Influence of the donor size in D-π-A organic dyes for dye-sensitized solar cells

Article abstract of DOI:10.1021/ja500280r

We report two new molecularly engineered push-pull dyes, i.e., YA421 and YA422, based on substituted quinoxaline as a π-conjugating linker and bulky-indoline moiety as donor and compared with reported IQ4 dye. Benefitting from increased steric hindrance w

Full text of DOI:10.1021/ja500280r

Subscriber access provided by UNIV OF OTTAWA
Article
Influence of the Donor Size in D-#-A
Organic Dyes for Dye Sensitized Solar Cells
Jiabao Yang, Paramaguru Ganesan, Joël Teuscher, Thomas Moehl, Yong Joo
Kim, Chenyi Yi, Pascal Comte, Kai Pei, Thomas W. Holcombe, Mohammad K.
Nazeeruddin, Jianli Hua, Shaik M. Zakeeruddin, He Tian, and Michael Grätzel
J. Am. Chem. Soc., Just Accepted Manuscript • DOI: 10.1021/ja500280r • Publication Date (Web): 24 Mar 2014
Downloaded from http://pubs.acs.org on March 29, 2014
Just Accepted
“Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted
online prior to technical editing, formatting for publication and author proofing. The American Chemical
Society provides “Just Accepted” as a free service to the research community to expedite the
dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts
appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been
fully peer reviewed, but should not be considered the official version of record. They are accessible to all
readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered
to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published
in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just
Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor
changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers
and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors
or consequences arising from the use of information contained in these “Just Accepted” manuscripts.
Journal of the American Chemical Society is published by the American Chemical
Society. 1155 Sixteenth Street N.W., Washington, DC 20036
Published by American Chemical Society. Copyright © American Chemical Society.
However, no copyright claim is made to original U.S. Government works, or works
produced by employees of any Commonwealth realm Crown government in the course
of their duties.

Page 1 of 11
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
Influence of the Donor Size in D-
Sensitized Solar Cells
π
-A Organic Dyes for Dye
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Jiabao Yanga,b, Paramaguru Ganesanb, Joël Teuscherb, Thomas Moehlb, Yong Joo Kimb,c,
Chenyi Yib, Pascal Comteb, Kai Peia, Thomas W. Holcombeb, Mohammad Khaja Nazeerud-
dinb, Jianli Huaa*, Shaik M. Zakeeruddinb*, He Tiana and Michael Grätzelb*
aKey Laboratory for Advanced Materials and Institute of Fine Chemicals, East China University of Science and
Technology, Shanghai 200237, People’s Republic of China.
bLaboratory for Photonics and Interfaces, Institute of Chemical Sciences and Engineering, Ecole Polytechnique
Fédérale de Lausanne, CH-1015 Lausanne, Switzerland.
cR & D Center DSC Team, Dongjin Semichem Co., Ltd., 445-935 Hwasung, South Korea.
ABSTRACT: We report two new molecularly engineered push-pull dyes, i.e. YA421 and YA422, based on substituted
quinoxaline as a π-conjugating linker and bulky-indoline moiety as donor and compared with reported IQ4 dye. Benefit-
ting from increased steric hindrance with the introduction of bis(2,4-dihexyloxy)benzene substitution on the quinoxaline,
the electron recombination between redox electrolyte and the TiO2 surface is reduced, especially in redox electrolyte
employing Co(II/III) complexes as redox shuttles. It was found that the open circuit photovoltages of IQ4, YA421 and
YA422 devices with cobalt-based electrolyte are higher than those with iodide/triiodide electrolyte by 34 mV, 62 mV and
135 mV, respectively. Moreover, the cells employing graphene nanoplatelets on top of gold spattered film as a counter
electrode (CE) show lower charge transfer resistance compared to platinum as a CE. Consequently, YA422 devices
deliver the best power conversion efficiency due to higher fill factor, reaching 10.65% at AM 1.5 simulated sunlight.
Electrochemical impedance spectroscopy and transient absorption spectroscopy analysis were performed to under-
stand the electrolyte influence on the device performances with different counter electrode materials and donor struc-
tures of D-π-A dyes. Laser flash photolysis experiments indicate that even though the dye regeneration of YA422 is
slower than that of the other two dyes, the slower back electron transfer of YA422 contributes to the higher device per-
formance.
dyes incorporating internal electron-withdrawing units
such as diketopyrrolopyrrole,16-19 bithiazole,20,21 isoindi-
go,22 benzothiadiazole,10,15,23 benzotriazole,24 and
quinoxaline25-27 into the traditional D-π-A structure were
reported to extend further the spectral response. Among
the non-ruthenium based sensitizers, a donor-acceptor
zinc-porphyrin co-sensitized DSSC in combination with a
cobalt based electrolyte obtained a new PCE record of
12.3%.28
1. INTRODUCTION
Significant attention has been attracted by dye-
sensitized solar cells (DSSCs) as alternatives to tradi-
tional solar cells, owing to their low-cost fabrication com-
bined with high solar energy to electricity power conver-
sion efficiency (PCE).1,2 The performance of DSSCs de-
pends on the nature of sensitizer(s), photoanode, coun-
ter electrode, electrolyte and their combination.3 It is
necessary to utilize dyes that possess broad absorption
spectra to absorb more photons of sunlight for efficient
solar energy to electricity conversion. Ruthenium sensi-
tizers based DSSCs in combination with volatile electro-
lyte and platinum counter electrodes show certified PCE
value of 11.9 %.4-7 At the same time, the research activi-
ty has increased in finding ruthenium free sensitizers
recently, as ruthenium is a noble and expensive metal.
Metal-free organic sensitizers have a number of ad-
vantages, such as high molar extinction coefficient, good
flexibility of molecular tailoring, tunable spectral proper-
ties, relatively high efficiency and low cost.8-13 In general,
most organic sensitizers are synthesized with the donor-
π-acceptor (D-π-A) configuration, due to the efficient
intramolecular charge transfer (ICT) properties of this
class of molecules.14,15 Based on this strategy, a series
of novel donor-acceptor-π-acceptor (D-A-π-A) organic
In terms of electrolyte development for the high DSSC
device performance, cobalt polypyridyl based electro-
lytes have an edge over the other redox electrolytes due
to the feasibility of easily tuning the redox energy of elec-
trolytes by varying the structure of the polypyridyl lig-
ands.29 The main advantage of cobalt based redox elec-
trolytes lies in obtaining higher device open-circuit pho-
tovoltage (Voc) values by tuning the redox energy of the
electrolytes and reducing the excess regeneration ener-
gy losses.30 Redox energy tuning is not possible with
traditional iodide/triiodide based redox electrolytes as the
redox energy of the electrolyte is almost fixed. However,
one of the disadvantages of the cobalt-based electrolyte
compared to iodide/triiodide redox electrolyte is its faster
recombination kinetics with the injected electrons in the
TiO2 film due to its one-electron charge transfer mecha-
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 2 of 11
nism. Faster interfacial charge recombination will lower quinoxaline moiety is depicted in Scheme 1. Suzuki cou-
1
2
3
4
5
6
7
8
the open circuit photovoltage (Voc) and reduce the
charge collection efficiency, thus decrease the short-
circuit photocurrent density (Jsc) and the power conver-
sion efficiency (PCE).31 Evidence from recent research
suggests that the introduction of long alkyl chain groups
into dye structure can effectively control the unwanted
recombination process.32 Inspired by the performance of
quinoxaline dye IQ427 employing iodide electrolyte and to
make an efficient use of the one-electron transfer cobalt
redox electrolyte, we introduced different size bulky do-
nor moieties into D-π-A organic sensitizers (YA421 and
YA422, as shown in Figure 1) to protect the approach of
redox electrolyte closer to the surface of TiO2 film. Two
newly developed dyes were investigated with cobalt and
iodide-based electrolytes and compared with IQ4 dye to
understand the donor size influence on the device per-
formance. The YA422-based DSSC with the biggest
donor group performed worst with iodide-based electro-
lyte but it was the best with cobalt-based electrolyte.
