Increasing the Efficiency of Organic Dye-Sensitized Solar Cells over 10.3% Using Locally Ordered Inverse Opal Nanostructures in the Photoelectrode

Article abstract of DOI:10.1002/adfm.201706291

3D inverse opal (3D-IO) oxides are very appealing nanostructures to be integrated into the photoelectrodes of dye-sensitized solar cells (DSSCs). Due to their periodic interconnected pore network with a high pore volume fraction, they facilitate electrolyte infiltration and enhance light scattering. Nonetheless, preparing 3D-IO structures directly on nonflat DSSC electrodes is challenging. Herein, 3D-IO TiO₂ structures are prepared by templating with self-assembled polymethyl methacrylate spheres on glass substrates, impregnation with a mixed TiO₂:SiO₂ precursor and calcination. The specific surface increases from 20.9 to 30.7 m² g^?1 after SiO₂ removal via etching, which leads to the formation of mesopores. The obtained nanostructures are scraped from the substrate, processed as a paste, and deposited on photoelectrodes containing a mesoporous TiO₂ layer. This procedure maintains locally the 3D-IO order. When sensitized with the novel benzothiadiazole dye YKP-88, DSSCs containing the modified photoelectrodes exhibit an efficiency of 10.35% versus 9.26% for the same devices with conventional photoelectrodes. Similarly, using the ruthenium dye N719 as sensitizer an efficiency increase from 5.31% to 6.23% is obtained. These improvements originate mainly from an increase in the photocurrent density, which is attributed to an enhanced dye loading obtained with the mesoporous 3D-IO structures due to SiO₂ removal.

Full text of DOI:10.1002/adfm.201706291

FULL PAPER
Solar Cells
www.afm-journal.de
Increasing the Efficiency of Organic Dye-Sensitized Solar
Cells over 10.3% Using Locally Ordered Inverse Opal
Nanostructures in the Photoelectrode
Lin Xu, Cyril Aumaitre, Yann Kervella, Gérard Lapertot, Cristina Rodríguez-Seco,
Emilio Palomares, Renaud Demadrille,* and Peter Reiss*
structures.[1] Since pioneering works by
O’Regan and Grätzel who introduced the
3D inverse opal (3D-IO) oxides are very appealing nanostructures to be
integrated into the photoelectrodes of dye-sensitized solar cells (DSSCs).
Due to their periodic interconnected pore network with a high pore volume
fraction, they facilitate electrolyte infiltration and enhance light scattering.
Nonetheless, preparing 3D-IO structures directly on nonflat DSSC electrodes
use of a mesoporous TiO film in the photo­
2
[
2]
electrode of DSSCs in 1991, research
on this type of solar cells has been contin­
uously and intensively pursued. Recently,
a power conversion efficiency (PCE) of
is challenging. Herein, 3D-IO TiO structures are prepared by templating with
2
1
3% has been achieved with DSSCs using
[
3]
self-assembled polymethyl methacrylate spheres on glass substrates, impreg-
nation with a mixed TiO :SiO precursor and calcination. The specific surface
purely organic dyes, while the highest
value (14.6%) has been obtained by using
a complex mixture of sensitizers and coad­
sorbents in a tandem structure. A DSSC
is typically composed of a dye­sensitized
photoelectrode, a counter electrode, and
an electrolyte serving as a redox shuttle
for regenerating the sensitizer. Accord­
ingly, in the last few years the highest PCE
values of DSSCs were achieved by syn­
2
2
2
−1
increases from 20.9 to 30.7 m g after SiO removal via etching, which leads
2
[
4]
to the formation of mesopores. The obtained nanostructures are scraped from
the substrate, processed as a paste, and deposited on photoelectrodes con-
taining a mesoporous TiO layer. This procedure maintains locally the 3D-IO
2
order. When sensitized with the novel benzothiadiazole dye YKP-88, DSSCs
containing the modified photoelectrodes exhibit an efficiency of 10.35%
versus 9.26% for the same devices with conventional photoelectrodes. Simi-
larly, using the ruthenium dye N719 as sensitizer an efficiency increase from
[
5]
thesizing more efficient sensitizers, by
[
6]
developing specific electrolytes, and by
designing novel mesoporous photoelec­
trodes. The latter can either enhance the
light scattering or increase the electron
mobility to facilitate the charge transport
5.31% to 6.23% is obtained. These improvements originate mainly from an
increase in the photocurrent density, which is attributed to an enhanced dye
loading obtained with the mesoporous 3D-IO structures due to SiO removal.
2
[
7]
to the electrodes.
N­type mesoporous photoelectrodes with a pore size around
1
. Introduction
2
0 nm are generally obtained by sintering TiO nanoparticles
2
Novel strategies to enhance the efficiency in dye­sensitized solar
cell (DSSC) are attracting a renewed interest in this research
topic, which expands from new electrolyte formulations to new
sensitizers and, less explored, the use of photonic inorganic
(NPs) deposited on an appropriate substrate. However, the dis­
ordered particle assembly and tight packing in the film reduce
electron transport and ion diffusion of the redox species from
the electrolyte solution, respectively. In this regard, novel strate­
gies are being proposed for engineering TiO photoelectrodes
2
[
8]
Dr. L. Xu, Dr. C. Aumaitre, Y. Kervella, Dr. G. Lapertot, Dr. R. Demadrille,
and achieving structures of better controlled morphology.
Dr. P. Reiss
University Grenoble Alpes
CNRS, CEA, INAC, SyMMES
3D inverse­opal (3D­IO) photonic crystals are one particular
interesting example of easily accessible, highly ordered
[
9a,b]
materials.
3D­IO nanostructures show a face­centered cubic
1
7 rue des martyrs, F-38000 Grenoble, France
periodic pore network containing interconnected macro­ and
mesopores of controllable porosity with a pore volume frac­
E-mail: renaud.demadrille@cea.fr; peter.reiss@cea.fr
C. Rodríguez-Seco, Prof. E. Palomares
Institute of Chemical Research of Catalonia (ICIQ)
Avenguda Països Catalans, 16, Tarragona 43007, Spain
[
9c]
tion of up to ≈55%. Due to these appealing features, 3D­IO
TiO appears to be a promising candidate for use as a photo­
2
Prof. E. Palomares
ICREA
Passeig Lluís Companys, 23, Barcelona E-08010, Spain
anode in DSSCs. Fully connected pores with extended diam­
eters could facilitate the infiltration of the electrolyte solution
and enhance light scattering, i.e., reduce losses due to photons
nonabsorbed by the sensitizer. Furthermore, 3D­IO structures
can promote electron transport since the highly ordered TiO2
network provides more efficient transport channels compared
The ORCID identification number(s) for the author(s) of this article
can be found under https://doi.org/10.1002/adfm.201706291.
DOI: 10.1002/adfm.201706291
Adv. Funct. Mater. 2018, 1706291
1706291 (1 of 12)
© 2018 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim

www.advancedsciencenews.com
www.afm-journal.de
to conventional electrodes made of sintered NPs. In particular,
this type of macroporous electrodes can be of great benefit for
the development of DSSCs with polymer or gel electrolytes,
which possess a high viscosity.[10]
graphene oxide for improving electronic transport) and 3D­IO
with optimized macropore size as the photoanodes.
[
1e]
Here, we report a new method to prepare photoelectrodes
incorporating locally ordered 3D­IO TiO2 nanostructures,
which overcomes the cited drawbacks. The straightforward
and scalable fabrication procedure of mesoporous 3D­IO TiO2
nanostructures is based on self­assembly of template beads on
a glass substrate, calcination, and etching. The obtained nano­
structures are then scraped from the surface, formulated as a
paste, and deposited by doctor blading for effective screening
and study of different photoanode configurations. The obtained
photoelectrodes were sensitized with a new organic dye (YKP­
88) containing a benzothiadiazole chromophore group. As we
have shown earlier with another benzothiadiazole­based dye
So far, hierarchically porous materials and 3D­IO TiO2
photoanodes are prepared within a self­assembled synthetic
opal template by sol–gel hydrolysis of metal alkoxide precur­
[11]
sors or by infiltration of TiO NPs. The obtained structure
2
is then converted into an inverted opal structure by means of
either thermal or chemical treatment.[
11a–d]
However, these
approaches bear several shortcomings and it remains highly
challenging to obtain a high­quality 3D­IO TiO structure suit­
2
able for use in DSSCs. First, due to the necessity of using a
smooth and clean surface in the self­assembling process of
the 3D opal template, the implementation of standard fabrica­
tion methods for photonic structures on the comparably rough
using a standard TiO photoanode, this family of molecules
2
[
18]
can lead to outstanding PCE values.
For comparison the
blocking layer covering the transparent electrode of DSSCs
well­known commercial ruthenium dye N719 has been used as
a sensitizer, and in all cases the standard iodine/iodide redox
couple was applied in the electrolyte. We demonstrate that
the use of the locally preserved 3D­IO TiO2 nanostructures,
playing the role of a scattering and sensitization layer on top
is often not possible.[
11b,12]
Besides, cracks induced by the
free­volume shrinkage appear inevitably during the template
removal process due to the incomplete infiltration of metal
precursors or TiO NPs into the template. This phenomenon is
2
known to deteriorate the contact with the underlying blocking
of the conventional mesoporous TiO photoelectrode, signifi­
2
layer and to negatively impact the integrity and continuity of
cantly enhances the device performances with both dyes com­
pared to conventional electrodes. Whereas in the case of N719
a PCE increase from 5.31% to 6.23% is achieved, in the case
of YKP­88 a maximum PCE value of 10.35% associated with a
the 3D­IO film.[
10,11b,13]
Therefore strategies have to be devel­
oped to minimize cracks in IO structures. Although efforts
have been made to find alternatives to the sol–gel method by
−
2
using TiO colloidal NPs or atomic layer deposition techniques,
high current density (J ) of 19.86 mA cm was reached. These
2
sc
a shrinkage of 10–30% in volume and crack formation are still
performances rank among the best results reported for DSSCs
with fully organic dyes and iodine­based electrolytes and to the
best of our knowledge, this is the highest PCE ever achieved
observed.[
13,14]
Finally, despite the fact that IO TiO materials
2
reveal a very large surface/mass ratio due to their hollow 3D
structure, the absolute specific surface area for a given film
employing locally ordered inverse opal TiO nanostructures in
2
thickness is lower than in sintered films of TiO NPs. This
the photoanode.
2
effect is a consequence of the larger pore size, on the order of
hundreds of nm for 3D­IO structures, and leads to a decrease
[
10,15]
of the dye loading and hence the efficiency of the DSSCs.
2
. Results and Discussion
For the above mentioned reasons, the reported PCEs of solar
cells containing 3D­IO TiO structures lie in the range of 3–7%
2.1. Preparation of Mesoporous 3D-IO TiO Photoanodes
2
2
when they are directly employed as scaffold for the sensitizer
[
14a,16]
in the working electrode,
in the range of 3–9% when used
The highest efficiency reported so
Figure 1 illustrates the experimental procedure for the fabri­
cation of the mesoporous 3D­IO TiO2 structures developed
in this study. An inverse­opal structure was produced by first
generating a self­assembled 3D opal using a colloidal array of
as a scattering layer.[
1e,12,17]
far (9.2%) has been obtained by Yu and co­workers using four
alternating layers of sintered TiO NPs (with added reduced
2
Figure 1. Schematic diagram of the preparation of mesoporous 3D-IO TiO nanostructures.
2
Adv. Funct. Mater. 2018, 1706291
1706291 (2 of 12)
© 2018 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim

www.advancedsciencenews.com
www.afm-journal.de
monodisperse polymethyl methacrylate (PMMA) spheres (450
nm, Figure S1, Supporting Information) as a sacrificial tem­
plate and next infiltration of the TiO precursor solution (TiCl )
to form the 3D­IO skeleton after a calcination step. With the
goal to increase the specific surface a small amount (3–8 mol%)
of tetraethyl­orthosilicate was added to the TiO2 precursor
solution. This approach implies that the 3D­IO skeletons are
further treated with a 10% hydrofluoric acid (HF) solution to
TiO (Figure S2a–c, Supporting Information) nor in the case
2
of 3D­IO TiO with 5 mol% SiO before etching (Figure S2d–f,
2
2
Supporting Information). The pore size distribution after SiO2
removal lies in the range of 2.5–17.5 nm with an average pore
size of 7.8 nm (Figure 3d). The crystallinity of the as­prepared
2
4
mesoporous 3D­IO TiO structures was investigated by means
2
of high­resolution TEM (HRTEM) and selected­area electron
diffraction (SAED). As shown in Figure 3e, the TiO2 nano­
structures exhibit inter­reticular distances of 2.43 and 3.52 Å,
which correspond to the (101) and (103) planes of anatase TiO2,
respectively. This is further confirmed by SAED (Figure 3f),
which reveals Debye–Scherrer rings characteristic of anatase
remove the SiO part and hence increase the roughness of the
2
walls. A subsequent annealing step at 500 °C affords the final
mesoporous 3D­IO TiO structures.
2
Figure 2 shows scanning electron microscopy (SEM) images
of the as­prepared 3D­IO structures. For comparison, pris­
[
19]
TiO in a wide area. These results indicate that a good crys­
2
tine 3D­IO TiO , only infiltrated with the TiCl4 precursor,
was also fabricated (cf. Figure 1a,b). It reveals a skeleton of
tallinity of the anatase phase is obtained, which is highly ben­
eficial for efficient electron transport.
2
densely packed TiO NPs and regular arrays of interconnected
To estimate the specific surface area, N adsorption–desorp­
2
2
macropores with a diameter of 300 nm. The macropores
shrank by 33% compared to the original PMMA spheres during
the thermal template removal. The mesoporous 3D­IO TiO2
tion isotherms of the pristine 3D­IO TiO and of mesoporous
2
3D­IO TiO nanostructures have been measured on powders
2
obtained after scratching from the surface (Figure S3, Sup­
porting Information). The specific Brunauer–Emmet–Teller
structure (5 mol% SiO removed, Figure 2c,d) shows marked
2
differences: It exhibits a much thinner skeleton with intercon­
(BET) surface areas of pristine 3D­IO TiO and mesoporous
2
2
−1
nected macropores of 320 nm in diameter and the TiO NPs are
3D­IO TiO (5 mol% SiO removed) are 20.9 and 30.7 m g ,
2
2 2
3
−1
packed in a looser way due to the removal of the SiO . This has
with a total pore volume of 0.06 and 0.13 cm g and a pore
diameter of 12.7 and 17.3 nm, respectively. The specific surface
2
important consequences on the specific surface area (vide infra).
To get more detailed structural information of the 3D­IO
structure, transmission electron microscopy (TEM) images
area of the mesoporous 3D­IO TiO calculated using the BET
2
equation is thus 1.5 times higher than that of pristine 3D­IO
were acquired. The mesoporous 3D­IO TiO structure (5 mol%
TiO , its pore volume per mass unit is two times greater, and
2
2
SiO removed) can be clearly observed on the low magnification
furthermore the pore size is effectively enlarged by 36% due to
2
image in Figure 3a. At higher magnification (Figure 3b,c), the
the SiO removal.
2
randomly distributed mesopores can distinguished as lighter
Generally, the increase of the specific surface area of 3D­IO
materials requires the use of smaller template particle sizes
spots on the skeleton of the 3D­IO TiO . We emphasize that
2
[
11a]
these features are found neither in the case of pristine 3D­IO
and porous packing of the NPs in the skeleton.
However,
Figure 2. SEM images of a,b) pristine 3D-IO TiO and c,d) mesoporous 3D-IO TiO (5 mol% SiO removed).
2
2
2
Adv. Funct. Mater. 2018, 1706291
1706291 (3 of 12)
© 2018 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim

www.advancedsciencenews.com
www.afm-journal.de
Figure 3. a–c) TEM images at different magnification of mesoporous 3D-IO TiO (5 mol% SiO removed); d) mesopore size distribution; e) HRTEM
2
2
image; f) selected area electron diffraction (SAED) pattern.
it is more challenging to synthesize monodisperse template
nanospheres (PMMA or polystyrene) with sizes below 200 nm
than for intermediate sizes (300–500 nm). In our approach we
propose another strategy, which does not require the use of
small­sized template NPs. According to the above results, the
during the transfer process. Besides, the SEM images indicate a
very good contact between the two layers, which is in stark con­
trast to the poor contact properties usually observed in conven­
tional direct 3D­IO synthesis methods due to the incomplete
[
10,11b,13]
infiltration of metal precursors.
surface area and pore volume of mesoporous 3D­IO TiO struc­
2
tures were successfully enhanced due to the removal of SiO2
from the initially obtained skeleton. First, as proven by the SEM
images, this step has an influence on the macroskeletal struc­
ture, which becomes much thinner and looser than in the pris­
tine 3D­IO TiO . We attribute this to small SiO NPs that were
2.2. Synthesis and Characterization of the Organic
Photosensitizer
Another key element to develop efficient DSSCs is the dye
itself. Until 2010, high performance devices have mostly been
2
2
randomly formed in the 3D­IO TiO2 skeleton. Moreover, as
[
20]
demonstrated by the TEM images, removal of SiO also induces
fabricated by employing ruthenium complexes, but the lim­
ited natural availability of ruthenium, the usually low molar
2
the formation of additional mesopores in the skeleton, which
further increase the surface area and pore volume.
The mesoporous 3D­IO TiO2 structures were scraped
from the substrate (cf. Figure S4, Supporting Information),
suspended in a paste and deposited by doctor­blading onto
[
21]
extinction coefficients of Ru­complexes and their suspected
toxicity have pushed the scientific community to develop
alternative dyes such as Zn­porphyrin dyes. This class of dyes
showed very competitive performances and great progress has
been made by combining them with cobalt­based electrolytes,
photoelectrodes containing TiO NPs (P25, Solaronix). In this
2
[
22]
configuration, the locally preserved 3D­IO nanostructures can
act as both a scattering layer and additionally as mesoporous
layer for dye sensitization. This procedure bears several advan­
tages over other methods of reporting the 3D­IO structure onto
the photoelectrode: the integrity of the mesoporous 3D­IO TiO2
structure is only kept locally, but the whole film exhibits a good
leading to PCEs up to 13%. However, the synthesis of these
dyes requires multistep synthetic routes with very low overall
yields and currently fully organic sensitizers appear as the most
promising materials related to the mass­market development of
this technology. Indeed they can be prepared and purified in
only few steps, their energy band gap and absorption properties
[
23]
contact with the lower layer of TiO NPs and crack formation is
can be tuned relatively easily, and they can lead to efficient
2
[
24]
prevented. As shown in Figure 4a, the as­prepared layer is con­
and robust solar cells even under harsh operating conditions.
tinuous on a very large scale. When zooming in (Figure 4b,c),
In this context, our group reported in 2015 several dyes
based on a benzothiadiazole chromophore unit that combine
high efficiency and long­term stability in solar cells when ionic
liquid is used as the electrolyte. Following this research axis,
here we report a new organic dye named YKP­88, which is
obtained in 21% yield after only six synthetic steps (Figure 5).
For the preparation of this dye, it is necessary to synthesize the
electron­donating unit (3) consisting of a fused triphenylamine
the locally intact tightly packed 3D­IO TiO structures with big
2
pore size in the layer can be clearly seen. This is further con­
firmed in the cross­section view of the photoanode (Figure 4d).
[
19]
Compared with the underlying mesoporous layer of TiO NPs
2
(
3
≈11 µm in thickness), the upper mesoporous layer made of
D­IO TiO nanostructures (5 µm) appears much more porous
2
due to the inverse opal structure that is maintained locally
Adv. Funct. Mater. 2018, 1706291
1706291 (4 of 12)
© 2018 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim

www.advancedsciencenews.com
www.afm-journal.de
Figure 4. a–c) Top-view, d) cross-section view SEM images of a photoanode containing a ≈11 µm thick layer of TiO2 NPs and a ≈5 µm layer of
mesoporous locally ordered 3D-IO TiO (5 mol% SiO removed). Substrate: fluorine-doped tin oxide (FTO)-coated glass.
2
2
with a thiophene through the formation of an indene ring. The
synthesis of this synthon is achieved in three steps thanks to a
Neigishi coupling of commercially available products followed by
a cyclization reaction and a Buchwald–Hartwig coupling between
diphenylamine and compound (2). The intermediate (3) is sub­
sequently stannylated by action of butyllithium and trimethyl
As shown in Figure 6, the UV–vis spectrum of YKP­88 shows
several absorption bands in the UV region corresponding to
π–π* transitions of the different aromatic and heteroaromatic
rings. The strong absorption feature located in the visible at 529
nm, with a molar absorption coefficient of 28 100 m cm , is
attributed to an internal charge transfer (ICT) band, resulting
in a red­pinkish color in solution. Once the dye is grafted on
−
1
−1
tin chloride and then directly involved in a Stille coupling with
bromo­formylbenzyl­benzothiadiazole.[
25]
From the resulting
TiO , the absorption spectrum of the ICT band is blueshifted
2
compound (5) YKP­88 is obtained through a simple Knoevenagel
reaction. The dye was characterized optically by UV–vis spectros­
shifted by 16 nm and broadened. Such a shift is often observed
with organic dyes and it can be attributed to the deprotonation
copy in solution and after grafting on TiO and electrochemically
of the carboxylic acid function after grafting on the TiO sur­
2
2
by means of cycling voltammetry to determine the position of
the Highest Ocuupied Molecular Orbital (HOMO) and Lowest
Unoccupied Molecular Orbital (LUMO) energy levels.
face whereas the broadening of the spectrum and the disap­
pearance of the discrete features related to the π–π* transitions
can be attributed to the formation of aggregates. The energy
Figure 5. Schematic diagram of the synthesis of YKP-88.
Adv. Funct. Mater. 2018, 1706291
1706291 (5 of 12)
© 2018 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim

www.advancedsciencenews.com
www.afm-journal.de
Figure 6. UV–vis spectra of YKP-88 measured a) in dichloromethane (10−5m) and b) after adsorption on a 2 µm thick TiO mesoporous layer.
2
level position of the frontier orbitals determined by cyclic vol­
tammetry (cf. Table 1) are compatible with those of the conduc­
tion band of the oxide (at ≈ꢀ−4.1 eV) and of the potential of the
The photovoltaic parameters short­circuit photocurrent (JSC),
open­circuit voltage (VOC), fill factor (FF), and PCE of the best
DSSCs sensitized with YKP­88 and the mean values obtained
from five devices are shown in Table 2. Alongside we report
those obtained for devices sensitized using the well­known
ruthenium dye N719 used as a reference. The photocurrent
density–voltage characteristics of the best performing cells
with YKP­88 are shown in Figure 7a and those of DSSCs using
N719 in Figure S5 in the Supporting Information. Importantly,
the mesoporous photoanode with local 3D­IO nanostructures
−
−
[1b]
I /I redox couple used in the electrolyte (at ≈ꢀ−4.95 eV): the
3
ΔEinj (0.6 eV) and ΔEreg (0.35 eV) are large enough to ensure a
good injection of the photoexcited electrons in the oxide and
the regeneration of the photo­oxidized dye molecules by the
redox mediator.
2
.3. Photovoltaic Properties
(5 mol% SiO removed) that revealed the highest porosity shows
2
−2
the best values of JSC (19.86 mA cm ) and V (0.74 V) leading
OC
The locally ordered, mesoporous 3D­IO nanostructures
modified photoelectrodes were immersed in a dyeing bath
constituted of YKP­88 and chenodeoxycholic acid (CDCA)
in acetonitrile/tert­butanol solution. The use of CDCA
was motivated by its capacity to decrease dye aggregation,
to the highest PCE of 10.35%. This value ranks among the
highest reported to date for DSSCs fabricated with an iodine­
based electrolyte. The photovoltaic parameters with the N719
[
26]
which is detrimental for the solar cell performance.
The
Table 2. Photovoltaic parameters of different DSSCs sensitized with
YKP-88 or N719 (highest performances and mean values with standard
deviation over five devices).
sensitized electrodes were assembled with Pt counter elec­
trodes and the devices were filled with a classical iodide/
tri­iodide­based liquid electrolyte. Pristine photoelectrodes
embedding 3D­IO TiO2 nanostructures and conventional
Jsc [mA cm−2
Photoanodes [Dye]
Pristine 3D-IO TiO2
]
Voc [V]
FF [%]
67.9
PCE [%]
8.83
TiO NPs based photoelectrodes (with a thickness of 12 µm
2
18.67
0.70
for the mesoporous layer and 4 µm for the scattering layer,
Solaronix) were also prepared for comparison. To be fully
comparable, the total thickness and area of all the photoan­
odes used in this work were fixed to 16 µm and 0.36 cm ,
respectively.
(
YKP-88)
18.86 ± 0,32
0.68 ± 0,02 67.8 ± 0,9 8.70 ± 0,20
2
Mesoporous 3D-IO TiO2
19.86
0.74
70.6
10.35
(5 mol% SiO2 removed)
19.55 ± 0.40
0.74 ± 0.01 70.7 ± 0.8 10.21 ± 0.30
(
YKP-88)
TiO2 NPs (Solaronix)
YKP-88)
18.29
0.73
69.0
9.26
Table 1. Optoelectronic properties of YKP-88.
(
18.25 ± 0.08
0.73 ± 0,01 68.8 ± 0,5 9.24 ± 0.10
a)
Eg TiO2b) HOMOd) LUMOd)
onset
Dye
λabs [nm] λonseta) Egb) λmax TiO2c)
λ
ε [M 1.cm−1
−
]
[nm]
[nm]
TiO c)
[eV]
2
[eV]
[eV]
[eV]
Pristine 3D-IO TiO2
N719)
12.70
0.59
69.7
5.23
[
nm]
(
12.55 ± 0,15
0.55 ± 0,05 68.3 ± 1.4 5.11 ± 0,20
YKP-88 315 (27800) 623 1.99
371
513
632
1.96
−5.30
−3.50
Mesoporous 3D-IO TiO2
13.43
0,63
73.3
6.23
3
74 (32700)
29 (28100)
(5 mol% SiO2 removed)
13.34 ± 0,11
0.60 ± 0,04 72.3 ± 1.0 6.16 ± 0.07
5
(N719)
a)Measured in dichloromethane (10−5 m); b)Calculated from 1241/λonset; c)Measured
d)
ona2µmthickmesoporouslayerofTiO2; Measuredbycyclicvoltammetryindichlo-
romethane with tetra-butylammonium hexafluorophosphate and using ferrocene/
ferrocenium couple as a reference.
TiO2 NPs (Solaronix)
N719)
11.47
0.62
74.3
5.31
(
11.37 ± 0.10
0.59 ± 0.04 73.1 ± 1.2 5.23 ± 0.08
Adv. Funct. Mater. 2018, 1706291
1706291 (6 of 12)
© 2018 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim

www.advancedsciencenews.com
www.afm-journal.de
effects, such as multiple scattering at disordered regions, Bragg
diffraction, dielectric mirror effects, and resonant modes, which
could enhance the effective optical path of light within the
[27]
active layer. We cannot exclude that some of these effects also
contribute to the enhancement of light absorption in the solar
cells with the mesoporous 3D­IO nanostructures locally intact.
To conclude this section, our novel approach gives access to
modified photoanodes with locally intact mesoporous 3D­IO
TiO nanostructures showing improved solar cell characteris­
2
tics, in particular enhanced photocurrent density. The crucial
importance of the SiO removal step is visible by the compar­
2
ison with pristine 3D­IO modified electrodes, which lead to a
much lower JSC improvement than the mesoporous 3D­IO
modified ones with respect to conventional TiO2 NPs based
electrodes. We explored this point in more detail by preparing
electrodes embedding local 3D­IO TiO2 nanostructures with
3
and 8 mol% SiO . In addition, modified layers obtained by
2
mixing the mesoporous 3D­IO TiO nanostructures with TiO
2
2
NPs (P25) using a weight ratio of 3:1 were also prepared and
tested (Figure S6, Supporting Information). Here again the total
thickness of the photoanodes was fixed at 16 µm. The photocur­
rent density–voltage characteristics are shown in Figure 8a and
Figure 7. a) Photocurrent density–voltage curves (AM 1.5 illumination,
active area 0.36 cm ) for solar cells containing mesoporous local 3D-IO
2
TiO2 (5 mol% SiO2 removed), pristine local 3D-IO TiO , and conven-
2
tional TiO NPs modified photoanodes sensitized with YKP-88 dye. The
2
photographs show the photoanodes modified with b) mesoporous local
3
D-IO TiO2 (5 mol% SiO2 removed), c) pristine local 3D-IO TiO , and
2
d) conventional TiO NPs after dye loading.
2
sensitizer are significantly lower. The photocurrent density only
reaches 13.43 mA cm−2 leading to a maximum PCE of 6.23%.
Globally, both YKP­88 and N719 sensitized solar cells
show very similar trends: the Jsc values of the electrodes with
mesoporous 3D­IO TiO (5 mol% SiO removed) are signifi­
2
2
cantly enhanced compared to electrodes with pristine 3D­IO
TiO nanostructures and the lowest values are obtained with
2
conventional electrodes based on TiO NPs. The highest gain
2
of the mesoporous 3D­IO nanostructures is achieved in the JSC
value, indicating an improvement of the light harvesting proper­
ties of the solar cell. In order to verify this hypothesis, we deter­
mined the dye loading of the different electrodes. We found that
the number of YKP­88 molecules adsorbed on the mesoporous
electrodes is almost the same for devices based on pristine
3
D­IO TiO and on TiO NPs with 195.6 ± 1.3 and 199.0 ±
2 2
2.0 nmol, respectively. However, for devices based on the
mesoporous 3D­IO TiO (5 mol% SiO removed) nanostruc­
2
2
tures, a slightly higher value 208.1 ± 1.4 nmol was found and
the photoelectrode appears a bit darker than the other two types
(Figure 7b–d). These findings can be understood on the basis of
the above­mentioned N adsorption–desorption measurements
2
(
Figure S3, Supporting Information); after SiO removal in the
2
mesoporous 3D­IO TiO structure the specific surface area is
2
enlarged by 46.9% compared to the pristine 3D­IO TiO . Fur­
2
Figure 8. a) Photocurrent density–voltage curves (AM 1.5 illumination,
thermore, the SiO removal also generates additional mesopores
2
2
active area 0.36 cm ) for different mesoporous 3D-IO TiO modified
2
in the skeleton (Figure 3c), which help to further increase the
loading of dye molecules. In addition to the improvement in the
dye loading the locally ordered, mesoporous 3D­IO nanostruc­
tures can also influence the light management in the DSSCs.
Several theoretical approaches report a variety of possible
photoanodes sensitized with YKP-88 dye. Comparison of the obtained
PCE and JSC values (including error bars) of photoelectrodes modified
with mesoporous 3D-IO TiO using different amounts of SiO b) without
2
2
and c) with the addition of TiO NPs (P25) into the IO structure using a
2
weight ratio of 3:1 (IO:TiO NPs).
2
Adv. Funct. Mater. 2018, 1706291
1706291 (7 of 12)
© 2018 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim

www.advancedsciencenews.com
www.afm-journal.de
Table 3. Photovoltaic parameters of DSSCs fabricated using different mesoporous 3D-IO 3D­IO TiO structure obtained with 3 mol%
2
modified photoelectrodes and sensitized with YKP-88 dye (highest performances and mean
values with standard deviation over five devices).
SiO2 (Figure 9a) shows a denser skel­
eton compared to that using 5 mol% SiO2
(
cf. Figure 2c,d), indicating a smaller avail­
Jsc [mA cm−2]
Photoanodes
Voc [V]
FF [%]
PCE [%]
able surface area for grafting dyes. In case
of the 3D­IO obtained with 8 mol% of SiO2
(
Mesoporous 3D-IO TiO2 (3 mol%
SiO2 removed)
19.64
0.72
69.5
9.86
19.52 ± 0.12
0.72 ± 0.02
68.9 ± 0.06
9.81 ± 0.05
Figure 9b), the structure becomes unstable
due to the excess holes after removal of the
large amount of SiO2 from the skeleton,
which makes the whole structure more
fragile.
Mesoporous 3D-IO TiO2 (3 mol%
18.51
0.73
70.7
9.60
SiO2 removed):P25 NPs = 3:1
18.62 ± 0.11
0.73 ± 0.02
70.3 ± 0.4
9.59 ± 0.03
As visible in the top­view images
Mesoporous 3D-IO TiO2 (5 mol%
SiO2 removed)
19.86
0.74
70.6
10.35
(
Figure 9c,d) when using 3 mol% SiO , the
2
19.55 ± 0.40
0.74 ± 0.01
70.7 ± 0.8
10.21 ± 0.3
local 3D­IO nanostructures appears well
intact in the blend similarly to 3D­IO TiO2
Mesoporous 3D-IO TiO2 (5 mol%
19.04
0.73
69.1
9.58
with 5 mol% SiO removed (Figure 3) while
2
SiO2 removed):P25 NPs = 3:1
18.89 ± 0.15
0.73 ± 0.01
69.0 ± 0.1
9.49 ± 0.09
with 8 mol% SiO , the original morphology
2
is not preserved. The more fragile nano­
structures presented in Figure 9b are prob­
ably compressed during the processing. As a
consequence these electrodes lead to a lower
JSC despite the fact that they have a larger
number of mesopores.
Mesoporous 3D-IO TiO2 (8 mol%
SiO2 removed)
19.20
0.73
68.4
9.54
18.89 ± 0.21
0.73 ± 0.01
68.8 ± 0.4
9.46 ± 0.08
Mesoporous 3D-IO TiO2 (8 mol%
18.50
0.73
69.8
9,40
SiO2 removed):P25 NPs = 3:1
18.67 ± 0.17
0.75 ± 0.02
69.0 ± 0.8
9.35 ± 0.05
Finally, to investigate in more detail the
origin of the better solar cell performances
the corresponding photovoltaic parameters are summarized in
Table 3 and compared in Figure 8b,c. This comparison shows
of cells sensitized with YKP­88 as compared to N719, we
measured their photocurrent response (incident photon­to­
electron conversion efficiency, IPCE). From the spectra pre­
sented in Figure 10a it can be evidenced that in the visible
and near ultraviolet wavelength region, i.e., between 350 and
750 nm, YKP­88 is highly efficient and reaches a maximum
IPCE of 80% in nearly the whole range of wavelengths. To the
contrary, N719 shows a much lower IPCE maximum of 65%
and this value is only obtained for a limited wavelength range.
Further understanding of the high photocurrent generation
with YKP­88 and concomitant elevated VOC has been gained
by means of charge extraction (CE) and transient photo­
voltage (TPV) measurements (see Figure 10b,c). It should be
reminded here that the Voc of a DSSC is determined by the
quasi­Fermi­level of the metal oxide semiconductor, which
that the best PCE value of the 3D­IO TiO based devices is
2
obtained with 5 mol% of SiO initially incorporated and subse­
2
quently removed. Both lower (0.3 mol%) and higher (0.8 mol%)
amounts of SiO lead to lower values even though it should
2
be underlined that all the solar cells show high performances
above 9.4%. It appears clearly that the variation of the PCE is
mainly driven by the JSC, as the VOC and FF values are almost
constant for all devices.
When 25 wt% of NPs TiO (P25) were mixed to the different
2
mesoporous 3D­IO TiO2 nanostructures, it can be observed
that the JSC and by consequence the PCE values are slightly
decreased, while otherwise the same trends are observed as for
the other cells, i.e., without added NPs. Figure S7 in the Sup­
porting Information shows the top view SEM images of the
is dependent on its conduction band edge (E ) and electron
c
corresponding modified photoanodes. TiO NPs are filling the
density. The electron density is itself influenced by the rate of
recombination between semiconductor electrons and oxidized
electrolyte species.
2
3
D­IO nanostructures, decreasing the effective pore size in the
whole IO layer. This effect is particularly noticeable in the case
of the mesoporous local 3D­IO nanostructures with 8 mol%
SiO2 removed (Figure S7c, Supporting Information). Moon
and co­workers studied the effect of electrolyte ion diffusion in
Electron lifetimes measured at identical electron densities
using TPV measurements are in good agreement with the
cell voltages listed in Table 2, where longer electron lifetimes
correspond to larger device voltages. The electron lifetime of
YKP­88 is significantly higher than that of N719 indicating
that recombination processes between electrons of the oxide
and redox species of the electrolyte are less prone to occur.
This could be explained by the presence of the lateral alkyl
chains on this dye that may prevent redox species in the
electrolyte to approach the TiO2 surface. Furthermore, the
charge density is not similar in both devices, indicating that
the conduction band potentials of the electrodes are different
after grafting the two different dyes and that its value is lower
with N719.
3
D­IO structures based and conventional P25 based photoelec­
trodes using as here iodine­based electrolytes. The electrolyte/
electrode heterojunction is an important factor determining the
device performance and the study showed that the electrolyte
ion­diffusion coefficient in the 3D­IO structure based photo­
anode was 1.5 times higher than in the P25 electrode.[ Accord­
9]
ingly, in our case, the introduction of TiO NPs may hinder the
2
electrolyte diffusion and thus decreases the performance.
Figure 9 shows SEM images of mesoporous 3D­IO TiO2
structures with 3 and 8 mol% SiO removed and top views
2
of the corresponding photoanodes. The morphology of the
Adv. Funct. Mater. 2018, 1706291
1706291 (8 of 12)
© 2018 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim

www.advancedsciencenews.com
www.afm-journal.de
(cf. Figure S8, Supporting Information), NiO
being an important candidate for the fabri­
cation of p­type photoelectrodes for use in
tandem DSSCs, quantum dot (QD), and pero­
vskite solar cells. Another salient and general­
izable feature of our approach is the efficient
and nondestructive transfer of the obtained
3D­IO nanostructures to the photoelectrode
of interest. Therefore, this study paves the
way for new works aiming at the develop­
ment of more efficient photoelectrodes for
use in DSSCs, perovskite solar cells or photo­
catalytic devices.
4
. Experimental Section
Materials: Titanium (IV) chloride (TiCl4,
≥
99%) was purchased from Fluka Analytical,
ethyl cellulose (5 mol% in toluene/ethanol),
terpineol (anhydrous), and tetraethyl-orthosilicate
(
99.999% trace metals basis) were purchased
Figure 9. SEM images of mesoporous 3D-IO TiO structures obtained with a) 3 and b) 8 mol%
2
from Sigma-Aldrich, microscopy glass slides were
purchased from VWR International France. Platisol
SiO removed and top views of the corresponding photoanodes c) 3 mol% and d) 8 mol%.
2
T, P25 NPs, and standard photoelectrodes were
purchased from Solaronix. Deionized water (R > 18 MΩ) was used for
all experiments.
3
. Conclusion
We have reported a versatile method to synthesize mesoporous
D­IO TiO nanostructures and to incorporate them in photo­
electrodes suitable for application in DSSCs. Their specific sur­
face area and porosity can be strongly increased by adding a
small amount of silica precursor during the synthesis and subse­
Synthesis of Mesoporous 3D-IO TiO2 Structures: The fabrication
process of the 3D-IO TiO2 nanostructures is shown in Figure 1.
3
2
First, monodispersed PMMA spheres with
a diameter around
4
50 nm (Figure S1, Supporting Information) were synthesized from
[28]
methyl methacrylate according to a published method. For the self-
assembling of the PMMA opal film, the glass slides were first cleaned by
immersion into a concentrated H SO solution for at least 2 h, followed
quent removal of the formed SiO domains. After scraping from
2
2
4
the substrate the obtained 3D­IO nanostructures have been inte­
grated into conventional DSSC photoelectrodes while preserving
locally their structural properties. These photoelectrodes have
been sensitized with a new purely organic dye (YKP­88) as well
as with the ruthenium complex dye N719 used as a standard
for comparison. The addition of the locally ordered mesoporous
by ultrasound treatment in deionized water. The monodisperse PMMA
spheres were then self-assembled on the as-cleaned glass substrates
through the vertical deposition process (step A in Figure 1) at 50 °C in
an oven until all the liquid was evaporated. Then the substrates (step B in
Figure 1) were sintered at 120 °C for 40 min aiming to reduce the distance
between the PMMA spheres and enhance the mechanical strength of the
PMMA opal array. Subsequently, the interstitial spaces between the PMMA
3
D­IO nanostructures in the photoelectrodes leads to a signifi­
cant efficiency gain of 17% for the organic dye YKP­88 and of
2% for the ruthenium dye N719 compared to standard DSSC
−3
spheres were filled by immersion in a 4.5 × 10 m TiCl solution (volume
4
ratio of ethanol/deionized water = 1:1) containing 0, 3, 5, or 8 mol% of
tetra-ethyl-orthosilicate (step C in Figure 1). After drying in ambient air
for one day, the films were annealed in a furnace with a ramp of 1 °C
1
electrodes of the same total thickness (16 µm) containing a TiO2
mesoporous layer and a scattering layer. In the case of YKP­88,
the resulting DSSCs employing classical iodine­triiodide liquid
electrolyte reached a maximum value of 10.35% when applying
locally ordered, mesoporous 3D­IO nanostructures obtained
by removing 5 mol% SiO . Both lower and higher SiO con­
−1
min to 500 °C and kept at this temperature for 3 h to completely remove
the polymer spheres by calcination and crystallize the TiO network, then
2
cooled down to room temperature naturally. 3D-IO TiO nanostructures
2
mixed with different amounts of SiO2 (step D in Figure 1, TiO -SiO )
2
2
were obtained. To obtain the mesoporous 3D-IO TiO nanostructure, the
2
samples were scraped from the glass substrates, and etched using a 10%
HF solution for 30 min. This step was followed by washing with ethanol
and centrifugal cleaning for three times (10 000 rpm for 5 min); finally the
obtained powder was dried at 70 °C in an oven.
Synthesis of the Dye YKP-88,86-Bromo-4,4’-bis(4-hexylphenyl)-4H-
indeno[1,2-b] thiophene (2): Under argon, (4-hexylphenyl)-magnesium
bromide was prepared from 1(4-bromophenyl)hexane (2.0 g, 8.29 mmol,
2.5 equiv) and magnesium (202 mg, 8.29 mmol, 2.5 equiv) in freshly
distilled tetrahydrofuran (THF) (10 mL). This reaction mixture was
stirred at reflux for 1 h. In a second flask, methyl 5-bromo-2-(thiophen-
2
2
tents led to slightly lower efficiencies, even though still ranging
between 9.40% and 9.86%. These performances rank among
the best published for this class of devices and 10.35% is the
highest value reported for DSSCs using inverse opal nanostruc­
tures as photoactive component in the photoelectrode. In this
work we have also demonstrated that the improved efficiency
of our devices is mainly related to an enhanced photocur­
rent generation originating from a higher dye loading on the
mesoporous 3D­IO electrodes and a lower rate of recombina­
tion. The presented approach is of great versatility and can be
applied for generating electrodes with 3D inverse opal struc­
tures of various types of metal oxides. As an example we applied
the same procedure to the synthesis of NiO 3D­IO structures
2-yl)-benzoate(1.0g, 3.37mmol, 1.0equiv)wassolubilizedinTHF(15mL).
Then the Grignard reagent was added dropwise and the solution was
heated to reflux for 5 h. After cooling to room temperature, the crude
mixture was poured into water. The organic layer was extracted twice with
ethyl acetate and washed by water and brine, dried over sodium sulfate,
filtered, and concentrated under vacuum overnight. The crude was
Adv. Funct. Mater. 2018, 1706291
1706291 (9 of 12)
© 2018 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim

www.advancedsciencenews.com
www.afm-journal.de
Figure 10. a) IPCE spectra, b) transient photovoltage data showing electron lifetimes versus electron density, and c) charge extraction data showing
electron density as a function of induced voltage for DSSCs containing mesoporous 3D-IO (5 mol% SiO removed) photoelectrodes, sensitized with
2
N719 and YKP-88, respectively.
dissolved in glacial acetic acid (40 mL). After 30 min, concentrated HCL
4 mL) was added dropwise and was refluxed at 120 °C for 5 h. After back
to room temperature, the acetic acid was removed by rotary evaporation,
and the crude was extracted by pentane. The organic layer was washed
several times by water and dried over sodium sulfate, filtered, and
concentrated under vacuum. The crude was chromatographed on silica
2 mol%) and P(o-tolyl)3 (4.4 mg, 14.55 µmol, 4 mol%). The products
were dissolved in anhydrous toluene (20 mL) and refluxed for 24 h. The
mixture was then poured into HCl aqueous solution (2 m). The organic
phase was extracted with methylene chloride, washed with water,
dried over Na SO , and concentrated. The crude solid was purified by
(
2
4
chromatography on silica using methylene chloride/n-hexane, 6:4 as
gel using n-hexane as eluent to afford colorless oil (1) (1.40 g, 73%).
eluent to afford dark red solid (5) (195 mg, 75%).
1
1
H NMR (CDCl , 400 MHz): δ (ppm): 7.47 (d, J = 1.7 Hz, 1 H),
H NMR (CD Cl , 400 MHz): δ (ppm): 10.09 (s, 1 H), 8.19
3
2
2
7
(
4
1
.40 (dd, J = 8.0, 1.8 Hz, 1 H), 7.31 (d, J = 4.9 Hz, 1 H), 7.30
(d, J = 8.3 Hz, 2 H), 8.15 (s, 1 H), 8.02 (d, J = 8.5 Hz, 2 H), 7.97
(d, J = 7.6 Hz, 1 H), 7.81 (d, J = 7.6 Hz, 1 H), 7.42 (d, J = 8.2 Hz, 2 H),
7.30–6.94 (m, 21 H), 2.62–2.48 (m, 4 H), 1.68–1.45 (m, 4 H), 1.40–1.18
d, J = 8.0 Hz, 1 H), 7.08 (d, J = 8.4 Hz, 4 H), 7.04 (d, J = 8.4 Hz,
H), 6.99 (d, J = 4.9 Hz, 1 H), 2.58–2.51 (m, 4 H), 1.62–1.52 (m, 4 H),
13
13
.38–1.23 (m, 12 H), 0.92–0.84 (m, 6 H). C NMR (CDCl , 100 MHz):
(m, 12 H), 0.94–0.79 (m, 6 H). C NMR (CD Cl , 100 MHz): δ (ppm):
3
2
2
δ (ppm): 156.2, 155.5, 141.4, 414.0, 139.7, 136.1, 130.5, 129.3, 128.4,
28.2, 127.4, 122.9, 120.3, 118.9, 35.3, 31.5, 31.1, 28.9, 22.4, 13.9. High
192.0, 156.2, 155.6, 154.1, 152.7, 147.9, 146.9, 143.5, 143.4, 142.03,
141.95, 141.5, 136.1, 131.6, 130.7, 130.0, 129.5, 129.3, 128.6, 128.4,
128.0, 124.6, 124.3, 123.3, 123.1, 122.1, 120.6, 63.6, 35.8, 32.0, 31.8,
29.4, 22.9, 14.2. Anal. Calcd for C H N OS : C, 80.23; H, 6.17; N, 4.68;
1
+
.
Resolution Mass Spectrometry (HRMS) (ESI): [M+H] = 571.2033
7
9
(1 ppm) (calcd. for C H BrS: 571.2029).
35 39
60 55
3
2
Synthesis of the Dye YKP-88,4,4-Bis(4-hexylphenyl)-N,N-diphenyl-4H-
indeno[1,2-b] thiophen-6-amine (3): Under argon, Pd dba (4 mg, 4.37
S, 7.14. Found: C, 80.19; H, 6.06; N, 4.59; S, 6.71.
2
3
Synthesis of the Dye YKP-88,2-Cyano-3-(4-(7-(6-(diphenylamino)-4,4-
bis(4-hexylphenyl)-4H-indeno[1,2-b]thiophen-2-yl)benzo[c][1,2,5]thiadiazol-
4-yl)phenyl)acrylic acid (YKP-88): Under argon, the compound (5)
(180 mg, 200 µmol, 1.0 equiv) and cyanoacetic acid (85 mg, 1 mmol,
5 eq.) were dissolved in a mixture of acetonitrile (9 mL) and chloroform
(9 mL). A catalytic amount of piperidine was added and the solution was
refluxed for 3 h. Solvent was removed under reduced pressure and the
solid was dissolved in chloroform. The organic phase was washed with
HCl aqueous solution (2 M), dried on Na SO , and concentrated. The
µmol, 1 mol%) and tri-tert-butylphosphine tetrafluoroborate (3 mg, 8.75
µmol, 2 mol%) were dissolved with anhydrous toluene (5 mL). After
stirred for 15 min, a solution of the compound (2) (250 mg, 437.3
µmol, 1.0 equiv) and diphenylamine (81.4 mg, 481.0 µmol, 1.1 eq.)
in anhydrous toluene (10 mL) was added. Before refluxed for 48 h,
potassium tert-butoxide was added (161.94 mg, 1.44 mmol, 3.3 eq.) and
the resulting mixture was stirred for 30 min at room temperature. The
crude was filtered through celite and poured into HCl aqueous solution
2
4
(
2 m). The organic phase was extracted with dichloromethane (DCM),
crude solid was purified by chromatography on silica using methylene
chloride first then methylene chloride/MeOH, 90:5 and then methylene
chloride/MeOH/acetic acid, 90:5:5 as eluent to afford the corresponding
dye YKP-88 a dark red solid (179 mg, 93%).
washed with water, dried over Na SO and concentrated. The crude oil
was chromatographed on silica using n-hexane/DCM, 9:1 as eluent to
afford pale yellow oil (3) (266 mg, 92%). H NMR (CDCl , 400 MHz): δ
(
7
(
1
1
2
4
1
3
1
ppm): 7.29 (d, J = 8.2 Hz, 1 H), 7.24–7.16 (m, 6 H), 7.13–7.02 (m, 8 H),
H NMR (THF-d8, 400 MHz): δ (ppm): 8.33 (s, 1 H), 8.31 (d, J = 8.3
.02–6.92 (m, 8 H), 2.57–2.49 (m, 4 H) 1.61–152 (m, 4 H), 1.37–1.24
Hz, 2 H), 8.27 (s, 1 H), 8.21 (d, J = 8.4 Hz, 2 H), 8.12–8.07 (m, 1 H),
8.00–7.95 (m, 1 H), 7.45 (d, J = 8.2 Hz, 1 H), 7.24–7.17 (m, 6 H), 7.15
(d, J = 8.2 Hz, 4 H), 7.08–7.01 (m, 8 H), 7.01–6.95 (m, 3 H), 2.59–2.52
(m, 4 H), 1.63–1.53 (m, 4 H), 1.38–1.27 (m, 12 H), 0.91–0.84 (m, 6 H).
1
3
m, 12 H), 0.92–0.82 (m, 6 H). C NMR (CDCl , 100 MHz): δ (ppm):
3
55.0, 147.5, 145.4, 141.7, 141.0, 131.9, 128.9, 127.9, 127.5, 126.7, 123.7,
23.01, 122.95, 122.5, 119.6, 62.6, 35.3, 31.5, 31.2, 29.5, 28.9, 22.4, 13.9.
+
.
13
HRMS (ESI): [M] = 659.3579 (0 ppm) (calcd. for C H NS: 659.3580).
C NMR (THF-d8, 100 MHz): δ (ppm): 164.2, 157.4, 156.8, 155.0,
4
7
49
Synthesis
bis(4-hexylphenyl)-4H-indeno[1,2-b]thiophen-2-yl)-benzo[c][1,2,5]-thiadiazol-
-yl)benzaldehyde (5): Under argon, the compound (3) (240 mg, 364
µmol, 1.0 eq.) was dissolved in distilled THF (15 mL) then n-BuLi
279 µL, 418 µmol, 1.15 eq.) was added at −78 °C. The solution was
stirred for an hour at −78 °C before adding n-hexane solution of Me SnCl
of
the
Dye
YKP-88,4-(7-(6-(Diphenylamino)-4,4-
154.4, 153.8, 149.1, 147.9, 144.8, 143.3, 142.9, 142.5, 133.0, 132.3, 131.3,
130.9, 130.4, 130.1, 129.4, 129.2, 125.7, 125.5, 124.3, 124.2, 123.5, 121.5,
116.7, 105.1, 64.7, 36.8, 33.1, 33.0, 31.1, 30.5, 23.9, 14.8. HRMS (ESI):
[M]+. = 964.3835 (0 ppm) (calcd. for C63H N O S : 964.3839).
4
56
4
2 2
(
Fabrication of Different DSSC Devices: To prepare the photoanode
containing the different 3D-IO TiO2 nanostructures, pastes were first
prepared according to published work.[ In a typical process, 0.5 g of
3
29]
(
545 µL, 545 µmol, 1.5 eq.) at −78 °C. The solution was allowed to reach
room temperature and stirred for 2 h. The reaction was quenched with
a saturated solution of ammonium chloride and the organic phase
was extracted with n-hexane, watches by water and dried on Na SO ,
filtered, and concentrated under vacuum. The resulting oil was engaged
without any further purification in a Stille coupling with 4-bromo-7-(4-
formylbenzyl)-2,1,3-benzothiadiazole (93 mg, 291 µmmol, 0.8 eq.). This
stannic product was put under argon with Pd dba (6.7 mg, 7.27 µmol,
the 3D-IO TiO powder was first added into 2 mL of ethanol and stirred
2
for 10 min, then the suspension was put into an ultrasonic bath for 3
s (repeated for five times) to make sure that the 3D-IO structures are
well dispersed. In addition, for comparison pastes mixed with 25 wt%
of TiO NPs (P25) were also prepared by mixing 0.125 g P25 TiO NPs
2
4
2
2
with 0.375 g of the 3D-IO TiO2 powders in 2 mL of ethanol. Then,
0.25 g of ethyl cellulose (10 wt%) was added and subsequently 17 µL
2
3
Adv. Funct. Mater. 2018, 1706291
1706291 (10 of 12)
© 2018 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim

www.advancedsciencenews.com
www.afm-journal.de
of acetic acid, and 80 mg of terpineol. Finally, the mixture was stirred
and concentrated on a hot plate until the desired viscosity was reached.
Afterward, the paste was deposited on FTO-coated glass electrodes
containing a 11 µm thick layer of mesoporous TiO2 (Solaronix), and
for the CTQ2013-47183-R project. The authors thank Dipankar Kalita for
carrying out transmission electron microscopy, Dominique Locatelli and
Sébastien Jacquinot for the BET measurements, and Dr. Maxime Godfroy
for his help during the characterization of the YKP-88 dye. This work
was partially funded by the European Union Research Executive Agency
(contract 315131) on the Research for SMEs program. E.P. acknowledges
the ICIQ and ICREA for financial support. R.D. and P. R. acknowledge
the LABEX Laboratoire d’Alliances Nanosciences-Energies du Futur
(LANEF, ANR-10-LABX-51-44001) for funding and the Hybriden facility at
CEA-Grenoble.
the thickness of the 3D-IO TiO film was controlled by the doctor-blade
2
technique to a value of 5.0 µm on a 0.6 cm × 0.6 cm area. After drying at
room temperature, the photoelectrodes were subsequently annealed at
−
1
5
00 °C for 30 min with a heating ramp of 12 °C min in air to improve
the interconnection between the 3D-IO structures and its contact with
the mesoporous TiO2 layer on the substrate as well as to remove any
residual solvents or organic compounds.
Before immersion into the dye solution, the 3D-IO porous TiO2
photoanode was impregnated with a 0.2 m TiCl4 aqueous solution at
Conflict of Interest
7
2
0 °C for 30 min. Then, it was heated to 500 °C in air with a ramp of
−
1
0 °C min and kept at this temperature for 30 min before cooling down
The authors declare no conflict of interest.
to room temperature. The as-prepared electrodes were immersed for
5 h at room temperature into a mixture of CH CN and t-BuOH (volume
1
3
−
3
−3
ratio = 1:1) containing 2 × 10 m CDCA together with 0.2 × 10 m of
Keywords
−
3
−3
YKP-88 or containing 5 × 10 m CDCA together with 5 × 10 m of N719
dye, Ru(dcbpy) (NCS) (N ) (dcbpy = 2,2′-bipyridine-4,4′-dicarboxylic
2
2
3
dye-sensitized solar cells, inverse opals, photoanodes, photovoltaics,
TiO2
acid). The Pt counter electrodes were prepared by drop casting Platisol
T onto washed FTO plates and sintering at 500 °C for 30 min under
air. The electrolyte was composed of 0.1 m LiI, 0.03 m I , 0.5 m, tert-
2
Received: October 30, 2017
Revised: January 4, 2018
Published online:
butylpyridine, 0.5
m 1-butyl-3-methylimidazolium iodide, and 0.1 m
guanidinium thiocyanate in acetonitrile. The sensitized photoanodes
were rinsed with dichloromethane and dried under argon flow. Then they
were assembled and sealed with the counter electrode using thermal
adhesive films (16 µm thick) as a spacer to produce sandwich-type cells.
The electrolyte solution was introduced through a hole drilled in the
counter electrode. Finally, this hole was sealed with a hot-melt film and a
[
1] a) J. W. Gong, K. Sumathy, Q. Q. Qiao, Z. P. Zhou, Renewable Sus-
tainable Energy Rev. 2017, 68, 234; b) T. W. Hamann, R. A. Jensen,
A. B. F. Martinson, H. Van Ryswyk, J. T. Hupp, Energy Environ. Sci.
008, 1, 66; c) B. C. O’Regan, J. R. Durrant, Acc. Chem. Res. 2009,
2, 1799; d) S. Hattori, Y. Wada, S. Yanagida, S. Fukuzumi, J. Am.
Chem. Soc. 2005, 127, 9648; e) Y. Yang, Q. Jin, D. Mao, J. Qi,
Y. Z. Wei, R. B. Yu, A. R. Li, S. Z. Li, H. J. Zhao, Y. W. Ma, L. H. Wang,
W. P. Hu, D. Wang, Adv. Mater. 2017, 29, 1604795.
2] B. Oregan, M. Gratzel, Nature 1991, 353, 737.
3] Z. Y. Yao, H. Wu, Y. Li, J. T. Wang, J. Zhang, M. Zhang, Y. C. Guo,
P. Wang, Energy Environ. Sci. 2015, 8, 3192.
4] K. Kakiage, Y. Aoyama, T. Yano, K. Oya, J. Fujisawa, M. Hanaya,
Chem. Commun. 2015, 51, 15894.
[5] a) B. Pashaei, H. Shahroosvand, M. Graetzel, M. K. Nazeeruddin,
Chem. Rev. 2016, 116, 9485; b) R. Sivakumar, R. Recabarren,
S. Ramkumar, A. Manivel, J. Alzate-Morales, D. Contreras,
P. Manidurai, New J. Chem. 2017, 41, 5605; c) Y. Z. Wu,
M. Marszalek, S. M. Zakeeruddin, Q. Zhang, H. Tian, M. Gratzel,
W. H. Zhu, Energy Environ. Sci. 2012, 5, 8261.
2
cover glass. The active area of all cells is 0.36 cm .
Photovoltaic and Photoelectrical Measurements: Photovoltaic
measurement of the DSSCs was recorded by applying an external
potential bias to the cells while recording the change of generated
current with a digital source meter (Keithley model 2400). The cells were
measured under irradiation using 1000 W m air-mass (1.5) simulated
solar light generated by an Oriel class AAA solar simulator. IPCE was
measured as a function of wavelength from 300 to 800 nm. A 75 W xenon
lamp was used as the light source for generating a monochromatic
beam. Calibration was performed using an National Institute of
Standards and Technology (NIST)-calibrated silicon photodiode as a
standard. IPCE values were collected at low chopping speed.
Other Types of Characterization: The morphology of the samples
was inspected using a ZEISS-ultra 55 field emission scanning electron
microscope (SEM) with accelerating voltage of 5 kV. Transmission
electron microscope (TEM) and high-resolution TEM (HRTEM) images
were recorded on a JEM-3010 transmission electron microscope using
a working voltage of 300 kV, equipped with a LaB6 gun and a Gatan
Orius SC 200 2 k × 2 k Charge Coupled Device (CCD) camera. The
specific surface areas of the 3D-IO samples were investigated using the
Brunauer–Emmett–Teller (BET) method (BEL Japan, Inc.).
2
4
−
2
[
[
[
[
6] a) J. N. de Freitas, A. F. Nogueira, M. A. De Paoli, J. Mater. Chem.
009, 19, 5279; b) H. Wang, J. Li, F. Gong, G. Zhou, Z. S. Wang,
2
J. Am. Chem. Soc. 2013, 135, 12627; c) M. Wang, X. Pan, X. Q. Fang,
L. Guo, W. Q. Liu, C. N. Zhang, Y. Huang, L. H. Hu, S. Y. Dai, Adv.
Mater. 2010, 22, 5526.
7] a) M. S. Ahmad, A. K. Pandey, N. Abd Rahima, Renewable Sus-
tainable Energy Rev. 2017, 77, 89; b) P. Ilaiyaraja, T. K. Das,
P. S. V. Mocherla, C. Sudakar, Sol. Energy Mater. Sol. Cells 2017, 169,
Supporting Information
[
Supporting Information is available from the Wiley Online Library or
from the author.
8
6; c) J. J. Wu, W. P. Liao, M. Yoshimura, Nano Energy 2013, 2, 1354.
[
8] a) Y.-J. Cheng, P. Müller-Buschbaum, J. S. Gutmann, Small 2007, 3,
1
3
379; b) Y. Park, J. W. Lee, S. J. Ha, J. H. Moon, Nanoscale 2014, 6,
105; c) G. Kaune, M. Memesa, R. Meier, M. A. Ruderer, A. Diethert,
Acknowledgements
S. V. Roth, M. D’Acunzi, J. S. Gutmann, P. Müller-Buschbaum, ACS
Appl. Mater. Interfaces 2009, 1, 2862.
9] a) E. Armstrong, C. O’Dwyer, J. Mater. Chem. C 2015, 3, 6109;
b) G. von Freymann, V. Kitaev, B. V. Lotsch, G. A. Ozin, Chem. Soc.
Rev. 2013, 42, 2528; c) S. J. Ha, J. H. Moon, J. Phys. Chem. C 2017,
121, 12046.
L.X. benefits from a scholarship of the “Fondation Franco-Chinoise pour la
Science et ses Applications” (FFCSA). The authors acknowledge financial
support from the French Research Agency ANR (grants ANR ANR-14-
OHRI-0003 ODYCE and ANR-16-CE05-0019-02 Persil) and ICIQ-FPI
grant (BES-2014-068795) and ICREA. E.P. and C.R.S. are also grateful to
Spanish Ministry of Economy and Competitiveness program (MINECO)
[
Adv. Funct. Mater. 2018, 1706291
1706291 (11 of 12)
© 2018 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim

www.advancedsciencenews.com
www.afm-journal.de
[
10] J. H. Shin, J. H. Kang, W. M. Jin, J. H. Park, Y. S. Cho, J. H. Moon,
Langmuir 2011, 27, 856.
[19] M. Adachi, Y. Murata, J. Takao, J. T. Jiu, M. Sakamoto, F. M. Wang,
J. Am. Chem. Soc. 2004, 126, 14943.
[
11] a) B. T. Holland, C. F. Blanford, A. Stein, Science 1998, 281, 538;
b) Y. G. Seo, K. Woo, J. Kim, H. Lee, W. Lee, Adv. Funct. Mater. 2011,
[20] M. Gratzel, Inorg. Chem. 2005, 44, 6841.
[21] C. Y. Chen, M. K. Wang, J. Y. Li, N. Pootrakulchote, L. Alibabaei,
C. H. Ngoc-le, J. D. Decoppet, J. H. Tsai, C. Gratzel, C. G. Wu,
S. M. Zakeeruddin, M. Gratzel, ACS Nano 2009, 3, 3103.
[22] A. Yella, H. W. Lee, H. N. Tsao, C. Y. Yi, A. K. Chandiran,
M. K. Nazeeruddin, E. W. G. Diau, C. Y. Yeh, S. M. Zakeeruddin,
M. Gätzel, Science 2011, 334, 629.
2
2
1, 3094; c) A. Stein, B. E. Wilson, S. G. Rudisill, Chem. Soc. Rev.
013, 42, 2763; d) M. C. Orilall, N. M. Abrams, J. Lee, F. J. DiSalvo,
U. Wiesner, J. Am. Chem. Soc. 2008, 130, 8882; e) X.-Y. Yang,
L.-H. Chen, Y. Li, J. C. Rooke, C. Sanchez, B.-L. Su, Chem. Soc. Rev.
2
017, 46, 481; f) I. Concina, Z. H. Ibupoto, A. Vomiero, Adv. Energy
Mater. 2017, 7, 1700706.
12] A. Mihi, C. J. Zhang, P. V. Braun, Angew. Chem., Int. Ed. 2011, 50, 5711.
[23] Y. K. Eom, S. H. Kang, I. T. Choi, Y. J. Yoo, J. H. Kim, H. K. Kim,
J. Mater. Chem. A 2017, 5, 2297.
[
[
13] Y. G. Seo, M. A. Kim, H. Lee, W. Lee, Sol. Energy Mater. Sol. Cells
011, 95, 332.
14] a) E. S. Kwak, W. Lee, N. G. Park, J. Kim, H. Lee, Adv. Funct. Mater.
009, 19, 1093; b) L. J. Liu, S. K. Karuturi, L. T. Su, A. I. Y. Tok,
[24] D. Joly, L. Pelleja, S. Narbey, F. Oswald, J. Chiron, J. N. Clifford,
E. Palomares, R. Demadrille, Sci. Rep. 2014, 4, 4033.
[25] A. S. D. Sandanayaka, Y. Taguri, Y. Araki, T. Ishi, S. Mataka, O. Ito,
J. Phys. Chem. B 2005, 109, 22502.
2
[
2
Energy Environ. Sci. 2011, 4, 209; c) Z. Z. Yang, S. M. Gao, W. Li,
V. Vlasko-Vlasov, U. Welp, W. K. Kwok, T. Xu, ACS Appl. Mater. Inter-
faces 2011, 3, 1101.
[26] J. H. Yum, S. J. Moon, R. Humphry-Baker, P. Walter, T. Geiger,
F. Nuesch, M. Graetzel, M. D. K. Nazeeruddin, Nanotechnology
2008, 19, 424005.
[
15] a) C. Y. Cho, J. H. Moon, Langmuir 2012, 28, 9372; b) J. W. Lee,
J. Lee, C. Kim, C. Y. Cho, J. H. Moon, Sci. Rep. 2014, 4, 6804.
16] a) S. H. Kang, H. S. Kim, J. Y. Kim, Y. E. Sung, Mater. Chem. Phys.
[27] a) G. Lozano, S. Colodrero, O. Caulier, M. E. Calvo, H. Miguez,
J. Phys. Chem. C 2010, 114, 3681; b) A. Mihi, H. Miguez, J. Phys.
Chem. B 2005, 109, 15968; c) R. Rengarajan, D. Mittleman, C. Rich,
V. Colvin, Phys. Rev. E 2005, 71, 016615; d) S. Colodrero, A. Mihi,
J. A. Anta, M. Ocaña, H. Miguez, J. Phys. Chem. C 2009, 113, 1150;
e) S. Colodrero, A. Forneli, C. Lopez-Lopez, L. Pelleja, H. Miguez,
E. Palomares, Adv. Funct. Mater. 2012, 22, 1303; f) S. Colodrero,
A. Mihi, L. Häggman, M. Ocaña, G. Boschloo, A. Hagfeldt,
H. Míguez, Adv. Mater. 2009, 21, 764.
[28] Y. Zhu, W. Xu, H. Zhang, W. Wang, S. Xu, H. Song, J. Phys. Chem. C
2012, 116, 2297.
[29] S. Ito, P. Chen, P. Comte, M. K. Nazeeruddin, P. Liska, P. Pechy,
M. Gratzel, Prog. Photovoltaics 2007, 15, 603.
[
2
010, 124, 422; b) H. N. Kim, J. H. Moon, ACS Appl. Mater. Inter-
faces 2012, 4, 5821; c) H. N. Kim, H. Yoo, J. H. Moon, Nanoscale
013, 5, 4200.
2
[
17] a) M. Jin, S. S. Kim, M. Yoon, Z. Li, Y. Y. Lee, J. M. Kim, J. Nanosci.
Nanotechnol. 2012, 12, 815; b) S. Guldin, S. Hüttner, M. Kolle,
M. E. Welland, P. Müller-Buschbaum, R. H. Friend, U. Steiner,
N. Tétreault, Nano Lett. 2010, 10, 2303.
[
18] D. Joly, L. Pelleja, S. Narbey, F. Oswald, T. Meyer, Y. Kervella,
P. Maldivi, J. N. Clifford, E. Palomares, R. Demadrille, Energy
Environ. Sci. 2015, 8, 2010.
Adv. Funct. Mater. 2018, 1706291
1706291 (12 of 12)
© 2018 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim