Electropolymerisable bipyridine ruthenium(II) complexes: Synthesis, spectroscopic and electrochemical characterisation of 4-((2-thienyl) ethenyl)- and 4,4′-di((2-thienyl) ethenyl)-2,2′ -bipyridine ruthenium complexes

Article abstract of DOI:10.1016/j.poly.2003.10.004

Four new ruthenium polypyridyl complexes with mono- or di-((2-thienyl) ethenyl) substituted bipyridines have been synthesized. The complexes were characterized by NMR, elemental analysis, UV-Vis absorption and electrochemistry (differential pulse and cyclic voltammetry). Electroactive polymer films of these complexes have been prepared by oxidative electropolymerisation and characterized by UV-Vis absorption spectroscopy and electrochemistry. The electrochemically induced polymerisation of the complexes resulted in a significant shift of the oxidation potential of the Ru(II)-Ru(III) process towards more positive potentials. Also, MLCT absorption band of the polymeric complexes is shifted towards shorter wavelengths. These results are interpreted in terms of an interruption of the conjugated system of the (2-thienyl)ethenyl-substituted bipyridine ligands due to a radical polymerisation mechanism affecting rather the ethenyl part of the ligand than the thienyl.

Full text of DOI:10.1016/j.poly.2003.10.004

Polyhedron 23 (2004) 589–598
www.elsevier.com/locate/poly
Electropolymerisable bipyridine ruthenium(II) complexes:
synthesis, spectroscopic and electrochemical characterisation of
4-((2-thienyl) ethenyl)- and 4,40-di((2-thienyl) ethenyl)-2,20-bipyridine
ruthenium complexes
a
Viviane Aranyos , Anders Hagfeldt , Helena Grennberg , Egbert Figgemeier
a
b
a,
*
a
Department of Physical Chemistry, University of Uppsala, P.O. Box 579, Uppsala S-75123, Sweden
b
Department of Organic Chemistry, University of Uppsala, P.O. Box 599, Uppsala S-75124, Sweden
Received 16 July 2003; accepted 21 October 2003
Abstract
Four new ruthenium polypyridyl complexes with mono- or di-((2-thienyl) ethenyl) substituted bipyridines have been synthesized.
The complexes were characterized by NMR, elemental analysis, UV–Vis absorption and electrochemistry (differential pulse and
cyclic voltammetry). Electroactive polymer films of these complexes have been prepared by oxidative electropolymerisation and
characterized by UV–Vis absorption spectroscopy and electrochemistry. The electrochemically induced polymerisation of the
complexes resulted in a significant shift of the oxidation potential of the Ru(II)–Ru(III) process towards more positive potentials.
Also, MLCT absorption band of the polymeric complexes is shifted towards shorter wavelengths. These results are interpreted in
terms of an interruption of the conjugated system of the (2-thienyl)ethenyl-substituted bipyridine ligands due to a radical poly-
merisation mechanism affecting rather the ethenyl part of the ligand than the thienyl.
Ó 2003 Elsevier Ltd. All rights reserved.
Keywords: Electropolymerisation; Ru-complexes; Thiophene; Differential pulse voltammetry; 4-((2-Thienyl)-ethenyl)-2,20-bipyridine; Redox-
polymers
1. Introduction
other side, the highly efficient nano-structured solar cells
based on TiO2 as electron acceptor are incorporating
liquid electrolytes for the transport of charge, which is
problematic in terms of stability and production. With
this background it can be very beneficial to combine the
advantages of conducting polymers and nano-structured
TiO2 in the design of a solar cell. In this context we are
currently investigating Ru-complexes, which can act as
dye sensitizers in NSCs, which are modified with thio-
phene substituents on the bipyridine ligands. These
thienyl groups are thought to be used as starting points
for a covalent link between the dye and a solid electro-
lyte as, e.g., polythiophene. Also, an extension of the
absorbtion spectrum together with an increased hydro-
phobicity improving stability of the NSC towards water
penetration gives these substituents a bright future in the
field of photoelectrochemical solar cells. However, in
order to apply this type of dyes in a complex system such
Nano-structured dye-sensitized photoelectrochemical
solar cells (NSCs) and solar cells based on conjugated
polymers are considered as promising routes to decrease
significantly the price of electricity by means of solar
power in the future [1,2]. Moreover, these types of solar
cells have features like flexibility and lightweight, which
give them an even higher market potential. Nevertheless,
polymer-based solar cells are currently using either
Buckminsterfullerenes or polymers as electron acceptors
which is costly or triggers stability problems. On the
* Corresponding author. Present address: Departement Chemie,
€
Universitat Basel, Spitalstrasse 51, Basel CH-4056, Switzerland. Tel.:
+46-18-471-3661; Fax: +46-18-471-3654.
E-mail address: egbert.figgemeier@fki.uu.se (E. Figgemeier).
0277-5387/$ - see front matter Ó 2003 Elsevier Ltd. All rights reserved.
doi:10.1016/j.poly.2003.10.004

590
V. Aranyos et al. / Polyhedron 23 (2004) 589–598
Scheme 1.
as the NSC, the study of models compounds 5–8
(Scheme 1) is a necessity.
under vacuum at 120 °C for 48 h before performing
electrochemistry. Syntheses were performed under a
nitrogen atmosphere. Transparent conducting oxide
electrodes of fluorine-doped tin oxide (FTO) on a 3-mm
glass support (sheet resistance 8 X/square) were pur-
chased from Liffey–Owens–Ford. Elemental analyses
were performed by the Analytische Laboratorien,
Lindlar, Germany.
In the past we have presented the synthesis and
properties of ruthenium tris-bipyridyl complexes with
methoxystyryl substituents and the characteristics of
electrochemically produced polymers [3]. Following up
this investigation, we here report the synthesis and the
study of some ruthenium polypyridyl complexes with
hydrophobic polymer precursors as ethylene-thiophene.
