Effect of the CF3 Substituents on the Charge-Transfer Kinetics of High-Efficiency Cyclometalated Ruthenium Sensitizers

Article abstract of DOI:10.1021/acs.inorgchem.6b02088

Six thiocyanate-free complexes, DUY1-DUY6, were synthesized, and their application in a dye-sensitized solar cell was studied to explore the effect of the CF₃ substituent positioned in the ancillary ligand and the structure of the anchoring ligand on the physicochemical properties, charge-transfer kinetics, and photovoltaic properties of ruthenium sensitizers. When the electron-withdrawing groups were installed on the cyclometalating ligands and their π conjugation of the ancillary ligand was extended, the frontier orbital energy levels of the ruthenium complex appeared to be sufficient for effective electron injection and dye regeneration, at the same time having high light-harvesting ability. Two electron-withdrawing CF₃ groups meta to the cyclometalated position reduce the electron density at the metal center less seriously than o-CF₃ and p-CF₃ groups. The sensitizers containing a m-CF₃ group also reveal a more favorable distribution of β lowest unoccupied spin orbital for interaction between the oxidized dyes and the iodide ion, which promotes dye regeneration. The absorption profiles of DUY1-DUY4 adsorbed a TiO₂ film extended to longer wavelength compared to those in an N,N-dimethylformamide solution, especially DUY1 and DUY2 dyes, which have λ_max red shifts of up to 30 nm. The DUY2-dyed cell exhibited the highest efficiency of 9.03%, while the power conversion efficiencies of DUY1-, DUY3-, DUY4-, and N719-based devices were 7.40%, 7.01%, 8.92%, and 8.63%, respectively. DUY5 and DUY6 (the side products of DUY3 and DUY4) without anchoring groups have very weak physical adsorption on a TiO₂ anode. The corresponding cells exhibit very low efficiency (<0.1%), although both dyes have high light-harvesting ability and proper frontier orbital energy levels.

Full text of DOI:10.1021/acs.inorgchem.6b02088

Article
pubs.acs.org/IC
Effect of the CF3 Substituents on the Charge-Transfer Kinetics of
High-Efficiency Cyclometalated Ruthenium Sensitizers
The-Duy Nguyen, Chun-Han Lin, and Chun-Guey Wu*
Department of Chemistry and Research Center for New Generation Photovoltaics, National Central University, Jhong-Li 32001,
Taiwan, Republic of China
S
* Supporting Information
ABSTRACT: Six thiocyanate-free complexes, DUY1−DUY6, were synthesized, and
their application in a dye-sensitized solar cell was studied to explore the effect of the
CF3 substituent positioned in the ancillary ligand and the structure of the anchoring
ligand on the physicochemical properties, charge-transfer kinetics, and photovoltaic
properties of ruthenium sensitizers. When the electron-withdrawing groups were
installed on the cyclometalating ligands and their π conjugation of the ancillary
ligand was extended, the frontier orbital energy levels of the ruthenium complex
appeared to be sufficient for effective electron injection and dye regeneration, at the
same time having high light-harvesting ability. Two electron-withdrawing CF3
groups meta to the cyclometalated position reduce the electron density at the
metal center less seriously than o-CF3 and p-CF3 groups. The sensitizers containing
a m-CF3 group also reveal a more favorable distribution of β lowest unoccupied spin
orbital for interaction between the oxidized dyes and the iodide ion, which promotes
dye regeneration. The absorption profiles of DUY1−DUY4 adsorbed a TiO2 film
extended to longer wavelength compared to those in an N,N-dimethylformamide solution, especially DUY1 and DUY2 dyes,
which have λmax red shifts of up to 30 nm. The DUY2-dyed cell exhibited the highest efficiency of 9.03%, while the power
conversion efficiencies of DUY1-, DUY3-, DUY4-, and N719-based devices were 7.40%, 7.01%, 8.92%, and 8.63%, respectively.
DUY5 and DUY6 (the side products of DUY3 and DUY4) without anchoring groups have very weak physical adsorption on a
TiO2 anode. The corresponding cells exhibit very low efficiency (<0.1%), although both dyes have high light-harvesting ability
and proper frontier orbital energy levels.
HOMO−LUMO energy gap (Egap).3 Another method to red-
INTRODUCTION
■
shift the absorption and enhance the molar absorption
A dye-sensitized solar cell (DSC), first reported by O’Regan
coefficient of the sensitizer is to extend the π-conjugation
and Gratzel in 1991, is one of the most promising photovoltaic
̈
length of the ancillary and/or anchoring ligands.1,4−6 Never-
theless, this method sometimes exhibits a trade-off of JSC and
VOC due to the fast charge recombination at the TiO2/
electrolyte interface caused by the poor TiO2 surface protection
of the bulky sensitizers with big ancillary ligands.7
technologies by virtue of its semitransparency, flexibility,
impressive photon-to-current conversion efficiency, noteworthy
stability, and low manufacturing cost. The device color can also
be varied by selecting dyes that absorb at various wavelengths,
which enhances the application versatility, such as those used in
indoor lighting.1 To obtain a high conversion efficiency,
optimization of the open-circuit potential (VOC) and short-
circuit current density (JSC) of the cell is essential. The value of
VOC depends on the edge of the TiO2 conduction band known
to have a Nernstian dependence on the pH value2 and the
redox potential of the mediator. Meanwhile, JSC is related to the
light-harvesting properties of the sensitizer, the interaction
between the sensitizer and TiO2 anode, and the charge-transfer
processes in the device. Both optical and electronic properties
of the complex sensitizers can be tuned in two ways: by the
introduction of a ligand with a low-lying π* molecular orbital or
by destabilization of the metal t2g orbital through the
introduction of a strong donor ligand. The former lowers the
energy of the lowest unoccupied molecular orbital (LUMO),
while the latter destabilizes the highest occupied molecular
orbital (HOMO) of the sensitizer, ultimately decreasing the
One of the most successful sensitizers is N719 (JSC = 17.73
mA/cm2, VOC = 846 mV, FF = 0.75, and η = 11.18%).8
However, the main drawbacks of N719 are the lack of
absorption in the red region of the visible spectrum3 and the
presence of monodentate thiocyanate (NCS) ligands. Although
the NCS ligand can encourage dye regeneration by interacting
with I− and red-shifting the absorption profile by destabilizing
the HOMO, it is a monodentate ligand, which makes
replacement by other ligands (such as N,N-dimethylformamide,
acetonitrile, 3-methoxypropionitrile, and 4-tert-butylpyridine)
possible, resulting in low cell stability under excessive thermal
stress and/or light soaking.9−12 Furthermore, because of its
structure and ambidentate nature, the NCS ligand is difficult to
Received: August 29, 2016
© XXXX American Chemical Society
A
DOI: 10.1021/acs.inorgchem.6b02088
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
Figure 1. Structures of DUY series sensitizers.
