Layered γ-zirconium phosphate intercalated with Ru(bpy)3-viologen dyads as unusual materials for dye-sensitised solar cells: Improving efficiency by double sensitisation

Article abstract of DOI:10.1071/CH13451

The intercalation materials prepared from the laminar salt α-zirconium phosphate and various dyads formed from Ru (bpy)₃ and viologens were tested as photoactive components in dye-sensitised solar cells. The efficiencies significantly increased, up to 0.2 %, when the materials were treated with the panchromatic N535 dye. A mechanism for this double sensitisation process is proposed, which should enable further improvement of these α-ZrP-based materials as a reasonable alternative to the usual Ti-based dye-sensitised solar cells. CSIRO 2014.

Full text of DOI:10.1071/CH13451

RESEARCH FRONT
CSIRO PUBLISHING
Aust. J. Chem. 2014, 67, 389–397
Full Paper
http://dx.doi.org/10.1071/CH13451
Layered g-Zirconium Phosphate Intercalated
with Ru(bpy)3-Viologen Dyads as Unusual Materials
for Dye-Sensitised Solar Cells: Improving Efficiency
by Double Sensitisation
A
B
A
Mar´ıa de Victoria-Rodriguez, Pedro Atienzar, Olga Juanes,
A
A C
,
Juan Carlos Rodr´ıguez-Ubis, Ernesto Brunet,
B C
,
and Hermenegildo Garc´ıa
A
Departamento de Qu´ımica Orga´nica, Facultad de Ciencias, Universidad Auto´noma de
Madrid, 28049 Madrid, Spain.
B
Instituto Universitario de Tecnolog´ıa Qu´ımica UPV-CSIC, Universidad Polite´cnica de
Valencia, Av. de los Naranjos s/n, Valencia, Spain.
C
Corresponding authors. Email: ernesto.brunet@uam.es; hgarcia@qim.upv.es
The intercalation materials prepared from the laminar salt g-zirconium phosphate and various dyads formed from Ru
(bpy)3 and viologens were tested as photoactive components in dye-sensitised solar cells. The efficiencies significantly
increased, up to 0.2 %, when the materials were treated with the panchromatic N535 dye. A mechanism for this double
sensitisation process is proposed, which should enable further improvement of these g-ZrP-based materials as a reasonable
alternative to the usual Ti-based dye-sensitised solar cells.
Manuscript received: 11 September 2013.
Manuscript accepted: 23 October 2013.
Published online: 25 November 2013.
Introduction
of the various positively charged species, intercalated between
the inorganic lamellae. The topotactic exchange is more com-
plex because it involves the replacement of the surface phos-
phates by other phosphorous entities such as phosphonates. In
this case, the phosphonates remain covalently linked to the
inorganic layer and their degrees of freedom are much more
limited.
The exploitation of these two processes, intercalation and
topotactic exchange, using molecules with different structures
and properties leads to a virtually unlimited number of solid
materials with a broad range of features. Figure 2 shows a
handful of examples of our own research.[3]
The versatile (g-ZrP) bears the molecular formula Zr(PO4)
[O2P(OH2)] ꢀ nH2O and its lamellar structure (Fig. 1) was long
ago refined by Rietveld methods from powder X-ray diffraction
(XRD) patterns.[1]
The structure reveals two kinds of phosphate groups. The
non-acidic internal phosphate groups (green tetrahedrons) are
the backbone of the layered structure, and remain bonded to four
different Zr atoms. In contrast, the acidic surface phosphates
(blue tetrahedrons) are responsible for the chemical properties
of g-ZrP. In short, this salt may enclose a wide variety of organic
components in the interlayer space either by reaction with basic
species leading to simple intercalation (ionic forces) or by
topotactic exchange with organic phosphorous derivatives
(covalent bonding).[2] In the case of simple intercalation, the
layered g-ZrP acts as a surface of multiple negative counterions
In the future, it is expected that DSSCs will play a key role in
ensuring the sustainability of energy conversion and/or produc-
tion processes. To this effect, considerable research effort has
been devoted to enhancing the efficiency of DSSCs based on
OϪ
OH
n
ϩ n
ϩ n
P
P
OϪ
OϪ
O
OH
OϪ
O
γ-ZrPO4[O2P(OH)2]
Fig. 1. Polyhedric structure of a layer of g-zirconium phosphate, as deduced by powder X-ray diffraction.
Journal compilation Ó CSIRO 2014
www.publish.csiro.au/journals/ajc

390
M. de Victoria-Rodriguez et al.
Variable porosity
(Catalysis, shape recognition, gas
separation...)
Cation recognition
(Entrapment of noxious metals...)
Supramolecular chirality
(Chiral recognition...)
C
A
B
Luminescent (chiral)
materials
(Optical devices, signaling, CPL...)
Drug and dye
confinement
(Controlled delivery...)
D
H
Photoactive materials
(Solar Cells...)
F
E
G
Electrochemical “noses”
(Detection of dangerous chemicals...)
Porous materials
(Storage of gases (hydrogen)...)
Fig. 2. Summary of the main application areas of g-ZrP-based materials, as explored by our research group.
titania, by far, the most used semiconductor in this area.[4] Yet,
despite these efforts, the maximum efficiency of titania-based
DSSCs has seldom exceeded 10 %, which is still too low for
commercial use. Thus, it would be of prime interest in this
context to explore alternative semiconducting materials. The
aforementioned versatility of organic–inorganic hybrid materi-
als based on g-ZrP makes these substances worth examining in
the DSSC field (Fig. 2, Path E).
Our recentlyperformed laserflash-photolysis studiesofg-ZrP
containing rather sophisticated arrangements of Ru(bpy)3 and
C60 derivatives[5] suggested semiconducting properties of g-ZrP
that had not been contemplated before. Thus, we seriously
considered the capabilities of laminar organic–inorganic g-ZrP-
based composites that have never been used for the construction
of dye-sensitised solar cells (DSSCs).
As mentioned earlier, the beauty of g-ZrP-based systems is
that the photoactive ruthenium complexes (electron donors) and
suitable electron acceptors (e.g. C60 or viologens) can be easily
placed in their intergalleries to ensure a good interfacial contact
between the photosensitive species and the semiconductor.