Electroactive polymers of these complexes have been
prepared by oxidative electropolymerisation. The re-
sulting materials have been characterized with UV–Vis
spectroscopy and by means of electrochemistry.
2.2. Physical characterisation
NMR spectra (1H NMR at 400 MHz and 13C NMR
at 100 MHz) were recorded on a Varian Unity 400
spectrometer for CDCl3 and CD3COCD3 solutions
unless otherwise stated. The chloroform signals at 7.26
1
2. Experimental
ppm (for H) or 77.0 ppm (for 13C) and the acetone
1
signals at 2.04 ppm (for H) or 29.8 ppm (for 13C) were
used as indirect reference to TMS.
2.1. Materials
UV–Vis absorption spectra were recorded using a
Hewlett-Packard 8453 diode-array spectrophotometer
and referenced against a solvent blank (CH3CN). The
fluorine-doped tin oxide (FTO) electrodes with polymer
films deposited on the surface were positioned against
the wall of a 1-cm quartz cuvette containing CH3CN.
The absorption of the FTO glass substrates was sub-
stracted from the measured spectra of the compound.
Electrochemistry was performed using a CH Instru-
ments Model 660. Electrochemical workstation and a
conventional three electrode cell with a silver wire as
pseudo-reference electrode. After the experiments fer-
rocene was added as an internal reference and all
potentials given are quoted against the ferrocene–
ferrocenium redox couple.
All reagents were from commercial sources and were
used without further purification. Solvents were HPLC
grade and were used without further purification with
the following exceptions: Tetrahydrofuran (THF) was
distilled from sodium, diisopropylamine (DIPA) and
triethylamine (Et3N), which were distilled from CaH2.
For electrochemical experiments, acetonitrile (Fluka,
H2O 6 0.001%) stored over molecular sieves was used.
Butyl lithium (BuLi) was stored in a desiccator under
nitrogen and used as fresh (no precipitate) 1.6 M solu-
tions in hexane. Trifluoroacetic acid (TFA) and trifluo-
roacetic acid anhydride (TFAA) were stored at )18 °C
and used as received. Tetrabutylammonium hexaflu-
orophosphate ((TBA)PF6) (Aldrich, 98%) was dried

V. Aranyos et al. / Polyhedron 23 (2004) 589–598
591
2.3. Synthesis
2.3.1. Ligands
2.3.1.3. 4-Methyl-40-(2-(2-thiophene) ethenyl)-2,20-bi-
pyridine (3).
4-Methyl-40-(2-hydroxy-2-(2-thienyl)
ethyl)-2,20-bipyridine (2.30 mmol) was dissolved in 10
mL of CH2Cl2 under nitrogen. As trifluoroacetic anhy-
dride (TFA) (3.90 mL, 27.6 mmol) was added dropwise,
the reaction mixture became orange. The mixture was
stirred overnight and then concentrated to dryness,
dissolved into EtOAc (50 mL), washed with water
(2 ꢀ 50 mL), dried over MgSO4 and concentrated to
dryness. The product was afforded as yellow crystals in
2.3.1.1. 4-Methyl-40-(2-hydroxy-2-(2-thienyl) ethyl)-
2,20-bipyridine (1). LDA was formed by mixing 1.54
mL BuLi (1.6 M in hexane) with 0.39 mL of DIPA in 10
mL THF at )60 °C under N2. The mixture was then
stirred at room temperature for 1/2 h. Thereafter, 4,40-
dimethyl-2,20-bipyridine (0.50 g, 2.48 ꢀ 10ꢁ3 mol) dis-
solved in 15 mL THF was added through a double
tipped needle after the temperature was cooled to
)40 °C. The black reaction mixture was then stirred at
room temperature during 1 h. The aldehyde (0.25 mL,
2.72 ꢀ 10ꢁ3 mol) was diluted in 10 mL THF and added
the same fashion to the reaction mixture to form a yel-
low precipitate. After stirring for 1 h, the reaction was
stirred into water and extracted with dichloromethane.
The organic phase was then washed with brine and dried
over MgSO4. The dried solution was then concentrated.
NMR of the crude product showed more than 85% of
conversion compared to the aldehyde and complete
conversion compared to the bipyridine. The crude
product was used in the next step without further puri-
1
96% yield (0.616 g, 2.21 mmol). H NMR (400 MHz)
(CDCl3) d 8.85 (1H, d, J ¼ 5:6 Hz, aromatic pyridine),
8.78 (1H, J ¼ 5:2 Hz, d, aromatic pyridine), 8.55 (1H,
broad s, aromatic pyridine), 8.35 (1H, broad s, aromatic
pyridine), 7.81 (1H, d, J ¼ 16 Hz, alkene), 7.60 (1H, dd,
J ¼ 5:6 Hz, J ¼ 1:2 Hz, pyridine), 7.52 (1H, d, J ¼ 5:2
Hz, pyridine), 7.42 (1H, d, J ¼ 5:6 Hz, thiophene), 7.33
(1H, d, J ¼ 3:6 Hz, thiophene), 7.11 (1H, dd, J ¼ 3:6
Hz, J ¼ 5:2 Hz, thiophene), 6.93 (1H, d, J ¼ 16 Hz,
alkene), 2.63 (3H, s, Methyl). 13C NMR, 154.7, 151.2,
148.3, 147.9, 146.2, 146.0, 140.6, 131.6, 131.7, 130.9,
128.5, 128.4, 127.3, 124.4, 122.7, 122.4, 119.6, 21.8.
