On the degradation process involving polyfluorenes and the factors governing their spectral stability

Article abstract of DOI:10.1021/ma2015003

This study deals with an investigation of the spectral stability of differently structured polyfluorenes (PFs), deprived of 9-H defects, embodying 9,9-dialkylfluorene (P1), 9,9-diarylfluorene (P2), or 9,9-diarylfluorene/9,9- dibenzylfluorene units in a 1:1 alternating fashion (P3). Thermal annealing or UV irradiation carried out on films of P1-P3 in air revealed that their typical blue photoluminescence is invariably stained, independently of their 9-substitution, by the appearance of the low-energy band (g-band) pointing out a remarkable effect of light on the degradation process. A more comprehensive picture of the degradation pathway is proposed, including as key step a light-promoted formation of a PF radical cation generated by aerobic oxidation (photoluminescence test) or p-doping (cyclic voltammetry test). The blue emission of P1-P3 could successfully be preserved by dispersing them into a higher band gap matrix, such as polyvinylcarbazole (PVK), indicating a fundamental role of the intermolecular interactions between PF chains in the appearance of the low-energy emission band. Comparison between the optical behavior of suitably prepared PFs containing either fluorenone moieties (PFK) or 9-(bis-methylsulfanyl-methylene)fluorene moieties (PFS) holds regions of planarity within the PF backbone (inducing local intermolecular interactions) and not the fluorenone charge-transfer emission as responsible of the g-band of degraded PFs.

Full text of DOI:10.1021/ma2015003

ARTICLE
pubs.acs.org/Macromolecules
On the Degradation Process Involving Polyfluorenes and the Factors
Governing Their Spectral Stability
Roberto Grisorio, Giovanni Allegretta, Piero Mastrorilli,* and Gian Paolo Suranna
Department of Water Engineering and of Chemistry, Polytechnic of Bari, Campus Universitario, via Orabona 4, 70125 Bari, Italy
ABSTRACT: This study deals with an investigation of the
spectral stability of differently structured polyfluorenes (PFs),
deprived of 9-H defects, embodying 9,9-dialkylfluorene (P1),
9,9-diarylfluorene (P2), or 9,9-diarylfluorene/9,9-dibenzylfluorene
units in a 1:1 alternating fashion (P3). Thermal annealing or
UV irradiation carried out on films of P1ÀP3 in air revealed that
their typical blue photoluminescence is invariably stained,
independently of their 9-substitution, by the appearance of
the low-energy band (g-band) pointing out a remarkable effect
of light on the degradation process. A more comprehensive
picture of the degradation pathway is proposed, including as key
step a light-promoted formation of a PF radical cation generated by aerobic oxidation (photoluminescence test) or p-doping (cyclic
voltammetry test). The blue emission of P1ÀP3 could successfully be preserved by dispersing them into a higher band gap matrix,
such as polyvinylcarbazole (PVK), indicating a fundamental role of the intermolecular interactions between PF chains in the
appearance of the low-energy emission band. Comparison between the optical behavior of suitably prepared PFs containing either
fluorenone moieties (PFK) or 9-(bis-methylsulfanyl-methylene)fluorene moieties (PFS) holds regions of planarity within the PF
backbone (inducing local intermolecular interactions) and not the fluorenone charge-transfer emission as responsible of the g-band
of degraded PFs.
’ INTRODUCTION
has concentrated on the synthesis of PFs functionalized with
bulky substituents in order to hamper intermolecular interac-
tions, thereby limiting the evolution of the g-band.6 It is worth
noting that the PF functionalization with bulky substituents can
be successful also under the alternative hypothesis explaining the
g-band emergence in terms of an emission from fluorenone-
based excimers,7 rather than from on-chain defects.
Parallel to the search for the origin of the g-band, the mechanisms
of PFs oxidative degradation, which are equally important in order
to rationalize the spectral instability of this class of polymers, have
also been extensively studied. Oxidation to 9-fluorenone is a
favorable process due to the resonance stabilization deriving
from the extension of the π-conjugated system. The preferred
sites of oxidation in PFs have been individuated in their 9-H
defects (i.e., the fluorene units embodying a hydrogen atom at
the C-9 position) invariably present in the polymer as a con-
sequence of an incomplete purification of the dialkylfluorene
monomer feed.8 To circumvent this drawback, two main strate-
gies have been proposed: (i) the thorough purification of the
fluorene monomers9 or (ii) the employment of specifically
designed synthetic approaches that avoid any fluorene alkylation
step to obtain fully C-9 disubstituted fluorene units.10 However,
the appearance of the g-band has been recently reported also for
defect-free oligofluorenes,11 highlighting that the great advantages
During the past years, the research interest concerning poly-
fluorenes (PFs)1 has produced numerous studies focused on the
appearance, under oxidative conditions, of an unwanted green
emission (the so-called “g-band”) at 520À535 nm accompanying
their typical blue emission. An intense research impulse has
therefore been directed to understand the origin of this spectral
instability of fluorene-based materials, with the aim of suppres-
sing it to successfully employ PFs as blue emitters in OLED
devices. After a decade-long debate, it is now generally accepted
that the appearance of the g-band is correlated with the formation
of 9-fluorenone units in the polymer backbone,2 though the exact
role of these units, named keto defects, remains so far not
completely clear. The hypothesis according to which the fluor-
enone unit constitutes a potential green-emitting site (on-chain
emission) has been supported by theoretical calculations3 asso-
ciating the low-energy emission band to a charge-transfer (CT)
state deriving from the presence of carbonyl moieties.4 On the
basis of these assumptions, the g-band emission intensity would
mainly be governed by the ease with which the fluorenone excited
states can be populated by energy transfer. However, the sole
formation of fluorenone units in the course of an oxidative degra-
dation of a PF is not sufficient to cause the appearance of the
g-band. For instance, it has been noted that the g-band is
suppressed in fluorenone-containing polymers bearing bulky
substituents. This observation was attributed to the frustration
of the interchain packing between PF units, hindering an efficient
energy transfer toward fluorenone emitters.5 Thus, a lively research
Received: July 1, 2011
Revised:
August 30, 2011
Published: September 26, 2011
r
2011 American Chemical Society
7977
dx.doi.org/10.1021/ma2015003 Macromolecules 2011, 44, 7977–7986
|

Macromolecules
ARTICLE
associated with the presence of the highly reactive C-9 on
fluorene (e.g., its easy functionalization) unfortunately also
represent the weakness point of PFs, indicating the necessity
of an ultimate effort for the rationalization of the degradation
mechanisms involving fluorene-based materials.
