Triarylamine-based hydrido-carboxylate rhenium(i) complexes as photosensitizers for dye-sensitized solar cells

Article abstract of DOI:10.1039/c9cp00856j

Two new dyes based on a dinuclear rhenium complex and (E)-3-(5-(4-(bis(2′,4′-dibutoxy-[1,1′-biphenyl]-4-yl)amino)phenyl)thiophen-2-yl)-2-cyanoacrylic acid (namely D35) have been investigated as sensitizers for dye sensitized solar cells (DSSCs). Two different pyridazine ligands have been used, namely 4-pyridazine-carboxylic acid for dye 2 ([Re₂(μ-H)(-D35)(CO)₆(μ-pyridazine-4-COOH)]) and 4-pyridazinyl-butanoic acid for dye 3 ([Re₂(μ-H)(-D35)(CO)₆(μ-pyridazine-4-C₃H₆-COOH)]). The performances of these new dyes have been compared with those of the dye containing the bare 4-diphenylaminobenzoic acid, namely TPA, as the ancillary ligand (dye 1). Compared to dye 1, dyes 2 and 3 show an impressive tenfold increase in the absorption intensity in the range of 487-493 nm on TiO₂ films, with great improvement of the light harvesting. Cyclic voltammetry experiments, performed on derivatives containing the methyl ester of the pyridazine ligands, show narrow electrochemical band gaps in the range of 1.36-1.84 eV. Solar cells with each dye have been prepared, using both iodide/triiodide and cobalt redox couples as the electrolytes, platinum or carbon as the counter electrodes, and TiO₂ or SnO₂ as the metal oxide photoelectrodes, respectively. The best DSSC results have been obtained using dye 3, with an overall solar-to-electric conversion efficiency of 3.5%, which greatly overcomes the previous result of 1.0% obtained for dye 1 in a not-optimized setup of the device. The performances of dye 3 are due to the presence of D35 ligand, which further suppresses the recombination of the injected electron with the electrolyte and with the oxidized state of the dye.

Full text of DOI:10.1039/c9cp00856j

PCCP
View Article Online
View Journal
PAPER
Triarylamine-based hydrido-carboxylate rhenium(I)
complexes as photosensitizers for dye-sensitized
solar cells†
Cite this:DOI: 10.1039/c9cp00856j
ab
a
c
Lorenzo Veronese, Elsa Quartapelle Procopio, Thomas Moehl,
Monica Panigati,
ab
d
d
*
Kazuteru Nonomura* and Anders Hagfeldt
0
0
0
Two new dyes based on a dinuclear rhenium complex and (E)-3-(5-(4-(bis(2 ,4 -dibutoxy-[1,1 -biphenyl]-
-yl)amino)phenyl)thiophen-2-yl)-2-cyanoacrylic acid (namely D35) have been investigated as sensitizers
for dye sensitized solar cells (DSSCs). Two different pyridazine ligands have been used, namely 4-pyridazine-
carboxylic acid for dye 2 ([Re (m-H)(-D35)(CO) (m-pyridazine-4-COOH)]) and 4-pyridazinyl-butanoic acid for
dye 3 ([Re (m-H)(-D35)(CO) (m-pyridazine-4-C -COOH)]). The performances of these new dyes have been
4
2
6
2
6
3 6
H
compared with those of the dye containing the bare 4-diphenylaminobenzoic acid, namely TPA, as the
ancillary ligand (dye 1). Compared to dye 1, dyes 2 and 3 show an impressive tenfold increase in the
2
absorption intensity in the range of 487–493 nm on TiO films, with great improvement of the light harvesting.
Cyclic voltammetry experiments, performed on derivatives containing the methyl ester of the pyridazine
ligands, show narrow electrochemical band gaps in the range of 1.36–1.84 eV. Solar cells with each dye have
been prepared, using both iodide/triiodide and cobalt redox couples as the electrolytes, platinum or carbon as
Received 12th February 2019,
Accepted 13th March 2019
2 2
the counter electrodes, and TiO or SnO as the metal oxide photoelectrodes, respectively. The best DSSC
results have been obtained using dye 3, with an overall solar-to-electric conversion efficiency of 3.5%, which
greatly overcomes the previous result of 1.0% obtained for dye 1 in a not-optimized setup of the device. The
performances of dye 3 are due to the presence of D35 ligand, which further suppresses the recombination of
the injected electron with the electrolyte and with the oxidized state of the dye.
DOI: 10.1039/c9cp00856j
rsc.li/pccp
4
5
is 12.1% with long-term stability, while an efficiency as high as
3% was achieved using a molecularly engineered porphyrin dye
Introduction
1
1
6
Dye sensitized solar cells (DSSCs) are hybrid nanostructured in combination with a cobalt-based electrolyte.
organic/inorganic cells that have attracted great attention so far
The most commonly used dyes in DSSCs are complexes of
from both the academic field and industry, due to the use of ruthenium, containing organic ligands functionalized with
7
abundant, cheap materials and simple, low cost fabrication carboxylic substituents as anchoring groups to the TiO
2
. These
processes that can be applied to both rigid (glass) and flexible poly-pyridine complexes are characterized by intense and broad
2
3
(
plastics, metals) substrates and easily scaled up.
absorption bands due to metal-to-ligand charge transfer
In its simplest assembly, a DSSC is an electrochemical cell (MLCT) transitions, longer excited state lifetimes, and long-
8
with two electrodes and an electrolyte filled in the space term chemical stability. Alternative dyes based on different
9
10
11
between them. The highest efficiency reported for a small area metal complexes, such as iron, copper and platinum
DSSC based on ruthenium dye and an iodide/triiodide electrolyte mainly containing bipyridine ligands, have been investigated
1
2
so far, while no example of rhenium complexes used as a dye
in operating solar cells has been reported in the literature. In
order to cover this gap and following some preliminary studies
concerning the electron transfer from Re-polypyridyl complexes
a
b
c
Universit a` degli Studi di Milano, Dipartimento di Chimica, Via Golgi 19,
0133 Milano, Italy. E-mail: monica.panigati@unimi.it
Istituto per lo Studio delle Macromolecole, Consiglio Nazionale delle Ricerche
ISMAC-CNR), Via E. Bassini, 15, 20133 Milano, Italy
2
1
3
(
to nano-crystalline TiO2, we recently reported the first exam-
ple of hydrido rhenium complexes employed in an operating
Department of Chemistry, University of Zurich, Winterthurerstrasse 190, CH-8057,
Z u¨ rich, Switzerland
Laboratory for Photomolecular Science, ISIC, E´ cole Polytechnique F ´e d ´e rale de
14
DSSC. The best results have been obtained for the hydrido-
d
carboxylato complex containing a triarylamine moiety (1, see
Chart 1), which suppresses the recombination of the injected
electron with the oxidized state of the complex, thus improving
Lausanne, CH-1015 Lausanne, Switzerland. E-mail: kazuteru.nonomura@epfl.ch
†
Electronic supplementary information (ESI) available. See DOI: 10.1039/
c9cp00856j
This journal is ©the Owner Societies 2019
Phys. Chem. Chem. Phys.

