Molecularly engineered Ru(II) sensitizers compatible with cobalt(II/III) redox mediators for dye-sensitized solar cells

Article abstract of DOI:10.1021/acs.inorgchem.6b00427

Thiocyanate-free isoquinazolylpyrazolate Ru(II) complexes were synthesized and applied as sensitizers in dye-sensitized solar cells (DSCs). Unlike most other successful Ru sensitizers, Co-based electrolytes were used, and resulting record efficiency of 9.53% was obtained under simulated sunlight with an intensity of 100 mW cm^-2. Specifically, dye 51-57dht.1 and an electrolyte based on Co(phen)₃ led to measurement of a J_SC of 13.89 mA cm^-2, V_OC of 900 mV, and FF of 0.762 to yield 9.53% efficiency. The improved device performances were achieved by the inclusion of 2-hexylthiophene units onto the isoquinoline subunits, in addition to lengthening the perfluoroalkyl chain on the pyrazolate chelating group, which worked to increase light absorption and decrease recombination effects when using the Co-based electrolyte. As this study shows, Ru(II) sensitizers bearing sterically demanding ligands can allow successful utilization of important Co electrolytes and high performance.

Full text of DOI:10.1021/acs.inorgchem.6b00427

Article
pubs.acs.org/IC
Molecularly Engineered Ru(II) Sensitizers Compatible with Cobalt(II/
III) Redox Mediators for Dye-Sensitized Solar Cells
Kuan-Lin Wu,†,‡ Aron J. Huckaba,§ John N. Clifford,∥ Ya-Wen Yang,† Aswani Yella,‡ Emilio Palomares,∥,⊥
Michael Gratzel,‡ Yun Chi,*,† and Mohammad Khaja Nazeeruddin*,§
̈
†Department of Chemistry and Low Carbon Energy Research Center, National Tsing Hua University, Hsinchu 30013, Taiwan
‡Institute of Chemical Sciences and Engineering, Laboratory of Photonics and Interfaces, Ecole Polytechnique Fed
́ ́
erale de Lausanne,
1015 Lausanne, Switzerland
§Group for Molecular Engineering of Functional Materials, Ecole Polytechnique Federale de Lausanne Valais Wallis, Rue de l’Indutrie
17, 1950 Sion, Valais, Switzerland
∥Institute of Chemical Research of Catalonia (ICIQ), The Barcelona Institute of Science and Technology, Avda. Països Catalans 16,
43007 Tarragona, Spain
⊥ICREA, Avda. Lluis Companys, 23, E-08010 Barcelona, Spain
S
* Supporting Information
ABSTRACT: Thiocyanate-free isoquinazolylpyrazolate Ru-
(II) complexes were synthesized and applied as sensitizers in
dye-sensitized solar cells (DSCs). Unlike most other successful
Ru sensitizers, Co-based electrolytes were used, and resulting
record efficiency of 9.53% was obtained under simulated
sunlight with an intensity of 100 mW cm−2. Specifically, dye
51-57dht.1 and an electrolyte based on Co(phen)3 led to
measurement of a JSC of 13.89 mA cm−2, VOC of 900 mV, and
FF of 0.762 to yield 9.53% efficiency. The improved device
performances were achieved by the inclusion of 2-hexylth-
iophene units onto the isoquinoline subunits, in addition to
lengthening the perfluoroalkyl chain on the pyrazolate
chelating group, which worked to increase light absorption and decrease recombination effects when using the Co-based
electrolyte. As this study shows, Ru(II) sensitizers bearing sterically demanding ligands can allow successful utilization of
important Co electrolytes and high performance.
electrons in the TiO2 under illumination and the Nernst
potential of the redox couple in the electrolyte. A redox shuttle
with a larger positive magnitude (more stabilized) redox
potential provides higher VOC in devices. Upon stabilization of
the redox shuttle oxidation potential, though, the driving force
for dye regeneration (ΔEregen) is diminished. This leads to
slower dye regeneration, which diminishes both VOC and JSC
through recombination pathways as well.10 Thus, fine-tuning of
the electrolyte composition is a critical issue for increasing the
performance of DSCs.11,12
INTRODUCTION
■
In a matter of hours, the sun bombards our planet with enough
energy to meet annual human consumption. Thus, photovoltaic
technology could prove a much-needed answer to the world’s
energy needs. Specifically, advances in dye-sensitized solar cells
(DSCs) have seen top all-organic, organometallic, and quantum
dot sensitizer efficiencies leap from 7.9%1 in 1991 to >13.0%
currently.2−6 If perovskite materials are also included as
inorganic sensitizers, efficiencies have skyrocketed to 20.2%
and are poised to leap higher.7
Even though thiocyanate-ligated Ru sensitizers yield stable
devices utilizing the iodide/triiodide redox couple system,
performance is limited by the relatively high overpotential for
dye regeneration associated with this couple.13 Furthermore,
the inclusion of thiocyanate ligands results in destabilization of
oxidized sensitizers through the monodentate SCN ligand as
well as tedious isomer separation of N- and S-bound linkage
isomers.14,15 Thus, the replacement of singly bound thiocyanate
While Ru sensitizers were the first to be employed in DSCs
and several sensitizers have yielded devices with >10%
efficiency, in general the champion dyes have suffered from
comparatively low short-circuit current (JSC) values. Even
though these sensitizers absorb strongly from 400 to 800 nm,
charge recombination from the semiconductor to the sensitizer
or redox shuttle means the overall current is diminished. Some
high (>20 mA/cm2) JSC sensitizers have been reported;
however, in general they suffer from low voltage at open
circuit (VOC) compared to other sensitizer classes.8,9 The DSC
VOC corresponds to the difference of the quasi-Fermi level of
Received: February 20, 2016
© XXXX American Chemical Society
A
DOI: 10.1021/acs.inorgchem.6b00427
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
Scheme 1. Structures of Ru Sensitizers Used in This Study, with Important Structural Differences Marked in Red, Blue, or
Green
with higher chelation number ligands, which are better able to
reduce recombinative and other nonproductive electron-
transfer pathways, is advantageous in designing high efficiency
examining the steric properties of triphenylamine-based organic
dyes in the presence of cobalt redox mediators showed a
decrease in interfacial charge recombination between TiO2 and
electrolytes with increasing donor sterics.21 In 2011, a power
conversion efficiency (PCE) of 9.6% was achieved with organic
sensitizer Y123 and Co(II/III)(bipyridine)3,22 and then Yella et
al. boosted the PCE to 12.3% by using a Co(II/III)(bipyridine)3-
based electrolyte in conjunction with a zinc porphyrin sensitizer
(YD2-o-C8).26 The gain in the device performance was largely
due to the molecular design of YD2-o-C8, which included
many nonconjugated substituents meant to block the semi-
conductor surface from the redox shuttle and led to an increase
sensitizers.16−18 In assessing whether to aim for CoIII- or I−/I3 -
−
based electrolyte systems, there is another important issue to
consider: The large mismatch between its oxidation potential
and the dye oxidation potential limits the open-circuit voltage
(VOC) to 0.7−0.8 V. Furthermore, the reduction of I2 and I3− is
a complex multistep reaction, each with a performance-
diminishing overpotential.19,20 Boschloo et al. reported that
the regeneration of oxidized dye with iodide leads to the
•−
formation of the diiodide radical (I2•−−) and then I2
disproportionates to I− and I3 (E° (I−/I3 ) = ca. 0.29 V and
−
in VOC
.
