Phosphonic acid anchored ruthenium complexes for ZnO-based dye-sensitized solar cells

Article abstract of DOI:10.1016/j.dyepig.2013.12.018

We report on the development of ruthenium dyes for the application in flexible ZnO based dye-sensitized solar cells. The ZnO based solar cells were prepared by electrodeposition using eosin Y as a structure-directing agent. The newly synthesized ruthenium dyes differed in the complexity of the extended π-donor-system and in their anchor moieties. As alternatives to carboxylic acids as anchor groups, dyes carrying phosphonic acids were studied. Comparison of the dyes contrary to what could be expected showed that the phosphonic acid anchored dyes were not superior to the carboxylic anchored dyes. A surprising second effect was that the dyes with the least sophisticated ligands performed best. Exploring possible reasons we established a simple model that allows pre-evaluation of the applicability of the dyes before testing them in real solar cells.

Full text of DOI:10.1016/j.dyepig.2013.12.018

Dyes and Pigments 104 (2014) 24e33
Contents lists available at ScienceDirect
Dyes and Pigments
journal homepage: www.elsevier.com/locate/dyepig
Phosphonic acid anchored ruthenium complexes for ZnO-based
dye-sensitized solar cells
,
b
c
a
b
Katja Neuthe a , Florian Bittner , Frank Stiemke , Benjamin Ziem , Juan Du ,
*
Monika Zellner c, Michael Wark d, Thomas Schubert c, Rainer Haag a
a Institut für Chemie und Biochemie e Organische Chemie, Freie Universität Berlin, Takustr. 3, 14195 Berlin, Germany
b Institut für Physikalische Chemie und Elektrochemie, Leibniz Universität Hannover, Callinstr. 3a, 30167 Hannover, Germany
c IOLITEC Ionic Liquids Technologies GmbH, Salzstr. 184, 74076 Heilbronn, Germany
d Institute for Chemistry e Photocatalysis and Sustainable Feedstock Utilization, Carl-von-Ossietzky University Oldenburg, Carl-von-Ossietzky-Str. 9-11,
26129 Oldenburg, Germany
a r t i c l e i n f o
a b s t r a c t
Article history:
We report on the development of ruthenium dyes for the application in flexible ZnO based dye-
sensitized solar cells. The ZnO based solar cells were prepared by electrodeposition using eosin Y as a
structure-directing agent. The newly synthesized ruthenium dyes differed in the complexity of the
Received 22 August 2013
Received in revised form
15 December 2013
Accepted 19 December 2013
Available online 28 December 2013
extended
p-donor-system and in their anchor moieties. As alternatives to carboxylic acids as anchor
groups, dyes carrying phosphonic acids were studied. Comparison of the dyes contrary to what could be
expected showed that the phosphonic acid anchored dyes were not superior to the carboxylic anchored
dyes. A surprising second effect was that the dyes with the least sophisticated ligands performed best.
Exploring possible reasons we established a simple model that allows pre-evaluation of the applicability
of the dyes before testing them in real solar cells.
Keywords:
ZnO
Dye sensitized solar cells
Ruthenium (II) complex
Phosphonate anchor
Photovoltaics, Evaluation model
Ó 2013 Elsevier Ltd. All rights reserved.
1. Introduction
usually carry two different ligand moieties, anchor and donor,
which may influence the cell performance.
In the context of renewable energies, dye sensitized solar cells
(DSSCs) are considered to be a useful alternative source of energy.
DSSCs employ a dye that is adsorbed onto mesoporous semi-
conductor nanoparticles with a high specific surface area. Upon
excitation of the dye by sunlight an electron can be injected from
the dye into the conducting band of the semiconductor. Besides
TiO2, the semiconductor material that is most commonly used, ZnO
is a promising alternative for application in DSSCs. In addition to its
higher electron mobility [1,2], it can be easily fabricated by elec-
trodeposition [3] at low temperatures and consequently also on
flexible plastic substrates, which is advantageous for later industrial
production processes and the exploration of novel application
areas. Different dyes can be used for sensitizing the semiconductor
material. Ruthenium complexes were introduced as sensitizers for
DSSCs by Grätzel [4] in 1991 and have become superior to other
metal complexes and organic dyes [1]. Ruthenium based dyes
Some of the selected conversion energy values reported for ZnO
based DSSCs prepared by low temperature methods on various
substrates are 5.6% on glass [3], 4.17% on PEN-ITO [5], on 2.5% CNT [6].
Our goal was to synthesize dyes for application in ZnO based,
flexible DSSCs. As shown in Fig. 1, ruthenium complexes were
designed, synthesized, and evaluated. As it is known that dyes
adsorbed via a carboxylic anchor might lead to dissolution of the
ZnO surface [7], dyes were synthesized with a phosphonate anchor
for better adsorptivity [8] and enhanced efficiency [9].
2. Experimental
2.1. Materials and reagents
Chemicals used for synthesis and electrochemical measure-
ments and deposition procedures were commercially available and
used without further purification. Distilled or reagent grade sol-
vents were preferred. Solvents for reactions under dry conditions
were taken from a MBraun MB SPS-800 solvent purification system
or purchased as ultra dry solvents from Acros Organics or Fluka. 4,40-
dimethyl-2,20-bipyridine was provided by IOLITEC. Column
* Corresponding author. Tel.: þ49 30 838 57288.
E-mail addresses: kneuthe@zedat-fu.berlin.de, kneuthe@zedat.fu-berlin.de
(K. Neuthe).
0143-7208/$ e see front matter Ó 2013 Elsevier Ltd. All rights reserved.
http://dx.doi.org/10.1016/j.dyepig.2013.12.018

K. Neuthe et al. / Dyes and Pigments 104 (2014) 24e33
C9H19
25
1
4
N
N
N
Cl
Cl
N
N
Ru+2
N
N
C9H19
N
Ru
N
N
PO3H2
PO3H2
PO3H2
PO3H2
C9H19
PO3H2
PO3H2
2
5
N
N
N
N
N
N
Ru+2
N
N
N
PO3H-
Ru+2
C9H19
N
PO3H-
Ph2N
N
N
PO3H2
PO3H-
PO3H-
PO3H2
C6H13
S
6 X = PO3H2
C101 X = COOH
COO-
3
N
N
N
N
NCS
Ru+2
N
COO-
N
N
NCS
X
Ru
Ph2N
S
C6H13
N
N
N
HOOC
COOH
Fig. 1. Molecular structures of the investigated dyes.
X
chromatography was carried out on silica gel 60 (230e400 mesh,
particle size 0.040e0.063 mm) and size exclusion chromatography
with Sephadex LH-20 from GE Healthcare.
Fabrication of ZnO films was carried out based on the commonly
used procedure [3] in aqueous solutions of KCl from Carl Roth
(ꢀ99.5%), ZnCl2 from ABCR (ꢀ98%) and eosin Y (EY) from Acros
Organics (ꢀ80%). FTO-coated glass substrates from Pilkington TEC
A8 (7 U sqꢁ1) were pre-treated in a mildly alkaline cleaning agent
(1% solution of deconexÒ 12 BASIC, Borer Chemie AG).
ZAHNER-Elektrik. The setup consisted of a main potentiostat Zen-
nium, an auxiliary potentiostat PP 211, a tunable light source TLS,
and a photosensor. The IPCE spectra were recorded in the wave-
length range from 430 to 700 nm. During the measurements the
solar cells were operated under short circuit conditions.
The IeV characteristics of the assembled DSSCs were measured
under simulated sunlight with an irradiance of 1000 W mꢁ2. A
Xenon arc lamp (Oriel Instruments 66901/69911), equipped with
filters to generate AM 1.5D conditions, served as the light source.
The light intensity was adjusted with a thermopile (Kipp & Zonen
CA2). An electrochemical workstation ZAHNER-Elektrik IM6e was
used to perform the measurements. The reported current densities
and conversion efficiencies refer to the photoactive area of 1.33 cm2
of the solar cells.
2.2. Analytical instruments and measurements
UV/Vis and emission spectra in solution were recorded in
methanol using a Scinco S-3100 UV/Vis spectrometer and an FP-
6500 spectrofluorometer (both from Jasco), respectively.
