Heteroleptic copper(i)-polypyridine complexes as efficient sensitizers for dye sensitized solar cells

Article abstract of DOI:10.1039/c4ta01755b

The synthesis and the physico-chemical characterization of HETPHEN based heteroleptic copper(i)-bis(diimine) complexes are reported. For TiO₂ based dye sensitized solar cells (DSCs), the latter display impressive photoconversion efficiencies (P

Full text of DOI:10.1039/c4ta01755b

Journal of
Materials Chemistry A
View Article Online
View Journal | View Issue
COMMUNICATION
Heteroleptic copper(I)–polypyridine complexes as
efficient sensitizers for dye sensitized solar cells†
Cite this: J. Mater. Chem. A, 2014, 2,
9944
Martina Sandroni,‡a Ludovic Favereau,a Aurelien Planchat,a Huriye Akdas-Kilig,b
Nadine Szuwarski,a Yann Pellegrin,a Errol Blart,a Hubert Le Bozec,b
Received 10th April 2014
Accepted 16th April 2014
a
a
*
*
Mohammed Boujtita and Fabrice Odobel
DOI: 10.1039/c4ta01755b
www.rsc.org/MaterialsA
The synthesis and the physico-chemical characterization of HETPHEN bipyridine onto which were attached two mesityl groups in
based heteroleptic copper(I)–bis(diimine) complexes are reported. For positions 6 and 60, providing the necessary steric bulk to avoid
TiO2 based dye sensitized solar cells (DSCs), the latter display the formation of homoleptic complexes. The ligands
impressive photoconversion efficiencies (PCEs), unprecedented for completing the coordination sphere of the copper(I) ion belong
first row transition metal coordination complexes.
to the family of 4,40-bis(styryl)-2,20-bipyridines, derivatized with
electron releasing moieties of different strength. Methyl groups
in a of the chelating nitrogen atoms confer rigidity to the nal
scaffold, preserving the excited state from exciplex quenching
and excessive attening upon excitation, to a certain extent.
Three complexes C2, C3 and C4, bearing respectively alkoxy,
N,N-diethylamine and N,N-diphenylamine moieties, were thus
isolated. For the sake of comparison, a fourth model complex
1
¨
Since 1991 and the discovery of DSC (Gratzel cells), many
attempts to replace the costly and toxic (albeit remarkably effi-
cient) ruthenium–polypyridine complexes have been repor-
ted.2,3 Copper(I)–bis(diimine) complexes have earlier shown
promising results in this eld.4–6 Lately, the use of heteroleptic
copper(I) complexes has afforded signicant PCEs thanks to an
improved extinction coefficient in the visible and electron
transfer vectorialization.6 The latter point is an essential crite-
rion to full in the design of efficient sensitizers for TiO2.
Indeed, each ligand is set to play one or more well-dened roles
such as anchoring, passivation of the surface and assisting
charge injection. Accordingly, ligands differ by their molecular
structures and therefore by their electronic nature. In the course
of our program on heteroleptic bis-diimine copper(I)
complexes,7–9 prepared according to the HETPHEN concept
developed by Schmittel and colleagues,10 we have prepared and
studied four new stable heteroleptic copper(I) complexes
[CuL0Ln]+ hereaer named Cn (n ¼ 1–4, Fig. 1).
[CuL0L1]+ (C1) was synthesized, with L1
¼
2,20,4,40-
tetramethylbipyridine.
The syntheses of all ligands are reported in the ESI.† The
HETPHEN modus operandi was used to isolate C1–4 and started
with the synthesis of the Cu(L0)]+ intermediate in DMF. An
equivalent of Ln was subsequently added dropwise, entailing an
immediate colour change of the medium from yellow to deep
red. Impurities were removed by size exclusion chromatog-
raphy. A similar protocol was used to isolate the dimethyl-ester
forms of each complex (named hereaer Cnester, n ¼ 1–4,
synthesis given in the ESI†).
The electronic absorption spectra of the complexes were
recorded in solution and on nanocrystalline TiO2 lms (Fig. 2
and S3†). All the complexes featured the classical MLCT
absorption band at ca. 500 nm (Table 1 and Fig. 2).11 The
increased conjugation of the p system on both L0 and Ln (n ¼ 2–
The anchoring ligand L0 (6,60-dimesityl-2,20-bipyridine-4,40-
dicarboxylic acid) is based on the classical 4,40-dicarboxylic acid
a
´
´
UNAM, Universite Nantes, Angers, Le Mans, CEISAM, Chimie Et Interdisciplinarite,
`
´
`
Synthese, Analyse, Modelisation CNRS, UMR CNRS 6230, 2, rue de la Houssiniere -
BP 92208, 44322 Nantes Cedex 3, France. E-mail: Fabrice.Odobel@univ-nantes.fr;
Mohammed.Boujtita@univ-nantes.fr
b
´
UMR CNRS 6226-Universite de Rennes 1, Institut des Sciences Chimiques de Rennes,
Campus de Beaulieu, 35042 Rennes Cedex, France
† Electronic supplementary information (ESI) available: Synthesis of the
complexes, absorption spectra on TiO2 and electron lifetime (sn), mean electron
transit time (str) and hcoll measured by IMVS and IMPS. See DOI:
10.1039/c4ta01755b
´
‡ Current address: CEMCA UMR CNRS 6521, Universite de Bretagne Occidentale,
6 avenue Victor Le Gorgeu, 29238 Brest, France.
Fig. 1 Molecular structures of Ln and Cn (n ¼ 1–4).
9944 | J. Mater. Chem. A, 2014, 2, 9944–9947
This journal is © The Royal Society of Chemistry 2014

