A new family of substituted triethoxysilyl iodides as organic iodide sources for dye-sensitised solar cells

Article abstract of DOI:10.1039/b925315g

This paper reports the synthesis and characterisation of a family of heterocyclic triethoxysilyl iodides of general formula, (EtO) ₃SiR²-N⁺R¹, where R¹ = -C₅H₁₁, -CH₂C₆H₅ or -C₁₆H₃₃; N⁺ = pyrrolidinium or piperidinium; and R² = -(CH₂)₃- and -(CH₂) ₁₁- and their applications as organic iodide sources for use in dye sensitised solar cells. The effect of three variations in the structure of these derivatives (viz., R¹, R² and N⁺) on DSSC performance has been investigated using a range of techniques, including I-V profiling at different light intensities and transient photovoltage and photocurrent measurements. The results of these analyses have been used to develop an understanding of how different organic cations interact with the DSSC on a molecular level and influence the overall performance of the device. Of the three parameters, the length of the spacer - R² - was found to have the greatest influence of device performance. Use of the organic iodides with the longer spacer - where R² = -(CH₂)₁₁- were superior in performance to those with a shorter spacer - where R ² = (CH₂)₃- with devices constructed using the silyl iodide with longer spacer showing larger open circuit voltages and longer recombination times in comparison to those prepared using the shorter one.

Full text of DOI:10.1039/b925315g

View Article Online / Journal Homepage / Table of Contents for this issue
PAPER
www.rsc.org/materials | Journal of Materials Chemistry
A new family of substituted triethoxysilyl iodides as organic iodide sources
for dye-sensitised solar cells
Naomi A. Lewcenko,a Matthew J. Byrnes,a Torben Daeneke,a Mingkui Wang,b Shaik M. Zakeeruddin,b
b
Michael Gratzel and Leone Spiccia
a
*
*
€
Received 2nd December 2009, Accepted 5th February 2010
First published as an Advance Article on the web 8th March 2010
DOI: 10.1039/b925315g
This paper reports the synthesis and characterisation of a family of heterocyclic triethoxysilyl iodides of
general formula, (EtO)3SiR2-N+R1, where R1 ¼ –C5H11, –CH2C6H5 or –C16H33; N+ ¼ pyrrolidinium
or piperidinium; and R2 ¼ –(CH2)3– and –(CH2)11– and their applications as organic iodide sources for
use in dye sensitised solar cells. The effect of three variations in the structure of these derivatives
(viz., R1, R2 and N+) on DSSC performance has been investigated using a range of techniques, including
I–V profiling at different light intensities and transient photovoltage and photocurrent measurements.
The results of these analyses have been used to develop an understanding of how different organic
cations interact with the DSSC on a molecular level and influence the overall performance of the device.
Of the three parameters, the length of the spacer – R2 – was found to have the greatest influence of
device performance. Use of the organic iodides with the longer spacer – where R2 ¼ –(CH2)11– were
superior in performance to those with a shorter spacer – where R2 ¼ (CH2)3– with devices constructed
using the silyl iodide with longer spacer showing larger open circuit voltages and longer recombination
times in comparison to those prepared using the shorter one.
increased short circuit current.40,41 Cations in the electrolyte have
Introduction
also been shown to influence the kinetics of cell degradation as
well as play an important role in the electron lifetimes of the
device.42–44 Moreover, the cation can also interact with the dye,
giving rise to a bathochromic shift in the absorption spectra of
the bound complex.45 These studies have stimulated interest in
the development of new types of cations for incorporation into
DSSC electrolytes.
The dye-sensitised solar cell (DSSC)1 is an intensively investi-
gated renewable energy technology. The design of the DSSC
utilises a high band-gap semiconductor material, coated with
a monolayer of a highly absorbing dye capable of injecting
electrons into the semiconductor upon irradiation with light.
These injected electrons then percolate through the semi-
conductor to the conductive glass and flow to the counter
electrode whereby reducing the redox mediator present in the
electrolyte. This reduced redox mediator in turn regenerates
the oxidised dye completing the circuit as demonstrated sche-
matically in Fig. 1.
The effect of each component of DSSC performance, viz.,
different semiconductors,2,3 dyes,4–17 redox mediators18–20 and
electrolytes21–39 continues to be widely studied. The electrolyte is
very important in determining the performance of the DSSC due
to the intimate contact it has with every other cell component.
For example, it has been established that lithium ions present in
the electrolyte effect cell performance by interacting with the
titania surface. Li+ ions have been shown to have a two-fold
effect: firstly, Li+-TiO2 interactions lead to a positively shifted
conduction band potential (CBP) decreasing the energy differ-
ence between the conduction band and the redox mediator,
giving rise to smaller open circuit voltages. As a result of this
positive shift, the driving force for injection (i.e., the energy
difference between the excited state of the dye and the CBP of the
titania) is increased, leading to enhanced injection and an
Fig. 1 Schematic diagram of the electron transfer processes occurring in
a DSSC. (A) Photon-induced dye excitation. (B) Injection of the
photoexcited electron into the conduction band of the titania. (C)
Movement of the electron through an external circuit. (D) Reaction of
the electron with the oxidised redox couple at a platinised counter elec-
trode. (E) Regeneration of the dye by the reduced redox couple.
aSchool of Chemistry, Monash University, Victoria, 3800 Australia.
E-mail: leone.spiccia@sci.monash.edu.au
bLaboratory for Photonics and Interfaces, EPFL, CH1015, Lausanne,
Switzerland. E-mail: michael.graetzel@epfl.ch
3694 | J. Mater. Chem., 2010, 20, 3694–3702
This journal is ª The Royal Society of Chemistry 2010

View Article Online
evaluated as organic iodide sources for DSSCs. Three structural
variations in the organic iodide were examined: (i) the organic
substituent on the quaternary nitrogen (pendant: R1), (ii) the
spacer between the siloxane and nitrogen (spacer: R2) and (iii) the
cation identity – pyrrolidinium or piperidinium (cation: N+).
These variations were examined to provide insight into how the
cations are able to interact with the titania surface and what
effects these interactions have on the overall DSSC device
performance. DSSC efficiency was determined for a number of
electrolyte compositions. Photovoltage and photocurrent decay
measurements were conducted on the optimised electrolyte
composition to elucidate the origin of the variations in device
performance.
Fig. 2 Examples of cationic alkylammonium and alkylpyrrolidinium
triethoxysilyl iodides previously used as organic iodide sources in DSSCs.