2.3. Synthesis of dyes
2.3.1. Anchor ligands
UV/Vis absorption spectra of sensitized ZnO films were recorded
with a Varian UV/Vis spectrophotometer Cary 4000 from Varian.
The spectra cover the wavelength range from 300 to 800 nm.
Cyclic voltammetry data were recorded on an Autolab
PGSTAT302N potentiostat (Metrohm Autolab BV). Measurements
were performed in a 3-cell compartment using a Pt disc working
electrode, a Pt wire counter electrode and a Ag/AgCl (1 M LiCl in
ethanol) reference electrode. The reference electrode was calibrated
against ferrocene before each measurement. Voltammogramms
were recorded in degassed DMF containing 0.1 M tetrabutylammo-
nium hexafluorophosphat (TBAP) as conducting salt.
2.3.1.1. 2,20-Bipyridine-4,40-diphosphonic acid (dpbpy), 7. This was
prepared following procedures reported by McKenna [10] and
Penicaud [11]. 4,40-dibromo-2,20-bipyridine (1.22 g, 3.88 mmol,
1.00 equiv.) was dissolved in a mixture of dry toluene (70.0 mL) and
triethylamine (1.30 mL) under an argon atmosphere. Diethyl
phosphite (1.15 mL, 8.92 mmol, 2.3 eq.), triphenylphosphine (10.2 g,
38.8 mmol, 10.0 equiv.), and tetrakis(triphenylphosphin)palla-
dium(0) (0.450 g, 0.400 mmol, 0.100 eq.) were added. The solution
was stirred at 110 ꢂC for 6 h. After cooling to room temperature the
reaction mixture was first extracted with ammonium hydroxide
The incident photon-to-current conversion efficiency (IPCE) was
recorded with a photoelectrochemical measurement setup from

26
K. Neuthe et al. / Dyes and Pigments 104 (2014) 24e33
solution and distilled water. The organic layer was dried over
MgSO4 and the solvent removed in vacuo. Purification by flash
column chromatography with DCM and MeOH afforded the
diphosphonic ethyl ester 8 as a white solid (1.42 g, 3.31 mmol,
23.3 mmol, 1.10 equiv.) was added dropwise. The solution was then
warmed to room temperature and stirred for another 1.5 h. After-
wards, trimethylstannyl chloride (8.75 g, 26.9 mmol, 1.28 eq.) was
added at ꢁ78 ꢂC, the mixture warmed to r.t. and stirred for 12 h. The
reaction was quenched with an excess of aqueous ammonium
chloride (150 mL, saturated). Adding an excess of water dissolved
the white precipitate formed. The phases were separated and the
aqueous phase extracted with DCM. The organic phases were
combined, dried over MgSO4, and the solvent removed in high
vacuo.
85.6%). 1H NMR (400 MHz; CDCl3):
d
in ppm ¼ 8.85 (t, J ¼ 5.0 Hz,
2H), 8.79 (d, J ¼ 14.2 Hz, 2H), 7.74 (dd, J ¼ 13.0, 4.8 Hz, 2H), 4.27e
4.14 (m, 8H),1.38 (t, J ¼ 7.1 Hz,12H), 31P NMR (500 MHz; CDCl3):
ppm ¼ 14.74 (s).
d in
Tetraethyl 2,20-bipyridine-4,40-diyldiphosphonate
8 (1.26 g,
2.94 mmol, 1.00 equiv.) was dissolved in dry DCM under an argon
atmosphere. Bromotrimethylsilane (4.50 g, 29.4 mmol, 10.0 equiv.)
was added and the yellow solution was stirred for 24 h at room
temperature. Afterwards, methanol was added dropwise until the
color of the mixture turned colorless. After extraction with water
2,20-bipyridine-4,40-diyldiphosphonic acid 7 was obtained as a pale
yellow solid (0.93 g, 4.49 mmol, quant.). 1H NMR (400 MHz; DMSO-
The obtained stannyl compound (9.56 g, 20.9 mmol,
3.8 equiv.) was then dissolved in degassed DMF (190 mL, abs.)
under an argon atmosphere and 4,40-dibromo-2,20-bipyridin
(1.74 g, 5.54 mmol,
1 equiv.) and Pd(PPh3)2Cl2 (0.220 g,
0.300 mmol, 0.054 equiv.) were added. The solution was stirred
at 85 ꢂC until a color change to dark red indicated completion of
the reaction (at least 12 h). The solvent was removed and the
crude product purified by column chromatography (SiO2, eluent:
CHCl3, then CHCl3:MeOH ¼ 20:1). Recrystallization from hexane
afforded the desired ligand 11 (1.24 g, 2.54 mmol, 51.0%.) 1H NMR
d6):
d
in ppm ¼ 8.85 (t, J ¼ 4.5 Hz, 2H), 8.67 (d, J ¼ 13.6 Hz, 2H), 7.73
(dd, J ¼ 12.4, 4.8 Hz, 2H), 31P NMR (500 MHz; CDCl3):
ppm ¼ 8.45 (s).
d in
2.3.1.2. 2,20-Bipyridine-4,40-dimethylenediphosphonic
acid
(500 MHz, CDCl3)
d
in ppm ¼ 8.64 (d, J ¼ 5.2 Hz, 4H), 7.53 (s, 1H),
(dmdpbpy), 9. 9 was prepared in accord with a literature procedure
[12]. 2,20-bipyridine-4,40 dimethylenephosphonic ethyl ester
(2.50 g, 5.50 mmol, 1.00 equiv.) were dissolved in a mixture of dry
acetonitril/DCM (200 mL, 1:1) under an argon atmosphere. Bro-
motrimethylsilane (10.0 mL, 75.7 mmol, 13.8 equiv.) was added and
the light orange solution was heated under reflux over night. The
reaction was quenched with water, resulting in a white precipitate.
Filtration and drying under high vacuum yielded the desired
bipyridine 9 as a white powder (1.60 g, 4.65 mol, 80%). 1H NMR
7.47 (d, J ¼ 6.4 Hz, 2H), 6.83 (d, J ¼ 3.6 Hz, 3H), 2.86 (t, J ¼ 7.6 Hz,
4H), 1.72 (p, J ¼ 7.6 Hz, 4H), 1.42e1.38 (m, 5H), 1.34e1.31 (m, 8H),
0.90 (t, J ¼ 7.0 Hz, 6H). HRMS (ESI) m/z: [M]þ calcd. for
(C30H36N2S2) 488.23; found 488.2.
2.3.2.3. (E)-4-(2-(40-methyl-[2,20-bipyridin]-4-yl)vinyl)-N,N-diphe-
nylaniline (mpabpy), 12. Diisopropylamine (2.10 mL, 13.3 mmol,
1.70 equiv.) was dissolved in dry THF (30.0 mL) under a nitrogen
atmosphere and the solution cooled to 0 ꢂC. n-butyl lithium in
hexane (2.50 M in hexane, 5.60 mL, 14.1 mmol, 1.80 equiv.) was
added and the solution stirred for 1 h. The solution was then
cooled to ꢁ78 ꢂC, 4,40-dimethyl-2,20-bipyridine (2.00 g,
10.9 mmol, 1.50 equiv.) added, warmed to 0 ꢂC and stirred for 2 h.
4-(N,N-diphenylamino)benzaldehyde (2.00 g, 7.30 mmol,
1.00 equiv.) was added at ꢁ78 ꢂC and the mixture was allowed to
warm to r.t. and stirred for 12 h. The reaction was quenched with
an excess of ammonium chloride solution (0.83 g, 15.5 mmol,
2.20 equiv.) and filtered. THF was removed in vacuo and the
residue extracted with DCM and water (3 ꢃ 50.0 mL). The organic
layer was dried over MgSO4 prior to removal of the solvent. The
crude product was purified by column chromatography (eluent
DCM:EtOAc ¼ 6:1) affording 1-(4-(N,N-diphenylamino)phenyl)-
2-(40-methyl- 2,20-bipyridin- 4-yl)ethanol) 12 as a yellow solid
(1.70 g, 3.72 mmol, 51.5%). 1H NMR (700 MHz; CDCl3): in
ppm ¼ 8.53 (ddd, J ¼ 1.9, 7.8, 3.9 Hz, 2H), 8.24 (dd, J ¼ 12.2, 1.9 Hz,
2H), 7.25e6.98 (m, 16H), 4.99 (dd, J ¼ 12.0, 5.6 Hz, 1H), 4.00e3.93
(m, 1H), 3.10 (d, J ¼ 5.5 Hz, 1H), 2.93e2.87 (m, 1H), 2.78e2.71 (m,
1H), 2.44 (d, J ¼ 2.3 Hz, 1H), 1.58e1.51 (m, 2H), 1.37e1.22 (m,
10H), 0.88 (t, J ¼ 6.8 Hz, 3H).