E° (I2•−/I−) < 0.93 V vs NHE in acetonitrile), giving a large
loss in potential energy.13 Therefore, the development of stable,
alternative redox couples with more positive (stabilized)
oxidation potentials can avoid the problems associated with
Using Ru(II) sensitizers N719 and Z907, and the
Co(II/III)(bpy)3 (bpy = bipyridine) redox shuttle, lower PCE
and JSC values were observed due to decreased charge collection
and increased charge recombination. Despite this drawback, a
PCE of up to 6.5% was achieved using Z907 with cobalt redox
mediators,25 while, with other sensitizer classes, one way to
improve the PCEs is to increase the bulkiness of the Ru(II)
sensitizer for blocking the approach of redox electrolyte to the
TiO2 surface.27−29 Here, three thiocyanate-free Ru(II)
sensitizers, 51-5ht, 51-57dht, and 51-57dht.1 (Scheme 1),
are compared to investigate the effect of steric hindrance,
recombination driving force, and dye regeneration kinetics on
performance in devices employing cobalt-based electrolytes.
the aforementioned I−/I3 system.21,22
−
Because of low molar extinction coefficients and tunable
redox potentials controlled by their substituents, cobalt
complexes are interesting alternative redox mediators for dye-
sensitized solar cells. Furthermore, because the redox shuttle
involves outer sphere single electron transfer (Co(II)/(III)), the
kinetics of dye regeneration are simplified greatly. However,
there are several drawbacks, such as slower diffusion through
the mesoporous TiO2 film to the counter electrode due to the
larger ionic radius, effective mass, and diffusion coefficient
−
compared to the I−/I3 electrolyte. This limitation enhances
EXPERIMENTAL SECTION
■
the electron recombination from the semiconductor to the
oxidized Co(III) species and restricts achievable photocurrent
densities at full sun intensity.23 Lastly, unlike the negatively
General Procedures. All reactions were performed under argon
atmosphere, and solvents were distilled from appropriate drying agents
prior to use. Commercially available reagents were used without
further purification unless otherwise stated. [Ru(dcbpy)(p-cymene)-
Cl]Cl was obtained from treatment of [Ru(p-cymene)Cl2]2 with
dcbpy (4,4′-bis(ethoxycarbonyl)-2,2′-bipyridine) in refluxing methanol
solution. 2-(5-Hexylthiophen-2-yl)-4,4,5,5,-tetramethyl-1,3,2-dioxabor-
olane was prepared from 2-hexylthiophene and 2-isopropoxy-4,4,5,5-
tetramethyl-1,3,2-dioxaborolane as documented in the literature.30 5-
Bromoisoquinoline was prepared by monobromination of isoquinoline
in concentrated H2SO4 using N-bromosuccinimide (NBS),31 while
5,7-dichloroisoquinoline was synthesized from the respective halo-
genated benzalaminoacetal reagents according to the Pomeranz−
Fritsch reaction.32,33 Acetylation of all isoquinolines was then
conducted using para-acetaldehyde, t-BuO2H, trifluoroacetic acid,
and FeSO4·7H2O.34 Next, all halogen substituents on isoquinoline
framework were replaced by a 2-hexylthienyl group (ht) using
dioxaborolane reagent with Suzuki coupling. Finally, conversions to
the isoquinolinyl pyrazoles were achieved using standard Claisen
condensation with ethyl trifluoroacetate or ethyl heptafluorobutanoate,
−
charged I−/I3 redox shuttle, both Co(III) and Co(II) species
are strongly attracted to the n-type TiO2 surface (approximately
−0.5 V vs normal hydrogen electrode, (NHE)), which also
exacerbates recombination problems.