Zennium CIMPS-1 potentiostat installed with a light-intensity-
modulated function to control the light intensity. Transient absorption
spectroscopy (TAS) measurements were performed on a Proteus
Ultrafast System with a custom-designed light path. The measure-
ments (EIS, IMPS, IMVS, and TAS) and relative calculations can be
found in our previous papers.6,19
modify and dye purification steps are virtually required to
separate the active N-coordinated isomer from the inactive S-
coordinated isomer.12 One of the common approaches for
replacing NCS ligands is using cyclometalating bidentate
ligands5,12−18 that can manipulate the electronic and optical
properties of the sensitizer by modifying the substituents on the
ligand. However, a ruthenium complex containing a cyclo-
metalating ligand generally has a high HOMO level, which may
decrease the driving force for dye regeneration.5,16−18 Adding
an electron-withdrawing group (such as CF3) on the cyclo-
metalating ligand is a strategy to lower the HOMO level to
facilitate dye regeneration. In this paper, we reported the
physicochemical and photovoltaic properties of four high-
efficiency NCS-free, CF3-substituted cyclometalated ruthenium
complexes (DUY1−DUY4) and two model complexes (DUY5
and DUY6). Their structures are displayed in Figure 1. This
study mainly focuses on the effects of the CF3-substituent
positions on the charge-transfer kinetics and the photovoltaic
performance of the ruthenium dyes. The ultimate purpose is to
understand the structure−property relationship of a ruthenium
complex containing a CF3-substituted cyclometalated ancillary
ligand in order to develop NCS-free high-efficiency cyclo-
metalated complex sensitizers for DSC.
Density Functional Theory (DFT) Calculation. DFT calcu-
lations were performed using the Gaussian 09 program.20 For ground-
state geometry optimization, the B3LYP (Becke, three-parameter,
Lee−Yang−Parr) hybrid functional21,22 was used. The LanL2DZ
effective core potential23 was used for the ruthenium atom, and the
split-valence 6-31G(d,p) basis set24 was applied for the hydrogen,
sulfur, carbon, oxygen, nitrogen, and fluorine atoms. The time-
dependent DFT (TD-DFT) excited-state calculation of the lowest
singlet−singlet electronic transitions was also performed using the
same basic sets. The solvent effects were included in the calculations
using the conductor-like polarizable continuum model.25−27
Device Fabrication and Photovoltaic Performance Charac-
terization. The preparation of TiO2 pastes, fabrication of a TiO2
anode, dye loading, assembly, and photovoltaic parameter measure-
ments of the cells were similar to what we reported previously6 unless
specified.
RESULTS AND DISCUSSION
■
Strategy for the Design of the Sensitizers. DUY1 and
DUY2 were derived from the replacement of two NCS ligands
in N719 with cyclometalating bidentate ligands named 2-[3,5-
bis(trifluoromethyl)phenyl]-4-[5-(hexylthio)-3,4-
ethylenedioxythiophenyl]pyridine (LD1) and 2-[2,4-bis-
(tri fluo ro met hyl)ph enyl ]-4 -[5-(hexylthio)-3, 4-
ethylenedioxythiophenyl]pyridine (LD2), respectively. The
electron-donating character of the negatively charged 2-
phenylpyridine (ppy−) unit could destabilize the HOMO
levels, reduce the HOMO−LUMO energy gap, and bath-
ochromically shift the absorption profile, especially the metal-
to-ligand charge-transfer transitions. To maintain a sufficient
driving force for dye regeneration (lowering of the HOMO
level), two electron-withdrawing CF3 substituents were
introduced to the benzene ring of the cyclometalated ligand.
Depending on the spatial closeness of the CF3 substituents to
the cyclometalated position, the o- and p-CF3 substituents of
DUY1 were expected to show a stronger influence on the
HOMO level of the dye compared to the two m-CF3
substituents for DUY2. DUY3 and DUY4 were designed by
extending the π-conjugation systems with the 5-(hexylthio)-3,4-
ethylenedioxythiophen-2-ide (EDOTSHexyl) moiety attached
on one of the anchoring ligands of DUY1 and DUY2,
respectively. The functions of EDOTSHexyl not only improve
the absorption of the dye and suppress dye aggregation but also
keep the hydrophilic water from approaching the TiO2 surface.
EXPERIMENTAL SECTION
■
Materials and Measurements. All reagents were obtained from
commercial sources and used as received unless specified. Solvents
were dried over 4 Å molecular sieves or CaH2 before use. The detail
preparation and structure characterization steps of the six ruthenium
complexes are collected in the Supporting Information (SI). NMR
spectra were recorded on a Bruker 300 MHz NMR spectrometer.
High-resolution mass spectrometry (HRMS) spectra were obtained
using a JMS-700 HRMS spectrometer. Elemental analyses were carried
out with a PerkinElmer 2400 CHNS/O analyzer. UV−vis spectra of
the dyes in N,N-dimethylformamide (DMF) and adsorbed on a TiO2
film were measured using a Cary 300 Bio spectrometer. Voltammetric
measurements were performed in a single-compartment, three-
electrode cell with a platinum wire counter electrode and a platinum
disk working electrode. The reference electrode was Ag+/Ag, and the
supporting electrolyte was 0.1 M tetrabutylammonium perchlorate in
DMF. The square-wave voltammograms of the sensitizers dissolved in
DMF (potential step increment = 5 mV; frequency = 25 Hz) were
recorded using a potentiostat/galvanostat (PGSTAT30, Eco-Chemie
Autolab, Utrecht, The Netherlands), and ferrocene was used as a
calibration internal standard. Electrochemical impedance spectroscopy
(EIS) measurements were carried out using an Eco Chemie Autolab
PGSTAT30 potentiostat/galvanostat in 50 mV voltage steps with a
sinusoidal potential perturbation of 10 mV in the dark. Intensity-
modulated photocurrent spectroscopy (IMPS) and intensity-modu-
lated photovoltage spectroscopy (IMVS) were recorded with a Zahner
B
DOI: 10.1021/acs.inorgchem.6b02088
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
DUY5 and DUY6, the side products of DUY3 and DUY4,
respectively, were used as model compounds to further confirm
the effects of the CF3-substituent positions on the photo-
physical and electrochemical properties of a ruthenium
sensitizer. They were also used to reveal the adsorption
behavior (on TiO2) and photovoltaic performance of sensitizers
having all ligands without a carboxylate group but with a CF3-
substituted cyclometalated ligand.
consistent with the studies by Berlinguette and co-workers16 as
well as Frey and co-workers,17 who proved that the
configuration of tris-heteroleptic cycloruthenated dyes had the
cyclometalating ring positioned trans to 2,2′-bipyridine-4,4′-
dicarboxylate based on the X-ray diffraction data.
