Effect of molecular weight on electronic, electrochemical and spectroelectrochemical properties of poly(3,3″-dioctyl-2,2′5′, 2″-terthiophene)

Article abstract of DOI:10.1039/b605504d

Poly(3,3″-dioctyl-2,2′5′,2″-terthiophene), obtained from its corresponding monomer by oxidative polymerization with FeCl ₃, has been fractionated into five fractions of reduced polydispersity, covering the M_n range from 1.50 kDa to 10.50 kDa (measured vs. polystyrene standards). The effect of M_n on spectroscopic, electrochemical, spectroelectrochemical and electrical transport properties has been investigated. Fractions of growing M_n show an increasing bathochromic shift of the band originating from the π-π* transition in the neutral polymer with the appearance of a clear vibrational structure for the two highest molecular fractions. The onset of oxidative doping determined from the cyclic voltammogram shifts towards lower potentials with increasing molecular weight. A similar trend is observed for doping induced near infrared bands, which shift towards lower energies (higher wavelengths) with increasing molecular weight and appear at lower potentials in spectroelectrochemical experiments. Finally, a comparison of the FET mobility in two transistors fabricated under identical conditions from polymer fractions differing in their molecular weight shows that a ca. fourfold increase of M _n (from 2.40 kDa to 10.50 kDa) results in a two orders of magnitude increase in the carriers' mobility (from μ_sat = 4 × 10 ^-5 cm² V^-1 s^-1 to μ_sat = 2 × 10^-3 cm² V^-1 s^-1). The obtained results underline the importance of the control of the macromolecular parameters in the preparation of electronic and electrochemical devices from poly(3,3″-dioctyl-2,2′5′,2″-terthiophene). The Royal Society of Chemistry 2006.

Full text of DOI:10.1039/b605504d

View Article Online / Journal Homepage / Table of Contents for this issue
PAPER
www.rsc.org/materials | Journal of Materials Chemistry
Effect of molecular weight on electronic, electrochemical and
spectroelectrochemical properties of poly(3,30-dioctyl-2,29:59,20-
terthiophene){
a
Rafal Pokrop, Jean-Marie Verilhac, Anna Gasior, Ireneusz Wielgus, Malgorzata Zagorska,*
b
a
a
a
c
Jean-Pierre Travers and Adam Pron
c
Received 18th April 2006, Accepted 25th May 2006
First published as an Advance Article on the web 14th June 2006
DOI: 10.1039/b605504d
Poly(3,30-dioctyl-2,29:59,20-terthiophene), obtained from its corresponding monomer by oxidative
polymerization with FeCl , has been fractionated into five fractions of reduced polydispersity,
3
covering the M
effect of M on spectroscopic, electrochemical, spectroelectrochemical and electrical transport
properties has been investigated. Fractions of growing M show an increasing bathochromic shift
n
range from 1.50 kDa to 10.50 kDa (measured vs. polystyrene standards). The
n
n
of the band originating from the p–p* transition in the neutral polymer with the appearance of a
clear vibrational structure for the two highest molecular fractions. The onset of oxidative doping
determined from the cyclic voltammogram shifts towards lower potentials with increasing
molecular weight. A similar trend is observed for doping induced near infrared bands, which shift
towards lower energies (higher wavelengths) with increasing molecular weight and appear at lower
potentials in spectroelectrochemical experiments. Finally, a comparison of the FET mobility in
two transistors fabricated under identical conditions from polymer fractions differing in their
molecular weight shows that a ca. fourfold increase of M
n
(from 2.40 kDa to 10.50 kDa) results in
2
5
2
21 21
s
a two orders of magnitude increase in the carriers’ mobility (from msat = 4 6 10 cm V
to
2
3
2
21 21
m
sat = 2 6 10 cm V
s
). The obtained results underline the importance of the control of the
macromolecular parameters in the preparation of electronic and electrochemical devices from
poly(3,30-dioctyl-2,29:59,20-terthiophene).
Introduction
very well suited for testing new molecular design concepts in
the fabrication of modified electrodes, electrochemical sensors
3,4
Attractive properties of poly(alkylthiophene)s have stimulated,
in the past two decades, extensive, multidisciplinary research
aimed at the elucidation of their electrical transport, electro-
chemical and spectroscopic properties as well as their applica-
tion as active components in various types of organic
or organic electronic devices.
The effect of the type of the regioregularity on various
properties of poly(alkylthiophene)s was first pointed out by
5
6
Souto Maior et al. and Zagorska and Krische, who demon-
strated that structural, spectroscopic and electrochemical
features of head to head–tail to tail (HH–TT) coupled
poly(3,39-dialkyl-2,29-bithiophene)s and poly(4,49-dialkyl-
1
,2
electronic devices.