NMR (101 MHz, CDCl3): δ 152.6, 139.1, 130.2, 126.2, 121.5, 121.1,
55.7, 40.2, 31.5, 29.6, 23.7, 22.6, 14.0.
2,7-Bis(4,4,5,5-tetramethyl-1,3,2-dioxaborolan-2yl)-9,9-
dihexylfluorene (2). To a solution of 1 (4.00 g, 8.12 mmol) in THF
(80 mL) kept at À80 °C, a solution of n-BuLi (1.6 M in hexanes,
15.2 mL, 24.36 mmol) was added dropwise. The obtained mixture was
vigorously stirred for 1 h at À80 °C before the addition of 2-isopropoxy-
4,4,5,5-tetramethyl-1,3,2-dioxaborolane (4.53 g, 24.36 mmol) in one
portion. The resulting solution was then allowed to reach room
temperature and to react for further 4 h. After removing the solvent
under vacuum, CH2Cl2 (50 mL) was added, and the obtained solution
was washed with water (3 Â 50 mL) and dried over Na2SO4. After
removing the solvent under vacuum, the obtained crude product was
purified by flash column chromatography (SiO2, petroleum ether
40À60 °C/CH2Cl2 = 2/1) to give 2 (3.55 g, 74%) as a white solid.
1H NMR (400 MHz, CDCl3): δ (ppm) 7.85À7.79 (m, 2H), 7.78À7.70
(m, 4H), 2.06À1.97 (m, 4H), 1.40 (s, 24H), 1.15À0.95 (m, 12H), 0.76
(t, J = 7.3 Hz, 6H), 0.61À0.50 (m, 4H). 13C{1H} NMR (101 MHz,
CDCl3): δ (ppm) 150.5, 143.9, 133.7, 128.9, 119.4, 83.8, 55.2, 40.1,
31.5, 29.7, 24.9, 23.6, 22.6, 14.1.
2,7-Dibromo-9,9-dibenzylfluorene (3). A suspension of 2,7-
dibromofluorene (1.00 g, 3.09 mmol) and tetrabutylammonium bro-
mide (1.61 g, 5.00 mmol) in a 50 wt % NaOH aqueous solution (20 mL)
was stirred for 15 min at 60 °C. To the red suspension, benzyl bromide
(6.18 g, 32.00 mmol) was added, and the mixture was kept under stirring
overnight. After cooling the solution, the product was extracted with
diethyl ether (3 Â 75 mL) and the organic layer dried over Na2SO4.
After removing the solvent under vacuum, the crude product was
submitted to flash chromatography (SiO2, petroleum ether 40À60 °C)
to afford 3 (1.12 g, 72%) as a white solid. 1H NMR (400 MHz, CDCl3):
δ (ppm) 7.53 (s, 2H), 7.34 (d, J = 8.2 Hz, 2H), 7.18 (d, J = 8.2 Hz, 2H),
7.06À6.95 (m, 6H), 6.68 (d, J = 6.3 Hz, 4H), 3.34 (s, 4H). 13C{1H}
NMR (101 MHz, CDCl3): δ (ppm) 150.2, 138.8, 136.1, 130.4, 130.2,
128.0, 127.4, 126.3, 121.2, 120.6, 57.2, 45.2.
In the course of the years, the original motivation for the
studies on PFs spectral instability (their use as blue OLEDs) has
gradually shifted toward the mandatory need to explore more
generally the fate of an organic semiconductor under p-doping
conditions which, in perspective, is of transversal interest for any
organic electronics application, including solar cells and field
effect transistors. On these basis, we decided to study the spectral
stability of suitably chosen PFs endowed with different C-9
functionalization focusing on the degradation mechanisms and
on the role of intermolecular interactions in PFs spectral instability.
We have addressed our attention on three PFs embodying 9,9-
dialkyl-, 9,9-diaryl-, or 9,9-dibenzylfluorene units. The polymers
were prepared starting from dibromofluorene monomers thor-
oughly purified from their 9-H defects, in order to correlate the
spectral behavior solely to the fluorene C-9 substitution. This study
allowed us to sketch new pathways for the degradation of PFs and to
suggest a pivotal role of intermolecular interactions between adjacent
backbones in the appearance of the low-energy emission band.
’ EXPERIMENTAL SECTION
All syntheses were carried out under an inert nitrogen atmosphere
using Schlenk techniques. All solvents were carefully dried and freshly
distilled. All reactants were purchased by commercial sources and used
without further purifications. 1H and 13C{1H} NMR spectra were
recorded at 295 K on a Bruker Avance 400 MHz spectrometer; chemical
shifts are reported in ppm referenced to SiMe4. UVÀvis spectra were
recorded on a Jasco V-670 instrument, and fluorescence spectra were
obtained on a Varian Cary Eclipse spectrofluorimeter. To prepare the
polymer films submitted to spectral stability test in the solid state, the
materials were deposited on quartz substrates by spin-coating (2000 rpm) a
drop of 5 a mg/mL solution of the polymer in CHCl3. UV-photo-
decomposition experiments were carried out by irradiating the sample
with a 150 W high-pressure Hg lamp for 30 min. Thermogravimetric
analyses (TGA) were carried out on a Perkin-Elmer Pyris TGA 6
thermobalance. GPC analyses were carried out on an Agilent Series 1100
instrument equipped with a Pl-gel 5 μm mixed-C column. THF solutions
for GPC analysis were eluted at 25 °C at a flow rate of 1.0 mL/min and
analyzed using a multiple wave detector. Molecular weights and molecular
weight distributions are relative to polystyrene. Cyclic voltammetry
(CV) was carried out under an inert nitrogen atmosphere with an
Autolab PGSTAT 100 potentiostat using a three-electrode cell consist-
ing of a platinum disk as a working electrode, a platinum wire as counter
electrode, and a Ag/Ag+ electrode as pseudoreference electrode. The
CV measurements were carried out in dry acetonitrile solutions of
tetrabutylammonium tetrafluoroborate (0.10 M).
2,7-Dibromo-9,9-dihexylfluorene (1). A suspension of 2,7-
dibromofluorene (10.93 g, 33.73 mmol) and tetrabutylammonium bro-
mide (3.62 g, 11.25 mmol) in a 50 wt % NaOH aqueous solution (75 mL)
was stirred for 15 min at 60 °C. To the red suspension, n-hexyl bromide
(11.69 g, 70.84 mmol) was added, and the mixture was kept under
stirring overnight. After cooling the solution, the product was extracted
with diethyl ether (3 Â 75 mL) and the organic layer dried over Na2SO4.
