The synthesis and properties of solution processable phenyl cored thiophene dendrimers

Article abstract of DOI:10.1039/b508683c

In this paper we describe the convergent synthesis of a new class of phenyl cored thiophene dendrimers, which are promising candidates for use in organic semiconductor devices. We have prepared dendrimers with three and four dendrons around the core as we

Full text of DOI:10.1039/b508683c

View Article Online / Journal Homepage / Table of Contents for this issue
PAPER
www.rsc.org/materials | Journal of Materials Chemistry
The synthesis and properties of solution processable phenyl cored thiophene
dendrimers
William J. Mitchell,* Nikos Kopidakis, Garry Rumbles, David S. Ginley and Sean E. Shaheen
Received 20th June 2005, Accepted 12th August 2005
First published as an Advance Article on the web 9th September 2005
DOI: 10.1039/b508683c
In this paper we describe the convergent synthesis of a new class of phenyl cored thiophene
dendrimers, which are promising candidates for use in organic semiconductor devices. We have
prepared dendrimers with three and four dendrons around the core as well as dendrimers with 1st
and 2nd generation dendrons. All the dendrimers were soluble in common organic solvents such
as chloroform, THF and toluene. The structures and size properties are confirmed by a number of
techniques including NMR, GPC and MALDI-TOF-MS. Decomposition was studied by TGA
with initial breakdown of the hexyl surface groups followed by the aromatic core. The
spectroscopic properties were studied by UV-vis and PL spectrometry which demonstrated
substantial differences between the dendrimers with three and four dendrons. Optical band gaps
varied between 2.34 and 2.60 eV for thin films of the dendrimers and electronic band gaps were,
on average, 0.3 eV greater than the optical band gaps. The smallest band gap was measured for
the dendrimer with four 2nd generation dendrons around the phenyl core. Fluorescence lifetimes
of the molecules in solution ranged from 200 to 560 ps. This range in values was attributed to
differences in internal conversion rates associated with varying degrees of flexibility of the
extended dendrons.
12b
bithiophene moieties as the surface groups and Burn et al.
Introduction
used a thiophene group to tune the emissive colour of a
12f
dendrimer, designed for use in OLEDs.
Dendrimers are well-defined, monodisperse macromolecules
with a highly branched structure. Typically they consist of a
core connected to a number of branching units, known
as dendrons, with the degree of the branching of the
dendrons being known as the ‘‘generation’’ of the dendrimer.
Dendrimers can have aliphatic, non-conjugated dendrons or
Here we report the convergent synthesis of four dendrimers
containing a phenyl core with thiophene dendrons. The phenyl
core was chosen so that we could directly compare dendrimers
with three (1,3,5 positions) and four (1,2,4,5 positions)
dendron arms attached around the core. In addition, 1st and
1
the dendrimer as a whole can be conjugated. Conjugated
2
dendrimers include dendrons containing phenylacetylene,
2
nd generation dendrimers were prepared and characterised in
order to investigate the effects of the additional generation on
the properties of the molecule. Adding the 2nd generation is
shown to increase the band gap of the molecule due to an
increase in conjugation length, however steric forces that
twist the dendrons out of plane may mitigate the increase in
conjugation, especially as the dendron becomes larger in
volume. For all the dendrimers, spectroscopic and photo-
physical properties, size information and electrochemical
properties are reported.
3
phenylene and phenylenevinylene moieties. More recently
4
dendrimers have been prepared that are conjugated through
5
thiophene repeat units. Possible advantages of dendrimers
over conjugated polymers include a well defined molecular
weight, improved batch-to-batch reproducibility and a higher
degree of purity as synthesised using standard laboratory
procedures.
Thiophene has been increasingly used as the basis of
6
polymers and oligomers over the previous decade. The
materials have been utilised in organic light-emitting diodes
Results and discussion
7
OLEDs), field-effect transistors (FETs) and in photovoltaic
8
(
(
9
PV) applications. The properties of the materials can be
Synthesis and characterisation
tuned by introducing a variety of substituents either onto
The structures and synthetic pathway to the dendrimers are
shown in Schemes 1 and 2 and Chart 1. The synthesis of the
dendrimers was carried out via a convergent method where
the cores and dendrons are prepared separately and united in
the final step.
1
0
the backbone or as sidegroups. The stereochemistry of the
materials can also play an important role, for example
regioregular poly(3-alkylthiophene)s have superior electrical
6
a,11
properties compared to the regio-random equivalent.
More recently thiophenes have also been employed in a
The two different cores 3 and 6 were prepared from
commercially available 1,3,5-tribromobenzene 1 and 1,2,4,5-
5
,12
number of dendrimers.
For example, Roncali et al. used
tetrabromobenzene
4 respectively. Initial coupling with
tributyl(thiophene-2-yl)stannane via a Stille reaction gave
thiophene functionalised cores 2 and 5 in yields of 47% and
National Renewable Energy Laboratory, 1617 Cole Boulevard, Golden,
Colorado, 80401-3393, USA. E-mail: william_mitchell@nrel.gov
4
518 | J. Mater. Chem., 2005, 15, 4518–4528
This journal is ß The Royal Society of Chemistry 2005

View Article Online
2 3 2
Scheme 1 (a) Tributyl(thiophen-2-yl)stannane, PdCl (PPh ) , DMF,
3
90 uC; (b) NBS, AcOH, CHCl , 0 uC then rt.
5
a
Chart 1 Dendrons used.
3
7% respectively. Bromination of the newly introduced
thiophene groups was achieved by reaction of 2 and 5 with
a slight excess of N-bromosuccinimide. After work-up and
purification, bromides 3 and 6 were isolated in 84% and 81%
yields respectively.
using Stille coupling of the relevant core and the stannanes of
a
7 and 8. The stannanes were prepared by literature methods
5
and were used without purification of the crude material. The
core 3 or 6, G1 or G2-stannane and dry N,N-dimethylform-
amide were initially degassed with argon followed by addition
of the palladium catalyst with heating at 90 uC for 17 hours.
Work-up and purification by column chromatography gave
dendrimers 9 (3G1), 10 (3G2) and 11 (4G1) in excellent yields
of 93%, 96% and 85% respectively. However, 12 (4G2) was
only isolated in 11% yield due to incomplete reaction of all
four arms of the core even after prolonged reaction time at
elevated temperatures with a large excess of dendron. This
contributed to make the purification of 4G2 difficult due to
the similar polarity of the side products, however repeated
column chromatography allowed the isolation of the desired
product.
Dendrons 7 and 8, Chart 1, were additionally prepared. The
5a
synthesis of the two dendrons followed literature procedures.
Alkylation of thiophene with 1-bromohexane followed by
bromination gave 1-bromo-5-hexylthiophene in an overall 86%
yield. This was coupled with 2,3-dibromothiophene via a
Grignard reaction to give the G1 dendron 7 in 52% yield.
Further reaction of 7 with butyllithium followed by tributyltin
chloride gave a stannane functionalised G1 dendron, which
undertook a Stille coupling with 2,3-dibromothiophene to give
G2 dendron 8 in 88% yield.
The dendrimers were prepared following the synthetic route
outlined in Scheme 2. All four dendrimers were prepared by
Scheme 2 (a) Stannane dendron of 7 or 8, Pd(PPh
3
)
4
, DMF, 90 uC, 17 h.
