Diastereoisomers of Ruthenium Dyes with Unsymmetric Ligands for DSC: Fundamental Chemistry and Photovoltaic Performance

Article abstract of DOI:10.1021/acs.inorgchem.5b01967

A new thiocyanic acid-free ruthenium sensitizer, CYC-B29, containing two unsymmetrical ancillary ligands, was synthesized, and its three diastereoisomers CYC-B29-CC, CYC-B29-TT, and CYC-B29-CT with significantly different optical, electronic, and electrochemical properties were carefully separated. CYC-B29-TT with the smallest size has the strongest absorption coefficient of the MLCT band, the shortest λ_max, the lowest highest occupied molecular orbital level and the highest dye loading. Therefore, dye-sensitized solar cell based on CYC-B29-TT has the highest efficiency, which is two times higher than that of CYC-B29-CC-sensitized device and 10% higher than that of N719-based cell. Time-dependent density functional theory-calculated transition bands for the three isomers are not identical, and only CYC-B29-TT has the calculated transition bands close to the experimental absorption profile. Although the calculated transition bands for CYC-B29-CC and CYC-B29-CT are not consistent with the experimental data, the ground-state vertical excitation energy with oscillator strength and electron-density difference map data combining with the dye loading predict correctly the order of the photocurrent for the three isomers sensitized devices.

Full text of DOI:10.1021/acs.inorgchem.5b01967

Article
pubs.acs.org/IC
Diastereoisomers of Ruthenium Dyes with Unsymmetric Ligands for
DSC: Fundamental Chemistry and Photovoltaic Performance
Jheng-Ying Li,‡ Chia Lee,† Chia-Yuan Chen,†,‡ Wen-Long Lee,† Ren Ma,† and Chun-Guey Wu*,†,‡
‡
†Department of Chemistry and Research Center for New Generation Photovoltaics, National Central University, Jhong-Li 32001,
Taiwan, ROC
S
* Supporting Information
ABSTRACT: A new thiocyanic acid-free ruthenium sensitizer, CYC-B29,
containing two unsymmetrical ancillary ligands, was synthesized, and its three
diastereoisomers CYC-B29-CC, CYC-B29-TT, and CYC-B29-CT with signifi-
cantly different optical, electronic, and electrochemical properties were carefully
separated. CYC-B29-TT with the smallest size has the strongest absorption
coefficient of the MLCT band, the shortest λmax, the lowest highest occupied
molecular orbital level and the highest dye loading. Therefore, dye-sensitized solar
cell based on CYC-B29-TT has the highest efficiency, which is two times higher
than that of CYC-B29-CC-sensitized device and 10% higher than that of N719-
based cell. Time-dependent density functional theory-calculated transition bands
for the three isomers are not identical, and only CYC-B29-TT has the calculated
transition bands close to the experimental absorption profile. Although the
calculated transition bands for CYC-B29-CC and CYC-B29-CT are not consistent
with the experimental data, the ground-state vertical excitation energy with oscillator strength and electron-density difference
map data combining with the dye loading predict correctly the order of the photocurrent for the three isomers sensitized devices.
groups. They used one 4,4′-carboxy-2,2-bypyridine and two 2-
pyridyl pyrazolate ligands to synthesize a series of neutral,
thiocyanate-free Ru(II) sensitizers (TFRS-x). DSC based on
TFRS-2 has the efficiency of 9.54% (vs 8.56% for N719-
sensitized device).19 They also used 2-pyridyl triazolate27 or
isoquinolinyl pyrazolate28 to prepare a series of NCS-free
heteroleptic tris-bidentate cyclometalated ruthenium sensitizers
and demonstrated that all these NSC-free ruthenium sensitizers
have good photovoltaic performance and long-term stabil-
ity.19−28
Most bidentate ancillary ligands used in NCS-free ruthenium
complexes have an unsymmetrical structure, and the resulting
complex has three diastereoisomers. These three diaster-
eoisomers (not easy to be separated) may have different
physicochemical properties and photovoltaic performance. A
paper, published by Chi29 et al. last year, has studied the
photovoltaic performance of two old dyes containing unsym-
metrical ancillary ligand. They isolated only two diaster-
eoisomers (of the three) and found that the physicochemical
properties of the two diastereoisomers are very similar.
Therefore, the variation in the photovoltaic performance of
the two isomers is mainly due to the different surface coverage
of the mesoporous TiO2 film, which was determined by the
molecular conformation of the isomers. We are developing
high-efficiency NCS-free ruthenium dyes for DSC and find that
the efficiency of the device is affected apparently by the purity
INTRODUCTION
■
Dye-sensitized solar cell (DSC) showed great promise for the
future renewable energy due to its possible low manufacture
cost and diverse materials available for every component in
DSC.1−3 Dye molecule (sensitizer), which absorbs photon and
produces movable electron, plays a vital role in the light
harvesting, charge separation, and the overall power conversion
efficiency (PCE) of the devices. Thousands of dye molecules
were scrutinized, and ruthenium complexes have been proved
to be one of the most effective sensitizers to be used in DSC.