2.3.1.4. 4,40-(2-(2-thiophene) ethenyl)-2,20-bipyridine
(4). Crude 2 (2.04 ꢀ 10ꢁ3 mol) was dissolved in 15 mL
of CH2Cl2 under nitrogen. As trifluoroacetic anhydride
(TFAA) (3.90 mL, 27.6 ꢀ 10ꢁ3 mol) was added drop-
wise, the reaction mixture turned orange. The mixture
was stirred overnight and then concentrated to dryness,
dissolved into EtOAc (50 mL), washed with water
(2 ꢀ 50 mL), dried over MgSO4 and concentrated to
dryness. The product was afforded as a yellow powder in
37% yield (0.281 g, 7.54 ꢀ 10ꢁ3 mol). 1H NMR (400
MHz) (CDCl3) d 8.73 (1H, d, J ¼ 5:2 Hz, aromatic
pyridine), 8.65 (1H, broad s, aromatic pyridine), 7.70
(1H, d, J ¼ 16 Hz, alkene), 7.41 (1H, d, J ¼ 4:8 Hz,
pyridine), 7.32 (1H, d, J ¼ 4:8 Hz, thiophene), 7.24 (1H,
m, thiophene), 7.06 (1H, dd, J ¼ 3:6 Hz, J ¼ 4:8 Hz,
thiophene), 6.94 (1H, d, J ¼ 16 Hz, alkene). 13C NMR,
148.1, 141.5, 131.1, 129.7, 129.4, 129.0, 128.4, 127.4,
124.2, 122.3, 119.3.
1
fication. H NMR (400 MHz) (CDCl3) d 8.54 (1H, d,
J ¼ 4:8 Hz, aromatic pyridine), 8.50 (1H, J ¼ 4:8 Hz, d,
aromatic pyridine), 8.30 (1H, broad s, aromatic pyri-
dine), 8.21 (1H, broad s, aromatic pyridine), 7.26 (1H,
multiplet, thiophene), 7.13 (2H, multiplet, aromatic
pyridine), 6.96 (2H, multiplet, thiophene), 5.29 (1H, dd,
J ¼ 7:6 Hz, J ¼ 6 Hz, benzylic alcohol), 3.21 (1H, d,
J ¼ 7:6 Hz, benzylic), 3.21 (1H, d, J ¼ 6 Hz, benzylic),
2.43 (3H, s).
2.3.1.2. 4,40-(2-hydroxy-2-(2-thienyl) ethyl)-2,20-bipyri-
dine (2). LDA was formed by mixing 3.72 mL BuLi (1.6
M in hexane) with 0.83 mL of DIPA in 20 mL THF at
)60 °C under N2. The mixture was then stirred at room
temperature for 1/2 h. Thereafter, 4,40-dimethyl-2,20-bi-
pyridine (0.50 g, 2.48 ꢀ 10ꢁ3 mol) dissolved in 15 mL
THF was added through a double-tipped needle after
the temperature was cooled to )40 °C. The black reac-
tion mixture was then stirred at room temperature
during 1 h. The aldehyde (0.52 mL, 5.96 ꢀ 10ꢁ3 mol) was
diluted in 10 mL THF and added in the same fashion to
the reaction mixture to form a yellow precipitate. After
stirring for 1 h, the reaction was stirred into water and
extracted with dichloromethane. The organic phase was
then washed with saturated aq. NaCl and dried over
MgSO4. The dried solution was then concentrated
yielding 0.836 g (2.04 ꢀ 10ꢁ3, 75%) of the product.1H
NMR (400 MHz) (CDCl3) d 8.58 (2H, m, aromatic
pyridine), 8.30 (2H, broad s, aromatic pyridine), 7.26
(2H, multiplet, thiophene), 7.15 (2H, multiplet, aromatic
pyridine), 6.99 (4H, multiplet, thiophene), 5.29 (2H, m,
benzylic alcohol), 3.21 (2H, m, benzylic).
2.3.2. Homoleptic complexes
2.3.2.1. Tris-(4-methyl-40-(2-thienylethenyl)-2,20-bipyri-
dine) ruthenium dihexafluorophosphate (5). RuCl3 ꢂ
xH2O (0.079 g, 0.38 mmol) and L (0.321 g, 1.15 mmol)
were stirred at reflux in a 20 mL mixture of EtOH and
H2O (1:1), under nitrogen during 24 h. The solution
turned dark red. NH4PF6 (1.24 g, 7.66 mmol) dissolved
in water was added to the reaction mixture, after which
a brown precipitate appeared. The precipitate was fil-
tered off, washed with water and ether. The product was
dried under reduced pressure during 24 h and was
afforded in 24% yield (0.115 g, 0.095 mmol). Anal. Calc.
for Ru(Bpy-Th1)3(PF6)2 + 4H2O (Mol Form.: C51H42

592
V. Aranyos et al. / Polyhedron 23 (2004) 589–598
F12N6P2RuS3 ꢂ 4H2O): C, 47.19; H, 3.88; N, 6.47.
tuted bipyridine, m, 4H), 7.98 (–CH@CH–, d, J ¼ 16
Hz, 1H), 7.91 (substituted bipyridine, d, J ¼ 6 Hz, 1H),
7.86 (substituted bipyridine, d, J ¼ 5:4 Hz, 1H), 7.64
(substituted bipyridine, dd, J1 ¼ 6:4 Hz, J2 ¼ 2 Hz, 1H),
7.62–7.55 (non-substituted bipyridine, m, 4H), 7.42
(substituted bipyridine, unresolved dd, 1H), 7.38
(substituted bipyridine, d, J ¼ 3:2 Hz, 1H), 7.13
(substituted bipyridine, dd, J1 ¼ 5:2 Hz, J2 ¼ 4 Hz, 1H),
2.59 (–CH3, s, 3H). Anal. Calc. for C37H30F12N6P2RuS ꢂ
2H2O: C, 43.66; H, 3.37; N, 8.26; Found: C, 43.82; H,
3.21; N, 8.09%.
Found: C, 46.82; H, 3.89; N, 5.85%.