A technologically interesting property of this family of
electroactive polymers is their solution processability, which is
rare in polyconjugated macromolecular systems. Moreover,
several structural features of these polymers such as chain
planarity, extent of p-stacking, type of supramolecular
aggregation etc. can controllably be tuned by changing the
type of macromolecule regioregularity. Since all these struc-
tural parameters govern the electronic and electrochemical
properties of polymer solid films, poly(alkylthiophene)s are
2
,29-bithiophene)s are significantly different from those of
their predominantly head to tail–head to tail (HT–HT)
coupled regioisomers. These findings were then additionally
corroborated by comparison with ‘‘purely’’ regioregular
7
poly(3-alkylthiophene)s
developed
by
Rieke
and
8
McCullough. Regioregular HT–HT poly(3-alkylthiophene)s
were found to be promising candidates for the fabrication of
9–14
field effect transistors
and, in a bulk heterojunction
configuration with fullerene derivatives, for the preparation
1
of photovoltaic cells. To the contrary, they show one order
5
a
Faculty of Chemistry, Warsaw University of Technology,
Noakowskiego 3, 00-664, Warsaw, Poland.
E-mail: zagorska@ch.pw.edu.pl
CEA Liten-DTNM-LCH, 17 rue des Martyrs, 38054, Grenoble cedex 9,
of magnitude lower electroluminescence quantum efficiency as
1
6
compared to their HH–TT coupled regioisomers. These
examples merely indicate the importance of molecular
design in the development of new organic electronic or
electrochemical devices.
b
France
CEA DRFMC, UMR 5819-SPrAM (CEA-CNRS-Univ. J.
c
Fourier-Grenoble I), 17 rue des Martyrs, 38054, Grenoble cedex 9,
France. E-mail: adam.pron@cea.fr
Basic electronic, spectroscopic and electrochemical proper-
ties are not only dependent on the type and degree of polymer
regioregularity but also on its molecular weight. The effect of
{
Electronic supplementary information (ESI) available: A diagram of
the transistor structure, synthetic details and product characterization.
See DOI: 10.1039/b605504d
This journal is ß The Royal Society of Chemistry 2006
J. Mater. Chem., 2006, 16, 3099–3106 | 3099

View Article Online
molecular weight on spectroscopic, electrochemical and
spectroelectrochemical properties of HT–HT coupled poly(3-
For cyclic voltammetry investigations, a thin layer of the
polymer was deposited on a platinum disc or plate electrode
by casting from chloroform solution. The experiments were
carried out in a one compartment electrochemical cell, in 0.1 M
1
7
alkylthiophene)s was first reported by Trznadel et al., who
showed that this dependence persisted to relatively high M
n
values. The authors rationalized the observed behavior by
identifying the types of propagation errors and chain termina-
tion mechanisms which led to lower chain regioregularity in
lower molecular weight fractions. More recent papers have
Bu
4
NBF
4
in acetonitrile solution, using a Pt counter-electrode,
in acetonitrile as a reference electrode.
and Ag/0.1 M AgNO
3
In the case of UV-vis-NIR spectroelectrochemical studies, a
thin polymer film was deposited on an ITO transparent
working electrode, the electrolyte and the counter and
reference electrodes remaining the same.
shown that M
poly(3-alkylthiophene)s and their crystallinity
consequence charge carrier mobilities in HT–HT poly(alkyl-
n
influences the supramolecular organization in
1
8,19
and by
1
4,20
thiophene) based organic field effect transistors
2
photovoltaic cells.
and
Fabrication of transistors and electrical measurements
1
The active semiconducting layer was deposited by spin-coating
on a specially prepared substrate, which together with the
The use of dialkylterthiophenes as macromonomers in the
preparation of poly(alkylthiophene)s constitutes an interesting
route to the preparation of processible, electroactive polymers
with tunable properties since it increases the number of
possible regioisomers of different chain microstructure.
Among them poly(3,30-dioctyl-2,29:59,20-terthiophene) has
recently been tested in the fabrication of organic field effect
poly(3,30-dioctyl-2,29:59,20-terthiophene) layer formed
bottom gate, bottom contact field-effect transistor (FET).
Thus, it consisted of a 320 nm thick SiO layer grown on a Nb
a
2
gate. Interdigitated (L = 20 mm, W = 9 mm) source and drain
electrodes, consisting of an adhesion layer of Cr (15 nm)
covered with an Au (70 nm) layer, were prepared by photo-
3
,4
transistors with promising results.