The high photovoltaic performance (i.e., efficiency > 11%)
ruthenium-based dyes (such as N719,4 CYC-B11,5 and C1066)
were all derived from N3 ([Ru(dcbpy)2(NCS)2]; dcbpy = 4,4′-
dicarboxy-2,2′-bipyridine) dye,7 which contains two thiocyanate
(NCS−) ligands. Despite their excellent efficiency, thiocyanate
is a monodentate ligand, which may be unstable under
excessive thermal stress and/or light soaking.8,9 This dis-
advantage has triggered the research for the effective chelating
ligands to replace thiocyanate in ruthenium-based sensi-
tizers.10−17
A first successful example for using NCS-free ruthenium
sensitizer in DSC to achieve high (10.1%) efficiency was
reported by Gratzel et al., who used a bidentated 2,4-
̈
difluorophenylpyridinato to replace two NCS to prepare a
cyclometalated tris-bidentate ruthenium complex, YE05.18
Subsequent studies on using cyclometalating chelates19−24
and other multidentate ligands25,26 to prepare NSC-free Ru(II)
complexes for DSC applications were reported by several
research groups, and Chi et al. is one of the representative
Received: August 28, 2015
© XXXX American Chemical Society
A
DOI: 10.1021/acs.inorgchem.5b01967
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
Figure 1. Structure of the three diastereoisomers: CYC-B29-CC, CYC-B29-TT, and CYC-B29-CT.
of the isomers. Therefore, the physicochemical properties of
the three diastereoisomers may not be the same. To really
understand the fundamental chemistry and photophysics of the
diastereoisomers, a new NCS-free ruthenium sensitizer (code
CYC-B29) with unsymmetrical ancillary ligands was prepared.
The three diastereoisomers (CYC-B29-CC, CYC-B29-TT, and
CYC-B29-CT) were carefully separated, and their photovoltaic
performance was investigated. The three diastereoisomers are
different not only in the photovoltaic performance but also in
the optical, electronic, and electrochemical properties.
RESULTS AND DISCUSSION
■
The structures of the three diastereoisomers of 4-(5-thiohexyl-
(3,4-ethylene- dioxythiophen-2-yl))-2-(3-(trifluoromethyl)-1H-
pyrazol-5-yl)pyridine (CYC-B29) are displayed in Figure 1:
their names (CYC-B29-CC, CYC-B29-TT, and CYC-B29-CT)
were defined by the pyridine unit in the ancillary ligand,
considering if it is trans or cis to the pyridine ring of the
anchoring ligand. The detailed two-step synthetic procedure,
isolation, and identification of the three isomers can be found in
the electronic Supporting Information. The 1H NMR spectra of
the three isomers are displayed in Figure S1, and the peak
assignment is based on the paper reported by Chi19et al. The
yields of CYC-B29-CC, CYC-B29-TT, and CYC-B29-CT are
8%, 17%, and 10%, respectively. CYC-B29-CC and CYC-B29-
CT isomers have a similar yield; nevertheless, the yield of CYC-
B29-TT, which has the two pyrazole (Paz) units trans to each
other, is almost two times of that for the other two isomers.
This may be because the trans effect of Paz is higher than that
of pyridine, which means the ligand trans to Paz is less stable
compared to that trans to pyridine. Therefore, two Paz units
trans to each other (in TT isomer) are more stable compared
those (CC and CT isomers) with two Paz units cis to each
other. We prepared several batches of CYC-B29 dye, and the
ratios of the yields for the three isomers are very close: TT
isomer has the yield always approximately two times of the
other two isomers.
Figure 2. Absorption spectra of the three diastereoisomers measured
in DMF and immobilized on the transparent TiO2 film (thickness is
ca. 3 μm).
diastereoisomers in DMF have significantly different absorption
profiles: CYC-B29-CC has a strong absorption peak at 462 nm
with a shoulder at longer wavelength, CYC-B29-CT has λmax at
489 nm with a shoulder at each side of the peak, whereas CYC-
B29-TT has two defined peaks centered at 434 and 540 nm,
respectively; all these bands may relate to the metal-to-ligand
charge transfer (MLCT) bands. Furthermore, the absorption
intensity at the visible region (the integration of the absorption
coefficient between 350 and 750 nm) for the three isomers is
also different: CYC-B29-TT and CYC-B29-CT have similar
integrated visible light absorption intensity, while the integrated
absorption intensity of CYC-B29-CC is much lower than the
other two isomers. The possible reason for the low light
harvesting ability of CYC-B29-CC isomer will be discussed
more in later paragraphs.
All isomers are blue-shifted when adsorbed on TiO2 thin
film, indicating that CYC-B29 undergoes a H-aggregation or a
conformational change induced by the deprotonation in the
self-assembly process. The absorption spectra of the specific
dyes varied when adsorbed on TiO2 anode. For example the
absorption profiles for CYC-B29-TT and CYC-B29-CT
adsorbed TiO2 films are similar, even though their absorption
profiles are different in DMF. On the contrary, the absorption
profiles of CYC-B29-CC and CYC-B29-TT in DMF are similar,
but they differ after adsorption on TiO2 film. Nevertheless, the
absorption of CYC-B29-CC adsorbed on TiO2 film is much
The electronic absorption spectra of the three diaster-
eoisomers measured in dimethylformamide (DMF) and
immobilized on the transparent TiO2 film (thickness is ca. 3
μm) are displayed in Figure 2, and the optical data were
summarized in Table S1, Supporting Information. The three
B
DOI: 10.1021/acs.inorgchem.5b01967
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
weaker (especially at long wavelength, which correlated to
MLCT band) compared to those of the other two isomers
adsorbed on TiO2, consistent with the smallest absorption
coefficient of CYC-B29-CC dye. The significant variation in the
absorption profile for the dye dissolved in DMF and adsorbed
on TiO2 was not observed in the high-performance ruthenium
dye with two NCS ligands we prepared.5,30,31 Interestingly, for
the dyes adsorbed on a TiO2 film, the absorption onset extends
to longer wavelength for all three isomers. The widening in
absorption profile is reasonably attributed to dye molecules
with various conformations coexisting on TiO2 surface, due do
the bulky nature of NCS-free ruthenium complexes.