2.3.2.2. Tris-(4,40-(2-thienylethenyl)-2,20-bipyridine) ru-
thenium dihexafluorophosphate (6). RuCl3 ꢂ xH2O (0.079
g, 0.89 mmol) and L (0.100 g, 0.268 mmol) were stirred
at reflux in a 20 mL mixture of EtOH and H2O (1:1),
under nitrogen during 24 h. The solution turned dark
red. NH4PF6 (0.145 g, 0.89 mmol) dissolved in water
was added to the reaction mixture, after which a brown
precipitate appeared. The precipitate was filtered off,
washed with water and ether. The product was dried
under reduced pressure during 24 h and was afforded in
68% yield (0.092 g, 0.45 mmol). Anal. Calc. for Ru(Bpy-
Th2)3(PF6)2 + 5H2O (Mol Form.: C66H48F12N6P2
RuS6 ꢂ 5H2O): C, 49.59; H, 3.66; N, 5.26; Found: C,
49.23; H, 3.83; N, 4.50%.
2.3.3.3. Bis-2,20-bipyridyl-(4,40-(2-thienylethenyl)-2,20-
bipyridine) ruthenium dihexafluorophosphate (8). Bis
2,20-bipyridyl ruthenium dichloride (0.050 g, 0.10 mmol)
and 4,40-(2-thienylethenyl)-2,20-bipyridine (0.038 g, 0.10
mmol) were stirred at reflux in a 20 mL mixture of EtOH
and H2O (1:1), under argon for 1 h. The solution turned
red. NaPF6 (0.250 g, ꢃ20 eq) was added to the reaction
mixture, which was stirred overnight (18 h), resulting in
an orange precipitate. The precipitate was filtered and
dried under reduced pressure during 24 h to give 0.065 g
(0.6 mmol, 60% yield) of bis-2,20-bipyridyl-(4,40-(2-thie-
nylethenyl)-2,20-bipyridine) ruthenium dihexafluoro-
2.3.3. Heteroleptic complexes
2.3.3.1. Bis-2,20-bipyridyl ruthenium dichloride [4].
RuCl3 ꢂ xH2O (0.200 g, 0.96 mmol), 2,20-bipyridyine
(0.316 g, 2.02 mmol) and LiCl (0.173 g, 4.08 mmol) were
stirred at reflux in 20 mL reagent grade DMF under
argon for 8 h. After the reaction had cooled to room
temperature, 100 mL of reagent grade acetone was ad-
ded. The solution was then refrigerated overnight at 0
°C. Filtration gave a red violet solution and dark green
crystals. Those were washed with water (3 ꢀ 15 mL) and
with ether (3 ꢀ 15 mL). The product was then dried at
1
phosphate. H NMR (CD3COCD3) d 8.98 (substituted
bipyridine, s, 2H), 8.81 (non-substituted bipyridine, d,
J ¼ 8 Hz, 4H), 8.21 (non-substituted bipyridine, t, J ¼ 8
Hz, 4H), 8.16 (substituted bipyridine, d, J ¼ 5:6 Hz,
2H), 8.05 (substituted bipyridine, d, J ¼ 5:2 Hz, 2H),
7.97 (–CH@CH–, d, J ¼ 16 Hz, 2H), 7.91 (substituted
bipyridine, d, J ¼ 8 Hz, 2H), 7.68 (dd, J1 ¼ 6:4 Hz,
J2 ¼ 2 Hz, 2H), 7.62–7.55 (non-substituted bipyri-
dine + substituted bipyridine, m, 4 + 2H), 7.40 (substi-
tuted bipyridine, broad d, J ¼ 3:8 Hz, 2H), 7.68
(substituted bipyridine + –CH@CH–, dd + d, J1 ¼ 5:2
Hz, J2 ¼ 3:8 Hz, J3 ¼ 16 Hz, 4H). 13C NMR, 157.7,
157.5, 152.0, 151.6, 146.8, 141.37, 138.2, 130.1, 129.8,
128.6, 128.1, 124.6, 124.1, 123.4, 122.6, 121.1. Anal.
Calc. for C42H32F12N6P2RuS2 ꢂ 8H2O: C, 41.35; H, 3.97;
N, 6.89; Found: C, 40.94; H, 3.41; N, 7.10%.
1
room temperature under reduced pressure for 24 h. H
NMR (CDCl3) d 10.07 (d, J ¼ 5:6 Hz, 2H, aromatic),
8.15 (d, J ¼ 8 Hz, 2H, aromatic), 7.99 (d, J ¼ 8 Hz, 2H,
aromatic), 7.84 (t, J ¼ 8 Hz, 2H, aromatic), 7.55 (m, 4H,
aromatic), 7.46 (t, J ¼ 8 Hz, 2H, aromatic), 6.87 (t,
J ¼ 5:6 Hz, 2H, aromatic). 13C NMR, 160.4, 157.9,
153.9, 152.6, 134.5, 133.4, 125.5, 124.7, 121.9, 121.7.
2.3.3.2.
Bis-2,20-bipyridyl-(4-methyl-40-(2-thienylethe-
nyl)-2,20-bipyridine) ruthenium dihexafluorophosphate
(7). Bis-2,20-bipyridyl ruthenium dichloride (0.050 g,
0.10 mmol) and 4,40-(2-thienylethenyl)-2,20-bipyridine
(0.027 g, 0.10 mmol) were stirred at reflux in a 20 mL
mixture of EtOH and H2O (1:1), under argon for 1 h.
The solution turned red. NaPF6 (0.250 g, ꢃ20 eq) was
added to the reaction mixture, which was stirred over-
night (18 h), resulting in an orange precipitate. The
precipitate was filtered and dried under reduced pressure
during 24 h to give 0.043 g (0.044 mmol, 42% yield) of
bis-2,20-bipyridyl-(4-methyl-40-(2-thienylethenyl)-2,20-bi-
3. Results and discussion
3.1. Synthesis
A two-step synthesis was used to afford 4-methyl-40-
(2-(2-thiophene) ethenyl)-2,20-bipyridine (ligand 3), or
4,40-(2-(2-thiophene) ethenyl)-2,20-bipyridine (ligand 4)
as pictured in Scheme 2.