(400 MHz, DMSO-D6)
d
8.56 (d, J ¼ 5.0 Hz, 2H), 8.30 (s, 2H), 7.35e
7.30 (m, 2H), 3.12 (d, J ¼ 21.9 Hz, 2H). 31P NMR (400 MHz; DMSO-
D6):
d
in ppm ¼ 19.76 (s). HRMS (ESI) m/z: [M ꢁ H]ꢁ calcd. for
(C12H13N2O6P2) 343.02; found 343.03.
2.3.2. Donor ligands
2.3.2.1. 4,40-dinonyl-2,20-Bipyridine (dnbpy), 10. 10 was prepared by
modification of a literature procedure [13].
Diisopropylamine (28.0 mL, 0.160 mol, 2.66 equiv.) was dis-
solved in dry THF (60 mL) under an argon atmosphere. n-buty-
lithium (107.0 mL, 0.160 mol, 2.66 equiv.) was added dropwise at
0 ꢂC and the solution was stirred for 1 h. Afterwards, a solution of
4,40-dimethyl-2,20-bipyridine (11.0 g, 0.060 mol, 1 eq.) in dry THF
(300 mL) was slowly added and the orange solution was stirred
for 3 h. The reaction mixture was kept at 0 ꢂC and bromooctane
(28.0 mL, 0.160 mol, 2.66 equiv.) in dry THF (20 mL) was added.
The reaction mixture was then allowed to warm to room tem-
perature. After stirring for 12 h the reaction was quenched first
with water (20 mL), then poured into an ice/water mixture and
extracted with diethylether. The resulting yellow oil was recrys-
tallized from hexane, redissolved in DCM, extracted with diluted
NaOH, dried over MgSO4, filtered and the solvent removed in
vacuo. Drying in high vacuum yielded the desired product
The alcohol (1.70 g. 3.70 mmol, 1.00 equiv.) was then added to a
solution of glacial acetic acid (150 mL) and stirred for 12 h at 110 ꢂC.
The resulting red solution was poured into 100 mL DCM and
extracted with 1 M sodium hydroxide solution (3 ꢃ 50.0 mL) and
water (3 ꢃ 50.0 mL). The organic layer was dried over MgSO4,
filtered and the solvent removed. The resulting oil was purified by
column chromatography and the desired product 12 obtained as
yellow solid (0.80 g, 1.82 mmol, 49%). 1H NMR (700 MHz; CDCl3):
(11.02 g, 0.027 mol, 45%). 1H NMR (500 MHz, DMSO)
d in
ppm ¼ 8.85 (s, 1H), 8.83 (s, 1H), 7.28 (s, 1H), 7.27 (s, 1H), 2.68 (t,
J ¼ 7.6 Hz, 4H), 1.62 (p, J ¼ 8.2, 6.9 Hz, 4H), 1.32e1.11 (m, 24H),
0.83 (t, J ¼ 6.5 Hz, 6H). HRMS (ESI) m/z: [M þ H]þ calcd. for
(C28H44) 409.358; found 409.3581.
2.3.2.2. 4,40-Bis(2-hexylthiophene-5-yl)-2,20-Bipyridine (bhthpbpy),
11. 11 was prepared by modification of a literature procedure
[14,13].
2-hexylthiophene (4.30 mL, 87%, 21.0 mmol, 1.00 equiv.) was
dissolved in dry THF (50 mL) under an argon atmosphere, cooled
to ꢁ78 ꢂC and n-butyl lithium in hexane (1.60 M in hexane, 14.6 mL,
d
in ppm ¼ 8.62 (d, J ¼ 5.1 Hz, 1H), 8.60 (d, J ¼ 5.2 Hz, 1H), 8.49 (s,
1H), 8.38 (s, 1H), 7.44e7.38 (m, 3H), 7.36 (dd, J ¼ 5.2, 1.8 Hz, 1H),
7.31e7.26 (m, 4H), 7.25 (dd, J ¼ 5.1, 1.7 Hz, 1H), 7.15e7.04 (m, 8H),
7.00 (d, J ¼ 16.3 Hz, 1H), 6.64 (dt, J ¼ 15.8, 6.9 Hz, 1H), 6.43 (d,
J ¼ 15.9 Hz, 1H), 2.30e2.23 (m, 2H), 1.54e1.46 (m, 2H), 1.36e1.24
(m, 8H), 0.89 (t, J ¼ 7.0 Hz, 3H).

K. Neuthe et al. / Dyes and Pigments 104 (2014) 24e33
27
2.3.3. Complexes
(d, J ¼ 5.8 Hz, 1H), 7.64 (d, J ¼ 5.0 Hz,1H). HRMS (ESI) m/z: [M ꢁ Cl]þ
calcd. for (C24H16ClN4O8Ru-) 477.0056; found 477.0128.
Bi- and trileptic ruthenium(II) complexes were synthesized by
modification of literature procedures [14,15,16,17].
In a ethanol/water (1:1) mixture RuCl2dcbpy2 14 and mpabpy 12
were first stirred for 12 h at 100 ꢂC, and then for 12 h at r.t. ethanol
was removed under reduced pressure and water added. A yellow
precipitate formed which was removed by filtration. The red filtrate
was extracted with DCM. The aqueous layer was concentrated in
vacuo and the resulting residue purified by HPLC (RP-18, 12.5 cm,
RSC, 1.5 mL minꢁ1, 450 nm, 60/40 water/acetone with 0.6% acetic
acid). Acetonitrile was removed under reduced pressure and the
product precipitated from the acidic solution (pH ¼ 2). Filtration
yielded the complex 3 as a red solid (46.8 mg, 0.046 mmol, 10.0%).
2.3.3.1. [Ru(bpy)2(dpbpy)], 1. To 2,20-bipyridine (2.00 g, 12.8 mmol,
2.00 equiv.), dissolved in degassed DMF, under argon, and shielded
from light, LiCl (1.90 g, 44.8 mmol, 7.0 equiv.) and RuCl3∙H2O
(1.33 g, 6.41 mmol, 1.0 equiv.) were added. The resulting mixture
was first stirred for 24 h at 175 ꢂC, and then for 12 h at r.t. DMF was
removed in vacuo and an excess of acetone was added to the res-
idue. The solution was kept for 1e2 d at ꢁ12 ꢂC. A precipitate
formed, which was collected by filtration and washed with 3 por-
tions of cold acetone, yielding ruthenium precursor RuCl2bpy2 (13)
as dark crystals (1.12 g, 2.31 mmol, 38.0%). 1H NMR (700 MHz;
1H NMR (500 MHz, CD3OD: CD3CN ¼ 2:1):
in ppm ¼ 9.14 (m, 2H),
d
8.76 (s, 1H), 8.67 (s,1H), 8.61 (s,1H), 8.40 (s,1H), 8.07 (d, J ¼ 5.8,1H),
8.01 (d, J ¼ 5.8, 1H), 7.97e7.95 (m, 2H), 7.90 (s, 1H), 7.89 (s, 1H), 7.86
(s, 1H), 7.83 (s, 1H), 7.73 (d, J ¼ 16.3, 1H), 7.62e7.60 (m, 2H), 7.57e
7.55 (m, 2H), 7.50 (d, J ¼ 6.1, 1H), 7.39 (s, 1H), 7.36 (d, J ¼ 7.5, 4H),
7.34 (s, 1H), 7.25 (d, J ¼ 16.3, 1H), 7.18 (s, 1H), 7.15 (m, 5H), 7.12 (t,
J ¼ 1.1 Hz, 1H), 7.02 (dd, J ¼ 8.6, 5.3 Hz, 2H), 2.60 (s, 3H). HRMS (ESI)
m/z: [M ꢁ H]ꢁ1: calcd. for (C55H38N7O8Ruꢁ): 1026.1846; found
1026.1913.