After removing the solvent under vacuum, the crude product was submitted
to flash chromatography (SiO2, petroleum ether 40À60 °C) to afford 1
in 92% yield as white solid. 1H NMR (400 MHz, CDCl3): δ 7.54 (d, J =
8.9 Hz, 2H), 7.50À7.45 (m, 4H), 1.99À1.91 (m, 4H), 1.21À1.02
(m, 12H), 0.83 (t, J = 7.0 Hz, 6H), 0.68À0.58 (m, 4H). 13C{1H}
2,7-Dibromo-9,9-bis(4-hydroxyphenyl)fluorene (4). A mix-
ture of 2,7-dibromofluoren-9-one12 (7.05 g, 20.9 mmol), phenol (19.7 g,
0.209 mol), and methansulfonic acid (30 mL) was heated to 50 °C
overnight. After cooling to room temperature, the mixture was poured
into water. The obtained solid was filtered and washed with water several
times. The desired product 4 (9.63 g, 91%) was obtained by precipita-
1
tion from ethyl acetate into petroleum ether as a brownish solid. H
NMR (400 MHz, CDCl3): δ 7.71 (d, 2H), 7.49 (dd, 2H), 7.44 (d, 2H),
6.94 (m, 4H), 6.68 (m, 4H). 13C{1H} NMR (101 MHz, CDCl3): δ (ppm)
158.2, 153.8, 137.9, 136.3, 130.7, 129.3, 129.0, 121.8, 121.6, 114.3, 64.4.
2,7-Dibromo-9,9-bis(4-hexyloxyphenyl)fluorene (5). A mix-
ture of 4 (8.40 g, 16.5 mmol), NaOH (1.44 g, 36.3 mmol), 1-bromohexane
(6.00 g, 36.3 mmol), and THF (30 mL) was stirred at 50 °C overnight.
After cooling the solution down to room temperature, diethyl ether
(60 mL) was added. The organic phase was washed with water (3 Â
60 mL) and dried over Na2SO4. After solvent removal, the crude product
was purified by flash chromatography (SiO2, petroleum ether 40À60 °C/
CH2Cl2 = 4/1 v/v) yielding 5(4.20 g, 38%) as a colorless oil. 1HNMR(400
MHz, CDCl3): δ (ppm) 7.56 (d, J = 8.5 Hz, 2H), 7.48À7.43 (m, 4H), 7.05
(d, J = 8.5 Hz, 4H), 6.77 (d, J = 8.5 Hz, 4H), 3.90 (t, J = 6.7 Hz, 4H), 1.75
(m, J = 6.7 Hz, 4H), 1.48À1.27 (m, 12H), 0.93À0.86 (m, 6H). 13C{1H}
NMR (101 MHz, CDCl3): δ (ppm) 158.2, 153.8, 137.9, 136.3, 130.7,
129.3, 129.0, 121.8, 121.6, 114.3, 68.0, 64.4, 31.6, 29.2, 25.8, 22.6, 14.1.
2,7-Bis(4,4,5,5-tetramethyl-1,3,2-dioxaborolan-2yl)-9,9-
bis(4-hexyloxyphenyl)fluorene (6). To a solution of 5 (2.97 g,
4.39 mmol) in THF (80 mL) kept at À80 °C, a solution of n-BuLi (1.6 M
in hexanes, 8.2 mL, 13.17 mmol) was added dropwise. The obtained
mixture was vigorously stirred for 1 h at À80 °C before the addition of
2-isopropoxy-4,4,5,5-tetramethyl-1,3,2-dioxaborolane (2.45 g, 13.17 mmol)
in one portion. The resulting solution was then allowed to reach room
7978
dx.doi.org/10.1021/ma2015003 |Macromolecules 2011, 44, 7977–7986

Macromolecules
ARTICLE
Scheme 1. Synthesis of P1ÀP3a
a Reagents and conditions: (i) NaOHaq (50 wt %), TBAB, n-hexyl bromide, 60 °C; (ii) n-BuLi, 2-isopropoxy-4,4,5,5-tetramethyl-1,3,2-dioxaborolane,
À80 °C; (iii) NaOH(aq) (50 wt %), TBAB, benzyl bromide, 60 °C; (iv) phenol, CH3SO3H, 80 °C; (v) n-hexyl bromide, NaOH, 50 °C; (vi) Pd(PPh3)4,
Na2CO3 2 M, reflux.
temperature and to react for further 4 h. After removing the solvent
under vacuum, CH2Cl2 (50 mL) was added, and the obtained solution
was washed with water (3 Â 50 mL) and dried over Na2SO4. After
removing the solvent under vacuum, the obtained crude product was
purified by flash column chromatography (SiO2, petroleum ether 40À
60 °C/CH2Cl2 = 1/1 v/v) to give 6 (1.05 g, 31%) as a white solid. 1H
NMR (400 MHz, CDCl3): δ (ppm) 7.78À7.74 (m, 6H), 7.14 (d, J = 8.5
Hz, 4H), 6.75 (d, J = 9.5 Hz, 4H), 3.90 (t, J = 6.7 Hz, 4H), 1.75 (m, J = 6.7
Hz, 4H), 1.50À1.25 (m, 36H), 0.91 (t, J = 7.3 Hz, 6H). 13C{1H} NMR
(101 MHz, CDCl3): δ (ppm) 157.7, 151.9, 142.6, 137.7, 134.1, 132.2,
129.4, 119.8, 114.0, 83.7, 67.9, 64.2, 31.6, 29.3, 25.8, 24.9, 22.6, 14.1.
Poly[2,7-(9,9-dihexylfluorene)] (P1). A mixture of 1 (123 mg,
0.25 mmol), 2 (147 mg, 0.25 mmol), and Pd(PPh3)4 (6 mg, 0.5 Â 10À2
mmol) in a 2.0 M Na2CO3 aqueous solution (2.5 mL) and toluene
(4.5 mL) was refluxed for 72 h. After cooling the solution, the reaction
mixture was poured into a beaker containing methanol (200 mL),
causing the precipitation of the polymer which was filtered off, washed
with methanol, and reprecipitated with methanol. The obtained solid
was then dried under vacuum at 60 °C for 24 h to give P1 as an off-white
powder in 69% yield. GPC (THF): Mn = 12 600 Da, Mw = 24 300 Da,
D = 1.9. 1H NMR (400 MHz, CDCl3): δ (ppm) 7.91À7.81 (m, 2H),
7.77À7.62 (m, 4H), 2.25À1.99 (m, 4H), 1.25À1.05 (m, 12H),
0.92À0.72 (m, 10H).
starting from 5 and 6 in 71% yield as an off-white solid. GPC (THF):
Mn = 12 000 Da, Mw = 28 400 Da, D = 2.4. 1H NMR (400 MHz, CDCl3):
δ (ppm) 7.81À7.71 (m, 2H), 7.63À7.47 (m, 4H), 7.24À7.11 (m, 4H),
6.82À6.72 (m, 4H), 3.98À3.87 (m, 4H), 1.81À1.70 (m, 4H),
1.49À1.38 (m, 4H), 1.37À1.26 (m, 8H), 0.94À0.85 (m, 6H).