Despite the relatively marginal efficiencies attained in our first
attempts,[6] we achieved a remarkable 20-fold improvement by
covalently tethering the photoactive Rucomplex and the acceptor
species, and intercalation of the resulting dyad into g-ZrP by very
simple procedures.[7] Although the best overall efficiency (0.1%)
of these materials was still well below the values currently
achieved with titanium dioxide, we believe our work may pave
the way to further research in layered organic–inorganic compo-
sites, the efficiency of which could be improved by the using a
more suitable dye with a stronger and wider light absorption
response and/or doping of the semiconducting salt.
In the present paper we address the issue by both modifying
the structure of the Ru(bpy)3-viologen dyads and adding a
supplementary ruthenium dye (Ruthenizer 535 or N3), which
has a wider panchromatic absorption range.
Results and Discussion
Scheme 1 shows the dyads and synthetic procedures used in the
present study. The dyads were intercalated individually intog-ZrP
by routine procedures leading to the materials listed in Table 1.
Figure 3 shows models of the synthetised dyads 1 4. The
]
measured interlayer distances by powder XRD are only com-
patible with an arrangement of the molecules aligning their
longest axis parallel to the layers as shown in Fig. 4 for dyad 1. It
should be noted that the lamellae are formed by two metal layers
(Figs 1, 4). Therefore, two consecutive slabs containing 100 Zr
atoms each enclose an interlayer space between two facing
surfaces, each equivalent to the effective area surrounding
50 metal atoms. The available space as per Zr in the g-ZrP

Unusual materials for dye-sensitised solar cells
391
O
ϩ
H
N
N
N
N
N
N
N
ii-v
Ru
ϩ
N
N
N
Dyad 1
i
ϩ
N
N ϩ
N
vi, iv-v
N
N
N
ϩ
ϩ
m
n
n
N
N
N
ϩ
Dyad
n
m
N
2
3
4
3
0
0
1
N
Ru
N
10
3
A
N
N
Scheme 1. Synthesis of dyads 1 4 (i: SeO , dioxane 1108C, 24h; ii: NaBH , MeOH, 3h; iii: PBr , CH CN, 3h;
]
2
4
3
3
iv: bromide of (A), CH3CN908C,48h;v: Ru(bpy)2Cl2, ethanol : H2O (1: 1), 1108C;vi: lithiumdiisopropyl amide(LDA; 1
eq for dyads 2 and 3 or 2 eq for dyad 4), THF ꢂ788C, 1,3-dibromopropane (for dyads 2 and 4) or 1,10-dibromodecane
(for dyad 3)). Dyad counterions were a mixture of chloride and bromide, which were lost in the intercalation (see text).
A similar situation would certainly be attained in the case of the
other dyads.
Table 1. Materials resulting from the intercalation of dyads 1-4 of
Scheme 1 into g-ZrP
(ZrPO4)100 (H2-axPO4)100 (Dyad)x (C3H6O)z ꢀ (H2O)w
The four resulting solid composites were deposited as thin
films in order to build two sets of four corresponding cells.
Scheme 2 shows the configuration of the cells used in the present
study. Set A was assembled without further modification, where-
as set B was prepared by additionally dipping the thin film into an
ethanolic basic solution of N535 dye for 12h before their final
assembly (see the Experimental section for details). Two addi-
tional cells were created from pristine g-ZrP with and without
N535 for comparison. It should be noted that it is unlikely that the
N535 dye would reach the interlayer space of the particles
because of the high occupation of this space by the dyads and
the mild dipping conditions. No significant changes were
observed between the powder X-ray patterns of the ‘dipped’
and ‘non-dipped’ films. Hydrogen bonding among the carboxylic
acid groups and the particles outermost phosphates should allow
the N535 dye to be easily adsorbed on the surface.
Table 2 summarises the results. The performance of the cells
was evaluated by measuring the relevant parameters: open-
circuit voltage (Voc), short-circuit current (Jsc), and fill factor
(FF ). The presence of co-sensitiser N535 did not significantly
affect either the Voc or FF values, whereas Jsc underwent a two-
to-three fold increase in all cases studied. This behaviour was
consistent in all the various studied replicas of the DSSCs sets,
each of which was constructed independently. The observed
general increment of Jsc can be explained in terms of the extra
MaterialA
Elemental analysisB
C [%] H [%] N [%] (Dyad)x
Molecular formula d100 [nm]
z
w/100
1 (a 5 4) 20.0 (20.3) 3.6 (3.6) 2.4 (2.7) (1)11 100
2
1.93
1.93
1.70
1.64
2 (a 5 4) 15.4 (15.1) 2.9 (2.8) 2.5 (3.0) (2)11
3 (a 5 4) 17.0 (16.9) 3.2 (3.7) 1.7 (1.8) (3)6
4 (a 5 6) 12.0 (11.9) 2.5 (2.6) 2.0 (2.2) (4)6
0
75
0
2.1
2
2
Aa denotes the number of positive charges of dyad or triad for computing the
molecular formula.
BData show found (calculated) values.
structure (Fig. 1) is ,0.36 nm2 or 50 ꢁ 0.36 ¼ 18 nm2 for two
consecutive lamellae containing 100 Zr atoms each. The models
in Fig. 4 show that the projection area of dyads 1–3 is ,1.7, 2.7,
and 3.5 nm2, respectively. These simple considerations and the
elemental analyses (Table 1) indicate that g-ZrP was able to take
in as much dyad as the available space allowed (,6–11 dyads
depending on the chain length). In the case of dyads 1 and 3,
some space remained that was occupied by acetone molecules
from the intercalation medium. In the case of dyad 4, a simple
qualitative model (Fig. 3) indicates that the experimentally
observed accommodation of six molecules per (50 þ 50) Zr
atoms (Table 1) requires some kind of cooperative arrangement.

392
M. de Victoria-Rodriguez et al.
1.82 nm
C11
0.99 nm
0.97 nm
3.77 nm
1.2 nm
1.0 nm
3.49 nm
C4
1.02 nm
0.99 nm
2.65 nm
C1
5.0 nm
1.09 nm
0.89 nm
1.93 nm
4.0 nm
Fig. 3. Molecular models of the dyads 1 3 (left) showing their calculated measurements and possible co-operative arrangement of the intercalated dyads 3 4
]
]
(right).