In 2001, the cobalt-based mediator, bis[2,6-bis(1′-butylben-
zimidazol-2′-yl)pyridine] cobalt(II) complex, was found to rival
the I−/I3 system in terms of dye regeneration. However,
−
serious diffusion problems limited performance under full
sunlight (1 sun).24 Later it was reported that cobalt
polypyridine complexes formed from structurally simple ligands
functioned as efficient electron-transfer mediators in DSCs and
concluded that added Li+ resulted in reduced recombination
between the electrons in the TiO2 conduction band and the
Co(III) species. This result contrasts with the observations
made with I−/I3 cells.25 A classic study in organic sensitizers
−
B
DOI: 10.1021/acs.inorgchem.6b00427
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
followed by hydrazine cyclization in refluxing ethanol solution (yield
>50%).35 The detail ligand synthetic procedures are reported in
Supporting Information. The syntheses of 51-5ht, 51-57dht, and 51-
57dht.1 sensitizers are similar, using similar experimental proce-
dures.28,36,37 All reactions were monitored using precoated TLC plates
(0.20 mm with fluorescent indicator UV254). Mass spectra were
obtained on a JEOL SX-102A instrument operating in electron impact
(s, 4F). Anal. Calcd for C82H78F14N8O4S4Ru·2H2O: C, 55.61; N, 6.33
H, 4.67. Found: C, 55.24; N, 6.49; H, 4.89.
Device Fabrication and Photovoltaic Characterization. The
cells consisted of a mesoscopic TiO2 film composed of a 4.1 μm thick
transparent layer of 20 nm sized TiO2 anatase nanoparticles (Ti-2105,
ETERDSC, 15−40 nm) onto which a second 4.5 μm thick scattering
layer of 400 nm sized TiO2 (CCIC) was superimposed. The double-
layer film was heated to 520 °C and sintered for 30 min, then cooled
to 80 °C and immersed into the dye solution (0.3 mM) containing the
20% DMSO in ethanol for 12 h. The iodine electrolyte (W95)
contains 0.45 M 1-methyl-3-propylimidazolium iodide (PMII), 0.15 M
I2, 0.15 M LiI, and 0.8 M tert-butylpyridine (TBP) in acetonitrile; the
cobalt electrolyte (W187) contains 0.6 M [Coll(bpy)3][bis-
(trifluoromethane)sulfonimide]2, 0.15 M [Colll(bpy)3][TFSI]3, 0.15
M LiTFSI, and 0.8 M TBP in acetonitrile [TFSI = bis-
(trifluoromethane)sulfonimide], and the cobalt electrolyte (W178)
contains 0.6 M [CoII(phenanthroline)3][TFSI]2 (phen = phenanthro-
line), 0.15 M [CoIII(bpy)3][TFSI]3, 0.15 M LiTFSI, and 0.8 M TBP
in acetonitrile. The cell was sealed with 15 μm thick transparent Surlyn
ring at 130 °C for 15 s to the counter electrode (fluorine-doped tin
oxide (FTO) glass, 7 Ω per square, thermal deposited carbon at 450
°C for 15 min). The cells were filled with an electrolyte solution
through a predrilled hole in the counter electrode. The hole was then
sealed with a Bynel disk and a thin glass to avoid leakage of the
electrolyte. The current−voltage characteristics, incident photon to
current efficiency, and transient photocurrent/photovoltage measure-
ments were performed as reported in our previous publications.28,36
Basic Characterization. Photophysical and Electrochemical
Properties. Figure 1 shows absorption properties of Ru-pyrazolate
1
(EI) or fast atom bombardment (FAB) mode. H NMR spectra were
recorded on a Varian Mercury-400 instrument. Elemental analysis of
the Ru(II) sensitizers was carried out with a Heraeus CHN-O Rapid
Elementary Analyzer.
Synthesis of 51-5ht Sensitizer. A mixture of [Ru(dcbpy)(p-
cymene)Cl]Cl (65 mg, 0.14 mmol), (L5ht)H (91 mg, 0.21 mmol, see
Scheme 2), and potassium acetate (50 mg, 0.51 mmol) in 30 mL of
Scheme 2. Structures of Ligands Used To Synthesize Ru
Sensitizers
xylene was refluxed for 5 h. After the solvent was removed, the residue
was extracted with CH2Cl2 (3 × 25 mL), washed with water, and
concentrated to dryness. The mixture of products was purified by silica
gel column chromatography (ethyl acetate/hexane = 1:3) to afford
black solid (42 mg, 20%).
Next, the above-mentioned solid (42 mg, 0.043 mmol) was
dissolved in a mixture of acetone (20 mL) and 1 M NaOH solution
(0.2 mL). The solution was heated to 60 °C under nitrogen for 3 h.
After then, the solvent was removed under vacuum, and the residue
was dissolved in H2O solution (10 mL), followed by acidification with
2 M HCl to pH 3 for inducing a dark brown precipitate. The collected
precipitate was rinsed with deionized water, acetone, and diethyl ether
in sequence, giving a dark brown solid (30 mg, 78%).
Other Ru(II) sensitizers, namely, 51-57dht and 51-57dht.1, were
prepared from Ru(II) source complex [Ru(dcbpy)(p-cymene)Cl]Cl
and respective free ligand, i.e., (L57dht)H and (L57dht.1)H, using
similar experimental procedures.
Spectral data of 51-5ht follow. MS (FAB, 102Ru): m/z 1202 (M +
1)+. 1H NMR (400 MHz, d6-DMSO, 298 K): δ 8.90 (s, 4H), 8.10 (d, J
= 6.0 Hz, 2H), 7.82−7.78 (m, 6H), 7.70−7.67 (m, 4H), 7.12 (d, J =
6.8 Hz, 2H), 7.06 (d, J = 3.2 Hz, 2H), 6.84 (d, J = 3.6 Hz, 2H), 2.76−
2.68 (m, 4H), 1.50−1.47 (m, 4H), 1.21−1.14 (m, 12H), 0.79 (t, J =
6.7 Hz, 6H). 19F NMR (376 MHz, d6-DMSO, 298 K): δ −58.16 (s,
6F). Anal. Calcd for C58H50F6N8O4S2Ru·H2O: C, 57.9; N, 9.34; H,
4.30. Found: C, 57.94; N, 9.32; H, 4.19.
Figure 1. UV−vis absorption spectra of Ru sensitizers (1 × 10−5 M) in
DMF.
sensitizers 51-5ht, 51-57dht, and 51-57dht.1 measured in dimethyl-
formamide (DMF), and the spectral data are summarized in Table 1.