Further optimization of YA422-based DSSC with differ-
ent counter electrodes (graphene nanoplatelets and
gold/graphene nanoplatelets instead of conventional
platinum) and cobalt based electrolyte, yielded the high-
est device PCE of 10.65% at full sunlight intensity. To
the best of our knowledge, this is the highest PCE ever
reported by employing a single metal free sensitizer.
Electrochemical impedance spectroscopy and transient
absorbance spectroscopy analysis were performed to
understand the influence of electrolyte and counter elec-
trode on the dye regeneration and recombination kinet-
ics of DSSC devices.
pling reaction of 5,8-dibromo-2,3-diphenylquinoxaline (1)
with (5-formylthiophen-2-yl)boronic acid gave mono-
aldehyde substituted compound 2. In the next step, Su-
zuki coupling of 2 with two different indoline boronic acid
pinacol esters afforded 3 and 4, respectively. Finally, the
target products (YA421 and YA422) were obtained via
Knoevenagel condensation reaction of the respective
two aldehydes with cyanoacetic acid in the presence of
ammonium acetate. All the intermediates and target
dyes were fully confirmed by 1H NMR, 13C NMR and
HRMS (see the Supporting Information).
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Scheme 1. The synthetic procedure of the dyes
YA421 and YA422.
Optical Properties. The absorption spectra of IQ4,
YA421 and YA422 in dichloromethane (DCM) solution
are shown in Figure 2. The extension of the π-system of
the electron-donating moiety does not lead to significant
changes in the position of the lowest energy absorption
band but has a stronger influence on higher energy ab-
sorption band. For all the three dyes, the lowest energy
absorption peak is around 522-534 nm. The absorption
band positions and their relative intensities are tabulated
in Table 1. Obviously, the insertion of the additional phe-
nyl units and extending the donor part of YA422 leads to
increased intensity of the high energy bands with a blue
shift in the spectral response, which is assigned to π-π*
charge transfer (CT) transitions of the donor moiety. The
shorter wavelength absorption band (360-440 nm) corre-
sponds to quinoxaline moieties as a spacer. The dyes
YA421 and YA422 adsorbed on 4 ꢀm transparent TiO2
films (Figure 2b) show advantage over the IQ4 dye being
red shifted in the spectral response. The blue shift in the
absorption spectra of dyes adsorbed on the surface of
TiO2 film compared to the absorption spectra in solution
(39, 20 and 26 nm for IQ4, YA421 and YA422, respec-
tively) might be ascribed to the deprotonation of the car-
boxylic acid.34
Figure 1. The molecular structures of quinoxaline-based
dyes.
2. RESULTS AND DISCUSSION
Synthesis. Compared to the small indoline as a donor
of IQ4 dye, the uniqueness of the bigger indoline donor
lies in the additional bis(2,4-dialkoxy)benzene substitu-
ents to terminate the indoline unit, which not only im-
prove their solubility, but also decrease electron recom-
bination and thus increase the open circuit voltage, es-
pecially in the presence of cobalt electrolyte.33 The syn-
thetic route to the dyes YA421 and YA422 containing
ACS Paragon Plus Environment

Page 3 of 11
Journal of the American Chemical Society
Table 1. Photophysical and electrochemical properties of dyes
1
2
3
4
E0 c/V
Φ0(S+/S*)e/V
(vs. NHE)
Amount/10-8
(mol*cm-2)
ox
Dye
λmaxa/nm (ε×10-4 M cm)
λmaxb/nm
E0–0d/V
(vs. NHE)
5
IQ4
526 (1.73), 423 (1.47)
487
8.35
0.96
1.97
-1.01
6
7
8
9
YA421
YA422
522 (2.14), 413 (1.83)
534 (2.74), 388 (5.30)
502
508
7.61
5.85
0.95
0.94
1.94
1.91
-0.99
-0.97
a
b
Absorption maximum in DCM at room temperature. Absorption maximum on 4 ꢀm transparent TiO2 films (co-
sensitized with 10 mM CDCA). c Redox potentials were measured in DCM with 0.1 M tetrabutylammonium hexafluorophos-
phate (TBAPF6) as electrolyte (working electrode: glassy carbon; reference electrode: Pt; counter electrode: Pt; calibrated
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
with ferrocene/ferrocenium (Fc/Fc+) as an internal reference and converted to NHE by addition of 0.69 V). E0-0 was esti-
d
e
mated from the absorption thresholds from absorption spectra of dyes in solution. Φ0(S /S*) is estimated by subtracting
+
E0-0 to E0
.
ox
All of the IPCEs do not show great difference due to their
similar absorption spectra and energy levels. For the
iodide-based electrolyte, the devices with IQ4, YA421
and YA422 dyes show similar Voc values whereas the FF
and PCE decreases in the trend of YA422 > YA421 >
IQ4. Due to the smaller size of triiodide spicies com-
pared to Co+3 complex the introduction of long alkyl
chains has less influence in retarding the recombination
in the iodide/triiodide redox electrolyte based devices.
(See the discussion of impedance part) The best PCE of
8.53% was obtained with IQ4 in AM 1.5 simulated solar
light (100 mW cm-2). It is interesting to note that the de-
vice performance trend with cobalt-based electrolyte is
just opposite to the trend that obtained with iodide-based
electrolyte. In case of cobalt-based electrolytes, the
IPCE values of YA421 and YA422 reached a value of 90%
at 550 nm whereas a maximum of 80% is reached for
IQ4 devices. The Voc and PCE of the devices increased
in the order of YA422 > YA421 > IQ4. The lower light
harvesting capability of IQ4 device (as shown in the Fig-
ure S2a) might be responsible for the lower Jsc obtained
for IQ4 devices than that of YA421 and YA422 devices.
Another reason for the lower Jsc of IQ4 devices with co-
balt based electrolyte than that of iodide-based electro-
lyte could be due to the faster electron recombination of
injected electrons from the TiO2 to the electrolyte. It is
known that the electron recombination with one electron
transfer cobalt redox electrolyte is more facile than that
with two-electron transfer iodide based-electrolyte.35
However, in case of YA421 and YA422 dyes the pres-
ence of bulky donor groups (containing long alkoxy
chains) shield the approach of cobalt complexes closer
to TiO2 surface more effectively due to their relatively
bigger size of cobalt complex compared to io-
dide/triiodide molecules. The decreased recombination
in YA422 based device contributes to an increase in the
Voc value by more than 100 mV compared to the IQ4
based devices. The best performing YA422 reached
9.22% PCE at full sunlight intensity in combination with
Pt counter electrode.
