Regiochemically well-defined fluorenone-alkythiophene copolymers: Synthesis, spectroscopic characterization, and their postfunctionalization with oligoaniline

Article abstract of DOI:10.1021/ma030171z

A new solution processable, regioregular, alternate copolymer of fluorenone and dialkyl-bithiophene, namely poly[(5,5′-(3,3′-di-n-octyl-2,2′-bithiophene))-alt-(2,7-fluore n-9-one)] (abbreviated as PDOBTF), was synthesized by three different preparation methods: chemical or electrochemical oxidation of 2,7-bis(4-octylthien-2-yl)-fluoren-9-one or polycondensation of 2,7-bis(5-bromo-4-octylthien-2-yl)-fluoren-9-one in the presence of Ni(0) reagent. Independent of the preparation method, the crude product is a mixture of high molecular weight fractions and short oligomers. It can be however easily fractionated into fractions differing in their molecular weight by sequential extractions with a series of solvents. The principal absorption band registered for the undoped polymer (λ_max = 384 nm for the THF solution and 389 nm for the solid state) originates from the π-π* transition of the conjugated backbone and is blue-shifted because of the chain torsion effects caused by steric hindrance. This band is accompanied by a peak of smaller intensity (λ_max = 476 nm for the THF solution and 485 nm for the solid state) attributed to the n-π* transition in the carbonyl group of the fluoren-9-one subunit. Preliminary photoluminescence studies show that PDOBTF exhibits a very large Stokes shift and emits red light (λ_max = 631 nm in THF solution and 643 nm in the solid state). Upon chemical p-type doping with FeCl₄^-, the polymer reaches the conductivity of σ_dc = 0.05 S cm^-1. Moessbauer spectroscopy studies of the doping process show that both structural subunits, i.e., the bithiophene subunit and the fluoren-9-one one, participate in the doping. PDOBTF can be relatively easily postfunctionalized by grafting aniline oligomers as pendant groups via the carbonyl groups of the fluoren-9-one subunit. By consequence, the spectrum of the modified polymer can be precisely tuned in the visible region by changing the grafting level.

Full text of DOI:10.1021/ma030171z

Macromolecules 2003, 36, 7045-7054
7045
Regiochemically Well-Defined Fluorenone-Alkylthiophene Copolymers:
Synthesis, Spectroscopic Characterization, and Their
Postfunctionalization with Oligoaniline
Ren a u d Dem a d r ille,† P a tr ice Ra n n ou ,† J oe1l Bleu se,‡ J ea n -Lou is Od d ou ,§ a n d
Ad a m P r on *,†
DRFMC, UMR 5819-SPrAM (CEA-CNRS-Univ. J . Fourier-Grenoble I), Laboratoire de Physique des
Me´taux Synthe´tiques, DRFMC, SP2M, Laboratoire “Nanophysique et Semiconducteurs”, and
DRDC, FRE2427 (CEA-CNRS-Univ. J . Fourier-Grenoble I), Laboratoire de Physico-chimie des Me´taux
en Biologie, CEA Grenoble, 17 Rue des Martyrs 38054 Grenoble Cedex 9, France
Ma lgor za ta Za gor sk a
Faculty of Chemistry, Warsaw University of Technology, 00 664 Warszawa Noakowskiego 3, Poland
Received March 14, 2003; Revised Manuscript Received J une 26, 2003
ABSTRACT: A new solution processable, regioregular, alternate copolymer of fluorenone and dialkyl-
bithiophene, namely poly[(5,5′-(3,3′-di-n-octyl-2,2′-bithiophene))-alt-(2,7-fluoren-9-one)] (abbreviated as
PDOBTF), was synthesized by three different preparation methods: chemical or electrochemical oxidation
of 2,7-bis(4-octylthien-2-yl)-fluoren-9-one or polycondensation of 2,7-bis(5-bromo-4-octylthien-2-yl)-fluoren-
9-one in the presence of Ni(0) reagent. Independent of the preparation method, the crude product is a
mixture of high molecular weight fractions and short oligomers. It can be however easily fractionated
into fractions differing in their molecular weight by sequential extractions with a series of solvents. The
principal absorption band registered for the undoped polymer (λmax ) 384 nm for the THF solution and
389 nm for the solid state) originates from the π-π* transition of the conjugated backbone and is blue-
shifted because of the chain torsion effects caused by steric hindrance. This band is accompanied by a
peak of smaller intensity (λmax ) 476 nm for the THF solution and 485 nm for the solid state) attributed
to the n-π* transition in the carbonyl group of the fluoren-9-one subunit. Preliminary photoluminescence
studies show that PDOBTF exhibits a very large Stokes shift and emits red light (λmax ) 631 nm in THF
solution and 643 nm in the solid state). Upon chemical p-type doping with FeCl4-, the polymer reaches
the conductivity of σdc ) 0.05 S cm-1. Mo¨ssbauer spectroscopy studies of the doping process show that
both structural subunits, i.e., the bithiophene subunit and the fluoren-9-one one, participate in the doping.
PDOBTF can be relatively easily postfunctionalized by grafting aniline oligomers as pendant groups via
the carbonyl groups of the fluoren-9-one subunit. By consequence, the spectrum of the modified polymer
can be precisely tuned in the visible region by changing the grafting level.
In tr od u ction
dialkylbithiophenes,5,6 poly[(5,5′-(3,3′-dialkyl-2,2′-bi-
thiophene))-alt-(2,7-fluoren-9-one)]s offer some advan-
tages unobtainable for the former (vide infra).
Our interests in this new family of polyconjugated
systems arise from the following reasons:
(i) Their highly regular chain microstructure, which
is a simple consequence of the fact that the fluoren-9-
one moiety possesses a symmetry axis and the alkyl
substituents in the bithiophene unit are regularly
distributed via “head-to-head” coupling. Such chain
regularity facilitates ordered supramolecular aggrega-
tion in the solid state.
(ii) Their luminescent properties combined with solu-
tion processability, the latter being assured by the alkyl
side groups.
The development of “plastic” electronic devices such
as field-effect transistors (FETs),1 light-emitting diodes
(LEDs),2 or photovoltaic cells3 resulted in a renewed
interest in polyconjugated macromolecular systems
observed in the past decade. One of the most important
advantages of conjugated polymers is a rather facile
tuning of their optical, spectroscopic, and electronic
properties by appropriate design of their chain structure
and by imposing favorable aggregation of these chains
in the solid state via supramolecular engineering. In this
paper we report on the first example of a new family of
copolymers with conjugated backbone, namely poly-
[(5,5′-(3,3′-dialkyl-2,2′-bithiophene))-alt-(2,7-fluoren-9-
one)]s. Poly(fluoren-9-one-2,7-diyl) homopolymer has
already been reported in the literature;4 however, we
were tempted to prepare copolymers of fluoren-9-one
with alkyl-substituted thiophenes with the goal to tune
its electronic and optical properties and to improve its
solution processability. Although being similar to al-
ready reported alternating copolymers of fluorene and
(iii) The possibility of the tuning of their electrical
transport properties via chemical or electrochemical
doping.
(iv) The possibility of their postpolymerization func-
tionalization via carbonyl groups7 present in the poly-
mer main chain. Thus, poly[(5,5′-(3,3′-dialkyl-2,2′-
bithiophene))-alt-(2,7-fluoren-9-one)]s can be considered,
for example, as polymer precursors for the preparation
of a whole new series of macromolecules whose absorp-
tion spectrum can be precisely tuned via branching
appropriate chromophores as side groups. This approach
† Laboratoire de Physique des Me´taux Synthe´tiques.
‡ Laboratoire “Nanophysique et Semiconducteurs”.
§ Laboratoire de Physico-chimie des Me´taux en Biologie.
* To whom the correspondence should be addressed: Tel 33-4-
38784389; Fax 33-4-38785113; e-mail pron@cea.fr.
10.1021/ma030171z CCC: $25.00 © 2003 American Chemical Society
Published on Web 08/21/2003

7046 Demadrille et al.
Macromolecules, Vol. 36, No. 19, 2003
Photoluminescence measurements of monomers and un-
doped polymers were carried out on a J obin-Yvon HR 460
monochromator equipped with a CCD silicium detector cooled
at 140 K. Excitation was performed with an argon laser at
303, 365, and 458 nm.
is of extreme importance in view of the use of conjugated
polymers as components of organic solar cells8-10 since
it may result in a much better match between the solar
radiation and the polymer absorption, leading in turn
to much better device efficiency.
The dc conductivity (σdc [S cm-1]) measurements of the doped
polymer films were performed at room temperature (RT), on
pressed pellets, in a standard four-probe geometry, using
pressure contacts.
In the first part of the paper we describe the synthesis
of poly[(5,5′-(3,3′-di-n-octyl-2,2′-bithiophene))-alt-(2,7-
fluoren-9-one)] (PDOBTF) and characterize its principal
properties in the undoped state. In the second part we
focus on its p-type doping using FeCl3 as the doping
agent and on the characterization of the doped polymer
by elemental analysis, dc conductivity measurements,
and Mo¨ssbauer spectroscopy. Finally, we give an ex-
ample of postpolymerization functionalization of the
above copolymer with the goal to demonstrate that its
UV-vis spectrum can be tuned by grafting aniline
tetramer to the conjugated main chain via the carbonyl
group present in the fluoren-9-one moiety.
Mo¨ssbauer spectra of FeCl3-doped polymers were measured
at 4.2 K. The samples were loaded in a drybox into a tight
polyamide container prior to the experiment and immediately
transferred to the cryostat of the spectrometer. A typical
thickness of the absorber was 5 mg cm-2 of natural iron. Co-
(Rh) was used as a Mo¨ssbauer source. An R-Fe absorber
operating at RT was used for the velocity calibration. The
spectra were recorded in a constant acceleration mode and
analyzed in a least-squares procedure.
Syn th esis. 5,5-Dim eth yl-2-(4-octylth ien -2-yl)[1,3,2]d i-
oxa bor in a n e (1). 6.8 mL (11 mmol) of n-butyllithium in a
form of 1.6 M solution in hexane was added dropwise, at -50
°C during 5 min, to a stirred solution of 1.96 g (10 mmol) of
3-n-octylthiophene in dry THF (20 mL). The color of the
mixture became slightly yellow. The mixture was additionally
stirred for 60 min at -50 °C and then cooled to -78 °C. At
this stage 9 mL (33 mmol) of tributylborate was added quickly
to the reaction vessel, and the solution was allowed to warm
to RT. Upon warming the reaction mixture became milky. It
was then poured into 1 M HCl containing ice (40 mL) and then
extracted with diethyl ether. The organic layer was washed
twice with brine and dried over magnesium sulfate in the
presence of 5.20 g (50 mmol) of neopentyl glycol for 30 min.
Subsequent separation of the drying agent by filtration and
removal of the solvent, using a rotary evaporator, gave a
transparent oil. Further purification by silica gel column
chromatography with a hexane-ether mixture (9:1) as eluent
afforded 2.34 g of a light yellow oil (yield 76%). 1H NMR
(acetone-d6, 200 MHz, ppm): δ 7.36 (d, 1H, J ) 1.21 Hz), 7.28
(d, 1H, J ) 1.08 Hz), 3.76 (s, 4H), 2.63 (t, 2H, J ) 7.8 Hz),
1.58-1.65 (m, 2H), 1.32-1.25 (m, 10H), 1.01 (s, 6H), 0.89-
0.84 (m, 3H).13C NMR (acetone-d6, 200 MHz, ppm): 146.01,
138.71, 128.37, 73.72 (2C), 33.60, 32.52, 31.56, 31.12, 31.00,
30.81, 30.42, 24.30, 22.87 (2C), 15.32, C-Boron not observed.
Elemental analysis: Calcd for C17H29BO2S: C, 66.23%; H,
9.48%; S, 10.40%. Found: C, 64.90%; H, 9.46%; S, 9.26%.
2,7-Bis(4-octylth ien -2-yl)-flu or en -9-on e (2). 436 mg of
2,7-dibromofluoren-9-one (1.29 mmol) and 890 mg of compound
1 (2.88 mmol) were placed in anhydrous DMF (8 mL). The
mixture was stirred under argon for 10 min, and then 610 mg
of K3PO4 (2.87 mmol) and 150 mg of Pd[P(Ph)3]4 (0.129 mmol)
in 8 mL of DMF were added. The mixture was kept for an
additional period of 12 h at 85 °C with constant stirring and
then allowed to cool to RT. At this stage 10 mL of diethyl ether
was added. After filtration an orange, strongly fluorescent solid
was recovered. This solid was then washed twice with ether
and then with pentane. The product was then separated from
K3PO4 by dissolving in dichloromethane. The solvent was
removed using a rotary evaporator to give 480 mg of the
product (yield 65.4%). 1H NMR (CDCl3, 200 MHz, ppm): δ 7.88
(d, 2H, J ) 1.61 Hz), 7.68 (dd, 2H, J ) 7.79 and 1.74 Hz), 7.47
(d, 2H, J ) 7.79 Hz), 7.22 (d, 2H, J ) 1.34 Hz), 6.90 (d, 2H, J
) 1.07 Hz), 2.61 (t, 4H, J ) 7.39 Hz), 1.56-1.70 (m, 4H), 1.20-
1.40 (m, 20H), 0.85-0.92 (m, 6H).13C NMR (CDCl3, 200 MHz,
ppm): 193.39 (CdO), 144.57 (2C), 142.67 (2C), 142.61 (2C),
135.64 (2C), 135.09 (2C), 131.46 (2C), 125.13 (2C), 121.37 (2C),
120.66 (2C), 120.18 (2C), 31.86 (2C), 30.59 (2C), 30.42 (2C),
29.40 (2C), 29.32 (2C), 29.22 (2C), 22.65 (2C), 14.05 (2C).
Elemental analysis: Calcd for C37H44OS2: C, 78.12%; H,
7.80%; S, 11.27%. Found: C, 77.86%; H, 7.83%; S, 11.46%.
2,7-Bis(5-br om o-4-octylth ien -2-yl)-flu or en -9-on e (3). A
solution of 2 (545 mg (0.96 mmol) in 10 mL of chloroform) was
slowly added to a solution of N-bromosuccinimide (NBS) (427
mg (2.40 mmol) in 10 mL of the same solvent). The mixture
was stirred overnight at RT under argon. It was then washed
with water, and the organic layer was dried over magnesium
Exp er im en ta l Section
Rea gen ts a n d Ch em ica ls. All reagents and chemicals
were purchased from Aldrich. Chloroform and nitromethane
were dried over CaCl2 and distilled in a vacuum line prior to
the use. FeCl3 was dried in a high-vacuum line by dynamic
pumping at the temperature of 60 °C for 2 h. Other reagents
and chemicals were used as received.
Ch a r a cter iza tion Tech n iqu es. All synthesized monomers
and polymers in the undoped state were characterized by H
1
and 13C NMR and elemental analysis. Undoped polymers were
also characterized by Fourier transform infrared (FT-IR) and
ultraviolet-visible (UV-vis) absorption spectroscopies. Poly-
mers doped with FeCl3 were subjected to elemental analysis,
room temperature dc conductivity measurements, and 57Fe
Mo¨ssbauer spectroscopy.
NMR were recorded on a Bruker AC 200 MHz or on a
Varian 400 MHz spectrometer. Acetone-d6 or chloroform-d
(CDCl3) containing TMS as an internal standard were used
as solvent depending on the solubility of the material.
Elemental analyses (C, H, N, and S in the case of monomers
and undoped polymers and additionally Fe and Cl in the case
of FeCl4--doped samples) were carried out by the Analytical
Service of CNRS Vernaison (France).
FT-IR spectra of undoped polymers were recorded on a
Perkin-Elmer paragon500 spectrometer (wavenumber range:
400-4000 cm-1; resolution: 2 cm-1) using the KBr pellet
technique. Solution (THF) and solid-state (thin films cast from
THF solutions) UV-vis absorption spectra were recorded in
situ during the SEC analyses using the diode array UV-vis
spectrometer of the 1100HP Chemstation and on a Perkin-
Elmer Lambda 2 spectrometer, respectively.
Molecular weight determinations of polymer fractions stud-
ied were measured using size exclusion chromatography (SEC)
on a 1100HP Chemstation equipped with a 300 × 7.5 mm
PLgel Mixed-D 5 µm/104 Å column. Detection was performed
by a diode array UV-vis detector and a refractive index
detector. The column temperature and the flow rate were fixed
to 313 K and 1 mL min-1, respectively. The calibration curve
was built using 10 polystyrene (PS) narrow standards (S-M-
10* kit from Polymer Labs). Two runs of 20 µL injection of ca.
2 mg mL-1 polymer in HPLC grade THF solutions were
typically analyzed for each sample with a UV-vis detection
located at 355 nm.