ϩ
N
N
N
N
N ϩ
2ϩ
N
1.9 nm
N
Ru
N
Fig. 4. Representative molecular model of two layers of g-ZrP at the experimental interlayer distance containing the appropriate number of molecules of dyad
1, as deduced from elemental analysis.
photons absorbed by the panchromatic N535 co-sensitiser that
should contribute to the increasing number of charge separation
events. Interestingly, the overall efficiency of the cells contain-
ing the two photo-sensitisers, i.e. the ruthenium–viologen dyads
in the intergallery space and the panchromatic N535 dye on the
external surface of the laminar particles, reached similar values
in the range of 0.15–0.18 %. The fact that the control cells, built
from pristine g-ZrP, showed low and similar efficiencies
(0.04 % and 0.06 % for sets A and B, respectively) points to a
synergistic effect between the interlayer dyads and the adsorbed
N535 dye. The measured photoresponse of some of the devices
(Fig. 5) supports this conclusion.

Unusual materials for dye-sensitised solar cells
393
Glass cover
Surlyn
Hundreds
of layers
Glass substrate
FTO layer
Practical cell
Pt layer
Electrolyte
Set A
Set B
Surlyn
N535
OR
Dyads 1–4
Active material
TiOx layer
γ-ZrP
0.5
0
FTO layer
Glass
substrate
Hundreds
of layers
Arrangement of
a practical cell
0
0.5
V [v]
Scheme 2. General diagram of the g-ZrP-based DSSCs assembled in this work.
Table 2. Performance of the two sets of DSSCs prepared with the
3
2
1
0
g-ZrP-based materials 1 4 upon illumination with a solar simulator
]
through a 1.5 a.m. filter at 1000 W m22
VOC, voltage at open circuit; JSC, current density at short circuit; FF, fill
factor.
b
Material (set)A Voc [mV] Jsc [mA cmꢂ2
]
FF [%] Efficiency [%]
a
g-ZrP(A)
g-ZrP(B)
1(A)
601.83
595.41
527.01
574.19
544.21
550.90
461.49
538.59
456.92
546.14
122.49
174.74
147.11
484.84
276.21
419.19
63.13
56.2
56.6
65.8
66.3
71.0
65.2
77.4
66.8
73.1
62.9
0.04
0.06
0.05
0.18
0.11
0.15
0.02
0.15
0.04
0.16
1(B)
2(A)
2(B)
3(A)
3(B)
427.85
132.32
475.01
300
400
500
600
700
800
4(A)
Wavelength [nm]
4(B)
ASet A without N535; set B with N535 (see Scheme 2).
Fig. 5. Representative IPCE spectra for the cell built from 1 without (a) and
with (b) N535 dye co-sensitisation.
The highest photocurrent occurs at ,350 nm. In previous
studies, we attributed this to direct light absorption by the
inorganic matrix, the dyad thus being responsible for the less
strong and discreet absorption at 475 and 690 nm. However,
when the N535 dye is present, the current is much more
effectively produced over a wider wavelength range, a finding
that can explain the efficiency increase in set B cells (Table 2).
A plausible explanation is illustrated in Scheme 3. When the
N535 dye is present on the surface of the particles,y sensitisation
should occur more effectively. All N535 molecules are readily
exposed to the incoming light, whereas only a few of the dyads
in the outermost layers could be excited. Our hypothesis is that
the effective N535 light-harvesting molecules could transfer the
absorbed energy to the ground state of the Ru(bpy)3 end of the
dyads that were not excited by the incident light. Therefore,
the role of the surface N535 molecules would be that of an
energy pump or conveyor belt, i.e. a kind of catalyst that makes
the entire light-absorption process more effective. Once the
dyad gets the energy from the N535 molecules, electron transfer
to the viologen takes place in either an intra- or intermolecular
fashion. The different performance of materials 1 3 (in the
]
absence of the N535 dye) with diverse spacer lengths suggests
that the intramolecular Ru(bpy)3-to-viologen electron transfer is
the favoured process. This was more pronounced with the four-
methylene spacer.
Following the dyad charge separation, electrons start being
injected into the g-ZrP semiconducting layers, resulting in the
generation of the light-induced current. Figure 6 supports the
semiconducting behaviour of g-ZrP. The cyclic voltammetry of
pure g-ZrP (Fig. 6a) exhibited a reversible peak at ꢂ0.6 V,
corresponding to the reduction of the material. The correspond-
ing reduction peak in cyclic voltammetry supports the possibi-
lity that the charge-separated dyad can inject an electron into the
g-ZrP layered salt. In addition, if a thin film of g-ZrP supported
on fluorine-doped tin oxide (FTO) glass is subjected to a 0-V
bias potential and the electrode is illuminated at 300 nm, a
photocurrent of ,0.5 mA can then be measured (see inset 1 of
Fig. 6a), a fact that further reinforces the behaviour of layered
g-ZrP as a semiconductor. Finally, Kubelka–Munk optical
yThe presence of the dyads or triad on the outer surface of the particles cannot be completely ruled out after the dipping process. Therefore, some dyads or triad
and the N535 dye would be in closer contact at the particles surface. This is a possible option that has to be investigated further.

394
M. de Victoria-Rodriguez et al.
∗
∗
∗
N535
N535 N535 N535
ZrP
N535
N535
ZrP
Electron
transfer
Energy
transfer
Ru(bpy)32ϩ- V2ϩ Ru(bpy)32ϩ- V2ϩ
∗
Ru(bpy)32ϩ- V2ϩ Ru(bpy)32ϩ- V2ϩ
(dyad) (dyad)
(dyad)
(dyad)
ZrP
ZrP
ZrP
Ru(bpy)32ϩ- V2ϩ
(dyad)
Ru(bpy)32ϩ- V2ϩ
(dyad)
hυ 460–600 nm
Ru(bpy)32ϩ- V2ϩ
(dyad)
Ru(bpy)32ϩ- V2ϩ
(dyad)
Sensitisation
ZrP
Electron
transfer
∗
Ru(bpy)32ϩ- V2ϩ Ru(bpy)32ϩ- V2ϩ
(dyad) (dyad)
Ru(bpy)32ϩ- V2ϩ
(dyad)
Ru(bpy)32ϩ- V2ϩ
(dyad)
Energy
transfer
ZrP
N535 N535 N535
ZrP
∗
N535
∗
∗
N535
N535
∗
∗
∗
∗
N535
∗
Electrons from
the circuit
Electrons to
the circuit
N535
N535
ZrP
N535
N535
ZrP
N535
∗
∗
Ru(bpy)3ϩ- Vϩ
Ru(bpy)3ϩ- Vϩ
Ru(bpy)32ϩ- V2ϩ
(dyad)
Ru(bpy)32ϩ- V2ϩ
(dyad)
3
3
(dyad)
(dyad)
Electron
transfer
to ZrP
ZrP
ZrP
hυ 460–600 nm
Sensitisation
Ru(bpy)32ϩ- V2ϩ Ru(bpy)32ϩ- V2ϩ
(dyad) (dyad)
Ru(bpy)32ϩ- V2ϩ Ru(bpy)32ϩ- V2ϩ
(dyad) (dyad)
Charge separation
Energy transfer
ZrP
ZrP
Energy
transfer
to ZrP
∗
∗
Ru(bpy)32ϩ- V2ϩ Ru(bpy)3ϩ- Vϩ
Ru(bpy)32ϩ- V2ϩ Ru(bpy)3ϩ- Vϩ
3
3
(dyad)
(dyad)
(dyad)
(dyad)
Electrons from
the circuit
Electrons to
the circuit
ZrP
ZrP
∗
∗
∗
∗
∗
N535
N535
N535
N535
N535
N535
Scheme 3. Proposed mechanism for the energy transfer process in the doubly sensitised materials.
