Synthesis, photophysics, and electroluminescence of new quinoxaline-containing poly(p-phenylenevinylene)s

Article abstract of DOI:10.1021/ma048903q

Four new soluble poly(p-phenylenevinylene) (PPV) derivatives containing one or two quinoxaline moieties per repeat unit, either in the main chain or as pendants to the main chain, were synthesized, characterized, and explored as emissive and electron transport materials in polymer light-emitting diodes (LEDs). Polymers containing one quinoxaline moiety per repeat unit (QXPV1 and QXPV3) showed low melting transitions (<100°C), whereas those with two quinoxaline moieties per repeat unit (QXPV2 and QXPV4) had relatively high glass transition temperatures (>140°C). The polymers emit blue to green light (404-536 nm) in dilute solution and blue-green to yellow light (470-563 nm) in the solid state. The photoluminescence emission was well-described by single-exponential decay with lifetimes ranging from 200 ps to 2.2 ns in both dilute solution and thin film, indicating lack of intermolecular emissive species in the solid state. PPV derivatives with quinoxaline moieties in the main chain (QXPV1 and QXPV2) showed facile reversible electrochemical reductions with electron affinities of 2.63-2.75 eV. As emissive materials in LEDs, greenish-yellow electroluminescence with a brightness of up to 450 cd/m² was obtained from single-layer diodes of QXPV1 with aluminum cathode in air.

Full text of DOI:10.1021/ma048903q

Macromolecules 2004, 37, 7867-7878
7867
Synthesis, Photophysics, and Electroluminescence of New
Quinoxaline-Containing Poly(p-phenylenevinylene)s
P a n a yiotis Ka r a sta tir is, J oh n A. Mik r oya n n id is, a n d Ioa k im K. Sp iliop ou los*
Chemical Technology Laboratory, Department of Chemistry, University of Patras, Patras, Greece 26500
Abh ish ek P . Ku lk a r n i a n d Sa m son A. J en ek h e*
Departments of Chemical Engineering and of Chemistry, University of Washington,
Seattle, Washington 98195
Received J une 3, 2004; Revised Manuscript Received August 5, 2004
ABSTRACT: Four new soluble poly(p-phenylenevinylene) (PPV) derivatives containing one or two
quinoxaline moieties per repeat unit, either in the main chain or as pendants to the main chain, were
synthesized, characterized, and explored as emissive and electron transport materials in polymer light-
emitting diodes (LEDs). Polymers containing one quinoxaline moiety per repeat unit (QXP V1 and QXP V3)
showed low melting transitions (<100 °C), whereas those with two quinoxaline moieties per repeat unit
(QXP V2 and QXP V4) had relatively high glass transition temperatures (>140 °C). The polymers emit
blue to green light (404-536 nm) in dilute solution and blue-green to yellow light (470-563 nm) in the
solid state. The photoluminescence emission was well-described by single-exponential decay with lifetimes
ranging from 200 ps to 2.2 ns in both dilute solution and thin film, indicating lack of intermolecular
emissive species in the solid state. PPV derivatives with quinoxaline moieties in the main chain (QXP V1
and QXP V2) showed facile reversible electrochemical reductions with electron affinities of 2.63-2.75
eV. As emissive materials in LEDs, greenish-yellow electroluminescence with a brightness of up to 450
cd/m2 was obtained from single-layer diodes of QXP V1 with aluminum cathode in air.
In tr od u ction
transport properties while preserving the desired high
luminescence quantum yields. Toward this end, several
copolymers have been synthesized based on p-type
backbones incorporating an electron-deficient unit such
as pyridine,8 cyano group,9 oxadiazole,10,11 quinoline,12
or quinoxalines,12-16 either in the main chain or as
pendants to the main chain. Quinoxaline is a useful
n-type building block with high electron affinity and
good thermal stability. It has been successfully incor-
porated in small molecules and polymers for use as
electron-transport materials in multilayer OLEDs based
on PPV.17-20 In contrast to the vast literature on
oxadiazole-containing donor/acceptor copolymers, there
are only a few reports on quinoxaline-containing co-
polymers based on p-type emissive backbones such as
polyfluorene,12,13 PPV,14,15 and polycarbazole.16 To the
best of our knowledge, quinoxaline-containing PPVs
have not yet been explored as emissive materials for
OLEDs.
Organic light-emitting diodes (OLEDs) based on
conjugated polymers such as poly(p-phenylenevinylene)
(PPV) and polyfluorene have attracted much attention
in the past decade as promising candidates for the next
generation of full-color, flat-panel displays.1-7 One of the
many important challenges that still remain is the
improvement of the external quantum efficiency (EQE)
of OLEDs. Electroluminescence (EL) from OLEDs arises
from the radiative decay of excitons generated by the
recombination of electrons and holes injected from two
opposite electrodes into the emissive polymer layer.
Balanced rates of injection and transport of both elec-
trons and holes are essential to achieving high EQE in
an OLED. However, most emissive conjugated polymers
such as PPVs have much higher hole mobility than
electron mobility and low electron affinities, causing an
imbalance in charge injection and transport and thus
poor EQEs from single-layer OLEDs.1-7 Several ap-
proaches have been explored to improve the device
performance with varying degrees of success, including
the use of low work function cathodes such as calcium
(Ca) to improve electron injection and the utilization of
a separate electron-transport (n-type) material in mul-
tilayers or blends with an emissive p-type polymer.1-7
One promising approach to improving the external
quantum efficiency of OLEDs is to incorporate the
functions of hole and electron transport and light
emission into a single polymer, thereby creating bipolar
(donor/acceptor) emissive polymers.8-16 Rational molec-
ular design of emissive polymers with improved electron
affinity and balanced charge transport would allow the
fabrication of efficient single-layer OLEDs. One of the
challenges of this approach is to improve the charge
In this paper, we report the synthesis and character-
ization of four new PPV derivatives with one or two
quinoxaline moieties incorporated either directly in the
polymer backbone or as pendants to the main chain.
Four dibromo-substituted quinoxalines, namely 2,3-bis-
(4-bromophenyl)quinoxaline,2,2′-bis(4-bromophenyl)-
3,3′-diphenyl-6,6′-biquinoxalinyl, 2-(2,5-dibromophenyl)-
3-phenylquinoxaline, and 2,2′-(2,5-dibromo-1,4-phen-
ylene)bis(3-phenylquinoxaline), were synthesized and
polymerized via Heck coupling. Systematic variation in
the molecular structure of the copolymers via attach-
ment of quinoxaline units at selective positions on the
polymer backbone allows for detailed study of structure-
property relationships. The photophysical properties of
the copolymers in dilute solution and thin film and their
electrochemical properties were investigated. The co-
polymers were used as both emissive and electron-
transport materials in OLEDs.
* Corresponding authors. E-mail: makspil@upatras.gr; jenekhe@
u.washington.edu.
10.1021/ma048903q CCC: $27.50 © 2004 American Chemical Society
Published on Web 09/16/2004