Figure 2. UV-Vis absorption spectra of IQ4, YA421 and
YA422 in dichloromethane solution (a) and adsorbed on 4
ꢀm transparent TiO2 films (b).
Electrochemical Properties. The redox potentials of
IQ4, YA421 and YA422 dyes, measured by cyclic volt-
ammetry (CV) are shown in Figure S1 and the values
are presented in Table 1. The redox potentials of the IQ4,
YA421 and YA422 are 0.96, 0.95 and 0.94 V vs NHE,
respectively, which are higher than that of the io-
dide/triiodide (0.40 V vs NHE) or Co(bpy)32+/3+ (0.57 V vs
NHE)31 redox electrolyte energy level, guaranteeing am-
ple driving force for the dye regeneration. The structural
differences in the donor part of IQ4, YA421 and YA422
have little influence on their redox potentials. By extend-
ing the donor conjugation of YA422, the redox potential
is increased by 20 mV compared to IQ4. Estimated from
the absorption thresholds from absorption spectra of
dyes in solution, the zero-zero transition energies (E0-0
)
were determined to be 1.97, 1.94 and 1.91 V, respec-
tively. The excited state reduction potentials of IQ4,
YA421 and YA422 were obtained by combining the re-
dox potentials with the energy of the transition (E0-0). The
calculated excited state potentials of these dyes (-1.01, -
0.99 and -0.97 V, respectively) are placed sufficiently
above the TiO2 conduction band edge to ensure no en-
ergetic barriers for the electron injection.
Photovoltaic Performance of DSSCs. The photovol-
taic performance of devices with IQ4, YA421 and YA422
dyes were investigated using double layer TiO2 films as
photo anodes (4 µm transparent TiO2 layer and 4 µm
TiO2 scattering layer) in combination with platinum (Pt)
as a counter electrode (CE) and the photovoltaic param-
eters are presented in Table 2. The photocurrent-voltage
curve as well the incident photon to current conversion
efficiency (IPCE) value of the IQ4, YA421 and YA422
dyes containing DSSC devices are shown in Figure S2.
Table 2. Photovoltaic performance of the DSSCs
based on IQ4, YA421 and YA422 a
Voc
Jsc
Dye
Electrolyte
FF
η (%)
(mV)
(mA
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 4 of 11
cm-2)
0.02)
15.58
2)
0.001)
0.03)
1
2
3
4
5
6
7
8
797
0.712
(0.709
0.003)
0.729
(0.728
0.001)
0.717
(0.712
0.005)
8.84
(8.73
0.11)
8.77
14.69
0.688
7.79
GNP
771
(775
4)
YA421
100 (15.55 (792
b
[Co(bpy)3]2+/3
(14.4
4
(0.706
(7.90
IQ4
YA421
YA422
IQ4
+
0.03)
7.81
5)
770
(765
5)
0.018
)
0.11
)
0.25)
15.76
50
10
(7.76
0.05)
1.56
(8.65
0.12)
7.78
0.712
9.00
803
(804
1)
[Co(bpy)3]2+/3
(15.4
6
(0.717
(8.91
+
695
(691
4)
9
0.005
)
0.09
)
0.30)
15.26
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
(1.54
0.02)
(7.61
0.17)
0.689
9.22
876
(853
23)
[Co(bpy)3]2+/3
(15.3
4
(0.702
(9.18
15.71
882
0.693
(0.698
0.016)
0.730
(0.728
0.011)
0.742
(0.746
0.023)
9.60
(9.45
0.45)
10.11
(9.85
0.64)
9.86
+
GNP
0.013
)
0.04
)
YA422
100 (15.70 (868
c
0.08)
15.33
0.62)
8.03
17)
862
(845
17)
0.755
8.53
737
(730
8)
(15.5
1
(0.758
(8.60
-
I-/ I3
50
10
(8.04
0.34)
1.64
0.003
)
0.07
)
0.18)
15.41
805
(780
25)
0.711
8.12
741
(744
3)
(1.63
0.06)
(9.68
0.47)
(15.0
8
(0.728
(8.16
-
I-/ I3
YA421
0.017
)
0.04
)
16.25
890
0.737
10.65
Au+
0.33)
14.40
100 (16.36 (887
(0.732 (10.62
GNP
0.682
7.28
b
741
(743
2)
0.11)
8.28
3)
870
(867
3)
0.005)
0.759
0.03)
10.93
(14.4
4
(0.686
(7.36
-
YA422
I-/ I3
0.004
)
0.08
)
0.04)
50
10
(8.23
0.05)
1.68
(0.756 (10.80
a
0.003)
0.779
0.13)
10.69
Measured under irradiation of AM 1.5 simulated solar
light (100 mW cm-2) at room temperature, iodide electrolyte
containing: 1.0 M 1,3-dimethylimidazolium iodide, 0.03 M
815
(811
4)
(1.72
0.04)
(0.752 (10.32
0.27) 0.37)
iodine, 0.1
M guanidinium thiocyanate, 0.5 M tert-
butylpyridine, 0.05 M lithium iodide in acetonitrile: valeroni-
trile (85:15, v/v), cobalt electrolyte containing: 0.22 M
Co(bpy)3[B(CN)4]2, 0.06 M Co(bpy)3[B(CN)4]3, 0.1 M LiClO4,
and 0.5 M 4-tert-butylpyridine in acetonitrile, the concentra-
tion of dyes is 3 × 10-4 M in chloroform/ethanol (v/v: 3/7)
mixed solvent with 10 mM CDCA, platinum is the counter
electrode. Double layer TiO2 film (4+4 um) was used with
active area of 0.159 cm2. Average values are based on two
replicate measurements.
Measured at room temperature, 0.159 cm2 working
area, electrolyte containing: 0.22 M Co(bpy)3[B(CN)4]2,
0.06 M Co(bpy)3[B(CN)4]3, 0.1 M LiClO4, and 0.5 M 4-tert-
butylpyridine in acetonitrile, the concentration of dyes is
0.3 mM in chloroform/ethanol (v/v: 3/7) mixed solvent with
10 mM CDCA, the thickness of TiO2 film is 4+4 ꢀm. b Av-
erage values are based on two replicate measurements. c
Average values are based on fourteen replicate meas-
urements.
a
Table 3. Photovoltaic performance of the DSSCs
based on the three dyes with GNP counter electrode
and YA422 with Au+GNP counter electrode a
Pt is known as the best catalyst for the regeneration of
-
I-/I3 based electrolyte. However, a number of recent
studies show that a carbonaceous type catalyst outper-
forms Pt with the cobalt redox electrolyte, due to the low
charge transfer resistance at the counter electrode.36,37
The influence of CE material on the device performance
was also explored by comparing Pt CE with the gra-
phene nanoplatelets (GNP) CE and the photovoltaic per-
formance at three different light intensities (100%, 50%
and 10%, respectively) are shown in Figure 3 and the
data are presented in Table 3. Applying either Pt or GNP
as CE there is no difference in the PCE trend (YA422 >
YA421 > IQ4) of the devices. Interestingly, in spite of
lower fill factor values obtained with GNP CE, the YA422
device PCE reached 9.60% at full sunlight intensity.