Mass spectra were acquired on a LCQ-ion trap Ther-
mofinnigan spectrometer equipped with an electrospray source
(MS-ESI). Electrospray full scan spectra, in the range of m/z
180-300 uma, were obtained by infusion through fused silica
tubing at 2-10 µL min-1. The solutions (CH2Cl2) were ana-
lyzed in the positive mode. The LCQ calibration (m/z 50-2000)
was achieved according to the standard calibration from the
manufacturer (mixture of caffeine, MRFA, and ultramark
1621). The temperature of the heated capillary of the LCQ was
set to 433 K.

Macromolecules, Vol. 36, No. 19, 2003
Fluorenone-Alkylthiophene Copolymers 7047
sulfate and filtered. Evaporation of the solvent followed by
recrystallization of the product from chloroform/methanol
afforded 540 mg of an orange solid (yield 77.5%). 1H NMR
(CDCl3, 200 MHz, ppm): δ 7.74 (dbr, 2H, J ) 1.61 Hz), 7.56
(dd, 2H, J ) 7.80 and 1.75 Hz), 7.41 (dbr, 2H, J ) 7.80 Hz),
7.04 (s, 2H), 2.54 (t, 4H, J ) 7.26 Hz), 1.53-1.64 (m, 4H), 1.26-
1.32 (m, 20H), 0.85-0.92 (m, 6H). 13C NMR (CDCl3, 200 MHz,
ppm): 192.93 (CdO), 143.44 (2C), 142.78 (2C), 142.14 (2C),
134.82 (2C), 135.06 (2C), 134.82 (2C), 131.15 (2C), 124.60 (2C),
120.90 (2C), 120.83 (2C), 109.11 (2C), 31.86 (2C), 29.65 (4C),
29.36 (2C), 29.24 (2C), 29.21 (2C), 22.65 (2C), 14.06 (2C).
Elemental analysis: Calcd for C37H42Br2OS2: C, 61.16%; H,
5.83%; S, 8.83%. Found: C, 61.53%; H, 5.83%; S, 8.51%.
P oly[(5,5′-(3,3′-d i-n -octyl-2,2′-bith iop h en e))-a lt-(2,7-flu -
or en -9-on e)] (P DOBTF ). Method A. A solution of 530 mg of
anhydrous ferric chloride (4.18 mmol) in a mixed solvent
consisting of 5 mL of nitromethane and 30 mL of chloroform
was added dropwise to a solution of 2 (450 mg, 0.79 mmol) in
30 mL of freshly distilled and degassed chloroform. The
addition was performed at 0 °C with constant stirring over a
period of 90 min. At the end of the addition, the mixture was
warmed to 10 °C and was maintained at this temperature for
an additional period of 60 min. The reaction mixture was then
allowed to warm to RT and stirred for 12 h. Subsequently, it
was concentrated by pumping under vacuum and then pre-
cipitated in 100 mL of methanol. The crude polymer was then
dissolved in 50 mL of chloroform and washed four times with
a 0.1 M aqueous solution of ammonia (150 mL each time). In
the next step the polymer was stirred for 48 h with the same
aqueous solution. As-synthesized polymer usually contains
minute amounts of dopants of unidentified chemical nature
and requires further dedoping. The dedoping was achieved by
washing the chloroform solution of the polymer with an EDTA
aqueous solution (0.05 M, 200 mL). The polymer was then
washed with water twice and then dried under vacuum to give
95%). 1H NMR (CDCl3, 400 MHz, ppm): δ: 7.91 (m, 2H), 7.69
(m, 2H), 7.49 (m, 2H), 7.29 (s, 2H), 2.60 (t, 4H, J ) 8 Hz), 1.64
(m, 4H), 1.27 (m, 20H), 0.87 (m, 6H). 13C NMR (CDCl3, 200
MHz, ppm): δ: 193.18 (CdO), 143.75 (2C), 142.60 (2C), 142.35
(2C), 135.07 (2C), 131.34 (2C), 128.76 (2C), 125.42 (2C), 125.12
(2C), 121.05 (2C), 120.66 (2C), 31.92 (2C), 30.72 (4C), 29.45
(4C), 29.24 (2C), 22.66 (2C), 14.08 (2C). IR (KBr, cm-1): 3054
(w), 2953 (m), 2924 (s), 2852 (s), 1720 (s), 1601 (w), 1583 (w),
1537 (w), 1472 (s), 1434 (m), 1376 (w), 1292 (w), 1260 (w), 1161
(w), 1124 (w), 1026 (w), 823 (m), 785 (m). Elemental analysis:
Calcd for C37H44OS2: C, 78.12%; H, 7.80%; S, 11.27%. Found:
C, 76.41%; H, 7.38%; S, 10.74%.
UV-vis Sp ectr oelectr och em istr y of Ch em ica lly P o-
lym er ized P DOBTF . For spectroelectrochemical investiga-
tions a thin layer of the polymer (method A) was deposited on
an ITO transparent electrode by casting from a THF solution.
The experiments were carried out in a one-compartment
spectroelectrochemical cell using a Pt counter electrode and
an Ag/0.1 M AgNO3 reference electrode. A 0.1 M solution of
tetrabutylammonium tetrafluoroborate solution in acetonitrile
was used as the electrolyte.
Ch em ica l Dop in g of P DOBTF . Chemical doping with
FeCl3 was carried out for the polymer prepared by method A
using a standard high-vacuum line technique. In a typical
procedure an appropriate amount of dry nitromethane was
distilled in a vacuum line to the sidearm of the doping reactor
containing vacuum-dried FeCl3 as to give a 0.3 M doping
solution. This solution was then transferred to the main arm
of the reactor and kept in contact with the polymer film for 2
h. During the doping reaction the film, floating initially on
the surface of the doping solution, fell to the bottom of the
reactor which was caused by the doping-induced increase of
the polymer density. The excess of the unreacted FeCl3, as well
as side-reaction products, was removed by repeated washing
with small amounts of dry nitromethane. Finally, the polymer
was pumped in a high-vacuum line for 2 h.
1
327 mg of a red-brown powder (yield 73%). H NMR (CDCl3,
200 MHz, ppm): δ: 7.78 (sbr, 2H), 7.59 (d, J ) 7.6 Hz, 2H),
7.35 (d, J ) 7.8 Hz, 2H), 7.19 (s, 2H), 2.53 (m, 4H), 1.55 (m,
4H), 1.18 (m, 20H), 0.80 (m, 6H). 13C NMR (CDCl3, 200 MHz,
ppm): δ: 193.29 (CdO), 144.55 (2C), 143.8 (2C), 142.59 (2C),
135.11 (2C), 131.34 (2C), 128.71 (2C), 125.48 (2C), 125.13 (2C),
121.11 (2C), 120.70 (2C), 31.89 (2C), 30.32 (2C), 29.67 (4C),
29.42 (2C), 29.24 (2C), 22.66 (2C), 14.08 (2C). IR (KBr, cm-1):
3054 (w), 2953 (m), 2924 (s), 2852 (s), 1720 (s), 1601 (w), 1583
(w), 1537 (w), 1472 (s), 1434 (m), 1376 (w), 1292 (w), 1260 (w),
1161 (w), 1124 (w), 1026 (w), 823 (m), 785 (m). Elemental
analysis: Calcd for C37H44OS2: C, 78.12%; H, 7.80%; S, 11.27%.
Found: C, 77.10%; H, 7.70%; S, 9.95%.
Method B. In a one-compartment, three-electrode electro-
chemical cell equipped with a flat platinum working electrode
(3 mm2), a Pt wire counter electrode, and an Ag/AgNO3
reference electrode, 10 mL of the electrolyte composed of 0.1
M tetrabutylammonium tetrafluoroborate solution in dichlo-
romethane was placed, and the cell was purged with argon
for 10 min. 11.36 mg (0.02 mmol) of monomer 2 was added to
this electrolyte solution to obtain 2 × 10-3 M concentration
with respect to 2. The film was produced by consecutive
potential cycling (100 scans) in the range from -0.2 to +0.9 V
vs Ag/Ag+ with the scan rate of 100 mV s-1. The deposited
polymer was then reductively dedoped by cycling 50 times in
the potential range of 0.0 to -0.2 V with the same scan rate.
The reduced (undoped) polymer deposited on the Pt working
electrode was then washed with acetonitrile, dried, and finally
dissolved in THF. This solution was then used for the SEC
characterization of the electrochemically synthesized polymer.
