Deep blue light-emitting polymers with fluorinated backbone for enhanced color purity and efficiency

Article abstract of DOI:10.1016/j.polymer.2012.02.010

Two fluorene-based copolymers (PF-33F and PF-50F) with p-difluorophenylene units in the backbone were synthesized. In comparison with the reference poly(9,9-dioctylfluorene) (PFO), the introduction of p-difluorophenylene units not only increased the fluorescent quantum yields, but also improved the spectra purity and stability of these deep blue emitting copolymers. The famous green emission band at 520 nm from fluorenone defects was never detected for these copolymers even after they were thermal annealed in air at 150 °C. Organic light-emitting diodes were fabricated using them as emitting layer and pure blue electroluminescence was obtained. It was observed that PF-33F based device exhibited much higher current density and brightness than PF-50F and PFO devices. A maximum external quantum efficiency of 1.14% (1.14 cd A^-1) and the CIE (0.16, 0.13) were achieved for PF-33F device, which are among the best performance for polyfluorenes reported so far.

Full text of DOI:10.1016/j.polymer.2012.02.010

Polymer 53 (2012) 1529e1534
Contents lists available at SciVerse ScienceDirect
Polymer
journal homepage: www.elsevier.com/locate/polymer
Deep blue light-emitting polymers with fluorinated backbone for enhanced color
purity and efficiency
,
Ting Zhang a, Renjie Wang a, Huicai Ren b, Zhiguang Chen b, Jiuyan Li a
*
a State Key Laboratory of Fine Chemicals, School of Chemical Engineering, Dalian University of Technology, 2 Linggong Road, Dalian 116024, China
b School of Chemistry, Dalian University of Technology, 2 Linggong Road, Dalian 116024, China
a r t i c l e i n f o
a b s t r a c t
Article history:
Two fluorene-based copolymers (PF-33F and PF-50F) with p-difluorophenylene units in the backbone
were synthesized. In comparison with the reference poly(9,9-dioctylfluorene) (PFO), the introduction of
p-difluorophenylene units not only increased the fluorescent quantum yields, but also improved the
spectra purity and stability of these deep blue emitting copolymers. The famous green emission band at
520 nm from fluorenone defects was never detected for these copolymers even after they were thermal
annealed in air at 150 ꢀC. Organic light-emitting diodes were fabricated using them as emitting layer and
pure blue electroluminescence was obtained. It was observed that PF-33F based device exhibited much
higher current density and brightness than PF-50F and PFO devices. A maximum external quantum
efficiency of 1.14% (1.14 cd Aꢁ1) and the CIE (0.16, 0.13) were achieved for PF-33F device, which are
among the best performance for polyfluorenes reported so far.
Received 10 November 2011
Received in revised form
18 January 2012
Accepted 5 February 2012
Available online 9 February 2012
Keywords:
Fluorene
p-Difluorophenylene comonomer
Organic light-emitting diodes
Ó 2012 Elsevier Ltd. All rights reserved.
1. Introduction
ability of the 9-position carbon in fluorene moieties and thus to
suppress the green emission. One example was to construct a rigid
Polyfluorene (PF)-type polymers have been established as most
promising blue light-emitting materials for application in organic
light-emitting diodes (OLEDs). They possess many merits such as
high fluorescent quantum yields, easy syntheses and structure
modification, good solubility in common organic solvents and
acceptable charge mobility [1e3]. However, an undesired low-
energy “green” band covering a broad spectral range from 500 to
600 nm is frequently generated in both photoluminescence (PL)
and electroluminescence (EL) of PFs during operation, which not
only damages the blue color purity and stability but also limits the
emission efficiency [4].
After the intensive debate in the last few years on the origin of
this long-range emission [5,6], it is more acceptable that it stems
from the emissive keto defects of PFs (i.e. fluorenone), which are
generated by chemical oxidation of the monoalkylfluorene impu-
rities at the 9-position carbon during polymer preparation or
operation in air for a long time [7,8]. Accordingly, ingenious
synthetic methods were used to prepare polyfluorenes without
monoalkylfluorene defects [9] and some attempts were also taken
to remove the generated impurities [10]. Furthermore, various
approaches have also been exploited to increase the anti-oxidation
backbone in the fluorene moieties to inhibit the oxidation of the
9-position carbon since the oxidation of the 9-position carbon
involves hybrid state conversion from sp3 to sp2 and thus the
change of bond angles. In this way Park et al. developed a stable
blue emitting PF derivative with an EL efficiency of 0.7 cd Aꢁ1 and
CIE coordinates of (0.17, 0.12) [11]. Another attempt was to incor-
porate electron-deficient groups as the endcappers [12] or side-
chains [13,14] of the fluorene-based polymers. Chen and coworkers
have achieved high stability and improved purity in blue emission
as well as improved device efficiency via capping polydioctyl-
fluorene (PFO) chain ends with electron-deficient moieties such as
oxadiazole and triazole, which can induce a minor amount of long
conjugating length species (regarded as
of energy transfer from amorphous matrix to the
b
phase) to control extents
phase [12]. In
b
our previous report [13], the fluorene copolymers with di(tert-
butyl)phenoxy)sulfonyl-phenylene comonomer showed stable
blue PL even after thermal annealing in air and pure blue EL. The
improved spectral stability of these PFs was suggested to benefit
from the strong electron-withdrawing nature of the sulphonate
side groups that facilitated the anti-oxidation ability. However, the
introduction of these sidechain groups also resulted in the decrease
in the charge mobility and the fluorescent quantum yields of the
copolymers. Similarly, the strong electron-deficient 2,8-
dihexyldibenzothiophene-S,S-dioxide-3,7-diyl units was also used
as the building block to synthesize pure blue light-emitting PF
* Corresponding author. Fax: þ86 411 84986233.
E-mail address: jiuyanli@dlut.edu.cn (J. Li).
0032-3861/$ e see front matter Ó 2012 Elsevier Ltd. All rights reserved.
doi:10.1016/j.polymer.2012.02.010

1530
T. Zhang et al. / Polymer 53 (2012) 1529e1534
derivatives. However, an EL efficiency of those polymers was only
0.24 cd Aꢁ1 with a blue color of (0.18, 0.15) [15].
performed on a CarloeEriba 1106 elemental analyzer. Thermog-
ravimetry analyses (TGA) and differential scanning calorimetry
(DSC) measurements were carried out using a PerkineElmer ther-
mogravimeter (Model TGA7) and a Netzsch DSC 204 at a heating
rate of 10 ꢀC minꢁ1 under a nitrogen atmosphere, respectively. The
UVevis absorption and fluorescence spectra measurements were
Apparently, the type of the functional comonomer and its
content ratio in polymers are two important factors to determine
both the luminescent and charge-transporting properties and thus
the EL performance of these polymers. The functional groups
should be properly selected and their ratio should be well tuned so
that the blue color purity and stability can be improved without
expense of luminescent quantum yield and charge mobility. In this
paper, we reported the synthesis and properties of two novel
fluorene-based copolymers containing p-difluorophenylene
comonomer, PF-33F and PF-50F (Scheme 1). The p-difluor-
ophenylene group was selected as an important building block with
the expectation to tune the electronic state of the polymer back-
bone and to improve spectral purity and stability of the polymers
[16,17]. For optimization purpose, the content ratio of the p-
difluorophenylene unit was varied from 33 mol% (PF-33F) to
50 mol% (PF-50F). For comparison, poly(9,9-dioctylfluorene) (PFO),
was also synthesized and studied under the same conditions. The
PL spectral stability of these polymers was investigated by moni-
toring the emission spectra before and after thermal annealing in
air. These polymers were also used as emitting layer to fabricate
OLEDs. The results show that, both the PL and the EL spectra of
these copolymers were never contaminated by that green emission
band, even after the samples were thermal annealed in air for 2 h
(for PL). Furthermore, the EL performance of these copolymers
depends on the ratio of the p-difluorophenylene units in copoly-
mers. A ratio of 33 mol% of p-difluorophenylene units in copolymer
led to the highest device current and brightness at a given voltage.
The PF-33F based OLEDs exhibited a maximum external quantum
efficiency of 1.14% (1.14 cd Aꢁ1) and the CIE coordinates of (0.16,
0.13), both which were much better than the reference PFO.
performed on
a PerkineElmer Lambda 35 UVeVisible spec-
traphotometer and a PerkineElmer LS55 fluorescence spectrom-
eter, respectively. The fluorescence quantum yields were
determined against quinine sulfate as the standard (
F
¼ 0.55, in
0.1 N H2SO4). Cyclic voltammetry of the spin-coated films of PF-33F
on ITO glasses was performed by using a conventional three-
electrode configuration and an electrochemical workstation (BAS
100B, USA) at a scan rate of 50 mV sꢁ1. A (0.10 M AgNO3)/Ag elec-
trode and a platimun wire were used as reference and counter-
electrodes. All measurements were made at room temperature in
acetonitrile solution, with 0.10 M tetra-n-butylammonium hexa-
fluorophosphate (Bu4NPF6) as supporting electrolyte.