7868 Karastatiris et al.
Macromolecules, Vol. 37, No. 21, 2004
through 0.2 µm syringe filters. Thin films of copolymers and
their 5 wt % blends with 1,1-bis(di-4-tolylaminophenyl)-
cyclohexane (TAPC) were spin-coated from these solutions onto
the PEDOT layer and dried at 50 °C in a vacuum overnight.
The film thicknesses obtained were ca. 40-50 nm which were
measured by an Alpha-Step 500 surface profiler (KLA Tencor,
Mountain View, CA). Finally, 130-150 nm Al electrodes were
thermally evaporated through a shadow mask onto the poly-
mer films using an AUTO 306 vacuum coater (BOC Edwards,
Wilmington, MA), typical evaporations being done at base
pressures lower than 2 × 10-6 Torr. The active area of each
EL device was 0.2 cm2. For the bilayer OLEDs with PPV, a
60 nm film of PPV was spin-coated directly on top of ITO from
its 0.6 wt % sulfonium precursor solution in water and cured
at 220 °C for 100 min under a vacuum. Thin films (∼25 nm)
of the quinoxaline copolymers were spin-coated on top of the
PPV layer from chloroform solutions and dried as aforemen-
tioned. Electroluminescence (EL) spectra were obtained using
a PTI QM-2001-4 spectrophotometer. Current-voltage char-
acteristics of the LEDs were measured using a HP4155A
semiconductor parameter analyzer (Yokogawa Hewlett-Pack-
ard, Tokyo). The luminance was simultaneously measured
using a model 370 optometer (UDT instruments, Baltimore,
MD) equipped with a calibrated luminance sensor head (model
211). The device external quantum efficiencies were calculated
using procedures reported previously.11,22 All the device fab-
rication and characterization steps were done under ambient
laboratory conditions.
Rea gen ts a n d Solven ts. Dimethylacetamide (DMAc) and
tetrahydrofuran (THF) were dried by distillation over CaH2.
Chloroform and triethylamine were dried by distillation over
P2O5 and KOH, respectively. 4,4′-Dibromobenzil and 1,2-
phenelynediamine were recrystallized from 1,4-dioxane and
toluene, respectively. 2,5-Didodecyloxy-1,4-divinylbenzene9a
and 1,4-diiodo-2,5-dibromobenzene23 were synthesized accord-
ing to known methods. All other reagents and solvents were
commercially purchased and were used as supplied.
Syn th esis of Mon om er s (Sch em es 1 a n d 2). 2,3-Bis(4-
br om op h en yl)qu in oxa lin e (1). A solution of 4,4′-dibro-
mobenzil (0.81 g, 2.20 mmol), 1,2-phenylenediamine (0.24 g,
2.20 mmol), and p-toluenesulfonic acid (23 mg) in CHCl3 (15
mL) was refluxed under N2 for 48 h. The solution was cooled
to room temperature and filtered. The solvent was subse-
quently evaporated under reduced pressure, and the residue
was recrystallized from acetonitrile to afford compound 1 (0.60
g, 62%) as a yellowish solid; mp 191-193 °C (lit.24 201-203
°C). Anal. Calcd for C20H12N2Br2: C, 54.58; H, 2.75; N, 6.36.
Found: C, 54.31; H, 2.77; N, 6.33. FT-IR (KBr, cm-1): 1586,
1556, 1448, 1390,1342, 1220, 1070, 1010, 976. 1H NMR (CDCl3,
ppm): 8.30-8.28 (m, 2H, at positions 5 and 8 of quinoxaline);
7.95-7.92 (m, 2H, at positions 6 and 7 of quinoxaline); 7.65-
7.63 (d, 4H, ortho to bromo); 7.55-7.53 (d, 4H, meta to bromo).
13C NMR (CDCl3, ppm): 152.31, 141.,67, 138.12, 132.08,
131.83, 130.78, 129.61, 124.10.
Exp er im en ta l Section
Ch a r a cter iza tion Meth od s. Melting temperatures were
determined on an electrothermal melting point apparatus
IA6304 and are uncorrected. IR spectra were recorded on a
1
Perkin-Elmer 16PC FT-IR spectrometer with KBr pellets. H
NMR (400 MHz) and 13C NMR (100 MHz) spectra were
obtained using a Bruker spectrometer. The NMR spectra were
recorded using DMSO-d6 or CDCl3 as solvent. Chemical shifts
(δ values) are given in parts per million with tetramethylsilane
as an internal standard. UV-vis absorption spectra were
recorded on a Perkin-Elmer model Lambda 900 UV/vis/near-
IR spectrophotometer. The photoluminescence (PL) emission
spectra were obtained with a Photon Technology International
(PTI) Inc. model QM-2001-4 spectrofluorimeter. GPC analysis
was conducted on a Waters Breeze 1515 liquid chromatograph
equipped with a 2410 differential refractometer as detector
(Waters Associate) and Styragel HR columns using polystyrene
as standard and THF as eluent. Differential scanning calo-
rimetry (DSC) and thermogravimetric analysis (TGA) were
performed on a DuPont 990 thermal analyzer system. Ground
polymer samples of about 10 mg each were examined by TGA,
and the weight loss comparisons were made between compa-
rable specimens. The DSC thermograms were obtained at a
heating rate of 10 °C/min in a N2 atmosphere at a flow rate of
60 cm3/min. Dynamic TGA measurements were made at a
heating rate of 20 °C/min in atmospheres of N2 or air at a flow
rate of 60 cm3/min. Thermomechanical analysis (TMA) was
recorded on a DuPont 943 TMA using a loaded penetration
probe at a scan rate of 10 °C/min in N2 with a flow rate of 60
cm3/min. The TMA experiments were conducted at least in
duplicate to ensure the accuracy of the results. The TMA
specimens were pellets of 8 mm diameter and 2 mm thickness
prepared by pressing powder of polymer for 3 min under 5-7
kpsi at ambient temperature. Elemental analyses were carried
out with a Hewlett-Packard model 185 analyzer.
To measure the PL quantum yields (Φf), polymer solutions
in spectral grade chloroform were prepared. The concentration
(∼10-5 M) was adjusted so that the absorbance of the solution
would be lower than 0.1. A 10-5 M solution of 9,10-dipheny-
lanthracene in toluene (Φf ) 0.93) was used as a standard.21
Cyclic voltammetry of the polymers was performed in aceto-
nitrile with 0.1 M tetrabutylammonium hexafluorophosphate
(TBAPF6) as the supporting electrolyte at scan rates of 20-
100 mV/s. Platinum wire electrodes were used as both counter
and working electrodes, and silver/silver ion (Ag in 0.1 M
AgNO3 solution, from Bioanalytical Systems, Inc.) was used
as a reference electrode. Using ferrocene as an internal
standard, the potential values obtained were converted to vs
SCE (saturated calomel electrode) and the corresponding
ionization potential (IP) and electron affinity (EA) values were
estimated from the onset redox potentials.
Tim e-Resolved P L Deca y Dyn a m ics. Fluorescence de-
cays of the polymers in solution and thin film were measured
on a PTI model QM-2001-4 spectrofluorimeter equipped with
a Strobe Lifetime upgrade. The instrument utilizes a nano-
second flash lamp filled with high purity nitrogen/helium (30/
70) mixture as an excitation source and a stroboscopic detec-
tion system. All measurements were done at room temperature.
The decay curves were analyzed using a multiexponential
fitting software package provided by the manufacturer. Re-
duced chi-square values, Durbin-Watson parameters, and
weighted residuals were used as the goodness-of-fit criteria.
F a br ica tion a n d Ch a r a cter iza tion of LEDs. The single-
layer OLEDs were fabricated as sandwich structures between
aluminum (Al) cathodes and indium-tin oxide (ITO) anodes.
ITO-coated glass substrates (Delta Technologies Ltd., Still-
water, MN) were cleaned sequentially in ultrasonic bathes of
detergent, 2-propanol/deionized water (1:1 volume) mixture,
toluene, deionized water, and acetone. A 50 nm thick hole
injection layer of poly(ethylenedioxythiophene) doped with
poly(styrenesulfonate) (PEDOT) was spin-coated on top of ITO
from a 0.7 wt % dispersion in water and dried at 150 °C for 1
h under a vacuum. The copolymer solutions in chloroform were
made at concentrations of 10-15 mg/mL and then filtered
1-Br om o-4-(p h en yleth en yl)ben zen e (2). 4-Bromoiodo-
benzene (1.00 g, 3.53 mmol), Pd(PPh3)2Cl2 (0.1240 g, 0.177
mmol), and CuI (0.0673 g, 0.353 mmol) were dissolved in THF
(20 mL), and triethylamine (10 mL) was added under nitrogen.
Phenylacetylene (0.36 g, 3.53 mmol) was added, and the
mixture was stirred at room temperature for 24 h. Then it was
filtered, and the volatile components were stripped off under
reduced pressure. The residue was purified by column chro-
matography with CH2Cl2/n-hexane (1/1 v/v) as eluent to afford
2 as a yellow solid (0.70 g, 77%); mp 86-88 °C (lit.25 87-88
°C). FT-IR (KBr, cm-1): 1488, 1442, 1392, 1068, 1008. 1H NMR
(CDCl3, ppm): 7.46-7.40 (m, 4H, ortho and meta to bromo);
7.33-7.27 (m, 5H, other aromatic). 13C NMR (CDCl3, ppm):
133.44, 132.92, 132.01, 128.93, 128.82, 123.32, 122.88, 122.66.
4-Br om oben zil (3). To a solution of 2 (0.53 g, 2.06 mmol)
in CH2Cl2 (10 mL), KMnO4 (1.03 g, 6.50 mmol), NaHCO3 (0.21
g, 2.50 mmol), and tetraethylammonium bromide (0.2 g) were
added by dissolving in water (10 mL), and the mixture was
rapidly stirred at room temperature for 48 h. The excess
KMnO4 was destroyed by addition of HCl and Na2SO3 until
the red color disappeared. The organic layer was separated,