Poly{2,7-[9,9-bis(4-hexyloxyphenyl)fluorene]-alt-2,7-(9,9-
dibenzylfluorene)} (P3). Following the procedure reported for P1,
this polymer was synthesized starting from 6 and 3 in 46% yield as an off-
white solid. GPC (THF): Mn = 15 000 Da, Mw = 36 000 Da, D = 2.4. 1H
NMR (400 MHz, CDCl3): δ (ppm) 7.95 (m, 2H), 7.75À7.63 (m, 4H),
7.57À7.45 (m, 6H), 7.36À7.29 (m, 4H), 7.01À6.78 (m, 14H),
3.98À3.87 (m, 4H), 3.60À3.43 (m, 4H), 1.81À1.70 (m, 4H),
1.49À1.38 (m, 4H), 1.37À1.26 (m, 8H), 0.94À0.85 (m, 6H).
Poly[2,7-(9,9-dihexylfluorene)-co-2,7-(fluoren-9-one)] (PFK).
Following the procedure reported for P1, this polymer was synthesized
starting from 2 (mole fraction = 0.50), 1 (mole fraction = 0.45), and 2,7-
dibromofluoren-9-one (mole fraction = 0.05) in 69% yield as a pale yellow
powder. GPC(THF):Mn =16000 Da, Mw = 38 000 Da, D=2.4) 1HNMR
(400 MHz, CDCl3): δ (ppm) 7.91À7.81 (m, 2H), 7.77À7.62 (m, 4H),
2.25À1.99 (m, 4H), 1.25À1.05 (m, 12H), 0.92À0.72 (m, 10H).
Poly{2,7-(9,9-dihexylfluorene)-co-2,7-[9-(bis-methylsulfanyl-
methylene)fluorene]} (PFS). Following the procedure reported for
P1, this polymer was synthesized starting from 2 (mole fraction = 0.50),
1 (mole fraction = 0.45), and 2,7-dibromo-9-(bis-methylsulfanylmethy-
lene)fluorene (mole fraction = 0.05) in 51% yield as a pale yellow
Poly{2,7-[9,9-bis(4-hexyloxyphenyl)fluorene]} (P2). Fol-
lowing the procedure reported for P1, this polymer was synthesized
7979
dx.doi.org/10.1021/ma2015003 |Macromolecules 2011, 44, 7977–7986

Macromolecules
ARTICLE
powder. GPC (THF): Mn = 10 500 Da, Mw = 23 500 Da, D = 2.2. 1H
NMR (400 MHz, CDCl3): δ (ppm) 7.96À7.83 (m, 2H), 7.77À7.49 (m,
10H), 7.22À6.67 (m, 10H), 3.60À3.43 (m, 4H), 2.28À2.11 (m, 4H),
1.30À1.13 (m, 12H), 0.91À0.73 (m, 10H).
’ RESULTS AND DISCUSSION
Polymer Synthesis and Characterization. The polymers
chosen to compare the effect of different C-9 substitutions on
the spectral behavior embodied (i) n-hexyl chains, which are
easily subject to radical degradation that ends up in fluorene C-9
oxidation; (ii) 4-hexyloxyphenyl chains, as bulky substituents
which also constitute a barrier to the radical C-9 oxidation; (iii)
4-hexyloxyphenyl and benzyl chains, the latter of which are bulky
substituents albeit easily subject to radical degradation. The three
PFs (Scheme 1) were thus the homopolymers poly[2,7-(9,9-di-
n-hexylfluorene)] (P1) and poly{2,7-[9,9-bis(4-hexyloxyphenyl)-
fluorene]} (P2) as well as the alternating copolymer poly{2,7-[9,9-
bis(4-hexyloxyphenyl)fluorene]-alt-2,7-(9,9-dibenzylfluorene)}
(P3). The obtainment of 2,7-dibromo-9,9-di-n-hexylfluorene (1)
and 2,7-dibromo-9,9-dibenzylfluorene (3) was achieved by a
straightforward alkylation of 2,7-dibromofluorene with n-hexyl
bromide or benzyl bromide, respectively. In order to evaluate the
effect of the C-9 substitution on the spectral stability of P1ÀP3
excluding any possible contribution of 9-H defects, care was
taken to obtain the above-mentioned polymers deprived of such
defects. To this purpose, monomers 1 and 3 were thoroughly
purified following Meijer procedure.9 After the standard column
chromatography, treatment of a dry THF solution of 1 and 3
(0.10 g/mL) with an excess of a strong base (potassium tert-
butoxide) promoted the abstraction of the 9-H of monoalkylated
fluorenes, testified by the strongly red coloration of the solution
due to the formation of the fluorenyl anions. The obtained
mixture was then purified by column chromatography using
neutral alumina as stationary phase. This procedure was repeated
until the color of the solution after the addition of the base
remained unaltered, which can be considered as a qualitative test
for the absence of 9-H fluorene defects in the monomer feed.13
The monomer 2,7-dibromo-9,9-bis(4-hexyloxyphenyl)fluor-
ene (5) was prepared starting from 2,7-dibromofluorenone by a
double FriedelÀCrafts reaction with phenol (Scheme 1) fol-
lowed by an alkylation with n-hexyl bromide. The synthetic
procedure followed for the introduction of the aryl groups at the
fluorene C-9 position did not leave 9-H defects in the target
monomer 5, as confirmed by the negative result of the “tert-
butoxide test” on this compound. The boronic esters 2 and 6
were prepared by metalation of the corresponding 2,7-dibromo
derivatives, followed by reaction with 2-isopropoxy-4,4,5,5-tetra-
methyl-1,3,2-dioxaborolane.
Figure 1. TGA plots of P1ÀP3.
Figure 2. UVÀvis spectra of P1ÀP3 recorded in chloroform and as
thin film on quartz.
position accelerates the thermal degradation of the corresponding
polymer, which is likely triggered by the homolitic C9ÀCH2Ph
bond cleavage.