View Article Online
Paper
PCCP
Chart 1 Molecular structure of the investigated dyes used as sensitizers.
2
the charge separation on TiO . Moreover, the bulky triarylamine
These new dyes have been characterized from the spectro-
moiety also reduces the back reaction of the injected electron scopic and electrochemical points of view. Solar cells with each
with the electrolyte. Even if the maximum power conversion dye have been prepared and their performances have also been
efficiency is about 1%, these results have provided a sufficiently reported, using both iodide/triiodide and cobalt redox couples
clear picture of the critical factors and of the lines to follow to as the electrolytes, platinum or carbon as the counter electrodes,
improve the performances.
and TiO (and partially SnO ) as the metal oxide photoelectrode.
2 2
Therefore, in order to follow these lines, we report here on the The performances have been compared to those obtained for
design and preparation, based on a novel and versatile three-step D35 and discussed in relation to the molecular structure of
procedure, of two new hydrido-carboxylate dinuclear organo- the dyes and the previous results obtained for dye 1 in a not-
metallic rhenium complexes, namely 2 and 3 (see Chart 1) with optimized setup of the device.
0
0
0
the more conjugated TPA linker (E)-3-(5-(4-(bis(2 ,4 -dibutoxy-[1,1 -
biphenyl]-4-yl)amino)phenyl)thiophen-2-yl)-2-cyanoacrylic acid
a.k.a. DN-F04 dye (from now on D35, see Chart 1). Interest in
organic molecules, characterized by a D–p–A electronic struc-
ture, as dyes in DSSCs has been growing quickly in the past
Results and discussion
Synthesis of the dyes
decade, due to their unlimited availability from feedstock, The hydrido-carboxylato derivatives 2 and 3 have been prepared
improved prerequisites for scalability, recycling issues and by exploiting the same three-step synthetic procedure already
14
economy on a large scale. The absorption spectrum of D35 developed for the synthesis of the TPA-based derivative 1. This
makes it a reference organic dye for DSSCs. Excellent device procedure, shown in Scheme 1, is based on the peculiar reactivity
stability has been obtained both under light and after at least of the tetrahedral hydrido-carbonyl cluster [Re
4
(m
3 4
-H) (CO)12] with
1
5,16
1000 h of storage in the dark at a temperature of 85 1C.
It is bridging donor ligands such as 2,5-diphenyl-1,3,4-oxadiazole
also interesting for indoor applications due to its good match (ppd), a sterically hindered heterocycle which contains two
with fluorescent light and its attractive orange color. Therefore, aromatic substituents in the a positions of the oxadiazole
1
8
we linked D35 as a bridging ligand to the dinuclear rhenium ring. In the first step, the [Re (m-H) (CO) (m-ppd)] inter-
2
2
6
complexes, to extend the light harvesting capacity and to allow mediate has been obtained in high yields by reacting
their usage in cobalt-based electrolytes. [Re (m -H) (CO)12] with two equivalents of ppd at room tem-
Looking at the anchoring group, one of the reasons for the perature in CH
observed low efficiency of the previously reported complexes is the carboxylato complexes [Re
excessive stabilization of the p* orbitals of the 4-carboxy-pyridazine, have been obtained by reaction (1), which exploits the reac-
4
3
4
1
8
2
Cl
2
solution. Then, the mixed hydrido-
(m-H)(m-OOCR)(CO) (m-ppd)] (Scheme 1)
2
6
14
which partially hampers the electron injection into TiO2. In order tivity of the hydride ligands towards the acidic proton of any
to overcome this problem, we report here a new dye (3 in Chart 1) carboxylic acid. Through a simple acid–base reaction, one of
containing a diazine ligand endowed with an alkyl chain. In fact, the two hydrides is removed as H and the carboxylate anion
2
1
4,19
the presence of electron-donor substituents in the b positions of takes its place, bridging the two metal centers.
the diazine ligand raises the LUMO level of the dye and then different carboxylic acids have been used for the synthesis of
Three
17
fosters the electron injection into the semiconductor.
the dyes and, in particular, 4(diphenylamino)benzoic acid
Phys. Chem. Chem. Phys.
This journal is ©the Owner Societies 2019

View Article Online
PCCP
Paper
Scheme 1 Synthetic pathways for the preparation of the rhenium hydrido-carboxylato dyes 1–3.
(
TPA-COOH) for dye 1 and the carboxylic acid of D35 for which is centered on the diazine ligand and whose chemical
dyes 2 and 3.
reversibility is confirmed by the presence of its anodic counter-
part. In contrast, for dye 3, the reduction peak is observed at
[
Re (m-H)(m-OOCR)(CO) (m-ppd)]
2
6
ꢀ
1.47 V, i.e. at the same reduction potential of the complex
2
0
+
RCOOH [Re (m-H)(m-OOCR)(CO) (m-ppd)] + H
[Re (m-Cl) (CO) (m-4-n-hexylpydz)] (ꢀ1.46 V) containing a pyr-
2
6
2
2
2
6
(1)
idazine ligand functionalized at one of the b positions with a
similar aliphatic chain. Also, in this case, the reduction is
chemically and electrochemically reversible. It is interesting
to note that the peak potential is slightly modulated by the
different nature of the ancillary ligand, in agreement with the
different localization of the ground-state energy levels.
Finally, substitution of the oxadiazole with properly function-
alized diazine ligands (namely the pyridazinyl-carboxylic acid for
dyes 1 and 2 and the pyridazinyl-4-butanoic acid for dye 3,
respectively) afforded the desired products.
All these features are in agreement with the localization
of the reduction process on the diazine ring, as also indicated
by the description of the LUMO provided by the DFT
Electrochemical characterization
1
7,20
Cyclic voltammetry analysis has been performed on all the computations.
Therefore, the LUMO level results stabilized,
derivatives 1-COOMe, 2-COOMe and 3-COOMe containing the and partially delocalized, due to the presence of the carboxylic
methyl ester of the corresponding diazine ligand, in order to group at the b position, as in the dyes 1 and 2. On the other
avoid the reduction of the acidic proton and H
cyclic voltammetry behavior in acetonitrile, at 298 K, is shown presence of the electron-donor alkyl chain, as in the case of 3.
in Fig. 1 and the peak potentials, together with the related In the anodic part, dyes 1–3 display a close sequence of three
2
evolution. The hand, a partial destabilization of the LUMO is observed in the
HOMO and LUMO energy values, are reported in Table 1. In the monoelectronic oxidation peaks. The two oxidation processes,
cathodic region, dye 2, containing the same diazine ligand of 1, observed at the highest potentials, are clearly localized on the
displays the same behavior as that reported previously: namely metal core. This attribution is confirmed both by the computa-
+
one monoelectronic reduction peak at about ꢀ0.9 V (vs. Fc |Fc), tional analysis and by comparison with the electrochemical
This journal is ©the Owner Societies 2019
Phys. Chem. Chem. Phys.