Macromolecules, Vol. 37, No. 21, 2004
Poly(p-phenylenevinylene)s 7869
Sch em e 1
Sch em e 2
washed with sodium dicarbonate and water, and dried (Mg-
SO4). Evaporation of the solvent afforded compound 3 (0.58 g,
97%). It was recrystallized from methanol; mp 88-89 °C (lit.26
89-90 °C). FT-IR (KBr, cm-1): 1668, 1582, 1210, 1174, 1070,
a mixture of methanol/1,4-dioxane; mp >300 °C. Anal. Calcd
for C40H24N4Br2: C, 66.68; H, 3.36; N, 7.78. Found: C, 66.14;
H, 3.32; N, 7.76. FT-IR (KBr, cm-1): 1614, 1588, 1476, 1342,
1
1072, 1054, 1010, 978. H NMR (CDCl3, ppm): 8.60 (s, 2H, at
1
1010. H NMR (CDCl3, ppm): 7.97-7.95 (d, 2H, at positions
positions 5, 5′); 8.32-8.23 (m, 4H, at positions 7, 7′, 8, 8′);
7.61-7.39 (m, 18H, other aromatic). 13C NMR (CDCl3, ppm):
141.84, 141.73, 141.43, 141.35, 139.12, 138.28, 131.91, 130.35,
130.22, 130.09, 130.04, 129.56, 128.92, 127.87.
2′ and 6′); 7.85-7.84 (d, 2H, at positions 2 and 6); 7.69-7.65
(m, 3H, at positions 3′, 4′, and 5′); 7.54-7.51 (d, 2H, at positions
3 and 5). 13C NMR (CDCl3, ppm): 194.10, 193.53, 135.30,
133.34, 132.80, 132.30, 131.57, 130.77, 130.28, 129.41.
2,5-Dibr om oiod oben zen e (5). To a solution of 2,5-dibro-
moaniline (2.60 g, 10.4 mmol) in glacial acetic acid (5 mL),
96% H2SO4 (2 mL) was added. The mixture was cooled to 0
°C, and a solution of NaNO2 (2.14 g, 31.0 mmol) in water (5
mL) was added dropwise. After 1 h of stirring at 0 °C, urea
2,2′-Bis(4-br om op h en yl)-3,3′-d ip h en yl-6,6′-biqu in oxa li-
n yl (4). Compound 4 was prepared (63%) as a yellow solid
from the condensation of 3 with 3,3′-diaminobenzidine accord-
ing to the procedure described for 1. It was recrystallized from

7870 Karastatiris et al.
Macromolecules, Vol. 37, No. 21, 2004
(1.2 g) in H2O (4 mL) was added dropwise to destroy the excess
of NaNO2. Aqueous KI (5.16 g, 31.1 mmol) was added dropwise
at room temperature, and the mixture was stirred at room
temperature for 1 h. The brown precipitate was filtered,
washed with water, and dried. It was purified by column
chromatography with n-hexane as eluent affording compound
5 (2.60 g, 69%) as a white solid; mp 38-39 °C. FT-IR (KBr,
cm-1): 1544, 1436, 1352, 1246, 1096, 1078, 1002, 868, 810, 742,
516. 1H NMR (CDCl3, ppm): 7.99 (s, 1H, ortho to I); 7.47-
7.45 (d, 1H); 7.33-7.31 (d, 1H). 13C NMR (CDCl3, ppm):
142.64, 133.92, 132.96, 129.03, 121.58, 102.42.
Compounds 6, 9 and 7, 10 as well as 8, 11 (Scheme 2) were
synthesized according to the procedures described for the
corresponding compounds 1, 2, and 3 (Scheme 1), respectively.
1,4-Dibr om o-2-(p h en yleth yn yl)ben zen e (6). It was puri-
fied by column chromatography using n-hexane/CH2Cl2 (3:1
v/v) as eluent. Compound 6 was obtained as a white solid in
97% yield; mp 55-57 °C. FT-IR (KBr, cm-1): 3070, 2215, 1595,
1569, 1488, 1460, 1439, 1377, 1075, 1031, 876, 805, 745, 683,
565. 1H NMR (CDCl3, ppm): 7.68 (s, 1H, at position 3); 7.57-
7.55 (d, 2H, ortho to the triple bond); 7.47-7.45 (d, 1H, at
position 6); 7.37-7.35 (m, 1H at position 5 and 3H meta and
para to triple bond). 13C NMR (CDCl3, ppm): 136.03, 134.07,
132.76, 132.18, 129.40, 128.83, 127.74, 124.68, 122.84, 121.02,
95.67, 87.15.
purged with argon. DMAc (15 mL) and triethylamine (5 mL)
were added with a syringe, and the mixture was heated at
120 °C for 36 h. After cooling to room temperature, the solution
was filtered and the filtrate was poured in methanol. The
yellow precipitate was filtered, washed with methanol, and
dried to afford polymer QXP V1 (0.87 g, 87%). The polymer
was purified by dissolution in THF, filtration, and reprecipi-
tation in methanol. Mn ) 12 700; polydispersity (PDI) ) 2.3
(by GPC). Anal. Calcd for (C54H68N2O2)n: C, 83.46; H, 8.82; N,
3.60. Found: C, 82.55; H, 8.94; N, 3.53.
P olym er QXP V2. It was prepared in 82% yield from the
polymerization of 4 with 1,4-didocyloxy-2,5-divinylbenzene. Mn
) 15 600; PDI ) 2.1 (by GPC). Anal. Calcd for (C74H80N4O2)n:
C, 84.05; H, 7.63; N, 5.30. Found: C, 83.24; H, 7.68; N, 5.26.
P olym er QXP V3. It was prepared from the polymerization
of 8 with 1,4-didocyloxy-2,5-divinylbenzene in 74% yield. Mn
) 8200; PDI ) 1.8 (by GPC). Anal. Calcd for (C54H68N2O2)n:
C, 83.46; H, 8.82; N, 3.60. Found: C, 82.52; H, 8.71; N, 3.55.
P olym er QXP V4. It was prepared according to the afore-
mentioned method from the reaction of 11 with 1,4-didodecy-
loxy-2,5-divinylbenzene in 88% yield. Mn ) 14 500; PDI ) 2.6
(by GPC). Anal. Calcd for (C68H76N4O2)n: C, 83.22; H, 7.81; N,
5.71. Found: C, 82.31; H, 7.78; N, 5. 77.
Resu lts a n d Discu ssion
2,5-Dibr om oben zil (7). Recrystallized from ethanol 95%
(yield 57%); mp 126-128 °C. FT-IR (KBr, cm-1): 3082, 1672,
1594, 1570, 1452, 1380, 1198, 1178, 12086, 1026. 1H NMR
(CDCl3, ppm): 8.09-8.07 (d, 2H, at positions 2′ and 6′); 7.92
(s, 1H, at position 6); 7.71-7.67 (d,1H, at position 4); 7.58-
7.54 (m, 3H, at positions 3′, 4′, 5′); 7.50-7.48 (d, 1H, at position
3). 13C NMR (CDCl3, ppm): 193.00, 191.02, 138.27, 137.46,
135.36, 135.13, 132.74, 130.83, 129.34, 122.43, 120.54.
2-(2,5-Dibr om op h en yl)-3-p h en ylqu in oxa lin e (8). Puri-
fied by recrystallization from CH3CN (yield 99%); mp 188-
190 °C. Anal. Calcd for C20H12N2Br2: C, 54.58; H, 2.75; N, 6.36.
Found: C, 54.21; H, 2.78; N, 6.35. FT-IR (KBr, cm-1): 3058,
1560, 1542, 1474, 1444, 1400, 1376, 1334, 1214, 1086, 1066,
1024, 980. 1H NMR (CDCl3, ppm): 8.26-8.18 (m, 2H, at
positions 6 and 9); 7.88-7.81 (m, 2H, at positions 7 and 8);
7.70 (s, 1H, at positions 6 of pendant dibromophenyl); 7.53-
7.51 (d, 2H, at positions 3 and 4 of pendant dibromophenyl);
7.39-7.32 (m, 5H, of pendant phenyl). 13C NMR (CDCl3,
ppm): 153.73, 152.28, 142.56, 142.31, 140.98, 138.39, 134.72,
133.67, 131.18, 130.69, 130.02, 129.83, 129.63, 128.61, 121.83.
1,4-Dibr om o-2.5-bis(p h en ylet h yn yl)b en zen e (9). Re-
crystallized from CH3CN; mp 145-147 °C. FT-IR (KBr, cm-1):
Syn th esis a n d Ch a r a cter iza tion . Schemes 1 and
2 show the synthesis of dibromides 1, 4, 8, and 10.
Compound 1 was easily prepared from commercially
available 4,4′-dibromobenzil. The other dibromides were
successfully prepared via multistep procedures. The first
substantial step of the procedure was the synthesis of
bromo-substituted (phenylethynyl)benzenes 2, 6, and 9
that were prepared from Sonogashira coupling of 4-bro-
moiodobenzene with phenylacetylene. Their synthesis
was based on the higher reactivity iodobenzene relative
to bromobenzene toward Sonogashira coupling. As a
result, the bromoiodobenzenes were exclusively alky-
nylated through the iodo group at ambient tempera-
ture.27 The second step of the procedure involved the
oxidation of the triple bond to diketones 3, 7, and 10.28
Finally, the condensation of the latter as well as of the
commercially purchased 4,4′-dibromobenzil with 1,2-
phenylenediamine or 3,3′-diaminobenzidine afforded
dibromoquinoxalines 1, 4, 8, and 11.
1
3072, 1496, 1440, 1354, 1064, 1024. H NMR (CDCl3, ppm):
Polymers QXP V1, QXP V2, QXP V3, and QXP V4
(Chart 1) were prepared from dibromides 1, 4, 8, and
11, respectively, and 1,4-didodecyloxy-2,5-divinylben-
zene via Heck coupling. They were obtained in 74-88%
yields, and their number-average molecular weights
(Mn) ranged from 8200 to 15 600 with polydispersity
indices (PDI) of 1.8-2.6 (Table 1). The polymers had
molecular weights comparable to those of other poly-
mers obtained through Heck coupling.8a,9a,11 The struc-
7.79 (s, 2H, ortho to Br); 7.59-7.57 (m, 4H, ortho to triple
bonds); 7.39-7.37 (m, 6H, meta and para to triple bonds). 13
C
NMR (CDCl3, ppm): 136.44, 132.22, 129.54, 128.87, 126.84,
124.13, 122.74, 97.08, 87.22.
1,1′-(2,5-Dibr om o-1,4-p h en ylen e)bis(2-p h en yleth a n ed i-
on e) (10). Recrystallized from ethyl acetate; mp 205-207 °C.
FT-IR (KBr, cm-1): 3084, 1693, 1593, 1451, 1358, 1273, 1183,
1085. 1H NMR (DMSO-d6, ppm): 8.18 (s, 2H, ortho to Br);
8.12-810 (d, 4H, ortho to triple bond); 7.85-7.65 (m, 6H, meta
and para to triple bond). 13C NMR (DMSO-d6, ppm): 192.00,
190.03, 141.48, 136.78, 135.40, 132.46, 130.92, 129.44, 120.81.
2,2′-(2,5-Dibr om o-1,4-p h en ylen e)bis(3-p h en ylqu in oxa -
lin e) (11). Recrystallized from DMAc/H2O (4/1); mp >300 °C.
Anal. Calcd for C34H20N4Br2: C, 63.38; H, 3.13; N, 8.69. Found:
C, 62.94; H, 3.10; N, 8.72. FT-IR (KBr, cm-1): 3056, 1556, 1538,
1478, 1442, 1396, 1358, 1210, 1088, 1048, 1028. 1H NMR
(DMSO-d6, ppm): 8.24-8.18 (m, 4H, at positions 5 and 8 of
quinoxaline); 7.88-7.82 (m, 4H, at positions 6 and 7 of
quinoxaline); 7.65 (s, 2H, ortho to Br); 7.50-7.30 (m, 10H, of
pendant phenyls).
Syn th esis of P olym er s. P olym er QXP V1. The prepara-
tion of polymer QXP V1 is given as a typical example. A round-
bottom flask equipped with magnetic stirrer, reflux condenser,
and gas input-output was charged with a mixture of 1 (0.57
g, 1.30 mmol), 1,4-didocyloxy-2,5-divinylbenzene (0.65 g, 1.30
mmol), Pd(OAc)2 (0.0120 g, 0.054 mmol), and tri-o-tolylphos-
phine (0.0910 g, 0.299 mmol). The flask was evacuated and
1
tures of the polymers were verified by FT-IR, H NMR
spectroscopy, and elemental analysis. The FT-IR spectra
of the polymers displayed absorption features that are
consistent with their structures, at about 2924, 2850
(aliphatic), 1620-1460 (aromatic), 1210 (ether bonds),
and 970 cm-1 (trans olefinic bonds). The 1H NMR
spectra of all the polymers exhibited peaks at about
8.60-7.00 (aromatic and olefinic protons), 4.00 (O-
CH2-), and 1.80-0.81 ppm (-(CH2)10-CH3). The peaks
of olefinic protons resonated at 7.20-7.00 ppm, sup-
porting the formation of trans olefinic bond. In addition,
no peak was observed at 6.5-5.5 ppm, indicating the
absence of cis olefinic bond and terminal vinyl groups.
Figure 1 shows the 1H NMR spectrum of polymer
QXP V1 as a typical example, and Figure 2 shows the
13C NMR of polymer QXP V3.