DMSO-d6):
d
in ppm ¼ 9.98 (d, J ¼ 5.6 Hz, 2H), 8.64 (d, J ¼ 8.2 Hz,
2H), 8.48 (d, J ¼ 8.3 Hz, 2H), 8.06 (t, J ¼ 7.8 Hz, 2H), 7.77 (t, J ¼ 6.6 Hz,
2H), 7.68 (t, J ¼ 7.8 Hz, 2H), 7.51 (d, J ¼ 5.1 Hz, 2H), 7.10 (t, J ¼ 6.6 Hz,
2H), HRMS (ESI) m/z: [M ꢁ Cl þ CO]þ calcd. for (C21H16ClN4ORu)
477.0056; found 477.0128.
To RuCl2bpy2 13 (20.0 mg, 0.041, 1.00 equiv.) dissolved in water
(15 mL) 2,20-Bipyridine-4,40-diphosphonic acid
7
(20 mg,
0.041 mmol, 1.0 equiv.) was added and was heated under reflux for
18 h. The solvent was removed in vacuo and an excess of acetone
was added. The resulting solution was kept for 1e2 d at ꢁ12 ꢂC. A
precipitate formed, which was filtered and washed with cold
acetone (3 ꢃ 15 mL) to give [Ru(bpy)2(dpbpy)] 1 as dark brown
crystals (15.3 mg, 0.021 mmol, 51.1%). 1H NMR (700 MHz; DMSO-
2.3.3.4. [Ru(dnbpy)(dpbpy)Cl2], 4. To Dichloro-p-cymene ruthe-
nium dimer (0.050 g, 0.090 mmol, 1.00 equiv.) dissolved in dry,
degassed DMF (20 mL), under argon, and shielded from light was
added donor ligand dnbpy 10 (0.073 g, 0.180 mmol, 2.00 equiv.).
The mixture was stirred at 75 ꢂC for 5 h, cooled to room tem-
perature, and stirred for 12 h. After adding dpbpy 7 (0.150 g,
0.360 mmol, 4 equiv.) the reaction mixture was refluxed for 4 h
resulting in a dark-red solution. The solvent was removed in
vacuo and the crude product purified by size exclusion chroma-
tography on a Sephadex LH-20 column using water. Aqueous HCl
(6 M, 25 mL) was added to the main fraction and the resulting
solution was refluxed for 15 h. The black solution was purified by
size exclusion chromatography on a Sephadex LH-20 yielding 4
d6):
d
in ppm ¼ 8.95 (s, 2H, H-30; H-37), 8.84 (d, J ¼ 8.3 Hz, 4H, H-5;
H-12; H-17; H-24), 8.19e8.15 (m, 4H, H-2; H-9; H-14; H-21), 7.86
(dd, J ¼ 5.4, 3.2 Hz, 2H, H-27; H-34), 7.77 (d, J ¼ 5.3 Hz, 2H), 7.71 (d,
J ¼ 5.3 Hz, 2H), 7.64 (ddd, J ¼ 11.7, 5.7,1.2 Hz, 2H; H-26; H-35), 7.55e
7.51 (m, 4H, H-1; H-10; H-13; H-22).
2.3.3.2. [Ru(mpabpy)(dpbpy)2], 2. Dichloro-p-cymene ruthenium
dimer (27.6 mg, 0.045 mmol, 1.00 equiv.) dissolved in dry DCM
(20 mL), under argon, shielded from light by aluminum foil, was
reacted with mpabpy 12 (39.60 mg, 0.090 mmol, 2.00 equiv.) at r.t.
for 12 h. DCM was removed in vacuo to afford the corresponding
[RuCymClL] precursor with L ¼ mpabpy as a red solid (61.6 mg,
95.6%).
(40.0 mg, 0.045 mmol, 28.0%). 1H NMR (500 MHz, CD3OD):
d in
ppm ¼ 8.60 (s, 4H), 7.66 (s, 4H), 7.40 (s, 2H), 7.33 (s, 2H), 4.02e
4.01 (m, 2H), 2.84 (t, J ¼ 5.8, 2H), 1.82e1.69 (m, 4H), 1.39e1.20 (m,
24H). 0.88 (t, J ¼ 6.5, 6H). HRMS (ESI) m/z: [M ꢁ Cl]þ: calcd. for
(C39H57ClN4O6P2Ruꢁ): 876.25.; found 876.28.
To the [RuCymClL] precursor (44.6 mg, 0.063 mmol,1.00 equiv.),
dissolved in distilled water (20.0 mL), under argon, shielded from
light by aluminum foil, dpbpy 7 (39.9 mg, 0.126 mmol, 2.00 equiv.)
was added. After refluxing for 24 h the dark red solution was
poured into DCM and extracted with a water/methanol (1:1)
mixture. After freeze-drying, ruthenium complex 2 was received as
a red powder (39.3 mg, 0.034 mmol, 53.3%). 1H NMR (700 MHz;
2.3.3.5. [Ru(dnbpy)(dmdpbpy) 2], 5. To Dichloro-p-cymene ruthe-
nium dimer (0.050 g, 0.090 mmol, 1.00 equiv.) dissolved in dry,
degassed DMF (20 mL), under argon, and shielded from light
donor ligand dnbpy 10 (0.073 g, 0.180 mmol, 2.00 equiv.) was
added. The mixture was stirred at 75 ꢂC for 5 h, cooled to room
temperature and stirred for 12 h. After adding anchor dmdpbpy 9
(0.160 g, 0.36 mmol, 4.00 equiv.) the reaction mixture was
refluxed for 4 h resulting in a dark-red solution. The solvent was
removed in vacuo and the crude product purified on a Sephadex
LH-20 column using water. Aqueous HCl (6 M, 25 mL) was added
to the main fraction and the resulting mixture was refluxed until
IR measurements confirmed disappearance of the P-O-alkyl
signal (1000e920 cmꢁ1). The crude product was purified by size
MeOD-d4):
d
in ppm ¼ 8.87 (s, 1H, H-12), 8.85 (s, 1H, H-2), 8.69 (d,
J ¼ 12.8 Hz, 5H, H-9; H-24; H-31; H-38; H-45), 8.66 (t, J ¼ 4.4 Hz, 5H,
H-5; H-27; H-34; H-41; H-48), 7.84 (dd, J ¼ 5.4, 2.9 Hz, 1H, H-4),
7.75 (dd, J ¼ 11.7, 4.7 Hz, 6H, H-17; H-21; H-23; H-32; H-37; H-46),
7.73e7.68 (m, 4H, H-55; H-57; H-60; H-62), 7.59e7.51 (m, 2H, H-10;
H-15), 7.34e6.97 (m, 9H, H-14; H-18; H-20; H-54; H-56; H-58; H-
59; H-61; H-63), 2.69 (s, 3H, CH3). 31P NMR (500 MHz; MeOD-d4):
d
in ppm ¼ 7.60 (s), 4.83 (d, J ¼ 32.7 Hz), 4.61 (s). HRMS (ESI) m/z:
[M calcd. for (C51H41N7O12P4Ru): 1170.0874; found
H]ꢁ1
1170.0701.
ꢁ
:
exclusion chromatography on a Sephadex LH-20 yielding
5
(160.0 mg, 0.134 mmol, 96.0%). 1H NMR (700 MHz, CD3OD):
d
in
ppm 8.63 (s, 5H), 7.68 (s, 5H), 7.42 (s, 4H), 7.34 (s, 2H), 2.89e2.83
(m, 4H), 1.78e1.73 (m, 4H), 1.46e1.22 (m, 24H), 0.93e0.87 (m,
2.3.3.3. [Ru(mdpabpy)(dcbpy)2],
3. Ruthenium
precursor
RuCl2dcbpy2 14 was prepared in the same manner as described for
precursor 13 and was obtained as black crystals (0.530 g,
6H). 31P NMR (400 MHz; MeOD-d4):
d
in ppm ¼ 21.07 (s). HRMS
(ESI) m/z: [M þ Li]þ: calcd. for (C70H72LiN6O12P4Ruþ): 1205.34.;
found 1205.8228. Anal. Calcd for RuC52H70N6O12P4: H, 6.06; C,
52.13; N, 7.01. Found: H, 6.69; C, 51.23; N, 6.05. IR (max, solid):
0.803 mmol, 79%). 1H NMR (700 MHz; MeOD-d4):
d
in ppm ¼
d 9.61
(d, J ¼ 5.3 Hz, 1H), 9.45 (d, J ¼ 5.5 Hz, 1H), 9.08e9.06 (m, 4h), 7.89 (s,
1H), 7.88 (s, 1H), 7.87 (d, J ¼ 1.6 Hz, 1H), 7.86 (d, J ¼ 1.7 Hz, 0H), 7.83
2921 n(CH3), 2850 n(HeOeP]O).