measurements (Fig. 6b) of pristine g-ZrP revealed a band gap
energy of 6.2 eV, which is surprisingly higher than expected.
In conclusion, the photovoltaic response observed for modi-
fied g-ZrP and the outcome of the photo-electrochemical
measurements for g-ZrP have two important implications: (1)
the absorption of light appears to be much more effective at the
surface of the particles; and (2) the semiconducting properties of
g-ZrP are not very good because its bandgap is relatively high.
Yet, the studies presented herein illustrate once more the high
versatility offered by layered g-ZrP in building composite
materials, which definitely have potential for use in the solar
cell field. For instance, although still quite preliminary, our very
recent measurements of g-ZrP containing 5–7 % of Nb in the
layers showed a reduced band gap of ,4.0 eV (Fig. 6b), 2.2 eV
lower than that of pure g-ZrP. The application of metal-doped
g-ZrP together with the co-sensitisation method described in
this paper, the investigation of other photosensitising dyes, and a
better control of the particle size would lead to a substantial
improvement of g-ZrP-based DSSCs, all of which are currently
under investigation.
Experimental
Synthesis of Dyads 1–4
Preparation of 1-ethyl-4-pyridin-4-ylpyridinium
bromide (A)
In a sealed tube, a mixture of 4,40-bipyridine (5 g, 32.0 mmol)
and bromoethane (2.6 mL, 32.0 mmol) was stirred at room
temperature for 96 h, yielding a pale yellow solid that was
filtered and washed with diethyl ether, toluene, and acetone.
The remaining solid was dissolved in acetonitrile, and the
solution was filtered and the solvent was evaporated. Yield

Unusual materials for dye-sensitised solar cells
395
(a)
(b) 0.00015
0.00013
0.00011
9E-05
Nb doped
Non-Nb
doped
(i)
0
1
2
(ii)
light On
light Off
7E-05
5E-05
3E-05
Ϫ5.0ϫ10Ϫ5
Ϫ1.0ϫ10Ϫ4
1E-05
0
02:00 04:00 06:00 08:00 10:00
Time [min]
300
400
500
600
700
Wavelength [nm]
Ϫ1E-05
1
2
2
4
5
6
7
Ϫ1.0
Ϫ0.5
0
0.5
1.0
E[eV]
Voltage [V]
Fig. 6. (a) Cyclic voltammetry of g-ZrP supported on FTO (i) and FTO substrate (ii) as reference in a 0.1 M LiClO4 acetonitrile solution. Inset 1 shows the
photocurrent measured for this thin film of g-ZrP as a function of the excitation wavelength at 0-V bias potential. Inset 2 shows the photocurrentduring repeated
on/off cycles upon illumination with monochromatic light of 300 nm wavelength. (b) Tauc’s graphs from Kubelka–Munk optical measurements of pristine
g-ZrP and of its 7 % Nb-doped counterpart.
(85 %). 1H NMR (300 MHz, DMSO) d (ppm): 1.57 (t, J 7.3, 3H),
4.71 (q, J 7.3, 2H), 8.05 (m, 4H), 8.65 (d, J 6.8, 2H), 8.84
(m, 2H), 9.31(d, J 7.1, 2H). 13C NMR (75 MHz, DMSO) d (ppm):
16.29, 55.89, 121.85, 125.31, 140.82, 145.09, 150.87, 152.08.
added to
a
solution of 4,40-dimethyl-2,20-bipyridine
(1 g, 5.43 mmol) in dry THF (38 mL) at ꢂ788C. After 1 h, at
788C, the resulting brown–red solution was treated with dibro-
mopropane (0.88 mL, 8.69 mmol). After stirring for 1 h, a pale
green colour developed. The mixture was allowed to reach room
temperature for 3 h and the reaction was quenched with water.
Phosphate buffer (1 M, 10 mL, pH 7.0) was added and the
mixture was extracted with diethyl ether. Usual work-up of
the organic layer yielded a crude product that was purified by
silica gel chromatography with 1 : 1 ethyl ether : hexane, pro-
ducing a colourless oil (0.69 g, 42 %). 1H NMR (300 MHz,
CDCl3) d (ppm): 1.89 (m, 4H), 2.44 (s, 3H), 2.74 (t, J 7.3,
2H), 3.42 (t, J 6.6, 2H), 7.14 (d, J 1.4, 2H), 8.24 (s, 2H), 8.54
(d, J 5.2, H), 8.57 (d, J 5.1, H). 13C NMR (75 MHz, CDCl3)
d (ppm): 156.0, 156.0, 152.9, 148.9, 148.8, 148.1, 124.6, 123.9,
122.0, 121.3, 34.26, 33.01, 31.93, 28.51, 20.95.
Synthesis of 1-ethyl-10-[(40-methyl-2,20-bipyridin-4-yl)
methyl]-4,40-bipyridinium dibromide (Bipy1-A)
A mixture of 4-(bromomethyl)-40-methyl-2,20-bipyridine[8]
(250 mg, 0.95 mmol), and A bromide (264 mg, 0.99 mmol) in
CH3CN (6.5 mL) was heated at 908C for 120 h in a sealed tube.