Macromolecules, Vol. 37, No. 21, 2004
Poly(p-phenylenevinylene)s 7871
Ch a r t 1
F igu r e 1. 1H NMR spectrum of polymer QXP V1 in CDCl3.
F igu r e 2. 13C NMR spectrum of polymer QXP V3 in CDCl3.
Ta ble 1. Molecu la r Weigh ts a n d Th er m a l P r op er ties of
P olym er s
a
a
polymer
Mn
Mw/Mn Tgb (°C) Tmc (°C) Tdd (°C) Yce (%)
QXP V1 12 700
QXP V2 15 600
2.3
2.1
1.8
2.6
n.o.f
140
n.o.
155
70
n.o.
90
355
370
340
380
30
52
41
60
QXP V3
8 200
QXP V4 14 500
n.o.
a
Molecular weights were determined by GPC using polystyrene
standards. Glass transition temperature. c Melting temperature.
b
d
Decomposition temperature in nitrogen (at which 1% weight loss
was observed). e Anaerobic char yield at 800 °C. f n.o. ) not
F igu r e 3. TMA traces of the polymers.
observed.
mL in THF or chloroform. The poor solubility made it
difficult to spin-coat uniform thick films of QXP V4 for
making LEDs; however, thin films for photophysical
studies could be solvent cast on glass substrates.
The thermal behavior of the polymers was investi-
gated by DSC and TMA, and their thermal stability was
evaluated by TGA. Figure 3 shows the TMA traces of
the four polymers. Polymers QXP V1 and QXP V3,
bearing one quinoxaline moiety per repeat unit, exhib-
ited melting transition (Tm) at 70 and 90 °C, respec-
tively. This transition was detected by both DSC and
TMA. Polarized microscopy showed that these polymers
turn into an isotropic melt and confirmed the absence
of thermotropic behavior. Compared to analogous di-
alkoxy-substituted PPVs, the Tm of the present polymers
is lowered by more than 60 °C with incorporation of the
quinoxaline moieties.29a No Tgs were observed for poly-
Qualitative studies demonstrated that the polymers
were soluble in common organic solvents such as THF,
CHCl3, CH2Cl2, 1,2-dichloroethane, 1,1,2,2-tetrachloro-
ethane, chlorobenzene, and toluene due to the long
dodecyloxy side chains. In addition, they were slightly
soluble in trifluoroacetic acid and partially soluble in
formic acid due to the protonation of the quinoxaline
rings. Polymers QXP V1 and QXP V3, which contain one
quinoxaline ring per repeat unit, exhibited higher
solubility than QXP V2 and QXP V4, which carry two
quinoxaline rings per repeat unit. In addition, the
polymers that bear quinoxaline moieties in the main
chain were more soluble than the polymers with pen-
dant quinoxalines units. QXP V4 was the least soluble
polymer due to the two bulky pendant quinoxaline
groups with a maximum solubility of only about 3 mg/