2.2. Device fabrication and measurements
The pre-cleaned ITO glass substrates (30
U ,
ꢁ1) were treated by
UV-Ozone for 20 min. A 40 nm thick PEDOT/PSS film was first spin-
coated on pre-treated ITO substrates from aqueous dispersion and
baked at 120 ꢀC for 40 min in air. Subsequently the polymer solu-
tions (12 mg mlꢁ1 in chlorobenzene) were filtered through 0.45
mm
PTFE filter and spin-coated on PEDOT/PSS film, the thickness of
which was controlled as 40 nm by adjusting the spin rate. The
substrate was transferred into a vacuum chamber to deposit the
TPBI layer at a base pressure less than 10ꢁ6 torr. Finally, the device
fabrication was completed through thermal deposition of LiF
(1 nm) and then capping with Al metal (100 nm) as cathode. The
emitting area of each pixel is determined by overlapping of two
electrodes as
9
mm2. The EL spectra, CIE coordinates, and
2. Experimental
currentevoltageeluminance relationships of devices were
measured with computer-controlled Spectrascan PR 705 photom-
eter and a Keithley 236 source-measure-unit. All the measurements
were carried out in ambient condition at room temperature.
2.1. Materials and methods
Chemicals, reagents and solvents from commercial sources are
of analytical or spectroscopy grade and used as received without
further purification. 1H NMR spectra were recorded on a Varian
INOVA spectrometer (400 MHz). Number-average (Mn) and weight-
average (Mw) molecular weights were determined by using a PL-
GPC220 instrument with THF as retention solvent and mono-
disperse polystyrenes (PS) as standards. Elemental analysis was
2.3. Synthesis of polymers
The monomers 1 and 2 were synthesized according to the
literature method [18], 3 was purchased from Alfa Aesar and used
without further purification. The following general procedure was
Scheme 1. Synthetic routes of polymers.

T. Zhang et al. / Polymer 53 (2012) 1529e1534
1531
Table 1
used for the preparation of all the polymers. To a 50 mL two-necked
Structural characterization, thermal and photophysical data of the polymers.
flask were added the appropriate diboronate (1 equiv), the appro-
priate dibromide (1 equiv), Pd(PPh3)4 (0.005 equiv) and toluene
(15 mL). The flask equipped with a condenser was then evacuated
and filled with argon three times to remove traces of air. The
mixture was stirred at 85 ꢀC for 5 min before an aqueous tetraethyl
ammonium hydroxide solution (20%, 4.3 equiv) was added. The
mixture was then stirred at 85 ꢀC under nitrogen for 24 h. Bro-
mobenzene (1 equiv) was added to the mixture and stirred for 12 h.
Phenylboronic acid (1 equiv) was added to the mixture after which
the mixture was stirred for 12 h. The mixture was cooled down to
room temperature and dropped by pipette into a stirred solution of
methanol (100 mL) in a beaker. The precipitate was isolated and
extracted with acetone for 12 h in a Soxlet apparatus, the resulting
polymers were collected and dried under reduced pressure.
PFO: Light yellow powder (yield: 81.5%). GPC (THF, PS standard):
Mn ¼ 202,215 g molꢁ1, Mw ¼ 269,000 g molꢁ1, PD ¼ 1.33; 1H NMR
abs
em
max
Polymer Yield Mn
Mw/Mn Tg
Td,5%
F
(%) a
l
l
(nm)b
max
(%)
(ꢀC) (ꢀC)
(nm)b
PFO
PF-33F
PF-50F
81.5 202,215 1.33
93.3 35,633 1.52
93.6 36,679 1.61
/
/
0.63
387/384 417/432
370/370 415/420
364/361 403/411
109 412 0.70
n.oc 421 0.73
a
Measured in CH2Cl2 solution with quinine sulfate as the standard (
The data before the slash is for dilute dicholoromethane solution, and the data
F
¼ 0.55).
b
after the slash are for solid films.
c
n.o. ¼ not observed.
transition behaviors of the copolymers were investigated by DSC.