Furthermore, the absorption intensity (in the visible region)
of CYC-B29-TT is close to that of CYC-B29-CT in DMF, but
CYC-B29-TT-sensitized TiO2 thin film has stronger absorption
than CYC-B29-CT-sensitized TiO2. This is due to CYC-B29-
TT with symmetric slender conformation occupying less space
when adsorbed on TiO2 compared to CYC-B29-CT, which has
an unsymmetrical more spherical-like conformation.
The graphical representation of the computed frontier
orbitals for the three isomers is displayed in Figures S2−S4.
The distribution of the frontier orbitals for the three isomers is
similar: highest occupied molecular orbital (HOMO),
HOMO−1, and HOMO−2 are mainly located on the metal
center and Paz units, while HOMO−3 and HOMO−4 are
contributed primarily from the ethylene−dioxythiophene
(EDOT)−pyridine moiety of the ancillary ligand. However,
the lowest unoccupied molecular orbital (LUMO), LUMO+1,
and LUMO+4 locate on the anchoring ligand, while LUMO+2
and LUMO+3 mainly distribute on the ancillary ligands.
Nevertheless, since the relative position of the two Paz units is
different, the contour of the electron density map of the frontier
orbitals is quite different in three isomers, even though the
orbitals are contributed by the same chemical identity. For
example the HOMO of both CC and TT isomers is mainly
contributed from Ru center and two Paz units. However, the
conformation of two Paz units in CC is bending, whereas that
for TT is linear. Similar situations happened in LUMO+2 and
LUMO+3 orbitals, which distribute predominantly on EDOT
units of the ancillary ligands. The significant difference in the
absorption spectra for the three diastereoisomers should be
attributed to the difference in the contour of the electron
density map of the frontier orbitals. Nevertheless, the subtle
different contour of the electron density map cannot be used to
explain the significant difference in the absorption profile of the
three diastereoisomers.
To gain more insight into the absorption bands, time-
dependent density functional theory (TD-DFT) calculation of
the singlet electronic transition was performed with DMF as
the solvent. The calculated absorption peaks and electron
density difference maps (EDDMs) as well as the experimental
absorption spectra for the three isomers are displayed in Figure
3, and the detailed transition assignments are listed in Tables
S2−S4. The computed singlet transitions for the three isomers
have no big difference. All isomers have a moderate transition
at 535−545 nm, belong to H-2→ L transition, and several
peaks occur at 400−460 nm although the relative intensity of
the peaks for the three isomers is not the same. However, only
CYC-B29-TT has the calculation result that agreed well with
the experimental spectrum, considering the energy and relative
intensity of the peaks. The results reveal the limitation of the
theoretical calculations when it was used to calculate the
difference in the optical property of the diastereoisomers. CYC-
Figure 3. Calculated (DT-DFT) absorption peaks and EDDMs as well
as the experimental absorption spectra for the three diastereoisomers
of CYC-B29.
B29-TT has four strongest transitions occur at 544.4 nm (S3
with oscillator strength (f) = 0.1558), 458.8 nm (S4 with f =
0.2446), 453.3 nm (S5 with f = 0.0725), and 420.8 nm (S13 with
f = 0.4901). Three of them (except that at 420.8 nm) with total
oscillator strength of 0.4729 are attributed to the beneficial
(MLCT) charge transfer from Ru orbitals (or Ru-Paz mixed
orbitals) to the anchoring ligand. These transitions are believed
to be contributing significantly to the photo absorption,
electron injection, and photocurrent production. Nevertheless,
the calculated and experimental spectra for CYC-B29-CC are
not the same. Peaks for the four strongest calculated
absorptions are 534.5 nm (S3 with oscillator strength (f) =
0.1231), 445.3 nm (S8 with f = 0.1442), 421.7 nm (S12 with f =
C
DOI: 10.1021/acs.inorgchem.5b01967
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
0.3437), and 402.8 nm (S18 with f = 0.6623). Only two of the
four peaks with total oscillator strength of 0.2673 are strongly
related to the metal-to-anchoring ligand charge transfer. The
calculated absorption bands are also not quite consistent with
the experimental absorption spectrum for CYC-B29-CT. The
four strongest calculated absorptions are 536.9 nm (S3 with
oscillator strength (f) = 0.1350), 462.0 nm (S5 with f = 0.1987),
436.8 nm (S9 with f = 0.2114), and 419.7 nm (S14 with f =
0.2162), and three of the four transitions (total oscillator
strength is 0.5451) are mainly contributed from Ru-Paz to the
bipyridine ring (not the COOH) of the anchoring ligand
charge transfer. Therefore, the excited electron (from the
absorption at 419.7 nm) may linger at the bipyridine unit,
cannot inject to TiO2 through COOH anchoring moiety. On
the basis of only the optical data (both theoretic and
experimental) without considering the dye loading, we may
expect that CYC-B29-CT-based DSC has the highest Jsc value,
and the Jsc of the cell sensitized with CYC-B29-CC is the
lowest.