First, a nucleophilic addition of the mono- or di-an-
ion of 4,40-dimethyl-2,20-bipyridine to 2-thiophenecar-
boxaldehyde (one or two equivalents) gave the
corresponding alcohol (1 or 2) [4]. Dehydration was
achieved using trifluoroacetic anhydride (TFAA) fol-
lowed by precipitation. The preparation of the homo-
1
pyridine) ruthenium dihexafluorophosphate. H NMR
(CD3COCD3) d 8.95 (substituted bipyridine, d, J ¼ 1:2
Hz, 1H), 8.81 (non-substituted bipyridine, d, J ¼ 7:2
Hz, 4H), 8.76 (substituted bipyridine, s, 1H), 8.23-8.18
(non-substituted bipyridine, m, 4H), 8.16 (substituted
bipyridine, d, J ¼ 0:8 Hz, 1H), 8.15–8.04 (non-substi-

V. Aranyos et al. / Polyhedron 23 (2004) 589–598
593
Scheme 2.
leptic complexes 5 and 6 was done in methoxyethanol
(MeOEtOH) by mixing one equivalent of the ruthenium
salt with three equivalents of the desired ligand (1 or 2).
After the reaction mixture became dark red, the solution
was cooled and the complex precipitated out of the so-
lution by addition of (NH4PF6)-saturated water. The
complex was filtered and dried under vacuum.
The heteroleptic complexes were prepared using a
two-step synthesis. The dichloro bis-bipyridyl ruthe-
nium complex was prepared according to a literature
procedure [5]. The third ligand was introduced by mix-
ing the dichloro bis-bipyridyl ruthenium salt to one
equivalent of the mono- or di-substituted ligands fol-
lowed by an exchange of the counter ion by addition of
(NH4PF6)-saturated water.
Fig. 1. Absorption spectra for both the conjugated and the non-con-
jugated thiophene substituted ligands 2 (-s-) and 4 (-d-) [7].
3.2. Absorption properties
The absorption spectra for the ligands 2 and 4 were
compared and clearly showed the substituent influence
1
(see Fig. 1): The very typical Ligand Centered Charge
Transfer (LCCT/p–pꢄ) for bipyridine ligands is repre-
sented around 280 nm in both cases, and the conjugated
thiophene gives an extra LCCT at ꢃ340 nm [6].
The absorption properties of the different complexes
have been recorded in acetonitrile and are shown in
1
Fig. 2.
The metal to ligand charge transfer (MLCT) corre-
lates very well with the number of substituents on the
molecules (see Table 1).
The more extended conjugated system is expected to
have a smaller energy difference between its HOMO and
its LUMO, which corresponds to a shift of the MLCT
towards longer wavelengths.
Fig. 2. Absorption spectra for the different complexes 5–8 (complex 5:
-s-; complex 6: -j-; complex 7: -}-; complex 8: -M-) (Footnote 1).
The extinction coefficients reflect as well the struc-
tures of the complexes: The more extended conjugation
on the complex leads to a shift of the MLCT towards
longer wavelengths and a higher e-value, although the
variation between eMLCT for 7 and 8 is not significant,
showing the little difference between the presence of one
or two thiophene groups as substituents.
Table 1
Absorption properties of complexes 5–8 in acetonitrile
a
a
Complex
LCCT
(nm)
eLCCT
MLCT
(nm)
eMLCT
(molꢁ1 cmꢁ1
)
(molꢁ1 cmꢁ1
)
5
6
7
8
362
368
350
369b
61 700
76 950
20 900
25 200
483
493
458
463
29 500
35 000
16 600
15 700
a Extinction coefficient for the absorption maximum in the visible
region.
b The relatively high value is due to an artefact from the absorption
spectrum. The average would be around 358.
1
The symbols are for orientation purposes only and the spectra were
measured with a 1 nm resolution.

594
V. Aranyos et al. / Polyhedron 23 (2004) 589–598
Table 2
Electrochemical data for complexes 5–8 as determined by differential
pulse voltammetrya;b
Complex
Oxidation (V)
Reduction (V)
3
4
1.08
1.08
)1.24
)1.24
(1.05)
1.12
5
6
7
8
(0.75)
0.72
)1.61 )1.80 )1.99
)1.58 )1.70 )1.92
)1.71 )1.91 )2.14
)1.62 )1.89 )2.12
)1.69 )1.90 )2.14
)1.75 )1.93 )2.17
(0.98)
1.14
(0.74)
0.74
(1.21)
1.35
(0.90)
0.85
(1.20)
1.51
(0.90)
0.85
0.93
[Ru(bpy)3]2þ
[Ru(4,40dsty-bpy)3]2þ
[Ru(Me2bpy)3]2þ
Fig. 3. Cyclic voltammogram of complex 7. A glassy carbon electrode
was used and the voltammogram was measured in acetonitrile with
0.83
0.82
(TBA)PF6 as supporting electrolyte at a scan speed of 0.1 V sꢁ1
.
a In Volts. A glassy carbon electrode was used as working electrode,
the redox couple Fc/Fcþ was used as internal reference and (TBA),
PF6 as electrolyte. The scan rate was 100 mV sꢁ1. The solvent was
acetonitrile.
Th), the separation of the effects on the electronic
structure within the complexes is difficult, but literature
data of related compounds point out that both substit-
uents are electron donating, which is stabilizing the
oxidized state of the complex (Table 2) [7,8].
b The values given in brackets are measured in dichloromethane
3.3. Electrochemical properties
Beside the Ru(II)–(III) transition another – irrevers-
ible – oxidation process was observed for all complexes
5–8. We assign these signals to a ligand-centred oxida-
tion of the ethylene–thiophene substituent. A compari-
son with the electrochemistry of the uncoordinated
ligand supports this assumption (see Table 2). Also in
the literature, an oxidation of thiophene substituted
polypyridine complexes was described at similar poten-
tials [9,10].