View Article Online
Paper
PCCP
potential is observed (see Table 1 and Fig. 1). This peak is
clearly attributed to the formation of a radical cation on the
triarylamine (TPA) moiety, as indicated by the comparison with
the electrochemical behavior of the free 4-diphenylamino-
1
4
benzoic acid (Ep,a = 0.70 V) and by the modulation of the peak
potential by the substituents on the TPA core. Indeed, dyes 2 and
3, containing a more conjugated TPA moiety, display a lower
oxidation potential than dye 1 (+0.66 V vs. +0.41 V, see Table 1).
Spectroscopic characterization
The UV-Vis absorption maxima and molar absorptivities are
reported in Table 2 both in solution and upon adsorption on
TiO , while Fig. 2 shows all the spectra in toluene solution. In
2
solution, all the dyes exhibit two broad and featureless absorp-
tion bands extending in a conspicuous part of the visible
spectrum between 300 nm and 600 nm.
The low energy absorption band is ascribable to the metal-
to-ligand dp(Re) - *p(diazine) charge transfer transition
1
(
MLCT), in agreement with the DFT computations and by
1
7
comparison with the analogous complexes of this family.
This broad MLCT band arises from the convolution of multiple
transitions, as testified by the more-or-less pronounced
shoulders observed at longer wavelengths. This attribution is
supported by their solvatochromic behavior and by the presence
of the predicted red-shift in the absorption maximum for the
D35-based dyes 2 and 3, compared to the bare TPA-based dye 1.
In addition, dyes 2 and 3 show an impressive tenfold increase in
the value of e compared to 1, therefore, greatly improving the
light harvesting. This is due to a partial superposition between
the MLCT band and the absorption band of the D35 ligand,
Fig. 1 Normalized CV features of the methyl–ester derivatives of dyes
ꢀ1
1
–3 on GC electrodes, in ACN + 0.1 M TBAPF
6
solution, at 0.2 V s with
ohmic drop compensation.
behavior already observed in the derivatives containing alcoholate, which displays an absorption band at 500 nm in CH Cl solution
2
ꢀ1
2
21
14
ꢀ1 16
phenolate and benzoate anions as ancillary ligands. Moreover, with a high molar extinction coefficient (31 300 M cm ).
these oxidation potentials are lower for dyes 2 and 3 than for dye 1
All the dyes also display another higher energy absorption
(+0.75 V vs. +1.07, see Table 1), and this is due to the presence of band around 335–350 nm, whose position is completely inde-
the strong electron-withdrawing acrylonitrile moiety bound to the pendent of the polarity of the solvent. In agreement with the
coordinating carboxylate group. These data still confirm that electrochemical and computational data already reported for
the oxidation process is markedly different in the case of the
oxygen-bridged derivatives compared to that observed for
1
7
22
halide or chalcogenide ones. This behaviour could be related Table 2 UV-Vis MLCT absorption data of the dyes in solution and on TiO
2
films
to the hard–soft nature of the bridging ancillary ligands. Indeed,
the softer halides or chalcogenides anions can better stabilize
the cationic products, favoring the simultaneous loss of two
electrons instead of two mono-electronic oxidations, as observed
in the case of the harder oxygen-based anions.
a
ꢀ1
ꢀ1
b
c
Complex
l
max /nm
e/M cm
l
max /nm
lmax /nm
3
1
2
3
483
493
486
5.2 ꢁ 10
425
486
487
443
487
493
4
3.1 ꢁ 10
4
3.5 ꢁ 10
Besides the two oxidation peaks centered on the metal
a
ꢀ5
b
ꢀ5
c
Toluene solution (2 ꢁ 10 M). EtOH solution (2 ꢁ 10 M). TiO
2
ꢀ
1
ꢀ1
core, another chemically reversible oxidation peak at a lower film. Absorption data for D35: 31 300 M cm at 500 nm.
a
Table 1 The first reduction and oxidation peak potentials (Ep,c and Ep,a) and electrochemical (DE
e
) and spectroscopic (DE
S
)
energy gaps of the dyes.
+
b
ꢀ1
Potentials are referred to the Fc |Fc couple in the operating medium (ACN, 0.1 M TBAPF
6
). Scan rate: 0.2 V s
Complex p,c [V] p,a [V] DE [eV]
E
E
e
DE [eV]
s
ELUMO [eV]
EHOMO [eV]
1
2
3
-COOMe
-COOMe
-COOMe
ꢀ0.97
ꢀ0.94
ꢀ1.42
0.66, 1.07, 1.32
0.41, 0.74, 1.10
0.42, 0.75, 1.10
1.64
1.36
1.84
2.56
2.52
2.55
ꢀ3.81
ꢀ3.85
ꢀ3.37
ꢀ5.45
ꢀ5.21
ꢀ5.21
a
1
The spectroscopic (DE
S
) energy gap is the energy associated with the electronic transition determined from the maximum of the MLCT
b
+
absorption band. Fc |Fc potential is 0.385 V vs. SCE in ACN solution.
Phys. Chem. Chem. Phys.
This journal is ©the Owner Societies 2019

View Article Online
PCCP
Paper
Fig. 3 Current–voltage curves for the optimized devices sensitized by
rhenium-based dyes 1–3 under 1 Sun illumination, AM 1.5.
Fig. 2 UV-Vis absorption spectra of dyes 1–3 in toluene solution.
1
4
dye 1, this band is attributed to the p–p* transition involving electrolyte: one based on the iodide/triiodide redox shuttle, and
2
+/3+
the triarylamine moiety. Noticeably, for dyes 2 and 3, contain- the other employing Co
metal complexes. Each electrolyte
ing the more conjugated D35 ligand, this band is moderately has been coupled with a suitable counter electrode: platinum
blue-shifted, as a consequence of the presence of two butox- for iodine and carbon for cobalt. Both the electrolytes have
yphenyl electron donating units on the TPA moiety of D35.
been developed in-house, the first one was labeled IE and the
latter one CE.
Devices
The compositions of both iodine-based and cobalt-based
electrolytes have been varied to obtain the best current–voltage
compromise. We have used cells with pure D35 as photosensi-
tizers as a reference for this comparison. Solid-state devices
(SSDs) have also been prepared to investigate if these dyes may
hold potential for this kind of devices. However, the generally
low efficiencies detected prevented us from carrying out further
investigations.