This polymer can
2
lithography. In the next step, the surface of SiO , not covered
formally be considered as an alternate copolymer of tail to
tail (TT) coupled dialkylbithiophene and thiophene. Due to a
different chain microstructure its optical and electrical trans-
port properties are distinctly different from those reported for
with the electrodes, was modified by exposure to hexamethyl-
disilazane (HMDS). Finally, the polymer layer was deposited
from CHCl3 solution. The fabricated device was pumped
under dynamic vacuum with the goal to fully remove even
minute amounts of the residual solvent and adsorbed oxygen
from its polymer layer. The exact scheme of the transistors
tested can be found in the Electronic Supplementary
Information (ESI{).
4
HT–HT and HH–TT coupled poly(alkylthiophene)s. In this
paper, we explore in detail: electrical transport, redox, spectro-
scopic and spectroelectrochemical properties of poly(3,30-
dioctyl-2,29:59,20-terthiophene) as a function of the polymer
molecular weight. Such studies, to the best of our knowledge
never reported to date, are important not only from the basic
research point of view but should also constitute the starting
point for the design of better poly(alkylthiophene) based
electronic devices, modified electrodes and electrochromic
devices.
All electrical measurements of the fabricated FETs were
2
performed in an inert (N ) atmosphere using a HP4156B
Semiconductor Parameter Analyzer. The mobility was
extracted from the saturation regime (Vds = 260 V) using
22
the standard equation:
2
) where Ids is the source–drain
I
ds = (WCox
m
sat/2L)(V
g
2 V
t
current, W the channel width, L the channel length, Cox the
dielectric capacity, msat the saturation field effect mobility, V
and V the gate voltage and the threshold voltage, respectively.
Experimental
g
t
Characterization techniques
Both the monomer and the resulting polymer were charac-
1
Synthesis
13
terized by H NMR, C NMR, FTIR and elemental analysis.
NMR spectra were recorded on a Varian Mercury 400 MHz
The applied synthetic pathway is depicted in Scheme 1. First,
the monomer (3) is obtained from 1 and 2 by Suzuki coupling,
it is then oxidatively polymerized to give poly(3,30-dioctyl-
2,29:59,20-terthiophene) (4). Full synthetic details and product
characterization can be found in the ESI.{
spectrometer with chloroform-d (CDCl
3
) solvent. FTIR
spectra were obtained on a FTIR Bio-RAD FTS 165 spectro-
21
21
meter (wavenumber range: 400–4000 cm , resolution: 4 cm
)
either on free standing films cast from chloroform or using the
KBr pellet technique. UV-vis-NIR spectra of thin solid films
or solutions in chloroform were recorded on a Cary 5000
Fractionation of poly(3,30-dioctyl-2,29:59,20-terthiophene) (4)
(
Varian).
As prepared, crude polymer exhibits a rather high poly-
dispersity index. In order to obtain sharper fractions we
modified the fractionation procedure used for HT–HT poly(3-
Molecular weights of the fractionated polymer were
measured using size exclusion chromatography (SEC), on a
LabAlliance chromatograph equipped with a Jordi Gel
DVB Mixed Bed column, using refractometric detection by
1
7
alkylthiophene)s. This method consists of sequential extrac-
tion of fractions with increasing molecular weight using an
experimentally established set of solvents. It turned out to be
very effective not only for poly(alkylthiophene)s but also for
copolymers of alkylthiophenes with other conjugated mole-
RI-2000-F refractometer, CHCl was used as the eluent. The
3
column temperature and the flow rate were fixed to 35 uC
2
1
and 1 mL min , respectively. The column was calibrated
using polystyrene standards provided by the Polymer
Standards Service.
2
3
cules. The polymer was fractionated in a Soxhlet apparatus,
by consecutive extractions using the following sequence of
3
100 | J. Mater. Chem., 2006, 16, 3099–3106
This journal is ß The Royal Society of Chemistry 2006

View Article Online
Scheme 1
solvents: acetone, hexane, dichloromethane, chloroform. Two
fractions were taken with dichloromethane, at the beginning
and at the end of the extraction, with the goal of increasing
the number of fractions with reduced polydispersity.
Macromolecular parameters of the obtained fractions are
listed in Table 1.
chromatography (SEC) (see Table 1). The ‘‘raw’’ SEC
results of polythiophene derivatives should however be
treated with caution, especially if polystyrene sharp fractions
are used as standards. Because of the significantly different
shape of the macromolecules, the exclusion mechanisms are
different in the flexible s-bonded main chain polymers and
the conjugated ones. As a consequence, the SEC M
n
data of
polythiophene derivatives are underestimated for low mole-
Results and discussion
2
3
cular weights (Mn , 2 kD) and overestimated for higher
2
ones (Mn > 2 kD). Assuming that the relationship between
5
Although tested only very recently, for application to the
3
fabrication of organic field effect transistors (FETs),
the M values determined by mass spectrometry (MALDI) and
n
poly(3,30-dialkylterthiophene)s have been known for more
SEC is the same as for HT–HT poly(alkylthiophene)s and for
2
4
than a decade. This polymer offers a very attractive chain
microstructure, at least concerning the extended conjugation,
since two TT-coupled alkylthiophene units are always
separated by an unsubstituted thiophene ring, which makes
the spacing between the alkyl groups larger. All these
structural features, combined with chain regioregularity,
promote the formation of ordered supramolecular aggregation
poly(3,30-dioctyl-2,29:59,20-terthiophene) we have calculated
the corrected M . The results are included in Table 1.