The resulting yellow precipitate was filtered off and washed
with hot acetonitrile and diethyl ether, yielding 381 mg (73 %) of
the product. 1H NMR (300 MHz, CD3OD) d (ppm): 1.73
(t, J 7.4, 3H), 2.47 (s, 3H), 4.80 (q, J 7.4, 2H), 6.16 (s, 2H),
7.31 (d, J 5.0, 1H), 7.58 (dd, J 1.8, 5.0, 1H), 8.23 (m, 1H), 8.31
(m, 1H), 8.47 (d, J 5.0, 1H), 8.70 (d, J 6.8, 2H), 8.77 (m, 3H),
9.30 (d, J 6.8, 2H), 9.44 (d, J 7.0, 2H). 13C NMR (75.5 MHz,
CD3OD) d (ppm): 16.71, 21.21, 58.85, 64.41, 121.47, 123.79,
124.43, 126.72, 128.46, 128.89, 144.69, 146.78, 146.95, 147.76,
150.01, 150.88, 151.63, 152.28, 156.08, 158.43.
Synthesis of 1-ethyl-10-[4-(40-methyl-2,20-bipyridin-4-yl)
butyl]-4,40-bipyridinium dibromide (Bipy2-A)
A mixture of Bipy2 (309 mg, 1.03 mmol) and A bromide
(290 mg, 1.08 mmol) in CH3CN (10 mL) was heated at 908C for
48 h in a sealed tube. The resulting yellow precipitate was
filtered off and washed with hot acetonitrile and diethyl ether,
Preparation of Dyad 1
A mixture of Bipy1-A (160 mg, 0.320 mmol) and cis-bis(2,20-
bipyridine)dichlororuthenium (170 mg, 0.351 mmol) in ethanol :
water (7 : 3, 11.8mL) was heated at 1108C for 24 h, protected
from ambient light. The mixture was concentrated under vacuum
to yield a dark solid that was used to prepare 1 without further
purification. 1H NMR (300 MHz, CD3OD) d (ppm): 1.73 (t, J 7.2,
3H), 2.60 (s, 3H), 4.82 (q, J 7.2, 2H), 6.26 (s, 2H), 7.36 (d, J 6.4,
1H), 7.34–7.58 (m, 4H), 7.63 (d, J 5.7, 2H), 7.77–7.85 (m, 3H),
7.89 (d, J 6.0, 1H), 7.98 (d, J 4.9, 1H), 8.08–8.16 (m, 4H),
8.68–8.81 (m, 9H), 9.02 (s, 1H), 9.30 (d, J 7.2, 2H), 9.54 (d, J 7.2,
2H). 13C NMR (75.5 MHz, CD3OD) d (ppm): 16.71, 21.25,
58.50, 63.50, 125.57, 125.62, 125.69, 127.40, 128.48, 128.91,
128.97, 129.05, 129.14, 130.34, 139.15, 139.27, 144.70, 146.88,
147.98, 151.17, 151.71, 152.22, 152.27, 152.49, 152.70, 152.77,
153.14, 153.43, 157.36, 158.32, 158.47, 158.49, 158.53, 159.60.
1
yielding 300 mg (52 %) of the product. H NMR (300 MHz,
DMSO) d (ppm): 1.60 (t, J 7.4, 3H), 1.71 (m, 2H), 2.04 (m, 2H),
2.40 (s, 3H), 2.77 (t, J 7.5, 2H), 4.74 (m, 4H), 7.29 (m, 2H), 8.22
(d, J 7.1, 2H), 8.52 (dd, J 4.8 Hz; 15.6, 2H), 8.78 (d, J 4.9, 4H),
9.41 (d, J 6.5, 4H). 13C NMR (75 MHz, DMSO) d (ppm): 16.27,
20.68, 26.21, 30.27, 33.80, 56.52, 60.58, 120.45, 121.21,
124.15, 124.89, 126.51, 126.61, 145.57, 145.73, 147.91,
148.48, 148.54, 148.90, 149.15, 151.46, 155.04, 155.30.
Preparation of Dyad 2
A mixture of Bipy2-A (300 mg, 0.526 mmol) and cis-bis
(2,20-bipyridine)dichlororuthenium (331 mg, 0.684 mmol) in
ethanol : water (7 : 3, 24 mL) was heated at 1108C for 24 h
protected from ambient light. The mixture was concentrated in
vacuum to yield 631 mg of a dark solid that was used to prepare 2
Preparation of 4-(4-bromobutyl)-40-methyl-2,20-
bipyridine (Bipy2)
1
without further purification. H NMR (300 MHz, CD3OD) d
(ppm): 1.73 (t, J 7.5, 3H), 1.92 (m, 2H), 2.24 (m, 2H), 2.59
(s, 3H), 2.97 (m, 2H), 4.85 (m, 4H), 7.32 (d, J 5.7, H), 7.41
(d, J 5.7, H), 7.44–7.56 (m, 4H), 7.59 (d, J 5.7, H), 7.64 (d, J 5.9,
H), 7.78–7.87 (m, 4H), 8.07–8.17 (m, 4H), 8.65–8.75 (m, 10H),
Butyllithium (1.6 M in hexane, 2.7 mL, 4.34 mmol) was
added to a solution of diisopropylamine (0.61 mL, 4.34 mmol)
in THF (9 mL) at ꢂ788C. The mixture was stirred for 45 min and

396
M. de Victoria-Rodriguez et al.
9.28 (d, J 6.4, 2H), 9.34 (d, J 6.4, 2H). 13C NMR (75 MHz,
CD3OD) d(ppm):16.86, 21.32, 27.48, 31.87, 35.24, 58.64, 62.70,
125.66, 125.91, 126.70, 128.33, 128.87, 128.92, 128.98, 129.16,
129.78, 139.01, 146.81, 147.10, 151.11, 151.52, 151.85, 151.92,
152.49, 152.54, 155.07, 157.82, 158.09, 158.33, 158.37, 158.45.
then added to a solution of 4,40-dimethyl-2,20-bipyridine
(1 g, 5.43mmol) in THF (38 mL) at ꢂ788C. After 1 h at 788C,
the resulting brown–red solution was treated with dibromopro-
pane (2.2mL, 21.68mmol) and, after stirring for 1 h, a pale green
colour developed. The mixture was allowed to reach room
temperature for 3 h and the reaction was quenched with water.