This journal is ß The Royal Society of Chemistry 2005
J. Mater. Chem., 2005, 15, 4518–4528 | 4519

View Article Online
The two G1 dendrimers 9 and 10 are solid whereas the two
G2 dendrimers 11 and 12 are very viscous oils. All of the
dendrimers are soluble in common organic solvents such as
chloroform, THF and toluene, but polar solvents such as
methanol and DMSO are very poor solvents. The structures
The 1st generation dendrimers, 3G1 and 4G1, both show a
well-defined peak in both solution and thin film at approxi-
mately 400 nm, which is 45 nm further to the red than the
1
3
maximum for a-terthiophene at 355 nm. Interestingly, the
peak absorption of the lowest energy transition is the same for
both dendrimers in solution and thin film. The absorption
peak of 4G1 is however broader than that for 3G1 resulting in
an apparent red-shifted band edge. We do not attribute the
broadening effect to intermolecular interactions from dimer
formation; instead, the red-shift is consistent with the greater
conjugation associated with the phenyl core. Based on these
absorption data alone it is difficult to infer if the red-shift in
the leading edge of absorption of 4G1 is simply due to a
broadening of the peak at 400 nm relative to 3G1, or is in fact
a new absorption state that is truly red-shifted. However,
fluorescence data discussed later clearly indicate that the red-
shift can be attributed to a new absorption state. The
absorption peak at 400 nm is indeed similar to the absorption
peak of 405 nm found for a similar star-shaped molecule which
had three arms around a phenyl core with each arm consisting
1
13
and purity were confirmed by H and C NMR, MALDI-
TOF mass spectrometry and micro-analysis.
Spectroscopic properties
Absorption. The UV-vis spectra of the dendrimers in DCM
solutions and as thin films are shown in Fig. 1. The thin films
2
3
were prepared from a 20 mg cm solution of the dendrimer in
toluene spin-coated onto a quartz substrate at 1000 rpm. All
four dendrimers show a broad absorption but the onset to
absorption and the overall shape of the spectra differ between
the two different generations of the dendrimers as well as
the number of dendrons surrounding the core. However, the
spectra did not change when the solvent was altered from
DCM to THF, indicating limited solvatochromism. The
linkages between each thiophene unit are either a–a or a–b
connections with the former providing the greater p-conjuga-
tion. Therefore, the longest conjugation sequence of 3G1 and
8h
of a terthiophene. The absorption maximum is also similar to
5
b
a peak at 392 nm for a dendrimer previously prepared which
had two G1 dendrons 7 attached via an a–a connection giving
four thiophenes in conjugation. Additionally, the onset to
absorption is red-shifted by 88 nm for 3G1 and by 119 nm for
4
G1 is an a-terthiophene unit whereas for 3G2 and 4G2 there
are four thiophene units joined by a–a connections.
5
b
4
G1 when compared with the 1st generation dendron,
indicating improved conjugation through the 4G1 core versus
the 3G1 core. The 1st generation dendrimers also show
maxima at shorter wavelengths at approximately 320 and
2
50 nm. These shorter wavelength peaks have been previously
attributed to excitations that are localised on shorter conjuga-
tion sequences and are also identifiable in the 3G2 and 4G2
5
b
dendrimers. The absorptions of the 2nd generation dendri-
mers 3G2 and 4G2 have spectra with no clear discernable
peaks and a featureless tail to long wavelength with the
exception of a broad peak at 370 nm for 3G2 in solution and as
5
b
a film. This is similar to the G2 dendron 8 and a dendrimer
consisting of two G2 dendrons joined via an a–a connection
where the absorption spectra had no clear peaks indicating the
higher generation dendrons and dendrimers have broad
overlapping absorptions leading to featureless long wavelength
tails in their absorption spectra.
Optical band gaps, taken as the low energy onsets of the
absorption, are shown in Table 1. Overall the band gaps are
red-shifted for the films when compared with the solutions
by between 18 and 29 nm, which can be attributed to the
difference in polarisability of the solvent in comparison to
1
4
that of the neat film, as well as potential intermolecular
Table 1 Optical band gaps, taken as the low energy absorption
onsets, for the dendrimers in solution and film, and extinction
coefficients for the dendrimers in solution
Optical band gap/eV (l/nm)
2
1
21
Dendrimer
Solution
Film
e
max/M cm
3
4
3
4
G1
G1
G2
G2
2.70 (460)
2.53 (491)
2.60 (478)
2.48 (501)
2.60 (478)
2.42 (514)
2.49 (498)
2.34 (530)
85 000
115 000
76 000
112 000
Fig. 1 UV-vis absorption spectra of the dendrimers (a) in DCM, (b)
as thin films spin coated onto a quartz substrate from a toluene
2
3
.
solution, 20 mg cm
4
520 | J. Mater. Chem., 2005, 15, 4518–4528
This journal is ß The Royal Society of Chemistry 2005

View Article Online
through the benzene core at the ortho and para positions,
which the four dendrons occupy, rather than the meta
positions through which the three dendrons are related. As
before, the 2nd generation dendron has a lesser effect on
the band gap that we attribute to twisting of the dendron
backbone because of steric effects.
Thermal annealing of films cast on glass substrates was
performed for 3G1, 3G2, 4G1 and 4G2 for 10 min at 90 uC
under argon atmosphere. No change in the UV-vis absorption
was seen as would be the case with the analogous polymer
poly(3-hexylthiophene) (P3HT). P3HT is known to undergo a
transition to a more ordered state upon thermal or solvent
annealing of films, leading to a red-shift and concomitant
1
5
increase in vibronic structure of the absorption spectrum.
This effect is presumed to be driven by steric forces between
the hexyl side chains that yield an interdigitated structure that
improves the planarity and intermolecular p-stacking of the
thiophene backbones. We do not at present see such an effect
in the dendrimers presented here. Annealing at higher
temperatures (up to 200 uC) induced crystallisation in the
dendrimers and a broadband loss of optical absorption.
Emission. The corrected, fluorescence spectra of the
dendrimers in DCM, normalised to concentration, are shown
in Fig. 2a and the maxima given in Table 2. The wavelengths
of the maxima are red-shifted as the generation increases and
also as the number of dendrons around the core increases. The
fluorescence spectra are independent of the excitation wave-
length over a broad range. We ascribe this to intramolecular
energy transfer from the shorter conjugation branches to the
longest branch in an energy-funneling effect, resulting in
emission from the branch with lowest optical band gap
regardless of excitation energy. A similar effect was observed
Fig. 2 (a) Corrected fluorescence spectra for dendrimers in degassed
DCM normalised to absorbance at the excitation wavelength (444 nm).
(b) Fluorescence decay profiles recorded with excitation at 438 nm.
5
b
Solid line through data for 4G1 represents a single-exponential fit to
the data convoluted with the instrument response function (IRF).
for a set of dendrimers consisting solely of thiophene units.
