Malachite green derivatives for dye-sensitized solar cells: Optoelectronic characterizations and persistence on TiO2

Full text of DOI:10.1246/bcsj.20170289

Advance Publication Cover Page
Malachite Green Derivatives for Dye-sensitized Solar Cells: Optoelectronic
Characterizations and Persistence on TiO2
Jean-Baptiste Harlé,* Shuhei Arata, Shinya Mine, Takashi Kamegawa, Van Tay Nguyen,
Takeshi Maeda, Hiroyuki Nakazumi, and Hideki Fujiwara*
Advance Publication on the web October 12, 2017
doi:10.1246/bcsj.20170289
© 2017 The Chemical Society of Japan

Malachite Green Derivatives for Dye-sensitized Solar Cells: Optoelectronic
Characterizations and Persistence on TiO2
Jean-Baptiste Harlé,*1 Shuhei Arata,1 Shinya Mine,2 Takashi Kamegawa,2 Van Tay Nguyen,3 Takeshi Maeda,3
Hiroyuki Nakazumi,3 and Hideki Fujiwara*1
1Department of Chemistry, Graduate School of Science, Osaka Prefecture University, 1-1 Gakuen-cho, Naka-ku, Sakai 599-8531
2Nanoscience and Nanotechnology Research Center, Osaka Prefecture University, 1-2 Gakuen-cho, Naka-ku, Sakai 599-8570
3Department of Applied Chemistry, Graduate School of Engineering, Osaka Prefecture University, 1-1 Gakuen-cho, Naka-ku, Sakai
599-8531
E-mail: jeanbaptiste.harle@gmail.com, hfuji@c.s.osakafu-u.ac.jp
Jean-Baptiste HARLE received his M. Sc. degree in 2013 from the university Pierre and Marie Curie (Paris VI). He
then joined the Fujiwara Laboratory at Osaka Prefecture University, Japan to perform his Ph.D degree. His research
interest encompasses the development of functional materials for dye-sensitized solar cells.
Hideki FUJIWARA received his Ph.D. (1999) from Kyoto University. He was appointed as a Research Associate at
the Institute for Molecular Science in 1996 and moved to the Research Institute for Advanced Science and Technology,
Osaka Prefecture University in 2003. Then he joined the Department of Chemistry, Osaka Prefecture University in
2005. He became a Lecturer (2006), an Associate Professor (2010) and a Professor (2016) of Osaka Prefecture
University. He was awarded the Chemical Society of Japan Award for Young Chemists (2003).
Abstract
Derivatives of malachite green,
triphenylmethine dye, have been adapted for third-generation
photovoltaic applications as dye-sensitized solar cells (DSSC).
for photovoltaic applications. Only a few examples like Eosin
Y,3 Rhodamine,4-8 Phloxine B and Bromophenol Blue9 or Rose
Bengal10 and Pyrogallol Red11-13 dyes were reported available
for sensitization in the standard nanocrystalline TiO2
anatase-based DSSC. Malachite green and a few other
a
well-known
–
The solar cells were developed based on a concentrated Br3
/Br– liquid electrolyte coupled to different trifluoroacetate
derivatives have been considered for sensitization in a
(TFA–), triflate (TfO–), bromide (Br–) and tetrafluoroborate
SnO2-liquid junction-based cell8 and recently inserted in
(BF4 ) malachite green salts as dye sensitizers and mesoporous
CdS-based7,
and Zr-doped SrTiO3 photovoltaic
–
14
TiO2 anatase as electron collector, and their optoelectronic
properties were characterized. The adsorption patterns of such
salts at the TiO2 nanoparticles surface were studied by zeta (ζ)
potential measurements on colloidal suspensions under neat
conditions, and compared to the desorption rates of the dyes
when exposed to the DSSC electrolyte. The different affinities
of the ionic pairs for the oxide surface and the bulk were found
crucial for the stability of the self-assembled monolayer of
carboxylic acid-anchored chromophores at the surface, and for
the photoconversion efficiency associated therewith. This study
aimed at depicting the behaviour of the ionic pairs at the
surface and gave insights for their physical and chemical
stabilization in the DSSC environment.
technologies.15 Sensitization using Ru-based complexes with
free or surface-bounded counterions have been carried out by
M. Grätzel and outlined the impact of their nature on the cells
performances,16, 17 but the studies neglected the impact of the
counterions on the ionic distribution within the electrical
double layer and their influence on the adsorption energetics of
the dyes at the surface. Regulation of the diffuse ion swarm
across the dye monolayer has a bearing on the motion of ions
and the concentration gradients of the electrolyte and hence
directly
impacts
the
electron
transfers
at
the
TiO2/dye/electrolyte interface.18 Therefore, to investigate the
role of the counterions under DSSC conditions, we synthesized
new malachite green dyes 1, 2 (Figures 1, 2) with a series of
–
free (BF4 , Br–) and adsorption-capable (TfO–, TFA–)
1. Introduction
counterions. The cationic structures of 1Me+, 1Dec+ and 2Me+
were derived from malachite green and adapted for n-type
dye-sensitized solar cell (DSSC) sensitization with different
4-carboxyphenyl (1) and 5-carboxythiophen-2-yl (2) groups as
acceptor/anchor moieties, and n-decyl hydrophobic moieties.
The photoinactive counterions were found to play a significant
role in the photoconversion capacity of the devices developed
for this study as well as on the adsorption/desorption rates of
the cationic chromophores.
Malachite green (MG) and its analogs are hindered
short-chain (monomethine) cyanine dyes, i.e. conjugated ionic
type chromophores with an odd alternant system with two
extreme resonance structures between two auxochromic
nitrogens (Figure 1). It absorbs light at longer wavelength
thanks to the strong bathochromic shift induced by the steric
hindrance observed between the two N↔N-intrachain phenyl
groups.1 Although widespread,2 mostly cost-effective and
allowing light-harvesting in the visible light and at longer
wavelength, triarylmethine dyes have been shown few interest
1