The optical properties of the polymers were studied both in
solution and in the solid state. In chloroform solution, the UVÀvis
spectra of P1ÀP3 exhibited a maximum in the range 379À
390 nm (Figure 2) while, in the solid state, their absorption
maxima are slightly blue-shifted (in the range 378À381 nm) with
respect to those observed in solution seemingly due to a back-
bone torsion, limiting the conjugation extension, or as the result
of aggregation.
Concerning their photoluminescence (PL), CHCl3 solutions
of P1ÀP3 showed emission wavelengths ranging between 415
and 417 nm, independently of their structure (Figure 3). In the
solid state, their emission relevant to the S1ÀS0 transitions
(421À425 nm) were not remarkably red-shifted with respect
to those recorded in solution.
Spectral Stability of the Polymers. The spectral behavior
toward aerobic thermal stress was investigated by studying the
solid-state PL spectra of P1ÀP3 after a thermal annealing in air at
130 °C for 2 h (Figure 4). After the annealing, the emission
profile of the dialkylfluorene P1 was only slightly modified by the
appearance of a g-band of very low intensity, while the thermal
annealing in air did not modify the emission profile of the fully
arylated P2. On the contrary, in the case of P3 a quite intense
The polymers P1ÀP3 were obtained by a Suzuki polycon-
densation in 46À71% yields starting from the suitable diboronic
esters and dibromofluorenes, as shown in Scheme 1. Their
number-average molecular weights were estimated by GPC in
the range 12 000À15 000 Da with polydispersities ranging from
1.9 to 2.4.
The thermal stability of the polymers was investigated by
thermogravimetric analysis (TGA) carried out under a nitrogen
atmosphere. The incipient decomposition temperature, corre-
sponding to the 5% weight loss, was recorded at 413, 423, and
361 °C for P1, P2, and P3, respectively (Figure 1). The lower
thermal stability observed for P3 with respect to P1 and P2
suggests that the presence of the benzyl groups at the C-9
7980
dx.doi.org/10.1021/ma2015003 |Macromolecules 2011, 44, 7977–7986

Macromolecules
ARTICLE
Figure 5. PL spectra of P2 after thermal annealing in air at 130 °C for 24
and 48 h in daylight.
Figure 3. PL spectra of P1ÀP3 recorded in chloroform solution and as
thin film on quartz.
Figure 6. PL spectra of P1ÀP3 after UV exposure in air for 30 min.
spectral degradation of P2 started after 24 h and became evident
after 48 h (Figure 5). This result points out an important
conclusion: notwithstanding a scale of stability can be sketched
for differently substituted fluorene units (9,9-diarylfluorene >
9,9-dialkyfluorene > 9,9-dibenzylfluorene), the results collected
so far point out that PFs are intrinsically susceptible of aerobic
oxidation, albeit with different rate, whatever the substitution at
the C-9 sites.
In order to study the influence of light on the degradation of
the PFs, the spectral stability of P1ÀP3 was also investigated by
submitting polymer films to UV photodecomposition in air for
30 min. The corresponding PL spectra (Figure 6) showed the
formation of a g-band of equal intensity for P1 and P2, suggesting
that under these conditions the kinetics of degradation is
independent of the alkyl or aryl substitution. Under these
conditions, a more intense g-band was again exhibited by P3.
Furthermore, since the g-band intensity shown by the poly-
mers after a UV irradiation was invariably higher with respect to
that caused by a thermal annealing, it can be stated that UV
irradiation accelerates the degradation processes of the fluorene
units. To gain insight into the influence of light on the PF
degradation, a thermal annealing (130 °C, 2 h) in air was carried
out on P3 (the material that exhibited the most intense g-band
under the same conditions) either in the dark or in daylight.
Figure 7 shows that when the aerobic thermal annealing was
carried out in daylight, a strong g-band was present in the PL
spectrum, while such a g-band was almost completely suppressed
Figure 4. PL spectra of P1ÀP3 after thermal annealing in air (130 °C,
2 h) in daylight.
g-band was observed. The same thermal annealing carried out
under a nitrogen atmosphere did not determine any change of
the absorption and emission profiles of the annealed films. As
stated above, since P1ÀP3 are deprived of 9-H defects, the
difference in behavior between P1 and P2 must be explained
invoking that the different C-9 substitution controls the evolu-
tion of the g-band for the corresponding polymer. In fact, while
for P1 the presence of alkyl side chains susceptible of oxidation
could favor the formation of fluorenones through well-estab-
lished degradation pathways,13,14 in the case of P2 the steric
hindrance of the aryl pendant groups at the C-9 position may
exert an influence in frustrating the intermolecular packing, thereby
limiting the efficiency of the energy transfer toward fluorenone
defects possibly formed during the aerobic oxidative stress.
Interestingly, the appearance of the g-band observed for P3
after annealing points out that the presence of benzyl substitu-
ents (which are bulky but, differently from 4-hexyloxyphenyl
ones, susceptible of radical attack) results in a rapid degradation
of the emission profile, suggesting that the steric hindrance of the
substituents does not play a central role in the issue of spectral
stability of PFs.
The spectral stability observed for P2 was not indefinite: in
fact, by prolonging the thermal annealing at 130 °C in air, the
7981
dx.doi.org/10.1021/ma2015003 |Macromolecules 2011, 44, 7977–7986

Macromolecules
ARTICLE
when the same aerobic thermal annealing was carried out in the
absence of a light source. This finding confirms a strong influence
of light on the PF degradation.
dialkylfluorene units can be hypothesized as described in the
literature.13,14
Under the harsh conditions of the aerobic thermal or radiative
stress, the generation of radical species initiates the mechanism
responsible for the degradation of the aliphatic side chains of the
fluorene units, as shown in pathway A of Scheme 2. Although a
radical could be able to attack any CÀH bond of the hexyl chain
of P1, causing its step-by-step degradation,11 the formation of a
radical at the Cα position followed by reaction with dioxygen can
rapidly lead to fluorenone units. The attack of radicals to the Cα
position of the side chains may also explain the fact that P3 is
more degradable with respect to P1 and P2 due to the mesomeric
stabilization of the formed benzylic radical (pathway B of
Scheme 2). However, this degradation pathway could not explain
the appearance of the g-band in polymers deprived of side-
degradable chains such as P2, for which the cleavage of the C(9)
sp3ÀC sp2 bond can be the only possible starting point for the
generation of fluorenones.
Mechanisms of the Degradation Process. The following
preliminary considerations on the degradation process of PFs
can be formulated: (i) the similar behavior of P1 and P2 upon
UV irradiation suggests that the mechanism should consider an
initiation step which takes place independently of the type of
substituent at the fluorene C-9 position; (ii) the aerobic degrada-
tion of PFs is strongly influenced by the presence of a light
source.
In the case of P1, the polymer functionalized with aliphatic
side chains, a radical mechanism for the degradation of 9,9-
A plausible degradation route for P2 which takes into account
the effect of light is reported in pathway C of Scheme 2. After the
absorption of light, at the excited state and in the presence of
hydrated dioxygen,15 the conjugated segment undergoes an
electron abstraction, forming a radical cation. The conjugated
system then rearranges to form the more stable benzyl radical
species by elimination of the C-9 substituent as a cation. The
fluorenyl radical, endowed with a high mesomeric stabilization
for the presence of an aryl substituent at the C-9 position of
the fluorene group, can be involved in a further oxidation to
fluorenone.
A similar rearrangement of a radical cation onto a π-conju-
Figure 7. PL spectra of P3 after thermal annealing in air at 130 °C for
2 h in the dark and in daylight.
gated system has been proposed for thiophene-based materials.16
Scheme 2. Proposed Degradation Pathways of Polyfluorenes
7982
dx.doi.org/10.1021/ma2015003 |Macromolecules 2011, 44, 7977–7986

Macromolecules
ARTICLE
Figure 8. Cyclic voltammetry of P1ÀP3. For each sample, two con-
Figure 10. UVÀvis spectra of a film of P1 on ITO before and after five
secutive oxidation cycles are reported.
anodic scans in acetonitrile.
Figure 11. PL spectra of P1ÀP3 in blend with PVK (10 wt %) after UV
treatment in air for 30 min.
Figure 9. Cyclic voltammetry of P1. Five consecutive oxidation cycles
are reported.
By contrast, in the case of P2 and P3, only an irreversible
anodic event was observed and the oxidation process completely
degraded the two polymers, since no electrochemical events were
observed during the second scan on the same sample. It is
reasonable to suppose that the anodic scan causes several con-
secutive electrochemical processes, inducing irreversible chemi-
cal modifications, one of which is a σ-bond cleavage involving the
C-9 position of the arylated fluorene units. A further investigation
on the electrochemical behavior of P1 was performed by carrying
out the same voltammetric scans at potentials lower than 1.70 V,
using as working electrode an ITO sheet on which P1 was drop-
casted, in order to measure the UVÀvis spectrum before and
after the scans, as shown in Figure 10.
This experiment allowed us to draw some conclusions about
the fate of the formed radical cation. In fact, the recovered films of
P1 after the voltammetry cycles were almost completely inso-
luble in chloroform. Furthermore, the UVÀvis spectrum of the
film showed a remarkable (∼20 nm) blue shift of the absorption
maximum. From these observations, it can reasonably be hy-
pothesized that after the formation of a radical at the C-9 position, in
the absence of dioxygen, this radical attacks the p-conjugated
backbone (pathway D of Scheme 2) resulting in a cross-linking of
the polymer chains, leading to a backbone distortion (causing the
blue shift of the absorption maximum) and to the formation of an
insoluble film. Furthermore, no solid-state PL of the resulting
It is worth remarking that the degradation pathway described so
far for P2 can occur also for P1 and P3. Since these findings may
be of considerable importance in order to rationalize possible
degradation routes of functionalized PFs upon positive charging
(p-doping), we have carried out an electrochemical investigation
aimed at gaining more insight into the behavior of the materials
upon p-doping.
To this purpose, we carried out cyclic voltammetry (CV)
measurements on P1ÀP3 under an inert atmosphere. The anodic
scans recorded for the three materials are shown in Figure 8.
The samples were analyzed as thin films obtained by drop-
casting from a 1.0 mg/mL chloroform solution onto a Pt disk
used as a working electrode. In the case of P1, two electro-
chemical events were clearly observed in the range of the applied
potential (0À2 V): the first event at Epeak = 1.59 V and the
second one at Epeak = 1.87 V. The irreversible character of the
latter event (indicated by the absence of an electrochemical
response in the second CV scan) suggests that further p-doping
of the conjugated backbone results in the loss of its conductive
properties probably by σ-bond cleavage.17 In fact, if the scan is
stopped at 1.70 V (Figure 9), the only anodic event was clearly
reversibile, suggesting that under these conditions the sole electro-
chemical process is the polaron formation by a one-electron
abstraction from the π-conjugated backbone of the polymer.
7983
dx.doi.org/10.1021/ma2015003 |Macromolecules 2011, 44, 7977–7986

Macromolecules
ARTICLE
Scheme 3. Synthetic Approach for the Obtainment of PFK and PFSa
a Reagents and conditions: (i) Pd(PPh3)4, Na2CO3 2.0 M, reflux.
film could be measured, suggesting that a higher packing degree
of the PFs cross-linked backbones may favor nonradiative decay
pathways for the excited state.
On the Origin of the g-Band. Although PFs have shown to be
intrinsically susceptible of aerobic oxidation, the synthesis of PFs for
which the appearance of the g-band is retarded as much as possible
is all the same highly desirable, and it would very much benefit from
a continuous investigation on the physical origin of the onset of the
g-band. Recalling that no consensus has been reached on this issue
and that the origin of the g-band has been explained either by a
favorable energy transfer toward the low-energy luminophors (the
so-called on-chain defect emission hypothesis) or by the formation
of fluorenone-based excimers, the following experiment was set up
to discriminate between the two assumptions.
The spectral stability of the synthesized PFs was studied
submitting dispersions (10 wt %) of P1ÀP3 into polyvinylcar-
bazole (PVK) to the same UV treatment in air described above.
The use of a semiconducting higher energy gap matrix such as
PVK was conceived to ensure a three-dimensional energy transfer
toward the low-energy luminophors. Thus, the eventual appear-
ance of the g-band for the UV-treated PF/PVK blends could be
taken as a clue of the occurrence of the on-chain defect emission
hypothesis, whereas a response of spectral stability would favor
the fluorenone-based excimers assumption.
Figure 12. PL spectra of PFK in chloroform solution, as neat film and as
a 3 wt % blend with PVK.
Since the results of this experiment are significant to our
purpose only if a complete energy transfer occurs from PVK
toward the polyfluorene emitters, preliminary tests were carried
out by recording the PL spectra of blends of P1ÀP3 in PVK
obtained by exciting at 320 nm, the absorption maximum of PVK.