The peak positions of the 3G2 (556 nm), 4G1 (547 nm) and
G2 (564 nm) fluorescence are seen to be red-shifted from that
4
interactions in the solid state. 3G2 exhibits an optical band gap
in solution that is 0.1 eV smaller than 3G1, while the reduction
in band gap between 4G2 and 4G1 is only 0.05 eV, indicating
that the higher generation has a greater effect when there
are only three dendrons around the core rather than four. We
attribute this to the orientation of the dendrons, since
accommodating four dendrons around the core results in
more steric hindrance. The dendrons bend or twist to alleviate
the congested structure, decreasing the orbital overlap of the
a–a linkages along the thiophene chain. Additionally, reduc-
tions in optical band gap in solution of 0.17 eV and 0.12 eV
are seen between 3G1 and 4G1 and between 3G2 and 4G2,
respectively. These reductions are due to greater conjugation
of 3G1 (508 nm). From this we conclude that adding a fourth,
1st generation dendron to 3G1 to form 4G1 (increased
conjugation through the core) or increasing the generation
(increased conjugation length of the dendron) to form 3G2
both result in a red-shift of approximately 30–40 nm. However
when both are done simultaneously to form 4G2, only a small
additional red-shift is seen, presumably due to steric hindrance
in the bulky structure leading to bending and twisting of the
dendrons that limits the degree of red-shift distortion.
The fluorescence spectra were also taken for thin films of
the dendrimers, however only very weak fluorescence was
observed. Low fluorescence efficiencies are often observed for
alkyl-substituted polythiophenes where y40% efficiency in
R f f
Table 2 Analysis of fluorescence spectra: peak position, quantum yield and lifetime. Natural radiative lifetime calculated from t = t /W
Fluorescence maximum
Fluorescence quantum
yield W (%)
Fluorescence
lifetime t /ps
Natural radiative
lifetime t /ns
Dendrimer
l/nm
Energy/eV
f
f
R
3
4
3
4
a
G1
G1
G2
G2
508
547
556
564
2.44
2.27
2.23
2.20
9 ¡ 1
17 ¡ 2
6 ¡ 1
4 ¡ 1
250 ¡ 10
560 ¡ 10
200 ¡ 10
340 ¡ 70
2.9 ¡ 0.1
3.4 ¡ 0.1
5.3 ¡ 0.2
6.2 ¡ 0.4
a
Value determined from average value of bi-exponential fitting function.
This journal is ß The Royal Society of Chemistry 2005
J. Mater. Chem., 2005, 15, 4518–4528 | 4521

View Article Online
Fig. 3 MALDI mass spectra of the four dendrimers (a) 3G1, (b) 4G1, (c) 3G2, (d) 4G2.
1
6
solution is reduced to ,2% in the solid state. This reduction
depends strongly on the regioregularity of the polymer and on
the nature of the species appended to the 3-position on the
yield and longest lifetime of the 4G1 dendrimer may arise from
the enhanced delocalisation effect through the ortho position
of the central phenyl ring, combined with the shorter dendron
resulting in a more rigid structure and hence lower rate of
internal conversion overall.
1
7
thiophene ring and is attributed to an ordering effect in
thin films leading to strong interactions between adjacent
chains, which can promote non-radiative decay. A similar
mechanism may explain the lack of fluorescence observed for
the dendrimers, where the disc-like structures could promote
close, parallel stacking in the solid state.
Mass properties
The mass properties were studied by Matrix Assisted
Laser Desorption/Ionisation (MALDI) mass spectrometry,
gel-permeation chromatography (GPC), and the molecular
decomposition was investigated by thermogravimetric analysis
Fluorescence quantum yields and lifetimes were determined
and the results are summarised in Table 2. The values fall in
the range 10 ¡ 7%, which are quite low, but these are
consistent with the short fluorescence lifetimes and follow the
same trend. The extracted natural radiative lifetimes indicate
that in each case, the radiative transitions are fully allowed,
consistent with the large molar decadic extinction coefficients.
Fig. 2b shows 4 decays recorded at the peak of the emission
spectrum. Also shown is the instrument response function and
a fit to the data of 4G1. With the exception of 4G2, most
decays measured across the emission profile were described
well by a single-exponential decay function with a charac-
teristic fluorescence lifetime. This simple decay function was
inapplicable to 4G2, however, where a biexponential decay
function was required to adequately describe the data. The
value quoted in Table 2 for this sample is a yield-weighted
average of the two extracted decay times, as the origin of the
biexponentiality is unclear. Although it does not appear to
be derived from purity or polarisation effects, it may be
attributable to effects such as spectral diffusion. However, a
more detailed study is required to resolve this issue and for the
purposes of this article, a larger uncertainty in the derived
fluorescence lifetime is quoted. The lower fluorescence
quantum yields and lifetimes of 3G2 and 4G2 may be
attributed to enhanced internal conversion associated with
the flexibility of the extended dendrons. The highest quantum
(
TGA). Fig. 3 shows the expanded MALDI spectra obtained
for the four dendrimers with the isotope distribution clearly
visible. Although the equipment used was not a dedicated
MALDI-TOF spectrometer, a high resolution spectrum was
observed for the smaller dendrimers. However, at higher
masses, the resolution decreases and extended acquisition
time was required to obtain the spectra. Table 3 shows the
calculated mass for the dendrimers along with the mass
obtained from the MALDI spectrometer. As expected, the
mass obtained from the MALDI measurements corresponds
closely to the calculated value. The inserts in Fig. 3a and b
show the first aggregates of the dendrimers at double the mass
of the parent dendrimer. Unfortunately, the third, fourth etc.
aggregates could not be observed for 3G1 and 4G1 and no
aggregates could be observed for the two 2nd generation
Table 3 Mass data obtained from MALDI and GPC analysis
Calculated
mass
MALDI mass
(MH )
GPC mass
(M
+
Dendrimer
W
)
3
4G1
G1
1566.5
2062.6
3054.9
4047.2
1567.6
2064.2
3056.0
4048.6
1790
2310
3400
4110
3
4
G2
G2
4
522 | J. Mater. Chem., 2005, 15, 4518–4528
This journal is ß The Royal Society of Chemistry 2005

View Article Online
Fig. 5 A typical TGA trace of the dendrimers, shown for 3G1.
carried out in dry DCM and the reductions performed in dry
THF with all measurements being carried out under argon.
The potentials shown are vs. SCE, and tetrabutylammonium
hexafluorophosphate was used as supporting electrolyte in a
0.1 M solution. Fig. 6 shows the solution cyclic voltammetry
results for the dendrimers. For 3G1, three waves are observed
during oxidation at 0.93, 1.15 and 1.43 V. The first wave is
reversible with a corresponding E1/2 of 0.87 V but the second
and third oxidations are irreversible. Upon reduction, five
waves are barely distinguishable at 21.79, 22.02, 22.38,
Fig. 4 GPC traces of dendrimers 3G1, 3G2, 4G1 and 4G2.
dendrimers due to the lower resolution of the higher masses.
This aggregation affect has previously been seen in dendrimers
5
consisting solely of thiophene units and was attributed to the
b
p–p stacking of the mostly planar aromatic units in the gas
1
8
phase after vaporisation.
The mass properties of the dendrimers were also studied by
GPC, and the weight averaged molecular weights (M ) were
W
2
2.64 and 22.96 V, all irreversible. Scanning 3G2 with
increasing potential shows two overlapping waves at E1/2 of
.81 and 0.93 V respectively which can be seen if the potential
is scanned up to 1150 mV. If the potential is scanned to
calculated against polystyrene standards. The traces are shown
in Fig. 4, and the molecular weights obtained are shown in
Table 3. Although deviations between the calculated masses
and the observed masses are seen, these are expected due to the
different shape and flexibility of the dendrimers in comparison
to the polystyrene standards used as calibrants. Indeed, it has
0
1
1
500 mV, two more irreversible waves are seen at 1.17 and
.41 V, and the reversibility of the first oxidation is lost.