UV/vis/near-IR measurement under air conditions.
N
N
Fabrication of the electrodes. Colourless 3.2 mm (TEC8,
Dyesol Ltd.) and 2.2 mm (TEC7, Dyesol Ltd.) thick soda-lime
glass coated with FTO were used as conductive substrate for
the front (photoanode) and back (counter electrode) contacts,
respectively. The glasses were successively cleaned with soap,
rinsed with tap water and washed with acetone and methanol
and underwent oxidation with a UV-O3 system for 20 min. The
TEC8 plates were subsequently immersed in a 40 mM aqueous
TiCl4 solution and the bath was heated to 70 °C for 30 min and
washed with deionized water and methanol and dried under
vacuum. A rheological paste of TiO2 nanoparticles (18NR-T
transparent, Dyesol Ltd.) was screenprinted with a 90T mesh
and dried for 5 min at 125 °C. After cooling to room
temperature the screen-printing procedure was repeated and the
plates were heated to 125 °C for 5 min. A third layer of
light-scattering nanoparticles (PST-400C active-opaque,
Dyesol Ltd.) was screenprinted over as the last photoactive
layer of the square 0.25 cm2 electrode, and the whole was
heated to 125 °C for 5 min and sintered at 475 °C for 45 min.
The photoanodes were treated once more with a 40 mM
aqueous TiCl4 solution, heated to 70 °C for 30 min, rinsed with
deionized water and washed with acetone, and calcinated at
500 °C for 15 min. Platina was coated at the back-contact as
follows: A 10–2 M hexachloroplatinic acid (H2PtCl6) ethanolic
solution was poured dropwise on the TEC7 plates and kept still
as the solvent evaporated, and the electrodes were heated for 15
min at 400 °C.
OH
O
N
N
N
1Me+
MG
C10H21
N
C10H21
HO
OH
S
O
O
C10H21
N
C10H21
N
2Me+
1Dec+
Figure 1. Chemical structures of malachite green MG and its
derivatives 1Me+, 2Me+ and 1Dec+ developed for the present
study and represented in their quinonoidal form.
2. Experimental
–
Photoelectrochemistry. The Br3 /Br– electrolyte used as
General procedures. All syntheses were performed
under nitrogen atmosphere for safety reasons unless otherwise
noted. Commercial reagents used for the syntheses were
purchased from Tokyo Chemical Industries Ltd., Nacalai
Tesque Inc. or Wako Pure Chemical Industries and used as
received. 18NR-T transparent and PST-400C opaque titania
pastes and FTO-coated TEC7 and TEC8 plates were purchased
redox shuttle for the solar cells was of the same composition as
for the above-described optoelectronic characterizations of the
dry TiO2 photoanode. Sensitization, sealing of the solar cells
and storage of the electrolyte and TiO2 photoanodes were
carried out under argon atmosphere in order to avoid exposure
to air and moisture. The TiO2 screens were sensitized over 6 h
with 3×10–4 M dye solutions with superdry CH2Cl2 as solvent
for 1Me-TfO, 1Me-TFA, 1Me-BF4 and 2Me-TfO, superdry
CH3CN for 1Me-Br and dry CCl4 for 1Dec-TfO for
maximization of the density of dyes at the surface and in
absence of coadsorbents. Subsequently, the cells were washed
with the respective sensitization solvent and dried under argon
flow. The sensitized photoanodes and the back-contacts were
sealed together with Surlyn film, and the electrolyte was
injected in the cells through pre-drilled holes at the
back-contact. At last, the devices were sealed with a thin glass
and Surlyn film before exposure to air for characterization. The
solar cells were irradiated with a CEP 2000 AM1.5 solar
simulator (Bunkoukeiki Co, Ltd., Japan) for every photocurrent
measurement (J-V curve under light and dark conditions, IPCE
spectra and electrochemical impedance spectroscopy (EIS)).
EIS measurements were performed with application of a
cathodic bias voltage equal to the VOC value under dark
conditions and without application of a cathodic bias voltage
under light (AM1.5 one sun) conditions, and collected with a
HZ-5000 automatic polarization system (Hokuto Denko).
1
from Dyesol Ltd. H NMR were recorded on JEOL JNM-400
400 MHz and VARIAN 400 MHz instruments with
tetramethylsilane (TMS, δ = 0 ppm) as reference. 13C NMR
spectra were recorded on a JEOL JNM-400 MHz instrument
with signal of solvents [CDCl3 (δ = 77.16 ppm) or CD3CN (δ =
116.79 ppm)]19 as reference. CDCl3 was used as a deuterated
NMR solvent of every synthetic intermediates and 1Dec-TfO.
All other malachite green's NMR samples were prepared with
dry CD3CN prepared as follows: 3Å molecular sieves20 were
heated in vacuo overnight to 200 °C, allowed to cool down to
room temperature and poured in a nitrogen-filled Schlenk flask.
A volume of commercial CD3CN up to twice the volume of
molecular sieves was poured inside under nitrogen flow, and
the system was left as-is for at least 24 h before use. High
resolution mass spectra (HRMS) were collected with a JEOL
mass spectrometer station JMS-700 through FAB+ and FAB–
mode using 3-nitrobenzyl alcohol as matrix.
Optoelectronic
characterizations.
UV/vis/near-IR
UV-3600
absorption spectra were recorded on
a
spectrophotometer (Shimadzu Corp.). Sample quartz cuvettes
with 1 cm path length sealed under argon atmosphere were
used absorption measurements of 10–5 M dye solutions in
superdry acetonitrile. Direct measurements of the absorption
spectra of sensitized TiO2 screens were performed under air
conditions, and the spectra were collected prior to and after
treatment with an electrolyte constituted of 1.00 M LiBr(H2O),
0.05 M Br2 in superdry acetonitrile. The dyes were desorbed
from the screen by dipping the photoanode into 0.06 mL of the
electrolyte solution, quenching the desorption at every 15 s
with dry acetonitrile and drying the cell under argon flow for
Zeta potential measurements. The dye@TiO2 samples
were prepared under argon atmosphere: into each 10–4 M dye
solution (10 mL) was injected a 10–8 M colloidal suspension of
TiO2 nanoparticles (0.5 mL), and the samples were left as such
after homogenization for at least 6 h for coverage equilibrium.
The resulting colloidal solutions containing 1 part of TiO2
nanoparticles for 104 parts of dye were kept in
ultrasonic-cleansed sealed flasks until use. Zeta potential
measurements were performed on an ELSZ-DN2 (Otsuka
Electronics, Japan) analyzer with bare and sensitized TiO2
anatase colloidal suspensions under neat conditions.
2

3. Results and Discussion
1Me-Br in 1:1 stoichiometry with the dye but did not appear in
any other dye characterization. Stabilization of the oxidized
form of the malachite green derivatives was observed under
argon atmosphere in polar, superdry solvents like acetonitrile,
benzonitrile or even methylene chloride. The low lightfastness
3.1 General synthetic route to malachite green derivatives.
The two-steps synthetic route followed broadly used ways to
synthesize malachite green, without any sharp modifications
(Figure 2). It provides a fast and inexpensive synthesis to small,
steady and panchromatic dye sensitizers: the first condensation
that was carried out with a catalytic amount of mild Lewis acid
ZnCl2,21, 22 at reflux for 24 h in an ethanolic solution gave the
Leuco compounds Leuco-Me, LeucoT-Me and Leuco-Dec in
65%, 52% and 39% yields, respectively. A solvent mixture of
ethanol/n-butanol (1/1, v/v) was used for the condensation of
the Leuco-Dec derivative for solubilization issues. The same
reaction carried out in pure butanol afforded the condensate in
10% yield only, suggesting that the hydrophilic character of the
solvent was of importance for the execution of the
condensation. The following two-electron oxidation process23
was merely carried out at room temperature with chloranil22, 24
as a mild oxidant with one equivalent of the corresponding
acids, namely triflic acid (TfOH), 42 wt% aqueous
tetrafluoroboric acid (HBF4), trifluoroacetic acid (TFA) or 48
wt% aqueous hydrobromic acid. Unlike the synthesis of
methine dyes carried out by Kast et al.24 in pure glacial acetic
acid as solvent (hence in great excess), only one equivalent of
each corresponding superacids was added to the medium in
order to avoid undesired protonation of the tertiary amine
function.25 The oxidation yields correlated neither with slowlier
kinetics due to the use of weaker acids and the size of the
N-substituents nor with any difficulty to oxidize due to steric
hindrance of the orthogonal substituents,26, 27 and most likely
relied to a large extent upon the different conditions for
recrystallization. Pure black flickering compounds were
obtained by slow elution with a solvent/non-solvent bilayer
solution: CCl4/hexane for 1Dec-TfO and CH2Cl2/hexane for all
2Me+ and 1Me+ derivatives except 1Me-Br that - due to
solubility reasons - was recrystallized in a benzonitrile/benzene
mixture. Oxidation trials with excess PbO2 in the same
conditions yielded the target compounds together with
decomposition impurities, consistent with the observations
made by Parvin, B. et al.28 Tetrachlorohydroquinone (Cl4HQ)
could not be isolated from the synthetic mixture in the case of
1Me-Br because of the poor solubility of the dye in low
polarity solvents, and was observed in the crystalline lattice of
products faded within a few hours in any solvent in the
29
presence of oxygen, in accordance with literature.2,
The
phenomenon was accelerated in low polarity solvents like
tetrahydrofuran or chloroform, consistent with a photolysis rate
rising with the concentration and lifespan of the dioxygen
excited singlet state30 or other radicals2 in low polarity solvents,
phenomenon also increased by moisture.2, 31
3.2 Molecular structure and spectral properties of the dyes.
An X-ray crystal structure analysis was performed on a
block-like single crystal of 1Me-Br recrystallized from
benzonitrile/benzene. The structure of this malachite green
dye was slightly modified in order to adapt for sensitization in
n-type Grätzel cells: A mere grafting of a carboxylic acid on
the para position of the phenyl group that is orthogonal to the
N1↔N2 intrachain provided the dye a strong adsorption
potential, red-shifted the S0→S1 (x band) absorption band32-34
by 10 nm (λmax(MG) = 619 nm,35 λmax(1Me+) = 629 nm) as
shown in Table 1 and induced a small distortion in the
triphenylmethine structure by comparison with the pure C2v
symmetry of the MG dye (Figure 4). In the solid state,
differences between bond angles of the central sp2 methine of
R
N
R
N
R
N
ZnCl2
R
R
R
O
+
Z
O
∆, alcohol
H
OH
Z
3a: R = Methyl
3b: R = n-Decyl
4a: Z = p-Phenylene (Ph)
O
OH
Leuco-Me: R = Me, Z = Ph
LeucoT-Me: R = Me, Z = Th
: 65%
: 52%
4b: Z = Thiophen-2,5-yl (Th)
Leuco-Dec: R = n-Decyl, Z = Ph : 39%
S
R
R
N
N
R
Chloranil
HA
R
–
1Me-BF4: R = Me, Z = Ph, A– = BF4
: 36%
1Me-TfO: R = Me, Z = Ph, A– = TfO–
1Me-TFA: R = Me, Z = Ph, A– = TFA–
1Me-Br: R = Me, Z = Ph, A– = Br–
A
: 62%
: 96%
: 46%
: 32%
: > 99%
Z
Figure 3. Single-crystal XRD-resolved coordinates of the
molecular structure of 1Me+ (from 1Me-BržCl4HQ salt) Top:
ORTEP drawing of the molecular structure with bond angles of
the N1- and N2-centered sp2 amine/aminium and C7-centered
methine systems. Hydrogens were omitted for clarity. Black =
carbon, Blue = nitrogen, Red = oxygen. Bottom: bond lengths
and dihedral angles between the three a, b and c phenyl planes.
The number in brackets for bond lengths and angles represents
the error on the last digit.
2Me-TfO: R = Me, Z = Th, A– = TfO–
1Dec-TfO: R = n-Decyl, Z = Ph, A– = TfO–
O
OH
Figure 2. General synthetic route to malachite green
derivatives 1 (Z = p-Phenylene) and 2 (Z = Thiophen-2,5-yl).
The monomethine dyes were isolated as triflate (TfO–),
tetrafluoroborate (BF4 ), trifluoroacetate (TFA–) and bromine
–
(Br–) salts.
3