Optical Properties. The absorption spectra of DUY
sensitizers and N719 dissolved in DMF solutions and adsorbed
on thin TiO2 films are depicted in Figure 2. All sensitizers (in
Synthesis and Characterization. For the preparation of
ligands LD1 and LD2, we chose 2-bromo-4-aminopyridine as
the starting material. Compared to the use of 2-bromo-4-
iodopyridine for the synthesis of 4-substituted ppy−,28 2-
bromo-4-aminopyridine exhibited some advantages, such as low
cost, stability at ambient temperature, insensitivity to moisture
or light, ease of handling, a few side products after coupling
reactions (because there are no competitive reactions), and
short purification time. Six complexes, DUY1−DUY6, were
prepared by modifying procedures developed by Berlinguette
and co-workers.13,14 The addition of C^N (C^N = bidentate
cyclometalating ligand, e.g., ligands LD1 and LD2) to [(η6-p-
cymene)RuCl2]2 at 45 °C led to both [Ru(CH3CN)4(C^N)]-
PF6 and [Ru(p-cymene)(CH3CN)2(C^N)]PF6 in approximate
equimolar amounts. Only [Ru(CH3CN)4(C^N)]PF6 is a useful
precursor for the later syntheses because [Ru(p-cymene)-
(CH3CN)2(C^N)]PF6 does not react with 2,2′-bipyridyl-4,4′-
dicarboxylic acid to an appreciable extent.15 We note that
[Ru(CH3CN)4(C^N)]PF6 is the main product at 55 °C
because [Ru(p-cymene)(CH3CN)2(C^N)]PF6 can convert to
[Ru(CH3CN)4(C^N)]PF6 at either higher temperature or
longer reaction time.14
The solubility of DUY1−DUY6 in organic solvents (e.g.,
CH2Cl2, CHCl3, CH3OH, CH3CN, and DMF) follows the
order of DUY1 < DUY2 ≪ DUY3 ≈ DUY4 < DUY5 ≈ DUY6,
indicating that their solubility is determined by the number of
alkyl chains. Therefore, DUY3−DUY6 could be simply purified
by conventional silica gel chromatography, while DUY1 and
DUY2 needed to be isolated by reversed-phase chromatog-
raphy after dissolution in a 0.2 M tetrabutylammonium/
methanol solution. NMR data indicated that all complexes exist
as a single isomer with structures displayed in Figure 1.29 DUY3
and DUY4 could have the Ru−C bond trans to either 2,2′-
bipyridine-4,4′-dicarboxylate or 4,4′-bis[5-(hexylthio)-3,4-ethyl-
enedioxythiophenyl]-2,2′-bipyridine (ligand B20). The result-
ing configurations of DUY3 and DUY4 could be expected by
the trans effect and the reactivity of the ligands in tha
coordination reaction. Et2dcbpy has a smaller molecular size
and could make a stronger synergistic bonding with the
ruthenium metal center than ligand B20,15 while the negatively
charged carbon atom of the cyclometalating ligand induced a
higher trans effect of the benzene ring compared to the pyridine
ring.30 As a consequence, the first step of the coordination
reaction is the replacement of two CH3CN molecules in
[Ru(CH3CN)4(C^N)]PF6 (one trans to the benzene ring and
one cis to both the benzene ring and the pyridine ring of the
cyclometalating ligand) with Et2dcbpy, and in the next step, the
remaining acetonitrile ligands were replaced with ligand B20 to
produce the ester form of DUY3 or DUY4. After hydrolysis,
DUY3 and DUY4 were obtained with the conformation shown
in Figure 1. The chemical shifts of the ortho protons in the
benzene ring of DUY4 and its ester form are almost the same
(Δδ = 0.013 ppm), also indicating that Et2dcbpy oriented trans
to the benzene ring and evidently the ligand B20 trans to the
pyridine ring of the cyclometalating ligand.31 These results are
Figure 2. (A) Absorption spectra of the DUY sensitizers and N719
dissolved in DMF and (B) absorption spectra of DUY1−DUY4 and
N719 anchored on transparent TiO2 films.
DMF) show a strong absorption band around 350 nm with a
molar absorption coefficient (ε) in the range of (3.0−7.0) ×
104 M−1 cm−1 assigned to the π−π* transitions of the
cyclometalated ligands. A series of broad absorption bands at
longer wavelength are primarily from the mixed-metal/ligand-
to-ligand charge-transfer (MMLLCT) transitions.14,17 The
extinction coefficients of the DUY sensitizers are all remarkably
higher than that of N719 in the whole absorption range. The
absorption maxima (λmax) of the lowest-energy band of the
sensitizers are in the order of DUY1 < DUY2 < N719 < DUY3
< DUY5 < DUY4 < DUY6. The electron-withdrawing nature of
the CF3 groups on the cyclometalating ligands makes λmax of
DUY1 and DUY2 shorter than that of N719. On the contrary,
DUY3−DUY6 sensitizers show a red shift in λmax compared to
N719 due to the presence of the electron-donating ancillary
ligand B20 on the DUY dyes.
Electrochemical Properties. The HOMO energy levels of
the DUY sensitizers were determined with square-wave
voltammetry, as shown in Figure 3A. The only oxidation
peak of the DUY1 and DUY2 dyes is assigned to the oxidation
of RuII to RuIII.17 DUY3−DUY6 exhibit an extra three-electron
oxidation peak at ∼1.47 V vs NHE, which is assigned to the
C
DOI: 10.1021/acs.inorgchem.6b02088
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
sensitizers are in the order of DUY1 > DUY2 > DUY3 > DUY5
> DUY4 > DUY6.