The driving force for electron injection from the photo-
excited dye into the conduction band of TiO2 and the
regeneration of oxidized dye by the electron from the redox
couple depends on the energy level of the frontier orbitals of
the dye molecule. The HOMO and LUMO energy levels of the
three isomers determined by the square-wave voltammograms
and the absorption thresholds are displayed in Figure 4. The
Figure 5. I−V and IPCE curves of CYC-B29 sensitized devices.
photovoltaic parameters as well as the dye loading are
summarized in Table 1. The IPCE spectra of the devices
sensitized with the three isomers all show a broad spectral
response in the range from 350 to 800 nm, reaching a
maximum value of 58%, 76%, and 66% for Device A (CYC-
B29-CC-sensitized cell), Device B (CYC-B29-TT-sensitized
cell), and Device C (CYC-B29-CT-sensitized cell), respectively.
The IPCE value of the Device B is higher than those of the
Device A and Device C at the whole measured wavelength,
consistent with the absorption data of the dye-adsorbed TiO2
thin film shown in Figure 2. Moreover, the dye loading for
CYC-B29-TT is higher than that of the other two isomers (see
Table 1). As shown in Figure 1 and Figure 3, the conformation
of CYC-B29-TT molecule with the two longer axes parallel to
the anchoring ligand occupies less TiO2 surface compared to
CYC-B29-CC or CYC-B29-CT isomers, which has two longer
axes perpendicular to the anchoring ligand or one perpendic-
ular, one parallel to the anchoring ligand. The absorption (dye
loading times absorption coefficient) of CYC-B29-TT is the
strongest among the three diastereoisomers. That is why the Jsc
of the Device B is higher than that for the Device C (based on
CYC-B29-CT) although the effective MLCT band for CYC-
B29-CT is stronger. Furthermore, CYC-B29-TT also has the
HOMO energy level slightly lower than that of CYC-B29-CT
(see Figure 4); therefore, it will have higher driven force for the
regeneration of the oxidized dye by the electrolyte. This may
also be one of the reasons that Device B has higher Jsc
compared to Device C. Jsc for the Device A listed in Table 1 is
much smaller than that of Device B. It is because CYC-B29-CC
has smaller absorption coefficient, less dye loading, and weaker
effective MLCT transition (from DT-DFT calculation)
compared to CYC-B29-TT. Similar reasons (low absorption
coefficient and weak effective metal-to-ligand charge transfer)
also explain why the dye loading of CYC-B29-CC-sensitized
Figure 4. Voltammograms and energy level of the frontier orbitals for
the three diastereoisomers of CYC-B29.
ground-state potentials of CYC-B29-CC, CYC-B29-TT, and
CYC-B29-CT dissolved in DMF are 0.67, 0.80, and 0.78 V
versus normal hydrogen electrode (NHE), respectively. The
HOMOs of the three isomers are all higher than those for the
complexes with NCS ligands we reported before5,30,31 but still
more positive than the redox potential (∼0.4 V vs NHE) of the
iodide/triiodide electrolyte used in this study. The optical
transition energy, E0−0, determined from its absorption onset
energy Eg measured in DMF is 2.16, 2.04, and 2.15 eV for
CYC-B29-CC, CYC-B29-TT, and CYC-B29-CT, respectively.
Therefore, the LUMO energy levels (LUMO = HOMO − Eg)
for the three isomers are all much more negative than the
conduction band of TiO2 (ca. −0.5 V vs NHE),32 benefiting the
electron injection from the excited dyes to TiO2 anode.
Nevertheless, the HOMO of CYC-B29-CC is only 0.27 V lower
than the redox potential of iodide electrolyte. The energy
difference may not provide sufficient driving force for the dye
to be regenerated by iodide ion, which may result in low short-
circuit current of the corresponding device.
The current−voltage (I−V; measured under the illumination
of AM 1.5G simulated sunlight (100 mW cm−2)) and incident
photon-to-current conversion efficiency (IPCE) curves for
CYC-B29 sensitized devices are shown in Figure 5, and the
D
DOI: 10.1021/acs.inorgchem.5b01967
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
Table 1. Photovoltaic Parameters and TiO2/Dye/Electrolyte Interphase Resistance (R2) of the Cells Sensitized with the Three
Diastereoisomers
dye
device
Jsc (mAcm‑2)
Voc (V)
FF
η (%)
dye loading (mole)
R2 (Ω)
CYC-B29-CC
CYC-B29-TT
CYC-B29-CT
N719
A
B
C
R
8.61
17.94
15.23
14.42
0.704
0.727
0.663
0.772
0.643
0.640
0.659
0.680
3.90
8.34
6.66
7.55
2.8 × 10−8
3.2 × 10−8
2.3 × 10−8
128
578
81.5
TiO2 is close to that of CYC-B29-CT dyed anode but the Jsc of
the Device A is significantly smaller than that of Device C.
Jsc of a cell is also affected by the kinetics of the dye
regeneration, which was influenced by the HOMO level of the
dye molecules. To probe the kinetics of dye regeneration,
transient absorption spectra (TAS, Figure 6) of the devices
the Device A listed in Table 2 is the smallest, suggesting that
the ineffective reducing of the oxidized dye by iodide is the
major reason for the poor photovoltaic performance of Device
A. TAS data can also explain why the cell sensitized by CYC-
B29-CC exhibits such low photocurrent density.