Previous studies into the effect of organic cations on DSSC
behaviour has predominantly focused on the planar, aromatic
imidazolium cation37 and tetraalkylammonium cations such as
tetrabutylammonium,43,46 mainly due to the fact that these
cations are well characterised and have clearly defined electro-
chemical properties. In comparison, we have recently reported
Experimental
Reagents and electrolytes
a
series of trialkylammoniumpropyltriethoxysilyl iodides
{(EtO)3Si(CH2)3N+R3}Iꢀ, including 1 (R ¼ ethyl) and 2
(R ¼ heptyl) (Fig. 2), which were found to function as efficient
iodide sources in DSSCs.47 Cells constructed using these
compounds as the sole iodide source showed reasonably large
open circuit voltages (Voc) and short circuit currents (Jsc) with
good overall efficiencies, particularly at low sunlight levels. These
types of siloxanes were used to develop ammonium-functional-
ised silica nanoparticles whose use as hybrid organic–inorganic
iodide sources in DSSCs led to good cell performances at low
sunlight levels.48
All chemicals and solvents used over the course of this investiga-
tion were of reagent grade or better and were purchased from
Aldrich, Merck or Fluka. Toluene and THF were pre-dried over
sodium and distilled prior to use according to the method of
Perrin.50 3-Iodopropyltriethoxysilane (12) and 11-iodoundec-1-
ene (13) were prepared according to literature procedures.51,52 1
H
and 13C{1H} (125.7 MHz) NMR spectra for all compounds
prepared in this study were measured on a Varian Inova
500 spectrometer at 500 MHz and 125.7 MHz, respectively, or,
unless otherwise indicated, a Varian Mercury 300 spectrometer
and referenced against the residual solvent peak. Electrospray
mass spectra of the N-alkylpyrrolidinium and N-alkylpiper-
idinium alkyltriethoxysilyl iodides were recorded on a Waters
Platform II Mass Spectrometer and elemental analyses were
carried out by Campbell Microanalytical Laboratory (New Zea-
land). ATR FTIR measurements were collected using a BioRad
ATR FTIR Spectrometer with a diamond sample window and
These promising results stimulated our interest in new
siloxane-based organic iodides for DSSCs. Two new N-pentyl-
pyrrolidiniumalkyltriethoxysilyl iodides, 3 and 4 shown in Fig. 2,
were prepared and used as iodide sources in DSSCs.49 The only
difference between 3 and 4 was that the cation was separated
from the siloxane moiety by either a 3-carbon or an 11-carbon
alkyl chain, respectively. DSSC testing with these compounds
indicated that better device performance was achieved when the
siloxane with the 11-carbon alkyl spacer was used as the iodide
source.
a tungsten carbide anvil, with 64 scans at a resolution of ꢁ4 cmꢀ1
.
To further correlate structural modifications with DSSC
performance, a new series of substituted triethoxysilyl iodides
(incorporating pyrrolidinium and piperidinium cations) have
been synthesised (Fig. 3) and these compounds have been
Synthesis of N-alkylpyrrolidines and piperidines
N-Pentylpyrrolidine (5). Following the methodology reported
by Hart and McGreal,53 pyrrolidine (5.00 g, 70.3 mmol, 3 eq),
pentyl bromide (7.08 g, 46.9 mmol, 2 eq) and K2CO3 (3.56 g,
25.8 mmol, 1.1 eq) were suspended in methanol (50 mL) and
heated to reflux with stirring for 16 h. After this time, water
(10 mL) was added to dissolve the inorganic salts, the methanol
was removed and the resulting crude product was extracted using
dichloromethane (3 ꢂ 50 mL). The combined organic fractions
were washed with a saturated brine solution (10 mL), dried using
sodium sulfate and the solvent removed to yield a pale yellow oil.
This oil was purified by vacuum distillation (70ꢃ C at ca.
10 mmHg) to give 3.04 g (31% yield) of 5 as a colourless oil. As
the 1H NMR spectral data reported below for 5 closely matched
that reported previously in the literature, no further character-
ization was performed.
The procedure for the preparation of 5 was also used to
prepare 6, 7, and 10 with the same number of moles of pyrroli-
1
dine or piperidine, alkyl bromide and K2CO3. The H NMR
Fig.
3 Cationic alkylpyrrolidinium and alkylpiperidinium propyl-
triethoxysilyl iodides used in this work.
spectra for these compounds also matched those in the literature.
This journal is ª The Royal Society of Chemistry 2010
J. Mater. Chem., 2010, 20, 3694–3702 | 3695

View Article Online
N-Hexadecylpyrrolidine (6). 5.62 g of 6 (27% yield)ꢃ was
obtained as a colourless oil after vacuum distillation (160 C at
10 mmHg).
dry toluene and 6 drops of ethanol. The flask was sealed and
heated to 115 ꢃC overnight. The reaction mixture was filtered
through a Celite pad and washed with dry toluene. The volatiles
were removed in vacuo by heating to approx. 70 ꢃC leaving
a light brown coloured oil. This material was vacuum distilled
(90–125 ꢃC at ca. 1 mmHg) to give 13 (10.59 g, 67% yield) as
a colourless oil.
N-Benzylpyrrolidine (7). 5.90 g of 7 (52% yield) was obtained as
a colourless oil after vacuum distillation (95 ꢃC at ꢄ10 mm Hg)
N-Pentylpiperidine (10). 3.93 g of 10 (43% yield) was obtained
as a colourless oil after vacuum distillation (72 ꢃC at 10 mm Hg)
1H NMR (300 MHz, CDCl3): d 0.61 (m, 2H, SiCH2CH2), 1.20
(t, 9H, J ¼ 6.9 Hz, OCH2CH3), 1.25 and 1.36 (br s and br m, 16H
total, CH2 of alkyl chain), 1.80 (quintet, 2H, J ¼ 6.9 Hz,
CH2CH2I), 3.18 (t, 2H, J ¼ 6.9 Hz, CH2CH2I), 3.80 (q, 6H, J ¼ 6.9
Hz, OCH2CH3). 13C{1H} NMR (75.42 MHz, CDCl3) 7.2, 10.3,
18.3, 22.7, 28.5, 29.2, 29.3, 29.4, 29.5, 30.4, 33.1, 33.5, 58.2. Anal.
CalcdforC17H37IO3Si:C, 45.94;H, 8.39. Found:C, 46.35;H, 8.58.
3-(N-Pentylpyrrolidinium)propyltriethoxysilyl iodide (3).
A
solution of 5 (0.636 g, 4.5 mmol, 1 eq) and 3-iodopropyl-
triethoxysilane (12) (2.00 g, 4.5 mmol, 1 eq) in toluene (10 mL)
was heated to 110 ꢃC in the dark for three days. After this time,
the toluene was removed in vacuo and the product dried at 80 ꢃC
at 10 mmHg for 3 h to yield 1.88 g of a pale yellow, highly viscous
liquid (4.0 mmol, 88% yield) that was used without further
purification.
1H NMR (CDCl3) d: 0.69 (2H, t, J ¼ 7.1 Hz, SiCH2), 0.931
(3H, t, J ¼ 6.5 Hz, CH3), 1.23 (9H, t, J ¼ 7.0 Hz, OCH2CH3),
1.39 (4H, m, CH2), 1.78 (4H, m, CH2), 2.32 (4H, m, CH2), 3.41
(8H, m, CH2N+), 3.82 (6H, m, OCH2CH3). 13C NMR (CDCl3)
d: 6.5, 13.4, 17.0, 17.9, 21.7, 22.9, 27.9, 58.3, 59.7, 60.7, 63.0.
Anal. Calcd for C18H40INO3Si (444.46): C, 45.66; H, 8.51; N,
2.96; I, 26.80. Found: C, 45.42; H, 8.52; N, 2.72; I, 27.13. ESI MS
m/z (M-Iꢀ) 346.3 (C18H40NO3Si+ requires 346.3).