Method C. Ni(COD)2 (218 mg, 0.79 mmol), COD (72 mg, 0.66
mmol), and 2,2′-bipyridyl (124 mg, 0.79 mmol) were mixed
together in 10 mL of anhydrous DMF. The solution was then
heated at 55 °C for 40 min and then slowly added (over 20
min) to a solution of 3 (480 mg, 0.66 mmol) in dry THF (10
mL) kept at the same temperature. The mixture was then
maintained at 55 °C under stirring in the dark for 48 h. It
was then allowed to cool to RT and precipitated with 300 mL
of methanol. The precipitate was filtered and dried under
dynamic vacuum to provide 355 mg of a red powder (yield:
Gr a ft in g of An ilin e Tet r a m er t o P DOBF T Ma in
Ch a in : P DOBTF -4EB1 a n d P DOBTF -4EB2. Aniline tet-
ramer in the oxidation state of emeraldine (4EB), which is not
commercially available, was prepared using a modification of
the method described in ref 11. The exact preparation proce-
dure used in this research can be found elsewhere.12
Grafting of the tetramer to the PDOBTF main chain can
briefly be described as follows. 86 mg of PDOBTF (0.15 mmol)
obtained by oxidative polymerization (dichloromethane frac-
tion Mn ) 5.48 kDa equiv PS and Mw/Mn ) 2.01) were dissolved
in 15 mL of hot dry toluene under argon. 100 mg (0.27 mmol)
of aniline tetramer was added, and the resulting mixture was
cooled to 0 °C. 34 mg (0.18 mmol) of TiCl4 was then added
dropwise, and the solution turned instantaneously greenish.
The mixture was left under stirring at 0 °C for an additional
10 min and then allowed to warm to RT. In the next step it
was heated at 110 °C overnight. After this period the reaction
mixture was cooled to RT and concentrated using a rotary
evaporator. After precipitation from methanol (350 mL), the
crude polymer (black powder) was recovered by filtration. It
should be noted here that, due to the acidity of the reaction
mixture, the oligoaniline side groups grafted to the PDOBTF
main chain are “doped” or, in other words, protonated and by
consequence must be dedoped, i.e., transformed from their salt
form to the base form. The dedoping was performed by stirring
the crude polymer in 50 mL of 0.05 M aqueous solution of
ammonia for 15 h. The dedoped polymer was then separated
by filtration and washed consecutively with water, 200 mL of
methanol, and 200 mL of acetone to remove the remaining
traces of “free” aniline tetramer which were not grafted to
PDOBTF. The polymer with grafted oligoaniline side groups
was separated into two fractions: the “low grafting” level
fraction which was extracted with CHCl3 in a Soxhlet ap-
paratus (PDOBTF-4EB1) and the high grafting level fraction
which remained insoluble (PDOBTF-4EB2). This fraction was
then dried under vacuum to yield 90 mg of a black powder
(65% yield). IR (KBr, cm-1): 3340 (w), 3013 (w), 2940 (m), 2916
(s), 2847 (s), 1717 (w), 1588 (s), 1494 (s), 1461 (s), 1376 (m),

7048 Demadrille et al.
Macromolecules, Vol. 36, No. 19, 2003
Sch em e 1. Ch em ica l F or m u la e of 1, 2, 3, a n d P DOBTF
a n d Th eir Syn th esis P a th w a ysa
F igu r e 1. Selected cyclic voltammograms (scan numbers are
indicated in the figure) corresponding to voltammetric elec-
troplymerization of 2 (2 × 10-3 M). Scan rate 100 mV s-1 (E
vs Ag/0.1 M AgNO3, electrolyte 0.1 M Bu4NBF4 in dichlo-
romethane). The two arrows showing the oxidation (arrow
pointing upward) and reduction (arrow pointing downward)
waves are only guidelines.
applying the former, it is convenient to use Andersson’s
procedure17 yielding, in may cases, higher molecular
weight polymers. In this method the potential of the
oxidative couple is adjusted by the Fe3+/Fe2+ ratio
following the Nernst equation. As a result, the polym-
erization can be carried out at relatively mild conditions
without overoxidation effects. Electrochemical oxidative
polymerization can be carried out either with constant
current or with constant potential or by potential
scanning (voltammetric electropolymerization). Since
the potential control is of crucial importance in view of
the possibility of overoxidation effects, the latter two
methods are preferable. Cyclic voltammetry scans ac-
companying the voltammetric electropolymerization of
2 are shown in Figure 1. In addition to the onset of the
irreversible polymerization peak, a redox couple in-
creasing in intensity with each consecutive scan can
clearly be seen. This couple is associated with the doping
and dedoping of the polymer deposited on the electrode.
PDOBTF can also be prepared via polycondensation of
3. This method is, in reality, based on Yamamoto’s
coupling18,19 using zerovalent nickel reagent, bis(1,5-
cyclooctadiene)nickel(0), Ni(COD)2, in the presence of
1,5-cyclooctadiene (COD) and 2,2′-bipyridyl (Bpy). Re-
cently,20 it has been found that the difference in the
sequence of the reagents and catalyst addition is of
crucial importance for obtaining high molecular weight
polymers. Inspired by the data presented in ref 17, we
have carried out the polycondensation reaction by slowly
adding the hot nickel catalyst dissolved in DMF to the
monomer solution in THF. This order of additions
essentially eliminates the formation of the dinickel-
substituted complex which is known to be unstable and
to undergo hydrolysis or decomposition,21 ending, in this
manner, the chain propagation. Since chain termination
can also be caused by the polymer precipitation, we have
changed the solvent, in comparison to the original
procedure described in ref 20, with the goal to increase
the solubility of the polymer in the reaction medium.
Molecu la r Weigh t. In a comparative study of the
preparation of new polymers by three different methods,
the first problem to be addressed is the influence of the
preparation procedure on the molecular weight of the
resulting polymer and its distribution. SEC chromato-
a
Reagents and conditions: (i) BuLi, THF; B(OC4H9)3, -78
°C; neopentyl glycol, ether; (ii) 0.5 equiv of 2,7-dibromofluoren-
9-one, Pd[P(Ph3)]4, K3PO4, DMF, 85 °C; (iii) 2 NBS, CHCl3;
(iv) chemical oxidative polymerization (method A): FeCl3,
CHCl3/CH3NO2, 0 °C; (v) electrochemical polymerization
(method B): CH2Cl2, Bu4NBF4; (vi) polycondensation polym-
erization (method C): Ni(COD)2, COD, Bpy, THF/DMF, 55 °C,
48 h.
1291 (m), 1252 (m), 1157 (m), 1124 (m), 1026 (w), 906 (w), 818
(m), 783 (w), 744(w), 693 (w).
Resu lts a n d Discu ssion
Syn th esis. Suzuki or Stille coupling are among the
most popular synthetic routes for the preparation of
conjugated copolymers. They can be used directly in the
copolymerization reaction5,13,14 or alternatively in the
preparation of monomers combining building blocks
originating from different homopolymers. Such “hybrid-
type” monomers can then be polymerized by other
methods to give copolymers with well-defined regio-
chemistry.5,6 In the presented research we have ex-
ploited the latter approach. The exact synthetic strategy
is based on the synthesis of a symmetric monomer
containing fluoren-9-one and 3-n-octylthiophene units
which upon homocoupling give polymer chains with
regioregular head-to-head sequence of the alkyl sub-
stituents. Both the synthesis of the monomer and its
polymerizations are outlined in Scheme 1. Monomer 2,
2,7-bis(4-octylthien-2-yl)-fluoren-9-one, can be obtained
from 2,7-dibromofluoren-9-one which is commercially
available and 5,5-dimethyl-2-(4-octylthien-2-yl)[1,3,2]-
dioxaborinane (1) prepared according to the literature
method.15 The reaction involves Suzuki coupling with
palladium tetrakis(triphenylphosphine) (Pd[P(Ph)3]4) as
the catalyst. Direct bromination of 2 with N-bromosuc-
cinimide affords 3 in good yield.16 Both 2 and 3 can be
polymerized to give PDOBTF (see Scheme 1). Oxidative
polymerization of 2 can be carried out chemically using
FeCl3 as the oxidizing agent or electrochemically. When

Macromolecules, Vol. 36, No. 19, 2003
Fluorenone-Alkylthiophene Copolymers 7049
Ta ble 1. MS-ESI vs SEC Ch a r a cter iza tion s of Mon om er 2
a n d of th e Eth er F r a ction of P DOBTF (Meth od A:
Ch em ica l Oxid a tive P olym er iza tion )a
Ta ble 2. Ma cr om olecu la r P a r a m eter s of Cr u d e P olym er s
(Meth od A: Ch em ica l Oxid a tive P olym er iza tion ; Meth od
B: Electr och em ica l P olym er iza tion (Bold ); Meth od C:
P olycon d en sa tion P olym er iza tion (Ita lic)), a n d Th eir
F r a ction s Obta in ed by Sequ en tia l Extr a ction sa
absolute
mass
m/z from
MS-ESI
Mpeak (Da equiv
PS) from SEC
oligomers
Fcorr
Mn (kDa
equiv PS) equiv PS) Mw/Mn
Mw (kDa
abundance
(wt %)
monomer 2
dimer
trimer
tetramer
pentamer
568.28
1134.55
1700.82
2267.09
2833.35
568.3
1134.6
1700.9
1134.7 (2+)
1417.7 (2+)
299.6
913.4
1645.1
2546.7
3645.6
0.53
0.81
0.97
1.12
1.55
crude
PDOBTF-A
P DOBTF -B
PDOBTF-C
ether-A
4.99
5.88
2.26
1.30
0.79
1.85
2.17
2.32
1.94
5.48
4.82
28.50
b
31.55
12.82
7.29
1.72
1.37
2.46
3.99
3.64
2.21
11.0
6.77
62.87
b
6.32
2.18
3.23
1.33
1.72
1.33
1.84
1.57
1.15
2.01
1.40
2.21
b
100
100
100
a
m/z denotes the mass obtained from MS-ESI, Mpeak denotes
the molecular weight at the peak apex obtained from SEC, and
Fcorr is the correction factor.