The electronic absorption spectra of DUY1−DUY4 and
N719 immobilized on the transparent TiO2 films displayed in
Figure 2B show a red shift of λmax (especially those for DUY1
and DUY2, which red-shift about 30 nm) along with a change
in the absorption profile compared to those in DMF. This is
probably because of the increasing electronic coupling between
the π and π* orbitals of the ancillary ligands.33 The spectra of
the DUY5 and DUY6 sensitizers have not been shown here
because they almost cannot anchor on the TiO2 surface. The
absorption profiles of DUY dyes adsorbed on TiO2 films are
also wider compared to those for the dyes in DMF. The
widening of the absorption profile is reasonably attributed to
the dye molecules with various conformations coexisting on the
TiO2 surface, due to the bulky nature of NCS-free ruthenium
complexes.6 This probably is one of the benefits for designing a
sensitizer with bulky ancillary ligand/structure. Furthermore,
compared to the absorption of N719 on a TiO2 thin film, the
absorption profiles shown in Figure 2B clearly reveal that
DUY2 and DUY4 are able to absorb more light than N719
when applied in DSC devices.
Theoretical Calculation of the Frontier Molecular
Orbital Distributions and Electronic Transitions. The
graphical representation of the computational frontier orbital
distribution is illustrated in Figures S1−S6 (see the SI). The
HOMOs of all DUY1−DUY6 are contributed primarily from
both the metal and ligands (cyclometalating and/or polypyridyl
ligands), while the LUMOs virtually distribute on the
polypyridyl (anchoring) ligands, suggesting that the broad
absorption bands of the sensitizers result predominantly from
the MMLLCT transitions.14 The partial contribution of ppy−
and/or 4-[5-(hexylthio)-3,4-ethylenedioxythiophenyl]pyridin-
2-yl units to the HOMOs indicates that the cyclometalating
ligands and ligand B20 could influence localization of the
HOMO levels. On the other hand, because the LUMOs
virtually distribute on the polypyridyl ligands, cyclometalation
of the metal center does not destabilize the LUMOs. Empty
molecular orbitals (see Figures S1 and S2 in the SI) of DUY1
and DUY2 are almost all localized on the anchoring ligands,
while for DUY3 and DUY4, the empty molecular orbitals (see
Figures S3 and S4 in the SI) are virtually contributed from the
4-[5-(hexylthio)-3,4-ethylenedioxythiophenyl]pyridin-2-yl
units. For this reason, both DUY1 and DUY2 may exhibit
higher electron injection efficiency than DUY3 and DUY4
(from the photoexcited dyes to the conduction band of TiO2).
To gain more insight into the absorption bands, TD-DFT
calculation of the singlet electronic transition was performed
using DMF as the solvent. The calculated oscillator strengths
and electron density difference maps as well as the experimental
absorption spectra for DUY dyes are displayed in Figures S7−
S12 in the SI, and the detailed transition assignments are listed
in Tables S1−S6 in the SI. The calculated oscillator strengths
are quite close to the experimental absorption spectra, although
the calculated λmax values are blue-shifted compared to the
experimental data for all dyes. The results imply that variation
in the electronic properties of these types of sensitizers could be
predicted by a routine theoretical calculation. DUY1 and DUY2
have six (S5, S6, S7, S11, S13, and S16) and five (S7, S8, S10, S11,
and S16), respectively, strong transitions with oscillator
strengths (f) > 0.1 (see Tables S1 and S2 in the SI). These
transitions are believed to contribute significantly to the
photocurrent because photoexcited electrons could be
Figure 3. (A) Square-wave voltammograms and (B) frontier orbital
energy-level diagram of DUY sensitizers.
oxidation of 4-[5-(hexylthio)-3,4-ethylenedioxythiophenyl]-
pyridin-2-yl units. The positions of the electron-withdrawing
CF3 groups on the cyclometalating ligand can give rise to
distinct impacts on the oxidation potentials of the sensitizers.
Two CF3 groups at ortho and para positions reduce the
electron density at the metal center more efficiently than two
m-CF3 groups. As a consequence, DUY1, DUY3, and DUY5
have HOMO levels 50−60 mV more positive than those of
DUY2, DUY4, and DUY6, respectively (see Figure 3A). Ligand
B20 is a strong donor that also plays a role in the tuning of the
HOMO levels of the sensitizers. Replacement of a bipyr-
idinedicarboxylic acid ligand with a ligand B20 leads to a 200−
210 mV rise in the HOMO level (such as DUY3 and DUY4),
while a 250 mV rise was obtained by substituting two
bipyridinedicarboxylic acids with two B20 ligands (such as
DUY5 and DUY6). The ground-state energy levels of the DUY
sensitizers are in a range of 0.83−1.14 V, which are all more
positive than the threshold (∼0.8 V vs NHE) for dye
regeneration by I− in the redox couple.18,32 The LUMO energy
levels (ELUMO) of the dyes were calculated by the formula
ELUMO = EHOMO − Egap, where Egap values are obtained from the
onset of their absorption spectra illustrated in Figure 2A. The
HOMO and LUMO energy levels of all DUY dyes are
summarized in Figure 3B. The LUMO levels of all sensitizers
are found to be at least 490 mV higher than the Fermi level of
the TiO2 film,15 providing a sufficient driving force for electron
injection. The energy levels of the frontier orbitals suggest that
DUY1−DUY6 are all appropriate dyes for application in
DSCs.2 Furthermore, the σ-donating ligands (LD1, LD2, and
B20) affect HOMOs more than LUMOs by transferring the
electron density to the metal center.5 In this sense, Egap values
also depend on the positions of the electron-withdrawing CF3
groups on the cyclometalating ligands and the number of ligand
B20 in the complexes. As a result, the energy gaps of the DUY
D
DOI: 10.1021/acs.inorgchem.6b02088
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
effectively injected into the conduction band of TiO2 from
these transitions through the anchoring ligands on which the
empty molecular orbitals located. On the contrary, the
transition assignments of DUY3 and DUY4 (see Tables S3
and S4 in the SI) indicate that the electron injection
corresponding to the strong transitions is less efficient because
some of the empty molecular orbitals locate on 4-[5-
(hexylthio)-3,4-ethylenedioxythiophenyl]pyridin-2-yl units in-
stead of on the anchoring ligand. On the basis of the estimated
electron injection and optical data without considering the dye
loading, we may expect that JSC of the corresponding cell is in
the order of DUY1 > DUY2 > DUY4 > DUY3.