Photovoltaic measurements have been carried out to evaluate
the potential of the new rhenium complexes as dyes in DSSC
devices. The main photovoltaic parameter performances of the
solar cells under AM 1.5 G at 1 sun (1000 W m ) illumination
are presented in Table 3 and the current–voltage ( J–V) curves
are displayed in Fig. 3.
Sets of cells have been prepared and measured, changing
one parameter at a time, in order to find the best operating
conditions for each of these parameters. Dye 1, already tested in
an optimized device, has also been re-tested after the engi-
neering of optimized cells. In particular, different semiconduc-
2
The values related to the performances of the cells, reported
in Table 3, clearly indicate that a careful engineering of the cells
has afforded a new maximum performance of 1.8% for dye 1,
1
4
1
4
nearly doubling the previous value obtained for such a dye.
2
3
A substantial increase in the photogenerated currents can be
observed (see Table 3) for dyes 2 and 3 compared to dye 1, due
to the presence of the highly light-absorbing D35 moiety. This
is also confirmed by the incident photon-to-current conversion
efficiency (IPCE) plots. Indeed, the monochromatic IPCE spectra,
displayed in Fig. 4, show that the highest IPCE is obtained
tor oxides (TiO
2
, SnO
2
)
have been tried in single or stacked-
layer architecture, along with their relative thickness. The dyes
were adsorbed on the semiconductors in different media to
find out the best solvent and the adsorption time. At the same
time, various counter electrode glasses with different sheet
resistances have been tested with our dyes. The nature of the
electrolytes has also been intensively investigated and the
behavior of each dye has been tested in two different types of
Table 3 Photovoltaic parameters for optimized and solid-state cells
sensitized by 1–3 compared to D35 using homemade iodine based (IE)
and cobalt (CE) electrolytes
ꢀ
2
CELL
J
SC/mA cm
VOC/V
FF
Z/%
1
1
1
2
2
2
3
3
3
IE/Pt
CE/C
SSD
IE/Pt
CE/C
SSD
IE/Pt
CE/C
SSD
ꢀ4.5
ꢀ1.4
ꢀ0.5
ꢀ7.9
ꢀ9.1
ꢀ0.4
ꢀ8.8
ꢀ7.6
ꢀ1.1
0.51
0.47
0.52
0.52
0.54
0.46
0.54
0.65
0.48
0.88
78
72
57
51
64
53
67
71
63
72
1.8
0.5
0.15
2.05
3.0
0.35
3.15
3.5
0.32
7
Fig. 4 IPCE spectra for optimized devices sensitized by rhenium-based
dyes 1–3.
D35 CE/C
ꢀ11
This journal is ©the Owner Societies 2019
Phys. Chem. Chem. Phys.

View Article Online
Paper
PCCP
from the solar cell sensitized by 3 which, however, is not
as red as 2.
The external efficiency values, however, remain well below
those obtained under similar conditions by cells assembled
Due to the introduction of D35, the new dyes appear to work using the conventional D35 dye, reaching 7% (CE).
best under CE/C conditions, while dye 1, with a simple triaryl- Even though dyes 2 and 3 have a similar or even wider light
amine remains more efficient under the IE/Pt condition. The absorption compared to D35, this superior property is not
performances of dye 2 is only slightly better than those of 1 under reflected in the device efficiency. This is the case for dye 2 that
IE conditions, affording 2%, while, under CE conditions, a value has the smallest energy gap (see Table 1). Dye 2 can harvest the
of 3% in efficiency can be achieved. The best performing cells are photons the most among the tested dyes. However, due to
sensitized by dye 3, giving an overall power conversion efficiency several reasons (discussed below), dye 2 did not show the best
of 3.15% (IE) and 3.5% (CE). The recombination between the properties, even for photocurrent. This can be explained by the
electrons in TiO and the oxidized dye or between the electrons in desorption of the dye and it is also confirmed by electrochemical
2
TiO and the electrolyte are crucial factors for the performances of impedance spectroscopy (EIS) results.
2
the cells. Introduction of the D35 moiety in dyes 2 and 3 clearly
contributes to further suppression of the former recombination, scopy, it was found that the charge transfer resistance at the
thus improving the charge separation on TiO . Moreover, the counter electrode for dye 2 is higher than that for other samples
higher steric hindrance of D35 also reduces the back reaction (data not shown). This implies the desorption of dye 2 from the
between the injected electron and the electrolyte. The recombina- surface of TiO and then, their adsorption on the surface of
From the analysis after electrochemical impedance spectro-
2
2
tion process with the oxidized dye is further suppressed in dye 3, the counter electrode. Such phenomena can explain one of the
in which the presence of the less electron-withdrawing diazine reasons for the lower fill factor for the sample with dye 2.
affords a gain both in JSC and VOC, leading to an efficiency of 3.5%. Moreover, the desorption of dye molecules from the surface of
Indeed, by replacing the direct carboxylic moiety with the TiO causes a negative influence on the solar cell properties,
2
4
-butanoic acid derivative, the electron injection into the conduc- such as reduction in light harvesting and an increase of the
24
tion band of TiO
2
becomes retarded, as proved by Durrant, et al.,
2
TiO surface area that is exposed to the electrolyte (Fig. 5).
to minimize the kinetic redundancy. This adjustment prolongs
the electron lifetime in the photoelectrode before recombining considerable amounts of Li in the electrolyte should also be
Together with the desorption of the dye, the presence of
+
+
with the oxidized dye and/or cation in the electrolyte.
taken into account. Indeed, the presence of Li shifts the
Fig. 5 Left: Resistance vs. potential (solid curves: Rct; solid straight lines: Rtrans: dotted curves: C) and right: electron lifetimes vs. capacitance (solid lines:
lifetime; dotted lines: transport time) plots for devices sensitized by dyes 1 (orange line), 2 (brown line) and 3 (red line) under IE/Pt conditions.
Fig. 6 Left: Resistance vs. potential (solid line: Rct; dotted line: C) and right: electron lifetimes vs. capacitance plots for devices sensitized by dyes 1
(
orange line), 2 (brown line) and 3 (red line) under CE/C conditions.
Phys. Chem. Chem. Phys.
This journal is ©the Owner Societies 2019

View Article Online
PCCP
Paper
conduction band of TiO
2
more for devices with Co as the to the reduced recombination and an optimization of the kinetic
electrolyte than those with iodide. This feature also affects the retardation. However, there are still challenges, such as dye-
open current voltage (VOC) which is calculated by the difference aggregation, thus requiring further tuning of the HOMO–LUMO
between the Fermi level of the TiO and the redox potential of energy level for realizing high-efficiency dye-sensitized solar cells.