20.70
7.75
2.40
1.40
0.40
0.80
33.00
24.75
31.50
8.75
12.00
56.55
ether-C
acetone-A
acetone-C
n-pentane-A
n-pentane-C
CH2Cl2-A
CH2Cl2-C
THF-A
THF-C
CH2Cl3-A
CHCl3-C
grams of crude polymers, independent of the polymer-
ization method used, are rather complex and on their
longer retention time side consist of several sharp peaks.
Mass spectrometry (MS-ESI) investigations carried out,
for the ether and acetone fractions, in parallel to the
SEC ones, unequivocally show that the above-mentioned
peaks correspond to oligomers that from dimer to
pentamer give clear maxima in the chromatograms. In
contrast to SEC, mass spectrometry gives access to
absolute values of the molecular weight. Thus, a com-
parison of the mass spectrometry data with those
obtained using SEC with polystyrene standards pro-
vides the values of the correction factors (Fcorr) for the
chromatographic determination of the molecular weight
of the polymer under investigation. The results of this
comparative study are summarized in Table 1. The
41.01
13.27
74.11
19.20
1.81
1.45
a
Mn denotes number-average molecular weight, Mw weight-
average molecular weight, and Mw/Mn the polydispersity coef-
ficient. The abundance of each fraction in the crude polymer is
b
expressed in wt %. Degraded after prolonged storage at RT.
correction factor determined as the ratio of Mn(SEC)
/
Mn(MS-ESI) increases with the increase of Mn(MS-ESI) from
0.81 to 1.55 in the range of the molecular weights
studied. This means that for short oligomers the SEC
method underestimates the molecular weight whereas
for longer ones it overestimates it with the cross-
over from underestimation to overestimation around
Mn(MS-ESI) ) 2000 g mol-1. Interestingly, for comparable
molecular weights, the calculated correction factors
determined here for PDOBTF are very similar to those
determined for regioregular poly(3-hexylthiophene-2,5-
diyl) (P3HT) by comparison of MALDI-TOF and SEC
data.22
As already stated, crude polymers can be considered
as a mixture of oligomeric and polymeric species. Mn,
Mw, and polydispersity coefficients determined for crude
polymers using polystyrene standards are collected in
Table 2. One should note that the crude polymers,
independent of whether they are obtained by chemical
oxidative polymerization of 2 or by polycondensation of
3, show rather large polydispersity coefficients. For this
reason we have undertaken the task of their fraction-
ation not only with the goal to remove the oligomer
fractions but also to obtain polymer fractions with
significantly lower polydispersity. Few papers have been
published on the fractionation of poly(thiophene)
derivatives.23-25 In particular, in ref 23 an efficient
method for the fractionation of regioregular P3HT has
been proposed, leading to relatively sharp polymer
fractions. The method is based on sequential extractions
of the fractions with increasing molecular weight using
an experimentally determined series of solvents. For the
purpose of this research we have slightly modified the
original procedure used for regioregular P3HT. In
particular, the crude polymer was first extracted with
ether in a Soxhlet apparatus until the filtrate was
colorless. The remaining part, insoluble in ether, was
then consecutively extracted with acetone, n-pentane,
F igu r e 2. SEC elugrams of monomer 2, of the crude polymer
PDOBTF prepared by chemical oxidative polymerization
(method A), and of its fractions obtained by consecutive Soxhlet
extractions (ether, acetone, n-pentane, CH2Cl2, THF, CHCl3).
The SEC elugram of the acetone fraction is omitted for clarity
reasons.
dichloromethane, tetrahydrofuran, and chloroform in
the same manner as in the case of the extraction with
ether. Macromolecular parameters of all fractions ob-
tained by sequential extractions are collected in Table
2. In the last column the content of each fraction in the
crude polymers, expressed in wt %, is indicated. The
elution curves recorded for monomer 2, the crude
polymer (method A), and each fraction are shown in
Figure 2.
It is clear that the ether, acetone, and n-pentane
fractions, which account for ca. 23 wt % of the polymer
prepared by method A, are in reality mixtures of
oligomers. The fractions obtained by extraction with
THF and chloroform exhibit the highest molecular

7050 Demadrille et al.
Macromolecules, Vol. 36, No. 19, 2003
F igu r e 3. UV-vis and photoluminescence (PL: λexc ) 365
nm) solution (s) and solid state (- - -) spectra of the THF
fraction of PDOBTF prepared by method A.
F igu r e 4. UV-vis-NIR spectroelectrochemical data recorded
for thin solid film of PDOBTF (method A, THF fraction)
deposited on an ITO electrode (E vs Ag/0.1 M AgNO3, electro-
lyte 0.1 M Bu4NBF4 in acetonitrile).
Ta ble 3. UV-vis a n d P h otolu m in escen ce (P L) Da ta of
Mon om er 2, P DOBTF , a n d Its Oligom er s (THF Solu tion s
a n d Th in Solid F ilm )
should be noted here. This shift can be interpreted as a
result of the conjugation of the carbonyl group with a
delocalized π-bonding system which in the case of the
monomer must extend to the thiophene units. The n-π*
band is very little dependent on the chain length. It
bathochromically shifts when going from the monomer
to the dimer, and then its position stabilizes.
λmax (nm)
λmax (nm)
λmax PL (nm)
π-π* transition n-π* transition (λexc ) 365 nm)
monomer 2
dimer
trimer
tetramer
pentamer
CH2Cl2
THF
CHCl3
film
370
374
376
378
380
382
384
384
389
468
476
476
476
476
476
476
476
485
608
In our further investigations we have used UV-vis-
NIR spectroelectrochemistry with the goal to corrobo-
rate the above-proposed assignments. The spectra re-
corded at increasing electrode potentials are presented
in Figure 4. First, we note that the absorption bands
measured for the thin solid film are significantly
broadened as compared to those recorded for the solu-
tion spectrum. Moreover, they are bathochromically
shifted to ca. 389 and 485 nm, i.e., by 5 and 9 nm with
respect to their positions in the solution spectrum. Such
a behavior is indicative of a small solvatochromic effect.
Solvatochromism is a common phenomenon in alkyl-
thiophene-based conjugated polymers.26 The hypsochro-
mic shift in the position of the bands occurring upon
the dissolution of the polymer is caused by macromol-
ecule-solvent interactions which change the conforma-
tion of the chain from more planar in the solid state to
less planar in solution. In the latter the conjugation is
less pronounced since the overlap of pz orbitals is less
effective.
Before the interpretation of the spectroelectrochemi-
cal data a short comment is needed. Electrochemical
oxidative doping of conjugated polymers with non-
degenerated ground state gives rise to a bleaching of
the band corresponding to the π-π* transition in the
conjugated backbone of the polymer. Simultaneously,
three (or two) bands grow in the less energetic part of
the spectrum associated with the formation of radical
cations (polarons) or dications (bipolarons), respectively.