the illumination of AM 1.5 G simulated sunlight, the power
conversion efficiencies of DUY5- and DUY6-based cells are
very low because very few dye molecules could adsorb on TiO2
(DUY5- and DUY6-adsorbed TiO2 anodes look white). These
results indicate that the weakly physical interaction between the
dye (without a carboxylate group) and TiO2 is not sufficient for
adsorbing the sensitizer on the TiO2 surface. It seems that even
the few DUY5 and DUY6 impurities existing in DUY3 and
DUY4 may not significantly affect the photovoltaic perform-
ance of DUY3 and DUY4 when applied in DSC. DUY1−
DUY4-based cells exhibit efficiencies of 7.40%, 9.03%, 7.01%,
and 8.92%, respectively, while that of the N719-based cell is
8.63%. Like the absorption spectra displayed in Figure 2B, the
high efficiency of DUY2 and DUY4 could be derived from the
strongest absorption of DUY2 and the widest absorption range
of DUY4 when they adsorbed on a TiO2 film. Interestingly,
DUY2- and DUY4-based cells display higher IPCE values at
590−800 nm than DUY1- and DUY3-sensitized cells because
the DUY2- and DUY4-dyed TiO2 films have a stronger
absorption at long wavelength (see Figure 2B). Data listed in
Table 1 show that VOC and FF of all cells are very close to each
other. As a consequence, the efficiency of the cell is mainly
determined by JSC. The order of the dye loadings (DUY2 >
DUY4 > DUY3 > DUY1; see Table 1) implies that the
positions of the CF3 groups on the cyclometalating ligand affect
the dye loading like the bulk ancillary ligand B20 does. DUY2
has the highest dye loading among all of the sensitizers because
of its small molecular size; therefore, the corresponding cell
achieves the highest JSC. The lowest dye loading of DUY1 could
originate from its poor solubility in organic solvents used for
dissolving dye. Nevertheless, JSC of a DUY1-sensitized device is
higher than that of a cell based on DUY3 because DUY1 has a
stronger absorption at 450−525 nm [in this range, the standard
sunlight (AM 1.5 G) has the largest photo number] than DUY3
(see Figure 2). Furthermore, the driving force for dye
regeneration of DUY1 is higher than that of DUY3 because
the former has a lower HOMO level (see Figure 3B).
Photovoltaic Performance. The J−V and IPCE curves for
DUY1−DUY4-sensitized devices are displayed in Figure 4. The
Charge-Transfer Kinetics inside the Cells. To obtain the
dynamics of the interfacial charge-transfer processes within the
devices, EIS, IMPS/IMVS, and TAS spectra were employed to
estimate the resistance at the TiO2/dye/electrolyte interface,
the transport time (τtr) and lifetime (τn) of the electron on
TiO2, as well as the dye regeneration kinetics, respectively. EIS
spectra (Nyquist plots) of the devices taken under illumination
and in the dark are illustrated in Figure 5. The second
semicircle of the EIS Nyquist plot taken in the dark typically
presents to the interfacial resistance (R2) of TiO2/dye/
electrolyte interfaces.34 At the illumination and open-circuit
conditions, the charge-transfer processes at the interface are
composed of electron injection from the light-excited dye to the
Figure 4. (A) J−V and (B) IPCE curves of DUY1−DUY4- and N719-
sensitized devices.
photovoltaic parameters and dye loading (also including those
of N719, DUY5, and DUY6) are summarized in Table 1. Under
Table 1. Photovoltaic Parameters, Dye Loading, TiO2/Dye/Electrolyte Interface Resistance (R2), and Regeneration Times of
the Sensitizers
dye
JSC (mA/cm2)
VOC (V)
FF
η (%)
dye loading (×10−7mol/cm2)
R2 (Ω, illum)
R2 (Ω, dark)
τ1 (μs)
τ2 (μs)
DUY1
DUY2
DUY3
DUY4
DUY5
DUY6
N719
15.01
18.40
14.29
17.74
0.13
0.733
0.737
0.737
0.741
0.391
0.381
0.744
0.67
0.67
0.67
0.68
0.42
0.42
0.66
7.40
9.03
7.01
8.92
0.02
0.02
8.63
4.96
5.85
5.08
5.29
19.9
17.2
22.6
18.9
100
104
104
110
7.47
6.63
11.8
7.59
402.1
378.4
385.2
383.1
0.13
17.54
E
DOI: 10.1021/acs.inorgchem.6b02088
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
Figure 6. (A) Electron-transport time and (B) electron lifetime as a
function of the light intensity for DSCs based on DUY1−DUY4.
Figure 5. EIS spectra (Nyquist plots) of DUY-sensitized cells: (A)
under AM 1.5 G illumination; (B) in the dark.
VOC values of all cells were measured at a light intensity (100
mW/cm2) much stronger than those (5−22 mW/cm2) used in
the IMPS and IMVS studies. The results of both the interfacial
resistance (from EIS in the dark) and electron lifetime (from
IMVS) support the analogous charge recombination of DUY1−
DUY4- sensitized cells, even though the dye loading of DUY1
is significantly lower than that of DUY2.
TiO2 conduction band, the charge recombination of electrons
from the TiO2 conduction band to the oxidized dye, and the
oxidizing component in the redox couple. The R2 values
calculated from the EIS spectra are also listed in Table 1. Under
illumination, the DUY2-dyed cell has the smallest R2,
indicating that the cell possesses the most facile electron-
transport cycle and therefore exhibits the largest JSC. The order
of the R2 values (DUY3 > DUY1 > DUY4 > DUY2) indicated
that both the dye loading and driving force of dye regeneration
are important parameters for determining the R2 value under
illumination. Higher dye loading and dye regeneration driving
force result in a smaller R2 value. On the other hand, in the
dark conditions, the R2 values for all cells are very similar and
much larger than the R2 values under illumination, suggesting
that there is no big difference in the charge recombination of all
DUY-sensitized cells; therefore, the VOC values for all cells are
also very close (see Table 1).
The kinetics of dye regeneration can be derived from the
TAS spectra of the cells illustrated in Figure S13 in the SI. Two
regeneration times, τ1 and τ2, corresponding to the reduction of
the oxidized dye by the electrolyte and by the electron on TiO2,
respectively,6 are also listed in Table 1. For all devices, τ1 is
much smaller than τ2, which means that reduction of the
oxidized dye by iodide in the electrolyte is the major process for
dye regeneration. The regeneration time is usually related to
the HOMO level of the dye molecule: the lower the HOMO
level, the shorter the regeneration time.6 This tendency was
distinctly observed when the HOMO levels and τ1 values of
sensitizers possessing the same CF3 positions (such as the
DUY1/DUY3 pair or the DUY2/DUY4 pair) were compared.