28
K. Neuthe et al. / Dyes and Pigments 104 (2014) 24e33
2.3.3.6. [Ru(bnthpbpy)(dpbpy)(SCN)2],
6. Dichloro-p-cymene
iodide (c ¼ 1 mol Lꢁ1) and iodine (c ¼ 0.1 mol Lꢁ1) in a 1:4 mixture of
ruthenium dimer (0.200 g, 0.320 mmol, 1.00 equiv.) was dis-
solved in degassed, dry DMF (122 mL). 4,40-Bis(5-
hexylthiophene-2-yl)-2,20-bipyridin (0.310 g, 0.640 mmol,
2.00 equiv.) was added and the mixture was first stirred at 80 ꢂC
for 4 h and then at room temperature for 12 h. Afterwards, dpbpy
7 (0.200 g, 0.640 mmol) was added and the reaction mixture
stirred at 150 ꢂC for 4,5 h. After cooling to room temperature an
excess of ammonium thiocyanate (2.02 g, 26.5 mol) was added,
the mixture heated to 150 ꢂC and stirred for 4 h and then at room
temperature for 12 h. The solvent was removed in vacuo, the
crude product was precipitated with water, and subsequently
purified by size exclusion chromatography on a Sephadex-LH 20
column. The desired complex was obtained as dark-red crystals
acetonitrile and ethylene carbonate.
3. Results
3.1. Synthesis
The characteristics of a dye affect the performance of a solar cell.
A good dye should feature: broad absorption, strong adsorption via
a suitable anchor group, reliable regenerability, and efficient elec-
tron injection [18] [1]. In the case of ruthenium complexes these
features depend upon the ligands attached to the metal center. For
example, current efforts to improve dye effectiveness focus on
increasing the absorption range and the extinction coefficient, by
incorporating complex substituents within the donor ligand
[19,20,21].
The structures of the ruthenium complexes we synthesized are
shown in Fig. 1. First of all, the prepared dyes differ in the
complexity of the donor ligands. While the bipyridine ligand of dye
1 carries no additional light harvesting moieties, dyes 2 and 3 are
substituted triphenylamine linked to the bipyridine ligand via a
(180.0 mg, 0.176 mmol, 27%). 1H NMR (700 MHz, CD3OD):
d in
ppm 8.96e8.92 (m, 2H), 8.83 (s, 1H), 8.13 (bs, 1H), 7.82 (d, J ¼ 3.7,
1H), 7.74e7.73 (m, 1H), 7.70e7.62 (m, 4H), 7.54 (d, J ¼ 6.3 Hz, 1H),
7.46 (dd, J ¼ 6.0, 2.1 Hz, 1H), 7.37e7.335 (m, 1H). 7.17 (d,
J ¼ 5.8 Hz, 1H), 6.91 (d, J ¼ 3.9, 1H), 2.86 (t, J ¼ 7.6 Hz, 4H), 1.72 (p,
J ¼ 7.6 Hz, 4H), 1.42e1.31 (m, 12H), 0.90 (t, J ¼ 7.0 Hz, 6H). 31P
NMR (400 MHz; MeOD-d4):
d
in ppm ¼ 6.88 (s). HRMS (ESI) m/z:
[M ꢁ 2H]ꢁ2: calcd. for (C42H44N6O6S4P2Ruꢁ2): 510.0367; found
510.0364. Anal. Calcd for RuC42H46N6O6P2S4: H, 4.54; C, 49.35; N,
8.22. Found: H, 5.01; C, 36.26; N, 5.64. IR (max, solid): 2921
double bond. This electron-rich, extended p-system is intended to
improve the absorptivity of the dye [22]. Complexes 4 and 5 are
equipped with a hydrophobic alkyl chain that should introduce
stability against water. In addition to this dye 6 has an electron rich
thiophene group that is known to enhance the molar extinction
coefficient and thus the overall efficiency [23]. In order to meet the
requirements of the ZnO, surface all dyes except 1 are equipped
with phosphonic acid anchors. Better adsorption behavior and
higher electron injection rates were expected for the phosphonate
anchors. Dye 3 and the benchmark dye C101 were used for com-
parison. They closely resemble dyes 2 and 6, respectively, differing
from them in the carboxylic anchor group.
n
(CH3), 2850
n(HeOeP]O), 2096 n(SCN), 1273 n(NSC).
2.4. Deposition of ZnO films
Electrodeposition of DSSC ZnO films was performed on a
rotating disc electrode (Metrohm Autolab RDE-2) with a fixed
deposition area of 1.54 cm2 with a zinc wire as counter electrode
(Alfa Aesar, d ¼ 1 mm). The films were deposited at ꢁ0.91 V vs. Ag/
AgCl. Compact ZnO films were deposited for 10 min, ZnO/EY hybrid
films on top of the compact ZnO films were deposited for 45 min.
Afterwards EY was desorbed by immersing the films for 24 h in an
aqueous solution of KOH (c ¼ 0.316 mM, V ¼ 500 mL, pH ¼ 10.5,
Applichem, ꢀ85%). The thickness of the ZnO films amounted to 8e
The 4-(N,N0-diphenylaminio)bipyridine ligand of complexes 2
and 3 was synthesized in two steps by addition of 4-(N,N0-diphe-
nylaminio)benzaldehyde to 4,40-dimethyl-2,20-bipyridin [24] and
subsequent elimination of the alcohol. Even though more than one
equivalent of base was used, the expected side reaction, i.e.
deprotonation of both alkyl groups of the bipyridine, did not take
place and only the mono-substituted product was obtained. In a
similar manner, the 4,40-dinonyl-2,20-bipyridine ligand 10 of com-
plexes 4 and 5 was formed by addition of lithiated 4,40-dimethyl-
2,20-bipyridine to bromooctane [13].
10 mm.
ZnO films for absorption measurements were deposited on
substrates with a deposition area of 38.9 cm2. Convection of the
electrolyte within the setup was facilitated with a paddle. The films
were deposited at ꢁ1.0 V vs. the rest potential vs. Ag/AgCl. The
compact and the hybrid film were deposited for 20 min and
120 min, respectively, giving a film thickness of about 5
m
m. The EY
Initially, the modular approach shown in Fig. 2 was planned for
complex formation of those ruthenium dyes, where two of the
three ligands were identical. It involves a precursor molecule
RuCl2rbpy2 formed from ruthenium(III)chloride 2 to which the
desired ligands were to be attached in a one step reaction. In this
manner an easy ligand screening should be facilitated. Dyes 1 and 3
were prepared following this route. It was however, not possible to
synthesize phosphonate-substituted bipyridine precursor mole-
cules RuCl2rby2 (with R ¼ POOH) or to add a phosphonate anchor
ligand to a RuCl2rbpy2 precursor, where r were bulky donor moi-
eties. We can only speculate about electronical or sterical reasons.
As shown in Fig. 3, the desired phosphonate complexes 2, 4, 5,
and 6 had to be synthesized following a different approach in two
steps, each one starting from dichloro(p-cymene)ruthenium(II)
dimer. In the first step, the dimer was cleaved by one bipyridin
donor ligand. In the second step, the remaining cymene and chlo-
ride moieties of the intermediate RuCymClL were exchanged by
the desired anchor ligands (or one anchor ligand and two thiocy-
anate ligands, respectively).
was desorbed for 24 h in an aqueous solution of KOH (V ¼ 3 L,
pH ¼ 11, Applichem, ꢀ85%), while the solution was stirred at a
rotation speed of 150 rpm.
2.5. Fabrication of ZnO-based DSSCs
The prepared films were dried for 2 h at 120 ꢂC, the ZnO films
were sensitized for 24 h in a solution of the respective dye
(c ¼ 0.5 mmol Lꢁ1) in methanol. The sensitized ZnO films were
subsequently rinsed with acetonitrile and dried at 80 ꢂC.
The counter electrodes were prepared by platinum sputtering
on FTO-coated glass for 120 s. The sputter coater (Cressington
Scientific Instruments Inc. sputter coater 108auto) was operated at
a current of 30 mA and an argon pressure of 0.1 mbar.
Pressure-sensitive adhesives (PSAs, tesaÒ 61562) with a circular
cutout, defining the active area to 1.33 cm2, were usedasspacers and, at
the same time, as sealants between the two electrodes. The assembled
cells were filled with electrolyte using the backfilling method through
a hole in the counter electrode. A special syringe (Solaronix Vac’n’Fill
Syringe) was used for this procedure. Afterwards the hole was sealed
with PSA. The electrolyte was a solution of tetrapropylammonium
All desired ligands and complexes (see Fig. 1) were obtained by
modification and optimization of standard procedures in good
yields.

K. Neuthe et al. / Dyes and Pigments 104 (2014) 24e33
29
Anchor
2 eq. BiPy(Anchor)2
Anchor
7 mol% LiCl
DMF
RuCl3*3H2O
N
N
150-160 °C, 8 h
Anchor
N
Ru
N
Anchor
Cl
Cl
RuCl2bpy2
Anchor
Anchor
Anchor
Anchor
1 eq. BiPy(Donor)2
EtOH/H2O
N
N
N
N
Anchor
Ru+2
reflux, 12 h
N
N
Anchor
N
Ru
N
Anchor
Anchor
N
N
Cl
Cl
RuCl2bpy2
Donor
Donor
2X-
Fig. 2. Modular pathway to complexes model and 1.
+
Cl-
2 eq. donorligand
DCM
rt, 12 h
Ru
Cl
Cl
Ru
Ru
Cl
Cl
Cl
N
N
RuCymCl
Donor
Donor
Anchor
2X-
2 eq. bpy(Anchor)2
or
1 eq. bpy(Anchor)2, 1eq. NH4SCN
Anchor
bpyAnchor/SCN
bpyAnchor/SCN
N
Ru
Cl
Ru+2
N
N
H2O, 120 °C, 24 h
N
N
N
Donor
Donor
Donor
RuCymCl
Donor
Fig. 3. Synthetic pathway.
3.2. Photophysical properties
However, we do not have an explanation for the even lower coef-
ficient of dye 2. Excitation of the dyes at the MLCT wavelength leads
to an emission centered around 630 nm in all cases (graphs are
provided in ESI). Overall, the dyes show the optical features pre-
requisite for successful application in dye sensitized solar cells.
The absorption data of all dyes are summarized in Table 1.
Spectra are shown in Fig. 4. Spectra were recorded in methanol. All
absorption spectra show the typical metal-to-ligand-charge-
transfer (MLCT) transition between 470 and 550 nm and a
pep*
transition between 290 and 210 nm. The molar extinction co-
efficients of the complexes investigated vary between 12.0 and
15.5 L molꢁ1 cmꢁ1 and are in the range of the values published for
the standard dyes N719 (14.9) and C101 (15.7) except for dyes 2 and
4, which show lower values (8.3 and 6.5, respectively). The low
extinction coefficient of 4 might be related to the fact that the dye
3.3. Electrochemical behavior
The electrochemical behavior of dyes is of interest because it
determines their ability to inject electrons into the conductive band
of the semiconductor used and their regenerability after electron
injection.
In order to efficiently inject electrons into the conductive band
of the semiconductor (here ZnO) the energy level of the lowest
unoccupied molecular orbital (LUMO) of the dye has to be higher
than the conduction band edge (Ec) of the semiconductor. In this
work the LUMO of the dye was determined using Equation (1)
carries two chloride ligands instead of
a bipyridine ligand.
Table 1
Photophysical and electrochemical properties.
Dye
3
[103L molꢁ1 cmꢁ1
lmax/[nm])
]
lem [nm] Eox [V vs. E0e0 [eV] ELUMO [V vs.
(
NHE]
NHE]
E(LUMO) ¼ Eox (vs. NHE) ꢁ E(0e0)
(1)
1
2
3
4
5
6
15.5 (450)
6.50 (461)
12.0 (470)
8.30 (465)
15.4 (462)
15.2 (474)
629
617
634
623
612
1.29
0.75
1.20
1.19
1.36
1.00
2.23
2.36
2.35
2.18
2.31
2.15
ꢁ0.94
ꢁ1.61
ꢁ1.08
ꢁ0.99
ꢁ0.95
ꢁ1.15
were EOx is the oxidation potential of the ruthenium center deter-
mined with cyclovoltammetric measurements, and E(0e0) is the
excitation transition energy determined from the interception of
normalized absorption and emission spectra. Even though both, the
LUMO values of a particular dye and the exact level of the
625
800 [14]
C101 15.7 (526) [14]
0.90 [14] 1.60 [14] ꢁ0.70 [14]

30
K. Neuthe et al. / Dyes and Pigments 104 (2014) 24e33
3.4. Preparation of ZnO cells (LUH)
ZnO films can be easily synthesized at low temperatures in
several morphologies and with a very high crystallinity. In general,
it is advantageous when films for application in DSSCs are
composed of a compact layer and a porous layer. In the assembled
DSSC the compact layer is supposed to prevent recombination
between electrons in the conductive substrate layer and triiodide
ions in the electrolyte. The nanoporous layer provides a large sur-
face area that is sensitized with the particular dye. Here, we have
fabricated both compact and porous ZnO by electrodeposition. The
temperature applied during film deposition is 70 ꢂC, showing that
the method is feasible for application on plastic foil substrates.
The depositions of the ZnO photoelectrodes were based on the
commonly used procedure as, for example, comprehensively
described by Yoshida et al. [3] The fabrication process is illustrated
in Fig. 5.
First, the blocking compact and highly crystalline ZnO layer was
deposited without additives on FTO-coated glass. The porous films
were deposited using eosin Y (EY) as structure directing agent
(SDA). Utilization of SDAs for electrodeposition of nanoporous ZnO
films is beneficial because it improves cell performance particularly
in term of fast electron transport compared to layers composed of
individual nanoparticles. EY is the SDA of choice because it not only
leads to formation of porous crystalline ZnO films but also accel-
erates film growth and can be completely desorbed afterwards
with KOH without dissolving the ZnO layer [27].
After desorption of EY, the films were sensitized with the
particular dyes. The adsorption strength was quantified with
transmission measurements of the sensitized films. Binding of the
dyes was indicated by a shift of the absorption maxima, determined
in solution, to higher wavelengths. The spectra of dye 4 are given as
example in Fig. 6(a). The relative intensity of the first maximum of
the transmission spectra recorded on sensitized ZnO films was used
to compare the adsorption behavior of the dyes. All adsorbed very
well, but as shown in Fig. 6(b) to (a) different extent. Clearly, the
adsorption behavior of phosphonic acid carrying dyes was of most
interest. Due to their higher ligand stability, we were interested in
Fig. 4. UV/Vis spectra of newly synthesized complexes in Methanol.
conducting band of the semiconductor may vary upon adsorption
(e.g., due to different levels of protonation) and can thus differ from
the values estimated from measurements in solution, the approach
can be used as a relative good model.
After successful electron injection sufficient driving force for
effective dye regeneration is provided when the energy level of the
highest occupied molecular orbital (HOMO) is lower than the redox
potential of the redox mediator used, i.e., the oxidation of the metal
center has to be at a more positive potential than the redox po-
tential of the redox mediator. The oxidation potential of the
ruthenium metal centers of all dyes was found to be at least 0.35 V
more positive than the redox potential of I-/I3- couple (E1/2(Iꢁ/
Iꢁ3 ) ¼ þ0.4 V vs. NHE) as shown in Table 1; thus, providing enough
driving force for dye regeneration.