1
9
During reduction, a reversible reduction wave with E1/2 of
been shown that GPC tends to overestimate the molecular
weights of rigid polymers such as P3HT. Interestingly, there is
a greater deviation for the smaller dendrimers than the largest
2
2
2.03 V and further irreversible reduction waves at 22.36 and
2.59 V are seen. Similar to 3G1, 4G1 shows an initial
reversible oxidation at E1/2 of 1.00 V and further irreversible
oxidations at 1.44 and 1.81 V. Additionally a reversible process
is observed while decreasing the potential at E1/2 21.75 V,
and further reductions are seen at 22.18 and 22.53 V, both
irreversible. 4G2 shows an initial broad oxidation at 1.15 V
comparable to that seen for 3G2. Two further oxidations at
4
G2 dendrimer following a similar trend for that observed for
5b
a different set of thiophene dendrimers. In that set the
smallest dendrimers were similarly overestimated in mass,
however the larger dendrimers were underestimated, indicating
that the higher generation dendrimers adopt a more spherical
structure leading to a greater retention time in the GPC. A
high degree of monodispersity is seen in the GPC traces for all
the dendrimers.
1
.48 and 1.68 V are observed while scanning increasing
potential but on decreasing the potential, the scan is ill
defined. During reduction, peaks can be seen at 21.92, 22.10
and 22.31 V which all show irreversibility. Interestingly, the
onset to oxidation for both the 2nd generation dendrimers is
less well defined than that for the 1st generation dendrimers.
This could be due to the larger number of possible orientations
the dendrons occupy in the 2nd generations dendrimers,
resulting in slightly different oxidation onsets and therefore a
broad onset for 3G2 and 4G2.
Decomposition of the dendrimers was followed by TGA. All
the dendrimers exhibited very similar decomposition curves. A
typical trace can be seen in Fig. 5. An initial decomposition
starts at about 400 uC with a corresponding weight loss of
approximately 35%. This is attributed to the loss of the hexyl
surface groups, which have weaker bonds than the aromatic
core and dendrons. These account for a third of the total mass
of the molecule. The remaining conjugated backbone of the
dendrimer begins to dissociate at about 500 uC and is fully
decomposed by 700 uC.
Organic semiconductors posses both an optical band gap
and an electronic band gap. Strictly speaking, the difference
between the two is equal to the exciton binding energy of the
2
0
molecule. Precise measurement of the exciton binding energy
Electrochemistry
has proven to be difficult, but values of y0.2 eV are typically
2
1
The electrochemical properties of the four dendrimers were
investigated by cyclic voltammetry. The oxidations were
reported for conjugated polymers. Here we estimate the
electronic band gaps of thin films of the dendrimers using
This journal is ß The Royal Society of Chemistry 2005
J. Mater. Chem., 2005, 15, 4518–4528 | 4523

View Article Online
Fig. 6 Cyclic voltammagrams of the dendrimers in solution. The shown potentials are vs. SCE. Oxidations were carried out in dry DCM,
n
21
4 6
reductions in dry THF, with 0.1 M [ Bu N][PF ] at 100 mV s under argon.
cyclic voltammetry. The cyclic voltammetry technique has the
advantage that both the oxidation and reduction waves can be
observed on the same voltammogram, and therefore a direct
observation of the electronic band gap energy is possible by
taking the difference of the two potentials multiplied by the
elementary charge. The resulting value can however only be
considered as an estimate of the true electronic band gap due
to subtleties of the electrochemical experiment such as swelling
of the films in the presence of solvent, the specific arrangement
of the electric double-layer at the interfacial region near the
film surface, or changes in polarisation in the material due to
the incorporation of ionic species into the film.
spectra listed in Table 1 and may be an indication of
the magnitude of electron–hole correlation effects in these
molecules.
Films were prepared by drop-casting
dendrimer dissolved in DCM onto
a
solution of
a
gold substrate.
Tetrabutylammonium hexafluorophosphate was used as a
supporting electrolyte in a 0.1 M acetonitrile solution. A
typical cyclic voltammogram can be seen in Fig. 7. The band
gaps were taken as the difference between the onsets of
oxidation and reduction potentials, with the onset being
defined as the crossing point of the tangent of the baseline
and the slope of the peak. Using this technique, 3G1 was
determined to have an electronic band gap of 2.85 eV, 4G1
Fig. 7 Example of a cyclic voltammogram obtained from a thin film
2
0
.83 eV, 3G2 2.70 eV and 4G2 2.74 eV. These are between
.21 and 0.42 eV larger than the onset of the optical absorption
of the dendrimer, shown for 4G1. Carried out in dry acetonitrile with
n
21
4 6
0.1 M [ Bu N][PF ] at 100 mV s under argon.
4
524 | J. Mater. Chem., 2005, 15, 4518–4528
This journal is ß The Royal Society of Chemistry 2005

View Article Online
Conclusions
(1110-8125 and 1110-8150). Chloroform was used as the
3
21
solvent for GPC with a flow rate of 1 cm min and poly-
styrene standards (Polymer Labs EasiCal PS-1) were used.
Emission spectra were recorded on a Jobin-Yvon Fluorolog 3
fluorescence spectrometer with a cooled CCD detector. All
spectra were corrected for the response of the detection system.
Fluorescence quantum yields were determined relative to a
perylene dye (BASF Lumogen yellow 083) in degassed
dichloromethane, which has a quantum yield close to 100%.
Fluorescence decays were recorded using time-correlated
single-photon counting, with a pulsed diode laser excitation
at 438 nm (Jobin Yvon IBH). Emission was monitored
perpendicular to excitation through a 0.25 m monochromator
and detected using a cooled photon counting photomultiplier
tube (Hamamatsu H6279), which gave a response function of
Dendrimers based on thiophene dendrons attached to a phenyl
core have been designed and synthesised using a convergent
synthesis. The dendrimers were soluble in organic solvents
such as toluene, chloroform and THF. The structures and
1
13
purity were confirmed by H and C NMR, MALDI-TOF-
MS and microanalysis. The dendrimers were prepared in high
yields, except 4G2 which suffered from incomplete reaction
even after prolonged reaction time at elevated temperatures.
The values obtained from MALDI-TOF mass spectrometry
were consistent with the proposed structures, while GPC gave
values which were consistently slightly higher than the
calculated values. This is normal for rigid aromatic molecules
when using polystyrene standards. TGA found initial decom-
position of approximately 35% of the dendrimers at 400 uC
accounting for loss of the hexyl surface groups followed by
decomposition of the aromatic core. UV-vis spectroscopy
showed an increasing red-shift of the onset to absorption and
hence decreasing optical band gaps from 3G1 to 4G1 and from
2
25 picoseconds. Data were analysed using a non-linear least
squares iterative reconvolution procedure using established
22
procedures. The finite width of the instrument response
function was effectively deconvoluted from the measured data
to give an overall temporal response of 20 picoseconds. Data
were fit to either a sum of one or two exponentials and the
3
G2 to 4G2 indicating extended conjugation through the
phenyl core for the four armed dendrimers. Additionally, a
red-shift was seen when increasing from the 1st to 2nd
generation dendrimer, although this was less marked for the
four armed dendrimer possibly due to steric constraints and
consequently twisting of the dendron. The smallest optical
band gap was observed for 4G2 and was found to be 2.34 eV
for a thin film of the dendrimer. A similar trend was
established for the maxima in the fluorescence spectra which
was independent of the excitation wavelength. The electro-
chemical properties were additionally studied by cyclic
voltammetry. Investigations in solution elucidated the redox
properties, while studies on thin films yielded the electronic
band gaps. These proved to be on average 0.3 eV greater than
the optical band gap, providing an indication of the electron–
hole correlation energy in these materials. We are currently
investigating the potential use of the prepared dendrimers as
the light-harvesting material in organic PV devices.