Nevertheless, the trend of τ1 was not consistent with the order
of the HOMO levels for sensitizers having different CF3
positions (like the DUY1/DUY2 pair or the DUY3/DUY4
pair). This phenomenon indicates that the kinetics of dye
regeneration also depends on other factors (such as the
positions of the CF3 groups on the cyclometalating ligands)
besides the HOMO levels. In the working condition of a DSC,
the oxidation of a dye would involve the excitation of one
electron from the HOMO to the LUMO and then excited
electron injection from the LUMO to the TiO2 conduction
band edge. After electron injection, the HOMO of the dye
becomes a single occupied molecular orbital (SOMO), which
involves an α highest occupied spin orbital (α-HOSO) and a β
lowest unoccupied spin orbital (β-LUSO).36,37 The calculated
Figure 6 displays the electron-transport time (τtr) and
electron lifetime (τn, also called recombination time) of cells
obtained from the IMPS and IMVS data. τtr and τn represent
the time that the electrons on TiO2 transit across the
−
photoanode and recombine with I3 in the electrolyte,
respectively. For a good performance cell, the electron-
transport time should be shorter than the electron lifetime.35
τtr values for all devices studied are much smaller than τn values,
suggesting that all cells could have high efficiencies. τtr is in the
ascending order of DUY2 < DUY4 < DUY1 < DUY3 (see
Figure 6A), which is consistent with the JSC order (DUY2 >
DUY4 > DUY1 > DUY3; see Table 1). Like the data shown in
Figure 6B, at low light intensity, the DUY4-sensitized cell has
the highest τn value. Nevertheless, when the light intensity
increases, the τn values of all cells tend to be very close. The
VOC values of all cells (see Table 1) are very close because the
F
DOI: 10.1021/acs.inorgchem.6b02088
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
β-LUSO of the oxidized DUY sensitizers, displayed in Figure 7,
shows the most orbital character on the ruthenium centers and
CONCLUSION
■
In summary, six NCS-free cyclometalated ruthenium complexes
were synthesized and examined in the context of structure−
property (including the photovoltaic performance) relation-
1
ships. The trans effect and H NMR spectra were successfully
used to define the configurations of the heteroleptic octahedral
complexes DUY3 and DUY4 to be the cyclometalating ring
positioned trans to 2,2′-bipyridine-4,4′-dicarboxylate. The light-
harvesting ability can be increased by extending the π
conjugation of the ancillary ligands; however, it will raise the
HOMO level of the sensitizer. Installing the electron-
withdrawing groups on the cyclometalating ligands can lower
the HOMO level for effective dye regeneration. The
experimental data indicate that two m-CF3 groups on the
cyclometalated position reduce the electron density at the metal
center less significantly than those at ortho and para positions.
The distribution of β-LUSO of the sensitizers reveals that the
dye containing two m-CF3 groups on the cyclometalated ligand
is more favorable for the interaction between the oxidized dyes
and the iodide ion because of less steric hindrence. DUY2 has
two CF3 groups in the meta positions, the low HOMO level
and the high dye loading. As a consequence, the corresponding
cell exhibited the best efficiency of 9.03%, which is 5% higher
than that of a N719-based cell fabricated at the same condition.
This study provides a new direction for designing new high-
efficiency NCS-free cyclometalated sensitizers for DSC.
Figure 7. DFT models of the oxidized DUY1−DUY4 sensitizers, along
with their β-LUSO distributions and the proposed approaches of I− to
the oxidized dyes.
some on the cyclometalated ligand, indicating that the positive
charge of the dye cations is more localized on the ruthenium.
Dye regeneration corresponds to the transfer of one electron
from the iodide ion to the β-LUSO of the oxidized dye, and the
more widely distributed β-LUSO will result in better interaction
between the oxidized dye and the iodide ion. Nevertheless,
once dye regeneration occurs through direct Ru−I interaction,
a seven-coordination Ru3+ complex (which appears not to exist
in the literature) will form.38 Additionally, all attempts by
Privalov and co-workers had failed to computationally optimize
a seven-coordinated structure of a RuIII complex.9 Therefore,
instead of direct Ru−I interaction, we consider that dye
regeneration should take place at other atoms that contribute to
the β-LUSO such as the cyclometalated ancillary ligand. After
carefully considering the relationship of the dye regeneration
time (Table 1) and the β-LUSO distribution on cyclo-
metalating ligands (Figure 7), we believe that the dye
regeneration time is strongly influenced by the β-LUSO
contribution from the 2,4-bis(trifluoromethyl)phenyl unit
more than that from the EDOTSHexyl unit because the
hexyl chain can prevent the approach of the iodide ion. At the
same isovalue = 0.02 (Figure 7), [DUY2]+ and [DUY4]+
possess a larger β-LUSO character on the 2,4-bis-
(trifluoromethyl)phenyl ring than [DUY1]+ and [DUY3]+,
respectively. Furthermore, the calculated β-LUSOs displayed in
Figure 7 reveal that the β-LUSO of the all oxidized dye
([DUY]+) is distributed more on the o- and p-carbon atoms
compared to the aromatic m-carbon atoms on the cyclo-
metalating ligand. The m-CF3 substituents of [DUY2]+ and
[DUY4]+ are more favorable for approaching of the iodide ion
because of the less spatial hindrance of the substituents.
Consequently, DUY2- and DUY4-based cells exhibit faster dye
regeneration and larger JSC than the DUY1- and DUY3-based
cells, respectively.
ASSOCIATED CONTENT
* Supporting Information
■
S
The Supporting Information is available free of charge on the
ACS Publications website at DOI: 10.1021/acs.inorg-
chem.6b02088.
1
Synthesis, H NMR, HRMS, elemental analysis, TD-
DFT-calculated absorption, oscillator strengths, orbital
transition probabilities of the singlet optical transitions,
frontier orbitals of complexes DUY1−DUY6, and TAS
spectra of DUY1−DUY4-sensitized devices (PDF)
AUTHOR INFORMATION
Corresponding Author
*E-mail: t610002@cc.ncu.edu.tw (C.-G.W.).
ORCID
Chun-Guey Wu: 0000-0001-8540-5602
Notes
The authors declare no competing financial interest.
■
ACKNOWLEDGMENTS
■
Financial support (Grant NSC 104-2113-M-008-002-MY3)
from the Ministry of Science and Technology (MOST),
Taiwan, is gratefully acknowledged. The device fabrications and
photovoltaic parameter measurements were performed at the
Advanced Laboratory of Accommodation and Research for
Organic Photovoltaics, MOST, Taiwan, Republic of China.
REFERENCES
■
(1) Sridhar, N.; Freeman, D. 26th European Photovoltaic Solar Energy
Conference and Exhibition, Hamburg, Germany, 2011; Fraunhofer-
Institut fur Solare Energiesysteme ISE: Breisgau, Germany, 2011; pp
̈
232−236.