Under these conditions, only the polyfluorene emission was
observed, without residual emission from PVK, thereby confirm-
ing the occurrence of a complete energy transfer. It should be
noted that, due to the amorphous nature of the polymers used in
this study and to the low concentration of PFs in the blend, the
degree of oxygen penetration in the blend films (a possible
variable influencing the extent and/or the kinetics of the aerobic
degradation) can be thought as independent from the PFs
primary structure. As is apparent from Figure 11, all of the
P1ÀP3 polymers in blend with PVK show an excellent spectral
stability upon UV irradiation in air for 30 min. After this
treatment, no g-band could be recorded either at λex = 320 nm
(i.e., exciting the donor PVK and promoting a three-dimensional
energy transfer) or at λex = 380 nm, thus directly exciting the
polyfluorene emitters. This result can be interpreted by admit-
ting that the isolation of the polymeric chains drastically frus-
trates the appearance of the low-energy emission.
Figure 13. PL spectra of PFS in chloroform solution, as neat film and as
a 3 wt % blend with PVK.
to investigate the PL behavior of a fluorenone-containing
polymer PFK, as neat film or in blend with PVK. The polymer
PFK was synthesized starting from 1 (mole fraction = 0.45),
2 (mole fraction = 0.50), and 2,7-dibromofluorenone (mole
fraction = 0.05) by a Suzuki polycondensation, as shown
in Scheme 3.
As already observed for similar fluorenone-containing poly-
mers, while in chloroform solution only the blue emission was
observed, the PL of PFK in the solid state was almost completely
In order to further clarify the role of the interchain interac-
tions in the appearance of the g-band, we deemed it worthwhile
7984
dx.doi.org/10.1021/ma2015003 |Macromolecules 2011, 44, 7977–7986

Macromolecules
ARTICLE
dominated by the g-band with an Ig/Ib ratio (Ib and Ig: maximum
intensity emission in the blue and green regions) independent of
the excitation wavelength (320 or 380 nm). The complete
absence of the g-band in a chloroform solution of PFK can be
explained, according to the on-chain defect emission scenario, by
the fact that the population of the excited states of fluorenone
luminophors is hindered due to a poorly efficient energy transfer
operating along the polyfluorene chain.
’ CONCLUSIONS
The spectral stability of three differently structured PFs under
aerobic oxidative stress was studied as a function of the C-9
substitution. The degradation of the PFs was studied in air either
by thermal annealing or by UV photodecomposition and allowed
us to support the following conclusions: (i) PFs show an intrinsic
tendency toward fluorenone formation independently of their
9-substitution and from the presence of 9-H defects; (ii) the
degradation pathways of PFs are greatly influenced by the
presence or the absence of a light source.
To improve the energy transfer toward the fluorenones while
simultaneously preserving the isolation of the polymeric chains,
we dispersed PFK (3 wt %) in PVK. Interestingly, in this case,
PFK exhibited the g-band (Figure 12) when the blend was
excited at 380 nm. However, the emission of the blend showed a
lower Ig/Ib ratio, when excited at 320 nm. The observation that
the Ig/Ib ratio remarkably changes as a function of the excitation
wavelength indicates the possible formation of an inhomoge-
neous blend, comprising both isolated and aggregated chains.
The PL emission of an inhomogeneous blend is the sum of the
photoluminescence of isolated polymeric chains (the contribu-
tion of which is higher at λex= 320 nm, since they can accept also
energy from the PVK host) and aggregated polymeric chains, the
contribution of which is higher at λex= 380 nm, since the energy
transfer from PVK is less efficient toward the aggregated chains.
This explains why the excitation at 380 nm increased the Ig/Ib
ratio. The cause of the formation of an inhomogeneous blend
may reside in the dipolar interactions between keto-groups,
inducing strong intermolecular interactions between adjacent
polymer chains.18 The entity of these interactions may depend
on several parameters including the polymer architecture and the
preparation of the films, thus explaining the variability reported
in the literature concerning the emission properties of fluore-
none-containing model systems. At the same time our results
prove the on-chain defect model as definitely insufficient to
interpret the results obtained for an inhomogeneous blend of a
fluorenone-containing PF in PVK as host.
The reproduction of the same behavior of fluorenone-contain-
ing polymers with a different planar moiety introduced into the
polymer chain could be useful to reinforce the hypothesis on the
excimeric nature of the g-band (Figure 13). To this purpose, a
poly(9,9-di-n-hexylfluorene) containing a 5% mol/mol of 9-(bis-
methylsulfanylmethylene)fluorene units (PFS) was prepared, as
reported in Scheme 3. The 9-(bis-methylsulfanylmethylene)-
fluorene was chosen as comonomer to warrant the presence,
within the PF backbone, of planar units devoid of carbonyl
moieties.19 The fluorescence behavior of PFS strictly resembles
that of PFK, since in chloroform solution PFS solely emits in the
blue region (λmax = 415 nm) while as film its fluorescence is
dominated by a lower energy (λmax = 537 nm) band. In the case
of the 3 wt % PFS/PVK blend, however, the low-energy emission
band is completely suppressed, exciting both at 320 and at
380 nm (Figure 13). It can tentatively be stated that the absence
of the strong dipolar interactions between the polymeric chains
facilitates the homogeneous dispersion of PFS in PVK. The fact
that the g-band emerges only under conditions promoting the
aggregation of PF chains is an indication that the interchain
interactions play a pivotal role in the appearance of the low-
energy emission band. By this line of reasoning, it is worth noting
that the oxidative degradation of PFs introduces not only the
keto-defects but also, through the fluorenone formation, regions
of planarity in the polymer backbone that can favor local
interchain interactions strongly influencing the optical behavior
in the solid state.
These experiments also allowed us to add additional pathways
to the known degradation mechanism of these materials, the
most important of which is the formation of a radical cation
which can occur by aerobic oxidation or by p-doping. Further-
more, this study contributed to the elucidation of the origin of the
g-band emission. By studying the behavior of the synthesized
PFs, our investigation points out that the sole on-chain defect
emission hypothesis is insufficient to explain the origin of the
g-band for fluorenone containing PFs. In this regard the
fundamental role of the interchain interactions was clearly
established. The formation, in the course of the PF degrada-
tion, of fluorenone defects so close in space to cofacially
interact with each other is unlikely because it would imply a
massive aerobic degradation under thermal or photochemical
stress. Conversely, it is safe to assume that the generation of
regions of planarity accompanying the fluorenone formation
along the polymer backbone can favor local intermolecular
interactions (aggregate and/or excimer formation) on which
the spectral stability of the polymer depend. The opinion
matured on the matter of the stability of fluorene-based
materials is that the great advantages associated with the
presence of the highly reactive C-9 on fluorene (e.g., its easy
functionalization) unfortunately also represent the Achille’s
heel of PFs. The research on alternative to the fluorene
building block seems to be a more promising approach to
warrant material stability under oxidative or, in general, under
p-doping exercise conditions.
