes that are efficient sensitizers for TiO2
DSSC. We showed that DPP1 bearing a cyanoacrylic acid anchor-
ing group is far superior to DPP2 with a rhodanine-3-acetic acid
group. This highlights the strong influence of the anchoring group
on the photovoltaic performances, which becomes critical when
the electron injection driving force is moderate. DPP1 constitutes
a suitable molecular building block to engineer more efficient
diketopyrrolopyrrole-based sensitizer for DSSC. To improve the
above systems, it will be necessary to enlarge the electron injection
driving force. This study provides valuable guidelines for optimiz-
ing diketopyrrolopyrrole-based sensitizer for DSSC.
8.2. Compound DPP1
Compound
2
(2.1 × 10−5 mol) and cyanoacrylic acid
(6.4 × 10−4 mol) were placed in 4 mL of dry THF before dis-
tilled piperidine (9.0 × 10−4) was added. The solution was heated
to reflux for 10 h. The solution’s colour turned to deep red and the
degree of advancement was followed by TLC. At room temperature
dichloromethane and a diluted solution of hydrochloric acid
were added and the organic phase was washed with water and
dried on Na2SO4. After filtration and concentration the crude was
purified by silicagel column chromatography mounted with a
mixture of dichloromethane and methanol (9/1). After the removal
of low polarity residues, a few amounts of triethylamine were
added to the eluent. The obtained red fraction was washed with a
hydrochloric acid solution (1 M), dried, filtered and concentrated
to furnish a red solid (96%). 1H NMR (300 MHz, THF-d8): ıH; 8.36
(1H, s); 7.96 (2H, d, 3J = 8.4); 7.88 (3H, m); 7.76 (2H, d, 3J = 8.5);
7.69 (1H, d, 3J = 4.0); 7.63 (2H, d, 3J = 8.5); 3.81 (2H, d, 3J = 7.3), 3.77
(2H, d, 3J = 7.5), 1.43 (2H, m), 1.11 (16H, m), 0.72 (12H, m) FT-IR
(KBr, cm−1): 2218, 1698, 1652, 1121. MALDI-TOF: m/z: Calcd for:
768.2465 [MH]+, Found: 768.2438 [MH]+, ꢀ = 3.5 ppm.
8. Experimental part
1H and 13C NMR spectra were recorded on a Bruker ARX
300 MHz. Chemical shifts for 1H NMR spectra are referenced
relative to residual protium in the deuterated solvent (CDCl3
ı = 7.26 ppm for 1H and ı = 77.16 ppm for 13C; THF-d8 ı = 3.57,
1.72 ppm for 1H). Spectra were recorded at room temperature,
chemical shifts are written in ppm and coupling constants in
Hz. MALDI-TOF analyses were performed on a Bruker Ultraflex III,
microTOF Q spectrometer in positive linear mode at 20 kV accel-
eration voltage with 2,5-dihydroxybenzoic acid (DHB) or dithranol
as matrix. Electrochemical measurements were performed with a
potentiostat-galvanostat AutoLab PGSTAT 302N controlled by resi-
dent GPES software (General Purpose Electrochemical System 4.9)
using a conventional single-compartment three-electrode cell. The
working electrode was a Pt electrode, the auxiliary was a Pt wire of
10 mm long and the reference electrode was the saturated potas-
sium chloride calomel electrode (SCE). The supporting electrolyte
was 0.1 N Bu4NPF6 in CH2Cl2 and solutions were purged with
argon before the measurements. All potentials are quoted rela-
tive to SCE. In all the experiments the scan rate was 100 mV/s.
UV–Visible absorption spectra were recorded on a UV-2401PC Shi-
madzu spectrophotometer. Fluorescence spectra were recorded on
a SPEX Fluoromax fluorimeter. Infrared spectra (IR) were recorded
on a BRUKER Vector 22 spectrometer; frequencies are reported in
8.3. Compound DPP2
Compound
2
(2.1 × 10−5 mol) and rhodanine-3-acetic acid
(2.1 × 10−4 mol) were placed in 4 mL of dry THF before distilled
piperidine (9.0 × 10−4) was added. The solution was heated to
reflux for 1.5 h. The solution’s colour turned to deep red and the
degree of advancement was followed by TLC. At room tempera-
ture dichloromethane and a diluted solution of hydrochloric acid
were added before organic phase was washed with water and dried
on Na2SO4. After filtration and concentration the crude was puri-
fied by silicagel column chromatography mounted with a mixture
of dichloromethane and methanol (9/1). After elimination of low
polarity residues, a few amounts of triethylamine were added to the
solvent of chromatography. The obtained red fraction was washed
with a hydrochloric acid solution (1 M), dried, filtered and concen-
trated to furnish a red solid (99%). 1H NMR (300 MHz, THF-d8):
ıH = 8.11 (1H, s); 7.98 (2H, d, 3J = 7.9); 7.90 (2H, d, 3J = 7.9); 7.79 (2H,
d, 3J = 8.4); 7.71 (1H, d, 3J = 4.0); 7.66 (2H, d, 3J = 8.4); 7.61 (1H, d,
3J = 3.9), 4.79 (2H, s), 3.83 (2H, d, 3J = 7.0), 3.79 (2H, d, 3J = 7.3), 1.43
(2H, m), 1.12 (16H, m), 0.75 (12H, m) MALDI-TOF: m/z: Calcd for:
873.1934 [MH]+, Found: 873.1902 [MH]+, ꢀ = 3.7 ppm.