View Article Online
Communication
Journal of Materials Chemistry A
measurements were performed on the diester forms of
complexes. The latter featured the expected, reversible copper-
centred oxidation around 1 V vs. SCE (Table 1). The voltam-
mograms of C3ester and C4ester displayed an additional
oxidation wave at 0.95 and 0.80 V vs. SCE respectively, corre-
sponding to the removal of one electron from the NR2 (R ¼ ethyl
or phenyl) amine moieties. Only differential pulse voltammetry
allowed discriminating the two close oxidation steps for
complex C4ester. The higher CuII/CuI potentials displayed by
C3ester and C4ester likely originate in the coulombic repulsion
between the copper cation and the electrogenerated hole on the
amine fragment.
The combination of electrochemical and UV-Vis data allowed
evaluating the Gibbs free energies associated with the various
charge transfer processes. In all cases, both charge injection
and dye regeneration are exergonic (ca. 300 meV, see Table S1 in
the ESI†). Energy-wise, C1–4 feature roughly the same behav-
iours. TiO2 electrodes were dipped while still hot for two days in
ethanolic solutions of C1–4 and the photovoltaic devices were
then assembled with a platinum counter-electrode, sealed with
a hotmelt polymer frame and their performances along those of
the reference benchmark N719 were evaluated under AM 1.5
calibrated articial sunlight (Table 2 and ESI† for details).
The weakest PCE is afforded by C1 based DSCs, grounded in
low photocurrent and photovoltage. The latter is assigned to a
lower light harvesting efficiency (LHE) and probably to an
exacerbated charge recombination with the electrolyte. Indeed,
the positive charge of C1–4 entails a coulombic repulsion
between them on the surface of TiO2, increasing the number of
unoccupied adsorption sites and thus recombination centres.
C3, C4 and C2 are a lot bulkier than C1, and thus passivate more
the surface of the semi-conductor, yielding a higher Voc. This is
further conrmed by the higher dark current displayed by C1-
based DSCs (see the ESI†). Besides, both C1 and C2 yield poor
photocurrents, likely because of their less intense absorption
coverages of the solar spectrum, leading to an overall weaker
LHE (see Fig. 3 and S4†).
Fig. 2 UV-Visible spectra of complexes C1 (dashed-dot), C2 (plain),
C3 (dot) and C4 (dash) recorded in dichloromethane.
Table 1 UV-Visible and electrochemical data for C1–4 recorded in
dichloromethane. * data collected with the methyl ester forms of
C1–4 and referenced versus SCE
l (nm) [3 (Mꢁ1 cmꢁ1)]
E (V)* [DE (mV)]
C1
C2
C3
C4
477 [4.7 ꢂ 103]
0.94 [96]
0.91 [96]
1.08a [—]
1.03b [—]
—
—
500 [9.8 ꢂ 103]
504 [1.3 ꢂ 104]
502 [1.4 ꢂ 104]
0.80a [—]
0.95b [—]
a
b
Ea, irreversible wave.
As the two waves merge, the square
voltammetry value is given.
4) induces stabilization of the p* orbitals, explaining the red-
shi of this transition compared to the benchmark bis–neo-
cuproine Cu(I) complex C5 (Fig. S7†).9 One notices that the
MLCT bands are more intense as well, because of the increased
ground state dipolar moment generated by the combination of
electron poor L0 and electron rich Ln. The complexes C3 and C4
present higher light harvesting efficiency in the visible than C1
and C2 because of an intense additional intraligand charge
transfer transition (ILCT), located at the edge of the visible
around 420 nm.9
This very intense ILCT transition corresponds to a shi of
the electron density from the electron rich amine moieties to
the electron poor pyridine. Such a band does exist for C2 too,
but is signicantly blue-shied compared to C3 and C4 because
of the poorer electron donating power of L2. Spectra recorded on
TiO2 transparent electrodes (Fig. S3†) feature the same patterns
as those recorded in the solution phase (Fig. 2). Overall, the
complexes displayed a rather broad and intense absorption over
a large wavelength frame (lonset ꢀ 620 nm), revealing their
potential as wide band gap semi-conductor sensitizers.
No luminescence was detected upon excitation in the MLCT
band, regardless of the conditions. This could be due to cis–
trans isomerization of the vinyl double bond12 or the lesser
rigidity of the Cu–bpy coordination cage compared to Cu–Phen,
facilitating the deleterious exciplex quenching.
The short circuit currents of C3 and C4 based DSCs are by far
the highest of the series, in part because of the presence of ILCT
bands in the visible domain, increasing the LHE. This is
conrmed by the incident photon to current efficiency (IPCE)
recorded for each DSC, where current generation is indeed
Table 2 Photovoltaic data for DSCs based on TiO2 sensitization by
C1–4 without (a) and with (b) CDCA. Voc: open circuit voltage; Jsc
:
short circuit current density; FF ¼ fill factor
Voc (mV)
Jsc (mA cmꢁ2
)
FF (%)
PCE (%)
C1(a)
C2(a)
C3(a)
C4(a)
N719
C1(b)
C2(b)
C3(b)
C4(b)
475
535
545
565
635
525
565
605
625
2.20
2.89
7.51
6.70
16.87
3.76
4.99
10.86
10.13
72.80
72.54
71.52
73.32
68.69
74.64
72.39
70.97
69.76
0.76
1.12
2.93
2.77
7.36
1.47
2.04
4.66
4.42
To record better resolved cyclic and pulsed voltammograms
(no adsorption of the dye on the electrode), these
This journal is © The Royal Society of Chemistry 2014
J. Mater. Chem. A, 2014, 2, 9944–9947 | 9945