The estimated levels of the dye’s LUMOS were above 2 eV for all
of them (Table 1). Because it is commonly assumed [25,26] that the
bandgap and the conduction band edge of ZnO are similar to that of
TiO2 anatase (lower conduction band edge ECB ¼ ꢁ0.50 V vs. NHE,
upper conduction band edge E ¼ ꢁ0.84 V vs. NHE) the LUMO levels
of the dyes are appropriate for electron injection into the ZnO
conducting band.
receiving preliminary insights into
a
structure-adsorption-
relationship. We expected them to attach better to the ZnO sur-
face than their carboxylic counterparts. The strongest binding was
observed for dye 4, a complex equipped with two phosphonate
anchors. Despite the fact that the strongest adsorption was
O2
O2
SDA
e-
e-
Zn2+
Zn2+
KOH(aq)
pH = 10,5
Solution of
Sensitizer
Adsorption
of
Sensitizer
Removal of SDA
by treatment with
alkaline solution
Fig. 5. Fabrication scheme for nanoporous ZnO-photoelectrodes.

K. Neuthe et al. / Dyes and Pigments 104 (2014) 24e33
31
a
b
0.25
0.2
0.15
0.1
0.05
0
20
18
16
14
12
10
8
ratio extinction coefficent/absorbance on ZnO
extinction coefficient
6
4
2
0
C101
1
2
3
4
5
6
Dye
Fig. 6. Adsorption behavior of the dyes on ZnO: (a) comparison of absorption of 4 in
solution and adsorbed on ZnO (b) relative intensity of first transmission maximum
recorded on sensitized ZnO films.
Fig. 7. (a) IPCE and (b) J/V curves of dyes investigated.
3 reached stable IPCE values of 4% from up to 600 nm. Dyes 4, 5, and
6 display IPCE values around 2.5% (1.8% for 6) with maximum ef-
ficiencies at 450 and 550 nm. The IPCE spectra of all dyes resemble
the absorption spectra of the dyes, when adsorbed on ZnO, i.e. local
maxima in the absorption spectra are reflected by local maxima in
the IPCE spectra.
ZnO cells of the same type were sensitized with the high per-
formance dye C101 for comparison. The cells reached IPCE values of
50% maximum. As already noted for 4, 5, and 6, the IPCE curve is not
constant over a broad wavelength range but exhibits maximum
values at 450 nm and 550 nm. The slow decay of IPCE to 750 nm
was remarkable.
observed for a phosphonate anchored complex, there is no clear
tendency between the two anchor types. While dye 2 (phospho-
nate) adsorbs stronger than dye 3 (carboxylate), it is, however, the
opposite for dyes 6 and its carboxylic anchored equivalent C101.
Furthermore, we were interested in seeing if the amount of phos-
phonate anchors affects the adsorption of the dyes. Here, too, no
trend was visible. The best dye with phosphonate anchors, i.e. dye 4
carries two groups. The next best, dye 2, has four groups and the
third best complex, dye 6, two anchors groups. Overall, no
structure-adsorption relationship could be revealed. All dyes
adhere tightly to the ZnO surface and can thus be applied in DSSCs.
As suspected from the IPCE data, the values for open circuit
potential, short circuit photocurrency, fill factor, and efficiency
3.5. Photovoltaic performance on ZnO
The prepared cells were illuminated under AM 1.5D conditions,
and photocurrent spectra and IPCE data were measured. The graphs
are shown in Fig. 7 The data of interest, i.e., IPCE values, short circuit
photocurrent (JSC), open circuit voltage (VOC), fill factor (FF), and
Table 2
Comparison of cell performance.
Dye
JSC [mA cmꢁ2
]
VOC [mV]
FF [%]
h [%]
1
2
3
4
5
6
0.63
0.26
0.1
0.78
0.70
0.41
7.81
443
209
454
374
333
392
538
55.2
25.7
39.7
57.2
48.7
47.6
31.1
0.15
0.01
0.02
0.17
0.11
0.08
1.31
efficiency (h) are presented in Table 2.
The complexes exhibit IPCE values below 10%. The curves are
shown in Fig. 7. Surprisingly, the simple dye 1, which did not carry
special donor moieties, gave the best results with IPCE values of
around 10% between 450 and 500 nm, thereafter decreasing and
plateauing at 2% until 600 nm. The carboxylate-anchored complex
C101

32
K. Neuthe et al. / Dyes and Pigments 104 (2014) 24e33
η(%)
η(%)
0.01
0.01
0.02
0.08
0.11
0.15
0.17
1.31
3
3
0.02
0.08
0.11
0.15
0.17
1.31
2
2
6
6
5
5
1
1
4
4
C101
C101
400 420 440 460
480 500 520 540
0
0.5
1
1.5
2
2.5
3
λmax (nm)
Absorbance on ZnO (AU)
η(%)
3
0.01
0.02
0.08
0.11
0.15
0.17
1.31
2
6
5
1
4
C101
Distance LUMO Level
ZnO-conduction Band edge (eV)
-0.2
0
0.2
0.4
0.6
0.8
1
1.2
Fig. 8. Factors influencing efficiency.
were lower than expected. The best results were recorded with the
reference dye C101, which gave
¼ 1.31; JSC ¼ 7.81, VOC ¼ 538, and
energetic overlap of LUMO and UBE, the more likely it is that an
electron is not injected into the conducting band but instead falls
back to the ground state or recombines with the electrolyte.
It should be mentioned that dye 4 is an exception from the rule.
Here the distance LUMO-UBE is larger than for 1 but without an
effect upon efficiency. However, as shown in the “absorption vs.
efficiency” plots in Fig. 8, more dye is bound to the ZnO surface than
1, leading to a bigger surface coverage and a positive influence on
the cell performance.
In summary, we propose that there is a strong correlation of the
overall cell efficiency and the electrochemically determined energy
separation between LUMO of the dye and UBE of ZnO. That pro-
posal adds a new aspect to the energy requirements a “good”
sensitizer has to meet. It suggests an upper energy level of the dye’s
LUMO, whereas to our best knowledge so far only a lower level was
defined as limiting factor for efficient electron injection.
h
FF ¼ 31.1. The two best dyes from the study were 1 and 4 with ef-
ficiencies of
h
¼ 0.17 and 0.15%, Jsc ¼ 0.78 and 0.63 mA cmꢁ2, and
Voc ¼ 374 and 443 mV. The corresponding data for all dyes vary over
a broad range e
h
¼ 0.17e0.01, JSC ¼ 0.8e0,2, and VOC ¼ 400e209.
The fill factors, however, are nearly constant with values around
FF ¼ 50, except for dye 2 with FF ¼ 25.7. Interestingly, the dyes with
carboxylic acid anchoring groups show better cell results than their
counterparts with phosphonic acid anchoring groups.
Keeping in mind that the factors influencing cell performance
like extinction coefficient and absorption range were comparable
with standard dyes, the cell performance of the dyes was weaker
than expected. Curiously, in this study, the dyes that performed
best were the dyes with the simplest structures as depicted in Fig.1,
which were neither superior in term of extinction coefficient nor in
the absorption range. Dye 4, for example, only has one donor ligand
while the bipyridine donor ligands of the dye 1 do not carry addi-
tional light harvesting moieties at all. So, we wondered if there is
any relation between the factors mentioned above and the per-
formance of the cells sensitized with our dyes.
4. Conclusion
New ruthenium complexes with different anchor groups have
been synthesized and applied in ZnO based solar cells.
In Fig. 8 the absorbance on ZnO, absorption maximum, and
LUMO level of a particular dye are plotted against the efficiency
yielded. We chose to display the absorbance measured on ZnO,
because it reflects the combination of the effects of extinction co-
efficient and adsorption strength of the particular dye. The graphs
show, that after exceeding a certain value, the impact of the optical
features of the dyes on the efficiency decreases. Note that the LUMO
levels of the dyes are shown in relation to the upper edge of the
conduction band of ZnO, i.e., the distance between the LUMO and
the upper band edge (UBE) vs. yielded efficiency. Positive values
mean that the LUMO is above the UBE of ZnO (“outside” the con-
ducting band). Negative values imply that the LUMO of the dye is
below the UBE of ZnO (“inside” the conducting band). This shows
that the relation between the distance LUMO-UBE and efficiency
strongly correlates. Only one dye features a LUMO that lies within
the conducting band - C101, which is the best dye in this study. The
efficiency of the other dyes decreases with increasing distance
between dye LUMO and the UBE.