The absorption spectrum of the parent complex 51-5ht exhibits two
peak maxima in the UV region at 306 and 354 nm, which are assigned
to π−π* transitions, and the absorption peak at 510 nm is due to
metal-to-ligand charge-transfer bands (MLCT) mixed with small
amount of ligand-to-ligand charge-transfer contribution (vide infra).
Upon introduction of 4-hexylthiophene substituents to 51-5ht in the
7-position to form 51-57dht, the lower energy MLCT absorption
undergoes a slight red shift to 521 nm with increased molar
absorptivity (ε). Also, the higher energy band exhibited a red shift
to 310 and 378 nm with a significantly enhanced molar extinction
coefficient due to extended conjugation of ligands. In addition, the
introduction of a heptafluoropropyl group did not alter the absorption
spectrum for 51-57dht.1.
Spectral data of 51-57dht follow. MS (FAB, 102Ru): m/z 1534 (M +
1
1)+. H NMR (400 MHz, d6-DMSO, 298 K): δ 8.94 (s, 2H), 8.73 (s,
2H), 8.16 (d, J = 5.6 Hz, 2H), 7.92 (s, 2H), 7.72 (d, J = 5.6 Hz, 2H),
7.61 (d, J = 6.4 Hz, 2H), 7.52−7.49 (m, 4H), 7.15 (d, J = 6.0 Hz, 2H),
7.01 (s, 2H), 6.85 (d, J = 13.6 Hz, 4H), 2.82−2.74 (m, 8H), 1.64−1.55
(m, 8H), 1.36−1.20 (m, 24H), 0.85−0.80 (m, 12H). 19F NMR (376
MHz, d6-DMSO, 298 K): δ −58.71 (s, 6F). Anal. Calcd for
C78H78F6N8O4S4Ru·H2O: C, 60.33; N, 7.22; H, 5.19. Found: C,
60.15; N, 7.13; H, 4.97.
Spectral data of 51-57dht.1 follow. MS (FAB, 102Ru): m/z 1735
(M + 1)+. 1H NMR (400 MHz, d6-DMSO, 298 K): 8.82 (s, 2H), 8.66
(s, 2H), 8.00 (d, J = 4.4 Hz, 2H), 7.81 (s, 2H), 7.62 (d, J = 6.0 Hz,
2H), 7.47−7.44 (m, 4H), 7.37 (s, 2H), 7.13 (d, J = 6.0 Hz, 2H), 6.95
(s, 2H), 6.74 (d, J = 6.8 Hz, 4H), 2.74−2.69 (m, 8H), 1.56−1.53 (m,
8H), 1.28−1.07 (m, 24H), 0.78−0.69 (m, 12H). 19F NMR (376 MHz,
d8-THF, 298 K): −79.86 (t, J = 8.3 Hz, 6 F), −107.20 (s, 4F), −126.78
Cyclic voltammetry was used to measure the oxidation potential of
the sensitizers (ES+/S), to see compatibility with the redox potential of
C
DOI: 10.1021/acs.inorgchem.6b00427
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
Table 1. Photophysical and Electrochemical Data of the
Studied Sensitizers in DMF
λabs [nm] (ε × 10−3
a
b
c
d
dye
[L mol−1 cm−1])
E°′ox
E0−0
E°′*
51-5ht
306 (44), 354 (39), 510 (23)
310(65), 378 (44), 521 (28)
309 (64), 379 (43), 518 (27)
0.90
0.95
0.93
1.92
1.86
1.86
−1.02
−0.91
−0.93
51-57dht
51-57dht.1
a
Molar extinction coefficients data were measured in DMF solution.
Oxidation potentials of sensitizers were measured in DMF with 0.1 M
b
[TBA][PF6] supporting electrolyte. The scan rate was 50 mV s−1, and
the potentials were calibrated using Fc/Fc+ as internal reference. The
c
values were converted to NHE by addition of 0.63 V. E0−0 was
determined from the intersection of the absorption and the tangent of
d
emission peak in DMF. E°′* was calculated as E°′ox − E0−0
.
the cobalt electrolyte. The redox potential data of sensitizers are
summarized in Table 1, which are more positive (more stabilized
HOMO) compared to the redox potential of [Co(bpy)3]2+/3+ (ca. 0.56
V vs NHE) and [Co(phen)3]2+/3+ (ca. 0.63 V vs NHE), warranting
enough (ΔEregen > 150 mV) driving force for dye regeneration. The
0−0 transition energy (E0−0) was determined from the intersection of
the absorption and the tangent of the emission peak in DMF, for
which their energy gaps are identical due to the similar core structure.
The excited-state oxidation potential (ES+/S*) values were estimated
from the difference in the ES+/S and E0−0 of 51-5ht, 51-57dht, and 51-
57dht.1 and are −1.02, −0.91, and −0.93 V, respectively, which are
more negative than the conduction band edge of the TiO2 electrode
(ca. −0.5 V vs NHE), confirming that the ES+/S* was well-matched
with the conduction band value of TiO2.
Device Performance. Several cobalt-specific device-engineering
techniques are required to suppress the mass transport limitations
present in those electrolyte systems, for example, using low viscosity
solvent, controlling the TiO2 film pore size, and minimizing the space
between TiO2 electrode and counter electrode. These are expected to
improve the mass transport and overall efficiency.
Figure 2. Electrochemical impedance spectra measured under dark
conditions at a forward bias of 0.85 V for the cells employing different
counter electrodes.
performance was observed with a TBP concentration of 0.8 M
(Figure 3, right).