Phosphate buffer(1M, 10mL, pH7.0)was added and the mixture
was extracted with diethyl ether. Usual work-up of the organic
layer yielded a crude product that was purified by silica-gel
chromatography with 1 : 1 ethyl ether : hexane to give a white
solid (1.27 g, 55%). 1H NMR (300MHz, CDCl3) d (ppm): 1.90
(m, 8H), 2.73 (t, J 7.3, 4H), 3.42 (t, J 6.4, 4H), 7.12 (d, J 4.9, 2H),
8.24 (m, 2H), 8.56 (d, J 4.9, 2H). 13C NMR (75 MHz, CDCl3) d
(ppm): 28.50, 31.95, 33.11, 34.26, 120.92, 121.77, 123.55,
124.48, 147.85, 148.68, 148.87, 151.487, 155.69, 156.02.
Preparation of 4-(11-bromoundecyl)-40-methyl-2,20-
bipyridine (Bipy3)
Butyllithium (1 M in hexane, 2.17 mL, 2.17 mmol) was
added to a solution of diisopropylamine (0.304 mL, 2.17 mmol)
in THF (4.5 mL) at 788C. The mixture was stirred for 45 min and
added to a solution of 4,40-dimethyl-2,20-bipyridine (0.5 g,
2.71 mmol) in THF (19 mL) at ꢂ788C. After 1 h at 788C, the
resulting brown–red solution was treated with 1,10-dibromode-
cane (1 mL, 4.34 mmol) and, after stirring for 1 h, a pale green
colour was developed. The mixture was allowed to reach room
temperature for 3 h and the reaction was quenched with water.
Phosphate buffer (1 M, 10 mL, pH 7.0) was added and the
mixture was extracted with diethyl ether. Usual work-up of
the organic yielded a crude product that was purified by silica
gel chromatography with 1 : 1 ethyl ether : hexane to give a
white solid (42 % yield). 1H NMR (300 MHz, CDCl3) d (ppm):
1.14–1.38 (m, 14H), 1.62 (m, 2H), 1.77 (m, 2H) 2.35 (s, 3H),
2.61 (t, J 7.7, 2H), 3.32 (t, J 6.8, 2H), 7.05 (m, 2H), 8.17 (m, 2H),
8.48 (m, 2H). 13C NMR (75 MHz, CDCl3) d (ppm): 156.0,
156.0, 152.9, 148.9, 148.8, 148.1, 124.6, 123.9, 122.0, 121.3,
35.5, 34.0, 32.8, 30.4, 29.4, 29.4, 29.3, 29.2, 28.7, 28.1, 21.2.
Preparation of 1,10-[2,20-bipyridine-4,40-diylbis(butane-
4,1-diyl)]bis(10-ethyl-4,40-bipyridinium) tetrabromide
(Bipy4-A)
A mixture of Bipy4 (1.27 g, 2.98 mmol) and A bromide
(1.66 g, 6.27 mmol) in CH3CN (38 mL) was heated at 908C for
48 h in a sealed tube. The resulting yellow precipitate was
filtered off and washed with hot acetonitrile and diethyl ether,
1
yielding 300 mg (62 %) of the product. H NMR (300 MHz,
DMSO) d (ppm): 1.60 (t, J 7.4, 6H), 1.72 (m, 4H), 2.06 (m, 4H),
2.77 (t, J 7.5, 4H), 4.76 (m, 8H), 7.32 (d, J 4.8, 2H), 8.24 (s, 2H),
8.55 (d, J 4.8 Hz; 2H), 8.81 (m, 8H), 9.43 (m, 8H). 13C NMR
(75 MHz, DMSO) d (ppm): 16.27, 26.21, 30.27, 33.80, 56.52,
60.58, 120.54, 124.27, 126.52, 126.61, 145.59, 145.75, 148.40,
148.47, 149.00, 151.46, 155.00.
Preparation of 1-ethyl-10-[4-(40-methyl-2,20-bipyridin-
4-yl)butyl]-4,40-bipyridinium dibromide (Bipy3-A)
A mixture of Bipy3 (702 mg, 1.05 mmol) and A bromide
(293mg, 1.10mmol) in CH3CN (10 mL) was heated at 908C for
48h in a sealed tube. The resulting yellow precipitate was filtered
off and washed with hot acetonitrile and diethyl ether, yielding
Preparation of Dyad 4
A mixture of Bipy4-A (489 mg, 0.511 mmol) and cis-bis
(2,20-bipyridine)dichlororuthenium (322 mg, 0.664 mmol) in
ethanol : water (7 : 3, 24 mL) was heated at 1108C for 24 h,
protected from ambient light. The mixture was concentrated
under vacuum to yield quantitatively a dark solid that was
used to prepare 4 without further purification. 1H NMR
(300 MHz, DMSO) d (ppm): 1.60 (t, J 7.2, 6H), 1.81 (m, 4H),
2.08 (m, 4H), 2.90 (t, J 6.8, 4), 4.77 (m, 4H), 4.89 (m, 4H), 7.43
(d, J 5.9, 2H), 7.50–7.60 (m, 6H), 7.69–7.75 (m, 6H), 8.12–8.19
(m, 4H), 8.90 (m, 12H), 9.49 (m, 4H), 9.65 (m, 4H). 13C NMR
(75 MHz, DMSO) d (ppm): 16.42, 25.54, 30.27, 33.32, 56.34,
60.07, 124.58, 126.62, 127.88, 128.03, 137.78, 145.66, 145.90,
148.39, 150.33, 151.09, 151.16, 153.20, 156.29, 156.51, 156.61.
1
394 mg (61 %) of the product. H NMR (300MHz, DMSO) d
(ppm): 1.17–1.35 (m, 14H), 1.60 (m, 5H), 1.95 (m, 2H), 2.40 (s,
3H), 2.67 (t, J 7.7, 2H), 4.70 (m, 4H), 7.26 (m, 2H), 8.20 (m, 2H),
8.52 (m, 2H), 8.78 (m, 4H), 9.40 (m, 4H). 13C NMR (75 MHz,
DMSO) d (ppm): 155.2, 155.1, 152.3, 149.0, 148.9, 148.5, 148.4,
147.9,145.7,145.6, 126.6, 126.5, 124.8,124.1, 121.2,120.4, 60.8,
56.5, 34.5, 30.7, 29.8, 28.8, 28.7, 28.5, 28.3, 25.4, 20.7, 16.27.