Besides the influence of the different ligands on the
oxidation potentials, an effect of the solvent was ob-
served: Whereas the Ru(II)–Ru(III) transition remains
at the same position when switching the solvent, the
peaks assigned to the ethylene–thiophene oxidation
were shifted by more than 100 mV to less positive
potentials when changing the solvent from acetonitrile
to CH2Cl2. To a smaller extent, such an effect was
observed for thiophene oligomers for which the authors
found a shift of 50 mV to less positive potentials when
switching the solvent from acetonitrile to dichlorome-
thane [11]. In the same contribution it was pointed out
that the radical produced due to electrochemical oxi-
dation is strongly stabilized by dichloromethane, which
also was described generally before [12]. Therefore, we
assign the shift of the second oxidation of the com-
plexes to a solvent induced energetic stabilisation of the
produced radical.
The electrochemistry of complexes 5–8 was mea-
sured. All complexes were electrochemically active in
CH2CL2 and acetonitrile solution with 0.1 M (TBA)PF6
2
as supporting electrolyte. For comparison purposes
also the ligands 3 and 4 were electrochemically investi-
gated. The results are summarized in Table 2.
The redox potentials listed were measured by differ-
ential pulse voltammetry (DPV). A significant difference
to numbers determined by cyclic voltammetry (CV) was
not observed.
3.3.1. Oxidation
All complexes 5–8 showed in CV experiments at least
one fully reversible peak in the oxidative region of the
potential window, which can be assigned to the oxida-
tion of Ru(II)–Ru(III). The position of the signal de-
pends on the substitution of the bipyridine ligands. This
oxidation characteristic is illustrated in Fig. 3.
As can be seen from Table 1, the Ru(II)–Ru(III) re-
dox process of the homoleptic complexes 5 and 6 is
shifted by slightly more than 100 mV towards less po-
sitive potentials in comparison to the heteroleptic com-
plexes 7 and 8 and 200 mV relative to [Ru(bpy)3]2þ. This
experimental result can be interpreted as a stabilisation
of the Ru(III)-ion in these complexes relative to
[Ru(bpy)3]3þ. Having two chemically very different
substituents on the bpy-ligands (–CH3 and –CH@CH–
3.3.2. Reduction
Beside the electrochemical characteristics in the oxi-
dative region of the potential window, the reduction of
complexes 5–8 and the ligands 3 and 4 were investigated.
The results are summarized in Table 2.
2
The complexes 5 and 6 showed only poor solubility in acetonitrile
and dichloromethane (<0.5 ꢀ 10ꢁ4 mol/l), which is reflected in the low
intensity of the electrochemistry shown in Fig. 3.

V. Aranyos et al. / Polyhedron 23 (2004) 589–598
595
Fig. 4. DPV for compounds 5–8 (reductive region of the potential
window). A glassy carbon electrode was used as working electrode, the
redox couple Fc/Fcþ was employed as internal reference and
(TBA)PF6 (0.1 M in acetonitrile) as electrolyte.
Fig. 5. Cyclic voltammogram for complex 6. A glassy carbon electrode
was used as working electrode, the redox couple Fc/Fcþ was employed
as internal reference and (TBA)PF6 (0.1 M in dichloromethane) as
electrolyte at a scan rate of 100 mV sꢁ1
.
3.3.3. Electrochemical polymerisation at positive poten-
tials
For all complexes at least three signals were detected
in CV and DPV (see Fig. 4) experiments at negative
potentials within the potential window of acetonitrile
and CH2Cl2 with (TBA)PF6 as the supporting electro-
lyte, which correspond to the successive reduction of the
three bipyridine ligands [13].
For complexes 5, 6 and 8, an increase in intensity of
the oxidation waves in the CV experiments were ob-
served from one cycle to the next (see Fig. 5): The first
scan gives an oxidation wave for the ruthenium center as
well as a strong oxidation corresponding to the irre-
versible oxidation of the thiophene groups [14].
As expected from this type of complexes, the differ-
ence in potential between the processes is close to 200
mV. A further reduction of the bpy-ligands was also
The oxidation waves merge together after the first
scan in combination with a substantial growth of the
peaks. One can also see a red-orange film deposited on
the working electrode after the experiment. These effects
was only seen when the potential window of the vol-
tammetry experiments reached the oxidation of the
complexes associated with the oxidation of the ethylene-
thiophene ligand substituents. We assign these obser-
vations to a radical induced polymerisation as other
authors described it for similar complexes leading to
three-dimensional networks with the Ru in the back-
bone [15–21]. This polymerisation process was mainly
observed in dichloromethane, whereas a stable effect in
acetonitrile could only be seen for complex 6. As dis-
cussed earlier, a stabilisation of radical cations was ob-
served in dichloromethane relative to acetonitrile. It
seems likely that besides the shift of the oxidation po-
tential, the stabilisation of radicals favours a successful
polymerisation, since the longer lifetime increases
the probability of finding a thiophene counterpart for
polymerisation. Also, the shift of the ligand centred
oxidation towards less positive potential in dichlorom-
ethane increased the accessibility of the process within
the potential window, decreasing the interferences with
traces of water or solvent related processes. A typical
example for the electrochemically induced polymerisa-
tion is given in Fig. 5 for complex 6.
3
observed as expected but not analysed [13].
Relative to the reduction potentials of [Ru(bpy)3]2þ
,
the values for the homoleptic complexes 5 and 6 are
significantly shifted to less negative potentials for all
three ligand centred processes. It can be assumed that
the expansion of the conjugated system by the substi-
tution of the bpy ligand with ethylene–thiophene
groups is responsible for the lowering of the LUMO.