Optimization of TiCl4 Concentration and TiO2 Film Thick-
ness. By tuning the pore size and porosity of TiO2 films, mass
transport limitations can be addressed in devices using cobalt-based
electrolytes. Post-treatment of TiO2 with TiCl4 not only enhances the
connections between TiO2 nanoparticles but also increases pore size
and porosity in the iodine/triiodide electrolyte system. Here, we tried
to control the pore size and porosity by simply optimizing the TiCl4
post-treatment.38 Device performances were obtained with the thin
(4.1 + 4.5 μm) double-layered TiO2 film consisting of an average
particle size of 15−40 nm. After film fabrication, different TiCl4
treatments of 15, 25, and 40 mM at 70 °C were administered for 30
min. The best device performance was observed at a concentration of
15 mM (Figure 4, left). Even though the performance of the devices
treated with 25 and 40 mM were similar, the mass transport limitations
were increased in the 40 mM case, as shown by the low VOC and FF.
The diffusion limitation problems in cobalt-based electrolyte can be
further investigated by recording photocurrent transients under
different light intensities (Figure 4, right). The experiment shows an
immediate rise of the photocurrent when the device is exposed to light,
with the magnitude of the current density directly proportional to the
light intensity. As illustrated by the two graphs in Figure 4, increasing
the TiCl4 concentration more than 15 mM lowered both the porosity
and pore size of the TiO2 photoanode, which suppressed Co(III)
diffusion and resulted in poor JSC and current−density profiles.
It is well-known that increasing the TiO2 film thickness can enhance
the photocurrent due to enhanced light absorption, but decrease the
VOC due to enhanced recombination and charge transport distance.
Because each sensitizer exhibits a different molar absorptivity profile,
film thickness should be optimized in each case. We attempted to
optimize film thickness, and the photocurrent transients with the
different thicknesses are shown in Figure 5. Unfortunately, serious
diffusion problems were observed even under low light intensity with a
thicker (7.9 μm) film. The photovoltaic performance of devices with
different light intensity is also described in Table 2. Due to diffusion
limitation problems accompanied by thicker TiO2 films, the photo-
current dropped significantly after light exposure.
Although platinum counter electrodes are generally employed for
−
the I−/I3 electrolyte, charge-transfer resistance in these devices is
high, which can lead to losses in the fill factor. In 2002, Sapp et al.
observed that gold and carbon are superior cathode materials to
platinum by cyclic voltammetry, especially for Ru sensitizers with
sterically demanding ligands.25 Here, we evaluate the performance of
devices modified by using either traditional thermally deposited
platinum or carbon counter electrodes. Electron impedance spectros-
copy (EIS) was used to investigate the properties and quality of the
DSC device. The EIS spectrum of devices with Pt and carbon counter
electrodes is presented as a Nyquist plot (Figure 2). The spectrum
exhibits a first semicircle corresponding to the charge-transfer
resistance (RCE) at the counter electrode/electrolyte interface. The
RCE is the resistance of regenerating Co(II) from Co(III) at the
counter electrode. The charge-transfer resistance was found to be 6.44
Ω (12.88 Ω cm−2) for Pt and 1.46 Ω (2.92 Ω cm−2) for the carbon
counter electrode, meaning that the resistance at the carbon counter
electrode is significantly lower than that of the Pt counter electrode.
Optimization of Co(II) and t-Butylpyridine (TBP) Concen-
trations in Electrolyte. The electron-transfer dynamics of a DSC
dictates that, after electron injection, the photo-oxidized dye should be
predominantly reduced by the Co(II) redox mediator (eq 1).
S+(on TiO2) + Co(II) → S(on TiO2) + Co(III)
(1)
Thus, the concentration of the Co(II) complex affects the dye
regeneration rate. Optimization of this key electrolyte component was
undertaken (Figure 3, left), and the best results were observed when
using the Co(II)/Co(III) ratio of 4:1 (i.e., 0.6 M Co(II)) with 0.1 M
LiTFSI and 1.6 M TBP as additives. Another additive, t-butylpyridine
(TBP), absorbs onto the TiO2 surface and decreases the charge
recombination in the TiO2/Co(III) interface. Thus, we also optimized
the concentration of TBP in the electrolyte. The best device
RESULTS AND DISCUSSION
■
Effect of Sensitizer Sterics and Driving Force for
Recombination on Photovoltaic Performance. We
examined the ruthenium sensitizers using Co(bpy)3 and
Co(phen)3 complexes as redox mediators to investigate the
influence of redox mediator energy levels on power conversion
D
DOI: 10.1021/acs.inorgchem.6b00427
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
Figure 3. Left: Optimization of Co(II) in the electrolyte. Right: Optimization of TBP in the electrolyte.
Figure 4. Left: Optimization of the concentration of TiCl4 during post-treatment. Right: Photocurrent transients of devices during TiCl4
optimization. The devices were measured using different incident light intensities, and the maximum intensity used was 100 mW cm−2.
Table 2. Performance of 7.9 + 4.5 μm Thick TiO2 Film DSC
Sensitized with 51-57dht.1 Employing [Co(phen)3]+2/+3
a
Electrolyte
light
dye
intensity [%] JSC [mA cm−2
]
VOC [mV]
FF
η [%]
51-
57dht.1
10
1.46
799
0.802
9.37
50
60
6.73
6.96
8.89
863
870
884
0.810
0.883
0.870
9.41
8.92
6.85
100
a
Devices were probed at various incident light intensities up to AM1.5
100 mW cm−2. All of the devices were fabricated using a 7.9 + 4.5 μm
TiO2 anode with an electrolyte that consists of 0.6 M Co electrolyte
(Co(II)/Co(III) ratio of 4:1), 0.1 M LiTFSI, and 0.8 M TBP in
acetonitrile.