22
quality of fit judged using stringent statistical procedures.
TGA measurements were taken using a TA instruments TGA
21
2
950 thermogravimetric analyzer (5 uC min ) under nitrogen
and melting points were recorded on a TA instruments DSC
21
910 differential scanning calorimeter (10 uC min ) and were
2
taken as the onset to the melting point transition. MALDI
spectra were recorded using a micromass Q-TOF spectrometer
with an AP/MALDI ion source. Mass calibration was
performed using a ProteoMass peptide MALDI-MS calibra-
tion kit purchased from Sigma-Aldrich. All samples were
run using dihydroxybenzoic acid (DHB) as the matrix with
THF as the solvent in a dried-droplet preparation method.
Cyclic voltammetry experiments were performed under argon
using an electrochemical cell driven by a PINE instruments
company biopotentiostat model AFCBP1. A platinum work-
ing electrode, platinum coil counter electrode and a silver
wire quasi-reference electrode were utilised with a scan rate
2
1
23
of 100 mV s . Decamethylferrocene was used as the internal
2
reference for DCM solutions and ferrocene was used in THF
4
Experimental
2
5
solutions and the potentials corrected vs. SCE.
General
All reactions were performed under argon, using glassware
that was flame dried and then cooled under a stream of argon.
Tetrahydrofuran was dried over sodium and freshly distilled
prior to use. All other chemicals were purchased from Sigma-
Starting materials prepared according to literature methods
26
Tributyl(thiophen-2-yl)stannane,
5a
5,50-dihexyl-[2,29;39,20]-
terthiophene 7, tributyl-(5,50-dihexyl-[2,29;39,20]terthiophen-
5a
5
9-yl)stannane, 2,3-di(59,590-dihexyl-[29,20;30,290]terthiophen-
5a
1
13
Aldrich and were used as received. H and C NMR spectra
were recorded on a Varian unity INOVA AS400 spectrometer
and are reported in ppm relative to residual solvent. All UV-vis
spectra were taken as solutions in DCM or THF or as thin
films spin-coated onto quartz substrates using a Hewlett-
Packard 8453 spectrophotometer. Optical band gaps were
taken as the low energy onset to absorption and were
calculated as the intersection of two tangents, one along the
baseline and the other at the point of inflection of the initial
absorption peak. Elemental analysis was performed by
Huffman Laboratories, Golden, CO. GPC analysis utilised a
Beckman 110B solvent delivery module, SPEX fluorolog
50-yl)thiophene
8
and tributyl-{2,3-di(59,590-dihexyl-
5a
29,20;30,290]terthoiphen-50-yl)thiophen-5-yl}stannane.
[
1
,3,5-Tri(29-thienyl)benzene (2)
A mixture of tributyl(thiophen-2-yl)stannane (99 mmol), 1,3,5-
tribromobenzene 1 (6.9 g, 22 mmol) and dry N,N-dimethyl-
3
formamide (50 cm ) was degassed with argon for 10 minutes.
Bis(triphenylphosphine)palladium(II) dichloride (1.0 g) was
then added and the mixture heated at 90 uC for 65 hours. The
solvent was removed in vacuo, water (150 cm ) added and the
3
3
product extracted with ether (2 6 150 cm ). The organic
extracts were combined, washed with aqueous potassium
1
680 0.22 m double spectrometer and Polymer Labs columns
This journal is ß The Royal Society of Chemistry 2005
J. Mater. Chem., 2005, 15, 4518–4528 | 4525

View Article Online
3
fluoride (2 6 50 cm ), dried over anhydrous magnesium
sulfate, filtered and the solvent removed in vacuo. Purification
10 minutes. Tetrakis(triphenylphosphine)palladium(0) (30 mg,
0.03 mmol) was then added and the mixture heated at 90 uC
for 41 hours. After allowing the mixture to cool, ether (50 cm )
3
by column chromatography (ethyl acetate–hexanes; 1 : 9) gave
1
(3.3 g, 47%) as a pale yellow solid which had identical H
2
was added and the organic layer was washed with aqueous
3
potassium fluoride solution (20 cm ) followed by water
2
7
NMR to that reported in the literature.
3
20 cm ). The organic layer was dried over anhydrous
(
1
,3,5-Tri(29-bromo-59-thienyl)benzene (3)
magnesium sulfate, filtered and the solvent removed in vacuo.
The residue was purified by column chromatography (ethyl
acetate–hexanes; 1 : 49) to give 9 (0.65 g, 93%) as a yellow–
To a mixture of 1,3,5-tri(29-thienyl)benzene 2 (2.0 g, 6.2 mmol),
3
3
chloroform (75 cm ) and glacial acetic acid (75 cm ) at 0 uC
was added portionwise N-bromosuccinimide (3.4 g, 19 mmol).
The mixture was then stirred at room temperature for 17 hours.
brown solid, mp 69 uC. (Found: C, 68.61; H, 6.92 C90
requires C, 68.91; H, 6.55%); Cl )/nm 250
4.76), 325 (4.71) and 399 (4.93);
d (400 MHz; CDCl ) 0.88 (18 H, t, J 6.8), 1.23–1.41 (36 H,
102 12
H S
lmax (CH
2
2
3
21
21
(
loge/dm mol
cm
3
The mixture concentrated in vacuo, hexanes (100 cm ) added
H
3
and the mixture filtered. The solid was washed with water
3
m), 1.64 (6 H, q, J 7.2), 1.66 (6 H, q, J 7.2), 2.76 (6 H, t, J 7.6),
.78 (6 H, t, J 7.6), 6.65 (3 H, d, J 3.6), 6.66 (3 H, d, J 3.6), 6.89
3
(3 6 100 cm ) followed by acetone (100 cm ) to give an off-
3
white solid, 2.6 g. Dichloromethane (300 cm ) was added to the
2
(
3 H, d, J 3.6), 6.94 (3 H, d, J 3.6), 7.12 (3 H, d, J 3.6), 7.19
3
filtrate and the mixture was washed with water (2 6 200 cm )
followed by aqueous sodium bicarbonate (2 6 100 cm ). The
organic layer was then dried over anhydrous magnesium
sulfate, filtered and the solvent removed in vacuo. The solid
residue was recrystallised from dichloromethane–hexanes to
give 0.3 g of additional product. The two solids were combined
(
3 H, s), 7.28 (3 H, d, J 3.6) and 7.62 (3 H, s); d (100 MHz;
C
3
CDCl ) 14.3, 22.8, 29.0, 30.4, 31.7, 31.8, 121.9, 124.3, 124.5,
3
1
1
24.8, 124.9, 126.5, 126.7, 127.7, 131.0, 132.3, 132.6, 134.7,
35.1, 135.5, 136.9, 142.3, 146.6 and 147.8; m/z (MALDI-
+
TOF-ES+) 1567.57 (100%) [MH ] (Calcd. 1566.46).