1Me+ (C8 - C7 - C4 = 125.7(2)°, C4 - C7 - C14 = 114.7(3)°,
C14 - C7 - C8 = 119.5(3)°) and torsional angles between planes
(a ∧ b = 39.8(1)°, b ∧ c = 66.5(1)°, a ∧ c = 80.1(1)°)
outlined the distortion of the triphenylmethine system induced
by the intrinsic dissymmetry of the carboxylic acid (Figure 3)
and the heterogeneous distribution of the counterions. The
bromine atoms, although all equivalent in the P21/c lattice of
1Me-BržCl4HQ (Figure S1, ESI) were unequally distant to the
N1- and N2-centered aminium moieties of 1Me+: N2 - Br1
[4.593 Å, 5.103 Å, 4.878 Å] and N1 - Br1 [5.123 Å, 5.439 Å].
Hence, the cationic charge was greater on N2 as witnessed by
the different bond lengths in the aniline moieties: the quinonoid
character of phenyl b was more important than for phenyl a (C7
- C8 [1.400(4) Å] < C7 - C4 [1.444(4) Å], C9 - C10 [1.361(5)
Å] < C3 - C2 [1.372(4) Å], C12 - C13 [1.357(4) Å] < C5 - C6
[1.364(4) Å], C11 - N2 [1.334(4) Å] < C1 - N1 [1.362(5) Å])
and at the same time participates at the destabilization of the
cationic π -system. The global distortion induced by the
electronic effect of the carboxylic acid was only small
compared to the environment heterogeneity observed in the
crystal, as observed by DFT studies: the distribution of the
HOMO isodensity was homogeneously spread over the cyanine
chain when optimizing the structure by DFT in adiabatic
conditions (Figure 4). Furthermore, the dihedral angle between
planes a and b of 1Me+ was observed much larger than in the
crystal structure (+ 9.4°) whilst dihedral angles a ∧ c and b
∧ c were reduced by c.a. 18.1° and 4.6°, respectively. As
seen in Figure 5, the substitution of the orthogonal phenyl ring
by a thiophen-2,5-yl moiety red-shifted λmax by 2 nm only (λmax
(2Me+) = 631 nm). However the substitution of the four
N-methyl groups by N-decyl substituents induced an 18 nm
red-shift (λmax (1Dec+) = 647 nm), consistent with literature36
and in accordance with the Dewar-Knott empirical rules for
spectral changes relative to steric hindrance and electronic
effects.1, 37 The observed λmax (x band) and y band red-shifts
were mostly induced by the more electron-inductive character
of the N-decyl substituents for 1Dec+ and the more
electron-accepting thiophene moiety in 2Me+, respectively. The
twisting of the a and b phenyl rings did not result in a more
localized electronic charge on the auxochromic nitrogens,
phenomenon observed by D. R. Waring & al26 and likely had
no bearing on the spectral parameters of these cyanine
chromophores. As observed by density functional theory
(Figure 4 and Table S3, ESI): the Mulliken charges on the
auxochromes were calculated equivalent (δ(N) = –0.497) in
2Me+ and 1Me+ despite a +4.2° bathochromic torsional strain
of the a ∧ b angle in the 2Me+ derivative. The N-decyl
substituents were responsible for the more important Mulliken
charges (δ(N1) = –0.528, δ(N2) = –0.526) on the auxochromes
in 1Dec+.
Figure 4. DFT-optimized molecular structures of the a) MG, b) 1Me+, c) 2Me+ and d) 1Dec+ cations. Top left: Mulliken atomic
charge numbers; the atoms are coloured over a range of colours from red (−0.650) to green (+0.650). The origin of the permanent
dipole moment vector is the center of the global electronic charge. Top right: front view of the HOMO (blue) and LUMO (red)
isodensities, computed by DFT calculations with B3LYP 6-31G**38 as exchange correlation functional and basis set; isodensity
contour (solid) = 0.02. Absolute values of the bond angles (bottom left, green) and dihedral angles (bottom right, orange) at the
planar aminium and central carbon atoms are displayed in degrees. The long alkyl chains of 1Dec+ were truncated in the present
figure for clarity.
4

Table 1. Optoelectronic parameters of the chromophores in solution.
x bandi
[nm]
εmax (x)
y bandi
[nm]
εmax (y)
E(S+/S)
E(0-0)
E(S+/S*)
j
k
l
Dye
[M-1cm-1]
[M-1cm-1] [V vs NHE] [V] (Abs/Em)
[V vs NHE]
629
647
631
50050
46190
29500
427
430
482
10940
8220
1.41
1.43
1.44
1.88
1.84
1.75
-0.47
1Me+
1Dec+
2Me+
-0.41
-0.31
11570
i Wavelengths of peak maxima collected from absorption spectra of dyes in superdry acetonitrile solutions.
j The ground state redox potentials, measured with cyclic voltammetry (CV) under the following conditions: Pt counter electrode and
Pt working electrode, Ag/AgCl reference electrode; electrolyte, 0.1 M recrystallized TBAClO4 in benzonitrile (the benzonitrile was
chromatographed on Alumina N Akt. I column before use. Potentials measured vs Ag/AgCl were corrected with ferrocene Fc/Fc+
(E(Fc/Fc+) = 0.48V) and converted to NHE by addition of +0.241V. Cyclic voltammograms of the dyes are shown in Figure S2 (ESI).
k 0-0 transition energy (E(0-0)) was issued from the energy value at 10% of the main peak intensity.
l LUMO energies (E(S+/S*)) estimated vs NHE from the ground state oxidation potential added to E(0-0)
.
chromophore and its counterion only depended on one another.
Long-term cell degradation involving the device sealing,
leakage and dye/electrolyte degradations were left outside of
the scope of this study. Figure 6 highlights the evolution of the
absorption patterns of the coloured TiO2 screens as a function
of the immersion time in the DSSC electrolyte (1.00 M
LiBr(H2O) / 0.05 M Br2 in superdry CH3CN): The screens were
dipped in the electrolyte to desorb the dyes for 15 s and rinsed
in pure, superdry CH3CN to quench the desorption to carry out
the UV-Vis measurement (no desorption was observed in pure
CH3CN). This simple operation was repeated over a total 3 min
time span to observe the desorption patterns of the dyes on a
short time scale. For given chromophores, the desorption rates
sharply differed from one counterion to another: for instance,
the 1Me-BF4-sensitized TiO2 screen was utterly bleached after
the
3 min time span, suggesting that nearly all the
chromophores were desorbed whereas the 1Me-TFA-sensitized
screen remained dark green. Although the chromophore's
structure (1Me+) remains unchanged and hence its
physico-chemical properties were identical, the desorption rates
were observed quite different. This observation underlined the
fact that the use of a lone carboxylic acid as the anchor leaves
these chromophores highly dependent to small changes in their
environment.
Figure 5. UV-Vis spectra of 10–5 M dye solutions in superdry
CH3CN. All measurements were performed under argon
atmosphere with sealed 1 cm path length quartz cuvettes.
In
a
mesoporous semi-infinite domain, namely,
3.3 Lability of the chromophores in DSSC conditions
nanocrystalline TiO2 considered as an impermeable solid and
the electrolyte as the bulk in the present case, the space
distribution of ions proximate to the interface is function of
cohesive and dispersive forces. Unfortunately, under such high
41
Unlike the black dye,39 N71940,
or most Ruthenium
polypyridyl complexes16, 42 developed for DSSC sensitization,
the majority of developed organic dyes are designed with a
lone carboxylic acid as an anchoring moiety for synthetic
simplicity and efficiency.43 Without any efficient chelation to
contribute to the stabilization of the dyes at the
electrolyte/dye/TiO2 interface, the dye has a propensity to be
driven away to the bulk. More specifically, the dye adsorbed at
the surface is in equilibrium with the dye dissolved in the
electrolyte. J. H. Park et al.44 directly addressed this problem by
merely adding some dye to the electrolyte prior to mounting the
cell and showed a significant stabilization of every parameter
of the cells. The thermodynamics of the dye
adsorption/desorption process are significantly dependent -
ionic strength (>
1 M) of the DSSC condition the
Debye-Hückel point-charge approximation47 doesn't hold and
the ionic interactions ought to be considered and thus neither
the Gouy-Chapman,48,
the Grahame Langmuir,50 the
49
hypernetted-chain theory based on the Ornstein-Zernike
equations51 nor any other theory can accurately describe the
system from a purely theoretical point of view, to the best of
our knowledge.52-55 Ionic adsorbates binding to the surface of a
particle form a quasi-monolayer governed by their relative size
and charge:56 the bounding process can be described by a
Langmuir-type adsorption isotherm relating fractional surface
coverage Θf to the adsorbate concentration C with respect to
the equation:
among many parameters - on heat, illumination, nature of the
45, 46
solvent and surface-active species41,
and therefore all
characterizations of the dye@TiO2 systems in this study were
performed at room temperature, without any surface-active
species in the electrolyte nor any coadsorbent other than the
chromophore and its counterion. The chromophore's behaviour
was thus function of the nature of its lone counterion and
reciprocally, i.e. the adsorption and desorption rates of the
ꢀꢁ
ꢀ! =
1 + ꢀꢁ
with κ as the ratio of adsorption/desorption for a given ionic
5