11-(N-Pentylpyrrolidinium)undecyltriethoxysilyl iodide (4).
Compound 4 was prepared in a similar manner to 3 using 2.00 g
of 12 (4.5 mmol) and 0.635 g of 5 (4.5 mmol) to give 2.21 g of 4
(84%) as a pale orange, waxy solid.
1H NMR (CDCl3) d: 0.71 (2H, t, J ¼ 7.2 Hz, SiCH2), 0.86
(3H, t, J ¼ 6.6 Hz, CH3), 1.24 (9H, m, OCH2CH3), 1.29–1.31
(26H, m, CH2), 1.72 (8H, m, CH2), 3.41 (8H, m, N+CH2), 3.82
(6H, m, OCH2CH3). 13C NMR (CDCl3) d: 9.2, 13.5, 17.9, 18.2,
21.6, 21.9, 22.3, 25.8, 26.0, 26.8, 28.8, 28.9, 29.2, 30.3, 34.4, 57.9,
58.6 Anal. Calcd for C26H56INO3Si (585.72): C, 53.32; H, 9.64;
N, 2.39; I, 21.67. Found: C, 52.99; H, 9.78; N, 2.22; I, 21.01.
ESI MS m/z (M-Iꢀ) 458.4 (C26H56NO3Si+ requires 458.8).
Compounds 8 and 9 were obtained by the procedure used to
prepare compound 3 using equivalent molar quantities and
similar reaction conditions.
11-(N-Pentylpiperidinium)undecyltriethoxysilyl iodide (11).
Compound 11 was prepared following the outlined above for 3
using 2.00 g of 12 (4.5 mmol) and 0.698 g of 10 (4.5 mmol) to give
2.43 g of 11 (90% yield) as a pale orange, waxy solid.
1H NMR (CDCl3) d: 0.56 (2H, t, J ¼ 8.3 Hz, SiCH2), 0.87 (3H, t,
J ¼ 6.9 Hz, CH3), 1.17 (9H, t, J ¼ 7.0 Hz, OCH2CH3), 1.20–1.37
(20H, m, CH2), 1.61 (4H, m, CH2), 1.82 (2H, m, CH2), 1.89
(4H, m, CH2), 3.95 (4H, m, CH2), 3.65 (4H, m, N+CH2), 3.75
(6H, q, J ¼ 7.0 Hz, OCH2CH3). 13C NMR (CDCl3) d: 9.5, 13.0,
17.5, 19.2, 19.9, 20.7, 21.0, 21.4, 21.9, 25.6, 27.6, 28.3, 28.4, 28.5,
28.6, 57.4, 58.5. Anal. Calcd. for C27H58INO3Si (599.32): C, 54.07;
H, 9.75; N, 2.34; I, 21.16. Found: C, 53.87; H, 9.81; N, 2.15; I,
20.65. ESI MS m/z (M-Iꢀ) 472.3 (C27H58NO3Si+ requires 472.8).
3-(N-Hexadecylpyrrolidinium)propyltriethoxysilyl iodide (8).
1.64 g of 8 (58% yield) was isolated as a pale brown, highly
viscous liquid.
1H NMR (CDCl3) d: 0.71 (2H, t, J ¼ 7.2 Hz, SiCH2), 0.86
(3H, t, J ¼ 6.6 Hz, CH3), 1.24 (9H, m, OCH2CH3), 1.29–1.31
(26H, m, CH2), 1.72 (8H, m, CH2), 3.41 (8H, m, N+CH2), 3.82
(6H, m, OCH2CH3). 13C NMR (CDCl3) d: 7.1, 14.3, 15.8, 18.6,
20.1, 20.7, 21.9, 22.9, 26.6, 28.1, 29.5, 29.7, 29.9, 31.4, 32.1, 58.9,
59.2. Anal. Calcd for C29H62INO3Si (627.80): C, 55.48; H, 9.95;
N, 2.23; I, 20.21. Found: C, 55.32; H, 9.96; N, 2.17; I, 20.85. ESI
MS m/z (M-Iꢀ) 500.4 (C29H62NO3Si+ requires 500.9).
3-(N-Benzylpyrrolidinium)propyltriethoxysilyl iodide (9). 1.98 g
of 9 (89%) was isolated as a pale yellow, waxy solid.
Dye sensitised solar cell assembly
1H NMR (CDCl3) d: 0.69 (2H, m, CH2), 1.22 (9H, t, J ¼ 9.7,
OCH2CH3), 1.68 (2H, m, CH2), 1.94 (4H, m, CH2), 2.50 (2H, m,
CH2), 3.35 (4H, m, CH2), 3.81 (6H, q, J ¼ 7.7, OCH2CH3), 3.98
(2H, m, CH2Ph), 7.46 (3H, m, CH), 7.64 (2H, m, CH). 13C NMR
(CDCl3) d: 6.0, 14.7, 16.4, 18.7, 19.4, 20.4, 21.4, 21.7, 49.6, 56.8,
58.1, 63.1, 125.5, 128.4, 129.5, 132.4. Anal. Calcd for
C20H36INO3Si (493.59): C, 48.68; H, 7.35; N, 2.84; I, 25.72.
Found: C, 48.57; H, 7.39; N, 2.77; I, 26.01. ESI MS m/z (M-Iꢀ)
366.5 (C20H36NO3Si+ requires 366.6).
Working electrodes were prepared from paste produced in-house
and consisted of a 7 mm layer of 20 nm particles, topped with
a 5 mm scattering layer of 400 nm particles. Working electrodes
were sintered at 500 ꢃC for 30 min, cooled to 80 ꢃC and immersed
for 16 h in a 0.3 mM solution of the amphiphilic ruthenium dye,
NaRu(4-carboxylic
acid-40-carboxylate-2,20-bipyridine)-(4,40-
dinonyl-2,20-bipyridine)(NCS)2 coded as Z907Na.38,54 After this
time, the films were removed from the dye solution, washed with
acetonitrile and constructed into DSSCs using standard solar cell
construction procedures.55
11-Iodoundecyltriethoxysilane
(13).
11-Iodoundec-1-ene
(10.05 g, 35.7 mmol) was placed in a Schlenk flask, which was
then evacuated and back-filled with nitrogen three times and
SiH(OEt)3 (7.4 mL, 40.0 mmol) was added using a syringe. To
this mixture, ca. 8 mg of RuCl3 was added followed by 6 drops of
Electrolytes
Initial testing was conducted using electrolytes with 1.0 M
organic iodide, 0.03 M I2, 0.5 M tert-butylpyridine (TBP) in 3 : 1
3696 | J. Mater. Chem., 2010, 20, 3694–3702
This journal is ª The Royal Society of Chemistry 2010

View Article Online
Fig. 5 Depiction of cationic triethoxylsilyl iodides showing the three
components varied in DSSC testing; the spacer (R2), the cation (N+) and
the pendant group (R1).