cm−1
.
Thin-layer chromatography (TLC) was performed on aluminium
sheets precoated with Merck 5735 Kieselgel 60F254. Column
chromatography was carried out with Merck 5735 Kieselgel
60F (0.040–0.063 mm mesh). Chemicals were purchased from
Sigma–Aldrich and used as received. Chenodeoxycholic acid and
titanium dioxide screen printing pastes were purchased from Sola-
ronix SA (Switzerland). Compound 1 was prepared according to
literature methods [12,13].
8.1. Compound 2 (adapted procedure from previous publication
[9])
Compounds 1 (3.0 × 10−4 mol), Pd(PPh3)4 (2.1 × 10−5 mol), and
sodium carbonate (5.3 × 10−3 mol) were solubilised in 4 mL of
THF + 2 mL of H2O, underargon atmosphere. The blend was heated

14
J. Warnan et al. / Journal of Photochemistry and Photobiology A: Chemistry 226 (2011) 9–15
9. Fabrication of the dye-sensitized solar cells
the optimal PBE0/6-311G(d,p) geometries determined for the first
excited-state though a PCM-TD-DFT optimization [26]. The con-
tour threshold selected to represent the molecular orbitals was
systematically set to 0.030 a.u. To estimate the charge-transfer, we
have used the procedure defined by Ciofini and coworkers [17], but
selected (Mulliken) partial atomic charges rather than electronic
densities as starting data. For both dyes, the lateral alkyl chains,
that are not expected to play a significant role in the optical prop-
erties and have been replaced by methyl groups (see Fig. 3) for the
sake of computational efficiency.
Conductive glass substrates (F-doped SnO2) were purchased
from Pilkington (TEC8, sheet resistance 8 ꢂ−2). Conductive glass
FTO substrates were successively cleaned by sonication in soapy
water, then ethanol for 10 min before being fired at 450 ◦C for
30 min. Once cooled down to room temperature, FTO plates were
rinsed with ethanol and dried in ambient air. TiO2 films were then
prepared in three steps. A first treatment is applied by immersion
for 30 min in an aqueous TiCl4 solution (50 mM) at 80 ◦C. Layers of
TiO2 were then screen printed with transparent colloidal paste Ti-
Nanoxide T20/SP and light scattering Ti-Nanoxide 300 as final layer,
with intermediate drying steps at 150 ◦C for 10 min between each
layer. The obtained substrates were then sintered at 450 ◦C, follow-
ing a progressive heating ramp (325 ◦C for 5 min, 375 ◦C for 5 min,
450 ◦C for 30 min). A second TiCl4 treatment was applied while cells
are still hot. Thicknesses were measured by a Sloan Dektak 3 pro-
filometer. The prepared TiO2 electrodes were soaked while still hot
(80 ◦C) in a 0.16 mM solution of each dye during 16 h. A mixture
of distilled solvents was used (dichloromethane/tetrahydrofuran
3/1, v/v) for bath preparation. In case of co-adsorption, required
quantity of chenodeoxycholic acid (0.6 mM) was added to the bath
before soaking.
Acknowledgements
The authors wish to thank the ANR HABISOL (program Asyscol,
n◦ ANR-08-HABISOL-002) for financial support. D.J. is indebted to
the Région des Pays de la Loire for financial support in the framework
of a recrutement sur poste stratégique. This research used resources
of the GENCI-CINES/IDRIS (Grant c2011085117) and of the CCIPL
(Centre de Calcul Intensif des Pays de Loire).