n
Conjugation length is the principal factor determining
electronic and electrochemical properties of organic con-
ductors and semiconductors. Better conjugation implies wider
bands and a narrower band gap in the resulting semiconduct-
ing polymer. The latter should be reflected in the UV-vis
spectra of thin solid layers of the neutral (undoped) polymer. It
is convenient to establish the concept of an effective conjuga-
tion length as probed by UV-vis spectroscopy. In the visible
part of the spectrum, oligo- and polythiophene derivatives
show a strong absorption band originating from the p–p*
electronic transition. For monodispersed regioregular oligo-
alkylthiophenes this absorption band is being red shifted, with
increasing degree of polymerization (DP), up to a certain
DP value above which no further shift is observed. This
corresponds to the effective conjugation length. Further
increase of the DP does not result in an increase of the
effective conjugation length. As shown in ref. 26 for
monodispersed oligoalkylthiophenes the limiting effective
conjugation length corresponds to a DP of ca. 20.
4
,24
in the solid state, clearly seen in X-ray diffraction studies.
As already mentioned, the effect of molecular weight on
electronic, electrochemical and spectroscopic properties in
regiochemically well defined poly(3-alkylthiophene)s persists
1
4,17,18,20,21
to relatively high molecular weights.
We were
tempted to verify whether a similar phenomenon can be
observed for poly(3,30-dioctyl-2,29:59,20-terthiophene). The
determination of such a relationship is of crucial importance
for any application of this polymer in organic electronics,
electrochemical sensors or electrochromic devices. Poly(3,30-
dioctyl-2,29:59,20-terthiophene) can be fractionated in a similar
manner as regioregular HT–HT poly(3-alkylthiophene)s,
giving fractions of distinctly different M
n
and reduced
polydispersity coefficient as demonstrated by size exclusion
Table 1 Macromolecular parameters of poly(3,30-dioctyl-2,29:59,20-terthiophene) fractions studied in this research
Correction
factor
Content of the
fraction /wt%
a
b
n
c
Solvent
M
n
(SEC)/kDa
M (SEC)/kDa
w
PI
M
n
corrected
DP
corrected
Acetone
Hexane
Dichloromethane I
Dichloromethane II
Chloroform
a
1.54
2.40
5.50
6.62
1.78
2.96
8.45
11.74
17.75
1.16
1.23
1.53
1.77
1.69
b
0.97
1.3
1.5
1.55
1.77
1.59
1.84
3.67
4.27
5.93
10
12
24
27
38
15.6
24.4
23.3
4.4
10.50
17.9
c
Calculated on the basis of the results presented in refs. 23, 25. Degree of polymerization normalized per thiophene ring. Insoluble part –
4.4 wt%.
1
This journal is ß The Royal Society of Chemistry 2006
J. Mater. Chem., 2006, 16, 3099–3106 | 3101

View Article Online
Table 2 Absorption maxima (nm) determined for the bands corre-
sponding to the p–p* transition in electronic spectra of fractionated
poly(3,30-dioctyl-2,29:59,20-terthiophene) (PDOTT) recorded for
a
3
CHCl solutions and for thin solid films
Solid state
CHCl
solution 0–0
3
Fraction
0–1
0–2
0–3
PDOTT/acetone
PDOTT/hexane
441
449
460
463
484
516
PDOTT/CH
PDOTT/CH
2
Cl
Cl
2
2
(I)
2
(II) 463
466
594
598
547
551
507
511
474
476
PDOTT/CHCl
a
3
2
2
The maxima of 2d A/dl are given in italics.
Polydispersed samples behave differently. For regioregular
HT–HT poly(3-hexylthiophene) the dependence of the effec-
n
tive conjugation length persists for DP values exceeding 50, as
shown by the comparison of the UV-vis spectra of fractionated
2
0
samples. Evidently, this difference of behaviour originates
from the polydispersity phenomenon and reflects different
distributions of the effective conjugation length for samples
Fig. 1 UV-vis spectra of thin solid layers of poly(3,30-dioctyl-
2,29:59,20-terthiophene) deposited on quartz plate.
n
with increasing DP . A similar phenomenon is observed
for poly(3,30-dioctyl-2,29:59,20-terthiophene) (see Table 2 and
As expected, the energy of this transition is the smallest for
the fraction DPn = 38. It should also be noted that, for
comparable DPn values, the spectra of poly(3,30-dioctyl-
2,29:59,20-terthiophene) are very similar to those of regio-
regular HT–HT coupled poly(3-hexylthiophene) with the
band of the 0–0 transition only slightly hypsochromically
Fig. 1). Spectral changes are observed for all DP studied, for
n
the two highest molecular weight fractions being small but
measurable. The spectra of these fractions also show a clear
vibrational structure. The vibrational structure of the electro-
nic spectrum, frequently observed for regioregular conjugated
polymers, has its origin in electron-lattice coupling. In
particular, the position of the band originating from the
transition from the ground state to the relaxed excited state
1
7
shifted (by ,10 nm).