Fig. 4 I–V profile of devices constructed using 1.0 M or 0.6 M 3 with
0.03 M I2, 0.5 M TBP in 3 : 1 AN : VN.
acetonitrile : valeronitrile (AN:VN) or 1.0 M organic iodide,
0.15 M I2, 0.5 M N-butylbenzimidazole (NBB) in 3-methox-
ypropionitrile (MPN) but the devices exhibited poor fill factors
(see Fig. 4). Consequently, the composition of the electrolytes
used in further testing were 0.6 M organic iodide, 0.03 M I2,
0.5 M TBP, 0.5 M N-ethyl-N0-methylimidazolium bis(tri-
fluoromethanesulfonyl)imide in 3 : 1 AN : VN or 0.6 M organic
iodide, 0.15 M I2, 0.5 M N-butylbenzimidazole (NBB) in MPN.
Scheme 1 Synthesis of N-alkylpyrrolidines and N-alkylpiperidines and
the corresponding N-alkylpyrrolidiniumpropyltriethoxysilyl iodides.
Reagents and conditions: (i) & (iii) amine (3 eq), alkyl bromide (2 eq),
K2CO3 (1.1 eq), methanol, reflux 12 h. (ii) alkylamine (1.1 eq), 3-iodo-
propyltriethoxysilane (1.0 eq), toluene, reflux 2 days.
Testing
DSSC devices were tested using simulated sunlight (AM1.5
1000 W mꢀ2) provided by an Oriel solar simulator with an AM1.5
filter. Current–voltage characteristics were measured using
a Keithley 2400 source meter. Cells were biased from +1000 mV
to ꢀ300 mV in 10 mV steps with a 40 ms settling time (delay
between application of the bias potential and the current
measurement). These techniques have been described in detail by
Nazeeruddin et al.56 Typical reproducibilities were ꢁ10 mV for
Voc and ꢁ0.3 mA cmꢀ2 for Jsc. IPCE measurements were per-
formed using a similar data-collection system to the I–V
measurements, using monochromatic light (from 800 to 300 nm).
This was achieved by focussing a 300 W Xenon lamp (ILC
Technology, USA) through a Gemini 180 double mono-
chromator (Jobin Yvon, UK) onto the photoanode side of the
device being tested.
(pyrrolidinium or piperidinium), and the spacer separating the
cation from the siloxane moiety R2 (propyl and undecyl).
The synthetic route to form the N-alkylpyrrolidines (alkyl ¼
pentyl (5), hexadecyl (6) and benzyl (7)) and N-pentylpiperidine
(10) is shown in Scheme 1. Alkylation pyrrolidine and piperidine
was carried out following the methodology reported by Hart and
coworkers53 to reduce the possibility of forming N,N-dia-
lkylpyrrolidinium and N,N-dipentylpiperidinium salts. Reaction
of 5, 6, and
7 with 3-iodopropyltriethoxysilane formed
3-(N-pentylpyrrolidinium)propyltriethoxysilyl iodide (3), 3-(N-
hexadecylpyrrolidinium)propyltriethoxysilyl iodide (8), 3-(N-
benzylpyrrolidinium)propyltriethoxysilyl iodide (9) in yields
ranging from 58–89% using conditions described by Manoso and
DeShong60 (Scheme 1).
Photovoltage and photocurrent decay characteristics were
measured using methodology developed by Zhang et al.57
The longer chain iodides were prepared by an adaptation of
a literature method52 using 11-iodoundecyltriethoxysilane (13) as
shown in Scheme 2. Reaction of equimolar amounts of 5 and 10
with 12 in toluene under inert conditions and exclusion of light
A
white light bias was applied to the photoanode side of the device
and a disturbance in the system was generated by the application
of a 200 ms red pulse. The intensity of the pulse was controlled to
keep the voltage change below 10 mV and a range of white light
bias intensities ranging from 150% to 0.1% Sun were used. After
this disturbance, the exponential photovoltage, photocurrent
decay and the photoinduced charge density at the same intensity
were recorded and the photoinduced charge density was calcu-
lated using the method of Frank.58 These measured parameters
were used to determine band-edge movements and recombina-
tion rates.59
afforded
11-(N-pentylpyrrolidinium)undecyl-triethoxysilyl
iodide (4) and 11-(N-pentylpiperidinium)-undecyltriethoxysilyl
Results and discussion
Synthesis and characterisation
Scheme 2 Synthesis of 11-iodoundecyltriethoxysilyl iodides. Reagents
and conditions: (i) N-pentylpyrrolidine (1.1 eq), 11-iodoundecyltriethox-
ysilane (1.0 eq), toluene, reflux 2 days, (ii) N-pentylpiperidine (1.1 eq),
11-iodoundecyltriethoxysilane (1.0 eq), toluene, reflux 2 days.
A family of triethoxysilyl iodides have been prepared which have
three major structural variations (Fig. 5): the pendant group R1,
(pentyl, hexadecyl and benzyl), the heterocyclic cation N+,
This journal is ª The Royal Society of Chemistry 2010
J. Mater. Chem., 2010, 20, 3694–3702 | 3697

View Article Online
iodide (11), respectively, in isolated yields between 84–90%. The
physical appearance of the isolated silyl iodides ranged in
consistency from viscous liquids to waxy solids and colouration
electrolyte thereby reducing the rate of recombination between
the photoinjected electron and the triiodide in the electrolyte.
The formation of this layer could be evident from the recombi-
nation rates, as a better insulator would give rise to longer
recombination rates. For this analysis, three 3-(N-alkylpyrroli-
dinium)propyltriethoxysilane iodides were studied which have
pentyl (3), hexadecyl (8) and benzyl (9) pendant.
ranging from pale yellow to pale brown. H and 13C{1H} NMR
1
spectroscopy, electrospray mass spectrometry and elemental
analyses were used to confirm the purity and composition of each
product.
The data clearly show that the use of the electrolyte containing
8 as the sole iodide source results in poorer efficiencies (Table 1),
cf. 3 and 9, especially at high light intensities (Fig. 6). At 1 Sun,
the values for Jsc, Voc and fill factors were all lower for 8 than
either 3 or 9, by ca. 1.5–4 mA cmꢀ2, 30–90 mV and 0.1–0.27,
respectively, for all three electrolyte solutions studied. The
combined effect of these differences are reflected in cell efficien-
cies that are almost half those measured for 3 and 9 even at
decreased iodide concentrations. At lower sunlight intensity, the
differences in these parameters are reduced, especially for the Voc
and fill factors. From these results it can be concluded that the
poor performance of these devices can be attributed to mass
transport limitations. In comparison, the I–V curves and corre-
sponding parameters for devices constructed using compounds 3
(pentyl pendant) and 9 (benzyl pendant) were very similar, even
DSSC testing
The I–V characteristics were measured for devices constructed
with each siloxyl iodide as the sole iodide source. Variations in
the organic solvent systems were investigated to establish
whether the solvent matrix influences device performance. The
examples of I–V profiles shown in Fig. 4 clearly show that higher
concentrations of the organic iodide (solid line) lead to poorer fill
factors, especially at high light intensities (1 Sun), in comparison
to DSSCs constructed using lower concentrations (dotted line).
Consequently, although values are quoted for devices con-
structed using 1.0 M organic iodide concentrations, most testing
has been conducted with organic iodide in 0.6 M concentrations.