As shown by the DSC thermograms in Fig. S2, one endothermic
peak due to glass transition was detected at 110 ꢀC (Tg) when
a powder of PF-33F was heated for the first run. This implies that
PF-33F was amorphous even when it was directly obtained from
organic solvents. When a sample of PF-50F was heated, an
exothermic peak was observed at 127 ꢀC, which was ascribed to the
crystallization. With increasing temperature, the crystal melts at
247 ꢀC. No glass transition was observed for PF-50F under the
present experimental conditions. The relatively high glass transi-
tion temperature for PF-33F is an essential merit that is favorable
for device stability, especially when it is used under high temper-
atures [19].
(400 MHz, CDCl3, TMS):
d 7.84e7.32 (m, ArH), 2.12 (br, CH2),
1.25e1.15 (m, CH2), 0.84e0.80 (m, CH2þCH3);
PF-50F: White fibrous solid (yield: 93.6%). GPC (THF, PS stan-
dard): Mn ¼ 36,679 g molꢁ1, Mw ¼ 59,070 g molꢁ1, PD ¼ 1.61; 1H
NMR (400 MHz, CDCl3, TMS):
d 7.87e7.36 (m, ArH), 2.12 (br, CH2),
1.23e1.13 (m, CH2), 0.84e0.81 (m, CH2þCH3); Elemental analysis
found: C 82.48, H 8.40.
PF-33F: White fibrous solid (yield: 93.3%). GPC (THF, PS stan-
dard): Mn ¼ 35,633 g molꢁ1, Mw ¼ 54,064 g molꢁ1, PD ¼ 1.52; 1H
NMR (400 MHz, CDCl3, TMS):
d
7.87e7.36 (m, ArH), 2.12 (br, CH2),
3.3. Photophysical properties
1.23e1.12 (m, CH2), 0.84e0.80 (m, CH2 þ CH3); Elemental analysis
found: C 85.28, H 9.27.
The photophysical properties of these polymers were investi-
gated by means of electronic absorption and photoluminescence
(PL) spectra in dilute dichloromethane solutions and solid films on
quartz plates. Fig. 1 illustrates the absorption and PL spectra of
these polymers in dilute dichloromethane solutions. Similar to the
reference PFO, the two copolymers exhibit structureless absorption
in dilute solutions with peaks at 364 and 370 nm for PF-50F and
3. Results and discussion
3.1. Synthesis
The general synthetic routes for copolymers PF-33F and PF-50F
and the reference polymer PFO are outlined in Scheme 1. They were
synthesized by Suzuki polycondensation reaction of monomer 2 (1
equiv.) with 1 and 3 (totally 1 equiv.) in the appropriate molar ratios
([3] ¼ 0, PFO; [1] ¼ 0, PF-50F; [1]/[3] ¼ 0.5, PF-33F). The reaction
was performed in toluene solution with [Pd(PPh3)4] (0.005 equiv.)
as catalyst and aqueous tetraethylammonium hydroxide (4.3 equiv)
as an emulsifying base. The polymers were end-capped with
phenyl groups. Monomers 1 and 2 were synthesized according to
literature methods as shown in Scheme 1 [18]. Copolymers PF-33F
and PF-50F were obtained as white fibrous solids in high yields
over 90%. They have good solubility in common organic solvents
such as tetrahydrofuran, dichloromethane and toluene. The
number-average molecular weights (Mn) and weight-average
molecular weights (Mw) of these copolymers were determined by
gel-permeation chromatography (GPC) using polystyrenes (PS)
standards as 35,633 and 54,064 for PF-33F, 36,679 and 59,070 for
PF-50F, respectively. The polydispersity indexes (PDI, PDI ¼ Mw/
Mn) were calculated as 1.52 and 1.61 for PF-33F and PF-50F,
respectively. A brief summary of these parameters is listed in
Table 1.