References
Electrolytes used are composed of: 0.6 M 1,2-dimethyl-3-
butylimidazolium iodide, 0.1 M LiI and 0.05 M I2 in acetonitrile,
with or without t-butylpyridine (0.5 M). Counter electrodes were
prepared by chemical deposition of platinum from hexachloropla-
tinic acid in distilled isopropanol (2 mg/mL) and subsequent firing
at 375 ◦C for 20 mn. The two electrodes were placed on top of each
other using a thin transparent film of Surlyn polymer (DuPont,
25 ␮m) as a spacer to form the electrolyte space. The empty cell was
tightly held, and the edges were heated to 110 ◦C to seal the two
electrodes together. A drop of electrolyte was introduced through a
predrilled hole in the counter electrode by vacuum backfilling, and
was sealed afterwards. The cells have an active area of ca. 0.25 cm2.
The current–voltage characteristics were determined by apply-
ing an external potential bias to the cell and measuring the
photocurrent using a Keithley model 2400 digital source meter.
The solar simulator is an Oriel Lamp calibrated to 100 mW/cm2.
The overall conversion efficiency (Á) of the photovoltaic cell is
calculated from the product of the photocurrent density (Jsc), the
open-circuit photovoltage (Voc), the fill factor of the cell (ff) divided
by the intensity of the incident light (100 mW/cm2).
[1] A. Mishra, M.K.R. Fischer, P. Bauerle, Angew. Chem. Int. Ed. 48 (2009) 2474.
[2] (a) A.S. Polo, M.K. Itokazu, N.Y. Murakami Iha, Coord. Chem. Rev. 248 (2004)
1343;
(b) B. O’Regan, M. Grätzel, Nature 353 (1991) 737;
(c) V. Yam, M. Nazeeruddin, M. Grätzel, Transition metal complexes for pho-
tovoltaic and light emitting applications, in: Photofunctional Transition Metal
Complexes, Springer, Berlin/Heidelberg, 2007, pp. 113–123.
[3] (a) S. Ito, H. Miura, S. Uchida, M. Takata, K. Sumioka, P. Liska, P. Comte, P. Pechy,
M. Gratzel, Chem. Commun. 5194 (2008);
(b) T. Bessho, S.M. Zakeeruddin, C.-Y. Yeh, E.W.-G. Diau, M. Graetzel, Angew,
Chem. Int. Ed. 49 (2010) 6646.
[4] (a) W. Zeng, Y. Cao, Y. Bai, Y. Wang, Y. Shi, M. Zhang, F. Wang, C. Pan, P. Wang,
Chem. Mater. 22, (2010) 1915;
(b) T. Horiuchi, H. Miura, K. Sumioka, S.J. Uchida, Am. Chem. Soc. 126 (2004)
12218.
[5] N. Koumura, Z.-S. Wang, S. Mori, M. Miyashita, E. Suzuki, K. Hara, J. Am. Chem.
Soc. 128 (2006) 14256.
[6] (a) M. Fukuda, K. Kodama, H. Yamamoto, K. Mito, Dyes Pigments 63 (2004) 115;
(b) Z. Hao, A. Iqbal, Chem. Soc. Rev. 26 (1997) 203.
[7] (a) D. Gendron, M. Leclerc, Energy Environ. Sci. 4 (2011) 1225;
(b) Y. Zou, D. Gendron, R. Neagu-Plesu, M. Leclerc, Macromolecules 42 (2009)
6361;
(c) H. Bronstein, Z. Chen, R.S. Ashraf, W. Zhang, J. Du, J.R. Durrant, P. Shakya
Tuladhar, K. Song, S.E. Watkins, Y. Geerts, M.M. Wienk, R.A.J. Janssen, T.
Anthopoulos, H. Sirringhaus, M. Heeney, I. McCulloch, J. Am. Chem. Soc. 133
(2011) 3272;
(d) J.C. Bijleveld, V.S. Gevaerts, D. Di Nuzzo, M. Turbiez, S.G.J. Mathijssen, D.M.
de Leeuw, M.M. Wienk, R.A. Janssen, J. Adv. Mater. 22 (2010) E242;
(e) M.-F. Falzon, A.P. Zoombelt, M.M. Wienk, R.A.J. Janssen, Phys. Chem. Chem.
Phys. 13 (2011) 8931.