Preparation of Dyad 3
A mixture of Bipy3-A(280 mg, 0.418 mmol) and cis-bis(2,20-
bipyridine)dichlororuthenium (264mg, 0.544 mmol) in ethanol :
water (7 : 3, 19 mL) was heated at 1108C for 24 h protected from
ambient light. The mixture was concentrated under vacuum to
yield 590 mg of a dark solid that was used to prepare 3 with-
out further purification. 1H NMR (300 MHz, DMSO) d (ppm):
1.20–1.34 (m, 14H), 1.59 (t, 3H), 1.64 (m, 2H), 1.96 (m, 2H),
2.51 (s, 3H), 2.75 (m, 2H), 4.79 (m, 4H), 7.37 (m, 2H), 7.53
(m, 6H), 7.71(m, 4H), 8.15(m, 4H), 8.86 (m, 10H), 9.48(m, 4H).
13C NMR (75 MHz, DMSO) d (ppm): 16.32, 20.64, 25.47, 28.46,
28.79, 28.84, 28.94, 29.51, 30.83, 34.43, 55.97, 60.71, 124.40,
124.50, 125.28, 126.60, 127.81, 128.56, 137.74, 137.69, 145.64,
145.79, 148.50, 149.76, 150.22, 150.40, 151.05, 151.12, 151.17,
154.03, 156.02, 156.17, 156.58, 156.62.
General Procedure of Intercalation of Dyads inside g-ZrP
g-ZrP (400 mg) was added to a 1 : 1 water : acetone solution
(40 mL) and the suspension was stirred at 808C for 35 min to
attain complete exfoliation of the layers. The dyad to be
intercalated was dissolved in a 1 : 1 water : acetone solution
(33.4 mL), and added to the g-ZrP suspension. The mixture
was maintained at 808C for 72 h, and the solid was centrifuged at
5–10g for 15 min at room temperature, washed thrice for 20 min
with distilled water and centrifuged under the same conditions
again. The washing–centrifugation procedures were repeated
twice. The resulting orange solid was dried in the oven at 1008C
for 24 h and conditioned for at least 3 h in a desiccator that
contained a saturated solution of BaCl2.
Preparation of 4,40-bis(4-bromobutyl)-2,20-bipyridine
(Bipy4)
Diffuse Reflectance UV-Vis Spectra
Butyllithium (1.59 M in hexane, 7.5 mL, 11.94 mmol) was
added to a solution of diisopropylamine (1.5mL, 11.94 mmol) in
THF (23 mL) at ꢂ788C. The mixture was stirred for 45 min and
Optical spectra in the 200–800 nm region were recorded on a
Cary 5G spectrophotometer for the powdered samples, using the

Unusual materials for dye-sensitised solar cells
397
diffuse reflectance mode and an integrated sphere. BaSO4 was
used as the reference. The remittance (R) from the instrument
was recorded and the spectra plotted as the Kubelka–Munk
function (F(R) ¼ (1-R)2/2R). The band gap of g-ZrP was esti-
mated from the onset of the absorption band (330 nm) by
applying the equation: E (eV) E 1240/l (nm).
Level in the Solid State’’, Chapter 9 (2012), 279–316; in Metal
Phosphonate Chemistry, From Synthesis to Applications, (Abraham
Clearfield, Kostantinos D. Demadis, Eds.) RSC Publishing.
[3] (a) E. Brunet, M. Huelva, R. Va´zquez, O. Juanes, J. C. Rodr´ıguez-Ubis,
Chem. –Eur. J. 1996, 2, 1578. doi:10.1002/CHEM.19960021217
Figure 2, path B (b) E. Brunet, H. M. H. Alhendawi, O. Juanes,
J. C. Rodr´ıguez-Ubis, J. Mex. Chem. Soc. 2009, 53, 155.
(c) G. Alberti, E. Brunet, C. Dionigi, O. Juanes, M. J. de la Mata,
J. C. Rodr´ıguez-Ubis, R. Vivani, Angew. Chem. Int. Ed. 1999,
38, 3351. doi:10.1002/(SICI)1521-3773(19991115)38:22,3351::AID-
ANIE3351.3.0.CO;2-V
Solar Cell Preparation
A series of photo-electrochemical cells were prepared by
depositing a micrometric layer of the corresponding g-ZrP
material as a paste onto a conducting transparent electrode
FTO, using the doctor blade technique. The surface of the
FTO electrode onto which the film is deposited is defined by
means of two parallel strips of adhesive Scotch tapes. The final
area obtained was 1 ꢁ 1 cm2. The film was dried at 1508C for
15 min to remove the terpineol from the paste. The paste was
obtained by stirring 100 mg g-ZrP material in a 3 : 1 mixture of
acetone : a-terpineol (4 mL). The electrode-containing material
was immersed in an ethanolic basic solution of N535 for 12 h.
The platinum counter electrode was produced by thermal
decomposition of H2PtCl6 in isopropanol solution. The ZrP
electrode was assembled with the platinum-based transparent
conducting glass using a double sided adhesive polymer film
(Surlyn, DuPont) that acts as a separator and sealing element.
The two electrodes were held together by hot melting the Surlyn
seal at 1008C while applying pressure. The electrolyte (0.5 M
lithium iodide and 0.05 M iodine in methoxypropionitrile) was
introduced into the cell through the two holes that were drilled in
the counter electrode.
(d) E. Brunet, O. Juanes, M. J. de la Mata, J. C. Rodr´ıguez-Ubis, Angew.
Chem., Int. Ed. 2004, 43, 619.
Figure 2, path C (e) E. Brunet, M. J. de la Mata, H. M. H. Alhendawi,
C. Cerro, M. Alonso, O. Juanes, J. C. Rodr´ıguez-Ubis, Chem. Mater.
2005, 17, 1424. doi:10.1021/CM048754Q
(f) E. Brunet, M. J. de la Mata, O. Juanes, H. M. H. Alhendawi, C. Cerro,
J. C. Rodr´ıguez-Ubis, Tetrahedron: Asymmetry 2006, 17, 347.
doi:10.1016/J.TETASY.2006.01.014
Figure 2, path D (g) R. Ferna´ndez-Ruiz, J. C. Rodr´ıguez-Ubis,
A. Salvador, E. Brunet, O. Juanes, J. Anal. At. Spectrom. 2010, 25, 1882.
doi:10.1039/C0JA00043D
(h) E. Brunet, O. Juanes, L. Jime´nez, J. C. Rodr´ıguez-Ubis, Tetrahedron
Lett. 2009, 50, 5361. doi:10.1016/J.TETLET.2009.07.027
(i) E. Brunet, H. M. H. Alhendawi, O. Juanes, L. Jime´nez,
J. C. Rodr´ıguez-Ubis, J. Mater. Chem. 2009, 19, 2494. doi:10.1039/
B817317F
(j) E. Brunet, M. J. de la Mata, O. Juanes, J. C. Rodr´ıguez-Ubis, Chem.