PF-33F, respectively, which are ascribed to the
pe
p* transition of
the polymer backbone. With gradually increasing the ratio of the p-
difluorophenylene units from PFO to PF-50F, the absorption spec-
trum reveals a general trend of a blue shift. The same blue-shift
order was observed in the absorption spectra of the polymer
films, as shown by the data in Table 1. This is probably because the
less coplanarity between the adjacent fluorene and phenylene rings
in copolymers than the two fluorenes in PFO reduces the conju-
gation extent in the polymer backbone and thus increases the
HOMOeLUMO energy of individual molecules in dilute solution
and the energy band gap in solid state. Upon photoexcitation at the
absorption maximum, all these polymers emit strong deep blue
fluorescence. As shown in Fig. 1b, the fluorescence spectra of these
polymers show three well-resolved vibronic bands. For example,
the three bands for PF-33F are located at 415, 438, and 472 nm, and
those for PFO are at 417, 440, and 475 nm. The three fluorescence
bands of each polymer can be ascribed to the electronic transitions
from the lowest vibronic level of the ground state (S0,0) to different
vibronic levels of the first singlet excited state (S1,0, S1,1, and S1,2). In
consistence with the blue shift in absorption spectra, a blue shift is
also observed in the PL spectra of these polymers when increasing
the ratio of p-difluorophenylene units.
3.2. Thermal properties
From the wavelength difference between the absorption
maximum and fluorescence maximum, the Stokes shifts of these
copolymers were determined as 45 nm and 39 nm for PF-33F and
PF-50F, respectively, which are both lightly larger than that of PFO
(30 nm). A larger Stokes shift is usually desired for a light-emitting
material since it means a weaker self-absorption and more efficient
The thermal properties of PF-33F and PF-50F were investigated
by thermogravimetry analysis (TGA) and differꢁe1ntial scanning
calorimetry (DSC) at a scanning rate of 10 ꢀC min in a nitrogen
atmosphere and the results are summarized in Table 1. PF-33F and
PF-50F exhibit high thermal stability with decomposition
temperatures (corresponding to 5% weight loss) over 400 ꢀC (see
Fig. S1 in the supporting information). Thermally induced phase
light output in OLEDs. The fluorescence quantum yields (
F) of the
copolymers were measured in dilute dichloromethane solutions

1532
T. Zhang et al. / Polymer 53 (2012) 1529e1534
a
a
Before
50
100
1.0
0.8
0.6
0.4
0.2
0.0
1.0
PF-50F
PF-33F
PFO
0.8
150
0.6
0.4
0.2
300
350
400
450
500
550
0.0
Wavelength (nm)
350 400 450 500 550 600 650
Wavelength (nm)
b
b
1.0
0.8
0.6
0.4
0.2
0.0
PF-50F
PF-33F
PFO
1.0
0.8
0.6
0.4
0.2
0.0
Before
50
100
150
350
400
450
500
550
600
Wavelength (nm)
Fig. 1. Absorption a) and fluorescence b) spectra of polymers in dilute dichloro-
methane solution.
400
450
500
550
600
650
Wavelength (nm)
using quinine sulfate (
They were determined as 0.70 and 0.73 for PF-33F and PF-50F,
respectively, which are both higher than that of PFO (
There have been similar reports that the introduction of fluorine
atoms or trifluoromethyl groups in organic molecules or metal
complexes resulted in the increase of the fluorescent or phospho-
rescent quantum yields [21]. The high fluorescence quantum yields
of the present copolymers indicate that they may be promising
blue light-emitting materials for application in OLEDs.
F
¼ 0.55 in 0.1 N H2SO4) as a reference [20].
c
1.0
0.8
0.6
0.4
0.2
0.0
F
¼ 0.63).
Before
50
100
150
3.4. Spectral stability
In order to evaluate the spectral stability of PF-33F and PF-50F,
copolymer films were spin-coated from chlorobenzene solutions
onto quartz plates and then thermally annealed in air at various
temperatures for 2 h. After the films were cooled to the ambient
temperature, PL spectra were measured. PFO films were also
treated and tested under same conditions for comparison. Fig. 2
shows the fluorescence spectra of these polymer films before and
after annealing. It is obvious that the PL spectra of copolymers PF-
33F and PF-50F before and after annealing are almost identical to
each other except for the discernable spectral position change that
is probably due to physical change or instability of measuring
environment [22]. The absence of any new emission band in PF-33F
and PF-50F films excludes the formation of any new emissive
species, even after annealing at such a high temperature of 150 ꢀC
in air for 2 h. However, under the same conditions, the PFO film
exhibited the well-known green emission band from 500 to
600 nm after annealing at and over 100 ꢀC. Furthermore, the
intensity of the green band became higher with increasing
350
400
450
500
550
600
650
Wavelength (nm)
Fig. 2. Fluorescence spectra of a) PF-50F, b) PF-33F) and c) PFO films before and after
thermal annealing in air at various temperatures.
temperature. It is obvious that the introduction of the p-difluor-
ophenylene units to the mainchains in copolymers PF-33F and PF-
50F prevented the formation of the green emission band and
improved the color purity and stability of their blue fluorescence.