These conclusions concerning the effective conjugation
length in fractionated poly(3,30-dioctyl-2,29:59,20-terthiophene)
are also confirmed by FTIR spectroscopy (see Fig. 2).
(
0–0 transition) is indicative of effective conjugation.
2
1
Fig. 2 a) FTIR spectra of different fractions of poly(3,30-dioctyl-2,29:59,20-terthiophene) in the range of 1400–1550 cm ; b) Ianti/Isym band ratio
plotted against DP for different fractions of poly(3,30-dioctyl-2,29:59,20-terthiophene).
n
3
102 | J. Mater. Chem., 2006, 16, 3099–3106
This journal is ß The Royal Society of Chemistry 2006

View Article Online
Fractions with increasing DP
n
significantly differ in the
27
compared to that measured for the lower molecular weight
21 21
2
spectral range 1400–1550 cm . According to Furukawa
1
25
2
hexane fraction (DP = 12, m = 4 6 10 cm V
s
). The
n
sat
2
1
the ratio of the band at 1497–1502 cm (CLC antisymmetric
obtained values slightly exceed those obtained for HT–HT
19,20
2
1
stretching) to that at 1463–1465 cm
stretching) is diagnostic of the conjugation length. The plot
of Ianti/Isym against DP , shown in Fig. 2, clearly indicates the
increasing average effective conjugation length with increasing
DP
. One must add here that the band ascribed to the CLC
symmetric stretching deformations embraces the contribution
from the CH bending deformations of the alkyl substituents.
Since, to a first approximation, the oscillator strength of the
latter is not dependent on DP and, for this reason remains
(CLC symmetric
poly(3-hexylthiophene) of similar DPn value.
Thus, the
molecular weight dependence of the charge carriers’ mobility is
not restricted to HT–HT-poly(alkylthiophene)s but seems to
be a more common phenomenon in the polythiophene family
of organic semiconductors. Moreover, this mobility strongly
n
n
increases even for polymer fractions of high DP , where the
n
2
observed increase of the effective conjugation length is very
small. Two factors seem to determine this behaviour: i)
evidently, independently of the effective conjugation length,
the carriers’ transport along an individual chain is easier than
their transport via interchain hopping, leading to increased m
n
constant for all fractions, the selected ratio properly describes
changes in the effective conjugation length.
2
9
As it has already been stated, the changes in the average
effective conjugation length persist in poly(alkylthiophene)s
for fractions of higher DP ; ii) as pointed out by Sirringhaus
n
the increasing chain length should also improve the interchain
hopping of carriers since the probability of this hopping
depends on the density of low activation energy pathways for
crossing between chains whose number should increase with
the chain length.
n n
to high DP values. Moreover DP has been found to be the
critical factor determining the charge carrier mobility in
polymer based field effect transistors (FETs). In fact, the
measured FET mobility in HT–HT poly(3-hexylthiophene)
steadily increases with DP
n
even for very high molecular
The above results seem to suggest that electrochemical
properties of poly(3,30-dioctyl-2,29:59,20-terthiophene) should
vary for different fractions since the effective conjugation
weights where the changes in the average effective conjugation
length are very small. This strong dependence was previously
1
verified up to DP = 85 and more recently up to DP = 163.
9
28
n
n
length and the charge mobility – critical parameters
Inspired by the findings described above, we have decided to
verify whether the same effect can be observed for poly(3,30-
dioctyl-2,29:59,20-terthiophene). We have fabricated two field
effect transistors (FETs) under identical technological condi-
tions (see Experimental and ESI{ for details) in which the
active semiconductor layer consisted of the hexane and
the chloroform fractions of the polymer, respectively. The
characteristics of these transistors obtained in the saturation
regime (Vds = 260 V) are presented in Fig. 3.
which determine electrochemical oxidative doping in thin solid
films – are strongly molecular weight dependent.