Transient photovoltage and photocurrent decay experiments,
which involve monitoring the short circuit current (Jsc) and open
circuit voltage (Voc) decay following a small perturbation to the
steady state, allow for in-depth probing of the electronic
behaviour of the DSSC. This provides a measure of changes to
the conduction band potential of the titania and recombination
rates caused by variations in the electrolyte, such as the effect of
structural variations of the siloxyl iodide on device performance.
Two key parameters determined in these experiments are the
density of states (DOS) and recombination rate. The DOS of the
device is a measure of the chemical capacitance in the titania film
at a given light intensity and when plotted as a function of Voc,
gives a measure of the conduction band potential of the titania.23
The recombination rate, calculated from the photocurrent decay,
is the rate at which the photoinjected electrons recombine with
the oxidised redox mediator present in the electrolyte; slower
recombination rates generally translate into better device
performance.
Table 1 Performance parameters for devices constructed using pentyl
(3), hexadecyl (8), or benzyl (9) substituted 3-(N-alkylpyrrolidinium)-
propyltriethoxysilyl iodide in electrolytes prepared with AN : VN or
MPN as solvent. (Measurements were made at 1.0 and 0.1 sunlight
intensities)
Voc/mV
Jsc/mA cmꢀ2
FF
% h
0.1 Sun 1 Sun 0.1 Sun 1 Sun 0.1 Sun 1 Sun 0.1 Sun 1 Sun
3a 639
8a 622
9a 644
3b 666
8b 655
9b 684
3c 703
8c 614
9c 701
694
669
697
720
704
742
768
675
761
1.4
1.3
1.5
1.2
1.6
1.5
1.4
1.3
1.4
12.8
8.8
0.75
0.66
0.73
0.75
0.67
0.74
0.79
0.66
0.80
0.58
0.43
0.53
0.52
0.41
0.55
0.69
0.42
0.67
7.3
5.6
7.4
7.9
7.3
8.1
8.0
5.0
8.1
5.2
2.6
5.0
5.5
3.1
6.0
6.9
3.3
6.9
11.3
14.6
10.8
14.8
13.1
11.6
13.6
a
1.0 M organic iodide, 0.15 M I2, 0.5 M N-butylbenzimidazole (NBB) in
3-methoxypropionitrile (MPN). b 1.0 M organic iodide, 0.03 M I2, 0.5 M
tert-butylpyridine (TBP) in 3 : 1 acetonitrile:valeronitrile (AN : VN).
Effect of the pendant group on DSSC performance
In the first series of DSSC experiments, the pendant group (R1)
attached to the quaternary nitrogen was varied. Three different
effects arising from changing the pendant group could be
envisaged. Firstly, the size can affect the efficiency of the iodide
diffusion as larger pendants are expected to result in slower
diffusion, which in turn leads to poorer fill factors and
lower efficiencies as a result of mass transport issues. Such
c
0.6 M organic iodide, 0.03 M I2, 0.5 M TBP, 0.5 M N-ethyl-N0-
methylimidazolium bis(trifluoromethanesulfonyl)imide (EMITFSI) in
3 : 1 AN : VN.
a
correlation was seen in previous experiments using
trialkylammoniumpropyltrimethoxysilyl iodides, where
increasing the chain length (up to twelve carbons) led to poorer
performances.47 Secondly, the steric demands of the pendant can
affect the nature of the cation–titania interactions, causing
changes in conduction band potentials. In this case, a positively
shifted conduction band potential would indicate a stronger
cation–titania interaction due to a greater transfer of positive
charge from the cation to the titania. Thirdly, the pendant group
could form an insulating layer between the titania and the
Fig. 6 I–V profile of devices constructed using 3, 8 or 9 in an electrolyte
containing 0.6 M 3, 8 or 9, 0.03 M I2, 0.5 M TBP, 0.5 M EMITFSI in 3 : 1
AV : VN.
3698 | J. Mater. Chem., 2010, 20, 3694–3702
This journal is ª The Royal Society of Chemistry 2010

View Article Online
improved cation–titania interactions is the generation of an
insulating layer that could reduce the recombination processes.
The results of I–V testing for 4 and 11, which vary only in their
heterocyclic nitrogen component are summarised in Table 2.
For the MPN-based electrolyte containing 1.0 M siloxyl
iodide, the similarity of the results suggests that the cation
identity – pyrrolidinium or piperidinium – has little effect on
device performance when using this solvent. At the same organic
iodide concentration, no measurable differences were observed,
even when the solvent was changed to the less viscous AN : VN
mixture. The large drop in fill factor and overall device efficiency
when increasing the light intensity from 0.1 Sun to 1.0 Sun
suggests that poor mass transport through the electrolyte was
impacting adversely on device performance, which was attrib-
uted to the long undecyl spacer separating the siloxyl group from
the cation. To minimise these effects, the concentration was
decreased to 0.6 M (cf. 1.0 M) and 0.5 M EMITFSI was added to
improve the conductivity of the electrolyte. The measured I–V
traces are shown in Fig. 8 and a comparison of the cation effect
using the undecyl compounds 4 and 11 at the reduced organic
iodide concentrations shows that the values for the Voc for the
piperidinium compound 11 are ꢄ50 mV higher than those
measured for the pyrrolidinium compound 4 at both high and
low light intensities (Table 2). These higher Voc values are
compensated for by a concomitant decrease in Jsc (0.5 mA cmꢀ2
and 0.2 mA cmꢀ2 at 1.0 and 0.1 Sun, respectively) such that the
device efficiency remains constant.
Fig. 7 Plots of: (a) density of states as a function of Voc: (b) recombi-
nation lifetime as a function of density of states for devices constructed
using 0.6 M 3 or 9, 0.03 M I2, 0.5 M TBP, 0.5 M EMITFSI in 3 : 1 AV :
VN.
for electrolytes containing reduced concentrations of these
siloxyl iodides (see Table 1).
The plot of density of states (DOS) as a function of open
circuit voltage (Voc) generated from the photovoltage/photo-
current decay analysis shown in Fig. 7a indicates that at the same
photoinduced electron density, the benzyl analogue gives rise to
a small, positive band-edge shift of ca. 20 mV in comparison to
the pentyl analogue, particularly at higher light intensities. This
result suggests that the benzyl pendant imparts a slightly greater
positive character to the titania than the pentyl which could arise
from stronger cation–titania interactions. Interestingly, the
results indicate that there is no discernible difference between the
two pendants with respect to recombination rates (Fig. 7b)
suggesting that although cation–titania interactions vary for the
different pendants, in this case, the pendant groups themselves
are unable to shield the surface from recombination reactions.