In this perspective the doping of PDOBTF should only
involve the aromatic π-system of the polyconjugated
backbone of the polymer chain, leaving the carbonyl
group intact. As seen from Figure 4, first spectroscopic
indications of the doping process appear at the potential
of E ) 0.7 V. The doping is manifested by a small
decrease of the band at 389 nm and the appearance of
a band at ca. 650 nm accompanied by a second band in
the near-infrared part of the spectrum, whose onset is
seen in Figure 4 in the form of steadily increasing
absorbance in the spectral range 850-1100 nm. Further
629
631
631
643
weight. Again, it should be pointed out that Mn and Mw
listed in Table 2, for all fractions higher than the
acetone one, are overestimated.
Op tica l a n d Sp ectr oelectr och em ica l P r op er ties.
THF solution and solid-state absorption as well as
photoluminescence spectra of PDOBTF prepared by
method A are presented in Figure 3. In the absorption
spectrum, measured in solution, three clear bands can
be distinguished at 310, 384, and 476 nm. We assign
the band at 384 nm to the π-π* transition in the
conjugated backbone of the copolymer. The position of
this band is slightly hypsochromically shifted as com-
pared to the corresponding band recorded for poly[(5,5′-
(3,3′-dialkyl-2,2′-bithiophene))-alt-(2,7-(9,9-dialkylfluo-
rene))], i.e., the polymer analogous PDOBTF but
containing two alkyl groups at the 9-position instead of
the carbonyl group.5 Following the assignment proposed
in ref 4 for poly(fluoren-9-one-2,7-diyl) homopolymer, we
attribute the band at 476 nm to the n-π* transition of
the carbonyl group. UV-vis absorption spectroscopy
data for the oligomers and for the highest molecular
weight fractions of PDOBTF prepared by method A are
collected in Table 3 and compared with that recorded,
in the same solvent (THF), for the monomer, i.e., 2,7-
bis(4-octylthien-2-yl)-fluoren-9-one (2). The increase in
the chain length results in a continuous bathochromic
shift of the π-π* band, as expected. Beginning from the
THF fraction, no further increase of this shift is
observed, consistent with the SEC results which show
that for these fractions no contribution from the oligo-
meric species exists.
A large bathochromic shift (>80 nm) of the peak
corresponding to the n-π* transition in the carbonyl
group for both the monomer and PDOBTF in compari-
son to the analogous band in fluoren-9-one molecule4

Macromolecules, Vol. 36, No. 19, 2003
Fluorenone-Alkylthiophene Copolymers 7051
increase of the electrode potential results in a gradual
bleaching of the band at 389 nm. To the contrary, with
the increasing electrode potential, the band at 485 nm
increases in intensity. The apparent growth of this band
is caused by the fact that it is superimposed on the low-
energy tail of the first doping-induced band whose
intensity grows with increasing doping level. Thus, the
spectroelectrochemical behavior of our copolymer is
consistent with the band assignment proposed in ref 4
for poly(fluoren-9-one-2,7-diyl) homopolymer with the
band at 389 nm attributed to the π-π* transition in
the polyconjugated chain and the band at 485 nm to the
n-π* transition of the carbonyl group.
To better characterize the newly synthesized polymer,
we have registered its photoluminescence spectra. Since
the presented results are only qualitative, they must
be considered as preliminary. Solution photolumines-
cence spectra of PDOBTF show one clear peak at 631
nm whose shape and position are independent of the
energy of excitation line (see Table 3 and Figure 3). In
the solid state the emission band is bathochromically
shifted by 12 nm. A comparison of the positions of the
luminescence peak of the monomer, i.e., 2,7-bis(4-
octylthien-2-yl)-fluoren-9-one (2), and the polymer shows
that the polymerization induces a significant batho-
chromic shift of 25 nm, much higher than that observed
in the absorption spectra (vide supra).
troscopy, and dc conductivity measurements.
With the goal to obtain the saturation or close to
saturation doping level large excess of the doping agent
was used. The most obvious phenomenological conse-
quence of the doping process is the increase of the
polymer conductivity over several orders of magnitude.
This is also observed for the copolymer described here,
whose RT dc conductivity reaches the value of 0.05 S
cm-1 (measured on pressed powder). The mechanism
of the doping of conjugated polymers with FeCl3 has
been described more than 20 years ago.33 It involves a
redox reaction accompanied by an acid-base one:
2FeCl3 + 1e- f FeCl4- + FeCl2
P - 1e- f P+
where P+ denotes the segment of the polymer being
oxidized to carbocation as a result of the doping. FeCl4
-
anions created according to the above reaction scheme
diffuse to the polymer matrix in order to compensate
the positive charge imposed on the polymer chain in the
course of the doping process. Elemental analysis of
doped PDOBTF is consistent with the above picture
since the analytically determined Fe:Cl ratio is, within
the experimental error, equal to1:4. The doping level
corresponds to 1.2 dopant anion (FeCl4-) per polymer
repeat unit or equivalently to 6 dopants per 5 repeat
It is instructive to compare the absorption and
luminescence spectra of PDOBTF with those reported
for poly(fluorene)s and poly(fluoren-9-one-2,7-diyl) ho-
mopolymers. Poly(fluorene)s are blue emitters with a
relatively low Stokes shift. The replacement of dialkyl
groups with carbonyl ones at the 9-position results in a
significant increase of the Stokes shift as observed in
random copolymers of fluorene and fluoren-9-one27 as
well as in oxidatively degraded poly(fluorene)s.28-30 In
both cases the keto groups act as “guest” emitters which
efficiently trap singlet excitons created on the conju-
gated backbone by different types of excitation energy
transfer mechanisms.29,30 Similar effects give rise to a
large Stokes shift in poly(fluoren-9-one-2,7-diyl) ho-
mopolymer31 whose absorption band due to the back-
bone π-π* transition is located at 375 nm whereas the
maximum of the photoluminescence band is at 580 nm.
Thus, a very large Stokes shift in PDOBTF of ca. 245
nm, resulting in red emission, is not unexpected.
-
units: Calcd for C37H42OS2(FeCl4
)
1.2: %C, 55.23; %H,
5.22; %S, 7.96; %Cl, 21.20; %Fe, 8.36. Found: %C, 54.60;
%H, 5.56; %S, 6.88; %Cl, 20.89; %Fe, 7.77.
Intuitively, one may expect that the charge introduced
in the doping process should be localized on the
bithiophene units which are easier to oxidize than the
fluoren-9-one ones. This would however imply an un-
usually high charge concentration in the bithiophene
segments, ca. 1 charge per 1.6 thiophene ring, which is
approximately twice higher than typically observed in
polythiophene homopolymers (1 charge per 3 thiophene
rings). Thus, it can be postulated that in PDOBTF doped
to saturation both subunits undergo the doping process
or at least some fraction of the fluoren-9-one segments
are doped. This postulate is consistent with Raman
spectroscopic studies of the polymer doped to saturation.
In this case the doping-induced shift toward lower
wavenumbers is observed not only for the band at 1460
cm-1 ascribed to the CR-Câ stretching deformations in
the thiophene ring but also for the band at 1603 cm-1
originating from C-C stretching deformations in the
aromatic ring of the fluoren-9-one subunit.
Finally, the influence of the head-to-head coupled
3-alkylthiophene rings on the optical properties of the
copolymer studied should be briefly discussed. Evi-
dently, the alkyl substituents create additional steric
hindrance which results in an increase of the torsion
angle along the polymer backbone. As a consequence,
the absorption corresponding to the π-π* of the polymer
backbone is hypsochromically shifted to 385 nm. Similar
behavior is observed not only for head-to-head, tail-to-
tail coupled poly(alkylthiophene) homopolymers32 but
also for poly[(5,5′-(3,3′-dialkyl-2,2′-bithiophene))-alt-(2,7-
(9,9-dialkylfluorene))]s which in their bithiophene seg-
ments are identical with our copolymer.5
The doping of both subunits should also be manifested
in 57Fe Mo¨ssbauer spectra of the doped polymer.34 It is
known that both the isomer shift (IS) and the quadru-
pole splitting (QS) values in FeCl4--doped conjugated
polymers are dependent on the chemical constitution
of the polycation neutralizing the charge of the dopant.35
In particular, the IS values in doped heterocyclic and
aromatic conjugated polymers are different. To the
contrary, both IS and QS values are little dependent
on the presence of alkyl substituents. These facts
facilitate the assignment of the Mo¨ssbauer lines. Figure
5 shows the Mo¨ssbauer spectrum of doped PDOBTF. It
can be deconvoluted into three doublets whose Mo¨ss-
bauer parameters are collected in Table 4. Two principal
doublets of equal intensity are characteristic of high-
Ch em ica l Dop in g a n d Mossba u er Sp ectr oscop y.