In general, in polythiophenes the electronic spectra of their
neutral (undoped) form which reflect their effective conjuga-
tion lengths, correlate well with the onset of the polymer
oxidative doping process (p-doping). In particular, in poly-
(alkylthiophene)s of lower conjugation, whose p–p* bands are
hypsochromically shifted, the oxidative electrochemical doping
6
(p-doping) starts at higher electrode potentials. With this in
The FET mobility, extracted from the above data, is almost
two orders of magnitude higher for the higher molecular weight
mind one should expect a lowering of the oxidative doping
potential for fractions with increasing DP . Cyclic voltamme-
n
2
3
2
21 21
s
chloroform fraction (DP = 38, m = 2 610 cm V
)
try is very well suited for this type of investigation since the
n
sat
Fig. 3 Transfer characteristics (Vds = 260 V) of field effect transistors fabricated from two different fractions of poly(3,30-dioctyl-
,29:59,20-terthiophene): a) hexane fraction, b) chloroform fraction.
2
This journal is ß The Royal Society of Chemistry 2006
J. Mater. Chem., 2006, 16, 3099–3106 | 3103

View Article Online
Fig. 4 Cyclic voltammograms obtained for different poly(3,30-dioctyl-2,29:59,20-terthiophene) fractions, in 0.1 M Bu
4
NBF –acetonitrile solution
4
2
at a scan rate of 50 mV s : a) acetone fraction; b) hexane fraction; c) dichloromethane (I) fraction; d) chloroform fraction.
1
polymer oxidative doping gives rise to one (or more) anodic
peaks, whereas consecutive reduction of the doped polymer
results in the appearance of its (their) cathodic counterparts. In
Fig. 4 cyclic voltammograms registered for four fractions of
poly(3,30-dioctyl-2,29:59,20-terthiophene), distinctly differing in
their molecular weight, are compared. First, it should be noted
p-doping anodic peak is relatively narrow and a significant
potential difference is observed between the positions of the
anodic doping and the cathodic dedoping peaks. These
features are characteristic of the voltammetric behavior of
6
poorly conjugated poly(alkylthiophene)s. With increasing M
n
,
the shape of the voltammogram starts to resemble that of
better conjugated derivatives of polythiophenes. For example,
the CV curve of the chloroform fraction of poly(3,30-dioctyl-
2,29:59,20-terthiophene) is similar to that of the corresponding
fraction of HT–HT coupled regioregular poly(3-hexylthio-
that with increasing M the onset of the p-doping anodic peak
n
n
shifts to lower potentials with increasing DP (see Fig. 5).
Second, significant changes in the shape of both anodic and
cathodic peaks are observed with increasing molecular weight.
For low molecular weight fractions (acetone, hexane), the
1
7
phene). Two broad, strongly overlapping redox couples of
higher electrochemical reversibility, superimposed on a pro-
nounced capacitive current background, can be distinguished.
For this fraction we have extended the cyclic voltammetry
investigations to E = 1.05 V i.e. to the maximum potential at
which the irreversible overoxidation phenomenon does not yet
take place.
UV-vis-NIR spectroelectrochemistry is especially interesting
in the study of conjugated polymers since the dominant doping
induced spectral changes are observed in the near infrared
spectral range. Figs 6a and b show the evolution of the UV-
vis-NIR spectra of poly(3,30-dioctyl-2,29:59,20-terthiophene),
imposed by increasing working electrode potential, recorded
n n
for the hexane (DP = 12) and the chloroform (DP = 38)
fractions, respectively. The observed spectral changes can
clearly be correlated with the registered voltammograms
(
Fig. 4), although some differences should be pointed out.
The discussed spectroelectrochemical experiment is quasi-
static i.e. the electrode potential is being changed step-wise in
small increments. At each imposed potential the spectrum is
Fig. 5 The onset of the oxidative doping peak (Eon/ox) plotted against
n
DP for different fractions of poly(3,30-dioctyl-2,29:59,20-terthiophene).
3
104 | J. Mater. Chem., 2006, 16, 3099–3106
This journal is ß The Royal Society of Chemistry 2006

View Article Online
extended over a much larger potential range, which is
consistent with the broad nature of the cyclic voltammogram
registered for this fraction. Two spectroscopically distinctly
different doping steps can be distinguished. In the potential
range from E = 0.15 V to E = 0.65 V the electrochemical
doping results in the correlated increase of the two bipolaronic
bands at 732 nm and 1850 nm, the latter being continuously
hypsochromically shifted with increasing electrode potential.
In the same potential range, the band originating from the
p–p* transition of the neutral polymer decreases in intensity,
being still visible, however, at E = 0.65 V. Above this potential
the intensity of the band at 732 nm starts to decrease whereas
the less energetic bipolaronic band continues to grow and to
broaden. Finally at E = 1.00 V both bands merge giving
an extremely broad, essentially featureless, absorption charac-
3
0
teristic of the metallic state.