Data listed in Table 1 also reveal that CYC-B29-TT-
sensitized device also has the highest Voc. It is known that the
theoretical Voc of a DSC is defined by the potential difference
between the quasi-Fermi level of TiO2 and the redox potential
of the electrolyte. Therefore, Voc was determined by the energy
level of TiO2 conduction band edge33 and the degree of the
electron recombination (from TiO2 to the oxidized dye or
electrolyte) of the cell.32,34 The former depends on the number
of the electrons injected from the excited dye molecules to
TiO2, and the latter relies on the surface protection of TiO2
anode. CYC-B29-TT has the largest dye loading and the largest
absorption coefficient, the Fermi level of TiO2 sensitized with
CYC-B29-TT should be the highest, results in high Voc. The
degree of the charge recombination (TiO2 surface protection)
for a DSC is influenced not only by the amount of adsorbed
dye molecules but also on the conformation and size of the
sensitizer. The information on the surface protection of TiO2
anode can be obtained from the electrochemical impedance
studies of the device under dark. The TiO2/dye/electrolyte
interphase resistances (R2, derived from the electrochemical
impedance spectroscopy (EIS) diagrams illustrated in Figure
S5) listed in Table 1 clearly reveal that TiO2/dye/electrolyte
interphase resistance of CYC-B29-TT sensitized device (Device
B) is much larger than those for the other two devices (Device
A and Device C) based on CYC-B29-CC and CYC-B29-CT
isomers. Less charge recommendation for the Device B would
also be the reason for the higher Voc. As a result CYC-B29-TT-
sensitized cell has the highest overall efficiency of 8.34%, which
is ca. 10% higher than that of the Device R based on N719 dye
fabricated at the same condition (see Table 1).
Figure 6. Transient absorption spectra of CYC-B29 sensitized devices.
were performed, and the calculated dye regeneration lifetimes
(τ1 and τ2 corresponding to the reduction of the oxidized dye
by I− in electrolyte and by the electrons in TiO2, respectively)5
as well as the amplitude ratios of the two reduction processes
(A1/A2: can be regarded as the ratio of the oxidized dye
reduced by electrolyte and electron in TiO2) are summarized in
Table 2. The driving forces (ED1 and ED2) for the dye
Table 2. Dye Regeneration Driving Force, Lifetime and the
Amplitude Ratio of the Two Dye Regeneration Paths
amplitude
ratio of τ1
a
b
ED1
ED2
and τ2
dye
(V vs NHE) τ1 (μs) (V vs NHE) τ2 (μs)
(A1/A2)
CYCB29-
CC
0.27
0.40
0.38
58.42
30.35
38.84
1.17
1.30
1.28
818.97
436.80
499.29
48/52
63/37
64/36
CYCB29-
TT
CYCB29-
CT
CONCLUSIONS
■
In summary, three diastereoisomers of a new NCS-free
ruthenium sensitizer, CYC-B29, were carefully isolated. These
isomers (CYC-B29-CC, CYC-B29-TT, and CYC-B29-CT)
have significant variation in the molecular conformation and
electronic, electrochemical, and photophysical properties and
therefore have tremendously different photovoltaic perform-
ance when applied in DSC. CYC-B29-CC isomer has the
lowest absorption coefficient, the smallest dye loading, and the
highest HOMO energy level. Device A sensitized with CYC-
B29-CC has the lowest Jsc and PCE. CYC-B29-CT has a
similar dye loading (compared to CYC-B29-CC), but the
corresponding cell has lower Voc, due to the poorer surface
protection of the dyed TiO2 since it has less bulky structure.
However, the efficiency of the Device C (based on CYC-B29-
CT) is still higher than that for Device A, because CYC-B29-
CT has higher absorption coefficient and therefore higher Jsc of
the corresponding cell. CYC-B29-TT (with a less-open
a
b
ED1 = (EHOMO − Eredox of electrolyte) ED2 = (ETiO2 CB − EHOMO
)
regeneration from two different paths are also listed in Table 2.
The lifetime for the dye regenerated by electrolyte or electron
in TiO2 is inversely proportional to the driving force: the higher
the driving force, the shorter the lifetime. Device A has the
longest τ1 indicating that the reducing of the oxidized dye by
iodide in the electrolyte is slower than the other two isomers,
due to the higher HOMO energy level of CYC-B29-CC, and
therefore the oxidized dye will be reduced in more negative
potential. The driving force (ED1) for the dye regeneration of
CYC-B29-CC is also smaller than that for the other two
isomers. Device A also has the longest τ2 value suggesting the
back electron transfer from TiO2 to the oxidized dye is not as
serious as that for Device B and C, because it has the smallest
driving force (ED2). However, the amplitude ratio (A1/A2) for
E
DOI: 10.1021/acs.inorgchem.5b01967
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
fabrication can be found in our earlier report.37 TiO2 film was
immersed into a 200 μM dye solution containing 2−30 mM
chenodeoxycholic acid as a coadsorbent in a mixture of DMSO,
acetonitrile, and tert-butyl alcohol (volume ratio 1:1:1) at room
temperature for 12 h. The electrolyte used for the device fabrication
contains 0.2 M LiI, 0.03 M I2, 0.6 M BMII, 0.1 M GuNCS, and 0.5 M
TBP in acetonitrile unless specified otherwise. The photocurrent−
voltage characteristics were obtained by applying an external voltage
bias to the devices exposed to an irradiance of AM 1.5G (100 mW
cm−2) simulated sunlight (Yamashita Denso, Japan) and measuring the
generated photocurrent using a Keithley 2400 digital source meter
(Keithley, USA). The illuminated active area was 0.25 cm2 using a
metal mask. The dye loading of the DSC was quantified by immersing
the practical TiO2 anodes in a 0.20 mM dye solution, and the
concentration difference before and after anode dye loading was
quantified by optical spectrophotometry.38
Electrochemical Impedance Spectroscopy. The EIS measure-
ment was performed with an Autolab PG30 potentostat in 50 mV
voltage steps with a sinusoidal potential perturbation of 10 mV. The
potential perturbation was applied over a frequency range from 1 MHz
to 0.1 Hz. The resulting impedance spectra were fitted according to
the transmission line model by Bisquert et al. with the ZView software
from Scribner.39 The EIS obtained under dark was used to probe
TiO2/adsorbed dye/electrolyte interphase resistance, which related to
the surface coverage of dyed TiO2 anode.