7872 Karastatiris et al.
Macromolecules, Vol. 37, No. 21, 2004
mers QXP V1 and QXP V3 by DSC or TMA. However,
polymers QXP V2 and QXP V4 carrying two quinoxaline
groups per repeat unit showed clear Tg at 140 and 155
°C, respectively. These Tgs were determined by the TMA
method utilizing a loaded penetration probe as the onset
temperature of the first inflection point of the corre-
sponding TMA traces (Figure 3). The Tgs of the present
polymers are more than 100 °C higher than those of
analogous dialkoxy-substituted PPVs.29a The absence of
melting and the high Tg of QXP V2 and QXP V4 reflect
their higher rigidity due to the presence of two quinoxa-
line moieties per repeat unit. Similar enhancements of
Tg were observed for polyfluorenes with the incorpora-
tion of the quinoxaline units in their main chains.12a
Wide-angle X-ray diffractograms of the “as-prepared”
powder of polymers revealed that QXP V1 possessed a
certain degree of crystallinity whereas QXP V2 was
completely amorphous. Besides, QXP V2 and QXP V4
exhibited a transition at about 50 °C, which was
attributed to side chain melting in accordance with
previous didodecyloxy-substituted PPVs.29a All the poly-
mers showed satisfactory thermal stability, being stable
up to 340-380 °C under nitrogen. Polymers containing
two quinoxaline moieties per repeat unit (QXP V2 and
QXP V4) were relatively more stable than those with
one quinoxaline per repeat unit (QXP V1 and QXP V3)
due to the excellent thermal stability characteristics
inherent to the quinoxaline group. The anaerobic char
yields of the polymers at 800 °C ranged from 30 to 60%.
The thermal properties of the polymers are collected in
Table 1.
P h otop h ysica l P r op er ties. The optical absorption
and photoluminescence (PL) emission spectra of the four
polymers in dilute (10-5 M) chloroform solution are
shown in Figure 4, parts a and b. All polymers except
QXP V4 showed a broad low-energy absorption band
around 400 nm, typical of the π-π* transition of the
conjugated PPV backbone.29 Polymer QXP V1 with one
quinoxaline unit in the backbone showed an absorption
peak at 402 nm, whereas polymer QXP V3 with one
pendant quinoxaline unit had the maximum located at
420 nm. QXP V1 had a higher absorption onset than
QXP V3, suggesting a smaller conjugation length in
QXP V1. This can be explained by the greater interrup-
tion in conjugation caused by the presence of the
quinoxaline unit in the PPV backbone in QXP V1 vs
pendant substitution in QXP V3. Polymer QXP V2 with
two quinoxaline units in the backbone had a higher
absorption maximum at 388 nm, while QXP V4 had the
most blue-shifted peak at 354 nm with a significantly
higher absorption onset at ∼390 nm. This suggests a
drastic reduction in electron delocalization in QXP V4
caused by substantial twisting of the backbone by the
two bulky pendant quinoxaline units, given that poly-
mer chains in dilute solution possess a high degree of
rotational freedom. The high-energy absorption peak at
∼350 nm was observed only in polymers QXP V3 and
QXP V4 with pendant substitution of quinoxalines,
suggesting that it could be associated with the aromatic
quinoxaline moieties rather than the PPV backbone.
Overall, we note that the absorption features in all four
polymers are blue-shifted compared to typical dialkoxy-
substituted PPVs,29 indicating that the conjugation
length is decreased by inserting the quinoxaline moi-
eties either in the main chain or as pendants.
F igu r e 4. (a) Normalized optical absorption spectra of 10-6
M solutions of the four copolymers in chloroform. (b) Normal-
ized PL emission spectra of 10-6 M solutions of the copolymers
in chloroform. Excitation wavelength is 380 nm in all cases
except for QXP V4 (360 nm).
mers. The introduction of an electron-donating or -ac-
cepting group directly affects the HOMO and LUMO
levels of the polymer. In our case, the insertion of the
electron-withdrawing quinoxalines on a PPV backbone
changed drastically its properties. The introduction of
one quinoxaline moiety caused a blue shift of the UV-
vis spectra of polymers relative to alkoxy-PPV. This shift
was larger in the case of QXP V1, which contains the
quinoxaline unit in the main chain, than QXP V3, where
the quinoxaline is laterally attached to the polymer
backbone. This might be due to the direct participation
of quinoxaline in the π-π* transition of the main chain
of polymer QXP V1. In the case of QXP V3, a reasonable
noncoplanar assembly of quinoxaline side group and the
polymer backbone may reduce the effects of quinoxaline
unit on polymer, hence a smaller blue shift of the UV-
vis spectrum than that of QXP V1. The introduction of
a second quinoxaline group either in the main chain or
as a side group brought about an additional blue shift
of absorption and emission of the polymer QXP V2 and
QXP V4, respectively. It is believed that it is a result of
steric effects due to insertion of the second bulk sub-
stituent rather than their effect on electronic properties
of the polymers. More insight into the effects of qui-
noxalines moieties on the polymers properties appears
in the Electrochemical Properties section.
It is known that the electronic character of a sub-
stituent influences the optical properties of the poly-

Macromolecules, Vol. 37, No. 21, 2004
Poly(p-phenylenevinylene)s 7873
Ta ble 2. P h otop h ysica l P r op er ties of P olym er s
a
b
a
b
λa,max in
λf,max in
λa,max in
λf,max in
polymer
solution (nm)
solution (nm)
Φfc
thin film (nm)
thin film (nm)
Egd (eV)
QXP V1
QXP V2
QXP V3
QXP V4
402
388
519
419, 536
504
0.55
0.34
0.21
0.19
404
397
350, 425
355
539
547
563
470
2.59
2.57
2.38
3.05
348,e 420
354
404
a
b
λ
a,max: the absorption maxima in chloroform solution or in thin film.
λf,max: the PL emission maxima in chloroform solution or in
thin film. c Φf: PL quantum yield in chloroform solution. Eg: optical band gap calculated from thin film absorption edge. e Italic numerical
values denote absolute maxima.
d
Figure 4b shows the PL emission spectra of the four
polymers in dilute chloroform solution. The excitation
wavelength used is 380 nm in all cases except for
polymer QXP V4 (360 nm). Polymers QXP V1 and
QXP V2 with quinoxalines in the main chain had
emission maxima at 519 and 536 nm, respectively.
However, QXP V3 and QXP V4 had blue-shifted emis-
sions at 504 and 404 nm, respectively. One would have
expected that QXP V3 would have the most red-shifted
emission due to its longer conjugation length. However,
polymer QXP V2 shows the lowest energy emission band
at 536 nm. Since QXP V2 has the long docecyloxy side
chains spaced farthest apart per repeat unit due to two
quinoxaline units, it is likely that there could be
planarization of the backbone in the excited state
leading to the bathochromic shift in its emission. The
additional peak at 419 nm could be likely due to the
exciton localization on the two neighboring quinoxaline
units. It is rather surprising that the presence of an
additional pendant quinoxaline unit in QXP V4 com-
pared to QXP V3 blue shifts its emission maximum
significantly by 100 nm. Compared to analogous di-
alkoxy-substituted PPVs,29 the PL emission spectra of
the present quinoxaline-containing PPVs are not only
blue-shifted but also lack any well-resolved vibronic
structure due to apparent lack of intrachain order. The
main photophysical properties of the quinoxaline PPVs
are summarized in Table 2.
The absorption spectra of thin films of the four
quinoxaline PPVs are shown in Figure 5a. They are very
similar to the dilute solution absorption spectra in terms
of the spectral shapes and peak positions. The lowest
energy transitions of polymers QXP V1 and QXP V3
were located at 404 and 425 nm, respectively. Polymers
QXP V2 and QXP V4 with two quinoxalines per repeat
unit showed blue-shifted transitions at 397 and 355 nm,
F igu r e 5. (a) Normalized optical absorption spectra of the
respectively. The similarity of the thin film spectra to
the dilute solution spectra suggests comparable ground-
state electronic structures with minimal interchain
aggregation even in the condensed state. The optical
band gaps derived from the absorption edge of the thin
film spectra gave similar values of 2.4-2.6 eV for all
polymers except QXP V4, which had a higher band gap
of ∼3.0 eV. Figure 5b shows the PL emission spectra of
the polymer thin films with excitation being done at the
low-energy absorption maximum of each polymer. The
polymers emit greenish-yellow light in thin film, except
QXP V4 which emits greenish-blue color. The emission
maxima of polymers QXP V1 and QXP V2 are red-
shifted by 10-20 nm compared to their corresponding
dilute solution spectra (Figure 4b). This can be ex-
plained as due to increased exciton delocalization in the
thin films due to the interchain π-stacking. The red shift
of ∼60-65 nm in the emission maxima of polymers
QXP V3 and QXP V4 in going from dilute solution to the
thin film state is much greater. Given that the rotational
freedom of the polymer chains in the solid state is much
four copolymer thin films. (b) Normalized PL emission spectra
of the four copolymer thin films. In each case, the excitation
wavelength is the absorption maximum in (a) above.
reduced, the pendant quinoxaline units could force the
PPV backbones in QXP V3 and QXP V4 to adopt more
planar conformations in the solid state than in solution,
leading to the much red-shifted emission peaks at 563
and 470 nm, respectively. Besides, we observe a slight
vibronic structure in the PL spectra of polymers QXP V1
and QXP V3, reminiscent of the typical emission spectra
of dialkoxy-substituted PPVs.29 This could be due to
some interchain ordering due to a certain degree of
crystallinity in these two polymer films as evidenced by
WAXD patterns (not shown).
The fluorescence quantum yields (φf) of the four
polymers in dilute chloroform solution are listed in
Table 2. The highest φf value of 55% was observed for
polymer QXP V1, which is higher than that of MEH-
PPV and comparable to CN-PPV.30 Polymers QXP V3
and QXP V4 had the lowest φf values of ∼20%. The