The best results were achieved with the standard dye C101 and
the new dye 4 (
h
¼ 1.31 and 0.17%).
Looking at the dye structures two points stand out: in terms of
efficiency carboxylic anchor groups are superior to phosphonic
anchor groups and the most simple dyes performed best.
Using a simple model, we have determined the LUMO level of
the dyes. We propose that the energy distance between LUMO of
the dye and the upper band edge of ZnO directly affects the cell
performance. The larger the distance of the LUMO to the upper
band edge the lower the efficiency of the cell. This model could
serve as a quick and easy pre-test to evaluate the applicability of a
potential dye without carrying out extensive and time-consuming
cell tests. More dyes are currently under investigation in order to
verify our proposal.
Acknowledgment
The German Federal Ministry of Education and Research (BMBF)
supported this work. Fundings were received within the frame of
the program “organic photovoltaics”. Members of the joint research
Supposedly this effect can be attributed to recombination pro-
cesses diminishing the injection efficiency. The smaller the

K. Neuthe et al. / Dyes and Pigments 104 (2014) 24e33
33
photophysical behavior and quenching with divalent metal ions. Inorg Chem
2000;39:76e84.
project “scalable dyes sensitized solar cells (FSZ Industrie)” were
BGT Bischoff Glastechnik AG, Fraunhofer IAO, Fraunhofer Institute
for Solar Energy Systems (ISE), Freie Universität Berlin, IOLITEC
Ionic liquids Technologies GmbH, Leibniz Universität Hannover,
Merck KGaA, MP&L-Produktions GmbH, Pröll KG, tesa AG, Zürcher
Hochschule für Angewandte Wissenschaften.
[13] Percec V, Barboiu B, Kim H. Arenesulfonyl Halides: a universal class of func-
tional initiators for metal-catalyzed “living” radical polymerization of sty-
rene(s), methacrylates, and acrylates. J Am Chem Soc 1998;120:305e16.
[14] Gao F, Wang Y, Shi D, Zhang J, Wang M, Jing X, et al. Enhance the optical
absorptivity of nanocrystalline TiO2 film with high molar extinction coeffi-
cient ruthenium sensitizers for high performance dye-sensitized solar cells.
J Am Chem Soc 2008;130:10720e8.
[15] Liska P, Vlachopoulos N, Nazeeruddin MK, Grätzel M. cis-Diaquabis (2, 200-
bipyridyl-4, 400-dicarboxylate) ruthenium(II)sensitizes wide band gap oxide
semiconductors very efficiently over a broad spectral range in the visible. J Am
Chem Soc 1988;110:3686e7.
[16] Sullivan BP, Salmon DJ, Meyer TJ. Mixed phosphine 2,2’-bipyridine complexes
of ruthenium. Inorg Chem 1978;17:3334e41.
Appendix A. Supplementary data
Supplementary data related to this article can be found at http://
dx.doi.org/10.1016/j.dyepig.2013.12.018
[17] Caramori S, Cristino V, Argazzi R, Meda L, Bignozzi CA. Photoelectrochemical
behavior of sensitized TiO2 photoanodes in an aqueous environment: appli-
cation to Hydrogen production. Inorg Chem 2010;49:3320e8.
[18] Vougioukalakis GC, Philippopoulos AI, Stergiopoulos T, Falaras P. Contribu-
tions to the development of ruthenium-based sensitizers for dye-sensitized
solar cells. Coord Chem Rev 2010;255:2602e21.
[19] Siu C, Ho C, He J, Chen T, Cui X, Zhao J, et al. Thiocyanate-free ruthenium(II)
cyclometalated complexes containing uncommon thiazole and benzothiazole
chromophores for dye-sensitized solar cells. J Org Chem 2013;748:75e83.
[20] Erten-Ela S, Ocakoglu K. Fabrication of thin film nanocrystalline TiO2 solar
cells using ruthenium complexes with carboxyl and sulfonyl groups. J Ind Eng
Chem [in press], http://dx.doi.org/10.1016/j.jiec.2013.05.004; 2013.
[21] Suresh T, Rajkumar G, Singh SP, Reddy PY, Islam A, Han L, et al. Novel
ruthenium sensitizer with multiple butadiene equivalent thienyls as conju-
gation on ancillary ligand for dye-sensitized solar cells. Org Electron 2013;14:
2243e8.
[22] Wang P, Zakeeruddin SM, Moser JE, Humphry-Baker R, Comte P, Aranyos V,
et al. Stable new sensitizer with improved light harvesting for nanocrystalline
dye-sensitized solar cells. Adv Mater 2004;16:1806e11.
[23] Jiang K, Masaki N, Xia J, Noda S, Yanagida S. A novel ruthenium sensitizer with
a hydrophobic 2-thiophen-2-yl-vinyl-conjugated bipyridyl ligand for effective
dye sensitized TiO2 solar cells. Chem Commun 2006;23:2460e2.
[24] Hewlins M, Salter R. The photochemical cyclodehydrogenation route to
polycyclic azaarenes. Synthesis 2007;14:2164e74.
References
[1] Hagfeldt A, Boschloo G, Sun L, Kloo L, Petterson H. Dye-sensitized solar cells.
Chem Rev 2010;110:6595e663.
[2] Zhang Q, Dandeneau CS, Zhou X, Cao G. ZnO nanostructures for dye-sensitized
solar cells. Adv Mater 2013;21:4087e108.
[3] Yoshida T, Zhang J, Komatsu D, Sawatani S, Minoura H, Pauporté T, et al.
Electrodeposition of inorganic/organic hybrid thin films. Adv Funct Mater
2009;19:17e43.
[4] O’Regan B, Grätzel M. A low-cost, high-efficiency solar cell based on dye-
sensitized colloidal TiO2 films. Nature 1991;353:737e9.
[5] Yin X, Liu X, Wang L, Bin Liu. Electrophoretic deposition of ZnO photoanode
for plastic dye-sensitized solar cells. Electrochem Commun 2010;12:1241e4.
[6] Du J, Bittner F, Hecht DS, Ladous C, Ellinger J, Oekermann T, et al. A carbon
nanotube-based transparent conductive substrate for flexible ZnO dye-
sensitized solar cells. Thin Solid Films 2013;531:391e7.
[7] Keis K, Lindgren J, Lindquist S, Hagfeldt A. Studies of the adsorption process of
Ru complexes in nanoporous ZnO electrodes. Langmuir 2000;16:4688e94.
[8] Reisner E, Powell DJ, Cavazza C, Fontecilla-Camps JC, Armstrong Fraser A.
Visible light-driven H2 production by hydrogenases attached to dye-sensitized
TiO2 nanoparticles. J Am Chem Soc 2009;131:18457e184666.
[9] Jakubikova E, Snoeberger III RC, Batista VS, Martin RL, Batista ER. Interfacial
electron transfer in TiO2 surfaces sensitized with Ru(II)-polypyridine com-
plexes. J Phys Chem A 2009;113:12532e40.
[25] Thavasi V, Renugopalakrishnan V, Jose R, Ramakrishna S. Controlled electron
injection and transport at materials interfaces in dye sensitized solar cells.
Mater Sci Eng R Rep 2009;63:81e99.
[26] Bisquert J. Dilemmas of dye-sensitized solar cells. ChemPhysChem 2011;12:
1633e6.
[10] Mc Kenna CE, Higa MT, Cheung NH, McKenna M. The facile dealkylation of
phosphonic acid dialkyl esters by bromotrimethylsilane. Tetrahedron Lett
1977;2:155e8.
[11] Penicaud V, Odobel F, Bujoli B. A facile and efficient syntheses of 2,20-bipyr-
idine-based bis(phosphonic) acids. Tetrahedron Lett 1998;39:3689e92.
[12] Montalti M, Wadhwa S, Kim WY, Kipp RA, Schmehl RH. Luminescent ruth-
[27] Yoshida T, Minoura H. Electrodeposition of ZnO/dye hybrid thin films for dye-
sensitized solar cells. Adv Funct Mater 2009;19:17e43.
enium(II)
bipyridylꢁphosphonic
acid
complexes:
pH
dependent