Figure 6. Plotting of the desorption of the dyes from the TiO2 nanoparticles screen over time when dipped into 0.06 mL of an
electrolyte constituted of 1.00 M LiBr(H2O) / 0.05 M Br2 in superdry CH3CN, followed by UV-Visible absorption spectra
measurement. The absorption spectra were recorded every 15 seconds over a 3 min time span, as the desorption was quenched by
rinsing the cells in superdry CH3CN.
species. Only in the case of an ionic pair A the adsorption of
one of the two ions encompasses the properties of both, and
κ can be expressed as:
(adsorption, electrostatic, van der Waals and so on) are
balanced out by the diffusive forces such as osmosis, solvation,
electrostatic repulsion. Sensitization with the TfO- and
TFA-based species (1Me-TfO and 1Me-TFA) shifted the zeta
potential towards more positive values whereas the BF4- and
Br-based species (1Me-BF4 and 1Me-Br) shifted the zeta
potential of the nanoparticles downward (Figure 8, left and
Table 2), indicating that the excess species in the diffuse layer
were the chromophores 1Me+ and 2Me+ in the case of TfO-
ꢀꢁ
ꢀ =
ꢀ ꢀ
with [AS] and [S] corresponding to the surface densities of all
occupied and unoccupied sites at the surface of the particle,
respectively. In the case where both the cation and the anion
possess an adsorption potential, the adsorption processes
cannot be merely described inasmuch as (i) the fractional
surface coverages may depend on one another, (ii) the versatile
adsorbates may occupy different adsorption sites and (iii) the
interaction between adsorbates may not be neglected. However,
in the globally electroneutral system the concentration of one
ion at the surface raises an electric potential Φ0 and implies the
building of an opposed nearby potential Φdiff (Figure 7). The
zeta (ζ) potential of sensitized nanoparticles in colloidal
suspensions, related by nature to those potentials, was
experimentally available to qualitatively determine the excess
species in the diffuse layer and thus at the surface. As
previously seen for other streptocyanine derivatives,57 the zeta
–
and TFA-based species and the photoinactive moieties BF4
and Br– for the compounds 1Me-BF4 and 1Me-Br, respectively.
–
BF4 is generally considered as a non-complexing anion58 and
likely to be indifferent to the TiO2 anatase surface in the
present study. The Br– anion may display weak interactions
with the surface in such anhydrous and concentrated
conditions,59 however these can be neglected compared to the
adsorption
potential
of
1Me+.
Therefore,
the
adsorption/desorption kinetics of their respective 1Me-BF4 and
1Me-Br salts mostly relied upon the adsorption strength of
1Me+ and the affinity to the bulk of both the anion and the
–
cation. After 1 min dip in the Br3 /Br– electrolyte, the
absorption intensity of the monomer peak of 1Me-BF4 (624
nm) decreased by 76% against a 23% drop for 1Me-Br (Figure
8, right). Zeta potential measurements of 1Dec-TfO-sensitized
nanoparticles suspended in superdry acetonitrile were
unsuccessful: the electro-optic sampling (EOS) signal was
nearly flat, probably accounting for the insolubility of such
lipophilic dye@TiO2 system, but this property was also likely
responsible for the best stability observed for such system. The
intensity of the monomer peak of 1Dec+ (654 nm) only
decreased by 3.5% after 1 min dip in the electrolyte and by
15% after 3 min dip against a 65% drop for the methyl
derivative 1Me-TfO. Interestingly, the 1Me+ chromophore was
best stabilized with TFA– as counterion: the derivative
1Me-TFA only displayed a 29% drop in intensity (Figure 8).
potential shift (Δζ) relative to bare TiO2 nanoparticles (ζTiO2
=
4.33 ± 2.05 mV) is directly correlated to the linkage patterns of
the dye and of the same sign (positive or negative) as the
excess species in the diffuse layer. Owing to electroneutrality,
the global concentrations of the cations and the anions are
equal and thus, the majority species in the diffuse layer is
minority adjacent to the surface and reciprocally. In one case,
both moieties of the ionic pair are potentially bounded to the
surface (Figure 7a), and in the other case only the cation
possess an adsorption potential and the free anion remains
attracted to its counterion via Coulombic (electrostatic)
interaction (Figure 7b): At the equilibrium, attractive forces
6

Figure 7. Representation of the surface potential Φ patterns with respect to the distance to the surface y and function of the ionic
distributions a) at the surface for TiO2 coadsorbed by cations and anions and b) at the surface where cations adsorbed at the TiO2
surface and free anions with ΔΦ as the difference between the surface potential Φ0 and the diffuse layer potential Φdiff. Only the
excess species are represented in the diffuse layer and the bulk.
Figure 8. Left. Averaged values of at least two zeta potential measurements of colloidal suspensions of bare - and sensitized with
2Me-TfO, 1Me-TfO, 1Me-TFA, 1Me-BF4 and 1Me-Br - TiO2 anatase nanoparticles in superdry acetonitrile. Right. Plotting of the
desorption of the dyes from the TiO2 screen when dipped in an electrolyte constituted of 1.00 M LiBr(H2O) and 0.05 M Br2, followed
by UV-Vis absorption measurements after quenching the desorption every 15 s by washing with dry acetonitrile (See also Figure 6).
The absorption values at λmax of the dyes@TiO2, namely 596 nm for 2Me-TfO, 624 nm for 1Me+-based dyes and 654 nm for
1Dec-TfO, were normalized with respect to the initial value before dipping.
The affinity for the bulk/affinity for the surface ratio
decreases in the order BF4– > TfO– > Br– > TFA– in these 1Me+
salts and the minimum ratio was obtained with a carboxylic
acid-anchored anion (TFA–). The most stable adsorption mode
of TFA– calculated by De Angelis et al.60 was a bridged
bidentate mode whereas the sulfonate of TfO– was likely
adsorbed at the surface via a weaker bridged bidentate mode
assisted by hydrogen bonding,61 also consistent with the lower
desorption rates of TFA- and TfO-based dyes. Although
stabilized thanks to a work on the nature and properties of its
counterions, the 1Me+ chromophore could not be fully
stabilized at the TiO2 surface and better results could be
obtained with n-decyl substituents as previously discussed with
1Dec+ or with the thienyl derivative 2Me+. The three
chromophores 1Me+, 1Dec+ and 2Me+ were linked to the TiO2
surface with a lone carboxylic acid, usually adsorbed onto the
surface in a dissociative bridging bidentate mode, with one
proton transferred to a nearby surface oxygen.60, 62 However,
FT-IR studies did not help witnessing the adsorption mode of
the dyes as the empirical Deacon and Philips attribution of any
νasCOO or νsCOO frequencies,63 that was hindered by the
overcharged spectra (Figure S3, ESI), and all peaks from 1800
to 3600 cm–1 in the dye@TiO2 spectra were flattened out after
adsorption on TiO2 with the exception of the νC-H in the range
2800 - 3000 cm–1.16 In the case where the counterion is a
triflate, the difference between desorption rates of all three
7