Transient Absorption Spectroscopy. The TAS measurements
were performed in Proteus Ultrafast System with the light path
designed by us. A laser light with the wavelength of 470 nm was used
to excite the dye molecules in DSC device, and a white light was
passed through the device to measure the absorption of the excited dye
molecules. The decay of the absorption at 820 nm was used to study
the regeneration kinetics of the oxidized dye. The rate constants (or
decay time) were extracted by fitting the data using a Surface Xplorer
software to obtain the rates for the reduction of the oxidized dye by I−
ions in the electrolyte and by charge recommendation with the
electrons in TiO2.40
structure (slender conformation)) therefore has the highest
dye-loading, the lowest HOMO energy level, and the largest
absorption coefficient, the corresponding device (Device B) has
the highest efficiency, which is two times higher than that of
CYC-B29-CC-sensitized cell. Furthermore, only CYC-B29-TT
isomer has the TD-DFT calculated transition bands close to the
experimental absorption spectrum, implying that the electronic
properties of the diastereoisomers may not be predicted
precisely simply by routine theoretical calculation. Never-
theless, the calculated ground-state vertical excitation energy
with oscillator strength and EDDM of the isomers afford the
useful information for understanding the origin of the Jsc when
applied in DSCs. This study provides not only the under-
standing of the basic chemistry of inorganic diastereoisomers
but also the strategy for designing high-efficiency ruthenium
dye for DSC.
EXPERIMENTAL SECTION
■
Materials, Synthesis, and Characterization. All reagents were
obtained from the commercial sources and used as received unless
specified. Solvents were dried over sodium or CaH2 before use. The
ancillary ligand B29 (4-(5-thiohexyl-(3,4-ethylene- dioxy-thiophen-2-
yl))-2-(3-(trifluoromethyl)-1H-pyrazol-5-yl)pyridine) for CYC-B29
was prepared according to the literature report,19 and CYC-B29 dye
was synthesized using two-step synthetic procedure. The detailed
preparation procedures, as well as the characterization of the ancillary
ligand and the final ruthenium complexes, can be found in the
electronic Supporting Information. The structure of CYC-B29 dye was
1
identified with H NMR spectrum, FAB-MS, and elemental analysis.
The structure of the diastereoisomers was mainly assigned according
to the paper19 that has single-crystal data of the modeling compound.
1H NMR spectra were recorded with a DRX-500 NMR spectrometer
in deuterated dimethyl sulfoxide (DMSO-d6.) The FAB-MS spectra
were obtained using JMS-700 HRMS. Elemental analyses were
performed with a Heraeus CHN−O-S Rapid-F002 analysis system.
UV−vis spectra were measured using a Cary 300 Bio spectrometer.
Electrochemical study of the metal complexes dissolved in DMF was
performed in a single-compartment, three-electrode cell with a Pt wire
counter electrode and a platinum disk working electrode. The
reference electrode was Ag/Ag+, and the supporting electrolyte was
0.1 M tetrabutylammonium hexafluorophosphate (TBAPF6) in DMF.
The square-wave voltammograms (potential step increment: 10 mV;
frequency: 25 Hz) were recorded using a potentiostat/galvanostat
(PGSTAT 30, Autolab, Eco-Chemie, The Netherlands), and ferrocene
was used as a calibration external standard. The thickness of TiO2
anode was determined by a surface profile meter from Veeco Co.
(Dektak 150), USA.
Semiempirical Computation. The DFT calculations were
performed to examine the electronic structure of the complexes with
the Gaussian 09 program. Ground-state optimization for the three
geometric isomers was performed with Becke, three-parameter, Lee−
Yang−Parr (B3LYP) exchange-correlations function with DGauss
Double-Zeta Valence Polarized (DGDZVP) basis set. The solvent
(DMF) effects were included in the calculations using the conductor-
like polarizable continuum model (C-PCM). The TD-DFT excited-
state calculation of the lowest 60 singlet−singlet electronic transitions
for the complexes was performed also at the B3LYP/DGDZVP level35
for the geometry optimization. For computing convenience, the
hexylthio-terminal chains in the molecules were modeled by
methylthio- substitutions. The overlap weighting factors were set as
a default value.36
ASSOCIATED CONTENT
■
S
* Supporting Information
The Supporting Information is available free of charge on the
ACS Publications website at DOI: 10.1021/acs.inorg-
chem.5b01967.