10. Theoretical calculations
[8] (a) C. Kanimozhi, P. Balraju, G.D. Sharma, S. Patil, J. Phys. Chem. C 114 (2010)
3287;
All simulations have been achieved with Gaussian09 program
[21], applying default procedures, integration grids, algorithms
and parameters, except for tighten SCF (10−9 a.u.) and internal
forces (10−5 a.u.) convergence thresholds. We have adopted a three
step strategy that is efficient to determine the UV/Vis features
of most organic dyes [22]. The computational protocol system-
atically (all steps) includes a modelling of bulk solvent effects
(here CH2Cl2) through the Polarizable Continuum Model (PCM)
[23] and proceeds as: (1) the ground-state geometrical param-
eters have been determined at the PBE0/6-311G(d,p) level [24],
via a force-minimization process; (2) the vibrational spectrum of
each derivatives has been determined analytically at the same
level of theory, that is PBE0/6-311G(d,p), and it has been checked
that all structures correspond to true minima of the potential
energy surface; (3) the first ten low-lying excited-states have been
determined within the vertical TD-DFT approximation using the
CAM-B3LYP/6-311+G(2d,p) level of approximation [25]; and (4) the
fluorescence wavelengths were estimated though the use of verti-
cal CAM-B3LYP/6-311 + G(2d,p) transition energies computed on
(b) F.-L. Guo, S.-Y. Qu, W.-J. Wu, J. Li, W.-J. Ying, J.-L. Hua, Synth. Met. 160 (2010)
1767.
[9] S. Qu, W. Wu, J. Hua, C. Kong, Y. Long, H. Tian, J. Phys. Chem. C 114 (2010) 1343.
[10] J.E. Kroeze, N. Hirata, S. Koops, M.K. Nazeeruddin, L. Schmidt-Mende, M.
Grätzel, J.R. Durrant, J. Am. Chem. Soc. 128 (2006) 16376.
[11] D. Cao, Q. Liu, W. Zeng, S. Han, J. Peng, S. Liu, J. Polym. Sci. Part. A 44 (2006)
2395.
[12] E.Q. Guo, P.H. Ren, Y.L. Zhang, H.C. Zhang, W. Yang, J. Chem. Commun. (2009)
5859.
[13] G. Zhang, K. Liu, Y. Li, M. Yang, Polym. Int. 58 (2009) 665.
[14] Y. Zhu, A.R. Rabindranath, T. Beyerlein, B. Tieke, Macromolecules 40 (2007)
6981.
[15] A. Hagfeldt, M. Grätzel, Chem. Rev. 95 (1995) 49.
[16] M. Nazeeruddin, M. Grätzel, Transition metal complexes for photovoltaic and
light emitting applications Photofunctional Transition Metal Complexes, vol.
123, Springer Berlin/Heidelberg, 2007, p. 113.
[17] T. Le Bahers, C. Adamo, I. Ciofini, J. Chem. Theory Comput. 7 (2011) 2498–2506.
[18] (a) S. Rühle, M. Greenshtein, S.G. Chen, A. Merson, H. Pizem, C.S. Sukenik, D.
Cahen, A. Zaban, J. Phys. Chem. B 109 (2005) 18907;
(b) Y. Liang, B. Peng, J. Chen, J. Phys. Chem. C 114 (2010) 10992.
[19] S.A. Haque, E. Palomares, B.M. Cho, A.N.M. Green, N. Hirata, D.R. Klug, J.R. Dur-
rant, J. Am. Chem. Soc. 127 (2005) 3456.
[20] J. Wiberg, T. Marinado, D.P. Hagberg, L. Sun, A. Hagfeldt, B. Albinsson, J. Phys.
Chem. C 113 (2009) 3881.

J. Warnan et al. / Journal of Photochemistry and Photobiology A: Chemistry 226 (2011) 9–15
15
[21] M.J. Frisch, et al., Gaussian 09 Revision A. 02, Gaussian Inc., Wallingford CT,
2009.
(d) J. Preat, C. Michaux, D. Jacquemin, E.A. Perpète, J. Phys. Chem. C 113 (2009)
16821.
[22] (a) D. Jacquemin, E.A. Perpète, I. Ciofini, C. Adamo, Acc. Chem. Res. 42 (2009)
326;
[23] J. Tomasi, B. Mennucci, R. Cammi, Chem. Rev. 105 (2005) 2999.
[24] C. Adamo, V. Barone, J. Chem. Phys. 110 (1999) 6158.
[25] T. Yanai, D.P. Tew, N.C. Handy, Chem. Phys. Lett. 393 (2004) 51.
[26] G. Scalmani, M.J. Frisch, B. Mennucci, J. Tomasi, R. Cammi, V. Barone, J. Chem.
Phys. 124 (2006) 094107.
(b) F. Labat, I. Ciofini, H.P. Hratchian, M.J. Frisch, K. Raghavachari, C. Adamo, J.
Am. Chem. Soc. 131, (2009) 14290;
(c) M. Pastore, S. Fantacci, F. De Angelis, J. Phys. Chem. C 114, (2010) 22742;