’ AUTHOR INFORMATION
Corresponding Author
*E-mail: p.mastrorilli@poliba.it.
’ REFERENCES
(1) (a) Scherf, U.; List, E. J. W. Adv. Mater. 2002, 14, 477. (b) Neher,
D. Macromol. Rapid Commun. 2001, 22, 1365. (c) Bernius, M. T.;
Inbasekaran, M.; O’Brien, J.; Wu, W. Adv. Mater. 2000, 12, 1737.
(2) (a) Bliznyuk, V. N.; Carter, S. A.; Scott, J. C.; Kl€arner, G.; Miller,
R. D.; Miller, D. C. Macromolecules 1999, 32, 361. (b) Lee, J. I.; Kl€arner,
G.; Miller, R. D. Chem. Mater. 1999, 11, 1083.
(3) Zojer, E.; Pogantsch, A.; Hennebicq, E.; Beljonne, D.; Brꢀedas,
J. L.; Scandiucci de Freitas, P.; Scherf, U.; List, E. J. W. J. Chem. Phys.
2002, 117, 6794.
(4) (a) Becker, K.; Lupton, J. M.; Feldmann, J.; Nehls, B. S.;
Galbrecht, F.; Gao, D. Q.; Scherf, U. Adv. Funct. Mater. 2006, 16, 364.
(b) Kulkarni, A. P.; Kong, X.; Jenekhe, S. A. J. Phys. Chem. B 2004,
108, 8689. (c) Romaner, L.; Pogantsch, A.; Scandiucci de Freitas, P.;
Scherf, U.; Gaal, M.; Zojer, E.; List, E. J. W. Adv. Funct. Mater. 2003,
13, 597.
(5) Chochos, C. L.; Kallitsis, J. K.; Gregoriou, V. G. J. Phys. Chem. B
2005, 109, 8755.
(6) See for example: (a) Wu, Y. S.; Li, J.; Ai, X. C.; Fu, L. M.; Zhang,
J. P.; Fu, Y. Q.; Zhou, J. J.; Li, L.; Bo, Z. S. J. Phys. Chem. A 2007,
7985
dx.doi.org/10.1021/ma2015003 |Macromolecules 2011, 44, 7977–7986

Macromolecules
ARTICLE
111, 11473. (b) Vak, D.; Chun, C.; Lee, C. L.; Kim, J. J.; Kim, D. Y.
J. Mater. Chem. 2004, 14, 1342. (c) Lee, J. H.; Hwang, D. H. Chem.
Commun. 2003, 2836.
(7) (a) Ferenczi, T. A. M.; Sims, M.; Bradley, D. D. C. J. Phys.:
Condens. Matter 2008, 20, 045220. (b) Jaramillo-Isaza, F.; Turner, M. L.
J. Mater. Chem. 2006, 16, 83. (c) Sims, M.; Bradley, D. D. C.; Ariu, M.;
Koeberg, M.; Asimakis, A.; Grell, M.; Lidzey, D. G. Adv. Funct. Mater.
2004, 14, 765.
(8) List, E. J. W.; Guentner, R.; Scandiucci de Freitas, P.; Scherf, U.
Adv. Mater. 2002, 14, 374.
(9) Craig, M. R.; de Kok, M. M.; Hofstraat, J. W.; Schenning, A. P. H.
J.; Meijer, E. W. J. Mater. Chem. 2003, 13, 2861.
(10) (a) Kwon, Y. K.; Kim, H. S.; Kim, H. J.; Oh, J. H.; Park, H. S.;
Ko, Y. S.; Kim, K. B.; Kim, M. S. Macromolecules 2009, 42, 887. (b) Kim,
J.; Jin, Y.; Song, S.; Kim, S. H.; Park, S. H.; Lee, K.; Suh, H.
Macromolecules 2008, 41, 8324. (c) Liu, J.; Zou, J.; Yang, W.; Wu, H.;
Li, C.; Zhang, B.; Peng, J.; Cao, Y. Chem. Mater. 2008, 20, 4499. (d) Cho,
S. Y.; Grimsdale, A. C.; Jones, D. J.; Watkins, S. E.; Holmes, A. B. J. Am.
Chem. Soc. 2007, 129, 11910. (e) Jacob, J.; Sax, S.; Gaal, M.; List, E. J. W.;
Grimsdale, A. C.; M€ullen, K. Macromolecules 2005, 38, 9933.
(11) Abbel, R.; Wolffs, M.; Bovee, R. A. A.; van Dongen, J. L. J.; Lou,
X.; Henze, O.; Feast, J. W.; Meijer, E. W.; Schenning, A. P. H. J. Adv.
Mater. 2009, 21, 597.
(12) Grisorio, R.; Piliego, C.; Striccoli, M.; Cosma, P.; Fini, P.; Gigli,
G.; Mastrorilli, P.; Suranna, G. P.; Nobile, C. F. J. Phys. Chem. C 2008,
112, 20076.
(13) Grisorio, R.; Suranna, G. P.; Mastrorilli, P.; Nobile, C. F. Adv.
Funct. Mater. 2007, 17, 538.
(14) Liu, L.; Qiu, S.; Wang, B.; Zhang, W.; Lu, P.; Xie, Z.; Hanif, M.;
Ma, Y.; Shen, J. J. Phys. Chem. B 2005, 109, 23366.
(15) Zhuo, J. M.; Zhao, L. H.; Png, R. Q.; Wong, L. Y.; Chia, P. J.;
Tang, J. C.; Sivaramakrishnan, S.; Zhou, M.; Ou, E. C. W.; Chua, S. J.;
Sim, W. S.; Chua, L. L.; Ho, P. K. H. Adv. Mater. 2009, 21, 4747.
(16) Koenen, J. M.; Bilge, A.; Allard, S.; Alle, R.; Meerholz, K.;
Scherf, U. Org. Lett. 2009, 21, 2149.
(17) Montilla, F.; Mallavia, R. Adv. Funct. Mater. 2007, 17, 71.
(18) Chan, K. L.; Sims, M.; Pascu, S. I.; Ariu, M.; Holmes, A. B.;
Bradley, D. D. C. Adv. Funct. Mater. 2009, 19, 2147.
(19) Song, S.; Jin, Y.; Kim, S. H.; Kim, J. H.; Lee, K.; Suh, H. Polym. J.
2009, 41, 138.
7986
dx.doi.org/10.1021/ma2015003 |Macromolecules 2011, 44, 7977–7986