On the other side, the methyl substituents seem to have
a smaller or even opposite effect since the reductions of
complex 5 are located at more negative potentials than
for complex 6. For the heteroleptic complex 7, no sig-
nificant shift of the reduction potentials relative to
[Ru(bpy)3]2þ were observed pointing out that methyl
groups are rather lifting the LUMO level. This can be
confirmed by a comparison with [Ru(bpyMe2)3]2þ
(Table 2). For complex 8, we measured a less negative
reduction process only for one ligand whereas the other
two reductions are taking place at the same potentials
as the equivalent processes in [Ru(bpy)3]2þ. This al-
lows us to attribute the first reduction of complex 8 to
a localisation on the ethylene-thiophene substituted
bpy-ligand.
3
Differential pulse voltammetry had to be deployed for the
By cyclic voltammetry and DPV, one can see a tre-
mendous shift (over 100 mV) of the oxidation potential
of the Ru(II)–Ru(III) redox wave, indicating a change of
determination of redox potentials since the cyclic voltammograms
were dominated for the compounds 5, 6 and 8 by polymerisation
processes as discussed in the following section.

596
V. Aranyos et al. / Polyhedron 23 (2004) 589–598
Table 3
Absorption and electrochemical properties of the complexes 5–8 (non-
polymerized and oxidatively polymerized on FTO-glass support)
E0 (V) (vs. Fc/Fcþ)a
Complex
kmax, MLCT (nm)
Monomer
Polymer
Monomer
Polymer
5
6
7
8
483
493
485
463
472
475
0.72
0.74
0.85
0.85
0.85
1.07
450
1.02
a E0 was determined in dichloromethane. The electrodes modified
with a polymer were removed from the complex solution after poly-
merisation, rinsed with dichloromethane and then measured in di-
chloromethane/(TBA)PF6 without the corresponding complex.
Fig. 7. Cyclic voltammogram of the polymer of complex 5 in CH2Cl2
with (TBA)PF6 as supporting electrolyte and scan rate dependency of
the peak current (inset).
The electrochemistry of the polymers was also mea-
sured in monomer-free electrolyte. A typical cyclic vol-
tammogram is presented in Fig. 7, and the values for the
oxidation potentials are given in Table 3.
In all cases we find only one oxidation peak, which
we relate to the Ru(II)–Ru(III) redox couple, whereas
no further ligand-located oxidation as for the monomer
was detected within the potential window. For a surface
bound species, the separation between the oxidation and
reduction peak is even at moderate scan rates rather
high for a surface bound species. For the given example
it is 130 mV at a scan rate of only 20 mV sꢁ1, which is an
indication of slow electron transport and transfer pro-
cesses [15]. On the other side, the scan rate dependency
of the peak current shows the linear behaviour expected
for surface bound electro-active material, where the scan
rate dependency is not influenced by its diffusion to and
Fig. 6. DPV for complex 6. A glassy carbon electrode was used as
working electrode, the redox couple Fc/Fcþ was used as internal ref-
erence and (TBA),PF6 (0.1 M in acetonitrile) as electrolyte. First scan –
solid line and second scan - - - - dashed line.
the ligandÕs substituents polarity (see also Table 3). This
effect can be seen most clearly for the polymerisation of
4
complex 6 in acetonitrile. In Fig. 6 two successive
5
DPVs are shown: The first scan shows two distinct peaks
for the Ru(II)–Ru(III) and the ligand centred process,
whereas the second scan only shows one strong peak –
presumably the oxidation of the polymeric material on
the electrode surface.
As discussed previously, the low oxidation potentials
of complexes 5–8 relative to [Ru(bpy)3]2þ has its origin
in the increased size of the conjugated system of the bpy-
ligand, therefore the strong shift of the oxidation po-
tential is indicative of a loss of this conjugation. At the
same time, the disappearance of the strong oxidation
peak at 1.42 V is indicative of either a consumption or a
cut of the thiophene groups from the conjugated system
moving the oxidation potential of the group outside the
potential window due to the polymerisation.
from the electrode (see Fig. 7, inset).
Two possible mechanisms for the polymerisation can
be taken into account and are pictured in Scheme 3.
Thiophenes are known to dimerize via a radical cat-
ion mechanism through the 2- and 5-positions (Scheme
3A) [22], which eventually leads for complexes with
multiple thiophene substituents on the ligands to poly-
thiophene type polymers with metal complexes in the
backbone. However, in the present case the positive
5
From the area under the redox signals, the charge and therefore the
absolute number of complex entities deposited on the electrode and
incorporated into the polymer network can be calculated. For
example, the evaluation of the redox peaks shown in Fig. 7 gives a
total charge of 3.5 ꢀ 10ꢁ5 C, which corresponds to 3.6 ꢀ 10ꢁ10 mol of
complex 5 deposited on the electrode. Taking into account the surface
of the electrode and that a monolayer of these complexes includes
about 7.7 ꢀ 10ꢁ11 mol cmꢁ2 (J. Am. Chem. Soc. 116 (1994) 5444), the
thickness of the polymer is equivalent to 66 monolayers and therefore
the thickness of the polymer layer is in the order of 50–100 nm after 16
4
The same effect was also observed for complexes 5, 6 and 8 in
dichloromethane, but due to the close vicinity of the ligand and metal
centred oxidations, the shift of oxidation due to polymerisation is less
obvious than for the electrochemistry of 6 in acetonitrile.
ꢀ
deposition cycles, assuming an average Ru–Ru distance of 15–20 A.

V. Aranyos et al. / Polyhedron 23 (2004) 589–598
597
4. Conclusions
We have reported the synthesis, characterisation and
polymerisation behaviour of four thiophene substituted
ruthenium complexes. The polymerisation behaviour of
the complexes at positive potentials showed the behav-
iour of a redox active film. No polymerisation was ob-
served at negative potentials showing the low reactivity
of the vinyl functionality in these complexes. The shift of
the MLCT band of the absorption spectra towards
shorter wavelengths in combination with an oxidation
potential at less positive potentials implied a loss of
conjugation within the ligands 3 and 4 due to a radical
polymerisation through the vinyl functionality after
delocalisation instead of the expected thiophene–thio-
phene coupling. Nevertheless, an interesting feature for
an application in the NSC is the broader absorption
spectrum of the dye compared to a ruthenium tris-
bipyridyl complex. Also, the introduction of a thio-
phene-group makes this type of dye a possible base to
form a solid-state solar cell by polymerisation. These
compounds, used as a model, lay the ground for the
development of new dyes possessing broader absorption
properties and better water tolerance for solar cell
application.