Mater. 2004, 16, 1517. doi:10.1021/CM0353106
Figure 2, path E (k) E. Brunet, M. Alonso, C. Cerro, O. Juanes,
J. C. Rodr´ıguez-Ubis, A. E. Kaifer, Adv. Funct. Mater. 2007, 17, 1603.
doi:10.1002/ADFM.200700048
(l) E. Brunet, M. Alonso, M. J. de la Mata, S. Ferna´ndez, O. Juanes,
O. Chavanes, J. C. Rodr´ıguez-Ubis, Chem. Mater. 2003, 15, 1232.
doi:10.1021/CM025778T
Figure 2, path F (m) E. Brunet, H. M. H. Alhendawi, C. Cerro, M. J. de
la Mata, O. Juanes, J. C. Rodr´ıguez-Ubis, Microporous Mesoporous
Mater. 2011, 138, 75. doi:10.1016/J.MICROMESO.2010.09.027
(n) E. Brunet, H. M. H. Alhendawi, C. Cerro, M. J. de la Mata, O. Juanes,
J. C. Rodr´ıguez-Ubis, Chem. Eng. J. 2010, 158, 333. doi:10.1016/J.CEJ.
2010.01.040
(o) E. Brunet, C. Cerro, O. Juanes, J. C. Rodr´ıguez-Ubis, A. Clearfield,
J. Mater. Sci. 2008, 43, 1155. doi:10.1007/S10853-007-2377-0
(p) E. Brunet, H. M. H. Alhendawi, C. Cerro, M. J. de la Mata, O. Juanes,
J. C. Rodr´ıguez-Ubis, Angew. Chem., Int. Ed. 2006, 45, 6918.
doi:10.1002/ANIE.200602445
Figure 2, path G: unpublished results. Figure 2, path H (q) H. M. H.
Alhendawi, E. Brunet, E. Rodr´ıguez-Paya´n, O. Juanes, J. C. Rodr´ıguez-
Ubis, M. Al-Asqalany, J. Incl. Phenom. Macrocycl. 2012, 73, 387.
doi:10.1007/S10847-011-0076-6
Photovoltaic Response Measurements
To determine the J–V plots, the cell was connected to a
source meter (Keithley 2601) using metallic clamps covered
with gold. The voltage scan was controlled using ReRa Tracer
software. The data were automatically transferred to a PC that
controlled the experiment and, at the same time, provided data-
storage capability to the system. The solar simulator (Sun 2000,
ABET Technologies) was adapted to the AM 1.5G filter and the
nominal power for the measurements was 100 mW cmꢂ2. The
same cells were used to record the incident photon-to-electron
conversion efficiency (IPCE) spectra. For the IPCE measure-
ments the sample was excited with a 150-W xenon lamp through
a Czerny–Turner monochromator. The current output at short
circuit was measured by a potentiostat (AMEL), which trans-
ferred the data through the A/D converter to the PC controlling
the monochromator apparatus. IPCE curves were calculated
using a Newport (818-UV-L) calibrated photodiode.
[4] (a) M. R. Hoffmann, S. T. Martin, W. Y. Choi, D. W. Bahnemann,
Chem. Rev. 1995, 95, 69. doi:10.1021/CR00033A004
(b) B. O’Regan, M. Gratzel, Nature 1991, 353, 737. doi:10.1038/
353737A0
(c) M. Gra¨tzel, J. Photochem. Photobiol., A 2003, 4, 145. doi:10.1016/
S1389-5567(03)00026-1
[5] E. Brunet, M. Alonso, M. C. Quintana, P. Atienzar, O. Juanes,
J. C. Rodr´ıguez-Ubis, H. Garc´ıa, J. Phys. Chem. C 2008, 112, 5699.
doi:10.1021/JP800026R
[6] L. Teruel, M. Alonso, M. C. Quintana, A. Salvador, O. Juanes,
J. C. Rodriguez-Ubis, E. Brunet, H. Garc´ıa, Phys. Chem. Chem. Phys.
2009, 11, 2922. doi:10.1039/B816698F
[7] P. Atienzar, M. Victoria-Rodr´ıguez, O. Juanes, J. C. Rodr´ıguez-Ubis,
E. Brunet, H. Garc´ıa, EnergyEnviron. Sci. 2011, 4, 4718.
[8] (a) B. M. Peef, G. T. Ross, G. J. Meyer, T. J. Meyer, B. W. Erickson, Int.
J. Pept. Prot. Res. 1991, 38, 114.
Acknowledgements
The group from Universidad Auto´noma de Madrid (UAM) is grateful to
ERCROS-Farmacia SA (Aranjuez, Spain) for sponsoring the UAM-
ERCROS Chair in Chemistry that generously funded this work, in part. The
group from Universidad del Pa´ıs Valenciano-Consejo Superior de Investi-
gaciones Cient´ıficas (UPV-CSIC) group is grateful to the Spanish former
Ministerio de Ciencia e Innovacio´n de Espan˜a (MICINN) for grant
CTQ2012–32315. P.A. thanks the MICINN for a Juan de la Cierva research
contract.
References
[1] D. M. Poojary, B. Zhang, Y. Dong, G. Peng, A. Clearfield, J. Phys.
Chem. 1994, 98, 13616. doi:10.1021/J100102A029
[2] E. Brunet, ‘‘How Laminar Zirconium Phosphates Can Make Organic
Molecules Display New Behavior and Properties at the Supramolecular
(b) L. Strekowski, J. L. Mokrosz, F. A. Tanious, R. A. Watson,
D. Harden, M. Mokrosz, W. D. Edwards, W. D. Wilson, J. Med. Chem.
1988, 31, 1231. doi:10.1021/JM00401A027