chromophores (1Me-TfO, 1Dec-TfO and 2Me-TfO) mostly
relied upon their interactions with adsorbed co-ions and their
affinity to the bulk. More specifically, the chromophores were
stabilized at the surface either isolated (monomers) or as H-
and J-aggregates, as witnessed by the UV-Vis absorption
patterns of the dyes adsorbed on TiO2. Except for 1Dec-TfO
for which only H-aggregation was observed with a shoulder
around 580 nm, all dyes exhibited a monomer peak around 640
~650 nm with two shoulders arising from H- and J-aggregation,
at shorter and longer wavelength, respectively. During the
desorption measurements, the intensity of the shoulder around
700 ~750 nm arising from the J-aggregation of the dyes
decreased quicker than both the monomer peak and the
shoulder of H-aggregated dyes around 550 nm. This drop was
pre-eminent for the 1Me-TFA, 1Me-Br and 2Me-TfO
derivatives and lesser for 1Me-TfO and 1Me-BF4. These
observations suggested that the aggregation modes of the dyes
may have a bearing on their stabilization at the surface and that
J-aggregated ones had a greater propensity to desorption.
characteristics curves of the cells (Figure 9, left) dipped in a
–
Br3 /Br–-based electrolyte. Owing to its higher lying redox
–
couple's energy level (E(Br3 /Br–) = +1.09 V vs NHE), the
open-circuit voltage (VOC) of the cells greatly improved from a
best 0.20 V to 0.50 V when using this redox shuttle instead of
–
the commonly used I3 /I– couple. The wider energetic gap,
allowing VOC values as high as 1.21 V for a DSSC with this
–
Br3 /Br–-based electrolyte as shown by Hanaya, M. et al,64
between the electrolyte and the TiO2 conduction band largely
compensated for the small loss in short-circuit current density
–
(JSC). Sugihara et al.3 first demonstrated in 2005 that a Br3 /Br–
–
redox shuttle should be preferred to I3 /I– for sensitizers having
a ground state oxidation potential more positive than that of
–
Br3 /Br–, which was the case of the three 1Me+, 1Dec+ and
2Me+ sensitizers: driving forces for electron recombination ∆E
were calculated to be 0.32, 0.34 and 0.35 V, respectively with
respect to the equation ∆E = E(S+/S) – E(Br3/Br), where the
ground state redox potentials of the dyes E(S+/S) were 1.41, 1.43
and 1.44 V, respectively, as displayed in Table 1. Moreover,
the photovoltaic activity of the cells were best improved at
higher concentrations of the electrolyte (1.00 M LiBr(H2O) /
0.05 M Br2) than the electrolyte used by Sugihara et al.,
constituted of 0.4 M LiBr and 0.04 M Br2 for their study with
the well-known Eosin Y dye.3 Higher concentrations of
electrolyte did not lead to any improvement of the cell
parameters due to limitations in recombination kinetics
between the electrolyte and the oxidized dye. Further trials for
Table 2. Influence of the counterion natures on the ζ potential.
Dye@TiO2
ζi (mV)
∆ζi (ζi − ζTiO2, mV)
+11.38 ± 3.45
2Me-TfO
1Me-TfO
1Me-TFA
1Me-BF4
1Me-Br
+15.71 ± 1.40
+7.98 ± 0.14
+8.88 ± 2.54
+2.71 ± 0.49
–4.08 ± 1.57
+3.65 ± 2.19
improvement of the VOC value with
a
well-known
CoII/IIItris(bipyridyl)65 electrolyte constituted of 0.25
M
+4.55 ± 4.59
–1.62 ± 2.54
[CoII(tbp)3](PF6)2 / 0.025 M [CoIII(tbp)3](PF6)3 in superdry
acetonitrile did not lead to any significant photocurrent. The
TiO2 screens were sensitized in 0.3 mM dye solutions without
any coadsorbent: for instance, the photoconversion was nearly
quenched with addition of chenodeoxycholic acid (CDCA) in
the dyeing bath. Furthermore, the addition of any coadsorbent
might have impaired the study of the influence of the different
TfO–, TFA–, BF4– and Br– counterions on the cell's parameters.
As witnessed by comparison of the UV/vis/near-IR
absorption spectra of the dye@TiO2 screens (Figure 6) and the
photocurrent action spectra (Figure 9, right), J-aggregates,
although allowing light harvesting at longer wavelength, barely
induced any photocurrent. J-aggregated cyanine dyes are
–8.41 ± 3.62
The zeta (ζ) potential values are the average of at least two
consecutive measurements in the same conditions with the
corresponding standard deviation, and the combined standard
deviation for the differential values ζi - ζTiO2 (ζTiO2 = 4.33 ±
2.05 mV).
3.4 Photovoltaic performances of DSSCs.
The photocurrent action spectra of the dye-coated
nanocrystalline TiO2 stacked electrodes are displayed in Figure
9, right together with the photocurrent (J) - voltage (V)
Figure 9. left: Photocurrent (J) -voltage (V) characteristics curves under AM 1.5 artificial sunlight source (plain) and dark conditions
(dotted line) and right: IPCE action spectra for nanocrystalline TiO2 solar cells sensitized under argon with the dye sensitizers dipped
in 1.00 M LiBr(H2O) / 0.05 M Br2 in superdry CH3CN. The error bars represent the standard deviation on the measurements obtained
with at least two cells prepared under the very same condition. No further additives were used in the electrolyte.
8