UV-vis-NIR spectroelectrochemical data unequivocally
show that the synthesized copolymer can be oxidatively
doped; however, they do not indicate whether both
subunits of the repeat unit, i.e., bithiophene and fluoren-
9-one, undergo the oxidation. To address this problem,
we have studied chemical doping of PDOBTF with
FeCl3, combining elemental analysis, Mo¨ssbauer spec-
spin Fe(III) complexes and can be ascribed to FeCl4 at
two different doping sites or, in other words, neutralized
by carbocations of different nature. We ascribe the
-

7052 Demadrille et al.
Macromolecules, Vol. 36, No. 19, 2003
Sch em e 2. P ostfu n ction a liza tion of P DOBTF w ith
4EB
F igu r e 5. Mo¨ssbauer spectrum registered at 4.2 K of a
-
PDOBTF thin film doped with FeCl4
.
Ta ble 4. Mo1ssba u er P a r a m eter s of P DOBTF Dop ed w ith
- a
F eCl4
IS (mm s-1
)
QS (mm s-1
)
W (mm s-1
)
Fe(III) site I
Fe(III) site II
Fe(II)
0.27 (2)
0.37 (2)
1.35 (2)
0.36 (2)
0.57 (2)
3.12 (2)
0.78 (2)
0.78 (2)
0.64 (2)
a
IS (isomer shift) relative to R-Fe. QS (quadrupole shift) )
1/2e2qQ, W ) line width. Errors are given in parentheses, in units
corresponding to the last digit of quoted values.
possibility of precise tuning of its spectroscopic proper-
ties by grafting of appropriate chromophores via the
carbonyl group present in the fluoren-9-one segment.
In our first attempt to functionalize PDOBTF, we have
selected aniline tetramer in the oxidation state of
emeraldine (4EB) as the grafted chromophore. The
presence of oligoaniline side groups not only influences
the UV-vis-NIR spectrum of the polymer but also
diversifies its electrochemical activity since in this case
both the conjugated main chain and oligoaniline side
chains are electroactive. Moreover, additional tuning of
spectroscopic properties is possible by selective doping
of the oligoaniline side chains via protonation.12
Oligoanilines can be conveniently attached to the
chain of PDOBTF using TiCl4-catalyzed coupling reac-
tion between the amine end group of the aniline tet-
ramer and the carbonyl group of PDOBTF (see Scheme
2). A very similar procedure37 has previously been
applied to the synthesis of fluorenimines from fluoren-
9-one and amines giving products in yields approaching
80%. In the case of the reaction between aniline tet-
ramer and PDOBTF high grafting levels are obtained
only after extended reaction times. The level of grafting
can be conveniently determined by elemental analysis
and in particular from the N/S ratio. An extended
reaction time, ca. 12 h, applied in this research leads
to a N/S molar ratio of 1.46, which corresponds to 73%
grafting level, taking into account that each repeat unit
of the polymer contains two sulfur atoms whereas each
grafted oligoaniline group contains four nitrogen atoms.
It should be noted here that “low grafting” level post-
functionalized PDOBTF (PDOBTF-4EB1) remains solu-
tion processable in classical solvents of the poly-
thiophene family of conjugated polymers (for example,
THF, CHCl3, etc.) while the highest grafting level
fraction (PDOBTF-4EB2: grafting level of ca. 73%) is
soluble in solvents dissolving aniline oligomers such as
N-methyl-2-pyrrolidinone (NMP).
doublet with lower isomer shift (IS) and quadrupole
shift (QS) values as originating from bithiophene doping
sites since their Mo¨ssbauer parameters are close (but
not identical) to those measured for poly(3-alkyl-
thiophene) homopolymers.35 The second Fe(III) doublet
-
can in turn be attributed to FeCl4 associated with
fluoren-9-one doping sites.36 Much larger line widths
measured for both doublets as compared to the case of
conjugated homopolymers doped with the same dopant
reflects the heterogeneous nature of the individual
copolymer chain as well as a rather complex supramo-
lecular structure of the doped polymer. Local changes
in the polymer chain conformation as well as the
coexistence of supramolecular aggregations of different
degree of structural order result, for both doping sites,
in a small distribution of the IS value around its mean
value and by consequence contribute to the line broad-
ening.
The third doublet of high isomer shift (IS ) 1.35 mm
s-1) and a high quadruple splitting (QS ) 3.12 mm s-1
)
values, which accounts for ca. 15% of the intensity of
the spectrum, is characteristic of Fe(II) and originates
from incomplete removal of the side-reaction product
(FeCl2). Such incomplete removal of FeCl2 is caused by
minute amounts of water present in the doping solution
which promote its precipitation within the polymer
matrix.
At the end it should be stated that, in principle, other
interpretations of the origin of the two observed prin-
cipal Fe(III) doublets are possible. The explanation
presented above seems to be the most plausible since it
is supported by complementary spectroscopic data and
is in accordance with Mo¨ssbauer spectroscopic data
obtained previously for the corresponding homopoly-
mers.
P ostp olym er iza tion F u n ction a liza tion : An ilin e
Tetr a m er Gr a ftin g to P DOBTF Ma in Ch a in . One
of the most interesting features of PDOBTF is the
In Figure 6a,b the IR spectrum of PDOBTF is
compared with that registered for PDOBTF-4EB2.

Macromolecules, Vol. 36, No. 19, 2003
Fluorenone-Alkylthiophene Copolymers 7053
F igu r e 7. UV-vis solution (THF) spectra of PDOBTF,
PDOBTF-4EB1 (“low grafting” level), and PDOBTF-4EB2
(grafting level 73%).
F igu r e 6. FTIR spectra (spectral range 500-1800 cm-1) of
PDOBTF (a) and postfunctionalized PDOBTF-4EB2 (grafting
level 73%) (b).
π-π* transition in the benzene ring. The band corre-
sponding to this transition merges with the band
characteristic of PDOBFT at 302 nm to give one broad
peak with the maximum at 304 nm. This absorption
peak strongly overlaps with the peak at 378 nm ascribed
to the π-π* transition in the conjugated main chain.
In the sample with the highest grafting level (73%)
(PDOBTF-4EB2), the peaks originating from electronic
transitions in the tetramer substituent dominate. The
band due to the π-π* transition in the benzene ring of
the tetramer merges in this case with the band origi-
nating from the π-π* transition in the main chain into
one broad peak located at 313 nm.
In addition to the expected increase of the intensity
of the bands originating from the tetramer with the
increasing grafting level, one should note their batho-
chromic shift by ca. 10 nm as compared to the analogous
bands registered, in the same experimental conditions,
for the “free” tetramer, i.e., the tetramer not attached
to the polymer chain. This shift unequivocally indicates
the conjugation between the π-systems of the side
groups and the main chain. Finally, the above-described
experiments clearly demonstrate that the grafting reac-
tion gives the opportunity of tuning the spectrum of the
polymer both in the UV and in the visible regions.
Apart from the appearance of a broad band in the
vicinity of 3340 cm-1 which is characteristic of N-H
stretching deformations (not shown here), one can notice
a significant decrease in the intensity of the band at
1717 cm-1 ascribed to CdO stretchings in the carbonyl
group, consistent with a rather high level of oligoaniline
grafting which results in the transformation of the
carbonyl linkages into the imine ones. Four most
intensive bands characteristic of aniline tetramer at
1588 cm-1, a doublet at 1494 and 1503 cm-1, and a peak
at 1291 cm-1 dominate in the spectrum of PDOBTF-
4EB2 (compare parts a and b of Figure 6). The bands
corresponding to the PDOBFT constituents, although
less intensive, are also clearly visible. These are bands
at 2916 and 2847 cm-1 (not shown here) originating
from the C-H stretching deformations in the dialkyl-
bithiophene units. Other bands characteristic of the
alkyl groups are present at 1471 and 1376 cm-1. The
peaks associated with the deformations of the conju-
gated main chain can also be found in the spectrum,
although in some cases they are obscured by the
presence of the bands originating from the aniline
tetramer deformations. This is for example the case of
the band at 824 cm-1 attributed to out-of-plane defor-
mations in the substituted thienylene ring which nearly
coincides with a band characteristic of aniline tetramer
Con clu sion s
at 818 cm-1
.