Synthesis of Ligand 29 and CYC-B29, tabulated optical
data of the three diastereoisomers CC, TT, and CT,
tabulated ground-state vertical excitation data for CC,
1
TT, and CT in DMF, H NMR spectra of CC, TT, and
CT, illustrated frontier orbitals of CC, TT, and CT, ESI
diagrams of devices A, B, and C sensitized with CC, TT,
and CT. (PDF)
AUTHOR INFORMATION
■
Corresponding Author
*E-mail: t610002@cc.ncu.edu.tw.
Notes
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
■
Device Fabrication and Photovoltaic Characterization.
Electrode for the photoanode is 7 Ω cm−2 fluorine-doped tin oxide
(FTO)-coated glass (Merck Glass). The FTO glass was cleaned by
detergent, acid, and sonicating in H2O and then stored in isopropanol
before used. A 12 μm (particle size: 20 nm) + 4 μm (particle size: 400
nm) thick TiO2 film was deposited on the conducting FTO with
screen printer. Detailed methods for TiO2 film preparation and device
Financial support from the Ministry of Science and Technology
(MOST), Taiwan, ROC (Grant No. NSC 101-2113-M-008-
008-MY3), is gratefully acknowledged. The device fabrications
and photovoltaic parameter measurements were performed in
Advanced Laboratory of Accommodation and Research for
Organic Photovoltaics, MOST, Taiwan, ROC.
F
DOI: 10.1021/acs.inorgchem.5b01967
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
(29) Hu, F.-C.; Wang, S.-W.; Chi, Y.; Robertson, N.; Hewat, T.; Hu,
Y.; Liu, S.-H.; Chou, P.-T.; Yang, P.-F.; Lin, H.-W. ChemPhysChem
2014, 15, 1207−1215.
REFERENCES
■
(1) O’Regan, B.; Gratzel, M. Nature 1991, 353, 737−740.
̈
(2) Gratzel, M. Acc. Chem. Res. 2009, 42, 1788−1798.
̈
(30) Chen, C.-Y.; Pootrakulchote, N.; Chen, M.-Y.; Moehl, T.; Tsai,
(3) Hagfeldt, A.; Boschloo, G.; Sun, L.; Kloo, L.; Pettersson, H.
H.-H.; Zakeeruddin, S. M.; Wu, C.-G.; Gratzel, M. Adv. Energy Mater.
̈
Chem. Rev. 2010, 110, 6595−6663.
2012, 2, 1503−1509 and unreported data..
(4) Gratzel, M. Inorg. Chem. 2005, 44, 6841−6851.
̈
(31) Li, C.; Wu, S.-J.; Wu, C. G. J. Mater. Chem. A 2014, 2, 17551−
(5) Chen, C. Y.; Wang, M.; Li, J. Y.; Pootrakulchote, N.; Alibabaei,
17560.
L.; Ngoc-le, C. H.; Decoppet, J. D.; Tsai, J. H.; Gratzel, C.; Wu, C. G.;
Zakeeruddin, S. M.; Gratzel, M. ACS Nano 2009, 3, 3103−3109.
(6) Sauvage, F.; Chen, D.; Comte, P.; Huang, F.; Heiniger, L.-P.;
Cheng, Y.-B.; Caruso, R. A.; Gratzel, M. ACS Nano 2010, 4, 4420−
4425.
(7) Nazeeruddin, M. K.; Kay, A.; Rodicio, L.; Humphry-Baker, R.;
Miiller, E.; Liska, P.; Vlachopoulos, N.; Gratzel, M. J. Am. Chem. Soc.
1993, 115, 6382−6390.
(8) Nguyen, P. T.; Degn, R.; Nguyen, H. T.; Lund, T. Sol. Energy
Mater. Sol. Cells 2009, 93, 1939−1945.
(9) Asghar, M. I.; Miettunen, K.; Halme, J.; Vahermaa, P.; Toivola,
M.; Aitola, K.; Lund, P. Energy Environ. Sci. 2010, 3, 418−426.
(10) Islam, A.; Chowdhury, F. A.; Chiba, Y.; Komiya, R.; Fuke, N.;
Ikeda, N.; Nozaki, K.; Han, L. Chem. Mater. 2006, 18, 5178−5185.
(11) Funaki, T.; Yanagida, M.; Onozawa-Komatsuzaki, N.; Kasuga,
K.; Kawanishi, Y.; Sugihara, H. Chem. Lett. 2009, 38, 62−63.
(12) Wadman, S. H.; Kroon, J. M.; Bakker, K.; Lutz, M.; Spek, A. L.;
van Klink, G. P. M.; van Koten, G. Chem. Commun. 2007, 1907−1909.
(13) Wadman, S. H.; Kroon, J. M.; Bakker, K.; Havenith, R. W. A.;
van Klink, G. P. M.; van Koten, G. Organometallics 2010, 29, 1569−
1579.
̈
(32) O’Regan, B.; Walley, K.; Juozapavicius, M.; Anderson, A.; Matar,
F.; Ghaddar, T.; Zakeeruddin, S. M.; Klein, C.; Durrant, J. R. J. Am.
Chem. Soc. 2009, 131, 3541−3548.
̈
̈
(33) Zhang, Z.; Zakeeruddin, S. M.; O’Regan, B.; Humphry-Baker,
R.; Gratzel, M. J. Phys. Chem. B 2005, 109, 21818−21824.