3.5. Electrochemical properties
In order to investigate the electrochemical properties of the
novel copolymers, the PF-33F films were made by spin-coating on
ITO glass substrates and were used as working electrode in the

T. Zhang et al. / Polymer 53 (2012) 1529e1534
1533
250
PF-50F
PF-33F
PFO
0.6
0.4
200
150
100
50
0.2
0.0
-0.2
0
0
2
4
6
8
10
12
14
-3
-2
-1
0
1
2
+
Voltage (V)
Potential (V vs Ag /Ag)
Fig. 3. Cyclovoltammogram of PF-33F film in acetonitrile at a scan rate of 50 mV sꢁ1
.
Fig. 5. The current densityevoltage characteristics for the PFO, PF-33F, and PF-50F
based OLEDs.
traditional three-electrode set-up. The cyclovoltammogram of PF-
33F film is shown in Fig. 3. PF-33F exhibits a reversible one-
electron oxidation and irreversible reduction processes. The irre-
versible reduction was frequently observed for polymers especially
when the sample was film [23,24], the real reason of which is still
not clear, but may be partially because of the film format. The onset
potentials of oxidation and reduction for PF-33F were determined
as 1.17 V and ꢁ1.95 V versus Agþ/Ag. The HOMO and LUMO energy
identical conditions for comparison. In these devices, PEDOT:PSS
acts as hole injection layer (HIL), TPBI as the electron-transporting
(ETL) and hole-blocking layer (HBL). The copolymer based devices
emitted brightdeep-blue light with emissionpeakat 422 nmandCIE
coordinates of (0.16, 0.13) for PF-33F, and 412 nm and CIE coordi-
nates of (0.17, 0.13) for PF-50F, as shown by the EL spectra in Fig. 4.
The EL spectra are almost identical in spectral features to their PL of
films, except with a slight red-shift of 1e2 nm. The red-shift in EL is
a commonly observed phenomena in most organic and polymeric
light-emitting diodes, which is probably caused by the influence of
the electrical field on the excited states [26]. Moreover, the EL
spectra and the CIE coordinates of PF-33F and PF-50F are observed
to be insensitive to the driving voltage, indicating a stable blue
emission color with respect to the excitation intensity. In contrast,
the EL spectra of the PFO device (Figure S3 in the supporting infor-
mation) always consist of the blue parts at 438, 466 and 494 nm,
which correspond tothe fine-structured bands in PL of the fresh PFO
films, and a broad green component with a peak at 520 nm even at
low driving voltage of 5 V. The intensity of the green emission band
increased with increasing the driving voltage. It should be noted that
the two long-wavelength blue bands also ascended along with the
green band. The green emission could be observed in EL of the fresh
PFO device, but not in PL of the fresh PF films. There have been
similar reports that the green emission from the fluorenone is
stronger in EL than in PL [8,27]. This is because the low-energy keto
defects can be excited not only by energy transfer from the fluorene
levels were estimated as ꢁ5.57 eV and ꢁ2.45 eV, respectively,
onset
according to the equations of HOMO (eV) ¼ ꢁ(E
þ 4.4 eV) [25]
ox
onset
and LUMO (eV) ¼ ꢁ(E
þ 4.4 eV). The electrochemical bandgap
red
of PF-33F was determined as 3.12 eV by calculating the difference
between the onset potentials of oxidation and reduction. Appar-
ently, a HOMO level at ꢁ5.57 eV is quite to that of the widely-used
hole injecting material PEDOT:PSS (ꢁ5.3 eV) and a LUMO level
at ꢁ2.45 eV is also close to that of the electron-transporting
material TPBI (ꢁ2.8 eV), implying that efficient hole and electron
injections can be expected when PF-33F was used in combination
with these functional layers in OLEDs.
3.6. Electroluminescence properties of OLEDs
Based on their pure blue fluorescence and relatively high
quantum yields, PF-33F and PF-50F were used as emitting layer to
fabricate OLEDs with a configuration of ITO/PEDOT:PSS (40 nm)/
polymers (40 nm)/TPBI (20 nm)/LiF (1 nm)/Al (200 nm). The fresh
PFO sample was also used to fabricate OLEDs and investigated under
500
1.0
PF-50F
PF-50F
PF-33F
PFO
PF-33F
400
0.8
300
200
100
0
0.6
0.4
0.2
0.0
0
2
4
6
8
10
12
14
300
400
500
600
700
800
Wavelength (nm)
Voltage (V)
Fig. 4. EL spectra for the PF-33F, and PF-50F based OLEDs.