View Article Online
Journal of Materials Chemistry A
Communication
upon CDCA treatment, along with a slightly broadened MLCT
transition (Fig. S4†). These subtle changes in the absorption
spectra of the chemisorbed complexes are in line with the
IPCE, and are probably grounded in a reorganization of the
dye monolayer upon CDCA adsorption. The role of CDCA is
oen associated with the disruption of dye aggregates and
certainly comes into play here, especially due to the presence
of organic styryl branches on the complexes C2–4. Based on
the effect of CDCA on both IPCE measurements and on the Jsc
enhancement, we conclude that the main role of CDCA with
these complexes is certainly to decrease the aggregation on the
TiO2 surface leading to higher LHE and injection quantum
yield. In these conditions, DSCs provided a maximum PCE of
4.66% for the C3-based device. This is to date the highest PCE
ever reported for a DSC based on a copper(I) complex sensi-
tizer, and holds great promise for the future of these cheap
Fig. 3 IPCE for DSCs sensitized with C1 (dashed-dot), C2 (dot), C3
(plain), and C4 (dash) recorded with CDCA.
monitored between 400 and 460 nm for C3 and C4 (around 43% solar cells. Most highly performing dyes, including ruthenium
at 410 nm). C1 and C2 based DSCs, being deprived of such ILCT complexes, are neutral species, while this rst series of cop-
above 400 nm consequently display lesser LHE and IPCE.
per(I) complexes are positively charged. This is certainly one
Spin coating a 0.1 M CDCA (chenodeoxycholic acid) etha- weak point of these dyes, which can be overcome by using new
nolic solution onto the photo-electrodes prior to the nal seal- ancillary ligands.
ing is anticipated to eliminate the deleterious self-quenching
process induced by aggregation. Rewardingly, unprecedented
improvements in the power conversion efficiencies (PCEs) of all
Conclusions
DSCs were observed upon such CDCA surface treatment. First of We successfully isolated four stable heteroleptic copper(I)–poly-
all, an increase of the photopotential was observed for all DSCs. pyridine complexes, using the HETPHEN concept. Through a
C1–4 based devices exhibited a 50–60 mV rise of Voc, together careful choice of ligands, unprecedented PCE was achieved,
with a decrease of the dark current. This improvement was reaching 4.66%. The new anchoring ligand L0 paves the route to
therefore assigned to the higher electron concentration in the prepare other sensitizers as it certainly forms stable heteroleptic
CB and to a passivation of recombination sites by the co- copper(I) with many unhindered diimine ligands. This contri-
adsorbent molecules. In the case of C2, the octyl chains may bution brings further credit to these molecular complexes as
provide a built-in, efficient protection for titanium dioxide's efficient sensitizers for DSCs, en route for a cheap and less toxic
surface, thus explaining the lesser increase of Voc (ca. 30 mV). substitute to ruthenium dyes.
The electron lifetime (sn) and mean transit time (str) of photo-
injected charge carriers were then recorded by intensity-
modulated photovoltage spectroscopy (IMVS) and intensity-
Acknowledgements
modulated photocurrent spectroscopy (IMPS). However, these The ANR agency is gratefully acknowledged for the nancial
measurements reveal that no signicant improvement of both support of these research studies through program HeteroCop
sn and sd was observed when CDCA was added in the prepara- (no. ANR-09-BLAN-0183-01).
tion of the series of solar cells (see the ESI†). As a result, the
charge collection efficiency hcoll measured as a function of
the illumination intensity is quite similar for all the dyes (see
Notes and references
the ESI†).
The most spectacular improvement of the PCE originates in
1 B. O'Regan and M. Gratzel, Nature, 1991, 353(6346), 737–
740.
the rise of Jsc for all devices. One calculates a 70% increase of the
photocurrent for C1 and C4 and 45 and 50% for C2 and C3
based photovoltaic devices, respectively. For a better under-
standing, IPCEs of DSCs with and without CDCA treatment were
compared (Fig. S5†). A signicant increase of the IPCE is
observed for CDCA-treated cells, regardless of the dye (Fig. 3),
together with broadening of the signals. Several reasons can be
invoked to rationalize this important result. First, CDCA mole-
cules release protons and this bends the conduction band
downwards, thus increasing the driving force of the electron
injection into the semi-conductor, and consequently improving
the electron injection yield.13 Second, a noticeable increase of
the absorbance of C1–4 based photo-electrodes was monitored
2 A. Hagfeldt, G. Boschloo, L. Sun, L. Kloo and H. Pettersson,
Chem. Rev., 2010, 110(11), 6595–6663.
3 A. Mishra, M. K. R. Fischer and P. Bauerle, Angew. Chem., Int.
Ed., 2009, 48, 2474–2499.
4 S. Sakaki, T. Kuroki and T. Hamada, J. Chem. Soc., Dalton
Trans., 2002, 840–842.
5 B. Bozic-Weber, E. C. Constable, S. O. Furer,
C. E. Housecro, L. J. Troxler and J. A. Zampese, Chem.
Commun., 2013, 49(65), 7222–7224.
6 B. Bozic-Weber, E. C. Constable and C. E. Housecro, Coord.
Chem. Rev., 2013, 257(21–22), 3089–3106.
7 Y. Pellegrin, M. Sandroni, E. Blart, A. Planchat, M. Evain,
N. C. Bera, M. Kayanuma, M. Sliwa, M. Rebarz, O. Poizat,
9946 | J. Mater. Chem. A, 2014, 2, 9944–9947
This journal is © The Royal Society of Chemistry 2014