There is a possibility to increase the PCE further by im-
proving the fill factor value of device. Instead of deposit-
I0
Jsc
(mA
(mW
cm-
Voc
(mV)
Dye
CE
FF
η (%)
cm-2)
2
)
14.83
767
0.666
(0.660
0.006)
0.692
(0.687
0.005)
0.672
(0.673
7.57
(7.44
0.13)
7.70
GNP
IQ4
100 (14.66 (769
b
0.17)
7.52
2)
739
(741
2)
50
10
(7.40
0.12)
1.48
(7.53
0.17)
6.45
651
(653
(1.46
(6.42
ACS Paragon Plus Environment

Page 5 of 11
Journal of the American Chemical Society
time (τn) and transport time (τtrans) have been calculated
ing the graphene nanoplatelets on FTO glass to produce
1
2
3
4
5
6
7
8
the CE, a sputtered gold layer (50 nm) was deposited on
FTO glass, which can be beneficial not only to reflect
back the sunlight at the CE (contributing to increased Jsc)
but also to decrease the charge transfer resistance at
the CE (contributing to increased FF). As far as we know,
this is the first time that a gold-sputtered GNP covered
CE is applied in DSSCs. For a champion cell applying
Au+GNP as the CE, the fill factor is improved largely and
combined with an increase in Jsc, the PCE is enhanced
dramatically to 10.65% under full sunlight. At half sun-
light intensity, the device performance even reaches to
10.93% PCE. To the best of our knowledge, this is the
highest device performance reported for any sole D-π-A
metal free dye.
by using the transport (Rtrans) and recombination (Rct)
resistance in conjunction with the chemical capacitance
(Cchem)
Rtrans×Cchem
of the TiO2 (τn
= Rct×Cchem and τtrans =
)
38-40. The iodide electrolyte based devices
show a comparable τn for the dyes YA421 and YA422
and a slightly lower τn for the IQ4, in accordance with
their attained Voc values (Figure 4A and Table 2). There
is a minor shift in the TiO2 conduction band edge posi-
tion (Ecb) of the devices (Figure S3C) as noticed in the
chemical capacitances (about 15 mV drop in the Ecb for
IQ4, YA422 to YA421). This is compensated by the in-
creased τn value of YA422 compared to YA421. The
variations in the FF can be explained by looking at the
counter electrode and the Warburg diffusion resistance
(Figure S4). The IQ4 device shows the highest FF fol-
lowed by YA421 and YA422 devices since the highest
Warburg diffusion resistance is observed for YA422
while the highest CE resistance is measured for YA421.
The differences in Jsc cannot be conclusively explained
by EIS experiments as these measurements were done
under dark current conditions.
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
In the case of cobalt redox electrolyte devices, the ob-
served very small difference in the Ecb position of the
mesoporous TiO2 for the devices with the different dyes
emphasises that the increases in the Voc results from an
increase in the electron lifetime (Figure 4B). The Voc dif-
ference between IQ4 and YA421 is ∼30 mV and be-
tween YA421 to YA422 it is ∼70 mV. The τn values are
in good agreement with the observed device Voc values
(a rough estimate from the diode equation ꢀVoc
=
(kBT/q)(ln(τn,dye1/τn,dye2) with kB as Boltzmann constant, q
as elementary charge and T as temperature leads to a
difference of about ꢀVoc = 20 mV between IQ4 and
YA421 and ꢀVoc = 68 mV YA421 to YA422). The bulkier
donor part of the dye shields the TiO2 interface in a bet-
ter way from approaching of the oxidized species of the
redox electrolyte thus reducing the recombination rate.
The concept of protecting the TiO2 surface by the bulkier
donor groups with the cobalt-based redox electrolyte has
been reported earlier.32,41 The higher Warburg diffusion
resistance of IQ4 (Figure 5 and S5) is responsible for the
lower FF, whereas the higher counter electrode re-
sistance of YA422 compared to YA421 is contributing to
its lower FF values.
Figure 3. (a) I-V curves for DSSCs based on dyes IQ4,
YA421, YA422 using GNP counter electrode and YA422
using Au-sputtered GNP counter electrode. (b) IPCE curves
for DSSCs based on YA422 employing two different counter
electrodes.
Electrochemical Impedance Analysis. To under-
stand the intriguing results obtained with two different
redox electrolytes with different CE materials, we applied
electrochemical impedance spectroscopy (EIS) analysis
to get a clear picture on the internal electrical parameters
of the devices. The measured dark currents and the re-
sults obtained by fitting the EIS measurements using the
transmission line model38-40 are presented in the support-
ing information (Figure S3). The apparent electron life-
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 6 of 11
(red), YA421 (blue) and YA422 (green). Black represents
the resistance of the device with Au+GNP CE.
1
A
2
3
4
5
6
7
8
9
Photoinduced charge transfer dynamics. Femto-
second pump-probe spectroscopy is used to identify
electron injection in the semiconductor. At first, we
measured the transient absorbance of dyes adsorbed on
insulating alumina films allowing observation of spectral
changes due to the formation of dye-excited state (see
SI for details and Figure S7-S9).
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
B
Figure 6. Transient spectra of YA422 adsorbed on titania
and measured in the presence of acetonitrile after excitation
at 580 nm. For IQ4 and YA421 dyes the data is presented
in Figures S10 and S11.
Figure 4. Electron lifetime (solid) and electron transport time
(dotted) determined from EIS measurements for the dyes
IQ4 (red), YA421 (blue) and YA422 (green). A: Iodine redox
system based devices; B: cobalt redox system based
devices with GNP CEs.
The situation is different if the dyes are adsorbed on
the titania film (Figure 6). The excited dye injects an
electron into the titania conduction band and then the
molar extinction of excited state (eS*) is mixed with oxi-
dized dye (eS+) spectra. On TiO2 film, the maximum of
the transient absorbance is slightly shifted towards the
red at 480 nm and this shift might be due to the differ-
ences in the substrate materials (titania or alumina) in-
ducing different electronic environment on the dye. This
spectral shift might also arise from the differences in the
spectral shape of the excited and oxidized forms of the
dye and therefore would reveal some ultrafast electron
injection during the initial excitation (<100 fs). Unfortu-
nately, even at the shortest times, one cannot see any
wavelength dependent dynamics, preventing the unam-
biguous assignment of different species over the investi-
gated wavelengths. For all the three dyes, the spectral
shape is similar (Figure 6, S10 and S11).
The champion cell fabricated with YA422 in combina-
tion with Au+GNP based CE yielded the highest PCE of
10.65% due to the best FF and Jsc values. In this device,
an increase in Jsc compared to GNP CE is most probably
due to an enhanced back light reflection from the CE
(enabling higher light harvesting mainly in the wave-
length region between 300 and 400 nm) while the FF
increase can be evidently assigned to the reduced total
series resistance as shown in Figure 5.