couple (see Table 3 for redox potential). Triphenylamine-based
donor−acceptor organic sensitizer, D35, has been shown to
minimize the recombination between electrons in the TiO2 and
the oxidized redox mediator in the electrolyte Co(III) due to
increasing number of substituents on the donor moiety. Under
full sunlight irradiation (AM1.5G, 100 mW cm−2), a device
sensitized with 51-57dht.1 and incorporating a [Co-
(phen)3]2+/3+ redox mediator exhibited the best performance
with the short-circuit photocurrent density (JSC) of 13.89 mA
Figure 5. Optimization of TiO2 film thickness by comparison of
photocurrent transient of devices sensitized with 51-57dht.1 employ-
ing [Co(phen)3]+2/+3 electrolyte. The devices were probed at various
incident light intensities up to 100 mW cm‑2.
efficiency of the dye-sensitized solar cells. The device current−
voltage characteristics demonstrate the VOC increased with
increasing (more stabilized) Nernst potential of the redox
E
DOI: 10.1021/acs.inorgchem.6b00427
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Table 3. Performance of Ru-Sensitized DSCs Employing Electrolytes Based on I−/I3 and CoII/Colll
Article
a
−
d
dye
redox mediator
JSC [mA cm−2
]
VOC [mV]
FF
η [%]
dye loading [×10−8 mol cm−2
]
b
2+/3+
51-5ht
Co(bpy)3
12.78
12.17
15.31
13.56
12.32
14.17
13.36
13.89
13.53
840
842
718
844
898
727
853
900
0.764
0.750
0.746
0.742
0.754
0.743
0.750
0.762
0.749
8.22
7.69
8.20
8.49
8.34
7.66
8.55
9.53
7.50
6.4
b
b
b
2+/3+
Co(phen)3
c
I−/I3
−
b
2+/3+
51-57dht
Co(bpy)3
4.3
3.7
2+/3+
Co(phen)3
c
I−/I3
−
b
2+/3+
51-57dht.1
Co(bpy)3
2+/3+
Co(phen)3
c
I−/I3
−
740
a
b
All of the devices were fabricated using a 4.1 + 4.5 μm TiO2 anode. Devices fabricated using an electrolyte that consists of 0.6 M Co electrolyte
c
(Co(II)/Co(III) ratio of 4:1), 0.15 M LiTFSI, and 0.8 M TBP in acetonitrile. Devices fabricated using an electrolyte that consists of 0.45 M PMII,
0.15 M I2, 0.15 M LiI, and 0.8 M TBP in acetonitrile. The dye loading on 4.1 + 4.5 μm TiO2 films was determined by desorbing the dye into a 0.5
d
M TBAOH solution in 1:1 (v/v) MeOH/H2O and then performing the UV−vis spectral analysis.
cm−2, open-circuit voltage (VOC) of 900 mV, and fill factor
(FF) of 0.762, corresponding to a power conversion efficiency
(PCE, η) of 9.53%. Cells using the phenanthroline electrolyte
exhibited better VOC than iodine/triiodide (160 mV) and
[Co(bpy)3]2+/3+ (46 mV)-based cells. The analogues 51-5ht
and 51-57dht provide lower device efficiencies of 7.69% and
8.34%, largely due to poor diffusion of cobalt complexes, which
induces losses in JSC. Thus, careful tuning of the dye structure is
very important.
Table 4. Current−Voltage Characteristics of the Device
Sensitized with 51-57dht.1 at Various Simulated Intensities
a
up to AM1.5G (100 mW cm−2)
intensity
JSC
dye
[mW cm−2
]
[mA cm−2
]
VOC [mV]
FF
η [%]
51-57dht.1
10
50
1.41
7.29
813
877
900
0.806
0.784
0.762
9.21
10.03
9.53
100
13.89
a
All of the devices were fabricated using a 4.1 + 4.5 μm TiO2 anode.
The current−voltage curves of the champion cell measured
under different light intensity are shown in Figure 6, and the
Devices fabricated using an electrolyte that consists of 0.6 M
Co(phen)3 electrolyte (Co(II)/Co(III) ratio of 4:1), 0.1 M LiTFSI,
and 0.8 M TBP in acetonitrile.
exponentially with increased VOC. The Cμ is directly propor-
tional to the density of state (DOS) (Cμ = q(e)DOS, where
q(e) is electron charge).39,40 It is clear that very similar electron
2+/3+
densities are measured for sensitizers with Co(bpy)3
and
Co(phen)32+/3+ redox mediators. The quasi-Fermi level of TiO2
with different sensitizers are revealed in the negative shift: 51-
57dht > 51-57dht.1 > 51-5ht. However, devices constructed
2+/3+
with the 51-57dht.1 sensitizer with Co(phen)3
electrolyte
show longer electron lifetimes (Figure 7b) due to the longer
fluoroalkyl chain efficiently suppressing the back-electron-
transfer rate (i.e., recombination); however, the quasi-Fermi
level of TiO2 is deeper than 51-57dht.
Influence of Sensitizer Driving Force on Regeneration
Kinetics. The influence of driving force between sensitizer
oxidation potential and the redox mediators oxidation potential
for dye regeneration using a cobalt redox mediator was
investigated by transient absorption spectroscopy (TAS), since
the regeneration is a clearer estimation of these devices to
monitor the kinetics of ground-state quenching because there is
no overlapping signal from TiO2 injected electron absorption in
the visible region. Figure 8 demonstrates the ground-state
quenching of the sensitizer at 550 nm without and with the
redox mediator. The decrease in negative absorbance signal
with redox mediator and inert electrolyte shows the
recombination of conduction band electrons with the oxidized
dye molecules. With the redox mediator, the enhanced signal is
strongly accelerated suggesting rapid dye regeneration. The 51-
5ht sensitizer shows a similar regeneration rate using I−/I3− and
cobalt redox mediator, while 51-57dht and 51-57dht.1
sensitizers both show a lower regeneration rate using the
cobalt redox mediator. Therefore, the dye regeneration rate is
not greatly affected by the increased steric bulk substituents of
dye structure. In addition, regenerations of the oxidized dye
Figure 6. Current−voltage characteristics of a 51-57dht.1-sensitized
dye-sensitized solar cell device measured under various light
intensities.
photovoltaic performance data are listed in Table 4. The cell
efficiency reached the higher value of 10.03% at 50% sun
intensity due to less mass transport limitations at low light
intensity.