View Article Online
Communication
Journal of Materials Chemistry A
C. Daniel and F. Odobel, Inorg. Chem., 2011, 50(22), 11309– 11 N. Armaroli, G. Accorsi, F. Cardinali and A. Listorti,
11322.
Photochemistry and Photophysics of Coordination
8 M. Sandroni, M. Kayanuma, A. Planchat, N. Szuwarski,
Compounds I, Top. Curr. Chem., 2007, vol. 280, pp. 69–115.
E. Blart, Y. Pellegrin, C. Daniel, M. Boujtita and F. Odobel, 12 V. Aubert, L. Ordronneau, M. Escadeillas, J. A. G. Williams,
Dalton Trans., 2013, 42(30), 10818–10827.
A. Boucekkine, E. Coulaud, C. Dragonetti, S. Righetto,
D. Roberto, R. Ugo, A. Valore, A. Singh, J. Zyss, I. Ledoux-
Rak, H. Le Bozec and V. Guerchais, Inorg. Chem., 2011,
50(11), 5027–5038.
9 M. Sandroni, M. Kayanuma, M. Rebarz, H. Akdas-Kilig,
Y. Pellegrin, E. Blart, O. Poizat, H. Le Bozec, M. Sliwa,
C. Daniel and F. Odobel, Dalton Trans., 2013, 42, 10818–
10827.
13 M. K. Nazeeruddin, R. Humphry-Baker, P. Liska and
¨
10 M. Schmittel and A. Ganz, Chem. Commun., 1997, 999–1000.
M. Gratzel, J. Phys. Chem. B, 2003, 107(34), 8981–8987.
This journal is © The Royal Society of Chemistry 2014
J. Mater. Chem. A, 2014, 2, 9944–9947 | 9947