̈
́
(34) Reynal, A.; Forneli, A.; Martinez-Ferrero, E.; Sanchez-Díaz, A.;
̈
Vidal-Ferran, A.; O’Regan, B.; Palomares, E. J. Am. Chem. Soc. 2008,
130, 13558−13567.
(35) Hirata, N.; Lagref, J. J.; Palomares, E. J.; Durrant, J. R.;
Nazeeruddin, M. K.; Gratzel, M.; Di Censo, D. D. Chem. - Eur. J. 2004,
̈
10, 595−602.
(36) Chen, C. Y.; Lu, H. C.; Wu, C. G.; Chen, J. G.; Ho, K. C. Adv.
Funct. Mater. 2007, 17, 29−36.
(37) Chen, C. Y.; Chen, J. G.; Wu, S. J.; Li, J. Y.; Wu, C. G.; Ho, K. C.
Angew. Chem., Int. Ed. 2008, 47, 7342−7345.
(38) Delley, B. J. Chem. Phys. 2000, 113, 7756−7764.
(39) Fabregat-Santiago, F.; Garcia-Belmonte, G.; Bisquert, J.; Zaban,
A.; Salvador, P. J. Phys. Chem. B 2002, 106, 334−339.
(40) Mozer, A. J.; Panda, D. K.; Gambhir, S.; Winther-Jensen, B.;
Wallace, G. G. J. Am. Chem. Soc. 2010, 132, 9543−9545.
(14) Wadman, S. H.; Lutz, M.; Tooke, D. M.; Spek, A. L.; Hartl, F.;
Havenith, R. W. A.; van Klink, G. P. M.; van Koten, G. Inorg. Chem.
2009, 48, 1887−1900.
(15) Bomben, P. G.; Robson, K. C. D.; Sedach, P. A.; Berlinguette, C.
P. Inorg. Chem. 2009, 48, 9631−9643.
(16) Koivisto, B. D.; Robson, K. C. D.; Berlinguette, C. P. Inorg.
Chem. 2009, 48, 9644−9652.
(17) Robson, K. C. D.; Koivisto, B. D.; Yella, A.; Sporinova, B.;
Nazeeruddin, M. K.; Baumgartner, T.; Gratzel, M.; Berlinguette, C. P.
̈
Inorg. Chem. 2011, 50, 5494−5508.
(18) Bessho, T.; Yoneda, E.; Yum, J.-H.; Guglielmi, M.; Tavernelli, I.;
Imai, H.; Rothlisberger, U.; Nazeeruddin, M. K.; Gratzel, M. J. Am.
̈
Chem. Soc. 2009, 131, 5930−5934.
(19) Wu, K.-L.; Hsu, H.-C.; Chen, K.; Chi, Y.; Chung, M.-W.; Liu,
W.-H.; Chou, P.-T. Chem. Commun. 2010, 46, 5124−5126.
(20) Bomben, P. G.; Gordon, T. J.; Schott, E.; Berlinguette, C. P.
Angew. Chem., Int. Ed. 2011, 50, 10682−10685.
(21) Bomben, P. G.; Theriault, K. D.; Berlinguette, C. P. Eur. J. Inorg.
Chem. 2011, 2011, 1806−1814.
(22) Dragonetti, C.; Valore, A.; Colombo, A.; Roberto, D.; Trifiletti,
V.; Manfredi, N.; Salamone, M. M.; Ruffo, R.; Abbotto, A. J.
Organomet. Chem. 2012, 714, 88−93.
(23) Colombo, A.; Dragonetti, C.; Magni, M.; Meroni, D.; Ugo, R.;
Marotta, G.; Lobello, M. G.; Salvatori, P.; De Angelis, F. Dalton Trans.
2015, 44, 11788−11796.
(24) Dragonetti, C.; Colombo, A.; Magni, M.; Mussini, P.; Nisic, F.;
Roberto, D.; Ugo, R.; Valore, A.; Valsecchi, A.; Salvatori, P.; Lobello,
M. G.; De Angelis, F. Inorg. Chem. 2013, 52, 10723−10725.
(25) Wu, K. L.; Li, C. H.; Chi, Y.; Clifford, J. N.; Cabau, L.;
Palomares, E.; Cheng, Y. M.; Pan, H. A.; Chou, P. T. J. Am. Chem. Soc.
2012, 134, 7488−7496.
(26) Wu, K. L.; Ku, W. P.; Wang, S. W.; Yella, A.; Chi, Y.; Liu, S. H.;
Chou, P. T.; Nazeeruddin, M. K.; Gratzel, M. Adv. Funct. Mater. 2013,
̈
23, 2285−2294.
(27) Wang, S.-W.; Wu, K.-L.; Ghadiri, E.; Lobello, M. G.; Ho, S.-T.;
Chi, Y.; Moser, J.-E.; De Angelis, F.; Gratzel, M.; Nazeeruddin, M. K.
̈
Chem. Sci. 2013, 4, 2423−2433.
(28) Wu, K.-L.; Ku, W.-P.; Clifford, J. N.; Palomares, E.; Ho, S.-T.;
Chi, Y.; Liu, S.-H.; Chou, P.-T.; Nazeeruddin, M. K.; Gratzel, M.
̈
Energy Environ. Sci. 2013, 6, 859−870.
G
DOI: 10.1021/acs.inorgchem.5b01967
Inorg. Chem. XXXX, XXX, XXX−XXX