Fig. 6. The luminanceevoltage curves for the PFO, PF-33F, and PF-50F based OLEDs.

1534
T. Zhang et al. / Polymer 53 (2012) 1529e1534
Table 2
synthesized and characterized as deep blue light-emitting mate-
EL performance of the polymers based OLEDs.
rials. The introduction of the fluorinated groups in backbone
increased the fluorescent quantum yield and improved the spectra
purity and stability. However, 33 mol% of fluorinated units was
proved to be the most appropriate ratio leading to satisfied charge
mobility. This is one successful example to improve the blue color
purity of fluorene-type light-emitting polymers without expense of
luminescent quantum yield and charge mobility. A quantum effi-
ciency of 1.14% and CIE (0.16, 0.13) were obtained for PF-33F based
OLED, among the best performance reported for fluorene-based
blue light-emitting polymers so far.
b
b
Polymer Von
L (cd mꢁ2
)
hL (cd Aꢁ1
)
hext (%) b lmax
CIE (x, y)
a
(V)
(nm)
PFO
PF-33F
PF-50F
4.8
4
5.9
452 (14 V)
486 (8 V)
61 (12 V)
0.88 (6 V)
1.14 (5 V)
0.092 (7 V)
0.53
1.14
0.086
438
422
412
0.18, 0.21
0.16, 0.13
0.17, 0.13
The data in the parentheses are the voltages at which the data are recorded.
Turn-on voltage, recorded at 1 cd mꢁ2
Maximum values of the devices.
.
a
b
moieties but also by direct charge trapping in the EL device made of
PFO, while only by energy transfer in PL of PFO films. Evidently the
introduction of the fluorinated moieties in the backbone of these
two copolymers also guaranteed the color purity and stability of the
EL, which is usually more sensitive to the purity of light-emitting
material than PL. It is suggested that the improved spectral purity
and stability in both PL and EL of these blue light-emitting copoly-
mers benefit from the presence of p-difluorophenylene units which
probably take away electron density from the 9-position carbon
Acknowledgments
We thank the National Natural Science Foundation of China
(20923006, and 21072026), the Ministry of Education for the New
Century Excellent Talents in University (Grant NCET-08-0074), and
the NKBRSF (2009CB220009) for financial support of this work.
through
withdrawing nature.
p-conjugation channels due to its appropriate electron-
Appendix A. Supplementary data
The current densityevoltage (JeV) characteristics of these
OLEDs are illustrated in Fig. 5. In spite of the identical device
configuration in the three OLEDs, the PF-33F and PF-50F based
devices exhibited much higher current densities than PFO device. It
is implied that the introduction of p-difluorophenylene units in the
copolymer backbone is favorable to increase the charge conduc-
tivity of the resultant copolymers. However, when the content ratio
of the fluorinated units increased from 33% to 50%, the current
density at a given voltage decreased dramatically for PF-50F device
in comparison to PF-33F device. It is proposed that too high ratio of
p-difluorophenylene units probably act as electron traps and
induce unbalance of negative and positive charges.
Fig. 6 shows the luminanceevoltage curves for the three OLEDs.
All the EL performance data are summarized in Table 2. The PF-33F
device turned on (to deliver a brightness of 1 cd mꢁ2) at a relatively
lowvoltage of 4 V, and reached a maximumluminance of 486 cd mꢁ2
at 8 V. The luminance at a given voltage decreased dramatically for
PF-50F device in comparison with the PF-33F device despite the
slightly higher fluorescent quantum yield for PF-50F. This is
consistent with the current density trend observed in Fig. 5. A
maximum external quantum efficiency of 1.14% (corresponding to
aluminance efficiencyof1.14 cdAꢁ1) was achieved for PF-33Fdevice
at 5 V. Under same experimental conditions, the PFO device turned
on at a higher voltage of 4.8 V and the maximum external quantum
efficiency reached to 0.53% (0.88 cd Aꢁ1 at 6 V), which is comparable
with the reported data for PFO in literatures [11,24]. Evidently, the EL
performances of PF-33F are much better than the parent PFO with
improvedblue color purity, highercharge mobility, and much higher
EL efficiency. This should benefit from the presence of the p-
difluorophenylene units and its proper content ratio. In addition, the
EL efficiency and color coordinates of present copolymers are
evidently better than those of most polyfluorene derivatives re-
ported in the literature [11,13,15]. As far as we know, a quantum
efficiency of 1.14%, a luminescence efficiency of 1.14 cd Aꢁ1 and CIE
(0.16, 0.13) achieved for PF-33F device are among the best perfor-
mance reported for blue light-emitting polyfluorenes so far.