7874 Karastatiris et al.
Macromolecules, Vol. 37, No. 21, 2004
Ta ble 3. F lu or escen ce Deca y Lifetim es of P olym er
Solu tion s a n d Th in F ilm s
longer than those observed for thin films of PPV or
MEH-PPV.30 Although we did not measure the fluo-
rescence quantum yields in thin films, visual observa-
tion of the films under excitation indicates higher
quantum yields than PPV, with QXP V1 being the most
fluorescent. The drastic reduction in lifetime from 1.61
to 0.19 ns in going from solution to thin films of QXP V3
suggests a greater number of channels for nonradiative
decay of excitons. The greater conformational defects
along the polymer backbone introduced by the bulky
pendant quinoxaline units could be preferential sites for
such nonradiative processes. We could not obtain good
fluorescence decay curves for films of QXP V4 due to the
poor quality of the solvent cast films and their low
fluorescence yields. The general approach of incorporat-
ing quinoxaline units in the main chain or as pendants
seems to reduce the propensity to form interchain
excitations and excimer-like emissions in thin films,
unlike in CN-PPV thin films where efficient excimer
emission is clearly observed.30a It is also interesting that
relatively high PL emission quantum yields were ob-
served in polymers QXP V1 and QXP V2 given that
these copolymers have a donor/acceptor architecture.
Strong intramolecular charge transfer in donor/acceptor
conjugated polymers is usually a major source of dra-
matic luminescence quenching in such materials.32 Our
photophysical results imply that there is little or no
contribution of intramolecular charge transfer to the
photoexcitations of the donor/acceptor copolymers QX-
P V1 and QXP V2.
a
b
λexc
λem
polymer
(nm)
(nm)
τc (ns)
ø2
DWd
solution
QXP V1
QXP V2
QXP V3
QXP V4
thin film
QXP V1
QXP V2
QXP V3
381
381
381
358
519
536
504
404
1.637
2.201
1.611
1.55
1.106
1.08
1.072
1.08
1.716
1.816
1.541
2.04
381
381
381
547
539
563
0.604
1.167
0.197
1.173
0.9267
0.9435
1.801
1.947
2.091
Excitation wavelength. b Monitored emission wavelength. c Flu-
orescence lifetime extracted from the single-exponential fits.
a
d
Durbin-Watson parameter for the fits. Concentration of all
solutions was 10-5 M in chloroform.
Ta ble 4. Red ox P oten tia ls a n d Electr on ic Str u ctu r e
P a r a m eter s of P olym er s
a
a
el
Eonset
Eonset
IP
EA
Eg
polymer oxidation (V) reduction (V) (eV)b (eV)c (eV)d
QXP V1
QXP V2
QXP V3
QXP V4
0.88
0.87
0.75
0.90
-1.77
-1.65
-1.75
-1.70
5.28
5.27
5.15
5.30
2.63
2.75
2.65
2.70
2.65
2.52
2.50
2.60
a
b
Eonset ) onset potential vs SCE. Determined from the onset
oxidation potential. c Determined from the onset reduction poten-
d
tial. Electrochemical band gap ) IP - EA.
polymers with quinoxaline pendant units seemed to
have lower φf values than the ones with backbone
substitution of quinoxalines. Thus, by subtle variations
in the number and point of attachment of the quinoxa-
line moieties on the PPV backbones, we observed
systematic variations in the steady-state photophysical
properties of the four polymers.
Electr och em ica l P r op er ties. To understand the
variation in the electronic structure of the PPV deriva-
tives by incorporation of the electron-deficient quinoxa-
line moieties, we performed cyclic voltammetry (CV)
measurements on polymer films. The redox scans of
polymers containing quinoxaline moieties in the main
chain, QXP V1 and QXP V2, are shown in Figure 6,
parts a and b, respectively. Several reduction scans were
taken at increasing scan rates ranging from 20 to 100
mV/s in steps of 20 mV. QXP V1 showed a one-peak
reversible reduction with a formal potential of -1.85 V
vs SCE and an onset potential of -1.77 V vs SCE.
However, the oxidation scan was completely irreversible
with onset potential of +0.88 V vs SCE. Taking -4.4
eV as the SCE energy level relative to vacuum,33
electron affinity (EA) and ionization potential (IP)
energy levels of 2.63 and 5.28 eV were estimated. The
presence of an additional quinoxaline moiety in the
main chain in polymer QXP V2 leads to the emergence
of a second reversible reduction peak with a formal
potential of -1.68 V vs SCE, in addition to the peak at
-1.85 V seen in QXP V1. The reduction scans on both
polymers are highly reversible with repeatable scans
made on the same thin film samples without any
changes in their peak characteristics. The onset reduc-
tion potential of -1.65 V leads to an EA value of 2.75
eV for QXP V2, which is slightly higher than that of
QXP V1 as would be expected with the additional
quinoxaline moiety. Compared to the EA values of
analogous dialkoxy-substituted PPVs such as MEH-
PPV (EA ) 2.9-3.0 eV), the EA values of the present
polymers are slightly lower despite the presence of
electron-deficient quinoxaline moieties in their back-
bones. Additionally, unlike in MEH-PPV, the oxidation
scans of the current polymers are irreversible, suggest-
ing that the ease of p-doping the PPV backbone is lost
on introduction of the quinoxaline moieties.
Tim e-Resolved P L Deca y Dyn a m ics. To further
shed light on the nature of the emission from the
quinoxaline-containing polymers, we investigated the
fluorescence decays of the polymer PL emission bands
in solution and thin film. A summary of the fluorescence
decay parameters is presented in Table 3. In dilute
solution (10-5 M), the decay of the lowest energy
emission band for all the polymers was found to be
single-exponential with lifetimes ranging from 1.5 to 2.2
ns. This suggests that the emission originates from
intrachain singlet excitons as would be expected for
polymer chains with minimal interchain interactions.
These lifetimes are longer than the typical lifetimes
observed for other PPV-based polymers such as MEH-
PPV and CN-PPV in dilute solution.30 Knowledge of
the fluorescence quantum yields in solution and the
fluorescence lifetimes can give information on the
natural radiative lifetimes (τ0) of the emission bands.
τ0 values of 2.98 and 7.67 ns were obtained for the PL
emission of QXP V1 and QXP V3, respectively. The
much smaller τ0 value in QXP V1 suggests a signifi-
cantly faster radiative rate constant leading to the
highest fluorescence yield observed among the four
polymers.
In going from the polymer dilute solutions to thin
films, the PL emission lifetimes decreased by more than
50%. However, the fact that the fluorescence was still
well-described by single-exponential decay and that the
steady-state PL spectra showed a slight vibronic struc-
ture ruled out any interchain related emissive species
such as excimers in the thin films.31 The fluorescence
lifetimes of QXP V1 (604 ps) and QXP V2 (1.16 ns) are

Macromolecules, Vol. 37, No. 21, 2004
Poly(p-phenylenevinylene)s 7875
F igu r e 6. Cyclic voltammograms of (a) QXP V1 and (b)
QXP V2 in 0.1 M TBAPF6 in acetonitrile. The scan rates for
the reduction waves range from 20 to 100 mV/s.
F igu r e 7. Cyclic voltammograms of QXP V3 and QXP V4 in
0.1 M TBAPF6 in acetonitrile: (a) reduction waves and (b)
oxidation waves. The scan rate used was 40 mV/s.
Figure 7, parts a and b, shows the reduction and
oxidation waves at scan rates of 40 mV/s, respectively,
of the polymers with pendant quinoxaline units QXP V3
and QXP V4. We observed a reversible reduction and
irreversible oxidation for QXP V3 with one pendant
quinoxaline unit. EA of 2.65 eV and IP of 5.15 eV were
estimated from the onset redox potentials. However, the
reduction waves were not as well-defined and repeatable
as with polymers QXP V1 and QXP V2. Similar char-
acteristics were observed in the redox waves of QXP V4
where complete reversibility in the reduction waves
could not be captured due to dissolution of the doped
polymer in the electrolyte after the first uptake of an
electron. The poor characteristics of the redox waves of
QXP V4 could be in part due to the poor quality of the
polymer thin films on the Pt electrodes. However,
comparing Figures 6 and 7, it is evident that the
polymers with quinoxaline moieties in the main chain
(QXP V1 and QXP V2) are more amenable to n-type
doping than those with pendant quinoxaline units
(QXP V3 and QXP V4). One plausible reason for this
trend could be the fact that the quinoxaline moieties in
the main chain directly participate in the conjugation
along the polymer backbone, whereas the pendant
quinoxalines do not. In fact, the redox scans in Figure
6 are very reminiscent of typical n-type conjugated
polymers such as the polyquinolines3,4,6 and polyben-
zobisazoles.7b The incorporation of electron-deficient
quinoxalines in the polymer backbones has apparently
transformed the PPV-based polymers into n-type poly-
mers, suggesting better electron-transport characteris-
tics than hole transport. A similar transformation of
PPV to n-type polymers was previously observed by
incorporation of one or two oxadiazole moieties in the
backbone of PPV-based polymers.11 Thus, we do not
realize true bipolar behavior in these donor-acceptor
type quinoxaline PPVs, which has implications in rela-
tion to charge injection properties for light-emitting
diodes (LEDs) based on these polymers as discussed
next.
EL Device P r oper ties. Light-emitting diodes (LEDs)
based on the quinoxaline copolymers as both emissive
and electron transport materials were made and inves-
tigated. Figure 8a shows the current density-voltage-
luminance (J -V-L) characteristics of single-layer QX-
P V1 LEDs of the type ITO/PEDOT/polymer/Al. Diodes
comprising of neat polymer layer had a turn-on voltage
of ca. 6 V with a maximum brightness of 120 cd/m2 and
maximum external quantum efficiency (EQE) of 0.012%
at 12.5 V (115 cd/m2, 260 mA/cm2). The corresponding
EL spectra shown in Figure 8b were similar to the thin
film PL spectra with EL maxima centered at ∼535 nm.
There was a slight blue shift from 535 nm (7.0 V) to
528 nm (13 V) with increase in applied bias. This
suggests that the greater local heating in the polymer
film at the high voltages probably leads to conforma-
tional changes of the polymer backbone, especially given
that QXP V1 has a melting transition at 70 °C (Table
1). Although the device performance of this simple