2
the electrolyte. For the cells with the Co redox couple, the VOC
As the last notable point, the three-step synthetic procedure
increases on going from 1 to 3 and the difference in the VOC of developed here has a wider scope than the preparation procedure
the cells with iodide is the smallest for complex 2. This is in for the compounds described here and opens the way for a large
agreement with the desorption of dye 2, which leaves a larger number of dinuclear complexes of this family, tailored for specific
+
surface area exposed to the electrolyte allowing, therefore, the Li
to come to the surface of TiO
band of TiO for the sample with dye 2 improves the electron
applications.
2
. The lowering of the conduction
2
lifetimes since the driving force between the conduction band of Experimental
TiO2 and the electrolyte is reduced and hence the transport
efficiency, which are the best for dye 2.
Materials and methods
All the reactions are performed under a nitrogen atmosphere
unless specified otherwise. The reagents are purchased from
Aldrich, Fluka and Lancaster and used as received. D35 is
2
purchased from Dyenamo, washed with H O and then dried
In the case of the cobalt-based electrolyte, the EIS shows a
similar tendency (Fig. 6). The conduction band of 2 is once
again the lowest among all the complexes, while 3 exhibits
a slightly higher one compared to 1, as expected after the
introduction of a non-conjugated chain. The lifetimes are, in
the case of the cobalt electrolytes, relatively similar between the
three different dyes, but also in this case, complex 2 shows the
longest one.
under vacuum before use. All the solvents are deoxygenated
and dried by standard methods before use; toluene is distilled
on Na(s) and CH Cl on P O , both under a N atmosphere.
2
2
2
5
2
Commercial deuterated solvents are used as received. Column
chromatography is performed using Alfa Aesar silica gel 60
2
5
(
0.032–0.063 mm). [Re
4
(m
3
-H)
4
(CO)12](C
6 6 2
H ) ,
[Re
2
(m-H)
2
(CO)
6
-
1
8
(m-ppd)] (ppd = 2,5-diphenyl-1,3,4-oxadiazole) and [Re
2
(m-H)-
Conclusions
(
m-4-OOC-TPA)(CO)
6
(m-pydz-4-COOH)] (1) (4-OOC-TPA = 4-(diphe-
nylamino)benzoate anion pydz-4-COOH = pyridazine-4-carboxylic
In our previous work, quantomechanical calculations using
density functional theory (DFT) and time-dependent density
functional theory (TDDFT) have been coupled with UV/vis
absorption spectroscopy and preliminary photovoltaic device
testing. Based on these results, herein we have designed and
synthesized new dirhenium complexed dyes. We have analyzed
their electronic/optical properties and characterized the nature
14
acid) are synthesized according to literature procedures. The IR
spectra in solution are acquired using a Bruker Vector 22 FT
spectrophotometer. The solution and thin-film UV/Vis absorp-
tion spectra are obtained using an Agilent 8453 spectro-
photometer, fused quartz cuvettes (10 mm optical path), and
CHROMASOL V grade solvents.
2
of their interaction with the TiO substrate. Dye-sensitized solar
cells were prepared and tested in operating solar cells. Light-to-
current conversion, which greatly overcomes the previous result
Synthesis
Synthesis of [Re (l-H)(l-D35)(CO) (l-ppd)]. 40 mg (0.046 mmol)
2
6
of 1.0% obtained for dye 1 in a not-optimized setup of the of solid D35 are added to a solution of 32 mg (0.042 mmol) of
device, has been observed for all the dyes here investigated. [Re (m-H) (CO) (m-ppd)] previously dissolved in 5 mL of anhydrous
2
2
6
These interesting results are mainly due to the introduction of toluene. The reaction mixture is left stirring under reflux overnight;
the triarylamine-based D35 as a carboxylate ligand, which then, the solution is evaporated to dryness under vacuum. The
further suppresses the recombination of the injected electron obtained solid is purified by column chromatography (eluent
2 2
with the oxidized state of the dye. This wider light-harvesting CH Cl /hexane 75 : 25), then dried under vacuum, yielding 30 mg
and bulkier triarylamine moiety also reduces the back-reaction (0.018 mmol) of a brick-red powder (yield 44%). IR (CH Cl ) n(CO):
2
2
ꢀ
1 1
of the injected electron with the electrolyte. The best results 2042 (s), 2024 (vs), 1937 (vs), 1917 (s) cm . H NMR: (CD Cl ,
2
2
have been obtained for the hydrido-carboxylato dye 3, with a 300 K, 400 MHz) d (ppm) 8.18 (d, J = 7.5 Hz, 4H, H
ppd), 7.90
H
ortho
power conversion efficiency of 3.5%. This feature resulted by (m, 2H, Hpara ppd), 7.79 (m, 4H, Hmeta ppd), 8.16 (s, 1H), 7.75 (d, J =
combining molecular design with optimization of the cell. 4.2 Hz, 1H), 7.60 (d, J = 8.88 Hz, 2H), 7.52 (d, J = 8.47 Hz, 4H), 7.37
Indeed, the problems concerning the lower efficiency of the (d, J = 4.2 Hz, 2H), 7.29 (d, J = 9.3 Hz, 2H), 7.21 (d, J = 8.47 Hz, 4H),
other two dyes, 1 and 2, have been represented by the excessive 7.17 (d, J = 8.88 Hz, 2H), 6.59 (s, 2H), 6.58 (m, 2H), 4.02 (t, J = 6.98 Hz,
2 2 2
stabilization of the p* orbitals of the diazine and by the 4H, CH ), 4.01 (t, J = 6.98 Hz, 4H, CH ), 1.79 (m, 8H, CH ) 1.57 (m,
desorption of the dye. These critical factors have been overcome 4H, CH ), 1.50 (m, 4H, CH ), 1.03 (m, 6H, CH ), 0.97 (m, 6H, CH ),
2
2
3
3
by the use of a more electron-rich diazine ligand, contained in ꢀ7.14 (s, 1H, hydride). Elemental anal. calcd for C H N O SRe :
74
68
4
12
2
dye 3, which provides a higher LUMO level and a better anchor- C, 54.67; H, 4.22; N, 3.45. Found: C 55.31, H 4.83, N 3.27.
ing group on the TiO surface. In summary, the presence of the Synthesis of [Re (l-H)(l-D35)(CO) (l-pydz-4-COOH)] (2).
bulkier triphenylamine moiety and the use of a more electron- Once 15 mg (0.009 mmol) of [Re (m-H)(m-D35)(CO) (m-ppd)]
rich ligand successfully improved the photovoltaic efficiency due are dissolved in 5 mL of freshly distilled and degassed THF,
2
2
6
2
6
This journal is ©the Owner Societies 2019
Phys. Chem. Chem. Phys.