Transient photocurrent and photovoltage decay measure-
ments were carried out in order to detail the rates of interfacial
recombination of electrons from the TiO2 conduction band to
the electrolyte. It is worth noting that VOC decays are
dependent on the accumulated charge in the TiO2 conduction
band, and so to obtain a fair comparison of the TiO2/
electrolyte+ recombination dynamics between different dyes,
the charge density on both dyes must be equal. As presented in
Figure 7a, the chemical capacitance Cμ of devices rises
F
DOI: 10.1021/acs.inorgchem.6b00427
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
Figure 7. Electron (a) capacitance and (b) lifetime determined with photocurrent and photovoltage decay measurements of devices sensitized with
51-5ht, 51-57dht, and 51-57dht.1 dyes.
Figure 8. Transient absorption kinetics of (a) 51-5ht-, (b) 51-5dht-, and (c) 51-5dht.1-sensitized TiO2 electrodes in the absence (black) and
presence of redox mediator. The transient optical signal was observed at 550 nm after the laser pulse excitation at 500 nm with high intensity probe
light.
with [Co(bpy)3]2+/3+ and [Co(phen)3]2+/3+ are similar and not
affected by the different driving force (60 mV difference) with
increasing (more stabilized) redox potential of cobalt redox
mediators.
AUTHOR INFORMATION
■
Corresponding Authors
*E-mail: ychi@mx.nthu.edu.tw.
*E-mail: mdkhaja.nazeeruddin@epfl.ch.
Author Contributions
All authors have given approval to the final version of the
manuscript.
CONCLUSIONS
■
We have designed and synthesized a series of new isoquinolinyl
functionalized ancillary ligands and incorporated them into
Ru(II) sensitizers to use in a cobalt-based electrolyte DSC
device. The increase in π conjugation in the heterocyclic
ancillary ligands of sensitizers significantly improves the light-
harvesting capabilities resulting short-circuit current (JSC) and
power conversion efficiency. Additionally, the sterically
demanding substituents on the ligands, with optimized TiO2
film, and electrolyte composition, yielded in a power
conversion efficiency of 9.5%, which is the highest reported
to date for the ruthenium sensitizer and a cobalt-based
electrolyte.
Notes
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
■
A.J.H. and M.K.N. thank the European Commission H2020-
ICT-2014-1, SOLEDLIGHT project, grant agreement 643791,
and the Swiss State Secretariat for Education, Research and
Innovation (SERI). Y.C. thanks the MOST of Taiwan for
funding (grant 104-2119-M-007-001).
REFERENCES
■
ASSOCIATED CONTENT
(1) Gratzel, M.; O’Regan, B. Nature 1991, 353, 737.
̈
■
(2) Mathew, S.; Yella, A.; Gao, P.; Humphry-Baker, R.; Curchod, B.
F. E.; Ashari-Astani, N.; Tavernelli, I.; Rothlisberger, U.; Nazeeruddin,
Md. K.; Gratzel, M. Nat. Chem. 2014, 6, 242.
S
* Supporting Information
The Supporting Information is available free of charge on the
ACS Publications website at DOI: 10.1021/acs.inorg-
chem.6b00427.
(3) Yao, Z.; Wu, H.; Li, Y.; Wang, J.; Zhang, J.; Zhang, M.; Guo, Y.;
Wang, P. Energy Environ. Sci. 2015, 8, 3192.
(4) Lee, B. Y.; Lo, Y. Adv. Funct. Mater. 2009, 19, 604.
(5) Wang, J.; Mora-Sero, I.; Pan, Z.; Zhao, K.; Zhang, H.; Feng, Y.;
́
Yang, G.; Zhong, X.; Bisquert, J. J. Am. Chem. Soc. 2013, 135, 15913.
Full experimental details and characterization of
compounds described (PDF)
G
DOI: 10.1021/acs.inorgchem.6b00427
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
(6) Zhao, K.; Pan, Z.; Mora-Sero,
Gong, X.; Wang, J.; Bonn, M.; Bisquert, J.; Zhong, X. J. Am. Chem. Soc.
2015, 137, 5602.
́
I.; Can
́
ovas, E.; Wang, H.; Song, Y.;
(37) Lobello, M. G.; Wu, K.-L.; Reddy, M. A.; Marotta, G.; Gratzel,
M.; Nazeeruddin, M. K.; Chi, Y.; Chandrasekharam, M.; Vitillaro, G.;
De Angelis, F. Dalton Trans. 2014, 43, 2726.
(38) Chandiran, A. K.; Casas-cabanas, M.; Comte, P.; Zakeeruddin, S.
M.; Graetzel, M.; Sauvage, F. J. Phys. Chem. C 2010, 114, 15849.
(39) O'Regan, B. C.; Bakker, K.; Kroeze, J.; Smit, H.; Sommeling, P.;
Durrant, J. R. J. Phys. Chem. B 2006, 110, 17155.
(40) O'Regan, B. C. O.; Durrant, J. R. J. Phys. Chem. B 2006, 110,
8544.
(7) Saliba, M.; Orlandi, S.; Matsui, T.; Aghazada, S.; Cavazzini, M.;
Correa-Baena, J.-P.; Gao, P.; Scopelliti, R.; Mosconi, E.; Dahmen, K.-
H.; De Angelis, F.; Abate, A.; Hagfeldt, A.; Pozzi, G.; Graetzel, M.;
Nazeeruddin, M. K. Nat. Energy 2016, 1, 15017.