and significantly blueshifted compared to the y bands in
solution, 482, 430 and 427 nm for 2Me+, 1Dec+ and 1Me+
(Figure 5), respectively, although a red-shift was observed at
high surface coverage on TiO2 for this very band around 450
nm for 1Dec+ and 1Me+ and plateaued from 440 to 520 nm for
2Me+ (Figure 6). This discrepancy can be explained from the
fact that the UV/vis/near-IR measurements were performed
under dry conditions and that the y band, with a strong
charge-transfer character is highly sensitive to its environment
and largely blue-shifted under DSSC conditions, namely, when
the solar cell is dipped into the highly ionic electrolyte. The
usually photoactive in DSSCs, as formerly demonstrated with
67
the efficient asymmetrical closed-chain cyanines,66,
69
asymmetrical streptocyanines,57 and hemicyanines,68,
outlining their utility for sensitization of n-type Grätzel cells. In
the present case their inactivity is presumably due to the
aggregation-induced lack of driving force for electron
migration from the lower-lying redox level of the J-aggregates
excited state to the conduction band of TiO2. The excited state
level of the monomer E(S+/S*) was determined to be −0.47,
−0.41 and −0.31 V vs NHE, respectively, for the chromophores
1Me+, 1Dec+ and 2Me+ by adding the zeroth-zeroth transition
energy E(0-0) to the ground state oxidation potential E(S+/S)
(Table 1). Owing to the ultrafast nonradiative relaxation
dynamics of malachite green derivatives impeding fluorescence
spectroscopy in the standard conditions as largely reported in
the scientific literature,2, 32-34, 70-72 E(0-0) was estimated from the
energy value equivalent to the wavelength at 10% of the
S0→S1 (x) band of the monomers in solution. Those values of
accepting/anchoring moiety
- orthogonal to the N↔N
intrachain - of every dye bears the largest isodensity coefficient
of the LUMO+1 (S2) state and allows facile injection into the
semiconductor's conduction band. The efficiency (η) of the
DSSCs sensitized with 1Me+ salts ranged from 0.37(1)% for
1Me-TFA to 0.13(4)% for 1Me-TfO, pointing out the role of
the counterion not only on the cell stability as previously
discussed but also on the cell's efficiency (Table 3). The better
efficiency obtained with carboxylate-anchored (TFA–) and
sulfonate-anchored (TfO–) counterions were consistent with the
results of Huang and coworkers69 with benzothiazolium
hemicyanine dyes, albeit these dyes' counterions were
covalently bound to the chromophore thanks to a short aliphatic
chain. Devices sensitized with the 1Dec+ and 2Me+
chromophores showed lower photovoltaic conversion
efficiencies (PCE) due to the thermodynamically unfavourable
electron transfer from the dyes' excited states to TiO2,
correlating with the down shift of the dark currents and lower
open circuit voltage (VOC) values despite longer electron
recombination lifetimes for both derivatives, respectively 0.465
and 0.527 ms (Fig. S7 and Table 3) that were estimated using
the Bode phase plot mentioned below. The best electron
lifetime at the TiO2/dye/electrolyte interface with 1Me+
derivatives was observed for 1Me-TfO and 1Me-Br with 0.379
ms. Fill factors as high as 0.65 could be achieved with both
1Me-TFA or 1Me-BF4 and barely dropped to 0.61 for both
E
(S+/S*), too close to the Fermi level of TiO2 at −0.50 V vs
NHE73 did not allow a saturation of the IPCE signal despite the
high lithium concentration (1.00 M) in the medium74 and
thereby highlighted the different contributions of every
electronic transition of monomeric and aggregated species to
give rise to a photocurrent in the experiment conditions. Still,
part of the low IPCE also accounted for the low light harvesting
efficiency (LHE) of the devices (Figure S5, ESI) in spite of the
good extinction coefficient of the dyes (Figure 5): Most optical
densities of the dye monolayer@TiO2 plateaued at 35%, except
1Me-Br and 1Me-TfO which displayed lower - 33% and 29%,
respectively - LHEs at the maximum absorption wavelength.
The best IPCE observed for every dye was observed around
330 nm, with a maximum value of 17% for the 1Me-TFA
derivative. At longer wavelength, the best photocurrents were
attributed to the blue-shifted vibronic shoulder with absorption
maxima ranging from 565 nm (1Me-BF4) to 590 (1Dec-TfO)
nm of the H-aggregated 1Me+ and 2Me+ chromophores, on
average 1.5 times higher than the IPCE arising from the
monomeric x band around 635 nm. The better photocurrents
observed were attributive to the higher oxidation potential of
the dyes in the H-aggregated regions of the self-assembled
monolayer (SAM).4-6, 12 Interestingly, the coadsorption of the
chromophore and its counterion did not impede nor favoured
the H- or J-aggregation. The photocurrents arising from the
S0→S2 (y band) in the 410 - 440 nm region were as high as the
photocurrents induced from the H-aggregates (Figure 9, right),
1Me-TfO and 1Me-Br, demonstrating
a
good charge
separation and screening at the TiO2/dye/electrolyte interface
and induced better series and shunt resistances for a malachite
green-based device compared to the reported SnO2-liquid
junction8 or porous CdS7 technologies of DSSC.
Table 3. Photochemical properties of the cells under simulated AM 1.5 irradiation.
Dye
VOC
(V)
JSC
VMAX
(V)
JMAX
τ
(ms)
η
(%)
ff
(mA cm–2)
(mA cm–2)
1Me-TFA 0.50 ± 0.01
1Me-BF4 0.48 ± 0.01
1Me-TfO 0.47 ± 0.00
1.15 ± 0.03
0.70 ± 0.05
0.48 ± 0.15
0.49 ± 0.00
0.24 ± 0.04
0.17 ± 0.02
0.39 ± 0.02
0.38 ± 0.00
0.36 ± 0.00
0.37 ± 0.02
0.29 ± 0.01
0.29 ± 0.04
0.95 ± 0.01
0.57 ± 0.01
0.37 ± 0.11
0.38 ± 0.00
0.21 ± 0.07
0.12 ± 0.02
0.65 ± 0.02
0.65 ± 0.04
0.61 ± 0.02
0.61 ± 0.01
0.58 ± 0.03
0.53 ± 0.04
0.290
0.366
0.379
0.379
0.465
0.527
0.37 ± 0.01
0.22 ± 0.01
0.13 ± 0.04
0.14 ± 0.00
0.06 ± 0.01
0.04 ± 0.01
1Me-Br
0.46 ± 0.00
1Dec-TfO 0.40 ± 0.02
2Me-TfO 0.40 ± 0.04
VOC: open circuit voltage; JSC: short circuit current density; ff = (VMAX × JMAX)/(VOC × JSC): fill factor; τ: electron recombination
lifetime; η: photoconversion efficiency. The error numbers correspond to the standard deviation of the measurements of at least two
–
solar cells performed under the exact same conditions. All cells were fabricated under argon atmosphere were performed with a Br3
/Br– electrolyte constituted of 1.00 M LiBr(H2O) and 0.05 M Br2 in superdry acetonitrile as electrolyte.
9

Electrochemical impedance spectroscopy was carried out
under AM 1.5 one sun and dark conditions to visualize the
charge-transfer/recombination processes in the devices, and
shown in Figure 10 as Nyquist plots and in Figure S7 as Bode
current densities: an increase in Jsc induces a concomitant
diminution of the electron lifetime and inversely.77
The observed VOC values did not follow the same trend and
most likely arose from
a
limitation in recombination
–
phase plots. In a device working with a Br3 /Br– redox shuttle,
kinetics between the dye and the electrolyte as formerly
observed by Detty, M. et al with chalcogenorhodamine
the large semicircle corresponds to the charge-transfer
resistances at the TiO2/dye/electrolyte interface and generally
5
dyes/iodine-iodide as sensitizer/electrolyte couples,4,
–
bigger than with an I3 /I– electrolyte.3 The charge-transfer
howbeit monotonically increased together with JSC and
hence with the electronic population in the conduction band
of TiO2. The influence of the counterion on the charge transfer
resistance was, in the present case more important than the
induced loss in injected electron lifetimes. The origin
resistances at the interface under both AM 1.5 one sun and dark
conditions increased, for the 1Me+ derivatives in the order
1Me-TFA < 1Me-Br < 1Me- TfO < 1Me-BF4, outlining the
strong effect of the counterion on the electron transfers.
1Me-TFA showed the lowest charge transfer resistance,
consistent with the increase in Jsc75 and contributed at
demonstrating that TFA– was best suited for this chromophore.
We also estimated the electron recombination lifetimes τ at the
TiO2/dye/electrolyte interface from the low angular frequency
peaks in the Bode phase plot according to the Kern equation,76
that leads τ = 1/(2πf), where f is the frequency peak of the
superimposed ac voltage (Fig. S7, ESI and Table 3). The
lifetimes of 1Me+ salts are estimated to be 0.366, 0.379, 0.379
of
the
difference
in
charge transfer/recombination
ratio at the electrolyte/1Me+/TiO2 interface in function
of the nature of the counterion could hardly be discussed
at the present state of studies because several factors like
partial dye desorption and difference in ionic distribution
of the diffuse ion swarm proximate to the surface could
not be set apart. Indeed, the impact of the electroactivity
of the counterions on both the electric double layer, the
surface quantum states and passivation of the surface were
correlated to each other, and those quantities had to be
considered together for discussion. Further experiments are
required to distinguish and clear out the implications of each
one of these parameters and explain the ordering of
the photoconversion efficiencies: 1Me-TfO < 1Me-Br < 1Me-
BF4 < 1Me-TFA. Anyway, the IPCE spectra showed a
global raise/drop of the photoaction over the whole visible
range while switching counterions, most likely ruling out
their impact on the propensity to aggregate and the type of
aggregation of the chromophores and impacting the general
thermodynamics of electron transfers.
–
and 0.290 ms for the BF4 , Br–, TfO– and TFA– derivatives
under AM 1.5 one sun illumination, respectively. They are
correlated to recombinations kinetics of the injected electrons
and in stark contrast with the cells parameters: The TfO– and
Br– salts showed the same, longest electron lifetimes (ca. 0.379
ms) and gave similar Voc, Jsc, desorption rates and
–
photoconversion efficiencies whereas the BF4 salt yielded a
slightly lower electron lifetime (0.366 ms) altogether with an
increase in Voc. Likewise, 1Me-TFA showed the best
stabilization on TiO2 and the best Voc, inconsistent with its
short electron lifetime. Evidently, in this case the lone electron
lifetimes did not reflect the cell behaviors with such different
Figure 10. Nyquist plots of the solar cells measured under AM 1.5 one sun (left) and under dark (right) conditions.
4. Conclusion
1Me-BržCl4HQ revealed, to the best of my knowledge the first
crystal structure determined for a malachite green derivative.
Determination of the potential by electrophoresis
Colouring of TiO2-based n-type DSSC cells was performed for
the first time with malachite green derivatives thanks to a mere
grafting of carboxylic acid on the electron accepting/anchoring
moiety of the chromophores. The 1Me+, 1Dec+ and 2Me+
chromophores were synthesized in a cost-effective two-steps
ζ
measurements on colloidal solutions of sensitized nanoparticles
qualitatively unveiled the electric double layer patterns at the
nanoparticle/bulk interface specific to the adorption-capable
ionic pairs such as the present dyes. A chromophore coupled to
a free, easily solvable anion like 1Me-BF4 quickly desorbed
synthetic route and isolated as TFA–, TfO–, BF4 and Br– salts.
–
DFT methods showed that the electronic isodensity of the
different ground and excited states of the dyes looked alike
from one chromophore to another and explained, to a large
degree their similar optoelectronic properties. Investigation of
the structure of 1Me+ by X-ray diffraction on a single crystal of
–
from the surface in an acetonitrile-based concentrated Br3 /Br–
DSSC solvent. On the other hand, the hydrophilicity of Br– and
the mild adsorption potential of TfO– allowed 1Me+ to desorb
at a lower rate, and the strong adsorption potential of TFA–
10