View Article Online
Paper
.5 mg (0.01 mmol) of 4-pyridazine-carboxylic acid are added to
PCCP
1
Electrochemical measurements. A cyclic voltammetry study
the reaction mixture. The reaction is set at 85 1C for 2 hours. of the complexes has been performed at scan rates typically
ꢀ
1
The solution is evaporated to dryness under vacuum and the ranging from 0.02 to 10 V s
in HPLC-grade acetonitrile
M concentration for each
ꢀ
4
crude product dissolved in CH Cl and re-precipitated with (MeCN) solutions at 2.5 ꢁ 10
2
2
n-hexane. The supernatant solution is removed and the remain- substrate, deaerated by N bubbling with 0.1 M TBAPF (TBA =
2
6
ing powder is washed with hexane (3 mL) five times and dried NBu
under vacuum affording 11 mg (0.072 mmol) of the desired drop was compensated by the positive feedback technique. The
product (yield 80%). IR (CH Cl ) n(CO): 2040 (m), 2020 (s), 1938 experiments were carried out using an AUTOLAB PGSTAT
s), 1920 (m) cm , H NMR: (CD Cl , 300 K, 400 MHz) d (ppm) potentiostat (EcoChemie, The Netherlands) run by a PC with a
.75 (s, 1H, H3pydz-4-COOH), 9.51 (d, J = 5.6 Hz, 1H, H6pydz-4-COOH), GPES software. The working electrode was a glassy carbon one
.39 (dd, J = 4.7, 2.9 Hz, 1H, H5pydz-4-COOH), 8.13 (s, 1H), 7.73 (AMEL, diameter 1.5 mm) cleaned with diamond powder
4
, Aldrich) as the supporting electrolyte, at 298 K. The ohmic
2
2
ꢀ
1 1
(
2
2
H
9
8
(m, 1H), 7.62 (m, 2H), 7.51 (m, 4H), 7.36 (m, 2H), 7.27 (m, 2H), (Aldrich, diameter 1 mm) on a wet cloth (STRUERS DP-NAP);
7
6
CH
6
.22 (m, 4H), 7.18 (m, 2H), 6.59 (s, 2H), 6.57 (m, 2H), 4.02 (t, J = the counter electrode was a platinum wire; the reference elec-
.98 Hz, 4H, CH ), 4.01 (t, J = 6.98 Hz, 4H, CH ), 1.79 (m, 8H, trode was an aqueous saturated calomel electrode, with a
) 1.53 (m, 4H, CH ), 1.49 (m, 4H, CH ), 1.03 (t, J = 7.4 Hz, difference of 0.385 V vs. the Fc |Fc couple (the intersolvent redox
H, CH ), 0.97 (t, J = 7.4 Hz, 6H, CH
), ꢀ6.75 (s, 1H, hydride). potential reference currently recommended by IUPAC) in our
Elemental anal. calcd for C65 14SRe : C, 51.10; H, 4.09; N, working medium.
.67. Found: C 52.09, H 4.42, N 3.19. DSSC and SSD preparation. The solution-processed and the
Synthesis of [Re (l-H)(l-D35)(CO) (l-4-pyridazinyl-butanoic solid-state devices were made using 18 NR-T and 30 NR-D
2
2
+
2
2
2
3
3
H
62
N
4
O
2
3
2
6
acid)] (3). 15 mg (0.009 mmol) of [Re (m-H)(m-D35)(CO) (m-ppd)] titanium oxide (TiO ) paste, respectively, obtained from Dyesol.
2
6
2
are dissolved in 5 mL of freshly distilled and degassed THF; Lithium iodide (LiI), lithium bis(trifluoromethane)sulfonimide
mg (0.012 mmol) of 4-pyridazinyl-butanoic acid are added to (LiTFSI), iodine, tert-butyl pyridine (TBP), and guanidinium
2
0
0
the reaction mixture. The reaction is maintained under reflux for thiocyanate (GuSCN) were purchased from Aldrich, and 2,2 ,7,7 -
2
0
hours. The solution is evaporated to dryness under vacuum, tetrakis-(N,N-di-4-methoxyphenylamino)-9,9 -spirobifluorene
the crude product dissolved in CH
2
Cl
2
, and re-precipitated with (Spiro-OMeTAD) was obtained from Solaronix. [Co(bpy)
3 4 2
(B(CN) ) ]
n-hexane. The supernatant solution is removed and the remain- and [Co(bpy)
3
(B(CN) ] were synthesized at EPFL. The homemade
4 3
)
ing powder is washed with hexane (3 mL) five times, then dried substrates for liquid-containing cells have a size of approximately
2
under vacuum affording 11 mg (0.007 mmol) of a red powder 1.4 cm ꢁ 1.6 cm, with an active area of 0.159 cm . The device
(
(
(
H
7
yield 78%). IR (CH Cl ) n(CO): 2042 (m), 2024 (s), 1943 (s), 1921 structure consists of a transparent fluorine doped tin oxide (FTO)
2 2
ꢀ1 1
m) cm , H NMR: (CD
d, J = 5.9 Hz, 2H, H64-pydz-BuCOOH), 9.21 (d, J = 5.9 Hz, 2H, substrate. After being cleaned and having undergone UV-O
34-pydz-BuCOOH), 8.12 (s, 1H), 7.78 (d, J = 2.0 Hz, 1H, H54-pydz-BuCOOH), treatment for 15 minutes, the FTO glass plates were immersed
.74 (d, J = 4.1 Hz, 1H), 7.60 (d, J = 8.88 Hz, 2H), 7.52 (d, J = 8.6 Hz, in a 40 mM aqueous TiCl solution at 70 1C for 35 min and then
H), 7.36 (d, J = 4.1 Hz, 2H), 7.28 (m, 2H), 7.21 (d, J = 8.8 Hz, 4H), washed with water and ethanol. Two screen-printed layers of
.17 (d, J = 7.4 Hz, 2H), 6.59 (s, 2H), 6.58 (d, J = 6.8 Hz, 2H), 4.02 nanocrystalline TiO particles were used as the photoelectrode.