1
to give 3 (2.9 g, 84%) as a white solid which had identical H
2
1
,3,5-Tri(29-{20,30-di(590,5900-dihexyl-[290,200;300,2900]terthio-
8
NMR to that reported in the literature.
,2,4,5-Tetra(29-thienyl)benzene (5)
phen-500-yl)thiophen-50-yl}thiophen-59-yl)benzene (10) (3G2)
1
The 3G2 dendrimer 10 was prepared in a similar manner to 9
using a mixture of tributyl-{2,3-di(59,590-dihexyl-[29,20;30,290]-
terthiophen-50-yl)thiophen-5-yl}stannane (720 mg, 0.6 mmol),
Tetrathiophene benzene 5 was prepared as described for
thiophene 2 from tributyl(thiophen-2-yl)stannane (250 mmol),
1
,3,5-tri(29-bromo-59-thienyl)benzene 3 (51 mg, 0.1 mmol), dry
1
,2,4,5-tetrabromobenzene 4 (14.1 g, 36 mmol), dry N,N-
3
3
N,N-dimethylformamide (1 cm ) and tetrakis(triphenylphos-
phine)palladium(0) (10 mg, 0.01 mmol). The residue was
purified by column chromatography (ethyl acetate–hexanes;
dimethylformamide (150 cm ) and bis(triphenylphosphine)-
palladium(II) dichloride (2.75 g). Purification by silica plug
(
(
dichloromethane) followed by column chromatography
1
: 19) to give 10 (260 mg, 96%) as a red highly viscous oil.
hexanes) gave 5 (5.35 g, 37%) as a pale yellow solid, mp
(
198
Found: C, 68.32; H, 6.82 C174H S24 requires C, 68.32; H,
2
17 uC. (Found: C, 64.99; H, 3.47 C22
H, 3.83%); lmax (CH Cl )/nm 256sh (loge/dm mol
.16), 287 (4.33) and 316sh (4.12); d (400 MHz; CDCl
.94 (8 H, m), 7.29 (4 H, dd, J 1.6, 5.2) and 7.65 (2 H,
(100 MHz; CDCl ) 126.6, 127.3, 127.7, 133.5, 133.6
and 141.8.
14 4
H S requires C, 65.25;
3
21
21
6
.52%); lmax (CH
Cl
2 2
)/nm 251 (loge/dm mol cm 5.06)
) 0.83–0.90 (36 H, m),
3
21
21
2
2
cm
and 370 (5.03); d
H
(400 MHz; CDCl
3
4
6
H
3
)
1
6
6
.23–1.38 (72 H, m), 1.58–1.68 (24 H, m), 2.70–2.78 (24 H, m),
.62–6.65 (12 H, m), 6.86 (6 H, d, J 3.6), 6.92 (3 H, d, J 3.6),
.93 (3 H, d, J 3.6), 7.16 (3 H, d, J 4.0), 7.19 (3 H, s), 7.20
s); d
C
3
(
3 H, s), 7.22 (3 H, s), 7.31 (3 H, d, J 4.0) and 7.65 (3 H, s);
d
C
(100 MHz; CDCl
3
) 14.4, 22.9, 29.1, 30.4, 31.8, 31.9, 121.7,
1
,2,4,5-Tetra(29-bromo-59-thienyl)benzene (6)
1
24.3, 124.4, 124.8, 125.3, 126.2, 126.5, 126.7, 127.7, 127.8,
Tetra(bromothiophene) benzene 6 was prepared in an identical
129.9, 130.2, 130.4, 131.8, 132.0, 132.1, 132.2, 132.3, 132.5,
132.7, 133.0, 134.7, 134.9, 135.0, 135.3, 136.2, 136.4, 142.7,
146.3, 146.5, 147.6 and 147.8; m/z (MALDI-TOF-ES+)
method to that of 3 from 1,2,4,5-tetra(29-thienyl)benzene 5
3
0.46 g, 1.1 mmol), chloroform (25 cm ), glacial acetic acid
(
(
(
3
25 cm ) and N-bromosuccinimide (0.87 g, 4.7 mmol) to give 6
+
3055.97 (100%) [MH ] (Calcd. 3054.88).
0.66 g, 81%) as a pale yellow solid, mp 223 uC. (Found: C,
3
6.59; H, 1.40 C H S Br requires C, 36.89; H, 1.61%); l
3
1,2,4,5-Tetra(29-{50,500-dihexyl-[20,290;390,200]terthiophen-590-
yl}thiophen-59-yl)benzene (11) (4G1)
2
2
10
4
4
max
21
21
(
CH Cl )/nm 254sh (loge/dm mol cm 4.34), 301 (4.53)
2 2
and 329 (4.35); d (400 MHz; CDCl ) 6.73 (4 H, d, J 4.0), 6.95
H 3
The 4G1 dendrimer 11 was prepared in a similar manner to
(
C 3
4 H, d, J 4.0) and 7.48 (2 H, s); d (100 MHz; CDCl ) 113.8,
9
using a mixture of tributyl-(5,50-dihexyl-[2,29;39,20]terthio-
phen-59-yl)stannane (1.69 g, 2.4 mmol), 1,2,4,5-tetra(29-
bromo-59-thienyl)benzene (289 mg, 0.4 mmol), dry
1
28.2, 130.4, 132.9, 133.4 and 142.4.
6
1
,3,5-Tri(29-{50,500-dihexyl-[20,290;390,200]terthiophen-590-
3
N,N-dimethylformamide (4 cm ) and tetrakis(triphenylphos-
phine)palladium(0) (30 mg, 0.03 mmol). The residue was
purified by column chromatography (ethyl acetate–hexanes;
1 : 99) followed by a second column (dichloromethane–
hexanes; 1 : 9) to give 11 (700 mg, 85%) as a orange/brown
yl}thiophen-59-yl)benzene (9) (3G1)
mixture of tributyl-(5,50-dihexyl-[2,29;39,20]terthiophen-
9-yl)stannane (1.7 g, 2.4 mmol), 1,3,5-tri(29-bromo-59-
thienyl)benzene (250 mg, 0.45 mmol) and dry
N,N-dimethylformamide (4 cm ) was degassed with argon for
A
5
3
3
solid, mp 62 uC. (Found: C, 68.44; H, 6.46 C111
134 16
H S
4
526 | J. Mater. Chem., 2005, 15, 4518–4528
This journal is ß The Royal Society of Chemistry 2005

View Article Online
6
(a) R. D. McCullough, Adv. Mater., 1998, 10, 2, 93–116; (b)
D. Fichou, Handbook of Oligo- and Polythiophenes, Wiley-VCH,
New York, 1998.