afforded the best stabilization for this chromophore.
Nonetheless, better results were achieved with the 2Me+ and
1Dec+ derivatives, and the substitution by n-decyl was found
critical for stabilization of the dye at the interface thanks to an
enhanced lipophilicity, disrupting its ability to diffuse away in
the highly ionic electrolyte. On the other hand, no further
experiments were carried out to fully explain the better
stabilization of the 2Me+ derivative. Antagonistically, the
stabilization of these derivatives did not improve the DSSC
photoaction due
a mismatch between energy levels: A
necessary driving force for electron injection of ca. 0.2 V
between the excited state of the dye and the Fermi level of TiO2
should be observed for achievement of high IPCEs.73 We
presently aim at designing and characterizing malachite green
derivatives with more negative excited state redox levels, more
adapted to DSSC use.
Supporting Information
Details of X-ray data collection and reduction for the single
crystalline sample of 1Me-BržCl4HQ. Detailed syntheses of all
1
compounds, H/13C NMR spectra of the dyes and syntheses
intermediates, cyclic voltammetry of the chromophores, FTIR
spectroscopy of the dyes in solid state dyes@TiO2 and
syntheses intermediates and raw plotting of the intensity of the
zeta potential values and Bode phase plots of the DSSCs.
Acknowledgement
This work was financially supported in part by Grants-in-
Aid for Scientific Research (No. 15K05483) from the
Ministry of Education, Culture, Sports, Science and
Technology of Japan.
11

References
1
M. J. S. Dewar, J. Chem. Soc. 1950, 2329-2334.
D. F. Duxbury, Chem. Rev. 1993, 93, 381-433.
Z.-S. Wang, K. Sayama, H. Sugihara, J. Phys.
Goldberg, Organometallics 2010, 29, 2176-2179.
20 D. B. G. Williams, M. Lawton, J. Org. Chem. 2010,
75, 8351-8354.
2
3
Chem. B 2005, 109, 22449-22455.
21
D. Zhu, Y. Mu, Q. Li, X. Pang, X. Wang, Y. Liu, G.
4
K. R. Mulhern, M. R. Detty, D. F. Watson, J. Phys.
Chen, H. Chen, Anal. Methods-UK 2016, 8, 6236-6243.
Chem. C 2011, 115, 6010-6018.
22
W. Müller, I. Hattesohl, H. J. Schuetz, G. Meyer,
5
K. R. Mulhern, A. Orchard, D. F. Watson, M. R.
Nucleic Acids Res. 1981, 9, 95-119.
Detty, Langmuir 2012, 28, 7071-7082.
23
J. E. Kuder, W. W. Limburg, M. Stolka, S. R.
6
M. W. Kryman, J. N. Nasca, D. F. Watson, M. R.
Turner, J. Org. Chem. 1979, 44, 761-766.
Detty, Langmuir 2016, 32, 1521-1532.
24
25
H. Kast, Patent US4000135 A, 1976.
7
H. Abdulelah, B. Ali, M. A. Mahdi, A. Q. Abdullah,
N. M. Cuong, S. Ishizaka, N. Kitamura, American
J. J. Hassan, H. F. Al-Taay, P. Jennings, Journal of Solar
Energy 2016, 2016, 7.
Journal of Analytical Chemistry 2012, 3, 18256 (7 pages).
26 G. Hallas, K. N. Paskins, D. R. Waring, J. Chem.
Soc., Perkin Trans. 2 1972, 2281-2286.
8
M. Shimura, K. Shakushiro, Y. Shimura, J. Appl.
Electrochem. 1986, 16, 683-692.
27
S. G. R. Guinot, J. D. Hepworth, M. Wainwright, J.
9
P. Ren, Y. Li, Y. Zhang, H. Wang, Q. Wang, Int. J.
Chem. Soc., Perkin Trans. 2 1998, 297-304.
Photoenergy 2016, 2016, 2135847.
28
A. G. Kraus, I. Jeon, M. Nilsen-Hamilton, M. A.
10
M. S. Roy, P. Balraju, M. Kumar, G. D. Sharma,
Awad, J. Banerjee, B. Parvin, Molecules 2008, 13.
Sol. Energy Mater. Sol. Cells 2008, 92, 909-913.
29
L. Yong, G. Zhanqi, J. Yuefei, H. Xiaobin, S.
11
G. Ramakrishna, H. N. Ghosh, A. K. Singh, D. K.
Cheng, Y. Shaogui, W. Lianhong, W. Qingeng, F. Die, J.
Hazard Mater. 2015, 285, 127-136.
Palit, J. P. Mittal, J. Phys. Chem. B 2001, 105, 12786-12796.
12
P. M. Sirimanne, Renew. Energ. 2008, 33,
30
Bensasson, J. Am. Chem. Soc. 1995, 117, 9159-9164.
31 G. Schwen, G. Schmidt, J. Soc. Dyers Colour. 1959,
75, 101-105.
J.-M. Aubry, B. Mandard-Cazin, M. Rougee, R. V.
1424-1428.
13
P. M. Jayaweera, S. S. Palayangoda, K. Tennakone,
J. Photochem. Photobiol. A 2001, 140, 173-177.
14 S. M. Reda, S. A. El-Sherbieny, J. Mater. Res. 2010,
25, 522-528.
32
M. Yoshizawa, K. Suzuki, A. Kubo, S. Saikan,
Chem. Phys. Lett. 1998, 290, 43-48.
15
M. Y. El Zayat, A. O. Saed, M. S. El-Dessouki, Int.
33
G. Li, D. Magana, R. B. Dyer, J. Phys. Chem. B
J. Hydrogen Energy 1998, 23, 259-266.
2012, 116, 12590-12596.
16
M. K. Nazeeruddin, R. Humphry-Baker, P. Liska,
34
P. Singhal, H. N. Ghosh, Phys. Chem. Chem. Phys.
M. Grätzel, J. Phys. Chem. B 2003, 107, 8981-8987.
2014, 16, 16824-16831.
17
M. K. Nazeeruddin, S. M. Zakeeruddin, R.
35
M. Ucak-Astarlioglu, S. Edwards, C. A. Zoto,
Humphry-Baker, M. Jirousek, P. Liska, N. Vlachopoulos, V.
Shklover, C.-H. Fischer, M. Grätzel, Inorg. Chem. 1999, 38,
6298-6305.
Journal of Laboratory Chemical Education 2017, 5, 32–39.
36
B. M. Fox, J. D. Hepworth, D. Mason, J. Sawyer, G.
Hallas, J. Soc. Dyers Colour. 1982, 98, 10-13.
18
Semiconductor Photochemistry and Photophysics
37
38
E. B. Knott, J. Chem. Soc. 1951, 1024-1028.
Vol. 10. V. Ramamurthy, K. S. Schanze, Marcel Dekker, Inc.,
New York, Basel, 2003.
M. J. Frisch, G. W. Trucks, H. B. Schlegel, G. E.
Scuseria, M. A. Robb, J. R. Cheeseman, G. Scalmani, V.
Barone, B. Mennucci, G. A. Petersson, H. Nakatsuji, M.
Caricato, X. Li, H. P. Hratchian, A. F. Izmaylov, J. Bloino, G.
19
G. R. Fulmer, A. J. M. Miller, N. H. Sherden, H. E.
Gottlieb, A. Nudelman, B. M. Stoltz, J. E. Bercaw, K. I.
12