To summarize, we describe in detail three methods
of the preparation of a new regioregular conjugated
copolymer, namely poly[(5,5′-(3,3′-di-n-octyl-2,2′-bithio-
phene))-alt-(2,7-fluoren-9-one)] (PDOBTF), which com-
bines electronic properties of poly(fluoren-9-one-2,7-diyl)
homopolymer with solution processability of poly(alkyl-
thiophene)s and in addition can be easily functionalized
via carbonyl groups present in its main chain. SEC
results show that, independent of the polymerization
method (chemical or electrochemical oxidation of (2) or
polycondensation of (3)), the resulting crude product is
a mixture of high molecular weight fractions and short
chain oligomers. It can however be fractionated into
sharper fractions by sequential extraction. Upon p-
doping the polymer becomes a conductor with the
Grafting of aniline tetramer gives also rise to a
significant modification of the UV-vis spectrum of the
polymer (see Figure 7). In the spectrum of the polymer
with an intermediate grafting level (PDOBTF-4EB1) we
notice a new broad band peaked at 565 nm, which is
characteristic of aniline tetramer in the oxidation state
of emeraldine. This band is attributed to an excitonic-
type transition between the HOMO level of the quinone
ring and the LUMO level of the benzene ring. Consistent
with a partial grafting reaction, this peak is super-
imposed on the peak at 476 nm, present in PDOBTF
and ascribed to the n-π* transition of the carbonyl
group. In the polymer with aniline tetramer side groups
another electronic transition is also expected, namely

7054 Demadrille et al.
Macromolecules, Vol. 36, No. 19, 2003
conductivity σ ) 0.05 S cm-1. We also demonstrate that
postfunctionalization of PDOBTF consisting of aniline
tetramer grafting to the main chain is a convenient
method for tuning its spectroscopic properties.
(15) Guillerez, S.; Bidan, G. Synth. Met. 1998, 93, 123-126.
(16) Bau¨erle, P.; Pfau, F.; Schupp, H.; Wurthner, F.; Gaudl, K.
U.; Balparda Caro, M.; Fischer, P. J . Chem. Soc., Perkin
Trans. 2 1993, 489-494.
(17) Andersson, M. R.; Selse, D.; Berggren, M.; J arvinen, H.;
Hjertberg, T.; Inganas, O.; Wennerstro¨m, O.; Osterholm, J .
E. Macromolecules 1994, 27, 6503-6506.
(18) Yamamoto, T. Bull. Chem. Soc. J pn. 1999, 72, 621-638.
(19) Shiraishi, K.; Yamamoto, T. Synth. Met. 2002, 130, 139-147.
(20) Zhang, Z. B.; Fujiki, M.; Tang, H. Z.; Motonaga, M.; Torimitsu,
K. Macromolecules 2002, 35, 1988-1990.
(21) Yong-J oo Kim, R.; Sato, R.; Maruyama, T.; Osakada., K.;
Yamamoto, T. J . Chem. Soc., Dalton. Trans. 1994, 943-948.
(22) Liu, J .; Loewe, R. S.; McCullough, R. D. Macromolecules 1999,
32, 5777-5785.
(23) Trznadel, M.; Pron, A.; Zagorska, M.; Chrzaszcz, R.; Pieli-
chowski, J . Macromolecules 1998, 31, 5051-5058.
(24) Hourquebie, P.; Olmedo, L. Synth. Met. 1991, 65, 19-26.
(25) Ishikawa, H.; Amano, K.; Kobayashi, K.; Satoh, A.; Hase-
gawa, E. Synth. Met. 1994, 64, 49-52.
(26) See, for example: Ingana¨s, O. In Hanbook of Organic
Conductive Molecules and Polymers; Nalwa, H.; Ed.; Wiley
& Sons: New York, 1997; Vol. 3, p 785.
Ack n ow led gm en t . The authors acknowledge
ADEME and CEA for a partial funding (Convention de
recherche Nï. 0105148) through the CSPVP “Cellules
Solaires Photovolta¨ıques Plastiques” research program.
We also thank C. Lebrun for her assistance in MS-ESI
characterizations and helpful discussions.
Refer en ces a n d Notes
(1) Bao, Z. Adv. Mater. 2000, 12, 227-230.
(2) Kraft, A. C.; Grimsdale, A. C.; Holmes, A. B. Angew. Chem.,
Int. Ed. 1998, 37, 402-428.
(3) Yu, G.; Hummelen, J . C.; Wudl, F.; Heeger, A. J . Science
1995, 270, 1789-1791.
(4) Uckert, F.; Setayesh, S.; Mu¨llen, K. Macromolecules 1999,
32, 4519-4524.
(27) Panozzo, S.; Vial, J . C.; Kervella, Y.; Ste´phan, O. J . Appl.
(5) Liu, B.; Yu, W. L.; Lai, Y. H.; Huang, W. Macromolecules
2000, 33, 8945-8952.
Phys. 2002, 92, 3495-3502.
(28) Bliznyuk, V. N.; Carter, S. A.; Scott, J . C.; Kla¨rner, G.; Miller,
R. D.; Miller, D. C. Macromolecules 1999, 32, 361-369.
(29) List, E. J . W.; Guentner, R.; Scandiucci de Freitas, P.; Scherf,
U. Adv. Mater. 2002, 14, 374-378.
(6) Pei, J .; Yu, W. L.; Huang, W.; Heeger, A. J . J . Chem. Soc.,
Chem. Commun. 2000, 1631-1632.
(7) Wong, K. T.; Chien, Y. Y.; Chen, R. T.; Wang, C. F.; Lin, F.
T.; Chiang, H. H.; Hsieh, P. Y.; Wu, C. C.; Chou, C. H.; Su,
Y. O.; Lee, G. H.; Peng, S. M. J . Am. Chem. Soc. 2002, 124,
11576-11577.
(30) Zojer, E.; Pogantsch, A.; Hennebicq, E.; Beljonne, D.; Bre´das,
J . L.; Scandiucci de Freitas, P.; Scherf, U.; List, E. J . W. J .
Chem. Phys. 2002, 117, 6794-6802.
(31) Uckert, F.; Tak, Y. H.; Mu¨llen, K.; Ba¨ssler, H. Adv. Mater.
2000, 12, 905-908.
(32) Lapkowski, M.; Zagorska, M.; Kulszewicz-Bajer, I.; Pron, A.
J . Electroanal. Chem. 1992, 341, 151-163.
(33) Pron, A.; Zagorska, M.; Kucharski, Z.; Lukasiak, M.; Su-
walski, J .; Mater. Res. Bull. 1982, 17, 1505-1510.
(34) Greenwood, N. N.; Gibbs, T. C. In Mo¨ssbauer Spectrocopy;
Chapman and Hall: London, 1971.
(8) Brabec, C. J .; Sariciftci, N. S.; Hummelen, J . C. Adv. Funct.
Mater. 2001, 11, 15-26.
(9) Brabec, C. J .; Cravino, A.; Meissner, D.; Sariciftci, N. S.;
Fromherz, T.; Rispens, M. T.; Sanchez, L.; Hummelen, J . C.
Adv. Funct. Mater. 2001, 11, 374-380.
(10) Padinger, F.; Rittberger, R. S.; Sariciftci, N. S. Adv. Funct.
Mater. 2003, 13, 85-88.
(11) Feng, J .; Zhang, W.; MacDiarmid, A. G.; Epstein, A. J . Proc.
Soc. Plast. Eng., Annu. Technol. Conf. (ANTEC’97) 1997, 2,
1373-1377.
(35) Kulszewicz-Bajer, I.; Pron, A.; Suwalski, J .; Kucharski, Z.;
(12) Dufour, B.; Rannou, P.; Travers, J . P.; Pron, A.; Zagorska,
M.; Korc, G.; Kulszewicz-Bajer, I.; Quillard, S.; Lefrant, S.
Macromolecules 2002, 35, 6112-6120.
Lefrant, S. Synth. Met. 1989, 28, C225-C229.
(36) Pron, A.; Fatseas, G.; Krichene, S.; Lefrant, S.; Maurice, F.;
Froyer, G. Phys. Rev. B 1985, 32, 1839-1842.
(13) J ayakannan, M.; Van Dongen, J . L.; J anssen, R. A. Macro-
molecules 2001, 34, 5386-5393.
(37) Dai, W.; Srinivasan, R.; Katzenellenbogen, J . A. J . Org. Chem.
1989, 54, 2204-2208.
(14) Asawapirom, U.; Gu¨nter, R.; Forster, M.; Farrell, T.; Scherf,
U. Synthesis 2002, 9, 1136-1142.
MA030171Z