Scheme 3.
charge can delocalize towards the olefinic double bond,
which will eventually dimerize (Scheme 3B) and there-
fore offers a alternative pathway for reaction of the
oxidized species [22–26].
Absorption spectra of the polymers on FTO-glass
have been recorded, and the influence of the polymeri-
sation method is clearly seen, as exemplified for 6 (see
Fig. 8) and listed in Table 3.
The blue shift of the MLCT, which we find for all
oxidatively produced polymers, points towards a loss of
conjugation as it is depicted in Fig. 8. The disappearance
of the LC-transition for the substituted ligand at 360 nm
confirms the loss of conjugation in the ligand structure
(see Fig. 8).
Acknowledgements
The authors thank the University of Uppsala and the
ꢀ
Angstrom Solar Center for support. E.F. acknowledges
€
the European Community for the financial support
(Marie-Curie Fellowship contract no. MCFI-2001-
01264).
References
[1] C.J. Brabec, N.S. Sariciftci, J.C. Hummelen, Adv. Funct. Mater.
11 (2001) 15.
€
[2] A. Hagfeldt, M. Gratzel, Acc. Chem. Res. 33 (2000) 269.
[3] V. Aranyos, J. Hjelm, A. Hagfeldt, H. Grennberg, J. Chem. Soc.
Dalton Trans. (2001) 1319.
[4] B.P. Sullivan, D.J. Salmon, T.J. Meyer, Inorg. Chem. 17 (1978)
3334.
[5] T.A. Heimer, S.T. DÕArcangelis, F. Farzad, J.M. Stipkala, Inorg.
Chem. 35 (1996) 5319.
[6] A. Juris, V. Balzani, F. Barigilletti, S. Campagna, P. Belser, A.
von Zelewsky, Coord. Chem. Rev. 84 (1988) 85.
[7] V. Skarda, M.J. Cook, A.P. Lewis, G.S.G. McAuli, A.J. Thom-
son, D.J. Robbins, J. Chem. Soc., Perkin Trans. 2 (1984) 1293.
[8] P.A. Mabrouk, M.S. Wrighton, Inorg. Chem. 25 (1986) 526.
[9] J. Hjelm, E.C. Constable, E. Figgemeier, A. Hagfeldt, R. Handel,
C.E. Housecroft, E. Mukhtar, E. Schofield, Chem. Commun.
(2002) 284.
Fig. 8. Absorption spectra of polymer films of 5, deposited on glass
with a conducting layer of fluorine doped tin oxide (FTO) (solid line),
and of the monomer 5 in solution (dashed line).
[10] T.M. Pappenfus, K.R. Mann, Inorg. Chem. 40 (2001) 6301.
[11] P. Hapiot, L. Gaillon, P. Audebert, J.J.E. Moreau, J.-P. Lere-
Porte, M. Wong Chi Man, Synth. Met. 72 (1995) 129.
ꢁ

598
V. Aranyos et al. / Polyhedron 23 (2004) 589–598
[12] H Lund, in: H Lund, M.M. Baizer (Eds.), Organic Electrochem-
istry, 3rd ed., Marcel Dekker, New York, 1991, p. 301.
[13] A. Juris, V. Balzani, F. Barigelletti, S. Campagna, P. Belser, A.
von Zelewsky, Coord. Chem. Rev. 84 (1988) 85.
[20] T. Shimidzu, H. Segawa, F. Wu, N. Nakayama, J. Photochem.
Photobiol. A: Chem. 92 (1995) 121.
[21] R.P. Kingsborough, T.M. Swager, J. Mater. Chem. 9 (1999) 2123.
[22] J. Roncali, Chem. Rev. 92 (1992) 711.
[14] B.J. MacLean, P.G. Pickup, J. Chem. Soc. Chem. Commun.
(1999) 2471.
[15] M.O. Wolf, Adv. Mater. 13 (2001) 545.
[23] V. Ruiz, A. Colina, A. Heras, J. Lopez-Palacios, R. Seeber, Helv.
Chim. Acta 84 (2001) 3628.
[24] Refs. [29] and [30] from previous reference: J. Heinze, P.
Tschuncky, A. Smie, J. Solid State Electrochem. 2 (1998) 102.
[25] Refs. [29] and [30] from previous reference: J.R. Lenhard, R.L.
Parton, J. Am. Chem. Soc. 109 (1987) 5808.
[16] A.J. Bard, L.R. Faulkner, Electrochemical Methods: Fundamen-
tals and Applications, Wiley, New York, 1980.
[17] R.C. Leidner, B.P. Sullivan, R.A. Reed, B.A. White, M.T.
Crimmins, R.W. Murray, T.J. Meyer, Inorg. Chem. 26 (1987) 882.
[18] P.D. Beer, O. Kocian, R.J. Mortimer, C. Ridgeway, J. Chem. Soc.
Faraday Trans. 89 (1993) 333.
[26] M. Pagels, J. Heinze, B. Geschke, V. Rang, Electrochim. Acta 46
(2001) 3943, Citation: ÔRecent studies of the oxidation of radical
ions containing olefinic double bonds have shown that such
radicalic species dimerises at one of the olefinic carbon atoms,
forming dimers. The dimerisation process may be reversible.’.
[19] H. Segawa, F.P. Wu, N. Nakayama, H. Maruyama, S. Nagisaka, N.
Higuchi, M. Fujitsuka, T. Shimidzu, Synth. Met. 71 (1995) 2151.