It is clear from spectroelectrochemical studies that both
polymer fractions are semiconductors of different band struc-
tures. The energy of the p–p* transition is higher for the
hexane fraction (2.56 eV) as compared to the chloroform
fraction (2.25 eV). The energies of the first and the second
bipolaronic transitions, determined for low electrode poten-
tials i.e. for low doping levels, are also more energetic for the
hexane fraction (1.71 eV and 0.86 eV) than for the chloroform
one (1.69 eV and 0.67 eV). Note also that, in the case of
the chloroform fraction, total disappearance of the p–p*
band leads to the metallic state whereas no such state is
observed for the hexane fraction, even after the disappearance
of the p–p* band.
Fig. 6 UV-vis-NIR spectra of poly(3,30-dioctyl-2,29:59,20-terthio-
phene) recorded for increasing electrode potential, E vs Ag/0.1 M
+
Small discrepancies between the potentials of the onset of
the doping process as probed by dynamic (CV) and quasi-
static (spectroelectrochemistry) methods has prompted us to
perform experiments in which spectral changes are recorded
simultaneously with the cyclic voltammogram. As a typical
example, we show here the results obtained for the chloroform
fraction. Five spectral lines were selected for these investiga-
tions: i) one line from the spectral range of the band
corresponding to the p–p* transition (510 nm); ii) one line
from the spectral range of the first doping induced band
Ag reference electrode, in 0.1 M Bu
4
NBF
4
-acetonitrile solution: a)
hexane fraction; b) chloroform fraction.
recorded after a certain time, when no more current and
spectral evolutions are observed. This quasi-static character of
the experiment causes the doping induced changes to appear at
lower potentials compared to cyclic voltammetry investiga-
tions which are dynamic in nature. The first spectroscopic
indication of the oxidative doping of the lower molecular
weight fraction (hexane fraction) appears at E = 0.23 V i.e. at
the potential 60 mV lower than the potential of the onset of
the anodic doping peak in the cyclic voltammogram. The
electrochemical doping gives rise to two doping induced bands
with maxima at 722 nm and 1470 nm, attributed to the
formation of bipolaronic states within the band gap, with
simultaneous decrease of the band associated with the p–p*
transition at 484 nm. The bipolaronic band of lower energy is
being hypsochromatically shifted with increasing electrode
potential. Note that the most pronounced spectral changes
occur in a rather narrow potential range from 0.25 V to 0.4 V
consistent with the shape of the corresponding cyclic
voltammogram. For the higher molecular weight fraction
(
750 nm); iii) three lines from the spectral range of the second
doping induced band (1500 nm, 1700 nm, 2000 nm). The
obtained absorbance–time responses were differentiated in the
form of 2dA/dt for the peaks decreasing in intensity with
increasing potential and as dA/dt for those increasing in
intensity with increasing potential. Their plot vs. E gives, so
called, differential cyclic voltabsorptograms (DCVAs). In
Fig. 7 five voltabsorptograms obtained for the chloroform
fraction, in the potential range from E = 20.3 V to E = +0.7 V,
are compared with the corresponding voltammogram. The
double peak nature of the oxidative doping, clearly seen in
the cyclic voltammograms is very well reproduced in all five
voltabsoptograms with only minimal differences in the peak
positions.
(
chloroform fraction) we have carried out UV-vis-NIR
spectroelectrochemical investigations up to E = 1.05 V, as in
the case of cyclic voltammetry. Spectroelectrochemical signa-
tures of the oxidative doping appear in this case at E = 0.18 V,
again at a slightly lower potential than that of the onset of the
corresponding cyclic voltammetry curve. Contrary to the case
of the hexane fraction, the doping induced spectral changes are
Conclusions
To summarize, we have demonstrated that by fractionation
of poly(3,30-dioctyl-2,29:59,20-terthiophene) into a few frac-
tions differing in molecular weight and showing reduced
This journal is ß The Royal Society of Chemistry 2006
J. Mater. Chem., 2006, 16, 3099–3106 | 3105

View Article Online
4
5
Y. Wu, P. Liu, S. Gardner and B. S. Ong, Chem. Mater., 2005, 17,
2
21.
R. M. Souto Maior, K. Hinkelmann, H. Eckert and F. Wudl,
Macromolecules, 1990, 23, 1268.
6
7
8
M. Zagorska and B. Krische, Polymer, 1990, 31, 1379.
T.-A. Chen and R. D. Rieke, J. Am. Chem. Soc., 1992, 114, 10087.
R. D. McCullough and R. D. Lowe, J. Chem. Soc., Chem.
Commun., 1992, 70.
9
H. Sirringhaus, N. Tessler and R. H. Friend, Science, 1998, 280,
1
741.
0 Z. Bao, A. Dodabalapur and A. J. Lovinger, Appl. Phys. Lett.,
996, 69, 4108.