Photovoltage/photocurrent decay measurements were carried
out to probe the origin of these differences in DSSC perfor-
mance. The plot of DOS as a function of Voc (Fig. 9a) shows that
the presence of a pyrrolidinium cation in the iodide source, leads
to a positive shift in the conduction band-edge potentials by
Effect of the cation on DSSC performance
In contrast to the possible multi-modal action of the pendant, the
major mode of action of the cationic pyrrolidinium or piper-
idinium groups (N+) is through electrostatic interactions with
ꢀ
anionic sites on the titania surface and the anionic Iꢀ/I3 redox
couple. The geometry of the cationic nitrogen centre has the
possibility of influencing the extent to which the cation to
interacts with the titania. However, a beneficial effect of
Fig. 8 I–V profile of devices constructed using 0.6 M 4 or 11, 0.03 M I2,
0.5 M TBP, 0.5 M EMITFSI in 3 : 1 AN : VN.
Table 2 Performance parameters for cells constructed using either 11-(N-pentylpyrrolidinium)undecyltriethoxysilyl iodide, 4, or 11-(N-pentylpiper-
idinium)undecyltriethoxysilyl iodide, 11, in electrolytes using AN : VN or MPN as solvent. (Measurements were made at 1.0 and 0.1 sunlight intensities)
Voc/mV
Jsc/mA cmꢀ2
FF
% h
0.1 Sun
1 Sun
0.1 Sun
1 Sun
0.1 Sun
1 Sun
0.1 Sun
1 Sun
4a
722
716
697
698
710
760
769
769
749
753
769
820
1.2
1.3
1.6
1.6
1.4
1.2
10.5
10.5
15.4
14.0
13.9
12.6
0.65
0.67
0.66
0.66
0.75
0.76
0.39
0.43
0.32
0.35
0.62
0.65
6.0
6.3
7.8
7.7
8.5
8.6
3.1
3.5
3.2
3.5
7.0
7.2
11a
4b
11b
4c
11c
a
1.0 M organic iodide, 0.15 M I2, 0.5 M NBB in MPN. b 1.0 M organic iodide, 0.03 M I2, 0.5 M TBP in 3 : 1 AN : VN. c 0.6 M organic iodide, 0.03M I2,
0.5 M TBP, 0.5 M EMITFSI in 3 : 1 AN : VN.
This journal is ª The Royal Society of Chemistry 2010
J. Mater. Chem., 2010, 20, 3694–3702 | 3699

View Article Online
Fig. 10 I–V analysis of devices constructed using 0.6 M 3 or 4, 0.03 M I2,
0.5 M TBP, 0.5 M EMITFSI in 3 : 1 AN : VN.
charge through the electrolyte when using 4 as the iodide source.
For the AN : VN solvent system, although the longer chain
analogue was hampered by diffusion limitations as evidenced by
their reduced fill factors, these devices showed larger short circuit
currents than the propyl analogue.
Fig. 9 Plot of: (a) density of states as a function of Voc: (b) recombi-
nation lifetime as a function of density of states for devices constructed
using 0.6 M 4 or 11, 0.03 M I2, 0.5 M TBP, 0.5 M EMITFSI in 3 : 1
AV : VN.
To improve diffusion, the iodide concentrations in the elec-
trolyte was decreased to 0.6 M and 0.5 M N-ethyl-N0-methyl-
imidazolium bis(trifluoromethanesulfonyl)imide (EMITFSI)
was added as a charge carrying assistor. For these electrolytes,
larger short circuit currents are achieved for 4, than for 3 at all
light intensities. Interestingly, at low light intensities, improve-
ments in fill factor lead to devices with larger overall device
efficiencies for the undecyl siloxyl iodide (Fig. 10, Table 3). It is
worth noting that % by mass of siloxyl iodide present in the
1.0 M solutions is 60% for 3 and 74% for 4 and the corresponding
almost 40 mV in comparison to the piperidinium cation, with
these effects becoming more pronounced at high light intensities.
Analysis of the plot of recombination rate as a function of DOS
(Fig. 9b) shows that there is no significant change in recombi-
nation lifetime when using the pyrrolidinium cation, suggesting
that while the pyrrolidinium cation can interact better with the
titania surface, it is unable to provide any additional insulative
effects.
Effects of the spacer length on DSSC performance
The length of the spacer (R2) could affect the nature of the cation/
pendant interaction with the titania and, depending on the effi-
ciency of shielding of the titania surface, could therefore influ-
ence recombination rates. To examine the effect of the spacer,
compounds 3 (R2 ¼ propyl spacer) and 4 (R2 ¼ undecyl spacer)
were used as iodide sources in DSSC electrolytes and the results
of the I–V analysis are summarised in Table 3. At iodide
concentrations of 1.0 M, the open circuit voltages are 30–80 mV
larger for the device constructed using 4 than those constructed
using 3 at all light intensities especially when using MPN as the
solvent, although this difference becomes negligible when
the concentration was decreased to 0.6 M. However, use of the
longer chain analogue results in much lower fill factors, giving
rise to reduced device efficiencies in comparison to the propyl
analogue, 3. These effects were more pronounced at high light
intensities (Table 3) and were attributed to poor diffusion of
Fig. 11 Plot of: (a) density of states as a function of Voc: (b) recombi-
nation lifetime as a function of density of states for devices constructed
using 0.6 M 3 or 4, 0.03 M I2, 0.5 M TBP, 0.5 M EMITFSI in 3 : 1 AN :
VN.
Table 3 Performance parameters for cells constructed using either 3-(N-pentylpyrrolidinium)propyltriethoxysilyl iodide, 3, or 11-(N-pentylpyrroli-
dinium)undecyltriethoxysilyl iodide, 4, in electrolytes using AN : VN or MPN as solvent. (Measurements were made at 1.0 and 0.1 sunlight intensities)
Voc/mV
Jsc/mA cmꢀ2
FF
% h
0.1 Sun
1 Sun
0.1 Sun
1 Sun
0.1 Sun
1 Sun
0.1 Sun
1 Sun
3a
4a
3b
4b
3c
4c
639
722
666
697
703
710
694
769
720
749
768
769
1.4
1.2
1.2
1.6
1.4
1.5
12.8
10.5
14.6
15.4
13.1
13.9
0.75
0.65
0.75
0.66
0.79
0.75
0.58
0.39
0.52
0.32
0.69
0.62
7.3
6.0
8.0
7.8
8.0
8.5
5.2
3.1
5.5
3.2
7.8
7.0
a
1.0 M organic iodide, 0.15 M I2, 0.5 M NBB in MPN. b 1.0 M organic iodide, 0.03 M I2, 0.5 M TBP in 3 : 1 AN : VN. c 0.6 M organic iodide, 0.03 M I2,
0.5 M TBP, 0.5 M EMITFSI in 3 : 1 AN : VN.
3700 | J. Mater. Chem., 2010, 20, 3694–3702
This journal is ª The Royal Society of Chemistry 2010

View Article Online
recombination rates in comparison to the piperidinium cation.
Although the pyrrolidinium cation may be geometrically better
suited to forming strong cation-titania interactions, this
improvement is not accompanied by any surface insulative
benefits. Compound 4, with the undecyl spacer R2 provided the
most effective surface insulation properties as evidenced by the
greatly improved recombination rates and short circuit currents
for these devices in comparison to ones constructed using the
propyl spacer 3.
Fig. 12 IPCE plots for devices constructed using either compound 3 or 4
as iodide sources (electrolyte composition 0.6 M 3 or 4, 0.03 M I2, 0.5 M
TBP in 3 : 1 AN:VN).