Zheng, J. L. Sonnenberg, M. Hada, M. Ehara, K. Toyota, R.
Fukuda, J. Hasegawa, M. Ishida, T. Nakajima, Y. Honda, O.
Kitao, H. Nakai, T. Vreven, J. A. Montgomery, J. E. Peralta, F.
Ogliaro, M. Bearpark, J. J. Heyd, E. Brothers, K. N. Kudin, V.
N. Staroverov, R. Kobayashi, J. Normand, K. Raghavachari, A.
Rendell, J. C. Burant, S. S. Iyengar, J. Tomasi, M. Cossi, N.
Rega, J. M. Millam, M. Klene, J. E. Knox, J. B. Cross, V.
Bakken, C. Adamo, J. Jaramillo, R. Gomperts, R. E. Stratmann,
O. Yazyev, A. J. Austin, R. Cammi, C. Pomelli, J. W.
Ochterski, R. L. Martin, K. Morokuma, V. G. Zakrzewski, G.
A. Voth, P. Salvador, J. J. Dannenberg, S. Dapprich, A. D.
Daniels, Farkas, J. B. Foresman, J. V. Ortiz, J. Cioslowski, D. J.
Fox, Gaussian 09, Gaussian, Inc., Wallingford, CT, 2010.
53
W. Schmickler, R. Guidelli, J. Electroanal. Chem.
Interfacial Electrochem. 1987, 235, 387-392.
54
M. Boström, B. W. Ninham, Langmuir 2004, 20,
7569-7574.
55
P. Attard, D. Antelmi, I. Larson, Langmuir 2000, 16,
1542-1552.
56
D. A. H. Hanaor, W. Chrzanowski, Y. Gan, J. Aust.
Ceram. Soc. 2015, 51, 88–93.
57
J.-B. Harlé, S. Mine, T. Kamegawa, V. T. Nguyen,
T. Maeda, H. Nakazumi, H. Fujiwara, J. Phys. Chem. C 2017,
121, 15049-15062.
58
B. J. Hathaway, D. E. Webster, Proc. Chem. Soc.
1963, 14.
39
M. K. Nazeeruddin, P. Péchy, T. Renouard, S. M.
59
M. Kosmulski, J. B. Rosenholm, J. Phys. Chem.
Zakeeruddin, R. Humphry-Baker, P. Comte, P. Liska, L. Cevey,
E. Costa, V. Shklover, L. Spiccia, G. B. Deacon, C. A.
Bignozzi, M. Grätzel, J. Am. Chem. Soc. 2001, 123, 1613-1624.
1996, 100, 11681-11687.
60
E. Ronca, M. Pastore, L. Belpassi, F. Tarantelli, F.
De Angelis, Energy Environ. Sci. 2013, 6, 183-193.
40
B. O'Regan, M. Gratzel, Nature 1991, 353,
61
L. Zhang, J. M. Cole, C. Dai, ACS Appl. Mater.
737-740.
41
Interfaces 2014, 6, 7535-7546.
J. Lim, Y. S. Kwon, S.-H. Park, I. Y. Song, J. Choi,
62
M. Pastore, F. De Angelis, ACS Nano 2010, 4,
T. Park, Langmuir 2011, 27, 14647-14653.
556-562.
63
42
A. Hagfeldt, G. Boschloo, L. Sun, L. Kloo, H.
G. B. Deacon, R. J. Phillips, Coord. Chem. Rev.
Pettersson, Chem. Rev. 2010, 110, 6595-6663.
1980, 33, 227-250.
43
A. Mishra, M. K. R. Fischer, P. Bäuerle, Angew.
64
K. Kakiage, T. Tokutome, S. Iwamoto, T. Kyomen,
Chem. Int. Ed. Engl. 2009, 48, 2474-2499.
M. Hanaya, Chem. Commun. 2013, 49, 179-180.
44
45
N. Heo, Y. Jun, J. H. Park, Sci. Rep. 2013, 3, 1712.
P. M. Sommeling, M. Späth, H. J. P. Smit, N. J.
65
A. Yella, H.-W. Lee, H. N. Tsao, C. Yi, A. K.
Chandiran, M. K. Nazeeruddin, E. W.-G. Diau, C.-Y. Yeh, S.
Bakker, J. M. Kroon, J. Photochem. Photobiol. A 2004, 164,
M. Zakeeruddin, M. Grätzel, Science 2011, 334, 629-634.
137-144.
66
W. Wu, J. Hua, Y. Jin, W. Zhan, H. Tian,
46
J. Lim, Y. S. Kwon, T. Park, Chem. Commun. 2011,
Photochem. Photobio. Sci. 2008, 7, 63-68.
47, 4147-4149.
67
X. Ma, J. Hua, W. Wu, Y. Jin, F. Meng, W. Zhan,
47
48
49
50
51
P. Debye, E. Hückel, Phys. Z. 1923, 24, 185–206.
H. Tian, Tetrahedron 2008, 64, 345-350.
M. Gouy, J. Phys. Theor. Appl. 1910, 9, 457-468.
D. L. Chapman, Philos. Mag. 1913, 25, 475-481.
D. C. Grahame, Chem. Rev. 1947, 41, 441-501.
L. S. Ornstein, F. Zernike, Koninklijke
68
F. S. Meng, Q. H. Yao, J. G. Shen, F. L. Li, C. H.
Huang, K. C. Chen, H. Tian, Synth. Met. 2003, 137, 1543-1544.
69
Wang, B.-W. Zhang, J. Mater. Chem. 2005, 15, 1654-1661.
70 Y. Nagasawa, Y. Ando, T. Okada, Chem. Phys. Lett.
1999, 312, 161-168.
Y.-S. Chen, C. Li, Z.-H. Zeng, W.-B. Wang, X.-S.
Nederlandsche Akademie van Wetenschappen Proceedings
1914, 17, 793–806.
52
G. Trefalt, S. H. Behrens, M. Borkovec, Langmuir
71
M. A. Bell, B. Crystall, G. Rumbles, G. Porter, D.
2016, 32, 380-400.
R. Klug, Chem. Phys. Lett. 1994, 221, 15-22.
13

2014, 1303-1311.
75 H.
72
B.-B. Xie, S.-H. Xia, L.-H. Liu, G. Cui, J. Phys.
Chem. A 2015, 119, 5607-5617.
Wang, B. Wang, J. Yu, Y. Hu, C. Xia, J. Zhang,
R. Liu, Sci. Rep.-UK 2015, 5, 9305 (9 pages).
73
K. Hara, T. Sato, R. Katoh, A. Furube, Y. Ohga, A.
Shinpo, S. Suga, K. Sayama, H. Sugihara, H. Arakawa, J. Phys.
Chem. B 2003, 107, 597-606.
76
R. Kern, R. Sastrawan, J. Ferber, R. Stangl, J.
Luther, Electrochim. Acta 2002, 47, 4139-4277.
74
T. Funaki, H. Kusama, N. Onozawa-Komatsuzaki,
77
A. Zaban, M. Greenshtein, J. Bisquert,
K. Kasuga, K. Sayama, H. Sugihara, Eur. J. Inorg. Chem. 2014,
ChemPhysChem. 2003, 4, 859-864.
14

Graphical Abstract
Malachite Green Derivatives for Dye-sensitized Solar Cells: Optoelectronic Characterizations and Persistence on TiO2
Jean-Baptiste Harlé,* Shuhei Arata, Shinya Mine, Takashi Kamegawa, Van Tay Nguyen, Takeshi Maeda, Hiroyuki Nakazumi, and
Hideki Fujiwara*
Derivatives of malachite green were developed for dye-sensitized solar cell, and conjugated to various coordinative (trifluoroacetate
–
[TFA–], triflate [TfO–]) and non-coordinative (tetrafluoroborate [BF4 ], bromide [Br–]) counterions. Throughout this study, the
bearing of the affinity for the surface/affinity for the bulk of both the chromophores and its counterions was scrutinized and found
crucial for the stability and efficiency of the cells.
<Diagram>
15