Supplementary data associated with this article can be found, in
the online version, at doi:10.1016/j.polymer.2012.02.010.
References
[1] Grimsdale AC, Chan K, Martin RE, Jokisz PG, Holmes AB. Chem Rev 2009;109:
897e1091.
[2] Pogantsch A, Wenzl FP, List EJW, Leising G, Grimsdale AC, Müllen K. Adv Mater
2002;14:1061e4.
[3] Xiao S, Nguyen M, Gong X, Cao Y, Wu H, Moses D, et al. Adv Funct Mater 2003;
13:25e9.
[4] Scherf U, List EJW. Adv Mater 2002;14:477e87.
[5] Teetsov JA, Vanden Bout DA. J Am Chem Soc 2001;123:3605e6.
[6] Montilla F, Mallavia R. Adv Funct Mater 2007;17:71e8.
[7] Romaner L, Pogantsch A, de Freitas PS, Scherf U, Zojer E, List EJW. Adv Funct
Mater 2003;13:597e601.
[8] Gong X, Iyer PK, Moses D, Bazan GC, Heeger AJ. Adv Funct Mater 2003;13:
325e30.
[9] Cho SY, Grimsdale AC, Jones DJ, Watkins SE, Holmes AB. J Am Chem Soc 2007;
129:11910e1.
[10] Craig MR, de Kok MM, Hofstraat JW, Schenning APHJ, Meijer EW. J Mater
Chem 2003;13:2861e2.
[11] Park SH, Jin YG, Kim JY, Kim SH, Kim JW, Suh HS, et al. Adv Funct Mater 2007;
17:3063e8.
[12] Hung MC, Liao JL, Chen SA, Chen SH, Su AC. J Am Chem Soc 2005;127:
14576e7.
[13] Li JY, Ziegler A, Wegner G. Chem Eur J 2005;11:4450e7.
[14] Charas A, Morgado J, Marinho JMG, Alcácer L, Lim SF, Friend RH, et al. Polymer
2003;44:1843e50.
[15] Kamtekar KT, Vaughan HL, Lyons BP, Monkman AP, Pandya SU, Bryce MR.
Macromolecules 2010;43:4481e8.
[16] Kameshima H, Nemoto N, Endo T. J Polym Sci Part A Polym Chem 2001;39:
3143e50.
[17] Assaka AM, Rodrigues PC, de Oliverira ARM, Ding L, Hu B, Karasz FE, et al.
Polymer 2004;45:7071e81.
[18] Jo J, Chi CY, Höger S, Weger G, Yoon DY. Chem Eur J 2004;10:2681e8.
[19] Ye SH, Liu YQ, Lu K, Wu WP, Du CY, Liu Y, et al. Adv Funct Mater 2010;20:
3125e35.
[20] Ren HC, Li JY, Wang RJ, Zhang T, Gao ZX, Liu D. Polymer 2011;52:3639e46.
[21] Wang R, Liu D, Ren H, Zhang T, Yin H, Liu G, et al. Adv Mater 2011;23:2823e7.
[22] Tao S, Peng ZK, Zhang XH, Wang PF, Lee CS, Lee ST. Adv Funct Mater 2005;15:
1716e21.
[23] Zeng DL, Chen JW, Chen Z, Zhu WH, He J, Yu F, et al. Macromol Rapid Commun
2007;28:772e9.
[24] Yang W, Hou Q, Liu C, Niu Y, Huang J, Yang R, et al. J Mater Chem 2003;13:
1351e5.
[25] Zhang T, Liu D, Wang Q, Wang RJ, Ren HC, Li JY. J Mater Chem 2011;21:
12969e76.
4. Conclusion
[26] Yamamoto H, Wilkinson J, Long JP, Bussman K, Christodoulides JA, Kafafi ZH.
Nano Lett 2005;5:2485e8.
[27] Lupton JM. Chem Phys Lett 2002;365:366e8.
In summary, two fluorene-based copolymers containing p-
difluorophenylene units in backbone, PF-33F and PF-50F, were