These results give insight into how these compounds interact
with the titania and the redox mediator and lead to the possibility
of intelligent design of new organic iodide sources for DSSCs.
values for the 0.6 M solutions are 36% and 45% (cf. the standard
imidazolium iodide, N-propyl-N0-methylimidazolium iodide
which is 32% and 19% by weight, respectively). Thus it is not
surprising to find a substantial improvement in cell performance
when the concentration is decreased and a charge-assisting
carrier is introduced and that this improvement is higher – in
relative terms – for devices constructed using the undecyl
derivative.
Acknowledgements
This work was supported by the Australian Research Council,
through the Linkage International, Discovery, Centre of Excel-
lence (Australian Centre for Electromaterials Science) Programs,
Victorian Organic Solar Cells Consortium and the Swiss
National Science Foundation.
Photovoltage/photocurrent decay analysis reveals a slight
positive shift (ꢄ 15 mV at low light intensity) in the conduction
band potential for devices using 3 (propyl spacer) as the iodide
source in comparison to those using 4 (Fig. 11a) however there
are sizeable differences in electron recombination rates
(Fig. 11b). 4 gives rise to devices with almost twice the recom-
bination lifetimes in comparison to 3, suggesting better insu-
lation with longer spacers. As improved insulation leads to
increased electron density and improved Voc, it is believed that
this is why devices constructed using the undecyl spacer show
slightly larger photovoltages in comparison to the propyl one.
IPCE analysis was performed using the best devices constructed
with 3 and 4 as iodide sources (Fig. 12). Although the two profiles
are almost identical, there is a small improvement in IPCE for 4
at shorter wavelengths which is reflected in a slightly higher
current for devices constructed using 4, cf. 3.
References
€
1 B. O’Regan and M. Gratzel, Nature, 1991, 353, 737–740.
2 M. Quintana, T. Edvinsson, A. Hagfeldt and G. Boschloo, J. Phys.
Chem. C, 2007, 111, 1035–1041.
3 A. B. F. Martinson, J. E. McGarrah, M. O. K. Parpia and J. T. Hupp,
Phys. Chem. Chem. Phys., 2006, 8, 4655–4659.
4 C.-Y. Lin, C.-F. Lo, L. Luo, H.-P. Lu, C.-S. Hung and
E. W.-G. Diau, J. Phys. Chem. C, 2009, 113, 755–764.
5 Y. Lee, S.-R. Jang, R. Vittal and K.-J. Kim, New J. Chem., 2007, 31,
2120–2126.
6 K. Wongcharee, V. Meeyoo and S. Chavadej, Sol. Energy Mater. Sol.
Cells, 2007, 91, 566–571.
7 X.-F. Wang, O. Kitao, H. Zhou, H. Tamiaki and S. Sasaki, J. Phys.
Chem. C, 2009, 113, 7954–7961.
8 M. K. Nazeeruddin, P. Pechy, T. Renouard, S. M. Zakeeruddin,
R. Humphry-Baker, P. Comte, P. Liska, L. Cevey, E. Costa,
V. Shklover, L. Spiccia, G. B. Deacon, C. A. Bignozzi and
€
M. Gratzel, J. Am. Chem. Soc., 2001, 123, 1613–1624.
9 S. Hayashi, Y. Matsubara, S. Eu, H. Hayashi, T. Umeyama,
Y. Matano and H. Imahori, Chem. Lett., 2008, 37, 846–847.
10 M. K. Nazeeruddin, T. Bessho, L. Cevey, S. Ito, C. Klein, F. De
Conclusion
€
Angelis, S. Fantacci, P. Comte, P. Liska, H. Imai and M. Gratzel,
A family of siloxane-based organic iodides of general formula
R1N+R2-Si(OEt)3 were investigated as iodide sources for DSSC
electrolytes. Variations to the pendant group (R1) revealed that
very long alkyl chains led to poorer performing devices origi-
nating from mass transport/diffusion problems. The benzyl
pendant gave rise to positively shifted conduction band poten-
tials in comparison to the pentyl pendant, without changing
recombination rates. Thus, although the benzyl group appears to
allow for a stronger cation–titania interaction, there are no
accompanying insulative effects. While the pendant group can
affect how the cation interacts with the surface – as evidenced by
changing Voc – it seems unable to interact itself with the titania
surface to any great extent, hence no differences in recombina-
tion rates were observed. This could be due to the short nature of
the propyl spacer, where steric demands would play a greater role
than the identity of the pendant. When analysing the cation (N+)
effect, the pyrrolidinium cation appears able to better interact
with the titania surface, giving rise to positive shifts in the
conduction band without any accompanying improvements in
J. Photochem. Photobiol., A, 2007, 185, 331–337.
11 D. Kuang, S. Ito, B. Wenger, C. Klein, J. E. Moser, R. Humphry-
€
Baker, S. M. Zakeeruddin and M. Gratzel, J. Am. Chem. Soc.,
2006, 128, 4146–4154.
12 S. M. Zakeeruddin, M. K. Nazeeruddin, P. Pechy, F. P. Rotzinger,
R. Humphry-Baker, K. Kalyanasundaram, M. Graetzel,
V. Shklover and T. Haibach, Inorg. Chem., 1997, 36, 5937–5946.
13 M. Yanagida, L. P. Singh, K. Sayama, K. Hara, R. Katoh, A. Islam,
€
H. Sugihara, H. Arakawa, M. K. Nazeeruddin and M. Gratzel,
J. Chem. Soc., Dalton Trans., 2000, 2817–2822.
14 S. Ruile, O. Kohle, P. Pechy and M. Graetzel, Inorg. Chim. Acta,
1997, 261, 129–140.
15 C. Baik, D. Kim, M.-S. Kang, S. O. Kang, J. Ko, M. K. Nazeeruddin
€
and M. Gratzel, J. Photochem. Photobiol., A, 2009, 201, 168–174.
16 W. M. Campbell, A. K. Burrell, D. L. Officer and K. W. Jolley,
Coord. Chem. Rev., 2004, 248, 1363–1379.
17 G. Ramakrishna, D. A. Jose, D. K. Kumar, A. Das, D. K. Palit and
H. N. Ghosh, J. Phys. Chem. B, 2005, 109, 15445–15453.
18 B. O’Regan and F. Lenzmann, J. Phys. Chem. B, 2004, 108, 4342–
4350.
19 B. A. Gregg, F. Pichot, S. Ferrere and C. L. Fields, J. Phys. Chem. B,
2001, 105, 1422–1429.
20 Z. Zhang, P. Chen, T. N. Murakami, S. M. Zakeeruddin and
€
M. Gratzel, Adv. Funct. Mater., 2008, 18, 341–346.
This journal is ª The Royal Society of Chemistry 2010
J. Mater. Chem., 2010, 20, 3694–3702 | 3701

View Article Online
€
40 S. Pelet, J.-E. Moser and M. Gratzel, J. Phys. Chem. B, 2000, 104,
1791–1795.
41 N.-G. Park, S.-H. Chang, J. Van de Lagemaat, K.-J. Kim and
A. J. Frank, Bull. Kor. Chem. Soc., 2000, 21, 985–988.
42 S. Nakade, T. Kanzaki, S. Kambe, Y. Wada and S. Yanagida,
Langmuir, 2005, 21, 11414–11417.