(2) Hagfeldt, A.; Boschloo, G.; Sun, L.; Kloo, L.; Pettersson, H. Dye-
sensitized solar cells. Chem. Rev. 2010, 110, 6595−6663.
G
DOI: 10.1021/acs.inorgchem.6b02088
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
(3) Yum, J.-H.; Baranoff, E.; Wenger, S.; Nazeeruddin, M. K.; Gratzel,
M. Panchromatic engineering for dye-sensitized solar cells. Energy
Environ. Sci. 2011, 4, 842−857.
(19) Li, G.; Wu, S. J.; Wu, C. G. Structural design of ruthenium
sensitizer compatible with cobalt electrolyte for a dye-sensitized solar
cell. J. Mater. Chem. A 2014, 2, 17551−17560.
̈
(20) Frisch, M. J.; Trucks, G. W.; Schlegel, H. B.; Scuseria, G. E.;
Robb, M. A.; Cheeseman, J. R.; Scalmani, G.; Barone, V.; Mennucci,
B.; Petersson, G. A.; Nakatsuji, H.; Caricato, M.; Li, X.; Hratchian, H.
P.; Izmaylov, A. F.; Bloino, J.; Zheng, G.; Sonnenberg, J. L.; Hada, M.;
Ehara, M.; Toyota, K.; Fukuda, R.; Hasegawa, J.; Ishida, M.; Nakajima,
T.; Honda, Y.; Kitao, O.; Nakai, H.; Vreven, T.; Montgomery, J. A., Jr.;
Peralta, J. E.; Ogliaro, F.; Bearpark, M.; Heyd, J. J.; Brothers, E.; Kudin,
K. N.; Staroverov, V. N.; Kobayashi, R.; Normand, J.; Raghavachari, K.;
Rendell, A.; Burant, J. C.; Iyengar, S. S.; Tomasi, J.; Cossi, M.; Rega,
N.; Millam, J. M.; Klene, M.; Knox, J. E.; Cross, J. B.; Bakken, V.;
Adamo, C.; Jaramillo, J.; Gomperts, R.; Stratmann, R. E.; Yazyev, O.;
Austin, A. J.; Cammi, R.; Pomelli, C.; Ochterski, J. W.; Martin, R. L.;
Morokuma, K.; Zakrzewski, V. G.; Voth, G. A.; Salvador, P.;
Dannenberg, J. J.; Dapprich, S.; Daniels, A. D.; Farkas, O.;
Foresman, J. B.; Ortiz, J. V.; Cioslowski, J.; Fox, D. J. Gaussian 09,
revision A.02; Gaussian, Inc.: Wallingford, CT, 2009.
(4) Chen, C.-Y.; Wu, S.-J.; Wu, C.-G.; Chen, J. G.; Ho, K. C. A
ruthenium complex with superhigh light-harvesting capacity for dye-
sensitized solar cells. Angew. Chem., Int. Ed. 2006, 45, 5822−5825.
(5) Hussain, M.; Islam, A.; Bedja, I.; Gupta, R. K.; Han, L.; El-Shafei,
A. A comparative study of Ru(II) cyclometallated complexes versus
thiocyanated heteroleptic complexes: thermodynamic force for
efficient dye regeneration in dye-sensitized solar cells and how low
could it be? Phys. Chem. Chem. Phys. 2014, 16, 14874−14881.
(6) Li, J. Y.; Lee, C.; Chen, C. Y.; Lee, W. L.; Ma, R.; Wu, C. G.
Diastereoisomers of ruthenium dyes with unsymmetric ligands for
DSC: Fundamental chemistry and photovoltaic performance. Inorg.
Chem. 2015, 54, 10483−10489.
(7) Yu, Q.; Liu, S.; Zhang, M.; Cai, N.; Wang, Y.; Wang, P. An
extremely high molar extinction coefficient ruthenium sensitizer in
dye-sensitized solar cells: The effects of π-conjugation extension. J.
Phys. Chem. C 2009, 113, 14559−14566.
(21) Becke, A. D. Density-functional thermochemistry. III. The role
of exact exchange. J. Chem. Phys. 1993, 98, 5648−5652.
(8) Nazeeruddin, M. K.; De Angelis, F.; Fantacci, S.; Selloni, A.;
Viscardi, G.; Liska, P.; Ito, S.; Takeru, B.; Gratzel, M. Combined
̈
(22) Lee, C.; Yang, W.; Parr, R. G. Development of the Colle-Salvetti
correlation-energy formula into a functional of the electron density.
Phys. Rev. B: Condens. Matter Mater. Phys. 1988, 37, 785−789.
(23) Hay, P. J.; Wadt, W. R. Ab initio effective core potentials for
molecular calculations. Potentials for the transition metal atoms Sc to
Hg. J. Chem. Phys. 1985, 82, 270−283.
experimental and DFT-TDDFT computational study of photo-
electrochemical cell ruthenium sensitizers. J. Am. Chem. Soc. 2005,
127, 16835−16847.
(9) Privalov, T.; Boschloo, G.; Hagfeldt, A.; Svensson, P. H.; Kloo, L.
A study of the interactions between I¯/I3¯ redox mediators and
organometallic sensitizing dyes in solar cells. J. Phys. Chem. C 2009,
113, 783−790.
(10) Hoggard, P. E.; Porter, G. B. Photoanation of the tris(2,2′-
bipyridine)ruthenium(II) cation by thiocyanate. J. Am. Chem. Soc.
1978, 100, 1457−1463.
(11) Nguyen, H. T.; Ta, H. M.; Lund, T. Thermal thiocyanate ligand
substitution kinetics of the solar cell dye N719 by acetonitrile, 3-
methoxypropionitrile, and 4-tert-butylpyridine. Sol. Energy Mater. Sol.
Cells 2007, 91, 1934−1942.
(12) Abbotto, A.; Coluccini, C.; Dell’ Orto, E.; Manfredi, N.;
Trifiletti, V.; Salamone, M. M.; Ruffo, R.; Acciarri, M.; Colombo, A.;
Dragonetti, C.; Ordanini, S.; Roberto, D.; Valore, A. Thiocyanate-free
cyclometalated ruthenium sensitizers for solar cells based on
heteroaromatic-substituted 2-arylpyridines. Dalton Trans. 2012, 41,
11731−11738.
(24) McLean, A. D.; Chandler, G. S. Contracted Gaussian basis sets
for molecular calculations. I. Second row atoms, Z = 11−18. J. Chem.
Phys. 1980, 72, 5639−5648.