7876 Karastatiris et al.
Macromolecules, Vol. 37, No. 21, 2004
F igu r e 9. (a) Current density-voltage-luminance charac-
teristics of LEDs from polymer QXP V2 and its 95 wt % blend
with TAPC. (b) EL spectra of the blend device.
F igu r e 8. (a) Current density-voltage-luminance charac-
teristics of LEDs from polymer QXP V1 and its 95 wt % blend
with TAPC. (b) EL spectra of the neat polymer device. The
inset shows the device schematic.
11 V (23 cd/m2, 75 mA/cm2). The EL spectra of the blend
LED are shown in Figure 9b. The EL maxima were
located at ∼538-543 nm and were comparable to the
thin film PL maximum of QXP V2 at 547 nm. It is rather
surprising that although there is no significant differ-
ence in the EA/IP levels and electrochemical character-
istics of QXP V1 and QXP V2, the former is a much more
emissive material than the latter. The poor performance
of QXP V2 as an emissive material in LEDs could be
explained as being partly due to the very low fluores-
cence quantum yields in the solid state. Although the
solid-state PL quantum yields were not calculated, we
note that QXP V1 had the highest PL quantum yield
(φf ) 0.55) in solution among the series of polymers.
The LED characteristics of QXP V3 are shown in
Figure 10; part a shows the J -V-L characteristics of
QXP V3 LEDs, and part b displays the EL spectra of
the neat polymer diode. The neat polymer device had a
low turn-on voltage of 4.0 V and a maximum brightness
of 35 cd/m2 at 12.0 V. The shape of the J -V curve was
typical of space-charge-limited currents in semiconduc-
tors with traps,34 possibly due to poor electron injection
and transport as evidenced by the inferior reduction
characteristics in CV. The EL spectra of the neat LED
shown in Figure 10b indicate yellow emission with
maxima at 550 nm. A small improvement in device
performance was obtained on addition of 5 wt % TAPC
to QXP V3. The maximum brightness was 50 cd/m2 with
very low EQE of 0.004% at 11 V (48 cd/m2, 420 mA/
cm2). As the IP of QXP V3 is already low at 5.15 eV,
single-layer LED is satisfactory, it could be further
improved by adding appropriate charge transport ma-
terials. The irreversible oxidation in the CV of QXP V1
(Figure 6a) suggests lack of bipolar character and poor
hole transport that could limit the device performance.
Thus, a blend of QXP V1 was made with a hole transport
molecule, 5 wt % 1,1-bis(di-4-tolylaminophenyl)cyclo-
hexane (TAPC), to enhance the hole transport through
the device. The J -V-L characteristics of the corre-
sponding LED are also shown in Figure 8a. A maximum
brightness of 450 cd/m2 and a maximum EQE of 0.06%
at 17.5 V (435 cd/m2, 225 mA/cm2) were obtained,
representing factors of 4-5 enhancement in perfor-
mance compared to the QXP V1 LED. Devices made
from similar blends of QXP V1 with a hole-transporting
PVK polymer showed slightly lower efficiencies due to
higher operating voltages. Further improvements in EL
device efficiencies are likely by optimizing the blend
composition. The EL spectra of the blend LED (not
shown) are very similar to those of neat QXP V1 with
yellow-green emission at all voltages.
Figure 9a shows the J -V-L characteristics of QX-
P V2 LEDs, both neat polymer and blends with TAPC.
The neat polymer device did not show any measurable
luminance until 13 V even at the high current densities
of 500 mA/cm2, although the weak EL was visible to the
eye. On addition of 5 wt % TAPC to QXP V2, a low
brightness of 23 cd/m2 was seen with EQE of 0.01% at

Macromolecules, Vol. 37, No. 21, 2004
Poly(p-phenylenevinylene)s 7877
F igu r e 10. (a) Current density-voltage-luminance charac-
teristics of LEDs from polymer QXP V3 and its 95 wt % blend
with TAPC. (b) EL spectra of the neat polymer device.
F igu r e 11. (a) Current density-voltage-luminance charac-
teristics of LEDs from single-layer PPV and bilayer with
QXP V1. (b) EL spectra of both devices at drive voltage of 11
V.
the addition of 5 wt % TAPC (IP ) 5.3 eV) apparently
did not improve hole transport in the diodes signifi-
cantly.
respectively, were obtained using a separate layer of
QXP V1. The EL spectra of the two types of devices
shown in Figure 11b show typical PPV emission with
two well-resolved peaks at 522 and 555 nm, suggesting
that QXP V1 solely serves as the ETM with recombina-
tion occurring in the PPV layer. No enhancements in
brightness were observed using QXP V2 as an ETM, in
spite of its similar energy levels to QXP V1. QXP V3 did
not work at all as an ETM with no measurable EL from
the bilayer PPV/QXP V3 LED. This suggests very poor
electron transport properties of QXP V3 as evidenced
in its lack of reversibility in CV measurements. As an
ETM for PPV-based LEDs, we note that QXP V1 is
inferior compared to previous polyquinoxalines20 and
some polyquinolines.3b,4,6b This could be explained as
due to its poorer hole-blocking and electron-transport
properties. The low IP of 5.28 eV in QXP V1 does not
present a significant barrier to hole injection at the PPV/
QXP V1 interface. Also, the electron mobilities through
QXP V1 would be much lower due to the long dodecyloxy
side chains preventing efficient interchain hopping of
charges. Both these factors were much improved in the
conjugated rigid polyquinolines and polyquinoxalines
leading to their superior performance as an ETM for
PPV LEDs.
In the case of polymer QXP V4, EL devices could not
be fabricated due to the very poor film quality caused
by the low solubility of the polymer in any solvent
combination. Overall, polymer QXP V1 with one qui-
noxaline moiety in the backbone is the best emissive
material among the four quinoxaline-containing PPVs,
most likely due to its high fluorescence quantum yield
and good electron transport characteristics. To the best
of our knowledge, this is one of the first systematic
investigation of EL properties of quinoxaline-containing
dialkoxy-PPVs. Compared to some previous studies on
oxadiazole-containing PPVs, the present quinoxaline
polymer QXP V1 is much superior as an emissive
material for LEDs.10a,b,11 Besides, the good reversibility
of the electrochemical reductions in QXP V1 and QX-
P V2 suggested that they could be used as electron-
transport materials (ETMs) in conjunction with other
p-type emissive polymers for multilayer LEDs.
The electron-transport properties of the quinoxaline-
PPVs were investigated by fabricating bilayer LEDs of
the type ITO/PPV/polymer/Al, with PPV serving as the
emissive material. The J -V-L characteristics of single-
layer PPV and bilayer PPV LEDs with QXP V1 are
shown in Figure 11a. Single-layer PPV diodes had a
maximum brightness of ∼20 cd/m2 and maximum EQE
of 0.001% at 11 V (22 cd/m2, 500 mA/cm2). Enhance-
ments in brightness and EQE by factors of 5 and 10,
Con clu sion s
Four new dialkoxy-substituted poly(p-phenylene-
vinylene)s, containing one or two quinoxaline moieties