2 2 H
Cl , 300 K, 400 MHz) d (ppm) 9.22 layer as the bottom electrode, supported on an NSG TEC C10 glass
3
4
4
7
2
(
t, J = 6.98 Hz, 4H, CH ), 4.01 (t, J = 6.98 Hz, 4H, CH ), 1.79 (m, The transparent mesoporous layer (tl), made of 18 nm sized TiO2
2
2
8
H, CH ) 1.56 (m, 4H, CH ), 1.47 (m, 4H, CH ), 1.03 (m, 6H, particles (Dyesol DSL18NR-T), was printed on the FTO conducting
2
2
2
CH
calcd for C68
3
), 0.97 (m, 6H, CH
3
), ꢀ6.65 (s, 1H, hydride). Elemental anal. glass, and the plates were then heated at 150 1C for 7 min. To
68 4
H N O14SRe
2
: C, 52.02; H, 4.37; N, 3.57. Found: C obtain a high PCE, an B5 mm scattering layer (sl, 400 nm
53.64, H 4.95, N 3.10.
diameter, Catalysts & Chemicals Ind. Co. Ltd (CCIC), HPW-400)
Synthesis of the methyl ester of pydz-4-COOMe and complexes was deposited on the transparent layer. A total film thickness
1-3)-COOMe. A sample of pyridazine-4-carboxylic acid (100 mg, of 7(tl) + 5(sl) mm was used. Sintering was carried out following a
.805 mmol) (for 1 and 2) or pyridazine-4-butanoic acid (134 mg, 4-step temperature ramp from 175 to 500 1C, with a residence
.805 mmol) (for 3) was dissolved in MeOH (2 mL) and was treated time of 30 min. The sintered glass plates were immersed again in
with H SO (40 mL, 96% w/w) at room temperature. The reaction 20 mM aqueous TiCl solution at 70 1C for 35 min and washed
(
0
0
2
4
4
mixture was heated at reflux temperature and was stirred overnight. with water and ethanol. The films were heated again at 500 1C for
Then, the solution was cooled at room temperature and the 30 min using a heat blower followed by cooling to 90 1C and
reaction was quenched by the addition of a saturated solution of dipping into 0.3 mM toluene solution of the dyes overnight at
2 3 2
Na CO up to pH 8. The product was extracted with Et O and the room temperature. To prepare the counter-electrode, the Pt
organic fractions were collected, washed with brine, dried with catalyst was deposited on a cleaned TEC 15 FTO glass by coating
Na SO and evaporated to dryness to afford a pale yellow solid with a drop of H PtCl solution (5 mM in 2-PrOH solution) with
2
4
2
6
1
(
isolated yield 30%). H NMR (DMSO, 300 K, 400 MHz) d (ppm) subsequent rapid thermal treatment at 400 1C for 15 min. For
H
9
(
.58 (s, 1H, Hortho), 9.51 (d, 1H, Hortho), 8.09 (dd, 1H, Hmeta), 3.94 carbon counter-electrodes the same procedure was followed using
s, 3H, CH ). This pyridazine was used to synthesize the corres- a graphite/acetone solution. For the assembly of DSSCs contain-
ponding derivatives 1-COOMe, 2-COOMe and 3-COOMe according ing a liquid electrolyte, the dye-containing TiO electrode and the
counter-electrode were assembled into a sandwich-type cell and
3
2
to the previous procedures.
Phys. Chem. Chem. Phys.
This journal is ©the Owner Societies 2019

View Article Online
PCCP
Paper
sealed with a hot-melt gasket with 25 mm thickness made of the potentiostat with a Zview software, at open circuit, both in the
ionomer Surlyn 1702 (Dupont). The redox electrolyte was driven dark and under illumination. Spectra were analyzed with a Zview
into the cells through two holes previously drilled in the counter- equivalent circuit modeling software, including the distributed
electrode. The iodine-based type of electrolyte used for rhenium- element DX11 (transmission line model). All the measurements
based dyes consists of 0.7 M LiI, 0.025 M I and 0.2 M TBP in ACN. were performed one day after cell preparation.
2
The cobalt-based type of electrolyte contained instead 1 M LiTFSI,
0
3 4 2 3 4 3
.33 M [Co(bpy) (B(CN) ) ], 0.06 M [Co(bpy) (B(CN) ) ] and 0.2 M
Conflicts of interest
TBP in ACN. Finally, the hole was sealed using Surlyn and a cover
glass (0.1 mm thickness). Solid state devices (SSDs) were
prepared following a similar procedure. An FTO glass plate
with 1.4 ꢁ 2.3 cm was laser etched to create a non-conductive
There are no conflicts to declare.
zone for the separation of the contacts, then it was cleaned Acknowledgements
and UV/O treated before undergoing flame spray pyrolysis. A
3
The use of instrumentation purchased through the Regione
Lombardia – Fondazione Cariplo joint SmartMatLab Project
2013-1766) is gratefully acknowledged.
solution containing 0.6 mL of titanium isopropoxide acetylace-
tonate [Ti(OiPr) (acac) ], 0.4 mL of acetylacetone and 9 mL of
ethanol was then sprayed (gas carrier: O , 0.8 bar) on the glass
plates at 450 1C, creating a non-porous transparent TiO layer.
2
2
(
2
2
The transparent mesoporous layer, in this case made of 30 nm
sized TiO2 particles (Dyesol DSL30NR-D), was printed with a
different mesh on the FTO conducting glass, followed by a no
scattering layer. Sintering was carried out as described above.
The titania films were then dipped into 0.3 mM toluene solution
of the dyes overnight at room temperature, then dried, and
transferred into a glove box. A solution of the hole transporting
material (HTM), Spiro-OMeTAD, was spin coated onto the
adsorbed dye. The final thickness of the whole cell was mea-
sured to be around 2 mm. The counter-electrode, metallic gold,
was deposed via physical vapor deposition, without sealing the
cell before measurements.
Notes and references
1 B. O’Regan and M. Gr ¨a tzel, Nature, 1991, 353, 737.
2 T. M. Brown, F. De Rossi, F. Di Giacomo, G. Mincuzzi, V. Zardetto,
A. Reale and A. Di Carlo, J. Mater. Chem. A, 2014, 2, 10788.
3 A. Fakharuddin, R. Jose, T. M. Brown, F. Fabregat-Santiago
and J. Bisquert, Energy Environ. Sci., 2014, 7, 3952.
4 M. Nazeeruddin, F. De Angelis, S. Fantacci, A. Selloni,
G. Viscardi, P. Liska, S. Ito, B. Takeru and M. Gr ¨a tzel,
J. Am. Chem. Soc., 2005, 127, 16835–16847.
5 D. Kuang, S. Ito, B. Wenger, C. Klein, J. Moser, R. Humphry-
Baker, S. Zakeeruddin and M. Gr ¨a tzel, J. Am. Chem. Soc.,
2006, 128, 4146–4154.
6 S. Mathew, A. Yella, P. Gao, R. Humphry-Baker, B. F. E.
Curchod, N. Ashari-Astani, I. Tavernelli, U. Rothlisberger,
Md. K. Nazeeruddin and M. Gratzel, Nat. Chem., 2014, 6,
242–247.
7 (a) M. Nazeeruddin, A. Kay, I. Rodicio, R. Humphry-Baker,
E. M u¨ ller, P. Liska, N. Vlachopoulos and M. Gr ¨a tzel, J. Am.
Chem. Soc., 1993, 115, 6382–6390; (b) K. Kalyanasundaram
and M. Gr ¨a tzel, Coord. Chem. Rev., 1998, 177, 347–414.