(25) Cossi, M.; Rega, N.; Scalmani, G.; Barone, V. Energies,
structures, and electronic properties of molecules in solution with the
C-PCM solvation model. J. Comput. Chem. 2003, 24, 669−681.
(26) Cossi, M.; Barone, V. Time-dependent density functional theory
for molecules in liquid solutions. J. Chem. Phys. 2001, 115, 4708−4717.
(27) Barone, V.; Cossi, M. Quantum calculation of molecular
energies and energy gradients in solution by a conductor solvent
model. J. Phys. Chem. A 1998, 102, 1995−2001.
(28) Coluccini, C.; Manfredi, N.; Calderon, E. H.; Salamone, M. M.;
Ruffo, R.; Roberto, D.; Lobello, M. G.; De Angelis, F.; Abbotto, A.
Photophysical and electrochemical properties of thiophene-based 2-
arylpyridines. Eur. J. Org. Chem. 2011, 2011, 5587−5598.
(29) In fact, a racemate (meaning a mixture of two enantiomers; also
called a racemic mixture) also cannot be distinguished by NMR
spectroscopy. Typical laboratory purification methods (normal-phase,
reversed-phase, and size-exclusion chromatography, recrystallization,
distillation, extraction, etc.) cannot separate two enantiomers because
of their same chemical and physical properties. Importantly, in an
achiral environment such as a DSC cell, two enantiomers behave in the
same way. Therefore, we support that the isolation of each enantiomer
is really unnecessary in DSC.
(13) Bomben, P. G.; Gordon, T. J.; Schott, E.; Berlinguette, C. P. A
trisheteroleptic cyclometalated RuII sensitizer that enables high power
output in a dye-sensitized solar cell. Angew. Chem., Int. Ed. 2011, 50,
10682−10685.
(14) Bomben, P. G.; Koivisto, B. D.; Berlinguette, C. P.
Cyclometalated Ru complexes of type [RuII(N∧N)2(C∧N)]z:
Physicochemical response to substituents installed on the anionic
ligand. Inorg. Chem. 2010, 49, 4960−4971.
́
(15) Bomben, P. G.; Theriault, K. D.; Berlinguette, C. P. Strategies
for optimizing the performance of cyclometalated ruthenium
sensitizers for dye-sensitized solar cells. Eur. J. Inorg. Chem. 2011,
2011, 1806−1814.
(30) Hartley, F. R. The cis- and trans-effects of ligands. Chem. Soc.
Rev. 1973, 2, 163−179.
(31) DUY2 and its ester form show a small δ difference (Δδ = 0.1
ppm) of their two ortho protons, related to the ester-to-acid
conversion, while two ortho protons of DUY4 and its ester form are
almost the same (Δδ = 0.013 ppm). This indicates that the ester-to-
acid conversion of ligand Et2dcbpy has not influenced the magnetic
environment around the ortho protons of DUY4 and its ester form, or,
in other words, Et2dcbpy orients trans to the benzene ring and ligand
B20 trans to the pyridine ring of the cyclometalating ligand.
(32) Boschloo, G.; Gibson, E. A.; Hagfeldt, A. Photomodulated
voltammetry of iodide/triiodide redox electrolytes and its relevance to
dye-sensitized solar cells. J. Phys. Chem. Lett. 2011, 2, 3016−3020.
(33) Chen, C. Y.; Pootrakulchote, N.; Chen, M. Y.; Moehl, T.; Tsai,
(16) Bomben, P. G.; Borau-Garcia, J.; Berlinguette, C. P. Three is not
a crowd-efficient sensitization of TiO2 by a bulky trichromic
trisheteroleptic cycloruthenated dye. Chem. Commun. 2012, 48,
5599−5601.
(17) Polander, L. E.; Yella, A.; Curchod, B. F. E.; Astani, N. A.;
Teuscher, J.; Scopelliti, R.; Gao, P.; Mathew, S.; Moser, J.-E.;
Tavernelli, I.; Rothlisberger, U.; Gratzel, M.; Nazeeruddin, M. K.;
̈
Frey, J. Towards compatibility between ruthenium sensitizers and
cobalt electrolytes in dye-sensitized solar cells. Angew. Chem., Int. Ed.
2013, 52, 8731−8735.
(18) Funaki, T.; Otsuka, H.; Onozawa-Komatsuzaki, N.; Kasuga, K.;
Sayama, K.; Sugihara, H. Systematic evaluation of HOMO energy
levels for efficient dye regeneration in dye-sensitized solar cells. J.
Mater. Chem. A 2014, 2, 15945−15951.
H. H.; Zakeeruddin, S. M.; Wu, C. G.; Gratzel, M. A new heteroleptic
̈
ruthenium sensitizer for transparent dye-sensitized solar cells. Adv.
Energy Mater. 2012, 2, 1503−1509.
H
DOI: 10.1021/acs.inorgchem.6b02088
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
(34) Wang, Q.; Moser, J. E.; Gratzel, M. Electrochemical impedance
̈
spectroscopic analysis of dye-sensitized solar cells. J. Phys. Chem. B
2005, 109, 14945−14953.
(35) Bisquert, J.; Fabregat-Santiago, F.; Mora-Sero,
́
I.; Garcia-
Belmonte, G.; Gimen
́
ez, S. Electron lifetime in dye-sensitized solar
cells: Theory and interpretation of measurements. J. Phys. Chem. C
2009, 113, 17278−17290.
(36) Merhi, A.; Zhang, X.; Yao, D.; Drouet, S.; Mongin, O.; Paul, F.;
Williams, J. A. G.; Fox, M. A.; Paul-Roth, C. O. New donor−acceptor
conjugates based on a trifluorenylporphyrin linked to a redox-
switchable ruthenium unit. Dalton Trans. 2015, 44, 9470−9485.
(37) Simon, S. J. C.; Parlane, F. G. L.; Swords, W. B.; Kellett, C. W.;
Du, C.; Lam, B.; Dean, R. K.; Hu, K.; Meyer, G. J.; Berlinguette, C. P.
Halogen bonding promotes higher dye-sensitized solar cell photo-
voltages. J. Am. Chem. Soc. 2016, 138, 10406−10409.
(38) Clifford, J. N.; Palomares, E.; Nazeeruddin, M. K.; Gratzel, M.;
̈
Durrant, J. R. Dye dependent regeneration dynamics in dye sensitized
nanocrystalline solar cells: evidence for the formation of a ruthenium
bipyridyl cation/iodide intermediate. J. Phys. Chem. C 2007, 111,
6561−6567.
I
DOI: 10.1021/acs.inorgchem.6b02088
Inorg. Chem. XXXX, XXX, XXX−XXX