7878 Karastatiris et al.
Macromolecules, Vol. 37, No. 21, 2004
(9) (a) Peng, Z.; Galvin, M. E. Chem. Mater. 1998, 10, 1785. (b)
Liu, M. S.; J iang, X.; Liu, S.; Herguth, P.; J en, A. K.-Y.
Macromolecules 2002, 35, 3532.
per repeat unit in the main chain or side chain, were
prepared by Heck coupling of dibromoquinoxalines with
2,5-didodecyloxy-1,4-divinylbenzene. Polymers contain-
ing one quinoxaline moiety per repeat unit (QXP V1 and
QXP V3) displayed melting at low temperatures, while
polymers with two quinoxalines per repeat unit (QXP V2
and QXP V4) had high glass transition temperatures.
The polymers emitted blue to green light (404-536 nm)
in dilute chloroform solution and blue-green to yellow
light (470-563 nm) in thin film. Time-resolved photo-
luminescence decay dynamics of the polymers revealed
single-exponential lifetimes ranging from 200 ps to 2.2
ns in both dilute solution and thin film. Typical n-type
characteristics were seen in polymers bearing quinoxa-
line moieties in the main chain (QXP V1 and QXP V2)
as revealed by their facile reversible reductions and
irreversible oxidations in cyclic voltammetry. As emis-
sive materials, greenish-yellow electroluminescence with
a brightness of up to 450 cd/m2 was obtained from
single-layer blend LEDs of QXP V1 with a triarylamine
molecule. As electron transport materials in PPV bilayer
LEDs, only moderate improvements in device perfor-
mance were achieved. These results constitute the first
systematic investigation of the photophysical and emis-
sive properties of quinoxaline-containing PPV polymers
for LED applications.
(10) (a) Peng, Z.; Bao, Z.; Galvin, M. E. Adv. Mater. 1998, 10, 680.
(b) Bao, Z.; Peng, Z.; Galvin, M. E.; Chandross, E. A. Chem.
Mater. 1993, 5, 1201. (c) Song, S.-Y.; J ang, M. S.; Shim, H.-
K.; Hwang, D.-H.; Zyung, T. Macromolecules 1999, 32, 1482.
(11) (a) Mikroyannidis, J . A.; Spiliopoulos, I. K.; Kasimis, T. S.;
Kulkarni, A. P.; J enekhe, S. A. Macromolecules 2003, 36,
9295. (b) Mikroyannidis, J . A.; Spiliopoulos, I. K.; Kasimis,
T. S.; Kulkarni, A. P.; J enekhe, S. A. J . Polym. Sci., Part A:
Polym. Chem. 2004, 42, 2112.
(12) (a) Zhan, X.; Liu, Y.; Wu, X.; Wang, S.; Zhu, D. Macromol-
ecules 2002, 35, 2529. (b) Sun, Q.; Zhan, X.; Yang, C.; Liu,
Y.; Li, Y.; Zhu, D. Thin Solid Films 2003, 440, 247.
(13) J ung, S. H.; Suh, D. H.; Cho, H. N. Polym. Bull. (Berlin) 2003,
50, 251.
(14) J onforsen, M.; J ohansson, T.; Inganas, O.; Andersson, M. R.
Macromolecules 2002, 35, 1638.
(15) Lee, B. H.; J aung, J . Y.; Cho, J .-W.; Yoon, K. J . Polym. Bull.
(Berlin) 2003, 50, 9.
(16) Morin, J .-.F.; Leclerc, M. Macromolecules 2002, 35, 8413.
(17) (a) Kanbara, T.; Yamamoto, T. Macromolecules 1993, 26,
3464. (b) Yamamoto, T.; Sugiyama, K.; Kushida, T.; Inoue,
T.; Kanbara, T. J . Am. Chem. Soc. 1996, 118, 3930.
(18) J andke, M.; Strohriegl, P.; Berleb, S.; Werner, E.; Bru¨tting,
W. Macromolecules 1998, 31, 6434.
(19) Schmitz, C.; Po¨sch, P.; Thelakkat, M.; Schmidt, H.-W.;
Montali, A.; Feldman, K.; Smith, P.; Weder, C. Adv. Funct.
Mater. 2001, 11, 41.
(20) (a) Fukuda, T.; Kanbara, T.; Yamamoto, T.; Ishikawa, K.;
Takezoe, H.; Fukuda, A. Appl. Phys. Lett. 1996, 68, 2346. (b)
O’Brien, D.; Weaver, M. S.; Lidzey, D. G.; Bradley, D. D. C.
Appl. Phys. Lett. 1996, 69, 881. (c) Cui, Y.; Zhang, X.;
J enekhe, S. A. Macromolecules 1999, 32, 3824.
(21) Heinrich, G.; Schoof, S.; Gusten, H. J . Photochem. 1974, 3,
315.
Ack n ow led gm en t. Work at the University of Wash-
ington was supported by the Army Research Office
TOPS MURI (Grant DAAD19-01-1-0676).
(22) (a) Kulkarni, A. P.; J enekhe, S. A. Macromolecules 2003, 36,
5285. (b) Kulkarni, A. P.; Kong, X.; J enekhe, S. A. J . Phys.
Chem. B. 2004, 108, 8689.
(23) Goldfinger, M. B.; Crawford, K. B.; Swager, T. M. J . Am.
Chem. Soc. 1997, 119, 4578.
(24) Ochoa, C.; Rodriguez, J . J . Heterocycl. Chem. 1997, 34, 1053.
(25) Kubel, C.; Chen, S. L.; Mullen, K. Macromolecules 1998, 31,
6014.
Refer en ces a n d Notes
(1) (a) Kraft, A.; Grimsdale, A. C.; Holmes, A. B. Angew. Chem.,
Int. Ed. 1998, 37, 402. (b) Friend, R. H.; Gymer, R. W.;
Holmes, A. B.; Burroughes, J . H.; Marks, R. N.; Taliani, C.;
Bradley, D. D. C.; Dos Santos, D. A.; Bre´das, J . L.; Lo¨gdlund,
M.; Salaneck, W. R. Nature (London) 1999, 397, 121. (c) Rees,
I. D.; Robinson, K. L.; Holmes, A. B.; Towns, C. R.; O’Dell, R.
MRS Bull. 2002, 27, 451.
(2) (a) Rothberg, L. J .; Lovinger, A. J . J . Mater. Res. 1996, 11,
3174. (b) Heeger, A. J . Solid State Commun. 1998, 107, 673.
(c) Sheats, J . R.; Antoniadis, H.; Hueschen, M.; Leonard, W.;
Miller, J .; Moon, R.; Roitman, D.; Stocking, A. Science 1996,
273, 884.
(3) (a) Tarkka, R. M.; Zhang, X.; J enekhe, S. A. J . Am. Chem.
Soc. 1996, 118, 9438. (b) Zhang, X.; Shetty, A. S.; J enekhe,
S. A. Macromolecules 1999, 32, 7422.
(4) (a) J enekhe, S. A.; Zhang, X.; Chen, X. L.; Choong, V.-E.; Gao,
Y.; Hsieh, B. R. Chem. Mater. 1997, 9, 409. (b) Zhang, X.;
J enekhe, S. A. Macromolecules 2000, 33, 2069. (c) Tonzola,
C. J .; Alam, M. M.; J enekhe, S. A. Adv. Mater. 2002, 14, 1086.
(5) (a) Gustafsson, G.; Cao, Y.; Treacy, G. M.; Klavetter, F.;
Colaneri, N.; Heeger, A. J . Nature (London) 1992, 357, 477.
(b) Greenham, N. C.; Moratti, S. C.; Bradley, D. D. C.; Friend,
R. H.; Holmes, A. B. Nature (London) 1993, 365, 628.
(6) (a) Zhu, Y.; Alam, M. M.; J enekhe, S. A. Macromolecules 2002,
35, 9844. (b) Zhu, Y.; Alam, M. M.; J enekhe, S. A. Macro-
molecules 2003, 36, 8958.
(26) McKillop, A.; Swann, B. P.; Ford, M. E.; Taylor, E. C. J . Am.
Chem. Soc. 1973, 95, 3641.
(27) Tao, W.; Nesbitt, S.; Heck, R. F. J . Org. Chem. 1990, 55, 63.
(28) Srinivasan, N. S.; Lee, D. G. J . Org. Chem. 1979, 44, 1574.
(29) (a) Bao, Z.; Chen, Y.; Cai, R.; Yu, L. Macromolecules 1993,
26, 5281. (b) Vaidyanathan, S.; Dong, H.; Galvin, M. E. Synth.
Met. 2004, 142, 1. (c) Brandon, K. L.; Bentley, P. G.; Bradley,
D. D. C.; Dunmur, D. A. Synth. Met. 1997, 91, 305.
(30) (a) Samuel, I. D. W.; Rumbles, G.; Collison, C. J .; Friend, R.
H.; Moratti, S. C.; Holmes, A. B. Synth. Met. 1997, 84, 497.
(b) Smilowitz, L.; Hays, A.; Heeger, A. J .; Wang, G.; Bowers,
J . E. J . Chem. Phys. 1993, 98, 6504.
(31) (a) J enekhe, S. A.; Osaheni, J . A. Science 1994, 265, 765. (b)
Osaheni, J . A.; J enekhe, S. A. Macromolecules 1994, 27, 739.
(32) J enekhe, S. A.; Lu, L.; Alam, M. M. Macromolecules 2001,
34, 7315.
(33) (a) Bre´das, J . L.; Silbey, R.; Boudreaux, D. S.; Chance, R. R.
J . Am. Chem. Soc. 1983, 105, 6555. (b) Agrawal, A. K.;
J enekhe, S. A. Chem. Mater. 1996, 8, 579.
(34) Campbell, A. J .; Bradley, D. D. C.; Lidzey, D. G. J . Appl. Phys.
1997, 82, 6326.
(7) (a) Tonzola, C. J .; Alam, M. M.; Kaminsky, W.; J enekhe, S.
A. J . Am. Chem. Soc. 2003, 125, 13548. (b) Alam, M. M.;
J enekhe, S. A. Chem. Mater. 2002, 14, 4775.
(8) (a) Fu, D.-K.; Xu, B.; Swager T. M. Tetrahedron 1997, 53,
15487. (b) Yang, W.; Huang, J .; Liu, C.; Niu, Y.; Hou, Q.;
Yang, R.; Cao, Y. Polymer 2004, 45, 865.
MA048903Q