For the IQ4, YA421 and YA422 dyes adsorbed on the
titania film the transients extracted at 480 nm after excit-
ing the dye at 580 nm are presented in Figure 7. Even
having a similar spectral shape, dynamics for titania films
are different to that observed on alumina. A first rapid
decrease is observed on TiO2, followed by a plateau
over the investigated timescale (1.6 ns), assigned to the
oxidized state for the three dyes. This indicates that a
difference between eS+ and eS* allows us for monitoring
electron injection at this wavelength. We excluded the
influence of a Stark signal at this wavelength thanks to
the absence of positive absorbance change in the pho-
toinduced absorption spectroscopy (PIA) spectra (see SI
Figure S6). The transient absorbance spectrum wit-
Figure 5. Sum of the series resistance originating from the
counter electrode and the Warburg diffusion resistance
plotted against the passed dark current for the dyes IQ4
ACS Paragon Plus Environment

Page 7 of 11
Journal of the American Chemical Society
tion. Oxidized IQ4 has a lifetime of 438 µs. Increasing
nessed on alumina samples also supports observation of
1
2
3
4
5
6
7
8
the excited dye species only. We extracted the electron
injection lifetime from a mono-exponential fit of the decay
with time constants of 4.0, 3.4 and 5.8 ps for IQ4, YA421
and YA422, respectively. At these short time spans, the
dye excited state population did not show a significant
decrease on alumina samples (Figure S9), confirming
that in this case there is no electron injection. Similar
dynamics are found above 700 nm, also thanks to a dif-
ference between eS+ and eS* (1.9, 1.4, 2.2 ps at 700 nm,
respectively, see Figure S12). This confirms that the
differences in the LUMO levels of these three dyes have
negligible influence on the electron injection into the TiO2
conduction band. The measured time constants for elec-
tron injection are significantly slower than the fastest
reports for ruthenium dyes42, however, the spread of the
dynamics is less important and comparable with other
reported D-π-A dye results.43 The observed plateau at
longer time period ensures that the oxidized dye is stable
over this timescale and no back electron transfer is oc-
curring. Over the investigated spectral range there is,
however, no simple wavelength to extract that would
allow to monitor directly either the oxidized dye or its
excited state.
the donor size leads to an increase in the lifetime of
YA421 and YA422 to 595 µs and 1070 µs, respectively,
confirming the hole stabilization by the bigger donor
group.
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 7. Normalised transient dynamics extracted at
480 nm for IQ4, YA421 and YA422 adsorbed on TiO2 film
after exiting at 580 nm in presence of acetonitrile solution.
Data are smoothed and solid lines are the result of a mono-
exponential fit.
To monitor the influence of dye donor group towards
the oxidized dye lifetime and its regeneration by electro-
lytes, nano-microsecond transient absorbance spectros-
copy was applied under open-circuit conditions (Figure
8). After excitation at 535 nm, with a low fluence ensur-
ing an average not more than one electron is injected
per nanoparticle, transient decays are monitored at 1100
nm, following the oxidized dye signature found with PIA
experiment (see SI Figure S6). Electrons also absorb at
this wavelength, but they have a much smaller extinction
coefficient compared to the oxidized dye species.44 As
traces significantly deviate from mono-exponential de-
cays, transient absorbance decays are fitted with a
stretched exponential to extract back electron transfer
and dye regeneration lifetimes.
Figure 8. Nanosecond normalized transient absorbance
decays of IQ4, YA421 and YA422 dyes at 1100 nm in
presence/absence of redox electrolyte (iodide or cobalt).
Oxidized dye regeneration occurs in the presence of
redox active electrolyte, and therefore a faster decay of
dye cation should be present for efficient device perfor-
mance. For IQ4, efficient dye regeneration occurs in the
presence of both redox electrolyte types. YA421 pre-
sents faster dye regeneration compared to YA422 by
cobalt-based electrolyte than that of iodide-based elec-
trolyte. The results of the fits and the regeneration yields
(ηreg) are calculated assuming first order rate for dye re-
generation and back electron transfer reactions (ηreg
=
kreg/(kreg + krec), ki = 1 / ti) (in this study, the lifetime ex-
tracted from the fit is used as an approximation for ti) are
summarized in Table 4.45 This estimation of the regener-
ation yield might overestimate the true regeneration effi-
ciency according to a recent study.46
In the absence of redox active electrolyte such as ace-
tonitrile, oxidized dye has no other pathway than recom-
bining with the injected electrons in TiO2 film. Thus, fol-
lowing the decay of the oxidized dye allows one to
measure the lifetime of the back electron transfer reac-
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 8 of 11
At open-circuit conditions, dye regeneration yields of creased 6% due to the decrease in the charge transfer
1
2
3
4
5
6
7
8
IQ4 and YA421 are close to unity with cobalt-based elec-
trolyte (ηreg = 0.99 and 0.98, respectively) and expected
to have excellent photovoltaic performance. Whereas in
case of iodide based electrolyte, the dye regeneration
yield is lower for YA421 (ηreg = 0.93) than that of IQ4
(ηreg = 0.98). However, the dye regeneration reaction is
much slower for YA422 with both the (iodide or cobalt-
based) electrolytes. Fortunately, the long lived oxidized
state of YA422 is responsible for preventing back elec-
tron transfer issues and permits for good dye regenera-
tion yields (ηreg = 0.94 and 0.90 for cobalt and iodide
electrolytes, respectively) which contributes to excellent
device photovoltaic performance. After charge injection,
the donor moiety of YA422 allows for delocalisation of
the hole further away from the TiO2 interface. Because of
the larger charge separation distance with increasing
size of the donor, the trend is similar for the dye regen-
eration reaction and the back electron transfer reaction,
in the presence of both types of electrolyte. This effect is
attributed to a steric hindrance of the dye blocking the
approach of the redox mediator similarly to what has
been observed for ruthenium dyes.47 The steric hin-
drance of the dye has same influence in the presence of
both the electrolytes that are investigated. As a conse-
quence, dye regeneration reaction lifetime is slowed
down by one order of magnitude along the dye series for
both the electrolytes. The cobalt-based electrolyte in-
volves a one electron dye regeneration reaction com-
pared to two electrons involved in iodide-based electro-
lyte, and this partly explains the observed differences for
the same dye with two different electrolytes.
resistance at the CE and the Jsc increased due to the
higher back reflection. Therefore the device performance
enhanced to a record efficiency of 10.65%. To the best
of our knowledge, this is the highest efficiency reported
by a single organic D-π-A dye. Laser flash photolysis
experiments indicate that even though the dye regenera-
tion of YA422 is slower than that of the other two dyes,
the slower back electron transfer of YA422 contributes to
the higher device performance. This work emphasizes
the significance of the size of the dye in controlling the
kinetic parameters, which are responsible in determining
the device PCE. Further work will be focussed on ration-
al dye design for this system towards red-shifting the dye
absorption, giving possibilities to harvest a broader
range of photons in the solar spectrum to increase the
device performance.