requires C, 68.62; H, 6.54%);
l
max (CH
2
Cl
4.79), 309 (4.75) and 396 (4.88);
d (400 MHz; CDCl ) 0.95 (24 H, t, J 6.6), 1.31–1.47 (48 H,
2
)/nm 247
3
21
21
(
loge/dm mol
cm
H
3
7 (a) D. Braun, G. Gustafsson, D. McBranch and A. J. Heeger,
J. Appl. Phys., 1992, 72, 2, 564–568; (b) F. Geiger, M. Stoldt,
H. Schweizer, P. Bauerle and E. Umbach, Adv. Mater., 1993, 5, 12,
m), 1.66–1.76 (16 H, m), 2.81 (8 H, t, J 7.6), 2.82 (8 H, t, J 7.6),
6
6
7
3
1
1
2
.70 (8 H, d, J 3.2), 6.93 (4 H, d, J 3.6), 6.94 (4 H, d, J 4.0),
.98 (4 H, d, J 3.2), 7.09 (4 H, d, J 3.6), 7.20 (4 H, s) and
.71 (2 H, s); d (100 MHz; CDCl ) 14.5, 23.0, 29.1, 30.5, 31.8,
9
22–925; (c) G. Horowitz, P. Delannoy, H. Bouchriha, F. Deloffre,
J. L. Fave, F. Garnier, R. Hajlaoui, M. Heyman, F. Kouki, P. Valat,
V. Wintgens and A. Yassar, Adv. Mater., 1994, 6, 10, 752–755; (d)
P. Dyreklev, M. Berggren, O. Inganas, M. R. Andersson,
O. Wennerstrom and T. Hjertberg, Adv. Mater., 1995, 7, 1, 43–45;
C
3
1.9, 124.3, 124.5, 126.6, 126.7, 127.7, 128.8, 130.9, 132.5,
32.6, 133.0, 133.7, 134.8, 135.1, 138.1, 140.6, 146.5 and
+
47.8; m/z (MALDI-TOF-ES+) 2064.22 (100%) [MH ] (Calcd.
(e) R. N. Marks, F. Biscarini, R. Zamboni and C. Taliani, Europhys.
Lett., 1995, 32, 6, 523–528; (f) P. Barta, W. R. Salaneck, M. Zagorska,
A. Pron and S. Niziol, Adv. Mater. Opt. Electron., 1996, 6, 5–6,
062.60).
4
06–408; (g) C. Vaterlein, H. Neureiter, W. Gebauer, B. Ziegler,
M. Sokolowski, P. Bauerle and E. Umbach, J. Appl. Phys., 1997, 82,
, 3003–3013; (h) M. Granstr o¨ m, Polym. Adv. Technol., 1997, 8, 7,
424–430; (i) U. Mitschke and P. B a¨ uerle, J. Chem. Soc., Perkin Trans.
, 2001, 740–753; (j) G. Gigli, M. Anni, M. Theander, R. Cingolani,
1,2,4,5-Tetra(29-{20,30-di(590,5900-dihexyl-[290,200;300,2900]terthio-
6
phen-500-yl)thiophen-50-yl}thiophen-59-yl)benzene (12) (4G2)
1
The 4G2 dendrimer 12 was prepared in a similar manner to
G. Barbarella, L. Favaretto and O. Ingan a¨ s, Synth. Met., 2001, 119,
581–582.
(a) F. Garnier, A. Yasser, R. Hajlaoui, G. Horowitz, F. Deloffre,
B. Servet, S. Ries and P. Alnot, J. Am. Chem. Soc., 1993, 115, 19,
3
G2 10 using a mixture of tributyl-{2,3-di(59,590-dihexyl-
8
[
29,20;30,290]terthiophen-50-yl)thiophen-5-yl}stannane (1.56 g,
1.3mmol), 1,2,4,5-tetra(29-bromo-59-thienyl)benzene6 (154 mg,
3
0.2 mmol), dry N,N-dimethylformamide (2 cm ) and tetra-
8
716–8721; (b) A. Dodabalapur, L. Torsi and H. E. Katz, Science,
1995, 268, 270–271; (c) A. J. Lovinger and L. J. Rothberg, J. Mater.
Res., 1996, 11, 6, 1581–1592; (d) Y. Wei, Y. Yang and J. M. Yeh,
Chem. Mater., 1996, 8, 11, 2659–2666; (e) A. R. Brown,
C. P. Jarrett, D. M. deLeeuw and M. Matters, Synth. Met.,
kis(triphenylphosphine)palladium(0) (25 mg, 0.02 mmol). The
residue was purified by repeated column chromatography
(
ethyl acetate–hexanes; 1 : 49) to give 12 (89 mg, 11%) as a
deep red highly viscous oil. (Found: C, 68.25; H,6.63
32 requires C, 68.16; H, 6.52%); lmax (CH Cl )/nm
55 (loge/dm mol cm 5.12), 308sh (5.05) and 360 (5.03);
d (400 MHz; CDCl ) 0.82–0.92 (48 H, m), 1.21–1.39 (96 H,
1
997, 88, 1, 37–55; (f) H. E. Katz, Z. N. Bao and S. L. Gilat, Acc.
Chem. Res., 2001, 34, 5, 359–369; (g) X. M. Hong, H. E. Katz,
A. J. Lovinger, B. C. Wang and K. Raghavachari, Chem. Mater.,
C
230
H
262
S
2
2
2
001, 13, 12, 4686–4691; (h) S. A. Ponomarenko, S. Kirchmeyer,
3
21
21
2
A. Elschner, B. -H. Huisman, A. Karbach and D. Drechsler, Adv.
Funct. Mater., 2003, 13, 8, 591–596.
(a) N. Noma, T. Tsuzuki and Y. Shirota, Adv. Mater., 1995, 7, 7,
H
3
9
m), 1.58–1.68 (32 H, m), 2.70–2.78 (32 H, m), 6.60–6.65 (16 H,
6
47–648; (b) M. Granstr o¨ m, K. Petritsch, A. C. Arias, A. Lux,
m), 6.84 (4 H, d, J 3.6), 6.85 (4 H, d, J 3.6), 6.92 (12 H, d, J
M. R. Andersson and R. H. Friend, Nature, 1998, 395, 257–260; (c)
L. Chen, L. S. Roman, D. M. Johansson, M. Svensson,
M. R. Andersson, R. A. J. Janssen and O. Ingan a¨ s, Adv. Mater.,
3
.6), 7.09 (4 H, d, J 3.6), 7.17 (4 H, s), 7.18 (4 H, s), 7.24 (4 H,
s) and 7.66 (2 H, s); d (100 MHz; CDCl ) 14.3, 22.8, 29.0, 30.4,
1.7, 31.8, 124.2, 124.4, 124.7, 126.4, 126.5, 126.6, 127.7, 127.8,
C
3
2
000, 12, 15, 1110–1114; (d) T. Fromherz, F. Padinger,
3
1
1
1
D. Gebeyehu, C. Brabec, J. C. Hummelen and N. S. Sariciftci,
Sol. Energy Mater. Sol. Cells, 2000, 63, 1, 61–68; (e) J. J. Dittmer,
E. A. Marseglia and R. H. Friend, Adv. Mater., 2000, 12, 17,
1270–1274; (f) W. Feng, A. Fujii, S. Lee, H. Wu and K. Yoshino,
J. Appl. Phys., 2000, 88, 12, 7120–7123; (g) S. E. Shaheen,
D. Vangeneugden, R. Kiebooms, D. Vanderzande, T. Fromherz,
F. Padinger, C. J. Brabec and N. S. Sariciftci, Synth. Met., 2001,
28.9, 130.0, 130.3, 130.4, 131.9, 132.0, 132.1, 132.2, 132.3,
32.4, 132.6, 132.9, 133.1, 133.7, 134.7, 134.8, 135.0, 136.1,
37.7, 140.9, 146.3, 146.5, 147.6 and 147.9; m/z (MALDI-
+
TOF-ES+) 4048.56 (100%) [MH ] (Calcd. 4047.16).