1 W. Fix, A. Ullman, J. Ficker and W. Clemens, Appl. Phys. Lett.,
002, 81, 1735.
1
1
1
1
2
2 J. F. Chang, S. Baoquan, D. W. Breiby, M. M. Nielsen, T. I. S o¨ lling,
M. Giles, I. McCulloch and H. Sirringhaus, Chem. Mater., 2004,
1
6, 4772.
1
1
3 B. H. Hamadani and D. Natelson, Appl. Phys. Lett., 2004, 84, 443.
4 R. J. Kline, M. D. McGehee, E. N. Kadnikova, J. Liu and
J. M. J. Frechet, Adv. Mater., 2003, 15, 1519.
Fig. 7 Cyclic voltammogram (solid line) and voltabsorptograms
obtained for the chloroform fraction of poly(3,30-dioctyl-2,29:59,20-
terthiophene) in 0.1 M Bu
21
mV s .
4
NBF
4
–acetonitrile solution at the scan rate
15 G. Li, V. Shrotriya, J. S. Huang, Y. Yao, T. Moriarty, K. Emery
and Y. Yang, Nat. Mater., 2005, 4, 864.
5
1
6 P. Barta, F. Cacialli, R. H. Friend and M. Zagorska, J. Appl.
Phys., 1998, 84, 6279.
17 M. Trznadel, A. Pron, M. Zagorska, R. Chrzaszcz and
polydispersity, it is possible to establish the effect of DP
n
on
J. Pielichowski, Macromolecules, 1998, 31, 5051.
several application-important properties of this polymer such
as: the onset of electrochemical doping, the potential range
of its spectroelectrochemical response and the FET charge
carriers’ mobility etc. Since in all cases the observed changes
are monotonic in character, fractionation can be considered as
a convenient tool for tuning the properties of poly(3,30-dioctyl-
1
8 A. Zen, M. Saphiannikova, D. Neher, J. Grenzer, S. Grigorian,
U. Pietsch, U. Asawapirom, S. Janietz, U. Scherf, I. Lieberwirth
and G. Wegner, Macromolecules, 2006, 39, 2162.
19 R. J. Kline, M. D. McGehee, E. N. Kadnikova, J. Liu,
J. M. J. Frechet and M. Toney, Macromolecules, 2005, 38, 3312.
0 A. Zen, J. Pflaum, S. Hirschmann, W. Zhuang, F. Jaiser,
U. Asawapirom, J. P. Rabe, U. Scherf and D. Neher, Adv.
Funct. Mater., 2004, 14, 757.
1 P. Schilinsky, U. Asawapirom, U. Scherf, M. Biele and C. J. Brabec,
Chem. Mater., 2005, 17, 2175.
2 G. Horowitz, Adv. Mater., 1998, 10, 365.
2
2
2
,29:59,20-terthiophene).
Acknowledgements
2
2
3 R. Demadrille, P. Rannou, J. Bleuse, J. L. Oddou, A. Pron and
M. Zagorska, Macromolecules, 2003, 36, 7045.
24 M. C. Gallazzi, L. Castellani, R. A. Marin and G. Zerbi, J. Polym.
Sci., Part A: Polym. Chem., 1993, 31, 3339.
This work was partially financed by the Polish Ministry of
Education and Science in the years 2006–2007 as scientific
project No. 1T09B04730.
2
5 J. Liu, R. S. Loewe and R. D. McCullough, Macromolecules, 1999,
2, 5777.
3
2
6 T. Izumi, S. Kobashi, K. Takimiya, Y. Aso and T. Otsubo, J. Am.
Chem. Soc., 2003, 125, 5286.
References
1
2
3
Handbook of Oligo- and Polythiophenes, ed. D. Fichou, Wiley-
VCH, Weinheim, 1999.
Handbook of Conductive Molecules and Polymers, ed. H. S.
Nalwa, vol. 1–4, Wiley, Chichester, 1997.
I. McCulloch, C. Bailey, M. Giles, M. Heeney, I. Love,
M. Shkunov, D. Sparrowe and S. Tierney, Chem. Mater., 2005,
27 Y. Furukawa, M. Akimoto and I. Harada, Synth. Met., 1987, 18,
151.
28 J.-M. Verilhac, G. LeBlevennec, D. Djurado, F. Rieutord,
M. Chouiki, J.-P. Travers and A. Pron, Synth. Met., in press.
29 H. Sirringhaus, Adv. Mater., 2005, 17, 2411.
30 T. C. Chung, J. H. Kaufman, A. J. Heeger and F. Wudl, Phys. Rev.
B, 1984, 30, 702.
17, 1381.
3
106 | J. Mater. Chem., 2006, 16, 3099–3106
This journal is ß The Royal Society of Chemistry 2006