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
EXPERIMENTAL SECTION
Materials. Tetra-n-butylammonium hexafluorophos-
phate (TBAPF6), 4-tert-butylpyridine (4-TBP), lithium
iodide, and 2-cyanoacetic acid were purchased from
Fluka. Starting
diphenylquinoxaline,
materials 5,8-dibromo-2,3-
4-(2',4'-bis(hexyloxy)-[1,1'-
biphenyl]-4-yl)-7-(4,4,5,5-tetramethyl-1,3,2-dioxaborolan-
2-yl)-1,2,3,3a,4,8b-hexahydrocyclopenta[b]indole and 4-
(4-(2,2-bis(2',4'-bis(hexyloxy)-[1,1'-biphenyl]-4-
yl)vinyl)phenyl)-7-(4,4,5,5-tetramethyl-1,3,2-
dioxaborolan-2-yl)-1,2,3,3a,4,8b-
hexahydrocyclopenta[b]indole were synthesized accord-
ing to the corresponding literature methods.19,27
Co(bpy)3[B(CN)4]2 and Co(bpy)3[B(CN)4]3 were prepared
in the same manner reported previously.48 The Thin-
layer chromatography (TLC) was conducted with Merck
KGaA precoated TLC Silica gel 60 F254 aluminum
sheets and visualized with UV and potassium perman-
ganate staining. Flash column chromatography was per-
formed on glass columns packed with silica gel using
Silicycle UltraPure SilicaFlash P60, 40–63 mm (230–400
mesh). Graphene nanoplatelets (GNP) were purchased
from Cheap Tubes, Inc (USA, grade 3). All other sol-
vents and chemicals were purchased from Aldrich and
used as received without further purification.
Table 4. Time constants extracted from the fits of
Figure 8 and calculated regeneration yields.
τreg iodide
(µs)
τreg cobalt
(µs)
τrec
(µs)
ηreg iodide
ηreg cobalt
IQ4
438
595
10.9
0.98
0.93
0.90
5.7
0.99
0.98
0.94
YA421
YA422
43.2
13.4
66.4
1070
120.7
Preparation of the solar cells. The dye-sensitized
TiO2 photoelectrode was prepared by following the pro-
cedure reported in the literature.49 The TiO2 transparent
photoelectrodes (4 ꢀm) were prepared by screen-printing
onto FTO (fluorine-doped tin oxide, Solar 4 mm thick-
ness, 10 ohms per square, Nippon Sheet Glass) con-
ducting glass. The screen-printing paste was composed
of ~20 nm diameter anatase particles that gave a meso-
porous layer with ~32 nm pores. In order to render high
PCE, a scattering layer (4 ꢀm) (400 nm diameter, Cata-
lysts & Chemicals Ind. Co. Ltd. (CCIC), HPW-400) was
deposited on the transparent layer. Sintering was carried
out at 500°C for 15 min. The films were soaked in the
solution of 0.02 M aqueous TiCl4 for 30 minutes at 70°C.
After being washed with deionized water and fully rinsed
with ethanol, the films were heated again at 500°C for 15
min. These electrodes were immersed into a 0.3 mM
solution of dye with 10 mM 3α,7α-dihydroxy-5β-cholic
acid (CDCA) in chloroform/ethanol (3:7 v/v) mixed sol-
vent and kept for 3 h at room temperature. To prepare
3. CONCLUSION
We successfully introduced two various degrees of
bulky size donors into quinoxaline-based D-π-A dyes as
a part of molecular engineering of dyes and to test them
in dye-sensitized solar cells. Compared to the reported
IQ4 dye bearing indoline donor, bulky indoline-based
dyes YA421 and YA422 acquire larger molar extinction
coefficient, which is beneficial to apply thinner semicon-
ductor films in DSSCs. It is interesting to note that the
photovoltaic performance of all the three dyes studied
with Pt as a CE show opposite trend with iodide/triiodide
(IQ4 > YA421 > YA422) and cobalt redox electrolytes
(YA422 > YA421 > IQ4). Electron recombination process
was efficiently impeded due to the presence of long
alkoxy chains on the bulky donor moieties, particularly
with cobalt-based redox electrolyte. For YA422 dye-
containing device, 9.60% PCE was obtained in combina-
tion with the GNP CE and cobalt-based electrolyte. Ap-
plying Au+GNP as a CE, the FF of YA422 device in-
ACS Paragon Plus Environment

Page 9 of 11
Journal of the American Chemical Society
plified Ti:Sapphire femtosecond laser at 1 kHz (CPA-
the counter electrode, a hole (0.8-mm diameter) was
1
2
3
4
5
6
7
8
drilled on the counter electrode by a Drill-press. For Pt or
GNP electrode,36,37 the Pt or graphene catalyst was de-
posited on the cleaned FTO glass by coating with four
drops of H2PtCl6 solution (0.02 M in 2-propanol solution)
or GNP solution (1 mg in 1 mL 2-propanol solution) with
the heat treatment at 400 °C for 15 min. For Au/GNP
electrode, a gold layer (50 nm thick) was deposited on
the FTO substrate by a sputter first, and then deposited
another graphene layer on the gold electrode with four
drops of GNP solution. About the assembly of cells, the
dye-contained TiO2 film and counter electrode were as-
sembled into a sandwich-type cell and sealed with a hot-
melt gasket of 25 ꢀm thickness made of the ionomer
Surlyn 1702 (Dupont). The redox electrolyte was placed
in a drilled hole in the counter electrode, and was driven
into the cell by means of vacuum backfilling. The iodide
electrolyte consists of 1.0 M 1,3-dimethylimidazolium
iodide, 0.03 M iodine, 0.1 M guanidinium thiocyanate,
0.5 M tert-butylpyridine, 0.05 M lithium iodide in acetoni-
trile: valeronitrile (85:15, v/v). The cobalt electrolyte con-
2001, Clark-MXR). It sourced a non-collinear optical par-
ametric amplifier tuned at 580 nm for the pump and a
white light continuum generated in a CaF2 crystal for the
probe. Typical instrument response measured by a Kerr
gating technique in a thin glass window was 50 fs and
spectra were corrected for the temporal chirp of the
white light pulse. Probe pulses were delayed with re-
spect to the pump pulses using a computer controlled
optical delay line. The probe was detected with a pair of
150 mm spectrographs (SpectraPro-2150, Acton Re-
search Corporation) equipped with 512x58 pixels back-
thinned CCD detectors (Hamamatsu S7030-0906). The
transmitted signal intensity was recorded shot by shot
and corrected for fluctuations against a reference beam.
4000 averaged pulses were needed to obtain satisfacto-
ry signal-to-noise ratio. Typical measurements were per-
formed over 1.6 ns with 400 data points and a smallest
increment of 20 fs close to time zero. The fluence on the
cells was kept low, typically at 400 ꢀJ cm−2. Samples
were prepared in a similar way to devices, omitting the
scattering layer and filled with acetonitrile. They were
excited from the counter-electrode side of the device.
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
sists of 0.22
M
Co(bpy)3[B(CN)4]2, 0.06
M
Co(bpy)3[B(CN)4]3, 0.1 M LiClO4, and 0.5 M 4-tert-
butylpyridine in acetonitrile. Finally, the hole was sealed
using Surlyn and a cover glass (0.1 mm thickness).