21 H. Matsui, K. Okada, T. Kawashima, T. Ezure, N. Tanabe,
R. Kawano and M. Watanabe, J. Photochem. Photobiol., A, 2004,
164, 129–135.
22 Q. Dai, D. B. Menzies, D. R. MacFarlane, S. R. Batten, S. Forsyth,
L. Spiccia, Y.-B. Cheng and M. Forsyth, C.R. Chimie, 2006, 9, 617–
621.
23 V. Jovanovski, E. Stathatos, B. Orel and P. Lianos, Thin Solid Films,
2006, 511–512, 634–637.
43 T. Kanzaki, S. Nakade, Y. Wada and S. Yanagida, Photochem.
Photobiol. Sci., 2006, 5, 389–394.
24 N. Yamanaka, R. Kawano, W. Kubo, N. Masaki, T. Kitamura,
Y. Wada, M. Watanabe and S. Yanagida, J. Phys. Chem. B, 2007,
111, 4763–4769.
25 J. A. Widegren, E. M. Saurer, K. N. Marsh and J. W. Magee,
J. Chem. Thermodyn., 2005, 37, 569–575.
44 S. Nakade, T. Kanzaki, W. Kubo, T. Kitamura, Y. Wada and
S. Yanagida, J. Phys. Chem. B, 2005, 109, 3480–3487.
45 K. Hara, T. Nishikawa, M. Kurashige, H. Kawauchi, T. Kashima,
K. Sayama, K. Aika and H. Arakawa, Sol. Ener. Mat. Sol., 2005,
85, 21–30.
26 H. Paulsson, L. Kloo, A. Hagfeldt and G. Boschloo, J. Electroanal.
Chem., 2006, 586, 56–61.
27 P. Wang, S. M. Zakeeruddin, P. Comte, G. Garcia-Belmonte and
46 C. A. Kelly, F. Farzad, D. W. Thompson, J. M. Stipkala and
G. J. Meyer, Langmuir, 1999, 15, 7047–7054.
€
47 S. Cerneaux, S. M. Zakeeruddin, M. Gratzel, Y.-B. Cheng and
L. Spiccia, J. Photochem. Photobiol., A, 2008, 198, 186–191.
48 S. Cerneaux, S. M. Zakeeruddin, J. M. Pringle, Y.-B. Cheng,
€
M. Gratzel, J. Am. Chem. Soc., 2003, 125, 1166–1167.
28 R. Komiya, L. Han, R. Yamanaka, A. Islam and T. Mitate,
J. Photochem. Photobiol., A, 2004, 164, 123–127.
29 H. Usui, H. Matsui, N. Tanabe and S. Yanagida, J. Photochem.
Photobiol., A, 2004, 164, 97–101.
30 F. Mazille, Z. Fei, D. Kuang, D. Zhao, S. M. Zakeeruddin,
€
M. Gratzel and L. Spiccia, Adv. Funct. Mater., 2007, 17, 3200–3206.
49 N. A. Lewcenko, M. J. Byrnes, Y.-B. Cheng, S. M. Zakeeruddin,
€
M. Gratzel and L. Spiccia, Chem. Commun., 2008, 3852–3854.
50 D. D. Perrin and W. L. F. Armarego, Purification of Laboratory
Chemicals, 3rd edn, 1988.
€
M. Gratzel and P. J. Dyson, Inorg. Chem., 2006, 45, 1585–1590.
31 M. Wang, X. Xiao, X. Zhou, X. Li and Y. Lin, Sol. Energy Mater.
Sol. Cells, 2007, 91, 785–790.
51 Geraud Dubois, R. Tripier, S. Brandes, F. Denat and R. Guilard,
J. Mater. Chem., 2002, 12, 2255–2261.
32 K. Suzuki, M. Yamaguchi, S. Hotta, N. Tanabe and S. Yanagida,
J. Photochem. Photobiol., A, 2004, 164, 81–85.
52 T. W. Baughman, J. C. Sworen and K. B. Wagener, Tetrahedron,
2004, 60, 10943–10948.
33 M. Wang, X. Yin, X. R. Xiao, X. Zhou, Z. Z. Yang, X. P. Li and
Y. Lin, J. Photochem. Photobiol., A, 2008, 194, 20–26.
34 V. Jovanovski, B. Orel, R. Jese, A. S. Vuk, G. Mali, S. B. Hocevar,
J. Grdadolnik, E. Stathatos and P. Lianos, J. Phys. Chem. B, 2005,
109, 14387–14395.
53 W. F. Hart and M. E. McGreal, J. Org. Chem., 1957, 22, 86–88.
54 P. Wang, S. M. Zakeeruddin, R. Humphry-baker, J. E. Moser and
M. Graetzel, Adv. Mater., 2003, 15, 2101–2104.
€
55 Q. Wang, S. Ito, M. Gratzel, F. Fabregat-Santiago, I. Mora-Sero,
J. Bisquert, T. Bessho and H. Imai, J. Phys. Chem. B, 2006, 110,
25210–25221.
56 M. K. Nazeeruddin, S. M. Zakeeruddin, J.-J. Lagref, P. Liska,
35 N. Papageorgiou, Y. Athanassov, M. Armand, P. Bonhote,
€
H. Pettersson, A. Azam and M. Gratzel, J. Electrochem. Soc., 1996,
€
P. Comte, C. Barolo, G. Viscardi, K. Schenk and M. Gratzel,
Coord. Chem. Rev., 2004, 248, 1317–1328.
57 Z. Zhang, N. Evans, S. M. Zakeeruddin, R. Humphry-Baker and
143, 3099–3108.
36 E. Stathatos, P. Lianos, S. M. Zakeeruddin, P. Liska and M. Gratzel,
€
Chem. Mater., 2003, 15, 1825–1829.
37 D. Kuang, P. Wang, S. Ito, S. M. Zakeeruddin and M. Gratzel, J. Am.
Chem. Soc., 2006, 128, 7732–7733.
€
€
M. Gratzel, J. Phys. Chem. C, 2007, 111, 398–403.
58 N. Kopidakis, K. D. Benkstein, J. Van de Lagemaat and A. J. Frank,
J. Phys. Chem. B, 2003, 107, 11307–11315.
59 S. Nakade, T. Kanzaki, Y. Wada and S. Yanagida, Langmuir, 2005,
21, 10803–10807.
60 A. S. Manoso and P. DeShong, J. Org. Chem., 2001, 66, 7449–7455.
38 P. Wang, S. M. Zakeeruddin, J. E. Moser, M. K. Nazeeruddin,
T. Sekiguchi and M. Graetzel, Nat. Mater., 2003, 2, 402–407.
39 Z. Fei, D. Kuang, D. Zhao, C. Klein, W. H. Ang, S. M. Zakeeruddin,
€
M. Gratzel and P. J. Dyson, Inorg. Chem., 2006, 45, 10407–10409.
3702 | J. Mater. Chem., 2010, 20, 3694–3702
This journal is ª The Royal Society of Chemistry 2010