8 S. Campagna, F. Puntoriero, F. Nastasi, G. Bergamini and
V. Balzani, Top. Curr. Chem., 2007, 280, 117–214.
9 (a) S. Ferrere and B. A. Gregg, J. Am. Chem. Soc., 1998, 120,
843–844; (b) P. Balraju, M. Kumar, M. S. Roy and
G. D. Sharma, Synth. Met., 2009, 159, 1325–1331.
DSSC and SSD characterization. All measurements were
carried out in air directly after the fabrication of the cells. A
black shadow mask with a fixed aperture was used on DSSCs
2
and SSDs, so that the active area was set to be 0.16 cm . The
current–voltage characteristics were recorded by applying an
external potential bias to the cell while recording the generated
photocurrent using a digital source meter (Keithley model
2400) connected to a pc. The light source was a 450 W xenon
lamp (Oriel) equipped with a Schott K113 Tempax sunlight
filter (Pr ¨a zisions Glas & Optik GmbH) to match the emission
spectrum of the lamp with the AM 1.5 G standard. Before each
measurement, the exact light intensity was determined using a
calibrated Si reference diode equipped with an infrared cutoff
filter (Schott KG-3). The incident photon to collected electron
conversion efficiency (IPCE) measurement was plotted as a 10 (a) T. Bessho, E. C. Constable, M. Gr ¨a tzel, A. H. Redondo,
function of wavelength by using light from a 300 W xenon lamp
ILC Technology), which was focused through a Gemini-180
double monochromator (Jobin Yvon) onto the photovoltaic cell
under test. A computer-controlled monochromator equipped
with automatic grating and wavelength selection and operating
C. E. Housecroft, W. Kylberg, M. K. Nazeeruddin, M. Neuburger
and S. Schaffner, Chem. Commun., 2008, 3717–3719; (b) C. L.
Linfoot, P. Richardson, T. E. Hewat, O. Moudam, M. M. Forde,
A. Collins, F. White and N. Robertson, Dalton Trans., 2010, 39,
8945–8956.
(
in the spectral range (300–800 nm) generates a photocurrent 11 (a) W. Wu, X. Xu, H. Yang, J. Hua, X. Zhang, L. Zhang,
action spectrum with a sampling interval of 10 nm and a current
sampling time of 4 s to reduce scattered light from the edge of
the glass electrodes of the dyed TiO2 layer. Electrochemical
Y. Longa and H. Tian, J. Mater. Chem., 2011, 21,
10666–10671; (b) E. A. M. Geary, L. J. Yellowlees, L. A.
Jack, I. D. H. Oswald, S. Parsons, N. Hirata, J. R. Durrant
and N. Robertson, Inorg. Chem., 2005, 44, 242–250.
ꢀ
1
ꢀ6
impedance spectroscopy (10 mV steps in the 10 –10
Hz
range) was carried out only on the transparent layer-only version 12 B. Bozic-Weber, E. C. Constable and C. E. Housecroft,
of the already assembled DSSC cells, using a BioLogic SP-300
Coord. Chem. Rev., 2013, 257, 3089–3106.
This journal is ©the Owner Societies 2019
Phys. Chem. Chem. Phys.

View Article Online
Paper
PCCP
1
3 (a) J. B. Asbury, E. Hao, Y. Wang and T. Lian, J. Phys. Chem. 20 The reduction potential is ꢀ1.34 V for the parent complex
B, 2000, 104, 11957–11964; (b) N. A. Anderson, X. Ai,
D. Chen, D. L. Mohler and T. Lian, J. Phys. Chem. B, 2003,
[Re
2
(m-Cl)
2
(CO)
6
(m-pyridazine)] and ꢀ1.58 V for [Re
2 2
(m-Cl) -
(CO(m-4,5-Mez-pyridazine))] containing an electron-rich pyrida-
zine ligand. See for instance M. Mauro, E. Quartapelle Procopio,
Y. Sun, C.-H. Chien, D. Donghi, M. Panigati, P. Mercandelli,
P. Mussini, G. D’Alfonso and L. De Cola, Adv. Funct. Mater.,
2009, 19, 2607–2614.
1
07, 14231–14239.
1
4 L. Veronese, E. Quartapelle Procopio, F. De Rossi,
T. M. Brown, P. Mercandelli, P. Mussini, G. D’Alfonso and
M. Panigati, New J. Chem., 2016, 40, 2910.
15 S. M. Feldt, E. A. Gibson, E. Gabrielsson, L. Sun, G. Boschloo 21 A. Raimondi, M. Panigati, D. Maggioni, L. D’Alfonso,
and A. Hagfeldt, J. Am. Chem. Soc., 2010, 132, 16714–16724.
6 X. Jiang, K. M. Karlsson, E. Gabrielsson, E. M. J. Johansson,
P. Mercandelli, P. Mussini and G. D’Alfonso, Inorg. Chem.,
2012, 51, 2966–2975.
1
M. Quintana, M. Karlsson, L. Sun, G. Boschloo and 22 L. Veronese, E. Quartapelle Procopio, D. Maggioni, P. Mercandelli
A. Hagfeldt, Adv. Funct. Mater., 2011, 21, 2944–2952. and M. Panigati, New J. Chem., 2017, 41, 11268.
7 M. Panigati, M. Mauro, D. Donghi, P. Mercandelli, P. Mussini, 23 the data of devices with SnO are reported in the ESI† only
1
2
L. De Cola and G. D’Alfonso, Coord. Chem. Rev., 2012, 256,
621–1643.
8 M. Mauro, M. Panigati, D. Donghi, P. Mercandelli, P. Mussini,
for the dye 1.
1
24 J. E. Kroeze, N. Hirata, S. Koops, Md. K. Nazeeruddin,
L. Schmidt-Mende, M. Gr ¨a tzel and J. R. Durrant, J. Am.
Chem. Soc., 2006, 128, 16376–16383.
1
1
A. Sironi and G. D’Alfonso, Inorg. Chem., 2008, 47, 11154–11165.
9 E. Quartapelle Procopio, V. Bonometti, M. Panigati, 25 (a) M. A. Andrews, S. W. Kirtley and H. D. Kaesz, Inorg.
P. Mercandelli, P. Mussini, T. Benincori, G. D’Alfonso and
F. Sannicol `o , Inorg. Chem., 2014, 53, 11242–11251.
Chem., 1977, 16, 1556–1561; (b) J. R. Johnson and H. D.
Kaesz, Inorg. Synth., 1978, 18, 60–62.
Phys. Chem. Chem. Phys.
This journal is ©the Owner Societies 2019