Nanosecond transient absorbance spectroscopy
(TAS). TAS was applied to the same samples than for
femtosecond measurements either with acetonitrile or
the desired electrolyte. Excitation (λ = 535 nm, 7 ns
pulse duration, 20 Hz repetition rate) was carried out by
Photovoltaic performance measurements. For pho-
tovoltaic measurements of the DSCs, a solar simulator
equipped with a 450 W xenon light source (Osram XBO
450) with a filter (Schott 113) was employed, whose
power was regulated to the AM 1.5 solar standard by
using a reference Si photodiode equipped with a colour-
matched filter (KG-3, Schott) to reduce the mismatch in
the region of 350–750 nm between the simulated light
and AM 1.5 to less than 4%. The measurement-settling
time between applying a voltage and measuring a cur-
rent density for the I–V characterization of DSCs was
fixed to 80 ms with a Keithley model 2400 digital source
meter. The photocurrent action spectra were measured
with an Incident Photon-to-Current Conversion Efficiency
(IPCE) test system. The modulation frequency used was
about 2 Hz and light from a 300WXenon lamp (ILC
Technology, USA) was focused through a computer con-
trolled Gemini-180 double monochromator (John Yvon
Ltd., UK). A white light bias (5%) was used to bring the
total light intensity on the device under testing closer to
operating conditions. A light mask was used on the
DSCs, so the illuminated active area of DSCs was fixed
to 0.159 cm2.
a
Powerlite 7030 frequency-doubled Q–switched
Nd:YAG laser (Continuum) pumping an optical paramet-
ric oscillator (OPO, GWU-355). A planoconcave lens
expanded the laser beam output to homogeneously irra-
diate the sample and the laser fluence was attenuated
with neutral density grey filters down to <50 µJ cm-2 per
pulse. The probe light was produced by a continuous
wave xenon arc lamp, first passed through bandpass
filters up to 1000 nm, focused onto the the sample, and
then through a monochromator (LOT-Oriel, Omni-λ 150),
before being detected by a fast InGaAs diode (Thorlabs,
SM05PD5a) across a 1 kꢁ    . Data waves were
recorded on a digital oscilloscope (Tektronix, DPO 7254).
Satisfactory signal-to-noise ratios were typically obtained
by averaging over 1000 laser shots.
ASSOCIATED CONTENT
Supporting Information. Synthetic procedures and charac-
terization data of all new compounds. Details for all physical
characterizations. This material is available free of charge
via the Internet at http://pubs.acs.org.
Electrochemical impedance spectroscopy (EIS).
The EIS analysis was done on DSC devices at room
temperature in the dark. A sinusoidal potential perturba-
tion with an amplitude of 15 mV was applied over a fre-
quency range 7 MHz – 0.1 Hz (Bio Logic SP300 potenti-
ostat) at constant potential bias. The bias potential var-
ied between 0 mV and Voc with about 50 mV step. The
spectra were fitted using ZView software (Scribner As-
sociates) using the transmission model line.38-40 If the
potential was corrected for the ohmic losses by the se-
ries resistance of the devices this is indicated in the fig-
ure caption.
AUTHOR INFORMATION
Corresponding Author
jlhua@ecust.edu.cn
shaik.zakeer@epfl.ch
michael.graetzel@epfl.ch
Notes
The authors declare no competing financial interest.
ACKNOWLEDGMENT
Femtosecond pump-probe spectroscopy. The
femtosecond transient spectrometer is based on an am-
This work was supported by NSFC/China (2116110444,
21172073, 91233207 and 21372082), the National Basic
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 10 of 11
(23) Zhu, W. H.; Wu, Y. Z.; Wang, S. T.; Li, W. Q.; Li, X.;
Chen, J.; Wang, Z. S.; Tian, H. Adv. Funct. Mater. 2011, 21,
756.
(24) Mao, J. Y.; Guo, F. L.; Ying, W. J.; Wu, W. J.; Li, J.; Hua,
J. L. Chem. Asian J. 2012, 7, 982.
(25) Chang, D. W.; Lee, H. J.; Kim, J. H.; Park, S. Y.; Park,
S.-M.; Dai, L. M.; Baek, J.-B. Org. Lett. 2011, 13, 3880.
(26) Pei, K.; Wu, Y. Z.; Zhang, Q.; Chen, B.; Tian, H.; Zhu, W.
H. Chem. Eur. J. 2012, 18, 8190.
(27) Pei, K.; Wu, Y. Z.; Islam, A.; Zhang, Q.; Han, L. Y.; Tian,
H.; Zhu, W. H. ACS Appl. Mater. Interfaces 2013, 5, 4986.
(28) Yella, A.; Lee, H. W.; Tsao, H. N.; Yi, C. Y.; Chandiran,
A. K.; Nazeeruddin, M. K.; Diau, E. W. G.; Yeh, C. Y.; Zakee-
ruddin, S. M.; Grätzel, M. Science 2011, 334, 629.
(29) Yum, J-H.; Baranoff, E.; Kessler, F.; Moehl, T.; Ahmad,
S.; Bessho, T.; Marchioro, A.; Ghadiri, E.; Moser, J-E.; Yi, C. Y.;
Nazeeruddin, M. K.; Grätzel, M. Nat. Commun. 2012, 3, 631.
(30) DeVries, M. J.; Pellin, M. J.; Hupp, J. T. Langmuir 2010,
26, 9082.
(31) Sapp, S. A.; Elliott, C. M.; Contado, C.; Caramori, S.;
Bignozzi, C. A. J. Am. Chem. Soc. 2002, 124, 11215.
(32) Feldt, S. M.; Gibson, E. A.; Gabrielsson, E.; Sun, L. C.;
Boschloo, G.; Hagfeldt, A. J. Am. Chem. Soc. 2010, 132, 16714.
(33) Gao, P.; Kim, Y. J.; Yum, J-H.; Holcombe, T. W.; Nazee-
ruddin, M. K.; Grätzel, M. J. Mater. Chem. A 2013, 1, 5535.
(34) Wang, Z. S.; Cui, Y.; Dan-oh, Y.; Kasada, C.; Shinpo, A.;
Hara, K. J. Phys. Chem. C 2007, 111, 7224.
(35) Nakade, S.; Makimoto, Y.; Kubo, W.; Kitamura, T.; Wa-
da, Y.; Yanagida, S. J. Phys. Chem. B 2005, 109, 3488.
(36) Kavan, L.; Yum, J-H.; Grätzel, M. Nano Lett. 2011, 11,
5501.
(37) Kavan, L.; Yum, J-H.; Nazeeruddin, M. K.; Grätzel, M.
ACS Nano 2011, 5, 9171.
(38) Fabregat-Santiago, F.; Bisquert, J.; Garcia-Belmonte, G.;
Boschloo, G.; Hagfeldt, A. Sol. Energy Mater. Sol. Cells 2005,
87, 117.
(39) Fabregat-Santiago, F.; Garcia-Belmonte, G.; Mora-Sero,
I.; Bisquert, J. Phys. Chem. Chem. Phys. 2011, 13, 9083.
(40) Wang, M. K.; Grätzel, C.; Moon, S.; Humphry-Baker, R.;
Rossier-Iten, N.; Zakeeruddin, S. M.; Grätzel, M. Adv. Funct.
Mater. 2009, 19, 2163.
(41) Tsao, H. N.; Yi, C. Y.; Moehl, T.; Yum, J-H.; Zakeerud-
din, S. M.; Nazeeruddin, M. K.; Grätzel, M. ChemSusChem
2011, 4, 591.
(42) Wenger, B.; Grätzel, M.; Moser, J.-E. J. Am. Chem. Soc.
2005, 127, 12150.
(43) Ziolek, M.; Martín, C.; Sun, L. C.; Douhal, A. J. Phys.
Chem. C 2012, 116, 26227.
(44) Rothenberger, G.; Fitzmaurice, D.; Grätzel, M. J. Phys.
Chem. 1992, 96, 5983.
Research 973 Program (2013CB733700), and the Funda-
mental Research Funds for the Central Universities
(WJ0913001). PG is gratefully acknowledges financial sup-
port from the Swiss confederation under Swiss Government
Scholarship programme and UGC, India. MG thanks the
Swiss National Science Foundation and European Re-
search Council (ERC) for an Advanced Research Grant
(No. 247404) funded under the CE-Mesolight project. Yang
would like to thank the funding of the visiting program,
which is supported by China Scholarship Council (CSC) and
Dr. Kuan-Lin Wu (National Tsing Hua University, Taiwan)
for fruitful discussion.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
REFERENCES
(1) O’Regan, B; Grätzel, M. Nature 1991, 353, 737.
(2) Grätzel, M. Acc. Chem. Res. 2009, 42, 1788.
(3) Hagfeldt, A.; Boschloo, G.; Sun, L. C.; Kloo, L.; Petters-
son, H. Chem. Rev. 2010, 110, 6595.
(4) Komiya, R.; Fukui, A.; Murofushi, N.; Koide, N.; Yama-
naka, R.; Katayama, H. Technical Digest, 21st International
Photovoltaic Science and Engineering Conference, Fukuoka
November 2011, 2 C-50-08.
(5) Chen, C. Y.; Wang, M. K.; Li, J. Y.; Pootrakulchote, N.;
Alibabaei, L.; Ngoc-le, C.; Decoppet, J. D.; Tsai, J. H.; Grätzel,
C.; Wu, C. G.; Zakeeruddin, S. M.; Grätzel, M. ACS Nano 2009,
3, 3103.
(6) Gao, F. F.; Wang, Y.; Shi, D.; Zhang, J.; Wang, M. K.;
Jing, X. Y.; Humphry-Baker, R.; Wang, P.; Zakeeruddin, S. M.;
Grätzel, M. J. Am. Chem. Soc. 2008, 130, 10720.
(7) Yu, Q.; Wang, Y.; Yi, Z.; Zu, N.; Zhang, J.; Zhang, M.;
Wang, P. ACS Nano 2010, 4, 6032.
(8) Qu, S. Y.; Hua, J. L.; Tian, H. Sci. China Chem. 2012, 55,
677.
(9) Yang, J. B.; Guo, F. L.; Hua, J. L.; Li, X.; Wu, W. J.; Qu,
Y.; Tian, H. J. Mater. Chem. 2012, 22, 24356.
(10) Wu, Y. Z.; Marszalek, M.; Zakeeruddin, S. M.; Zhang, Q.;
Tian, H.; Grätzel, M.; Zhu, W. H. Energy Environ. Sci. 2012, 5,
8261.
(11) Bessho, T.; Zakeeruddin, S. M.; Yeh, C-Y.; Diau, E. W-
G.; Grätzel, M. Angew. Chim. Int. Ed. 2010, 49, 6646.
(12) Marszalek, M.; Nagane, S.; Ichake, A.; Humphry-Baker,
R.; Paul, V.; Zakeeruddin, S. M.; Grätzel, M. RSC Adv. 2013, 3,
7921.
(13) Mao, J. Y.; He, N. N.; Ning, Z. J.; Zhang, Q.; Guo, F. L.;
Chen, L.; Wu, W. J.; Hua, J. L.; Tian, H. Angew. Chim. Int. Ed.
2013, 51, 9873.
(14) Hong, Y. P.; Liao, J.-Y.; Cao, D. R.; Zang, X. F.; Kuang,
D.-B.; Wang, L. Y.; Meier, H.; Su, C.-Y. J. Org. Chem. 2011, 76,
8015.
(15) Haid, S.; Marszalek, M.; Mishra, A.; Wielopolski, M.;
Teuscher, J.; Moser, J.-E.; Humphry-Baker, R.; Zakeeruddin, S.
M.; Grätzel, M.; Bäuerle, P. Adv. Funct. Mater. 2012, 22, 1291.
(16) Qu, S. Y.; Wu, W. J.; Hua, J. L.; Kong, C.; Long, Y. T.;
Tian, H. J. Phys. Chem. C 2010, 114, 1343.
(17) Qu, S. Y.; Qin, C. J.; Islam, A.; Wu, Y. Z.; Zhu, W. H.;
Hua, J. L.; Tian, H.; Han, L. Y. Chem. Commun. 2012, 48, 6972.
(18) Holcombe, T. W.; Yum, J-H.; Yoon, J.; Gao, P.; Mar-
szalek, M.; Di Censo, D.; Rakstys, K.; Nazeeruddin, M. K.;
Grätzel, M. Chem. Commun. 2012, 48, 10724.
(19) Yum, J-H.; Holcombe, T. W.; Kim, Y. J.; Rakstys, K.;
Moehl, T.; Teuscher, J.; Delcamp, J. H.; Nazeeruddin, M. K.;
Grätzel, M. Scientific Reports 2014, 3, DOI: 10.1038/srep02446.
(20) He, J. X.; Wu, W. J.; Hua, J. L.; Jiang, Y. H.; Qu, S. Y.;
Li, J.; Long, Y. T.; Tian, H. J. Mater. Chem. 2011, 21, 6054.
(21) He, J. X.; Guo, F. L.; Wu, W. J.; Li, X.; Yang, J. B.; Hua,
J. L. Chem. Eur. J. 2012, 18, 7903.
(45) Zhang, Z.; Ito, S.; Moser, J.-E.; Zakeeruddin, S. M.;
Grätzel, M. ChemPhysChem 2009, 10, 1834.
(46) Li, F.; Jennings, J. R.; Wang, Q. ACS Nano 2013, 7,
8233.
(47) Wang, P.; Zakeeruddin, S. M.; Moser, J.-E.; Nazeerud-
din, M. K.; Sekiguchi, T.; Grätzel, M. Nat. Mater. 2003, 2, 402.
(48) Aribia, K. B.; Moehl, T.; Zakeeruddin, S. M.; Grätzel, M.
Chem. Sci. 2013, 4, 454.
(49) Wang, P.; Zakeeruddin, S. M.; Comte, P.; Charvet, R.;
Humphry- Baker, R.; Grätzel, M. J. Phys. Chem. B 2003, 107,
14336.
(22) Ying, W. J.; Guo, F. L.; Li, J.; Zhang, Q.; Wu, W. J.; Tian,
H.; Hua, J. L. ACS Appl. Mater. Interfaces 2012, 4, 4215.
ACS Paragon Plus Environment

Page 11 of 11
Journal of the American Chemical Society
TOC
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
11
ACS Paragon Plus Environment