(8) Bomben, P. G.; Robson, K. C. D.; Koivisto, B. D.; Berlinguette,
C. P. Coord. Chem. Rev. 2012, 256, 1438.
(9) Ozawa, H.; Sugiura, T.; Kuroda, T.; Nozawa, K.; Arakawa, H. J.
Mater. Chem. A 2016, 4, 1762.
(10) Urbani, M.; Gratzel, M.; Nazeeruddin, M. K.; Torres, T. Chem.
̈
Rev. 2014, 114, 12330.
(11) Hamann, T. W.; Ondersma, J. W. Energy Environ. Sci. 2011, 4,
370.
(12) Sun, Z.; Liang, M.; Chen, J. Acc. Chem. Res. 2015, 48, 1541.
(13) Boschloo, G.; Hagfeldt, A. Acc. Chem. Res. 2009, 42, 1819.
(14) 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. Dalt. Trans. 2012, 41, 11731.
(15) Wadman, S. H.; Kroon, J. M.; Bakker, K.; Havenith, R. W. A.;
van Klink, G. P. M.; van Koten, G. Organometallics 2010, 29, 1569.
(16) Dragonetti, C.; Colombo, A.; Magni, M.; Mussini, P.; Nisic, F.;
Roberto, D.; Ugo, R.; Valore, A.; Valsecchi, A.; Salvatori, P.; Lobello,
M. G.; Angelis, F. De. Inorg. Chem. 2013, 52, 10723.
(17) Sinn, S.; Schulze, B.; Friebe, C.; Brown, D. G.; Jager, M.; Kubel,
̈
̈
J.; Dietzek, B.; Berlinguette, C. P.; Schubert, U. S. Inorg. Chem. 2014,
53, 1637.
(18) Li, G.; Hu, K.; Robson, K. C. D.; Gorelsky, S. I.; Meyer, G. J.;
Berlinguette, C. P.; Shatruk, M. Chem. - Eur. J. 2015, 21, 2173.
(19) Ardo, S.; Meyer, G. J. Chem. Soc. Rev. 2009, 38, 115.
(20) Boschloo, G.; Gibson, E. A.; Hagfeldt, A. J. Phys. Chem. Lett.
2011, 2, 3016.
(21) Feldt, S. M.; Gibson, E. A.; Gabrielsson, E.; Sun, L.; Boschloo,
G.; Hagfeldt, A. J. Am. Chem. Soc. 2010, 132, 16714.
(22) Tsao, H. N.; Yi, C.; Moehl, T.; Yum, J.; Zakeeruddin, S. M.;
Nazeeruddin, M. K.; Gratzel, M. ChemSusChem 2011, 4, 591.
̈
(23) Nelson, J. J.; Amick, T. J.; Elliott, C. M. J. Phys. Chem. C 2008,
112, 18255.
(24) Moser, J.; Zakeeruddin, S. M.; Nazeeruddin, M. K.; Gratzel, M.;
Nusbaumer, H. J. Phys. Chem. B 2001, 105, 10461.
(25) Sapp, S. A.; Elliott, C. M.; Contado, C.; Caramori, S.; Bignozzi,
C. A. J. Am. Chem. Soc. 2002, 124, 11215.
(26) Yella, A.; Lee, H.-W.; Tsao, H. N.; Yi, C.; Chandiran, A. K.;
Nazeeruddin, M. K.; Diau, E. W.-G.; Yeh, C.-Y.; Zakeeruddin, S. M.;
Gratzel, M. Science 2011, 334, 629.
(27) Wu, K.-L.; Hu, Y.; Chao, C.-T.; Yang, Y.-W.; Hsiao, T.-Y.;
Robertson, N.; Chi, Y. J. Mater. Chem. A 2014, 2, 19556.
(28) Wu, K.-L.; Clifford, J. N.; Wang, S.-W.; Aswani, Y.; Palomares,
E.; Lobello, M. G.; Mosconi, E.; De Angelis, F.; Ku, W.-P.; Chi, Y.;
Nazeeruddin, M. K.; Gratzel, M. ChemSusChem 2014, 7, 2930.
̈
(29) Chi, Y.; Wu, K.-L.; Wei, T.-C. Chem. - Asian J. 2015, 10, 1098.
(30) Park, J.-H.; Chung, D. S.; Park, J.-W.; Ahn, T.; Kong, H.; Jung,
Y. K.; Lee, J.; Yi, M. H.; Park, C. E.; Kwon, S.-K.; Shim, H.-K. Org. Lett.
2007, 9, 2573.
(31) Brown, W. D.; Gouliaev, A. H. Synthesis 2002, 2002, 83.
(32) Smith, A. L.; DeMorin, F. F.; Paras, N. A.; Huang, Q.; Petkus, J.
K.; Doherty, E. M.; Nixey, T.; Kim, J. L.; Whittington, D. A.; Epstein,
L. F.; Lee, M. R.; Rose, M. J.; Babij, C.; Fernando, M.; Hess, K.; Le, Q.;
Beltran, P.; Carnahan, J. J. Med. Chem. 2009, 52, 6189.
(33) Hendrickson, J. B.; Rodriguez, C. J. Org. Chem. 1983, 48, 3344.
(34) Giordano, C.; Minisci, F.; Vismara, E.; Levi, S. J. Org. Chem.
1986, 51, 536.
(35) Chung, M.-W.; Lin, T.-Y.; Hsieh, C.-C.; Tang, K.-C.; Fu, H.;
Chou, P.-T.; Yang, S.-H.; Chi, Y. J. Phys. Chem. A 2010, 114, 7886.
(36) 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.
H
DOI: 10.1021/acs.inorgchem.6b00427
Inorg. Chem. XXXX, XXX, XXX−XXX