1
21, 1–3, 1583–1584; (h) C. J. Brabec, N. S. Sariciftci and
Acknowledgements
J. C. Hummelen, Adv. Funct. Mater., 2001, 11, 1, 15–26; (i)
L. Tan, M. D. Curtis and A. H. Francis, Macromolecules, 2002, 35,
1
Adv. Funct. Mater., 2003, 13, 1, 85–88; (k) D. Chirvase,
Z. Chiguvare, M. Knipper, J. Parisi, V. Dyakonov and
J. C. Hummelen, Synth. Met., 2003, 138, 299–304.
The support of this work by the NREL DDRD program and
the Department of Energy is gratefully acknowledged. We also
thank Bill McMahon for help in the MALDI-TOF-MS
measurements.
2, 4628–4635; (j) F. Padinger, R. S. Rittberger and S. Sariciftci,
1
1
0 (a) J. Roncali, Chem. Rev., 1997, 97, 173–205; (b) H. Saadeh,
T. Goodson and L. P. Yu, Macromolecules, 1997, 30, 16,
4608–4612; (c) W. Huang, H. Meng, W. L. Yu, J. Gao and
A. J. Heeger, Adv. Mater., 1998, 10, 8, 593–596; (d) A. Greiner,
Polym. Adv. Technol., 1998, 9, 7, 371–389; (e) U. Mitschke and
P. B a¨ uerle, J. Mater. Chem., 2000, 10, 1471–1507.
1 (a) R. D. McCullough and R. D. Lowe, J. Chem. Soc., Chem.
Commun., 1992, 70–72; (b) R. D. McCullough, R. D. Lowe,
M. Jayaraman and D. L. Anderson, J. Org. Chem., 1993, 58,
904–912; (c) T.-A. Chen, X. Wu and R. D. Rieke, J. Am. Chem.
Soc., 1995, 117, 233–244.
References
1
(a) O. A. Matthews, A. N. Shipway and J. F. Stoddart, Prog.
Polym. Sci., 1998, 23, 1, 1–56; (b) A. Adronov and J. M. J. Fr e´ chet,
Chem. Commun., 2000, 1701–1710.
2
(a) C. Devadoss, P. Bharathi and J. S. Moore, J. Am. Chem. Soc.,
1996, 118, 9635–9644; (b) Q.-S. Hu, V. Pugh, M. Sabat and L. Pu,
J. Org. Chem., 1999, 64, 7528–7536; (c) Z. Peng, Y. Pan, B. Xu and
J. Zhang, J. Am. Chem. Soc., 2000, 122, 6619–6623.
A. J. Berresheim, M. M u¨ ller and K. M u¨ llen, Chem. Rev., 1999, 99,
3
4
12 (a) J. Nakayama and J.-S. Lin, Tetrahedron Lett., 1997, 38, 34,
6043–6046; (b) R.-M. Sebastian, A.-M. Caminade, J.-P. Majoral,
E. Levillian, L. Huchet and J. Roncali, Chem. Commun., 2000,
507–508; (c) J. Alvarez, L. Sun and R. M. Crooks, Chem. Mater.,
2002, 14, 3995–4001; (d) M.-H. Xu and L. Pu, Tetrahedron Lett.,
2002, 43, 6347–6350; (e) M. J. Frampton, E. B. Namdas, S.-C. Lo,
P. L. Burn and I. D. W. Samuel, J. Mater. Chem., 2004, 14,
2881–2888; (f) J. P. J. Markham, E. B. Namdas, T. D. Anthopoulos,
I. D. W. Samuel, G. J. Richards and P. L. Burn, Appl. Phys. Lett.,
2004, 85, 9, 1463–1465.
1747–1785.
(a) M. Halim, J. N. G. Pillow, I. D. W. Samuel and P. L. Burn,
Synth. Met., 1999, 102, 922–923; (b) J. N. G. Pillow, M. Halim,
J. M. Lupton, P. L. Burn and I. D. W. Samuel, Macromolecules,
1
999, 32, 19, 5985–5993.
(a) C. Xia, X. Fan, J. Locklin and R. C. Advincula, Org. Lett.,
002, 4, 12, 2067–2070; (b) C. Xia, X. Fan, J. Locklin,
R. C. Advincula, A. Gies and W. Nonidez, J. Am. Chem. Soc.,
004, 126, 8735–8743.
5
2
2
This journal is ß The Royal Society of Chemistry 2005
J. Mater. Chem., 2005, 15, 4518–4528 | 4527

View Article Online
1
1
1
1
1
1
1
2
3 C. V. Pham, A. Burkhardt, R. Shabana, D. D. Cunningham,
H. B. Mark and H. Zimmer, Phosphorus Sulfur, 1989, 46, 153–168.
4 J. B. Bicks, Photophysics of Aromatic Molecules, John Wiley &
Sons, Chichester, 1970, pp. 113–118.
5 A. Breeze, D. Ginley and B. Gregg, Proc. NCPV Solar Program
Review Meeting 2003, 2003, 690–692.
6 Y. Li, G. Vamvounis and S. Holdcroft, Macromolecules, 2002, 35,
21 I. H. Campbell, T. H. Hagler, D. L. Smith and J. P. Ferraris, Phys.
Rev. Lett., 1996, 76, 11, 1900–1903.
22 D. Phillips and D. V. O’Connor, Time-Correlated Single-Photon
Counting, Academic Press, London, 1984.
23 T. Gennett, D. F. Milner and M. J. Weaver, J. Phys. Chem., 1985,
89, 13, 2787–2794.
24 N. G. Connelly and W. E. Geiger, Chem. Rev., 1996, 96, 877–810.
25 A. J. Bard and L. R. Faulkner, Electrochemical Methods: Funda-
mentals and Applications, John Wiley & Sons, Chichester, 1980.
26 J. T. Pinhey and E. G. Roche, J. Chem. Soc., Perkin Trans. 1, 1988,
8, 2415–2421.
27 A. Pelter, I. Jenkins and D. E. Jones, Tetrahedron, 1997, 53, 30,
10357–10400.
28 F. Cherioux and L. Guyard, Adv. Funct. Mater., 2001, 11, 4,
305–309.
6900–6906.
7 M. Theander, O. Ingan a¨ s, W. Mammo, T. Olinga, M. Svensson
and M. R. Andersson, J. Phys. Chem. B, 1999, 103, 37, 7771–7780.
8 S. Hoger, J. Spickermann, D. L. Morrison, P. Dziezok and
H. J. Rader, Macromolecules, 1997, 30, 3110–3111.
9 L. Jinsong, R. S. Loewe and R. D. McCullough, Macromolecules,
1999, 32, 5777–5785.
0 B. A. Gregg, MRS Bull., 2005, 30, 1, 20–22.
4
528 | J. Mater. Chem., 2005, 15, 4518–4528
This journal is ß The Royal Society of Chemistry 2005